ABSTRACT Climate change is more and more considered to be a major global environmental risk. To s t i m u l a t e the participation of Dutch scientists in the i n t e r n a t i o n a l r e s e a r c h effort a r e s e a r c h p r o g r a m m e was established jointly by Ministries involved in Dutch policy actions. The aim of this scientific long term, policyoriented p r o g r a m m e (NRP) was to support Dutch and i n t e r n a t i o n a l climate change policy. To conclude the first phase of the NRP an international conference was held in Maastricht (The Netherlands) from 6 through 9 december 1994. The proceedings of this conference cover a wide range of subjects including: * key note papers of internationally leading scientists on relevant aspects of the climate problem; * assessments of NRP research on the climate system, the causes of potential change in the system, the possible effects and consequences of climate change, and possible alternative policy actions (including technological and/or social); * short papers of the NRP projects and other ongoing research projects, with final conclusions per project.
vii PREFACE
The Proceedings of the International Conference on Climate Change Research: Evaluation and Policy Implications give an excellent impression of both the state of the art of climate change research in general as well as the research projects carried out during the first phase of our National Research Programme on Global Air Pollution and Climate Change (NRP). A large number of experts gathered in Maastricht, one of the most hospitable cities in our country. During discussions with participants it became clear t h a t the quality of the research presented and the organization of the conference itself were considered to be well above average. For this excellent achievement I would like to commend the Programme Committee, the Organising Committee and above all the Conference Secretariat for a job well done! The results of the conference laid down in these proceedings, the result of the international expert review brings us to our conclusion: we have to proceed on the road chosen. Also in the second phase of the continued programme (till 2001) we will put a bit of emphasis on carefull programming and accurate evaluation and presentation of the research projects. In particular the incorporation of the projects and their results within international joint efforts will be promoted. Again my appreciation for these proceedings that will attract the attention of a large research (and policy) community, also due to the timely production and distribution by Elsevier Science B.V.
T. Schneider Programme Director of the Dutch National Research Programme on Global Air Pollution and Climate Change
Editorial The D u t c h N a t i o n a l R e s e a r c h P r o g r a m m e on Global Air P o l l u t i o n and Climate Change The International Conference on Climate Change Research: Evaluation and Policy Implications, held from 6 t h r o u g h 9 December 1994 in M a a s t r i c h t , The N e t h e r l a n d s , concluded the first phase of the Dutch N a t i o n a l R e s e a r c h Programme on Global Air Pollution and Climate Change (NRP). The second phase of this programme started in 1995 and will last to 2001. The conference covered a wide range of subjects, including the climate system, the causes of potential change in the system, the possible effects and consequences of climate change and possible a l t e r n a t i v e responses ' w i t h i n the context of s u s t a i n a b l e development. About 350 scientists, r e s e a r c h m a n a g e r s a n d policymakers from in and outside the Netherlands participated in this succesfull conference. These proceedings contain the texts of the opening statement made by the Dutch Minister of Housing, Spatial Planning and the Environment, invited papers of internationally recognised experts which give a state of the art assessment for several areas of climate change research (part one), assessment reports of the various parts of Dutch climate change research and short papers about Dutch and foreign ongoing research projects (part two). History Climate change is regarded as a serious global problem. Over the years the insight has grown that climate change may pose a serious threat to the world (and also to the Netherlands). To know more both about the nature and seriousness of the problem as well as about the possibilities for countering its effects an intensive international research effort is needed. The NRP was established in 1990 with the aim of providing a scientific basis for the development of climate change policies, and to increase the involvement of the Dutch research community both nationally and internationally in this field. The design of the programme around specific policy-relevant goals distinguishes it from many traditional approaches. Central questions for climate change policy and research The threat of climate change has posed governments a new problem and, in terms of its size and nature, one which is very difficult to manage. Central questions for policy making are: what is going on; what is at stake; what can be done about it, how and with what consequences; what should be the timeframe and what actions should be t a k e n by whom? Such questions have to be translated into research terms in order to arrive at a research programme. A first question with respect to the global climate problem is: How does the climate system work? W h a t processes are going on? Which and w h a t kind of climate
fluctuations can be expected? When are such deviations abnormal and do they have an anthropogenic origin? How predictable is all this? A further question is how seriously the global carbon cycle is disturbed? What is the role of human activity in this? What is the influence on the global system of the greenhouse gases that are emitted? Another question is related to the impacts and consequences of climate change on nature and society. Are the risks associated with climate change larger than those related to other changes in society, especially in the developing countries? How does all this fit in with the goal of a world-wide sustainable development? And finally there is also the important question related to the way in which society deals with the climate change problem. Is it possible to mitigate emissions by the introduction of new technology? Or will it also demand the adaptation of societal structures and institutions, together with changes of lifestyle.
Organisation of the Dutch programme The nature and scale of the climate problem demands a far broader approach than hitherto has been used. The programme comprised 150 research projects, the contents of which were distributed over five themes. The research effort represents approximately 700 man-years, of which 60% is contributed by the research institutions while 40% is funded through NRP. Over 30 research institutes and universities partcipated in the programme. The NRP is presenting itself in the international arena with "Change", a research and policy newsletter on global change. Because of its policy-oriented mission, the NRP has a broad framework. The programme embodies fundamental scientific research to study the m a n n e r in which the climate system works and the physical and chemical processes t hat may produce climate change. Research on causes (emissions) contributes also to the knowledge of the climate system. The programme includes research towards the potential impacts of climate change and possible responses (technical, economic and behavioural/social options). The assessment component of the programme involves the synthesis, integration and communication of research results which provides the basis for decision making and policy actions. In practice NRP is structured according to five themes, which also constitute the structure of these proceedings: 9 the climate system: functioning, modelling and monitoring 9 greenhouse gases: underlying causes of changes in the climate system 9 impacts and consequences of climate change 9 sustainable solutions 9 integration of climate change research. Besides these proceedings, final project leaders and the final report the end of 1995. The report of the 1995) which was held during and Programme Office.
reports of all projects are available from the of the first phase as a whole will be published by international review of the programme (S&PA, just after the conference, is available from the
Acknowledgements We would like to thank Marianne Vonk who did an excellent job in organising the Maastricht conference. She also took care of the preparations for this proceedings.
xi Ottelien van Steenis and Mini Schneider's assistance was a welcome contribution both before and during the conference. We also like to recognise the work of the chairman of the conference, dr. B. Metz, the chairs of the different sessions and the rapporteurs. Last but not least we are greatful for the work of the Programme Committee and the assistance of Sue Postle H a m m o n d and Chris Bernabo of Science and Policy Associates.
The Editors Bilthoven, August 1995 NRP Programme Office P.O. Box 1 NL-3720 BA Bilthoven The Netherlands e-mail:
[email protected]
OPENING ADDRESS M A R G A R E T H A DE BOER M I N I S T E R OF HOUSING, SPATIAL P L A N N I N G AND THE E N V I R O N M E N T Ladies and Gentleman, It was w i t h g r e a t p l e a s u r e t h a t I accepted the invitation to address you at the s t a r t of this i n t e r n a t i o n a l conference. You are here to evaluate the results of the first N e t h e r l a n d s ' R e s e a r c h P r o g r a m m e on Global Air Pollution a n d C l i m a t e Change, N R P I. And you will also be discussing strategies for the second N R P I have come here today as a politician, to tell you w h a t the Government is expecting from the scientific c o m m u n i t y over the next few years. M a n y of the world's climate scientists are contributing to the invaluable work of the I n t e r g o v e r n m e n t a l P a n e l on Climate Change. The IPCC has concluded t h a t the serious risk of human-induced climate change justifies immediate action. Going on t h e s t r e n g t h of the p r e c a u t i o n a r y principle, the i n t e r n a t i o n a l c o m m u n i t y , therefore, took steps to minimize these risks by establishing the U n i t e d Nations' F r a m e w o r k Convention on Climate Change. The developed countries c o m m i t t e d t h e m s e l v e s to curbing their emissions of carbon dioxide to 1990 levels by the y e a r 2000 as a first step. The N e t h e r l a n d s h a d a l r e a d y i n t r o d u c e d a climate change policy before t h e Convention was finalised. It aims at a 3% reduction for Carbon Dioxide (CO2) by the y e a r 2000 relative to the 1990 level, a 10% reduction of m e t h a n e emissions and a stabilization of Nitrous Oxide (N20) emissions. This policy has always rested on a broad consensus in P a r l i a m e n t and society. M e a s u r e s to support the 2000 t a r g e t are also beneficial from other points of view. In the N e t h e r l a n d s , we w a n t to do as m u c h as we can to link the i n t e r e s t s of economy and the environment. With the m e a s u r e s t a k e n so far, we have certainly m a d e a good start. Nevertheless, our most recent forecasts indicate t h a t we will need an incentive tax on energy if we are to meet our commitment. Of course, I am still hoping for a common energy tax w i t h i n the E u r o p e a n Union. B u t the new cabinet has committed itself to the introduction of such a tax in the N e t h e r l a n d s in 1996 if the preferred option fails. An i m p o r t a n t e l e m e n t in our climate policy has been the N a t i o n a l R e s e a r c h P r o g r a m m e , the reason for this conference. M a n y of you have participated in this p r o g r a m m e which was set up to increase our knowledge on the climate system, the causes and effects of rapid climate change, and s u s t a i n a b l e solutions. I t s unique integration of strategic and applied research has received international acclaim. It was m e a n t to encourage our scientists to p a r t i c i p a t e in i n t e r n a t i o n a l r e s e a r c h programmes.
These p r o g r a m m e s have greatly enhanced our u n d e r s t a n d i n g of the climate system. And the results so far have not changed the basic conclusions of the 1990 IPCC report. Of course, there are still many uncertainties left, but we must realize t h a t every answer may raise new questions. We have to learn to cope with this. Climate change is a difficult problem to handle for everyone: for scientists, for politicians, for the media, and for the general public. The atmosphere is a very complex system, and the links between h u m a n activities and the climate are not always clear. There are m a n y things we do not yet know precisely. Furthermore, there are long time lags between causes and effects. The picture is very confusing, especially for the ordinary man in the street. Let me say a little bit more about this. Until now, the emphasis on uncertainties has dominated the public discussion about the greenhouse effect. And as I said, this leaves m a n y people, in the N e t h e r l a n d s too, confused about the seriousness of the matter. Some react by playing down the probability and consequences of climate change or by denying the problem outright. Others argue t h a t we can afford to wait and see, and rely on adopting measures should threat of flooding become a reality. The u n c e r t a i n t i e s could contribute to a wait-and-see attitude. I t h i n k such attitudes are fundamentally flawed. We m u s t realize that by the time scientific uncertainties have been resolved it will be too late. The consequences of rapid climate change could become very costly and serious -maybe even irreversible- as time goes on. In a country like the Netherlands we may see damage to ecosystems and a decreasing supply of fresh water. Developing countries may even have to endure worse problems. We cannot possibly afford to be indifferent about these issues. In general, however, I t h i n k t h a t the public expressions of doubt t h a t I j u s t mentioned should be a signal for us to use a different type of communication. I n s t e a d of e m p h a s i s i n g u n c e r t a i n t i e s , we need to express our scientific understanding in clear terms of risk. That is the type of information policymakers, but also decision-makers in the business community and ordinary citizens are accustomed to handling. Most of our economic decisions are taken in the light of various risks. So that is the type of information politicians and the public need to receive from the scientific world! I t h i n k t h a t scientists involved in the National Research Programme should take the lead in changing tracks to a risk-based discussion of the n a t u r e and the consequences of climate change. This implies translating the results of effects studies into operational terms, attuned to the most vulnerable areas of the world. There is another fact that is becoming increasingly important for the progress of international climate policy. In international discussions there are many questions t h a t need to be answered on the basis of the latest scientific findings. Research efforts should therefore be closely attuned to the questions t h a t arise in the international policy arena. So, let me now give you an impression of what is going on in the run-up to the first meeting of the Conference of the Parties to the Climate Convention, which is to take place in March next year in Berlin.
Emission stabilization by the developed countries is only the first step towards achieving the ultimate objective stabilization. And it is a very small step if we look at the m e a s u r e s t h a t will ultimately be necessary! At their first meeting, the Parties will have to decide what the next step should be. IPCC has been requested to bring t o g e t h e r and assess the scientific information t h a t would help in determining the need for future commitments. Preliminary results show t h a t the development of global emissions over the next decade or so is crucial to w h a t we can ultimately achieve. If we take no further action we will probably not attain the u l t i m a t e objective of the Convention. Consequently, Prof. Bert Bolin, rightly advised INC delegates to address the question of commitments for the first decades of the next century as soon as possible. Now, how are we going to set out our course. We do not yet know exactly w h a t a dangerous level of greenhouse gases in the atmosphere means. What we can say right now, is that it would be foolish to cut off certain options at an early stage. We m u s t keep in mind t h a t the risks are two-way. There is not only the risk of inaction. There is also the risk of being more aggressive than necessary in dealing with the problem. The challenge is to strike the right balance by designing a kind of step-by-step risk optimization process. We need to adjust our course at regular intervals on the basis of the best available information at the time. We must try to m a n a g e the risk of climate change -even if the problem t u r n s out to be more serious t h a n we think- nor should we disregard the risk of major changes in our economies. Let me illustrate the dilemma as follows. When you are driving a car in the mist you cannot look far ahead. So what do you do? Of course, you drive slower and more carefully so t h a t you can anticipate in time. You also look attentively to any signs by the roadside to see what lies ahead. One example of applying this approach would be to assess potential investment decisions in certain areas with respect to their effect on the overall energy intensity. In areas such as the transportation infrastructure, buildings and energy supply systems we should avoid development in the direction of more energy and Carbon Dioxide (CO2) intensive patterns. Everything should be done to think ahead in this areas and launch innovative approaches in them. During the first meeting of the Conference of Parties, therefore, I will actively support initiatives aimed at reaching international agreement ion specific areas, such as energy conservation and economic instruments. For m a n y developed countries, including the Netherlands, it is probably already going to be difficult to achieve the Carbon Dioxide (CO2) target for the year 2000. Nevertheless any credible strategy m u s t soon include a scenario beyond 2000. Investors with a long-term horizon need guidance with respect to long-term developments in climate policy. I would urge participants in the NRP to provide policy-makers with the basic information they need to design the policy options for the period beyond 2000. Finally, I m u s t conclude t h a t ongoing scientific research into all the relevant aspects of the climate change problem is needed to lay a solid basis for climate
policy. The second National Research Programme should thus continue to investigate the workings of the climate system. It should also provide ammunition for developing strategies and measures for adaptation and mitigation for the period beyond the year 2000. Finally, I would strongly recommend the risk-based approach as a new dimension in effect research and in the communication of results. Both politicians and citizens urgently need information of this kind to clarify their thinking and bring the climate change problem closer to solutions. Ladies and Gentleman, I hope that you will be able to shed light on the issues I have mentioned during the days ahead. I wish you a very fruitful and inspiring conference. Thank you very much!
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
Current Progress in the Study of Global Biogeochemical Cycles Michael H. Unsworth and Gordon Wolfe College of Oceanic and Atmospheric Sciences Oregon State University, Corvallis, Oregon, 97331-6511, USA
Abstract
Aspects of global cycles of carbon, nitrogen and sulfur are reviewed. New work defining carbon source and sink strengths in oceans, northern forests and tundra, and wetlands is discussed. Effects of CO2 and N fertilization on the carbon cycle may be large but are currently ill-defined. Recent changes in rates of increase of C02, CH4 and CO in the atmosphere are probably related to volcanic eruptions. The global nitrogen cycle is grossly disturbed by human activity. Land use change and fertilizer use in the tropics may be major sources of N20. Ammonia merits further study as a regional pollutant and because of its role in tropospheric aerosol formation. Sulfate aerosols are now recognized as having significant negative forcing on climate. Whilst the direct (radiative scattering) effect of aerosols is well-understood, the indirect effect (altering cloud properties) is very uncertain, and inclusion of aerosol effects in climate models is limited by lack of data on aerosol regional distributions. In the future, industrial growth in developing countries will alter the amounts and global distribution of greenhouse gases and aerosols, and the change and distribution of aerosols will have particularly important implications for future regional climate change.
1. INTRODUCTION
Studies of biogeochemical cycles play an intrinsic part in research programs into climate change. In particular, cycles which involve radiatively active gases and particles have received a great deal of attention in recent years, partly to establish the strengths of sources and sinks and how these strengths are altered by human activity, and partly to investigate the processes by which the cycles are driven. Increasingly, it is being recognized that biogeochemical cycles of elements cannot be treated as entirely independent; improved knowledge of the chemistry which takes place in the atmosphere, soils and the oceans is revealing many ways in which cycles are intricately linked.
l0
In the space available in this paper, we have chosen to concentrate on some of the current issues associated with the biogeochemical cycles of three elements: carbon; nitrogen; and sulfur. Inevitably this assessment is biased by our personal interests, but we have attempted to include examples of some of the exciting recent developments in the atmospheric, oceanic, and terrestrial sciences. 2. THE CARBON CYCLE
Three issues have dominated research into the carbon cycle in recent years. First, balancing the C02 budget - establishing the strengths of sources of atmospheric C02 arising from human activity and natural systems, and of sinks in the ocean and on the land. Second, understanding feedbacks by which the sources and sinks of a number of carbon trace gases interact with climate change and with increasing C02 concentration. Third, most recently, investigating possible causes of sudden changes in the rates of atmospheric accumulation of a number of trace gases that have been observed in the early 1990s. 2.1
Balancing the CO 2 budget
Table 1 shows the global C02 budget for the decade of the 1980s as proposed by the IPCC (1994). Emissions are relatively well-established, and occur mainly (90%) in the northern hemisphere from fossil fuel combustion and cement production. Emissions from tropical land use change remain poorly quantified, and improved data from southern Asia, Africa, and tropical South America, preferably collected with a common methodology such as high-resolution satellite imagery, are urgently needed. Until recently it was believed that tundra ecosystems were globally a net accumulator of carbon at a rate of about 0.1 to 0.3 Gt yr 1, but recent work by Oechel et al (1993) suggests that warming in the Arctic may have changed these regions to sources with a global strength of about 0.2 Gt yr -1. The accumulation of carbon dioxide in the atmosphere is very well defined by a global network of monitoring stations. Analysis of the stable isotope 13C02 shows clearly that the seasonal amplitude in concentration which varies around the globe is dominated by the activity of the terrestrial biosphere in the northern hemisphere, rather than by seasonal changes in fossil fuel emissions or in ocean sink strength. Recent work by Farquhar et al. (1993) on the isotope composition of oxygen in atmospheric C02 leads to the possibility of distinguishing influences of different terrestrial biomes in the global C02 monitoring network, and when this approach is combined with atmospheric mixing models, it may be possible to resolve some of the present conflict about the relative role of the oceans and the terrestrial biosphere in the net uptake of C02.
1!
Table 1
Annual average budget for anthropogenic carbon for 1980-1989 in GtC/yr
Sources Fossil fuel Changes in tropical land use Total emissions Partitioning to Reservoirs Storage in the atmosphere Ocean uptake Northern Hemisphere forest regrowth Other terrestrial sinks (CO2 and N fertilization, climate effects)
GtC/yr 5.5 + 0.5 1.6_+ 1.0 7.1 _+ 1.1 3.2 2.0 0.5 1.4
+ 0.2 + 0.8 + 0.5 _+ 1.5
Data from IPCC 1994
Estimates of carbon uptake by the oceans have been made by t w o independent approaches. Methods using radiocarbon, produced naturally in the upper atmosphere or artificially during nuclear weapons testing, as a tracer give larger estimates of uptake by oceans than methods based on air-sea exchange. A recent analysis (Sesshaimer et al. 1994) suggests that estimates by the radiocarbon tracer method may need to be reduced by about 25 %. The air-sea exchange method, which is based on differences in the partial pressure of CO2 between the water and the air, and on an exchange function described in terms of wind speed and temperature, has also been reassessed (Robertson and Watson 1992) These authors pointed out that the upper 1 mm or so of the oceans is generally cooler than the bulk mixed layer by about 0.3 o C, and when this thermal skin effect is taken into account, the air-sea exchange method results in CO2 uptake estimates that must be increased by about 0.7 Gt C yr 1 . Thus recent work has brought the two approaches into much closer agreement in defining ocean uptake. There has been debate about whether increases in ocean phytoplankton productivity may have increased the ocean sink strength for CO2 over the last 100 years or more. Analysis by Falkowski and Wilson (1992) of changes in phyto- plankton biomass in the north Pacific Ocean over the last 70 years shows that changes are too small to have a significant effect on the sink strengths for atmospheric CO2. Although there are very few historical data for other main ocean basins, it seems likely that this conclusion applies on a global scale. In contrast to the open ocean, coastal zones have undoubtedly experienced large increases in nutrients associated with human population changes, but Falkowski and Wilson concluded that there is no conclusive evidence yet to suggest that these coastal zones represent a significant new sink.
12 As an example of interacting biogeochemical cycles between the atmosphere and ocean, it has been proposed that, because iron concentrations limit phytoplankton productivity in some parts of the ocean, deposition of iron from either human activities or volcanic eruptions might increase the ocean sink strengths for CO2. An experiment has recently been conducted in the equatorial Pacific Ocean to test this hypothesis, and Watson et al. (1994) reported some early results. When a small patch (8 x 8 km) was enriched with iron, there was a significant depression in the surface concentration of CO2 within 48 hours of the iron release, but the effect was only a small fraction (about 10%) of the CO2 drawdown that would have occurred if the enrichment had resulted in the complete utilization of all other available nutrients. Reasons why the fertilization procedure was much less effective in the open ocean than in the laboratory are not yet clear, but at present the results do not support the idea that iron fertilization significantly affects the oceanic CO2 sink strength. The weight of evidence at present suggests that in order to balance the sources of CO2 in Table 1 against sinks and atmospheric storage, there must be an illdefined terrestrial sink in northern temperate latitudes. This conclusion is reached by inversion of the observed atmospheric CO2 distribution combined with atmospheric tracer models, making constraining assumptions about ocean and terrestrial sinks (Tans et al. 1990, Enting and Mansbridge 1991). The likeliest terrestrial processes contributing to this sink are: changing forest management; and enhancement of productivity due to atmospheric CO2 increases and/or nitrogen fertilization from atmospheric deposition. A number of recent carbon inventories of temperate forest systems have concluded that such systems may have been a sink for about 0.5 GtC/yr in the last 20 years or so, partly because of natural regrowth and replanting after forest harvesting, and partly through fire suppression (IPCC 1994, Auclair, personal communication). Although there is much evidence from laboratory and controlled environments to show that plant productivity can be increased by 30% to 40% when CO2 is doubled, there is no conclusive evidence from the field to show long-lasting increases in northern temperate ecosystem productivity in response to increased CO2. There are sound biochemical reasons indicating that the interaction of CO2 and temperature is such that the benefits of CO2 fertilization cannot be achieved at low temperatures (Long, 1991 ). This may explain partly why, when Oechel et al. (1994) exposed natural arctic tundra to doubled CO 2 concentration, there was no long-term boost in carbon sequestration. An alternative explanation is that, after an initial burst of productivity, plants exhaust the supply of soil nutrients, and this limits future productivity. There have been insufficient long-term experiments with perennial systems such as forests and tundra to determine whether the equilibrium longterm response involves eventual changes in soil nutrient availability that would allow productivity to be enhanced. IPCC (1994) estimated that CO2 fertilization may have accounted for a sink of 0.5 to 2.0 GtC/yr during the 1980s, but such estimates must be regarded as extremely tentative.
13 In and around industrialised regions, ecosystems may receive substantial inputs of nitrogen, arising from fossil fuel burning and agriculture, and this input can act as a fertilizer. IPCC (1994) speculated that this fertilizer effect could have increased terrestrial carbon storage by 0.2-1.0 GtC/yr in the 1980s. One of the most important current programs aimed at understanding the interaction between boreal forests and the atmosphere is the international multi-agency BOREAS Program taking place in northern Canada (Sellers, et al. 1995). Analysis of the many detailed records of CO2 exchange collected during BOREAS should help in quantifying the scale of net carbon input to northern forest systems. 2.2 Feedback processes
We have already mentioned a number of studies which indicate interactions between changing climate and atmospheric CO2 concentrations and the net exchange of CO2 between the surface and the atmosphere. There has also been interest in the sensitivity of methane fluxes to climate change and CO2 concentration. Whiting and Chanton (1993) found that, for wetlands of varying productivity around the world, higher net primary production was associated with higher emissions of methane. It has therefore been suggested that, if CO2 increased the productivity of wetland vegetation, some of the benefits of carbon sequestration would be lost because of increased methane emissions. Dacey et al. (1994) recently presented results supporting this view. They studied methane emissions from a marsh that had been exposed in open-top chambers to twice ambient atmospheric CO2 for the previous 7 years, and found that methane emission over a 10-day period from the CO2-enriched sites was nearly 80% higher than in control sites. If confirmed in longer-term work, the implications of this observation are important, not only for methane fluxes from natural ecosystems but also for fluxes from wetland rice production where much effort is put into increasing productivity. Most soils that are not flooded consume methane, but the extent varies with soil water content and land use. A number of recent reports have shown that inputs of nitrogen in the form of ammonium to soils strongly inhibit soil methane consumption (King and Schnell 1994). Ammonium concentrations in many soils have increased in recent years as a result of land use changes and increases in the ammonium concentration of precipitation. Similar responses are not observed in soils treated with nitrate-based fertilizer or farmyard manure (Goulding et al., in press). Goulding et al also analyzed soils from long-term experiments at Rothamsted, England, and showed that extended (150 years) cultivation of land for arable crops reduced methane uptake rates by 85% compared to those in soil under calcareous woodland. King and Schnell argued that past increases in atmospheric methane concentration may have increased the inhibitory effect of ammonium on soil methane uptake, and this mechanism would provide a positive feedback on future atmospheric methane concentrations.
14 2.3
Recent changes in atmospheric accumulation of trace gases
One of the most puzzling and yet instructive aspects in the study of trace gas biogeochemistry occurred in the early 1990s. Until this time, CO2 concentrations around the world had increased rather consistently over the previous 30 years at about 0.5 - 1.5 vpm per year, with the rate tending to increase With time (Fig. 1).
.-..
3.0
2.5 "~ 2.0 o o
c? O
"6
~.5
1.0 0.5
~! ii
ii I !, ii I
o
;5 1960 1965 1970 1975 1980 1985 1990 1995 Year
Figure 1
Growth rate of CO2 at Mauna Loa, Hawaii. The smooth curve is filtered to suppress short-term (-~10 yr) variations. (From IPCC 1994)
After mid-1991, continuing throughout 1992 and 1993, the growth rate of atmospheric carbon dioxide slowed by an unprecedented amount (Keeling, 1993, Sarmiento, 1993). Concentrations of carbon monoxide CO and methane CH4, which had also been increasing steadily up to 1991, grew slowly from 1991 to 1993 (Fig 2). The cause of these large changes was almost certainly the eruption of Mount Pinatubo in the Philipines in June, 1991, but the mechanisms that brought about the changes are a matter for debate. As we discuss later, it seems likely that the carbon dioxide anomaly is associated with the volcanic aerosols that were injected into the stratosphere, reducing solar radiation at the surface and producing cooling on a global scale. Cooling, and
15 perhaps associated changes in rainfall and evaporation, could alter the balance between photosynthesis and respiration on the land, and the sink strengths of the ocean for carbon dioxide.
175O ..-.. > JO
o_ 1725
g "=
1700
co
1675
c 0 o r-
1650
*
1625 I
'83
'84
I
'85
I
'86
!
'87
I
'88
'89
I
'90
I
'91
I
'92
'93
Year
Figure 2
Globally averaged CH 4 concentration showing low growth rates during 1992 and 1993. (From IPCC 1994)
The changes since 1991 in atmospheric methane and carbon monoxide are probably even more complex, because both gases have a major atmospheric sink by reaction with the hydroxyl radical OH. A recent analysis by Bekki et al. (1994) suggested that an unprecedentedly large depletion of stratospheric ozone over this period may have contributed to the sharp decrease in growth rates of both gases. The decrease in stratospheric ozone would allow more ultraviolet radiation to reach the troposphere and this would have resulted in increased concentrations of OH. Bekki et al. concluded that they could account for almost half of the 1992 decrease in growth rates of both gases by this mechanism, but there may also have been changes in source and sink strengths at the surface to account for the remainder. It seems likely that the ozone depletion in the stratosphere was also caused by the Pinatubo eruption, because the fine aerosol of sulfuric acid droplets resulting from the injection of 15 to 20 million tonnes of sulfur dioxide into the stratosphere interacted with other stratospheric chemicals to destroy ozone. The consequences of this natural event provide an excellent test of our understanding of the carbon cycle, and serve as a reminder of the complex interactions that are contained within the carbon cycle.
16 3. THE NITROGEN CYCLE In comparison to the global carbon cycle, the global nitrogen cycle has been much more grossly disturbed by human activities. Table 2 summarizes annual terrestrial fluxes prior to substantial human alteration, and lists current changes arising from human activities.
Table 2 Global fluxes of nitrogen (in Tg/yr) in unperturbed terrestrial ecosystems and as a consequence of human activity
(i)
(ii)
(iii)
Unperturbed systems Biological N fixation Fixation by lightning Denitrification Perturbed systems Biomass burning Tropical land clearance New fixation Manufacture of fertilizer N N fixation by legume crops N fixation by fossil fuel combustion
TaN/vr 100 10 9O 40 22 80 30 25
Data from Vitousek and Matson, 1995
Since the late 1970s, the production of nitrogen fertilizer has probably increased by about one-third (Vitousek and Matson, 1995). Prior to that time, most of this fertilizer was used in developed countries of the temperate zones, but since then, the rate of increase in nitrogen fertilization has been extremely rapid in the tropics, so that by the late 1980s, more than half of the global nitrogen fertilizer use was in the developing world (including China). This trend is likely to continue, with the implication that the distribution of trace gas fluxes to the atmosphere, discussed in the following sections, will change substantially. Changes in the source strengths of t w o trace gases as a result of the disturbed nitrogen cycle have attracted particular attention in recent years. Nitrous oxide emissions are influenced by fertilizer use and land use changes; and ammonia arising from both fertilizer use and animal production has been shown to create important regional problems, especially in Europe.
17 3.1
Nitrous Oxide
The global budget of nitrous oxide has been revised recently by the IPCC (1992), based on new information on soil fluxes from tropical ecosystems and temperate forests, further work on cultivated soils, and new estimates of emissions from biomass burning. Large tropical sources are required to explain the N20 latitudinal gradient revealed by atmospheric monitoring over the last ten years. Keller and Matson (1994) recently reviewed sources of N20 in the tropics and evaluated the effects of land use changes. Wet undisturbed tropical forests appear to account for the largest natural terrestrial source, and this is associated with the large rates of nitrogen transformation in the soil and cycling through vegetation in these systems. Even so, it seems necessary to include additional tropical sources to balance the global budget, and Keller and Matson proposed that tropical land use change and intensification of tropical agriculture may be significant contributors towards this missing source. In particular, the creation of young pastures from previously undisturbed systems seems likely to considerably increase N20 fluxes to the atmosphere, and the increasing use of nitrogen fertilizer in tropical systems appears from the few measurements available to cause larger fluxes of N20 than would be found from similar practice in temperate crop systems. It seems likely that crop and soil management practices can be manipulated to control nitrous oxide flux and there is clearly a need in temperate and tropical areas for further work to explore this possibility. Since nitrogen fertilization increases emission of nitrous oxide to the atmosphere and may decrease absorption of methane by soils (see earlier), the potential of improved soil management for slowing the buildup of radiatively active trace gases in the atmosphere is very important. 3.2
Sources and Sinks of Ammonia
Concern about ammonia fluxes from intensive agriculture in Europe and eastern North America arises principally because dry and wet deposition of reduced nitrogen compounds can make a substantial contribution to the acidification and nitrogen eutrophication of semi-natural systems (Fowler et al. 1989). The importance of ammonia fluxes in global radiative forcing has not been adequately explored yet, but reactions between ammonia and sulfur dioxide to create ammonium sulfate aerosols, discussed in the next section, are an important contributor to the global anthropogenic aerosol burden. One of the most important advances in recent years has been the development of micrometeorological methods for studying ammonia fluxes between vegetation and the atmosphere. These techniques have allowed investigations over grazed pastures, fertilized agricultural crops, and natural ecosystems. Sutton et al. (1995) recently reviewed this work. The measurements have clearly demonstrated that there is an NH 3 'compensation point' associated with plant tissue, so that emission from plants occurs when atmospheric concentrations are below the compensation point, and deposition occurs at
18
higher atmospheric concentrations. Figure 3 summarizes measurements over agricultural crops and semi-natural vegetation (Sutton et al. 1995), showing that at low NH 3 concentrations, agricultural crops are a source of NH 3 to the atmosphere and semi-natural vegetation is a sink. The figure also indicates that the compensation point is different for the two systems, as would be expected as a result of nitrogen fertilization of crops. It seems likely that natural vegetation on which there has been substantial NH3 deposition could also have a higher compensation point, and so the sink strengths of natural vegetation for NH 3 may decrease in polluted regions with time.
30i
8~
20
o '7,
E u.. z v
00
OOC~
9 9
9
agricultural croplands
0
semi-natural vegetation
:
-10 0
0
00~
-20 -30
o
Oo
O 0
-
o~
o
00 0
0
O0
-40 -50 -60 O0 -70 0.0
i
i
i
I
i
I
0.5
1.0
1.5
2.0
2.5
3.0
3.5
NH3 (~g m "~)
Figure 3
Micrometeorological measurements of the variation of NH 3 fluxes (positive = away from the surface) with NH 3 concentration over agricultural crops and semi-natural ecosystems (From Sutton et al., in the press)
The global ammonia budget has been relatively neglected in studies of the nitrogen cycle. Schlesinger and Hartley (1992) concluded that the major uncertainties and emissions were associated with fluxes from undisturbed soils. They also particularly commented that deposition of ammonium in rain in the eastern United States declined over the 20-year period from 1963 to 1982, which they regarded as surprising, given the increasing use of urea fertilizer. They concluded that atmospheric interactions with sulfur dioxide to form sulfate
19 aerosol may account for this observation, but it also seems likely from recent work that ammonia deposition on natural vegetation close to agricultural fields has substantially increased, and the consequences of this nitrogen input deserve more attention. 4. THE SULFUR CYCLE
The global sulfur cycle has been severely perturbed by human industrial activities for many years. Because world industrial activity and fossil-fuel consumption have been concentrated in the northern hemisphere, there are virtually t w o global sulfur cycles: a relatively "natural" cycle in the southern hemisphere, and a vastly perturbed cycle in the northern hemisphere. Some consequences of the perturbed cycle, such as acid rain, have been recognized for decades. But we have only in the past few years begun to appreciate the potential importance for global climate of human impacts on the sulfur cycle. In the following sections we will briefly review recent findings about the climatic roles of both the natural and perturbed sulfur cycles. 4.1
Sulfur, aerosols, and climate
Scattering of solar radiation by atmospheric aerosols effects the atmospheric radiation balance because it reduces the amount of energy from the sun that reaches the surface of the earth. Scattered radiation may be either lost to space or absorbed by aerosols, but both processes result in a net loss of energy to earth's surface and therefore net cooling. Thus, aerosol scattering has a negative forcing influence on climate - the opposite of greenhouse gases such as CO 2. Although the mass of sulfur in the atmosphere is only about 10 .5 that of carbon, it dominates scattering in the atmosphere. This is because sulfur oxidizes and hydrates to form sulfate aerosol particles that are in the size range (0.1 - 1pro) that scatters visible radiation very effectively. Such scattering is roughly 105 times more efficient per atom as a climate forcing mechanism than the radiation absorption of greenhouse gases such as CO2 (Shaw, 1983), but the radiative influence of aerosols on climate is reduced by their much shorter atmospheric residence time, less than a week, compared to an effective residence time of decades for CO2. Sulfate aerosols contribute to radiation scattering in two distinct ways: by scattering directly and by modifying cloud optical properties and thereby influencing radiation scattering by clouds. The latter "indirect" effect has been appreciated as an important feature in global climate for some years, but is an extremely complex process and is still poorly understood.
20 4.1.1
The natural sulfur cycle and climate
We have surprisingly little quantitative understanding of the role of the natural sulfur cycle in climate. This is partly because the subject has only recently attracted much attention, and partly because the natural cycle has been so perturbed by human activity. There are few data on the distribution of sulfate aerosols even today, much less from prior eras, and trying to find the climate signal of natural sulfate aerosols in a vastly perturbed world is extremely difficult. Nonetheless, there are several likely mechanisms by which the natural sulfur cycle plays a role in climate. The marine dimethyl sulfide (DMS) effect - the brightening of marine stratus clouds hypothesized by Charlson et al. (1987) - may have been the dominant process by which the natural sulfur cycle affected climate before the industrial revolution. An intriguing aspect of this hypothesis is the suggestion of a potential feedback loop which, if negative, would act to stabilize climate. However, studies of ice-core concentrations of methane sulfonic acid and non-seasalt sulfate (atmospheric oxidation products of DMS) (Legrand et al., 1988) suggest instead that there was a positive feedback during previous ice ages. To date, evidence supporting the marine DMS hypothesis has been difficult to gather (Ayers and Gras 1991, Ayers et al. 1991, Bates et al. 1987) because of the difficulties associated with measuring simultaneously all the necessary variables (Bates et al., 1990) and the interference from anthropogenic sulfate aerosols (Falkowski et al. 1992, Schwartz 1988). Field studies planned for 1995 - 96 in the southern Pacific ocean are aimed at determining whether the 'cloud brightening' mechanism actually exists, and its potential strength. Volcanic eruptions (e.g. Mt. Pinatubo in June, 1991 in the Philippines) have important impacts on the natural sulfur cycle by injecting huge pulses of sulfur into the stratosphere which oxidize to form sulfate aerosols. Unlike tropospheric aerosols, these stratospheric aerosols remain in the atmosphere for several years due to the lack of aqueous removal mechanisms in the extremely dry stratosphere. They therefore achieve circumglobal distributions and may cause global cooling for a significant period. Such aerosols also deplete stratospheric ozone, as discussed earlier.
4.1.2 Anthropogenic changes to the S cycle Sulfur is an integral element in all biological materials, and all biogenic oil and coal contain approximately 1% - 10% S by mass. Therefore, production of sulfur gases is an inevitable by-product of fossil fuel combustion.
21 Emissions of sulfur to the atmosphere from human activities are now 2 - 3 times natural emissions annually. In the past few decades, the major areas of sulfur flux to the atmosphere have been the eastern United States and western and central Europe, and globally, about 94% of the emissions are in the northern hemisphere (Schwartz 1988). Because aerosols are removed from the atmosphere before they can be transported across the equator, this leads to vastly different distributions of aerosols in the northern and southern hemispheres (Langner et al. 1992). The sulfate haze that blankets much of the northern hemisphere is now recognized to be a direct result of fossil fuel burning rather than from natural sources. Further evidence for the disparate hemispheric cycles comes from 200% - 300% increased sulfate deposition to Arctic ice but not to Antarctic ice in the past hundred years (Mayewski et al. 1990). Because no global or even regional aerosol sampling network exists, and satellite observations are lacking, our best estimates of aerosol distributions come from computer models which begin with known sources and simulate atmospheric transport, chemistry, and physics to predict aerosol distributions (Langner et al. 1992). Although the mechanism of cooling by direct aerosol scattering is relatively simple, its impact was until recently underestimated, largely because it was not realized how much of the sulfate aerosol haze in the northern hemisphere is actually from industrial emissions. Although one recent estimate suggests that perhaps less than 10% of sulfur emissions result in the formation of new aerosol particles (Langner et al. 1992), the rate of new sulfate particle formation may have doubled since pre-industrial times. Recent re-evaluations of the direct climate forcing by radiation scattering from anthropogenic aerosols have suggested cooling of similar magnitude to the C02 warming (Charlson et al. 1992, IPCC 1994), leading to speculation that the "greenhouse signal" predicted in the late 1980's has been partially masked by concomitant sulfate aerosol production. Because the cooling due to aerosol scattering is localized, it is thought to be heavily concentrated around eastern North American and central Europe. Aerosol modification of cloud albedo (indirect climate forcing by aerosol) is a much more difficult problem. The distribution and radiative properties of clouds are probably the major uncertainties in climate prediction models, and there are no models today which treat clouds in a wholly realistic manner. Worse, the increased reflectivity of clouds from sulfur aerosol condensation nuclei is extremely non-linear and poorly understood. Nonetheless, several recent attempts to estimate the impact of human sulfur emissions on cloud properties (Jones etal. 1994, Wigley 1989) suggest cooling which may be similar to that produced from aerosol scattering - that is, roughly comparable to C02 warming.
22 Climate models are only just beginning to include sulfur emissions and direct and indirect effects of aerosols (Jones et al. 1994, Kiehl and Briegleb 1993, Taylor and Penner, 1994, Wigley 1989). Initial results suggest that predicted climate responses when aerosols are included may be quite different than for radiative gases alone. Feedbacks within the climate system may lead to cooling not just in regions of sulfur emission but also in far-removed areas, such as in the sub-Arctic (Taylor and Penner, 1994). One consequence of the northern-hemisphere enhancement of sulfur aerosols is that warming associated with increases in greenhouse gases may occur more quickly in the southern hemisphere, where it is not partially offset by aerosol cooling. 5. CONCLUDING REMARKS
Changes in global economic and social systems are likely to have profound effects on emissions of radiatively active gases and particles in future decades, with implications for global climate change. The breakup of the Soviet Union, explosive growth in the third world, and the economic emergence of Asia, in particular the "industrial revolution" in China, will lead to geographically changing patterns of fuel consumption over the next decade that are unlike anything in the past 25 years. Carbon and sulfur emissions from rapidly industrializing nations are likely to soar. At the same time, emissions in the developed world may decrease, as more stringent controls take effect. In general, we may expect an increase in emissions from low latitudes on both sides of the equator, and a possible stabilization in the higher-latitude emissions from North America and Western Europe. Emission changes in Eastern Europe and the countries of the former Soviet Union are major uncertainties. Geographic patterns of fossil fuel emissions of greenhouse gases are not particularly important, since these species are well mixed around the globe and have lifetimes of years to decades in the atmosphere. Although climate response to greenhouse gas forcing will certainly vary between regions, the distribution of responses is likely to be relatively insensitive to where the gases are emitted. The situation is very different for short-lived sulfur aerosols. Their climatic effects are intrinsically regional, since they do not exist in the atmosphere long enough to be dispersed globally. Consequently, changes in the distribution of sulfur emissions will result in different local climatic impacts. However, it is likely that regional climate response will not be limited only to areas of strong forcing, because of feedbacks in the climate system (Taylor and Penner 1994). We have barely begun to explore the complex interactions between the climatic forcing of industrial carbon, sulfur, and other emissions, as well as our other diverse impacts on global biogeochemical cycles. If world economic changes are more rapid than scientific advances necessary to understand the climatic effects of these coupled emissions, we may be chasing a "moving target" of
23 climate forcing. Climatic change policy decisions based on today's economic and social scenarios may be wrong for tomorrow's world unless we understand the effects on climate of our modifications of the major biogeochemical cycles. 6. REFERENCES
Ayers, G. P. and Gras, J. L. (1991). Seasonal relationship between cloud condensation nuclei and aerosol methanesulphonate in marine air. Nature 353: 834-835. Ayers, G. P., Ivey, J. P. and Gillett, R . W . (1991). Coherence between seasonal cycles of dimethyl sulphide, methanesulphonate and sulphate in marine air. Nature 349: 404-406. Bates, T. S., Charlson, R. J. and Gammon, R. H. (1987). Evidence for the climatic role of marine biogenic sulfur. Nature 329: 319-321. Bates, T. S., Clarke, A. D., Kapustin, V. N., Johnson, J. E. and Charlson, R. J. (1990). Oceanic dimethylsulfide and marine aerosol: difficulties associated with asssessing their covariance. Global Biogeoch. Cycles 3: 299-304. Charlson, R. J., Lovelock, J. E., Andreae, M. O. and Warren, S. G. (1987). Oceanic phytoplankton, atmospheric sulphur, cloud albedo and climate. Nature 326: 655-661. Charlson, R. J., Schwartz, S. E., Hales, J. M., Cess, R. D., Jr., J. A. C., Hansen, J. E. and Hofmann, D. J. (1992). Climate forcing byanthropogenic aerosols. Science 255: 423-430. Dacey, J.W.H., Drake, B.G. and Klug, M.J. (1994) Stimulation of methane emission by carbon dioxide enrichment of marsh vegetation. Nature 370: 47-49. Falkowski, P.G. and Wilson, C. (1992) Phytoplankton productivity in the North Pacific ocean since 1900 and implications for absorption of anthropogenic CO2. Nature 358: 741-743. Falkowski, P. G., Kim, Y., Kolber, Z., Wilson, C., Wirick, C. and Cess, R. (1992). Natural versus anthropogenic factors affecting low-level cloud albedo over the north Atlantic. Science 256:1311-1313. Farquhar, G.D., Lloyd, J., Taylor, J.A., Flanagan, L.B., Syvertsen, J.P., Hubick, K.T., Wong, S.C. and Ehleringer, J.R. (1993) Vegetation effects on the isotope composition of oxygen in atmospheric CO2. Nature 363: 439-443.
24 Goulding, K.W.T., Hutsch, B.W., Webster, C.P., Willison, T.W. and Powlson, D.S. (1995) The effect of agriculture on methane oxidation in soil. Philosophical Transactions, Royal Society of London: in the press. Hesshaimer, V., Heimann, M. and Levin, I. (1994) Radiocarbon evidence for a smaller oceanic carbon dioxide sink than previously believed. Nature 370: 201-203. IPCC (1992) Climate Change 1992. Houghton, J.T., Callander, B.A. and Varney, S.K. (Eds.) Cambridge, England: Cambridge University Press. IPCC (1994) Radiative Forcing of Climate Change. Geneva, Switzerland: Intergovernmental Panel on Climate Change. Jones, A., Roberts, D. L. and Slingo, A. (1994). A climate model study of indirect radiative forcing by anthropogenic sulphate aerosols. Nature 370: 450453. Keller, M. and Matson, P.A. (1994) Biosphere-atmosphere exchange of trace gases in the tropics: evaluating the effects of land use changes. In: Prinn, R.G. (Ed.) Global Atmospheric-Biospheric Chemistry. pp. 103-117. New York: Plenum Press Kiehl, J. T. and Briegleb, B. P. (1993). The relative roles of sulfate aerosols and greenhouse gases in climate forcing. Science 260:311-314. King, G.M. and Schnell, S. (1994) Effect of increasing atmospheric methane concentration on ammonium inhibition of soil methane consumption. Nature 370: 282-284. Langner, J., Rodhe, H., Crutzen, P. J. and Zimmermann, P. (1992). Anthropogenic influence on the distribution of tropospheric sulphate aerosol. Nature 3 5 9 : 7 1 2 - 7 1 5 . Legrand, M. R., Delmas, R. J. and Charlson, R. J. (1988). Climate forcing implications from Vostok ice-core sulphate data. Nature 324: 418-420. Mayewski, P. A., Lyons, W. B., Spencer, M. J., Twickler, M. S., Buck, C. F. and Whitlow, S. (1990). An ice-core record of atmospheric response to anthropogenic sulfate and nitrate. Nature 346: 554-556. Oechel, W.C., Cowles, S., Grulke, N., Hastings, S.J., Lawrence, B., Prudhomme, T., Riechers, G., Strain, B., Tissue, D. and Vourlitis, G. (1994) Transient nature of CO2 fertilization in Arctic tundra. Nature 371: 500-503.
25 Oechel, W.C., Hastings, S.J., Vourlitis, G., Jenkins, M., Riechers, G. and Grulke, N. (1993) Recent change of Arctic tundra ecosystems from a net carbon dioxide sink to a source. Nature 361: 520-523. Robertson, J.E. and Watson, A.J. (1995) Thermal skin effect of the surface ocean and its implications for CO2 uptake. Nature 358: 738-740. Sarmiento, J.L. (1993) Atmospheric CO2 stalled. Nature 365: 697-698. Schlesinger, W.H. and Hartley, A.E. (1992) A global budget for atmospheric NH s. Biogeochemistry 15:191-211. Schwartz, S. E. (1988). Are global cloud albedo and climate controlled by marine phytoplankton? Nature 336: 441-445. Sellers, P., Hall, F., Margolis, H., Kelly, B., Baldocchi, D., den Hartog, J., Cihlar, J., Ryan, M., Goodison, B., Crill, P., Ranson, J. and Lettenmaier, D. (1995) The Boreal Ecosystem-Atmosphere Study (BOREAS): an overview and early results from the 1994 field year. Bulletin, American Meteorological Society: in the press. Shaw, G. E. (1983). Bio-controlled thermostasis involving the sulfur cycle. Climatic Change 5: 297-303. Sutton, M.A., Schjorring, J.K. and Wyers, G.P. (1995) Plant-atmosphere exchange of ammonia. Philosophical transactions, Royal Society of London: in the press. Taylor, K. E. and Penner, J. E. (1994). Response of the climate system to atmospheric aerosols and greenhouse gases. Nature 369: 734-737. Vitousek, P.M. and Matson, P.A. (1995) Agriculture, the Global Nitrogen Cycle, and Trace Gas Flux. In: Proceedings, l Oth International Symposium on Environmental Biogeochemistry, pp. 193-207. Watson, A.J., Law, C.S., Van Scoy, K.A., Millero, F.J., Yao, W., Friederich, G.E., Liddicoat, M.I., Wanninkhof, R.H., Barber, R.T. and Coale, K.H. (1994) Minimal effect of iron fertilization on sea-surface carbon dioxide concentrations. Nature 371 : 143-145. Wigley, T. M. L. (1989). Possible climate change due to SO2-derived cloud condensation nuclei. Nature 339: 365-367.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
27
The potential effects of climate change in a riverine hydrological system in northwestern Canada S. J. Cohen
Environmental Adaptation Research Group, Atmospheric Environment Service, Environment Canada, 4905 Dufferin Street, Downsview, Ontario M3H 5T4, Canada
Abstract Assessments of climate change risks, and system vulnerabilities, may benefit from a focus on a watershed. A riverine hydrological system is an integrator of natural and human systems, and so constitutes an appropriate setting for determining the regional effects of climate change scenarios. The Mackenzie Basin Impact Study (MBIS) is presented as a case study that illustrates the challenges and opportunities presented by integrated regional assessment of climate change scenarios in a large watershed in northwestern Canada.
1. Climate Change Impact Assessment: The Integration Challenge The Framework Convention on Climate Change (FCCC) is now a part of international law, committing more than 80 nations to action. Its ultimate objective is to stabilize global concentrations of carbon dioxide and other 'greenhouse gases' at a level that does not represent 'dangerous anthropogenic interference' to the atmosphere. At issue, however, is the definition of the term 'dangerous.' This is an important challenge for climate impact assessment at the regional scale. Global scale atmospheric anomalies, such as E1 Nifio-Southern Oscillation (ENSO), are known to produce region-specific impacts (e.g. Glantz et al., 1987). The same thing is likely to happen with an enhanced greenhouse effect. What might these impacts be and what might (or should?) be the nature of the adaptive responses? Regional impact assessment is a complex multidisciplinary research challenge. To make matters even more difficult, we are considering an assessment not of an observed climatic event (such as the 1993 Mississippi River flood) but of a theoretical warming of the earth's climate by increased concentrations of greenhouse gases. There are many uncertainties associated with the data and methods used to construct scenarios of a future warmer world, and some have argued for the use of analogues (Glantz, 1988; Kearney, 1994) as an alternative to scenarios based on climate model simulations, population projections, and other forecasting tools. There is little doubt, however, that if climate warming occurs, the earth and its people will feel its effects through a variety of "pathways" and "filters," and the impact assessment needs to account for these.
28
1.1 What is integration.9 In order to capture the complex linkages between climate and regions, a research framework is needed which effectively combines information about individual sectors so that the result is more than the sum of the parts. The Intergovernmental Panel on Climate Change (IPCC) defines integrated assessment as "the most comprehensive treatment of the interactions of climate and society" (Carter et al., 1992). It addresses the "net" effect of climate-related stress, so that the indirect linkages between atmosphere, land and water resources, resource management and other policy matters, can be considered in a way that can be understood by decision makers. This would prevent the implementation of strategies or policies which assist one group or sector at the expense or detriment of others. Working at the regional scale is important because at larger scales, impacts may offset each other, and the final result may hide critical details (e.g. Rosenzweig and Parry, 1993). There are two main approaches to integrated assessment: a) models, including integrated system models such as IMAGE (Alcamo, 1994), and a new model being developed by Battelle/Pacific Northwest Laboratories in the United States (Frederick and Rosenberg, 1994), and b) assessments of policy instruments (e.g. development plans, conservation plans (e.g. Inuvik, 1993)), or regulatory bodies (e.g. river authorities (e.g. Arnell et al., 1994)). There are a wide range of options available within these two sets. Rather than relying on only one approach, a regional study could make use of several integrating techniques if they provide unique and complementary assessments. This could be called the "family of integrators" approach. With the growing interest in global-scale assessment models, and their potential application in policy gaming exercises, there remains a need for detailed information on smaller scales, which could provide the foundation for global models to produce regional simulations that are plausible to stakeholders. Integrated regional assessments could provide this information. At the same time, however, there is a requirement to convince policy makers that these decision support tools are useful for assessing response options. Policy makers do not represent a global constituency, so there is a need to address issues at their regional/national scales of interest. It is suggested that an integrated assessment should not rely exclusively on integrated system models, since most of these do not necessarily involve the stakeholder, nor make direct use of the stakeholder's perception of the climate change issue. This perception is not based solely on whether a climate change has been noticed, but on whether any observed or simulated changes in the landscape or economic production can be linked to observations or simulations (scenarios) of climate change. Stakeholders can be an important source of "ground truth," and that is the frame of reference they would use when considering responses to future scenarios of climate change (e.g. Aharonian, 1994; Bielawski, 1994). Henderson-Sellers (1993) warns that integrated impact assessments might still be circumvented in the rush towards responding to the climate change threat, and that uncertainties make such assessments premature. If full integration is impossible to achieve due to insufficient information, what about partial integration, in which there are some aspects that remain outside of the assessment, and assumptions have to be made about their level of influence. For example, the study of the Corn Belt in the United States (Crosson and Rosenberg, 1993) considered an area bounded by four states (Missouri-Iowa-Nebraska-Kansas or MINK). Water resources issues were still addressed even though upstream sub-basins were located outside the MINK region. Despite some obvious problens, can partial
29 integration provide useful input to the debate on policy responses to climate change? How could this more limited framework be designed so that climate-society issues could still be addressed, while recognizing the limitations that prevent consideration of all factors? 1.2 Purpose In order to attract the breadth of expertise and interests needed for an integrated assessment with stakeholder collaboration, some common ground must be laid out. Many impact assessments have focused on individual sectors (e.g. agriculture, wildlife, water resources), and while these can provide important technical information on direct 'first-order' impacts (IPCC 1990; Tegart and Sheldon, 1993), a wide range of external factors are often assumed to remain unchanged (Carter et al., 1992). Regional and national assessments have been produced elsewhere (e.g. Henderson and Coils, 1993; Hulme et al., 1992; Liverman, 1992; New Zealand Climate Change Programme, 1990; Ninh et al., 1991; Nishioka et al., 1993; Smith and Tirpak, 1990), but these have generally consisted of parallel sectoral studies. Crosson and Rosenberg (1993) and Parry et al. (1992) have attempted integrated assessments based on regions defined by sectoral dominance (e.g. agriculture) and/or political borders. The purpose here is to suggest that riverine hydrologic systems may provide an appropriate setting for producing an integrated regional assessment of climate change scenarios. What follows is a description of a watershed-based case study from northwest Canada, which is still in progress. The regional/watershed focus has been used to attract scientific expertise and diverse stakeholders with local knowledge. The common ground for all of them is an interest in the future of this place, with water serving as an important link.
2. Watersheds as Integrators of Natural and Human Systems The choice of study area can influence many aspects of an impact assessment, including the identification of issues and the collection of data. There are various administrative and ecological settings that might be considered (Carter et al. 1992), but the focus here is exclusively on watersheds as integrators. Land cover and land use affects hydrology and water quality, so water users (e.g. hydroelectric utilities, fisheries, navigation, domestic users, wildlife, agriculture, recreation) are necessarily linked with forces that modify the landscape (e.g. agriculture, forestry, hydroelectric utilities, industrialization, fire, pests). Governments have used watersheds as the basis for the creation of customized management structures (e.g. basin commissions, water boards). These attempt to reconcile the goals of competing interests while providing direction for regulation, water allocations and other matters. For the purpose of climate impact assessment, this is an important source of information on regional issues and their stakeholders. Although economic data are rarely collected on a watershed basis, it should be possible to at least partially address this requirement through the use of census data or other local/regional sources of information. Watershed-based assessments have been tried for the North American Great Lakes (Smith and Tirpak, 1990; Mortsch et al., 1993) and a series of international basin studies including the Nile, Indus and Zambezi (Strzepek and Smith, forthcoming). Arnell et al. (1994) provide a case that focusses on a water management authority in the United Kingdom. It is expected that climate warming would lead to an acceleration of the water cycle, with increased throughputs along the pathways linking atmosphere, ocean, landscape, freshwater
30 bodies, and society (Falkenmark, 1991). Climate warming may impose its most significant effects on water sensitive sectors, through changes in a) the frequency and severity of extreme events (floods, drought, etc.), b) timing of seasonal and annual events (e.g. spring runoff peak, autumn low flow, ice formation and break up, etc.), c) thresholds and ranges (e.g. maximum summer water temperatures), and d) land cover (e.g. erosion, fire, etc). Riverine hydrological systems will exhibit basin-specific adjustments to global climatic changes. Most warming scenarios tend to show increases in precipitation, but this does not necessarily mean wetter land surfaces or more soil moisture. Gleick (1993) concludes that if the future climate will not look like the past, there will be a great increase in the overall uncertainty associated with water management and supply. Understanding impacts is a necessary prerequisite for determining the kind of measures that could promote both limitation of greenhouse gas emissions and adaptation to environmental stresses. Assessing adaptation options requires a greater understanding of how individuals, companies and governments operate when faced with environmental stresses (Smit, 1993).
3. Case Study: Mackenzie Basin, Canada The Mackenzie Basin Impact Study (MBIS) is part of the Government of Canada's Green Plan, and has passed the halfway point in its six-year mandate to assess the potential regional implications of global climatic change (Cohen, 1993, 1994). The program includes studies on water resources, permafrost, vegetation, wildlife, economic activities, resource-based and subsistence-based communities, and applications of remote sensing and geographic information systems (GIS). Attention is also given to the challenges of producing an integrated assessment, and to incorporating traditional ecological knowledge into the MBIS.
3.1 Setting This region was chosen because it is a major high latitude watershed, 1.7 million km 2 in area, with many climate-sensitive landscapes and transition zones: tree lines (Arctic, montane, aspen parkland), discontinuous permafrost, wetlands and deltas, the edge of multiyear sea ice, and the northern limits of commercial forestry and agriculture. Freshwater and terrestrial migratory wildlife might be sensitive to climate-induced changes in the landscape. The study area is defined by the watershed boundary of the Mackenzie River and its ti'ibutaries, plus the southern Beaufort Sea and coastal zone north and east of the Mackenzie Delta. The large size precludes detailed study of all areas, though there are some activities that are conducted on the Basin scale. For the MBIS to achieve its objective, however, the range of impact-related policy questions are limited to the following: a) interjurisdictional water management, b) sustainability of native (aboriginal) lifestyles, c) economic development opportunities, d) maintenance of infrastructure, and e) sustainability of ecosystems. Additional focus has been provided by defining critical regions within the study area (Figure 1). Each of these represent potential flash points due to the intersection of potential biophysical changes with human activities. For example, the Upper Peace River region includes a major hydroelectric facility (Bennett Dam), agriculture, expanding forestry operations and communities with a history of flooding. The operation of the dam has led to concerns about the viability of the freshwater wetlands and delta, and consequently, the
32 wildlife and subsistence-based communities in the Peace-Athabasca Delta region, located downstream. How would a scenario of climate warming affect dam operations, water levels at the Delta, fisheries, migrating waterfowl, agriculture and forestry operations? Would resource-based and native communities experience the same impacts, or would climate change be felt in different ways depending on lifestyle (wage economy, subsistence/non-wage economy)?
3.2 Objective If climate warming occurs, governments and their constituents will need advice on how to adapt to the new climate. Since decision making occurs in an environment where different stakeholders compete for resources, any response options will have to account for tradeoffs between these various interests. Land and water use patterns today represent the result of historic and current compromises between these various interests, combined with knowledge gained from research and personal experience. At the scale of most current GCM-based impact assessments (e.g. grid sizes larger than 2 ~ latitude x 2 ~ longitude), land in a grid cell is not necessarily assigned to a single optimal use today, so it is unlikely that this would be different in the future. The assessment, therefore, should not restrict itself to changes in physical capability to support a particular activity (e.g. crop production). The objective of MBIS is to provide an integrated regional assessment of scenarios of climate warming for regional stakeholders and the scientific community. As a high latitude watershed, the Mackenzie Basin has been seen as an area that might benefit in certain ways by a warmer climate. These include a) longer growing season for agriculture, b) greater productivity for forestry, c) longer ice-free season for navigation, d) reduced energy demand for space heating, e) longer summer tourist season, and f) reduced cold weather stress on infrastructure. Taken individually, economic impacts could be quantified, and these might show substantial benefits for the region. Other factors need to be considered, however, and some of these may constrain the potential benefits. This list includes: a) current use of land for subsistence hunting and trapping, b) current system of land transportation, much of which is based on a stable ice and snow cover for winter roads, c) current ranges and habitats of wildlife, which underpin conservation plans and native land claims (currently being negotiated between aboriginal people and governments in Canada), and d) scientific uncertainty which hampers anticipatory responses to projected beneficial conditions. Potential negative impacts of climate warming must also be considered, because they may offset possible benefits. Examples are: a) increased erosion due to permafrost thaw, b) increased frequency and severity of forest fires, c) extension of mid-latitude pests and diseases into high latitudes, and d) reduction of habitat suitable for cold climate species of vegetation and wildlife. 3.3 Study Framework MBIS is attempting to produce an integrated regional assessment of global warming scenarios, as a way of identifying the indirect linkages between climate and regional policy concerns, such as land and water management. Several exercises are being tried, including 1) resource accounting with input-output modelling, 2) land assessment (including goal programming and multiobjective program modelling), 3) review of water resources policy instruments and their sensitivity to hydrologic changes, and 4) study of settlement patterns and their sensitivity to landscape changes. Each of these utilize the outputs of various
33 individual studies in order to address some of the human dimensions of climatic change (Cohen, 1993, 1994). All of these approaches are being tried because there is no consensus on which method is best for producing an integrated study. System models (1 and 2 above) provide a closed integrated model or set of linked models that describe particular components of the system. Analyses based on planning/management instruments (3 and 4 above) consist of a mixture of models and expert judgement. These instruments (e.g. plans, policies, regulations, indices) represent the integration of scientific information and stakeholders' preferences, and their performance under climate change scenarios would provide an important measure of impact. There is a difference between the level of control exerted by the researcher in these approaches. While the idea of "megamodels" (Frederick and Rosenberg, 1994) is growing in popularity in Europe and North America, the family of integrators concept presented here serves to provide an opportunity for other forms of input to contribute to the assessment, particularly those which are difficult to quantify. Policy analysis has both quantitative and qualitative aspects, and may be preferred by stakeholders who are leery of 'black box' models. There are several opportunities to facilitate linkage between individual study components. Within MBIS, integrated system models, economic models, and other similar tools, are being used to address complex issues related to land use and economic growth (e.g. Lonergan, 1994; Yin and Cohen, 1994; Huang et al., 1994). These mathematical or statistical techniques require a wide range of inputs, including census data, outputs of other models, and/or indices obtained from remote sensing, thereby serving as integrators of information obtained from other disciplines.
3.4 Preliminary Results One theme that has clearly emerged in the MBIS is that climate is a complex agent of change. Although scientific and political discussions have tended to focus on atmospheric change, the land and its people will likely experience climate warming through changes in streamflow, water levels, ice and snow cover, permafrost, plant growth, wildlife patterns, fire, pests and diseases. Some changes may occur gradually while others may come in the form of large steps or new extremes. The linkage between changes in air temperature and regional socio-economic concerns is largely through these landscape 'filters.' Biophysical changes are what people will notice before they pay attention to climate statistics. Has the winter road season changed? Is anything new with the caribou migration? Are current fire management strategies still working satisfactorily? What is the status of permafrost along the Mackenzie Valley and the Beaufort coastal zone? Some preliminary indications of landscape and socioeconomic impacts for the scenarios being assessed by MBIS are shown in Tables 1 and 2, respectively. Many MBIS activities are not yet at the stage where scenario results can be reported, but some information is available.
34 Table 1 MBIS Preliminary Summary of Landscape Im )acts of Climate Warming Scenarios
PARAMETER
DETAILED IMPACTS
Permafrost thaw occurs, but rate of change varies with site
*thaw would occur primarily in discontinuous zone *seasonal active layer would increase *rate of thaw in wetland areas would lag behind other sites *slopes and Beaufort Sea coastal zone may experience accelerated erosion
Water Supply changes slightly, with earlier spring peak
*annual Basin runoff changes -7 % to -3 % in GCM-based scenarios, +7 % in composite analogue scenarios *increased precipitation offset by increased evapotranspiration in many subbasins *spring snowmelt peak begins up to 1 month earlier *longer snowmelt season, lower peak in some subbasins (including Williston, upstream of Bennett Dam)
Peace River Ice Cover reduced in duration and extent
9ice cover reduced by up to 4 weeks 9upstream progression of ice reduced by up to 200 km 9runoff reduction (or reduction of discharge from Bennett Dam) would offset effects of temperature increase on ice cover
Soil Capability for Agriculture increases
9increase in availability of marginal and suitable land for spring seeded small grains and forages due to longer growing season and frost free period 9decrease in soil moisture supply
Pine Weevil Hazard increases
9increase in temperature-based pine weevil hazard index 9low elevation sites particularly vulnerable 9non-temperature factors not yet included
Fire Weather Index increases
9median index for four GCM-based scenarios corresponds to change o f - 15 % to + 81% in burned area
Summarized from Cohen (1993, 1994).
35 Runoff for the Basin was obtained using a square grid model (Soulis et al., 1994), and for the Williston subbasin with the UBC Watershed Model (Chin and Assaf, 1994). Although increased runoff was anticipated (e.g. see Miller and Russell, 1992), this does not appear to be the case for the GCM-based scenarios (Canadian Climate Centre or CCC, Geophysical Fluid Dynamics Lab or GFDL (R30 version)) for the Basin as a whole. Only the composite analogue scenario shows an increase. Newton (1994) has therefore concluded that scenario spring flood risks for vulnerable communities may not be that different from current climatic conditions. What is not clear as yet is the implication of hydrologic and landscape changes on water management agreements currently being negotiated by various governments (Felton, 1994). Peace River ice cover, for example, will be affected by both temperature changes and changes in outflow from the Bennett Dam at Williston subbasin (Andres, 1994). This may not be the final word on runoff impacts, since the Global Energy and Water Cycle Experiment (GEWEX) is pursuing a research programme in the Mackenzie (Lawford, 1994). It would appear that the other main threats to the Mackenzie landscape are a) accelerated erosion caused by permafrost thaw, especially in sloping terrain and the Beaufort Sea coastal zone (Aylsworth and Egginton, 1994; Solomon, 1994), b) increased fire hazard (Kadonaga, 1994), and c) invasion of new pests and diseases from warmer regions (Sieben et al., 1994). These landscape impacts could lead to changes in plant succession (Wein et al., 1994), thereby affecting wildlife habitat and subsistence activities of native communities. Additional information on ecosystem impacts should become available for the MBIS Final Report in 1997. First-order and second-order impacts eventually lead to others which are considerably more difficult to address. Will land claims or water resources agreements be affected? Could there
Table 2 MBIS Preliminary Summary of Socio-Economic Impacts of Climate Warming Scenarios SECTOR/LOCATION
DETAILED IMPACTS
Tourism/Nahanni National Park would experience mixed impacts
9little impact from projected minor changes in streamflow 9extended season for water-based recreation would provide economic benefits to communities near the Park 9increased Fire Weather Index (fire frequency and severity) could affect runoff, landscape character, visitor safety
Community Vision of Impacts depends on vision of lifestyle
9response to flood hazard varies by community, according to the interplay of individual, community and government responses 9significance of landscape impacts depends on whether community maintains subsistence lifestyle, or switches to wage economy
Summarized from Cohen (1994).
36 be new conflicts over land use, especially if agriculture expands northward to take advantage of improved soil capability to support crop production (Brklacich and Curran, 1994)? What might be the effects on parks and other protected areas (Pollard and Benton, 1994)? Could climate change affect the economics of oil and gas production in the Beaufort Sea (Anderson et al., 1994)? Expressing socio-economic impacts in monetary terms is going to be difficult, but it should be possible to do so for agriculture, forestry, energy, and some aspects of tourism. In the case of Nahanni Park located in the Liard subbasin (see Figure 1), water-based recreation is expected to benefit from the longer summer, but this could be offset by the threat of increased fire (Staple and Wall, 1994). There is no assessment, yet, on the potential costs of increased fire or fire protection. Community impacts could be quantified, but the effects of climate warming scenarios may vary depending on whether a traditional aboriginal lifestyle of hunting and trapping is maintained, or a shift to greater reliance on the wage economy occurs. Aharonian's (1994) case study of Aklavik, in the Mackenzie Delta region (see Figure 1), shows that residents can provide detailed visions of both "futures." In their view, community vulnerability to climate warming scenarios will change if their lifestyles changes. This may parallel circumstances that could be experienced in some developing countries during the next several decades. The integration component is currently focussed on data collection. One activity is on the development of a resource accounting framework, including a Mackenzie Basin input-output model. This will be used to determine impacts of changes in energy and forestry on the region's employment and economic productivity (Lonergan, 1994). A second modelling exercise is the integrated land assessment framework or ILAF. Its purpose is to compare changes in land capability with stakeholders' goals in order to identify possible land use conflicts in a climate warming scenario (Yin and Cohen, 1993, 1994). Potential expansion of commercial agriculture and forestry could create a conflict with existing subsistence activities, so there is a need to determine whether this is possible within the scenarios. Additional activities in multiobjective programming (Huang et al., 1994), and a study of the non-wage economy in a native community, will complement ongoing MBIS socioeconomic studies in agriculture, forestry, energy, tourism and community development (Cohen, 1994). Impacts and responses will not be felt by individual sectors in an isolated manner. A unit of land (at a scale comparable to GCM output) is not likely to end up becoming exclusively devoted to one kind of land cover or use. This set of research activities will hopefully enable MBIS to address some important cross-cutting issues at a scale comparable to regional stakeholders' interests.
4. Conclusions
A riverine hydrologic system is presented as an appropriate setting for integrated regional assessment of climatic warming scenarios. The Mackenzie Basin Impact Study (MBIS) illustrates the application of the "family of integrators" approach, consisting of several integrated system models and analyses of policy instruments. We have considered the difficulties in producing a fully integrated assessment of climate warming scenarios, and acknowledge that in the case of the MBIS, several aspects are not covered (e.g. marine wildlife in the Beaufort Sea, native communities in Alberta, future
37 economic linkages with the rest of Canada and other countries). MBIS includes population and economic growth scenarios (Lonergan and Difrancesco, 1993), but technological and institutional change scenarios have not been constructed. Although it is unlikely that MBIS can achieve full integration, we hope that partial integration can provide relevant information on sectoral and cross-cutting regional impacts. MBIS is an exercise in interdisciplinary research with stakeholder collaboration. Maintaining linkages between researchers and stakeholders has been a challenge. It may be difficult at this stage to appreciate the long term value of the MBIS experience, but it is clear that collaboration with stakeholders is vital for there to be any hope of producing an assessment that could be useful and relevant to the region of interest. In fact, partially or fully integrated assessments may be impossible without stakeholder involvement during all phases of research. For example, stakeholders participating in MBIS planning meetings contributed to the selection of economic growth scenarios, and the identification of communities and individuals willing to be interviewed as part of surveys conducted by MBIS investigators. During the remainder of the MBIS program, investigators will be completing biophysical and socio-economic impact studies, transferring information to the "integrators" (i.e. systems modellers, policy analysts, etc.), and completing integration exercises. There will be a workshop on water management, and a larger gathering in 1996 similar to the event that facilitated the production of MBIS Interim Report #2 (Cohen, 1994). MBIS investigators are expected to exchange information with each other before and after their components are completed. There are also plans for more discussions on the MBIS within the region, before and after publication of the final report in 1997.
5. ACKNOWLEDGEMENTS My thanks to Krystyna Czaja for producing Figure 1. Any opinions expressed in this chapter are my own, and not necessarily those of Environment Canada.
6. REFERENCES Aharonian, D. (1994). Land use and climate change: an assessment of climate-society interactions in Aklavik, NWT. In Cohen, S.J. (ed.), 410-420. Alcamo, J. (ed.) 1994. IMAGE 2.0: Integrated modelling of global climate change. Water, Air and Soil Pollution, 76, 1/2 (special issue), 1-318. Anderson, W.P., R. Difrancesco and M. Kliman. 1994. Potential impacts of climate change on petroleum production in the Northwest Territories. In Cohen, S.J. (ed.), 433-441. Andres, D. 1994. Peace River ice regime: an interim report. In Cohen, S.J. (ed.), 237-245. Arnell, N.W., A. Jenkins and D.G. George. 1994. The Implications of Climate Change for the National Rivers Authority. Institute of Hydrology R&D Report 12, National Rivers Authority, Bristol, United Kingdom. Aylsworth, J.M. and P.A. Egginton. 1994. Sensitivity of slopes to climate change. In Cohen, S.J. (ed.), 278-283. Bielawski, E. 1994. Lessons from Lutsel k'e. In Cohen, S.J. (ed.), 74-76.
38 Brklacich, M. and P. Curran. 1994. Climate change and agricultural potential in the Mackenzie Basin. In Cohen, S.J. (ed.), 459-464. Carter, T.R., M.L. Parry, S. Nishioka and H. Harasawa. 1992. Preliminary Guidelines for Assessing Impacts of Climate Change. Environmental Change Unit, Oxford, and Center for Global Environmental Research, Tsukuba. Chin, W.Q. and H. Assaf. 1994. Impact of global warming on runoff in Williston Basin. In Cohen, S.J. (ed.), 210-236. Cohen, S.J. (ed.). 1993. Mackenzie Basin Impact Study Interim Report #1. Environment Canada, Downsview, Ontario. Cohen, S.J. (ed.). 1994. Mackenzie Basin Impact Study Interim Report #2. Environment Canada, Downsview, Ontario. Crosson, P.R., and N.J. Rosenberg. 1993. An overview of the MINK study. Climatic Change, 24, 159-173. Egginton, P. 1993. Permafrost south of the Beaufort coastal zone. In S.J. Cohen (ed.), 5258. Falkenmark, M. 1991. The Ven Te Chow memorial lecture: Environment and development: urgent need for a water perspective. Water International, 16, 229-240. Felton, G. 1994. A review of interjurisdictional water management in Canada. In S.J. Cohen (ed.), 67-73. Frederick, K.D. and N.J. Rosenberg (eds.). 1994. Assessing the impacts of climate change on natural resource systems. Climatic Change, 28, nos. 1-2 (special issue), 1-219. Glantz, M.H. (ed.) 1988. Societal Responses to Regional Climatic Change: Forecasting by Analogy. Westview Press, Boulder. Glantz, M.H., R. Katz and M. Krenz (eds). 1987. The Societal Impacts Associated with the 1982-83 Worldwide Climate Anomalies. National Center for Atmospheric Research, Boulder. Gleick, P.H. 1993. Water in the 21st century. In Gleick, P.H. (ed.), Water in Crisis." A Guide to the World's Fresh Water Resources. Oxford University Press, New York, 104-113. Henderson-Sellers, A. 1993. An antipodean climate of uncertainty. In Henderson-Sellers and Colls (eds.), 203-224. Henderson-Sellers, A. and K. Coils (eds.). 1993. Climatic impacts in Australia. Climatic Change, 25, nos. 3-4 (special issue), 201-438. Huang, G.H., Y.Y. Yin, S.J. Cohen and B. Bass. 1994. Interval parameter modelling to generate alternatives: a software for environmental decision-making under uncertainty. In Brebbia, C.A. (ed.), Computer Techniques in Environmental Studies. Kluwer Academic Publishers, Dordrecht. Hulme, M., T. Wigley, T, Jiang, Z.-c Zhao, F. Wang, Y. Ding, R. Leemans, and A. Markham. 1992. Climate Change due to the Greenhouse Effect and its Implications for China. World Wide Fund for Nature, Gland, Switzerland. Intergovernmental Panel on Climate Change (IPCC). 1990. Climate Change: The IPCC Impacts Assessment. W.J.McG. Tegart, G.W. Sheldon and D.C. Griffiths (eds.). Australian Government Publishing Service, Canberra. Inuvik, Community of. 1993. Inuvik Inuvialuit Community Conservation Plan. Available from Wildlife Management Advisory Council, Inuvik, Northwest Territories. Kadonaga, L. (1994). Fire in the environment. In Cohen, S.J. (ed.), 329-336.
39 Kearney, A.R. 1 9 9 4 . Understanding global change: a cognitive perspective on communicating through stories. Climatic Change, 27, 419-441. Lawford, R.G. 1994. Knowns and unknowns in the hydroclimatology of the Mackenzie River Basin. In Cohen, S.J. (ed.), 173-196. Liverman, D. 1992. The regional impact of global warming in Mexico: Uncertainty, vulnerability and response. In Schmandt, J., and J. Clarkson (Eds.), The Regions and Global Warming: Impacts and Response Strategies. Oxford University Press, New York, 44-68. Lonergan, S. 1994. Natural resource/environmental accounting in the Mackenzie Basin. In Cohen, S.J. (ed.), 39-42. Lonergan, S., and R.J. Difrancesco. 1993. Baseline population and economic growth simulation. In Cohen, S.J. (ed.), 131-139. Miller, J.R. and G.L. Russell. 1992. The impact of global warming on river runoff. Journal of Geophysical Research, 97, D3, 2757-2764. Mortsch, L., G. Koshida and D. Tavares (eds.). 1993. Adapting to the impacts of climate change and variability. Proceedings of the Great Lakes - St. Lawrence Basin Project Workshop, 9-11 February, 1993, Quebec City. Environment Canada, Downsview, Ontario. Newton, J. 1994. Community response to episodes of flooding in the Mackenzie Basin. In Cohen, S.J. (ed.), 421-430. New Zealand Climate Change Programme. 1990. Climatic Change: Impacts on New Zealand. Ministry for the Environment, Wellington. Ninh, N.H., M.H. Glantz and H.M. Hien. 1991. Case Studies of Climate-Related Impact Assessment in Vietnam. UNEP Project Document No. FP/4102-88-4102. United Nations Environment Programme, Nairobi. Nishioka, S., H. Harasawa, H. Hashimoto, T. Ookita, K. Masuda and T. Morita (Eds.). 1993. The Potential Effects of Climate Change in Japan. Center for Global Environmental Research, Tsukuba, CGER-I009-'93. Parry, M.L., M. Blantran de Rozari, A.L. Chong and S. Panich. 1992. The Potential SocioEconomic Effects of Climate Change in Southeast Asia. United Nations Environment Programme, Nairobi. Pollard, D.F.W. and R.A. Benton. 1994. The status of protected areas in the Mackenzie Basin. In Cohen, S.J. (ed.), 23-27. Rosenzweig, C. and M.L. Parry. 1994. Potential impacts of climate change on world food supply. Nature, 367, 133-138. Sieben, B.G., D.L. Spittlehouse, R.A. Benton and J.A.McLean. 1994. A first approximation of the effect of climate warming on the white pine weevil hazard in the Mackenzie River Drainage Basin. In Cohen, S.J. (ed.), 316-328. Smit, B. (Ed.). 1993. Adaptation to Climatic Variability and Change. Report of the Task Force on Climate Adaptation, Canadian Climate Program. Department of Geography, University of Guelph, Occasional Paper No. 19. Smith, J.B. and D.A. Tirpak (Eds.). 1990. The Potential Effects of Global Climate Change on the United States. Report to Congress, United States Environmental Protection Agency, Washington. Solomon, S. 1994. Storminess and coastal erosion at Tuktoyaktuk. In Cohen, S.J. (ed.), 286-292.
40 Soulis, E.D., S.I. Solomon, M. Lee and N. Kouwen. 1994. Changes to the distribution of monthly and annual runoff in the Mackenzie Basin using a modified square grid approach. In Cohen, S.J. (ed.), 197-209. Staple, T. and G. Wall. 1994. Implications of climate change for water-based recreation activities in Nahanni National Park Reserve. In Cohen, S.J. (ed.), 453-455. Strzepek, K.M., and J.B. Smith (eds.), forthcoming. As Climate Changes: The Potential International Impacts of Climate Change, Cambridge University Press, New York. Tegart, W.J.McG., and G.W. Sheldon. (eds). 1993. Climate Change 1992: The Supplementary Report to the IPCC Impacts Assessment. Australian Government Publishing Service, Canberra. Wein, R., R. Gal, J.C. Hogenbirk, E.H. Hogg, S.M. Landh~iusser, P. Lange, S.K. Olsen, A.G. Schwarz and R.A. Wright. 1994. Analogues of climate change - fire vegetation responses in the Mackenzie Basin. In Cohen, S.J. (ed.), 337-343. Yin, Y., and S.J. Cohen. 1993. Integrated land assessment framework. In Cohen, S.J. (ed.), 151-163. Yin, Y., and S.J. Cohen. 1994. Identifying regional policy concerns associated with global climate change. GlobalEnvironmental Change, 4, 246-260.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
41
Assessing the impacts of climate: The issue of winners and losers in a global climate change context Michael H. Glantz Environmental and Societal Impacts Group National Center for Atmospheric Research* PO Box 3000 Boulder, Colorado USA 80307-3000
1. INTRODUCTION Most reviews of the greenhouse issue begin with the works in the mid-1890s of Swedish scientist Arrhenius. The physical processes have been well known for more than a century. Interest in the possible impacts on climate of CO2 emissions as a result of human activities has waxed and waned since that time, with temporary peaks of interest appearing in the mid1930s (Callendar, 1938), the mid-1950s (Revelle and Suess, 1957), and again in the mid1970s (e.g, Kellogg, 1977). Today we are inundated by assessments of the prospects of a global warming and its possible impacts on society and the environment produced by national, international, and nongovernmental organizations. Discussions of such a prospect have steadily increased during the past twenty years, reaching very high political levels in the late 1980s and early 1990s. The century-long interest in the science and impacts of the human-induced enhancement of the greenhouse effect has been interrupted partly by other more pressing and urgent historical events such as two world wars, a worldwide depression, decolonization, the Cold War and then its demise, and a temporary global cooling; and partly by the fact that the impacts of a CO2-induced global warming were originally believed to be beneficial to society. For example, Callendar (1938) suggested that a greenhouse warming would help to thwart the emergence of an apparently imminent Ice Age. Scientific and anecdotal evidence was cited to suggest that the earth was coming to the end of an interglacial period, and that soon processes leading to an Ice Age would begin again. From about 1940 to the late 1960s, the global atmosphere underwent a yet-to-beunderstood cooling. Scientists provided scientific as well as anecdotal evidence (convincing both to the lay public and segments of the scientific community) to support the view that the earth was possibly on the threshold of an Ice Age: the growing season in England had been shortened by two weeks, fish species formerly caught off the northern coast of Iceland began
*The National Center for Atmospheric Research is sponsored by the National Science Foundation.
42
appearing only off its southern coast, sea ice in the North Atlantic had increased in its southward extent in the early 1970s and was appearing in shipping lanes that were normally ice-free; and hay production in Iceland declined by 25% as a result of less hospitable weather. In the United States, the fact that the armadillo, which had migrated as far north as Kansas in wanner decades, was starting to retreat toward the south was also used as evidence to support the Ice Age hypothesis. Geologic records were invoked as well, to show that an Ice Age was near. During this brief period of concern with global cooling, one issue widely considered was how it might affect the relative economic and political positions of different countries. Even the US Central Intelligence Agency undertook a set of studies to show how the cooling might affect the agricultural production and energy demand in the USSR (CIA, 1974, 1976). The Ecologist examined the potential impacts of a few degrees of cooling on agriculture in the Canadian prairies (Goldsmith, 1977). Some books and articles on the topic went so far as to identify specific countries that would become climate-related world powers in the event of a cooling. For example, Ponte (1976) suggested that "adapting to a cooler climate in the northern latitudes, and to a drier climate nearer the equator, will require vast resources and almost unlimited energy .... A few countries, such as equatorial Brazil, Zaire, and Indonesia, could emerge as climate-created superpowers." He also suggested that "We can say with high probability today that the global monsoon rainfall will be below average for the remainder of the century." Another book on the possibility of a global cooling (Impact Team, 1977) suggested that with a cooling "there would be broad belts of excess and deficit rainfall in the middle latitudes; more frequent failure of the monsoons that dominate the Indian subcontinent, south China, and western Africa; shorter growing seasons for Canada, northern Russia, and north China. Europe could expect to be cooler and wetter. Of the main grain-growing regions, only the United States and Argentina would escape adverse effects." There was no reluctance whatsoever to discuss who might win and who might lose, or to identify specific countries or specific economic sectors within a country as winners and as losers in the event of a global cooling. There is a striking difference between the scientific and political responses in the 1970s to a potential cooling and those of today to a warming. Today there is a strong reluctance, if not opposition, within scientific as well as policymaking circles to recognize (or address or discuss) the existence and identity of specific winners and losers, especially winners. When he was a US Senator, US Vice President Albert Gore (1992), for example, argued that there would be no winners in the event of a global warming, a view that is apparently also held by the US Environmental Protection Agency (EPA). Soviet scientist Mikhail Budyko (1988), in contrast, asserted that everyone would benefit from a global warming based on scenarios, plausible from his perspective. Perhaps the comments that US Senator Tsongas (1982) made about diametrically opposing views on the energy crisis of the 1970s and 1980s apply to the views of Gore and Budyko on winners and losers: Both of these approaches are equally absurd, equally rhetorical, and equally successful. When talking to the convinced, they are very powerful. And that is basically how most people address the issue: we are awash in rhetoric, not to mention hypocrisy, when what we need is a careful sorting and weighing of the facts and values involved in making ~ or not making m a decision. Many people seem to believe that discussing winners and losers (or, as some prefer to call them, advantaged and disadvantaged) will be divisive and could ultimately undermine
43 efforts to put together a global coalition truly intent on combating anthropogenically induced global warming. Reaction to a 1989 speech by Barber Conable, then-President of the World Bank, illustrates that discussion of winners and losers has, at least up to the recent past, been politically taboo. Environmental groups, which have been marching lock-step on this particular issue, opposed his public comments. In addition, some US congressmen even went so far as to suggest the need for a closer scrutiny of the World Bank's activities and budget. For example, the Washington Post (12 September 1989) reported, "In a letter to Conable, Wisconsin Senator Robert Kasten, Jr. wrote, 'the Bank's failure to be on the front lines of efforts to fight global warming threatens the Bank's long-term financial support from Congress.'" A similar argument was raised with respect to preventive versus adaptive response strategies. When the US EPA released two reports in 1983 suggesting that global warming was inevitable (Seidel and Keye s, 1983) and, as a result, people should plan for a rising sea level (Hoffman et al., 1983), the Friends of the Earth publication Not Man Apart denounced the Agency for "throwing in the towel," while at the same time, President Reagan's science adviser denounced the EPA reports as "alarmist." There was a feeling that "premature" discussions about adaptive strategies with respect to global wanning would break down the development of a united effort to support the pursuit and enactment of preventive strategies. Proponents of preventive strategies wanted attention to focus mainly on prevention as the best way to cope with global warming. There is, however, one projected impact of global warming for which one is allowed to identify specific winners and losers M sea level rise. This is probably because it is the one impact of a global warming for which there may be no obvious winners at the national level. No one has been reluctant to identify specific losers associated with sea level rise (papers have identified winners at the subnational level, such as coastal engineering firms and people who would have beachfront property as a result of a neighbor's misfortune). In this regard, one could argue that the sea level rise problem is similar to the stratospheric ozone depletion problem m no readily apparent national winners can be identified. Such would probably not be the case for changes in rainfall distribution, water resources availability, agricultural production, fisheries productivity, and energy production and consumption. The purpose of this presentation is to foster discussion of issues associated with the process of identifying winners and losers. What factors, for example, must be taken into account in labeling a region, an activity, an economic sector, or a country a winner or a loser? How do perceptions compare with reality? Can wins and losses be objectively and reliably identified? What are the costs and benefits of not addressing this issue as opposed to addressing it openly? My intention is not to label specific countries as winners or losers. To do that, one could simply use any of the GCM-generated scenarios, the scenarios generated by paleoecological reconstructions, or assessments of recent environmental changes and label specific countries and regions within countries accordingly. My purpose is to draw attention to the importance of addressing the winner-loser issue. As a note of caution, any attempt to identify potential winners and losers could only be viewed as a preliminary first step, because of the possibility of climate change surprises. For example, when I sought to include an assessment of the impacts of freezes on citrus production in the state of Florida as part of a larger set of analogues to possible global warming regional impacts, EPA advised me to drop that case study, asserting (not suggesting) that with a global warming "there would be no more freezes in Florida"! We did the study
44 anyway (Miller, 1988). As it happened, the 1980s, cited by scientists as the warmest decade in North America on record, witnessed the largest number of freezes in central Florida in its 130 years of record. Thus, regional counter-intuitive climate surprises must be expected. Nevertheless, identification of winners and losers is happening behind the scenes and should be brought out into the open. I realize that there is a risk associated with identifying winners and losers. If winners and losers are identified with some degree of reliability, the potential for unified action against the global warming may be reduced. Winners will not necessarily want to relinquish any portion of their benefits to losers in order to mitigate the impacts of their losses. On the other hand, there is also a risk in not making such a distinction between winners and losers. While scientists and policymakers formally discuss only losses associated with a global wanning, others may perceive that there will be positive benefits as well. The result is that the proponents for action on global warming could be likened to the fable about the emperor's new clothes, professing there are no winners, while everyone agrees with them in public but privately believes the opposite. This could sharply reduce the credibility of the proponents for taking action, lessening the chances for any response, preventive, mitigative, or adaptive.
2. SCENARIOS OF WINNERS AND LOSERS In the following section, the notion of winners and losers is discussed in terms of climatic conditions. These conditions include today's global climate regime, an altered climate regime, and varying rates of change. Winners and Losers with Today's Global Climate Regime It seems obvious that, say, fifty years hence there will be some societies that benefit from whatever climate exists at the time. After all, with today's climate, we can identify climaterelated winners and losers. As an example, the following map (Figure 1) shows droughtprone regions in sub-Saharan Africa, some of which could be considered climate-related losers. Such maps, depicting drought-prone and flood-prone areas, exist for other regions around the globe. Gains and losses at all levels of social organization, from local to international, may result directly from climate changes or from human responses to those changes. While there are several spokespersons for the extreme views (i.e., that all will win or all will lose), in all cases of changes (both relative and absolute), some will benefit, while others will be adversely affected. In addition, some nations, sectors, and groups may be in a better position to respond or adapt to climate change, turning this to their future advantage. The currently identifiable relative advantages and disadvantages of different nations, sectors, and groups result from a combination of climatic factors (such as climate variability and the frequency and intensity of extreme meteorological events) and a wide range of unique (by country, region, sector, or group) economic, social, and political factors that must be taken into serious consideration in any analysis (for more discussion of this issue, see ESIG, 1990). The differences, attributable to climate factors (e.g., recurrent droughts or floods), are likely to persist, although the relative positions of those affected might change. Furthermore, if such differences become extreme, they can lead to population movements by the disadvantaged (i.e., generating environmental refugees) and to conflicts either within national borders or across them).
45
~ M O S T CRITICALLYAFFECTEDBYTHEDROUGHT As of June 1985
ALGERIA
I
I ...... LIBERIA GUINEA-BISSAU
GAMBIA
RWANDA. BURUNDI 7
ZAIRE
~!.:~. . . . . . . . ~iili~i~i U
Critically affected
~
Most critically affected
Localized drought is prevalent in several other countries.
~ w York Tim~, 8120185
Figure 1 One could easily argue that there has been little sustained (or effective) effort to date by climate-related winners to assist those who might be considered climate-related losers. Examples that reinforce low expectations about adequate humanitarian assistance from the industrialized countries are not difficult to find. We have seen, for example, that in the past several decades, foreign assistance has been frequently tied to political considerations (e.g., aid to Cambodia and South Vietnam in the 1960s and 1970s, or to Ethiopia in the 1980s). In the early 1970s when there were widespread droughts throughout the world (except in the United States), then-US Secretary of Agriculture Earl Butz spoke about how food exports from the United States would be a new tool in the nation's foreign policy negotiating kit. Despite public statements to the contrary, few leaders in countries chronically affected by the adverse impacts of today's climate believe that they can rely on long-term, politically neutral assistance from those favored by today's global climate. The Colorado River Compact of 1922 provides an example of a recent "climate change" in which winners and losers have been identified. The Colorado River Basin in the southwestern United States was divided into two parts, the Upper and Lower Basins. The flow in the system was estimated at about 15 million acre-feet (mar), based on the record for the previous 20-year period, 1900-20. The representatives of the various states in the basin agreed to divide in absolute terms 15 maf average annual flow equally between the two
46 basins: 7.5 maf for each basin (75 maf over a ten-year period). However, because the Upper Basin states thought that there would be more water in the system than 15 maf, they agreed to provide the lower basin states with 7.5 maf, thinking that they would benefit from any surplus that might exist (see Brown, 1988). Shortly after the agreement was signed, however, the Colorado River entered a period of low streamflow, setting record lows in the 1930s (referred to as the US Dust Bowl decade). Today, the average annual streamflow (Figure 2) is estimated at about 13.5 maf. The loss of streamflow has to be absorbed by the Upper Basin. Thus, in this situation, one can
LEE'S
/
m ~
13.5 Million Acre-ft./Year Long-term Average Flow
/'vvx. /'1~,
,,,UJ~oO:'~'----4 1- ~ k0. , ~
V
Compact Drafted
20
O~z 15. w~
FERRY
5 0 , 1564
,
-"%/ ,
I 1600
,
,
,
~J ,
I 1650
i
i
_..
~b/ ~ , . / i
l
I 1700
r,,
"%.I i
~ t
^ k./
~ I 1750
i-, ~
~ J i
i
,,,k,
"~#k/''"
~'j
I
I
I
l
1800
A k/N-,/
I
I 1850
I
...~
k.~-I
I
v I
I 1900
,
~ ~ ,
I 1950
Figure 2 (from Stockton and Jacoby, 1976) identify winners and losers as a result from what might be considered a climate change that has, to date, lasted about six decades. Carrying this analysis a step further, one might ask what those who benefited from the Compact have done to compensate those who have not? What lessons for climate change responses by society might be drawn from this situation? Should future water compacts be based on proportional divisions of a variable resource, instead of absolute amounts? What does this case study suggest about when to reach agreement on a variable resource m before or after winners and losers are identified? Finally, an important question that merits attention, but has yet to be addressed among discussions about possible strategic responses to global wanning, is the following: who loses and who wins if no action is taken and the global climate regime remains as it is today? If it could be ascertained that no global wanning were to occur, what actions would today's climate-related winners take to alleviate the climate-related problems of today's climaterelated losers?
Winners and Losers with an Altered Global Climate Regime While we do not yet know the global, let alone regional, specifics of the havoc (or windfall) that a climate change will bring, we can assume that there will be winners and losers with a global climate warming. Some researchers and policymakers who are primarily concerned about regional impacts believe that, compared to the present climate of their region, it is possible that their climate could improve rather than worsen with a global warming. Saudi Arabia is one such example; Ethiopia might be another. Given their current climate, they might consider the risk of change worthwhile. Bandyopadhyaya (1983), an Indian social scientist, as well as Budyko
47 (1988) of the Russian Federation have made this argument at length in favor of a climate warming. Often, when people talk about the possibility of increased rainfall in a given region, a counter-argument is raised that ambient temperatures (and, therefore, evaporation rates) will also increase. This would tend to negate any benefits that might come from additional rainfall. Yet, history shows that societies have devised ways to capture rainfall and reduce evaporation, thereby improving the percentage of rainfall that they can effectively use (Glantz, 1991). Can we trmd examples of environmental conditions that different societies might have to cope with in the advent of a global warming? Are there existing climate change analogues for most places in the world? For example, in the United States, it has been suggested that the state of Iowa would become hotter and drier. Might Nebraska or Kansas provide a glimpse at Iowa's possible future environmental setting and, therefore, a glimpse of Iowa's future? Attempts to identify climate analogues are not new. The following maps of the former USSR (CIA, 1974) and of China (Nuttonson, 1947), Figure 3 and 4, respectively, depict agroclimate analogues from North America. Similar analogue maps could be created pertaining to climate warming, once we have an improved picture of the regional impacts of a global warming.
Analogies "Forecasting by analogy" provides social scientists with another approach to identifying possible societal scenarios associated with climate-related environmental change. The objective of this approach is not to forecast future states of either the atmosphere or society. Its purpose is to identify present-day societal strengths and weaknesses in human responses to environmental changes in order to forecast society's ability to cope with stresses that might accompany an unknown climate future. It can provide researchers with a low-tech approach to scenario development that encourages researchers to rely on existing, thus reliable, information about the regional impacts of extreme meteorological events. It can also provide a first approximation of societal preparedness for coping with an as-yet-uncertain climate future. Each methodological approach to develop a global warming scenario generates highly speculative glimpses of the future. To date, no one has successfully identified a method to forecast with any degree of reliability future states of the atmosphere. It would, therefore, be misleading to rely on any one of these scenarios as a basis for making specific policy recommendations in a specific region or locale. Such scenarios should not be taken as predictions or forecasts. They can, however, be used to create awareness among policymakers of the need to assess the regional consequences of climate change (e.g., Glantz, 1988). Winners and Losers and Rates of Change Many environmental changes with which decisionmakers are concerned today derive from human activities: climate change, tropical deforestation, desertification, mangrove destruction, and varying lake and inland sea levels. Climatologists, environmentalists, and policymakers have sought to obtain numbers that characterize the rates of these changes at global, national, and regional levels. These rates of current environmental change are determined directly or indirectly from space and ground observations, combined with
49
,~
,i
~
Alberta .
9
Climatic Analogues Wyoming ....~.
'~: i~:. j.i .....
Minnesota ~: ~ ~ ..:o
North Monta~ 4::.
The NorthAmedcan ~ n s |ndcsted have temperatures aml minfail somewhatsimilarto areas in China, C o ~ n s such as these can only be suggestive.
Figure 4
Puerto ~ Rico
statistical measurements. Projections of future rates are obtained from modeling activities, extrapolations of present-day trends, and subjective "guesstimates." These rates are extremely important to the development of scenarios about environmental conditions (including climate) in future decades. They also have an important impact on the particular policies pursued to mitigate or adapt to those rates and processes of environmental change, as well as on attempts to fine-tune the methods of detection. Perceptions about rates of change of global warming can affect one's views of the costs or benefits of such a climate change. Rates of environmental changes are often as important as the magnitude of those changes over the long term. When rates take on a crisis element (e.g., high stakes involved, perceived threat, short time to act, thereby challenging a society's ability to adjust), decisionmakers appear to take a more serious view of environmental changes that might affect them. High rates are more likely to cause alarm, while slow rates foster a "business as usual" attitude. One would even argue that it is often perceptions about the rates of change that prompt human action and not necessarily the order of magnitude of that change, even if that magnitude of change is unprecedented and beyond recent human experience. Environmental change appears to be used frequently as a bargaining chip in domestic as well as international negotiations. More specifically, in this context rates of change will affect response by decisionmakers. Negotiating tactics and strategies will clearly be affected by the rates of environmental changes that are proposed or used in intra-national and international negotiations. Varying rates are used by different protagonists in debates about
50 policy responses that might be required to deal with such changes. Thus, the importance of taking the subjective elements out of rates and processes can lead to more informed, more integrated, and more objective decisionmaking and perhaps to a more realistic view of what regional and local gains or losses might accompany global warming. It is also important to note that for any group, relative advantages and disadvantages are likely to change over time and that what might appear to be an advantage from climate change in the near term may, in the long run, turn into a disadvantage and vice versa.
3. RELATED QUESTIONS Before any attempt to identify specific winners and losers from a global wanning, there are several "prior" questions that must be addressed. In this section, some of these questions are posed and only briefly discussed to stimulate more critical examination. The following is meant to be suggestive of the kinds of concerns that must be raised when assessing the societal impacts of a global wanning. What Do W e Mean By a Win or a Loss?
It is not sufficient, meaningful, or realistic to equate more rainfall than normal with a win, and less rainfall than normal with a loss. In reality, the actual amount of rainfall in a given location does not by itself tell much about agricultural production. There are numerous articles about definitions of drought (e.g., Wilhite and Glantz, 1985). Researchers have identified differences between meteorological, agricultural, and hydrologic droughts. For example, if the expected annual amount falls (no meteorological drought) but is distributed throughout the growing season at the wrong time with respect to crop growth and development, a sharp decline in agricultural production (an agricultural drought) could occur. Likewise, defining a win or a loss according to changes in evaporation rates also may not be very useful. If evaporation rates increase, and all else remains the same, then there will be a depletion of water resources. However, as noted earlier, people in many add and semiadd areas have devised ways to minimize the impacts of high evaporation rates by the way they collect, store, and use their available, often scanty, water resources. Thus, the dependence on a single physical parameter to identify the costs or benefits to a society of a climate change has severe limitations. How Does One Measure a Win or a Loss? One might suspect that Canada will be a winner because, as temperatures increase and the growing season lengthens, agricultural productivity will improve. But what will be the impacts on Canadian fisheries, the timing of seasonal snowmelt, or the Canadian ski industry? Another local-scale example of the difficulty associated with measuring wins and losses is provided by historical attempts to augment precipitation in a semiarid part of central Colorado (USA). Cloud seeders were hired to suppress hail, augment rainfall during the growing season, and reduce rainfall during harvest, in order to improve the productivity of hops for beer production. Another group of farmers growing other crops (e.g., lettuce) and ranchers with different moisture requirements in the same valley opposed these cloud-seeding activities. The conflict between the two factions became violent, and the operation was eventually halted. Thus, even within small areas there can be different responses to changes in rainfall, making an objective determination of a win or a loss exceedingly difficult.
51
Finally, if one group loses, but loses less than others, should they be considered as an absolute loser or relative winner?
Can Wins and Losses Be Aggregated? While wins and losses can be "added" together to produce a net figure, one must question the value of that figure. The wins (or losses) are not shared commodities. Those who lose may not benefit in any way from those who win. For example, when the Peruvian fishery collapsed, those fishermen who had focused their activities (fishing gear, fishmeal processing factors, etc.) on exploiting anchoveta were not prepared to take advantage of exploiting the sharp increase in shrimp populations that appeared along the Peruvian and Ecuadorian coasts. A country can expect to have both winners and losers within its borders in the event of a climate change. While the winners may be in a position to take care of themselves, someone will have to help the losers. Wins and losses cannot be meaningfully aggregated. A win is a win, and a loss is a loss. What Is the Relationship between Perceptions of Wins and Losses and Actual Wins and Losses? Given the uncertainties surrounding the regional impacts of a global warming, actual winners and losers within and between countries cannot be identified with any degree of confidence. Perhaps we will learn that in reality everyone will lose (or win) with a global wanning of the atmosphere. However, as long as some regions or countries perceive themselves to be winners (or losers), they will act according to this perception. Thus, the issue of winners and losers must be addressed openly, objectively, and scientifically, if we wish to minimize the chance that actions taken in response to a global wanning will be based on misperceptions (Jamieson, 1994). How Should One Deal with the Issue of Intergenerational Equity? Identifying winners and losers spatially, as well as temporally, must become a concern of those dealing with the global warming issue. Arguments about intergenerational equity have been invoked to generate support for taking action now against global warming. We are asked to take actions today to protect future generations from the environmental insults wrought by the present generation. But how can we generate support for inter-generational equity when we cannot even achieve intra-generational equity among the various groups and generations now living? It appears that we have come to believe that any change in the status quo is, by definition, a bad change. But the real answer to this question will depend on who is asked to respond. A Saudi Arabian might believe that any change in the current climate regime will most likely be better for future generations of Saudi Arabians than the existing one. The opposite belief might be held by a farmer in the US Great Plains. The truth of the matter is that most people fear change (e.g., Hoffer, 1952).
4. CONCLUSION Every discipline has dealt with the concept of winners and losers in one way or another n biology, political science, history, sociology, economics, geography, law, ecology, conflict resolution, risk assessment, game theory, and so on. Climate-related impact assessment as
52 a result of global warming is only the latest topic that requires consideration of winners and losers. There have been conflicting views on whether to identify specific countries as winners or losers in the event of a global warming of the atmosphere. There has also been a reluctance to discuss the possibility that there may be any winners at all. It is time to get beyond that reluctance and to ask questions that need to be addressed so that the notion of winners and losers can be assessed on a more objective and realistic level. There is a calculated risk in such a discussion. Once specific winners have been reliably identified, there may be a reluctance on their part to lend support for global action to combat a greenhouse warming. We must take this risk. Many issues must be resolved before we will be in a position to identify with any degree of confidence who those specific winners might be. In the meantime, other issues, such as equity, definition, measurement, and perception versus reality, must be addressed if we ever hope to identify with some degree of confidence how specific countries, economic sectors, and regions within countries can develop response strategies to climate change in the 21st century. With such information in hand, governments and nongovernmental organizations would be in a position to devise tactics and strategies for coping with global-warming-induced national and regional changes.
The Regionalization of Environmental Problems Given the resurgence of worldwide concern about "global" environmental issues and their regional causes or consequences, regional organizations could provide an effective arena for discussing, resolving, or averting regional conflicts related to environmental change. Perhaps the 1990s provide a "window of opportunity" for a review of regional organizations and their potential contribution toward resource-related cooperation and conflict resolution at the regional level. Whereas regional international organizations (functional as well as geographic) have often been relegated to marginal roles in the international political arena with regard to resources issues, this could change in the future. Such a change can be expected because, as climate regimes shift in response to global warming, so too will the location of some highly valued natural resources, sometimes across national boundaries, such as water resources and fish populations. As resources and people dependent on them "migrate" on land and in the marine environment, the risks of regional conflicts, as well as the opportunities for regional cooperation, are likely to increase. The time may be fight to talk about the "regionalization of environmental problems."
5. REFERENCES Bandyopadhyaya, J., 1983: Climate and World Order: An Inquiry into the National Causes of Underdevelopment. New Delhi, India: South Asian. Brown, B.G., 1988: Climate variability and the Colorado River Compact: Implications for responding to climate change. In M.H. Glantz (ed), Societal Responses to Regional Climate Change: Forecasting by Analogy. Boulder, Colorado: Westview Press, 279-305. Budyko, M.I., 1988: Anthropogenic climate changes. Paper presented at the World Congress on Climate and Development, 7-10 November 1988, Hamburg, Germany.
53 Callendar, G.S., 1938: The artificial production of carbon dioxide and its influence on temperature. Quarterly Journal of the Royal Meteorological Society, 64, 223-241. CIA (Central Intelligence Agency), 1974: Potential Implications of Trends in World Population, Food Production, and Climate. Report OPR-401, August. Washington, DC: CIA. CIA (Central Intelligence Agency), 1976: USSR: The Impact of Recent Climate Change on Grain Production. Report ER 76-10577 U. Washington, DC: CIA. ESIG (Environmental and Societal Impacts Group), 1990: On Assessing Winners and Losers in the Context of Global Warming. Report of Workshop held 18-21 June 1990 in St. Julians, Malta. Boulder, Colorado: ESIG, National Center for Atmospheric Research. Goldsmith, E., 1977: The future of an affluent society: The case of Canada. The Ecologist, 7, 160-194. Glantz, M.H., 1991: The use of analogies in forecasting ecological and societal responses to global warming. Environment, 33, No. 5, 10-33. Glantz, M.H. (ed.), 1988: Societal Responses to Regional Climatic Change: Forecasting by Analogy. Boulder, Colorado: Westview Press. Gore, A., 1992: Earth in the Balance: Ecology and the Human Spirit. New York: Houghton Mifflin Co. Hoffer, E., 1952: The Ordeal of Change. New York: Harper and Row Publishers. Hoffman, J.S., D. Keyes, and J.G. Titus, 1983: Projecting Future Sea Level Rise: Methodology, Estimates to the Year 2100, and Research Needs. EPA 230-09-007. Washington, DC: US EPA, Office of Policy & Resource Management. Impact Team, 1977: The Weather Conspiracy: The Coming of the New Ice Age. New York: Ballantine Books. Jamieson, D., 1994: Global environmental justice. In R. Attfield and A. Belsey (eds), Philosophy and the Natural Environment. Cambridge: Cambridge University Press. Kellogg, W.W., 1977: Effects of Human Activities on Global Climate: A Summary with Considerations of the Implications of a Possibly Warmer Earth. WMO Tech. Note 156 (WMO No. 486). Geneva, Switzerland: WMO. Miller, K.M., 1988: Public and private sector responses to Florida citrus freezes. In M.H. Glantz (ed), Societal Responses to Regional Climatic Change: Forecasting by Analogy. Boulder, Colorado: Westview Press, 375-406.
54 Nuttonson, M.Y., 1947: Ecological crop geography of China and its agro-climatic analogues in North America. International Agro-Climatological Series, Study No. 7. American Institute of Crop Ecology. Ponte, L., 1976: The Cooling. Englewood Cliffs, NJ: Prentice-Hall. Revelle, R., and H.E. Suess, 1957: Carbon dioxide exchange between atmosphere and ocean and the question of an increase of atmospheric carbon dioxide during the past decades. Tellus, 9, 18-27. Seidel, S., and D. Keyes, 1983: Can We Delay a Greenhouse Warming? Washington, DC: US Environmental Protection Agency. Stockton, C.W., and G.C. Jacoby, 1976: Long-term surface water supply and streamflow trends in the Upper Colorado River Basin. Lake Powell Research Bulletin, 18. Tsongas, P.E., 1982: Foreword. In K.A. Price (ed), Regional Conflict and National Policy. Washington, DC: Resources for the Future, xi-xiv. Wilhite, D.A., and M.H. Glantz, 1985: Understanding the drought phenomenon: The role of definitions. Water International, 10, 111-120.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
55
Sustainable Development and Climate Change R.K. Turner Centre for Social and Economic Research on the Global Environment (CSERGE), University of East Anglia, Norwich and University College London, United Kingdom
Abstract Sustainability is defined in terms of four overlapping positions, ranging from very weak to very strong sustainability. The core idea is of a non-declining capital stock (including natural capital) over generational time. Weak sustainability positions emphasise capital substitution possibilities and the power of technical process to mitigate resource depletion and pollution problems. Climate change and its associated risks and strong uncertainty are characterised by features which f a v o u r ' a strong sustainability approach incorporating the precautionary principle. Strong sustainability positions recognise constraints on substitution processes and incorporate ethical concerns such as intergenerational equity as a moral duty. Cost-benefit analysis is moderated via safe minimum standards which set GHGs concentrations and emissions abatement targets.
Introduction: Sustainable Development Concept Economists define sustainable development in terms of non-decreasing levels of utility, or income per capita, or real consumption per capita over time. In broad terms it involves providing a bequest from the current generation to the next of an amount and quality of wealth which is at least equal to that inherited by the current generation. This requires a non-declining capital stock over time and is consistent with the intergenerational equity criterion. The most publicised definition of sustainable development credited to the World Commission on Environment and Development also included an intragenerational equity criterion (WCED, 1987). Sustainability therefore requires a development process that allows for an increase in the wellbeing of the current generation (with particular emphasis on the welfare of the poorest members of society), while simultaneously avoiding uncompensated and 'significant' costs (including environmental damage costs) on future generations. Such a cost liability would reduce the 'opportunities' for future generations to achieve a comparable level of well-being (Pearce, Barbier and Markandya, 1990). The sustainability approach therefore is based on a long-term perspective, it incorporates an equity as well as an efficiency criterion, and it may also emphasise the need to maintain a 'healthy' global ecological system (Costanza et al., 1992). A spectrum of overlapping sustainability positions (from very 'weak' to very 'strong') can be distinguished, see Figure 1 (Turner, 1993). Weak sustainability requires the maintenance of the total capital stock- composed of
57 K~ (manufactured or reproducible capital); Kh (human capital, or the stock of knowledge and skills); K (natural capital: exhaustible and renewable resources, together with environmental structures, functions and services) - through time with the implicit assumption of infinite substitution possibilities between all forms of capital. The Hartwick Rule (Hartwick, 1978) is also used to buttress the weak sustainability position by regulating the intergenerational capital bequests. The rule lays down that the rent obtained from the exploitation of the natural capital stock by the current generation, should be reinvested in the form of reproducible capital which forms the future generations' inheritance. This inheritance transfer should be at a sufficient level to guarantee non-declining real consumption (well-being) through time. The implicit capital substitutability assumption underpins the further argument that extensive scope exists over time for the decoupling of economic activity and environmental impact. The decoupling process is mediated by technical progress and innovation. While total decoupling is not possible, and with the important exception of cumulative pollution, society's use of resources can be made more efficient over time (i.e. the amount of resources used per unit of GNP goes down faster than GNP goes up and the aggregate environmental impact falls). From the weak sustainability perspective a key sustainability requirement will be increased effective research and development, i.e. new knowledge properly embodied in people, technology and institutions. From the strong sustainability perspective some elements of the natural capital stock cannot be substituted for (except on a very limited basis) by manmade capital and therefore there is a concern to avoid irreversible losses of environmental assets. Some of the functions and services of ecosystems in combination with the abiotic environment are essential to human survival, they are life-support services (e.g. biogeochemical cycles) and cannot be replaced. Other multi-functional ecological assets are at least essential to human wellbeing if not exactly essential for human survival (e.g. landscape, space and relative peace and quiet). We might therefore designate those ecological assets which are essential in either sense as being 'critical natural capital'. Supporters of the "deep ecology " [VSS] position argue for a particular type of nonsubstitutability based on an ethical rejection of the trade-off between man-made and natural capital. The strong sustainability rule therefore requires that we at least protect critical natural capital and ensure that it is part of the capital bequest. The combination of the risk of irreversible environmental losses and a high degree of uncertainty surrounding past rates and future trends in resource degradation and loss, as well as the full structural and functional value of ecosystems (Gren, Folke, Turner and Bateman, 1994), leads strong sustainability advocates to adopt the precautionary principle. Conservation of natural capital and the application of a safe-minimum standards (Bishop, 1993) approach are therefore important components of a strong sustainability strategy. This message is that environmental degradation and loss of natural resources represent one of the main ways in which today's generation is creating uncompensated future costs. Hence restoration and conservation of natural resources and the environment is crucial to achieving sustainable development.
58 A number of sustainability rules (which fall some way short of a blueprint) for the sustainable utilisation of the natural capital stock can be outlined:
II)
III)
IV) V)
Market and policy intervention failures related to resource pricing and property rights should be corrected. The regenerative capacity of renewable natural capital should be maintained, i.e. harvesting rates should not exceed regeneration rates; and cumulative pollution Which could threaten waste assimilation capacities and life-support systems should be wherever feasible avoided. Technological changes should be steered via an indicative planning system such that switches from non-renewable natural capital to renewable natural capital are fostered; and efficiency-increasing technical progress should dominate throughput-increasing technology. Resources should, wherever possible, be exploited, but at a rate equal to the creation of substitutes (including recycling). The overall scale of economic activity must be limited so that it remains within the carrying 'capacity of the remaining natural capital. Given the uncertainties present, a precautionary approach should be adopted with a built-in safety margin.
Figure 2 summarises some of the measures and enabling policy instruments that would be involved in any application of a very weak sustainability (VWS) through to a very strong sustainability (VSS) strategy (Turner, 1993). From our review of sustainability, the emphasis on equity and social issues in sustainability as well as on the physical constraints is important. For development to be sustainable it must incorporate (under the strong sustainability view) non-depletion of natural capital; both intergenerational and intragenerational equity principles; and in the latter context must be capable of providing sustainable livelihoods to those whose livelihoods are primarily natural resource dependent. Agenda 21 sets out principles for sustainable development without advocating any explicit definition of sustainability and with a tendency for focusing on global issues which may not be of greatest concern to those poorest sections of the world. The implicit definition of sustainability within Agenda 21 however would seem to be closely related to the concept of strong sustainability discussed above, though the lack of operational details and the prevailing obstacles to change mean that implementation of such an agenda represents a very formidable task. In the context of climate change the sustainability concept would favour the adoption of a general response strategy that was based on the following ethical arguments: there is an obligation to avoid harm to future generations, either in an absolute sense, or so long as the avoidance measures themselves do not impose unacceptable cost on society;
59
Figure 2 Sustainability Mode (overlapping categories)
VWS
WS
SS
VSS
Sustainability Practice Management Strategy (as applied to projects policy or course of action)
Conventional Cost-Benefit Approach: Correction of market and intervention failures via efficiency pricing; potential Pareto criterion (hypothetical compensation); consumer sovereignty; infinite substitution Modified Cost-Benefit Approach: extended application of monetary valuation methods; actual compensation, shadow projects etc; systems approach, "weak' version of safe-minimum standard F i x e d S t a n d a r d s Approach: Precautionary Principle, recognition of the full value of natural capital; constant natural capital rule; 'strong' version of safe minimum standard A b a n d o n m e n t of Cost-Benefit Analysis: or severely constrained cost-effectiveness analysis; bioethics (i.e. an acceptance of the rights and interests of non-human species which then constraints human activity on moral grounds, e.g. the loss of tropical forests is in some circumstances morally wrong)
Policy Instruments (most favoured)
Pollution Raw Control and Materials Waste Policy Management e.g. pollution taxes, elimination imposition of property rights
Conservation and Amenity Management of subsidies,
e.g. pollution taxes, permits, deposit-refunds; ambient targets
e.g. Ambient standards; conservation zoning; process technology-based effluent standards; permits; severance taxes (i.e. taxes on resource extraction); assurance bonds (a sort of market-based insurance fund to mitigate environmental damage impacts) standards and regulations; birth licences
Source" Turner (1993)
the avoidance of harm to future generations is important because the future has no power to influence decisions taken now which may cause them harm; and current generations have moral obligations to future generations (either via overlapping generations or via the acceptance of interests/rights for future people). Climate change may strain an economy's capacity to achieve sustainable development by imposing unpredictable and significant damage, damage mitigation and adaptation costs. Resource investment and development planning may then be badly disrupted, pushing the sustainability goal further into the future. Developing economies will be faced with disproportionality severe dislocation costs because of their 'vulnerable' socio-economic systems and supporting ecological systems. Climate change is only one component of global environmental change (i.e. a complex flux of factors - population growth, increasing urbanisation, increasing industrialisation and intensification of agriculture, increasing rate of economic growth and international economic interdependency, the globalisation of information transfer and communications
60 and an increasing rate of attitudinal and lifestyle changes - the impacts of which can manifest themselves at a number of different spatial and temporal scales). Many developing countries, and to a lesser extent some regions (e.g. coastal zones) of developed countries, are already under heavy environmental pressure and potential climate change impacts on, in particular, agricultural sectors and coastal zone resources will further exacerbate their developmental problems. Climate risk is therefore very much an equity issue because the cost of riskbearing is not evenly distributed across societies, and more significantly across countries. Developing countries wiI1 face an especially high risk-bearing cost burden. We now turn to a closer examination of climate change risk and its implications for sustainable development.
Climate Change Risk Climate risk and other Global Environmental Change risks are shrouded by strong uncertainty and this is especially so at the regional level. The global scope of the potential changes means that there is collective risk which affects very large numbers of people. These risks are not statistically independent and the effectiveness of risk "pooling" is reduced. They are also endogenous risks in the sense that the global systems changes are being driven by human economic activity, because of its sheer 'scale' (Chichilnisky and Heal, 1993). The economy and the environment are jointly determined systems and the overall scale of economic activity is now very significant. Climate change impacts are potentially therefore part of a wider set of impacts and consequences. There is also a degree of permanent unpredictability present because the dynamics of the jointly determined system (coevolutionary process) are characterised by discontinuous change around critical threshold values both for biotic and abiotic resources, and for ecosystems functions. The stability of the jointly determined economy-environment systems depends less on the stability of individual resources, than on the resilience of the system, i.e. the ability of the system to maintain its self-organisation in the face of stress and shock. Unfortunately even if the critical threshold values could be discovered, neither the transition time to a new system state, nor the form of the new system state could be predicted. It is now a matter of some debate in the context of climate states whether the most 'natural' behaviour to be expected is a gradual warming trend process, or an abrupt phase change, as one climate region gives way to a new one (either globally or regionally). The characteristics of climate and related risks and the pressure of strong uncertainty provide a compelling rationale for the deployment of strong sustainability/precautionary instruments in ecological economic systems. The existence of possible threshold effects involving irreversible loss of potential productivity, and the failure of markets to signal the nearness of such thresholds, both imply the need for instruments that maintain economic activity and its pollution and waste generation consequences within appropriate bounds. The economic perspective, in principle, suggests the following analytical sequence for GHGs abatement and mitigation of other global environmental change effects - a general acceptance and application of extended cost-benefit analysis; recognition and quantification wherever feasible of environmental
5! risks; deployment of the precautionary approach via safe minimum standards (subject to their social opportunity costs) in the presence of strong uncertainty; deployment of a portfolio of enabling policy instruments to meet the chosen GHGs concentrations abatement targets and other sustainability goals.
Climate Change Decision-Making Strategies Greenhouse gases (GHGs) induced climate change poses a multifaceted challenge which has to be addressed via a collective decision-making framework operating at both the national and international levels. The decision-making contexts are characterised by 'strong' uncertainty and irreversibility and therefore favour the adoption of more, rather than less, risk averse strategies. A priori, a strong sustainability approach would seem therefore to be appropriate since it recommends the avoidance of those options which may generate the worst outcomes ('unacceptable' cost burdens) and encompasses the precautionary principle. What is and what is not an acceptable cost, from the strong sustainability, perspective is only partly measured by reference to individuals' preferences (conventional 'economics approach). Individuals may not be well informed about climate risk and expert opinion is constrained by strong uncertainty. Further, human preferences may not fully capture intrinsic values in nature - see Figure 3. In the market place, a product's value is encapsulated in its market price which in turn is determined in part by consumers' willingness-to-pay. But environmental resources often have no price tag and information is lacking concerning their 'true' value and significance. Many of these environmental assets are also public goods and this is another characteristic that makes it difficult for markets to evolve in such assets. To make the comparisons of environmental and other costs and benefits, within cost-benefit analysis, economists have therefore to impute a value for non-market environmental assets. A range of valuation methods and techniques have been deployed in order to estimate the value of various components of the environment. Environmental economists have developed a terminology of valuation which distinguishes between individual (private) use value (direct and indirect use of the environment), option value, quasi-option value, bequest and existence (non-use) value. Debate continues over the precise boundaries between these different components of total economic value. The social value of environmental resources is then simply the aggregation of private values. However, ecological economic research findings indicate that the social value of environmental resources committed to some use may not be equivalent to the aggregate private value of the same resources in any given system, because of the following factors: o
The full complexity and coverage of the underpinning 'life support' functions of healthy evolving ecosystems is currently not precisely known in scientific terms. A number of indirect use values within systems therefore remain to be rediscovered and valued.
62
Figure 1.
3
A General
Value
Typology
Anthropocentric I n s t r u m e n t a l Value
This is equivalent to "Total economic v a l u e " = use + non-use value. = DUV + [direct use value]
IUV -+ [indirect use value]
OV [option value]
+
QOV + [quasi-option value
BV [bequest value]
+
EV [existence value]
The non-use category is bounded by the e x i s t e n c e v a l u e concept which has been the subject of much debate. Existence value may therefore encompass some or all of the following motivations: interpersonal altruism, resource conservation to ensure availability for others; vicarious use value linked to self-interested altruism and the "warm glow" effect of purchased moral satisfaction; intergenerational altruism ( b e q u e s t motivation and value), resource conservation to ensure availability for future generations; stewardship nature;
motivation, h u m a n responsibilities for resource conservation on behalf of all
" Q - a l t r u i s m " , motivation based on the belief t h a t non-human resources have rights and/or interests and as far as possible should be left undisturbed. If existence is defined to include stewardship and "Q-altruism" then it will overlap into the next value category outlined below 2.
Anthropocentric Intrinsic Value _
This value category is linked to "Weak anthropocentrism" in a subjectivist sense of the term value. It could be culturally dependent. The value attribution is to entities which have a "sake" or "goods of their own", and instrumentally use other parts of nature for their own intrinsic ends..." It remains an anthropocentrically related concept because it is still a h u m a n valuer that is ascribing intrinsic value to non-human nature ("Q-altruism"). 3.
Non-Anthropocentric I n s t r u m e n t a l Value
In this value category entities are assumed to have sakes or goods of their own independent of h u m a n interests. It also encompasses the good of collective entities, e.g. ecosystems, in a way t h a t is not irreducible to t h a t of its members. But this category may not demand moral considerability as far as h u m a n s are concerned. 4.
Non-Anthropocentric Intrinsic Values _
This value category is viewed in an objective value sense, i.e. "inherent worth" in nature, the value t h a t an object possesses independently of the valuation of valuers. It is a meta-ethical claim, and usually involves the search for constitute rules or t r u m p cards with which to constrain anthropocentric i n s t r u m e n t a l values and policy. Source" Adapted from Hargrove (1992)
53
.
o
.
Because the range of use and non-use value that can be instrumentally derived from an ecosystems is contingent on the prior existence of such a healthy and evolving system, there is in a philosophical sense a 'prior value' that could be ascribed to the system itself. Such a value may not, however, be measurable and may not be commensurate with the economic (secondary) values of the system. The continued functioning of a health ecosystem is more than the sum of its individual components. There is a sense in which the operating system yields or possessed 'glue' value, i.e. related to the structure and functioning properties of the system which holds everything together. A healthy ecosystem also contains a redundancy reserve, a pool of latent keystone species and processes which are required for system maintenance in the face of stress and shock.
The adoption of a systems perspective, the recognition of primary ecosystem value (in addition to secondary value related to components of the system) and the nature of much environmental risk, i.e. high cost, low probability risks, emphasise the need for policy instruments that safeguard the range of options to future generations. Such precautionary instruments ensure that irrespective of the actual outcome of current activity, the next generation is left with an equivalent resource endowment (allowing for some trading between different forms of capital-physical capital, human capital and natural capital) and opportunities for economic development. These are commonly identified as sustainability constraints, e.g. safe minimum standards. Uncertainty about system boundaries and the effects of scale and thresholds underline the value of a precautionary approach, and many sustainability instruments have the property that they are precautionary. Sustainability requires each generation to maintain the self-organising systems that provide the context for all human activity and therefore possess 'primary' value. This does not imply that all assets should be preserved. Rather it implies conservation of opportunity. Thus one criterion, for example, for decision-making under 'strong' uncertainty is the 'maximim' criterion (i.e. minimise the worst outcome strategy). This is also complementary to the Rawlsian equity criterion (i.e. maximise the conditions of the least well off). On the other hand, since the uncertainties are so great and potential mitigation costs are so high, a better strategy might be to 'wait and see' and not to adopt any extensive policy interventions. As scientific, economic and technological data cross some of the uncertainties may diminish and we will be better able to discern and quantify the GHGs abatement cost and damage cost functions. Policymaking could then be aided by the application of the costbenefit method and techniques. Such an approach would be more in line with the 'weak' sustainability perspective. For some commentators the 'wait and see/business-as-usual' stance is attractive because they argue that if the forecast of a gradual trend rate of temperature rise (+0.3k per decade) is accurate (Houghton et al., 1990), then the global temperature signal will be discernible sometime between 2010 and 2020.
54 Policymakers therefore ought to defer any significant GHGs emission abatement measures to the next generation, and take out 'partial cover' by encouraging insurance schemes applicable to individuals and nations (e.g. Alliance of Small Island States International Insurance Pool proposal). Critics have countered that the mere existence of strong uncertainty cannot justify a no policy response option. The potential climate-induced damage costs could be very high, and in any case are only one possible element in the aggregate global environmental change impacts, many of which have already put heavy pressure on ecosystem resilience and adaptation capacities: The futures' 'opportunities' set is therefore being threatened, especially as GHG impacts may be irreversible. Whether or not future generations do possess moral rights or interests, most people would support the view that the present's 'coefficient of concern' for the future is not zero. Finally, a range of policy options are either 'no regret' negative net cost options, or are moderate cost options because once implemented they carry with them secondary benefits in addition to avoided climatic change damages. 9 The decision-making context and process are complex because uncertainty and irreversibility characteristics are compounded by the existence of a range of conflicting decision criteria, e.g. economic efficiency, intragenerational/ intergenerational equity, sustainability and precaution. The process probably has to be both hierarchical and sequential. Taking the weak sustainability position, we might assume that given moderate rates of technical progress (1 to 2% per annum) and actual global warming adaptation costs of up to 3% of GNP, future generations will be substantially better off than the current generation. So if the future benefits outweigh the present abatement costs, the future should pay those costs. But recall the weak sustainability capital substitution axiom, which in this context assumes that atmospheric capital is substitutable by manmade capital. If no such extensive substitution is possible then delaying GHG abatement measures in favour of providing a capital bequest (and investing is knowledge) for future generations cannot be justified. Alternatively, taking the strong sustainability position, possible significant future damage costs and irreversible impacts suggest that the future may be made worse off than the present. The passing on of a 'net liabilities' bequest to the future is morally questionable. Therefore a global level GHGs concentrations/emissions abatement target and interregional allocations need to be exogeneously set, guided by the precautionary principle. Once the commitments have been made then the search should be for cost-effective enabling measures, which should be adopted sequentially - 'no regret' energy conservation and efficiency improvement measures first, followed by fuel switching and other measures requiring longer lead times and significant capital investments. W e a k Versus Strong S u s t a i n a b i l i t y P o l i c y O p t i o n P o r t f o l i o s In principle, the full set of climate change response measures include the following: (I)
science-based research to reduce climate change and impact uncertainties;
65
(II)
technological research focused on more cost-effective GHGs mitigation measures (energy conservation and efficiency measures etc);
(III)
reversal of policies which in the past have encouraged the inefficient use of resources and waste sinks (i.e. correction of market and intervention 'failures');
(IV)
joint implementation, technology transfer and other forms of international cooperation to limit climate change;
(V)
measures to reduce GHGs emissions and/or to increase sequestration of GHGs;
(VI)
measures to enhance adaptation capacities of both socio-economic and natural systems facing the consequences of climate change;
(vii) insurance schemes to hedge against climate risk and the costs of adaptation; (VIII) policy interventions directed at the main drivers of global environmental change (such as population growth) and its related pressures. The different sustainability perspectives encourage the adoption of different policy option portfolio configurations. A weak sustainability strategy would seek to promote a portfolio based on measures (I), (II), (III) and (VII). It would be a more reactive rather than a proactive strategy and would lay great stress on the ability of research and development to reduce uncertainty and promote efficient resource usage. Cost-benefit thinking and analysis would be used as an i m p o r t a n t aid to decision-making. A strong sustainability strategy would seek to implement a more proactive and a more comprehensive policy portfolio including all the measures (I) to (VIII) in the list above. It would lay stress on the need for a precautionary approach and would recognise an obligation to future generations, not to pass on net liabilities. It would further seek to incorporate climate change impacts within a more general recognition of the GEC 'scale' problem. In this context integrated resource m a n a g e m e n t strategies would be highlighted (e.g. integrated coastal zone management). Cost-benefit analysis would be constrained by the precautionary principle via a safe-minimum standards (SMS) approach. The latter would be applied in an absolute sense (regardless of costs) when and if 'critical' natural capital assets were identified as being under threat; and more generally in a relative sense depending on the social acceptability of the SMS's own cost implications. References
R. Bishop, Economic efficiency, sustainability and biodiversity. Ambio 22 (1993) 69-73.
55
10
G. Chichilnisky and G. Heal, Global environmental risks. Journal of Economic Perspectives 7 (1993) 65-86. R. Costanza et al., Ecosystem Health: New Goals for Environmental Management, Island Press, Washington, 1992. I-M. Gren, C. Folke, R.K. Turner and I. Bateman, Primary and secondary values of wetland ecosystems. Environmental and Resource Economics 4 (1994) 55-74. C. Hargrove, Weak anthropocentric intrinsic value. The Monist 75 (1992) 183-207. J. Hartwick, Substitution among exhaustible resources and intergenerational equity. Review of Economic Studies 45 (1978) 347-354. J.T. Houghton et al. (eds), Climate Change: The IPCC Scientific Assessment, Cambridge University Press, Cambridge, 1990. D.W. Pearce, E. Barbier and A. Markandya, Sustainable Economic Develpment, Edward Elgar, Aldershot, 1991. R.K. Turner (ed), Sustainable Environmental Economics and Management, Belhaven Press, London, 1993. World Commission on Environment and Development, Our Common Future, Oxford University Press, Oxford, 1987.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
67
Global Climate change" social and institutional options M.R. Redclii~
Global E n v i r o n m e n t a l C h a n g e P r o g r a m m a (ESRC), W y e College, Wye, K e n t T N 2 5 5AH, U n i t e d K i n g d o m
Introduction In exploring the available social and institutional options over global climate change we need to address some fundamental questions about individual behaviour, social responsibility and 'globalisation'. Before doing this, however, it might be useful to establish some points of entry.
(1) Responses to the condition of the environment We need to know more about how environmental problems are perceived. For example, in the case of climate change, policies to combat global warming, to be effective, require some understanding of the links between individual behaviour and climate (both atmospheric concentrations and emission levels). Much better public information and media attention are essential before people can assess their responsibility for what is happening and what they can do about it. There are a number of social mechanisms which enable us to distance ourselves from the full implications of our behaviour. These need to be looked at - how our 'underlying social commitments' help establish this distance - before behaviour can be changed.
(2) Responses to existing policies We also need to know more about public responses to existing policies, many of which are not viewed as 'environmental'. Societies are not homogenous. Some ecological benefits carry distributive costs. How do people become enrolled in more sustainable practices like recycling and companies in green accounting? Pressure points exist where public opinion and values are more amenable to change. We should not forget that societies are reflexive systems; unlike inanimate objects, what we do reflects what we understand. More work needs to be done on sustainability indicators, to enable us to place normative goals into an operational context. Before going further, however, we need to examine global environmental change itself.
68 Global Environmental Change
Although it is usually conceded that values play a large part in the way we approach the environment, particularly the environment on our doorstep, the same concession is rarely made for the global environment. Global environmental change is often identified with physical processes "out there", such as ozone depletion, biodiversity losses and, particularly, global warming. The global environmental agenda has, to some extent, been established by the natural sciences, working within a positivist tradition (Newby 1993). The reports of the Intergovernmental Panel on Climate Change (IPCC) are a case in point. The authority of the IPCC's deliberations stems, to some extent, from its "scientific" objectivity, which influenced people like the former British Prime Minister, Mrs Thatcher, in lending it their support (Boehmer-Christiansen 1993). This paper examines whether global environmental change is as free from value considerations as many people believe, or hope. It goes on to explore three clusters of issues which suggest that an alternative approach needs to be taken. Global environmental change can be understood in terms of three sets of issues, each of which forces us to examine our part in its construction: human relations with "Nature", the need to live with increased uncertainty, and the extent to which our management of the environment reflects essentially human, rather than environmental, concerns. Each of these issues influences not merely the way we understand environmental problems, but also the way in which we can act to change them. They are also represented in the three major policy initiatives to have developed out of the Earth Summit in Rio de Janeiro in 1992: the Framework Climate Convention; the Biodiversity Convention and the institutions responsible for establishing more sustainable practices at the international level (particularly the Commission for Sustainable Development and the Global Environment Facility)(Grubb 1993). In the final part of this paper their relevance to the global climate agenda is considered. Human Relations and "Nature"
The nineteenth century was a period in which the physical sciences saw spectacular progress, and most of the scientific disciplines assumed the identity they possess today. It was also a period, in Europe and North America, of enormous economic growth, and with economic progress came confidence. Looking back from the vantage point of the end of the twentieth century this belief in progress, and the confidence that went with it, are the hallmarks of modernism (Redclift 1993). Relatively rapid industrialisation, and the growth of towns, were "global" phenomena because they served to incorporate other economic systems, and other cultures. Globalisation in the latter part of the twentieth century has served to underline these links, changing the international economic division of labour, using technology and communications to provide global images, as well as markets, and seeking to preserve the exotic and unfamiliar ("the other") whether through tourism or environmental campaigning, as items of consumption (Featherstone 1990, King 1991).
69
During the late nineteenth century, and early twentieth century, the opposition between nature and culture, made room for the social sciences as autonomous disciplines, they grew up in the interstices between the ethical concerns of the humanities and the positivism of the "hard" sciences. The insistence that human cultures were distinctive brought into question both the 'external' environmental determinism of some of the new sciences, and the "naturalism" of others, which saw human behaviour as the outcome of "internal" biological forces, equally beyond our control (Benton and Redclift 1994). Both of the imperatives provided by nature, the external environment and the human biological condition, were found wanting. It is not an accident that many of the issues which proved (and still prove) difficult for the social sciences to confront, such as eugenics, racism and the measurement of intelligence, lie at the crossroads of biology and social conditioning. In this sense the nature/culture dichotomy was both the springboard for the social sciences' advance, and the irresoluble problem they confronted (Benton and Redclift 1994). Within the social sciences the discipline which benefited most from its identification with human purposes in the nineteenth and twentieth centuries was economics. Neo-Classical economics grew out of the increasing confidence of industrial capitalism with its own success, and its refinements were linked to the problems faced by twentieth century industrial economies (welfare economics, Keynesianism, development economics). Many of the issues which will confront us as we approach the twenty first century - the relationship between the production of goods and services and the satisfaction of our needs, as well as the social and environmental consequences of their production - elude mainstream economists. Many of the underlying assumptions which influenced economic reasoning, such as the effects of scarcity, now appear much less important than issues like the environmental consequences of economic behaviour, which played little part (Yearley 1991). In the view of many Neo-Classical economists the significance of scarcity could be grasped through concepts like the economic costs of resource acquisition. Pollution and the proliferation of so-called "externalities" can be seen as manifestations of profligacy, rather than scarcity, and our inability to manage its consequences. As our dependence on economic techniques increases, the need for more inclusive systems of thought appears more urgent. We are forced back, inevitably, to consider our relations with nature, from which resources derive. Our increasing knowledge of biological systems has not enabled us to utilise them sustainably, and this is due in some part to the divorce which was effected in the nineteenth century between our understanding of the laws of nature and those of "man". We are faced by an interesting paradox. On the one hand the degradation of nature has called into question some of the values which contributed to the Promethean successes of the past. The rights of non-human species, and the primary obligations which we have to nature, are now regarded as politically important, and not merely by Deep Ecologists. At the same time, many of those who espouse environmental concerns refuse to acknowledge that it is the way that human societies are organised, and structured, which determines environmental problems. What are the values generated from the management of the environment today? They clearly reflect the interface between society and nature, and the difficulty we experience in dealing with this interface. Environmental management itself suggests a mastery of nature, and an ability to control the environmental consequences of our behaviour. The growing
70 importance of scientific knowledge, and "rationality" as the coda for this knowledge, together with our institutionalised behaviour and social commitments, has served to increase the appeal of technical solutions to human-induced problems. To provide solutions to environmental problems, however, we need look no further than the human societies which produce them; something which we seldom do (Beck 1992).
Living With Uncertainty: the Importance of Time and Space Another consequence of the growing confidence of science has been the expectation of certainty. With the development of scientific techniques and methods the status of scientific prediction rose, and with it the status of scientists. Predicting environmental consequences has proved to be difficult, however, partly because of the complexity of environmental systems, and partly because of the unpredictability of human actions. Science is apparently successful in offering predictions which reduce uncertainty. However, science also collapses time and space, and increases the flow of knowledge and information available. This, in turn, tends to increase uncertainty, and to fuel speculation about the basis on which decisions have been taken. We need to give close attention to the factors have buttressed the claims of science to reduce uncertainty (Brown 1989). First, many environmental problems involve high levels of human anxiety, associated with risks to human health, which appear to increase with the expansion of our knowledge. Second, since environmental science is an essential part of the solution to environmental problems, it follows that improved regulation, and greater technical expertise in addressing environmental problems, also serve to demonstrate the shortcomings in the application of science (Read 1994). Global environmental problems, in particular, such as ozone depletion and global warming, are not only complex in terms of their chemistry or biology, they are also apparently inaccessible to technical "fixes". Unlike the administration of antibiotics, or the inoculation of patients against the risk of contracting life-threatening diseases, changes in behaviour induced by environmental awareness, such as the purchase of aerosols free from CFCs, or the use of lead-free petrol, do not ensure environmental safety. We know more but we are able to do less. In addition, there is evidence from the growth of campaigning groups, around environmental issues, that the gap between "lay" perceptions of the environment and "expert" opinion, is actually widening (Yearley 1991). Faced with a barrage of increasingly complicated, and contradictory, information about environmental risks, the layperson is likely to question the authority of science, and the confidence politicians place in scientists. It soon becomes clear that the "critical thresholds" which are endorsed by political leaders and expert witnesses, are themselves political compromises, framed to manage public apprehension. The more that the official environmental discourse may seek to dampen public apprehensions, the more it becomes clear that "certainty" does not prevail. Public anxiety is only part of the picture. If the environment exists in the specialist knowledge that we possess about it, there is less for the "non-expert" to regard as their area of competence. This effects the "ownership" of environmental issues. Research from developing countries has shown that the growth of specialist knowledge is related to the
71
growth of non-specialist "ignorance", and this observation is equally appropriate in the North. Doubts about the degrees of certainty associated with formal scientific knowledge are matched by alternative, holistic models of human relations with nature, which interpret "facts" differently, and which seek new ways of understanding, rather than an enlarged databank of information. It is clear that different values are held by different groups of people. Some groups, at least, are using the opportunity presented by scientific uncertainty to re-evaluate their values (Thompson et al. 1986). The two dimensions of uncertainty which deserve particular attention are the spatial and the temporal. We are accustomed to make most decisions on the basis of present time, and any future consequences play a smaller part in our calculations than immediate consequences. However, environmental choices often bear little relationship to the decision-making and dislocation of everyday life. They require an imaginative leap into the future, to the next generation or subsequent generations. The timescale of ecological processes, particularly those operating at the global level, makes it imperative that we attach weight to the future, and that what economists call the rates of discount reflect this importance. Many environmental changes are also "systemic" in the sense that they can only really be understood through the way that systems change. Biodiversity is a case in point, since threats to individual species carry serious implications for ecosystems as a whole. The loss of one plant variety from a local ecosystem can jeopardise the survival of animal populations which are dependent upon it. Since the timescale of ecological processes bears so little relationship to everyday decision-making, it is important that we attach value to the loss of flexibility and variety in future environments. The spatial dimensions of the environment are also important in any consideration of values. The environmental consequences of human activity are often experienced at several removes, not only in time but in space. The economic development of the industrialised countries, their diets and lifestyles, have been responsible for transforming the environments of developing countries located thousands of miles away. The "ecological footprints" left by industrialisation, and consumer wants in the North, are not easily erased. This serves as a reminder that while in the North we tend to regard the protection of nature as a fundamental ingredient of environmentalism, in the developing countries environmental issues often present themselves in terms of protection from nature. Perhaps we need to consider whether the driving forces behind global environmental change, including industrial growth and consumerism, increase the environmental security of people in the South or seriously threaten it? The values generated in our society carry implications for the environment that are only dimly perceived most of the time. Consumerism implies a commitment to aspiration, to "improve" one's lifestyle. A desire to own the fruits of technoscience is apparently within everyone's grasp. At the same time we are concerned should the environmental costs of progress arrive on our doorstep. The response of local communities to environmental problems - "Not In My Back Yard" (NIMBY) - is a product of contemporary lifestyles in the industrialised countries, every bit as much as concern about protecting the whale or tropical forests. The process through which we are removed from the consequences of our actions, sometimes called "distanciation", is illustrated in a number of ways. Among them is the way
72 the enhanced greenhouse effect, through its impact on climate, is likely to increase perturbations in weather conditions, especially in the tropics, with increased occurrence of freak storms, drought and sea-level rise. The measures necessary to avert these risks are not difficult to determine, but the political will to act confronts widespread apathy and indifference.
Economic Values and Environmental Management Neo-Classical economics developed through making a number of assumptions about the environment. Natural resources such as water, soil and clean air, were often depicted as "free goods", meaning that they were available freely; they did not involve a charge. However, it is clear that environmental "goods" are qualitatively different, in significant ways, from goods for which we do pay a charge. Clean water and air, unpolluted soils, are not available "freely" in nature once human beings have had a hand in economic development. Environmental economics has been forced to consider the costs of cleaning up the environment, and of conserving natural resources to ensure their supply (Winpenny 1991). Ecological economics is also concerned with wider questions which have eluded most economists since the nineteenth century. Attempts are being made to distinguish between "wants" and "needs", and between the way our needs are satisfied, for example through more consumer goods, and the needs themselves. The conditions under which goods and services are produced is a key question. At the same time the social and ecological consequences of their production is a concern to Green economists. Many argue that we should develop methodologies for arriving at "utilisation values", that is, the value of goods and services throughout their lifetime. Such values would include the cost of waste disposal, the benefits from reuse or recycling, and the pollution or resource degradation associated with the use of raw materials in their manufacture. Within environmental economics there are broadly three camps. The first camp argues that there is nothing to prevent us from placing economic value on the environment. Using prices and market instruments we can assign the real costs of environmental degradation. What is required is further refinement of methodologies such as contingent valuation, which enable us to approximate individual preferences for environmental goods and services. In the view of these economists the "logic" of economic rationality can be used to manage the "randomness" of nature (Pearce 1993). A second camp takes a very different view. They argue that we cannot place a value on the environment, like that for human-made goods. Natural capital, in their view, is qualitatively different from human-made capital, and should be treated as qualitatively different. Following Oscar Wilde's famous aphorism, we are in danger of knowing "..the price of everything and the value of nothing". In the view of radical ecologists the logic of nature cannot be geared to the randomness of the market. As human beings we are part of nature, and cannot subject nature to our laws as we are subjected to natural laws (Ekins and Max Neef 1992). Between these apparently irreconcilable positions are others which probably attract considerably more support than is immediately evident. Some institutional economists like Jacobs argue that we can, and should, develop economic methodologies which, in effect,
73 "value" nature (Jacobs 1991). However, we should also recognise that Neo-Classical economics is itself a social construction, and its development reflects the preoccupations of industrial capitalism. We can develop methodological tools which place more, or less, emphasis on the importance of market forces. If we wish we can propose guidelines, indicators for "sustainability planning", which allow radical shifts in economic policy and thinking. Unlike some radical political ecologists, people in this third camp, propose that we intervene and regulate the environment, essentially to meet human purposes rather than follow imperatives in nature itself. They also agree that we will all be the richer if we examine the underlying social commitments which govern our lives, the maintenance of our present "lifestyles" and patterns of consumption. However, unlike Deep Ecologists, for example, mainstream environmental economists believe changes in human behaviour can be induced through policy instruments and interventions. It is also important to distinguish between analytical positions like those found within environmental economics or the sociology of science, and the value commitments of a society. To some extent analytical positions can play the part of, or even displace, other systems of values. We have only to reflect on the central role which Neo-Liberal economics has attributed to the "choices" of individuals in the market-place, or what Huber has called "ecological modernisation", through which business has sought to incorporate environmental costs in its range of products and services (Mol and Spaargaren 1990). These are examples of the close relationship between the values of the wider society and those that govern environmental questions. It would be surprising if core values such as "individualism", "private property", "choice" and "independence", the political values which govern everyday actions and desires, were unrelated to the way in which we interact with our environment. However, it is much more difficult to specify the nature of this interaction, the variables at work, and the lines of causation. These positions themselves reflect a modernist discourse that still sees the human subject as universal and all knowing (Redclift 1993). They do not address the fallibility of human beings, most notably in our inability to reflect upon the increased knowledge we possess about the wider universe. If science is continually widening the frontiers of what we know, it is also revealing the extent of what we do not know. We are, in fact, seeking to interpret what we do not know in terms of what we know. At the very least this is a hazardous procedure. Global environmental agreements and human values
The international agreements which were signed at the Earth Summit in 1992 give expression to environmental values, many of them widely shared. At the same time these agreements, if they are to succeed in changing the way we manage our resources globally, require that we pay more than lip-service to the values we espouse. The institutional apparatus established at Rio de Janeiro, as much as the agreements themselves, provides evidence of the difficulty in providing a consensus for global environmental management (Thomas 1993).
74 It is clear that values are implicit in what we take for granted from natural systems, as well as what we propose to do to protect these systems. At the same time, the process of economic development enshrines a different set of values. The Brundtland Commission, which reported in 1987, sought to enlarge this debate, and to make our value preferences more explicit (Brundtland 1987). Unlike the reports of the I.P.C.C. it did not purport to be a value-free document, but freely admitted to political objectives, many of which were subsequently incorporated in Agenda Twenty One. The idea of "sustainable development" as a way of informing policy cannot be divorced from the attempt to integrate quite different systems of values. Much of the confusion accompanying the discussion of sustainable development, and the drawing up of international agreements, stems from the relationship between our values and our knowledge about global environmental problems. The scientific controversy accompanying global climate change, and the deliberations of the I.P.P.C., has suggested that increasing ou/" knowledge about future climate change, and its impacts, will enable us to adopt more appropriate values, emphasising long-run sustainability over short-run economic gain. However, the evidence for this assumption is weak. Rather, it might be asserted that until we address the environmental problems associated with our current values, there is little likelihood that we will be able to make much use of the knowledge which is accumulating about the global environment. As Tickell argues, "... our ignorance of species and ecosystems is profound, not only of present ecosystems and species, but of their future uses and services. It is an understatement to refer to this level of ignorance as mere uncertainty" (Tickell 1994, 4). The major provisions of the Framework Convention on Climate change mark an important watershed in international agreements to protect the environment. The Convention established the principle that action to start addressing the problems of climate change should not wait upon the full resolution of scientific uncertainties. It also asserts that developed countries should take the lead in introducing measures to reduce the threat of global warming. Finally, it endorses the idea that developed countries should compensate the developing countries for any additional costs that they might incur in taking measures under the Convention. Superficially, at least, the goal of sustainable development is one publicly espoused by most governments. Most of the governments in the North have signed, and in some cases ratified, agreements which endorse a set of principles and values that place global sustainability above vested interests and short-term economic advantage. However, at a more profound level there is little agreement about the "values" which need to inform sustainable development. The "natural services" provided by the environment are acknowledged, but the assumption that they will continue to be provided, is still made. Real environmental costs and benefits are scarcely acknowledged in the day-to-day economic management that determines their use. Similarly, rather than using the precautionary principle to help provide for more flexible responses to uncertainty, most policy is still formulated against unsustainable assumptions, about population, military expenditure and economic growth. Global inequalities, particularly between North and South, are part of the "taken-for-granted" assumptions behind international agreements in "non-environmental" areas such as the liberalisation of trade. Inequalities within developing countries, we are regularly told, are part of the price that such countries pay for the absence of "development". However, evidence that economic growth has particularly
75 adverse effects on the Newly Industrialising Countries' environments, should lead us to question whether successes in market economies really are a prerequisite for better environmental management in these countries. This paper has argued that the options available to us over global climate change need to be placed in their context; our societies. Environmental consciousness is indelibly linked to social and political unease; it does not spring from the physical 'environment' alone. It follows that measures to combat possible climate change need to be located within socially meaningful categories, and we need to develop a better understanding of the reasons people assume social responsibilities towards the environment in the first place.
REFERENCES Beck, U. (1992). Risk Society, Sage, London. Benton, T., and Redclift, M.R. (1994). 'Introduction' in Social Theory and The Global Environment, Routledge, London. Boehmer-Christiansen, S. (1993). 'Scientific consensus and climate change: the codification of a global research agenda'. Energy and Environment 4(4). Brown, J. (1989)(edited). Environmental Threats, Belhaven Press, London. Brundtland (1987). World Commission on Environment and Development, Our Common Future, Oxford University Press. Ekins, P., and Max-Neef, M. (1992)(edited). Real-Life Economics, Routledge, London. Featherstone, M. (1990)(edited). Global Culture, Sage, London. Grubb, M. (1993). The Earth Summit Agreements: a Guide and Assessment, Earthscan/Royal Institute of International Affairs, London. Jacobs, M. (1991). The Green Economy, Pluto Press, London. King, A. (1991)(edited). Culture, Globalization and the World System, Macmillan, London. Mol, A.P.J., and Spaargaren, G. (1990). 'Sociology, Environment and Modernity', Paper presented to International Sociological Association Conference, Madrid. Newby, H. (1993). 'Global Environmental Change and the Social Sciences: Retrospect and Prospect' Economic and Social Research Council, Swindon. Pearce, David (1993). Blueprint Three - measuring sustainable development, Earthscan, London. Read, P. (1994). Responding to Global Warming, Zed Books, London. Redclift, M.R. (1993), "Sustainable Development: Needs, Values, Rights', Environmental Values 2(1), Spring. Thomas, C. (1993)(edited). 'Rio: Unravelling the Consequences', Environmental Politics (Special Issue) 2(4), Winter. Thompson, M., Warburton, M., and Hatley, T. (1986). Uncertainty on a Himalayan Scale, Milton Ash Edition, London. Tickell, C. (1994). 'Socio-political perspectives on Biodiversity', Green College, Oxford. Winpenny, J.T. (1991). Values for the Environment, HMSO, London. Yearley, S. (1991). The Green Case, Harper Collins, London.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
77
National and International Economic Instruments for Climate Change Policy J. B. Opschoor Department of Spatial Economics, Vrije Universiteit, De Boelelaan 1105, 1081HV Amsterdam, Netherlands
Abstract Policy instruments addressing sources and sinks of climate change can be applied at two levels: the national and the international level. The international focus on costeffective policies points at the need to evaluate the merits of "economic" or incentive based instruments in particular. An overview is given of experiences with such instruments at the national level and prospects for applying similar instruments internationally are explored. Although a tradable system of (net) emissions quota may have high potential benefits, its acceptance meets with relatively large problems. Carbon taxes or charges may be a better alternative. An important recommendation is that countries and regions could move ahead in locally optimal but partial ways, to be extended and harmonised after some more trial-and-error. Experimentation with joint implementation is a promising avenue for the long run.
1.
INTRODUCTION
Global environmental issues arise when specific forms of environmental degradation turn into a problem in which a large number of countries have a stake" they may be the results of activities in specific individual countries but often the welfare of other countries is adversely affected, potentially or factually. In the absence of a global environmental regulator international coordination through voluntary agreements is required. Such co-ordination may extend to the principles to be collectively applied, targets with respect to environmental quality to be achieved, individual countries' shares in these targets, and instruments to be applied. On the latter issue, a distinction can be made between instruments to be applied at the international level, and instruments to be applied within national jurisdictions. Below, we shall look at possible instruments at both levels and their relative merits and demerits. We will look at some more promising instruments at the international level in particular: a carbon/energy charge and tradable emission quota.
2. INTERNATIONAL INSTRUMENTS
2.1
ENVIRONMENTAL
POLICIES:
PRINCIPLES
AND
International Environmental Policies: General Principles
Transboundary environmental issues are the results of actions in individual countries that adversely affect the welfare of other countries. In the case of significant transboundary externalities, one possible approach is that some countries undertake
78 unilateral action (see Hoel 1991 for some caveats in this respect), but normally international coordination through agreements is more effective. This requires decisions by countries to become part of such agreements, which may imply a weighing of costs and benefits by the countries concerned. Individual countries might opt for not entering such agreements and to become free riders: then they would not share in the costs of measures whilst enjoying the benefits of the collective efforts of the other countries. In the case of unidirectional externalities such free riding behaviour cannot be countered by others polluting more in order to punish the free rider (OECD 1991a); hence other forms of establishing reciprocity would have to come in force. Even in cases of reciprocal externalities, countries might decide to not participate in a collective effort, if their valuation of environmental quality is relatively low. In such cases, incentives may be needed to make such countries engage in the agreement, such as side payments (compensations in one way or another). Also, efficiency arguments might be raised in favour of some countries financing environmental measures in other countries. Generally, therefore, some sort of international cooperation is needed if global issues such as climate change are to be addressed effectively and this may include a mechanism for burden sharing. The main perimeters of international environmental policy in the decades ahead, have been set by the principles as laid down in the Rio Declaration on Environment and Development (UN 1992)and specific international agreements such as the Framework Convention on Climate Change. The principles of the Rio Declaration most relevant to the development of national and international policy instruments, are: - t h e souvereign right of states to exploit their own resources and the responsibility to ensure that activities within their jurisdiction do not cause damage to the environment of other states or areas beyond the limits of national jurisdiction (Principle 2); -the duty to cooperate (with common, but differentiated responsibilities) to conserve, protect and restore the health and integrity of the Earth's ecosystems (Principle 7); - the application of "the precautionary approach": where serious or irreversible damage may occur, cost-effective measures to prevent environmental degradation must be taken (Principle 15); - the promotion of the internalisation of environmental costs and the use of economic instruments, taking into account the approach that the polluter should, in principle, bear the cost of pollution (Principle 16). The Framework Convention on Climate Change aims at stabilising concentrations of greenhouse gases in the atmosphere at levels that would prevent dangerous anthropogenic interference with the climate system are prevented, in a time frame that allows ecosystems to adapt and economic development to continue sustainably (Art. 2). Developed countries have to take the lead (cf. Principle 7) and proceed (cf. Principe 1 5 ) w i t h cost-effective measures and policies so that they generate global advantages at least cost (Article 3); efforts to abate climate change may be implemented jointly by parties to the treaty (ibid.).
2.2
E n v i r o n m e n t a l Policy I n s t r u m e n t s
So far, agreements
and conventions
have often
sought consensus
about
79 national efforts in terms of targets for emission reductions (e.g. the SO2-"clubs", CFCs). Given such agreed targets it would be a matter for national policies to decide how these targets would translate into national policies and instruments. However, from an economic perspective such agreed and fixed national targets might be a costly way of achieving overall environmental quality (Hoeller et al 1992), especially when (marginal) abatement costs differ between countries or regions. We thus have two levels at which climate change policy instruments should be considered: the national and the international. These shall be discussed in Sections 3 and 4; here we proceed with a general introduction to environmental policy instruments. Environmental policy can make use of two basic strategies (Fig. 1, routes a and b). Firstly, public projects and programmes can be set up that aim at preventing, compensating and eliminating environmental degradation or at providing substitutes for traditional behavioural patterns, such as: collective treatment facilities, environmental sanitation and (re)construction programmes, afforestation, etc.; the costs of such programmes would have to be borne by the relevant agents (governments, firms, individuals) according to some mechanism for burden sharing (e.g., from public funds, through levies and charges, etc.). Secondly, the decision making process may be influenced at the micro level through policy intervention. The second strategy is discussed in more detail below. Rational decision makers will base their decisions about their activities on a comparison of the various options open to them. They will compare the costs and benefits of these options, defined as all (dis)advantages relevant to the decision maker as aggregated by his/her individual weighing system. In such a situation, decisions can basically be influenced in three different ways (Fig. 1, left hand side): 1) alteration of the set of options open to agents; 2) alteration of the cost and/or benefits relevant to agents; 3) alteration of the priorities and significance agents attach to environmental change (i.e. changing the structure of agents' costs and benefits). Route 1) is referred to as direct regulation, defined as: institutional measures aimed at directly influencing the environmental performance of polluters by regulating processes or products used, by abandoning or limiting the discharge of certain pollutants, and/or by restricting activities to certain times, areas etc. The polluter is left no choice: he has to comply, or face penalties in judicial and administrative procedures. Route 2)entails economic incentives or market stimuli. The motivation relied upon here is that if environmentally more appropriate behaviour is made more rewarding in the eyes of the agent involved, then attitudes and behaviour will 'automatically' shift in favour of these socially more desirable alternatives. Options can be made more or less (financially or economically) attractive by applying charges or levies, granting subsidies, implementing tax differentiation etc. Such instruments will be referred to below as economic instruments. In this way environmental concerns can in a certain sense be 'internalised' by altering the agent's context rather than the agent's value structure or preferences. Route 3) includes approaches such as: education, information extension, training, but also: social pressure, negotiation and other forms of 'moral suasion' leading to a change of perceptions and priorities within the agent's decision framework. It aims at full 'internalisation' within the preference structure of the agent.
80 In this paper we will concentrate on economic instruments, i.e. route 2. Basically, one may distinguish the following categories of economic instruments: 1) charges, 2) subsidies, 3) deposit-refund systems, 4) market creation, and 5) financial enforcement incentives (OECD 1989). Charges may, to some extent, be considered as a "price" to be paid for pollution. Polluters have to pay for their implicit claim on environmental "services", which thereby enters at least in some part into private cost-benefit calculations. There are various types of charges, including charges on emissions (e.g. CO2), and charges on products (e.g. fossil fuels, CFCs, cars). In deposit-refund systems a surcharge is laid on the price of potentially polluting products. When pollution is avoided by returning these products or its residuals to a collection system, a refund of the surcharge follows. Markets can be created where actors might buy "rights" for actual or potential pollution or where they can sell "pollution rights" or their process residuals (for reuse or recycling). In emissions trading, dischargers operate under some multi-source emission limit and trade is allowed in permits adding up to that limit. Given certain quota under agreed emission reductions, also countries could trade. Criteria for selecting specific instruments relate to (Opschoor and Turner 1994; OECD 1994): (environmental) effectiveness, economic efficiency, and (social and political) acceptability. Amongst the acceptability criteria distributional considerations are especially important, particularly in the context of global environmental issues. Economists have argued that economic instruments are to be preferred to especially direct regulation, as they tend to invoke least cost technical and economic responses; moreover, it is believed by many of them to also generate a more effective incentive for technocal innovation. Using simple models, it can be shown that a pro rata agreed emissions reduction approach is less efficient in economic terms, than applying either charges or trade in permits to achieve the same reduction (e.g. Barrett in OECD 1991b) - but this is theory: it ignores much of the lack of data and knowlegde on effectiveness and efficiency, and overlooks the acceptability problem and all that it stands for (Opschoor and Turner 1994). Environmental policy instruments typically come in "cocktails" or "mixes" of the pure elements described above. Thus, a carbon trading system is a combination of a regulatory measure (i.e. the total volume accepted) and an economc one (trading quota within that maximum). Economic instruments do have a part to play in tackling global environmental issues. The wider the geographical extent within which they are applied, the larger the efficiency and flexibility-benefits of applying these instruments are likely to be. However, it is also fair to observe that economic instruments meet with a number of problems. In fact, one can discern a set of different dilemmas: Firstly, the 'administrator's'-dilemma: direct regulation may be inefficient but yet efficient economic alternatives may be inacceptable to the policy makers for cultural or political reasons. Charges and trading schemes will then not be used where this could be practical and soacially advantageous. Secondly, the 'second best' -dilemma : a first best approach making use of economic instruments may, in a distorted world with government and market failures, create more inefficiency and be less effective than introducing second best instruments such as direct and uniform regulations. Thirdly, the 'revenue'-dilemma: returning the revenues of charges and levies to
81 the sector that they came from (e.g. for subsidising CO2-reduction by fuel shifts) enhances acceptability and effectiveness but is socially inefficient. This often results in discussions between public finance and fiscal experts on the one hand, and environmental managers on the other. Finally, there is the 'leverage point'-dilemma: policy instruments may affect economic agents and processes in different stages of the product life cycle. Subtle and tailor-made emissions charges may be optimally effective without loss of environmental potentials for economic use, but at the same time they may be administratively costly. However, administratively easier but more clumsy inputs charges (e.g. on energy or Chlorine) may be environmentally and economically inefficient.
2.3
Climate Change and Climate Change Policy Instruments Climate change, CC-policies and Cooperation
The overall objective of the Climate Convention has been given in Par. 2.1. It is assumed here that under a "Business as Ususal"-scenario of global economic development, GHG-concentrations would rise to levels beyond those aimed at within the Climate Convention. This then would imply levels of damage to ecosystems and to development that are to be avoided at least partially. One of the principles of the Convention is to deal with climate change in a cost-effective way; that is, to achieve objectives at least social costs. The costs of climate change include avoidance or abatement costs (cost of emissions reduction and sink enhancement) and accommodation costs (costs of coping with residual GHC-concentration rise). Both abatement and accommodation costs include direct costs (i.e. the cost of measures and of policies aimed at reducing climate change or coping with it) and indirect costs (the net socio-economic effects elsewhere in society as a spinoff of these measures and policies). One may wonder whether the level of admissible GHG-concentration could not also be determined in a least-cost approach, i.e. by minimising the sum of abatement and accommodation costs associated with alternative levels of abatement (this is one of the hot issues in the IPCC Working Group III at the time of this Conference). This would imply that all relevant elements of accommodation costs can be measured with a reasonable degree of accuracy - including damage costs due to exposure to concentrations- over a very long time horizon. However, according to many researchers (including the present author) the uncertainties inherent in such calculations justify a more political approach, where the overall net emission levels are to be below some negotiated maximum path, satisfying conditions such as those of Article 2 of the Convention. Within such an overall maximum, various different patterns of national net emissions (i.e. emissions corrected for sink enhancement) could be accepted. This allows for negotiation over, or trade in emissions quota, etc., to achieve cost-effectiveness.
Climate Policy-related Economic Instruments In terms of environmental impact, there are several ways in which environmental policies could operate technically in order to become or remain compatible with some environmental target or standard. One is, to move or relocate sources (i.e.
82 activites) to areas where they will contribute less to the global issue at stake. In cases such as the climate issue, relocation does not provide a real solution. Environmental degradation can also be avoided by reducing source strengths through reduction of activity levels, diffusion of existing cleaner technology and/or innovation. A final strategy is that of circumventing the environmental impact by enhancing the environment's capacity to absorb or otherwise deal with the pollution. Sink enhancement e.g. through afforestation is one relatively inexpensive method of dealing with notably carbon dioxide emissions -the major greenhouse forcing substance. In order to trigger economic agents responsible for GHG-sources and sinks to move towards either one of these technical options, they will have to be commanded or convinced, and especially in the global context, economic approaches to this may warrant attention. Prospective modelling work suggests that the impacts of instruments (charges and permits) on revenues and income transfers between countries can be very important (Hoeller et al 1992). In the case of (transboundary and) global environmental problems a special issue arises out of the combination of the criteria of effectiveness and efficiency. It may well be more effective to allocate a certain amount of money to financing environmental activities elsewhere ("joint implementation"). Several types of economic instruments for addressing global environmental issues have been proposed: i) Emissions charges or products charges (e.g. taxes on energy use) or combinations such as an international carbon/energy charge; ii) Global permitting systems for emissions or for using a global environmental resource, allowing for trade betweeen countries in such permits or in emissions offsets (cf. the Montreal Protocol); iii) Sanctions against free riding or non-compliance in relation to environmental treaties, including trade sanctions; iv) Motivation to participate in agreements by compensation payments and by socalled "joint implementation' programmes; v) Deposit-refund systems. Below, we shall first look at national policy instruments (section 3) and then move on to options for climate policy instruments at the international level (section 4).
3.
REVIEW OF NATIONAL POLICY INSTRUMENTS Currently, most countries operate a range of environmental policy instruments including economic instruments. In this section climate change related economic instruments as currently in use or under consideration in OECD countries, will be briefly reviewed (after OECD 1994): charges and trading schemes. 3.1
Charges Charges on processes and products that generate pollutants contributing to global environmental problems could make these inherently less attractive. Several existing charges on products relate to global environmental problems (e.g. CFCs, fuelrelated CO2). Product charges have several advantages. First, the administrative framework for collecting the charge may be relatively easy to conceive, or may -in some cases, already function in some parts of the world: in most countries and for many fuels, systems already exist for taxing them. Second, changing the level of a charge or
83 tax is a relatively simple intervention, compared with altering other instruments (such as tradeable permits). In the area of charges on emissions most air pollution charges relate to non-GHG emissions, especially of acidifying substances (US, Canada, France, Japan, Scandinavia, Portugal). One succesful example is the Swedish NO x-charge on heat and power producers: the charge has speeded up compliance to sharper emission standards to be imposed in 1995. The accelerating mechanism was the rechanneling of the charge's revenues to the producers according to their final energy production; thus, heavy emitters subsidised clean energy producers and this provided an incentive for rapid innovation. More than half of the OECD countries have differentiation in car sales tax rates or annual vehicle tax, according to the levels of emissions (based on e.g. car weight, catalytic converter, emission standards compliance). Several countries have explicit and implicit carbon taxes with an intended incentive impact. Carbon taxes are now applied in the Scandinavian countries (including Denmark), Italy and the Netherlands, but only in the Norwegian and Swedish case are they significant enough to have an incentive impact. All OECD countries have energy taxes, though at different levels and using different operational systems. Some countries have or are considering effective energy/carbon charges, and others look at raising energy tax levels. Charges on Ozone depleting chemicals exist in Australia, Austria, Denmark and USA, possibly with incentive impacts in the latter two countries.
Domestic Policy Impacts National charges will have nationally relevant policy impacts, notably in the fields of income distribution, sectoral activity levels, public finance, macroeconomic policies (Piacentino 1994). Energy taxes are applied to a commodity with fairly low short term price elasticities. Hence, in order to achieve a given quantity impact on its consumption, the price rises must be high. In order to achieve a levelling off of CO2 emissions in 2020 at the 1990 level, calculated emission taxes vary between $30 to $150 per ton of carbon. This implies high tax revenues and raises the question as to the distributional impacts of such a taxation policy. There is a case to be made for using these revenues as part of the overall public finance, which, in the case of an assumed fiscal neutrality would imply that other taxes could be reduced accordingly. In addition to the fiscal aspect, the income effects of the charge are to be considered and perhaps compensated. Emissions reduction efforts may induce significant feedbacks to the economic process. Research has been done to explore the economic costs of e.g. reducing carbon emissions in various countries. Achieving large reductions of energy-related CO2-emissions may depress growth rates of world GDP by .2 percentage points, but this can nevertheless imply reductions in the long run levels of global GDP of some 3-8% (between 2025 and 2050); some national models (e.g. for Norway, Netherlands, Sweden) predict higher growth rate depressions with higher GDP-impacts earlier on (2000-2010) at much lower levels of reduction (Hoeller et al 1992). Moreover, charges could lead to shifts in the industrial structure. In one country's exploration of the economic impacts of unilateral and joint action via carbon/energy charges (without exemptions), national (and even some collective)
84 sectoral implications could be dramatic (CPB 1992); in the very energy-intensive sectors there might be relocation of industries to other countries or regions. Calculated reductions in energy use might result from reduced or replaced activity, rather than from fuel shifts, energy conservation or new (leaner) technology (ibid.). Unless agreements could be reached with all important trading partners, countries will not easily impose extra costs on their industries through high energy or carbon taxes. To protect their international competitive positions, countries usually consider tax schemes only if associated with substantive exemptions for domestic energy-intensive industries operating on international markets. In the context of regionally applied charges, one alternative to such exemptions could be: mitigating measures on transboundary transactions correcting for cost differentials due to nonparticipation in the charge scheme. The proposals for an EC CO2/energy tax appears to favour an approach based on such exemptions. Providing exemptions and mitigation may easily lead to different consequences in terms of internal and external support for policies aimed at global environmental issues and little is known of these and other indirect effects of such corrective measures.
3.2
Other Policy Instruments
Apart from charges (or their less efficient and effective reverse: subsidies), countries may use instruments such as trading schemes and of deposit-refund systems. On national trading schemes, there is very little experience outside the US and it has been reviewed elsewhere (OECD 1994). Tradable permit systems exist in the US, Canada, Australia and Germany; the Canada and US schemes include trading in CFC-quota, air pollution emissions, and some car and fuel-related emissions. There are no direct climate change related trading schemes. An extensive literature exists on the efficiency of trading and its prospects for air pollution abatement and prevention (see, e.g. NAPA 1994; Klaassen and Pearce 1994). In theory emissions trading is as efficient as charges, and there may be less uncertainty as to its environmental effectiveness than would be the case with charges. We shall come back to this below, in Section 4 on international instruments. A second approach could be that of deposit-refund systems in relation to greenhouse gases. Deposit-refund systems would imply putting a charge or tax on bringing a unit of e.g. CO2 into the atmosphere whilst reimbursing for removal, disposal or sequestering of a unit of CO2. CO2 removal and fixation might enable economies to seek least cost options for GHG-reductions by comparing the costs of combustion reductions with those of fixation, etc. Compensiations for CO2-fixation could thus be combined with a CO2 charge into a deposit-refund analogue. This could be done both nationally (e.g. when fixation in new forests is financially facilitated) and internationally (see e.g. Huppes et al. 1993).
4.
CLIMATE POLICY INSTRUMENTS: INTERNATIONAL OPTIONS
As in the case of national policy instrumems, international instrumems may aim at an incentive effect or they could be intended to raise financial resources to
85
undertake other activities including emission abatement. In fact, the revenue raising character of some of these instruments might appear as an advantage given the large amounts of financial resources needed to address adequately the global problems and the distributional aspects that would be encountered in attempts to obtain global commitments. One argument against this is that if prices are correctly reflecting environmental costs in consequence of a "proper" tax level, then earmarking would lead to a distortion of the optimal allocation.
4.1
Emission charges
Especially in the case of charges there are grounds for engaging in an international dialogue on account of possible repercussions on international trade and investment. In relation to global warming, taxation may be based on energy content or on Carbon content (or some combination of these two). All OECD countries have energy taxes, though at different levels. Some countries have or are considering effective energy/carbon charges, and others look at raising energy tax levels. Carbon taxes are now applied in the Scandinavian countries (including Denmark) and the Netherlands. The European Community is considering a carbon/energy tax on energy raising primary energy prices initially with $3, going up to $10 per barrel of oil equivalent some 7 years after the scheme becomes operational. This might result in a reduction of CO2 emissions in 2000 in the order of 6-7% (DRI as quoted by Carraro and Siniscalco 1993). The charges/tax option is an appropriate one given the wish to improve market signals and to raise public awareness. Due to the elasticities involved however, they may not be the most effective way to modify behaviour in the short run. Macro economic impacts or the fear for such impacts, might make individual countries reluctant to move ahead of others in introducing such charges, or in making them high enough to have an incentive impact (see par. 3.2). Especially in the case of charges there are very solid grounds for engaging in an international dialogue on them, explicitly taking into account impacts on international trade and investment, on relocation etc., as resulting from a substantial tax.
4.2
Tradeable Permits Systems (TPS)
The main idea of tradeable permits is achieve an environmental target at least cost to society, by setting an emissions reduction target, distributing or auctioning permits up to the total set by the reduction target, and by allowing trade in these permits. Compared with charges, permit trading has the advantage of a more certain result in terms of emissions reductions: the amount of permits issued sees to that, if it is enforced. These permits can be subjected to market forces: there can be more of them for sale if further technological innovation results in cleaner technologies; demand -if mobilised- will competetively force prices down to their appropriate level. There must be an information system (with information about potential buyers and sellers) and an auction procedure. Also, there must be some agreed initial endowment and this is one of the main difficulties with a TPS. In addition, there are operational conditions to be met e.g. on the definition of the market, the 'size' of the market in terms of numbers of buyers and sellers, of real possibilities for trade in terms of actual or potential cost differentials, etc. (see e.g., UNCTAD 1992). Several of these issues will be briefly reviewed.
86 The commodity in which trading is to occur, has to be defined clearly. For obvious economic reasons the flexibility-and hence the potential cost-effectiveness- of the scheme would be enhanced if apart from emissions also sink-enhancement strategies could be incorporated, and if the whole range of GHgasses (i.e., not only CO2) could be considered. Even in the case of greenhouse forcing alone, there are several options (e.g. fossil CO2 emissions, ibid. plus C-sequestration by plantation, net CO2-emission, equivalent CO2 including other gases etc.). The choice of definition is likely to reflect the performance to be expected on a number of criteria, such as compatibility with sustainable development, efficiency, etc. There seems to be a growing consensus that limiting trade to emissions of energy-related carbon dioxide is the most feasible initial option (Swart, in OECD1992a), with that of including other CO2sources and sinks as a very promising alternative especially as it may do more justice to the claims of developing countries (Agarwal and Narain 1991). With limited numbers of market actors there is the risk of parties being capable of manipulating permit prices, or to influence prices in related commodity markets (Tietenberg, OECD 1992a) in order to affect the distribution of rents. According to Roland (in OECD 1992a), this risk is relatively low (see also Bohm 1991). Yet, there is another distributional issue related to market power: with unequal purchasing power and in a buyers' market, the possession of permits might concentrate in the portfolios of a few rich nations. One solution to the problem of market power confusing the performance of trading schemes would be to limit the period during which rights to emission remain valid: under such circumstances hoarding and accumulating emission rights would be much less economically rewarding (Boorsma et al 1988, OECD 1994). The question of how to distribute initial rights is a crucial one in obtaining international support for any large scale trading system. This initial endowment or distribution could be based on e.g. current emissions level, past responsibility, equality of effort, GDP, population, etc. The initial allocation will have to be a compromise, based e,g. on both a per capita allocation and current emission levels.
4.3
Deposit Refund Systems
Countries, apart from being sources of substances or other interventions giving rise to global environmental problems, may also engage in activities that enlarge the environment's capacity to absorb (or otherwise handle) human activities. In the case of climate change, sink enhancement e.g. through agricultural and reafforestation policies is a case in point. Economic instruments may support such activities. In the case of charges: if only net emissions are charged, or emissions corrected for the annual impact in terms of sink enhancement, then the latter would be economically attractive; and so would, of course, innovation. A similar argument holds for allowing trade in credits built up by sink enhancement or technological innovation. However, the issue of the measurement of the contribution to resolving global environmental problems through sink-enhancement is such an intricate one, that it might preclude advancement in international environmental agreements, unless put aside until better monitoring is possible. In principle, however, if countries engage in activities that enhance their environments' capacities to absorb or buffer global pollutants, then they could be compensated for that within a trading system if such activities would yield additional permits or credits to them that could subsequently be offered on the
87 permits market. Future extensons of the Climate convention might include provisions for compensating countries for carbon removal and fixation, presumably certified, or by allowing these to be credited to in the form of equivalent additions to their allowed emissions.
4.4
Joint Implementation
A step towards a full permit trading situation, may be that of "joint implementation". Under a regime of joint implementation, countries might find it in their interest to participate in an agreement and to take on emissions reductions responsibilities, whilst the financial consequences would be shared with or adopted by other participants. From Western Europe, acid deposition offsets are being sought and financed in Central and Eastern Europe as an alternative to carry out costly abatement programmes in e.g. the Netherlands (1993). The Climate Convention allows certain parties (notably OECD countries and economies in transition) to implement jointly with other parties (including developing countries) to the Convention, of activities to reduce GHG-emissions and enhance sinks ("joint implementation", Article 4.2.a). For countries that have accepted emissions targets, joint implementation would effectively allow more efficient emissions reduction within the total target. If joint implementation is undertaken with countries that are not (yet) committed to any target, then the effectiveness becomes more dubious, whilst no doubt average (and most likely total) emission abatement costs would be lower. From the side of the developing countries a risk of joint implementation would be, that it might be a disincentive to developed countries to change their patterns of production and consumption. Special attention should be given to the conditions under which such schemes are appropriate (see e.g. Kuik et al 1994). Criteria to be applied include: (i) additionality of net emissions reduction, (ii)certified environmental effectiveness, (iii) complementarity to reduction of own emissions. In a way joint implementaton schemes can be regarded as justified by the fact that often the countries providing these compensations are those that now and in the past have been largely responsible for the present state of the environment; they could be regarded as paying their 'environmental debt'. In fact, joint implementation enables one category of polluters to engage in environmental policies when its income constraint is an over-severe impediment. Such schemes, while avoiding some of the problems posed by straightforward trading schemes, may lead to other difficulties: i) In as far as co-operation of developing countries is required, these might show reluctance in going along with programmes that apparently imply financial transfers with "new conditionalities" (the proviso that they be spent on specific activities targeted to deal with specific reductions in emissions) involved. ii) Difficulties in establishing the environmental effectiveness are likely, as this involves difficult assessments of time paths of emissions with and without the programmes, against the background of often relatively impredictable developments at the level of the underlying economic activity levels. Certification arrangements may at least partially address this issue. iii) Some of these schemes entail the exchange of property rights on natural resources in other countries; it could be difficult to disentangle environmental effectiveness from considerations in terms of expected capital gains. A growing literature (e.g. Kuik et al 1994) is geared towards designing effective,
88
efficient and equitable arrangements for joint implementation. A tradable carbon quota system could easily arise out of a joint implementation situation, when the countries involved all develop targets for (net) carbon emissions.
5.
A TENTATIVE ASSESSMENT AND CONCLUSIONS
Climate change policies may lead to substantial social costs; attempts to identify least-cost approaches can therefore have high social benefits. Economic approaches to global issues are important to build into the emerging conventions and institutions aspects such as flexibility and efficiency. Several types of instrument may be considered. Recently most attention has been given to systems of charges and tradeable permit systems. These systems have different characteristics in terms of their environmental and economic performance (see e.g. IPCC 1994). But their most important common characteristic is that they will help in minimising overal emissions reduction costs by shifting effective emissions reductions to countries with lowest marginal costs; that is, if we are comparing trading in GHG-permits with GHGcharging (e.g. a CO2-tax) and not with an energy tax. Energy taxes would be relatively inefficient ways of achieving climate-relate objectives compared with GHG-emissions charges (Carraro and Siniscalco 1993; Zhang 994). Advantages of TPS over a charges system include (Bohm 1991): (i) the relative certainty of meeting emissions standards; (ii)fewer complications with non-convertible currencies than when handling charges' revenues; (iii) TPS does suffer from harmonisation problems with existing national taxes, as carbon/energy charges would. In addition, a tradeable quota regime would put developing countries in a position where transfers would be based on agreed upon rights, whereas a tax-cum-transfers system might keep developing countries in a situation of structural dependency on industrial countries. Finally, a tradable permit system would probably induce a forward market with associated intertemporal efficiency gains. Relative advantages of a tax system over TPS are (ibid.): (i) the revenue raising nature of the instrument; (ii)the familiarity of governments and other actors with the excise tax principle; (iii) low transaction costs; (iv) tax systems would not give rise to a compromising dominant position of large industrial countries. Furthermore, appropriate international institutional frameworks to operate a charges system are relatively easy to conceive. Both systems, that of tradeable permits and of charges, suffer from difficulties in obtaining wide support. Obviously, introducing an international system of charges will not be easy, given the differences in the levels at which energy and -implicitly carbon are charged in different countries. Trading schemes appear to, at best, pose a future option only: practical difficulties, especially related to establishing an acceptable initial endowment, provide impediments to their rapid introduction. A rapid and full-scale introduction of any one of these theoretical alternatives is therefore very unlikely. Rather, one should expect experimentation on smaller scales and with partial approaches, from which a broader and harmonised approach might develop later. Various ways can be envisaged: 1) Second-best compensatory introduction.
89 In the case of tradeable permits, side payments, an "equitable" initial endowment and temporary exemption of poorer countries could help in (gradually) introducing a worldwide system. In the case of taxes the revenue raised by it could be (partially) recycled on the basis of income effects of the tax, initial income differentials, efforts in developing or installing cleaner technology or augmenting pollution sinks, etc., so as to enhance international support. 2) A less than across-the-board-approach which focuses on certain main elements of the climate change problem, operates in groups of contributing countries or sources, with selected compounds, gradually building up from there towards a more complete system. For a number of reasons (OECD 1992b), it appears that a system of national taxes on energy or carbon could be achieved easiest at the regional level. This might hold particularly in the European region (some details on an EC carbon/energy charge were given above). Subsequent introduction of similar charges elsewhere could lead to harmonisation towards e.g. an OECD-wide charge. The larger the geographical scope of the charge, the less need there is to exempt industrial sources on the basis of distortions in long term comparative cost differentials and competitive positions, or to use other complicating additional measures. A system of trading in emissions permits or offsets would presumably only start with a group of developed economies as well, on the basis of internationally agreed emission targets. Such countries could be allowed to buy reductions elsewhere (offsets). It is important to design procedures that could facilitate the gradual development of a true market for entitlements out of bilateral transfers of such entitlements in the initial phases (Roland in OECD 1992a). 3) Mixed systems might arise with charges in some regions and emissions permit trading elsewhere. Other instruments of a mixed or hybrid nature may exist: (i) national or international funds drawn from the revenue of national charges for e.g. energy use and/or carbon emission could be set up to finance environmental expenditure in developing countries; (ii) tradable credits could be built up by abating emissions in other countries, based e.g. on unit rates to be decided in international agreements; etc. The partial and hybrid systems discussed here, may have institutional advantages in that they often build on well known procedures such as negotiations about reduction efforts. Conclusion
Looking at the economic efficiency (and disregarding a number of institutional and political impediments), fully fledged schemes of emissions charges or tradeable emission permits appear most (and equally) attractive. If their effectiveness were to be the criterion, tradeable permits hold the promise of more certainty in comparison with charges. However, such schemes are very unlikely to come about in the near future. In terms of environmental effectiveness a step-by-step approach starting with some system applied by a small number of countries, gradually increasing its geographical scale and incorporating more elements of charging and/or trading, with a gradual level of price rise in the case of a charge, appears to be the most promising and practicable avenue for an international climate change policy. Considering the acceptability of the various instruments, there are difficulties
90 Fig. 1. Environmental Policy Approaches
I control activities
facilities/ options i
I
i
I b. 1: regulation I" . . . . . . . . . . . . . .
[ I
!
l--
!
I
I
I
i 1
i
1
perceived options
decision
--"
behaviour
ul
I
I
1
L
environmental impacts
t * win.-
mmu
I
F I I I !
1I
i
I
',b.2: incentives
1
I
options costs and benefits
I
1
t
I_ . . . . .
i P 1
other impacts
I
I
I
',b.3: internaliI
9
zatlon ...,..
=.,=,.== . . . . .
!
other impacts of options
1 I
I I
[ I
1
. . . . . . .
I I b.3: information
approach b
approach a [
policy choices | I I
institutional context
~,fter: Opschoor 1991
j
.J
91 with the equity aspects (both initial endowments and issues related to accumulation) and the institutionalisation rendering tradeable permit schemes politically ill-acceptable compared with charges or taxes on energy. On-going discussions about a carbon/energy charge in Europe, perhaps to later on be introduced OECD-wide as a second step to a global scheme, are very important in this respect. "Joint implementation" of emissions reduction where some countries participate financially and technically in other countries' abatement or sink enhancement efforts, could be a way of developing towards an international trading system. However, given the current status of joint implementation it will not be a major instrument in achieving targets for the year 2000.
6.
REFERENCES
Agarwal A. and S. Narain (1991). Global Warming in an Unequal World: a Case of Environmental Colonialism. Centre for Science and Environment, New Delhi. Bohm, P. (1991). Policy Problems of International Environment. Paper presented at the International Environment in the '90s",Neuchatel, Aug. 1991.
Cooperation Conference:
on the Global "Economy and
Boorsma P., P.C. Gilhuis, B.M.S. van Praag and J.B. Opschoor (1988). An Anti Acidification Fund. (in Dutch). Ministry of Public Housing, Planning and Environmental Mnagement, Publication Series Air, Nr. 77, 55 pp. Carraro C. and D. Siniscalco (1993). The European Carbon Tax: An Economic Assessment. Kluwer Ac. Press Dordrecht. CPB (Central Planning Bureau) (1992).Long Term Economic Consequences Energy Charges (in Dutch). CPB Working Documents No. 43, The Hague._
of
Hoel M. (1991). "Global environmental problems: the effects of unilateral actions taken by one country". Journ. Envir. Econ.& Man., Vol. 20 Nr. 1, Jan. 1991: 55-71. Hoeller P.,A. Dean and M. Hayafuyi (1992). "New Issues, New Results: the OECD's Second Survey of the Macroeconomic Costs of Reducing CO2 Emissions". Working Papers, No. 123. OECD Economics Department, OECD, Paris. Huppes G., E. van der Voet, W. van der Naald, P. Maxson and G. Vonkeman (1993). New Market-OrientedInstruments for Environmental Policies. Graham & Trotman Ltd, London and Kluwer Ac. Publ. Norwell MA. IPCC (intergovernmental Panel n Climate Change) (1994). Climate Chane: Policy Instruments and Their Implications. Proceedings of the Tsukuba Workshop of IPCC Working Group III. Ceenter for Global Environmental Research, National Institute for Environmental Studies, Environmental Agency Japan.
92 Klaassen G. and D. Pearce (1994). "Economic Incentives and Air Pollution Control". Environment and Resource Economics Special Isue on Economic Incentives and Air Pollution Control, forthcoming. Kuik, O. P. Peters and N. Schrijver (1994). Joint Implementation to curb Climate Change: Legal and Economic Aspects. Kluwer Ac. Press, Dordrecht/Boston/London. NAPA (National Academy of Public Administration (1994). The Environment Goes to Market: the Implementaton of Economic Incentives for Pollution Control . NAPA Washington, July 1994 OECD (1989). Economic Instruments for Environmental Protection. Paris, 1989. OECD (1991a). How to Apply Economic Instruments. OECD, Paris OECD (1991b). Responding to Climate Change: Selected Economic Issues. OECD. Paris 1991. OECD (1992a). Climate Change: designing a tradeable permit system. Paris 1992. OECD (1992b). Climate Change: designing a practical tax system. Paris 1992. OECD 1994. Manageing the Environment: the Role of Economic Instruments. Paris, 1994 Opschoor J.B. (1991). "Economic Instruments for Controlling PMPs". Hans Opschoor and David Pearce (eds) (1991). Persistent Pollutants: Economics and Policy. Kluwer Ac. Press Dordrecht. Opschoor J.B. and R.K. Turner (eds) (1994). Economic Incentives and Environmental Policues: Principles and Practice Kluwer Ac. Publ. Dordrecht?london/Boston. Piacentino D. (1994). "Carbon Taxation and Global Aspects". In Opschoor and Turner 1994.
Warming:
Domestic
Policy
UN (1992). Declaration on Environment and Development. UNCED, A/CONF.151 /PC/ WG.III/L.33/Rev. 1. UNCTAD (1992). Combating Global Warming: Study on a Global Tradeable Carbon Emission Entitlements. UN New York
System of
Zhang Z.X. (1994). "Setting Targets and the Choice of Policy Instruments for Limiting CO2 Emissions". Wageningen Economic Papers 1994-2, Wageningen Agricultural University
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
93
M A G I C C a n d S C E N G E N " I n t e g r a t e d m o d e l s for e s t i m a t i n g r e g i o n a l c l i m a t e c h a n g e in r e s p o n s e to a n t h r o p o g e n i c e m i s s i o n s T.M.L. Wigley University Corporation for Atmospheric Research, P.O. Box 3000, Boulder, CO 80307-3000, USA EXTENDED ABSTRACT
MAGICC and SCENGEN are a suite of models that determine the regional details of future climatic change for specified emissions scenarios, together with estimates of their uncertainties. These models follow through and compare the consequences of a "policy" emissions scenario and a "reference" scenario. MAGICC (Model for the Assessment of G__reenhouse-gas I_nduced Climate Change) converts emissions to concentrations, to radiative forcing, to globalmean temperature and sea-level change. SCENGEN uses this temperature change output together with information from General Circulation Models (GCMs) to develop regional scenarios for climate change. Input emissions data (from an editable "library") are required for CO2 (fossil and land-use emissions separately), CH4, CO, NOx, NMHCs, halocarbons and fossil SO2. CO2 concentration changes are calculated using the carbon cycle model of Wigley (1993), which uses CO2 fertilization to give a contemporary carbon budget consistent with observations. CH4 concentrations are determined using the variable-lifetime model of Osborn and Wigley (1994). For both CO2 and CH4, user, best-guess, low and high projections are given to allow an assessment of uncertainty. N20 and halocarbon concentrations are computed using simple constant-lifetime mass-balance models. For the halocarbons, input is required only for four key species, C F C l l , CFC12, HCFC22 and HFC134a. A scaling method calibrated against a range of more comprehensive analyses is used to account for other halocarbons. The effect of halocarbon-induced stratospheric ozone depletion is included using a modification of the chlorineloading method of Wigley and Raper (1992). Fossil-based SO2 emissions are used to determine both the direct and indirect radiative forcing effects of sulfate aerosols, following the method of Wigley and Raper (1992). Tropospheric ozone and carbonaceous aerosol effects are also accounted for, albeit in a relatively simple way. For the gas-cycle and radiative forcing models, all parameters are consistent with the latest (1994 and 1995) recommendations of Working Group 1 of the Intergovernmental Panel on Climate Change (IPCC). Radiative forcing values are transformed to global-mean t e m p e r a t u r e changes and oceanic thermal expansion using t h e upwelling diffusion energybalance climate model of Wigley and Raper (1992). The temperature change values are used to drive ice-melt models for Greenland, Antarctica, and small glaciers and ice caps in order to obtain total sea level rise. The models currently
94 used are those of Wigley and Raper (1993), but these are in the process of being updated as a part of the 1995 IPCC exercise. Uncertainty ranges for globalmean temperature and sea level change are also calculated. The temperature and sea level results from MAGICC are being used by IPCC for their 1995 assessment of climate c h a n g e ~ i n this sense, the models used represent the current state of the art. MAGICC t e m p e r a t u r e output is used to drive the SCENGEN climate scenario generator. Regional patterns of climate change, AC(t) (the underlining here indicates a two-dimensional pattern), are calculated using AC(t)=AT(t)AC* where AT(t) is the global-mean temperature change and AC* is the normalized pattern of climate change. AC(t) may be temperature, precipitation, humidity, cloudiness, etc. on monthly, seasonal and annual time scales. AC* (based on more than 10 GCMs) may be either for single models or averages of a number of models. AC* values are obtained by dividing the results from individual GCMs by the corresponding global-mean temperature change values. This scenario generation method allows time-dependent patterns of climate change to be developed from either equilibrium GCM or transient coupled A/OGCM results (or both), under the assumption of a time-invariant "signal" pattern ( which can be justified by analysis of coupled A/OGCM results). It also allows results from models with widely different climate sensitivities to be combined. For the globe, SCENGEN gives output at the 5 ~ latitude by 5 ~ longitude level. For Europe and the USA, output is available at 1~ by 1~ or better. To obtain the higher resolution, the 5 ~ by 5 ~ data are smoothly i n t e r p o l a t e d t o 1~ by 1 ~ and added to high-resolution, high-quality baseline climatologies. Uncertainties are quantified at two levels, either by driving SCENGEN with low, mid or high global temperature changes from MAGICC, and/or by using the 90% confidence bands for AC* obtained from an analysis of inter-model differences. The whole system is embedded in a user-friendly shell, and is designed to run rapidly on a high-level (e.g. 80486-based) microcomputer. A full simulation can be completed in a few minutes.
Acknowledgments MAGICC and SCENGEN were developed using funds from the U.S. Department of Energy, the European Community (DGXI), the U.K. Department of the Environment, and the Electric Power Research Institute. Most of the work was carried out in the Climatic Research Unit, University of East Anglia, Norwich, UK, by the author, Sarah Raper, Mike Hulme, Mike Salmon, Jiang Tao and Tim Osborn.
REFERENCES I T.M.L. Wigley, Tellus 45B (1993) 409-425. 2 T.J. Osborn and T.M.L. Wigley, Climate Dynamics 9 (1994) 181-193. 3 T.M.L. Wigley and S.C.B. Raper, Nature 357 (1992) 293-300. 4 T.M.L. Wigley and S.C.B. Raper, (In) Climate and Sea Level Change: Observations, Projections and Implications (eds. R.A. Warrick, E.M. Barrow and T.M.L. Wigley), Cambridge University Press, Cambridge, U.K., (1993) 111-133.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
95
The Process of Developing Policy Based on Global Environmental Risk Assessment D J Fisk Department Environment UK
Introduction I have been invited to give a short presentation on developing policy based on a global environmental risk assessment. I very much look forward to an exchange of views later in the morning about how policy and science interact, and what we have learnt from the process so far. For my part I am going to take the title literally and look at the global warming issue as if it were a formal problem in risk assessment. I want to use that framework to make one or two suggestions as to how the interaction of policy and research might evolve as the climate convention progresses.
Background The Intergovernmental Panel on Climate Change which began its work in 1988 and gave its first report in 1990, was a landmark in the development of technically based global environmental policy. Of course the assessment process has moved on since then. The Framework Convention for Climate Change has come into force, and national plans by most Annex 1 countries have been sent to the interim secretariat. In Berlin this March the Conference of Parties will meet for the first time. Amongst its tasks will be to set up the Subsidiary Body on Science and Technology Assessment. Characteristic of the stepby-step approach of a modern international environmental instrument, the convention has set the Conference of Parties a review deadline of 1998 to set post 2000 goals. With all this new process coming on board, I very much commend the conference's wish to look at how interactions between policy and research can be improved. IPPC90 as a Risk Assessment The first IPPC process was in many ways a classic risk assessment the report identified the hazard - IR absorption and the gases which exhibited this property
2)
the risk of these gases accumulating in the atmosphere through scenarios and chemistry
96
3)
assessed the consequences - the impacts on sea level, agriculture, health etc
4)
looked at the options for risk management
This IPPC process, which in its formal risk assessment form is familiar to most of us handling micro-environment problems, has not been without its critics. Certainly, since 1990, the peer review process has been improved, but at the penalty of a much more time consuming process. Gone for ever are the pre-IPPC days when a small group of experts could sit down in Bellagio and write their conclusions up in an afternoon! There is the perennial problem of immediacy that plagues any risk assessment that is based on an active area of research. The early steps in the risk assessment can become outdated by the time that the final steps have been completed. IPPC90 you will recall had to use the earlier National Academy Science review for its consequenceanalysis because the IPPC90 risk assessment was still in progress. There was no opportunity to test the effect of the risk management responses in the climate and impact models. Despite these criticisms the IPPC has proved a powerful tool to assemble and assimilate the state of research. Indeed it picks up an enviable number of citations in its own fight in learned journals. Other global environmental areas, such as the state of the global oceans, may not be any poorer in underpinning science, but are disadvantaged because of this lack of synthesis machinery. While constructing a synthesis of the science is vital, it leaves the issue of how the science is to fed into the actions which might follow from it. The IPPC has more recently been wrestling with this aspect, and I will also focus on this issue.
Development of National Hans In charting the development of international action, it would be wrong to ignore the development represented by the preparation of national plans under the convention. It is a common experience that risk management takes on a different character once we move from analysis to trying to actively manage the hazards. Those disadvantaged by the action exercise their fight to question the risk assessment and consequences analysis. Often Governments or at least democratic Governments - find that they have inadequate instruments to deliver action by others unless they are equally convinced. The national plans submitted to the convention underline this point. Governments can set the framework for action, but they need the co-operation of others to deliver real changes in emissions trends. In that sense Dr Brenabo's paper on communications between scientists policy makers and society at large is especially important.
Hazard Identification and Risk Analysis Most people I suppose ask 'Is climate change a problem?' and if the answer is yes 'what could I do that would have any effect?' These are not bad questions for the policy making process at any level. Answering these questions through risk assessment begins by establishing the hazard. No one has seriously challenged that the infra-red absorption
97
property of greenhouse gases is a hazard. Identification that there is a hazard in a risk assessment is usually sufficient to establish the case for 'best practice' in handling the hazard, or in climate change parlance 'no regrets' measures. I shall have something more to say about 'no regrets' when I come to risk management options. Hazard identification is only the start of the analysis. The next step. Risk analysis has proved more difficult. S c e n a r i o s are S c e n a r i o s
In a traditional risk analysis, situations are envisaged which might realise the hazard. The probability of each situation is assessed and the overall probability of the hazard being realised computed. Superficially IPPC have worked in a similar m a n n e r . Four scenarios were exhibited in IPPC90. The number expanded in IPPC92 to six. To these might be added the scenarios developed by World Energy Council. IPPC have often been pressed to identify the most likely scenario, or attribute probabilities to the set of scenarios. This would certainly permit a conventional risk analysis. However close scrutiny of the time axis of these scenarios which extends to 2100 shows that IPPC would be right to stand its ground - a 'scenario is a scenario not a probability weighted forecast'. Let me argue this point by looking at the oft quoted IS92a scenario. Suppose I were to treat this as a forecast. Then I can make a number of other deductions about the long term future. First, taking into account the implied cost of nuclear power in IS92a it would be clear that despite the passage of a 100 years and an ever widening technical and scientific base we had found no cure for cancer that trivialised incidence of the disease. The world, although incredibly richer would still not be at peace and would still be concerned at nuclear proliferation. Treating IS92a as a forecast we do not appear to have found a room temperature super-conductor which would of course have revolutionised energy storage and transport. No doubt with that knowledge we could save a guilder or two elsewhere in the Dutch national research budget! If the choice of technical revolutions look as if I am biasing IS92a downwards perhaps I might add it also implies that we do not seem to have cracked the biochemistry of ageing either in 100 y e a r s . If we had it would be difficult to guess what the population driver figures might look like. Anyone of us could associate a subjective probability to these events. However the likelihood of consensus amongst 5 billion people as to what those subjective probabilities would be seems rather remote. This is in contrast to forecasts in the shorter term - or at least the shorter term to the climate scientist. These forecasts limit themselves to a time span in which even if these technical shocks were to be realised their probability of influencing the forecast is vanishingly small. For the sake of a name we might call the end of such a time span a Schumpeter horizon to acknowledge that beyond it Joseph Schumpeter's creative destruction implies that we can no longer rely in any sense on extrapolation. The recent IEA forecast for global carbon dioxide emissions, for example, falls within the Schumpeter horizon. Such a horizon is also the natural time span in which to set step by step legally binding commitments in conventions, at least for those who want to take their commitment seriously. The scenario process is inescapably normative. This is less of a problem than might be supposed at the stage I have reached in the risk analysis. If for example you turn to the
98 Brundtland Commission Report you will find the development of a normative scenario for the economic development of the world's nations. It may not happen but the scenario embodies widely held aspirations for the future. Thus while IPPC could happily construct an infinite number of scenarios, it is only those that express our aspirations that we believe we want to see actively pursued that need be included in the initial risk analysis. The scenarios ought for example to show the property of sustainable development. I deduce from this argument that the policy making process in the convention needs in due course to address which scenarios reflect the aspirations of its parties. It ought to be these scenarios which make up the feedstock of climate models and impact estimates. In the first stage of a risk assessment the key scenarios are those which reflect aspirations without being fettered by considerations of climate impact. It is a matter of taste whether the term 'business as usual' quite captures that flavour. Consequence Analysis It has become rather popular to open discussions on climate change with a recital of the uncertainties in climate modelling. From a risk assessment point of view this narrow focus is not altogether healthy. Admittedly the IPPC90 key index of modelling uncertainty - the climate sensitivity - ranges over a factor of 3 from lower to upper bound. But this is no larger than the range of climate forcings arising from the IS92a scenarios themselves. It is therefore not just a range of climate science possibilities that need to be explored. The point to note is that all but one of the scenarios have rising climate forcing, and that all estimates of climate sensitivity are positive non-zero. Thus under these scenarios the modelling uncertainty simply changes the time at which a certain climate change condition takes place. Uncertainties in climate modelling influence the risk management not the initial risk analysis. The question that is seldom answered by professional sceptics is just what scale of climate change matters, and whether that degree of climate change is within the range of scenarios, taking into account uncertainty in climate sensitivity. These are key questions that the consequence analysis must address. The degree of precision that we need from climate modellers depends critically on the degree of precision required by the impact assessment. Types of Impact Assessment In collective environmental decision making, the 'least helpful' outcome for a consequence analysis is that changes are found to be gradual. It may be gard to find consensus on a trade-off. In contrast sharp changes, sometimes called comer solutions from optimisation theory, are very important findings for gaining a consensus. By their nature they bring together a coincidence of different interests. For example there may be a rate of change at which temperate forests decline, or a sea temperature at which the Antarctic ice sheet begins to shelve, or the thermohaline circulation stops. I hope that it not too self evident if I suggest that these classes of impacts deserve special priority in impact research as a basis for collective decision taking.
99 It may of course be that such sudden changes do not exist and that climate impacts are gradual in their effect. There have been some attempts to tackle gradualist change by normalisation to some valuation criteria as a basis for contracting trade-offs. The new IPPC assessment will be reviewing some of these approaches. Personally I have some doubts that we have fully worked out how to use this methodology in the context of a long term issue like climate change. In particular it is not clear how much prior context has to be agreed before the figures have a hope of gaining a consensus. However the approach teases out one difficulty in a reductionist approach to impacts. We simply do not know what it would feel like to be living during a time that climate change was so apparent that we lacked confidence in how the climate might change around us. It is common experience in environmental policy that society's response to a consequence, changes once the consequence is realised. In climate change this state of mind presumably sets in when we are confident that we can detect the enhanced greenhouse gas signal in the global climate record. I would argue that will be an important marker in the development of the convention. I conclude that impact studies have a special importance in a risk assessment because they define the precision demanded of climate models, and their structure determines the likelihood of a consensus to respond to the risk. Risk M a n a g e m e n t - Yet to be Begun
What I have discussed so far is simply establishing the climate change consequences of pursuing our aspirational scenarios. For any scenario that breaches the conditions of the climate change convention - adaptable rate of change to a safe stable concentration - the risk management component of a risk assessment comes into play and requires us to revisit our scenario. It would certainly be true to say that we have hardly begun to articulate in the convention how the next steps in risk management analysis could be undertaken. The conventional history of risk management in an environmental instrument starts with a few suspect hot spots. These lead to some generalised early action. By the end date of this agreed action the underlying science is clearer and usually substitute technology has been developed. The final stage of the instrument is then played out. The parties in the climate convention are clearly struggling with the first stage, which we label hazard management. Countries with the technical and social means to devise ways of abating emission have drawn up national plans. The unusual aspect of the convention is its timescale. We may still have not detected man-made climate change by the early part of the next decade. Although there are some good ideas in the national plans the dream substitute technologies have yet to come into play. It is of course difficult to judge how the Conference of Parties will take this issue forward. Most countries have found that the store of 'no regrets' measures is difficult to unlock, not least because those who have interests in the older policies often take a 'I regret nothing' stance. The Conference is therefore likely to be interested in looking
100 more closely at means of better co-ordinating national measures, possibly through a Protocol to the Convention. Germany has already submitted some ideas on these lines. If risk management were to become the underlying principle through which the Conference of Parties developed its work, then progress would undoubtedly be step-bystep. Measures would be assembled to take effect over my Schumpeterian horizon against specific commitments. As each target end-date was reached, the Parties would review the effect of their measures, assess the improved knowledge of the climate science, and inspect the concentrations levels of greenhouse gases that had actually been reached. Let us suppose in conclusion that the convention was to take this route through its subsequent meetings. What kind of dialogue with science might be needed? I would suggest on the basis of the points made: For Improving Hazard Analysis (1)
Continuation of IPPC reviews of global atmospheric chemistry, as the convention attempts to treat greenhouse gases in a comprehensive fashion.
For Improving Risk Analysis (2)
Clear rationals from IPPC and SUBSTA for scienarios.
For Improving Consequence Analysis (3)
Continued Scouting for 'comer solutions' in impacts.
(4)
Differential impacts analysis (i.e. comparing impacts between different risk management scenarios).
For improving Risk Analysis
(5)
Analysis of the comparative performance of differing measures in national plans.
I have not meant to exclude work on either large scale climate modelling or extensive impacts research. I thought however that this important work might be referenced in a different context. I pointed to the time when there was general agreement that man-made global warming had been detected in the climate record. It would be an important turning point in the development of the convention, but also in the nature of file public debate. It is difficult not to have noticed how extreme local climate events in the recent past have spurred public interest and debate in this issue. In the past our meteorological advisers have been able to re-assure us that these events were not distinguishable from natural variation. In the future it may be more difficult to make that assertion. The research into large scale
101 modelling enterprises may then become important not just for projecting future change, but interpreting the change that will be seen around us.
The views expressed in this paper are those of the author, and do not necessarily represent those of the UK Department of the Environment
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
103
Communication Among Scientists, Decision Makers and Society: Developing Policy-Relevant Global Climate Change Research. J. Christopher Bernabo Science & Policy Associates, Inc. Suite 400 West Tower, 1333 H Street N.W., Washington, DC 20005, USA Abstract
Defining the research most relevant to policy is not simply a technical task that can be answered by scientists. Decision makers need and value information differently than curiosity-driven scientists. In order to link science more effectively to policy, the two communities must gain a greater mutual understanding. Decision makers must define their needs so that scientists can determine how, and by when, research can address these needs. This vital dialogue between communities typically has been more ad hoc than systematic. The complexity and urgency of the global climate change issue necessitate ongoing communication between scientists and decision makers on the information needed for policy development and what research can provide. The results of relevant science policy dialogues are discussed herein.
1. INTRODUCTION Effective communication between researchers and decision makers is a crucial ingredient for successfully addressing society's pressing environmental concerns. The increase in policy makers' demands for research that is relevant to solving societal issues highlights the communication gap between the technical and policy communities. The gap, largely caused by lack of mutual understanding, results in flawed and inadequate communication that hinders decision making and confuses the public. This paper examines the cause of this communication gap and describes the significance of recent efforts to develop more fruitful science-policy dialogues on the issue of global climate change. First, the post-Cold War shift in government priorities for research funding is described; then the underlying relationship between science and policy is explored to identify key sources of ongoing miscommunication. The paper then explains the importance of defining policy-relevantscience questions that research can address. Finally, three projects are described involving the elicitation of decision makers' information needs in The United States, The Netherlands, and internationally.
2. POLICY RELEVANT RESEARCH Fifty years after World War II, the major political, social, and economic changes sweeping the globe are causing an historic shift in the emphasis of research funded by governments. In many nations, such as the United States, national security was a major societal justification for massive public funding of the natural sciences and engineering. The end of the Cold War military competition has caused a wide spread reevaluation of science funding priorities [ 1,2].
104 Furthermore, the public's faith in science as an unquestioned source of ever increasing material living standards has been shaken by the emergence of many technologically-induced environmental problems [3]. With economic constraints to growth and global competition rapidly increasing, there are greater demands to direct government-funded science and engineering toward solving pressing societal problems. The emerging post-Cold War rationale for government funding of research has five priority factors: 9 Emphasizing science that provides societal benefits; 9 Linking research programs to the needs of decision makers; 9 Providing economic development and competitive advantages; 9 Developing partnerships with diverse stakeholders; and 9 Leveraging international research activities and programs. None of these factors are new, but the increased emphasis upon them in guiding research investments is a major development for science in the post-Cold War period. This greater attention to investment return and the societal relevance of research will require enhanced efforts to improve the communication between scientists, decision makers, and the public.
3. RELATIONSHIP OF SCIENCE TO POLICY "Science has the first word about everything and the last word about nothing," Victor Hugo observed. The truth of this is inherent in the relative roles that both "objective" scientific information and "subjective" human values inevitably play in decision making. Environmental policies are developed by interpreting and applying technical information in light of the needs and human values of society (Figure 1). Viable policies must not only be technically sound but also socially, politically, and economically acceptable.
SOCIETAL FACTORS POLICY DEVELOPMENT TECHNICAL INFORMATION
ACTION
7
~., Figure 1. The relationship of science and human values in policy development, showing primary interactions and feedbacks.
105 Science alone cannot provide answers to policy makers' ultimate questions because science necessarily is silent on the human values that underlie the decisions societies make. The scientific method itself is designed to screen out the value preferences and biases of the subjective human beings who conduct research. Technical information is useful in identifying issues, developing options, providing understanding and evaluating consequences for policy actions. But in the end, human values must be applied to determine what is "good" policy for a given society on a specific issue. Take the example of nuclear energy: is promoting it a good or bad policy'? On the surface this appears to be a scientifically answerable question yet nations with access to the same technical information have made different choices about the best policy for their societies. Indeed, there are Nobel Laureates that staunchly argue opposite sides of the case because the question ultimately involves human values. Science can only approximate the risks and benefits, but a subjective value judgement must be applied to decide what ratio between the two is acceptable to a given individual or society [4]. Many of the difficulties scientists and policy makers face in communicating and working together arise from differences in their professional cultures (Table 1). Both the scientific and decision making communities experience frustration over the paradoxical relationship between information development and policy development. The public and policy makers often perceive that science is more effective at identifying uncertain problems than it is at providing certain solutions. On the other hand, the technical community becomes frustrated by the perceived inability of policy makers to grasp the facts and take what they personally judge is the "logical" action. Table 1
Contrasting Professional Cultures of Scientists and Policy Makers Science
Policy
Objective Facts Proof Rational Measurements Incremental Progress
Subjective Values Beliefs Emotional Perceptions Deadlines and Crises
Applying technical information to decision making is a fundamentally different type of activity than discovering new knowledge. Alvin Weinberg coined the term "trans-science" to describe the process of using technical information in making decisions that inherently transcend the bounds of science [5]. He points out that facts alone are not sufficient even for weighing the benefits and costs in policy issues, because subjective values must be applied in choosing what facts to use and how. Harvey Brooks concludes that, "the facts that are selected and the way they are presented to the public may have a greater political impact than the facts themselves" [6].
106 4. SCIENTIFIC UNCERTAINTY AND POLICY DECISIONS Environmental policy debates typically involve discussion of uncertainties and how much certainty is "enough" to justify a proposed action. The question of how much information is adequate for a given policy always involves a value judgement and cannot be answered by scientific research alone. There is no objective point in science that defines enough certainty for policy. Research can only quantify the uncertainty in the science, and even that with great difficulty, but policy involves many other types of inherent uncertainties. Policy decisions must consider uncertainties about matters such as the significance of facts, the perceptions of the issue (opinion polls), the economic and social viability of the proposed solutions, and the actual versus intended consequences of the action. The degree of scientific consensus is just one part of the information needed for decision making. Brooks cautions that we "should be careful not to expect that scientific consensus should be a necessary condition for policy consensus, an expectation to which scientists tend to be too prone" [7]. For instance, we might have no fiscal policies if action required consensus on economic predictions. The policy makers' roles include making subjective judgements about which information should be acted on and how much certainty is enough for decision making. The degree of certainty that is adequate for policy can be viewed as an equation balancing scientific uncertainty and political uncertainty. Two general principles apply to environmental issues: The greater the societal consensus on an issue, the less scientific certainty required for action.
II.
The higher the societal costs of a policy, the greater the scientific certainty required for action.
The inverse of these principles also is true. They imply that enough certainty in the science is always defined relative to the political certainty in the issue. Therefore, enough scientific certainty in the policy process is a dynamic factor, not a static end point from research. Two examples illustrate these principles. The United States and Canada fully shared scientific information on acid deposition; they had joint monitoring programs and the same degree of technical certainty on the issue. Nonetheless, lower scientific certainty was required to justify policy action by Canada because there was much higher political certainty than in the United States. Over 90% of Canadians believed that acid rain was a serious problem, while there was no such consensus in the United States. Canadians saw a threat to their major industries--timber, fisheries, and tourismnfrom the potential damages. In the United States, pollution control costs were instead perceived to be a threat to industry and jobs. In essence, all details of politics aside, the reason for the national differences in the thresholds of scientific certainty required for action was simple and predictable. In the United States, chlorofluorocarbons (CFCs) were banned as spray can propellants back in 1978. At that time no ozone hole had appeared and scientific certainty about the issue was lower than about acid rain in 1980 or global climate change in 1994. The threshold
107 of scientific certainty was low for the initial CFC ban because there was political consensus that the risks of skin cancer were not judged to be worth the benefits of protecting a few jobs. Further banning CFCs from all other uses awaited higher scientific certainty because of the greater societal costs involved.
5. POLICY-RELEVANT SCIENCE QUESTIONS For scientists to assist effectively in the development of policy, their research needs to be focused on the questions of greatest value to decision makers. Examining past experiences in applying science to address environmental issues helps illustrate the importance of defining the policy-relevant science questions to guide research. Policy relevancy is determined by the specific needs of the information user (policy maker) not the interests of the information producer (researcher). Unfortunately, the questions investigated by curiosity-driven science are often different than those required to provide the most policy-relevant information. This occurs because decision makers only require the information that can materially assist their specific deliberations, whereas scientists seek greater fundamental understanding of their subjects. Other mismatches exist because of the different values attached to information in the research and decision-making realms, and because policy makers need information that cuts across fields of research. There are three general ways to define policy-relevant research questions: 9 Educated guesses: This has been the traditional means whereby scientists who study an issue presume to formulate what questions they deem relevant to decision makers. Although quick, this investigator-driven approach fails to examine the real needs of the policy users. Curiosity-driven questions tend to dominate these agendas without the benefit of decision makers' input. 9 Multi-stakeholder dialogues: This approach involves systematically eliciting the information needs of decision makers in the various stakeholder groups for the issue. Interviews and meetings are utilized to determine what the information users' need. Then scientists are involved in examining and responding to these needs in a facilitated process that ensures results reflecting the best input from both co~rununities. This process can be accomplished over several months and builds direct dialogue between the participants, helping bridge the science-policy communication gap. A limitation of this method is that it does not allow distinguishing what information participants say they need from what they may use in practice. 9 Social science research: This is the most intensive approach and goes beyond eliciting the expressed needs of decision makers to study their actual behavior in applying information. It involves carefully designed research and field studies observing the behavior of subjects involved in decision making. This approach provides valuable insights into the use of technical information in policy development. This scholarly approach requires extended periods, usually years, during which the policy relevant questions may shift. Moreover, it does not necessarily build ongoing dialogue between the science and policy communities. Whereas the educated guess approach has typically been used, a combination of the multistakeholder dialogues and social science research is most effective. The dialogues facilitate timely development of broad policy-relevant science questions and build mutual understanding as a basis for consensus between the participants. This approach directly
108 enhances the effectiveness of linking science and policy. The longer-range and more intensive social studies of decision makers' and scientists' behaviors help provide deeper understanding for designing more effective communication. Interactions between these two types of approaches is valuable in assisting each to reach its goal. The remaining sections of this paper describe three projects that represent multi-stakeholder dialogues aimed at defining policy-relevant research questions for global climate change. The general significance of the results of a pioneering study done in the United States in 1992 are reported. The second study was done in 1994 for The Netherlands, and it improved on the methods in the initial project. The third study is being conducted in 1995 by a joint team of the investigators from the U.S. and Dutch projects and applies the previously developed approaches to an international context.
6. U.S. DECISION MAKERS' CLIMATE INFORMATI()N NEEDS In 1990, a number of U.S. research organizations became concerned that the governmentsponsored U.S. Global Change Research Program (USGCRP) may not provide an adequate basis for the inevitable information demands of future policy development. They decided that a first step in moving toward a policy-relevant research agenda was to determine generally what information decision makers needed, and they launched the "Joint Climate Project to Address Decision Makers' Uncertainties" [8]. This unique private-federal partnership was sponsored by the Electric Power Research Institute (EPRI), U.S. Environmental Protection Agency (EPA), the U.S. Forest Service (USFS), and the U.S. Departments of Energy (DOE), Agriculture (USDA), and Interior (DOI). The project was designed and conducted by Science & Policy Associates, Inc. The Joint Climate Project established a multi-stakeholder dialogue to help identify some major questions U.S. decision makers had about global climate change and then had scientists determine what research and time frames would be required to address those questions. 6.1. Focusing on the Needs of Decision Makers
The Joint Climate Project identified policy-relevant research using two interactive phases: U.S. decision makers first defined their information needs, then scientists gave feedback on these needs and determined the research required to address the policy-relevant questions. During the first phase of the project, the needs of the users of climate information were identified through interviews, workshops, and focus groups involving national-level decision makers. These individuals included dozens of U.S. government and private sector officials, ranging from working-level experts to members of Congress, Administration officials, and industry CEOs. They were invited to participate in the project on the basis of their active roles in climate change policy and their diverse perspectives, from federal regulators and resource managers, to industrial representatives and environmental groups. The interactive process lasted six months and resulted in a consensus set of policy-relevant general questions for researchers to address. Then, leading experts in climate-related fields were convened at a workshop to discuss the specific questions developed by the decision makers. The scientists were chosen for their activities in research or in the synthesis of research results. They represented a broad range
109 of expertise, including climate system modeling and monitoring, managed and unmanaged ecosystems, energy and technology, as well as economics and social sciences. The workshop participants examined the research needed to address the questions and the expectations for providing better information over the next two, five, and ten years, and beyond.
6.2. Findings of the Joint Climate Project The consensus-identifying approach of this project yielded several key findings that reflect the general concerns of decision makers and the responses of the research co~ruuunity. In discussions with these two communities, several common themes emerged for enhancing communication and increasing the value of research results.
6.3. The Concerns of Decision Makers The participating decision makers identified several general principles that define policy-relevant questions for research. The project was conducted during the year before the United Nations Conference on Environment and Development (UNCED). Talks were well underway to craft a Framework Convention on Climate Change. Therefore, many government policy makers focused on these and other ongoing international negotiations and conferences. The officials specifically asked for information to support follow-up actions to UNCED and preparations for future events. For their part, non-government decision makers expressed concern with the possible regulatory implications of proposed actions. 9 Climate Change Impacts and Human ReL~ponses are Key to Decision Making: Aside from pressing international policy issues, decision making is driven by concerns about the potential impacts of changing climate at the regional level, rather than predictions of changing global mean values of climate variables. Specifically, input is needed from the economic, social, and ecological sciences on the potential regional impacts of climate change and the consequences of possible response strategies. Any response to the threat of climate change must be measured against what is at stake. Therefore, more information is needed on the ecosystems, regions, and human populations that are most at risk from potential climate changes, even if atmospheric research is still unable to provide reliable predictions of the specific changes that will drive effects. 9 Implications of Uncertainties Need Clarification: Researchers need to clarify the sources and implications of policy-relevant scientific uncertainties and estimate time frames for reducing them. Many uncertainties, although scientifically profound, may be relatively insignificant for developing policies. There is a need to define better which uncertainties are most important for policy development and resource management, and the practical implications of these uncertainties for decision makers. 9 Certainty is Not a Prerequisite for Action: During the project, several decision makers stressed that the resolution of all scientific uncertainties is not a prerequisite for policy action. Decisions are regularly made in the face of some uncertainty. Decision makers will apply their constituents' values to determine how much certainty they judge is enough to take political action. 9 International Perspectives Drive Policy:
110
6.4. The Response of Researchers In the next phase of the project, a diverse group of U.S. experts in climate-related fields were convened to examine how research could best address the questions posed by decision makers. Specifically, the scientists examined what types of research are needed to reduce the uncertainties in the policy-relevant questions and estimated the time frames for possible results. 9 Timely Results: Some of the key questions decision makers have about climate change can be addressed within a short time frame on the basis of analysis and interpretation of currently available scientific information. Although more complete scientific understanding of climate change may be decades away, much of the information needed to begin addressing decision makers' questions can be provided within two to five years. This could include a comprehensive evaluation of indicators of global climate change, a preliminary vulnerability analysis for systems and regions most sensitive to climate change, and an assessment of the sources and levels of greenhouse gas emissions for use in identifying potential mitigation and adaptation options. 9 Parallel Approach to Climate and Human Responses Research: Scientists need not wait for accurate climate predictions before beginning their research on potential impacts and response options. It is neither necessary nor practical for research to progress sequentially from the climate system, to the impacts, and then to the potential human responses in order to provide useful results for decision makers. Much can be done to improve the understanding of impacts without waiting for accurate regional climate predictions. For example, integrated regional and multi-sectoral models~using climate, ecological, demographic, economic, and social data collected at the regional level--can provide essential information on potential climate responses, the vulnerability and adaptability of key systems, the extreme ranges of change, and the impacts of climate change on the global marketplace. 9 Greater Emphasis on Impacts and Human Responses Research: Information on climate change impacts and response strategies has the greatest potential for assisting decision makers, yet these fields are the least researched. Many of the key questions identified by decision makers involve a significant amount of new socioeconomic, behavioral, and ecological research. However, only modest increases in funding for these disciplines would be necessary to achieve useful information for policy within a few years. Social science and economic research, in particular, receive a small percentage of federal funding, but are critical for making decisions about climate change. 9 Integrated Assessments and Case Studies: Integrated assessments of the causal linkages from emissions through impacts and human responses would help structure information for effective use in decision making. Such assessments would incorporate natural and physical sciences, economics, and social factors, including technological change and adaptation. In addition, a coordinated examination of case studies of regional climate variability is needed--based on historically documented events that show how societies have responded to past climatic variations. This information would provide valuable insights on how to treat future events. 9 Expect the Unexpected: Multi-disciplinary research on potential surprises is also important, given their potentially serious implications for decision making (i.e., climate change could be much worse than anticipated, or it could be insignificant). Decision makers and scientists should frequently re-examine research on potential surprises, given that scientific progress is
111 incremental and new information may become available. Based on this information, contingency plans could be developed to prepare for unforeseen events. 9 International Perspective: Because of the global dimensions of the issue, an international perspective for research is essential. Although decision makers may be most concerned with regional and local consequences, developing world issues (such as population and economic development as well as the pace, quality, and sustainability of development) will be critical. Assessing the ability of the international community to implement mitigation and adaptation measures is important for evaluating the effectiveness of response strategies on the climate system. The project asked researchers to identify the potential types of information that research could provide to address decision makers' concerns in two, five, and ten years. The participants provided educated estimates of the potentially available information for time frames of interest to decision makers. These estimates were developed without regard to financial or other resource constraints. Furthermore, the researchers suggested what research could do, and not what currently planned efforts will do. 6.5. Lessons in Communication
Discussions during the Joint Climate Project with representatives of both communities provided ample evidence that decision makers and researchers are uncomfortable with the present situation. Both are anxious to develop and sustain a productive dialogue. Both would like to increase the effectiveness of the research community in the decision-making process. Both agree that a two-way bridge must be developed to span the communications gap between the two communities. But to truly close this gap, to construct a bridge between the two communities, will take more than wistful expressions and lofty pronouncements. There is no substitute for sustained effort and innovative institutional arrangements. The decision makers and researchers who participated in the project agreed that greater attention must be paid to the development of systemic communications processes. In particular, both sides need to recognize the following points. 9 N o t an Either~Or Decision: Decision makers' choices are not simply between pursuing research or implementing response strategies. Rather, the challenge is to define the appropriate levels of each over time. Researchers need to provide a broad array of information to address the complex and interacting decisions on global climate change. Decision makers, for their part, need to recognize the long time scales involved in research and, thus, the importance of continuity of funding and program goals. 9 Global Climate C h a n g e in a Relative Risk Context: Prediction of changes in mean global temperatures does not give an adequate picture of the societal risk that can be related to every-day experiences. The risk of global climate change needs to be compared to the risks of other economic, social, and environmental issues. Because the public tends to respond to perceived crises, assigning relative risk would help decision makers distinguish between verifiable serious threats and possibly misplaced public concern. Given that risk is a function of both the probability and the magnitude of the expected consequences, better data on possible impacts are critical to better estimates of societal risk. 9 Urgent N e e d f o r Education: A concerted effort is needed to educate decision makers on the facts and uncertainties of global climate change. Since public concern is often the
112 impetus for formulating policy, scientists need to communicate technical information to the public more effectively and more frequently. In addition, scientists need to learn more about the decision-making process and the types of information most useful for policy. Frequent, two-way communication between decision makers and researchers is essential if research is to play an effective role in the decision-making process. 9 Research Does Not Always Provide the Answer: Decision makers should understand that additional research can increase the amount of uncertainty in some areas. Researchers should inquire about how much certainty decision makers require to take a specific action. To this end, uncertainties that are not relevant to decision making should be identified early in the process. Decision makers and researchers should also seek ways to manage continuing uncertainties. For example, building resilient institutions would provide a flexible response to any future changes in climate, albeit at potentially significant costs. Contingency plans allow decision makers to prepare for possible climate outcomes through R&D on response technologies, without needing to deploy them. 9 Develop an Ongoing Assessment Process for Research: To improve communication and better inform decision makers, research efforts should include an iterative assessment process. These assessments not only help to identify the relevant questions, but also serve to structure the research results and, thus, facilitate clearer communication between the two communities. Furthermore, the assessment process provides valuable input to the planning of policy-relevant research.
6.6. Project Significance The Joint Climate Project represents a preliminary step in determining how researchers can assist U.S. decision makers over the coming years and decades, thereby helping to bridge the communication gap between these two corrununities. A more frequent and systematic twoway dialogue will be needed between decision makers and researchers in order for research to inform the decision-making process. Discussions with decision makers and researchers during the project revealed that both communities are very interested in developing and sustaining a productive dialogue. Both would like to increase the effectiveness of the research community in the decision-making process. Following the successful dialogue established by the Joint Climate Project, other similar efforts were initiated for climate change in The Netherlands and for biodiversity in the United States [9]. These types of dialogues also need to be supplemented by more in-depth social science studies to elicit greater understanding of the behavior of decision makers in applying science. A better mutual understanding of the professional cultures of researchers and decision makers is required to enhance the effectiveness of linking science to policy.
7. NETHERLANDS POLICY OPTIONS STUDY "Policy Options Addressing the Greenhouse Effect," a climate change project conducted in The Netherlands, had an approach and goals that were consistent with the Joint Climate Project. The Policy Options study was conducted for the Dutch National Research Programme on Global Air Pollution and Climate Change (NRP) by Prof. Pier Vellinga with his colleagues at the Institute for Environmental Studies (IVM) at the Free University of
113 Amsterdam and Prof. Jan Klabbers, with consultation by Dr. Chris Bernabo [10]. Within the project a dialogue has been initiated between policy makers, scientists, and other societal actors to look at how Dutch society can cope with the risks of climate change and the challenge of sustainable development. The project produced two types of results. The first included various policy options and related actions. The second, and probably more important, results were related to the process itself. There was an improvement in the communication and discussions among all the stakeholders which, over the longer term, can lead to a more solid foundation for action.
7.1. Project Objectives The Policy Options study was designed to bridge the gap between perceptions of policy makers, researchers, and public interest groups. The specific objectives were to: 9 Reinforce communication between the three communities; 9 Illustrate the perceptions of the communities; 9 Examine policy development options; and 9 Inject the options into the Dutch policy development process.
7.2. Project Approach The first step in the process to develop climate policy options was to identify the issues through interviews and workshops with policy makers. Natural and social science researchers then assessed the issues in position papers and workshops. Next, round table discussions linked the science and policy perspectives. The outcomes of these discussions provided the basis for the development of a range of policy options and related actions.
7.3. Resulting Policy Options The key policy options that emerged from the study were: no-regrets (actions which may be economical regardless of climate change considerations, although they may not be considered no-regrets by every country), least regrets (actions which adopt the precautionary principle), acceleration (encouraging reductions through subsidies or taxes), technological innovation, and institutional(ised) cultural change. The five options that have been generated effectively illustrate the complexity of the climate change issue with respect to causes, uncertainties, international relationships, and the values and norms that are at stake. They acknowledge the divergence of the views and interests of all players, and encourage working towards convergence of actions. Full details of the policy options may be found in the final report of the Dutch study [10].
7.4. Recommendations Recommendations are based on the observation that within the natural sciences it was relatively easy to reach a shared view on climate change, but the bridge between the natural and social sciences was rather difficult to make. The project's recommendations include: 1. Enhance the Communication Between Scientists: Discussions in the science workshops revealed that economists, sociologists, philosophers still show large discrepancies in their
114 view of the problem and the paradigms used in their approach to the problem. As climate change and sustainable development require open, interdisciplinary minds, much work has to be done to improve the dialogue between the disciplines concerned. 2. Improve the Science-Policy Interface: The improvement of the science-policy interface will promote the assimilation of scientific results by policy actors and will also help in identifying the relevant research questions. This implies a broadening of the communication between the science community and policy actors from the private sector, the national and local government, and public interest groups. It is recommended that attention be paid to sector-specific constraints and opportunities of climate change and sustainable development. 3. Integrate Climate Policy into Broadened Socio-Economic and Environmental Policies: Various key societal groups do not perceive climate change as a problem that warrants stringent measures. All groups seem to agree, however, that environmental policies, including those relevant to climate change, should be integrated into broad socio-economic policy. 4. Address Two Fields of Priorities: The project revealed two fields of priorities, that necessarily need to be addressed: improving the dialogue between natural and social sciences; and improving the dialogue within the social sciences.
7.5. Project Significance The Dutch project laid a solid basis for a continuation of the fruitful communication between the researchers attached to the NRP and the policy actors from all sectors of society. During the project it became clear that there is a vast body of knowledge available outside the scientific community. Different but valid perceptions exist about the various aspects of climate change and climate change policy outside the scientific community. Through the project, these have been initiated and can now serve as an important source of information both for policy actors and researchers.
8. INTERNATIONAL CLIMATE CHANGE PROJECT The success of the U.S. and Netherlands studies encouraged the development of a project applying a similar process at the international scale. The project on "Enhancing the Effectiveness of Research to Assist International Climate Change Policy Development" (International Climate Change Project) is designed to determine the range of uncertainties and information needs of decision makers in relation to global climate change in an international context [ 11]. The project also assesses the research needed to help answer the associated questions and facilitates dialogue between scientists and policy makers at the international level. S&PA, IVM, Prof. Jan Klabbers, and Professor Bill Moomaw of the Tufts University Fletcher School of Law and Diplomacy in the United States have undertaken a project designed to address these issues at the international level. The initial phases of the project are funded jointly by the Dutch NRP and the U.S. EPA.
115 8.1. Project Goals The goals of the International Climate Change Project are to: 9 Identify and scope the range of policy options under consideration by representative countries, for which future research information is needed. 9 Determine the research required to address the information needs relevant to the range of policy options identified and to help guide the planning of policy-relevant research. 9 Enhance the dialogue between the decision making and research communities at the international level for the climate issue. The international dialogue fostered by this effort will promote a better understanding between decision makers and scientists both nationally and internationally. 9 Facilitate the planning of research that is more relevant and usable by decision makers. Exercises such as this project make a lasting contribution to improving the utilization and linking of science with policy development.
8.2. Project Approach The project compliments the current international activities relating to climate change and explores the long-term policy questions that require research-based information needs to support the range of policy options identified. The International Climate Change Project utilizes and updates the results obtained from the similar studies in the United States and Netherlands. The project will be undertaken in three phases: 9 Phase I - Project Design and Analysis: Activities in Phase I covered project planning and design to formulate the scope and tasks of the project. A project Steering Committee provided guidance and recommendations on the scope of the project. Selection criteria for choosing participating countries were developed together with the procedures and approach for undertaking interviews and workshops in the subsequent phases. These selection criteria were designed to promote the selection of countries that would provide a wide range of policy options and the information and research needs to support these options. The Steering Committee considered five types of criteria identified by the project team: environmental, economic, political, cultural/geographical and feasibility. Initially four countries, in addition to the United States and The Netherlands were chosen to be included in this pilot project: Brazil, China, India, and Poland. 9 Phase H - Identification of Policy Options: The objective of Phase II is to determine the range of policy options under consideration by policy stakeholders in the climate change issue. This will be undertaken through in-country interviews with representatives from the participating countries in February and March 1995, including an update of the information obtained during the national projects in The Netherlands and the United States. An international decision makers' workshop will be held in June 1995 to elaborate on the range, motivations, and substance of policy options. The output of Phase 1I will be a final report detailing the policy options identified and other results of the interviews and workshop. 9 Phase III- Identification of Research Needs to Address Policy Options Identified and InterCommunity Dialogue: Phase III will present these options identified in Phase II to the
research management community from the participating countries through the presentation of briefing papers and a workshop. A dialogue between the policy-making and research communities will be established through round tables or another workshop. The purpose of
116 this exercise is to identify research priorities and agendas and consider their implementation. Output from Phase III will be a report identifying key areas of research to address the policy questions identified as priorities in Phase II, synthesizing the dialogue between the policymaking and research communities, and summarizing the key areas for research and information needs identified during this dialogue.
8.3. Project Significance The selection of Brazil, China, India, and Poland as pilot countries allows for an examination of climate change policy and research options in situations that are markedly different than those found in more developed countries. Integrating the findings of these efforts with those of the U.S. and Netherlands studies will provide a preliminary picture of how the decision-making and research communities can work together to address climate change at the global level. The process used in the project can be tailored to the needs of other countries to help them establish a dialogue between the science and policy communities.
9. CONCLUSIONS Developing policy-relevant research requires the involvement of both scientists and decision makers in framing the appropriate questions. Policy users of the research results must articulate their information needs and consult scientists on the feasibility of research providing meaningful answers. Scientists can examine those initial requirements to determine the limitations and strengths of investigation and to meet them within available budgets and time frames. An iterative process between the users and producers of the information is desirable to refine and then periodicall~ update the policy-relevant research questions as both the science and policy evolve.
10. REFERENCES 1 2 3 4 5 6 7 8
Enabling the Future: Linking Science and Technology to Societal Goals, Carnegie Commission, Washington, 1992. Environmental Research and Development: Strengthening the Federal Infrastructure, Carnegie Commission, Washington, 1992. Report of the Task Force on the Health of Research, Committee on Science, Space, and Technology, U.S. House of Representatives 102nd Congress, Washington, 1992. J.C. Bernabo, Science and Policy: Notes from a Former Congressional Fellow, in Proceedings of the American Geophysical Union, EOS, 7 (1986) 82. A.M. Weinberg, Science and Trans-Science, in Minerva 10 (1972) 207. H. Brooks, Expertise and Politics: Problems and Tensions, in Proceedings of the American Philosophical Society, 119 (1975) 257. H. Brooks, The Resolution of Technically Intensive Public Policy Disputes, in Science and Human Values, 9 (1984). J.C. Bernabo and P. Eglinton (eds.), Final Report of the Joint Climate Project to Address Decision Makers' Uncertainties, EPRI Technical Document No. TR- 100772 (1992).
117 9
T.B. Carter and K.D. Smythe, Biodiversity Uncertainties and Research Needs: Interim Report, Science & Policy Associates, Washington, 1993. 10 J. Klabbers, P. Vellinga, et al., Policy Options Addressing the Greenhouse Effect, National Research Programme on Global Air Pollution and Climate Change, Bilthoven, The Netherlands, 1994. 11 S.P. Hammond, J.C. Bernabo, et al., Project Plan for Enhancing the Effectiveness of Research to Assist International Climate Change Policy Development, Science & Policy Associates, Washington, 1994.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
119
CLIMATE CHANGE, POLICY O P T I O N S A N D R E S E A R C H IMPLICATIONS P. Vellingaa, M. HisschemSllera, J.H.G. Klabbersb, M.M. Berkc, R.J. Swartc and A.P. van Uldend
a
Institute for Environmental Studies, Vrije Universiteit (IVM/VU), De Boelelaan 1115, 1081 HV Amsterdam, The Netherlands
b
Klabbers Management & Policy (KPMC), Oostervelden 59, 6681 WR Bemmel, The Netherlands
c National Institute for Public Health and Environment (RIVM), P.O. Box 1, 3720 BA Bilthoven, The Netherlands d
Royal Netherlands Meteorological Institute (KNMI), P.O. Box 201, 3730 AE De Bilt, The Netherlands
ABSTRACT Policy options for climate change and their research implications are discussed in this paper. Instead of science telling policy actors what to do, this project started by asking policy actors how they perceived the climate change issue, how it could best be handled and what science can do to address their questions and concerns. Through a series of interviews and workshops five different options have been constructed and for each a corresponding research agenda has been developed. I m p o r t a n t findings of this project are, 1) it is easier to reach consensus about actions t h a n about the nature of the problem; 2) issue linkage is crucial as the problem of climate change is complex and the benefits of emission limitation are too remote to be the single motivator for action; 3) an i m p o r t a n t condition for progress in climate change policy is a strengthening of the science-policy interface. This project is an illustration of how this can be achieved. 1.
INTRODUCTION
There are m a n y questions surrounding the climate issue. The project on 'Policy options addressing the greenhouse effect' took a look at whether and in what way our society can cope with the risks of climate change and the challenge of sustainable development. A dialogue was initiated between scientists and the broadest range of policy actors such as members of parliament, governmental
120 policy-makers, representatives of trade unions, employers' organisations, business, environmental and consumers' NGOs etc. This dialogue was organised in such a way that the risk of climate change and opportunities to deal with these risks were dealt with simultaneously. The project was set up to bridge the gap between the perceptions of policy makers, the private sector, societal organisations and scientists. Its objectives were to: * enhance the communication between the various parties; * make the various perceptions of the problem visible; * explore options for policy development; * introduce the options into the Dutch policy formation process; * identify the information needs and related research strategies per option and discuss the results in the research community. 1.2 Steps towards identification and exploration of policy options Theproject has been carried out in a number of steps as indicated in Figure 1. Step one is a serie~ of interviews, which led to three hundred questions relevant to the problem, formulated by policy actors. Through a workshop with the policy actors the questions were articulated and the number was reduced to 35. In step two these questions were assessed by natural and social scientists. Next, in step three the results were fed back to the policymakers, private sector and societal organisations (all relevant policy actors) through round-table discussions. The results generated through this process formed the basis of a workshop of all parties, including the scientists. This meeting developed options for Dutch (long-term) policy aiming at sustainability in terms of solutions to the climate question. A variety of options and related actions were formulated ranging from no-regrets to social-cultural change. As a follow-up, round-table discussions were held in which policy actors identified information needs and research strategies for each of the options to be addressed by the research community. This paper describes the process, the results of the various stages, the various policy options and the related set of research strategies.
~'dentification~~ sc'ientific~~ policy of issues assessment linking * interviews and workshop to identify major questions
* position papers and workshops with natural and social scientists
* round-table discussions linking science and policy
Figure 1. Subsequent steps in the research project
policy options * elaboration of policy options, related actions and research implications
121 2.
I N T E R V I E W S AND ELABORATION OF INTERVIEW R E S U L T S
Rather than science telling policy actors what to do, this project started by asking policy actors what they thought about the climate issue, how it could best be handled and what science could do to address their questions and concerns. The project primarily focused on the longer term policy options, with 2025 as time horizon. The people interviewed took the opportunity to blow off steam about politics and the way the issue is handled by various governmental departments and about the scientific community, which should produce "consistent signals instead of generating ambiguity and controversy". Some examples of statements that several times cropped up in the interviews are listed in Box 1. 2.1 The m e s s a g e : t w o - w a y c o m m u n i c a t i o n With the statements available, two clear messages were conveyed to the scientific community. The first was that (perceived) controversy among scientists on the nature of the problem is the main hindrance in convincing decision makers and the public at large of the urgency of emission control. Scientists tend to focus on remaining uncertainties, rather than communicating what is known and agreed upon. It was accepted that progress in science thrives on controversy. It was stressed, however, that the dissemination of results should not reinforce already existing confusion about the greenhouse effect, since this would weaken the initial commitment to action. The second message was that scientists should not restrict themselves to working in their institutes and communicating their results only through scientific publications. It was stressed that a precautionary approach can only be based on a broadly shared understanding of the nature of the problem and that this can only be gained through active communication by the scientific community with the public at large. Likewise, the societal and technological science community should not just perform disciplinary desk studies: they should carry out a number of demonstration projects in which their claims about the feasibility of certain solutions can be demonstrated to the decision makers and the general public. Finally, the interviews made it clear that knowledge of climate change is not the monopoly of the scientific community. Among the people interviewed there was a very broad and often detailed knowledge of the greenhouse effect and of the various response strategies. 2.2 One g o v e r n m e n t , different v o i c e s On the basis of the interviews, it was observed that the various governmental departments had positioned themselves in different phases of the development of climate change policies. One department (Economic Affairs) was still in the phase of conceptualising the policy issue (Is there a problem?); two were in the second phase: having accepted the risks of the problem, exploring response actions and building coalitions (the d e p a r t m e n t of Agriculture, Fisheries and N a t u r e M a n a g e m e n t and the department of Transport and Public Works. A fourth department (Environment) was in the next phase: efforts are made to implement policies; societal groups are persuaded and resources are allocated to solve the problem. As a consequence, the government did not speak with one voice. The resulting inconsistency in environmental policy development caused confusion amongst actors. Some considered the policy measures too weak, while others thought they lacked any basis and went too far. The following associated response
122 p a t t e r n s were identified: reactive (defensive response to g o v e r n m e n t policy), receptive (receptive response to government policy), constructive (acceptance of one's own responsibility) and pro-active (internalising in one's strategic planning). 2.3 I d e n t i f i c a t i o n of i s s u e s , i n t e r v i e w s a n d w o r k s h o p Questions put to the policy actors were: 1. "What are your perceptions of the greenhouse effect?" 2. "What potential impact may the issue have on your organisation, on both a short and long term basis?" 3. "How is your organisation responding to it?" 4. "Can the research community help to address your questions?"
Examples of statements from the first round of interviews * The greenhouse problem is probably the biggest environmental problem that we shall face in the twenty-first century. * Early action on emission reduction is imperative. * The m a i n challenge is how to get everybody on board for the far-reaching measures t h a t will be necessary. * Climate change is a non-issue, pushed by science and embraced by politics * If it would eventually prove to be a problem, then adaptation would be the best strategy. * Even if the Netherlands were to be in favour of reducing emissions, a unilateral Dutch policy would never succeed because of the global scale of the issue. * Strong emission reduction measures in the N e t h e r l a n d s would w e a k e n the industrial sector in international competition. Box 1. Examples of statements from the first round of interviews. The research team grouped the questions and s t a t e m e n t s t h a t were g a t h e r e d through the interviews. Next, the results were articulated through a workshop with the policy actors. 3.
SCIENCE ASSESSMENT
3.1 S c i e n t i f i c c o n t r o v e r s y d i s c u s s e d The next step in the project was a logical consequence of the previous one. Position papers were drawn up around the questions from the previous phase, and these were then discussed in two working conferences in order to provide a scientific assessment of the enhanced greenhouse effect. The first conference gathered experts from the n a t u r a l sciences and focused on biogeochemical cycles, responses of the climate system to changes in greenhouse gas concentrations and the effects of climate change. The basic question put to the n a t u r a l scientists was: does the greenhouse effect exist and, if so, w h a t are the risks? Some controversial positions were discussed and evaluated, including a
123 recent report written by professor BSttcher, an outspoken dissenter from the 'climate consensus' in the Netherlands. The second conference included experts from the social sciences and focused on technological solutions, economic aspects of Dutch greenhouse policy options and psychological, sociological as well as philosophical/ethical aspects. The m a i n question which had to be answered was: how do you perceive the climate problem and how can we cope with it from the socio-economic, technological and behavioural points of view?
3.2 S c i e n t i f i c a s s e s s m e n t of t h e g r e e n h o u s e e f f e c t P l e n a r y sessions during both gatherings resulted in confrontations between r e p r e s e n t a t i v e s of different disciplines. These confrontations resulted from 'language problems' and differences in underlying assumptions among the related disciplines. After these 'language problems' were solved and assumptions were thoroughly discussed, a consensus was reached on a total of 90 statements. A few of these are listed in Box 2. The n a t u r a l scientists (the first three s t a t e m e n t s in Box 2) discussed and evaluated the (un)certainties related to the dynamics of the climate system. The experts from the social sciences were inclined to accept the problem and tried to find ways and means to deal with it. This is reflected in some of their statements.
S e l e c t e d s t a t e m e n t s f r o m n a t u r a l a n d s o c i a l s c i e n t i s t s a b o u t the g r e e n h o u s e effect * The concentration of greenhouse gases, CO2, CH4, and N 2 0 , CFCs and tropospheric ozone have increased since 1960 as a result of h u m a n activities. * Experimenting with the global climate is not a feasible option. Risk assessment should include the possibility of irreversible changes to the climate. * The greenhouse effect is only one of man's disturbances of the t e r r e s t r i a l system. If possible, i n s t r u m e n t s to reduce the greenhouse effect should therefore also reduce other disturbances. * Estimations of the effects of a substantial reduction of greenhouse gases show t h a t it is possible t h a t macroeconomic effects will be relatively small and sectoral relatively large (employment, profitability and production). * Sustainable lifestyles should be promoted as positive changes. As a general rule, it can be concluded that lifestyle changes need to be based on the three A's: they m u s t be Achievable, Acceptable and Attractive. * Sustainable technology requires a shift in ideological, cultural and societal values within society. Ultimately, individuals will have to find new modes of behaviour within the limits of the ecospace. Box 2. Selected s t a t e m e n t s from the n a t u r a l and social scientists about the greenhouse effect
124 4.
FIVE POLICY OPTIONS
4.1 Linking science and policy In the third stage of the project, the results of the scientific assessment were fed back to the policy actors. In six round table discussions the results of the scientific assessment were linked with policy and institutional actors. Participants included representatives of several ministries, the chemical industry, the electricity sector, t r a n s p o r t and agricultural organisations, political parties, trade unions, environmental NGOs and consumer organisations. The round table discussions once again revealed the wide variety of perceptions about climate change. This part of the project showed that policy and institutional decision makers in general accept the scientific statements. The debate primarily focused on the societal aspects of a range of climate change policies. This implied a change in the perceptions of the different actors; at this stage there was more convergence than there had been at the beginning of the project (workshop 1). Differences of opinions were primarily related to the proposed policies for dealing with the problem. Six round table discussions produced five rough drafts of policy options for dealing with climate change. These were further developed in five working groups. The policy options are: * N o regrets: it is uncertain whether climate change will occur and, if so whether
substantial reductions in greenhouse gas emissions will be necessary; this implies no action regarding climate. * L e a s t regrets: climate change is a most serious problem with potentially
irreversable effects. Since the effects as yet are unclear, a risk approach should be taken. A trade-off is made between risks linked with intervention and non-intervention; this implies action now; the uncertainties are an important motivator for pre-cautionary measures. * A c c e l e r a t i o n : climate change is a serious problem, but too complex to address
head on; the climate problem can best be addressed through generally recognized short term problems in related fields; the focus is on issue linkage and on synergies and positive feedbacks presently existing in society; this implies action now but only in the context of other issues. * Technological innovation: climate change is a serious problem; but technological
development is the only way to match the demands of an ever increasing world population with the carrying capacity of the environment; this implies action focusing on technology, research, development, demonstration and diffusion. * I n s t i t u t i o n a l ~cultural change: within this option it is assumed that technological
solutions will not be sufficient to reach a sustainable society. Major societal, cultural and institutional changes are required to create a sustainable society; this implies action in all areas, not necessarily related to climate. In this project the policy options are defined not as a single type of action or instrument, but as the whole of opinions and suppositions about the climate issue, the effects and the possible solutions. The five policy options represent five different mainstream perceptions present in society. The options are typical views
125 which r e p r e s e n t a mixture of problem perception and solution perception. An i m p o r t a n t observation in this project is t h a t the perception of the 'problem' per actor is strongly linked with the perception of the 'solution'. For each of the five policy options a n u m b e r of actions were formulated. The interesting result of the project is t h a t there is much more agreement on the type of actions t h a n on the policy options. It was possible to identify eight different fields of action t h a t were mentioned under all policy options. The main difference between the options is the intensity and geographic scale of the implementation of the listed actions. The 'common' fields of action are indicated in Box 3.
Common actions * Towards an eco-tax system. * Low carbon transport systems/infrastructure. * Energy efficient housing/offices. * Redesign of industrial processes and products. * Towards renewable energy sources and renewable materials. * Joint implementation. * Towards closing the substance cycles at smallest geographical scale. * Stimulate technological and cultural innovation. Box 3. List of common actions. 5.
RESEARCH IMPLICATIONS
5.1 I n t r o d u c t i o n The next step in the project was to investigate the research implications of the various options. For each option a round-table conference was held. In these meetings the information needs and related research areas (including their focus) were identified. The r e s e a r c h implications are discussed in the following p a r a g r a p h s per policy option. This part of the project has not yet been finished. Nevertheless, a n u m b e r of observations with respect to information needs and research implications can tentatively be made. 5.2 N o - r e g r e t s Within the no-regrets option it is considered uncertain whether climate change will occur and, if so, whether substantial reductions in emissions of greenhouse gases will be necessary. Priority is given to instruments t h a t serve other (socio-economic and e n v i r o n m e n t a l ) objectives, s i m u l t a n e o u s l y r e s u l t i n g in a reduction of greenhouse gas emissions. No-regrets i n s t r u m e n t s will, irrespective of climate change, pay off anyway. Key words for this policy option are scarcity and real
politics. - Climate change is considered not to be a real problem. However, scarcity of fossil fuel resources is a problem. Moreover, policies t h a t address scarcity are likely to get much more support. Simultaneously, such policies lead to a reduction of greenhouse gas emissions.
126 In addition, greenhouse gas emission reductions in the Netherlands, including high cost for the Dutch economy, would not have a significant effect on global greenhouse gas concentrations, since the biggest countries in the world continue to grow both with respect to their populations and emissions. - Finally, there is no strong indication that the Dutch economy would significantly suffer from a changing climate. -
The following information needs and research areas are identified by the policy actors supporting the no-regrets view. 1. Scarcity of fossil fuel and mechanisms that can help to increase the efficiency of resource use are important research fields. Mechanisms and institutional arrangements that help to remove intersectoral barriers should be investigated (such as full cycle management). 2. Possibilities and impossibilities of demand side management should be investigated. What are the demands of the people and what are their priorities. Research should be carried out in the fields of price-elasticity and into the question of whether people at all are willing to adjust their consumption patterns on the basis of uncertain long-term changes in the climate system that may have both positive and negative effects. 3. Global, particularly Third World, growth of fossil fuel use and C O 2 emissions should be investigated. Special attention should be given to ways and means to increase energy efficiency in developing countries. A special point of interest is research into the leakage of greenhouse gases during exploitation and transportation of oil and gas (the total global quantity of flared natural gas is equal to the total European consumption of fuels by cars). 4. Population growth and ways to control this growth is an important area for research. 5. Regarding climate system research, priority should be given to monitoring and process analysis. 5.3 Least
regrets
Within the least regrets option climate change is perceived as a serious problem with potentially irreversable effects. As the effects are unclear a risk approach should be taken. A trade-off should be made between risks linked with the occurrence and non-occurrence of climate change in relation to the policies selected. The policies include all no-regrets instruments supplemented with anticipatory policies aimed at limitation of risks resulting from climate change. Policies anticipate a substantial reduction of greenhouse gas emissions. If reduction proves to be unnecessary, part of the effort is lost. If reduction proves necessary, further reductions will be more efficient than they would have been if only a no-regrets policy was implemented. The least regret option thus includes hedging strategies. From its inception, this option offers a long-term perspective. Keywords in this option are probability and insurance. - In this option climate change is recognized as a real problem. However it is not known how large the problem is. Therefore it is necessary to take immediate
127 action to reduce the risks. Such actions can be seen as an insurance premium. The least regrets option is seen as a rational policy based on a quantitative risk analysis. - The N e t h e r l a n d s as an energy and emissions intensive country has the responsibility to take actions according to its historic and present contribution and according to its economic and technical capabilities. The information needs and research areas as identified by the policy actors supporting the least regrets option are the following. 1. Action research in communicating the risks of climate change and the range of anticipatory actions is important. Especially the question of how all societal actors can be encouraged to implement a least regrets approach, should be addressed. 2. Regarding the climate system, the most i m p o r t a n t t a s k is to describe the uncertainties in terms of probabilities. In particular, the possibility of non-linear behaviour of the oceans and of the sources and sinks of greenhouse gases should be studied. 3. R e s e a r c h into the effects of climate change should not j u s t look at the N e t h e r l a n d s . The p r i m a r y focus should be on the global ecological and socio-economic systems such as ecosystems, food production and e x t r e m e events. Adaptation research should investigate the possibilities of decreasing the vulnerability and thus increasing the robustness of society, infrastructure and other socio-economic systems. 4. Research is also required to increase the understanding of the response capacity of society in r e l a t i o n to climate change scenarios including s u r p r i s e s . I n v e s t m e n t cycles and rates of m a r k e t penetration of new technologies need to be studied for a range of climate change scenarios. A special field of research concerns the possibilities and the potential of lifestyle changes. 5. Research to investigate and develop a range of options for drastic emission control: both in the field of rapid implementation of existing technologies as in the field of development of new technologies (renewable energy sources, sink e n h a n c e m e n t ) . Simultaneously, research should be carried out for drastic emission control through institutional changes such as new fiscal regimes and new i n t e r n a t i o n a l regimes such as joint i m p l e m e n t a t i o n and t r a n s f e r of technology. 5.4 A c c e l e r a t i o n The acceleration option focuses on synergies and positive feedbacks presently existing in society. Forces and currents t h a t are consistent with climate change policy are accelerated and barriers are removed. All policies should take into account the different time cycles of society. Key words in this option are opportunities and issue linkage. - Also in this option climate change is considered to be a real problem, but the government and intellectual elite is not capable to convince the major economic actors and society at large to substantially invest in emission control measures.
128 The only way to achieve something is "hitch hiking": riding with other issues. The direction to go is clear, but the (climate) vehicle does not have enough power on its own to get things moving so any opportunity that comes along should be grasped. - This option takes other environmental and societal problems as its point of departure. The climate issue is thus linked to other problems such as: e m p l o y m e n t , congestion, technology co-operation, u r b a n i s a t i o n , individualisation.. The following information needs and research areas are identified by the policy actors that adhere to the acceleration option. 1. Climate system research should focus on the relations between the greenhouse effect, the effects of aerosols, acidification, ozone formation, the effects of land-use changes and the various problems related to the human interference with the bio-geo-chemical cycles. Hence, climate research should be fully embedded in the global change research. Research should identify the common sources of a range of environmental problems. 2. Impacts and adaptation research should start by identifying which natural and socio-economic systems have the attention of the government and the public at large. Particularly those systems should be systematically studied for the potential of issue linkage. Research should focus on measures that help to reduce the vulnerability vis-a-vis climate change but that are originally envisaged for other reasons. 3. Similarly, the emission control type of research should start with an analyses of the various relations between climate change limitation measures and the range of technological, infrastructural, economic and other presently perceived problems that society wants/needs to address: urbanisation (housing, work, recreation and infrastructure), employment, transport/congestion, energy supply, communication, fiscal regimes, liberalisation of energy markets in Europe, agricultural problems and land surplus, world trade arrangements, development co-operation, population growth, international debt etc.. 4. Systematic research into issue linkages and development of strategies based on issue linkage, not just in the technical sense but also, perhaps even more so, in the communication domain. 5. Research into the question of how the various actions primarily driven by other issues, can be orchestrated in such a way that the total result with regard to climate is satisfactory. Research on how the climate change momentum can be maintained in a policy strategy that primarily focuses on other problems. 6. Systematic research into what people value, with the aim to identify those issues and actions that have sufficient support for implementation. 5.5 Technological
innovation
According to this strategy, technological development is the only way to match the demands of an ever increasing world population with the environment's carrying
129 capacity limited. This requires a long-term co-operation between government and private enterprise. In this option it is required that government plays a very active role in directing technological development by providing opportunities and constraints (e.g. subsidies, fiscal incentives, regulations, etc.) to stimulate the required development. The key word in this option is innovation. - The technological innovation option builds on (i) the assumption that there is sufficient technological creativity available in our society to address the climate change problem without loss of economic welfare and (ii) on the notion that technology development and implementation can be accelerated by removing existing barriers and creating positive incentives. The information needs and research areas identified by the policy actors supporting this option are the following. 1. The various relations between the climate problem and other effects of resource use should be investigated, as it is of crucial importance that new technologies address all problems and not just one single (climate) problem while increasing or creating other problems. Similarly, research into climate effects and adaptation should be relatively broad and linked with research in other environmental fields for the development of a robust technology strategy. 2. Research into the time dimension of emissions, concentrations and climate change impacts is very important for the implementation of the right technology at the right time and for minimization of the overall cost for society (e.g. a hundred year strategy for emission control and technology development). I n v e s t i g a t e the possibilities of bifurcation problems (one type of technology/infrastructure precluding the implementation of a better technology at a later stage). 3. Investigate and identify the conditions that are optimal for technology development and implementation (adoption and diffusion): e.g. market, infrastructural, institutional and cultural conditions. 4. Research and development programmes in the field of energy efficiency (supply and demand side, renewable energy, materials and redesign of industrial processes (the potential of a shift from non-renewable fossil resources to agro/biological renewable resources), and energy systems research (centralized versus decentralized systems and storage systems). The research programmes to be developed should be based on careful analyses of (potential) competitive advantages for the Netherlands. 5. Research with the aim to identify the sectors/technologies for which small incremental changes can produce large results, for example avoidance of leakages, process integrated energy efficiency schemes, more efficient cars, land use management systems etc.
5.6 Institutional/Cultural Change Within this option it is assumed that technological solutions will not be sufficient to create a sustainable society. Social, cultural and institutional changes are required
130 to reach such a goal. Furthermore, it is assumed that sustainable development can only be achieved through processes of change within society. The role of government is limited to setting conditions and providing support. Changes are promoted through support of concerted actions within society, mobilization of social organizations and the removal of institutional barriers. The focus within this option is on the achievement of the desired situation (positive motivation) r a t h e r than on the avoidance of a non-desired situation (negative motivation). Keyword in this option is quality. Quality of life is the central theme in this option, as opposed to quantity and speed. The following information needs and research areas were identified by the policy actors supporting this option. 1. With regard to climate change research, it was noted that some research will be neces-sary in order to illustrate the impact of wasteful h u m a n activities on the life support systems. However, it should be realized that the mechanisms that are advocated in climate research, such as ever larger computers trying to predict the inherently unpredictable and conventions to manage this are part and parcel of the same societal systems that are causing climate change such as ever increasing global trade and transport. Since closing the material cycles at the smallest possible geographical scale is the aim under this option, not much research would be needed to illustrate that large scale fossil fuel use and large scale landuse changes are detrimental for the environment and thus should be avoided. 2. Effects and adaptation research should focus on the relation between social, cultural, economic systems and the local climate. This relation is probably more important than presently realized, research should investigate this. 3. Research into the question of what a sustainable lifestyle looks like. 4. Research into technological innovation in support of sustainable lifestyles: technology supporting a local closure of material cycles (including carbon and nutrients). This includes research into the institutional, infrastructural and cultural systems that support a sustainable lifestyle. Research into the driving mechanisms for unsustainable production and consumption. 5. Research into the phenomenon of defensive/compensating consumption (skiing as required to compensate for stress work and intensive travelling; far away eco-tourism to compensate for lack of nearby n a t u r a l ecosystems t h a t are destroyed to make way for international airports; driving your children to school because the traffic is to dangerous for them to walk or to go on bicycle). 6. Research into the higher order effects of institutional and technological changes both to explain what has happened in the past, what is happening at present and what may happen in the future. 7. Action research and local experiments to investigate the feasibility of different lifestyles. The aim is to demonstrate that a large diversity of social/technical configurations are possible within the domain of sustainable life styles. The idea
131 of social learning and the idea of demonstration is important in the design of these experiments. P a r t of this type of research should also be how such experiments can be encouraged through generic i n s t r u m e n t s or removal of (generic) b a r r i e r s (including research into w h a t we m a y learn from earlier idealistic/utopian movements and their ideals, including research into the role of elite behaviour as a change agent, and research in the role of examples as media/agents communicating the necessity of change). 6. C O N C L U S I O N S AND R E C O M M E N D A T I O N S The evaluation of the results of this research project has not yet been finalized. Still a number of conclusions can be drawn. One conclusion came forward rather clear: the interviews and the analysis indicate that it will be very difficult, if not impossible, to reach concensus on the nature and the seriousness of climate change. It is not likely t h a t one of the five options discussed above will emerge as a concensus option as the way the risk of climate change is perceived appears to depend strongly on the societal and economic interests of the policy actors and on the individual values. For example, all participants from the energy intensive industries were convinced t h a t IPCC is trying to fool t h e m and t h a t some politicians join this game out of publicity interests. They typically favour the no-regrets option. The majority of the actors from the private sector, with interests that are relatively neutral vis-&-vis climate change policy, are in favour of either a least regrets, an acceleration or a technological innovation policy. A small part of the policy actors are in favour of the socio-cultural change option. The conclusions and recommendations that can be formulated at this phase of the project are listed below. 1. Based on the results of the project the researchers believe t h a t it is more fruitful to seek consensus about actions t h a n to seek consensus about the nature of the climate issue. 2. For all p a r t i c i p a n t s technology is an i m p o r t a n t and valued p a r a m e t e r in addressing the climate issue. However, preferences about the type of technology and the role of technology in society may differ. Still, all participants, also the ones favouring social-cultural change, favour technological research. All the promoters of technology agree that the societal needs and concerns should play the major role in technology development. 3. With regard to the climate problem, it appears as if most of the actors are fairly well informed about the n a t u r e of the problem, although the interpretation of the information differs. It seems as if society as a whole is waiting for the scientists to come with the ultimate answer about the risks. However, for m a n y of the scientists the expected response of the climate s y t s t e m and the uncertainties involved are a major reason for action. 4. A large p a r t of the policy actors implicitly favour some kind of acceleration
132 policy. Issue linkage as a basis for common action looks like a promising approach to climate change policy. However, this reveals a paradox. The general thrust of the acceleration option is that society is not willing or not ready to address the climate issue in its own right. This is expressed by the statement often made that climate change measures should piggy back on other issues. The paradox is that other issues, for example fossil fuel scarcity, usually are not perceived as sufficiently urgent to generate effective and long-term policies. If the climate issue had not been raised, all energy efficiency programmes would have been stopped in the late 1980's. It is only because of the climate issue that combined heat and power could become a success in the Netherlands. Consequently, in order to develop and implement an effective strategy, existing problems and concerns in the society other than climate change, may need to be the starting point for greenhouse gas control policies. Nevertheless, the climate issue needs to be raised continuously to keep the action going. So climate change may serve as a long-term argument for change, while short-term problems and concerns are the practical, day to day motivator for business. It should be realized that both problems are real and both arguments are needed for coherent and long-term action. This implies that the framework convention for climate change should be regarded as a meta-level policy driver, whereas more sectoral agreements should serve as implementation agents of the broader goal of limiting greenhouse gas emissions. 5. Perhaps this also holds for the research agenda: climate change research and the related research for climate change policy, play an important role as an overarching and integrating element in many fields of research. Still, their role in guiding the overall research may not be sufficiently addressed yet. The overall agenda should probably focus on global environmental change including biodiversity, land use, resource use (energy, water, etc.) and demographic issues, while climate change research should be embedded in such a global change research agenda. 6. The project has revealed that crucial societal actors may have different perceptions about climate change and the desirability of certain responses. As a consequence, they also have different information needs. In order to adequately address these needs, it is important to involve various policy actors in the development and evaluation of climate change research programmes. In this respect it is recommended to give more attention to the science-society dialogue. 7. Finally, the project was evaluated as particularly fruitful for the scientists who participated in the workshops. Many of them had never before exchanged ideas beyond existing disciplinary boundaries. This project revealed that scientists from different disciplines may come to opposite conclusions with regard to the feasibility of controlling greenhouse gas emissions. In most cases it was apparent that it is not the scientific elaborations that cause the differences, but the differences in paradigms and related assumptions that are taken for granted a priory. It is therefore recommended to stimulate multidisciplinary dialogues as a formal part of climate and global change research programmes. The effectiveness of climate change research can particularly be enhanced when also the policy actors are involved in the multidisciplinary dialogue.
133 7. D I S C U S S I O N ON " D E V E L O P M E N T OF (POTENTIAL) P O L I C Y O P T I O N S IN T H E N E T H E R L A N D S " Rapporteur: P.A. Boot Ministry of Economic Affairs, P.O. Box 20101, 2500 EC Den Haag, The Netherlands This research project did have four goals: 1) to enhance the communication between policy makers, third parties and scientists; 2) to identity and explore a range of options; 3) to input these options to the Dutch policy making process and 4) to generate a series of questions and concerns for the second phase of NRP. The activities of the project took place in a dialogue between policy and research, between policy makers themselves and researchers of different disciplines. It was organised as follows. In interviews, policy makers were asked to identify the problem. This resulted in some 300 questions for scientists. In workshop-I these were reformulated into 50 questions. Together with several position papers from scientists these questions were input into two workshops: workshop II A for the n a t u r a l scientists and workshop I I B for the social scientists and technological experts. The natural scientists felt obliged to re-arrange some of the questions to bring them more in agreement with their views. They reached consensus on 41 s t a t e m e n t s about observations on and possible effects of changes in the climate system. In the workshop of the social scientists (IIB), it proved to be more difficult to create understanding among sociologists, economists and philosophers, because of different presuppositions and disciplinary frames of reference. Once they reached consensus on the assumptions underlying the statements, they were able to formulate 49 s t a t e m e n t s conveying their joint views on causes, impacts and solutions. However, a straightforward scientific assessment could not be given and was even considered to be impossible. The statements underlined the necessity to strive for a common reference framework. The results of the workshops II A and B were fed back to policy makers and third parties during Round Table meetings. During this stage rough drafts of five policy options were formulated, which served as input for the final workshop III. This workshop did not aim at recommendations, but at the formulation of internally consistent policy options. Different perspectives on climate change response strategies were grouped into five options: 1. No regrets. In this option priority is given to i n s t r u m e n t s t h a t serve other objectives and which simultaneously result in reduction of greenhouse gas emissions. 'Climate' is no issue, and even if it eventually would prove to be a problem, then adaptation would be the best strategy. Because of the eventual scarcity of fossil fuel reserves, energy efficiency is worthwile to strive for, however. Keywords: Realism, Scarcity. 2. Least regrets. A climate problem exists, but what it looks like is uncertain. We have to m a n a g e risks. Policy has to anticipate substantial but no absolute reduction and has to provide a long-term perspective. Science m u s t support action. Keywords: Probability, Insurance.
134
3. Acceleration. Climate problems are tackled by linking it with other issues. Policy has to focus on synergies. Measures developed to contribute to other environmental and societal problems should be strenghtened in order to address climate change. Climate problems are used to solve other problems. Acceleration might be called an active form of hitch hiking or active version of no-regrets, as the climate problem is accepted as a problem. Keyword: Opportunities, Issue linkage. 4. Technological innovations. This option is based on the a s s u m p t i o n t h a t technology is the only way to match the demands of an ever increasing world population with the carrying capacity of the environment. A long-term cooperation between g o v e r n m e n t s and private e n t e r p r i s e s is required. Governments have to provide opportunities and remove constraints in order to enhance R, D & D and implementation. However, innovation itself comes from the private sector. Keyword: Innovation. 5. Institutional and cultural change. Technological solutions will not be sufficient to create a sustainable society. F u n d a m e n t a l changes are required. Societal organizations have a potential that has to be fully mobilized. Keyword: Society, Quality of life. Each of these options is most cost-effective in its own way of thinking. It might be impossible to strive for a narrowing of the range of these basic beliefs, but a promising similarity in actions may be observed t h a t are proposed by the representatives of different options. These possible actions range from low carbon t r a n s p o r t s y s t e m s and energy efficient buildings to ecotaxes and j o i n t implementation. Of course there is a difference in timing and scope of the proposed actions, but the fundamentals differ less with regard to actions that are considered useful than with regard to basic beliefs. A continuing improvement in inter- and intradisciplinary communication among scientists and policy makers is necessary to look for a more solid foundation for action. The way forward leads to utilization of shared opportunities, more than to a possibly fruitless search for consensus in basic beliefs. However, scientists and policy m a k e r s should more actively communicate those viewpoints they agree upon. The general idea in the workshop was that this important project should be known better internationally. Ideas on an international 'repetition' with representatives from different groups of countries exist. Learning-by-doing might be a fruitful way of constructing knowledge for both scientists and policy makers.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
STABILIZING GREENHOUSE CONSEQUENCES
GASES:
GLOBAL
135
AND
REGIONAL
Joseph Alcamo*, Maarten Krol, Rik Leemans E n v i r o n m e n t a l Forecasting Bureau, National Institute of Public Health and the Environment, P.O. Box 1, 3720 BA BILTHOVEN, The Netherlands *With Contributions From: Andr~ van Amstel, Johannes Bollen, Gert J a n van den Born, Alex Bouwman, Kees Klein Goldewijk, Eric Kreileman, Jelle van Minnen, Jos Olivier, Sander Toet, Bert de Vries, G~ Zuidema ABSTRACT This p a p e r assesses the environmental consequences of two t a r g e t s for C02 stabilization: 350 ppm by the year 2150 (367 ppm by 2100), and 450 ppm by 2100. As a tool for this investigation we use the IMAGE 2 integrated model of climate change. It was found t h a t these targets lead to much lower regional impacts on crop productivity, natural vegetation, and sea level rise as compared to the baseline case. Nevertheless some negative impacts do occur, and to further reduce these impacts would require more stringent stabilization targets. It was also found t h a t to achieve these stabilization targets in the atmosphere, global emissions should not s u b s t a n t i a l l y increase at any time in the future, and eventually they must be significantly reduced. 1. I N T R O D U C T I O N Article 2 of the F r a m e w o r k Convention on Climate Change proclaims the goal of achieving "stabilization of greenhouse gas concentrations in the atmosphere t h a t would prevent dangerous anthropogenic interference with the climate system." The purpose of this brief report is to review some of the consequences of two scenarios for stabilizing greenhouse gas concentrations. It is thought t h a t this information can be used in the process of selecting i n t e r n a t i o n a l policies for complying with the objectives of the Convention. Our analysis concentrates on two scenarios in p a r t i c u l a r because they have been adopted for study by Working Group I of the IPCC, as will be explained later. Our analysis draws on results of the IMAGE 2.0 model, an integrated model of climate change and the global environment1. Information about IMAGE 2.0 is given in Appendix 1. 2. W H A T W I L L H A P P E N IF NO A C T I O N IS TAKEN? In order to evaluate scenarios for stabilizing greenhouse gases, a baseline is needed for comparison. Our baseline scenario uses i n t e r m e d i a t e a s s u m p t i o n s about
136 population and economic growth.2 We note that this is not meant to be a "most likely" scenario. This scenario also assumes that no actions are taken to mitigate climate change; this allows us to estimate the possible incremental improvements that could come from stabilization versus no action. Under baseline (i.e. no action) conditions, the IMAGE 2.0 model computes that by 2100 global CO2 emissions could reach 24 Gt C/yr (within the range of IPCC emission scenarios3) and global average CO2 concentration 777 ppm. At the same time global average surface temperature could increase by 2.50C between the years 1990 and 2100. During the same period, temperatures increase by about 1.80C in the tropics and around 3 to 50C in the high latitude regions (Figure 1). Such changes to temperature and precipitation could lead to a variety of impacts. We focus on three in this p a p e r - - changes in crop productivity, disturbance of natural vegetation patterns, and sea level rise. These were selected because they are related to risks to food production, ecosystems, and economic development, which are the three impacts specifically mentioned in Article 2 of the Convention. 8
7~
" - - Base
/
-'"
/
$450
6
/
$350
.x:. ~5
~4
\
~3
/
...-"-
E2
1
,,.,
I : ;""
0" -90
. . . .. .. .. .
~'fJ1.,,''"
-60
South pole
"~i'-" "','",", .--: ..... ;
,
,
,"
-30
0
degrees latitude
30
60
90
North pole
Figure 1. Zonal Temperature Increase. The increase in surface air temperature computed by the IMAGE 2.0 model. The average temperature increase in 100C latitudinal zonal bands is given. Crop Productivity. As a result of baseline changes in climate, large portions of currently cultivated areas could experience reductions in crop yields. As one example, the potential rainfed productivity of wheat could be substantially decreased in 32% of current wheat growing areas between the years 1990 and 2100 (Figures 2a and 3a). During the same period, millet productivity could be substantially reduced in 37% of current millet growing areas (Figures 2b and 3b).
137
~ IncreasinYigeld
~~j StableYield
II
Decreasing
Yield
Figure 2 A and B. Changes in Crop Yield of Current Crop Growing Areas According to Baseline Scenario: (a) Wheat, (b) Millet. Shown are "substantial" decreases or increases in the potential rainfed productivity of w h e a t and millet over the period of the simulation, 1990 to 2100. Substantial is defined as follows: For wheat -- Substantial is taken as an increase or decrease of 0.5 t/ha/yr or more. This amounts to a roughly 10% change in the c u r r e n t level of potential rainfed productivity in c u r r e n t wheat-growing areas. For comparison, the current net yield of w h e a t is s u b s t a n t i a l l y lower -- 2.6 t/ha/yr, globally averaged. (Agrostat PC, FAO, Rome, Computerized Information Series no 1, October, 1992). Note that impacts on only current wheat growing areas are shown. New areas might be become productive for w h e a t u n d e r climate change. This is of course a very limited definition of risk to wheat growing areas, but does indicate where there is increased risk to production in current areas. For millet-- Substantial is t a k e n as an increase or decrease of 0.25 t/ha or more. This threshold is set lower t h a n wheat because millet is grown more often t h a n wheat by subsistence farmers who obtain low net yields. Indeed the current net yield of millet (0.8 t/ha/yr globally averaged, FAO, 1992, op cit.) is substantially lower t h a n t h a t of wheat. Hence, a smaller change in potential productivity for millet is of importance. It should be noted t h a t these calculations do not take into account the possible CO2 fertilization effect which could increase future yield estimates.
138 During the same period, millet productivity could be substantially reduced in 37% of current millet growing areas (Figures 2b and 3b). On the other hand, potential yield may increase in other areas, although this will not necessarily compensate for the disruption in yield elsewhere. The main areas affected would be current wheat growing areas in China, Western Europe, and parts of North America; and millet growing areas of Africa, the Middle East, India, and China (Figure 2). Area with decreasing yield of wheat
35
Base - - ' " S450 - - - -
,~ 3o~ 25c ~ 20 ~
....
iiiiiiiiiiiiii
- - - -
,,.:......
15-
..
..............................
j.J~'" I0-
2
,P~ "
5-
0
1o.
~ooo
~o~
~o~o
~o,~
~1oo
time in years
40
3s-
Area with decreasing yield of millet -----
30-
Base $450 S350
~ 25-
2o"1"t "
r lO-
t. ~ j~ ,t"o."
o
1990 2000
o................................
2025
20150
2075
2100
time in years
Figure 3. Changes in Cultivated Areas Affected by Decreasing Yields" (a) Wheat, (b) Millet. Shown are the currently cultivated areas with "substantial" decreases in potential rainfed productivity over the period of simulation, 1990-2100. As in Figure 2, "substantial" is taken as a decrease of 0.5 t/ha/yr or more of wheat, and 0.25 t/ha/yr of millet.
Natural vegetation. IMAGE 2.0 calculates global potential vegetation patterns by determining the occurrence of different plant types such as needle and broadleaved trees, shrubs and grasses. Each plant type has it typical distribution as a response to local climate and soil characteristics. Using this approach it was estimated that the baseline climate change would change the potential vegetation
139 in 42% of t h e w o r l d ' s l a n d a r e a by t h e y e a r 2100 (fig. 4), a n d in 44% of its c u r r e n t n a t u r e r e s e r v e a r e a s (Table 1). C o n s e q u e n t l y , t h e c u r r e n t n a t u r a l v e g e t a t i o n in t h e s e a r e a s will n o t be well a d a p t e d to t h e s e c h a n g e d c l i m a t e conditions. C h a n g e s of v e g e t a t i o n a t s u c h a l a r g e s c a l e could l e a d to s e v e r e d i s r u p t i o n of n a t u r a l vegetation succession, the main process through which vegetation can respond a n d a d a p t to n e w conditions. T h e s e c h a n g e s will t h e r e f o r e i m p a c t s t r o n g l y on local a n d r e g i o n a l biodiversity.
•
Soco i-Econom Facictors
I
ClimatC ehange
I
Combn ied
Figure 4. Threat to Natural Vegetation According to Baseline Scenario (1990-2100). Changes in natural land cover stemming from two main factors (i) socio-economic, (2) climate change. "Socio-economic" refers to current areas of natural vegetation that may be used for new agricultural land or fuelwood to satisfy the future food and fuel demands of the baseline scenario. These agricultural demand and land cover calculations are described in: (i) Alcamo, J., van den Born, G.J., Bouwman, A.F., de Haan, B., Klein Goldewijk, K., Klepper, O., Leemans, R., Olivier, J.A., de Vries, B., van der Woerd, H. and van den Wijngaard, R., 1994. Modeling the global society-biosphere-climate system, Part 2: computed scenarios. Water, Air and Soil Pollution, 76: 37-78, and (ii) Zuidema, G., van den Born, G.J., Alcamo, J. and Kreileman, G.J.J., 1994. Simulating changes in global land cover as affected by economic and climatic factors. Water, Air and Soil Pollution, 76: 163-198. "Climate change" refers to areas in which the potential vegetation is estimated to change because of climate change. The potential vegetation calculations employ a global vegetation model, "BIOME", described in: Prentice, I.C., Cramer, W., Harrison, S.P., Leemans, R., Monserud, R.A. and Solomon, A.M., 1992. A global biome model based on plant physiology and dominance, soil properties and climate. Journal of Biogeography, 19: 117-134. The model BIOME is embedded in IMAGE 2.0 as described in: Leemans, R. and van den Born, G.J., 1994, Determining the potential global distribution of natural vegetation, crops and agricultural productivity. Water, Air and Soil Pollution, 76: 133-161.
140 Climate is not the only factor t h a t will threaten natural vegetation patterns, and thus biodiversity. Another major factor will be the expansion of agricultural land stemming from population and economic growth (which will occur despite the more intensive use of current agricultural land). This is taken into account by IMAGE 2 in all scenarios (Figure 4). According to the baseline scenario, 23% of the world's c u r r e n t n a t u r e reserve areas m a y be t h r e a t e n e d by a g r i c u l t u r a l expansion between 1990 and 2100 (Figure 4). Also according to baseline calculations, 12% of the world's n a t u r a l reserve areas may be threatened by both climate change and agricultural expansion during this period. This includes large areas of Africa and Asia. The main point is that there is a close connection between policies to address climate change, world food production, and land use, and they will have overlapping effects on the world's natural vegetation cover and its level of biodiversity. Threat to natural vegetation in nature reserves due to climate change
-_.: s4Baseo
4540"13 r
35-
r
30-
t,_ r
(D 0
2520-
O) 1 5 t-
ml00!,-
0.. 5 0
I
1990 2000
I
I
I
2025
2050
2075
21 O0
time in years
Figure 5. Area of Nature Reserves Affected by Climate Change. Shown is the area of n a t u r e reserves where potential vegetation changes because of climate change. Calculations are performed as in Figure 4 and are then overlayed with the current location and area of nature reserves.
Sea Level Rise. Another consequence of not acting to mitigate climate change will be sea level rise due to melting of glaciers and ice caps, and thermal expansion of sea water. By year 2100 sea level is computed to be 20 to 60 cm higher t h a n in 1990, depending on location (Figure 6). Much of South Asia's coastline m a y experience a sea level rise of between 25 to 30 cm. Island states in the Caribbean could experience a sea level rise of the same magnitude, and those in the South Pacific between 20 to 25 cm (Figure 6).
141
m40.0.6o.0 30,0.40.0 ~20.0.3o.0
----~:~ ~::~ii ............. iiii~~ !7::
/~.0.
~.0
i~.o.~, o
,+~71~
~b,~:,,L
I
~:~ii::ii~i~7~iii~: i!
.':'.:~'. .;:.i. .i.iT.ii:i:i:i:i:.7..ii,lii~,:,~/~ iii!L::i~ii:i:ii!!!il;.....:::~i~.~:i.i.:.:.:~i . .i::: ........'~il~:~7!ii~
i i~i:::Li:iTii::ii,,,,,:~,:~:,,:i,,i::::}% .....
,:
~20.O -30.0 Figure 6 A and B. Regional Sea Level Rise Between 1990 and 2100 Corresponding to Baseline Scenario: (a) North America, (b) Asia and Oceania. Shown are mean increases in sea level over the period 1990 to 2100. These calculations take into account melting of ice caps, glaciers, and regional differences in the t h e r m a l expansion of sea water. They do not take into account the shifting of ocean currents nor differences in coastal wind velocity t h a t may accompany climate change.
142 2. W H A T
STABILIZATION
SCENARIOS
ARE CONSIDERED?
One way to mitigate climate change would be to stabilize the levels of CO2 and other greenhouse gases in the atmosphere. Results from the IMAGE 2.0 model show t h a t this could be an effective approach, depending on the target level and date of stabilization. In this p a p e r we examine two t a r g e t scenarios of stabilization: 9 CO2 stabilized at 350 ppm in 2150 (reaching 367 ppm in 2100). 9 CO2 stabilized at 450 ppm in 2100. These scenarios are of interest from the policy standpoint because CO2 would stabilize at about its current level (around 354 ppm), or moderately above this level. These scenarios were also p a r t of an i n t e r n a t i o n a l modeling exercise sponsored by Working Group I of the IPCC.4 For both scenarios, the atmospheric levels of CO2 are assumed to follow a smooth pathway from 1990 to their future t a r g e t date and concentration. Other greenhouse gases are also a s s u m e d to stabilize within this time frame.5 Global sea level rise
25-
///"
---- Base ----$450 9 9 9 S350
/
/
20-
t"
E 15.o. .I
._~ -6 lO-
j" .I"
j
.d"
t
I
o.
."
.t"
I .o~
.."
.,"
."
,"
eb 5-
0 1990 2000
~o~
~oso
~o~
~,oo
time in years
Figure 7. Global Average Sea Level Rise. Same calculations as in Figure 6, but averaged for the globe.
3. H O W
EFFECTIVE
ARE THE STABILIZATION
SCENARIOS?
Because of the unavoidable uncertainties of these model estimates, it is more informative to examine the relative differences between the baseline and the stabilization scenarios (Figures 1, 3, 5, 7, 8 and Table 1) r a t h e r t h a n their exact numbers. The stabilization scenarios have the following effects: 9 Regional t e m p e r a t u r e increases are subtantially smaller t h a n the baseline scenario. 9 The crop and n a t u r a l vegetation areas affected by climate change do not increase after 2050, whereas they do in the baseline scenario. 9 The total amount of area affected by climate change is significantly lower t h a n in the baseline scenario.
143 9 Sea level continues to rise beyond 2100 despite CO2 stabilization (it also does in the baseline scenario). This is because of the slow response time of the atmosphere-ocean system. 9 However, the r a t e of sea level rise is much lower than in the baseline scenario. These results show t h a t the stabilization scenarios have an overall lower negative impact t h a n the baseline. However, they also show t h a t they are not "risk-free". Impacts still occur because it takes several decades to stabilize greenhouse gases in the atmosphere, and in the meantime climate change occurs. To further reduce these impacts it would be necessary to adopt even more stringent stabilization targets.13 4. E M I S S I O N L E V E L S TO A C H I E V E STABILIZATION OF C O 2
A key question is how to achieve the stabilization of CO2 and other greenhouse gases in the atmosphere. Specifically, w h a t level of emissions would be allowed from the world's energy and industrial system? After accounting for the uptake of CO2 by vegetation and the ocean, this has been e s t i m a t e d by several global models as p a r t of an IPCC Working Group I exercise6. Results from the IMAGE 2.0 model are shown in Figure 8, and are consistent with results from other models7: 9 In order to stabilize CO2 levels by 2150 at 350 ppm, it will be necessary to i m m e d i a t e l y stabilize and t h e n s h a r p l y reduce global e n e r g y / i n d u s t r i a l emissions towards the end of the 21st century. 9 For the a l t e r n a t i v e scenario of stabilizing CO2 at 450 ppm by 2100, global energy/industrial emissions will be allowed to increase slightly above current levels, and t h e n m u s t be significantly reduced after the middle of the next century. Global CO2 emissions
25
Base
-.-s45o .
~ O15-
...s3
o
~ ~ . . . .
~ " ~ ....
/
/
I
I
I -- _|: ___ I
ffl e(~.~E ~s 110
1
~ .....
~ ~ " ~ ~ ~ " ""
~ ~ 9 ~
~ !
o
I
~176176
~176176 ~176176176
time in years
Figure 8. Allowable Global Emissions from E n e r g y / I n d u s t r y to Achieve CO2 Stabilization Goals. Shown are global CO2 emissions from energy and i n d u s t r y only (land use emissions are not included).
144
Put another way, large increases in emissions would be unacceptable at any time for either scenario. This is an i m p o r t a n t point because the allowable global emissions for stabilizing CO2 and other greenhouse gases are far lower t h a n baseline emissions (Figure 8). In the absence of policy measures, emissions are expected to sharply increase along with economic development in developing c o u n t r i e s 8 . Hence t h e r e exists a large "policy gap" between the allowable emissions for stabilizing greenhouse gases, and the emissions that will occur if no action is taken. The last issue to be raised in this report is whether emission strategies can be found to achieve the stabilization scenarios. It is possible t h a t some proposed global energy scenarios, for example from Johannson, et al.9, Shell 10, or Working Group II of the IPCCll, produce emissions low enough to achieve the stabilization scenarios. This is a key unresolved issue that needs to be resolved by the research community and reported to policymakers.12 5. S U M M I N G U P This brief p a p e r highlights some of the consequences of two scenarios for stabilizing greenhouse gases: (i) CO2 stabilized at 350 ppm in 2150 (367 ppm by 2100), (ii) CO2 stabilized at 450 ppm in 2100. Among its main findings: 9 To achieve t h e s e s t a b i l i z a t i o n ~ t a r g e t s , emissions are not allowed to substantially increase at any time, and eventually they m u s t be significantly reduced. : 9 Because of the current upward trend in global emissions, there is a large policy gap between the allowable emissions for stabilizing greenhouse gases, and the emissions that will occur if no action is taken. 9 Stabilization scenarios lead to much lower impacts on crop productivity, natural vegetation, and sea level rise as compared to the baseline case. 9 Although the stabilization scenarios show lower impacts t h a n a baseline, they are not "risk-free". Some impacts do occur, and to further reduce these impacts would require more stringent stabilization targets. 9 With regards to threats to natural vegetation and biodiversity, there is a strong need to connect policies that address climate change, world food production, and land use.
6. A C K N O W L E D G E M E N T S
The IMAGE Project is supported by the Dutch Ministry of Housing, Spatial P l a n n i n g and the E n v i r o n m e n t (VROM), and the Dutch National Research P r o g r a m m e on Global Air Pollution and Climate Change (NRP). This paper was p a r t l y funded u n d e r NRP contracts 853129, 853130, 853131, and 853132. An earlier version of this paper was prepared as a background report for the Dutch Delegation to the F i r s t Session of the Conference of P a r t i e s to the U.N. F r a m e w o r k Convention on Climate Change, Berlin, Germany 28 March - 4 April, 1995. Authors are grateful to M.M. Berk, B. Liibkert-Alcamo, B. Metz and R.J. Swart for reviewing this manuscript.
145 A p p e n d i x 1. O v e r v i e w of t h e IMAGE 2.0 m o d e l ES~ROV- ~NDVSrRY S~VmH ii .... i .......... i i.......... ~
i ,.,-=., L__J
J
EL
ouu, A r ~ s p , E~.~:~AN sYstEM
Input
' IP | " G~oo~ho~o G ~ I
"-* ~A
!
.oooo;
h " ~X,~ Zonal |
I
L gases
cl
.
L.ControlPolicies
.J
=~
TERRESTRIAL ENV1RONMENT SYSTEM
J
~~lj aiospher e
.O~an
"GreenhouseGase~
- J "Ozone-.re/ated | L gases ___j
! !X \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \
UF~ba=s
J
i!.;;,~.......~......
La,d u= ! .........:~"~ .']~ ~'::
rEm'~~
]
/ "~176 L gases
/ J
"~ r
NewLandUse
LPattern
F Ao~u,u~
I Impac~ / Ecosystemand
-I
LOther Risks
Figure 9. Box Diagram of the IMAGN 2.0 Model. Each box represents a submode]. The IMAGE 2.0 model is a multi-disciplinary, integrated model designed to simulate the dynamics of the global society-biosphere-climate system. The objectives of the model are to investigate linkages and feedbacks in the system, and to evaluate consequences of climate policies. Dynamic calculations are performed from year 1970 to 2100, with a spatial scale ranging from grid (0.50 x 0.50 latitude-longitude) to world regional level, depending on the sub-model. The model consists of three fully linked subsystems: Energy-Industry, Terrestrial Environment, and Atmosphere-Ocean. The Energy-Industry models compute the emissions of greenhouse gases in 13 world regions as a function of energy consumption and industrial production. End use energy consumption is computed from various economic/demographic driving forces. The Terrestrial Environment models simulate the changes in global land cover on a grid-scale based on climatic and economic factors, and the flux of carbon dioxide and other greenhouse gases from the biosphere to the atmosphere. The Atmosphere-Ocean models compute the buildup of greenhouse gases in the atmosphere and the resulting zonal-average temperature and precipitation patterns. The fully linked model has been tested against data from 1970 to 1990, and after calibration can reproduce the following observed trends: regional energy consumption and energy-related emissions, terrestrial flux of carbon dioxide and
146 emissions of greenhouse gases, concentrations of greenhouse gases in the atmosphere, and t r a n s f o r m a t i o n of land cover. The model can also s i m u l a t e current zonal average surface and vertical temperatures. For f u r t h e r i n f o r m a t i o n consult: Alcamo, J. (Editor), 1994. I M A G E 2.0: Integrated Modeling of Global Climate Change. Kluwer Academic Publishers, Dordrecht, 318 pp. Also published as Special Issue of Water, Air and Soil Pollution, 1994. Volume 76, Nos 1-2.
147
Table 1. Regional I m p a c t s of Climate Change. Region !
Area With Substantial D e c r e a s e in W h e a t Yielda
Area With S u b s t a n t i a l Decrease in Millet Yieldb (% of C u r r e n t Millet Areas)
(% of C u r r e n t W h e a t Ares)
Areas of N a t u r e Reserves With C h a n g e in Potential Vegetation Due to Climate Changec (% of C u r r e n t N a t u r e R e s e r v e Areas)
1
2
3
1
2
3
1
2
3
Global
32
24
16
37
21
8
44
17
9
Canada
68
62
48
-
-
-
66
34
16
USA
59
52
43
65
52
33
68
31
16
Latin America
26
6
2
53
27
1
32
13
9
Africa
16
11
3
46
19
3
42
12
7
OECD Europe
46
37
27
38
38
38
67
38
26
Eastern Europe
35
31
21
43
44
30
58
43
27
CIS
10
9
5
18
16
11
48
23
9
Middle East
14
11
6
6
4
1
16
6
3
India + S. Asia
2
1
1
40
15
2
43
9
4
China + C.P. Asia
43
21
3
22
20
12
37
18
10
E a s t Asia
-
-
-
4
2
0
31
12
8
45
20
11
52
46
3
56
14
8
Oceania
N O T E S for TABLE 1 = Baseline Scenario 2 = Stabilization of CO2 at 450 ppm in 2100. 3 = Stabilization of CO2 at 350 ppm in 2150. a I n d i c a t e s t h e a r e a s t h a t e x p e r i e n c e a " s u b s t a n t i a l " d e c r e a s e in t h e p o t e n t i a l r a i n f e d p r o d u c t i v i t y of w h e a t over t h e period of t h e s i m u l a t i o n , 1990 to 2100. S u b s t a n t i a l is defined as a d e c r e a s e of 0.5 t/haJyr or more. This a m o u n t s to a roughly 10% change in the current level of potential rainfed productivity in c u r r e n t
148 wheat-growing areas. For comparison, the current n e t yield of wheat is substantially lower-- 2.6 t/ha/yr, globally averaged. (Agrostat PC, FAO, Rome, Computerized Information Series no 1, October, 1992). Note that impacts on only current wheat growing areas are shown. New areas might be become productive for wheat under climate change. bIndicates the areas t h a t experience a "substantial" decrease in the potential rainfed productivity of millet over the period of the simulation, 1990 to 2100. Substantial is defined as a decrease of 0.25 t/ha/yr or more. This threshold is set lower than wheat because millet is grown more often than wheat by subsistence farmers who obtain low net yields. Indeed the current net yield of millet (0.8 t~a/yr globally averaged, FAO, 1992, op cit.) is substantially lower than that of wheat. Hence, a smaller change in potential productivity for millet is of importance. eIndicated are areas in which the potential vegetation is estimated to change because of climate change over the period of simulation, 1990 to 2100. The potential vegetation calculations employ a global vegetation model, BIOME (Prentice, I.C., Cramer, W., Harrison, S.P., Leemans, R., Monserud, R.A. and Solomon, A.M., 1992. A global biome model based on plant physiology and dominance, soil properties and climate. Journal of Biogeography, 19: 117-134.), which is embedded in IMAGE 2.0 (Leemans, R. and van den Born, G.J., 1994. Determining the potential global distribution of natural vegetation, crops and agricultural productivity. Water, Air and Soil Pollution, 76: 133-161.) ENDNOTES
1 The IMAGE 2.0 model used for calculations in this report is fully documented in Alcamo, J. (Editor), 1994a. IMAGE 2.0: Integrated Modeling of Global Climate Change. Kluwer Academic Publishers, Dordrecht. Also published as Special Issue of Water, Air and Soil Pollution, 1994. Volume 76, Nos 1-2. 2 The baseline scenario is based on the Conventional Wisdom scenario documented in: Alcamo, J., van den Born, G.J., Bouwman, A.F., de Haan, B., Klein Goldewijk, K., Klepper, O., Leemans, R., Olivier, J.A., de Vries, B., van der Woerd, H. and van den Wijngaard, R., 1994b. Modeling the global society-biosphere-climate system, Part 2: computed scenarios. Water, Air and Soil Pollution, 76: 37-78. This scenario takes population and economic growth assumptions from the intermediate emissions scenario (IS92a) of the IPCC (1992). The population assumptions correspond to median estimates of the U.N. F u r t h e r assumptions of the Conventional Wisdom scenario are given in Alcamo, et al., Ibid.
Leggett, J., Pepper, W.J., and Swart, R.J., 1992. Emission scenarios for the IPCC: an update, in Houghton, J.T., Callander, B.A., and S.K. Varney (eds) Climate Change 1992: Supplement to the IPCC 1990 Assessment. Cambridge University Press, Cambridge, pp.71-95. 3
4 Enting, I.G., Wigley, T.M.L. and Heimann, M., 1994. Future emissions and concentrations of carbon dioxide. Technical Paper No. 31., CSIRO, Australian
149 Division of Atmospheric Research, Mordialloc, Australia. 5 The IPCC Working Group I exercise o n C 0 2 stabilization (Enting, et al., 1994, op cit.) did not specify the trend of non-CO2 gases. Therefore, we make the following assumptions: (i) CFC emissions are assumed to be phased out according to international agreements as interpreted in the intermediate IPCC emission scenarios (Leggett, 1992, op.cit.); (ii) other greenhouse gas concentrations (pX) are a s s u m e d to have a similar historical and future p a t h w a y of CO2 gas concentrations: pX(t) - pX(1990) pX(1990) - pX(1900)
pCO2(t) - pCO2(1990) pCO2(1990) - pCO2 (1900)
6 For these calculations it was assumed that sinks of greenhouse gases would not be artificially enhanced by large geoengineering projects such as pumping CO 2 t o low levels of the ocean. Also, it was assumed that land use emissions would not be reduced. 7
Enting, et al., 1994, op cit.
8 See, for example, Conventional Wisdom scenario of IMAGE 2.0, Alcamo, et al., 1994b, op cit., or IPCC reference scenarios in Leggett, et al., 1992, op cit. 9 Johannson, T., Kelly, H., Reddy, A. and Williams, R. (eds), 1993, Renewable Energy, Island Press, Washington.
10 Kassler, P., 1994. Energy for development. Selected paper, Shell International Petroleum Company, Shell Centre, London. 11 Ishitani, H. and Johanssson, T., 1995. Energy supply mitigation options. In: R.T. Watson and R. Moss (Editors), IPCC Working Group II: Impacts, Adaptation, and Mitigation. Cambridge University Press, Cambridge, (in review). 12 One of many important questions regarding these scenarios is whether there will be adequate land to provide the biofuels specified in these scenarios. Calculations with the IMAGE 2.0 model indicate that there could be spatial limitations in some regions (Alcamo, et al., 1994b, op. cit.) 13 Moreover, even if greenhouse gases were immediately stabilized, we still expect some climate change due to the historical build-up of greenhouse gases in the atmosphere, and because of the momentum of the climate system. Hence, a certain amount of climate impacts may be very difficult to avoid.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
155
ASSESSMENT REPORT ON NRP SUBTHEME "ATMOSPHERIC
PROCESSES
& UV-B RADIATION"
R. Guicherit TNO Institute of Environmental Sciences P.O. Box 6011 2600 JA Delft The Netherlands
With contributions by: H. ten Brink W. Ruijgrok, M. Vosbeek M. Allaart, R. van Dorland, A.J. Feijt, F. Kuik, A.C.A.P. van Lammeren, E. van Meijgaard, P. Stammes, G.H.L. Verver F.C. van Duyl, H.J. Lindeboom, R. Osinga J.P. Beck, J. Bordewijk, W.A.J. van Pul, H. Reinen, E. Schlamann, H. Slaper, D.P.J. Swart, T.H.P. The, D.L. Veenstra L. Dijkhuizen, W.W.C. Gieskens, T.A. Hansen, M.J.E.C. van der Maarel, P. Quest, J. Stefels B. Bregman, J. Vil~-Guerau de Arellano A. Baart, R. Bosman, P.J.H. Builtjes, P. Esser, R. Guicherit, K.J.M. Kramer R.L.J. Kwint, M. Roemer
ECN, Netherlands Energy Research Foundation, Petten KEMA, Environmental Services, Arnhem KNMI, Royal Netherlands Meteorological Institute, De Bilt NIOZ, Netherlands Institute for Sea Research, Texel RIVM, National Institute of Public Health and Environment, Bilthoven RUG, University of Groningen RUU, University of Utrecht TNO, Institute for Environmental Sciences, Delft
156
Contents Abstract 1. 0
0
0
Introduction Clouds, aerosols and radiation 2.1 Formation and air/sea exchange of dimethylsulphide (DMS) from marine sources 2.2 The role of aerosol of anthropogenic origin in the radiation balance of the atmosphere 2.3 Clouds-radiation-hydrologic interactions in a limited-area model 2.4 The effect of clouds on UV radiation and photodissociation rates in the troposphere The role of atmospheric ozone in global change 3.1 Introduction 3.2 Dutch participation in the international network for the detection of stratospheric change (NDSC) 3.3 Spectral UV radiation measurements and correlation with atmospheric parameters 3.4 Determination of the UV-B climate in the Netherlands: high resolution spectral measurements, monitoring and modelling from the perspective of risk-analysis 3.5 STREAM: Stratosphere Troposphere Experiment, study by Aircraft Measurements Tropospheric budget of non-CO2 greenhouse gases 4.1 Introduction 4.2 GLOMAC: Subproject of EUROTRAC related to climate modelling of KNMI 4.3 Application of 2-D global models 4.4 Global modelling of atmosphere trace gases-application of a global three dimensional model 4.5 Continental ozone issues; monitoring of trace gases, data analysis and modelling of ozone over Europe "Dutch contribution to The EUROTRAC-TOR project"
5.
Acknowledgements
6.
References
157 ABSTRACT The a t m o s p h e r e is a very complex, open, dynamic and multi-causal relation s y s t e m in which non-linear processes and feedback m e c h a n i s m s play an important role. Research within this subtheme in NRP-1 focused on uncertainties in our understanding of three issues i.o. the role of clouds and aerosols on the radiation budget the role of atmospheric ozone in global change and the effect of atmospheric change on I_W-B climatology tropospheric budgets of non CO2 greenhouse gases These issues are being dealt with in 12 projects. On average good progress has been made in m a n y of the projects and through t h e r e are some weaknesses, the overall quality of science and technological d e v e l o p m e n t s is good and in some cases even excellent in comparison to international standards. Links with international programmes such as IGBP, CEC, EUREKA/EUROTRAC are well established. 1.
INTRODUCTION
M a n y different gases can interfere with the earth's radiation budget. Some are highly stable and have residence times of decades or even a century or more. The most i m p o r t a n t of these gases are w a t e r vapour (H20); carbon dioxide (CO2), m e t h a n e (CH4), n i t r o u s oxide (N20), ozone (03) a n d t h e ( H ) C F C ' s (hydrogen-containing chlorofluorocarbons). O t h e r atmospheric constituents, including certain aerosols and notably clouds, m a y also affect radiation regimes. Most scientists agree t h a t this m a y have significant effect on regional climates. A model of global warming simulates the effects of various policy strategies on the built-up of greenhouse gases in the atmosphere (see Figure 1.1). First the built-up of radiation influencing constituents is linked to emission scenarios, which are based on assumptions about future population levels, global and regional energy resources, energy and material use, land-use changes (e.g. due to deforestation, regrowth and biomass burning), agricultural activity, industrial activity, certain policy d e p e n d e n t a s s u m p t i o n s about taxes, income and price elasticities of demand, and other economic factors. In Table 1.1 an overview of activities and emissions of radiatively active compounds are given.
158
Input & assumptions Productions & emissions
Demographics; Technical & Macro-economics; Policy etc. Anthropogenic / Natural (biogenic) Atmospheric (retention) models
Atmospheric concentrations
N20, C02, (H)CFC's, 03, CH4, aerosols (SO2]DMS) Clouds, H20-vapour; radiation Equilibrium radiative forcing models
$
Radiative effects
$
Estimated global warming
Figure 1.1 Schematic structure of the model of global warming In the second stage of analysis, atmospheric retention models are used to simulate chemical and biophysical processes by which the relevant emitted compounds are removed from the atmosphere, resulting in projections of their future concentrations. Table 1.1 Anthropogenic activities and emissions of radiatively active compounds
1:
2:
3: 4:
Fossil-fuel combustion a: CO2 emission (infrared IR trapping) b: CH4 emission by natural gas leakage (IR trapping and 03 changes, which on its turn leads to changes in I_W absorption and IR trapping) c: NOx emission (alters 03) & Carbonaceous soot emission (efficient solar absorption) e: SO2-sulfate emission (solar reflection and IR trapping) f: VOC/CO emission (alters 03) Land-use changes a: Deforestation (releases CO2 and increases albedo) b: Regrowth (absorbs CO2 and decreases albedo) c: Biomass burning (releases CO2, NOx and aerosols) Agricultural activity a: Release of CH4 (IR trapping and changes 03) b: Release of N20 (IR trapping and changes 03) Industrial activity a: Release of CFC's and their substitutes (IR trapping and 03 destruction) b: Release of SF6, CF4 and other ultra-long lived gases (IR trapping virtually forever) c: Release of VOC (03 changes)
159 In the third stage, radiative effects of the projected concentration changes are estimated and translated into equilibrium global and regional radiative forcing, which may be translated into temperature changes. The effects of the radiatively active constituents will not register immediately as a change in surface temperature. The oceans large thermal mass will cause a lag in warming effects. Nonetheless radiative active constituents will cause an eventual or "equilibrium" warming after some time, perhaps several decades after a certain atmospheric concentration has been reached. Since the industrial revolution an additional global average radiative forcing, due to an increase in atmospheric CO2, N20, CH4, stratospheric water vapour, CFC's and tropospheric 03 of 2.4 W/m2 has been calculated in this way, which is equivalent to a temperature increase of about 0.7 K The s u b t h e m e "Atmospheric processes and UV-B radiation" focuses on atmospheric processes. The atmosphere is a very complex, open, dynamic and multi-causal relation system in which nonlinear processes and feedback m e c h a n i s m s play an i m p o r t a n t role. One should note t h a t the feedback mechanisms themselves might influence the emissions in a direct way. The major problems we are faced with can be summarized as follows: 9 Radiative forcing by gases is a function of their concentration. The relation between emissions and concentration is usually not a linear one, but is determined by chemical processes in the atmosphere. The chemistry of the atmosphere, however, will change if the composition of the atmosphere, as is the case, changes. The consequence e.g. being t h a t changes in the atmosphere's methane concentration since the turn of the century does not reflect changes in its emission pattern. This means that for projections of atmospheric concentrations of radiatively active gases for different emission scenarios, one should also take into account the chemical processes occurring in the atmosphere. 9 Aerosols may change the earth's albedo in a direct way through absorption and backscatter of solar radiation and indirectly as condensation nuclei in cloud formation. Aerosol cloud condensation nuclei (CCN) may namely increase cloud droplet concentration and cloud reflectance (albedo) of incoming solar radiation. Atmospheric aerosol particles of concern are both biogenically derived from dimethylsulphide (DMS) oxidation and by anthropogenic activities, particularly sulphates from SO2 emissions, organic condensates and soot from biomass burning. Recent applications of coupled atmospheric chemical/radiative transfer models, by utilizing empirical aerosol scattering properties, have shown that anthropogenically derived sulphate aerosols cause clear-sky climatic forcing, that when averaged over the globe is comparable in magnitude, but opposite in sign, to forcing by CO2. Unlike CO2, anthropogenically derived aerosol particles are not uniformly distributed over the globe but are mainly found over industrialized areas in the Northern Hemisphere. As a matter of fact the present day aerosol forcing in many regions of the Northern Hemisphere, as an annual average, may offset the combined greenhouse effect of CO2; CH4; N20; CFC's, and 03 (Figure 1.2).
160
Radiation
Sulphate haze CCN Greenhouse gas trapping
Emissions CH 4 N20 CO 2 (H)CFC's
VOC NO x CO Halons
.. -lP"
'V....
I l
... - - " l ......
"
1'
SO 2 gas
t
DMS
Figure 1.2 Interference of radiation with gases, aerosols and clouds The effect of clouds is even more complex. Depending on cloud physics and cloud dynamics, clouds may exert a positive or negative feedback in radiative forcing. It is assumed, and this certainly holds for low level stratos- and s t r a t o c u m u l u s clouds over the oceans, t h a t the average net effect is a negative one. Cloud statistics are very sensitive to minor changes in atmospheric circulation patterns and the hydrological cycle. Also the n u m b e r of condensation nuclei may play a role, especially over the oceanic areas of the Southern Hemisphere and to a lesser extent also over oceanic areas of the N o r t h e r n Hemisphere. At the moment it is not clear in w h a t way cloud statistics will react to climate change. The uncertainties projected increase in average global t e m p e r a t u r e (ranging between 2-5 K) for doubling the atmospheric CO2 concentration is merely due to the way clouds are treated in the climate models.
161 The role of atmospheric ozone needs special attention in global change. Depletion of s t r a t o s p h e r i c ozone m a y alter the UV-B climatology, while an increase in t r o p o s p h e r i c ozone m a y have serious adverse effects on the biosphere. Tropospheric ozone is not emitted, it is formed in the atmosphere by chemical reactions involving compounds such as NOx, CO and VOC, which are called ozone precursors. The role of tropospheric ozone in climate change is significant. A complicated factor being, t h a t the magnitude of the O3-forcing effect is height dependent with a maximum around the tropopause and an "opposite" effect above 30 km. Since the effect of precursor emissions and atmospheric chemical processes on tropospheric ozone levels depend on varying regional atmospheric conditions, it is difficult to predict future global changes in tropospheric ozone concentrations accurately. This also holds for changes in stratospheri~ ozone depletion. Although the ozone depletion substances, due to their long atmospheric residence times, are distributed evenly over the globe, there exists large regional differences in the depletion of the ozone layer, notably in the antarctic and arctic regions, over the tropics and at medium latitudes. One may state that next to the hydrological cycle and the direct and indirect effects of tropospheric aerosol, changes in the ozone column density distribution pose the largest uncertainty in model calculations of climate forcing due to anthropogenic induced changes in the composition of the atmosphere. Related to stratospheric ozone depletion are, as was mentioned before, possible changes in the UV-B climatology. This needs special attention since UV-B plays an i m p o r t a n t role in the chemistry of the atmosphere and may enhance u r b a n photochemical smog and UV-B may also exhibit negative effects on the biosphere. The UV-B climatology is dependent on many factors of which the most important are: 9 Stratospheric ozone destruction. 9 Changes in tropospheric ozone. 9 Scattering by aerosols and interaction with clouds. Insight in these processes is crucial to predict changes in the I.W-B climatology to be expected near the surface of the earth. The NRP s u b t h e m e "Atmospheric processes and UV-B radiation" is aimed at s t u d y i n g the aforementioned processes. Changes in the composition of the atmosphere on a global scale and related changes in radiative forcing may have far reaching social-economic consequences especially with regard to preventive and adaptive measures to be taken. For this reason projects belonging to this theme, cannot predominantly be qualified as basic science, but are at the same time policy oriented. The a s s e s s m e n t report will focus on uncertainties in our u n d e r s t a n d i n g of 3 priority issues i.e.: I. Clouds, aerosols and radiation II. The role of atmospheric ozone in global change III. Tropospheric budgets of non-CO2 greenhouse gases The projects to be assessed and the priority issue they are linked with, are listed Table 1.2.
162 Table 1.2 List of projects in the NRP subtheme on "Atmospheric processes and UV-B radiation" Title
Projectleader
Clouds, aerosols and radiation Formation and air/sea exchange of dimethylsulphide R.Guicherit (DMS) from marine sources
Number
850026
The role of aerosols of anthropogenic origin in the radiation balance of the atmosphere
H.M. ten Brink
852066
Clouds-radiation-hydrological interactions in a limited area model
A. van Lammeren
851058
The effect of clouds on ultraviolet radiation and photodissociation rate in the troposphere
H. van Dop
850018
The role of atmospheric ozone in global change Dutch participation in the international network for D.P.J. Swart the detection of stratospheric change (NDSC)
850024
H. Kelder
852088
H. Slaper Determination of the UV-B climate in The Netherlands: high resolution spectral measurements, monitoring and modelling from perspective of risk-analysis
852087
Spectral UV radiation measurements and correlation with atmospheric parameters
P.J.H. Builtjes
VvA205
Budget studies Climate modelling/Global Modelling of Atmospheric Chemistry (GLOMAC)
H. Kelder
851050
Application of 2-D global models
M.G.M. Roemer
852072
Global modelling of atmospheric trace gases "Application of a global three dimensional model"
J.P. Beck
852070
Continental ozone issues: monitoring of trace gases, J.P. Beck data analysis and modelling of ozone over Europe "Dutch contribution to the EUROTRAC-TOR project"
852094
Stratosphere Troposphere Experiment: Study by aircraft measurement
163 2.
C L O U D S , A E R O S O L S AND R A D I A T I O N
Aerosol particles play an important role in the radiation budget of the atmosphere because of t h e i r direct interaction (absorption and scattering) of solar and t e r r e s t r i a l radiation, as well as t h r o u g h their influence on cloud formation processes. Aerosol particles have a lifetime of a few weeks in the troposphere and occur in highly variable concentrations. A large proportion of the particles, which are of interest for the radiation balance and for cloud processes are derived from n a t u r a l sources, a n t h r o p o g e n i c gaseous s u l p h u r emissions, emissions of carbonaceous and organic particles, and biomass burning. It is complicated to determine the direct effect of aerosols. This is due to the fact t h a t the effect is dependent on the aerosol absorption to backscattering ratio, surface albedo, total aerosol optical depth and solar elevation. Their is, however, general agreement, that anthropogenically generated sulphate aerosols will reduce solar irradiance, leading to a net negative change in radiative forcing and thus regionally offsetting the effect of w a r m i n g due to increased concentration of greenhouse gases. The global radiation (energy) balance is very sensitive to cloud albedo, particularly for marine stratus clouds which cover a substantial part (about 25%) of the earth. Cloud albedo itself is sensitive to changes in droplet n u m b e r concentration. The droplet n u m b e r depends on the concentration of cloud condensation nuclei (CCN) which in turn depends on condensation nuclei (CN) or on the aerosol concentration. The ability of an aerosol particle to act as a CCN under low supersaturation found in clouds depends on its size and its chemical composition (notably its w a t e r solubility and substances t h a t influence surface tension). There is a significant non-linearity in the effect on cloud formation and cloud microphysics of changes in CCN concentration; depending on the starting CCN concentration. The effect is far more pronounced in a r e a s (such as clean oceanic sites) where low aerosol concentrations prevail, compared to more polluted areas, e.g. areas over the continents which are strongly influenced by anthropogenic emissions. The hypothesis by Charlson et al (1987) of a connection between climate and p h y t o p l a n k t o n activity in ocean surface waters is based on the fact t h a t CCN concentrations in air over oceans far from land are low, t h a t CCN available in clean maritime air are composed almost totally of sulphate particles, and t h a t this sulphate originates almost entirely from emissions of dimethylsulphide (DMS) from the ocean surface (see Figure 2.1). This subtheme comprises 4 projects.
164
O
indirect._ Radiation ~
Cloud condensation nuclei I'
+
'~xter
Sulfate aerosol I'
budget
Global temperature Climate feedbacks
+
S~ 2 +
+ or-?
DMS
I
Atmosphere
+
Ocean
/
DMSP~
I Phyt~176
I
CO 2 / S042-
H2S / CH 4
Figure 2.1 Proposed feedback cycle between climate and marine DMS production (adapted from M.O. Andreae, 1990) 2.1 F o r m a t i o n a n d a i r / s e a e x c h a n g e of d i m e t h y l s u l p h i d e (DMS) f r o m marine sources
K.J.M. Kramerl, A. Baartl, L. Dijkhuizen2, R. Guicherit (coordinator)l, W.W.C. Gieskes2, T.A. Hansen2, R.L.J. Kwint 1, H.J. Lindeboom3, R. Osinga3, P. Quist2, J. Stefels2, M.J.E.C. van der Maarel2 and F.C. van Duyl3 1 TNO, Institute of Environmental Sciences (IMW) P.O. Box 6011, 2600 JA Delft University of Groningen, Department Microbiology, P.O. Box 14, 9750 AA Haren 3 Netherlands Institute for Sea Research (NIOZ), P.O. Box 59, 1790 AB Den Burg 2
Introduction
Global climate and the "greenhouse effect" can be influenced by several feedback mechanisms. The present project focuses on dimethyl sulphide (DMS) as a likely candidate for important negative feedback processes. DMS is the most important biogenic precursor of non-seasalt sulphate (NSS-SO4) which is a major source of
165 cloud condensation nuclei (CCN). DMS production in marine environments is linked to p h y t o p l a n k t o n a n d m a c r o - a l g a e ; it is formed from t h e p r e c u r s o r B-dimethylsulphoniopropionate (DMSP), which is believed to be an osmoregulating or cryoprotecting agent. It is i m p o r t a n t to realise, t h a t the ultimate flux of DMS to the atmosphere is dependent on m a n y biological and physical/chemical factors. The production of DMSP by algae is subject to species abundance and composition, which is a function of the environmental conditions. The conversion of DMSP to DMS by algae and microorganisms, the degradation of DMS and DMSP by bacteria, loss of compounds due to sedimentation, photochemical processes or escape to the atmosphere, all processes are intricately linked to each other. The various paths of formation and degradation of DMSP and DMS, and their relative importance are presented in Figure 2.2a, where the state of the art at the start of the project is depicted. Quantification of the DMS emitted to the atmosphere has been found difficult as large temporal and geographic differences occur. For a better estimation of the fluxes, a proper u n d e r s t a n d i n g of the (micro)biological processes in the w a t e r column related to the biogeochemical cycle of DMS(P) is essential. Although the possible effects upon the world climate will to a large extent be a function of open ocean processes, coastal waters (and sediments) may play an i m p o r t a n t role on a regional scale, not in the last place because of its enhanced algal productivity. If we consider the algal species Phaeocystis sp. and Emiliania huxleyi as typical and i m p o r t a n t representatives for coastal and open ocean waters respectively, Figure 2.3 illustrates schematically the DMS(P) cycles t h a t play a role in the various marine compartments. The objectives of the present project are: 9 To assess these processes and rates of biological production. 9 To assess the processes of degradation and t r a n s f o r m a t i o n in the w a t e r column and the role of sedimentation therein. 9 To estimate the fluxes from water to the atmosphere.
166
Air Water; sediments
CO2, sulfate
DMSO
~,~. ~~
DMSP (in algae) ~ I I
~
photochemical
algae: lyase?
aerobic
bacterial egradation
anaerobic ~i~ ~ DMS ~ CH4, CO2, sulfide (+ acrylate) bact. degr.
excretion sedi- " ~ , m e n - ~si~nescence . ~
tation
/~
oxidation ~~ bacterial ~~/red
bacterial
degradation
r MMPA, MPA rDMSP DMSP (in algae) autolysis (free) bacterial bacterial degradation degradation
Figure 2.2a The understanding of major DMS pathways before the start of the NRP/DMS project
167
Air Water; sediments
| CO 2, sulfate
DMSO ~ ~
DMSP (in algae) ~ I
QI
I
,!,
algae: lyase? ._, ~,)
~1~ anaerobic ~ DMS ~ CH 4, CO 2, sulfide (+ acrylate) bact. degr. Q
sedi-~~ excretion men"%~enescence tation O ~grazing
(in algae)
"
aerobic bacterial egradation
"
autolysis ( ~ bacterial degradation
|
I I " algal: lyase + I I bacterial degradation
(free)
bacterial degradation
Figure 2.2b Our understanding of the major DMS pathways including results of the NRP/DMS project
c'~;&--~-cc~-, --__~ . . . . . . .
rivers-)~N+P intertical mats
I I
I
I T N O (field + lab)
i TNO
I
~ phaeocystis --)~DMS(P) ~ S O 4 , .... I ~ " -- ..
n emilania --~DMS(P) ~
~,1.-~_~
"---
o s/P)
coastal zone
c"T~_L~ coN:, ~__.-
.
~ DMS(P) ~
SO4n
ocean + northern North Sea
Figure 2.3 The sulphur cycle in different parts of the marine environment
SO4n
168 To investigate the different parts of the DMS(P) cycle, experiments were performed under well defined laboratory conditions, in large experimental enclosures (mesocosms) and in the natural environment (Dutch coastal waters). In this project research focussed on species typical for the Dutch coastal environment because it was considered that although the rates may not be comparable to those in oceanic systems, the processes as such will probably be analogous in both systems. Within the project a close cooperation was established between different marine scientific disciplines, in order to be able to tackle the various routes of production/elimination and the process rates involved. The project is a combined scientific effort of the: 9 University of Groningen, Department of Marine Biology and Department of Microbiology. 9 TNO Institute of Environmental Sciences, Department of Biology, and Department of Environmental Chemistry. 9 Netherlands Institute for Sea Research, Department of Applied Scientific Research. In the following sections the DMS(P) production/elimination paths will be discussed in the framework of the activities performed within the present project. References will be made to the numbered routes (paths) in Figure 2.2b, which represents the situation as we understand it now.
Laboratory studies
Phytoplankton. DMSP is produced intracellularly by phytoplankton. Two mechanisms are to be considered for the formation of DMS, either direct (through algal lyase) or via DMSP that is released into solution (by excretion, senescence and/or predation). Until now, little attention has been paid to the influence of environmental stress factors on the DMSP content of algal cells, and the possible role of algae in the conversion of DMSP to DMS (paths 1 and 2). As it is often suggested that bacteria are most important in this conversion (see below), evidence for an important role of phytoplankton itself was never presented. As Phaeocystis sp. is considered an important producer of DMSP in coastal waters, this species was investigated in detail in laboratory studies. It is often hypothesized t ha t DMSP production in algae is stimulated under nitrogen limited conditions, when DMSP substitutes for other- nitrogen containing - osmolytes. This was tested by inducing osmolyte production in axenic Phaeocystis sp. cultures at different salinities (from 25 to 45x10-3) and different N:P ratios. Although DMSP per cell did increase with salinity, indicating its function as an osmolyte, no significant difference was found between the N and P limited cultures. In Phaeocystis sp. cultures, that were free of bacteria (axenic) the transformation of extracelluar DMSP to DMS and acrylate was measured (path 2). It was proven t h a t this conversion could be correlated with an enzymatic (lyase) reaction associated with the cells, which could be inhibited by heating. The lyase activity in
169 living cultures increased with temperature: at 5 ~ activity was 50% of the activity at 20 ~ During the growth phase of an axenic Phaeocystis sp. culture leakage of DMSP or DMS was very low. DMS production started in early stationary phase, when only small a m o u n t s of dissolved DMSP appeared to be present, indicating a rapid conversion of DMSP by the algal lyase. In a completely lysed culture, approximately 75 % of total DMSP was found to have been converted to DMS. In experiments with crude extracts of Phaeocystis sp. cells, the DMSP-lyase activity exhibited an alkaline optimum, which is in contrast to results from experiments with other organisms as reported in the literature. This points into the direction of a species specific enzyme.
Bacteria. Once DMSP or DMS are liberated in the w a t e r column, bacterial d e g r a d a t i o n / t r a n s f o r m a t i o n m a y occur. Previous studies on the microbial metabolism of DMSP had shown that there are two pathways for degradation of DMSP. The first p a t h w a y involves a cleavage to dimethyl sulphide (DMS) and a c r y l a t e (path 4). The second p a t h w a y involves the d e m e t h y l a t i o n to 3-S-methylmercaptopropionate (MMPA) and subsequently to 3-mercaptopropionate (MPA) (path 3). DMS may be further degraded to CO2 and SO42- (path 5) under aerobic conditions. F r o m a v a r i e t y of m a r i n e sources ( p h y t o p l a n k t o n cultures, s e d i m e n t s , macro-algae and mesocosms) bacteria capable of degradation DMS or DMSP in aerobic environments could be isolated and characterized during the present project. Common characteristics of these strains are the ability to utilize a broad substrate spectrum, their motility and their coccoid rod-like morphology. Apart from aerobic conversion of DMS(P), anaerobic degradation of both DMS and DMSP is possible. With a few exceptions, anaerobic processes are not considered important for the water column. However, anoxic sediments can be a b u n d a n t in many coastal waters, where they may influence the cycling of DMS(P). Under anoxic conditions, DMS may be converted by sulphate-reducing bacteria or by m e t h a n o g e n s to m e t h a n e (path 6). The degradation step from DMSP via MMPA and MPA (path 3), can be followed by a conversion to m e t h a n e by m e t h a n o g e n s . Except for D M S - m e t a b o l i z i n g m e t h a n o g e n s , no anaerobic microorganisms t h a t are responsible for one of the above mentioned conversions had been isolated from anoxic marine sediments. During this study it was found that demethylation of DMSP to MMPA in anoxic W a d d e n s e a sediments is coupled to sulphate reduction. A pure culture of a DMSP-demethylating sulphate-reducing bacterium, which was identified on basis of physiological characteristics and positive hybridisation with genus-specific molecular RNA probes as a Desulfobacterium sp, was isolated from the sediment. Certain other Desulfobacterium species were also found to demethylate DMSP. It is, however, not yet clear w h a t the ratio of DMSP cleavage over DMSP d e m e t h y l a t i o n is at n a t u r a l (anoxic) concentrations of DMSP. Recently a DMSP-cleaving anaerobic bacterium was isolated which ferments acrylate to propionate and presumably CO2 but which is not able to utilize DMS. This strain
170 showed a high DMSP-lyase activity (approximately 10 ~imol/min.mg protein) and a rather high Km value for DMSP (150 ~tM). The methanogens Methanosarcina sp strain MTP4 and Methanosarcina acetivorans were found to be able to convert MMPA to MPA and methane during growth. Methanogenic conversion of MMPA to MPA was only found in anoxic Waddensea sediment when certain antibiotics, acting against bacteria but not methanogens, were used. Under non-inhibited conditions MMPA was rapidly converted to methanethiol, which was subsequently converted to methane. It was concluded that conversion of MMPA in anoxic marine sediments directly or indirectly results in the formation of methane. The anaerobic metabolism of DMS was also studied in anoxic Waddensea sediment. In this sediment DMS appeared to be converted to methane, as had been described for other anoxic marine sediments. The conversions of DMS to methane seems to be restricted to Methanosarcina species. The population of DMS-metabolizing methanogens appears to be in the same range as the total population of methylotrophic (i.e. methanol- and trimethylamine-utilizing) methanogens (1-6x106 cells per gram dry weight). Both pathways indicate that methane can be produced under anaerobic conditions; there may thus exist a positive flux of methane to the atmosphere. Thus, in anoxic marine environments, not only the counteractive effect of global warming through DMS may take place, but also the production of the potent greenhouse gas methane. Mesocosm studies Apart from the direct production of DMS by algae, DMSP may be introduced into the water column by various other mechanisms (path 1). Living algae may excrete DMSP, the DMSP may be liberated when the algal cells die (senescence), or zooplankton may release DMS(P) during or after digestion of the algal cells (predation, sloppy feeding). Particulate DMSP may be transported out of the surface layers by sedimentation.
In order to study the role of phytoplankton and zooplankton (and their interaction) in the production of DMSP and the release of DMS to the water column (and eventually to the atmosphere), three series of large scale experimental systems (pelagic mesocosms) were used to study the development of phytoplankton blooms as function of the presence of zooplankton, different nutrient regimes, species composition and varying general environmental characteristics. Measurements on densities and activities of aerobic DMS(P) metabolizing bacteria were included. In one mesocosm experiment the deposition of particulate m a t t e r (containing DMSPpart) was quantified. The first two mesocosm experiments showed that the DMS is generally released in the water column directly after the peak of the phytoplankton bloom, during the senescence phase. This was observed for a diatom bloom. For the bloom of Phaeocystis sp. in the last mesocosm experiment, a maximum DMS production was found during the stationary phase; no direct correlation was found with chlorophyll-a. The decline of the Phaeocystis bloom did not lead to an increased DMS concentration. However, but just before the release of DMS a dramatic drop in cellular DMSPpart was observed. The latter species produced substantially
171 higher amounts of DMS(P) than the diatom species, which confirmed earlier findings reported in the literature. Due to boundary effects, the duration of the chlorophyll-a peak appeared to be compressed in time in our mesocosm studies, as was the peak of [DMS]water. It appeared that the change in DMS concentration could occur very fast. In a matter of days the [DMS]water can change by orders of magnitude. Both increase and decrease of [DMS]water showed this dynamic character. In m e s o c o s m - s y s t e m s with a p p a r e n t l y similar characteristics (both in chlorophyll-a, nutrients, plankton assemblages) the production of DMS did not always follow the same pattern. Considering the results of the last mesocosm experiment the role of Phaeocystis may have been underestimated in the first two experiments. Another reason could be that the importance of the microbiological activity is not fully understood. In order to study the potential effect of zooplankton upon the production of DMS, in a number of mesocosm experiments the zooplankton was inhibited or removed. A positive influence of zooplankton e.g. of grazing (path 1) upon the release of DMS into the watercolumn could not be found, however. This could indicate t h a t zooplankton does not play an important role in the DMS(P) dynamics. However, as the time axis of the phytoplankton bloom in the mesocosm experiments is compressed, a mismatching of life stages of algae and zooplankton may have biased these experiments to some extent. During this last experiment, deposition of suspended particulate m a t t e r was studied (path 7). At least in the mesocosms, sedimentation is an important sink for biomass of Phaeocystis sp. The sedimentation of Phaeocystis cells showed a good correlation with Phaeocystis cell numbers in the water column, suggesting that the sinking rate of Phaeocystis cells was rather constant. The daily sedimentation of living cells being nearly equal to the standing stock at that time. At the decline of the Phaeocystis bloom in the mesocosm, sedimentation accounts for only 50 % to the observed loss in biomass. This strongly suggests that cell lysis in the water column must have been important at this time. Although this cell lysis does increase the DMSPdiss concentration, this does not lead to a peak in the DMS concentration: possibly the production and consumption are of a similar magnitude. The importance of bacteria in the conversion of DMSP and DMS was studied in the same mesocosm experiments. The DMSP-degrading population consisted of a variety of microorganisms. Bacterial populations are strongly stimulated by the collapse of an algal bloom, as their substrate becomes abundant from the decaying algal cells. The DMS-utilizing bacterial population reacted upon the increase of DMS with a delay of about a week. The DMS assimilating ability seemed to need some induction. The bacterial activity found in the mesocosm systems, 10-1000 ~tmol DMS per mg protein per day, was in the range of activities of laboratory cultures. The biological turnover rate ranged between and 0.2 and 0.6 days, whereas the turnover by DMS effiux ranged between 2 and 5 days. In these experiments, the contribution of the bacteria to the loss of DMS was calculated to be approximately
172 90% of the total DMS loss. This figure should be taken with some care, as only few observations were made. Nevertheless, the importance of this process should not be underestimated. Although limited data on the turnover of DMSP exist, they indicate a key role for the demethylation of DMSP via MMPA (path 3) rather than the cleavage of DMSP (path 4). This means that only a fraction of the DMSPwater seems to contribute to the formation of DMS. In benthic mesocosm studies carried out separately, addition to the sediment of fresh algal material (dominated by Phaeocystis sp., path 7) resulted in a strong (benthic) microbial response after 2 days, which indicated that freshly deposited algae become subject to rapid bacterial degradation. In shallow waters, benthic processes are therefore likely to have a strong influence on the processes in the water column in the pelagic mesocosms. Since m a n y Phaeocystis cells reach the mesocosm floor intact, there is also a continuous downward transport of considerable amounts of particulate DMSP. Sedimentation of DMSP may significantly contribute to the production of DMS in sheltered, shallow systems like the mesocosms, as a result of autolysis or bacterial d e g r a d a t i o n on the bottom (path 8). Despite the relatively high d o w n w a r d transport of DMSP, and the potentially fast transformation to DMS, there was no evidence that DMS in the water column correlated with the sedimentation rate. In deep, oceanic systems, path 6 may be a sink for DMSPpart. Since the rates of bacterial transformation were not sufficiently well determined, it seems premature to base budget studies on these results. Another u n k n o w n box in such budget studies forms the potential formation of dimethylsulphoxide (DMSO) by photochemical oxidation and/or bacterial activity (path 9). This compound has worldwide received little attention, due to the problematic analysis. By closing of one mesocosm system, the actual flux of DMS over the water-atmosphere interface could be determined. The m e a s u r e d flux of DMS to the atmosphere (path 10), under constant wind speed conditions (1.3 m/s) in one mesocosm system, confirmed t h a t the m a x i m u m efflux of DMS from the waterphase to the atmosphere took place at the same period of the development of the phytoplankton bloom, linked to the (elevated) water concentrations. The measured flux of DMS (at this low wind speed) agreed well with the calculated flux according to the Liss-Slater model (see next Section).
Field o b s e r v a t i o n s The laboratory observation t h a t the DMSP-lyase activity could result from a species specific enzyme, was confirmed by field experiments. During the spring bloom of 1993, a DMSP-lyase-assay was applied to n a t u r a l s e a w a t e r samples off the Dutch coast. A very good correlation was observed between the DMSP-lyase activity and Phaeocystis sp. n u m b e r s , the most abundant species found during the cruise (r2 = 0.9660, n = 23). No correlation was found with either one of the other species. From these results a potential DMS production via route 2 by Phaeocystis sp. cells in the field was calculated. It appeared t h a t DMS production rates were in the same order of magnitude as the
173 total abiotic loss factors (fluxes of DMS to the atmosphere and photochemical degradation). In order to try to link the processes in the mesocosms to the natural environment, water samples were collected regularly in the Marsdiep tidal channel over a period of 1.5 years. M e a s u r e m e n t s of DMS(P) and other organic sulphur compounds in the coastal waters of the Marsdiep tidal inlet, showed that also here only during or shortly after the phytoplankton peak the majority of the DMS was released. In only one week the [DMS]water increased or decreased five to ten fold. M a x i m u m concentrations were up to 20 nM, as compared to 450 nM in the mesocosms in the same period. Other sulphur compounds, like carbonylsulphide (COS) and dimethyldisulphide (DMDS) were found to change not much with the time of the year, w i t h concentrations s u b s t a n t i a l l y lower t h a n those of DMS (for COS and DMDS respectively: 1.2 nM and 0.35 nM in the field, comparable to 1 nM and 0.4 nM found in the mesocosms). Maximum concentrations of particulate DMSP observed, were about 1500 nM in the Marsdiep, which is substantially less than those found in the mesocosms (7500 nM). The ratio of p a r t i c u l a t e DMSP/Chl-a, however, was 20 nmol/~g in the M a r s d i e p which compared well with the 10-60 nmol/~g in the mesocosm experiment. The Marsdiep experiment (and the last mesocosm experiment) showed, t h a t from the total a m o u n t of DMSP produced by the phytoplankton, only 5 to 10% can be detected as DMS in the watercolumn. The sea to air transfer-velocity of DMS has been experimentally determined. This p a r a m e t e r is of major importance in calculating the DMS flux from the DMS concentration in the w a t e r (flux = concentration in water x transfer velocity). Liss and Slater showed that the transfer velocity (of CO2) primarily to be dependant on the wind velocity. Their empirical relations are being applied to DMS as well. To our knowledge there have been no report of the experimental determination of the sea-air transfer velocity of DMS in the open literature. In this study the transfer velocity of DMS from water to the atmosphere has been determined by two different methods : 9 Enclosure (at low, artificial wind velocity of 1.3 m/s). 9 Concentration vs. height gradient method (field measurements, wind velocities over a range from nearly 0 up to 11 m/s). Good agreement was found between the Liss-Slater relations and the experimental data at lower wind velocities (up to 4 m/s). At higher velocities the Liss Slater relations u n d e r e s t i m a t e s the transport velocity up to 20 %., assuming, however, t h a t the collected w a t e r in the field observations contained a r e p r e s e n t a t i v e a m o u n t of DMS for the surface w a t e r s t h a t d e t e r m i n e d the a t m o s p h e r i c concentrations.
Climate modelling The influence of the sea to air t r a n s p o r t of DMS upon the climate has been estimated by mathematical modelling.
174 In the atmosphere DMS is converted to primarily sulphate and methanesulfonic acid (MSA). The climate effect has been calculated using a global zonal averaged 2D atmospheric chemistry and transport model. This global model has a resolution of 10 x latitude and 20 layers (of 500m and l k m height) in a vertical dimension from 0 to 16 km. Transport is described by using seasonally averaged data for advection and eddy-diffusion and also for DMS emissions (literature data). The data base contained over 130 different chemical compounds and over 200 different chemical reactions. The climate effect has been calculated from the direct climate effect of sulfate aerosols, and the effect of DMS emissions has been compared to t h a t of anthropogenic SO2 emissions. The radiative forcing resulting from DMS emissions amounts to a global annual averaged value of-0.4 W/m2. It was shown, that at northern hemisphere mid-latitudes, the radiative forcing by sulfate aerosols is governed more by anthropogenic sulfurdioxide (SO2) emissions than DMS emissions. DMS, however, still accounts for 20 to 30 % of the annual averaged radiative forcing of sulfate aerosols. At the other latitudes and especially on the southern hemisphere, DMS is the major precursor of sulfate aerosols and the relative contribution of DMS emissions to the direct radiative forcing was calculated ranging from 50-99 %. Conclusion Many routes of formation/degradation of DMS(P) and related compounds were studied in this project. Figure 2.2a presents the situation as we understood it at the start of the project; Figure 2.2b summarizes our present understanding, based on our experimental results. Bold arrows indicate that these paths are considered important. It became possible to get a better insight in the production and fate of DMSP and DMS, and the relative importance of the DMS flux to the atmosphere and bacterial turnover as the main loss factors for DMS during an algal bloom. The strong relation with the development of algal blooms was confirmed. It was found t h a t the efflux of DMS to the atmosphere was definitely episodic in character: closely connected to (only) the occurrence of an algal bloom, fast increases and decreases of DMS in the water column were observed in all experiments. The importance of the processes, as indicated by the arrows in Figure 2.2b, may temporarily be enhanced due to physical effect: wind speed and turbulence. This emphasizes the significance of the processes that may occur over short time periods. Because not all results fit easily into one model, and because the importance of the bacterial compartment could only in a late stage of the project be confirmed, it seems too early to quantify the rates of the various processes. Based on the availability of DMSPpart, the DMSwater, and the calculated and measured DMS fluxes, it was calculated that only part of the DMSP produced is transformed into DMS, and that only part of this DMS produced actually escapes to the atmosphere. This implies t h a t minor changes in the bacterial and phytoplankton population density or composition, could have pronounced effects on the global S budget, with all of its consequences.
175
Assessment This project has given insight in some important processes governing the oceanic S-cycle. A major finding is, that only part of the DMSP produced, is transformed into DMS and that only a small part of this DMS actually escapes to the atmosphere. This implies that minor changes in the bacterial and phytoplankton population density or composition, could have pronounced effects on the global sulphur budget, with all its consequences. For the transfer of DMS a good agreement was found between the Liss-Slater relations and the experimental data. It was also found that the global annual radiative forcing due t.o DMS derived aerosols amounts to -0.4 W/m2. It was further found that at northern midlatitudes DMS accounts for 20-30% of the radiative forcing due to sulphate aerosols. For the other latitudes and especially over the southern hemisphere DMS is the dominant precursor of sulphate aerosols. The indirect effects due to cloud formation are much h a r d e r to deal with, no attempts are made. Although progress in understanding the oceanic S-cycle has been achieved, the state of the art is such, that at his moment no predictions can be made, with respect to expected changes in DMS emissions as a result of expected environmental changes, and the consequence thereof on climate change. 2.2 The role of aerosol of anthropogenic origin in the radiation balance of the a t m o s p h e r e H.M. ten Brink ECN, P.O. Box 1, 1755ZG Petten, The Netherlands
Direct effect The present study aims at assessing the influence of anthropogenic aerosol particles on the solar radiative flux in Europe. Aerosol particles reflect shortwave radiation and absorb little infrared terrestrial radiation. They thus exert a cooling forcing over surfaces with low albedo, the "direct" forcing effect. Aerosol particles originate both from natural and anthropogenic sources. The anthropogenic particles form an (extra) forcing factor introduced by mankind. They mostly result from fossil fuel combustion. Estimates for the forcing by the anthropogenic aerosols indicate a regional forcing in the heavily polluted regions of Europe of up to 10 W.m-2 with an uncertainty of the same order of magnitude. This means that, after the forcing by clouds, the forcing effect of aerosols is the second largest climatic forcing term. The forcing by aerosols is a regional effect, because of the limited residence time of the particles in the atmosphere. Better quantification of the "direct" aerosol effect in Europe is studied by ECN in a separate European project, of which the progress report is in press. A main preliminary conclusion is that nitrate and, possibly, carbonaceous aerosol are more important than sulfate in the local "direct" effect. In the above mentioned estimates sulfate was used as the only anthropogenic aerosol component. The actual local forcing is therefore higher due to nitrate and carbonaceous aerosol. These aerosol components are not directly emitted but almost exclusively formed in the atmosphere from gaseous emissions, which makes source apportionment difficult.
176
I n d i r e c t effect In de indirect effect, anthropogenic aerosols act as a radiative cooling forcing via clouds. Aerosol particles serve as the nuclei on which the clouds in the atmosphere form. In the present study, ECN assesses the influence of anthropogenic aerosol particles on cloud structure. Anthropogenic aerosol particles act as extra cloud nuclei, in addition to the n a t u r a l nuclei. The extra cloud droplets change the microstructure of the clouds and thus their radiative transfer as well as their lifetime. Simple radiative models show t h a t this leads to an increase in the reflectivity of the clouds. The influence on the infrared absorption is, however, small. Thus anthropogenic aerosol particles exert an "indirect" radiative cooling effect. Sensitivity studies show that the "indirect" effect is most i m p o r t a n t in marine stratus near polluted continents.
Cloud c h a m b e r In a first approach to assess the effect of anthropogenic aerosols on clouds the differences in the microphysics of clouds formed in clean and polluted marine air were investigated. This is done by artificial cloud formation using a cloud chamber, drawing in ambient air and comparing the number of cloud droplets formed. The site of the cloud chamber is ideally located to study marine clouds, since it is situated at the coast of the North Sea in The Netherlands. In arctic North-West airflows only natural aerosols occur and extra anthropogenic aerosol is present when the air travels more southerly over the UK. In the cloud chamber the supersaturation can be controlled in the range of 0.1 to 0.3% s u p e r s a t u r a t i o n values, which correspond to the s u p e r s a t u r a t i o n s in the marine s t r a t u s of interest. The particle size and number-concentration of the ingoing aerosol and the unactivated aerosol (particles not grown to droplets) are also compared. First, laboratory generated aerosols were used for testing the p e r f o r m a n c e of the chamber. Aerosol particles w i t h a size l a r g e r t h a n approximately 0.07 ~tm were found to grow into cloud droplets. This "critical" size was in agreement with the calculated size for the supersaturations used. From these tests it was concluded that the chamber has the proper characteristics for simulating marine stratus. Results The n u m b e r of cloud droplets in the polluted air is higher by a factor of three compared to that in the clean air. This is less than the increase in particle number. A substantial fraction of the particles larger t h a n the critical size do not grow, presumably since they are not soluble, see below. In order to compare results obtained with the cloud chamber with actual clouds, ECN cooperates with one of the leading i n s t i t u t e s (Brookhaven N a t i o n a l L a b o r a t o r y ) in the ARM p r o g r a m of US-DoE, in the e v a l u a t i o n of the microstructure of clean and polluted marine stratus off the coast of Nova Scotia. ECN has also participated in a recent cloud study near the English west coast. In special campaigns, droplets and aerosol particles were collected for chemical analysis. Unfortunately these campaigns were on days with continental air. In the polluted continental air, the number concentration of particles is about twenty times higher than that in the clean marine air, but the number of cloud droplets is
177 not proportionally larger. Droplet concentrations were typically 1100 per cm-3, independent of the aerosol concentration. This is indicative of a saturation effect of the cloud n u m b e r in continental air. This phenomenon is in part explained by the fact t h a t particles with the proper size are non-soluble and therefore cannot act as cloud nuclei. M e a s u r e m e n t of the actual amount of non-soluble particles is very difficult, because most cloud nuclei are small and thus contain very little mass. Preliminary conclusions1 2 9 The local "direct" aerosol effect can be assessed by m e a s u r e m e n t s and is thus easier to quantify t h a n the "indirect" effect. 9 Nitrate seems of more importance for the local direct effect t h a n sulfate. 9 Non-soluble, presumably carbonaceous, particles form a central factor, both in the direct and the indirect effect.
Assessment The project is one of the so called "late starters". Progress is questionable. No substantial new results are as yet available. The use of a cloud chamber seems promising. At this m o m e n t it cannot be judged if the goals of the project will be met. 2.3 C l o u d s - r a d i a t i o n - h y d r o l o g i c i n t e r a c t i o n s in a l i m i t e d - a r e a m o d e l
A.C.A.P. van Lammeren, A.J. Feijt, R. van Dorland, E. van Meijgaard, P. Stammes KNMI, Royal Netherlands Meterological Institute, P.O. Box 201, 3730 AE De Bilt
Introduction Clouds play an important role in our climate. Clouds produce precipitation which is an essential ingredient of the hydrological cycle. Clouds modify the earth-radiation budget. Thin cirrus clouds have a warming effect while low clouds have a distinct cooling effect ( R a m a n a t h a n , 1989). Clouds dominate the vertical t r a n s p o r t of energy, m o m e n t u m and trace gases in the free troposphere. Despite t h e i r importance, clouds are represented only rudimentary in climate as well as weather forecast models. It appears t h a t the model r e p r e s e n t a t i o n of clouds in climate models has a major impact on model predictions for climate change. Cess et al. (1989) showed t h a t cloud feedback is a major source of u n c e r t a i n t y in model responses to climate forcing. There are two main reasons why the uncertainties with respect to clouds are so large. The first reason is t h a t cloud processes are extremely complicated. A proper
1. In the p r e s e n t s u m m a r y the comments by the cluster coordinator on presentation of 19 May 1994 are incorporated. These comments centred on m a g n i t u d e a n d the u n c e r t a i n t y of the aerosol forcing r e l a t i v e to m a g n i t u d e / u n c e r t a i n t y of the forcing by clouds and on the values for anthropogenic forcing in recent literature.
the the the the
2. In a special investigation, which was not part of the European project, the local "direct" aerosol effect in November 1993 was studied. The data indicate a local daytime radiative forcing by the aerosol in polluted air of 20 W.m2.
178 representation of clouds requires the parameterization of subgrid processes both on t h e m a c r o s c a l e (cm - km) and on the microscale (<< cm). Cloud parameterization is still in its infancy. The second r e a s o n is the lack of good q u a n t i t a t i v e observations of cloud characteristics (cloud cover, cloud structure, optical depth, droplet spectra). This lack of good data h a m p e r s the development and validation of models. Satellites begin to provide useful data on global cloud statistics and corresponding radiation budgets (ISCCP, ERBE). However, these data sets still have to be validated. To validate these global data sets on a regional scale, detailed m e a s u r e m e n t s of the cloud cover and s t r u c t u r e are necessary. It is i m p o r t a n t to m e a s u r e the variability of the cloud characteristics within one datapoint of the global dataset. High resolution m e a s u r e m e n t s are also crucial for the d e v e l o p m e n t a n d i m p r o v e m e n t of p a r a m e t e r i z a t i o n schemes for clouds and cloud-radiation interactions. The aim of the project "Clouds-Radiation-Hydrologic interactions in a limited-area model" is threefold: 9 Provide a detailed regional dataset on cloud cover and cloud characteristics. 9 Analyze the ISCCP and ERBE datasets with respect to validity and possible applications for climate model verification. 9 Create a model environment for the enhancement of regional data analysis and for the improvement of parametrizations of clouds and radiation. These items will be described below.
Hybrid cloud detection system A proper description of clouds and cloud-radiation interactions in climate models requires the parametrization of the subgrid processes which determine cloud cover and cloud structure. To develop and improve these p a r a m e t r i z a t i o n s a set of measurements is required, so results of the model can be validated. At the KNMI a project is ongoing which yields accurate and high resolution cloud m e a s u r e m e n t s over the Netherlands. A hybrid cloud detection system in which both groundbased and satellite remote sensing instruments data are combined has been built. This cloud detection system is described in full detail by S t a m m e s et al. (1994). An outline is given below. The cloud detection system is used to retrieve the following cloud characteristics: cloud cover fraction, cloud top t e m p e r a t u r e , cloud base height, cloud base t e m p e r a t u r e , reflectivity and optical thickness; it provides information on the 3-dimensional structure of cloud ensembles. The cloud detection system consists of a n e t w o r k of s t a t i o n s for g r o u n d b a s e d remote sensing and a processing e n v i r o n m e n t for AVHRR and METEOSAT m e a s u r e m e n t s . In the 120x120 km target area a network of stations for groundbased remote sensing has been built. Each station consists of a LIDAR ceilometer, narrowband IR radiometer and a pyranometer. On two stations more extensive radiation m e a s u r e m e n t s are done. This provides data for the analysis of cloud-radiation interactions. In order to obtain a complete description of the geometry and height of the clouds the signals from the various instruments are correlated. O t h e r available tools for i n t e r p r e t a t i o n of the m e a s u r e m e n t s are data on the actual atmospheric conditions from model analysis and radiosonde. Radiative transfer calculations are performed using Lowtran-7 (Kneizys, 1988) for longwave
179 and DAK (Stammes, 1994) for shortwave radiation. AVHRR data will be analysed using the Apollo retrieval scheme (Saunders and Kriebel, 1988), which has been implemented at the KNMI. The use of state-of-the-art retrieval techniques is guaranteed by the international cooperation with other institutes with a tradition in satellite remote sensing, among others: D.L.R. (Oberpfaffenhofen), L.M.D. (Paris), Universit~it Berlin. The a) b) c) d)
following data is archived: Satellite instruments: NOAA/AVHRR, Meteosat 13 Groundstations: LIDAR, IR-radiometer and solar radiation Radiation stations: Direct, diffuse and total downward SW flux, downward and upward LW flux Other datasources: 3-hourly analysis data of the operational HIRLAM (High Resolutional Local Area Model) weather forecast model, 6-hourly rawinsonde data in De Bilt
The aim is to obtain continuous data on clouds and radiation for two years. The hybrid cloud detection system will also be used to create a dataset which will be compared with the results of the ISCCP algorithm. Global data sets
The E a r t h Radiation Budget Experiment (ERBE) dataset contains invaluable information about the radiation budget at the top of the atmosphere. The I n t e r n a t i o n a l Satellite Cloud Climatology Project (ISCCP) d a t a s e t contains information on clouds, their height, temperature, optical thickness, etc. Both global datasets are derived from satellite measurements and cover many years. E R B E . During the NPR project the literature on ERBE dataprocessing scheme was studied (Feijt, 1992). The main questions were: Can ERBE data be used for climate model validation and how can the data be used most reliably? Special a t t e n t i o n has been given to a s s u m p t i o n s about a t m o s p h e r i c and surface conditions, field-of-view homogeneity, spectral variability, time and space variability of cloud cover. It is concluded that the errors in radiative fluxes show much variation related to geographical and atmospheric conditions and the viewing geometry, the positions of the sun and the satellite relative to the area. Nevertheless, monthly mean values data are useful for climate model validation if used properly. Individual values should be used with caution, because the accuracy is highly dependent on the viewing geometry and surface/cloud type observed. It is advisable to users of the ERBE dataset to spend some time on the merits of the data. Likewise it would be good practise if global datasets which are to be used over a wide range of research areas were accompanied by an indication of the errors involved, written down in a way readable by the non-specialist investigator. ISCCP. In the course of 1993 parts of the ISCCP-cloud detection algorithm were implemented at the KNMI. Numerous case-studies have been performed on the merits of these modules for North-West Europe. This yielded a feel for the sensitivity of the algorithm for atmospheric conditions and possible related error-sources. Furthermore, the literature on the algorithm was studied in order to obtain insight into the coherent set of thresholds m a k i n g up the detection algorithm. Recently, Rossow (1993a, b, c) published a validation study on the cloud
180
detection algorithm. In general his findings agree well with our own investigations (to a p p e a r in KNMI technical report in 1994). The algorithm performance decreases with: increasing variability of the surface albedo and t e m p e r a t u r e , decreasing c o n t r a s t between clouds and surface and decreasing t e m p o r a l variability of the cloud fields. As the ISCCP algorithm is based on spatial and temporal coherence tests this is not surprising. In terms of geographic areas: The algorithm performs well over ocean, somewhat worse over landsurfaces and worst over polar regions. Moreover the algorithm is hampered by areas of persistent cloudfields and stormtracks. Two cloud types are missed frequently: low broken cloud fields and high semi-transparent cirrus clouds. Both cause little contrast in albedo and brightness temperature and are hard to identify especially over cold bright surfaces (polar regions) or highly variable land surfaces. The cloud climatology over the period 1982-1988 shows a global mean a n n u a l cloudcover of 63%, which is on the higher end of the range of values which are r e p o r t e d in l i t e r a t u r e . Rossow (1993c) states t h a t this value is still an underestimation, because the derived cloud fraction is too low about 10% over land and 10-15% in polar regions. The derived cloud fractions are approximately correct over ocean. The h y b r i d cloud detection system will provide a d a t a s e t which will be intercompared with the results of the ISCCP algorithm. This comparison will yield information on the merits of the basic assumptions and the used thresholds, and hopefully give directions for improvement of the ISCCP cloud detection algorithm. The m o d e l e n v i r o n m e n t A Regional Atmospheric Climate MOdel (RACMO) has been i m p l e m e n t e d at KNMI (Christensen and van Meijgaard, 1992). This is a descendant of the High Resolution Limited Area Model (HIRLAM) and the Global Climate Model developed at the Max Planck Institute in H a m b u r g (ECHAM). HIRLAM is the optional weather forecast model at the KNMI which employs a 50 km grid and 16 layers. RACMO inherited the dynamics and initialization environment from HIRLAM. All physics modules, including the liquid water and radiation modules, stem from the ECHAM model. There is a close cooperation with the research groups involved in the development of HIRLAM and ECHAM. The research environment supports flexible module management, so new versions of cloud and radiation modules can be activated and compared. The Mocrette radiation scheme is included in the RACMO (Morcrette, 1991). This scheme has been extended to include trace gases and aerosols. In addition to the Morcrette scheme a radiative transfer code for the shortwave spectrum was developed. This KNMI Doubling-Adding radiative transfer model (DAK, Stammes, 1994) is a line-by-line model which yields so-called exact results provided exact information on the atmospheric conditions and constituents is given. DAK was adapted to calculate the radiative properties of a plan-parallel cloud of which the user can define the cloud droplet size distribution and optical thickness. Assessment The cloud detection system has been completed and m e a s u r e m e n t s have been started. It appears t h a t the combination of ground-based and satellite-based
181 remote sensing provides unique information on clouds, amongst others on the small scale structure of clouds and on cloud bottom and cloud top. The ground stations may well serve as a prototype for a global cloud detection network. A global network consisting of 13 stations / 120x120 km (as was discussed in the section dealing with the Hybrid cloud detection system) must, however, be precluded in my opinion. An assessment of ERBE data has been completed. The strong and weak points of ERBE are highlighted. A similar assessment of ISCCP data will be finished this year. Major conclusions are given in the summary report. The model e n v i r o n m e n t has been created and is in its testing phase. This environment includes newly developed radiation codes, which account for the important radiatively active trace gases, for aerosols and for plan-parallel clouds (see s u m m a r y report). In s u m m a r y it may be stated t h a t this project is highly relevant for u n d e r s t a n d i n g the earths' radiation budget. The results seem very promising. International cooperation is an important aspect. 2.4 T h e e f f e c t o f c l o u d s o n u l t r a v i o l e t r a d i a t i o n a n d p h o t o d i s s o c i a t i o n r a t e s in t h e t r o p o s p h e r e
J. Vil~-Guerau de Arellano University of Utrecht, Institute for Marine and Atmospheric Research (IMAU), Princetonplein 5, 3584 CC Utrecht, The Netherlands Introduction Most chemical t r a n s f o r m a t i o n s in the a t m o s p h e r e are i n i t i a t e d by the photodissociation of molecules due to UV radiation. The intensity of the UV radiation field at different altitudes in the atmosphere is determined by the incoming solar flux, solar zenith angle, ground albedo and atmospheric composition. In the atmosphere the relevant processes include Rayleigh scattering, aerosol extinction, gaseous absorption (mainly by ozone) and the presence of clouds. Changes in these physical parameters will alter the amount of solar UV radiation available to dissociate of molecules in the atmosphere. Several studies (Tsay and Stamnes, K. 1992; Madronich, 1987; van Weele and Duynkerke 1993) have shown t h a t clouds are the main cause of changes in the radiation field throughout the troposphere. Since m a n y i m p o r t a n t chemical species such as ozone, hydrogen peroxide and nitrogen dioxide are photodissociated by ultraviolet radiation, it is of fundamental importance t h a t we should be able to u n d e r s t a n d and adequately estimate the effect of clouds on the radiation field and as a consequence the effect on tropospheric chemistry, both on a global and regional scale. Methods Within the framework of NPR-I, in 1990 IMAU started a study to determine this effect. The strategy of the study was based on a combination of radiometric m e a s u r e m e n t s and radiative transfer modelling. To carry out the measurements, we developed a photo-electric instrument to measure actinic flux (Van der Hage et al., 1994). The actinic flux is defined as the radiative flux incident on a sphere with unit area. With this radiometric quantity one can calculate photodissociation rates
182 of molecules and also estimate the biological effects. Vertical profiles were taken and surface measurements were made within the framework of the international experimental campaign ASTEX (Vil~-Guerau de Arellano 1994). During this campaign detailed measurements were made of cloud properties (liquid water content, cloud-base height, cloud-top height, effective radius, etc.) by other research institutes. These observations constitute a complete data base for the validation of radiative transfer models and the development of parameterizations. In addition, actinic flux surface measurements were made in Antarctica (to study the effect of surface albedo) and in De Bilt in collaboration with KNMI (Van Weele et al. 1994). The latter experimental campaign include measurements of UV-A and UV-B irradiances and global radiation. Results The comparison of these vertical profiles m e a s u r e m e n t s and surface measurements of actinic flux are in very good agreement with the model results calculated with the radiative transfer model based on the delta-Eddington approximation designed at IMAU, see Vil~-Guerau de Arellano et al. (1994). The m e a s u r e m e n t s and model calculations revealed that the key p a r a m e t e r s governing the effect of layered clouds on the UV radiation are: cloud optical depth and solar zenith angle. A parameterization based on the experimental and model results concerning the effect of layered clouds on the photodissociation rates is currently being implemented in the regional atmospheric chemistry model LOTOS at TNO (Van Weele et al., 1993). Assessment This project aims at the measurement of the influence of clouds on the profiles of the actinic flux. An instrument has been developed and the actinic flux has been measured during field campaigns. The expected enhancement of this flux at the cloud top and reduction below the cloud were reported. A quantitative comparison between observed data and radiative transfer calculations are in very good agreement. Much has been achieved in the project and results have been published in refereed journals. For the troposphere as a whole, the net effect of clouds on tropospheric chemistry, however, is expected to be small. Locally and temporary the production- and destruction rate of chemical species may be altered. 3.
THE ROLE OF A T M O S P H E R I C OZONE IN GLOBAL C H A N G E
3.1 I n t r o d u c t i o n Nearly all the earth's atmospheric ozone is found in the stratosphere (about 90%) and the remainder in the troposphere (10%). In the stratosphere ozone levels largely depend on competing (photo)chemical reactions. Ozone in the stratosphere is formed by photolysis of oxygen by short-wavelength solar radiation and destroyed by longer wavelength sunlight. Various trace constituents also react with ozone. Ozone formation and destruction processes are normally in dynamic equilibrium, however, when the concentration of constituents capable of destroying ozone increases, the balance is upset and a new equilibrium is reached which sustains less ozone. In the troposphere, next to transport of ozone-rich air from the stratosphere and destruction of ozone at the earth's surface, ozone is also being formed and destroyed by photochemical reactions in which volatile organic
183 compounds (mainly m e t h a n e and n o n m e t h a n e hydrocarbons), carbonmonoxide and nitogenoxides play a dominant role. Depending on the chemical regime (i.e. the NOx mixing ratio) there may be a net production or net destruction of ozone. The combined a m o u n t of ozone in the stratosphere and troposphere, known as total column ozone, occupies on average 300 Dobson units (or milli-atmospheric centimeters, equivalent to 3 mm of ozone at standard temperature and pressure). The main region of ozone formation is found at heights around 25 km in tropical regions. The level at which the maximum concentration of ozone occurs is much lower n e a r e r the poles t h a n n e a r the equator. For example, h i g h e s t ozone concentrations are found at about 18 km at 75, N and S, and at 15 km over the Arctic. The distribution of ozone around the globe is by no means uniform. Ozone at all latitudes other than around the equator shows a marked seasonal variation, with a m a x i m u m in spring and a minimum in autumn. The highest concentrations are found in the Arctic region in spring. The lowest concentrations occur at the equator, where there is little change of ozone with latitude or with season. Besides the regular seasonal variations, ozone values exhibit short-term, local variations of equal m a g n i t u d e which can take place within a few days. These variations are closely associated with meteorological conditions in the u p p e r a t m o s p h e r e , p a r t i c u l a r l y at higher latitudes. Finally the total ozone column also varies n a t u r a l l y from year to year at any particular location. The natural variability in ozone makes trends in total ozone column difficult to detect. From the point of view of live on earth, the most important property of ozone is its ability to absorb UV radiation emitted by the sun. UV radiation is harmful to h u m a n s and ecosystems, and is classified according to wavelength UV-A (320-400 nm), UV-B (280-320 nm) and UV-C (< 280 nm). The subdivision of the UV waveband into A and B is not unequivocally. Sometimes the division between AV-B and UV-A is set at 315 nm, and the UV-A region is sometimes defined as 315-380 nm. In this document, unless specifically s t a t e d otherwise, the aforementioned divisions will be taken. Atmospheric oxygen (and a small fraction of ozone) block all UV-C radiation completely at higher levels in the atmosphere. The ozone layer also absorbs most, but not all, of the UV-B radiation. Ozone absorption of UV-A radiation can be neglected, which means that ozone depletion is of little consequence where UV-A is concerned. The intensity of the UV-B radiation to which m a n and the ecosystem is being exposed depends on: 9 Sun's angle 9 Cloudiness 9 Stratospheric ozone column density 9 Tropospheric aerosol 9 Ozone content in the troposphere and in the PBL 9 Surface albedo 9 NO2- and SO2 total column The change in ozone column thickness with latitude and the difference in prevailing suns' angles combine to produce a very large gradient in natural solar UV from the tropics to high latitudes. At higher latitudes, at which e.g. The N e t h e r l a n d s is situated, also a large difference between winter and summer UV levels exists. As
184 shall be discussed later the period of maximum ozone depletion over the Northern Hemisphere, occurring in winter/spring, happens to coincide with the period when absolute UV irradiance values are at their lowest and therefore least harmful. Imposed upon the annual cycle in UV, are long term variations arising from trends in stratospheric and tropospheric ozone, in cloud cover and aerosols (which scatter and diminish UV) and upward trends in tropospheric pollution which absorbs UV. Ozone absorption is i m p o r t a n t in both ultraviolet and infrared p a r t s of the spectrum. The radiative forcing due to ozone depends on w h e t h e r it is in the s t r a t o s p h e r e or troposphere. The r a d i a t i v e forcing is g r e a t e s t n e a r the tropospause and so the ozone trends of most interest are those observed in the upper troposphere and lower stratosphere. M e a s u r e m e n t s of atmospheric ozone over the last t w e n t y years has shown a significant decrease of stratospheric ozone as depicted in Figure 3.1. In the northern Hemisphere the losses are most pronounced in winter and spring (6% per decade) and less pronounced in the other seasons (3% per decade). Trend in % per decade oo.oooO... . . . .
/ -2
_
-4
_
/
-6 -8
...'..X .................
o..O~
9
-..... -..
-12
,.,
..'s -
i
Dec-March
f
---
i.."
-10
~176176
...." ...."
-
__A
_-
............ Sept-Nov
~.."
ti
--
Mav-~un IkA
o
-14
"1
-60 S
I
I
I
I
I
I
-40
-20
0
20
40
60
equator
N
Figure 3.1 Ozone trends in % / decade in dependence of latitude and time of the year. Results are from TOMS-measurements taken in the period 1978-1991 Recently the trends have (temporarily) been increasing, with total ozone in 1992 and 1993 at exceptionally low levels world wide. Chemical and dynamical processes in the tropopause region have an i m p o r t a n t bearing both on total ozone, the vertical distribution of ozone and on the exchange of other trace gases. Some of the observed decreases in total ozone, especially at mid-latitudes, is simply caused by transfer of air masses containing markedly different amounts of ozone and are not the result of ongoing chemical destruction processes. A major fraction of the ozone
185 decrease, both long-term and in 1992/1993, has occurred in the lower stratosphere at altitudes starting almost at tropopause level up to 25 km. The reason for this behaviour is the occurrence of aerosols or PSC in this part of the atmosphere. The presence of PSC and aerosols can affect the chemical balance by providing surfaces on which chemical reactions eventually leading to ozone loss can proceed, an effect which is enhanced at lower temperatures. The anomalous behaviour of ozone in 1992/1993 has been attributed to the eruption of Mt Pinatubo in the Philippines in J u n e 1991 which injected large amounts of gaseous SO2 into the stratosphere which then chemically was transformed into sulphuric acid droplets i.e. aerosols. The state of knowledge regarding ozone trends in the troposphere is not good as in the stratosphere. There are a limited number of ozonesonde stations with records suitable for long-term trend analysis and even at these sites data quality remains questionable. The stations are located mainly in the Northern Hemisphere. Over the last 20-30 years the largest change in free tropospheric ozone seems to have occurred over Europe with an increase of about 50% since the end of the 1960s. The change over North America seems to be less. i.e. about 10-15%. The change has become less pronounced at the second halve of the 1980s, from whereon no change has been observed over North America and also over Europe the observed trends were smaller t h a n before; actually levels of tropospheric ozone over North America and Europe may now be on the decline. Ground-based m e a s u r e m e n t s both in the boundary layer and at mountain sites have revealed a significant rise in tropospheric ozone since pre-industrial times both in N o r t h e r n and S o u t h e r n Hemispheric sites (Figure 3.2). PPB 0 3 60
[] Cape Grim, 1985
-
...~i.~i~i!i!iiii!i!i!i!iii!ii!i!i!i!ii...... iiii~
9 Kloosterburen, 1980-1984 + 1987-1989 O Cordoba, 1886-1892
50 -
$ Montsouris, 1876-1886 ................................................................ .... ::~,,~,~,~,,,~,,,~,,~,,,~,,,~,,,~,~,~,,~,~,,,~,~,~,~,~,~,~,
:~iiiii!iiiiiii!ii!i!iiiiiiiii!iii!ili!iiii!iliiiiiiiiiiiiiii!iiiiiiiiili!~ ~....
................................................................... ........~iiiii!i!iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii!iiiii!ili}iiiii!iiiiiiiii~ ....
~iiiiiii!iiiiii!!iiiii!i!i~iiiiiiiiiiii!!i!ii~!!i~!!i!~iiiiiii~iiiii!iiiiiii~ ..... ~iiiii!ii!~ili;iiii ....................................... ~i~i~i~i~i~i~.......... i~!~i~i~i~
40 1980 30 1980 20 1900
10 -
0
I
J
I
F
I
M
I
A
I
M
I
J
I
J
I
A
I
S
I
O
I
N
D
Figure 3.2 Ozone t r e n d s in the free troposphere. Also depicted are ozone levels over continental mid-western Europe
186 Observations conducted in recent years have confirmed that tropospheric ozone has increased at many stations in the Northern Hemisphere. In the Southern Hemisphere tropospheric ozone seems to have remained constant or may have even decreased (Figure 3.3). The latter tendency now also seems to be the case at stations in the Northern Hemisphere. Ozone Trend [%/yr) 3 wIZ
2
A
A
H~
1
0
-1
7~
I
I
-90
-60 South
A
I
-30
I
I
I
i
0
30
60 North
90
Latitutes
Figure 3.3 Trends in tropospheric ozone observed at different latitudes. Only coastal and high-altitude observatories were included. South Pole 90~ 2800 m a.s.l., start 1975; CP: Cape Point, 34~ 1982; S. Samoa, 14~ 1975; ML: Mauna Loa, 20~ 3400 m, 1973; WM: Whiteface Mountain, 43~ 1600 m, 1973 Z: Zugspitze, 47 ~ 3000 m, 1978; W: Wank, 47~ 1800 m, 1978; HP: Hohenpeissenberg, 48~ 1000 m, 1971; A. Arkona, 53~ 1956; B: Barrow, 70~ 1973. The solid line is a linear fit through the data, excluding the sites in southern Germany (HP, W and Z). Source: Volz-Thomas (1993) The role of atmospheric chemistry in the steady state chemical composition of the atmosphere is of great importance in order to assess the associated radiative forcing on climate. The atmospheric abundance of a component is governed by its rate of production and removal. A number of atmospheric feedback mechanisms are known which have the potential to change the atmospheric removal rate, principally through perturbation of the UV radiation field and most notably the global hydroxyl field. Even an chemical species like CO2, which is regarded as being "inert" influences atmospheric chemical processes. CO2 tends to cool the stratosphere. This has an effect on rate constants of chemical reactions, in such a way that stratospheric ozone levels may increase. This will lead to a negative feedback of temperature p e r t u r b a t i o n s in the stratosphere. On the other hand lower stratospheric temperatures potentially favour the formation of PSC. This may worsen lower stratospheric ozone depletion and allowing a deeper penetration of UV radiation into lower altitudes.
187 For a reactive compound such as CH4 the relations are far more complex. An increase in CH4 emissions will lead to a tropospheric ozone increase, notably in northern midlatitudes during s u m m e r and in the tropical upper troposphere. The tropospheric OH abundance on the other hand will decrease, leading to a longer residence time of chemically and/or radiatively active trace gases such as CH4 itself, CO, HCFC's, etc. In the s t r a t o s p h e r e , an increase in w a t e r v a p o u r a b u n d a n c e and significant perturbations of ozone and chlorine species can be expected. The reason is t h a t H20 is the source of odd hydrogen (HOx), which accounts for ozone destruction, notably below 30 km and above 50 km. At other altitudes NOx is the domant species where ozone loss is concerned. In the presence of HOx, more NOx is converted in less reactive HNO3 and HO2NO2; consequently, an ozone increase can be expected at this altitude range. Finally CH4 reacts efficiently with chlorine atoms to form the less reactive reservoir species HC1, counteracting ozone destruction by CFC's. An important feedback mechanism that affects the lifetimes of m a n y atmospheric constituents is the change of UV radiation reaching the troposphere due to changes in the stratospheric ozone column density. Reduction in stratospheric ozone leads to enhanced penetration of UV radiation into the troposphere. This in t u r n will yield increased production of O(1D) and OH and enhance reactions t h a t produce or destroy tropospheric ozone. Whether this results in a net increase or decrease of tropospheric ozone depends on the NOx mixing ratio. Chemical and physical processes occurring in clouds and aerosols play a major role in d e t e r m i n i n g the chemical composition and the radiative properties of the a t m o s p h e r e . Aerosols also affect the e a r t h ' s h e a t b a l a n c e d i r e c t l y by backscattering of shortwave radiation and, indirectly, by their influence on the reflectivity of clouds. Clouds are considered to be an important removal pathway for some atmospheric trace constituents, and the lifetime of these compounds are determined to a large extend by the frequency of encounters with precipitating clouds. For example in-cloud oxidation of SO2 accounts for about 50% of the total SO2 sink on a global scale. In some areas, direct interception of clouds or fogs with the surface of the earth is an important removal pathway. Moreover, u p d r a u g h t s within clouds are important means of vertical t r a n s p o r t of atmospheric constituents, particularly between the PBL and the free troposphere. In strong convective systems, of the cumulus nimbus type, trace constituents may even be transported into the lower stratosphere. This subtheme comprises 4 projects.
188
3.2 D u t c h participation in the international n e t w o r k for the d e t e c t i o n of stratospheric change (NDSC) D.P.J. Swart N a t i o n a l I n s t i t u t e of Public Health and E n v i r o n m e n t a l Protection (RIVM), P.O. Box 1, 3720 BA Bilthoven, The Netherlands
Introduction The Network for Detection of Stratospheric Change is a set of 5 high-quality, remote sounding research stations for observing and understanding the physical and chemical state of the stratosphere. These stations, where ozone and key ozone-related p a r a m e t e r s are measured, are complemented by both secondary stations and satellite m e a s u r e m e n t s . Currently, over 65 scientists from 15 countries are involved with NDSC. The NDSC is a major component of the upper atmosphere research effort and has been endorsed by national and international scientific agencies, including the International Ozone Commission, The United Nations E n v i r o n m e n t a l P r o g r a m m e (UNEP), and the World Meteorological Organisation (WMO). The goals of NDSC are: 9 To make observations through which changes in the physical and chemical state of the stratosphere can be determined and understood. In particular to m a k e the earliest possible identification of changes in the ozone vertical distribution and to discern the cause of these changes. 9 To provide an independent calibration of satellite sensors of the atmosphere. 9 To obtain data t h a t can be used to test and improve m u l t i d i m e n s i o n a l stratospheric chemical models, thereby enhancing confidence in the predictive and assessment capabilities of these models.
Project description The NRP project described in this abstract comprises the development of a stratospheric ozone lidar, one of the primary NDSC research instruments, and implementation of this instrument at one of the 5 NDSC primary sites, the station of Lauder in New Zealand. This groundbased instrument is capable of obtaining a vertical concentration distribution in the altitude range of 15 to 45 kilometers. Vertical resolution is 1 kilometer at 20 kilometers altitude, increasing to 5 kilometers at 45 kilometers altitude. Measurement error is typically about 1% at 20 km, increasing to 10-20% at 45 kin. These figures hold for a single vertical profile obtained in a measurement time of about I hour. Errors are of a statistical nature, and will decrease when a number of profiles is averaged in e.g. a monthly or yearly mean profile. As of this writing, June '94, the development and preliminary validation are nearly completed, and the system will be shipped to New Zealand in August-September. It will be implemented in the network before the project expires at the end of the year. A typical profile obtained by the lidar is given in F i g u r e 3.4. A balloon profile t a k e n in Uccle, Belgium is also shown for intercomparison.
189
Ozone profiles 14 February 1994
Ozone (molecules / m 3) 7E+18 6E+18
-
1
5E+18
3E+18
_
t
I
II
- - KMI sonde ....... RIVM Lidar near channel, 2 km - - RIVM Lidar far channel, 5 km
i'
I
2E+18 -
"~.~. ~ ~ '," ~
J v"".. '..",
'',
w
;j i.....~
..~ 4E+18
I.:."
,,
'
,
,~
\ ~ -I
i ~'
~ .~.,/
,
1E+18
0
I
1
t
I
I
I
I
5000
10000
15000
20000
25000
30000
35000
]"~ 40000
~-" 45000
Altitude (meters) Figure 3.4 Ozone profiles measured at February 14, 1994 with lidar and sonde. Lidar data were taken in Bilthoven, the Netherlands, at 24:00 MET. Sonde data were taken in Uccle, Belgium at 12:00 MET (de Muer and de Backer, private communication). Uccle and Bilthoven are separated by approximately 160 km
International collaboration This project has provided us an excellent opportunity to enter the international research on stratospheric ozone and related species, a field virtually non-existent in the Netherlands only a few years ago. During the development of the instrument we have established close links with researchers and i n s t i t u t e s t h a t have lidarprograms in this field: NASA and NOAA in the United States, NIES and MRI in J a p a n , York University in Canada, CNRS in France, to mention some. In the framework of the NDSC network, international relations are to n u m e r o u s to indicate all individually, and range from universities and national environmental research institutes to political bodies like WMO and UNEP. A special set of relationships is generated by the international team of principal investigators with i n s t r u m e n t s at the Lauder station, with e.g. NIWA from New Zealand, and again different researchers from the US, Australia, J a p a n and Italy.
Project results and spin-off By the end of this year, the project will have met the goals we set out for five years ago: 9 A state of the art measuring lidar i n s t r u m e n t for stratospheric ozone has been developed successfully. 9 We have succeeded in obtaining a strong position for this instrument in a (the) most prominent international stratospheric monitoring network. The system will be part of a multicomponent, multiyear daily monitoring scheme.
190 9
9
We will contribute to the NDSC database, cohosted by NOAA, NILU and Rutherford Appleton Laboratory. We will take part in the establishment of a long term dataset on the status of the stratosphere. As a contributor, we will have early access to all NDSC data worldwide. We have established a firm position and good contacts in the international research community in this field.
The future As the NRP project draws to an end, the work with the developed lidar will enter a new phase. A scientific collaboration has been set between RIVM and the Free University of Amsterdam (VU) and a Dutch PhD student will work in New Zealand on r e s e a r c h with this i n s t r u m e n t . A rigorous intercomparison and quality assurance program will be conducted using groundbased lidar (NASA), balloon sondes (NIWA) and microwave instruments (NOAA). A routine monitoring scheme will be implemented, and a database will be generated for climatological purposes and trend analyses. The system will take part in the validation of the performance of the GOME i n s t r u m e n t on ESA'S ERS-2 satellite to be launched early 1995. Many other scientific uses of the system are foreseen. Assessment This project is technologically and scientifically of high quality. A state-of-the-art m e a s u r i n g lidar i n s t r u m e n t for s t r a t o s p h e r e ozone has successfully been developed. International cooperation via NDSC is ensured. The goals of the project will certainly be met. In order to be a complete success funds for prolonged m e a s u r e m e n t s (10-20 years) should become available. 3.3 S p e c t r a l UV r a d i a t i o n atmospheric parameters
measurements
and
correlation
with
F. Kuik KNMI, Royal Netherlands Meterological Institute, P.O. Box 201, 3730 AE De Bilt, The Netherlands
Introduction The goal of the research of the KNMI in the NPR-I project was to develop a UV-monitoring network, to study the relation between UV and meteorological parameters like ozone and clouds, and to analyse trends in the UV-climatology in the N e t h e r l a n d s . For this purpose U V - m e a s u r i n g i n s t r u m e n t s have been developed. Theoretical research was performed using a spectral radiative t r a n s f e r code coupled to an atmosphere model. Both the UV-measurements and the theoretical methods are discussed, and some results are shown. UV-measurements The UV-measurements consist of spectral and narrow band measurements. Both types of measurements will be discussed.
Spectral UV-measurements. Are required to determine the relation between ground level UV and various kinds of atmospheric constituents. For example, 03 and SO2
191 exhibit similar wavelength dependent absorption structures in the UVB, although the optical thickness for both species is quite different. Because the absorption structures are roughly 2 nm wide, spectral m e a s u r e m e n t s should be performed with a wavelength step of maximal 0.5 nm with a resolution of 0.1 nm or better. Extinction by aerosol is assumed to be continuous but different in strength in the UVA and the UVB. To distinguish extinction due 03 and aerosol, m e a s u r e m e n t s should start below 295 nm and reach far into the UVA. For such spectral m e a s u r e m e n t s a diode a r r a y detector coupled to a single m o n o c h r o m a t o r was employed. This i n s t r u m e n t took p a r t in the E C - S T E P intercomparison of UV measuring instruments in Garmisch-Partenkirchen in July 1993. Compared to the double monochromators equipped with photomultipliers, the diode a r r a y i n s t r u m e n t behaved as good as could be expected. The limited dynamical range of approximately 103 (cf. a photomultiplier having a dynamical range of 106) and the poor straylight rejection make it possible to perform m e a s u r e m e n t s between 300 and 420 nm. For measurements below 300 nm the straylight signal exceeds the instrumental noise. In Figure 3.5 m e a s u r e m e n t of the spectral UV taken in Garmisch-Partenkirchen are shown. The curve denoted by AI was recorded with the double monochromator of the group from Austria, Innsbruck. This i n s t r u m e n t was considered as 'the s t a n d a r d ' together with 2 other instruments. It is seen t h a t the UV spectrum obtained with the KNMI diode array instrument deviates at most approximately 10% from the AI spectrum. Using a better UV filter or an alternative method to substract straylight, this may be improved to a deviation smaller t h a n 5%. This will be tested in an intercomparison to be held in Bilthoven in August 1994, in which KNMI, RIVM and the AI group will take part. 1.5
'
'
'
I
'
'
'
I
'
'
'
I
'
'
'
I
'
'
'
I
'
'
'
I
_..
:~
,
I
'
KNMI 9 ~-- 10
-
0.5-
,."
p~"
~:
,.,~" ~
L. ~
~
~. 9 .
t i-'~i 0.0 ................... !, , 2!)0 310 330
Figure 3.5
,
,
I
350
,
,
,
I
,
370
,
,
I
390
,
,
,
410
Wavelength (nm)
UV spectra measured with the KNMI diode array i n s t r u m e n t and the AI double monochromator (Garmisch-Partenkirchen, ,July 24, 1993, 12:48 UT)
192 In December 1993, Brewer #100 arrived at the KNMI. This is a s t a n d a r d i n s t r u m e n t for continuous O3-column measurements, and since this i n s t r u m e n t is the version equipped with a double monochromator, it is particularly suitable for U V B - m e a s u r e m e n t s as well (a s i m i l a r i n s t r u m e n t also took p a r t in the G a r m i s c h - P a r t e n k i r c h e n intercomparison and performed so well t h a t is was considered as one of the three s t a n d a r d i n s t r u m e n t s . Since J a n u a r y 1994 this i n s t r u m e n t is continuously measuring the ozone column in The Bilt. In Figure 3.6 Brewer ozone m e a s u r e m e n t s for 1994 for The Bilt are shown together with TOMS m e a s u r e m e n t s for The Bilt averaged over the period 1980-1992. In between the ozone m e a s u r e m e n t s the spectral LW-irradiance is m e a s u r e d from 286.5 to 365 nm w i t h a 0.5 nm step. This i n s t r u m e n t thus provides almost s i m u l t a n e o u s m e a s u r e m e n t s of the O3-column and the UV-irradiance, so t h a t the relation between these two quantities can be ideally correlated. The Brewer #100 is our p r i m a r y ozone and UV m o n i t o r i n g i n s t r u m e n t . It is i n t e n d e d to produce semi-annual KNMI-reports with combined UV and ozone measurements. In Figure 3.7 m e a s u r e m e n t s of the UV-irradiances of F e b r u a r y 19 and 20, t a k e n at 12:05 UT on both days, are shown. There was a large difference in the ozone column on these days: 386 and 431 DU for February 19 and 20, respectively. Also the atmospheric aerosol load on F e b r u a r y 20 was much greater t h a n on the 19th. The difference in the UV-irradiances for wavelengths larger t h a n approximately 330 nm is due to the difference in aerosol, whereas for wavelengths below 330 nm it is the combined effect of differences is ozone and aerosol.
~"
450
I-T
"-~ 400
' ..,:,..111'lllllfl~
- - - T O M S avg. '80-'92
~.,!z!~ ~'.,.."~,~, ~ .... ~ii ~ii~}%.,,..,
~t,4".,-:
~ 350 '~ !,
'~';,~,,:
-
o
0 300
250
~"9..,,f..,,, ~ ~"~ ~ ~, .~'~ ~ " t ~ ' _
0
60
120
180
240
300
360
Day number M e a s u r e m e n t s of the ozone column (in Dobson Units) m a d e in The Bilt with Brewer #100 compared with TOMS satellite m e a s u r e m e n t s for The Bilt averaged over the period 1980-1992 Figure 3.6
193
0.25
il._ eb. 19,1, ,--,"" . 0.20 E LiL Feb. I 20 , ~, ~ L~~~i ! ! ~ ' :~l~ ~04,
E 0.15
o 010 ._~ -u 0.05 c--
LJ
".]'?!
':3
-
#
m
0.00 ~ 300 310
......... 320 330 340 350 360 Wavelength(rim)
o
Figure 3.7 Measurements of the UV irradiance for February 19 and 20, 1994, 12:05 UT. The ozone columns were 386 DU and 431 DU for February 19 and 20, respectively, and on February 20 there was much more aerosol in the boundary layer Narrow band U V instruments. Spectral UV measuring instruments are difficult to operate and expensive. To set up a network of UV instruments in The Netherlands more simple i n s t r u m e n t s are needed. Therefore narrow band i n s t r u m e n t s were developed aside the spectral UV instrument. Specifications and requirements for the instruments were formulated by the science department of KNMI, and Kipp & Zonen (Delft, The Netherlands) constructed the prototypes. After testing the instruments extensively for a period of 3 years they will be installed in The Bilt in the middle of 1994 as regular UV-monitoring instruments.
Two types of measuring devices have been chosen: irradiance and direct solar UV meters, one in the UVB and one in the UVA wavelength region. The irradiance meters record the vertical UV-flux, similar to the s t a n d a r d pyranometer. The direct UV m e t e r is mounted on a solar tracker and detects the UV-radiation coming from the direction of the Sun. From a combination of the vertical flux and the direct measurements, the ratio between the amount of scattered and directly t r a n s m i t t e d UV is determined. This ratio typically depends on the a m o u n t of aerosol in the atmosphere in the UVA, and again in the UVB absorption by ozone, especially for the direct UVB m e a s u r e m e n t s , is important. When these UVB m e a s u r e m e n t s are combined with the ozone measurements of the Brewer, effects due to ozone can be eliminated. This way it will be possible to really distinguish absorption by ozone and scattering by aerosol in the UVB. The sensors basically exist of a UV-sensitive photodiode, a UV-interference filter, t e m p e r a t u r e stabilization, and some electronics. The filters can be chosen freely, and for the UVA we use a filter with a central wavelength of 367 nm and a full width half maximum (FWHM) of 10 nm. In the UVB a narrower filter is needed to avoid errors in the measurements because of the steepness of the UV spectrum in this w a v e l e n g t h region. At his m o m e n t the UVB i n s t r u m e n t s contain an
194 interference filter with a central wavelength of 306 nm and a F W H M of 4 nm. However, Kipp & Zonen recently found filters with the same central wavelength, but with a F W H M of 2 nm. These filters will be installed and tested as soon as possible. There are several ways of analyzing the m e a s u r e m e n t s of the n a r r o w band instruments. One way is to determine the ratio between the direct and the diffuse (= scattered) UV-radiation in the UVA and UVB. Moreover, combining the data from the narrow band i n s t r u m e n t s with the spectral UV m e a s u r e m e n t s and the ozone measurements can provide additional information on the aerosol contents of the atmosphere. Combining all m e a s u r e m e n t s also provides the possibility to check the calibration consistency of the instruments. Another simple application of the direct and irradiance measurements is that they immediately show whether or not it was a cloudless day. However, due to lack of time and man power, analysis of the data has to be postponed to a later time.
UV radiative transfer and atmospheric modelling For UV radiative transfer calculations we use the Doubling/Adding method KNMI (DAK). This is an 'exact' multiple scattering model, which at the moment is coupled to an a t m o s p h e r e model. In the model approximately 32 layers are used for atmospheric gases for Rayleigh scattering (0 to 100 km altitude). Up to now aerosol optical properties and profiles have been used from the LOWTRAN model atmosphere because better data for The Netherlands is not available at this time. Profiles and columns of the UV-absorbing gases ozone and NO2 are already t a k e n into account in the model, and gases such as SO2 and 02 will be added in the future. As an example of a comparison of a m e a s u r e m e n t and calculations, Figure 3.8 shows a m e a s u r e m e n t of the UV irradiance recorded by the AI-group during the Garmisch P a r t e n k i r c h e n intercomparison (July 1994), together with 2 computed 1-61: . . . . .
i .....
i .....
i .....
i .....
i .....
i,,
",:-. 1.4 ~
. . . . . . rural aerosol I 7............ urban a e r o s o l I
,,;.§
9 ~'~ " ""
'
(~, ;..~:'r ,;, .~ ,. I$ I
1.2
=E" 1.0
~08
"
8
,;i
0.6
:'~ "
'
~" '~-" . . . . . .
9 ~,~ ~i::!.'i'.
0.4
0.2
0.0
290
320
350
380 410 W a v e l e n g t h (rim)
440
470
500
Figure 3.8 Measurements and calculations of the UV irradiance for Garmisch Partenkirchen, J u l y 24, 1994, 12:48 UT. The ozone column was 285 DU and in the calculations only the boundary layer aerosol was varied
195 curves: one for an u r b a n and one for an rural boundary layer aerosol. From comparing such model calculations with measurements we learned that especially variations in b o u n d a r y layer aerosol can have effects on the UVB irradiance similar in magnitude to those due to the daily variation in the ozone column.
KNMI UV-calibration laboratory It is a well known fact in the UV-measuring world t h a t calibration of the instruments is a difficult task and requires special equipment. Because the KNMI has planned to continue the UV-research, and calibration is so important, an investment has been made by the institute to realize a high quality UV calibration laboratory. The design of the lab is flexible so that several types of UV i n s t r u m e n t s can be calibrated. It measures 5x6 m, is divided into a lamp room and an instrument room (both separately airconditioned), and will be completely painted with a special absorbing black paint. Also a special high accuracy lamp power supply together with calibration lamps will be installed, and the calibrations will be performed and registered by a computer. According to the plans this facility should also become available for calibrations for non-KNMI work.
Assessment During NPR-I facilities of a technologically high level have been created to monitor ground level UV in The Bilt in various ways, producing a fairly complete data set. At the end of NPR-I all measuring equipment will be beyond the experimental phase and become operational. Also facilities for proper UV-calibrations and experiments will then be available. The KNMI radiative t r a n s f e r code (DAK) has been a d a p t e d for use in the UV-region and various p a r a m e t e r s such as profiles and columns of 03, NO2, aerosols and aerosol optical properties, t h a t affect the UV t r a n s f e r in the atmosphere have been investigated. However, the model needs improvement, and also a better characterization of the atmosphere is needed to compare local measurements with computed UV-spectra. This work is still in an initial stage. M o n i t o r i n g of UV i r r a d i a n c e s and ozone, combining m e a s u r e m e n t s with computations, etc., is long term work and requires a lot of time in the near future. The main result of NPR-I thus can be summarized by stating that the required measuring facilities have been realized, and now the phase of measuring and data analysis has arrived. International cooperation is an important aspect. Cooperation on a national level with RIVM (see next project) should be strengthened. With respect to radiative t r a n s f e r modelling in general, it m u s t be recommended t h a t groups in The Netherlands who use and/or develop these models (IMAU, KNMI, RIVM) should closely cooperate.
196
3.4 Determination of the UVB climate in The Netherlands: high resolution s p e c t r a l m e a s u r e m e n t s , m o n i t o r i n g and m o d e l l i n g from the perspective of risk analysis H. Slaper, H. Reinen, E. Schlamann, J. Bordewijk, N a t i o n a l I n s t i t u t e of Public H e a l t h and E n v i r o n m e n t a l Protection (RIVM), P.O. Box 1, 3720 BA Bilthoven, The Netherlands
Introduction Increased levels of biologically effective UV could lead to a variety of harmful effects for m a n and the environment, among which: skin cancer, impairment of the i m m u n e system, reduced biomass production by phytoplankton, and changes in aquatic and terrestrial ecosystems and food chains (UNEP 1991, Gezondheidsraad 1993). F u r t h e r m o r e tropospheric photochemistry could be influenced by changes in UV-irradiances. Therefore knowledge of the UV-climate and changes caused by h u m a n activities is highly relevant. Risk assessment Research at RIVM is focussed on integrated risk assessment of environmental changes. An integrated source risk model was developed for the effects of ozone depletion, the UV-chainmodel, which was applied for the evaluation of skin cancer risks in relation to various halocarbon production scenarios (Slaper et al 1992, RIVM 1993 ). Skin cancer is already one of the most common cancers among the Caucasian population, with 15-20 thousand cases each year in The N e t h e r l a n d s (around 1200 cases per million year). The model predicts an increased level of yearly effective UV doses in The Netherlands due to decreases in ozone. Around the year 2000 the increase in yearly effective UV at ground level is estimated to a m o u n t to approximately 15% (Slaper et al 1992 ). After the year 2000 a slow recovery could occur if all countries in the world fully comply with the Copenhagen Amendments of the Montreal protocol for a reduction in production and emission of ozone depleting substances. The consequences of the increased levels of UV can, at present only be estimated quantitatively for skin cancer. Elevated risklevels are expected to peak approximately 40 years after the peak exposures, thus reflecting the delay between exposure and the occurrence of tumours. The additional number of cases of skin cancer in The Netherlands associated with the increased UV-levels amount to around 200 new cases per million per year for the Copenhagen scenario (Slaper et al 1994, Slaper 1994). In terms of increased mortality an additional 2-3 per million per year is estimated for the Copenhagen scenario (RIVM 1993). An excess death risk of one per million per year is often referred to as the Maximum Permissible risk level for chemical substances and (ionizing) radiation sources by policymakers in The Netherlands. Such a risk level is associated with an increase in surface effective I.W of 5% (3-6%, depending on the different estimates for the melanoma risks). It should be noted that it is not a priori clear t h a t skin cancer is the p r i m a r y adverse effect. Nevertheless, changes in effective UV-exposure of no more t h a n 5% are relevant from the perspective of skin cancer risks. Relevance a n d objectives o f UV-measurements I n t e g r a t e d models are by definition simplifications and extrapolations, and therefore need to be verified and updated by means of monitoring and detailed analysis of UV-transfer in relation to atmospheric changes. When aiming at risk
197 assessments for the UV-related effects of man made changes in atmospheric composition it is highly relevant to investigate the biologically relevant UV irradiances and the relation between the UV irradiance and atmospheric parameters. Ozone measurements and monitoring has been established at many sites and by satellites, however, a worldwide network of spectral l.W-monitoring is lacking (UNEP, 1991). This lack of UV-monitoring data, especially the lack of spectral data, is generally recognized as a gap in the present scientific knowledge. During the past five years several (inter)national initiatives have emerged to improve the situation. RIVM and KNMI participate in a project from the EC STEP and Environment program. That project focuses on international standardisation of calibrations and intercomparisons of measurement systems, since at present no standardized measurement technique is established. The UV-monitoring project at RIVM, which is supported by the National Research Program, contributes to the specification of systems for future UV-monitoring networks and provides high quality spectral monitoring data. The main objectives of the RIVM-project are: 9 Design and building of a spectral monitoring system capable of measuring the biologically and photochemically relevant solar UV at groundlevel in the wavelength range: 295-450 nm. 9 High resolution spectral measurement of present UV-irradiances in The Netherlands. 9 Validation of UV-transfermodels and determination of primary important atmospheric parameters.
Measurement system The RIVM UV-monitoring system is designed to obtain data that could be used (both now and in the future) for effect evaluations and model validation. Specifications of the system are directly derived from these goals in a definition study (van Sonderen, et al, 1991). In order to be able to evaluate a variety of effects, and in view of the strong wavelength dependence of biological effectiveness and atmospheric UV-absorption, knowledge of the spectral distribution of the UV is required. The biological effectiveness of sunlight is primarily determined by the UVB. The solar irradiance in the UVB range changes over more than 6 orders of magnitude due to the strong wavelength dependence of the ozone absorption. This puts high demands on measurement systems with respect to dynamic range, linearity, stray light rejection and spectral resolution. From the monitoring perspective long term stability and reliability are important, and in order to reduce effects of changing atmospheric conditions during a recording of the spectrum the duration of a measurement must be minimized. It was concluded that the optimal spectral system consists of two complementary instruments" 9 A scanning spectroradiometer with a double monochromator and photon counting detection for the UVB region of the spectrum. 9 A spectrograph with photodiode array detection for the longer wavelengths in the UVB and the UVA. In addition to these two spectral instruments two broad band detectors were added: a pyranometer that records integrated irradiances from 300-3000 nm, and a Roberson Berger biometer which records the erythemally (sunburn) weighted UV
198 dose. These integrated measurements are used to determine changes during a scan and to fill in gaps in the monitoring data set.
Measurement protocol The monitoring system is operational since april 1993. The first year is used to improve and establish the system and operational procedures. The RIVM monitoring system participated in a European intercomparison in J u l y 1993 in G a r m i s c h P a r t e n k i r c h e n . Daily sequences of m e a s u r e m e n t s are t a k e n since October 1993. The m e a s u r e m e n t s are collected fully automatic each twelve minutes during day time (Reinen et al, 1993). The spectral structures of the m e a s u r e m e n t s with the RIVM-monitoring system reproduce the Fraunhofer structures found in the extraterrestrial spectrum (see Figure 3.9a). A technique was developed to use the Fraunhofer structures for an internal wavelength calibration. This technique is applied to all measured spectra and serves as a validation criterium. It was found t h a t the stability of the w a v e l e n g t h calibration is generally within 0.1 nm. For irradiance calibration NIST-calibrated 1000 W lamps are used.
Results a n d d a t a analysis Evaluation of effects requires the calculation of effective UV-irradiances. Effective irradiances are obtained by multiplication of spectral irradiances with the proper action spectra, and a subsequent integration over the effective wavelength region. Action spectra are sets of weighting factors reflecting the effectivity of various wavelengths. Many action spectra have a shape similar to the one shown in Figure 3.9b. It should be realized, however, t h a t effective doses/irradiances are strongly dependent on the exact shape of the action spectra, and can therefore be different for different effects. F i g u r e 3.10 shows the diurnal variation of effective UV (weighted with the STESLA-action spectrum (Slaper 1987)) for two days in October. October 29 (1993) was a cloudless day and the m e a s u r e m e n t s compare very well with modelcalculations for clear day situations, applying ozone values as m e a s u r e d by KMI in Uccle for that day. October 21 (1993) had some variation in cloud cover, as is clearly shown by the reduction in I.W during parts of the day. The s p e c t r a l m e a s u r e m e n t s are used to s t u d y the influence of v a r i o u s a t m o s p h e r i c p a r a m e t e r s . Factors of p r i m a r y i m p o r t a n c e for the effective UV-irradiance are: solar height, ozone, cloud cover and aerosol content. It was found t h a t on average cloud cover and aerosol load reduced clear sky effective UV irradiances by 20-30%.
199 Irradiance
(m Wlm2. nm) ............... ,..,.. ,... ~ ~ ,,....,""~,...:~./..fl"*; ,. /"... ~ "..... " ~.......:....,...,..,s ...... '. / '.l'.'":"" ".. ".. . . . "- "* '. ~ ", :'..'v'.. ,. , :.:"'.. ~" ~ v . -. A./~N ~/V, -~ .~/ WA I I
103
,.
"
i]v
VVl
102
101
100
_
10-1 280
]
......... Extraterrestrial solar spectrum
f
Solar spectrum (RIVM)
i
I
I
I
I
I
300
320
340
360
380
Wavelength
400
(nm)
Figure 3.9a E x t r a t e r r e s t r i a l and surface solar UV-spectrum, and effective UV
Irradiance
103
(m W/m2.nm)
~176 ......... ~176 ~
102
101 Solar spectrum (RIVM) .......... Action spectrum (SCUP) Effective UV 100
10-1 !_ 280
300
320
340 Wavelength
Figure 3.9b Sample solar UV-spectrum and effective UV
360 (nm)
380
400
200 Effective UV-irradiance (mW/m2.nm) 40 9 35 - O
RIVM 2 9 - 1 0 - 1 9 9 3 (cloudless) RIVM 21-10-1993 (variable clouds)
30 25201510 5 0
"O t t t i I I t t 08:00 09:00 10:00 11:00 12:00 13:00 14:00 15:00
Time (UT)
Figure 3.10 Diurnal variation in effective UV Clouds and aerosols reduce both effective UV and visible light. Therefore, we expected a correlation between a cloud/aerosol induced reduction of ground level irradiances of the pyranometer and spectral l.W-measurements. This relationship could be used to add an empirical correction to model calculations for cloudless conditions, in order to estimate the effective UV. The first analysis shows some promising results (Reinen et al,, 1994, Bordewijk et al, 1994). F u r t h e r analysis is needed to s u b s t a n t i a t e these findings. The observed relationships provide an o p p o r t u n i t y to e s t i m a t e effective U V - i r r a d i a n c e s on the basis of model calculations in combination with ozone values and pyranometer m e a s u r e m e n t s . At present uncertainties in the method are in the order of 10-20%. Figure 3.11 shows a comparison between daily sums of effective UV doses obtained from m e a s u r e m e n t s , as compared to the modelled effective UV, applying the empirical correction for cloud and aerosol reduction to a clear sky model calculation (De Leeuw, 1988). Daily sums were based on at least 10 spectral m e a s u r e m e n t s , and the ozone values obtained from KMI in Uccle (Belgium) were applied in the clear sky model calculations. Applying the method l.W-doses could be e s t i m a t e d in periods and places where UV-measurements are lacking. It should be noted t h a t f u r t h e r long t e r m m e a s u r e m e n t s are needed to establish the findings. F u r t h e r a n a l y s i s is also aimed at revealing similar relationships between spectral m e a s u r e m e n t s and a broad band Robertson Berger biometer. Such methods could contribute to a reduction in the n u m b e r and complexity of m e a s u r e m e n t systems required in a UV-monitoring network. F u r t h e r m o r e the methods can be applied to estimate l.W-irradiances/doses over time periods where detailed UV-monitoring is lacking. High quality and resolution spectral monitoring remains necessary at a number of sites in order to establish the longterm stability of the empirical relationships and serve as reference instruments.
201 M o d e l calculation of effective U V (J/m 2)
o 1993 (RIVM)
..-6" ~.~.(~.""
10 3 -
.~...-
. . . "" ..~" oo.4"" 10 2 .
.
.
.
.. ~
.....-'% .......~o ...,5
101 ::'"
I 10 2
101
t 10 3
M e a s u r e d effective I I V [dmlv .~llm', I/m 2)
Figure 3.11 Daily effective UV-sums: model versus measurements
The method described above is applied to estimate the total yearround effective I.W received in the past year in The Netherlands. Figure 3.12 shows monthly totals of effective UV-doses at ground level in Bilthoven for the year 1993. The open bars on top show the reduction in effective UV by clouds and aerosols. Applying the method for the year 1991 to 1993, we see that 1992 was estimated to have 14% higher doses than 1991 and 1993 11% higher doses (Figure 3.13). The main reason for the increases in 1992 and 1993 are differences in the thickness of the ozone layer, explaining a 9% increase in 1992, and a 15% increase in 1993. Compared to 1991 the reduction by clouds was estimated to be smaller in 1992 and larger in 1993. Thus the real UV-burden was higher in 1992 than in 1993, although based upon ozone measurements higher doses were expected in 1993. A preliminary evaluation of the first months of 1994 showed lower doses than the calculated doses in the same periods in 1993 and 1992. Both ozone levels and cloud reduction are responsible for the lower irradiances. Effective UV (JIm2) r
I
surfacedose
1.4.105 -~l-1 reductionbyclouds1993 1.2.105 -
[--]
1.105 8.104 6.104 4.104 2.104 J
F
M
A
M
J J A Months
Figure 3.12 Monthly totals of effective UV in Bilthoven
S
0
N
D
202
Effective UV (J/m2) Effective UV at groundlevel r-I Reduction by clouds and aerosols 8105
6105
4.105
2105
1991
F i g u r e 3.13 Y e a r l y t o t a l e f f e c t i v e I.W
1992
1993
Ozone column change (%) 85 75 65 55 45 35 25 15 5 -5 -15 -25 -35 -45 -55 -65 -75 -85
9.. -10 -
............"
...........................
" ...............................
. ...... - 5
""
- s ........................................~
o_s
2
~
2
~
_
---------------- 0
0~
9.".............................................5 ................................................................................................................-5 .............................. ........................................... ~o ........................................................................................ :::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
..............................- 15
......:201 J
i F
,............. ...................................... 20 ..........
~ M
i. A
, M
, J
, J
.,i(~-3~ A
S
......i. O
N
D
Month Figure 3.14 O z o n e c o l u m n c h a n g e (%). P r e i n d u s t r i a l / p r e s e n t . ( H a u g l u s t a i n e e t al., 1 9 9 4 )
203
Assessment A spectral monitoring system was designed, built and applied to obtain information on the effective UV-irradiances in The Netherlands. An analysis of spectral m e a s u r e m e n t s has shown an empirical relationship between the reduction of UV by clouds and aerosols and a reduction of global radiation as m e a s u r e d with a p y r a n o m e t e r . This relationship is applied to estimate the UV-burden in the previous years. Compared to 1991 the effective UV dose received at ground level was 14% h i g h e r in 1992 and 11% higher in 1993. The increase is caused by decreased ozone levels, and is also influenced by variations in cloud cover and aerosol loads. A preliminary evaluation of data from the first months in 1994 shows t h a t effective UV doses are lower t h a n in the same period for the previous two years. The goals of the project are met. From a technological and scientific point of view the results are of high quality. International cooperation is envisioned. Interaction with Project No. 85/20/88 should be strengthened. The role of aerosols and tropospheric pollution in the UVB radiation transfer model should be dealt with in more detail. Note by the subtheme c o o r d i n a t o r R. Guicherit Calculated, m e a s u r e d and future projections of increased UVB levels should be regarded in perspective. A m a x i m u m increase in yearly effective UV of 15% at groundlevel around 2,000 is comparable with moving about 400 km n e a r e r to the equator (e.g. Groningen - Maastricht) or moving to an altitude of 500 m compared to sealevel. The consequences of this, in terms of effects and risks are beyond the scope of this part of the national research programme and shall be dealt with elsewhere. Of g r e a t importance in my view are effects of increased UV on atmospheric chemistry notably at middle and upper tropospheric levels, leading to changes in chemical reactivity, lifetimes of pollutant etc. F u r t h e r m o r e local and regional scale tropospheric pollution and clouclcover also affect UV levels and their trends. Although tropospheric 03 accounts only for about one t e n t h of the total ozone column, it is a more effective absorber of UV radiation t h a n stratospheric ozone (on a molecule by molecule basis) under high sun conditions (see e.g. Briihl and Crutzen, 1989). Model calculations (see e.g. Hauglustaine et al., 1994) have shown, t h a t despite stratospheric O3 loss, due to tropospheric O3 increase, the ozone column density m a y have increased since the industrial revolution over a large p a r t of the globe, i.e. from 15~ to 55~ (Figure 3.14). This suggests t h a t changes in tropospheric ozone alone (i.e. at constant stratospheric ozone, aerosols and cloud cover) m u s t have had considerably reduced UV radiation over large areas of the globe. Other factors such as changes in cloud cover, cloud optical properties, and troposphere aerosol load may also have affected trends in UV radiation. Although there are indications t h a t cloudiness over both hemispheres are on the increase since the t u r n of the century and emissions of aerosols and precursors of aerosols have dramatically changed since the turn of the century and m a y either increase or decrease in the n e a r future, there are at the p r e s e n t insufficient data to assess whether these changes will offset predicted UV trends. My impression is t h a t major parts over the N-Hemisphere are still experiencing UV irradiances below p r e i n d u s t r i a l levels due to troposphere pollutants and possibly changes in cloud cover. Although UVB levels may be on the increase in the future, preindustrial levels will not be reached for many part of the globe.
204
3.5 STREAM; Stratosphere Troposphere Experiment, study by Aircraft Measurements B. Bregman and P.J.H Builtjes University of Utrecht, Institute for Marine and Atmospheric Research (IMAU), Princetonplein 5, 3584 CC Utrecht, The Netherlands Introduction A total of three flights were carried out at 16, 17 and 18 February 1993 with a two engine jet, a Cessna Citation II, which operated from Kiruna airport in Northern Sweden. In situ measurements were performed of different trace gases in the free troposphere and lower stratosphere. The measurements were performed in the afternoon, at different levels between 6 and 12.3 km altitude; one flight level in the troposphere and three in the stratosphere. The aircraft maintained for 10 minutes at every level, except at the maximum level were it was maintained for 45 minutes. The flights were carried out North and North-East of Kiruna to 73 N.
Data analysis have been carried out for preliminary HNO3 data, 03 and N 2 0 . 0 3 was measured by a chemiluminescence monitor, with ethylene as reaction gas. Optimalization of the technique, including pressure independence down to 300 mb has been described by Gregory et al. (1983). N20 was measured using a tunable diode laser (TDLAS) spectrometer which is described in detail by Wienhold et al. (1994). HNO3 was measured with a Ne cooled quadruple mass spectrometer. The air was sampled through a 4 cm diameter inlet tube. The technique is described in detail by Arnold and Knopp (1987). 03 concentrations ranged from 60 ppbv in the free troposphere to 700 ppbv in the stratosphere at maximum altitude. HNO3 concentrations were 4-5 ppbv at maximum altitude, decreasing with altitude to less than 100 pptv at the tropopause level. N20 concentrations varied between 240-290 ppbv in the s t r a t o s p h e r e and were occasionally less t h a n 220 ppbv. Tropospheric concentrations were 300-310 ppbv. Low O3/N20 ratios have been found in the lower stratosphere for N20 concentrations down to 200 ppbv. The air parcels with the lowest ratios were measured during the flight on 17-02-93, accompanied by relatively low HNO3/N20 ratios. The 10 days back trajectories of these air parcels follow the inner vortex structure, suggesting that the air has descended to the lower stratosphere. The "shift" in the O3/N20 ratio on the 17th extends to theta=315 K, implying that descending vortex air has reached the tropopause. Further, the low O3/N20 ratios on the 17th suggest that 03 loss has occurred. Based on comparison of the O3/N20 ratio on the 17th with that on the other days, 36% 03 loss is calculated. Although there are no previous simultaneously measured in situ data of HNO3, N20 and 03 available, the HNO3 data is compared with estimated HNO3 from measured NOy from the AASE I and II missions, assuming that NOy contains 80% HNO3. Our results show significantly higher HNO3 concentrations relatively to 03 and to N20 compared to the results of the AASE I and AASE II campaigns. The ratio NOy/O3 measured during AASE I is constant with altitude in the stratosphere, while our HNO3/O3
205 ratio shows much more variability in the stratosphere with increasing ratios with altitude. The measurements during AASE I were performed before the eruption of Mt. Pinatubo, so that the higher ratio values measured during STREAM could be caused by enhanced sulphuric acid aerosol abundance in the lower stratosphere, leading to enhanced HNO3 concentrations by heterogeneous reactions on the aerosols. However, the STREAM data also show significant higher HNO3/O3 ratios relatively to N20 compared to the results of AASE II, which was performed after the eruption of Mt. Pinatubo. These differences could partly be explained by the suggested 03 loss, since the ratio HNO3/O3 relatively to N20 is lower for the 17th compared to the other days. However, different meteorological conditions during the period of the measurements, e.g. less influences by vortex air of mixing of air from lower latitudes, affecting the ratios of the trace gases, should be taken into account. Our higher HNO3/O3 and HNO3/N20 ratios may also indicate enhanced heterogeneous chemistry on sulphate aerosols compared to 1993, due to enhanced sulphuric aerosol abundance in the lower stratosphere by gravitational settling of the aerosols. An alternative explanation of the enhanced HNO3 concentrations may be sedimentation of earlier formed PSC's, which evaporated in the lower stratosphere, releasing HNO3. The vortex was relatively cold and stable during the winter, allowing particles to sedimentate into the lower stratosphere. Also low HNO3 concentrations relatively to N20 were found on 17-02-93, which were lower than previous measurements during AASE I and AASE II in the Arctic vortex (Loewenstein et al., 1993; Kawa et al., 1990). These low HNO3 concentrations, corresponding to low N20 concentrations suggest t h a t denitrification has occurred during descent of air in the vortex. This suggestion is supported by the fact that the 10-days back trajectories of all concerned air parcels follow the inner vortex structure, preventing mixing with HNO3 rich air. Model calculations
10 Days back trajectories, ending at the flight track have been used as input for a lagrangian chemical box model to simulate the STREAM data. The model is based on a chemical scheme originally developed by Mfiller et al. (1993). Numerical calculations were carried out using the FACSIMILE code (Curthis and Sweetenham, 1987). The chemical scheme is coupled to a radiative transfer model, developed for stratospheric conditions, including multiple scattering at low zenith angles. Detailed description of the model can be found in Lary and Pyle (1991). Heterogeneous reactions on sulphuric aerosols have been included (Tolbert et al., 1988). Initial concentrations were derived by a 2-D photochemical stratospheric model (Bruhl and Crutzen, 1988), originally developed by Gidel et al. (1983) and Crutzen and Gidel (1983). Also HALOE data have been used as initial data for calculations at mid northern latitudes and EASOE data (Crewell, pers. comm., 1994; Von Clarmann, 1993). Aerosol concentrations were taken from Deshler et al. (1993). P r e l i m i n a r y r e s u l t s indicate discrepancies between model r e s u l t s and measurements. The model shows higher 03 relatively to N20 and lower HNO3 relatively to 03. Recent measurements during the AASE I and II missions show
206 t h a t the ratio O3/N20, NOy/N20 and NOy/O3 are constant in the lower stratosphere with small latitudinal variation. Therefore these discrepancies may not be an input problem and the chemical description of the model will further be investigated. Assessment The objective of the project is to measure trace gas concentrations in the Arctic lower stratospheric vortex. However, because only 3 components were measured, the information obtained is scarce, making the interpretation of the results difficult. In future experiments, therefore more attention should be paid to measuring strategy with regard to the minimal number of parameters to be measured. Finally one should reflect upon how the results can optimally be used in other modelling studies. 4.
T R O P O S P H E R I C B U D G E T OF NON-CO2 G R E E N H O U S E GASE
4.1 I n t r o d u c t i o n The n o n C 0 2 greenhouse gases, excluding w at er vapour, currently have concentration levels in the troposphere well above their natural background. The concentration of CH4 is about a factor 2.5 above its n a t u r a l background, tropospheric ozone also about a factor 2.5. The N20-concentrations is about 10% above its n a t u r a l value, and the CFC-concentrations are nearly completely anthropogenic in origin. The aim of the study of the tropospheric budget of non CO2 greenhouse gases is to understand and determine in a quantitative way the processes which govern these budgets. An overall understanding leads to a model with which in a reliable manner the relation is given between anthropogenic and biogenic emissions and the concentrations of these non CO2 greenhouse gases. Ambient measurements are required to investigate the relevant processes, for evaluation of the results of the modelsimulations and for trend measurements. Models, if they have proven to be reliable, can be used to investigate the causes of historical trends, and to calculate future trends as a result from emission scenarios and abatement policies. These type of models can be used both scientifically and for policy making. Budget studies combine different elements. In the first place, and often of primary importance, is accurate knowledge concerning anthropogenic and biogenic emissions. The overall tropospheric models contain meteorological and chemical submodels, as well as dry and wet deposition and radiation models. The overall models are directed to budget studies of CH4, N20, 03 and CFC, but consequently also address CO, NOx, VOC, HCFC and include clouds and aerosols. Both global and regional modelling are essential, regional modelling is important for the greenhouse gases with a relative short residence time like 03, and also aerosols. And last but not least, an essential element of budget studies is the evaluation of modelresults versus reliable and representative ambient measurements.
207 Global and regional tropospheric models are directed to the simulation of the troposphere as a whole, and consequently integrate all available knowledge in a consistent and coherent way. Tropospheric models have considerable prognostic capabilities, because they are based on the solution of the in principle 'complete' diffusion equation. In this cluster of budget studies, the main focus was on the development and i m p r o v e m e n t of tropospheric models, and the creation of a reliable and representative ambient measurement data base. Work has been directed to the improvement and development of both 2-D and 3-D global models, and 3-D regional models the driving concept was the setup of a hierarchy of models with different balances between chemistry and meteorology. The outcome of this cluster of budget studies is a hierarchy of models capable of addressing a range of questions concerning the relationship between emissions and the concentrations of non CO2 greenhouse gases. A key element, also for the future, remains the evaluation of the models against ambient data, and the reliability of the models in their evaluation of abatement strategies and future trends. This subtheme comprises 4 projects.
4.2 GLOMAC; Subproject of EUROTRAC related to climate m o d e l l i n g of KNMI M. Allaart KNMI, Royal Netherlands Meterological Institute, P.O. Box 201, 3730 AE De Bilt, The Netherlands
Stratosphere / troposphere exchange The goal of this research is to quantify the amount of ozone reaching the troposphere from the ozone layer. This research is motivated by the fact that ozone is a very reactive gas that plays an important part in the chemistry of the troposphere. Most of the ozone in the troposphere is (today) created in the troposphere itself, as a result of the oxidation of hydrocarbons. Human activities are important sources of these hydrocarbons. A part of the ozone in the troposphere is of stratospheric origin. It is brought there from the ozone layer by a process t h a t is called stratosphere/troposphere exchange. It is currently thought that the stratosphere/troposphere exchange is driven by the global scale circulation cell in the stratosphere. In the equatorial zone, or more precise over Indonesia, tropospheric air can pass through the tropopause and so enter the stratosphere. A large circulation cell through the stratosphere, and possible the mesosphere brings the air towards the poles. In the winter season this air can pass the tropopause again in the polar and mid-latitude zones.
208 The p r o c e s s c u r r e n t l y d o m i n a t i n g the d i s c u s s i o n over m i d - l a t i t u d e s t r a t o s p h e r e / t r o p o s p h e r e exchange is the so called tropopause folding. It is essential to realize that the height of the tropopause over the mid-latitudes is not a fixed property, but has strong fluctuations with a typical timescale of one day. The actual heights vary between 5 and 12 kilometres. These variations are closely linked to the tropospheric weather; a well developed ridge of high pressure will cause the tropopause to be relatively high, a developing low can cause the tropopause to sink to very low values. In the last case, a new tropopause can form above the old one, isolating a blob of stratospheric (ozone rich) air in the troposphere. This air will eventually dissolve into the troposphere. The very changeable height of the tropopause is also responsible for day to day fluctuation in the total ozone values over the mid-latitudes. With 'total ozone' the column integrated amount of ozone above a certain point of the globe is meant. This quantity is in the order of 5 to 12 grams per square meter. The best way to estimate the stratosphere/troposphere exchange is probably to quantify the mass-flux in the stratospheric circulation cell mentioned before. Direct m e a s u r e m e n t s are not possible. An indirect method has been used in the past by Holton. One can also look at the details of the process of tropopause foldings. One m u s t realize, however, that the tropopause foldings are not the driving force behind the circulation cell in the stratosphere, nor are the tropopause foldings directly caused by this cell. One can study the details of stratosphere/troposphere exchange with observations. Here we m u s t make use of the fact t h a t the stratospheric and tropospheric air have slightly different properties, and so one can distinguish between the two. The most important are the mixing ratios of ozone and water vapour. The ozone mixing-ratio is 10 to 100 times higher in the stratosphere t h a n in the troposphere. Near the tropopause it is a reasonably well conserved property, so it can be used as a tracer for stratospheric air. Unfortunately it is not possible to get a good three dimensional field of the ozone distribution near the tropopause from a satellite. The only m e a s u r e m e n t s with sufficient resolution are in situ measurements, for example from balloon or aircraft based instruments, or lidar measurements. In either case m e a s u r e m e n t s are far too infrequent to obtain a total hemispheric mass flux without extensive extrapolations. We do have the possibility to study stratosphere/troposphere exchange in the global w e a t h e r prediction models (GCM). These models, however, do not (yet) contain ozone as a trace gas. So we will have to parametrize the ozone mixing ratio in some way. In the current study we have looked at a quantity called the potential vorticity (PV). PV is relatively easy to compute, and behaves a lot like ozone. It is high in the stratosphere and low in the troposphere, and a conserved quantity. It has been
209 observed t h a t PV can be used to identify intrusions of stratospheric air in the troposphere. It would seem logical that, if there is a linear relation between ozone mixing ratio and PV, then there must also be a relation between column integrated ozone (total ozone) and column integrated PV. To investigate this we did a multiple variable regression analysis between TOVS total ozone maps and the PV on 15 pressure levels computed from ECMWF forecast fields. It was found that the results were indeed consistent with a linear relation between ozone mixing ration and PV, between levels of 400 and 50 hPa. The regression coefficient of this linear relation proved to have a seasonal cycle, and a large inter annual variability (Allaart et al., 1993).
Figure 3.15 Synoptic m a p of NOAA-TOVS total ozone data valid for May 4, 1992, 12 UT. Notice the extended structure of high ozone values from the N o r t h - E a s t to South-West over central Europe
82
Figure 3.16 Column integrated potential vorticity scaled linearly to show the r e m a r k a b l e resemblance to the Ozone values in Figure 4.1. Data computed from ECMWF global vorticity and temperature fields
210 This result has been very fruitful; it enabled us to distinguish between dynamical and other (possible chemical) effects on a regional scale on the ozone layer. For example we could prove that the very low ozone values recorded in J a n u a r y 1992 over Europe were caused for 50% by dynamical effects on a regional scale. Also this method enables us to forecast total ozone amounts for a few days ahead, based on the weather forecasts from the ECMWF. This method is now the basis for the daily predictions of damaging UV radiation made at the KNMI. In order to compute the mass flux through the tropopause, first a practical definition of the tropopause is needed. The official WMO definition, which defines the tropopause in terms of the temperature gradient is only one of the possible definitions. It is generally assumed that for atmospheric chemistry, the dynamical tropopause, defined in terms of constant PV, is a more suitable definition. We have investigated this and found that there is a very poor agreement between the two definitions of the tropopause. We also started an observational programme. At least once a week we launch balloon borne ozone sondes which can, with a very high vertical resolution, detect ozone rich, and dry air in the troposphere. We assume these are the signatures of tropopause foldings. Again we found t h a t there is only a very poor a g r e e m e n t between the observed structures and the folding events found in the ECMWF forecasts.
Meteorological aspects of global modelling of atmospheric chemistry Introduction. The work of Allaart, described in the previous section was supported by NPR. In the following sections we will describe the general progress of GLOMAC related work at KNMI.
Boundary-layer parametrization (G.H.L. Verver).In c u r r e n t t r a n s p o r t and chemistry models the parametrization of turbulent transport of chemical reactive gases is the same as for inert gases. Most models use a down gradient description similar to molecular diffusion. The turbulent exchange coefficient then depends on stability. However, for reactive gases, this will introduce errors for two reasons. The first is that it is assumed that each horizontal layer is well mixed. This will not be the case for all species. Correlated or anti correlated concentration fluctuations on a subgrid scale will enhance or diminish the effective reaction rates. The second reason is t h a t turbulent fluxes of one species will depend on the flux of the others. This is not incorporated in the usual K-diffusion description. Both effects depend on the ratio of the chemistry timescale and the turbulent timescale. It was found t h a t a second order model is able to describe these phenomena, and t h a t in the case of n o n - n e u t r a l b o u n d a r y - l a y e r s the c o n c e n t r a t i o n and t e m p e r a t u r e covariance m u s t be taken into account (Verver, 1994). A second order model t h a t explicitly describes fluxes and (co-)variances is now being developed. The aim is to deduce simpler p a r a m e t r i z a t i o n s for use in global transport models, that take into account the effects mentioned above.
211
Deep Convection (P. Siebesma). Pieter Siebesma studied the different convection parametrizations and constructed a new one.
Vertical diffusion by dynamic instabilities (H. Kelder). Vertical diffusion by instabilities was studied with a theoretical model. We found a new type of propagating instabilities (Lott et al., 1992). Observations of small scale transport of ozone was studied in the framework if the AESOE campaign (Teitelbaum et al., 1994). Unresolved non-turbulent transport (M.A.F. Allaart, H. Kelder, L.C. Heijboer, G.J.M. Velders, P.J.F. Velthoven). At KNMI a global atmospheric circulation and chemistry model ("TMK") was implemented. This is an off line t r a n s p o r t model coupled to the ECMWF (European Centre for Medium-Range Weather Forecasts) analyzed windfields. TMK is one of the models used within GLOMAC. As KNMI is responsible for the meteorological part, much effort has been put into upgrading the meteorological input of the model. With this model we have done studies of the transport of passive traces (Velders et al., 1994). A n u m b e r of chemical reactions has also been added in collaboration with other members of GLOMAC: P. Crutzen and J. Lelieveld. We have used the upgraded model to study the transport of aircraft emissions (Van Velthoven et al., 1994). The model was also used to study the exchange between troposphere and the stratosphere. For this purpose we increased the number of vertical levels near the tropopause. We studied the influence of the horizontal resolution on the cross-tropopause mass-fluxes in the model (Kelder et. al., 1994). In order to compute the cross-tropopause ozone fluxes a new ozone climatology is being constructed (Fortuin et al., 1994). Assessment Potential Vorticity (PV) can be used to identify intrusions of stratospheric air into the troposphere. This has been published before and was reconfirmed by this project. By using this method one may hope to discriminate between transport and chemistry as the cause of changes in ozone column density in the future. To be able to do so is a long way to go and it is not sure if it can be done at all. The correlation between PV and O3 column after all is an empirically one of which the validity is limited. The author does not discuss how the stratosphere/troposphere flux is determined and how it should be dealt with in models, which is one of the objectives of the project. Finally this method can be used to forecast total O3 for a few days ahead and its consequences for the V B radiation climatology. From the abstract it seems t h a t also progress has been made for those parts of the project not funded by NPR i.e. the meteorological aspects of global modelling. The project as a whole is relevant in perspective of global change and scientifically of good quality.
212
4.3 Application of 2-D global models M. Roemer TNO, Institute of Environmental Sciences (IMW) P.O. Box 6011, 2600 JA Delft, The Netherlands This a b s t r a c t s u m m a r i z e s the main findings obtained in this project so far. A global two dimensional model, the TNO version of the Isaksen model, was used to study the following issues: 1 the relation between emissions and concentrations of CO and CH4 on a global scale. ; 2 the tropospheric oxidising capacity of the troposphere (03, OH); 3 analysis of differences between results of the 2-D model and other models; 4 scenario calculations using the EDGAR]GEIA data base. Due to the fact t h a t the EDGAR/GEIA data base is not yet accessible the last item has not been addressed until now.
Introduction The last 10-15 years global scale models have been developed to describe the emissions, t r a n s p o r t , d e g r a d a t i o n and removal processes of gases in the troposphere. These models vary in dimensionality (0-D ,ZE 3-D), in the way the chemistry is represented, the climatological fields, photodissociation rates, basic emissions, etc. The p e r f o r m a n c e of models is u s u a l l y v a l i d a t e d a g a i n s t observations. However, on a global scale there are not enough m e a s u r e m e n t s to have sufficient knowledge of the distributions of most of the gases, except for m e t h a n e and to some extent ozone and carbon monoxide. Consequently, the validation of tropospheric global scale models is poor and primarily based upon the comparison for the three above mentioned species. A striking example of the fact t h a t agreement between observations and model results on just a few species and measurements is insufficient as validation test is given in Figure 4.3 (Guthrie and Yarwood, 1991). The five different models predict a wide r a n g e of m e t h a n e levels in the year 2100 based on the same emission scenarios.
Model intercomparisons In 1992 and 1993 two intercomparison workshop were organised to identify the causes in the diversity of model responses. TNO's 2-D model was one of the models which were c o m p a r e d in t h e s e workshops. An overview of some of the characteristics of these models is given in Figure 4.4 and Tables 4.1 and 4.2. Table 4.1 shows t h a t the emissions are within about 40 percent of each other which is consistent with the generally accepted u n c e r t a i n t y ranges of the global scale emissions. Figure 4.4a and 23b show a wide range of photodissociation rates used by the different models under the same conditions. A closer inspection indicated t h a t some of the discrepancies in the photodissociation rates are due to the way clouds are treated. Given these differences (and m a n y others) it is not surprising t h a t the global averages of some key species differ in the order of a factor of two (CO and 03) or more (NOx). Only for methane there is hardly any difference.
213 CH4 concentration (ppm) 6.0
IPCC Scenario A
x OSLO 5.0
O GSFC
4.0
"-I- HAR
* GISS
[] MPI
3.0
Mean
J= 2.0
1.0. 1980
2000
2020
2040
2060
2080
2100
Year
Figure 4.3 Calculated CH4 concentrations 1985-2100. Source Guthrie&Yarwood, 1991
Table 4.1 Base case emissions used in the models Model
CH4 TgCH4/yr
CO (Tg C/yr)
NOx (surface) (Tg N/yr)
Cambridge Harwell TNO MPI-CNRS-3D MPI-1D FSFC LLNL
405 485 55O fixed m.r. 540.5 400 453
828 675 616 729 537 785 fixed m.r.
36 34 32 30 34 29 39.4
214 Table 4.2 Average tropospheric mixing ratios calculated by the models for the base case Model
Cambridge Harwell TNO MPI-CNRS-3D MPI-1D FSFC LLNL a
CH4 (ppm)
CO (ppb)
OH (E5 cm-3)
03 (ppb)
NOx (ppt)
1.71 1.63 1.66 1.97 1.75 1.75 1.59
131 88 75 97 84 90 67
7.1 8.5 10.5 8.7 10.8 7.5a 8.3
51 77 49 36 44 54 67
143 40 66 55 114 56 53
approximate value derived from mixing ratio
rate (s-l) 1.2E-05
TNO x CAMB
9 HAR O MPI-1D A GSFC [] LLNL
1E-05 l
8E-06
6E-06
4E-06
2E-06 0
4
8 12 Altitude (kin)
16
20
Figure 4.4a Photodissociation rate (s-l) as a function of altitude (km); 50~ O(1D))
21 June. J(O3 n
215 rate (s- 1) ~: TNO 0.007
x CAMB
I .~
9 HAR O MPI-1D 9 GSFC [] LLNL
0.005
0.003
0.001 0
I 4
I 8
I 12
I 16
20
Altitude (km)
Figure 4.4b Photodissociation rate (s-l) as a function of altitude (km); 50~ NO+O(3P))
21 June. J(NO2 -~
Also the chemistry behaves quite differently. Under fixed conditions four chemical schemes were tested (Fig. 3.19 and 3.20). This illustrates furthermore the variety in model responses. Unfortunately for most of the parameters no yardsticks exist to identify which parameter values or chemical schemes are acceptable and which ones are flawed. In the future more intercomparison activities for tropospheric models are foreseen and needed to narrow uncertainties in model input and output. (03) (ppb)
(NOz) = 1 ppb
52, 9 TNO [3 LLNL
50
9 Harwell 48
O GSFC
46 44
42 40 ~ 700
Figure 4.5 Daily trend of the
t 1300
I 1900
t 100
700
Hour 03
concentration; NOz = I ppb (NOz = NO
+ NO2 +
N03 + 2N205)
216
08 a n d OH Since p r e i n d u s t r i a l times the tropospheric ozone content has i n c r e a s e d substantially. Comparing the Montsouris series (1876-1910) with current ozone records, a doubling to tripling of ground level ozone concentrations at northern midlatitudes is found. The indication is that the role of tropospheric chemistry in the formation of ozone has changed from a small net sink in the preindustrial era to a source comparable in size with the influx of ozone from the stratosphere (Table 4.3). The net chemical ozone production (or destruction) is the small difference between two large terms: ozone production and ozone destruction. The third term in the budget is dry deposition. Also calculations are presented in this table for scenarios in which NOx and NMHC emissions are doubled. It should be noted t h a t the assumed influx from the stratosphere is now regarded as being somewhat too high; demonstrating once more the importance of a net free troposphere source.
(0 3) (ppb)
(NOz) = 100 ppt
52 9 TNO
[] LLNL
50
9 Harwell
48
0 GSFC
46 44 42 40
700
1300
1900
1O0
700
Hour Figure 4.6 Daily trend of the 03 concentration; NOz~= 100 ppb.
Table 4.3 Global average ozone budget (1010 molec, cm-2.s-1) Runs
Preindustrial 1990 2.0 * NOx 2.0 * NMHC
Chemical production
Chemical destruction
Net. Ch. Prod./Destr.
Deposition
20.0 44.8 55.0 49.9
21.5 39.2 47.5 42.8
- 1.5 5.6 7.5 7.1
5.9 13.0 15.9 14.5
Flux from stratosphere 7.4 7.4 7.4 7.4
217
A m o r e d e t a i l e d a n a l y s i s of t h e s e p r o c e s s e s r e v e a l l a r g e r e g i o n a l d i f f e r e n c e s ( F i g u r e 4.7) T h e tropical m i d d l e t r o p o s p h e r e is a l a r g e s i n k region to ozone d u e to: 1 t h e p h o t o d i s s o c i a t i o n of ozone into O(1D) followed by t h e r e a c t i o n w i t h H 2 0 to p r o d u c e OH, a n d 2 t h e r e a c t i o n of ozone w i t h HO2. T h e l o w e s t levels of t h e n o r t h e r n m i d - l a t i t u d e s is a c h e m i c a l s o u r c e r e g i o n b e c a u s e of t h e o x i d a t i o n of V O C s in a NOx-rich e n v i r o n m e n t .
SH June* 16.250 15.250 14.250 13.250 12.250 11.250 10.250 9.250 8.250 7.250 6.250 5.250 4.250 3.250 2.750 2.250 1.750 1.250 0.750 0.250 * altitude
NH
-80.0-70.0-60.0-50.0-40.0-30.0-20.0-10.0
0.0
10.0
20.0
30.0
40.0
+++ +++ ++++ ++++ ++++ +++ +++ ++ ++ ++ +++ ++++ +++++ 0 0 0 + + + + + + + + + + 0 0 0 + + + + + + + + + + 0 0 0 + + + + ++ ++ ++ ++ + + 0 0 0 + + + ++ ++ ++ ++ ++ ++ ++ 0 0 0 + + ++ ++ ++ ++ ++ ++ ++ ++ 0 0 0 + + ++ ++ ++ ++ ++ ++ ++ ++ 0 0 0 + + ++ ++ ++ ++ ++ + + 0 0 0 + + + + + ++ . . . . 0 0 0 + + + . . . . . . . . . . + . . . . . . . . . . . . . . . . . 0 0 0 0 0 0 0 0 + . . . . . . . . . . . . . . . . . . . 0 0 0 0 + . . . . . . . . . . . . . . . . . . . . 0 0 0 0 + . . . . . . . . . . . . . . . . . . . . . . . . 0 0 0 0 + . . . . . . . . . . . . . . . . . . . . . . . . 0 0 0 0 + . . . . . . . . . . . . . . . . . . . . . . ++ 0 0 0 0 + . . . . . . . . . . . . . . . . . . . . . +++ 0 0 0 0 ++ . . . . . . . . . . . . . . . ++++ 0 0 0 ++ ++ +++ +++ +++ +++ ++ ++++ +++++ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
50.0 +++++ + + + ++ ++ ++ ++ + 0 0 .
. ++ ++ +++ ++++ ++++ +++++ . . . . .
70.0
80.0
++++ ++++ 0 + + + + + + ++ ++ ++ ++ ++ ++ ++ ++ ++ + + + +
60.0
++++ + + + ++ ++ ++ ++ ++ ++ +
+ + ++ + ++ + ++ + +++ + ++++ + ++++ + ++++ +++ . . . . . . . . . .
(km)
T h e classification is as follows (NP: n e t production; ND: n e t d e s t r u c t i o n ) . +++++ ++++ +++ ++ + 0 ---....
: : : : : : : : : 9 :
5.1012 1.1012 5.1011 1.1011 1.1010 -1.101o -5.1010 -1.1011 -5.1011 -1.1012
< 5.1012 < 1.1012 < 5.1011 < 1.1011 <+1.1010 < -1.1010 < -1.1011 < -1.1011 < -5.1011 < -1.1012
F i g u r e 4.7 P r o d u c t i o n a n d d e s t r u c t i o n r e g i m e s for ozone in t h e t r o p o s p h e r e . Unit: molec.cm3/month
+ ++ ++ ++ ++ ++ ++ ++ .
218
Assessment This study clearly demonstrates the uncertainties in scenario modelcalculations, even when state of the art models are being used. E.g. CH4 concentrations in 2100 for the IPCC scenario "A" may range between 3 and >5 ppm. Consequently the atmospheric residence time may vary from 8.5 to over 13 years. The international community of modellers should agree which 'standard' model to be adopted for policy issues of a supranational nature, more or less the way it was decided for the ECE-EMEP-project on t r a n s b o u n d a r y pollution over Europe. For tropospheric ozone, it is now accepted that a major part is produced in the troposphere itself. The net production is of the same order of magnitude as the influx from the stratosphere3. Depending on the future emission scenario used, tropospheric ozone production will further increase with serious consequences for m a n and the ecosystem and furthermore for the radiation budget of the atmosphere. This project is highly relevant for policy making; the model seems to perform well, especially for components with atmospheric residence times of a few weeks or more, when compared to other state of the art models. 0 3 influx from the s
t r a t o s p ~
-4770 MT ~-5450 MT~ [ - ~ .c: f3.
+(70-100 MT)
o o 9-
v | 0 3 loss due to - 1580 MT -
Figure 4.8
,~deposition T
/ ~
~/
| I,
0 3 influx from the | PBLinto the / free troposphere I land
oceans
4.4 G l o b a l m o d e l l i n g of a t m o s p h e r e t r a c e g a s e s "application of a g l o b a l t h r e e d i m e n s i o n a l model" J.P. Beck, T.H.P. The, D.L. Veenstra N a t i o n a l I n s t i t u t e of Public Health and E n v i r o n m e n t a l Protection (RIVM), P.O. Box 1, 3720 BA Bilthoven, The Netherlands Introduction The research in this project is performed in collaboration with TNO-MW. The broad objective of this project is the production of the 'base-document CH4' and a
3. From discussion with PIs cooperating in this priority issue the 0 3 budget, at present, most probably is as being depicted in Figure 4.8.
219 document on aircraft and air pollution, both also by order of the Ministry of the Environment. To be able to make the necessary scenario runs and assessments we had to put into operation a global three-dimensional model (the MOGUNTIA model) at RIVM. With this MOGUNTIA model the relations between emissions and concentrations of trace gases relevant to ozone and its relates species on a global scale are studied. In this extended abstract the focus will be on the study on effects of aircraft emissions. The CH4 study will be reported in our final NPR report. Considerations Aircraft emissions affect the chemical composition of the atmosphere in several different ways, depending on flying altitude and location. The conventional subsonic fleet grew extensively in the 1970s and 1980s and is expected to growth by a further 100% during the next two decades. This has led to several model studies focusing on effects of aircraft emissions on ozone in the upper troposphere (Hidalgo and Crutzen, 1977; Isaksen, 1980; Derwent, 1982; Ehhalt et al., 1992; Beck et al., 1992). These studies have predicted an increase in upper tropospheric ozone levels due to emissions of NOx varying between 5 and 12%. The aircraft-induced changing ozone profile could have implications for the atmospheric radiative balance. The previous modelled estimates of the ozone increases probably suffer from considerable errors due to uncertainties in the magnitude and distribution of e m i s s i o n s and differences b e t w e e n models. F u r t h e r m o r e , a n u m b e r of simplifications were made in the r e p r e s e n t a t i o n of aircraft emissions in atmospheric chemistry models. Firstly, in these studies it was assumed that the present subsonic fleet emits all e x h a u s t gases below the tropopause. However, new estimates predict t h a t a considerable part (25 to 50%) of the emissions from the subsonic fleet are deposited directly into the lower stratosphere. Residence times in the lower stratosphere are expected to be longer than those in the upper troposphere due to less efficient vertical overturning. So, the mixing and chemistry in the lower s t r a t o s p h e r e are different from t h a t in the upper troposphere. Obviously, the effects of emissions in the lower stratosphere will also be different from those in the upper troposphere. Secondly, in the previous work it was assumed that immediate large scale mixing of all emissions occurs. For this reason, the impact of flight corridor effects on the predictions has not been assessed. Again, in the previous work, the chemical composition of the source gases during largescale mixing is considered identical to that of the emission at the tail pipe. Initially, the exhaust plume from an aircraft is isolated and highly turbulent. The exhaust entrains ambient air and expands to dimensions of several h u n d r e d metres. Finally, it is subject to dispersion and mesoscale t r a n s p o r t before it becomes zonally mixed. This is illustrated by the estimate t h a t in the upper troposphere over the North Atlantic, emissions of a single aircraft are mixed in about a day in a volume of 5000 km along the flight path, 1000 km wide and 2 km in altitude. However, about 500 aircrafts per day use this corridor and burn 100 tonnes of fuel each. This implies that the emissions may be highly concentrated in heavily travelled flight corridors, causing zonal differences and local impacts of aircraft emissions s e p a r a t e from global perturbations due to long-range transport of emissions. Thirdly, another process omitted is the possible involvement of nonlinear reactions involving NOy species and heterogeneous chemistry occurring on ice particles in
220 the contrail. Soot and sulphur particles in the exhaust favour formation of ice crystals. The response to emitted NOx is expected to change if heterogeneous chemistry occurs. The N O x - 03 chemistry is very nonlinear and results in a greater ozone production per unit of NOx for lower NOx concentrations (Liu et al., 1987). In most models immediate largescale mixing of aircraft emissions is assumed. However, a part of the NO x had probably already been converted to NOy before the plume reached grid dimensions. Therefore the ozone production caused by the NOx emissions is overestimated. For this reason, the role of chemical plume processes in aircraft emissions need to be assessed. The aim of our present work is to determine the effects of emissions of NOx and other exhaust components from aircraft with the MOGUNTIA model, in particular with respect to changes in the upper troposphere and lower stratosphere. Focusing on the considerations mentioned above, we take into account the chemical processes in an aircraft plume in the upper troposphere before largescale mixing occurs. Therefore an exhaust plume model is being developed and the plume model study is expected to result in a parametrization of the sub-grid, in particular NO x NOy. chemistry of the plume. Using this parametrization the global NOx emission fields of aircraft can be translated into new processed fields, in which the sub-grid effects have been taken into account. These new emission fields will be used as input for a MOGUNTIA study on the effects of aircraft emissions at cruising altitudes on the global atmosphere.
The aircraft exhaust plume model In the aircraft exhaust plume model the kinetics and diffusion of an exhaust plume are simulated from almost immediately after the 'tailpipe' until the concentrations in the plume are equal to the background concentrations, or until the plume has grown to 3D-model grid dimensions. The model consists of two sets of differential equations representing the mass balance equations in and outside the exhaust plume. The FACSIMILE package (Curtis and Sweetenham, 1985), which applies a version of Gears' method, has been used to calculate the arithmetic solutions of the chemical equations. The chemistry of the plume is calculated using a set of reactions given by Beck et al. (1992). This set consists of 42 species, 73 gas-phase reactions and 16 photolytic reactions. Six primary NMHCs represent the emitted species and their degradation pathways. The concentrations in the ambient air are also calculated using this set of reactions. The photolysis constants are calculated from the intensity of the sun, the absorption cross-section and the quantum yields of the compounds. Background concentrations are based on the measurements of the STRATOZ III campaign (Drummond et al., 1988). Currently, the set of reactions does not contain any heterogeneous reactions, but it will be extended with heterogeneous chemistry in a short while. Also, for extension to the lower stratosphere the tropospheric chemistry will be replaced with a stratospheric reaction set. The diffusion part of the model was based on the theory of Gelinas and Walton (1974), which states that the kinetics and chemistry of an exhaust plume can be described separately. In this approach, the plume is described by a box of variable volume, which always encompasses the exhaust species. So, the explicit spatial dependence is suppressed, and the resultant time dependent equations can be readily solved using standard kinetic methods. The dimensions of an exhaust plume are about 3 km high and about 100 km wide after 24 hours expansion. The
221 simulation of only one exhaust plume will be extended to the simulation of several plumes later. The exhaust plume model was shown to be a good instrument to study plume chemistry. Generally speaking, maximum perturbations (caused by the emission of an aircraft) of the plume concentrations exist only in the first few hours after injection and disappear after about a day. We found that these perturbations strongly depend on the amount of emission, hour of the day and the season when injection takes place, as well as the background concentrations. The 03 concentration in the plume rapidly decreases after the injection of emissions, in particular caused by the reaction of emitted NO with 03. The 03 concentration is reestablished to a level similar to the background concentration in about an hour due to mixing with background air. Subsequently, some production of ozone occurs and the resulting concentration in the plume is somewhat higher than the background level for a few hours. The high NOx concentration occurring immediately after injection decreases rapidly due to mixing with background air. Subsequently, some production of ozone occurs and the resulting concentration in the plume is somewhat higher than the background level for a few hours. The high NOx concentration occurring immediately after injection decreases rapidly due to mixing with background air. However, a part of the NOx is also converted to NOy components. At first, HONO is formed; this then undergoes rapid photolytical decomposition. Subsequently, HNO3 is formed and after some hours also HO2NO2, from which formation was initially inhibited through a lack of HO2 (caused by the reaction of NO with HO2). The calculated NO and NO2 concentrations in the plume are quite similar to measured amounts in a young plume (Arnold et al., 1992). But the measured perturbations of HONO and HNO3 are 10 to 20 times greater than calculated in the exhaust plume model. A reason for this discrepancy may be our limited set of reactions. The formation of HONO and HNO3 in our exhaust plume model is only by reaction of respectively NO and NO2 with OH. However, the concentration of OH in the plume is too low for a very rapid transformation. For this reason, it seems that there must be another reaction pathway for the formation of HONO and HNO3. Heterogeneous reactions are a possibility for this pathway.
Translation of aircraft emission fields The parametrization of sub-grid chemistry was initially in the form of a simple conversion factor: an emission of x molecules of NOx is converted after expansion to grid-dimension into y molecules of NOx and z molecules ofNOy (HNO3, HONO, HO2NO2, N205, PAN), with x = y+z. The transformation of NOx into NOy components depends strongly on the altitude of emission and background concentration (e.g. NOy, OH and HO2) components, and less on the time of injection (time of day, season) and amount of emission. The higher the altitude of aircraft emissions the slower the conversion of NOx into NOy. Here the lower temperatures are the primary reason, followed by the lower OH and HO2 concentrations.
222 As a first parametrization, it is assumed that after one day the exhaust plume has reached grid dimensions and the plume model predicts: f(NOy) f(NOx)
= 0.7 = 0.3
So, in the new emission fields the NOx emissions will be differentiated into NOx (30%) and NOy (70%) components, the latter group divided in specific components contributing to NOy. MOG UNTIA- mode ! runs
The processed emission fields are used as input for the MOGUNTIA model. MOGUNTIA is a global tropospheric 3D-model incorporating t r a n s p o r t and chemistry and it was originally developed at the Max Planck Institute in Mainz (Zimmermann, 1988). It has grid dimensions of 10 ~ x 10 ~ x 100 hPa, starting at the surface and extending to the 100 hPa level. The dynamics of the model are calculated with ECMWF fields, assuming turbulent mixing. The reaction set used for the chemistry is from Dentener (1993). The calculations are performed with aircraft emission fields and corresponding anthropogenic NOx emission fields based upon data of McInnes and Walker (1992) and Mfiller (1992), processed by Olivier (1994). Preliminary results of unprocessed 1990 emission fields show an increase in the ozone concentration at the location of the North-Atlantic flight corridor at the cruising altitude of 8%. The processed emission fields show an increase of about 6% at the similar location. Assessment
The exhaust plume model was shown to be a good i n s t r u m e n t to study plume chemistry. A comparison of the calculations with m e a s u r e m e n t s shows a discrepancy for HONO and HNO3 concentrations in the plume caused perhaps by missing (heterogeneous) reactions in the reaction set. A parametrization of the sub-grid chemistry has been developed to take into account sub-grid effects (found in the study using the exhaust plume model) in a study of the global effects of NOx emissions from aircraft. This parametrization is, at first, a conversion factor. As a first estimate, it is assumed that after one day the exhaust plume has reached grid dimensions. After this, a 70% conversion of NOx into NOy is calculated. Preliminary results with the MOGUNTIA model show that this conversion results in a lower ozone production than due to unconverted emission fields. It is not clear which organic components are used to calculate ozone production. Usually the reaction set is limited to avoid excessive long computer calculation time. This may lead either to over- or underestimation of chemical species mixing ratios. The PI has furthermore clearly indicated what the main problems are when dealing with the effects of aircraft emissions on atmospheric chemistry i.e.: 9 The assumption by modellers t h a t emissions by subsonic aircraft mainly occur in the upper troposphere while in fact may be up to 50% occurs in the lower stratosphere, where the effects should be different. 9 The assumption by m a n y modellers, t h a t large scale mixing of emissions immediately occur. 9 The negligence of n o n l i n e a r reactions involving NOy species a n d heterogeneous chemistry occurring on (ice)particles in the contrail.
223
For these reasons the calculated effects given in recent literature of aircraft emissions on e.g. the 03 column density distribution are questionable. The CH4 study has not been dealt with in this s u m m a r y but will be reported at a later stage in the final report. An assessment of this part of the project has therefore been omitted.
4.5 C o n t i n e n t a l o z o n e issues; m o n i t o r i n g of trace g a s e s , d a t a a n a l y s i s a n d m o d e l l i n g o f o z o n e o v e r E u r o p e " D u t c h c o n t r i b u t i o n to t h e E U R O T R A C - T O R project" J.P. Beck1, W.A.J. van Pull, P.J.H. Builtjes2, M.G.M. Roemer2, R. Bosman2, P. Esser2, M. Vosbeek3 and W. Ruijgrok3 1 National I n s t i t u t e of Public H e a l t h and E n v i r o n m e n t a l Protection (RIVM), P.O. Box 1, 3720 BA Bilthoven 2 TNO, Institute of Environmental Sciences (IMW) P.O. Box 6011, 2600 JA Delft 3 KEMA, P.O. Box 9035, 6800 ET Arnhem
Introduction The Dutch p a r t of the EUROTRAC-TOR project consists of contributions from TNO, KEMA and RIVM. The research activities concentrate on m e a s u r e m e n t s at the high quality observatory 'Kollumerwaard', interpretation of the data, hosting the i n t e r n a t i o n a l TOR data base and chemistry and t r a n s p o r t modelling. The interpretation activities centre around four basic tasks as they were posed in the NRP-TOR project proposal: 1. How much higher is the mean ozone concentration in the boundary layer over Europe t h a n t h a t averaged over n o r t h e r n mid-latitudes, and w h a t is the seasonal, latitudinal and vertical variation of ozone within the adjacent free troposphere? Is there a secular trend in the concentrations of ozone and precursor molecules in the boundary layer or in the background atmosphere ? 2. W h a t are the emissions and distribution of the precursors responsible for the particular excess of ozone ? 3. Can we m e a s u r e how much of the excess ozone in the boundary layer spills over into the b a c k g r o u n d a t m o s p h e r e ? Is it possible to q u a n t i f y by co-measurements of ozone and other tracers the proportion of ozone produced in the t r o p o s p h e r e to t h a t t r a n s f e r r e d from the s t r a t o s p h e r e to the troposphere at our location ? 4. How much ozone and how m a n y precursors are transported across regional boundaries ? A full assessment of all the questions raised in the four tasks is beyond the scope of this extended abstract. In the following sections we discuss the Kollumerwaard observatory, the internal production of ozone in Europe, the exceedance of critical levels of ozone in Europe and exchange processes between the boundary layer and the free troposphere. A more elaborate evaluation of the above m e n t i o n e d questions is expected to be given in our final NRP report.
224
The K o l l u m e r w a a r d observatory The K o l l u m e r w a a r d o b s e r v a t o r y is one of the m a i n sites w i t h i n the EUROTRAC-TOR network. P a r t l y due to the funding from the c u r r e n t N R P project the site is now equipped with instruments measuring 03, NO, NO2, PAN, VOCs, CO, CO2, CH4, j(NO2), and meteorological parameters. Observations are generally performed on an hourly basis. In order to establish a coherent network and homogeneous observations several intercalibration experiments were carried out. The Dutch groups participated in round robins on VOCs, PAN, CO, CO2, CH4, NOx and O 3. The data are reported regularly to the TOR data base which is hosted by RIVM. F u r t h e r m o r e the data are reported to the E M E P programme and the WMO climate d a t a archive. The spatial distribution of the sites in the TOR network allows the observation of the chemical processing of polluted air typical of the European continent. The Kollurnerwaard site, bordering the North Sea, is one of the sites representing the 'European boundary layer background'.
The K o l l u m e r w a a r d d a t a (Abstract from a TOR report by M.G.M. Roemer, J.P. Beck, P. Esser and L.S. de Waal). The ozone and oxidant (=NO2 + 03) concentrations measured at Kollumerwaard and averaged over the period March 1989 - F e b r u a r y 1993 were analyzed as a function of wind direction. The wind sectors are taken at intervals of 10 degrees. The analysis demonstrates t h a t for most wind directions the ozone concentrations at Kollumerwaard are strongly influenced by ambient NOx concentrations. The ozone concentrations in the sector 090~ ~ (European continent) are 10-20 ppb lower t h a n the oxidant concentrations reflecting a shift in the photostationary state reaction due to elevated NOx concentrations. The lowest (daily averaged) oxidant concentrations are observed in the continental sector 090~ ~. The ozone and oxidant concentrations at ground level are determined by a n u m b e r of processes; influx from the free troposphere, advection from other regions, chemical production and destruction, and deposition. It will be clear t h a t air coming from the continent is more polluted with ozone precursors (VOC, CO, NOx) than marine air and that therefore the ozone forming potential is higher. However, this does not necessarily have to lead to higher ozone/oxidant concentrations. The minimum in oxidant concentrations coming from the continent can be a result of: 1. lower ozone background concentrations (in the free troposphere) over the continent than over the Atlantic Ocean; 2. continental air being so rich of NOx that it suppresses the OH concentrations and consequently blocks ozone production and/or 3. dry deposition. The 03 and Ox concentrations were also evaluated as a function of wind speed. Except for the s u m m e r season the ozone and oxidant concentrations tend to converge to some upper value with increasing wind speed. Stronger wind implies more t u r b u l e n t mixing and consequently a better exchange of air with the free troposphere. Since from these m e a s u r e m e n t s alone we cannot establish w h e t h e r at high wind speeds the observations represent truly free tropospheric air we call it pseudo free tropospheric air. In the northern wind sector the free tropospheric oxidant concentrations are 30-33 ppb in winter. During the s u m m e r months a different behaviour is observed. Ozone and oxidant concentrations increase at low
225
wind speeds until a m a x i m u m and decrease thereafter. A possible explanation m a y be found in the factor t h a t bad w e a t h e r and associated full cloud cover during the passage of depressions considerably lowers the ozone production. The ozone d a t a from the former Kloosterburen site and the K o l l u m e r w a a r d site were analyzed on trends. The trend in the ozone concentration from all d a t a is -1.2% per y e a r (1979-1993) and of oxidant it is -1.2% per year. The s a m e downward feature is observed from the dataset from the clean sector in the north. However, the annual trend in this sector is less downward (-0.6 and -0.9%.year for ozone and Ox respectively) then the trend averaged over all wind directions.
The i n t e r n a l p r o d u c t i o n o f ozone within Europe ( S u m m a r y from Beck and Grennfelt, 1994). The i n t e r n a l production of ozone within Europe was e s t i m a t e d by selecting monitoring stations from the TOR network which were found to be very little affected by local sinks as well as by European ozone production and destruction. Emission of NOx from traffic in the vicinity of a monitoring station is an important sink for ozone, in particular during inversions when high NO concentrations m a y consume a s u b s t a n t i a l p a r t of the ozone. Another sink is the deposition to vegetation during nighttime inversions. Both sinks m a y cause systematic diurnal v a r i a t i o n in the concentration of 03. In this evaluation the ratio m a x i m u m to m i n i m u m (O3max/O3min) hourly concentrations from the mean diurnal variation for the s u m m e r and w i n t e r half year respectively were used. The ratios varied considerably; from 5 to 8 in the areas most influenced by local sinks to 1.04 - 1.40 in the areas with the smallest influence from local sinks. The second criterion was t h a t the stations should very seldom experience ozone episodes. Only four lowland (<500 m a.s.1.) monitoring stations were found fulfilling these criteria. These were JelSy and Svanvik in Norway, Mace Head in Ireland and Strath Vaich in Scotland. At these stations ozone episodes very seldom occurred. The m e a n diurnal variation from the four background stations were a s s u m e d to represent the background ozone concentration in Europe. The s u m m e r t i m e m e a n c o n c e n t r a t i o n at the four background sites was 32 ppb. The i n t e r n a l ozone production was then calculated as the difference between the diurnal curves from the actual monitoring station and the mean for the four background sites. At most sites this approach gives a production of ozone during daytime (mainly between 9:00 and 21:00h) and a d e s t r u c t i o n of ozone during the night. Since most v e g e t a t i o n effects are caused by s t o m a t a l u p t a k e of ozone, it m a y be more appropriate to look only at daytime differences. Based on these assumptions the internal ozone production in Europe has been calculated to be of the order of 10-30 ppb at remote lowland sites in central Europe. In the south of Scandinavia the internal ozone production is of the order of 5-8 ppb. In the south of the UK, as well as at sites on the continent more influenced by local and mesoscale pollution, several sites show a m e a n ozone destruction, which at some sites m a y exceed 10 ppb. Table 4.5 summarizes the analysis on internal ozone production in Europe.
226 Table 4.5 I n t e r n a l ozone production in Europe calculated from the TOR network as the difference in daytime ozone concentration between the actual monitoring station and the mean of the concentrations at four 'boundary layer background' sites Monitoring area
Continental Europe; remote sites Continental Europe; urban sites British Isles - south of 54~ British Isles - north of 54~ Scandinavia - south of 62~ Scandinavia- north of 62~
Number of stations Internal ozone production (ppb) 6 8 9 6 12 6
5 -30 -5-+10 -13-+8 0-3 5-8 -3-+3
Ozone concentrations in Europe in relation to the c r i t i c a l level concept (The following section is a summary from a paper by Grennfelt and Beck presented at a UN-ECE meeting in Geneva, 1993). In order to develop effect oriented control strategies for regional air pollution the concept of critical loads and levels was developed. The concept was accepted in Europe for the assessment of the effects of acidic deposition and it will probably be the basis for the development of control strategies for ozone as well. It is now based on the exceedance of ozone concentrations over a threshold concentration and it is expressed in ppb.hours units. One important issue is to w h a t extent episodic ozone and background ozone contribute to the exceedance of the proposed 40 ppb base concentration. If episodes make the largest contribution, it seems reasonable to concentrate control strategies to eliminate episodic ozone but if the background ozone makes an important contribution it may be necessary to direct strategies to the free tropospheric background. The importance of episodes for a selected number of TOR sites was evaluated. If we define episodes as hours when the ozone concentration exceeds 60 ppb, episodes contribute to the exceedances by more t h a n 50% at most sites, where the exceedance of less t h a n 2000 ppb.hours per growing season, the contribution from episodes is at most sites less than 50% and at remote sites less t h a n 30%. So the central European oxidant problem is mostly an episode problem while episodes play a minor role for the exceedance of the critical levels at the outskirts of Europe. It is obvious t h a t the ways the critical levels are formulated may influence the control strategies for photochemical ozone in Europe. If critical levels are set to an exceedance of 50 ppb, the control of ozone will almost entirely be directed towards central Europe and one may assume that the exceedance may disappear at the outskirts by controlling the central European problem. If, on the other hand, the base level is set to 30 ppb, ozone becomes an all - European problem and control strategies may be directed towards all Europe. Since non-episodic conditions play an i m p o r t a n t role for the exceedance of the 30 ppb level it may be necessary to direct control m e a s u r e s to sources outside Europe and also to longer lived components like CH4 and CO.
227
Exchange of ozone between the atmospheric boundary layer and the free troposphere (This section s u m m a r i z e s a p r e s e n t a t i o n by J.P. Beck at the EUROTRAC symposium in Garmisch Partenkirchen, April 1994). The ozone b u d g e t in the t r o p o s p h e r e is composed of t r a n s p o r t from the s t r a t o s p h e r e , photochemical production, deposition at the earth's surface and photochemical destruction. The net photochemical production includes ozone formation both in the free troposphere (FT) and in the atmospheric boundary layer (ABL). There is transfer of air between the FT and the ABL; therefore a separate ozone b u d g e t of both layers m a y be d r a w n up. Despite the fact t h a t the importance of t r a n s p o r t processes on the FT and ABL ozone budgets has been known for long, there is still a large quantitative uncertainty about the fluxes involved. Except for transfer of ozone between the ABL and the FT, the t r a n s p o r t of VOC and NOx is also of importance because vertical mixing of these components may, due to the non-linearity in ozone formation, cause more efficient ozone production from the same emissions in the FT than would occur directly in the ABL. The influence of the diurnal cycle in the ABL depth, caused by convective activity, on exchange between the ABL and the FT was studied using a combination of the g r o u n d based TOR network, vertical profiles of ozone and meteorological parameters, synoptical information over landcover data. These 'ingredients' were i n t e g r a t e d with p a r a m e t r i z a t i o n s of atmospheric processes in a d a t a analysis model. The convective exchange processes and their influence on the ozone budget at the TOR station Uccle (B) were evaluated with this data analysis model for every hour covering the full year of 1989. The analysis revealed a flux directed from the ABL to the FT at the Uccle (B) location averaged over June, July and August 1989 of 0.77 x 10-3 mole.m-2 .day. The downward FT to ABL flux was 0.33 x 10-3 mole.m-2.day in the same period. So the net flux is directed upward from the ABL to the FT and has a m a g n i t u d e of 0.43 x 10-3 mole.m-2.day. If we multiply this n u m b e r with the surface of the countries in northwestern Europe (2,442,510 x 106 m2) it results in a net ABL to FT t r a n s p o r t of 1.0 Gmole.day-1. This corresponds to 4 - 6% of the daily cross tropopause flux of ozone in the northern hemisphere in summer4. Besides convective growth of the ABL depth, the exchange of air between the ABL and the FT takes place through several meteorological phenomena like clouds, frontal systems, heat island phenomena etc. The influence of these processes is not assessed in the present analysis.
Continental modelling The 1) 2) 3)
ozone budget over Europe consists of three terms: transport in and from other areas including the free troposphere; chemical production and destruction of ozone in the European boundary layer loss of ozone due to dry deposition.
Footnote added by the coordinator (R. Guicherit). Roemer (1990, report R 90/223) has calculated by means of a 2D-model a yearly net ABL to FT flux of 70 Mton 03 (0.25 Gmole.day-1) for an zonal area covering 30-55 N.
4
228 The LOTOS (LOng Term Ozone Simulation) model is one of the few models covering the whole of Europe and describing the major t r a n s p o r t and chemical processes important to photochemistry. Model simulations show that in Western Europe the influence of transport of ozone from elsewhere, i.c. The Atlantic Ocean plays an i m p o r t a n t role for the plant growing season ozone concentrations in this part of Europe. A reduction of 20% of the ozone values advected from over the Atlantic Ocean r e s u l t e d in 5-10T reduction of growing season ozone concentrations over Western Europe. This reduction is somewhat larger t h a n reduction achieved by a 30% reduction of anthropogenic European VOC and NOx emissions and much more t h a n the effect of a 30% CO emission reduction. The effect of Atlantic inflow of course becomes less over the more continental parts of Europe.
Model intercomparisons Three E u l e r i a n models [EURAD (Univ. of Cologne), REM (Univ. of Berlin) and LOTOS (TNO) and one Lagrangian model (EMEP) have been used and are being used in an intercomparison of European models. If the model results are evaluated on a r a t h e r general level, for instance daily or weekly averages and tendencies, the four models behave in quite a similar fashion with respect to the calculated ozone values. However, if the comparison is carried out in more detail (time specific values, wind roses) the discrepancies can become rather large. It appears t h a t this also holds for the comparison with observations. Time series of observations and model results are found to be in a band of 20-30 ppb width. A g r e e m e n t or disagreement with observations are evenly distributed among the models. None of the models appear to be of better or even superior quality t h a n the others. Given the differences among the models it seems t h a t they have features in common t h a t limits their success in simulating ozone time series at specific sites. The coarseness in resolution, spatially as well as temporal, and the lack of detailed input information (emissions) seems likely candidates in creating discrepancies between models and observations.
Assessment In this project amongst other, the fluxes into and out of the E u r o p e a n free troposphere are studied and determined. In this way the contribution of Europe to the global or hemispherical budget of 03, CO2, CH4 and NOx can be determined. The regional aspect of 03 as a greenhouse gas will also be evaluated at a later stage. Although continental in n a t u r e this project is of extreme importance to u n d e r s t a n d production and destruction processes of one of the i m p o r t a n t greenhouse gases i.e. 03. (As a m a t t e r of fact 03 next to CO2 may become, within a few decades, the most important tropospheric greenhouse gas over the Northern H e m i s p h e r e at our latitudes). Moreover "photosmog" has now become an environmental issue of global dimensions of which adverse effects are worse t h a n for e.g. CO2. Finally, 03 impacts on the lifetimes and, consequently on the troposphere concentrations of most atmospheric trace gases, including CH4 and hydrogen containing replacements of the CFC's, which are themselves greenhouse gases. For this reason results obtained by this project are of importance for the consistency of environmental policy making. The project has already given some insight in the processes mentioned before; the 03 trends in the lower (free) troposphere and the 03 budget in the free troposphere. Interesting to note t h a t for
229 some wind sectors there seem to be a decrease in 03 levels for the period (1979-1993) while in others a slight increase? This is a peculiar situation because other European stations show 03 levels which are still on the increase. However, the percentage increase seem to have slowed down in recent years. 5.
ACKNOWLEDGEMENTS
The theme coordinator wishes to thank the 'platform' members : P.J.H. Builtjes, F. de Leeuw and A.P. van Ulden for their help, contribution and useful discussions. Also the PI's are gratefully acknowledged for making the coordinators job more easy by promptly submitting their extended summaries. 6.
REFERENCES
Allaart, M.A.F., Helder, H., Heijboer, L.C. 1993. On the relation between ozone and potential vorticity. Geophysical Research Letters 9: 811-814. Allaart, M.A.F., Heijboer, L.C. Kelder, H., 1994. On the transport of trace gases by extra-tropical cyclones. Proceedings of the Quadrennial Ozone Symposium 1992, p. 82-84. Andreae M.O., 1990. Proposed feedback cycle between climate and marine DMS production. Marine chemistry 30: 1-29. Arnold, F. and G. Knop, 1987. Stratospheric trace gas detection using a new balloon-borne ACIMS method. Int. Journ. of Mass and Ion Processes 81: 33-44. Arnold, F., J. Scheid, Th. Stilp, H. Schlager, M.E. Reinhardt, 1992. Measurements of jet aircraft emissions at cruise altitude I" the odd-nitrogen gases NO, NO2, HNO2 and HNO3. Geophys. Res. Lett. 12: 2421-2424. Beck, J.P. and Grennfelt, 1994. The internal production of ozone within Europe. Atm. Env. Vol. 28: 129-140. Beck, J.P., C.E. Reeves, F.A.A.M. de Leeuw, S.A. Pencket, 1992. The effect of aircraft emissions on tropospheric ozone in the northern hemisphere. Atm. Env. 25A: 17-29. Bernhard, G., 1993. Einflul] von Diffusor-Eigenschaften auf die Bestimmung von Bestrahlungsst~irken im UV-Bereich: Versuchsaufbau, Messung und Korreckturverfahren, IFU Schrii~enreihe, Bd. 22, Garmisch-Partenkirchen. Bordewijk, J.A., Slaper, H., Reinen, H.A.J.M., Schlamann, E., 1994. Reduction factors of effective-UV and global radiation; a method to calculate effective-UV daily doses from measurements of global radiation (In press). Bruhl, C., and P.J. Crutzen, 1988. Scenarios of possible change in atmospheric temperatures and ozone concentration, due to man's activities, estimated with a one-dimensional coupled photochemical climate model. Climate Dynamics 2" 173-203. Brfihl, C. and P.J. Crutzen, 1989. High sun conditions, Geoph. Res. Lett. 16: 703-706. Cess, R.D. et al., 1989. Interpretation of cloud-climate feedback as produced by 14 atmospheric General Circulation Models, Science 245: 513-516. Charlson, R.J., J.E. Lovelock, M.D. Andreae and S.G. Warren. 1987. Nature 326: 655-661;
230 Christensen, J.H. and E. van Meijgaard, 1992. On the construction of a regional atmospheric climate model, KNMI TR-147. Crutzen, P.J. and L.T. Gidel, J., 1983. A two dimensional photochemical model of the admonsphere. 2: the tropospheric budgets of the anthropogenic chlorocarbons, CO, CH4, CH3CL and the effects of various NOx sources on tropospheric ozone. Geoph. Res. 88: 6641-6661. Curtis, A.R., W.P. Sweetenham, 1985. FACSIMILE Release H User's Manual. Report AERE R 11771. Curtis, A.R., and W.P. Sweetenham, 1987. Facsimile/checkmat., Harwell Lab. Report No. AERE R 12805. De Leeuw, F.A.A.M., 1988. Modelmatige berekening van fotolyse snelheden relevant voor troposferische chemie. RIVM report 228603003 (in Dutch). Dentener, F., 1993. Heterogeneous chemistry in the troposphere. Ph.D. Thesis, Universiteit Utrecht. Derwent, R.G. 1982. Two-dimensional model studies of the impact of aircraft exhaust emissions on tropospheric ozone. Atm. Env. 16: 1997-2007. Deshler, T., B.J. Johnson and W.R. Rozier, 1993. Balloon-borne measurements of Pinatubo aerosol during 1991 at 450 vertical profiles, size distributions and volatility. Geoph. Res. Lett. 20, 14: 1435-1438. Drummond, J.W., D.H. Ehhalt, A. Volz, 1988. Measurements of Nitric Oxide Between 0-12 km Altitude and 67 N to 60 S Latitude Obtained During STRATOZ III. J. Geophys. Res. 93, 15: 831-849. Ehhalt, D.H., F. Rohrer, A. Wahner, 1992. Sources and distribution of NOx in the upper troposphere at Northern Mid-Latitudes. J. Geophys. Res. 97: 3725-3738. Feijt, A.J., 1992. The e a r t h radiation budget experiment" overview of data-processing and error sources, KNMI TR-146. Fortuin, P., Langematz, U., Dorland, R. van, Kelder, H., 1994. Concurrent ozone and temperature trends derived from sonde measurements, submitted to the proceedings of "NATO-ASI atmospheric ozone as a climate gas". Gardiner, B.G., and Kirsch, P.J., 1993. Setting Standards for European Ultraviolet Spectroradiometers, Draft report, Cambridge. Gelinas, R.J., J.J. Walton, 1974. Dynamic-kinetic evolution of a single plume of interacting species. J. Atm. Sciences p. 1807-1813. Gezondheidsraad 1993. UV straling uit zonlicht; Advies van een commissie van de Gezondheidsraad. Gidel, L.T., P.J. Crutzen and J. Fishman, 1983. A two dimensional photochemical model of the atmosphere. 1: chlorocarbon emissions and their effect on stratospheric ozone. Geoph. Res. 88" 6622-6640. Gillotay, D., The 'Transportable Lamp System', a relative reference for the future. Final Report EC-STEP 076, Brussels. Gregory, G.L., Hudgins, C.H. and R.A. Edahl, 1994. Laboratory evaluation of an airborne ozone i n s t r u m e n t t h a t compensates for altitude/sensitivity effects.Env. Sci. Technol. 17: 100-103, 1983. Hauglustaine et al., 1994. Model calculations. J. of Geoph. Res. 99: 1173-1186. Heijboer, L.C. and Kelder, H. A nonlinear analytical model of developing extra-tropical cyclones with the associated flow regimes and transport of trace gases. Submitted to Quarterly Journal of the Royal Meteorological Society.
231 Hidalgo, H. and P.J. Crutzen, 1977. The tropospheric and stratospheric composition perturbed by NOy emissions of high-altitude aircraft. J. Geophys. Res. 82: 5833-5866. Hfibler, P.C. Murphy, 1987. Ozone production in the rural troposphere and the implications for regional and global ozone distributions. J. Geophys. Res. 92: 4191-4207. Isaksen, I.S.A., 1980. The tropospheric ozone budget and possible man made effects. Proc. of the Quadrennial Ozone Symposium. WMO, Boulder, 4-9 August. Kawa, R.S. et al., 1990. Measurements of total reactive nitrogen during the airborne arctic stratosphere expedition. Geoph. Res. Lett. 17, no. 4: 485-488. Kelder, H. and Teitelbaum, H., 1992. A note on instabilities of as Kelvin-Heimholtz velocity profile in different approximations. I1Nuovo Climento 14: 107-118. Kelder, H., Velthoven, P. van, Allaart, M., 1994. Exchange between the troposphere and the stratosphere. Accepted for publication in the proceedings of the EUROTRACK symposium 1994. Kneizys, F.X. et al., 1988. Users' guide to Lowtran-7, Air Force Geophysics Laboratory, Hanscom AFB(MA, AFGL-TR-88-0177. Lary, D.J. and J.A. Pyle, 1991. Diffuse radiation, twilight and photochemistry. Journ. of Atm. Chem. 13: 373-392. Liu, S.C., M. Trainer, F.C. Fehsenfeld, D.D. Parrish, E.J. Williams, D.W. Fahey, G. Loewenstein, M. et al., 1993. New obervations of the NOy/N20 correlation in the lower stratosphere. Geoph. Res. Lett. 20, (22): 2531-2534. Lott, F., Kelder, H., Teitelbaum, H., 1992. A transition from Kelvin-Helmholtz instabilities to propagating wave instabilities, Physics of Fluids 4: 1990-1997. Madronich, S. 1987. Photodissociation in the atmosphere: 1. Actinic flux and the effects of ground reflections and clouds. J. Geophys. Res. 92: 9740-9752. McInnex, G., and C.T. Walker, 1992. The global distribution of aircraft air pollutant emissions. Warren Spring Laboratory, LR 872 (AP). Morcrette, J., 1991. Radiation and cloud radiative properties in the European Centre for Medium Range Weather Forecasting system, J. Geophys. Res. 96: 9121-9132. Mfiller, R., and P.J. Crutzen. Chlorine activiation mechanisms during polar night. J. Geoph. Res. 98: 20483-20490. Mfiller, J.F., 1992. Geographical distribution and seasonal variation of surface emissions and deposition velocities of atmospheric trace gases. J. Geophys. Res. 97: 3787-3804. Ramanathan, V. et al., 1989. Cloud radiative forcing and climate: Results from the earth radiation experiment, Science 243: 57-63. Reinen, H.A.J.M., Schlamann, Ec., van Sonderen, J.F., Slaper, H., 1993. A new solar LW radiation monitoring station in The Netherlands. Stamnes, K.H. (ed.): In Atmospheric Radiation, Proc. SPIE 2049: 340-349. Reinen, H.A.J.M., Schlamann, E., Bordewijk, J.A., Slaper, H. 1994. Monitoring UV radiation in The Netherlands set-up and preliminary results. In press in proceedings of the European symposium on effects of environmental UV radiation Munich 27-29 okt. 1993. RIVM 1993. Nationale Milieuverkenning 3: 1993-2015; ISBN 90 6092 679 X. Rossow, W.B. and L.C. Garder. 1993. Validation of ISCCP cloud detections. J. Climate 6: 2370-2393.
232 Saunders, R.W. and K.T. Kriebel. 1988. An improved method for detecting clear sky and cloudy radiances from AVHRR data, Int. J. Remote Sensing 9: 123-150. Siebesma, A.P. and Cuijpers, J.W.M.: Evaluation of parametric assumptions for shallow cumulus convection. Submitted to Journal of atmospheric sciences 1993. Slaper, H., den Elzen, M.G.J., van der Woerd, H., de Greef, J., 1992. Ozone depletion and skin cancer incidence: an integrated modelling approach; June 1992; RIVM report 749202001. Slaper, H., den Elzen, M.G.J., van der Woerd, H.J., 1994. Ozone depletion and skin cancer risks: a scenario study. (In press). In proceedings of the European symposium on effects of environmental I_W radiation Munich 27-29 okt. 1993. Slaper, H., 1993. Ozone depletion and skin cancer incidence: science and policy implications of risk-oriented approaches: in press in proceedings of the US-Dutch symposium: Comparative Risk Analysis and Priority Setting of Air Pollution Issues, June 7-11, Keystone, USA. Slaper, H., 1987. Skin cancer and UV-exposure: investigations on the estimation of risks, Thesis, University of Utrecht. Sonderen, J.F. van, Slaper, H., Swart, D.P.J., Smetsers, R.C.G.M., 1991. D e f i n i t i e s t u d i e s p e c t r o m e t r i s c h e UV-meetopstelling. RIVM r e p o r t 249104001. Stammes, P. et al. 1994. TEBEX observations of clouds and radiation - potential and limitations. KNMI TR-162. Teitelbaum, H., Ovarlez, J., Kelderman H, Lott, F., 1994. Some observations of gravity-wave-induced structure in ozone and water vapour during EASOE. Geophysical Research Letters 21: 1483-1486. Tolbert, M.A., M. Rossi and D.M. Golden, 1988. Hetrogeneous reactions of CIONO2, HCL and HNO3 with sulfuric acid surfaces at stratospheric temperatures. Geoph. Res. Lett. 15. No. 8 847-850. Tsay S.Ch. and Stamnes K., 1992. Ultraviolet radiation in the Arctic: the impact of potential ozone depletion and cloud effects. J. Geophys. Res. 97: 7829-7840. UNEP 1991. Environmental effects of ozone depletion. Panel Report Pursuant to article 6 of the Montreal Protocol on Substances that deplete the ozone layer under Auspices of the UNEP. Von Clarmann, T. et al., 1993. Retrieval of stratospheric 03, HNO3, and CLOnO2 profiles form 1992 MIPAS-B limb emission spectra, results and error analysis. Journ. of Geoph. Res. 98: No. D l l . Van der Hage, J., Boot W., van Dop, H., Duynkerke, P.G. and Vil~-Guerau de Arella, J., 1994. A photo-electric detector suspended under a balloon for actinic flux measurements. Journal of Atmospheric and Oceanic Technology 11: 674-679. Van Weele, M. and Duynkerke, P.G., 1993. Effect of clouds on the photodissociation of NO2: observations and modelling. Journal of Atmospheric Chemistry 16: 231-255. Van Weele, M., Vil~-Guerau de Arellano, J. and Kuik, F., 1994. Combined measurements of UV-A actinic flux, I.W-A irradiances and global radiation in relation to photodissociation rates. Submitted to Tellus.
233 Van Weele, M., Vil~-Guerau de Arellano, J. and Duynkerke, P.G., 1993. A parameterization of the effects of clouds on photodissociation rates; comparisonwith observations. Proceedings XX ITM on Air Pollution Modelling and its Application, Valencia, Spain. Velders, G.J.M., Heijboer, L.C., Kelder, H., 1994. The simulation of the transport of aircraft emissions by a three-dimensional global model. Annales Geophysicae 12: 385-393. Velthoven, P.F.J. van, Wauben, W.M.F., Kelder, H. 1994. A comparative study of global 3-D transport of NOx emitted by aircraft. Proceedings of "Impact of Emissions from aircraft and spacecraft upon the atmosphere", 174178. Verver, G.H.L., 1994. Comments on: "A modified K-model for chemically reactive species in the planetary boundary layer" by Fujihiro Hamba. Accepted for publication in Journal of Geophysical Research Letters. Vil~-Guerau de Arellana, J., Duynkerke, P.G. and van Weele, M., 1994. Tethered-balloon measurements of actinic flux in a cloud-capped marine boundary layer. J. Geophys. Res. 99: 3699-3705. Volz-Thomas, A., 1992. Proceedings of the EUROTRAC Symposium '92. In Borrell P.M. et al. (eds). Ac. Publ. B.V. The Hague. Wienhold, F.G., T. Zenker, and G.W. Harris, 1994. A dual channel two tone modulation tunable diode laser spectrometer for ground bases and airborne trace gas measurements, to be submitted to Appl. Opt. Zimmermann, P.H. 1988. Moguntia: a handy global tracer model. In: Van Dop, H. Air Pollution Modelling and its applications VI. NATO/CCMS, Plenum, New-York.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
237
U l t r a v i o l e t radiation and p h o t o c h e m i s t r y in clouds: o b s e r v a t i o n s a n d m o d e l l i n g Jordi Vil5-Guerau de Arellano, Peter G. Duynkerke and Michiel van Weele IMAU, Utrecht University, Pfincetonplein 5, 3584 CC Utrecht, The Netherlands
Abstract Ultraviolet (UV) radiation plays a key role in the chemistry of the troposphere since it is in these wavelengths of the solar spectrum that radiation breaks molecules into reactive atoms and free radicals by photodissociation. It is for that reason that the Institute for Marine and Atmospheric Research Utrecht (IMAU) started in 199() a research project within the frame of the Dutch National Research Programme on Air Pollution and Climate Change (NRP). The strategy of the research project was to combine measurements and model calculations in order to determine accurate photodissociation rates for chemical species. The study focused on the actinic flux, a radiometfic quantity which determines the photodissociation rates for chemical species. Measurements and model calculations were focused on study the behaviour of the actinic flux under the presence of clouds and for various ground albedo.
1. I N T R O D U C T I O N In the troposphere key chelnical species such as ozone, hydrogen peroxide and nitrogen dioxide are dissociated by sunlight in the ultraviolet spectral region. As UV radiation passes through the atmosphere it is modified by absorption and scattering by gases and aerosols. In addition, UV radiation and therefore the effects on chemical species ale influenced by the solar zenith angle, the ground albedo and the presence of clouds (Figure 1).
sun
ozone. . layer ,.... . . 9:'..~..'.;. . "-
-"
.~ '.: "-':
i.$ .....
:
",
~ .: 9
"".:"
"-~
,,:
: . ' - , . , , . . ~ 2 - - , - , , - . 4 : : . . . - . , . ~ ; - - :....,.,-~--., ..-'.'e,,g." 9 : 9 9 - . . . -.,% - ' ~ " , , . . . ". ., '.5. '.% . -~ - -.q. ...--. . . . . : . . . .. . . . " 9 ' -" ' t-.;. - . , ". -. -' ,". -
"'
9
".~- % . .'.
"" ~
,.
.
.
.
.
"
...
"
"
"
~ :;.,'~.,.r-"
:"~.~4:;
9
f.'~,.:.:.'.
'.~',".-
:"
.
~
)lecular scattering
' / aerosol parti
.
clouds .
\t
.
~ .
<,Xv
.....
ocean
land
Figure 1. The physical processes relevant for the transfer of UV radiation in the atmosphere.
238
The photodissociation rates fl~r chemical species i (Ji) can be derived from
Ji (),,,T) = f
(Yi
(X,T) Oi (~,,T) F (X) dPv
(1)
where (Yi ()v,T) is the absorption cross section and ~i()v,T) is the quantum yield. Both quantities are properties of the chemical species and are dependent on the wavelength ~ and the temperature T. However, the actinic flux F is a radiometric quantity which is defined as the radiance integrated over all solid angles, i.e. over 4rt, per unit area. Consequently, the actinic flux differs fiom the usually measured irradiance E, which is the radiance incident on a horizontal surface per unit area [1] and [2]. Vertical profiles of actinic flux measured under clear and cloudy conditions by means of a tethered-balloon are presented. The measurement campaign was carried out during the Atlantic Stratocumulus Transition EXperiment (ASTEX) [3]. The strategy of ASTEX is based on a combination of modelling activities and satellite, airborne, ship and surface observations. Therefore, the campaign yielded complete and detailed meteorological, oceanographic and atmospheric chemistry data sets. In addition to the above-mentioned observations, ground level measurements of actinic flux were taken at different locations in the world: Azores, Antarctica and De Bilt (The Netherlands). These observations were made to study the influence of ground albedo in the actinic flux and consequently in the photodissociation rates. A comparison was made between the actinic flux measurements and radiative transfer calculations performed with a 8-Eddington model [1] and [2]. The comparison shows that these observational data are an excellent basis for the evaluation of radiative transfer models in the UV-A spectral region, i.e. between 315 and 4()() nm.
2. PHOTO-ELECTRIC DETECTOR TO MEASURE ACTINIC FLUX The actinic flux was measured with a photo-electric detector developed at IMAU [4]. This instrument measures the incoming radiation from all directions (actinic flux), whereas the standard UV radiometer measures only the UV radiation falling on a horizontal plate (irradialace). Briefly, the instrument consists of six diodes covered by diffuse filters, projecting from a cylinder, and arranged perpendicular to each other. The directional response of the photo-electric detector deviates less than 5% from perfect isotropy. The instrument covers the spectral range 33(~ to 39{)nm. It is in this range that e.g. N()2 photodissociates in the following way: NO2 + hv (X < 42{)nm) ---) N() + O
(2)
0 + 02 + M --~ (i)3 + M,
(3)
where hv denotes the radiation incident on a molecule, M is the molecule which absorbs the excess of energy and NO, (i)2 and (i)3 are nitric oxide, oxygen and ozone, respectively. The photodissociation of N()2 is key reaction in the formation of (i)3 in the troposphere. The photo-electric detector was calibrated against the irradiance measured by a UVradiometer. The instrument measures the total and direct downward irradiance at 367 nm. To calibrate the photo-electric detect~w, all the incoming radiation has to be diffuse. Under this condition, the actinic flux is linearly related to the irradiance. The unit of the actinic flux measurements is W m -2 nln-1. The instrument was designed to be used under a tethered balloon but can as well be used to make surface measurements.
239
3. A C T I N I C FLUX M E A S U R E M E N T S AND M O D E L R E S U L T S The first actinic flux measurements were carried out within the framework of ASTEX [5]. The observations were made between 1 and 3() June 1992 on Santa Maria island (36.99 ~ N, 25.17 ~ W), the Azores. More than 34 tethered-balloon soundings were carried out. Figure 2 shows the measured vertical profiles of actinic flux. These observations are compared with model calculations perforrned with the above mentioned 8-Eddington model. The cloud characteristic input necessary for the model calculations was provided by other research groups participating in ASTEX. For the same zenith angle the vertical profile of actinic flux is also calculated under clear sky conditions. An excellent agreement is found between the observations and the model calculations. The figure also shows the different behaviour of the actinic flux (photodissociation rates) under cloud and clear sky conditions. Under clear sky conditions, a slightly increasing tendency for the actinic flux in the whole atmospheric boundary layer is observed. Under cloud conditions, the actinic flux values are lower below the clouds than in a cloud-free sky because part of the radiation is shielded by the cloud. Inside the cloud the actinic flux increases ahnost linearly. Above the cloud the value of actinic flux was found to be higher than on a clear day because the cloud albedo has a higher value than the albedo of Earth's surface. Both observations and model calculations show that the different vertical profiles are dependent on the solar zenith angle and the cloud optical depth.
1000
,
I
1
I.
,
I
,
1
i Cloudtop ~
800-
" +_,,~_~
600- Clou base ~
400200'
I
0.5
I
I
1
1.5
F [Win- 2nm
I
1]
2
2.5
Figure 2. Observed vertical profiles of actinic flux: crosses represent upsounding; open circles downsounding. Solid line is the vertical profile calculated with the model under the presence of a cloud. Dotted line is the vertical profile for a clear sky situation. At ground level actinic flux measurement were taken in the Azores, Antarctica (74.58 ~ S, 11.12 ~ W) and De Bilt (52.()()~ N, 5.18 ~ W) [6]. The measurements at the different locations were all performed around the summer solstice. Figure 3 shows the different behaviour of actinic flux compared with the standard radiometric quantity global radiation G (irradiance) under clear sky conditions at different sites. The figure shows the large effect that the high surface albedo (about ().9) has on the actinic flux measurements. The albedo on the other two sites was ().()5.
240
2.5 o
Azores; 19 Jun 1992 ~I~ Antarctica; 31 Dec 1992 ~i~ De Bilt; 30 Jun 1993 ~Y'N De Bilt; 1 Jul 1 ~
A
2
+ x
~,~I.5AA
0.50 "I"-
0
'
I
200
'
I
'
I
400 600 G(Wln -2)
'
I
800
'
1000
Figure 3. Measured UV actinic flux as a function of measured global radiation for clear sky conditions. In summary, both measurements and model calculations stress the necessity to describe accurately clouds and take ground albedo into account to obtain accurate photodissociation rates for chemical species. Based on these measurements and model calculations, a parameterization which describes the influence of clouds and ground albedo on photodissociation rates was developed [7]. This paralneterization can be readily implemented in atmospheric chemistry models. 5. R E F E R E N C E S
[ 1] Madronich, S. 1987b. Photodissociation in the atmosphere, 1, Actinic flux and the effect of ground reflections and clouds, J. Geophys. Res., 92, 974()-9752. [2] Van Weele, M., and Duynkerke, P. G. 1993. Effect of clouds on the photodissociation of NO2: observations and modelling, J. Atmos. Chem., 16, 231-255. [3] ASTEX 1992. ASTEX Operations Plan. Available from the FIRE Project Office, MS 483, NASA Langley Research Center, Hampton, VA, USA. [4] Van der Hage, J. C. H., Boot, W., Van Dop, H., Duynkerke, P. G. and Vil'h-Guerau de Arellano, J. 1994. A photo-electric detector suspended under a balloon for actinic flux measurements, J. Atmos. Ocean Technol., 11, 674-679. [5] Vilh-Guerau de Arellano, J., Duynkerke, P. G., and Van Weele, M. 1994. Tetheredballoon measurements of actinic flux in a cloud-capped marine boundary layer, J. Geophys. Res., 99, 3699-37(i)5. [6] Van Weele, M., Vilh-Guerau de Arellano, J., and Kuik, F. 1995. Combined measurements of UV-A actinic flux, UV-A irradiance and global radiation in relation to photodissociation rates, Tellus , 47B, vol. 3, (in press). [7] Van Weele, M., Vilh-Guerau de Arellano, J. and Duynkerke, P. G.1994. Evaluation of a parameterization of the effect of clouds on photodissociation rates with actinic flux measurements. Proceedings of the VI European Symposium on the "Physico-chemical behaviour of Atmospheric pollutants", Varese, Italy, pp 104()-1(i)45.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) o1995 Elsevier Science B.V. All rights reserved.
T H E M A R I N E S U L F U R - C Y C L E : I M P O R T A N C E OF P H A E O C Y S T I S
241
SP. IN
DMS-PRODUCTION DURING A NEARSHORE SPRINGBLOOM Jacqueline Stefels 1, Lubbert Dijkhuizen 2, Winfried W.C. Gieskes I University of Groningen, Dept. of Marine Biology I and Dept. of Microbiology 2 P.O. Box 14, 9750 AA Haren, the Netherlands Abstract Potential enzymatic conversion of dimethylsulfoniopropionate (DMSP) to dimethylsulfide (DMS) was measured in natural seawater samples during the 1993 spring bloom off the Dutch coast. Good con'elations were found with Phaeocvstis sp. numbers, indicating the presence of a specific DMSP-lyase in this often so dominant algal species. The results suggest an impol~ant role of Phaeocvstis sp. in the conversion of DMSP to DMS. INTRODUCTION Dimethylsulfide (DMS) is thought to be involved in the biological regulation of the climate: 90 to 95% of the aerosols found above remote oceans consist of non-seasalt sulfate that is formed by gas-to-particle conversion of the oxidation products of organosulfur gases (principally DMS). Aerosols serve as cloud condensation nuclei (CCN). The amount of DMS released into the atmosphere influences the number of CCN and thereby cloud droplet size, cloud albedo and, consequently, climate (Andreae 1990, Charlson et al. 1987, Charlson & Wigley 1994, Malin et al. 1992). The flux of DMS from the ocean into the atmosphere is determined by its concentration in the water, which is the result of several production and removal processes (fig.l). In seawater DMS is produced from dimethylsulfoniopropionate (DMSP), a compound that is found in many lnm'ine microalgae (Keller et al. 1989). Conversion of DMSP into DMS and acrylic acid is thought to occur mainly after its release from the cells. Then, part of the DMSP is cleaved through enzymatic activity. Until now, attempts to quantify the processes as discribed in fig.1 have been scm'ce, but with today's state of knowledge, it is tempting to believe that only a small part of the DMSP-sulfur will ever reach the atmosphere (Bates et al. 1994). On the other hand, the processes described in fig. 1, of course, never occur all at the same time and with the same strength. Therefore, it is irnportant to investigate at what time which process is important, and what species are involved. In the literature it is mainly suggested that DMSP-lyase activity stems from bacteria. On the other hand, it was recently shown that an important DMSP containing species, Phaeocystis sp., also possesses a DMSP-lyase (Stefels & Van Boekel 1993). The objective of this study was to investigate whether or not there exists a relation between the potential DMSP-lyase activity in natural waters and the occurrence of Phaeocystis sp. or any other algal species present.
242
,~, ~;'/./
Icloud albedo I ,
,,.
,,
IDMSl
v ~ .,-.=.-= ~ 7 ~ v,--.-. -.-...,r .-.-. .-:,-vr.-, =.*...--,.v .--..-v..-~ r ~ v - ~ r Z . ~
y : ~ 7 . ~ T ~ 7 ~..= r . ~ ~
r.v ~
r.~ =.~ ~
physiological release ?
& lysis~.....~.nzymatic cleavac
;Ioppy feedi
DMSP dissolved
DMSP grazers J------
L
I methiolpropionate ]
~ " ] D M S P fecal pellets r
\
FIG. 1" Main production and consumption processes of DMS in the mmine environment. M A T E R I A L AND M E T H O D S
With axenic Phaeocystis sp. cultures an enzyme essay was developed, and applied on natural seawater samples off the Dutch coast, taken during the spring bloom of 1993 with RV "Pelagia". To this end, surface water samples were taken. The particulate matter of 350-1000 ml was subsequently fractionated (in > 100 gm and 10-100 gm fi'actions) and concentrated. Of each fraction subsamples were taken for cell counts; the remainder was immediately frozen at -80 ~ In the lab the samples were thawed; particulate matter was harvested through centrifugation and destroyed using a French 15"essure Cell. To the crude extract a buffer with DMSP was added after which DMS evolution was measured in the headspace using a gas chromatograph equipped with a sulfur-specific Hall Electrolytic Conductivity Detector. Part of the crude extract was used for protein measurements. RESULTS During the cruise, the northern and offshore waters were dominated by the colony forming alga
Phaeocystis sp., whereas the southern and central coastal waters mainly consisted of diatom species, giving a good impression of a typical spring phytoplankton bloom off the Dutch coast. DMSP-lyase activity measured in the samples proved to be mainly restricted to the > 100 jam fractions in which the Phaeocystis sp. colonies were trapped. DMSP-lyase activity showed a
243 very good con'elation with Phaeocystis sp. numbers (r2=0.9660, n=23), but no conelation with any other abundant species, nor with total diatom numbers, total diatom biovolume or total protein. DISCUSSION In the literature, the conversion of DMSP to DMS has mostly been attributed to bacterial activity. Our results show, however, that the alga Phaeocystis sp. has a very active DMSPlyase, specific for this species, which is potentially responsible for the conversion of DMSP to DMS during the early spring bloom off the Dutch coast. DMS production rates by Phaeocystis can be calculated for these waters, using the production rates measured in axenic cultures, Phaeocvstis numbers and dissolved DMSP concentrations found during the cruise, and a mean depth of the n'fixing layer of 5 m. In the northern part of the study area, values ranged from 47 to 131 btlnol l-n-2 day -1. We have compared these production rates with the main abiotic loss factors. Loss by air-sea exchange was estimated to be 16.6 btmol m -2 day-i; photochemical oxidation of DMS to DMSO is comparable with the flux to the atmosphere. Total abiotic loss rates can therefore be estimated to be approximately 30 l.tmol m -2 day -1. This is in the same range as DMS production by Phaeocystis; indeed a 1.5 to 4.5 times overproduction of DMS can be calculated, potentially available for bacterial consumption. Considering the conservative estimates of the pm'ameters used, production by Phaeoo, stis may even be higher. Several field studies have shown lm'ge seasonal vm'iations of DMS in the southern North Sea, with a maximum in fi'ont of the Dutch coast dm'ing the Phaeoo'stis bloom. Ore" study has made it plausible that Phaeocvstis itself plays an important role in this production of DMS. References: Andreae, M.O. (1990). Ocean-atmosphere interactions in the global biogeochemical sulfur cycle. Mar. Chem. 30:1-29 Bates, T.S., Kiene, R.P., Wolfe, G.V., Matrai, P.A., Chavez, F.P., Buck, K.R., Blomquist, B.W., Cuhel, R.L. (1994). The cycle of sulfur in surface seawater of the northeast Pacific. J. Geoph. Res. 99 (C4): 7835-7843 Charlson, R.J., Lovelock, J.E., Andreae, M.O., Warren, S.G. (1987). Oceanic phytoplankton, atmospheric sulfur, cloud albedo and climate. Natme 326:655-661 Charlson, R.J., Wigley, T.M.L. (1994). Sulfate aerosol and climate change. Scientific American 270 (2): 28-35 Keller, M.D., Bellows, W.K., Guillard, R.R.L. (1989). Dimethyl sulfide production in marine phytoplankton. In: Saltzman, E.S., Cooper, W.J. (eds.) Biogenic sulfur in the environment. ACS Symp. Set. 393, Washington DC. p. 167-182 Malin, G., Turner, S.M., Liss, P.S. (1992). Sulfur: the plankton/climate connection. J. Phycol. 28:590-597 Stefels, J., van Boekel, W.H.M. (1993). Production of DMS fi'om dissolved DMSP in axenic cultures of the marine phytoplankton species Phaeocystis sp. Mm. Ecol. Prog. Ser. 97: 11-18
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
245
Clouds-Radiation-Hydrologic interactions in a limited-area model A.C.A.P. van Lammeren, A.J. Feijt, R. van Dorland, E. van Meijgaard, P. Stammes, A.P. van Ulden
Royal Netherlands Meteorological Institute (KNMI) P.O. Box 201, 3730 AE De Bilt
Abstract In this project work has been directed towards the improvement of the knowledge on clouds and the way they influence our climate. The activities include measurement of cloud properties on a regional scale (120x120 km:), analysis of global satellite datasets and the development of a model environment to enhance the regional data analysis and the improvement of parametrizations of clouds and radiation.
1. Introduction
Clouds play an important role in our climate. They produce precipitation which is an essential ingredient of the hydrological cycle. Clouds modify the earth-radiation budget. Thin cirrus clouds have a warming effect while low clouds have a distinct cooling effect [1]. Clouds dominate the vertical transport of energy, momentum and trace gasses in the free troposphere. Despite their importance, clouds are represented only rudimentary in climate as well as weather forecast models. It appears that the model representation of clouds in climate models has a major impact on model predictions for climate change. Cess et al. [2] showed that cloud feed-back is a major source of uncertainty in model responses to climate forcing. There are two main reasons why the uncertainties with respect to clouds are so large. The first reason is that cloud processes are extremely complicated. A proper represenation of clouds requires the parameterization of subgrid processe both on the macro-scale ( c m - km) and on the microscale (<
246 characteristics (cloud cover, cloud structure, optical depth, droplet spectra). This lack of good data hampers the development and validation of models. Satellites begin to provide usefull data on global cloud statistics and corresponding radiation budgets (ISCCP, ERBE). However, these data sets still have to be validated. To validate these global data sets on a regional scale, detailed measurements of the cloud cover and structure are necessary. It is important to measure the variabillity of the cloud characteristics within one datapoint of the global dataset. High resolution measurements are also crucial for the development and improvement of parameterization schemes for clouds and cloud-radiation interactions. The aim of the project "Clouds-Radiation-Hydrologic interactions in a limited-area model" is threefold: -
-
Provide a detailed regional dataset on cloud cover and cloud characteristics. Analyse the ISCCP and ERBE datasets with respect to validity and possible applications for climate model verification. Create a model environment for the enhancement of regional data analysis and for the improvement of parameterizations of clouds and radiation.
These items will be described below.
2. H y b r i d cloud detection s y s t e m
A proper description of clouds and cloud-radiation interactions in climate models requires the parametrization of the subgrid processes which determine cloud cover and cloud structure. To develop and improve these parametrizations a set of measurements is required, so results of the model can be validated. At the KNMI a project is ongoing which yields accurate and high resolution cloud measurements over the Netherlands. A hybrid cloud detection system in which both groundbased and satellite remote sensing instruments data are combined has been built. This cloud detection system is described in full detail by Stammes et al. [3]. An outline is given below. The cloud detection system is used to retrieve the following cloud characteristics: cloud cover fraction, cloud top temperature, cloud base height, cloud base temperature, reflectivity and optical thickness; it provides information on the 3-dimensional structure of cloud ensembles. The cloud detection system consists of a network of stations for groundbased remote sensing and a processing environment for AVHRR and METEOSAT measurements. In the
247
120x120 km 2 target area a network of stations for groundbased remote sensing has been built. Each station consists of a LIDAR ceilometer, narrowband IR radiometer and a pyranometer. On two stations more extensive radiation measurements are done. This provides data for the analysis of cloud-radiation interactions. In order to obtain a complete description of the geometry and height of the clouds the signals from the various instruments are correlated. Other available tools for interpretation of the measurements are data on the actual atmospheric conditions from model analysis and radiosonde. Radiative transfer calculations are performed using Lowtran-7 [4] for longwave and DAK (KNMI Doubling Adding radiative transfer model [5]) for shortwave radiation. AVHRR data will be analysed using the Apollo retrieval scheme [6], which has been implemented at the KNMI. The use of state-of-the-art retrieval techniques is garanteed by the international cooperation with other institutes with a tradition in satellite remote sensing, among others: D.L.R. (Oberpfaffenhofen), L.M.D. (Paris), Universit~it Berlin. The following data is archived: a) Satellite instruments: b) 13 Groundstations: c) Radiation stations: d) Other datasources:
NOAA/AVHRR, Meteosat LIDAR, IR-radiometer and solar radiation Direct, diffuse and total downward SW flux, downward and upward LW flux 3-hourly analysis data of the operational HIRLAM (High Resolution Local Area Model) weather forecast model, 6-hourly rawinsonde data in De Bilt
The aim is to obtain continuous data on clouds and radiation for two years.
3. Global datasets
The Earth Radiation Budget Experiment (ERBE) dataset contains invaluable information about the radiation budget at the top of the atmosphere. The International Satellite Cloud Climatology Project (ISCCP) dataset contains information on clouds, their height, temperature, optical thickness, etc.. Both global datasets are derived from satellite measurements and cover many years.
3.1 ERBE During the NRP project the literature on ERBE dataprocessing scheme was studied [7]. The main questions were: Can ERBE data be used for climate model validation and how can the data be used most reliably? Special attention has been
248 given to assumptions about atmospheric and surface conditions, field-of-view homogeneity, spectral variability, time and space variability of cloud cover. It is concluded that the errors in radiative fluxes show much variation related to geographical and atmospheric conditions and the viewing geometry, the positions of the sun and the satellite relative to the area. Nevertheless, monthly mean values data are useful for climate model validation if used properly. Individual values should be used with caution, because the accuracy is highly dependent on the viewing geometry and surface/cloud type observed. It is advisable to users of the ERBE dataset to spend some time on the merits of the data. Likewise it would be good practise if global datasets which are to be used over a wide range of research areas were accompanied by an indication of the errors involved, written down in a way readable by the non-specialist investigator.
3.2 ISCCP In the course of 1993 parts of the ISCCP-cloud detection algorithm were implemented at the KNMI. Numerous case-studies have been performed on the merits of these modules for North-west Europe. This yielded a feel for the sensitivity of the algorithm for atmospheric conditions and possible related errorsources. Furthermore, the literature on the algorithm was studied in order to obtain insight into the coherent set of thresholds making up the detection algorithm. Recently, Rossow [8] published a validation study on the cloud detection algorithm. In general his findings agree well with our own investigations (to appear in KNMI technical report in 1994). The algorithm performance decreases with: increasing variability of the surface albedo and temperature, decreasing contrast between clouds and surface and decreasing temporal variability of the cloud fields. As the ISCCP algorithm is based on spatial and temporal coherence tests this is not surprising. In terms of geographic areas" the algorithm performs well over ocean, somewhat worse over landsurfaces and worst over polar regions. Moreover the algorithm is hampered by areas of persistent cloud fields and storm tracks. Two cloud types are missed frequently: low broken cloud fields and high semi-transparent cirrus clouds. Both cause little contrast in albedo and brightness temperature and are hard to identify especially over cold bright surfaces (polar regions) or highly variable land surfaces. The cloud climatology over the period 1982-1988 shows a global mean annual cloud cover of 63%, which is on the higher end of the range of values which are reported in literature. Rossow [8] states that this value is still an underestimation, because the derived cloud fraction is too low about 10% over land and 10-15% in polar regions. The derived cloud fractions are approximately correct over ocean. The hybrid cloud detection system (paragraph 2.1) will provide a dataset
249 which will be intercompared with the results of the ISCCP algorithm. This comparison will yield information on the merits of the basic assumptions and the used thresholds, and hopefully give directions for improvement of the ISCCP cloud detection algorithm.
4. The model environment
A Regional Atmospheric Climate MOdel (RACMO) has been implemented at KNMI [9]. This is a descendant of the High Resolution Limited Area Model (HIRLAM) and the Global Climate Model developed at the Max Planck Institute in Hamburg (ECHAM). HIRLAM is the operational weather forecast model at the KNMI which employes a 50 km grid and 16 layers. RACMO inherited the dynamics and initialization environment from HIRLAM. All physics modules, including the liquid water and radiation modules, stem from the ECHAM model. There is close cooperation with the research groups involved in the development of HIRLAM and ECHAM. The research environment supports flexible module management, so new versions of cloud and radiation modules can be activated and compared. The Morcrette radiation scheme is included in RACMO [10]. This scheme has been extended to include trace gases and aerosols. In addition to the Morcrette scheme a radiative transfer code for the shortwave spectrum was developed. This KNMI Doubling-Adding radiative transfer model [5] is a line-by-line model which yields so called exact results provided exact information on the atmospheric conditions and constituents is given. DAK was adapted to calculate the radiative properties of a plan-parallel cloud of which the user can define the cloud droplet size distribution and optical thickness.
References
1 Ramanathan V. et al.: Cloud radiative forcing and climate: Results from the earth radiation experiment, Science, 243, 57-63, 1989. 2 Cess R.D. et al.: Interpretation of cloud-climate feedback as produced by 14 atmospheric General Circulation Models, Science, 245, 513-516, 1989 3 Stammes P. et al.: TEBEX observations of clouds and radiation - potential and limitations, KNMI TR- 162, 1994. 4 Kneizys, F.X. et al., Users' guide to Lowtran-7, Air Force Geophysics
250 Laboratory, Hanscom AFB(MA), AFGL-TR-88-0177, 1988. 5 Stammes P.: Influence of clouds on radiative flux profiles and polarization, in "IRS '92: Current problems in atmospheric radiation", Eds. S. Keevallik and O. K~imer, 129-132, A. Deepak, Hampton, VA 1993. 6 Saunders R.W. and K.T. Kriebel: An improved method for detecting clear sky and cloudy radiances from AVHRR data, Int. J. Remote Sensing, 9, 123-150, 1988. 7 Feijt, A.J." The earth radiation budget experiment: overview of dataprocessing and error sources, KNMI TR- 146, 1992. 8 Rossow, W.B. and L.C. Garder: Validation of ISCCP cloud detections, J. Climate, 6, 2370-2393, 1993. 9 Christensen, J.H. and E. van Meijgaard: On the construction of a regional atmospheric climate model, KNMI TR- 147, 1992. 10 Morcrette, J.-J.: Radiation and cloud radiative properties in the European Centre for Medium Range Weather Forcasting system, J. Geophys. Res., 96, 9121-9132, 1991.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
251
The reduction of solar radiation by anthropogenic aerosol in the Netherlands H.M. ten Brink, A. Khlystov, J.P. Veefkind, C. Kruisz# and A. Berner# Unit Fossil Fuels, Netherlands Energy Research Foundation (ECN) # Institut fCir Experimental Physik, Universit,~t Wien
Abstract Nitrate appeared to be dominant aerosol species in the local "direct" aerosol radiative cooling forcing, which was of order -5 W.m 2 on cloudless days. A first approach to the local influence of anthropogenic aerosols on cloud microphysics and radiation transfer in clouds was the measurement, in a large cloud chamber, of the number and composition of the cloud droplets formed in marine air. In anthropogenically "polluted" marine air, five times more cloud nuclei were formed than in clean (arctic) marine air. Approximately half of the anthropogenic aerosol particles with the right size were not soluble in water and did not serve as extra cloud nuclei.
1. INTRODUCTION Aerosol particles in the atmosphere reflect short-wave radiation of the sun and absorb very little infrared terrestrial radiation. Both natural and anthropogenic aerosols are a cooling factor, however, it are the aerosols of anthropogenic origin which exert a forcing, known as the "direct" aerosol effect [1]. This forcing has a local/regional character because of the limited life time of aerosol particles in the atmosphere and thus requires local studies [2]. Aerosols also act in an indirect way as a coolant. They are the nuclei on which the water vapour condenses when clouds form. Anthropogenic aerosol particles serve as extra cloud nuclei in addition to the natural nuclei. Clouds with extra cloud droplets reflect more solar radiation but do not intercept more terrestrial radiation. The increase in reflectance (albedo) of anthropogenically influenced clouds is known as the "indirect" aerosol effect. Its magnitude is very uncertain [ 1,2].
2. DIRECT EFFECT Reflection of solar radiation on anthropogenic aerosol in Europe [1] is estimated to (over)compensate the "warming" by the manmade greenhouse gases. However, these estimates are based on the assumption that sulfate is the sole aerosol component and that the same relation between sulfate and light-reflection, as measured (at ground-level) in the US, was applicable for Europe. The relation between sulfate and light-reflection in West-Europe was investigated. Also the magnitude of the reduction of solar radiation by aerosol on cloudless days was measured and compared to the amount extrapolated from the ground-level measurements, in so-called "closure" experiments.
252 2.1 Experimental In the study automated monitors for the dominant aerosol species in the Netherlands, viz. sulfate and nitrate, were applied. In this way the relation between these parameters and the light-reflection by the aerosol could be determined for the relevant (midday) hours. During a period with cloudless skies the aerosol optical depth (OD) was compared to the OD extrapolated from the light-reflection measured at the ground in integrating nephelometers. LIDAR-data (RIVM) gave the height of the aerosol layer. The growth of the hygroscopic aerosol, caused by the increase in humidity with height, and the resulting increase in scattering were taken from measurements with a laboratory "humidograph" [3] in front of the nephelometer, in which the humidity can be varied. 2.2 Results
Extrapolated and measured aerosol OD were comparable, see fig 1. On the measuring days the concentrations of nitrate were (up to five times) higher than sulfate and thus nitrate dominated the aerosol extinction and the hygroscopic growth of the aerosol. This is in contrast to the current opinion [1] that sulfate is the dominant radiative forcing aerosol-component. Carbonaceous aerosol seemed as important as sulfate. The measured OD's were translated to aerosol radiative forcing, which then was of order -20 W.m 2 at midday. The 24-hour averaged aerosol forcing was -5 W.m -2. Aerosol Optical Depth 0.2
0.1
Figure 1. Comparison between measured and nephelometry derived aerosol optical depth Nov- 17-93
Nov- 18-93
Nov- 19-93
3. INDIRECT EFFECT 3.1 Introduction Sensitivity studies show that the "indirect" effect is most important in marine stratus clouds near polluted continents. The reason is that these clouds have small numbers of droplets because there are very few natural cloud nuclei and thus can be a maximally influenced by small numbers of extra anthropogenic cloud nuclei. 3.2 Aim and approach Actual measurement of the "indirect" effect seems impossible as yet and in a first approach to estimate the maximum local "indirect" effect here, the influence of the local anthropogenic aerosol particles on cloud droplet number has been assessed.
253
3.3 F.xperimental The study was performed in a large cloud chamber. The size of the chamber allows collection of particles and cloud droplets for chemical speciation. The effect of anthropogenic aerosols is determined by comparison of the number of cloud droplets formed in clean and polluted air. The cloud chamber is near the coast of the North Sea in the Netherlands. In arctic marine air the concentration of aerosols is low. Anthropogenic aerosol is present in air flows which have passed over the UK; this is with SW to W winds. Artificially generated aerosols with similar number concentrations served as reference cloud nuclei in experiments in which the cloud formation was tested. Supersaturation used in the study are of order 0.1%, typical for the relevant clouds. Apart from the number of droplets, the size and composition of the aerosol particles which serve as cloud nuclei are measured. In some occasions also the droplets were collected according to size inside the chamber and the chemical composition of the droplets was measured.
Figure 2. Sketch of Cloud Chamber; note size
Number of Particles
Figure 3. Results of test on 16 J u n e 1994
254 3.4 Results and preliminary conclusions The number of cloud droplets in the polluted air is higher by a factor of five compared to that in the clean air. The increase in total particle number is a factor of ten. The difference in the increase in aerosol number as compared to cloud nuclei is explained by the fact that particles with the proper size for cloud nuclei are non-soluble and therefore cannot act as cloud nuclei. Measurement of the actual amount of non-soluble particles, presumably of carbonaceous material, is very difficult because cloud nuclei are small and thus present very little mass.
4. GENERAL CONCLUSIONS AND RECOMMENDATIONS -Additional experiments are required to validate estimates of the local "direct" aerosol effect -Nitrate of is of more importance for the "direct" effect than sulfate in the Netherlands. -Carbonaceous material is a key component both in the direct and indirect aerosol effect.
5. ACKNOWlM.DGEMENT Financing by the department of Economic Affairs of the Netherlands; additional funding: -Indirect effect: National Research Program on Global Air Pollution and Climate Change; project No. 852066 -Direct effect: CEC-Environmental Research Programme/Climatology and Natural Hazards; project No. EV5V-CT92-0171
6. REFERENCES 1 R. Charlson, J. Langner, H. Rodhe, C.B. Levoy and S.G. Warren: "Perturbation of the Northern Hemisphere Radiative Balance by Backscattering from Anthropogenic Sulfate Aerosols"; Tellus 43AB (1991), 152-163. 2 J.E. Penner et al. "Quantifying and Minimizing Uncertainty of Climate Forcing by anthropogenic Aerosols" Bull. Americ. Meteorol. Soc. 7_55(1994), 375. 3 J.P. Veefkind, J. van der Hage and H.M. ten Brink, " Nephelometry Derived versus Measured Aerosol Optical Depth in the Netherlands", Manuscript for Climate Research. Based on "Scattering of Solar Radiation by Aerosol in the Netherlands; Master Thesis"; J.P. Veefkind, June 1994, ECN-R-94--013, Netherlands Energy Research Foundation.
PARTNERS Max Planck Institut for Meteorologie, Hamburg Centro di Studio per I'Applicazione della Informatica in Agricoltura, Firenze University College, Dublin Meteorologisches Institut, Universittit Mhnchen Institute of Astrophysics, T6ravere Tartu, Estonia
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
255
Global modelling of atmospheric trace gases: Application of a global three dimensional model T.H.P.The, D.L.Veenstra, J.P.Beck
RIVM-LLO, P.O.Box 1, 3720 BA Bilthoven, The Netherlands
Abstract The global MOGUNTIA model was applied in two studies: 'Aspects of atmospheric methane' and 'Effects of aircraft emissions'. Results of both studies with respect to the 'European Renaissance' and a 'Joint Implementation' scenarion will be presented. Methane plays a key role in the atmospheric radiative balance and chemistry. Per mass unit it is 58 times more effective than CO 2. In a chemical sense is is the primary sink for OH and it therefore detemines the oxidising capacity of the atmosphere to a large extent. We studied model results from 3 methane scenarios representing a uniform regional growth, a joint implementation option, and the European Renaissance assumptions. Previous studies on effects of aircraft emissions have indicated an increase in the upper tropospheric ozone concentration of 5 - 12% due to aircraft emissions. Since ozone at the upper troposphere level is an effective greenhouse gas, the atmospheric radiative balance may be (further) disturbed by this process. The additional amount of ozone at the cruising altitude is modelled to result in a radiative forcing of 0.03 - 0.07 W/m 2. We assessed the importance of homogeneous chemical processes in aircraft plumes before large scale mixing of the emissions occurs. The exhaust-plume-model developed resulted in a parameterisation of the sub-grid plume effects which was finally implemented in the global MOGUNTIA model. The parameterisation is currently a conversion factor showing a 60% conversion of NO x into NOy when the plume reaches the grid dimensions of MOGUNTIA. Preliminary results with the the global model show that this conversion results in a lower ozone production (6%) than due to unconverted emission fields (8%).
1. Introduction The broad objectives of this project are the production of the 'base-document CH 4' and a document on aircraft and air pollution, both also by Environment Ministry order. In order to make the necessary scenario runs and assessments we put into operation a global three-dimensional model (the MOGUNTIA model) at RIVM. With this MOGUNTIA model the relations between emissions and concentrations of trace gases relevant to atmospheric chemistry on a global scale are studied. In this extended abstract both the work on aircraft emissions and on methane will be described.
256
2. Aspects of atmospheric methane Methane (CH4) plays a key role in the atmospheric chemistry and radiative balance. The actual contribution to the increase of global warming or so called climate forcing cannot be disregarded. It has been estimated that the climate forcing during the past decade, caused by CH 4 together with CFCs and N20 was nearly as large as that for CO 2. Per mass unit, CH 4 is 58 times more effective as a greenhouse gas than CO 2. This figure does not entirely express the contribution to the greenhouse effect due to increased emissions, because the lifetimes are not taken into account. Obviously, long-lived components contribute more than shorter-lived ones. Therefore, another parameter, known as the Global Warming Potential (GWP) was introduced The GWP is defined as the time-integrated commitment to climate forcing of a kilogram of gas relative to a kilogram of CO 2. The GWP values of CH 4 are 35 after a 20years period, and 11 after a 100 years period. Methane is the most abundant trace gas after CO 2 and opposed to the latter gas it is chemically reactive. As a result CH 4 together with CO is, even though it is not extremely reactive, the primary sink for hydroxyl (OH). These OH-radicals play a key role in atmospheric chemistry. Hydroxyl, 03 and H20 2 are the most important oxidants in the atmosphere. Their total atmospheric burden is called the oxidising capacity of the atmosphere. They react with several other trace gases and radicals and control their concentrations. Since OH reacts in large quantities with CH 4, a further major increase in atmospheric CH 4 could lead to a reduction of the atmospheric oxidising capacity. Furthermore, oxidation of CH 4 is an important source of formaldehyde (HCHO). About 810 Tg CO per year comes from the oxidation of CH 4, compared to 1930 Tg/y from other sources Logan et al., 1981. The effects of CH 4 in the atmosphere are not limited to the troposphere but also extend to the stratosphere. About 50 Tg CH 4 per year enters the stratosphere, mostly through the inter-tropical convergence zone (ITCZ). Oxidation of CH 4 in the stratosphere leads to the formation of HC1, acting as a reservoir of C1, considered to be responsible for the destruction of stratospheric ozone. On its turn stratospheric 03 determines the amount of ultraviolet radiation entering the troposphere. At higher altitudes, the greenhouse gas properties of methane are causing an additional cooling of the stratosphere, as methane acts as a radiator. As stratospheric CH 4 dissociates, water vapour is produced. Methane, thus becomes a major source of stratospheric water vapour, since there is little direct transport of water vapour from the tropophere into the stratosphere. We studied three different CH 4 emission scenarios: 1a uniform regional emission growth, in all source categories and regions minut 0.5%/year; 2Joint Implenentation assumptions" A focus on Non_OECD countries, where options for emission reduction are cheapest; all sources and regions differentiated; 3the European Renaissance scenario: reference case; all sources and regions different. Furthermore, we compared our results to the TM2 model (Hein, 1994). The results will be discussed extensively in our final NRP report. Briefly, a comparison of the results of scenarios 1 and 2 shows that CH 4 concentrations are, after 10 years simulation, more than 7% lower in the northern hemisphere in the Joint Implementation case. Ozone is about two percent lower in the global atmosphere and the OH concentrations are about three percent higher.
257
3. Aircraft emissions 3.1 Considerations with respect to aircraft emissions Aircraft emissions affect the chemical composition of the atmosphere in several different ways, depending on flying altitude and location. The conventional subsonic fleet grew extensively in the 1970s and 1980s and is expected to grow by a further 100% during the next two decades. This has led to several model studies focusing on effects of aircraft emissions on ozone in the upper troposphere (Hidalgo and Crutzen, 1977; Isaksen, 1980; Derwent, 1982; Ehhalt et al., 1992; Beck et al., 1992). These studies have predicted an increase in upper tropospheric ozone levels due to emissions of NO• varying between 5 and 12%. The aircraft-induced changing ozone profile could have major implications for the atmospheric radiative balance, since ozone is an effective greenhouse gas at the tropopause level. Model calculations show that a radiative forcing of 0.04 - 0.07 W/m 2 may result from the predicted additional amount of ozone at the cruising altitude (Mohnen et al., 1993; Fortuin et al., 1994). The previous modelled estimates of the ozone increases probably suffer from considerable errors due to uncertainties in the magnitude and distribution of emissions. Furthermore, a number of simplifications were made in the representation of aircraft emissions in atmospheric chemistry models. Firstly, in these studies it was assumed that the present subsonic fleet emits all exhaust gases below the tropopause. However, new estimates predict that a considerable part (25 to 50%) of the emissions from the subsonic fleet are deposited directly into the lower stratosphere. Obviously, the effects of emissions in the lower stratosphere will be different from those in the upper troposphere. Secondly, in the previous work it was assumed that immediate large-scale mixing of all emissions occurs. For this reason, the impact of flight corridor effects with the occurring specific chemistry has not been assessed. Again, in the previous work, the chemical composition of the source gases during large-scale mixing is considered identical to that of the emission at the tail pipe. Model studies of the first 4 seconds of the lifetime of a plume show that not more than 1% of the emitted NO x underwent conversion to HNO 3. However, this does not exclude that larger conversions from NO x to NOy will be found if a longer period of the lifetime of a plume is modelled. Thirdly, another process omitted is the possible involvement of non-linear reactions involving NOy species and heterogeneous chemistry occurring on ice particles in the contrail. The response to emitted NO x is expected to change if heterogeneous chemistry occurs. The NOx---O 3 chemistry is very non-linear and results in a greater ozone production per unit of NO x for lower NO x concentrations (Liu et al., 1987). A part of the emitted NO x had probably already been converted to NOy before the plume reached grid dimensions. Therefore the ozone production caused by the NO x emissions may be overestimated in the older work. Focusing on the considerations mentioned above, in our work we assess the importance of the chemical processes in an aircraft plume in the upper troposphere before large-scale mixing occurs. Therefore an exhaust plume model is developed and the plume model study is expected to result in a parameterisation of the sub-grid, in particular NO x N Oy, chemistry of the plume 9 Using this parameterisation the global NO x emission fields of aircraft can be translated into new processed fields of effective emissions, in which the subgrid effects have been taken into account. These new emission fields were used as input for a MOGUNTIA study on the effects of aircraft emissions at cruising altitudes on the global atmosphere.
258
3.2 The aircraft exhaust plume model The model consists of two sets of differential equations representing the mass balance equations in and outside the exhaust plume. The FACSIMILE package (Curtis and Sweetenham, 1985), which applies a version of Gears' method, has been used to calculate the arithmetic solutions of the chemical equations. The chemistry of the plume is calculated using a set of reactions given by Beck et al. (1992). This set consists of 42 species, 73 gas-phase reactions and 16 photolytic reactions. Background concentrations arc based on the measurements of the STRATOZ III campaign (Drummond et al., 1988). Currently, the set of reactions does not contain any heterogeneous reactions, but it will bc extended with heterogeneous chemistry in a short while. Also, for extension to the lower stratosphere the tropospheric chemistry will bc replaced with a stratospheric reaction set. The diffusion part of the model was based on the theory of Gclinas and Walton (1974), which states that the kinetics and chemistry of an exhaust plume can bc described separately. In this approach, the plume is described by a box of variable volume, which always encompasses the exhaust species. This 'gaussian' approach seems justified since detailed model cxccrcises have shown that interaction of airplane wing vortices with the exhaust trail is of secondary importance in the dispersion of the plume (Louisnard ct al., 1994). The modelled dimensions of an exhaust plume arc about 3 km high and about 100 km wide after 24 hours expansion. The 0 3 concentration in the plume rapidly decreases after the injection of emissions, in particular caused by the reaction of emitted NO with O 3. The O 3 concentration is reestablished to a level similar to the background concentration in about an hour duc to mixing with background air. Subsequently, some production of ozone occurs and the resulting concentration in the plume is somewhat higher than the background level for a few hours. Only minor chemical changes of emissions during the first 4 seconds of the lifetime of a plume - 'the vortex regime' - were found by Louisnard ctal. (1994). However, our work predicts that if the plume is followed for several hours the high NO x concentration occurring immediately after injection decreases rapidly duc to mixing with background :air. A part of the NO x is also converted to NOy components. At first, HONO is formed; this then undergoes rapid photolytical decomposition. Subsequently, HNO 3 is formed and after some hours also HO2NO 2, from which formation was initially inhibited through a lack of HO 2 (caused by the reaction of NO with HO2). The calculated NO and NO 2 concentrations in the plume are quite similar to measured amounts in a young plume (Arnold et ai.,1992). But the measured perturbations of HONO and HNO 3 are 10 to 20 times greater than calculated in the exhaust plume model. A reason for this discrepancy may be our limited set of homogeneous reactions. The exhaust plume model was shown to be a good instrument to study plume chemistry. Generally speaking, maximum perturbations (caused by the emission of an aircraft) of the plume concentrations exist only in the first few hours after injection and disappear after about a day. We found that these perturbations strongly depend on the amount of emission, hour of the day and the season when injection takes place, as well as the background concentrations.
3.3 Translation of aircraft emission fields The parameterisation of sub-grid chemistry was initially in the form of a simple conversion factor: an emission of x molecules of NO x is converted after expansion to griddimension into y molecules of NO x and z moleculcs of NOy (HNO 3, HONO, HO2NO 2, N20 5,
259 PAN), with x = y + z. The transformation of NO x into NO y components depends strongly on the altitude of emission and background concentration (e.g. NOy, OH and HO2) components, and less on the time of injection (time of day, season) and amount of emission. The higher the altitude of aircraft emissions the slower the conversion of NO x into NOy. Here the lower temperatures are the primary reason, followed by the lower OH and HO 2 concentrations. As a first parameterisation, it is assumed that after one day the exhaust plume has reached grid dimensions and the plume model predicts that in the new emission fields the NO x emissions should be differentiated into NO x (40%) and NOy (60%) components, the latter group divided in specific components contributing to NOy.
3.4 M O G U N T I A - m o d e l r u n s ; effects of a i r c r a f t The processed emission fields are used as input for the MOGUNTIA model. MOGUNTIA is a global tropospheric 3D-model incorporating transport and chemistry and it was originally developed at the Max Planck Institute in Mainz (Zimmermann, 1988). It has grid dimensions of 10~ x 10~ x 100 hPa, starting at the surface and extending to the 100 hPa level. The dynamics of the model are calculated with ECMWF fields, assuming turbulent mixing. The reaction set used for the chemistry is from Dentener (1993). The calculations are performed with aircraft emission fields and corresponding anthropogenic NO x emission fields based upon data of McInnes and Walker (1992) and MUller (1992), processed by Olivier (1994). Preliminary results of unprocessed 1990 emission fields show an increase in the ozone concentration at the location of the North-Atlantic flight corridor at the cruising altitude of 8%. The processed emission fields show an increase of about 6% at the similar location.
REFERENCES Arnold, F., J.Scheid, Th.Stilp, H.Schlager, M.E.Reinhardt, 1992. Measurements of jet aircraft emissions at cruise altitude I: the odd-nitrogen gases NO, NO2, HNO2 and HNO3. Geophys. Res. Lett., 12, 2421-2424. Beck, J.P., C.E.Reeves, F.A.A.M.de Leeuw, S.A.Penckett, 1992: The effect of aircraft emissions on tropospheric ozone in the northern hemisphere. Atm. Env., 26A, 17-29. Curtis, A.R., W.P. Sweetenham, 1985: FACSIMILE Release H User's Manual. Report AERE R 11771. Dentener, F., 1993" Heterogeneous chemistry in the troposphere. Ph. D . Universiteit Utrecht.
Thesis,
Derwent, R.G., 1982: Two-dimensional model studies of the impact of aircraft exhaust emissions on tropospheric ozone. Atm. Env., 16, 1997-2007. Drummond, J.W., D.H.Ehhalt, A.Volz, 1988: Measurements of Nitric Oxide Between 0-12 km Altitude and 67 N to 60 S Latitude Obtained During STRATOZ IN. J. Geophys. Res., 93, 15,831-15,849.
260 Ehhalt, D.H., F.Rohrer, A.Wahner, 1992. Sources and distribution of NOx in the upper troposphere at Northern Mid-Latitudes. J. Geophys. Res., 97, 3725-3738. Gelinas, R.J., J.J.Walton, 1974: Dynamic-kinetic evolution of a single plume of interacting species. J. Atm. Sciences, 1807-1813. Hidalgo, H. and P.J.Crutzen, 1977: The tropospheric and stratospheric composition perturbed by NO x y emissions of high-altitude aircraft. J. Geophys. Res., 82, 5833-5866. Isaksen, I.S.A., 1980: The tropospheric ozone budget and possible man made effects. Proc. of the Quadrennial Ozone Symposium. WMO, Boulder, 4-9 August. Liu, S.C., M.Trainer, F.C.Fehsenfeld, D.D.Parrish, E.J.Williams, D.W.Fahey, G.Hiibler, P.C.Murphy, 1987. Ozone production in the rural troposphere and the implications for regional and global ozone distributions. J. Geophys. Res., 92, 4191-4207. Mclnnes, G., and C.T.Walker, 1992: The global distribution of aircraft air pollutant emissions. Warren Spring Laboratory, LR 872 (AP). MUller, J.F., 1992: Geographical distribution and seasonal variation of surface emissions and deposition velocities of atmospheric trace gases. J. Geophys. Res., 97, 3787-3804. Olivier, J.G.J., 1994. RIVM, Personal communication. Zimmermann, P.H., 1988. Moguntia: a handy global tracer model. In: Air Pollution Modelling and its applications VI, edited by H.van Dop, NATO/CCMS, Plenum, New-Y0rk.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
261
Spectral Ultraviolet Radiation M e a s u r e m e n t s and Correlation with A t m o s p h e r i c P a r a m e t e r s F. Kuik 8 and H. Kelder ~ 8Royal Netherlands Meteorological Institute (KNMI), P.O. Box 201, 3730 AE De Bilt, The Netherlands
Abstract At KNMI an experimental infrastructure for ground based monitoring of ultraviolet radiation (UV) and ozone has been realized since the beginning of 1994. In this article the measuring instruments will be discussed, and some preliminary results from our data analysis are shown. Spectral UV-radiative transfer calculations have been compared with measurements, but those results are presented elsewhere [1].
1.
Introduction
The amount of ultraviolet radiation (UV) reaching the Earth surface depends on parameters such as solar elevation, total column ozone and ozone profile, aerosol, clouds, ground albedo, etc. Since many of these parameters change when atmospheric conditions change, ground level UV also is highly variable. However, if long term ozone depletion occurs, a gradual increase of ground level UV may be observed. At this moment UV-measurements provide contradicting evidence for slightly decreasing and increasing UV-levels [2-5]. The 'long term' spectral UV-measurement series are still too short to provide definite answers on the issue of increasing/decreasing UVlevels. Ultraviolet radiation measured at sea level spans a wavelength region from 290 to 400 nm. To obtain a single number representing the energy in a UV-spectrum, the dose rate d D / d t (mWm -2) is introduced as the weighed wavelength integrated UV by [6] dD dt
Jloo
where ~ is the wavelength in nm, I(~) is the spectral UV-irradiance in mWm-2nm -1. A(~) is the CIE-action spectrum (also called McKinlay-Diffey action spectrum [7]), which is a mathematical representation of the sensitivity of 'the human skin' to sunburn. The total dose, D, is obtained by integrating Eq.(1) over the time the exposure takes place. From hereon the dose rate will be called Damaging UV (DUV). DUV is used here to represent amounts of UV for all kinds of processes in which UVB is a dominant factor.
262
250 :
:
:
i
i
i
i
i
i
i
!
:
:
'
;
'
:
'
:
4
200 "#
U
150
o
. J ~ M o d e 1 3 0 0 D U [ _ i -,, ........ ........ !,, ........ ........ [,, ........ ........ [,, ........ ........ [,, ....... ....... ...... i,,...... .... ~ ,.-~ ~. : [ ~"" ]~ Model 350 DU I~ i i ~ ~ ~ ~ ~ ' ~ ~ 9
........ !-i........ ........ ! ........ o
....
0
L,
,~= ....
'
10
15
.
.
.
.
.
.
.
,o, ,~;., . % , ~ , , ,
25 30 35 40 45 50 55 60 Solar elevation (~ Figure 1. All DUV=measurements from January to September 1994, as a function solar elevation. The curve that agrees best with the maximum observed values is a model calculation using total column ozone of 300 D U.
2.
5
Measurements
20
and Calibrations
To study the correlation between ozone and UV, spectral measurements should be performed at least every 0.5 nm. In medical and biological sciences various action spectra are frequently used to describe effects of UV-radiation on different systems. This can also only be done if spectral UV-measurements are available. However, for monitoring purposes it is more convenient to have additional 'simple' instruments that require little maintenance, are (relatively) cheap, and easy to handle in the sense that they do not require highly educated operators. Therefore, both types of instruments are used at KNMI. Measurements performed at KNMI concerning UV-research are: S p e c t r a l U V Spectral UV-monitoring between 285 and 365 nm is performed with Brewer spectrophotometer # 100. The optical part of the instrument consists of a scanning double monochromator. Total
column ozone m e a s u r e m e n t s The same instrument is employed to measure total colu m n ozone with an accuracy between i and 5%. There are at the moment over 100 Brewers spread over the globe for ozone measurements, which are collected by the World Meteorological Organisation (WMO).
Narrow
band U V - m e a s u r e m e n t s Four narrow band instruments are use to monitor UVradiation at 306 n m (UVB, FWHM: 3 nm) and 367 n m (UVA, FWHM: 10 nm). For both the wavelenghts we perform global measurements (irradiance passing through a horzontal plane) and direct sun measurements (flux in the direct solar beam). The instruments for the direct sun measurement are mounted on a solar tracker. radiation a n d m e t e o r o l o g i c a l m e a s u r e m e n t s Broad band global, direct and diffuse solar radiation monitoring is performed at the same radiation measurement setup at the KNMI-rooftop as that for the UV-measurements. Cloud cover observations and all other meteorological parameters measured at KNMI, are used or available for UV-research.
Addltonal
I.rradiance calibrations of (spectral) UV-measurement instruments are very difficult, especially in the UVB-region of the spectrum. Therefore, special requirements have to be met to do reliable and reproducable calibrations. Moreover, the scientific UV-community is still argueing
263
1.2 I> D 1.0 rm -u 0.8
I
I
I
r
!
$
! .............
i .............
i .............
i ..............
:..............
!
i
i
i
i
~
1
2
3
5
6
!
:
(l.)
9 - - -
o.6
E 0.4 o
7
0.2
.............
0.0 0
4
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7
.
.
.
.
.
.
.
.
8
Cloud cover in 8th
Figure 2. Normalized DUV as a function of cloud cover for the period January-September 1994.
about the best procedures, and at the KNMI-calibration laboratory we participate in this research. The main features of the lab, that was fmished after the summer in 1994, are discussed elsewhere [1].
3.
Results
In Fig. 1 all DUV measurements recorded in the period January-September 1994 are shown as a function of solar elevation. These measurements thus include all kinds of possible cloud cover conditions, i.e. from clear skies to completely overcast conditions. From this figure it can be seen that for a certain solar elevation almost every value for the DUV is possible within a certain maximum. In the same graph three curves are shown. These were computed employing an empirical model that uses total column ozone and solar elevation as input [8,9]. For the Netherlands the maximum value for the DUV can be approximated using this model with a total column ozone of 300 DU. D UV measurements recorded at the KNMI with the Brewer during the first eight months of 1994, are depicted in Fig. 2 as a function of cloud cover. The measurements are normalized using maximum values for clear sky conditions that were obtained from the empirical model. Because both solar elevation and ozone are taken into account in this model, the normalized measurements are independent of these parameters. From Fig. 2 it can be seen that for 0/8th cloud cover all ratios are within 0.8 and 1.1. The model does not take into account effects of aerosol, haze, ground albedo, UV-absorption by S02, and all other parameters that may influence ground level UV-irradiance for clear sky conditions. The magnitude of the fluctuations in the normalized DUV thus demonstrates that variations in UV-irradiances up to 20% can be expected for a given solar elevation and total column ozone. The highest values for the DUV do not occur for clear sky conditions, but for situations with 1/Sth to 6/8th cloud cover. This effect is caused by reflections at the side of clouds. When (cumulus) clouds are located near the position of the Sun in the sky, a large amount of almost direct solar radiation can be reflected in the direction of the detector. The amount of observed global UV-radiation can then increase beyond values measured under clear sky conditions [10,11]. This increase in irradiance in the presence of cumulus clouds is also observed in measurements of the narrow band and the short wave (broad band) instruments. In Fig. 3, 10-minute averaged measurements of the global radiation and the UVA and UVB-global instruments are shown for 2 days in the summer of 1994. Between 11:00 UT and 14:00 UT on the partly cloudy day, cloud cover varied between 1/8th and 5/8th. This resulted in an increase of the global radiation of
264 Global radiation (Wm 2)
1000
12L 10
~
UVA-global (Wm 2)
:
!
.
.
.
! .
.
.
.
i .
UVB-global (Wm 2)
030 ::i:::::::i:::: ::::::::::i:ii .
.
.
~ .
.
0.25 0.20
"
i
i
600
6
400
4
0.10
2OO
2
0.05
0
3 5 7 9 11 13 15 17 19 Time (UT)
0
3 5 7 9 11 13 15 17 19 Time (UT)
0.15
0.00
3 5 7 9 11 13 15 17 19 Time (UT)
Figure 3. Broad band global radiation, UVA and UVB global radiation, for June 28, 1994 (solid line). The sharp increases in radiation relative to the clear sky measurements (July 12, 1994) are caused by reflection of solar radiation at the side of cumulus clouds. 38%, and approximately 15% for both the UVA and UVB global measurements, relative to the clear sky measurements.
4.
Acknowledgements
The work described in this report has been supported by NOP (project no. 852088), KNMI, and EU (STEP project 76). Furthermore we thank Wout Slob, Casper Hofman, Piet Starnmes, Wiel Wauben, Jaap Boot, Cor Schuurmans, Andre van London, Rob van Krimpen, Jan Middelkoop, Nanne de Jong, and several people from the KNMI-workshops, who all made important contributions to the work described here.
5.
References
1 2 3 4 5 6
Kuik, F., and Kelder, H., Final Rep. NOP project no. 852088 (1994). Scotto, J., Cotton, G., Urbach, F. Berger, D., and Fears, T., Science 239, 762-764 (1988). Kerr, J.B., and McElroy, C.T., Science 262, 1032-1034 (1994). Watson, R.T., Prather, M.J., and Kurylo, M.J., NASA RP 1208, Washington DC (1988). Blumthaler, M., and Ambach, W., Science 248, 206-208 (1990). Dahlback, A., Henriksen, T., Larsen, S.H., and Stanmes, K., Photochemistry and Photobiology 49, 621-625 (1989). 7 McKinlay, A.F., and Diffey, B.L., in W.F. Passchier and B.F.M. Bosnajakovic (eds.), Human Exposure to Ultraviolet Radiation: Risks and Regulations, Elsevier, pp. 83-87 (1987). 8 Wilson, L., Vall~e, M., Tarasick, D., Kerr, J.B., and Wardle, D., Research Report No. (MSRB/ARQX) 92-004 (1992). 9 Kerr, J.B., McElroy, C.T., Tarasick, D., and Wardle, D.I., in Ozone in the Troposphere and Stratosphere, Part 2, Proc. Quadr. Ozone Symp. Charlottesville 1992, NASA CP-3266, 794-799 (1994). 10 McKenzie, R.L., Matthews, W.A., and Jonston, P.V., Geophys. Res. Lett. 18, 2269-2272
(1991). 11 ~ HI, F.M., and Frederick, J.E., Nature 371,291 (1994).
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
265
Continental ozone issues; monitoring of trace gases, data analysis and modelling of ozone over Europe J.P.Beck a, W.A.J.van Pul a, D.DeMuer b, P.Grennfelt c, P.J.H.Builtjes d, M.G.M.Roemer d, R.Bosmand, P.Esse~, M.E.J.P.Vosbeek e, W.Ruijgroke
aRIVM-LLO, P.O.Box 1, 3720 BA Bilthoven, The Netherlands
bKMI, Ringlaan 3, 1180 Brussels, Belgium
ClVL, P.O.Box 47086, 40258 G6teborg, Sweden
dTNO-MW, P.O.Box 6011, 2600 JA Delft, The Netherlands
eKEMA, P.O.Box 9035, 6812 AR Arnhem, The Netherlands
Abstract
The poster and this paper consist of two parts: (1) Concentrations and internal production of ozone in Europe in relation to different concepts of the critical levels and (2) Exchange of ozone between the atmospheric boundary layer and the free troposphere. The first part describes the evaluation of 1989 ozone data from the EUROTRAC Tropospheric Ozone Research (TOR) project and the Co-operative Programme for Monitoring and Evaluation of Long-range Transmission of Air Pollutants in Europe (EMEP). The data mostly from central and northwestern Europe, show that independent of ways of formulating the critical levels, there are excegAances over large parts of Europe. The second part shows results from a Data Analysis Model that evaluates the transfer of air from the atmospheric boundary layer (ABL) to the free troposphere (FT) and vice versa. This quantity is of importance since it determines the effective emissions of shorter lived species to the global atmosphere. The model was run for the full year of 1989 to study the exchange processes at the location of Uccle (B). For the summer months June, July and Au~gust ~e results releal a net ABL to FT flux of 0 3 with a magnitude of 0.43 * 10-3 mole m" day". This number is in agreement with values derived from transport models.
266
1. Introduction The Dutch part of the EUROTRAC-TOR project consists of contributions from TNO, KEMA and RIVM. The research activities concentrate on measurements at the high quality observatory 'Kollumerwaard', interpretation of the data, hosting the international TOR data base and chemistry and transport modelling. The interpretation activities centre around four basic tasks as they were posed in the NRP-TOR project proposal: 1. How much higher is the mean ozone concentration in the boundary layer over Europe than that averaged over northern mid-latitudes, and what is the seasonal, latitudinal and vertical variation of ozone within the adjacent free troposphere? Is there a secular trend in the concentrations of ozone and precursor molecules in the boundary layer or in the background atmosphere?; 2. What are the emissions and distribution of the precursors responsible for the particular excess of ozone?; 3. Can we measure how much of the excess ozone in the boundary layer spills over into the background atmosphere? Is it possible to quantify by co-measurements of ozone and other tracers the proportion of ozone produced in the troposphere to that transferred from the stratosphere to the troposphere at our location?; 4. How much ozone and how many precursors are transported across regional boundaries? A full assessment of all the questions raised in the four tasks is beyond the scope of this extended abstract. In the following sections we discuss the internal production and the exceedance of critical levels of ozone in Europe and exchange processes between the boundary layer and the free troposphere. A more elaborate evaluation of the above mentioned questions is expected to be given in our final NRP report.
2. The internal production of ozone within Europe The internal production of ozone within Europe was estimated by selecting monitoring stations from the TOR network which were found to be very little affected by local sinks as well by European ozone production and destruction. Emissions of NO x from traffic in the vicinity of a monitoring station is an important sink for ozone, in particular during inversions when high NO concentrations may consume a substantial part of the ozone. Another sink is the deposition to vegetation during nighttime inversions. Both sinks may cause systematic diurnal variation in the concentration of 0 3. In this evaluation the ratio maximum to minimum (O3max / O3min) hourly concentrations from the mean diurnal variation for the summer and winter half year respectively were used. The ratios varied considerably; from 5 to 8 in the areas most influenced by local sinks to 1.04 - 1.40 in the areas with the smallest influence from local sinks. The second criterion was that the stations should very seldom experience ozone episodes. Only four lowland (<500 m a.s.1) monitoring stations were found fulfilling these criteria, these were Jeliiy and Svanvik in Norway, Mace Head in Ireland and Strath Vaich in Scotland. At these stations ozone episodes very seldom occurred. The mean diurnal variation from the four background stations were assumed to represent the background ozone concentration in Europe. The summertime mean concentration at the four background sites was 32 ppb. The internal ozone production was then calculated
267 as the difference between the diurnal curves from the actual monitoring station and the mean for the four background sites. At most sites this approach gives a production of ozone during daytime (mainly between 9:00 and 21:00h) and a destruction of ozone during the night. Since most vegetation effects are caused by stomatal uptake of ozone, it may be more appropriate to look only at daytime differences. Based on these assumptions the internal ozone production in Europe has been calculated to be of the order of 10-30 ppb at remote lowland sites in central Europe. In the south of Scandinavia the internal ozone production is of the order of 5-8 ppb. In the south of the UK, as well as at sites on the continent more influenced by local and mesoscale pollution, several sites show a mean ozone destruction, which at some sites may exceed 10 ppb. Table 1 summarises the analysis on internal ozone production in Europe.
Table 1 Internal ozone production in Europe calculated from the TOR network as the difference in daytime ozone concentration between the actual monitoring station and the mean of the concentrations at four 'boundary layer background' sites. Monitoring area
Number of stations
Internal ozone production (ppb)
Continental Europe; remote sites
6
5-
30
Continental Europe; urban sites
8
-5 - +10
British Isles - south of 54~
9
-13 - +8
British Isles - north of 54 ~ N
6
0 - 3
Scandinavia - south of 62 ~ N
12
5 - 8
Scandinavia - north of 62 ~ N
6
-3 - +3
3. Ozone concentrations in Europe in relation to the critical level concept In order to develop effect-oriented control strategies for regional air pollution the concept of critical loads and levels was developed. The concept was accepted in Europe for the assessment of the effects of acidic deposition and it will probably be the basis for the development of control strategies for ozone as well. It is now based on the exceedance of ozone concentrations over a threshold concentration and it is expressed in ppb.hours units. One important issue is to what extent episodic ozone and background ozone contribute to the exceedance of the proposed 40 ppb base concentration. If episodes make the largest contribution, it seems reasonable to concentrate control strategies to eliminate episodic ozone but if the background ozone makes an important contribution it may be necessary to direct strategies to the free tropospheric background. The importance of episodes for a selected number of TOR sites was evaluated. If we define episodes as hours when the ozone concentration exceeds 60 ppb, episodes contribute to the exceedances by more than 50% at most sites, where the exceedances are above 2000 ppb.hours per growing season. For sites having an exceedance of less than 2000 ppb.hours per growing season, the contribution from episodes is at most sites less than 50% and at remote sites less than 30%. So the central European oxidant problem is mostly an episode
268 problem while episodes play a minor role for the exceedance of the critical levels at the outskirts of Europe. It is obvious that the ways the critical levels are formulated may influence the control strategies for photochemical ozone in Europe. If critical levels are set to an exceedance of 50 ppb, the control of ozone will almost entirely be directed towards central Europe and one may assume that the exceedance may disappear at the outskirts by controlling the central European problem. If, on the other hand, the base level is set to 30 ppb, ozone becomes an all European problem and control strategies may be directed towards all Europe. Since nonepisodic conditions play an important role for the exceedance of the 30 ppb level it may be necessary to direct control measures to sources outside Europe and also to longer lived components like CH4 and CO.
4. Exchange of ozone between the atmospheric boundary layer and the free troposphere The ozone budget in the troposphere is composed of transport from the stratosphere, photochemical production, deposition at the earth's surface and photochemical destruction. The net photochemical production includes ozone formation both in the free troposphere (FT) and in the atmospheric boundary layer (ABL). There is transfer of air between the FT and the ABL; therefore a separate ozone budget of both layers may be drawn up. Despite the fact that the importance of transport processes on the FT and ABL ozone budgets has been known for long, there is still a large quantitative uncertainty about the fluxes involved. Except for transfer of ozone between the ABL and the FT, the transport of VOC and NO x is also of importance because vertical mixing of these components may, due to the nonlinearity in ozone formation, cause more efficient ozone production from the same emissions in the FT than would occur directly in the ABL. The influence of the diurnal cycle in the ABL depth, caused by convective activity, on exchange between the ABL and the FT was studied using a combination of the ground based TOR network, vertical profiles of ozone and meteorological parameters, synoptical information and land-cover data. These 'ingredients' were integrated with parameterisations of atmospheric processes in a data analysis model. The convective exchange processes and their influence on the ozone budget at the TOR station Uccle 03) were evaluated with this data analysis model for every hour covering the full year of 1989. The analysis revealed a flux directed from the ABL to the FT at the Uccle (B) location averaged over June, July and August 1989 of 0.77 x 10-3 mole.m'2.day. The downward FT to ABL flux was 0.33 x 10-3 mole.m'2.day in the same period. So the net flux is directed upward from the ABL to the FT and has a magnitude of 0.43 x 10-3 mole.m2.day. If we multiply this number with the surface of the countries in northwestern Europe (2,442,510 x 106 m 2) it results in a net ABL to FT transport of 1.0 Gmole. day "1. This corresponds to 4 - 6% of the daily cross tropopause flux of ozone in the northern hemisphere in summer. Besides convective growth of the ABL depth, the exchange of air between the ABL and the FF takes place through several meteorological phenomena like clouds, frontal systems, heat island phenomena etc. The influence of these processes is not assessed in the present analysis.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) r Elsevier Science B.V. All rights reserved.
269
K-Gill propeller vane observations for the Cabauw parametrization experiment M. Bottema a & J.W. Verkaik b aEcole Centrale de Nantes, Div. Mecanique Energetique et Environnement, Laboratoire de M~canique des Fluides, 1, Rue de la No~, 44072 Nantes Cedex 03, France. bWageningen Agricultural University, Dept. of Meteorology, Duivendaal 2, NL 6701 AP Wageningen, The Netherlands.
Abstract Some calibration results of the K-Gill propeller vane are discussed. The propellers static and dynamic response were determined. The two propellers show perfect linearity in calibration factor. Starting speed is found less than 0.18 ms -1. Response lengths of the propellers is 2.5 m. Effective response length in K-Gill configuration is 3.3 m. A new function for propeller response to a drop to zero in wind speed is suggested.
1.
Introduction
In the TEBEX mesoscale project K-Gill propeller vanes [1] will be used at the 215 m mast of the Royal Netherlands Meteorological Institute (KNMI) at Cabauw. Measurements of turbulent heat and momentum fluxes over inhomogeneous terrain will be made in all weather situations. TEBEX will provide valuable data for e.g. validation of k-e-models. K-Gills are chosen for a number of reasons: K-Gills can be used to measure fluxes in all weather situations [2], K-Gills have stable calibration [3] and are less expensive than e.g. a sonic anemometer. Major Picture of the K-Gill propeller vane. disadvantages are the large response length (compared to e.g. a sonic) and the rather complicated dynamics of the K-Gill. This summer wind tunnel tests as well as field comparison tests, using the KGill and a sonic anemometer at the 20 m mast at the Haarweg test site in Wageningen, were carried out to measure the K-Gills static and dynamic response. Vane properties have been determined by the method of Wieringa [4]. The vane was
270
too large to do wind tunnel tests so the vane properties will not be discussed now. An extensive report of this project is written by Bottema_[5].
==
Description of the instrument
The K-Gill (manufactured by R.M. Young; model 35301) consists of two propellers, which are oriented 45 ~ upwards and downwards, and a vane. (see picture on the previous page) From thel~)ropeller velocities the horizontal wind (Uh) and the vertical wind can be computed [ . Meanwhile the wind direction is recorded by the vane, so that U h can be decomposed into an west-east and south-north component. Complications arise in recovering the desired along-wind, across-wind and vertical wind components as a results of non-perfect cosine response of the propellers and of the inertia of the propellers and vane. In this rather complex propeller-vane interaction both overspeeding and underspeedino are possible. An extensive analysis of propeller-vane interaction is given by Zhang [q.
3.
Static propeller response
The K-Gill propellers were tested in the wind tunnel of the Group Meteorology of Wageningen Agricultural University (WAU). The test section of this wind tunnel is too small for the K-Gill to fit in it, but the propellers could be tested separately.
3.1
Calibration factor The calibration factor K and the starting speed Us were determined for wind parallel to the propeller axis. The calibration relation is: u=
(1)
K . f . U,
U is the 'real' wind speed (ms -1)'f is the pulse rate (Hz). Pulse rates measurements were made with U from 0.1 ms -1 to 18.5 ms -1. After averaging several runs K and U s were determined from linear regression. The calibration measurements were repeated after two months of duty on the 20 m mast of the Haarweg test site in Wageningen this summer. Both propellers showed excellent agreement and almost perfect linearity and stability in calibration. Differences between propellers were less than 0.4%, the accuracy to which K was determined. The repeated calibrations after the field experiments did not show any significant deviations either. The starting speed however remains uncertain, because the flow blockas correction for propellers was not known. We could only conclude that Us -<0.18 ms-' for both propellers.
3.2
Angular response The output of propeller anemometers for oblique flow with angle t3 is generally less than cos(e)Ule=o. The angular response C(8) is defined by" U~u,, ~ = U,
cos(e) 9 C(0)
(2)
From literature it can be concluded that angular response of propellers of similar shape at least depends on the propeller material and on its dimensions [1'7'8'9].
271
C(0) has been determined with U from 1 ms -1 to 15 ms -1 at angles between the wind and the propellers axis from 0 ~ to 90 ~ As expected, both propellers showed similar behaviour with regard to cosine response. A slight dependence of cosine response on wind speed was probably caused by the fact that for small wind speeds and large 0, the axial wind component approximates the starting speed. The propeller units now used do not discriminate between flows with angle 90+5 ~ and 90-5 ~ which may be important in convective conditions. The best fit to the measured C(e) was: C{8) = 1 - 0.3sin2(8) + 0.02sin(88)
(3)
This formula is accurate within 0.005, except for 15 ~ and 60 ~ where the fit is within +0.02 of average experimental data.
4.
Dynamic
response
Generally, a propeller is assumed to be first-order system which satisfies the following equation dU(t)
= I(U " _ U(t))
(4)
dt where U~o is the full wind tunnel speed and "~ a time constant. Usually x is written as ~;
= 0 1 u.
(5)
where D is the distance constant. Solution to eq. 4 is a exponential function
u{t) = (Uo - u=)~-,I, + u=
(6)
where U 0 is the wind speed at t = 0 s.
4.1
Wind speed increase and angular dependence of D
D can be determined from the length of the air column after 63% speed up of the propeller (D63%) or from linear regression (DLR). In all our observations, there was a trend for D63 % to be larger than DLR (largest overestimation of 20%). One cause is the time discretization. With x = 0.63 s and At = 0.1 s the average overestimation is 8%. The remaining 12% must be caused by initial friction. We used linear regression when calculating D. All regressions were taken at the interval from 20% to 70% adaption. For flow parallel with the propellers axis, we found for both propellers D = 2.5 + 0.2 mo Repeated measurements after the field experiment did not show any significant differences. A propeller responds slower when it is inclined to the mean flow. The angular dependence we observed corresponds well with the formula D(8) = D(0~ Effective D in K-Gill configuration corresponds to the measured D at 45 ~ 3.3 m.
(7)
272
1
[]
measured Uo exp(-t/tau)
Uo/( l + t/T) E E3 v
9
" .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
0.01 -1.0
.
.
.
.
.
.
,
.
.
.
.
.
.
.
.
.
.
.
,
1.0
.
.
.
.
.
.
.
.
.
.
.
"- ..... 3.0
.
.
.
.
.
.
.
.
.
.
.
.
.
.
;
5.0
time (s)
Figure 1 Propeller response to a drop to zero in wind speed. 4.2
Wind
speed
decrease
Propellers deceleration is usually described by eq. 4. In case Uoo = 0 ms -1, ~: can not be defined as in eq. 5. Now let us assume the propellers deceleration is proportional to the drag forces on the propellers, so
d~Xt) d~
_ _I_ ( U = - U(t)) 2 ~.
(8)
where X is some length scale. Instead of an exponential function, eq. 6 yields a hyperbola: U(O = U~ - U = + U. I +tiT
(9)
where T is a new time constant: T = ~l(Uo -
U=)
(~o)
Measurements were carried out in a wind tunnel with fixed wind speed at Uoo = 0, 1, 2 and 3 ms -1. The propeller was speeded up like a top, using a cotton wire. This way Uo's were reached from 6.5 to 9.0 ms -1. The best type of function was determined from correlations coefficients of the regression.
273 -'
~
'
,
[]
measured
Uo oxp(-t/tau) Uo/( l + t/T) v
E
v
o, ,,,,,,,,,,,,,,,,.~,,6
i
9 .0
'"
i
1.0
i
i
3.0 time (s)
i
i
5.0
7.0
Figure 2 Propeller response to a drop to 1.2 ms -1 in wind speed. Some results are summarized in table 1. The values of D, ~:, ;L and T are averaged over several runs and the average and standard deviation for several values of U~ are printed. In the bottom row the ratio of the of the standard deviation and the mean values for D, ~:, X and T for all U , ' s are printed. Clearly D is not constant but increases with U . From the bottom row it can be seen that in this wind speed range the relative differences in D are larger than those in ~:, so rather ~: than D should be called a constant. Table 1 D, ~:, X and T for several Uoo's Uoo (ms 1)
D = xUoo (m)
I: (s)
X (m)
T (s)
.0
.0 + .0
.84 + .04
3.17 + .07
.37 + .01
1.2
.92 + .05
.76 + .04
2.2 + .2
.35 + .03
2.0
1.32 + .08
.65 + .04
1.6 + .1
.29 + .02
3.0
1.59 + .04
.54 + .02
1.27 + .07
.23 + .02
std./avg.
.22
.17
.35
.17
274
The ratio T/'c is 0.44 for all Uoo's. This is a result from the chosen regression interval. From the correlation coefficients of regression it could be concluded that equation 9 provides a much better fit than the exponential function only in case U = 0 ms -1. (fig. 1 and 2) For small Uoo the exponential function is equally well or better. The value for ;L found this way decrease from 3.17 m at Uoo = 0 ms -1 to 1.27 m at U= = 3.0 ms -1.
5.
References S.S. Atakt0rk & K.B. Katsaros, 1989, The K-Gill: A twin propeller-vane anemometer for measurements of atmospheric turbulence, J. of Atmospheric and Oceanic Techn. 6, p. 509-515. K.B. Katsaros, M.A. Donelan, W.M. Drennan, 1993, Flux measurements from a SWATH-ship in SWADE, J. of Marine Systems 4, p. 117-132. N.E. Busch, O. Christensen, L. Lading, S.E. Larsen, Cups, vanes, propellers and laser anemometers, Air Sea Interaction: Instruments and methods (eds. F. Dobson, L. Hasse, R. Davids, Plenum Press, NY, 1980, 11-46. J. Wieringa, 1967, Evaluation and design of wind vanes, J. of Applied Meteorol. 6, p. 1114-1122. M. Bottema, 1994, Evaluation of the K-Gill propeller vane, Report 94, Dept. of Meteorol. Wageningen Agricultural University, The Netherlands. S.F. Zhang, A critical evaluation of the von Karman constant from a new atmospheric surface layer experiment, PhD thesis, University of Washington, Dept. Atm. Sciences, AK-40, Seattle, Washington 98195, USA, 1988, 133 p. T.W. Horst, 1973, Corrections for the response error in a three-component propeller anemometer, J. of Applied Meteorol. 12, p. 716-725. R. Drinkrow, 1972, A solution to the paired Gill-anemometer response function, J. of Applied Meteorol. 11, p. 76-80. B.B. Hicks, 1972, Propeller anemometers as sensors of atmospheric turbulence, Boundary-layer Meteorol. 3, p. 214-228.
This project is part of the Dutch National Research Programme on Global Air Pollution and Climate Change.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
275
Empirical Orthogonal Function (EOF) Analysis of Ozone Variability F.J.M. Alkemade National Institute of Public Health and Environmental Protection (RIVM) P.O.Box 1, 3720 BA Bilthoven, The Netherlands
Abstract Empirical orthogonal function (EOF) analysis was applied to the monthly averages of the TOMS ozone data set from November 1978 to October 1993 for two latitude ranges. The EOF method yields a set of orthogonal (spatial) functions (or base-patterns) wich are, unlike in most other decomposition methods (where the set is usually predefined), derived from the data itself. This is done in such a way that the EOFs, ordered by the amount of explained variance, form the most efficient basis for the variance decomposition. In a typical case, using 15 years of monthly averaged ozone observations, rebinned to gridcels of 15~ longitude x 4 ~ latitude, up to 83 % of the total observed variance between latitudes 66~ and 66~ could be accounted for by only the first 3 EOFs. Taking into account only the tropical regions (between 30~ and 30~ the first 3 patterns (associated with the seasonal variation and the QBO-cycle) could explain 90 % of the total variance. The fact that a small number of coefficients is sufficient to characterize the monthly averaged ozone-fields reflects the existence of only a few dominant long-term patterns. Also the long term ozone depletion rate was analysed in terms of changes in the coefficients of the base-patterns. 1. I n t r o d u c t i o n During the last twenty years concern has been growing about the influence of human activity on the global climate. As ozone is one of the key constituents in the stratosphere, related directly to UV-radiation and the atmospheric temperature-profile, it has been monitored extensively by ground-based and orbiting instruments. Since 1978 the Total Ozone Mapping Spectrometer (TOMS) instruments on the Nimbus 7 and Meteor 3 satellites have produced daily measurements of the ozone column with a high spatial resolution. It has been shown that the resulting global ozone time series are mainly influenced by long-term (anthropogenic) decrease in ozone, the ll-year solar cycle, the annual seasonal cycles, the quasi-biennial oscillation (QBO), and the E1 Nino/Southern Oscillation (ENSO). In order to separate these effects usually a multiple linear regression method is used, for which a priori information is needed about the frequency and phase of the cycles of which the c o e f f i c i e n t s a r e t o be found. (Herman et al.,1993 [1], de W i n t e r - S o r k i n a , 1994 [2]) By applying an EOF decomposition method instead of linear regression, no assumptions need to be made in advance, as the only requirement for the construction of each successive base-pattern is that it explain the maximum possible proportion of the variance in the data not yet accounted for. The cost is the fact that, although seasonal and QBO patterns can be easily detected, a clear identification of the less significant EOFs is not always possible. 2. E O F d e c o m p o s i t i o n
technique
The method of Empirical Orthogonal Function analysis, also called Principal Component analysis, is used to study the temporal variability of data over a large area. It was first introduced by Lorenz (1956), and since then has mostly been used in oceanographic and meteorological research.
276 The EOFs described here are a set of orthogonal spatial patterns, which indicate the dominant deviations from the mean ozone-field. They are derived from the eigenvectors of the data covariance matrix R: if vm, indicates the difference of the observed ozonecolumn from its long-term mean at place xm and time t , (m = 1...M, n = 1...N), then R is defined as R ,1,., . ~ -- - ~a2 .v-.N -, ,, .~n=lVan 9 Vbn It can be shown that for each EOF the corresponding eigenvalue is a direct measure of the explained fraction of the total amount of variance in the data. In the following sections the EOFs will be sorted by this criterion, index 1 indicating the most significant EOF. After finding the EOFs, the observed ozone-fields can be rewritten as linear combinations of these EOFs, by calculating each EOF's contribution to the total field at any time. The resulting coefficients represent the time dependence of the EOFs. A comprehensive description of the EOF technique is given in Preisendorfer et al., 1981 [3].
3. T h e o z o n e - f i e l d f r o m 66~
to 6 6 ~
The EOF-analysis of the monthly averages of the TOMS ozone data set from November 1978 to October 1993, gridded at a resolution of 150 longitude x 4 ~ latitude, yields a very strong first component, explaining 60.1% of the total variability. This pattern can clearly be identified to the main seasonal variation (at midlatitudes typically in the order of some 100 Dobson Units) by its characteristic North/South asymmetry (fig. 1), and by the cyclic nature of its coefficient (fig. 2). At higher latitudes the longitudinal variation strongly suggests the influence of planetary waves, wich are known to affect the distribution of ozone (Garcia and Hartmann, 1980 [4]). The coefficients of the second and third EOF, explaining 13.4 % and 8.9 % of the total variance respectively, also clearly show a yearly periodicity (EOF 2 more clearly than EOF 3), but with smaller amplitudes, and some 90~ out of phase (in quadrature) with regard to the first EOF. These second and third basepatterns, with the same frequency as (but independent of) the first one, can be interpreted as phase-corrections to the main seasonal variation: The combination of the first three EOFs not only gives the amplitude but also the phase of the seasonal ozone-cycle for each gridcel, explaining 82.4 % of the total variance. The remaining EOFs have much less significance (e.g. EOF 4 explains 2.6 % of the variance, EOF 10 only 0.5 %), and their association with some specific physical process is not evident.
F,oF 1, 66~
Coefficients 166~ to 66~ E0F 1 (solid) and 2 (dashed)
to 66~ .
0.15
10
b 7
5
"~
0
o
-10 1983
..................................................... 1984
1985 Time
1986
1987
(year)
figures 1 and 2, showing EOF 1, and the timedependence for EOF's 1 and 2
1988
277
4. T h e o z o n e - f i e l d f r o m 30~
to 30~
Using the same spatial resolution and time-frame as before, the EOF-analysis of only the tropical regions results in three dominant EOFs, explaining 90.0 % of the total variance. The first two, both with a very strong year-periodicity, but in quadrature with one another, again describe the amplitude and phase of the seasonal variation. As was expected this signal is somewhat weaker then in the case where higher latitudes were included, reflecting the fact that in the tropics the ozone-layer is less influenced by the seasons. The third EOF (fig. 3 ) c a n clearly be associated with the QBO cycle: it is very symmetric with respect to the equator, and its coefficient shows a periodicity of somewhat over two years, closely matching the Singapore (2~ 30 mbar zonal wind values (fig. 4). Coeff. for E0F 3 (300S to 30~ (solid) and Singapore Winds (dashe,t) T
:30"S t o ~O~
~OlV 3,
2
0. I o
0. ~
.......................
10 9
E
0.10
o
.
~
1
0.5
~
o
o.o "~
o.OO - 1
-0.5
o
"
.~o.O 5
c
...0.10
~.~.
.~-~ ~
.r
~-~ ~ . :~>~.~
figures 3 and 4, showing EOF 3 and its time-depence compared with tropical zonM winds. 5. T r e n d a n a l y s i s The TOMS-data show a global long-term ozone depletion of about 0.3 % per year. This trend, however, has a strong geographical dependence. In the tropics little or no change has been detected during the last fifteen years, whereas at higher latitudes areas can be found where a very significant depletion has occurred. The EOF-decomposition can model this ozone-depletion pattern by calculating the linear trend of each of the EOFcoefficients. Multiplication by this trend transforms each EOF into a pattern of derivatives. By combining only the first three EOFs the ozone depletion map can be approximated quite well (fig. 5), showing again that a few coefficients (the linear trends for EOF 1 to 3) are sufficient to characterize a major part of the large scale ozone variability. 6. C o n c l u s i o n s These first applications of the EOF-technique to analyse the variability of observed atmospheric ozone, allow us to conclude that the decomposition in empirically found basepatterns as described above, is highly efficient, and that the most significant EOFs appear to be physically meaningfull. This suggests that at a large scale (in both space and time) the ozone sytem has only a limited number of degrees of freedom, a conclusion which would be in agreement with recent results of P. Yang et al. (1994, [5]), who applied nonlinear dynamical theory to the behavior of the ozone-layer and found evidence of low-dimensional attractors.
278 It is expected that, apart from producing efficient parametrizations, further EOF-analysis of the TOMS-data for different geographical ranges, different time-frames, and different spatial and temporal resolutions will also give some understanding of how the complex chemical and physical processes that produce, distribute, and destruct the ozone, are related to the basic global patterns of variability. This might help to identify those variables and equations that are essential in the simulation of long-term stratospheric ozone behavior. .......:.~9.....~.... fig. 5a, Ozone depletion trend ....... i;.;.::":.::.a-: 0.50 nov. 1978 to oct. 1 9 9 4 ~ .from the TOMS data -1.00 o o
::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: ::i:i:i ~i:~:~i~:~:~:i:~i:i:i:~:~:~:~:i:i:~:~:~:~:i:i:i:i:i:i:~:i:~:.~ !:!".:i:i:!:
-2.50 p..
,::~:i:~:~:!:~:~:!:i:!:~:~:~:~:~:~:i:~:~:~ ...........................~:::::::~::*~:
Wi!i!ii!ili!iiiiii!i!#!iiiii::
::i~':: -4.00
,.<
....:':'::~!~ii!~i~i~:~:~i~i~iN:.::~:~:~:~:i:;'.:~:~:~:~:i
-5.50
Ozone
t~g. 5b, Combined trend patterns nov. 1978 to oct. 1994 from EOFs 1 to 3,
Depletion
NASA/RIVM
.........._.....~........... -". ............ i;.;.7~.h:a"ii....
0.50
o
-1.oo ......
c
-2.50
r
iiiii~,~:~]NiiNiiNiiiiiii,: ~!~!~."..':;.'~i~i~!!i~i~i~i~!~!~!~iii#.~i~i~i~:.
References
-4.00
1:2
-5.50
~-"
..::~'~'N"............ NNiii ~:~
lon,lat
grid:
24
9
45
NASA/RIVM
[1] J.R. Herman and R. McPeters, Ozone Depletion at Northern and Southern Latitudes: J. Geoph. Res., Vol. 98, no. D7, pages 12,783-12,793 July 20 1993 [2] R. de Winter-Sorkina, Total Ozone Trend Analysis from the TOMS Data, Proceedings of Climate Change Research, Maastricht, The Netherlands, 1994 (publ. Elsevier Science). [3] R. W. Preisendorfer, F. W. Zwiers, and T. P. Barnett: Foundations of principM component selection rules. SIO Re/. Series 81-4. Scripps Inst. of Oceanography, 192 pp. [4] R. R. Garcia and D. L. Hartmann, The role of planetary waves in the maintenance of the zonally averaged ozone distribution of the upper stratosphere. J. Atmos. Sci. 37 1980) 2248 - 2264. ] P. Yang, G. P. Brasseur, J. C. Gille, S. Madronich, Dimensionalities of ozone attractors and their global distribution. Physica D 76 (1994) 331 - 343
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
279
Total Ozone Trend Analysis from the TOMS Data R. de Winter-Sorkina Air Research Laboratory, RIVM, P.O. Box 1, 3720 BA Bilthoven, The Netherlands
Abstract
A multiple linear regression model was applied to describe the monthly mean Nimbus-7 and Meteor-3 TOMS column ozone data in the period November 1978 to June 1994. Year-round and monthly mean ozone trends, QBO coefficients and solar cycle coefficients were calculated with a high spatial resolution for all grid cells (1 ~ latitude by 1.25 ~ longitude) of the TOMS column ozone data, as well as for zonal means over the 10~ latitude bands. After the eruption of Pinatubo, ozone trends became more negative (about 1% per decade) in the Tropics and the northern hemisphere.
1. Introduction The depletion of the ozone layer has a direct influence on climate due to changes in temperature, circulation and UV -radiation. Since the early 1970s there has been an increasing interest in the assessment of atmospheric ozone data for changes related to human activity. The satellite-based instruments have the advantage of full spatial coverage. The longest continuous global satellite ozone measurements have been obtained by the TOMS instrument. TOMS ozone trends were derived for zonal mean ozone over the 10~ latitude bands for November 1978 through May 1990 [1] and until the end of the 1991 [2]. In [3] the TOMS ozone trends were calculated on a monthly basis in the 10~ latitude x 10~ longitude blocks for the time period November 1978 to May 1990.
2. TOMS ozone dataset and model description TOMS is a satellite-based Total Ozone Mapping Spectrometer. TOMS measures the UV sunlight, backscattered from the atmosphere at six wavelengths between 312.5 and 380 nm on a daily basis over the entire world. The sunlight in the UV is absorbed by ozone molecules in the atmosphere. Total ozone columns are calculated by fitting the ratios of the radiances of the wavelength pairs (where one wavelength is strongly absorbed by ozone while the other is only weakly absorbed) with the ratios calculated by the radiation transport model. Ozone is measured with high spatial resolution; individual measurements have been averaged into grid cells of 1~ latitude x 1.25 ~ longitude. The TOMS instrument was active on board the Nimbus-7 satellite from November 1978 until May 7 1993. Version 6 of the TOMS data set is corrected for the diffuser plate degradation and is estimated to be precise to +1.3% (2~) at the end of 1989 relative to the beginning of the record. A second TOMS instrument was
280 launched on-board the Meteor-3 spacecraft in 1991. We used the ozone data after April 1993 from the Meteor-3/TOMS. Total ozone has a strong annual variation. On interannual time scales ozone varies in response to a number of competing factors. These factors include long-term changes in the atmospheric content of anthropogenic trace gases such as the chlorofluorocarbons (CFCs), quasi-biennial oscillations (QBO) of the zonal winds in the Tropics, 11 years solar cycle, volcanically ejected trace gases and aerosols, the E1 Nifio - Southern Oscillation (ENSO) and the solar proton events. The last two turn out to be non-significant for the determination of the long-term trends. The temporal dependency of the ozone column can be described by the following multiple linear regression model: 03(0 = ~t(i) + ~ ' t + I3Qao'XqBo(t-L) + 13stm-Xstm(t) + N(t) N(t) = ~-N(t- 1) + e(0 where 03 is the monthly mean ozone, t the time in months, Ix(i) a seasonal term for the ith month of the year (i=1,2 ..... 12); XQBo(t) is the Singapore (2 ~ N) 30 mbar zonal wind representing the tropical QBO, L the optimum phase lag between XQBo(t) and 03(0, Xstm the solar 10.7 cm flux used as a proxy for the ultraviolet solar irradiance; N(t) is a residual noise term, modelled as a first-order autoregressive process because there is a month-to-month correlation in ozone values; e(t) is a residual uncorrelated series with the mean equal to zero, although the variance of e(t) depends on the month [oa(t)=o'z(t-12)]; ~ , [~Q~ and 13s~ are, respectively, the linear trend, QBO and solar cycle coefficients to be determined by least squares regression. The monthly mean ozone values are used because of their relative independence, the daily data strongly correlated within the month. To determine the monthly coefficients and seasonal term, 12 monthly models were fast solved. Then the year-round model was fitted by a weighted estimation procedure (to account for different variability of ozone in different months), with weights for each month inversely proportional to the estimated monthly variances of the residual noise. In both cases, the initial application of the model yielded the estimation for N(t-1); the coefficients were determined in the second application of the model. Calculations were made with a high spatial resolution for all grid ceils (1 ~ latitude x 1.25 ~ longitude) of the TOMS column ozone data, as well as for zonal means over the 10~ latitude bands. The Singapore 30-mbar wind values were lagged to produce maximum positive correlation with the ozone time series in each latitude band.
3. Results and discussion Zonal means of the TOMS monthly mean ozone over the 10~ latitude bands were fit with the model described above. For the latitude bands centred at 75~ 85~ 75~ and 85~ there are some missing values as the satellite could not observe ozone over the polar region at night. Zonal means were calculated only in the latitudes 70~176 For the zonal mean ozone in this region, year-round and monthly ozone trends, QBO coefficients and solar cycle coefficients, together with their standard errors, were calculated. Ozone trends were calculated separately for the period before the eruption of the Pinatubo volcano in June 1991. The results are in good agreement with the previous calculations [2]. Calculations were then repeated for
281 the time period through June 1994. The ozone trend is not significant (at the 2t~ error level) in the Tropics from 300S to 30~ before Pinatubo and from 10~ to 10~ for the update through June 1994. The trend becomes negative and its absolute value increases as we move along the latitude towards both poles, faster in the southern hemisphere. The ozone trend has a strong seasonal variation depending on latitude. In Table 1 the year-round trends, the minimum and maximum monthly trends are listed for the latitude bands centred at 55~ (the Netherlands is situated in this area) and 650S for the time period before the eruption of Pinatubo and for an update through June 1994. After the eruption of Pinatubo ozone trends became more negative (about 1% per decade) in the Tropics and the northern hemisphere. The rise in ozone depletion is due, at least in part, to the cumulative effect of heterogeneous chemistry occurring on enhanced aerosol surfaces in the Pinatubo cloud; radiative and dynamical perturbations could induce reductions in the Tropics [4]. Table 1 Total ozone trend estimates for the latitude bands centred at 550N and 65~ (% per decade) Year-round trend 05/1991 06/1994
Minimum monthly trend 05/1991 06/1994
Maximum monthly trend 05/1991 06/1994
55~
-3.5+1.4
-4.4+1.4
65~
-7.2+1.8
-6.8+1.6
-6.0-2_3.1 February -19.8+4.9 October
-2.3+1.8 August -4.6+2.4 February
-8. +2.~.4 February -20.8+4.6 October
-2.9-&1.5 August -4.8+1.9 February
Year-round and monthly mean ozone trends, QBO coefficients and solar cycle coefficients were calculated with a high spatial resolution for all grid cells (1 ~ latitude x 1.25 ~ longitude) of the TOMS column ozone data. Figure 1 (upper part) shows the year-round total ozone trends worldwide. In addition to latitude and seasonal dependency, it is possible to see the longitudinal dependency of ozone depletion and its geographical location. Strong ozone depletion occurs in the high latitudes of the southern hemisphere and near the south pole, the area corresponding to the Antarctic ozone hole appearing from September to November. The monthly trends for this period show strong ozone depletion (more than 10% per decade) almost the whole area south of 60~ The ozone depletion reaches its maximum in October, when trends in the polar region are about 30% per decade and those above the most southern part of South America and the Falkland Islands are about 10% per decade. The high negative trend area moves northwest in November. Year-round ozone trends are significantly negative in the northern hemisphere above 30~ reaching absolute values of up to 6% per decade. The maximum negative year-round trends in this area apply to above the UK, south of Norway and Sweden and some northern parts of Russia. An example of the monthly ozone trends for March is shown in Figure 1 (lower part). Here, we see a fairly strong longitudinal dependency of ozone trends in the northern hemisphere. The areas above the UK, northern France, Benelux, northern part of Scandinavia and large parts of Russia have large absolute trend values equal to about 10% per decade and more. At the same time there is a region between the coasts of Canada and Greenland at about the same latitude where a trend is almost nonsignificant. Finally, we report that the zonal mean ozone seasonal term described above was compared
282 with the ozone values calculated with the 2D Cambridge/RIVM stratosphere model. A good overall agreement was found. 4. References
1 R.S. Stolarski, P. Bloomfield, R.D. McPeters, J.R. Herman, Geophys. Res. Lett., 18 (1991) 1015. 2 L.L. Hood, J.P. McCormack, Geophys. Res. Lett., 19 (1992) 2309. 3 X. Niu, J.E. Frederick, M.L. Stein, G.C. Tiao, J. Geophys. Res., 97 (1992) 4661. 4 J.M. Rodriguez, M.K.W. Ko, N.D. Sze, C.W. Heisey, G.K.Yue, M.P. McCormick, Geophys. Res. Lett., 21 (1994) 209.
0.4
-0.1 C~
-0.6
~
o~ -1.0
-1.5 ~-~ ...........
zone Trends 0.3
- 0.3
:~:~:~:~:~:~:!:~:~:~:
:-'!:!i~k'~:~:~:!;~:~-'!:~:i' :!'.!!:i:!' '!-'! i:i~ -- 0.8
iiiiiiiiiiii!iiiiiiiiiiiiii!!i!iiiii!iii ,,,....
i~ii
iiiiiiijl
iiiil~' /
-1.a x
-1.8 ~-~ Ozone Trends
for March
Figure 1. The year-round (upper part) and the monthly ozone trends for March (lower part).
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
283
ASSESSMENT REPORT ON NRP SUBTHEME " O C E A N S A N D T H E C L I M A T E SYSTE1W'
L. Otto Netherlands Institute for Sea Research P.O. Box 59 1790 AB Den Burg The Netherlands
With contributions by: H.A. Dijkstra, W.P.M. De Ruijter, C.B. Vreugdenhil
R.J. Haarsma, A. Kattenberg, G.J. Komen, W.A. Oost, A. Sterl H.J.W. de Baar, W. Helder P. Westbroek
IMAU/RUU, Institute for Marine and Atmospheric research/ University of Utrecht KNMI, Royal Netherlands Meteorological Institute, De Bilt NIOZ, Netherlands Institute for Sea Research, Texel RUL, University of Leiden
284 Contents Abstract 0
0
0
4.
Oceans and climate 1.1 Introduction 1.2 Ocean physics 1.3 Biogeochemistry Ocean physics and climate 2.1 Introduction 2.2 Models 2.3 Ocean observations 2.4 Ocean-atmosphere interaction B i o g e o c h e m i s t r y of t h e o c e a n s 3.1 Introduction 3.2 Chemical exchange ocean-atmosphere 3.3 Greenhouse gases and DMS 3.4 The phytoplankton system Conclusions
ABSTRACT This report reviews our present ideas on the physical and biogeochemical role of the oceans in the climate system and gives an a s s e s s m e n t of the r e l e v a n t research in the framework of the NRP and of the related N e t h e r l a n d s research activities. 1.
O C E A N S AND C L I M A T E
1.1 I n t r o d u c t i o n The importance of the oceans for the global climate lies both in their role in the t r a n s p o r t and accumulation of heat and moisture and in the influence of their b i o g e o c h e m i s t r y on the composition of the a t m o s p h e r e . W i t h o u t a n y a n t h r o p o g e n i c influence the feedbacks in the oceanic s y s t e m can cause instabilities and "natural" climate variations such as have been deduced from palaeoclimatic evidence. This is reason enough for investigating this system. The a c t u a l concern for the a n t h r o p o g e n i c influences, however, m a k e s t h e s e investigations even more urgent. This task is of a scope t h a t requires concerted international action in the field of ocean observations and ocean modelling, both in physics and in biogeochemistry. The Netherlands effort in those fields, the subject of this review, is to be judged against this background. The oceans are playing an important role in the climate system, by their physical role in the transport and accumulation of heat, but in addition as the environment
285 for a n u m b e r of biogeochemical processes t h a t are acting in the exchange of important greenhouse gases and condensation nuclei (DMS) with the atmosphere. A review of the present opinions with respect to this role of the oceans and a checklist of research needs, composed by an International Meeting of Scientific and Technical Experts, was published by UNESCO (Anon. 1991). The oceans, in combination with the atmosphere, are transporting the excess of heat received by the earth in the tropical regions to higher latitudes t h a t have a heat deficit. This role can be shown by budget calculations comparing the zonal radiation budget with the meridional transport by the atmosphere, for the direct e s t i m a t e s of the oceanic t r a n s p o r t are still less accurate. This meridional t r a n s p o r t accounts for about half the t r a n s p o r t required for a stable climate (depending on latitude). A recent estimate (Trenberth and Soloman, 1993) is given in the WOCE Science Plan for Ocean Modelling (WOCE International Project Office, 1994). Although different direct measurements of the oceanic transport still are showing large discrepancies, the finding t h a t the oceanic contribution is important is, in order of magnitude, confirmed by these estimates. However, we need b e t t e r observations and better models to compute the t r a n s p o r t with an accuracy comparable with the atmospheric estimates. It is clear that the oceans m u s t have a significant effect on the distribution of heat over the globe and thus on the global and on the regional climate. For western Europe especially the effect of the oceans on the regional conditions should be noted. Especially for the North Atlantic area the use of coupled ocean-atmosphere models appears to be essential in predicting a possible regional climatic change (Gates et al., 1992). The rate of climate change under the influence of an increased greenhouse effect is determined by the heat storage of the ocean. This again depends on the vertical exchange processes in the sea. An increase of the downward radiative flux of one or two watts, an a m o u n t corresponding with a doubling of the atmospheric CO2, reaches a new stationary state after one year if only the upper 10 m of the ocean would be involved. However, this is 10 years for a 100 m thick layer, and if the full depth of the ocean is involved, the equilibrium is only reached after centuries. As the ocean is a strongly layered medium with different vertical exchange processes, we have to do with different short-term and long-term "climatic memories". In the IPCC reports estimates are given for the exchange of greenhouse gases between ocean and atmosphere. This role appears to be very important for CO2. For other greenhouse gases the review given by IPCC (Watson et al., 1990; Watson et al., 1992) indicates that the oceans sometimes act as a "minor source" (methane), sometimes as a "significant, but not dominant source" (N20). Yet there are m a n y u n c e r t a i n t i e s , and b e t t e r e s t i m a t e s of the different n a t u r a l greenhouse-gas budgets are necessary before one can make good estimates of the influence of anthropogenic sources. The same review shows that oceanic DMS (Dimethyl sulphide), produced by algae, is an i m p o r t a n t item in the atmospheric sulphur budget. As DMS in the marine a t m o s p h e r e d i s t a n t from land is supposed to be a major source for cloud
286 condensation nuclei (CCN) this DMS production by marine organisms deserves attention in climate prognoses. Changed climatic conditions are likely to cause changes in the contribution of the oceanic sources and sinks of greenhouse gases and DMS, causing feedback processes. Therefore not only improvement of the present budgets but also research of the biogeochemical processes determining the role Of the ocean as a source or a sink, and of physico/chemical processes i n f l u e n c i n g the ocean-atmosphere exchange is necessary. Although the concern for anthropogenic disturbance of climatic conditions is the main impulse for the present interest in the influence of the oceans on the physical and chemical conditions of the atmosphere, ongoing work cannot be isolated from the more general question of the interactions between ocean and atmosphere and their role on the stability of climate. Palaeoclimatic studies of ice cores show t h a t past "natural" climatic change has been accompanied by changes in the atmospheric composition of greenhouse gases (e.g. Raynaud et al, 1993), pointing to feedback processes t h a t are probably of oceanic origin. U n d e r s t a n d i n g and modelling the physical and biogeochemical systems in the world ocean and their interrelations must be an ultimate goal in the study of past and future climate change. This is a task that will last far into the next century. For the moment most studies concentrate on both systems separately and consequently we too discuss them separately. N e t h e r l a n d s oceanographers of different disciplines, both in physics and in biogeochemistry, have experience in various studies of this kind. They are able to contribute to the large international programmes going on or planned for the coming years and in fact are doing so already. The NRP programme has enabled this contribution to be strengthened. In the following the main elements of the physical and biogeochemical systems are given first. Then, in the next sections, the international research background for the respective problems is sketched, and the relevant Netherlands activities are reported. In the following a list is given of the NRP projects within the subtheme "Oceans and the climate system" with the sections where they have been reported (Table 1.1). At the end of the various sections a brief assessment is given of the position of the NRP projects in the national and international research efforts. The activities initiated in the framework of the related "integrated plans" VvA-2 ("The coupled atmosphere-ocean system on time-scales from 0.1 to 100 years"), VvA-3 ("Ocean circulation and climate"), and VvA-9 ("Carbon-balance in the oceanic mixed layer and air-sea exchange") as well as other related oceanic research programmes supported by the Netherlands Science Foundation (NWO) via its subsidiary foundations (SOZ, SRON) have been mentioned when appropriate.
1.2 Ocean physics In the structure and circulation of the oceans we observe variations that are likely to have their effects on the climate. Such variations may influence local or regional atmospheric conditions, but they also can alter the distribution of heat and moisture on a global scale. The processes of i n t e r a c t i o n between a t m o s p h e r e and oceans should be differentiated according to their temporal and spatial scales. On one side one
287 encounters the i n t e r a n n u a l variability of "El Nifio" with its influence on regional climate, on the other side the problem of major shifts found in palaeoclimatic studies and the possibility of multiple equilibria of the oceanic circulation, such as the "conveyor belt" that could be "on" and "off' with dramatic consequences for the global climate. In different studies the physics of this variability is investigated. Although the oceans are much more stable t h a n the atmosphere, changes in properties and circulation do occur. The "El Nifio" phenomenon, discussed later, is one example, the recent "Great Salinity Anomaly" in the North Atlantic (Dickson et al., 1988) is another. The subtropical North Atlantic shows a significant increase of t e m p e r a t u r e since the late fifties (Parilla et al, 1994). Such oceanic variations can have effects on the global or regional climate system, irrespective of any anthropogenic influence. The i n t e r n a t i o n a l scientific c o m m u n i t y has to e s t a b l i s h t h e s e n s i t i v i t y of t h e c l i m a t e for such v a r i a t i o n s in t h e atmosphere-ocean system and to assess the risks of interfering with this system, e.g. by an anthropogenic increase of atmospheric CO2. The oceans redistribute heat over the globe. This transport of heat takes place at different levels. Although it is difficult to make sharp distinctions, we can discern two patterns. First, the upper layers, t h a t are part of the thermocline in a layer several hundreds of meters deep and that are mainly driven by the wind field. Only to a minor degree is the interaction with the atmosphere d e t e r m i n e d by the f r e s h w a t e r balance. The interaction may affect regions of the size of an ocean basin and is at time scales up to decades. The transport in the deeper ocean is driven by thermohaline effects (thus by differences in t e m p e r a t u r e and salinity) and is therefore also determined by the freshwater balance that, however, has a longer reaction time. The interaction may be at time scales of decades and over, and can affect the global ocean. The TOGA p r o g r a m m e is the m a i n oceanographic p r o g r a m m e in the World Climate Research Programme that is dealing with the shorter (interannual) time scales (TOGA Scientific Steering Group, 1985). For the longer time scales the WOCE p r o g r a m m e deals with the ocean circulation at global scales (WOCE Scientific S t e e r i n g Group, 1986). Both p r o g r a m m e s are i m p o r t a n t for our u n d e r s t a n d i n g the role of the oceans in the climate system: they cover different scales in time and space and the basic physics are different, but they are both contributing to the necessary oceanographic database and to the development of oceanographic modelling. The t r a n s p o r t by the oceans at a global scale and over its full depth means a net surface flow to regions of deep convection (the N o r t h Atlantic and a r o u n d Antarctica) and a deep r e t u r n flow to areas with (less localised) upward flow. This is the "conveyor belt" t h a t has its place in palaeo-climatic studies (e.g. Broecker, 1987, Broecker, 1991), and that could be "on" and "off', depending on the different equilibrium conditions of the ocean circulation.
288 Table 1.1 List of projects in NRP subtheme "Oceans and the climate system" Title
Project leader
Number
Phytoplankton and the oceanic carbon cycle; Emiliania huxleyias a model system
P. Westbroek
850003
VIERS-1 field experiment
W.A. Oost
850005
Variability of the North Atlantic sea surface temperature
A. Kattenberg
850007
H.J.W. de Baar
850021
Ocean circulation and climate
W.P.M. de Ruijter
850025
Air Sea exchange of DMS
R. Guicherit
850026
Water/atmosphere exchange of N20 in marine s y s t e ms
W. Helder
850027
ASGASEX
W.A. Oost
852082
Nonlinear dynamics of the equatorial oceanatmosphere system.
H.A. Dijkstra
853110
CO 2
exchange between ocean and atmosphere
Natural variability in a coupled ocean-atmosphere R.J. Haarsma model
853134
Mesoscale mixing processes and watermass transformation in the Greenland Sea
H.A. Dijkstra
VvA3-209
Stability of the general ocean circulation
C.B. Vreugdenhil
VvA3-234
The role of mesoscale eddies in the ocean/ atmosphere heat exchange and meridional heat transport
W.P.M. de Ruijter
VvA3-244
Distribution, chemical composition and isotopic ratios of dissolved organic carbon in the ocean. Part: dissolved organic carbon
H.J.W. de Baar
VvA9-219
The partial pressure of CO2 in sea water as a control on marine biological productivity and calcification
M.J.W. Veldhuis
VvA9-224
289 In theoretical studies of ocean circulation the possibility of multiple equilibria has been discussed for some decades. An important task now is to discover how and why transitions can occur in practice. But then the actual complexity of the ocean circulation as compared with the simple "conveyor belt" picture becomes an important issue. The suggestion of a transport by one, continuous loop should be amended. The stratification, with the water mainly moving along surfaces of equal density (isopycnic surfaces) and the subdivision of the ocean by continents and oceanic ridges in separate basins causes a layering and segmentation of the t r a n s p o r t in m a n y separate loops, each of them linked in a complex way to its neigbours. And the interaction with the atmosphere is along different lines, as discussed above. Oversimplification of this complex system could result in erroneous conclusions with respect to the climate change at time-scales ranging from tens of years to centuries. The thermohaline processes in the North Atlantic are thought to be critical for the operation of the "conveyor belt". In the North Atlantic the variability of salinity at decadal time-scales has received special attention since the "Great Salinity Anomaly", mentioned above, but also in the Pacific decadal variability is observed (Anon., 1992). Combined with palaeo-climatic data from ice-cores t h a t indicate that the climatic stability during the recent geologic past might be "the exception rather t h a n the rule" (Dansgaard et al., 1993), this stimulates the research in the m e c h a n i s m s of these decadal-scale ocean variations, especially in the North Atlantic. In the IPCC analysis the effect of anthropogenic d i s t u r b a n c e s has been investigated with oceanic models, wherein the ocean circulation is not differing much from the present conditions. These models give indications of the response time of the global and regional climatic indicators, such as t e m p e r a t u r e , to changing g r e e n h o u s e gas concentrations. Recently modelling studies have indicated that for certain scenario's of the CO2 input the thermohaline circulation decreases, and they have shown how this circulation may or may not restore in the long run (Manabe, Stouffer and Spelman, 1994). Still these results give no more t h a n indications. Practice shows that there are m a n y difficulties to overcome before models can attain a predictive quality, and this certainly is the case with long-term climate models. Improved models should give more insight into the short-term "natural" climate variations, in the longer-term response of the ocean-atmosphere system to anthropogenic forcing, and in the stability of the present climate system. Better models ask for a variety of observations, of process studies and of theoretical and numerical work. The international TOGA and WOCE programs are m a k i n g a co-ordinated effort aiming at this model-improvement..
1.3 B i o g e o c h e m i s t r y of the Ocean For different greenhouse gases estimates of the role of the ocean in the global budget have been made. However, the input/output values have considerable error-bars. Better figures are required. But without knowledge of the processes governing the exchange between ocean and atmosphere the actual budget is of limited use: interaction between changes in climate and input/output figures can cause a negative or a positive feedback, both with important consequences.
290 For an anthropogenic contribution to the different relevant greenhouse gases it is CO2 t h a t is considered the most important with respect to its effects (Shine et al, 1990). Now the CO2 concentration and exchange are strongly influenced by the biogeochemical processes in the oceans. The ocean continuously exchanges CO2 with the atmosphere, and because the increased anthropogenic input is only partially reflected in an increase of atmospheric CO2 it is likely that a large part is also taken-up by the oceans. Scenarios for future atmospheric concentrations of greenhouse gases are based upon our u n d e r s t a n d i n g of the actual global CO2 budget as presented by Watson et al. (1990) in the IPCC Report. Adaptations to this budget t h a t are expected from the next IPCC Report still leave open various questions on the oceanic uptake. IPCC e s t i m a t e s an extra oceanic u p t a k e of 2.0 + 0.8 GtC/y compared with ocean-atmosphere and atmosphere-ocean fluxes of 90 GtC/y. Other estimates of the redistribution of anthropogenic carbon dioxide result in an imbalance of 1.6 + 1.4 GtC/y , the "missing sink" (Watson et al, 1990, 1991). In a recent review (Siegenthaler and Sarmiento, 1993) this imbalance has slightly increased: 1.8 + 1.3 GtC/y. The different suggestions for the cause of this imbalance have to be explored for the development of better CO2 scenario's. This includes re-assessment of the role of the ocean. Therefore the available figures of the ocean-atmosphere exchange of CO2 have to be carefully scrutinized. I n d e p e n d e n t d a t a are i m p o r t a n t because p r e s e n t e s t i m a t e s are based essentially upon the same d a t a set. For the i m p o r t a n t ocean-atmosphere exchange an international programme (JGOFS: Joint Global Ocean Flux Study) has been initiated to tackle this problem over a broad front (Anon. 1990). Also in the global budgets of some other greenhouse gases the oceans act as a source. Better estimates of this source function for different oceanic environments is important to improve the actual budgets, and thus the concentration scenarios. Moreover, for CO2, as well as for the other greenhouse gases, the possibility of feedback mechanisms by changed environmental conditions is a serious problem. Changes in ocean t e m p e r a t u r e , in ocean circulation, and in the meteorological conditions at the sea surface may influence the oceanic sources and sinks of greenhouse gases. Only if the f u n d a m e n t a l biogeochemical processes are well understood we can take these feedback mechanisms into account. The climatic role of DMS production by algae was already mentioned in the introduction. In the report on atmospheric processes of the NRP programme the effects of DMS on the radiation budget are considered. (Guicherit, this volume). In this report the m a r i n e aspects of DMS production and its exchange with the atmosphere are discussed together with the other marine research. A problem here, t h a t is shared with some other biogeochemical processes, is the strong influence of short-term blooms as compared with more stationary conditions and the different role of various marine algae in its production. For better biological detail an organism-centered approach has been advocated as an alternative. The philosophy here is that different processes (carbon cycle, DMS production) are investigated in combination. Because of its important, sometimes
291 crucial, role in the different processes, and its palaeoclimatic significance the alga
Emiliania huxleyi has been chosen as a "model" organism in the international "Global Emiliania Modelling Initiative" (GEM) and the related European EHUX programme. Solutions for the the anthropogenic CO2 increase have been advocated t h a t in some way or a n o t h e r propose to enhance the storage in the oceans, e.g. by fertilizing the oceans or by deep sea disposal. It is clear that, as long as such technological remedies are popping up, it is necessary to have a broad knowledge of the processes involved. On the basis of present understanding these solutions are considered inefficient or unfeasible (De Baar and Stoll, 1989 and De Baar, 1992). 2.
O C E A N P H Y S I C S AND C L I M A T E
2.1 I n t r o d u c t i o n Model studies as well as different types of observations (e.g. oceanographic surveys, d a t a from a u t o m a t i c buoys, satellite observations, etc.) are both n e c e s s a r y in the p r o g r a m m e s t h a t investigate the m a i n components of the physical system of the oceans and their interaction with the atmosphere. The most i m p o r t a n t short-term variability is investigated in TOGA (the influence of i n t e r a n n u a l variations in the tropical oceans - E1 Nifio - on the climate) and the longer-term, global variability (including the role of the deep ocean circulation) in WOCE. The latter has different sub-programmes, aiming at the investigation at different spatial and time scales. Much work still has to be done in the modelling of the oceans alone, with prescribed boundary conditions. This includes the study of different processes t h a t have a crucial role in the system, but t h a t often only can be parameterized. Yet the modelling of ocean-atmosphere exchange is the ultimate goal. But coupled ocean-atmosphere models again have their specific problems.
The i n t e r n a t i o n a l p r o g r a m m e s TOGA and WOCE are i m p o r t a n t steps in the solution of these problems. They also m a r k a development towards a system of co-ordinated observational systems, generating data that, in combination with data-assimilation models, should be able to give early warnings and predictions of climatic developments. The basis for our u n d e r s t a n d i n g of the physical system is the development of models r e p r e s e n t i n g the essentials of the system. Usually one thinks here of numerical models, and only such models can cope with the complexity of the whole system, but they are the result of an integration of m a n y different theoretical and observational studies t h a t have their place in the assessment of the processes involved. F u r t h e r m o r e we need data on the actual and past conditions in the oceans, their structure and circulation for initialization and validation of models. Data-sets from the past often do not attain required accuracy and the distribution of observations over time and place is uneven. Therefore parallel to the modelling effort there is need for a new observational p r o g r a m m e t h a t u l t i m a t e l y m a y develop into an operational observing system. Most of the physical oceanographic research in the framework of climate change is c o - o r d i n a t e d in the i n t e r n a t i o n a l TOGA and WOCE p r o g r a m m e s . Both programmes concentrate on the development of models (stand-alone ocean models
292 or coupled ocean-atmosphere models), but they require a large effort in the international development of observational strategy and the organization of different process-studies, including the maintenance of quality criteria for the data and the organization of data banks. For the investigation of interannual variations of climate we need models of only the upper layers of the ocean. This is the approach of TOGA ("Tropical Oceans, Global Atmosphere"). Starting point for TOGA is the observation that the atmosphere, especially at lower latitudes, shows (atmospheric) climatic variations at periods that are usually between 3 to 5 years with no strict regularity: the Southern Oscillation. Later investigations indicated the connection between this phenomenon and oceanic variability in the tropical Pacific, along the Peruvian coast known as the "El Nifio" phenomenon, that involves large, multi-annual, variations in sea-surface temperature. This connection is, for instance clear from data shown by the Climate Analysis Centre, Washington, (published by the World Climate Data Programme ofWMO, s.d.). The idea that interaction between certain progressive wave-like processes typical for the tropical oceans and the atmosphere could be responsible for the climate variations led to the ENSO (El Nifio - Southern Oscillation) concept. The TOGA programme, expanding this concept, tries to arrive at the predicting of short-term climatic changes on the basis of a combination of oceanic models and oceanic observations (TOGA Scientific Steering Group, 1985). Although the most pronounced climatological effects of ENSO are in the Pacific region, there are indications of effects of ENSO on the European climate as well (Fraedrich, 1994). However, for longer-term variability t h a t incorporates an increased greenhouse-effect (together with long-term instabilities) we need models for the whole ocean and the complete thermohaline interaction process. In particular simulation and study of time-dependent changes requires that these models are coupled in a realistic way with atmosphere models (Bretherton, Bryan and Woods, 1990). The issue here is not so much prediction at longer time scales, but simulation of climate variation under influence of different external conditions (e.g. an increased greenhouse effect): sensitivity studies. The emphasis is on the development of global or ocean-wide General Circulation Models (GCM's). These GCM's depend on observations for their initialisation and validation, on either historical or new data (temperature-, salinity-, density-distribution, tracers). The modelling of the oceans that is possible by the development of increased computing power, and the new instrumental possibilities for observation of ocean circulation are equally important for the oceanographic part of climate research. This is the basis for WOCE (World Ocean Circulation Experiment), the international oceanographic programme in the WCRP. This programme has a field phase between 1990 and 1997, contemporary with the period of the new ocean-observing satellites (ERS-1 and TOPEX-POSEIDON). These satellites play an essential role in the WOCE programme are supplementing the many in-situ oceanographic observations made during this period. In the modelling phase of WOCE these data will be indispensable until far into the next century (Koblinsky, Gaspar and Lagerloef, 1992).
293 The WOCE p r o g r a m m e has three "core projects". Core Project 1, the Global Description, aims at the "zeroth order" description of the role of the oceans in the p l a n e t a r y climate system. Core Project 2, the Southern Ocean, concentrates on the formation of Antarctic w a t e r masses, and on the connections between the different ocean basins. Core Project 3, the Gyre Dynamics Experiment, focuses on the role of smaller-scale processes (quasi-geostrophic eddies, topographic effects and subduction of water masses) on the circulation of one ocean basin: the North Atlantic. Whereas Core Project 1 is essential for the long-term modelling (say 100 years or more), Core project 3 also is important for the modelling of shorter-term variability (of the order of 10 years). The development of ocean models is one step in the understanding of the climate system, the coupling of oceanic and atmospheric models is an other one. The physical coupling of the atmosphere and the ocean is by the exchange of heat (sensible and latent, and the balance of long-wave radiation), of m o i s t u r e (evaporation and precipitation) and momentum (wind stress). Indirectly also the short-wave radiation contributes, because atmospheric conditions determine the radiation reaching the sea surface. E s t i m a t i n g the heat, moisture and m o m e n t u m fluxes is done by p a r a m e t e r i z e d relations between bulk properties of the atmospheric boundary layer, and the surface layer of the sea. In these estimates different spatial and temporal scales m a y be used. The observations over the ocean are scattered in space and time and are often inaccurate. Occasionally this can result in error bars of the fluxes of the order of 100 W/m2. Compared with an expected (long-term) greenhouse effect of about 5 W/m2 this indicates the urgent need for better estimates. The use of atmospheric general circulation models through the assimilation of surface data a p p e a r s the best way for better estimates of the the fluxes with the relevant resolution in space and time. Here also the oceanographic satellites are a powerful resource. For e s t i m a t e s of the flux of m o m e n t u m all over the oceans the scatterometer data from satellites offer new possibilities. Coupling of ocean and a t m o s p h e r e models is also a modelling problem. Atmospheric and oceanic models t h a t operate well on their own produce the so-called "climatic driW' when coupled. This is because of inadequacies in the ocean models, because of uncertainties in the fluxes and lack of feedbacks to correct for these. Flux corrections are applied to reduce the climatic drift, but how to do this without predetermining the result is a serious problem. WOCE and TOGA are not the final programmes in this field. They also are a first step towards a system of oceanic observations and data-assimilation models, comparable to the meteorological system (with differences in spatial and time scales). Within the limits of predictability such a system should be an essential tool for the a s s e s s m e n t of n a t u r a l and a n t h r o p o g e n i c climate changes. E x p e r i m e n t a l predictions of the "El Nifio/Southern Oscillation" (ENSO) phenomenon have a moderate skill at lead times of one year. As to the problem of longer-term climatic trends, t h a t presently are not exceeding the "noise", it has been indicated by McBean (1994) that future predictive models may be able to reduce this "noise".
294 2.2 M o d e l s
Models are being developed for prediction of climate variations at interannual timescales. Models have also been used to assess the consequences of CO2 forcing, and the possible changes in ocean circulation because of changes in other external factors (occurrence of multiple equilibria). The number of models capable of realistic assessments is limited. However, further development of these models requires various additional modelling excercises, with less elaborate models. The different lines of Netherlands research in this field are referred to in the context of the international efforts. Models are essential tools for testing hypotheses on the role of the oceans in the climate system. But this asks for a critical attitude towards the quality of these tools. Model solutions can have different structures, depending on their parameterization (e.g of the exchange with the atmosphere, or of oceanic sub-grid processes). Developing adequate ocean models for different conditions is an important aspect of climate research. Since the start of the TOGA programme considerable progress has been made in the modelling, understanding and predicting of ENSO. In an evaluation of 1990 it was stated that it was predictable, but the skill of predictions was indicated as "modest" (WMO/IOC Intergovernmental TOGA board, 1990). The official TOGA programme is completed on 31 December 1994, but the development of predictive models is not finished. Short-term modelling furthermore is important in improving modelling skill in general. Models covering the whole ocean are the objective of WOCE. They have to answer the questions about multiple equilibria and those about the time response of the ocean to anthropogenic disturbances. In the 1992 IPCC report four coupled ocean-atmosphere models were used to investigate the transient results of increasing CO2 forcing: GFDL (Geophysical Fluid Dynamics Laboratory, Princeton, USA),MPI (Max-Planck Institute, Hamburg, Germany), NCAR (National Center for Atmospheric Research , Boulder, USA), and UKMO (Meteorlogical Office, Bracknell, UK). The development of OGCM's (Oceanic General Circulation Models) capable of this type of experiments is a major task and is therefore only done at a few places. However, these models are the result of an extensive international modelling effort, supported by a "family" of intermediate models that concentrate on special regions or special aspects of the ocean circulation. For instance, a question in the development of ocean models is the role of mesoscale eddies in the ocean. For coarse-grid models the role of these eddies in the horizontal transports has to be parameterized. Also the parameterization of other locally occurring oceanic processes as deep convection is critical. A question that occurs in the coupling of atmospheric and oceanic models is t hat of the best estimates of the ocean-atmosphere fluxes (of heat, moisture and momentum). The sensitivity of the system for different effects can be explored in intermediate models, sometimes with relatively modest means. Outside the centres where the C~M's are developed important modelling work can be done in these fields.
295 The WOCE Numerical Experimentation Group (WOCE International Project Office,1994) has developed a strategy for ocean modelling. This group concludes that for the further development of ocean models one should: 1. Encourage greater support for the development of small-scale process models. Areas of particular importance are the oceanic mixed layer, air-sea fluxes, deep convection, bottom water formation, overflows between ocean basins and mixing within the ocean. 2. Continue encouraging high-resolution studies of the ocean. It should be emphasised that what is required of these studies are models with the highest possible realism. Such studies will need the fastest of modern array processor computers. 3. Encourage the development of more accurate fully coupled models, the validation of such models using the high resolution models and the small scale process models,and the use of such models to study the circulation of the ocean, its variability and effect on climate change. In addition the group draws conclusions as to the strategy with respect to data assimilation and the setting-up of operational oceanographic analysis and forecast centres. Here models will have an important role (Smith, 1991). Work on ocean models in the framework of NRP and in related Netherlands p r o g r a m m e s is directed at (1) the modelling of s h o r t - t e r m coupled o c e a n - a t m o s p h e r e models, at (2) the study of the role of different p a r a m e t e r i z a t i o n s for the model solutions of long-term stability of the thermohaline ocean circulation and (3) at the development of small-scale process models that should form fundamental building blocks for the final large-scale models. In the following these activities and the perspectives and results are reported.
Short-term modelling This point in the first place addresses not the possible anthropogenic climate change, but the natural variability at interannual to decadal time scales. The question whether it is possible to predict short-term (interannual) climatic change is of direct interest for society. The adequate coupling of atmosphere and ocean is crucial for this short-term modelling. In most models a flux correction is applied to suppress so-called climatic drift. Much work is being done to investigate the influence of the various coupling procedures. It is clear that a flux correction easily may limit the applicability of such coupled models to variability studies. In Neelin and Dijkstra (1994) it is shown that multiple equilibria can be produced by coupled models as artefact of the flux correction. An objective of TOGA is to develop operational predictive models. However, notwithstanding the progress in modelling the E1 Nifio phenomenon, the fundamental reasons for its irregularity still are under dispute, and the reliability of predictions over longer periods only can be improved if this point is clarified. For the study of these problems there is place for relatively simple analytical models, parallel to the "brute force" numerical models (e.g. the E1 Nifio results reported in Science 264:70-74 from the numerical model of Jin et al., 1994, and the analytical model ofTziperman et al., 1994).
296 In the NRP/VvA project NRP project 853110 "Nonlinear dynamics of the equatorial ocean-atmosphere system" (H.A. Dijkstra, Institute for Marine and Atmospheric Research, Utrecht - IMAU) this point is tackled by such a parallel approach with a combination of such models. The aim is to determine the role of nonlinear wave structures in the interannual variability. The results could be a starting point of a systematic validation of certain components of a predictive GCM. In a number of interconnected modelling projects the KNMI, especially in co-operation with the Max Planck Institute of Hamburg contributes to the modelling of short-term variations. In Germany the interest in short-term modelling led to a concentrated effort for developing a coupled ocean-atmosphere high-resolution model for climate predictions over 1 year. This model has been developed by the Max-Planck Institute in co-operation with the Netherlands KNMI and the ECMWF ( E u r o p e a n C e n t r e for M e d i u m - R a n g e W e a t h e r ForecastsForecasting). (Latif et al., 1994, Stockdale et al., 1994). Although this modelling is not specifically directed to the TOGA region but to a more general applicability, the KNMI contribution to this project is directed to the modelling of the tropical Pacific Ocean in relation to the TOGA programme. Parallel to this in the NRP project "Variability of North Atlantic Sea Surface Temperature" of Kattenberg (NRP project 850007) the aim is to find adequate procedures to incorporate the mixed layer in an (existing) North Atlantic model (Kattenberg, 1993). This work is further discussed below in Section 2.1. At longer time-scales variability in the thermohaline processes becomes important. It was shown by Lenderink and Haarsma (1994) that regions can be identified in a simplified N.Atlantic circulation model in which unstable conditions prevail where deep convection may, or may not, occur. So this work in some respects is comparable at basin-scale with the modelling of the large-scale multiple equilibria that may determine the operation of the global "conveyor belt", and it may operate on decadal or centennial time scales. However, it is not about a conveyor belt being "on" or "off', but about the intensity and the regional extension of the deep circulation. The NRP/VvA project NRP project 853134 "Natural variability in a coupled ocean-atmosphere model" (R.J. Haarsma, Royal Netherlands Meteorological Institute - KNMI - De Bilt) continues this work by coupling the ocean model with an atmospheric model for the hydrological cycle and a sea-ice model. The aim is to investigate the natural variability of this system in the range of 10 to 100 years.
Long-term modelling For long-term modelling the whole ocean and the whole range of interacting processes should be taken into account. The number of Global General Circulation Models capable of this is small. But developing the scientific background, necessary for improving these models is an international task. One activity is the development of small scale process models to be discussed later. An other approach is the development of less elaborate models, with a higher degree of parameterization. These models should be capable of highlighting particular aspects of the modelling of the circulation.
297 In certain respects the VvA-2 project "Natural v a r i a b i l i t y in a coupled ocean-atmosphere model" discussed above belongs to this category, but this considers more the shorter time-scales. In the VvA-3 project "Stability of the general ocean circulation" (C.B. Vreugdenhil), originally initiated as a part of NRP project 850025 "Ocean circulation and climate" (W.P.M. De Ruijter, Institute for Marine and Atmospheric research, Utrecht - IMAU) a simplified - initially 2-d meridional - model of the ocean circulation is being developed for this type of studies. This model, as the more complex models, shows multiple equilibria. By a n a l y s i n g the s t r u c t u r e of the equilibrium model-solutions for different parameterizations in a s y s t e m a t i c w a y t h e s e n s i t i v i t y for t h e s e parameterizations is being investigated. This, for instance, refers to the role of convective adjustment (Section 2.1) and of the atmosphere-ocean coupling. The results thus obtained have to be tested finally in more realistic, 3-d models. Process models
Mixed layer models. The oceanic mixed layer is an important factor in linking atmospheric and oceanic models. The mixed layer is formed either by convection (during cooling) or by turbulent mixing (by wind stress and current shear). The applicability of the different existing theoretical models appears to be restricted, and the parameterization of the mixing in the usual GCM's is still unsatisfactory. This is one of the problems in the development of coupled ocean-atmosphere models. In the N R P project 850007 "Variability of N o r t h Atlantic Sea Surface T e m p e r a t u r e s " (A. Kattenberg, Royal N e t h e r l a n d s Meteorological I n s t i t u t e , KNMI, De Bilt), an explicit Mixed Layer Model was developed. This model was combined with the MPI North Atlantic model, and its performance was tested (Sterl and Kattenberg, 1993, 1994). An improved representation of the sea surface temperatures was found, especially under conditions of wind stirring. However, in the MPI model convection is not treatde adequately, and this cannot be cured by the combination with the mixed-layer model. This leads to the conclusion that our understanding of convective processes and their modelling has to be improved. Yet outside the convective regions, in the (sub-)tropics, the introduction of the mixed-layer model shows a more realistic modelling result 6). A further result is the reduction of errors in the heat flux, and this can improve the sensitivity for climate drift.
Deep convection. Is a small-scale process that yet forms an important link in the global thermohaline circulation of the oceans and therefore should be represented accurately in the climate models. In the North Atlantic the Greenland Sea and the Labrador Sea are important locations for deep convection. Deep convection results from extreme cooling events at high latitudes. To penetrate down to the ocean bottom the water mass formed should (in-situ) exceed the densities at that depth, depending on its t h e r m o d y n a m i c properties as determined by the acquired temperature and salinity. Observations have shown that, before the deep convection starts, a series of processes of different scales are contributing to the preconditioning of the water in the upper layers for a deep convective event (Schott, Visbeck and Send, 1994). An
298 important stage is the formation of "chimneys" of some 100 meters deep and with diameters of the order of 100 km. These chimneys again are thought to be the result of the formation of smaller-scale (order 100 meters) convective "plumes" (Marshall, Whitehead and Yates, 1994). In this small-scale process the formation of ice and the resulting brine ejection appear to be important. A problem is that small-scale processes cannot be observed directly in the sea. Conjectures, made on the basis of physical reasoning and founded on tests in laboratory models and by numerical experiments, should finally result in a picture of the process of convection as a whole and a parameterization of this process in GCM's. In the VvA-3 project "Mesoscale mixing processes and w a t e r mass transformation in the Greenland Sea" (H.A. Dijkstra, Institute for Marine and Atmospheric Research, Utrecht-IMAU) the aim is to contribute to a further understanding of the succession of processes. It concentrates on the role of preconditioning by freezing (Dijkstra and Molemaker, 1994, Dijkstra and Kranenborg, 1995) and on the role of differential diffusion of heat and salt (salt-fingering and the formation of "staircases") in competition with plume-formation, and their consequences for the appropriate parametrization of the convective process. Eddies. In many respects eddies in the ocean are comparable with depressions in the atmosphere. Atmospheric transport of heat and moisture is strongly dependent on the contribution by mid-latitude depressions, and it is therefore probable that, mutatis mutandis, this also is the case for the oceanic eddies. However, their size is about one tenth of that of the atmospheric depressions, estimating their effect by measurements is difficult and a model that explicitly takes them into account (eddy-resolving) is very demanding in computing power. The study of the role of eddies therefore is important for the development of oceanic climate models. Their effect is on the circulation pattern and on the transport of heat (and other properties). This role also appears to be different for different ocean regions (e.g. Gill, 1983). The incorporation of the role of eddies in GCM's by adequate parameterization, is a major task for further ocean modelling. Such a parametrization should consist of the eddy-transport of heat, salt, (potential) vorticity and momentum, and the eddy-induced changes in mean-flow transport of these quantities. A recent parameterization (Danabasoglu, McWilliams and Gent, 1994, Neelin and Marotzke, 1994) accounts for eddy-transport of heat and salt by linear eddies, which is believed to occur along isopycnals with the constraint of no density transport, while at the same time the isopycnal slope is reduced. This parameterization has been shown to improve steady state modelling. However, it does not account for the transport of isolated eddies (rings) and the transport of vorticity and momentum by eddies. Therefore it should be considered only as a first step towards a complete parameterization. The question investigated in the VvA-3 project "The role of mesoscale eddies in the ocean/atmosphere heat exchange and the meridional heat transport" (W.P.M. De Ruijter, Institute for Marine and Atmospheric Research, Utrecht - IMAU, and G.J. Komen, Royal Netherlands Meteorological Institute - KNMI- De Bilt) is directed to the role of eddies on the heat transport and the consequences for the
299 parametrization. The exchange with the atmosphere on eddy-scale appears an essential element and is explored in this study. This work extends earlier studies by Drijfhout (1990, 1992, 1994). In the VvA-3 study the role of eddies in a combined isopycnic/mixed-layer model is investigated.
Circumpolar current. The "conveyor belt" scheme of the ocean circulation shows that the Atlantic has to receive surface water from the Indian and Pacific Oceans, t h a t is returned at greater depths. From the Indian Ocean the Atlantic receives water transported by the Agulhas rings, dicussed later (Section 2.2). From the Pacific the transport is via the Drake Passage by the Circumpolar Current. The physics of this current have been described by Nowlin and Klinck (1986). By connecting the different (Atlantic, Pacific and Indian) ocean basins it is a critical link in the global circulation. Its study is a central element in the WOCE Core Project 2. Different models of the Southern Ocean and the Circumpolar Current exist. Eddies play an i m p o r t a n t role in the dynamic balance. One of the programmes in NRP project 850025 "Ocean circulation and climate" (W.P.M. De Ruijter, Institute for Marine and Atmospheric research, Utrecht - IMAU) is investigating the dynamical balance of the Circumpolar C u r r e n t and its relation to the Drake Passage transport in an eddy-resolving numerical model. (For the numerical methodology used: see Walsteijn, s.d.). The sensitivity of the current for stratification and windforcing and their role for different sectors of the current are investigated in this model. It appears t h a t there are important variations at timescales of several years. Assessment Different reasons can be given for a strengthening of Netherlands ocean modelling effort by the NRP programme. Development of a separate national global general circulation model cannot be the objective: it would be inefficient and unrealistic. Association with one or more of the groups that develop such a model is a more promising option. F u r t h e r possibilities are those contributing to the international modelling effort, e.g. along the lines of the WOCE modelling strategy (WOCE I n t e r n a t i o n a l Project Office, 1994), or a critical analysis of the modelling premisses. We note that the NRP project 850007 is done in association with the German MPI modelling development, and that it also is contributing to the mixed layer modelling (especially important for developments at interannual time-scales). The modelling efforts s t a r t e d in NRP project 850025 are more contributing to the last two options mentioned above.
The work in NRP project 850007 has already given interesting results, and in the combination with the MPI modelling has the promise of at least a b e t t e r u n d e r s t a n d i n g of the interannual climate variations in the North Atlantic area. The work on TOGA can be considered as a useful excercise for an area with strong interannual signals. In addition: there are indications that the E1 Nifio phenomenon has certain climatic effects at a global scale (e.g. McPhaden, 1994). Part of the modelling activities in NRP project 850025 had a late start. They are continued in the VvA-3 programme. Published results are only available on the
300 model development. A critical systematic analysis of certain modelling approaches is certainly important. The modelling of the Antarctic Circumpolar Current fits with other activities undertaken in the framework of this NRP project project (see below). Also the NRP projects 853110 and 853134 have been initiated recently. Earlier work on these subjects is promising. The different modelling projects strengthen the existing Netherlands experience in the field. They follow different lines, but are sufficiently connected to w arrant a good overall contribution to the international modelling efforts. 2.3 O c e a n o b s e r v a t i o n s In spite of the available oceanographic data the modelling of the role of the ocean in climate change requires a new, high-quality global data set. The WOCE programme is designed to generate this set, and also to provide a basis for long-term monitoring of the ocean and its interaction with the atmosphere. Netherlands observational programmes in the North Atlantic are contributing to the WOCE in-situ data set. Some of these programmes can also provide support for the biogeochemical observations discussed in Section 3. Furthermore Netherlands efforts in the processing and analysis of altimeter data from ERS-1 and TOPEX-POSEIDON not only contribute to the application of altimetry techniques to oceanography, but also provide information on the actual developments in the southern Oceans. Direct observations of changes in the ocean require long series of regular data of critical oceanographic parameters. Such observations have a warning function t h a t something is going-on (e.g. the "Great Salinity Anomaly"). Regular observations improve the predictive capability of models by data assimilation. Existing data programmes are only partially useable for these purposes. The number of long-term oceanographic data-series from one location is small, and most are from non-representative coastal seas. Also ocean data sets are needed for initialisation and validating of ocean models. At present the available oceanographic data (those stored in data centres and screened) that are our source of information are not adequate. The Levitus data set (Levitus, 1982) is often used, but it has not the quality that is needed in many cases. The WOCE hydrographic programme has the purpose to create a better and more "synoptic" data-set of the actual conditions and to support the development of a more p e r m a n e n t observational system (GOOS: Global Ocean Observing System, and GCOS: Global Climate Observing System). Also TOGA is generating a contribution to such a system. Efforts are now underway to set-up an adequate system for this purpose on an international basis. (Ocean Observing System Development Panel, 1993). The scientific background of the WOCE observational programme, edited by the WOCE International Planning Office (1988), formulates in detail the different components in the different "Core Projects". The whole programme is the most extensive observational effort in oceanography and with the available means it only can be accomplished within a number of years: between 1990 and 1997. It also has been agreed that progammes in the framework of WOCE can support JGOFS activities and vice-versa, by offering facilities for observations. It also is
301 clear that the detailed WOCE observations offer a physical framework for different biogeochemical studies. In view of the variability of less than 10 years a programme of repeated sections, fixed stations and regular XBT programmes is planned to resolve these time scales. In addition the satellite data are used. The altimetry of the sea-surface is particularly important in this respect. The observations from the ERS-1 and TOPEX-POSEIDON satellites form an essential contribution to the WOCE programme. They provide a high-density (space and time) picture of oceanic variability that, together with the in-situ observations gives a fairly complete picture of the ocean circulation. The accuracy expected from these satellites is adequate for the identification of the main oceanographic processes. Earlier (Geosat) data already show the potential of these observations in regions with strong sea-level signals. Different programmes in the framework of WOCE Core Project 3 in the North Atlantic have a Netherlands contribution. As to altimetric observations from satellites, this is being stimulated by the Netherlands Space Research Foundation (SRON). This contributes to a fruitful co-operation between experts in satellite altimetry and oceanographers from Delft and Utrecht Universitites, and it has resulted in application of these techniques from Geosat, and, more recently, ERS-1 and T O P E X - P O S E I D O N (see H a a g m a n s , Naeije and Feron, 1993) to oceanographic studies of the Southern Oceans (WOCE Core Project 2). The S o u t h e r n O c e a n The planning of the Southern Ocean programme of WOCE (Core Project 2) started in 1987 (Anon., 1987). The observations should be directed to the Antarctic Circumpolar Current, including interbasin exchange, the meridional transport, including that in and out of the Southern Ocean, and the ocean-atmosphere fluxes. Especially for this area, with its vastness and difficult environmental conditions, satellite data are very important.
The exchange between the Indian and the Atlantic Oceans south of Africa appears to be an involved process. The dynamics of the Agulhas Current, that runs westward south of the Cape, are important in this connection. Most of the transported water retroflects eastward, but large rings shedded from the current transport water to the south Atlantic. Estimates of this transport, its variability and understanding the dynamics of the shedding process are necessary elements in the modelling of the global circulation. As one of the sub-projects in the programme (NRP project 850025) "Ocean Circulation and Climate" (W.P.M. de Ruijter, Institute for Marine and Atmospheric Research, Utrecht - IMAU) an analysis was started of satellite data from the Agulhas retroflection, for a better description of the meso-scale processes occurring. Originally working with Geosat data techniques have been developed that allow to extract the characteristic spatial and temporal scales of the currents in this region. The large warm eddies that are formed in the Agulhas retroflection area and their movement to the northwest can be observed (Feron, 1992). This results in a better description of the ring shedding process (Feron, De Ruijter and
302 Oskam, 1992). The analysis of the altimetric data applied was compared with results from the FRAM model (Feron, 1993, 1994a, 1994b). A further phase of the programme is to investigate the relation between the formation of the rings and other variations of the Agulhas system (De Ruijter, Van Leeuwen and Lutjeharms, (submitted), Van Leeuwen and De Ruijter, (submitted). For this continued work the new r e s u l t s from the ERS-1 and TOPEX-POSDEIDON missions are available. The higher accuracy and higher resolution of these data enlarges the possibilities of satellite altimetry for this work. The N o r t h A t l a n t i c For the oceanographic changes in the Atlantic the WOCE Core-project 3 ("Gyre dynamics") provides the framework. An important aspect here is the interannual variability. At basin-scale the circulation in the North Atlantic is composed of a Sub-Tropical (anticyclonic) and Sub-Polar (cyclonic) Gyre. In the Sub-Tropical Gyre water that has exchanged properties with the atmosphere is subducted in the permanent thermocline, to be transported as "18 ~ Water" and "Sub-Polar Mode Water" over periods of several years around the Sargasso Sea until surfacing again in the extension of the Gulf Stream, when the cycle may be repeated, or crossing-over to the Sub-Polar Gyre takes place. In the Sub-Polar Gyre the water is gradually conditioned by cooling until deep convection is realised.
Recent provisional results of this part of WOCE show that there is a warming of the upper 3000 m over the past 35 years, and cooling below 3000 m in the subtropics (Parilla et al, 1994), while marked variations in the conditioning in the Sub-Polar Gyre have been found by Read and Gould, 1992. It is clear that such changes at decadal time-scales have to be modelled adequately before one can hope to predict climate changes over such periods. In the deeper parts of the North Atlantic deep water masses flow southward as part of the long time-scale ("conveyor belt") circulation. Regular observations of the variability of the oceanic structure between Europe and the Antilles (Sub-Tropical Gyre) have been made since 1991 in the VvA-3 project "Repeated XBT sections and analysis of variable meso-scale structures in the N.Atlantic in the framework of WOCE" (L. Otto, Netherlands Institute for Sea Research - N I O Z - Texel). This is a co-operative programme of the Royal Netherlands Navy and the Netherlands Institute of Sea Research (NIOZ). These observations can be used to estimate the variability in the subduction and in the transport by a branch of the Sub-Tropical Gyre (Azores Current), and provide information on the structure of the Mediterranean outflow (De Bruin et al, 1992). In the "DUTCH-WARP" programme financed by the former SOZ ("Stichting Onderzoek der Zee" or Marine Sciences Foundation) observations have been made to describe and quantify the circulation in the eastern part of the Sub-Polar Gyre, where the main transport to areas of deep convection takes place. In this framework also drifters have been released ( Otto, Van Aken and de Koster, 1992), and this part of the programme is continued as the VvA-3 project "Determination
303 of the current field and the diffusivity in the NE Atlantic with ARGOS drifters in the framework of WOCE". As a result the picture of the circulation and of the eddy kinetic energy in this part of the Sub-Polar Gyre is improved (Otto and Van Aken, submitted). As p a r t of the "DUTCH-WARP" programme also the deep circulation and especially the overflow from the Norwegian Sea across the Scotland-Faroe-Iceland Ridges was observed. This overflow is an important element in the deep return branch of the "conveyor belt" circulation. Both long-term current measurements (Van Aken, 1993) and hydrographic analysis (Van Aken, submitted) are used to evaluate this circulation and its variability (De Boer, van Aken and Van Bennekom, 1991, De Boer, 1993). Assessment Next to the various modelling activities reported above, good ocean data are necessary. Therefore a s u b s t a n t i a l support from the N e t h e r l a n d s to the programme of WOCE observations is important for international research . Although not incorporated in the NRP-1 programme, such a contribution to WOCE is partly being made via the VvA-3 programme. As the WOCE programme and subsequent observational or monitoring programmes need contributions from many countries, a continued Netherlands contribution in the framework of NRP and/or VvA certainly has its place. This also supplements the biogeochemical programmes in the oceans. For the past period the NRP support in NRP project 850025 to application of satellite altimetry to the observation of the Agulhas retroflection has resulted a number of publications that give valuable oceanographic information. Moreover the development of experience in oceanographic applications of satellite altimetry is important for future monitoring of the oceans in relation to the Global Climate Observing System.
It is desirable to consider a further focussing of the observational programmes (including regional JGOFS programmes, see below) to areas of common interest. Application of altimetry to the eastern North Atlantic now appears to be possible with the higher-accuracy satellites, and the southern oceans are of special interest to JGOFS. Also in view of the modelling activities in the North Atlantic and the Southern Oceans this could result in an effective exchange of experience and ideas between Netherlands oceanographers.
2.4 O c e a n - a t m o s p h e r e i n t e r a c t i o n Our limited knowledge of the exchange between ocean and atmosphere is a problem for climate studies. Higher-quality observations are necessary and a better understanding of the physics of the exchange processes. Satellite data are becoming more important for estimating global ocean-atmosphere exchange. A "stand-alone" ocean model (such as are found in long-term modelling) is constrained by prescribed conditions at the ocean-atmosphere interface. If realistic values of the wind stress and of the fluxes of moisture and heat can be given, the model performs according to its inherent quality. However, the estimate of the wind stress and the fluxes of moisture and heat is still a major problem, because of the wide range of spatial and temporal scales involved, and the
304 inadequacy of observations. The first task is therefore to improve these estimates. In WOCE accuracies of 10 W/m2 are sought in estimates of the mean heat flux components, averaged over m o n t h l y and longer time-scales, of 1 mm/d in evaporation and precipitation and of 10% or at least 0.01 Pa in surface stress (WOCE I n t e r n a t i o n a l Planning Office,1988). This has to be accomplished by optimizing the regular meteorological observations. Application of satellite data t h a t are matched with in-situ data, and data-assimilation procedures are further ways of improvement. For estimates of the water budget direct observations over the oceans are too inaccurate to be used. Only indirect estimates are available. For a comparatively well-known area such as the North Atlantic estimates of the total evaporation minus precipitation differ by several times the Amazon run-off (Schmitt and Bryan, 1991). Satellite data are translated into physical quantities with algoritms that formulate the relation between observed values and the values sought-for. These algoritms as a rule are derived from process-models based upon micro-meteorological observations. In this way small-scale local studies are important for obtaining reliable large-scale global results. The a v a i l a b i l i t y of good facilities to s t u d y air-sea i n t e r a c t i o n processes (observation platform, flume) and the experience in modelling air-sea interaction is a basis for N e t h e r l a n d s contributions to these problems. The different programmes in this field are discussed below. It is logical that the field programmes also incorporate observations of the chemical fluxes.The physical observations are also needed for the interpretation of the chemical fluxes (see Section 3.1). Small-scale interaction The VIERS-1 programme t h a t aims at developing algoritms for obtaining wind stress data from scatterometer observations has been running since the eighties. It is a joint programme of a number of Dutch and foreign institutes in the course of which a number of field experiments have been conducted at the open sea research platform "Meetpost Noordwijk".
In connection with this programme a NRP programme (NRP project 850005) "VIERS-1 (Preparation I n t e r p r e t a t i o n ERS-1) field experiment". (W.A. Oost, Royal Netherlands Meteorological Institute - KNMI - De Bilt) was performed in which, besides the fluxes of momentum, sensible heat and water vapor, also the flux of carbon dioxide was measured, using the eddy-correlation technique. The CO2 flux m e a s u r e m e n t s by eddy-correlation are difficult to make, but valuable experience was obtained and this resulted in the s u b s e q u e n t ASGASEX experiment (further discussed in Section 3.3). The other observations result in a set of data of the relevant physical parameters under a variety of conditions. An i m p o r t a n t aspect of the project is that the flow-distortion by the Noorwijk platform has been carefully studied. This has resulted in the characteristics of this distortion being well-known, which adds to the value of this platform among the few facilities for all kinds of flux measurements by eddy-correlation techniques. (Oost, 1994, Oost et al., 1994)
305
Large-scale observations The assessment of large-scale fluxes between ocean and atmosphere (necessary for using coupled models), remains a considerable problem, as the necessary meteorlogical and oceanographic data cannot be obtained in the conventional way. The distribution in time and place of observing in-situ platforms never will be a d e q u a t e to obtain the n e c e s s a r y r e s o l u t i o n in space and time. Only remote-sensing data, combined in data-assimilation schemes with large-scale numerical models have, in principle, this capability. Experience in wave-modelling has given an i m p o r t a n t contribution to the development of data-assimilation techniques for estimating the momentum flux at a global scale from satellite (scatterometer) data. In a joint programme with the ECMWF (European Centre for Medium-Range Weather Forecasts), supported by the Netherlands Remote Sensing programme (BCRS), this has already resulted in good-quality wind-stress fields over the oceans. (Janssen, 1994). In the VvA-2 project "Assimilation of wind and wave data to produce global flux fields" this approach is being developed further.
Assessment The NRP project 850005 is a limited project, mainly a precursor of the ASGASEX project (Section 3.1). The work on small-scale physical exchange is part of a long series of s i m i l a r e x p e r i m e n t s t h a t in different ways have contributed to u n d e r s t a n d i n g the interaction processes close to the sea surface. Adequate p a r a m e t e r i z a t i o n of t h e s e exchange processes depends on good-quality observations of the small-scale processes just above the sea surface under a variety of meteorological conditions. Because of the technical problems connected with the study of air-sea exchange processes expertise is restricted to a few groups. The facilities in the Netherlands are an important factor for maintaining this expertise. Support of the use of the Noordwijk platform for solving the different air-sea interaction problems in the field of climate research is a recommendable course. 3.
B I O G E O C H E M I S T R Y OF T H E O C E A N S
3.1 I n t r o d u c t i o n Reducing the error-bars in the global CO 2 b u d g e t needs a co-ordinated oceanographic research programme. In JGOFS such an effort is u n d e r t a k e n at international level. Also the role of the ocean in the production of N20 and DMS is believed to be i m p o r t a n t . The expectation t h a t complex biogeochemical interactions with physical conditions can be an important factor t h a t regulates climatic change makes a better u n d e r s t a n d i n g of the biogeochemical system necessary. An approach at such a better understanding for one (model) organism (Emilianya huxleyi) is undertaken in the GEM.
The oceans are an important reservoir for C02 that can act as a source or a sink for the atmosphere. The budget is evaluated in terms of the carbon flows and storage. Different fluxes and reservoirs can be defined in this budget. The exchange of carbon (as CO2) is with the surface layer, and between the surface layer and the
306 deeper parts of the ocean direct transport of dissolved inorganic carbon takes place by a localised process of deep convection downward and a more ocean-wide process of upward movement across the permanent thermocline. In addition, however, biogeochemical processes play a role ("the biological carbon pump") by the production of organic carbon (dissolved and particulate) and of particulate calcium carbonate. The downward transport of dissolved organic carbon appears negligible compared with that of the inorganic carbon. However, the biological processes are important by their production of particulate material. This particulate carbon sinks down, the organic carbon to be remineralised in the deep ocean, the calcium carbonate to be partly dissolved, but partly deposited on the ocean floor. These processes should have their place in models of the carbon cycle. Estimates of the uptake of carbon by the oceans of the increasing CO2 burden recently have been reviewed by Siegenthaler and Sarmiento (1993). Using more or less complex box models with their mutual exchanges (recently also GCM models) and suitable tracers (radiocarbon) estimates of the net oceanic uptake of carbon result in about 2 GtC/y. However, the error bars of these estimates are comparatively large. They depend on the models for the global carbon cycle discussed above. Improvements can come from better and possibly more detailed models and from better constraints, such as regional differences in ocean/atmosphere fluxes. Presently the estimates of these fluxes rely on a restricted data-set for the CO2 partial pressure difference atmosphere/water and there is much activity to improve this by the present programme of observations. Also observations of the atmospheric oxygen content can give additional means to constrain seasonal models (Keeling and Shertz, 1992). These models should form the starting point for longer-term budgets. Once a model is obtained, it is possible to estimate the development of the atmospheric CO2 concentration with time for different scenarios, as done in the IPCC report (Watson et al., 1990). A further application is to investigate the possibility of feedbacks. This asks for more details in the parameterization. For instance, a long-term gas exchange coefficient, relating the flux and the partial pressure difference may be wrong for changed climatic conditions. To assess this, one has to turn to the basic micrometeorological processes, that presently give values for the exchange coefficient differing considerably from the coefficient presently used. Although the difference between short-term and long-term values may offer (partly) an explanation for this, a further investigation of this problem appears appropriate. For the scenario-studies the role of biota on the carbon budget is not explicitly taken into account. It is assumed that changes in different marine biota do not critically depend on changed carbon concentrations, but rather on nutrients (nitrogen, phosphorus and silicium), light and the physical environment (although this assumption remains under critical scrutiny). However, this being so, the changes in climate may result in changes in biota and this can result in feedbacks that could be important, also because in a biological system different opposing feedback mechanisms could occur.
307 The task to solve the most important problems of the biogeochemical role of the oceans in the climate system is considerable. JGOFS (Joint Global Ocean Flux Study) is the m a i n p r o g r a m m e to solve these problems by an i n t e r n a t i o n a l approach. JGOFS has the goals (Anon., 1990): 1. To determine and understand the processes controlling the time-varying fluxes of carbon and associated biogenic elements in the ocean, and to evaluate the related exchanges with the atmosphere, sea floor, and continental boundaries. 2. To develop a capability to predict on a global scale the response of oceanic biogeochemical processes to anthropogenic perturbations, in particular those related to climate change. For this J G O F S is concerned with the development of models and with an observational programme, consisting of a mixture of global surveys and process studies, m e a s u r e m e n t s of fluxes between atmosphere and ocean, with benthic s e d i m e n t s , and b e t w e e n the oceans and coastal regions. In a d d i t i o n palaeooceanographic studies have to contribute to the understanding of the carbon cycle in the geologic past. Netherlands scientists have been involved in J G O F S right from the beginning (De Baar et al, 1989). The relation between the carbon cycle in the oceans and the production of biogenic products closely links the different biogeochemical studies for N 2 0 and DMS with the JGOFS programme. A complicating factor is t h a t the uptake of CO2 and the production of other greenhouse gases and DMS are the result of concurring biogeochemical processes and t h a t the m u t u a l relation between these processes can be fairly different for different organisms. This indicates t h a t there are limitations to large-scale modelling that is not taking into account such idiosyncrasies. Therefore an other, in principle complementary approach is t h a t of the Global Emiliania Modelling Initiative (GEM) a co-operative programme of different scientists involved in studies of the pelagic alga Emiliania huxleyi. GEM concentrates on the biogeochemical role of this one significant organism, with the objective to evaluate its present and past climatic relations and its role in the different chemical budgets. The GEM research strategy is that of process studies and modelling at different scales, from the biomolecular to the global level (Westbroek et al., 1993)
3.2 Chemical exchange ocean-atmosphere The chemical exchange between ocean and atmosphere is driven by the partial pressure difference of the gases near the sea surface and in the atmosphere. For assessment of the chemical fluxes more data are needed on a world-wide scale, but it is also important that the exchange processes are better understood. The data, as constraint on the modelling of variations of atmospheric greenhouse gases, can give better estimates of the oceanic fluxes. Direct (micrometeorological) measurement of the exchange of chemicals between ocean and a t m o s p h e r e is difficult, and certainly not feasible as a routine observation. The estimates of chemical fluxes depend on parameterizations. From
308 the difference of partial pressure, near the sea surface and in the overlaying atmosphere, Dp, the flux is estimated, using a transfer velocity Kg. This parameter appears to depend on the wind speed, and on the specific mass of the gas exchanged, expressed in a Schmidt number. Two often used parameterizations are those of Liss and Merlivat (1986) and of Wanninkhof (1992). The uncertainty in these parameterizations is about a factor 2. There are considerable discrepancies between large-scale fluxes of C02 based on geochemical modelling and estimates of fluxes from the few direct measurements. (Broecker et al., 1985; Smith and Jones, 1986; Wesely, 1986). This brings to the questions whether some of the implicit assumptions in the models are applicable, and about the accuracy and generalisation of direct measurements, and illustrates the need for systematic measurements of ApCO2 and for reliable values of Kg. In the programme (NRP project T 850021) "C02 exchange ocean/atmosphere" (H.J.W. De Baar, Netherlands Institute for Sea Research - NIOZ- Texel), later in VvA-9 continued as "Carbon budget in the oceanic mixed layer and air/sea exchange" measurements are made on transects across the (South) Atlantic and Antarctic waters. They constitute an important amplification of the existing datasets. Together with other data these observations can improve the estimate of the exchange of CO2 over those sea areas (Bakker and De Baar, 1992). First results indicate that these areas are an important sink for CO2. The methodology for measuring the partial pressure (equilibrator system) is important also for the measurement of partial pressure of other greenhouse gases (see Section 3.2). The parameter Kg stands for a complex exchange process in which meterological conditions (especially the wind) and the state of the sea play their role. In fact the problems that are encountered in estimating the fluxes of heat, moisture and momentum, discussed in Section 2.3, all are met again in estimating the chemical fluxes. The value of Kg for different gases and its dependence on the external conditions are therefore still under dispute. For other greenhouse gases the problems are the same. For the N20 emission from the sea (see "Water-atmosphere exchange of N20 in marine systems, NRP project 850027) the estimates are also based on ApN20 m e a s u r e m e n t s and a Kg parameter. The dependence of this parameter on the wind is a critical factor in upwelling areas where N20 concentrations are high and winds are strong. For DMS flux measurements were made in enclosures, in the framework of the NRP project "Formation and air/sea exchange of DMS from marine sources" (NRP project 850026). These measurements indicate agreement with theoretical values of the fluxes for lower wind speeds, but larger fluxes than theoretical at stronger winds. The flux-measuring programme (NRP project 852082) "ASGASEX" = Air Sea GAS EXchange (W.A. Oost, Royal Netherlands Meteorlological Institute, KNMI, De Bilt) is designed to improve the parameterization as described above. In the Dutch r e s e a r c h p l a t f o r m "Meetpost Noordwijk" facilities are a v a i l a b l e for
309 micrometeorological flux observations, and they serve not only to investigate the physical (see Section 2.3), but also the chemical fluxes. Different research groups from various countries are working together, performing measurements of a broad scala of micrometeorological and oceanographic measurements. Via the eddy-correlation technique measurements are made of the C 0 2 fluxes. A large difficulty here is the accurate measurement of the fluctuating CO2 signal. Comparison of estimates from the ocean-atmosphere pressure difference with the results of the eddy-correlation measurements indicates that the thickness of the layer over which the pressure difference ApCO 2 (and on which the Kg parameter is applied) is r a t h e r critical. A similar point is the evaluation of the skin effect discussed by Robertson and Watson (1992). An other micrometeorological m e t h o d is to m e a s u r e the g r a d i e n t of gas concentration in the layer of air directly above the water. Combination with the projects NRP 850021 and NRP project 850027 (for CO2 and N20 respectively, see above and Section 3.2) gives the possibility to apply this method and to compare different estimates. Also for DMS, be it not on occasion of ASGASEX the gradient method was attempted. Assessment Systematic collecting of ApCO2 data in the framework of JGOFS is an important point in the evaluation of the greenhouse-gas problem, and the NRP project 850021 forms a direct contribution to this effort. Results of this project are interesting in highlighting the importance of the southern ocean in the CO2 budget. For future modelling it will become i m p o r t a n t to obtain better insight in the physics of the exchange processes, but accurate data are difficult to obtain. By supporting the ASGASEX measurements (NRP project 852082) NRP contributes to developing techniques for direct flux measurements, using facilities t h a t are very suitable for this purpose. The co-ordination with the projects for CO2 and N20 (NRP project 850021 and NRP project 850027) provides a basis for valuable results. In case of further exchange measurements also co-ordination with possible future DMS exchange studies is to be considered. 3.3 G r e e n h o u s e g a s e s a n d D M S
The role of the oceans as a sink of anthropogenic CO 2 is one of the most important issues in the oceanographic climate research. The J G O F S p r o g r a m m e t h a t focuses on this problem from the beginning received substantial N e t h e r l a n d s support. The role of biological systems (especially of the "biological pumps") is studied in JGOFS by exploring different "biogeochemical provinces" in the ocean. Biological processes also are responsible for the role of the oceans as a source of N20 and DMS. No special international programmes have been organised for these items. But the N20 programme is closely linked with JGOFS, while for the DMS research at national level a programme has been organised within NRP t h a t covers most aspects of its production.
310
The C02 system At short terms only the uptake of excess CO2 in the upper layers of the ocean is significant. But over longer periods the transport of excess of dissolved carbon to sub-surface layers becomes important. This transport, by convection and by the action of the m a r i n e biota t h a t act as a carbon pump t r a n s f e r r i n g particulate carbon to deeper w a t e r s cannot be neglected in a balanced approach also considering other possible variations and long-term feed-backs. The role of the biological "pumps" is an important JGOFS issue. Various factors: nutrient concentrations, light, water stratification and zooplankton grazing are the classical agents regulating these pumps (although further studies on other factors as availability of iron, concentration of CO2 still are required). These factors again, including n u t r i e n t concentration, are determined by physical processes (e.g. upwelling, watermasses, etc.). In JGOFS different "biogeochemical provinces" are defined that primarily depend on the physical forcing functions. It will be clear that one of the possible feedback mechanisms is a shifting of these provinces. Therefore process studies are planned for these different provinces on which models are to be based of the shorter-term developments. At the other hand remote sensing has to expand the local and regional results to global scales, together with a programme of large-scale surveys from ships of opportunity, from fixed stations and by joining WOCE campaigns, etc. Furthermore a benthic programme should link the J G O F S results with palaeo-oceanographic work. Right from the beginning the Netherlands have played an active role in JGOFS. Since the s t a r t of the NRP and VvA programmes additional work in the framework of JGOFS has been supported. The N e t h e r l a n d s VvA-9 programme ("Carbon budget in the oceanic mixed layer and air/sea exchange") is now the core of the Netherlands participation in JGOFS. Yet a brief survey over the JGOFS participation should not be limited to NRP and VvA alone. In 1989 the first North Atlantic Pilot Study took place, and since a programme started in the Southern Oceans and in the Indian Ocean. These programmes were supported by the Marine Science Foundation (SOZ) of t h a t time. Below they are discussed below separately. The regular observations of CO2 partial pressure in the South Atlantic t h a t now are p a r t of the VvA programme (but t h a t started as NRP project 850021) have already been discussed in Section 3.3. Their contribution to an assessment of the role of the southern oceans to the global carbon budget is an important element of the JGOFS implementation. Process studies aim at the evaluation of parameters t h a t are required in models for the biogeochemical state of the ocean. We note t h a t within the Emiliania huxleyi project of Section 3.3 (NRP project 850003) a number of studies is made t h a t are directly relevant as JGOFS process studies. Two of the four processes of t r a n s p o r t of carbon from the surface waters to the deep sea, the processes of settling of particulate organic m a t t e r (1) and of calcium carbonate (2) are being investigated in the E. huxleyi project. To the study of the downward transport by the ocean circulation of dissolved CO2 (3) and organic carbon (4) by the ocean circulation contributions are made in the VvA-9 programme.
311 Actual observations made with sediment traps at different levels (Van Hinte and co-workers) produce figures of the fluxes of particulate carbon (1 and 2) and of the transformation processes in- and outside the euphotic zone. Sediment cores relate the actual fluxes with conditions in the geologic past. Special attention is given to the transport of dissolved organic carbon in the VvA-9 programme "Distribution, chemical composition and isotopic ratios of dissolved organic carbon in the ocean". The methodology for the c o n c e n t r a t i o n measurements of organic carbon has been investigated by De Baar et al. (1993). The values for dissolved organic carbon and dissolved organic nitrogen appear to be much higher than previously measured. For the organic carbon found in oceanic sediments an important question is their provenance: are they of marine or terrestrial origine. In the framework of this VvA programme this is investigated by means of certain biomarkers. Also in the framework of the VvA-9 programme a study is undertaken on "The partial pressure of CO2 in seawater as a control on marine biological productivity and calcification". This study investigates in how far the principle that in the sea (in contrast to conditions on land) CO2 partial pressure should be unimportant in regulating the phytoplankton growth (see above) is to be amended. Also this programme is directly related to the E. huxleyi project (Section 3.3) because of the opposing effects by photosynthesis and calcification on the CO2 partial pressure accompanying E.huxleyi blooms.
The JGOFS North Atlantic studies. The international JGOFS Pilot Study in 1989 had as scientific objective the understanding of the evolution in time of the spring bloom: the mechanisms for its generation, maintenance and decay. The main part was by multiple-ship (Canada, FRG, Netherlands, UK, USA) observations along a 20~ section between March and October. An important aspect of the North Atlantic is the transport of water to the source areas for deep convection, and thus for t r a n s p o r t of dissolved carbon to the deep ocean. Also d u r i n g the DUTCH-WARP programme (see Section 2.2) additional observations were made that resulted in estimates of the transport of dissolved anorganic carbon in the Iceland Basin. Plans for future North Atlantic JGOFS programmes are being considered. The Netherlands participation to this programme, with support from the SOZ (Marine Science Foundation) of that time, has resulted in a number of publications (De Baar et al., 1993, Veldhuis, Kraay and Gieskes, 1992, Stoll et al., 1993, Stoll, 1994.
The J G O F S I n d i a n Ocean studies. The n o r t h w e s t I n d i a n Ocean is a biogeochemically i m p o r t a n t area because of the strong seasonal monsoon variations. In the framework of the Netherlands Indian Ocean Expedition with the RV Tyro of 1992/93 (organized by the SOZ-Marine Science Foundation) a number of exploratory studies was made for the JGOFS studies that are planned from 1994 onwards. In Section 3.2 results of N20 and methane obsevations from this programme are discussed.
312 The J G O F S Southern Ocean studies. This programme started in 1992 with cruises of the RV Polarstern (FRG) and Discovery and James Clarke Ross (both UK) to be followed in 1993 by cruises of the RV Marion Dufresne (France). Netherlands scientists participate in these expeditions, with the objective to quantify the spring plankton bloom. This work continues earlier work in the framework of the EPOS (European Polarstern Study) of 1988 (Lancelot et al., 1991, Lancelot, Veth and Mathot, 1991).
Recently it was found that dissolved iron is crucial in developing plankton blooms in Antarctic waters and thus cause CO2 undersaturation of surface waters (De Baar et al., 1995).
Other greenhouse gases The study of the production of other greenhouse gases by marine biota logically is closely connected with the JGOFS programme. Estimates of the oceanic input as a rule are based on the principles discussed in Section 3.1. The available data are subject to improvements of the parametrization of the sea-atmosphere exchange. The most important oceanic contribution of the other greenhouse gases (excluding water vapour) is nitrous oxyde. In the IPCC report (Watson et al., 1990) an estimate is given of the oceanic input of 1.4 - 2.6 Tg N per year, on the basis of an estimated gas exchange coefficient and available data of the difference in partial pressure near the sea surface and in the atmosphere. But we note that regional and temporal variability in the oceanic concentrations make an estimate difficult. Two microbiological processes produce N20 in the open oceans: nitrification by oxydation of ammo n i um in an aerobic environment, and denitrification by reduction of nitrate and nitrite in an anaerobic environment. The former conditions occur in large parts of the oceans, but most intensively in upwelling regions, the latter occur only in special, often restricted regions. Upwelling and the occurrence of anaerobic conditions in the ocean are very sensitive to climatic variations. This indicates the possibility of feedback between N20 input from the oceans and climate change. Improving our estimates of present production and our insight in possible feedback mechanisms asks for the study of N20 production in different environments. In the NRP project 850027 programme "Water/atmosphere exchange of N20 in marine systems" (W. Helder, NIOZ) observations of N20 are made together with other r e l e v a n t oceanographic observations (hydrographic s t r u c t u r e , oxygen concentration) in different marine environments. These include the Somali basin (upwelling region) and the northern Arabian Sea (low-oxygen conditions), both in the Indian Ocean, and in different locations in the North Sea and Netherlands coastal waters. Earlier findings that indicate the NW Indian Ocean as a significant source of N20 (Law and Owens, 1990) could be confirmed. Upwelling waters were found to be important sources. Coastal waters (Scheldt estuary) can be an important source in local N20 budgets. The input into the atmosphere from the North Sea does not differ from the global average.
313 Parallel to the N20 observations observations of methane are made. According to IPCC estimates (Watson et al, 1.c.) the oceans "are thought to be a minor source of atmospheric CH4. However, the estimated flux of CH4 from the oceans is based on a limited data set, taken in the late sixties/early seventies when the atmospheric concentration of CH4 was about 20% lower. There are inadequate recent data from either the open oceans or coastal w a t e r s to reduce the u n c e r t a i n t y in these estimates". For this reason the newer CH4 data give an opportunity to check the IPCC estimates. DMS The role of oceanic-produced DMS in the climate system is t h a t it is a source of cloud condensation nuclei (CCN) and so m a y influence the cloudiness. This especially holds for conditions over the oceans: the atmospheric lifetime of most sulfur compounds is limited, and over the oceans the role of other sources of CCN is less important. The different aspects of changes in the concentration of CCN on the formation of clouds and on climate are discussed elsewhere (See: Guicherit, A s s e s s m e n t Report subtheme "Atmospheric processes & UV-B radiation", this volume). Here we discuss the production of DMS in surface w a t e r s by m a r i n e plankton, its exchange with the atmosphere, and the question w h e t h e r these processes can be at the basis of a feedback mechanism as suggested by Charlson et al., 1987). E s t i m a t e s of the input of DMS (as Tg S per year) is 10 - 50 (Table A1.10 in Watson et al., 1992). Because of the large variations of concentration of DMS in place and time these estimates are r a t h e r uncertain. The concentrations in the ocean are related to the variable biological activity, and the process of formation d e p e n d s on the d o m i n a n t p h y t o p l a n k t o n species and on the chemical transformations in the water. This means strong variations of DMS production in time. Also the relationship between dissolved DMS and chlorophyll concentration is not unique, and the possibility to make concentration estimates of DMS in w a t e r by means of remote sensing of chlorophyll still has its uncertainties. If by future RS methods it might be possible to discriminate between various organisms, and when we have gained further understanding of their production of DMS, this might change. The production of DMS by different algae and at different stages of their growth cycle is a first step in the assessment of their role in the process. A review of this subject has appeared by Liss, Malin and Turner (1993). In the project "Formation and air/sea exchange of DMS from marine sources" of Guicherit (NRP project 850026) the objectives are to assess (1) the processes and rates of production of DMS and its precursor DMSP, and (2) the processes of their degradation and transformation in the water column and by sediments, and (3) to e s t i m a t e the fluxes to the atmosphere (Guicherit, 1993). This project is also assessed in the framework of the subtheme "The Climate System". The work is being done in cultures, boxcosms and enclosures, but also field m e a s u r e m e n t s are made to be compared with production models. As Emiliania huxleyi is an important producer of DMS there is a link with the work described in Section 3.3. However, the work also is directed to Phaeocystis, also an important producer of DMS, and frequently blooming in Dutch coastal waters. As both
314 organisms play a different role in the carbon fixation and concurrent DMS production, their climatological significance is expected to be different. As the project has shown, only a fraction of the in-situ production of DMS(P) enters the atmosphere, in fact less than thought before. The production is mainly during blooms and in view of the rather short atmospheric lifetime of sulfur compounds, DMS presently might be more a factor of regional than of global climatic significance. However, in the conclusions of the assessment for "The Atmosphere" it is suggested that: "... minor changes in the bacterial and phytoplankton population density or composition could have pronounced effects on the global S budget, with all of its consequences". There appear to be reasons for further study of the DMS problem in the context of climate change. Assessment The different projects in this group (NRP project 850021, NRP project 850027 and NRP project 850026) are closely connected, and also results of the E.huxleyi project (NRP project 850003) can be considered against the background of the study of the systems of carbon and DMS in the sea. They contribute in different ways to the Netherlands JGOFS participation. The methodology for measuring the concentrations of dissolved gases in (sea)water with the equilibrator technique could be implemented via NRP project 850021 and NRP project 850027 and is available now for continued work. Combination of the different programmes resulted in common data sets, important for the study of the different systems.
The work in NRP project 850027 produced new data for the study of the production of N20 by different marine environments. Values for the transport of particulate carbon to the deep sea are being obtained in NRP project 850003 as the result of the E.huxleyi programme, but they are of direct relevance to JGOFS. Apart from the judgements given for NRP project 850026 in the assessment report by Guicherit, we may state that here the study of the DMS system from its production at microbiological level to its role in the formation of cloud condensation nuclei has brought together quite different research groups.
3.4 The phytoplankton system Studying the biogeochemical interaction with climate of one organism is an approach that is different from the JGOFS approach, although in actual research there is a considerable overlap. In an international group of investigators E.huxleyi was chosen as the object of this approach (Global Emiliania Modelling Initiative). This organism is important both for the CO2 exchange and for DMS production. The biochemical investigation of the role of the oceans in the budget of CO2 and some other greenhouse gases and as a source for DMS primarily is of interest in an evaluation of the present-day conditions and in the assessment of their development for reduction policies. But also these investigations will provide more insight in the possibility of feedback mechanisms that can be expected and that, judging from palaeoclimatological evidence, appear to have acted in the past. The JGOFS approach, as descrbed above, is to establish global values and global relations by considering large-scale biogeochemical provinces. To what extend the critical feedback mechanisms thus can be described adequately remains a question that only can be answered in the long run.
315 A different approach is an organism-centered approach. The NRP-supported project 850003 "Phytoplankton and the oceanic carbon cycle; Emiliania huxleyi as a model system" (Westbroek) takes this approach. The philosophy is that by considering separate phytoplanktonic organisms that play an important role in the biochemical transformations the interactions can be evaluated better. This philosophy is tested on the coccolithophore Emiliania huxleyi in the co-ordinated studies of the GEM ("Global E.huxleyi Modelling Initiative"). This organism contributes significantly to the biological pump by producing coccoliths, small calcium carbonate shells that during blooms can occur in huge quantitites that can be observed on remote-sensing images. At the other side it is an important producer of DMS (Westbroek, 1992, Westbroek et al., 1993). The programme consists of studies of the biochemistry at cellular level, experiments in the laboratory and in enclosures, and open-ocean studies. Remote sensing and geological observations add a global and palaeoclimatological dimension. The study aims at modelling of the system at various integration levels (cellular, population, global levels). By the choice of E. huxleyi an organism is selected that has a wide distribution and has existed for a long geological period. Results of the programme therefore are applicable not only under actual, but also under palaeoclimatological conditions. Further palaeoclimatic implications of this work come from the discovery that long-chain alkenones found in marine sediments and that are likely to come from E. huxleyi, reflect the growth temperature of the organism. This programme and the studies of the C02 system and of the DMS production are interlinked: coccolithophores rank high as DMS producers. Results of this programme can be considered, together with other results, in the relevant earlier Sections 3.2 and 3.2. The organism-centered approach discussed here should reveal the important interconnections between different parts of the carbon budget, the DMS production, and possible other relevant factors. Emiliania huxleyi is chosen to represent calcifying plankton in general. These organisms have their typical role in climate-related processes. Furthermore they are special in that their growth may be limited by the inorganic C supply in the surface waters. Blooms of E. huxleyi can be observed by remote sensing, and appear to have mainly an effect on a regional scale, compared with standing-stock estimates (Brown and Yoder, 1994). They exhibit a special succession in the various carbon fluxes. Their net effect is likely to be strongly non-linear. How to parameterize the standing-stock and bloom effects further can be clarified by the E. huxleyi project. The production of DMS and the carbon cycle are biologically related. For typical organisms this relation has to be established. Summarizing: The effect of these studies is twofold: First the dynamism of this system can be better understood. Also the methodology of this approach probably can be applied to similar studies of other crucial organisms. The biological detail obtained in this way can result in better parameterization of the more general biogeochemical models. Assessment
A separate assessment of the NRP project 850003 should consider the success of the underlaying concept, that is: studying the different interactions of one
316 organism with its environment. At this moment this is not possible. Immediate questions, such as: "how accurate are our present views of the carbon budget in the oceans ?", are only addressed indirectly. But specific results obtained in this project, are contributing to JGOFS (such as the data on downward transport of particulate carbon, have been mentioned under Section 3.2). At the other hand, in VvA-9 work of direct relevance to this project (see above) is continued as a JGOFS activity. The o r g a n i s m - c e n t e r e d approach is i m p o r t a n t for focussing on complex interactions t h a t may receive less attention when more immediate goals as the budgets of geenhouse gases are the issue. The choice of coccolithophores for this approach also provides a link with palaeoclimatic studies. 4.
CONCLUSIONS
The review given above describes the role of the oceans in the problem of climatic change, the international appreciation concerning the most important scientific issues and the Netherlands activities in these fields, with emphasis on the projects supported by the Dutch National Research Program on Global Air Pollution and Climate Change (NRP). The Netherlands oceanographic research that is discussed here addresses selected areas from a large scientific and societal problem. Scientifically, integration of the results takes place at three levels: first there is the integration at the level of the specific subject, the physical or biogeochemical processes, and the methodology of model development. Then there is the integration at modelling level: observations, processes and modelling methodology converge in model-simulations of climatic change. This is the aim of the large research projects as TOGA, WOCE and JGOFS. Finally the results of different simulations are integrated at IPCC level to obtain a scientifically broadly accepted opinion on the climatic change problem. In this review the emphasis is on the description of the different NRP projects in relation to the rest of the Netherlands and international research of the role of the oceans in climate. The quality of this work only can be judged in very general terms. Individual principal scientists have to report their results separately in the appropriate scientific communities. Firm conclusions concerning the ultimate contribution of specific oceanographic NRP projects to a better understanding of the climate problem cannot be given at this place. The importance for climate research of a strong international oceanographic research effort is evident. Netherlands oceanographers contribute in different fields and are supported by different sources in this effort. The NRP projects t h a t have been evaluated here form a substantial part of this. Many of them have not yet been finished: either the final scientific reporting has to be done, or the project has not yet been concluded and is continued in another form. In the NRP evaluation by SPA/HCG (1992) critical r e m a r k s are made on the co-ordination within this theme. It has to be said that the theme "water" is a heterogeneous one. There is certainly co-ordination at project level. J G O F S
317 observations have been made at WOCE cruises and vice-versa. The ASGASEX programme offers facilities for investigators of the ocean-atmosphere exchange of CO2 and N20. The studies of the CO2 budget, the work on E.huxleyi and on DMS production to some extent are intermingled. But generally the level of interaction between biogeochemistry and physics is low. Unless there is an explicit wish for more interaction, the need for future combined programming of both sectors can be questioned. F u r t h e r internal co-ordination within the sectors "physics" and "biogeochemistry" can, however, be profitable. The above review is not dealing with the IMAGE model, that is being developed in NRP as a tool for integrated assessment of climate change. We note, however, t h a t also in the IMAGE 2.0 effort a modelling of the coupled ocean and atmospheric energy balances is undertaken (Anon., 1993). For the ocean physics a "quasi 2-D" model is applied. In the IMAGE model an attempt is made to evaluate the effect of various, mainly biogeochemical, processes on the storage of anthropogenic carbon (Klepper et al., 1993). These results suggest that over one century predictions of atmospheric CO2 concentrations can be made without consideration of complex biogeochemical feedbacks. Question is how serious one has to take such ideas, considering the large efforts within and outside NRP to understand better the climatic role of the oceanic system. It is clear that the application of the oceanographic module of IMAGE could gain from exchange of ideas and results with the research groups working in the field, and that this could give weight to future judgements made on the basis of IMAGE. 5.
ACKNOWLEDGEMENTS
The present assessment gives the opinions of the theme co-ordinator with respect to the different projects. However, his ideas only could be developed thanks to the discussions with the investigators and especially with H.J.W. De Baar, G.J. Komen and W.P.M. De Ruijter. 6.
REFERENCES
Anon., 1987. WOCE Core Project 2 Planning Meeting. The Southern Ocean Bremerhaven, F.R.G., may 1986. WMOfrD 181 (WOCE Report 09/87), 53 pp. Anon., 1990. The Joint Global Ocean Flux Study, JGOFS, Science plan. JGOFS rep. 5. SCOR, 61 pp. Anon., 1991. The oceans and climate: A guide to present needs. IOC Technical Series 38. UNESCO 1991.39 pp. Anon., 1992. Oceanic interdecadal climate variability. IOC Technical Series 40. UNESCO 1992.40 pp. Anon., 1993. Report International Review Meeting IMAGE 2.0. Dutch National Research Programme on Global Air Pollution and Climate Change. Rep. 00-09. Bakker, D., and H.J.W. De Baar, 1992. Exchange of CO2 between the oceans and the atmosphere. Change 10: 3 - 4 .
318 Bretherton, F.P., K. Bryan and J.D. Woods, 1990. Time-dependent greenhouse-gasinduced climate change. In: Houghton, J.T.G.J. Jenkins and J.J. Ephraums, eds.). Climate Change. The IPCC Scientific Assessment. Cambridge Univ. Press, Cambridge, p. 173-194. Broecker, W.S., J.R. Ledwell, T. Takahashi, R. Weiss, L. Merlivat, L. Memery, T-H. Peng, B. Jahne, K.O. Miinnich, 1986. Isotopic versus micrometeorologic ocean CO2 fluxes: a serious conflict. J. Geophys. Res. 91: 10517-10527. Broecker, W.S., 1987. Unpleasant surprises in the greenhouse? Nature 328: 123-126. Broecker, W.S., 1991. The great ocean conveyor. Oceanography 4: 79-89. Brown, C.W., and J.A. Yoder, 1994. Coccolithophorid blooms in the global ocean. J. Geophys. Res. 99: 7467-7482. Charlson, R.J., J.E. Lovelock, M.O. Andreae and S.G. Warren, 1987. Oceanic phytoplankton, atmospheric sulphur, cloud albedo and climate. Nature 326: 655-661. Danabasoglu, G., J.C. McWilliams and P.R. Gent, 1994. The role of mesoscale tracer transports in the global ocean circulation. Science 264: 1123-1126. Dansgaard, W., S.J. ~I~ohnsen, H.B. Clausen, D. Dahl-Jensen, N.S. Gundestrup, C.U. Hammer, (~.S. Hvldberg, J.P. Steffensen, A.E. SveinbjSrnsdottir, J. Jouzel and G. Bond, 1993. Evidence for general instability of past climate from a 250-kyr ice-core record. Nature 364: 218-220. De Baar, H.J.W., H.G. Fransz, G.M. Ganssen, W.W.C. Gieskes, W.G. Mook, J.H. Stel, 1989. Towards a Joint Global Ocean Flux Study: Rationale and objectives. In: Ayala-Castafiares, A., W. Wooster and A. Y~ifiez-Arancibia, eds., Oceanography 1988. UNAM Press, M~xico D.F., 11: 34. De Baar, H.J.W., and M.H.C. Stoll, 1989. Storage of CarbonDioxide in the Oceans. In: P.A. Okken, S.J. Zwart, S. Zwerver, eds. Climate and Energy: the feasibility of controlling CO2 emissions Kluwer Ac.Publ., Dordrecht, p. 143-177. De Baar, H.J.W., 1992. Options for enhancing the storage of Carbon Dioxide in the oceans: a review. Energy Convers. Mgmt., p. 635-642. De Baar, H.J.W., C. Brussaard, J. Hegeman, J. Schijf and M.H.C. Stoll, 1993. Sea-trials of three different methods for measuring non-volatile dissolved organic carbon in seawater during the JGOFS North Atlantic pilot study. Marine Chemistry 41: 145-152. De Baar, H.J.W., J.T.M. De Jong, D.C.E. Bakker, B.M. LSscher, C. Veth, U. Bathmann and V. Smetacek, 1995. Iron is essential for plankton blooms and carbon dioxide drawdown in the Southern Ocean, Nature (in press). De Boer, C.J., H.M. Van Aken and A.J. Van Bennekom, 1991. Hydrographic variability of the overflow water in the Iceland Basin. ICES C.M.1991/C: 13, 11 pp. De Boer, C.J., 1993. Water mass distribution in the Iceland Basin calculated with an optimal parameter analysis. ICES C.M.1993/C: 16, 18 pp. De Bruin, T. F., R.X. De Koster, S. Ober and L. Otto, 1992. Netherlands XBT programme along the WOCE AX-5 section. NIOZ Data-report 1992-3, 40 pp. De Ruijter, W.P.M., P.J. Van Leeuwen and J.R.E. Lutjeharms, submitted. Triggering mechanisms for Natal Pulses in the Agulhas Current. Subm.to J.Geophys.Res. Dickson, R.R., J. Meincke, S.A. Malmberg and A.J. Lee, 1988. The "Great Salinity Anomaly" in the northern North Atlantic 1968-1982. Prog. Oceanogr., 20: 103-151.
319 Drijfhout, S.S.,1990. Ring genesis and the related transport of heat, momentum and vorticity: A parameter study. J.Phys.Oceanogr. 20: 1645-1665. Drijfhout, S.S., 1992. Ring genesis and the related heat transport. Part II: A model comparison. J. Phys. Oceanogr. 22: 268-285. Drijfhout, S.S., 1994. Heat transport by mesoscale eddies in an ocean circulation model. J. Phys. Oceanogr. 24: 353-369. Dijkstra, H.A., and M.J. Molemaker, 1994. Double diffusive and direct instabilities below growing sea-ice. Int. J. Heat Mass Transfer, 37:2547-2559 Dijkstra, H.A., and E.J. Kranenborg, 1995. The structure of (linearly) stable double diffusive flow patterns in a laterally heated stratified liquid. Phys. Fluids (in press). Feron, R.C.V., 1992. Satellite altimetry: its use for monitoring the ocean. Change 11: 6-7. Feron, R.C.V., W.P.M. De Ruijter and D. Oskam, 1992. Ring shedding in the Agulhas Current system. J. Geophys. Res. 97:9467-9477 Feron, R.C.V., 1993. Comparing the UK Fine Resolution Antarctic Model (FRAM) with 3-years of Geosat altimeter data. In: R.Rummel and F.Sanso eds. Satellite altimetry in geodesy and oceanography. Springer Verlag : 467-479. Feron, R.C.V., 1994. Eddies and mean flow in the ocean. Applications of satellite altimetry. Thesis, Univ. Utrecht. 126 pp. Feron, R.C.V., 1995. The Southern Ocean western boundary currents: comparison of FRAM-model results with Geosat-altimeter data. Accepted in: J. Geophys. Res. Fraedrich, K., 1994. An ENSO impact on Europe? A review. Tellus 46A, 541: 541-552. Gates, W.L., P.R. Rowntree and Q-C. Zeng, 1990. Validation of climate models. In: Houghton, J.T., G.J. Jenkins and J.J. Ephraums, eds. Climate Change. The IPCC Scientific Assessment. Cambridge Univ.Press, Cambridge, 1990: 93-130. Gates, W.L., J.F.B. Mitchell, G.J. Boer, U. Cubasch and V.P. Meleshenko, 1992. Climate modelling, climate prediction and model validation. In: J.T. Houghton, B.A. C a l l a n d e r and S.K. Varney, eds. Climate Change 1992. The supplementary report to the IPCC Scientific Assessment. Cambridge Univ. Press, Cambridge, 1992: 97-134. Gill', A.E., 1983. Eddies in relation to climate. In: Robinson, A.R. ed. Eddies in ma~ine science Springer-Verlag, p. 441-445. ~ui'cherit, R:, 1993. Ocean dimethylsulphide (DMS) and climate. Change 16: 7-9. Haagmans, R., M.C. Naeije, and R.C.V. Feron, 1993. TOPEX/POSEIDON and ERS-I: New dimension in satellite altimetry, part 1 & 2. Geodetical Info Magazine 5, nov-dec. 1993. Janssen, P.A.E.M., 1994. Results with a coupled wind wave model. ECMWF Technical Report 71, 60 pp. Jin, F.F., J.D. Neelin, M. Ghil, 1994. E1 Nifio on the devil's staircase: annual subharmonic steps to chaos. Science 264: 70-72. Kattenberg, A., 1993. The upper layer of the ocean. Change 14: 18-19. Keeling, R.F., and S.R. Shertz, 1992. Seasonal and interannual variations in atmospheric oxygen and implications for the global carbon cycle. Nature 358: 723-727.
320 Klepper, O., B.J. De Haan, P. Saager and M.S. Krol, 1993. Oceanic uptake of anthropogenic CO2: mechanisms and modelling. RIVM report 481507004.48 pp. Koblinsky, C.J., P.Gaspar and G. Lagerloef, 1992. The future of spaceborne altimetry. Oceans and climate change. A long-term strategy. Joint Oceanogr. Inst. Incorporated, Washington D.C., 75 pp. Lancelot, C., G. Billen, C. Veth, S. Becquevort and S. Mathot, 1991.Modelling carbon cycling through phytoplankton and microbes in the Scotia-Weddell Sea area during ice retreat. Marine Chemistry 35: 305-324. Lancelot, C., C. Veth and S. Mathot, 1991. Modelling ice-edge phytoplankton bloom in the Scotia-Weddell sea sector of the Southern Ocean during spring 1988. J. Marine Systems 2: 333-346. Latif, M., T. Stockdale, J-O. Wolff, G. Burgers, E. Maier-Reimer, M.M. Junge, K. Arpe, L. Bengtsson, 1994. Climatology and variability in the ECHO coupled GCM. Tellus 46A: 351-366 Law, C.S., and N.J.P. Owens, 1990. Significant flux of atmospheric nitrous oxide from the northwest Indian Ocean. Nature 346: 826-828. Lenderink, G., and R.J. Haarsma, 1994. Variability and multiple equilibria of the thermohaline circulation associated with deep-water formation. J. Phys. Oc. 24: 1480-1493. Levitus, S., 1982. Climatological atlas of the world ocean. NOAA Prof. paper 13, 173 pp. Liss, P.S., and L. Merlivat, 1986. Air-sea gas exchange rates: Introduction and synthesis. In: P. Buat-M6nard, ed. The role of air-sea exchange in geochemical cycling, p. 113-127. Liss, P.S., G. Malin and S.M. Turner, 1993. Production of DMS by marine phytoplankton. In: G. Restelli and G. Angeletti, eds. Dimethylsulphide: Oceans, Atmosphere, and climate ECSC, EEC, EAEC, 1993:1-14 Manabe, S., R.J. Stouffer and M.J. Spelman, 1994. Response of a coupled ocean-atmosphere model to increasing atmospheric carbon dioxide. Ambio 23: 44-49. Marshall, J., J.A. Whitehead and T. Yates, 1994. Laboratory and numerical experiments in oceanic convection. In: Manalotte-Rizzoli and Robinson eds. Ocean processes in climate dynamics: Global and Mediterranean examples. Kluwer Ac. Publ. 1994: 173-201. McBean, G.A., 1994. Global change models - A physical perspective. Ambio 23: 13-18. McPhaden, M.J., 1994. The eleven-year E1 Nifio? Nature 370: 326. Neelin and Dijkstra, 1995. Coupled ocean-atmosphere interaction and the tropical climatology. Part I: The dangers of flux-correction. Part II : Why the cold tongue is in the east. J. Climate (in press). Neelin, J.D., and J. Marotzke, 1994. Representing ocean eddies in climate models. Science 264: 1099-1100. Nowlin, W.D.jr. and J.M. Klinck, 1986. The physics of the Antarctic Circumpolar Current. Rev. Geophysics 24: 469-491. Ocean Observing Development Panel, 1993. Interim design for the ocean component of a Global Climate Observing System. Dept. Oceanogr. Texas A&M Univ. 105 pp. Oost, W.A., C.W. Fairall, J.B. Edson, S.D. Smith, R.J. Anderson, J.A.B. Wills, K.B. Katsaros and J. DeCosmo, 1994. Flow distortion calculations and their
321 application in HEXMAX. Proc. 5th conf. on Meteorlogy and Oceanography in the coastal zone. J. Atmospheric and Oceanic Technology 11: 366-386. Oost, W.A., 1994. Errors in eddy correlation measurements of momentum fluxes and their correction. J. of Marine Systems, 4: 171-181. Otto, L., H.M. Van Aken and R.X. de Koster, 1992. Use of drifters in the "DUTCHWARP" programme, 1990 and 1991. Programme description and preliminary results. NIOZ Data-report 1992-2, 12 pp. Otto, L., and H.M. Van Aken, submitted. Surface circulation in the North-East Atlantic as observed with drifters. Subm. to Progress in Oceanography. Parilla, G., A. Lavin, H. Bryden, M. Garcia and R. Millard, 1994. Rising temperatures in the subtropical North Atlantic over the past 35 years. Nature 369: 48-51. Raynaud, D., J. Jouzel, J.M. Barnola, J. Chappellaz, R.J. Delmas, C. Lorius., 1993.The ice record of greenhouse gases. Science 259: 926-259. Read, J.F., and W.J. Gould, 1992. Cooling and freshening of the subpolar North Atlantic Ocean since the 1960s. Nature 360: 55-57. Schmitt, R. and K. Bryan, 1991. Mysteries of the ocean water budget. Annex IV to Committee on Climatic Changes and the Ocean (CCCO), 12th session Woods Hole, June 1991, 17 pp. Schott, F., M. Visbeck and U. Send, 1994. Open ocean deep convection, Mediterranean and Greenland Seas. In: Manalotte-Rizzoli and Robinson eds. Ocean processes in climate dynamics: Global and Mediterranean examples. Kluwer Ac.Publ. 1994: 203-225. Shine, K.P., R.G. Derwent, D.J. Wuebbles and J.J. Morcrette, 1990. Radiative forcing of climate. In: J.T. Houghton, G.J. Jenkins and J.J. Ephraums, eds. Climate Change. The IPCC Scientific Assessment. Cambridge Univ.Press, Cambridge, 1990: 41-68. Siegenthaler, U., and J.L. Sarmiento, 1993. Atmospheric carbon dioide and the ocean. Nature 365: 119-125. Smith, N., 1991. The role of models in an Ocean Observing System. OOSDP Background Report 1. Joint CCCO-JSC, 85 pp. Smith, S.D., and E.P. Jones, 1986. Isotopic and micrometeorological ocean CO2 fluxes: different time and space scales. J. Geophys. Res. 91: 10529-10532. SPA and HCG, 1992. Evaluation of the technical emphasis, policy relevance and management performance of the Dutch National Research Program on Global Air Pollution and Climate Change. Washington and Amsterdam 66 pp. Sterl, A., and A. Kattenberg, 1993. Study of North Atlantic SST evolution with a coupled OGCM-Mixeld Layer model. In: G.J. Boer ed. Research activities in atmospheric and oceanic modelling WMO-TD 533: 838-839. Sterl, A. and A. Kattenberg, 1994. Embedding a mixed layer model into an OGCM of the Atlantic: the importance of surface mixing for heatflux and temperature. J. Geophys. Res. 99: 14139-14157. Stockdale, T., M. Latif, G. Burgers, J-O. Wolff, 1994. Some sensitivities of a coupled ocean-atmosphere GCM. Tellus, 46A: 367-380 Stoll, M.H.C., J.W. Rommets and H.J.W. De Baar, 1993. Effect of selected calculation routines and dissociation constants on the determination of total carbon dioxide in seawater. Deep-Sea Res. 1 40: 1307-1322. Stoll, M.H.C., 1994. Inorganic carbon behaviour in the North Atlantic Ocean. Thesis, Univ.Groningen. 193 pp.
322 TOGA Scientific Steering Group, 1985. Scientific plan for the Tropical Ocean and Global Atmosphere Programme. WMOfrD 64, 146 pp. Trenberth, K.E. and A. Soloman, 1993. A new look at atmospheric and oceanic poleward heat transports. In: Fourth Int. Conf. on Southern Hemisphere meteorology and oceanography. Hobart, Australia. Am.Met.Soc.Boston USA. Tziperman, E., L.Stone, M.A. Cane, H. Jarosh, 1994.E1 Nifio chaos: overlapping of resonances between the seasonal cycle and the Pacific ocean-atmosphere oscillator. Science 264" 72-74. Van Aken, H.M., 1993. Current measurements in the Iceland Basin. ICES C.M.1993/C: 11, 13 pp. Van Aken, H.M., (submitted) Hydrographic variability in the bottom layer of the Iceland Basin. Subm. to J. Phys. Oc. Van Leeuwen, P.J., and W.P.M. De Ruijter, submitted. Natal pulses and the formation of Agulhas rings. Subm.to J. Geophys. Res. Veldhuis, M.J.W., G.W. Kraay and W.W.C. Gieskes, 1992. Growth and fluorescence characteristics of ultraplankton on a north-south transect in the eastern North Atlantic. Deep-Sea Res. 40: 609-626. Walsteijn, F.H., 1994. Oceans and climate: robust numerical methods for 2D turbulence. J. Comput. Phys. 114: 129-145. Also: Dutch Nat.Res. Progr. on Global Air Pollution and Climate Change. Report 001/project 850025.34 pp. Wanninkhof, R., 1992. Relationship between wind speed and gas exchange over the ocean. J.Geophys. Res. 97: 7373-7382. Watson, R.T., H. Rodhe, H. Oeschger and U. Siegenthaler, 1990. Greenhouse gases and aerosols. J.T. Houghton, G.J. Jenkins and J.J. Ephraums, eds. The IPCC Scientific Assessment. Cambridge Univ. Press, Cambridge, 1990 : 1-40. Watson, R.T., L.G. Meira Filho, E. Sanhueza and A. Janetos 1992. Sources and sinks. In: J.T. Houghton, B.A. Callander and S.K. Varney, eds. Climate Change 1992. The supplementary report to the IPCC Scientific Assessment. Cambridge Univ. Press, Cambridge, 1992: 25-46. Wesely, M.L., 1986. Response to "Isotopic versus micrometeorologic ocean CO2 fluxes: a serious conflict" by W. Broecker et al.. J. Geophys. Res. 91: 10533-10535. Westbroek, P., 1992. Global Emiliania modelling initiative (GEM), an international programme to study the role of life in the global climate. Change 9: 8-10. Westbroek, P., C.W. Brown, J. v. Bleijswijk, C. Brownlee, G.J. Brummer, M. Conte, J. Egge, E. Fernandez, R. Jordan, M. Knappertsbusch, J. Stefels, M. Veldhuis, P. v.d.Wal, J. Young, 1993. A model system approach to biological climate forcing. The example of Emiliania huxleyi. Global and Planetary Change, 8: 27-46. World Climate Data Programme, s.d. Climate System Monitoring. The Global Climate System. A critical review of the climate system during 1982-1984. Unofficial Report. WOCE International Planning Office, 1988. World Ocean Circulation Experiment, Implementation Plan. Vol.1 Detailed Requirements, Vol.2 Scientific Background. WMO/TD 242 & 243 (WOCE Report 20/88 & 21/88). WOCE International Project Office, 1994. WOCE Strategy for Ocean Modelling. WOCE NEG Science Plan for Ocean Modelling 1993. WOCE Report 112/94, 35pp. WOCE Scientific Steering Group, 1986. Scientific Plan for the World Ocean Circulation Experiment. WMO/TD 122 (WOCE Report 06/86), 83 pp.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
325
Global Emiliania Modeling Initiative (GEM) P. Westbroek Leiden Institute of Chemistry, Gorlaeus Laboratories, Leiden University, P.O. Box 9502, 2300 RA Leiden, the Netherlands
Abstract Biological climate forcing is important, but difficult to assess in quantitative terms, as it depends on the ecological and evolutionary success of many different interacting organisms. The Global Emiliania Modeling Initiative (GEM) addresses this problem by focusing on a single key organism, the u n i c e l l u l a r calcifying alga Emiliania huxleyi and by s u b s e q u e n t generalisation to include the climatic effects of larger biological entities, such as the total calcifying plankton. In this paper, the general strategy of GEM is worked out. We believe that our approach may be more widely applicable. 1. THE PROBT,EM
Three major forcing functions are distinguished by which the marine pelatic biota may influence the global climate system [1]: (1) the drawdown of CO2 from the upper mixed water layer and the atmosphere to intermediate and deep waters by the formation, export and remineralisation of particulate organic carbon (POC) (the organic carbon pump); (2) the removal of dissolved inorganic carbon and alkalinity from the upper mixed layer and its partial regeneration in the deep ocean by the formation, sinking and partial dissolution of calcium carbonate (the carbonate pump); and (3) large-scale albedo effects mainly owing to the formation of highly reflecting clouds following gaseous emissions of dimethyl sulphide (DMS) from algal blooms. The climatic effects of these mechanisms vary geographically and through time, as they depend on the abundance and the behaviour of the participating organisms. Many species of algae, protozoa, bacteria, viruses and animals interact with the physico-chemical environment and with each other, and it is from all these interactions that the climatic effects emerge. In view of these complexities, the crucial question is: Can we formulate general rules that adequately describe these interactions and that at the same time are simple enough to be useful for climate modelling? The Global Emiliania Modelling Initiative (GEM)provides a 'model system' approach [2] whereby this fundamental problem is addressed.It assumes that, at all levels of organization, biological systems obey a limited set of fundamental rules. GEM is an attempt to track down some of these rules and to implement them in models of biosphere-climate interactions. Although this 'model system' approach is primarily applied to the calcifying pelagic biota, it is in principle applicable to many other systems [3]. The approach tends to reveal the conserved and widely distributed attributes of living systems involved in biogeochemical cycling and climate forcing, at
326 levels of biological organisation ranging from (macro)molecules to global-scale phenomena. Diversity is considered as variation on this common theme. We therefore expect this methodology to lead to the formulation of stylized and simplified biogeochemical models, well adapted to the resolution of physical and chemical models of ocean dynamics.
(climate) < latent heat
> storms
1
temi~at
~
organiccarbon
carbonate
Is
~ 002
respiration
archive
Figure 1. Overview of the Emiliania huxleyi system
2. G E M - A S Y S T E M ' S A P P R O A C H
The concept underlying the Global Emiliania Modelling Initiative (GEM) is clear and attractive [2]: a single dominant organism, the coccolithophore alga Emiliania huxleyi, is selected as an exemplary model system, much as the bacterium Escherichia coli became a key to the biochemistry of all organisms. Climate forcing by E. huxleyi is studied in depth in an interactive modelling and experimental investigation and the acquired understanding of
327 this system then serves as a key to investigate the wider implications of biological climate forcing. legend ,m~ DMSP CH,O
precursors
,,v
=.=. =.- ~- )lb
CaCO 3
chemical reaction influence experimentaland modelling procedure
phys/chem
environment
Model 1
gene
[1
lisation
'recurs~176
~ellulcrganisationJB
I~MH?(~ pr~ CaCq
/
>,''--" phys/chem environ~nt
1
~ ".
"
, . Moael2
~
DMSo /
c,~o/
~ J
I
/
~1 e~176 cme:]n~ I~ model J
Model 3 calcifyingplanktonecology
'
Ic~ ~
/ I / /
I I
I t~r~
model
"
J
| /
/
|
J Figure 2. GEM modeling of net calcium carbonate, organic carbon and DMS production
Fig. 1 is an overview of the major oceanic and atmospheric processes associated with the E. huxleyi phenomenon. E. huxleyi is a major component of virtually all open marine plankton world-wide, and forms very large blooms, particularly at mid latitudes. The figure shows that in E. huxleyi all three climate forcing functions are coupled, and that the history of this system can be reconstructed from the rich archive at the floor of the deep sea. Major
328 assets of the GEM strategy are the following: (1) it offers a key to the problem of biological diversity, as in-depth information on E. huxleyi may be extended to photic ocean
[
Emi liania] 2
plankton
L3
production
;:i 2ea
calcifying
. . . . . . . . . . . . . . . . . . . . . . .t. . .. .. .. .. .. . . . . . . . . . . . . . . . . . . . . . . . .
archive
|
[~miliania]
calcifying
plankton J~8|Oft core8
paleoclimatology
Figure 3. GEM modeling of (1) net calcium carbonate, organic carbon and DMS production in the photic ocean (see fig. 2), (2) export production into and transformation in the dark ocean of calcium carbonate and organic carbon, and (3) preservation of these materials in the geological archive. include a wide range of other organisms; (2) it highlights the highly nonlinear character of the biological intervention in the climate system; (3) it allows the intimate coupling of the three forcing functions to be modelled; (4) it optimally exploits the unique accessibility of E. huxleyi for laboratory and field research; and (5) the integration of biological and stratigraphic studies permits quantification of carbonate fluxes to the ocean floor as well as extensive model validation by comparing biological responses to diverse climate regimes. 3. G E M S T R A T E G Y I N O U T L I N E A simplified overview of the GEM research strategy is worked out in Fig. 2 and 3 (not all relationships are shown). Fig. 2 shows how a combination of
329 dedicated physiological and molecular genetical research on Emiliania huxleyi leads to a model simulating the CaCO3, CH20 and DMSP output of a single cell, as it depends on physico-chemical environmental conditions (model 1)(NB. DMSP is the precursor of DMS). Laboratory studies additional to the work on E. huxleyi on the major represented organisms lead to a generalised model of CaCO3, CH20 and DMSP productivity (model 2) by the calcifying biota in the pelagic ocean.
I E.calcifying huxleyisplankton ubspeci&es
spatialand temporaldistributionL~~patial and tempora,?,str,ouz,?
prediction
f .....
validation cycle
ut
GEM models
prediction
(palaeo)fluxes and climate forcing
Fig. 4. Validation cycle of GEM models required prior to incorporation into GCM's and estimation of climate forcing by pelagic calcifying biota. By combining model 2 with a global ocean environment model, the potential gross production of CaCO3, CH20 and DMSP by the calcifying plankton in general is predicted. Competing organisms (preferably classified in functional groupings such as phytoplankton taxa thriving under phosphate, nitrate or silica limitation) are known to modify the physicochemical conditions and thus the output of the calcifying plankton. To account for these effects, an extra competition model is included. Furthermore, special models are introduced to implement the effects of consumers on the proliferation of the calcifiers and on the fate of their products ('consumers' and 'product processing' models). These interacting factors are combined in model 3, which predicts the actual global net production of CaCO3, CH20 and DMS by the calcifying biota. Model 3 is validated by laboratory cultures in
330 chemostats, mesocosms in the Norwegian fjords, in situ studies of blooms in the ocean and by remote sensing. Models 1 - 3, which apply to production in the photic ocean, provide an input for export production and transformation models (mainly simulating CaCO3 and biomarker sedimentation and dissolution fluxes in the dark ocean and the top sedimentary layer) and subsequent carbonate palaeoflux (Fig. 3). The latter models allow, vice versa, an interpretation of the geological archive in terms of global productivity of carbonate, organic carbon and DMS by the calcifying biota in the geological past. This geological part of GEM research concentrates on laboratory work of coccolith dissolution and biomarker degradation, as well as sediment trap and core profiling at chosen localities in the N. Atlantic and the Pacific. Comparison of the geological and biological data with model predictions will allow extensive validation of the GEM models (fig. 4). Optimisation of the models will finally lead to increasingly realistic predictions of (paleo)fluxes and climate forcing by the marine calcifying plankton.
4. R E F E R E N C E S P.M. Holligan. In: P.G. Falkowski and A.D. Woodhead (eds.), Primary productivity and biogeochemical cycles in the sea, Plenum, New York (1992) 487-501. P. Westbroek et al. Global and Planetary Change 8 (1993): 1-20 P. Westbroek et al. In: Doumenge, F. (ed.), Past and present biomineralization processes. Considerations about the carbonate cycle. Monaco Mus~e Oc~anographique (1994) num. sp., 13: 37-60.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
331
Evaluating the role of the biological p u m p in the Northeast Atlantic through paleo p r i m a r y productivity reconstruction
Janneke Ottens, Shirley van Kreveld, Gerald Ganssen and Jan E. van Hinte. Center for Marine Earth Sciences, Free University, De Boelelaan 1085, 1081 HV Amsterdam, The Netherlands Abstract
The general objective of the project calcareous plankton as a monitor of a natural C 0 2 experiment during the last glacial-interglacial cycle (Geological subprogram of "Calcareous plankton and the oceanic carbon cycle: Emiliania huxleyi as a model system") is to quantify natural changes in plankton production and its role in the global carbon cycle. More specifically the research concentrates on 1) the reconstruction of long-term carbonate production during the last glacial cycle (150,000 years) at 48~ and 25~ and 2) the determination of regional trends in carbonate production in the Northeast Atlantic during times of rapid climatic change. To estimate past primary productivity using biogenic carbonate accumulation on the seafloor, one needs to quantify the amount of detrital and biogenic carbonate as well as the loss of calcium carbonate by dissolution. We developed a new method to quantify dissolution using recent sediments from a range of oceanic water depths which represent different degrees of dissolution. This can be applied to deep water piston cores to determine temporal variation in carbonate dissolution. Our time series of reconstructed surface ocean productivity of the last 150,000 years, based on biogenic carbonate accumulation at 48~ 25~ indicates that during interglacials (warm periods) primary productivity is generally high, whereas during glacials (cold periods) it is low. This corresponds to times of high and low atmospheric CO2 respectively, as suggested by the VOSTOK ice core record. The regional reconstruction of oceanic primary productivity shows that the contrast in productivity values between warm and cold periods decreases going from 60~ to 45~ Published reconstruction's at low latitudes (i.e. upwelling off NW Africa) indicate a continuation of this trend, ending in the tropics with high productivity during glacials and low productivity during interglacial periods which is the opposite of the higher latitude situation. This will be of importance when modelling global oceanic carbon flux. 1.
INTRODUCTION
The atmospheric stock of carbon dioxide continues to increase and with it, the potential for greenhouse induced global warming. However, the rate of carbon dioxide increase is less than the amount of CO2 released to the atmosphere since a portion is taken up by the ocean (1). Biological productivity is one of the mechanisms responsible for partitioning carbon within the ocean by drawing down both the atmospheric and surface water pCO2 (2). This biological pump depends on photosynthesis for carbon fixation, therefore on primary producers (i.e. coccolithophorids and diatoms) and on the amount of the primary production which escapes recycling in the mixed layer. The efficiency of this pump is further enhanced by the amount of carbon stored in deep-sea sediments either as particulate organic carbon (POC) or as carbonate (particulate inorganic carbon, PIC). In the present ocean calcifying pelagic organisms such as coccolithophorids and foraminifera are primarily responsible for the long term storage in deep
332 sea sediments. Some coccolithophorids, particularly Emiliania huxleyi, are important contributors of carbonate since they are capable of forming gigantic blooms at mid-latitudes (3). Since biological productivity plays an important role in the atmospheric CO2 uptake, we want to quantify the temporal and spatial variation in primary productivity for the Northeast Atlantic during the last glacial-interglacial cycle. By doing so we expect to gain a better understanding of the role of this part of the open ocean and to provide input for numeric models of the carbon cycle. In three piston cores from above the lysocline we determined temporal and regional variations in primary productivity during the past 150 ka (ka=1000 years) for the high to mid latitude (60-45 ~ N) Northeast Atlantic. Paleo primary productivity was derived from biogenic carbonate accumulation rates according to the method developed by Brummer and Van Eijden (4). We did not use organic carbon burial flux, since the little that is randomly preserved in deep water sediments of an open-ocean environment, does not represent the original production. Yet, these sediments cover orders of magnitudes large basin areas (5) and form a major link in the dynamics of the global carbon cycle. When using biogenic carbonate in deep-sea sediments as a primary productivity proxy two problems have to be addressed. Firstly, because carbonate dissolves increasingly below the lysocline, the use of its accumulation rate as a proxy for productivity may result in too low an estimate. We therefore developed a method for quantifying carbonate loss due to dissolution, using recent deep-sea sediments recovered from a range of water depths representing different degrees of dissolution (6). Secondly, carbonate in North Atlantic sediments can originate from both, calcifying organisms in the surface water and from ice rafting (7,8). The biogenic and detrital components must be separately quantified before biogenic carbonate can be used as a primary productivity proxy (9). 2.
M A T E R I A L AND METHODS
The studied sediments (box and piston cores) were recovered from the Northeast Atlantic (Fig. 1) during APNAP (Actuomicropaleontology Paleoceanography North Atlantic Project, 1988) and JGOFS (Joint Global Ocean Flux Studies, 1990) cruises with the Dutch R/V Tyro.
70 W~
60
I
I
50 I
I
40 i
I
30 I
I
20 t
I
10
C
I
-
".
!iiii!i!iiii.i.ji!ijiiiiiiiiii iiiiiiiiiiiiiiiiii!!i!iii!ii!iiiiiii!ii i i!!iiiiiii !iiiiiiii i~i~i~i~i~!i~i~i~i~.i~.i.~....................
0- 2 P
'
I
'
i~:i!i!i!i!i#! i:ii!ili!!::
50 0'/_ ._5,.T~ ' ~ T 9O-16 B ......... ':~i!:i~:~:;:i: ~ P~-e T 90~~9 P 990_8B ~ 04 -B ~ ~ ~ i ! i l i ~'k... ~ ,~ ~ ~ T 9 0 - 1 1 B ~ ~ I ! ~ ~ ~ o ~ ~ ' L ~ ~ f / ~T)90- l_A3B~))I ~ii~ S~~AN l~i! ~~:~ii~ :~:~~:i:~:~:~:~:~40N ~i :~:~:~:~~:~:~
-';, "............... ~ -: ...................... _~~-!ili!iiii~;iii::::!! :~:~:~:. .:........ ~%%o -i ~._......-,,.r--~~_.) ) _-~ ~ ~f~x'q_~_L I
60
,
_
'
-
I
'
I
'
I
'
I
'
Fig. 1 Physiographic map of the Northeast Atlantic Ocean showing the position of the studied box and piston cores.
333 For quantifying carbonate dissolution we used six box cores recovered between 44 to 47~ latitude and 20 to 24~ longitude and ranging from 3208-4375 m water depths, thereby representing different degrees of carbonate dissolution. To calculate the carbonate accumulation rates, we took a sample from the top and from a deeper interval in each box core and determined its carbonate weight percent and age. Dry bulk density was analysed on corresponding samples. Carbonate weight percent of the bulk sample, of the coarse (>32~m) and of the fine fraction (<32 l.tm) were determined by a gasonometric technique, with a precision of + 2%. Dry bulk density was determined by weighing 5 cm 3 of sediment after drying at 50~ Accelerator mass spectrometry (AMS) 14C measurements dated twelve samples consisting of hand-picked, >250 l.tm sized, excellently preserved, mixed planktic foraminifera. To reconstruct regional variations in primary productivity during the last 150 ka we used three piston cores, which were recovered from above the lysocline between 53 to 45~ Samples were taken at about 10 cm intervals for floral, faunal, carbonate, dry bulk density, terrigenous component and stable isotope analyses.Weight percent carbonate and dry bulk density for these samples were measured according to the method given above. Samples were dry-sieved through 125 l.tm mesh screens and then repeatedly split until about 500 grains remained. The biogenic carbonate component of the coarse (> 125 l.tm) fraction is composed mostly of foraminifera and pteropods while that of the <32 ~m fraction mainly consists of coccoliths. The detrital carbonate content of the coarse fraction was counted. The amount of detrital grains in the <32 ~tm fraction was visually estimated on smear slides of the fine fraction/bulk sample using a polarising microscope. Since these counts compare well with visual estimates, we used them both to approximate the relative frequencies of the bulk fraction. The relative frequencies were converted into absolute weights for the calculation of accumulation rates by assuming that detrital and biogenic carbonate grains have the same densities. For the build-up of a solid time frame, radiocarbon measurements using accelerator mass spectrometry (AMS) were used to date a selection of samples consisting of >250 ~tm sized mixed planktic foraminifera of an age of <50ka. Moreover stable isotope stratigraphy and two major North Atlantic ash falls were used in combination with the 14C dating to complete the stratigraphic framework for the studied sediments.
9
B I O G E N I C C A R B O N A T E A C C U M U L A T I O N AS A P R I M A R Y P R O D U C T I V I T Y PROXY
Biogenic carbonate accumulation rates can be used as a primary productivity proxy based on the observation, that the organic carbon flux and the carbonate fluxes in present day oceans are highly correlated when normalised to a 3200 m water depth (4,5,9,10,11,12; fig. 2a). When we assume a temporally constant carbon/carbonate rain ratio at a given depth, we can translate the carbonate flux in organic carbon flux (at 3200 m; fig. 2b)) and then reconvert this to surface ocean primary productivity using Suess' (13) formula, which corrects for organic carbon consumption with depth in the water column. The assumption of a constant rain ratio is supported by various other studies (14, 15, 16) and for different oceanic productivity regimes varying over an order of magnitude, which makes biogenic carbonate accumulation a viable proxy for primary productivity in open ocean environments. When carbonate dissolution can be quantified this primary productivity proxy can be applied on sediments recovered from greater water depths. We quantified carbonate loss by taking the difference in the carbonate mass accumulation rate of deeper cores from the shallowest one (6). It is assumed, that sediments in the shallowest core are undissolved, and represent the initial carbonate rain rate. We also assume that the initial carbonate rain rate is the same for all cores. These assumptions are viable since the studied box cores were recovered from a small geographic area governed by the same water masses and ecological regimes. Also, the shallowest core contains pteropods and juvenile foraminifera indicating that it lies above the
334
PRIMARYPRODUCTION ~-.ca. 11
0201
Suess, 1980 model
LL
~176 ~ /
d~
i
t'-
~~
o~ ~~t'xl o m
51
I /
0
o
.
*
I carbonate
0
*o~
o~ ~ o~
o.A ~ - 0 ~ , ~ " ~ 0
_
~
o
o -
I flux - - ~
0
_
,
50
9
8
"--
3
v
,
.......................................
- - - UT . - j ~
t'3 t~
j C-orgflux
,
100
9
..........................................................................................
,
150
C a r b o n a t e flux ( C a C O 3 F p )
in mg CaCO3/m2day
Fig. 2a Plots of organic carbon normalized to 3200 m according to Suess (13) versus carbonate export production, as intercepted in deep moored sediment traps in mid- and high latitude North Atlantic. We used the data set illustrated in (4) and added sediment trap data from the North Atlantic (17), including Lofoten Basin, Greenland Sea and Fram Strait (18), and equatorial Pacific (19). The linear regression is expressed as: organic carbon flux (3200 m) = 0.043*carbonate flux + 0.353 (r2=0.60). b. Explanation of how primary productivity can be reconstructed from biogenic carbonate accumulation rates (step 1 to 4) using the above described correlation between organic carbon flux and carbonate flux at 3200 m and the equation of Suess (13) to relate export production at depth with primary production. aragonite compensation depth and has undergone little if any carbonate dissolution. Thus, its carbonate accumulation rate (fig. 3) can be taken as the original carbonate rain rate. The carbonate dissolution rate increases with increasing water depth. A remarkable increase in carbonate loss is observed at 4,000 m which mark the lysocline (Fig. 3). It is not possible to define which fraction is most sensitive to dissolution since with increasing dissolution the fine fraction (<32 pm) becomes enriched with fragments from the coarse fraction (>32 lam). 4. R E S U L T S Primary productivity estimates based on biogenic carbonate mass accumulation rates are generally high during interglacials (warm periods), moderate during ambient glacials (cold periods) and low during Heinrich events (Fig. 4A). In our time series of surface ocean productivity of the last 150,000 years, based on biogenic carbonate accumulation at 48~ 25~ interglacials have an average productivity of 80 gC/m2/a, whereas glacials reach their maximum average at 49 gC/m2/a. Short intervals of reduced productivity during cold as well as warm periods correspond to "Heinrich events", which are short periods of massive iceberg discharge in the North Atlantic. This primary productivity reconstruction shows a positive correlation with atmospheric CO2 recorded in the Vostok ice core, with high primary productivity and atmospheric CO2 during interglacial periods and vice versa during glacial periods (Fig. 4B). The regional reconstruction of oceanic primary productivity based on biogenic carbonate accumulation at three sites (53 ~ 48 ~ and 45~ gives similar results namely high productivity during interglacials and low productivity during glacials. At 53~ primary productivity during interglacials reaches averaged values of 120 gC/m2/a, whereas at 45~ only 25-30 gC/rn/a is produced therewith creating a strong gradient in productivity. Furthermore, the contrast between "warm" and "cold" productivity decreases significantly from 60 gC/m2/a at 53~ to 510 gC/m2/a at 45~
335 CARBONATE MASS ACCUMULATION RATE ACD
2.0"
CARBONATE DISSOLUTION RATE
lysocline
ysocline
4
1.5 ,r
~
iiiiiiii!i!iiiiiiiiiiiiiii!iiiiiiiiiiiiiiiiiiiiiiiiiiiiii!iiiiiiiiiiiiiiiiiiiii iiiiii,iiiii,ii
0
::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
o ::::::: ::::::::::::::::: :::::i:i:::i:::: :::
o
1.0
o
0.5"
]
3200
3600 4000 Seafloor depth (m)
4400
0.0 3200
|
.
.
3600
1
4000
!
I
4400
Sealloor depth (m)
Fig. 3a Carbonate mass accumulation rate in g/cm2ka of the coarse (>32 ~tm) and fine (<32 ~tm) carbonate fractions. We used two age-models to determine carbonate accumulation rate, one based on calendar-calibrated radiocarbon ages (20) and the other on reservoircorrected (-400 a) radiocarbon ages (21, 22), for calculating carbonate mass accumulation rates. Accumulation rates based on calendar-calibrated ages are 14 to 20% lower than those based on reservoir-corrected radiocarbon age. b. Carbonate dissolution rate given in g/cm2ka as a function of water depth. Modified from (6).
5. CONCLUSIONS & DISCUSSION Although the North Atlantic mid latitudes are not assumed to be a prime region for "biological pumping" (drawing down of CO2 by increased primary productivity), a preliminary comparison of our reconstructed surface water productivity with the Vostok ice core atmospheric CO2 record shows that during times of high atmospheric CO2 ocean surface waters were blooming and vice versa. This is in agreement with the suggestion that phytoplankton growth at high- to mid latitude oceans is stimulated by increased CO2 concentrations when light and nutrient conditions are optimal (23). However, they do not necesarily indicate that huge amounts of CO2 were drawn down, since we did not calculate actual amounts with respect to the covered area. During Heinrich events, which generally correlate to times of low or decreasing atmospheric CO2, extensive iceberg shedding made surface waters unfavourable for primary producers. During these periods reconstructed primary productivity of the two northernmost cores (53 ~ and 48~ is lower than in the ambient periods. Exceptions to this pattern are the periods around Termination I and II (12 and 125 ka), during which a dramatic increase in CO2 concentration and Heinrich events occurred. Regional productivity reconstructions reveal a decreasing contrast in primary productivity between warm and cold periods going from 60~ to 45~ Published reconstruction's at low latitudes (15, 24) confirm a continuation of this trend ending in the tropics with high productivity during glacials and low productivity during interglacial periods. More detailed age-models are needed for both ice cores as well as deep sea cores to determine lead and lag times and answer the question whether variation in primary productivity is one of the causes or an effect of changes in atmospheric CO2 concentration. Furthermore, the strong geographical and temporal variation in primary productivity demonstrates the need for a higher coverage of deep sea cores with similar paleoproductivity estimates to arrive at a detailed picture of the global pattern necessary for Global Circulation Models.
336 To further improve the quantitative paleoproductivity estimates, research is required to validate primary productivity proxies world wide, and to update existing recent primary productivity maps. Our results suggest that during interglacials spring blooms at mid- to high latitudes act as one of the major biological pumps, whereas during glacials maximum productivity is restricted to low latitude upwelling areas. A
B
Piston core T88-9P Paleoproductivity estimates
Vostok ice core atmospheric CO 2 record
t-
._o
v
t-
(D "1-
gC/m2ka
<
20
40
ppmv
60
80
100
.,:::::::::::::::
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: .....................
::::::::::::::.:
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
hl === h2
,,,,r-
;",3
a
..... :::::::::::::::::::::::::::::::: ....
h3
180200 220 240 260 280
..
::::::::::::::::::::::::::
. - - ,,,,~
h4 50.
h5 ==
........o.....,.........:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:,: +:... ,..............,.. ..................
,,,,~_ ,,,,, .,,, -,, "~ .,,, .,.
h6
\
4
-) ..............................
75 ~ r \
!
:
h7 h8 h9 D
100 .
Suess
1
I
Betzer
125
150 "b
6
hl O ~ 175 9
1
h11== h12== h 13 ==
200 '
Fig. 4
(A) Primary productivity estimates for piston core T88-9P based on biogenic carbonate mass accumulation rates given in gC/m2a using Suess (13) and Betzer (25) models to convert the carbonate flux in primary productivity. (B) The 160 ka atmospheric CO2 record of Vostok ice core (26). After Kreveld et al. (9). Numbers (1 to 7) refer to isotopic stages.
337
PALEOPRODUCTIVITY ESTIMATES BASED ON BIOGENIC CARBONATE ACCUMULATION hl h2h3
h4
h5
h6
h7 h8
~0~'~ ~~ i i i::i:@: 100
~i
h9
i
80 60 40
0 I,,..
t'NI
E
cO
,
o
140
,
20
40
hl h2 h3
h4
i
,
60
Rm h5
do
lb0
17o
h6
h7
1~o
h8 h9
f ii
120
v
>
1:4o
100
o ~
"o o 13_ L_
I,,.,.
E
8O 60 40
",'3'13
.-_.:;-:,
i,._
13.. O (1) 13_
20
_=._- . .
o-;:
0 0
2~
4~
|
8~
|
80
|
100
i
120
|
140
160
140 120 100
_ -:
_ -:
_
_
-_
-_
- _
:
-_
-_-.
2934 m
-:
_
.
.
.
.
.
.
.
.
.
.
.
.
..'-i:
_
80 60 40
_
_
_
20 0 0
26
"4d
"66
,
80
9
i
100
9
l
120
--
9
140
-
.
160
age (calendar year "1000 BP) Fig. 5 Paleoproductivity estimates based on biogenic carbonate accumulation rates (4, 9, 13, 15) for three Northeast Atlantic cores, which were all recovered well above the lysocline. Note a decrease in productivity going form 53~ to 45~ as well as in the contrast between warm and cold productivity. Lines represent oxygen stable isotope stage boundaries; dotted areas represent cool periods.Rectangles h l-h9 mark intervals interpreted as Heinrich layers.
338 6. ACKNOWLEDGEMENTS This research has largely been funded by the Dutch Research Program Global Air Pollution and Gobal Change. The studied material was collected during Dutch APNAP and JGOFS cruises and were financed by the Foundation for Oceanic Research (SOZ). The radiocarbon dating was executed at the Van der Graaff Laboratory, State University, Utrecht. Michael Knappertsbusch is gratefully acknowledged for his supply of sediment trap data. Jakko van Waarden and Ron Kaandorp are thanked for their work on core T90-9P. 7. R E F E R E N C E S
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
Sundquist, E.T., The Carbon Cycle and Atmospheric CO2 variations Archean to Present, (eds. Sundquist E.T. and Broecker, W.S.), Washington, D.C., Geophys. Monogr. 32 (1985) 5-59. Berger, W.H., Fischer, K., Lai, C. and Wu, G., Univ. Calif., SIO 87-30, (1987) 67. Brown, C.W. and Yoder, J.A., Continental Shelf Res., 14 (1994) 175-197. Brummer, G.J.A. and Van Eijden, A.J.M., Mar. Micropaleontol., 19 (1992) 99-117. Milliman, J.D., Global Biochem. Cycles, 7, 4 (1993) 927-957. Van Kreveld, S., Ganssen, G., Van Hinte, J., Melkert, M., Troelstra, S., Van der Borg, K. and de Jong, A., Proc. 15th Intern Radiocarbon Conf., Glasgow, Scotland (submitted). Bond, G., Heinrich, H., Labeyrie, W., McManus, J., Andrews, J., Huon, S., Jantschik, R., Clasen, S., Simet, C., Tedesco, K., Klas, M., Bonani, G. and Ivy, S., Nature, 360 (1992) 245-249. Andrews, J.T. and Tedesco, K., Geology, 20 (1992) 1087-1090. Van Kreveld, S.A., Knappertsbusch, M., Ottens, J.J. and Ganssen, G., Van Hinte, J.E., Mar. Geol., (in press). Emerson, S., Bender, M., J. Mar. Res., 39 (1981) 139-161. Deuser, W.G., J. Foraminiferal Res., 17 (1987) 14-17. Deuser, W.G., Ross, E.H., J. Foraminiferal Res., 19 (1989) 268-293. Suess, P., Nature 288 (1980) 260-263. Lisitzin, A.P., SEPM Special Publ. 17 (1972) 218pp. Lyle, M., Murrey, D.W., Finney, B.P., Paleoceanogr. 5 (1988) 719-742. Rea, D.K., Pisias, N.G., Newberry, T., Paleoceanogr. 6 (1991) 227-244. Honjo, S. and Manganini, S.J., WHOI Technical Report, WHOI-92-15 (1992) 77pp. Honjo, S., Polar Oceanography, Part B: Chemistry, Biology and Geology (ed: Smith, W.O.J.), Academic press, San Diego, (1990) 687-739. Dymond, J. and Collier, R., Global Biochem. Cycles, 2, 2 (1988) 129-138. Stuiver, M. and Reimer, P., Radiocarb., 35 (1993) 215-230. Bard, E., Paleoceanogr., 3 (1988) 635-645. Stuiver, M., Pearson, G.W. and Braziunas, T., Radiocarb., 28 (1986) 980-1021. Riebesell, U., Wolf-Gladrow, D.A., Smetacek, V., Nature 361 (1993) 249-251. Sarnthein, M., Winn, K., Duplessy, J.C., Fontugne, M.R., Paleoceanogr. 3 (1988) 361-399. Betzer, P.R., Showers, W.J., Laws, W.A., Winn, C.D., DiTullio, G.R., Kroopnick, P.M., Deep-Sea Res., 31 (1984) 1-11. Barnola, J.M., Raynaud, D., Korotkevich, Y.S., Lorius, C., Nature 329 (1987) 408414.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
339
Surface water fCO2 in the Equatorial Atlantic Ocean D.C.E. Bakker a, H.J.W. de Baar a and E. de Jong a aNetherlands Institute for Sea Research, P.O. Box 59, 1790 AB Den Burg, Texel, The Netherlands
Abstract Fugacity of carbon dioxide (fCO2) in surface water was related to salinity and temperature for the equatorial Atlantic Ocean in late northern spring. The complex pattern of currents was unravelled using the ship's drift direction. Biological activity was low. Surface water fCO 2 was closely related to watermasses and changed abruptly at oceanographic fronts. Upwelling was characterised by high fCO 2, high salinity and low temperature. Undersaturation of fCO 2 north of the equator was ascribed to heavy precipitation and a thin mixed layer in the Intertropical Convergence Zone. The region between 15~ and 15~ was a slight source for atmospheric CO 2.
I. I N T R O D U C T I O N The oceans take up 20 to 40 % of the carbon dioxide (CO2) emitted by fossil fuel burning and cement making. Causes of the uncertainty are the enormous inorganic carbon reservoir, which the oceans constitute, and the large dynamic CO 2 air-sea exchange, variable with time and location. Continuous measurements of the fugacity of CO 2 (fCO2) in marine air and surface water in relation to salinity, water temperature and chlorophyll a were performed during five cruises in the (South) Atlantic Ocean at R.V. Polarstern. Aim of the project is to improve the estimate of CO 2 air-sea exchange and our understanding of its mechanisms for the South Atlantic Ocean. The paper presents surface water fCO 2 and CO 2 air-sea exchange in the equatorial Atlantic on a straight track between 15~ 3~ and 15~ 22~ from May 26 to June 3 1994 during one of those cruises, ANT XI/5. CO 2 air-sea exchange can be estimated as the product of a wind speed dependent gas transfer velocity and the difference of the concentration dissolved CO 2 at the sea surface and in the mixed layer. Biological, chemical and physical processes affect dissolved CO 2 in the mixed layer. The equatorial current system with "currents" and "countercurrents" is variable over ranges of time and space and responds strongly to changes in the wind field (1). Surface air ascents and precipitation is high in the lntertropical Convergence Zone (ITCZ), which shifts between 2~ and 8~ during the year. Trade winds come from the southeast south of the ITCZ, from the northeast noah of it. Trade winds are relatively steady in direction and speed and generally weaker than 11 m-s-1.
340 2. M E T H O D S
fCO 2 of surface water and air, total inorganic carbon and skin temperature were measured continuously. A showerhead type equilibrator was used for sampling of surface water fCO 2. Gaseous CO 2 was determined by a custom-built gaschromatograph with a methaniser and an FID-detector. The temperature correction of Copin-Mont6gut (2, 3) was applied. Total CO 2 was detected by coulometry (4). Alkalinity and nutrients were determined at regular intervals. CO 2 air-sea exchange was estimated from the concentrations dissolved CO 2 in the mixed layer and at the sea surface using the Wanninkhof (5) relationship negelecting the skin effect. The ship's data acquisition system supplied water temperature, salinity, fluorescence, wind speed and direction, atmospheric pressure and moisture content, air temperature, the ship's drift speed and direction. Dr. K. Schaumann (AWl) determined chlorophyll a. The region was characterised by steady trade winds from the southeast (140 ~ south of the ITCZ at 6~ and from the northeast (30 ~ further north. Ship's drift partly into the wind indicated a persistent current; a method already used by Renneli (6).
3. RESULTS The complex pattern of currents and countercurrents in the equatorial region was reflected in salinity, water temperature and fCO 2 (figure 1). The ship's drift suggested southward currents roughly between 9.3 ~ and 8.2~ and between 5.8 and 4.7~ (both the South Equatorial Countercurrent), eastward between 0.7~ and 0.6~ (Equatorial Divergence) and between 3.7 ~ and 5.7~ (North Equatorial Countercurrent). Trends for fCO 2 and salinity were strikingly similar, water temperature had a reverse pattern. Biological activity was low throughout the area as indicated by fluorescence (not shown) and chlorophyll a (Schaumann, personal communication). Surface water fCO 2 and, hence, CO 2 air-sea exchange were affected by the same physical processes as salinity and temperature, either directly or indirectly. Upwelled water had relatively low temperature, high salinity and high fCO 2. Upwelling had clearly occurred on both sides of the equator in the Equatorial Divergence and probably near the North Equatorial Countercurrent too. The salinity minimum at 5~ coincides with a minimum of water temperature and could have been caused by heavy precipitation of the ITCZ. The considerable undersaturation of surface water fCO 2 roughly between 0.7 and 9.0~ was unlikely to have been caused by the past or actual low biological activity. Thus, physical and/or chemical processes were responsible for the undersaturation. Upwelling would have elevated fCO 2, rather than have lowered it. Cooling of the water with 1 to 3~ necessary to cause the undersaturation thermodynamically, was unlikely with currents flowing parallel to the equator in an area of maximum surface water temperatures in late northern spring. Dilution of a thin mixed surface layer by heavy rains of the ITCZ could have caused the undersaturation.
341
15 A
10 oJ
L ~
"
0
v
E E
-5
9
!
9
i
9
!
9
i
9
i
!
Water
400
"-~
E
9
~
" ....................
300
36 "E
34
~
28
~
24
I-r
360
0 o
-~- ~" 180 v "0 ....
CI 0 -15
9 -1'0
-~
~
Latitude (~
~ +, ~
1'0
1'5
-)
Figure 1. C02 air-sea exchange, fC02 in water and air, salinity, water temperature and ship's drift direction between 15 ~ and 15~ ANT Xl/5.
342 4. CONCLUSION CO 2 air-sea exchange was closely related to oceanic watermasses, as demonstrated by reflection of the complex pattern of currents in the equatorial region in salinity, temperature and fCO 2. Abrupt changes of surface water fCO 2 occurred at oceanographic fronts. The CO 2 signature of each watermass was due to the common history of mainly physical, chemical and biological processes. These data will be combined with data from four additional cruises and compared with results by other researchers. Spatial and seasonal variation will be studied in relation to watermasses. Specific aspects of the equatorial region, like the pattern of currents and countercurrents, upweiling and undersaturation will receive attention.
5. ACKNOWLEDGEMENTS The work has been performed in close co6peration with the Alfred Wegener Institute for Polar and Marine Research and the crew of R.V. Polarstern. Dr. A.J. Watson of Plymouth Marine Laboratory helped us with the fCO 2 detection technique. The project is funded by the Dutch National Research Programme on Global Air Pollution and Climate Change (NOP) and the Netherlands Research Programme on Changes of Earth Systems (VvA).
6. REFERENCES
R.G. Peterson and L. Stramma, Progress in Oceanography, 26 (1991) p 1-73. C. Copin-Mont6gut, Marine Chemistry, 25 (1988) p 29-37. C. Copin-Mont6gut, Marine Chemistry, 27 (1989) p143-144. M.H.C. Stoll, Thesis Rijksuniversiteit Groningen, Groningen, 1994, 193 p. R.H. Wanninkhof, Journal of Geophysical Research, 97 nr.C5 (1992) p 7373-7382. J. Renell, In: J.G. and F. Rivington (eds.), London, (1832) 299 p.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
343
On the importance of surface mixing in an OGCM Andreas Sterl Royal Netherlands Meteorological Institute, Division of Oceanographic Research, P.O. Box 201, NL-3730 AE De Bilt, Netherlands Abstract An ocean general circulation model is coupled with a bulk mixed layer model and a sea ice model. It is forced both by daily varying fluxes and by the monthly means derived from these daily data. Comparing the respective results reveals the importance of highfrequency forcing. The incorporation of a sea ice model allows to study the interaction between the mixed layer and the sea ice. Sea ice extent appears to be very sensitive to mixing. Daily varying forcing makes it possible to analyse the mechanisms of mixed layer dynamics. 1. I N T R O D U C T I O N Recently, Sterl & Kattenberg (1994) (hereafter SK) showed that coupling a mixed layer model (MLM) to an OGCM improves the thermodynamic performance of the OGCM. Using monthly mean climatological forcing, they obtained better agreement with climatological sea surface temperature (SST) and needed a smaller correction to the applied heat flux than without MLM. Here these findings are extended by investigating the impact of using daily-varying forcing instead of the monthly means. The interaction between mixed layer and sea ice is also studied. 2. E X P E R I M E N T A L
SETUP
The HOPE ocean model (Wolff & Maier-Reimer, 1994), which includes a Hibler-type sea ice model, and the MLM of SK are used. The model is forced by a one-year set of daily values of air/sea fluxes obtained from an integration of the ECMWF atmosphere model at T63 resolution (Janssen, 1994). These fluxes are used as a new 'climatology', i.e., each year the model is forced by the same fluxes. The importance of high-frequency forcing is investigated by comparing runs forced by daily fluxes with runs forced by the corresponding monthly means. In all cases, SST and surface salinity are relaxed toward climatology to prevent model drift. Figure 1 shows wind stress (as a proxy of wind stirring energy available), mixed layer depth (MLD), and heat flux at 40~ 43~ to illustrate the working of the MLM. Clearly, heat flux and wind stress are highly anti-correlated, so that their respective impact on MLD is hard to separate. The mixed layer (ML) mainly deepens as a response to negative heat flux (cooling), but during summer deepening is possible with positive heat flux if wind stirring is high enough. During such episodes of ML deepening the heat flux error
344 300
800 4OO
t
I
2OO
0 -400
-8~840
100
5100 5100
5220 5280 5340 5400 Day
5(~90 5220
5250
Day
5280
Figure 1: MLD (thick solid, in m), wind stress (thin solid, in mPa), and heat flux (thin dashed, in m W / m 2) at 40~ 43~ Left: year 15, right: summer (JJA) enlarged. (the relaxation term for heat flux, see SK) decreases and the heat content increases (not shown), highlighting the importance of the IVIL as a sink for the atmospheric heat flux. 3. D A I L Y V E R S U S M O N T H L Y F O R C I N G SK used climatological fluxes to force their model when investigating the impact of the MLM on model performance. To test whether the values for the MLM parameters they recommended are also suited if daily forcing is used, the model has been run with and without MLM for both kinds of forcing. Figure 2 shows the rms SST and heat flux errors, respectively, from these four runs, averaged over the subtropics (10~ --. 30~ and the subartic (50~ --- 70~ As explained in Sterl (1994), the model fails to produce enough ice, which leads to large errors in areas that are affected by ice. Excluding subarctic winter from the analysis, Figure 2 confirms the finding of SK that the MLM mainly improves the model during summer, i.e., when ML formation is governed by wind-stirring. Furthermore, the improvement gained from adding the MLM is roughly the same for both forcings. Thus the parameter values recommended by SK are also suited for daily varying (high-frequency) forcing. 4. I N T E R A C T I O N B E T W E E N M I X E D L A Y E R A N D I C E Figure 2 also shows that the errors are smaller with daily forcing than they are with monthly forcing. The differences are enhanced if the MLM is present. Although they seem to be small, Figure 3 shows that these differences correspond to a large change in meridional overturning. Its maximum is situated at about 60~ the approximate latitude of the sea ice edge in the Labrador Sea (Figure 4). Further investigation indeed shows that the differences displayed in Figure 3 originate in the Labrador Sea and are related to the amount of sea ice there. Figure 4 shows the sea ice compactness for the two runs with MLM, obtained with daily and with monthly forcing, respectively. Although in both cases the model does not succeed in reproducing the actual sea ice distribution (see Sterl, 1994) it appears that the kind of forcing (daily/monthly) alters the sea ice compactness to some extend. This is also true for runs without MLM, even though for them the differences are smaller (not shown). With daily forcing, instantaneous fluxes can be much larger than the monthly average,
345 SST
RMS 1.6
.....
~._.: ....
1.4
.........
,,~" ~ ........
/
error
K
I .....
]_.:
i .........
/'
.... '~ .......
:
20
~.o 0.8
0.6
'
~.o
..........
2.2
- =."~ .....
1
"
.8
.....
1.4
..... ....
I
y:~
"'"
o
9rsg
........
/
-.; '/
! :-. .
......
i -.--"-': ....
]" /
,/
.....
"... " ~ . ::
"..
-20
.......
i
-40
-
- 40
- - -
-80
" .
i .........
o
/
error
...........
:C ~- ......
"".
- -
flux
, . . . . . . .
W,/rr~**2
, , , , .~
"--~
. . . . . . . .
- .....
,~ ......
,~ ........
.':
i . . . . . . . .
i
-60 . . . . . . . . . .
i ..........
.
Heat
r
:I
..........
" ..........
i .........
.....
,
:i
.........
._, .....
.~
") :i
:i
/~
-:--)
.
i .........
---~
."
i 6 i . . . . i68
i~i
Month
I
' i a o - ~%g'
is,
'
io~
'i~4
Month
i~o
Figure 2: Two years of rms SST error (left) and heat flux error (right), averaged over the areas between 1 0 ~ and 3 0 ~ (top) and 5 0 ~ and 70~ (bottom). (Thick" with MLM, thin" without; solid: daily forcing, dashed: monthly.)
leading to mixing reaching deeper at times. Warm water is brought to the surface, melting the thin ice at the ice edge. These changes in ice cover induce the changes in overturning shown in Figure 3. Without sea ice model there are hardly any such differences. This again shows the great sensitivity of the sea ice model with respect to changes in mixing and the subsequent impact on overturning that was already described in Sterl (1994). 5. C O N C L U S I O N S The HOPE OGCM with and without MLM has been forced with both daily and monthly varying fluxes. The results show that the MLM deals with both kinds of forcing
elOOO ,=2000 0-,
[]!;ii;iiiii!i!i~!;;i!~!i!i;;!;!iii!i!i;i!i!i!i!~;i!i!;i;!i!;;;~i;i!;!;!;!i!;~;!;;;!i!;i;!i;;iiii!;!ii;!;!i!;!;{;;!;iiiiii!i;i!i!;ii! ~ ' " ' ~ ! , . k ~ ~ . ~ . ~ v ~ . ~ . ~ . ~ . ~ . ~ Z ~ c ~ v ~ v ~ . ~ . ~ ~ Z : ~ Z ~ :
",-'.,'/,',',Z,_I~Z,',Z.2~.',',',22>'.'_
-/,-///.'/////'//////////..///.'.'//.'/.'//I'//,'////.'///.',""
"/////,//"/.
",'////
-
~_
.,.,,.,.v..,..,_, ",'/
-////-'//"//
~I~YY;~?~`:;~::~;~Z~;~Z~(~:~:~:~?Z4~Z9;:9Z~?:~Z<~Z~9:~Z~Z~Z~:~>Z~Z~Z~Z~9~~:~:j/i
~3000
4000
~~
~ ~~.F~~
~w ~ ~ ~ ~ ~T~~
~Y~ ~ ~ - ~
~~
~C~F ~ ~ ~ Z ~ ~ ~ ~ ~
:~ ~ Z ~ ~ ~ ~ ~
L','.'.',',~ . ","U.,Z,,~',"T',"T..~"/.~'. "%.~,","U.,~, ..'~,'~,'.
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: :9::::::
-20
0
20
40
IIc ~_ _
60
- , .~',~ ~ "
,
~';';'.',',',',','),: ~ l I :'T'.'~,','," .'Z,,., ~
::::::::;:: :::::;:':::::;:::l::
80
deg N Figure 3" Difference in annual-mean meridional overturning stream function between a run with daily and one with monthly forcing, both with MLM. Contour interval is 1 Sv, negative values are stippled.
346
80'm - ' - - ~ ~ ,-~>--~-'............
~
.---+=-~r . . . . - ~- =--I-/,, = ) ~ ' 7I.~~ %i . ~ - - ~ - ~ ~~~ '
70
<,~\<%
< >
50
I
'N
.
.
.
.
.
.
_~"
~'-'
I
60 'W .... ~ ~ c
70 'N
-
6 0 'N
-
40
50
~N
-,
I
~- . . . .
~
. . . . . L
:-L~'~.%]
.
.
~a,2,1~,.d~54<:_-_/
I
"W
4 ............ ~
L.
~
.
- -
:4-<"~t
80' N
,,
i
.
. . .-.-. .
~
20
}
'W
t _
,-~.
0 '
I
20
'E
_ _ _ =_--2-~ -~:L= - _~,_"
~
]
I
I
T-
!,........ 2~-i-~-=~.---- - , ~:.: ~ - ]\- ~ _ J~i.*~ _
i
~
I
"Jl, t
. ,-z,,
I
Figure 4: Sea ice compactness for run with daily (top) and monthly (bottom) forcing; MLM included. Contour interval is 10%. in the same way, i.e., it is not necessary to use different values for the parameters of that model if high-frequency forcing is used. This conclusion is important for the use of the MLM in a coupled ocean-atmosphere model, where the oceanic part is exposed to the high-frequency variability of the atmosphere. The results show further that the difference between daily and monthly forcing result in substantial differences in meridional overturning. At least in part, these differences correspond to and may be caused by changes in sea ice distribution. The sensitivity of overturning and sea ice distribution to the frequency characteristics of the forcing calls for care when testing OGCMs in stand-alone mode, which is usually done with monthly (climatological) forcing. 6. R E F E R E N C E S
Janssen, P.A.E.M. 1994, Results with a coupled wind wave model, Techn. Rep. No. 71, European Centre for Medium-Range Weather Forecasts, Reading, UK. Sterl, A., A. Kattenberg 1994, Embedding a mixed layer model into an ocean general circulation model of the Atlantic: The importance of surface mixing for heat flux and temperature, J. Geophys. Res., 99, 14,139-14,157. Sterl, A. 1994, Comparison of an explicit mixed layer model with a simple parameterization of surface mixing in an OGCM, submitted to J. Geophys. Res. Wolff, J.-O., E. Maier-Reimer 1994, H O P E - The Hamburg Ocean Primitive Equation Model. Cycle 1, DKRZ manual, Deutsches Klimarechenzentrum, Hamburg, Germany.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
347
Oceans and Climate" Circulation and interbasin exchanges in the Southern Ocean. Wilhelmus P.M. de Ruijter (project leader), Fred H. Walsteijn, and Raymond C.V. Feron Institute for Marine and Atmospheric Research, Utrecht University Princetonplein 5, 3584 CC Utrecht, The Netherlands Introduction The relatively warm climate of western Europe is strongly related to the Gulf Stream that branches north-westward as the warm North Atlantic Current. Evidence is piling up that in the recent past the climate state changed between warm and cold modes on time scales of decades only (e.g., Boyle and Keigwin [4], Dansgaard et al. [7], Harvey [22]). These are directly related to switches in the global scale thermohaline circulation in the ocean. This study focuses mainly on the Southern Ocean part of the circulation, where major interbasin transports take place. Variations in these exchanges may lead to significant global responses, particularly in the North Atlantic sector (Gordon [18]). Main theme of this project is the study of interbasin and intergyre exchanges in the Southern Ocean by performing sensitivity studies of the Antarctic Circumpolar Current in a numerical model and by determining mesoscale eddy exchanges and the associated mean circulation from satellite altimeter data. Results of both project components are summarized below in subsequent sections of this short paper.
T h e Antarctic Circumpolar Current The Antarctic Circumpolar Current (ACC) is an essential component of the climate system in that it accomplishes a significant poleward flux of heat and exchanges water masses and properties between the South Atlantic, Pacific, and Indian oceans (see, e.g., Nowlin and Klinck [29]). Therefore, reliable prediction of global climate change requires knowledge of the dynamical balance of this vital component and its sensitivity to changes in external and internal parameters. Measurements and model studies have provided the following characterization of the dynamical balance of the ACC (Nowlin and Klinck [29], McWilliams et al. [26], Marshall et al. [25]). Momentum input by wind stress is concentrated horizontally into a few narrow circumpolar jets by a high eddy activity in the upper ocean. Zonal momentum is transported downward, to the deep ocean, by isopycnal form stress induced by geostrophic eddies. Momentum is finally removed from the ACC by topographic form stress due to meanders in the deep ocean. Poleward heat transfer is directly related to this downward momentum transfer: if it is assumed that the momentum input is transferred downward completely, then the equivalent heat transport is close to the estimated loss of heat to the atmosphere south of the Polar Front (Marshall et al. [25]). However, key properties, such as the magnitude and variability of the ACC volume transport, still have not been explained satisfactorily. For example, the dependence of the ACC volume transport on the strength of the windforcing has been predicted by a number of simplified models, but there is no consensus at all. Baker [2] predicts a linear
348 relation (see also Godfrey [17], Straub [31]), while the concept of Johnson and Bryden [23] yields a transport proportional to the square root of the zonal wind stress. Marshall et al. [25] and Straub [31] give transport estimates that have no direct dependence on the wind stress. In the latter case only density stratification emerges as the dominant factor. As a second example, bottom friction plays no significant role in the momentum balance, but it is the most important energy sink of the system. Therefore, the sensitivity of model results to the value of this parameter has to be established. 1.1
Sensitivity
study
of the
ACC
The present study uses eddy-resolving simulations. Such high resolution is mandatory because transient eddies are a vital link in the ACC dynamics, as described above. The ACC has been modeled as a limited area quasi-geostrophic flow in a channel with schematic continents (Walsteijn [33]), similar to an earlier approach of McWilliams et al. [26]. As external parameters were varied: density stratification and magnitude and meridional structure of the applied zonal wind stress. Internal parameters that were varied are: bottom friction magnitude, position and width of Drake Passage, height and shape of topography, domain size, and type of lateral viscous terms (Laplacian or hyperviscosity). Each parameter was varied for two variants of the lateral viscous boundary conditions: either "no-slip" or "slip" conditions were used. 1.2
Results
When slip boundary conditions are applied, the model ACC shows a vacillation between a blocked state and an unblocked state with westward and eastward transports, respectively (Fig. la). Clearly, variability is large in comparison with observed Drake Passage transports. In this case an eastward zonal wind stress has been applied with a sinusoidal structure. The zero of its curl coincides roughly with the northern edge of Drake Passage. A reduction of the wind stress magnitude causes the ACC to become unblocked, i.e., it resides predominantly in the unblocked state. When the wind stress magnitude is increased or bottom friction reduced the flow no longer succeeds in sustaining a prolonged separation from the tip of "South America". Consequently, it predominantly resides in the blocked state (Walsteijn [33]). This reluctance to separate is a physically unrealistic feature of the case with slip boundary conditions. The blocked and unblocked states themselves are not sensitive to variations in the parameters except the density stratification, in the same way as for the case with no-slip boundary conditions (described below). Interestingly, the model ACC is hardly affected by mild topographic height variations and the width and position of Drake Passage. When no-slip boundary conditions are used, the model ACC resides persistently in an unblocked state (Fig. lb). The jet always separates from the south edge of "South America". The mean ACC transport is insensitive to variations of most of the abovementioned parameters. If the wind stress magnitude is increased or the bottom friction reduced only the amplitude of barotropic fluctuations is enhanced. A reduction of bottom friction by a factor of 16 is sufficient to obtain an energy balance where lateral and bottom friction dissipate energy at equal rates. Reducing bottom friction by a factor of 400 leads to
349
(a) Slip ~"
100
50 o
-50 -100 0
5
I0
20
15
time (years) (b) No-slip ~ 5o o 0 ~ -50~ ~. - 1 0 0 E-~
7'
,
0
,
,
i
5
,
,
,
,
10 time (years)
,
,
,
15
F i g u r e 1" Time series of eastward circumpolar volume transport (in 106 m 3/s) obtained from simulations with two layers and a sinusoidal zonal wind stress. Panel (a): slip boundary conditions; panel (b): no-slip boundary conditions.
a flow in which lateral friction constitutes the dominant energy sink. However, in the momentum and vorticity balances interface stresses and topographic form drag remain dominant (see Walsteijn [33]). The most important factors that determine the volume transport are stratification and the meridional structure of the wind stress. Reducing the stratification leads to a reduction of the transport almost linearly with the first internal deformation radius (Fig. 2). (See Walsteijn [33] for a discussion.) Shifting the zero-contour of the wind stress curl to the north gives a more realistic position of this line relative to the tip of "South America". The average meridional position of the ACC coincides with this line, except where the ACC is affected by topography or continents (Fig. 3a). Surprisingly, a northward shift turns out to have no significant effect on the ACC volume transport. If the wind stress curl is zero (i.e., the wind stress is uniform) the ACC mainly follows an alternative "southern" path, dictated by the iso-lines of ambient potential vorticity, although weaker "northern" paths can be distinguished as well (Fig. 3b). The volume transport is increased with respect to cases where a significant wind stress curl determines the ACC's position entirely (Fig. 2). Also, the sensitivity of the transport to the stratification is larger. In summary, the meridional distribution of wind stress determines the path(s) followed by the ACC. Preferred paths lie along ambient potential vorticity contours and at the latitude where the wind stress curl crosses zero, i.e., at the wind stress maximum. The transport is largest if both paths are fully active, e.g., when the wind stress maximum occurs at a northern latitude while the forcing is still large at southern latitudes (Figs. 3c and 4).
350 T r a n s p o r t (Sv) . . . . . . . . .
I
. . . . . . . . .
I
. . . . . . . . .
i
. . . . . . . . .
i
. . . . . . . . .
!
100 80
60
40
20
. . . . . . . . .
0
i . . . . . . . . .
10
! . . . . . . . . .
20
i . . . . . . . . .
30
i . . . . . . . . .
40 D e f o r m a t i o n radius (km)
i,,,
50
Figure 2" Time-mean circumpolar volume transport as a function of the internal Rossby deformation radius (i.e., the strength of the density stratification), for two-layer flow in a small domain with no-slip boundary conditions. Line A: the zonally directed wind stress has a sinusoidal meridional structure. Line D: the wind stress is uniform.
2
E d d i e s a n d m e a n flow in t h e S o u t h e r n a n a l y s i s of s a t e l l i t e a l t i m e t e r d a t a
Ocean:
Satellite systems are most effective in providing observations necessary to understand the oceans on a global scale. Observations of the sea surface height can be obtained by a satellite altimeter (e.g. Douglas and Cheney [9], Feron [12]). Sea surface height is directly related to ocean currents and their variations. In this part of the study three years of Geosat-data (GEOdetic SATellite) were used from its Exact Repeat Mission (ERM) to determine eddy fluxes and the associated mean circulation in highly energetic regions of the Southern Ocean One of these areas lies around South Africa where the Agulhas Current system interconnects the warm Indian Ocean with the relatively cold Southern Atlantic (Gordon et al. [19], Lutjeharms et al. [24]). The heat and fresh water fluxes between these two oceans takes place largely by enormous Agulhas rings (with diameters of order 300 km) that pinch-off from the Agulhas Current and penetrate the Atlantic (Boudra and De Ruijter [5], Gordon et al. [19], Wakker et al. [32]). Part of the Agulhas volume transport leaks directly into the South Atlantic. This leakage of warm Indian Ocean water, which is not well defined, together with the import of Agulhas rings may be a critical link in the global thermohaline ocean circulation and thus also in the climate system (De Ruijter [8], Gordon [18], Broecker [3]). Although the detection of Agulhas rings in the relatively quiet South Atlantic subtropical gyre is rather straightforward (Gordon and Haxby [20], Naeije et al. [27]), the shedding process is difficult to extract from altimeter measurements. In ocean areas with strong currents, like the Agulhas Current, the inability 1 to determine the total flow field 1This inability is caused by the fact that the exact shape of the geoid is unknown (Cheney and Marsh [6], Nerem et al. [28], Rapp and Wang [30]).
351
(a
(b o
9
(c Figure 3: Time-mean upper layer streamfunction (left panels) in a model of the Antarctic Circumpolar Current for three different meridional distributions of the zonally directed wind stress (right panels). In the top row (a) the wind stress has a sinusoidal structure, with a maximum about 400 km north of Drake Passage and a zero plateau at southern latitudes. In the middle row (b) the wind stress is uniform. The lower row (c) gives a case where the wind stress is a mixture of the two other cases. Both "southern" and "northern" paths are now clearly present. No-slip boundary conditions were used, except for slip conditions at the northern domain boundary. Tickmarks indicate edges of Drake Passage, where periodic boundary conditions hold. The model has three layers, and first internal Rossby deformation radius of 24.31 km. Domain size is 6275 • 1412 km~ Bottom topography consists of mounts in Drake Passage and at the center of the domain. Contour interval is 3 Sverdrups (3 • 106 m3/s). leads to artificial eddy-like structures in the relative satellite altimeter observations. In those ocean regions it is difficult to distinguish between rings and current meanders. Mesoscale eddies also play a significant role in the meridional exchange between the Antarctic Circumpolar Current (ACC) and the subtropical wind-driven gyres. The area of interaction between the ACC and the subtropical gyres is characterized by thermocline outcropping and has strong frontal features. This is also the area where the thermocline interacts directly with the atmosphere and where the t h e r m a l and haline properties of the ocean are established. Quantifying the heat and fresh water exchanges across these confluence fronts should be an essential ingredient of a climate observation system. We are presently at a stage of developing the analysis m e t h o d s t h a t will lead to such quantifications from satellite observations. In view of the above the following main goals of this part of the study emerge: 1) to detect mesoscale ring formations and study their subsequent trajectories and evolution from radar altimeter observations without knowing the exact shape of the geoid and 2) to derive the m e a n surface flow field from satellite altimeter observations of horizontal eddy fluxes.
352 Transport 200 150 100 50 0 -50 -100
' ,
0
,
5
.
' .
.
(Sv)
l"' .
"' i
" .
10 time (years)
.
.
.
'
' " ,
,
"["~ ,
,
15
Figure 4: Time series of eastward circumpolar volume transport (in 106 m3/s) for the case in Fig. 3c. 2.1
Results
Combination of harmonic analysis and principal component analysis enables the extraction of typical Agulhas Current frequencies of the sea surface dynamic topography from the 3-year Geosat altimetric data set. Applying these techniques to successive sea level anomaly maps in the Agulhas retroflection area indicates that 18 events took place over a 3-year analysis period. Variations occur from year to year. The combined analysis shows that the first three modes (principal components) of variability are statistically dominant. The structure of the spatial and temporal scales leads to the hypothesis that these modes are associated with periodic Agulhas front movements, culminating in the formation of large Agulhas rings. Sharp changes (pulses) in the sea level pattern then determine the time of pinch-off. If these pulses can be connected to the formation of large Agulhas Rings, they are of great importance for the large-scale circulation because the rings contribute significantly to the energy and fresh-water flux between the Indian Ocean and South Atlantic (Feron et al. [11]). The validity of our hypothesis was confirmed by applying the same techniques to the UK, Fine Resolution Antarctic Model (Webb et al. [34], Feron [14]). As a logical extension the method has been also applied to two other western boundary current systems in the Southern Ocean, the East Australian Current and the BrazilMalvinas (or Falkland) Confluence, leading to a catalogue of Geosat-observed ring shedding events in both current systems (Feron [13], Feron [14]). Active research is ongoing, also in our group, to estimate the impact of these eddies on the global meridional fluxes of momentum, heat and fresh water (Drijfhout [10], Feron [16], Van Ballegooyen et al. [1]) In the final part of this project we have developed a method to calculate mean sea surface dynamic topography (MSSDT) and thus the ocean surface circulation from the observed temporal variability (Feron [15]). Averaging of the potential vorticity balance leads to a relation between the mean divergence of the eddy-vorticity fluxes and the time-averaged vorticity field. The former can be derived from the satellite altimeter observations. Our method appears to be applicable to areas of the ocean with sufficiently strong mesoscale variability such as the major western boundary currents and their extensions and to the frontal regions of the Antarctic Circumpolar Current (Feron et al. [15]). Surprisingly realistic results followed for the Southern Ocean (Fig. 5, Feron [15]). An im-
353
portant application is to combine the newly derived averaged flow field with the observed eddy field to derive the total time-varying surface velocity field. As a striking example this has been applied to the Agulhas Current retroflection where the repeated shedding of large rings can now be reconstructed as a continuous process (Fig. 6). In principle, our method could form the basis of a monitoring system for such highly energetic areas. Levitus MSSDT
Divergence of eddy stresses
Improved MSSDT (GEOSAT)
Figure 5: (a) The Mean Sea Surface Dynamic Topography (MSSDT) in the Agulhas Extension in dynamic cm from Levitus, 1982 climatology, relative to 1000 dbar, (b) the mean divergence of the eddy vorticity flux, (units: x10-13s-2), derived from satellite altimeter observations,(c) improved MSSDT solution (i.e., surface flow field) from our new approach, contour interval is 10 cm.
3
S u m m a r y and Conclusions
In the Earth's climate system the redistribution of heat between equatorial and polar regions takes place in approximately equal amounts by the ocean and the atmosphere. Variations in these transports on decadal, and larger, time scales are mainly due to changes in the ocean's thermohaline circulation. In the Southern Ocean cross frontal exchanges, largely by mesoscale eddies, form an essential component of the global meridional overturning circulation of the ocean. Observing and realistically modeling those exchanges and their variability in time should therefore be an integral part of a Global Climate
354 total
sea
~!,,<~':.s:t:p~t%~q:~:~i~::~::~!;~. 5 ~ :--
~
:
:
:
:
:
:~--
total
~;~::.:
sea
surface
topography
,~..
" :"::
surface
~
~
-
S " ~ ' ~ > : . i : i
topography
(week
i
30)
~
........::: . . . . .
~ - :: :::-: ........
(week
" .......~
.... ~-~"::a":
:
31) 1~0 150 14,0 1:30 CM 1~0 110 100 90
total
sea
surface
topography
(week
32)
80 70 60 50
~>,~
s::: ~2~i~,~ .....
,'~. . . . .
::~:: :: ,:,-~ .
~ 40 30 20
Figure 6: A time sequence of the absolute dynamic height in the Agulhas Retroflection at weekly intervals determined by applying the method of (Feron et al., 1994) to Geosat altimeter height observations. In this sequence the actual shedding of large Agulhas rings south of Africa can be observed. Analysis of the 3-year observational period shows that 18 similar ring-shedding events occurred. These rings carry a large amount of heat into the South Atlantic, and thus establish one of the interbasin links in the global heat and fresh water circulation. Time is relative to 8 november 1986 and the contour interval is 10 cm. Observing and Forecasting System. In this project a few fundamental steps have been taken toward developing such a system. The first part of the present study has revealed the sensitivity of a new eddy-resolving model of the Antarctic Circumpolar Current (ACC) to its external and internal parameters. The ACC is most sensitive to the vertical density stratification and the wind stress profile (Walsteijn [33]). Depending on the wind stress profile the ACC consists of one or more jets. When wind stress has a significant curl, an ACC jet emerges that is forced towards the latitude where the curl crosses zero, i.e., where the wind stress is maximal. An alternative or second jet occurs if the wind stress is large at latitudes south of Drake Passage. Then the volume transport increases significantly as does its sensitivity to the stratification. The time-mean transport of simulations with a large domain and broad wind stress profile (e.g., Fig. 4) is of the same order as the observed value (cf. Nowlin and Klinck [29]). The transport history of the model shows realistic time scales, such as mild modulations with periods on the order of 6 months to several years. However, barotropic fluctuations with a period of roughly 25 days have an amplitude which is about three times larger than corresponding fluctuations in observations. Using a more realistic topography (e.g., by including small-scale "bottom roughness") would probably reduce the
355 magnitude of these fluctuations to more realistic levels. The dynamics of ACC jets are determined by transient eddies and the density stratification. Eddy activity is strongly affected by changes in stratification (Walsteijn [33]). The strength of the reverse feedback is still unknown, i.e., it is unclear to which degree the eddy heat transport is coupled with the stratification and thermohaline overturning. This is an area of active research. A full description of (sensitivities of) the ACC variability and transports requires both eddy-resolution and more complete thermodynamics than is present in a quasi-geostrophic model. For future climate research it is, therefore, important to extend the present sensitivity study to a system of primitive equations. The possibility to synoptically study the ocean's eddy field and related mean circulation from satellite altimeter observations has been verified and explored in the second part of this study. Progress has been made concerning the generation mechanism, formation rates, trajectories, translation speeds, and lifetime/dissipation rates of eddies in the Southern Ocean western boundary currents. Over the 3 year observational period of Geosat approximately 18 large rings pinched-off from the Agulhas Current, south of Africa. The majority of these rings reaches the South Atlantic. Just after being formed Agulhas rings have the largest translation speed, approximately 8 cm/s. When they move out of the highly active Agulhas retroflection area their translation speed reduces to approximately 4 cm/s. The associated volume transport on a yearly basis is at least 7 • 106m3/s, approximately half of the total exchange between these two oceans. We therefore conclude that Agulhas rings contribute significantly to the Indian-South Atlantic connection and the associated heat and fresh water flux. Van Ballegooyen et al. [1] estimated from a combined hydrographic-altimeter study that the net heat fluxes (300 Wm -2) and evaporative losses (1 cm/day) to the atmosphere due to eddies within the Agulhas region are appreciable larger than the summer climatological means for this region. Their volume flux 6 . 3 - 7.3 • is consistent with other estimates. Similarly, they estimate fluxes of heat and salt into the Atlantic Ocean via the Agulhas eddy field of 0.045 PW and 78 • 1012 kg per year, respectively. In the concluding part of this work, which is still ongoing, we try to better determine how the eddy field interacts with and modifies the mean circulation and how it is coupled to the deeper circulation. Coupling with numerical models will lead to improved estimates of meridional and cross frontal fluxes. First results are promising and show that it appears to be possible to improve the estimates of the mean sea surface dynamic topography by using satellite altimeter observations of horizontal eddy momentum and vorticity fluxes (Feron et al. [15]). Partly due to the success of Geosat, new altimeters are now operational. With the successful European Remote Sensing Satellite (ERS1) and Topex/Poseidon new altimeter data is becoming available with increased absolute accuracy and precision (Haagmans et al. [21]). Aside from major improvements in the new altimeter observations, the continuity of observations and their accurate analysis is a high priority for future climate studies aiming at Global Ocean and Climate Forecasting. Acknowledgment. This investigation was supported by the Dutch National Research Programme on Global Air Pollution and Climate Change, projectnumber 850025, and the Stichting Ruimte Onderzoek Nederland (SRON). Computations were performed on the CRAY Y-MP and CRAY C90 at the Academic Computing Centre (SARA), Amsterdam, The Netherlands. Use of these supercomputer facilities
356 was sponsored by the Stichting Nationale Computerfaciliteiten (National Computing Facilities Foundation, NCF), with financial support from the Nederlandse Organisatie voor Wetenschappelijk Onderzoek (Netherlands Organization for Scientific Research, NWO). Delft University of Technology (section for Space Research and Technology and Faculty of Geodesy) is acknowledged for the altimeter data processing.
References [1] Ballegooyen, R.C. van, M.L. Grfindlingh, and J.R.E. Lutjeharms, J. of Geophys. Res., 99, 1405314070, 1994. [2] Baker, D.J., Jr., J. Mar. Res., 40, suppl., 21-26 (1982). [3] Broecker, W.S., Oceanography, 4, 79-89, 1991. [4] Boyle, E.A. and L.D. Keigwin, Nature, 330, 35-40, 1987. [5] Boudra, D.B. and W.P.M. De Ruijter, Deep Sea Res., 33, 447-482, 1986. [6] Cheney, R.E. and J.G. Marsh, EOS Trans, AGU 62(45), 743-752, 1981. [7] Dansgaard W., J.W.C. White, and S.J. Johnson, Nature, 339, 532-533, 1989. [8] De Ruijter, W.P.M., J. Phys. Oceanogr., 12, 361-373, 1982. [9] Douglas, B.C. and R.E. Cheney, J. of Geophys. Res., 95, 2833-2836, 1990. [10] Drijfhout, S.S., PhD thesis, Utrecht University, 1992. [11] Feron, R.C.V., W.P.M. De Ruijter, and D. Oskam, J. Geophys. Res., 97, 9467-9477, 1992. [12] Feron, R.C.V., Change, 11, 6-7, 1992. [13] Feron, R.C.V., Satellite Altimetry in Geodesy and Oceanography, Springer-Verlag, Berlin, 1992. [14] Feron, R.C.V., accepted in J. Geophys. Res., 1994. [15] Feron, R.C.V., W.P.M. de Ruijter, and P.J. van Leeuwen, Submitted to J. Geophys. Res., 1994. [16] Feron, R.C.V., PhD thesis, Utrecht University, 1994. [17] Godfrey, J.S., Geophys. Astrophys. Fluid Dynamics, 45, 89-112, 1989. [18] Gordon, A.L., J. Geophys. Res., 91, 5037-5046, 1986. [19] Gordon, A.L., J.R.E. Lutjeharms, and M.L. Griindlingh, Deep Sea Res., 34, 565-599, 1987. [20] Gordon, A.L. and W.F. Haxby, J. Geophys. Res., 95, 3117-3125, 1990. [21] Haagmans, H.N., M.C. Naeije, and R.C.V. Feron, Geodetical In]o Magazine, 5, Nov-Dec 1993 [22] Harvey, L.D.D., Quaternary Science Reviews, 8, 137-149, 1989. [23] Johnson, G.C. and H.L. Bryden, Deep-Sea Research, 36, 39-53, 1989. [24] Lutjeharms, J.R.E., W.P.M. De Ruijter, and R.G. Peterson, Deep Sea Res., 39, 1791-1807, 1992. [25] Marshall, J., D. Olbers, H. Ross, and D. Wolf-Gladrow, J. Phys. Oceanogr., 23, 465-487, 1993. [26] McWilliams, J.C., W.R. Holland, and J.H.S. Chow, Dyn. Atmos. Oceans, 2, 213-291, 1978. [27] Naeije, M.C., Wakker K.F., R. Scharroo, and B.A.C. Ambrosius, ISPRS, J. Photogramm. Rein. Sensing, 47, 347-368, 1992. [28] Nerem, R.S., B.D. Tapley, and C.K. Shum, J. Geophys. Res., 95, 3163-3179, 1990. [29] Nowlin, W.D., Jr. and J.M. Klinck, Rev. Geophys. Space Phys., 24, 469-491, 1986. [30] Rapp, R.H. and Y.M. Wang, Geophys. J. Int., 117, 511-528, 1994. [31] Straub, D.N., J. Phys. Oceanogr., 23, 776-782, 1993. [32] Wakker, K.F., R.C.A. Zandbergen, M.C. Naeije, and B.A.C. Ambrosius, J. Geophys. Res., 95, 2991-3006, 1990. [33] Walsteijn, F.H., In preparation, 1994. [34] Webb, D.J., P.D. Killworth, A.C. Coward, and S.R. Thompson, Natural Environment Research Council, 1991.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
Determination of A R G O S drifters.
NE.
the
Atlantic
357
current
field
with
L.Otto, H.M. Van Aken and R.X. de Koster. Netherlands Institute for Sea Research, P.O.Box 59, 1790 AB den Burg, Texel, The Netherlands. The patterns of near-surface currents in the ocean as we know them today only give a very general picture, because they are based upon data that and calculations that are not adequate in every respect. Regional and temporal variability cannot be resolved in sufficient detail and this is for instance reflected in uncertainties in the estimates of the transport of mass and heat in the ocean. Yet these estimates are important for the assessment of the role of the oceans in the climate system. For example, in the northern part of the North Atlantic, north of latitude 53 ~, the near-surface circulation follows an anticyclonic pattern (the Sub-Arctic Gyre), with northward flow in the eastern part, and southward flow in the west. But how much of the water flows into the Norwegian Sea beyond the Scotland-Iceland Ridges, and along what routes, and how much of it turns westward south of Iceland, is still a matter of dispute. As the Sub-Arctic Gyre is an important link in the thermohaline ocean circulation (the "Conveyor Belt") better information is required for a realistic modelling of the oceanic part of the climate system.
,F o,._ ~
Gyre
f .... IJ /
50 N
i
60 W Fig.1.
/
~
,
ql
". . . . . . . . .
.---
4o N
]
Sub_Trropica I
Gy e )
"
i
40 W
i
20 W
/
0
General pattern of the surface circulation of the NE. Atlantic and the "DUTCH-WARP" study area.
358
Since about a decade the use of satellite-followed drifters offers a new possibility for observing ocean currents. The principle is that drifters that are released in the ocean can be followed regularly from satellites, and the analysis of the tracks thus obtained gives information on the mean current pattern, and the variability in time, the occurrence of eddies, etc. In the framework of the "DUTCH-WARP" programme ("Deep and Upper Transport, Circulation and Hydography, WOCE Atlantic Research Programme"), aiming at a better description of the circulation of that part of the Atlantic, a series of drifter observations was initiated, that has been supported by the VvA-3 programme. In the years 1990-1993 in total 19 drifters were released from the RV "Tyro" and the weathership "Cumulus" in the NE Atlantic. The drifters were drogued at 15 m, and were followed over periods between 43 and 365 days. In total the data cover some 10 drifter-years. The mean drift velocity for the area is typically of the order of 2 cm/s, to the northeast. However, the tracks reveal an important effect of the submarine topography and the main thermohaline structure on the regional current pattern. During the summer mean westward flow is observed over the western part of the Iceland Basin, and northeastward flow over most of the eastern parts. Over the Rockall Plateau the currents are variable and smaller. It is also interesting to compare the drift observed across the WOCE AR-7E section (IrelandGreenland) with the hydrographic structure observed during "DUTCHWARP" 1991. As for many ocean areas the eddy kinetic energy in the area is much higher than the mean kinetic energy. This means that the role of eddies in the transport of heat and salt cannot be neglected. An interesting result is that there is a marked difference between the high levels of eddy kinetic energy over a deep (> 2000 m) region as the Iceland Basin, and the much lower levels over the shallower (< 1000 m) parts of the Rockall Plateau. The results found here show that transport and exchange of water over parts of this area can be quite different. As there are indications that the convection over the Rockall Plateau is an important mechanism in the formation of the so-called "Mode Water", the transport and exchange between this area and the surrounding waters is an important point in the air-sea exchange of the NE Atlantic region. In the coming years drifter data will be used to obtain improved maps of surface currents. The programme reported here is a contribution to this international effort.
"Jewwns u! 8(s/wo) ssew l!un Jed AfJeue o!leU!>l Appe eoetJns to uo!lnq!als!a 00
I.ooo.,,r It-
k te ,' ~~9~
o0 ~
~ #
~,
~."
'.~~x
,.'M'
'Y
,'(
~
.,
/-"
i~
'
,
i "
"
.',,.
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: ii::::::::f::::::::;? ,-7" :::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: .; ::::::::::::::::::::::::: :::::::::::::::::::::::::::::::::::: .....
~
X~i::~:,~i~-~,~; -
,.;>;
,..:;;
/ ./ ....
. - .... /
. . . . . . ,,-
o
~
./
?:, .
!;
/,
o
/'i
Y.
;')
,, i..;.. <"
....
.Ol
~/
,
,,
o
.OZ
',
.___.___
~"
@?<<~:-~:
ii~i:.ii~i
~ ( ~
~
...~-,,,
..... .,x~
", ,
8"6!-t
o:
t_ '
~
/.___s ~ >&.
i
,~-
~
~'-"
'~
--I<..
//"
~-
'
/
j "
,
~"AeDAo nq s L< #i:i:i:i:i:!~iT~ijLo 9 . I=
..:. . . . /
" W e "
s
"
"
,'i ~owwnS
~7
/
,0 .-~
/
\
!
'-,,
OZ
j
"
/
I
#' /
"J
/
~
~
~
"
.."
~ "~
""
U'-
O~
_
,r~ I
i
~-'~"
I/k..- ' ~ " ~
,-',
_A"
I
~
\\
,
,/ ,_?',~% e
/
,
"\
0
~;6!-i
.--
..:...... ~i..::i:i:i:!:i::
::~::::::::::::::::::
i!i ~!~i f:~::~::i::i::f:i:;::~::ii~::~ii F:~!::ii !!~i i~g !i!> P
:;
"---"~
~":' , ~:::::::::.....
~ ,..
/"/~o
,~
'
"
~
.,.,/,
00s
~
~
/
'
/ .i .~ . .
..--__-. ~
t _--
;; ;c>~,C-
~
/ .(3
,OZ
""~> ; , oOO_o..::iij!i!#' . ~ "~ ':'::i%...
.~,"b..~ I~
iI
#t ....
..
09
c', v
,
W/
k" > 5 , ' ,,
"(epnl!BUOlo8 x epnl!lelo L) u!q Jed 866 ~-066k elep ~euJwns LUOJt pe^!Jep lueJsno eoe~Jns u e e ~ o
i'tOQ9N ~
J, , ,
I o9-I
i
I" J
...... ' ~ ~ ' ~
?
/~--7
*
/,.-'
0s
6S'~
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
361
Repeated XBT sections in the framework of WOCE T.F. de Bruin, L.Otto, S. Ober, R.X de Koster, H.M. van Aken Netherlands Institute for Sea Research (NIOZ) P.O. Box 59, 1790 AB Den Burg, The Netherlands The upper layers of the ocean have an important role in the coupled oceanatmosphere system. They exchange heat with the atmosphere on seasonal and interannual time scales and the meridional transport a substantial part in the global heat budget is via the upper layers of the oceans. In the observational programmes of T O G A and W O C E special attention is given to a co-ordinated system of regular observations of the thermal structure of the upper ocean by m e a n s of XBT (eXpendable BathyThermograph) measurements. XBT observations can be made from ships underway. The XBT sonde consists of a thermistor probe that is launched from the ship and that sinks with a well-known falling speed, transmitting its temperature signal to the recorder on board via a thin unwinding copper wire. As a result a temperature-depth registration is obtained down to depths of about 400, 700 or 1800 m (depending on the type of probe). When completely unwound the wire breaks. The advantage of this method is that data can be obtained by "ships of opportunity" without interfering with their normal duties.
TABLE 1. Overview of XBT measurements in the joint Navy-NIOZ project. Royal Netherlands Navy frigate
Period
Number of XBT's Total
T7
T5
Banckert Piet Heyn
14-05-91 / 24-05-91 25-05-91 / 04-06-91
85 82
85 82
0 0
Kortenaer Banckert
05-11-91 / 13-11-91 17-11-91 / 26-11-91
113 117
94 94
19 23
Philips van Almonde Kortenaer
19-05-92/ 29-05-92 31-05-92 / 09-06-92
114 104
94 78
20 26
Callenburgh Philips van Almonde
03-11-92 / 12-11-92 16-11-92/ 24-11-92
115 108
93 91
22 17
Bloys van Treslong Callenburgh
25-05-93/ 01-06-93 07-06-93 / 15-06-93
95 106
75 91
22 15
Karel Doorman Bloys van Treslong
09-11-93 / 18-11-93 21-11-93/ 30-11-93
111 116
88 92
23 24
Willem van der Z a a n Karel Doorman
17-05-94/ 23-05-94 04-06-94 / 08-06-94
92 61
72 61
20 0
Karel Doorman Willem van der Z a a n
22-11-94 / 30-11-94 05-12-94/ 13-12-94
114 125
91 104
23 21
1658
1385
273
Total:
362
Naval ships normally make XBT observations in connection with anti-submarine programmes, but as a rule not at high spatial resolution. The programme reported here is a co-operation between the Royal Netherlands Navy and the Netherlands Institute of Sea Research, with support from the VvA-3 programme. The regular relieving of the frigates stationed at Curas offers an opportunity to carry out XBT observations (T-7 probes, down to 700 m) on the route English Channel-Antilles (the WOCE AX-5 section) at 30 miles intervals, twice in spring and in fall. This spatial density ("highdensity mode") resolves much of the meso-scale structure, while the time schedule gives an opportunity to investigate seasonal variability in heat content and heat transport. Fig. 1. MAP OF XBT MEASUREMENTS
MAY 1991- JUNE 1994
50 40 -1l-.rr" O 3O z uJ a D !- 20
~
...-,, ........
.
...........e e ~
I
I
oee .......:.:.o.
. . . . . .. . . . . .
~
r--
9 .. ~ "'r
...., .,~..-~ -
f
I-< "-J 10
-0
. ....:':
i
,f"
I
-80
,~,'t'..~~
L
"~
-60
I
-40
.
I
-20
0
LONGITUDE (WEST) Since the beginning of the programme in 1991 16 sections have been accomplished (see Table 1). Figure 1 shows the tracks followed. Although for operational reasons the routes followed by the ships may differ, most of them are sufficiently close together to permit comparison and analysis of variability in time. Figure 2 shows a typical spring and autumn section. Features that can be recognized are the stratified (spring) and vertically mixed (autumn) conditions northeast of longitude 25~ the frontal structure near the Azores near 35~ the thickness of the 18 ~ mode water in the Sargasso Sea and the shallow thermocline in the Antilles area, southwest of 50~ In addition to the T-7 probes intermittently T-5 probes, that go down to 1800 m are launched along the NE part of the section. These observations (not represented on the results shown here) show the extent of the Mediterranean outflow. Preliminary results are published as NIOZ data reports [1]. A programme like this has its potential in its continuation over many years. Similar programmes should make part of a future ocean component of the GCOS (Global Climate Observing System). 1.
T.F.
de Bruin, R.X. de Koster, S.Ober and L. Otto, 1992.
Netherlands XBT programme along the WOCE AX-5 section. NIOZ Data-report 1992-3.
363
Temperature section
Hr. Ms. Banckert
May 1991
,oo 200
1 U
300 r
~- 400 a
500 600
7oot
.
I
'
I
-65
'
-60
;
-55
-50
'
-45
'
-40
I
I
-35
-30
-25
I
I
-
-20
-15
-10
Longitude (~ Temperature section 26
0
Hr. Ms. Karel Doorman
24
'
2
'
'
November 1993
22
0
'
11
/
(
.. fT-x 4"]}./-'x
100 200 ._. 300 E r
~- 400 a
500 600 700 '
I
I
-60
-55
'
I
-50
'
I
-45
'
'
-40 -35 Longitude (~
'
'
-30
I
-25
'
I
-20
'
I
I
-15
-10
Figure 2. Typical spring (top) and fall (bottom) temperature contourplots. The main difference between these plots is the clear presence of a well mixed surface layer of about 70 meters during the fall. The sharp frontal structure at about 35 ~ is the Azores Front. The cold water at 60 - 65 ~ is Antarctic Intermediate Water (AAIW), transported northwards along the South-American east coast into the Caribbean Sea.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
The ASGASEX
365
program
W.A.Oost Royal Netherlands Meteorological Institute (KNMI), De Biit, the Netherlands. Abstract There is a difference of sometimes more than an order of magnitude in the transfer velocity of CO2 at the sea surface if data based on concentration measurements are compared with direct flux measurements with the eddy correlation technique. To explain this discrepancy we performed the international VIERS-I and ASGASEX (for Air Sea GAS EXchange) experiments at the research platform Meetpost Noordwijk off the Dutch coast. The data indicate that the solution of the controversy may lie in the presence of an unexpected vertical concentration (fugacity) gradient of CO2 in water. Acting on this result another experiment (ASGAMAGE) is planned for 1996 to confirm or refute this conclusion. On that occasion the fluxes of a number of other trace gases beside CO2 will be measured with more methods and the CO2 profile will be monitored over the full water column. 1.THE PROBLEM The oceans play a crucial r61e in the carbon dioxide balance of the earth. Over 71)% of the earth's surface is covered with water. The CO2 transport (or CO2 flux) between the oceans and the atmosphere is not well known. Two types of methods have been used to measure CO2 fluxes: those based on chemical concentration measurements, and those using the eddy correlation technique. The chemical methods all need one or more assumptions that are possibly not always fulfilled and they require measurement times in excess of 24hrs. The eddy correlation method measures fluxes directly without such assumptions and within about half an hour, but suffers from a lack of accuracy, because the instruments used are functioning at the very limit of what's technically possible. The differences between the results of the two types of methods, however, are sometimes orders of magnitude, which is way outside the expected uncertainties [1]. 2.METtlOD
AND RESULTS
The 1990 VIERS-1 experiment and the 1993 experiment of the Air Sea GAS EXchange (ASGASEX) project, both supported by NOP, were designed to establish the feasibility of COz eddy correlation measurements from a relatively bulky platform and to explain the differences just mentioned by comparing the results of the two types of methods. The VIERS-I results showed that COz eddy correlation flux measurements were possible at Meetpost Noordwijk (MPN), a research platform 9km off the Dutch coast. ASGASEX'93 took place in September 1993 on the same platform. A large number
366 of environmental parameters, were measured to compare both types of methods.
Fig. 1 Research platform Meetpost Noordwijk The A S G A S E X '93 CO2 flux data, plotted as a function of dCO2 (the difference between pCO2, the concentration in the water, and the concentration in the air) showed no relationship at all. Plotted versus the wind speed, however, they showed a fairly systematic behavior, indicating trustworthy data. Looking more closely into the data we found a correlation of pCO2 with the tide. pCO2 had been measured using a pump, fixed to the platform. The depth from which the water was brought up therefore changed with the tide, so the correlation could indicate a vertical pCO2 gradient at the level of the pump. 120 ]"[-:~::::::.......................................................................... i:i:i:i:::!~:i:ii:i: ::i+:i:i:i::i:i:i:i:i:i:i:i:i:i:i:i:i:i:i:::i:i:i:iiiii:i::: i iii~:iii:i:ii:i~iiiiiii!iiiiiii~:iii:!i~iiiiiiiiiii!iiiiiiii~ii~iiiiiii!iiiii!i:i!i!ii iiiiiii 600
,ooii i~!i ii~iiiiiiil ~!
~ili
~
i
i!
i~
8o i',iiiiiiiiiiil ~!ii!',iiiiiiiiiiiiiiiiiiiiiii!iiiiiiiiii ~ii~iii?,!,'iliii~i~,-~ili!iiiiiii~:iiiiiiiilili~i::!iiiiiiiiiii!iiiiiiU"~!iiill i!i 550 i!iiiiii!i',ii!!iiii,'~i~,iii'ii!iiiliiiiiiiii',i~iiii iii~iii!ii~' ii~i/~i i,!i!iii!i!ii!iiili~i!i!iii,'i,'ii~,!i~!!i ,'iliiliiiiiii,'i~i!i~' ,~i 'ii,'i~i ~o iiiiiii~!',ii!!!iii!i~,i:i~' ii,'iii!iiiii~,~~!?~ii!iiiiiiiiii i:ilili!',!ii!ii!'~.i ii!ii!ii!',iiiii~!i !illi!,'i:,!i!i!i,'iii~' ii~/ :~:ili!:J:iiill iii!:,i:.iiiii:-:.iiiU: "ii!::i!ii::::ii:~i ~ ::!!::!ii~.::~iiii!!ii::i:~i i.~ ~iiiii::!::!:#.ilI~
,o~ ~ ~ o iiiii'ii !iii?
iliiii~iill iii!ii!ii!i'
~!~i~i ~ ",iiiil !iiiiiiii
iiii~#:iiii~:i::i i::i:/ji:::/j::i:~!iiiiii:~!iiiii ii:.
500
i~!i~
~!iiii'i:ilili!iii,i~! 9 ~ii i!i~,ii iii!iii',!ii: ii~ii!i
- o iiii!i:i.iiiiiiiii::i~:: ~..... ..ii iiiiiii!i::,, iiiiiiii .i~iiiiiiiiiiiiiiiiiiii:~!iiill~ ii!iiii i::i::i::i::iii!~iiiili:::.ii ................. ~i:!:i~i!!iiiiiiiiiiiii~ii; ~ii::i!::i::i!i!!ii ::i:::::.i ::i!:.!i
!i,ili~!! 'i~!i',iiii~,~ 9 iiiiiiiiiiiiiiiii i i i'~i i i i!ii i!i~i ~i~i~ ~:~:~'
21.5
22.5
o ~o.
:::!ii::iiiii!iii~::i 400 .....:::i:~:i~ii::~!:::.i!
-80 !:igi!~!~!i~!iiiii!ii::!ii::il;: :~i~5~i!::!i::~!gi1)i!~ii~!ig!!i::)==!g{~iii~)~i!i~:::~:.:::!?::~i~!i!~i!i;ig:!::;iigi:.~ii:: 20.5
450
23.5
350
Fig.2 pCO2 (see text) and the tide (heavy single line) versus time Our measurements also indicated that the wind speed affects the COz concentration in the water. This "wind effect" may be assumed to be absent at wind speed zero.
367 Extrapolating to that value we therefore find the "deep" concentration difference actuaily driving the CO2 transport, in this case 272#atm. 3oo i~iii!i:iiiiiiiliiiiiiii!!iiiiii!!iiiiii 250
!I
i ~ii i!i i i!i i iilii!~i ~;:~iiiiiii!i!~ii i i!i i i i !i !i i i i i i i ii i i ii~ii!i i i !i i i i i i ~i ii il i!i!i i i i!iii~l i~i i ~ i !i !i i i !i i i i i i i i i i i i ~!ii iiiiiiilii!i 200
c~
iiiliiiilli! i i i i i i;i;i iiiiiii;iii~.!~!iiiiiiiiiiiii~F ilii2?
!ii!i!iliiiiiiiiliiii!i!!ii i il i iiiiil;iill!ii
iii
-
::::::::::,:::::
0
...-::::::::~:::::==================================================== . . . . ::::::::::~: -
,:::::~::
iiii ii iiii !i:i i isi~i!iiisi!iss/iss:::::::::::::.:.:.:.
0
"...........
4
::,i:i
:-
8
::
.:.:::,::::::::-~((i:~s~:~s~s~ssii!siiis:~s~s~ ....
12
iiii
:~:",:s~si
~rssl
16
wind speed (m/s)
Fig.3 CO2 concentration difference (sea-air) against wind speed. With this value of the concentration difference we find a good correspondence with parametrisations for the CO2 flux based on concentration measurements (here the one of Wanninkhof [2]). This in principle could solve the controversy mentioned in [1] and so pave the way for improved accuracy in the measurement of COz transport over the oceans.
0.s ~i~ii!~ii~!ii~i~i~i!!~iii~ii~i!i!ii~i!i;;~ii~i;iii!i!~ii~iiiiii~iii~iiii~iiiiiii~i~!;i!i~i~i!ii~i~i~!i~ii~!wa"~"kh~f ~oael w th opco2.272 .,mat., ~iiiiiii!ii!iiiii!iiii!iiiil iil;iii!iiii!i;;;I
i 0!i!i!!!i!!!!!i!!!!! 0o -0.5
-1
0
5
10
15
20
wind speed (m/s)
Fig.4 Measured flux values (open squares) compared with the results of the Wanninkhof parametrization. If, as the ASGASEX'93 results suggest, the CO2 concentration (fugacity) in water can vary with depth in the first few meters below the surface, the use of the CO2 concentration close to the surface when comparing the flux methods - as has been done so far! - will lead to spurious values for the transport coefficients. This insight can explain the differences between the two types of methods and may lead to a greatly improved reliability and accuracy of CO2 transport between sea and air.
368 3. F U R T H E R
NEEDS
The ASGASEX '93 results require confirmation. The crucial data were obtained almost by accident and their interpretation is not unequivocal. More information is needed on the presence or absence of a COz gradient. To improve the accuracy of gas flux data over the sea we furthermore need better and generally applicable measurement methods. These methods must be tested and compared to assess their value. The eddy correlation method, as used presently, can measure the transport of only a single gas simultaneously and these measurements can so far only be done from a stable platform i.e. near the coast. However, due to the high quality of present-day motion sensing devices, it has become possible to allow with sufficient accuracy for the movements of e.g. a ship. Another new technique, the eddy accumulation method, holds promise of simultaneous and direct measurement of the fluxes of a number of gases. This would bring the direct measurement of gas fluxes on the high seas within reach. 4. T H E F U T U R E :
ASGAMAGE
In 1996 there will be another experiment at Meetpost Noordwijk, called ASGAMAGE (MAGE, for Marine Aerosol and Gase Exchange, is a working group of the IGBP IGAC (International Global Atmospheric Chemistry) program. There will be two five week measurement periods (one starting in May, the other in October). Several eddy correlation systems for COz fluxes will be functioning simultaneously and one or two eddy accumulation systems will also be present. One or two ships will make measurements in the MPN area to detect horizontal concentration gradients, one of them may be used for an attempt to perform eddy transport measurements from on board. Fluxes will be measured not only of COz, but also of DMS (with the eddy correlation technique!), CH4 and N20. 5. R E F E R E N C E S 1 W.S.Broecker et al, J. Geophys.Res. 91, (1986), 10517-10527. 2 R.Wanninkhof, J.Geophys.Res.. 97, (1992), 7373-7382.
Institutes (in alphabetical order) participating in A S G A S E X / A S G A M A G E : Bedford Institute of Oceanography (BIO) Dalhousie University Institute of Oceanographic Sciences (IOS) Riso National Laboratory Royal Netherlands Meteorological Institute (KNMI) Netherlands Institute for Sea Research (NIOZ) Southampton University TNO-Physics and Electronics Laboratory (TNO-FEL) University College Galway
Canada Canada UK Denmark the Netherlands the Netherlands UK the Netherlands Ireland
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
369
E1 Nifio and mixed-layer processes M a r c v a n E i j k I en G e r r i t B u r g e r s Royal Netherlands Meteorological Institute, Division of Oceanographic Research, P.O. Box 201, NL-3730 AE De Bilt, Netherlands Abstract E1 Nifio, the dominant form of interannual variability of climate, is caused by interactions between the tropical ocean and atmosphere which exchange heat and momentum. How the tropical ocean responds to atmospheric forcings depends sensitively on its mixing properties. So the nature of the E1 Nifio depends on the oceanic mixing properties too. The present description of mixing processes in ocean general circulation models is not fully satisfactory. Parameters are tuned to the best representation of E1 Nifio without much regard how well they represent small scale processes or not. Even then the models have difficulties in simulating both the background state and E1 Nifio's correctly. We want to use ideas from atmospheric general circulation models, in particular from non-local boundary-layer model schemes, and apply them to ocean models. So far, this has been hampered by a lack of data, but we hope that recent experiments like TOGACOARE will provide enough new information.
1. I N T R O D U C T I O N El Nifio, Spanish for the Christ child, has historically been associated with a weak, warm current appearing along the coast of Ecuador and Peru annually around Christmas, replacing the usual cold waters of the Peru current. Nowadays, the name El Nifio tends to be used for a much larger scale phenomenon that occurs not annually, but every three to seven years in which the normally cold waters over the entire eastern Pacific Ocean show a dramatic warming. Also, very large anomalies in the oceanic and atmospheric circulations and in the global weather are associated with these changes in the equatorial sea surface temperature. E1 Nifio is the dominant form of interannual variability of climate. Severe effects occur all over the world, like droughts in Australia, and reduced fishing near the coast of Peru. At the time of this conference, it looks like that a new E1 Nifio is about to start. One of indicators is the low pressure difference anomaly between Darwin and Tahiti over the last half year, another indicator is that the sea surface temperature (SST) around the date line is higher than normal and that this area of warm SST is moving westwards. 1Supported by the Nederlandse Organisatie voor Wetenschappelijk Onderzoek (NWO), project VvA-2-250.
370 2. O C E A N - A T M O S P H E R E
INTERACTIONS
The E1 Nifio phenomenon (Philander, 1990) is caused by interactions between the tropical ocean and atmosphere which exchange heat and momentum. The atmosphere influences the ocean mainly through the stress exerted by the surface winds, and to a lesser extent through the heat flux and the precipation and evaporation. The ocean in turn influences the atmosphere through the sea surface temperature. During an E1 Nifio, weakened easterlies cause warm western Pacific water to flood over the cold eastern water and to slow down the upwelling of cold water at the coast of South America. The barrier between warm surface waters and cold deep ocean waters, the thermocline, which normally slopes upward from west to east, becomes more fiat. This also contributes to the temperature rise of the waters in the east. The shift in sea surface temperature is accompanied by a shift of the major rain zone to the east. Related adjustments in the atmosphere cause a further weakening of the easterlies in the central Pacific. In this way all the coupled influences amplify each other, until eventually one can speak of a full-blown E1 Nifio.
3. F R I C T I O N A N D M I X I N G IN T H E T R O P I C S Friction plays a more important role in the dynamics of the tropical ocean-atmosphere system than in mid-latitude systems. This is because at the equator pressure gradients can be balanced only by frictional forces, while at higher latitudes there is the possibility of a balance between the Coriolis force and the pressure force. Indeed, strong ocean currents are found along the equator. In some places, the currents have such a strong horizontal or vertical shear that mixing is much enhanced. Horizontal friction is the main limiting mechanism for equatorial currents (Crawford 1982). Vertical mixing also affects the strength of the current, but, more important, controls the shape of the thermocline (Pacanowski and Philander, 1981). In a large region in the Western Pacific, there is a deep (about 200m), mixed layer of warm water (around 30 C). The heat which can be stored in this deep warm pool plays an essental role in triggering and maintaining E1 Nino events. The evolution of the deep warm pool depends on mixed-layer processes. All this has been confirmed by numerous studies, including some where KNMI was participating (M. Latif et al., 1994). At the equator, in the Pacific, a strong eastward countercurrent, the Equatorial Undercurrent, flows below the westwards flowing surface waters. The core of this current follows approximately the thermocline. In the high-shear regions around the undercurrent, mixing is substantially enhanced. The dynamics of the undercurrent is quite sensitive to this mixing. General Circulation Models without a proper representation of the undercurrent have difficulties in making good SST simulations. Mixing properties affect both the E1 Nifio's, and the background mean annual cycle. In general, ocean general circulation models (OGCM's) have difficulties in reproducing both good E1 Nifio's and a good mean annual cycle. But this is necessary if one wants to study how global climate change might affect the character of E1 Nifio.
371 4. M I X I N G F O R M U L A T I O N IN O C E A N M O D E L S The present description of mixing processes in OGGM's is not fully satisfactory. The usual mixing formulation for the shear around the undercurrent (Pacanowski and Philander, 1981) is rather ad hoc and parameters are tuned to the best representation of El Nifio without much regard how well they represent small scale processes. Even then the models have difficulties in simulating both the background state and E1 Nifio's correctly. For the surface mixed layer, mixed-layer schemes of the Niiler-Kraus (1977) type often are used, but they were originally developed for mid-latitude situations, and whether they can simulate the deep warm pool in the West-Pacific and the surface waters above the undercurrent is questionable.
5. N E W M I X I N G S C H E M E S Models of the atmospheric boundary layer, both stable and unstable, are more versatile and seem to have reached a more advanced state of development (Holtslag and Nieuwstadt 1986). Holtslag and Moeng (1991) have found a way to parameterize countergradient heat transport in the convective boundary layer. A module based on this parameterization was made for the NCAR Community Climate Model by Holtslag and Boville (1993). We want to apply these ideas, in particular the non-local schemes, to the oceanic boundary layer. First a 1-D model for a column of water will be made which can simulate well the detailed evolution, including day-night contrast and how the mixed layer responds to a shower. So far, testing such models has been hampered by a lack of data, but we hope to be able to do that with data from recent experiments like TOGA-COARE. Next a parameterization will be devised which is suitable for implementation in an OGCM with a resolution of the order of 15m in the first 150m, and the consequences for E1 Nifio will be investigated.
6. R E F E R E N C E S
Crawford, W.R. 1982: Pacific equatorial turbulence. J. Phys. Ocean. 12, 1137-1149. Holtslag, A.A.M., and Nieuwstadt, F.T.M. 1986: Scaling the atmospheric boundary layer. Bound.-Layer Meteor. 36, 201-209. Holtslag, A.A.M., and Moeng, C.-H. 1991: Eddy diffusivity and countergradient heat transport in the convective atmospheric boundary layer. J. Atmos. Sc.48, 16901698. Holtslag, A.A.M., and Boville, B.A. 1993: Local vs. non-local boundary-layer diffusion in a global climate model. J. Clim. 6, 1825-1842. Latif, M., T. Stockdale, J. Wolff, G. Burgers, E. Maier-Reimer, M.M. Junge, K. Arpe and L. Bengtsson 1994: Climatology and variability in the ECHO coupled GCM. Tellus 46A, 351-366. Niiler, P.P. and Kraus, E.B. 1977: One dimensional models of the upper ocean. In: Mod-
372 elling and prediction of the upper layers of the ocean, E.B. Kraus, ed., Pergamon Press, 143-177. Pacanowski, R.C. and Philander, S.G.H. 1981: Parameterization of vertical mixing in numerical models of the tropical oceans. J. Phys. Ocean. 11, 1443-1451. Philander, S.G.H., 1990: El Nifio, La Nifia and the Southern Oscillation. Academic Press, 291 pp. Yin, F.L., and E.S. Sarachik 1993: Dynamics and Heat Balance of Steady Equatorial Undercurrents. J. Phys. Ocean. 23, 1647-1669.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
375
LAND ICE AND SEA L E V E L CHANGE I n s t i t u t e for Marine and Atmospheric Research, Utrecht University in co-operation with the University of A m s t e r d a m and the Free University (Amsterdam) sponsored by - Dutch National Research P r o g r a m m e on Global Air Pollution and Climate Change - N e t h e r l a n d s Organization for Scientific Research (NWO) - N e t h e r l a n d s Antarctic Research P r o g r a m m e - E u r o p e a n Commission (ENVIRONMENT) Abstract Sea-level change is an important issue in the greenhouse problem. All workers agree t h a t predictions made so far have a high degree of uncertainty. Comparable contributions to this u n c e r t a i n t y come from limited knowledge of future emissions of greenhouse gases, different opinions concerning the response of the climate system, and inadequate knowledge of the sensitivity of land ice to climate change. The goal of this project is to improve estimates of the contribution of land ice masses to sea-level change in the coming 150 years.
MASS B ~ C E
OF THE GREENLAND ICE S H E E T
The mass balance of the Greenland ice sheet has been studied with an energy balance model. The mass balance is generated from climatological input. Data from several field experiments have been used to improve the p a r a m e t e r i z a t i o n of energy transfers between atmosphere and surface. The grid resolution is 20 km. The picture below shows the calculated surface mass balance for the "reference case". Sensitivity tests reveal that a 1K warming implies a 0.30 mm/year sea-level rise a 1K warming (+ dP) implies a 0.21 mm/year sea-level rise a 10% increase in cloudiness implies a 0.02 mm/yr sea-level drop a 0.02 decrease in albedo implies a 0.17 mm/year sea-level rise [dP is a change in precipitation in proportion to saturation vapour pressure]
376
m of water equivalent per year (nonlinear scale) +2.0
+1.0
+ 0.5
+ 0.25
+o.o
- 2.0
Conclusion:
Greenland ice sheet, +IK: 0.21 mm/year sea-level RISE (best estimate) G L A C I E R S AND SMALL ICE CAPS A model has been designed t h a t s i m u l a t e s m a s s b a l a n c e profiles on glaciers. It has been tested on 12 glaciers for which good observations exist. After careful calibration a large n u m b e r of sensitivity tests have been carried out. There appears to be a significant correlation between glacier sensitivity and precipitation regime. The figure below shows the result for an experiment with u n i f o r m 1K w a r m i n g and an i n c r e a s e in p r e c i p i t a t i o n p r o p o r t i o n a l to s a t u r a t i o n vapour pressure of the air. The m e a n loss of ice, averaged over the entire glacier, is shown for the 12 glaciers studied.
377
1 " ~>, E
0.8
.,_....
0.6
.o_ ,,.,,_. 0
0.4
__o t--
0.2 E ~
0
1
0
2
3 4 5 annual precipitation (m)
,!
I
6
7
,,
8
Extrapolation of this result to all glaciers and small ice caps outside Greenland and Antarctica yields a sea-level rise of 0.46 mm/yr for a uniform 1K warming (this includes increasing precipitation). This is about half the value of the 1.2 + 0.6 mm/yr used in IPCC-1990. The difference is due to an earlier overestimation of glacier sensitivity in the dry subpolar regions, where a large amount of glaciers and small ice caps are located (see below). 100
~E o
o
80
!
I
i
I
100 Qlacier reqions
~, .-
60
o
~
4O
@
20 0 -6o
-30
0
30 latitude
60
90
(")
Conclusion: Glaciers and ice caps, +IK: 0.46 mm/year sea-level RISE (best estimate)
SNOW ACCUMULATION ON THE ANTARCTIC ICE SHEET A simple meteorological model has been developed to simulate the temperature and precipitation distribution over the Antarctic continent. The model is two-dimensional (vertical plane, see figure below), and has 4 layers: stratosphere, troposphere, boundary layer and surface of the ice sheet. It has a detailed radiation scheme for short and long wave radiation.
378 The k a t a b a t i c outflow is explicitly calculated and drives the circulation over the ice sheet. The b o u n d a r y layer has two shear zones: one at the ice sheet surface a n d one at the top of the b o u n d a r y layer, where significant e n t r a i n m e n t t a k e s place. B o u n d a r y layer depth is a prognostic variable.
stratosphere (only radiation)
troposphere D~E
:: .... humid a ir
OCE/N Moisture is brought to the ice sheet by the r e t u r n flow in the free troposphere. Precipitation occurs because of cooling of the air (due to uplift and a negative r a d i a t i o n balance). The moisture budget at the surface has four contributions: - precipitation - riming - evaporation - divergence of snow drift W h e n r u n with a p p r o p r i a t e b o u n d a r y conditions (annual m e a n insolation a n d t e m p e r a t u r e at the ocean boundary), the model gives a satisfactory s i m u l a t i o n of the meridional profiles of t e m p e r a t u r e and a c c u m u l a t i o n on the Antarctic ice sheet (annual m e a n state). In case of a w a r m e r climate, snow a c c u m u l a t i o n increases because the "moisture pump" intensifies. The increase is p a r t l y c o m p e n s a t e d by l a r g e r evaporation on the steep slopes of the ice sheet, however. For a uniform 1K w a r m i n g , the model predicts an increase in snow a c c u m u l a t i o n t h a t is equivalent to a 0.27 m m / y e a r sea-level drop.
Conclusion: Antarctic ice sheet, +IK: 0.27 m m / y e a r sea-level DROP (best estimate)
379 PAPERS FROM THIS PROJECT (printed or accepted, status November 1994)
R S W van de Wal, J Oerlemans and J C van der Hage (1991): A study of ablation variations on the tongue of Hintereisferner, Austria. Journal of Glaciology 38, 319-324. J Oerlemans (1992): Climate sensitivity of glaciers in southern Norway: application of an energy-balance model to Nigardsbreen, Hellstugubreen and Alfotbreen. Journal of Glaciology 38, 223-232. J Oerlemans and J P F Fortuin (1992): Sensitivity of glaciers and small ice caps to greenhouse warming. Science 258, 115-117. 4. J Oerlemans and H F Vugts (1992): A meteorological experiment in the melting zone of the Greenland ice sheet. Bulletin of the American Meteorological Society 74, 355-365. J P F Fortuin and J Oerlemans (1993): An axi-symmetric atmospheric model to simulate the mass balance and temperatue distribution over the Antarctic ice sheet. Z. Gletscherk. Glazialgeol. 26, 31-56. J Oerlemans (1993): Modelling of glacier mass balance. In: Ice in the Climate System (ed. W R Peltier), NATA ASI Series, Vol. 1-12 (Springer), 101-116. M R van den Broeke, P G Duynkerke and J Oerlemans (1994): The observed katabatic flow at the edge of the Greenland ice sheet during GIMEX-91. Global and Planetary Change 9, 3-15. P G Duynkerke and M R van den Broeke (1994). Surface energy balance and katabatic flow over glacier and tundra during GIMEX-91. Global and Planetary Change 9, 17-28. .
R S W van de Wal and A J Russell (1994): A comparison of energy balance calculations, measured ablation and meltwater runoff near Scndre Strcmi~ord, West Greenland. Global and Planetary Change 9, 29-38.
10 A Meesters, E Henneken, N J Bink, H F Vugts and F Cannemeijer (1994): Simulation of the atmospheric circulation near the Greenland ice margin. Global and Planetary Change 9, 53-67.
380 11 E Henneken, N J Bink, H F Vugts, F Cannemeijer and A Meesters (1994): A case study of the daily energy balance at the VU-GIMEX camp. Global and Planetary Change 9, 69-78. 12 W Greuell and T Konzelmann (1994): Numerical modelling of the energy balance and the englacial temperature of the Greenland ice sheet. Calculations for the ETH-Camp location (West-Greenland, 1155 m a.s.1.). Global and Planetary Change 9, 91-114. 13 R S W van de Wal and J Oerlemans (1994): An energy balance model for the Greenland ice sheet. Global and Planetary Change 9, 115-131. 14 T Konzelmann, R S W van de Wal, W Greuell, R Bintanja, E A C Henneken and A Abe-Ouchi (1993): Parameterization of global and longwave incoming radiation for the Greenland ice sheet. Global and Planetary Change 9, 69-78. 15 J Oerlemans (1994)" Quantifying global warming from the retreat of glaciers. Science 264, 243-245. 16 M R van den Broeke, P G Duynkerke and E A C Henneken (1994): Heat, m o m e n t u m and moisture budgets of the katabatic layer over the melting zone of the West-Greenland ice sheet in summer, Boundary-Layer Meteorology, in press. 17 A G C A Meesters (1994): Dependence of the energy balance of the Greenland ice sheet on climate change: influence of katabatic wind and tundra. Quarterly Journal of the Royal Meteorological Society 120, 491-517. 18 F G M van Tatenhove, C Roelfsema, G Blommers, A van Voorden (1995): Change in position and altitude of a small outlet glacier during the period 1943-1992, Leverett glacier, West Greenland. Annals of Glaciology, in press. 19 F G M van Tatenhove and O B Olesen (1995) Ground temperature and related permafrost characteristics in west Greenland. Permafrost and Periglacial Processes, in press About 12 additional papers have been submitted
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
381
Stresses in the lithosphere caused by glacial loads P. Johnston and S. Cloetingh
Faculteit der Aardwetenschappen, Vrije Universiteit, De Boelelaan 1085, 1081 HV, Amsterdam
Abstract Simple elastic theory shows that horizontal stresses caused by a surface load which has a similar wavelength to the flexural wavelength of the lithosphere exceeds the vertical stress by several times. The enhancement factor is very sensitive to the wavelength. For glacial timescales, the wavelength of the British ice sheet was close to the flexural wavelength, while the Fennoscandian one was much larger. For models which allow time-dependent glacio-isostatic rebound, it is inferred that the stresses due to both the British and Fennoscandian ice loads were large until the end of glaciation, but very small by the present.
1
T w o - l a y e r e l a s t i c over fluid fiat E a r t h m o d e l
For a uniform, incompressible elastic layer overlying an inviscid fluid halfspace, the deformation in response to surface loads of arbitrary wavelengths can be calculated analytically. To determine the response to more general loads in two dimensions or with axial symmetry, the load may be expressed as a sum of Fourier or Hankel components and the total response is the sum of responses to each component. The continuum version of Newton's second law is (eg [5, 8]) , (o)
t!~ )+tp,juj),~z3,3
p(A)g(O)
~ -p
(0) ~(A)
~
=0
(1)
where t is the Cauchy stress tensor, p is the pressure (= --tkk/3), p the density and g the gravity. The superscripts (0), (5) and (A) indicate the initial, material incremental and local incremental fields respectively. If perturbations of the gravitational field are ignored, and incompressibility assumed, the last two terms do not enter the equation. Calculations have been made for the deformation and stress caused by a load of given wavelength for a 65 km thick elastic incompressible plate of density 3000 kg/m 3 and shear modulus 4 • 1010 Pa overlying an inviscid half-space. The elastic thickness is chosen to coincide with estimates of elastic thickness determined from postglacial rebound in the British Isles [4]. The fiexural wavelength of the elastic layer in the thin-plate approximation is A] -- 27c(Eh3/(12(1
-
l]2)pg))
1/4 --
660 km
(2)
where E is the Young's modulus, h is the thickness of the elastic plate and ~ is Poisson's ratio for the elastic layer. In Figure 1, the dimensionless horizontal stress is plotted versus wavelength of the load and depth within the elastic layer. The vertical stress for the same load is unity at the surface and
382
attenuates with depth, more sharply for short wavelengths than long wavelengths. The maximum amplification of the horizontal stress occurs at wavelengths close to the flexural rigidity where the bending of the lithosphere is greatest.
0
~
.....
"
. ....................
,
9
,.,
, 9............ -..,
.- .
... ...
9. ~
,.......................
,.
.................
:
"~..:
.....
~
oe,
Figure 1: Horizontal and vertical stress as a function of wavelength and depth within the lithosphere normalised by the weight of the load. The maximum response factor for the horizontal stress is close to the flexural wavelength of the lithosphere.
1.1
Results for axisymmetric
ice l o a d s w i t h e l l i p t i c p r o f i l e
The deformation and stress field has been calculated for two ice sheet models which :represent approximately the British ice sheet and Fennoscandian ice sheet at the last glacial maximum. Of particular interest is the maximum shear stress, which is equal to half the difference between the maximum and minimum principal components of stress. In this geometry, they will always be the radial horizontal stress and vertical stress. The vertical stress is constrained to be equal to the load at the surface and for ice loads of moderate to large lateral extent, there is little variation with depth within the lithosphere. The horizontal stress is much more dependent on the horizontal extent of the load compared with the flexural wavelength of the lithosphere. Because the smaller ice sheet (diameter 660 kin) is much closer in lateral extent to the flexural wavelength, it produces larger stresses, despite being just over half the thickness. The maximum shear stress is related to the likelihood of seismicity and faulting occurring. If the shear stress is in excess of 10 MPa, then pre-existing faults may be re-activated [3]. Because the horizontal stress is much larger in magnitude than the vertical stress, it is the main contributor to the maximum shear stress. In Figure 2, we compare the radial stress predicted for two ice sheets of elliptic profile, one with radius 330 km to model the British ice sheet and the other with radius 1000 km for the Fennoscandian ice sheet.
2
Spherical Maxwell viscoelastic model
The deformation of the lithosphere has been calculated using the full equation (1) above, without any approximations, a spherical Earth model, and an elastic lithosphere overlying a Maxwell
383
g 3000
i 2000 .o
1000 0
o
0
i
20
[ "
i
I i~.,/
I
I
I
i
I
o ...................... i ................. ~ ................. i .......................................... i ..........................................
a=
.qp
60
0
.... I
I
I
I
0
500
1000
1500
2000
0
500
1000
Radial distance (km)
Figure 2: Horizontal radial stress (MPa) as a function of distance from the centre of the load and depth within the lithosphere for an axisymmetric ice load with elliptic semi-profile for two ice sheets of different lateral extent. The maximum compressional horizontal stress occurs at the centre of the ice sheet with a small amount of extension outside the edge of the ice sheet. viscoelastic mantle, with seismically determined elastic properties [2] and mantle viscosities determined from fitting relative sea-level observations [4]. A glaciation/deglaciation cycle has been used to approximate the growth and decay of the Fennoscandian ice sheet. Figure 3 shows the maximum stress difference at the end of deglaciation and at the present. Because the Maxwell viscoelastic theology behaves elastically on short timescales and viscously on long timescales, the effective flexural wavelength is time-dependent but with a lower bound of about 660 km as in the model above. Therefore, the maximum stress difference is somewhat smaller than in Figure 2. The values of stress are quite strongly dependent on the elastic properties of the various layers as seen by the sharp variation in stress at 15 and 25 km depth. Because most of the postglacial rebound is complete by the present, there remains little residual stress difference near the surface in the model. However, at the end of the glaciation, the stresses are still large enough to cause seismicity. This is consistent with observations of late glacial faulting [6] and the predominance of the NW-SE regional stress field in Fennoscandia [7] rather than a radial pattern which would be caused by postglacial rebound.
3
Conclusions
The maximum shear stress in the lithosphere has been calculated using the same models which fit relative sea-level observations. The calculations indicate that the stresses were large enough to cause faulting during and after the end of deglaciation, but the residual stress today is probably too small to be observed in comparison with the prevailing NW-SE pattern in Europe due to ridge push from the Mid-Atlantic ridge. The stresses at glacial maximum may have been larger for the British Isles than for Fennoscandia. The results of the modelling are consistent with
384
0 I
,
I
--U
I
I
I
I
IL ~
I
I
I
I
I
L
40 60 0
500
1000 ' 1500 Distance (km)
2000
0
' 500
' 1000 ' 1500 ' 2000 Distance (km)
Figure 3: Maximum shear stress (MPa) for the Fennoscandian ice sheet at the end of deglaciation (left) and at the present (right) for an axisymmetric ice sheet with maximum radius of 1000 km at 18 thousand years before present (kaBP) and finished melting at 8 kaBP. observations of late glacial faulting in both Fennoscandia [6] and Great Britain [1], and with the observed stress field and seismicity pattern in Scandinavia today [7].
References [1] C. A. Davenport, P. S. Ringrose, A. Becker, P. Hancock, and C. Fenton. Geological investigations of late and post glacial earthquake activity in Scotland. In S. Gregersen and P. W. Basham, editors, Earthquakes at North-Atlantic Passive Margins: Neotectonics and PostglaciaI Rebound, pages 175-194. Kluwer, Dordrecht, 1989. [2] A. M. Dziewonski and D. L. Anderson. Preliminary reference Earth model. Phys. Earth Planet. Inter., 25:297-356, 1981. [3] A. C. Johnston. The effect of large ice sheets on earthquake genesis. In S. Gregersen and P. W. Basham, editors, Earthquakes at North-Atlantic Passive Margins: Neotectonics and PostglaciaI Rebound, pages 581-599. Kluwer, Dordrecht, 1989. [4] K. Lambeck. Glacial rebound of the British Isles. II. A high resolution, high-precision model. Geophys. J. Int., 115:960-990, 1993. [5] L. E. Malvern. Introduction to the mechanics of a continuous medium. Prentice-Hall, Inc., New Jersey, 1969. [6] R. Muir Wood. Extraordinary deglaciation reverse faulting in northern Fennoscandia. In S. Gregersen and P. W. Basham, editors, Earthquakes at North-Atlantic Passive Margins: Neotectonics and PostgIacial Rebound, pages 141-173. Kluwer, Dordrecht, 1989. [7] B. Miiller, M. L. Zoback, K. Fuchs, L. Mastin, S. Gregersen, N. Pavoni, O. Stephansson, and C. Ljunggren. Regional patterns of tectonic stress in Europe. J. Geophys. Res., 97:1178311804, 1992. [8] D. Wolf. Viscoelastodynamics of a stratified, compressible planet: incremental field equations and short- and long-time asymptotes. Geophys. J. Int., 104:401-417, 1991.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
385
The response of permafrost ecosystems to climate change Eduard A. Koster a a Department of Physical Geography, University of Utrecht, P.O.Box 80.115, 3508 TC Utrecht, The Netherlands
Abstract Permafrost areas are extremely sensitive to change as permanently frozen ground is directly temperature dependent. The heat exchange interactions of climate and permafrost in the socalled "buffer layer" are highly complicated, as ground temperatures are strongly influenced by local factors (snow cover thickness and duration, vegetation, organic layer and soil characteristics), which are interrelated with climate. Variations in these variables may either enhance or counteract each other, which makes it difficult to predict the accumulated effect of all changes. However, several simulation experiments indicate large shifts of permafrost boundaries due to a temperature increase, resulting in extensive permafrost degradation (thermokarst). Geothermal profiles of the upper 100-200 metres of permafrost, which yield a temporally integrated record of air temperature changes in the past decades to centuries, show significant changes. However, the quantitative relationships between permafrost degradation and biogeochemical processes, including the generation or uptake of carbon dioxide and methane are still largely unknown. 1. INTRODUCTION Approximately 25 % of the land surface of the Northern Hemisphere is underlain by permafrost. A major part of this huge area is designated as discontinuous permafrost (approx. 17.3 million square kms), the southern boundary of which roughly coincides with a mean annual air temperature of-1 to -2~ Near its southern boundary it occurs in isolated patches or islands and is sometimes referred to as sporadic permafrost. Approximately north of the -6 to -8~ isotherm continuous permafrost (approx. 7.6 million square kms) occurs. Moreover, an area of approximately 2.3 million square kms, mainly at lower latitudes, is covered by Alpine or mountain permafrost. Permafrost areas will be among the most heavily affected parts of the world in the event of accelerated future warming [1, 2, 3, 4]. The objectives of this review paper are: 1) to emphasize the complex interrelations in the atmosphere-"buffer layer"-permafrost system, 2) to summarize permafrost response to past and future temperature changes and 3) to indicate the uncertainties with respect to permafrost ecosystems as sources or sinks of carbon dioxide and methane. 2. HEAT EXCHANGE AND THE ACTIVE LAYER In permafrost areas several types of temperatures are defined (Fig.l), depending on where they are measured [1, 2]. The mean annual air temperature (MAAT) usually is several degrees lower than the mean annual ground temperature (MAGT), the latter being defined as the ground temperature at a depth where temperature fluctuates by less than 0.1~ per year. Above this depth the ground is subjected to strong seasonal fluctuations. Nevertheless, mean annual ground surface temperature (MAGST) can be deduced by upward extrapolation of the geothermal gradient, provided the measured gradient has achieved equilibrium and there are no recent climatic changes. Extrapolation of the geothermal gradient downwards will lead to an approximation of the depth of the permafrost base. Where the geothermal heat flow is
386 constant, the geothermal gradient is inversely proportional to conductivity. The thermal conductivity in its turn strongly varies depending on soil properties and sediment texture. The water or ice content is especially critical. The geothermal gradient in different types of sediment ranges from about l~ for sandy, relatively ice-rich material (high conductivity ~ low gradient ~ thick permafrost) to about 1~ for fine-grained, relatively ice-poor material (low conductivity - high gradient ~ thin permafrost).
/ MAAT 2 M(A)SST^ MAGSTu MAPST2
/ 5~'~.*~
,a,er
] [ [ ]f'J
" f - ~ l ~J~-~ X I~ -"
~**]~r'~..~ Unfrozen ground
, Tmean
. / ".~" " ;~t1~ betweenbaseof .'~" :'~l~l~Jll p3t~Vm~IpYs~rtae&bl
4 MAGT 6 8 10 12 14 16 20 2
ItHH-----/~~--iliiiila~mepi
IIIlll ,IIIIIL I IIL~
P2eertm[ IaJ fl lrlOJ lsl tl ~u
+ !
IIIIIII
I[l l IJlllJlll[l
I (5-1bin)
JJJlJJJ~Unfrozenground
,
withinpermafrost
IlflllJ ~J~L"~ mafrotUnfr~176 bellowpe
120 122
-;o
-~
o
Temperature
~ (~
,'o
Figure 1. Permafrost terminology and ground temperature profile. The heat exchange interactions of climate and permafrost through the active layer are strongly influenced by the vegetation cover, the (seasonal) snow cover thickness and duration, the organic soil (if present) and the mineral soil. Variations in these variables may either enhance or counteract each other [1, 2]. The snow cover thickness and duration probably is the most important factor due to its insulating properties. The conductive capacity of dry new snow is 1/5 of that of old compacted snow or dry sand, or 1/20 of that of wet sand or 1/25 of that of ice. Generally, a snow cover keeps the ground warmer, as it can heat up in summer, but is hampered in cooling down in winter. Snow cover thickness and duration depend to a large extent on the presence and nature of vegetation. Vegetation mainly has an effect upon surface temperatures by shading, thus cooling the ground. Moreover, vegetation prevents radiation back into the sky at night, and the soil is dried by evapotranspiration and this decreases its heat capacity. The organic layer strongly influences ground temperatures by its differential insulation capacities. In summer conductivity of a dry peat layer is extremely low (< 0.1 WmK). In winter, however, the organic layer freezes leading to conductivities many times higher (> 1.0 WmK), and consequently the ground cools off. To a lesser extent the properties of the mineral soil also influence the conductivity. Both changes in snow regime and in vegetation will determine the moisture condition of the soil and thereby the thermal conductivity of the materials.
387 3. HISTORIC CHANGES IN G E O T H E R M A L R E G I M E The above-mentioned uncertainties notwithstanding, in principle a rise in MAAT and consequently in MAGST will have the following effects. Firstly, the thickness of the active layer will increase. Secondly, the temperature profile within the permafrost will adjust itself to the new MAGST. The rate with which this happens depends on the thermal conductivity of the permafrost and the ice content of the ground. Response times of the active layer are in the order of years to tens of years. Eventually permafrost will decrease in thickness. During Pleistocene glacial episodes the permafrost area was probably twice as extensive as the present-day extent, whereas during the "climatic optimum" of the Holocene the southern limit of discontinuous permafrost in the Soviet Arctic was up to 600km north of its present position. In historic times significant changes in permafrost zonation have also been documented; e.g. in the southern part of the discontinuous permafrost zone in Manitoba (Canada) the southern limit has shifted northwards over the past 150-200 years and the areal extent of permafrost terrain has diminished strongly. In the Mackenzie Valley (Arctic Canada) MAGT values increased by about 3~ during a recent warm period (late 1800s to the 1940s) and have since decreased about I~ resulting in a shift in the continuous-discontinuous permafrost boundary of several hundreds of kms. Thus, the analysis of permafrost temperature as a function of depth appears to yield an integrated record of air temperature changes in the past. This has been well-documented by temperature profiles obtained from boreholes in the Alaskan Arctic Coastal Plain [5]. A vast number of these temperature profiles shows a distinct curvature towards higher temperature near the surface. The exact onset of warming seems to vary between locations, but they all indicate a warming in the range of 1.5-3~ during the last century, which seems to be in agreement with a similar trend in air temperatures as shown by regional weather records. Time series of annual permafrost ground temperatures in shallow drill holes (<60m) in northern Alaska show that these temperatures have cycled in the decade from 1983 to 1993. Whatever the causes for this cyclic event may be (changes in air temperature, snowfall, solar irradiance), this detailed record clearly illustrates the rapid response of permafrost ground temperatures to external factors [6]. 4. SIMULATING PERMAFROST ZONATION CHANGES The greatest changes can be expected to occur in those parts of the discontinuous permafrost zone where MAGT values are close to 0~ A model of heat and water transport has been constructed to simulate changes in the extent of permafrost in the USSR [7]. The model results based on a mean 2~ global warming (assumed to represent a 7-9~ increase in winter and a 4-6~ increase in summer in the latitudinal zone 55-70~ suggest that in about 50 years the area occupied by continuous permafrost might be 15-20 % smaller than at present. The vertical movement of the permafrost table is expected to be 0.5-0.8m/yr during the first year of the model run and to decrease exponentially over time. After 50 years of the model run the maximum depth of the thawed layer was 7m. Comparable rapid changes in active layer thickness have been computed by others [8]. All model simulations suggest that due to climate warming the active layer in continuous permafrost will thicken, but that permafrost will remain, in contrast to discontinuous permafrost which might locally disappear completely
[9]. 5. P E R M A F R O S T ECOSYSTEMS The cold climate, short growing season and nutrient-poor soils in northern ecosystems have led to adaptations in plants and in soil organisms. In addition, the low rate of decomposition relative to primary production has caused a large accumulation of carbon as dead organic matter during the Holocene [10]. According to various sources northern (tundra and taiga) ecosystems contain 400-550Gt of carbon stored as soil organic matter [2]. Present-day
388 emission of methane is dominated by northern wetlands, especially between 50-70~ and particularly in western Siberia and east-central Canada where extensive bog and fen areas occur. Wet coastal arctic tundras could become a net carbon dioxide source at higher temperatures. Even slight decreases in water level due to higher evapotranspiration could strongly accentuate the temperature effect. This effect seems to be stronger than the direct effect of increasing atmospheric carbon dioxide content on carbon uptake in the tundra. Net primary production will increase somewhat due to increasing carbon dioxide levels, provided that nutrients are not limiting growth. However, in periglacial environments nutrients usually are a limiting factor. On the other hand, if temperature rises and drier conditions prevail, decomposition of organic matter may accelerate, causing more nutrients to become available. Warming of Arctic and boreal wetland ecosystems will almost certainly be followed by increased methane emissions, which currently account for the release of about 25-40Mt of methane to the atmosphere annually. It is estimated that a 4~ rise in these regions could lead to a 45-65% increase in methane release. If warming is accompanied by drying, then there may ultimately be a reduction of methane release. The only valid conclusion at present can be that the net results of these interactive effects are poorly known [11, 3]. 6. CONCLUDING REMARKS Widespread thermokarst development is likely to occur as a consequence of permafrost degradation, particularly of ice-rich ground [7]. Regional lowering of the permafrost table will cause the development of large unfrozen near-surface aquifers with perennial groundwater flow. In practical terms, increased terrain instability, especially in the first phases of permafrost degradation, would lead to major concerns for the integrity and stability of roads, railways, airstrips, dams, reservoirs and other foundations in affected areas. Deepening of the active layer would subject foundations to continuing deformations as a result of thaw settlement. Many of the technologic and socio-economic implications of climate change at higher latitudes were recently reviewed [12, 13]. However, there is an acute need for a circumpolar network of ground temperature stations, ideally at or near ITEX (international Tundra Experiment) locations, to provide information about the nature, extent and rate of permafrost change that will be necessary for realistic regional and global planning activities [3, 121 . 7. REFERENCES 1 E.A. Koster and M.E. Nieuwenhuijzen, Catena Supplement 22 (1992) 37-58. 2 E.A. Koster, In: N. Roberts (ed.) The changing global environment. Blackwell (1994) 127149. 3 F.E. Nelson, A.H. Lachenbruch, M.-k. Woo, E.A. Koster, T.E. Osterkamp, M.K. Gavrilova and Cheng Guodong, Permafrost Sixth Intern. Conf. Proc. (Vol.2) South China Univ. of Techn. Press (Wushan, Guangzhou, China) (1993) 987-1005. 4 E.A. Koster, M.E. Nieuwenhuijzen and A.S. Judge, Glaciological Data Report GD-27 (1994) World Data Center A for Glaciology (Snow and Ice), Boulder, USA. 5 A.H. Lachenbruch and B.V. Marshall, Science 234 (1986) 689-696. 6 T.E. Osterkamp, T. Zhang and V.E. Romanovsky, Permafrost and Perigl. Proc. 5 (1994) 137-144. 7 F.E. Nelson and O.A. Anisimov, Permafrost and Perigl. Proc. 4 (1993) 137-148. 8 D.L.Kane, L.D. Hinzman and J.P. Zarling, Cold.Reg.Sc. and Techn. 19 (1991), 111-122. 9 A.V. Pavlov, Permafrost and Perigl. Proc. 5 (1994) 101-110. 10 G.M. Marion and W.C. Oechel, The Holocene 3 (1993), 193-200. 11 T.P. Kolchugina and T.S. Vinson, Permafrost and Perigl. Proc. 4 (1993) 149-163. 12 B.Blair Fitzharris, IPCC update (in press): Impacts Assessments Report Chapter IIA6: The Cryosphere. 13 T.S. Vinson and D.W. Hayley (eds.), J. Cold Regions Engin. 4 (1990).
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
391
Climate change scenarios for impact studies in the Netherlands A.M.G. Klein Tank, T.A. Buishand, J.J. Beersma and G.P. Ktinnen Royal Netherlands Meteorological Institute (KNMI), P.O. Box 201, 3730 AE De Bilt, e-mail: kleintan @knmi.nl
Abstract Observed relations between meteorological elements in the present-day climate are used to transform an observed daily time series into representative daily time series of a possible future climate. For point precipitation the method needs reasonable guesses of the large-scale changes in the seasonal means of surface air temperature (or the combination of surface air temperature and surface air pressure). The transformation method is described and an example is given for De Bilt (the Netherlands).
1. INTRODUCTION A climate change scenario can be defined as a meteorologically consistent picture of a possible future climate that can be used to study the response of environmental and social systems to future climate change. General Circulation Models (GCMs) are the natural tools for obtaining such a picture, since they synthesize existing knowledge of the physical and dynamical processes of the climate system and allow for many of the complex interactions between the various climate components. Nevertheless, impact studies generally need more detailed information about the present-day and possible future climate conditions for the site or region of interest than can be derived from the GCM simulations. The mismatch is caused by limited computer resources preventing global climate simulations with a gridmesh distance smaller than 100-200 km, combined with incomplete description of various atmospheric processes and interactions between the earth surface (land, ocean) and the atmosphere. As a result, the variability in GCM simulated climates at small spatial scales and short time scales is not adequately represented. Over the last decades several methods have been developed to supplement the GCM simulations with a range of climate change scenarios at the appropriate resolution for impact studies (Giorgi and Mearns, 1991; Wilks, 1992). Four approaches can be distinguished: 1) transformation of a base-line climate series conform GCM predictions of large-scale changes in the seasonal means; 2) stochastic generation of local climate time series by adjusting the parameters in a time series model according to predicted changes in long-term means and variances; 3) stochastic generation of local climate series conditional on large-scale atmospheric circulation patterns (= statistical downscaling);
392 4) deterministic simulation of regional climates using a high resolution limited area model nested in a GCM (= deterministic downscaling). The first approach has been followed to obtain daily precipitation scenarios in the KNMI Climate Scenario project. The KNMI method makes use of GCM predictions about large-scale changes in the seasonal means of the climate elements which are best reproduced in the GCM simulations, like surface air temperature and surface air pressure. Daily precipitation is transformed indirectly using relations in the present-day climate between precipitation, surface air temperature, surface air pressure and other important elements for impact studies (air humidity and solar radiation). The relations are modelled by weighted non-linear regression techniques (Buishand and Klein Tank, in press; Klein Tank and Buishand, 1993; 1995). Direct GCM predictions about changes in the seasonal means of precipitation are not used. One reason for this is that precipitation is poorly represented in the current GCMs. Another reason is that the application of GCM predictions about large-scale precipitation changes to higher resolution data (e.g. daily station data) may not be appropriate as a result of the high spatial variability of precipitation. Transformation of a base-line climate in the form of observed series is preferred to stochastic weather generators, because it is simple and it automatically provides a realistic variability on daily as well as longer time scales. This paper focusses on the derivation of the KNMI precipitation scenarios. The results of background activities like the analysis of GCM climate simulations are not discussed here.
2. TRANSFORMATION OF OBSERVED PRECIPITATION SEQUENCES The increase of the maximum concentration of water vapour with temperature (Clausius Clapeyron relation) causes fronts and other weather systems to produce more rain at higher temperatures. This effect explains part of the relation between mean precipitation amount R and surface air temperature T for wet days (threshold 0.1 mm) at De Bilt in Figure 1. A notable exception is the behaviour in the intermediate temperature regime, where the mean precipitation amount decreases with temperature. This is caused by a decreasing activity of large-scale precipitating systems at temperatures > 15~ These systems become rare at T > 18~ but then convective showers become increasingly active and the mean amounts rise again. KNMI Scenario 1 is based on Figure 1. It is obtained by transforming the precipitation amounts on wet days. The procedure is as follows: 1) apply a GCM-predicted change in seasonal mean temperature to all observed daily temperatures T; 2) determine for each wet day the resulting relative change in the mean precipitation amount R from Figure 1; 3) multiply the observed daily amounts by the calculated relative changes (multiplying factors).
393
De Bilt (1906-1981 ) '
I
'
I
'
I
'
I
/'
r ~" E rr
6 i
v
c0
E
4-
t--
._o
§
._o. ft.
I
2-
. / ~ 1
0 -2(
m
I -10
10 I
,
I 0
,
I 10
J
I 20
,
Temperature T (~
Figure 1. Mean precipitation amounts R (dots) at surface air temperature T class intervals of 2~ for wet days at De Bilt (1906-1981). The smooth curve represents the fitted regression relation; the error bars the standard deviations of the means. After Buishand and Klein Tank (in press).
In Scenario 1 only the effect of a prescribed atmospheric warming is taken into account. The relative changes in Steps 2 and 3 assume implicitely that the atmospheric circulation changes according to its present-day dependence on T. More flexible scenarios can be obtained by prescribing also a change in the atmospheric circulation. In Scenario 2 the daily mean surface air pressure P is included in the analysis for this purpose. Figure 2 presents the relation between R, T and P for wet days at De Bilt. The procedure for transforming precipitation amounts on wet days into a consistent scenario (Scenario 2) for the case of both a prescribed warming and a prescribed change in surface air pressure (atmospheric circulation) is similar to that for Scenario 1, but now Figure 2 is used instead of Figure 1. In Scenario 1, it is assumed that the number and the sequence of wet and dry days in the future climate time series remains the same as in the observed record. Since the occurrence of a wet day is linked to the atmospheric circulation, Scenario 2 with a systematic change in P must account for a change in the sequence of wet and dry days. This was done as follows: 1) assign probabilities of rain to each day using Figure 3; 2) compute for each season the change in the number of wet days from these probabilities in the present-day and future climate; 3a) if the number of wet days in a season decreases by n: assign n wet days in the series as dry on the basis of their probability of rain, e.g. using a Monte-Carlo method. 3b) if the number of wet days in a season increases by n: assign n dry days in the series as wet on the basis of their probability of rain and determine their amounts using Figure 2.
394
De Bilt (1906-1981) lo5o
.
,
.
,
9
9
,
De Bilt (1961-1990) .
,
:
.
9
~
......... i .......... ! ..........
.t:: ,.i.
r "~
i .......
- ......... ! ..................
J..........
! ..........
. 7 5 . . . . . . . . . . -.. ......... ! .......... i ......... : . . . . . . . . . i ......... i ......... i .......... i .......... I :: i i : : : low solar radiation .............................................................
ii,,i
}
liiiii iiii:iii ii =:iiiiii:.. iiiii:ii i,, i iiii ,, i i....,'.......... ,iiii: .
~
~..........i~.........~i..........!.........:..........:.........i .........i ........i..........t
,
*2-4
........... .!.......... !....... ~::~.:.:J;i :a:)~i::~.!...'i .............
.0-2 960
, -20
I -10
,
I 0
,
I
,
10
I 20
,
, 30
960
Temperature T (~
990
1020
I 1050
Surface air pressure P (hPa)
Figure 2. Mean precipitation amounts R (dots) at surface air temperature T class intervals of 2~ and surface air pressure P class intervals of 6 hPa for wet days at De Bilt (1906-1981). The isolines represent the theoretical mean amounts from a fitted regression model. After Buishand and Klein Tank (in press).
Figure 3. Fitted logistic regression relations for the probability of rain at De Bilt (19611990).
3. E X A M P L E F O R DE BILT The daily precipitation amounts in the 1961-1990 record of De Bilt were transformed by the above methods. The prescribed changes in T and P were taken from the Canadian Climate Centre G C M predictions of large-scale changes in the seasonal means (Table 1; 2xCO2 - l xCO2 experiment). Figures 4 and 5 illustrate the effects of the transformation on the July 1962 precipitation data for Scenarios 1 and 2, respectively. The multiplying factors (Scenarios 1 and 2) and the probabilities of rain (Scenario 2) are also shown. Note that in Scenario 2 two wet days (6 and 21 July 1962) were assigned dry. Table 1 Predictions of the large-scale changes in the seasonal mean temperature and surface air pressure over Western Europe for the 2xCO2 Canadian Climate Centre GCM experiment.
AT(~ AP (hPa)
Winter
Spring
+3.0 -3.4
+2.3 - 1.1
Summer
Autumn
+3.7 +0.3
+3.4 -0.1
395 D e Bilt
AT=+3.7~ 40
f2.00 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
~
~0 .................................................................. t~.~0 ~ 9 Multiplying factor
8
~
~
a
95. 1 o ................................................................... .... ,,m-.._
.5o
:-Ji:::f
........................
0
o.9,
5
3
13
9
7
11
17 15
21 19
25
23
(total 69 mm) M Scenario 1 (total 80 mm)
0.00
1
mObserved
29
27
31
J u l y 1962
Figure 4. Precipitation amounts in the observed and transformed July 1962 month at De Bilt for Scenario 1. The solid squares represent the multiplying factors. D e Bilt
ZXT=+3.7~
AP=+0.3 hPa
c'-
40
-2.00
E E -'30 t-"
---
~__ ~ o
:3
v
1.50 ~
E t~ 20
~t
c-, lO
ii] "
F O
1.00
t-" O
0 0
"0 O
.n.,
O
11t 50
-~-
~L_
13..
0
~0.00 1
5
3
13
9
7
11
17 15
21 19
25 23
29 27
o Probability of rain 9 Multiplying factor mObserved (total 69 mm) Scenario 2
0
(total 74 mm) m.
31
:3
J u l y 1962
Figure 5. Precipitation amounts in the observed and transformed July 1962 month at De Bilt for Scenario 2. The solid squares represent the multiplying factors and the open squares the probabilities of rain.
396 The largest precipitation changes in the two scenarios occur in winter (+20% and +44% for Scenarios 1 and 2, respectively) and the smallest in autumn (+5% for Scenario 1) or summer (+8% for Scenario 2). The annual mean amount increases by 10% in Scenario 1 and 20% in Scenario 2. These values differ considerably from the precipitation changes as predicted directly by the GCM itself (+28% in winter, -26% in summer and no change on average over the year).
4. APPLICABILITY OF THE SCENARIOS
An attractive feature of the KNMI approach is that the scenarios and their updates can easily be implemented by impact groups. The transformed daily series have a realistic variability on daily as well as longer time scales. Extreme case scenarios for sensitivity studies can be constructed from past (extreme) episodes. Scenarios in the form of monthly, seasonal or annual time series can be obtained directly from the transformed daily series. The scenarios facilitate integration of NRP climate change impact studies. After consulting the potential users a follow-up project is planned to construct a wider range of scenarios in which changes in air humidity and solar radiation are included.
5. REFERENCES
Buishand, T.A. and A.M.G. Klein Tank, (in press). Regression model for generating time series of daily precipitation amounts for climate change impact studies. Stochastic Hydrology and Hydraulics. Giorgi, F. and L.O. Mearns, 1991. Approaches to the simulation of regional climate change: A review. Reviews of Geophysics, 29, 191-216. Klein Tank, A.M.G. and T.A. Buishand, 1993. Modelling daily precipitation as a function of temperature for climate change impact studies. KNMI Scientific Report WR 93-02, De Bilt. Klein Tank, A.M.G. and T.A. Buishand, 1995. Transformation of precipitation time series for climate change impact studies. KNMI Scientific Report WR 95-01, De Bilt. Wilks, D.S., 1992. Adapting stochastic weather generation algorithms for climate change studies. Clim. Change, 22, 67-84.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
397
Climate Change Scenarios for Great Britain and Europe M.Hulme, E.M.Barrow, O.Brown, D.Conway, T.Jiang, P.D.Jones and C.Turney
Climatic Research Unit, School of Environmental Sciences, University of East Anglia, Norwich NR4 7TJ, United Kingdom
1. I N T R O D U C T I O N The objective of the work described here was to develop future climate change scenarios for Great Britain and for Europe related to global emissions of greenhouse gases. These scenarios were to be used by a variety of ecosystem and hydrological modellers in a research project titled 'Landscape Dynamics and Climate Change', a project sponsored by the UK Natural Environment Research Council (NERC) under their TIGER (Terrestrial Initiative in Global Environmental Research) programme. The work on the scenarios is now complete and the project as a whole will report its findings during 1995. This paper describes three stages to the scenario construction: the construction of gridded baseline climatologies for 196190 using station observations; the construction of the patterns of future climate change using results from General Circulation Model (GCM) experiments; and linking the previous two steps to generate estimates of future climate for specified decades in the future. At all stages work progressed at two spatial resolutions - a 10km resolution for Great Britain and a 0.5 ~ latitude/longitude resolution for Europe.
2. T H E 1961-90 C L I M A T O L O G I E S Since GCMs are generally not regarded as accurate enough to provide useful descriptions of current climate at local or regional scales, one of the essential components of any future climate scenario is an adequate description of the current climatology of the region of interest based on observed data. Mean monthly climatologies were therefore constructed for the two TIGER domains (Great Britain and Europe) for the following surface climate variables: mean, minimum and maximum temperature, precipitation and raindays, sunshine hours, vapour pressure, wind speed and ground frost days. These climatologies used station data for the period 1961-90 collected from National Meteorological Agencies (NMAs) across the region. The distributions of European stations for which 1961-90 data were obtained are shown in Figure 1 for temperature. The interpolation of the station data to the respective grids used partial thin-plate splines as developed by Mike Hutchinson from the Australian National University. Since elevation was one of the predictor variables, three climate surfaces were produced for each variable reflecting the 'minimum', mean and 'maximum' elevation within each 10km or 0.5 ~ cell. Month-by-month anomalies on these grids for the period 1961 to 1990 were also calculated for the variables mean temperature and precipitation.
398
80 70 .__
....
Z
o v
9
60
-
-
".E'
-
el) na
~ e..,-
~,
50
-
-
-
"~ ""
..... . :
~ .. "
9
.'"." : . . .
::
-
I
I
-20
(
" *
~
*.
0
"
"
" ~
""
"
"
{;
9 9
,.
"
9
"
-
"
7 -
~'-'.-:,'
.=
"':,\ "
'
"
" 9
.~' 42
"
-
~'
9
~
.
. , , ~ . . . . . . . . :
..... t'-"---'~-__.~-.-- 9
"
,
-10
"
-
,~ "~-;[.'~.~',~,_..~.,~-,..,Ci.~:..~.~..~.,.9 '.,.% _,o.)/..
"
9 9
9
'
";" "." .'" 9 I '..
..
~%..:-,.,.
j ?
""
'~
-
1
" ."
I,**~'k,
.---.r " ' - ~ . " ' " " ; . . . . .
:2-..~,.t, ".~X" . ~ - ~
,,,:,.:::.:-:..~*.~;';,~,),.-',<"-,':( ~. 9
-,.30
9 ~ "~'.~.t : ..~......~k'~%,
.~~'.,~:~."...v
40 3O
"f '
e/f, ),. j~" . ~, "-,, :"
. ~ a ~ , ~; . - . - ~ , _ 9r
o _J
~
"
"
9
,
,~
10
"~i
9
~
~
20
*
"*
30
\
"
9
9 t ~q, \
"\
40
'k
.
9
. 9 '~
9 "
"
"
~.~
9
" "~" '. ,~
.
I 9
I'~-.;~,
50
~'----T-"
60
Longitude
Figure 1" Distribution of sites across Europe for which 1961-90 monthly normals were obtained for temperature.
The accuracy of the various interpolated surfaces was assessed using validation sets of independent station data (i.e., station data not used in the interpolation). Estimated mean absolute errors (MAEs) for the European surfaces ranged from under 5% for vapour pressure to about 15% for wind speed to up to 20% for precipitation in some regions. The accuracy of the interpolated surface for maximum temperature was greater (MAE ---0.4~ than for minimum temperature (MAE ---1.0~ Some other validation statistics for the European climatology are shown in Table 1.
Table 1 Validation statistics for the European climate surfaces for January and July. None of the validation sites (n = 100) were used in generating the climate surfaces. Frostdays for July are omitted since so few occur. January July MAE Obs. mean Mean bias Obs. mean Mean bias MAE Max. temp. (~ 5.4 -0.1 0.6~ 24.9 0.0 .6~ Min. temp. (~ -1.9 0.1 0.8~ 15.4 0.0 0.6~ Precipitation (mm). 66.1 -1.6 21.7% 49.9 -2.4 13.0% Sunshine (hours) 72.9 1.4 6.2% 268.2 1.8 3.8% Vapour press. (mb) 6.1 0.1 15.2 0.0 4.1% 3.4% Wind speed (ms 1) 3.4 0.0 15.7% 3.0 0.0 17.4% Frostdays (days) 16.1 0.1 10.6% Raindays (days) 16.9 -0.1 6.0% 7.0% 12.0 -1.3
399 3. G E N E R A L
CIRCULATION
MODEL
EXPERIMENTS
General Circulation Models (GCMs) provide the most comprehensive method of investigating the response of the global climate system to various types of internal or external forcing. A considerable number of GCMs have now been used to simulate the effect of increasing atmospheric concentrations of greenhouse gases (GHGs) on global climate. GCM climate change experiments have fallen into one of two types: equilibrium and transient experiments. In equilibrium experiments, oceans are represented by a simple specified mixed layer with no deep vertical mixing. After integrating the model under a control concentration of GHGs, concentrations are usually instantaneously doubled and the model integrated until a new quasi-equilibrium climate state is reached. The difference between the control and perturbed climates then represents the equilibrium pattern of GHG-induced climate change. In transient experiments the situation is more complicated owing to the fully three-dimensional representation of the oceans. These ocean-GCMs are linked to the atmospheric-GCMs and the two models integrated in parallel allowing the ocean circulation respond to the changed fluxes between ocean and atmosphere. The perturbed experiment usually introduces a progressive increase in GHG concentrations, often a 1% per annum increase. In our work we have used results from three GCM experiments: two equilibrium and one transient. The equilibrium experiments were those performed in 1989 at the Hadley Centre (UKHI) and the Canadian Climate Centre (CCC) using relatively high-resolution atmospheric GCMs. The transient experiment (UKTR) was performed at the Hadley Centre in 1991/92 using their coupled ocean-atmosphere GCM. The mean monthly fields from these three GCM experiments were extracted for the same range of climate variables as existed in the baseline climatology and the changes between the control and perturbed integrations calculated. The change fields were interpolated down to the same resolutions as the baseline climatology using a simple Gaussian space filter.
4. C R E A T I N G
CLIMATE
CHANGE
SCENARIOS
For a climate change scenario to be used most effectively in an impacts analysis, it needs to be identified with a particular set of assumptions about the future GHG emissions path which might cause it. It is also necessary to assign a future year or decade by which the scenario might be realised. For a variety of reasons, using direct results from GCM experiments does not allow either of these conditions to be met. We therefore used a further type of climate model, a simple one-dimensional upwelling-diffusion model, to assist in this task. The model used is called MAGICC (Model for the Assessment of Greenhouse gas-Induced Climate Change) and was developed by Tom Wigley, Sarah Raper and Mike Salmon of the Climatic Research Unit. MAGICC determines the global-mean temperature and sea-level change implications of specified emissions scenarios for the various trace gases that may affect the Earth's climate. There are a number of model parameters which can be adjusted by the user, the most important one being the climate sensitivity which can take on a value between 1.5 ~ and 4.5~ for a doubling of CO2. Using the set of six GHG emissions scenarios published by the Intergovernmental Panel on Climate Change (IPCC) in 1992, estimates of future global warming were calculated using MAGICC (note here that we are ignoring the cooling effect of sulphate aerosols on global temperature since the forcing due to such aerosols was not included in the GCM experiments from which we derive our patterns of change). These estimates for three of the IPCC emissions
400 scenarios (IS92a, IS92c and IS92e) are shown in Figure 2 as projections between 1990 and 2100 with respect to 1990. The effect of three different climate sensitivities on these estimates is also shown. The IS92a scenario, with a 'central' climate sensitivity of 2.5~ yields a global warming of just under 2.5~ by 2100. Table 2 presents these calculations in a different way as the estimated date by which I~ of global warming, with respect to 1990, will have been reached. For the IS92a scenario this date may vary between 2023 and 2055 depending on the climate sensitivity. Using the results from MAGICC, together with the standardised GCM change fields described above, and then projecting these changes onto the 1961-90 baseline climatology, a variety of future climates can be generated. These scenarios relate to a range of future dates and are based on different assumptions about both the GHG emissions path the world will follow and the sensitivity of the climate system to GHG forcing.
Table 2 Estimated dates by which I~ of global warming would be reached, with respect to 1990, under different GHG emissions scenarios and assuming different climate sensitivities (results from MAGICC). Effect of sulphate aerosols are i~nored. 1992 IPCC Greenhouse Gas Emissions Scenarios IS92f IS92e IS92d IS92c IS92b IS92a Climate sensitivity 2047 2048 After 2100 2084 2057 2055 1.5~ 2032 2035 2047 2049 2037 2036 2.5~ 2025 2025 2034 2033 2027 2028 3.5~ 2021 2021 2027 2024 2028 2023 4.5~
c
5
...................................................................................................................................................................................................... ~':
r-
,
od
= 4 l- ....................................................................................................................................................................... , e~
.
3 ,,W"
E c
2
E -
1
T ~ , .
~+ 0
-a2.5 ---=---c2.5 e2.5 ...... al.5
(no (no (no (no
s) s) s) s)
......
cl.5
( n o s)
...... .... ....
el.5 a4.5 c4.5 e4.4
(no (no (no (no
s) s) s) s)
. . . . . . . . .
--
0
(.9 0 0 0") O~
0 0 0
0 v0
0 ~ 0
0 r 0
0 ~ 0
0 tO 0
0 CO 0
0 I",. 0
0 IX) 0
0 O~ 0
0 0 v-
Figure 2" Global-mean warming projections from 1990 to 2100 for the IPCC emissions scenarios IS92a, IS92c and IS92e assuming three different climate sensitivities: 1.5 ~ 2.5 ~ and 4.5~ Results are from the MAGICC model. Sulphate aerosol effect ignored.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
401
T h e R o s s b y s o l i t o n - a r o b u s t n o n - l i n e a r s t r u c t u r e in the e q u a t o r i a l ocean
T.R.F. Feitsma and H.A. Dijkstra Institute for Marine and Atmospheric research Utrecht (IMAU), Utrecht University, Princetonplein 5, 3584 CC Utrecht, The Netherlands
Abstract
A non-linear reduced gravity model of the equatorial Pacific ocean is used to study the characteristics of one special solution: the first baroclinic Rossby soliton. We have found it to travel westward and lose 25% of its energy in the reflection at the west coast. It returns as a non-linear Kelvin wave which steepens as it travels eastward. Reflection of this structure at the east coast generates a new Rossby soliton. Again about 25% of the energy is lost. Interaction of the non-linear Kelvin wave with a Rossby soliton leaves both structures unchanged. Finally, a temporal westerly wind patch in the centre of the basin generates a same kind of non-linear Kelvin wave, which returns after reflection as a Rossby soliton.
1. Introduction
This research is part of the NOP project 853110, entitled: "Non-linear dynamics of the coupled equatorial ocean-atmosphere system". In particular we are looking for the physical mechanisms that determine the behaviour of the tropical coupled system on interannual time scales of 2-9 years (El Nifio/Southern Oscillation). Focus is on persistent oscillatory structures in the coupled equatorial Pacific oceanatmosphere system, developing through non-linear interaction of unstable coupled modes. As a first approach we study these structures in a non-linear model of the ocean only. The coupling to the atmosphere is to be added later. Then we will study how weak coupling affects the solutions as found for the ocean model. Boyd (1980) has already made a rigorous study of the non-linear development of equatorial waves in an uncoupled homogeneous ocean model. This study shows that for dispersive Rossby waves eventually a balance between dispersion and non-linearity (inertia) develops, resulting in equilibrium structures like solitons. The importance of these Rossby solitons for ENSO is also stressed by Kindle (1983). He showed that temporal relaxation of trade winds can generate an internal Kelvin wave front, which reflects at the eastern oceanic boundary as one or more solitons. In this study we want to take another look at these Rossby solitons by means of a direct numerical time integration. We want to know what happens when the Rossby soliton reflects at the western boundary. What kind of structure results? What happens to the energy of the system? How is it redistributed, how much energy is absorbed by the boundary? Can a Rossby soliton be generated by an arbitrary initial disturbance?
402
2. Model
We use a reduced gravity model on an equatorial beta plane to describe the ocean dynamics. In this 1.5-layer model, the upper layer is bounded from above by a rigid lid. The lower layer is infinitely deep and inert. The interface between the two layers lies at a depth of about 100m and models the tropical thermocline. The set of partial differential equations describing the ocean dynamics consists of two momentum equations describing the time evolution of the zonal (u) and meridional (v)flow velocities, and the continuity equation describing the time evolution of the thermocline depth (h). At the eastern and western boundary we apply no-slip boundary conditions. The northern and southern boundary of the domain of calculation are open. To avoid any reflection from these artificial boundaries we apply a sponge filter, which damps out any signal approaching them.
3. Results
We initialise the dynamical fields u, v and h with the first baroclinic Rossby soliton given by Boyd (1980), in a basin extending from 140~ to 280~ and from 20~ to 20~ (fig. 1). Following the soliton as it travels westward, we find that the isolated structure moves across the basin without changing its shape. We can therefore conclude that in our numerical code the right balance between non-linear and dissipative effects, necessary for the existence of soliton solutions, is maintained.
10-
Z
-10 -
i
140
Fig. 1
I
'
I
180 220 longitude (~
/
I
260
The thermocline depth perturbation h at day 0: the Rossby soliton.
403
10Z o v
"o :3
00.5
. i
i
0.5
-10 -
' 140
Fig. 2
I J I 180 220 longitude (~
'
I 260
The thermocline depth perturbation h at day 120: the non-linear Kelvin wave.
10 - ~ Z o v
"o c}
0-
. i
-10 -,0.5 _
I
140
Fig. 3
I ' I 180 220 longitude (~
I
1 260
The thermocline depth perturbation h at day 200.
After some time the Rossby soliton reflects at the west coast. The reflected signal is clearly a non-linear Kelvin wave (fig. 2) with its maximum at the equator. Its meridional velocity is two orders of magnitude smaller than the zonal velocity, so
404
practically zero. A second product of the reflection are high frequency Rossby waves which are visible along the west coast. However, the wavelengths of these waves are too small to be properly resolved on the numerical grid that is used. Calculating the sum of the total kinetic and potential energy, we find that about 25% of the total energy is lost in the reflection. As the non-linear Kelvin wave travels eastward, it steepens because of non-linear effects. After reflection at the east coast we see that a new Rossby soliton appears (fig. 3). Furthermore we see coastal Kelvin waves travelling poleward along the east coast, leaving the basin. Again, about 25% of the energy is found to be lost in the reflection. In a second simulation we let the Kelvin wave which appeared in our first simulation interact with a second Rossby soliton, initialised in the east part of the basin. We find that the two structures add up almost as if the superposition principle were valid, although we are dealing with non-linear waves. After interaction the two structures separate again and regain their initial shape. This behaviour is typical for solitary waves. But as the non-linear Kelvin wave is not a soliton, for it changes shape as it propagates, this behaviour was not expected. In a third simulation we let a patch of westerly winds blow between 200~ and 220~ for three months. This generates an eastward travelling Kelvin wave, which reflects at the east coast and returns as a Rossby soliton. This result is the same as was found by Kindle (1983).
4. Conclusions Rossby solitons turn out to be robust solutions in a non-linear 1.5-layer equatorial ocean model. Not only are they maintained when explicitly initiated, but they can also be caused by reflection of any Kelvin wave, irrespective of how it was generated. In particular, a temporal weakening of easterly trade winds, modelled here by a temporal patch of westerly winds, generates such a Kelvin signal. As weakening of trade winds tends to coincide with the initiation of El NiSo events, we might expect to detect Rossby solitons a few months after the onset of an El NiSo, as the Kelvin wave has reflected at the coast of Peru. They have however never been observed so far. It might be that the signal is so weak in the real ocean that it disappears in other wave signals. Another explanation may be that nobody has ever attempted to extract a Rossby soliton signal from observational data.
References Boyd, J. P., 1980: Equatorial solitary waves. Part I: Rossby solitons. J. Phys. Oceangr., 10, pp. 1699-1717. Kindle, J.C., 1983: On the generation of Rossby sofitons during El Ni#o. In: J. C. J.Nihoul (editor), Hydrodynamics of the equatorial ocean, Elsevier Oceanography Series, 36, pp. 335-351.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
405
A nonlinear stability analysis of the coupled equatorial ocean-atmosphere system P.C.F. van der Vaart Institute for Marine and Atmospheric research Utrecht (IMAU), Utrecht University, Princetonplein 5, 3584 CC Utrecht, The Netherlands
Abstract A simple model of the coupled ocean atmosphere is considered to investigate the r61e played by nonlinearities in the evolution of large scale instabilities related to ENSO. In this paper we focus on the nonlinear evolution of unstable waves. The initial disturbances develop to finite amplitude on a long space- and time scale and their evolution is governed by amplitude equations of so called complex Ginzburg-Landau type. The coherent structures which emerge show how important the associated space- and time scales are in equatorial dynamics. 1. Introduction This research is part of the NOP project 853110, entitled: "Non-linear dynamics of the coupled equatorial ocean-atmosphere system". In particular we are looking for the physical mechanisms that determine the behaviour of the tropical coupled system on interannual time scales of 2-9 years (El Nifio/Southern Oscillation). Focus is on persistent oscillatory structures in the coupled equatorial Pacific ocean-atmosphere system, developing through non-linear interaction of unstable coupled modes. As a first approach we study these structures in a non-linear model of the equatorial ocean atmosphere system. Boyd (1980-1987) has already made a rigorous study of the non-linear development of equatorial waves in an uncoupled homogeneous ocean model. On the other hand Hirst (1986) thoroughly investigated the linear stability of the coupled model. The first study shows the presence due to nonlinearities of coherent structures, such as solitons, the second that coupling introduces instabilities which grow in time. Both phenomena have been related to (the onset of ) ENSO. In this study we want to explore the effect of nonlinearities on the initially unstable waves in a weakly nonlinear analysis. Their resulting finite amplitude behaviour is then described by a Ginzburg-Landau equation. 2. Model To study the weakly nonlinear properties of the equatorial pacific in a qualitative way we use an intermediate coupled ocean-atmosphere model on an equatorial 13-plane. This model has a reduced gravity 1.5 -layer nonlinear ocean, describing the temporal and spatial evolution of the horizontal velocities and thermocline depth. The atmosphere is a linear Gill model with horizontal momentum equations and an equation for the evolution of geopotential height. Atmospheric horizontal motions drive oceanic flow, the coupling is completed by a thermodynamic equation describing the evolution of the sea surface temperature (SST), which is optionally nonlinear. The main features of this last equation are advection of SST through zonal oceanic motion and changes in SST trough local thermal processes such as upwelling. The model has a simple climatological background state. Both the oceanic and the atmospheric components have zonally periodic boundary conditions and are meridionally unbounded.
406
3.Linear stability Due to the positive feedbackloop induced by coupling in our model ultra long oceanic waves of low frequency are the first to destabilize. We start our linear stability analysis with a simplified thermodynamics equation, considering only local thermal processes associated with upwelling (fast SST-limit). As the coupling is increased the oceanic Kelvin wave becomes unstable, other waves are damped. In fig. 1 the neutral curve, which separates regions of stable (damped) solutions from exponentially growing ones, for this situation is presented. At criticality, at the nose of the neutral curve ( k=kC=.058, coupling m=mC=. 1065 ) the wave has a wavelength of 26.000 km, approximately twice the basin width of the equatorial pacific. This unstable coupled Kelvin wave is now dispersive and moving eastward with phase velocity about 66% reduced in comparison to no coupling. In figure 2 eigenfunctions of the Kelvin at critical conditions are shown. In the case of purely zonal advection of SST, oceanic Rossby waves of wavelengths of about 32.000 km are destabilized, other waves are damped. When the full linear SST equation is considered the dominantly unstable mode is related to the uncoupled mode of this SST equation. Model parameters such as atmospheric damping, and the ocean- atmosphere Kelvin wave speed ratio or a steady state atmosphere affect the solutions only in a quantitative way (cf. Hirst).
0.6-
unstable
0.5-
== m.
o
0.4-
0.3,
0.2"
stable
p=p " 0
k=k
9 0.1
0.2
0.3
0.4
0.5
wavenumber k
fig. 1
NeutralCurve of the Coupled Kelvin wave
0.6
407
~, ~;~,,:~<~,:~,
,~,~,~ ~ ~,~ ,,
~,~, ~ ~~,i~ ~
~~~
I~,~
H -~
0
0
i ~ii~ii}!i!i ::
;:i:t
ii::i
;~i~!~:~::i~!:i' ~~i-ii!~i!iN~ ~ :,~: ~,~ ~ fit ,~ ; :.:i~ ::::iiiiii::: ..:.i... . . . . . i~,ii'~i~,i'~:~i;i:,~:',, ,:.:~
ii~i!. ii~..~i.:j::'~::::::iili::i:~i:~~ii:l .: %~ ~". . . . . / . . i:.ii:i~ ::! ~) ii~:!i'~:,:,:!~i: ", .. -" :: i !i[
! i ~ ''_: N ~)~: : i l : ~ ~ - ~ ~ ~
- 500 0 I
'
-20000.0
'
I
'
I
0.0 longitude (km)
......
I ......
'
....
I
'
20000.0
I
-0.250 0.000 0.250 In units of scaled mixed layer depth Fig. 2 The eigenfunctions of the Kelvin wave at critical conditions. The coloured background represents the position of pressure highs (light) and lows (dark). The contours indicate mixed layer depth, solid lines are associated with downwelling and positive SST perturbation. Note the positioning of pressure lows just east of positive SST perturbations. 4.
Non
linear
stability
In the previous section results of the linear stability analysis have been sketched, indicating that for different types of SST dynamics a spectrum of wavelike perturbations with exponentially growing amplitudes will develop. This description is valid in the initial growth stage, with very small amplitudes. To investigate the behaviour of these modes on long time- and space scales, the nonlinear terms in the model have to be taken into account. This is done using standard nonlinear analysis techniques. At criticality the method of multiple scales is used to derive a modulation equation which incorporates the nonlinear self-interaction of the neutral modes. The amplitude A of the initially unstable waves is then governed by the complex Ginzburg-Landau equation
0.4 02A O---T= ~ - ~ + fl2a - fl3alalZ
(4.1)
Here X and T have been defined on long time and space scales, the complex coefficients ~j depend on all model parameters. To obtain finite amplitude solutions of (4.1) the real parts of [32 and [33 have to be positive. In fig.3 the coefficients have been listed for each of the unstable waves. Though the coefficients of the linear terms are accurate, the nonlinear coefficients still need to be checked thoroughly. This and, subsequently, the verification of solutions obtained from equation (4.1) can be done by using a numerical model. At this moment investigation of the amplitude equation for the Kelvin wave is in
408
progress showing periodic solutions, while finite dimensional models of equation reveal that a-periodic and chaotic solutions are also possible. unstable mode crit. wavenumber Kelvin Rossby SST
.058 .048 .042
this
13] 4.833+2.065i 0.998-1.021i 9.685-2.5278i
0.245+0.0574i 0.00499-0.0331i 0.00133+0.00212i -9.929+15.075i 0.0516-0.00474i 0.029-0.0046i
Fig. 3 The coefficients 13jfor the unstable waves. Note that the Rossby wave has Re(133) < 0, this implies an unbounded amplitude in finite time.
5. Conclusions
We have performed a linear stability analysis of an intermediate coupled equatorial ocean atmosphere model and re-established the most important results found by Hirst (1986). These include the unstable coupled Kelvin, Rossby and SST-modes. Consequently the weakly non-linear behaviour of the model is investigated resulting in the derivation of an amplitude equation of complex Ginzburg-Landau type. This equation governs the evolution of the amplitude of each of the most unstable waves on long time and space scales and inhibits rich dynamical behaviour. The first results show that the amplitude of a Kelvin wave that becomes unstable through local thermal processes grows to a finite value; the solution is then periodic in time and space. Also there are regions in wavenumberspace in which aperiodic and chaotic behaviour occurs. Similar results can be expected for the modulations of the other unstable waves. However, the focus of our study will be to extend this model to include more realistic features of the coupled equatorial pacific, such as zonal boundaries and a spatially varying climatology. Furthermore the oceanic part will be improved by adding more complete thermodynamics and mixed layers. The already established results and developed techniques will be useful in understanding the more complicated models.
References
Boyd, J. P., 1983: Equatorial solitary waves. Part 2 Envelope Solitons. Phys. Oceangr., 13, pp. 428-449.
J~
Hirst, A. C., 1986: Unstable and damped equatorial modes in simple coupled ocean - atmosphere models. J. Atmos. Sci., 443, 606-630. Doelman, A. 1990: On the nonlinear Evolution of Patterns, Modulation Equations and their Solutions (Thesis)
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
411
Surface fluxes of water vapour, heat, momentum and CO2 o v e r a s a v a n n a h i n N i g e r : A c o n t r i b u t i o n to H A P E X SAHEL. A. Verhoef, H.A.R. de Bruin and R. Klikke
Wageningen Agricultural University, Department of Meteorology, Duivendaal 2, 6701 AP Wageningen, The Netherlands
ABSTRACT This study concerns a Soil-Vegetation-Atmosphere Transfer (SVAT) model, which consisted of the "best" parts of the SVAT's as developed by Choudhury and Monteith (1988) and Deardorff (1978). Model simulations are compared with data collected in the context of the HAPEX-SAHEL project (Gourtorbe et al., 1994) for day 273, which is about 9 days after the last rainfall in 1992.
1. I N T R O D U C T I O N For large scale models such as Global Circulation Models (GCM's) the lower boundary condition is often provided by a Soil-Vegetation-AtmosphereTransfer (SVAT) model. A wide range of SVAT's is in use nowadays, varying from models based on the simple big-leaf concept by Monteith (1965) to complicated multiple source models. Obviously, a SVAT intended to provide the lower boundary condition in GCM's needs to be able to describe a wide range of surface types, varying from completely vegetated to sparsely vegetated or completely bare surfaces. Especially sparse canopy surface types exhibit rather demanding features with respect to the exchange of momentum, water vapour, CO2 and heat between the surface and the atmosphere. In this paper attention will be focused a sparse canopy. We will compare SVAT model simulations with data collected in 1992 at a savannah site, in the framework of the HAPEX-SAHEL project. Two existing SVAT models are considered, notably the model by Choudhury and Monteith (1988) and by Deardorff (1978). In a separate study these models have been tested. Using the results of the latter, we constructed a combined model, consisting of the "best" parts of the original SVAT's. Some preliminary results will be presented.
412 2. E X P E R I M E N T A L S E T U P
The HAPEX-SAHEL experiment was executed in Niger during 1991-1992 with an intensive observation pe~iod in August-October 1992. The measurements of the Department of Meteorology of the Wageningen Agricultural University were carried out at the West-Central super site, about 50 (km) east of Niamey. The vegetation consisted of a ground layer of annual herbs and grasses with scattered multi-stemmed shrubs (Guiera senegalensis). The sensible and latent heat fluxes were determined by Eddycorrelation (EC) methods. The momentum flux by EC and profile methods. CO2-fluxes were obtained by the EC technique. The EC setup consisted of a Solent 3-D sonic anemometer, a Lymann-alpha humidiometer, a fast response thermocouple and the LICOR C02/H20 gas analyzer. Cup anemometers and psychrometers provided the data for the P method. All instruments were mounted on three masts of 5 (EC equipment), 6 (psychrometers) and 10 (anemometers + thermocouples) metres height. The slow data were processed with Campbell 21-X dataloggers, the fast were processed by p.c's. All raw and calculated data were stored on tapes.
3. M O D E L D E S C R I P T I O N
In this section we will describe briefly the various SVAT models used in this study. For more detailed information the reader is referred to the cited papers as well as to Van den H u r k et al. (1994).
3.1. T H E M O D E L OF C H O U D H U R Y - M O N T E I T H
The model of Choudhury and Monteith (1988) (CM88), can be regarded as an extension of the model by Shuttleworth and Wallace (1985). Two components are distinguished: a canopy layer and the underlying bare soil. In CM88 soil heat flux density and the resistance regulating soil evaporation are explicitly parameterized. The radiant energy available to the canopy is parameterized as total net radiation minus soil heat flux density and minus the net radiation reaching the bare soil. Measured values of total net radiation are needed as forcing. The partition of available energy over sensible and latent heat flux is solved using the Penman-Monteith equation. In the soil two layers are present, of which the lowest is assumed to be saturated with water. Soil evaporation takes place at the intersection of the two layers (depth z). The transfer of water vapour through the upper soil layer is regulated by a resistance, proportional to depth z and water diffusivity. The soil heat flux density is parameterized using the temperatue difference at the surface and depth z and a resistance, which is a function of the soil heat conductivity and depth z. The depth z increases at a rate proportional to the soil evaporation rate, in order to simulate the retreat of the saturated zone as soil water is evaporated into the air. The Penman-Monteith equation is used to
413 solve simultaneously for the surface temperature, t e m p e r a t u r e at depth z, the soil heat flux and soil evaporation. The soil layers have no heat capacity. It is assumed t h a t m o m e n t u m and scalars are transferred by the same mechanism. An aerodynamic resistance between the canopy air and the substrate is introduced, which depends on the roughness length of the bare soil surface, and which is parameterized assuming an exponential decay of the eddy diffusivity within the canopy layer. The level between the canopy and the soil sources is regarded to be the effective level of the m o m e n t u m sink. This level varies with LAI and canopy height (Shaw and Perreira, 1982). A boundary layer resistance between the leaf surface and the canopy air is also introduced, reflecting the absence of bluff body forces for heat and water vapour exchange. No equivalent resistance is present in the pathway between the canopy air and the soil surface.
3.2. T H E M O D E L OF D E A R D O R F F
In modified form the scheme of Deardorff (1978) (D78) is in use in various GCM's (Dickinson et al., 1986; Noilhan and Planton, 1989). As in CM88, a dual-source surface is postulated in D78: a canopy layer and the underlying soil. The net radiation is computed, whereas in CM88 the measured value is taken. Separate energy balances are drawn up for the canopy and the substrate, and the incoming radiant energy is distributed over the two components using the fraction of vegetation cover. The canopy t e m p e r a t u r e Ts is evaluated iteratively, while the soil temperature Tg is predicted from the previous time step, where a rate of change of T~ is computed. A weighting procedure is used to obtain the temperature and humidity deficit within the canopy layer, and use is made of the explicit values of the t e m p e r a t u r e and humidity at the soil surface, the canopy surface and the reference level. This weighting is included in the iterative solution of the radiation balance. The boundary layer resistance for heat and water vapour exchange between the canopy air and the canopy surface is treated somewhat more rigorously t h a n in CM88. The aerodynamic exchange within the canopy is parameterized as a weighted average for the exchange of a completely bare soil and for a dense canopy using u. as the characteristic wind speed within the canopy. The analytic stability correction of Louis (1979) is used for the aerodynamic resistance above the canopy. Soil evaporation is treated as function of the relative humidity in the top soil layer. Both the soil heat and soil moisture transport are described with the force-restore method by B h u m r a l k a r (1975). A forcing of the sensible heat and moisture transport at the surface is modified by a restoring t e r m depending on the t e m p e r a t u r e and soil moisture deeper in the soil. Both the forcing and restoring term are functions of coefficients which depend on the soil type and moisture content (Noilhan and Planton, 1989)
414 3.3. T H E C O M B I N E D
MODEL
A comparison with our data collected at the savannah site in the framework of HAPEX-SAHEL revealed that CM88 underestimates the soil heat flux, while this important term of the surface energy balance is described satisfactorily by D78. On the other hand the description of the various resistances by CM88 appears to be better than t h a t used in D78. On the basis of these results we constructed a combined model consisting of the best parts of D78 and CM88 (Annex I). In this paper some results of the combined model are presented.
4. R E S U L T S AND D I S C U S S I O N In Figures 1 to 4 a comparison for net radiation, latent, sensible and soil heat flux density is made between model simulations and observations. Also, in figure 4 the simulated soil heat flux in CM88 is depicted. DAY OF YEAR 273
DAY OF YEAR 273
7OO 600
500 ....................................................
~'500
-.
.
.
i!!!!i!!!! iiiiii ! ii i l
. '3od 200
rr
7
:.:...../ .............. .... ii.iii..i
200
'~176 .................................
,
0 -100
0
2
4
6
8
[~
10
12 14 T I M E (hr)
estimated
16
18
20
22
9 observed I
Fig.(1); Estimated net radiation by the combined model compared to observed values.
24
-100
.
0
.
.
.
.
2
.
.
.
.
.
4
.
.
.
.
.
6
.
.
.
.
8
[--
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
10 12 14 T I M E (hr) estimated
.
.
.
.
.
.
16
.
.
.
18
.
.
.
.
.
20
22
9 observed ]
Fig.(2); Estimated latent heat flux by the combined model compared to observed values.
These results refer to 9 days after the last rainfall. It is seen t h a t the combined model describes net radiation and latent heat flux fairly well. The results for sensible heat are less satisfactory. The combined model, which is using (as D78) the force-restore method by B h u m r a l k a r (1975), describes soil heat flux considerably better t h a n CM88. It should be noted that in our case of a sparse s a v a n n a h vegetation, soil heat flux is an important term of the surface energy balance. At present, the flux of C02 is not described in the combined model. In the near future we will include this very important feature, making use of the results by Verhoef et al. (1994), Moncl-ieff et al. (1994) and Jacobs (1994).
415
D A Y O F Y E A R 273
DAY OF YEAR 273 300
3OO
iiiiiiiiiii
~.200 x
100 100 o
~;
Q3
o
~0-)-100
.~oo t. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
2
4
6
8
[
10
12 14 T I M E (hr)
estimated ,
16
18
20
22
24
..,
.. observed]|
Fig.(3); Estimated sensible heat flux by the combined model compared to observed values.
-200 [
.......
.
0
.
.
.
. ............................
.
2
.
.
.
.
.
.
4
L~ , c o m b i n e d
.
.
6
.
.
.
.
.
8
model ~
.
.
.
10
.
.
.
.
.
.
.
.
12 14 T I M E (hr)
cm88
.
.
.
.
.
16
.
.
18
..
.
.
.
.
.
20
observed
.
22
24
]
Fig.(4); Estimated soil heat flux by the combined model and CM88 compared to observed values.
ACKNOWLEDGEMENTS We thank Ad van den Berg for his assistance. We are grateful to everybody involved in the 1992 HAPEX-SAHEL field campaign. The first author received a grant of NWO/MFO (NWO 750.650.37). The EU (former CEC) supported a part of this research (EPOC-0024-C(CD) and EV5V-CT91-0033). Also, the Dutch National Research Programme on Global Air Pollution and Climate change financed part of this project.
REFERENCES B h u m r a l k a r , C.M. (1975): Numerical expelfiments on the computation of ground surface temperature in an atmospheric general circulation model; J. Appl. Meteorol. 14, 1246-1258. C h o u d h u r y , B.J. a n d M o n t e i t h , J.L. (1988): A four-layer model for the heat budget of homogeneous land surfaces; Q.J.R. Meteorol. Soc. 114, 373-398. D e a r d o r f f , J.W. (1978): Efficient prediction of ground surface temperature and moisture, with inclusion of a layer of vegetation; J. Geophys. Res. 83, 18891903. D i c k i n s o n , R.E., et al., (1986): Biosphere-Atmosphere Transfer Scheme (BATS) for the NCAR Community Climate Model; NCAR Technical Note NCAPdTN-275+STR, 69 pp. G o u r t o b e , J-P, et al., (1994): HAPEX-Sahel: a large scale study of land atmosphere interactions in the semi-arid tropics; Annales Geophysicae 12, 53-64. J a c o b s , C.M.J. (1994): Direct impact of atmosphe~ic C02 enrichment on regional transpiration; Wageningen Agric. Univ., The Netherlands; 179 pp.
416 Louis, J.-F. (1979): A parametric model of vertical eddy fluxes in the atmosphere; Boundary-Layer Meteorol. 17, 187-202. Moncrieff, J.B., et al. (1994); Submitted for publication to Special Issues (HAPEX-SAHEL), J. of Hydrol. M o n t e i t h , J.L. (1965): Evaporation and the environment; Symp. Soc. Exp. Biol. 19, 205-234. N o i l h a n , J. a n d P l a n t o n , S. (1989): A simple parameterization of land surface processes for meteorological models; Monthly Weather Rev. 117, 536-549. S h a w , R.H. a n d P e r r e i r a , A.R. (1982): Aerodynamic roughness of a plant canopy: a numerical experiment; Agric. Meteorol. 26, 51-65. S h u t t l e w o r t h , W.J. a n d Wallace, J.S. (1985): Evaporation from sparse crops an energy combination theory; Q.J.R. Meteorol. Soc. 111,839-855. V a n d e n H u r k , B.J.M. et al., (1994): An intercomparison of 3 vegetation/soil model for a sparse vineyard canopy; Submitted for publication to Q.J.R. Meteorol. Soc. V e r h o e f , A. et al., (1994); Submitted for publication to Agric. Meteorol.
ANNEX I
TI~
i '! -Reference height k E i ~r,,,,
To
C~an~
,
-. . . . . . . . . . . . . . . . . . .
Force-Restore
l
%
i - - ~r~c q c = r,ic
[q,(Canopyairstream
X-E~
Soil sm4'ace
Force-Restore
1
q,
Fig.(5); Schematic diagram of the Combined Model.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
417
Climate change and deforestation in West Africa: a space-time trend analysis of rainfall series from C6te d'Ivoire and Liberia Renaat S.A.R. van Rompaey P r o g r a m m e Office of the Dutch National Research Programme on Global Air Pollution and Climate Change NOP, c/o RIVM (pb 59), Postbus 1, NL-3720 BA Bilthoven, The Netherlands, tel: +31-30-743781; fax: +31-30-25 19 32; e-mail:
[email protected] Abstract
Rainfall series (1920-1990) from 11 stations from CSte d'Ivoire and 19 stations from Liberia have been analysed for pseudo-cyclic trends. Standardized deviations from the station's mean through time were plotted separately for the Ivorian stations where extensive deforestation occurred and the Liberian stations where deforestation was less extensive. The 5 year r u n n i n g m e a n t h r o u g h these deviations did not show a different pattern for the two sets of stations. It is thus concluded that the droughts of the seventies and eighties in West Africa are to be r e l a t e d to decadal v a r i a t i o n s in ocean surface t e m p e r a t u r e , and not to deforestation or more intensive land use on the African continent. 1. I N T R O D U C T I O N
The Sahelian droughts in the seventies and eighties have been reported by m a n y a u t h o r s (Faure & Gay 1981, Lhomme 1981, Nicholson et al. 1988, Adejuwon et al. 1990, Olivry et al. 1993). Analysing rainfall and river discharge time series dating back from 1920 and 1840 respectively, Faure & Gay concluded that decadal rainfall is subject to a pseudo-cycle with a period of approximately 30 years, with the last minima in the decades 1910-1920, 1940-1950 and 1973-1983. Hubert et al. (1989) stressed the p s e u d o - c y c l i c nature of this oscillation by calling them segments with varying length and amplitude. Many authors related the decrease in rainfall between 1950 and 1985 to large scale deforestation in the rain forest zone and to changed land use in the savanna zone in West Africa (Monteny & Casenave 1989, Savenije 1995, Bruijnzeel 1995). The hydrological cycle is interrupted because evapo-transpiration from deforested land is less than from full grown forest, and as a consequence runoff increases and more rainfall water is drained back directly to rivers and to the ocean in stead of re-evaporating and p e n e t r a t i n g further north into the continent. It r e m a i n s questionable w h e t h e r the m a g n i t u d e of evaporated moisture is significant compared to the moisture t r a n s p o r t e d inland by the m a r i t i m e air masses. Savenije estimated the horizontal velocity of these air masses at 17 km per day. I believe t h a t this velocity (average wind speed) may be 10 times faster, thus m u l t i p l y i n g by 10 the a m o u n t of m a r i t i m e moisture t r a n s p o r t e d onto the
418 continent and reducing by a factor 10 the impact of evaporated moisture. Wet and dry years in West Africa may be more related to oscillations in ocean surface t e m p e r a t u r e (SST; KSnnen pers. comm.). In this paper I will compare rainfall series from Liberia with those from CSte d'Ivoire. In the former country deforestation was much less extensive and for stations near the coast, the influence of deforestation inland is of course nihil. If the d r o u g h t s in the seventies have also been recorded in Liberia, t h e n deforestation cannot have been the cause. Correlation can then be sought with decadal variations in SST, possibly related with ENSO-events in the Atlantic ocean (Folland et al. 1986). Although the determinants of rainfall remain ill identified, there is a need to predict rainfall quantities and distribution in the coming decades. This prediction can be based on extrapolation of trends detected in measured rainfall time series. This regression type of approach supposes that no major change occurred in the climate system and in driving forces of rainfall in West Africa which in the light of global climate change remains questionable. Prediction can use cyclic or linear trends, or both. Does the pseudo-cyclic trend allow to predict the nineties to be rainier again, with a probable next drought in 2005-2010? 2. S T U D Y R E G I O N
The West-african rain forest block covers south Sierra Leone, all of Liberia, southern CSte d'Ivoire and south-west Ghana. Most of it is situated in lowland, 0 to 500 m above sea level. This biome has an equatorial climate at 4 to 8 ~ North and receives between 46 dm rain near the Liberian coast and 13 dm near the forests a v a n n a boundary. The dry season is centred on J a n u a r y and lasts one to three months (van Rompaey 1994). Eighty percent of the rain forest cover has been cleared since 1900, most of it since 1950 (van Rompaey 1993). This makes the region the most deforestated in the tropical world. The largest remaining forests are found in east Liberia and western CSte d'Ivoire. Ghana has smaller forest reserves but in better condition and under more professional management. 3. R A I N F A L L S E R I E S
Thirty rainfall stations were studied, some with series starting in 1920. The period 1950-1980 is best documented. Year to year variability (standard deviation divided by mean in %) ranges from 10 to 20 % without a clear gradient over the region. Yearly totals and their five year running mean have been graphed station by station. To study the homogeneity between stations annual rainfall figures have been normalised by subtracting the station's mean and dividing by the station's s t a n d a r d deviation. The normalised deviations were plotted in two graphs: one for all Liberian and one for all Ivorian stations (Fig. 1). For a given year deviations were averaged and a five year r u n n i n g mean was calculated through these averages to detect major trends.
419
plus = C o t e d'lvoire 3_deviation/std
+ t
+ +
T
+
+
+
+
+
+ +
_
*+
+
+
+
+
+
+
+
+ .
+ # +
+
+
+
~
+
+
+~
+
+
+
+
_
*§
+"
;
I §
*
+
'
.
§
+ +
*:
§
**
+
[
.
:
,*
.+ - u . . +, + : ; T i : : +
.
+[~]!
",
*
"+
*§
+ +
:
,
§+
+
§
*+
+ "9 i " "
+ "
.
+
-2-
*
+
9 +,;
+
*
"
§
. . . .
.
,
I**
+ +
+,~+,~
+
.
-1-
I
"
+;
+
§
.
+
. **
.
.
+ . L_I - U .+ +
. +
++ .
+ ,
,
+ ;
+ + +
-3
i
+
I
1920
i
I
1930
I
I
1940
i
I
1950
i
I
1960
1
I
1970
1980
1990
year x = station in Liberia 3_deviation/std
^ x x x x
x x
x
x
x
x
x
x
x
x
x
t
x
x
x
x
x
_
x
~
x
xX
x
~
~
^
x x X X
x
x
X x X ~
x
x
x x
x x
x
~
x
~t
x
x
Xx
x
x x
x
_
x x
LFU
-1-
x
•
x
,
x
x
,
*'m~':~*
:~,~;
.** x
x
~
~
x
x
x
x
~
x x
x.U:x~**
: !
~
x~X
~
~ Ix
:x
x
t•
~
x
~
~x
x
-2-
•
I
-5
t
1920
i
I
19S0
i
I
1940
i
I
L
l
1960
19S0
t
1970
'
1
1980
1990
year Figure 1: Standardized yearly deviations from the station's mean, above for 11 stations in CSte d'Ivoire(extensively deforestated), below for 19 stations in Liberia (less deforestated). The stepped curve gives the yearly mean for all stations. The bold curve gives the 5 yr running mean of the latter mean.
420 4. R E S U L T S W h e n considering the y e a r s s e p a r a t e l y , drought or rainfall a p p e a r e d to be regionally coherent. To a large extent the stations are consequently wet or dry in given years. Between s u b s e q u e n t years large differences can occur. There seems not to be a n y s h o r t - t e r m i n t e r a n n u a l autocorrelation. Wet and dry y e a r s m a y occur i m m e d i a t e l y after each other. At a decadal scale the r u n n i n g m e a n shows wet and dry decades. The decades '25-'35, '55-'65 h a d more wet years; the decades '40-'50, '75-'85 had more dry years a n d drier years. This decadal v a r i a t i o n was found both in the deforested CSte d'Ivoire as in the still forested Liberia. Many individual dry years coincide however with E N S O events as listed by Quinn & Neal (1987) for the Pacific. 5. C O N C L U S I O N S It is likely t h a t after the past decadal rainfall m i n i m u m 1975-1985 more w e t y e a r s will occur in the 1990-2000 decade. Given the d e f o r e s t a t i o n - i n d e p e n d e n t decadal variability, downward rainfall trends caused by deforestation are v e r y h a r d to detect. Studies based on 1950-1980 d a t a lead to false conclusions because both causes were cooperative in reducing rainfall. Rainfall m o n i t o r i n g in 19902000 m a y provide a definitive answer as in this decade both causes are opposing. The decadal v a r i a t i o n seems to be more related to ENSO-events, sea surface t e m p e r a t u r e and ocean currants. If these are changed by global climate change, t h e n the above prediction of future rainfall in West Africa m a y have to be revised. 6. R E F E R E N C E S Adejuwon J.O., Balogun E.E. & Adejuwon S.A. (1990). On the annual and seasonal patterns of rainfall fluctuations in Sub-Saharan West Africa. Int. J. of Climatology 10: 839-848. Bruijnzeel L.A. (1995 (in press). Predicting the hydrological impacts of tropical forest conversion: the need for integrated research. Proceedings ABRACOS Symposium on Amazonian Deforestation & Climate, Brasilia Sept. 1994. Faure H. & Gac J.-Y. (1981). Will the Sahelian drought end in 1985? Nature 291: 475-478. Folland C.K., Palmer T.N. & Parker D.E. (1986). Sahel rainfall and worldwide sea temperatures, 19011985. Nature 320: 602-607. Hubert P., Carbonnel J.P. et Chaouche A. (1989). Segmentations des s~ries hydrom~t~orologiques application h des s~ries de pr~cupitations et de d~bits de l'Afrique de l'Ouest. Journal of Hydrology 110: 349-367. Lhomme J.P. (1981). L'~volution de la pluviosit~ annuelle en CSte d'Ivoire au cours des soixante derni~res ann~es. La Mdt~orologie Vie S~rie 25: 135-140. Monteny B.A. & Casenave A. (1989). The forest contribution to the hydrological budget in Tropical West Africa. Annales Geophysicae 7(4): 427-436. Nicholson S.E., Kim J. & Hoopingarner J. (1988). Atlas of African rainfall and its interannual variability. Dept. of Meteorology, Florida State Univ. Tallahassee, Florida, USA. Olivry J.C., Bricquet J.P. & Mahe G. (1993). Vers un appauvrissement durable des ressources en eau de rAfrique humide? Dans: Gladwell J.S. (ed.). Hydrology of warm humid regions. IAHS publ. 216 (Proceedings of the Yokohama Symposium, July 1993), p. 67-78. Quinn W.H. and Neal V.T. (1987). E1 Nifio occurrences the past four and a half centuries. Journal of geophysical research 92 (C13): 14449-14461. Savenije H.H.G. (1995). New definitions for moisture recycling and the relationship with land-use changes in the Sahel. Journal of Hydrology (in press). van Rompaey R.S.A.R. (1993). Forest gradients in West Africa. A spatial gradient analysis. Doctoral thesis, Department of Forestry, Agricultural University Wageningen, 142 pp. van Rompaey R.S.A.R. (1994). Climat. Dans: Riezebos E.P., Vooren A.P. et Guillaumet J.L. (~ds). Le Parc National de Ta~, CSte d'Ivoire. Volume I: Synth~se des connaissances. Tropenbos s~ries 8, la Fondation Tropenbos, Wageningen, p. 42-50.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
421
Regionalization and parameterization of exchange processes at the land surface-atmosphere interface P. Kabat, A.J. Dolman, W.G.M. Bastiaanssen, M.J. Ogink-Hendriks & J.A. Elbers
DLO-Winand Staring Centre for Integrated Land, Soil and Water Research (SCDLO), P.O. Box 125, 6700 AC Wageningen, the Netherlands
Abstract
Regionalization of local scale modelling concepts is of paramount importance for realistic predictions of climate by Global Climate Models and for improved accuracy in regional hydrological modelling. A series of land surface experiments are providing a methodology and a unique dataset to develop and test regionalization algorithms. Most of this data is successfully used in analysis of local scale phenomena and the physical interpretation of remote sensing data. SCDLO is currently integrating these results into regional atmospheric and hydrological models.
1. S C O P E
Coupled hydrological and meteorological models are the basic tools to study the effects of changes in land use on water resources and climate. These models operate on spatial scales many times larger than for which the physics was originally developed, and at which the land surface changes appreciably. It is important to test the applicability of small scale physics at large scales and to develop parameterizations that take into account the heterogeneity of the landscape. In the development of the models, aggregation and disaggragation methods meet through a physically based description of the relevant processes (Fig. 1). The World Climate Research Program (WCRP) and the International Geosphere Biosphere Program (IGBP) have initiated a series of experiments to provide the required data to test and develop these parameterizations. The experiments have taken place or are planned in areas which suffer from human induced or climatic pressure on natural resources such as the Mediterranean (Bolle et al., 1992), the Sahel (Goutorbe et al., 1993) and the Amazon basin (Dolman et al, 1994). The basic
422
Disaggreg
C
/ Aggregat,on
l
o
u
~
Meso - scale Meteorological Model
L] Macro- scale Hydrological Model
Figure 1. Domain of mesoscale and macro scale models
strategy in these experiments in which the Staring Centre is participating is to combine simultaneous measurement of water, CO 2 and energy fluxes across the landscape at various scales with interpretive modelling and remote sensing at these scales. The basic aim of the modelling is to test and develop rules for deriving "effective" parameters for large scale applications from a knowledge of the small scale variability and the physics involved in defining the area averaged fluxes.
2. M E A S U R E M E N T S
Surface fluxes of evaporation, sensible heat, momentum and CO 2 are measured by eddy correlation; a technique which measures turbulent fluxes by correlating high frequency deviations from a mean for vertical wind speed and temperature, CO 2 an humidity. HAPEX-Sahel (Goutorbe et al., 1993) was the first experiment where these techniques were applied for such a long period under harsh environmental conditions. This data froms the basis for the development of detailed Soil Vegetation Atmosphere Transfer schemes (SVATs) which will link atmospheric exchange processes of energy and momentum with biophysiological control mechanisms in the vegetation. Preliminary results of SC-DLO measurements in HAPEX-Sahel show that the land-surface heterogeneity presented by two typical land cover types in the Sahel appears to have little effect on average seasonal evaporation (Fig. 2). In contrast,
423
!
E v
175 -
'-- 150 6_ r 125
Fallow savanna. ~ ....
E 200
"
150
t-
~ ~oo 0 Q. 75 > 50 E 0
Fallow savanna, Tiger bush "
....
.,i.-.,,
.c: E
25 0 230 240 250 260 270 280 D a y of y e a r
E~ :3
(O
100 50
o
-
i
-
i
-
i
-
i
-
i
-
230 240 250 260 270 280 D a y of y e a r
Figure 2. Cumulative latent and sensible heat fluxes at West-Central Supersite during the HAPEX-Sahel field experiment.
the sensible heat release into the atmosphere differs considerably between both land cover types. This means t h a t high resolution parameterization schemes for mesoscale modelling are needed to be able to describe the interaction of the land surface with the regional climate.
3. M O D E L L I N G
Models describing the control the vegetation exerts over transpiration and the flow of water through the u n s a t u r a t e d soil are calibrated for homogeneous land surfaces with the m e a s u r e m e n t s obtained by micrometeorological techniques. These SVAT models form the basic unit of the larger scale models (mesoscale) which operate with grid resolutions from several hundred meter to tens of kilometers. A non-hydrostatic mesoscale model (KAMM, Adrian and Fiedler, 1992) is used to study the effect of landscape heterogeneity on area averaged fluxes. The area averaged fluxes produced by this 3-D model are then compared with simpler representations (parameterizations) in which the heterogeneity is integrated into a single "effective" p a r a m e t e r for a large area.
4. R E M O T E S E N S I N G
Remote sensing provides the third corner in the strategy of these experiments by providing the means for areal extrapolation of locally obtained results. A new algorithm (SEBAL, Bastiaanssen et al., 1994) has been developed and tested for
424 the EFEDA (Bolle et al., 1992) and HAPEX-Sahel field experiments to produce the regional distribution of the latent and sensible fluxes.
5. CONCLUSIONS The series of land surface experiments are providing a methodology and a unique dataset to develop and test regionalization algorithms. Most of this data is successfully used in analysis of local scale phenomena and the physical interpretation of remote sensing data. SC-DLO is currently integrating these results into regional atmospheric and hydrological models.
6. R E F E R E N C E S
Adrian, G. and Fiedler, F., 1991. Simulation ofunstationary wind and temperature fields over complex terrain and comparison with observations. Beitr. Phys. Atmosph. 64: 27-48. Bastiaanssen, W.G.M., Hoekman, D.H. and Roebeling, R.A., 1994. A methodology for the assessment of surface resistance and soil water storage variability at mesoscale based on remote sensing measurements. Wallingford, IAHS, IAHS Special Publication No. 2. Bolle, H.-J., Andre, J.-C., Arrue, J.L., Barth, H.K., Bessemoulin, P., Brasa, A., de Bruin, H.A.R., Cruces, J., Dugdale, G., Engman, E.T., Evans, D.L., Fantechi, R., Fiedler, F., van de Griend, A., Imeson, A.C., Jochum, A., Kabat, P., Kratzsch, T., Lagouarde, J.-P., Langer, I., Llamas, R., Lopez-Baeza, E., Melia Miralles, J., Muniosguren, I.S., Nerry, F., Noilhan, J., Oliver, H.R., Roth, R. Saatchi, S.S., Sanchez Diaz, J., de Santa Olalla, M., Shuttleworth, W.J., Scgaard, H., Stricker, H., Thornes, J., Vauclin, M., and Wickland, D., 1993. EFEDA: European field experiment in a desertification-threatened area. Ann. Geophys. 11: 173-189. Dolman, A.J., Kabat, P., Gahs, J.H.C., Noilhan, J., Jochum, A.M. and Nobre, C., 1994. A large scale field experiment in the Amazon Basin (LAMBADA/ BATERISTA). Proc. CEC Climatic Change Symposium. Brussels, in press. Goutorbe, J.P., Lebel, T., Tinga, A., Bessemoulin, P., Brouwer, J., Dolman, A.J., Engman, E.T., Gash, J.H.C., Hoepffner, M., Kabat, P., Kerr, Y., Monteny, B., Prince, S., Said, F., Sellers, P. and Wallace, J.S., 1994. HAPEX-Sahel: a large sclae study of land-atmosphere interactions in the semi-arid tropics. Ann. Geophys. 12: 53-64.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
425
F o u r i e r a n a l y s i s of t i m e s e r i e s of N O A A - A V H R R N D V I c o m p o s i t e s to m a p i s o g r o w t h z o n e s M. Menenti a, S. Azzali a and W. Verhoef b aDLO-Winand Staring Centre for Integrated Land, Soil and Water Research (SCDLO), Wageningen, The Netherlands bNational Aerospace Laboratory (NLR), Emmeloord, The Netherlands
Abstract A measurement of the quality of vegetation zones in dynamic terms is provided by the Fourier analysis of time series of AVHRR]NDVI monthly observations over 10 years. The present approach is shown to be a powerful way to classify and extract various dynamic parameters of the vegetation in Southern Africa. The resulting m a p o f i s o g r o w t h z o n e s of Southern Africa was highly correlated to a radiational index of dryness or Budyko ratio as well as to the spatial distribution of vegetation types according to White vegetation map.
1. I N T R O D U C T I O N The timing and distribution of phytophenological events reflect the dynamic nature of the biosphere. Phenology is an extremely sensitive indicator of the many factors, such as climate, soils, and land management, that affect natural vegetation as well as agricultural crops. Satellite remote sensing provides a way to measure and monitor phytophenology on a global scale: time series of monthly values of NOAA-AVHRR Normalized Difference Vegetation Index (NDVI) extending over 10 years (1981-1992) for Southern Africa and South America were used in this study. The main constraint remains how to summarize the huge quantity of satellite information: time series analysis techniques provide an opportunity to identify and describe coherent structures in the observations using few parameters. The NDVI is obtained with the radiances observed by the Advanced Very High Resolution Radiometer (AVHRR) on board the polar meteorological satellites of the NOAA series. Radiances in the red (Chl) and near-infrared (Ch2) spectral regions are used:
426 - 200 ~
0.4
o 0.3
150 ~
0.2
IO0 ~
0.0
.
1
// n .
2
.-
3
.
4
5
u
~
~
6
7
8
I-I n
0 0
9
Figure 2. Measurements of different attributes obtained by the FFT algorithm analyzing monthly NDVI data from August 1981 till July 1990 from the same subarea of Figure 1.1 = average NDVI value, 2 = value of amplitude at 9 years, 3 = value of amplitude at 4.5 years, 4 = value of amplitude at 1 year, 5 = value of amplitude at 6 months, 6 = value of phase in degrees at 9 years, 7 = value of phase at 4.5 years, 8 = value of phase at 1 year, 9 = value of phase at 6 months.
2.2. C l a s s i f i c a t i o n o f i s o g r o w t h z o n e s The Fourier transform of the NDVI time series gives a rather large number of possible choices of attributes, viz. amplitude and phase values in different combinations, to apply either numeric classification or qualitative interpretation. A detailed intercomparison of classification results is necessary to understand which attributes (amplitude, phase, frequency) are most significant and whether the latter depends on differences between ecosystems. A classification test has been developed to establish which attribute combination and which classification procedure give the best classification performance. This test establishes the efficacy of the classifier and of the pattern features used to train the classifier. The performance of the classifier is measured in terms of separability, accuracy and reliability. The separability indicates if n spectral vectors can be distinguished in i classes. Separability is quantified by estimating signature divergences. Accuracy indicates how m a n y pixels have been classified correctly whereas reliability indicates the probability that a spectral vector actually belongs to the class to which has been assigned. An optimal classifier should maximize the values of separability, accuracy and reliability. Briefly, the procedure to map isogrowth zones was based on the following steps: - selecting dominant components in the observed Fourier spectra (amplitude values vs. period) e.g. 1 year, 6 months; - defining classes; class attributes are amplitude and phase values; - applying different numerical classification methods; - analysing the statistical significance of classification results. The assessment of classification procedures was done by considering each procedure to be a vector Ci with three coordinates: # Reliability was calculated as the integral of the curve obtained by plotting the
427 NDVI = Ch2 - Chl (1) Ch2 + Chl NDVI values fall in the range (-1, 1) and are rescaled to (0, 1023) for binary codification (10 bits). Spectral vegetation indices, like the one given in Eq. (1), provide a measure of the amount of green (photosynthetically active) vegetation [1,2]. Some authors [1,2] observed simple relationships between NDVI and primary productivity. Because of seasonality in forcing environmental parameters, e.g. rainfall, time series is a quasi-periodic signal as illustrated by the example given in Figure 1. 0.6
0.5[ 0.4 0.3 z
0.2 0.1 .0
s
o IIIllUUUununnuuluuluuunuhnnnnuunnnuluuuuululuuunlhuuununnlnnnnnnnul I
NOAA 7
NOAA 9
NOAA 1 1
Figure 1. Time series of monthly average values of NOAA-AVHRR NDVI from August 1981 till July 1990 for a sub-area in Africa (Okawango Delta, Botswana).
2. M E T H O D O L O G Y 2.1. A l g o r i t h m A time series of NDVI images, I(x,y,t) can be represented as a Fourier series of the form:
I(x,y,t) = ~
An(x,y) expi(cont - (~n(X,y))
(2)
When a Fast Fourier Transform (FFT) algorithm is applied to the time series of NDVI images, these images are decomposed pixelwise in a set of periodic functions with different periods (or frequency). The amplitude of each component An(x,y) accounts for a portion of the observed NDVI variability; for each function the phase lag (readily converted into a time-lag taking the period into account), ~n with respect to the origin of time in the observations is obtained. The amplitude values of different periodic functions for a given area are objective measures of variability of vegetation growth. Finally, time-lag values provide a measure of earliness or lateness of vegetation growth. Since the technique is applied on a pixel by pixel basis, images of amplitude and time-lag are obtained for each period [3]. Figure 2 shows values of different attributes obtained by the FFT algorithm analyzing monthly NDVI data from August 1981 till July 1990.
428 fraction of pixels classified at a given confidence level vs. the confidence level. Values ranged between 0 and 1. # Separability was the signature divergence (calculated with Jeffries-Matusita distance). Values ranged between 0 and 1.414. # Accuracy was obtained as (1-error) where the error is the number of pixels in the training set not correctly classified divided by the number of pixels in the training set. It is assessed by calculating the error of the classified pixels. Values range between 0 and 1. Next, a performance indicator (IP) was calculated as:
~ i ~ ~ref
IP=
(3)
~rei ~
where ere f is the classification vector with coordinates equal to the maximum values (1, 1.414, 1).
3. R E S U L T S The results of 15 different classifications have been compared assessing the performance indicator via the methodology of Section 2.2. The results are shown in Figure 3; scores of Procedures 1, 2, 6, 7, 9 and 12 were rather high.
IP[-] 0.9
iiiiiiiiiii!
...:
........... ,
iiiiiiiij iii~iiiiiill ..... :::.l ..... , ......... ;
.....
.iiiii-
ili!i
i:!:i:i:i
~i~iiiiii!
0.7 0.6
......... ......... ,........
i~iii~iiii
0.8
,
::.::
:~
',!ii!i!!i!!
............
:.:.:.:.: ......
.....ii9
i!iiiiiill .......... ii~!i!i i~i !~ii!i
iiiiiiiii! ~i:::/:i~i::i:4
0.5 0.4 0.3
.....
I
:i:i:i:!:i ......
0.2
0
:i:i:i:i:
.................... .... . . . . . . . . . . . . . . . . ................. ... :::::::::: :::::::::: ...........:.:.:.:.:
.............. ..... 9. . . . . . . . . . . . . . . . . . ...... . . . . . : . : . : . : .
.................. ,........., ..........
9. :-:-:.:-: :.:-:.:.: ........ 9. . . . . . . . . . . . . . . . . , ....,......... ..... ..................
:::::::::: :::::::::: :::::::::: :,:.:.:.:..:.:.:.:.: :.:.:.:.:. ..... ::::: :.:.:.:.:..:.:.:.:.: ..... ................:.:.:.:.: ...,..............,. ..... ......,................., .... ............,......, ,:.:.:.:.: ........... ......... ..... ..... , ..... ...........,.. ..... :::::::::: :::::::::: ::::::::::
i!!iii!i!i . . . . . . . . . . . . . . . . .i!iii!iii! ..
iii!iiii~ !iiii!iiiii :iiiiiiiiii i!i!i!iii! iiiiii!i .i!iiiiiiii ~ ~ii!~i~iii~i!!!iiiiii! .......... :i:i:i:i:i
:.:.:.:.:.:
.:.:.:.:.:
:::::::::::
!:!:i:i:i:
iiiii!i!i i!!!i!ii!i i!iii!iiii iiii~i~!ii
ii!?!i!ii! ,........
......... . . . . ... .. . . ....... . . .. . .. . ... . . . .
!iiiiiill
.... :iiii!iil
iii!iiiiii . . . . . . . . . . . . . . . . .!i!i!iii!i . . . . . . . . . . . . !iiiiiiii! .... . . . . .. .. .. .. .. .. .. . . . . . . . . . , .. .. .. .. .. .:...:.. : . : . :
iiiii!ililiii~ili!i~'"'"~"!
!iiiii!il
::::::::::
::::::::::
::::::::::
.
ili!iiiiiii!iiiiii ;ii!ilililiiiiiiiil iiiiiii;ii !iiiii!iiiiiiiiiii i~i~i~iiiiiiiii ~i~i~ ~i~i~i i~ii~~
i i i 1
::::::::::
ii!iiiilil i~!i!~!~!~ :.:.:-:.:. .9:.:.:.:.: ......... ...................
.....9q i::i::!::ii!::i ......:] ::::i:/:::i::i:/:
0.1
::::: :.:.:.:+ :.:.:.:.:. .9.:.:.:.:.: ........ ....... . . . . ......... .................... .. ........ ::::::::::
!:i:i:i:i: :.:.:.:.:,
2
3
4
5
6
iiii!i!i!ii
ii!!i!iil
7
8
..... ...........
:.:.:.:.:
:.:.:.:.:. . ....
::::ii::::~:i::]
i!i!i!iii
iiii!!!~i!
iiiii!ii!i
!iii:iiiii
i!i!iiiii!
9
10
11 12 13 14 15
Figure 3. Values of the classification performance indicator (IP) for 15 classification procedures. The map shown in Figure 4 was obtained by applying the classification procedure 2, combining a high classification performance with a good qualitative performance. Qualitative performance was assessed by comparing the outcome of
429 the numerical classification with: the map of Budyko index (ratio net radiation/ precipitation on the surface) [4], the White map (vegetation map) [5] and the Global GRASS database [6]. Since the qualitative correlation, especially of the A n images for the yearly and half-yearly components, was rather good, a regression analysis was done to assess more precisely the relationship of the NDVI time series with aridity (Budyko values) and vegetation type. High correlation was found between amplitude values at 6 and 12 months with the distribution of the types of the vegetation and the Budyko index values. The highest regression coefficients were calculated for the values of amplitude at 6 months. Figure 5 shows these results obtained using the best fitting curves (a, b) The correlation values for a. and for b. were 0.93 and 0.87 respectively.
Figure 4. Map ofisogrowth zones in Southern Africa; contour lines (black) indicate Budyko index values.
430 Z
0.16 -
"7' 0.16
D O
E <
E <
0.12
0.12
0.08
0.08
0.04
0.04
O
o~ B
0.00
i
0
o
o i
20
40
~
~
60
80
100 Budyko index value [-]
0.00
o
o
0 ,
~
~
0
10
20
30
i
i
40
50
i
i
60 70 White class[-]
Figure 5a. Best fitting curve (linear divided by linear) of the amplitude values at 6 months vs. Budyko index values; b. Best fitting curve (exponential) of amplitude values at 6 months vs. vegetation classes according to White.
COOPERATIVE FRAMEWORK Satellite data sets have been provided by C.J. Tucker, NASA, Goddard Space Center, USA. Data elaboration, methodology and results have been performed in the Netherlands by M. Menenti and S. Azzali at the DLO-Winand Staring Centre (SC-DLO) and by W. Verhoef and R. van Swol at the National Space Laboratory (NLR). D. Fuller and S. Prince (Dept. of Geography, Univ. of Maryland, USA) have contributed in the data interpretation on African natural ecosystems. Results are being evaluated in consultation with several international patners: UNEP-GRID in Nairobi, Kenya; CRICYT in Mendoza, Argentina, and others in six Latin American countries. This research is being sponsored by the Netherlands Board of Remote Sensing (BCRS) and by the Commission of European Communities.
4. R E F E R E N C E S 1. C.J. Tucker, C.L. Vanpraet, M.J. Sharman and G. Vanittersum. Satellite remote sensing of total herbaceous biomass production in Senegalese Sahel: 1980-1984. Remote Sensing of Environment, 17 (1985): 223-245. 2. S.D. Prince and C.J. Tucker. Satellite remote sensing of rangelands in Botswana II. NOAA AVHRR and herbaceous vegetation. International Journal Rem. Sens. 7,11 (1986): 1555-1570. 3. M. Menenti, S. Azzali, W. Verhoef, R. v. Swol. Mapping agro-ecological zones and time lag in vegetation growth by means of Fourier analysis of time series of NDVI images. Adv. Space Res. 13,5 (1993): 233-237. 4. D. Henning and H. Flohn. Climate aridity index map, in Explanatory note of United Nations Conference on Desertification (1977) A/CONF 74/31: 7-9. 5. F. White. The vegetation of Africa: a descriptive memoir to accompany the UNESCO/AETFAT/UNSO vegetation of Africa. Natural Resources Research, 20 (1983) UNESCO, Paris: 356 p. 6. Global GRASS database. CERL and Rutgers University (1993).
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
431
Fluxes over non-uniform vegetation: a numerical study C.M.J. Jacobs, J.P. Nieveen and A.F.G. Jacobs Agricultural University, Department of Meteorology, Duivendaal Wageningen.
2, NL-6701
AP
Abstract
A detailed, two-dimensional, second-order closure model for the Planetary Boundary Layer is used to study theoretically the fluxes and the state of the PBL over non-uniform terrain. Simulations are performed for a number of hypothetical surfaces, consisting of alternating patches of idealized forest and grassland. Inhomogeneities are chosen at different length scales, Lx, with Lx=2, 8 and 32 km. The results suggest that, in the absence of meso-scale circulations, the average fluxes are almost independent of Lx. The simulations confirm that, in order to compute an effective roughness length, the concept of blending height leads to satisfactory results.
1. INTRODUCTION The exchange of momentum, heat and mass (e.g., water vapour) between the surface and the atmosphere is rather well understood for homogeneous areas. However, large parts of the Earth's surface are inhomogeneous. The grid scale in large-scale models, such as climate models, often does not allow to resolve these inhomogeneities. Thus, an important question is as to how the processes in the Planetary Boundary Layer (PBL) are integrated, that is, how can the relevant fluxes or the surface characteristics be averaged so that correct boundary conditions for large-scale models are obtained ? The present study is performed in the framework of the Surface Layer Integration Measurement and Modelling project (SLIMM; [1-2]). Its major goal is to study the vegetation-atmosphere exchange for inhomogeneous terrain theoretically at the landscape scale (- 10 km).
2. MODEL AND METHOD The detailed, two-dimensional (2-D), second-order closure model of Rao and coworkers [3-6] is used to simulate fluxes and the state of the PBL. We extended the model
432 with the Penman-Monteith big-leaf model [7] to include a description of the surface fluxes. The surface layer part includes stability corrections [8] and a distinction between the roughness length for momentum, z0, and that for scalars, z0~ [9]. The reader is referred to the cited literature for a detailed description of the model. The model consists of a 1-D part and a 2-D part. The 1-D part simulates the diurnal evolution of the PBL over homogeneous terrain, with a time step of 2.5 s. The vertical grid is logarithmic between 10 and 421 m, and linear between 421 and 2000 m, with a maximum grid spacing of 79 m in the linear part. Initial profiles of temperature, humidity and windspeed must be provided, along with boundary conditions like geostrophic windspeed and solar radiation. The 1-D part generates the initial profiles and boundary conditions for the 2-D part. In its turn, the 2-D model describes a stationary situation for a terrain with horizontal extension x, using a step size of 1 m and with the same vertical grid as in the 1-D part. The simulations presented here were performed for fair-weather summer conditions in the mid-latitudes (45~ initial potential temperature, 0, and specific humidity, q, at 6 LT: 20~ and 9 g&g, respectively; geostrophic windspeed: 10 m/s; solar radiation at noon: 855 W/m2). In the initial profiles, a strong inversion was assumed at a height of 875 m (d0/dz=6 K/km; dq/dz=0.5 (g/kg)/km). The 1-D model was run for idealized grassland (roughness length, z0=0.05 m; albedo, a=0.2; surface resistance to evapotranspiration, rs=50 s/m) and the 2-D model was initialized with the 1-D output for 12 LT. In the 2-D simulation air was advected over 30 km of homogeneous grassland in order to certify that the fluxes had become independent of position before the air reached the first inhomogeneity. Next, the air enters the inhomogeneous part assumed to consist of alternating patches of forest (z0=l m; a=0.1; r~=100 s/m) and grassland (characteristics as before). The horizontal length scale of the inhomogeneities, Lx, was taken 2, 8 and 32 km, respectively (Fig. 1).
. . . . . . . .
,
~
Lx
--
l..x
--
. . . . . . . . . . .
Lx
2
km
8
k ~
32
km
~ ,
--
Figure 1. Arrangement of surface patches in the hypothetical inhomogeneous terrain, assumed in the simulations. Trees denote forest (50% of total), flat parts denote grass (50% of total). The total distance of the inhomogeneous part is 32 km.
433
3. RESULTS AND DISCUSSION
Simulated fluxes of momentum and sensible heat in the heterogeneous part are depicted in Figs. 2 and 3, respectively, for the various values of Lx. The influence of changes in the surface characteristics can clearly be seen. Above forest, the momentum flux is higher than above grass, because of the higher roughness length of forest. Due to 1.2
1.2
.... 1.0
.
,t:::l 9"
IN
~
.
.
1.0
.
~0.10
-
0.10
0.8
0.6 -"
0.4 0.2
~
. . . - - - 2- ~
o.~o
," J, - - - : 'o. ~O ~ Ls _~).,.
s
~.-,,
,..-,,,
,,,
,,-'x
-
.
o.~o
o.4o~ ,,,.-
/
~ /~ . .
,,~,,~,,
.
~
o:4o~"~/~
~ s/ ~ 11 . . . .
~ ....
~ s/" . .
i
~ ....
,
sr
,=
1 .0
e~i 0 . 8 ""
~
0.4
-'"
o.4o_/",,,_~ o . 4 ~
% 1~''~ . . . .
i""t
,,
i''t
,. . . . . . .
s""
,,
,,
.
,,
,
1-~ . .
.
,,
0.2
I-~
,
0.0 [-
1.2
f
-" . . . . . . . . . . :...--o.,o ~o_,o
I - ~ Fo.'~~
I-
--'"
- - ~ _
~,/'"
o~ p..-" ,
. . . . . . . . .
-
.'"
~o.,o~o.,o
~
0.20
.....
...s /
.-" ,,,
,
:
1.0
-
I,
_ o.2.o ~
~ . - - - - "~
.."
_o ", ..o., , 0 ~.
'5 .'" ,
.:.,,, , , , / . - ..~ . . ~ - , ~ , ,,,.,,) ,
o.o~ ~ . ~ ~ 1.2
~
..,_.,7~.~._.-~..._
"-, ~
. . . . . . . . . . . . . .
~ o . .
_ 'z.o ~
- . ~ . - : - . ~,.,,..,I , ,,, { . ...
....
.
.
.
q
"..
,'"
,
=:
.
.... ,"
o
O
0.8
- . . . . . . . .
",,
o , ..,o
1.0
.~
- . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
o . z o o . .
1 - - - .
o.~ k , . < > ~..~ - , ' k
o.
0.6 o.~o
0.0 1.2
IZ
0.10
0.8
o
Z
.
",
.
04
.
r~., , / . ~ .... ~:,~,/,
:,,.,<,~ ~ ~ o . 2
~~~5"/~_~~
to.o 1.2
...... "2i~;o
.-
~_..,~.~i
o,o;
"l"
___
o.1o ~ - :
o_.Eo. . . . . . . - - - , ' - % _
o
.-<.. .~. : : i
:
1.0
:{!7 \
"'b.~o ----0.80 -.,', --" . . . . . . . . . . . . . . . . '~! --- 0.70 ---------,, :-',1,'/ I
I---
-=.
"'.,"0 d/~.' il~
, \
~ ,o .~_ N
',
,s
/
~
~
,
"--
~:~
0.4 ~.
"20
,
: I
Z
0
0.8
--...
0.6 t).ou . . . . . . . . . . . . . . . . .
-"'.
.....................
--
0.2
_
__ .
i
I
I
24.0
~d.B-~
Distance (km) Figure 2. Contour plots of simulated momentum flux over the surfaces depicted in Fig. 1. Upper panel: Lx = 2 km; middle panel: Lx = 8 km; lower panel: Lx = 32 km.
434
the lower albedo and the higher surface resistance of forest the sensible heat flux is higher as well. Lower albedo implies less reflection of solar radiation. Thus, more energy is available to the sensible and latent heat fluxes. In addition, the higher surface resistance leads to less evapotranspiration (not shown here), so that more energy is used for the sensible heat flux.
1.2 1.0 ,z:
1.2
._•.• .
9o d::
0.8
"~o
0.6
_
--0-.05 .....
-0.05
[o.oo
-o.o5
"--o_00
[6.b~ . . . . . .
---
0.05
--. ~ _ o : ! o .
-0.05"
-0.-05
1.0
Z O
0.8
t
-0.05
b.oo
o.oo
o.o5
o.o5
~_o:~o
o.~o
0.6
N
N
0.'~ u
. - - - - .
.
.
--'" ...._..i"-~0.15 ~
i
0.4
Z
0.2
.
.
.
.
.
.
.
.
.
.
.
.
_. -...
0.15 - - ' - ' ~ - ~ - - ' - - - - ' ~ 0.15
-_
0.0
I
-o.0.~~
~-o~
I.o L
.z=o 9~
i,.~o
0.2
0.0 1.2
::r' 0.4
. . . . . . .
V__~~__-_~
-o.o5
o.8 ~o.oo ~ _ ' - "
~
:-o.~--
.
.
.
o.oo .
..........
.
-~o5
o.os -
-- ~ . o o - - ~ _ i ~
~
~
1.2
-o.os"
.... - ........
-0.0_5_
~
b.oa --~_-_-_~ - 6 . o o
_ __~__"....
1.0
Z o
0.8 N
0.6
9-~
3
,'"
i
Z
0.2
",
.,"
i
o "', . "
,-"
I
~ ~'•, ~%-"
'
' ~ ', .'
~-'"
" ..,
r
m/,'Z.'L.-.,g ~'" w, :
~
o",
~
,~
o
'
~6---,'/'
~
o.~ .~. ,I,/
....
0.2
,
o.o
0.0
~.2
-0.9_9_ - -0.05""-
-
-~.05
-----;4-'_:-0.8
.=
~
c-
1.2
~-----'0.05 _-.---'>
"= 9~
~
-
/
m/,'z.-" ',
m'/ . / ~ ""- 7 . ~ , ' , ~'1 ' 9
,.,
,"
-
--0.0~
---~
.___ ~ ,
1.0
___
.
~Z o
9
_o.os.--:~~-.q~
5-
~,,.o
0.6 --;.~..-:_::6:~-5-:~--
_ . . . . . . .
O.
_.
i: = 9
;:r.
0.4
-
""
,'"
. .........
0.2 O.Oo.~_~6"---7 :
~
.
~
..........
; o.~o 8.0
, 16.0
24.0
32.."
Distance (km) Figure 3. Contour plots of simulated sensible heat flux over the surfaces depicted in Fig. 1. Upper panel: Lx = 2 km; middle panel: Lx = 8 km; lower panel: Lx = 32 km.
435 Isolines are tilted in the direction of the horizontal wind (from the left to the fight) indicating the influence of one surface on the other. Furthermore, a larger part of the PBL is directly affected by a new surface if the patch size becomes larger, that is, if Lx increases. For Lx = 2 km the influence of the inhomogeneities is felt in the lower 40% (300-400 m) of the PBL only. Above this level, the PBL is able to average out the effects of the individual patches. If Lx is 8 km, the influence of the inhomogeneities extends almost to the top of the PBL. For Lx - 32 km the entire PBL is affected. This scale is in fact the regional scale at which feedback between the PBL and the surface becomes significant [10]. Furthermore, meso-scale circulations around inhomogeneities may become important at this scale [11]. The present model is not able to simulate such features and therefore, results for Lx -- 32 km should be interpreted with caution. Average fluxes over the inhomogeneous region are shown in Fig. 4. Lx has hardly any influence on the averages. This suggests that, in the absence of meso-scale circulations, the flux profiles can be calculated using a single set of surface characteristics, independent of the scale of the inhomogeneities. Such an approach has been tested here. An average surface was created by taking a=0.5"0.1+0.5"0.2--0.15 and rs=1./(0. 5/100 + 0.5/50)=67 s/m. An effective roughness length, z0.~, was calculated using the concept of blending height [12-14], by taking [13]:
t 1 ~) l t 05 ln(/b/Zo,
ln(/b/Zol)
05 l
(1)
ln(l b/z02)
1.2 - ~ , , .
1.2
9-" 1.0
,~1.0 o1.,~
.~
,.~ 0.8 o
o 0.6
........-..,
"~ 0.4
o~..,
"..............
0.8 K,, .... "'-...
0.6
"\
:,,.
~ 0.4,
@
NxN
Z o.2
Z
\
oo~ 9
0
, 0.1
""".... 0.2 0.3 0.4 0.5 0.6 0.7 Momentum flux (m2/s 2) .
:,
.
9
O.Od.1
0.0
0.1
0.2
0.3
Sensible heat flux (Krn/s)
Figure 4. Profiles of momentum flux (left) and heat flux (fight), averaged over the inhomogeneous part, over homogeneous forest and grassland, and over the surface with "average" characteristics (see text). Lines with squares: Lx-2 km; Lines with crosses: Lx-8 km; Lines with hourglasses: Lx=32 km; Solid line: average surface; Dotted lines: homogeneous forest; Dashed lines: grassland.
436 where z01 and z02 denote the roughness length of forest and grassland, and 1b is the blending height. According to (1), z0.e depends rather strongly on lb for lb<10 m, but is rather insensitive to 1b for lb>100 m. Thus, the results given above suggest lb>100 m. Therefore, we used the parameterization for 1b suggested by Claussen [14], according to which lb= 0.7z0(Lx/z0) 4/5, leading to lb = 249 m and 746 m for Lx = 2 and 8 km, respectively. Note that these values are in reasonable agreement with the results presented in Figs. 2 and 3. For Lx = 32 we obtain lb=2240 m, which would suggest that the entire PBL is strongly affected by the inhomogeneity. However, this value of Lx may be associated with the occurrence of meso-scale circulations, in which case the concept of blending height breaks down [14]. Using the values of 1b given above leads to z0.e=0.35 m. An additional run was made with the inhomogeneous area replaced by a homogeneous area with a---0.15, r~=67 s/m and z0=0.35 m. Results are also shown in Fig. 4. A satisfactory representation of the average momentum flux is obtained, but the average heat flux is somewhat underestimated. The latter feature corresponds to a slight overestimation of the latent heat flux (not shown here). This result can perhaps be improved if an alternative way to calculate average surface resistance and perhaps albedo would be used. Also, the calculation of roughness length for scalars might have to be adjusted.
5. R E F E R E N C E S
3 4 5 6 7 8
H.F. Vugts, A.F.G. Jacobs and W. Klaassen, This Volume (1995). J.P. Nieveen, C.M.J. Jacobs and A.F.G. Jacobs, This Volume (1995). R.S. Rao, J.C. Wyngaard and O.R. Cot6, Boundary-Layer Meteorol. 7 (1974), 331. R.S. Rao, J.C. Wyngaard and O.R. Cot6, J. Atm. Sci., 31 (1974), 738. J.C. Wyngaard, Boundary-Layer Meteor., 9 (1975), 441. J.C. Wyngaard & O.R. Cot6, Boundary-Layer Meteorol. 7 (1974), 289. J.L. Monteith, Symp. Soc. Exp. Biol., 19 (1965), 205. A.J. Dyer and B.B. Hicks, Q. J. R. Meteorol. Soc. 96 (1970), 715.
9
J.R. Garrat & B.B. Hicks, Q. J. R. Meteorol. Soc., 99 (1973), 680.
10 11 12 13 14
C.M.J. Jacobs and H.A.R. de Bruin, J. of Clim. 5 (1992), 683.
1
2
J.P. Pinty, P. Mascart, E. Richard and R. Rosset, J. Appl. Meteorol. 29 (1989), 976. J. Wieringa, Q. J. R. Meteorol. Soc. 112 (1986), 867. P. Mason, Q. J. R. Meteorol. Soc. 114 (1988), 399. M. Claussen, Atmos. Environ. 24A (1990), 1349.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
437
Sensible and latent heat flux over natural bog vegetation R.S. Singh a, J.P. Nieveen b, C.M.J. Jacobs b and A.F.G. Jacobs b "Central Arid Zone Research Institute, Jodhpur-342 003, India. Presently working as Research Fellow at Department of Meteorology, Duivendaal 2, 6701 AP Wageningen, The Netherlands bDepartment of Meteorology, Duivendaal 2, 6701 AP Wageningen, The Netherlands
Abstract Goudriaan's crop model is tested for its ability to describe the energy balance of a natural bog vegetation. Data used to drive the model were collected during the summer of 1994. Model results were compared with field data obtained by means of the eddy correlation technique. The present study indicates that the model results are satisfactory as both sensible and latent heat fluxes are overestimated by 6% only. Furthermore, sensitivity analyses have shown that the model results are sensitive to the soil surface resistance to evaporation. Hence, there is a need to incorporate the soil surface resistance to evaporation as a function of time.
1. I N T R O D U C T I O N Many physical and biological models have been developed to simulate the heat and mass exchange between arable crops and the atmosphere. Most of them were validated for different field crops in the past. Application of these crop models to natural vegetation will further identify areas of their strength and weaknesses under varying soil, vegetation, and climatic conditions. With this aim, in the present paper, Goudriaan's crop model [1] is tested for its suitability to describe the physical processes of a ~atural vegetation from a bog region.
2. MODEL, MATERIALS AND METHODS The model fomaulated by Goudriaan [1] simulates the micro-weather within crops as a function of vegetation, soil, and boundary conditions. Feedback of vegetation and soil on their environment is included in the model. The partitioning of the absorbed radiant energy into sensible heat, latent heat, and photosynthesis is calculated using the energy balance equations [2, 3]. For details about the model, the readers are referred to Goudriaan's monograph [1]. The study area is located in a bog region of the northern part of The Netherlands. The dominating type of natural vegetation is grass locally known as Pijpestrootje (Molinia Caerulea) growing in irregular humps over the region. The soil of the area is waterlogged peat mainly formed due to decomposition of dead organic matter. The canopy and the boundary weather observations were carried out under the SLIMM project [4, 5] during the summer 1994. Two kinds of meteorological data were collected. An input set, to drive the model, includes wind speed, vapour pressure, air temperature at 4.0 m (reference level) above the soil surface and global as well as net radiation above the vegetation. In another set, to compare the model results, fluxes of sensible and latent heat were measured by the eddy correlation technique. The grass and soil physical parameters were taken either from the measured data from the field or from the available literature. Some major input parameters used to drive the model for the standard run are shown in Table 1.
438 Table 1 Grass and soil characteristics used in the model Parameters
Unit
Grass height Leaf area index (LAI) Average leaf width Surface roughness (z 0) Internal regulatory CO2 concentration External CO2 concentration Thermal conductivity of the top soil layer Volumetric heat capacity of the top soil layer Water stress in the soil Soil surface resistance to evaporation (RESS)
m m2n1-2 m m vpm vpm Wm-'K -~ Jm-~K-1 bar sin-'
Value 0.75 1.50 0.005 0.07 240.0 350.0 0.08 2.4* 106 -0.1 500.0
Most of the area of the bog region was covered by the humps. The top layer of the humps, which constitute dead organic (grass) matter, was partially dry during the period of simulation. This partially dry layer of the dead grass on the humps reduced the water loss from the soil through evaporation. Therefore soil surface resistance to evaporation (RESS) was taken arbitrarily 500 sm ~ as input for the model. Measured data of the soil thermal conductivity, using the nonsteady-state probe method [6], were also used in this study. Further, a sensitivity analysis was done to judge the irnportance of the input soil and grass variables for the behaviour of the model. The model was run at different input values of the RESS: at 0 sm -1 to represent wet soil surface [1] and at 250 sin", which could be possible due to occurrence of rainfall and dewfall over the bog region. Besictes these the model was also run at different input values of LAI to see its influence in comparison to RESS for the same magnitude of variation.
3. RESULTS AND DISCUSSION The model was run for two consecutive days (August 30 and 31, 1994). The weather conditions used as input to drive the model, are shown in Figure 1. The simulated energy balance over natural vegetation is presented in Figure 2. The temporal variation of measured and simulated fluxes of sensible and latent heat are shown in Figures 3 and 4. The root mean square error (RMS) of the simulated sensible heat flux equals 13.9 Wm -2. The RMS is calculated using equation (1) in which H,,,~ and }-1~,are the nleasured and simulated
RMS=
/
i=i
N
/
sensible heat fluxes, respectively, at half an hourly intervals i, and N=96 is the number of half an hourly measurements. Similarly the RMS of the simulated latent heat flux equals 20.9 Wm -2. The Regression line forced through the orioine indicated that both sensible and latent heat fluxes were overestimated by about 6% (Figure 5 and 6). The results of the sensitivity analysis are presented in Table 2. It was found that soil surface resistance to evaporation, RESS has much influence on latent and sensible heat fluxes [7].
439 244
5
L~ t
TA.....
/..%
450
4 c~
.............
.E~ 2 5 0
S
144 ,-" ,-/,
;~.~,
. . . . . . . 4"15. . . . . ....... . ..... , / . , . , . . . . . . . . .............
2
-50
0
...............................................
6
12
18 24 30 Time (hour)
T i m e (hour)
Figure 1. Boundary weather conditions used in the study. The scale for air temperature (TA), water vapour pressure in the air (VPA) are given on the left ordinate and for wind speed (WS) on the right ordinate.
36
42
48
Figure 2. Net radiation used in the study (R,) together with sin-related sensible heat flux (H), latent heat flux (LE) into the air above the vegetation and soil heat flux (G) at 1 cm depth. 250
&'-" I E250 200
~
150
[
100
0
S,mulated/%",'"
I
............
0
6
12
18 24 30 T i m e (hour)
36
R M S = 20.9 W/m e
c-
42
.?22?rt:
48
Figure 3. Variation of measured and simulated sensible heat fluxes over natural bog vegetation during two consecutive days.
50 0 ...................... -50
,~
A
/ '-,,
.,...,
__1
............................
_50 1..............................................
A
~
Measured
I/~'
a::: A~/~ 50 1 ///'" r~, ~
t::::: 200g x 150 ----=
R M S = 13.g W/m 2
0
...............................................
6
12
18 24 30 T i m e (hour)
36
42
'
48
Figure 4. Variation of measured and simulated latent heat fluxes over natural bog vegetation during two consecutive days. 250
, H, = 1.0SH,. E '50
LE, = 1.06LEr,,
200-
iRe=~
~" 4,
I
g ,oo:r2
l
,,
~
.:.-..
I
~(.
§
§
.~ "
"
§
~-.- 100~
{.o~
5o-
§
§
o-
O
-50
.4, o.7,,," "" " t
***
150
*~
50.
-50
E
7"*'1
R e = 0.90
&..--
0
50
100
150
200
250
-50
-50
50
100
150
200
250
LEm (W/m2)
Hm (W/m2) Figure 5. Measured sensible heat flux (H,,,> versus simulated sensible heat flux (H,). The linear regression line forced through the origin is ~ = 1.06H m (N = 96).
0
Figure versus linear origin
6. Measured latent heat flux (LE,,) simulated latent heat flux (LEs). The regression line forced through the is LE.~ = 1.06LE m (N = 96).
440
The 50% variation in RESS value is quite possible under peat soil situations of the bog region. Also, unlike grass parameters, the value of the RESS even can vary significantly within a day due to occurrence of rainfall or dewfall. Therefore, it is likely to get an improve model results with time variable of RESS. Obviously, an increase in LAI of grass caused a substantial increase in CO2 assimilation. Table 2 Influence of some parameters on daily fluxes. The change with respect to the standard run is indicated by the arrow Variables
Unit
DNCO2 DNRAD DLHF DSHF DLHFB DSHFB DSOILF DNCO2 DNRAD DLHF DSHF
kg CO2ha -1 106jn1-2 106jm -2 106jm -2 106jm -z 106jm "2
106jm -2 : : : :
Standard
RESS
run
500-+0
500-->250
1.5-->3.0
1.5--+2.25
189.0 7.50 4.69 2.04 1.51 1.30 0.51
193.0 7.50 7.78 -0.75 5.29 -1.47 0.20
190.0 7.50 5.39 1.40 2.35 0.68 0.44
228.0 7.50 5.41 1.34 1.11 0.21 0.44
218.0 7.50 5.14 1.59 1.31 0.59 0.47
Net CO2 assimilation Net radiation Latent heat flux above canopy Sensible heat flux above canopy
LAI
DLHFB DSHFB DSOILF
: Latent heat flux at bottom : Sensible heat flux at bottom : Soil heat flux
4. S U M M A R Y AND CONCLUSIONS Goudriaan's model simulated sensible and latent heat flux reasonably well over the natural grass in the bog region. It is likely that an incorrect value of soil surface resistance to evaporation may lead to an error in the simulated fluxes. Therefore, incorporation of soil surface resistance to evaporation as a function of time may further improve the model results, particularly in a case of a longer period of simulation.
5. A C K N O W L E D G E M E N T S R.S. Singh has been supported by CEC, Brussels and DST, New Delhi.
6. R E F E R E N C E S 1 2 3 4 5 6 7
H.L. Penman, Proceedings of Royal Soc. of America, 193 (1948) 120. J.L. Monteith, Principles of environmental physics. Edward Arnold, London, 1973. J. Goudriaan, Simulation Monographs, Pudoc, Wageningen, The Netherlands, 1977. H.F. Vugts, A.F.G. Jacobs and W. Klaassen, This Volume (1994). J.P. Nieveen, C.M.J. Jacobs and A.F.G. Jacobs, This Volume (1994). W. van Loon, I. van Haneghem and J. Schenk, Int. J. Heat Mass Transfer, 32 (1989) 1473. R.S. Singh and A.F.G. Jacobs, Neth. J. Agric. Sci., Submitted (1994).
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
441
SLIMM-PROJECT
H. Vugts", A.F.G. Jacobs b, W. Klaassen ~ a) Free University of Amsterdam (VUA), Department of Meteorology, De Boelelaan 1085, NL 1081 HV Amsterdam, The Netherlands b) Wageningen Agricultural University (WAU), Department of Meteorology, Duivendaal 2, NL 6701 AP Wageningen, The Netherlands c) State University of Groningen (RUG), Department of Physical Geography, Kerklaan 30, NL 9751 NN Haren, The Netherlands
1. INTRODUCTION
An important aim of BAHC (Biological Aspects of the Hydrological Cycle) is to find the average atmospheric exchange at the gridscale of General Circulation Models (GCM's). To achieve this goal, scaling-up rules must be developed to average local observations. At an intermediate scale of landscapes, scaling-up rules depend on advection in the atmospheric surface layer above individual patches. The available scaling-up rules at this scale show considerable scatter and have hardly been validated. Moreover, recent studies indicate that scaling-up rules may strongly underestimate the influence of surface heterogeneities on the average landscape exchange. The general aim of the SLIMM project (Surface Layer Integration Measurement and Modelling) is to determine the soil-vegetation-atmosphere exchange of momentum, heat (sensible and latent) and carbon dioxide at the landscape scale (...10 km). At this scale most landscapes are inhomogeneous. Average fluxes at the landscape scale are at present simply estimated by direct averaging of the fluxes of the elements, or, by estimating average values using the concept of blending height. The method of using blending height is to be extended and tested in a heterogeneous landscape. The resulting method should be the first step in scalingup local observations to areal averages for GCM's. To achieve the general aim of the SLIMM project, an intensive cooperation has been started between three Dutch universities (Amsterdam, Wageningen and Groningen. In this cooperation an intensive two years' measurement programme is carried out over an inhomogeneous terrain. Moreover, various computer simulations have been initiated in which the experimental evidence is used to validate these models.
442 2. SITE CONDITIONS
The region from Norg to Fochtelo~rveen in the north of The Netherlands has been selected as experimental site for collecting data. This location has been indicated in Figure 1.
Foc hte
ar ea
loot
t
Figure 1.Fochtelo~rveen location
............. ! ili!:ii:i!:!ii ~i
in The Netherlands.
The whole area of interest can be subdivided into three sub-sites; a forest area (Groningen), a bog area (Wageningen) and an area mainly consisting of arable land (Amsterdam). Each subregion has a principle investigation group notated in brackets. The forest site is a combined coniferous/deciduous forest located about 2 km NE from the natural bog site. The agricultural site consists of grass and berry bushes and is situated about 2 km NE from the forest location. This specific region has been selected for the following main reasons: 1) The region has a marked surface heterogeneity within the 10 km scale, our scale of interest. 2) This region is far away from major surface heterogenieties like land-sea interfaces. 3) In the same area a second hydrological experiment is executed in the same period which allows an intensive cooperation with other groups. 4) The region is situated close to one of the universities (Groningen) which guarantees supplementary manpower during the experimental period. 5) The Fochtelo~rveen area is the largest bog relict in the Netherlands and is extremely useful to study the exchange of greenhouse gases like CO2 and CH4.
443
3. TIME SCHEDULE
The main observations will take place continuously in a two year's period (1994 - 1995) above all the three subregions. The radiation fluxes, the turbulent fluxes of heat, mass (water vapour and carbon dioxide) and momentum and the soil fluxes of heat will be measured continuously. Moreover, during this period, the vegetation and soil characteristics of all three areas will be monitored like, for example, the Leaf Area Index (LAI), foliage area distribution, ratio between dead and living material and soil water content. In addition to the continuous measurement programme, at least three Intensive Field Experiments (IFE's) are planned to investigate areal variations, local advection and regional averages in more detail. One of the IFE's will be focused on the change in surface conditions around the bog-forest interface. Here special attention will be focused on the within-forest and up-wind flow field and the static pressure [1, 2] around the bog-forest interface. During the IFE's but also incidentally during the continuous measurement period, the following observations of landscape averages will be executed: 1) Boundary layer observations of 3-Dim wind and structure parameters by using a SODAR. 2) Boundary layer observations of the temperature by using a so-called RASS system.
4. MODELLING
The observation results will be used to validate various existing models and, if necessary, to extend these models with the goal to develop advection rules for application in meso-scale models. For example the models of Klaassen [3] and the model of Meesters [4] will be used. Extending local advection to the 10 km scale implies that the influence of multiple step changes will be analyzed as well. Primarily a smooth-rough-smooth transition will be studied with smaller variations within these elements. For this study the second order model of Kroon [5], and the extended model of Rao [6] will be used. In day-time, the planetary boundary layer (PBL) has a height of about 1000 m and has horizontal variations of temperature and windspeed that can be neglected at the 10 km scale. At night and possibly after rainy periods, however, a much shallower boundary layer occurs. This means for the boundary layer that
444
can influence the integration rules of the land surface-atmosphere exchange and will be investigated. Existing, but not yet calibrated, up-scaling rules are based on the concept of a blending height. At this height, the local variations are thought to merge into a regional average. In literature [7, 8,9 ] a first estimate of this height shows a variation of an order of magnitude. The height of blending might be estimated from elevated measurements at different locations in the landscape and is expected to relate to the scale of heterogeneity and atmospheric stability. Attention will be focused on a technique to arrive at an accurate measure for this height. 5. FUNDING
The SLIMM research programme is part of the Dutch National Research Programme on Global Air Pollution and Climate Change. 6. REFERENCES
[1] Jacobs, A.F.G., 1984: Static pressure around a thin barrier. Archiv Meteor. Geoph. Bioclim., Ser. B35: 127-135. [2] Jacobs, A.F.G., Van Boxel, J.H. and Shaw, R.H., 1992: The dependence of within-canopy stratification parameters on within-canopy turbulence properties. Boundary-Layer Meteorology, 58: 247-256. [3] Klaassen, W., 1992: Average fluxes from heterogeneous vegetated regions. Boundary-Layer Meteorology, 58: 329-354. [4] Meesters, A., 1991: Thermally-forced meso-scale circulation in tidal areas. PhD thesis, Free University Amsterdam, pp 180. [5] Kroon, J.L.M., 1985: Profile derived fluxes above heterogeneous terrain: a numerical approach. PhD. thesis, Agric. Univ. Wageningen, the Netherlands. pp 159. [6] Jacobs, C.M.J., Nieveen, J.P. and Jacobs, A.F.G., 1995: Fluxes over nonuniform vegetation: a numerical study. This volume. [7] Wieringa, J., Roughness-dependent geographical interpolation of surface wind speed averages. Quart. J. Royal Meteorol. Soc., 112: 867-889. [8] Mason, P.J., 1988: The formation of areally averaged roughness lengths. Quart. J. Royal Meteorol. Soc., 114: 399-420. [9] Claussen, M., 1991: Estimation of areally-averaged surface fluxes. BoundaryLayer Meteorology, 54: 387-410.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
445
Exchange processes of a natural bog vegetation; SLIMM measurements J.P. Nieveen, C.M.J. Jacobs and A.F.G. Jacobs Department of Meteorology, Wageningen Agricultural University, Duivendaal 2, 6701 AP Wageningen, The Netherlands
Abstract The Surface Layer Integration Measurement Modelling project (SLIMM)is set up to determine the soil-vegetation-atmosphere exchange of momentum, heat, water vapor and carbon dioxide at a regional scale. Starting point of the joint experiment is that the exchange processes above a heterogeneous landscape is not a simple weighed sum of these processes above the homogeneous parts of the composing elements. Therefore an experiment is set up during which the exchange processes of the different components are measured.
1. INTRODUCTION The experimental area is in the region from Appelscha to Norg in the north of the Netherlands. The site of the Wageningen Agricultural University (WAU)is located at the prevailing windward end of the experimental area, above a bog landscape type. The continuous experiment takes place during a two year's period (1994-1995). During this period flux measurements, representative for the bog area, will be performed as good as possible on a routine basis. In addition to the continuous measurements, three more detailed so called Intensive Field Experiments (IFE) are planned. The main goal of the IFE's is to provide insight into the mechanisms that are responsible for the areal fluxes from the complex bog area as for the total fluxes of the whole measurement site.
446
The contribution of the WAU to the SLIMM project include a continuous monitoring of the surface fluxes of radiation and profile measurements of temperature and wind speed. Second, simultaneous monitoring momentum, heat, water vapor and carbon dioxide fluxes during the two year's period. Third, temperature, water vapor and carbon dioxide fluctuations and mean water vapor and carbon dioxide concentration will be measured, and four, a detailed description of the temporal changes of the architecture of the vegetation and soil features. These points will be discussed here.
2. STANDARD METEOROLOGICAL MEASUREMENTS Two masts are placed at a representative location within the bog area (20 meters apart). To reduce shading effects, the radiation sensors are spread out over the two masts. All components of the radiation balance are measured using radiometers (Kipp & Zonen, CM 5) for short-wave incoming and reflected radiation (Albedo), a Funk net radiometer (Middleton) and a pyrgeometer (Kipp & Zonen, CG1) for Iongwave radiation. Figure 1, shows an example of the measured radiative fluxes.
8ool 700]
o,,.",/"".. ',
6001 ~" 500E 400v X Z3
300 200100-
O-100
Figure 1
0
300 600 900 1200150018002100 Time (GUT) Radiative fluxes measured at the 3rd of August 1994. Rn = Net, Sin-- Short wave incoming, Sour = Shortwave outgoing, Lin = Longwave incoming and Lout = Longwave outgoing radiation.
447
Apart from the radiation measurements, the following standard meteorological quantities will be measured: dry and wet bulb temperatures (aspirated Pt-100 psychrometers) at three heights as well as wind speeds at the same heights (sensitive cup anemometers, length constant l m), at four depths soil temperature using Pt-100 thermometers (0.05, 0.2, 0.5 en 1 m) and soil heat flux using a TNO transducer (WS 31 CP). A wind vane is used to measure the mean wind direction. The profile measurement of temperature and wind enable, by using various meteorological techniques (for example: the aerodynamic and Bowen ratio energy budget approach), to make assessment of fluxes of momentum, heat and water vapor.
3. SURFACE FLUXES To make a comparison of the surface fluxes calculated with the Bowen ratio or aerodynamic technique and direct measurements of the surface fluxes, an eddycorrelation system has been installed. The system consists of an ultrasonic anemometer/thermometer (Gill Instruments Ltd.), a fast response thermometer and a CO 2/H20 infrared gas analyzer (LiCor, Li 6262). This technique also allows us to add two relatively new techniques to the measurement program, namely: 1. The standard deviation or fluctuation technique [1] for heat, water vapor and carbon dioxide. 2. The structure parameter method [2] for heat, water vapor and carbon dioxide. Figure 2 shows the surface fluxes as measured with the eddy-correlation system and the net radiometer. Large errors can occur when the covariances are directly calculated from the measurements, so several corrections should be carried out to obtain the correct values (eg. Axis rotation and frequency response correction). These corrections are important, and should be carefully looked at in this experiment [3,4].
4. CARBON DIOXIDE The increase of carbon dioxide is the main cause for global warming, possibly resulting in climate change. The global carbon dioxide cycle is only partly understood [5], as the exchange between the atmosphere and oceans and vegetated
448
soils is still poorly quantified. A main aim of the SLIMM project is to obtain further insight into COa-exchange processes. 600
~
500 40004
E O0 X
300-
H
fX
~~,,,'; LE
/
200-
~\i"."., ~ ,.',;i~ :
,,.,
. :.,. ,, ~ ,
,
/ ..d..'"~ W"",v i~\ /.I....
m
I.L
100
-100
an
i i i 1 1 1 1 1 1 1 1 1
0
iiii
iii
|111111
i .....,.f,, ,", ',\
iii
ii11]
i l l i l
iii
iiii
i
300 600 900 1200150018002100 Time (GMT)
Figure 2:
Measured sensible heat (H) and latent heat (LE) flux and net radiation (Rn) on the 3rd of August 1994.
Carbon dioxide is a so-called spore gas. This means that measuring the COa flux density very serious errors emerge for which correction should be carried out. When CO 2 fluctuations are directly correlated to the vertical wind component fluctuations by eddy-correlation, density fluctuations affect the vertical velocity fluctuations. Density fluctuations are caused by heat and water vapor quantities. To be able to carry out the corrections properly, together with the CO2 flux the fluxes of heat and water vapor have to be measured. A so-called fast suction technique is used for taking air samples. These air samples are directly analyzed in the field for their water vapor and CO 2 components and the results are correlated with the vertical wind component near the sampling intake at 7 meters above the soil surface. Density corrections or Webb-corrections are applied later [6]. The CO2 concentration and flux will be measured all year round, so some estimates of the soil contribution to the total flux could be made.
5. SURFACE CHARACTERISTICS Throughout the growing season the state of the vegetation will be monitored.
449
This means that at representative locations within the bog area the leaf area index (LAI) and the vertical distribution will be estimated. Moreover estimates will be made about the horizontal variability. For the LAI measurements two techniques will be used; first, a direct method where leaf area is measured optically (PC hand scanner), second an indirect technique where the extinction of direct irradiation within the canopy is measured (Delta-T sunfleck ceptometer). By selecting a representative measurement site for the meteorological station, the obtained data will provide adequate information about two important surface characteristics: the roughness length, z o and the displacement height, d and their courses during the experiment. Moreover estimates will be made of the roughness length for heat, Zo,, for this characteristic can differ much from the z o for momentum. At various locations in the bog area, soil samples will be taken to obtain assessment of important soil parameters like: thermal conductivity, moisture content, soil composition and heat capacity.
6. C O N C L U S I O N S Because the measurements have just started, but will continue for an other year no hard results can be shown here. Some of the results are used for canopy simulation models and will be used in a numerical boundary layer study. It's clear that the eddy correlation method is not a very straight forward method but significant corrections should be applied. Especially when CO 2 fluxes are measured, the Webb corrections are of major importance. Attention should be given to soil and vegetation cycle and their contribution to and influence on the surface fluxes.
7. R E F E R E N C E S
[1] [2] [3]
De Bruin, H.A.R., 1994; Boundary-Layer Meteorology, 68, pp. 427-432. Kohsiek, W, 1982; Boundary-Layer Meteorology, 24, pp.89-107.
[4]
Moore, C.J., 1986; Boundary-Layer Meteorology, 37, pp. 17-35. MCMillen, R.T., 1988; Boundary-Layer Meteorology, 43, pp. 231-245.
[5]
Watson, R.T., Rodhe, H., Oeschger, H and Siegenthaler, U., 1990; In: Climate change: the IPCC scientific assessment, Cambridge.
450 [6]
Webb, E.K., Pearman, G.I. and Leuning, R., 1980 Quart. J. Royal Meteorol. Soc., 106, pp 85-100.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
453
ASSESSMENT REPORT ON NRP SUBTHEME "GREENHOUSE
GASES"
SOURCES AND SINKS OF CO2 CH4 AND N20, DATABASES AND SOCIO-ECONOMIC CAUSES
J.J.M. Berdowskil A.F. Bouwman2 W.M. Kieskamp3 J. Slanina3
1Netherlands Organization for Applied Scientific Research (TNO-MW) P.O.Box 6011, 2600 JA Delft, The Netherlands 2National Institute of Public Health and Environmental Protection (RIVM) P.O.Box 1, 3720 BA Bilthoven, The Netherlands 3Netherlands Energy Research Foundation (ECN) P.O.Box 1, 1755 ZG Petten, The Netherlands
With contributions by: H.G. van Faassen, P.J. Kuikman
W. M. Kieskamp W. Ruijgrok, M. Vosbeek, H. Spoelstra G.M.J. Mohren N.H. Batjes, E.M. Bridges, C.R. Oldeman
AB-DLO, Research Institute for Agrobiology and Soil Fertility, Wageningen ECN, Netherlands Energy Research Foundation, Petten KEMA Environmental Services, Arnhem IBN-DLO, Institute for Forestry and Nature Research, Wageningen ISRIC, International Soil Reference and Information Centre, Wageningen
454 J.A.M. de Bont, H. Denier van der Gon, A. van Dasselaar, B.O.M. Dirks, J. Goudriaan, H.J. Heipieper, P. Hofschreuder, P. Leffelaar, J. Lelieveld, S.W.M. Kengen, J.C. Koops, O. Oenema, R. Segers, A.J.M. Stams, D. van Veenhuysen, G. Velthof
LUW, Wageningen Agricultural University
C.G.M. Klein Goldewijk, C. Kroeze, RIVM, National Institute of Public R. Leemans, C.W.M. van der Maas, Health and Environmental J.G. van Minnen, J.G.J. Olivier, W.L.M. Smeets, Protection, Bilthoven R.J. Swart J. Oonk, J.I. Walpot
H.P. Baars, J. Baas, H.S.M.A. Diederen, J.H. Duyzer, J.C.Th. Hollander J.G. de Beer, A.P.C. Faaij
TNO-M&E, Institute for Applied Scientific Research Environmental and Energy Technology, Apeldoorn TNO-MW, Institute for Environmental Sciences, Delft RUU, University of Utrecht
455
Contents Abstract 0
0
0
0
0
0
7.
Introduction 1.1. Aim and organization 1.2 Assessment of the uncertainties in sources and sinks of greenhouse gases at the time of the start of NRP 1.3 Assessment of the developments in the knowledge on sources and sinks of greenhouse gases during the course of NRP phase I Carbon d i o x i d e (CO2) 2.1 Overview of the C02 cluster 2.2 Methodology 2.3 Results 2.4 F u t u r e research M e t h a n e (CH4) 3.1 Preparatory studies and organization 3.2 Methods 3.3 Results 3.4 Integration of results 3.5 F u t u r e research needs N i t r o u s o x i d e (N20) 4.1 Preparatory studies and organization 4.2 Methods 4.3 Results 4.4 Integration of results 4.5 F u t u r e research needs Emission database development 5.1 World Inventory of Soil Emission potentials (WISE) 5.2 Emission Database for Global Atmospheric Research (EDGAR) 5.3 F u t u r e research needs with respect to database development Socio-economic causes 6.1 Methods 6.2 Results 6.3 F u t u r e research needs R e f e r e n c e s and p u b l i c a t i o n s
456 ABSTRACT The aim of the subtheme Greenhouse gases of the Dutch National Research programme on (NRP) is to quantify the sources and sinks of the major greenhouse gases to enable estimates of the future atmospheric concentration. The major part of the projects in this theme is focused on the Dutch situation, but the results can be extrapolated countries or regions. The information gained will be used for Dutch policy decisions regarding abatement of greenhouse gases. S e c t i o n 1 deals with the aim and organization of Causes of climate change, and relates the scope to increased awareness of uncertainties in sources and sinks of greenhouse gases: at the start of the National Research Programme the general consensus of the scientific community was t h a t these uncertainties were not extreme large, it is nowadays accepted that these uncertainties are larger t h a n assumed before. The aim the Cluster CO 2 ( S e c t i o n 2) was devoted to study the exchange between t e r r e s t r i a l ecosystems and the atmosphere to gain more knowledge of the "fertilization" flux. The research was mainly focused on the development of a CO2 exchange model for grassland describing diurnal and seasonal fluxes, and on the validation of this local scale model on a regional and national scale. In both the clusters CH4 and N20 (respectively S e c t i o n 3 and S e c t i o n 4) anthropogenic and biogenic sources were studied. Major criteria to study sources were the source strength, but also the uncertainty in the source estimate and the potential emission reduction, all projected on the Dutch situation. Exception were the projects on CH4 emission from rice fields, and the sea/air exchange of N20 in oceans; expertise was available in The Netherlands to carry out these studies. As in the sub-theme CO2 the study of processes in grasslands was given a high priority in the sub themes CH4 and N20 in order to quantify emission the mentioned greenhouse gases. Moreover, in the CH4-sub theme projects were performed to evaluate and validate the strength of various sources. The two remaining clusters (limited in extend) were aimed at the development of emission databases and geographic quantification of soil processes controlling greenhouse gas fluxes (cluster Database Development, S e c t i o n 5), and on national inventories (cluster Socio-economic Causes, S e c t i o n 6). In the framework of the first cluster two databases were developed, one was the World Inventory of Soil Emission potentials (WISE), a global gridded database of the primary soil factors controlling soil greenhouse gas emissions, and the other was Emission Database for Global Atmospheric Research (EDGAR) aimed to describe the processes as land use, energy consumption etc, which control the emissions of greenhouse gases and other air pollutants. The goal of the other sub theme was to develop and apply methodologies to compile national inventories of greenhouse gas emissions in The Netherlands, focused on the compounds CH4 and N20.
457 1.
INTRODUCTION
1.1 Aim and organization
The scope of this theme was the a s s e s s m e n t of the causes of Dutch National Research P r o g r a m m e Global Air Pollution and Climate Change (NRP). This obviously includes the cycle of the greenhouse gases but also the description of anthropogenic activities resulting in changes in atmospheric concentrations. The aim of the causes of climate change was to provide information needed to quantify the sources and sinks of the major greenhouse gases in order to enable more accurate e s t i m a t e s of future atmospheric concentrations. While most of the programmes within National Research Programme on were defined in a bottom-up process, it was decided to organise the research on cycles of greenhouse gases in a different way. First, a first order analysis of the uncertainty in the cycles of CO2 (carbon dioxide), CH4 (methane), and N 2 0 (nitrous oxide) was made by a small group of Dutch scientists, active in this field. The conclusions of this analysis were t h a t the main u n c e r t a i n t y in the cycle of CO2 was caused by insufficient information on i m p o r t a n t sinks such as u p t a k e by oceans and t e r r e s t r i a l ecosystems. More knowledge on the level of mechanistic descriptions was clearly needed, but the process of integrating locally derived information regarding important sinks of CO2 up to the level of relevant descriptions of CO2 exchange on regional, continental, or global scale also introduce large uncertainties. The same situation was observed r e g a r d i n g the emissions of methane. Also in this case local m e a s u r e m e n t s are extrapolated to regional and global scale, introducing large errors. For both CO2 and CH4 the scaling problem, this is generalisation of local m e a s u r e m e n t s to regional and (sub)continental scale, was seen as a serious problem. In the case of N20 the state of knowledge was worse, information on the mechanisms of e.g. emission of N20 during nitrification or denitrification was not really available. Second, it decided t h a t coherent programs would be formulated for each of these greenhouse gases, which should address the mentioned m a i n problems. These programs should be formulated in such a way t h a t on one h a n d typical Dutch aspects of the cycle of greenhouse gases would be emphasised and t h a t on the other hand the information gained in the research would constitute a worthwhile contribution to international programs, primarily within the scope of IGAC. Based on these considerations the following clusters of projects were developed: s t u d y of the CO2 exchange between grasslands and the atmosphere, and development of methodology to validate CO2 exchange models for larger areas; investigation of CH4 emissions of selected sources, thought to contribute significantly and with a large margin of error, and development of methodology to validate CH4 exchange models for larger areas; quantification of N 2 0 emissions from fossil sources and i m p o r t a n t biogenic systems, in particular from grasslands, sewage t r e a t m e n t systems and from freshwater and marine systems.
458 Table 1.1 List of projects in the NRP subtheme "Greenhouse gases" by clusters Title
Project leade r
Numbe r
Carbon dioxide (C02) ecosystem studies, model development and validation The seasonal cycle of the CO2 exchange between J. Goudriaan 852062 atmosphere and vegetated surfaces Quantification of carbon fluxes in grassland
P.J. Kuikman
852063
A feasibility study for aircraft based flux measurements of CO2
W. Ruijgrok
852065
The development of a geographically explicit dynamic carbon cycle model
R Leemans
852067
Quantification of carbon fluxes in Dutch forests (Part 1: Desk study)
G.M.J. Mohren
852071
Determining relative importance of sources and sinks of carbon dioxide using carbon isotope measurements
W.M. Kieskamp
852076
Measurement of the exchange of C02 between the atmosphere and a grassland
W. Ruijgrok
853116
R.J. Nielen
850008
Methane (CH4) biogenic sources, fossil sources Quantification of methane emissions due to natural gas losses and petroleum production The influence of soil parameters on the production and emission of methane in/by wet rice paddies
N. van Breemen 850009
Greenhouse gases from landfills in the Netherlands
C. Verschut
850023
Methane formation by anaerobic consortia in organic grassland soils
A.J.M. Stams
853120
Programming study for methane research Validation of source strengths of atmospheric methane using carbon isotope ratios.
J.J.M. Berdowski 852068 W.M. Kieskamp 852097
Quantification of methane emissions in the exploration and production of natural gas and petroleum The Netherlands
J. Oonk
853104
459 Measurement study landfill gas production emission and recovery
J. Oonk
853105
Effects of grassland management on the emission of CH4 from grassland on peat soils
O. Oenema
853121
The methane consumption by indigenous grassland micro flora
J.A.M. de Bont
853122
From methane formation and oxidation to methane fluxes in organic grassland soils: modelling
P.A. Leffelaar
853123
Evaluation and validation of the CH4 emissions in the Netherlands and contributions from various sources
J.C.Th. Hollander 853124
Determination of emissions of methane in rural areas
P. Hofschreuder
853125
N20 emission from fossil fuel combustion in power plants
H. Spoelstra
850006
Preliminary study on N20 flux measurements
H.S.M. Diederen 850012
Investigation of the contribution of traffic to N20 emissions both now and in the future
J. Baas
850030
Effects of nitrogen fertilization and grazing on the N20 emission from grassland.
O. Oenema
852073
Factors influencing the ratio N ~ 2 0 as nitrate is removed from the soil by denitrification The emission of N20 from grassland
P.A. Leffelaar
852074
H.G.v. Faassen
852078
Modelling of soil emissions of nitrous oxide for global studies
A.F. Bouwman
852079
Measurement of atmospheric emissions of N20 from biogenous surface sources in general and grasslandecosystems in particular
J.H. Duyzer
852096
Nitrous Oxide (N20) biogenic sources, fossil sources
Database development emission database, geographic quantification of soil controlling gas fluxes Global emission database
J.J.M. Berdowski 850032
Geographic quantification of soil factors and soil processes that control fluxes of greenhouse gases (Currently used acronym: WISE, World Inventory of Soil factors and processes that control Emissions of greenhouse gases)
E.M. Bridges / N.H. Batjes
851039
460 Emission Database for Global Atmospheric Research (EDGAR); Phase 2: data collection and implementation
J.G.J. Olivier
851060
R.J. Swart
850019
Socio-economic causes national inventory, policy analysis Social causes of the greenhouse effect and emissions inventories
The information gained in these clustered projects should contribute to existing or future emission data bases. In the CO2 cluster, a specific project was formulated to e n s u r e t h a t the results of this cluster was t r a n s f e r r e d to the EDGAR and IMAGE data base (See Annex 2 for acronyms). In the CH4 and N20 clusters this t r a n s f e r was regulated in a less formal way, mainly because the estimate was at the time of the start of this work, t h a t insufficient information would be available after 3 years to perform this task rigorously. These clusters started fairly late in the NRP phase-I, so only some two and a halve years were available for these studies, in stead of five for most of the other NRP-I projects. 1.2 A s s e s s m e n t of t h e u n c e r t a i n t i e s in s o u r c e s a n d s i n k s o f g r e e n h o u s e g a s e s at t h e t i m e of t h e start of N R P At the time of the s t a r t of NRP-I (1989-1990) general consensus existed in the scientific community as expressed in the 1992 IPCC report, on the uncertainties in the predictions of climatic changes. It was widely accepted t h a t it was possible to predict future concentrations of greenhouse gases, with a considerable margin of errors of course. An important problem was perceived in the translation of these changes in concentrations of the radiative active gases in changes in the radiative balance of the earth. But the main problem, as perceived in t h a t period (as reflected in the discussions during the Chamrousse conference in 1989) were the effects of these shifts in the radiative budgets expressed in t e r m s of possible climatic changes.
The Carbon cycle T h a t the above mentioned consensus existed regarding the Carbon cycle was r e m a r k a b l e , in view of the fact t h a t already m a n y observations were available which indicated t h a t the uncertainty in sources and sinks of greenhouse gases including CO2 were much larger than thus far assumed. This was made very clear during a n u m b e r of scientific meetings, e.g. in the proceedings of the IUPAC workshop 'Assessment of uncertainties in the projected concentrations of carbon dioxide in the atmosphere' (Slanina et al., 1991). In these proceedings it is stated t h a t the uncertainties in the most important sinks of CO2 are very large. The estimates of uptake by the oceans vary from 1-2.5 Pg C per year, with a value of b e t w e e n 1-2 Pg as the most probable range. The increase of CO2 in the atmosphere will induce enhanced growth of vegetation and part of the carbon, fixed in this m a n n e r , will be present in the form of enlarged root systems. A certain fraction of the extra carbon, present in roots, will remain in the soil after the decay of the vegetation and be stabilised for periods between 50 and 500 years. Estimates of the value of this so-called 'fertilization' flux range from i to 3 Pg C y-1
461 (Table 1.2) (Goudriaan, 1989 and Tans et al., 1990). These uncertainties have a decisive influence on future environmental policy decisions (Slanina et al., 1991). If a large fertilization effect exists, abatement measures are feasible. If uptake by the oceans and fertilization effect are in the range of the lower estimates, stabilisation of the atmospheric CO2 concentration is very difficult. If we assume a n n u a l fluxes of 6 Pg C, 1.5 Pg C, 2 Pg C, and of 2.5 Pg C, for respectively fossil fuel, landuse changes, oceanic uptake, and terrestrial fertilization, the difference of 3 Pg C between sinks and sources accounts for the accumulation of CO2 in the atmosphere. If we are able to stop the deforestation and induce increase of forests (corresponding to -0.2 Pg C y-l), the difference between sinks and sources would be in the order of 1.3 Pg C, corresponding to approximately 25% reduction of the emissions by fossil fuels. Measures directed to stop deforestation and to optimise agricultural production are probably cheaper than reductions in the order of 60% or more of emissions by fossil fuel. The relatively low agricultural productivity per unit land area of the countries containing the large tropical forests (0.1 to 0.2 of the potential production per unit of land area, to be compared with 0.6 to 1.1 for E u r o p e a n countries) leave room for such a policy. A combination of different measures, reduction of the use of fossil fuel, reforestation, and optimisation of a g r i c u l t u r a l activities could be effective in this case and will leave room for extension of emissions by the developing countries.In the case t h a t no fertilization effect exists and the uptake by oceans is only moderate, a very different picture emerges. A reduction of at least 60% of the emissions of fossil fuel could be necessary to stabilise the present CO2 concentrations in the atmosphere. Any increase in the emissions of the third world countries would ask an even more s t r i n g e n t emission a b a t e m e n t in order to prevent a f u r t h e r increase of the a t m o s p h e r i c CO2 concentration. In the worst case even the most s t r i n g e n t emission reductions in the industrial countries could be insufficient to counteract increasing emissions in the developing countries. One would be tempted to invest in adaptation strategies r a t h e r than in abatement of emissions if this latter situation proves to be true. It is clear t h a t the uncertainties in the sinks of CO2 will enormously influence future political developments and t h a t the reduction of these uncertainties is of prime importance. The conclusions of the IUPAC workshop made very clear that predictions of future CO2 concentrations had far larger uncertainties as was assumed until then, and t h a t additional research on the role of terrestrial systems on the CO2 budget was urgently needed.
462 Table 1.2 Sources and sinks of C O 2 (IPCC, 1990) Flux (Tg y-l)
Source Fossil fuel combustion Deforestation/landuse
5.4 + 0.5 1.6 + 1.0 n
Sink U p t a k e by oceans "Terrestrial fertilization"
2.0 + 0.8 1.6 + 1.4
Atmospheric increase
3.4
+
0.2
Sources and sinks of CH4 The same situation as for C O 2 existed, in essence, regarding the sources and sinks of CH4. In Table 1.3 the estimates are given of the source strength for the most i m p o r t a n t m e t h a n e emissions, as presented by IPCC in 1990 (IPCC, 1990). It was felt t h a t these estimates were, of course with a degree of uncertainty, fairly well established. The fact t h a t the emissions of methane were nearly equal to the sum of m e t h a n e oxidized in photo-chemical reactions and the a m o u n t tied in with increasing atmospheric concentrations, was regarded as an objective proof for this assessment. The conclusions of the IUPAC workshop on uncertainties of in the projected concentrations of m e t h a n e in the atmosphere (Slanina et al., 1994) among other scientific meetings, were very different: the latest reports (Slanina et al., 1994) about the emissions of N o r t h e r n Wetlands, indicated t h a t these emissions could be substantially lower as formerly assumed. The emissions of Northern wetlands were extrapolated to be in the order of 20 Tg y-1 instead of 80 Tg y-1 ; an unexpected conclusion of the workshop was that the possibility exists t h a t India, one of the main rice growing countries, contributes no more t h a n 7% of global emissions from rice crops, because most Indian paddy crop is t a k e n from irrigated field and only a small portion from water-logged fields (Slanina et al., 1994); the u n c e r t a i n t y in atmospheric oxidation is so large (in the order of 40% according the estimates of Calvert, 1994) t h a t this m e c h a n i s m cannot be used to check our emission inventories; the emissions of landfills could be much larger as previously assumed, based on the first results of m e a s u r e m e n t s in Canada, China and other countries (Slanina et al., 1994). In these proceedings it is very clearly concluded t h a t the extrapolations of very local m e a s u r e m e n t s to emission fluxes on regional, continental and global scale is probably one of the m a i n sources of errors and uncertainties. The specific recommendations is made to develop methods and m e a s u r e m e n t strategies for specific v a l i d a t i o n m e a s u r e m e n t s to e v a l u a t e emissions on regional and
463 (sub)continental scale. These validation m e a s u r e m e n t s are required to check whether the measurements of methane emissions carried out on very small scale, have been extrapolated correctly in the past to global dimensions. This assessment of the sources and sinks of methane makes very clear t h a t the state of knowledge at that time did not provide a suitable scientific fundament for predictions of future development and hence for abatement policies. This a s s e s s m e n t already mentions that the trends are changing (Khalil et al., 1994). For an extended period a exponential growth, in the order of about 1% per year has been observed. This trend has abruptly changed and is much less t h a n formerly observed (Steele et al., 1992). Rigorous explanations were not offered, a clear proof of the lack of knowledge in this area. Table 1.3 Sources and sinks of CH4 (IPCC, 1990) Flux (Tg y-l)
Range (Tg y-l)
Source Natural Wetlands Rice paddies Enteric fermentation Gas drilling, venting, transmission Biomass burning Termites Landfills Coal mining Ocean Freshwaters CH4 hydrates destablilization
115 110 80 45 40 40 40 35 10 5 5
(100-200) (25-170) (65-100) (25-50) (20-80) ( 10-100) (20-70) (19-50) (5-20) (1-25) (0-100)
30 500
(14-45) (400-600)
44
(40-48)
Sink Removal by soils Reaction with OH in the atmosphere
Atmospheric increase
S o u r c e s a n d s i n k s of N20 The assessment of the sources of nitrous oxide was at the time different from the other gases. It was assumed t h a t all estimates of sources were quite uncertain. Tropical forest soils were regarded as the single most important source of nitrous oxide to the atmosphere. N20 is also emitted by a large number of smaller sources, such as biomass burning, agricultural activities leading to nitrification and denitrification processes and specialised industrial processes (Table 1.4). The conclusion was that most of these sources were very difficult to evaluate. As a consequence the uncertainties in the emission estimates were estimated to be large. In Europe, the production of N20 by agricultural systems with high loads of
464 nitrogen, the emissions of electricity generation plants and the exhausts of cars equipped with catalysts were seen as major sources. The high e s t i m a t e s of emissions of electricity generation plants were caused by artifacts in sampling and analysis of flue gases, as was proven in the first stages of NRP (Spoelstra, 1992). Table 1.4 Sources and sinks of N20 (IPCC, 1990) Flux (Tg y-l)
Source Ocean, estuaries Fertilizer (including ground water) Soils (tropical forest) (temperate forest) Combustion Biomass burning
1.4- 2.6 0.01 - 2.2 2.2 - 3.7 0.7- 1.5 0.1 - 0.3 0.02 - 0.2
Sink Removal by soil Stratospheric loss
? 7 - 13
Atmospheric increase
3 -4.5
1.3 A s s e s s m e n t o f t h e d e v e l o p m e n t s in t h e k n o w l e d g e o n s o u r c e s a n d s i n k s of g r e e n h o u s e s gases d u r i n g the course of N R P p h a s e I During the last 5 years, the period of NRP phase I, the scientific community has accepted t h a t the uncertainties in sources and sinks of greenhouse gases are much larger as assumed in 1989. The warnings, exemplified by the proceedings of the IUPAC workshops on the uncertainties of in the projected concentrations of carbon dioxide and methane in the atmosphere (Slanina et al., 1991 and 1994) that these uncertainties are a major in future predictions, have now widely be accepted. That a large uncertainties exist has been made clear by the fact t h a t the trends in the concentration of greenhouse gases in the atmosphere have changed drastically and t h a t no reasonable explanation can be provided for these changes in the trends. So the present state of affairs can be summed up as follows: The fair amount of research on sources and sinks of greenhouse gases, carried out internationally during this period of 5 years, has led to the conclusion that the u n c e r t a i n t y in these sources and sinks is much larger t h a n formerly assumed. This may seem a r a t h e r negative conclusion, but it must be born in mind t h a t a proper evaluation of the state of affairs is essential to formulate effective research in the future to remedy this problem. The long-term average growth rate of atmospheric CO2 concentration has increased since the s t a r t of the m e a s u r e m e n t s at M a u n a Loa. This rate was about 0.8 ppmv, 1.3 ppmv, and 1.6 ppmv for respectively the 1960s, the 1970s, and the 1980s. Systematic higher CO2 concentration growth rates have been observed during the years 1988-90, which exceeded the level of 2.0
465 ppmv y-l, while in the subsequent years (1991, 1992, 1993) very low growth rates have been observed, in the order of 0.6 ppmv y-1. Indications exist, based on the most recent data, that the trend is returning towards long-term g r o w t h rates. It m u s t be kept in mind t h a t the a b r u p t decrease in atmospheric CO2 growth rate in the period 1991-1993 exceeds any previous variation in the existing time series of atmospheric CO2 concentration.
co~ The research in this area has expanded, leading to an increase in knowledge of the oceanic and t e r r e s t r i a l sinks of the carbon dioxide. New insights have been obtained on the problem of "unidentified" terrestrial sink (indicated by Tans et al., 1990 and the IUPAC report Slanina et al., 1994, as "fertilization effect") to specific processes. The uptake of CO2 by terrestrial systems is governed, most likely by two important processes: 1) Changes in land use. The present estimate is t h a t the net emissions by changes in land use total 1.1 Pg with an uncertainty of 1.2 Pg. This net emission flux consists of the sum of emissions by tropical sources (1.6 Pg) minus a mid-latitude uptake due to forestation, etc. of about 0.5 Pg, according to recent estimates. 2) The existence of the CO 2 fertilization is increasingly accepted as an important sink. But there is increasing evidence that different interactions play a role. The direct CO2-enhanced plant growth could provide a sink with a strength of 0.5-2.0 Pg C y-1. E n h a n c e d supply of nutrients, e.g. by t r a n s p o r t and deposition of sulphur and nitrogen compounds leads to the so-called nitrogen fertilization, which could contribute to an uptake of 0.2-1.0 Pg C y-1. As a result, the future effects of CO2 are difficult to predict. Climatic change could have, on a global scale lead to a net uptake, equivalent to 0-1.0 Pg C y-1. The resulting picture is that indeed the knowledge of these processes has been increased. In 1989 sinks such as the "fertilization effect" were still hotly debated. But the uncertainty is still very large and remains in boundaries as indicated by the IUPAC report. This is the most i m p o r t a n t reason, why no consistent explanations can be offered for the variations in the yearly trends, observed in the last decade. The debate between scientists, who contribute most of the terrestrial sink to changes in land use, and those who claim that fertilization effects are the main cause, is still raging, see articles of Tans et al., (1990) and Smith et al., (1992). It is very clear that no reliable predictions of future CO2 concentrations can be made and t h a t the development of optimal strategies for a b a t e m e n t is severely hindered, until these questions are resolved.
CIt4 Essentially the same situation exists for CH4 as described for CO2. The changes in the trends for CH4 have in fact been more pronounced as observed for CO2. Results from different networks indicate that the globally averaged growth rates for m e t h a n e have declined from approximately 20 ppbv y-1 in the period 1979-1980 to 13 ppbv y-1 in 1983, to 10 ppbv y-1 in 1990 and to about 5 ppbv y-1 in 1992 (Steele et al., 1992 and Khalil et al., 1993d en 1993e). The trend in southern hemisphere has halved and the increase in 1991-1992 in the northern hemisphere was close to zero (Dlugokencky et al., 1992). The cause of this change
466 in methane growth rates is unknown and still a matter of speculation. (Khalil et al., 1992c and Steele et al., 1992), (Dlugokencky et al., 1992 and (Dlugokencky et al.) A wide range of explanations are given: decreasing CH4 emissions from the former Soviet Union; lowering of lowered the t e m p e r a t u r e of the northern wetlands and thereby decreasing methane emissions; indirectly caused by emission of the Pinatubo Volcano; lowering of the w a t e r table increases the thickness of the layer over which m e t h a n e oxidation can take place, so northern wetlands appear to be more sensitive to changes in moisture than temperature; t e r m i n a t i o n of the one-to-one correlation between m e t h a n e emissions and growth of the global population, as result of lack of suitable areas for rice cultivation or cattle raising; increasing OH-radical concentrations, caused by increasing UV-B radiation, could have shortened the life time of CH4. This wide range of hypotheses demonstrates quite clearly the lack of information on the s t r e n g t h and the variability of sources of m e t h a n e and this situation is acknowledged widely within the scientific community.
N2,0 The development in views about the emissions of N20 have been slightly different compared to the other two gases mentioned. U n c e r t a i n t y exists about the atmospheric concentration of N20 in the pre-industrial period. Estimates range from 260 to about 290 ppbv, compared with a concentration of 310 ppbv in 1993. So, the estimates of the yearly trend show a considerable uncertainty (Khalil et al., 1992C) Recent results indicate that the trend of N20 has been smaller in the last years t h a n the average of the last two decades and decreased from about 0.8 ppbv to 0.6 ppbv. The general opinion in the scientific community is that N20 plays only a minor role in the changes of the radiative balance of the e a r t h and t h a t the increase in concentration has been much less spectacular compared to the other greenhouse gases. Questions are raised however on the impact of new industrial processes (large scale introduction of catalytic devices for cars and catalytic NOx reduction in industry), of changes in agricultural practice (large scale application of nitrogen fertilizers) or changes in the water tables of wetlands and agricultural areas on future N20 emissions. The developments of the last five years can be s u m m e d up in the following statements: the considerable a m o u n t of new information has m a d e clear t h a t the uncertainties in sources and sinks of greenhouse gases were much larger than assumed in the recent past; major causes for the uncertainties in sinks and sources of greenhouse gases have been identified. This increased knowledge will provide the necessary fundament for effective research in the future. -
467 This s u m m a r y of the recent results of could give the impression t h a t less risk for climatic change is present, compared with a few years ago, as the concentrations of greenhouse gases are increasing slower as expected. This would be a very d a n g e r o u s assumption. As the changes to lower t r e n d s in the a t m o s p h e r i c concentrations of greenhouse gases are not well understood it is impossible to indicate w h e t h e r this situation will last. To the contrary, it is very well possible t h a t the t r e n d s could change in u p w a r d direction very fast by the impact of changes in industrial and agricultural practice. A certain lack of knowledge regarding the contribution of different sources of greenhouse gases is not a problem in the first stages of abatement policies. A wide range of so-called no regret options, which will not only reduce the emission of g r e e n h o u s e gases but also contribute to a b a t e m e n t of other e n v i r o n m e n t a l problems, is available and environmental policies can been adapted accordingly. This state of knowledge, however, is not a good basis for developing a b a t e m e n t policies over longer periods. For this reason a better understanding of the sources and sinks of greenhouse gases must have a high priority on the scientific agenda.
2.
C A R B O N D I O X I D E (CO2)
2.1 Overview
o f t h e CO2 c l u s t e r
The considerations in chapter 1 led to the decision t h a t additional research on sources and sinks of CO2 should be directed to the role of terrestrial ecosystems in the CO2 cycle, a p a r t from the already on-going activities on the exchange of CO2 between the oceans and the atmosphere. Reports on the impact of the "fertilization flux" t h a t were published at t h a t time, h a d m a d e clear t h a t the exchange between t e r r e s t r i a l ecosystems and the atmosphere was very important and m u s t be known better in order to be able to model the CO2 cycle and to predict future CO2 concentrations in the atmosphere. It was proposed to study the exchange of CO2 between g r a s s l a n d s and the atmosphere for the following reasons: The Netherlands are to a large extend covered by grasslands. Pastures are a major component in European land use and the amount of grassland has been considerably extended on a global scale during recent decades; grasslands exhibit the same behaviour as forests as far as the fertilization effect is concerned: net primary production increases, allocation to roots as well as losses into soil, and potentially more C will be stored as organic m a t t e r with large residence time. F a r m l a n d s are amongst the most productive ecosystems (in terms of net photosynthesis). The soils of grasslands contain generally large amounts of carbon, and the carbon content of soils increases at higher concentration of CO2 in the atmosphere. less knowledge was available on the exchange of CO2 between grasslands and the atmosphere compared to forest ecosystems; expertise was available in The Netherlands. It was decided to develop a coherent program, dedicated to formulate and validate a improved model simulating the exchange of CO2 between grassland and the
468 atmosphere. This model encompass diurnal to seasonal fluxes, and the exchange of carbon between the soils of grasslands and the atmosphere. Better knowledge on the gross exchange of carbon between grasslands and the atmosphere is urgently needed to: assess the effects of changes in land use on the global carbon cycle; u n d e r s t a n d short time to yearly trends of CO2 concentrations in the atmosphere. (Analysis of these trends are essential tools to u n d e r s t a n d the CO2 cycle); assess the potential contribution of fertilization effect to sequestering of carbon in soils. -
-
In order to reach this goal the following activities were incorporated in a number of projects:
1)
Development of an improved model describing the exchange of CO2 between grassland and the atmosphere (LUW-TPE, project no. 852062). This model will provide a good description of diurnal and seasonal fluxes, and will include the exchange of carbon between the soils of grasslands and the atmosphere.
2)
Measurements of the exchange flux of CO 2 over the most important types of grasslands (soil) in The Netherlands (KEMA, project no. 853116; ECN, project no. 852076). It was conceived that the different kinds of soil of grasslands (clay and peat) would have a strong effect on these exchange fluxes. The results of these measurements would be used to parameterize better models and to validate them on a local scale (ECN, project no. 852076).
3)
I n v e s t i g a t i o n of the fertilization effect on g r a s s l a n d by m e a n s of pulse-labelling by 14C02 (IB-DLO -at present AB-DLO-, project no. 852063). Grass is exposed to 14C02 during short periods and the distribution of 14C between different parts of the vegetation and the soil is determined and fluxes of carbon are calculated.
4)
M e a s u r e m e n t s of the exchange flux of C02 over larger areas of grassland, using eddy-correlation, gradient m e a s u r e m e n t s at higher elevation (ECN, project no. 852076), and eddy-correlation m e a s u r e m e n t s from aircraft (KEMA, project no. 852065).
5)
M e a s u r e m e n t of changes in CO 2 concentration and isotopic composition (13C/12C and 14C/12C) at an altitude of 200 meters on a tower, with the objective to obtain regional validation of sources and sinks of CO2 with emphasis on the role of terrestrial systems (ECN, project no. 8520786). The isotope ratios are dependent on sources and exchange processes.
An co-ordination group monitored the progress of the different projects and facilitated that coherent results could be obtained. The sub-theme CO2 encompasses two additional projects. One was a desk study to assess the role of (Dutch) forests as apart of the carbon cycle (IBN-DLO, project no. 852071), while the other project was on the development of a geographically
469 explicit dynamic carbon cycle model t h a t will be incorporated in more complex, integrated models as IMAGE 2.0 (RIVM, project no. 852067).
2.2 Methodology Measuring CO2 concentrations with sufficient precision and accuracy does not present severe problems in the present state of methodology development. Several monitors, based on IR absorption methods, are commercially available. They can m e a s u r e atmospheric concentration with an accuracy of b e t t e r t h a n 1 ppmv. Accuracy is d e p e n d e n t on the quality of calibration and quality a s s u r a n c e s t a n d a r d s , but generally an accuracy of 1 ppmv or better is attainable without major problems. The situation is different when flux measurements must be applied in the field to study in detail the exchange fluxes of CO2 between grasslands and the atmosphere. Three available methods to measure gas exchange between the atmosphere and biosphere have been used in the described projects, i) enclosure, ii) eddy-correlation, and iii) gradient.
Enclosure methods A box is placed over vegetation, water, or soil. Air is pumped through the enclosure and the difference in concentrations measured at inlet and outlet is used to assess deposition or emission rates. Enclosure methods suffer from two problems: one, the enclosure can alter the behaviour of vegetation or soil, and two, the deposition or e m i s s i o n m e a s u r e m e n t is e x t r e m e l y local. The a d v a n t a g e of e n c l o s u r e m e a s u r e m e n t s is t h a t the present state of i n s t r u m e n t a t i o n can be applied in nearly all studies. In m a n y cases enclosure methods have to be used as no other alternative is available. In view of the extreme local effects and as alternatives are available for CO2, it was decided to not to apply enclosure techniques for the s t a n d a r d CO2 exchange m e a s u r e m e n t s . Box m e a s u r e m e n t s are applied in this cluster to investigate the exchange of CO2 between atmosphere and grass with exclusion of soil respiration. In order to quantify the CO2 flux by grass (in contrast to the integrated CO2 flux: grass + soil) an enclosure system was developed which m e a s u r e s continuously the CO2 flux u n d e r conditions of overpressure. This overpressure prohibits the exchange of CO2 between the soil and the atmosphere. The m e a s u r e d CO 2 flUX is, under these conditions, related to net CO2 assimilation of a grass canopy u n d e r field conditions - one of the two components of the integrated CO2 flux. Moisture content, t e m p e r a t u r e and CO2 concentration of the circulating air are regulated to avoid the already mentioned artifacts caused by deviation in the box from local conditions.
Eddy-correlation Eddy-correlation m e a s u r e m e n t s are based on the covariance of fluctuations in a m b i e n t concentrations and vertical windspeed. T u r b u l e n t t r a n s p o r t in the a t m o s p h e r e t a k e s place by eddies. In eddy-correlation m e a s u r e m e n t s the difference in concentration of the investigated compound is measured with a time resolution of 1 to 10 hertz in air moving downward to the surface and moving u p w a r d from the surface. The upward moving air has been in contact with the surface and the concentration has altered due to exchange at the surface. As eddy correlation measures directly, it is very often the preferred method. The regarding speed and precision of the instrumentation for eddy correlation are very often so
470 extreme t h a t the method cannot be used for m a n y trace gases. However, the method can be applied for CO2, and was used in the CO2 cluster. Gradient measurements
Depletion or emissions of pollutants at the surface results in a g r a d i e n t in concentration. Air concentrations of compounds, temperature, and windspeed are m e a s u r e d at different heights over the surface. F r o m these g r a d i e n t s the turbulence of the atmosphere is derived and the fluxes can be calculated. The problem with the gradient method is t h a t a high precision is required of the m e a s u r e m e n t method, as the concentration gradients are often in the range of a few percent of the atmospheric concentration. The i n s t r u m e n t a t i o n for CO2 m e a s u r e m e n t s can fulfil these requirements and gradient m e a s u r e m e n t s have been applied in two projects in the CO2 cluster to measure fluxes. Micrometeorological methods (eddy-correlation and g r a d i e n t m e a s u r e m e n t ) enables to estimate fluxes over a certain area as function of the height of the measurements. If local exchanges are studied, gradient measurements are carried out at heights between 1 and 5 m. The integrated exchange flux over an area of some hundreds of square meters is characterized this way. To study the exchange over an area of a few hectares, the gradient is measured between 1 and 20 m. The precision of flux measurements are limited, typically a precision in the order of 20 % can be reached in most cases. The consequence is that direct m e a s u r e m e n t s of the fertilization flux is not possible. This can be illustrated easily if the total exchange flux of terrestrial ecosystems with the atmosphere, in the order of 100 Pg C y-1 is compared with a high estimate of the fertilization flux of 2 Pg C. The quality of flux measurements was validated in an intercomparison experiment, organized by ECN. The participants of the CO2 cluster ECN and KEMA, and also KNMI and TNO, the latter two institutes are engaged in flux m e a s u r e m e n t s at sea, took part in this one-week experiment in November 1993 at Cabauw, The N e t h e r l a n d s . Two methods were used: the eddy-correlation technique and the gradient technique. Unfortunately, the t e m p e r a t u r e was about 0 ~ and reduced the flux of CO2 considerably. Although the small magnitude of the fluxes makes comparison difficult, it seems t h a t the gradient method tends to result in larger fluxes compared to the eddy-correlation method. Moreover, this experiment made clear t h a t there was a considerable differences between the calibration standards t h a t were used. This implies t h a t direct comparison of absolute concentration values of different set-ups as at the Cabauw experiment can only be achieved by inter-calibration of the standards. Validation of exchange fluxes over larger regional areas (The N e t h e r l a n d s and surroundings) were investigated by two different methods, a dynamic method (aircraft measurements) and a static method (tower measurements). Eddy-correlation m e a s u r e m e n t s in aircraft were applied as a method to obtain integrated exchange fluxes over larger areas. Variations in CO2 concentration and isotopic composition were m e a s u r e d at an tower (200 m) at Cabauw, The Netherlands. Uptake and emission of CO2 over large areas changes not only the CO2 concentrations, but also the isotopic composition. Emissions of fossil fuel contains no 14C and the 13C-12C ratio is dependent on the sources of CO2. The
471 concentrations were measured with a non dispersive infrared spectrometer with an accuracy better t h a n 0.1%. Working standards were calibrated against so called NOAA station standards. An wet a n n u l a r rotating denuder, filled with a N a O H solution, was used to extract CO2 from the air quantitatively. In the laboratory the formed carbonate was isolated (using barium chloride), stored, and before analysis re-converted into CO2. The 13C/12C ratio and the 14C/12C ratio were determined at E C N and the U n i v e r s i t y of U t r e c h t respectively. Meteorological d a t a were provided by the Royal N e t h e r l a n d s Meteorological I n s t i t u t e (KNMI). An 2-dimensional 2-compartment mesoscale transport model was developed at ECN. D a t a on the spatial distribution as well as descriptions of CO2 exchange (both biogenic and anthropogenic) will be used as i n p u t - p a r a m e t e r s to model the observed CO2 concentration and carbon isotopes. To quantify the potential fertilization effect directly, gross a n n u a l carbon flows were estimated in grasslands with 14C pulse labelling. 14CO2 was supplied to grass plants as a single pulse (1-2 hours in a plastic bag covering the plants growing within a soil column) and subsequently the distribution within the plant and soil c o m p a r t m e n t s was m e a s u r e d after a 21-days period in which carbon allocation was completed. This labelling was repeated on 13 representative moments during the growing season. Moreover, the decomposition of shoots and roots and the remaining carbon in soil organic matter was estimated by adding uniformly labelled dead shoots to planted soils in the field and by leaving pulse-labelled plants in the field and m e a s u r i n g the dynamics of the r e m a i n i n g carbon over 18 m o n t h s following the addition and labelling of shoots and roots, respectively. The fate of carbon compounds t h a t are exuded from living roots within 21 days after being a s s i m i l a t e d was followed by adding 'model-rhizodeposits' and m e a s u r i n g the remaining carbon. 2.3 R e s u l t s
Overview o f the results As the cluster of projects started late (some were stated at the end of 1992, others in the middle of 1993) only initial results are available. The most important results are summarized below:
1)
A detailed model is developed to describe the exchange of CO 2. This new model provides information on exchange fluxes with a resolution of 30 m i n u t e s ( F i g u r e 2.1), and thus be used for mechanistic studies. The process to incorporate this model in IMAGE has been started.
2)
Exchange fluxes with a time resolution of I hour or better have been measured over meadows on clay, and peat soils. The results have been transferred to the modellers. The flux measurements are of good quality and provide a good basis for the parameterization.
3)
Gradient m e a s u r e m e n t of exchange fluxes on a scale of hectares have been carried out near Cabauw by measuring concentrations at altitudes between 1 and 10 m over clay/peat soil. A first comparison between model results and actual m e a s u r e m e n t s is given in Figure 2.1. The difference between observed and calculated values are most probably caused by oxidation of peat in the soil
472 regulated by the water table. The combination of better models and good flux measurements enables the study of these important phenomena.
4)
Eddy-correlation measurements of CO2 exchange using an aircraft have been tested. The resolution of the method is insufficient to be useful under the prevailing conditions in The Netherlands.
5)
The experiments with labelled 14C02 have been carried out and the distribution of carbon has been measured. The interpretation of the results is in progress.
6)
Regional validation of models, describing the exchange of CO2 between vegetation and atmosphere, by monitoring variations in CO2 concentrations and isotopic composition at an altitude of 200 m is not possible yet. Distribution of sources and probably very large homogeneous processes like u p tak e in oceans can be studied by this method, but a more detailed assessment will be difficult in view of the limitation in the present models.
Detailed results 1) A dynamic simulation model was developed to calculate the CO 2 flux related to net CO2 assimilation of a grass canopy. The existing carbon cycle model WCCM2 (Goudriaan, 1989) operates with annual time steps and does not consider the precise seasonal and diurnal pattern of CO2 exchange. This study has incorporated these cycles while retaining the final result of the net annual exchange rate. To this end an existing simulation model for crop growth (SUCROS) has been utilized as a basis, in combination with other models for carbon dynamics in the soil (CENTURY). The model first generates the diurnal cycle to obtain the net diurnal assimilation rate, and diurnal soil respiration. These diurnal rates follow a seasonal cycle and are integrated to generate a net annual uptake. The net annual uptake of the above ground vegetation is called the Net Primary Productivity. Factors such as green soil cover, progress in the growing season on basis of accumulated temperature, soil wetness, partitioning of assimilates between plant organs, root dynamic are considered. Respiration rate of plant and soil have been modelled on basis of temperature, biomass and growth rate. The model has the potential to drive a 3-D model for atmospheric CO2 content, first to generate a diurnal cycle in the vertical profile, second to obtain net CO2 exchange rates of a region on a seasonal basis.
A first comparison of the model calculations (grass component) with the i n t e g r a t e d CO2 flux m e a s u r e m e n t s (grass and soil organic m a t t e r components) at C a ba uw (The N e t h e r l a n d s ) , in c o m b i n a t i o n w i t h environmental conditions is presented in Figure 2.1, in which the measured and calculated (grass component - potential) CO2 flux for Cabauw is given (Period March 18 to 20, 1993). LAI2 and LAI4 represent different leaf area index used in the model. The results indicate larger emission fluxes for CO2 as calculated by the model. The difference between calculated and measured CO2 fluxes may well be attributed to oxidation of soil organic matter, as discussed in the section dealing with the integrated flux measurements performed in
473 Cabauw. The combination of modelling activities and experimental approach provides a very useful tool to develop mechanistic descriptions of exchange processes between soil and vegetation and the atmosphere. The development of a model for the calculation of the CO2 flux related to oxidation of soil organic matter was initiated. 0.4 I 0.2
9
y
9~
- urea9
:..:..
:ram "
.
m
. _ _ _ _ . _ I9149149149149149149 9 mn
0
,,,, ..........
~ -0.2
~,,,i, l l l l l l l ..... ~
-d ~
d o
c~ ,~ , d ' ~
mr-
-
g
................... d
cq d
c~ ,~
T,,,IIIIIIIIIIIIIIII',, d
~o ~ 9
,o
~0 n
--
-1,2
--
-1.4
--
~m" ~ 9
" _9
c~ d
~
,~
9
d
~o~ 9
o
0
9
cq ~
,d ~
r"-mr--
d ~
o~
~
9
o 9
===o
[]
9
9
9 009
9 =.,
9 nn
n
9 9 O
-1
9
9
~
-~ -0.6 - -
-0.8 - -
:
W"m,= ,, e e
....... , ....... ,,, ..... ,,~,,,, I l I',',',',', I l I I',',',',',,*,I l IIl III,,
N ~: ,d,,,~ ~
n o
0
9
O O
e
9
~176 o E -0.4 o~
9 e
~
o 0 9
9
9 m ~0
mm
9
O, No O9 O
0
: e9
[]
~176
[] "
9
9
9
."
9
daytime (GMT+ I) .
.
I " measured
Figure
2.1
.
.
.
n calculated (LAI 2)
.
I
* calculated (LAI 4) t I
Measured (integrated) and calculated (grasscomponent-potential) CO2 flux on the experimental site Cabauw (The Netherlands) form March 18 to March 20, 1993
2)
The project on the development of a geographically explicit dynamic carbon cycle model has been fully integrated within the IMAGE 2 projects and data, experience and personnel and results were exchanged and model developments were immediately implemented in the IMAGE 2 framework. A geographic explicit C cycle model has been developed based upon a series of global databases with topography, soil, climate and land cover characteristics. The model identifies globally different land cover types, each of which is divided into its appropriate compartments for C storage and dynamics. The model is driven by Net Primary Production, that is a function of local climate, soil and land use. Several feedback processes are implemented in a mechanistic manner. Sensitivity analysis was carried out and some specific applications to analyze the importance of different feedback processes and the influence of a transient dynamics vegetation response. Large regional differences were obtained for different feedback processes. For example, CO2 fertilization was the dominant feedback in tropical regions, while the temperature response on growth and respiration became dominant in boreal regions. Although changes through feedbacks processes are important determinant of C cycle properties,
474 changes in land use will probably more dominant in the near future. The model is appropriate to also assess the impact of land use change on the C cycle. C o m p o n e n t s of the project have now been reviewed twice d u r i n g the international IMAGE review meetings and adjustments in the approach have been added after recommendations of the review committee. Together with the P o t s d a m I n s t i t u t e for Climate I m p a c t Research an improved version of the IIASA Climate database has been developed. This d a t a b a s e (CLIMATE: Cramer, L e e m a n s I n t e r p o l a t e d Meteorology for Applications in Terrestrial Ecology) now forms the basis for several global modelling efforts (e.g. IGBP-GAIM). The structure of the modelling approach and its main databases used within IMAGE 2.0 Terrestrial Environment Subsystem is accepted by IGBP-GCTE as a valid contribution to their core-project research.
3)
Local scale exchange fluxes with a time resolution of I hour or better have been m e a s u r e d over grasslands on clay and peat soils using both the eddy correlation method and the gradient method. Eddy-correlation was performed at Zegveld (grassland over peat soil), gradient m e a s u r e m e n t s at Lelystad (grass on clay) and at Cabauw. The latter m e a s u r e m e n t s were carried out between 1 and 10 m, so integrated fluxes were measured over several km2 of surface. The soil at Cabauw (peat covered with a layer of typically 20 cm of clay) is not as well characterized as is the case in Zegveld and Lelystad. The sites at Zegveld and Lelystad are p a r t of experimental farms, so soil characteristics and all agricultural treatments are very well documented. In contrast, the meadows around Cabauw are commercially farmed, so little information is available about any application of fertilizer, grazing, pasture, etc. The basic idea was to use the d a t a from Zegveld and Lelystad to p a r a m e t e r i z e the models and carry out validation by m e a n s of the flux m e a s u r e m e n t s at Cabauw over an integrated area. As all the infrastructure at C a b a u w was a l r e a d y in condition, it was possible to e v a l u a t e the m e a s u r e m e n t s of fluxes, convert the results in a format which is suitable for the modellers and transfer them in a very short period. A nearly complete data set of fluxes for Cabauw with a time resolution of 30 minutes for the period March 1993 to March 1994 has been transferred to the modellers and is used for p a r a m e t r i s a t i o n and validation (Figure 2.2). Examination of the results indicate t h a t the fluxes are measured with, for this kind of measurements, a very good precision, in the order of 10% relative. Evaluation and validation of the raw data at Zegveld and Lelystad has not been completed yet, as time was needed to e s t a b l i s h the n e c e s s a r y i n s t r u m e n t a t i o n , to carry out the actual m e a s u r e m e n t s , and to perform intercomparison m e a s u r e m e n t s in order to compare the results form both locations etc. Results, as far as available, have been t r a n s f e r r e d to the modellers. The evaluation of the total data set of both sites is n e a r l y completed, so they will be transferred to the modellers in the near future. The flux m e a s u r e m e n t s performed at the Cabauw site are of such a quality t h a t they indeed provide a good basis for the intended parameterisation and validation. The comparisons of models and experimental results, as reported later in this section, is therefor based on the results of Cabauw only.
475 Normally flux exchange measurements are carried out during relatively short periods. The only m e a s u r e m e n t s performed over longer periods were m e a s u r e m e n t s of dry deposition of NH3 and SO2, carried out by ECN and RIVM over forests and grassland. These m e a s u r e m e n t s were based on gradient methods. The experience at Zegveld, where eddy-correlation was applied, has led to the conclusion t h a t gradient methods are much more suitable for m e a s u r i n g over long periods t h a n eddy-correlation based methodology.
4)
CO2 fluxes on km2 scale have been estimated by measuring concentration gradients up till 10 m altitude at Cabauw, The Netherlands. The results are available to validate the exchange models. The measurement of the exchange of CO2 between grass-on-peat and the atmosphere showed that fluxes caused by photosynthesis as well as soil respiration due to oxidation of peat can be assessed (Figure 2.2). A net uptake only takes place in March, April and May, while during all other months a net emission occurred. The total net emission for the period March 1993 to March 1994 was calculated to be about 3000 kg C ha-1. Two sources contributing to the soil respiration are oxidation of peat layers and animal waste. The CO2 emission due to animal waste was estimated at approximately 600 kg C ha-1 y-1 (about 20%) of the net estimated emission. This leaves 2400 kg C ha-1 y-1 for the oxidation of soil organic m a t t e r (e.g. peat) and is of the same order as the potential emission estimated for the oxidation of "shallow-drained" peat soil (about 2300 kg C ha-1 y-l, Wolff). This is approximately 2.5% and 4% of the total anthropogenic CO2 emission in the N e t h e r l a n d s respectively with and without emission from animal waste. Because the peat layer at Cabauw is covered with clay, it is expected t h a t emissions for grass-on-peat will be larger due to the larger amount of peat that can be oxidized. This type of m e a s u r e m e n t , in combination with models describing the exchange of CO2 between the atmosphere and grass and soils, is in principle capable to provide the necessary information on uptake or loss of carbon as a function of changes in land use.
476 Emission of CO~ at Cabauw March'93 to February'94 200
1200
150
I900
_~ 100
600
~UD .~
O
~
~
50
3oo
~ o~
O "" ~
~
0
0
-5o
-300
< -100 Z -150 -200
=
E
-6oo
........... i
1 i
J
E ~
[m Cumulative Emission [ [] Monthly Net Diurnal Emission
-900 -1200
Figure 2.2 Emissions of CO2 at Cabauw (The Netherlands) from March 1993 to February 1994
5)
The instrumentation has been developed and tested to m e a s u r e C O 2 exchange fluxes by eddy correlation m e a s u r e m e n t s using aircraft. Airborne flux measurements provide information on fluxes over very large areas, depending on height and on large scale geographical distribution of these fluxes. Two measuring flights were performed to test an airborne eddy correlation system developed in the Netherlands. The results of these tests show that the accuracy of the system is insufficient to be useful to support model development under the prevailing conditions in the Netherlands. Fluxes are measured with a precision of 30 to 50% relative and parametrisation and validation of models require data of a better quality. The generally very inhomogeneous landscape of The Netherlands is a severe handicap for aircraft measurements, even if they are carried out at a minimum altitude to reduce the surface area which is observed. It was concluded that the airborne flux measurements were not feasible for application in the Netherlands taking a number of considerations into account: inaccuracy of the measurements, the heterogeneous character of the Dutch landscape and the relatively high costs of this type of measurements.
6)
The experiments with labelled 1 4 C 0 2 have been carried out and the distribution of carbon has been measured. The interpretation of the results is in progress. Preliminary conclusion is that in peat soiis decomposition (mg C added per mg soil C) is twice as fast as in sand and clay, and results in 5-10 times higher incorporation of carbon in biomass and microbial products. Structural organic carbon was retained more in clay (80%) than in sand (58%) whereas the extract in clay was decomposed as fast as in sand. The latter is
477 surprising given the generally higher retention capacity of clay soils than of sandy soils. Care should be taken since these data concern similar additions to all soils wh er eas the carbon input (total C, s t r u c t u r a l and soluble rhizodeposits) to soil might differ between the soils studied.
7)
Regional validation of models, describing the exchange of CO 2 between vegetation and atmosphere, by monitoring variations in CO2 concentrations and isotopic composition at an altitude of 200 m has not been possible yet, although a mesoscale transport model was developed. Also "mesoscale" m e a s u r e m e n t of CO2 concentration were performed and carbon isotopes ratios were determined. Main problem in this project was the availability of detailed databases to be used as input for the model. Also good model descriptions of CO2 exchange were lacking. Both problems will be (partly) solved in the near future. Trajectory analysis show that in periods with enhanced C 0 2 concentrations air was transported over the continent i.e. industrialized areas, especially over G e r m a n y (South East). These periods mostly occurred in winter. The concentrations of CO2, and both the carbon isotopes were also strongly correlated during these periods indication that the main source for CO2 was combustion of fossil fuel. By use of the carbon isotopes the relative contribution of the anthropogenic sources can be estimated. An example of such an estimate is given for December 24, 1992. The concentration and 14C changed from 380 to 425 ppmv, and from 113 to 107 pmC (percent modern carbon) respectively. The air was coming from the north of Germany during this period. The anthropogenic contribution of approximately 50 % was calculated using the increase in concentration and the decrease in 14C. It is assumed here that the CO2 exchange by vegetation in this winter period was negligible, and consequently, soil respiration and litter decomposition are the only biogenic sources for CO2. These sources were assumed to contribute for the other 50% in the change in concentration. The 13C of the emitted CO 2 (anthropogenic and biogenic) was calculated on -24 promilles. The absolute emission of anthropogenic CO2 was estimated using a emission inventory and the trajectory of that particular day. An anthropogenic flux contributing to the change in atmospheric CO2 was estimated to be 9 g CO2 m2 d-1. Using the calculated ratio between anthropogenic and biogenic sources (50%/50%) a biogenic flux of 9 gram CO2 m-2 d-1 is calculated. As comparison the estimated flux of the grasslands in the surroundings of the site Cabauw is 2 g CO2 m-2 d-1. This value for the flux is calculated for the November 1992, when the assimilatory and respiratory processes of the vegetation were still active. (In the period November 1993 to December 1993, estimated emissions by grasslands range from 1 to 10 gram CO2 m-2 d-l). A better estimate of the anthropogenic emission of CO2 will be obtained in the near future with better description of spatial distributions of sources combined with the transport model.
478 8)
The present stock of carbon in living biomass, litter and stable humus and the annual accumulation of carbon in stems for fifteen forest types has been quantified from inventory data on growth and standing volume, and forest soil information in combination with literature data on forest biomass. The forrest area in The Netherlands is about 330000 ha, mainly young plantations of conifers. The present standing volume is 170 m3 ha-1 and the average volume increment was 9.0 m3 ha-1 y-1 over the period 1984 to 1989. At present approximately 63.7 Tg C is stored in the entire forest, including dead organic matter in the forrest soil. About 60% of the carbon is stored in the humus of the soil compartment. The average carbon stock in the stable h u m u s is approximately 110 Mg C ha-l, whereas only 60 Mg C ha-1 and 20 Mg C ha-1 is contained in respectively the living biomass and the litter layer. About 0.66 Tg C of atmospheric carbon is stored annually (by means of stem volume increment). About 50% of the annual storage is harvest each year. This implies that the Dutch forests act as a sink with a strength of approximately 0.33 Tg C y-1. The nett accumulation for the whole forest area amounts at present about 1 Mg C ha-1 y-1. The current sink acting of the Dutch forest can most likely be explained by the fact that the forests are young and still in building phase. However, this sink is not always as strong as reported here. The latest forest inventory reported an average annual volume increment of 7.8 m3 ha-1 y-1. The net storage rate as reported here, decreases correspondingly. The presented results therefore, depend very much on year tot year variation in growth of forest caused by climatic variability. The net annual sequestration probably varies in between 0.2 and 0.4 Tg C y-1. According to the investigators long rotation with species as oak, beech, and Douglas-fir are most suitable for long-term storage.
2.4 F u t u r e r e s e a r c h The uncertainties in the estimates of the so-called fertilization flux is still very high. Another problem is the uptake and loss of carbon by changes in land use. The m e a s u r e m e n t s of NRP I have indicated t h a t a substantial amount of CO2 (equivalent to 3000 kg C ha-1 y-i) is emitted by peaty soils induced by lowering of the water table. It is to be expected that the reverse process, uptake of carbon by peat formation will take place if the water level in peaty meadows is at lower depth than present. The plans to restore the former conditions in many areas in The Netherlands of very high water tables have clear consequences regarding CO2 emissions. Emissions due to peat oxidation will be stopped and an enhanced uptake of CO2 will take place not only due to the fertilization effect but also due to peat formation. This situation indicates that is very important to improve our knowledge regarding the exchange of CO2 between the atmosphere and ecosystems which are able to sequester large amount of carbon in their soils. Intensified research on the exchange of CO2 between grass lands and the atmosphere, including fertilization effect, peat formation and peat destruction has a high priority in this respect.
479
3.
M E T H A N E (CH4)
3.1 Preparation studies and organization The programming of the CH4 cluster has been based on three preparatory studies. One study presented estimates of the CH4 emissions and their uncertainty ranges for The Netherlands based on literature (Born et al., 1991). The other two studies (Leffelaar et al., 1991 and Diederen, 1992) discuss priorities and criteria for the CH4 research in The Netherlands.
Inventory of Dutch CH4 emissions A first a t t e m p t to quantify Dutch CH4 emissions and to estimate ranges of uncertainties was made by Van den Born et al. (1991). The major conclusions of this work are p r e s e n t e d in Table 3.1. This inventory indicated t h a t enteric fermentation, landfills, the oil and gas industry, and organic soils cover about 90% of the total national m e t h a n e emissions. On a global scale these sources are relatively less important, covering about 47% of the global m e t h a n e emissions. Sources like rice paddies, biomass burning and coal mining are i m p o r t a n t on a global scale, but are absent or minor sources in the Netherlands. The large uncertainties in emission estimates are large, both on a national and global scale.
Criteria and priorities Starting point of the research program on methane within the NRP was to achieve a significant reduction of uncertainties in knowledge on important emission sources (Leffelaar et al., 1991 and Diederen, 1992). Important criteria developed for the planning of the programme were the relative importance of the sources, the ranges of uncertainty and the availability of specific expertise on emission sources. The source s t r e n g t h of m e t h a n e from enteric fermentation in The N e t h e r l a n d s is relatively well known. It was therefore decided not to plan any research activities on this item, although it is the largest source on a national scale. The uncertainty of this source in developing countries, however, is large, but the specific expertise on the differing diet situation and the impact on physiology was too poor to plan a research project. Rice paddies are of no importance on a national scale, but contribute significantly on a global scale. As the appropriate expertise on this subject was available, it was decided to formulate a research project on this topic. Furthermore, it was decided to plan research on the emissions from landfills, the oil and gas industry, and organic soils. Also, a n u m b e r of research projects were formulated to validate local emission m e a s u r e m e n t s , and to extrapolate the information to a larger spatial scale.
480
Table 3.1 Relative contribution (%) of national and global sources tot CH4 emissions in The Netherlands in 1989/1990 and the world (Van den Bom et al., 1991) Source Animals - enteric fermentation Landfills Oil & gas industry/distribution Wetlands/organic soils Ocean/coastal waters Freshwater Animal waste Waste Water t r e a t m e n t Rice paddies Termites Biomass burning Coal mining Other Emission range (weight units)
Netherlands
Globe
40 27 16 7 4 2 2 0.3 NE NE 2
13 7 8 19 2 1 6 4 18 7 9 6 1
710-1230 Gg CH4 y-1
310-990 Tg CH4 y-1
NE: Not Estimated, not zero; -: Not applicable
Organization of the CH4 cluster The overview of the coherence between the research projects within the cluster is presented in Figure 3.1. Three relative important national sources were studied: the oil and gas industry, the landfills and the organic grassland soils. The organic grassland-soil related projects aimed at understanding the processes of m e t h a n e formation and consumption in the organic soils. The results of experiments and m e a s u r e m e n t s are integrated in a model of the methane flux from the soil to the atmosphere. The following projects are part of the CH4 cluster: B. Biogenic sources
BRP. BMF. BMC. BGM. BMMF.
Soil parameters controlling methane production and emission from rice paddies (LUW; project no. 850009) Methane formation by anaerobic consortia in organic grassland soil (LUW; project no. 853120) Methane consumption by indigenous grassland microflora. (LUW; project no. 853122) Effects of grassland m a n a g e m e n t on the emission of m e t h a n e from grassland on peat soils (LUW; project no. 853121) Modelling methane fluxes from and to grass covered peat soils (LUW; project no. 853123)
481 A. Anthropogenic sources ALl.
AL2. AOG1. AOG2.
Greenhouse gases from landfills in The Netherlands. (TNO-ME; project no. 850023) Landfill gas formation, emission and recovery in The N e t h e r l a n d s (TNO-ME; project no. 853105) Q u a n t i f i c a t i o n of CH4 emissions due to n a t u r a l gas losses and petroleum production (TNO-ME; project no. 850008) Quantification of methane emissions in the exploration and production of natural gas and petroleum in The Netherlands. (TNO-ME; project no. 853104)
V. Evaluation and validation
VCI. VEV. VUA.
Validation of source strengths of atmospheric CH4 using carbon isotope ratios (ECN; project no. 852097)) Evaluation and validation of the CH4 emissions in The Netherlands and contributions from various sources ( TNO-MW; project no. 853124) Methane emission of the Amsterdam urban area (LUW; project no. 853125)
VALIDATION
OIL & GAS INDUSTRY
LANDFILLS
ORGANIC SOILS INTEGRATION FLUX MODELLING F O R M A T I O N
C O N S U M P T I O N
M A N A G E M E N T
Figure 3.1 Schematic overview of the research of the projects within the m e t h a n e cluster of the NRP in The Netherlands 3.2 M e t h o d s Biogenic sources
Rice paddies (BRP). In this project the impact of various soil related parameters on the CH4 emission from wetland rice fields was studied. Methane fluxes from wetland rice fields in the Philippines were monitored with a closed chamber
482 technique as described by Schfitz et al. (1989) during two wet seasons (1991 and 1992) and one dry season (1992). The effects of soil-sulphate, soil-salinity, and organic m a n u r e on CH4 emission were studied in experiments where gypsum, salt and green m a n u r e were added respectively. The effect of a calcareous soil was studied by a comparison with a non-calcareous soil. Methane oxidation in the rhizosphere was studied using a specific inhibitor of m e t h a n e oxidising bacteria. The research was done in close co-operation with a project of the I n t e r n a t i o n a l Rice Research Institute (IRRI) in the Philippines which aims at collecting base-line CH4 emission data from Asian rice fields.
The integrated CH4 grassland projects. The main aims of the CH4 research projects on grassland are the understanding and quantification of m e t h a n e formation and consumption in grassland on peat soils, and of the net fluxes of m e t h a n e between soil and atmosphere by experiments and simulation modelling. Four different scales are distinguished, i.e. (i) micro organisms in pure culture studies; (ii) batch experiments with homogenised soils; (iii) intact soil columns; (iv) field scale. The modelling aims to inter-relate the data obtained from the different scales. Grasslands cover more t h a n 35% of the total surface area in the Netherlands, of which 32% is on peat soils. The study sites are located in the major peat area of the w e s t e r n p a r t of the N e t h e r l a n d s , around Zegveld (52~ 4~ The p r e d o m i n a n t l y eutrophic peat originates from sedges, reeds and wood, and generally have a clayey top-layer. Maximum peat depth is about 6 m. The organic m a t t e r content ranges from about 40% in the top 10 cm to about 90% below a depth of 60 cm, generally. Soil pH ranges between 3.5 and 5.0. A lowered m e a n ground w a t e r level, fertilizer application and removal of the grass crop via grazing and mowing are the major measures t h a t take place on intensively m a n a g e d grassland. On extensively managed grassland the vegetation is cut once a year in summer. The studied sites include both intensively managed, drained grassland and extensively managed natural grasslands. On intensively managed grassland, located at Zegveld, two typical sites have been chosen, i.e. site '8B' with a m e a n ground w a t e r level of 30 cm and site 'Bos 6' with a mean ground water level of 60 cm. Next to the effect of ground w a t e r level, effects of fertilizer application and grazing versus mowing on net exchanges of CH4 between peat soil and atmosphere are investigated. On extensively managed grassland, three typical sites have been chosen in the Nieuwkoopse Plassen area with mean ground w a t e r levels of 5, 10 and 15 cm. Data on ground water level, soil and air temperatures, soil water filled pore spaces, soil nitrate contents and net CH4 fluxes have been monitored on a weekly basis from September 1993 onwards. M e a s u r e m e n t s will continue till about August 1995.
Methane formation grassland soils (BMF). Soil profiles were t a k e n from the two Zegveld grassland sites, with water tables of 30 cm and 60 cm below surface. The soil profiles were sectioned, t a k e n to the laboratory in sealed plastic bags, and stored at 4~ Inside an anaerobic glove box soil samples (20 g wet weight) from each section were transferred to 300 ml serum bottles, containing 40 ml of anoxic
483 distilled water. The stoppered bottles were incubated under a N2 atmosphere (50 kPa overpressure) at 15~ in the dark. The initial pH of the suspended soils ranged from approximately 4.8 to 5.5. At certain time intervals samples were taken from the head space as well as the liquid phase and analyzed for gases (CH4, CO2, H2) and fatty acids or alcohols, respectively.
Methane consumption grassland soils (BMC). Soil samples from different depths (0-5, 5-10, 10-20, 30-40 cm) were taken from Zegveld. To investigate the kinetics of methane oxidation of these different depths the soil was placed in bath cultures in 300 ml flasks with gas tight septa and incubated with 1, 10, 100 and 10,000 ppmv methane, respectively, in artificial air with 1% (v/v) CO2 (Bender et al., 1992). For the enrichment of methanotrophic bacteria with different affinities for methane, soil (100 g) was incubated in a system receiving a continuous gas-flow of 4 ml/min containing methane at 4 different concentrations (1, 10, 100 and 10,000 ppmv).
Grassland management (BGM). Net CH4 emissions from grassland on peat soils in The Netherlands have been monitored with vented closed flux chamber (Hutchinson et al., 1981) from September 1993 onwards. Monitoring will continue in 1994 and 1995. At Zegveld, intensively managed grassland on peat soil with a mean ground water level of 30 cm and intensively managed grassland on peat soil with a mean ground water level of 60 cm have been investigated. Also, on both Zegveld sites the effects of nitrogen fertilization and grazing versus mowing on net CH4 emissions have been investigated. Finally, CH4 fluxes from three extensively managed grasslands at Nieuwkoop have been measured as well. Modelling methane fluxes grassland soils (BMMF). This project started in September 1993 and will last for 4 years. It aims at developing a process model for methane fluxes to and from organic grassland soils. Water dynamics at Zegveld will be obtained to be used as input for a gas transport model for Zegveld. With this gas transport model oxygen dynamics and methane transport will be described. The oxygen dynamics will be used as input for the methane production model. Results of the integrated model of methane production (results from project BMF), consumption (results from project BMC) and transport will be compared with field fluxes measured in project BGM. This comparison in combination with a sensitivity analysis of the model will show which aspects need most attention in further research. The model for methane production has already been developed. It calculates the dynamics of biomass, acetate, and methane formation. Moreover, it describes experimental data from the BMF project quite well. In this model, besides oxygen, a time lag is incorporated before methane production can start. This lag period could, in a later stage, be specified in terms of the presence of electron acceptors like nitrate and sulphate. If we succeed in incorporating all major processes in a realistic way, it will be possible to test two hypothesis: in-situ methane emission is low, because production is limited by the short duration of the anaerobic periods during wet periods; during dry periods, methane uptake by the soil is controlled by m e t h a n e t r a n s p o r t from the atmosphere to the methanotrophs.
484 Anthropogenic sources
Landfills - emission measurements (ALl). This preparatory study compared two m e a s u r e m e n t methods at three landfills from J u n e until S e p t e m b e r 1991 (Verschut et al., 1991) a dynamic closed-chamber method Balfour et al., 1987, Reinhart et al., 1992) measuring concentration differences between air entering and leaving a closed chamber system (10 m2; 8-10 replicates per landfill; during >_24h) and a micrometeorological method (Fowler et al., 1989) measuring concentration gradients and wind velocity along a pole of 10 m height situated for about two weeks at the centre of a landfill, combined with vertical and horizontal flux calculations and wind speed profiles. Landfills- gas formation, emission and recovery (AL2). To improve the reliability of the emission quantification from landfills the methane material balance (Emission = Formation - Oxidation - Recovery; no accumulation assumed) is investigated. Landfill gas formation is determined in two ways: by emission m e a s u r e m e n t s at various landfill sites (micrometeorological method Oonk et al., 1995) and adding the amount of landfill gas recovered and oxidized to the amount emitted; by evaluating recovery efficiency from the results of recovery projects in relation with landfill geometry, composition of the top liner system and the lay-out of the recovery system - landfill gas formation is subsequently obtained as the product of the a m o u n t of landfill gas recovered and the recovery efficiency. Oxidation data used are from literature (UK-DoE, 1993, US-EPA, 1990, Oonk, 1993). The formation of landfill gas is subsequently modelled, by correlating the formation to waste composition, age and amount of waste landfilled. Natural gas losses and petroleum production (AOG1). A first estimate of CH4 emissions due to the production and treatment, the high pressure transport, the distribution and consumption of natural gas and the production and transportation of petroleum in The Netherlands was made by an engineering approach (Nielen, 1991). This estimate was based on available information of m e a s u r e m e n t s , emission factors and production and consumption data for 1989. Exploration and production of natural gas and petroleum (AOG2). Three types of m e t h a n e emissions related to oil and gas production were examined. The continuous emissions due to leakages of systems used and from off-gases of various gas t r e a t m e n t installations have been quantified by an engineering approach using: emission factors, material and energy balances; knowledge on maintenance and testing procedures; information from NOGEPA on amounts of associated gas produced; information on process equipment, compressors, turbines etc. The irregular emissions due to periodic tests and maintenance of installations were quantified partly as a result of an engineering approach as described above, partly by a combination of measurements and dispersion modelling. The third type of m e t h a n e emissions, the incidental emissions due to failures of devices also are estimated by the combination of m e a s u r e m e n t s and dispersion modelling. The a m b i e n t CH4 concentration at Kollumerwaard, located N o r t h - w e s t of the Groningen gas field, which has been registered permanently since July 1991, was screened on indications of events elevating the background concentration. This was done in combination with information on meteorology, potential source location and distribution calculations by the Plume model.
485 Valuation and validation
Validation by carbon isotope ratios (VCI). At two sites a t m o s p h e r i c CH4 concentration has been monitored continuously (Eisma et al., 1995 and Kieskamp et al., 1995): the 200 m meteorological tower at Cabauw and the 'Vuurtoren' island near Amsterdam. The latter to examine the emissions of the Amsterdam area (see VUA). Carbon isotopic analysis in atmospheric CH4 t a k e n at Cabauw has been performed as well. However, due to interference from near-by 14CH4 emissions of Pressurised Water Reactors (PWR) 14C could not be used as a tracer for fossil and biogenic CH4 in Europe. Instead, from the 14CH4 record at Cabauw, emission factors from PWRs have been determined. A laboratory intercomparison was organized for ambient CH4 m e a s u r e m e n t s exercise, with ECN, TNO, KEMA and LUW as participants. Results obtained for Cabauw are interpreted by analysis of air mass trajectories, meteorological d a t a and the application of a L a g r a n g i a n t r a n s p o r t model in c o m b i n a t i o n w i t h CH4 emission i n v e n t o r i e s and c o m p a r i s o n of isotopic m e a s u r e m e n t s with characteristic isotopic values of CH4 sources. Also, the CH4 concentration records m e a s u r e d simultaneously at Cabauw and D u r g e r d a m are compared. As a first approximation a GIS application has been developed in which the excess CH4 concentration at Cabauw (> 1.75 ppmv) is related to the area over which the air mass was t r a n s p o r t e d (using 36 h backwards air trajectories at Cabauw and assuming a constant, 1000 m mixing layer height). The total area covered by the air trajectory was calculated as a reverse plume. The CH4 emission flux required for explanation of the observed excess CH4 was s u b s e q u e n t l y assigned to all points in a 0.1 ~ x 0.1 ~ grid over Europe. Methane emission from w a t e r surfaces was a s s u m e d to be zero w i t h exception of the oil-and gas-production sites on the North Sea. Evaluation and validation from various sources (VEV). This project validates present knowledge of CH4 emissions by comparing measured concentrations with t h o s e c a l c u l a t e d by a t m o s p h e r i c dispersion models from e m i s s i o n a n d m e t e o r o l o g i c a l data. C o n t i n u o u s m e a s u r i n g is p e r f o r m e d by u s i n g gas c h r o m a t o g r a p h y on two monitoring sites, namely, Arnhem and K o l l u m e r w a a r d from 1990 and 1991 onwards, respectively. Dispersion modelling has not s t a r t e d yet. Urban area (VUA). The u r b a n m e t h a n e emissions have been quantified by modelling methane air concentrations and comparison of calculated and measured imission concentrations. M e a s u r e d concentrations were obtained from E C N (project VCI, E i s m a et al., 1995) from the m e t h a n e m o n i t o r i n g station at Vuurtoreneiland, about 2 km east of Amsterdam. Methane emissions estimates from the urban area were based on literature data Amstel et al., 1993 Veldt et al., 1993) road traffic, the natural gas distribution network, and industrial sources. The D a n i s h OML model was selected for calculation of immission concentrations (Lofstom et al., 1992). This Gaussian plume model has a preprocessor to calculate dispersion height, atmospheric stability and turbulent mixing from synoptic and balloon meteorological measurements. The synoptic m e a s u r e m e n t s were obtained from Schiphol Airport, the balloon data from De Bilt. Emission data and dispersion data were fed to the OML model to calculate immission concentrations for the
486 continuous methane monitoring station. The calculated emissions were compared with data collected at the monitoring site. To obtain an estimate of the increase in ambient methane concentrations due to the urban plume, methane concentrations of Cabauw (some 40 km south of Amsterdam) (data obtained from ECN) were subtracted from the Vuurtoreneiland data. 3.3 R e s u l t s Biogenic sources
Rice paddies (BRP) Soil-sulphate. The methane emission from plots amended with 6.66x103 kg ha-1 gypsum (CaSO4) was reduced by 55-70% compared to non-amended plots (Figure 3.2). The reduced CH4 emission upon gypsum application was most likely due to inhibition of methanogenesis by sulphate-reducing bacteria. Observed SO42concentrations in the soil solution of gypsum-amended plots were well above minimum concentrations reported in the literature for successful competition of sulphate-reducing bacteria with methanogens. The data indicate t h a t CH4 emissions from rice grown on high-sulphate containing soils or gypsum-amended soils is low compared to non or low-sulphate containing soils. However, fertilization of rice fields with (NH4)2SO4 will not necessarily result in lower CH4 emissions because the amounts of sulphate added are relatively low. Soil-salinity. Rice is often grown on saline soils. To investigate whether the presence of salinity results in lower CH4 emissions NaC1 salt was added to a rice field. Pore water EC increased to about 4 dS m-l, which caused a reduction by 25% only in CH4 emission. It was shown that the CH4 production in the salt-amended field was strongly reduced compared to the control field (Table 3.2). However, CH4 oxidation in the salt-amended plot was even more inhibited than CH4 production. This resulted in about equal net CH4 fluxes from both salt-amended plots and non-amended plots. The data illustrate the importance of knowledge of both CH4 production and CH4 oxidation when estimating CH4 emission and show that a reduction in CH4 production does not necessarily lead to reduced CH4 emissions.
487 Methane emission (mg.m-2.day -1 ) 5,000
/~
J G. Manure
G. Manure + Gypsum
4,000
3,000
2,000
1,000
0
2
3
4
5
6
7
8
Days after transplanting Methane emission (mg.m-2day-1) 1,200 1,000 800 600 400 200
0
I 20
i 40
i 60
I 80
100
Days after transplanting Figure 3.2 The impact of Gypsum application on methane emission from wet rice fields during the wet season in 1992 at The Philippines Calcareous soils. CH4 emissions from rice grown on a calcareous soil were higher t h a n from a non-calcareous soil. The seasonal p a t t e r n of CH4 emission differed, with a more pronounced emission peak early in the season, probably due to the favourable pH for CH4 production in the calcareous soil. The difference in emission between the two soil types was no longer observed when the fields were fertilized with green manure, indicating that the "soil"-factor may be overruled by the input of organic matter. Application of organic manure. Application of green m a n u r e s t i m u l a t e d CH4 emissions. CH4 emission was highest during the first half of the growing season in plots t h a t received more t h a n 11x103 kg ha-1 of green m a n u r e . Ebullition contributes significantly to total CH4 transport, if rice fields receive high inputs of organic matter. The impact of organic m a n u r e on CH4 emissions, at different locations of the world, can be described by a dose-response curve if CH4 emission from the organically amended plots is expressed relative to CH4 emission from mineral fertilizer t r e a t m e n t s . Such an approach may prove to be useful when e s t i m a t i n g CH4 emissions from larger regions if information on the amounts of organic m a n u r e used in the region becomes available.
488 Plant-mediated gas transport. Plant mediated CH4 transport was shown to be described by diffusion only. The results combined with data from the literature suggest that the rate limiting step in plant-mediated methane transport is diffusion of CH4 across the root/shoot junction. CH4_ oxidation in the rice rhizosphere. CH4 oxidation in the rice rhizosphere was studied using methyl fluoride, a specific inhibitor of methane oxidising bacteria. CH4 oxidation in the rice rhizosphere depended on the growth stage of the rice plant and becomes much less important when the rice plant reaches the ripening stage. Therefore seasonal patterns of CH4 emission in rice fields do not only depend on changes in CH4 production but also on changes in CH4 oxidation. These findings indicate that methanotrophs do not oxidise a constant percentage of the CH4 produced throughout the growing season. Table 3.2 Average methane flux from triplicate soil cores of rice fields with and without salt amendment, during anaerobic and aerobic incubation, and percentage CH4 oxidized Sampling date a) Flux)
CH4 flux (nmol cm-2 h-l) anaerobic no salt salt
76 DAT 11.02 2,75 96 DAT 24.53 7.04 110 DATb) 20.82 7.42 a) b)
CH4 oxidized (nmol cm-2 h-l)
CH4 oxidized (% of anaerobic
aerobic no salt
salt
no salt
salt
no salt
1.31 3.29 2.75
1.32 2.17 2.99
9.71 21.24 18.07
1.43 4.87 4.43
88 88 87
salt 52 70 60
DAT = days after transplanting 110 DAT =1 week after harvest
Methane formation grassland soils (BMF) Formation of CH4__Methane formation was observed almost exclusively in the upper 10 cm of the soil, with the upper 5 cm of the soil being most active. Below 10 cm, methane formation decreased drastically. The compact peat layers, below 30 cm did not show any methane formation (Figure 3.3). In the upper two soil layers (0-5 cm, 5-10 cm) formation of CH4 was exponential indicating that there was no substrate limitation. The difference observed for both layers therefore probably results from a difference in the number of methanogens originally present. A doubling time of approximately 2-3 days could be determined. In the top soil methane formation started immediately. The highest rate of CH4 formation was reached after approximately 40 days for the low water table soil type (0.41 mmol 1-1; 0.042 mg CH4 g-1 dry soil d-l). Ultimately, CH4 formation more or less paralleled CO2 formation.
489
0
0 1: Eb Acetate ~
10
20
Concentration 30
..~E]I
(mmol]l) 40
I
~
,~ CH
,
........
"~"
9 ~-~ ................................................;~-.:.:.: ...............................................
-d
-g
-6O
" ..... ~ . .............................................................................................................. ,
x::
:
121 -80
-I~ - - ~ ..................................................................................................................
G
- [ I.}x 1 ................. ................................................................................................... g
,
9[ ~ , ,
-[
~ .....................................................................................................................
Figure 3.3 The relationship between formation rates of CH4, C O 2 , acetate and soil depth of various soil slurries after incubation under a N2 atmosphere. All concentrations expressed in mmol 1-1. Data obtained for the low water table series (60 cm) Formation of CO2-- A l s o , the formation of CO was highest in the top soil (0 - 10 cm). The initial rate of CO2 formation amounted up to 1.3 mmol.l-1 d-1 (0.36 mg CO2 g-1 dry soil d-l). CO2 formation showed no lag-phase; the rate of CO2 formation decreased in time. After 40 days a small increase was observed again, due to an increased activity of aceticlastic methanogens that produce CO2 and CH4 from acetate. Formation of acetate. Analysis of the soil suspensions for fatty acids showed that the two upper layers (0-5 cm, 5-10 cm) produce considerable amounts of acetate. The top layer produced up to 4.8 mM acetate, which was rapidly degraded by methanogens after 35 days. Small amounts of propionate and butyrate were formed as well.
Methane consumption grassland soils (BMC) Methane oxidation. All 4 applied concentrations were biologically degraded by this type of grassland soil. The highest oxidative activities, especially for lower concentrations (1-100 ppmv), were observed between 5 and 20 cm. One reason for the lower activity in the highest depth (0-5 cm) may be that this region was very wet at the time of sampling, resulting in hampered gas diffusion. The time course of methane degradation is plotted in Figure 3.4 for the initial concentrations of both
490 10,000 (A) and 1 (B) ppmv methane. A correlation between CH4-concentration and degradation rates was observed (0.19 nmol and 1.9 ~tmol g-1 dry soil d-1 for 1 and 10,000 ppmv, respectively) and, most important, it is demonstrated that this soil acts as a sink for methane even at concentrations well below 1 ppmv.
0 ~"
1.00
E Q.
0.90
vo .
0.80
._
0.70
24 ' ~ f
48
72
96
120
r
9
r
9
144
168
0.60 c:
0.50
c'-
0.40
0
0.30 0.20 O.lO
E
~
,"~Y'~'~l~l-
.
9
o.oo 0
24
48
72
96
time -e-
0-5 cm
-l-
120
144
168
(h)
5-10 cm - T -
10-20 c
---
30-40 c
Figure 3.4 Degradation rate of methane in samples from organic grassland soil from different depths in batch cultures at initial concentrations of 10,000 ppmv (A) an 1 ppmv (B) CH4
Identification of methanotrophic bacteria. A decrease of the effiux concentration was observed after 14 days of incubation in the columns incubated with 10,000 ppmv. Here the degradation rate (caused by microbial growth) increased to 19 ~tmol g-1 dry soil d-1 within 7 days. In the column incubated with 100 ppmv the increase was observed after 35 days of incubation. At lower concentrations (1, 10 ppmv) the efflux concentrations remained constant for 40 days, but then these methane concentrations were also degraded. The main goal of this part of work will be to isolate and identify the strains of methanotrophic bacteria which are responsible for the degradation of atmospheric methane concentrations.
Grassland management (BGM) The effect of ground water table. The results presented here are based on observations until July 1994. Grassland with a high ground water level and a relatively thin aerobic layer is expected to show more CH4 emission and/or less immission than grassland with a low ground water level and a relatively thick aerobic layer. However, the site with the relatively high ground water level showed equal or only slightly higher net CH4 emissions than the site with the relatively low
491 ground water level during the measuring period. Net CH4 fluxes were low in the period October 1993 - July 1994, in general less than 0.1 mg CH4 m-2 d-1. The effect of N-fertilization and mowing/grazing. Nitrogen fertilization could decrease the consumption of CH4. Mowing or grazing could effect CH4 emissions by influencing the amount of organic material that is added to the soil annually. However, there were no clear differences between the t r e a t m e n t s during the measuring period. The effect of m a n a g e m e n t intensity. Differences between the different sites were quite large (Figure. 3.5) as were the spatial variations at each of the sites. The site with the lowest CH4 emission had a somewhat lower ground water level t h a n the other sites. In the period J a n u a r y - June 1994, CH4 emission ranged from 0 to 185 mg CH4 m-2 d-1. CH4 fluxes were much higher at the Nieuwkoop area t h a n at Zegveld.
Modelling methane fluxes grassland soils (BMMF) (See before) Anthropogenic sources
Landfills- emission measurements (ALl) D y n a m i c closed-chamber m e a s u r e m e n t s . The spatial variation in m e t h a n e emissions proved to be large on each landfill: differences over a factor of 1000 were m e a s u r e d (e.g. 0.09 g m-2 d-1 to 225 g m-2 d-1 for one landfill and 7.2 g m-2 d-1 to 3150 g m-2 d-1 for another). It is clear t h a t emissions are s p a t i a l l y very inhomogeneous and t h a t local coincidental factors determine the emission rate through the top-layer. So, the m e a s u r e m e n t data are not representative for the total landfill area. Micrometeorological measurements. Depending on the type and age of the landfill, the amount of waste available, the height of the waste tip, the efficiency of landfill gas recovery and the composition of the top layer emissions were m e a s u r e d ranging from 2 to 150 g m-2 d-1. The method as such yielded quite reliable and uniform emission data. Decay rate. A waste decay rate constant (k) of k = 0.1 could be calculated on the basis of these emission measurements, combined with data on amounts of waste, landfill composition and age. This implicates a half-life of 7 years.
492
Nieuwkoopse Plassen, 1994 mg CH4 per m2 per day 200
150
100
50 /r,
Jan
Febr ,,n
Koole
March -
DBZ
April ~
May
June
Brampjesgat
Figure 3.5 Time course of mean CH4 fluxes (mg CH4 m-2 d-l) at three different sites in the Nieuwkoop area
L a n d f i l l s - g a s f o r m a t i o n , e m i s s i o n a n d r e c o v e r y (AL2)
Landfill gas formation. A first - very preliminary - result of the modelling of landfill gas formation by correlating the formation to waste composition, age and amount landfiUed is: {2t =~ 1,87 Q Co kl e-kit where, Dutch (0.094 waste;
Co = amount of degradable organic carbon in the waste in kg ton-1 (the m e a n value for Co is 112 kg ton-l; kl = rate constant of biodegradation y-l); Q = amount of waste landfilled in ton; t = time after dumping of the a t = formation of landfill gas in m3 y-1 (with a mean methane content of 57
vol%); ~ = 0.58 (formation factor).
Landfill gas recovery. In 1993 about 124 million m3 of landfill gas was extracted; 85 million m3 was utilized. In 1992 m e t h a n e emission reduction was 57 Gg (Adviescentrum Stortgas, 1994). During the exploitation period of the landfill, landfill gas formation increases with increasing amounts of waste in place. After closure of the waste tip, landfill gas formation gradually declines. Landfill gas recovery normally starts when the landfill is closed. The effectiveness of recovery is increased when a top-liner system is applied. Normally this is done 5 years after closure of the landfill. The environmental impact of landfill gas recovery is closely connected to its integral recovery efficiency, being the ratio of formation and recovery throughout the years. High integral efficiencies can only be obtained,
493 when landfill gas recovery starts during the exploitation period of the landfill. The technology for doing this is available, and proves to be very cheap, when landfill gas recovery is reckoned in the design phase of the landfill (J. Oonk 1993 and 1994). Emission estimates. Landfill gas emissions have been estimated using the material balance as described in 3.2.2. The CH4 emission estimates from landfills in The Netherlands range from 400-500 Gg y-1. Uncertainties in these estimates are due to uncertainties in the amounts of waste landfilled, amounts of m e t h a n e formed per ton of waste and amounts of methane oxidized in the top-soil of the landfill.
N a t u r a l gas losses and petroleum production (AOG1) The results of this study are summarized in Table 3.3. Total CH4 emissions have been estimated between 127 and 220 Gg y-1. At this point it m u s t be emphasised t h a t none of these estimates have been based on actual measurements, which are required for more accurate quantification.
Exploration and production of natural gas and petroleum (AOG2) The project has not been finalized yet. Preliminary results of the engineering study are partly in a qualitative form still. Table 3.4 presents the sources and source strengths of methane in the oil and natural gas exploration and production. Table 3.3 Methane emissions from natural gas losses and petroleum production in The Netherlands in 1989 (Gg y-l) Sector Natural gas Production/treatment High pressure transport Distribution Consumption Petroleum Production Total
Methane emission
40-70 6.5 65-79 15-30 1-35 127-220
494 T a b l e 3.4 Sources of m e t h a n e emission in the oil a n d n a t u r a l gas exploration a n d production
Sources
Strenghtl
Emission during exploration - drilling - well t e s t s
minor moderate/major
Emission during exploitation of natural gas continuous - vents - flares - e x h a u s t g a s e s of t u r b i n e s - e x h a u s t gases of reciproking engines - e x h a u s t g a s e s of f u r n a c e s - chronic leaks in production - chronic leaks in g a t h e r i n g a n d t r a n s p o r t - glycol d e h y d r a t i o n - t r e a t m e n t of f o r m a t i o n w a t e r - u s e of p n e u m a t i c devices - condensate treatment - condensate storage - p u r g e gas d r o m v e n t i n g s y s t e m s
major moderate minor moderate minor major minor major moderate moderate/major moderate minor moderate
non-continuous
- m a i n t e n a n c e in production - m a i n t e n a n c e of g a t h e r i n g a n d t r a n s p o r t pipelines - incidents a n d accidents in production - incidents a n d accidents pipelines
minor minor moderate minor
Emission due to exploration of oil continuous
- flaring of associated gas - e x h a u s t gases of reciproking engines - t r e a t m e n t of production w a t e r
minor minor minor
non-continuous
- n o n - e x h a u s t engine emissions
minor
Abandoned phase - chronic leaks from a b a n d o n e d wells
none
1minor: <2 Gg y-l; moderate: 2-10 Gg y-l; major: 10-100 Gg y-l; T h e emission e s t i m a t e s are p r e s e n t e d for the t h r e e p h a s e s exploration (drilling a n d testing), e x p l o i t a t i o n (production) a n d a b a n d o n m e n t (closed wells) w h i c h can be d i s t i n g u i s h e d within the oil a n d gas industry. Venting, chronical leaks in production a n d glycol d e h y d r a t i o n a r e m a j o r s o u r c e s d u r i n g n a t u r a l g a s e x p l o i t a t i o n . W e l l - t e s t s a n d t h e u s e of p n e u m a t i c devices can be m a j o r s o u r c e s as well. T h e
495 stationary measurements show several times a year extremely high methane concentrations (e.g. Figure 3.6). The peak of 10 July 1991 was probably caused by an emission of 150x103 kg of methane at a distance of 25 km from Kollumerwaard. Three mobile measurement campaigns were carried out, each covering one day. The first campaign was around the Groningen gas field and consisted of several gas production locations. The second campaign consisted of two exploration activities in the very near neighbourhood of measuring location Kollumerwaard. The third campaign consisted of some of the production locations of the first campaign and two exploration locations of the second campaign. On each of these three occasions no elevation of background concentration was measured.
I0
8
. . . . . . . . . . . . . . . .
J u l y
199!
i
,--=. . . . . . .
-
-- .
.
.
.
.
--
,
.......
6
b
4
3
i
9
!
.
-
- * * o ,
-
0
~4
6 " Time
|hour.~)
Figure 3.6 Daily course of air methane concentration at Kollumerwaard, The Netherlands, at July 10, 1991 Evaluation and validation
Validation (VCI) Laboratory intercomparison exercise. A CH4 laboratory intercomparison exercise was organized, with ECN, TNO, KEMA and LUW as participants. From the exercise, concluded that interlaboratory variation of ambient CH4 analysis (2.0 ppmv) is better than 2% (R.S.D.). The within-laboratory variation varied from between 0.1 - 0.9% (R.S.D.).
496 Trajectory modelling. Figure 3.7 shows CH4 emissions calculated from the CH4 concentration record at Cabauw in combination with air trajectory analysis. A strong preference of air trajectories coming from the south-west direction is observed. Accordingly, the calculated CH4 emissions from that direction will be obtained with satisfactory accuracy whereas for other wind directions, prolonged CH4 monitoring is still required. 'White spots' can be observed in Figure 3.7 for which no emissions have been calculated because since the start of Cabauw CH4 monitoring only I or 2 air trajectories have crossed these areas. Comparison of monitoring sites. By comparing the concentration records of Durgerdam and Cabauw, CH4 sources around Amsterdam have been analyzed. In particular, elevated CH4 emissions from the areas south-east and north of Amsterdam have been observed. These sources have not yet been positively identified due to lack of sufficient overlapping data but may result from cattle and wetland activity in these areas. Isotopes. The applicability of 13C/12C analysis in atmospheric methane is demonstrated in Figure 3.8 which shows a CH4 concentration record at Cabauw on July 22 (1994) and its corresponding change in 13C/12C isotopic ratio. The isotopic ratio of the source of the CH4 peak is calculated from the excess CH4 (i.e., above 1.70 ppmv background) and the measured 13C. In this case, the average of the CH4 sources in the peak amounts of 55 promilles, which indicates a 'normal' composition of fossil and biogenic methane in the peak. From the 14CH4 record at Cabauw, emission factors from Pressurised Water Reactors (PWR) have been determined. The 14CH4-PWR-emissions are higher by a factor 1.5 + 0.3 as compared to the present knowledge (Eisma et al.). In effect, a larger fossil methane contribution is needed in order to obtain the average global 14CH4 concentration. Evaluation and validation from various sources (VEV) Levels and phenomenology of methane concentrations. The measurements at Arnhem and Kollumerwaard have revealed the characteristic levels and main variables determining variation in atmospheric ground level concentrations. Average concentration levels on both monitoring sites were 2.0 ppmv (compared to 1.5 ppmv in 1979 at Terschelling Hollander, 1979) and range from 1.8 to generally 3.0 ppmv and incidentally higher values. These incidental short-lasting high concentrations are most probably due to local sources close to the monitoring site and their occurrence differed therefore at both monitoring sites (Vosbeek, 1993a and 1994b.
Ground level atmospheric concentrations appeared to relate mainly to wind direction and diurnal variation. The variation with wind direction (Figure 3.9) mainly reflected the presence of up wind source areas, modified to a small extent by systematics in meteorological variables like wind speed and mixing height. The patterns for both monitoring sites are striking similar. When conditions with higher wind speed only were selected from the data set, the wind direction profile remained largely unchanged, only resulting in a slightly lower concentration level. These lower concentrations were most pronounced in the maximum of the profile.
497
Overlay Tra|eelodel
-
Me,bane (kgtkm2tyr) Ms;dr 'OS - FeM 1 4
It
Figure 3.7 Methane emissions (kg CH4 km-2 y-l) calculated from CH4 concentration records from March 1993 until March 1994 at the 200 m meteorological tower at Cabauw, The Netherlands, in combination with air trajectory analyses Cabauw
200 m CH4
cone
2.2
- - - = - - d13C
-47.2
I
2.15
-47.4
2.1
-47.6
,....-, 2.05
-47.8
E (3L
2
9,r "r- 1.95 (.3
,-
o
0
-48 ,
,
-
1.9
0 -48.2 ,r..co "o
~
---------
-48.4
1.85
-48.6
1.8
-48.8
1.75
-49
1.7
I 22-07-94 0:00
I 22-07-94 3:00
I 22-07-94 6:00
-49.2
I 22-07-94 9:00
22-07-94 12:00
Date time
Figure 3.8 Methane concentrations (ppmv, left axis) and 13C of peak source the t3C values have been calculated from the measured 13C and the excess of CH4 in the peak (I.E. > 1.70 ppmv). 13C are expressed relative to PDB international carbon isotopic standard
498
Wind distribution
ppm 2"31
lol umerwaar~ --~-- T l l
2.2
I
IArnhemI
2.1 o to
~2.0l-
1.9-
1.8
' ' ' i ' ' ' 1 ' ' ' i ' ' ' 1 ' ' ' 1 ' ' ' 1 ' ' ' 1 ' ' ' 1 ' ' ' 1 ' ' ' 1 ' ' ' 1 ' ' ' i
0
30
60
90
120
150
180
210
240
270
300
330
360
Wind direction
Figure 3.9 The effect of wind direction on the methane concentration in air at A r n h e m and Kollumerwaard, The Netherlands
Diurnal variation mainly reflected the effect of mixing height on the concentrations resulting essentially from ground level sources and probably to a lesser extent from diurnal variations in source strength. The shift in the m a x i m u m at A r n h e m compared to K o l l u m e r w a a r d is related to traffic emissions. The m a x i m u m is delayed to rush hour time and coincides with a maximum in CO concentrations. Source contributions. This part of the project cannot be reported in this stage. U r b a n area (VUA)
U r b a n m e t h a n e emission estimate. The total CH4 emission for the A m s t e r d a m area, based on the use of recent emission factors, was estimated to be about 2 Gg-CH4 y-1 (Veenhuysen et al., 1994). The overall m e t h a n e emission from an urbane area in the Netherlands is low when compared to other sources of methane. Comparison of m e a s u r e d and calculated concentrations. Both calculations and m e a s u r e m e n t s showed a low increase in m e t h a n e concentrations due to u r b a n emissions (some t e n t h s of a ppmv). This was supported by some incidental windward and leeward m e a s u r e m e n t s of concentration levels of m e t h a n e around Amsterdam. The m e a s u r e m e n t s indicated incidental elevated concentrations of some ppmv, t h a t could be a t t r i b u t e d to n e a r b y h i g h w a y s a n d special meteorological conditions. The calculated concentrations were in general higher t h a n the measured concentration. This should be seen in the perspective t h a t for Gaussian dispersion modelling a margin of error of about +_ 100% is usually being accepted. Even if this margin is considered and uncertainty in roughness length is
499 taken into account, the calculated averages were high with respect to the observed ones. This points to the use of too high (although generally accepted) emission factors for road traffic.
3.4 Integration of results Biogenic sources
Rice paddies. Up to now, global estimates of CH4 emission from rice agriculture do not take differences in soil types into account. Bachelet and Neue (Bachelet et al., 1993) proposed to include soil characteristics to revise these global estimates. Using three published techniques to estimate global CH4 emissions Bachelet and Neue found that emission estimates were reduced by 25% if a distinction according to soil type was made. This indicates that including soil characteristics in the global estimates could significantly alter the global estimate. The results from project BRP provide a legitimate basis for the use of'soil type' correction factors to estimate CH4 emissions from flooded rice fields. On the other hand, the results of BRP showed that in the pre-flooding period and the period after harvest considerable amounts of methane may be released from flooded rice fields. In previous monitoring studies these periods were not included, which may cause an underestimation of total seasonal emission by 10-15%. Furthermore, the BRP results show that a reduction in CH4 production does not necessarily cause a proportional reduction in CH4 emission. The data illustrate the importance of both CH4 production and CH4 oxidation when estimating CH4 emission, which emphasises the relevance of the lay-out chosen for the grassland research projects (BMF, BMC, BGM and BMMF). The data also show that the ratio between production and oxidation may depend on environmental conditions. This is important when looking for mitigation options to reduce methane emission from rice fields.
Organic grasslands Formation. By choosing an experimental set-up with certain physical parameters a microbial community is selected, which might represent the natural community only partly. Since this project aimed at the microbial groups that lead to CH4 formation, anaerobic conditions were applied (N2-atmosphere). In all incubations immediate formation of CO2 was observed, indicating that an immediate anaerobic mineralization of organic matter took place. Mineralization was highest in the top soil. Comparable CO2 formation rates have been described for other organic soil types Magnusson, 1993). In this initial stage no H2 and almost no CH4 was formed, suggesting that the reducing equivalents produced during mineralization may be transferred to alternative electron acceptors, like NO3- or SO42-. However, the contribution of reduction of NO3-or SO42- cannot be deduced from the experiments, yet. In the absence of alternative electron acceptors, H2 is used to form methane or acetate by methanogens or homoacetogens respectively. Addition ofH2/CO2 resulted in the top soil in the formation of high amounts of acetate, indicating that active homoacetogens are present. It has been shown before that homoacetogens play a significant role in H2 turnover (Conrad et al., 1989) and CO2 conversion to acetate (Thebrath et al., 1992). It cannot be
500
concluded from this projects data, however, to what extent homoacetogens and hydrogenotrophic methanogens contribute to the conversion of H2 and CO2. The accumulation of acetate in the top soil indicates that the activity of aceticlastic methanogenesis is limiting during the first 30 days. However, between 30 and 50 days the degradation of acetate started to exceed the rate of acetate formation in this layer. From that moment on methanogenesis is limited by the supply of acetate and the rate of CH4 formation parallels the rate of CO2 formation. This 1:1 ratio is expected when organic matter (CH20) is fully degraded to CO2 and CH4 under methanogenic conditions. The accumulation of acetate observed here, without immediate conversion to CH4, has also been described for a Beech forest soil (Kfisel et al., 1994) and for a waste water disposal pond (Kotsyurbenko et al., 1993). This phenomenon appears to occur especially at low t e m p e r a t u r e s , in anoxic environments, where homoacetogenic bacteria contribute to the turnover of H2 (vide supra). At the start of the experiment the soil samples did not contain detectable amounts of acetate. This indicates that the conditions that were applied in the experiments, differ from the conditions in situ. Apparently, in the field the top soil is aerobic enough to prevent the accumulation of acetate. Consumption. The degradation of atmospheric concentrations of methane has often been observed (King, 1993 and Whalen et al., 1990). The found depth profile with a very high methanotrophic activity in depths between 5 and 15 cm has been found by Jones and Nedwell as well (1993). The evidence for different types of methanotrophic bacteria, a low affinity type for the degradation of high concentrations of methane and a high affinity type for the degradation of low (atmospheric) concentrations of methane could be described in kinetic studies by Bender and Conrad (1992) But until now, the high affinity strains could not be isolated and described. Management. Measured net CH4 fluxes were low in the period October 1993 - July 1994, in general less than 0.1 mg CH4 m-2 d-1. Literature data for comparable sites range from 0.1 for a poorly drained grassland soil in winter (Jarvis et al., 1994) to 0.8 mg CH4 m-2 d-1 for an unfertilized pasture (Mosier et al., 1991). Soil analyses in Zegveld showed relatively high sulphate and nitrate concentrations in the soil, especially in the top soil. Both nitrate and sulphate, will have blocked CH4 production. Low soil temperatures in winter will also have contributed to low microbial activities in the soil. It is suggested that flux measurements in summer will provide better insight in possible differences in CH4 fluxes due to grassland management. The measured emission rates from extensively managed grasslands ranged from 0-185 CH4 m-2 d-1 over the period January-June, 1994. In a review and assessment report of methane emissions from wetlands, Bartlett & Harris (1993) arrived at an estimate of 90 mg CH4 m-2 d-1. The results presented here indicate that, in the period September 1993 - J u l y 1994, the impact per hectare natural grassland on cycling of CH4 was much higher than the impact per hectare intensively managed grassland. It has to be emphasised that methane production by cattle is not included in these estimates.
501 Anthropogenic sources
Landfills F o r m a t i o n . There are numerous factors, t h a t influence either the a m o u n t of degradable organic carbon in the waste, the emission factor, or the gas formation rate. In general these factors are all connected to: waste composition, waste t r e a t m e n t and site management. The waste composition determines the a m o u n t of degradable organic carbon in the waste, which is the raw material for landfill gas. Besides t h a t waste composition determines numerous other factors, such as the presence of n u t r i e n t s or inhibitors and the humidity of the waste. Mechanical p r e - t r e a t m e n t , homogenisation, particle size reduction and baling, the extent of composition, the dumping method and the addition of water have significant effects on landfill gas formation. Finally, site geometry, landfill gas recovery, percolate w a t e r m a n a g e m e n t and top-liner system d e t e r m i n e gas formation. W a s t e composition is rapidly changing in the Netherlands, due to separate collection of the vegetable, fruit and g a r d e n (VFG) waste. The r e m a i n i n g waste, after separation, still contains large amounts of degradable organic carbon. This waste will be less humid, and the organic carbon left is relative slow degradable. So as a result of this separate collection, landfill gas formation will decrease, without any doubt, but to what extent is not yet clear. Recovery. Due to efforts of energy companies, waste treaters, NOVEM, and the 'Adviescentrum Stortgas' [the landfill advisory board] the n u m b e r of landfill gas projects is increasing rapidly. About 30 landfill gas projects will be operational by the end of 1995. Developments after 1995 are uncertain. Due to the s e p a r a t e collection of VFG, landfill gas formation becomes unpredictable and landfill gas projects are economically less attractive. This is counteracted by increased legislation. Landfill gas projects are more and more requested in allowances, and in future high (integral) efficiency recovery projects will be prescribed. This m e a n s t h a t in future landfills are obliged to start gas recovery gas during the exploitation of the landfill. Oil and natural gas. The results of the AOG projects allow for an engineering study in which first order estimates of methane emissions can be made. The main result of such an inventory will not be the emission estimate as such, but it facilitates the choice of future research needs of specific sources. It is not a simple m a t t e r to compare emissions from the Netherlands with those reported from abroad. Partly, this is because the quantifications are not equally detailed and partly because regional circumstances make any comparison difficult. With this proviso, one can conclude from Table 3.5 t h a t the relative m e t h a n e emissions from oil and gas production in the Netherlands are comparable with emissions in other countries of n o r t h - w e s t e r n Europe. The emissions per cubic m e t e r of gas produced are somewhat higher in the United States, while the emission situation in the former Soviet Union is far worse.
502 Table 3.5 Emission factors for natural gas production (in percentage of production) Netherlands - on shore - off shore Total
0.03-0.05 0.6-1.0 0.15-0.25
Western Europe United States Former USSR
0.15-0.3 0.2-0.3 +2
Validation Source strength. In general, the CH4 emissions which have been estimated by trajectory modelling in this project are in reasonable agreement with results from national emission inventories, except for the North Sea area. For this region, the CH4 emission as calculated from the Cabauw record, is higher as compared to the emission inventory for this region. Interestingly, the CH4 emission calculated for the Netherlands from the Cabauw record (Figure 3.7) is about 1100 Gg y-l, which corresponds quite well with the national data of the Dutch Emission Inventory (Berdowski, 1994). Detailed interpretation of Cabauw data using a mesoscale transport model has not yet been performed since this requires emission data on a European scale which are not yet available. Isotopic ratios . The applicability of the 13C/12C analysis of atmospheric CH4 is demonstrated in Figure 3.8. For this example it can be concluded from the 513C data that the excess CH4 is of mainly fossil origin. Detailed stable isotopic analysis of a t m o s p h e r i c CH4 peaks (time series) can be a promising tool in source a p p o r t i o n m e n t studies. By contrast, isotopic analysis in air samples t a k e n at Cabauw at a fixed time of the day was found to be of little use. Due to interference of 14CH4 emissions of Pressurised Water Reactors (PWR), 14C could not be used as a tracer as discussed before. 'Local' m o n i t o r i n g sites. Average concentration levels at the monitoring sites Arnhem and Kollumerwaard were 2.0 ppmv and ranged from 1.8 to generally 3.0 ppmv and incidentally higher values. These incidental short lasting high concentrations are most probably due to local sources close to the monitoring site. The lower limit generally is measured at higher wind speeds and western wind direction and it agrees with the northern hemisphere background concentrations as reported recently (Khalil et al., 1993D, 26: 59-62]). Although a G a u s s i a n type of model is not well suited to predict immision concentrations for exceptional situations, the model was used in the Amsterdam Urban Area study for more detailed analysis. It has been investigated whether the elevated concentrations, which were observed incidentally, resulted merely from extreme meteorological conditions and/or source configuration, or were the results of unknown sources of methane. Analysis indicated a possible influence of nearby
503 motorways on the monitoring site although calculated concentrations were higher t h a n measured concentrations. With the large uncertainty of Gaussian dispersion modelling kept in mind the following conclusions can be made: the sources in the urban area still remain only roughly constrained; the emission of methane from the Amsterdam area is only a minor contributor to the national methane emission; no major unknown urban sources of methane are present in the A m s t e r d a m area;
-
traffic is the dominant source of urban methane; the gas distribution system is of such a high quality, t h a t losses are an unimportant source of methane in the Amsterdam urban environment. Finally, t h e r e are indications that, given the u n c e r t a i n t y and v a r i a t i o n in roughness length, the source strengths were even lower than estimated. 3.5 F u t u r e r e s e a r c h n e e d s
Integration and validation Most research during NRP phase-I involved reduction of u n c e r t a i n t y of source estimates. Extrapolation, integration and validation have been formulated, but could have been performed to only limited extent so far. Extrapolation to national and w e s t e r n E u r o p e a n scale have to be carried out using the results achieved during the NRP phase-I, combined with international data. This requires a GIS approach and the development of emission models of the m e t h a n e sources. Such an approach will lead to a database, including the relevant information required for flux extrapolation. The validation of source strengths has to be performed by a combination of several modelling approaches (e.g. empirical and t r a j e c t o r y modelling) on a national and regional scale, concentration monitoring at a number of p e r m a n e n t sites and the results of emission maps, produced by extrapolation.
Specific research needed for CH4 Biogenic sources Global sources. For the estimate of global emissions from rice paddies, already several techniques have been applied in which differences in soil types were accounted for. These techniques have to be developed further, also based on the results achieved in project BRP. Combining soil-type specific CH4 production p o t e n t i a l s w i t h the WISE and EDGAR d a t a b a s e s ( C h a p t e r 5) will allow identification of areas where high CH4 emission can be expected and mitigation options may be especially successful. No attention has been paid to m e t h a n e emissions from ruminants, as the level of knowledge on a national scale is already quite high. However, on a global scale further examination of this sources would be very valid, especially in relation to the food composition in different regions of the world. So far, it has not been able to extrapolate national knowledge on this item successfully. G r a s s l a n d s . The results p r e s e n t e d here are m a i n l y p r e l i m i n a r y r e s u l t s of m e a s u r e m e n t s and analysis. Only parts of the a n n u a l flux courses have been established yet, for one base year only. The monitoring started in September 1993 and has to be continued for a longer period to be able to indicate the a n n u a l contribution of peat soils to the emission of CH4 to the atmosphere. Also, proper
504 interpretation and integration of results has not been done yet, due to the phase the research projects were in at the time of the preparation of this report. This has to be accomplished to yield m a x i m u m profit of the work done until so far. Especially, the flux modelling can prove to be an i m p o r t a n t tool for the data processing of these projects. The water dynamics at Zegveld could serve as input for a gas transport model for organic grassland soils. With this gas transport model oxygen dynamics and methane transport can be described. The oxygen dynamics can be used as input for the methane production model. Results of an integrated model of m e t h a n e production, consumption and transport van be compared with measured field fluxes. This comparison in combination with a sensitivity analysis of the model will show, which aspects need most attention in further research. When the major processes are incorporated in a realistic way, it will be possible to test two hypothesis: during wet periods in situ methane emission is low, because production is limited by the short duration of the anaerobic periods; during dry periods methane uptake by the soil is controlled by transport from the atmosphere to the methanotrophs. These results may be interpreted in relation to future management ideas for the grassland area in The Netherlands. Traditional pig farming. In the definition phase of this methane cluster, emissions from traditional pig farming systems were not considered as an i m p o r t a n t contributor to total national CH4 emissions. Lately, there have been several indications t h a t such systems might be important point sources. As these pig systems are quit abundant in The Netherlands, especially, in a few areas, it is to be considered to at least perform a pilot study to the relative importance of this source type. Anthropogenic sources
Landfills. The level of uncertainty in emission estimates from landfills still is high. Especially, the limited knowledge of the oxidation processes in the top-layer and the i m p a c t of w a s t e composition on formation processes d e t e r m i n e this uncertainty. F u r t h e r research on these items will reduce uncertainty in emission estimates from landfills considerably. Oil & gas exploitation/transportation. During NRP phase-I it was clearly indicated which sources were of major or moderate importance for the exploration, exploitation and transportation of oil and natural gas. However, these sources have been far from quantified yet. further research of the major sources is required to obtain better quantification. At least, first order estimates and uncertainty ranges will be required to use data on this source for regional extrapolation purposes.
4.
NITROUS OXIDE (N20)
4.1 Preparatory studies and organization The programming of the N20 cluster was based on two preparatory studies. One study involved comparison of measurement techniques (TNO; project no. 850012). The second study was the inventory of emissions of N20 (and CH4) from the Dutch territory based on literature (RIVM; project no. 850019).
505
Comparison of measurement techniques The aim of this study was to compare and assess the applicability of incubation techniques, static and dynamic enclosure techniques and the gradient method, and to assess the spatial and temporal variability of N20 fluxes and soil N-contents in peat soils under grassland. Five research groups participated. Two comparison campaigns indicated t h a t static enclosure techniques show large spatial and temporal variability, probably caused by the variability in soil parameters and processes. Static enclosures are, therefore, most useful in combination with studies of the r e g u l a t i n g factors and processes of N20 production and consumption. The gradient technique integrates fluxes over larger spatial scales and this method is most suitable for determining the mean flux from a field plot.
Inventory of Dutch emission The inventory by Van Den Born et al. (1991) was a first attempt to quantify Dutch N20 emissions and to assess the uncertainty in estimates. The major conclusions of this study summarized in Table 4.1 show that only a few sources may dominate the Dutch N20 emissions. Grasslands, particularly those on peat soils, are major sources of N20 in The Netherlands, even if the low end of the range is considered (Table 4.1). The tentative estimates suggested that other potential major N2 0 sources are transportation, arable lands, fresh water and marine aquatic systems and sewage water treatment.
Organization of the N20 cluster In the programming of the N20 cluster the source strength played an important role. Other criteria were the uncertainty in the source estimate, the available expertise, the availability of technical options to achieve emission reductions (policy relevance), and expected future developments in society that will have an effect on emissions. Grasslands were given a high priority, because they cover about 30% of the Dutch land area, and because of the high expected emission rates (Table 4.1). Detailed studies of N20 from arable lands and forests were not included in the N20 cluster. A great number of studies of N20 fluxes from arable lands were done in the period 1979-1986, mostly in the U.S.A., U.K., and Germany. Therefore, only exploratory measurements in arable lands were included (project 852096). Forests were given a low priority because they cover less than 10% of the Dutch land area. This does not imply th at forests may not be a potentially important source, since in The Netherlands N-inputs from deposition are among the highest worldwide. Transportation, aquatic systems and waste water treatment are research topics in the N20 cluster, because of the uncertainty regarding their source strength. One project investigating N20 from combustion processes was included because of the uncertainty t h a t exists since the discovery of sampling artifacts in studies performed before 1988 (Muzio et al., 1988). Current knowledge was synthesized in a global inventory emphasising biogenic soil emissions, but also including inventories for all other sources of N20 identified so far. All components of the cluster aim at verifying the results of the preparatory study, reducing the uncertainty and assessing options for reduction of emissions. The following components are part of the N20 cluster:
506 F. Fossil sources
FT. FC.
Traffic. Investigation of the contribution of traffic to the N20 emissions both now and in the future (IMW-TNO; project no. 850030) Combustion. N20 emissions in combustion processes with respect to the generation of electricity (KEMA; project no. 850006)
B. Biogenic sources
BAQ. BAQ1. BAQ2 BST. BGR. BGR1. BGR2. BGR3. BGR4.
Aquatic systems, including: Water/atmosphere exchange of nitrous oxide in marine systems (NIOZ; project no. 850027) Measurements in fresh water systems (IMW-TNO; project no. 852096) Sewage treatment. The generation of N20 in sewage treatment plants (BKH; project 853133) Grasslands. Research was organized in the integrated N20 grassland project, consisting of 4 sub-projects investigating soil fluxes" Measurement of the atmospheric concentration of N20 from biogenic surface sources in general and grassland ecosystems in particular (TNO-IMW; project no. 852096) Factors affecting the emission of nitrous oxide from grasslands in the removal of nitrate from soil by denitrification (LUW-TPE; project no. 852074) The effects of nitrogen fertilization and grazing on the emission of N20 from grasslands (NMI; project no. 852073) The emission of N20 from grasslands (IB-DLO; project no. 852078)
G. Global inventory
Modelling of soil emissions of nitrous oxide for global studies (RIVM; project no. 852079). In Section 4.2 the methods and research activities of individual projects are discussed. Section 4.3 is a summary of the results of the individual projects and Section 4.4 describes the integration of the results over the Dutch territory. Section 4.5 discusses future research needs.
507 Table 4.1 Emission of N20 in The Netherlands for 1990 made by Van den Bom et al. (1991) and estimates based on results of NRP-I. Emission (Gg N y-l) Van den Born et al. (1991)
Fossil sources energy sector industry commercial & residential transportation
0.2-1.2 0.1-0.6 0.1-0.6 1.4-4.1
Biogenic sources grassland, mineral soils grassland, peat soils arable land & horticulture natural soils fresh w a t e r systems coastal w a t e r s sewage t r e a t m e n t
2.8-5.7 3.0-11.8 2.2-4.1 0.0-0.4 0.7-2.2 2.2-2.8 1.6-3.2
Total
NRP-I1
0.1-0.3 3.4
NE NE/C NE NE <0.3
14.4-36.8
4.2 M e t h o d s
Validation of results In the p r e p a r a t o r y comparison of m e a s u r e m e n t techniques the concentration m e a s u r e m e n t s by various research groups was compared. In order to compare methods two standard gas mixtures were distributed to all participants involved in N20 m e a s u r e m e n t s . The results of this comparison will be analyzed according to ISO standards. Fossil sources
Traffic (FT). On the basis of the available literature emission factors for various types of automobiles and types of engines were estimated. On the basis of this future emissions from transportation were assessed. Combustion (FC). M e a s u r e m e n t s were carried out at power plants, chemical industries, an oil refinery and at a waste incineration plant. A sampling method was developed to prevent N20 formation in the sample caused by the above mentioned "sampling artefact". M e a s u r e m e n t s were carried out within 24 hours after sampling. The N 2 0 emissions were obtained from N 2 0 concentration m e a s u r e m e n t s in stack gases and total stack flow calculated from the fuel throughput.
508 Biogenic sources
Aquatic systems (BAQ) BAQ1. M e a s u r e m e n t s were carried out in the Northwest Indian Ocean, the open North sea and the Dutch coastal zone. The m e a s u r e m e n t s in the Indian Ocean were conducted in 1992 during cruises in the Somali Basin, the northern part of the Arabian Sea, and a cruise between both areas. North Sea m e a s u r e m e n t s were made during a cruise in the central and northern North Sea in 1990. During other North Sea cruises discrete samples were collected. Measurements in the Scheldt river and estuary were done in a u t u m n 1993 and spring 1994. The N20 measurements included analysis of discrete samples from various depths and monitoring of surface waters and atmospheric concentrations. In addition, oxygen and nutrient concentrations were determined to obtain information about the relation between local N-cycling and N 2 0 formation. For the calculation of concentration anomalies and air-sea fluxes, water temperature, salinity and wind speed were measured. The N20 concentration m e a s u r e m e n t s were done with a computerised gas chromatograph. BAQ2. A m e a s u r e m e n t campaign with the gradient technique of TNO was done performed along the Dutch Waddenzee. Fluxes from the fresh w a t e r of the Ketelmeer were determined in spring 1994. The N20 flux was calculated from the concentration gradient measured over the water surface and wind speed. S e w a g e t r e a t m e n t (BST). This project involved a l i t e r a t u r e s t u d y a n d m e a s u r e m e n t s in sewage t r e a t m e n t plants. The objective of the literature survey was to assess the potential for N20 production and emission in the different types of t r e a t m e n t plants in The Netherlands as a function of specific process variables, as well as an inventory of the nitrogen influent and removal for each plant type in The Netherlands. E x p l o r a t o r y m e a s u r e m e n t s were done in installations in Capelle a/d IJssel (covered carrousel with point-aeration) and Alblasserdam (installation with separated nitrification and denitrification), in combination with m e a s u r e m e n t of oxygen concentrations in the water and determination of the amount of influent water and nitrogen. Grasslands (BGR). The integrated grassland project includes 4 projects carried out by different research groups, t h a t focused on flux measurements, modelling N20 production and consumption in grassland soils and assessment of the effect of different agronomic practices on N20 emission. The soils studied included peat, sand and clay soils. Measurements were made at different levels of integration, w i t h soil i n c u b a t i o n t e c h n i q u e s , enclosure t e c h n i q u e s a n d a t m o s p h e r i c m e a s u r e m e n t s with the concentration gradient technique to cover the field scale. Parallel to the integration levels of measurements, the modelling approaches cover the micro-scale to u n d e r s t a n d the processes at the chemical physical level, the rhizotron level, and the field scale. BGR1. This project consisted of two parts: A) Comparison of m e a s u r e m e n t techniques in two campaigns as a continuation of the work in the preparatory study, to investigate the spatial variation of
509
B)
N 2 0 fluxes m e a s u r e d with the static enclosures of NMI and dynamic enclosures and the gradient method of TNO. Continuous m e a s u r e m e n t s of N20 fluxes from grasslands with the gradient technique, daily in the afternoon from April 1993 onwards. Additional m e a s u r e m e n t s in arable lands were made in spring 1994, but results are not yet available.
BGR2. This project concentrated on the processes affecting N 2 0 emission from g r a s s l a n d s by denitrification. The levels of i n t e g r a t i o n r a n g e from the chemical/physical level via the rhizotron scale to the field scale. Incubation experiments were conducted with peat samples collected at Zegveld to investigate the effect of sample location and depth, temperature, initial incubation condition (anaerobic/aerobic), and kind and a m o u n t of fertilizer applied. The rhizotron experiment was designed to study the effect of the depth of the ground water and fertilizer application on N20 losses from a sand soil covered by grass. The N 20 loss to the a t m o s p h e r e was determined with the closed chamber technique by a photo-acoustic spectroscopic infra-red gas analyzer and a multi-sampler described by Velthof and Oenema (1994). Subsoil N20 profiles were determined with a gas chromatograph provided with an Electron Capture Detector (ECD). BGR3. The effects of N-fertilization, grazing cattle and soil type on the N20 emission from grassland were investigated on the basis of field m e a s u r e m e n t s , including: A) Monitoring study during a period of 2 years at 4 sites, including two peat soils with different ground water levels (Table 4.3) in Zegveld, a sand soil in Heino and a clay soil in Lelystad. At each site N20 fluxes were measured weekly with 6 replicates in unfertilized and mown, N-fertilized and mown, and N-fertilized and grazed grassland. The N-fertilizer applied was calcium ammonium nitrate. Additional m e a s u r e m e n t s were done of soil mineral N, herbage yield and N-uptake, soil organic N-mineralization rates and N-inputs from cattle excreta. At all sites denitrification potentials and mineralizable C content were d e t e r m i n e d at 5 depths between 0 and 60 cm. At the Zegveld sites denitrification rates were monitored. Temporal and spatial variability were assessed with detailed measurements on the sand, clay and peat soils. In the sand soil and one of the peat soils N20 concentration profiles in the soil were determined regularly during 1 year. The atmospheric N20 loss was determined with the closed chamber technique, a photo-acoustic spectroscopic infra-red gas analyzer and a multi-sampler described by Velthof and Oenema (1994). B) Assessment of the effect of type and level of N-fertilizer application on N20 emission. Fertilization was done with a m m o n i u m (NH4+), nitrate (NO3-), combinations of NH4 + and NO3-, urea, NH4 + + nitrification inhibitor, and injected and surface applied cattle urine and cattle slurry. C) The development of a management scheme to minimise N20 losses, based on the above experiments and literature data. BGR4. This project concentrated on the effect of urine spots on N20 emission at the field scale, and consisted of an experimental and a modelling component. In the experimental part oxygen and N20 profiles were m e a s u r e d in homogeneous grassland soils as a function of soil depth and distance to field drains. These
510 simultaneous oxygen and N20 profiles provide information on the relation between N20 production and soil moisture. The spatial variability at the 0.001-0.1 m scale was studied in undisturbed soil columns. An experiment with application of urine to undisturbed sandy grassland columns was done in autumn 1993. The results of another identical experiment are not yet available. A model is being developed that describes the controls of N20 fluxes to the atmosphere. The model is based on the field scale model originally developed by Verberne (1992), and it describes C and N transformations in the soil in four interacting sub-models representing:
i) ~) . o o
m
iv)
grass development including C and N turnover and effects of grazing; transport of water and dissolved N; soil microbial C and N transformations; gaseous transport.
Global inventory. A preliminary global inventory of N20 emissions from soils was compiled. The objective was a first validation of global soil emissions. It became clear that for that purpose all sources need to be known. Therefore, grid based emission estimates derived from the EDGAR database (chapter 5) and from the IMAGE project (NRP Theme "Integration of climate change research)" were used. For N20 emission from mineral N-fertilizer and animal excreta a preliminary estimate of 1% of the N-input from was used, based on literature and supported by results of the integrated grassland project. These inventories were tested against inverse atmospheric modelling results for 4 latitudinal zones covering the globe. An attempt was made to explain high observed N20 fluxes after forest clearing observed in the tropics. Soil and plant samples were taken in a tropical rain forest zone in Costa Rica. A model developed by Wageningen Agricultural University, Department of Soil Science and Geology will be used to assess the amount of N liberated by mineralization of soil organic matter in the course of time after forest clearing and pasture development. 4.3 R e s u l t s Fossil
sources
Traffic (FT). Emissions are based mainly on data from the Institute Fran~ais du Petrole (IFP). The major factors influencing N20 emissions are (i) the presence and type of catalyst; (ii) the age of the catalyst; (iii) the air-fuel ratio; (iv) the temperature of the exhaust gas and (v) other factors including tuning and maintenance of the engine. The emission factors derived (Table 4.2) show that aged catalysts give rise to emission factors exceeding those for new catalysts by a factor 3. The study indicated that it is difficult to give uncertainty ranges for these emission factors.
511 Table 4.2 Emission factors for N20 from different types of engines Diesel, light Diesel, heavy Otto, without catalyst Otto, with new catalyst Otto, with aged catalyst
0.031 0.2 0.015 0.035 0.12
Combustion (FC). The observed stack gas concentrations of N 2 0 found for power plants fired with coal, oil and gas were close to ambient air concentrations (0.3 ppmv N20). The fluxes estimated on the basis of stack gas concentrations were low and in most cases negligible. For some types of gas turbines the concentrations in stack gases were somewhat higher, while the N20 concentration for an oil refinery with oil as fuel, vacuum residue and refinery gas was 0.3 ppmv. The N 2 0 concentrations in stack gases from boilers and gas turbines fired with high caloric gas and m e t h a n e gas in the chemical industry were below 0.2 ppmv. Depending on the t e m p e r a t u r e in the combustor, the N20 loss from fluidized bed combustors fired with several types of coal was found to be 4.5-49 g N20-N GJ-1, with higher N20 concentrations at lower temperatures. The N20 concentration at a domestic waste incinerator averaged 2.4 ppmv with peak values of 43 ppmv. The estimated emission from this incinerator was 13 mg N20-N kg-1 domestic waste. Biogenic sources
Aquatic systems (BAQ) BAQ1. In the Somali Basin N20 concentrations in water showed a strong negative correlation with 02 concentrations. Supersaturations of up to 700% were observed in the oxygen minimum zone. The calculated N 20 fluxes to the atmosphere ranged from 1.1 to 3.1 kg N20-N ha-1 y-1. The water column of the Arabian Sea showed zones of extremely low oxygen concentrations, giving rise to N20 consumption by denitrification. As this N20 consumption zone was surrounded by w a t e r masses with high N20 production by both nitrification and denitrification, surface waters were on balance supersaturated. The N20 supersaturations in the w a t e r column ranged from almost zero to 1300% in surrounding waters. In the North Sea the measured N20 saturations varied from 98 to 105%, yielding an atmospheric flux of between -0.04 and 0.2 kg N20-N ha-1 y-1. In the Scheldt river and estuary a close correlation was found between N-concentrations and N20 production. BAQ2. Measurements carried out along the Dutch Waddenzee showed high surface w a t e r N20 concentrations compared to the open sea, with values of around 1000 ng 1-1. The observed emission from the Waddenzee was equivalent to about 1 kg
512 N20-N ha-1 y-1. This is similar to fluxes observed from agricultural lands as shown in the integrated grassland project for mineral soils.
Sewage treatment (BST). The N20 emission from treatment plants from various studies reported in the literature ranges between 0.01 and 6% of the total N-influent. The highest N20 formation occurs in systems where nitrification and denitrification occur simultaneously, such as in aeration tanks, or if conditions for nitrification and denitrification are sub-optimal. During denitrification N20 formation is inevitable. The N20 formation during nitrification can be prevented if process conditions are optimal. Creation of conditions that are optimal for both denitrification and nitrification is not possible. The conclusion of the literature survey was that about 0.4% of the influent-N evolves as N20 averaged over all types of installations. However, the exploratory measurements indicated a loss of N20 of only 0.01% in two installations (Capelle a/d IJssel and Alblasserdam), both having a low influent load and very efficient N-removal. Grasslands (BGR) BGR1. There was a wide range in observed fluxes in two measurement campaigns of the method comparison of BGR1 and the p r e p a r a t o r y study. No firm conclusions could be drawn regarding systematic differences between the various methods. The reason may be that methods representing smaller areas tend to give a larger variation than methods for larger regions, reflecting the scale of N20 production processes t ha t occur in so-called "hot spots". The m e a s u r e m e n t comparison campaigns in BGR1 showed t h a t fluxes measured with static chambers vary at least one order of magnitude within a field. Observed night-time fluxes were lower than daytime fluxes. The fluxes measured with the gradient method showed that the flux was not constant with height, possibly due to inhomogeneity of the source area. In the method comparison study the 24-hours measurements with the static enclosure method seemed to give somewhat lower fluxes, although the order of magnitude was similar to fluxes determined with the gradient method. The results from the monitoring studies with both methods give similar annual results. Continuous measurements from April 1993 onwards showed that during rainy periods the flux determinations are more variable than during dry periods. Shortly after rain events the N20 flux may increase rapidly by accelerated denitrification and nitrification and then decrease again. The N20 pulses may also be caused by soil air that is forced out of the soil pores by infiltrating rain water and by negative fluxes by deposition of N20 on wet surface soils, observed occasionally with both the gradient method and static chambers. However, during the second half of 1993 no strong N20 pulses were measured, because of the high precipitation and high frequency of rainfall events. The estimate based on the gradient method for the flux from fertilized peat soils at Zegveld is 25 kg N20-N ha-1 y-1. BGR2. The peat soil incubation experiments indicated that net N20 production by denitrification occurs during part of the incubation period for all treatments. No net N20 loss occurs during nitrification after initial aerobic conditions. Relatively high amounts of N2 evolved during the incubations. The differences found between soils in different plots and locations were only minor. The denitrification activity decreased with depth and increased with temperature.
513 In the r h i z o t r o n e x p e r i m e n t s the highest N20 fluxes of 17 + 2 g N 2 0 - N ha-1 d-1 were observed for the highest ground water level and highest N-applications, fluxes being mostly h i g h e r after a grass cut t h a n before. Low N 2 0 fluxes were found for high ground w a t e r levels in containers where no N was applied (4 +_2 g N20-N ha-1 d-l) and for intermediate N-application (4 + 3 g N20-N ha-1 d-l). For these cases the fluxes before and after the cut were not different. No significant N20 loss (0 +_5 g N20-N ha-1 d-l) was observed in containers with low ground w a t e r level and no N applied. M o i s t u r e c o n t e n t w a s found to be the m a j o r factor affecting N 2 0 concentration in the soil atmosphere.
Table 4.3 E s t i m a t e d losses of N 2 0 for all t r e a t m e n t s on sand and clay soil and two peat soil w i t h different ground w a t e r levels. The % N20 loss of the N-applied as calcium a m m o n i u m n i t r a t e is calculated as the different between the N 2 0 loss from the fertilized plot and the unfertilized plot, presented as a % of N-fertilizer application. The fertilizer induced loss for grazed g r a s s l a n d is calculated for 1992 as the difference between the N20 loss from fertilizer and animal N-excretion. Amounts of N excreted via urine and dung for 1992 were calculated using s t a n d a r d procedures 1992 1994 Soil input treatment
1993
N-input N20 % of loss
1 9 9 2-
N-input N20 % of
% of N-
loss N-input
N-input
Sand soil unfertilized; mown N-fertilized; mown N-fertilized; grazed
0 313 743
1.2 3.1 7.3
0.6 0.8
0 426 4261
1.0 6.6 13.2
1.3 -
0 277 557
1.0 5.0 10.6
-
0.5
-
1.4 1.7
437 4371
2.6 16.1
1.5 -
0 266 521
2.1 8.0 11.9
2.2 1.9
0 464 4641
1.8 9.6 17.3
1.7 -
1.9
0 161 356
12.9 20.2 36.0
4.5 6.5
0 323 3231
4.2 15.9 41.0
3.6 -
3.9
1.0
Clay soil unfertilized; mown N-fertilized; mown N-fertilized; grazed
0
0.9
Peat soil, grondwatertrap II2 unfertilized; mown N-fertilized; mown N-fertilized; grazed
Peat soil, grondwatertrap III3 unfertilized; mown N-fertilized; mown N-fertilized; grazed
1 The N-input for 1993 does not include N from animal excreta; therefore the N20loss as % of the N-input is not calculated.
514 Grondwatertrap II = means annual low ground water level at 50 - 80 cm below surface. 3 Grondwatertrap III = means annual low ground water level at 80 - 120 cm below surface; the mean level was about 20 cm lower than for the site with grondwatertrap II 2
BGR3. The results of the monitoring study showed peak fluxes of N20 during the first 1-2 weeks after fertilization and grazing, except during dry conditions. High fluxes were observed during the wet period from July till the end of September 1993. Fluxes from the peat soil with low ground water level were much higher than from the other soils, and for all soil types the fluxes were higher for grazed plots then for ungrazed plots. Winter fluxes were low in all treatments, except those from the peat soil with low ground water level (grondwatertrap II, Table 4.3), where occasionally significant N20 fluxes were observed in winter. Fluxes of N20 from the clay soil in Lelystad were low during the wet period between July and September 1993, while fluxes from the other sites were high during this period. The results of the monitoring study between March 1992 and March 1994 show losses of N 2 0 of up 36 (in 1992-1993) to 41 (in 1993-1994) kg N 2 0 - N ha-l, observed in grazed plots on peat soils with deep ground water tables, while the losses for the corresponding periods for high ground water levels were 12 and 17 kg N20-N ha-1 (Table 4.3). These results are confirmed by the estimated 25 kg N20-N ha-1 y-1 based on the concentration gradient method of project BGR1. The N20 losses from the grazed grassland on clay soil in Lelystad were also high: 11 and 16 kg N ha-1 for 1992-1993 and 1993-1994, respectively. For the grazed sand soil in Heino the losses were 7 and 13 kg N ha-1 for 1992-1993 and 1993 and 1994, respectively (Table 4.3). In fertilized and mown plots the N20 losses for the 2-years period were about 1% of the N-input from synthetic fertilizer for the sand and clay soils, and 2% and 4% for the peat soils with respectively high and low ground water level (Table 4.3). In fertilized and grazed plots the N20 loss for 1992 1993 amounted to 0.8% of the total N-input from synthetic fertilizer and animal excreta for sand soils, 1.7% for clay soils, 2% for the peat soil with high ground water level, and 6.5% for the peat soil with low ground water. In the calculation of the % losses no correction was made for mineralized N. BGR4. Losses from urine spots calculated from field measurements appear to be less t h a n 1% of the amount of urine N-input, even for doses of up to 1300 kg N ha-1. The N20 loss was promoted by high concentrations of mineral N and by wet soil conditions, during at least 2 months after the urine application. Direct damage of the grass by urine (urine scorch) led to less re-growth, lower N-uptake and less transpiration at the urine spot, and these conditions are favourable for N20 losses via nitrification and denitrification.
Global inventory. The global estimated N20 emissions for different sources were calculated for 4 latitudinal zones covering the globe from the original 1~ x 1~ grid. The results were compared with inverse modelling estimates for these zones. The sources included: soils under natural vegetation (4.3-4.5 Tg N20-N y-l), grassland soils (1.4-1.5 Tg), cultivated soils (1.8-1.9 Tg), direct emissions from s a v a n n a
515 burning and deforestation (0.1 Tg), post burn effects on N20 emissions caused by deforestation (0.4 Tg), agricultural waste burning (0.1 Tg), emissions from animal excreta (1 Tg), commercial energy use (0.2 Tg), fuelwood combustion (0.1 Tg), adipic acid production (0.2-0.4 Tg), nitric acid production (0.2 Tg), and oceans (2.0-3.8 Tg). The calculated ratio northern hemispheric : southern hemispheric emissions of 1.5-1.8 is consistent with results obtained in other studies. The comparison with results of inverse modelling techniques showed that an estimation of 1% of N20 loss induced by synthetic fertilizer inputs is probably more appropriate than the lower N20 loss rates assumed by IPCC (1991). The results of this study also show that animal waste and tropical land clearing are significant N20 sources. The calculation of the effect of land clearing accounted for generally observed gradually declining N20 fluxes along with the age of the clearing, resulting in a lower estimate than that of IPCC (1990). A higher global oceanic emission than the 2 Tg N20-N y-1 assumed by IPCC (1990) is probable, with high contributions from the Antarctic ocean and the 0~176 tropical zone. The latter conclusion is supported by oceanic measurements of project BAQ1 and literature. 4.4
I n t e g r a t i o n of results
Fossil sources
Traffic (FT). The 1990 emission may be 3.4 Gg N20-N y-l, based on the emission factors in Table 4.2 and statistics on Dutch transportation activity levels, (RIVM, 1993). An increase for this source to 9.7 Gg N20-N y-1 was projected as a result of increasing mobility and the further penetration of catalyst-equipped vehicles. It is difficult to determine the uncertainty of this estimate, since the published emission factors in literature vary by more than a factor 10. However, it is obvious that the current emission is close to the high end of the range given in the preparatory inventory by Van Den Born et al. (1991) (Table 4.1), and that this source will increase in the future. Combustion (FC). The N20 emission from fossil fuel combustion may amount to only 0.1-0.2 Gg N20-N y-l, based on the emission factors presented in Section 4.3. The emission from waste incineration is an insignificant amount of less than 0.05 Gg N20-N y-1. This confirms recent publications (Dlugokencky et al.) showing that fossil fuel combustion is an insignificant source of N20. These estimates are lower than those made in the preparatory inventory of Dutch emissions (Table 4.1). Biogenic sources
Aquatic systems (BAQ). The N20 fluxes of 1.1-3.1 kg N20-N ha-1 y-1 observed in the Somali Basin and in the Arabian Sea are much higher than the mean global oceanic flux of 0.04-0.2 kg N20-N ha-1 y-1 (Butler et al., 1992). This indicates the importance of these waters for regional emissions. This conclusion is supported by the global inventory (see below). The North Sea, excluding the coastal regions, may be a net source of N 20 of about the same magnitude as the global mean. Although fluxes have not been calculated yet, the emissions observed in the Scheldt river and estuary must be higher than those in the North Sea, as the N20 concentration increased one order of magnitude from the mouth of the estuary upstream towards
516 Antwerp. Here a close correlation has been found between the N-loading and N20 fluxes. Extrapolation of the emission rate of i kg N20-N ha-1 y-1 to the area of 250000 ha of the Dutch Waddenzee yields a figure of 0.25 Gg N y-1. The observed emission rate is much higher than the 0.0042 kg N20-N ha-1 y-1 from earlier measurements in intertidal sediments in the Western part of the Waddenzee (Kieskamp et al., 1991). It is not clear what causes this difference. The gradient method applied by TNO is an actual measurement of fluxes, while Kieskamp et al. (1991) calculated the flux on the basis of N20 concentration in the surface water. Assuming that the gradient technique yields the actual flux, further work on this subject may yield insight in the relation between the total water soluble N-concentration in the water and N20 fluxes.
Sewage treatment (BST). The total Dutch emission from sewage t r e a t m e n t plants may be 0.33 Gg N20-N y-l, based on a total of 80 Gg y-1 of N-influent to sewage water t r e a t m e n t plants in The Netherlands and a N20-N loss rate of 0.4% of the N-influent derived from the literature. The exploratory measurements indicate a much lower N 2 0 emission of 0.01%. The number of measurements was, however, too limited to use this number for extrapolation. They indicate, however, that the estimate based on literature data is possibly an overestimation, and t h a t sewage water treatment plants are of relatively minor importance (Table 4.1). Grasslands (BGR). It is interesting to note that field m e a s u r e m e n t s for the peat soils at Zegveld showed highest N20 losses from plots with deep ground water, while in rhizotron experiments with sand the highest fluxes observed were found for high ground water. This is clearly a result of differences in soil conditions. With low ground water the organic soil material quickly decomposes and huge amounts of N are mineralized. The high microbial activity, relatively poorly drained soils and high available N, created conditions favourable for nitrification and denitrification. In the well aerated sand in the rhizotrons of project BGR2 a high water table also created anaerobic conditions, causing a significant increase in N20 flux relative to the rhizotrons with low ground water. The model development in BGR2 was not planned in this phase of NRP, while the model of project BGR4 is not yet completed. First conclusions on the basis of the integrated grassland project are that the N20 emission from peat soils under grasslands are important in The Netherlands. The measurements are of the same order or even higher as the preparatory estimates by Van Den Born et al. (1991). Another important finding is that urine spots have a long lasting effect on N20 fluxes and need to be included in field measurements of grasslands. However, it is not yet possible to reduce the uncertainty of the estimate (Table 4.1) and to extrapolate the results (see Section 4.5). A m a n a g e m e n t scheme for good agricultural practice to minimise N20 emission should restrict periods with increased mineral N contents in the soil. Possible m e a s u r e s include type of N-fertilizer, timing of application, reduction of the N-input according to the needs of the crop or grass, use of nitrification inhibitors, soil drainage, reduction of the period of grazing and increase of N-use efficiency by animals to achieve a reduction of N-excretion.
517
Global inventory. The tentative estimates for the various N 2 0 sources confirm t h a t at the global scale soils are dominant sources of N20, seconded by the oceans. Agricultural activities may be responsible for the major p a r t of the observed a t m o s p h e r i c increase. There are m a n y u n c e r t a i n t i e s to be resolved. Major uncertainties are in the estimates of N 2 0 emissions from tropical soils and the role of legumes in natural ecosystems and leguminous crops (covering about 10% of the world's cultivated area), N20 emissions from oceans, inland waters, coastal waters, and continental shelves, and industrial sources. E s t i m a t i o n of these sources is h a m p e r e d by a lack of m e a s u r e m e n t data. The global a m o u n t of N-fixation by leguminous crops m a y be of the same order of m a g n i t u d e as synthetic nitrogen fertilizer use and N from animal excreta. Aquatic systems may be very i m p o r t a n t sources of N20 t h a t are increasing due to ever increasing a n t h r o p o g e n i c N - i n p u t s in the e n v i r o n m e n t , such as l e a c h i n g of N from agricultural lands and h u m a n waste. The N20 loss of 1% of the N-input used for N-fertilizers and animal waste seems to be a reasonable assumption, and is in agreement with the results obtained in the integrated grassland project for sand and clay soils. This suggests t h a t fertilized fields are major sources of N 2 0 at the global scale, and this offers possibilities for reducing emissions in order to control the atmospheric increase. The t e n t a t i v e estimates of the N 2 0 from animal excreta and those caused by deforestation show t h a t these sources may be significant. 4.5 F u t u r e r e s e a r c h n e e d s
Integration and validation Most r e s e a r c h in the causes of climate change of the N R P - p h a s e I involved improvement and reduction of uncertainty of source estimates, and identification of sources. Extrapolation and validation of results were not major research topics. Extrapolation to the regional or national scale require a GIS approach, based on m e a s u r e m e n t results and, particularly for the biogenic CH4 and N20 sources, process and m a n a g e m e n t models. These models were not planned in phase I of NRP, or are not yet operational. The geographic information for extrapolation is available in various locations in electronic form, but not as one complete set of data with common scale and format. A general research need t h a t supports source and sink m e a s u r e m e n t s is, therefore, the development of a database including all basic information required for flux extrapolations. Validation of source and sink estimates can only be done at regional or national scales, i.e. scales larger t h a n point source scale or field scale. The emission maps produced by extrapolation and emission inventories for sources outside the Netherlands and sources not investigated, would form the basis for such validation efforts.
Specific research n e e d e d for N 2 0 Fossil sources. C o m b u s t i o n is an insignificant source of N20. H o w e v e r , transportation probably contributes significantly to Dutch emissions, and may be an important global source. With further penetration of catalyst-equipped vehicles and with ageing of existing catalysts, this source m a y show a rapid growth. However, the uncertainties of the emission factors are very high. M e a s u r e m e n t s
518 under real conditions of N 2 0 emission from vehicles with and without catalyst are required to verify and improve the estimates presented in Table 4.2. In addition, this research would have explore ways to prevent these N20 emissions.
Aquatic systems. The preliminary m e a s u r e m e n t s in the Dutch Waddenzee, the North Sea and the Indian Ocean, and the global inventory, indicate t h a t the knowledge on fluxes from aquatic systems is still inadequate to produce reliable estimates. However, N20 fluxes from the Waddenzee may be of the same order of magnitude as those from heavily fertilized agricultural lands. These relative high fluxes are probably caused by h u m a n activities. Part of the NO3- and other N-compounds t h a t are lost from soils by leaching finally end in surface waters. F u r t h e r inputs of N, and of organic C, to aquatic systems include h u m a n waste, effluent from sewage water treatment plants and industrial waste. This indicates t h a t research is required into the relation between organic C- and N- input into in rivers, coastal systems and estuaries, and CH4 and N20 fluxes.
Sewage treatment. The exploratory m e a s u r e m e n t s indicated t h a t sewage t r e a t m e n t may have only a minor contribution to the N20 emission from The N e t h e r l a n d s , and t h a t possible technical measures to reduce emissions are relatively well known. F u r t h e r study into this source is not an urgent research priority.
Grasslands. In the first phase of NRP an a t t e m p t was made to refine the estimates of CO2, CH4 and N 2 0 fluxes from grasslands as a function of soil properties and physical/chemical factors in the soil, fertilization, grazing, and soil and water management. Part of the modelling work needs to be completed and continued in order to make better extrapolations of the results. This is needed in order to investigate the effect of land use transformations on fluxes of greenhouse gases. In Europe i m p o r t a n t changes in land use are envisaged, particularly affecting arable lands and grasslands. Conversion of grasslands to arable lands, as well as rotations with 7 years of grassland followed by some years of maize or fodder beets is increasingly applied to obtain "virgin" soils. Such conversions may have effects on fluxes of CO2, CH4 and N20. Fluxes of one gas may be reduced while fluxes of another gas may increase at the same time. Because of the complex interrelations between controlling factors of CO2, CH4 and N20 production and consumption, future research on trace gas exchange between grasslands and the atmosphere can best be done as a combined research effort. Extrapolation of the results for grasslands requires models in combination with a GIS with information on soils, ground water tables, land use, N-fertilizer inputs, animal densities, N-excretion by animals, length of grazing periods, etc. Currently these data are not readily available for such calculations. In addition there are a number of research topics specific for N20 emission from grasslands that need research and policy attention, including: the effect of techniques to reduce ammonia volatilisation on dairy farms on emission of other compounds; there are indications that reduction of ammonia losses from animal waste may induce increases in N20 and CH4 emission; nitrate leached from soils may be denitrified in subsurface environments, producing N20. Nitrate may also enter surface waters, as discussed below.
519 Losses of N 2 0 dissolved in percolating water may be lost to the atmosphere by degassing from e.g. ditches and canals. There is great uncertainty regarding the magnitude at the global scale, but available publications show that this is a potentially i m p o r t a n t global source. Considering the current excessive inputs of N in Dutch agriculture, this may be very important at the national scale.
Sources not i n v e s t i g a t e d in N R P I As indicated in 4.1 a number of potentially important sources were not included in the N20 cluster. These include arable lands, forests and industrial processes. Arable lands and forested lands. The N 2~ cluster of phase I of NRP did not include detailed studies of N 2 0 losses from arable land. Significant N20 losses were m e a s u r e d from the mineral soils in the integrated grassland project. In the Netherlands there are no quantitative estimates and good field measurements for N20 losses from arable lands, including maize land with excessive N-inputs from animal waste, and horticulture.
Recent German research showed that acid forest soil that receive N-inputs from deposition may show emissions in the order of 5-6 kg N20-N ha-1 y-1 (Brumme et al., 1992). Dutch forests receive even higher N-inputs from deposition, and m a n y forest soils are in the process of acidification. Therefore, arable lands and forests may be much more important than suggested by the inventory in the preparatory study. I n d u s t r i a l sources. IPCC (1992) mentioned two i n d u s t r i a l N20 sources, i.e. production of adipic acid and production of nitric acid, that had not been identified when the preparatory inventory was made. Experts expect that all industrial and chemical processes in which nitrogen oxidation steps in overall reducing conditions are involved, are potential sources of N20 (Olivier, 1993). For these sources no e s t i m a t e s have been r e p o r t e d in the l i t e r a t u r e . The N e t h e r l a n d s is an industrialized country with m a n y chemical industries. An inventory of these i n d u s t r i e s coupled with collection of d a t a on N 2 0 losses and a d d i t i o n a l m e a s u r e m e n t s is required to improve Dutch estimates and to develop techniques and policies to reduce these sources.
Global inventory There are a number of major knowledge gaps in the global N20 budget. The major topics where research is required include: Emissions from soils under natural vegetation and from the world oceans. In particular, there is uncertainty in the emissions from the tropical zones. Furthermore, the contribution of episodic emissions from temperate soils in winter, early spring and autumn is poorly known. The contribution to global N 2 0 fluxes of N-inputs by N-fixing leguminous crops. The global amount of N-fixation by these crops is of the same order of magnitude as mineral N-fertilizer use and N from animal excreta. Considering that a part of this nitrogen ends in soils, this may be an important global N 2 0 source.
520
-
Part of the N-losses from various sources to ground water and aquifers and via aquifers to surface waters may finally end via rivers in estuaries, coastal waters, continental shelves and oceans. As shown for the Waddenzee, these aquatic systems may be a very important and very local sources. At the global scale knowledge about the amount of nitrogen transported this way is inadequate. Research involving an inventory of global N-transport though aquatic systems will yield the basic information required to extrapolate models of the relation between N-loading in the water and N20 emission. Quantification of N20 emission from industrial processes.
5.
EMISSION DATABASE DEVELOPMENT
-
5.1 W o r l d I n v e n t o r y of S o i l E m i s s i o n p o t e n t i a l s (WISE)
Methods The aim of this project (project 851039; ISRIC) was to make a global gridded d a t a b a s e of the major soil factors t h a t play a role in the production and consumption of greenhouse gases, such as CO2, CH4, N20 and NO, and to apply this database to assess CH4 emissions from rice cultivation. The project consisted of two phases. In the first phase an inventory based on available literature was made of the soil parameters and processes t h a t determine the production and consumption of CO2, CH4, N20, and of existing models to describe these processes. On the basis of this inventory a list of soil factors was made t h a t should be included in a global soil database. A workshop was organized with soil and rice experts participants from all over the world. The development of the database in the second phase was based on the 1/2~ Y2~version of the FAO/Unesco soil map of the world prepared by FAO-AGL in collaboration with WISE staff. The soil type information was coupled with soil information from representative profiles. In total 3000 detailed soil descriptions from all continents were included with analytical data. Secondary soil characteristics required for soil models will be derived from climate, parent material and land use.
A global inventory of the potential for CH4 production of rice soils will be made. In collaboration with the I n t e r n a t i o n a l Rice Research I n s t i t u t e (IRRI) and Wageningen Agricultural University, Department of Soil Science and Geology, a model for describing CH4 emission from wet rice fields will be developed. This model will be adapted to make use of the data available in the soil database. If this model proves a useful tool, it will be applied to make a global estimate of CH4 emission from rice cultivation and natural wetlands. Results The software to import soil profile information into the WISE d a t a b a s e is operational. About 3000 profiles have been imported and validated. The final version of WISE will include information for 3500-4500 profiles. A 1/2~ 1/2~ resolution version of the soil map of the world is being linked to the soil profile data. A start has been made with the development of algorithms to derive soil data for modelling purposes. Work on the development of simple models in co-operation with IRRI and Wageningen Agricultural University is ongoing.
521 There have been unforeseen problems in this project that have caused delays: the r e s p o n s e of n a t i o n a l soil s u r v e y i n s t i t u t e s to m a k e a v a i l a b l e representative soil profiles was disappointing; the a m o u n t of time needed for proper checks of the imported d a t a was underestimated; the p r e p a r a t i o n by FAO of the 1/2~ 1/2~d a t a b a s e of soil types has been proceeding more slowly than expected; development of the models to calculate CH4 emission from wet rice fields has revealed the complexity of the relationships between the process-regulating factors in wet rice fields, i.e. soils, land use, climate and hydrology.
5.2 E m i s s i o n D a t a b a s e for Global A t m o s p h e r i c R e s e a r c h (EDGAR) Methods This project (RIVM projectno. 851060) involved the development of a global emissions database with base year 1990 with l~ 1~ spatial resolution and altitude resolution of 1 km for the following compounds: CO2, CH4, N 2 0 , CO, NOx, n o n - m e t h a n e VOC, SOx, of NH3, and ozone depleting compounds (halocarbons) from all known sources. The database development consisted of information analysis, system design and software development by means of prototyping. Major processes described in EDGAR are land use, energy consumption, other industrial production and consumption, and waste handling. The emissions from processes are e s t i m a t e d from so-called activity levels and emission factors. Activity levels comprise demographic data, social and economic factors and land use and vegetation distributions. Country emissions are distributed on the basis of the co-ordinates of major point sources or by using so-called allocation functions (maps on grid, such as h u m a n population, a n i m a l populations and e n e r g y consumption). As far as possible statistical data are from internationally accepted sources to ensure comparability and efficient updating. For emission factors, representativity and availability of data as well as compliance with the Global Emission Inventories Activity (GEIA) (project included in the I n t e r n a t i o n a l Global A t m o s p h e r i c Chemistry Programme, IGAC), OECD and European emission database systems are i m p o r t a n t criteria. To g u a r a n t e e the quality of the data, checks on consistency, anomalies and completeness are performed on the result of the final data processing. A n u m b e r of inventories produced by GEIA and other institutes are included in EDGAR. In the framework of GEIA TNO and RIVM co-ordinate the work on global biogenic sources of N20 and NOx, and inventories of all sources of NH3 and VOC emissions. Results The first version of EDGAR containing activity levels and emission factors for the above sources and compounds will be completed in December 1994. A preliminary version will be made available in summer 1994. The first version of EDGAR allows for:
522 definition of source categories, data sets for variables connected to each category, including allocation functions for distributing emissions to the lOx 1 ~ resolution, and references for all data included in the database; importing, inspection by value and visual checks of activity levels, either by country or on grid, emission factors by country of region, and allocation functions to distribute national emissions to grid cells; calculation of emissions on grid and by region of selected source categories and inspection of results through maps (on grid and by country) and tables (by region or country); exporting of emission data in a specified format. Although for users only emission estimates for source categories will be made available, which is sufficient to meet the user requirements as expressed during the EDGAR user workshop, a full data source description of the underlying data is available for all data included in EDGAR. Additional EDGAR results are contributions to: GEIA: inventories completed so far of anthropogenic and biogenic sources of N20, and of VOC sources. IPCC National Greenhouse Gas Inventory programme: a default calculation scheme and default emission factors for CH4 and N20 from fuel combustion and industrial processes. IMAGE 2.0: aggregated regionally calibrated emission factors. For some sources, such as oceans, volcanoes and lightning, the emissions for some compounds cannot be generated in this version of EDGAR caused by lack of reliable data. 5.3 Future
research
needs
with respect
to database
development
The development of the WISE d a t a b a s e and EDGAR has t a k e n place in N R P - p h a s e I, but there has been very limited opportunity for developing applications of the products. For the WISE soils and derived databases, a number of possible research activities are apparent: Distributions of soil characteristics and derived properties, including soil texture, soil carbon and nitrogen, soil reaction, soil fertility, presented as 1/2~ 1/2~ resolution databases, form an important improvement compared to the generally used l~ 1 ~ soil type information in e.g. the model of the terrestrial carbon cycle. In addition, the derived data form important inputs in dynamic assessments of natural emissions of CH4 and N20. The inventory of soil properties important in CH4 production in wetland rice soils and n a t u r a l wetlands could be used in attempts to develop temporal emission distributions, and in extrapolations of emissions as done in the EDGAR project. Possible applications of the EDGAR database are related to validation of the emission inventories. The first version of EDGAR is based on measurements and extrapolated with models or with so-called allocation functions. Each global inventory has its specific uncertainty, related to either the m e a s u r e m e n t s or the methods of extrapolation. Uncertainty analysis and validation can only be done by using atmospheric chemistry and transport models. Such assessments have to be done on the basis of temporal distributions, since atmospheric models all have
523 their specific requirements in terms of temporal distribution. Hence, the production of temporal distributions of emissions and u n c e r t a i n t y analysis are i m p o r t a n t research needs related to validation. F u r t h e r m o r e , there are a n u m b e r of major omissions in the EDGAR database, such as emissions of ammonia, aerosols and compounds involved in aerosol formation (sulphate and DMS). 6.
SOCIO-ECONOMIC CAUSES
6.1 M e t h o d s The objective of the project (project 850019; RIVM) was to develop and apply methodologies to determine national inventories of greenhouse gas emissions. The project focused on m e t h a n e and nitrous oxide, because least was known about emissions of these greenhouse gases in the Netherlands. In three phases, three reports were planned: a first inventory of Dutch emissions (see Section 4.1), and two subsequent detailed overview reports on the state-of-the-art of knowledge about emissions of m e t h a n e and nitrous oxide. A secondary objective was the identification of knowledge gaps, which could be used by the NRP for programming e x p e r i m e n t a l emissions research. The main research method was l i t e r a t u r e survey and discussions with experts. Initially emission estimates were primarily based on international literature, later experimental data from Dutch research could increasingly be used. The focus of the project shifted rapidly from social causes to methodologies for national emissions estimates as a contribution to the IPCC methodology development. One of the contributions consisted of the organization of an international workshop on CH4 and N20. 6.2 R e s u l t s The first result was an inventory of the current and an estimate of future Dutch emissions for all greenhouse gases. This inventory was used as a basis for the development of greenhouse gas policies in the 2nd National Environmental Policy Plan and for setting priorities in selecting experimental projects for the NRP. The experience led to active RIVM-participation in the IPCC-OECD P r o g r a m m e on Developing Guidelines for National Emissions Inventories. A background study on m e t h a n e updated the previous estimate of national m e t h a n e emissions, partly based on new experimental data. It was found t h a t m e t h a n e emissions in the N e t h e r l a n d s would probably decrease by more t h a n 10% in 2000 relative to current levels as a result of policies in the area of waste and m a n u r e control. M a n u r e related emissions were identified as an i m p o r t a n t "new" and growing source. Specific policies were identified t h a t could f u r t h e r decrease national emissions. Currently, a similar report on nitrous oxide is being finalised. In this report, an i m p o r t a n t discrepancy is identified between the methodology proposed by IPCC and a more detailed method, developed in this project. The main reasons for this discrepancy are differences in emission factors and the omission of a n u m b e r of significant sources in the IPCC methodology. Future Dutch emissions of N20 are believed to increase, while increasing emissions due to the introduction of car catalysts are not fully compensated by decreasing emissions due to limitation of nitrogen deposition (atmospheric and t h r o u g h manure and fertilizer).
524 6.3 F u t u r e r e s e a r c h n e e d s There are two major research topics which merit research attention. First, the background reports on CH4 and N20 show that the IPCC methodology for national emission inventories needs to be complemented with methods to derive estimates for a number of identified but unquantified anthropogenic sources. Secondly, the IPCC method leaves a great liberty in using emission factors. The national or regional results need to be compared with estimates based on global methods, such as in EDGAR, or with results of validation efforts with atmospheric models. On the basis of such comparisons the proposed national methods can be revised.
7.
R E F E R E N C E S AND P U B L I C A T I O N S
Adviescentrum Stortgas, 1994. Annual report 1993, (in Dutch), Utrecht. Baas, J., 1991. Literatuuronderzoek naar de lachgasemissie door het wegverkeer (in Dutch). r e p o r t R 91/322, N e t h e r l a n d s e O r g a n i s a t i e voor Toegepast-Natuurweten-schappelijk Onderzoek, Instituut voor Milieuwetenschappen (TNO-MW), Delft, The Netherlands, 65 p. Baas, J. and H.S.M.A. Diederen, 1992. Voorstudie naar de meetmethoden voor de bepaling van de lachgasemissie uit de bodem. (in Dutch). Report No. R 92/211, N e t h e r l a n d s e O r g a n i s a t i e voor T o e g e p a s t - N a t u u r w e t e n s c h a p p e l i j k Onderzoek, Instituut voor Milieuwetenschappen (TNO-MW), Delft. 71 pp. Bachelet, D. and H.U. Neue, 1993. Methane emissions from wetland rice areas of Asia. Chemosph. 26: 216- 238. Balfour, W.D., C.E. Schmidt and B.M. Eklund, 1987. Sampling approaches for the measurement o volatile compounds at hazardous waste sites. J. Haz. Mat. 14: 135- 148. Bartlett, K.B., and R.C. Harris, 1993. Review and assessment of m e t h a n e emissions from wetlands. Chemosph. 26: 261-320. Batjes, N.H., 1993. World Inventory of Soil Emission Potentials: Guidelines for soil profile selection and protocol for completing the WISE data entry sheets. Working Paper and Preprint 93/02, ISRIC, Wageningen. 32 pp. Batjes, N.H., 1993. World Inventory of Soil Emission Potentials: Profile database user's manual (Version 1.0). Working Paper and Preprint 93/04, ISRIC, Wageningen. 28 pp. Batjes, N.H., 1994. Report of a joint ISRIC-IRRI Questionnaire on Site Characteristics of Irrigated Rice lands of Relevance to Methane Production and Emission. Working Paper and Preprint 94/02, ISRIC, Wageningen. Batjes, N.H. (ed.), 1992. World Inventory of Soil Emissions: Report of working group discussions and recommendations. Proceedings of international workshop organized within the framework of the Netherlands National Research Programme on Global Air Pollution and Climate Change. WISE Report 1, ISRIC, Wageningen. 20 pp. Batjes, N.H. and E.M. Bridges, 1992. A review of soil factors and processes that control fluxes of heat, moisture and greenhouse gases. Technical Paper 23 / WISE Report 3, ISRIC, Wageningen. 201 pp. Batjes, N.H. and E.M. Bridges, 1995. Potential Emissions of Radiatively Active Gases from Soil to Atmosphere with Special Reference to Methane: Development of a Global Database (WISE). J. Geophys. Res. (in press).
525 Batjes, N.H. and E.M. Bridges (eds.), 1992. World Inventory of Soil Emission potentials. Proceedings of an international workshop organized in the framework of The Netherlands National Research Programme on Global Air Pollution and Climate Change. WISE Report 2, ISRIC, Wageningen. 122 pp. Batjes, N.H. and E.M. Bridges, 1993. World Inventory of Soil Emission Potentials: Development of a Global Soil Database. In: Report of the Third Workshop of the Global Emissions Inventory Activity (GEIA), A.F. Bouwman (ed). National Institute of Public Health and Environmental Protection (RIVM), Bilthoven. report 481507002, p. 51-55. Batjes, N.H., E.M. Bridges and F.O. Nachtergaele, 1994. World Inventory of Soil Emission Potentials: Development of a global soil data base of process controlling factors. Working Paper and Preprint 94/01, ISRIC, Wageningen. In: Proceedings of International Symposium on Rice and Climate Changes International Rice Research Institute, Los Banos, 14 -18 March 1994, (in press). Batjes, N.H., V.W.P. van Engelen, J.H. Kauffman and L.R. Oldeman, 1995. Development of global databases for environmental research. Transactions of 15th ISSS-Congress, Acapulco (in press). Bender, M. and R. Conrad, 1992. Kinetics of CH4 oxidation in oxic soils exposed to ambient air of high CH4 mixing rations. FEMS Microbio. Ecol. 101:261 - 270. Berdowski, J.J.M., 1994. Emission Inventory in the Netherlands. Emissions to air and water in 1990. Summary, Publikatiereeks Emissieregistratie, VROM, Den Haag, nr. 16. Berdowski, J.J.M., L. Beck, S. Piccot, J.G.J. Olivier, and C. Veldt, 1993. Working Group Report. Methane emissions from fuel combustion and industrial processes. R.A. van Amstel (ed.). In: Proceedings of the International Workshop Methane and Nitrous Oxide: Methods in National Emission Inventories and Options for Control. Amersfoort, The Netherlands February 3-5, 1993, p. 231-237. Berdowski, J.J.M., J.G.J. Olivier, and C. Veldt, 1993. Methane emissions from fuel combustion and industrial processes.R.A, van Amstel (ed.). In: Proceedings of the International Workshop Methane and Nitrous Oxide: Methods in National Emission Inventories and Options for Control. Amersfoort, February 3 - 5, 1993, p. 131-141. Berdowski, J.J.M. and C. Veldt, 1992. Factors contributing to uncertainty in biogenic NMHC emission estimates. Report 92/022, TNO-MW, Delft. Bosinger, M.T.R., M. Schmidt and I. Levin, 1993. The regional budget of atmospheric methane of a highly populated area. Chemosph. 26:143 - 160. Bouwman, A.F., 1993. The global source distribution of nitrous oxide. R.A. van Amstel (ed). In: Proceeding of the international workshop Methane and nitrous oxide. Methods in national emissions inventories and options for control. Amersfoort, 3-5 February 1993, Report 481507003, National Institute of Public Health and Environmental Protection (RIVM), Bilthoven, The Netherlands, p. 261-272. Bouwman, A.F. (ed.), 1993. Report of the third workshop of the Global Emissions Inventory Activity (GEIA). report 481507002, National Institute of Public Health and Environmental Protection (RIVM), Bilthoven.
526 Bouwman, A.F., 1994. Estimated global distribution of emissions of nitrous oxide. In: CH4 and N20: Global emissions and controls from rice fields and other agricultural and industrial sources. K. Minami, A.R. Mosier and R. Sass (eds.) NIAES series 2, National Institute of Agro-Environmental Sciences, Tsukuba, Japan, p. 147-159. Bouwman, A.F., I. Fung, E. Matthews and J. John, 1993. Global analysis of the potential for N20 production in natural soils. Global Biogeochem. Cycles 7: 557-597. Bouwman, A.F. en K.W. van der Hoek, 1991. Analysis of soil and Water Borne emissions of N20 and CH4 in The Netherlands. report 76301010,National Institute of Public Health and Environmental Protection (RIVM), Bilthoven. Bouwman, A.F., K.W. van der Hoek and J.G.J. Olivier, 1994. Uncertainty in the global source distribution of nitrous oxide. Submitted to Journal of Geophysical Research. Bouwman, A.F., G.J. van den Born and R.J. Swart, 1992. Land-use related sources of CO2, CH4 and N20. Report 222901004, National Institute of Public Health and Environmental Protection (RIVM), Bilthoven. Bremer E. and P.J. Kuikman, 1994. Microbial utilization of 14C[U] glucose in soil is affected by the amount and timing of glucose additions. Soil Biol. Biochem. 26: 511-518. Brumme, R. and F. Beese, 1992. Effects of liming and nitrogen fertilisation on emissions of CO2 and N20 from a temperate forest. J. Geophys. Res. 97: 12851 - 12858. Butler, J.H., J.W. Elkins and T.M. Thompson, 1989. Tropospheric and dissolved N20 of the West Pacific and East Indian Oceans during the E1 Nifio Southern Oscillation event of 1987. J. Geophys. Res. 94:14865 - 14877. Calvert, J.C., 1994. Tropospheric CH4: uncertainties in its chemical lose rate. J. Slanina, P. Warneck, N.M. Bazhin, H. Akimoto and W.M. Kieskamp (eds). In: Assessment of uncertainties in the projected concentration of methane in the atmosphere. Proceedings of the IUPAC/IIASA workshop, Moscow, Russia, July 20-21, 1992. Pure & Appl. Chem. 66:137 - 200. Conrad, R., F. Bak, H.J. Seitz, B. Thebrath, H.P. Mayer and H. Schfitz, 1989. Hydrogen turnover by psychrotrophic homoacetogenic and mesophilic methanogenic bacteria in anoxic paddy soil and lake sediment. FEMS Microbio. Ecol. 62: 285- 294. Denier van der Gon, H.A.C., 1993. Methane emission from rice agriculture. Change, 15: 12-14. Denier van der Gon, H.A.C., 1994. Influence of Organic Matter Incorporation on the Methane Emission from a Wetland Rice Field, paper presented at EC-China Workshop on "Current status and Perspectives of Quantification of Methane and Nitrous Oxide Emissions from Agricultural Ecosystems in China", Garmisch-Partenkirchen, Germany, February 21-25, 1994. Denier van der Gon, H.A.C and N. van Breemen, 1993. Diffusion-Controlled Transport of Methane from Soil to Atmosphere as Mediated by Rice Plants. Biogeochem. 21: 177-190. Denier van der Gon, H.A.C. and H.-U. Neue, 1994. Impact of gypsum application on the methane emission from a wetland rice field. Global Biogeochem. Cycles, 8: 127-134.
527 Denier van der Gon, H.A.C. and H.U. Neue, 1995. CH4 emission from a wetland rice field -Impact of gypsum application. Poster presented at International Symposium Climate Change and Rice 14-18 March 1994, Los Banos, The Philippines. International Rice Research Notes (in press). Denier van der Gon, H.A.C. and H.U. Neue 1995. Influence of organic m a t t e r incorporation on the methane emission from a wetland rice field. Global Biogeochem. Cycles (in press). Denier van der Gon, H.A.C. and H..U. Neue, 1995. Methane emission from a wetland rice field as affected by salinity. Plant and Soil (in press). Denier van der Gon, H.A.C., H.-U. Neue, R.S. Lantin, R. Wassman, M.C.R. Alberto, J.B. Aduna and M.J.P. Tan, 1992. Controlling factors of methane emissions from rice fields. In: World Inventory of Soil Emission Potentials. Proceedings of an International Workshop, ISRIC, Wageningen. Diederen, H.S.M.A., 1992. Programming study methane research. TNO-MW, Delft, Report 92/353. Dirks, B.O.M., 1993. Diurnal and seasonal carbon dioxide exchange and its components in temperate grasslands in the Netherlands- An outline of the methodology. Water Air Soil Poll. 70, 1-4: 425-430. Dirks, B.O.M and J. Goudriaan 1994. Diurnal and seasonal CO2 fluxes between g r a s s l a n d ecosystems and the atmospheric b o u n d a r y layer in the Netherlands. LUW-TPE, Wageningen. Dlugokencky, E.J., K.A. Masarie, P.M. Lang, P.P. Tans, L.P. Steele and E.G. Nisbet, 1992. A dramatic decrease in the growth rate of atmospheric methane in the northern hemisphere during 1992. Geophys. Res. Lett. Dlugokencky, E.J., P.L. Steele, P.M. Lang and K.A. Masarie, The growth rate and distribution of atmospheric methane. J. Geophys. Res. submitted. Eisma, R., A. Vermeulen, W.M. Kieskamp, 1994. Determination of European m e t h a n e emissions using concentration and isotope m e a s u r e m e n t s . Proceedings of the International Symposium Non-CO2 Greenhouse Gases. Why and how to control?., J. v. Ham, L.J.H.M. Janssen and R.J. Swart (eds). Maastricht. 13 - 15 December, 1993. Environ. Monit. Assess. 31: 197-202. Eisma, R., A. Vermeulen, W.M. Kieskamp, G.P. Wyers and A.C. Veltkamp, 1995. Methane Emissions in North-West Europe. ECN-report (in preparation), ECN, Petten. Eisma, R., A. Vermeulen, W.M. Kieskamp, G.P. Wyers and A.C. Veltkamp, 1995, (in prep.). Methane emissions in North-West Europe. ECN, Petten. Eisma, R., A. Vermeulen and K. van der Borg, (submitted for publication in RadioCarbon). Fowler, D. and J.H. Duyzer, 1989. Micrometeorological techniques for the measurement of trace gas exchange. M.O. Andreae and D.S. Schimel (eds). In: Exchange of trace gases between terrestrial ecosystems and the atmosphere. John Wiley & Sons, p. 189-207. Goudriaan, J., 1989. Modelling biospheric control of carbon fluxes between atmosphere, ocean and land in view of climatic change. A. Berger, S. Schneider and J. C.I. Duplessy (eds). In: Climate and Geo-Sciences. NATO-ASI Series C, Vol. 285, Kluwer Academic Publishers Dordrecht. Graedel, T.E., T.S. Bates, A.F. Bouwman, D. Cunnold, J. Dignon, I. Fung, D.J. Jacob, B.K. Lamb, J.A. Logan, G. Marland, P. Middleton, J.M. Pacyna, M. Placet and C. Veldt, 1993. A compilation of inventories of emissions. Global Biogeochem. Cycles 7: 1-26.
528 Guenther, A., C. Hewit, G. Allwine, J.J.M. Berdowski, and C. Veldt, 1993. Global Natural VOC Emissions: An IGAC-GEIA Inventory and Assessment of Uncertainties. 37th OHOLO Conference, Global Atmospheric-Biospheric Chemistry: The First IGAC Scientific Conference, Eilat, Israel, April 1993. Hensen, A., A.T. Vermeulen, W.C.M. vd Bulk, W.M. Kieskamp and G.P. Wyers, 1995. CO2 Emissions in North-West Europe. ECN-report ECN, Petten, (in prep.). Hollander, J.C.Th., 1979. Measurement data IMG-TNO. Unpublished results, TNO, Delft. Hutchinson, G.L., and A.R. Mosier, 1981. Improved soil cover method for field measurement of nitrous oxide fluxes. Soil Sci. Soc. Am. J. 45: 311-316. IPCC, 1990. Climate Change: The IPCC Scientific Assessment. J.T. Houghton, G.J. Jenkins, J.J. Ephraums, (eds), Cambridge University Press, Cambridge. IPCC, 1991. Estimation of greenhouse gas emissions and sinks. Final report from OECD experts meeting, 18-21 F e b r u a r y 1991. P r e p a r e d for Intergovernmental Panel on Climate Change, OECD, Paris. IPCC, 1992. Climate change 1992. The supplementary report to the IPCC scientific assessment. J.T. Houghton, B.A. Callander and S.K. Varney (eds). University Press, Cambridge. Jarvis, S.C., D.R. Lockyer, G. Warren, D.J. Hatch and G. Dollard, 1994. Preliminary studies of the exchanges of methane between grassland and the atmosphere. L. 't Mannetje and J. Frame (eds.). In: Grassland and Society. Proceedings of the 15th General Meeting of the European Grassland Federation, p. 408-412. Jones, H.A., and D.B. Nedwell, 1993. Methane emission and methane oxidation in land-fill cover soil. FEMS Microbiol. Ecol. 102: 185-195. Khalil, M.A.K. and R.A. Rasmussen. a. Trends of atmospheric methane. In: Assessment of uncertainties in the projected concentration of methane in the atmosphere. Khalil, M.A.K. and R.A.Rasmussen, 1990b. Atmospheric methane: recent global trends. Environ. Sci. Technol. 24: 549-553. Khalil, M.A.K. and R.A. Rasmussen, 1992c. The global sources of nitrous oxide. J. Geophys. Res. 97: 14,651-14,660. Khalil, M.A.K. and R.A. Rasmussen, 1993d. Decreasing trend of methane: unpredictability of future concentrations. Chemosph. 26: 803-814. Khalil, M.A.K., R.A. Rasmussen and F. Moraes, 1993e. Atmospheric methane at Cape Meares: Analysis of a high-resolution data base and its environmental implications. J. Geophys. Res. p. 14,753-14,770. Khalil, M.A.K. and R.A. Rasmussen, 1994. Trends of atmospheric methane. J. Slanina, P. Warneck, N.M. Bazhin, H. Akimoto and W.M. Kieskamp (eds). In: Assessment of uncertainties in the projected concentration of methane in the atmosphere. Proceedings of the IUPAC/IIASA workshop, Moscow, Russia, July 20-21, 1992. Pure & Appl. Chem. 66: 137-200. Kieskamp, W.M., Hensen, A., Vermeulen. A.T. and Slanina, J., 1994. The atmospheric CO2 programme of the Netherlands Energy Research Foundation (ECN), The Netherlands. G.I. Pearman and J.T. Peterson (eds). In: Report of the seventh WMO meeting of experts on carbon dioxide concentration and isotopic measurement techniques, Rome, Italy, 7 - 10 September 1993,WMO, Geneva, Switzerland.
529 Kieskamp, W.M., L. Lohse, E. Epping and W. Helder, 1991. Seasonal variation in denitrification rates and nitrous oxide fluxes in intertidal sediments of the Western Waddenzee. Mar. Ecol. Prog. Ser. 72: 145-151. Kieskamp, W.M., A.T. Vermeulen and A.C. Veltkamp, 1995. Quantification of methane emission of an urban area by atmospheric measurements. ECN, Petten. (in prep.). King, G.M., 1993. Ecophysiological characteristics of obligate methano-trophic bacteria and m e t h a n e oxidation in situ. In: Microbial growth on C-1 compounds J.C. Murrell and Kelly (eds.) Proceedings of the 7th symposium in Coventry, 1992. Intercept Ltd., Andover. Kotsyurbenko, O.R., A.N. Nozhevenikova and G.A. Zavarzin, 1993. Methanogenic degradation of organic matter by anaerobic bacteria at low temperature. Chemosph. 27: 1745-1761. Kroeze, C., 1994. Background document nitrous oxide. National Institute of Public Health and Environmental Protection (RIVM), Bilthoven (in prep.) K u i k m a n P.J. and A. Gorissen, 1993. Carbon fluxes and Organic M a t t e r Transformations in Plant-Soil-Systems. S.C. van de Geijn, J. Goudriaan and F. Berendse (eds). In: Climate change: crops and terrestrial ecosystems. Agrobiologische Thema's 9, CABO-DLO, Wageningen. p. 97-108. Kfisel, K., K. and H.L. Drake, 1994. Acetate synthesis in soil from a Bavarian Beech forest. Appl. Environ. Microbiol. 60: 1370-1373. Langeveld, C.A. and J.E. Hofman, 1994. Nitrous oxide production and consumption in peat soils. In: Proceedings of the International Symposium Non-CO2 Greenhouse Gases. Why and how to control?. J.v. Ham, L.J.H.M. Jansen and R.J. Swart (eds). Maastricht. 13 - 15 December 1993. Environm. Monit. Assess. Leffelaar, P.A. and A.J.M. Stams, 1991. Raamwerk en Achtergronden voor methaanemissie onderzoek in bodemecosystemen t.b.v, de Ad-Hoc commissie van de programmeringsgroep methaan. (in Dutch). Lofstom, P., and H.R. Olesen, 1992. User's guide for OML-MULTI, An air pollution model for multiple point and area sources. N.E.R.I., Roskilde, Denmark. MS-LUFT-A12. Maas, C.W.M. van der, J.J.M. Berdowski, A.F. Bouwman and J.G.J. Olivier, 1994. Results Information Analysis EDGAR, Emissions Database for Global Atmospheric Research. RIVMfrNO report 776 010 001 (in prep.). Magnusson, T, 1993. Carbon dioxide and methane formation in forest mineral and peat soils during aerobic and anaerobic incubations. Soil Biol. Biochem. 25: 877-883. Mosier, A., D. Schimmel, D. Valentine, K. Bronson and W. Parton 1991. Methane and nitrous oxide fluxes in native, fertilized and. Muzio, L.J., and J.C. Kramlich, 1988. An artefact in the measurement of N20 of combustion sources. Geophys. Res. Lett. 15: 1369-1372. Nabuurs, G.J. and G.M.J. Mohren, 1993. Carbon in Dutch forest ecosystems. N. J. Agricul. Sci. 41: 309- 326. Nabuurs, G.J. and G.M.J. Mohren, 1993. Carbon stocks and fluxes in Dutch forest ecosystems. IBN-DLO, Wageningen. report 93/1, 85 pp. Nielen, R.J. 1991. Kwantificering van m e t h a a n e m i s s i e s als gevolg van aardgasverliezen en oliewinning in Nederland (Quantification of methane emissions caused by natural gas losses and petroleum exploitation in The Netherlands, (in Dutch), Report 91-210, TNO-ME, Apeldoorn.
530 Olivier, J.G.J., 1993. Nitrous oxide emissions from industrial processes. R.A. van Amstel (ed.). In: Proceedings of the International Workshop Methane and Nitrous Oxide: Methods in National Emission Inventories and Options for Control. Amersfoort. February 3-5, 1993, p. 339-341. Olivier, J.G.J., 1993. Working Group Report. Nitrous Oxide emissions from fuel Combustion and Industrial Processes. A Draft Methodology to Estimate National Inventories. R.A. van Amstel (ed.). In: Proceedings of the International Workshop Methane and Nitrous Oxide: Methods in National Emission Inventories and Options for Control, Amersfoort. February 3-5, 1993, p. 347-361. Olivier, J.G.J. and J.J.M. Berdowski 1993. EDGAR/IMAGE: Database and integrated scenario model to assess the impact of greenhouse gas emissions, Paper presented at the Advisory Group Meeting on Assessment of Databases on Energy Demand and Supply in Terms of their Adequacy for Use in Studies of Greenhouse Gas Emissions, IAEA Headquarters, Vienna, Austria, April 5-7, 1993. (623-I1-AG-832). Olivier, J.G.J., A.F. Bouwman, C.W.M. van der Maas and J.J.M. Berdowski, 1994. Emission d a t a b a s e for Global Atmospheric Research (EDGAR). In: Proceedings of the International Symposium Non-CO2 Greenhouse Gases. Why and how to control?, J.v. Ham, L.J.H.M. Jansen and R.J. Swart (eds). Maastricht, The Netherlands, 1 3 - 15 December 1993. Environm. Monit. Assess., p. 93-106. Oonk J., A. Weenk, O. Coops and L. Luning, 1994. Validation of landfill gas formation models. Report 94/315, TNO-ME/Grontmij, Apeldoorn. Oonk J., A. Boom, and J.H. Duyzer, 1995. Landfill gas formation, emission and recovery in the Netherlands, a measurement study. TNO-ME, Apeldoorn. (in prep.). Oonk J., 1993. Landfill gas formation, emission and recovery. : In: Proceedings of the International Symposium Non-CO2 Greenhouse Gases. Why and how to control? J.v. Ham, L.J.H.M. Jansen and R.J. Swart (eds). Maastricht. 13 - 15 December, 1993. Environm. Monit. Assess. p. 323-328. Oonk, J. ,1994. Landfill gas formation, emission and recovery. J.v. Ham, L.J.H.M. Jansen and R.J. Swart (eds). In: Proceedings of the International Symposium Non-CO2 Greenhouse Gases. Why and how to control?, Maastricht, 13 - 15 December 1993. Environm. Monit. Assess. p. 323-328. Oonk, J., A. Boom and J.H. Duyzer, 1995. Landfill gas formation, emission and recovery in the Netherlands, a measurement study. (In prep.), TNO-ME, Apeldoorn. Oonk, J., 1993. Stortgaswinning met hoog rendement (High efficiency landfill gas recovery, in Dutch). Report no. 93-248. TNO-ME, Apeldoorn. Oonk J., 1993. High efficiency landfill gas recovery (in Dutch). TNO-ME, Apeldoorn. report 93/248. Oonk, J. and M.E.J.P. Vosbeek 1995. Methane emissions due to exploration and production of natural gas and oil in The Netherlands. TNO-ME/KEMA- report (in press). Reiners, W.A., A.F. Bouwman, W.F.J. Parson and M. Keller, 1992. Tropical rain forest conversion: changes in vegetation and soil properties. Ecological Applications (in press).
531 Reinhart, D.R., and D.C. Cooper, 1992. Flux chamber design and operation for the measurement of municipal solid waste landfill gas emission rates. J. Air Waste Manag. Assoc. 42: 1,067-1,070. Schfitz, H., R. Holzapfel-Pschorn, R. Conrad, H. Rennenberg and W. Seiler, 1989. A 3-year continuous record on the influence of daytime, season and fertilizer treatment on methane emission rates from an Italian rice paddy. J. Geoph. Res. 94: 16,405- 16,416. Simmonds, P.G., D.M. Cunnold et al., 1993. Evidence of the phase-out of CFC use in Europe over the period 1987-1990. Atmosph. Environ. 27A: 1397 - 1407. Simmonds, P.G. and R.G.Derwent, 1991. Measurement of ozone and other radiatively active gases at Mace Head in the republic of Ireland. Atmosph. Environ. 25A: 1795-1808. Slanina, J. and P. Okken (eds), 1991. Assessment of uncertainties in the projected concentrations of carbon dioxide in the atmosphere. Proceedings of the IUPAC workshop, Petten, July 4-6, 1990. Pure & Appl. Chem. 63: 763-796. Slanina, J., P. Warneck, N.M. Bazhin, H. Akimoto and W.M. Kieskamp (eds), 1994. Assessment of uncertainties in the projected concentration of methane in the atmosphere. Proceedings of the IUPAC/IIASA workshop, Moscow, Russia, July 20-21, 1992. Pure & Appl. Chem. 66: 137-200. Smith, T.M., R. Leemans and H.H. Shugart, 1992. Sensitivity of terrestrial carbon storage to CO2 induced climate change: Comparison of four scenarios based on general circulation model. Clim. Change, 21: 367-384. Spoelstra, H., 1992. N20 emissions from combustion processes used in the generation of electricity. KEMA report, 10143-KES/MME 92-4029. Spoelstra, H., 1993. N20 emissions from combustion processes used in the generation of electricity. KEMA, Arnhem, report 10143-KES/MME 92-4029, 27 pp. Steele, L.P., E.J. Dlugokencky, P.M. Lang, P.P. Tans, R.C. Martine and K.A. Masarie, 1992. Slowing down of the global accumulation of atmospheric methane during the 1980s. Nature, 359: 313-316. Swart, R.J., A.F. Bouwman, J. Olivier and G.J. van den Born, 1993. Inventory of greenhouse gas emissions in the Netherlands. Ambio 22" 518-523. Swinnen J., 1994. Evaluation of the use of a model rhizodeposition technique to separate root and microbial respiration in soil. Plant and soil (in press). Swinnen J., 1994. Production and turnover of root-derived organic matter in the rhizosphere of wheat and barley under field conditions. Ph-D thesis no. 247, Catholic University Leuven, Belgium, 126 pp. Tans, P.P., I.Y. Fung and T. Takahashi, 1990. Observational constraints on the global atmospheric CO2 budget. Science, 247: 1431-1438. Tempel, P., 1994. Automated Soil Data Transfer Facility Between Disparate Soil Databases. Working Paper and Preprint (in press), ISRIC, Wageningen. Thebrath, B. H.P. Mayer and R. Conrad, 1992. Bicarbonate-dependent production and methanogenic consumption of acetate in anoxic paddy soil. FEMS Microbiol. Ecol. 86: 295-302. UK-DoE., 1993. An assessment of methane emissions from UK landfills. UK. DoE-report CWM 063/93. US-EPA, 1993. Anthropogenic methane emissions in the Unites States: estimates for 1990. Report to Congress EPA-430-R-93-003.
532 Van Amstel, A.R. (ed.), 1993. International IPCC Workshop on Methane and Nitrous Oxide: Methods in National Emissions Inventories and Options for Control. Report 481507003, National Institute of Public Health and Environmental Protection (RIVM), Bilthoven. Van Amstel, A.R., Swart, R.J., Krol, M.S., Beck, J.P., Bouwman, A.F. and K.W. van der Hoek, 1993. Methane the other greenhouse gas, Research and policy in The Netherlands. Report no.481507001, RIVM, Bilthoven. Van den Born, G.J., A.F. Bouwman and J.G.J. Olivier, 1991. The emission of greenhouse gases in The Netherlands. Report 222901003, RIVM, Bilthoven. Van Faassen, H.G., 1993. Modelling N20 emission from (grazed) grassland, a literature review. Nota 269, IB-DLO, Haren. 19 pp. Van Faassen, H.G., 1994. Dinitrogen oxide (N20) emission from grassland soil columns after application of artificial cow urine. Submitted to Netherlands Journal of Agricultural Science. Veenhuysen, D., and P. Hofschreuder, 1994. Methane emission of the Amsterdam urban area. Report R 669, LUW, Wageningen. Veldt, C.1992. GEIA note on residential biomass burning. A short communication about emission factors for the residential combustion of biomass fuels, November 1992. Veldt, C., 1992. Anthropogenic volatile organic compounds. Global Emission Inventory Status Report, December 1992. Veldt, C., 1993. VOC - A little learning is a dangerous thing. In: Proceedings of TNO/EURASAP workshop on the reliability of VOC emission databases. Delft. 9 -10 June 1993. Veldt, C. and J.J.M. Berdowski 1993. IGAC-GEIA anthropogenic VOC inventory. March, 1993, WO-563. Veldt, C., and P.F.J. van der Most, 1993. Emissiefactoren vluchtige organische stoffen uit verbrandingsmotoren (in Dutch) Publikatiereeks Emissieregistratie Nr. 10, VROM, Den Haag. Velthof, G.L., 1993. Lachgasemissie uit grasland (in Dutch). Praktijkonderzoek, 93-1: 55-56. Velthof, G.L., and O. Oenema, 1994. Nitrous oxide emission from grasslands on sand, clay and peat soils in The Netherlands. In: Proceedings of the International Symposium Non-CO2 Greenhouse Gases. Why and how to control?, J.v. Ham, L.J.H.M. Jansen and R.J. Swart (eds). Maastricht, 13 - 15 December 1993. Environm. Monit. Assess. p. 439-445. Velthof, G.L. and O. Oenema, 1995. Effects of nitrogen fertilizer type and urine on nitrous oxide flux from grassland in early spring. In: Proceedings of the 15th General Meeting of the European Grassland Federation, June 1994, Wageningen, The Netherlands (in press). Verberne, E., 1992. Simulation of the nitrogen and water balance in systems of grassland and soil. Nota 258: 56. Verschut, C., J. Oonk and W. Mulder, 1991. Broeikasgassen uit vuilstorts in Nederland (Greenhouse gases from landfills in the Netherlands, (in Dutch). Report 91-444, TNO-ME, Apeldoorn. Vosbeek, M.E.J.P., 1993. Evaluation and integration of CH4-, CO-, and CO2m e a s u r e m e n t s in Arnhem and Kollumerwaard (in Dutch). Report 63625-KES/MLU 93-32412, KEMA, Arnhem. Vosbeek, M.E.J.P., 1993. Quantification of methane emissions from oil and gas exploitation (in Dutch). Report 63630-KES/MLU 93-3241, KEMA, Arnhem.
533 Vosbeek, M.E.J.P., 1993. Methane emissions in the Netherlands (in Dutch). Report 63637-KES/MLU 94-3227.KEMA, Arnhem. Vosbeek, M.E.J.P., 1993a. Evaluatie en interpretatie van CH4-, CO-, CO2 metingen te Arnhem en Kollumerwaard (Evaluation and Interpretation of CH4-, CO-, CO2 measuraments at Arhem and Kollumerwaar, in dutch). Report 63625-KES/MLU 93-32242, Kema, Arnhem. Vosbeek, M.E.J.P., 1993b. Kwantificering van methaanemissies bij olie- en gaswinning (Quantification of methane emissions from oil and gas exploitation, in Dutch). Report no. 63630-KES 93-3241, KEMA, Arnhem. Vosbeek, M.E.J.P., 1994. Methane emissions due to the production of oil and natural gas in the Netherlands (in Dutch). Report 63630-KES 94-3223, KEMA, Arnhem. Wassmann, R., H. U. Neue, R.S. Lantin, J.B. Aduna, M.C.R. Alberto, M.J. Andales, M.J.P. Tan, H.A.C. Denier van der Gon, H. Hoffmann, H. Papen, H. Rennenberg and W. Seiler, 1994. Temporal patterns of methane emissions from wetland rice fields treated by different modes of N-application. J. Geoph. Res. 99:16457-16462 Weis, R.F. (in prep.). Whalen, S.C., and W.S. Reeburgh, 1990. Consumption of atmospheric methane by tundra soil. Nature, 346: 160-162. Wolff, J., Inventarisatie van niet-fosiele koolstofstromen en voorraden in terrestrische systemen in Nederland. LUW rap. (in Dutch).
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
535
Discussion on the NRP assessment report "Greenhouse gases" J. Goudriaan, chairman and A.F. Bouwman, rapporteur I n t r o d u c t i o n (J. Slanina) The u n c e r t a i n t i e s in the global C02 budget were discussed, in p a r t i c u l a r the relation between the fluxes of CO2 between the various compartments, i.e. oceans, atmosphere and terrestrial biosphere. Reduction of the CO2 emissions from fossil fuel combustion is necessary to stabilize the atmospheric CO2 concentration. The percentage reduction of the current emission depends on the amount of CO2 t h a t is t a k e n up by the world's oceans and the terrestrial biosphere (the CO2 fertilization effect). The uncertainties in these sinks is of great importance for future mitigation policies. The major global sources and atmospheric destruction of CH4 are still very uncertain. Regarding the emission estimates, the major source of uncertainty is in the extrapolation of m e a s u r e m e n t s to larger scales. In the past generally point m e a s u r e m e n t s made at one or a few sites were used to calculate the regional, continental or global emission. Recent extrapolation techniques develop towards validation of flux measurements at regional scales to arrive at more reliable source estimates. The most recent IPCC global N 2 0 budget shows t h a t the u n c e r t a i n t y in the various source estimates is very uncertain, and there may yet be unidentified sources. The trends in the concentrations of CO2, CH4 and N 2 0 have changed in recent years. The global increase of the mean CO2 and CH4 concentrations were lower in the early 1990s t h a n in the 1980s, and the trend in N20 in 1992 was about half t h a t in the 1980s. For N 2 0 the global observed t e m p e r a t u r e decrease caused by aerosols from the eruption of Mount Pinatubo m a y have lead to a decrease in biogenic emissions from soils and oceans. For the other gases it is more difficult to explain the observed changes in atmospheric growth rates. The major criteria used in the programming of the three clusters were: -
The relative importance of a source of emissions, both now and in the future The range of uncertainty in the available estimates (in the Dutch inventory, by Van den Born et al., 1991) Availability of techniques or policies to reduce emissions Availability of specific experience in The Netherlands.
Theme "Greenhouse Gases" of the Dutch National Research Programme on Global Air Pn|lutinn ~nd Climate Chan~e consisted of five clusters. Three clusters
536 concentrated on fluxes of C02, CH4 and N20. The two other clusters, d a t a b a s e development and social causes of climate change, were not discussed during this session. An overview of the CO2, CH4 and N20 clusters was given during this session. The individual projects, institutes and project leaders are listed in the assessment report of Theme B (Berdowski et al., in prep.), and will not be given in this report. CO2 cluster ( J . Slanina) The research in the C02 cluster concentrated on the exchange of CO2 between terrestrial ecosystems and the atmosphere, with a focus on grasslands. Grassland cover about one-third of the Netherlands. Moreover, grasslands have received much less research attention regarding CO2 exchange than e.g. forests. The
Research topics The research topics in the CO2 cluster included: -
-
-
Inventory of carbon stocks in Dutch forests The mechanism of soil carbon (C) accumulation in sand, clay and peat soils, u s i n g labelled 14C02 as a tracer to study the distribution of carbon in vegetation and soil. Development of models describing the exchange of CO2 Validation of models at different scales, on the basis of aircraft m e a s u r e m e n t s u s i n g eddy correlation t e c h n i q u e s , towers ( g r a d i e n t t e c h n i q u e ) a n d concentration m e a s u r e m e n t s and determination of the isotopic composition of atmospheric CO2. Integration of results.
Results An inventory was made of the current stocks of C in living biomass and soils for Dutch forests. One of the conclusions was t h a t Dutch forests are currently taking up 0.33x1012 g C yr-1. The mechanistic research showed t h a t retention of C in clay and sand soils was higher t h a n in peat soils. The difference between m e a s u r e d and modelled CO2 exchange between the grassland and the atmosphere was attributed to oxidation of the peat. The CO2 fluxes at the scale of a hectare were estimated with the gradient technique at Cabauw. Net uptake occurred in March-May, while a net emission to the atmosphere occurred in all other months. The estimated annual net emission was 3000 kg C ha-1. Most of this is caused by peat oxidation. Aircraft m e a s u r e m e n t s are not applicable to Dutch conditions. The heterogeneity of the landscape makes it very difficult to attribute fluxes to a certain area or component of the landscape. It was noted t h a t in places with vast areas with h o m o g e n e o u s l a n d s c a p e and vegetation, a i r c r a f t m e a s u r e m e n t s m a y be successful, as shown in e.g. the U.S.A. Low ground water tables cause oxidation of the peat material. The total historic Closs resulting from peat oxidation in The Netherlands is of the same order of
537 magnitude as the total historic C-injection into the atmosphere by Dutch fossil fuel combustion. By implementing higher water tables in the Dutch peat soils the loss of C can be halted and soils may even become a net sink by C accumulation. However, as was noted in the CH4 and N20 clusters, changes in the w a t e r table has i m p o r t a n t consequences for fluxes of all trace gases, where fluxes of one gas may decrease while the emission of another gas increases. The
CH4
cluster
(J. Berdowski)
Research topics The p r o g r a m m i n g of the CH4 cluster was based on the results of a Dutch inventory of sources (Van den Born et al., 1991). The research on CH4 sources concentrated on the following sources: -
-
Rice fields Grasslands on organic soils (integrated CH4 grassland project) Landfills Exploration of oil and gas Evaluation and validation
Not all the projects have been completed. The results of the different projects were briefly discussed.
Results Research on CH4 from rice fields paddies carried out in the Philippines has concentrated on soil factors influencing fluxes of CH4 from irrigated rice fields. The effect of additions of gypsum was given special attention. The general conclusion was t h a t soil p a r a m e t e r s need to be included in global estimates of CH4 from rice paddies. The sulphate added in the form of gypsum causes a reduction of CH4 emission, caused by competition between s u l p h a t e reducing b a c t e r i a and methanogens. The integrated CH4 grassland project consisted of different projects, investigating CH4 formation, CH4 oxidation, modelling of CH4 fluxes and effects of soil and water m a n a g e m e n t on CH4 fluxes. The conclusion drawn on the basis of the results of this project was t h a t CH4 emission from intensively managed grasslands on peat soils is lower t h a n for extensively managed systems. The major cause of lower fluxes in intensively managed systems is the lower ground water table. This causes a more extensive oxic layer in the peat surface soil, with higher oxidation of CH4 than in soils with high ground water tables. The major sources of m e t h a n e from oil and gas exploration in The N e t h e r l a n d s include the venting of gas, chronic leaks during collection and transport. Testing of wells is probably a moderate to major source in The Netherlands. The evaluation and validation studies included p e r m a n e n t m e a s u r e m e n t of the atmospheric concentration of CH4, modelling of air trajectories and estimation of urban sources of CH4. P e r m a n e n t ground based (20 m) measurements are made in
538 Delft, Arnhem, K o l u m m e r w a a r d and Amsterdam. M e a s u r e m e n t s at 200m are done at Cabauw. The m e a s u r e m e n t s show a background concentration of about 2 ppm, b u t very high peaks occur occasionally. In the N o r t h e r n p a r t of The N e t h e r l a n d s these peaks are associated with fluxes from gas fields d u r i n g exploration. Observations of peak CH4 concentrations correspond to SE winds, indicating a source SE of The Netherlands. The modelling of air trajectories and backcalculation of emissions resulted in estimates t h a t are consistent with the inventory of CH4 emissions. T h e N 2 0 c l u s t e r (A.F. B o u w m a n )
Research topics For the N 2 0 cluster two preparatory studies were carried out, i.e. an inventory of Dutch N 2 0 emissions and a comparison of m e a s u r e m e n t s techniques. Based on these studies, the following sources were selected for further research: -
-
Mobile and stationary combustion Aquatic sources Sewage water t r e a t m e n t plants Integrated N20 grassland project Global N20 inventory and validation
The results of the N20 cluster were used to compile a new inventory and scenarios of Dutch emissions in the Background document on N 2 0 (Kroeze, 1994) (part of the cluster on Social Causes).
Results Various types of vehicles and engines with/without catalysts were included in the study on mobile N20 sources. Cars equipped with catalysts emit more N 2 0 t h a n cars without catalysts. The N20 emission increases along with the age of the catalyst. At p r e s e n t mobile combustion contributes about 10% to D u t c h emissions. With increasing automobility and f u r t h e r p e n e t r a t i o n of c a t a l y s t equipped cars, the importance of mobile combustion will increase in the future. M e a s u r e m e n t s carried out in Dutch power plants, chemical industries, an oil refinery, waste incineration plant indicate t h a t stationary combustion is a minor direct source in The Netherlands. Indirect emissions from soils induced by NOx emission and deposition may be more important. M e a s u r e m e n t s in the Indian Ocean indicated very high fluxes from upwelling zones. The calculated fluxes from the North Sea were much lower. In the Scheldt estuary a close relation was found between N concentrations in the water and N 2 0 fluxes. Two different sewage water treatment plants were studied, representing the major types of i n s t a l l a t i o n s in the N e t h e r l a n d s . L i t e r a t u r e r e s e a r c h and m e a s u r e m e n t s indicate t h a t N 2 0 fluxes from these t r e a t m e n t installations are much lower than assumed previously (Van den Born et al., 1991). The integrated grassland N20 project included flux measurements, modelling and studies on the effects of management, and focused on grasslands on peat soils. The
539 results indicate that intensively managed grasslands on peat soils show much higher N 2 0 emissions than mineral soils. As was found in the preparatory inventory, grasslands form the major individual source of N 20 in The Netherlands. The N20 emission associated with food production makes up about 50% of Dutch emissions. Reduction of emissions from soils is possible by implementing higher ground water tables and by reducing N inputs. However, this will induce higher CH4 emissions. The overall result of extensification of grasslands on peat soils based on Global Warming Potentials (GWP) has not been estimated yet. The global inventory indicates t h a t 2/3 of the global increase of atmospheric N 2 0 is associated with food production, with an important part from animal production. It was noted that at the Kolummerwaard station the N20 concentrations are fairly constant. This is seen in all monitoring stations where N20 is measured worldwide. The explanation may be that N20 fluxes are negligible relative to the background concentration, so that variation in emission has hardly any effect on the air concentration. References Berdowski, J.J.M., A.F. Bouwman, J. Slanina and W.M. Kieskamp (in prep.) A s s e s s m e n t report, Theme B, "The Causes", Dutch National Research Programme on Global Air Pollution and Climate Change. Van den Born, G.J., A.F. Bouwman, J.G.J. Olivier and R.J. Swart, 1991. The emission of greenhouse gases in The Netherlands. Report 222901003, National Institute of Public Health and Environmental Protection, Bilthoven. Kroeze, C., 1994. Background document N20. Report 222901003, National Institute of Public Health and Environmental Protection, Bilthoven.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
543
Modelling of CO2 exchange between grassland ecosystems and the atmospheric boundary layer B.O.M. Dirks and J. Goudriaan
Department of Theoretical Production Ecology, Wageningen Agricultural University, P.O. Box 430, 6700 AK Wageningen, The Netherlands
Abstract
To calculate and analyse diurnal and seasonal patterns of CO 2 exchange between grassland ecosystems and the atmospheric boundary layer, a dynamic simulation model was developed. It distinguishes between a vegetational component, based on crop growth model SUCROS, and a soil component, based on soil organic matter model MOSOM, and calculates CO 2 exchange as a function of half-hourly values of air and soil temperature, shortwave irradiance and atmospheric [CO2]. As compared to measured CO 2 fluxes in a grassland ecosystem in Cabauw, The Netherlands, measurements and preliminary model calculations agreed better for nighttime fluxes than for daytime fluxes. This discrepancy suggests incorrect model assumptions. The CO 2 emission from cattle and manure, not yet included in the simulation model, is estimated to be approximately one tenth of the maximum daytime CO 2 flux in July.
1. INTRODUCTION It has been widely suggested that atmospheric CO 2 could serve as a climatic factor [1,2]. Apart from a general trend of increasing atmospheric [CO2] [3], spatially fluctuating cycles in atmospheric [CO2] have been observed [4,5,6]. The vertical dimensions of the cycles very much depend on the time scale [4]. The biosphere is thought to exert a major influence on these cycles [4,5,7,8,9]. A decrease in atmospheric [CO2] is observed during spring and summer and an increase during autumn and winter [5]. In the southern hemisphere the cycle's amplitudo is considerably less than in the northern hemisphere [4]. Within the global biosphere grasslands have an important position, in surface area, long term soil C storage and net primary productivity [7,10]. Therefore, grasslands could be a significant factor in cycles in atmospheric [CO2]. Despite having substantially different characteristics when compared with most of the world's grasslands, pasture land in The Netherlands also displays a high productivity and C storage [11]. This study aims at model development for the diurnal and seasonal cycles of total CO 2 exchange and their components between grassland ecosystems and the atmospheric boundary layer in The Netherlands, using CO 2 flux measurements for validation.
544 2. M E T H O D O L O G Y In pasture land near Cabauw, The Netherlands, situated on a 1 m thick layer of alluvial clay on peat, the Netherlands Energy Research Foundation (ECN) and the Royal Netherlands Meteorological Institute (KNMI) measured CO 2 fluxes between the vegetated surface and the atmospheric boundary layer [12], and environmental variables. Measurements used here were taken at the meteorological site of KNMI, from March 1993 up to February 1994. CO 2 flux measurements were done using the CO 2 gradient method, coveting a fetch of approximately 1.5 km length. To avoid, as much as possible, disturbance of the measurements as a result of nearby orchards and built-up area, only measurements taken at wind angles ranging from 195 up to 250 ~were used for analysis. A preliminary dynamic simulation model for CO 2 exchange and its components between a grassland ecosystem and the atmospheric boundary layer was developed (figure 1). The model distinguishes between a vegetational component, based on SUCROS, a model for crop growth [ 13 ], and a soil component, based on MOSOM, a model for soil organic matter dynamics [ 14]. The dynamics of the vegetational and soil component are a function of species characteristics and half-hourly values of shortwave irradiance, air temperature and atmospheric [CO2], and of soil characteristics and half-hourly values of soil temperature, respectively. r ........................................... ii . . . . . . . , , , air temperature radiation 9I ,
a~ssimilation~ . . . .
(rate
~--
: eI ; ,
lass'imilation rate
' I
canopy
............
...............
"i 0
~
~I CO~ ~ "---~" maintenance respiration
~ /~ 11
I "
T
I
I I ~
re-allocation ~
I
.
.
.
.
.
.
.
.
.
.
.
.
.
~,
r .......
',
!
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
l ~-. . . . . . . . . . . . . "~ stabilized l I
I-
i .... I~
I
.
.
.
.
.
.
.
.
~ "
' I
:
:
',
/ \
rate'
.
:
~
I I
.
soil tern )erature
atmospheric CO 2 ' ~ I
_ ........
~ I
.
.........
VEGETATION
I
deaihJ rate I ', ',
t
/
~~ - - " - ~ e
%uraCO~"
~ ,r ~ t
/-..,
...... Ires,,. ;
lire.'ooe "
]ire.'oo.
~ LITTER
ition
/
CO2, I'~'~--'--~
-
....
.....
/
SOIL
Figure 1. Schematic representation of a dynamic simulation model for CO 2 exchange between a homogeneous grassland ecosystem and the atmosphere. Boxes represent state variables, valves rate variables, closed lines mass flows and dashed lines information flows.
3. RESULTS AND DISCUSSION The absence of a distinct seasonal pattern of atmospheric [CO2] (figure 2) corresponds to similar observations in industrialized and densely populated areas [3]. The average diurnal pattern (figure 3) displays a relative low during daytime and a concentration gradient inversion
545 400
400 t",l
t",l
o
.
390
-: : - ~ - -
.~ ~_ 3 8 0 - , ~ -9
z: ~ 370
9
9
! m2.._mie
9 -~
9 ~-cx:,~=
-= - ,c
O
390 , . . . . . ' "
u ~
380
"=
370 '
".
.... -.
-
fi
350 1
-:-
360
"'- .....
5
1
3
5
7
9
11 13 15 17 19 21 23
daytime ( G M T + 1)
week
Figure 2. Seasonal pattern of atmospheric [CO2] at 1 m (11) and 10 m (V1), as weekly averages from half-hourly values, Cabauw, from March 1993 up to February 1994 (source data: ECN).
Figure 3. Diurnal pattern of atmospheric [CO2] at 1 m (m) and 10 m (I--1), as a yearly average of half-hourly values, Cabauw, from March 1993 up to February 1994 (source data: ECN).
-
0.25 ~ 0
--
350
" 9 13 17 21 25 29 33 37 41 45 49
"~
_~
~1,..--,-~ m
-
360
0.75,
9
:
o l
~
` l l " ....
~
--=~
-- d
S~
_-
-_ i
l"
f
"
l ~
'
l ' ID
0.5
o
;
._
=,..
•
::3
-0.25
= -o.5 c~ 9 -0.75
_--~
ipa t --- a i
_e
-.qllr t
~'
M
9
-
-i i "~
]
-1
.
-1.5 0
6
12 18 24
6
12 18 24
6
12 18 24
daytime ( G M T + 1)
Figure 4. Calculated (n) and measured (I-q) half-hourly CO 2 exchange, from 16 up to 18 July 1993, Cabauw (source measurements: ECN).
-1
-0.5
0
0.5
1
measured CO2 flux (mg/m2.s)
Figure 5. Comparison between measured and calculated half-hourly CO 2 exchange, in March, May, July and December, 1993, Cabauw (source measurements: ECN).
at transitions between daytime and nighttime, the latter reflecting transitions between upward and downward CO 2 fluxes. Depletion and replenishment of atmospheric CO 2 are thought to be governed by canopy CO 2 assimilation and respiratory processes, respectively. Patterns of calculated and measured CO 2 exchange (figure 4) during a selected period in July 1993, show a better agreement for CO 2 fluxes during nighttime than during daytime. Comparison for several months (figure 5) indicated the consistency of this discrepancy. A reasonable description of the nighttime processes - plant maintenance and soil organic matter dynamics - is suggested, especially during the 2nd half of the nighttime time interval (figure 4). Further validation would be required, encompassing the determination of the significance of the degree of detail in the process descriptions, CO 2 flux measurements under more homogeneous conditions (i.e. at a smaller spatial scale) and measurements of the CO 2 flux components. In addition it needs to be established which fraction of the variation in the measurements is inherent in the method of measurement. The different patterns of calculated and measured CO 2 fluxes during the 1st half of the nighttime time interval and the differences between calculated and measured daytime CO 2 fluxes (figure 4), seem to point to incorrect or incomplete model assumptions. CO 2 emission
546 from cattle and manure was not yet included in the calculations, but can be estimated. An uptake of 20 kg dry matter per cow per day, a dry matter C content of 40%, a 2.5 cows per ha, an equal division between actual uptake and excretion, a steady state in manure supply and decomposition, and a relatively negligible C release through CH 4 [15] results in a CO 2 emission of 0.07 mg.m-2.s -1 - approximately one tenth of the maximum daytime CO 2 flux in July 1993. In this additional CO 2 emission the manure provides a continuous background source, whereas the cattle acts as point sources.
4. A C K N O W L E D G M E N T S The Netherlands Energy Research Foundation (ECN) and the Royal Netherlands Meteorological Institute (KNMI) are acknowledged for providing their measurement data. This study was partly funded by the National Research Programme on Global Air Pollution and Climate Change (NRP).
5. R E F E R E N C E S
1 J.T. Houghton, G.J. Jenkins & JJ. Ephraums (eds.), Climate Change - the IPCC Scientific Assessment, Cambridge University Press, 1990. 2 S. Manabe, R.J. Stouffer, M.J. Spelman & K. Bryan, J. Climate, 4 (1991) 785. 3 T.A. Boden, R.J. Sepanski & F.W. Stoss (eds.), Trends '91 - a Compendium of Data on Global Change, Carbon Dioxide Information Analysis Center, 1991. 4 J. Goudriaan, Neth. J. Agric. Sci., 35 (1987) 177. 5 M.R. Manning, in: M. Heimann (ed.), The Global Carbon Cycle, NATO ASI Series I15, Springer, 1993. 6 M. Heimann, C.D. Keeling & I.Y. Fung, in: J.R. Trabalka & D.E. Reichle (eds.), The Changing Carbon Cycle, Springer, 1986. 7 J. Goudriaan, J. Exp. Bot., 43 (1992) 1111. 8 P.P. Tans, I.Y. Fung & T. Takahashi, Science, 247 (1990) 1432. 9 E.T. Sundquist, Science, 259 (1993) 934. 10 K. Minami, J. Goudriaan, E.A. Lantinga & T. Kimura, in: Proceedings of the 17th International Grassland Congress, 8-21 February 1993, Palmerston North, New Zealand, 1993. 11 J. Wolf & L.H.J.M. Janssen, Neth. J. Agric. Sci., 39 (1991) 237. 12 W.M. Kieskamp, A. Hensen, W.C.M. van der Bulk, A.T. Vermeulen, A.C. Veltkamp & G.P. Wyers, Sources and Sinks of Atmospheric CO2, report Netherlands Energy Research Foundation ECN, 1994. 13 J. Goudriaan & H.H. van Laar, Modelling Potential Crop Growth Processes, Kluwer, 1994. 14 E.L.J. Verberne, J. Hassink, P. de Willigen, J.J.R. Groot & J.A. van Veen, Neth. J. Agric. Sci., 38 (1990) 221. 15 L. Bakken, K. Refsgaard, S. Christensen & A. Vatn, in: L. 't Mannetje & J. Frame (eds.), Grassland and Society - Proceedings of the 15th General Meeting of the European Grassland Federation, June 6-9, 1994, Wageningen Pers, 1994.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
547
A S I M P L I F I E D M O D E L F O R E V A L U A T I N G THE R E S P O N S E OF T H E C L I M A T E S Y S T E M TO THE I N C R E A S E OF G R E E N H O U S E G A S E S I N C L U D I N G T H E S I M U L A T I O N OF THE G L O B A L C A R B O N C Y C L E . M. MAZZINI and F. VANTAGGIATO Dipartimento Costruzioni Meccaniche e Nucleari (DCMN) - University of PISA- ITALY
Abstract A simplified climate model which links a Box Advection Diffusion Model and a Global Carbon Cycle Model, has been set-up at DCMN of the University of PISA. The biosphere is represented by six ecosystems: tropical and temperate forests, grassland, land used for agriculture, urban areas, tundra and semi-desertic areas. In the vertical direction the model considers four fields: terrestrial surface, litter, humus and stable charcoal. The effect of varying various model parameters affecting oceanic circulation and biosphere is discussed. The analysis shows that, on a two centuries time frame starting from unperturbed 1860 equilibrium, the atmospheric CO2 concentration, and ultimately climate, will be prevalently determined by energy policies.
1. Introduction The aim of the research programme established some years ago, is the study of climate response to various energy scenarios. As first, specific objective, a global climate model has been set-up coupling a Box Advection Diffusion Model (BADM) with a Global Carbon Cycle Model (GCCM).The model contains a simple parametrization which anticipates a warming-induced transient response of the Global Thermohaline Circulation (GTC). This instrument is useful for studying the role that the ocean and the biosphere play during climate transients on the decades to centuries time frame that is of interest to the greenhouse problem. Section 2 presents the model itself, focusing on GCCM. Section 3 summarizes the results of a series of sensitivity calculations. Observations on the work already done conlude the paper.
2 The model The main features of the climate model set-up at DCMN are presented below in figures 1 to 3. In particular, figures 1 and 2 indicate schematically the main reservoirs and the oceanic and biospheric subdivision of BADM and GCCM. An overview of the model is presented in fig. 3, which focuses on the links between BADM, GCCM, and the proposed parametrization for the response of the GTC to the global warming. The 0 subscript indicates climatological mean unperturbed values.
548
Atmosphere
L Mixed layer
A'r~na
I
I ATMOSPHERE~__1
i T ~___L
Qb
IFOSS,LFUELIB~OSPHER~IOCEANI
Qb+l(1"~)Q~i Qb
Fig,2a
T~A ~
A
A
Figure 1 Main compartments and circulation scheme of BADM
LAYERS Inter.mediate
Qb A ntarcti c sensitivity
!
-
USD
ATcr T
BADM
~'"
DEEP LAYERS _____J Fig. 2b
Fig. 2c
AToff
ATs
/ [ I~%~Ni T ? I HUMUS~ - -
se21-ice
~ -~GC
R
cI
Figure 2 Main reservoirs and CO2 fluxes in GCCM
M
Figure 3 - Links between BADM, GCCM and GTC
2.1 The Box Advection Diffusion Model
The climatic model assumes that the surface temperature difference between the mixed layer and the atmosphere ATma does not appreciably vary during the transient; the reference value ATm, =4.8~ is assumed. Based on this approximation and on the continuity of temperature across the mixed layer and the deep ocean interface, the model is reduced to the equation: 81" --=K c2
02T
cTI' +w-cgz2 cgz
(1)
and the associated upper and lower boundary conditions (see [1] for details on BADM as well as for the parametrization for the AABW formation). The heat diffusion in the interior of the ocean is parametrized in B ADM by the so called deep ocean thermal diffusivity K [m2/s]. Due to the competing downword diffusion and upward advection of heat, the model predicts a steady-state exponential temperature profile in
549
2.2 The Global Carbon Cycle Model The GCCM considers four main reservoirs: atmosphere, ocean, biosphere and fossil fuel (fig. 2a). The atmospheric carbon content is driven by the flux of CO/emitted due to fossil fuel consumption (Ff,a), the net terrestrial biosphere output from both human disturbances (DEFOR) and Total Net Ecosystem Production (TNEP), and by FmLa, Fm2,a which are the ocean-atmosphere fluxes respectively from Warm Surface Water (WSW) and to Cold Surface Water (CSW). These are calculated through the model of oceanic circulation shown in fig. 2b. The relationship used is: dN~ = Ff~ + Fro,. + Fro2a + DEFOR + TNEP dt ' ' '
(2)
In this paper, the attention is focused on the terrestrial biosphere (Fig. 2c) which is represented by six ecosystems: tropical forest, temperate forest, grasslands, agricultural areas, urban zones, tundra and semi-desert; vertically, the subdivision has been fixed in four dominions: the first is partitioned in leaves, branches, stems and roots; the other three are litter, humus and stable carbon in soil. The outflow from each component is equal to its content divided by its average life-span. The model accounts for anthropogenic impacts (forest cutting, burning and shifting agriculture), wich allow to calculate DEFOR in eq. (2). The driving input of the ecosystems is NPP (Net Primary Production), given by the difference between the annual assimilation of carbon and respirational losses. Global NPP is partitioned according to fixed distribution coefficients eqk; j denotes the ecosystem number (16), k the vertical part number (1-4). The carbon originating from biomass and litter is not entirely channeled to the atmosphere, but a fraction ek is converted into long lasting charcoal. Shifting agriculture practices can be mathematically represented by a time dependent exchange matrix aij, denoting the gross annual transfer of land from ecosystem j to ecosystem i. The differential equations for the change of the carbon content are those suggested by Ketner [2]. Starting from the unperturbed climatological equilibrium, inputs to the model are provided in terms of C02 release to the GCCM atmosphere from fossil fuel reservoir. As illustrated in fig. 3 the coupling is firstly achieved by forcing BADM with a radiative perturbation term R calculated on the basis of the C02 level attained by the GCCM atmosphere. Then, the main output from BADM (referred to below as global mean surface temperature change ATs), determines the response of the oceanic circulation both in B ADM and in GCCM.
3. Values and ranges of the model parameters. To the basic BADM parameters K and w0, are assigned the reference values first proposed by Hoffert et al.[3]: 6.34 10-s m2/s (2000 m2/y) and 1.27 10.7 m/s (4 m/y) respectively, giving a scale lenght K/wo of 500 m. BADM was set-up by imposing a ramp radiative forcing dR/dt=l.27 10 .9 W/(m2.s), which falls in the range of interest of the greenhouse problem. The planetary albedo change is simply
550 estimated by assuming a climate sensitivity da/dT equal to 0.002/~ With this feedback, the BADM sensitivity to a C02 doubling is approximatively 3.4~ The GTC parametrization is focused on bottom water forming processes in Antarctica, with the possibility of a decrease of Qb or, equivalently, of the model parameter w. The weakening of the bottom circulation was evaluated on the basis of different scenarios for the Antarctic sea-ice sensitivity, defined by the couple AT,:r, ATofr (see [ 1] for details). With regard to the intermediate circulation loop, the assumed lower and upper boundary for the unperturbed volumetric flow Qio are 35 and 140 Mm3/s, while the recirculation parameter o~ is allowed to vary in the range 1/3 to 1. The sensitivity of the intermediate circulation, as defined by the model parameter IFS, is investigated assigning to IFS values from -0.6 to 0.5. The initial conditions are defined by the model equilibrium at an atmospheric CO2 level of 285 ppmv in 1860. The energy scenarios are the most important element in determining future CO2 concentrations. Both the standard and high IIASA scenarios [4] were applied.
3.1 Main results of the sensitivity analyses. The results of the sensitivity analyses are reported in [1]. They can be summarized as follows: 9 BADM is quite insensitive to the values of K and w0 provided the scale lenght K/wo of the unperturbed profile is mantained, as already observed by Harvey and Schneider [5]; 9 some differences in the atmospheric CO2 build-up are observed when either doubling or halving the Qi value; the model is almost insensitive to the intermediate circulation scheme (or parameter); 9 a reduction in the global mean upwelling rate has a significant impact on the transient response, corresponding to a lag-time (as measured with respect to the no feedback response) of several years after 100-200 years of integration even in the GTC lowsensitivity scenario; 9 this extra-lag time (additional with respect to the anticipated effect of the oceanic thermal inertia) develops progressively until the upwelling feedback is eventually exhausted; it ranges from approximatively 5 to over 10 years, as far as years around 2000 are concerned in intermediate (I) and high (H) sensitivity cases respectively; 9 the case of GTC full stop, although very crude, illustrates the intrinsic character of this feedback: its actual role is to delay the response without absorbing the perturbation; 9 inherently in the model dynamics, the impact of the flow weakening increases with time: around year 2030, extra-lag times rise to approximatively 15y in case I and 30y in case H. Qb variations are on the whole of the order of-15% (I) and -27% (H); 9 the model indicates an almost complete insensitivity of R to variations in GTC, when a 200 year time frame is considered; significant differences take a longer time to develop, up to approximatively 25 ppmv after 400 years. Referring to present years, the anticipated flow reductions and ATs variations are overwhelmed by the current uncertainty about the basic B ADM parameters K and w0. With regard to the biosphere response, the results of sensitivity analyses for GCCM linked to BADM are listed in table 1.
551 Table 1 Sensitivity of the atmospheric CO2 concentration (ppmv) to variations of main GCCM parameters or policy assumptions 1990 2030 -5,1 -14 Oceanic diffusion coefficient doubled +3,5 +12 Stop to deep water formation (Qb=0) +0,8 -6,3 NPP doubled +17 +36 Life-span biomass doubled -5,5 -15 Deforestation rate halved 0 -9,8 Reforestation started in 1990 with 10 ]] m2/y 0 -28 Stabilization of CO2 emissions to 1990 value +0,1 +53 High IIASA energy scenario
It is clear that the range in projected CO2 concentrations due to energy scenarios is much larger than that due to parameter uncertainties. The net effect of doubling the NPP is small because it involves not only the doubling of biomass but also the increase of CO2 releases by deforestation. Limiting the discussion to the time span of our interest (until 2030), a reference time history for the radiative forcing can be defined according to the case of unperturbed volumetric flows in GCCM, while fixing reference Qi0 and ot values of 70 Mm3/s and 0.5. Figure 4a, shows the CO2 atmospheric level (given by the fully coupled model of fig.3, with an intermediate sensitivity of GTC to the earth heating) and corresponding radiative forcing. The other curves of fig. 4a were obtained in a similar way, by considering the various policies outlined in table 1. The corresponding surface temperature responses are shown in fig. 4b. In 2020 the surface temperature increases differ of 0.3 degrees if one jump~from a policy of emission control and reforestation to the high IIASA scenario in energy consumption.
R~tiw forci~ [W/m2]
CO2~ ]
............
500
~,6
1,6
1,4
1,4
1,2
1,2
B
450
1 3
1
400
0,8 0,6
0,6
350
0,4
0,4
0,2
0,2
2
1
0,8
300 -
18(50
1890
|
1920
-
i
i
19~0
|
|
I
1990
i
i
i
018~,O-- 1890
1920
19.80
1950
2:O10
2020
Fig 4a At,mosphencC021evals Fig 4b Surfacetemperaturechange[~ A)- referenceatmosphericCO2level B)- high IIASAenergy scenario C ) - CO2 errlssion control+ reforestation+ deforestationrate halved(1990 values) Figure 4 Sensitivity of climate response to various policy scenarios
0
552
4. Concluding remarks The model presented in section 2, although rather simple, is already an useful instrument for sensitivity analyses of climate changes related to various energy scenarios. In this respect, the following observation can be recalled: . the low sensitivity of GCCM to changes in the ocean circulation flows, at least in a century time scale and for the examined range of variation of the various parameters; 9 the significant role of the GTC perturbation in delaying the climate response to the anthropogenic forcing, as pointed out also by several authors. 9 the demonstration that the effect of environmental policy (reforestation) is fairly small (around - 10 ppmv in 2030); its effectiveness is very low, compared with that obtainable by changing the typology of energetic consumptions. The DCMN model has some innovative features (coupling of BADM and GCCM, explicit modeling of Antarctic Bottom Water formation), which should contribute to improve the reliability of simulated climate response. However, further improvements in the model are needed before the intended goals are reached.
Abbreviations 9 AABW: Antarctic Bottom Water 9 BADM: Box Advection Diffusion Model 9 CSW: Cold Surface Water 9 DCMN: Dipartimento Costruzioni Meccaniche e Nucleari 9 D E F O R : Human disturbances to global carbon cycle 9 GCCM: Global Carbon Cycle Model 9 GTC: Global Thermohaline Circulation 9 IIASA: International Institute of Applied System Analysis 9 IFS: Intermediate Flow Sensitivity 9 NPP: Net Primary Production of ecosystem 9 TNEP: Total Net Ecosystem Production 9 WSW: Warm Surface Water
Symbols 9 9
9 9 9 9 9 9 9 9 9 9
a: earth mean surface albedo a..: rate of area transition from ecosystem j to i Ij Ff,,: Flux of carbon from fossil reservoir to atmosphere [Pg/y] Fm,l~:Flux of carbon from cold mixed layer to atmosphere [Pg/y] Fma~: Flux of carbon from warm mixed layer to atmosphere [Pg/y] K: ocean mean diffusivity [m2/s] N a : Mass of carbon in atmosphere [Pg] Qb: Deep water flux [Mm3/s] Qi: Intermediate water flux [Mm3/s] Q,i: mean volumetric flow-rate of sea-ice formation [MmS/s] R: Radiative forcing [W/m 2] t:time [s]
9 9
w: upwelling mean rate [m/s] z: vertical coordinate [m]
9 o~: coefficient of Qi partition between CSW and WSW 9 O~jk:distribution coefficient of NPP 9 ATr value of ATs producing the start of GTC reduction [~ 9 ATm: Mean temperature difference between atmosphere and ocean surface water [~ 9 ATo~: value of ATs producing the stop of GTC [~ 9 AT,: Temperature change of earth surface [~ 9 ek: Carbonization fraction of component k upon oxidation
References [1] Nicolini L., Vantaggiato F., Mazzini M. (1991) "Climate Impact of Changes in the Global Thermohaline Circulation". NATO ASI on "Energy and Water Cycles in the Climate System". Glucksburg (Germany), Sept. 30-Oct. 11, 1991. [2] Ketner P. and Goudrian J. ".4 Simulation Study for the Global Carbon Cycle, Including Man's Impact on the Biosphere". Climatic Change, n. 6, pp. 167-192, 1989 [3] Hoffert M.I., Callegari A.J., Hsieh C.T., "'The role of deep sea heat storage in the secular response to climatic forcing". J. Geophys. Res., vol. 85, pp.66676679, 1980. [4] IPCC-Scientific Assessment of Climate Change (1990). [5] Harvey, D., Schneider, S.H., "'Transient climate response to external forcing on 10~ year time scales part 1: experiments with globally averaged, coupled, atmosphere and acean enerKv balance models", J. Geophys. Res., v. 90, pp. 2191-2205, 1985,
554 tree carbon using basic wood density figures, a stem biomass to total biomass conversion factor (e.g. Cannell 1982) and a carbon content of 50% (Ajtay et al. 1977). Stem dry weight increment was not coverted to whole tree dry weight increment because the forest is already of considerable age at which it can be assumed that net increase in the compartments branches, foliage and roots is negligible. The stock of carbon in the forest floor and in the stable humus was derived from standard Dutch soil descriptions (Beuving 1984) and from available literature data on dry weight of the forest floor (e.g. Cannell 1982).
3. RESULTS The average age of the Dutch forest amounts to 50-60 years and the standing volume is small and still increasing. The HOSP-inventory showed an average standing volume of 170 m 3 ha ~ and an average current volume increment of 9.0 m 3 ha -a yr-~ for all species (HOSP 1991). Half of the current increment is harvested annually and the wood is mainly used for paper, packing wood and particle board. The standing forest of Scots pine is by far the most important forest ecosystem in the Dutch forest when carbon stock is concerned (see Figure 1).
Biomass
Litter
T
Humus
5 0 . 0 0 0 ha'
]]llm ::':':".':9:+,'::<::."
>5
.. +':.+"+:,.%,%"
....... :~,;-:.:.:+:-:~+x~::~x+,.-:+:~
-
iii!iiii!i ,,,.,,.,,
:i:!:i:i:i: ,,, ,., ,,,
ii!ii!i ,.,=;
.". , " . . ," ...,
,:.:-:.:,:. ?????i
:,;,;,: :':::
........ ""
'i'!'!??
i%
. . . .
:::::;::::
[::::':;::•
::::;:::
:
. 9.
i:i:i:iil,, ,.,,,., :.:.:.; :,. ::':"
......... v,,,-,,,,,,,,.,:,:,:,:.:,:-:-:, .........
..,
i~i~i~:::i~iliiiiiiiil iiiiiil.... ii::i:,i::: iii::iiiii::i~:~:~:~:~! ,,.,,
_
l ....../... 1
rcu~sru
!
Pinus sylvestris
]
Pinus nigra
] Larix kaempferi
uercus robur
P~
P i c e a abies Pseudotsuga menziesii
,,.,,,.
!)ii))iiiiiii)ill
i~ii!:!!
iiiiiiiiiiiiiiii}
.................. iiiiiiii!ii:i!!iii!!i:ii!!iiiiiil ~
,:,:,:,:,:-:-:,:, ......... ,:-:,:,:,:,:,:.:, ,:,:,:,:,:-:,:,:, ......... ':':':-:':-:':-:" .........
/liliii/ii!ili [:i:?:i:!:!:i:i:?
Spontaneous ~rest
/ Coppice
/ t
F a g u s sylvatica
/
/ Natural r e g e n e r a t i o n
Betula pendula Alnus glutinosa
Figure 1. Carbon stocks for three compartments of fifteen forest types of the Dutch forest. The horizontal axis gives the area for each forest type, and the vertical axis the amount of carbon per ha.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
553
Carbon relations of Dutch forests G.J. Nabuurs and G.M.J. Mohren Institute for Forestry and Nature Research, PO Box 23, NL-6700 AA Wageningen, The Netherlands
Abstract
Present stock of carbon in living biomass, stable humus and litter and annual accumulation of carbon in stems of fifteen forest types of the Netherlands was quantified. Estimates were based on the most recent forest inventory, yield tables, biomass measurements and soil profile descriptions. The present stock of carbon in the living biomass, soil stable humus and forest floor of the Dutch forest amounts to 64 Mt C. 58% of this is stored in the soil stable humus. The average carbon stock in the living biomass amounts to some 59 Mg C ha 1. The net annual carbon sink was roughly estimated at 0.33 Mt C yr-1. Due to variation in annual growth of the forest caused by e.g. climatic variability, net annual sequestration may vary between 0.2 and 0.4 Mt C yr-~ for the entire area.
1. INTRODUCTION The biosphere plays an important role in the regulation of the global carbon cycle. Especially forests are important as a storage pool of atmospheric carbon and as regulator of the carbon cycle although the precise role of temperate and boreal forests remains one of the largest unsolved questions in the global carbon budget. Forests of the temperate and boreal zone may play a substantial role in sequestering the missing carbon (Sampson et al. 1993). Although Dutch forests are insignificant in the temperate forest carbon cycle, determining the magnitude of its present source/sink activity is relevant from the point of view of national climate policy and for improved understanding of vegetation-atmosphere interaction.
2. METHODS
For each forest type and tree species, the following parameters were quantified: carbon stock in the living biomass, forest floor and soil stable humus and current annual accumulation of carbon through stem volume increment per site class. Harvesting data were gained from the last inventory and product decomposition was roughly estimated. Major source of information was the latest forest volume and growth inventory which covered the period 1984-1989 (HOSP 1991). To assess growth for three site classes, available yield tables (Janssen and Sevenster 1991) were used in combination with the forest inventory data. Standing wood volume figures were converted to whole
555 The total stock of carbon stored in the forest biomass, forest floor and soil organic matter equals (see Table 1) 54.2 Mt. 58% of the carbon is stored in the stable humus, which makes this compartment rather important. Assuming that the 85% of the Dutch forest which was covered in this study, is representative for the total afforested area (approx. 330,000 ha), the total stock of carbon can be estimated at 64 Mt. The carbon pool in the living biomass, totaling 16.5 Mt, is small, both because of the limited area of forest in The Netherlands and because of the young average age. The average carbon stock per hectare in the living biomass amounts to about 59 Mg C ha -1. Of the forest types considered, beech stands have the highest average stock with 125 Mg C ha -1. Figure 1 presents the results for each forest type.
Table 1. Total stock of carbon in living biomass, forest floor and stable humus of 85% of the Dutch forest.
Standing forest: Scots pine Austrian and Corsican pine Douglas-fir Japanese larch Norway spruce indigenous oak beech poplar and willow ash alder birch red oak Other forest types: spontaneous forest coppice natural regeneration forest
Total stock of carbon
Area (ha)
Living biomass (Mt C)
Forest floor (Mt C)
Stable humus (Mt C)
98 213 16 302 15 722 18 015 13 262 27 084 7 150 15 280 3 411 967 5 506 7 856
5.44 0.87 0.92 1.11 0.72 2.26 0.89 0.55 0.13 0.03 0.25 0.70
3.80 0.18 0.15 0.26 0.13 0.74 0.10 0.11 0.03 0.01 0.06 0.11
9.5 1.5 1.7 2.1 1.6 3.0 0.9 1.5 0.4 0.2 0.9 0.5
23 970 22 250 13 084
0.93 0.70 0.38
0.80 0.36 0.20
2.8 2.7 2.0
15.87
7.04
31.3
The present gross annual carbon accumulation through stem volume increment for the total area was estimated to be 0.66 Mt C yr-1. Through harvesting and decomposition of woody products, about half of this is immediately returned to the atmosphere. The resulting net annual accumulation amounts to 0.33 Mt C yr-1 (see Nabuurs and Mohren 1993 for more details). The average net rate of carbon sequestration for all forest types equals some 0.97 Mg C ha-lyr-1. The gross values per forest type varied from 0.8 to 4.6 Mg C ha -1 yr-1. Of the forest types considered, beech forests appeared to have the highest net carbon flux density.
556 4. DISCUSSION At present, forests in the Netherlands appear to act as a carbon sink. This in accordance with other studies reporting a noticable sink activity of the temperate forests (Kauppi et al. 1992). In the Netherlands, total sequestration amounts to 0.7% of the annual emission of carbon. These results are based on growth measurements over the period 1984-1989. More recent data indicate an average annual volume increment of 7.8 m3ha-ayr-1 instead of 9 m 3 ha-lyr-1as was previously reported. As a consequence, the net storage rate as reported here, will decrease accordingly. This indicates that these results strongly depend on yearto-year variation in growth of the forest as caused by e.g. climatic variability. Due to this, the net annual carbon sequestration rate is expected to vary between 0.2 and 0.4 Mt C yr-1. The reported volume increments are much higher than could be expected from earlier inventories and available yield tables. This discrepancy is most likely due to underestimation of growth in the yield tables and due to an overall growth increase because of enhanced nitrogen availability. However, from analysis of growth data of longterm yield-plots, no obvious increase in growth rate could be detected. It can be expected that the rate of carbon accumulation will continue for several decades because of the young age and limited standing volume at present.
5. REFERENCES Ajtay, G.L., Ketner, P. & P. Duvigneaud. 1977. Terrestrial primary production and phytomass. In: Bolin, B., Degens, E.T., Kempe, S. & P. Ketner (eds.) The global carbon cycle. Published on behalf of the Scientific Committee on Problems of the Environment (SCOPE) of the International Council on Scientific Unions (ICSU). SCOPE 13. pp. 129-181. Beuving, J. 1984. Moisture, percolation, density and composition of soil profiles in sand, loam, clay and peat soils. Instituut voor Cultuurtechniek en Waterhuishouding. Rapport 10. Wageningen, 26 pp (in Dutch). Cannell, M.G.R. (ed.) 1982. World forest biomass and primary production data. London, Academic Press, 391 pp. HOSP. 1991. Wood harvest, stocks and annual increment of the Dutch forest. Committee Harvesting statistics and prognosis harvestable wood. Utrecht, State Forest Service. 15 pp + app (in Dutch). Janssen, J. and J. Sevenster. (eds.) 1994. Yield tables for important tree species of the Netherlands. Committee Yield tables. 188 pp (in Dutch). Kauppi, P.E., K. Mielik/iinen & K. Kuusela 1992. Biomass and carbon budget of European forests, 1971 to 1990. Science 256: 70-74. Nabuurs, G.J. & G.M.J. Mohren. 1993. Carbon stocks and fluxes in Dutch forest ecosystems. Institute for Forestry and Nature Research. Wageningen, Netherlands. IBN-Research Report 93/1.85 pp. Sampson, R.N., Apps, M.J., Brown, S., Cole, C.V., Downing, J., Kauppi, P., Ojima, D.S., Smith, T.M., Solomon, A.M., Twilley, R.R. & J. Wisniewski. 1993. Terrestrial Biospheric carbon fluxes: quantification of sinks and sources of CO 2. Workshop statement. Bad Harzburg, Germany. 1-5 March 1993. 15 pp.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
557
E f f e c t s of c l i m a t e c h a n g e on d e c o m p o s i t i o n of soil o r g a n i c m a t t e r in a boreal ecosystem P. Verburg and N. van Breemen Department of Soil Science & Geology, Wageningen Agricultural University, P.O. Box 37, 6700 AA Wageningen, The Netherlands
Abstract In the CLIMEX project, two catchments are manipulated to investigate the effects of elevated temperature and CO2 on decomposition of soil organic matter. Litterbag studies do not show a clear treatment effect after five months of manipulation. Due to the treatment, however, the contribution of organic anions to the total charge balance of the soil solution appears to decrease. Laboratory experiments show that C loss from soil columns as CO2 and TOC increases with temperature. Because N uptake by vegetation is absent, nitrification increases with temperature resulting in increased acidification of the outflow water.
1. INTRODUCTION Yearly, at global scale, 50-100 Gt of C cycles along the photosynthesisdecomposition pathway against an annual net addition of 2.8 Gt C to the atmosphere as CO2 [1]. Therefore, changes in decomposition rate of Soil Organic Matter (SOM) due to climate change might significantly influence the net exchange of C between the atmosphere and the land surface. An increase in atmospheric CO2 is most likely to affect decomposition by influencing Net Primary Productivity and chemical composition of the litter. Also changes in soil physical conditions might occur due to a change in water use by the vegetation [2]. An increase in temperature could affect decomposition by stimulating biological activity [3] and by changing the physical environment through increased evapotranspiration. Increased decomposition could result in higher N mineralization. Especially in areas where plant growth is N-limited, as in many temperate ecosystems, plants might benefit both from increased N availability and from CO2 fertilization. If N mineralization exceeds plant uptake, N could leach to surface and/or groundwater resulting in acidification and eutrophication.
558 2. M E T H O D S AND R E S U L T S
2.1. Site description Field experiments take place within the CLIMEX project in Norway [2]. In this experiment two small forested catchments are being manipulated. The largest catchment (KIM; 1,200 m 2) is enclosed by a greenhouse and receives rain cleaned from acidifying components. In this catchment, CO2 is increased by 200 ppmv and temperature by 5~ above ambient conditions. In a second, covered, catchment receiving acid rain (EGIL; 800 m2), soil temperature is increased 5~ above ambient conditions. In both catchments, one-third of the catchment is not manipulated and acts as control section. Two non covered catchments (METTE; 650 m 2 and ROLF; 220 m 2) are used as reference to account for possible roof effects. These catchments receive ambient, acid, rain. The climate treatments started in June 1994. 2.2. Decomposition of fresh litter Using litterbags, we investigate the influence of temperature and substrate quality on the decomposition rate of fresh litter under field conditions. Before the start of the manipulation, we carried out a pilot experiment with litter from Scots pine to detect possible site differences. Mass loss of needles incubated under heather in KIM, EGIL and ROLF aider one year was 307_+53, 250_+39 and 222_+43 mg/g respectively. In April 1994, we incubated birch litter produced at 350 and 700 ppmv CO2 in the treatment and control parts of KIM and EGIL. This setup allowed for separation of'treatment-effect' from 'substrate-quality effect'. Aider six months of incubation, we found a statistically higher decomposition rate in KIM corresponding with the pilot study. Within the catchments, we could not find any effect due to differences in substrate quality or climate treatment. 2.3. Soil solution chemistry Cleaning of the rain has caused a clear change in runoff chemistry over the past eight years [4]. The percentage of the total positive charge balanced by organic anions proved to be sensitive to the rain treatments. Therefore, we used this parameter to detect whether the treatment has affected soil solution chemistry. In the control sections of the manipulated catchments, the contribution of organic anions to the charge balance increased significantly in the two consecutive years (Table 1). This increase was less pronounced in the treatment sections. We cannot fully explain the observed trend yet. We speculate that increased N mineralization, followed by nitrification will cause an increase in nitrate. If plants do not take up this nitrate, the contribution of organic anions to the charge balance decreases. However, field data do not show a statistically significant increase in N mineralization in the treatment sections.
559 Table 1 Positive charge balanced by organic anions during June-August (%) 1993
1994
KIM
control treatment
34 (10) b 32 (13) b
56 (12) c 35 (19) b
EGIL
control treatment
9 (8) a 9 (7) a
27 (7) b 20 (12) d
METTE
control
25 (15) b,d
33 (16) b
Standard deviation in parentheses. Different letters show significant differences at the 95% confidence level.
2.4. C m i n e r a l i z a t i o n in i s o l a t e d soil c o l u m n s Carbon mineralization can be estimated by measuring C O 2 evolution from soil columns without vegetation. In the field, interference of C02 originating from root respiration makes interpretation of C09 emission measurements difficult. We have incubated undisturbed soil columns (16 cm wide and 60 cm long) at 5, 10 and 17~ to follow the effect of temperature on C mineralization and chemistry of the outflow water.
Table 2 Average values for selected parameters in soil column experiment after 10 weeks
CO 2 emission pH in outflow TOC 1 in outflow A1"§ in outflow NOa in outflow
gCO2 coltmm "1 week "1 mg 1"1 laeq 1"1 laeq 1"1
5~
10~
17~
0.85 4.52 3.7 143 120
1.18 4.53 3.7 242 288
1.69 3.96 6.0 484 344
1 Total Organic Carbon
These results show increased C loss both as CO2 and TOC in drainage water with increasing temperature (Table 2). Simultaneously, more N is mineralized which is nitrified resulting in increased acidification of the drainage water. Since plants are absent, the acidification due to N mineralization followed by nitrification is probably more severe than in systems with vegetation. The increase in inorganic A1 in the outflow with temperature reflects the increasing
560 acidification.
3. D I S C U S S I O N
Background data obtained before the start of the treatments suggest that SOM does not behave similarly in all catchments. Some differences can be ascribed to the 'roof-effect' (litterbag studies), others to the rain treatment (soil solution chemistry). After half a year of treatment, field data do not show a clear treatment effect except the soil solution chemistry. However, the treatments only started in June 1994. We expect changes in decomposition rate of SOM to be more pronounced in winter because at lower temperatures biological processes respond more clearly to a change in temperature than at high temperatures [3]. One major difficulty with the interpretation of field data is that they result from an assembly of environmental conditions. Therefore, additional laboratory studies under controlled conditions are needed to separate different processes occurring simultaneously. Based on the column experiments we can expect an increase in N mineralization at increased temperature as well as higher TOC levels in runoff. Acidification of soils and surface waters will only occur if N mineralization exceeds plant uptake.
4. A C K N O W L E D G E M E N T S
We thank M. Vreeken-Buys and B. Berg for assistance with the litterbag studies, L. Rou and N. Nskken for carrying out most of the analyses and, R. Nieuwenhuis and M. Ent for their contributions to the soil column experiments. Financial support for this study was provided by the National Research Programme (NOP 853098) and the Commission of European Communities (EVSVCT91-0047).
5. R E F E R E N C E S
1 J. Goudriaan, J. Exp. Botany 43 (1992) 1111. 2 E.D. Schulze and H.A. Mooney (eds.), Design and execution of experiments on C02 enrichment, Brussels, 1993. 3 M.J. Swii~, O.W. Heal and J.M. Anderson, Studies in Ecology. 5 (1979). 4 R.F. Wright, E. Lotse and A. Semb, Nature, 334 (1988) 670.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
561
Soil carbon turnover in subalpine systems and its dependence on climate M. Kleber and K. Stahr Universit~it Hohenheim, Institut fLir Bodenkunde und Standortslehre (310), D-70593 Stuttgart, Germany
Abstract
In the "Westallg~iuer HLigelland" in southern Germany with a mean annual precipitation of 1400 mm and a mean temperature of 6.5~ soil respiration has been monitored since spring 1993 by means of the modified Lundegardh-method. Air temperatures in the months june to august 1994 were by an average 5~ higher than summer 1993 and long-term monthly means. This enabled the authors to determine the response of soils different in soil type, carbon content and slope position to a temperature increase under field conditions.
1. INTRODUCTION The worlds soils contain about twice the amount of carbon present in the atmosphere, mainly in the form of soil organic carbon. Mineralization of litter and humus materials returns CO2 to the atmosphere. Since the pool of organic matter on land is large, small changes in its size yield large impacts on the atmosphere. It is critical to understand, whether and how this reservoir takes part in a possible feedback loop with climate. Of special interest are soils, which developed under a moist and cool climate regime. These soils usually are on a high soil organic carbon level, when in equilibrium. Their response to potential climate change needs to be thoroughly understood. Field measurements of the flux of CO2 from the soil surface provide an estimate of the total respiration in the soil, and an approach for estimating turnover of the humus pool. Unfortunately, the respiration of living roots makes it difficult to use estimates of CO2 flux in calculations of turnover of the soil organic matter pool (Schlesinger, 1991). Up to date, only rough estimates (53-90% of total rhizosphere respiration; Johansson 1992) for root respiration are available. In spite of this uncertainty, soil respiration is regarded as an indicator for the overall physiological activity of a given soil. The CO2 efflux rates presented here are considered as being equivalent to soil respiration, but it has to be emphasized, that soil-CO2 efflux rates, not actual soil respiration, are measured. A major methodical constraint exists in this context: there is no technique capable of exposing intact soils and sites to controlled conditions to
562
evaluate their response to changing climatic conditions. A possible solution lies in the long-term monitoriong of soil carbon turnover on site. This offers the chance to compare the behavior of soil and site under the influence of seasons of varying intensity, and it should render insight into the principal behavior of a soil at a given site in case of changing climatic conditions. At the Siggen research site, soil respiration has been monitored since May 1993. The fact, that summer 1994 was much warmer than summer 1993 will be regarded as an on site experiment concerning the response of the soil system to changing climate.
2. MATERIALS AND METHODS
2.1 Site and soils The "Westallg~iuer H~gelland" belongs to the humid and cool areas in southwest germany. Rainfall increases from 700 mm per year at UIm to 2-3000 mm at the perimeters of the alps, with the Siggen area receiving 1400 mm. Mean annual temperature at Siggen is 6,5~ The area is dominated by temperate grassland, that receives up to five slurry applications per year and is cut 4 to 5 times. Grazing occurs usually only for a few weeks in late autumn, when cutting the gras would not be efficient any more. The research site Siggen was set up in 1987 on the northwestern slope of a terminal moraine of the W~rm-glaciation. Siggen is one of several sites contributing to a project investigating the effects of intensive agriculture on the environment. It was initially selected to determine the effects of slurry induced nitrate leaching on the little lake "Neuweiher". The two soils regarded here are a non-fertilized histic gleysol (SR) close to the lake and a gleyic cambisol (HFV), which is situated about 200 m upslope and is elevated about 30 m. The cambisol (HFV) is treated with fertilizer (= slurry) and cut 4-5 times a year according to local practice, the histic gleysol is cut one time per year if soil stability permits operation of harvesting machines, but does not receive any fertilizer. Vegetation consists mainly of filipendula and phragmites at SR, and is a Iolio-cynosuretum at HFV. The main properties of the soils are given in Table 1. 2.2 Soil respiration The most commonly used methods for measuring respiration rates employ one of many alkali absorption techniques or infrared gas analysis. The simple soda-lime technique described here was very appealing because measurement of soil CO2 efflux in several plots scattered over a large area and 200 km away from the University Hohenheim was to be achieved. The method was recommended and tested by Tesarova & Gloser (1976) and Edwards (1982). CO2 is absorbed by granules of soda-lime weighed before and after a 7 days exposure in plastic dishes under a metal cylinder (diameter 23 cm, height 31 cm), covering a surface area of 415.5 cm2. Prior to each determination, all green vegetation is clipped out inside the cylinders. The cylinders are pushed 2-3 cm into the ground and covered with a 50x50 cm PVC-sheet to prevent excessive heating through solar radiation. Each variant consists of 8 replications, and new patches within the plots are selected for
563
each measurement. Plots within the research site are of uniform soil characteristics and span an area of 400 square meters. The amount of CO2 absorbed in soda lime is calculated from its weight increment. Details to the procedure are given in Kleber et al. (1994).
Table 1: Soil properties for the first horizon, all soils non-calcareous. Parameters determined according to Schlichting & Blume (1966). Soils are classified according to Spaargaren (1994).
soil type C in fine earth (%) bulk density (g cm-3) stone content (%vol) depth (cm) stock of C (kg m-2) pH sand/silt/clay (%) total pore space (%~)
SR histic Gleysol 19.99 0.31 0 10 6.24 6.0 13/43/44 84
HMV gleyic Cambisol 3.94 0.9 1.0 10 3.49 5.O 25/46/29 64
2.3 Soil temperature and moisture measurements Soil temperature (at the depth of 5 cm) is measured with SKTS 200-Sensors of UP Umweltanalytische Produkte GmbH at two continuously recording weather stations. One weather station is located at the lakeside in the vicinity of site SR, the other uphill close to HMV. Soil moisture is determined gravimetrically from samples taken weekly from the upper soil layer, it is expressed as % of maximum water retention capacity (WRC), which was determined for each plot with ten replications according to Schlichting & Blume (1966).
3. RESULTS
As illustrated by Table 2, summer 1994 was unusually warm. Both 1993 and 1994 show precipitation values below yearly mean values. In 1994 air temperature was far above long-term monthly means. Air temperatures are generally higher in the basin than uphill. Depending on the month, air temperature in 1994 exceeded values of 1993 by 3-8~ uphill and 1-6 ~ at the lakeside, with absolute values highest at the lakeside. On a global scale, soil respiration rates correlate significantly with mean annual air temperatures, mean annual precipitation and with the interaction of these two variables (Raich & Schlesinger, 1992). If smaller temporal and spatial scales are investigated, interactions of these factors are not quite as easy to be identified.
564
Figure 1 shows soil respiration in 1993 and 1994 at the site HMV. Soil temperature in 1994 noticeably exceeds the values of the previous year, but soil respiration remains at about the same level (Table 3). Obviously the lower soil moisture content in 1994 hampers respiration.
Table 2 Mean air temperature and precipitation at Wangen/AIIg~iu from 1951 to 1980; together with corresponding values at Siggen (about 10 km away from Wangen) for 1993 and 1994. (SR) denotes weather station at lakeside, (HMV) weather station uphill.
June
temperature (~ 1951 - 1980 Wangen ~ / mm 14.1 / 150 HMV SR
/ precipitation (mm) 1993 1994 Siggen Siggen ~ / mm ~ / mm 12.1 / 90 15.9 / 52 16.2 /-"17.3 / -"-
July
15.8 / 156
HMV SR
12.1 / 117 15.9 /-"-
20.2 / 94 21.8 /-"-
August
15.1 / 149
HMV SR
13.2 / 43 16.7 /-"-
18.3 / 127 19.5 /-"-
A rather different situation can be observed at the lakeside (Figure 2). Temperatures in 1994 exceed the 1993-values not as far as at site HMV. Respiration however increased by 50%. Again, soil moisture has to be held responsible for an explanation. While soil moisture at HMV was stable at or below 50% WRC (maximum water retention capacity), it varied between 50 and 70% WRC at SR. This soil moisture condition is usually regarded as optimal for soil carbon turnover and therefore recommended for incubation experiments (Isermeyer, 1952). It can be concluded, that at site SR the combined effects of a warm summer (increased soil temperature and lowered soil moisture) shifted the soil ecosystem to a state optimal for soil respiration.
Table 3 Cumulated soil respiration (g C m2) for period from june - august in absolute values and as fraction of annual turnover (%). 1993 HMV SR
~lCm2 494 311
1994 % 44 45
~lCm2 478 467
% 42 68
565
Jun
125 t t,J D~ 3
--
Aug
199,,.3
Jun
--
Aug
1994-
HMV
100
q
75
o--__.9~_
~o
- j e
~ .
5O '
..... -a 9'~ I
'
1
'
'
I
'
'
I
20
25
V D
20-
E 0 ~150
,4 E
15 m
e m
0
E~ I0I I
5
'
180
21 0
24
'
!
'
'
1 80
Julian Days
I
'
210
I
240
Julian Days
Figure 1. Respiration rates (heavy line), soil temperature in 5 cm depth (thin line) and soil moisture (dotted line) expressed as fraction of max. water retention capacity (100% max. WRC = pF 0.6) at site HMV. Jun
o I ~ 11251 00 3:
--
Aug
S~.R~ - - ~ \
75
199,..3
/
[
Jun
--
Aug
1994-
"/"
L
/
5O ,
,
,
,
,
,
2O
.- 25 "~
20 15..~~
~
10 5
I
1 80 dulian
I
21 0 Days
[
2-'
,
I
1 80
21 0
24-0
Julian Days
Figure 2. Respiration rates (heavy line), soil temperature in 5 cm depth (thin line) and soil moisture (dotted line) expressed as fraction of max. water retention capacity (100% max. WRC = pF 0.6) at site SR.
566
4. CONCLUSIONS
Soil respiration on the northwestern slope of the terminal moraine of the WLirmian glaciation in a temperate grassland area in southwestern germany has been monitored since May 1993. A comparison of soil CO2 efflux between the "standard" summer of 1993 and the warm summer of 1994 showed no difference in CO2 - efflux at the site HMV on the upper slope of the moraine. At the site SR on the basis of the moraine, cumulated respiration was increased by 50% and reached 68% of the annual carbon turnover in 1994 versus 45% in 1993. Lower soil moisture contents at both sites during summer 1994 led to a moisture deficit uphill, but shifted soil moisture content in the basin into the optimum range for carbon turnover. Therefore, the following can be stated: 1. Carbon-rich soils do not necessarily respond to higher temperatures with higher respiratory activity. 2. Increased physiologic activity of a soil requires not only higher temperatures, but also adequate water supply. 3. Soil respiration will increase dramatically (50% over a three month period) if both soil temperature and soil moisture are shifted to optimum conditions. 4. Within a distance of 200 meters, soils differed considerably in their sensitivity to climate. Attempts to model carbon turnover on a large (global) scale have to take this into account.
5. REFERENCES
EDWARDS, N.T. 1982: The use of soda lime for measuring respiration rates in terrestrial systems. Pedobiologica 23:321-330. ISERMEYER, H., 1952: Eine einfache Methode zur Bestimmung der Bodenatmung und der Karbonate im Boden. Z. Pflanzenern~ihr. DLing., Bodenkde. 56:26-38. JOHANSSON, G. 1992: Release of organic C from growing roots of meadow fescue (Festuca Pratensis L.). Soil Biol. Biochem. 24:427-433. KLEBER, M., STAHR, K. und I. HENNING-MULLER, 1994: The influence of exposure time on the magnitude of soil respiration under employment of the Lundegardh-procedure. Journal of Plant Nutrition and Soil Science (in print). RAICH, J.W., und W.H. SCHLESINGER, 1992: The global carbon dioxide flux in soil respiration and its relationship to vegetation and climate. Tellus 44B 2:81-99 SCHLESINGER, W.H., 1991: Biogeochemistry: an analysis of global change. Academic Press, San Diego: 443 S. SCHLICHTING, E. und H. P. BLUME, 1966: Bodenkundliches Praktikum. - Paul Parey, Hamburg, Berlin: 209 S. SPAARGAREN, O.C. (ed.) 1994: World reference base for soil resources. International soil reference and information centre, Wageningen/Rome: 161 pp. TESAROVA, M., und J. GLOSER, 1976: Total CO2 output from alluvial soils with two types of grassland communities. - Pedobiologica 16: 364-372.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
567
Remote Sensing based modelling of the terrestrial carbon cycle of Europe G.W. Heil a, J.C.J.H. Aerts "~, A.M.J.V. van BoxteP, W.P.A. van Deursen ", R. Leemans b and J.G. van Minnen b "~Resource Analysis, Zuiderstraat 110, 2611 SJ Delft, The Netherlands bNational Institute of Environmental Protection, P.O. Box 1, Bilthoven, The Netherlands
Abstract Using simulation models on terrestrial carbon cycling is an important way to get sufficient insight into impacts of climate change. Remote Sensing (RS) techniques have been used to develop an operational method to derive Net Primary Production (NPP) data for European ecosystems. The method has been developed on the concept of finding so-called fingerprints of photosynthetic activity of different ecosystems. Normalized Difference Vegetation Index (NDVI) values from satellite images were used to calculate NPP values of these ecosystem per grid cell of 20x20 km 2. In a later phase, the methodology will be applied to a grid cell size of l xl km 2 and the results will be implemented in the terrestrial carbon submodel of IMAGE-2.
1. INTRODUCTION The increase of CO2 is the product of a series of interactions between terrestrial ecosystems, oceans, atmosphere, and human activities. As a result of changes in land use, e.g. by urbanization, the terrestrial carbon cycle will be disturbed, and as a consequence climate change will be influenced. Reversely, climate change will influence net primary production of ecosystems, and this might lead to changes in natural vegetation patterns [ 1]. A methodology has been developed on the application of RS-techniques to obtain data of biomass by NDVI-values [2]. It
Carbon cycle module of
IMAGE II
Figure
i. O v e r v i e w
of the m e t h o d
568 has been shown that these NDVI-values could be used to determine stocks of carbon inecosystems, which is important to calibrate carbon cycling models on terrestrial ecosystems. Carbon cycling models are used to obtain better insight into impacts of Yearly NDVI curves climate change. (6 ecosystems) 40 However, these type of -=- Agriculture simulation models can i -.- D e c i d e o u s f o r only be calibrated with 30 j -A- M e a d o w crude estimations for i -.- Pine f o r e s t net primary production. r ~ 20 I -v- Medit. Agric. Within the framework Z ~ / ~ ~ e Alpine meadow of the non fossil carbon 10 cycle, these net primary production data determine the reliability 0,~, to forecast effects of 1 2 3 4 5 6 7 8 9 10 11 12 Month climate change to great extent. This problem F i g u r e 2. Y e a r l y NDVI curves of 6 e c o s y s t e m s . occurs especially extrapolating small scale effects to regional areas.RS-techniques can be used to estimate photosynthetic activity cq. NPP. The aim of this RS project is to develop an operational methodology to derive NPP estimates for European ecosystems using satellite images, preferably from NOAA, to calibrate carbon cycling models, such as in IMAGE-2. For this, the study has been focused on the classification of satellite images to determine the spatial distribution of different ecosystems and on the determination of a possible relationship between NDVI values and NPP.
2. CLASSIFICATION The terrestrial carbon submodel of IMAGE-2 is based on a grid cell raster which covers the whole earth. One of the characterizations of each grid cell is its land cover type. In this submodel, 17 main ecosystems are defined. Remote sensing techniques can be used to determine NDVI values in relation to NPP rate, because these values are characteristic for the photosynthetic activity. The analysis of the curves of the multi temporal NDVI values throughout the year per land cover type showed the typical characterization of the different land cover types. As an example, NDVI curves of different ecosystems in Germany are shown in Figure 2. A comparison of NDVI curves of the same type of ecosystems in other regions of
569
egr icul Qgricul
reed
Meadow Meadou fert i Oeci d,,,,
f or
Pine for 9~ ~ E . .
._.7
Boreal
for
Medi t .
for
-----~-
M e a d o u / { o r es "~i!i:i!!!!.,i~!i
-~iiiiiiii~i~i !i
,iii~'~~-?~
"~-~
Scrub 1
....
Figure
3.
Classification
of
14
I I
Desert
defined
ecosystems
in
Europe.
western Europe show similar results. In combination with so-called training sites, the different ecosystems are classified using bi-weekly NDVI values [3]. The preliminary results of the 20x20 km2 ecosystem map of western Europe are shown in Figure 3.
3. NET PRIMARY PRODUCTION An analysis on the relationship between annual cumulative NDVI and annual production was performed for different sites in Europe. From these results, a significant relationship was shown between NDVI and NPP. However, it was also shown that this relation differs for different regions. For this climatic data were used to adjust the NPP values. The results are shown in Figure 4. The results of the annual NDVI values for various ecosystems in different regions in Europe are in agreement with the expectations. For example fertilized meadows always have a relatively high NPP, as a result of an elongated growing season due to a high nutrient availability.
4. CONCLUSIONS The results of this study show the applicability of Remotely Sensed data to determine the spatial distribution of NPP of the defined ecosystems in Europe. Different methods have
570
0
-
100-
200
........
100
200 -
450
iiiiiii!iiitiiiiiiiiiii!~ii~ ~ .......
i!!!ii!!]~ii!iiiiiiiiiiii!iiiiiiiiiiiiiiiiiiiiiiii!iit,,~,, ~ o iiiii!i!iiii!iiiiii!t,,i~i~~
zoo
zoo - soo
!l!iiiillilliil~i~Ji;#.~41[iii!iiiiiiiii~-~ml~.':.'~
'::!:!i:~!%iiiii!!ll!iiil~:"~/"-41~..~,.r
. 9
I150- 1350 J,~mm~]J~-~
1350 - 1550
............................ 1550 - 1800 ~@~4~'{~ii!JiiHiilJ~!!!i!i 1800 - 2050
dl,
......!,~'":"~'~' ~
2050
8 ~ ~ ~i~
2 1 5 0 - 2250
2150
........................ ii{i!i!ii~ii 2250-2,~50 ' ~ : [4i!!!i!{i!!iii{i:~:~ii'~il 2 4 5 0 - 2700
Figure
4. M a p
of net p r i m a r y
production
of d i f f e r e n t
/
1
I
ecosystems.
been tested to obtain good results within certain reliability limits. The most reliable method tumed out to be a combination of a supervised classification with training sites and an unsupervised classification. The results will be validated and implemented in the terrestrial carbon submodel of IMAGE-2 in a further phase of this project.
5. ACKNOWLEDGEMENTS This study has been commissioned by the Dutch Remote Sensing Board (BCRS), project no. NRSP-2 4.2/TO-08.
6. REFERENCES
R. Leemans. Possible changes in natural vegetation patterns due to a global warming. In: A. Hackl (ed.), Der treibhauseffekt: das Problem- Mogliche Folgen - Erfordliche Massnahmen. Akademie fur Umwelt und Energie, Laxenburg, Austria, (1989) 105- 122. G.W. Heil, W.P.A. van Deursen, R. Bobbink, N. Trigo Boix and W. Dijkman. Remote Sensing biomass and land use: modelling the non fossil carbon cycle. Ambio (submitted). Resource Analysis. Non-fossil carbon flux modelling for Europe. Interim report II (1994).
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
573
The integrated CH 4 grassland project: Aims, coherence and site description R. Segers a and A. van Dasselaar b aDepartment of Theoretical Production Ecology, Wageningen Agricultural University, P.O. Box 430, 6700 AK Wageningen, The Netherlands bDepartment of Soil Science and Plant Nutrition, Wageningen Agricultural University, P.O. Box 8005, 6700 EC Wageningen, The Netherlands
Abstract The integrated CH 4 grassland project aims to understand and quantify methane formation and consumption in grasslands on peat soils, and the resulting net fluxes of methane between grassland and atmosphere. In two subprojects methane production and consumption are studied in the laboratory. Field fluxes and major environmental variables are monitored in a third subproject. In a fourth subproject methane production and consumption is related to methane fluxes via a process based model. Both intensively managed, drained grasslands and extensively managed grasslands in a nature preserve with a water table near the surface are studied.
1. I N T R O D U C T I O N AND AIMS
Soils can act both as a source and as a sink of methane. The contribution of soils to the global methane balance is significant (19-39% of the total source (rice paddies excluded) and 39% of the total sink; IPCC, 1992). The uncertainty about net fluxes is large as, with the current knowledge, it is difficult to relate methane fluxes quantitatively to management factors and basic environmental variables, like water table, temperature, electron acceptors and vegetation (e.g. Hogan, 1993). The integrated CH 4 grassland project aims to understand and quantify the methane formation and consumption in grasslands on peat soils, and the resulting net fluxes of methane between grassland and atmosphere. To achieve this aim a case study was done on peat soil at both intensively managed, relatively well-drained grasslands and extensively managed grasslands in a nature preserve with a water table near the surface. This paper describes the coherence of the four subprojects. Furthermore, the main characteristics of the five study sites are described.
574 2. COHERENCE In Figure l the set-up of the integrated CH 4 grassland project is given. Methane production is studied at the department of Microbiology (Wageningen Agricultural University, WAU) by laboratory experiments with soil samples (Kengen and Stams, this volume). Methane oxidation is studied at the department of Industrial Microbiology (WAU) by laboratory experiments with soil samples (Heipieper and De Bont, this volume). Field fluxes and major environmental variables are monitored by the department of Soil Science & Plant Nutrition (WAU) and NMI (Van Dasselaar and Oenema, this volume). Using the results of the small scale laboratory experiments, the department of Theoretical Production Ecology (WAU) develops process based models to describe production and oxidation. Transport models are used to relate the small scale models to field fluxes (Segers and Leffelaar, this volume).
/
,,
t experimental1" NMI, Soil Science &
PI. Nutrition, WAU
,,
%
/
%
(methane methane production
environrnenta t,actors transport A
t
"-
methane consumption
' % I I I
experimental:
%
modelling"
Theoretical Production Ecology, WAU
I"
Microbiology,
i WAU
I I experimental: I
Industrial Microbiology,"1 WAU
Figure 1. Set-up of the integrated CH4 grassland project.
The project participants cooperate via the integrated CH4 grassland working group. There is also cooperation with scientists working on nitrous oxide emissions at the intensively managed, drained sites, as the same soil physical and biological processes affect the fluxes of both trace gases. Besides, research methods show similarities.
575
3. S I T E D E S C R I P T I O N
3.1. Introduction Grasslands cover more than 35% of the total surface area in the Netherlands. About 30% of the grasslands are situated on peat soils. Our study sites include both intensively managed, drained grasslands and extensively managed grasslands in a nature preserve. They are located in the major peat area of the western part of the Netherlands, around Zegveld (52o07'N, 4o52'E). In the Netherlands average air temperatures vary from 1 ~ in January to 17 ~ in July and August. Precipitation is distributed homogeneously over the year, with an average of 60-70 mm month -1 (K6nnen 1983). 3.2. Intensively managed, drained grasslands on peat soil At the experimental farm R.O.C. Zegveld, two typical sites have been chosen: - one with a mean ground water level of 35 cm (site 8B); - one with a mean ground water level of 50 cm (site Bos 6). The vegetation of the sites is dominated by perennial ryegrass (Lolium perenne L.). Ground water levels vary greatly during the year. The soil of both sites consists of clayey peat (Table 1). More information about these sites is given by Schothorst (1982) and Otten (1985).
Table 1 Chemical and physical properties of the 0-20 cm layer of the soils at Zegveld (Velthof and Oenema, 1994 and Van Dasselaar, not published), site 8B: relatively high ground water level; site Bos 6: relatively low ground water level. Property
Site 8B
Site Bos 6
Loss on ignition, % Clay, % Total C content, g kg -l pH-KC1 Ground water level, cm, mean range
38 28 156 5.0 35 2-70
45 29 223 4.7 50 15-85
At both sites there were three different treatments: (i) mowing, no nitrogen (N) application; (ii) mowing, N application; and (iii) grazing, N application. Fertilizer N was applied as calcium ammonium nitrate in six or seven dressings. Cumulative application rates were on average about 400 kg N ha -1 yr -1 for site 8B and about 350 kg N ha -1 yr -1 for site Bos 6.
3.3. Extensively managed grasslands on peat soil in a nature preserve In the Nieuwkoopse Plassen area (a nature preserve close to Zegveld) three typical sites have been chosen: Koole, Brampjesgat and Drie Berken Zudde (Table 2). The vegetation of these sites is dominated by grass, moss, sedges, rushes and reed. It is mown once every year in
576 summer. The sites Koole and Drie Berken Zudde have not been fertilized for more than 20 years and before that have only received incidentally some farm yard manure. Brampjesgat receives every second year some farm yard manure.
Table 2 Characteristics of the 0-20 cm layer of the sites in the Nieuwkoopse Plassen area in the period January - June 1994 (Van Dasselaar, not published). Property
Koole
Brampjesgat
Drie Berken Zudde
Ground water level, cm, mean range pH-H20 Loss on ignition, %
5 0-15 4.7 55
10 5-20 5.3 50
15 10-20 3.9 90
4. REFERENCES Heipieper, H.J. and J.A.M. De Bont, Methane consumption by indigeneous grassland microflora, this volume. Hogan, K.B., Current and future methane emissions from natural sources, US Environmental Protection Agency. Office of Air and Radiation, Washington, 1993. IPCC, Climate Change 1992. The supplementary report to the IPCC scientific assesment, 1992. Kengen, S.W.M. and A.J.M. Stams, Methane formation by anaerobic consortia in organic grassland soils, this volume. K6nnen, G.P., Het weer in Nederland, Zutpen, The Netherlands, 1983. Otten, W., Nader onderzoek naar oxidatie van veengronden, literatuuroverzicht en metingen aan veenmonsters, ICW, Wageningen, 1985. Schothorst, C.J., Proceedings of the symposium on peat lands below sea level (1982) 130-163. Segers, R. and P.A. Lefffelaar, Methane fluxes from and to a drained grassland on a peat soil: modelling methane production, this volume. Velthof, G.L. and O. Oenema, In: J. van Ham et al. (eds), Non CO 2 Greenhouse Gases, (1994) 439-444. Van Dasselaar, A. and O. Oenema, Effects of grassland management on the emission of methane from grassland on peat soils, this volume.
Acknowledgements This project was partially financed by the Dutch National Research Programme on Global Air Pollution and Climate Change. The authors thank O. Oenema, P.A. Leffelaar, S.W.M. Kengen H.J. Heipieper, and R. Rabbinge for their comments on an earlier version of this paper.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
Effects of grassland management o n p e a t soils
577
on the emission of methane from grassland
A. van Dasselaar a and O. Oenema b aDepartment of Soil Science and Plant Nutrition, Wageningen Agricultural University, P.O. Box 8005, 6700 EC Wageningen, The Netherlands bNMI, P.O. Box 8005, 6700 EC Wageningen, The Netherlands
Abstract Net methane (CH4) emissions from managed grassland on peat soils in the Netherlands have been monitored with vented closed flux chambers in the period January - June 1994. Net CH 4 emissions from two intensively managed grasslands were low, in general less than 0.1 mg CH 4 m -2 d -1. On these sites, the effect of management was negligibly small. CH 4 emission from three extensively managed grasslands in a nature preserve ranged from 0 to 185 mg CH 4 m -2 d -1. The results presented here indicate that CH 4 emissions are 2-3 orders of magnitude higher on extensively managed grasslands than on intensively managed grasslands.
1. I N T R O D U C T I O N Methane (CH 4) is a greenhouse gas and as such it contributes to the greenhouse effect. Soils can either be a source of CH 4, for example in the case of wetlands and rice paddies, or a sink, for example in the case of well-drained soils (Van Amstel, 1993). Grasslands are generally considered to be a net sink for atmospheric CH 4, especially when they are well-drained. The CH 4 consumption rate depends on grassland management (Mosier et al., 1991) and environmental conditions (Bartlett and Harris, 1993). Peat soils often show CH 4 emission, as they have a high organic matter content and are anoxic at some depth (Bartlett and Harris, 1993). Agricultural utilized peat soils are partially drained, so the oxic top layer is much thicker than in undrained peat soils. As a consequence, the sink-source balance of these soils for CH 4 alters drastically. In the Netherlands, about 30% of the grasslands are situated on peat soils. It is unknown whether these sites emit or consume CH 4. To assess the effects of grassland management on CH 4 emission rates from grassland on peat soils, a monitoring study was conducted. This study investigates both intensively and extensively managed grasslands.
2. M E T H O D S Five grassland sites in the major peat soil area of the western part of the Netherlands have been investigated (for a more detailed site description: see Segers and Van Dasselaar, this volume):
578 two typical sites on intensively managed, reasonably well-drained grassland at the experimental farm ROC Zegveld: * site 8B, with a mean ground water level of 35 cm; * site Bos 6, with a mean ground water level of 50 cm. At both sites there were three different treatments: (i) mowing, no nitrogen (N) application; (ii) mowing, N application; (iii) grazing, N application. three typical sites on extensively managed grassland in a nature preserve, the Nieuwkoopse Plassen area: * Koole, with a mean ground water level of 5 cm; * Brampjesgat, with a mean ground water level of 10 cm; * Drie Berken Zudde, with a mean ground water level of 15 cm. The vegetation of these three sites is quite diverse, but consists mainly of grass, moss, sedges, rushes and reed. They are mown once every year in summer. Net CH 4 emissions have been monitored with vented closed flux chambers (Hutchinson & Mosier, 1981) from autumn 1993 onwards. In general measurements took place once every week or every two weeks with six flux chambers at each site. In the Nieuwkoopse Plassen area, boardwalks and steelen frames were installed to prevent artificially induced fluxes due to the very soft topsoil. Gas samples were taken from the headspace of the chambers with glass syringes and analysed for CH 4 by gaschromatography (relative standard deviation: 0.08%). Monitoring will continue till November 1995. Results obtained in the period January - June 1994 are presented here.
3. RESULTS
3.1. Intensively managed, drained grasslands in Zegveld In Zegveld, net CH 4 emissions were low, in general less than 0.1 mg CH 4 m -2 d -1. Effect of mean ground water level in the range of 35 to 50 cm was negligible; site 8B gave equal or only slightly higher net CH 4 emissions than site Bos 6 (Figure 1). There were also no clear effects of N fertilization and grazing versus mowing on net CH 4 emissions from the soil (not shown).
3.2. Extensively managed grasslands in a nature preserve Net CH 4 emissions from extensively managed grasslands in the Nieuwkoopse Plassen area ranged from 0 to 185 mg CH 4 m -2 d -1 (Figure 2). Differences between the different sites were quite large, as were the spatial variations at each of the sites. Drie Berken Zudde, the site with the lowest CH 4 emission, had a lower ground water level than the two other sites.
4. DISCUSSION Net CH 4 emissions were low on the intensively managed grasslands (Figure 1). Literature data for comparable sites range from 0.1 for a poorly drained grassland soil in winter (Jarvis et al., 1993) to 0.8 mg CH 4 m -2 d -1 for an unfertilized pasture (Mosier et al., 1991). Soil analyses in Zegveld showed relatively high nitrate and sulphate concentrations in the soil, especially in the top soil. Both nitrate and sulphate will have blocked CH 4 production. Low
579
Zegveld, 1994
mg OH 4 m-2d -1 0.8
0.6
0.4-
0.2-
U~ -0.2-
-0.4
T
Febr
Jan
I--
i
March
I .
.
April
site Bos 6
-s
.
.
.
.
T
-
May
-
-
-
-
June
site 8B
Figure 1. Time course of mean CH 4 emissions (in mg CH 4 m -2 d -1) from intensively managed grassland with a mean ground water level of 50 cm (site Bos 6) and 35 cm (site 8B).
Nieuwkoopse Plassen, 1994
mg OH4 m- 2 d- 1 200 = 1751501251007550-
0
)
---
i
Jan
Febr -~
Koole
March
April
-0-
+
DBZ
May
June
Brampjesgat
Figure 2. Time course of mean CH 4 emissions (in mg CH 4 m -2 d -1) at three different sites in the Nieuwkoopse Plassen area: Koole, Drie Berken Zudde (DBZ) and Brampjesgat.
580 soil temperatures in winter and spring will also have contributed to low microbial activities in the soil. Grassland soils with a high ground water level generally have a relatively thin aerobic layer. These soils are expected to emit more CH 4 than grassland soils with a relatively low ground water level and a relatively thick aerobic layer. However, site 8B in Zegveld (relatively high ground water level) gave equal or only slightly higher net CH 4 emission than site Bos 6 (relatively low ground water level) (Figure 1). Nitrogen fertilization may decrease CH 4 consumption (Mosier et al., 1991; HUtch et al., 1993); mowing or grazing could affect CH 4 emissions by influencing the amount of organic material and nitrogen that is added to the soil annually. However, there were no clear differences between the treatments at the two sites. It has to be emphasized that CH 4 production by cattle is not included in these estimates. Results presented here indicate that, for intensively managed grasslands, the effect of management on CH 4 emissions is negligibly small. Compared to the intensively managed grasslands, net CH 4 emissions from the extensively managed grasslands in the Nieuwkoopse Plassen area were 2-3 orders of magnitude higher. In the period January - June 1994, CH 4 production ranged from 0 to 185 mg CH 4 m -2 d -1 (Figure 2). Literature data also show great variations in CH 4 emissions from wetlands. In a review, Bartlett and Harris (1993) arrive at a mean estimate of 87 mg CH 4 m -2 d -1 for boreal wetlands (standard error of mean: 18; range 0-664 mg CH 4 m -2 d -1. Even in a relatively small area as the Nieuwkoopse Plassen differences between sites were quite high. The data suggest that ground water level is a major controlling factor. The national government intends to set aside intensively managed grasslands and turn them into more natural ecosystems. The ground water level of these grasslands will then be raised again. As the CH 4 emission from extensively managed, 'natural' grasslands was 2-3 orders of magnitude higher than from intensively managed grasslands, the contribution of peat soils in the Netherlands to the total CH 4 emission will then increase significantly.
5.
ACKNOWLEDGMENTS
The authors gratefully acknowledge the support of ROC Zegveld and Vereniging Natuurmonumenten. This investigation was supported financially by the Dutch National Research Program on Global Air Pollution and Climate Change.
6. R E F E R E N C E S
- Bartlett, K.B. and R.C. Harris, Chemosphere 26 (1993) 261-320. - Jarvis, S.C., D.R. Lockyer, G. Warren, D.J. Hatch and G. Dollard, In: L. 't Mannetje and J. Frame (Eds.), Proc. of the 15th Gen. Meeting of the Eur. Grassl. Fed. (1994) 408-412. - HtRch, B.W., C.P. Webster and D.S. Powlson, Soil Biol. Biochem. 25 (1993) 1307-1315. - Hutchinson, G.L. and A.R. Mosier, Soil Sci. Soc. Am. J. 45 (1981) 311-316. - Mosier, A., D. Schimel, D. Valentine, K.Bronson and W.Parton, Nature 350 (1991) 330-332. - Van Amstel, A.R., R.J. Swart, M.S. Krol, J.P. Beck, A.F. Bouwman, K.W. van der Hoek, Methane the other greenhouse gas, RIVM report 481507001, Bilthoven, 1993.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
581
The i n t e g r a t e d CH4 grassland project: Methane consumption by indigenous grassland microflora H.J. Heipieper and J.A.M de Bont Division of Industrial Microbiology, Department of Food Science, Wageningen Agricultural University, P.O. Box 8129, 6700 EV Wageningen, The Netherlands.
Abstract Soil samples were taken from the test farm in Zegveld. In batch cultures the kinetics of methane oxidation by soil from different depths were investigated. Soil was incubated in 300 ml flasks with 1, 10, 100 and 10.000 ppmv methane, respectively. All 4 applied concentrations of methane were biologically degraded by this type of grassland soil. The highest oxidative activities, especially for lower concentrations (1-100 ppmv), were observed between 5 and 20 cm soil depth. Most importantly, these experiments demonstrated that this soil acts as a sink for methane even at concentrations well below 1 ppmv. In continuous cultures soil was incubated in columns receiving a continuous gasflow of 4 ml/min containing methane at 4 different concentrations. Thereby, all concentrations of methane were degraded continuously by this type of soil.
1.
INTRODUCTION
In soils not only the formation of methane but also its degradation has an important function in the net production of this greenhouse gas. While the formation of CH 4 is a strictly anaerobic process performed by methanogenic bacterial consortia, the degradation is oxygen dependent and therefore occurs only in the upper, aerobic regions of soils. For net fluxes of CH 4 in many types of soil, methanotrophic bacteria can play a very important role, because up to 80 % of the formed CH 4 can be degraded in the aerobic zones (2,5). Methanotrophic bacteria not only oxidize methane formed in the anaerobic parts of soil. Recently, it was found that some types of soil can function as sinks for atmospheric methane (1,5). In the global CH 4 budget the microbial degradation of atmospheric methane is believed to amount to about 10-20% of the sink strength (1). In kinetic studies it was shown that bacteria with different affinities for methane are responsible for degradation of high (methanogenic) and low (atmospheric) concentrations of methane in soils, respectively (1). Within the integrated grassland project we investigate the ability of Dutch grasslands to consume both methanogenic and atmospheric methane and we try to isolate and describe the methanotrophic bacteria responsible for this degradation.
582 2.
RESULTS
Kinetic experiments. Soil samples from different depth were taken from the test farm in Zegve/d. The soil was placed in bath cultures in 300 ml flasks with gas tight septa and incubated with 1, 10, 100 and 10.000 ppm methane, respectively, in artificial air with 1 % (v/v) CO2, according to the method described by Bender and Conrad (1). All 4 applied concentrations were biologically degraded by this type of grassland soil. The time course of methane degradation is plotted in Fig. 1 for initial concentrations of both 100 (A) and 1 (B) ppmv methane. The highest oxidative activities, especially for lower concentrations (1-100 ppmv), were observed between 5 and 20 cm. There were no great differences between the oxidation profiles and kinetics between soil samples taken in autumn, spring or summer. A linear correlation between CH4-concentration and degradation rates was observed (0,003 - 30 nmol g soil ~ h 1 - 0.19 nmol - 1.9 ~mol g dry soil ~ d -1 for 1 - 10.000 ppmv, respectively). But most importantly, it is demonstrated that this soil acts as a sink for methane even at concentrations well below 1 ppmv.
degradation rate (pmol/h g soil)
degradation rate (nmol/h g soil) ).0 i
0.2 9
i
0.4 9
i
,
0.8
i
i
1.0 9
0.0
1
2.0
4.0
i
I
6.0 ,
80
i
/
.S
10
~"
0.6
2O
o
r-
"10 m
"~ v}
30
40
B
A .
50
1.
9
i
I
9
x
.
1
.
9
v
9
I
Fig. 1: Depth profile of the degradation rates of 100 (A) and 1 ppmv (B) CH 4 in batch cultures.
583
Continuous experiments.
For the enrichment of methanotrophic bacteria with low affinity for methane, soil (100 g) was incubated in column systems receiving a continuous gas-flow of 4 ml/min. Methane was supplied at 4 different concentrations (1; 10; 100; 10,000 ppmv). A decrease of the efflux concentration was observed after 14 days of incubation in the columns incubated with 10,000 ppmv (data not shown). In the column incubated with 1, 10 and 100 ppmv, respectively, the efflux gas concentrations remained constant for about 30 days, but then these methane concentrations were also degraded (Fig. 2).
1.0
~0-000
00~ \
0.9
\
>
E
0.8
t~
0.7
•
0.6
|
0.5
o. CL
d c
0.4
c9
0.3
0 o
\ O~ 0
t"-
9.~
E
0.2
0.1 0.0
=
I
30
=
1
60
=
I
90
=
!
=
120
I
150
~
1
180
i
I
210
,
,!
240
time ( d a y s )
Fig. 2: Time course of the degradation of I ppmv CH 4 in a soil column filled with 100 g soil with a continuous gas flow of 4 ml/min. These installations were continuously running for about 300 days and demonstrated that also very low concentrations of methane were continuously degraded by this type of grassland soil. These enrichment cultures will be taken to isolate methanotrophic bacteria which are responsible for the degradation of methane at different concentrations.
584 3.
DISCUSSION
A relatively high methanotrophic activity was observed in depths between 5 and 20 cm. Similar results have also been reported by other authors (3,6). Such depth profiles of methanotrophic activities can be explained by the high density of organisms in the higher layers and/or by simultaneous presence of sufficiently high concentrations of the two substrates, oxygen and methane in the region of the soil were the greatest activities are measurable. The fact that the fluxes measured for Zegveld soil are very low and even negative can be explained by the fact that there is a balance between CH 4 formation and degradation in this grassland soil (see reports of the other members of the Integrated grassland
project). The evidence for different types of methanotrophic bacteria, a low affinity type for the degradation of high concentrations of methane and a high affinity type for the degradation of low (atmospheric) concentrations of methane were described in kinetic studies by Bender and Conrad (1). But until yet, the high affinity strains could not be isolated and described (4). Therefore, isolation and identification of the strains enriched in the continuous enrichment systems especially at low CH4-concentrations will be the aim of the second part of our work.
4. 1 2 3 4
REFERENCES
M. Bender and R. Conrad, FEMS Microbiology Ecology, 101 (1992) 261-270 J.A.M. de Bont, K.K. Lee, D.F. Bouldin, Ecol. Bull. (Stockholm), 26 (1978) 91-96 H.A. Jones and D.B. Nedwell, FEMS Microbiology Ecology, 102 (1993) 185-195 G.M. King, in: J.C. Murrell and D.P. Kelly (Eds.) Microbial growth on C-1 compounds. Intercept Ltd., Andover., (1993) 303-313 5 S.C. Whalen and W.S. Reeburgh, Nature, 346 (1990) 160-162 6 S. Schnell and G.M. King, Appl. Environ. Microbiol., 60 (1994) 3514-3521
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
METHANE FLUXES FROM AND TO A DRAINED GRASSLAND P E A T SOIL: M O D E L L I N G M E T H A N E P R O D U C T I O N
585
ON A
R. Segers and P.A. Leffelaar Department of Theoretical Production Ecology, Wageningen Agricultural University, PO Box 430, 6700 AK Wageningen, The Netherlands. Email: REINOUD.SEGERS@
[email protected]
Abstract A process based model for methane fluxes from or to a drained grassland on a peat soil was set up. The methane production submodel was developed with the aid of experimentally determined time series of methane and acetate in anaerobically incubated soil samples. This submodel was calibrated by the first part of the time series and validated by the second part. The modelling activity indicated that, for explaining field fluxes, knowledge about the decay of methanogenic bacteria during aerobiosis is needed, as well as knowledge about the role of alternative electron acceptors and about anaerobiosis dynamics.
1. I N T R O D U C T I O N At a drained grassland on a peat soil in Zegveld, The Netherlands, methane fluxes have been monitored by Van Dasselaar and Oenema (1994) during one year. They found that emission did not exceed 0.3 mg m -2 d-1. In this paper a process based model to describe methane fluxes at the studied site is presented. Emphasis will be on the first results of the methane production submodel. With the model we intend to investigate, whether low emission, even during wet periods, can be explained by the duration of the anaerobic periods, which could be too short to allow the methanogenic consortium in the top soil to grow sufficiently to produce significant amounts of methane. The framework of this study and the studied site are described elsewhere in this volume (Segers and Van Dasselaar, 1994).
2. MODEL SET UP The model comprises four submodels: water dynamics, gas transport, methane production and methane consumption (Figure 1). Water dynamics is mimicked by Richards' Equation using the Van Genuchten Parameters (Van Genuchten, 1980) from the Staring Series (W6sten et al., 1994) for the soil water retention and conductivity curves. The water model will be validated with measured water contents. From the water model gas filled porosities as a function
586 of depth will be derived and used as input for the gas transport model. Initially we will assume that each soil layer is homogeneous and that gas transport does occur only in the soil gas phase by diffusion. Analysis of first results could lead to the study of soil structure and the inclusion of gas transport in (nearly) water saturated aggregates and soil respiration to describe partial anaerobiosis. Methane consumption will be included to describe consumption of methane, that is produced in the soil or that is withdrawn from the atmosphere. The methane production submodel is described in this paper.
weather data
drainage i(/c~176
'~ soil water hydrological _ dynamics P ~ ' ~ soil physical properties ~
gas filled porosity(z,t) f
vertical
diffusion
gas tlnsport
coefficients
i
intra aggregate ~ gas transport
~
\ methane( z,x, t ) .~"
I methane production
-~methane(z,x,t ) en( z,x,t)
~
. ~ ___1 j"w.~......~ biological ~ [ k , properties
methane
consumption
1
Figure 1. Sketch of the methane model for the drained site, Zegveld. z denotes depth, x the distance from the gas continuous phase, and t the time. For clarity the influence of temperature is omitted. However, the influence of temperature(z,t) on biological processes will be incorporated.
3. M E T H A N E P R O D U C T I O N The methane production model is based on theoretical insights and experimental data of Kengen and Stams (pers. comm.). They collected soil samples from various depths, made slurries, and incubated these samples anaerobically at 15 ~ Experimental details are given in Kengen and Stams (1994). The quickest and highest methanogenic activity was observed in the samples from the 0-5 cm layer and, somewhat less, from the 5-10 cm layer. Deeper layers did
587
hardly show any methane production. Therefore, experimental and modelling research has been focussed on samples from the top soil. In Figure 2 time series of methane, acetate, and carbon dioxide upon anaerobic incubation are given. After a lag phase of several days methane in the headspace of the flasks started to increase exponentially concurrent with the accumulation of acetate. After a few weeks the increase in methane concentration became rather constant. At about the same time acetate disappeared. So, three phases can be distinguished. In the first phase methanogenesis is not observable, because a) growth is below detection limit, b) methanogens are poisoned or outcompeted as a result of the presence of electron acceptors, or c) the bacteria need time to adapt to changed conditions. In the second phase methanogenesis is limited by biomass and finally it is substrate limited. Generally it is assumed that most methane is formed from acetate or hydrogen and carbon dioxide (e.g. Conrad, 1989 or Cicerone and Oremland, 1988). Here, acetate seems to be the major substrate for the bacteria as no hydrogen accumulation was observed. This is supported by other experiments of Kengen and Stams (1994) with the same slurries
low water table (average 60 cm) 1000 gmol/
30 []o []
=.. -
(gdw),
/
-
,
9
~
on
D•n
~
,',
high water table (average 30 cm)
-~]
,
,
[]
[]
:
'I '-~-~__~._
"-
----
_ _ _ .
0.001
_
~
_
___
.......
400 l
~mol/
|
[]
(gdw) |
[]
200 ~
[] oooooO
~
100 #moll
[]
(gdw)
[]
[]
[]
-
~
[]
:
~
~
9
,,
A
~~
50 .,--9 ~ ,6 o
0
40
-
'
day 80
=
A v
120
0
40
day 80
120
Figure 2. Measured and simulated time series of CH4, Ac, and CO 2 in anaerobically incubated soil samples from the 0-5 cm layer A = CH 4 (meas),9 = Ac (meas), O = CO 2 (meas), = CH4 (sim), ......... Ac (sim) .The same data are plotted three times above each other. In the figures with the logaritmic axis zero concentrations could not be plotted. Measured values were assigned the value zero, when they were below the detection limit. Acetate was monitored less frequently than the other species.
588 Because of the time dependence of the methane production, a dynamic model is appropriate. During anaerobic periods the methane production potential will grow, during aerobiosis it will decay. If P represents production and B the methanogenic biomass, the general concept of the methane production model can be described by equation (1) and (2):
P
= fan(B,a);
P
= O;
dB > 0 dtdB < 0 dt-
anaerobiosis
(1)
aerobiosis
(2)
In this set of equations fan is a function to be determined. ~ represents a set of variables, influencing methanogenesis. Using the experimental data from Figure 2 the anaerobic part of the model (eq. 1) can be further developed (eq. 3-6). All three phases, which have been discussed earlier, are covered by the model. t < tlag:
P =0 dB= 0 dt dAc= 0 dt
t > tlag:
//B
P=~(1-Y)
1 WAc
dB_ dt - Ft B V dAc = Z
(3) (4)
~tB y
Ac l-t- l't m A c + KA c
(5)
(6)
Here, t represents the time and tlag the time during which methanogenesis does not occur yet. ~t represents the relative growth rate of, acetate using, methanogenic biomass B. Y is the yield of methanogenic biomass on acetate and WAc is the molar weight of acetate. A c is the concentration of acetate, V is the volume of the water phase,/~m is the maximum relative growth rate and KAc the Monod constant for acetate consumption. Z is the acetate production, which is assumed to be constant after the lag time. For Y and KAc values from literature have been taken: Y=0.04 (g biomass/g acetate), Pavlostathis and Gomez (1991) and KAc = 0.05 (g/m3) (assuming pH =5 and species M. barkeri), Fukuzaki et al. (1990). These parameters were determined for possibly different methanogenic, acetate using, bacteria at, certainly, different temperatures than in our case. However, given the range of the other parameters, the model is not sensitive for these two parameters. So neglecting these differences does not seem too dangerous. tlag and Z have been determined by linear regression on the data for acetate, using only the points, where methane produced was still negligible compared to the acetate produced. This implies that it is assumed that acetate is consumed by methanogenic bacteria only. In the case of low water table two data points (on day 10 and 20) and in the case of the high water table three data points (on day 10, 20 and 34) were available for this fit.
The maximum relative growth rate,/~m, and the initial amount of methanogenic biomass after the lag time, Bi, were determined from the data on methane in the headspace during exponential
589 increase of methane in the headspace. Because it is rather arbitrary during which part of the experiment exponential growth occurred, two fits have been carried out for determining maximum relative growth rate and initial amount of methanogenic biomass, using a different number of data points. However, the results of the simulations did not differ much. Therefore, only the result of one simulation is given in Figure 2.
4. RESULTS AND DISCUSSION In Table 1 the fitted parameters for the model are presented. The depth dependence seems to be stronger than the site dependence. Table 1. Fitted parameters for the methane production model. "low" refers to the site with low water table. "high" refers to the site with high water table. 0-5 means slurries derived from soil samples taken from the 0-5 cm soil layer. -1 and -2 refer to the number of the fit.
sample
llm (1/d)
tlag (d)
Z (llmol/(g dw d)
Bi (g bio / g dw soil)
low (0-5)-1 (0-5)-2 (5-10) high (0-5)-1 (0-5)-2 (5-10)- 1 (5-10)-2
0.237 0.281 0.179 0.277 0.264 0.259 0.192
7.77 7.77 30.7 9.23 9.23 29 29
1.39 1.39 0.604 1.31 1.31 0.401 0.401
3.95E-08 3.40E-08 5.05E-09 1.60E-08 1.66E-08 2.54E-08 3.82E-08
In Figure 2 the simulations with the fitted parameters are given for the two samples from the 0-5 cm layer. In the case of the low water table, the transition from the exponential to the linear phase is well described by the model. In the other case another limitation than substrate was probably present around day 40, because acetate was still present, when the increase of methane in the headspace slowed down. It is surprising that in both cases the simulated final amount of methane (after 4 months) deviates only 30% from the measured amount. So acetate production seems to have been quite constant after the lag time. With sophisticated parameter estimation methods and more data points a better estimate of the parameters could be made. However, these simulations have only been used to get an insight in the order of magnitude of the processes involved. Before focussing on the exact values and variation of parameters the relative importance of the parameters in the overall model should be clear. Especially more knowledge about the processes during the lag time and during aerobiosis (eq. 2), and (an)aerobiosis dynamics should be obtained. The large CO 2 production during the lag time should originate from some process, which needs electron acceptors. Therefore, the experiment will be repeated (by Kengen and Stams) with the monitoring of alternative electron
590
acceptors, NO 3, S O 2 - a n d Fe 3+. Possibly methanogenesis is poisoned or o u t c o m p e t e d by the presence of one or more of these electronacceptors during the lag time. To investigate the decay of the bacteria responsible for m e t h a n e production during aerobiosis, e x p e r i m e n t s with changing oxygen conditions will be carried out. Insight in anaerobiosis dynamics will be gained with the gas transport model (driven by the water dynamics). In summary, the methane production model was calibrated by the first part of the time series and validated by the second part. Together with cooperators from a related project we intend to further develop the model by understanding the black box parameter tlag and the decay of the methanogenic consortium during aerobiosis. Acknowledgements This project was partly financed by the Dutch National Research Programme on Global Air Pollution and Climate Change. The authors thank S.W.M. Kengen for providing the experimental data. The comments of S.W.M. Kengen, R. Rabbinge and the members of the PhD discussion group "Dynamics of water, nutrients and toxins in soils with different structures" of the C.T. de Wit Graduate School Production Ecology have been used to improve the paper.
REFERENCES Cicerone, R.J. Oremland, R.S. (1988). Biogeochemical Aspects of Atmospheric Methane. Global Biogeochemical Cycles 2_(4) p.299-327. Conrad, R. (1989). Control of Methane Production in Terrestrial Ecosystems. Exchange of Trace Gases between Terrestrial Ecosystems and the Atmosphere p.39-58 Andrea (Ed.) M.O.. Dasselaar, A. van; Oenema, O. (this volume). Effects of Grassland Management on the Emission of Methane from Grassland on Peat Soils. Climate Change Research: Evaluation and Policy Implications - Proceedings of International Conference. Elsevier, Amsterdam, The Netherlands Fukuzaki S.; Nishio, N." Nagai, S. (1990). Kinetics of the Methanogenic Fermentation of Acetate. Applied and Environmental Microbiology 5__6_6p. 3158-3163. Genuchten, M.Th. van (1980). A closed form equation for predicting the hydraulic conductivity of unsaturated soils. Soil Science Society Am. J. 4_44p. 892-898. Kengen, S.W.M. and Stares, A. F.J.M. (this volume). Methane formation by anaerobic consortia in grassland on peat soils.Climate Change Research: Evaluation and Policy Implications - Proceedings of International Conference. Elsevier, Amsterdam, The Netherlands Pavlostathis,S.G.; Giraldo-Gomez,E (1991). Kinetics of anaerobic treatment.. Water-Science-and-Technology. 2_4_4(8) p. 35-59. Segers, R. and Van Dasselaar, A. (this volume). The integrated CH4 grassland project: aims, coherence and site description. Climate Change Research: Evaluation and Policy Implications - Proceedings of lnternational Conference. Elsevier, Amsterdam, The Netherlands W6sten, J.H.M.; Veerman, G.J." Stolte, J. (1994). Waterretentie-en doorlatendheidskarakteristieken van bovenen ondergronden in Nederland: de Staringreeks. (in Dutch) Winand Staring Centre for Integrated Land, Soil and Water Research. PO Box 125, 6700 AC Wageningen, The Netherlands.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
591
Quantification of m e t h a n e e m i s s i o n s in the exploration and p r o d u c t i o n of natural gas and p e t r o l e u m in The N e t h e r l a n d s Hans Oonk 1 and Mari~lle Vosbeek 2 1 TNO-ME, P.O. Box 342, 7300 AH Apeldoorn, The Netherlands 2 KEMA Nederland B.V., P.O. Box 9035, 6800 ET Arnhem, The Netherlands
Abstract Methane emissions from the oil and gas industry contribute significantly to the total methane emissions. For this reason, methane emissions from this sector are further quantified for The Netherlands. This quantification, based on both a detailed engineering study and on measurements, indicates Dutch methane emissions to be about 30 to 50 ktonne higher than previously expected. The main reason for this difference is, that in this quantification emissions during exploitation and fugitive and incidental emissions are incorporated, whereas they have been neglected earlier.
1. I N T R O D U C T I O N
The exploration and production of natural gas and oil belong to the major sources of methane emissions. The magnitude of these emissions is still considerably uncertain, both world wide and in The Netherlands. Various estimates of methane emissions have been made by both government (Nielen, 1991; Elzenga en Smit, 1993) and the producing industry (in Elzenga and Smit, 1993; NAM, 1993). As indicated in table 1 results differ significantly. In these studies however several sources were not included such as possible emissions in the exploration phase, during drilling and well tests. Neither have fugitive and incidental emissions during exploitation been considered, such as those incurred during maintenance or as a consequence of accidents. The interest of formulating a climate change policy requires more and better information on methane emissions due to oil and gas production because it may affect future use. This information includes a better quantification as well as an identification of opportunities for emission reductions. A joint project of TNO-ME and KEMA has the objective to provide this information.
592 Table 1
Quantification of methane emissions in the oil and natural gas production (in ktonne/y)
Government
Industry
Industry
1989
1990
1992
14 69
15 99
83
114
Year Total Total Total Total Total
onshore offshore gas prodn. oil prodn. sector
40- 70 1 - 35 41 -105
2. M E T H O D O L O G Y
The methodology used in this study is designed to meet the nature of methane emissions of oil and gas productions. These methane emissions can be divided into three types: 1. continuous emissions due to leakages of systems used and from off-gases of various gas t r e a t m e n t installations 2. operational emissions due to periodic tests and maintenance of installations 3. incidental emissions due to failures of devices. The continuous emissions are best quantified in an engineering study; information on periodic emissions can be given as well, although less accurate. The quantification of operational and incidental emissions will mainly come from a combination of m e a s u r e m e n t s and dispersion modelling. This method is less suited for estimating continuous emissions.
3. E N G I N E E R I N G
STUDY
In an engineering study all possible sources in the production process are identified and for every single source emissions are quantified. Due to the heterogenity of emission sources a variety of methods is used, for example: emission factors along with material and energy balances are used for estimating fugitive emissions during exploration maintenance and test procedures are used to assess operational emissions process simulations are used to calculate emissions from specific processequipment, e.g. from glycol dehydrators. Table 2 shows the sources and relative source strengths of methane emissions in the oil and natural gas exploration and production. Because this estimate is not completed yet, the individual source strengths are only given indicatively.
593
Table 2
S o u r c e s of m e t h a n e e m i s s i o n s in t h e oil a n d n a t u r a l gas e x p l o r a t i o n a n d production
Sources
S t r e n g t h 1)
during exploration
minor moderate/major
- drilling - well-tests
emissions during exploitation of natural gas * continuous - vents - flares - e x h a u s t g a s e s of t u r b i n e s - e x h a u s t g a s e s of r e c i p r o k i n g e n g i n e s - e x h a u s t g a s e s of f u r n a c e s - c h r o n i c l e a k s in p r o d u c t i o n - c h r o n i c l e a k s in g a t h e r i n g a n d t r a n s p o r t - glycol d e h y d r a t i o n - t r e a t m e n t of f o r m a t i o n w a t e r - u s e of p n e u m a t i c devices - condensate treatment - condensate storage - p u r g e gas f r o m v e n t i n g s y s t e m s
major moderate minor moderate minor major minor major moderate minor/moderate moderate minor moderate
* non-continuous m a i n t e n a n c e in p r o d u c t i o n m a i n t e n a n c e of g a t h e r i n g a n d t r a n s p o r t p i p e l i n e s non-exhaust engine emissions i n c i d e n t s a n d a c c i d e n t s in p r o d u c t i o n incidents and accidents pipelines
-
-
-
-
-
minor minor moderate moderate minor
emissions due to exploration of oil * continuous - f l a r i n g of a s s o c i a t e d gas - e x h a u s t g a s e s of r e c i p r o k i n g e n g i n e s - t r e a t m e n t of p r o d u c t i o n w a t e r
minor minor minor
* non-continuous -
non-exhaust engine emissions
minor
abandoned phase - chronic l e a k s f r o m a b a n d o n e d wells
none
1) minor: < 2 ktonne/y; moderate: 2 - 10 ktonne/y; major: 10 - 100 ktonne/y
594 Vents are the largest source of methane from the oil and gas industry in The Netherlands. Off-shore venting is common practice, while onshore most of the methane containing streams are flared. For this reason, most of the Dutch emissions from the industry occur off-shore. Other major sources of methane are glycol dehydrators, chronic leakages and maybe the testing of wells. The latter two sources are not incorporated in the emission inventories, as given in table 1. As a result of this, the result of this new emission estimate will be about 30 to 50 ktonne higher than the estimates in table 1. It is not a simple m a t t e r to compare emissions in The Netherlands with those reported abroad. Partly because the quantifications are not equally detailed and partly because regional circumstances make any comparison difficult. With this proviso, one can conclude from Table 3 that the relative methane emissions from oil and gas production in The Netherlands are comparable with emissions in other countries of Northwestern Europe. The emissions per cubic meter of gas produced are somewhat higher in the United States, while the emission situation in the former Soviet Union is far worse. Table 3
Emission factors (in percentage of production)
Netherlands onshore Netherlands offshore Netherlands total
0.03 - 0.05 0.6- 1.0 0.15 - 0.25
Western Europe United States Former USSR
0.15 - 0.3 0.2 - 0.3 approx. 2
4. M E A S U R E M E N T S
Since July 1991 methane is measured in Kollumerwaard, located to the northwest of the Groningen gas field. The background concentration is around 1.8 and 2.0 ppm methane, depending on season and wind direction (KEMA, 1993a, b; KEMA, 1994a). South-easterly winds bring on average the highest methane concentrations to the measuring site due to sources in a large part of the European continent. Once or twice a month elevated methane concentrations are measured at Kollumerwaard, most likely due to anthropogenic emissions of methane probably from onshore natural gas production activities. Mobile measurements were carried out to support the continuous measurements in Kollumerwaard. Three campaigns were carried out, each covering one day. The campaigns were performed around the Groningen gas field and consisted of several natural gas production locations and several exploration locations. During all three campaigns no elevated methane concentrations were measured. Under normal circumstances it is not possible to measure elevated methane concentrations near production or exploration locations.
595 Estimates of methane emissions, using the continuous measurements in Kollumerwaard from July 1991 until December 1993, were made in the following way: screening methane concentrations for events with elevated concentrations rejection of extreme events due to meteorological conditions identification of production locations possibly causing the events identification of exploration locations possibly causing the events estimation of amounts of methane emitted with dispersion calculations. The applied method resulted in an average estimate of incidental methane emission for an exploration location (127 tonne methane per year) and for a production site (225 tonnen of methane per year) for the year 1991. Taking into account the total of exploration and production locations active in 1991 in The Netherlands, this results in an estimate of the total incidental CHn-emissions during onshore exploration and production activities in 1991 of 7.8 ktonne CH4. This estimate still has to be considered with great care. A comparison is needed with previous published emission data from the industry (see table 1). For this purpose it was assumed that the elevated methane concentrations, measured in Kollumerwaard, are caused by onshore natural gas production activities. This assumption seems to be valid, since the off-shore facilities are located far away on the continental shelf. The calculated emission of 7.8 ktonne methane is 55% of the published on-shore emission. A comparison between the calculated emission and the emission, as obtained in the engineering study (table 2) was also made. The order of magnitude of both estimates is about the same.
5.
C O N C L U S I O N S
Methane emissions from the Dutch oil and gas industry are quantified both in an engineering study and by interpretating methane concentration measurements, using dispersion models. The engineering-study proves to be the most convenient way to quantify the continuous emissions, where a good indication of incidental emissions may be obtained from monitoring measurements. National emissions are probably about 30 to 50 ktonne higher t h a n estimated before. The main reason for this is, that several significant sources, e.g. emissions in the exploration phase, fugitive emissions, were neglected in previous inventarisations.
6.
R E F E R E N C E S
Amstel A. van, et al., Methane the other greenhouse gas. RIVM-report 481507001, RIVM Bilthoven, April 1993.
596 2.
Elzenga H.E. and J.K. Smit, Inventariserend overzicht van enkele afvalstoffen en emissies uit der Nederlandse olie en gaswinning, basisjaar 1990. RIVM-report 736201027, RIVM Bilthoven, July 1993 (Inventarisation of several waste streams from the exploration and production of Dutch oil and gas, reference year 1990; in Dutch). 3. EPA, Anthropogenic methane emissions in the United States: Estimates for 1990. Report to Congress, EPA 430 R-93-003, April 1993. 4. EZ, Olie en gaswinning in Nederland, 1990; 1991; 1992; 1993. Ministry of Economic Affairs, The Hague 1991; 1992; 1993; 1994 (Exploitation of oil and gas in The Netherlands, 1990; 1991; 1992; 1993; partly in English). 5. KEMA, 1993a (Vosbeek, M.E.J.P.). Kwantificering van methaanemissies bij olie- en aardgaswinning, rapportnr. 63630-KES 93-3241 (Quantification of methane emissions in the exploration and exploitation of oil and gas; in Dutch). 6. KEMA, 1993b (Vosbeek, M.E.J.P.). Evaluatie en integratie van CH4-, CO-, CO2-metingen te Arnhem en Kollumerwaard, rapportnr. 63625-KES/MLU 93-3242 (Evaluation and integration of CH4, CO and CO2-measurements at Arnhem and Kollumerwaard; in Dutch). 7. KEMA, 1994 (Vosbeek, M.E.J.P.). Methaanemissies in Nederland, concept (Methane emisisons in The Netherlands; in Dutch). 8. NAM, Milieujaarverslag 1991, NAM B.V., Assen, 1992 (Environmental year report, 1991; in Dutch). 9. NAM, Milieujaarverslag 1992, NAM B.V., Assen, 1993 (Environmental year report, 1991; in Dutch). 10. NAM, Milieujaarverslag 1993, NAM B.V., Assen, 1994 (Environmental year report, 1991; in Dutch). 11. Nielen, R., Kwantificering van methaanemissies als gevolg van aardgasverliezen en oliewinning in Nederland. TNO-report 91-210, TNOME, Apeldoorn, July 1991 (Methane emissions due to oil and gas operations in The Netherlands; in Dutch).
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
597
V a l i d a t i o n of l a n d f i l l g a s f o r m a t i o n m o d e l s Hans Oonk, Tonnie Boom TNO Institute of Environmental and Energy Technology, P.O. Box 342, Apeldoorn, The Netherlands
Abstract Three models for landfill gas formation were validated and the model parameters were estimated for Dutch landfills. Landfill gas formation could be well described, using a first-order and a multi-phase model. The half time of biodegradation was estimated to be about 7 years; ultimately 60% of the organic carbon is estimated to be converted into landfill gas.
1. I N T R O D U C T I O N When solid waste is deposited, its organic content is converted into landfill gas: a mixture of about 40-50 vol% carbon dioxide and 50-60 vol% methane. Landfills are considered to be one of the main methane sources in The Netherlands, contributing significantly to the total emissions of about 1200 ktonnes [1]. Emissions of methane from landfills may be quantified, starting from the methane material balance. Assuming no accumulation this material balance goes: Emission = F o r m a t i o n - Oxidation- Recovery In order to enable a quantification, landfill gas formation has to be described as a function of the amounts of waste, waste age and composition. For this purpose, existing models for landfill gas formation were validated and the values of the model parameters were estimated. In this paper it is described how landfill gas formation is determined for several landfill sites. This is done in two ways: by performing emission measurements, and by interpretation of the results of landfill gas formation projects. The observed landfill gas formation is subsequently correlated to the characteristics of the waste in place.
2. LANDFILL GAS FORMATION MODELS Several models for landfill gas formation are described in literature [2]. Three models are listed in table 1, along with the values of the model parameters, as
598 determined in this study. The zero-order model (1) and the first-order model (2) are the most well-known, and are frequently used in national emission estimates [e.g., 1,3]. The multi-phase model (3) is often used in the design of Dutch landfill gas recovery projects [4]. Table 1 Landfill gas formation models
- Zero order model
0~t =
~1.87 x koA
- First order model
(zt = ~1, 87ACok 1 x e -kit 3
- M u l t i - p h a s e m o d e l ~ t - ~ Z l'87ACo, ikl, i • j =1
ko = 2.4 kg tonne -1 y-1
kl
O.58 = 0.094 y-1
_kl ' it ~ = 0.58 kl, 1 =0"185y~;k1,z=0"100y 1
kl, 3 = 0 . 0 0 3 y-1
In this table (~t is the landfill gas formation in m 3 per year, ~ is the generation factor. This generation factor expresses the part of the organic carbon t h a t is ultimately converted into landfill gas. Excavations show t h a t not all of the organic material is converted into landfill gas, so this generation factor is less t h a n one. The factors kn a r e the model parameters, A is the amount of waste in place in tonne; Co and Co.i are the amount of organic carbon in the waste and the amount of organic carbon of a specific fraction in kg per tonne, respectively; t is the time in years, elapsed since the depositing of the waste; the factor 1.87 has the dimension m ~ per kg. Of the models in table 1, the zero-order model is the most simple one, a s s u m i n g landfill gas formation to be equivalent to the amounts of waste in place. The firstorder model reckons also with waste age and carbon content. This results in a better estimate of formation, provided that the amounts of waste, waste age and the origin of the waste are sufficiently well known. In the multi-phase model, a fast, a moderate and a slow degradable fraction are distinguished in the waste, each decaying in a first-order way, but with different rate-constants of biodegradation. The a m o u n t s of carbon in the waste and the degradability of the carbon (slow, moderate or fast) can be calculated, using the mean composition of landfilled waste, as it has been determined in The Netherlands. This composition is given in table 2.
599 Table 2 Carbon contents of Dutch municipal waste [4]
Fraction
C0,1
C0,2
C0,3
Degradability
fast
moderate
slow
C0,tot
(in kg tonne -1) household waste industrial waste demolition waste agricultural waste
51 9
66 51
61
61
mean
66 51 11 13
66 111 11 135 112
3. E S T I M A T I N G F O R M A T I O N F R O M E M I S S I O N S As stated in the introduction, formation models are validated by correlating the waste characteristics with the observed landfill gas formation. One way for determining landfill gas formation is by m e a s u r i n g landfill gas emissions. Landfill gas emissions are n o t affected by oxidation, since the total amount of m e t h a n e and carbon dioxide emitted remains constant. Assuming no accumulation, landfill gas formation can be estimated as the sum of the amounts emitted and recovered. Several methods for m e a s u r i n g emissions are described in literature [5]. Two of these methods were tested: the closed chamber method and the micrometeorological method [6]. In the closed chamber method, the small area (10 m 2) is sampled and emissions are directly obtained. In a micrometeorological method, m e t h a n e concentrations are m e a s u r e d along with wind-velocities at various heights. From these wind-velocity and concentration profiles, vertical fluxes from a large area (> 2000 m 2) can be calculated [7]. Landfill gas emissions proved to occur in a very inhomogeneous way: using the closed chamber method, emissions varied over a factor of 1000 at different spots on a single landfill. Besides t h a t landfill gas emissions also varied widely on a timescale. As a result of this, a large n u m b e r of m e a s u r e m e n t s on different spots, pertained for longer times is required in order to get a reliable impression of emissions. Doing this, using the closed-chamber method results in a very timeconsuming and labour-intensive procedure, because of the large n u m b e r of transpositions required. The micrometeorological method does not have this disadvantage. Using this method - equipped with a u t o m a t e d sampling, analysing and data-acquisition - it proved to be possible to measure emissions from a large part of a landfill, within about two weeks. In this way emission estimates were obtained, which were fairly well in line with expected emissions.
600 This micrometeorological method is subsequently used in a m e a s u r e m e n t campaign, measuring emissions from about 15 to 20 landfills in The Netherlands
[8]. 4. ESTIMATING F O R M A T I O N FROM R E S U L T S OF R E C O V E R Y PROJECTS At about 22 landfill sites in The Netherlands landfill gas is recovered and utilised. For recovery of the gas, a well-system is installed which the gas is collected with. The amounts of gas recovered may be used to estimate landfill gas formation, provided the recovery efficiency is known. Amounts of gas recovered, and specific information about the landfill site, and its well-system was obtained from nine Dutch landfills [9]. The recovery efficiencies were estimated by Grontmij, a consultant with over 10 years of experience with landfill gas. These estimates were based on landfill geometry, composition top liner system and lay-out of the recovery system. In several landfills the waste is completely wrapped in gas-tight liner systems and the efficiency was assumed to be 100%. From the amounts of landfill gas recovered and this estimated recovery efficiency, the landfill gas formation was estimated: the formation is the ratio of the recovery and the recovery efficiency.
5. VALIDATION Both the emission m e a s u r e m e n t s and the results of the landfill gas projects provide information about landfill gas formation on the individual landfills. On the other hand, landfill gas formation may be calculated, starting from the waste characteristics, using the formation models in table 1. Comparison of the results of the models with the observed landfill gas formation for all the landfills, provides information about the accuracy of the model. Since the calculated formation is a function of the model parameters, the model accuracy is also a function of the accuracy of these parameters. This provides a tool for estimating the values of the model parameters. Assuming a type of model (e.g. a first-order model), the best-estimate for the model-parameters yields a m a x i m u m accuracy of the calculated and the observed landfill gas formation. In this study, this m a x i m u m accuracy was obtained numerically using SAS (software for statistical analysis). The resulting best-estimates of the values of the model p a r a m e t e r s are given in table 1. The result of the first-order model is illustrated in figure 1. The mean relative difference between observed and calculated formation was 44% for the zero-order model; 22% for the first order model and 18% for the multiphase model.
601
30
J
E o
cO
v
J
E cO
20
J
E J
O
N--.
6U.. ..J "O
/
10 []
,,I,-,
/
o []
o
0
10
20
30
observed LFG-formation (million m3/y) Figure 1. Observed versus calculated formation (using the first-order model)
6. D I S C U S S I O N A N D C O N C L U S I O N S The resulting values of the rate-constants of biodegradation correspond with a half-time of landfill gas formation of about seven years. The calculated generation factor indicates, t h a t ultimately about 60% of the organic carbon is converted into landfill gas. Landfill gas formation, both on a single landfill as on a national scale m a y be calculated, using the models in table 1. Use of the zero-order model results in less reliable estimate t h a n use of the first-order or multi-phase model. The accuracy of a national formation estimate depends upon the accuracy of the amounts of waste and the waste composition. In The Netherlands, these amounts are known within an accuracy of about 10-15%. Using the first order model, as described above, will result in a formation estimate with a relative error of about 25%. It has to be stressed, t h a t landfill gas formation is affected by waste composition, landfill-site m a n a g e m e n t and climatological conditions. For this reason, the results obtained in The Netherlands can not be used for estimating landfill gas formation in other countries, without any proviso.
602 7. A C K N O W L E D G E M E N T S
This work is funded by NOP-MLK (Dutch National Research Programme Global Air Pollution and Climate Change) and NOVEM (Netherlands Agency for Energy and the Environment).
8. R E F E R E N C E S
1. 2. 3. 4. 5. 6. 7.
8. 9.
Amstel A. v., Swart R.J., Krol M.S., Beck J.P., Bouwman A.F., Hoek K.W. v.d.: 1993, Methane, the other greenhouse gas, RIVM-Report no. 481507001, RIVM, Bilthoven, The Netherlands; Augenstein, D., Pacey J.: 1991, Modelling landfill gas generation, Sardinia 91, 3 r~International Landfill Gas Symposium, 14-18 october 1991, Sardinia, Italy. US-EPA: 1993, Anthropogenic methane emissions in the Unites States: estimates for 1990, Report to Congress EPA-430-R-93-003; Bogner J. and Spokas P.: 1994, Landfill CH4 emissions, guidance for field measurements, Argonne National Laboratory, USA. Scheepers M.J.J., Zanten B. van: 1994, Handleiding stortgaswinning, Adviescentrum Stortgas, Utrecht. (Manual landfill gas recovery, in Dutch) Verschut C., Oonk J., Mulder W.: 1993, Broeikasgassen uit vuilstorts in Nederland, TNO-report 91-444, TNO-ME, Apeldoorn, The Netherlands (Greenhouse gases from landfills in The Netherlands, in Dutch). Fowler D., Duyzer J.H.: 1989, Micrometeorological techniques for the measurement of trace gas exchange, in Exchange of trace gases between terrestrial ecosystems and the atmosphere, Andreae M.O, Schimmel D.S., eds., John Wiley & Sons, pp. 189-207. Oonk J., Boom A.: 1995, Landfill gas formation, emission and recovery in The Netherlands, a measurement study, TNO-report, to be published early 1995. Oonk J., Weenk A., Coops O., Luning L.: 1994, Validation of landfill gas formation models, TNO-report no 94-315, TNO-Apeldoorn, The Netherlands.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
603
M e t h a n e e m i s s i o n of the A m s t e r d a m u r b a n area.
D. Veenhuysen and P. Hofschreuder
Department of Air Quality, Wageningen Agricultural University P.O.Box 8129, 6700 EV Wageningen, The Netherlands.
Abstract Within the Dutch National Research Programme on Climate Change, the emission of methane from urban areas was considered to be one of the uncertain sources in the national methane budget. Amsterdam was chosen to be the urban area for methane emission assessment. An emission inventory was made and concentration fields were calculated using the Danish OML model. Calculated concentration values were compared to concentrations measured continuously by ECN. The methane emission of the Amsterdam area is low. Main sources are traffic emissions (1.3 ktons.y ~) and the gas distribution network (0.45 ktons.y-~). The emission inventory tends to overestimate the total emission by a factor of about ten. 1. INTRODUCTION
Methane is one of the non-CO2 greenhouse gasses, that gets increased attention last years. This attention is triggered by the relatively short lifetime of the gas, the rapid growth rate of ambient concentrations of about 1 % per year, and the low reduction of emissions (10 %) to stabilize atmospheric concentrations [1]. The global emission of methane of 520 Tg/year, based on estimates on source and sink strength by atmospheric chemistry [1] is generally accepted. The relative importance of source strengths for particular source categories can only be assessed with large uncertainty when global models are used. Assessment of the fluxes of greenhouse gases from important and accessible sources is useful for constraining the source strengths of unknown sources and for policy making. Determination of the source strength of methane emissions with high uncertainty and detection of unknown sources was part of the Dutch National Research Program on Climate Change. An overview of important Dutch methane sources is given by van Amstel [2]. The production of gas and oil is the third largest source of methane in the Netherlands after enteric fermentation and landfills. Production of natural gas is coupled to the distribution of the gas. The dense articulated distribution networks in urban areas was considered to be a large potential source of methane by leakage. Verification of urban methane emissions was the aim of this project. Instead of a costly approach of intensive measurements, we chose for a combined program of making an emission inventory, modelling methane concentrations and monitoring methane concentrations at a fixed site. The Amsterdam urban area (1,000,000 inhabitants) was chosen to test the emission inventory and eventually detect unknown sources of methane. A permanent monitoring station was installed by ECN at Vuurtoreneiland, a little isle in a lake east of Amsterdam to have a uniform fetch and no local sources next to the monitoring site.
604 2. E M I S S I O N I N V E N T O R Y
To obtain an estimate of the source strength for methane of urban areas an emission inventory was made. The natural gas distribution network, road traffic, landfills, high industrial stacks, wastewater treatment plants and some minor sources like deep water and air traffic were considered as possible sources for methane. Dairy farming of ruminants around Amsterdam will also cause emissions due to enteric fermentation. This emission, however, was not taken into account. Their influence on immision concentrations at the monitoring site could be cancelled out by subtracting the background Cabauw methane levels from the Amsterdam data. Cabauw is high level (200 m.) measuring station in the centre of the country. Around both monitoring sites, a comparable agricultural situation exists. Although the soil in the rural surroundings was very humid, one may not consider these meadows as wetland, because the topsoil is aerobic. The topsoil is also covered with grass, which does not have like rice hollow stems to convey rapidly methane from anaerobic soil layers to the atmosphere. Table 1 list the methane emissions for the Amsterdam urban area. Most recent emission factors for gas leaks, traffic [3] and industrial sources [4] were used. Table 1 Methane emissions in the Amsterdam urban area (ktons/y).
Gas distribution network Road traffic Industry (except power station) Power station Landfills Wastewater treatment Canals
0.45 1.30 0.06 confidential minor minor ?
Emissions of the gas distribution network were based on a number of 700 leaks discovered per year, an average leak rate of 0.04 m3.h-1 and incidental venting of pipes (35 km.y -1 O 0.2 m) for reconstruction, venting of new main lines (30 km.y 1) and regulating. These emissions are low because of the high quality polyethylene tubing used. They are 0.08 % of the throughput of 751,500 m3.y-1. The methane content of the natural gas is 83 %. For road traffic detailed data on traffic density on highways were available [5] on an hourly basis. For traffic in the city districts only daily mean values were available. We used the hourly mean traffic densities for work days and weekend to obtain one hour resolution for the traffic density in the city districts. This resolution is needed for dispersion calculations as traffic density and meteorology have strong diurnal trends. Emission factors for traffic categories and percentage of catalyst equipped cars were from recent date (1993). Industrial emissions play a minor role in Amsterdam. The largest source is a power plant, but emissions take place from a high stack and do not influence the concentrations at the monitoring site. Emissions of landfills are highly uncertain, but no recent landfills were found in the sector from the Amsterdam urban area to Vuurtoreneiland. No emissions were fed to the
605 dispersion model. Waste water treatment plants produce methane, which is reused in the process. Incidently a surplus of gas occurs. This gas is flared. Emissions from deep water surfaces and canals are highly uncertain. They depend on the thickness and anaerobic situation in the sludge and the availability of a oxidizing zone in the upper part of the sludge and in the water. No emissions were fed to the dispersion model. The Airport was that far away and emission estimated so low, that these emissions were not taken into account. The total emission for the urban area was estimated to be about 1.8 kton methane per year.
3. M O D E L L I N G M E T H A N E CONCENTRATIONS
The Danish OML model [6] was selected for calculation of immission concentrations after an inventory of appropriate models for this purpose. This Gaussian plume model has a preprocessor to calculate dispersion height, atmospheric stability and turbulent mixing from synoptic and radiosonde measurements. The synoptic measurements were obtained from Schiphol Airport, the radiosonde data from de Bilt. The model can handle up to 100 emission sources of point type or area type. Line emissions like highways were implemented as multiple point sources. Traffic emissions from a district were located in the centre of the district. Emissions from gas leaks were evenly distributed over the urban area. The surface roughness is a very uncertain factor to be put in the dispersion of [H~ (ppm),231293-2802% model. A town is very I+96- _i,/ / \ heterogenous and a Gaussian model only allows for a constant value. Calculated 9942 ' . results are very sensitive to this parameter. For the comparison /,90] / / )U \\ 1 of calculated and measured data a value of 1 m. was chosen. Emission data and dispersion data from the meteorological preprocessor were fed to the OML model to calculate isoimmission concentration lines 482t ~.o/=O/-~ \ j "~e" .@'tYirJ I and concentrations at the continuous methane monitoring station at Vuurtoreneiland. Only the frequency distribution 47B ' of hourly concentrations was 476L3~ calculated because of the 111 113 11's 117 123 127 statistical nature of the model Amersfoort coordinates km and the large uncertainty in Figure 1. Long term average distribution of [CH4] individual calculated hourly calculated with OML and a roughness length of l m. values.
IV
E
/I
606 This was done for a 180-320 degree windsector, being the sector in which the monitoring station is influenced by the urban plume and results were related to stability. Long term average CH4 concentrations around Amsterdam are given in figure 1.
4. OBSERVED METHANE CONCENTRATIONS
The calculated emissions were compared with data gathered at the permanent methane monitoring site run by the Netherlands Energy Research Foundation (ECN). To obtain an estimate of the increase in ambient methane concentrations due to the urban plume, methane immission concentrations of Cabauw (some 40 km south of Amsterdam) were subtracted from the Vuurtoreneiland data. Results will be published separately.
5. RESULTS AND CONCLUSIONS
Both calculations and measurements resulted in a low increase in methane concentrations due to urban emissions (some tenths of a ppm). The measurements indicated some incidental elevated concentrations of some ppm. Although Gaussian models are not well suited to predict immision concentrations for exceptional situations, the model was used for more detailed analysis. The elevated concentrations resulted merely from extreme meteorological conditions and the vicinity of major highways south of the monitoring site. No major unknown sources of methane could be detected. Calculated concentrations were higher than measured concentrations. Elevated concentration levels showed the same magnitude. One should realize that in Gaussian dispersion modelling a margin of error of about +100% is usual. Even if this margin is considered and uncertainty in roughness length is taken into account the calculated averages are a factor of five to ten higher than the observed ones. This may be the result of a to high emission factor for traffic or a higher percentage of catalyst equipped cars and methane oxidation in the soil around gas leaks. The urban contribution to the national methane budget is small.
6. R E F E R E N C E S
1 J. Lelieveld, P.J Crutzen, C. Brfihl, Chemosphere,Vol 26,1-4 (1993). 2 A.R. van Amstel, et al. Methane the other greenhouse gas, Research and policy in the Netherlands, RIVM report no: 481507001, (1993). 3 C. Veldt, P.F.J. van der Most, Publikatiereeks Emissieregistratie Nr 10, (1993). 4 J.J.M. Berdowski, W.J. Jonker, Publikatiereeks Emissieregistratie Nr 14, (1993). 5 Adviesdienst Verkeer en Vervoer, Verkeersgegevens, Jaarrapport 1992, Ministerie V&W 6 P. Lofstom, H.R. Olesen, User's guide for OML-MULTI, An air pollution model for multiple point and area sources, MST-LUFT-A126, National Environmental Research Institute, Roskilde, Denmark, (1992).
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
607
Soil parameters controlling m e t h a n e e m i s s i o n from rice paddies. H.A.C. Denier van der Gon and N. van Breemen Department of Soil Science and Geology, Agricultural University Wageningen, P.O. Box 37, 6700 AA Wageningen, The Netherlands
Abstract The influence of soil parameters on methane emissions from rice paddies is studied in a collaborative project of Agricultural University Wageningen and the International Rice Research Institute (IRRI). Methane fluxes from wetland rice fields in the Philippines were monitored with a closed chamber technique during two wet seasons (1991 and 1992) and one dry season (1992). Gypsum addition and salinity were found to reduce CH 4 emissions whereas green m a n u r i n g strongly enhanced the CH 4 fluxes from rice fields. Laboratory experiments with planted pots showed t h a t plant-mediated CH4 transport is diffusion controlled and t h a t rhizospheric CH 4 oxidation may depend on the plant growth stage.
1. I N T R O D U C T I O N About 20% of the global anthropogenic methane (CH 4) production comes from wetland rice [IPCC, 1992]. However, the uncertainty in the contribution of rice agriculture to the global methane budget is large. To better estimate the total CH4 source strength of rice and to develop mitigation strategies, factors controlling CH 4 emission from rice m u s t be better understood. In collaboration with the International Rice Research Institute (IRRI) in the Philippines, parameters t h a t influence CH4 emission from rice fields were screened. Methane fluxes from wetland rice fields with different treatments are monitored with a closed chamber technique as developed by Schiitz et al. [1989]. This paper gives an overview of the project activities and examples of results.
2. O V E R V I E W OF S U B J E C T S (Numbers following a subject refer to Table 1) 9 9 9 9 9 9 9
Effects of sulfate on CH 4 emissions from rice fields [4, 5]. The impact of salinity on CH 4 emissions from rice fields [6]. Mechanism of rice plant-mediated CH 4 transport [2]. Influence of organic amendments on CH 4 emissions from rice fields [1, 7]. Importance of CH 4 oxidation in the rice rhizosphere [8]. Influence of soil types on CH4 emissions [4, 6, 9] Assessment of total CH 4 emission from a rice crop cycle [1, 9]
608 Table 1
Project Publications
[1]
Denier van der Gon, HAC, H-U Neue, RS Lantin, R Wassman, MCR Alberto, JB Aduna and MJP Tan, Controlling factors of methane emissions from rice fields, In: World Inventory of Soil Emission Potentials, Proceedings Internat. Workshop, ISRIC, Wageningen, 1992.
[2]
Denier van der Gon, HAC and N Vanbreemen, Diffusion-controlled transport of methane from soil to atmosphere as mediated by rice plants, Biogeochemistry 21, pp. 177-190, 1993.
[3]
Denier van der Gon, HAC, Methane emission from rice agriculture, Change 15, pp. 12-14, 1993.
[4]
Denier van der Gon HAC and H-U Neue, Impact of gypsum application on the methane emission from a wetland rice field. Global Biogeochem. Cycles 8, pp. 127134, 1994.
[5]
Denier van der Gon HAC and H-U Neue, CH4 emission from a wetland rice field Impact of gypsum application. Climate Change and Rice 14-18 March 1994, Los banos, Philippines. International Rice Research Notes (in press), 1994.
[6]
Denier van der Gon HAC and H-U Neue, Methane emission from a wetland rice field as affected by salinity, Plant and Soil (in press), 1994.
[7]
Denier van der Gon HAC and H-U Neue, Influence of organic matter incorporation on the methane emission from a wetland rice field, submitted to Global Biogeochem. Cycles.
[8]
Denier van der Gon HAC and H-U Neue, Oxidation of methane in the rhizosphere of rice plants (in prep.)
[9]
Denier van der Gon HAC, H-U Neue and N Vanbreemen, Methane emission from wetland rice fields as effected by soil types (in prep.)
3. R E S U L T S T h e project yielded both m e c h a n i s t i c knowledge of t h e rice-soil-water e c o s y s t e m a n d d a t a to improve CH4 source s t r e n g t h e s t i m a t e s of rice a g r i c u l t u r e . 3.1 L a b o r a t o r y e x p e r i m e n t s A set-up was developed to s t u d y p l a n t - m e d i a t e d CH 4 t r a n s p o r t . It w a s shown, by u s i n g b i n a r y diffusion coefficients of C H 4 in o t h e r gases, t h a t CH4 t r a n s p o r t t h r o u g h rice p l a n t s is diffusion-controlled (Table 1; [2]). F u r t h e r m o r e , CH 4 oxidation in the rice r h i z o s p h e r e w a s q u a n t i f i e d w i t h t h e use of a specific i n h i b i t o r for CH 4 oxidation. The r e s u l t s indicate t h a t the rice g r o w t h stage is a n i m p o r t a n t factor d e t e r m i n i n g the rhizospheric CH4 oxidation (Table 1; [8]). R a t e s of CH 4 emissions from i n t a c t soil cores w e r e m e a s u r e d d u r i n g oxic a n d anoxic i n c u b a t i o n s u s i n g a methodology developed by King et al. [1990] (Table 1; [6]). T h e anoxic CH 4 fluxes from soil cores of a s a l t - a m e n d e d plot w e r e 3-4 t i m e s lower t h a n from cores of the control plot, w h e r e a s the oxic CH4 fluxes w e r e a b o u t equal. T h e CH 4 production in the s a l t - a m e n d e d field w a s s t r o n g l y r e d u c e d c o m p a r e d to t h e control field b u t CH 4 oxidation in the s a l t - a m e n d e d plot w a s even m o r e i n h i b i t e d t h a n CH 4 production. The n e t r e s u l t w a s a b o u t equal oxic CH 4 fluxes from both s a l t - a m e n d e d plots a n d n o n - a m e n d e d plots.
609
3.2 Field e x p e r i m e n t s Methane emission (mg.m'~.day ") Gypm~m is a common 1,200 amendment on alkaline I Green manure Green manure I soils. Adding 6.66 tons.ha q 1,000 of gypsum (CaSO4) reduced CH 4 emission by 55-70%, 800 most likely by inhibition of methanogenesis by sulfate600 reducing bacteria [Fig. 1]. Amounts of SO 2" in the soil 400 solution of gypsum-amended plots were sufficient for 200 sulfate-reducing bacteria to outcompete methanogens. I I I I 0 20 40 60 80 100 CH4 emissions are much Days after transplanting lower from rice fields on high-sulfate containing soils F i g u r e 1. CH 4 emission f r o m w e t l a n d rice fields w i t h a n d or gypsum-amended soils w i t h o u t g y p s u m addition [Denier v a n d e r Gon a n d N e u e , t h a n from low-sulfate soils. 1994]. The presence of non-sulfate salt also depressed CH4 emission, but much less so than gypsum did. The depression of CH 4 emission CH, emission (mg.m2.day -1) by (non-sulfate) salinity is .ooo , iiiiii!i!i!i! iiiii!iill Urea ., !:~:iiiiiiiii~iii!i!iii!~iiii~i~iii!!iii::::::::':-:':':':':-: also due to inhibition of 2,500 I~ ---0--methanogenesis. However, in saline rice fields CH 4 oxidation was inhibited even $~ ".~ :i:i:!:i::::::il!i!iiiiiiiiiiiii!iiiiii~iiiiiiiiiiiiiiiiiiiiil more t h a n methanogenesis. 1.soo ~ :~ s~est i~i,!!ii!iiiiiiiiiiiiii!iiiiii!i!i!i!iiii!!ii!l * ~ end of irrlga~on :~:~:.:-:.:.:.:-:.:-:.:':-:-:.:':':':-:':-:.:-| Therefore, the reduction in ~, y : , /~i~!~ii~!~iiiii!iiiiii!iiiii!ili!i!iiiiiiiiiiili!l , ~ : , ~i~!!~ii!~!i,!i~i!!i!~i~!i!i~!~i~i~i!ii!ii~!ii~!~ CH 4 e m i s s i o n is not , ~),, ~~ ~:i:i:i:i:i:i::~"i~:i:i:i:i:i:i:!:i:i:i:i:i:i:!:i:i:i:i:i:i:i _~) i , ::::::::::::::::::::::::::::::::::::::::::::::::::::::::: 1 1O00 , (~ , ,~ @^ :::::::::::::::::::::::::::::::::::::::::::::::::::::::::: r ...:.:.:.:.: ............................................ proportional to the reduction in CH 4 production eb f ;:;. :~i~!i~i~i!!~:iiii!i!i!i!!!!i!i!i!!iiiiiiii ~iiiiiii!i!! ;~ili!iii!i!i!ii!!i~ii;i!i!i!ii!i~i (Table 1; [6]). .....!i!~i~ :i{i;i!;ii~ii~i;i;i!ii;ii!i~iii~! Fig. 2 illustrates t h a t '~ o , !!i!!!!!i!ii!!i!!!!!~!~!i!i!il 0 20 40 60 80 100 ~ 120 140 (1) CH 4 emission is strongly Days after transplanting ~eU,dry enhanced by organic amendments, (2) organic a m e n d m e n t s cause a large F i g u r e 2. C H 4 emission from w e t l a n d rice fields fertilized w i t h release of CH 4 early in the urea or green manure. season and, (3) a f t e r harvest, when the fields are drained, considerable amounts of soil-entrapped CH4 are released. In previous studies this flush of CH 4 was not included, which causes an underestimation of total seasonal emission. ,
iiiiiiiiiiiiiiii!iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiliiiiiiii
iiiiiiiiiiiiii!iiiiiiii iiiii!!ii
o
,v I
!!!i!i!ii!i!?i!ii!i!?!Nii! !!!ii!iiigii?ii! . . . .
§
610 4. G E N E R A L D I S C U S S I O N The laboratory experiments stress the importance of both CH4 production a n d CH 4 oxidation w h e n e s t i m a t i n g CH 4 emission. They show t h a t the ratio b e t w e e n CH 4 production a n d CH 4 oxidation is not fixed b u t m a y depend on e n v i r o n m e n t a l conditions. F u r t h e r m o r e , fluctuations in the m e t h a n o t r o p h i c activity in the rice rhizosphere m a y contribute greatly to the observed seasonal fluctuation in CH 4 emissions from rice fields. W h e n looking for mitigation options to reduce CH 4 emission from rice fields, both CH 4 production a n d CH 4 oxidation have to be t a k e n into account. Global e s t i m a t e s of m e t h a n e emission from rice agriculture do not yet t a k e differences in soil types into account. Bachelet and Neue [1993] found t h a t emission e s t i m a t e s were about 25 % smaller if soil differences were considered. So, including soil characteristics could significantly alter the global estimate. I n f o r m a t i o n on CH 4 emission from rice fields on different soil types is scarce. Our r e s u l t s indicate t h a t "soil type" correction factors can be used to e s t i m a t e CH 4 emissions from flooded rice fields. For example, we showed t h a t CH 4 emissions from w e t l a n d rice fields on saline, low-sulfate soils are lower t h a n CH 4 emissions from otherwise comparable non-saline rice fields. On soils n a t u r a l l y high in sulfate or softs a m e n d e d with large a m o u n t s of sulfate-containing substances m e t h a n e emissions are reduced even more. Fertilization of rice fields with (NH4)2SO 4 will not depress CH 4 emissions m u c h because the a m o u n t s of sulfate added are relatively low. D u r i n g pre-flooding and post-harvest considerable a m o u n t s of CH 4 can be released from flooded rice fields. In previous monitoring studies these periods were not included, which m a y cause an u n d e r e s t i m a t i o n of total seasonal emission by 10-15%. The project improved our knowledge concerning CH4 emission from rice a n d revealed errors in the procedures to e s t i m a t e CH 4 emissions t h a t both justify a reduction and an increase of the e s t i m a t e d CH 4 source s t r e n g t h from rice fields. 5. A C K N O W L E D G M E N T S Without the support and active involvement ofH.-U. Neue, R. Wassmann and R.S. Lantin this project would not have been possible. Thanks are due to J.B. Aduna, M.C.R. Alberto and R. Eugenio for help with the experiments. The project was funded by the Dutch National Research Programme on Climate Change and Global Air pollution (NOPproject X103-1-B-90-9) and collaborated with the IRRI-EPA project on CH 4 emissions from wetland rice fields.
6. R E F E R E N C E S Bachelet, D. and H.U. Neue, Chemosphere, 26, pp. 219-238, 1993. Denier van der Gon H.A~C. and H.U. Neue, Global Biogeochem. Cycles 8, pp. 127-134, 1994. IPCC Climate Change, Cambridge University Press, Cambridge, 200 p., 1992. King G.M., P. Roslev and H. Skovgaard, Appl. Environ. Microbiol. 56, pp. 2902-2911, 1990. Schfitz, H., .4_ Holzapfel-Pschorn, R. Conrad, H. Rennenberg and W. Seiler, J. Geophys. Res., 94, pp. 16405-16416, 1989.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
613
Testing high resolution nitrous oxide emission estimates against observations using an atmospheric transport model A.F. Bouwman" and J.A. Taylor b a National Institute of Public Health and Environmental Protection, P.O. Box 1, 3720 BA Bilthoven, The Netherlands b Australian National University, Canberra ACT 0200, Australia
Abstract
Global budgets of atmospheric nitrous oxide (N20) - an important greenhouse gas - show considerable uncertainty in the estimates of sources and sinks. Here we present forward runs of an atmospheric transport model to test hypotheses on recently derived source and sink estimates. Most atmospheric N~O monitoring stations are not well situated for verifying the simulated seasonality in atmospheric N20. In addition, the precision of N20 measurements is not adequate for resolving seasonal trends. The global N20 inventory used does not account for soil-N20 consumption in temperate nitrogen-limited ecosystems and observed episodic emissions in temperate ecosystems during thaw periods, autumn and from snow-covered soils. These potential errors and possible underestimation of N~O emissions from combustion and industrial sources may exaggerate the simulated seasonal trends.
Nitrous oxide occurs in the atmosphere in minute quantities (310 ppb) and is increasing by -43.8 ppb yr-1 1, although the rate is not constant. In 1992 the increase was about half that of the 1980s 2. The uncertainty factor for most N20 source estimates is at least 2 (ref. 1). The causes of this increase and its variability are not well understood. The major global sources of N20 - soils and oceans - may show seasonal variation in emissions 3,4 that may be reflected in the atmospheric N20 concentration 5. Currently, there are 10 monitoring stations worldwide where atmospheric N20 concentrations are measured 6,7. These stations were established in remote locations where the air is thought to be well mixed and representative of large-scale air masses. There is only one continental station (Niwot Ridge, Colorado). Monthly inventories of N20 emission from soils under natural vegetation and from the oceans were used where available, while annual estimates for other sources were calculated on a monthly basis using different criteria (Table 1). The monthly N20 emission estimates (Figure 1) were used to prescribe a 2.5 ~ resolution global Lagrangian atmospheric transport model 5. The simulated N20 concentrations were --1 ppb higher in the Northern Hemisphere (NH) than in the Southern Hemisphere (SH). This result is consistent with measured concentrations 6, suggesting that the overall latitudinal source distribution is correct. The modelled latitudinal gradient varied with time; smaller differences and the highest N20 concentrations in the tropics were seen from November to April, and the maximum difference and highest values over Northern mid-latitudes from May to October. Simulated continental concentrations in the order of 310-315 ppb over strong-source regions (Figure 2) agree with high N20 concentrations observed at the Colorado station in July-September 1991 7.
614
Table 1. Global monthly 1~ x 1~ resolution inventories of N20 sources used to prescribe the atmospheric transport model * Source
Annual N20 Time emission scale (Tg N20-N yrl)t
Soils under natural vegetation Arable lands and synthetic N fertilizer use Grasslands and animal excreta Savanna burning and forest clearing Agricultural waste burning Post-clearing enhanced soil flux Fossil-fuel combustion and traffic Biofuel combustion Industry Oceans Total sources Stratospheric loss Atmospheric increase
4.3 ~t
month
1.8 month/year 2.4 month/year 0.1 month 0.1 year 0.4 year 0.3 year 0.1 year 0.5 ~'~" year 3.6 ~:~t month 13.6 10.5 3.1
Reference
Major criterion used to produce monthly N20 fluxes
18,19
__ w
18,19 18 22,23 23,24 18 18 18 18 4
Growing seasons 82 Soils under natural vegetation # __ w Growing seasons ** Soils under natural vegetation Constant flux Constant flux Constant flux __ w
* "Best" estimate of eight different cases 18. t Tg = teragram; 1 Tg - 10TM g ~t Model result 3,19for different broad ecosystems, including the N20 emission of 2.3 Tg N20-N yr"1 for closed tropical forests 18 consistent with the estimated 2.4 Tg N20 yr"1 for "lowland tropical forests" 25; 1 Tg N20-N yr"1 from open tropical forests 18, which is within the range (0.4-1.3 Tg N yr1) for "dry tropical forests" 26; and 0.5 Tg N20-N yr1 for temperate forests 18 consistent with the range of other estimates 12,15.No literature estimates are available for comparing the remaining 0.5 Tg N20-N yr"1 from "other lands" 18. w Monthly estimates are included in the inventory. 82"Background" emission (0.9 Tg N,O-N yr1) from unfertilized arable lands calculated on a monthly basis. Monthly distributions of the fertilizer-induced emission (1.25% of N application 18) based on the growing season 2o in each agricultural grid cell ,1, with 60% of N,O losses within 1 month of N application for growing periods < 180 days. For longer growing seasons allowing cultivation of more than one crop, the fertilizer-induced emission is assumed to occur as a constant flux during the growing period. # "Background" N20 emission from grasslands (1.4 Tg N20-N yr1) calculated on a monthly basis; N20 emission from animal excreta is given a monthly distribution identical to the emissions from soils under natural vegetation. ** For growing seasons < 300 days, burning of agricultural waste is assumed to occur at the end of the season (25% during the last month, 50% in the succeeding 30 days and 25% in the 4 weeks thereafter). For growing seasons > 300 days, the burning is assumed to be constant throughout the year. t"t" Including nitric and adipic acid production. :~t The original inventory at -2.8 ~ x 2.8~ resolution was converted to 1~ x 1~ resolution. Due to the scale difference the estimated global oceanic emission is 0.2 Tg N20-N yr"1 lower than the original estimate 4.
The flask sample concentration measurements show stronger oscillations than the high frequency measurements (Figure 3a-d). Neither concentration measurement shows seasonal trends in the NH and SH. Modelled results for monitoring stations in NH temperate zones showed somewhat higher summer than winter concentrations, according to the seasonality of N20 emissions (Figure 3a-b). The seasonal variation in the modelled concentrations for the SH was s m a l l e r - and more consistent with N:O concentration measurements - than in the NH (Figure 3c-d), because of lower emission rates and smaller seasonal variation in emissions in the SH relative to the NH (Figure 1).
615
mg N
90
yr 1
m 2
<0 0 - 0.5
0.5-
1
60
1 -
2
23-
3 4
45-
5 6-~-
30
o o
t,,.
o "o "o
-30
_.1
-60
-90 J
F
M
A
M
J
J
A
S
O
N
D
Month Figure 1. Seasonal-latitudinal distribution of annual emission rates from all sources listed in Table 1.
JULY ~:~i~......
..~
':'-:::'~"~. . . . . . . . . . i~:~.:i....... ~
"
..:i":!?..'..:. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
..
~i-
-.
~
~,.):::::.~ ..... " " ; : ..... .
.....~]i.
i I ................ ".......................................................
".............. :
I
~
,.,
~~~~.,.:.:~:,~:,,~i:i~!'~7; i ~ . .. ~::-Y.:II......
....:..-.~:"~" .
................................................................................. :__i:i:i::::
....:,.~_.~,~-.:::.......................................
.................................... :~ ......
Figure 2. Simulated global distribution of N20 concentration for July.
(. :.... .,.~31-o:"~:~' ~'ii/~
616 a. Oregon (45N, 124W)
314
314 t
3
o
....
cu 306 z
.... o - - ~
C87 --m- C88 - ~ - C89 - 4 - C90 --)1(- C91
304
.... ~
--0- AG87 ..E]-- AG88 --'~-- AG89 --o-- AGg0 ---x-- AG91 - - Modeled
302
b. Colorado (4ON, 106W)
1
.
"=',~
z
304 302300 Month
Month
.o
~
:}o8-1. 9 ~ " 7
I
300
314
0
d. Tasmania (41 S, 145E) 314 ,
c. Samoa (14 S, 171 W)
312
312"
310
310-1
308 -~.....~ .....~.....-x....."x~S"~"~":::~: : ' ' ' r ' ' ~ ' ~
308
:.,./ y,,~----_~:-~= ..... r-,>...~.~:~-~ ;<,L'.'.~.~..... : ~t--,.x...~_z:...u-,,~~y..qJ:.. ~ ....~...-'W 0 : : G ~ : ! ~ ~ o .... v "~ ~ ,IF"4 z 306 304 ' x . _ . / I t U 302
X ./
\\
~I{'
I XOI
--
!:......~>..... ~....~::-.,.~'.,.~....o..~#::::~:....~T,~
v
o z
306 -~,~ . ~ . ~ : : - . ~ : - - o - .... ""~" o...o..--o ....o----~ 304 302 300
300 Month
Month
Figure 3. Monthly mean N20 concentration for different years from Atmospheric Lifetime Experiment and its successor the Global Atmospheric Gases Experiment (AG) high frequency (4 12 measurements per day) real-time in situ gas chromatographic measurements (not corrected for pollution events) 6 and Climate Monitoring and Diagnostics Laboratory (C) measurements in weekly flask samples 7 as well as simulated monthly concentrations at the locations: Cape Meares, Oregon (a); Niwot Ridge, Colorado (b); American Samoa (c); Cape Grim, Tasmania (d).
Because the model correctly reproduces the latitudinal gradient and seasonality of CFC-11 and other tracers 8 and is in general agreement with other N20 modelling studies 4, we assume that significant model inconsistency with atmospheric observations is not caused by .biases in the model. The simulated seasonal variability of concentrations may be in error due to misrepresentations in the emission inventories. Net annual consumption of N20 in soils is reported for grasslands 9 and temperate forest soils 10, while other studies report episodical soil N20 uptake ~1.12.The uptake is attributed to consumption of N20 by denitrification under conditions of low nitrification, leading to low nitrate concentrations 9. In
617
nitrogen-limited cold temperate and boreal forests preferential immobilisation of N occurs in the forest floor and its associated decomposers 13. Immobilisation and slow mineralisation of N may create conditions conducive to N20 consumption. Uptake of N20 may also occur in nitrogen-poor wetlands 3. In our model, N~O consumption in the spring to autumn period would lead to a lower seasonal amplitude in the NH. The model used to calculate monthly N20 emissions from soils (Table 1) was based on the relationships between N~O fluxes and soil temperature and moisture, leading to the highest fluxes in temperate forests in spring and summer. However, elevated N20 fluxes were observed in temperate forests in autumn 14,15, in arable lands during thaw periods 16 and from snow covered soils 17. These episodical emissions from vast temperate forest areas may account for an important global contribution, causing a decrease in variability of the simulated NH N20 emission and concentrations. Industrial N20 emissions may not show a significant seasonality, but combustion-related N20 emission may be higher in winter than in summer. Higher estimates for these sources prevailing in the NH 18 could lead to a decrease in the simulated seasonal NH oscillation. The major part of the N20 from other source candidates, such as industrial and chemical processes, medical and industrial use of N20, production and use of explosives, corona power losses from electric transmission lines, and coastal and freshwater systems, stems from the 30~176 zone is. Emissions of N20 from most of these sources probably lack seasonal patterns. Their inclusion would further reduce the simulated seasonal amplitude of the NH N20 concentration. The precision of the N20 concentration measurements is 0.15% or 0.5 ppb, but the standard deviation in the daily and monthly averages is much higher 6.7. Different concentration measurements vary considerably in the amplitude and temporal pattem of the oscillations (Figure 3c). As CO~ is considered not to interfere with N20 measurements 6 (j. Elkins, personal communication), the inconsistencies between different observations could be the result of other interferences (e.g. H~O) in the N~O concentration measurements. The predicted seasonal variations are in the order of 1 ppb for the NH and <1 ppb for the SH. Resolving an armually averaged latitudinal gradient may be testing the limits of current measurement technology. In addition, there is a lack of N20 monitoring stations in continental interiors. Testing hypotheses on sources and sinks of atmospheric N20 with the available observational data is therefore difficult.
Acknowledgements This investigation was supported by the Dutch National Research Programme on Global Air Pollution and Climate Change (contracts N -~ 852079 and 851060), and the Netherlands Ministry of Housing, Physical Planning and Environment (project N -~771060). We thank Fred Alyea and Derek Cunnold for making available the ALE/GAGE data, Jim Elkins for helpful discussions, and Cynthia Nevison for sending the ocean N20 flux fields. The calculations on which the atmospheric model results are based were carded out using the Fujitsu VP-2200 at the ANU Supercomputer Facility. We are grateful to Ruth de Wijs-Christensen for editorial comments.
References 1 2
3
Khalil, M.A.K. & Rasmussen, R.A.J. geophys. Res. 97, 14651-14660 (1992). Swanson, T.H., Elkins, J.W., Buffer, J.H., Montzka, S.A., Myers, R.C., Thompson, T.M., Baring, T.J., Cummings, S.O., Dutton, G.S., Hayden, A.H., Lobert, J.M., Holcomb, G.A., Sturges, W.T. & Gilpin, T.M. in Climate Monitoring and Diagnostics Laboratory No. 21, Summary Report 1992 59-75 (U.S. Department of Commerce, National Oceanic and Atmospheric Administration, Environmental Research Laboratories, Boulder, Colorado, 1993). Bouwman, A.F., Fung, I., Matthews, E. & John, J. Global biogeochem Cycles 7, 557-597 (1993).
618 4 5 6 7
8 9 10 11 12 13 14 15 16 17 18 19 20 21
22 23 24 25 26
Nevison, C. Cooperative PhD thesis, No. 147 (Stanford University and National Center for Atmospheric Research, 1994). Taylor, J.A. Mathematics and Computers in Simulation 33, 597-602 (1992). Prinn, R., Cunnold, D., Rasmussen, R., Simmonds, P., Alyea, F., Crawford, A., Fraser, P. & Rosen, R. J. geophys. Res. 95, 18369-18385 (1990). Montzka, S.A., Elkins, J.W., Buffer, J.H., Thompson, T.M., Sturges, W.T., Swanson, T.H., Myers, R.C., Gilpin, T.M., Baring, T.J., Cummings, S.O., Holcomb, G.A., Lobert, J.M. & Hall, B.D. in Climate monitoring and Diagnostics Laboratory No. 20, Summary Report 1991 60-81 (U.S. Department of Commerce, National Oceanic and Atmospheric Administration, Environmental Research Laboratories, Boulder, Colorado, 1992). Taylor, J.A. in Computer modelling in the environmental sciences (eds D.G. Farmer & Rycroft, M.J.) 233-241 (Clarendon Press, Oxford, 1991). Ryden, J.C.J. Soil Sci. 34, 355-365 (1983). Castro, M.S., Steudler, P.A., Melillo, J.M., Aber, J.D. & Millham, S. Biogeochemistry 18, 119-135 (1993). Keller, M., Goreau, T.J., Wofsy, S.C., Kaplan, W.A. & McElroy, M.B. Geophys. res. Lett. 10, 1,156-1,159 (1983). Bowden, R.D., Steudler, P.A., Melillo J.M., & Aber, J.D.J. geophys. Res. 95, 13,997-14,005 (1990). Vogt, K.A., Grier C.C. & Vogt, D.J. Adv. ecol. Res. 15, 303-377 (1986). Goodroad, L.L. & Keeney, D.R.J. environ. Qual. 13, 448-452 (1984). Schmidt, J., Seiler, W., & Conrad, R. J. atmosph. Chem. 6, 95-115 (1988). D~rsch, P., Flessa, H., & Beese, F. Mitt. Deutschen Bodenkundl. Gesellsch. 72, 495-498 (1993) Sommerfeld, R.A., Mosier, A.R. & Musselman, R.C. Nature 361, 140-142 (1993). Bouwman, A.F., van der Hoek, K.W. & Olivier, J.G.J.J. geophys. Res. (in the press). Kreileman, G.J.J. & Bouwman, A.F. Water air soil Pollut., 76 231-258 (1994) Leemans, R. & Van den Bom, G.J. Water air soil Pollut. 76, 133-161 (1994). Olson, J.S., Watts, J.A. & Allison, L.J. ORNL 5862. Environmental Sciences Division Publication No. 1997 (Oak Ridge National Laboratory, Tennessee, National Technical Information Service, U.S. Depl. Commerce, 1983). Hao, W.M., Liu, M.H. & Cmtzen, P.J. in Fire in the Tropical Biota. Ecological Studies 84 (ed. J.G. Goldhammer) 440-462 (Springer Verlag, Berlin, 1990). Cmtzen, P.J. & Andreae, M.O. Science 250, 1669-1678 (1990). Andreae, M.O. in Global biomass burning (ed. Levine, J.S.) 3-28 (MIT Press, Cambridge, U.S.A., 1991). Matson, P.A. and Vitousek, P.M. Bioscience 40, 667-672 (1990). Vitousek, P.M., Matson, P., Volkman, C., Mass, J.M., and Garcia, G. Global Biogeochem. Cycles 3, 375-382 (1989).
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
619
M o d e l l i n g nitrous oxide e m i s s i o n from soils" a tool for e x p l o r i n g e m i s s i o n reduction strategies C.A. Langeveld, P.A. Leffelaar and J. Goudriaan Department of Theoretical Production Ecology, Wageningen Agricultural University, P.O. Box 430, 6700 AK Wageningen, The Netherlands, Phone: +31-8370-82140/82141; Fax: +31-8370-84892, E-mail: TPELANGEVELD @RCL.WAU.NL
Abstract Possibilities to reduce nitrous oxide (N20) emission from soils can be explored with simulation models. In this study, principal assumptions and hypotheses underlying such a simulation model under development are presented. It uses soil water content profiles as main input data and describes production, consumption and transport of N20. Observations on sandy grassland plots in the Wageningen Rhizolab will be used in the model development. The model is to explain the shape of the nitrous oxide profiles and the relation between subsurface nitrous oxide gradients and the observed fluxes.
1. I N T R O D U C T I O N Soils are estimated to be responsible for more than 50 % of the global emission of N20 [ 1]. Soil emission data show a large variation. This variation results from the influence of various factors on the fundamental processes denitrification, nitrification and transport that determine N20 emission from soils. These factors are temperature, aeration (related to moisture content), nitrogen mineralisation rate, amount and kind of added N fertiliser and content of readily decomposable carbohydrates. The current research is focussed on quantifying and modelling the influence of these factors on N20 emission from grassland soils. Some of the abovementioned factors, like aeration and fertilisation, can be manipulated in soils. These possibilities can be explored in experiments. Simulation models can be a powerful tool to guide this experimental work and to minimise the amount of work needed for giving well-based proposals for strategies to reduce N20 emissions from soils. In this paper the planned development, calibration and validation of a simulation model to describe nitrous oxide dynamics in and emission from relatively homogeneous grassland soils are outlined. In these modelling steps results of experiments in the Wageningen Rhizolab will be used [2,3].
2. M O D E L C O N C E P T S
2.1. Nitrous oxide production and consumption in the soil Nitrous oxide production and consumption in soils is attributed to the nitrogen transformation processes denitrification and nitrification (Figure 1). Before the modelling, a
620 quantitative inventory will be made of the relative importance of these processes for nitrous oxide production and consumption under various circumstances.
ii!~!~!!!hi!~iiiiiii iiii iilil Jiiiiiiiiiiiiiiiiiiiiiiiiiiiiii iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii ~
u
iiii iii
i
iii i
!ii
;; !i! iiiiiiiiiiiiiiii!iii Figure 1. Hole-in-the-pipe model for nitrous oxide dynamics in and transport into or out of soils (Adapted from [4]). 2.1.1. D e n i t r i f i c a t i o n Denitrification may occur at low partial oxygen pressures. Three types can be distinguished: biological denitrification by either heterotrophic microorganisms or autotrophic microorganisms and chemodenitrification. In all cases nitrate or nitrite is reduced. Heterotrophic biological denitrification is the most important for N20 emission from soils [5], and will be the only type taken into account in our model. The overall reaction equation is:
5 (CH20) + 4 NO 3- + 4 H + ~ >
5 CO 2
+
7 H20 + 2 N 2 + energy.
The reduction of nitrate (NO 3- ) in heterotrophic denitrification occurs stepwise; N20 is one of the intermediate compounds that can be either further reduced to N 2 or transported to the soil surface and eventually the atmosphere: N O 3 - - - > N O 2 - - - > NO ~ >
N20 m > N2"
The extent to which N20 is further reduced principally depends on the aeration of the soil, the amount of nitrate present and the residence time of N20 in a reduced soil portion. The process-based denitrification model accounting for these phenomena of [6] will be used as a starting point for model development. It essentially describes the dynamics of several (nitrogen) compounds as a result of the biological processes performed by two groups of heterotrophic bacteria: one group of strict aerobes, using only oxygen as electron acceptor, and one group of denitrifying bacteria that use either oxygen or, under anaerobic conditions, nitrate, nitrite and nitrous oxide as electron acceptors. Mineralisation and immobilisation are described in a very simplified manner. 2.1.2. N i t r i f i c a t i o n Nitrous oxide production in relatively dry soils, i.e. at moisture contents below field capacity, is generally attributed to nitrification [7]. The overall reaction of nitrification is given as:
621 NH4 + + 2 02 - - > NO3- + 2 H + + H20 + energy. In our opinion, however, N20 production attributed to nitrification should also be classified as N20 production in denitrification, since it is essentially due to denitrification of intermediate products in nitrification. Nitrification is performed by autotrophs as well as by heterotrophs, of which autotrophs are the most important. In fact, nitrification takes place in two steps. First ammonium (NH4 +) is oxidized to nitrite (NO2-) (ammonium oxidation). Next nitrite is oxidised to nitrate (nitrite oxidation). Nitrous oxide production from nitrification is ascribed to two processes [5]: 1. nitrifier denitrification: ammonium oxidisers use NO 2- as an alternative electron acceptor when 02 is (locally, temporarily) limiting and produce N20 (in fact this process is pure denitrification which occurs in nitrifying organisms), 2. a type of chemodenitrification: chemical decomposition of intermediates between NH4 + and NO2-, or NO 2- itself, to N20. We are not aware of any existing process-based model of nitrifier denitrification. However, this process could be described analogously to the description of denitrification by [6]. The second process can probably be described by elementary chemical reaction kinetics as soon as the reactions involved are identified.
2 . 2 . Transport of nitrous oxide within the soil and at the soil-atmosphere interface Several mechanisms of gas transport in soils have been distinguished [8]. Ordinary diffusion is the most important in the continuous gas phase of an unsaturated soil [9]. We assume that our system can be described as a one-dimensional system, basically governed by the reaction diffusion equation (Fick's Second Law plus production/consumption): iOc= 0 {Deff0__cc} + S , Ot ~)z iOz where:
(1)
c = nitrous oxide concentration, t = time, z = depth (z increasing with depth, z=0 at the surface), D etf = (modelled) effective diffusion coefficient, S = volumetric nitrous oxide production or consumption strength.
A more refined description of gas transport in the soil can be found in [ 10]. The equation describing nitrous transport at the soil-atmosphere interface is a special case of equation 1.
3. C A L I B R A T I O N AND V A L I D A T I O N P R O C E D U R E The model that will be developed has to be calibrated and validated. For this purpose, a twoyear experiment is performed in the Wageningen Rhizolab [2]. Regularly, nitrous oxide fluxes and belowground profiles of the soil moisture content and several gases and nutrients were determined on 4 grassland plots on a sandy soil. A portion of the results will be used for calibration. Results for the other dates will be used for validation. For validation, probably also results from other experiments, like the experiments of the Nutrient Management Institute (NMI; Velthof et al., this volume), will be used.
622 4. P O T E N T I A L REDUCTION STRATEGIES Experiments suggest that nitrous oxide emission from grassland soils could be reduced by applying suitable management practices like split application of nitrate fertilisers during dry periods [11, 12]. The simulation model could be very helpful to explore the possibilities to reduce emissions by measures like this.
5. ACKNOWLEDGEMENTS We gratefully acknowledge the support of J E Hofman, M H van den Bergh, A M van Dam, the staff of the Wageningen Rhizolab, the Nutrient Management Institute (NMI) and the members of discussion group 4 of the C.T. de Wit Graduate School Production Ecology during the research and preparation of this paper. It is part of the integrated N20 grassland research project in which also participate the NMI, and the Research Institute for Agrobiology and Soil Fertility (AB-DLO), Wageningen and Haren, The Netherlands. This Project is financially supported by the Dutch National Research Program on Global Air Pollution and Climate Change (Project 852074).
6. REFERENCES
1 A.F. Bouwman, In A. F. Bouwman (ed.), Soils and the Greenhouse Effect, John Wiley, Chichester (1990) 61. 2 C.A. Langeveld, P.A. Leffelaar and J.Goudriaan, Submitted to the Proceedings of the 8th Nitrogen Workshop, Ghent, Belgium, 5-8 September 1994. 3 S.C. Van de Geijn, J. Vos, J. Groenwold, J. Goudriaan and P.A. Leffelaar, Plant and Soil, 161 (1994) 275. 4 E.A. Davidson, In J.E. Rogers and W.B. Whitman (eds.), Microbial production and consumption of greenhouse gases: methane, nitrogen oxides, and halomethanes, American Society for Microbiology, Washington, D.C. ( 1991) 219 5 T. Granli and O.C. Beckman, Norwegian J. Agric. Sci., 12(Supplement) (1994), 1. 6 P.A. Leffelaar and W. Wessel, Soil Sci., 146 (1988) 335. 7 D.W. Bergstrom, M. Tenuta and E.G. Beauchamp 1994, Biology and Fertility of Soils 18 (1994) 1. 8 E.A. Mason and A.P. Malinauskas, Gas transport in porous media: The dusty-gas model, Elsevier, Amsterdam, 1983. 9 D.B. Jaynes and A.S. Rogowski, Soil Sci. Soc. Am. J. 47 (1983) 425. 10 P.A. Leffelaar, Soil Sci. 143(1987)79-91. 11 O. Van Cleemput, A. Vermoesen, C.J. De Groot and K. Van Ryckeghem, In J. Van Ham, L.J.H.M. Janssen and R.J.Swart (eds.), Non-CO2 greenhouse gases. Why and how to control?, Kluwer Academic Publishers, Dordrecht (1994) 145. 12 C.J. De Groot, A. Vermoesen and O. Van Cleemput, In J. Van Ham, L.J.H.M. Janssen and R.J.Swart (eds.), Non-CO2 greenhouse gases. Why and how to control?, Kluwer Academic Publishers, Dordrecht (1994), 183.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
623
Measurements of the atmospheric emission of N20 from biogenic sources in general and by grassland ecosystems in particular Jan Duyzer IMW TNO, P.O. Box 6011, 2600 JA Delft, The Netherlands Abstract
The project is part of the 'Integrated N 2 0 grassland project'. The project carried out at TNO aims to determine the atmospheric emissions of N20 from biogenic surface sources in the Netherlands. The following activities were part of the project: 9 determination of N20 emissions from grassland ~ comparison of methods to measure N 2 0 emissions in the field 9 determination of N20 emissions from arable land ~ determination of N20 emissions from fresh water systems and coastal waters 9 organisation of a round robin to compare measurements of N20 concentrations in air 1. INTRODUCTION
The objective of this study was to measure the N20 emission from relevant sources in the Netherlands with an emphasis on grassland. Other ecosystems studied include agricultural land, sea water an fresh water bodies. It was intended to measure these fluxes on an ecosystem scale ie. using the aerodynamic gradient method wherever possible. In other cases the dynamic enclosure technique was to be employed. This project was part of the so called 'Integrated N 2 0 grassland project' in which several Dutch groups (including RIVM, NMI, LUW and AB-DLO) participated. In the framework of this project a round robin was organised in which a comparison was made between the methods that are used in the programme to measure N 2 0 concentrations in air [ 1]. 2. THEORY AND METHODS
Dynamic enclosures designed by TNO [2] and NMI design static enclosures [3] and the aerodynamic gradient method were used. The dynamic enclosures cover a surface of 20 x 100 cm whereas the static enclosure cover a surface with a diameter of 20 cm.
624 In the gradient method the flux is derived from the concentration gradient of the trace gas in the air above the surface and measurements of the turbulence intensity. The method is described in detail in [4]. 3. RESULTS 3.1
Comparison studies
In two comparison campaigns the enclosure methods and aerodynamic methods were applied simultaneously. The aim of this experiment was specifically to investigate the spatial distribution of N 2 0 sources on a field scale. To this purpose 42 static enclosures were operated continuously by the NMI [3]. The enclosures were placed along a transect between the two dikes bordering the site at distances of one meter. Close to the dikes the observed fluxes were smaller, often around 0.5 mg N/m2/hr whereas in other areas some chambers showed fluxes as high as 10 mg/m2/hr. More than 80 % of the variation in N 2 0 fluxes could be attributed to variations in soil moisture and nitrate concentrations in the soil. The equipment used for the gradient method was located in such a way that the flux determination was expected to be representative for the area studied using the enclosures. Figure 1 shows the results of the experiment. The flux estimated using this method varied with time between 0 and 2.5 mg N/m2/hr. At night fluxes showed much less variation and varied between 0.5 and 1 mg N/m2/hr. During the day fluxes as high as 2.5 mg N/m2/hr were observed but the variation was much stronger. The dynamic enclosures showed fluxes with a median value of around 1 mg/m2/hr. During or after rain the direction of the flux indicated deposition, i.e. flux of N 2 0 towards the surface. These phenomena were also observed with the enclosures. 2.5
I
!
1.5
o E 0r z 0.5 z
~ "--4
0.9 -0.8 -0.7 -0.6
ra'nl
-0.5 r._
-0.4
!i
,i
o
-0.3 -0.2 -0.1 0
I1
"0.5
-1.5
1800
()
6()0
1:;:;00
1800
()
6()0
1200
1800
Time (hour)
Figure 1 Fluxes of N 2 0 observed during the inter comparison experiment over peat grassland at Zegveld in June 1993
625
In a second experiment in November the static enclosures and the gradient method were used. With the 42 static enclosures, this time located in a 6X7 meter grid showed fluxes varying between zero and 1.3 with an average of 0.25 mg N/m2/hr. With the gradient method hourly averaged fluxes varying between 0.15 and 0.7 mg/m2/hr were observed. The results of these experiments are not easy to generalize. The results of the first phase study were confirmed showing that static chamber results can show large variation probably linked with the scale of variation of relevant soil parameters and processes. It seems justified to conclude that on a day scale the fluxes of N20 measured using the static chambers lie within a factor of 2 to 3 from fluxes measured with the gradient method. 3.2
Continuous measurements of fluxes from grassland
On the peat grassland site at Zegveld N20 fluxes were determined every day from April 1993 to June 1994. To this purpose samples were taken every afternoon. Figure 2 shows the monthly averaged emissions calculated from all observations from April 1993 to June 1994 expressed in kg N/ha/year. The average emission is roughly 10 kg N/ha over this period with peaks in the summer. In general two peaks are found midsummer and November 1993. The December period was quite wet. The annual fertilizer input on the site was roughly 100 kg N/ha.
2~I
T
150
c"
..r
I
100-
50-
g ,,...
-~
o-
-50
-I00
-
-
.
Figure 2 3.4
.
.
.
.
.
g
g
g
~;
g
~;
Median emission of N 2 0 observed over peat grassland at Zegveld.
Emissions of N20 from coastal waters and freshwater systems.
A campaign was carried out along the Dutch Waddensea. To estimate the flux the concentration of N20 was determined at two heights 0.3 and 4 m above the sea surface. The N20 concentration in the sea was high compared to open sea values ie. around 1000 ng/1. The observed emission from the sea was equivalent to 2 kg N20 N/ha/yr. If this emission rate would be observed all year the Dutch Waddensea would emit roughly 500 ton N20 N per year.
626 In April and May several campaigns were carried out over the freshwater lake Ketelmeer. This lake was chosen because of its high nitrate concentrations. In one case significant gradients were detected with relatively high fluxes equivalent to 3 kg N/ha/yr. In those cases high N20 concentrations of more than 2000 ng/1 were observed in the water phase. Further water samples were taken from sweet as well as salt water bodies in the Netherlands. The results indicate a quite strong relation between nitrate and ammonium content on one hand and N 2 0 concentration on the other. This result is shown in Figure 3 which shows the N 2 0 concentration observed as a function of the ammonium and nitrate content for several salt and flesh water systems in the Netherlands. 9
.-.
fresh
1600
1600
1400
1400
1200
~.
salt
1200
c-
vc
0 c~
o
1000
O
z
1000
z
800
800
600
600
O9
0
0
Figure 3 3.5
2
4
6 8 1'0 1'2 1'4 NO 3 (mg/I)
6
0
0'.1
0'.2 NH 4 (rng/I)
013
0.4
The concentration of N 2 0 as a function of Nitrate and ammonium content
Emissions from arable land
Measurements of the emission rates of N20 from arable land were started in spring 1994. Two sites were chosen. One set of experiments was carried out at the agricultural experimental station at Nagele in the Noordoostpolder on a sea clay soil. The other experiments were 9 . carried out over a sandy soil in Vredepeel. Measurements were carried out over potatoes and onions over the clay soil and over peas, potatoes and grassland on a sandy soil. Measurements using two enclosures were carried out over bare soils and during crop growth. Around two weeks after fertilizer application a maximum in the emission rate is found. The emission from the onions upon which N fertilizer was applied was around a factor of three lower. 4. REFERENCES
1. Verhagen, H.L.M. and J.C.Th. Hollander (1994). TNO-MW report R 94/273. 2.
Baas, J. H.S.M.A. Diederen (1992). IMW-TNO report R 92/211. Delft, The Netherlands.
3.
Velthof, G.L., O.Oenema (1994). Draft report C 94.193 Nutrient Management Institute Wageningen, the Netherlands
4.
Duyzer, J.H. (1994). Draft scientific report. Emission of N20 from biogenic sources.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
627
Effects of nitrogen fertilization and grazing on the emission of nitrous oxide from grassland G.L. Velthof, A.B. Brader and O. Oenema NMI, Department of Soil Science and Plant Nutrition, Wageningen Agricultural University, P.O. Box 8005, 6700 EC, The Netherlands.
Abstract In the Netherlands, managed grasslands are potentially a large source of nitrous oxide (N20), because of the large nitrogen (N) input and the relatively high groundwater levels. To provide insight into the major factors that contribute to N20 emission from grassland and to provide quantitative N20 emission rates, a monitoring study was carried out on four sites, during March 1992 to March 1994. Fluxes of N2O increased after N fertilizer application and grazing, especially during wet conditions. Fluxes were higher from peat soils than from sand and clay soils. Fluxes were low during the winter periods. Total N20 losses were 2 to 4.5 times higher on grassland fertilized with 160-460 kg N ha 1 yr 1 than on unfertilized grassland. Losses from grazed grasslands were 1.5 to 3.5 times higher than losses from mown grassland. This study shows that management practice of grassland and soil type are major factors controlling N20 emission from grasslands.
1. I N T R O D U C T I O N
On a global scale, soils are a major source of nitrous oxide (N2O). In soils, N2O is produced during the microbial processes nitrification and denitrification, primarily controlled by the availability of mineral nitrogen (N), oxygen (02) and mineralizable carbon (C) in the soil [1]. In the Netherlands, intensively managed grasslands are possibly a large source of N20 because grasslands cover 30 % of the total surface area, the N input via fertilizer and animal excretions is high and many soils are relatively wet due to the shallow groundwater level. Furthermore, about 30% of the grasslands are situated on peat soils. Due to the high contents of organic N and C and due to the shallow groundwater levels, it is expected that especially grasslands on peat soils are a major source of N20. In the present study, the effects of N fertilization, grazing, and soil type on N20 emission from grasslands were investigated. The aim was to provide insight into the major factors that contribute to N20 emission from managed grassland and to provide quantitative N20 emission rates, obtained from field measurements.
628 2. M A T E R I A L S A N D M E T H O D S Fluxes of N20 were measured weekly in the period March 1992 to March 1994 on four contrasting grassland sites in the Netherlands, namely on a sand soil in Heino, a clay soil in Lelystad and two peat soils in Zegveld [2]. Peat soil I had a mean groundwater level of 35 cm below soil surface and peat soil II had a mean groundwater level of 50 cm below soil surface. Perennial ryegrass (Lolium perenne L.) was the dominant grass species in all swards. At each site, the experiments had a complete randomized block design, with three treatments in three replicates. The plots were 2.5 x 20 m. The treatments were mown grassland without N fertilizer applications, mown and N fertilized grassland and predominantly grazed and N fertilized grassland. Fertilizer N was applied as calcium ammonium nitrate (CAN), in six or seven dressings during the growing season. The economic optimum application rates of N fertilizer were assessed by using an interactive fertilization system based on a combination of modelling and measuring soil mineral N and N uptake. The application rates for the grazed grasslands were equal to those of the mown grasslands. Total N input by urine and dung of the grazing cattle was calculated using standard calculation procedures [3]. Fluxes of N20 were measured using vented closed flux chambers made of PVC cylinders with an internal diameter of 20 cm and height of 15 cm. Concentration of N20 in the headspace of the flux chambers was measured in the field at 0, 10, 20 and 30 minutes after closing, using a photo-acoustic spectroscopic infra-red gas analyzer, directly attached to the flux chambers. All fluxes were measured in six replicates. Mean N20 fluxes were calculated as arithmetic means and total N20 losses were calculated by integration of the mean N20 fluxes over time [2].
3. R E S U L T S A N D D I S C U S S I O N Fluxes of N20 increased after application of N fertilizer and grazing. The N20 flux pattern of fertilized grassland depicted in Figure 1 is typical for N20 fluxes from grasslands fertilized in several N dressings. This pattern is mainly due to fluctuations in mineral N content in the soil, caused by a combination of fertilizer N application, N uptake by the grass roots and microbial biomass and N losses by leaching, denitrification and volatilization. The magnitude of the N20 flux after N application was also dependent on soil moisture content, because during dry periods the effect of N application on N20 loss was much smaller than during wet periods. During the winter periods, fluxes were much lower than during the growing seasons, probably due to the low temperatures and low mineral N contents in the soil (not shown). Generally, both flux magnitude and duration were higher for the peat soils t h a n for the sand and clay soils and those from peat soil II were higher than those from peat soil I. The higher N20 fluxes from the peat soils than the sand and clay soils were likely due to higher organic C and N contents, higher denitrification potentials and higher groundwater levels of the peat soils (data not shown).
629 N20 flux, kg N ha-1 day -1 0.6
l lLll i
1 1[ 1
0.4
-'-'Unfertilized + mown - V - ' N fertilized + rnown ~'N
fertilized + g r a z e d
0.2-
0.0-
M'A'M'J 'J ' A ' S ' O I N ' D ' J ~F ' M ' A ' M ' J ' J ~A'S'O'N'D'J 'FfM ~
1992
!
Month
1993
i 1994
Figure 1. Time course of N20 flux from grassland on peat soil I, for the three m a n a g e m e n t practices. Arrows indicate time of N application and grazing. Dotted arrows indicate grazing without N application. For all sites, both the order in total N input via fertilizer, dung and urine, and the order in total NzO losses was: unfertilized and mown < N fertilized and mown < N fertilized and grazed grasslands (Figures 2A and 2B). Total NzO losses were 2 to 4.5 times higher on N fertilized grassland t h a n on unfertilized grassland and 1.5 to 3.5 times higher on grazed grassland t h a n on mown grassland. Losses from the peat soils were higher t h a n from the sand and clay soils, for all treatments. Remarkably, N20 losses from peat II, the soil with lowest N input, were highest, indicating the large effect of soil type on NzO losses. On the sand soil, 1.0% of the fertilizer N applied on mown grassland and 1.5% of the urine and dung N deposited on grazed grassland was lost as N20, during the two year period. This was 0.9 and 3.3% on the clay soil, 1.9 and 2.3% on peat soil I and 3.9 and 9.8% on peat soil II, respectively. The higher grazing derived losses t h a n fertilizer derived losses, in terms of % of the N input, suggests t h a t the effect of grazing on N20 losses was not only an effect of the higher N input. Several mechanisms may have contributed to the high grazing derived N20 losses: - the much higher nitrate contents of grazed grasslands t h a n mown grasslands (data not shown) may have increased N20 emission more t h a n proportionally, because N20 becomes a more important end product of denitrification at increasing nitrate contents [1]; - compaction of the soil by treading of the grazing cattle decreases 02 diffusion into the soil and may enhance production of N20; - high ammonia concentrations in urine patchess may inhibit nitrification, which may enhance production of N20; - urine and dung contain mineralizable C, which may increase denitrification rate.
630 In conclusion, this study shows that both soil type and practice of grassland management are major factors controlling N20 losses from grasslands. The high grazing derived N20 losses indicate that, besides N fertilizer application, also grazing has to be considered in N20 budget studies.
N input, kg N ha -1 year -1
N20 loss, kg N ha -1 year -1
A
B
I
800
4O
600
30
400
2O
|
"m m !
2ooi o
Sand
Clay
Peat l
] Unfertilized + mown
Peat II
0 _ 7/ Sand
N fertilized + mown
~ / Clay
, Peat l
/c Peat II
N fertilized + grazed
Figure 2. Total annual N input via fertilizer, urine and dung (A), and total annual N20 losses (B), for all sites and treatments. Averages of two years. 4. A C K N O W L E D G E M E N T S
The authors thank ROC Heino, ROC Zegveld, Waiboerhoeve in Lelystad and the colleagues of the Research Station for Cattle, Sheep and Horse Husbandry in Lelystad for their support. This investigation was supported financially by the Dutch National Research Program on Global Air Pollution and Climate Change. 5. R E F E R E N C E S
1 E.A. Davidson, Fluxes of nitrous oxide and nitric oxide from terrrestial ecosystems, p. 219-235 In: J.E. Rogers and W.B. Whitman, Microbial production and consumption of greenhouse gases: methane, nitrogen oxides, and halomethanes, American Society for Microbiology, Washington D.C (1991). 2 G.L. Velthof and O. Oenema, Nitrous oxide fluxes from grassland in the Netherlands: II. Effects of soil type, nitrogen fertilizer application and grazing. Submitted to European Journal of Soil Science. 3 D.W. Bussink, Relationships between ammonia volatilization and nitrogen application rate, intake and excretion of herbage nitrogen by cattle on grazed swards, Fertilizer Research 38 (1994), 111-121.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
631
Modelling the emission of dinitrogen oxide from mown and from grazed grassland J. Bril, H.G. van Faassen and H. Klein Gunnewiek Institute for Agrobiological and Soil Fertility Research, P.O. Box 129, 9750 AC Haren, The Netherlands
Abstract The integrated model SONICG was developed to simulate the emission of dinitrogen oxide (N20) from grassland soil. The model comprises modules on soil physico-chemical conditions and processes, and on soil microbial carbon and nitrogen turnover and makes use of data from an existing model on grass development. Some typical model results are shown for the daily emission of N20 from mown and from grazed grassland, with special attention for effects of urine. 1. INTRODUCTION
Based on literature data on N20 emission, intensively managed grazed grasslands were expected to be a major biogenic source of N20 emission in The Netherlands. A high contribution of intensively managed grassland to N20 emission might be explained from the low efficiency of high nitrogen (N) inputs in dairy farming. Furthermore, high concentrations of mineral N in urine spots in grazed grassland form active centres for N loss by nitrification-denitrification. To get a better understanding of the complex relationships that result in N20 emission from grazed grassland the model SONICG -Simulation Of the Nitrogen Cycle in Grassland soil- was developed. 2. THE SIMULATION MODEL SONICG
SONICG mechanistically simulates the relevant soil physical, chemical and biological processes in grassland soil layerwise. The model considers 30 layers of 2.5 cm, each with its own properties. Above the soil a gaslayer of 2.5 cm is present in the model as a kind of gasbuffer between the soil and the atmosphere. Figure 1 shows schematically the nitrogen cycle processes that play a central role in the model SONICG: Nitrification and Denitrification, including the production and reduction of N20. A separate M.I.T. module simulates mineralizationimmobilization-turnover of organic matter. Inputs of organic matter in the form of dead grass litter and roots, as well the uptake of mineral N by the grass are derived from a separate mechanistic model on grass development (Verberne, 1992).
632
I INPUTDATA b HYDROLOGY MODULE
GRASS development model (Verberne) I Roots ..........I N~46~EC_!.... ~
~
NH3~NH~
11
.
.
.
.
.
.
'
.
.
.
.
I I __.~ 5 _._s~ water pha_se _
I-
M.I.T. MODULE 1 =Nitrification 2= Denitrification 3= Mineralization 4=Immobilization 5 = Plant uptake 6=Cation exchange 7=water/gas exchange
Urine Fertilizer Manure
Figure 1. Nitrogen transformations and exchanges modelled in SONICG The order of the different modules and processes in a timestep (one day) of the simulation is shown in Figure 2: input data are read from a hydrology module (daily weather data; soil data, including water-filled pore space fractions for each layer) as well as from a grass growth model (organic C and N inputs from and potential uptake of mineral N by the grass). Next the temperature profile of the soil is calculated. Thus the actual process rates of M.I.T., nitrification and denitrification can be calculated next, including temperature and moisture effects. At the field scale soil water-filled pore space is a major rate controlling factor for the production and reduction of N20 (Groffman, 1991). Figures 3 and 4 show how M.I.T., nitrification and denitrification depend on WFPS in the model. After the M.I.T. module the actual plant uptake of mineral N is calculated. Then nitrification rates-production of N20 and nitrate- are calculated, followed by denitrification rates -nitrate and N20 reduction. Next gas transport and transport of dissolved substances are calculated. As a final steps chemical equilibrium calculations are made for the following components: H20, H§ Ca2§ K§ NH4§ NO3-, N20, CO32, CI-, CEC (Cation Exchange Capacity) and inert gas (all gasses except the explicitly modelled gasses CO2, NH3, and N20). For each layer cation exchange, complexation in solution, precipitation/dissolution of calcite and the exchange of gasses between the soil water- and gasphase are considered. As the main parameter of interest here SONICG can calculate the daily emission of N20 from the soil surface, as well as that of CO2 and NH3. Escape of gasses with drainwater at the lower boundary of the soil is also calculated. For details on SONICG see the endreport of NOP-project 852078 (Van Faassen and Bril, 1994).
633
~
Re t t i v e 1.0
t
i np u t
READ
1
,,,._ ,...-
1,2
1
PLANT
,,.._
1, 3, 5, 6._~
1
pp.2- 8
~
M.I.T.
~
ID 6, 7, 9
UPTAKE
~
ID~ 9
0.4
7, 9
DEN I TR I F I CAT ION
~
ID~ 7, g
~
ID~ 9
~
lp~ 9
TRANSPORT
D I SS.
TRANSPORT
EQUI L I BR I A
.
,
.
0
~
GAS
0.2
o
N I TR I F I CAT I ON
1 1 1
0.8
0.6 3, 5, 9
TEMPERATURE
,,.._ y
1~2,3
1 1 1
~
Rate
,2
.
.
.
0,4
.
.
,
0,6
Water-Filled
0.8
Pore
,0
Space
Figure 3. Dependance of organic matter decay and denitrification rates on soil water-filled pore space.
b Re
at
i ve
rate
1.0
OUTPUT t
=
t
+
1
F
0.8
0.6
0.4
012
o
I
0
I
012
I
9
i
o,,q Water-
F i I Ied
I
0.6 Pore
9
I
I
0.8 Space
Figure 2. Flow chart of SONICG
Figure 4. Dependance of nitrification, nitrate and N20-production rates on soil water-filled pore space.
I
1.0
634
3. SIMULATED EMISSION OF N20 FROM MOWN AND GRAZED GRASSLAND
As an example of SONICG, mown grassland was compared with grazed grassland, fertilized with 480 and 360 kg of N per ha per year, respectively. To simulate grazed grassland, at least two situations have to be considered: urine spots and area unaffected by urine. Grazed grassland gets less fertilizer N than mown grassland, but in urine spots an excess of mineral N will be present over a long period. Urine spots were simulated to get additional N from urine equivalent to 420 kg per ha. To get an overall picture of grazed grassland, simulation results of urine spots and of area without urine have to be added on an areal basis. Large differences are found in the simulated emission of N20 from mown grassland and from urine spots in grazed grassland. As shown in Figure 5, several months passed between the main N20 emission and the deposition of urine.
N 2 0 flux in kg N ha -1 day -1 0.70.60.5
with urine ~ t b . .
0.4 Re
0.3 urine 420 kg N/ha 0.2
0.1 _ 0
J~L '
0
'
I 91
'
'
I 182
'
'
I 273
'
'
I 364
Day of the year
Figure 5. Simulated daily emission of N20 from mown grassland and from urine spots in grazed grassland. Urine was deposited on day 140. 4. REFERENCES
1 E. Verberne, 1992. Simulation of the nitrogen and water balance in a system of grassland and soil. Nota 258. IB-DLO, Haren. 2 P.M. Groffman, 1991. Ecology of nitrification and denitrification in soil at scales relevant to atmospheric chemistry. In: J.E. Rogers & W.B. Whitman, eds., Microbial production and consumption of greenhouse gases, Am. Soc. Microbiol., p.201-218. 3 H.G. van Faassen & J. Bril, 1994. Modelling N20 emission from grazed grassland. Endreport of NOP-project 852078. AB-DLO, Haren.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
635
Emission of greenhouse gases from wastewater treatment processes J.G. Bruins, H.D. Oostergo & M.A. Brinkhorst BKH Consulting Engineers, P.O. Box 5094, 2600 GB Delft, The Netherlands
Abstract
Biological wastewater treatment processes are a source of emission of the greenhouse gases CO2, CH4 and N20 into the atmosphere. Studies were carried out to quantify the emissions of these gases from the municipal wastewater treatment plants in The Netherlands, which in 1987 handle a total waste load of 17,049,000 population equivalents. On the basis of detailed calculations for the different treatment methods and assumptions for the formation of N20 the following emissions were calculated: CO2 - 880 million kg/yr (0.5% of the total CO2-emission in The Netherlands), CH412.5 million kg/yr (1% of the total CH4-emission) and N20 - 829,000 kg N per year (0.9% of the total N20-emission). Measurements of N20-emissions from 2 low load activated sludge plants were carried out. First results indicated that the N20 formation and emission amount to 0.1% of the total N-load with the influent to the wastewater treatment plants. 1. INTRODUCTION Biodegradation processes in soil and water are an important source of emission of greenhouse gases into the atmosphere. Municipal wastewater treatment plants incorporate various types of processes, which involve the biodegradation of organic matter and biological reduction and oxidation of nitrogen compounds. In general terms a biological wastewater treatment system involves the following processes, causing emission of greenhouse gases: - aeration for bio-oxidation of organics and of nitrogen compounds (nitrification), which results in emission of CO2 and possibly N20; CH4 and NH 3 may also be emitted in these processes when process conditions are not optimal; anaerobic digestion of wastewater treatment sludge with production of biogas, mainly consisting of CH4 and CO2; the biogas is mostly used as an energy source in wastewater treatment plants, which positively effects the energy balance and hence the CO2-emission from the wastewater treatment plant; reduction of nitrate (denitrification) with formation of N2 and N20; disposal of wastewater treatment sludge, e.g. by use in agriculture, dumping at solid waste disposal sites or incineration, with the possible formation of CO2, CH4 and N20;
636 - biodegradation
(aerobic and anaerobic) of residuals (organic and nitrogen compounds) discharged with treatment plant effluent into surface water, with formation of COs, CH4, N2, NH3 and N20; - consumption of external energy resources, in particular for aeration, causing the emission of COs. In the period 1990-1994 several studies were carried out with the objective to quantify the emissions of greenhouse gases from the municipal wastewater treatment plants in The Netherlands. The results of these studies are summarized here. 2. E M I S S I O N OF G R E E N H O U S E GASES FROM THE NETHERLANDS
WASTEWATER TREATMENT PLANTS In 1987 a number of 491 municipal wastewater treatment plants was in operation in The Netherlands treating a total organic waste load of 886,210 tonnes COD (equal to 17,049,000 population equivalents). In 1987 the most applied methods for treatment of municipal wastewaters were respectively: Plant type
Organic waste load (population equivalents)
Trickling filters Aeration tanks (high load activated sludge) Oxidation tanks (low load) Carrousel (low load) Other Total
1,259,000 6,111,000 833,000 3,537,000 4,457,000 17,049,000
For each type of wastewater treatment systems the emissions of greenhouse gases were calculated on the basis of the overall treatment efficiencies of the different systems and the consumption of external energy resources. The total calculated CO2-emission from the wastewater treatment plants in The Netherlands equals about 880 million kg CO2 per year (about 0.5% of the total estimated COs-emission in The Netherlands). The specific emission equals 1.2 kg CO2 per kg COD removed. These emissions originates from the following sources: Source of COs
Aerobic biodegradation Use of external energy sources Use of biogas Wastage of biogas Biodegradation of organic residues after discharge Sludge disposal
Percentage
48 27 8 1.5 8.5 7
C H 4 in wastewater treatment processes is generated by anaerobic sludge digestion and anaerobic decomposition of sludge disposed of to solid waste dumping sites. The total calculated CH4-emission from wastewater treatment processes amounts to about 12.5 million kg C H 4 per year (1987), equivalent to about 1% of the total CH4-emission in
637 The Netherlands. The specific emission equals 17 g CH4 per kg COD removed. Most of the digestion gas, generated in the Netherlands wastewater treatment plants, is utilized as a source of energy, but some CH 4 is wasted or flared off. The CH4-emission into the air originates for 20% from wastage of digestion gas and for 80% from anaerobic decomposition of sludge after disposal. N20 in wastewater treatment processes is produced in nitrification and denitrification processes. Research results indicate that N20 is formed in the nitrification process as a result of non-optimal process conditions, and that in denitrification always formation of N20 takes place in conjunction with formation of N 2. The proportion between the quantities N 2 and N20 depends on the process conditions and the presence of particular micro-organisms. Literature data on N20emissions from wastewater treatment processes show large differences, varying from 0.01 to 6% of the total N-load to the wastewater treatment plant, being converted into N20. For calculation of the N20-emission from the municipal wastewater treatment plants in The Netherlands the assumption was made that by nitrification 0.3% of the N-load in the influent minus the N-load in the effluent is converted into N20 and that 0.3% of the nitrate-N, formed by nitrification, is converted into N20. On the basis of these assumptions the total NzO-emission from the Netherlands wastewater treatment plants would amount to 330 t N per year. On the basis that 1% of the residual N, discharged with the effluent, is converted into N20 , it was calculated that the N20emission from surface waters as a result of effluent discharge amounts to 415 t N per year. On the assumption that 1% of the N in sludge, disposed of to agriculture and solid waste disposal sites, it was calculated that the N20-emission from sludge disposal amounts to 84 t N per year. The estimated N20-emission from wastewater treatment processes equals about 0.9% of the total estimated N20-emission in The Netherlands.
3. MEASUREMENT OF N20-EMISSION FROM WASTEWATER TREATMENT PROCESSES In 1994 indicative measurements regarding the production and emission of N20 by two municipal wastewater treatment plants in The Netherlands were carried out. The photo-acoustic method for NzO-analysis was applied. The measurements were carried out at the wastewater treatment plant in Capelle aan de IJssel, a carrousel plant with a covered aeration circuit andsimultaneous nitrification and denitrification, and at the wastewater treatment plant in Alblasserdam, a Schreiber type plant with separated nitrification- and denitrification compartments. Samples of the air above the nitrification and denitrification compartments were taken according to a standardized method. The analyses were carried out over a period of about six hours, during which air samples were analyzed at intervals of six minutes. On the basis of the analysis results it was calculated that at the plant in Capelle aan de IJssel 0.006% of the Nload in the influent was emitted as N20 and 0.07% at the Alblasserdam plant.
638 4. R E D U C T I O N OF THE E M I S S I O N OF G R E E N H O U S E GASES F R O M WASTEWATER TREATMENT P R O C E S S E S
Measures for reduction of the emission of greenhouse gases from wastewater treatment processes should be considered in view of future effluent discharge quality standards, which gradually become stricter especially for the concentrations of N- and P-compounds. In The Netherlands the new limit values for discharge of effluent into surface waters are 10 mg N-total/1 and 1 mg P-total/1. Such values can only be realized in highly efficient wastewater treatment plants with special process steps for nitrification/denitrification and P-removal (chemical or biological). For this purpose usually ultra low-load activated sludge systems, with aerobic sludge mineralization, are applied, since these systems are the most suitable for nitrification/denitrification and incorporation of biological P-removal processes. However these low load systems are also characterized by the highest specific CO2-emission (kg CO2/kg COD removed), due to the high consumption of external energy. Hence optimization of these wastewater treatment processes should be considered in order to reduce the use of external energy resources. Measures to this effect may include: - incorporation of less energy consuming oxidation methods, e.g. trickling filters or rotating biological contactors in the wastewater treatment process; - application of anaerobic sludge digestion, with efficient use of the digestion gas, where possible. Very little is known about the exact mechanisms for formation of NzO in biological wastewater treatment processes and there are not many representative data on measurements of the N20-emission from wastewater treatment processes. Research data indicate that the emission of NzO can be reduced by creating optimum process conditions for nitrification and denitrification. 5. R E F E R E N C E S
1 Emission of greenhouse gases from wastewater treatment plants (in Dutch) Ministry of Housing, Spatial Planning and Environment BKH Consulting Engineers November 1990 2 Study regarding the formation of N20 in wastewater treatment plants (in Dutch) National Research Programme on Global Air Pollution and Climate Change (N.O.P) BKH Consulting Engineers June 1994
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
N20 E M I S S I O N S
FROM COMBUSTION
639
PROCESSES
H.Spoelstra KEMA Nederland
B.V.,
P.O.Box
9035,
6800
ET Arnhem,
Netherlands
Abstract
N20 e m i s s i o n s from the D u t c h p o w e r plants, c h e m i c a l industry, oil r e f i n e r y and a w a s t e i n c i n e r a t i o n p l a n t w e r e d e t e r m i n e d . D u r i n g s a m p l i n g s p e c i a l p r e c a u t i o n s were t a k e n in o r d e r to a v o i d the well k n o w n artifacts. N20 e m i s s i o n s at the level of a m b i e n t air c o n c e n t r a t i o n s of 0.3 ppm(v) w e r e d e t e r m i n e d at p o w e r p l a n t s f i r e d w i t h coal, oil and gas, r e f i n e r y f u r n a c e s fired w i t h a v a r i e t y of d i f f e r e n t fuels and at c h e m i c a l industry f u r n a c e s fired w i t h a v a r i e t y of h i g h c a l o r i f i c fuels. Low N20 c o n c e n t r a t i o n s w e r e d e t e r m i n e d in flue g a s e s from g a s t u r b i nes w i t h o u t a d d i t i o n a l f i r i n g in a s u b s e q u e n t b o i l e r and also at a w a s t e i n c i n e r a t i o n plant.
I.
INTRODUCTION
In 1987 EPA e s t i m a t e d that the N20 c o n c e n t r a t i o n in flue g a s e s from coal c o m b u s t i o n a m o u n t e d 20-25% of the NOx c o n c e n tration. For gas a v a l u e of 3-7% was e s t i m a t e d [i]. In 1988 it became c l e a r that t h e s e high v a l u e s w e r e due to s a m p l i n g artifacts [2]. For r e l i a b l e e s t i m a t i o n of the N20 e m i s s i o n s from p o w e r p l a n t s a m e t h o d was d e v e l o p e d in w h i c h t h e s e samp l i n g a r t i f a c t s w e r e avoided. T h e n N20 c o n c e n t r a t i o n s in the flue gases of d i f f e r e n t p o w e r p l a n t s were m e a s u r e d . Secondly the N20 e m i s s i o n s of s e v e r a l o t h e r c o m b u s t i o n p r o c e s s e s w e r e determined. This last part of the w o r k was c a r r i e d out and f i n a n c e d w i t h i n the f r a m e w o r k of the N a t i o n a l R e s e a r c h Prog r a m m a on G l o b a l Air P o l l u t i o n and C l i m a t e Change.
2.
DEVELOPMENT
OF
AN
ARTIFACT
FREE
SAMPLING
METHOD
A m e t h o d was c h o s e n in w h i c h g r a b s a m p l e s w e r e t a k e n from flue gases w h i c h w e r e a n a l y s e d on a g a s c h r o m a t o g r a p h a f t e r some time. D u r i n g a series of e x t e n s i v e tests c a r r i e d out w i t h flue gases from a p i l o t p l a n t gas b u r n e r in w h i c h a d d i t i o n a l g a s e s c o u l d be a d d e d an o p t i m a l m e t h o d was achieved. The SO2 in the flue gases was r e m o v e d w i t h a s e r i e s of w a s h i n g b o t t l e s f i l l e d w i t h a H202 solution. Then the flue g a s e s w e r e d r i e d by means of a p e r m e a t i o n dryer and s t o r e d in g l a s s sampling bottles under overpressure (0.5 bar). By a n a l y z i n g the flue gas d i r e c t l y and a f t e r s e v e r a l p e r i o d s of s t o r a g e no i n c r e a s e
640
in N20 c o n c e n t r a t i o n s (up to one week) c o u l d be detected. This was also the case d u r i n g the e m i s s i o n m e a s u r e m e n t s in w h i c h the s a m p l e s w e r e a n a l y z e d as soon as p o s s i b l e (mostly w i t h i n 24 hours) and as a c h e c k also a f t e r 3-7 days. The tests s h o w e d also t h a t s t o r a g e in s t a i n l e s s steel c i l i n d e r s in some c a s e s give a rise in N20 c o n c e n t r a t i o n .
3. N20 E M I S S I O N S
FROM ELECTRICITY GENERATION
M e a s u r e m e n t s w e r e c a r r i e d out at a v a r i e t y of p o w e r p l a n t s f i r e d w i t h coal, gas and oil. A l s o the i n f l u e n c e of load, f i r i n g m e t h o d and flue gas c l e a n i n g e q u i p m e n t was i n v e s t i g a ted. 3.1. Coal f i r e d p o w e r p l a n t s M e a s u r e m e n t s w e r e c a r r i e d out at six p o w e r p l a n t s w i t h a c a p a c i t y in the range b e t w e e n 115 and 650 MW. N20 c o n c e n t r a t i o n s w e r e b e l o w 0.2 ppm for t h r e e b o i l e r s w i t h a c o n v e n t i o n a l f i r i n g m e t h o d at full load (90 - 96%). For t w o - s t a g e c o m b u s t i on at t h r e e o t h e r b o i l e r s (180 - 520 MW) N20 c o n c e n t r a t i o n s r a n g e d from b e l o w 0.2 ppm to 0.4 • 0.2 ppm in the load r a n g e b e t w e e n 45 - 95%. No e n h a n c e m e n t of N20 c o n c e n t r a t i o n s was found at two b o i l e r s (650 and 520 MW) e q u i p p e d w i t h a flue gas d e s u l p h u r i z a t i o n p l a n t ( c o n c e n t r a t i o n s <0.2 ppm). One b o i l e r (115 MW) e q u i p p e d w i t h a DeNOx i n s t a l l a t i o n ( s e l e c t i v e c a t a lytic r e d u c t i o n - SCR) s h o w e d no i n c r e a s e in N20 c o n c e n t r a t i o n (<0.2 ppm). As the N20 c o n c e n t r a t i o n s are as well as b e l o w or s l i g h t l y above a m b i e n t air c o n c e n t r a t i o n s an e m i s s i o n f a c t o r of 0 • 0.I g N20-N/GJ was c a l c u l a t e d . 3.2. Gas f i r e d p o w e r p l a n t s The N20 c o n c e n t r a t i o n from a 640 MW c o n v e n t i o n a l gas fired p o w e r p l a n t r a n g e d from b e l o w 0.i ppm to 0.4 • 0.2 ppm N20 at a loads b e t w e e n 22% and 95%. The N20 c o n c e n t r a t i o n from a c o n v e n t i o n a l fired p l a n t w i t h n a t u r a l and b l a s t f u r n a c e gas (460 MW) was b e l o w 0.2 ppm. At one b o i l e r (180 MW) w i t h t w o - s t a g e f i r i n g N20 c o n c e n t r a t i o n s were 0.2 - 0.3 (• 0.2) ppm w i t h a load b e t w e e n 33% and 95 %. N20 c o n c e n t r a t i o n s at five g a s t u r b i n e s (40 - 130 MW) w e r e in the r a n g e b e t w e e n 0.4 and 2.8 ppm. H o w e v e r w h e n g a s t u r b i n e s w e r e u s e d in c o m b i n a t i o n w i t h a b o i l e r ("boiler w i t h t o p p i n g g a s t u r b i n e " ) the N20 c o n c e n t r a t i o n s w e r e b e l o w 0.2 ppm. A p p a r e n t l y any N20 p r e s e n t in the flue gases from the g a s t u r b i n e is d e s t r o y e d in the boiler. The e m i s s i o n factor for b o i l e r s is 0 • 0.i g N20-N/GJ. The e m i s s i o n factor for g a s t u r b i n e s was e s t i m a t e d to be in the r a n g e b e t w e e n 0.8 - 3.2 g N20-N/GJ. W h e n the total a m o u n t of gas b u r n t in the D u t c h p o w e r p l a n t s is t a k e n in a c c o u n t than the e m i s s i o n f a c t o r for gas fired p o w e r p l a n t s (boilers plus g a s t u r b i n e s ) is 0.i - 0.4 g N20-N/GJ. 3.3. Oil f i r e d p o w e r p l a n t s A l t h o u g h oil is a l m o s t not f i r e d in the N e t h e r l a n d s in p o w e r plants, m e a s u r e m e n t s c o u l d be done at one b o i l e r (180 MW) at
641
three different loads (33% - 95%). N20 c o n c e n t r a t i o n s ranged f r o m b e l o w 0.2 p p m to 0.4 + 0.2 ppm. The N20 c o n c e n t r a t i o n f r o m one oil f i r e d g a s t u r b i n e (23 MW) w a s 0.9 + 0.2 ppm. F r o m t h e s e d a t a an e m i s s i o n f a c t o r of 0 + 0.i g N 2 0 - N / G J for oil f i r e d b o i l e r s was e s t i m a t e d .
4. N20 E M I S S I O N S
FROM OTHER COMBUSTION PROCESSES
4.1. R e f i n e r y The N20 e m i s s i o n s f r o m a s e r i e s of 13 r e f i n e r y f u r n a c e s f i r e d w i t h v a c u u m r e s i d u e , r e f i n e r y gas ( c o n s i s t i n g of m a i n l y H2 and CH4) a n d fuel oil w e r e all 0.3 + 0.i ppm. 4.2. C h e m i c a l i n d u s t r y At a c h e m i c a l i n d u s t r y N20 e m i s s i o n s of <0.2 p p m w e r e m e a s u r e d at t w o b o i l e r s f i r e d w i t h e i t h e r m e t h a n e gas (98% CH4) or h i g h c a l o r i f i c gas (88% CH4 p l u s h i g h e r h y d r o c a r b o n s ) . A l s o N20 concentrations of <0.2 p p m w e r e m e a s u r e d at a g a s t u r b i n e w i t h s t e a m i n j e c t i o n and at a c o m b i n a t i o n block. 4.3. W a s t e i n c i n e r a t i o n At a m u n i c i p a l w a s t e i n c i n e r a t i o n samples were taken every h o u r d u r i n g t w o days. The 14 m e a s u r e m e n t s s h o w e d v a r y i n g N20 concentrations b e t w e e n 0.5 and 5.5 p p m w i t h an a v e r a g e of 2.8 ppm. F r o m t h i s an e m i s s i o n f a c t o r of 20 g N20 p e r t o n n e of domestic waste was calculated. 4.4. F l u i d i z e d b e d c o m b u s t i o n D a t a w e r e o b t a i n e d f r o m the c o m p a n y w h i c h r u n s a FBC. F r o m t h e s a m p l i n g p r o c e d u r e t h a t was u s e d it was c o n c l u d e d t h a t no artifacts were present. As e x p e c t e d a s t r o n g d e p e n d e n c e u p o n t h e f r e e b o a r d t e m p e r a t u r e w a s found. N20 c o n c e n t r a t i o n s in the range between 8-85 p p m w e r e m e a s u r e d at the f r e e b o a r d t e m p e r a t u r e r a n g e of 7 9 0 - 9 4 0 ~ F r o m t h e s e d a t a a N20 e m i s s i o n f a c t o r of 4.5 - 49 g N20/GJ was c a l c u l a t e d .
5. C O N C L U S I O N S The measurements i n d i c a t e t h a t t h e N20 c o n c e n t r a t i o n s from m o s t c o m b u s t i o n p r o c e s s e s s u c h as in p o w e r p l a n t s , r e f i n e r i e s and c h e m i c a l industry are v e r y low and a r o u n d ambient air concentration (0.3 ppm). N20 c o n c e n t r a t i o n s are low for g a s t u r b i n e s and w a s t e i n c i n e r a t i o n . N20 c o n c e n t r a t i o n s r a n g e f r o m 0.4 to 2.8 p p m for g a s t u r b i n e s w i t h o u t a d d i t i o n a l f i r i n g and 0.5 to 5.5 p p m for w a s t e i n c i n e r a t i o n . If the g a s t u r b i n e s are u s e d with additional firing such as in a c o m b i n a t i o n block N20 concentrations are v e r y low and a r o u n d or e v e n b e l o w a m b i e n t air c o n c e n t r a t i o n (0.3 ppm). As e x p e c t e d N20 c o n c e n t r a t i o n s for fuid bed combustion (FBC) are h i g h and r a n g e f r o m 8 - 85 p p m depending upon the freeboard temperature. The N20 e m i s s i o n s f r o m the D u t c h p o w e r p l a n t s are in w e l l
642 a g r e e m e n t w i t h a J a p a n s e study in w h i c h 43 p o w e r p l a n t s w e r e i n v e s t i g a t e d . Y o k o y a m a [3] found for coal, oil and gas fired p o w e r p l a n t s v a l u e s of 0.5, 0.3 and 0.i ppm r e s p e c t i v e l y . As in a n u m b e r of cases the N=O c o n c e n t r a t i o n s w e r e just b e l o w a m b i e n t air c o n c e n t r a t i o n s , N=O from the a m b i e n t air c o u l d be d e s t r o y e d in the c o m b u s t i o n p r o c e s s and thus r e s u l t i n g in a n e g a t i v e N=O emission. As a s u m m a r y the e m i s s i o n f a c t o r s are shown in f i g u r e i.
FB|C~ Chemical industry T
up to 49
Refineryt Oill~
Gasj mm(all installations)
Gasturbine~ Gas| Coal _
-2
~ _ ~ _ ~
0
2
+
4
L
l
,
I
6
'
'
'
I
8
10
several
combustion
N20 emission factor (g N20-N/GJ) F i g u r e i. R a n g e processes.
6.
of N20 e m i s s i o n
factors
for
REFERENCES
1 M. K a v a n a u g h , A t m o s . E n v i r o n . vol. 21, no. 3 (1987) 463. 2 L.J. M u z i o and J.C. Kramlich, Geophys. Res. Lett. Vol. 15, no. 12 (1988) 1369. 3 T. Y o k o y a m a and S. N i s h i n o m i y a , Environ. Sci. Technol. 25 (1991) 347.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
645
W o r l d I n v e n t o r y of Soil Emission Potentials (WISE): g e o g r a p h i c q u a n t i f i c a t i o n of soil factors that control fluxes of g r e e n h o u s e gases N.H. Batjes and E.M. Bridges International Soil Reference and Information Centre (ISRIC), P.O. Box 353, 6700 AJ Wageningen, The Netherlands
Abstract
Scientists at the International Soil Reference and Information Centre (ISRIC) have been developing a digital database to permit a better quantification of the role of soils in controlling processes of global change. The database is a combination of soil areadata and soil attribute-data. The area-data, based on a digitized and edited version of the 1:5 M Soil Map of the World prepared by the Food and Agriculture Organization (FAO), specify the type and extent of the main soil units for each 1//2~latitude by 89 longitude grid. The properties of these soil units can be characterized using representative soil profiles held in the WISE attribute database, for example soil organic carbon pools, moisture holding capacities, and soil fertility. In first instance, however, the WISE database has been used to make an inventory of the world's poorly drained soils, in order to provide the geographical basis for an improved estimate of methane emission potentials using auxiliary models and GIS.
1.
INTRODUCTION
Climatologists, oceanographers and atmospheric chemists involved in modelling the effect of the increase of radiatively active trace gases in the atmosphere, have taken into consideration the interaction between the ocean and the atmosphere [1]. However, relationships between processes controlling the production, absorption and emission of CO2, N20 and CH 4 in soils, and the relationships with atmospheric gas concentrations were less well understood and given less importance. As a result, difficulties were encountered in balancing the values found in the atmosphere with known values for emission and destruction of the natural trace gases in the global equation. The World Inventory of Soil Emission Potentials (WISE) project was conceived to attempt a geographical quantification of soil factors and processes that control fluxes of greenhouse gases, at a spatial resolution of 1/2~latitude by 89176 longitude, for primary use in Global Change research [2].
646 2.
THE W I S E DATABASE
2.1
Database structure
The three-year WISE project has been accomplished in two distinct phases. The first phase was devoted to a literature study of the chemical, physical and biological factors controlling the gaseous exchanges involved in order to identify the soil attributes needed to develop the WISE database [3], for further discussions during an international workshop [4]. The soil attributes to be collected may be considered in three groups: general information, physical data and chemical data (see ref. 2 for further details). These attributes, which are common to both the European soil database, and subsequent proposals for an IGBP-DIS World Soil Database appear to have gained wide support amongst the scientific community. The second phase of the WISE project started with the compilation of guidelines for the selection of soil profiles, and development of a relational data management system for handling the data. In parallel with this, colleagues at FAO have developed the griding algorithms necessary to generate the basic spatial data for the WISE database. The cartographic base has been built up mechanically, by identifying the soil units which occur in each 5' x 5' grid-cell according to FAO's composition rules [5]. The next step involved computing the percentage area of each soil unit present in the 36 cells which make up the 89 x 89 grid cell [Nachtergaele, unpublished data]. This information has then been used by WISE project staff to prepare the soil area data relevant to each terrestrial grid. Each area on the map is linked by the FAO-Unesco map legend descriptors to a suite of representative soil profiles held in the soil profile data files, which increases the value of the geographical information.
I
Representative~
SOTER World Soil and
soil profile datal
...%
Terrain Database Site data
[,5 x .5 deg. gridl
Lab. methods
;omposltion I ~oU unit 1 ~)1 un~ 2 (%) I :,~1 unit 3 (%) I ;oil unit 4 (%) I .
.
.
.
.
n~ule
I
.
_ _
SPATIAL DATA .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
i F^O~,u~
I
I. . . . . . I Sek~ted I derived =oil I properties per I deplh layer I
l l I I l
ISIS ISRIC pedon database
JOTHER SOURCES I - Reportl, etc. __ - Digital data sets
i A'I-I'RIBUTE DATA
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
" L
Data soume
L
1:5 M FAO Soil
Map of the World
I
Horizon data
grid cell
.
.
.
.
.
- . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
WISE SOIL DATABASE
.
AUXILIARY SYSTEMS
Figure 1. Schematic representation of the WISE database.
647 The attribute data, and the area data form two major elements of the WISE database (Fig. 1). In order that this information can be effectively used, it can be linked to a geographical information system (GIS) from where the soil information can be brought together with other global databases containing climatic, land use and hydrological information for subsequent modelling work and scenario analyses. 2.2. D a t a sources
An international workshop confirmed that development of the WISE database should take place, based on the present international data sets held by ISRIC's Soil Information System (ISIS), the National Resources Conservation Service (NRCS) of the United States Department of Agriculture (USDA), and FAO's Soil Database System (SDB), together with regional and national data sets where these were available and accessible. The resources of the ISRIC library collection would also be used [4]. As part of the WISE programme of activities, ISRIC initiated the development of an automated facility to transfer data from one digital profile database to another. Mechanical transfer of data is a very efficient procedure when many profiles are involved and it permits uniform standards of treatment to be maintained. The transfer program has been used to bring the data from ISRIC-ISIS, FAO-SDB and USDA-NRCS holdings into the uniform WISE format [Tempel, unpublished data]. All mechanical transfers of data have been subject to routine quality checks. Additionally, the inferred quality of the data has been flagged in the database, and source of material documented. Table 1 gives an overview of the distribution of the profiles available by broad geographic area. It will be clear that the distribution throughout the world is not uniform. Two areas significantly lacking in information are Siberia and China, and it is not surprising that few soil profiles have been described and analyzed from the central parts of the Sahara. In the absence of information, or at least its unavailability, the geographical spread of representative profiles is the best that could be achieved in the period available for compilation. Table 1. Number of profiles in the WISE database per broad geographic area'. Geographic Area
No. of profiles
Africa Australia and Pacific Islands China, India, Indonesia, and Philippines Europe North America South America and Caribbean S.W. and N. Asia
1595 85 424 490 122 486 478
" An additional 650 profiles are available from the NRCS data set, including 260 profiles from the USA.
648
2.3. Quality control Great stress has been laid upon the systematic collection and recording of data as well as careful consideration of the laboratory methods by which the analytical results were obtained. For each individual set of data the laboratory name and the methods used have been recorded. In parallel with the work of the WISE data compilation, ISRIC staff carried out a review of the comparability of different soil analytical methods, with particular attention for pH, CEC and organic carbon determinations [Vogel, unpublished
data]. Four criteria have been defined for accepting profiles into the WISE database: (a) completeness and apparent reliability of data; (b) traceability of source of data; (c) classifiable in the FAO-Unesco legend [6]; and (d) geo-referenced within defined limits. All profiles included in the WISE database have been printed-out for a visual check on the data, to complement the computerized data checking procedure.
3.
DATABASE APPLICATIONS
In view of the range of soil attributes considered, the WISE database can be used for a wide range of environmental studies at the macro level, which include assessments of crop production potentials, soil gaseous emissions and soil pollution studies. Experience gained in the collection and compilation of the profile data has shown that certain values, in most cases the physical values, are often missing. In order for the WISE database be of optimum usefulness for modelling purposes, it will first be necessary to develop pedotransfer rules or functions to fill-in eventual missing values --to to be stored in a separate derived database file---, for example bulk density data. Upon the advice of the expert panel consulted about the activities of the WISE project [4], collaboration was sought with institutes which had facilities for laboratory and field measurements of methane production, absorption and emission, and experience in CH4model development. So far an inventory of soils with high methane emission potentials has been prepared, using the WISE database linked to a GIS. This inventory forms one of the GIS data-layers necessary to refine global estimates of methane emissions from poorly drained soils. A preliminary development has been to assign soil units to one of a number of potential methane emission categories [7], but this information is still too limited for a sound spatial extrapolation using the WISE database [8, 9]. Over the past few years, much progress has been made in understanding the environmental controls on methane production, absorption and emission at site-level spatial and temporal scales. However, it remains difficult to explain regional differences in methane emission, ecological responses to climate change, and the effect of other perturbations such as drainage and flooding regimes in a global model [8, 10]. Results of this type of research are required before the potential of the WISE database for making a refined calculation of soil methane emissions can be fully utilized [9]. A recent, added scientific challenge has been to find an explanation for the currently decreasing annual increase rate of atmospheric methane concentrations [11].
649 4.
CONCLUSIONS
Although attempts have been made in the past to use the FAO-Unesco Soil Map of the World as a basis for determining the nature of the soil cover for modelling purposes, the resolution used was coarse and the potential of the map as a source of information was only partially exploited. The griding procedure developed by FAO staff as part of the WISE project has greatly increased the amount of information about the world soil pattern which can be derived from the FAO-Unesco Soil Map of the World at a scale of 1:5 M. The WISE database has assembled, in a uniform manner, over 4000 representative soil profiles with their morphological, chemical and physical attributes in a single user-friendly database. It represents a major achievement in soil science and it should soon provide a useful international soil profile data set for global modelling purposes. An important development has been the request by the Global Soil Data Taskgroup of the International Geosphere-Biosphere Project's Data and Information System (IGBP-DIS) for a subset of the WISE profile database as a foundation for their soil data gathering activity [12]. While some immediate applications have been investigated by ISRIC staff, many of the opportunities provided by the existence of the WISE database still remain to be exploited in the future. A project proposal has been formulated, in close collaboration with modellers of the IMAGE group [13], for possible implementation during the Second Phase of the Netherlands National Research Programme.
5.
ACKNOWLEDGEMENTS
Grateful acknowledgement is made to the Netherlands National Research Programme on Global Air Pollution and Climate Change (NOP project 851039) for financial support, to staff of the Land and Water Development Division of FAO (Rome) and staff of the USDA-NRCS (Lincoln) for their contributions to the WISE project. The contributions made by many individuals who provided information and their organizations are gratefully acknowledged; without them the scope of this project would have been limited.
6.
REFERENCES
1 Houghton, T.J., G.J. Jenkins, and J.J. Ephraums (eds), Climate Change: The IPCC's Scientific Assessment. Inter-governmental Panel on Climate Change, Cambridge University Press, Cambridge (1990). 2 Batjes, N.H. and E.M. Bridges, Potential Emissions of Radiatively Active Gases from Soil to Atmosphere with Special Reference to Methane: Development of a Global Database (WISE), J. Geophys. Res. 99(D8): 16,479-16,489 (1994). 3 Batjes, N.H. and E. M. Bridges, A Review of Soil Factors and Processes that Control the Fluxes of Heat, Moisture and Greenhouse Gases, p. 149-158, Technical Paper 23, ISRIC, Wageningen (1992).
650 4 Batjes, N.H. and E.M. Bridges (eds), World Inventory of Soil Emission Potentials. WISE Report 2, ISRIC, Wageningen (1992). 5 FAO, The Digital Soil Map of the World, Volume 1 Africa (Release 1.0), World Soil Resources Report No. 67/1, FAO, Rome (1991). 6 FAO-Unesco, Soil Map of the World. Vol. 1, Legend, Unesco, Paris (1974). 7 Van der Gon, H.A.C., H-U. Neue, R.S. Lantin, R. Wassman, M.R.C. Alberto, J.B. Aduna, and M.J.P. Tan, Controlling factors of methane emissions from rice fields. In WISE Report 2, pp. 81-92, ISRIC, Wageningen (1992). 8 Neue, H.-U. and R.L. Sass, Rice Cultivation and Trace Gas Exchange. In: International Global Atmospheric Chemistry (IGAC) Project: The Operational Plan, A.A.P. Psenny and R.G. Prinn (Eds), pp. 43-49, Global Change Report 32, Int. GeosphereBiosphere Programme, Stockholm (1994). 9 Batjes, N.H., E.M. Bridges and F.O. Nachtergaele, World Inventory of Soil Emission Potentials: Development of a global soil database of process controlling factors. Proceedings of International Symposium on Rice and Climate Change, International Rice Research Institute (In press). 10 Bubier, J.L. and T. R. Moore, An ecological perspective on methane emissions from northern wetlands, TREE 9, 460-464 (1994). 11 Prinn, R.G., The interactive atmosphere: Global atmospheric-biospheric chemistry, AMBIO 23, 50-61 (1994). 12 Scholes, R.J., D. Skole, and J.S. Ingram, A Global Database of Soil Properties: Proposal for Implementation. Working Paper No. 10. International GeosphereBiosphere Program Data and Information System, Paris (1994). 13 Leemans, R. and G.J. van der Born, Water, Determining the potential distribution of vegetation, crops and agricultural productivity, Water, Air, and Soil Pollution 76: 133-161 (1994).
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 1995 Elsevier Science B.V.
EMISSION DATABASE (EDGAR): V E R S I O N 2.0
FOR GLOBAL ATMOSPHERIC
651
RESEARCH
J.G.J. Oliviera, A.F. Bouwmana, C.W.M. Van der Maasa and J.J.M. Berdowskib a National Institute of Public Health and Environmental Protection (RIVM), P.O. Box 1, NL-3720 BA Bilthoven, The Netherlands
Netherlands Organization for Applied Scientific Research (TNO), P.O. Box 6011, NL-2600 JA Delft, The Netherlands
b
Abstract To meet the urgent need of atmospheric chemistry and climate modellers a global emissions source database called EDGAR has been developed jointly by TNO and RIVM to estimate for 1990, on a regional and on a grid basis, annual emissions of greenhouse gases (CO2, CH4, N20, CO, NOx, non-methane VOC, SOx), of NH3, and of ozone depleting compounds (halocarbons). The aim was to establish the global emissions from both anthropogenic and biogenic sources: a complete set of data required to estimate the total source strength of the various gases with an lo x lo resolution (altitude resolution of 1 km), as agreed upon in the Global Emissions I n v e n t o r y Activity (GEIA) of the I n t e r n a t i o n a l Atmospheric C h e m i s t r y Programme (IGAC). The data comprise demographic data, social and economic factors, land use distributions and emission factors (with due emphasis on the uncertainty). As understanding in this field is still changing, due attention is paid to flexibility regarding the disaggregation of sources, spatial and temporal resolution and species. The objective and methodology chosen for the construction of the database are presented, as well as the type and sources of data and the approach used for data collection. As an example, the construction of the N20 inventory is discussed. 1. I n t r o d u c t i o n Atmospheric chemistry and climate modellers require gridded global emissions data as input into their models. Within RIVM there is a need to monitor globally the climate affecting emissions and to gather the basic underlying data as input to the emissions calculation modules of RIVM's climate model IMAGE 2.0 (Integrated Model to Assess the Greenhouse Effect) as well as to support the validation of this model (Alcamo et a l . , 1994). On a regional scale, emission data are required for tropospheric ozone modelling, e.g. as performed with the LOTOS model (Builtjes, 1992). To meet these needs, a global emission source database called Emission Database for Global Atmospheric Research (EDGAR) has been constructed, which is able to generate for the base year 1990 on a regional and on a grid basis the annual global emissions of the greenhouse gases CO2, CH4, N20, CO, NOx (NO and NO2), non-methane VOC and SOx (SO2 and SO4), of NH3, and of ozone depleting compounds (halocarbons) from both anthropogenic and biogenic sources. The finest spatial resolution of the data is loxlo (with an altitude resolution of 1 km for aircraft emissions), as agreed upon in the Global Emissions Inventory Activity
652 (GEIA) of the International Atmospheric Chemistry Programme (IGAC), which is part of the International Geosphere-Biosphere Programme (IGBP). The temporal resolution is monthly for most natural sources, supplemented with information on the temporal variation of the other sources. The data comprise demographic data, social and economic factors, land use distributions and emission factors. As insights in this field are still changing, due attention is paid to flexibility regarding the disaggregation of sources, spatial and temporal resolution and species. Based on the conclusions of a feasibility study performed by TNO (Baars et al., 1991), RIVM and TNO have carried out a project to establish this global database. The work consists for one part of data selection, collection and processing, and for the other part of implementation of the database system (information analysis, system design, software development). Version 2.0 of the database system has been completed in J an ua r y 1995. The database is located at RIVM and serves as an analysis tool, as an emissions generator for other atmosphere modelling groups, both within RIVM and TNO and externally, and acts as the database to provide the IMAGE model with the basic data to drive the model calculations on emissions. In this short paper we will, subsequently, discuss the functionality as requested by the future users, the approach used in data selection and collection, as an example the construction of the gridded N20 inventory, and the structure of the database system. For a more detailed description we refer to Olivier et al. (1994). 2. User's r e q u i r e m e n t s EDGAR is designed to be used by modelling groups involved with atmospheric chemistry, for scenario studies and for policy assessments. The needs of Dutch modelling groups within the Dutch National Research Programme on Global Air Pollution and Climate Change (NRP) were identified in the feasibility study. The functionality proposed to fulfil the needs of potential users was discussed at a workshop held in 1992, in which participants from different atmospheric modelling groups were present. Users of the inventories are institutes and universities in the Netherlands such as RIVM, TNO, the Royal Netherlands Meteorological Institute (KNMI), the Netherlands Energy Research Foundation (ECN), the Agricultural University Wageningen (LUW), and others in particular those co-operating within the NRP, as well as international research groups. Access of publicly available data will be by FTP (File Transfer Protocol) from RIVM's anonymous FTP site (internet address:
[email protected]; node address: 131.224.1.22), by WWW, or through e-mail. 3. D a t a s e l e c t i o n and quality a s s u r a n c e Data selection on activities has been done on the basis of internationally accepted statistical data, assembled by an international organization which has performed consistency checks of the data. With respect to emission factors, representativity and availability of data as well as compliance with GEIA, OECD and European emission database systems are important aspects taken into consideration. Although for outside users only emissions estimates for major source categories will be made available, a full data source description of the underlying data is available for all data included in EDGAR. In order not to duplicate activities worldwide, RIVM and TNO are co-operating
653 a m o n g s t others with activities in the f r a m e w o r k of the Global Emissions Inventory Activity (GEIA), which is an activity of IGAC in which inventories are developed and exchanged between the participating international group. TNO and RIVM coordinate the GEIA inventories on anthropogenic VOC and on N20 emissions. Besides these inventories/contributions from EDGAR to GEIA, results from early finished inventories by GEIA and other institutes are included in EDGAR. Inventories of NO x from soils and of NH3 are also performed within the GEIA framework. Although it was not planned to include specific country inventories, national greenhouse gas inventories such as submitted to the Conference of Parties under the F r a m e w o r k Convention on Climate Change (FCCC) may also be included in the database, since an option will be provided to include and select alternative data sets (e.g. national sets) for emissions calculations. This facilitates comparison of data sets and creation of new global totals and maps by combining different sets. To g u a r a n t e e the quality of the data, besides full reference to and a careful selection of the data source, quality checks are performed on the result of the final data processing. Quality assurance will be implemented following fixed procedures on data processing (e.g. by detailed logging of all quality checks) (Van der Maas et al., 1994). Although all sources are dealt with the research was focused on the major source categories. To the extent possible, compliance with developed GEIA inventories is pursued.
4. S o u r c e c a t e g o r i e s and related data EDGAR consists of landuse-related processes, partially on a grid basis and partially on a country basis, energy related processes on a country basis, other i n d u s t r i a l production and consumption processes, waste h a n d l i n g and other sources on a country basis. Major point sources are included or used as distribution p a r a m e t e r s ; allocation functions (thematic maps) are used to convert country emissions to gridded emissions. Activity data are taken from statistical data available, e.g. from IEA (energy data), UN (industrial production and consumption), FAO (agricultural data). For biogenic sources we use gridded data, e.g. of soil types, as the basic activity data. This also defines the basic source categories used in EDGAR (Table I). Emission factors were evaluated separately. Existing inventories, e.g. of LOTOS, Corinair and NAPAP, are used to derive national or regional emission factors for surface sources. Emission factors for biogenic sources are often a function of the local climatic conditions, such as temperature, so here a more advanced approach is required. In the latter case the effective emission factors, e. g. per grid cell, have been calculated by the emissions module of IMAGE and the results have been imported in the EDGAR database. The allocation of emissions on a grid is dealt with in three ways. For major point sources, such as power plants and large basic materials producing industries, point source data are used to allocate part of national activities. If these are not present or not available - and also in case of a remaining part of national activities (while attributing the other part to point sources) - a distribution function will be used to allocate national emissions to the grid cells. Finally p a r t of the activity data were already collected at grid level. For
654 e s t i m a t i n g regional emissions we use the regional sub-division of the world as defined within the IMAGE 2.0 model as defaults (Alcamo et al., 1994). Table I Major source categories used in EDGAR and dominant sources of trace gases.
Source category
C02 CH4
N20
CO
NOx
VOC
Transportation Power generation O t h e r combustion (industry, residentials) Fossil fuel production Gas distribution Solvent use Halocarbon use O t h e r industrial processes W a s t e disposal Agriculture Live stock ( r u m i n a n t s ) Biomass b u r n i n g T e r r e s t r i a l ecosystems (soils) Oceans O t h e r n a t u r a l sources
x x x
x
x
x x x
x
x
SOx
HCs
x x
x x x x
x x x
x x x x
x
x
x x
x
x x x x
x
x
x
x
x
5. Example: construction of the N20 inventory The construction of the N20 emissions inventory is a good example of how different types of emission sources have been combined to arrive at a spatially d i s t r i b u t e d inventory. The m a j o r source categories for N 2 0 are s u m m a r i z e d in Table II and include fuel combustion and i n d u s t r i a l processes, biomass burning, arable land, grasslands and animal excreta, and n a t u r a l emissions from soils. For soils under natural vegetation we used the results o f B o u w m a n et al. (1993) a n d K r e i l e m a n a n d B o u w m a n (1994), because of the b e t t e r a g r e e m e n t w i t h a t m o s p h e r i c models and observations. For more details we refer to B o u w m a n et al., 1995. For grassland soils the N20 emission was e s t i m a t e d w i t h the above described m e t h o d for soils u n d e r n a t u r a l vegetation, using Olson et al. (1983) to allocate g r a s s l a n d s (Bouwman et al., 1995).
Soils (natural) Grasslands Arable lands4 Animal excreta Large scale burning Agricultural waste Post-burn effects Fossil fuels Fuelwood Adipic Acid Nitric Acid
Terr. ecosystems
5.2-18.1
2.7-9.7 ? 0.1-3.0 } }0.2-1.0 } 0.3-0.9 see biomass 0.4-0.6 0.1-0.3 1.4-2.6 10.6
}0.6 } 0.8 0.4 ibid. 0.5 0.2
}6.1 } 2.0
Pep per et al.
Houghton et al. (1992)
1 Used as activity defined on grid. 2 For chicken we used arable land cover on grid. 3 Used as surrogate thematic m a p for distribution of country totals. 4 Cultivated soils. * Subject to latest revisions.
Total
Oceans
I n d u s t r i a l processes
Combustion
Biomass b u r n i n g
Agriculture
S ub-category
Type
Table II Major sources of N20, global strength (in Tg N20-N y-l) and location type.
9.2
4.3 1.4 1.8 1.0 0.1 0.1 0.4 0.3 0.1 0.3 0.3 p.m.
This study*
land cover on grid1 land cover on grid1 arable land cover on grid1 animal density2 on grid3 from 5x5o grid distribution1 arable land cover on grid3 from 5x5o grid distribution1 population density on grid3 population density on grid3 point sources point sources/population density3
Location type
656 For arable land the estimates of the emissions from fertilized fields show a great u n c e r t a i n t y (Eichner, 1990; Bouwman, 1990). The method t h a t gave the best correlation with atmospheric observations was included in EDGAR (Bouwman et al., 1995). Recently, Khalil and Rasmussen (1992) identified animal excreta as a source of N20. Our estimate for nitrogen excretion for different animal categories is described in Bouwman et al. (1995). For large scale biomass burning during forest clearing, s a v a n n a burning and shifting cultivation, and for agricultural residue burning, we adopted the emission factors proposed by Crutzen and A n d r e a e (1990). The amounts of biomass burned and the grid distribution are from Hao et al. (1990). For agricultural waste burning we use a total volume of 909 Tg C yr-1 (Andreae, 1991). Emission factors are from Crutzen and Andreae (1990). The estimated emission was distributed tentatively over the arable land area within grid cells. The so-called post-burn effects of h u m a n activity in tropical forests, which m a y change emission rates of biogenic trace gases, are decribed in (Bouwman et al., 1995). Studies and m e a s u r e m e n t s of enhanced N20 emissions following savanna burning are too scarce to make even a tentative estimate for this process. Our emission estimate for fossil fuel combustion has been based on national energy d a t a from IEA (1992) and emission factors from De Soete (1993) and (Olivier, 1993). The calculated country totals are distributed according to the gridded population densities of Logan (1993). For fuelwood combustion, we used a tentative estimate based on consumption data of FAO (1991) and IEA (1992) and an tentative emission factor (see De Vries et al., 1994). Here too, we used the population density of Logan (1993) as allocation function for the national totals. Adipic Acid (AA) production data are primarily based on the production capacity and locations of plants given by Castellan et al. (1991). Emission factors are based on Reimer et al., 1992. For manufacturing of Nitric Acid (HNO3 or NA), which is mainly used as feedstock in fertilizer production, global production estimates from UN statistics (UN, 1993) and by the industry (McCulloch, 1993, pers. comm.) are inconsistent. Therefore, we adopted statistics of N-fertilizer production as a correlate for NA production (IFA, 1992). Emission factors were selected from ranges described in Bouwman et al. (1995). In the absence of point source data, distribution of emissions was done using population density as a surrogate. This example shows t h a t the various sources of N20 need to be t r e a t e d differently. Our estimates per category compare r a t h e r well with global totals estimated by Pepper et al. (1992) (see Table II). It also clearly illustrates t h a t by taking the process type of approach combined with activities defined either for areas (countries), points or grid cells, the estimation of emission factors as discussed above, and the use of selected thematic maps, we are able to construct an integrated gridded emissions inventory.
6. Structural design of the database system The central concept of the database system is the process approach. In principle, no emissions data are stored in the database, but the underlying processes t h a t cause the emissions: activity levels, environmental factors and locations (Van der Maas et al, 1994). A process is defined as an activity in which a product or waste material is transformed into another product or waste material, where energy is used and emissions are produced. This definition is quite general and can be used in
657 many ways. A process can be the production of steel, the manufacturing industry in general, the production of corn, public transportation, painting of boats, the production of electricity etc. Emissions, energy use and waste stream are calculated by multiplying the process level with the load factors. Thus, the emissions of compound x for a list of processes are calculated as: P Emissionx =E ALi * EFix i=1 where: P = the total number of processes ALi = process level (activity level) of process i EFix = is the environmental load factor (here emission factor) of compound x for process i in a historical year. Basic functionalities of the software are: system set-up (emission source categories, correspondence tables, definition of activity levels, emission compounds etc.), data import (conversion, checking, analysis, and processing), emissions calculations, inspections and exporting of results (maps, tables, aggregated input data). The development of this global database framework is part of the development of a comprehensive system of environmental databases at RIVM/LAE called RIM+, which have a national, European and global scope (Van der Laan and Bruinsma, 1993). EDGAR differs mainly from the other parts by the quantity of data to be processed, the emphasis on the location of activities (countries, regions, points or grid cells), and by the specific import/export and data analysis facilities.
7. Limitations Version 2.0 of EDGAR, which is in operation from January 1995, includes a data set covering all major sources, but does not yet include fully assessed uncertainty estimates, and has a limited functionality in modelling and calculating past and future emissions. Also time profiles to distribute emissions over months are only partially included in version 2.0 of EDGAR. The main goal has been to create a database with the information necessary to calculate globally gridded emissions in the base year 1990. We envisage that later versions will be expanded with a linkage to the IMAGE model for scenario calculations on grid. Acknowledgements The project is part of the Global Emissions Inventory Activity (GEIA), which is an activity of the International Global Atmospheric Chemistry Programme (IGAC), and of the Dutch National Research Programme on Global Air Pollution and Climate Change (NRP). This project is funded by the Climate Change Department of the Dutch Ministry of Housing, Physical Planning and Environment (Global Biosphere project, no. MAP-481507/771060) and by the Dutch NRP (no. 851060).
658 References
Alcamo, J., Kreileman, G.J.J., Krol, M., and Zuidema, G.: 1994, 'Modeling the Global Society-Biosphere-Climate System. Part I: Model description and Testing. Water, Air and Soil Pollution, 76, 1-35; special issue "IMAGE 2.0". Andreae, M.O.: 1991, 'Biomass burning: its history, use and distribution and its impact on environmental quality and global climate', In: Levine, J.S. (Ed.), Global biomass burning, MIT Press, Cambridge, pp. 3-28. Baars, H.P., Berdowski, J.J.M., and Veldt, C.: 1991, 'Preliminary study on a global emissions database (EDGAR)', TNO Institute of Environmental Sciences, Delft, TNO report R 91/136, June 1991. Bouwman, A.F.: 1990, 'Exchange of greenhouse gases between terrestrial ecosystems and the atmosphere', In: Bouwman, A.F. (Ed.), Soils and the greenhouse effect, Wiley and Sons, Chichester, pp. 61-127. Bouwman, A.F., Fung, I., Matthews, E. and John, J.: 1993, 'Global analysis of the potential for N20 production in natural soils', Global Biogeochem. Cycles 7, 557-597 Bouwman, A.F., van der Hoek, K.W. and Olivier, J.G.J.: 1995, 'Uncertainty in the global source distribution of nitrous oxide'. J. Geophys. Res. In the press. Builtjes, P.J.H." 1992, 'The LOTOS - Long Term Ozone Simulation - project. Summary report', TNO Institute of Environmental Sciences, Delft, report IMW-R92 /240. Castellan, A., Bart, J.C.J. and Cavallero, S.: 1991, 'Industrial production and use of adipic acid', Catalysts Today 9, 237-254. Crutzen, P.J. and Andreae, M.O.: 1990, 'Biomass burning in the tropics: impact on atmospheric chemistry and biogeochemical cycles', Science 250, 1669-1678. De Soete, G.G.: 1993, 'Nitrous oxide from combustion and industry. Chemistry, emissions and control', In: Van Amstel, R.A. (Ed.), Proceedings of the International Workshop Methane and Nitrous Oxide: Methods in National Emission Inventories and Options for Control, Amersfoort, The Netherlands, February 3-5, 1993, pp. 287-337. De Vries, H.J.M., Olivier, J.G.J., Van den Wijngaart, R.A., Kreileman, G.J.J. and Toet, A.M.C.: 1994, 'Model for Calculating Regional Energy Use, Industrial Production and Greenhouse Gas Emissions for Evaluating Global Climate Scenarios', Water, Air and Soil Pollution, 76, 79-131; special issue "IMAGE 2.0". Eichner, M.J.: 1990, 'Nitrous oxide emissions from fertilized soils: summary of available data', J. of Environ. Qual. 19, 272-280. FAO: 1991, 'Agrostat PC, Computerized Information Series 1/3: Land use', FAO Publications Division, Food and Agricultural Organization of the United Nations, Rome. Hao, W.M., Liu, M.H. and Crutzen, P.J.: 1990, 'Estimates of annual and regional releases of CO2 and other trace gases to the atmosphere from fires in the tropics, based on the FAO statistics for the period 1975-1980', In: Goldhammer, J.G. (Ed.), Fire in the Tropical Biota. Ecological Studies 84, Springer Verlag, Berlin, pp. 440462. IEA: 1992: 'Energy balances of OECD countries 1989-1990'. OECD/IEA, Paris. Data diskettes dated 15-04-1992. IFA: 1992, 'Nitrogen fertilizer statistics 1986/87 to 1990/91, Information & Market Research Service, A/92/147', International Fertilizer Industry Association, Paris, France. Khalil, M.A.K. and Rasmussen, R.A.: 1992, 'The global sources of nitrous oxide', J.
659 Geophys. Res. 97, 14651-14660. Kreileman, G.J.J. and Bouwman, A.F.: 1994, 'Computing land use emissions of greenhouse gases', Water, Air and Soil Pollution 76, 231-258; special issue " IMAGE 2.0". Logan, J.: 1993: personal communication. Olivier, J.G.J.: 1993, 'Working Group Report. Nitrous Oxide Emissions from Fuel Combustion and Industrial Processes. A Draft Methodology to Estimate National Inventories', In: Van Amstel, R.A. (Ed.). Proceedings of the International Workshop Methane and Nitrous Oxide: Methods in National Emission Inventories and Options for Control, Amersfoort, The Netherlands, February 3-5, 1993, pp. 347-361. Olson, J.S., Watts, J.A. and Allison, L.J.: 1983, 'Carbon in live vegetation of major world ecosystems. ORNL 5862', Oak Ridge National Laboratory, Oak Ridge, Tennessee, Environmental Sciences Division Publication No.1997. National Technical Information Service, U.S. Dept. Commerce. Pepper, W., Leggett, J., Swart, R., Wasson, J., Edmonds, J. and Mintzer, I.: 1992,' Emission scenarios for the IPCC; an update. Assumptions, methodology, and results. Prepared for IPCC Working Group 1', May 1992. Reimer, R.A., Parrett, R.A. and Slaten, C.S.: 1992, 'Abatement of N20 emission produced in adipic acid', Proceedings of 5 th International Workshop on Nitrous Oxide emissions, Tsukuba (JP), July 1-3, 1992. UN: 1993, 'Industrial Commodity Production Statistics', UN-ECE Statistical Division through International Environmental Data Service (IEDS), Geneva, Switzerland. Data on diskette. Van der Laan, W.P.M. and Bruinsma, P.H.: 1993, 'Environmental Information and Planning Model RIM+', Toxocol. and Env. Chemistry 40, 17-30. Van der Maas, C.W.M., Berdowski, J.J.M., Olivier, J.G.J., Bouwman, A.F.: 1994,' Emission Database for Global Atmospheric Research (EDGAR). Background Document', RIVM, Bilthoven, RIVM report 776 010 001 (in prep.).
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
663
Emissions inventories and options for control R.J. Swart, A.R. van Amstel, G.J. van den Born and C. Kroeze National Institute of Public Health and Environmental Protection, Bilthoven, the Netherlands
1. Introductionand background In 1990, little was known about the emissions of greenhouse gases in the Netherlands, notably those of the non-CO2 greenhouse gases. Uncertainties included the causes, the emissions factors and the regional distribution of emissions. The main objectives of the project at t h a t time were formulated as follows:
a) b) c)
provide information for prioritizing greenhouse gas emissions research in the Netherlands provide input data for global models (later shifted to the EDGAR-project) support national and international policy development
The emphasis of the project was on non-CO 2 greenhouse gases, notably methane (CH 4) and nitrous oxide (N20). While state-of-the-art information from international research would be used and analyzed, the focus of the project was on the Dutch emissions and their causes. Information was drawn from literature research, discussions with national and international experts, and experimental information from NRP and other projects. We refer to the products of the project listed in the reference list for more detailed lists of relevant literature.
2. National inventory In 1990, a first inventory of current and future emissions of greenhouse i gases in the Netherlands was performed. In this inventory, data from the international literature and expert opinions were used to arrive at estimates of national emissions (van den Born et al., 1991). Because of the relatively small area of the Netherlands, emissions from industry, energy consumption and transport generally dominate over land-related natural and agricultural emissions. Carbon &'oxide
As a result of policies, emissions of carbon dioxide were estimated to be reduced slightly by 2000, consistent with the 3-5 % government target. Options for further reduction of carbon dioxide emissions would be found primarily in energy conservation, reduction of demand and efficiency improvements, followed by a shift to non-fossil fuels. Because of the small size of the Netherlands, sequestration by afforestation programmes can only play a minor, albeit psychologically important role. Additional, technologically feasible measures in the energy sector were suggested to achieve at least another 20 % reduction, but the cost-effectiveness of these measures was concluded to be subject to controversy, as it indeed still is.
664 Methane As most important sources of methane in the Netherlands, cattle, landfills, gas distribution and the production of oil and natural gas were identified. Because of policies in the areas of waste disposal, livestock management and distribution of natural gas, emissions of methane were estimated to decrease by approximately 10 % by 2000 as compared to 1990. Figure 1" Methane emissions 1988-2000 accoording to van den Born et al. (1991) and 1990-2015 according to van Amstel et al. (1993): A = current policies and B = additional policies Methane emission scenarios 1988-2015 (kton/yr) 1988-2000van den Bornet al (1991) 1990-2015VanAmstelet al (1993) 1400
1200
1000 1-
800
600 .,_.,
e--
.e
o c,l
z
400
200
1988 ~
2000A
1990
2000B
2000A
2015A
Wetlandsand others I
Gasproduction
Gasdistribution
Ruminants
Manure
Landfills
~
2000B
2015B
Nitrous oxide As major Dutch sources of nitrous oxide the following were identified: grassland (both on organic and mineral soils), agricultural lands, transportation, and surprisingly - coastal and inland waters and waste water treatment plants. The source strengths were found to be extremely uncertain and therefore more research would be necessary to better quantify the emissions level and to evaluate response options. Through the year 2000, emissions reductions of nitrous oxide because of planned reduction of fertilizer application were estimated to be balanced by emissions increases because of the introduction of three-way catalysts in transport.
665 Halocarbons
The Netherlands' programme to phase out CFCs and reduce the associated emissions was found to effectively reduce their contribution to the enhanced greenhouse effect. On the long term it was concluded to be important that HCFCs and HFCs with a remaining (though smaller) global warming potential should be regarded as useful products only in a (short) transition period. Ozone precursors
Carbon monoxide contributes to the enhanced greenhouse effect as precursor of ozone and carbon dioxide, and competes with methane for hydroxyl radicals (OH) in the troposhere, therewith enhancing methane's lifetime. Main sources are the transportation sector and industry, notably the basic metals producers. Present policies in the transportation sector are likely to reduce emissions by 40 % in 2000. Emissions of nitrogen oxides and non-methane hydrocarbons (or volatile organics compounds, VOCs) in the Netherlands contribute significantly to climate change because of their role as precursors of ozone and the fact that the national emissions are relatively high due to intensive traffic and industrialization. It was found that the present policies to abate acidification and photochemical air pollution are also important from the point of view of climate change.
3. Support of the joint I I ~ C ~ E C D Progrsmme on Guidelines for National Emissions Inventories The expertise acquired with the fLrst national inventory was used to support the joint IPCC/OECD Programme on Guidelines for National Emissions Inventories. Amongst others, an international workshop was organized in support of the IPCC process (van Amstel, 1993a and 1994). More than 100 particants from about 30 countries discussed methods for emissions inventories and options for control for 10 source categories: methane from oil and gas, methane from coal mining, methane from ruminants, methane from animal waste, methane from landfills and sewage treatment, methane from combustion and industry, methane from rice production and wetlands, methane from biomass burning, nitrous oxide from agricultural soils and nitrous oxide from combustion and industry (Van Amstel and Swart, 1994). Also, RIVM participated actively in the International Liaison Group supporting the Programme. Jointly with the Institute for Environmental Studies of the Amsterdam Free University, RIVM staff helped the implementation of regional workshops, and performed an in-depth study for OECD to compare submissions of national inventories (van Amstel, 1993b). The ideas developed in this project have to a large extent influenced the structure and contents of the international Symposium "Non-CO2 Greenhouse Gases: Why and Howto Control?" and its Conference Statement (van Ham et al., 1994).
4. Methane emissions reductions: a side-effect of waste disposal and agricultural policies A background study on methane updated the 1991 estimate of national emissions, applying new information on methane sources (van Amstel et al., 1993).
666 The specific objectives of the methane study were: o o o
to test the IPCC methodology on 1990 data, to further refine the emissions estimates, to assess the effects of curent policies in more detail, and to assess the needs of additional policies in reaching the government 10 % reduction target.
The most important of the methane sources in the Netherlands were confirmed to be landfills, cattle, manure and the exploration, transport and distribution of oil and gas (figure 1). In van Amstel et al. (1993), the emissions of methane were determined both by applying the draft version of the IPCC/OECD Guidelines for National Emissions Inventories and by using more detailed information on the situation in the Netherlands. Even for a developed country as the Netherlands the application of the draft Guidelines for estimating the emissions from the production of oil and gas appeared to be difficult, primarily because of the absence of sufficient information about the number of wells in operation and the composition of the oil/gas mixture. A national method, which makes use of emissions factors expressed as volume percentages, is recommended. The government emissions control target, set at a 10 % reduction of the 1990 emissions by 2000, was estimated to be reached only if the planned measures to reduce landfilled waste, acid deposition and the manure surplus would be fully effective. While the relative reduction would remain as estimated by 1991, the absolute emissions levels were upgraded considerably. Notably, potential emissions from manure were found to be considerably higher t h a n estimated in the 1991 report. Additional policies for up to 30 % methane emissions reductions were found to be feasible, amongst others in the recovery of landfill gas and in the utilization of vented gas on North Sea platforms.
5. Nitrous oxide: losses at different places of the nitrogen cycle makes control difficult The study on nitrous oxide (Kroeze, 1994) had similar objectives of t h a t on methane (see section above). A comparison was made between the IPCC Guidelines and the application of a more detailed methodology, taking recent experimental data from the Netherlands into account. The detailed method, an update of the methodology of the 1991 report, led to considerably higher emissions estimates (13 - 70 Gg N/yr, central estimate 37 Gg N) t h a n in the 1991 report (638 Gg N/yr, central estimate 17 Gg N/yr). The reasons were: 1) higher local emission factors were used for key sources and 2) in the IPCC Guidelines some sources appeared not to be taken into account (figure 2).
Currently, no specificnitrous oxide reduction policies have been developed to achieve the government goal of stabilizing emissions, primarily because stabilizationwas expected to be autonomous. Kroeze (1994) however shows that emissions reductions from soils,as caused by decreasing nitrogen deposition and fertilizerapplication,willprobably be outweighed by expected emissions increases from especially the transport sector: emissions by 2000, 2010 and 2015 emissions
667 are estimated to be 1%, 6 % and 10 % higher than in 1990. The technical potentials of additional control options to reduce N20 emissions (figure 2) have been estimated for emissions from stationary combustion (15%), industry (70%), agriculture (35%) and waste (50%). Before 2000, technological options to reduce Dutch emissions include: a catalytic reduction step in nitric acid production; improved fertilzer-N use efficiency by, e.g. the use of slow-release fertilizers; improved combustion of municipal solid waste and sewage treatment, and improved combustion in power plants. Conceivable technologies that will not be implementable before 2000 include a number of promising options, such as low NOx engines in vehicles, electric vehicles, NOx reduction with low N20 formation in stationary combustion and development of modified combustors. Figure 2: Total Dutch emissions of N~O in 1990, the theoretical potential (TP) to reduce 1990 emissions (if nothing else would change) and the 2015 projections based on current policy (M 3a), National Environmental Policy Plan 2 (NMP-2) and additional policy (AP) Dutchemissionsof N20
I
40 I-
'~
30
z ..,,_., v t.2
E o z
20
1990 Natural
1990TP ~
Agriculture ~
Energy Waste
2015MV3a ~
2015NMP2
2015AP
Industry Other
6. Conclusions and recommenchfions for further work The project "Emissions inventories and options for control" has facilitated the development of the national capabilities to comply with the requirement of the Framework Convention on Climate Change that countries should regularly report about their emissions of greenhouse gases and national policies. The methodologies
668 developed and applied in the project, will be used for future national National Environmental Outlooks and national communications to comply with the requirements of the Convention. The project has laid out the special circumstances that govern the emissions of greenhouse gas emissions in the Netherlands: because of the high intensity of population and economic activities - including intensive agriculture, industry, transport - the vast majority of emissions are of anthropogenic origin. The implementation of policies in areas other than climate policies - notably the prevention, disposal and treatment of solid waste and manure and the abatement of acidifying nitrogen emissions - has a counteracting effect on the tendency of growing non-CO2 emissions that would result from the growth of economic activities.
7. Selected
publications (full references contained in these reports):
Amstel. A.R., van, R.J. Swart, M.S. Krol, J.P. Beck, A.F. Bouwman and K.W. van der Woerd: "Methane: the Other Greenhouse Gas", report no. 481507001, RIVM, Bilthoven, 1993 Amstel, A.R. van (ed.): "Proceedings International IPCC Workshop on Methane and Nitrous Oxide: Methods in National Inventories and Options for Control", RIVM, Bilthoven, 1993a Amstel, A.R. van: "Transparency of National Emissions Inventories", in: van Amstel, A.R. (ed.): "Proceedings International Workshop on Methane and Nitrous Oxide: Methods in National Inventories and Options for Control", RIVM, Bilthoven, 1993 Amstel, A.R. van and R.J. Swart: "Methane and Nitrous Oxide: an Introduction", in Fertilizer Research, vol. 37, pp. 213-225, 1994 Amstel, A.R., R.A.W. Albers, C. Kroeze, A.C. Matthijsen, J.G.J. Olivier and J. Spakman: "Greenhouse Gas Emissions in the Netherlands 1990, 1991, 1992 and Projections for 1990-2010", report nr. 773001003, RIVM, Bilthoven, 1994 Amstel, A.R.: "Methane emissions and control in the Netherlands", in: J. van Ham, L. Janssen and R.J. Swart (eds.): "Non-CO2 Greenhouse Gases: Why and How to Control?", pp. 515-525, Kluwer, 1994 Bouwman, A.F., G.J. van den Born and R.J. Swart: "Landuse-related Sources of CO2, CH 4 and N20: Current Global Emissions and Projections for the Period 1990-2100", report no. 222901004, RIVM, Bilthoven, 1992 Born, G.J. van den, A.F. Bouwman, J.G.J. Olivier and R.J. Swart: "The Emissions of Greenhouse Gases in the Netherlands", report no. 222901003, RIVM, Bilthoven, 1991 Ham, J. van, L.J.H.M. Janssen and R.J. Swart: "Non-C02 Greenhouse Gases: Why and How to Control?", Kluwer, 1994 Kroeze, C.: "Nitrous Oxide: Emissions Inventory and Options for Control in the Netherlands", report no. 773001004, RIVM, Bilthoven, 1994 in preparation Olivier, J.G.J.: " Inventory of Aircraft Emissions: a Review of Literature", RIVM report no. 736301008, Bilthoven, 1991 Swart, R.J., A.F. Bouwman, J.G.J. Olivier and G.J. van den Born: "Inventory of Greenhouse Gas Emissions in the Netherlands", Ambio, vol. 22, no. 8, 1993
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
669
The role of population growth in global CO2 emissions J.P. van Ypersele a and F. Bartiaux b alnstitut d'Astronomie et de G6ophysique G. Lemaitre, Universit6 catholique de Louvain, Chemin du Cyclotron 2, B-1348 Louvain-la-Neuve, Belgium. Internet: vanYpersele@astr, ucl.ac.be blnstitut de D6mographie, Universit6 catholique de Louvain, Place Montesquieu 1, B-1348 Louvain-la-Neuve, Belgium. Internet:
[email protected]
Abstract
The principle of "differentiated responsibilities" of North and South in protecting the climate system against global warming is recognized in the Rio Framework Convention on Climate Change. We focus here on the quantification of one aspect of this issue: population growth versus growth in COJcapita emissions. We first mention several problems raised by the way the Ehrlich-Holdren equation (Environmental impact = Population times Per capita impact) is used in the context of greenhouse gas emissions. In particular, we remind the importance of using the lowest possible aggregation level with this equation. We then apply this equation to population and fossil fuel-related CO2-emission data for nine regions of the world over the 1950-1990 period. The results of a scenario analysis using these data show that the increase in developed countries CO2 emission per capita had a significantly larger impact on world total emission increase than LDC's (Less Developed Countries) population growth during that period. It is also shown that population growth in developed countries had a larger effect than LDC's population growth.
1. INTRODUCTION In order to allow a 'sustainable climate' (Gouz6e and van Ypersele, 1992), i.e., a climate that does not change faster than the speed at which the economy, and the natural and agricultural ecosystems can adapt, the concentration of greenhouse gases in the atmosphere must be stabilized as soon as possible, which implies a global reduction of emissions. To achieve this, the Rio Framework Convention on Climate Change (FCCC) states in its Article 3: 'The Parties should protect the climate system for the benefit of present and future generations of humankind, on the basis of equity and in accordance with their common but differentiated responsibilities and respective capabilities.' This principle of differentiated responsibilities raises the question of the role of population growth as a factor in increasing global emissions of greenhouse gases. This paper explores a way of quantifying the importance of this factor compared to the CO2/capita increases observed over the 1950-1990 period.
670 2. USE AND AND MISUSE OF THE EHRLICH-HOLDREN EQUATION (I = PF) A classical way of quantifying the links between environmental issues and population size is illustrated by the following equation, originally published by Ehrlich and Holdren (1971): I = P F, where I is a negative impact on the environment, P the population size, and F is a function which measures the per capita impact. That equation was subsequently rewritten as (Ehrlich and Ehrlich, 1990, p. 58): I = P A T, where I is a negative impact on the environment, P the population size, A stands for affluence or per capita consumption, and T is 'an index of the environmental disruptiveness of the technologies that provide the goods consumed.' Although the I =PAT equation relates CO2 emissions to three factors, population is often seen as the main one. It may be because population statistics are more readily available to climatologists and ecologists than economic data, or because many MDCs (More Developed Countries) writers believe that it is more important or easier to control LDCs population growth than consumption patterns in their own countries. For example: '[...]The atmospheric concentrations of these [greenhouse] gases are tightly tied to population size. Consequently, there is no practical way to achieve the necessary reduction in greenhouse emissions without population control' (Ehrlich and Ehrlich, 1990, p. 58-59). However, this 'consequence' cannot be justified: statistical association does not necessarily mean a causation. Indeed, the fact that population and carbon dioxide emissions grew over the same period does not imply that there is a simple causal link between the two. This logical error is related to the problem of aggregation. The I = PAT equation can certainly yield useful indications on the roots of environmental degradation, but only if applied to a specific homogeneous region or country. If it is applied to the world as a whole, average values will hide the deviations from the mean and the variations of the distributions. Since the fraction of the world population which grows at the fastest rate is also the one with the lowest per capita emissions, using average population growth rate and average per capita emissions will overestimate world emissions. An example of faulty aggregation made by the Ehrlichs is again taken as granted and cited in full in a recent UNFPA-published book (UNFPA, 1991, p. 17): the Ehrlichs (1990, p. 59) say: 'to illustrate how this interaction [between reduction in greenhouse emissions and population control] works, suppose that, by dint of great effort, humanity managed to reduce the average per-capita consumption of resources on the planet (A in the I =PAT equation) by 5 percent and improved its technologies (T) so they did 5 percent less damage, on the average. This would reduce the total impact (I) of humanity by roughly 10 percent.' This calculation is meaningful only if one assumes the distribution of technology and per capita consumption to be uniformly distributed among every humans, which is not presently the case. But the Ehrlichs continue with: 'unless population growth (P) were restrained, however, its growth would bring the total impact back to the previous level in less than six years.' Since the main part of the population growth occurs in developing countries, it is incorrect to multiply this increasing population by a decreasing consumption and an improving technology that are both supposedly occurring in a developed country in this example. A similar faulty aggregation is made in 'Earth in the Balance', the book written by US Vice-President Gore (1992, p. 309-310).
671 3. THE ROLE OF POPULATION GROWTH IN PAST CO2 EMISSIONS
To try to quantify the role of population growth in past CO2 emissions, we now present the results of a set of scenarios that show how important the world CO2 emissions from fossil fuel and cement would have been had the population and/or the CO2 per capita followed different paths of evolution during the period 1950-1990. The observed emissions used as a basis for the simulations come from Marland and Boden (1991). They are given for nine regions I of the world between 1950 and 1990. Figure 1 displays the evolution of population and CO2 emissions in MDCs and in LDCs during the period of study. tCO2/capita
Population (Billions) 6 " - P o p . MDC
-~- Pop. World 4- tCO2/cap. MDC
14
5 ::::::::::::::::::::::::::~.:S-::..
~
~ ' ~/ = - : " " - ~ 1 2
. . . . . . . . . .~~.,r :".....
8
3 :.-- " : 7 i~
6
-x- tCO2/cap. LDC
2 .........
-~-tCO2/cap. World
1
1950
4
" ~'*::::::
1960
. . . . . . . . ~ : :::: : : : : : : : ' : : : : : :'::: ::_- 2
1970
1980
1990
Year
Figure 1. Evolution of population (continuous lines) and CO2 emissions per capita (dashed) in MDCs and in LDCs from 1950 to 1990 (based on data from Marland and Boden, 1991). Table 1 presents the results for four scenarios (additional scenarios and background analysis can be found in Bartiaux and van Ypersele, 1993). Although the results are summed up for the MDCs and LDCs, the simulations were performed separately for the nine regions with their corresponding population figures and CO2 emissions per capita. This disaggregation into nine regions is more accurate than performing the calculations for two groups only (MDCs and LDCs) but is yet probably too crude to fully avoid the heterogeneity within the nine regions (see Lutz (1992) for additional comments on the aggregation problem). For each scenario Table 1 also compares the simulated total CO2 emission to the
1 Marland and Boden (1991) subdivided the world into the following nine regions: North America (USA and Canada), Western Europe, Eastern Europe (including former Soviet Union), Centrally Planned Asia (including People's Republic of China, Viet Nam, North Korea, and Mongolia), Far East (including India, South Korea, Indonesia, Taiwan, Thailand, Pakistan, Malaysia, the Philippines .... ), Oceania (mostly: Japan, Australia, and New Zealand), Developing America (the American continent less USA and Canada), Middle East (Saoudi Arabia, Iran, Turkey . . . . ), and Africa.
672 observed value for the corresponding year and the difference is indicated in per cent. It thus gives the influence of the factor(s) considered in that scenario. Admittedly, these simulations also neglect a number of factors, including emissions from deforestation (for which data are more uncertain), emissions of other greenhouse gases, and the effect of trade. Table I. The role of population growth in past CO2 emissions: scenarios For each scenario and for each year, CO 2 emissions from fossil fuel and cement are given in million tonnes of CO 2 (MtCO2) for the more developed countries (MDC), the less developed countries (LDC), and the world total (TOT). The calculations have been made with nine separate regions. The last line (D%) shows the relative change in world emissions compared to the observed data for the same year.
Observed data
1970 MtCO2
1980 MtCO2
1990 MtCO2 14665 6676 21341
Source: Marland and Boden (1991)
MDC LDC TOT
5358 462 5821
7604 1589 9193
11781 2411 14192
14152 4415 18567
1. Pop. MDC = real data Pop. LDC = 1950 value CO2/ca p MDC = real data CO2/cap LDC = real data
MDC LDC TOT D%
5358 462 5821 0
7604 1301 8905 -3.1
11781 1517 13297 -6.3
14152 2223 16375 -11.8
2. Pop. MDC = 1950 value Pop. LDC = real data CO2/ca p MDC = real data CO2/ca p LDC = real data
MDC LDC TOT D%
5358 462 5821 0
6608 1589 8197 -10.8
9171 2411 11582 -18.4
10123 4415 ! 14538 -21.7
3. Pop. MDC = 1950 value Pop. LDC = 1950 value CO2/ca p MDC = real data CO2/cap LDC = real data
MDC LDC TOT D%
5358 462 5821 0
6608 1301 7909 -14.0
9171 1517 10687 -24.7
10123 2223 12346 -33.5
9778 2786 12565 -41.1
7586 4415 12001 -35.4
8149 6676 14825 -30.5
!
i
!
14665 2786 17452 -18.2 !
9778 6676 16454 -22.9
!
4. Pop. MDC = real data Pop. LDC = real data CO2/cap MDC = 1950 value CO2/cap LDC = real data
The first set of scenarios deals with the effect of population growth alone. Scenario 1 tests the effect on the CO2 emissions of an hypothetical stationarity of the LDCs populations at their 1950 level, with the observed evolution of the CO2 per capita emissions in the different regions: the result would have been 18 % less world emissions than the real figure for 1990. Conversely if the MDCs populations had not grown since 1950, the difference would have been higher: -23 %, as shown by scenario 2. Of course these two gains would be added to each other had both the LDCs and MDCs populations been blocked (scenario 3). Thus for the period 1950-1990, the MDCs population growth played a greater role in the increase of the world CO2 emissions than the LDCs population growth did, because the CO2 emissions per capita multiplier is so large in MDCs. To demonstrate the effect that aggregating can have on such calculations, we also computed scenario 3 without disaggregating the world in nine regions, by using a world average for COJcapita: the decrease in total emissions in 1990 compared to 1950 is then -52.5 % (instead of -41.1% without aggregation). This shows clearly how aggregation contributes to overestimate the role of population growth.
673 The fourth scenario addresses the effect of CO2 per capita. If we block the CO2 per capita in the MDCs at their 1950 values and have the MDCs and LDCs populations grow as they really did, the world emissions would have been significantly less than observed: -15 % in 1960, -30% or more since 1970 (scenario 4). The comparison between scenarios 4 and 3 indicates that the effect of the populations growth in both MDCs and LDCs could have been offset until 1980 by the sole blocking of MDCs CO2 per capita rates to their 1950 values. 4. CONCLUSION These scenarios results illustrate in a striking manner the concept of 'differentiated responsibilities' introduced in the Rio Convention. They tend to show that, for fossil fuel CO2 emissions between 1950 and 1990, past population increases in developing countries have contributed much less to CO2 increases than either increases in consumption in MDCs or even population growth in MDCs did (scenario 1 compared to scenarios 4 and 2, respectively). Since changes in consumption patterns are affected by a weaker structural inertia than population is, it may be argued (see Meadows et al., 1992) that it would be more rapidly effective to put the emphasis on changing the energy and resource consumption pattems, specially in rich countries. By doing so, the growth of the world CO2 emissions could be braked while waiting for the outcome of the reduction of population growth. Such reduction may indeed be necessary, not as much because of past responsibilities of population growth in global warming, as shown by Table 1, but to allow a sustainable development. As most demographers have long recognized, the relationships between demographic, social, economic, and environmental factors are complex. In a report made in preparation of the Rio Conference, the Overseas Development Administration of the United Kingdom Government (1992), addresses the issue of population growth in the following terms: 'the more the scope for an induced reduction in population growth is seen to be real and significant -- for example by meeting an unmet demand for family planning -- the more this encourages the view that population is a direct causal factor which should be manipulated. On the other hand, the more the fertility decisions are seen to be a function of a complex of structural socio-economic factors, the less it can be seen as a simple lever to be used, and therefore the less population growth can be seen as an independent cause' (p. 10). [...] 'Indeed, there are risks attached in tying the rationale for family planning too closely to poorly understood relationship between population growth and environmental degradation. The links are too complex and uncertain, and risks of overemphasis too great, for environment to be highlighted as a major reinforcing rationale for family planning. In these circumstances it may be better to recognize that slowing population growth does not constitute a short-term solution to environmental degradation, and that working with the multiplicity of positive linkages between measures which benefit women and environmental management represent the constructive way forward' (p. 21).
Acknowledgements We would like to thank Thomas A. Boden (CDIAC, Oak Ridge National Laboratory) for providing us with C02 emission data in computer format.
674 5. REFERENCES
Bartiaux, F., and J.P. van Ypersele, 1993, The role of population growth in global warming, in: International Population Conference, International Union for the Scientific Study of Population (IUSSP), ISBN 2-87108-030-5, Liege, Belgium, 4, 33-54. Ehrlich, P.R. and Ehrlich, A.H., 1990, The population explosion, Simon and Schuster, New York, 320 pp. Ehrlich, P.R. and Holdren, J.P., 1971, Impact of population growth, Science, 171, 12121217. Gore, A., 1992, Earth in the Balance, Earthscan, London, 408 pp. GouzEe, N. and van Ypersele, J.P., 1992, Objectif 9 un climat 'soutenable', La Revue Nouvelle, Brussels, BE ISSN 035-3809, vol. 95, No. 4., 124-133. Lutz, W., 1992, 'Population and Environment- What do we need more urgently: better data, better models or better questions?', paper presented at the annual conference of the British Society for Population Studies, 'Population and Environment', September 911, at Exeter College, Oxford University, 16 pp. Marland, G. and Boden, T., 1991, COz emissions - modem record, in: Trends '91: A compendium of data on global change, ORNL/CDIAC-46. Boden, T.A., Sepanski, R.J., and Stoss, F.W., (eds.), Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, Oak Ridge, Tennessee, USA, 386-425. Meadows, D.H., Meadows, D.L., and Randers, J., 1992, Beyond the Limits, Earthscan, London, 300 pp. Overseas Development Administration of the UK Government, 1992, 'Population, Environment and Development: an Issues Paper for the Third UNCED Preparatory Comittee', London, 22 pp. UNFPA, United Nations Population Fund, 1991, Population, resources, and the environment: The critical challenges, New York, 154 pp.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
677
A S S E S S M E N T R E P O R T ON N R P SUBTHEME
'rEFFECTS OF CLIMATE CHANGE TERRESTRIAL ECOSYSTEMS"
ON
S.C. van de Geijn DLO-Research Institute for Agrobiology and Soil Fertility, (AB-DLO) P.O.Box 14 6700 AA Wageningen The Netherlands
With contributions by: P. Dijkstra, A. Gorissen, A.H.C.M. Schapendonk K. Kramer, G.M.J. Mohren
AB-DLO, Research Institute for Agrobiology and Soil Fertility, Wageningen IBN-DLO, Institute for Forestry and Nature Research, Wageningen
F. Berendse, Th. Jetten, C. de Kovel, W. Takken
LUW, Agricultural University of Wageningen
H. Lankreijer, A.W.L. Veen
RUG, University of Groningen
W. Bouten, M. Schaap J. Rozema, A.Visser
UvA, University of Amsterdam VUA, Free University Amsterdam
678 Contents Abstract 0
Introduction 1.1 General outline 1.2 Ecosystem interactions 1.3 Types of ecosystems considered 1.4 Ecosystem stability 1.5 Ecosystem evolution and succession P l a n t g r o w t h and carbon allocation in a soil-plant s y s t e m 2.1 Introduction 2.2 Primary production 2.3 Effects on transpiration at the level of stomata and canopy 2.4 Carbon partitioning and soil organic matter dynamics
0
I m p a c t s on forest e c o s y s t e m s 3.1 Introduction 3.2 Growth of Douglas fir trees 3.3 Changes in organic matter profiles and availability of water in forest soils 3.4 Water balance of a forested area 3.5 Phenological reactions of Dutch tree species and frost sensitivity
0
0
0
0
0
8.
Carbon e c o n o m y of grasses 4.1 Introduction 4.2 Primary production 4.3 Water use 4.4 Degradation of organic matter grown in elevated CO2 P h y s i o l o g y and productivity of arable crops 5.1 General 5.2 Basic plant physiological studies 5.3 Effects of the OTC's 5.4 Biomass formation and yield I n t e g r a t i o n of effects of climate c h a n g e on terrestrial e c o s y s t e m s 6.1 Introduction 6.2 The model 6.3 Progress and plans I m p a c t on ecology and distribution of malaria vectors 7.1 Introduction 7.2 Simulation models 7.3 Distribution of the malaria vector 7.4 Concluding remarks Closing remarks
679 9.
References
ABSTRACT In the projects fostered by the NRP the effects of changed climate (atmospheric CO2 concentration, temperature) on different terrestrial ecosystems were studied. For forests it was concluded that the initial stimulation of tree growth in general did not persist after two years, and therefore care must be taken not to overestimate the potential contribution of increased carbon sequestration by forests. On the other hand, shifted patterns of carbon distribution in the tree-soil system may lead to a higher soil organic matter content, which will contribute to an improved soil structure and availability of soil moisture. A sensitivity analysis revealed that, for the poor sandy forest soils, improved rooting depth is however more effective for drought prevention than higher soil organic matter. From the model exercises it was also inferred that with increased precipitation, as predicted u n d er the projected future climate, runoff and recharge of the groundwater will be increased, especially in winter and for deciduous forest stands. Changed seasonal temperatures will change the timing of phenology but will, in Dutch conditions, not lead to a higher risk for spring frost damage in the period of bud burst. However, competition between tree species may change as the duration of the closed and functional canopy is differentially influenced. For instance for Larix and oak the effective growing season is extended, whereas beech and Tilia cordata will have a shortened leaf area duration. Mechanistic forestry production models, adapted to include also the changes in [CO2], (e.g. in a transient climate scenario), showed that after a transient increase in production, a double-CO2 climate from GCM calculations caused a subsequent decline in productivity. A high variability of the growth and production enhancement by rising [CO2] was also detected in the OTC (Open Top Chambers) and Rhizolab experiments for the crop species studied (potato, wheat, faba bean). The physiological parameters (photosynthesis and respiration) and full season canopy and soil gas exchange m e a s u r e m e n t s showed no growth stage or light and t em perat ure dependent CO2-enhancement effect. An analysis using crop growth models produced clues as to the origin of the existing confusion about the v a r i a b i l i t y of the CO2-enhancement factor for biomass production and yield. Using the growth and weather data of different years it could be shown that interactions between growth stage, light and especially temperature in the early growth stages could explain a large par t of the variation. Also about half of the differences in growth enhancement between e.g. (winter) wheat (16 - 34%) and faba bean (35 - 56%) could, according to the model outcome, be ascribed to temperature differences in the early (spring) growth stages. An important and for the carbon cycle very significant finding was that the roots of grass, grown at 700 ppm CO2 were degraded much slower by the soil organisms
680 t h a n reference root material. It was shown that this change in properties m a y fully offset the stimulation of the decomposition of soil organic m a t t e r by the projected temperature rise. In a pilot study the potential distribution within Europe of mosquitoes that can act as a vector for malaria transfer was investigated. The combined effects in the various growth stages of the mosquito, as brought together in a simulation model, indeed point to a highly increased risk for infectious individuals. The probability of an epidemic is considered low, as the European health care system is expected to be sufficiently effective in picking up disease incidence. The concept of "infection potential", developed in this project, offers excellent possibilities to quantify risks also for other, more vulnerable, areas or world wide studies. The methodology developed in this project can be used for other (also agricultural) pests and diseases. 1.
INTRODUCTION
1.1 General outline In the Dutch National Research Programme on Global Air Pollution and Climate Change NRP several aspects of the functioning of widely differing t e r r e s t r i a l ecosystems are considered, varying from physiology, development and productivity of agricultural crops, input and turnover of soil organic matter, characteristics of forest soils and impacts on trees, to the risk of changes in d i s t r i b u t i o n of mosquitoes as vectors for the spreading of malaria. A project aiming at the integration of knowledge at the level of semi-natural ecosystems and thereby to estimate the possible disturbance of the natural succession in such systems was only recently started. In this assessment a selection of processes, tools and (sub-)systems is considered, and results are discussed t h a t may contribute to a further development of our understanding. Each of the contributions has a merit in its own right, but its added value is realised by linking to the knowledge base of the international scientific community. After an introduction this report loosely follows the projects within the subtheme "Effects of climate change on terrestrial ecosystems" and discusses the findings in the general context outlined below. 1.2 E c o s y s t e m i n t e r a c t i o n s Terrestrial managed- or n a t u r a l ecosystems are by definition functioning and developing in a p e r m a n e n t interaction with external environmental conditions. During their course of development they loose or acquire minerals and organic matter, and thereby modify their internal regulation of the structure and further characteristics. As such they are never in a condition of a stable equilibrium, but e s s e n t i a l l y in a dynamic and t r a n s i e n t state, moving from one stage of development to another. From this general notion it follows that changes in the environment, be it natural or man-made, may perturb or disrupt the development path, destabilize the system or affect the rate of development. Changes m a y be g r a d u a l and difficult to assess, or sudden and s u b s t a n t i a l when related to
681 instabilities in the system. Transitions to a new quasi-stationary state may be triggered by incidents like storms, droughts, fires or pests. Also socio-economic conditions may be of great influence. Climate change in that respect adds to the environmental factors that determine and change the variability and predictability of ecosystem functioning and stability. Table 1.1 List of projects in the NRP Subtheme "Effects of climate change on terrestrial ecosystems" Title
Project leader
Number
The role of organic matter profiles in the effects of climatic changes on the availability of water in forests
W. Bouten
850010
Phenological reactions of the main Dutch tree species to climate change described by a simulation model of the annual cycle
G.M.J. Mohren 850014
Interactions between atmospheric CO2-concentration, A.H.C.M. temperature and environmental factors with respect Schapendonk to photosynthesis, assimilate distribution and development rhythm of three agricultural crops
850020
Distribution of carbon over plant and soil compartments during the growth of perennial plants under increased CO2 concentrations
A. Gorissen
850029
Potential impact of climatic changes on the ecology and distribution of malaria vectors
W. Takken
851057
Integration of effects of climate change on terrestrial ecosystems
F. Berendse
853126
Effect of an increase of greenhouse gases on the water balance of the forested land surface
A.W.L. Veen
850015
Not only the intrinsic value of (natural) ecosystems to mankind (although poorly defined like beauty, ecosystem function) is relevant to the issue of climate change. Changes in ecosystem composition or activity will inevitably have an impact on the pools and fluxes of water and carbon dioxide in the environment. Such changes also may modify greenhouse gas fluxes (N20, CH4, water), production as well as absorption, may be influenced, as the magnitude of the fluxes varies between ecosystem types and with ecosystem functioning. Moreover, characteristics like surface roughness and albedo are of direct relevance to the climatologists. The information on the impact of climate change on terrestrial ecosystems is therefore
682 also of direct relevance for the estimation of feed-backs in the global climate system (see Theme " The climate system").
1.3 Types of ecosystems considered It may be clear that an assessment of the impacts of climate change on terrestrial ecosystems requires differentiation, depending on the type of system considered. Intensively managed systems like agricultural production systems may require a change in choice of crop type or crop cultivars and a modification of the cropping system, management and market structure. Agricultural systems may also change in productivity and (economic) viability, especially in marginal production areas. Although impacts of climate change on agriculture in the highly developed technologically advanced Western economies possibly can be absorbed without major risk for food security or shortage, the situation on local and regional markets and with respect to international and intercontinental trade and transportation of food and feed may be affected substantially. A more dramatic situation may arise in regions where local economies do not have access to the economical and technological means to counteract or reduce the threats posed by climate change. Food shortage may lead to starvation or trigger the migration of whole populations. Similarly, measures to protect extensively or unmanaged ecosystems or to counteract unintended developments may also require a substantial economic basis. However, to judge the risks for the various ecosystems properly, we have to look primarily for the changes in driving forces determining functioning, growth and production potential of e.g. various plant types, crop species, vegetations and forests. Physiological processes, water relations, soil organic matter dynamics and competitive relationships are altered by elevated atmospheric CO2 concentrations, modified (seasonal) temperatures and precipitation patterns.
1.4 Ecosystem stability At the next higher integration level from physiological processes in plants and soil, the impact at the level of ecosystem processes should also be taken into account. Forests and semi-natural ecosystems have to be considered as being largely dependent on their internal coherence and consequent resilience to perturbations. As stated, climate change as such is superimposed on top of the existing natural and manmade environmental changes, be it variability or trendwise developments. A proper estimation of the impact therefore requires a complete and coherent picture at a system level and a quantitative description of the selected systems in their present state, their dominant processes and relationships. The projects on plant- and crop physiology, soil science and soil organic matter dynamics find their application in estimating agricultural and forest productivity, ecosystem functioning and hydrological relationships. Another basis for the estimation of changes in forest ecosystem development is provided by the study of tree phenology in a changed climate.
683
1.5 E c o s y s t e m e v o l u t i o n and s u c c e s s i o n Apart from short term changes and risks, a long term perturbation may happen. Pools and fluxes of carbon and mineral nutrients in nutrient poor systems largely determine the dynamics of vegetation succession. The understanding of the effect of climate change on such ecosystem processes, or ecosystem physiology, is r e q u i r e d to e s t i m a t e the i n t e g r a t e d effect on the stability or the r a t e of development. A s e p a r a t e study has addressed the potential change in the distribution of the mosquitoes t h a t can act as a vector for malaria. Interpretation of results of the population ecology and other ecological interactions still has to be combined with socio-economic scenarios to estimate the relevance in terms of health risks. As such this type of studies clearly shows the complexity of long-term interactions and the need for caution when interpreting the outcome of sectorial and specialised studies. 2.
PLANT G R O W T H AND CARBON ALLOCATION IN A S O I L - P L k N T SYSTEM
2.1 I n t r o d u c t i o n P r i m a r y photosynthesis globally takes up as much as 60 - 70 Pg carbon per a n n u m . In addition the oceans have a yearly exchange of the same order of magnitude. Above ground standing biomass is estimated to contain about 500 Pg C, whereas the atmosphere contains about 700 Pg C. These n u m b e r s compare with a yearly emission of 6 Pg C from fossil fuel burning. It m e a n s t h a t the equivalent of the total carbon content of the standing biomass is t u r n e d over by these processes every 8 - 10 years, and of the atmosphere every 6 - 7 years. The total carbon storage in soil organic m a t t e r amounts to approximately 1500 Pg C, of which about half is almost inert. Residence times in various compartments of soil organic m a t t e r m a y range from very short (days to weeks) to extremely long (thousands of years). At the estimated input into the soil the m e a n turnover time in the active part would be of the order of 30-50 years, but this n u m b e r has only limited value, given the range of residence times. Although the dimensions of these numbers give a fair idea about the magnitude of the interactions, the net fluxes are much smaller. Most ecosystems are in a type of steady state, where m e a n losses of carbon by respiration and decay of organic m a t t e r are roughly in equilibrium with C-gains through photosynthesis or other inputs. Even small changes in the balance between C-fixation and release from the storage pools of a system may however have i m p o r t a n t consequences for the long-year cumulative outcome. Especially the pools in the soil t h a t have long residence times are of interest. Pools and fluxes differ between various ecosystems, and depend on activity and structure of communities. Inputs consist of litter accumulating on the soil surface and in the surface layers and of root material derived from living and dead root systems.
684
2.2 P r i m a r y p r o d u c t i o n The changed composition of the atmosphere will have a direct effect on plant processes. The most significant impact will be on photosynthesis. As carbon dioxide is the p r i m a r y s u b s t r a t e for the p h o t o s y n t h e t i c process a r i s i n g concentration in general will enhance the production of assimilates, although not proportional. This is true for plants with the so-called ca pathway, where the concentration of carbon dioxide at the site of photosynthesis inside the leaf at present ambient concentrations is limiting. The enhancement is small or absent for C4 plants. At present most plant species in t e m p e r a t e areas have a c a p a t h w a y , whereas C4 plants are predominantly found in w a r m e r climates and tropical areas. For the Dutch and West-European situation the most prominent exception to the general dominance of ca plants is maize. 2.3 Effects on transpiration at the level of stomata and c a n o p y E l e v a t e d levels of CO2 not only enhance the supply of CO2 as s u b s t r a t e for photosynthesis, but also modify the gas exchange properties of leaves in a canopy in a different way. The opening status of the stomates is a compromise between water loss and uptake of CO2 from air. The increased levels of carbon dioxide cause a partial closure of stomates, and consequently reduce water loss by transpiration. This is also reflected at a leaf level (30 - 60% lower water loss). However, results at a canopy level are not always clear as m a n y other processes interfere. Because of a r e d u c e d t r a n s p i r a t i o n the e v a p o r a t i v e cooling is reduced, a n d canopy t e m p e r a t u r e s rise. Also, the growth e n h a n c e m e n t by elevated CO2 leads to a larger standing biomass, and an increased evaporating surface. As a result various outcomes are possible, and very much depend on periods of water shortage and the timing thereof. 2.4 Carbon partitioning and soil organic matter d y n a m i c s An increased uptake of CO2 by the plant biomass may result in a changed supply of root derived products to the soil microbial biomass. The proper quantification of the g r a d u a l changes in pools and fluxes of soil organic carbon requires special techniques, as only minor changes may occur, t h a t can accumulate however to significant levels over the years. Q u a n t i t a t i v e l y , up to 40% of the gross a m o u n t of carbon assimilated in the photosynthetic process is allocated directly to the root system and rhizosphere during the season. Root respiration and breakdown of easily decomposable organic compounds in the root rhizosphere r e t u r n the carbon almost i n s t a n t l y as CO2. Also p a r t of the biomass built from assimilates produced by photosynthesis is r e t u r n e d in a r a t h e r short cycle. In forest systems e.g. the equivalent of the complete fine root system is t u r n e d over within one y e a r (Olsthoorn 1991, Olsthoorn & T i k t a k 1991). With m a n y annual species of arable crops the full non-harvested biomass is r e t u r n e d as crop residue, and together with the root mass (2 to 4 tons of dry m a t t e r per ha), is subject to decay. Up to 60% is easily decomposable, and returns within one year to the atmosphere as CO2. P a r t of the r e m a i n i n g material is more resistant to degradation, or is converted into partly humified m a t e r i a l by soil organisms, thereby entering the pool with longer residence times in the soil.
685 It is of great importance to assess the direction and magnitude of changes in the different types of pool sizes and fluxes as a consequence of changed atmospheric CO2 concentrations and induced by a change in the climate (temperature, moisture). Such changes may be triggered not only through a direct effect on primary productivity, but also through changes in the quality (composition) of the plant material. This quality, reflected e.g. in carbon to nitrogen ratio (C/N) or lignin content of the material, determines the degree of resistance to decomposition by soil fauna and flora (meso-fauna, bacteria, fungi). The quality of the plant material is also of importance for animals (herbivorous macro-fauna) feeding on the plants. The relationship is even more intricate, as the below-ground processes like mineralization re-supply plants with mineral nutrients, thereby closing the nutrient cycle. These pools and fluxes determine to a large extent the productivity and the composition of the vegetation. The complex interactions with moisture, temperature and mineral nutrient availability require a cautious and multi-faceted approach. 3.
I M P A C T S ON F O R E S T E C O S Y S T E M S
3.1 I n t r o d u c t i o n Because of the importance of forest ecosystems a major question in the framework of climate change research within the NRP has been the estimation of the sensitivity of these systems. This applies to growth of individual trees, species composition and stability of the forest ecosystem.
Climate change may affect forests in a complex way (Figure 3.1). Both direct and indirect effects can be envisaged. As stated above, rising CO2 concentrations will enhance primary production while simultaneously improving water use efficiency. Moreover, rising temperatures may alter phenological development and changing precipitation patterns will also affect growth and hydrological relations. Changes in growth strategy may affect carbon allocation and litter production and distribution as well as litter quality, resulting in modifications of the soil organic matter content, quality and distribution over the soil profile. It is virtually impossible to create a comprehensive picture of all interactions and processes, and answer in general terms to the questions at hand. Therefore efforts have been concentrated on some major issues. In the sub-theme "Effects of climate change on terrestrial ecosystems" following aspects of impacts on forests have been specifically studied: * Douglas fir growth and organic matter dynamics: productivity, quality, conversion rates and pools of organic matter; * hydrological aspects of changes in organic matter content in forest floors; * atmospheric exchange and hydrological relations for a forested land surface; * Phenological development and forest productivity.
686
CLIMATIC
Soil ~ 0rg.
Water/
CHANGE Figure 3.1 Schematic interactions between climate change and forest characteristics
3.2 G r o w t h of D o u g l a s fir trees
P r o d u c t i v i t y a n d carbon allocation To contribute to an improved interpretation of the highly variable results reported in l i t e r a t u r e concerning persistence of the e n h a n c e m e n t of photosynthesis, Douglas fir trees Pseudotsuga menziesii were grown at ambient and elevated (double present) CO2 concentrations. The long term growth enhancement and the partitioning and utilisation of assimilates was followed in time. Using special equipment allowing periodic labelling of the atmosphere with 14CO2 at different CO2 concentrations short t e r m fluxes from the plant to the root and root r h i z o s p h e r e were studied. The hypothesis was tested t h a t the s u p p l y of a s s i m i l a t e s to the root and root environment is stimulated at elevated CO2 concentrations. To ascertain t h a t the effects studied were not (only) t r a n s i e n t p h e n o m e n a , trees and grasses were p r e - t r e a t e d for 14 m o n t h s at the two experimental CO2 concentrations. Subsequently, the plants were exposed for a 4-week period to the 14C-CO2 labelled atmosphere, to trace the fluxes of assimilates. In the experiments with Douglas fir the initial CO2 uptake, expressed in biomass gain was stimulated by about 25%, but this stimulation fell to lower levels with time. Exposure to elevated CO2 for 14 months resulted in 12% more needle biomass and 16% more roots. Root biomass gain is therefore stimulated more t h a n t h a t of branches or needles. Three year old trees appear to have a relatively higher root biomass gain than four year old trees. However, d u r i n g the 4 weeks of exposure to a 14C-labelled a t m o s p h e r e , m e a s u r e m e n t s showed t h a t the total CO2 uptake of the trees p r e - t r e a t e d and
687 treated at 700 ppm was higher than that at 350 ppm, but t h a t the stimulation of photosynthesis of trees pre-treated at 350 ppm exposed to 700 ppm during the m e a s u r e m e n t was higher. This was true in both absolute uptake and the more so w h e n expressed per unit needle biomass. When expressed in this w a y the photosynthesis was reduced by 14% after a p r e t r e a t m e n t at 700 ppm CO2. The reduced photosynthetic activity per unit needle biomass is also found w h e n switching trees from 700 to 350 ppm. Such changes m a y have i m p o r t a n t consequences for the potential for carbon storage to be expected in forests. Q u a n t i t a t i v e conclusions have to wait for the results of additional studies and integration thereof.
Soil a n d root respiration In the high-CO2 t r e a t m e n t at both age classes a higher specific activity was m e a s u r e d in the respiration fluxes from the soil compartment. This m a y be explained postulating t h a t the soil and rhizosphere organisms in this situation have a preference for the (increased) direct supply of energy rich assimilates from the roots, as compared to the older non-labelled organic material in the soil. It would also m e a n t h a t the microbial biomass and the organic m a t t e r fraction involved in this rapid turnover is not subjected to effects of changes in C/N ratios in the material, or lack of mineral nutrients due to enhanced growth of the trees. W a t e r use a t e l e v a t e d C02 Cumulative w a t e r use is changed hardly or not at all, but is lower per unit needle biomass (14-16% lower). Water used for the production of a unit new biomass is also reduced (32% lower). Consequently in unchanged precipitation conditions more biomass can be formed and supported in situations where w a t e r poses a serious limitation, or otherwise more water is becoming available to recharge groundwater reserves or contribute to run-off and surface water storage. Coneluding remarks The general picture can be summarised concluding that initial growth stimulation of Douglas firs does not persist over longer periods. The larger biomass built up initially does not result in a continued higher specific growth rate. It is not realistic to assume t h a t the larger biomass would require proportionally more energy for maintenance, or t h a t tree architecture would lead to a substantial reduction in light interception per unit needle biomass. The latter fact may be at the basis of the high variability of the growth enhancement factor often reported for seedlings in small scale and short duration experiments. From the detailed results of the present experiments it has been concluded t h a t a limited sink strength of the root system, as proposed in literature, does not play a role. A similar conclusion was also drawn from experiments with crop plants (see Section 5.2). Physiological and morphological adaptations and effects of changes in the nutrient requirements and n u t r i e n t availability have not been considered here, but may have consequences t h a t modify p r e s e n t results, but are difficult to assess systematically.
688
3.3 Changes in organic matter profiles and availability of w a t e r in forest soils Introduction Because of the importance of forest ecosystems a major question in the framework of climate change research is the estimation of the sensitivity of these systems. This applies to growth of individual trees, species composition and stability of the forest ecosystem. The question arises also what would be consequences of changes in content and type of soil organic matter due to changed growth characteristics and biomass quality of the trees and of altered organic matter decay rates under climate change (temperature, moisture). The above discussed potential changes in soil organic matter (Section 2.4) therefore are not only relevant for the global carbon cycle and an improved estimation of the projected changes in carbon storage in terrestrial ecosystems. The organic matter in (forest) soils is also very important for the general soil characteristics, especially for structural properties, fertility and water holding capacity. In other words, changes in the amount and quality of the organic matter deposited on and in the soil have an effect on the quality of the soil for various uses. Higher organic matter content will lead to better hydrological properties. Depending on the soil type (particle size and size distribution) the effective water storage capacity e.g. is increased. This is especially true for sandy soils with a deep water table, where most of the Dutch forests are located. For most of these forest soils, where the organic matter content is below 2 to 3% the effect of increases in organic matter can be substantial. Organic matter also lowers the bulk density, and enhances possibilities for deep penetration of roots. Again, this helps to supply trees with water in dry periods.
Role o f organic m a t t e r in sandy forest soils The organic matter in forest soils is distributed over the profile, and varies in properties from barely decomposed leaf litter and branches, decaying roots and partly decomposed material on and in the forest floor to roots and partly or completely humified organic matter in the mineral soil. Soil organic matter influences forest hydrology by affecting evaporation from the forest floor, but also by enhancing soil water retention in the root zone. For most conditions on Dutch sandy soils with a deep water table, the amount of required water for evapotranspiration (ca. 400 mm/year) is not available from storage plus summer precipitation, although the year-total of precipitation may be around 800 mm. This water shortage periodically leads to water stress and thus limited growth. The present project aimed at combining the general understanding of soil physical properties and collecting data in such a way as to allow the evaluation of the sensitivity of the forest hydrological system to climate change. The evaluation itself is not part of this project.
689 From the soil physical characterisation of a range (8 types) of forest soil profiles q u a n t i t a t i v e r e l a t i o n s have been derived t h a t allow a classification a n d quantitative description of most Dutch sandy soils. Organic m a t t e r lowers the bulk density, and thereby enhances the development of soil structure and its stability. As a consequence the shape of the water retention curve is changed. Dependent on the soil particle characteristics the effective water storage capacity is increased. The effect is stronger for former drift sand soils with uniform particle size distributions t h a n in soils with a non-uniform distribution. This difference is brought about by a better pore size distribution in the latter. For mineral soils with organic m a t t e r content below 2 or 3 volume percent the effect is substantial. An additional effect of soil organic matter is its influence on soil wetting and drying characteristics: the hysteresis gap becomes larger with higher soil organic m a t t e r content. The consequences for water availability of this phenomenon still have to be evaluated.
Forest floor evaporation Direct evaporation from the forest floor can constitute a substantial fraction of the available water. Contrary to expectations, the formation of a dry top layer of the organic forest floor does not limit evaporation losses. Apparently the low-density organic material forming the forest floor acts as a (dense and inverted) canopy, where evaporation takes place at all depths. Moreover water movement from the mineral soil to the organic cover by capillary rise can not be neglected. The thickness of the forest floor is found to be an important factor for forest floor hydrology. The supply of w a t e r down to the mineral soil does not depend on the drainage rate of the forest floor, probably because of the widely diverging time constants. The e s t i m a t e d evaporation in a dense Douglas fir forest was 85 mm/y, but is expected to be higher in more open forests where both mean radiation level and wind speed at soil level may be higher. Water availability for evapotranspiration depends not only on water storage, rainfall and retention characteristics (3.2), but also on rooting patterns and root growth strategies. An evaluation of the results in an extended and coupled version of a model for forest hydrology (FORHYD) indicated that the sensitivity of forests for differences in w a t e r availability caused by organic m a t t e r dynamics is small. Probably trees mostly escape conditions of water shortage by extending roots to greater depths. However, also here organic m a t t e r makes a positive contribution, as root growth is enhanced by a lower soil bulk density down the profile. Ongoing sensitivity analysis, still to be completed, concerning forest floor hydrological dynamics related to future climate scenarios will help to estimate better impacts on forests.
690 Rainfall 750 m m / y
Transpiration
:~.I'~!:~200
- 400 m m / y
....: ,,~ ~:~,
Interception 100-350 m m / y
Evaporation ;0 - 150 m m / y
Storage 50 - 300 mm Drainage 50 - 200 m m / y
Figure 3.2 Generalised forest hydrological cycle for the Dutch situation
3.4 W a t e r b a l a n c e
of a forested area
The w a t e r balance of a forested a r e a very m u c h depends on the s t r u c t u r e of a forest, forest canopy and the availability of soil water. A study at a l a r g e r spatial scale was also conducted, using a one dimensional computer simulation model. The results are primarily reported in the sub-theme "Regional Hydrology". F o r t h e s t u d y at a l a r g e r s p a t i a l scale a physiologically b a s e d model w a s i n c o r p o r a t e d to e v a l u a t e the direct effect of elevated CO2 c o n c e n t r a t i o n s on photosynthesis, s t o m a t a l conductance and w a t e r use. To e v a l u a t e the i m p a c t of a climate change, a sensitivity analysis and a climate scenario according to the KNMI-2 scenario (with a relatively strong increase in precipitation, especially in winter) were applied. In a g r e e m e n t w i t h the r e s u l t s from other studies in this subtheme, a low increase of productivity was applied by i n c r e a s i n g t h e leaf a r e a index by 5%. The r e s u l t s show a s e n s i t i v i t y to t h e availability of soil water. Total seasonal w a t e r use of forests m a y change relative to p r e s e n t conditions w i t h b e t w e e n -20% and +10%. Forests suffering w a t e r l i m i t a t i o n show an increase in w a t e r consumption. This m a y even be s t r o n g e r w h e n soil w a t e r storage is increased as concluded by the s t u d y on soil organic m a t t e r . Forests with no w a t e r limitation consume less w a t e r in a changed climate. In general w a t e r shortage is reduced, and more of the excess w a t e r in the w i n t e r period is available to recharge the groundwater.
691
3.5 P h e n o l o g i c a l r e a c t i o n s of D u t c h tree s p e c i e s in r e l a t i o n to frost sensitivity, growing season and primary productivity Introduction Trees are suggested to be in particular vulnerable to climate change because of their long life span and the period of several decades to reach the reproductive stage. Genetic adaptation is therefore too slow and much depends on the plasticity of i n d i v i d u a l trees and tree species to respond physiologically a n d / o r morphologically to changed local conditions. Important phenomena in the annual life-cycle of deciduous trees are bud burst and leaf area development, and the date of leaf fall or preceding senescence. To study the sensitivity to climate of these events for different species, data were collected and used to develop a phenological model. A differential response of trees of different species will not only have a direct effect, but may also, or even more importantly, result in changed competitive relations in a mixed stand. This aspect would need further attention as a follow-up of the present study, which has been concentrated on phenology and productivity changes of individual species.
A n n u a l life-cycle o f deciduous trees Important phenomena in the annual life-cycle of deciduous trees are the moment of bud burst and leaf area development, and the date of leaf fall or preceding senescence. These define the physiologically active period, where light intercepted by the foliage is converted into biomass. This period is for deciduous trees also coinciding with a high water demand. The timing of leaf unfolding is mainly regulated by temperature. For temperate tree species chilling and forcing temperatures are both required to induce leaf unfolding. In other words, a minimum low-temperature exposure and a m i n i m u m h i g h - t e m p e r a t u r e exposure are both required. It is not a priori clear w h e t h e r w a r m e r winters will advance or delay the date of leaf unfolding: the chilling requirement may be attained later, while the critical t e m p e r a t u r e sum for leaf unfolding is likely to be reached earlier in spring. Such shifts m a y have consequences for the occurrence of frost damage. To test the sensitivity of the timing of bud burst to climate several types of models were developed or improved and tested.
Phenology models The phenology models were calibrated and tested using long-year records of phenological development of beech Fagus sylvaticus in the N e t h e r l a n d s and adjacent parts of Germany. Additional data sets of 13 other species but with a shorter record were obtained from phenological gardens in Europe. The models with the best prediction (modified "sequential" and "alternating" model) were used to test the probability of spring frost damage in the bud burst period under various climate change scenarios. Results were compared with similar studies in the UK and Finland.
692
Bud burst and frost damage Based on an analysis of tree clones transferred over Europe, it is argued t h a t the survival of trees is curtailed by spring frosts, and t h u s t h a t the lowest t e m p e r a t u r e occurring around leaf unfolding may be a sensitive indication for the geographical distribution of species. Furthermore it is found that trees possess a considerable plasticity, which enables them to accommodate by phenological adaptation a significant change in climate. With respect to climate warming, in general the probability of frost damage is predicted to be reduced in Dutch, German and UK conditions, for the scenarios and models used (temperature rise uniform throughout the season or temperature rise season dependent). In Finnish conditions however probability of frost damage is found to be increasing. This disparity is due to local climatic conditions, causing predicted bud burst in Finland to be advanced much more than at the other locations. The response of bud break of the earlier tree species to temperature is higher t h a n of those which unfold their leaves in the first weeks of May. This enhanced separation in timing may have consequences for the competitive relationships, as competition for light in early spring is changed.
Length of the growing season As stated, not only bud break but also date of leaf fall is of high significance in tree species performance. Some species (Larix decidua, Quercus robur) are found to end up with a shortened growing season, while others (Fagus sylvatica, Tilia cordata) extend it at higher temperatures as a result of an earlier leaf unfolding without concomitant early leaf fall. Such effects may also have consequences for the competitive relationships.
P r i m a r y productivity models Using a mechanistic model, the combined effect of elevated atmospheric C02, temperature and water shortage was explored for Norway spruce. Results showed t h a t the CO2 response is enhanced in conditions of water shortage. This is due to the fact t h a t water shortage is partly alleviated by the effect of CO2 on stomatal closure. Similarly, the larger increase of respiration of the standing biomass with rising t e m p e r a t u r e s as compared to photosynthesis results in a reduced productivity of cool t e m p e r a t e species when temperatures increase. (6% resp. 2% for potential growth, but 14% resp. 6% in water limited conditions at 350 and 700 ppm CO2 respectively). Climate change scenarios were used to evaluate the sensitivity of three tree species (birch, beech and oak) as predicted by process-based tree growth models. Modified current weather data as well as synthetic weather data, modified using GCM output, were used. A transient climate scenario was constructed using a 100 year r a m p to the 2 x CO2 temperature level (mean GCM outcome). Rainfall was
693 modified according to temperature relations, keeping the number of rainy days constant. In these conditions forest productivity was found to increase initially in the transient scenarios, but was reduced at the 2 x CO2 level, because of an increasing negative effect of temperature on productivity (Figure 3.3).
Net Primary Production 15
B. pubescens
F. sylvatica
Q. robur
o~,~,,.,, v--? QFDL
,!-.-
'~,10
OI86
'7"
N5
NN
! I transient I 2 x CO 2
OSU
UKMO
transient 2 x CO 2
I transient I 2 x CO 2
Figure 3.3 Model prediction of productivity of birch, beech and oak forest under various (GCM-)climate change scenarios
Concluding remarks Primary productivity models, largely based on correlative relationships, can be used to estimate forest ecosystem productivity in present conditions. More complex mechanistic models of forest growth may also be used to explore the range of effects of climate change on biomass production and carbon storage to be expected. Validation of such models is only partly possible as very few time series of sufficiently detailed forest growth data have been collected. Such more physiological models are however extremely useful tools to explore the possible outcome of the complex reaction of a forest ecosystem to atmospheric [C02] and temperature changes. Basic relations derived in this way may subsequently be simplified in summary functions and included in primary production models like the model FORGRO. Some points should be taken into consideration with respect to these findings. The outcome of these physiological-based models should be interpreted cautiously. A serious drawback is t ha t data to calibrate and validate the forest productivity models are very scarce. Moreover, dry matter partitioning, nutrient relationships and water use are only incorporated in a descriptive way, and the quantitative relationships may itself be modified by climatic conditions. The impact of temperature on respiration is important when considering long-term effects. Temperatures could deviate considerably from present patterns. It is however still unclear if the short term response of respiration to higher temperature, as used in the model, is also applicable after long-term exposure to higher temperatures. The sensitivity determined in short term experiments
694 amounts to about 7% per ~ and therefore lower values, (e.g. 3% as suggested by Gifford, 1994) might change the picture importantly. Nevertheless the model exercise and literature study has revealed t h a t i m p o r t a n t changes in forest ecosystem productivity may occur, although the probability of spring frost damage (in Dutch and UK conditions) is decreased.
4.
C A R B O N ECONOMY OF G R A S S E S
4.1 I n t r o d u c t i o n Grasslands dominate an important part of the terrestrial ecosystems, and do play an important role in carbon storage, in some respects with a potential comparable to forests (Goudriaan 1993a, b; Fischer et al, 1994). Especially the below-ground storage of carbon is quantitatively of great importance. Experiments similar in setup to t h a t with Douglas fir (Section 3.2) were also done with different species of grasses (Lolium perenne and Festuca arundinacea). In these experiments the effect of nitrogen supply was studied using two rates of fertiliser application. Plants were grown in a pretreatment for 14 months at 350 resp. 700 ppm CO2 in a greenhouse. After acclimation in a special growth cabinet groups of plants were exposed for 24 hours to a 14C-C02 labelled atmosphere. During the 24 hour t r e a t m e n t and the succeeding 3 weeks of the experiment the plants were hermetically sealed at the base, to allow the determination of (combined) root and soil respiration separate from CO2 exchange of the plant tops. Carbon dioxide produced in the soil c o m p a r t m e n t was trapped in sodium hydroxide, and both absolute a m o u n t and specific activity were determined. This allowed the separation of the contribution of different sources of carbon: respiration of labelled recent assimilates from the root and associated soil biomass, and combined non-labelled material from the root and from native soil organic matter. 4.2 P r i m a r y p r o d u c t i o n Elevated C02 does stimulate the growth of grasses, but similar to what was found with Douglas fir, the effect is falling over the period of a prolonged exposure. The initial growth enhancement was about 25%, but it dropped to about 16% over a period of 66 weeks. At the final harvest yield differences were almost absent. Both grass species behaved similarly, although the yield response of F. arundinacea was somewhat higher. As the effect was similar at both nitrogen levels (135 and 400 kg N/ha) N-limitation as a cause for the falling growth enhancement does not seem likely. At the highest N - t r e a t m e n t root biomass at final harvest was still strongly stimulated by CO2, in contrast to the low-N application, where CO2-enhancement of both root and shoot growth disappeared over time. The 14C label was used to discriminate between biomass g r o w t h d u r i n g pretreatment, and during and after the labelling period. The results show that most of the larger root systems of the 700 ppm-400 kgN-plants had been formed prior to the labelling. However, as the labelling took place in September an interaction with the season can not be excluded. The analysis of the respiration data showed t h a t
695 the percent distribution of assimilates over the compartments did not change with CO2 concentration. The higher application of N-fertilisers however reduced root respiration with 29%. Also the percent 14C in the microbial biomass dropped (from 2 to 1% at 350 ppm CO2). Also at 700 ppm CO2 the 14C labelled fraction dropped (by 20%). Residual labelled organic material increased by about 100% in the 700 ppm treatment. 4.3 W a t e r u s e The p r e t r e a t m e n t had an effect on the w a t e r use t h a t persisted during the t r e a t m e n t . Grass grown at 700 ppm CO2 used less water: both total (-16%) and per unit leaf mass (-25%) when exposed to 350 ppm CO2. At 700 ppm CO2 the w a t e r use was reduced by as much as 35%, independent of p r e t r e a t m e n t . F. a r u n d i n a c e a was more sensitive, reducing water loss per unit leaf area with 47%, compared to 21% for L. perenne. A first extrapolation could be t h a t F. a r u n d i n a c e a is capable of taking better advantage of the conditions with higher CO2 and the more so during periods of water stress. 4.4 D e g r a d a t i o n of o r g a n i c m a t t e r g r o w n in e l e v a t e d CO2 The above discussed experimental results are obtained with organic m a t t e r derived from plants while growing in ambient or double-present CO2 concentrations. The carbon sequestered or released is newly fixed, and only within growing season effects are considered. This might give results t h a t differ in some respects from a normal, more representative growing cycle, where plant residues are left behind at the end of a growth cycle. To test the effects of potential differences in quality of the plant material grown at both CO2 concentrations, the rate of decay of such material has been analysed. The 14C-labelled plant material therefore has been incorporated into the soil to determine the effect of changes in the composition (quality) of the m a t e r i a l caused by growth conditions at elevated CO2. Two t e m p e r a t u r e s (14 and 20oC) and soil moisture levels were used. Roots were t a k e n from L o l i u m p e r e n n e grown and uniformly labelled with 14C for 4 weeks. The different growing conditions were clearly reflected in the C/N ratio. Roots of L. p e r e n n e grown at 700 ppm CO2 had a C/N ratio of 32, whereas for the control plants C/N was 18.
The total production of C02 at 2 ~ consistently exceeded t h a t at 14 ~ accumulating to 30% difference after 64 days. Root derived respiration increased by 26%. The two levels of soil moisture imposed in the e x p e r i m e n t s had no significant effect. The decomposition of roots grown at 700 ppm CO2 was accelerated during the first two days, but the r a t e decreased m a r k e d l y after 8 days. At the end of the incubation high-CO2 roots had released 24% less 14C-CO2. The decomposition of the high C02 roots at 20 ~ essentially paralleled t h a t of 350ppm root material at 14 ~ The rate of decomposition of the root material had no effect on the decomposition of the native soil-organic matter.
696 These results show that the change in quality of the plant material as shown here, can h a m p e r decomposition, and possibly partly negate or even a n n u l the accelerating effect of a temperature rise as induced by greenhouse gases. 5.
P H Y S I O L O G Y AND P R O D U C T I V I T Y OF ARABLE C R O P S
5.1 General Introduction The persistence of the growth enhancement of plants at elevated concentrations of atmospheric CO2 is still under debate. Although mechanisms have been studied and clarified, the transient nature of the growth enhancement and the variable nature are much less understood. Part of the problems in the experimental studies can be related to experimental conditions, where growth environments have to be created to allow controlled exposure to elevated CO2 concentrations. In such e n v i r o n m e n t s plant characteristics m a y be modified, leading to c h a n g e d sensitivity. One aim of the experimental work therefore was to determine the changes in growth characteristics and yield of crop plants in near-field conditions. Therefore three selected arable crops were exposed during the entire growing season to ambient and double-present CO2 concentrations in OTC's (Open Top Chambers). The research was aimed at answering such questions as the nature of species differences and seasonal variations in plant, crop and vegetation response to climate change conditions ([CO2], temperature). Detailed studies of source-sink relations, daily and seasonal variations in assimilation, assimilate distribution and total biomass production have been done. The more detailed studies with individual plants and different growth stages and growing conditions were used to support the seasonal studies. The results of the experiments were combined and mechanistic simulation models were used to evaluate the various findings and extrapolations.
S e l e c t e d crop species and e x p e r i m e n t a l facilities Rising CO2 concentrations enhance photosynthesis and thereby the availability of assimilates for various plant processes. Three crop types differing in assimilate utilisation were selected: * wheat was selected as small grain crop, being a type of "reference crop" used in much of the international climate change research; * faba beans can develop a symbiosis with nitrogen fixing bacteria which convert atmospheric nitrogen to a plant-available form. To s u s t a i n this process the plants have to supply the bacteria in the root nodules with energy in the form of carbohydrates. As such the nitrogen fixation is a drain on the assimilates of the plant; * potato plants can be characterised as strong starch accumulators during tuber bulking, and were expected to profit specifically from the enhanced assimilate availability. The effects of elevated [CO2] were studied experimentally in Open Top Chambers, greenhouses and a Rhizolab facility.
697
5.2 Basic plant p h y s i o l o g i c a l studies P o t size To elucidate some aspects of the effect of smaller pot sizes as reported in literature, special tests were set up aiming specifically to avoid shortage of minerals and water. Although for winter wheat a strong reduction in growth was seen for smaller pot sizes (range used: 0.8 to 10L), CO2-enhancement of growth was not affected. In experiments with faba bean results were more variable, but the general conclusion is t h a t at optimal supply of water and nutrients pot size effects can be avoided. I n t e r a c t i o n w i t h UV-B Special attention was given to the effect of the presence of UV-B in the solar radiation reaching the plants. To this end OTC's were constructed from special material t r a n s m i t t i n g UV-B. Results point to an interaction between UV-B and CO2. Also some effects on plant morphology were detected. Assimilate distribution and photosynthesis In the detailed experiments the distribution of sugars over various parts of wheat plants was studied. No differences could be detected in assimilate and nitrogen distribution, apart from an increased carbohydrate content in the wheat leaves in March. Crop a r c h i t e c t u r e was also u n c h a n g e d . Detailed p h o t o s y n t h e s i s m e a s u r e m e n t s at the leaf level with wheat and faba beans grown in OTC's and greenhouses as used in Amsterdam (and Wageningen at the canopy level) did not show the occurrence of photosynthetic acclimation.
5.3 Effects of t h e OTC's The temperature inside Open Top Chambers is higher than in open field conditions, a l t h o u g h this varies over the day. These higher t e m p e r a t u r e s do have an accelerating effect on plant development, and in general lead to an earlier harvest (about 2 weeks earlier for wheat). The chamber effect on biomass production is however very much influenced by the weather conditions of the specific season. For instance early (or late) high temperatures or drought will affect the crop inside the chamber in a different way, as the development stage or standing biomass is different. For wheat this resulted in 1993 in a clear chamber effect for both winter wheat and faba bean, but in 1994 only for faba bean. A full analysis of the data with crop growth simulation models may reveal whether plants differ in various growth stages regarding their sensitivity to weather conditions. An unexpected r e s u l t was the striking effect of the OTC's on the quality (composition) of the plant material. The carbon content was lower (37% as compared to 38.5%), but the effect on nitrogen content was even more prominent. For wheat N-levels at harvest dropped for stems from 8 to 4 mg/g, for leaves from 10 to 6.4 mg/g but for grain it went up from 21 to 23.3 mg/g. Such differences are very important when using plant material for decomposition studies (see above), and may result in misleading outcomes. In general there was also a slight increase of the harvest index (fraction of the total dry m a t t e r in the harvested product). For wheat the effect was stronger t h a n for
698
faba bean. In 1993 there was no significant effect for faba bean. Differences in harvest index may for faba bean also be influenced by the stage of maturation. It was observed that both with potato and faba bean maturation of the plants tended to be delayed, whereas the faba bean seeds and pods ripened at the same time. Slight differences in moisture content, combined with differences in remaining leaves (including shed leaves) may influence the outcome. A special aspect of the use of OTC's is the existence of (micro meteorological) gradients within the OTC. In pot experiments, as done in Amsterdam, regular rotation of plants can compensate for non-uniform conditions. However, for field grown plants especially the light gradient causes growth differences that put limits to sampling of plants within the chamber. As general conclusion it can be stated that although OTC's can be a helpful (and necessary) tool for experimentation in elevated CO2 conditions, changes in plant growth and development can not be avoided, requiring a cautious interpretation of the magnitude and type of results. In most situations OTC use should be combined with verification experiments at different scales.
5.4 B i o m a s s formation and yield
Growth o f w i n t e r wheat in winter The hypothesis was tested that the higher photosynthesis at double-present C02 concentrations will result in enhanced growth in autumn, or in higher accumulation of reserve carbohydrates in the plant material. Both from detailed carbohydrate analysis and from biomass determinations it appeared that during the winter period no CO2-enhancement of growth occurred. Apparently temperatures from late October till early April are prohibitive. This was corroborated by the results with winter rye grown in the Rhizolab. The rye was sown late August, and grown till April. In the winter period no significant differences in growth or rate of photosynthesis could be observed. However, a stepwise increase in temperature in February resulted in a 21% higher biomass in April. In parallel experiments with grasses it appeared t h a t also there no CO2-enhancement of growth could be detected, and only in May a clear growth stimulation was found. From switching experiments (between CO2-1evels) it appeared that the higher accumulation of carbohydrates during winter did not result in a clear advantage.
Differences in response between species Comparing ambient and double-present CO2 concentrations with wheat, potato and faba bean (Vicia faba) it appeared that the increase in biomass productivity between species and years varied from 16 to 55% (Table 5.1). In general faba bean showed the highest effect, and potato and wheat reacted about equal. Growth enhancements for OTC-grown potato plants were low, and rather variable. In some years no statistically significant differences could be shown.
699 An analysis of the data using simulation models showed that variations in the weather pattern, but also effects related to light and temperature conditions in the period following emergence could explain a significant part of the year to year variability and also of between-species differences. Model performance for potato crops was rather poor. Although models could not explain all results, they clearly helped to show the great importance of the timing of local weather (temperature levels, drought) relative to the development stage of the crop. The generally observed reduction of nitrogen concentrations in the plant biomass was confirmed in our experiments. Therefore the nitrogen harvest (per land area) increased less than the biomass did. Water use
efficiency
In the experimental conditions (optimal water and nutrients) water use efficiency (biomass produced per unit water transpired) was enhanced, but total seasonal water use (per square meter) did not significantly change because of a larger (transpiring) biomass. More detailed analysis showed that nitrogen content in the biomass dropped, but that the harvest index did not change. Full details on the growth of roots and production of CO2 by respiration of the roots and soil throughout the season can not yet be given. Table 5.1A Biomass and yield (g.m-2 dry matter) of crops grown in near field conditions in Open Top Chambers, exposed to ambient and double presen [CO2] in 1994 Crop Winter Wheat total biomass grain yield Potato total biomass yield (dm) tuber fresh wht F a b a bean total biomass seed yield
ambient [CO2]
double [CO2]
statist. signif,
ratio
2302 1009
2709 1206
p
1.18 1.20
1202 1036 5011
1499 1284 5725
p<0.001 p<0.05 p<0.1
1.25 1.24 1.14
1116 614
1641 733
p<0.001 p<0.05
1.47 1.20
700
Table 5.1B [C02] response of biomass production for various years and crops as compared with simulated yield response experiental year ratio spring wheat winter wheat faba bean
1991 1993 1992
simulated year ratio
1.35 1.19 1.58
1975-1987 1993 1975-1987
1.42 1.35 1.52
Canopy gas exchange For the crops grown in the Rhizolab facility gas exchange has been followed throughout the season, allowing the detailed analysis of various interactions with environmental factors. At the canopy level the often reported down-regulation of photosynthesis after prolonged exposure to elevated CO2 was not detected. The growth enhancement due to CO2 did not or very little depend on the development stage of the crop. An interaction with light level or temperature during the growing season could not be detected (Figure 5.1). The absence of a growth enhancement at low (winter) t e m p e r a t u r e s could not be proved, as the photosynthetically active biomass (winter wheat) was too small at that stage. More detailed analysis of data from grass swards and winter rye canopies may shed some light on this point. CCER (750) 100 A o
/
8o
o
o Io.,"=~"
60
~
o
--:
4O
S
0
(
i
0
....
i
,
20
i
40
,
i
. . . . .
60
CCER (350)
Figure 5.1 Scatter plot of canopy photosynthesis of spring wheat and the season at 350 and 700 ppm C02
faba bean throughout
701 6.
INTEGRATION OF EFFECTS TERRESTRIAL ECOSYSTEMS
OF
CLIMATE
CHANGE
ON
6.1 I n t r o d u c t i o n Changes in soil acidity and nitrogen supply are held responsible for the strong decline in species diversity in Europe over the past 40 years. Succession of vegetation types is intimately related to changes in nitrogen mineralization, soil acidity and plant biomass production levels, that are a result of soil organic m a t t e r accumulation. It is to be expected t h a t changes in CO2 concentration in the a t m o s p h e r e will have i m p o r t a n t effects on the rate of soil organic m a t t e r accumulation and nitrogen turn-over. To study the effects on succession in ecosystems, a simulation model is being developed that will help to evaluate effects of changes in [CO2] and temperature on the succession and biological diversity in nutrient poor ecosystems.
Data for the calibration and validation of the model are collected in an existing chronosequence (0 to 120 years) on a nutrient-poor sandy soil. 6.2 T h e m o d e l The model simulates carbon and nitrogen flow through plant biomass and soil. It consists of three main modules: a water balance, a soil module, and a plant module. Decomposition and mineralization are calculated in the soil module. In the model species compete for light, water and nitrogen.
The plant model is still under development. Because species will show different responses to climate change because of specific plant traits, "competition m a y change and lead to a modified development course for the ecosystem. 6.3 P r o g r e s s a n d p l a n s The project was started only in late 1993. Although progress is along expected lines, as yet no results are available. Model outcome will be related to biodiversity using the correlation between species diversity and variables like soil moisture content, nitrogen content, pH, vegetation height and vegetation structure. The duration of the development stages of the succession series of the ecosystem will be used as an indication for the potential for maintaining biodiversity during the succession.
P a r a m e t e r values are obtained from field experiments in five stages in the chronosequence. Among others biomass quantities and decomposition rates are measured for different sources of organic matter. 7.
IMPACT ON VECTORS
ECOLOGY
AND
DISTRIBUTION
OF
MALARIA
7.1 I n t r o d u c t i o n This project was started as a pilot study for vector borne diseases. Specifically it aimed at evaluating the effects of climate change on the complex ecology of
702 mosquitoes and, from there, estimate the risk of a future spread of diseases. In itself it is not directly linked to the other terrestrial ecosystem studies. The main goal of the project was to use detailed information on the habitat and bionics of the relevant Anopheles species to develop a simulation model for the population dynamics of malaria vectors in Europe, allowing the estimation of potential changes in location and size of areas with a health risk. 7.2 S i m u l a t i o n m o d e l s
The model MOSQSIM, describing the population dynamics of vector species, is linked to a model simulating water availability including changed patterns in precipitation. The consequent changes in habitat (moisture, temperature) favouring or hampering the successive stages of mosquito development are calculated. Critical conditions for population development like availability of fresh water pools, and of humans or other blood sources for the blood meal prior to hatching are also quantified in the model. Cur.
Cli,
>1E+8 m >IE+7 [] >1E+6
~
>lE+5 [E >lE+4 []
'
'
'~
'
i
. . . . .
,
'
'
'1
'
'
|
'
i
'
'
U |
'
'
'
............
yl
>1E+3 I~1 >IE+Z ffl >1E+1 [ ] >IE+9 V]
v1ND
El
Figure 7.1 Simulated distribution of Plasmodium vivax infectious mosquitoes under the current temperature (+ 2 ~ climate scenario
703 C.C. >IE+8 >IE+7 >IE+6 >IE+5 >IE+4 >IE+3 >IE+Z >IE+I >IE+8
+ZK.
.
.
.
.
LA'
'
9
[]
I'[]
[~]
FU C]
EI
[
Figure 7.2 Simulated distribution of Plasmodium vivax infectious mosquitoes under increased temperature (+ 2 ~ climate scenario 7.3 D i s t r i b u t i o n of t h e m a l a r i a v e c t o r On basis of a climatological database the model correctly generates the vector distribution and malaria incidence in Europe before World War II. Also, realistic values for the vector distribution were found for the last decade. W h e n applied to climate change scenarios the validated model predicts a significant change in the vector distribution. Not only present mosquito species will extend in distribution area, but also more dangerous species will move to higher latitudes over Europe. A s u m m e r t e m p e r a t u r e increase of 2 ~ may result in a h u n d r e d fold or more increase in the number of infected mosquitoes in Southern Europe. The epidemic potential of vectors already present in an area thus may rise drastically. The techniques developed are suitable for estimating the regionalised malaria risk. As an index the vectorial capacity has been defined, being the n u m b e r of potentially infective contacts made by a mosquito population per infectious person per day. This index is a powerful tool, as it may also be used to express the regionalised malaria risk at a world-wide scale. 7.4 C o n c l u d i n g r e m a r k s In the project models have been developed for the distribution of mosquitoes as a vector of malaria. The models from this pilot study can be adapted to assess the impact of climate change on other harmful as well as beneficial insects, provided the proper technical coefficients describing critical stages and processes are
704 available. Evaluation of changes in biodiversity as well as questions related to integrated pest management in crop protection could be studied in this way. 8.
CLOSING REMARKS
The projects summarised above have not yet been completed at this stage. The project on integration of effects in fact has only started a year ago. As a consequence an assessment can only be preliminary. Further interactions and progress in the analysis and interpretation of the results will help to complete the contribution to the scientific basis for the estimation of the impact of global climate change on terrestrial ecosystems. Although the subjects and objects under study were different in many respects, the knowledge developed under the programme has made a significant contribution to improve the understanding of the complex terrestrial ecosystems, and has been instrumental for the development of new tools. It has also helped to establish excellent networks with the international scientific community, and to provide policy makers both nationally and internationally with the latest results. However, the full value of the effort will only be achieved in a follow-up of the programme, where missing information can be added, but certainly a transfer and further integration of knowledge based on the expertise built up in the NRP, and in International Programmes (like IGBP-GCTE, IGBP-BAHC, HDP-LUCC) should take place. 9.
REFERENCES
Arp, W. and F. Berendse, 1993. Plant growth and nutrient cycling in nutrient-poor ecosystems. In: Van de Geijn, S.C., J. Goudriaan and F. Berendse, (eds.). Climate Change: crops and terrestrial ecosystems Agrobiologische Thema's 9: 109-123. Berendse, F., 1993. Ecosystem stability, competition and nutrient cycling. In: Schulze, E.D. and Mooney, H.A. (eds.). Biodiversity and Ecosystem Function Ecological Studies 99: 409-431. Springer Verlag, Heidelberg. Bouten, W. and J. Goudriaan, 1995. Effects of CO2 fertilisation on evapotranspiration. In: Oliver H.R. (ed.). The role of water and the hydrological cycle in global change. ASI NATO. (In press). Dijkstra, P., A.H.C.M. Schapendonk and S.C.van de Geijn, 1994. Response of spring wheat canopy photosynthesis to CO2 concentration throughout the growing season: effect of development stage and light intensity. In: Veroustraete, F. and R. Ceulemans (eds). Proceedings of the Workshop "Modelling and climate change". SPB Academic Publishing bv, The Hague, p. 53-62. Fisher, M.J., I.M. Rao, M.A. Ayarza, C.E. Lascano, J.I. Sanz, R.J. Thomas and R.R. Vera, 1994. Carbon storage by introduced deep-rooted grasses in the South American savannas. Nature 371: 236-238.
705 Gorissen, A, P.J. Kuikman and H .van de Beek, 1995. Carbon allocation and water use in juvenile Douglas fir under elevated CO2. New Phytol. 129: 275-282. Gorissen, A., J.H. van Ginkel, J.J.B. Keurentjes and J.A. van Veen, 1995. Grass root decomposition is retarded when grass has been grown under elevated CO2. Soil Biol. Biochem. 27: 117-120. Goudriaan, J., 1993a. Interaction of ocean and biosphere in their transient responses to increasing atmospheric CO2. Vegetatio 104/105: 329-337. Goudriaan, J. 1993b. Carbon cycle and ecosystem productivity on a global scale. In: Van de Geijn, S.C., J.Goudriaan and F. Berendse (eds). Climate change; crops and terrestrial ecosystems. Agrobiologische thema's no 9. CABO-DLO, Wageningen, p. 125-138. IGBP, 1993. Biospheric Aspects of the Hydrological Cycle: The operational Plan, IGBP-report No. 27. Jetten, T.H., and W. Takken, 1994. Anophelism without malaria in Europe. A review of the ecology and distribution of the genus Anopheles in Europe. Wageningen Agricultural Papers 94-5, 69 pp. Kramer K., 1995. Phenotypic plasticity of the phenology of seven European tree species in relation to climatic warming. Plant Cell and Environment 18: 93-104. Kramer K., 1995. Modelling comparison to evaluate the importance of phenology for the effects of climate change on growth of temperate-zone deciduous trees. Climate Research, in press. Kramer K., 1994. A modelling analysis of the effects of climatic warming on the probability of spring frost damage to tree species in The Netherlands and Germany. Plant Cell and Environment 17: 367-377. Lankreijer, H.J.M., The effects of an increase in CO2 on the hydrology of forests. Thesis in prep., Department of Physical Geography, University of Groningen. Lankreijer, H.J.M., 1994. The effects of an increase in CO2 on the hydrology of forests. Interim report, Department of Physical Geography, University of Groningen. M.G. Schaap, 1995. The role of soil organic matter in forest hydrology. Thesis: Landscape and environmental research group, University of Amsterdam. (In prep.) Martens, W.J.M., L.W. Niessen, J. Rotmans, T.H. Jetten and A.J. McMichael, 1995. Potential impact of global climate change on m a l a r i a risk. Environmental health perspectives 103: 458-464. Olsthoorn, A.F.M., 1991. Fine root biomass of two Douglas-fir stands on sandy soils in the Netherlands. 1. Root biomass in early summer. Neth.J.Agric.Sci. 39: 49-60. Olsthoorn, A.F.M. and A. Tiktak, 1991. Fine root biomass of two Douglas-fir stands on sandy soils in the Netherlands. 2. Periodicity of fine root growth and estimation of below-ground carbon allocation. Neth.J.Agric.Sci. 39: 61-77. Reilly, J., S.C. Van de Geijn, R. Rogasik, C. Rosenzweig, and S.K. Sinha, (Lead Authors) 1995. (In review). Climate change and agriculture: Contribution of IPCC WGII Subgroup Agriculture II.B.5. 1994 to the IPCC Scientific Assessment Report 1995. Steffen, W., B.H. Walker, J.S.I. Ingram and G.W. Koch (eds), 1992. Global Change and Terrestrial Ecosystems: The Operational Plan. IGBP-report No. 21.
706 Tosserams, M., A.J. Visser, M. Groen, E. Magendans and J. Rozema, 1995. The combined effect of CO2 and solar UV-B radiation on faba bean. II. Effect on physiology.Submitted to Plant Cell & Environment. Turner II, B.L., R.M. Moss and D.L. Scole (eds), 1993. Relating Land Use and Global Land-Cover Change. IGBP-report 24 (published jointly with HDP). Van de Geijn, S.C., and J. Goudriaan, 1994. (In press). The effects of elevated CO2 and temperature change on transpiration and crop water use. Proc. FAO Expert Consultation, Rome, 10-14 December 1993. Van de Geijn, S.C., P. Dijkstra, J. van Kleef, K.Groenwold and J.Goudriaan. 1993. An experimental facility to study effects of CO2 enrichment on the daily and long-term carbon exchange of a crop/soil system. In: Schultze, E.D. and H.A. Mooney (eds). Design and execution of experiments on CO2 enrichment. Ecosystems reports 6, Commission of the European Communities, p. 167-174. Van de Geijn, S.C., J. Vos, J. Groenwold, J. Goudriaan and P.A. Leffelaar, 1994. The Wageningen Rhizolab - a facility to study soil-root-shoot-atmosphere interactions in crops. Plant Soil 161: 275-287. Visser, A.J., M. Tosserams, M. Groen, E. Magendans and J. Rozema, 1995. The combined effect of CO2 and solar UV-B radiation on faba b e a n . I.Growth and allocation. Plant Cell & Environment (submitted). Visser, A.J., H. Bleeksma and J. Rozema, 1995.Growth and physiology of spring wheat and faba bean. Effects of differing rooting volumes. Submitted to New Phytol.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
709
CLIMATE C H A N G E AND A G R I C U L T U R A L P R O D U C T I V I T Y A.H.C.M. Schapendonka, P. Dijkstraa, M.J.H. J a n s e n a , C.S. Pota, S.C. van de Geijna, A. Visserb and J. Rozemab a Research Institute for Agrobiology and Soil Fertility (AB-DLO), P.O. Box 14, 6700 AA Wageningen, The Netherlands b D e p a r t m e n t of Ecology and Ecotoxicology, Free University of Amsterdam, de Boelelaan 1087, 1081 HV Amsterdam, The Netherlands 1. A B S T R A C T Research performed at the AB-DLO and the Free University was intended to quantify the effects of a doubled CO2 concentration on some key agronomic species and grasslands. A set of physiological and morphological processes was studied and related to above- and below ground cycling of carbon. The research was based on experiments and simulation studies at the level of plant, crop and soil in laboratory facilities and semi-field conditions with controlled CO2 supply. A g r i c u l t u r a l crops were grown in "Open Top C h a m b e r s " or greenhouses and grasses in transparent tunnels made of Lexan. Soil processes and root respiration were studied in the Wageningen Rhizolab. Photosynthesis and assimilate partitioning were m e a s u r e d simultaneously in the photosynthesis laboratory. S i m u l a t i o n models a p p e a r e d to be useful tools to quantify the consequences of elevated CO2 and climate change on the productivity of grasses and crops with different growth strategies.
2. R E S U L T S 2.1 P h o t o s y n t h e s i s and p r o d u c t i v i t y In general, a doubled CO2 concentration enhanced photosynthetic rates of leaves by 25-50%. This stimulated the development of leaves and therefore the amount of interecepted light increased. The combined effect of CO2 on photosynthesis and the increased light interception was an improvement of the yield by 30-60%. It was established t h a t the stimulation was not a temporary one. Results from experiments with s u m m e r wheat and faba bean, showed that the CO2 response of wheat remained equal over the season and for faba it even increased towards the end of the season. A remarkable observation was the decreased respiration of soil and roots for grasses under elevated CO2. The consequence of this might be a net positive effect on productivity. The r a n k i n g for the CO2 effect was: faba bean, spring wheat, grass, potato and winter wheat, in a decreasing order. The increase of the harvested products, pods, grain etc. was equal to the total dry m a t t e r increases, which m e a n s t h a t the
710 h a r v e s t index was unaltered by C02. The CO2-driven increase in production, however differed over seasons and between years. The positive effects on grasses and w i n t e r w h e a t were much smaller during w i n t e r t h a n d u r i n g s u m m e r . Differences between years and species were explained in terms of differences in timing on which developmental stages were reached until h a r v e s t [1,2]. Thus environment interacted differently for the respective crops.
2.2 Simulation models Simulation studies were of great help to resolve these interactions. A simulation model for potato production predicted a significant interaction between the CO2 effect and a specific cultivar property, called earliness. In comparison with an early cultivar a late cultivar is characterised by later tuber filling and later leaf senescence. A v e r a g e d over the y e a r s 1988-1991, a s i m u l a t e d i n c r e a s e of the CO2 concentration from 350 vpm to 700 vpm increased tuber dry m a t t e r production by 22% for late cultivars and 29% for early cultivars. The effects were smaller for late cultivars, irrespective of the occurrence of a drought period. The higher benefit for early cultivars compared with late cultivars is due to their relatively low Leaf Area Index (LAI) and the concomitant higher positive effect of CO2 on the l i g h t utilization efficiency [2]. This effect is enhanced by a relatively higher positive effect CO2 on of the Leaf Area Duration (LAD) for early cultivars. 2.3 R e g r o w t h from storage P a s t u r e s are regularly defoliated by grazing or mowing. As a consequence of the removal of photosynthetic tissue, the plants have to remobilize carbon from reserves t h a t were stored in the stubble. These reserves consists mainly of sugars t h a t were obtained from photosynthesis prior to the defoliation. It was found t h a t elevated CO2 increased the amount of reserves. Strong interactions exist between the a m o u n t of biomass harvested and the CO2 effect on the biomass produced in the next cut. There is an optimum for the cutting height to gain the largest effect for double CO2. In the first days after cutting, growth of grasses is mainly sink limited. The a m o u n t of stored carbohydrates in the stubble only set the lower limits for potential regrowth. After about one week of regrowth, the supply of assimilates for regrowth becomes limiting due to photosynthetic restrictions. Doubled CO2 h a d a strong positive effect both on tillering and the rate of refilling the storage pools to be used aider a next defoliation. 2.4 Effect of t e m p e r a t u r e Higher t e m p e r a t u r e s were synergistic with elevated CO2 on photosynthetic rates of leaves. Doubled CO2 stimulated photosynthesis of grasses and w i n t e r w h e a t much more in s u m m e r t h a n in winter, partly due to the temperature- and partly due to light interactions. This tentatively would lead to the conclusion t h a t elevated t e m p e r a t u r e would enlarge the beneficial effect of doubled CO2 on plant productivity. This however appeared not to be the case. Simulation models of grasses and potato, predicted a shorter growing period at higher temperatures due to an accelerated senescence of the leaves. The a d v a n t a g e of a f a s t e r leaf
711 appearance rate and leaf elongation rates in spring were adversely affected by a faster senescence of the leaves. In addition annual crops developed faster, which shortened the time between emergence and harvest and therefore decreased the amount of intercepted radiation and thus the amount of accumulated biomass. 2.5 N i t r o g e n i n t e r a c t i o n As an overal response, high C02 leads to a decrease of the nitrogen content in the biomass m a i n l y due to an increase in the a m o u n t of s t a r c h a n d o t h e r carbohydrates. This effect however is relatively small and in most cases it is of minor importance compared with the effect of other environmental p a r a m e t e r s (Figure 1).
3
N (g rn -2 )
A
2.5 2 1.5 1 0.5
Top
0 0
3
0.2
0.4 0.6 0.8 relative height
1
N (g m - 2 )
B
2.5 2 1.5 1 0.5
Top
0
,
0
0.2
|
=
0.4 0.6 0.8 relative height
,
1
Figure 1: Nitrogen concentration (g m-2 leaf area) of leaves of faba bean against relative plant height as affected by CO2 concentration (LCO2 squares, HCO2 triangles) at low (A) and high (B) plant density.
712 A shortage of nutrients reduces the positive effect of elevated C02. Because large areas in the world are characterised by low amounts of available nitrogen, the expected positive effects on plant production on a global scale have to be considered in this respect. We found t h a t part of the CO2 stimulation might be retained under low nitrogen because the plant seemed to require less nitrogen in the tissue. This was concluded from the findings t h a t the nitrogen contents of wheat and faba bean were reduced under doubled CO2 and optimal nitrogen supply. However, the total amount of nitrogen t h a t was taken up on a surface area basis was higher under double CO2. This can be explained by the bigger root system of those plants with which a larger volume of soil could be explored for nutrients. Obviously, this can only be the case when the soil volume t h a t is concerned contains any nitrogen. In other cases nitrogen becomes limiting even sooner. 2.6 UV-B e f f e c t s F a b a bean was shown to be very sensitive for I_W-B radiation (280-320 nm). C u r r e n t outdoor I_W-B levels already had an effect on the growth of this p l a n t , although UV-B levels in t h a t open top chamber experiment were very low. Both CO2 and UV-B had a significant effect on biomass at all three harvests [3]. The highest UV-B level decreased biomass with 44% in ambient CO2 and 55% in elevated CO2 at the final harvest (41 days). Without UV-B total biomass was s t i m u l a t e d 34% by elevated CO2. This stimulation was only 7% at the highest UV-B level. The 10 k J t r e a t m e n t already decreased total biomass at the final h a r v e s t of ambient grown plants whereas this t r e a t m e n t had no effect on total biomass of plants grown under double CO2. At the highest UV-B t r e a t m e n t total biomass of elevated grown plants was less than the biomass in the ambient grown plants in the 0,6 and 10 k J treatments. 2.7 U n c e r t a i n t i e s While plant growth is stimulated under elevated C02, not all species react in the same way. Interactions between developmental stages of different species and e n v i r o n m e n t a l factors are involved such as light, t e m p e r a t u r e and n u t r i e n t supply. The background of this is not known with sufficient certainty. This aspect is of major importance for plants which grow in competition with each other especially in natural vegetations. A project for further research is in preparation, involving simulation models and "Open Top" experiments. E l e v a t e d CO2 m a y lead to more carbon stored in the vegetation. T h r o u g h an increased root growth it may lead to an elevated carbon pool in the soil, too. The consequences for the total carbon balance may be considerable. The question remains w h e t h e r carbon or nitrogen is the controlling factor in this process. This will be the challenge and the subject of forthcoming projects. 3. G E N E R A L C O N C L U S I O N S 3.1 A b o v e g r o u n d We conclude that the response to doubling the CO 2 concentration varies to a great e x t e n t and t h a t this v a r i a t i o n can p a r t l y be explained by differences in
713 development of specific crop species during the season and interactions with weather conditions. Apart from the uncertainties given above, it is expected that a doubling of the CO2 concentration will result in an increase in production of 15-50 %. An increase in temperature will partly neutralise the effect of CO2 because it decreases the length of the period for growth of annual crops and it increased the turn-over of life leaves into dead litter for grasses. In order to achieve an increase in production, increased input of nutrients is required.
3.2 Below ground When nitrogen is not limiting, a doubling of the C O 2 concentration stimulates the allocation of carbon to the roots and the soil. This may increase the storage capacity of the soil for carbon. The ultimate effect for the carbon cycle will depend on the stimulation in the long term, the availability of nutrients, the effects of higher temperature and the decomposition rates of soil organic matter. References 1 Dijkstra, P., Schapendonk, A.H.C.M. and S.C.van de Geijn, S.C. 1994. Response of spring wheat canopy photosynthesis to CO2 concentration throughout the growing season: effect of developmental stage and light intensity. In: Vegetation, Modelling and Climatic Change Effects. (Eds. P. Veroustraete, R. Ceulemans, I. Impens & J. Van Rensbergen.) SPB Academic Publishing. The Hague, ISBN 90-5103-090-8. pp 53-62. 2 Schapendonk, A.H.C.M., Pot, C.S. and J. Goudriaan, (1995) Simulated Effects of Elevated Carbon Dioxide Concentration and Temperature on the Productivity of Potato. Interaction with Maturity Classes. In: Ecology and Modelling of Potato Crops under Conditions Limiting Growth. (Eds. A. Haverkort & D.K.L. MacKerron). Series Current Issues in Production Ecology. Kluwer Scientific Publishers, Dordrecht. pp 101-114. 3 Visser, A.J., Tosserams, M., Groen, M.W., Kalis, G., Kwant, R. and J. Rozema, 1994. Simultaneous application of CO2 and UV-B radiation (280-320 nm) to Vicia faba: I. CO2 effects on growth and physiology. Proceedings of the 9th Congress of the Federation of European Societies of Plant Physiology. 3-8 july 1994, in press.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
715
Integrating the effects of climate change on terrestrial ecosystems C.G.F. de Kovel a and Y.J.O Wilms a aDepartment of Terrestrial Ecology and Nature Conservation, Wageningen Agricultural University, Bornsesteeg 69, 6708 PD Wageningen, The Netherlands
Abstract
A model of carbon, nitrogen and water flows within an ecosystem in relation to changes in species composition is nearing completion. This model, which applies to succession on sandy soils, will be used to investigate the effects of temperature, nitrogen deposition and CO2 concentration in the atmosphere on vegetation succession. The model applies to primary succession on sandy soils. Parameters are being collected in a Dutch drift sand area (Hulshorsterzand). Preliminary simulation show strong interactions between the effects nitrogen deposition and CO2 concentration: The rate nitrogen is supplied to the ecosystem influences whether species replacement is accelerated, not affected or delayed by increasing CO2 concentrations. This implies that relative abundances of species may change in response to climate change.
1. INTRODUCTION One important component of vegetation succession is the modification of the environment brought about by the plants themselves. This modification encompasses changes in the availability of nutrients, light and water for the different plant species. These changes alter the competitive balance between species, which can lead to shifts in their relative abundances. The carbon, water and nutrient flows in ecosystems can be seriously affected by the various aspects of climate change: rising CO2, changing precipitation patterns and changing temperatures as well as by the changing amounts of nitrogen deposition. This can have pronounced consequences for species composition and biodiversity in natural ecosystems (Pastor & Post, 1986). Short term experiments with one or a few species are insufficient to enable the eventual effects of climate change on plant communities and biodiversity to be predicted with confidence. The solution is to integrate knowledge acquired in these experiments into models. We are therefore in the process of developing a dynamic simulation model to investigate the effects of climate change on succession, which calculates how dominant plant species in a primary succession mediate carbon and nitrogen flows. The final model furthermore will include a water balance. Species diversity is often found to be correlated with ecosystem parameters such as primary production, vegetation structure, nitrogen mineralization and soil pH (Pausas, 1994). Other variables might be identified as being relevant in this context too. These ecosystem parameters are mainly determined by the external environmental conditions, the successional
716 age of the ecosystem and the dominant plant species. The aim of the model presented is to calculate effect of climate change on the mentioned ecosystem parameters and on the biomass of the dominant plant species. Based on the changes in these variables probable consequences for biological diversity will computed using descriptive models.
2. M O D E L DESCRIPTION
The succession model simulates the dynamics of populations of dominant plant species at a certain site through time. In the model plant populations compete for light and nitrogen (Figure 1). A water module will enable competition for water to be investigated too. The model has been specially designed to study the effects of different scenarios of atmospheric nitrogen deposition, CO2 concentration and temperature on succession. The driving variables of the main model are temperature, CO2 concentration in the atmosphere and nitrogen deposition. The water module will use additional weather data, such as rainfall and irradiation as driving variables. The time step of the model is one year and the run time is about 150 years.
FI
mortality
r .....
E'~ 0
,~
:
::
,~,~,~,~
7~
~
ar~
i
" .....
..... . . . . . . . . . . . .
,~
:
~
I~~K
---
_l
,,::-..
d'~/l~l
I,,,
,~,
I! v
I
/.=--I
~ / ~. \
~'
oo
-q'~
~
..........
11
~'~--
--II
L . . . . . .
~
~'I~1
9 ~
!
~
I
.~ ~
.-i
"---"~L% ~'~
/~\
assil'nilation
.~ ! ,.
'~
:
! ~=.- In-',. . . .~. . .t. . . . i:,: E
~/L~
:
i
Fig. 1. General structure of a succession model simulating competition for light and nitrogen as influenced by cBmate change. SoBd Bnes: carbon /nitrogenflow, dotted Bnes: information flow, grey circles: driving variables.
The model computes at each time step the amounts of carbon and nitrogen in the plant parts of the different species. Furthermore, the amounts of carbon and nitrogen contained in litter and soil organic matter and the amount of mineral nitrogen in the soil are calculated. The latter information is useful because on sandy soils with high rainfall the amount of mineral nitrogen in the soil is a measure of the leaching of nitrogen. Vegetation height, light at soil surface and several other derived values can be produced by the model.
717 The model parameters are currently being measured in a chronosequence in a Dutch drift sand area (Hulshorsterzand, 52020 ' N, 5044 ' E). Validation data are being collected in independent measuring campaigns.
3. RESULTS TO DATE As the model is still being developed, the results reported here are preliminary. Parameters were tuned in such a way that a doubling of the CO2 concentration would under optimal conditions result in a 30% increase of production for all species. At intermediate levels of nitrogen input, simulations with a high CO2 concentration (of 560 ppm) showed lower N to C ratios in living tissue and in fresh litter. This in turn resulted in lower decomposition and mineralization rates (Figure 2). Due to reduced N availability, in these cases CO2 enrichment did not lead to increased production, compared to simulations with present-day CO2 concentrations (of 350 ppm). At low N input, little changed; at high N input simulations showed increased total production. Simulations with parameters for Calluna vulgaris (L.) Hull and Deschampsia flexuosa (L.) Trin. (from Bakema et al. 1994) showed that the relative abundances of species may be influenced by increased CO2 concentrations (Figure 3).
kg N ha 4 y e a r ~
kg C / h a 20 _
t
450
II ~,,, ,, CV +
/
350
/
15
350 ppm 250 /
/
/
{-
/
/
150
560 ppm..--"
J"
,/
t
,/ /
/
i
/
,
I
/
/..,.9
5
i Df
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
~., I - .
.
50 ." ~"..~ -50 25
I
I
I
50
75
100
,,,I 125
/ ~
CV
Df+ /
\
0 0
25
50
75
1 oo
year
Fig 2. N mineralization rates under a Calluna Deschampsia vegetation at two C 0 2 concentrations and 30 kg N / h a / y e a r as atmospheric deposition.
Fig 3. Biomass development o f C. vulgaris and D. flexuosa at two C 0 2 concentrations and 30 kg N ~ha~year as atmospheric deposition.
4. DISCUSSION AND CONCLUSIONS So far, we have only investigated the effects of nitrogen deposition and CO2 concentration on vegetation and soil processes. However, the water balance module has not yet been used, so simulations have not included competition for water. The effects of changes in temperature have not yet been studied either. The simulations to date have shown that enhanced CO2 concentration results in a lower nitrogen supply if one of the species is light-limited. This light-
125
718 limited species will produce lower quality litter than under present-day CO2 levels and this will slow down mineralization. Decreased nitrogen availability will limit the production of species that are nutrient-limited. Although the production of the light-limited species usually increased, in most simulations total production did not increase. In some cases increased CO2 concentration led to a shift in dominance if a light-limited species were suppressed by a nutrient-limited species. However, the light-limited species profited most from the increase in CO2 concentration. This effect is largely opposite to the effect of increasing N deposition. In heather ecosystems heavy atmospheric N deposition has been shown to benefit the nutrient-limited grasses more than the more light-limited heather species, which may lead to an accelerated expansion of grasses (Berendse & Elberse, 1990). Clearly, there are strong interactions between the effects of CO2 enrichment and nitrogen deposition. In a dry, nutrient-poor ecosystem, like that of our study site on drift sand, the impact of CO2 on plant water use efficiency may be a very important effect of climate change (Arp, 1991). This aspect of CO2 enrichment will have strong interactions with temperature. During succession nitrogen availability often increases, and concomitantly the competition for light becomes more important. Often early species in the succession suffer from light shortage more than later species. An increase in CO2 concentration might slow down succession in some stages by benefiting early species more than late species. However, an increase in total production might accelerate succession, for example canopy closure may be attained sooner. Our preliminary conclusions are that increasing CO2 concentration may seriously affect succession patterns in poor ecosystems, but probably to a lesser extent in extremely nutrientpoor ecosystems. The direction and magnitude of the changes will differ between ecosystems and successional stages. CO2 enrichment and nitrogen deposition operate antagonistically in some aspects, though both may increase production and both may accelerate succession. The relations between different ecosystem parameters and species richness vary between vegetation types and between successional stages. These relations will have to be established for actual situations. However, it seems probable that a faster disappearance of early successional stages will reduce total species richness on a landscape scale. REFERENCES Arp, W.J. (1991). Plant water relations in elevated atmospheric CO2. Vegetation of a North American Salt Marsh and Elevated Atmospheric Carbon Dioxide. Ph.D. thesis. Vrije Universiteit Amsterdam. Bakema, A.H., Meijers, R. Aerts, R. Berendse, F. & Heil, G.W. (1994). HEATHSOL: a Heathland Competition Model. RIVM, Bilthoven. Berendse, F. & Elberse, W.T. (1990). Competition and nutrient availability in heathland and grassland ecosystems. Perspectives on Plant Competition (Ed. Grace, J.B. & Tilman, D.) Academic Press, Inc., San Diego. Pastor, J. & Post, W.M. (1986). Influence of climate, soil moisture, and succession on forest carbon and nitrogen cycles. Biogeochemistry, 2, 3-27. Pausas, J.G. (1994). Species richness patterns in the understorey of Pyrenean Pinus sylvestris forest. Journal of Vegetation Science, 5, 517-524.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
719
Selection of bio-indicators to assess the possible landscape ecological effects of climate change Rudolf S. de Groot a, Pieter Ketner b and Arjan H. Ovaa a
aWageningen Institute for Environment & Climate Studies, Wageningen Agricultural University, P.O. Box 9101, NL-6700 HB Wageningen, The Netherlands
bDepartment of Terrestrial Ecology and Nature Conservation, Wageningen Agricultural University, P.O. Box 9101, NL-6700 HB Wageningen, The Netherlands
Abstract Past climate changes have led to considerable changes in the species composition of ecosystems. The recent increase in average global temperature is rather strong compared to previous warming periods and, if climate models are correct, future warming will be even stronger. Especially in Europe where the landscape has been greatly fragmented by human activities, the ongoing and projected changes in climate will pose an additional stress on the natural biodiversity. This paper will discuss a method for the selection of bio-indicators to assess the possible landscape-ecological effects of climate change, and presents some preliminary results of the selection of indicator species for the Netherlands.
1. INTRODUCTION
In this paper, some results are presented of a short pilot-study to find suitable bioindicators to assess effects of climate change on biodiversity in Europe. Since species determine the structure and functioning of ecosystems, these bio-indicators can be used to assess the climate-sensitivity of natural end semi-natural ecosystems and landscapes. The main purpose of the study is to develop criteria for the selection of species (bioindicators) that are suitable for assessing the (potential) response of natural ecosystems to climate change "in the field". The complexity of the potential effects of climate change on species as well as on landscape-ecological processes, together with the simultaneous occurrence of many other influences on the landscape (e.g. land use, eutrophication, acid precipitation) [1], makes it essential, but at the same time quite difficult, to select appropriate indicator-species.
720 2. S O M E CRITERIA FOR SELECTING INDICATOR SPECIES
In order to select potential indicator-species, they must be screened on a number of criteria based on the response features and practical constraints. As a start a few criteria are listed in table 1 , but future research will probably refine these further and may add more.
Table 1" Some criteria for selecting bio-indicators for climate change Climate sensitivity Habitat constraints Position within distribution range Presence and abundance Dispersal capacity Functional position in the ecosystem Suitability for monitoring
Depending on the aim for which indicator species are to be used, i.e the responsefeature to be monitored, which m a y be physiology, phenology, inter-specific interactions or abundance/distribution or a combination of these features, a selection of criteria should be used. These criteria could then serve as a type of "filter" to reduce the available species-list in a given region to a few suitable indicator-species.
J i
(1) Climate s e n s i t i v i t y
i
(2) Habitat c o n s t r a i n t s i
i
(3) Position in distr, range i
(4) Dispersal c a p a c i t y i ( ) etc.
'
I
I
i i
Bioindicator(s) Figure 1" "Filter" showing the succesive application of criteria to select indicator species for climate change
721 3. SELECTION OF POSSIBLE INDICATOR SPECIES FOR CLIMATE CHANGE IN THE NETHERLANDS The expected climate change in the Netherlands will probably result in winters that are more atlantic (higher temperatures and wetter), while the summer-period will be more continental (higher temperatures and drier). Changes in climate and weather will particularly influence plant and animal species which: (a) occur at the border of their geographical distribution, (b) do not occur in their preferemial habitat (marginal habitats), and (c) are able to adjust rapidly by changing their distribution area or which are just not able to do so and thus will be threatened by extinction. Various studies have been carried out in the Netherlands on the possible effects of global (esp. climate) change on ecosystems and species to various degrees of detail, including on terrestrial plants [3-5], aquatic plants [6], mammals [7-8], birds [9-10], butterflies [11-12] and dragonflies [13]. Most of these studies were carried out within the framework of a project initiated and sponsored by the Dutch Ministry of Housing, Physical Planning and Environment under the title Flora and Fauna 2030 [14] and the socalled LICC project [1]. Using the above mentioned publications and applying some of the criteria described in section 2 of this paper, a preliminary list of indicator species was prepared for herbaceous plants, butterflies, breeding birds and mammals. Only species that are considered as native or well established in the Netherlands have been used. Method of selection For the four groups, all species found in the Netherlands were used as a starting point: 1,253 higher plants, 71 butterflies, 179 breeding birds, and 49 mammals. The selection took place in four steps"
Step (1)" climate sensitivity In the first step, a survey based on expert judgements reveiled that 175 of the 1,253 selected plant species in the Netherlands are expected to respond mainly to climate change and much less to other global changes, whereby 165 plants are expected to benefit from climate-warming and 10 are expected to decline [4]. This list was narrowed down further by using the "Ellenberg-value" for temperature and only those species were selected which have a low value (1-3 = plants which indicate cold growing conditions) or high value (7-9 = plants which indicate warm habitats) [ 15-16]. For animals, such indicator values do not exist and here the climate-sensitivity of the species was qualitatively assessed on the basis of known ecological relations between the species performance and/or distribution and climate factor(s).
Step (2)" habitat constraints Only those species were selected for which adequate habitats can be found at all latitudes in The Netherlands.
722
Step (3): dispersal capacity During step 3, only those species which have a good dispersal capacity or of which (local) expansion is already taking place were selected.
Step (4): suitability for monitoring Finally, the number of suitable indicators was narrowed down further by leaving out those species which may be difficult to monitor, for example because they are too small to recognize, because they lead a "hidden life" or because of taxonomical difficulties or other reasons. Table 2 shows the results of this selection-procedure for the Netherlands for the four species-groups.
Table 2 Selection of bio-indicators for detecting climate change in the Netherlands
Number: Criterion (1) (*) (2) (3) (4)
Number:
Vascular plants 1253 100% + 165 31 18 18 13
14.0% 2.6% 1.4% 1.4% 1.4%
71
9 3 3 3
-
1 0 0 0
-
4
0
5.6%
2
0 0 0
2.8% 2.8% 2.8%
2 2
Mammals 49 100% +
5.6% 1.7% 1.7% 1.7%
*) based on Ellenberg-values (see text)
100%
+
Breeding birds 179 100% +
Criterion (1) (2) (3) (4)
10 1 0 0 0
Butterflies
0 0 0 0
-
0 0 0 0
0% 0% 0% 0%
+ = increase, - = decrease
Table 3 gives a preliminary list of species which could be used as climate indicator species in the Netherlands. All are species which are expected to increase in abundance and/or expand their distribution range. It should be emphasised that possible changes in phenology were not considered in the selection procedure.
723 Table 3 Examples of possible indicator species for climate change in the Netherlands
Vascular plants Ca species:
C__4species:
Anchusa officinalis (Alkanet) Hordeum murinum (Mouse barley) Medicago minima (Bar medick) Poa bulbosa (Bulbous meadow-grass) Sagina apetala Senecio vernalis Trifolium scabrum Trifolium striatum (Knotted Clover)
Abutilon theophrasti (Velvetleaf) Cynodon dactylon (Bermuda grass) Eragrostis minor (Little lovegrass) Eragrostis pilosa (Indian lovegrass) Portulaca oleracea (Purslane)
Butterflies Papilio machaon (Swallowtail) Nymphalis antiopa (Morning cloak)
Breeding birds Asio otus (Long-eared Owl) Cettia cettia (Cetti's Warbler) Cisticola juncidis (Fan-tailed Warbler)
When research and monitoring proceed, species may be added or deleted on the basis of difficulties encountered or based on new insights in the ecology of the species. No mammals were selected as bio-indicators, because most potential indicator-species in this group do not only react strongly to climate change, but also to other environmental changes. However, some species such as the Common Hamster (Cricetus cricetus) or Common Rabbit (Oryctolagus cuniculus) may be added to the list later on. 4. PRELIMINARY CONCLUSIONS A first application of the proposed method for the selection of bio-indicators showed that only 0-3 % of the species within the investigated taxonomic groups (vascular plants, breeding birds, butterflies and mammals) are potentially suitable as indicators for climate change in the Netherlands. All the selected indicator-species are expected to increase in abundance or expand their distribution range. It must be investigated further if this is a bias in the selection method or wether it is an indication of the possibly predominant positive effects of climate change for the investigated species groups in this part of Europe.
REFERENCES 1 M.M. Boer & R.S. de Groot (eds.), Landscape-ecolgical Impact of Climatic change, Proceedings of a European Conference, 429 pp. IOS Press, Amsterdam, Washington, Tokyo, 1990. 2 R.S. de Groot, Biodiversity as indicator for measuring sustainability: A case study on the use of indicator species. In: RMNO (Advisory Council for Nature and
724
3
4 5 6 7
8
9
10
11 12
13
14
15 16
Environment) Towards environmental performance indicators based on the notion of environmental space. 113-131 (1994) P. Ketner, P., Impact of climate change on flora and vegetation inWestern Europe with special emphasis on The Netherlands. In: J.I. Holten (ed), Effects of climate change on terrestrial ecosystems. Report from a seminar in Trondheim, 47-60 (1990). R. van der Meijden, Flora en Fauna 2030: Hogere planten, een geannoteerde soortenlijst. FLORON, 1993 F. Mohren & K. Kramer, Reactions of trees and forests to climate change. Change 9: 14-15, 1992 Th.C.M. Brock & W. van Viersen, Climatic change and hydrophytic-dominated communities in inland wetland ecosystems, Wetl. Ecol. and Man. 2 (1/2) 37-49, 1992 C.P.M. Zoon, Flora en Fauna 2030: Nederlandse zoogdieren in 2010: aantalsontwikkelingen tot 2010 van Nederlandse zoogieren (excl. vleermuizen) op grond van de gevoeligheid voor milieuthema's, 125 pp. Report Vereniging voor Zoogdierkunde en Zoogdierbescherming, Netherlands, report no. 8, 1993. V. Martens, Flora en Fauna 2030: Een studie naar het populatieverloop van vleermuizen in Nederland aan de hand van aantalsontwikkelingen in winterverblijven over de periode 1943-1992. Vleermuizen, 54 pp. Stichting Vleermuiswerkgroep Nederland, 1993 H.T.H.M. Meekes & S.C.V. Geelhoed, Climatic change: The possible consequences for migratory birds. A case study contribution to the LICC-Conference, December 1989. Lunteren, The Netherlands. Nature Conservation Department, Agricultural University Wageningen, 1989. H.P. van der Jeugd, Flora en Fauna 2030: Toekomstperspectief van de Nederlandse Broedvogels. Een extrapolatie van aantallen naar het jaar 2010, 74 pp. SOVON, Netherlands, 1993. Vlinderstichting, Flora en Fauna 2030, Achtergrondreeks, deel 4: Dagvlinders, 38 pp., Vlinderstichting, Wageningen, 1993 C.A.M. van Swaay, The Changing Climate: possible effects on butterflies.A case study contribution to the LICC-Conference, 2-7 December 1989. Lunteren, The Netherlands. Nature Conservation Department, Agricultural University Wageningen, 1991. M.T. Wasscher &J. van Tol, Flora en Fauna 2030: Veranderingen in het voorkomen van libellen (Odonata) in relatie tot geselecteerde milieuparameters, 40 pp. Stichting European Invertebrate Survey, Leiden, Netherlands, 1993 A.H. Ovaa, J. Latour & R. Reiling, Flora and Fauna 2030, Hoofdrapport: Lange termijn effecten van milieubeleid op flora en fauna: schattingen door middel van expert judgement, 31 pp. Wageningen Agricultural University & National Institute of Public Health and Environmental Protection, 1993. H. Ellenberg, Zeigerwerte der Gef~isspflanzen Mitteleuropas, 2nd edition. 122 pp. Scripta Geobotanica IX, G6ttingen, 1979. H. Ellenberg, H.E. Weber, R. Drill, V. Wirth, W. Werner & D. Paulissen Zeigerwerte von Pflanzen in Mitteleuropa, Scripta Geobotanica XVIII, 258 pp., Goltze Verlag, G6ttingen, 1992
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
725
Phenological reactions of D u t c h tree species to climate change described by a simulation model of the annual cycle
K. Kramer and G.M.J. Mohren
Institute for Forestry and Nature Research (IBN-DLO), P.O. Box 23, 6700 AA Wageningen, The Netherlands
Introduction
The relationships between climate and both phenology and growth of Dutch tree species were studied to evaluate the potential impacts of climate change on trees and forests in the Netherlands. In order to make such assessments, insight is required on the mechanisms how climatic variables interact with plant processes. The topics addressed in this study are: (1) the modelling of phenology; (2) the consequences of climate change on spring frost damage; (3) the importance of phenotypic plasticity; and (4) the impacts of climate change on growth.
1. MODELLING PHENOLOGY To evaluate the impacts of climate change on growth of temperate deciduous tree species, the onset and cessation of the growth must be accurately described. It was concluded that the timing of leaf unfolding could best be described by a model in which the effects of chilling temperatures (-5 to + 10~ and forcing temperatures (>0~ operate sequentially in time, according to a triangular and logistic function, respectively (Kramer 1994a).
2. SPRING FROST DAMAGE The effects of climatic warming to the probability of spring frost damage of Larix decidua, Betula pubescens, Tilia platyphylla, Fagus sylvatica, Tilia cordata, Quercus rubra, Quercus robur, Fraxinus excelsior, Quercus petraea, Picea abies and Pinus sylvestris in the Netherlands and in Germany were studied. It was concluded that for these species the probability of spring frost damage will decrease, provided the variability in temperature does not change (Figure 1, Kramer 1994b).
726
P(T< O ~ 0.6-
I
l e sequential 1 o alternating ]
0.4-
0.2-
I
-2
I
I
0
I
I
*2
1
U
*4
/~ T (~ Figure 1. Shift of the probability of sub-zero temperatures around the date of leaf unfolding of Fagus sylvatica on uniform temperature increase according to the sequential and alternating model.
3. PLASTICITY
To evaluate the potential response of individual trees to climatic warming, phenological observations of clones of Larix decidua, Betula pubescens, Tilia cordata, Populus canescens, Quercus robur, Fagus sylvatica and Picea abies transferred over a large latitudinal range in Europe were analyzed (Figure 2). It was found that these tree species possess a considerable plasticity and are able to respond phenotypically to a major change in their local climate. For the clones of Larix decidua and Quercus robur the growing season may shorten with increasing temperature, because leaf fall is advanced more than leaf unfolding. In Betula pubescens and Populus canescens, leaf unfolding and leaf fall are advanced equally, whereas in Tilia cordata and Fagus sylvatica the date of leaf fall seems to be unaltered but leaf unfolding advances with increasing temperature (Table 1). These differences in the duration of the growing season in response to increasing temperature may alter the competitive balance between the species in mixed stands (Kramer, in press).
727 Table 1 Shift of leaf unfolding (U) on mean winter temperature (Tw) and leaf fall (F) on mean summer temperature (Ts)
sty/sr,,
sF/sr=
9
d: ~J
S
s"
t/~
9
.j"
Larix decidua Betula pubescens Tilia cordata Populus canescens Quercus robur Fagus sylvatica Picea abies
-2.8 -3.7 -2.8 -3.0 -1.7 -2.4 -3.5
-8.5 -3.0 -1.4 -3.8 -5.6 0.0
~,_.~
~ .i'
_
.~_.o
<~ ~'
._ ~ - % - - ~ ~ " ~i~.ill
~.,_.......~...-~ 9
9 Oo oo llll-
9
9 ollll
"~-
9
9
. ~ -,,,O
....~'--------------~' '~
~ /
. . . ~ . ~ ',
~
.~_,
9
~
,"
~ ~0
;'
9 9
L"\ ~ .';~~II
9
9
9 ,~
"-', ~ ~ ~,~,
, .>~.'-..
<~
('/f
Figure 2. Location of the International Phenological Gardens.
4. GROWTH
The direct effects of increased atmospheric CO2, together with indirect effects,/.e. water relations, on growth of Picea abies were quantified, using a process-based forest growth model (FORGRO). The results indicated an increased response on growth of Picea abies to increasing CO2 under conditions of water shortage compared to the response in a potential growth situation. This effect is enhanced when the CO 2 increase is combined with a temperature rise possibly associated with the CO 2 induced climate change. These results were qualitatively confirmed by a similar study on Betula pubescens, Fagus sylvatica and Quercus robur, using several mechanistic models of photosynthesis and allocation with different levels of detail, and four climate change scenarios. However, quantitatively large differences were found between both the species, the scenarios and the models. The importance of phenology on the effects of climate change on growth of deciduous temperate-zone tree species was evaluated by a model comparison. The photosynthesis models used diverge in their response of annual gross photosynthesis the CO 2 concentration doubles, and the temperature increases by 2~ to 7~ they agree of differences caused by phenology ranging from 5% if the temperature rises 2~ to 20% if it rises 7~ (Figure 3). The corresponding figures are 5% and 13% when a mechanistic model for allocation of assimilates over the plant organs is used.
728
FORGRO
2.0
OSUM
GISS GFDL
1.5-
UKMO
1.0-
0.5-
0.0
I
I
I
I
I
I
I
I
0
1
2
3
4
5
6
7
8T (~ rq Betula
A Fagus
9 Quercus
Figure 3. Change in NPP relative to the current climate, given a 2x [CO2] and a range of temperature increases 5. REFERENCES
Kramer K., 1994a. Selecting a model to predict the onset of growth of Fagus sylvatica. Journal of Applied Ecology 31: 172-181. Kramer K., 1994b. A modelling analysis of the effects of climatic warming on the probability of spring frost damage to tree species in The Netherlands and Germany. Plant, Cell and Environment 17: 367-377. Mohren G.M.J., 1994. Modelling Norway spruce growth in relation to site conditions and atmospheric CO 2. In: Vegetation, Modelling and Climate Change Effects, F. Veroustraete & R. Ceulemans (eds.) SPB Academic Publishing bv. The Hague, The Netherlands. Kramer K., (in press). Phenotypic plasticity of seven European tree species, in relation to climatic warming. Plant, Cell and Environment. Kramer K. & Mohren G.M.J., (submitted). Sensitivity of FORGRO to climatic change scenarios. A case study on Betula pubescens, Fagus sylvatica and Quercus robur.
Climatic Change. Kramer K., (submitted). A modelling comparison to evaluate the importance of phenology on the effects of climate change on growth of temperate-zone deciduous trees. Climate Research.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
729
The Role of Organic Soil Profiles on Water Availability in Forests: Sensitivity Analyses M.G. Schaap, W. Bouten and L.C. Kuiper Landscape and Environmental Research Group, University of Amsterdam, Nieuwe Prinsengracht 130, 1018VZ Amsterdam, The Netherlands
Abstract This research concerns the soil organic matter linked availability of water in forests on dry, sandy soils. Results show that changes in soil organic matter (SOM) storage influences soil water retention, especially at low SOM contents. Further the organic forest floor allows a significant evaporation flux from the forest floor (yearly max. 30% of the evapotranspiration). Sensitivity analyses show that forest transpiration is only indirectly affected by SOM. A change in water retention does hardly change the transpiration but can alter the strategies of water uptake by roots. Global changes can thus indirectly affect the vitality of forests although the effects may vary with soil type.
1. Introduction Forest growth and development is related to soil organic matter (SOM) dynamics and water supply. In order to estimate the sensitivity of forest development to climatic change, it is necessary to have the relations in figure 1 quantified. As a part of this research effort, we study organic matter linked water availability in forests on sandy soils. We presume that these forests are sensitive to climate change because 1) changing meteorological conditions may alter the supply of water and the transpiration demand, and 2) altered primary production and decomposition may change the accumulation of soil organic matter in the mineral soil and organic debris on the forest floor, both changing the water availability in the soil profile.
CLIMATIC
f
Forest
Development
~ W'~ater
Soi~l
org. matter
"~
~
availabilit!
CHANGE Figure 1. The relations between forest development, soil organic matter and water availability. Our project is represented by a black arrow.
730 In this project we focused only on the effects of different SOM levels on soil water availability. A high water availability leads to optimal transpiration and carbon assimilation. Soil organic matter influences forest hydrology in three ways"
A~
Changes in hydraulic functions of the mineral soil Changes in the amount of dead organic debris on the forest floor Adaption of the root system
B.
C.
Input nodes
Hidden nodes
1..J
Output nodes
1..K Wik
1..L
_
Wkl
x, Y2
X2--~
_
-
1.0
.........
Bias
Figure 2. General structure of the neural network as used in this project. Model input is represented with X, model output with Y. Using selected examples an optimizing algorithm takes care that the neural network 'learns' to predict Y from X without the need for an a-priory model.
2. Results We used a neural network to link soil water characteristics to organic matter contents, the density and textural properties of the soil. The new technique of neural networks allowed us to establish a more reliable model for the retention characteristic than with other methods. Retention characteristics were predicted for two soils of existing Douglas fir sites in the central part of the Netherlands. Both soils mainly differ in soil texture, resulting in a lower water retention in site 1 than in site 2 (figure 3). The organic matter contents were changed from 0 to 200 % of the present level. Water availability appears to increase with SOM, especially at low content. Figure 3 shows that the
731 difference between field capacity (0Fc) or the point of water uptake reduction (0R~D) and the wilting point (0w) is higher on site 2. The forest floor (FF) can be significant in water supply to trees because it can store large amounts of water (up to 20 mm) which can either be discharged to the root system or evaporated. With a two year monitoring program of FF water contents and evaporation we estimated a FF evaporation of 85 mm per year in a dense Douglas fir forest (site 2). Higher amounts are expected for forests with a lower canopy density. Root development is linked to soil organic matter because of the supply of water and nutrients and because of the mechanical limitations for root penetration at low porosities which only occur at very low SOM.
W a t e r retention c u r v e s site 1
site 2
4.0
4.0 0% SOM
\ \
100% SOM 3.0
|
,--m
1i
3.0 200% SOM
c-
,2.'. o m
2.0
,..._,
9
2.0
E)RED
u_
|
o.
1.0
1.0
0,0 0.00
0.10
0.20
0.30
Water content [cma/cm 3]
0.40
0.0 0.00
0.10
0.20
0.30
0.40
Water content [cm3/cm 31
Figure 3A and B. Modeled water retention characteristics for site 1 and 2 as well as hydrologically relevant water contents. 0Fc: field capacity water content after free drainage in winter, 0RED point of reduced water uptake, 0w: wilting point.
3. Sensitivity analyses It is difficult to predict a-priory what the integrated effects SOM on the water availability will be because root water uptake is a process with many feedbacks. Answers can only be obtained with dynamic models calculating forest transpiration from altered soil properties and meteorological variables. We used the FORest HYDrological package (FORHYD) to perform sensitivity analyses of the transpiration of two Douglas fir forests
732 while varying the amount of soil organic matter, using the meteorological conditions of the past 30 years. Water availability was evaluated as the transpiration reduction (RED), calculated from the potential Makkink-transpiration (PT) and the actual transpiration (AT):
RED
PT :
- AT
I00
PT
Figures 4A and B show the transpiration reduction as influenced by hydraulic properties which change with organic matter content. To produce these figures, only one property was varied at the time while the others were kept constant. It appears that the point of root water uptake reduction is the dominating factor. Further, the transpiration reduction increases with lower SOM. An increase of SOM above the present level hardly affects the reduction. The impact of SOM is greatest in site 2 which has a higher water availability. This is confirmed in the three dimensional diagrams which show the transpiration reduction when SOM is varied simultaneously with forest floor thickness or with root depth. A decrease in SOM increases the reduction more in site 2. This can be explained by the difference in effective storage: in figure 3A and B the mineral soil water content difference between field capacity and the wilting point is far greater in site 2 than in site 1. Therefore, the thickness of the forest floor is more important in site 1 than site 2 (fig. 5A and 5B) because this layer can supply a part of the desired water uptake. For the same reason the transpiration reduction decreases much more in site 1 than in site 2 if the rooting depth is varied (fig. 5C and 5D). A deeper extending root system allows more water uptake from the mineral soil. Due to the better water retention, site 2 has already sufficient water storage with a limited rooting depth, an increase of roots beyond 50 cm decreases the reduction less than site 1. Whether trees can extend roots to deeper soil layers depends mainly on penetration resistance or soil chemical limitations. The penetration resistance is determined by bulk density and organic matter content and to a lesser extent the soil texture. In fact, the rooting depth in site 1 was only 30 cm due to mechanical resistance and 200 cm in site 2, which leads to large differences in present transpiration reduction between the sites.
4. Conclusions Our research lead to quantification of the relation ships between soil water retention characteristics, soil organic matter content and soil texture. Soil texture was the main determining factor in the water retention characteristic, while SOM increases effective water storage, especially at low SOM contents (fig. 3A and B). However, sensitivity analyses show that forest transpiration is linked to SOM due to alteration of root water uptake effectiveness rather than altered soil water retention. Measurements showed that forest floor evaporation is higher than previously estimated. Yearly amounts may be as high as 30% of the evapotranspiration or 80-150 mm per year. Forest floor transpiration can thus cause a significant decrease in water
733 availability, especially in more open types of forest where wind speeds and radiation levels at the forest floor are high. Concluding it is expected that forests on uniformly textured poor (drift sand) soils are sensitive to forest floor thickness and root depth. Forests or, broadly textured (fiver or glacial) soils are more dependent on the amount of SOM in the mineral soil but have a high water holding capacity. Possible future changes in meteorological circumstances (rainfall and evaporation patterns) would therefore have a larger impact in the water availability of forests on uniformly textured soils. Changes in accumulation and decomposition of SOM would have a larger effect of forests with soils having broader textures. The research results still need to be integrated with predictions of SOM accumulation or decomposition dynamics and altered meteorological conditions due to climatic change.
Transpiration reduction Site 1
Site 2
60
~
35
'
'
BD
55
--e- -
pF
--e--
K(h)
\ \
30
X 25
~~
rr
~
45
40 .0.00
"~
~.~'
20
15 0.50
1.00 SOM amount
1.50 [%]
2.00
0
50
' 100 SOM amount
"
" 150
207.
[%]
Figure 4A,B. Transpiration reduction as affected by soil hydraulic properties varying with soil organic matter content. One property is varied at the time while the rest is kept constant, pF: water retention characteristic, K(h) hydraulic conductivity, BD: soil bulk density, 0RED point of water uptake reduction, All: all properties varied.
734
Site 1 .
.
.
Site 2 .
.
.
g
o
,'.._
,'
', ,'
',
'
'__
o
o ',
,
,
'',
,
,
"'
'
g
o m
.
',
',
~
o
o
7
m
o
"~
~
u3
~
m
[~o1 uop,0nP ~
o
mo
m
"
o
'
g
,,
)
.~>~
o
mm m ~ [%] uop,0npaEI
o
:
,
,
K~ l
,i
,~
~ g
'<
o
,
00 ~o [%] uo!10npal::l
g
'
o
/ .~
o
~
o
~
"*
"~
[%] uo!10npeEI
o
g
o
o9
~o
Figure 5A-D. Transpiration reduction dependent on the simultaneous variation of mineral soil SOM and forest floor thickness (5A and B) and mineral soil SOM and root depth (5C and D).
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
735
Carbon allocation in mature grass ( L o f i u m p e r e n n e ) under elevated CO2 at two soil nitrogen levels. A. Gorissen, J.H. van Ginkel and H. van de Beek DLO-Research Institute for Agrobiology and Soil Fertility, P.O. Box 14, 6700 AA Wageningen, The Netherlands
Abstract The uptake of atmospheric carbon by terrestrial ecosystems may play an important role in the global carbon cycle since every 6-7 years the whole atmospheric carbon content passes through the plant biomass. Major uncertainties in this area concern the persistency of growth stimulation by elevated CO/and effects on carbon allocation to the soil compartment. In this study the effect of elevated C O / o n growth and carbon allocation of Lolium perenne was investigated. Plants were pretreated at 350 and 700/~L L ~ at two nitrogen levels (135 and 400 kg N ha -~yr -~) for 14 months and subsequently crosswise transferred to ESPAS-phytotrons for a short-term treatment at 350 and 700 ~L L -~ CO: for three weeks. The pretreatment stimulated total shoot growth until the end of the experiment, although no CO: pretreatment effects were observed in the yields of the last cut. The higher nitrogen level almost doubled shoot yield throughout the experiment. The fact that nitrogen stimulated shoot growth until the end of the experiment suggests that the disappearance of the growth stimulation of shoots by elevated CO/was not primarily caused by exhaustion of nitrogen or other nutrients in soil. The nitrogen effect on root growth showed an interaction with the CO2 pretreatment. At the lower nitrogen level root dry weight was not increased at 700/zL L -1 CO:, whereas at the higher nitrogen level a strong increase was observed. This interaction indicates that nitrogen may have important implications for stimulating effects of elevated CO: on root growth on the long-term and thus on carbon allocation to the soil.
1. INTRODUCTION The concentration of CO2 has steadily increased since the Industrial Revolution from 270 ttL L -I to the present value of about 355 #L L -1. The effects on crop yield have been surveyed by Kimball (1983), Cure & Acock (1986) and several others, who concluded that a doubling of the CO: concentration will probably result in an average increase of 30% in crop yield under optimal conditions. Bazzaz (1990) questioned whether such a stimulation would be prolonged in time, since plants adapt to changing circumstances and soil nutrients may be exhausted on the long-term. Although numerous studies have investigated the effects of elevated CO2 on plants, only few have also focussed on roots (Rogers et al., 1994). They described plant reactions on elevated CO2 with emphasis on the soil compartment and concluded that exploration of CO: effects on roots and soil needed strong attention since many uncertainties still have to be solved. The objectives of this study were i) to study the effect of elevated CO: on biomass production and carbon allocation in a perennial species exposed to elevated CO/for a long period, ii) to determine how the effects of elevated CO2 are affected by different nitrogen application rates.
736 2. MATERIALS AND METHODS
2.1. CO, treatments Lolium perenne cv 'Barlet' plants were cultivated from seed in 2-L columns. The columns were filled with Ede loamy sand with a bulk density of 1.2 kg 1-1 (dry weight) and soil moisture was adjusted to about 14% (w/w). All columns received 100 mg N (KNO3) at the start of the experiment. The plants were grown in two adjacent greenhouse compartments with ambient (about 350/zL L -~) and elevated (700 #L L ~) CO2 levels for 14 months (June 1992 September 1993) without additional light. During this pre-incubation period, in the following referred to as pretreatment, the CO2 levels were measured by an Ari-P analyzer (Siemens). The columns were watered weekly and at several dates readjusted to their initial weight with de-ionized water. In spring 1993 the nitrogen treatment was started, the lower nitrogen treatment receiving an amount of nitrogen corresponding with 135 kg N ha-~ yr-~ and the higher treatment corresponding with 400 kg N ha-1 yr-~. After the pretreatment 24 Lolium perenne plants were randomly selected and transferred to the ESPAS growth chambers (Experimental Soil Plant Atmosphere System. Half of the plants from each greenhouse compartment were placed in one ESPAS chamber and vice versa to distinguish between pretreatment and treatment effects. After three days of acclimatization, the plants were pulselabelled for one day with 14CO2. Due to a technical flaw the specific activities in the growth chambers are unknown. For this reason, the effects of the CO2 treatment on the total net CO2 uptake could not be measured. Nitrogen effect and CO2 pretreatment effect on total net uptake were not affected by this flaw, since these treatments were equally distributed among the systems. After the 1-day pulse-labelling the growth cabinets were flushed with fresh air and the plants were further treated with 350 or 700 ~L L -~ CO2 for three weeks. The preset atmospheric CO2 levels were maintained either by injecting CO2 or by removing it by carbosorb filters (Sodasorb, Grace). CO2 was supplied from gas cylinders (100% CO2) and the inflows were controlled automatically by means of Brooks flow controllers. CO2 levels were measured by an URAS 10E infrared analyzer (Hartmann & Braun). Temperatures in the growth chambers (shoot 20~ at day; 15~ at night; roots 16~ at day; 11~ at night) were measured by a platinum resistance thermometer Ptl00, relative humidity (70% at day; 80% at night) using a capacitive humidity sensor and irradiation (300 #mol m-2 s-~ at plant level) by means of a PAR-meter. Wind velocity was set at 0.1 m s-1. All environmental variables were checked with a third independent meter to assure identical conditions. Day/night rhythm was 16/8 h. Prior to CO2 treatment, the column lids were sealed with a silicon rubber (Q3-3481, Dow Chemical) to prevent exchange of x4CO2between growth chamber and soil columns. During the experiment, columns were flushed every 6 hours with CO2-free air at a flow rate of about 40 L h -a for 15 minutes, to prevent O2 deficiency and to remove CO2 from the soil. Root/soilrespired CO2 was trapped by conducting the air through a 300 mL 2 N NaOH solution. 2.2. Analyses Root/soil respiration was measured every third day by taking an aliquot of 1 mL of the NaOH solution and precipitating the HCO3-and CO32-ions with excess BaCI2. Total CO2 was determined by titrating the remaining NaOH with 0.2 N HCI. 1'CO2 was determined in a subsample by liquid scintillation counting using Ultima Gold (Packard). The plants were harvested after 21 growth days in the ESPAS growth chambers. Dry
737 weights of shoots and roots were determined after drying at 80~ for 24 hours. Dried plant material was ground and homogenized and a wet combustion procedure was used to determine total C and 14C. Plant material (30 mg) and soil (1 g) were digested with a 5 ml solution of 2.0 g K2Cr207 in 25 ml H2SO4 and H3PO4 (3/2 v/v) at 160~ for 2 hours. Released CO2 was trapped in 10 ml 1.0 N NaOH, and processed as described above.
3. RESULTS Figure 1 shows the cumulative shoot yields of Lolium perenne during the pretreatment. The first cut after sowing showed that the CO2 pretreatment had increased shoot growth by about 20% (P < 0.04). In the second season, after the N treatment started, this CO2 pretreatment effect could be observed until the end of the experiment, although the increase in total yield reduced to about 16% (P<0.01). The last cut before the treatment in the ESPAS growth chambers revealed no differences in yield anymore. From the very start of the nitrogen application in the second season, the 400 N treatment increased the shoot yield of the first cut by about 18% (P<0.01), independent of the CO2 pretreatment. The average increase amounted to about 65 % (P < 0.001) at the end of the experiment. The results after the CO2 treatment in the ESPAS growth chambers are shown in table 1. During this treatment the 400 N treatment increased shoot yield by about 91% compared with the 135 N treatment. The CO2 pretreatment strongly increased dry root weight at the end of
Lolium perenne
Yield (g) 30 f 25
*
.................................................................
"/~o"" J
20
............................................................
f ":.':: ..... /..'" .&
r 15 !- ......................................................
4~.....:... ~.. . . . .
f .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6,*
0
s
.
.
........... : 0"-
.
.
.
.
.
.
.
.
.
.
'
10 135N-350C
20
.
.
.
.
.
.
.
.
.
.
.
.
.
.
I
30 40 Plant age (weeks) 135N-700C
o~
400N-350C
50
60
70
400N-700C
Figure 1. Yield of Lolium perenne during the pretreatment at 350 and 700 #L L -1 CO2 at two nitrogen levels (135 kg N ha-1 yr-1).
738 Table 1 Biomass (g) and shoot/root ratio after treatment of Lolium perenne with two nitrogen levels (135 and 400 kg N ha -1 yr -~) and with 350 #L L -1 CO2 and 700 #L L -1 CO2 for three weeks after applying a 14C-pulse-label at day 1 (n=3). Nitrogen level 135 350 350
Leaves Roots Shoot/root ratio
Statistics'
Leaves Roots Shoot/root ratio
5.4 4.5 1.2
700
4.9 4.5 1.1
N
< 0.001 < 0.001 ns
400 Pretreatment 700 350 700 Treatment 350 700 350 700 350 700
Plant biomass 5.5 10.5 4.6 7.5 1.2 1.4
6.8 4.5 1.5
P
11.2 10.2 1.1
10.4 14.9 0.7
T
interactions
ns 0.02
ns ns
N* P
ns
ns
10.9 18.2 0.6
1 N = Nitrogen; P-- Pretreatment; T = Treatment; ns---not significant.
the experiment, dependent on the nitrogen level. At 135 N no increase was observed at 700 CO2, whereas at 400 N dry root weight was strongly increased. The percent distribution of 14C among the different compartments is shown in table 2. The percentage recovered in the shoots and roots was not affected by any of the treatments. The percentage in the root/soil respiration decreased in the 400 N treatment by 29% compared with the 135 N treatment, but was unaffected by the other treatments. Also the percentage 14C recovered in the microbial biomass decreased from about 2 % to 1% in the 400 N treatment compared with the 135 N treatment. The residual I'C in soil was also decreased in the 400 N treatment compared to the 135 N treatment, whereas it was increased by 95 % in the 700 CO2 treatment compared with the 350 CO2 treatment. Although the main effect of the CO2 treatment can not be estimated (see above), the main effect of nitrogen showed a 150% increase at 400 N compared with 135 N. The CO2 pretreatment had no effect on total net CO2 uptake.
739 Table 2 Distribution of 14C (% of net total uptake) among different plant/soil compartments after treatment of Lolium perenne with two nitrogen levels (135 and 400 kg N ha -1 yr -1) and with 350/zL L -~ CO2 and 700/zL L -I CO2 for three weeks after applying a ~'C-pulse-label at day 1 (n-3). Nitrogen level 135
400 Pretreatment
350
Leaves Roots Root/soil respiration Microbial biomass Soil residue Shoot/root ratio 1'C Total net uptake (kBq)
700
350
700
350
700
350
Treatment 700 350
47.0 27.8 21.5
47.1 21.5 24.8
48.6 18.4 29.0
% distribution 48.4 56.1 12.9 24.4 33.6 16.9
1.9 1.8
2.1 4.5
2.3 1.7
2.7 2.5
1.2 1.4
1.2 2.4
1.5 0.7
0.5 1.5
1.9 178
2.4 94
2.7 160
3.7 92
2.5 435
3.0 171
3.2 362
1.6 327
Statistics 1
N
P
T
Leaves Roots Root/soil respiration Microbial biomass Soil residue
ns ns 0.03
ns ns ns
ns ns ns
< 0.001 ns
ns ns
ns ns
14C shoot/root ratio Total net uptake (kBq)
0.07 < 0.001
ns ns
ns nr
700
350
700
54.0 20.2 22.2
55.7 22.0 20.1
48.6 31.1 18.2
interactions
1 N = Nitrogen; P = Pretreatment; T = Treatment; ns = not significant; n r - not relevant
4. D I S C U S S I O N
Stimulating effects of CO2 have often been reported, but frequently doubts have been raised about the persistency on the long-term. Adaptation of the plants or exhaustion of soil nutrients could eventually reduce the initial stimulation. In this study some evidence was found that an initial stimulation of shoot growth was reduced during the second season. After 66 weeks the
740 cumulative shoot yield was still increased by 16% after an initial increase of about 20%. However, the last cut before the treatment in the ESPAS growth chambers revealed no differences in yield indicating that the growth stimulation, due to the pretreatment, had disappeared. The slopes in figure 1 support this conclusion since the rates of shoot biomass production are almost equal at the end of the experiment. It seems not plausible that nitrogen plays an important role in this reduction in growth stimulation, since it occurred at both nitrogen levels almost at the same time. Exhaustion of other soil nutrients or morphological/physiological adaptations of the plants are more likely explanations. The 400 N treatment stimulated shoot growth throughout the pretreatment and treatment. Although the amount applied was three times higher than in the 135 N treatment, the yield was almost doubled. The fact that nitrogen stimulated plant growth until the end of the experiment suggests that serious exhaustion of other nutrients did not occur. This implies that, with regard to the disappearance of the growth stimulation by elevated CO2, probably other mechanisms play a role than exhaustion of soil nutrients. In contrast to shoot yield, a strong interaction between nitrogen and CO2 pretreatment was observed for root dry weights. In the 400 N treatment root dry weight strongly increased by 46 % at 700 CO2, whereas no increase was found at 135 N. Summarizing, at the low nitrogen level growth stimulation by CO2 disappeared both in shoots and roots, whereas at the high nitrogen level only the root growth was still stimulated by elevated CO2. This was also expressed in the s/r ratio (Table 1) which was lowest in the 400 N/700 CO2 pretreatment. The interactions with the CO2 treatments indicate that nutrients such as nitrogen, apart from acclimation, may have important implications for stimulating effects of CO2 on the longer term especially with regard to the carbon storage capacity of terrestrial ecosystems. Also with regard to a possible decrease in decomposition rates of litter (CoOteaux et al., 1991; Cotrufo et al., 1994) and roots (Gorissen et al., 1994) the availability of nitrogen may play a key role in the sequestration capacities of natural ecosystems.
5. REFERENCES
Bazzaz FA. 1990. The response of natural ecosystems to the rising global CO2 levels. Annual Review of Ecological Systems, 21, 167-196. Cotrufo MF, Ineson P. and Rowland AP. 1994. Decomposition of tree leaf litters grown under elevated CO2: Effects of litter quality. Plant and Soil, 163, 121-130. Co0teaux MM, Mousseau M, C616rier ML and Bottner P. 1991. Increased atmospheric CO2 and litter quality: decomposition of sweet chestnut leaf litter with animal food webs of different complexities. Oikos 61, 54-64. Cure JD and Acock B. 1986. Crop responses to carbon dioxide doubling: A literature survey. Agriculture and Forest Meteorology, 38, 127-145. Gorissen A, Van Ginkel HJ, Keurentjes JJB and Van Veen JA. 1994. Grass root decomposition is retarded when grass has been grown under elevated CO2. Soil Biology & Biochemistry (in press). Kimball BA. 1983. Carbon dioxide and agricultural yield. An assemblage and analysis of 430 prior observations. Agronomy Journal, 75, 779-788. Rogers HH and Runion GB. 1994. Plant responses to atmospheric CO2 enrichment with emphasis on roots and the rhizosphere. Environmental Pollution, 83, 155-189.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
741
WHEAT AND MAIZE PRODUCTION IN HUNGARY UNDER DOUBLED ATMOSPHERIC CO2 CONCENTRATION M. Hunk~ Zemankovics a and ZS. Bacsi b a Agrometeorological Research Station of the Hungarian Meteorological Service, Keszthely, P.O.Box. 80. H-8361, Hungary b Pannon Agricultural University, Keszthely, P.O.Box.71. H-8361 ,Hungary Abstract
The present investigation aims at asseseing the impacts of three well known climate change scenarios - the carbon dioxide doubling scenarios from the GISS, GFDL and UKMO general circulation models - on wheat and maize yields in Hungary. For this purpose the monthly outputs for global radiation, temperature and precipitation of the above three GCMs were used to create daily weather time series for Hungary. Climatic data scenarios were chosen from the gridbox which covers the location of Keszthely. Historical weather data of 16 years from Keszthely were used as baseline. These were used together with the CERES-Wheat and CERES-Maize crop growth simulation models. The validation of crop models is based on field experiment data from Keszthely. Statistical analysis of simulated crop data is presented. Comparing to the baseline the average wheat yield shows 13-25 % decrease in spite of the slightly increase of total biomass production according to different GCM scenarios. In case of maize the GISS scenario resulted in a small 8 % yield increase on the average, while the GFDL and UKMO resulted in 7% and 14 % yield decrease respectively. Yield decreasing mainly due to the shortened length of growing period. 1.THE CREATION OF THE CLIMATE CHANGE SCENARIOS Outputs from three General Circulation Models were used to create climate change weather scenarios. These are: -GISS (Goddard Institute for Space Studies, New York,U.S.A. [ 1], -GFDL (Geophysical and Fluid Dynamics Laboratory, Princeton, U.S.A. [2], - UKMO ( United Kingdom Meteorological Office, Bracknell, U.K. [3]. All the three applied GCM are so called equilibrum models and they have their limitations regarding spatial resolution. The GISS handles the earth's surface as gridboxes of the size 10 ~ latidude x 7.9 ~ longitude. The climate of each gridbox is considered to be homogeneous, and the climate characteristics are allocated to the gridbox centre. The GFDL works similarly with gridboxes of 4.5~ x 7.5~ and the UKMO with gridboxes of 5~ x 7.5~ In the present study the gridbox covering the location of Keszthely (46.4~ 17.3~ was used with each
742 GCM, thus for GISS the gridbox with the centre 50~176 for GFDL the gridbox with the centre 46.7~176 for UKMO the gridbox with the centre 47.5~176 Regarding temporal resolution, the three GCMs are capable of calculating the weather of a year as monthly average values at most detail. Since the GCMs are inaccurate in simulation of the present climate, so the climate charactherstics simulated under changed greenhouse gas concentrations are also rather unreliable. Therefore we used the rate of changes instead of the absolut values of predicted climatic variables. The "baseline" weather was taken from the location of Keszthely from the period 1975-1990. The change factor was calculated as the difference between the 'future' and 'present' values for temperature, and as the ratio for radiation and precipitation, in agreement with the recommended methodology of many similar impact assesment studies [4],[5],[6]. As the crop growth simulation models require daily weather data and the climate model outputs present the results only in monthly resolutions, the baseline daily weather data were modified by the change factor of the corresponding month to create climate change scenarios with daily resolutions. 2. CROP GROWTH SIMULATION MODELS
The growth and development of the winter wheat and maize crop were assesed using the CERES Wheat and CERES Maize models [7],[8]. Soil and plant characteristics required as input data were chosen according to typical ones for the location of Keszthely as well as the agrotechnology data like the date of sowing. Before starting the climate change experiment the validation of the models had been carried out. In the case of CERES Maize the average differences between the predicted and observed plant variables were less than 4 % [9] and for CERES Wheat the average difference is about 6 % [ 10]. In order to compare the impact of the changed weather on the crops the agrotechnology was assumed constant for all simulation experiments. In the simulation runs the following agrotechnology parameters were used" sowing on 10 October with MV-4 winter wheat variety for the wheat model, and sowing on 20 April with Pioneer 3901 variety for the maize model. For both crops optimal nitrogen supply was assumed, and no irrigation was applied. 3. RESULTS The outputs from the CERES crop models were used for the impact assessments, that is, for both wheat and maize the simulated values of resultant variables under the baseline weather were compared to the simulated values of the same variables under the climate change weather scenarios generated from the three ~ s . The following resultant variables
743 were considered in the assessment: the grain yield (t/ha), the amount of above ground biomass (t/ha), maturity date. 3.1 MAIZE
In the maize experiments altogether 16 baseline weather years were available together with 16 GISS-years, 16 GFDL-years and 16 UKMO-years. The climate change scenarios resulted in maturity occuring much earlier, and thus the growing season became significantly shorter for all the three different ~ s . Biomass and grain yield show somewhat less unanimous results. The GISS scenario resulted in a small, 8% yield increase on the average, while the GFDL resulted 7%, and the UKMO in a 14 % yield decrease. In biomass production the GISS and the GFDL scenarios resulted small increase while the UKMO scenario resulted small decrease. For all the three GCM scenarios the standard deviations of the yield are somewhat smaller than for the baseline weather (table 1.). Table 1 The averages and the standard deviations of the simulated resultant variables of maize for different scenarios Maturity date (day of the year) avg. Base GISS GFDL UKMO
248 225 212 207
Grain yield (t/ha)
Biomass (t/ha)
std.
avg.
std.
avg.
std.
18.0 5.3 3.8 7.0
8.57 9.29 7.96 7.35
3.35 1.07 0.90 2.20
15.95 17.29 16.10 15.25
3.89 1.20 1.16 3.25
3.2 WINTER WHEAT In the wheat experiments altogether 15 baseline weather years were available together with 15 GISS-years, 15 GFDL-years and 15 UKMO-years. Results showed that similar to maize the maturity dates occured earlier and the growing season became significantly shorter for all of the climate change scenarios. The fastest crop development occured in the case of the UKMO scenario, with an average of 42 clays shorter period from sowing to maturity in comparison to the baseline, while under the GFDL scenario maturity occured 25 days earlier on the average, and under GISS scenario 22 days earlier in average than under the baseline weather. The average biomass production slightly increased for all the three climate change
744 scenarios, and the standard deviation decreased. In the case of grain yield the GISS scenario resulted in a yield decrease of 13 %, the GISS scenario resulted 28 % decrease and the UKMO scenario 25 % decrease. Standard deviation of crop variables decreased for all the three GCM scenarios comparing to the baseline. (see table 2.) Summarizing that equilibrium GCM scenarios give a significant warmer climate with moderately higher precipitation amounts. These conditions seem to be favourable for total biomass production but unfavourable for grain filling processes. Table 2 The averages and standard deviations of the simulated resultant variables of winter wheat for different scenarios Maturity date (day of the year)
Base GISS GFDL UKMO
Grain yield (t/ha)
Biomass (t/ha)
avg.
std.
avg.
std.
avg.
std.
193 171 169 151
5.5 5.5 5.1 5.3
6.28 5.44 4.49 4.73
0.85 0.62 0.50 0.68
17.45 18.34 17.75 18.58
2.02 1.47 1.82 1.21
4. REFERENCES
9 10
J.Hansen, I. Fung, A. Laces, S. Lebedeff, D. Rind, R. Ruedy and G. Russet, J.Geophysic. Res., 93 (1988) 9341. R.T. Wetherald, and S. Manabe, Climatic Change, 8 (1986) 5. C.A. Wilson and J.F.B. Mitchell, J. Geophysic. Res. 92 (1987) 13315. M.L. Parry, T.R. Carter and N.T. Konijn (eds.), The Impact of Climatic Variations on Agriculture. Vol. 1. Kluwer Academic Press, 1988. J.B. Smith and J.A. Tirpak (eds.), The Potential Effects of Global Climate Change on the United States. EPA, Washington, DC., 1989. R.M. Adams, C.Rosenzweig, R.M. Peart, J.T.Ritchie, B.A. McCarl, J.D. Glyer, R.B. Curry, J.W. Jones, K.J. Boote and L.H.Jr. Allen, Nature, 345 (1990) 219. D.C. Godwin, J.T. Ritchie, U. Singh and L. Hunt, A User's Guide to CERES Wheat-V2.10. Muscle Shoals, Alabama: International Fertilizer Development Centre, 1989. S.A. Jones and J.R. Kiniry, CERES Maize: A Simulation Model of the Growth and Development of Maize. Texas A&M University Press, College Station, Texas, 1986. M. Hunk~,Id6j~is, 98 (1994)37. ZS. Bacsi and M. Hunker,, Id6j~is, 98 (1994) 119.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
745
Effects of climate change on yield potential of wheat and maize crops in the European Union J. Wolf a and C.A. van Diepen b a Department of Theoretical Production Ecology, Wageningen Agricultural University, Bornse steeg 65, P.O. Box 430, 6700 AK Wageningen, the Netherlands b
DLO the Winand Staring Centre, P.O. Box 125, 6700 AC Wageningen, the Netherlands
Abstract Yields of winter wheat, silage maize and grain maize in the main arable areas of the European Union (E.U.) were calculated with a simulation model, WOFOST, using historical weather data and average soil characteristics. The sensitivity of the model to individual weather variables was determined.Subsequent analyses were made using climate change scenarios with and without the direct effects of increased atmospheric CO 2. The impact of crop management in a changed climate was also assessed. The various climate change scenarios used appear to yield considerably different changes in yield, both for each location and for the E.U. as a whole.
1. Introduction Any changes in climate which may result from increasing concentration of greenhouse gases in the atmosphere could have dramatic consequences for the agricultural yield potential. In this study that formed a part of the E.U. funded EPOCH project, the effects of climate change on the yields of winterwheat, silage maize and grain maize were analysed. In the EPOCH project as a whole the effects of climate change on agriculture and horticulture in the E.U. have been investigated by scientists from nine research institutes in the E.U. The main tasks in this project were: 1. construction of scenarios of climate change; 2. broadscale analysis of the effects of climate change on selected crops; 3. site-specific analysis of the effects of climate change on crop phenology and yields with crop growth simulation models; 4. experimental investigation of the effects of CO 2 enrichment and temperature on crop growth.
2. Model and input data A dynamic crop growth model, WOFOST, developed for calculating agricultural yield potential [ 1] was used. In the model, crop growth is simulated from sowing to maturity on the basis of physiological processes as determined by the crop's response to environmental conditions. The simulation is carried in time steps of one day. The major processes considered are CO 2 assimilation, respiration, partitioning of assimilates to various plant organs, transpiration and phenological development. Two levels of crop yield are calculated: the water limited yield which can be realized in situations where the supply of plant nutrients and crop management are optimum and secondly the potential yield which can be realized if also the water supply is optimum. In order to apply the model data that specify crop growth and phenological development are required including information on properties that determine assimilation and respiration processes and response to moisture stress, partitioning of assimilates to plant organs, life span of leaves and death rates of plant organs. This information is
746 derived from the literature and field experiments. Calculations of crop growth were made for historical sets of daily weather data from 20 meteorological stations. For each of the stations that are representative of the main arable land areas in all E.U. countries, the weather data set covered generally the period 1966-1985. In order to calculate the soil water balance, the soil's physical characteristics such as effective soil depth, soil moisture characteristics and hydraulic conductivity must be known. For each meteorological station the main soil types that occur on arable land areas within a radius of 100 to 150 km around the station, were obtained from the soil map of the European Communities. By interpreting the information per soil type quantitative terms for use in the simulation model could be obtained.
3. S c e n a r i o data In order to calculate crop growth in future the historical weather data were changed on the basis of climate change scenarios. Two different types of scenarios have been applied. Firstly, the composite time-dependent scenarios that were based on the average standardized output of seven equilibrium GCM's.This output resulted in a regional pattern of climate change. By using estimates of global mean warming for the various IPCC emission scenarios (in this study scenario A (i.e. business-as-usual emission scenario) for the years 2010, 2030 and 2050 and scenario A High (high estimate of climate change for scenario A) for year 2050 [2] ) which were calculated with a simple climate model, the GCM-derived regional patterns could be scaled up to obtain the time-dependent changes in climate [3]. Secondly, the individual scenarios that were based on output from three equilibrium 2"CO2 GCM's, i.e. GFDL, GISS and UKMO-L models. 4. R e s u l t s Variables that determine crop yield directly, are atmospheric CO 2 concentration, solar radiation and temperature and those that affect the water balance and hence the degree of drought stress are rainfall, windspeed, vapour pressure and again the other three variables. These variables were adjusted separately, in a stepwise manner, in order to gauge the sensitivity of crop yields to changing values of each. This sensitivity analysis was carried out for winter wheat on three locations, i.e. Kinloss, Orleans and Brindisi that cover the main differences in climate in the E.U. (Table 1). Table 1 Sensitivity of potential (POT) and water-limited (WAT) grain yields of winterwheat in Kinloss, U.K., Orleans, France and Brindisi, Italy to increasing values for atmospheric CO 2 concentration (C), temperature (T), rainfall (R), solar radiation (S), windspeed (W) and vapour pressure (V) 1 (Source = [4]).
POT WAT
C
T
R
S
W
V
++ ++
--,02 __,+3
0 ++
+,++ _
0 __
0 ++
1 0, +, ++ : no, moderate, strong increase in grain yield; - -- : moderate, strong decrease in grain yield. 2 Temperature effect varies from about zero in Kinloss to strongly negative in Brindisi. 3 Temperature effect varies from strongly and moderately negative in Kinloss and Orleans, respectively to moderately positive in Brindisi.
747
10000
]
.oooi 6000 r
4000 2000
0
|
|
Present
A2010
A2030
A2050 Ahigh2050
Figure 1. Average potential (Pot.) and water limited (Wat.) grain yields of winter wheat cultivated at current and at future climate conditions (based on composite scenarios A and A High) in Kinloss, U.K., Orleans, France and Brindisi, Italy, taking into account the direct effect of increasing atmospheric CO 2 in future (Source = [4]).
<-1500
-1500<
<-500
<.v) -500<
500<
<500
<1500
> 1500
o-
. . . . . F------. .~
lc,
__/
?.
Figure 2. Changes in water limited grain yield (kg ha "l dry matter) of winter wheat in the main arable land areas in the E.U. if the weather is changed on the basis of the GISS equilibrium 2'CO 2 scenario and the direct effect of increased CO2 is taken into account (Source = [4]).
748 Potential and water limited grain yields of winter wheat were calculated for historical weather data that were changed on the basis of composite scenarios A and A High (Fig. 1), taking also into account the direct effect of increasing atmospheric CO 2 concentration. Both potential and water-limited yield levels appear to increase greatly over time, mainly as a result of increasing CO 2. For scenario A High that gives the largest increase in temperature, the yield increase is somewhat less than that for scenario A for the same year. If water limited grain yields were calculated for the individual scenarios and the direct effect of increased CO 2 (from 353 to 560 ppm) was taken into account, the GISS scenario gave major yield increases for most locations in the E.U., mainly as a result of CO 2 enrichment (Fig. 2). The GFDL scenario on the other hand, gave yield decreases for the western part of the E.U. (Fig. 3).
<-1500
9 - 1500<
-500<
t-<..../
500<
<500
< 1500
> 1500
9 "--~....,.//
!
<-500
--,. ...... /
/
/
o~o_%_ ~oo .~oo 400 ~~o ~r, 110'
15" L
Figure 3. Changes in water limited grain yield (kg ha -1 dry matter) of winter wheat in the main arable land areas in the E.U. if the weather is changed on the basis of the GFDL equilibrium 2'CO 2 scenario and the direct effect of increased CO 2 is taken into account (Source = [4]). If climate changes, present crop management may be inadequate for the new climate conditions. A number of management responses and their potential usefulness in adapting to the adverse effects of climate change were evaluated. Firstly, yields were calculated for crop varieties that differ with respect to their temperature sum requirements for phenological development. Secondly, amounts of irrigation water required to attain the potential yield level were determined (Table 2). They increase with climate change but decrease if also the direct effect of increased atmospheric CO 2 is taken into account. Finally, the impact of changes in sowing date on the yield level of for example silage maize
749 were determined (Fig. 4). In a changed climate the optimum day for sowing appears to move to an earlier date. Table 2 Required amounts of irrigation water (mm) for attaining the potential yield level of grain maize on three locations in the E.U., both for historical climate and for scenario climate with and without direct CO 2 effect (Source = [5]). Location
Hist. climate
Brindisi, Italy Mannheim, Germany Orleans, France
Scen. A 2050
382 92 157
Scen. A 2050 + CO 2
400 109 211
326 66 147
24000
22000
20000
18000 Orleans Hist.
---o----
Orleans A
16000
Orleans A
14000
. 40
. 6'0
.
. 8'0
High . 1;0
. 120
140
160
Sowing date
Figure 4. Sensitivity to changes in sowing date of the average potential total yield of silage maize cultivated in Orleans, France. Yields have been established for historical weather data (Hist.) and for composite scenarios A and A High for the year 2050 (A and A High) (Source = [6]). 5. C o n c l u s i o n s - Various climate change scenarios appear to yield considerable different changes in yield both for each location and for the E.U. as a whole; - The direct effect of increasing atmospheric CO 2 on wheat grain yield appears to be much greater than the effect of climate change; The average water limited grain yield of winter wheat in the E.U. increases by 1000, 1300 and 2300 kg ha -1 dry matter for the UKMO-L, GFDL and GISS scenarios respectively (based on calculated yield changes and the present area distribution in the E.U.) if the direct effect of increased CO 2 is taken into account; - The average potential grain yield of maize in the E.U. decreases by 2000, 2800 and 2800 kg ha -1 dry matter for the GISS, GFDL and UKMO-L scenarios respectively; -
750 -
The average water limited total yield of silage maize in the E.U. decreases by 600, 2500 and 6300 kg ha -1 dry matter for the GISS, UKMO-L and GFDL scenarios respectively if the direct effect of increased CO 2 is taken into account.
For more information on this study it is referred to [4], [5] and [6]. Complete reports on this study can be obtained from the first author. Results from the EPOCH project are given in a special issue of the European Journal of Agronomy (1993-2(4) ). 6.
References
1 Diepen C.A. van, WolfJ., Keulen H. van and Rappoldt C., 1989. WOFOST: a simulation model of crop production. Soil Use and Management 5:16-24. 2 Houghton J.T., Jenkins G.J. and Ephraums J.J. (Eds.), 1990. Climate change: the IPCC scientific assessment. Report of working group I of the intergovernmental panel on climate change. Cambridge university press, Cambridge. 3 Barrow E.M., 1993. Scenarios of climate change for the European Community. European Journal of Agronomy 2: 247-260. 4 Wolf J., 1993. Effects of climate change on wheat production potential in the European Community. European Journal of Agronomy 2:281-292. 5 Wolf J. and Diepen C.A. van, 1994. Effects of climate change on grain maize yield potential in the European Community. Climatic Change, in press. 6 Wolf J. and Diepen C.A. van, 1994. Effects of climate change on silage maize production potential in the E.C. Agricultural and Forest Meteorology 71: 33-60.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
751
Mediterranean soilscapes and climatic change. An overview J.J. Ib~ifieza, A. Garcia-Alvarez a, J.L. Gonz~ilez Rebollar a and A.C. Imeson b aCentro de Ciencias Medioambientales (CSIC), Serrano 115 bis, 28006 Madrid, Spain. bFysisch Geografisch Bodemkunding Laboratorium, Universiteit van Am~erdam, Nieuwe Prinsengracht 130, 1018 VZ Amsterdam, The Netherlands
Abstract Despite the reduced extension and spatial fragmentations of the Mediterranean region, its biogeodiversity (biologic, edaphic, geomorphic, climatic and hydrologic) is among the highest of the planet. Quantitative data published recently show great links between arid and Mediterranean soilscapes. Mediterranean soilscapes are extremely fragile and therefore potential changes due to climatic and anthropic causes may lead to desertification. Thus, future CO2 warming will increase current desertization processes. Land degradation in arid and semiarid Mediterranean regions (rainfall less than 600 ram) is highly analogous to desertification. 1. GENERAL CHARACTERISTICS: MEDITERRANEAN LANDSCAPES AND CLIMATE CHANGE
The Mediterranean climate currently affects approximately 9,000,000 km 2 (4.6% of the continental lithosphere or 3.3% of the global pedosphere). Mediterranean ecosystems are distributed over Mediterranean Basin, California, Chile, South Africa and Australia. About 3/4 of Meditarreanean basin area enjoys a Mediterranean climate whilst 1/4 suffers greater aridity. The Mediterranean camlot be regarded as a homogeneous morphoclimatic region. A relatively wide range of marked climatic zones exist between wet and arid. In general, the Mediterranean region is characterized by generally warm temperatures, winter or spring and autumn-dominated rainfall, dry and hot summers (of at least two or three months and up to 11 months) and a profusion of microclimates due to local terrain. Mean annual rainfall can vary from 20-25 mm in the "Mediterranean deserts" to 2000-2500 mm on mountain slopes or local maritime areas exposed to rain-bearing winds. Temperature varies widely with latitude, altitude and continentality: the annual mean may drop below 10 C or rise above 25 C. In parallel with these mean conditions the intra-annual and inter-annual climatic variability is very high, with sudden and high intensity rainfall. According to the World Bioclimatic Classification System proposed by Rivas-Martinez [1], the Mediterranean climate has the highest diversity of bioclimatic belts in the world. The Mediterranean area is especially vulnerable to climatic change. The impact of CO 2induced climatic change is potentially serious because of the aridity, the low rates of primary production in the drier areas, the sensitivity of the soils to degradation and erosion, and the high risk of drought and fire. The increase in aridity however is not necessarily caused by decreases in annual precipitation. Increasing aridity can be due to: (a) higher
752 evapotranspiration; (b) a change in the frequency and magnitude of rainfall events; (c) processes of soil degradation that decrease the ability of soil to retain water [2]. Simulation models of the atmosphere's general flow predict that a doubling of the greenhouse gas concentration compared to pre-industrial periods would bring a mean worldwide temperature increase between 3 and 5.5 C [2]. Precipitation predictions are different for different GMC models. The GISS model indicates an increase in general precipitation for the Mediterranean Basin. The BMO model predicts a decrease in precipitation. Both models suggest an increment in potential evapotranspiration of 200 mm~ which may cause an extension of the period with a rainfall deficit and a very large increase in soil moisture deficit [2]. 2. M E D I T E R R A N E A N SOILSCAPES AND SOIL GENESIS 2.1. Mediterranean soils and climate
Duchaufour [3] distinguishes two types of soil characteristic in Mediterranean regions. Thus, when phytoclimatic conditions are relatively moist, the development of Alfisols [4] is common. On the other hand, hi arid and semiarid Mediterranean environments, the most representative soils are Haploxeroll, Argixeroll and Calcixeroll [4]. In view of the large variety of subclimates that can be included within the Mediterranean climate, it is not surprising that pedoclimates and, more specifically, humidity regimes, may be aridic, xeric and even udic in the high mountahls. The dry period causes characteristic rubification of many Mediterranean soils. Incomplete profile leaclfing also allows frequent horizon genesis with accumulation of carbonates raider the Bt horizons [3]. More acidic, weathered profiles develop on the oldest, most stable geomorphological Palaeosurfaces. These are or the Xerult and Ultic Xeralf [3, 4]. In these circumstances, the base of the Bt horizons has become silted up with clay. Impenneabilization induced by this process has the genesis ofhydromorphic features and aquic regimes as a corollary. In arid and semiarid Mediterranean subclimates highly developed, shallow calcic horizons are also characteristic. Likewise, wetting-drying cycles encourage hardening of the horizons with accumulation of carbonates, giving rise to the formation of petrocalcic horizons or calcretes [3]. In the most arid subclimates, caliche accumulation can resuk from the deposition of dust material. Under special conditions a distinct caliche layer can develop within several months. Where petrocalcic horizons are situated at or near the soil surface they can have serious implications for hydrology. Salt accumulation is most severe in arid and semi-arid subclimates (receiving between 300 and 600 mm of rainfall). However, this salt accumulation becomes less impoltant in areas where precipitation drops below this level [5]. Another typical characteristic of Mediterranean environments is the low organic matter content of surface horizons, so they rarely overcome the state of ochric epipedon. This is why unevolved soils conserve a large part of the properties of their original material. In some cases, the effects of small changes in the supply and mineralization of organic matter could have a large impact on soil structure and greatly influence soil and hillslope hydrology. Soils having low organic matter contents are less stable and therefore have a lower infiltration capacity. Soils that contain low amounts of organic matter can respond by slaking, swelling, dispersion, cracking and mellowing, when they are wetted, and because of this can develop surface seals or crust. Ill dispersive soils, physico-chemical characteristics may trigger piping erosion which creates a positive feedback of pipe collapse producing rill and gully incision [6]. Finally, undeveloped shallow soils (Entisols with umbric epipedons, adjacent to Inceptisols
753 and Histosols) predominate in Mediterranean mountain climates. [7].
2.2. Quantitative analysis of Mediterranean soilscapes Ib~iiiez et al. [8], have studied the Mediterranean soil landscapes, on a global level using the data compiled by the FAO [9, 10]. This information served to draw up a matrix on which Minimum Spanning Tree and cluster analysis were performed (Figs. 1 and 2). Close connections between soilscapes of arid and Mediterranean environments become apparent [8]. There are marked, though lesser, sinailarities between Mediterranean and other bordering climate zones. Moreover, a common trait of Mediterranean and arid climate zones is that they both present intermediate evolution soils (Calcisols and Cambisols). These latter two climate types however, are identified by the greater abundance of shallow and saline soils in arid environments and by that of other more developed ones in Mediterranean environments (Luvisols and Planosols) [8]. Results of Minimum Spanning Tree for climate zones (Fig. 2), clearly reflect how soilscapes link in a very well defined latitudinal gradient. References are frequently found in the literature to most characteristic pedogenetic processes of certain environments [3], [11]. However, these canonic types of pedogenesis are not necessarily those underlying soil typological units of greatest land representation. Soilscapes of Meditenanean climates are usually associated to the presence of Luvisols and Vertisols [3], [9]. Cambisols are also frequent [11]. However, Mediterranean soilscapes are very rich and quite diverse for the small surface area covered by this type of climate [8]. Likewise, if the Soil Map of the World at 1:25,000,000 scale [9] is analyzed, different Mediterranean regions in the world can be seen to have quite different soilscapes. They are products of their respective geological and palaeo-environmental histories [8]. In fact, Ve~isols (3.5%) and Luvisols (15.6%) otfly occupy a small area of these territories, with Calcisols (20.3%) and Cambisols (16.2%) being the dominant soil types. Thus, when analyzing global soil patterns, the concept of canonic soil types and/or canonic pedogenetic processes should be supplemented with that of territorially canonic soils types (those with greater territorial representation) [8]. On a World level, despite the small area of the Mediterranean landscapes, their pedodiversity or richness (number of soil typological units -STU- according a FAO keys of 1989) is outstanding. Moreover, after tile mountain soilscapes, the STU density (richness/area unit) of the Mediterranean pedosphere is the greatest [8]. In other words, Mediterranean and mountain climates have a STU density far higher than that of the climatic types of the rest of the world considered by these authors [8]. Therefore these two climatic types, at global level, have a diversity which is not in consonance with the area they occupy. Ibfifiez eta/. [12] also show that soil richness of countries in the Mediterranean Basin are higher than those in temperate and cold environments in central and northern EU countries (Fig. 3). 2.3. Soils and greenhouse effect Climatic changes associated with all increase in CO 2 are comparable to differences in climate that occur locally as a result, for example, of relief and exposure. Unless threshold factors are involved, it is not thought likely that CO2-induced climatic changes will result in a major shift ill the boundalies between soil types. Tile soil contains elements which vary in age from a few weeks (e.g. nutrient concentrations) to tens of thousands of years (e.g. oxic horizons). Many soil characteristics, therefore, reflect the operations of soil-forming processes under a wide range of former
754 climatic conditions; because these conditions differed much more t~om the present climate than the climate predicted for a doubling of CO 2 concentration, these characteristics are not expected to be influenced to a degree that requires consideration [2]. The envisaged changes in climate are likely to have the greatest short-term (50 years) impact through the effect they have on: (a) the salt balance and salt composition of the soil; (b) the chemical precipitation of Ca/Mg carbonates in soil; (c) processes associated with the supply and breakdown of orgalfiC matter [2]. The changes in these processes might not be very significant for the long-term morphological evolution of the soil, but they are undoubtedly important with respect to soil ecology, soil hydrology, soil erosion and land uses. For example, the response of soil aggregates to wetting by rainfall is extremely sensitive to the chemistry of the soil solution and to microbial activity [5]. Slight changes brought about, for example, by longer dry seasons or changing atmospheric deposition may lead to changes in the soil structure which dramatically alter the hydrology and nutrient balance of the soil. For instance, destabilized soil aggregates break down to form surface seals or crusts. When soil surfaces are sealed by chests, the infiltration capacity becomes reduced and overland flow occurs at lower threshold amounts of rainfall. Thus, all increase in aridity would increase the area of soils having unfavourable infiltration characteristics and higher soil erodibility values. In sub-humid and semi-arid climatic zones, a general decrease in precipitation or increase ill evapotranspiration will cause an increase in the area of soils affected by saline or sodic conditions. These changes in salt accumulation could lead to an increase in the area of soils affected by vertic conditions and physical deterioration. 2.4. Soil erosion in the Mediterranean Basin and greenhouse effects In the past, the Mediterranean environment has proved to be sensitive and vulnerable to environmental change, and consequently the landscape in many places has been irreversibly degraded. Currently in tile Mediterranean region, bedrock is exposed over very large areas and most soils are stony. Maximum rates of denudation concur in those semi-arid areas where it rains sufficiently to develop surthce runoff but not to maintain plant cover sufficiently dense to protect the soil [13]. This is the case of semi-arid and arid Mediterranean regions. Erosion is highest in areas receiving between 200-600 mm rainfall per year or with great importance of extreme precipitation events [2, 14, 15]. In sub-humid and semi-arid Mediterranean environments, erosion appears to be closely correlated with rainfall within each area [6], whereas runoff is not significantly correlated with erosion [ 14, 15]. This means that the response of the soil surface to rainfall is of greater importance for achieving erosion than is nmoff alone, suggesting a marked influence of extreme rainthll events again [14]. According to hneson and Emmer [2], an increase in the temperature of 3 to 4 C is argued to have the following impacts on erosion process in the Mediterranean environments: (a) higher rates of erosion on slopes and either relative or absolute increase of overland flow during extreme rainfall events; (b) higher sediment concentration in water runoff; (c) higher sediment and solute supply fiom badlands and/or marl areas; (d) an increase in the risk of rill and gully erosion, (e) an increase in the discontinuity of runoff and sediment transport in drainage basins; (f) lower rates of sediment transport in the major rivers; (g) adjustment of river channels, especially those canting coarse load (braiding will increase and fans will grow); (11) A shift fiom certain perennial streams to ephemeral streams); (i) An increase in wind
755 erosion. 2.5. Desertification and climatic change
At the present time, it would appear that there is a consensus with respect to the fact that the Meditenanean region is the most sensitive in Europe to the climate change which appears to be drawing near [2]. Although the cause of desertification of Mediterranean environments is a debatable matter. Nevertheless, all the evidence seems to point to natural processes strengthening the trend of human ones. Likewise, this region has undergone the most intense, prolonged exploitation known over the whole planet. Simulation models of the atmosphere's general flow predict that a doubling of the greenhouse gas concentration compared to pre-hldustrial periods would bring a mean worldwide temperature increase between 3 and 5.5 C. Current calculations would seem to indicate that this event could occur half way through the twenty-first century. A global change of this magnitude would increase aridity and desertification in the Mediterranean Basin. 3. R E F E R E N C E S 1 S. Rivas-Martinez, Bioclimatic Classification System of the World, 21 a Aproximation,
(lnanusclipt), 1993. 2 A.C. Imeson and I.M. Emmer, Implications of climatic change on land degradation in the Meditenanean. In: L.Jefiic, J.D.Milliman and G.Sestini, Climatic Change and the Mediterranean. E.Amold, 1992. 3 Ph. Duchaufour, Pedologie, 1. Pddogenrse et Classificattion, Masson, Paris, 1977. 4 Soil Suvvey Staff, Soil Taxonomy. A Basic System of Soil Classification for Making and Interpreting Soil Surveys, USDA-SCS Agric. Handbook n ~ 436, Washington, D.C., 1975. 5 A.C.lmeson and R.S. De-Groot, Landscape-Ecological Impact of Climatic Change; Discussion Repport on Meditenanean Region. LICC, VROM-DGMH, Amsterdam, 1989. 6 G. Benito, M. Gutirnez and C. Sancho, Z.Geomorph., 37 (1993) 199. 7 J.J. Ibafiez and R. Jimdnez-Ballesta, An. Edafol. Agrobiol., 45 (1986) 961. 8 J.J. Ib~ifiez, V. Zucarrello, S. De-Alba and A. Garcia-,~dvarez, Catena (Submitted). 9 FAO-UNESCO, World Soil Resources. An Explanatory Note on the FAO World Soil Resources Map at 1:25.000.000 Scale, FAO World Soil Resources Reports 66, Rome, 1991. 10 FAO-UNESCO, Mapa Mundial de Suelos. Leyenda Revisada, FAO, Roma, 1989. 11 E.M. Bridges,World Soils. (2nd edition), Cambridge Univ. Press, Cambridge, 1978. 12 J.J.lbafiez, S. De-Alba and J. Boixadera, The pedodiversity and its measurement: Aplications to soil information system. In: D. Khlg, R.A.J. Jones and A.J. Thomasson (eds), The Development of an EU Land Information System for Agro-Environmental Monitoring. EU, Bnlssels, (in press), 1994. 13 W.B. Langbein and S.A. Schumm, Transactions of the America Geophysicial Union, 39 (1958) 1076. 14 M. Sala, Hillslope nmoff and sediment production in two Mediterranean mountain environments. In: A.C. hneson and M. Sala (eds.), Geomorphology in Enviroments with Strong Seasonal Contrast. Catena Supplement, 12, (1988). 15 H.N. Le-Houdrou, Vegetation and land-use in the Mediterranean Basin by the year 2050: A prospective study. In: L. Jefiic, J.D. Milliman and G. Sestini, Climatic Change and the Mediterranean, E. Arnold. 1992.
756
European Union m e m b e r countries (E: 1/1000000) (Drawn up from INRA & Joint Res. Center, 1992) BB Mediterranean area
10
5
F
SP
I
D
ENG
GR
P
SCOS-IRL DK
NL
B
N-IRL
L
Figure 1. Richness in Major Soils Groups (FAD &UNESCO; 1974) for the soil map of EU countries
MOUNTAINOUS
HUMID TROPICS AND TROPICS
BOREAL
'~ SEASONALLY DRY TROPICS AND SUBTROPICS
COLD
TEMPLATE
~
MEDITERRANEAN
ARID
Figure 2. Minimum spanning tree analysis for climatic zones
ARID
a ,= ,:, , - r ,= ,=, ,=, ,=, , ~ N ,= A ,,,
,
I
TEMPERATE COLD BOREAL MOU SEASONALLY DRY AND SUBTROPICS
NTAI
NOUS
TROPICS
HUMID TROPICS SUBTROPICS
AND
Figure 3. Dendrograms grouping climatic zones with respect to the number of their Major Soils Groups (FAD & UNESCO, 1989)
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
757
A mathematical model for predicting the impact of climate changes on Mediterranean plant landscapes J.L. Gonzfilez-Rebollar, A. Garcia-Alvarez and J.J. Ib~,fiez Centro de Ciencias Medioambientales, CSIC, Serrano 115 bis, 28006 Madrid, Spain
Abstract
The main properties of a phytoclimatic model are explained. It is a technique for theoretic simulation but a field work verification is essential. The methodology involves the transformation of a territory's general phytoclimate data into particular phytoclimate estimates. We are dealing with a simple mathematical climate-soil-relief-vegetation model offering the possibility of predicting changes in vegetation liable to occur at any point in the Iberian peninsula's territory, when the climate data values of a site change as compared with data currently estimated. It consequently enables both alterations and sensitivity of potential plant communities to possible climate changes to be detected and predicted at a certain point in the territory. Using this methodology, it is possible to analyze the repercussions which it is assumed the green house effect would cause according to General Circulation Models (GMCs). A representative case for Le6n province in the NW of the Iberian peninsula is analyzed. 1. INTRODUCTION In addition to the direct influence of the climate on vegetation (the macroclimate) it also exerts an indirect influence via the soil and other environmental factors (the meso/microclimates). The vegetation in the Mediterranean region results from the complex interaction between the climate and other environmental factors. Mathematical models however, cannot accurately represent these interactions and can only identify and simulate general patterns of behaviour. The model presented here has been used in a number of studies to interpret Mediterranean vegetation of the Iberian peninsula [ 1-4]. This study focuses on the variation in vegetation resulting from temperature increases of 1 and 2.5 degrees, guided by predictions resulting from the application of General Circulation Models (GMCs). One must bear in mind however, that edapho-topo-climatic variations in the Mediterranean region are based on a wide range of situations that reflect great diversity and complexity. This study was carried out in Leon province in the NW of the Iberian peninsula. This province is situated on the border between the temperate regions of West Europe (in the N and NW of Spain) and the Mediterranean region. In this study we collected mean monthly precipitation and temperature data from 47 sampling sites distributed over the whole province. This model enabled us to work with different hypotheses that attempt to reproduce distinct geomorphological and edaphic situations. In this case however we have only considered the effects of an increase in temperature.
758 2. M E T H O D O L O G Y The mathematical model used for prediction was described some years ago by one of our team [1] and used to model phytoclimatic-vegetation relationships. The methodology consists of different stages (Fig. 1): 2.1. Phase 1.
The climatic data (precipitation, temperature and evapotranspiration) are transformed to: A)postdictive estimates, B)predictive estimates or C) remain unchanged. This phase allows us to simulate macroclimatic modifications in time in the fight ofpaleoclimatic territorial studies, predictive models or using current data. 2.2. Phase 2.
Transformation of climatic data to phytoclimatic estimates by means of a theoretical simulation of the conditions of the biotope. These are simulations of A) exposure (N or S), B) water loss due to drainage and C) hydric retention in soil. Variations in exposure affect the temperature (T) and the evapotranspiration data; lateral drainage is estimated by means of a runoff coefficient (W), and hydric retention by means of the retention capacity (CR). Monthly precipitation (P), evapotranspiration, runoff" and retention capacity are integrated into the water balance in each location, which allows us to evaluate the monthly availability of water in the soil (D). This evaluation is carried out for each CR,W hypothesis. Data P from each location "i" are thus transformed to estimates D. But D is a parameter that depends on which hypothesis (CK, W) is being considered, there is therefore the same number of monthly estimates of D as there are formulated hypotheses CR, W (Fig. 2). D = DO + P - W P - E
where:
D = Available water P - Precipitation W = Rate of runoff DO = Available water retained CR = Retention capacity E = Evapotranspiration After the water balance, the direct bioclimatic repercussion of the temperature will be evaluated: IBP = f (T - t) where: IBP-- Potential Bioclimatic Intensity T = Temperature t = Minimum vegetative temperature
759
[
DATA BASE
DISCIIIMINANT SIMULATION ANALYSIS ~ DATA " ~ TRANSFORMATION
]
MeteorologicalH ~ ! Potential Variables ] [ ~ '1 'h~l-;-cl)m"-lir
CLIMATE
j[
~_~ 1~ - - ~
~.~
eo,..,,.,
,
~
Figure 1. General scheme of the model
i
I
":
,"~ I
t .,
'.
'>
I '~
:.,
1 i '
~o=
,:,(c r)
D=
Do+P--coP--E
[ w~.TER8ALA.cE!
,,.m
Biologically
15o
o p e n A v a i l a f h a s eIb tDe ~>~ EE> ~> ~E~ >e e water
.
_
2>~E"' r
~
/
~
/Rechargephase
Percolation
/ 1 sO i e c h a Soil moisture storage
cr ] F M A M ]
]
" ~ o A S 0 N D
Figure 2. Basis of the model
rge
Deficit of water
760 Attempts to measure the degree of vegetative activity/inactivity taking into account the double concept of temperature and hydric availability: IBR = Cp * IBP
where:
IBR = Real Bioclimatic Intensity Cp = f (D - e / E - e) e = Minimum vegetative evapotranspiration The general formulas of this hygro-thermic process of bioclimatic estimation can be found in [ 1]. For each hypothesis CR, W, the data P, T and E of a site are transformed into a phytoclimatic vector (F). The vector F obtained from the climatic data of a location "i", using the hypothesis CR = "J" and W - "K", will be called F (i,[J,K]). Likewise we can assume the hypothesis CR, W (eg. CR = Maximum and W = 0) and modify the T values as we have shown here. In this case an increase or decrease in temperature will change the phytoclimatic vector F (in this case F [i, Max CR/W=0]). 2.3. Phase 3. This analyzes the answers of the test. We try to identify the sequential phytoclimatic answers in different temperature hypotheses ( F [i, Max CR/W=0], by means of its degree of similarity (probability) to reference phytoclimatic types: F (Tactual), which are correlated with the potential vegetation Vp. As a consequence, this phase includes two different aspects: - The obtention of reference phytoclimatic types - Identification/classification of hypothetical F (i, Max CRAV=0). We consider that both aspects can be solved using Discriminant Analysis. By means of this analysis, we can establish the zonal phytoclimatic typology in the territory and its Discriminant Functions, and we can refer the rest of the simulated phytoclimates to this typology (mathematically represented by the centroids of the zonalphytoclimatic group). In conclusion: - The methodology involves the transformation of a territory's general phytoclimatic estimates (in time and in space). The transformation is carried out by means of a theoretical simulation of site conditions (precipitation, temperature, geomorphology, soil, etc.) For each location, one must differentiate the theoretical phytoclimatic environment, which is similar to the zonal conditions, from the other estimates, obtained by nonzonal simulation (eg. increase in temperature, from the perspective of climatic change) - The vectors F (i, Max CRJW=0) are grouped according to the type of zonal vegetation to which the location '~ belong. Each group configures a numerical environment "type" (Discriminant analysis). - The intrazonal phytoclimates F (T1, T2 .... Tn) are classified/identified by their similarity to the zonal phytoclimatic typology which is discriminated numerically in the last phase. -
-
=
~
4
~ e
o
9
k a
o
9
Z
9
9
o
o
~
0
II
O,,r
,
~e
"~
@,.,
0
<1
k~
.
761
Q.
E 0
m m
E
m m
0 0
m
k o
m m
0
0
"0 e-
a~
B 0 0 .J
e,i L.
i m
it_
762 3. TEMPERATURE INCREASE AND CLIMATIC CHANGE MEDITERRANEAN REGION. AN ASSAY IN LEON PROVINCE
IN
THE
According to current General Circulation Models (GMCs), predictions of climatic change are expressed as an increase in mean temperature. On a planetary scale this implies a significant change from the present conditions in large areas of the biosphere. Our study however also enables prediction of changes on a regional scale or more detailled level. The vegetation in Leon province has been divided into four phytoclimatic units, each one representing the presence of a characteristic tree species. Two of these are representative of temperate regions (Quercus robur and Fagus sylvatica) and two of Submediterranean and Mediterranean environments (Quercus pyrenaica and Q. rotundifolia respectively). The model can be calibrated by using observations of current vegetation in the feld. Each site can be allotted to one of the previously cited units according to its discriminant fimction obtained from phytoclimatic estimations determined from precipitation (P), temperature (T) and evapotranspiration (E) data. This therefore enabled us to reconstruct a current phytoclimatic map (Fig. 3) representing the phytoclimatic units of the region. This model has been used to predict changes occurring in compliance with hypotheses of temperature increases as a consequence of climatic change. Here, predictive phytoclimatic maps (Fig. 4) have been obtained for a mean temperature increase of 1 or 2.5 C. In the former case, 11 of 47 sites, 23% of the total, experienced an increase in aridity of their phytoclimatic environment. The mean temperature increase of 2.5 C caused an increase in aridity of the phytoclimatic environment in 21 sites (45% of total). This percentage of change is within the range obtained from several general circulation models (eg. Geophysical Fluid Dynamics Lab. (GFDL) or Goddard Institute of Space Studies (GlSS) [5]. 6. REFERENCES 1 J.L. Montero de Burgos and J.L. Gonzfilez-Rebollar, Diagramas Bioclimfiticos, Madrid 1983. 2 J.L. Gonzfilez-Rebollar, Estudios Geogrfificos, 177 (1984) 401 3 J.L. Gonzfilez-Rebollar and J.L. Montero de Burgos, E1 paisaje vegetal a la luz de los modelos fitoclimfiticos. M6todos nuevos para viejas cuestiones, Jaca, Huesca, 1988 4 J.L. Gonzfilez-Rebollar, Monograf. Europ. Cofer. LICC, Lauteren, Holland, 1989 5 D.L. Urban, A.J. Hansen, D. O. Wallin and P.N. Halpin, in Biodiversity and Global Change, Wallinford, UK, 1994
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
763
Vulnerability of Mediterranean ecosystems to Climatic Change, study of soil degradation under different climatological conditions in an altitudinal transect in the south east of Spain C. Boix a, A. Calvob, A. Cerdfi a, A.C. Imesona, M.D.Soriano b and I.R. Tiemessen a aFysisch Geografisch en Bodemkundig Laboratorium. Landscape and Environmental Research Group. Universiteit van Amsterdam. Nieuwe Prinsengracht 130. 1018 VZ. Amsterdam. The Netherlands. bDepartamento de Geografia. Universidad de Valencia. Avenida Blasco Ibafiez 28. 46010 Valencia. Espafia. Abstract To investigate the potential response of soils to climatic change, measurements of soil physical and chemical properties were carried out during a year in a mountain zone in Alicante (Spain), along an altitudinal and climatological gradient. Hydrological properties (infiltration runoff and sediment concentration) were measured under winter and summer conditions. Chemical and physical soil properties were analyzed for reference soil profiles along the transect. The erosional response of the soils as well as soil properties like organic matter and CEC are found to be under the direct influence of the climate, and as a result they have to be considered as important factors in the desertification processes.
I. INTRODUCTION To investigate the relationship of climate to the erosion hazard and soil degradation in the Mediterranean area, a climatological and altitudinal gradient in Alicante (Spain) was selected as a study zone. The criteria for the selection of the zone follow the approach used already by some authors [ 1-3 ] performing soil studies and experiments along climatological and altitudinal gradients, trying to isolate the impact of the Climatic Change on soil erosion. Two slopes (south and north facing) were selected at three sites situated within a zone of only 30 km and showing a high variation in the range of precipitation (Table 1). The lithology is uniform for the three sites: Upper Cretaceous limestone. The study slopes have had an intensive land use in the past: abandoned agricultural terraces appear in Benidorm (BE) and Callosa (CS) (the lowest and intermediate site, respectively) and signs of grazing and forest fires appear in Cocoll (CC) (the highest site in the gradient). The objective of this work is to determine the changes in properties and erosional response of the soil caused by climate. The experimental design of the field work, extensively explained in other papers [3,4], was as follows: Rainfall simulation experiments were carried out on the six slopes in winter and in summer using a rainfall simulator producing rain at 55 mmh-1 of intensity and during one hour. Several soil profiles were taken along each slope and described in the field. Texture, organic matter, the EC and CEC were sampled and analyzed in the
764 laboratory. This paper summarizes some of the findings of this research, that will be described in detail elsewhere. Table 1 Main characteristics of the study zones Location
Lithology
Vegetation series
Aspect Altitude Slope Average (degrees) (meters)(degrees) precipitation
Benidorm
Upper Cretaceous limestone Upper Cretaceous limestone
ChamaeropoRhamnetum lyciodes QuercococciferaePistacietum lentisci Rubiolongifoliae quercetum
N 75 S 210
74-90 74-106
15 20
350 mm
N 10 S 240
280-360 25 282-344 30
550 mm
N 10 S 120
99420 1026 18 850-910
850 mm
Callosa
CocoU
Upper Cretaceous limestone
2. SOIL HYDROLOGICAL PROPERTIES In general, runoff production in summer shows an inverse trend with increasing altitude, a higher runoff is found at the lowest site and a lower runoff at the highest one (Figure 1). 0,60 -
-
0,50 -
1,00
~Rcs
-- 0,90
= c~
~Rcw
-- 0,80 0 , 70
," o "t~ L_
0,60
*"
-- 0,50
o tO
e--
9~ r
0,40-
o r
0,30-
O
'tY
--o- S c s _ _
--a-- S c w .-+.....= .::: . . . . . . . . . .... . . . . . . . . .
...........
0,20-
iiiiiiiiii'iii
0,10 -
!~
0,00 BEs
BEn
CSs
CSn
Fi!iiiiiiiiiiiiiiii
CCs
.
- 0,40
l.==
o
- 0,30
a~"
-
0,20
.E_ "a
-
0,10
09
-
0,00
cD
CCn
Figure 1 Runoff coefficients and sediment concentration in summer (s) and winter (w) (BE: Benidorm, CS: Callosa, CC: Cocoll, n: north slope, s: south slope)
765 In the case of Benidorm, the infiltration capacity is limited by the presence of a crust which generates more runoff and. However, the runoff production in winter does not follow clearly the altitudinal and climatological gradient, as shown also in Figure 1. Callosa, the intermediate site, shows the most runoff in winter due probably to the combined effect of a high soil moisture content at the time of the experiments and a high percentage of stones at soil surface. Looking at the slope aspect, the runoff coefficients and the sediment concentration (understood as an indicators of the soil erosion), were found always to be higher at the south facing slopes, in summer and in winter (Figure 1). South slopes present, in general, more degraded soil conditions and lower values of soil moisture content.
3. CHEMICAL AND PHYSICAL SOIL PROPERTIES The results obtained demonstrate that there are several soil parameters which show a trend following the climatological gradient. At the most arid site (Benidorm) the soils are shallow and poorly developed (lithic Leptosol) while in the more humid areas (Callosa and Cocoll, respectively) better developed soils can be found (lithic Leptosol, haplic Calcisol and chromic Luvisol). The maximum contents in organic matter, clay percentage and CEC values occur at the highest site: upwards along the transect. Considering aspect, the maximum values are always found on the north facing slopes. 40 35 30
45
~OM % .002 mm -=-- CEC
40 35 30
25
25
20
20
"7, O
E
O
15
15
10
10
5
5
0
0 BEn
BEs
CSs
CSn
CCs
CCn
Figure 2. Organic matter, clay content and CEC (BE: Benidorm, CS CaUosa, CC CocoU, n: north slope, s: south slope)
4. DISCUSSION AND CONCLUSIONS The runoff and the sediment concentration values obtained from the rainfall simulation experiments are, in general, very low. But, paying attention to the differences between the three zones, the hazard of erosion of the soils is found to be higher when the climatological conditions become more arid and under dry soil conditions (summer). Under wet soil
766 conditions (winter) the runoff coefficients and the sediment concentration are higher in the intermediate site (Callosa), but this result coincides with the high soil moisture conditions in the soil at the moment of the winter experiments. For the south facing slopes the sediment concentrations and the runoff coefficients decrease upwards along the transect. Among the studied soil properties only the organic matter content, the clay content, and as a consequence of that the CEC, increases upwards along the transect while the electrical conductivity decreases. When the climatological conditions are less favourable (more dry) and the soils less well developed, the rate and the spatial variability, of parameters such as the runoff coefficient, sediment concentration and erosion rate, increases. With more favourable climatological conditions (in the case of Cocoll) the soils show better physical and chemical characteristics, with very low runoff coefficients and sediment concentrations. In the case of Callosa (an intermediate situation), the soils can have high runoff coefficients and sediment concentrations when they are very wet, but lower values when they are dry. In general the soils are more fertile and less erodible when the climatological conditions become more humid. Even on the smaller scale of the difference in micro climate between south and north slope there is found in general a better developed vegetation cover and soil properties as well as less erodible soils on the north slopes. 4. ACKNOWLEDGMENT This work was financially supported by the Commission of the European Communities in the Climatology and Natural Hazards program, EV5V-CT91-0023 ERMES project. Thanks very much to J.M. Schoorl for his valuable help reviewing the English manuscript.
5. REFERENCES
1 Lavee, H., Imeson, A.C., Pariente, S., Benyamini, Y., 1991. The response of soils to simulated rainfall along a climatological gradient in an arid and semi-arid region. Catena 19,1937. 2 Imeson, A.C., Calvo, A., Lavee, H., Perez-Trejo, F., 1993. Modelling and exploring the impact of Climatic change on ecoystem degradation, hydrology and land use along a transect across the Mediterranean. Paper presented in the EC Meeting in Copenhagen, to be published in the Proceedings in 1994. 3 Calvo, A., Soriano Soto, M.D., Boix Fayos, C., Tiemessen, I., 1994: Suelos y procesos geomrrficos en un gradiente climatico altitudinal (Alicante). Actas de la III Reunirn Nacional de Geomorfologia, Logrofio. 4 Soriano, M.D.; Boix, C.; Calvo, A.; Imeson, A.; Cerd~i, A.; Perez-Trejo, F., 1993. Metodologia y disefio del campo experimental en ecosistemas degradados en un transecto altitudinal en la provincia de Alicante, Cuadernos de Geografia, 54, 268-284.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
767
The impact of Climatic Change and land use on the hydrological response of Mediterranean soils; a study along a climatological gradient in Crete (Greece) C. Boix a, A. Calvob, A.C. Imeson a, J.M. Schoorl a, M.D. Soriano b and I.R. Tiemessen a aFysisch Geografisch en Bodemkunding Laboratorium. Landscape and Environmental Research Group. Universiteit van Amsterdam. Nieuwe Prinsengracht 130. 1018 VZ. Amsterdam. The Netherlands. bDepartamento de Geografia. Universidad de Valencia. Avenida Blasco Ibafiez 28. 46010 Valencia. Espafia.
Abstract
To help understand the impact of Climatic Change on the soils of the Mediterranean area, measurements of physical soil properties were carried out in a mountain zone in Crete (Greece), following a climatological gradient. Four experimental slopes were chosen, south facing and situated on limestone lithology. Soil hydrological properties including infiltration, runoff and sediment concentration, were measured and the percentage of waterstable microaggregation in the soil was calculated and used as an indicator of soil degradation. It was found that as well as climate, soil properties were highly affected by the extensive land use of the area, intensive grazing by goats and small scale wildfires.
1. INTRODUCTION One of the approaches proposed by some authors [ 1-4] to evaluate the effects of Climatic Change on soil processes and erosion is to study soil properties and the erosion response of soils along climatological gradients. These climatological gradients have to be carefully selected, joining areas of comparable geology, soils, and vegetation characteristics but having different ranges of precipitation. In this way, it is possible to evaluate the effect of climate on the water-soil-vegetation system. The effect of land use is also important affecting the erosional behaviour of an ecosystem. Following this approach, a climatological and altitudinal gradient was selected in Crete (Greece). This gradient was chosen to study the potential changes generated by changes in climate on soil properties like hydrological and erosional response as well as physical properties. This gradient was one of three gradients selected for study in a transect across the Mediterranean. The Crete transect differed from those in Alicante and Israel with respect to the high density of grazing. Four sites on Upper Cretaceous limestone were selected along this gradient (Table 1). The slopes are used as grazing land for goats, with the more intensive
768 grazing at the highest site (Omalos) and the less intensive grazing at the lowest site (Afrata). The farmers frequently burn patches of shrub vegetation on the slopes to improve the pasture for grazing. In the whole area, old agricultural terraces at the foot slope positions have been abandoned. Rainfall simulation experiments were carried out on the south facing slope of each site using a portable sprinkling rainfall simulator [6]. Rain was applied at an 55 mm h-1 during 55 minutes. At each slope a total of eight simulations on different soil surfaces and land uses categories were carried out. Soil samples were collected from comparable situations to analyze the water stable microaggregation. The waterstable microaggregation and the clay and silt proportion in the fraction <0.105 mm were determined. The methods used are described fully in the cited literature [4]. Table 1 Main characteristics of the four selected slopes (RFS: Rainfall simulation experiments) Location
Altitude Aspect Slope Precipitation (m) (degrees) (degrees) mm 1100 200 25 1000 800 180 35 800 400 175 18 350 100 165 12 200
OMALOS LAKI RODOUPOS AFRATA
Number of RFS 6 8 6 8
2. RESULTS 2.1. Soil hydrological and erosional response A clear decrease in soil erosion and other related parameters, such as the sediment concentration and the sediment yield, with the decreasing altitude and thus a more dry climate was found.
Table 2 Average values of the hydrological response of soils (OM: Omalos; LA: Laki; RO: Rodoupos; AF: Afrata)
OM LA RO AF
Runoff coefficient
Sediment concentra, (gr 1-1)
Sediment yield (gr)
0.66 0.68 0.69 0.37
9.26 1.95 1.72 1.40
61.41 15.74 12.53 6.00
Soil Erosion Bulk Porosity moisture rate density (%) (grm-2h "1) (gr cm -3) content (%) 356.19 97.59 70.99 33.40
0.94 0.90 0.95 0.91
64.45 65.70 64.35 65.80
15.43 17.69 22.41 10.04
Considering the average values for each slope (Table 2), Omalos, the highest and most humid site, shows the highest values of erosion rate, sediment concentration and sediment yield
769 as well as a quite high bulk density. However, the highest runoff coefficient appears in Rodoupos, an intermediate site. In this case the runoff production was clearly influenced by the high soil moisture content at the moment of the experiments (Table 2), also Rodoupos shows the highest bulk density among the studied sites. The lowest and most arid site, Afrata, shows the lowest value of runoff coefficient, erosion rate and sediment yield and concentration. Analyzing the values independently, according to the specific characteristics of the soil surface and land use, bare and burnt plots show the highest runoff coefficients, with the highest erosion rates in bare plots. Abandoned terraces also present high values of runoff and erosion, but taking into account that the gradient of the slope at the terraces is very low, these values are always lower than in bare patches on the slope itself but higher than the vegetated patches on the same slope. 2.2. Soil aggregation
A positive influence of the clay content on the waterstable microaggregation, already pointed out by some authors has been found. Afrata, the lowest site, seems to have the best soil aggregation among the studied soils. The soils in Affata present a more uniform aggregate size distribution for all the measured fractions and the highest value of waterstable microaggregation, besides the highest clay content in the soil. The lowest clay content is found in the Laki soils, an intermediate site, together with a lower value of waterstable microaggregation. Figure 1 shows how the waterstable microaggregation and the clay content follow parallel trends while that in the runoff coefficient is the opposite direction. The waterstable microaggregation is higher in the plots with the highest clay content and the runoff coefficient is higher in plots with lower waterstable microaggregation and clay content. ~,
7O
tO "10
6O
O
~
50
.0
t--
0,9 0,8
0,7 E
t-
o
40
N ~
30
~
20
2 tO E
10
.m O 0,6 iE 0,5 o O
0,4
O
0,3 ,-~
0,2 rv 0,1 0
0 t-~
O
>
O
t-~
~ o
E] Clay %
-,--
0
t'~
>
.s
t'~
-,--,
>
5 3 5 3
t'~
0
.,.,
>
t'~
.t)
~
.Q
>
0
"0
> < o
0
u_
u_
.," , <
o
m Waterstable
9 Runoff coefficient
microaggregation %
Figure 1. Clay content, waterstable microaggregation and runoff coefficients under different plot conditions (OM: Omalos; LA: Laki; RO: Rodoupos; AF: Afrata; b: bare patch; v: vegetated patch; bb: bare and burnt patch; vb: burnt patch with new vegetation; t: abandoned terrace).
770 3. DISCUSSION AND CONCLUSIONS Soil structure and erosion rates seem to be the result of the combination of two main factors in this case: the climatological conditions and the land use. Omalos, the highest site, has nowadays more biomass production due to the very humid climatological conditions. As a consequence of that, the grazing is more intensive and the erosion rates higher. Due to the grazing the soil is trampled and this produces an increase of the bulk density of the top layer. The goats produce also a displacement of material that remains on the soil surface and this material is easily transported by the runoff water and so increasing the erosion rates. Afrata, the most arid site with less biomass production, less grazing, lower values of soil moisture and a high clay content and waterstable microaggregation shows the lowest erosion rate. Specific soil surface conditions produce a different erosional response. Vegetated patches always show lower runoff coefficients and higher values of waterstable microaggregation than burnt and bare patches. Abandoned terraces at the medium sites have erosion rates higher than vegetated patches, but lower than expected, probably due to a high clay content and a low gradient. In some cases, along studied gradients in Spain and Israel [2-4] high values of erosion have been found when the conditions become more arid, but in the case of Crete, the behaviour of the water-soil-vegetation system is clearly dependent on factors derived from the land use.
4. ACKNOWLEDGMENT This work was financially supported by the Commission of the European Communities in the Climatology and Natural Hazards program, EV5V-CT91-0023 ERMES project.
5. REFERENCES
1 Eybergen, F.A. and Imeson, A.C., 1989. Geomorphological processes and climatic change. Catena 16, 306-319. 2 Lavee, H., Imeson, A.C., Pariente, S., Benyamini, Y., 1991. The response of soils to simulated rainfall along a climatological gradient in an arid and semi-arid region. CATENA 19, 19-37. 3 Imeson, A.C., Calvo, A., Lavee, H., Perez-Trejo, F., 1993. Modelling and exploring the impact of Climatic change on ecoystem degradation, hydrology and land use along a transect across the Mediterranean. Paper presented in the EC Meeting in Copenhagen, to be published in the Proceedings in 1994. 4 Calvo, A., Soriano Soto, M.D., Boix Fayos, C., Tiemessen, I., 1994: Suelos y procesos geom6rficos en un gradiente climatico altitudinal (Alicante). Actas de la III Reuni6n Nacional de Geomorfologia, Logrofio. 6 Calvo, A., 1988. Un simulador de lluvia portatil de fiicil construccirn. In: M. Sala & F. Gallart (eds.),. Mrtodos y Trcnicas para la medicirn en el campo de procesos geomorfolrgicos. Monografia n~ 1, Sociedad Espafiola de Geomorfologia. Zaragoza.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
771
Climate change and malaria risk W.J.M. Martens University of Limburg, Department of Mathematics, P.O. Box 616, 6200 MD Maastricht, The Netherlands / National Institute of Public Health and Environmental Protection (RIVM), P.O. Box 1, 3720 BA Bilthoven, The Netherlands Abstract The biological activity and geographic distribution of the malarial parasite and its vector are sensitive to climate influences, especially temperature and precipitation. In this paper the effects of an increase in temperature on the epidemic potential of malaria are explored. Assessment of the potential impact of global climate change on malaria risk suggests a widespread increases of risk due to an expansion of areas suitable for malaria transmission. The health impact will be most pronounced in populations living in the less economically-developed temperate areas in which endemicity is low or absent. 1.
THE ISSUE
The occurrence of vector-borne diseases ranges from the tropics and subtropics to the temperate climate zones. With a few exceptions, vector-borne diseases do not occur in the cold climates of the world. In Table 1 some of the world's most important vector-borne diseases are listed. Table 1:
Global status of the mqjor vector-borne diseases in 1990. ~
disease
vector
populations at risk (millions)
prevalence of infection (millions)
malaria schistosomiasis lymphatic filariasis onchocerciasis
mosquito snail mosquito blackfly
22(X) 600 900 90
270 2(X) 90.2 17.8
The extent to which vector-borne disease transmission occurs in a specific area is determined by two factors: first, the presence of (an abundance of) vectors capable of transmitting the disease, and second the presence of the relevant parasite. Any factor influencing these two determinants hence influences disease transmission. Direct effects of the anticipated changes in global and regional temperature, precipitation, humidity and wind patterns rcsulting from anthropogenic climate change are the factors which have an impact on the vectors' reproduction habits and on their longevity, and are thus associated with changes in annual vector density. In general, the rate of development of a parasite accelerates as the temperature rises. An increase in temperaturc may therefore result in the completion of the life cycle of a parasite in areas in which previous temperatures were too low for the parasite to reach maturity. Indirect effects of climate change include changes in vegetation and agricultural practices which are mainly caused by temperaturc changes and trends in rainfall patterns. Another indirect effect of climate change is associated with the rise in sea level and the resulting coastal flooding. The proliferation of brackish water lagunae influences the availability of habitat and either encourages or discourages vector species depending on whether they prefer brackish water. Generally speaking, drought and dcsertification, including a migration or extension of global desert belts, could be expected to decrease vector-borne disease transmission. It is thus evident that major changes in the incidence of vector-borne diseases associated with a climate change might be expected, and that the manifestation of these changes is closely related to socioeconomic developmcnt and the provision of health services.
772
2.
MALARIA: A GLOBAL PROBLEM
One of the world's most important vector-borne diseases is malaria, and there are few infectious diseases which have as great an impact on the social and economic development of societies. Out of a world population of approximately 5,300 million people in 1990, some 2,200 million are regarded as being at risk of contracting malaria. Roughly 270 million people are actually infected with the malaria parasite. At present, the distribution of malaria is mainly restricted to the tropics and sub-tropics, although before the Second World War malaria was a common disease in many temperate areas of the world. Malaria eradication campaigns and socio-economic development caused malaria to disappear from areas in which it had previously been endemic, although mosquito densities still allow transmission in these areas. 2 The incidence of malaria is determined by various factors: the abundance of Anopheline species, the propensity of the mosquitoes to bite human beings, the longevity of the mosquitoes and the rate at which the Plasmodium parasite in the mosquito develops. 3.
T E M P E R A T U R E AND E P I D E M I C P O T E N T I A L
A unit of measurement which encapsulates many of the important processes in the transmission of infectious diseases is the basic reproduction rate (Ro), defined as the average number of secondary infections produced when a single infected individual is introduced into a potential host population in which each member is susceptible. 3 The basic reproduction rate allows us to calculate the critical density threshold of host populations necessary to maintain parasite transmission. The critical density for malaria transmission can be expressed as:
Nz=k 19 -log____~)
(1)
N,
where Nz/N1 is the number of malarial mosquitoes (N2) per human (N1); p the survival probability of the mosquito; a the frequency of taking human blood meals; n the incubation period of the parasite in the vector. The term kl is a constant, incorporating variables assumed to be temperature independent (including the efficiency with which an infective mosquito infects a susceptible human and an infected human infects a susceptible mosquito, the number of blood meals a mosquito takes from man, and the recovery rate in man). The epidemic potential of malaria is defined as the reciprocal of the host density threshold. This epidemic potential can be used as a comparative index in estimating the effect on the risk of malaria represented by a change in ambient temperature. In Table 2, a number of temperatures which are critical to malarial transmission are set out.
Table 2:
Crucial temperatures in malarial transmission. extrinsic incubation cycle Plasmodium species
digestion of blood-meal Anopheline species
vivax
falciparum
maculipennis
culicifacies
stephensi
degree-days (*C day)
105
111
36.5
29.7
43.4
threshold temperature (*C)
14.5-15 16-19
9.9
12.6
8.9
The most direct effect of temperature is on n. The incubation period of the parasite in the malarial mosquito must have elapsed before the infected vector can transmit the parasite. The relation
773
between ambient temperature and latent period is calculated using a temperature sum as described by MacDonald. a The frequency of feeding depends mainly on the rapidity with which a blood meal is digested, which increases as temperature rises, and can be calculated by means of a thermal temperature sum. 5 The female mosquito has to live long enough for the parasite to complete its development if transmission is to occur. Between certain temperature thresholds, the longevity of a mosquito decreases with rising temperature. The optimal temperature for mosquito survival lies in the 20-25~ range. Temperatures in excess of these will increase mortality and there is a threshold temperature above which death is inevitable. By the same token, there is a minimum temperature below which the mosquito cannot become active. Relying upon data reported by Boyd 6 and Horsfall 7, we assume a daily survival probability of 0.82, 0.90 and 0.04 at temperatures of 9~ , 20 ~ and 40 ~ C, respectively. The epidemic potential is most sensitive to changes in host mortality rates and development time of the parasite. In Figure 1 the influence of increasing temperature on the epidemic potential and the effects of different values of mosquito longevity and minimum temperature requirements for parasite development on the epidemic potential are illustrated. 1.0 reciprocal critical mosquito density
reciprocal critical mosquito density
...... ,,, " * - - ."."~ " -
centralestimate
,," " / / ~ " ' ~ ' ,
,,/,'
,,,/,,,
--
,,\,
,,,\,,
,/,,
0.8-
t
A
17
19
21
23
27
29
31
/
,' / ,',,/ ,, ,' / ,' ,,/ ,, . ' Y
/
"~" ~
i
,",\,, ~
~
\
,
I
\
\~ ~ \l
,'
-,\,
',~, ',~,
',~,
,,\ ',~
,, ,- .. ,. st
-
13
33 35 37 39 temperature (~
/ /
o.,o.~-
/
tt~ ' ~" "
/ ~ "
I
,'/
,,\ 25
/
t
0,,
15
,
i"
0.6-
,,\,,
,< ,."
t
,
,,\',
,'/,,
13
centralestimate
15
17
19
21
23
25
27
29
31
33 35 37 39 temperature (~
B
Figure 1: Epidemic potential (valued as 1 as a maximum) for A) P. vivax (left-hand curve p(20~ = 0.8 and Tmin = 14.5~ central estimate p(20~ = 0.9 and Tmin = 14.5~ fight-hand curve p(20~ = 0.95 and Tmin = 15~ B) P. falciparum (left-hand curve p(20~ = 0.8 and Tmin = 16~ central estimate p(20~ = 0.9 and Tmin = 16~ fight hand curve p(20~ = 0.95 and Tmin = 19~
A high epidemic potential indicates that despite a smaller vector population, or alternatively, a less potent vector population, a given degree of endemicity may be maintained. As temperature increases, epidemic malarial potential increases until a maximum is reached. At high temperatures, the accelerated development of the parasite and the increased biting rate can no longer compensate for the decreasing mean life expectancy among the mosquitoes. The distributions shown in Figure 1 indicate that, in temperate climates, small increases in temperature can result in large increases in epidemic potential, irrespective of the values chosen for the survival probability or minimum temperatures assumed for parasite development. The effect of an increase in temperature will be more pronounced on the epidemic potential of less potent mosquito populations. It should be noted that, although the maximum values for epidemic potential are found in the ranges 29-33~ for malaria, the actual transmission intensity also depends on vector abundance. The optimal temperature for the rapid expansion of a population of malarial mosquitoes is found to lie in the range 20-30~ 8 Therefore, within this range, an increase in mosquito numbers may cause an additional increase of the epidemic potential.
774
4. MALARIA RISK DUE TO CLIMATE CHANGE The concept of the basic reproduction rate, discussed in the previous section, is used in an integrated linked-system model to study the effects of projected changes in temperature and precipitation on malaria epidemic potential next century. 9 Here, some major conclusions of this study are presented. With a global mean temperature increase of approximately 3~ the simulation runs on the model show a projected worldwide increase in transmission potential of the mosquito population and an extension of the areas conducive for malaria transmission. The risk of introduction of malaria transmission in non-malarious areas, including large parts of Australia, the United States, and Southern and Central Europe, associated with imported cases of malaria is a real one, since the former breeding sites of several Anopheles species are still available. Given the fact that in the most developed countries, effective control measures are economically feasible, it is not to be expected that human-induced climate changes would lead to a return of a state of endemicity in these areas. A different situation can be expected in currently endemic areas and areas bordering on them, especially in the subtropics. In the highly endemic malarious areas of tropical Africa, the malaria prevalence may increase. In the malarious areas of lower endemicity, however, the prevalence of infection is far more sensitive to climate changes. Therefore, a human induced climate change may have profound effects on numbers of people suffering from malaria in such areas. In this study, the direct effects of a changing temperature and precipitation on malaria transmission were considered. Additional research on the biological, ecological and socio-economic factors important in malaria transmission will be required for a more complete analysis of the impact of a human-induced climate change on this vector-borne disease. 5.
REFERENCES
1. World Health Organization, Potential Health Effects of Climatic Change, Geneva, Switzerland, 1990. 2. World Health Organization, 'World malaria situation in 1990', World Health Statistics Quarterly, Vol 45, 1992, pp 257-266. 3. R.M. Anderson and R.M. May, Infectious Diseases of Humans: Dynamics and Control, Oxford University Press, New York, USA, 1991. 4. G. MacDonald, The epidemiology and control of malaria, London, Oxford University Press, 1957. 5. T.S. Detinova, Age-Grouping Methods in Diptera of Medical Importance, World Health Organization, Monograph 47, Geneva, Switzerland, 1962. 6. M.F. Boyd (ed.), Malariology, Vol 1, W.B. Saunders Company, Philadelphia, USA, 1949. 7. W.R. Horsfall, Mosquitoes: Their Bionomics and Relation to Disease, Hafner Publishing Company, New York, USA, 1955. 8. L. Molineaux, 'The epidemiology of human malaria as an explanation of its distribution, including some implications for its control', in: W.H. Wernsdorfer and I. McGregor (eds.), Malaria, Principles and Practice of Malariology (volume 2), Churchill Livingstone, New York, USA, 1988, pp 913-998. 9. W.J.M. Martens, J. Rotmans and L.W. Niessen, Climate Change and Malaria Risk: An Integrated Modelling Approach, GLOBO Report Series no. 3, RIVM Report no. 461502003, Bilthoven, The Netherlands, 1994.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
775
Will malaria return to Europe under the greenhouse effect? W. Takken', J. van de Wege' and Th.H. Jetten" "Department of Entomology, Wageningen Agricultural University, P.O. Box 8031, 6700 EH Wageningen, The Netherlands
Abstract Malaria risk is determined by environmental and socio-economic factors. The predicted climate change under the greenhouse effect is likely to affect the epidemic potential of malaria due to a change in vector mosquito phenology and distribution. This effect was simulated using a computer model incorporating mosquito life stages and parasite infections in the mosquito and human host. It was found that both air and water temperature are the most important factors determining mosquito phenology and density. A temperature rise of +4~ shows major changes in mosquito distributions and densities at a worldwide scale, but more so in temperature regions than near the equator. The European situation was taken as an example to study epidemic potential under climate change. Malaria risk, in particular that of P l a s m o d i urn vivax, would increase under climate change. There is little risk for transmission of P. f a l c i p a r u m in currently temperate areas because the local anophelines are refractory to this parasite. In areas adjacent to malaria endemic regions, however, climate change may cause a dramatic shift in P. falciparum risk.
1. INTRODUCTION The predicted climate change as a result of the greenhouse effect is expected to cause major shifts in the distribution and epidemiological risk of vector-borne diseases (1), of which malaria is undoubtedly the most important. At present, some 400 million new cases of malaria appear each year with an estimated one million deaths (2). The world distribution of malaria is to a large extent determined by the geographical distribution of the anopheline vectors. In temperate regions malaria was eradicated due to a judicious use of control methods, which was made possible because of the absence of transmission in the winter months (3). Malaria eradication has failed in the tropics because of technical failures and inadequacies, environmental factors, and low level of socio-economic development. It is feared that under the predicted climate change the risk of malaria might return to areas where it was formerly endemic, as well as to areas adjacent to currently endemic regions. In the present study we investigated the risk of malaria transmission in Europe under climate change, using a simulation model for mosquito population dynamics and vectorial capacity.
776 2. THE MODEL A weather driven, stochastic life-table simulation model (MOSQSIM) was developed to simulate the phenology and population dynamics of mosquito species. The core of the program is the so-called fractional boxcar train. This numerical scheme simulates the development cycle of an entire population during one or more stages. Both development rate as well as dispersion of development rate can be changed during the simulation (4). The model calculates the development of the mosquito population with a daily output of all life stages. Water temperature, affecting development of the aquatic mosquito stages, is calculated via an energy exchange model. Malaria transmission is simulated by allowing fractions of mosquitoes to feed on malaria infectious hosts. 2.1 Meteorology Daily weather data on temperature, humidity and sunshine duration used for case studies were obtained from routine meteorological measurements for locations studied. Monthly averaged minimum, maximum and mean temperature, mean relative humidity, wind speed and sunshine duration were derived from a global climatological database (5). The database is used to define the various climate scenarios due to the greenhouse effect. Temperatures during the day have been simulated using the maximum and minimum air temperature. The temperature in the aquatic habitat is determined by the meteorological conditions (temperature, humidity, radiation, wind) and the characteristics of the habitat itself (depth, horizontal water flow, radiation absorption, vegetation). Estimated surface water temperatures were calculated using a mathematical model based on a model described by Losordo and Piedrahita (6). This model calculates temperature variation and thermal stratification in shallow aquaculture ponds.
3. RESULTS 3.1 Case study From a sensitivity analysis it was found that the temperature of the aquatic habitat plays an important role in larval development and survival. As far as we have been able to ascertain, only one reliable data set of both aquatic and terrestrial stages is available, namely from Aguas de Moura in Portugal (7). The larval and pupal countings reported by these authors were used to verify model computations for An.
777
1000000000
-=.
100000000
~
I0 0 0 0 0 0 0
F~-
1000000
= = k_
. . . . . . . . . . . . ..... ..... cocoa cmoom
S a m p l e s 1939 S a m p l e s 1940 S i m u l a f e d air f e m p e r a t u r e Simulated air+water femperature
-1
L
100000
10000
7. I
~
-i
1000 ~
,,,(/i
''
' ,/
E
!
I00~_ F-
~!
~,,i\%
i~ ~
-I
/ v ~ "
"/
x~
!
,~
--]
]
P '
1o
~
"
!
Aguas
1000000
"
i
120
de
"
'
I
'
!
~
:
'
i
I
240 Time
10000000
I
'
~]
.560
(days)
Moura"
Larval
stage
IV
~. . . . . . . . . . . . . . . . ~ ........... lF c o c o a S i m u l a f e d a i r t e m p e r a f u r e CC ff . . . . . Sim~lo,ed water+air temperature CC
-= -~
ooooo!
j _--
c~
I 0000 F=_-
100 }
\
F-
I
i~ rii~) . . . .
~
I
:
'
'
~ b
i
I
, '
I
'
!
I
' ~_
120 Time Aguas
',
40
l
I
i
I
I
I
560
(days)
de M o u r a "
Adulfs
Fig. 1 - Comparison of simulated and field collected data in Aguas de Moura, Portugal, of (A) larval populations of Anopheles atroparvus and of adult populations. Simulations are shown for air temperature only and air and water temperature combined.
778
atroparvus. Mean monthly maximum temperatures, minimum temperatures and relative humidities for Aguas de Moura during the period 1937-1940 were derived from (8). The simulations have the same pattern as the observations (Fig. 1). After two generations the number of emerging adults starts to decrease due to an increased mortality of the aquatic stages. At the same time adult mortality increases due to higher vapour pressure deficits. At the end of August the number of female adults reaches a minimum value and starts to increase again till November. A first peak of adults in May-June and a second peak in September-November was typical for southern Europe. 3.2 Distribution of malaria vectors in Europe
The distribution limits of the various species of the An. maculipennis complex are determined by environmental conditions. Mortality is closely related to aquatic and ambient temperatures. The model parameters found in literature have been used to calculate the population dynamics of members of the An. maculipennis complex over Europe using climate data. Incorporation of the water surface model (see above) enabled us to simulate the distribution of the various species. Results from simulations, using the weather data of the various stations in the Miiller climate-database, predicted where in Europe malaria was endemic. The model had a good fit with historic data from malaria endemic areas in Europe. Based on these simulations it is possible to use the MOSQSIM model to predict the effect of climate change on malaria. 3.3 European distribution An. atroparvus
An. atroparvus is the most widespread malaria vector in Europe. Under increasing temperatures (+4~ the model predicts an increase in adult mosquito densities in northern Europe but a decrease in southern Europe. At the southern distribution limit of An. atroparvus the number of adults sharply decreased. The highest densities were calculated for central Europe. Under increasing temperatures the model predicts an 100 fold or more increase in infectious mosquito densities in central Europe but a decrease in south-eastern Europe. 3.4 Epidemic potential
A comparative index to estimate malaria risk was derived from the vectorial capacity and is expressed as epidemic potential. This is defined as the reciproke of the critical density threshold of mosquitoes resulting in more than one new potentially infective contact per infectious person per unit time. Simulated global changes of yearly mean epidemic potential for Plasmodium falciparum caused by a temperature increase of 4~ show a projected worldwide increase and an extension of the areas conducive for malaria transmission as climate changes. The highest changes are seen at the northern and southern distribution limits of P. falciparum and at higher altitudes within malarious areas (9). Particularly in temperate zones where malaria was once widespread, epidemic potential is likely to increase due to a increase of
779
Cur. C l i . .
9
-." . "
. ..." 9 -.. +ZK:
i
X
.x
< .00111 < .0111
9 x
<
.01
x
<
.1
>----
.1
i II
~
x~ x,x xx X X
X
~x
..L
I
Fig. 2 - The impact of a temperature increase on the simulated epidemic potential of P. vivax malaria in Europe for September. Cur.Cli. = Current Climate; +2K = temperature increase of 2~ +4K = temperature increase of 4~
parasite development rate in the mosquito, mosquito longevity and biting rate. This concerns large parts of the world including Europe, the Asiatic part of Russia and the continental USA. In these regions potential malaria transmission will largely be due to P. vivax because many anopheline species of these regions are refractory to P. falciparum. The model predicts that under a temperature rise of +4~ epidemic potential of P. vivax malaria increases 10 fold in southern Europe and 100 fold further north (Fig. 2).
4. DISCUSSION In the present study it is shown that the population dynamics and distribution of European malaria vectors can be simulated accurately provided the models incorporate environmental data from a large number of geographical locations. The significant difference between surface water temperature and ambient temperature as derived from this study, shows that the impact of climate change can affect insect populations in two ways. First, higher temperatures may cause increased mortality of aquatic stages in species adapted to the Mediterranean climate and second, higher temperatures lead to accelerated development and extra generations in the more northern vector species without the observed mortality in the aquatic stages. The study clearly
780 shows that the winter temperatures in Europe were the limiting factors for malaria transmission of the past. Under the predicted climate change, with higher annual temperatures, there will probably be a shift of anopheline distribution in Europe, with the southern species moving further north, and the northern species extending their phenological duration and density. For these reasons the risk of malaria transmission will increase provided infectious individuals are present in the human population. Endemic malaria has been eradicated from Europe since the nineteen seventies, and in the absence of a Plasmodium reservoir, we do not expect a return to a state of endemic malaria due to an increasing potential transmission intensity caused by climate change. The high level of socio-economic development in Europe, in particular the health care system and current animal husbandry, will prevent a reintroduction of endemic malaria. However, large numbers of imported cases of malaria are being registered in Europe each year due to the increased travel to and from endemic countries. Therefore, with the densities and distribution of European anopheline species increasing, there is an increased risk of incidental P. vivax cases, especially in those areas in south-eastern Europe where socio-economic conditions have deteriorated. The present study focused on the impact of climate change on malaria vectors in Europe. Preliminary investigations on a global level demonstrated that the models can be used to predict malaria risk and vector distribution on a larger scale (9; 10). Such studies will be especially useful to assess the risk of malaria (and other vector borne diseases) under climate change in areas that are bordering endemic regions and to predict the potential shift in transmission risk in endemic areas. Such studies are needed in order to prepare for the consequences of climate change at the socioeconomic level, particularly in developing countries.
5. REFERENCES
1
World Health Organization, Potential health effects of climatic change, report of a WHO task group. Geneva, 1990. 2 G.T. Strickland and S.L. Hoffman, Sci. Am. (Science and Medicine) 1 (1994) 24. 3 L.J. Bruce-Chwatt and J. de Zulueta. The rise and fall of malaria in europe. Butler and Tanner Ltd., London, 1980. 4 J. Goudriaan, pp 453-473 in J.A.J. Metz and O. Diekman (eds.), The dynamics of physiologically structured populations 68 (1986). 5 M.J. Miiller, Selected climatic data for a global set of standard stations for vegetation science, W. Junk Publishers, The Hague, 1982. 6 T.M. Losordo and R.H. Piedrahita, Ecol. Mod. 54 (1991) 189. 7 F.J.C. Cambournac and J.M. Simoes, Separata dos Anais do Instituto de Medicine Tropical (Lisbon) 1 (1944) 229. 8 F.J.C. Cambournac, Sobre a epidemiologia do sezonismo en Portugal. Sociedade Industrial de Tipografia, Lisboa, 1942. 9 Th.H. Jetten and W. Takken, Change 18 (1993) 10. 10 Th.H. Jetten, W.J.M. Martens and W. Takken, J. Med. Entomol. (submitted).
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
781
A S S E S S M E N T R E P O R T ON N R P S U B T H E M E
"IMPACT OF C L I M A T E C H A N G E ON T H E W A D D E N SEA"
W.J. Wolff Institute of Forestry and Nature Research (IBN-DLO) P.O.Box 23 6700 AA Wageningen The Netherlands
With contributions by: A.G. Brinkman, K.S. Dijkema, B.J. Ens C.J. Smit, G. Wintermans E.J. Houwing, N. Dankers J.A.J. Terwindt
J.J. Beukema, P.J.C. Honkoop
IBN/DLO, Institute of Forestry and Nature Research, Texel IMAU/UU, Institute for Marine and Atmospheric Research/ University of Utrecht NIOZ, Netherlands Institute for Sea Research, Texel
F. Gerritsen, T. Louters, J.P.M. Mulder
RIKZ, National Institute for Coastal and Marine Management, The Hague
G.M. Lenssen, J. Rozema
VUA, Free University of Amsterdam
782
Contents Abstract 1.
Introduction
2.
S h o r t description of the Wadden Sea
E x p e c t e d effects of climate c h a n g e on the Wadden Sea and c h o i c e of r e s e a r c h subjects
0
3.1 3.2 3.3 3.4 3.5 0
Impact of increased CO2 concentrations 4.1 4.2
0
5.3 5.4
Impact of sealevel rise on the morphology of tidal fiat environments Impact of sealevel rise on erosion, sedimentation and plant dynamics in salt m a r s h e s Impact of sealevel rise on benthic animals of tidal fiats Impact of sealevel rise on shorebirds
Impact of c h a n g i n g water t e m p e r a t u r e s 6.1 6.2 6.3
0
Impact on s a l t m a r s h plants Evaluation
Impact of sealevel rise 5.1 5.2
0
Expected effects of an increase of C02 Effects of changes in t e m p e r a t u r e and other meteorological factors Effects of an increase rate of sealevel rise Effects of increased UV-B radiation Integration of effects
The effects of winter t e m p e r a t u r e on the reproductive success on some bivalves in the Dutch Wadden Sea Experimental tests of the effects of winter t e m p e r a t u r e on reproductive success of bivalves Evaluation
Integration 7.1 7.2 7.3
Integration of effects of climate changes on estuarine ecosystems Results of some scenario studies Evaluation
8.
Evaluation of the Wadden Sea studies of N R P I
9.
References
783 ABSTRACT This section summarizes studies on the effect of climate change on the estuarine Dutch Wadden Sea. Increased concentration of carbon dioxide is expected to have a r a t h e r small effect on s a l t m a r s h vegetation and a negligible effect on the functioning of the Wadden Sea ecosystem. Rates of sealevel rise of 60 and 85 cm per century are not expected to have a major impact on the geomorphology of tidal basins because of increased sedimentation. Also saltmarsh accretion may be able to keep pace with rates of sealevel rise of this magnitude, if erosion at the seaward edge can be controlled. Shorebird population sizes, which appear related to the area of tidal flats in their winter quarters, are not expected to decline because of change in the area of tidal flats in the Wadden Sea. However, because reproduction of several species of bivalve shellfish m a y be i m p a i r e d by h i g h e r w i n t e r temperatures, recruitment of these species may strongly decrease and this m a y affect shorebird populations. Model calculations do not show a major effect of higher winter temperatures on ecosystem functioning, however.
1.
INTRODUCTION
The Wadden Sea is a shallow tidal sea of about 8 000 km2 situated along the coasts of The Netherlands, Germany and Denmark. The Dutch part of the Wadden Sea is considered to be the most important nature area of The Netherlands. It has received recognization as such in the Physical Planning Decision (PKB) for the Waddenzee, about 2 000 km2 of it are protected under the N a t u r e Conservation Act, together with the Danish and German parts it has been given the status of 'Wetland of international importance' under the Convention of Ramsar, and it is recognized as a Biosphere Reserve by UNESCO. The Wadden Sea is one of the best studied wetlands of the world. Boekschoten (1973) estimated that, at that time, already 4,150 scientific publications described this area. This number has increased considerably since then. Hence the structure and functioning of the 'normal' Wadden Sea system are well known. Beukema et al. (1990) arrived at some preliminary conclusions about the possible effects of climate change on the Wadden Sea and other coastal areas. Table 1.1 lists the studies commissioned by the Dutch National Research P r o g r a m m e on Global Air Pollution and Climate Change (NRP I) on the effects of climate change on the Wadden Sea as well as some other relevant studies. In addition research on the effects of climate change has been carried out in the German Wadden Sea. This report aims at an assessment of the results of studies on the Wadden Sea commissioned by NRP I, taking in account any other relevant study.
784 Table 1.1 List of projects in the NRP subtheme "impact of climate change on the Wadden Sea" Title
Project leader
Number
Effects of an increased sealevel rise on geomorphology and ecological functioning of the Wadden Sea
T. Louters
850011
Salt marshes and sealevel rise: plant dynamics in relation to accretion processes and accretion enhancement techniques
J.H.J. Terwindt
850033
Effects of climate change on bird migration strategies
C.J. Smit
850034
Winter temperature and reproductive succes in bivalves living on tidal flats in Western Europe
J.J. Beukema
851053
Integration of effects of climate change on estuarine ecosystem communities
A.G. Brinkman
853127
Non-NRP funding Effects of atmospheric CO 2 enrichment, salinity and flooding on the ecology of C3 and C4 saltmarsh plants
J. Rozema/ G.M. Lenssen
Subsidence of a coastal area due to gas extraction as a model for sealevel rise
N. Dankers/ K.S. Dijkema
2.
S H O R T D E S C R I P T I O N OF T H E W A D D E N S E A
The Wadden Sea is a shallow estuarine area situated along the North Sea coasts of Denmark, Germany and The Netherlands. Average tidal ranges vary between 1.4 and 3.4 m. The Wadden Sea occupies about 8,000 km2 and about half of this area consists of bare tidal flats. Salt marshes occur only above mean high w a t e r m a r k and cover about 300 km2. The Dutch part of the Wadden Sea covers about 2500 km2. The Wadden Sea is considered to be one of the major n a t u r e areas of w e s t e r n Europe. Its tidal landscapes where n a t u r a l processes such as erosion and accretion are visibly active, seem relatively unaffected by humans. The extensive tidal flats contain extremely large numbers of benthic invertebrates and are characterized by high biomasses. Also the subtidal areas are rich in individuals and biomass. On this invertebrate biomass fish and birds feed in very large numbers. Many North Sea fish and crustacean species use the Wadden Sea as a nursery.
785 The bird population includes both breeding birds, such as gulls, terns, and several species of shorebirds, and non-breeding migratory species. The l a t t e r species mainly breed in the Arctic and visit the Wadden Sea as a stopover site during migration or as a wintering site. Non-breeding migratory species include geese, ducks, m a n y species of limicoles, and several other species. It is estimated t h a t altogether about 7 million shorebirds visit the Wadden Sea annually. Wolff (1983) gives an extensive description of the ecology of the Wadden Sea. The past ecological development of the Wadden Sea has been summarized in two review papers by Wolff (1992a) and Wolff et al. (1994). His conclusions are summarized below. In a completely n a t u r a l situation a Wadden Sea landscape would show the following sequence of belts of different landscape types when going from the North Sea to the higher inland areas (Figure 2.1): beaches, dunes and salt marshes on barrier islands tidal flats and channels salt and brackish marshes freshwater marshes, swamps, and peat bogs high-lying soils with forests and other dry vegetation types. It may be assumed t h a t these belts have moved shoreward under the influence of the rising sealevel in the geological past (Zagwijn, 1986). An increased r a t e of sealevel rise due to climate change would in principle have the same effect. Until about 1000 years ago m a n hardly interfered with the geomorphological development of the Wadden Sea. At t h a t time, however, the invention of dikebuilding introduced a major change (Wolff, 1992b). At the end of the 12th century the larger part of the vegetated landscape was separated from the tidal Wadden Sea by a contiguous system of dikes. Thus the connection between the Wadden Sea and the adjacent belt of brackish and fresh wetlands was broken. The result of all embankments and reclamations is a very marked and firm separation between the tidal Wadden Sea on the one hand and the non-tidal land and freshwater areas on the other hand. This hard boundary will prevent further shoreward movement of the coastal ecosystems e n u m e r a t e d above, which leads to the question w h e t h e r these systems can survive in their present position. FLATS .
.
.
.
.
.
.
.
,:,:,:-.,:.- . . . .
I BARRIER ISLAND .
.
.
.
.
.
SALT MARSH FRESH PEAT HILLSI
DIKE .
.
.
.
TIDAL FLATS
POLDERS
HILLSI
.
............................................................................................. ~:................................. ~ ~i~i~i~
Figure
DIKE
~i~~i:i:i
Lw
9,.1
Cross-section t h r o u g h the Wadden Sea landscape in the Middle Ages and at present
786 Many species are known to have been introduced from abroad into the Wadden Sea area by h u m a n intervention. Most species were inadvertently introduced; some like the cord grass Spartina townsendii and the Japanese oyster Crassostrea gigas deliberately for, respectively, stimulation of accretion and oyster farming. Other new species were able to colonize the Wadden Sea because their habitat was introduced there. The best example is the construction of artificial rocky shores in the form of b r e a k w a t e r s , moles, and stone-covered dike slopes. It m a y be concluded that the Wadden Sea nowadays lodges quite a number of species of algae and invertebrates which did not occur there before. Most of these species have in common t h a t they are relatively small and short-living and that they have short reproduction cycles. They did not change the ecosystems very much. With respect to climate change it may be wondered whether any further immigrations due to an amelioration of the climate will have equally little effect. On the other hand, h u m a n activities have had such a strong impact on the Wadden Sea ecosystem, that several species have become extinct. Wolff (1992a) lists three species of marine mammals, six species of birds, seven species of fish, and three species of molluscs. Climate change might lead to f u r t h e r local extinctions because species cannot cope with the changed conditions. The following general conclusions may be drawn on the past changes of the Wadden Sea (see also Wolff 1992a; Wolff et al., 1994): in a seemingly natural area h u m a n s appear to have had a very strong impact on the ecosystems; already in the Middle Ages about half of the ecosystems had been destroyed; m a n y large, long-living and slowly reproducing species (K-selected species) have disappeared from the Wadden Sea, whereas many small, short-living and rapidly reproducing species (r-selected species) have colonized the area or increased in numbers. At present the Wadden Sea is managed as a nature reserve or national park, but at the same time several h u m a n activities are still possible. Among these are fisheries for fish, shrimps, and shellfish, mussel culture, extraction of shells, sand, and n a t u r a l gas, sailing and other forms of water sport, military exercises, and grazing of the salt marshes. Moreover, the Wadden Sea is strongly influenced by land drainage containing nutrients and toxic substances. Especially the influence of the river Rhine is quite important in the Dutch part of the Wadden Sea.
0
E X P E C T E D E F F E C T S OF CLIMATE C H A N G E O N T H E W A D D E N S E A AND C H O I C E OF R E S E A R C H S U B J E C T S
In November 1988 a workshop on the expected effects of climate change on marine coastal ecosystems was held at the island of Texel (Beukema et al., 1990). In several contributions the Wadden Sea served as a case study and m a n y of the conclusions of the workshop are applicable to the Wadden Sea. 3.1 E x p e c t e d effects of an i n c r e a s e of CO2 In the workshop cited above Long (1990) and Rozema et al. (1990) showed t h a t elevated CO2-concentrations have a clear impact on s a l t m a r s h plants. This observation has since been studied in the project 'Effects of atmospheric CO2 enrichment, salinity and flooding on the ecology of C3 and C 4 salt m a r s h plants' by
787 J. Rozema and G.M. Lenssen (Table 1.1). CO2-effects on the aquatic ecosystems were considered to be less important (Brouns, 1988) and consequently have not been a subject for further study. 3.2 E f f e c t s of c h a n g e s in t e m p e r a t u r e a n d o t h e r m e t e o r o l o g i c a l f a c t o r s In the 1988 workshop t e m p e r a t u r e and related climate factors were clearly identified as a factor which might strongly influence the distribution of plant and animal species (Van den Hoek et al., 1990; Breeman, 1990; De Vooys, 1990; Beukema, 1990; Wilson, 1990; Costa, 1990). Changed meteorological conditions on one hand will lead to the local extinction of species for which the habitat conditions become unsuitable, and on the other hand will enable other species to colonize new areas because the conditions have become suitable for them. These conclusions have not been followed up by detailed studies since it was reasoned that comparison of the Wadden Sea with estuarine areas more southerly would give enough information on the expected changes of occurrence of plant and animal species (De Vooys, 1990; Costa, 1990). Less attention was paid to shifts in the abundance of species due to changed conditions. However, one possible shift was singled out because of its possible importance for the Dutch shellfish cultures. This concerns the possibility t h a t a relatively small change of w a t e r t e m p e r a t u r e s might negatively affect the reproduction and subsequent survival of the larvae of a n u m b e r of shellfish species. This has been studied in greater detail in the project 'Winter temperature and reproductive succes in bivalves living on tidal fiats in Western Europe' by J.J. Beukema and P.J.C. Honkoop (Table 1.1). However, this study will continue for at least a year after the conclusion of NRP I, so definite conclusions are not yet possible. 3.3 E f f e c t s of an i n c r e a s e d r a t e of s e a l e v e l rise In the 1988 workshop an increased rate of sealevel rise was recognized as a factor with potentially great effects on the geomorphology and the ecology of the Wadden Sea and similar areas (Siefert, 1990; Misdorp et al., 1990; Westerhoff and Cleveringa, 1990; Lefeuvre, 1990; Huiskes, 1990; Dijkema et al., 1990; GossCustard et al., 1990). Salt marshes and tidal fiats could be inundated and this in turn could lead to impacts on the plant and animal species occurring. Consequently several projects addressed this possible problem (Table 1.1). The potential effects on the geomorphology and benthic fauna of tidal fiats were studied in the project 'Effects of an increased sealevel rise on geomorphology and ecological functioning of the Wadden Sea' by J.P.M. Mulder and T. Louters. The possible effects on the geomorphology and vegetation of salt marshes were investigated in the study 'Salt marshes and sealevel rise: plant dynamics in relation to accretion processes and accretion e n h a n c e m e n t techniques' by J.H.J. Terwindt, K.S. Dijkema and E.J. Houwing. The consequences of the possible inundation of tidal flats for shorebird populations were the subject of the study 'Effects of climate change on bird migration strategies' by C.J. Stair, B.J. Ens and G. Wintermans. In addition a non-NRP study on 'Subsidence of a coastal area due to gas extraction as a model for sealevel rise' by N. Dankers and K.S. Dijkema could be used as an experimental analogue of an increased rate of sealevel rise.
788 3.4 Effects of i n c r e a s e d UV-B radiation At the 1988 workshop UV-B was identified as another factor which might have consequences on the Wadden Sea ecosystem (Kramer, 1990; Van de Staay et al., 1990). In the latter study effects on saltmarsh plants were clearly demonstrated. The former study, however, concluded that increased UV-B radiation would have little effect on the biota of the turbid and turbulent Dutch coastal waters. An exception was made, however, for the organisms of tidal flats and shallow pools in the tidal zone during low tide. In view of this conclusion no further studies were commissioned. However, Peletier et al. (submitted) have shown since that benthic diatoms occurring on the Wadden Sea tidal flats show marked changes of abundance under the influence of increased UV-B. 3.5 I n t e g r a t i o n of effects Already in the 1988 workshop it became clear that many biota could become subject to the influence of several aspects of climate change at the same time (e.g. Van de Staay et al., 1990). To address this aspect of the impact of climate change the study 'Integration of effects of climate change on estuarine ecosystem communities' was started by A.G. Brinkman. Based on an already developed simulation model of the ecosystem of the Wadden Sea the various aspects of climate change had to be investigated simultaneously. 4.
IMPACT OF INCREASED CO2 CONCENTRATIONS
4.1 I m p a c t on s a l t m a r s h plants CO2 is essential for the growth of terrestrial and some aquatic plants. Other aquatic plant species are dependent on ions derived from dissolved carbon dioxide. In the photosynthetic process CO2 is chemically reduced to carbohydrates and part of these become plant material. How plants respond to increased CO2 concentrations depends on the photosynthetic pathway of the plant in question. In so-called C3 plants the C3 photosynthetic pathway will result in an increased photosynthetic rate and higher biomass production under higher levels of carbon dioxide. In C4 plants these effects are small or absent (Strain and Cure, 1985; Rozema et al., 1993). In Wadden Sea salt marshes both C3 and C4 plants occur. Increasing CO2 concentrations therefore may result in different rates of biomass production and, hence, in changed competitive abilities. This could lead to changes in the composition of saltmarsh vegetation. An effect on the amount of dead plant material exported from the marsh is also possible.
Experimental CO2 enrichment studies in salt marshes with C3 and C4 species in the Chesapeake Bay, USA, analysed the competitive relationships between the two types of plants (Arp, 1991). The CO2 experiments were done in the field for five continuous years. The C3 species Scirpus olneyi showed increased biomass under elevated atmospheric CO2, in contrast with the C4 grass Spartina patens. This long-term field research was one of the first experimental studies providing evidence that the competitive balance between C3 and C4 plants will shift in favour of the former ones. At the end of the five-year research period the C3 plants
789 demonstrated the same increased rates of growth and photosynthesis as at the start. This implies t h a t under field conditions these plant species do not show photosynthetic acclimation, that is a down regulation of the rate of photosynthesis as a result of end product negative feedback. In The Netherlands Lenssen (1993) investigated the response of three C3 species (Aster tripolium, Elymus athericus, Puccinellia maritima) and one C4 species (Spartina anglica)to CO2 enrichment in the project 'Effects of atmospheric CO2 enrichment, salinity and flooding on the ecology of C3 and C4 salt m a r s h plants' (Table 1.1). In addition he investigated whether the response to CO2 enrichment was modified by other environmental factors, viz. light, t e m p e r a t u r e , UV-B radiation, salinity, and flooding. In his experiments Lenssen (1993) found t h a t at a CO2 concentration of 720 ~tmol.mol-1 the C3 plants showed an increase of plant biomass of 19-33% relative to the ambient CO2 concentration. The only C 4 plant, however, was not stimulated by an higher concentration. These results confirm the earlier results of Curtis et al. (1990) and Arp (1991) for American salt marshes. However, the outcome of the experiments appears to be dependent on the other environmental factors studied. Higher temperatures as a result of climate change probably will favour the C4 species more. Also the other environmental factors influence the effect of increased CO2, but not in correlation with the C3- C4 separation. It may be concluded that higher atmospheric CO2 concentrations will result in higher biomass production of the C3 species and a loss of competitive ability of the C4 species. 4.2 E v a l u a t i o n Changes in competitive ability of saltmarsh plants will probably be reflected in a changed composition of the vegetation on Wadden Sea salt marshes. The higher biomass production probably will have an effect on the exchange of plant organic m a t t e r between saltmarsh and adjacent estuary, but in view of the magnitude of this exchange (Dankers et al., 1984) and the relatively small area of Wadden Sea salt marshes (see section 2) the effect on the Wadden Sea system will be negligible. In view of the conclusion drawn in Section 3.1 about the relative unimportance of CO2-effects below high-tide level, it is concluded here that no further studies on carbon dioxide effects are needed for the Wadden Sea. 5.
I M P A C T OF S E A L E V E L RISE
5.1 I m p a c t of s e a l e v e l r i s e o n t h e m o r p h o l o g y of tidal fiat e n v i r o n m e n t s
A geological evolutionary model of tidal basin development Van der Spek and Beets (1992) studied the evolution of a Wadden-Sea like tidal basin, the so-called Holland tidal basin, between 7000 BP and 3500 BP. Their results have been used to derive a geological model for tidal basin evolution under the influence of a rise in sealevel. The model emphasizes the balance between the extra storage capacity of the basin created by a sealevel rise, and the amount of sediment available for a geomorphological response. The model shows t h a t if the rate of sealevel rise exceeds the rate of sediment supply, the innermost parts of the basin will not receive sufficient sediment for an intertidal morphology to be
790 preserved. Eventually, sand will be deposited only in tidal channels and in the flood tidal delta through which the sediment is supplied; mud deposition will occur in the interchannel areas, and salt marshes will disappear.
An e m p i r i c a l model of the morphological behaviour a n d s t a b i l i t y of channels and fiats in tidal basins D e p a r t i n g from the geological model Eysink (1992) has defined an empirical morphological model MORRES (acronym for MORphological RESponse model) which is b a s e d on a set of empirical equilibrium r e l a t i o n s h i p s b e t w e e n hydrodynamic and geomorphological characteristics of tidal basins. This model is a sediment balance model describing the long-term geomorphological development of the Wadden Sea as a result of increased sealevel rise on the scale of a tidal basin. It covers the geographical units outer delta, inlet and basin. The outer delta is characterized by its sand volume. The inlet is characterized by a cross-sectional area and its length. The tidal basin is characterized by the curve representing the hypsometry. The area and elevation of the tidal flats, defined as the area between m e a n high w a t e r (MHW) and mean low water (MLW), and the depth of the basin are derived from the hypsometric curve. The basin is defined as the surface area at high water. Furthermore, the water motion in the basin is characterized by the tidal prism of the basin. The tidal range in the basin is assumed to remain constant during the morphological adaptation process. The exchange of sediment between the tidal basin and the adjacent sea is based on the principle of a sand trap. An (accelerated) rise in sealevel will cause a regional rise of the average water level in the Wadden Sea. In other words, the depth of the Wadden Sea will increase somewhat. Hence, the dynamic equilibrium will be slightly disturbed. This slight depth increase will cause a slight deceleration of the average current speeds in the channels and over the flats. Since the capacity for sediment transport is a power function of the current speed, the sediment t r a n s p o r t capacity will drop much more as the current speed slows. The flood stream carrying sediment will continue to deposit sediment in the basin. However, the ebb current does not have enough force to lift and remove the total quantity brought in. Thus, over a longer period, this creates a net sand t r a n s p o r t towards the tidal basin. This p r o p e r t y of deepened tidal basins to demand large quantities of sand is termed sand demand or 'sand hunger'. The total quantity of sand required to restore dynamic equilibrium is directly proportional to the depth increase. Hence, sealevel rise will result in filling in of the tidal basin until a new equilibrium is reached. The sand will be derived from the outer deltas of the inlets and eventually from the coasts of the barrier islands. Other changes in the morphology of a tidal basin, e.g. sand extraction, subsidence of the seafloor due to gas extraction, reduction of the tidal basin due to engineering works or accretion of salt marshes, will cause similar reactions of the tidal basin. To predict the behaviour of the Wadden Sea under conditions of sealevel rise on the basis of the empirical r e l a t i o n s h i p s b e t w e e n a n u m b e r of morphological characteristics, it has been assumed t h a t the system will strive for equilibrium. This assumption can be validated by the observations made after two tidal inlets have been changed drastically by reducing their tidal area through the building of dams (1932: Zuiderzee; 1969: Lauwerszee). In both cases the system shows a development towards a new equilibrium.
791 The knowledge of the processes governing the morphology of the tidal flats is rather limited. Observations have shown that, generally speaking, the present rate of flat growth can keep up with the current rise in sealevel of about 20 cm per century. The m a x i m u m r a t e s of increase of tidal flat level derived from measurements in the period 1925-1987, appear to be around 8-13 mm annually. It seems t h a t the tidal flat system has the capacity to compensate for a wide range of rates of rise in sealevel by raising its level. It remains to be seen whether the flats can keep up this growth rate in an increasingly rough wave climate at an accelerated rise in sealevel. This requires more knowledge of the process. To make a prediction of the effects of sealevel rise on the Wadden Sea tidal flats, Louters and Gerritsen (1994)used three different rates of sealevel rise: (1)present rate of 20 cm per century, (2) a predicted rate of 60 cm per century, and (3) a worst-case scenario of 85 cm per century. In addition the effects of other h u m a n interventions, such as sand and shell extraction, and subsidence due to gas extraction, are taken into account.
Consequences o f some likely scenarios o f sealevel rise for the t i d a l fiats and channels of the Wadden Sea The Wadden Sea was created by the rising sealevel. Should this rise slow down or cease in combination with inflow of sediment, the Wadden Sea will silt up. If, on the other hand, the sealevel rises too fast, or too fast in proportion to the inflow of sediment, the tidal flats will become inundated. The future of the Wadden Sea depends on the balance between the supply and the demand of sediment, both of which nowadays are largely under the influence of mankind. On a regional and local scale, the d e m a n d for sediment is partly determined by the effects of mineral extraction (gas, sand and shells) and by the size reduction of the basin caused by e m b a n k m e n t and reclamation projects. Table 5.1 shows t h a t under rates of sealevel rise of 60 and 85 cm per century the Wadden Sea system is still able to track sealevel (nearly) without any time lag. Under this assumption the consequences of these rates of sealevel rise for the Dutch Wadden Sea system have been analyzed (Louters and Gerritsen, 1994). Huge quantities of sediment are transported by the tides between North Sea and Wadden Sea. The flood transports annually about 40-60 million m3 of sand and 100-200 million m3 of silt and clay through the tidal inlets of the Wadden Sea. The ebb carries quantities of the same order of magnitude, but the variance of the data is such t h a t no conclusions can be drawn from the flood and ebb transports about any net transport. It is clear, however, that the quantities transported are much larger t h a n those required to explain sedimentation and/or erosion rates observed in the area.
792 Table 5.1 Amounts of sediment required annually (in million m3 per year) to compensate for changes in the morphology of the tidal basins of the Dutch Wadden Sea Sediment demand (106 m3 per year) 1990 - Present rate of sealevel rise (20 cm/100 year) - Past engineering works Extraction of sand and shells Extraction of natural gas Accretion of salt marshes - Total sediment requirement at present rate of rise -
-
-
- Extra required at sealevel rise of 60 cm/100 yrs - Extra required at sealevel rise of 85 cm/100 yrs
2040
2090
4-5
4-5
4-5
2-3 8-9 0.3 0-9
1-2 6 1-2 0-9
1-2 6 0 0-9
14-26
12-24
11-22
4-5
6-7
6-7
9-10
Louters and Gerritsen (1994) attempt to quantify the amounts of sand needed for the various changes in basin morphology in the Dutch Wadden Sea with use of the empirical model MORRES (Table 5.1).
Effects of engineering works. To restore the disturbed dynamic equilibrium caused by the damming of the Zuiderzee (1932) and Lauwerszee (1969) a long term supply of sand is needed. In the remaining Zuiderzee basin (= the westernmost part of the Wadden Sea) about 100-200 million mS sediment have been deposited in the past 60 years. Louters and Gerritsen (1994) estimated that the total restoration of the basin morphology will require another 700-900 million mS, of which about 70% will be needed in the first 300 years. Therefore the westernmost part of the Wadden Sea will continue to require 1-2 million m3 annually. The much smaller Lauwerszee basin requires about 50 million m3 of sediment of which about 60% had been deposited up to 1987. It is expected t h a t about 1 million mS will be deposited annually in the next 30 years. Altogether the compensation required because of past engineering works will demand 2-3 million m3 annually; at a later stage this will become less.
Extraction of sand and shells. Louters and Gerritsen (1994) report t h a t in the period 1960-1990 annually about 8-9 million m3 of sand have been derived from the Wadden Sea. They expect, based on present government policy, t h a t in the future this amount will consist of about 6 million m 3 of sand and 0.14 million m3 of shells. About 75% of this will be derived from the Ems estuary by dredging in relation to shipping.
793
Extraction of gas. Especially in the e a s t e r n p a r t of the Dutch W a d d e n Sea subsidence of the seafloor due to extraction of n a t u r a l gas will require some 1 million m3 per year for the coming decennia.
Accretion of salt marshes. N a t u r a l accretion of salt m a r s h e s occurred in the past centuries. Although the exact rates of horizontal accretion are not known, an indication m a y be derived from the rate of reclamations, since reclamation of salt m a r s h e s m a y be assumed to have kept pace with horizontal accretion of marshes. Hence, in the past the rate of accretion m a y have been 1.5 km2 per year (Dijkema, 1987). This rate will have required a compensatory sand supply to the tidal basin of about 7-9 million m3 annually. Nowadays, however, the salt m a r s h e s do not show horizontal accretion any longer; in fact the s a l t m a r s h area has been stable since about 1970. Consequently the sand demand of the W a d d e n Sea caused by salt m a r s h accretion m a y be expected to be much less or even nil nowadays. This situation is dependent on h u m a n m a n a g e m e n t of the marshes. For the long t e r m the value of 7-9 million m3 per year may be considered as a ceiling.
Sealevel rise. The present rate of sealevel rise in the Wadden Sea (about 20 cm per century) requires 4-5 million m3 of sand per year for compensation of the depth changes in the tidal basin. An increased rate of sealevel rise of 60 cm per century will require an additional 6-7 million m3 per year in 2090, and a rate of 85 cm per century will need 9-10 million m3 extra per year. The r a t e s of s e d i m e n t a t i o n predicted are much smaller t h a n the a m o u n t s of s e d i m e n t transported. Hence, it is expected t h a t the tidal basins of the Dutch Wadden Sea can keep up with rates of sealevel rise of 60 and 85 cm per century, respectively. Thus it becomes possible to predict the change in the area of tidal flats in the next 50-100 years. It is concluded t h a t a rate of sealevel rise of 60 cm per century will result in a loss of about 0.5% of the area of tidal flats, provided t h a t the supply of sand r e m a i n s constant; a rate of 85 cm will result in a loss of about 1%. These insignificant changes are more t h a n compensated by the expected development of new tidal flats in those tidal basins which still are restoring from the effects of engineering works in the past. For the Marsdiep basin a development of new flats of about 30-50 km2 is possible (2.5-4% of the Wadden Sea total). However, for the individual tidal basins of the Marsdiep and the Ems-Dollard the total a m o u n t of sediment required at rates of sealevel rise of 60 cm and more per century, m a y come very close to the potential supply. Evaluation The study about the effect of accelerated sealevel rise on the flats in tidal basins has a n u m b e r of aspects which require attention. In the first place the model MORRES has been based on empirical relationships determined under the present rate of sealevel rise. It can not be verified w h e t h e r these relationships, such as the assumption of a dynamic equilibrium, will also hold at much higher rates of sealevel rise. Secondly, the results have been based on the assumption of a sediment supply larger t h a n the sediment demand. Above it was already indicated t h a t this is not necessarily true.
794 Thirdly, the model operates at the scale of the tidal basin, which makes translation of the result to ecologically meaningful scales difficult. On the other hand the study provides insight in one of the probably major impacts on the present Wadden Sea system by climate change. Hence, it will be useful to continue the studies on the future development of tidal flat morphology, especially at smaller scales.
5.2 I m p a c t of sealevel rise on erosion, s e d i m e n t a t i o n and plant d y n a m i c s in salt m a r s h e s S a l t m a r s h e s , s a l t m a r s h w o r k s a n d sealevel rise Salt marshes are areas covered with terrestrial vegetation under the influence of seawater. In the Wadden Sea these marshes are found at levels above about mean high tide level at neap tides. Originally salt marshes occurred up to the highest level of regular saltwater flooding. This is still the case on parts of the barrier islands, but elsewhere, especially along the mainland coast, salt m a r s h e s are bounded by the seawall. Moreover the salt marshes along the larger part of the m a i n l a n d coast have been created by h u m a n intervention t h r o u g h m e a s u r e s stimulating accretion (Figure 5.1). This activity, originally aimed at gaining new land for agriculture, has been abandoned in 1980; nowadays the status quo is maintained for these marshes in order to preserve them as nature reserves. Hence, the former 'land reclamation works' have been coined now ' s a l t m a r s h works'. The present area of salt marshes, including the saltmarsh works, in the Dutch part of the Wadden Sea is 73 km2 (Dijkema et al., 1990). The vertical growth of the marsh is determined by the rates of minerogenic and organogenic sedimentation, the frequency and period of tidal flooding and the overall compaction of the sediment (Allen, 1990; Craft et al., 1992). The s e d i m e n t a t i o n rate on the m a r s h is controlled by the m e a n tidal amplitude (Stevenson et al., 1986) and is a function of the height of the saltmarsh in relation to mean high water level (Dijkema et al., 1990). For partly or wholly minerogenic marshes the transport of fine sediment to the marsh surface is mainly dominated by the flood tidal currents (Allen, 1990; Postma, 1967). In addition, the sedimentation rate on flats and marshes is high in those areas which are sheltered, frequently overflooded and where the sediment supply is high.
795
3 rd
field
sedimentation field
....
'---4
.....
I i
2 nd
~
......
i
!
field
!
(400xqOOm)
1 ......
'
, i
-----II IlIHIHi
"
subfield (|OOxlOOm)
....... ~l]HlnT]n
.~e~ ,n~ ,o~ measurements of elevation
Ii
IIIIIIIIIIIIIi_ ....._______,
. .......
~u~,.~o ~o,n. earth groyne ~main
....
1st
.
"
.......
: nnnnH
field
I
i.......nlnHIl
ditch
ditch , open field drain
o
,oo
2.__~.__==oo
m
[...... 7nIH ll ~
IIIIIIIIIIIII
Figure 5.1. Scheme of the s a l t m a r s h works along the m a i n l a n d coast of the Dutch W a d d e n Sea. The Wadden Sea is at the top of the figure, the land at the bottom W a d d e n Sea salt m a r s h e s grow t h r o u g h deposition of s e d i m e n t which leads to relatively slow (mm's - cm's per y e a r ) v e r t i c a l growth of the m a r s h a n d at the s a m e t i m e relatively rapid (dm's - m's per year) horizontal growth at the s e a w a r d edge. W a d d e n Sea salt m a r s h e s decrease t h r o u g h erosion, vertically all over the m a r s h as well as horizontally, often in the form of cliffs, m a i n l y at the s e a w a r d side. An horizontally eroding m a r s h can be growing vertically at the same time. On the m a r s h the balance between sedimentation and erosion is usually positive. The balance b e t w e e n erosion and accretion at the s e a w a r d edge of the m a r s h , w h e r e pioneer v e g e t a t i o n e s t a b l i s h e s i t s e l f d u r i n g accretion, b u t w h e r e t h e s a m e vegetation is destroyed during erosion, is m u c h more unstable. This s e a w a r d edge or pioneer zone, apparently is the most vulnerable zone of the marsh. Most s a l t m a r s h e s along the m a i n l a n d coast h a v e a static b o u n d a r y at t h e l a n d w a r d side in the form of a huge seawall; the s e a w a r d b o u n d a r y of m a n y W a d d e n Sea m a r s h e s is fixed by m e a n s of the construction of brushwood groynes. Because of the seawall the m a r s h e s cannot shift their position l a n d w a r d during an increase in sealevel rise. Their only way to survive is to h e i g t h e n up the bed level and so keep pace w i t h the increase in sealevel. It depends on the outcome of the change in processes and s e d i m e n t t r a n s p o r t on and towards the salt m a r s h e s , if these m a r s h e s eventually will submerge or keep pace with an increase in sealevel rise. Since the salt m a r s h e s in the Netherlands Wadden Sea are a result of the dynamic equilibrium between the sedimentation and the erosion processes, changes in the
796 hydrodynamic parameters which might be changed by sealevel rise and climate change, like tidal amplitude, mean high water and wave height distribution, are expected to change the sediment transport pattern in the Wadden Sea and the net result of the processes on the salt marshes. In order to predict the effect of sealevel rise on s a l t m a r s h development, the hydrodynamics and the sedimentation and erosion processes have been studied during NRP I at three locations in the saltmarsh works along the mainland coast. At the same places the adaptation of the vegetation cover to changes in the hydrodynamic parameters has been investigated.
H i s t o r i c a l development o f Wadden Sea salt marshes In the mediaeval Wadden Sea large areas of salt m a r s h e s occurred. H u m a n impact was negligible and probably consisted mainly of building artificial mounds for building houses and of grazing cattle. E m b a n k m e n t of m a r s h e s probably started in the 10th century and by 1300 large areas had been surrounded by dikes. After this time saltmarsh reclamation continued and more or less kept pace with the accretion of new m a r s h e s (Dijkema, 1987). On average about 1.5 kin2 of m a r s h was reclaimed annually. From the 17th century onwards man started to build sand dikes on the barrier islands. In the lee of the dune-like constructions new salt marshes developed. Some of these areas have been reclaimed since, but on several areas large areas of saltmarsh still occur. Along the mainland coast reclamations gradually overtook the natural accretion. One reaction of the coastal population was to stimulate drainage by digging ditches. Thus the marsh area was better drained which enabled the establishment and development of vegetation which again stimulated accretion and growth of the marsh. Subsequently the new marshes could be embanked. From 1930 onwards, brushwood groynes have been build along the mainland coast of the Dutch Wadden Sea to stimulate accretion even more. The construction of the brushwood groynes significantly increased the rate of sedimentation in the areas in between (Dijkema et al., 1988, 1990). Digging of ditches occurred already before 1930. Through the combination of brushwood groynes and ditches the m a r s h area expanded seaward with 8.2 m yr-1 for the Friesland coastal area and 4.7 m yr-1 for the Groningen area during the period 1960-1985 (Bakker et al., 1993). The sedimentation rate and seaward expansion of the saltmarsh was higher during the first years of the construction of the brushwood groynes (1960-1978). Later saltmarsh expansion arrested and erosion occurred from 1978 to present. In 1980 the Netherlands government decided that the existing salt marshes would not be embanked and instead managed as nature reserves. The s a l t m a r s h works had to be aimed at maintaining the status quo. Since, hardly any further accretion of the m a r s h e s has occurred. Partly this can be ascribed to a reduction of the m a n a g e m e n t effort, but other factors have to play a part as well. Van Malde (1992) found for different gauge stations in the Netherlands part of the Wadden Sea an increase in mean sealevel, according to the trend line computed for the period from 1900 till 1960, of about 0.18 cm yr-1. Recent years have shown an accelerated increase in mean high water level of 0.44 cm yr-1 from 1961 to 1983
797 (Dijkema et al., 1990). It has been suggested that this recent rapid rise of m e a n high-tide level could be responsible for the present standstill of the m a r s h growth thus illustrating the vulnerability of the m a r s h for increased rates of sealevel rise (Bossinade et al., 1993). The present m a n a g e m e n t of the man-made marshes or saltmarsh works along the mainland aims in the first place at a reduction of turbulence on and in front of the salt marshes. This is done by the construction of the brushwood groynes. The result is twofold: reduction of the near-bed turbulence leads to less erosion of the bed and it leads to possible increase in sedimentation of mud in and in front of the m a r s h zone. Secondly, m a n a g e m e n t of the m a r s h area aims at protection and improvement of the cover of the vegetation. This results partly from the construction of the groynes: protection of the vegetation against waves and currents, and partly from digging ditches. This ensures a good drainage of the bed which reduces sediment mobility and improves the growth of the vegetation (Reed and Cahoon, 1992).
Transport of sediment and sedimentation The sediment dynamics on and in front of the salt-marshes are determined by the tidal current and the waves. Waves determine the amount of (re)suspended matter and currents will transport the sediment in suspension. The sedimentation on the salt m a r s h e s is d e p e n d e n t on the near-bed turbulence and the grain-size distribution of the transported sediment. Suspended sediment concentrations have been measured during the NRP I study s i m u l t a n e o u s l y with hydrodynamic m e a s u r e m e n t s . The s u s p e n d e d sediment consists of the mud fraction (<50 ~m) only. It is concluded that mud is transported as suspended load throughout the entire water column. Sand, on the contrary, is resuspended by wave action and transported close to the bed as bed load. The data show a decline in the sediment concentration from flood towards ebb. Thus it seems that suspended sediment is transported into m a r s h area. However, the amount of transported sediment is a function of the concentration times the current velocity and the net transport is also dependent on the current direction. It is to be expected that the vegetated parts will show less turbulence in the near-bed water column and most sediment will be deposited here. Advective t r a n s p o r t of suspended sediment into the s a l t m a r s h works has been found during moderate weather conditions as well as just before storms and shortly after a storm period. During storm conditions sediment in the s a l t m a r s h works is r e s u s p e n d e d by wave action and t r a n s p o r t e d from the m a r s h by the strong currents. In this case, the brushwood groynes prevent the loss of large quantities of sand out of the sedimentation fields. These data confirm the earlier results of Dankers et al. (1984) and Asjes et al. (1992) who found a net transport of sediment to Wadden Sea salt marshes on an annual basis. These authors also report that transport towards the m a r s h occurs during quiet weather, whereas seaward transport may occur during storms. Net sedimentation rates have been reported by Dijkema et al. (1990). They vary in time and space and range from about - 1 cm per year to + 3 cm per year. Most marshes show positive vertical accretion rates.
798
Erosion A recent loss in saltmarsh area in the Wadden Sea might be due to an accelerated increase in mean high water level of 0.44 cm yr-1 from 1961 to 1983 (Dijkema et al., 1990), of which between 1976 and 1983 about 75% is calculated to be due to an increase in average windspeed (Bossinade et al., 1993). During the NRP I study much attention has been given to erosion processes. It is concluded from field m e a s u r e m e n t s t h a t the tidal c u r r e n t velocity perpendicular to the shore during moderate (summer) weather conditions never exceeds 10 cm per second inside the brushwood groynes. During storm conditions the waterdepth increases significantly due to wind set up and the current velocity can reach up to 20 cm per second. The ebb current is stronger t h a n the flood current. The alongshore tidal current velocity is extremely low during moderate (summer) conditions. However, wind induced alongshore current velocities can range up to 40 cm per second (as much as 10 times the tidal current component). Wind waves grow during an increase in windspeed and fetch length. It is concluded t h a t the w a t e r d e p t h is the limiting factor for wind induced wave growth at the m a r s h sites. The construction of brushwood groynes, in this situation, does not have any effect on wind wave growth at the m e a s u r e m e n t location some 200 m e t e r distance from the groynes. This conclusion is confirmed by model calculations showing that the growth of wind waves is completed within 30 meters distance from the groynes (Elfrink and Houwing, in prep). A new in-situ erosion meter has been developed to measure the in-situ bed-shear strength of cohesive beds. This in-situ erosion flume (ISEF) is a circulating flow system in the vertical plane. The bottom part of the horizontal test-section is open over a length of 0.9 meter. Current velocity and sediment concentration sensors are fixed in the flume. The shear stress, exerted on the bed by the unidirectional current in the ISEF, causes an erosion of the bed. This erosion is measured by the concentration sensor. It is assumed t h a t the current velocity, at initiation of motion of the top layer of the bed, can be used to compute the critical bed-shear stress. This critical shear stress is assumed to be the characteristic p a r a m e t e r which defines the different types of sediment beds. The results from the ISEF m e a s u r e m e n t s show only small differences in the erosion resistance of the bed within the s a l t m a r s h works. This m e a n s that, theoretically, the bed strength does not differ spatially within the fields and erosion of the bed will start, if the shear-stress is high enough, simultaneously at different places. However, visual observations at low tide show that erosion of the bed is very local. It starts at specific locations only, for instance where the structure of the bed is disturbed by bioturbation. These small-scale erosion p a t t e r n s can generate larger scale erosion forms.
Vegetation and climate change Sea-level rise could become a t h r e a t to coastal marshes by affecting the m a r s h vegetation through an increase of the frequency of tidal flooding and wave energy. If the accretion of sediment on the marsh is less than the increase in sea-level, the m a r s h will be flooded more frequently and vegetation deterioration will increase (Reed and Cahoon, 1992). Similarly, erosion of the marsh will also result in a loss of vegetation and subsequently in a loss of saltmarsh area. Salicornia dolichostachya is the most i m p o r t a n t pioneer species found in the foreland salt marshes of the Wadden Sea. The first important and perhaps most critical phase in the life cycle of Salicornia is the germination and survival of the
799 seedlings (Ungar, 1978). It determines, in combination with growth, the plant density at the end of the growing season and biomass of the fully grown Salicornia population. The second important phase in the life cycle is the seed production. If the a m o u n t of seeds remaining in the pioneer zone after the winter is too low, the size of the Salicornia population will decline which might have a great influence on the development of the saltmarsh as a whole. Although S. dolichostachya is able to reduce wave energy and thus create suitable circumstances for sediment to settle, this annual species is not able to fix mud and therefore does not seem to be of great value for accretion on salt m a r s h e s . However, Salicornia fulfils an i m p o r t a n t role in the spreading and s e a w a r d expanding of Puccinellia maritima. This very important species for silting up is spread mainly by vegetative parts torn from the parent plant. P a r t of these tillers get caught behind living, or in winter dead, Salicornia plants and can establish themselves especially on the higher parts of the pioneer zone (Kamps, 1962). Hence, the presence of Salicornia in the pioneer zone is very i m p o r t a n t for the expansion of salt marshes. The size of the glasswort dominated pioneer zone might be an indication of the quality of the saltmarsh. Therefore, more information has been collected about the life cycle and boundary conditions for establishment of this species. Like Jeffries et al. (1981) no seed bank was found in October before seedfall. Like Beeftink (1985) found for Salicornia procumbens, the population of Salicornia dolichostachya is thinned especially during the seed phase in winter time (99% of the produced seeds was removed from the pioneer zone) and during the growth from established seedlings to fully grown plants (65 % of the seedlings did not survive). Nevertheless, on both sites the development of the Salicornia population does not seem to be limited by a lack of seeds. In contrast to Joenje (1978), it was found in this study t h a t seeds do not act as in water suspended sediment, but are transported as bed load like sand grains (6072~). Finally, m a n y seeds will end up in creeks bottoms where they cannot germinate because of lack of oxygen and a high salt concentration. Evaluation
At the time of writing of this assessment report the s a l t m a r s h study was not yet completed. Especially the relationships between the different parts of the study had not yet been sufficiently developed. Nevertheless the study m a k e s clear t h a t s a l t m a r s h erosion is a complicated process and that further research is required on this subject. The s a l t m a r s h study under NRP I considered processes at the scale of the m2. This makes comparison with the results of the other studies a difficult task. A very detailed in-depth study of some important processes in the saltmarsh, necessarily leads to a lack of overview. On the basis of the results it is not yet possible to make predictions about the impact of sealevel rise on the entire s a l t m a r s h area of the Dutch Wadden Sea. On the other hand some conclusions on the effectivity of the brushwood groynes in stimulating s a l t m a r s h accretion emerged. These conclusions may be incorporated in the m a n a g e m e n t plans for these areas. As a preliminary conclusion from the NRP I study emerges t h a t Wadden Sea salt marshes may survive an increased rate of sealevel rise if erosion processes can be
800 controlled. Sediment supply and accretion rates seem to be sufficient for higher rates of increase of sealevel than at present. This conclusion is in line with a recent study (Oost and Dijkema, 1993) on the possible effects of soil subsidence due to gas extraction on W a d d e n Sea salt m a r s h e s . They concluded t h a t the salt m a r s h e s at the b a r r i e r islands could accomodate a soil subsidence (= equivalent to a sealevel rise) of 0.5 cm per year (= 50 cm per century), whereas the mainland marshes could survive a rate of 1.0 cm per year. However, they drew attention to the vulnerability of the pioneer zone of the m a r s h e s with respect to erosion processes. The conclusions reached above might be tested by the large-scale field study of the effects of soil subsidence due to extraction of n a t u r a l gas (see Table 1.1). This 'experiment' gives information on the behaviour of several km2 of salt marshes in a 'simulated' rapid rise of sealevel. Unfortunately the results of this study have not yet been completely analyzed. 5.3 I m p a c t of s e a l e v e l r i s e o n b e n t h i c a n i m a l s of tidal fiats
Results of a literature study Louters and Gerritsen (1994) report the results of a literature study on the impact of sealevel rise on benthic animals of tidal flats. They conclude t h a t most of the tidal flat species are characterized by a broad ecological amplitude with regard to salinity, w a t e r movements, sediment characteristics, and period of emergence of the tidal flats. Hence, they conclude t h a t sealevel rise as such will not have an i m p o r t a n t ecological effect on tidal flat organisms as long as the present a r e a of tidal flats does not change very much. F r a g m e n t a t i o n of large tidal flat areas because of sealevel rise might result in a situation with more smaller tidal flats, each with an low-tide edge relatively poor in benthic animals. This might have consequences for the food supply for birds and fish.
Evaluation This literature study lacks experimental verification, but there is little to indicate t h a t e x p e r i m e n t s on the effects of higher w a t e r levels would result in other conclusions for the individual species. Also the study did not address the impact of sealevel rise on new species which might colonize the Wadden Sea because of higher temperatures. 5.4 I m p a c t of s e a l e v e l rise on s h o r e b i r d s
Defining the problem of sealevel rise for shorebirds Along coasts m a n y bird species can be seen feeding on the tidal flats. In the Dutch W a d d e n Sea this includes 15-20 a b u n d a n t species from different taxonomic groups: geese, ducks, waders, and gulls. The various species breed in an area ranging from N.E. Canada through Greenland, Iceland, Spitsbergen, and Europe to n o r t h e r n Russia and Siberia. After the breeding season these birds migrate to the Wadden Sea either to stay there all winter or to remain there for a shorter period to feed and to moult, after which the birds fly on to wintering areas further South, e.g. in western Africa. Next spring these populations return to the breeding areas, often after another stopover in the Wadden Sea.
801 D u r i n g t h e i r life cycle these shorebirds are dependent on two very different habitats. They breed in terrestrial environments which differ according to the species concerned. Some species breed in salt m a r s h e s or dune areas, other in agricultural landscapes, still others in heathlands or forests, and several species have their breeding habitat in the Arctic tundra. At first sight it seems t h a t many, if not all species occupy huge areas as breeding ranges. Outside the breeding season, however, all these birds concentrate in a r a t h e r small n u m b e r of tidal flat areas where most of them feed on benthic animals such as shellfish and worms, and some graze algae and seagrasses. The total area of tidal flats along the flyway these bird populations use is small, only about 7,000 km2 of which about 50% are situated in the Wadden Sea. Hence, it has been suggested t h a t the numbers of these birds are primarily governed by competition for food present in their tidal flat habitat in winter. If so, disappearance of tidal flats due to sealevel rise could have a very large effect on these bird populations. When density-dependent processes operate on both the wintering grounds and the breeding grounds of shorebird populations a reduction of either habitat will lead to a decrease in population size (Fretwell, 1972; Goss-Custard, 1980). This implies t h a t when sealevel rise will cause a change in the area of tidal flats available as feeding grounds, population size of shorebirds may be changing too. Furthermore, m a n y species of shorebirds are concentrated in only a few widely separated estuaries during w i n t e r and migration. It is thought t h a t this makes these species especially vulnerable to h a b i t a t loss, as this distribution p a t t e r n breaks the normal link between the a b u n d a n c e of a species and its i m m u n i t y to extinction (Myers et al., 1987; Davidson and Piersma, 1992). Among the bird species dependent on tidal flats limicoles or w a d e r s keep a prominent position. Most wader species breed in arctic or subarctic environments and spend the time outside the breeding season on tidal flats south of their breeding areas. The are several indications t h a t competition in the w i n t e r q u a r t e r s does occur. Several species exhibit leapfrog migration, whereby more northerly breeding populations w i n t e r south of more southerly breeding populations, which is generally attributed to the avoidance of competition (Greenberg, 1986). There is also direct evidence for both competition for food on the wintering grounds, through interference and prey depletion, and for competition for territorial space on the breeding grounds (Goss-Custard, 1986). Several studies exist t h a t report on the effects of loss of tidal flats due to h u m a n interventions. Evans (1978) reports t h a t following a reclamation of 60 % of the intertidal flats of the Tees estuary in north-east England, the wintering numbers of the majority of wader species decreased by a similar percentage or more in the year following the reclamation. Similar results are reported for reclamation of intertidal areas of the Forth estuary in eastern Scotland (McLusky et al., 1992). The construction of a dam across the entrance of the Grevelingen estuary in The N e t h e r l a n d s ended tidal movements in the e s t u a r y and resulted in complete disappearance of the tidal flats. Consequently, nearly all waders disappeared from the a r e a (Wolff et al., 1975); for at least two species it could be proven t h a t n u m b e r s in the n e a r b y Oosterschelde estuary had increased at the same time (Van Latesteijn and Lambeck, 1986; Lambeck et al., 1989). The former studies show t h a t local occurrence of waders is influenced by loss of tidal flats, but they do not show any population effects. Meire (1991), however,
802 reports that after a 30% decrease of intertidal area in the same Oosterschelde estuary due to the building of a storm-surge barrier and secondary dams about 15 years later, the total number of Oystercatchers in the unaffected parts of the estuary did not increase. Moreover, a temporary reduction of the tidal range in the estuary caused by partly closure of the storm-surge barrier in combination with extensive fisheries for shellfish, led to strongly increased m o r t a l i t y of oystercatchers in winter. In the face of the strong fidelity of adult waders to their wintering grounds (Myers, 1984) it is actually surprising t hat the wintering numbers so often quickly readjusted to a reduction in feeding area. Thus, the results of the 'field experiments' corroborate the expectations derived from the studies on competition that reduction of tidal flat habitat will lead to reduction of wader population size. Because bird populations are subject to processes both in the breeding areas and in the winter quarters it is hard to predict the effect of sealevel rise on wader populations. Moreover, the various species and populations normally use several estuaries along their flyway during the non-breeding period and will be influenced by changes in all of these estuaries. There is no shortage of speculations on the possible consequences of climate change. Although not all possible effects are considered, -apparently nobody seriously suspects an effect of increased levels of UV-B on migratory birds for instance-, the diversity of hypotheses is nonetheless bewildering. Therefore, instead of formulating complex scenarios we must seek to answer a set of interrelated precise questions. At least three questions can be distinguished: Which aspects of the climate change or secondary effects will have the greatest impact: global warming, storms, rainfall patterns, sealevel rise? At wh at stage of the life cycle of the birds does the climate change or climate-induced effect operate: breeding season, autumn migration, spring migration, nonbreeding season? Which processes must be considered: vagaries during the migratory flight, interspecific competition between resident and non-resident species, intraspecific competition on the breeding grounds, intraspecific competition on the wintering grounds, phenology of the food supply, evolution of the migration schedule? Second, a formal structure must be devised that will allow us to find an answer to these questions. This formal structure includes mathematical models and a methodology to assess important parameter values. Hence, it was decided to study the effect of sealevel rise by building mathematical models of the population dynamics and the migratory behaviour of two well-studied species: the Oystercatcher Haematopus ostralegus and the Knot Calidris canutus. The Oystercatcher serves as a model for a short-distance migrant, whereas the Knot is a clear example of a long-distance migrant. In the short-distance migrant the migratory phase itself is insignificant in time, and attention focusses on how processes on the wintering and breeding grounds affect the population. In contrast, the long-distance migrant spends a significant amount of time preparing for migration and the model specifically focusses on this part of the annual cycle. The next step is to use the models to investigate the consequences of various climate scenarios.
803
Model development The basic idea of the model is that both the rate of reproduction in summer and the rate of mortality during winter are dependent on the number of individuals in the population, i.e. the population density. The p r i m a r y source of density-dependence in s u m m e r is assumed to result from territorial exclusion, i.e. relatively fewer birds breed when the density of potential breeding birds increases. The proportion excluded is expressed as a k-value: k - aT + bTlogl0N where N is the n u m b e r of potential breeding birds and the slope bT measures the compressibility of the territories. W h e n bT = 1, the territories cannot be compressed any further so that, above the numbers set by the intercept aT, a constant n u m b e r of pairs breed, irrespective of the numbers a t t e m p t i n g to do so. Thus, bT = 0 implies no density dependence and bT = 1 implies perfect 'contest' density dependence. It is not allowed that bT < 0. The primary source of density dependence in winter is decreased survival chances due to competition for food. At low densities, there is no competition, but only a d e n s i t y - i n d e p e n d e n t proportion (mw) starving. When bird density increases, eventually a point, Cw, is reached at which mortality begins to increase and so becomes density-dependent. From then on, mortality increases by bw for every unit increase uin bird density. For the Oystercatcher four sub-populations have been distinguished: Continental coastal, Continental inland, Atlantic coastal, and Atlantic inland. The p a r a m e t e r s defined above have been estimated for each sub-population and, where necessary, age class. F u r t h e r m o r e the size is known of each of the four sub-populations. Within the Continental and Atlantic regions, inland-breeding and coastal-breeding sub-populations use the same coastal areas in winter. Finally, the sometimes quite substantial annual fluctuations in the main production and mortality p a r a m e t e r s were generally not correlated across sites within a sub-population. This allows the s t a n d a r d deviations of the annual variations in these parameters to be estimated for both sub-populations in each region, so t h a t realistic annual variations could also be included in the model. More details are to be found in Goss-Custard et al. (1994). During their m i g r a t o r y journeys, birds m a y cover distances of up to several thousand kilometers in a non-stop flight. The migration of m a n y species, especially waders, consists of a series of such non-stop flights i n t e r r u p t e d by stopovers at tidal flat areas during which large quantities of fat are deposited. This fat serves as fuel for the next leg of the journey. It seems likely t h a t waders migrating between their widely separated wintering and breeding grounds can choose among a number of potential stopover sites using different itineraries. However, the factors determining stopover site choice and departure fuel loads from the stopover sites are not well understood. It has been argued t h a t waders sometimes store more fat t h a n is necessary to fuel the flight to the next site, a phenomenon known as overloading, and t h a t suitable sites are skipped. When only energetic flight costs are considered neither the
804 skipping of potential sites should occur nor overloads be deposited, because flight costs increase rapidly with increasing body mass. Using the equations of Alerstam and LindstrSm (1990) and Weber et al. (1994) developed optimality models of d e p a r t u r e fat loads and stopover site use with time spent m i g r a t i n g as the currency to be minimised. Under this assumption the only cost is the reduction of marginal rate of gain of flight range with increasing body mass. The models identify conditions under which potential stopover sites are skipped or overloads occur. In these models the conditions leading to a storage of overloads are very restrictive but skipping can occur over a wide range of parameters. However, several potentially important factors shaping migration schedules and fuel loads cannot be addressed using the aforementioned models. These factors are: 1.
2.
.
At stopover sites or on the wintering grounds birds m a y face a trade-off between gaining energy for fat deposition and avoiding predation. Predation risk is likely to be influenced by several factors, including flock size, high body mass impairing flight kinematics and consequently the escape response when attacked and the time exposed to predation risk while foraging. Time of arrival in the breeding area and the subsequent timing of the breeding a t t e m p t have a strong effect on breeding success. F u r t h e r m o r e , extra body stores at arrival may guard against adverse conditions or pay some of the costs of producing eggs. The rewards associated with time of arrival m u s t not necessarily be always decreasing with later arrival; arriving on the breeding grounds too early could also be disadvantagous, for instance due to snow cover. Foraging at a stopover site and d e p a r t u r e decisions are also likely to be influenced by s t o c h a s t i c i t y of the e n v i r o n m e n t , including the social e n v i r o n m e n t . Stochasticity in food supply and abiotic conditions, like t e m p e r a t u r e and wind speed and direction affect energy intake and energy expenditure and may therefore lead to unpredictable daily fat deposition rates or even to the risk of starvation. Unpredictable wind conditions during flight m a y cause uncertainty in expenditure when flying from one stopover site to the next.
The model describes the behavioural decisions of a bird having to migrate from its breeding grounds to its wintering grounds, or vice versa. It is a s s u m e d t h a t decisions are made in such a way t h a t fitness upon arrival at the destination is maximised. Fitness expectations upon arrival may depend on both the state of the individual and the time of arrival. Fitness is defined as the life-time contribution of the individual to the next generation. The environment the bird has to migrate through consists of N sites, i.e. the initial site (e.g. the wintering ground), N-2 discrete stopover sites and the final site (e.g. the breeding ground). For simplicity, it is assumed t h a t these sites are a r r a n g e d linearly, so t h a t the distances between the sites are additive. The complete migration period is divided into T time units; one time unit corresponds to one day. The state of a bird is characterised by its reserve level and its location. Location is referred to as a state variable, because it is affected by the decisions of the bird. The reserve level can take any integer value from 0 up to a m a x i m u m reserve level. If the reserves fall to 0 it is assumed t h a t the birds i m m e d i a t e l y die of starvation.
805 At the s t a r t of each time unit the bird m u s t decide on the best possible course of action, given its location and reserves. It can either decide to stay in the site and feed, or to fly to another site in the direction of the final destination. If it decides to feed, it also h a s to decide on its foraging i n t e n s i t y u, which can t a k e a n y value between 0 and 1. W h e n feeding there is also a risk of predation. This risk can be influenced by two variables: the mass of the individual, which is a linear function of reserves, and the foraging intensity u. If the bird decides to depart it m u s t decide on the site w h e r e it will land. It is a s s u m e d t h a t no m o r t a l i t y occurs d u r i n g flight, while flight costs are modelled according to the e q u a t i o n s of A l e r s t a m a n d L i n d s t r 6 m (1990). E x p e n d i t u r e during flight can be stochastic due to influence of unpredictable wind conditions. To find the optimal behaviour, the method of dynamic p r o g r a m m i n g (Mangel and Clark 1988) h a s been applied. For each combination of reserves, site and time, the optimal decision and expected future success is found using b a c k w a r d iteration. Forward iteration then allows to simulate the fate of individual birds, as well as to find population m e a n s and standard deviations. To investigate the effects of h a b i t a t change, e.g. because of sealevel rise, a simple s i t u a t i o n is c o n s i d e r e d w h e r e the fitness consequences of c h a n g e s in t h e environment are determined. These changes could take place at two different time scales: if changes are slow the migrating birds are allowed to adjust their behaviour to the new circumstances, i.e. the birds behave optimally in an e n v i r o n m e n t t h a t deteriorated as compared to an earlier reference situation. Alternatively changes could h a p p e n m u c h faster and the birds do not have the chance to a d j u s t t h e i r b e h a v i o u r to the new e n v i r o n m e n t . Birds t h e n use the b e h a v i o u r which w a s o p t i m a l u n d e r b e t t e r c i r c u m s t a n c e s , i.e. t h e i r m i g r a t o r y b e h a v i o u r is now s u b o p t i m a l in the new environment. In both scenarios fitness can be affected in several ways: either by changes in m o r t a l i t y en route or by delayed arrival at the breeding site. The birds can also be forced to stay at an i n t e r m e d i a t e site. Evans et al. (1992) discuss the consequences h a b i t a t loss at s t a g i n g posts can h a v e for m i g r a t o r y w a d e r s . H a b i t a t loss is modelled by decreasing the m a x i m u m fuel deposition the birds can achieve at a site. D e p e n d i n g on t h e s i t u a t i o n e n v i s a g e d h a b i t a t c h a n g e can h a v e d i f f e r e n t consequences. If the birds can adjust their behaviour to the new conditions h a b i t a t destruction at the wintering grounds are most severe. If the birds use a suboptimal policy fitness losses are more pronounced the closer the affected site is to t h e breeding site. Fitness losses are also more severe when the reward at the breeding site is decreasing over the whole time interval under consideration. The results show clearly t h a t predation is a potentially i m p o r t a n t factor s h a p i n g m i g r a t o r y s t r a t e g i e s . The r a t e s of m a s s gain are also strongly influenced by predation risk. Overloads are deposited w h e n the e n v i r o n m e n t is stochastic. U n c e r t a i n t y in e x p e n d i t u r e leads to a level of overloads which are able to deal w i t h the worst possible case of h e a d w i n d s independent of its probability of occurrence, because if such a s i t u a t i o n arises d e a t h is certain. The model does not allow to m a k e departure decisions like fuel load and the choice of the target site to depend directly on a p a r t i c u l a r wind condition; this would m a k e the inclusion of a n o t h e r s t a t e variable necessary. Intraspecific differences in site use b e t w e e n years m a y t h u s be a consequence of unpredictable meteorological circumstances.
806
Results of model calculations Since the final model calculations will be carried out after this assessment report has been finished, only some examples of calculations will be given. Figure 5.2 gives the r e s u l t s of model calculations for the N.W. E u r o p e a n population of the Oystercatcher. Four subpopulations are distinguished: coastal breeders of the British Isles and those of the continent, and inland breeders of the British Isles and of the continent. The simulations show t h a t most populations decrease with reduction of their wintering area, but t h a t the size of the reduction is strongly dependent. Especially the density dependence of winter mortality (bW) is a very i m p o r t a n t parameter. Fig. 4 gives the results of calculations with the dynamic version of the migration model. Loss of tidal flats has been modelled as decrease of fattening rates. In the top panel it has been assumed that habitat loss is a slow process and t h a t the bird populations can adapt. In the bottom panel h a b i t a t loss is rapid and the populations cannot adapt. The simulations show that under the former scenario only changes in the wintering area have an important effect. The latter scenario shows t h a t changes in any site have an effect, although the most severe effect result from changes in the last stopover area before the breeding area is reached. Hence, also for the long-distance migrants habitat loss appears to have negative consequences for population size. The d e t e r m i n a t i o n o f the a r e a o f suitable h a b i t a t The bird study started under the assumption t h a t the area of suitable h a b i t a t after a rise of sealevel could be predicted from the present topography of tidal flat areas. The basic idea was t h a t future tidal flat areas could be derived from the height distribution of the present fiats, because by higher sealevels lower areas would be drowned. This scenario could be elaborated with the data available. However, the studies on the morphology of the Wadden Sea (Sections 5.1 and 5.2) have shown t h a t tidal fiats and salt marshes are able to accomodate an increased rate of sealevel rise. For the most likely scenarios even no i m p o r t a n t change in tidal fiat area is predicted for the first 100 years. On the other hand the quality of the tidal fiats as a feeding area for shorebirds might change with sealevel rise. Hence, predictions on the area of tidal flats available as feeding areas along the flyways have become very difficult. For the Wadden Sea this is possible, but not for the other areas. Evaluation The model of the short-distance m i g r a n t , even though it was tied to the Oystercatcher, can easily be applied to other species when relevant p a r a m e t e r s are known. Habitat loss due to sealevel rise is predicted to be small (see 5.1), hence effects on wader populations are predicted to be small too, even though the exact magnitude of the effect is hard to predict for such small changes. Other effects are likely to be more important, like shell-fisheries and t e m p e r a t u r e effects on dynamics of the prey (see Section 6). The current model cannot cope with these effects. The model of the long-distance migrant, even though the model was constructed with the Knot in mind, is couched in sufficiently general terms, to be applied to other species for which the relevant p a r a m e t e r s are known. The model can be incorporated as a migration module for the population dynamics model (i.e. the model for the short-distance migrant), but this was not done, since nothing is
807 c u r r e n t l y k n o w n on density-dependent processes on wintering and breeding grounds in the case of Knots. If it is known how habitat loss will affect fattening rates on the stopover sites, the present model can predict the effect on survival rate and reproductive success. The prediction depends on the time scale: if sufficient time has elapsed for adaptation to the new circumstances, habitat loss on the first stopover site on spring m i g r a t i o n will have the most serious consequences, otherwise, habitat loss on the last stopover site will have the most serious consequences During the study it became clear that two basic assumptions were not valid. In a earlier section was already discussed t h a t tidal flat areas do not simply drown during sealevel rise, but t h a t they may be expected to restructure or even to keep pace with moderate levels of sealevel rise. Secondly the developments in the breeding areas cannot be left out of consideration. Especially the Arctic breeding species will face huge changes in their breeding quarters in the case of climate change. These changes should be taken into account if meaningful conclusions are to be drawn about the future of our wader populations.
808
CONTINENTAL
CONTINENTAL
COASTAL
,+0 t . . . . . . . . . . . . o 160 o
. "~
. . . . . .~1~.r r t. :..'vl" . . ~. .
~ . ~ '
..~ ~. . "" ,.. o~.
~1~0
~\
.
250
""-
"X
o
-
i
!
"
.,.
-'" ~ . } . . : ~ : . .
. 200
z 120
INLAND
i
-",',
150
~80o
.
-
_,am~" 60 -
:.
~ 4o-
i *'~
200
:11 i
t
I
k
~ ~,. ,,,.
,
I
_
100
li
"-...~.
'
i
I
i
5O
9
i.. %
~ ~ ~ .. .. . .
~,
.
0
t
!
t
I
t
,
,
,
I
I
,
0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 90 % REDUCTION IN OVERWINTERING AREA % REDUCTION IN OVERWINTERING AREA
ATI_ANTIC
COASTAL
ATLANTIC
INLAND
100 dfl 250 O O
t',]
200 _ Z 0 <150
'~1
\\
,.A
o 100
80
~.~g~~*"':'~ . . ' . . , , ...:'t: \ : " l : . .':l". " + \ . -,,..,, -.-,, ~
-
~
:
'.,\
\\
_
am,,.-.,.,
le
.,.
IP
,
..
60
x :\ kx
40
~
-
i
~ 5o [.., k t
f
I
I
t
~
t
~_" '
9 , ._t .
0 I0 20 30 40 50 60 70 80 90 % REDUCTION IN OVERWINTERING AREA BW.~0.10
0
I
I
!
t
t
t
t
t
t
t
.
0 10 20 30 40 50 60 70 80 90 % REDUCTION IN OVERWINTERING AREA
BW = 0.05 BW = 0.01 ,11~
BW =_0.005 BW = 0.002 BW = 0.001 ,40 F i g u r e 5.2 S i z e o f t h e e q u i l i b r i u m p o p u l a t i o n for f o u r s u b p o p u l a t i o n s of t h e O y s t e r c a t c h e r a s a f u n c t i o n o f h a b i t a t l o s s a t d i f f e r e n t i n t e n s i t i e s of d e n s i t y - d e p e n d e n t winter mortality (bW)
809 1.01
1-
o = t:
0.99 -
o.i e~ 0.98 -
0.97 -
0.96 0.75
i
I
i
I
I
1.00
1.25
1.50
1.75
2.00
2.25
1.05 tel. fitness site 1 ..... 0 ....
tel. fitness site 2
.... 0----
tel. fitness site 3
1..~~176
/
s o,~
.~
0.95 -
0.9$
,,
//
9s a
/
0.85 -
6 l o,
6
0.8
,o~
-
0.75 0.75
I
I
I
I
I
1
1.25
1.5
1.75
2
2.25
maximum gain
Figure 5.3 The effect of habitat loss (expressed as maximum possible gain of body fat load) on production of young (expressed as relative fitness) by a long-distance migrant. Site I is the wintering area, 2 and 3 are stopover sites (3 is the last site before the breeding area is reached). In the simulation one site has been changed whereas the other ones remained the same. In the top panel it has been assumed that changes occurred slowly (e.g. sealevel rise) and that the bird population was able to adapt. In the bottom panel the change occurred quickly (e.g. reclamation) so that the bird population could not adapt
810 6.
IMPACT OF CHANGING WATER TEMPERATURES
6.1 T h e e f f e c t s o f w i n t e r t e m p e r a t u r e o n t h e r e p r o d u c t i v e s u c c e s s o n s o m e b i v a l v e s in the D u t c h W a d d e n Sea Since 1969 the numbers and biomass of several species of benthic invertebrates are monitored on tidal flats in the westernmost part of the Wadden Sea (Balgzand). The common mussel (Mytilus edulis), the Baltic tellin (Macoma balthica) and the cockle (Cerastoderma edule) are among the most important species. These three species account for a high proportion of the total biomass. Their lifecycle starts with eggs from which planktonic larvae hatch. These larvae spend a period in the water column after which they settle on the tidal flats. The young individuals which thus recruit to the parent population often settle first in a habitat different from t h a t of their parents. At a later stage they move again to settle finally in the habitat of their parents. Winter t e m p e r a t u r e appears to be a factor diminishing recruitment. The data series of the Netherlands Institute for Sea Research (NIOZ) show t h a t after mild winters with a mean water temperature in the period J a n u a r y - M a r c h of about 6 ~ the r e c r u i t m e n t was lower t h a n after cold winters with a m e a n w a t e r t e m p e r a t u r e of about 1 ~ This results in high recruitment and abundance after cold winters, but relatively low abundance after mild winters, and particularly after a series of mild winters.
Two hypotheses may be formulated to account for this diminished recruitment. The first hypothesis assumes predation of just-settled larvae of these bivalves by brown shrimps (Crangon crangon) and shorecrabs (Carcinus maenas) resulting in a diminished recruitment. During a cold winter a large proportion of the predators die, whereas they do not in mild winters. This means t h a t after a mild winter the predators arrive early and they are already on the tidal flats at the moment the larvae of the bivalves settle, resulting in heavy mortality of these newly settled larvae. After a cold winter the larvae are ahead of the predators and have grown to a less vulnerable size at the time the predators appear. The second hypothesis states t h a t the metabolic rate of adult bivalves is on a higher level at higher water temperatures in winter. Since food is scarce in this period, this goes at the expense of stored reserves. As a consequence less energy is available for gametogenesis. As a result the eggs could be smaller, or the n u m b e r of eggs could be lower, or both. The first hypothesis is studied by Beukema at NIOZ. This study is dependent on field observations after mild and cold winters and is necessarily long-term. The NRP-I-period has been too short to come up with meaningful results. The second hypothesis has been studied, mainly experimentally, during NRP I. Four questions have been asked: 1. Are t h e r e differences in g o n a d a l d e v e l o p m e n t at different w i n t e r temperatures? 2. Are eggs after a mild winter smaller than after a cold winter? 3. Is the amount of eggs dependent on the winter temperature? 4. Is the survival of the larvae different at different winter temperatures?
811 6.2 E x p e r i m e n t a l tests of the effects of winter temperature on r e p r o d u c t i v e s u c c e s of b i v a l v e s So far successful experiments have only been carried out with Macoma balthica a n d Cerastoderma edule. Three groups of M a c o m a have been kept in an experimental set-up during the winter months of 1993. One group was kept at ambient temperatures, the second at t e m p e r a t u r e s about 2 ~ higher, and the third group at temperatures 2 ~ lower. The first experiment was not completely successful due to the development of an early plankton bloom in the basins, but it nevertheless d e m o n s t r a t e d a significant correlation between condition of the animals (expressed as ~g ash-free dry weight / (mm shell length)3 ) and egg size in ~lm. A second group of experiments was run in the winter of 1994. Two t e m p e r a t u r e s were used: cold (3.0 ~ and mild (5.6 ~ in combination with two periods of submergence: continuous (24 h per day) and periodically (16 h per day). The length of the period of submergence determines the possibilities for food consumption in the experimental set-up. Longer submergence should mean more food and a higher condition index. Indeed, the condition index of the group permanently submerged was higher t h a n t h a t of the group submerged for 16 h a day (Table 6.1). In accordance with the hypothesis formulated the condition index of the cold winter groups was higher t h a n t h a t of the mild winter groups (Table 6.1). Also the number of eggs varied in accordance with the hypothesis (Table 6.1). The same experiment was run for Cerastoderma edule with similar results (Table 6.1).
Table 6.1 Condition index (mg ash-free dry weight cm-3), mean egg number, and mean egg size (~t) of Macoma balthica and Cerastoderma edule after a winter and spring at cold and mild w a t e r t e m p e r a t u r e s . P a r t of the animals lived p e r m a n e n t l y submerged (sub), whereas the other group was submerged for 16 h and dry for the remaining 8 h per day (int) Condition index
egg number
egg size
Cerastoderma edule cold/int cold/sub mild/int
8.7 + 0.1 11.4 + 0.2 7.8 + 0.7
214,892 201,083 65,212
78.2 + 0.3 79.3 + 0.2 76.7 + 0.8
mild/sub
I0.0 + 0.I
Macoma balthica cold/int cold/sub mild/int mild/sub
9.5 13.8 8.1 9.8
+ + + +
0.3 1.3 0.4 0.4
9,876
29,743 63,561 19,182 20,330
75.3 + 0.i
106.9 104.6 104.1 106.8
+ + + +
1.0 1.0 0.2 0.8
812 6.3 E v a l u a t i o n As already indicated in Section 3.2 this study is not finished at the end of NRP I, so this evaluation is based on a limited amount of information. The hypothesis tested so far is confirmed by the data obtained. Whether the same conclusions can be drawn for populations in the field remains to be seen. In the meantime, nature carried out its own experiment by providing a series of very mild winters on a row. This led to unprecedented low levels of cockles and mussels in the Wadden Sea, thus confirming once more the basic hypothesis t h a t w i n t e r t e m p e r a t u r e s influence the r e c r u i t m e n t of bivalve populations (Stair, 1994). Two conclusions can be drawn from the present data. In the first place the relationship between t e m p e r a t u r e and ecological function is not a simple one. Through the interaction with food, predators, or competitors the abundance of species can change in unexpected ways. For those species which are a b u n d a n t in the Wadden Sea nowadays, we can try to investigate in well-planned studies if their n u m b e r s might be influenced strongly by some interaction of different factors. However, the reverse is also possible. For rare species, or even species which do not yet occur in the Wadden Sea, it is possible t h a t they become very numerous because of similar interactions. Such cases will be very hard to predict. The second preliminary conclusion concerns the Wadden Sea shellfish cultures. These might face a bleak future if the conclusions obtained for Macoma and Cerastoderma hold also for Mytilus. On the other hand it is also possible t h a t other ecotypes of these species develop or are introduced from more southern latitudes which might thrive in a warmer Wadden Sea.
7.
INTEGRATION
7.1 I n t e g r a t i o n of e f f e c t s of c l i m a t e c h a n g e s o n e s t u a r i n e e c o s y s t e m s All aspects of climate change have a larger or smaller impact on the Wadden Sea at the same time. In this Wadden Sea m a n y different processes interact, as do populations of plants and animals. Other h u m a n impacts also play their part. The studies under NRP I, however, are mostly concerned with only one aspect of climate change and one or a few processes or species. Although the subjects under investigation have been chosen carefully, taking into account the importance of the expected effect and the position of the process or species investigated in the total Wadden Sea system, it will be difficult to generalize the results to the entire system. For t h a t reason it has been a t t e m p t e d to bring together all studies in the ecosystem model ECOWASP developed by the DLO-Institute for Forestry and N a t u r e Research and the Netherlands Institute for Sea Research. ECOWASP is a formal description of the relations between a n u m b e r of abiotic factors (light, t e m p e r a t u r e , n u t r i e n t s etc.) and a n u m b e r of species groups (micro-algae, zooplankton, benthic filterfeeders etc.). It covers the western part of the Dutch Wadden Sea; it takes relations with the North Sea and the hinterland into account. U n f o r t u n a t e l y the ECOWASP study was commissioned long after the other studies had s t a r t e d and it will continue to March 1995. So there were few opportunities to steer the other studies, and at the moment of this assesment not all results are available. Especially the integration of the various NRP I studies is not yet completed.
813 7.2 R e s u l t s o f s o m e s c e n a r i o s t u d i e s
Scenario studies have been carried out with the original configuration of the E C O W A S P model, i.e. a model without salt m a r s h e s and birds. Scenarios computed concern addition and reduction of nutrients, changes in turbidity of the water, and increase of the w a t e r temperatures. Especially the later scenario is useful for this a s s e s m e n t , but the other ones m a y serve as i n t e r e s t i n g comparisons. The water t e m p e r a t u r e scenario was computed for both the normal annual course of t e m p e r a t u r e and for a temperature which was increased 20% (based on degrees Celsius). R e m a r k a b l e enough, this r a t h e r drastic increase of t e m p e r a t u r e had relatively little effect on the results of the model computations. Only slight changes occurred. The n u t r i e n t scenarios had much stronger effects, suggesting t h a t the effects of climate change for the Wadden Sea might be less important t h a n the effects of direct h u m a n interventions. The turbidity scenarios gave contradictory results. 7.3 E v a l u a t i o n
Evaluation of the integration of the studies by means of the ECOWASP model is hardly possible at this stage. A general r e m a r k is that this kind of simulations of ecosystem behaviour should be verified by experiments with ecosystems. For the Wadden Sea opportunities to do so exist. 8.
E V A L U A T I O N OF THE W A D D E N SEA S T U D I E S OF N R P I
Did the NRP I studies address the right questions? This question may be answered starting from the different aspects of climate change, but also starting from those aspects of the Wadden Sea considered to be important, either because of their key position in the system, or because of the values society attaches to them. With the present knowledge it seems that the effects of the increase of CO2 have been covered sufficiently. With respect to t e m p e r a t u r e it seems t h a t more a t t e n t i o n to the effects of t e m p e r a t u r e and other meteorological factors on key species might have been useful. For example, t e m p e r a t u r e effects on the eider duck, the most i m p o r t a n t avian predator on the benthic fauna, may be important, because in the Wadden Sea the species is n e a r the southern limit of its distribution. Also the effects on various fish species might be important, both for n o r t h e r n and for s o u t h e r n species. From the viewpoint of sealevel rise the right aspects of the Wadden Sea system seem to have been covered. Finally, it is believed t h a t UV-B increase should have had more attention, especially with regard to the organisms living on the tidal flats. From the viewpoint of key elements in the system it may be concluded t h a t tidal flat benthic microalgae and fish have received insufficient attention. Do the NRP I studies provide a coherent answer to the question of the effects of climate change? The answer is no, chiefly because the studies were conceived as separate research projects by different research groups. Hence, the studies have been carried out using very different spatial scales, respectively of the size of tidal basins, m2 's, entire Wadden Sea, entire flyway population, and again entire
814 Wadden Sea. Also the temporal scales varied widely from single tide events to centuries. Nevertheless it is possible to integrate them to some extent by using the extensive background information. Doing this it becomes possible to draw a few preliminary conclusions. The effects of increase of atmosferic CO2 are expecteded to be slight. Some changes may occur in the salt marshes, but on the scale of the Wadden Sea no important effects are expected. Changes in temperature and other meteorological factors are expected to lead to considerable changes in the species composition of the ecosystem with northern species disappearing and (more) southern species colonizing the area. No important changes are expected in ecosystem functioning, unless one or more key species change drastically in abundance. The NRP I study on bivalve molluscs provides an example of this. It is expected that the most likely scenarios for sealevel rise will not result in disappearance of tidal flats and saltmarshes within the first 100 years. Consequently the plants and animal species depending on tidal flats will not experience large changes. An exception has to be made for the migratory shorebirds which not only will undergo the influence of sealevel rise in their tidal flat habitat in winter, but also the effects of climate change in their arctic and subarctic breeding areas. UV-B increase may lead to changes in the biota of the tidal flat areas and in the saltmarshes but it is expected that this will mainly affect balances between species. It has to be stressed that the expectations summarized above are based on assumptions which could not always be verified and on the study of a limited part of a large and complicated ecosystem. In many respects further study is needed to underscore the present conclusions. The results for the Wadden Sea in general will be applicable to other European estuaries. Only sealevel rise may have very different consequences in some estuaries because of differences in the hydrography and the sedimentology. 9.
REFERENCES
Alerstam, T. and LindstrSm, A., 1990. Optimal bird migration: the relative importance of time, energy, and safety. In: Gwinner, E. (ed.). Bird Migration, Springer-Verlag, Berlin, p. 331-351. Allen, J.R.L., 1990. Salt-marsh growth and stratification: a numerical model with special reference to the Severn Estuary, SW Britain. Mar. Geol. 95: 77-96. Arp, W.J., 1991. Vegetation of a North American saltmarsh and elevated atmospheric carbon dioxide. Thesis, Free Univ. Amsterdam, 181 pp. Asjes, J., K. Zegers, P.W. van Leeuwen, A. Sleutel, M. Binsbergen, F. Biliam, E. Wagenaar, B. Brinkman, N. Dankers and K.S. Dijkema, 1992. Comparative studies on saltmarsh processes, The Netherlands. Rep. to the European Commission, 40 pp. Bakker, J.P., J. de Leeuw, K.S. Dijkema, P.C. Leendertse, H.H.T. Prins and J. Rozema, 1993. Saltmarshes along the coast of The Netherlands. Hydrobiologica 265: 73-95. Beukema, J.J., W.J. Wolff and J.J.W.M. Brouns (eds.), 1990. Expected effects of climatic change on marine coastal ecosystems. Kluwer Acad. Publ., Dordrecht, 221 pp.
815 Beukema, J.J., 1990. Expected effects of changes in winter temperatures on benthic animals living in soft sediments in coastal North Sea areas. In: Beukema, J.J., W.J. Wolff and J.J.W.M. Brouns (eds.). Expected effects of climatic change on marine coastal ecosystems. Kluwer Academic Publ., Dordrecht, p. 83-92. Boekschoten, M., 1973. Lijst van publikaties over het Waddengebied. Report Rijksinstituut voor Natuurbeheer, Leersum, 191 pp. Bossinade, J.H., J. van den Bergs and K.S. Dijkema, 1993. De invloed van de wind op het jaargemiddelde hoogwater langs de Friese en Groninger waddenkust. Rijkswaterstaat, Directie Groningen, Nota Gran 1993-2009, 22 pp. Breeman, A.M., 1990. Expected effects of changing seawater temperatures on the geographic distribution of seaweed species. In: Beukema, J.J., W.J. Wolff and J.J.W.M Brouns (eds.), 1990. Expected effects of climatic change on marine coastal ecosystems. Kluwer Academic Publ., Dordrecht, p. 69-76. Brouns, J.J.W.M., 1988. The impact of elevated carbon dioxide levels on marine and coastal ecosystems. Report Rijksinstituut voor Natuurbeheer, Texel. RIN-rapport 88/58, 101 pp. Costa, M.J., 1990. Expected effects of temperature changes on estuarine fish populations. In: Beukema, J.J., W.J. Wolff and J.J.W.M. Brouns (eds.), 1990. Expected effects of climatic change on marine coastal ecosystems. Kluwer Academic Publ., Dordrecht, p. 99-104. Craft, C.B., E.D. Seneca and S.W. Broome, 1993. Vertical accretion in microtidal regularly and irregularly flooded estuarine marshes. Estuarine Coastal and Shelf Science 37: 371-386. Curtis, P.S., B.G. Drake, P.W. Leadley, W.J. Arp and D.F. Whigham, 1989. Growth and senescence in plant communities exposed to elevated CO2 concentrations on an estuarine marsh. Oecologia 78: 20-26. Dankers, N., M. Binsbergen, K. Zegers, R. Laane and M. Rutgers van der Loeff, 1984. Transportation of water, particulate and dissolved organic and inorganic matter between a salt-marsh and the Ems-Dollard estuary, The Netherlands. Estuarine, Coastal and Shelf Science 19: 143-165. Davidson, N. and T. Piersma, 1992. The migration of Knots: conservation needs and implications. Wader Study Bull. 64 Suppl.: 198-209. Dijkema, K.S., 1987. Changes in salt-marsh area in the Netherlands Wadden Sea after 1600. In: Huiskes, A.H.L., C.W.P.M. Blom and J. Rozema (eds.)Vegetation between land and sea. Junk, Dordrecht. p. 42-49. Dijkema, K.S., J.H. Bossinade, P. Bouwsema and R.J. de Glopper, 1990. Saltmarshes in The Netherlands Wadden Sea: rising high-tide levels and accretion enhancement. In: Beukema, J.J., W.J. Wolff and J.J.W.M. Brouns (eds.), 1990. Expected effects of climatic change on marine coastal ecosystems. Kluwer Academic Publ., Dordrecht, p. 173-188. Dijkema, K.S., J. van den Bergs, J.H. Bossinade, P. Bouwsema, R.J. De Glopper and J.W.Th.M. van Meegen, 1988. Effecten van rijzendammen op de opslibbing en de omvang van de vegetatiezones in de Friese en Groninger landaanwinningswerken, (in Dutch). Rijkswaterstaat, Directie Groningen, Nota Gran 1988-2010, 130 pp. Evans, P.R., 1978. Reclamation of intertidal land: some effects on shelduck and wader populations in the Tees estuary. Verh. orn. Ges. Bayern 23: 147-168.
816 Eysink, W.D., 1992. Impact of sealevel rise on the morphology of the Wadden Sea in the scope of its ecological function - Proposed set-up of a dynamic morphological model for Wadden Sea basins and estuaries based on empirical relations. Rep. H1300, vol. 1, ISOS*2project, phase 3, Delft Hydraulics, Delft. Fretwell, S., 1972. Populations in a seasonal environment. Princeton Univ. Press, Princeton. Goss-Custard, J.D., 1988. Foraging behaviour of wading birds and the carrying capacity of estuaries. In: Sibly, R.M. and R.H. Smith (eds.). Behavioural Ecology: ecological consequences of adaptive behaviour. Blackwell, Oxford, p. 169-188. Goss-Custard, J.D., S. McGrorty and R. Kirby, 1990. Inshore birds of the soft coasts and sealevel rise. In: Beukema, J.J., W.J. Wolff and J.J.W.M Brouns (eds.). Expected effects of climatic change on marine coastal ecosystems. Kluwer Academic Publ., Dordrecht, p. 189-194. Goss-Custard, J.D., R.T. Clarke, K.B. Briggs, B.J. Ens, K.-M. Exo, C.J. Smit, A.J. Beintema, R.W.G. Caldow, D.C. Catt, N. Clark, S.E.A. le V. dit Durrell, M.P. Harris, J.B. Hulscher, P.L. meininger, N. Picozzi, R. Prys-Jones, U.N. Safriel and A.D. West, 1994. Population consequences of habitat loss in a migratory shorebird. I. Estimating model parameters. J. appl. Ecol. (in press). Greenberg, R., 1986. Competition in migrant birds in the non-breeding season. Current Ornithol. 3: 281-307. Huiskes, A.H.L., 1990. Possible effects of sealevel changes on salt-marsh vegetation. In: Beukema, J.J., W.J. Wolff and J.J.W.M. Brouns (eds.). Expected effects of climatic change on marine coastal ecosystems. Kluwer Academic Publ., Dordrecht, p. 167-172. Kamps, L.F., 1962. Mud distribution and land reclamation in the eastern Wadden shallows, Rijkswaterstaat communications nr. 4. Kramer, K.J.M., 1990. Effects of increased solar UV-B radiation on coastal marine ecosystems: an overview. In: Beukema, J.J., W.J. Wolff and J.J.W.M. Brouns (eds.). Expected effects of climatic change on marine coastal ecosystems. Kluwer Academic Publ., Dordrecht, p. 195-210. Lambeck, R.H.D., A.J.J. Sandee and L. de Wolf, 1989. Long-term patterns in the waderusage of an intertidal flat in the Oosterschelde (SW Netherlands) and the impact of the closure of an adjacent estuary. J. appl. Ecol. 26: 419-431. Louters, T. and F. Gerritsen, 1994. Het mysterie van de Wadden. Hoe een getijdesysteem inspeelt op de zeespiegelstijging. Report Rijksinstituut voor Kust en Zee (in Dutch). Lefeuvre, J.C., 1990. Ecological impact of sealevel rise on coastal ecosystems of Mont-Saint-Michel Bay (France). In: Beukema, J.J., W.J. Wolff and J.J.W.M. Brouns (eds.). Expected effects of climatic change on marine coastal ecosystems. Kluwer Academic Publ., Dordrecht, p. 139-154. Lenssen, G.(M.), 1993. Responses of C3 and C4 species from Dutch saltmarshes to atmospheric CO2 enrichment. Thesis, Vrije Universiteit, Amsterdam, 113 p. Long, S.P., 1990. The primary productivity of Puccinellia maritima and Spartina anglica: a simple predictive model of response to climatic change. In: Beukema, J.J., W.J. Wolff and J.J.W.M Brouns (eds.). Expected effects of climatic change on marine coastal ecosystems. Kluwer Academic Publ., Dordrecht, p. 33-40. Mangel, M. and Clark, C.W. 1988. Dynamic modelling in behavioral ecology. Princeton University Press, Princeton, New Jersey.
817 McLusky, D.S., D.M. Bryant and M. Elliott, 1992. The impact of land-claim on macrobenthos, fish, and shorebirds on the Forth estuary, eastern Scotland. Aquatic Conservation 2" 212-222. Meire, P.M., 1991. Effects of a substantial reduction in intertidal area on numbers and densities of waders. Acta XX Congr. Internat. Ornithol. p. 2219-2227. Misdorp, R., F. Steyaert, F. Hallie and J. de Ronde, 1990. Climate change, sealevel rise and morphological developments in the Dutch Wadden Sea, a marine wetland. In: Beukema, J.J., W.J. Wolff and J.J.W.M Brouns (eds.). Expected effects of climatic change on marine coastal ecosystems. Kluwer Academic Publ., Dordrecht, p. 123-132. Myers, J.P., 1984. Spacing behaviour of nonbreeding shorebirds. Behavior of Marine Animals 6: 271-321. Myers, J.P., R.I.G. Morrison, P.Z. Antas, B.A. Harrington, T.E. Lovejoy, M. Sallaberry, S.B. Sennar and A. Tarak, 1987. Conservation strategy for migratory species. Amer. Scientist 75: 19-26. Oost, A.P. and K.S. Dijkema, 1993. Effecten van bodemdaling door gaswinning in de Waddenzee. DLO-Instituut voor Bos- en Natuuronderzoek, Texel. IBN-rapport 025, 133 pp. (in Dutch). Peletier, H., W.W.C. Gieskes and A. Buma, submitted. Paper on UV-B effects on benthic diatoms on tidal flats. Postma, H., 1967. Sediment t r a n s p o r t and s e d i m e n t a t i o n in e s t u a r i n e environments. In: G.H. Lauff (ed), Estuaries, Am. Ass. Adv. Sci., Washington, 83: 158-179. Reed, D.J. and D.R. Cahoon, 1992. The relationship between m a r s h surface topography, hydroperiod and growth of Spartina alterniflora in a deteriorating Louisiana saltmarsh. J. Coastal Res. 8: 77-87. Rozema, J., H. Lambers, S.C. van de Geijn and M.L. Cambridge (eds.) 1993. CO2 and biosphere. Kluwer, Dordrecht, 484 pp. Rozema, J., G.M. Lenssen, R.A. Broekman and W.P. Arp, 1990. Effects of atmospheric carbon dioxide enrichment on salt-marsh plants. In: Beukema, J.J., W.J. Wolff and J.J.W.M. Brouns (eds.). Expected effects of climatic change on marine coastal ecosystems. Kluwer Academic Publ., Dordrecht, p. 49-54. Siefert, W., 1990. Sea-level changes and tidal-flat characteristics. In: Beukema, J.J., W.J. Wolff and J.J.W.M. Brouns (eds.). Expected effects of climatic change on marine coastal ecosystems. Kluwer Academic Publ., Dordrecht, p. 105-112. Smit, C.J., 1994. Alternatieve voedselbronnen voor schelpdier-etende vogels in Nederlandse getijdewateren. DLO-Instituut voor Bos- en Natuuronderzoek, Texel. IBN-rapport 077, 80 pp. (in Dutch). Stevenson, J.C., L.G. Ward and M.S. Kearney, 1986. Vertical accretion in marshes with varying rates of sealevel rise. In: Wolfe, D.A. (ed.). Estuarine variability. Acad. Press, San Diego, p. 241-259. Strain, B.R. and J. Cure (eds.), 1985. Direct effects of increasing carbon dioxide on vegetation. US Dept. of Energy, Washington DC, 287 pp. Ungar, I.A., 1978. Halophyte seed germination. Botanical Review 44: 233-264. Van Latesteijn, H.C. and R.H.D. Lambeck, 1986. The analysis of monitoring data with the aid of time-series analysis. Environm. Monit. Assess. 7: 287-297.
818 Van den Hoek, C., A.M. Breeman and W.T. Stam, 1990. The geographic distribution of seaweed species in relation to temperature: present and past. In: Beukema, J.J., W.J. Wolff and J.J.W.M. Brouns (eds.). Expected effects of climatic change on marine coastal ecosystems. Kluwer Academic Publ., Dordrecht, p. 55-68. Van Malde, J., 1992. Relative rise of mean sea-levels in The Netherlands in recent times. In: Tooley, M.J. and S. Jelgersma (eds.), Impact of sea-level rise on European Coastal Lowlands, Blackwell Publishers, Oxford, UK, p. 36-55. Van der Spek, A.J.F. and D.J. Beets, 1992. Mid-Holocene evolution of a tidal basin in the western Netherlands. Sediment. Geol. 80: 185-197. Van de Staay, J., J. Rozema and M. Stroetenga, 1990. Expected changes in Dutch coastal vegetation resulting from enhanced levels of solar UV-B. In: Beukema, J.J., W.J. Wolff and J.J.W.M. Brouns (eds.). Expected effects of climatic change on marine coastal ecosystems. Kluwer Academic Publ., Dordrecht, p. 211-218. Vooys, C.G.N. de, 1990. Expected biological effects of long-term changes in t e m p e r a t u r e s on benthic ecosystems in coastal w a t e r s a r o u n d The Netherlands. In: Beukema, J.J., W.J. Wolff and J.J.W.M. Brouns (eds.), 1990. Expected effects of climatic change on marine coastal ecosystems. Kluwer Academic Publ., Dordrecht, p. 77-82. Weber, T., Houston, A.I. and B.J. Ens. in press. Optimal departure fat loads. Proceedings of the Royal Society. Westerhoff, W.E. and P. Cleveringa, 1990. Sea-level rise and coastal sedimentation in central Noord-Holland (The Netherlands) around 5000 BP: a case study of changes in sedimentation dynamics and sediment distribution patterns. In: Beukema, J.J., W.J. Wolff and J.J.W.M. Brouns (eds.). Expected effects of climatic change on marine coastal ecosystems. Kluwer Academic Publ., Dordrecht, p. 133-138. Wilson, J.G., 1990. Effects of temperature changes on infaunal circalittoral bivalves, particularly Tellina tenuis and T. fabula. In: Beukema, J.J., W.J. Wolff and J.J.W.M Brouns (eds.), 1990. Expected effects of climatic change on marine coastal ecosystems. Kluwer Academic Publ., Dordrecht, p. 93-98. Wolff, W.J. (ed.), 1983. Ecology of the Wadden Sea. Balkema, Rotterdam, c. 2000 pp. Wolff, W.J., 1992a. Ecological developments in the Wadden Sea until 1990. In: Dankers, N., C.J. Smit and M. Scholl (eds.) - Proceedings 7th International Wadden Sea Symposium, Ameland, 1990. Net. Inst. Sea Res. Publ. Ser. 20: 23-32. Wolff, W.J., 1992b. The end of a tradition: 1000 years of reclamation of wetlands in The Netherlands. Ambio 21: 287-291. Wolff, W.J., N. Dankers, K.S. Dijkema, P.J.H. Reijnders and C.J. Smit, 1994. Biodiversity of the Wadden Sea (Denmark, Germany, The Netherlands): recent changes and future projections. In: Boyle, T.J.B. and C.E.B. Boyle (eds.) Biodiversity, temperate ecosystems and Global change. NATO Advanced Research Workshop. Springer, Berlin, p. 337-356. Wolff, W.J., A.M.M. van Haperen, A.J.J. Sandee, H.J.M. Baptist and H.L.F. Saeijs, 1975. The trophic role of birds in the Grevelingen estuary, as compared to e their role in the saline Lake Grevelingen. In: Persoone, G. et al. (eds.). Proc. 10th Eur. Symp. Mar. Biol., Ostend, Belgium, Sept. 17-23, 1975, vol. 2: 673689.
819
Discussion on the NRP assessment report "Impact of climate change on the Wadden Sea" R.S.A.R. van Rompaey Programme Office of the Dutch National Research Programme on Global Air Pollution and Climate Change NOP, c/o RIVM (pb 59), Postbus 1, NL-3720 BA Bilthoven, The Netherlands, tel: +31-30-743781; fax: +31-30-25 19 32; e-mail:
[email protected] After the clear presentation by W. Wolff (IBN-DLO), coordinator of the 'Wadden' project cluster, a short discussion took place. W. Roeleveld (Free University Amsterdam) commented that the geological sedimentation record in the Wadden Sea is worth studying from the palaeo-ecological point of view. J. Terwindt (Utrecht University) commented that further research from his team will focus on fine elements deposition and on spatial analysis of vegetation gradients. Van Rompaey asked whether climate change may cause loss of biodiversity in the Wadden Sea. Wolff answered that changes in species composition are likely to occur, but that the system does not have as high biodiversity as, e.g. dune ecosystems, and that great losses are not to be expected. However, there may occur important shifts in population size of a number of species, like mussels, which do change the appearance and functioning of the system considerably. De Boer (DGM) asked whether temperature changes on the breeding grounds in the Arctic may also have consequences for nature values in the Wadden Sea. Wolff affirmed that this certainly will be the case and that the interrelations of the Wadden Sea with the global system are part of the research questions of the NRPII project, especially considering birds.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
823
Salt M a r s h e s and sea level rise: m a r s h d y n a m i c s in relation to accretion p r o c e s s e s and accretion e n h a n c e m e n t techniques Erik-Jan Houwing and Joost H.J. Terwindt Institute for Marine and Atmospheric Research Utrecht, dept. Physical Geography, Utrecht University, P.O. Box 80115, 3508 TC Utrecht, The Netherlands
Abstract Sea-level rise will become a worldwide threat to coastal marshes by affecting the marsh vegetation through an increase in number of tidal flooding and an increase in wave energy. The survival of the salt marshes depends on the accretionary-erosional balances in both the marsh zone itself and in the pioneer zone in front of the marsh. The salt marshes along the main coast of the Netherlands Wadden Sea are stimulated by human intervention (siltation constructions) in the natural processes. Evaluation of the effects of climatic change in this area is necessary not only for predictions of future situations but also for the active management and the elaboration of mitigating measures in order to prevent salt marsh erosion.
1.
INTRODUCTION
The Wadden Sea is a shallow tidal sea of about 8,000 km 2 situated along the coast of the Netherlands, Germany and Denmark. An essential characteristic of the Wadden Sea system is the emergence of large tidal flats at low water. A part of special interest are those higher flats along the mainland and the Wadden Sea face of the barrier islands, because on these parts salt marshes are found. Much research has been done concerning the conditions of growth, accretion and erosion processes and vegetational succession in the salt marsh area. Accordingly, the vertical growth of the mudflat-marsh is determined by the rates of minerogenic and organogenic sedimentation, the frequency and period of tidal flooding and the overall compaction of the sediment (Allen, 1990; Craft et al., 1992). The sedimentation rate on the marsh is controlled by the mean tidal amplitude and is a function of the height of the mud flat and salt marsh to mean high water level (Dijkema et al., 1990). In addition, the sedimentation rate on mud flats is high in those areas which are sheltered, frequently overflooded and where the sediment supply is high. The zonation (succession) of the vegetation on the salt marsh is dependent on the surface elevation, the tidal amplitude and the drainage of the area (Armstrong et al., 1985). If the accretion of sediment on the marsh itself remains behind an increase in sea-level rise the marsh will be more frequent overflooded and the vegetation detoriation will increase (Reed and Cahoon, 1992). The Wadden Sea is a unique nature area with wildlife of many different species containing fishes, birds and mammals. The tidal flats and salt marshes act as nursery grounds for
824 migrating birds and contain extremely large numbers of benthic invertebrates and are characterized by high biomass. Most of the salt marsh area in the Wadden Sea can be found attached to the main coast. Here, the marshes stabilize the shoreline and afford protection to developed areas during storms by absorbing and dissipating wave energy, and by storage of water. A rapid increase in mean high water levels might lead to severe erosion of salt marshes and tidal fiats. A loss in the marsh area will affect the coastal protection to storm surges and will have major consequences for migratory birds and for the estuarine ecosystem in the Wadden Sea. The effects of sea level rise on the development of salt marshes can be described in a general manner as follows. The higher mean sea level causes an increase in the tidal storage volume of the flat and marsh area resulting in an extended drainage and associated outbuilding (erosion) of the gully system. In addition the distribution of wave heights over time may shift towards the higher wave domain because the increase in water depth allows higher waves to penetrate towards the fiat-marsh area. The higher wave attack normally will result in flatter profiles. Furthermore the seaward boundary of the pioneer vegetation may shift landward. In the other hand the increased frequency of tidal flooding of the marshes will result possibly in larger sedimentation rates and thus a vertical upbuilding of the salt marsh. Hence the heightening of mean sea level is expected to be associated with an erosional tendency of the fiats and the transition zone of the pioneer vegetation and with an accretional tendency of the higher marsh area. Small changes in the boundary condition for the development of tidal fiats (hydrodynamic parameters like mean sea level, wind induced wave heights and tidal currents) will have consequences for the sedimentation- and erosion-processes on the tidal fiats and in the salt marshes. The pioneerzone, in this respect, is of major importance for the salt marsh. It is transitional to the tidal fiats and it is situated on a level which is most affected by wave action. Erosion of the pioneerzone during an increase in rising high-tide levels will lead to cliff formation and marsh erosion from the seaward edge. The pioneer zone contains the pioneer vegetation Salicornia dolichostachya and an erosional tendency of this zone will be indicated by a landward retreat of this species. The present research addresses the hydrodynamic processes and expected changes in the fiat-salt marsh area in order to develop management strategies and mitigating measures.
2.
RESEARCH AREA
As early as 1930, brushwood groynes have been build along the mainland coast of the Netherlands Wadden sea to stimulate accretion of sediment. The construction of the brushwood groynes significantly increased the rate of sedimentation in the areas in between (Dijkema et al., 1990). The excavation of ditches in the sedimentation area ensured an increase in dewatering in such a way that the succession of pioneer vegetation towards higher order vegetation originated. The sedimentation rate and seaward expansion of the salt marsh area was higher during the first years of the construction of the brushwood groynes (1960-1978). After this period the salt marsh expansion arrested and even erosion within the sedimentation fields occurred. Partly this can be ascribed to a reduction of the management effort, but other factors have to play a part as well. Recent loss in salt marsh area in the Wadden sea might be due to an accelerated increase in mean high water to 0.44 cm yr -1 from 1961 to 1983 (Dijkema et al., 1990), from which about 75 % is calculated to be due to an increase in wind
825 speed between 1976 and 1983 (Bossinade et al., 1993). It is expected that changes in the hydrodynamic boundary conditions will first be noticeable in the pioneer zone. The research, therefore, is focused on the hydrodynamic processes in the pioneer zone and investigates the boundary conditions for the growth and survival of the pioneer vegetation.
3.
MEASUREMENTS
The sediment bed in the pioneer zone consists of mud and fine sand and can be classified as a cohesive bed. Erosion of a cohesive bed or deposition of cohesive material will be confined to a near bed layer and will be hard to measure in situ. Both erosion- and sedimentation- processes are controlled by the near-bed hydrodynamic processes as represented by the bed-shear stress (Partheniades, 1986). The net result of erosion- and sedimentation-processes is determined on one hand by the exerted shear stress (induced by waves and currents) and on the other hand by the strength of the bed. An intensive field campaign was started in April 1993. During this field campaign the shear stress and shear strength determining hydrodynamic parameters were measured in the pioneer zone. Two field sides have been chosen along the Dutch Wadden Sea coast which show a difference in the development of the pioneer zone and in the expansion of the salt marsh. One area is characterized by a salt marsh retreat in the last ten years whereas the other area shows a steady sedimentation in the same period. Simultaneous experiments in both areas will characterize the differences between the repetition and magnitude of the hydrodynamic processes that take place in both salt marsh areas. The hydrodynamic measurements in the pioneer zone includes the determination of wave height distribution throughout the year, tidal current velocities, wave induced current velocities, related erosion- and deposition processes and sediment transport. The germination and growth of the pioneer vegetation is monitored in both areas. Field experiments as well as laboratory experiments were carried out in order to determine the boundary conditions for the establishment of the pioneer vegetation, for the effects of future changes in hydrodynamic processes on the vegetation and the effect of pioneer vegetation on sedimentation- and erosion-processes in the salt marshes. The results are used to draw up and calibrate a Salicornia computer simulation model which describes the life cycle of this pioneer vegetation species.
4.
RESULTS
The results from this research show that sand is transported as bed load in the pioneer zone whereas mud is transported in suspension. In this respect, it is important for the salt marsh as well as for the pioneer zone how much sand or mud is available in front of the flatmarsh zone for the transportation into the salt marsh works. The Wadden morphology in front of the salt marshes partly determines the rate of sedimentation in the pioneer zone. Differences in the Wadden morphology consequently will lead to spatial differences in salt marsh development along the Wadden Coast. It was assumed that the construction of the brushwood groynes would limit wind wave growth in the pioneer zone and thereby decrease the turbulence in the near bed region and as a result decrease erosion of the sediment bed. However, measurements have shown that this
826 is not the case. Wind waves are fully developed within a distance of thirty meters from the groynes. The distance between the groynes measures at present 400 meter. The construction of groynes, however, significantly increased sedimentation. It seems likely that the groynes reduce the current induced turbulence and thus increase the deposition of suspended sediments. The germination, growth and survival of the pioneer vegetation is determined in great extend by the hydrodynamic parameters. For instance the wash away of seeds and seedlings during the first week of gemination is a function of the mobility of the top layer of the bed. This means that the seaward expansion of the pioneer vegetation is determined not only by the frequency and duration of overflooding but also by the composition (mud/sand ratio, shear strength) of the bed. This research is carried out in order to investigate the effects of the present management techniques on erosion- and sedimentation-processes during an increase in sea level rise. The results of the present investigations indicate that variable arrangements and outlines of the brushwood groynes may be applied in order to stimulate the deposition in the area. As such it may combat the negative effects of sea level rise in these areas.
5.
REFERENCES
Allen, J.R.L., 1990. Salt-marsh growth and stratification: A numerical model with special reference to the Severn Estuary, Southwest Britain. Marine Geology, vol 95, pp 77-96. Armstrong, W., E.J. Wright, S. Lythe and J.T. Gaynard, 1985. Plant zonation and the effects of the spring-neap tidal cycle on soil aeration in a Humber salt marsh. Journal of Ecology, vol 73, pp 323-339. Bossinade, J.H., J. van den Bergs and K.S. Dijkema, 1993. The influence of the wind on the annual mean high water along the Frisian and Groninger Wadden Sea coast (in Dutch). Rijkswaterstaat, Directie Groningen, Nota Gran 1993-2009, 22 p. Craft, C.B., E.D. Seneca and S.W. Broome, 1993. Vertical accretion in microtidal regularly and irregularly flooded estuarine marshes. Estuarine Coastal and Shelf Science, vol 37, pp 371-386. Dijkema, K.S., J.H. Bossinade, P. Bouwsema and R.J. de Glopper, 1990. Salt marshes in the Netherlands Wadden Sea: rising high tide levels and accretion enhancement. In:J.J. Beukema, W.J. Wolff and J.J.W.M. Brouns (eds.). Expected effects of climatic change on marine Coastal ecosystems, Kluwer, Dordrecht, pp 173-188. Partheniades, E., 1986. The present state of knowledge and needs for future research on cohesive sediment dynamics. Proceedings, 3rd International Symposium on River Sedimentation, vol III, University of Mississippi, pp 3-25. Reed, D.J. and D.R. Cahoon, 1992. The relationship between marsh surface topography, hydroperiod and growth of Spartina alterniflora in a deteriorating Louisiana Salt Marsh. Journal of Coastal Research,vol 8 ,no 1 , pp 77-87.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
827
The effect of sea level rise on a migratory wader B.J. Ens a and J.D. Goss-Custard b aInstitute for Forestry and Nature Research (IBN-DLO), P.O. Box 167, 1790 AD Den Burg, The Netherlands blnstitute for Terrestrial Ecology (ITE), Furzebrook Research Station, Wareham, Dorset BH20 5AS, United Kingdom
Abstract To investigate the possible effects of sea level rise on the migratory waders that depend on the Wadden Sea for their survival, we chose the well-studied Oystercatcher (Haematopus ostralegus) as a representative of a short-distance migrant. The total population size is modelled as an equilibrium between density-dependent mortality during winter and density-dependent reproduction during summer. Loss of winter habitat will increase mortality in winter and reduce population size. Even under extreme current scenarios of sea level rise, the negative effects on population size seem small. However, the current model ignores effects of temperature, which may affect winter habitat quality through its known effects on the population dynamics of the prey, and the possible effects of changes in breeding habitat.
1. INTRODUCTION The Wadden Sea is renowned worldwide for its importance as a breeding, wintering, and, especially, staging area for migratory waterbirds. A significant proportion of at least 52 distinct populations of 41 species of waterbirds utilize the Wadden Sea at some stage in their annual cycle and in 8 cases this involves no less than the entire world population [1]. For these birds the Wadden Sea is a vital link in their migrations along what is generally referred to as the East Atlantic Flyway. To conserve this rich nature area, the three Wadden Sea countries Denmark, Germany and the Netherlands have taken many protection measures over the last two decades. For instance, the Wadden Sea is recognized as the most important nature area of the Netherlands in the Physical Planning Decision (PKB) for the "Waddenzee" and 2000 km 2 are protected under the Nature Conservation Act. Internationally, it has been given the status of "Wetland of international importance" under the Convention of Ramsar, and it is recognized as a Biosphere Reserve by UNESCO. Very recently, a new potential threat to the system has appeared that might have disastrous consequences: global climate change. Due to the magnitude of its effects, global climate change could easily wipe out all progress that has been made so far in conserving the Wadden Sea. As formulated by Peters [2] " What is clear .. is that the climatological changes would have tremendous impact on communities and populations isolated by development and, by the middle of the next century,
828 may dwarf any other consideration in planning for reserve management." In an early appraisal of the problem for marine coastal ecosystems, Beukema et al. [3] took the necessary first steps of making an inventory of present knowledge, identifying uncertainties and formulating research needs. Although apparently nobody seriously suspects an effect of heightened levels of UV-B on migratory birds, the diversity of hypotheses on the effects of climate change on bird migration is nonetheless bewildering [4-6]. To fight this chaos we need a conceptual framework that allows us to answer a set of interrelated but precise questions, like: 1. Which aspects of the climate change or secondary effects will have the greatest impact: global warming, storms, rainfall patterns, sea level rise? 2. At what stage of the life cycle of the birds does the climate change or climate-induced effect operate: breeding season, autumn migration, spring migration, nonbreeding season? 3. Which processes must be considered: vagaries during the migratory flight, interspecific competition between resident and non-resident species, intraspecific competition on the breeding grounds, intraspecific competition on the wintering grounds, phenology of the food supply, evolution of the migration schedule? We apply the conceptual framework described by e.g. Goss-Custard [7] and Ens et al. [8], which emphasises the importance of individual-based concepts for understanding population processes, to construct mathematical models with a strong empirical basis. Ideally, the models allow us to calculate the effect on the population for a given climate scenario. Space limitations do not permit us more than a brief description of one example: the potential effect of sea level rise on the population dynamics of the oystercatcher Haematopus ostralegus, i.e. we do not report on our work on the long-distance migrants.
2. MODEL DESCRIPTION The basic idea of the model is that both reproduction in summer and mortality during winter are density-dependent. The primary source of density-dependence in summer is assumed to result from territorial exclusion, i.e. relatively fewer birds breed when the density of potential breeding birds increases. We express the proportion excluded as a k-value: k = aT + bTlog~oN, where N is the number of potential breeding birds and the slope bT measures the compressibility of the territories. When bT = 1, the territories cannot be compressed any further so that, above the number set by the intercept aT, a constant number of pairs breed, irrespective of the numbers attempting to do so. Thus, bT = 0 implies no density dependence and bT = 1 implies perfect 'contest' density dependence. We do not allow bT < 0. The primary source of density-dependence in winter is decreased survival chances due to competition for food. At low densities, there is no competition, but only a densityindependent proportion (mw) starving. When bird density increases, eventually a point, Cw, is reached at which mortality begins to increase and so becomes density-dependent. From then on, mortality increases by bw for every unit increase in bird density. We estimated these parameters for each sub-population and, where necessary, age class. Furthermore, we knew the size for each of the four sub-populations that we distinguished: Continental coastal, Continental inland, Atlantic coastal and Atlantic inland. Within the Continental and Atlantic regions, inland-breeding and coastal-breeding subpopulations use the same coastal areas in winter. Finally, the sometimes quite substantial
829 annual fluctuations in the main production and mortality parameters were generally not correlated across sites within a sub-population. This allowed the standard deviations of the annual variations in these parameters to be estimated for both sub-populations in each region so that realistic annual variations could also be included in the model. For more details we refer to [9].
3. MODEL SIMULATIONS In the model the effect of loss of winter habitat is simulated by removing an increasing proportion of the winter habitat and thus, initially, increasing the density of birds on the wintering grounds. Simulations over a range of probable parameter values show that the density at which winter mortality becomes density dependent, Cw, simply determines the point at which population size is affected as habitat is gradually removed. The population is affected sooner in the more widely fluctuating Continental sub-populations than in the less variable Atlantic sub-populations. Once winter density reaches Cw, the consequences depend on the slope, bw, of the density-dependent winter mortality function. In all sub-populations, the reduction in population size increases sharply as bw increases, but only at low values; above a certain level, further increases in bw make less difference. Because of their higher reproductive rate, inland sub-populations are initially less affected by winter habitat loss than coastal sub-populations. These conclusions are robust over a range of assumptions about competition for territories in summer and age differences in mortality in winter [10]. However, the conclusion that inland sub-populations are initially less affected by winter habitat loss depends very much on the assumption that coastal and inland birds compete on equal terms on the wintering grounds. Work is in progress to test the hypothesis that coastal birds may actually be at an advantage, because, as residents, they are likely to dominate their migratory conspecifics. It is unlikely though that our conclusions on the effects of winter habitat loss on the total population size would be greatly altered if this hypothesis was found to be true.
4. THE EFFECT OF SEA LEVEL RISE Given these and other uncertainties in parameter values it is not possible to arrive at a precise prediction of the change in population size under different climate scenarios. However, within the large range of parameter values assessed, there were no cases where population size increased following loss of winter habitat, a theoretical possibility according to Fretwell [11]. Furthermore, it was generally true that the equilibrium population size decreased by a smaller percentage than the reduction in surface of winter habitat. This was especially true for small reductions of winter habitat. Thus, if habitat loss is small, reduction in population size will be even smaller. Under the most extreme scenario of a sea level rise of 85 cm during the next century, the calculations of Louters and Gerritsen [12] indicate that the overall surface area of the Dutch Wadden Sea fiats will not change significantly in the next fifty to a hundred years; in the order of 1% of the intertidal area will be lost. Even if this estimate is out by a factor of 10 it seems unlikely on the basis of the above that the equilibrium population size would be reduced by more than 5 %.
830 5. CONCLUSION Although we think it unlikely that oystercatchers will suffer much from sea level rise, it would be premature to conclude that they are not affected by climate change. The model is able to handle changes in breeding habitat, but we did not include such changes in the scenario calculations. More importantly, we did not allow for possible changes in the quality of the remaining habitat. Simulations with an individual-based physiologically-structured game theoretic model show that deteriorating habitat quality can greatly reduce population size [13]. Thus, our current model cannot handle changes in the population dynamics of the bivalve prey. Recruitment in these prey species depends strongly on temperature [14]. We are currently extending the oystercatcher population model to include the dynamics of the prey. The extended oystercatcher population model will also allow evaluation of the effects of other man-induced habitat changes, like shellfisheries.
6. REFERENCES
10 11 12 13 14
H. Meltofte, J. Blew, J. Frikke, H.-U. R6sner, and C.J. Smit, Wader Study Group Bulletin, 74 (1994) 1. R.L. Peters, pp. 99-118 in R.L. Wyman (ed.), Global Climate Change and Life on Earth, Chapman and Hall, New York, 1991. J.J. Beukema, J.J. 1990. pp. 83-92 in J.J. Beukema et al. (Ms.), Expected Effects of Climatic Change on Marine Coastal Ecosystems, Kluwer Academic Publishers, Dordrecht, 1990. J.D. Goss-Custard, S. McGrorty and R. Kirby, pp. 189-193 in J.J. Beukema et al. (eds.), Expected Effects of Climate Change on Marine Coastal Ecosystems, Kluwer Academic Publishers, Dordrecht, 1990. R.T. Lester, and J.P. Myers, J.P., pp. 119-133 in R.L. Wyman (ed.), Chapman and Hall, New York, 1991. P. Berthold, Bird Migration: a general survey, Oxford University Press, Oxford, 1993. J.D. Goss-Custard, Bird Study 40 (1993) 81. B.J. Ens, T. Piersma and R.H. Drent, Ophelia Suppl. 6, (1994) 127. J.D. Goss-Custard, R.T. Clarke, K.B. Briggs, B.J. Ens, K.-M. Exo, C.J. Smit, A.J. Beintema, R.W.G. Caldow, D.C. Catt, N Clark, S.E.A. le V. dit Durell, M.P. Harris, J.B. Hulscher, P.L. Meininger, N. Picozzi, R. Prys-Jones, U.N. Safriel and A.D. West, J. appl. Ecol. 32 (1995) in press. J. D. Goss-Custard, R. T. Clarke, S.E.A. le V. dit Durell, R.W.G. Caldow and B.J. Ens, J. appl. Ecol. 32 (1995) in press. S.D. Fretwell, Populations in a seasonal environment, Princeton University Press, Princeton, 1972. T. Louters and F. Gerritsen, report RIKZ-94.040. J.D. Goss-Custard, R.W.G. Caldow, R.T. Clarke, S.E A. le V. dit Durell and W.J. Sutherland, J. Anim. Ecol., 64 (1995) in press c. J.J. Beukema, K. Essink, H. Michaelis and L. Zwarts, Neth. J. Sea Res. 31 (1993) 319.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
831
Winter temperature and reproductive success in shell-fish in the Dutch Wadden Sea P.J.C. Honkoop, J.J. Beukema and D. Kwast Netherlands Institute for Sea Research, P.O. Box 59, 1790 AB Den Burg, The Netherlands
Abstract Effects of mild and cold winters on the reproductive success of some species of bivalves were studied under experimental conditions. Egg production in the Balthic clam M a c o m a balthica and the cockle Cerastoderma edule was smaller after high than after low winter temperatures. Egg production in M. balthica was positively correlated with their body weight and they lost a higher proportion of their weight during mild than during cold winters. The eggs of the cockle were smaller after a mild than after a cold winter, giving rise to smaller Dlarvae. The lower quantity and quality of eggs after mild as opposed to cold winters can explain why reproductive success in bivalves is generally higher after cold than after mild winters.
1. I N T R O D U C T I O N Ever since 1970, numbers and biomass of bottom animals are recorded annually on the Balgzand, a tidal-flat area in the westernmost part of the Dutch Wadden Sea. Strong year-toyear fluctuations in biomass were observed in most bivalve species, resulting in a low stock size in some years, which then creates a conflict between fishery and nature conservancy. Fluctuations were synchronised for several species and failure of recruitment occurred in particular after mild winter in the cockle, the mussel, the Balthic clam and the gaper clam. In these species reproduction is usually highly successful after cold winters [ 1]. Two factors can be responsible for these fluctuations: 1. Due to an early appearance of predators (shrimps, shore crabs) after mild winters on the tidal flats, the predation pressure on juvenile bivalves will be higher after mild winters than after cold winters [2,3], 2. At high winter temperatures metabolic processes in bivalves proceed at a relatively high rate. As a consequence, weight loss is high during mild winters and energy available for gametogenesis is lower after mild than after cold winters. Higher weight losses during mild winters have been found in mussels and other species of bivalves [4,5]. Low adult weights may affect viability of their offspring by the reduction of size of eggs and larvae[6]. Egg size appears to be important for the survival of bivalve larvae [7]. Our aim is to find out whether such processes can be observed also in Wadden Sea bivalves. By simulation of mild and cold winters, the influence of winter temperature on quantity and quality of eggs is being studied in bivalves such as the Balthic clam, the cockle and the mussel.
832
2. MATERIAL AND METHODS To study effects of winter temperature on bivalves, we kept groups of mussels, cockles and Balthic clams at different levels of water temperature during the winter months until spawning. One group followed the water temperature in the field, the other was kept at a 2.5~ lower level. At each of the two temperature treatments, an intertidal (intermittently immersed) and a subtidal (permanently immersed) group was run. Longer immersion times result in longer daily feeding periods and less weight loss. We used the animals of the four different groups for the determination of body weights prior to spawning, egg production, egg size and size of the larvae.
3. RESULTS The results of measurements of condition index and egg production are shown in Table 1. Egg size and shell length of D-larvae are shown in Table 2. Combined results of body weights and egg production from the years 1993 and 1994 are shown in Figure 1 Table 1 Weight (standardised for animals of exactly 10 mm shell length) and egg production (means _+ SE) after four treatments. Experiments of 1994. Balthic clam cockle
weight
warm tidal cold tidal warm subtidal cold subtidal
egg no.(* 104)
8.1 _ 0.4 9.5 _+0.3 9.8 _+0.4 13.8 _+ 1.3
1.9 3.0 2.0 6.4
_+0.4 + 0.5 + 0.6 + 0.8
weight
7.8 _+0.7 8.7 _ O. 1 10.0 +_0.3 11.4 _+0.2
egg no.(* 105)
0.7 __ O. 1 2.2 _+0.8 O. 1 2.0 _+0.3
Table 2 Egg size of the Balthic clam and egg size and larval shell length of the cockle after four treatments. Experiments of 1994. Balthic clam cockle
warm tidal cold tidal warm subtidal cold subtidal
egg size (~tm)
egg size ( ~ t m )
D-larvae shell length (tam)
104.1 106.9 106.5 104.6
76.7 78.2 75.3 79.3
106.0 + 0.7 122.4 + 0.7
_+0.2 _+ 1.3 + 0.7 + 1.4
+ + + +
0.8 0.4 0.1 0.3
123.0 _+3.3
833
EGG
NUMBER xl0
4
7, 6
[]
1993
9
1994
5 4 3-J 2
o
1 0
9
[] |
7
9
!
9
8
i
9
BODY
9
|
9
i
9
I
9
i
9
|
9
10 11 12 13 14 WEIGHT
(mg)
Figure 1 Egg production in the Balthic clam as a function of body weight (at standard shell length) just before spawning. Data from experiments in two winters combined. Body weights varied in the expected way, the higher temperatures and shorter feeding time resulted in lower body weights. Egg production was in both species higher at the lower than at the higher winter temperatures (Table 1). Egg size of the Balthic clam was not influenced by winter temperature. In the cockle the eggs of the groups kept at higher winter temperatures were somewhat smaller. The shell lengths of the D-larvae were much smaller after mild than after cold winters (Table 2). The plot of egg numbers against body weights (Figure 1) showed a similar correlation as reported by others [4,5]. High body weights resulted from either good feeding conditions or low winter temperatures (compare Table 1).
4. CONCLUSIONS Apart from diminishing predation pressure, cold winters are beneficial to reproductive success in bivalves by the production of higher numbers of larger eggs, giving rise to more and bigger larvae (which are expected to be more viable than small ones).
5. REFERENCES 1 J.J. Beukema, Neth. J. Sea Res., 30 (1992) 73-79 2 J.J. Beukema, J. Exp. Mar. Biol. Ecol., 153 (1991) 97-113
834 3 4 5 6
J.J. Beukema, Mar. Ecol. Prog. Ser., 83 (1992) 157-165 U. Maung Myint and P.A. Tyler, Mar. Biol., 67 (1982) 209-223 L. Zwarts, Neth. J. Sea Res., 28(3) (1991) 231-245 B.L. Bayne, D.L. Holland, M.N. Moore, D.M. Lowe and J. Widdows, J. Mar. Biol. Ass. UK, 58 (1987) 825-841 7 J.N. Kraeuter, M. Castagna, R. van Dessel, J. Exp. Mar. Biol. Ecol., 56 (1982) 3-8
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
835
A S S E S S M E N T R E P O R T ON N R P S U B T H E M E
' ~ R E G I O N A L H Y D R O L O G Y ~'
J. Dronkers Rijkswaterstaat, National Institute for Coastal and Marine Management, RIKZ P.O. Box 20907 2500 EX The Hague The Netherlands
With contributions by: G.P. KSnnen F.R. Rijsberman B. Parmet, M. Raak
H.J.M. Lankreij er N.E.M. Asselman, H.J.A. Berendsen, J. Kwadijk and H. Middelkoop J. Postma, L.C.P.M. Stuyt and P. Kabat M. van der Drii~
KNMI, Royal Netherlands Meterological Institute, De Bilt Resource Analysis, Delft RIZA, Institute of Inland Water Management and Waste Water Treatment, Arnhem RUG, University of Groningen RUU,University of Utrecht SC-DLO, Research Institute,The Winand Staring Center, Wageningen UvA, University of Amsterdam
836
Contents Abstract 1.
Introduction
2.
Climate Scenarios 2.1 Introduction 2.2 The need for standardization of scenarios within NRP 2.3 The KNMI approach to precipitation scenarios 2.4 Conclusions
3.
L a n d use scenarios 3.1 Introduction 3.2 Method 3.3 Results 3.4 Implications
4.
Forests 4.1 Introduction 4.2 Method 4.3 Sensitivity and climatic scenario analysis 4.4 Results and conclusions
5.
Lowland hydrology 5.1 Introduction 5.2 Method 5.3 Results of simulations 5.4 Implications
6.
River d i s c h a r g e 6.1 Introduction 6.2 Method 5.3 Results 6.4 Implications
7.
Erosion 7.1 Introduction 7.2 Method 7.3 Results of simulations
8.
T r a n s p o r t and s e d i m e n t a t i o n 8.1 Introduction 8.2 Method 8.3 Results 8.4 Implications
837
0
Economy, safety, environment 9.1 9.2 9.3 9.4 9.5
Introduction Climate scenarios Impacts on safety against flooding Economic impacts Conclusions
10. References
ABSTRACT The major part of the Netherlands consists of a low-lying river delta which is very sensitive to hydrological conditions in the North-Western part of the European continent. The rivers Rhine, Meuse and Scheldt carry through this delta to the North Sea annually nearly 100 km3 of fresh water. This water originates from a drainage basin of about 185 000 km2, which is 6 times the country area. The present geography of the Netherlands has largely been shaped by this river inflow and by the sediments which are carried along. Interaction of these fluxes with North Sea hydrodynamics in a period of rising sea level has produced large lowlands, which in the past millennium have been reclaimed for agricultural, urban and industrial purposes. At present an extended hydrological infrastructure is required to contain high waters and to keep ground water and surface water tables permanently under control. Otherwise more than half of the Netherlands would be permanently or frequently flooded. In the Netherlands water management is a matter of permanent concern. The abundance of water is at the same time source of prosperity and source of vulnerability. Waterways for shipping and water supply for agriculture, industry and domestic use are essential resources for economy. The Dutch wetlands also represent a great environmental value. Changes in water supply and river discharge therefore have important impacts. More frequent occurrence of low discharge is detrimental to fluvial transport and agriculture. More frequent occurrence of high river discharge affects the safety of population against flooding and causes economical damage. Additional sedimentation raises the river beds with respect to the surrounding lowlands with possible consequences for safety and river management. The previous considerations form the basis for the sub-theme "REGIONAL HYDROLOGY" of the Dutch climate change research programme NRP. Seven research projects dealing with different aspects of the hydrological system have been selected in order to study the vulnerability of the Netherlands to climate related hydrological change. Although not all components of the hydrological system could be studied, the most important processes and relationships in the hydrological system have been addressed (see Figure 1). In the present phase of research most of the results still have an indicative character. Some projects have been completed, but others have started only recently. Nevertheless several important conclusions already emerge, which are summarized below:
838 Recent scenarios indicate greater regional climate changes t h a n a s s u m e d in the past. Most r e m a r k a b l e are the strong expected increase in w i n t e r precipitation and the increased drought in summer. With increasing CO2 concentrations forests will become more r e s i s t a n t to droughts. If biomass does not increase, the evapotranspiration will decrease, causing an increase in drainage to ground water and run-off. Drought damage to crop production will increase. Agriculture will be less affected in the low lying river delta t h a n in neighbouring regions. This m a y yield a comparative advantage for the Dutch economy. In coastal lowlands no substantial increase in saline seepage is expected, even if the sea level rises more than one meter. The discharge regime of the Rhine will change drastically. The a n n u a l variability of the discharge will strongly increase; w i n t e r discharges will increase and summer discharges will decrease. Periods of low river discharge will become more frequent and more prolonged. This will diminish the transport function of the Rhine, with serious economic consequences. The availability of cooling water for power plants will also be affected. A shortage of Rhine water will cause further intrusion of saline water in the lower river delta. The frequency of high discharges will increase. This causes more frequent inundation of the embanked floodplains in the Netherlands. It is not yet clear to what extent this will change safety from flooding. Sensitivity of soils to erosion is more affected by the expected changes in land use t h a n by climate change. The production of sediment by soil erosion may be substantially increased by changes in precipitation. Sedimentation rates on the embanked floodplains in the N e t h e r l a n d s will increase. As a consequence, polluted sediments will become an increasing environmental concern. Pollution in the catchment basins of Rhine, Meuse and Scheldt has to be kept under control. A general conclusion from the different sub-theme studies is t h a t the most serious impacts of climate change result from a shift in extreme conditions more t h a n from a shift in average climate conditions. At present no reliable indications can be given regarding changes in frequency of extreme conditions. This question should be addressed with priority in the future NRP. It appears that, in particular, a more f r e q u e n t occurrence of periods of drought will have significant economical consequences.
839
I
....
~' '"
!
I 1" sCENARIOs CLIMATE CHANGE temperature - precipitation - sea level 9~
-
i
["
3 . FORESTS
watera'nci
NARIOS ,,
cachm as'n" t
5 RIVER DISCHARGE
!cl
7. SEDIMENTATION I lowland rivers I
~
I
8.
NOMY, SAFETY, ENVIRONMENT
Figure 1.1 Structure of NRP subtheme Regional Hydrology and flow-chart of climate change impacts through Regional Hydrology 1.
INTRODUCTION
Climate change is the driving force in a chain of processes which by altering the hydrology of the Netherlands affect important economical and social values. Changes in temperature and precipitation/evapotranspiration in the catchment basin of the rivers Rhine, Meuse and Scheldt and changes in sea level will affect the water and sediment balances of the Dutch coastal lowlands. Change in land use should also be taken into account. Change in land use may occur in response to climatic factors, but also in response to social or economic factors. Changes in climate and land use both affect the retention capacity and evapotranspiration of the catchment basins and by this the variability of water supply and river discharge. There is also an impact on soil erosion and sediment loading of the rivers with possible consequences for river management. Climate related change of regional hydrology therefore has an important impact on The Netherlands, physically, economically, socially and environmentally. This leads to the questions:
Are the hydrological infrastructure and water management practices in the Netherlands adequate for coping with climate change? Which adaptation measures could be considered? The chain of hydrological processes by which climate change affects socioeconomic and environmental values is depicted in the scheme. This scheme forms the structure of the sub-theme Regional Hydrology. Each process in the chain has been addressed by a specific NRP-project. Starting point are scenarios for relevant climate related variables. Different scenarios are considered, each representing an internally consistent set of projections of global climate change at the regional
840 scale. They form basic input for the different research projects within the subtheme. Much attention has been paid to the consistency of scenario assumptions used in the different projects. To the climate change scenarios are added scenarios for change in land use. Changes in soil erosion and river discharge are studied as p r i m a r y consequences of the climate and land use scenarios. A special underlying study is devoted to the water balance of forest areas. Another special study deals with the water balance of lowlands which are affected by salt w a t e r seepage and sea level rise. As a secondary consequence of climate and land use change the sedimentation in the lowland river system is investigated. C h a n g e in river morphology lags behind the aforementioned processes and therefore is a concern for the longer term. In a final chapter some indications are presented with respect to the economic, social and environmental consequences of the impacts of climate change through the hydrological system. The results obtained sofar in the sub-theme Regional Hydrology do not yet cover the complete hydrological system of the Netherlands including all its catchment basins. Ground w a t e r has not been dealt with in a general way; only for a p a r t i c u l a r type of lowland areas the ground w a t e r balance has been studied. Impacts of climate change on the coastal marine system are not considered in the sub-theme. Table 1.1 List of projects in the subtheme "Regional Hydrology" Title
Project leader
Number
Impact of climate change on the discharge of the river Rhine
B.W.A.H. P a r m e t
850001
H.J.A. Berendsen Influence of climate change on the sedimentation of the rivers Rhine and Meuse (embanked floodplains)
850002
A.W.L. Veen
850015 9
Effect of climate change on groundwater and surface hydrology in the low-cosastal regions of The Netherlands
P. Kabat/ L.C.P.M. Stuyt
851041
The effects of climate change on transport and sedimentation of suspended load of Rhine/Meuse
H.J.A. Berendsen
851059
The impact of temperature and rainfall changes on land degradation in source areas of the suspended sediment load of the Rhine
F.J.P.M. Kwaad/ A.C. Imeson
852089
Effect of increase of C O forested landsurface
2 on
the balance of the
841 2.
CLIMATE SCENARIOS
G.P. KSnnen Royal Netherlands Meterological Institute, KNMI P.O.Box 201, 3730 AE De Bilt,The Netherlands
2.1 I n t r o d u c t i o n Climate impact studies require knowledge of the evolution of climate variables on scales ranging from local to global. Of particular importance in m a n y studies are variables like mean precipitation and its variability, sometimes down to timescales of one day or less. The present-day GCM simulations are still unable to produce reliable predictions for these. One reason for this is the coarse spatial resolution of the models in comparison which the typical size of precipitation areas, which prevents the inclusion of a physically correct description of precipitation; another reason is the present incapability to represent the large-scale atmospheric flow correctly enough to estimate from it even the large-scale precipitation, which depends essentially on the first derivative of the gradient in the flow pattern. To fill this gap, climate scenarios can be constructed. They can be defined as follows (Viner and Hulme 1993) : " A climate scenario is an internally consistent r e p r e s e n t a t i o n of a possible future climate [...], t h a t can be used to get a better understanding of the consequences of climate change [...]". 2.2 The n e e d for s t a n d a r d i z a t i o n of scenarios w i t h i n N R P Integration of results of impact studies requires a certain level of standardization of climate scenarios, particularly because different effects m a y be sensitive to different climate variables, or even to a combination of such variables. Among such variables are means, variances, autocorrelations and extremes of elements like temperature, precipitation, radiation, evaporation and so on. For hydrological studies, precipitation is a key variable. Due to the reasons mentioned before, direct GCM-prediction can not be used. A common approach to obtain absolute values of monthly means is to apply the relative change in precipitation, as obtained from GCMs for a future climate to the observed monthly means (Bultot 1988, Kwadijk 1993); an evolution in time is then obtained by scaling of the change with the GCM-predicted global temperature change. This procedure preserves in first order the internal consistency of the precipitation-global temperature relation. To obtain a scenario on e.g. daily time scale, an observed daily time series m i g h t be transformed into a time series for a future climate, just by scaling all daily values with the relative difference in monthly means. However, it is not a priori g u a r a n t e e d t h a t such a procedure results in a scenario t h a t is meteorologically consistent and hence in e.g. realistic extremes. 2.3 The KNMI a p p r o a c h to p r e c i p i t a t i o n scenarios In the KNMI method, it was attempted to produce precipitation scenarios in the form of daily time series t h a t can be considered to be representative for a future climate. These time series are obtained by transformation of an observed daily time series of t e m p e r a t u r e (T) and precipitation (RR). If required, a scenario for monthly m e a n s follows immediately from the averages of the transformed time series.
{]42 The t r a n s f o r m a t i o n is done as follows. First, a s e a s o n - d e p e n d e n t c h a n g e in t e m p e r a t u r e is applied to all observed daily t e m p e r a t u r e s T in the series. The m a g n i t u d e of the change is based on GCM information. This leads to a new series of t e m p e r a t u r e T*. Then, the daily precipitation a m o u n t s RR are changed. This is done on the bases of the observed relation b e t w e e n RR a n d T in t h e p r e s e n t climate (Figure 2.1) on wet days (threshold 0.1 mm/day), r a t h e r t h a n on bases of GCM-information. The d a t a were fitted by the following piecewise polynomical function R(T) with Temperature in Celsius degrees and daily Rainfall in mm). R(T) = exp [ 0.76 + 0.083 T ] T < 7~ R(T) = exp [ 0.76+0.083 T-0.014 (T-7)2+ 0.0007 (T-7)3 ] T > 7 ~ and T < 21 ~ R(T) = exp [-0.47+0.103 T] T > 21~
Eg. 1.
a n d the t r a n s f o r m e d precipitation RR* in the series were obtained by applying to the observed amounts each day a factor F t h a t depends on T and T* : RR* = JR(T*) / R(T)] RR = F . RR(2) This scenario, referred to as KNMI-1, a s s u m e s implicitly t h a t the m e a n T-RR relation r e m a i n s in first approximation preserved in a changing climate. Table 2.1 shows the dependence of F on T and T* for T* - T= 3~ We note t h a t the coefficient of v a r i a t i o n ( s t a n d a r d deviation divided by the mean) of the original series a n d the t r a n s f o r m e d one is the same. The plausibility of such a r e s u l t is s u p p o r t e d by t h e fact t h a t the coefficient of v a r i a t i o n of RR shows no clear dependence on temperature.
Table 2.1 Multiplying factor F for various values of t e m p e r a t u r e T w h e n there is a c o n s t a n t increase in t e m p e r a t u r e of 3~ on every day (T*=T+3~ The s t a n d a r d error of F is given in parentheses T
T*
F
-12
-9
1.28
:
:
:
:
4 6 8 10 12 14 16 18 20 22 24
7 9 11 13 15 17 19 21 23 25 27
1.28 1.21 1.08 .98 .95 .96 1.02 1.14 1.26 1.28 1.28
(.02) (.01) (.01) (.01) (.01) (.02) (.03) (.06) (.09) (.10) (.10)
(.02)
843 A second scenario, KNMI-2, a s s u m e s t h a t the R-P-T r e l a t i o n is in first a p p r o x i m a t i o n constant, where P is surface air pressure. This leads to the application of a R-T relation t h a t is explicitly pressure dependent. An additional assumption of KNMI-2 is that the difference in pressure is taken to be zero. This is analogous to the assumption t h a t in first approximation the circulation remains unchanged in a future climate. We note, t h a t this assumption does not apply to KNMI-1, as there is an implicit variation of P with T present in Figure 2.1 according to the present-day climatology. Table 2.2 compares the KNMI scenarios for seasonal means with other sources. The T* - T are taken from the Canadian Climate Centre GCM-model. The Bultot and Kwadijk scenario are based on direct model output of GCMs. In general, there is agreement between the various methods, apart from the summer. The latter is probably r e l a t e d to the absence of an accurate description of convective precipitation in the GCMs, which in our method is described by the right hand part in Figure 2.1. The scenarios can be based on the R - Tmax relation too, in which Tmax is the daily m a x i m u m temperature. This leads to other constants in Eq. 1. For further details of he various scenarios we refer to Klein Tank and Buishand (1993a; 1993b; 1994) and Buishand and Klein Tank (1994). Although the KNMI scenarios can give a plausible description of the variability of rain and t e m p e r a t u r e down to very short time scales, its spatial validity is restricted. Over an area of the Netherlands inland the R - T relations are similar, so t h a t a de Bilt-based scenario can be applied in that region. It is presumely also valid in the German lowlands. Extension to other areas like the middle and south basin of the Rhine requires analysis of the R - T relation in those regions.
844
Table 2.2 Mean seasonal changes (2* CO2-1" CO2) in t e m p e r a t u r e T (~ and precipitation R (%) in 2 KNMI climate scenarios for the Netherlands. The values are compared to model predictions of the C a n a d i a n Climate C e n t r e GCM (Boer et al., 1992; McFarlane et al., 1992) and the scenarios used by Bultot et al. (1988) and Kwadijk (1993) Season
Year
DJF
MAM
JJA
SON
KNMI-1,2 1)
+ 3.0
+ 2.3
+ 3.7
+ 3.4
+ 3.7
CCC-GCM 2) Bultot 3) Kwadijk 4) High Best Low
+ + + + +
3.0 3.2 5.7 4.3 3.2
+ 2.3 + 3.1
+ + + + +
3.7 2.5 4.3 2.7 1.7
+ 3.4 + 2.7
+ + + + +
KNMI-1,5) KNMI-25)
+ 20 + 19
+ 8 + 10
+ 6 + 10
+ 5 + 10
+ 10 + 12
CCC-GCM2) Bultot 3) Kwadijk4) High Best Low
+ + + + +
28 15 34 19 4
+ 21 + 11
+ + -
- 11 + 6
+ 2 + 7 + 32 + 11 - 10
AT (~
3.7 2.9 5.0 3.5 2.5
~WR (%)
26 10 25 4 20
1) T a k e n from the CCC-GCM. (see 2) 2) C h a n g e s predicted by the CCC-GCM; T values are for 16 l a n d g r i d p o i n t s s u r r o u n d i n g the N e t h e r l a n d s w h e r e a s R values are for 2 l a n d g r i d p o i n t s covering the Netherlands; two other high resolution GCMs (Geophysical Fluids Dynamics Laboratory GFHI and United Kingdom Meteorological Office UKHI) show s o m e w h a t higher t e m p e r a t u r e increases: about +5~ (Houghton et al., 1992). 3) Three monthly values of the Bultot scenario are averaged. The scenario is valid for Belgium and compiled from various early-GCM sources ( M a n a b e a n d Stouffer, 1980; Manabe et al., 1981; Washington and Meehl, 1983). 4) Scenario based on monthly m e a n changes predicted by the ESCAPE model for gridpoints covering the northern part of the Rhine-basin (using the IMAGE and STAGGER climate modules in ESCAPE; Rotmans, 1990; Wigley et al., 1991). The values presented are for the IPCC Business-as-Usual emission scenario (BaU). High, Best and Low denote upper 90% confidence limit, m e a n and lower
845
90% confidence limit of temperature and precipitation changes obtained with 6 different GCMs. 5) A t e m p e r a t u r e - d e p e n d e n t multiplying factor is applied to each individual rainday in the record at De Bilt (1961-1990). For KNMI-1 this factor is derived from the observed precipitation-temperature relation on wet days at De Bilt (1906-1981) as presented in Figure 2.1. The factor represents the relative change in the mean daily precipitation a m o u n t given a certain t e m p e r a t u r e perturbation AT; it ranges from 0.9 for days with T--12~ to 1.3 for days with T below 4~ and T above 21~ KNMI-1 implicitly assumes t h a t the surface air pressure P changes according to the present-day relation between P and T. For KNMI-2 this factor is derived from the observed p r e c i p i t a t i o n temperature/surface air pressure relation on wet days at De Bilt (1906-1981). The factor r e p r e s e n t s the relative change in the m e a n daily precipitation a m o u n t at distinct v a l u e s of surface air p r e s s u r e P given a c e r t a i n t e m p e r a t u r e p e r t u r b a t i o n AT; it ranges from 0.7 for days with T--17~ and P-1000 h P a to values > 2 at low and high temperatures. KNMI-2 explicitly keeps the daily surface air pressure P unchanged.
0
I
I I
9
I
i
-20
i
-10
i
I
i
0
I
10
i
i
20
i
30
T (~ Figure 2.1 Mean precipitation amounts at t e m p e r a t u r e T class intervals of 2~ for wet days (threshold 0.1 ram) at De Bilt (1906-1981). The figure is based on 15897 wet days (57% of the total n u m b e r of days). The n u m b e r of wet days in a t e m p e r a t u r e interval is 2104 for the T= 6~ class and decreases to about 10 at the extreme t e m p e r a t u r e s . The error bars indicate the estimated s t a n d a r d deviations of the means. The smooth curve represents the fitted regression relation (Equation 1)
846 2.4 C o n c l u s i o n s Precipitation scenarios with a time resolution of one day can be obtained from the empirical relation between observed precipitation amounts and temperature. The corresponding change in seasonal precipitation compares well with GCM-based scenarios, apart from the summer. The scenario can be refined by taking pressure into account as predictor, to account for systematic changes in circulation in case there are clear indications of these in GCM output. The scenarios obtained in these ways are meteorological consistent and provide a plausible description of extremes. Extension to other regions in Europe requires study of the local time series to find the geographical dependence in the results of Eq. 1.
3.
LAND USE SCENARIOS
B. P a r m e t R i j k s w a t e r s t a a t , I n s t i t u t e of I n l a n d W a t e r M a n a g e m e n t and W a s t e W a t e r T r e a t m e n t , RIZA P.O.Box 9072, 6800 ET Arnhem, The Netherlands Abstract Land use is an important p a r a m e t e r in hydrological and morphological processes. Climate change can induce changes in land use because the production and w a t e r use of crops is influenced. In the framework of a project of the I n t e r n a t i o n a l Commission for the Hydrology of the Rhine Basin, land use scenarios have been developed for the Rhine area. Besides climate change, autonomous developments were t a k e n into account, since these determine for a major p a r t the land use changes. A biophysical classification system has been designed and in combination with a crop simulation model geo-referenced information on land use potentials under present and possible future conditions is generated. The influence of climate change is mainly positive, the production increases. Autonomous developments were expressed in a Central Projection with a Plus and a Minus variant. In the Central Projection about one million hectare (10%) is vacated and comes available for other purposes t h a n agriculture or urban land. In the Minus v a r i a n t this is 3 million and in the Plus variant zero. Changed climate adds 0.2 million hectare to this, because less land is required due to the higher production. 3.1 I n t r o d u c t i o n Land use determines interception of precipitation, influences the ratio between i n f i l t r a t i o n a n d surface r u n o f f and d e t e r m i n e s to a l a r g e e x t e n t t h e evapotranspiration. It is therefore an important p a r a m e t e r in hydrological and morphological processes. An increased CO2-content and associated climate change might induce changes in land use, since growth and evapotranspiration of plants are influenced, see also Section 4 and 5. For n a t u r a l vegetation this could m e a n t h a t existing ecosystems move, alter in t h e i r species composition or even completely disappear. For agricultural crops the most i m p o r t a n t aspect is t h a t crop production may increase. Furthermore cropping patterns can change and new v a r i e t i e s can be introduced, t h a t cannot be grown u n d e r p r e s e n t climate
847 conditions. W h e t h e r changed climate conditions lead to changes in land use as described above, depends for a major part on economic, political, demographic and technical, socalled autonomous developments. As there are large uncertainties, both with respect to climate change and to autonomous developments, possible changes in land use have to be expressed in alternative scenarios. In the project 'Influence of climate change on the discharge of the river Rhine', that is coordinated by the International Commission for the Hydrology of the Rhine Basin (CHR), also the effects of land use changes are considered. Land use scenarios taking into account the effects of climate change in combination with autonomous developments were not available and have been developed as part of the CHR-project. The study has been carried out by the Winand Staring Centre at the request of The I n s t i t u t e of Inland Water M a n a g e m e n t and Waste W a t e r Treatment. In this chapter the methodology and the results of the study are presented. 3.2 M e t h o d To d e t e r m i n e the possible impacts of climate change on crop production a preliminary study was carried out (Wolf en van Diepen, 1991). The study showed t h a t the effects of a doubling of the CO2-concentration and an increase in temperature are mainly positive. Most crops grown in Western Europe are of the socalled C3 type, for which the CO2-concentration is sub-optimal. An increase in CO2 acts as a fertilizer and the assimilation rate increases. For socalled C4 crops, of which maize is the only important representative, the CO2-concentration is optimal and the increase in assimilation rate does not occur. An increase in t e m p e r a t u r e enhances the CO2 growth stimulation and increases production where temperature conditions are sub-optimal. Besides production, CO2 influences the water use efficiency. With higher CO2-concentrations, the s t o m a t a of crops have to be opened less to take up the same ammount of CO2. The water loss per s t o m a t a is less. For the overall water use of crops the increase in production counterbalances for a part the increase in water use efficiency, because the leaf surface increases.
An important conclusion of the preliminary study is that m o r e CO2, an increase in temperature and a small change in precipitation during the growing season, does not bring about limitations and even improves the circumstances for the cultivation of presently grown crops. Moreover possibilities for other crops arise. Climate change itself will however not directly generate changes in land use in the Rhine Basin. Although the changed climate boundary conditions will play a role, land use changes will be determined by autonomous developments. A farmer will only grow another crop if it is economically more profitable. It follows t h a t the autonomous developments are very important with respect to changes in land use. The study to land use scenarios for the Rhine basin was therefore divided in two parts. A biophysical and a socio-economic part. The target period is around the mid of next century when, according to the Business as Usual emission scenario of IPCC, the CO2-concentration has doubled. A best guess climate scenario for this period was derived from Kwadijk (1993). The scenario assumes an increase in t e m p e r a t u r e of 1.5~ in s u m m e r and 2~ in winter. Precipitation r e m a i n s unchanged during summer and increases with 10% during winter.
848 The biophysical part is aimed at assessing the effects of a doubling of the CO2concentration and a changed climate on crop production, crop w a t e r use and cropping calendar (Roetter, 1994; Roetter en van Diepen, 1994). The specific aim is to give geo-referenced information on land use potentials under present and possible future conditions. To cover the regional differences in climate and soil in the Rhine basin, a biophysical classification system has been developed. The changes in potential (optimal supply of water, nutrients and pesticides) and water limited yields (optimal use of nutrients and pesticides) and water use of agricultural crops have been investigated using a crop growth simulation model. Simulation results for present and possible future climate were combined into changes in land suitability and attainable yields in the Rhine Basin. The socio-economic part examines the influence of autonomous developments on land use and combines this with the results from the biophysical p a r t into scenarios or projections (Veeneklaas et al, 1994). A Central Projection describes the long-term tendency in land use and is based on secular historic trends, f u n d a m e n t a l scientific and technical principles and basic assumptions. Secular t r e n d s have been used to u n d e r p i n q u a n t i t a t i v e s t a t e m e n t s about f u t u r e developments. Scientific and technical restrictions refer mainly to a t t a i n a b l e agricultural production levels and land suitability and follow from the biophysical part. The basic assumptions are the most controversial. By referring to other studies on future developments they can be made plausible to a greater or lesser degree. In case of great u n c e r t a i n t y a Plus v a r i a n t and a Minus v a r i a n t is constructed. For the Plus variant maximum, and for the Minus variant m i n i m u m u r b a n and agricultural claims on land are assumed. The socio-economic part results in two types of land use projections. For unchanged and changed conditions a Central Projection is constructed with, if necessary a Plus and a Minus variant. 3.3 R e s u l t s
B i o p h y s i c a l p a r t ; changes in l a n d use p o t e n t i a l s A biophysical classification system containing the elements climate and soils and adapted for present and possible future conditions was not available for the Rhine basin and had to be developed. First a bioclimatic classification was designed, which was combined with a soil classification and integrated in a Geographical Information System (GIS) (Roetter, 1994). Climatic, agroclimatic and agroecological m a p s show t h a t a n n u a l m e a n t e m p e r a t u r e , precipitation and annual t e m p e r a t u r e amplitude are the m a i n factors to describe the regional differentiation of agricultural crops and n a t u r a l vegetation. The bioclimatic classification system was based on meteorological data for 53 stations and a digitized altitude map. Regression equations were derived between meteorological variables as dependent variables and combinations of altitude, longitude and latitude as independent variables. Based on the regression analysis and known classification systems the set-up of the bioclimatic system for the Rhine basin is based on: 1) annual mean temperature (seven classes); 2) annual mean temperature amplitude (four classes); 3) a n n u a l mean temperature of the coldest month (five classes); 4) annual mean precipitation (five classes).
849 The first three levels are based on regression equations and the fourth level is based on a digitized precipitation map. The equations have been derived both for p r e s e n t and possible future conditions. They are i m p l e m e n t e d in a GIS and consequently bioclimatic d a t a surfaces can be easily obtained. In total 700 combinations are possible, but only 90 occur at present in the Rhine Basin of which 25 have a surface area larger t h a n 1%. With the a s s u m e d scenario, the climate becomes more maritime/less continental and warmer. The soil suitability classification was based on a digitized soil map. Soil mapping units were clustered in four soil suitability groups based among others on slope class, soil texture, depth, moisture retention characteristics and soil genesis. Biophysical types were generated with the GIS by combining the bioclimatic types with soil suitability groups. This was done for present and possible f u t u r e conditions. It has been assumed that a change in climate does not affect the soil characteristics used in defining suitability groups. With the crop growth model WOFOST, potential and water-limited yields have been computed for seven major crops in the Rhine basin; w i n t e r wheat, silage maize, barley, oil seed, potato, sugar beet and rye grass, for present and possible future conditions (Roetter en van Diepen, 1994). Computations were carried out with meteorological data from 18 weather stations, representing the predominant base-line climatic types, and for two soil types representing the soil moisture and retention characteristics of the soil suitability groups. The crop characteristics were adapted for future conditions according to state of the art knowledge. In line with the preliminary study, the simulations with WOFOST showed that, in general, production increases. U n d e r water-limited situations, besides the CO2-fertilizer effect, the increased w a t e r use efficiency causes the production to increase. For the group of soils with an available w a t e r capacity of 70 mm, the average production for the Rhine area increased for winter wheat with 40%, of rye-grass with 33%, of sugar beets with 25% and of silage maize with 12%. It can be derived from the simulations t h a t soil and t e r r a i n characteristics in combination with a change in mean annual temperature are the main determining factors with respect to land suitability. Based on these criteria, five land suitability classes were defined: Very high, high, moderate, m a r g i n a l and unsuitable. If climate changes according to the described best guess scenario, the a r e a l percentages of land suitability classes change as described in table 3.1. The class "very high" increases from 1.3 to 38.6%. The percentages of the other classes decrease. The a s s u m e d climate change has a positive effect on the overall suitability of land for cultivation of current crops and tree species. S o c i o - e c o n o m i c p a r t ; l a n d use p r o j e c t i o n s Starting point for the land use projections is the present land use in the Rhine basin. The Rhine basin has been divided into 13 regions based on the NUTS-1 division of the European Union (EU). Land use was derived from statistics. Half of the total area of the Rhine basin is used for agriculture and about one third is covered with forest. The basin is densely populated with about 55 million people, consequently a relatively large share, 11%, is built-up land. The Central Projection is based on secular trends in the past, other surveys of the future and basic a s s u m p t i o n s including technical and scientific restrictions
850
(Veeneklaas et al., 1994). Looking at past secular trends in land use, it seems t h a t we enter a period of contraction of the agricultural area. This is founded on the ongoing productivity increases and s t a g n a t i n g demand following from the low expected population growth. Furthermore, there are m a n y parallels with other historic periods of contraction. A decline in agricultural area is also the outcome of other surveys of future land use. The rate of decline in these surveys depends on the scenario assumptions, for example free trade versus protected markets. Table 3.1 Areal percentages of land suitability classes for unchanged and changed climate conditions (Roetter en van Diepen, 1994) Land suitability class
Percentage of total area (%) Unchanged climate
Very high High Moderate Marginal Unsuitable
1.3 28.1 41.8 8.8 20.0
Change (%)
Changed climate 38.6 3.7 37.3 0.7 19.7
+ 37.3 - 24.4 - 4.5 - 8.1 - 0.3
Basic assumptions in the Central Projection for urban land use are that population growth is marginal, but the amount of urban land per i n h a b i t a n t will increase, although at a slower rate t h a n during the last 40 years. For agriculture it is a s s u m e d t h a t technical progress will go on and t h a t regional differences in ratio between actual and water limited production will level out. Around the mid of next century yield levels will have reached 90% of the water limited yield in all regions. The common m a r k e t of agricultural products within the EU will remain. Because food r e m a i n s a strategic good a completely free m a r k e t will not develop. Consequently, world trade in agricultural products will not expand dramatically and protection of own markets for food will not disappear. For agricultural production stricter environmental regulations are expected, which will however not prevent approaching the water-limited yields. F u r t h e r m o r e it is assumed t h a t in the long r u n there will be a tendency to grow crops in those parts of the Rhine Basin t h a t have the highest yields. A certain degree of diversification within the regions will however remain. To construct the projections a hierarchical scheme is applied. U r b a n land needs and n a t u r e claims as defined in national policy plans have the highest priority. Second in line are agricultural land requirements and the lowest priority is given to forest and other land use. This hierarchy is based on the price of land paid by the different categories. For agriculture a second hierarchical scheme is nested, based on the profitability and the required quality of land; Horticulture and p e r m a n e n t crops, root crops, cereals and, with the lowest priority, grassland and fodder crops. For the Rhine basin as a whole the changes are listed in table 3.2. The basic assumptions of the Central Projection result in an increase in urban land use. The
851 Plus v a r i a n t assumes an increased population growth and more u r b a n sprawl and results in a larger increase of u r b a n land use. In the Minus v a r i a n t it decreases because of a decrease in population and lower land claims per i n h a b i t a n t . N a t u r e conservation claimed by policy plans has the same position in the h i e r a r c h y as u r b a n land use. In the N e t h e r l a n d s explicit claims have been f o r m u l a t e d in the N a t u r e Policy Plan of about 10% of the agricultural area. The a r e a used for agriculture decreases in all projections. With changed climate conditions this decrease is even larger, because production levels increase a n d hence, less land is needed. The m a i n decrease is found for cereals. Outside the E U production costs are lower and f u r t h e r m o r e the physical production conditions of the R h i n e b a s i n w i t h i n t h e E U are not optimal for cereal production. The production will therefore partly shift to outside the Rhine Basin. Next in line are potatoes. For this crop strong competition is expected with E a s t e r n Europe. Only for beets a small increase in area is expected for the Central Projection and the Plus Variant, for u n c h a n g e d climate. This is mainly caused by an increase in the production of fodder beets, t h a t will be used in cattle feed in line with a development of more self-sufficiency in dairy farming. The changes in a r e a of u r b a n and a g r i c u l t u r a l land use can differ for the 13 d i s t i n g u i s h e d regions. For the region N e d e r l a n d - O o s t ( N e t h e r l a n d s - E a s t ) for example u r b a n land use increases with 37% in the Central Projection. If n a t u r e reserves are included the increase is 72%. Agriculture decreases w i t h 16% for unchanged and with 21% for changed climate conditions. In the Minus v a r i a n t the agricultural land use decreases with 35%. Besides grassland, the acreage of cereals and potato decreases.
Table 3.2 Changes in areas of u r b a n and agricultural land use for u n c h a n g e d and changed climate conditions, for three v a r i a n t s with respect to the basic a s s u m p t i o n s , for the decade 2040-2050, in million ha and percentages (Veeneklaas et al., 1994) Central Projection Land use
unchanged
Agriculture -
Urban
Urban+ agriculture
Plus v a r i a n t
Minus variant
changed unchanged
changed unchanged
changed
1.57
-
1.83
- 2.67
- 2.84
- 1.26
-
20%
-
24%
-
-
-
-
34%
37%
16%
1.52 20%
+ 0.68 32%
+ 0.68 + 32%
- 0.18 - 9%
- 0.18 - 9%
+ 1.39 +66%
+ 1.39 + 66%
-
-
1.15
-
-
-
0.13
-
-
12%
- 29%
+ 1%
-
-
0.89
9%
2.85
3.02
-31%
0.13 1%
In the C e n t r a l Projection about one million hectare would become available for other use, in the Minus v a r i a n t this is 3 million hectare and in the Plus v a r i a n t no s u b s t a n t i a l s u r p l u s w o u l d be available. C h a n g e d c l i m a t e conditions add
852 approximately 0.2 million hectare. In Germany and the French part of the Rhine basin large parts will be vacated, mainly the areas were presently cereals are grown. The vacated areas could be used for afforestation, especially if different functions like timber production, recreation and nature can be combined. Other plausible possibilities are nature reserves or mixed designation, like dispersed housing, hobby farming, etc. The production of industrial crops does not require large amounts of land and biofuel production is economically not viable. These are therefore less realistic options for the vacated land. 3.4 I m p l i c a t i o n s A doubling of the CO2-content and an increase in t e m p e r a t u r e seem to have a positive influence on crop production. The implications of a climate change as assumed in this study are however small for land use, compared to the influence of autonomous changes. In general, also without climate change, it may be expected that the area built-up land will increase but the agricultural area will decrease at a faster rate. This may offer possibilities for nature development and afforestation. Possible implications for morphological processes in the Rhine basin are briefly discussed in Sections 7 and 8 and for hydrological processes in 6.
It should be noted t h a t in this study only average changes in climate were considered. Changes in for example frost risk or extreme events such as hail storms have probably a larger influence on average yields and yield variability and consequently on land use. However, due to lack of information on changes in these phenomena, they were not taken into account. 4.
FORESTS
H.J.M. Lankreijer Department of Physical Geography, University of Groningen Kerklaan 30, 9651 NN Haren, The Netherlands Abstract The possible impact of an increase in CO2 on the hydrology of forests is evaluated using sensitivity analysis and a climate scenario on an one-dimensional model of forest hydrology. Water use of forests is affected by plant physiological and meteorological variables. Doubling of CO2 leads to a decrease of s t o m a t a l conductance, resulting in a decrease in transpiration of 10 to 30%. The evaporation of rainfall interception by the canopy is increased due to a higher leaf area index and higher temperatures. Total interception increases, but the ratio between interception and precipitation decreases. Simulating a small increase in forest canopy increases the evapotranspiration only weakly and the higher precipitation in the scenario is mainly passed on to drainage. Drought damage in summer should reduce, but winter discharge may strongly increase. 4.1 I n t r o d u c t i o n A change in the concentration of C 0 2 as well as a possible climate change will have direct and indirect effects on the water use of plants, including trees. The changing
853 concentration of ambient CO2 directly effects physiological processes in the plant. The indirect effect results from the change in meteorological variables. Forests are aerodynamically rough and are therefore strongly coupled to atmospheric conditions. As a result, changes in the atmosphere might affect forests stronger than other vegetation types. The aim of this study is to estimate the consequences of a climatic change associated with a doubling of the atmospheric CO2 concentration, for the water balance of forests. The results may be used in other studies in the subtheme Regional Hydrology. Given the direct effect of CO2 on plant physiology, the project is also part of subtheme Terrestrial Ecosystems. The water flow in forests can be divided into interception of rainfall, transpiration by the canopy and drainage to groundwater. Interception and transpiration depend on meteorological variables and characteristics of the canopy. Transpiration is regulated by the stomatal conductance. Because of the turbulent flow of air in the canopy the dependency of interception and transpiration on meteorological conditions is much stronger for forests than for low vegetation. Soil characteristics determine in general the availability of water and the rate of drainage. A simultaneous change in atmospheric CO2 concentration and in climate influences the forest ecosystem in a complex way. Photosynthesis, water use efficiency, growth, canopy structure, nutrient circulation, species composition and phenology are all affected by a climate change. The interrelated and partly unknown processes involved make an analysis of the effects difficult and the results uncertain. Also the different reactions per species makes it difficult to generalize results. Some species, like several coniferous trees, show no reaction of s t o m a t a l conductance to changed CO2 concentrations. In general, pl ant physiological studies show that an increase in CO2 results in higher growth rates, lower stomatal conductance and increased water use efficiency. To simulate the water use of a forest, a realistic model of stomatal conductance (Gs) is needed. However, the exact relation between stomatal regulation, plant physiological processes and environmental variables is not fully known. This has resulted in a variety of empirical models simulating Gs. In this study a well known empirical parameterization of Gs is applied. Given the available data and the existing uncertainties in stomatal behaviour this parameterization is believed to be adequate. However, it is expected that in the near future the stomatal regulation will be simulated more realistically. 4.2 M e t h o d
Model A one-dimensional model is developed to simulate the water balance of a forest on an one hour time scale. The model is based on the model used by Dolman (1988) to simulate the water balance of a coniferous forest and is devided into three main submodels. Transpiration is simulated using the Penman-Monteith equation. The Gash-Rutter (Gash,1979) approach is applied to simulate the interception of rainfall. The soil water balance is simulated on a daily time scale by a simple
854 bucket type model. The amount of water exceeding field capacity is considered as precipitation excess and drained. Actual s t o m a t a l conductance is calculated from solar radiation, a t m o s p h e r i c humidity, air t e m p e r a t u r e and soil w a t e r deficit using the regression equation according to Jarvis (1976) and Stewart (1988). Data Five data sets of different forests in Europe were available to calibrate the model. These have been analyzed for their potential use in this study. The data sets of the Thetford forest (1976) in England and Ede (1988/1989) in The N e t h e r l a n d s are used. The calibration of the coniferous forest in Thetford is described by S t e w a r t (1988), and the calibration of the deciduous forest in Ede is derived from Hendriks et al. (1990) and Ogink-Hendriks (1994). The datasets of Ede did not include winter measurements. As water use during winter is limited due to low t e m p e r a t u r e s and low irradiation, this restriction of data is permissible to calibrate the model. The w a t e r balance is simulated over 5 years using the KNMI-data set of'De Bilt'. This d a t a set covering 1974 - 1978, consists of hourly values of air t e m p e r a t u r e , air h u m i d i t y , global radiation, windspeed and precipitation. The totals of precipitation of these years were 992, 635, 536, 813 and 643 mm respectively; on average 724 mm. The average over 1961- 1990 is 802 mm. The period of 5 y e a r was r e l a t i v e l y dry, with 1974 a wet y e a r and 1976 a very dry one. The meteorological variables of the KNMI data set were measured above grass and are transformed to above forest conditions according to Nonhebel (1987). The forest c h a r a c t e r i s t i c s are described by the calibrated p a r a m e t e r s of the 'Ede' and Tnetford' forests.
4.3 Sensitivity and climatic scenario analysis The influence of the main model p a r a m e t e r s on interception and t r a n s p i r a t i o n were analyzed by sensitivity analysis. To integrate the results with the results of other impact research groups within the National Research Program, the scenario KNMI-2 as described in Section 2 is applied to the 'De Bilt' data. In this study this scenario is n a m e d scenario-2. The changes in t e m p e r a t u r e and precipitation are given in table 4.1. In the scenario the relative humidity is held identical to the relative humidity of the unchanged climate. Other meteorological variables are not changed. The amount of precipitation in the scenario increases strongly compared to climate scenarios described by IPCC or Kwadijk (1993). The increase in precipitation in the scenario is regarded as an increase in precipitation intensity, and not as an increase in duration. In order to apply the scenario an e s t i m a t i o n m u s t be m a d e of the forest p a r a m e t e r s in a changed climate. In particular the leaf are index (LAI) is an i m p o r t a n t parameter. According to the review by Idso and Idso (1994), doubling CO2 increases dry weight by 24% when water is not limiting, and by 58% w h e n w a t e r is limiting. Trees may even be more responsive to a CO2 increase t h a n herbaceous plants, although most experiments are done on seedlings, leafs or small trees. According to the same authors, average increase in dry weight w h e n nutrients are limiting, still amounts to 48%. With limited nutrients and high CO2 the increase will be concentrated in the roots. It is unclear whether the increase in growth is sustainable. Due to the use of unacclimated plants and leaves in most experiments, and the short periods over which m e a s u r e m e n t s are made, it is h a z a r d o u s to transfer the results of these studies to forest (Eamus and Jarvis,
855 1989). For instance, the response of the assimilation rate of acclimated plants seems to be 50% lower than of unacclimated plants due to the lack of active sinks for the assimilation products (Cure and Acock, 1986). In the scenario a modest increase of 5% in LAI and storage capacity is applied. It is expected that growth of the canopy will be limited by low nutrient availability and the maximum LAI possible considering the radiation in the canopy. The direct effect of increased CO2 is simulated by a decrease in stomatal conductance of 30%. Based on present knowledge, these changes are regarded as realistic, though variations due to varying species composition and forest site may be large. Table 4.1 Scenario use as input in model simulations. Change of actual temperature per hour in ~ and hourly precipitation in %. Number of precipitation days is unchanged
Temperature change Precipitation change scenario 2
Winter
Spring
Summer Autumn Year
3.0 19.2
2.3 9.5
3.7 16.2
3.4 10.4
3.1 13.8
4.4 R e s u l t s a n d c o n c l u s i o n s
Interception The interception of rainfall is especially sensitive to changes in evaporation rate, leaf area and related storage capacity of the canopy. Changes in temperature, air humidity and windspeed strongly affect interception. A change in air humidity of 20% results in a change in average air humidity deficit from 0.6 g/kg to 1.8 g/kg. Interception changes by about 50% for coniferous forest and by 60% for deciduous forests. An increase in storage capacity of 20% results in an increase in interception of 10% for coniferous forest and 7% for deciduous forests. In applying the scenario, the small increase in interception for both forest types (Figure 4.1) is mainly caused by the increase in storage capacity. Relative humidity is unchanged and evaporative demand of the air is hardly increasing. Although precipitation increases strongly, it hardly affects interception because the increase is concentrated in winter, when evaporation is low. As a result both forest types show a small decrease in the ratio of interception and total precipitation.
856 Precipitation
4 ,
1974
I
I
1975
1976
el
;.
1
Deciduous forest
i
i
1978
Avg
Coniferous forest
Interception
E 50
I
I
I
I
I
i
I
l
1974
197s
197e
1977
197s
--
Avg
1974
I~7s
197e
197r
lgTe
Aw
I
1975
1976
1977
1978
Avg
1974
1975
1976
1977
1978
Avg
Transpiration
1974
Precipitation excess E~ 35o
Normal
~
Scenario2
Figure 4.1 Yearly totals of precipitation, interception transpiration and excess simulated with normal climate and scenario 2
Transpiration Using the P e n m a n - M o n t e i t h equation, t r a n s p i r a t i o n of forests is sensitive to changes in m a x i m u m stomatal conductance, temperature, air humidity and soil water availability. The sensitivity of transpiration is caused by soil water. Due to limited soil water availability, transpiration is reduced after some time. So in most years, when transpiration is enhanced during winter and spring, water shortage occurs and reduces the t r a n s p i r a t i o n in summer. On the other hand, w h e n t r a n s p i r a t i o n is decreased, more water is available and t r a n s p i r a t i o n during s u m m e r is not so often limited. For some years, this results in a higher total transpiration when transpiration enhancing parameters have lower values. This includes interception. When interception is increased, less water is available for transpiration. Application of the scenario shows a small decrease in transpiration over the five year period. In dry years transpiration is limited due to low soil water content during summer. The strong increase in precipitation with a climate change leads to a higher availability of soil water, so transpiration in those years increases. But
857
due to the lower stomatal conductance, transpiration decreases by 10-30% most of the time when water availability is not limiting (Figure 4.1). Forest water balance
The annual water use of forest is simulated to change between -20% and +10 % depending on water availability. The average change over the 5 years is close to zero. Water use is increased when the forest stands on soil with low water availability. The increase in precipitation results in large precipitation excesses and a reduction in the number of days with water shortage (Table 4.2 and 4.3). In winter, when the evapotranspiration of the forest is low, the large increase in precipitation will drain almost completely to the groundwater. Large discharges can be expected, especially in deciduous forests. Figure 4.2 shows a total increase in precipitation excess of 60%, with high peaks during winter. Thetford/coniferous
5oo--
"~2oo o_ loo
... 9
....
9
o
Uljllllll,ll~'l 1
5
.
...
.
I r I I ' l I' I ~'11 I I ' l ' l ' l " l I"1 ~ 9
13
17
21
25
29
33
lll]r'll 37
41
45
It 49
53
Week
Ede/deciduous 4oo
t 350
9. . .
..
.
~ loo1 5O o -T-~ I i l l 1
5
I I 1 ] I 1 1 1 1 F7 i ~ |--I I l-I I ~ t I t .I..I.(1 9
13
17
21
25
29
33
I I vil 37
it 41
I"l
I1,
I 45
I ~.1.
II
49
53
Week
9
Scenario
2
Normal
Figure 4.2 Cumulative precipitation excess per week for Thetford and Ede forests Table 4.2 Number of days with soil water deficit above maximum for Ede deciduous forest
Normal Scenario 2
1974
1975
1976
1977
1978
Avg
0 0
39 18
46 31
13 0
22 0
24 9.8
858 Table 4.3 N u m b e r of days with soil w a t e r deficit above m a x i m u m for Thetford coniferous forest
Normal Scenario 2
1974
1975
1976
1977
1978
Avg
7 0
42 18
91 59
15 10
46 17
40.2 20.8
Implications According to the simulation study winter discharge will increase strongly and s u m m e r droughts will decrease. The increase in winter discharge results from the strong increase in precipitation and not from the decrease in transpiration, which is low in winter. Compared to deciduous forests, coniferous forests diminish winter discharge. It should be noted t h a t the results of the study are strongly dependent on the expected increase of precipitation, the decrease in stomatal conductance, the small increase in leaf area and the assumed constant relative humidity. On the short t e r m the s t o m a t a l conductance of most C3 species at elevated CO2 levels decreases, but long t e r m effects are h a r d l y known at present. Given the uncertainties in these parameters, the limits of confidence of the present study are very wide. The p r e s e n t policy to replace coniferous forest by deciduous forest to limit evaporation, will further increase drainage in a greenhouse climate. This m e a n s that frequent flooding can be expected in winter when the soil is already saturated. Like the prediction of future climate, the prediction of the impacts of climate change on w a t e r use of forest systems is hazardous. An important reason is t h a t data to validate the model are scarce. The response of trees to elevated CO2 levels might mirror t h a t of other C3-plants, but may also differ because trees are woody and perennial. Experiments with increased CO2 concentration on fully grown forest trees are recommended to improve the confidence level of the present studies. 5. L O W L A N D H Y D R O L O G Y J. Postma, L.C.P.M. Stuyt and P. Kabat Research Institute: DLO-The Winand Staring Center (SC-DLO) P.O.Box 125, 6700 AC Wageningen, The Netherlands
Abstract Dynamic computer simulation models were used to carry out scenario studies, forecasting the possible effects of sea level rise and climate change on physical processes which are crucial in regional- and agro-hydrology. These effects call for w a t e r m a n a g e m e n t m e a s u r e s on a regional scale. Attention was focused on changes in hydrology in the upper soil layers where these effects interfere with soil
859 water dynamics. A modified version of the two-dimensional groundwater flow model MOC of Konikow & Bredehoeft was used to simulate density-dependent deep groundwater flow and salt transport. Soil water dynamics and salt transport in the u n s a t u r a t e d zone were simulated with the one-dimensional model SWAP. A sea level rise of 1.2 m (worst-case scenario of IPCC, 1990), gradually imposed during a century, affects the seepage rate into polders in the studied area almost instantaneously but at a negligible rate. During the simulated period, the salinity of the seepage w a t e r r e m a i n s unaffected due to the low flow velocities of the g r o u n d w a t e r and the great path lengths to be travelled by the g r o u n d w a t e r between the coastal area and the polders. In contrast, climate change significantly affects crop production, viz. potential and actual transpiration. 5.1 I n t r o d u c t i o n
Climate change will interfere with low-coastal hydrology int two different ways, namely sea level rise and altered meteorological conditions near the land surface. Sea level rise will probably cause increased seepage rates in low-coastal regions, leading to s a l i n i z a t i o n of (shallow) ground- and surface waters. Altered meteorological conditions will affect the exchange of water and energy at the soil surface, and t h u s soil w a t e r dynamics in the u n s a t u r a t e d zone and crop production. In low-coastal regions of The Netherlands, integral water management influences the open water and the shallow groundwater systems. Agriculture, horticulture, n a t u r e conservation, domestic and industrial w a t e r supply are involved on a regional scale. As climatic change is likely to interfere with integral water supply and demand, an investigation of its possible consequences is called for. The climatic change was simulated using meteorological relationships from the KNMI (see Section 2), based upon a temperature rise of 1 ~ Simulations of the proposed t e m p e r a t u r e rise of 3 ~ (IPCC, 1990) was abandoned, because of sensitivity of the available crop varieties to the changes in t e m p e r a t u r e sums. Adapting these and other physiological plant parameters to such comparatively extreme conditions was considered to be unreliable at this time. It is to be expected t h a t varieties suited to changed conditions will be available when they become necessary. The consequences were assessed through a series of scenario studies, made with dynamic computer simulation models which were modified for this study. These studies were made in a vertical cross-section through the island of Voorne-Putten in the SW-Netherlands. This island was selected because it lies below sea level, and there is a certain amount of saline seepage there already. Also, investigations were made here earlier, providing essential data. The effect of sea level rise on saline seepage was simulated with the 2-D groundwater flow model MOC (Konikow & Bredehoeft, 1978) in cooperation with G. Oude Essink of the Technical University of Delft. Soil water dynamics and crop production were s i m u l a t e d using the SWAP model. SWAP is an i n t e g r a t e d s i m u l a t i o n tool consisting of SWACROP, a quasi 2-D model of the water (plus soluble salt) balance of a cropped soil including drainage and irrigation (Feddes et al., 1994), and WOFOST: a water-limited crop production model (van Diepen et al., 1988) made at the DLO-Centre for Agrobiological Research (CABO-DLO).
860 5.2
Methods
C l i m a t e scenarios To create a climate scenario, the methods were used t h a t are discussed in Section 2. Radiation, humidity and wind and the pattern of rainfall are assumed to r e m a i n unchanged. The increase in t e m p e r a t u r e used in the scenarios is 1 ~ resulting in a change to a n n u a l precipitation of-2% to +9%, depending on the temperature. The meteorological files of the years 1966, 1976, 1979, 1985 and 1986, r a n g i n g from very dry to very wet, were selected as input for the changed climate. Crop production and water use were calculated for these years, first without, then with the climate scenarios. The differences in production show the effect of climate change.
C a l c u l a t i n g crop p r o d u c t i o n with SWAP The island of Voorne-Putten, surface area 19025 ha., was divided into 761 subareas. Soil physical properties, open water levels, drainage properties, salinity and seepage rates, and land-use were collected for each subarea, 461 of which are cultivated. Production of the most frequently grown crops, potatoes, sugarbeets, w i n t e r w h e a t and grass, was calculated of the 461 subareas, for the five selected years, and calibrated with estimated actual harvests. Production and water use for the changed climate was then calculated by using the same years, changed by the climatic change. Higher temperatures will cause: maintenance respiration to increase; - plant organs to age faster, inhibiting daily increase and harvest total of dry matter; - t e m p e r a t u r e sums to increase faster, causing the crop to flower, m a t u r e and/or ripen (too) early. Higher atmospheric CO2 affects the crops (of the C3 plant type) by 4 i m p o r t a n t mechanisms (Wolf & van Diepen, 1993): - Leaf thickness increases, meaning specific leaf area decreases, - Light-use efficiency (crop production per unit radiation) increases, - Maximal assimilation rate increases, - The crop can absorb sufficient quantities of CO2 in a shorter time, keeping the s t o m a t a open for a shorter time, and so reducing transpiration. W a t e r use efficiency is increased this way. The simulations were done with the same crops, but with different crop-varieties assessed to give a realistic yield under the associated climatic conditions, by changing the physiological p a r a m e t e r s of the crop models, cf. Table 5.1 (BoonsPrins et al., 1993).
861 Table 5.1 C h a n g e s in p l a n t physiological p a r a m e t e r s to a d a p t to h i g h e r t e m p e r a t u r e s a n d raised CO2-1evels (from Wolf & van Diepen, 1993) specific leaf
light-use maximal temp. sum area efficiency assimilation (m 2.kg-1) (kg.ha-l.h-1 rate /J.m-2.s-1) (kg.ha-l.h-1)
temp.sum before flowering (~
surface until maturity (~
resistance (s m-l)
Winter wheat l'CO 2 2"CO 2
18.0 14.4
0.45 0.55
40 80
1048 1290
1258 1171
40
Potatoes 1"CO2 2"CO2
18.0 14.4
0.45 0.55
40 80
150 150
1550 1800
30
Grass l'CO 2 2"CO2
25.0 20.0
0.45 0.55
40 80
-
-
65
Sugarbeets 1"CO2 2"CO2
18.0 14.4
0.45 0.55
40 80
573 483
1909 2194
30
Groundwater flow modelling with MOC After several experiments, 3-D simulation of g r o u n d w a t e r flow was discontinued due to severe limitations of the available models. Instead, g r o u n d w a t e r flow was s i m u l a t e d in a v e r t i c a l l y oriented, 2-D cross-section t h r o u g h V o o r n e - P u t t e n , r u n n i n g w e s t to east, w i t h dimensions 200 m (depth) by 25 k m (length). The g r o u n d w a t e r flow model used was MOC (='Method Of Characteristics'), version 3.0 of 1989, w h i c h w a s developed by the US Geological S u r v e y (Konikow a n d Bredehoeft, 1978) as a t r a n s i e n t solute t r a n s p o r t model, including h y d r o d y n a m i c dispersion, t h r o u g h the horizontal plane. In order to suit the model for application in v e r t i c a l l y oriented cross-sections, it was a d a p t e d for d e n s i t y differences of g r o u n d w a t e r (Oude E s s i n k , 1993). In the model, the chloride c o n c e n t r a t i o n d e t e r m i n e s g r o u n d w a t e r density. The n u m b e r of grid cells is 100 (horizontal direction) by 20 (vertical direction); all cells are 250 m long by 10 m high. The g e o m e t r y of the geohydrological s y s t e m at the cross-section t h r o u g h VoorneP u t t e n is depicted in Figure 5.1. Geohydrological p a r a m e t e r s of the subsoil, initial salinities and b o u n d a r y conditions for g r o u n d w a t e r flow were derived from Wit (1987), DGV-TNO (1984), Oude E s s i n k (1993) and P o m p e r (1983). MOC requires the ratios t r a n s v e r s a l to longitudinal conductivity and dispersivity to be constant in the entire modelling domain; these were set to 0.1. Initial salinities following are shown in Figure 5.2. The following boundary conditions were imposed. The base is a no flow b o u n d a r y . Along the seaside and inland b o u n d a r i e s w h e r e h y d r o s t a t i c conditions are assumed, constant piezometric levels and salinities are m a i n t a i n e d , determined by m e a n sea level, w a t e r levels in bordering channels and the density of the water. Along the upper boundary, constant phreatic levels are m a i n t a i n e d in polder areas. These levels are determined by the w a t e r levels in open channels and
862 collector drains. At the sand dune areas a constant rate of groundwater recharge is maintained ( 180 mm.yr- 1).
C a l i b r a t i o n o f MOC The geometry of the island of Voorne-Putten imposes restrictions to the modelling of groundwater flow. Simulation accuracy in a 2-D vertical cross-sectional area is hampered by the fact that important boundary conditions to groundwater flow, i.e. pressure heads at the nearby n o r t h e r n and southern shorelines of the island cannot be incorporated in the model. Hence, calculated seepage rates will be lower t h a n observed ones, particularly in the central area of the domain where the effect of the inland and seaside boundary conditions of the groundwater velocity field are c o m p a r a t i v e l y insignificant. It was therefore decided to concentrate model calibration in the area bordering the seaside boundary where the effect of sea level rise was to be simulated. In addition, the area used for calibration was confined to the bottom layer of the upper aquitard and the first aquifer because of the high r e s i s t a n c e to flow of the lower a q u i t a r d (10000 d.; P o m p e r , p e r s o n a l communication). The calibration was made for seepage rates through the upper a q u i t a r d for the reference case, using the rates established by Wit (1987), by varying the kSAT of (groups of) model cells within ranges, derived from existing information (Figure 5.3). All seepage rates are averaged for specific subareas, mainly polders, with uniform open water levels. j -." / / ,~: ~ / y .
.......... Northsea ..... i"
dikes
dunes / '\ "~ / \ \ polders 9 t ~ ~ - . . . , - . - - , - ~ ~J//////////J//'///////J
(0.0 to -2.0 m) b e l o w s e a level ,___ \ 9 .... - .... ~,.......-;~,---,,---.................. = , , - ~ ~ ~ - ~ - ~ Spul uu_inK er_K e__ P_:_ ! _m_ult i p l e k's!j ~/~/z'/7-/7-/z~'/z/~/7"/7~,/7~/z/~, .........
~ X X X X X X X X X X X X X X X X X X ~ - : I i f f I f l I ! [ I I ! : : , C a l agl s ( t ~])
Kreftenheye
>.
,
C :D
;
4--'
....:i: ...'.:"
m/d)!fi!:~///X/////y//J/J///'///'///'///'SJJ~J~
F. ( 2 0
'-
m/d)
~J 4_.,
~/////////////////////////////~/, Kedichem
F. ( 0 . 0 0 2
m/d)
m
o
,5 ""-7- c
d)
,. 9
.
.
..~ MOC cells (200~ 10
m)
<
25
~:'.~:f.~.'
SWAP
-.:! ,.~'. "" ....X :::::::::::::::::
prescribed
flux
prescribed
pressure
marine
Maassluis
profiles
F. ( 1 0
for
crop
production
and
salt
boundary
and
salt
m/d):
of
> ~2.ZZ~.ZT/7".~.~ " aquitard
(MOC)
boundaries
name
km
(MOC)
formation
---------(with
no--flow
boundary
hydrobia
layer
(0.01
m/d)
permeability)
Figure 5.1 The sub-soil of Voorne-Putten partitioned into aquifers and aquitards. The h y d r a u l i c conductivities were found after calibration. Vertical scale = 100*horizontal scale
863
5.3 R e s u l t s of s i m u l a t i o n s
Results of simulations of g r o u n d w a t e r flow and solute transport with MOC The effect of sea level rise on saline seepage was assessed by c o m p a r i n g the modelling results of two scenarios. Both simulations r u n for 100 years, from 1994 onward, one with a gradual sea level rise of 1.2 m, and one without sea level rise. Of these simulations, the differences in velocity and salinity of the g r o u n d w a t e r are compared in the upper soil s t r a t a where the effect on land use is most pronounced. Figure 5.4 shows the differences in g r o u n d w a t e r velocities between the two 100 y e a r s i m u l a t i o n s in mg/1. Velocities below 1.10-4 m d-1 are indicated by dots. Seepage i n t e n s i t y does increase, but r e m a i n s of m i n o r i m p o r t a n c e in the total w a t e r balance compared to the a m o u n t of fresh w a t e r deliberately let in to flush saline surface water. It was decided not to incorporate any changes in seepage rate a n d g r o u n d w a t e r salinity in the associated b o u n d a r y conditions for the crop production scenario studies.
Results of Crop Production Simulation Climatic change appears to have two opposite effects on crop production: while it tends to decrease due to higher respiration at higher temperatures, it increases due a longer growing season and to higher water-use efficiency at higher CO2 levels. The net result is increased crop production, as shown in table 5.2. Table 5.2 Crop production (tons (dm) ha-l) for 4 crops in 5 years, w i t h and w i t h o u t 1 ~ climate change, calculated with SWAP. (din = dry m a t t e r , min = m i n i m u m , m a x = m a x i m u m , avg = average, w a r m e r = with w a r m e r climate) Sugarbeets (t.ha-1 dm) year & climate
Potatoes (t.ha-1 dm)
Wheat (t.ha-1 dm)
min m a x
avg
min
max
avg
1966 11.1 15.4 w a r m e r 11.6 20.0
13.8 15.8
11.2 11.6
14.4 20.0
14.4 15.8
5.4 7.4
7.1 7.1 9.7 9.4
1.5 11.9 10.3 4.8 15.3 13.5
8.2 8.2
4.0 5.1
1.7 2.5
6.0 8.2
3.4 5.1
0.3 0.1
4.1 2.1 6.1 3.2
7.4 11.5 10.0 10.2 15.4 12.8
1979 7.7 14.6 w a r m e r 10.4 17.4
12.4 14.7
9.7 9.7
15.0 20.8
13.0 14.7
4.7 6.9 6.8 8.1 10.410.1
1.0 10.8 9.8 3.8 13.3 12.0
1985 7.9 14.5 w a r m e r 10.0 15.8
12.2 13.7
9.0 10.2
15.0 21.0
11.8 13.9
6.5 7.3 7.3 5.3 10.910.1
1.6 11.0 10.0 4.1 14.6 12.6
1986 warmer
9.8 12.3
4.4 4.5
12.2 14.3
6.9 7.4
2.8 7.3 6.3 2.3 10.0 7.2
3.2 11.2 10.0 7.5 15.3 13.4
1976 warmer
1.2 2.5
7.1 14.5 9.9 15.1
min m a x avg
Grass (t.ha-1 dm) min m a x
avg
864 ~lti~ti!i[:::~iiiiiii!;i!iL ~~~:.:~
-,~..-.-.._~/
...~:.::.:~.:~ . . . . . . . . . ~:~;;;- ::;
r'd " . . . . . . . . . . . . .
~
~ e
.~~_
~o~~
...............................
..... . . .
~,,
..
........
:..
:!ii!::ii:i:!ii:!:!!i":::i::!:!::::!:::i:!- ::::!.i.:::::i:: :.::!:!:::: !:i:::":: . : i ::i : !::: .: ::":": ':.: " '::"::.":i:::::i: i:i:i:ii: ::::::""~_....'"":":' : :: mo.o
.-.
-
:::::::::::::::::::::::::ii.::i:i:iii!iii:.ii:i:!:~i!.:::.:.':::..:.i:i .: ..i:: :: i:..i ~!i::.: i: " : .::::i. .:..i::ii: .i::-:~i :ii!::i:i..i:!'::: :i.i.:i:i::... : ~-~".:::: ii:.
ii~i:i~;~i:i:~;:i:i.~:::~i.i.!~i i~:ii;!:.ii::~i .:. ~.~!:i:i :i~ i:::. :::~! i:i :~ .rT: ; :: :r:~..: ~.
!iiiiii"i
i!ili
:: .:ii:i :.:r%:i:!. Maa:sMs:-i
::!:'.~!:.
;i !:..::~::i.. :.:i!::. ::
i:ii ii:.::i::ii.:..i:ii
i.i:
i !:!:i:i ! ii !ii!!'li! !.i!~~~~,iil~i!i ii i':~1
',ii i!!i!ii!i!iii ii !~'
:::::::::::::::::::::::: :::::::::::::::::::::: :i.i:i:i: :..i:i:::::: :i.i-.:i:.:i..:i..:::i ::::::::::::::::::::::: :::::::::::::::::::::-:-:::~,~: :i.i: .i: :i:i:.:i:i.:.i:'.:i:i-'.:':i-i-i-:-...::'::::i:i:.-:: ::.. :i!!!iii!!:i!ili!!!i!il !! ! !:::!i!iiii!!i!i!iiiii!iii i:: i :ii :!iii!i!iiiii!:ii:i!.il :!! i i:i!!iii!i-iiiiii:iiiii:iii:i i:iil !ii: ii:!iii::!i:ii:ii!ii: iii:i !: .!!i :i!:i:!ii:ii:iiii:ii:i :::~i ii::!i!ii:iiiii:ii!:i[ ii !!! i~.:..!:!.'~i~~i.:i:.::iii..:!! I :: ! i: i: :~"~ i:!::ii::ii!i:il i~:~ iii:[ o.o
~.~
~.o
~.s
m.o
~2.~
is.o
i~.~
~o.o
~2.~
:,~.o
Distance from Coastline (Kin)
Figure 5.2 Isohalines of the groundwater (initial conditions) in the model MOC (mgfl)
2-
dunes
polders
1.5-
1 -
(D 03 0Q_
0.5-
c~176
1............. : :............: i
(D r,P) -0.5 qD (_ 0 -lQ_
[ Lt ........................ ................... i ............... ! '----:
-1 .5 21
0.0
I
.5
5.0
I
7.5
OLstonce
I
I
10.0
12.5
to
I
]5.0
CoostLLne
I
17.5
I
20.0
I
22.5
I
25.0
(km)
Figure 5.3 MOC-simulated seepage is compared to the seepage computed by Wit (1988) from field data. Calibration was only done for the polders close to the sea, to minimize the influence of the ignored raised open water to the north and south
865
:::::::::::::::::::::::::::::::::
:-:.:: :-:-].:!:!:!:i:i:!:!:i:!:i:i:!:?.iiiili!
-\
::~!~ii!ii~::~i~":i::~i~!ii;ii i ili::i~i!~ .i ;i~!!;~;i!~i~!i!!!:i:i:i:!:i:i:!:!
--/ . . . . . . . . . . . . . . . .
. . . . . . . . . . .
:!:!:!:!:!:!:]:!~i:::'::!~!~i#~!~#:
' .........
9. . . .
[
I ........
~!:!:i:!:i:"i.-..!iii!~ii !:i:i:i:!:!:i:i:i
~ ....
'"
9 ........
............
I" . . . . . . . . . . . . . . . . . . . . . . . . .
........
......
[
.,.
.......
[ ......... ~.-
:.!.[:.[:t.[!.::[...~::.[,.i::;~i:.:!..:"t.;.
-. ".. i.:y "..=.11"...* .r,:-).~..:...::.::::..;
9 .
.
.
.
.
.......
.
.
.
.
.
..
.
.
9
.
.
.
:.;::.s~ : ..... >.:-.
.
..
:= ....
. ...........
__~
.... ....
.
.....
..... 9. . . . . . .
.
"
...... 9
.
::.~ .... :
...
.
9
._~....._~..
:...:;; .......:
..
..
I.
:....................
..'
:.
.....
9. . . . . . .
". . . . .
i..-
.............
~-,.._.
......
.
.
"
"
"
"
..............
. . :
- _ ... . .
- .......
i .....
". . . . . . . . . . . .
i!i ;i:iii!i. iii:i.li!ii.i.i.li:iii,!i.iil.i:;ii;:il i;iiiiill iiiiiiiil iiiii?ili .!.i.iii!ii! i;i.iiiiii ii.iiiil.i 9
_ _ _
9
......
. . . . . . . . .
.. . . .
. - ! _ . ! . - :
. . . . . . . . . . . .
9 , .....--.. ~.. ~ , ~ , . . . . . .
9. . . . .
.........
- _ . _ :
_..-!
_-~ [. r--.-. [...~, [ i . . . . . . . . . . . . . .
.....
.
.
:
.
~i . . . . . . . . . . . . . . . . . . . .
:.
9__
..
...
'..
..
.
.
_}
200.0-'
.. 9
..
~
.
" 0.0
2.5
6.0
" 7.5
"
~176 ~
12.6
20.0
~
~
i
...
~oo~__.
~__. 6 o o o
i.:.::'::~
velocity-fields
. . . .
.
~ooo.... . . . . .
i~o
,.......... 4 ~
- - - - - - - ~ ' ~ - ~ o o ~
:~:~'~!:.i~.~:
::;..:.~ ~.~:~.::.
:.:::".:".:.:..::r;i.L[k!...v';:i:[.::[:~;.:.:::::::i~.:~i;:;.i : i:.:!.!;9 ..i.::. :...~::..t:.:.:.~..~ i"[:.[i...:..:". . . .
.
.
. . . . . . ..
.
.
9
after
simulating
......
.. ~=J
~s
.
.
:~.~4sooo
~
o
::.:: 9~ .
~I,--~
~
',~.~ :. ":
7
.
.
.
.
.
.
.
~
:
9
~
~
'"
O
" .
9.
....~
.
9
.
.
9 .
9
.
.
.
..
.
: ......
:
-
..
......
:.
k9
9 9
..
9
'..
.
~ -, ~-7%=/-~
"
9 '
.'~:::~:~""::f":".~~'~".--"";, ,~.... :
"
".':: . " "
'-.
. . . .
'
9
~
~------,:.-:-,::-:-:~:.,---:.:--.--:--:::--.--,---: .... %--:.
....:..k.. ":
:.....~......:
:: ~
'.:.'..:i:.
:"?:.:.:::...:i::..
::.%
::"..
9 : ......~::.. ~:.....i. :"ii.. ;i:
. .:...
:
......,ed~.
~..:
[ . ..-,,:....
--,---:
.:.. :
:ii:i.: ~:..i....:?< '
....,
.,......:
9
.-.
.
,...
I
.
.
\
~
\
9 >-,
%
~,
%.
"
.. "-,.,.,~ " " .'"i-4,
. . . . . .
.
..
.
9
. .
"
'
...9
.
. . .
9
" 9 9
~ ' %...k~.
"
9 .
:.. .
9
.:
;
i".
9
" . ..
:
~,,,,,
...
"
"m_..~L... " '
9 Maa,,
sluis
'
.
9
"',%o.._ .
~"
"
"
'
.. '
..
'
"~. "'"%,
"
"
"--~oo_ .
.~,~
. ~ -
..:
.
" " 9
. .
"
::.
.
' '
.
.
"b%.
..
9
"
" "
. i...
"
9 "
.
: " "
\%,
X.~
..:
.
.
\"~. ..
_~.~
[..
9 .
.
-~\ 9
. . . . " i......
o
9
"
" ..".".i '.".
\
--%
..
.
.
.
1~o.o
leO
~:~:: . . . . . . : ~i ,=
.
. 9
9
and
.
..
......
o
"
..
.
with
. 9
140.0
o
i
i.
9
\
.
.
.
9
.
26.0
100 years,
~
~0 -::---: 9 ~.-: .- .~.~: ,- ''~- .: ~
'
. .
o
~_~ i::
~i~...... :...."?.'~
.~ ::
.
~
i 9 e o . o
lOO
22.,5
.
.
.
20.0
(Km)
: ': eo.o
17.6
MOC
~
....................................................................................
16.0
Coastline
ll~~~':iiI"~::::, :~:.~...:., ~:::~".i"":~:.... i: ~ ~ . ~.~...'... ~ ' "
~
..~
......
from
~ ~: :.... , . ~ - ~ - ~ _ ~ ~ ~ , 9
..
" "
10.0 Distance
Figure 5.4 Differences in groundwater without sea-level rise, with
L
..
.
.
9. . . . . . . . . . . . . . . . . . . . . .
: 9
9
"
"
. " % o 4 ~ . ~
:.
~ooo <
"
--~
-'~......
...
.
,~,,,,~
"~
..
~oo
..~ ~ - - L _ _ = _
.
~
.
"
.
9
Figure 5.5 Groundwater salinity differences (%) after simulating 100 years, with and without sea-level rise, with MOC
866 5.4 I m p l i c a t i o n s A sea level rise of 1.2 m during the next century does not bring about a significant increase of saline seepage in the polders of Voorne-Putten. Instead, h u m a n intervention, i.e. m a i n t a i n i n g deep phreatic levels in these polders, and the associated land subsidence, are decisive in this respect. A temperature increase of 1~ and increased precipitation are favourable conditions for increased crop production rates, probably with only an occasional need of additional fresh water supply. If the increase of average temperature is greater than the assumed I~ it is likely that additional fresh water will be needed to balance increased evapotranspiration. Possibly the climate induced changes in the discharge regimes of the rivers Rhine and Meus will make it necessary to increase fresh water storage capacity. See also Section 6. Effects of changes of land use on agricultural water demands have been studied in Section 3. 6.
RIVER DISCHARGE
B. ParmetD, M. Raakl) and J. Kwadijk2) 1) 2)
Institute of Inland Water Management and Waste Water Treatment RIZA P.O.Box 9072, 6800 ET Arnhem, The Netherlands University of Utrecht P.O.Box 80.115, 3508 TC Utrecht, The Netherlands
Abstract Climate change influences the water balance of drainage basins in several ways. In a project of the International Commission for the Hydrology of the Rhine basin the possible consequences for the discharge regime of the Rhine are investigated. In the first phase of this project detailed models have been developed and applied for selected sub-basins and a rough water balance model has been developed for the whole Rhine basin. In this study results are presented for climate scenarios a s s u m i n g an increase in t e m p e r a t u r e of about 3~ and an increase in a n n u a l precipitation. The consequences of such a climate change are largest in the Alpine part of the Rhine basin, but are also considerable for the basin as a whole. In general the Rhine changes towards a rain-fed river. The winter discharge increases, which can have consequences for safety, and s u m m e r discharge decreases with consequences for shipping, industry, agriculture and nature. 6.1 I n t r o d u c t i o n Climate change influences the components of the water balance of drainage basins in several ways. Precipitation patterns may change and because of a higher t e m p e r a t u r e also the accumulation and melt pattern of snow. Evapotranspiration is directly influenced by an increase in temperature. In this respect also adaptions of the physiological behaviour of plants to an increased CO2-concentration are important. Changes in these water balance components will of course affect the discharge. Furthermore climate change may induce changes in land use, which is an important factor in evapotranspiration and runoff processes.
867 The river Rhine is economically the most important river of Western-Europe. Its drainage basin, see figure 6.1, covers from the source in Switzerland to the mouth in the North sea, an area of 185.000 km2 and is habitat to over 50 million people. The river is one of the most intensively navigated inland waterways in the world and is of major importance for the supply of w a t e r to large socio-economically i m p o r t a n t areas. Changes in the discharge regime can have consequences for safety and for the w a t e r availability for shipping, i n d u s t r y , domestic use, agriculture, the n a t u r a l e n v i r o n m e n t and recreational purposes. If possible changes are known, counter measures can be formulated to minimize negative effects. Against this background, the International Commission for the Hydrology of the Rhine basin (CHR) initiated in 1989 a project to assess the consequences of climate and land use changes for the discharge regime of the river Rhine. Since a proper tool for this was lacking, the main purpose of the project was to develop a water m a n a g e m e n t model for the whole Rhine basin. This model should be suitable to analyze the changes of total discharge and its distribution over the year as well as changes in height and frequency of discharge peaks. At the same time it should also be used to examine the effectiveness of counter measures. Several institutes from the Rhine r i p a r i a n states cooperate in the project (Parmet, 1993a). The Netherlands contribution is incorporated in the NRP. 6.2 M e t h o d The w a t e r m a n a g e m e n t model for the Rhine basin h a s to m e e t several requirements. To have a certain g u a r a n t e e it is also valid u n d e r the changed conditions it will be used for, the model m u s t have a physical basis. This is especially true for those processes t h a t are directly influenced by changes in climate and land use. This implies that the spatial variability within a basin must be t a k e n into account. To be able to simulate peak flows, a m i n i m u m temporal resolution of one day is required. In mountainous areas this has to be even smaller. In view of the complexity of the area and consequently of the model, it has been decided to phase the project. In the first phase different models for representative drainage basins of characteristic parts of the Rhine basin m u s t be developed. The Rhine basin has therefore been divided in three more or less distinct areas, the "alpine", the "middle mountains" and the "lowland" area. The relevant hydrological processes differ within these areas, for example snowmelt in mountainous areas versus g r o u n d w a t e r flow in the lowland. To give in short t e r m p r e l i m i n a r y estimations on the effects of climate change, in the first phase also a rough model for the whole Rhine basin m u s t be developed. In a second phase the models it is planned to improve, extend and combine the models.
Since reliable information on climate change is not yet available, climate scenarios have to be used (see Section 2). Developments in land use are, if possible, even more uncertain t h a n changes in climate. Hence also for this p a r a m e t e r scenarios have to be used, which are described in Section 3. The first phase of the CHR project is almost finalized. Several hydrological models have and are being developed but also existing models have been applied for relatively small representative basins. For the alpine area an existing hydrological model, the IRMB model, was applied for several small drainage basins (Bultot,
868 1992, Sch~idler 1992). This model computes the evapotranspiration, snow cover and melt and discharge on a daily basis. Because of the temporal resolution, it is not suited to simulate peak flows in mountainous areas. Therefore a w a t e r m a n a g e m e n t model is being developed with a time step of one hour, based on a hydrological forecasting model for the Swiss part of the Rhine basin. In the middle mountains area, a hydrological model is being developed for S a u e r basin, a sub-basin of the Mosel. This model has a very detailed spatial resolution and can operate on hourly and daily basis. The model is not yet operational. For another sub-basin, the Saar, in the near future an existing model will be applied. The lowland model is being developed for the drainage basin of the Overijsselsche Vecht ( P a r m e t , 1993b). The hydrological component of the model is used to compute the daily evapotranspiration and discharge for sub-basins. It consists of a g r o u n d w a t e r model, an u n s a t u r a t e d zone model and a rainfall-runoff model. The flow-routing component of the model combines the sub-basins and routes their discharges towards the mouth of the Overijsselsche Vecht. For the Rhine basin as a whole the w a t e r b a l a n c e model R H I N E F L O W was developed (Kwadijk, 1993). This model is designed to study the sensitivity of the discharge of the Rhine and its main tributaries for a climate change. It is a simple w a t e r balance model based on a Geographical Information System. Computations of evapotranspiration, snow melt and discharge are carried out for grid cells of 3*3 km on a monthly basis. 6.3 R e s u l t s
Effects of climate change, representative basins; Alpine area With the IRMB model the effects of a climate change for several components of the w a t e r balance were simulated for three drainage basins, Murg, Ergolz and Broye. A climate scenario as defined by Bultot was applied (Bultot, 1988). The m o n t h l y t e m p e r a t u r e and precipitation changes are given in table 6.1. The average t e m p e r a t u r e increases with 2.8~ and the annual precipitation with 54 mm (5%). Also changes in net terrestrial and global solar radiation and cloudiness are assumed. Changes in physiological behaviour of plants were not t a k e n into account. Computations were carried out for the period 1981 to 1988. Table 6.1 Monthly t e m p e r a t u r e (T) and precipitation (P) according to the Bultot scenario, used for representative alpine basins and a scenario based on the method developed by the KNMI, used for the representative lowland basin Month
J
F
M
A
M
J
J
A
S
O
N
D
TBultot, ~ 3.1 PBultot, % 1) 10 TNRP, ~ 3.0 PNRP, % 21
3.4 14 3.0 20
3.4 11 2.3 15
3.1 10 2.3 13
2.8 -1 2.3 5
2.7 -2 3.7 12
2.5 -2 3.7 11
2.3 -2 3.7 9
2.3 0 3.4 3
2.7 6 3.4 8
2.8 10 3.4 19
3.2 10 3.0 18
1) The percentual change is an average for the three basins Murg, Ergolz and Broye
869 According to the computations annual potential evapotranspiration increases with about 10%. Actual evapotranspiration increases somewhat less because during the s u m m e r period there is a slight decrease in soil moisture. Discharge increases during the winter period with about 10%. This is due to the fact that the amount of precipitation increases and less precipitation is stored as snow. F u r t h e r m o r e the a c c u m u l a t e d snow melts faster. The duration of the snow cover decreases considerably, especially below an altitude of 1500 m. Discharge in spring decreases slightly with 1%, and in summer discharge decreases with about 15%. This follows from less snowmelt, a larger e v a p o t r a n s p i r a t i o n and a slight decrease in precipitation. The total annual discharge hardly changes. The daily m a x i m u m discharge increases and the daily minimum discharge decreases.
Effects of climate change, representative basins; lowland area For the lowland area, climate scenarios were generated with the method developed by the KNMI within the NRP (see Section 2). In the same way as for the study to the water balance of forest (Section 4), a precipitation scenario was constructed based on the KNMI-2 method, with a t e m p e r a t u r e increase of about 3~ and unchanged air pressure. This resulted in an increase in annual precipitation of 13%. The monthly changes in precipitation and temperature are given in table 6.1. Compared to the other scenarios used in this study, this scenario is r a t h e r wet. Computations were carried out for a sub-basin of the Overijsselsche vecht basin, the Radewijkerbeek, for the period 1965-1990. On the one hand to stay in line with the results of other CHR-studies presented here, and on the other h a n d to illustrate its possible effects, computations have been carried out without and with taking into account changes in plant physiological characteristics. Without considering adaptions of plant physiological properties, referred to as scenario 1, the increase in t e m p e r a t u r e results in an increase of the actual evapotranspiration with 11% (55 mm). The annual increase in precipitation of 13% (102 mm) exceeds the increase in evapotranspiration, hence the precipitation excess increases. Consequently the annual discharge increases. From 6.1 it can be seen t h a t this increase is 16% (scenario 1). The winter discharge increases with 21% and the summer discharge increases too, with 9%. Although the effect of the increase in evapotranspiration is largest during summer, it does not exceed the increase in precipitation. As in the alpine areas the annual amplitude of the discharge regime increases. With this wet climate scenario no increase of problems with water shortages is expected. However the maximum discharge for the period 1965-1990 increases considerably with 29%. With this scenario problems with water surpluses could therefore be expected. As an example the daily discharge is given in figure 6.2 for present and for scenario 1 for the year 1981. The figure clearly shows the increase in the peak flow in March.
870 40 35 3O e- 25 .,.."~i .i
r (3
9"~20 0 .__ (D
c~ 15 E r-
o
10
year
winter
summer
maximum
III scenario 1 ~ scenario 2 9 scenario 3 t Figure 6.1 Changes in discharge characteristics for the Radewijkerbeek, for computations with changed climate, NRP scenario (scenario 1), with changed climate and adapted plant physiology (scenario 2) and with changed climate and changed land use (scenario 3), for the period 1965-1990 12
10
-
"0
E E 6 133 L_
-f•
I
9
I
Ib
4
__
0 1
2
3
4
5
6
7
8
9
10
11
12
Month
I-------Reference
situation "
"
" Scenario 1 I
Figure 6.2 Daily discharge of the Radewijkerbeek for the reference situation and for computations with changed climate, NRP scenario (scenario 1), for 1981
871 An increased CO2-concentration influences plant physiology. For most plants the w a t e r use efficiency increases and the biomass production increases. An increase in t e m p e r a t u r e for the t e m p e r a t e zones generally leads to an increase in production too (see also Sections 3, 4 and 5). Whether the increase in production exceeds t h e i n c r e a s e in w a t e r use efficiency a n d c o n s e q u e n t l y how e v a p o t r a n s p i r a t i o n changes, is not yet clear. Present knowledge indicates, for doubled CO2-concentrations and an increase in t e m p e r a t u r e of about 1.5 ~ C, a small decrease in evapotranspiration for most crops and forests (Roetter en van Diepen, 1994; Hendriks, 1994). For forests, even with a t e m p e r a t u r e increase of 3~ e v a p o t r a n s p i r a t i o n m a y decrease, as shown in Section 4. Based on this knowledge, plant physiological p a r a m e t e r s were provisionally a d a p t e d in the lowland model. Computations with the same climate scenario as used for scenario 1 (see table 6.1), but with adapted plant physiological characteristics, were carried out for the period 1965-1990. These computations are referred to as scenario 2. The actual evapotranspiration increased with 7%. This increase is 4% less t h a n for scenario 1, which can be explained from the increased water use efficiency of crops. Consequently the increase of the net precipitation excess is larger. The a n n u a l discharge increases therefore more, instead of 16, with 23%. The differences between scenario 1 and 2 with respect to the winter discharge are relatively small, as can be seen from figure 6.1. During this season evapotranspiration plays a minor role (see also Section 4). The increase of 25% is somewhat larger. The effects of a smaller increase in evapotranspiration strongly affect the s u m m e r discharge. Compared to scenario 1 this increases considerably more, 21% instead of 9%. The influence on the m a x i m u m discharge is small, since in such s i t u a t i o n s evapotranspiration plays a minor role.
Effects of climate change, Rhine basin C o n s e q u e n c e s for the whole Rhine b a s i n have been c o m p u t e d w i t h the R H I N E F L O W model. The sensitivity of the discharge regime was examined with a wide range of climate scenarios for the period 1956 to 1980 (Kwadijk, 1993). Here the results of computations with one scenario, the socalled BAU-best scenario, are presented. As already indicated in Section 2, this scenario in general agrees with the scenarios used for the representative basins in the alpine and lowland area. It a s s u m e s an a v e r a g e rise in t e m p e r a t u r e of 3.5~ and a small change in precipitation in s u m m e r and an increase of winter precipitation (see table 6.2). Changes in the physiological characteristics of plants were not taken into account.
872 Table 6.2 BAU-BEST scenario for t e m p e r a t u r e (T) and precipitation (P), for different areas in the Rhine basin (Kwadijk, 1993) P a r t of Rhine basin
North Middle South
Year
Summer
Winter
T,~
P, %
T,~
P, %
T,~
P, %
3.5 3.5 3.5
11 8 7
2.9 2.9 2.9
4 -1 -4
4.3 4.2 4.1
19 19 19
For the alpine part of the Rhine basin, the changes as computed with R H I N F L O W have the same direction as the results for the representative alpine basins. As can be seen from figure 6.3, the discharge during winter increases. This is caused by increased precipitation and snow melt. During s u m m e r the discharge decreases due to a smaller contribution of melt water, increased evapotranspiration and a slight decrease in precipitation. The increase in winter discharge is much larger t h a n for the representative basins, up to 100% with an average of 60%. This can be explained partly from the used scenarios. Both the increase in t e m p e r a t u r e and in precipitation is smaller for the Bultot scenario compared with the BAU-best scenario. F u r t h e r m o r e it can be explained by differences in model components, especially the snow component, and off course the considered area is not the same. The changes during s u m m e r are comparable, both for the alpine area as a whole and for the representative basins in the alpine area, a decrease of about 15% was computed. The changes for the area downstream, the middle and lowland part, are much less pronounced. The discharge increases during winter and spring and decreases during s u m m e r and a u t u m n , as can be derived from figure 6.3. The i n c r e a s e in evapotranspiration causes the soil water deficit to increase. As a result, s u m m e r d i s c h a r g e decreases, but because p a r t of the w i n t e r s u r p l u s is stored as groundwater, not until July. The water surplus during a u t u m n is partly used to replenish soil water, which explains the decrease of discharge during a u t u m n . Because the scenario used for the representative basin for the lowland is wetter, especially for the s u m m e r period, t h a n the BAU-best scenario, the changes in discharge are not directly comparable.
873
100 80 g
60
o
40
~
20
c
~
0
-20
-40
t
11
t
I
t
I
I
I
I
I
I
I
I
12
1
2
3
4
5
6
7
8
9
10
Month Alpine part
-
-
-
Middle part & lowland - ' -
-
Rhine basin outlet I
Figure 6.3 Changes in monthly discharge for the alpine part, the middle and lowland part and the outlet of the Rhine Basin, for computations with the BAU-best scenario, for the period 1956-1980 Where the Rhine enters The Netherlands, near the basin outlet, the changes in the alpine, middle and lowland part are combined. The annual changes are small, the discharge increases with 2%. However, winter and spring discharge increase with about 15%, and s u m m e r and a u t u m n discharge decrease with about 10%. Due to the changes in the alpine area the character of the river Rhine changes from a combined rain-fed/snow-fed into a rain-fed river. The discharge pattern will become less smooth and the difference between m a x i m u m and m i n i m u m flows will increase. The n u m b e r of months with low flows will increase. For example the n u m b e r of months with an average discharge below 1000 m3/s increases for the period 1956 to 1980 with 13, which is about 60%. To make an assessment about peak flows with the monthly discharges computed by R H I N E F L O W , a relation b e t w e e n a v e r a g e m o n t h l y flows and peak flows was derived (Kwadijk en Middelkoop,1994). This relation can be applied with a sufficiently small reliability interval for discharge peaks up to 7000 m3/s. Therefore the relation can be applied for the s t u d y to s e d i m e n t a t i o n processes in the river and floodplains (see Section 8). The considered period is however too short to give fundated results for the design discharge, and hence about consequences for safety. However, model results indicate t h a t the critical discharge (recurrence time of 1250 years) related to the safety s t a n d a r d of the river dikes may increase, with a m a x i m u m of 1500 m3/sec (Kwadijk en Middelkoop, 1994).
Effects o f l a n d use change The effects of changes in land use on discharge characteristics have been studied only very roughly. A first assessment showed t h a t for the entire Rhine basin the impact of land use changes on the river Rhine discharge were smaller t h a n climate
874 changes according to the Business as Usual scenario (Kwadijk,1993). The land use schematization of the model RHINEFLOW is very coarse, and in its present form it is not very suitable to study the effects of land use changes. F u r t h e r m o r e the land use scenarios for the Rhine basin as described in Section 3, were developed only very recently. Computations have therefore just been carried for the lowland area, for the sub-basin of the Radewijkerbeek. A land use scenario was generated based on the "Minus" projection for the region Netherlands- East. In this scenario in total 35% of the agricultural land is vacated. This is divided over potatoes (75%), s u g a r beets (-60%), cereals (-35%), maize (-50%) and grass (-30%). In total 10% of the agricultural area is changed into nature, the remaining 25% into forest, equally divided over coniferous and deciduous forest. The urban area, which is very small in the Radewijkerbeek sub-basin remains constant. Computations with this land use scenario were carried out for the period 1965-1990. Climate conditions w e r e a d a p t e d according to the KNMI-2 scenario. P l a n t physiological characteristics were not adapted. The c o m p u t a t i o n results, referred to as scenario 3, show an i n c r e a s e in e v a p o t r a n s p i r a t i o n of 15%. Compared to scenario 1 this is 4% larger, which is m a i n l y caused by the increase of the area coniferous forest. In Figure 6.2 the changes in discharge characteristics are given for scenario 3. Not surprisingly the increase in a n n u a l discharge is smaller t h a n in scenario 1. The increase is 9% compared to 16% for scenario 1. The effect of the increased evapotranspiration is of course largest during summer. Although the increase in precipitation still exceeds the increase in evapotranspiration, the increase in s u m m e r discharge is only 3% for scenario 3, compared to 9% for scenario 1. The effect of the increased a r e a of coniferous forest is also visible during winter. The total a m o u n t of interception of precipitation increases, and therefore the increase in discharge for scenario 3 is smaller t h a n for scenario 1, 15 respectively 21%. For the flat lowland area changes in land use mainly influence evapotranspiration. Since the extreme discharges are mainly determined by precipitation, the increase in m a x i m u m discharge for scenario 3 is comparable with scenario 1. The effects of the land use scenario are smaller t h a n those resulting from the climate scenario, but are still considerable, especially for total discharges.
6.4 Implications I n t e r i m r e s u l t s of the CHR project show t h a t climate change can h a v e considerable effects for the discharge regime of the Rhine. With the a s s u m e d scenario for the Rhine basin, the winter discharge increases considerably. This could have consequences for safety, but the models are not yet suitable to assess consequences for maximum peakflows. The contribution of water originating from snow melt from the Alps during the s u m m e r period decreases, which is an i m p o r t a n t r e a s o n for a d e c r e a s e in s u m m e r d i s c h a r g e . F u r t h e r m o r e evapotranspiration is expected to increase, which contributes also to a decrease in s u m m e r discharge. Consequently the frequency of periods with low flows increases. It should be noted that in the RHINEFLOW study the effects of increased CO2 on w a t e r use have not yet been taken into account. This would probably result in a reduced increase in evapotranspiration of agricultural crops and for forests even in a decrease as concluded by Lankreijer (1994). However this is not expected to have a large influence on the conclusions with respect to low flows, since the w a t e r originating from snow melt is dominant in such periods and Lankreijer (1994)
875 shows t h a t for dry years even for forests evapotranspiration increases. For water m a n a g e m e n t in The Netherlands an increased frequency of low flows implies increasing costs for shipping. Ships can be loaded less and have to wait longer for sluices and bridges. Costs of electricity production will increase too. To avoid e n v i r o n m e n t a l problems with the t e m p e r a t u r e of cooling water, other more expensive, production units have to be brought into operation. More frequent intrusion of salt water can cause problems for intake of water of certain polders. This m a y cause damage to agriculture. In general the changed discharge regime will also influence river morphology (see Section 8). The effects of land use changes have not yet been studied in detail. From first computations it can be concluded that for lowland areas the total discharge rather than peak discharges will be affected. For the alpine and middle mountains area it is expected that also the peak flows are influenced. Further study is required. The model R H I N E F L O W in its present form is a useful tool for sensitivity analysis. However, the simple process descriptions and the poor quality of the underlying database, limit its applicability. On the other hand the detailed models are only available for a relatively small part of the Rhine basin. To extend these models for the whole basin is a time consuming task. Therefore a promising direction is to couple the rough and the detailed models, for example with transfer functions. RHINEFLOW has to be refined in time and space for this. Furthermore the detailed models have to be applied also in other characteristic areas, to cover the variability within the Rhine basin in a better way. It is recommended to investigate the possible effects of climate change also for other i m p o r t a n t river systems, like the Meuse. A similar approach as in the CHR project can be applied. The largest u n c e r t a i n t y in climate change impact studies has to do with the climate scenarios. For decision makers the u n c e r t a i n t y interval of possible impacts should be as small as possible. It is therefore very i m p o r t a n t t h a t the development of consistent and plausible climate scenarios continues. 7.
EROSION Catchment basin
M. van der Drift and F.J.P.M. Kwaad Laboratory of Physical Geography and Soil Science, University of Amsterdam. Nieuwe Prinsengracht 130, 1018 VZ Amsterdam, The Netherlands Abstract Major source areas of the suspended sediment load of the river Rhine have been identified and the relationship between soil structure stability and climate in selected s e d i m e n t source areas has been studied. The location of the m a i n sediment source areas has been assessed with a mass balance method. Important source areas appeared to be the Aare basin and the Neckar basin. The influence of temperature on soil erodibility was investigated by comparing relevant properties of the same loess derived soil types under different meso-climatological and land
876 use conditions. No correlation of soil structural stability with climate could be established. 7.1 I n t r o d u c t i o n Each y e a r on average 3.1 million tons of suspended s e d i m e n t e n t e r The Netherlands via the Rhine. About two thirds of this sediment is deposited in The Netherlands. The other one-third reaches the North Sea, part of which moves along the Dutch coast to the Wadden Sea. This suspended sediment has important consequences for management and policy development which relate to: - the sediment budget of the lower courses of the Rhine, including the embanked floodplains ('uiterwaarden'), - the sediment budget of the Delta waters, the Dutch part of the North Sea and the Wadden Sea, - the scale of m u d dredging works in the Rhine, including the R o t t e r d a m harbours, - problems of water quality and pollution caused by chemicals (heavy metals, toxic organic compounds, nutrients) adsorbed to suspended sediment particles and present in recent mud deposits.
The suspended sediment load of the Rhine derives from erosion of the beds and banks of the river and its tributaries, but mostly from the valley side slopes and the sloping interfluve areas between the numerous first and second order branches of the Rhine system. The sediment from the valley side slopes and the interfluve areas is t r a n s f e r r e d to the river channels by a set of processes, collectively referred to as the 'slope forming processes'. These include processes of slope wash under n a t u r a l vegetation and processes of accelerated soil erosion (sheet and rill erosion) on agricultural land. The occurrence and rate of soil erosion is controlled by a number of factors, viz. ability of rainfall to cause erosion (erosivity), resistance of the soil to erosion (erodibility), length and steepness of slopes and land use. Climate change may, directly or indirectly, affect all of these factors. For this study the impact of climate change on soil erodibility was selected, because this is an under-developed research area. Soil erodibility is primarily controlled by soil structural stability. Climate affects soil structure through its influence on the organic m a t t e r status of the soil, which depends on biomass production and soil (micro)biological activity. Soil structure is characterized by the presence of soil aggregates, clusters of soil particles which m u t u a l l y adhere by chemical and physical binding forces. In surface soils, these forces are mainly controlled by organic matter. Macroaggregates (>250 ~m) are mainly stabilized by plant roots and larger fungi. Microaggregates (20-250 ~m) are bound together by decomposed organic substances. Micro-aggregation in the size class 2-22 ~m is mainly caused by clay particles, and to a lesser extent by organic materials. The rate of structure development and structure breakdown is dependent on the dynamics of soil organic matter, which, in turn, is controlled by soil moisture and soil temperature regime. The impact of climate change on soil erosion will be largest where soils are most susceptible to erosion. A class of soils that are highly sensitive to erosion, are loess soils, which are wide-spread in the Rhine basin. Objectives of research were: (a) to identify the main source areas of the suspended s e d i m e n t load of the Rhine under present-day climatic conditions, and (b) to
877 analyze the influence of t e m p e r a t u r e on soil erodibility in selected source areas of suspended sediment. The approach was a comparative study of soil erodibility in different parts of the Rhine basin with different temperature conditions under the current climate. Existing differences in erodibility between areas with different t e m p e r a t u r e regimes are an indication of the change in soil erodibility t h a t will possibly occur as a consequence of climate change. Starting date of the research project was 1st J a n u a r y 1993. Duration is two years. In this report results of the first year of study are summarized. On the global level the research project is related to three IGBP core projects which deal with the hydrological and geomorphological study of soil erosion and river basin dynamics: - L a n d - O c e a n I n t e r a c t i o n s in the Coastal Zone (activity: c a t c h m e n t basin dynamics and delivery), - Biospheric Aspects of the Hydrological Cycle (activity: biospheric control of waterborne transport, and integrating waterborne transport at the river-basin scale), - Past Global Changes (effects of climate and h u m a n impacts on the biosphere). On the national level there is a strong relationship with the research projects carried out at the State University of Utrecht and Rijkswaterstaat/RIZA. This concerns the study of the w a t e r balance and water discharge of the Rhine basin, the movement of sediment in the Rhine, the sedimentation rate of suspended solids on the e m b a n k e d floodplain of the Rhine and Meuse, and their sensitivity to climate change (Asselman and Middelkoop; Parmet et al., this volume). 7.2 M e t h o d s
Location o f s u s p e n d e d sediment sources The suspended sediment t h a t enters The Netherlands at Lobith originates from the part of the Rhine basin between the Bodensee and Lobith (see figure 7.1). This part of the Rhine basin has a surface area of roughly 159,000 kin2 which shows a wide range of climatic, geologic, geomorphologic and pedologic conditions. The river itself is regulated by man. Sediment is trapped by weirs and in n a t u r a l and manmade lakes. In some cases sediment is removed from the river system by dredging. It was outside the scope of the project to locate the origin of the suspended s e d i m e n t w i t h the aid of t r a c i n g or f i n g e r - p r i n t i n g t e c h n i q u e s involving mineralogical or chemical analyses. Instead, a mass balance approach was followed. The basis of such an approach is given by long-term m e a s u r e m e n t s of daily w a t e r discharge and suspended sediment concentration at a n u m b e r of stations along the Rhine and its main tributaries. The differences in sediment load between m e a s u r e m e n t points are due to contributions from sediment sources between m e a s u r e m e n t points or to losses due to sedimentation (and dredging) between points. This method does not differentiate between n a t u r a l and h u m a n sources of sediment. Therefore, additional data of the suspended sediment output of the w a s t e w a t e r s of the F r e n c h P o t a s s i u m Mining (MdPA) were used to complete the balance.
878 t"~
I 120 km
THE
/ NETHERLANDS ~'" 9 LO61TH~."
N
Laco~
9
Neuchatel SWITZERLAND
Figure 7.1 -~-~" The Rhine basin u p s t r e a m of Lobith (The Netherlands). Figure adapted from Kwadijk and Middelkoop (1994)
I d e n t i f i c a t i o n o f c l i m a t e effects on soil s t r u c t u r e s t a b i l i t y The effect of climate on soil structure was studied on a meso-scale by comparing loess soils on north and south-facing slopes and on a macro scale by comparing loess soils from a part of the Rhine basin with a more continental climate (Kraichgau) to a part with a more maritime climate (Nordrhein-Westfalen), while keeping other environmental factors, such as geology and topography, constant. Besides climatologically induced variations, also differences in soil s t r u c t u r e between land use types (arable land and forest) and between topsoil and subsoil were investigated (see figure 7.2). Soil samples were tested on presence of lime, organic carbon content, aggregate stability (drop test), soil texture and micro-aggregation (Microscan). With statistical methods of data analysis (cluster analysis, Analysis of Variance, MannWhitney U-test) differences between groups of samples were evaluated (Van der Drift, 1994).
879
Continental climate loess soils
Different aspect and altitude
[arable soils
forest soils I
> Isubsoilsl
Maritime climate loess soils
Different aspect and altitude
l arable soils
forest soils I
> Isubsoils]
Figure 7.2 Schematic representation of field sampling strategy 7.3 R e s u l t s a n d c o n c l u s i o n s Sediment
sources
The major sources of suspended sediment in the Rhine have been identified from the s h a r e of the s e d i m e n t a m o u n t at Lobith in table 7.1. Rate of erosion (ton/ha/year) is calculated by dividing this share by the area of the river basin. This is an indication of the severity of erosion in the catchment. The sediment loads of the sources u p s t r e a m of Iffezheim (km.334) are reduced with 40% for sediment retention upstream of the dams in the Rhine. Of the h u m a n sediment sources, the industrial waste-waters from the French potassium mining and G e r m a n soda industries contribute to nearly 0.5 Mton suspended sediment/year at Lobith, or 15% (corrected for sedimentation). The n a t u r a l sediment sources in the Rhine basin are characterized by a yearly cycle: in s u m m e r the Aare is the main contributing river; in winter and spring the Neckar, Main and Mosel have a strong influence. If we look at the annual totals of table 7.1, the Swiss Aare system is an important source of suspended sediment, with a corrected mean annual contribution of 0.55 Mton/year, which is equal to an average soil loss of 0.3 ton/ha/year. The Main has the second-highest sediment contribution (0.5 Mton/year), but a slightly lower erosion figure t h a n the Neckar, which delivers a mean annual sediment output of 0.45 Mton, or 15% of the load at Lobith. This m e a n s a soil loss of 0.3 ton/ha/year. No data were available for a separate calculation of the contribution of the Mosel and the Lahn.
880 Table 7.1 Major sources of suspended sediment in the Rhine
Source
Sediment load Mt/y
A m o u n t at Lobith* Mt/y
erosion+ %
Bodensee Thur Aare,Reuss,Limmat Trib. Basel-Maxau Neckar Main Mosel and L a h n Trib. Maxau-Rees MdPA Other industries Households Algae
0 0.2 0.91 0.16 0.45 0.50 0.70 0.17 0.65 0.09 0.02 0.01
0 0.12 0.55 0.096 0.45 0.50 0.70 0.17 0.39 0.09 0.02 0.01
0 3.9 18 3.1 15 16 23 5.5 13 2.9 0.65 0.32
Total
4.32
3.1
100
ton/ha/year
0 0.7 0.3 9 0.3 0.2 0.2 9 n.a. n.a. n.a. n.a.
*: corrected for sedimentation u p s t r e a m of dams (40%) .: unknown +: calculated from 'amount at Lobith' n.a.: does not apply D a t a w e r e received from t h e B u n d e s a n s t a l t ffir Gew~isserkunde, Koblenz, Germany
Soil s t r u c t u r a l stability a n d soil erodibility S u m m a r i e s of aggregate stability analyses with the drop test m e t h o d are given in tables 7.2 - 7.5. No statistically significant difference in aggregate stability w a s found b e t w e e n two climatologically different p a r t s of the Rhine b a s i n (Van der Drift, 1994), a l t h o u g h the d a t a s u g g e s t t h a t soils in a n a r e a w i t h a m o r e continental climate have more stable aggregates t h a n soils under a more m a r i t i m e climate. The t e s t s give an indication of the p r e s e n t state of soil s t r u c t u r e . This does not exclude a difference between the development of aggregate.
881 Table 7.2 S u m m a r y of drop-test data for Kraichgau and NRW topsoil samples, May-June 1993. Statistics are listed for all samples, and arable land only Region Average AS St. dev. AS St. skewness AS Number of samples
Kraichgau
arable
NRW
arable
1.08 0.873 3.39 25
1.27 0.557 - 0.0867 15
1.55 1.02 0.630 14
1.75 0.961 -0.135 4
Table 7.3 Summary of drop-test data for topsoil samples from arable and forest soils, MayJune 1993 Land use Average AS St. dev. AS St. skewness AS Number of samples
Arable
Forest
1.46 0.874 2.65 25
0.850 0.943 2.64 14
Table 7.4 Summary of drop-test data for topsoil and subsoil samples, May-June 1993 Soil horizon Average AS St. dev. AS St. skewness AS Number of samples
Topsoil
T-arable
Subsoil
S-arable
1.011 0.690 1.68 34
1.37 0.661 0.657 19
2.49 1.22 -0.405 6
2.65 1.58 -0.607 3
882 Table 7.5 S u m m a r y of drop-test data for topsoil samples from North- and South-facing slopes, May-June 1993 Aspect
South
S-arable
North
N-arable
Average AS St. dev. AS St. skewness AS N u m b e r of samples
0.771 0.543 0.612 12
1.16 0.593 -0.142 9
1.05 0.556 0.942 12
1.47 0.725 1.07 8
AS = index of aggregate stability; NRW = Nordrhein-Westfalen (Source of tables 7.2 - 7.5: Van der Drift, 1994)
Stability, due to differences in climate On a micro scale, aggregates are slightly more stable on south- and southwestfacing slopes (warm, dry) compared with northerly- and n o r t h e a s t e r l y exposed slopes (cold, moist). However, this difference was not large enough to become s t a t i s t i c a l l y significant. If soils with different land use are compared, the conclusion d r a w n from the M a n n - W h i t n e y test (Van der Drift, 1994) is t h a t aggregate stability is highly influenced by differences in land use. Soils with agricultural land use have a less stable structure t h a n forest soils. Forests are different from arable fields: they have a more shady, constant, t e m p e r a t e and moist climate, which results in more organic matter, with a different composition. Not only the amounts of organic matter, but also the composition and dynamics of organic materials, including soil biologic activity are important controlling factors of soil structure. This is in agreement with the fact that the topsoils have a more stable s t r u c t u r e t h a n subsoils. The soil processes which are responsible for soil structure formation and stabilization, are strongly dependent on fluctuations and range of soil temperature and soil moisture.
I m p l i c a t i o n s a n d recommendations From the results of the first year of study it can be concluded, t h a t a t e m p e r a t u r e change of 3~ in the Rhine basin probably will have no profound effect on soil structural stability and soil erodibility. This does no exclude an impact of climate change on soil erosion and sediment production in the Rhine basin. Sediment production on agricultural land will increase in s u m m e r due to higher rainfall intensities which strongly control the rate of soil erosion (Kwaad, 1991). For instance, a 40% increase of m e a n hourly rainfall intensity will occur, when mean day t e m p e r a t u r e rises from 20 to 23~ (Klein T a n k and KSnnen, 1993). Runoff and soil loss from agricultural land will increase in w i n t e r due to an increased probability of occurrence of s a t u r a t i o n overland flow (Kwaad, 1991), caused by a 19-20% increase of w i n t e r rainfall (KSnnen, this volume). Land use changes, such as foreseen for the decade 2040-2050 by P a r m e t (this volume), viz. a 20-24% decrease in area of agricultural land, will lead to a decrease of sediment production in the Rhine basin, if this 20-24% surface area is forested
883 and if the reduction of the area of crop land is not counteracted by an increased rate of soil loss per ha of remaining crop land. Increased winter discharge volumes and increased daily m a x i m u m discharges of the Rhine, such as mentioned by P a r m e t et al. (this volume), m a y lead to accelerated remobilization of sediment stored as alluvium along the Rhine and its tributaries, e.g. the so called 'Auelehm'. On the other hand, decreased s u m m e r discharges and decreased daily m i n i m u m discharges m a y lead to increased s e d i m e n t a t i o n in the channels of the Rhine system, because these decreased discharges coincide with increased sediment production on the slopes of the Rhine basin due to more intense and more frequent local summer thunderstorms under a warmer climate. A first s e m i - q u a n t i t a t i v e approximation of the impact of climate change on sediment production in the Rhine basin has been made by Van der Drift et al. (1994). From this it appears that sediment production is very sensitive to climate warming. It is therefore recommended to investigate more quantitatively the effect of climate change on the processes, factors and rate of soil erosion in the sediment contributing landscape units of the Rhine basin. 8.
T R A N S P O R T AND S E D I M E N T A T I O N
N.E.M. Asselman, H. Middelkoop and H.J.A. Berendsen Department of Physical Geography, University of Utrecht Heidelberglaan 2, 3508 TC Utrecht, The Netherlands Abstract Erosion, transport and deposition of fine suspended sediments are both directly and indirectly influenced by changed climate conditions. Changes in sediment transport rates were studied using sediment rating curves in combination with flow duration curves, developed using the BaU-climate scenario and four sediment t r a n s p o r t scenarios. All sediment transport scenarios show t h a t an increasing part of the yearly sediment load will be transported at discharges over 4000 m3/s: about 20% under present climate conditions increasing to about 40% when climate changes in accordance with the BaU-scenario. Three aspects of floodplain sedimentation have been studied: (1) past and present sedimentation rates, (2) the impact of climate change on future sedimentation rates and (3) heavy metal pollution of sediment. Floodplain sedimentation shows a high variability in time and space. Depending on site characteristics, present sedimentation rates range between 0.5 and 15 m m per year. At the beginning of floodplain formation, sedimentation rates probably were 3 to 4 times as high as at present. A climate change according to the BaU scenario will lead to a considerable increase in floodplain sedimentation rates in The Netherlands. Depending on the floodplain morphology, however, local changes in sedimentation rates will vary strongly; the expected increase will therefore range between 1% and >100%. The quality of the sediment is still a m a t t e r of concern. Although the heavy metal contamination has considerably decreased since 1970,
884 accelerated future sedimentation will accumulate considerable a m o u n t s of pollutants on the floodplains in The Netherlands. 8.1 I n t r o d u c t i o n The expected climatic change will affect erosion, transport and deposition of suspended sediments of the river Rhine. Concerns are not only related to the impact of environmental and climatic changes on transport and sedimentation of suspended sediments, but also to transport and deposition of sediment associated pollutants. Within the scope of the National Research Program (NRP 1) the impact of climate change on discharge, production, transport and sedimentation of suspended sediment by the river Rhine have been studied. The BaU-best scenario as given by Kwadijk (1993) was used to represent future climate conditions. The impact of climate change on the hydrology of the river Rhine was studied by Kwadijk (1993) and Parmet et al. (Section 6). The study on the effects of climate change on the suspended sediment budget of the river Rhine can be subdivided in three stages. (1) Erosion or production of sediment, that can subsequently be transported into the river. Van der Drift identified the major source areas of the suspended sediment transported by the river Rhine. He also studied the effect of climate and land use change on soil erodibility. (2) A sediment transport stage, d u r i n g which the s e d i m e n t particles are t r a n s p o r t e d d o w n s t r e a m . (3) Sedimentation in the lower course of the river and the delta area and, during high discharge periods, on the embanked floodplains along the river.
A rough estimate of changes in sediment production by soil erosion was obtained by using the USLE rain erosivity factor in combination with the BaU climate scenario (Section 2) and land use scenario (Section 3). This study was carried out by Van der Drift, Middelkoop and Asselman (1994). Transport and deposition of suspended sediment were investigated in two separate NRP studies which are reported here. In the first project, carried out by Asselman, the relation between suspended sediment transport rates and discharge is investigated. The objectives of this study are: to investigate the processes of sediment transport through the river Rhine, to assess the effect of climate change on suspended sediment transport rates, depending on changes in discharge and sediment supply to the rivers. In de second project, carried out by Middelkoop, the sedimentation on the embanked floodplains is investigated. Past and present sedimentation rates are reconstructed using various methods, and the possible effects of climate change on future floodplain sedimentation are assessed. The objectives of this study can be summarized as follows: Assessment of the rate of sedimentation on the embanked floodplains in The Netherlands in relation to flood-frequencies during the past decennia, and centuries. Assessment of quantitative relationships between (1) floodplain morphology, (2) the characteristics of flood periods and (3) sedimentation rates. Ev alu atio n of possible effects of climate change on future floodplain sedimentation.
885 8.2 M e t h o d s
Transport of suspended sediment The amount of fine suspended sediment (wash load) transported by the river Rhine depends on the availability of loose material and to a lesser extent on the capability of the river to transport this material. Unlike bed material load, wash load is a non-capacity load, which implies that sediment transport rates cannot be calculated using stream power related transport formulas. Instead, the so-called rating curve technique was used to study the effect of changes in discharge on the amount of suspended sediment transported through the river. A sediment rating curve describes the average relation between discharge and suspended sediment concentration. This relation is often described by a power function. In this study a power function with additive constant term was used: C=P+a*Qb where c is suspended sediment concentration (mg/1), Q is river discharge (m3/s) and a, b and p are regression coefficients. The sediment rating curves can be used to obtain information on the availability of sediment in a certain area in combination with the erosive power of the river itself. Steep rating curves (low a- and high b-values) are characteristic for river sections with little sediment transport taking place at low discharge. An increase in discharge results in a large increment of suspended sediment concentrations, indicating that either the power of the river to erode material during high discharge periods is great, or that important sediment sources become available when the water level rises. Flat rating curves are characteristic for river sections with intensively weathered materials or loose sedimentary deposits, which can be transported at relatively low discharges. The constant p-coefficient can be seen as a background concentration, a minimum concentration of suspended sediment occurring at very low discharges. In this study, sediment rating curves were developed for various locations along the river Rhine, using the daily measurements of water discharge and suspended sediment concentrations, measured by the Bundesanstalt fiir Gewasserkunde (BfG), Germany. The sediment rating curves were combined with flow duration curves to obtain sediment discharge curves, showing the effectiveness of different discharge intervals in transporting suspended sediment. Changes in the sediment discharge regime were studied using the changes in monthly discharges of the river Rhine given by Kwadijk (1993) and Parmet et al (Section 6). A relationship between monthly and daily water discharges was obtained following the method used by Kwadijk and Middelkoop (1994). The newly obtained flow duration curve was combined with different sediment rating curves to obtain the sediment discharge curves. Different sediment rating curves were used, corresponding to assumed changes in sediment production in the Rhine basin. A rough estimate of long term average changes in suspended sediment production by soil erosion in the Rhine basin under BaU climate and land use conditions was made using the Universal Soil Loss Equation (USLE), developed by Wischmeier and Smith (1978). It was assumed that only arable land substantially contributes to the production of fine sediment by soil erosion. The effect of climate change is calculated from changes in the rain erosivity factor (R) in the USLE. The
886 estimations of future soil erosion under changed rain erosivity are therefore assessed by changes in both the total amount and intensity of rainfall under the BaU climate scenario. F u t u r e rainfall intensities are calculated from changes in t e m p e r a t u r e s according to Klein T a n k & KSnnen (1993). The effect of changes in land use (Veeneklaas et al., 1994; Section 6) on soil erosion is calculated from the expected changes in the total area of arable land. The results are described in Van der Drift et al. (1994); the changes in annual suspended sediment load t h a t are used for the sediment rating curves are shown in table 8.1. For Rees, n e a r the D u t c h - G e r m a n border (figure 8.1), the following s e d i m e n t transport scenarios were used: 1) Sediment loads are determined by hydraulic properties of the river, sediment production in u p s t r e a m areas has no direct effect, la) The p r e s e n t r a t i n g curve remains valid under changed climate conditions; total yearly sediment load will change, lb) Background concentrations will change u n d e r changed climate conditions; total yearly sediment load remains constant. 2) Sediment loads are determined by the erosion rates in u p s t r e a m parts of the river basin. Changes in erosion rates are influenced only by changes in precipitation and temperature, no land use scenario is used. 3) Sediment loads are determined by the erosion rates in u p s t r e a m parts of the river basin. Differences in precipitation, temperature, and land use are t a k e n into account. 4) Sediment loads are determined by the erosion rates in u p s t r e a m parts of the river basin. Changes in erosion rates are the result of changes in land use. No climate change is t a k e n into account. This scenario is used as a reference scenario to evaluate the effect of climate change under changed land use conditions. In this case land use changes are assumed to be independent from climate change.
"~
THE
Figure 8.1 v Location of the gauging stations
'' /
~
I~O krn
887
Analysis o f s e d i m e n t from dike-breach p o n d s For the reconstruction of temporal variations in sedimentation rate during the past 200 300 years, sediment accumulated in dike-breach ponds was analyzed. These sediments can have a thickness of more t h a n 5 m. They often show a lamination of light and dark coloured humic clay. These laminations are believed to represent (yearly) floods. In order to correlate these laminae to floods and minor climate changes in the past, they were dated using the Pb-210 method. Analysis of the h e a v y m e t a l contamination of samples from the dated sediment profiles allowed to make a reconstruction of the pollution history of the river Rhine. -
Floodplain s e d i m e n t a t i o n rates on various time scales For a better interpretation of expected future changes in floodplain sedimentation rates, the spatial and temporal variability of the present and past sedimentation rates were investigated first (figure 8.2).
zwo,,e
.
E
=
~
.E_
9
X
X ,,
X
e
X
X
o
X
X
. . ~'
O
m
X X
~~" ~ Gorkum
. , ~ , ~ ~ , , ~ ~ ~ ~"
O ~ .~ " .-.
Lobith .."....
,
X
X = method applied at floodplain section
floodplains
Figure 8.2 Study area and location of investigated floodplains A. Present sedimentation rates. Present floodplain sedimentation rates and their spatial variability were m e a s u r e d after the two floods of 1993 and one in 1994 using about 800 sediment traps made of artificial grass. In the laboratory, the (dry) sediment from the traps was weighed, and the grain size distribution and organic m a t t e r content were determined. The results from the individual traps were interpolated to create r a s t e r maps of sediment accumulation, sand content and organic m a t t e r content. The patterns shown on the maps were correlated with floodplain morphology and sedimentation mechanisms (Asselman & Middelkoop, 1993). B. Sedimentation rates during the past decennia. In several floodplain sections with different elevations and distances to the main channel, soil samples from vertical profiles were collected and their heavy metal content was measured. The heavy metal content of floodplain soils was related to floodplain morphology and flood frequency. The sedimentation rates during the past 50 - 100 years were
888 reconstructed by comparing the heavy metal profiles in the floodplain soils with the pollution history of the Rhine (determined from the dike-breach ponds). In addition, the sedimentation rates of several profiles were assessed using the Pb210 method.
C. Sedimentation rates since the formation of the embanked floodplains. Old river maps provide a rough indication of the beginning of sedimentation on the enclosed floodplains. The total amount of accumulated sediment can be assessed by means of corings. Flood durations can be calculated from records of historic water levels. Using this information a simple model was made to estimate the average yearly sedimentation on several floodplains (Middelkoop & van der Perk, 1991).
A s s e s s m e n t o f the i m p a c t o f c l i m a t e change on f l o o d p l a i n s e d i m e n t a t i o n rates A. Assessment of the impact on floodplain inundation times. U s i n g the R H I N E F L O W model, the changes in monthly Rhine discharges were assessed for the BaU scenario (Kwadijk, 1993; Parmet et al., Section 6). From the relationship between monthly discharges and daily discharges, changes in peak discharge probabilities and exceedance times were calculated (Kwadijk & Middelkoop, 1994). These were used to assess future floodplain inundation times. B. Assessment of local sedimentation rates using a sedimentation model. The effect of climate change on the s e d i m e n t a t i o n rate on one floodplain was i n v e s t i g a t e d u s i n g the ( 2 - d i m e n s i o n a l ) W A Q U A - D E L W A Q m o d e l of R i j k s w a t e r s t a a t (Dutch M i n i s t r y of T r a n s p o r t , Public Works and W a t e r Management). Calibration of the model was carried out by simulating the flood of J a n u a r y 1993 of which the sedimentation rates were measured using the sediment traps. Also, the average sedimentation rates over the past 50 years reconstructed from heavy metal profiles of the floodplain soil were used to calibrate the model. The sedimentation rates under changed climate conditions were assessed by using the sediment discharge curves in correspondence with the BaU climate scenario. C. Sensitivity of large scale potential sedimentation rates.Complementary to the detailed model study for a small area using WAQUA-DELWAQ it was tried to estimate the sensitivity of floodplain sedimentation rates for the Rhine (Waalbranch) embanked floodplains as a whole. At this scale it is not possible to take physical flow and sedimentation processes into account. Instead, two estimators for potential floodplain sedimentation were introduced. The term potential s e d i m e n t a t i o n is used because the estimators do not calculate real sediment deposition, but they are a measure of the a m o u n t of sediment available for deposition. The first estimator calculates for each discharge interval the product of the corresponding (1) total floodplain area over which sediment flows, (2) the suspended sediment concentration and (3) the relative frequency of occurrence. The summed totals for all discharges gives the average yearly figure, expressed in km2*kg/m3/yr. The second estimator uses total volumes of water over floodplain areas and calculates the total average yearly load of suspended sediment present over floodplains, expressed in tons/yr.
889 The effect of the BaU discharge scenario and four sediment rating scenarios on the estimators has been calculated to investigate the possible impact on floodplain sedimentation. 8.3 R e s u l t s
Transport of suspended sediment The relationship between discharge and suspended sediment concentration shows considerable scatter. Some of this scatter can be the result of inaccuracies in the field or in the laboratory, seasonal effects, antecedent conditions in the river basin and differences between falling and rising stages. To reduce the scatter separate rating curves were developed after subdividing the data according to season, stage and wet or dry years. However, since this hardly improved the rating relationship, rating curves were developed using all data. The sediment rating curves developed for various gauging stations are shown in figure 8.3. It can be seen that the steepness of the rating curve decreases in downstream direction, indicating that near the Dutch-German border large quantities of fine material are available for transport at relatively low discharge. The importance of high discharge on suspended sediment transport decreases in downstream direction.
250
-
/ / ."
/" /
2O0
,"
,"
/,."
/,'" E
l.l : "
/
///
_
100
Rhelnfelden (kin 145)
,27" ,~"
150
-
/
..,.:'"
.)" .
/"
.
, . ' / ' / "/
I -o
/
i
///
. . . . .
Maxau (krn 362)
.........
BacharacWKaub (kin 545)
. . . . . . .
Welssenthurrn/Andernach
. . . . . . .
Rees/Emmedch (kin 837)
(kin 607)
.'/"
/ ." / 5O
I
0 o
I
~ .
I
o .
o
o .
.,..-
I o
o
o
,,.-
~
.
I ~ ~
I o o
o
o~
I o
I
I
II
o
oo
oo
o
~
o
o
o
o3
,,r
~
dlscharge
o
I
I
I
o
oo
oo
o
~
o
o
o
u3
co
~
I o
I
I
o
oo
~
o
I
~
2
o
o
r-.
co
~
I o
o
o
o
o~
o
~
o P
(m3/s)
Figure 8.3 Sediment rating curves developed for various gauging stations along the River Rhine
890 Sediment discharge curves for the river Rhine near Rees were developed using the flow duration curves in combination with the sediment rating curves developed for present and future climate conditions. The sediment discharge curves are shown in figure 8.4. The results of the different sediment transport scenarios are also given in table 8.1. Table 8.1 Sediment transport rates using four sediment transport scenarios scenario
P
Present la lb 2 3 4
29 29 24 32 15 18
p total load >4000 >6000
= = = =
Total load % and (Mt/yr) 100 115 100 122 78 71
>4000 % and (Mt/yr)
(3.09) (3.55) (3.09) (3.77) (2.41) (2.20)
20 35 37 35 42 22
>6000 % and (Mt/yr)
(0.62) (1.24) (1.14) (1.32) (1.01) (0.48)
5 (0.15) 17 (0.60) lS (0.56) 16 (0.60) 20 (0.48) 6 (0.13)
background concentration (mg/l) % of present total yearly sediment transport percentage of total sediment load transported at Q>4000 m3/s percentage of total sediment load transported at Q>6000 m3/s
Steepness of the rating curve is kept constant with a = 1.96"10-5 and b = 1.93 500000000
T
450000000 ~L
/ / ~
"c" 400000000 T
/,~" ~ ' \ . X
~ ~ooooooo~ om ~oooooooo~
/'<" ~'-'A .'\ li"/ .--"'--.. "%.,
2sooooooo t
,I,'/
/
,//,' /
#.'/,-"
~oooooooo Jr -o
, ,
\
\
-\\'\
-,~<\
.... - \
,,,,,/," 1.'
c
. ooooooo
oooooooo +
\:,.~ -.
"
/!//
.
" \\
\
h',~ ,~
5ooooooo +A~S//
\ : " ~-'~-~.~. ~
~
~ -':- L:"":~
- :2..: .'~. ~____:.~_..,._,.~:~
o in
Discharge
..................... present
. . . . .
scenario la
.......
~
scenario 4
scenario 3
(m3/s)
.........
scenario lb
.......
scenario 2
Figure 8.4 S e d i m e n t discharge curves developed for Rees according to t h r e e s e d i m e n t transport scenario's
891 Scenario 1. According to the first sediment transport scenario, suspended sediment concentrations only depend on hydraulic properties of the river and t h u s on discharge. The effect of climate change can therefore be s i m u l a t e d using the sediment rating curve based on measurements of the last 15 years in combination with the flow duration curve according to the BaU scenario. The resulting sediment discharge curve shows t h a t more suspended sediment will be transported both at very low discharge (Q < 1500 m3/s) and at high discharge when inundation of low lying floodplains occurs (Q > 4000 m3/s). If the present sediment rating curve is used (scenario la), total suspended sediment load will increase by 15%. If the total load r e m a i n s the same (scenario lb), background concentrations will decrease from 29 to about 24 mg/1. Scenario 2. When changes in soil loss due to climate change are investigated using the BaU scenario for changes in rainfall amounts and rainfall intensities while land use is assumed to r e m a i n unchanged, sediment production will increase by 22% (van der Drift et al., 1994). The increased sediment load can result in h i g h e r background concentrations, occurring at low discharges. If the steepness of the sediment rating curve remains the same, background concentrations will increase from 29 to about 32 mg/1. Scenario 3. W h e n changes in soil loss u n d e r the BaU climate scenario are combined with a land use scenario with a 23 to 48% decrease in area of arable land, sediment production will decrease by 22%. This means t h a t even if the steepness of the rating curve changes, the background concentration will decrease. This can be explained by the trapping of sediment behind weirs. During s u m m e r discharge will decrease, which might result in increased sedimentation behind weirs during s u m m e r time and lower base flow concentrations near Rees. During the winter, more high discharge periods will occur and much of the sediments deposited behind weirs will be flushed out, leading to higher concentrations in the river during high discharge events, as shown by the increased value of the b-coefficient. If the steepness of the rating curve changes, background concentrations will decrease to 18 mgfl or less. If the steepness of the rating curve remains the same, expressed by the unaltered values of a and b, the background concentration (p) will decrease to 15 mg/1. Scenario 4. In scenario 4 the effect of changes in land use without a climatic change is studied. As a result of a 25 to 42% decrease in arable land, the total suspended sediment load will decrease by about 29%. If the steepness of the rating curve r e m a i n s the same, background concentrations will be about 18 mgfl. A similar background concentration was found for scenario 3. The most i m p o r t a n t difference occurs in the sediment discharge curve. According to scenario 4 only 22% of the total sediment load (0.48 Mt/yr) is transported at discharges over 4000 m3/s. When changes in land use occur in combination with changes in climate (scenario 3) about 42% of the total yearly load (1.02 Mt/yr) is t r a n s p o r t e d at discharges of 4000 m 3]s or more. This difference is mainly the result of changes in the flow duration curve. High discharges will occur more often under BaU climate conditions t h a n under present climate conditions.
From the scenarios is can be concluded t h a t the expected change in land use has a stronger impact on the yearly production of suspended sediment t h a n the BaU
892 climate scenario. However, due to the changes in the flow duration curve, a climate change will have a much stronger effect on possible floodplain sedimentation. The effect of climate change under changed land use scenario will be about the same as under present land use conditions.
A n a l y s i s o f sediment f r o m d i k e - b r e a c h p o n d s Although the sediment shows a clear lamination, it is difficult to correlate this to individual floods and hence to fluctuations in flood frequencies. This is because local sedimentation rates are the results of a combination of flood duration, inundation depth, sediment concentration and local flow patterns. It is virtually impossible to reconstruct these in detail for historic times. Using historic dates, Pb-210 dates, palynological information and variations in sediment compaction a time/depth curve of the sediment could be made. Using this as time control the heavy metal pollution of the Rhine sediment during the last 200 years was reconstructed (figure 8.5). This diagram shows t h a t the heavy metal pollution strongly increased during the first half of this century; maximum pollution occurred between 1960 and 1970, since 1970 the heavy metal pollution has strongly decreased. Cd,
~ 50
0
Cu,
Pb (mg/kg)
100
200
. . . .
!
~!~~1,
....
F; .... i ..... i--,::]-"-,~,-:~
100
150
:
',
k..- ,'~
,,
,
!:',.-r, ~ i
200
,, j
~. . . . . 4. . . . . 4-%,.-; . . . . . ',
. . . . ; . . . . . ~ : , . - - ~. . . . . . r - - - r ,, ~ , ;-~ : . : E'q l I * ',m l :
~~
..... i ..... i ..... ! ..... ! ..... 1 1980
---~ . . . . . ~. . . . . 4. . . . . . 14
13
300
', ....
',
.;-"
l l
.--; ..... ,, : l l :
i
i
,
,
',
1960 I
~. . . . . ~. . . . . . ,, ,, : : : :
i
i
i
h~ 1940
102o
..... !......i ..... i..... I-.+-::ti ..... t 400
450
Ii_IIilii i i . . . .
i . . . . . i . . . . . . ',:-. . . . . i . . . . .
............................
--"--
z.
i ....................
i iI
""" .....
500 0
Figure
8.5
500
1000
1500
2000
2500
zn (mg/kg)
Heavy metal pollution history of the river Rhine reconstructed from a dike-breach pond near Wamel
F l o o d p l a i n sedimentation rates on various time scales A. Present sedimentation rates. The sediment accumulation maps show clear patterns in spatial variability (figure 8.6). During a flood, large amounts of sandy s e d i m e n t with low organic m a t t e r content are deposited on the levees. The sediment accumulation decreases exponentially with distance from the river. This
893
,,,iiiiiiiiiiiiiiiiiiiiiiiiiii ,,
::!9
if::
iiii!iiiiiiiii!!i!
iiiiiiiiiiiiiiiiiii ii!iiiiiii!ii~ii!!i
:!:!:~',!:!:~: :.:..,:.:.:,.,:.:.: ::::::::::::::::::: .........
> 0.8 kg/m2 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.00
,.::::~
Study
m~t
'Keenf
a. Keent.
iiiiiiiiiiii!iiiiiiiiiiiiiiiiiiimiiiiiiiiii :: ii!i!i?!!iii!iii!i;iii)iiiiiiiiiiiiiiiii?iiii.;ii:: :
2.26 kg/m2 1.99 1.73 1.46 1.20 0.93 0.67 0.40 0.14
. . . . . . .
Study
area
~ili~: i:i=:i;:ii
'V'triksche Plaat'
b. Variksche Plaat. floodplain
~
......
minor river dike
former floodplain protected by river dike
{...........'".. escarpment
major river dike
~
.
.--
woe
groyne
_
_ V
a
r
i
~
Figure 8.6 E s t i m a t e d s e d i m e n t a c c u m u l a t i o n after block kriging. Grid cell size = 10 x 10 m2
b''h
894 gradient is mainly caused by the amounts of deposited sand; no gradient was found in the amounts of silt and clay. The amounts of sand deposited j u s t behind the levee show a high spatial variability. If a floodplain has been completely s u b m e r g e d for several days, local depressions have almost no effect on the sediment deposition. The amounts of sediment deposited on different floodplains can vary considerably. Observed average values after the flood of december 1993 range between 1.0 and 6.6 kg/m2, equivalent to 0.8 respectively 5.4 mm. These differences between floodplains seem mainly related to the flow pattern of the water during inundation. During minor floods, when a floodplain is inundated only during a few days, the p a t t e r n of sediment accumulation is much more related to local differences in elevation and inundation times than during large floods. The amount of deposited sediment is far less than proportional to the magnitude of the flood. This implies t h a t minor floods contribute considerably to the yearly floodplain sedimentation rate. From the results of the individual floods, the present average yearly sedimentation rates will be estimated for different floodplain sections. This part of the study is still in progress.
B. Sedimentation rates during the past decennia. The heavy metal profiles obtained from the floodplain soils generally have the same shape as the pollution curve reconstructed from the dike-breach pond sediments. Profiles from floodplains with high sedimentation rates have a higher pollution in the upper 10 cm and have a higher total pollution. Also the maximum concentration is greater and is found at greater depth. The results show that the total soil pollution can be very different from the pollution in the upper 10 cm. A one-dimensional sedimentation model is being developed that simulates the floodplain sedimentation with contaminated sediments. Using this model the sedimentation rates will be reconstructed. The average sedimentation rates during the past decades obtained from the Pb-210 samples range between 1.6 and 0.1 cm/yr. C. Sedimentation rates since the floodplain formation. This method applies to lateral bars and can be used in a limited number of undisturbed floodplains. At the beginning of the floodplain formation the sedimentation rates varied between 1 and 3 cm/year, which is 3 to 4 times as great as at present. Two examples are shown in table 8.2.
895 A. Klompenwaard: 1800 - 1990
B. Variksche Plaat: 1860 - 1990
year
elev. (m a.s'l.)
sed. rate (mm/yr)
year
1800 1850 1900 1950 1990
10.6 11.2 11.7 12.0 12.3
16.0 11.0 7.6 6.2 4.8
1860 1900 1950 1990
N
i
elev. (m a.s.I.)
sed. rate (mm/yr)
4.5 5.4 6.1 6.5
30.0 18.0 11.0 8.5
Ame.m 2 KL'*I~ ~
~
~iTme~e'/ .....~::
.::~-".....
Lobith ,,........
[iiiiiiiiii
i] floodplains
KL = Klompenwaard VP = Variksche Plaat
Table 8.2 Reconstruction of average yearly sedimentation rates using historic data; two examples
A s s e s s m e n t o f the i m p a c t o f c l i m a t e change on f l o o d p l a i n s e d i m e n t a t i o n rates A. Assessment of impact on floodplain inundation times. The effects of climate changes on the Rhine peak discharge frequencies and exceedance times according to the AP and BaU scenarios are given in Section 6. The BaU scenario causes a significant increase in inundation times of floodplains. B. Assessment of local sedimentation rates using a sedimentation model. A proper calibration of the WAQUA/DELWAQ model requires more d a t a on actual sedimentation than those obtained of the flood of J a n u a r y 1993, and preferably for l a r g e r a r e a s t h a n the V a r i k s c h e Plaat. Therefore, also the s e d i m e n t measurements of the December 1993 flood will be used for calibration. Preliminary results of the model indicate that the sedimentation rates on the Variksche Plaat under the BaU climate scenario increase by only 1% if the sediment concentration of the Rhine only depends on hydraulic conditions. Sedimentation rates increase by almost 20% when increased sediment production is taken into account. The effect of climate change for this particular floodplain seems to be insignificant. The main reason for this is that in floodplains directly bordering the main channel only minor amounts of sediment are deposited during high discharges since the flow velocities are then too high for sedimentation. The increase in sedimentation rates will be
896 larger on floodplains t h a t are separated from the main channel by a s u m m e r dike and w h e r e relatively large a m o u n t s of sediment are deposited d u r i n g high discharges.
C. Sensitivity of large scale potential sedimentation rates. The BaU scenario has a strong impact on the potential floodplain sedimentation (figure 8.7 and 8.8). According to the estimator using sedimentation areas, potential sedimentation rates will increase by a factor 1.5 to 2, depending on the sediment rating scenario. The effect on the estimator using sediment volumes is even larger: potential sedimentation rates will increase by a factor 1.7 to 3. When scenario 4 (only land use change) is t a k e n as a reference, potential sedimentation m a y increase by a factor 2.5 respectively 4.5. As demonstrated by the WAQUA/DELWAQ model study of the Variksche Plaat, the increase in the amounts of sediment t h a t really will be deposited may be less large. This strongly depends on the morphological properties of the floodplain sections. For a proper estimate a further analysis of t h e s e d i m e n t a t i o n process u s i n g s e d i m e n t t r a p s and models such as WAQUA/DELWAQ is required in different types of floodplains. 25
:i~~... ~:-i ::-!--!-i-::,/], ~. ~-~;i~i
20
:
:
:
:
! .......... '~ ' ~' ~ - ~ ~ -~
=--~176176 " =~
:',:ILII~,:"~-~L!"::"',
. .:_ .:- - :. - -:- - :- - -:- - : - i ~ )
\-:
- ~ ~ , ~ \ -
.~
9
,. - -:- - : - -:- - :- -
=
=~
--!--:--;--:--:---:-';A;~~-i--~--:-:--<~!--:--:--i ! i . ---i _i._:_ ii_.',. .... _::__ .::__::__....
i_
5
i_
0
o
~
~
o
o
o
o
o
o
o ,~
o
~
o ~~ ,o
o ,~
o
o
o
discharge at Lobith
Rhine
o ,o
o
o
o
o
o ,o
o
o
o
o
o
(m31s)
Figure 8.7 Expected yearly area of sediment influx * sediment concentration as a function of discharge for different scenario's - river Waal 0.05
.
.
v~
o.o,
0o,
8.8
.
.
.
i
:: ~ ! :: :: :: ::
.
.
. '.
- , - - . - - : ~ - : - - : - - , . ...... ? ~ / : - ,'~<~ - : - -
oo o ~ Figure
.
I~'esen!
.
~::::
.
.
.
. . ' .
:_. i.: :..i__ : ' :~~i--:--~--:' "u
: .
.
.'
---
,
~:~~~
.
.
:::::= Rhine discharge
.
.
.
'
.
:!
1
'
- "'" : - - : 9-
.
.
~ ! ~ ~
: ~ : at Lobith
'
.
.
'
:
.
~~.
~ - : ~
(m3/s)
Expected yearly sediment load present above inundated floodplains as a function of discharge for different scenario's - river Waal
897
8.4 Implications According to all sediment transport scenarios a larger part of the yearly sediment load will be transported at discharges over 4000 m3/s, when inundation of low lying floodplains occurs. Under present climate conditions about 20% of the yearly sediment load is transported at discharges over 4000 m3/s, this will be about 40% when climate changes according to the BaU scenario. Under present climate conditions about 5% of the total sediment load is transported when all floodplains bordering the Dutch part of the river are inundated (Q>6000 m3/s); this will be over 15% when climate changes. If no measures against soil erosion are taken by reducing the area of arable land or by soil conservation programmes, the production of sediment will increase. Increased, sediment production will lead to about 20% higher concentrations of suspended sediment at low discharges. When the area of arable land decreases, background suspended sediment concentrations will decrease by some 40%. The amounts of suspended sediment transported at discharges when all floodplains are inundated is hardly affected by the expected changes in land use. The increase in sediment load, transported at high discharge is mainly the result of changes in the discharge regime. Floodplain s ed i me nt a t i on is a complex process. It depends on floodplain characteristics, discharge frequency distributions and sediment concentrations, and therefore shows a high variability in time and space. Depending on site characteristics, present sedimentation rates range between 0.5 and 15 mm per year. Present sedimentation rates are 3 to 4 times as low as they were in the past when the floodplain surface was 1 to 2 m lower and flooding occurred more frequently. This implies that removal of summer dikes and lowering of floodplain surfaces as proposed in nature rehabilitation plans will strongly increase sedimentation rates. The effects of climate change on floodplain sedimentation, as evaluated using the BaU scenario can be considerable. This is mainly caused by an increased frequency of high discharges. As the relationship between effective discharge and the area of inundated floodplains is non-linear, even minor changes in discharge and sediment concentrations will have a strong effect. As a result, the potential floodplain sedimentation is very sensitive for the BaU discharge and sediment rating scenario conditions. The effective sedimentation rates may be different, depending on the morphology and type of floodplain. On floodplain sections directly bordering the main channel and without a summer dike, changes in sedimentation rates are relatively unimportant. However, the effect on floodplains t h a t are situated behind a summer dike, is expected to be relatively high. As it was found that floodplain sedimentation is highly variable in space and during individual floods, predictions on future sedimentation rates cannot yet be done with high accuracy. Further analysis on the effective sedimentation on different types of floodplain sections is therefore required. A p a r t from the sediment quantities, sediment quality is a matter of concern. A reconstruction of the heavy metal pollution of the river Rhine during the past century shows that maximum pollution occurred between 1960 and 1970. The present distribution of heavy metals in floodplain soils is related to this pollution history and the floodplain sedimentation in the past. At many locations the soil
898 quality of the upper 10 cm gives an underestimation of the total pollution of the whole profile. Although the heavy metal content has considerably decreased since 1970, an accelerated sedimentation will still accumulate large amounts of heavy metals on the floodplains in The Netherlands. As average sedimentation rates on most floodplains are in the order of 1 mm/yr, the quality of the upper 50 cm of the floodplain soils will only little improve by the accumulation of sediment of improving quality during the forthcoming century. 9. ECONOMY, SAFETY, ENVIRONMENT F.R. Rijsberman Resource Analysis Zuiderstraat 110, 2624 SJ Delft, The Netherlands 9.1 I n t r o d u c t i o n
In the following pages the results of the hydrology-related projects in the NRP, and their implications for economy, safety and environment, are compared with the results and conclusions of earlier studies. The objective of this comparison is to evaluate whether the NRP-sponsored research has led to changed insights into the impacts of hydrology-related impacts of climate change on the economy, safety and environment in The Netherlands. During the period when IPCC produced its First Assessment Report a comprehensive study was undertaken in The Netherlands to assess the Impacts of Sea Level Rise on Society (ISOS). The ISOS study (Peerbolte et al., 1991) focused largely on sea level rise, but attempted to assess the impacts of changes in hydrology on various aspects of the economy, safety against flooding and the environment as well. The study attempted to integrate data, knowledge and modelling results available in 1988-1991 to produce an assessment of the impacts of climate change for The Netherlands that can be considered as a state-of-the-art review at that point in time; as such it is used in the following pages as a baseline against which the progress made through the NRP programme can be evaluated. It has to be noted at the outset that the approach adopted in the ISOS and NRP studies is quite fundamentally different. The former study made relatively quick and dirty scenario-assumptions for local changes in climate and for a series of physical (intermediate) effects of such changes, e.g. on discharge of the River Rhine, and subsequently attempted to estimate socio-economic impacts resulting from these changes. Impacts were expressed, where possible, in monetary terms for damages such as production losses in agriculture, flooding damages for industry, additional costs of provision of cooling for power plants, and increased costs of water management measures such as pumping drainage water. The hydrology related NRP studies focus largely on the development of regional climate change scenarios (Section 2) and the direct physical effects of these changes in terms of soil erosion, river discharge, sediment transport, sediment deposition, saline seepage and the hydrology of forests. The hydrology-related NRP studies do not assess the socio-economic impacts of these physical effects directly,
899 w i t h the exception of the analysis of changes in agricultural production caused by changes in t e m p e r a t u r e , evaporation, precipitation and CO2-concentrations. 9.2 C l i m a t e s c e n a r i o s One of the more i m p o r t a n t steps forward in the last few y e a r s is the improved regionalization of the climate change scenarios. Both earlier studies as well as the N R P studies are based on the IPCC Business as U s u a l (BaU) scenario, b u t the regionalization of the hydrology-related climate variables is quite different. For c o m p a r i s o n t h e scenarios u s e d in the I S O S - s t u d y are p r e s e n t e d h e r e a f t e r , t o g e t h e r w i t h the scenarios on which the N R P studies are based. One ISOSscenario is the so-called Average (AV) scenario in which all variables, including hydrology-related variables, have their expected values. Another scenario, in which the e s t i m a t e d s t a n d a r d deviations were deducted from the expected values for all variables, is referred to as the Unfavourable (UNF.ALL) scenario. The hydrology related NRP-studies refer to either the regional climate scenarios for The N e t h e r l a n d s developed in NRP (Section 2) or the scenarios for the Rhine Basin developed by Kwadijk (1993). These are also expected values and should therefore be compared with the AV scenario in ISOS. Both scenarios are presented in Table 9.1, together with the two earlier scenarios. Table 9.1 Comparison of regionalized climate change scenarios used in the hydrology related NRP studies as well as in earlier studies Variables Ta Tw Ts Pa Pw Ps Qw Qs Ea=Ew=Es SLR
ISOS AV
ISOS UNF.ALL
NRP
Kwadijk
+3 +3 +3 + 10% + 10% + 10% + 5% -5% + 10% 0.6
+3 +3 +3 +15% + 5% + 10% - 10% + 15% 0.85
+3 * 2.3-3.7 * 2.3-3.7 + 13% * 3%-21% * 3%-21% ** + 15% * * - 10%
+3.5 + 4.3 + 2.9 + 11% + 19% +4%
* m o n t h l y values ** computed values from N R P studies ( P a r m e t et al., 1994) Where: Ta = average a n n u a l T e m p e r a t u r e increase in degrees C Ts = s u m m e r T increase in degrees C Tw = winter T increase in degrees C Pa = average precipitation change in % Ps = s u m m e r precipitation change in % Pw = winter precipitation change in % Qs = s u m m e r river discharge change in % Qw = winter river discharge change SLR = sea level rise in m Ea = average a n n u a l evaporation change in %
900 As can be seen from Table 9.1, the changes in precipitation assumed in the recent scenarios (Section 2) are considerably greater than in the earlier (ISOS-AV) scenario; even greater than in the scenario that was considered "extreme" a few years ago (UNF.ALL). Similarly, the expected changes in river Rhine discharge resulting from the recent NRP RHINEFLOW study considerably exceed the average changes expected earlier (AV scenario). In addition, the NRP studies (Section 6) conclude that the frequency of low flow months (Q< 1000 m3/s) increases with 60%. The conclusion can be drawn from the scenarios presented above that the regional climate change studies undertaken in recent years have shown that the hydrological changes associated with the IPCC-BAU scenario are considerably greater than expected before. Consequently, the hydrology-related impacts of climate change on the economy, safety and the environment can also be expected to be relatively more important than estimated before.
9.3
Impacts on safety against flooding
In the ISOS study it was assumed that the design discharge used for flood protection infrastructure (a discharge of 16,500 m3/s at Lobith, with a recurrence interval of 1250 years) would increase 10%. This resulted in increased design dike crest levels for the river dikes that varied between 0.1 and 0.5 m for different dike sections to maintain the same safety against flooding. The cost of raising river dikes in the non-tidal part of the country (necessitated both by sea level rise and increased river discharges) was estimated to be in the order of 2,500 million guilders over a 100 year period. The cost of raising sea and river dikes was the largest direct economic impact of sea level rise determined in the ISOS-study. The NRP RHINEFLOW study concludes that the average winter discharge increases 15% (Qw= +15%), but u since only a 25 years period was simulated does not allow conclusions to be drawn concerning changes in discharges at frequencies relevant for flood analyses, e.g. discharges with a recurrence interval of 1250 years (Section 6). In a related study, Kwadijk and Middelkoop (1994) used the RHINEFLOW model to assess the probability of exceedance for discharge peaks under possible future climate conditions. For peak discharges relevant for flooding of floodplains (up to Q= 6.500 m3/s) they found that a precipitation increase of 20% leads to a 30% higher two-year peak discharge. For larger design discharges, used for the design of flood protection structures, reliable results could not be obtained because the length of the discharge record was too short. This implies that the hydrology-related NRP studies have not, so far, improved the assessment of floodsafety related impacts. Further analyses of the changes in the peak discharges of the River Rhine are therefore recommended.
9.4
Economic impacts
I m p a c t s on a g r i c u l t u r a l p r o d u c t i o n a n d l a n d use The NRP study (Section 3) analyzed changes in crop yield potential for a time horizon of 2040-2050 with a biophysical simulation model (WOFOST) for Ps=0% en PW=+10%, Ts=l.5, Tw=2, double CO2 concentrations, and computed (Penman) evaporation rates (Ea= +15%). It concludes that the changes would result in average increases in yields of wheat, grass, sugarbeets and corn on sandy soils in the Rhine river basin of 40, 33, 25 and 12 %, respectively. Based on analysis of autonomous developments such as population growth, agricultural production
901 levels and developments in the world and EU market, land use patterns have been simulated, with and without climate change. Without climate change, the range of land use scenarios for the Rhine Basin in the decade 2040-2050 shows that the land used for agriculture will decrease by 0 to 3 million hectares. With climate change, land used for agriculture decreases an additional 0.2 million hectares. Both earlier studies and analysis carried out in the NRP (Section 5) conclude that possible increases in saline seepage through the groundwater due to sea level rise are negligible. Parmet et al. (Section 6) conclude that the summer discharge of the River Rhine would decrease by about 10% and the number of low flow months (Q< 1000 m3/s) increases with about 60%. It has not been analyzed in the NRP what the surface water related implications of these drier summer conditions are, but these could be considerable. The analyses of earlier studies concerning climate change impacts on agricultural production focused on surface water management related issues, including agriculture drought damage, agriculture salinity damage, sprinkler irrigation cost; (discharge) pumping capacity and cost. Conclusions were, for instance, that hydrological dryness influences drought damage substantially, e.g., Ps=-7% (and assumed Es=+15%) resulted in 38% increase of drought damage cost from about 350 to about 500 million guilders per year (Peerbolte et al., 1991). It was also concluded that reductions in average summer discharge of the River Rhine, e.g., Qs=-7%, had only minor impacts on agricultural production. Summarizing, the NRP studies have concluded that there is a potential for significant increases in agricultural production as well as a potential of significant increases in drought damage. From a perspective of the socio-economic impacts on agriculture, the forecasted increase in productivity could increase the comparative advantage of agriculture in The Netherlands, but this will depend on a host of other factors as well (e.g., surface water related drought damages, and w a t e r management costs). The surface water related impacts of the considerably lower summer discharges of the Rhine have not been investigated in the NRP studies, but based on earlier studies it is expected that these could be quite significant for The Netherlands.
Increased costs o f electricity p r o d u c t i o n The ISOS study concluded that the additional costs of electricity production of a decrease in summer discharge of the River Rhine (Qs=-7%) would be in the order of 6 million guilders per year. In Section 6 it is concluded that the average reduction in summer discharges could be higher (Qs=-10%) and that the frequency of low flow months, which can be critical for design of the cooling system, increases considerably. The additional costs of electricity production are therefore likely to be significantly higher than estimated in the earlier studies. Water m a n a g e m e n t costs The larger decrease in average summer Rhine discharges, and increased frequency of low flow months, can also be expected to impact several surface water management factors, particularly the effort required to control the salinity intrusion through the Nieuwe Waterweg (harbour of Rotterdam), navigation, and flushing of polder areas. Increased costs for shipping have not been analyzed in the
902 climate change impact studies to date. Through other studies it is known, however, that the costs for navigation increase rapidly when discharges decrease. Shipping costs in The Netherlands increase from 85 million guilders per week to 130 million guilders per week as the discharge decreases from 1600 to 1000 m3/s (Anonymous, 1990).
F l o o d i n g in f l o o d p l a i n s Earlier studies by Licht (1990) estimated t hat for SLR= 0.85, Qs=-10%, en Qw=+10%, the expected annual economic losses for the brick industry are in the order of 2 million guilders per year due to increased flooding, whereas for dairy farming the increased flooding in winter is more or less compensated by decreased flooding in summer. The NRP study results (Section 6) of Qs=-10% and Qw=+15% would probably result in somewhat higher losses for both the brick industry and dairy farming, but these impacts are still quite small. I m p a c t s on the environment The NRP studies evaluated changes in the following variables that may, in turn, cause environmental impacts" sediment production (Section 7) and suspended sediment t r a n s p o r t and suspended sediment deposition on embanked river floodplains (Section 8); - hydrology of forests (Section 4). -
Earlier studies for The Netherlands evaluated a number of climate change impacts on the environment, but most of these related to coastal environmental aspects. Where, for instance, the ISOS study did look at environmental impacts (impact of flooding on natural areas in river floodplains; decreased biomass phytoplankton and chlorophyll in fresh water ecosystems), the assumed changes in hydrology were too small to determine significant impacts on the environment. The NRP studies showed that sediment production in the River Rhine catchment area may increase by about 20% due to climate change (Section 8), but this increase will be more than balanced by autonomous land use changes that cause a decrease in sediment production. The NRP studies also show that a much higher part of the suspended sediment will be transported at higher discharges, which will lead to a considerably higher deposition of suspended sediment on the embanked floodplains (Section 8). If this sediment remains as polluted as it has been during the last decades than this would lead to a buildup of pollutants on the floodplains which will have serious consequences for the environment. Fortunately, there is a trend towards improving sediment quality in the River Rhine that may mitigate this potential impact. The studies on the hydrology of forests shows t h a t increasing CO2 concentrations may lead to lower evapotranspiration and hence reduced drought damages in summer. Generally speaking, the value of the conclusions of these studies would increase if they were integrated into a common framework for analysis. The changes in the hydrology-related climate variables that follow from the scenarios discussed above, as well as the computed changes in river discharge, are likely to have significant consequences for natural ecosystems, such as desiccation of floodplains in summer. Such ecosystem related consequences have not been looked at in the hydrology-related NRP studies, however.
903 9.5 C o n c l u s i o n s
The changes in hydrology-related climate variables t h a t follow from NRP regionalized climate scenarios are significantly greater those assumed in earlier studies. NRP studies have analyzed the changes in average summer and winter discharges of the River Rhine, as well as the changes in the 2-year peak discharges and frequency of low flow months. The results of these studies show that the discharges of the River Rhine change considerably more than assumed earlier. Peak discharges at recurrence intervals relevant for analysis of impacts on flood safety have not yet been analyzed. Further research in this area is recommended. NRP studies have shown that the expected changes in temperature, precipitation and CO2-concentration have the potential for increased agricultural productivity in The Netherlands. The economic consequences of such changes also depend on changes in drought damages and water management costs. The hydrology related NRP studies have forecasted relatively greater decreases in average summer discharges of the River Rhine than assumed earlier, and significantly increased frequencies of low flow months. This is likely to have considerable impacts on drought damages in agriculture, water management costs, costs of navigation, and costs of electricity production. These surface-water related drought effects may well be the most important economic impacts caused by changes in hydrological climate variables. Several NRP studies have assessed physical effects due to changes in hydrologic variables, such as sediment production, sediment transport and deposition, and hydrology of forests. The expected change in hydrology is likely to lead to increased sediment production, a larger deposition of sediment on the embanked floodplains, and a decrease in the evapotranspiration of forests. These physical effects are likely to have impacts on the environment, but these have not been estimated directly. It is recommended that the analysis of impacts of the various physical effects studied in the future hydrology-related NRP projects is conducted in an integrated framework. 10. R E F E R E N C E S
Anonymous. 1990. Beleidsanalyse waterhuishouding scheepvaart, bevindingen van de werkgroep scheepvaart (in Dutch). RIZA-report 90.005. Asselman, N.E.M. and H. Middelkoop, 1993. Floodplain sedimentation; quantities, patterns and processes. Geopro-93.09. Dept. of Physical Geography, Utrecht University. Accepted for publication in E a r t h Surface Processes and Landforms. Boer, G.J., N.A. McFarlane and M. Lazare, 1992. Greenhouse gas-induced climate change simulated with the CCC second-generation general circulation model. J. Climate, 5: 1045-1077. Boons-Prins, E.R., G.H.J. de Koning, C.A. van Diepen and F.W.T. Penning de Vries, 1993. Crop specific simulation parameters for yield forecasting across the European Community. Simulation reports CABO-TT, no. 32, CABO-DLO. Wageningen, The Netherlands.
904 Buishand, T.A. and A.M.G. Klein Tank, 1994. Regression model for generating time series of daily precipitation amounts for climate change impact studies (submitted to Stochastic Hydrology and Hydraulics). Bultot, F., Coppens, A., Dupriez, G.L., Gellens, D. and F. Meulenberghs, 1988. Repercussions of a CO2 doubling on the water cycle and on the water balance, A case study for Belgium. Journal of Hydrology 99: 219-347. Bultot, F., Gellens, D., Spreafico, M. and B. Sch~idler, 1992. Repercussions of a CO2 doubling on the water balance - a case study in Switzerland. Journal of Hydrology, 137: 199-208. Cure, J.D. and B. Acock, 1986. Crob responces to CO2 doubling: a literature survey. Agri. For. Meteor., 38" 127-145. DGV-TNO, 1984. Grondwaterkaart van Nederland, (Groundwater map of The Netherlands), Rotterdam 37 oost, 37 west). (In Dutch). Rapport GWK 35, Dienst Grondwater-verkenningen TNO, Delft-Oosterwolde. Dolman, A.J., Stewart, J.B. and Cooper, J.D., 1988. Predicting forest transpiration from climatological data. Agri. For. Meteor., 42: 339-353. Drift, J.M.W., H. Middelkoop & N.E.M. Asselman, 1994. Estimation of the effects of climate and land use change on the production of fine sediment by soil erosion in the catchment area of the river Rhine. Internal report, Laboratory of Physical Geography and Soil Sciences, University of Amsterdam; Department of Physical Geography, University of Utrecht, 12 pp. Eamus, D. and Jarvis, P.G., 1989. The direct effects of increase in global atmospheric CO2 concentration on natural and commercial temperate trees and forests. Adv. Ecol. Res., 19: 1-55. Feddes, R.A. and P. Kabat, (eds.) In prep. 1994. SWAP. A model to simulate the Soil Water Atmosphere Plant interactions. Part I. Theory and model description. Simulation Monograph, Pudoc, Wageningen. Hendriks, C.M.A., 1994. Biophysically-based, spatial analysis of possible climate change impacts on forest yield potentials and water use in the Rhine basin, volume 3. (In press). Hendriks, C.M.A., 1994. Biophysically-based analysis of possible climate change impacts on forest yield potentials and water use in the Rhine basin. Volume 3 of Land use projections for the Rhine basin based on biophysical and socioeconomic analysis. Winand Staring Centre-RIZA report 85.3. Hendriks, M.J., Kabat, P., Homma, F. and Postma, J. 1990. Research into the evaporation of a deciduous forest. Measurements and simulations. Report 90, Staring Centrum, Wageningen, The Netherlands, 95 pp, (in Dutch). Houghton, J.T., Callender, B.A. and Varney, S.K. (eds.), 1992. Climate Change 1992, The Supplementary Report to the IPCC Scientific Assessment. Cambridge University Press, Cambridge. Idso, K.E. and Idso, S.B., 1994. Plant responces to atmospheric CO2 enrichment in the face of environmental constraints: a review of the past 10 years' research. Agri. For. Meteor., 69: 153-203. IPCC, 1990. 'First Assessment Report'. In: Land and Water International no. 69. Isarin, R.F.B. & H.H.A. Berendsen, 1992. Morfodynamiek van de rivierduinen langs de Waal en de Lek. Rapport GEOPRO-92.08. Vakgroep Fysische Geografie Universiteit Utrecht. Jarvis, P.G., 1976. The interpretation of variations in leaf water potential and stomatal conductance found in canopies in the field. Phil. Trans. R. Soc. Lon. B., 273: 593-610.
905 Klein Tank, A.M.G. and T.A. Buishand, 1993. Modelling daily precipitation as a function of temperature for climate change impact studies. Scientific Reports WR 93-02, KNMI. De Bilt, The Netherlands. Klein Tank, A.M.G. and T.A. Buishand, 1993b. The occurrence of rain in a changing climate. KNMI Memorandum 93-04. Klein Tank, A.M.G. and T.A. Buishand, 1994. Daily precipitation amounts in a future climate derived from temperature and surface air pressure. KNMI Memorandum 94-03. Klein Tank, A.M.G. & G.P. KSnnen, 1993. The dependence of daily precipitation on t e m p e r a t u r e . In: Proceedings of the 18th annual climate diagnostics workshop, Bolder, Colorado. US Dept. of Commerce. KNMI, De Bilt, The Netherlands. Konikow, L.F. and J.D. Bredehoeft, 1978. Computer model of two-dimensional solute transport and dispersion in ground water. U.S. Geol. Surv. Techn. of Water Resour. Investigat., Book 7, Ch C2, 90. Kwaad, F.J.P.M., 1991. Summer and winter regimes of runoff generation and soil erosion on cultivated loess soils (The Netherlands). Earth Surface Processes and Landforms, Vol. 16, 653-662. Kwadijk, J.C.J., 1993. The impact of climate change on the discharge of the river Rhine. Thesis, University of Utrecht. Kwadijk, J. and H. Middelkoop. 1994. Estimation of the impact of climate change on the peak discharge probability of the River Rhine. Climatic Change. 27: 199-224. Licht, P.M., 1990. Beleidseffecten evaluatie binnen rivierbeheer (in Dutch). Thesis, University of Twente. Delft Hydraulics report Q1065/H472. Manabe, S. and Stouffer, R.J., 1980. Sensitivity of a global climate model to an increase of CO2 concentration in the atmosphere. J. Geophys. Res., 85: (C10), 5529-5554. Manabe, S., Wetherald, R.T. and Stouffer, R.J., 1981. Summer dryness due to an increase of atmospheric CO2 concentration. Climatic Change, 3: 347-385. McFarlane, N.A., G.J. Boer, J.P. Blanchet and M. Lazare, 1992. The Canadian Climate Centre second-generation general circulation model and its equilibrium climate. J. Climate, 5: 1013-1044. Middelkoop. H, 1991. Impact of climatic change on sedimentation on the bottomlands ("uiterwaarden") in The Netherlands. Progress-report GEOPRO91.023. Vakgroep Fysische Geografie Universiteit Utrecht. Middelkoop, H. & N.E.M. Asselman, 1993. Assessment of sedimentation rates on floodplains, a case study in The Netherlands. Poster presentation at the 5th International Conference on Fluvial Sedimentology, July 1993, Brisbane, Australia. Middelkoop, H., E.L.J.H. Faessen & H.J. Huizinga, 1992a. Historische morfologie, hydrologie en ecologie van de Waal tussen Pannerden en Nijmegen. Inundatie en landgebruik van de uiterwaarden en morfologie van het zomerbed tussen 1770 en 1830. Rapport GEOPRO-92.06. Vakgroep Fysische Geografie Universiteit Utrecht, (in Dutch). Middelkoop, H., N.J. van den Berg, E.L.J.H. Faessen & H.J.A. Berendsen, 1992b, Morfodynamiek van nevengeulen van de Waal: een historisch overzicht. Rapport GEOPRO-92.07. Vakgroep Fysische Geografie Universiteit Utrecht.
906 Middelkoop, H. & M. Deurloo, 1993. Geomorfologische en Historisch geografische waardering van het uiterwaardengebied rond Sint Andries. Toetsing van een inrichtingsschets. Rapport GEOPRO-93.13. Vakgroep Fysische Geografie Universiteit Utrecht, (in Dutch). Middelkoop, H., & W.P.A. van Deursen, 1993. Modellering inundaties uiterwaarden Case study Gelderse Poort. Rapport GEOPRO-93.01. Vakgroep Fysische Geografie Universiteit Utrecht, (in Dutch). Middelkoop, H. & H.J. Huizinga, 1992. Assessment of suspended sediment concentrations in the rivers Rhine and Waal during the high water period on April 2nd, 1988 using LANDSAT TM data. Rapport GEOPRO-92.01, Vakgroep Fysische Geografie Universiteit Utrecht. Middelkoop, H. & M. van der Perk, 1991. Een reconstructie van de opslibbing van uiterwaarden. Rapport GEOPRO-91.06. Vakgroep Fysische Geografie Universiteit Utrecht. Middelkoop, H., E.L.J.H. Faessen & H.J. Huizinga, 1992a. Historische morfologie, hydrologie en ecologie van de Waal tussen Pannerden en Nijmegen. Inundatie en landgebruik van de uiterwaarden en morfologie van het zomerbed tussen 1770 en 1830. Rapport GEOPRO-92.06. Vakgroep Fysische Geografie Universiteit Utrecht, (in Dutch). Nonhebel, S., 1987. Water use of Dutch forests: a simulation study. Report 7G, Studiecommissie Waterbeheer Natuur, Bos en Landschap, (in Dutch). Ogink-Hendriks, M.J., 1994. Modelling surface conductance and transpiration of a oak forest in The Netherlands. Agri. For. Meteor., (Submitted). Oude Essink, G.H.P., 1993. Effect of Sea Level Rise on the Groundwater Flow System through Amsterdam Waterworks and Haarlemmermeer polder, The Netherlands. Proc. UNESCO Conf. on Sea Level Changes and their Consequences for Hydrology and Water Management, Noordwijkerhout, The Netherlands. Parmet, B., 1993a. Impact of climate change on the discharge of the Rhine. Change 15: 1-3. Parmet, B. and M. Mann, 1993b. Influence of climate change on the discharge of the River Rhine - a model for the lowland area. IAHS publication 212: 469477. Peerbolte, E.B., J.G. de Ronde, L.P.M. de Vrees, M. Mann, G. Baarse, 1991. Impact of Sea Level Rise on Society. A case study of The Netherlands. Report GWAO 90-016. Rijkswaterstaat, Den Haag. Pomper, A.B., 1983. Geohydrological situation and observations on the hydrochemical groundwater situation of the western Netherlands. Geologie en Mijnbouw 62: 3/4. Roetter, R., 1994. Biophysical classification of the Rhine basin as a frame for land use projections. Volume 1 of Land use projections for the Rhine basin based on biophysical and socio-economic analysis. Winand Staring Centre-RIZA report 85.1. Roetter, R. and C.A. van Diepen, 1994. Biophysically-based analysis of possible climate change impacts on crop yield potentials and water use in the Rhine basin. Volume 2 of Land use projections for the Rhine basin based on biophysical and socio-economic analysis. Winand Staring Centre-RIZA report 85.2.
907 Roetter, R. and C.A. van Diepen, 1994. Biophysically-based, spatial analysis of possible climate change impacts on crop yield potentials and water use in the Rhine basin, volume 2. In press. Rotmans, J., 1990. IMAGE: An integrated model to assess the greenhouse effect. PhD Thesis, Kluwer, Dordrecht. Sch~idler, B., Spreafico, M., Bultot, F. and D. Gellens, 1992. Evaluation Wasserhaushaltmodelle. Vorstudie Nationales Forschungsprogramm 31: "Klima~inderungen und Naturkatastrophen". Stewart, J.B., 1988. Modelling dependence of surface conductance on environmental conditions. Agri. For. Meteor., 43" 19-35. Van Diepen, C.A., C. Rappoldt, J. Wolf and H. van Keulen, 1988. Crop growth simulation model WOFOST version 4.1, documentation. SOW-88-01. Center for World Food Studies, Wageningen, The Netherlands. Van Dijck, S.J.E., 1993. Palaeomagnetic research of the sediments from the dike burst pond in the river foreland of the Waal at Wamel (The Netherlands). Student report, Vakgroep Fysische Geografie Universiteit Utrecht. Van Dinter, M., 1993. Palynologisch onderzoek naar wielopvullingen in het Nederlandse rivierengebied. Student report, Vakgroep Fysische Geografie Universiteit Utrecht, (in Dutch). Van Dinter, M., H. Middelkoop, & B. Derks, 1992. Pollen analysis of dike burst ponds near Nijmegen, The Netherlands. Poster presentation at the 8th International Palynological Congress, Aix-en-Provence, 1992. Van der Drift, J.W.M., 1994. The impact of temperature and rainfall changes (climate change) on land degradation in source areas of the suspended sediment load of the Rhine. NRP Project no. 852089, Interim report no.l, Laboratory of Physical Geography and Soil Science, University of Amsterdam, 33 pp. Van der Drift, J.W.M., Middelkoop, H. and N.E.M. Asselman, 1994. Estimation of the effects of climate and land use change on the production of fine sediment by soil erosion in the catchment area of the river Rhine. Internal report, Laboratory of Physical Geography and Soil Science, University of Amsterdam, Department of Physical Geography, University of Utrecht, 12 pp. Veen, A.W.L. and Dolman, A.J., 1989. Water dynamics of forests: one-dimensional modelling. Progress in Physical Geography, 13: 19-35. Veeneklaas, F.R., L.M. van de Berg, D. Slothouwer and G.F.P. IJkelenstam, 1994. Rhine Basin study: Land use projections based on biophysical and socioeconomic analyses. Volume 4, Land use: past, present and future. Report 85.4, Winand Staring Centre (SC-DLO), Wageningen, The Netherlands. Viner, D. and M. Hulme, 1993. Climate change scenarios for impact studies in the UK: General circulation methods and applications for impact assessment. Climatic Research Unit, University of East Anglia, Norwich. Washington, W.M. and Meehl, G.A., 1983. General circulation model experiments on the climatic effects due to a doubling and quadrupling of carbon dioxide concentration. J. Geophys. Res., 88 (Cll): 6600-6610. Wateren-de Hoog, B. van der, & H. Middelkoop, 1992. Floods in the River Rhine and atmospheric circulation patterns. Rapport GEOPRO-92.02. Vakgroep Fysische Geografie Universiteit Utrecht. Wigley, T.M.L., T. Holt and S.C.B. Raper, 1991. STUGE, an interactive greenhouse model: Users manual, Climate Research Unit, Norwich.
908 Wischmeijer, W.H. and D.D. Smith, 1978. Predicting rainfall erosion losses - a guide to conservation planning. US Dept. of Agriculture, Agriculture Handbook No. 537. Wit, K, 1987. Wateraanvoerbehoefte Zuidhollandse Eilanden en Waarden (Fresh water requirements of the south-western island and polders of Zuid Holland), I.C.W. Nota nr. 1801, Winand Staring Centre, RIZA., Wageningen (in Dutch). Wolf, J, and Diepen, van, C.A., 1993. Effects of climate change on crop production and land use in the Rhine basin. In: Geijn, S.C., Goudriaan, and Berendse, F., editors. Climate change; crops and terrestrial ecosystems. Agrobiologische Thema's 9. 1993, CABO-DLO, Wageningen, The Netherlands. Wolf, J. and van Diepen, C.A., 1991. Effects of climate change on crop production in the Rhine basin. Report 52, Winand Staring Centre, RIZA, Wageningen.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
911
IMPACT OF CLIMATE CHANGE ON THE DISCHARGE OF THE RIVER RHINE
B. Parmet "), J. Kwadijkb) and M. Raak a) a)
RIZA, Institute of Inland Water Management and Waste Water Treatment Section Rivers, P.O. Box 9072, 6800 ED Arnhem, the Netherlands
b)
University of Utrecht, Department of Physical Geography P.O. Box 80115, 3508 TC Utrecht, the Netherlands
Abstract Climate change influences the water balance of drainage basins in several ways. In a project of the International Commission for the Hydrology of the Rhine basin the possible consequences for the discharge regime of the Rhine are investigated. In the first phase of this project detailed models have been developed and applied for selected sub-basins and a simple water balance model has been developed for the whole Rhine basin. Results are presented for climate scenarios assuming an increase in temperature of about 3~ and an increase in annual precipitation. The consequences of such a climate change are largest in the Alpine part of the Rhine basin, and are also considerable in other parts of the basin and for the basin as a whole. In general the Rhine changes towards a rain-fed river. The winter discharge increases, which can have consequences for safety, and summer discharge decreases with consequences for shipping, industry, agriculture and ecology.
1. Introduction
Climate change influences the components of the water balance of drainage basins in several ways. Precipitation patterns may change and because of a higher temperature also the timing and amount of snow fall and snow melt. Evapotranspiration is directly influenced by an increase in temperature. An increased CO2-concentration affects plant physiology and may lead to changes in water use too. Furthermore climate change may induce changes in land use, which is an important factor in evapotranspiration and runoff processes. Changes in these water balance components will of course affect the discharge. The River Rhine is economically the most important river of Western-Europe. Its drainage basin, see figure 1, stretches from the source in Switzerland to the mouth in the North sea, covering an area of 185.000 k n l 2 and is habitat to about 55 million people. The river is one of the most intensively navigated inland waterways in the world and is of major importance for the supply of water to large economically important areas. Changes in its discharge regime can have consequences for safety and for the water availability for shipping, industry, domestic use, agriculture, the natural environment and recreational purposes.
912 Against this background, the International Commission for the Hydrology of the Rhine basin (CHR) initiated a project to assess the consequences of climate and land use changes for the discharge regime of the River Rhine. Since a proper tool for this was lacking, the main purpose of the project was to develop a water management model for the whole Rhine basin. This model should be used to analyze the changes in average and extreme discharges and the effectiveness of mitigating measures. Several institutes from the Rhine riparian states cooperate in the project (Parmet, 1993). The Netherlands contribution to this project is incorporated in the Dutch National Research Program on Global Change (NRP).
0# \
120 km
o
\
........--"
,..,
,,) ,J
~~176 \
.J
-\(y :_ )
( "",3
Figure 1. Drainage basin of the River Rhine
2. Method
The model will be developed in two phases. In the first phase model development takes place along a bottom-up and a top-down approach. In the second phase both approaches will be combined. Along the bottom-up approach hydrological models are developed for selected sub-basins. These models are detailed in time and space and can be used to simulate the effects of climate and land use changes on average and peak flows. Sub-basins were selected in three distinct areas; the Alpine area, the Middle Mountains and the Lowland area. Different model concepts are used because the relevant hydrological
913 processes differ. Snow accumulation and melt, for example, is very important in the Alpine area and groundwater flow for the Lowland. Along the top-down approach a GIS-based water balance model is developed with a course resolution in time and space. This model can be used to study the sensitivity of the average discharge regime of the river Rhine and its main tributaries for climate change. The first phase of the CHR project is almost finalized. Along the bottom-up line, several detailed models are developed and also existing models are applied for representative subbasins. For the Alpine area an existing hydrological model, the IRMB model, was applied for several small drainage basins (Bultot, 1992, Sch/idler 1992). In the Middle Mountains area, a model will be developed for the Saar basin, a sub-basin of the Mosel. Results are not yet available. The Lowland model is developed for the drainage basin of the Overijsselsche Vecht (Parmet and Mann, 1993, Parmet and Raak, in prep). This model consists of a hydrological component, that is used to compute the daily evapotranspiration and discharge for sub-basins, and a flow-routing component, that combines the sub-basins and routes their discharges towards the mouth of the Overijsselsche Vecht. For the Rhine basin as a whole the water balance model RHINEFLOW was developed (Kwadijk, 1993). It is a simple water balance model based on a Geographical Information System.
3. Results 3.1. Effects of climate change in representative basins in the Alpine area With the IRMB model the effects of a climate change for several components of the water balance were simulated for three drainage basins, Murg, Ergolz and Broye. A climate scenario as defined by Bultot was applied (Bultot, 1988). The monthly temperature and precipitation changes are given in table 1. Changes in physiological behavior of plants were not taken into account. Simulations with changed climate were carried out for the period 1981 to 1988. The simulations showed an increase of annual potential evapotranspiration of 10%. Actual evapotranspiration increased somewhat less because during the summer period there is a slight decrease in soil moisture. Discharge increased during the winter period with 10%. This is due to the fact that the amount of winter precipitation increases and less precipitation is stored as snow. Furthermore the accumulated snow melts faster. The duration of the snow cover decreases considerably, especially below an altitude of 1500 m. Discharge in summer decreased with about 15%. This follows from a reduced contribution of melt water, a larger evapotranspiration and a slight decrease in precipitation. The total annual discharge hardly changes. The daily maximum discharge increases and the daily minimum discharge decreases. 3.2 Effects of climate change in representative basins in the Lowland area The climate scenario that was used for the Lowland area was based on a method developed in the framework of the NRP (Klein Tank and Buishand, 1995). The monthly changes in temperature and precipitation are given in table 1. Compared to the other scenarios used in this study, this scenario is rather wet. Computations were carried out for the basin of the Overijsselsche Vecht for the period 1965-1990. Changes in plant physiological characteristics were taken into account.
914 An increased CO2-concentration influences plant physiology. For most plants the water use efficiency increases and the biomass production increases. An increase in temperature for the temperate zones generally leads to an increase in production too. Present knowledge indicates, for doubled CO2-concentrations and an increase in temperature of about 1.5~ a small decrease in evapotranspiration for most crops and forests (Roetter en van Diepen, 1994; Hendriks, 1994). Based on this knowledge, plant physiological parameters were provisionally adapted in the Lowland model (Parmet and Raak, in prep).
Table 1 Monthly temperature (T) and precipitation (P), used for representative Alpine basins and for the representative Lowland basin Month
J
F
M
A
M
J
J
A
S
O
N
D
Talpine, ~ Palpine, % 1)
3.1 10
3.4 14
3.4 11
3.1 10
2.8 -1
2.7 -2
2.5 -2
2.3 -2
2.3 0
2.7 6
2.8 10
3.2 10
Tlowland, ~ Plowland, %
3.0 21
3.0 20
2.3 15
2.3 13
2.3 5
3.7 12
3.7 11
3.7 9
3.4 3
3.4 8
3.4 19
3.0 18
1) The percentual change is an average for the three basins Murg, Ergolz and Broye (Sch~idler, 1992)
700
E d L.. ~m
6oo 500
t-
.-
400
E m E
300
E
2oo
r-
100
U
f
f
f
~
Reference --
- Scenario
r
< 0
2
i
I
i
I
I
I
5
1o
.50
1 oo
200
500
Reccurence
1250
period, [years]
Figure 2. Discharges for different recurrence periods for reference and scenario conditions, as computed with the lowland model for the drainage basin of the Overijsselsche Vecht.
915
Simulations for the lowland showed an increase in actual evapotranspiration of about 7%. The CO2 effect on water use does not compensate for the increase in temperature of 3 ~ The annual discharge increased with 22%. Winter discharge increased with 25%. The increase in evapotranspiration in summer does not exceed the increase in precipitation. Discharge in summer increased with 19%. The maximum discharge increased considerably. The distribution of annual maxima can be described using a Gumbel distribution (Mendel, 1993). According to fitted Gumbel functions for the reference (r 2 = 0.98) and scenario (r 2 = 0.97) simulations, peak flows with different recurrence periods change as indicated in figure 2. River dikes in the Netherlands are designed for a discharge with a recurrence period of 1250 years. The figure shows an strong increase of 26% in this design discharge for the river Vecht. 3.3 Effects of climate change, Rhine basin Consequences for the whole Rhine basin have been computed with the RHINEFLOW model. The sensitivity of the discharge regime was examined with a wide range of climate scenarios for the period 1956 to 1980 (Kwadijk, 1993). Here the results of computations with one scenario, the so-called BAU-best scenario, are presented. The scenario is based on the IPCC Business as Usual scenario (IPCC, 1991). It is given in table 2. Changes in land use and physiological characteristics of plants were not taken into account.
Table 2 BAU-BEST scenario for temperature (T) and precipitation (P), for different parts of the Rhine basin (Kwadijk, 1993) Year
Summer
Winter
Part of Rhine basin
T, ~
P, %
T, ~
P, %
T, ~
P, %
North Middle South
3.5 3.5 3.5
11 8 7
2.9 2.9 2.9
4 -1 -4
4.3 4.2 4.1
19 19 19
For the Alpine part of the Rhine basin, the changes as computed with RHINFLOW have the same direction as the results for the representative Alpine basins. As can be seen from figure 3, winter discharge increases. This is caused by increased precipitation and snow melt. During summer the discharge decreases due to a smaller contribution of melt water, increased evapotranspration and a slight decrease in precipitation. The increase in winter discharge is much larger than for the representative basins, upto 100% with an average of 60%. This can be explained partly from the used scenarios. Both the increase in temperature and in precipitation is smaller for the Bultot scenario compared with the BAUbest scenario. Furthermore it can be explained by differences in model components, especially the snow component, and off course the considered area is not the same. The changes during summer are comparable, both for the alpine area as a whole and for the
916 representative basins in the alpine area, a decrease of about 15% was computed. The changes for the area downstream, the Middle and Lowland part, are much less pronounced. The discharge increases during winter and spring and decreases during summer and autumn, as can be derived from figure 3. The increase in evapotranspiration causes the soil water deficit to increase. As a result, summer discharge decreases, but because part of the winter surplus is stored as groundwater, not untill July. The water surplus during autumn is partly used to replenish soil water, which explains the decrease of discharge during autumn. Because the scenario used for the representative basin for the lowland is wetter, especially for the summer period, than the BAU-best scenario, the changes in discharge are not directly comparable. 120
100
o~~ 80 ,.._,
Alpine part -- - Middle & lowland part
40
- - - Outlet basin 20 l-9
~
R li"~
~
~
~
.c
0
0 -20 -40
o>
Z
~
O
~
~
~LL
~
~
~_
<:
~=-
:~
==
"~
~
==
<
~
CO
O
Month
Figure 3. Changes in monthly discharge for different parts of the Rhine basin and near the basin outlet, for simulations with the BAU-BEST climate scenario, for the period 19651980.
Where the River Rhine enters the Netherlands, near the basin outlet, the changes in the Alpine, Middle and Lowland part are combined. The annual changes are small, the discharge increases with 2%. However, winter and spring discharge increase with about 15%, and summer and autumn discharge decrease with about 10%, see figure 3. Due to the changes in the alpine area the regime of the River Rhine changes from a combined rainfed/snow-fed into a rain-fed river. The discharge pattern will become less smooth and the difference between maximum and minimum flows will increase. The number of months with low flows will increase. For example, the results indicate that the number of months with an average discharge below 1000 m3/s increases for the period 1956 to 1980 with 13, which is about 60%. To make an assessment about peak flows with the monthly discharges computed by RHINEFLOW, a relation between average monthly flows and peak flows was derived. Using this relation it was shown that the probability of discharge peaks upto 7000
917
m3/s increases significantly. This is important in respect to sedimentation on floodplains, because these are inundated for such discharges. The relation does not give well-founded results for the design discharge, and hence about consequences for safety.
4. Implications Interim results of the CHR project show that climate change can have considerable effects for the discharge regime of the Rhine. With the assumed scenario for the Rhine basin, the winter discharge increases considerably. This could have consequences for safety, but the models are not yet suitable to assess consequences for design discharges. The contribution of melt water originating from the Alps during the summer period decreases, which is an important reason for a decrease in summer discharge. Furthermore evapotranspiration is expected to increase, which contributes also to a decrease in summer discharge. Consequently the frequency of periods with low flows increases. For water management in the Netherlands an increased frequency of low flows implies increasing costs for shipping. Ships can be loaded less and have to wait longer for sluices and bridges. Costs of electricity production will increase too. To avoid environmental problems with the temperature of cooling water, other more expensive, production units have to be brought into operation. Other industries that use water for cooling purposes may have to limit production or built cooling towers. Lower water levels will also have undesirable effects on ecological development in floodplains, dessication problems are expected to increase. Furtermore the increased frequency in low flows wil result in a more frequent intrusion of salt water. This can cause problems for intake of water of certain polders and may cause damage to agriculture. In general the changed discharge regime will also influence river morphology. For the Rhine basin land use scenarios have been developed in the framework of the CHR project (Roetter, 1994; Roetter and v. Diepen, 1994; Hendriks, 1994; Veeneklaas et al, 1994). The effects of land use changes have not yet been studied in detail. From first computations it can be concluded that for Lowland areas the total discharge rather than peak discharges will be affected. For the Alpine and Middle Mountains area it is expected that also the peak flows are influenced. First estimates show that land use changes are very important for the production of suspended sediment, sediment transport and sedimentation processes (Asselman, 1994; Middelkoop, 1994). Further study is required. The model RHINEFLOW in its present form is a usefull tool for sensitivity analysis. However, the simple process descriptions and the poor quality of the underlying database, limit its applicability. On the other hand the detailed models are only available for a relatively small part of the Rhine basin. To extend these models for the whole basin is a time consuming task. Therefore a promissing direction is to couple the rough and the detailed models, for example with transfer functions. RHINEFLOW has to be refined in time and space for this. Furthermore the detailed models have to be applied also in other characteristic areas, to cover the variability within the Rhine basin in a better way. It is recommended to investigate the possible effects of climate change also for other important river systems, like the Meuse. A similar approach as in the CHR project can be applied.
918 5. References
Asselman, N., 1994. The impact of climate change on suspended sediment transport in the river Rhine. Report University Utrecht, Department of Physical Geography. Bultot, F., Coppens, A., Dupriez, G.L., Gellens, D. and Meulenberghs, F., 1988. Repercussions of a CO2 doubling on the water cycle and on the water balance - A case study for Belgium. Journal of Hydrology 99, p. 219-347. Bultot, F., Gellens, D., Spreafico, M. and B. Sch~idler, 1992. Repercussions of a CO 2 doubling on the water balance - a case study in Switzerland. Journal of Hydrology, 137, p. 199-208. Hendriks, C.M.A., 1994. Biophysically-based analysis of possible climate change impacts on forest yield potentials and water use in the Rhine basin. Volume 3 of Land use projections for the Rhine basin based on biophysical and socio-economic analysis. Winand Staring Centre-RIZA report 85.3. IPCC, 1991. Climate Change: The IPCC Response Strategies. Cambridge. University Press. Klein Tank, A.M.G. and Buishand, T.A., 1995. Transformation of precipitation time series for climate change impact studies. KNMI report WR 95.01. De Bilt. Kwadijk, J., 1993. The impact of climate change on the discharge of the River Rhine. PhD Thesis. Netherlands Geographical Studies no. 171. Utrecht. Kwadijk, J.C.J. and Middelkoop, H., 1994. Estimation of impact of climate change on the peak discharge probability of the river Rhine. Climatic Change 27, p. 199-224. Mendl, H.-G., 1993. Verteilungsfunctionen in der Hydrologie. CHR/KHR report II-8. Middelkoop, H., 1994. The impact of climate change on the sedimentation rates on the embanked floodplains in the Netherlands. Report University Utrecht. Department of Physical Geography Parmet, B., 1993. Impact of climate change on the discharge of the Rhine. CHANGE 15, p. 1-3. Parmet, B. and M. Mann, 1993. Influence of climate change on the discharge of the River Rhine - a model for the lowland area. IAHS publication 212, p. 469-477. Parmet, B. and Raak, M. In prep. Impact of climate change on the discharge of the river Rhine - A case study for the lowland drainage basin of the Overijsselsche Vecht. Rijkswaterstaat RIZA report. Roetter, R., 1994. Biophysical classification of the Rhine basin as a frame for land use projections. Volume 1 of Land use projections for the Rhine basin based on biophysical and socio-economic analysis. Winand Staring Centre-RIZA report 85.1. Roetter, R. en C.A. van Diepen, 1994. Biophysically-based analysis of possible climate change impacts on crop yield potentials and water use in the Rhine basin. Volume 2 of Land use projections for the Rhine basin based on biophysical and socio-economic analysis. Winand Staring Centre-RIZA report 85.2. Schfidler, B., Spreafico, M., Bultot, F. and D. Gellens, 1992. Evaluation Wasserhaushaltmodelle. Vorstudie Nationales Forschungsprogramm 31: "Klima~.nderungen und Naturkatastrophen". Veeneklaas, F.R, van den Berg, L.M., Slothouwer, D. en G.F.P. IJkelenstam, 1994. Land use: Past, present and future. Volume 4 of Land use projections for the Rhine basin based on biophysical and socio-economic analysis. Winand Staring Centre-RIZA report 85.4.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
919
The effects of an increase in CO2 on the hydrology of forests H.J.M. Lankreijer Department of Physical Geography, University of Groningen, Kerklaan 30, 9751 NN Haren, The Netherlands
Abstract The possible impacts of a climatic change associated with an increase in CO2 on the hydrology of forests is evaluated by applying sensitivity analysis and a climatic scenario to a one-dimensional model. Water consumption of forests is affected by changes in plant physiology and meteorological environment. Changes in species composition are not taken into account. Increase in CO2 leads to a decrease of stomatal conductance, resulting in a decrease in transpiration of 10-30%. The evaporation of rainfall interception by the canopy is increased due to a higher leaf area index and higher temperatures. Application of a wet scenario shows an increase in total interception, but the ratio between interception and precipitation decreases. Simulating a small increase in leaf area increases the evapotranspiration only weakly and the higher precipitation in the scenario is mainly passed on to drainage. Drought damage in summer should reduce, but winter discharge may strongly increase.
1. INTRODUCTION Forests are not only important in the global carbon balance but have also a relevant influence on local and regional hydrology. A change in the concentration of CO2 as well as a possible climatic change will have direct and indirect effects on the water use of plants, including trees. The direct effect of CO2 increase will be most significant on the photosynthesis. Besides, forest are strongly coupled to the atmospheric conditions because of their roughness. Changes in the atmosphere will affect forests stronger than other vegetation types. The aim of this study is to estimate the consequences of a climate change with an increase of atmospheric CO2 concentration for the water balance of forests.
2. SIMULATION OF THE FOREST WATER BALANCE A one dimensional model was developed to simulate the water balance. The model, derived from the model by Dolman et al. [1], is devided in three main parts. Transpiration
920 through the leaves is simulated using the Penman-Monteith equation, which contains the stomatal conductance. Interception and evaporation of rainfall by the canopy is modelled according to the Gash-Rutter [2] approach. And thirdly, the soil is considered as a simple bucket and water exceeding the field capacity is drained. Transpiration is calculated on a hourly time basis, whereas interception and soil water content are calculated with a daily time step. Transpiration of forests is strongly determined by the stomatal conductance. As stomata are sensitive to changes in the CO2 concentration, a model simulating the responses of stomata is essential [3]. The correlation between stomata, meteorological variables and plant physiological processes is evident, but the exact mechanistic process of stomatal functioning is still unknown. This situation has resulted in several possible empirical simulation models. In this study the model developed by Leuning [4] was chosen. To incorporate the effect of soil moisture deficit on the stomatal conductance, the soil moisture function of Stewart [5] is used in combination with the Leuning model. The model describes a relation of stomatal conductance with the assimilation rate and air humidity deficit. To simulate the assimilation rate the biochemical model of Farquhar et al. [6] is used. 2.1 Climate scenario The influence of the main model parameters on interception and transpiration were analyzed by sensitivity analysis. To integrate the results with results of other groups with in the National Research Program, the scenario KNMI-2 as described by K6nnen [7] was applied. The amount of precipitation increases strongly (with a yearly of average 13 %) as compared to climate scenarios described by IPCC. In order to apply the scenario, an estimation must be made of the forest parameters in a changed climate. Changes in the parameters are derived from studies to responses of trees in a changed climate. An excellent review is given by Ceulemans and Mousseau [8]. The reaction of trees depends on species and nutrient environment with the possibility of the tree to develop sinks for the extra carbon. A general trend is the increase of growth and, with some exceptions, of the water use efficiency, but the increase is limited when no sinks are available. With limited nutrients, which is the case for many forests on sandy soils, the increase of biomass will be concentrated in the roots. Therefore, in the scenario a modest increase of 5 % in leaf area index (LAI) of the canopy is applied. The onset of leaf growth and leaf fall are not changed. Also the nitrogen content is not changed to keep the approach simple. Most species show a decrease in stomatal conductance caused by both smaller stomatal opening and lower stomatal density. In this study doubling of CO2 resulted in a decrease of stomatal conductance of 31%. Based on Morison [9], who compared the stomatal conductance of various species under ambient and double CO2 concentrations, these changes are regarded as realistic. Although, variations due to varying species composition and forest site may be large. 2.2 Data The model was calibrated on data of a decidiuos forest (Quercus rubra ) located near Erie in the Netherlands and of the coniferous forest (Pinus sylvestris) of Thetford in England. The model had to be fitted on measured transpiration fluxes and soil moisture content because assimilation data were not available. The simulated variation in the
921 Precipitation
1000
--
i
i
i
1974
1975
1976
1
1977
1978
avg
i Normal
1961-1990
Year
Decidious
Coniferous
Interception 250 -
501974
1975
1976
1977
1978
avg
1974
1975
1976
1977
1978
avg
1975
1976
1977
1978
avg
Transpiration 500 45O 400
..........................
T
...........
300 250 1974 Precipitation
650
1975
1976
1977
1978
avg
1974
excess
i
-
l
~ 350
.
'
1974
1975
1976
1977
.
.
.
.
.
.
.
.
.
.
.
.
t
1978
avg
1974
1975
1976
1977
1978
avg
Year
D
Normal climate
I~
Scenario
Figure 1. Yearly totals of precipitation, interception, transpiration and precipitation excess simulated with normal climate and scenario for both forest types.
assimilation rate with the Farquhar model was judged to be realistic. To simulate the water balance over longer periods, the climate scenario study was conducted with a data covering the years 1974-1978, including dry and wet years. The variables were measured above grass and transformed to above forest conditions. The characteristics of Ede and Thetford forests were used to describe the modelled forests.
922 4. RESULTS AND CONCLUSIONS Interception of rainfall is especially sensitive to changes in temperature, air humidity, windspeed and leaf area. With no change in the relative humidity when temperature increases, the small increase in leaf area index raises the interception with a small amount (see figure 1). Although precipitation increases strongly, the effect on interception is low. The ratio of interception and total precipitation is decreasing. This means that a larger part of the gross precipitation reaches the soil. Using the physiological based model for stomatal conductance, transpiration appears to be sensitive not only to changes in temperature, air humidity and soil water availability but also the nitrogen availability and the photosynthetic capacity have a strong influence. The available soil water determines the change in transpiration compared to the actual climate. Lower transpiration during winter and spring due to decreasing stomatal conductance, diminish water shortage during summer and the total yearly transpiration may increase. When water is not limiting transpiration decreases with 10-30%. Taking the uncertainties of the applied scenario and used model into consideration, the results must be used with care. The increase in precipitation, especially high during winter, will drain almost completely to the groundwater. This means that frequent flooding can be expected in winter when evapotranspiration is low and the soil is already saturated. Summer droughts will decrease. The present policy to replace coniferous forests with deciduous forests to limit evapotranspiration, will further increase discharge in a greenhouse climate.
5. REFERENCES
1 Dolman, A.J., Stewart, J.B. and Cooper, J.D., 1988. Predicting forest transpiration from climatological data. Agri. For. Meteor., 42: 339-353. 2 Gash, J.H.C., 1979. An analytical model of rainfall interception by forests. Q.J.R. Meteorol. Soc., 105:43-55. 3 Lankreijer, H.J.M. and Veen, A.W.L., 1989. Possible changes in forest hydrology following a global climatic change. Proc. WMO Conf. Climate and Water, Publ. Acad. Finland, Helsinki, 1989; vol. I, pp. 187-196. 4 Leuning, R., 1994. A critical appraisel of a combined stomatal-photosynthesis model for C3 plants. Submitted to Plant, Cell and Environment. 5 Stewart, J.B. 1988. Modelling dependence of surface conductance on environmental conditions. Agri. For. Meteor., 43: 19-35. 6 Farquhar, G.D., von Caemmerer, S., and Berry, J.A., 1980. A biochemical model of photosynthetic CO2 assimilation in leaves of C3 species. Planta 149: 78-90. 7 K6nnen, G.P., 1994. Scenarios climate change. Assessement Report NRP, sub theme Regional Hydrology, pp 9-13. 8 Ceulemans, R. and Mousseau, M., 1994. Tansley review No. 71. Effects of elevated atmospheric CO2 on woody plants. New Phytol., 127: 425-446. 9 Morison, J.I.L., 1987. Stomatal responses to CO2. In: E. Zeiger, G.D. Farquhar and I.R. Cowan (Eds), Stomatal Function. Stanford University Press, Stanford California, 229-251.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
923
The effect of temperature change on soil structure stability Drs. J.W.M. van der Drift
University of Amsterdam, Landscape and Environmental Research Group, Nieuwe Prinsengracht 130, 1018 VZ Amsterdam, The Netherlands
Abstract The influence of variations of temperature, organic carbon content and land use on soil structure stability was investigated by statistical analysis of field and soil data of two loess covered areas in the German part of the Rhine Basin. Aggregate stability as an indicator of structure stability was determined and related to land use, slope aspect and soil organic carbon content. Because aggregate stability showed no relation with aspect, there is little evidence t h a t soil structure stability is related to temperature. Relations between aggregate stability and soil organic carbon content and between aggregate stability and land use were found to be significant.
1. I N T R O D U C T I O N
The occurrence and rate of soil erosion is controlled by a n u m b e r of causes and factors, viz. ability of precipitation to cause erosion (erosivity), length and steepness of slopes, resistance of the soil to erosion (erodibility), and land use (Wischmeier and Smith, 1978; Dikau, 1986). Climate change may, directly or indirectly, affect some of these causes and factors. This study concentrates on the impact of climate change on soil structure stability, because structure stability controls the rate of soil structure degradation, and it is still in an under-developed research area. Climate affects soil structure stability through its influence on the organic m a t t e r status of the soil, which depends on biomass production and decomposition
924 of organic materials by soil (micro)biological activity (Tate, 1987). Soil structure is characterized by the presence of soil aggregates, clusters of soil particles which mutually adhere by chemical and physical binding forces. In surface soils, these forces are mainly controlled by organic matter. Macro-aggregates (>2501~m) are mainly stabilized by plant roots and larger fungi (Tisdall and Oades, 1982). Microaggregates (20-2501~m) are bound together by decomposed organic substances and bacteri. Micro-aggregation in the size class 2-22~m is mainly caused by clay particles, and to a lesser extent by organic materials (Oades, 1993; Tisdall and Oades, 1982). The rate of structure development and structure breakdown is dependent on the dynamics of soil organic matter, which, in turn, is controlled by soil moisture and soil temperature regime (Chaney and Swift, 1984; Tate, 1987; Jenkinson and Ayanaba, 1977). The study of the impact of climate change on soil structure stability is carried out in two areas covered with loess in the catchment of the River Rhine (Dikau, 1986; Clemens and Stahr, 1994; Schalich, 1981). Objective of research is: to analyze the influence of temperature variation on soil structure stability by establishing relationships between aggregate stability and environmental factors like land use and temperature, and soil properties like soil organic matter content and particle size distribution for loess soils in the German part of the Rhine Basin. Based on a comparative study of aggregate stability in different parts of the Rhine basin with different temperature conditions under the current climate, existing differences in structure stability between areas with different temperature regimes might be an indication of the change in soil structure stability that will possibly occur as a consequence of climate change.
2. STUDY AREA AND METHODS
In the German part of the Rhine basin areas covered with loess were selected to compare the structure stability of the soils and analyze the relationship with environmental factors and soil properties: "the Kraichgau" in the federal state Baden Wiirttemberg in the drainage basin of the river Neckar and the "Jfilicher B5rde" in the federal state Nordrhein Westfalen, stretching from the drainage area of the Lippe to the west bank of the Rhine into the basin of the Meuse. These areas differ amongst others with respect to climate, relief and mean elevation. The soil types present in these areas are Braunerden grown with forest, Parabraunerden (agriculture use or forest) and Pararendzina's (agricultural use). The climate of the Kraichgau is warm and dry (semi-continental). Mean annual precipitation as displayed in table 1 is 806 mm. The precipitation of the period June till August is dominated by high-energy storms. Snow cover and soil frost
925 m a y occur from December till March.
Table 1 Climatological
data for station H e i d e l b e r g
Month
J
F
M
A
M
J
J
A
S
0
N
D
YEAR
Tmean Tmax Tmin Prec
1.3 3.6 -1.4 66
2.4 5.3 -0.8 52
6.7 10.9 2.6 45
10.7 15.6 6.1 61
15.0 20.4 9.9 73
18.1 23.4 12.9 90
19.8 25.1 14.8 87
19.0 24.4 14.8 90
15.8 20.8 11.6 65
10.6 14.5 7.2 62
6.1 8.5 3.6 60
2.4 4.5 0.0 55
10.7 14.8 6.7 806
~ ~ ~ mm
Source: Mfiller (1979)
T e m p e r a t u r e d a t a from table 1 are t a k e n from Mfiller (1979). M e a n a n n u a l t e m p e r a t u r e for station Heidelberg is 10.7 ~ The w a r m e s t m o n t h is July, average t e m p e r a t u r e 19.8 ~ the coldest m o n t h is J a n u a r y , average t e m p e r a t u r e 1.3 ~ The a n n u a l m e a n m i n i m u m t e m p e r a t u r e is 6.7 ~ the a n n u a l m e a n m a x i m u m t e m p e r a t u r e is 14.8 ~ Climatic d a t a of the Jfilicher BSrde are t a k e n from Mfiller (1979) as displayed in table 2.
Table 2.2 Climatological
data of station Aachen.
Month
J
F
M
A
M
J
J
A
S
0
N
D
YEAR
Tmean Tmax Tmin Prec
1.8 4.2 -0.9 72
2.2 5.1 -0.7 59
5.6 9.9 2.0 49
8.9 13.5 4.8 63
12.9 17.8 8.1 67
16.0 20.8 11.4 77
17.6 22.3 13.3 75
17.2 22.2 13.3 82
14.5 19.5 10.9 68
i0.i 14.0 6.9 64
6.0 8.6 3.6 67
3.1 5.2 0.7 62
9.7 13.6 6.1 805
Source: Mfiller (1979)
The m e a n a n n u a l precipitation is 805 mm, with a m a x i m u m m e a n in A u g u s t of 82 m m , and the driest m o n t h is March with 49 mm. The period from J u n e to August is dominated by high-intensity storms, with a m o n t h l y m a x i m u m total of 82 m m in August. The a n n u a l m e a n t e m p e r a t u r e is 9.7 ~ the m e a n t e m p e r a t u r e of the w a r m e s t m o n t h (July) is 17.6 ~ the coldest m o n t h is J a n u a r y with a m o n t h l y m e a n t e m p e r a t u r e of 1.8 ~ The effect of climate on soil s t r u c t u r e was studied on a meso-scale by comparing loess soils on north- and south-facing slopes with m i n i m u m variation of geology and topography. Besides climatologically induced variations, also differences in soil s t r u c t u r e between land use types (arable land and forest) and between surface soil and subsoil were investigated.
926 Soil samples were tested on presence of lime, organic carbon content (Allison, 1935), aggregate stability (Low, 1954; Grieve, 1979; Imeson and Vis, 1984), soil texture and micro-aggregation (Edwards and Bremner, 1967; Imeson and Vis, 1984). Results of the drop test of Low were transformed to the AS-index of aggregate stability (Hollemans and Van Dijk, 1988). Aggregate stability can be expressed as the reciprocal of AS. With statistical methods of data analysis (cluster analysis, Analysis of Variance, Mann-Whitney U-test; see Davis 1986) differences between groups of samples were evaluated.
3. R E S U L T S
The statistical analysis of the relations between environmental factors and soil structure stability of loess soils shows that the direct and indirect effects of temperature on
20 o~
16
.~-
12
_.< o ~
,s ,. ................................................................
I--A-ARABLE I .................................. -*-FOREST .......
,,
;,'.~ ...............................
.',, .............................................................................................
o o
,<>..............
9
8
~
4
~'
',,
,
............... i:: ........................................................................................................
........................ i ............ ;:;" .......
.i
,"
<
;- ........................
_
9 ...........................
, ...........................
9 ...........................
:
:
}
~.
k,
.,,
North Northeast
; ...........................
~
~
"'--~'"
1 ......................................................
.~
East Southeast South Southwest West Northwest Slope aspect
Figure 1 Aggregate stability of agricultural soils and forest soils grouped to aspect
927 aggregate stability is difficult to recognize in the field9 The mean values of aggregate stability of arable surface soils do not vary for slopes having different aspects (figure 1). For forest topsoils the values of aggregate stability vary widely with aspect. The results as displayed in figure i suggest that forest soils are stable on south-east facing slopes, and have minimum stability on north facing slopes. There is no clear recognizable pattern of variation of aggregate stability with slope aspect. There is a strong and significant relationship (R = -0.82) between the index of aggregate stability (AS) and soil organic carbon content. After log-transforming the data to stabilize the variances of both variables (Davis, 1986), the regression equation is: (1)
LOG(AS) = 0.099 - 1.138 LOG(%C)
AS decreases with increasing carbon content, which means that aggregates become more stable with increasing carbon content of the soil (figure 3).
9 8
o~
7
9 o
6
>,,
...................
! ..........................................................................
I ......................................
1 ......................................
( ..................
i ......................................
i~ ...................
iiii!iiill............. iiiiiiiii ii i
.m
5 4 3 ,L
<
2
1 0
..................
i
arable Topsoils
............
forest
pasture LAN DUSE
Figure 2 Aggregate stability grouped to land use
sub arable
sub forest Subsoils
928 From figure 2 the influence on aggregate stability of land use is obvious: forest surface soils are most stable, while arable surface soils have more stable aggregates t h a n subsoils. The aggregate stability of surface soils under pasture is between t h a t of forest and arable soils. The variation of temperature and soil moisture due to different aspect of valley slopes showed no good correlation with aggregate stability.
3
~ , I Q I
0i
2 ........... ;"-""~~ o-=.--.~, .... i" ~ ' ~o'~" ~..i
1 ~ ~"
0
,
~
Oo,
i
:
:
....
i ....
i ....
i ....
i ....
i ....
o~{,~ ............. i ................. } ................ + ................ i ................. {.................
""-.".
i
i
-..~...,ooo o
i
i
,
....... o ~ ~
i
,
,
.................~.................~.................
................. } ................ !---:-: ......... !.........
............
.................
.................................................................................................~ ' ~ ]
~-1 ___0
'
]
i
L
i
'
~9"oi
.....
-2
...................................................g ................................................................... o ....... i....."-:.7
-3
.................................................... ].................................................................... ~.o.............~................
ooo
o
i
o
-4 ............................................ -2.5 -2 -1.5 -1 -0.5 log(organic
~ 0
0.5
carbon)
%
1
1.5
2
Regression 95%confid.
Figure 3 Relation between organic carbon content and aggregate stability index
Table 1 Index land use
l a n d use Arable Forest Pasture subsoil arable subsoil forest all s a m p l e s
of aggregate
stability
of soils
JQlicher B~rde
Kraichgau
AS
AS
1.56 0.43 0.57 4.58 6.60 2.75
1.61 0.29 0.61 5.41 6.61 2.91
from
two
sample
areas
grouped
to
929 The aggregate stability of soils of two different areas is listed in table 1. The difference of means of aggregate stability calculated for 5 groups of land use was not significantly different for areas under different climate. The mean of aggregate stability for all groups indicated that soils in the Kraichgau area (with w a r m e r and drier climate) are somewhat less stable than in the Jtilicher BSrde, but this difference was not significant on the a=0.05 level. However, as the tests give an indication of the present state, there still can be a difference between the development of aggregate stability, due to differences in climate. The present study of the effect of temperature variations on aggregate stability of loess soils shows that temperature does not have a significant influence on the structure stability of soils. Soil structure stability is strongly correlated with the soil organic carbon content, which is determined by the land use. According to the study of Veeneklaas et al (1994), climate change may have a profound effect on land use, thereby implicitly influencing soil properties as well.
4.
R
E
F
E
R
E
N
C
E
S
Allison, L.E. (1935): Organic carbon by reduction of chromic acid. Soil Sci. 40:pp.311-320 - Chaney, K. and R.S. Swift (1984): The influence of organic matter on aggregate stability in some British soils. Journal of Soil Science, 35, pp.223-230 - Clemens, G. and K. Stahr (1994): Present and past soil erosion rates in catchments of the Kraichgau area (SW-Germany). In: Catena 22 (1994) pp.153-168 - Davis, J.C. (1986): Statistics and data analysis in geology. 2nd edition. John Wiley & Sons, New York Chichester - Dikau, R. (1986): Experimentelle Untersuchungen zu Oberfl~ichenabfluss and Bodenabtrag von Messparzellen und landwirtschaftlichen Nutzfl~ichen. Heidelberger Geographischen Arbeiten Heft 81. Heidelberg, Germany - Edwards, A.P. and J.M. Bremner (1967): Dispersion of soil particles by sonic vibration. J. Soil Sc.18:47-63 - Grieve, I.C. (1979): -
930 Soil aggregate stability tests for the geomorphologist. Technical Bulletin 25. British Geomorphological Research Group Hollemans, W. and P. van Dijk (1988): Faktoren van invloed op infiltratie en bodemerosie. Regionale vergelijking tussen drie gebieden in zuid-limburg. Fysisch Geografisch & bodemkundig laboratorium, Universiteit van Amsterdam (In Dutch) Imeson, A.C. and M. Vis (1984): Assessing soil aggregate stability by water-drop impact and ultrasonic dispersion. Geoderma, 34:85-100 - Jenkinson, D.S. and A. Ayanaba (1977): Decomposition of Carbon-14 Labeled plant material under tropical conditions. Soil Sci. Soc. Am. J., Vol 41, pp912-915 - Low, A.J. (1954): The study of soil structure in the field and in the laboratory. Journal of Soil Science, 5, pp57-74 Mfiller, M.J. (1979): Handbuch ausgewahlter Klimastationen der Erde. Forschungsstelle Bodenerosion der Universit~it Trier Mertesdorf (Ruwertal) Oades, J.M. (1993): The role of biology in the formation, stabilization and degradation of soil structure. Geoderma 56, pp 377-400 - Schalich, J. (1981): Boden- und Landschaftsgeschichte in der westlichen Niederrheinischen Bucht. In: Fortschr. Geol. Rheinld. u. Westf. 29, pp 505-518 Tate III, R.L. (1987): Soil organic Matter. Biological and ecological effects. J. Wiley & sons Tisdall, J.M. and J.M. Oades (1982): Organic matter and water-stable aggregates in soils. Journal of soil science 33 pp 141-163 Veeneklaas, F.R., L.M. van den Berg, D. Slothouwer and G.F.P. IJkelenstam (1994): Rhine Basin study: Land use projections based on biophysical and socio-economic analyses. Volume 4, Land use: past, present and future. Report 85.4, Winand Staring Centre (SC-DLO), Wageningen, the Netherlands - Wischmeier, W.H. and D.D. Smith (1978): Predicting rainfall erosion losses - a guide to conservation planning. U.S. Department of Agriculture, Agriculture Handbook No. 537
-
-
-
-
-
-
-
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
931
T h e i m p a c t o f c l i m a t e c h a n g e on the s e d i m e n t a t i o n rates on the e m b a n k e d floodplains in the Netherlands H. Middelkoop Department of Physical Geography, Utrecht University, P.O. Box 80.115, 3508 TC Utrecht, The Netherlands
Abstract
Sedimentation on the embanked floodplains in the Netherlands was studied on different time scales. Many sections of the embanked floodplain along the river Waal were formed only during the past centuries. Spatial differences in floodplain sedimentation rates were investigated using heavy metals as a tracer. Average sedimentation rates during the past decennia range between 2 and >15 mm/yr, depending on the floodplain topography. Present sedimentation rates were measured using sediment traps. Sediment accumulation decreases exponentially with distance from the river channel; relatively large amounts of sediment are deposited in depressions. The effects of the IPCC BaU climate scenario on floodplain sedimentation were evaluated in detail for a small floodplain section, and for the entire Waal floodplain. The results indicate that a climate change alone will increase floodplain sedimentation by at least 50%; the yearly average sediment load transported over the entire Waal floodplain becomes more than three times as large as at present.
1. INTRODUCTION The increased emission of CO2 and other greenhouse gases into the lower atmosphere is expected to lead to a warming of the earth's surface. This greenhouse effect may cause a world-wide climate change in the forthcoming decades, resulting in a.o. changes in temperature, precipitation and evaporation. A climate change will affect the river Rhine discharge regime and the suspended sediment load transported into the Netherlands. As a result, the sedimentation rates on the embanked floodplains in the Netherlands may change. For the assessment of the impact of climate change the IPCC scenario BaU-best (Kwadijk, 1993) is used. Kwadijk (1993) investigated the expected changes in the discharge of the river Rhine and its tributaries using the RHINEFLOW model. Asselman (1994) described the relation between discharge and suspended sediment load by a sediment rating curve. Changes in sediment production may change the present Sediment rating curves. Asselman (1994) defined four sediment rating scenarios associated with different combinations of the BaU climate scenario and land use changes, resulting in different sediment rating curves for the Rhine at the Dutch-German border. In this project the sedimentation on the embanked floodplains is investigated. Past and present sedimentation rates are reconstructed using various methods, and the possible effects of climate change on future floodplain sedimentation are investigated (Middelkoop, 1994). The objectives of this study were to assess: - the sedimentation rate on the embanked floodplains in the Netherlands in relation to flooding-frequencies during the past decennia, and centuries,
932 the spatial variability of floodplain sedimentation, quantitative relationships between sedimentation rates and (1) floodplain morphology and (2) the characteristics of flood periods, and - possible effects of climate change on future floodplain sedimentation. -
-
The sedimentation on the embanked floodplains was studied from different points of view and on various temporal and spatial scales. The investigated floodplain sections are shown in figure 1. The effects of the BaU climate scenario were investigated for a small floodplain section using the WAQUA-DELWAQ model, and for the entire Waal floodplain using the average sediment discharge as an estimator for potential sedimentation.
0
100km
... ~
~" ~
Z voile E
":~:~:~:~:~:~:~:::
~:
N
~> ,
.E "1o
X
X
X
X
X
:
::::i~:::::: ........ ::::::::::::::::::::::
::iii.i:...
.=
' ~.-
:i:~i:iJ:i::~
~..
X Arnhem
Variksche Plaat
9
.
J~::.-|~
o_
Lobith
....... ' ~ I ~ ~ r
~ ~
~
~ ~
-
f "
~se .... K e e n t /
o .....!..
Nijmdt]en
..~_~ --
X
.:::, o
~o ~m
~''";~':=
x = method applied at floodplain section [ ij~iii~i~i~i~!~i] floodplains
Figure 1. Study area and location of investigated floodplain sections. 2. FLOODPLAIN
GEOMORPHOLOGY
AND
GENESIS
The genesis and geomorphology of the embanked floodplains were investigated by geomorphologic mapping and corings, and by analysis of old river maps and historic records of water levels. Old river maps provide a rough indication of the beginning of sedimentation on the enclosed floodplains. The present embanked floodplains along the rivers Rhine and Meuse consist of point bars of meandering rivers as well as lateral bars separated by side channels. Many lateral bars were formed only during the last two to three centuries. Lateral bars are mainly found along the river Waal. The development of lateral bars is related to land reclamation from the river bed. Using old maps, a succession scheme was developed showing the stages of development of lateral bars, side channels and vegetation (Middelkoop et al., 1992). Assuming that the yearly sedimentation rate is proportional to the product of inundation time and suspended sediment concentration it was found that at the beginning of the floodplain formation the sedimentation rates were 3 to 4 times as great as at present, and varied between 1 and 3 cngyear.
933 3. R E C O N S T R U C T I O N OF F L O O D P L A I N S E D I M E N T A T I O N RATES USING HEAVY METALS AS A T R A C E R The river Rhine sediments that are deposited on the embanked floodplains in the Netherlands are contaminated with pollutants, including heavy metals. The heavy metals in the floodplain soil profiles were used as a tracer to calculate floodplain sedimentation rates. Sediment accumulated in dike-breach ponds was analysed to reconstruct changes in heavy metal pollution of the river Rhine sediment during the past 200 - 300 years. From historic dates, Pb-210 dates, palynological information and variations in sediment compaction, a timedepth control of the sediment fill of the dike-breach ponds was obtained. The reconstructed changes in the heavy metal pollution of the Rhine sediment are shown in figure 2. The heavy metal pollution increased during the first half of this century; maximum pollution occurred around 1960; since 1970 the heavy metal pollution has strongly decreased. From various sections of the embanked floodplains, samples from vertical soil profiles were collected and their heavy metal content was measured. The samples were taken from sites with different inundation frequencies, local elevations, and distances to the main channel. The heavy metal content of floodplain soils was related to these factors. Average sedimentation rates during the past decades were reconstructed by comparing the heavy metal profiles in the floodplain soils with the pollution history of the Rhine. The heavy metal profiles obtained from the floodplain soils generally have the same shape as the pollution curve reconstructed from the dike-breach pond sediment. Depending on the sedimentation rate, the vertical profile is more or less stretched, and the total heavy metal content varies (figure 3). The total soil pollution can be very different from the pollution in the upper 10 cm. The sedimentation rates for the past decennia that were estimated from the heavy metal profiles range between 2 and > 15 mm/yr.
(Jrl 0 0
0 0 0
._L 01 0 0
I~ 0 0 0
tO 0 0
0 1980 1
1960 I
1940 2
...... ,~-;~- ....... ..i-:........... J,........... 4............ 1920 1900
3 average sedimentation rates (mm/yr):
,50 4 o "13 :3"
2-4 -
15-18
s
~s
v3
o 0
o 0
o 0
200
9
i
Cd, Cu, Pb (mg /kg )
Figure 2. Heavy metal pollution history of the river Rhine reconstructed from a dike-breach pond along the river Waal.
Figure 3. Zinc profiles of floodplain soils as a result of different sedimentation rates.
934
4. PRESENT FLOODPLAIN SEDIMENTATION RATES Present floodplain sedimentation rates and their spatial variability were measured after the two floods of 1993 and one flood in 1994 using a total of about 800 sediment traps of artificial grass. Measurements from the individual traps were interpolated using block-kriging to create raster maps of sediment accumulation (figure 4). The patterns shown on the maps were correlated with floodplain morphology and sedimentation mechanisms (Asselman & Middelkoop, in press.).
0,
....
>1.0 kg/m 2 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 500 m
....~:~;!!!!!!i!i;ii~.:.. ===================================== 9 . ~ .-.
>9.0 k g / m 2 8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0
sand sheet
B. K e e n t - D e c e m b e r 1993.
A. K e e n t - J a n u a r y 1 9 9 3 .
5.5 k g / m 2
2.4 kg/m 2 1.8 9~ i
'iiii!ili%~.iiiiiiiiii!iiii!:' ....
4.5 5.0 4.0 3.5 3.0
i .......-:i:i:i" "
1.5
1.2 0.9 0.6 0.3 0.0
i
C. VarikschePlaat- January1993.
2.5
:.:.!
, iiiiiil}iiP"i iiiiiiiIiiii} 500 m
:.
..-.-.:,-.:.:.:.-.'..-,:.-. D. V a r i k s c h e P l a a t - D e c e m b e r
1993.
Figure 4. Sediment accumulation after a single flood, measured by sediment traps. January 1993: minor flood; December 1993: extreme flood.
During the flood, sand is locally deposited directly behind a levee. The amount of deposited (fine) sand decreases exponentially with distance from the levee. This pattern agrees well with existing diffusion models. At distances larger than 50 to 100 m from the river channel the accumulated sediment consists mainly of material < 53~. Here, differences in sediment accumulation are determined by the local topography, which causes differences in inundation times and ponding (figure 4c). The total amounts of suspended sediment < 531.t deposited in the investigated floodplain sections during the high flood of December 1993 ranged between 1.20 and 3.98 kg/m 2 along the river Waal, and between 1.0 and 2.0 kg/m 2 in the study areas along the river Meuse. Equivalent thicknesses of sediment accumulation during the flood ranged between 0.8 mm and 3.2 mm in the central parts of the floodplain sections. The estimated total amount of suspended sediment deposited during the flood on the entire river Waal floodplain was 0.24 Mton. This is about 19% of the total suspended sediment load transported through the river Waal in the same period. It is 7.7% of the average yearly load of suspended sediment transported by the river Rhine into the Netherlands. Comparison of sediment accumulation of different floods shows that the amount of sediment deposited on a floodplain increases less than proportionally with the flood magnitude.
935 5. I M P A C T OF C L I M A T E CHANGE ON FLOODPLAIN SEDIMENTATION RATES 5.1 L o c a l s e d i m e n t a t i o n rates
Sedimentation on a small floodplain section was simulated using the 2-dimensional WAQUA-DEWAQ model (Ubels, 1986). For a series of stationary discharge stages, the average yearly sediment accumulation was calculated by multiplying the sedimentation rates by the associated frequency of occurrence and suspended sediment concentration. The total yearly sedimentation is the summed total of the products. Sedimentation rates under changed climate conditions were calculated by using the BaU flow duration curve (Kwadijk & Middelkoop, 1994) and using the sediment rating curves provided by Asselman (1994). The WAQUA-DELWAQ model demonstrates that sedimentation becomes very ineffective at high discharges (figure 5). Preliminary results indicate that under the BaU climate scenario the sedimentation rate in the test area increases by only 3% if both climate and land use change. A climate change alone, however, increases the sedimentation rates by about 50%. Sedimentation rates will increase much more on floodplain sections behind a minor dike and where sedimentation only occurs during relatively high discharges.
tonh
kg/m2/yr
reference scenario 8.00
i i [ [ i ........i - - - - ~ ..................................
present auton, land use change only
6.00
4.00
climate scenario =
BaU-climate
.t
BaU-climate + land use
2.00
0.00 3500
4500
5500
6500
7500
.
o o
8500
discharge (m3/s)
Figure 5. Effective sedimentation on a low floodplain section for different discharges.
.
o o
.
o o
.
.o o
.
o o
o o
~
~
o o o
o o o
Rhine discharge (m3/s)
Figure 6. Sediment load transported over the embanked floodplains for different scenarios.
5.2 Sensitivity of large scale potential s e d i m e n t a t i o n rates
The potential sensitivity of the sedimentation rate on the entire embanked Waal floodplain for a climate change was investigated from changes in the annual amount of river sediment load that is transported over the embanked floodplain. The 1-dimensional SOBEK model (Huyskens & Barneveld, 1994) was used to calculate for a series of discharge stages the percentage of the river discharge transported over the embanked Waal floodplain. At each stage, the sediment load transported over the floodplain was calculated from the product of (1) the discharge, (2) the suspended sediment concentration and (3) the relative frequency of occurrence. The summed totals of the product for all discharges gives the average yearly load,
936 expressed in Mton/yr. This is the amount of sediment is potentially available for deposition. The effect of the BaU discharge scenario and four sediment rating scenarios on this sediment load was calculated to investigate the possible impact on floodplain sedimentation. The BaU climate scenario will increase the average sediment load transported across the floodplains by a factor 3.5 compared to the present situation and by a factor 3.2 compared to the situation with autonomous land use changes (figure 6). The changes in effective sediment deposition will be smaller than the changes in the sediment load. 6. CONCLUSIONS The present floodplain geomorphology comprises both pointbars of meandering rivers and lateral bars that have been formed during the past centuries. Floodplain sedimentation depends on floodplain characteristics, discharge frequency distributions and sediment concentrations, and therefore shows a high variability in time and space. Present floodplain sedimentation rates range between 0.5 and 15 mm per year. The effect of the BaU climate change on floodplain sedimentation are considerable. High discharges will occur more frequently and sediment concentrations are expected to increase as a result of climate change. Under the BaU climate scenario, the yearly amount of sediment transported over the embanked Waal floodplain is more than three times as large as under present climate conditions. The increase of the effective sedimentation rates will be smaller, depending on the type of floodplain section. Sedimentation rates are expected to increase at least by about 50% on floodplain sections directly bordering the main channel and without a summer dike. The sedimentation rate on floodplains that are situated behind a summer dike is expected to increase more than 50%. If also the effect of changes in land use are taking into account, changes for low lying floodplains are insignificant. Nevertheless, the average sediment load transported over the entire floodplain increases still by a factor 2.8.
REFERENCES
Asselman, N.E.M., 1994: The impact of climate change on suspended sediment transport in the river Rhine. Department of Physical Geography, Universiteit Utrecht. Asselman, N.E.M. & H. Middelkoop, in press.: Floodplain sedimentation: Quantities, patterns and processes. Accepted for publication in: Earth Surface Processes and Landforms. Huyskens, R.B.H. & H.J. Barneveld, 1994: Sobek-model voor de Nederlandse Rijntakken. Rapport Q 1895, WL. Kwadijk, J.C.J., 1993: The impact of climate change on the discharge of the River Rhine, Thesis. Universiteit Utrecht. Kwadijk, J.C.J. & H. Middelkoop, 1994: Estimation of the impact of climate change on the peak discharge probability of the River Rhine. Climatic Change, Vol.27, issue 2, pp. 199-224. Middelkoop, H., N.J. van den Berg, E.L.J.H. Faessen & H.J.A. Berendsen, 1992: Morfodynamiek van nevengeulen van de Waal: een historisch overzicht. Vakgroep Fysische Geografie, Universiteit Utrecht, Rapport GEOPRO 1992.07. Middelkoop, H., 1994: The impact of climate change on the sedimentation rates on the embanked floodplains in the Netherlands. Report of the NRP-project NOLK/002/90. Department of Physical Geography, Universiteit Utrecht. Ubels, J.W., 1986: Verantwoording van het hoogwateronderzoek op de Boven-Rijn, de Waal, het Pannerdensch Kanaal, de Neder-Rijn, de Lek en de IJssel. Nota nr. 86.035. DBW/RIZA, Arnhem.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
937
The impact of climate change on suspended sediment transport in the river Rhine Nathalie E.M. Asselman Department of Physical Geography, Utrecht University, PO Box 80.115, 3508 TC Utrecht, The Netherlands
Abstract Erosion, transport and deposition of fine suspended sediments are both directly and indirectly influenced by climate conditions. In this study, the suspended sediment transport regime of the river Rhine under present and future climate conditions was assessed. The impact of climate change on the sediment transport regime was investigated using sediment rating curves in combination with flow duration curves, developed using the BaU-climate scenario, and three sediment production scenarios. The results indicate that a climate change under changed land use conditions will result in a 14% increase in total annual suspended sediment load. A larger part of the yearly sediment load will be transported at discharges over 4000 m3/s. This probably results in increased floodplain sedimentation rates.
1. INTRODUCTION The increased emission of C O 2 and other greenhouse gases during the 20th century, is expected to enhance the greenhouse warming of the lower atmosphere. This may cause a world-wide climate change in the forthcoming decades, resulting in changes in temperature and precipitation. The climate induced changes in vegetation cover and water discharge in turn will affect erosion, transport and deposition of suspended sediments in the river Rhine. Within the scope of the National Research Program (NRP 1) the impact of climate change on discharge, and the suspended sediment transport regime of the river Rhine was studied. The IPCC "Business as Usual" (BaU) scenario projected on the Rhine catchment, was used as climate change scenario (Kwadijk, 1993). The aim of this study is to investigate the processes of sediment transport through the river Rhine under actual climate conditions, and to assess the effect of climate change on the suspended sediment transport regime, depending on changes in discharge and sediment supply to the rivers.
938 2. M E T H O D S
Since wash load is a non capacity load, the amount of fine suspended sediment transported by the river Rhine depends on the availability of loose material and to a lesser extent on the capability of the river to transport this material. Therefore, sediment transport rates cannot be calculated using stream power related transport formulas. Instead, the so-called rating curve technique can be used. A sediment rating curve describes the average relation between discharge and suspended sediment concentration. The most commonly used relationship between discharge and suspended sediment concentration is a rating curve in the form of a power function (a.o. Walling, 1974). In this study a power function with additive constant term was used: c = /9 + a . Q b where c is suspended sediment concentration (mg/1), Q is river discharge (m3/s) and a, b and p are regression coefficients. Sediment rating curves were developed for five gauging stations along the river Rhine (figure 1), using daily discharge and suspended sediment concentration data, measured by the Bundesanstalt fir Gewisserkunde (BfG), Germany.
iiiiiii iiii!iiii \
o
,00k=
ERMANY ~ / /
t...i,
o
..
N .
ANDERNACH-WEISSEI~THURM~ ~ k~.,/..~-"'--" ,J 2'/"
KAuB
~
....I,..'
"t. <. III
,.~,,.~.,,'~
....
",-..,,,,,. .........-.,t',~.
777 J
~.. L|
7/
~.., ........
Figure 1. Location of the gauging stations.
""
,
"'"-/)' \ "',,~n
i
< ~ , < ~ '-7/
.,
~'~.'\.,.~
"
939 The amount of sediment transported at a certain discharge is given by the sediment rating curve. The frequency of occurrence of a certain discharge is given by the discharge frequency distribution. Both curves were combined to obtain the sediment discharge curve. This curve shows at which discharges most suspended sediment transport takes place. Changes in the sediment discharge regime were studied using the flow duration curve and sediment rating curves, developed in accordance with the BaU climate scenario. Monthly scenario discharges were calculated by Kwadijk (1993). The monthly average discharge values were converted into daily discharges using statistical methods based on the method described by Kwadijk and Middelkoop (1994). Changes in suspended sediment production by soil erosion in the Rhine basin under BaU climate and land use conditions were assessed using the Universal Soil Loss Equation (USLE) (Wischmeier and Smith, 1978). Three sediment production scenarios were used, each resulting in a different sediment rating curve. These sediment rating curves were combined with the BaU flow duration curve to obtain sediment discharge curves. The sediment discharge curves show the changes that are expected to occur in the sediment transport regime of the Rhine, when climate changes in accordance with the BaU scenario. For Rees, near the Dutch-German border (figure 1), the following sediment transport scenarios were used: 1) Sediment loads are determined by hydraulic properties of the river, as shown by the sediment rating curve developed for present climate conditions. This rating curve remains valid under changed climate conditions. Changes in sediment production in upstream areas have no direct effect on the sediment load near Rees. 2) Sediment loads are determined by the erosion rates in upstream parts of the river basin under BaU climate, and present land use conditions. 3) Sediment loads are determined by erosion rates in upstream parts of the river basin, as expected for the BaU climate and land use scenario. 4) Sediment loads are determined by changes in sediment production due to autonomous changes in land use, no climate change is taken into account. This scenario is used as a reference scenario to evaluate the effect of climate change under changed land use conditions.
3. RESULTS The sediment rating curves developed for the selected gauging stations are shown in figure 2. The sediment rating curves can be used to obtain information on the availability of sediment in a certain area in combination with the erosive power of the river itself. Steep rating curves, as observed near Rheinfelden, are characteristic for river sections with little sediment transport taking place at low discharge. An increase in discharge results in a large increment of suspended sediment concentrations, indicating that the power of the river to erode material during high discharge periods is great, or that important sediment sources become available when the water level rises. Flat rating curves, as found near Rees, are characteristic for river sections with intensively weathered materials or loose sedimentary deposits, which can be transported at all discharges. The constant p-coefficient can be seen as a background concentration, a minimum concentration of suspended sediment occurring at very low discharges. Since the
940
steepness of the rating curve decreases in a downstream direction, it can be concluded that (1) the importance of high discharge events in transporting suspended sediment decreases in a downstream direction, and that (2) near the Dutch-German border large quantities of fine material are available for transport at all discharges.
250
............ Rheinf. (km 145) ...... Maxau (km 362) - - - Bach./Kaub (km 545) -- -- Weiss./And. (km 607) Rees/Emm. (km 837)
/
%
200
-
...
.,I
i
-
= 150
,4 4"
r
o ,~..~
"~ 100
.." .."
r
. O
," ~
,"
sj
s"/
~.f
50
0
2000
Figure 2.
4000 6000 Discharge (m3/s)
8000
10000
Sediment rating curves developed for several gauging stations along the river Rhine, under actual climate conditions.
C (mg/1) & f(Q) (d/yr) 160 C
*s S . S
120
""A
T (Mt/yr) Suspended sediment transport (Mt/yr) - 0.5 0.5 present scen. 1 ........ - 0.4 0.4 f~ scen. 2 -- __.. ,,~ scen. 3 - scen4
80 -
0.2 0.2
-
0.1 0.1
4O i
0 mo
,
,
,
i
|
2000 4000 6000 8000 1000012000 Discharge (m3/s)
Figure 3.
0
/i,;
-
-2-
"
....."\
".....~
"'.
~ :""'~ " -- ......
.~--~.~..
0 '~,'r . . . . . -. . . . 0 2000 4000 6000 8000 10000 12()00 B. Discharge (m3/s)
Sediment discharge curves developed for Rees under present and BaU climate conditions. a) present climate conditions (n: suspended sediment concentrations (mg/1); B: frequency of occurrence of daily discharges (d/yr); c: suspended sediment transport (Mt/yr)). b) according to several sediment transport scenarios.
941 Figure 3a shows the sediment rating curve (A), the discharge frequency distribution (B), and the sediment discharge curve (C), developed for Rees under present climate conditions. The sediment discharge curve shows that under present climate conditions, most suspended sediment is transported at moderate discharges, with a high frequency of occurrence. The sediment discharge curves developed for the different sediment transport scenarios are shown in figure 3b. The results are summarized in table 1. According to all scenarios, the importance of high discharge events in transporting suspended sediment will increase. This is mainly related to the expected changes in the discharge frequency distribution. The total annual sediment loads are different for each scenario. The impact of climate change on sediment transport rates under future land use conditions can best be assessed by comparing scenarios 3 and 4. This comparison shows that a climate change as assumed in the BaU climate scenario, will result in a 14% increase in the total annual suspended sediment load near Rees. The suspended sediment load transported at discharges over 4000 m3/s, when several floodplains along the Dutch part of the river Rhine are inundated, will increase by about 47% from 0.66 to 0.97 Mt/yr.
Table 1 Sediment transport rates using four sediment transport scenarios Scenario
present la lb 2 3 4
p
29 29 27 38 15 14
Total load
Q > 4000
Q > 6000
%
Mt/yr
%
Mt/yr
%
Mt/yr
100 106 100 125 72 64
3.04 3.21 3.04 3.81 2.20 1.93
28 38 38 36 44 34
0.85 1.22 1.16 1.37 0.97 0.66
12 17 18 16 21 16
0.36 0.55 0.55 0.61 0.46 0.31
p = background concentration (mg/1) total load = annual suspended sediment load in % of present transport and Mt/yr Q > 4000 = percentage of average annual sediment load transported at Q > 4000 m3/s Q > 6000 = percentage of average annual sediment load transported at Q > 6000 m3/s Steepness of the rating curve is kept constant with a = 1.96"10 -6 and b = 1.93
4. CONCLUSIONS When climate changes in accordance with the BaU scenario, the annual suspended sediment load near Rees is expected to decrease by about 28% compared to the present load. However, the impact of climate change on the suspended sediment loads is best assessed by comparison of scenarios 3 and 4. This comparison shows that a climate change in addition to autonomous changes in land use will result in a 14% higher annual suspended sediment load near Rees.
942 According to all sediment transport scenarios a larger part of the yearly suspended sediment load will be transported at discharges over 4000 m3/s, when inundation of low lying floodplains occurs. Under present climate conditions about 28% of the yearly sediment load is transported at discharges over 4000 ma/s, this will be over 40% when land use and climate change in accordance with the BaU scenario.
5. REFERENCES Asselman, N.E.M. (1994): The impact of climate change on suspended sediment transport in the river Rhine. Dept. of Physical Geography, Utrecht University. Kwadijk, J.C.J. (1993): The impact of climate change on the discharge of the River Rhine. Thesis, Utrecht University, 201 pp. Kwadijk, J.C.J. and H. Middelkoop (1994): Estimation of the impact of climate change on the peak discharge probability of the River Rhine. Climatic change, Vol. 27, 2, 199-224. Walling, D.E. (1974): Suspended sediment and solute yields from a small catchment prior to urbanization. In: K.J. Gregory and D.E. Walling (Ed.), Fluvial processes in instrumented watersheds, Institute of British Geographers special publication no. 6, London. Wischmeier, W.H. and D.D. Smith (1978): Predicting rainfall erosion losses - a guide to conservation planning. US Dept. of Agriculture, Agriculture Handbook No. 537.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
943
Effect of sea-level rise and climate change on groundwater salinity and agro-hydrology in a low coastal region of the Netherlands L.C.P.M. Stuyt, P. Kabat, J. Postma and A.B. Pomper DLO Winand Staring Center for Integrated Land, Soil and Water Research, P.O. Box 125, 6700 AC Wageningen, The Netherlands.
Abstract Scenario studies to predict the effects of doubled CO 2 levels, a 1 ~ temperature increase and a 1.2 m sea-level rise on seepage, groundwater and crop production were carded out. Climatic change was simulated, showing increased precipitation. Simulation of effects of sea-level rise on groundwater flow and salt transport showed changes in seepage to be negligible. Simulated crop growth was increased significantly by temperature and CO 2 increase, without increased demand for irrigation.
1. INTRODUCTION Climate change will affect open water and groundwater levels and quality in coastal lowlying regions through sea-level rise and precipitation, and through changes in crop wateruse, enhanced by doubled CO 2 . These changes may cause changes in water management. Climate change was simulated using statistical weather relationships from the KNMI (3). Sea-level rise was simulated with the 2-D transient density-driven model MOC (4,5). Soil water dynamics and crop production were simulated using SWAP (2, 6).
2. METHODS 2.1. Climate Scenarios A climate scenario 1 ~ warmer then present was created with the climatic relationships from the KNMI, using data from the years 1966, 1976, 1979, 1985 and 1986, and called scenario(W). Radiation, humidity and wind and the pattern of rainfall are assumed to remain unchanged, CO 2 levels doubled. The present climate is scenario(P).
2.2. Groundwater Flow Modelling with MOC Voorne-Putten, in the SW of the Netherlands, was selected for modelling as it lies at the coast, below sea level, with saline seepage. Dunes form the Western end. Groundwater flow was simulated with MOC 3.0 of 1989, a transient solute transport model. To suit the model for vertical application, density driven groundwater flow was added. The grid used is 100 by 20, west-to-east, 200 m deep and 25 km long. transversal to longitudinal conductivity and dispersivity ratios were set to 0.1. The boundary conditions imposed were:
944 -
The vertical boundaries have constant piezometric levels. The upper boundary has constant phreatic levels in the polders, in the dunes there is a constant groundwater recharge (180 mm.yr-1). The geometry of the cross-section is shown below. -
/// dunes / f ~
dikes I ~ l ~ r s (0.0 to - 2 . 0 m) below ~ea level Dumkerke F (multtple ks)
Northsea
}
KrefteN~ye F. (20 m/d)
"
~/////////////////////////////~ Kedichem F. (0.002 m/d)
Spui
8
8
to
i~lll
1
rr~ine Me~ssluis F. 110 m/d)
AI[NIIIIII
MOC
(200--x- 10 rrO
25
<
SWAP profiles for e r ~
E
km
......... >
I~eX:lu~tlon
~".~/z~
aqud:ard
boundary . . . . . . . . . prescribed pressure and salt b o ~ r t e s (MOC) . . . . . . . hyclrobia layer (0.01 m/d} marine Maassluts F, (10 m/d): name of forrnatton (wtth 10ermeablltty)
Fig. 1 The sub-soil of Voorne-Putten partitioned into aquifers and aquitards. Calibration was done for seepage with measured rates (7) by varying ksAT.
2.3. Calculating Crop Production with S W A P
Voorne-Putten was divided into 761 subareas, 461 of which are cultivated. Soil physical properties, open water levels, hydrological properties and land-use were established for each. Different crop-varieties chosen for scenario(W) to give a realistic yield, by adapting some of the physiological parameters of the crop models. An example is Table 1 for wheat. Table 1. Plant physiological parameters for scenario(W) in winter wheat (8,1). spec. light-use eft. max. assim pre-anthesis temp. sum surface leaf area (kg.ha-l.h -1 rate temp. sum until maturity resisi)(s (m2.kg-1) /J.m-2.s -1) (kg.ha-l.h -1) (~ (~ mScen.(P) Scen.(W)
18.0 14.4
0.45 0.55
40 80
1048 1290
1258 1171
40 44
Production of important crops was calculated, with scenario(P), calibrated with estimated
945
actual harvests and then calculated again with scenario(W).
3. RESULTS OF SIMULATIONS The effect of sea-level rise on seepage was assessed by comparing the results of two simulations. Both ran for 100 years, one with and one without sea-level rise. Figure 2 shows the resulting differences in groundwater salinity (mg/1). Seepage intensity increases, but is unimportant compared to the fresh water let in to maintain water quality. On crop growth, climatic change has opposing effects: higher respiration, a longer growing season and higher water-use efficiency. The net result is increased crop production, as an example shows in Table 2. Table 2. Average crop production (tons (dm) ha-l), for 1979, for scenario's (P)&(W), calculated with SWAP. sugarbeets potatoes scen.(P) scen.(W)
12.4 14.7
13.0 1.4.7
wheat
grass
6.8 10.1
9.8 12.0
REFERENCES
1. Boons-Prins, E.R., G.H.J. de Koning, C.A. van Diepen and F.W.T. Penning de Vries. 1993. Crop specific simulation parameters for yield forecasting across the European Community. Simulation reports CABO-TT, no. 32, CABO-DLO. Wageningen. 2. Feddes, R.A. and P. Kabat, (eds.) 1994. SWAP: a model to simulate the Soil_Water_Atmosphere_Plant interactions. Part I: Theory and model description. Simulation Monograph, Pudoc, Wageningen. (in prep). 3. Klein Tank, A.M.G. and T.A. Buishand. 1993. Modelling daily precipitation as a function of temperature for climate change impact studies. Scient. Rep. WR 93-02, KNMI. De Bilt. 4. Konikow, L.F. and J.D. Bredehoeft. 1978. Computer model of two-dimensional solute transport and dispersion in ground water. U.S. Geol. Surv. Techn. of Water Resour. Investigat., Book 7, Ch C2, 90 pp. 5. Oude Essink, G.H.P. 1993. Effect of Sea Level Rise on the Groundwater Flow System through Amsterdam Waterworks and Haarlemmemleer polder, The Netherlands. Proc. UNESCO Conf. on Sea Level Changes and their Consequences for Hydrology and Water Management, Noordwijkerhout, The Netherlands. 6. Supit, I., Hooijer, A.A., van Diepen, C.A. (Eds.), 1994. System description of the WOFOST 6.0 crop simulation model. SC-DLO, Wageningen (in prep). 7. Wit, K., 1987. Wateraanvoerbehoefte Zuidhollandse Eilanden en Waarden (Fresh water requirements of the south-western island and polders of Zuid Holland). I.C.W. Nota nr. 1801. SC-DLO, Wageningen. 8. Wolf, J., and Diepen, van, C.A. Effects of climate change on crop production and land use in the Rhine basin. In: Geijn, S.C., Goudriaan, J. and Berendse, F., editors. Climate change; crops and terrestrial ecosystems. Agrobiologische Thema's 9. 1993, CABO-DLO, Wageningen.
946
9
d "t:
9
0 ~
i:
o~.~
0
0 0
0 b~
0
~
.~-~
o~..~
6. ~q ~
~,~~,~,~ ,~,~,~,~
.....
...
.
",
..,
-..
9
.,
'
9
.....
,
.,,,
",
,
"
,,
"...
. -...
.
.
9 .
'
9
9
.
.
..
9
.
9
.
,..
-...
.
.-.
.
.
9
.
..
......
.
..
i.. . .
.
.
...
"...,..
.
.'.-I
. . .
i
9
.
.
i
..
.
,...,
.
..
....
.
i
9
.
.
.
,
- . . . "
,
,
1!!iiiii!ii!'
. /
.
~
I<
d
~i~ii~i~l!,iii ~I ~
:.:i-::.: :-i:.::-: ...:-::.: :.:::i ,:ij:::.:i::
'~:~', .... ~ ......... I
::::::::: -:::::I "
.......... ~.......... : .
'
i
~...... i).~ : ! : : . . . " . I " " . " ,
0
9
.
"
i
I , , , . ,
,
".
i
~!::::~i, :. i0,
~
~!:.
.:..
~ii
I ".
::!i;
{:. ':
.
"........ " 9
9
......
'
~,,~'!!,,,~,~i,~'~!i,~~ i li~i ! !!i!! t~
i!:..
...
:i::?
1.......,-::....-
:
"7..:::
'.
~ ~ .... :-""
0
o
~.~0i-~ ,~-i~-~:~.::~,., .!::iiii!ii:.i:.-. .
(!)
(I0-~
~
c~
i -~~
.7X ~ -
::::"'::::%L~,
lliiii-i<......
9
9
i:ii:i.)i:.. 2... i:i:}::}: "
"" '
....":.....': . . . 9
~~
d
0
d
0
d
r
~ ~:=
d
9
0
d
<)
.............ii')~'~,),~il,i ~ ~ ~ I~:~~ i ~~~~ ~: ~:I~ ~~ =
d
r~==
@ ~=
0
~ i=~ x~
d
Q 11)
d
~ ~
d
0 ~
d <~I
6:)
UI)
d
~@IiI U ~ |
~#l|
<S)
I[
~-~
~=~
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
947
Impacts of changes in climate and socio-economic factors on land use in the Rhine basin" projections for the decade 2040-49 R.P. R6tter, F.R. Veeneklaas and C.A. van Diepen DLO, Winand Staring Centre, P.O. Box 125, 6700 AC Wageningen, NL
Abstract
The purpose of this study was to develop land use projections for the middle of the next century. To separate the influence of climate change from other factors on land use, projections (a Central Projection and two variants) were made under both unchanged and changed climate. They cover the plausible range of alternative land claims of agriculture and urbanization. Impacts of climate change on land suitability as well as overall changes in the acreages of several land use categories are presented.
1. BACKGROUND
AND
OBJECTIVES
Expected climate change might have a great impact on the discharge pattern of the river Rhine with consequences for agriculture, fresh-water supply, shipping, etc [1]. In addition to the direct effects of changes in temperature and precipitation, changes in land use may also affect the discharge regime through, for instance, changes in water extraction patterns over the season and shifts in the area of arable land. The current study is a contribution to the Climate Project of the International Commission for the Hydrology of the Rhine (CHR). At the request of of the Institute for Inland Water Management and Waste Water Treatment (RIZA), land use projections for the middle of the next century based on biophysical and socio-economic analyses were developed. Several projections were constructed for both unchanged and changed climate. Results will be used by hydrologists examining the possible consequences of changes in climate and land use for the discharge regime of the river Rhine. This paper presents some of the results from the Rhine basin study on land use projections, prepared by the DLO Winand Staring Centre [2-5]. The specific objectives of the Rhine basin study were: - to subdivide the Rhine basin into biophysically differentiated land types (current and future climate/soil/terrain combinations); - to calculate production potentials and associated water use of major crops (and forest types) for the predominant biophysical land types; - to assess actual land use; - to construct different projections of land use for the decade 2040-49: (a) without climate change, i.e. driven by socio-economic factors only, and (b) with climate change, i.e. driven by both climate change and socio-economic factors.
948 2. A P P R O A C H
The impact assessment approach combines results from biophysical and socio-economic analyses at various stages of constructing the projections. 2.1 Land suitability, crop yields and water use First, a biophysical classification system for the Rhine basin was designed, combining a bioclimatic classification adapted for present and possible future climates and a soil/terrain suitability grouping integrated in a geographical information system [2]. Subsequently, potential yields and water use were simulated for both unchanged and changed climate for the predominant biophysical land types and crops [3,4]. Under changed climate we refer to Table 1 and a CO2 concentration of 560 ppmv, based on a Business-as-Usual emission scenario [6], best estimate (BaU-best) as interpreted by [7] for the decade 2040-49.
Table 1 Expected changes in temperature and precipitation applied to the entire Rhine basin
Temperature (~ Precipitation (%)
Spring
Summer
Autumn
Winter
+1.75 0.00
+1.50 0.00
+1.75 0.00
+ 2.00 +10.00
2.2 Construction of land use projections For thirteen administrative regions, actual land use data were obtained from the latest land use statistics [5]. Based on analysis of technical restrictions, historical-trend analysis and basic assumptions regarding technical, political, demographic and political factors, a trend projection was carried out, the Central Projection of land use for both unchanged and changed climate. As much uncertainty occurred with respect to the basic assumptions, two variants to the Central Projection were constructed. The Plus Variant combines the highest claims of urbanization and agriculture, and the Minus Variant represents the lowest claims [5]. In constructing the projections it was assumed, that land use in the long run is demand-induced, and that claims on land of the various demand categories follow an explicit hierarchy: 1. Urban land needs (including nature claims formulated in official policy plans) 2. Agriculture - 2.1 Outdoor horticulture, permanent crops - 2.2 Root crops - 2.3 Cereals and oil-seeds. - 2.4 Grassland and fodder crops 3. Forestry & nature
The rationale behind this 'hierachy' is to be found in the price of land paid by these different land use categories and, for agriculture, in the profitability (per ha) of the various crops and in their land quality requirements.
949 3. RESULTS AND CONCLUSIONS An increase in CO2 concentration to 560 ppmv combined with increased temperature and little change in precipitation (Table 1) does not curb crop yields, but improves the conditions for cultivating crops and tree species presently grown. Under baseline climate some ninety biophysical land types have been distinguished in the Rhine basin. Each of them is characterized by soil and terrain characteristics, three temperature characteristics and precipitation [2]. Shifts in their compositions, locations and acreages are considerable under changed climate. One illustrative example is type C2, L-mc/c4, characterized by soil/ terrain group C2 and bioclimatic type L-mc/c4. In soil group C2, orthic Luvisols and eutric Cambisols are clearly predominant and the average slope of land is less than 25 %. The bioclimatic type is characterized by 9.0-9.9 ~ annual mean temperature (L-), a mean annual temperature amplitude of 16.0-17.9 ~ (mc), January mean temperature of 0.0-1.4 ~ (c) and 600 - 799 mm mean annual precipitation (4). Under baseline climate, this biophysical land type covers 6426 km 2, i.e. about 3% of the total Rhine basin area, mainly found in France-Est, the Neckar region, Hessen and Nordrhein-Westfalen. Under changed climate (BaU-best), its area is reduced to 45 km 2, as soil group C2 located in France-Est, etc. then combines with warmer and less continental climates. In other areas with soil group C2 occurring (e.g. Alpenvorland), future temperature and/or precipitation conditions do not meet the criteria (L-mc/c4). Table 2 illustrates changes in average rainfed potential yields and water use of winter wheat for predominant soil/terrain/climate combinations.
Table 2 Selected biophysical land types (as defined in [2]) and associated simulated yields (rainfed potential yields) and crop evapotranspiration of winter wheat under baseline and changed climate (scenario BaU-best, 2040-2049) Biophysical land type
Grain yield (t/ha DM)
Evapotranspiration (mm)
baseline
changed c.
baseline
changed c.
baseline
changed c.
C1, Umc/d2 C2, L-mc/d2 C2, L-mc/c4 B, L-m/c4 B, Umc/d3 B, Mmc/c3 B , L-mc/c4 B, Mcm/d4 C2, L-m/b4
C1, Mmc/d2 C2, L*mc/c2 C2, L+m/b4 B, L+m/b3 B, Mmc/d3 B, L*mc/b3 B, L+m/b3 B, L*cm/c4 C2, L+m/a4
8.7 8.0 7.0 5.8 7.5 6.2 5.5 5.0 7.7
11.1 10.1 10.0 8.3 10.0 8.7 8.2 9.6 10.6
391 398 386 328 361 343 319 320 368
346 343 341 291 334 309 287 294 320
Table 3 summarizes the overall land use changes in the Rhine basin. It is mainly the acreage of cereals (including oil-seeds) leading to decreases in agricultural area: in the
950 Central Projection (CP) and the Plus Variant areas under these crops will be halved. In the Minus Variant the decrease ranges from 62% (unchanged climate) to 66% (changed climate). Climate change according to the BaU-best scenario adds a surplus (some 0.2 * 106 ha) in comparison with unchanged climate, thanks to higher agricultural yields.
Table 3 Projected changes in urban and agricultural land u s e ( 1 0 6 ha), decade 2040-49, under baseline and changed climate (scenario BaU-best) in the CP and two variants (Source: [5])
agric urban total
Central Proj. basel, ch.cl,
Minus Variant basel, ch.cl,
Plus Variant basel, ch.cl.
-1.57 -1.83 +0.68 +0.68 -0.89 -1.15
-2.67 -0.18 -2.85
-1.26 +1.39 +0.13
-2.84 -0.18 -3.02
-1.52 +1.39 -0.13
The given explorations of future land use are largely "status quo-oriented". No surprising dynamics in climate, economic conditions or consumption patterns, to name a few, are taken into account. Therefore, we may conclude that changes in the order of-0.18 to +1.39 * 10 6 ha for urban land use and -2.84 to -1.26 * 10 6 ha for agricultural land use can be expected, if climate change, biophysical impacts and socio-economic factors develop as described by [2-5]. Projected conversions of agricultural land into built-up area on the one hand and into secondary forest, grassland or a mix of the two on the other would have distinct impacts on the discharge regime of the river Rhine. Most favourable in hydrological terms (reduced run-off) appears to be the Minus Variant. 4. R E F E R E N C E S
1 2
3 4 5 6 7
Parmet, B. (1993): Impact of climate change on the discharge of the Rhine.- Change 15: 1-3. R6tter, R.P. (1994): Rhine basin Study: Land Use Projections based on biophysical and socio-economic analysis. Vol.1, Biophysical classifcation as a general framework.- SCDLO Report 85.1, Wageningen and Lelystad, 104 pp. R6tter, R.P. and C.A. van Diepen (1994): Rhine basin Study. Vol.2, Climate change impact on crop yield potentials and water use.- SC-DLO Report 85.2, 145 pp. Hendriks, C.M.A. (1994): Rhine basin study. Vol.3, Climate change impact on forest yield potentials and water use.- SC-DLO Report 85.3, 40 pp. Veeneklaas, F.R., Van den Berg, L.M., Slothouwer, D. and G.F.P. Ijkelenstam (1994): Vol.4, Land use: Past, present and future.- SC-DLO Report 85.4, 155 pp. Houghton, J.T., Jenkins, G.J. and J.J. Ephraums, eds. (1990): Climate Change: The IPCC Scientific Assessment.- Cambridge University Press, Cambridge, 365 pp. Kwadijk, J. (1993): The impact of climatic change on the discharge of the river Rhine.- PhD Thesis, University of Utrecht, 172 pp.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
951
ASSESSMENT REPORT ON NRP SUBTHEME "EFFECTS
OF INCREASING
UV-B RADIATION"'
J.C. van der Leun University Hospital Utrecht, Institute of Dermatology Heidelberglaan 100 3584 CX Utrecht The Netherlands
With contributions by: L. van Liere, H. van Loveren
A. Buma,W.W.C. Gieskes F.R. de Gruijl E. Magendans, J. Rozema, M. Tosserams
RIVM, National Institute of Public Health and Environment, Bilthoven RUG, University of Groningen RUU, University of Utrecht VUA, Free University of Amsterdam
952 Contents Abstract 0
0
0
0
Introduction 1.1 Why research on UV-B effects? 1.2 Selection of projects I n v e s t i g a t i o n of t h e effects of UV-B on t h e i m m u n o l o g i c a l r e s i s t a n c e to tumours and infections 2.1 Introduction 2.2 Basic interspecies comparisons 2.3 Infection models in rodents 2.4 Immunological changes in relation to uv-induced skin cancer 2.5 Basic mechanisms 2.6 Conclusions 2.7 Scientific cooperation 2.8 Effects of UV-B on the immunological resistance to t u m o u r s and infections I m p a c t o f e n h a n c e d s o l a r UV-B r a d i a t i o n on p l a n t s f r o m t e r r e s t r i a l ecosystems 3.1 Introduction 3.2 Methodology of UV-B experiments in terrestrial ecosystems 3.3 Responses of terrestrial plants to enhanced solar UV-B-radiation 3.4 Ecological effects of ozone depletion; implications for e n v i r o n m e n t a l policy 3.5 National and international cooperation 3.6 Impact of enhanced solar UV-B irradiation on plants from t e r r e s t r i a l ecosystems E f f e c t s o f i n c r e a s e d UV-B r a d i a t i o n on s t r u c t u r e a n d f u n c t i o n i n g o f a l g a l c o m m u n i t i e s in d i f f e r e n t c l i m a t i c z o n e s : r i s k s o f a g l o b a l d e c r e a s e in s t r a t o s p h e r i c o n z o n e 4.1 S u m m a r y 4.2 Introduction and problem definition 4.3 Aim of the project 4.4 Results 4.5 Conclusion and perspectives of results obtained during NRP I 4.6 Cooperation with other research groups within/outside NRP 4.7 Effects of increased UV-B radiation on structure and functioning of algal communities in different climatic zones: risk of a global decrease in stratospheric ozone
5.
Integration
6.
R e l e v a n c e for p o l i c y m a k i n g
7.
Outlook: f u t u r e w o r k
953 8.
Conclusions
9.
References
ABSTRACT Solar UV-B radiation has profound effects on many organisms. Increases in this type of radiation may, therefore, be expected to have impacts. In this NRP-theme, investigations were carried out on effects of increased UV-B radiation on the immune system, plants from terrestrial ecosystems and algal communities. These three topics were selected on the basis of expertise available in The Netherlands; the topics chosen ranked high in several i n t e r n a t i o n a l priority listings. The projects are described, and the results assessed. Each of the projects produced new knowledge on the effects studied. Several results were taken up in the 1994 UNEP-Assessment of Environmental Effects of Ozone Depletion. Reasons for continuing and intensifying this type of research are given. Other reasonings, implying that such research would not be necessary, are weighed and found wanting. Effects of increasing LW-B radiation rank high among the impacts of global atmospheric change on organisms. 1.
INTRODUCTION
1.1 Why r e s e a r c h o n UV-B effects? Changes in the atmosphere become relevant to society only if they have effects. This places attention for effects in a crucial position in a research programme on atmospheric change. The present assessment deals with effects of increased penetration of solar UV-B radiation to the earth's surface, as a consequence of decreasing total-column ozone. UV-B radiation is the most energetic component of the sunlight reaching the surface. Even without change it has profound influences on humans, animals, plants, microorganisms, materials, and on the chemistry of the atmosphere. A possible increase of this type of radiation demands careful attention. 1.2 S e l e c t i o n of p r o j e c t s Effects of sunlight on organisms are studied in the science called photobiology. The Netherlands have a strong tradition in photobiology. The emphasis has been on h u m a n health from the time between the two world wars. Interest in effects on plants and aquatic organisms developed especially in connection with the problem of depletion of the ozone layer.
954 Selection of topics for NRP-I was guided by the expertise available in the country. This occurred in a spontaneous way, because it was in these areas t h a t project proposals were submitted. This led to the three projects listed in Table 1.1. Table 1.1 List of projects in NRP Subtheme "Effects of increasing UV-B radiation" Title
Project leader
Number
Effects of UV-B on the immunological resistance to tumours and infections
H. van Loveren
850017
Impact of enhanced solar UV-B radiation on plants from terrestrial ecosystems
J. Rozema
850022
Effects of increased UV-B radiation on structure L. van Liere and functioning of algal communities in different climatic zones: risk of a global decrease in stratospheric ozone
851054
The Netherlands provide about I percent of the research efforts in the world, and in some areas where it is relatively active a few percent. It was not considered necessary, therefore, t h a t the NRP-projects would provide complete coverage of this worldwide problem. Important areas, such as effects of UV radiation on the eyes, and on materials, were lacking. On the other hand, the topics t h a t were covered belonged without exception to the "key areas of uncertainty" defined by the U n i t e d Nations E n v i r o n m e n t Programme, and the priorities defined by SCOPE, the Scientific Committee on Problems of the Environment, formed by the International Council of Scientific Unions. 2.
I N V E S T I G A T I O N OF T H E E F F E C T S OF UV-B ON T H E I M M U N O L O G I C A L R E S I S T A N C E TO T U M O I g ~ S AND I N F E C T I O N S
H. van Loveren RIVM, N a t i o n a l I n s t i t u t e of Public H e a l t h and E n v i r o n m e n t a l Protection, Laboratory of Pathology, P.O.Box 1, 3720 BA Bilthoven 2.1 I n t r o d u c t i o n Scientific tools to perform investigations on the health effects of increased UV-B exposure are available. From experimental studies it can be concluded t h a t UV-B causes an increased incidence of tumours by at least two separate mechanisms: 1) by genotoxic activity of UV-B and 2) by affecting the immunologically mediated resistance to tumours. In laboratory animals and also in h u m a n s evidence has
955 been provided of UV-B induced damage of the immune system. For this reason it is reasonable to expect t h a t also immunological resistance to infections may be at risk after UV-B-exposure. It is known that resistance against infectious agents t h a t enter the body via the skin, such as Leishmania, and Herpes simplex, is affected by UV-B irradiation. Whether resistance to infectious agents that enter the body via other routes t h a n the skin is affected is as yet not known. Because UV-B can induce systemic immunosuppression, also suppression of i m m u n i t y against non-skin associated infections (diseases), may be suspected. The main goal of this project was to study/analyze the effects of UV-B on the immune system and resistance to tumours and infections. In these studies dependency of UW-B induced alterations of the i m m u n e system on UV-B dose, UV-B wavelength (spectrum) and species (rat, mice, man) were investigated. This information can serve as a basis for estimation of the health risk of UW-B exposed population groups.
2.2 Basic interspecies comparisons Experiments have been performed in which mice and rats are exposed to UV-B using the newly developed and standerized equipment (FS40 UV-B sunbanks). The MED (minimal erythema, or edema, dose) values were determined for mice and rats. A quantitative scoring method for the determination of UV-B effects on the skin was developed. Using this method it was possible to q u a n t i t a t e edema, redness and swelling reactions after UV-B exposure. In addition to these macroscopical studies also histological studies were performed. Pathological effects such as acanthosis, hyperkeratosis, parakeratosis, and inflammation were induced by UV-B irradiation. These pathological parameters were UV-B dose and time dependent. Similar experiments were performed with a Kromayer lamp as an UVB source. This lamp was used for local/acute UV-B irradiation. Results obtained with this equipment are compared with results obtained with the FS40 lamps (total body/semi-chronic). K r o m a y e r lamps are widely used in clinics for UV-exposure of h u m a n s because it is easily to expose h u m a n s with this lamp. Results obtained with the Kromayer lamp in animal species and humans are being compared in order to be better able to extrapolation animal data. Parameters that are being analysed are: acanthosis, parakeratosis, hyperkeratosis, inflammation and macroscopic changes due to UV-B exposure. Based on these initial studies (characterization of primary skin effects due to UVB exposure in m a n and different animal species), the effects of UW-B exposure on several basal specific and non-specific i m m u n e p a r a m e t e r s were tested. Investigations with respect to changes in the number and subtype of i m m u n e competent cells in the skin as well as in lymphoid organs are being carried out. In addition to these descriptive studies also immune function studies are being carried out. Because the mixed lymphocyte reaction (MLR), which is a general parameter for T cell immunity, is often used in studies that deal with UV-B induced changes of the immune system (in rodents and man) we developed the MLR technique at the RIVM. This model is now operable in rats, mice and man. UV-B exposure (in vivo) of mice and r a t s inhibited the MLR significantly, d e p e n d e n t upon UV-B irradiation-dose and irradiation-time. The MLR will be one of the parameters to be used for the comparison of UV-B effects on T cell i m m u n i t y between different species (mice, rats, man). For this purpose MLR experiments with h u m a n cells were carried out. The effect of in vitro UW-B irradiation of rodent and h u m a n blood
956 cells was also studied. UV-B suppressed significantly MLR responsiveness of blood cells in h u m a n s and rodents after in vitro UV-B exposure of lymphocytes. Recently we developed (in collaboration with dr. M. Mommaas, AZL), a method in order to investigate the capacity of skin cells to induce MLR responses of allogenic blood lymphocytes. This assay is called the Mixed Skin Lymphocyte Reaction (MSLR). In vivo, in vitro, as well as in situ UV-B exposure of skin cells from different species including m a n are being carried out. The in vitro irradiation experiments have recently been finished and the comparison of UV-B sensitivity with respect to this i m m u n e p a r a m e t e r for the capacity to induce a lymphocytic reaction was analysed, and accepted for publication in "Photochemistry and Photobiology". Histological parameters, obtained in the same species, will be added to these MLR and MSLR results for a better comparison of UV-B sensitivity/susceptibility, between the different species. Other immune function tests such as n a t u r a l killer cell function and mitogenic responses to several different mitogens were also significantly altered in rats after in vivo UV-B exposure. Mice and h u m a n studies are p l a n n e d with respect to these p a r a m e t e r s . In vitro UV-B exposure of lymphocytes from rodents and man inhibited mitogenic responses significantly. 2.3 I n f e c t i o n m o d e l s in r o d e n t s The effects of UV-B exposure on (immunological) resistance against infectious diseases (bacteria,parasites and viruses) is studied in rats and mice. For this goal several infection models in rodents are available at the RIVM. At this moment a Listeria model (bacteria) and a Trichinella model (parasite) and a Cytomegalo virus infection model are operational. In collaboration with the EPA (Selgrade; Environmental Protection Agency, Health Effects Research Laboratory, section of Immunotoxicology, Research Triangle Parc, NC, USA) an additional virus model is used. U s i n g this collaborative s t u d y the effect of UV-B exposure on the immunological resistance to Influenza infections in the rat will be investigated. Recently experiments were performed in order to test the effect of I.rV-B on the resistance to malaria infections in the rat. This part of the resistance study is done in collaboration with Dr. Luebke from EPA, Research Triangle Park, NC, USA. All infection models that are used are in principle suitable to test immunosuppression, and hence are also suitable to test possible immunosuppression induced by UV-B. If all the available infection models are used almost all the different aspects of immunological resistance against infective agents are investigated.
We d e m o n s t r a t e d t h a t UV-B exposure (suberythemal doses) can inhibit the immunological resistance against Trichinella spiralis infections in rat. Especially UV-B exposure during the second week after Trichinella infection inhibits r e s i s t a n c e a g a i n s t this pathogen indicating t h a t T cell d e p e n d e n t i m m u n e responses are probably altered. In addition, specific IgE titers in serum of UV-B exposed animals are decreased if the animals were i r r a d i a t e d 3 weeks after Trichinella infection Additionally, we demonstrated t h a t suberythemal doses of UV-B inhibited the resistance a g a i n s t a Listeria monocytogenes infection. Especially the specific T cell response to this pathogen was inhibited by UV-B exposure. A manuscript describing these results is now accepted for publication in "Environmental Health Perspectives". As was found for the parasitic and bacterial infection models also resistance to cytomegalovirus was inhibited by UV-B exposure of rats.
957
Therefore, the important new finding is that resistance against several non-skin-associated infectious diseases can be inhibited by UV-B exposure. 2.4 I m m u n o l o g i c a l c h a n g e s in r e l a t i o n to U V - i n d u c e d s k i n c a n c e r Extensive experimental studies on UV-induced skin carcinomas have been carried out with albino hairless mice (SKH:HR1) in order to extract basic quantitative dose-response relationships for modelling of h u m a n skin cancer risk. From other animal models it was known t h a t UV tumours are highly antigenic, i.e., these t u m o u r s are rejected by a strong i m m u n e reaction when t r a n s p l a n t e d into a genetically identical animal. Somehow the UV-exposed animals loose this ability. In the p r e s e n t project lymphoid organs and skin of the hairless mice were i n v e s t i g a t e d in the course of chronic UV-B exposures, leading up to skin carcinomas. I m p o r t a n t shifts in subpopulations of lymphoid cells were found very early in the experiment in the spleens and especially in the lymph nodes of the animals. CD8-positive cells migrated into the superficial layers of skin in the course of the experiment. By challenging the mice after different times and levels of UV exposure it was ascertained how the observed shifts in subpopulations of lymphoid cells correlated to the failure to reject UV turnout cells. This failure sets in long before the occurrence of macroscopically observable, UV-induced primary tumours, it gradually develops aider the initial shifts in subpopulations of lymphoid cells. The relatively rapid suppression of the immunity against the UV carcinomas m a y suggest t h a t this is not a serious rate-limiting step in the genesis of these cancers (and thus difficult to prevent in high risk individuals). Papers on this subject are in p r e p a r a t i o n and will be submitted this year. E a r l y detection of corresponding immunological changes in humans could serve to identify individuals t h a t r u n a high risk for UV-related skin cancers and perhaps also for decreased immunity against certain infections. 2.5 B a s i c m e c h a n i s m s Besides the aforementioned interspecies comparisons of UV sensitivities, a basic u n d e r s t a n d i n g of the mechanisms involved in UV-induced immunosuppression is likely to further a well-directed research on UV-related immunological risks in man. Although fundamental research on mechanisms is carried out by several groups, this research has not yet yielded data t h a t provide good answers on how the results obtained in animals translate to humans. Hence, experiments were carried out to contribute to this animal-to-human extrapolation. We did this by exploring the in vivo murine model for fundamental immune parameters that would directly signify a high sensitivity to UV-suppression, or a state of UV-induced i m m u n e suppression, and that would also be measurable in man.
To capitalize on a well-established assay at the RIVM on the delayed type of hypersensitivity (dth), we studied the effect of [ W irradiation on the sensitization for the e a r l y (initiation) p h a s e of the h y p e r s e n s i t i v i t y r e a c t i o n a g a i n s t picrylchloride applied to the skin. Besides the already-known suppression of the classical dth reaction (swelling at 24 h after the application), we found a suppression by UV radiation of the early phase of dth (swelling at 2 h). Both phases in the dth reaction are mediated through T cells, and we found - contrary to our expectation - t h a t suppression of the late phase did not fully correspond with the suppression of the early phase (the early phase could be made to occur in conjunction with an LW-suppressed late phase dth reaction; see Y. Suntag et al.,
958 1994). An antigen-specific factor (in this case picrylchloride-specific) 'arms' mast cells in the skin for the induction of increased permeability of bloodvessels, causing the edema of the early phase dth reaction. Such a factor should in principle also be m e a s u r a b l e in humans, and might be used to measure a person's UV suppressibility. A very interesting finding stemmed from the combination of the present research on UV immunological effects with a parallel running research project on DNA damage induced by UV radiation. UV-specific DNA damage can be detected through the binding of a monoclonal antibody (H3, supplied by Dr. L. Roza at TNO in Rijswijk) to UV-induced thymine dimers in the DNA. Upon UV irradiation the DNA damage is normally confined to the most superficial layers of skin (the damaging UV radiation does not penetrate any deeper), but UV-damaged DNA was also detected in some cells the draining lymph nodes, already 1 hour after I_W exposure. By a cumbersome combination of techniques, it was ultimately established t h a t these UV-damaged cells from the skin were in fact antigen-presenting cells (APCs). UV-damaged APCs are known to be (partially) dysfunctional, which can result in unbalanced, aberrant immune reactions (no stimulation of Thl cells, but only of Th2 cells exerting suppressive action). It is now of great importance to establish whether UV-damaged APCs (more particularly, Langerhans cells) migrate from the human skin upon I.W irradiation. Such a follow up study is currently pursued with a high priority. Enhanced migration of UV-damaged APCs in some individuals (e.g. people who had skin cancers removed) could signify an increased risk for UV-induced immune suppression. 2.6 C o n c l u s i o n s
The present NRP-I project has been extremely rewarding in that it produced a good many new data on which to base an animal-to-man extrapolation of UV-induced suppression of immune reactions, such as involved in combating infections and skin cancers. Animal experiments widened the spectrum of infections that could be adversely affected by UV radiation, by including infections not entering through the skin. Thus, the possible impact of an increase in UV-B radiation has been importantly broadened. The well-directed study has identified some promising immunological parameters by which to establish an individual's UV-immunosuppressibility, or a state of immunosuppression against certain (infectious) agents. Further studies are needed to establish the feasibility of measuring these parameters in humans for the proposed diagnostic purposes. Clearly, this research has not yet fully evolved to the point t h a t quantitative assessments of UV immunosuppression, and especially the repercussions on infections and vaccinations can be reliably made. The mathematical centre of the RIVM is, however, carrying out preliminary studies on developing a suitable risk model based on the data generated in the present project. In conjunction with this study, the centre of epidemiology at the RIVM is exploring the possibilities of providing epidemiological data (e.g. influenza data in relation to sun hours during summer) to support assessments of UV effects on infections or vaccinations. As such epidemiological data may not be available in a suitable form, plans are also being made for dedicated surveys in which people will be interviewed to investigate a possible relationship between
959 (solar) UV exposure (during vacations) and subsequent risk or severity of infectious diseases. Although it is not part of our daily experience, animal experiments show that UV radiation from the sun may interfere with our immunity against infections. The relevancy of this UV-immunosuppression for infections and vaccinations in a human population still remains to be established: it is unknown whether we are all indiscriminantely at risk of l.W-immunosuppression, or whether the risk pertains solely to high risk individuals (e.g. people already immunocompromised in some way), or whether the risk is nonexistent. Judging by the available experimental data, also supplied by the present project, the latter alternative does not appear to be the most plausible one. Hence, a continuation of this line of research is well warranted, because our ignorance on the subject hampers a good appreciation of health effects of UV radiation, and of increases in I.W radiation due to an depletion of the ozone layer.
2.7 Scientific cooperation In this NRP project the research groups of Dr. F. de Gruijl (RUU, laboratory for dermatology) and of Dr. H van Loveren (RIVM, laboratory for pathology, section of immunobiology) have been and still are collaborating intensively. Knowledge with respect to Ultraviolet radiation and the immune system are connected due to this collaboration. In addition to this formal collaboration five times a year a scientific meetings among researchers in The Netherlands, concerned with UV-B effects on the immune system are organized. During these meetings plans/results from the participants with respect to research concerning UV-B induced alterations of the immune system are discussed. The participants of this national UV-B network are: 1. RUU, Laboratory for Dermatology, Utrecht. 2. RIVM, Laboratory for Pathology, section Immunobiology. 3. LU, Laboratory for Dermatology, Leiden. 4. UvA (AMC), Laboratory for Dermatology, Amsterdam. In addition to these meetings regarding health effects of UV-B two times a year a meeting is arranged in which the other UV effects projects, financed by NRP, participate (Free University, Amsterdam; RIVM, Bilthoven; State University, Groningen; State University, Utrecht). The first of these meetings was in November last year at the Free University in Amsterdam. The second UVB-effects cluster meeting was at the RIVM in April 1993. This meeting was very successful with respect to the integration of the different UV-B "effects studies". The third and fourth meeting were held in Utrecht respectively Groningen (November 1993 and june 1994). In addition to the national collaborations as mentioned above, the RIVM is coordinating a project sponsored by the EC, entitled: "A basis for better evaluation of risk of increasing UV-B exposure for public health". For this purpose collaboration with several groups in Europe is established: Dr. Dall'acqua, Padua, Italy; Dr. Cerimele, Rome, Italy; Dr. Norval, Edinburgh, Scotland; Dr. Gibbs, Dundee, U.K.; Dr. de Gruijl, Utrecht, The Netherlands. The knowledge/expertises of
960 these different labs will serve to further evaluate the risk of increasing UV-B exposure on public health. The accessibility to h u m a n data with respect to UV-B induced effects obtained by several of the participants is now possible. 2.8 E f f e c t s o f U V - B o n t h e i m m u n o l o g i c a l infections
resistance
to t u m o u r s
and
Assessment It was known from experimental work by other groups t h a t UV radiation could influence the course of certain infections. That were infections where the skin was involved, either as the site where the infection entered the body, such as in Leishmania, or where the infection came to expression, such as in Herpes simplex. In the p r e s e n t project it was shown t h a t UV-B i r r a d i a t i o n of the skin of e x p e r i m e n t a l animals can also aggravate the course of an infection t h a t has nothing to do with the skin, for instance, a worm infection t h a t enters the body with the food and has its damaging effects deep inside the body. This surprising finding broadens the range of infectious diseases that require attention as diseases possibly to be influenced by increasing solar UV-B radiation. The result is being included in the U N E P Assessment on Environmental Effects of Ozone Depletion (1994), as has been the case with several experimental results from this group in the earlier U N E P assessments. Another valuable result obtained in this project was the finding of cells with I_W-B induced damage to their DNA in the deep lymph nodes of mice. Because of the limited penetration of UV-B radiation in animal tissues, this damage m u s t have been inflicted while the cells were in the skin. This finding confirms an important p a r t of the theoretical model on how the i m m u n e system is influenced by UV radiation, and strengthens the basis for the prediction of the ultimate effects. : 3.
I M P A C T O F E N H A N C E D S O L A R UV-B R A D I A T I O N O N P L A N T S FROM TERRESTRIAL ECOSYSTEMS
J. Rozema, M. Tosserams and E. Magendans Free University Amsterdam, Department of Ecology and Ecotoxicology, Faculty of Biology, De Boelelaan 1087, 1081 HV Amsterdam
3.1 I n t r o d u c t i o n
Background Decrease in stratospheric ozone over the last decade has been revealed by satellite m e a s u r e m e n t s . As a result a gradual increase in solar ultraviolet-B radiation occurs. M e a s u r e m e n t s of UV-B radiation in the Swiss Alps indicate an a n n u a l increase of UV-B-radiation of 1% since 1981 (Blumthaler and Ambach, 1990).
S o l a r UV-B a n d terrestrial ecosystems Solar UV-B radiation has important biological and ecological effects. Many crops and n a t u r a l plant species demonstrate reduced growth, under enhanced UV-B
961 radiation but there is a wide range of sensitivity to UV-B radiation between plant species and cultivars. A major part of research of UV-B effects relates to crop species and only a few UV-B effect studies are known of plant species of n a t u r a l ecosystems (SCOPE 1992). Most studies of UV-B effects on plants have been conducted in greenhouse or in controlled e n v i r o n m e n t cabinets with UV-B lamps. Relatively low levels of Photosynthetic Active Radiation (PAR), in controlled environment studies m a y prevent induction of photo-repair of UV-damage. This has led to an overestimation of growth reduction of plants exposed to enhanced levels of UV-B. Outdoor experimental UV-B supplementation systems, with natural levels of PAR provide more realistic responses of plants to enhanced UV-B.
Scientific aims O u r r e s e a r c h a i m e d at d e v e l o p m e n t and a p p l i c a t i o n of outdoor UV-B supplementation and UV-B filtrations systems and to assess effects of solar UV-B radiation to plant species of terrestrial ecosystems.
3.2 Methodology UV-lamp systems In experimental studies with enhanced levels of UV-B irradiance UV-lamps are used t h a t are pre-burnt and filtered with cellulose acetate as a cut off filter of wavelengths smaller t h a n 280 nm. The Cellulose Acetate filters are renewed twice a week to avoid reduced transmission of UV-B related to ageing of the cellulose acetate. Mylar polyester filters absorb UV-B and are used to cover UV-B lamps over control t r e a t m e n t t h a t receive no UV-B irradiance, but where the UV-B levels are the same as in the t r e a t m e n t t h a t receive no with enhanced UV-B (see Figure 3.1).
Figure 3.1 Philips TL12/40 lamps are applied both in indoor and outdoor studies of UV-B effects on plants. The plots shown in the photograph, which are exposed to solar UV-B + UV-B supplied by the lamps, refer to monocultures and mixed cultures of the dune grassland species Calamagrostis epigejos and Holcus lanatus. Thus the effect of enhanced UV-B on competitive relationships between plant species is analysed
962 We use Philips TL12/40 lamps both in indoor controlled environment studies and in outdoor studies in an experimental field and in natural ecosystems. Two outdoor unenclosed systems are applied: A. U~-B Supplementation system and B. UV-B filtration systems.
UV-B Supplementation system. A lamp system to supplement ambient solar UV-B over experimental plots and over natural vegetation has been developed and applied (Figure 3.1). At present this outdoor UV-B supplementation system is applied in a "square wave" mode of enhanced UV-B radiation. This implies switching UV-B lamps on for a fixed period around midday, when natural solar UV-B is greatest. Currently an outdoor solar tracking UV-B supplementation is developed and installed. This involves continuous m eas ur em ent of solar UV-B levels using appropriate UV-B sensors (Yu et al. 1991). This allows accurate simulation of enhanced UV-B throughout the day. There is evidence that plant responses to enhanced UV-B supplied in the solar tracking modulated mode differs both qualitatively and quantitatively from responses to UV-B supplied in the square wave mode. The development and testing of this solar tracking UV-B supplementation is in close cooperation with Dr. Andy McLeod, Institute of Terrestrial Ecology, Huntingdon, United Kingdom and Dr. Gaetano Zipoli, IATA-CNR, Florence, Italy. UV-B lamps are mounted above the vegetation and are maintained at a constant height above the canopy. UV-B sensors will measure and track ambient solar UV-B radiation. Dual UV-B sensors are used beneath both treatment and control racks with UV-lamps, spaced such that one is unshaded from direct sunlight by lamps or frame. The unshaded sensor is selected by the control system preventing shading effects from interfering with feedback control of output of UV-lamps. The lamp output is adjusted to give a constant multiple of UV-B radiation above ambient UV-B radiation. Increases of UV-B radiation relative to ambient UV-B irradiance relating to 20% ozone depletion under clear sky are realized. UV-filtration system. Various types of solid plastic filters are installed above precultured plants in pots or above natural vegetation. These filters absorb either (a) little solar UW-B (and UV-A) (Acrylate filter), (b) all UV-B (Mylar polyester film on top of acrylate filters) and (c) all UV-A + UV-B (Lexan filter). The transmission spectra of these three filter types are given in Figure 3.2.
963
1.0
' ~ii':ii,,e, 0.8 .
.
.
.
.
.
.
.
[
=:=
I
0.6
8
0.4
|
.-,
....,
-~
o.2
0.0
-0.2 280
,
,
300
,
320
!
340
,
,
360
380
400
Wavelen~th ( n m )
Figure 3.2 Solar UV-irradiance (August 1992, 13.00 h), on a clear sky day, beneath the various filter types (A = acrylate, L = Lexan, M = Mylar), compared to full sunlight (no filter). Measurement with a Optronic spectroradiometer (OL 752, Optronics Laboratories, Florida, USA) The development and testing of low-cost solid plastic UV-B filters allow outdoor and ecosystem studies of reduction of ambient levels of solar UV-B radiation. In practice it will allow research of responses of plant from nat ural terrestrial ecosystems (Figure 3.3).
3.3 Responses of terrestrial plants to enhanced solar UV-B irradiation Solar UV-B effect studies have been performed on a number of crop and native plant species both in controlled environment (greenhouse), unenclosed outdoor (experimental garden) and field studies in a dune grassland ecosystem (Tables 3.1 and 3.2). A wide range of sensitivity to enhanced UV-B measured in indoor controlled environment experiments, exists among crop species. Cultivars of crops such as soybean (G[ycine max) vary greatly in sensitivity to UV-B. Of the n a t u r a l plant species from Dutch coastal ecosystems: the salt m a r s h ecosystem and dune grassland ecosystem, no apparent sensitive plant species were found both in indoor and outdoor studies. It should be noticed that no natural plant species of the Leguminosae (a plant group with many plant species sensitive to UV-B) have been subjected to enhanced UV-B radiation levels in outdoor experiments. The woodland understorey species Alliaria petiolata indicates to be sensitive to enhanced UV-B. It may be that plant species occurring in open habitats (Calamagrostis epigejos, Spartina anglica) and being exposed to relatively high levels of ambient solar UV-B have evolved UV-B adaptation mechanisms. Other species occurring in the woodland understorey with reduced PAR and reduced ambient UV-B, seem to be relatively UV-B sensitive. This hypothesis is studied in current research.
964
Figure 3.3 Application of solid plastic UV-B filters in the study of reduction of ambient solar UV-B radiation on plant species of a dune grassland ecosystem. The height of the plastic UV-B filters can be adjusted according to seasonal development of the vegetation. At the background a rack of UV-lamps over the dune grassland vegetation is shown as a UV-B supplementation system
3.4 Ecological effects of ozone depletion; implications for e n v i r o n m e n t a l policy 1. Responses of terrestrial plant species to enhanced UV-B range from very sensitive to tolerant or even a positive response. This broad range type of response possibly reflects the intrinsic heterogeneity of the various terrestric ecosystems. Grassland ecosystems are much more exposed to full solar UV-B than under storey plants from forest ecosystems. .
.
Adaptation to enhanced UV-B may consist of two types of mechanisms. a. UV-B tolerant plant species avoid enhanced UV-B radiation flux at sensitive metabolic sites by absorption of UV-B by epidermal structures and dissolved, UV-B absorbing compounds such as flavonoids. b. UV-B tolerant plant species have effective UV-B repair mechanisms, in contrast with UV-B sensitive plant species. The n a t u r e and occurrence of these adaptation mechanisms has been analysed only for a few plant species. Understanding of UV-B responses of all groups of terrestrial plant species needs research priority. Crop plant species tend to be more sensitive to enhanced UV-B than wild plant species. This may indicate that during the breeding procedure domesticating wild species to crop plants adaptation to (high) UV-B has been lost. Maybe a trade off exists between high productivity and sensitivity to UV-B. This is an intriguing hypothesis. It might indicate that plant species may loose (genetically determined) adaptations to solar UV-B. This implies that certain crops or certain cultivars of crops will be less productive with increasing solar UV-B. It is unknown whether selection for
965 increased tolerance to enhanced UV-B will be successful and over w h a t time period. .
.
There are m a n y uncertainties in the knowledge of UV-B effects on plant life in terrestrial ecosystems: effects of UV-B on reproductive biology are largely unknown, some of the reproductive features will be vulnerable to enhanced UV-B; little is known of UV-B effects on "lower plants" like mosses ferns, fungi and algae; the effect of an enhanced UV-B flux is unknown in a "multiple stress" environment such as UV-B air pollution; UV-B x n u t r i e n t and w a t e r deficiency. T h e r e are some indications t h a t enhanced UV-B will cause i n c r e a s e d occurrence of plant diseases by fungi, bacteria and viruses. Consequences of enhanced solar UV-B to terrestrial ecosystems are expected to be substantial. There will be a shift to dominace of UV-B resistant plant species. The altered biochemistry of plant species exposed to high UV-B fluxes may not only affect plant-herbivore relationships but also change the process of decomposition of leaf litter (see Figure 3.4).
3.5. National and international cooperation N a t i o n a l a n d I n t e r n a t i o n a l contacts. Between the different national UV-B research
groups exchange of experimental results and methods occurs. Every six months a meeting is organized by one of the participating groups. D u r i n g these meetings results are presented and discussed. Close collaboration between different groups exists on a s s e s s i n g the impact of UV-B on DNA Damage. Immunological techniques used in skin cancer research are now also applied to assess DNA damage in marine algae and terrestrial plants. Contacts exist with research groups in Germany (prof. M. Tevini, Karlsruhe) and the United States (prof. A.H. T e r a m u r a and Dr. J. Sullivan, Univ. of Maryland, Washington), who are well known for their work on UV-B effects on terrestrial plants. Results were presented and discussed during visits to these research groups in 1992, 1993, 1994.
National cooperation 1. Dr. L. van Liere, Dr. A. Veen, RIVM, UV-B and aquatic Ecosystems. 2. Dr. W. Gieskes, Dr. A. Buma, RUG, DNA-UV-B-damage, UV-B dosimetrie.
International cooperation 3. Dr. A. McLeod, ITE, Huntingdon, U.K., UV-B experimental set-up. 4. Prof.Dr. M. Tevini, Dr. J. Ros, TU, Karlsruhe, UV-B spectroradiometrie. 5. Prof.Dr. A.H. Teramura, Un. Maryland & Hawaii, Dr. J. Sullivan, USA, advanced cooperation and interchange.
International workshop UV-B and Biosphere The groups involved in this project and in the project on UV-B and algal communities (nr. 851054) will organize an i n t e r n a t i o n a l workshop on the influences of UV-B radiation on terrestrial and aquatic ecosystems. T h a t will be held towards the end of 1995.
966
Sponsors: the Dutch Ministry of Housing, Physical Planning and the Environment, The Dutch Organisation for Scientific Research, the Royal dutch Academy of Sciences and the European Community.
TERRESTRIAL ECOSYSTEMS PAR
UV-B
PAR INTERCEPTION
PLANT
1
!
-~r
LEAF ANGLE CANOPY ARCHIITECTURE LA!
NUTRIENTS
y 2 2
HERBIVORY GRAZING
UV-B ABSORBING COMPOUNDS SECUNDARY METABOLITES
RGR BIOMASS PROD. PRIM. PROD.
UV-B DAMAGE PHOTOREPAIR
/
[DEFENSE AGAINST
SEC. PROD. DECOMPOSITION
3
LEAF LONGEVITY SPROUTING BUDBURST
COMPETITIVE ABILITY
Figure 3.4 Conceptual model structure Impact of Enhanced Solar UV-B on Terrestrial Ecosystems: Physiology, Functioning and Dynamics. Ecosystem Functions: 1. Prim. Prod. = Primary productivity; 2. Sec. Prod. = Secondary Productivity; 3. Decomposition. Ecosystem structure: 4. Competitive ability. PAR = Photosynthetic Active Radiation; Pn = Net Photosynthesis; RGR = Relative Growth Rate; LAI = Leaf Earea Index
967 Table 3.1. Survey of qualitative plant responses to enhanced UV-B in controlled environment studies. Daily biologically effective dose of UV-B relating to 20-50% ozone depletion under clear sky. Response to enhanced UV-B negative very sensitive
response sensitive
no significant response
positive
Crop species Pisum sativum Phaseolus vulgaris Vicia faba Lycopersicon esculentum Cucumis sativus Triticum aestivum Zea mays
X X
X X
X X X
Natural plant species Verbascum thapsus Calamagrostis epigejos Plantago lanceolata Alliaria petiolata Aster tripolium Elymus athericus Spartina anglica Holcus lanatus Silene vulgaris
X X X X X X X X X
The reader is referred to Rozema (1993) for more detailed information
968 Table 3.2 Survey of qualitative plant responses to enhanced UV-B in controlled environment studies. Daily biologically effective dose of UV-B relating to 20-50% ozone depletion under clear sky Response to enhanced UV-B negative very sensitive
Crop species Vicia faba Triticum aestivum
response sensitive
positive
X X
Zea mays Natural plant species Calamagrostis epigejos Plantago lanceolata Urtica dioica Holcus lanatus Verbascum thapsus Silene vulgaris
no significant response
X X X X X X X
The reader is referred to Rozema (1993) for more detailed information.
3.6 Impact of enhanced solar UV-B irradiation on plants from terrestrial ecosystems Assessment The project gives valuable additions to the international efforts in this field: the inclusion of plants from n a t u r a l ecosystems, and the investigation of mixed cultures of these plants. These are necessary steps towards the difficult analysis of the influence of increased UV-B irradiance on natural terrestrial ecosystems, a priority topic in several international listings.
The r e s e a r c h e r s in this project are m a k i n g progress in identifying those ecosystems, or sub-ecosystems, for which the UV-B irradiance is an i m p o r t a n t factor. Cultured crops tend to be more sensitive to enhanced UV-B radiation t h a n wild plants, and plants occurring in woodland understoreys more sensitive t h a n plants occurring in open habitats.
969
0
E F F E C T S OF I N C R E A S E D UV-B R A D I A T I O N ON S T R U C T U R E A N D FUNCTIONING OF ALGAL COMMUNITIES IN DIFFERENT C L I M A T I C ZONES: R I S K S OF A G L O B A L D E C R E A S E IN STRATOSPHERIC ONZONE
L. van Lierel and W.W.C. Gieskes2 1 RIVM, Department of Water and Drinking Water Research, National Institute of Public Health and Environmental Protection, P.O.Box 1, 3720 BA Bilthoven 2 RUG, D e p a r t m e n t of Marine Biology, University of Groningen, P.O Box 14, 9750 AA H a r e n
4.1 S u m m a r y During NRP-I new insights were gained with respect to UV penetration in the aquatic e n v i r o n m e n t and UV stress p h e n o m e n a in microalgae after chronic exposure to realistic levels of UV-B. UV-B radiation and penetration modelling has revealed t h a t UV-B penetration can be high in natural waters, particularly in the open ocean where the concentration of humic acids is low. Here, the UV-B attenuation coefficient approaches that of pure water. Large differences in UV-B sensitivity were found when microalgae were subjected to chronic UV-B exposure. Especially the growth of species occurring in offshore waters, such as the important bloom forming alga Emiliania huxleyi was already inhibited at ambient UV-B levels: for Emiliania huxleyi growth inhibition was found to occur at least up to a depth of 15 m, as calculated for 52~ in the Atlantic Ocean. This indicates t h a t UV-B is a n a t u r a l selective factor in plankton assemblages and t h a t an increase in incident UV-B levels will strongly affect the performance of m a n y organisms in the field. Furthermore, thanks to our approach of measuring chronic UV-B effects over several generations, we can now present convincing evidence t h a t L ~ - B effects are underestimated when only short term effects are t a k e n into account, as done so far in UV-B effect studies. This means t h a t the observed 12% decrease in phytoplankton production in Antarctic waters, t h a t is used often for risk evaluation of ozone depletion, strongly needs to be reevaluated, because this study was based on short term photosynthesis measurements. Moreover, growth r a t e reduction was found to be a more reliable and realistic p a r a m e t e r to be considered t h a n the production m e a s u r e m e n t s done so far, because growth is a resultant of all metabolic processes under UV-B stress, including DNA damage, i.e. not only the photosynthetic process. Therefore chronic UV-B mediated growth inhibition phenomena urgently need to be included in future UV-B effect studies and modelhng. The results obtained during NRP-I have indicated t h a t DNA damage and its subsequent effect on the cell cycle play a major role in UV-B mediated growth rate reduction. This approach is new in aquatic UV-B effect studies and f u r t h e r underlines the importance of studying UV-B effects over several generation times. DNA damage occurs at realistic UV-B levels in most species tested so far, and the ability to overcome DNA damage is now thought to be a p r i m a r y factor causing difference in UV-B sensitivity in microalgae. The UV-B induced cell cycle arrest, t h a t we observed, causes shifts in mean cell volume and the contents of structural compounds (protein) and storage products. This will definitely result in changes in the food web through altered grazing activities by herbivores with specific food size
970 requirements, as well as altered sinking rates, resulting in a proportional increase in vertical carbon fluxes out of the euphotic zone. The experimental results have provided basic insight into the key parameters to be included in effect models, that are designed for risk assessment in specific marine habitats. First modelling on large scale can be realised within the coming years if funding is obtained through NRP-II. The simulation model that we developed recently is focussed on the vulnerability of the Atlantic Ocean ecosystem for integrated climate change factors, including UV-B, temperature and carbon dioxide, and is so far unique within the international scientific community. 4.2 I n t r o d u c t i o n and p r o b l e m definition At present no conclusive picture exists on the impact of enhanced UV-B radiation on the structure and functioning of aquatic ecosystems. The limited knowledge on UV-B leaves enough room for speculation on dramatic effects but also trivialization of the problem. During the past few years evidence for both can easily be found in the media and official reports. It is naive to consider the UV-B problem a passed station following the Copenhagen amendments, first of all because observed ozone depletion levels do not always fit the predicted ozone scenarios and secondly because the effects on ecosystems are still poorly quantified. Therefore, a scientifically sound picture of the effects of UV-B enhancement on ecosystems is required. In aquatic ecosystems microalgae form the first trophic level. Since microalgae are known to be sensitive to UV radiation, and since UV penetrates to ecologically significant depths in aquatic environments, effects on global primary production upon ozone depletion are likely to occur. Moreover, UV induced changes in primary production will affect the functioning and structure of aquatic food webs. Laboratory experiments have shown that UV-B affects almost all metabolic processes. This is not surprising: microalgae are unicellular, so the complete organism is subjected to UV radiation. However, as laboratory light conditions usually deviate from the underwater light environment with respect to applied radiation levels and spectral conditions, extrapolation of laboratory findings to the actual field situation is extremely difficult. Therefore, it is crucial for effect studies to assess first of all the primary targets which determine the net result in situ, i.e. the changes in primary production. Available experimental data refer almost exclusively to short term effects. Natural phytoplankton assemblages or species isolated for laboratory experiments are generally exposed to UV-B radiation for only a few hours, while mechanisms operating on a time scale of more than several hours, such as repair of DNA damage and adaptation are not accounted for in such experiments. Nevertheless, these mechanisms may well determine the sensitivity of an organism to chronic enhancement of UV-B radiation. Therefore, experimental data reported so far were not adequate to support a realistic risk analysis. Obtaining such risk analysis was the ultimate goal of our research effort. Collection of experimental data for the purpose of assessing the quantitative impact of UV-B radiation on natural microalgal assemblages is obstructed by several other problems. UV-B effect studies are complicated by spatial and temporal variations in damaging radiation as well as its interaction with UV-A and
971 visible light, as occurring in the field. Therefore, natural daily courses of UV-B radiation are generally significantly different from experimental exposure regimes. Because several UV induced biological effects appear to be nonreciprocal between dose rate and exposure time, an accurate dosimetry is essential for mimicking natural UV conditions. Furthermore, the spectral distribution of solar light differs significantly from that of most artificial light sources. Since biological effects are known to be strongly dependent on wavelength, biological weighting functions are required for the conversion of radiation spectra to biological effective radiation. So far, no specific "microalgae" action spectrum had been d e t e r m i n e d for the inhibition of chronic UV-B radiation on algal production. Finally, only few data are available on the penetration of UV (especially UV-B) in seas and oceans under different hydrographical and climatological conditions. Spectral data are needed for the estimation of attenuation coefficients in the ultraviolet range, the assessment of biological effective dose rates at different water depths in the euphotic zone and a realistic assessment of in situ UV responses. This type of data should also form basic knowledge to be used in the establishment of an appropriate dosimetry in laboratory experiments. Summarizing, at the start of the project no quantitative data were available to assess the effects on aquatic ecosystems. Additionally, the high accuracy necessary for the quantification of UV-B effects made traditional culture methods only poorly applicable. So the generation of accurate and relevant quantitative data required both adequate infrastructure and method development. 4.3 Aim of the project Considering the above mentioned uncertainties, this project was directed towards the study of long-term effects (several generation times) of enhanced UV-B radiation on algal production. To this end knowledge on the most relevant UV-sensitive metabolic processes had to be gained. This information would then be essential for selecting the most appropriate biological weighting functions (action spectra) for translation of experimental results to field conditions. Furthermore, dose effect relationships had to be established for different species. Also, field measurements of the underwater UV-B light regime were needed to gain knowledge of the vertical extinction of UV-B in various natural waters. The results of both experimental and field m e a s u r e m e n t s would supply basic information to be incorporated in exposure models to assess the effects of various scenarios of UV-B irradiance on specific habitats. Two main lines of research were pursued: A. Analysis of in situ UV-B light regimes of different water types (both freshwater and marine). This included: development of a radiation model to calculate surface UV-B irradiation and daily doses; verification of the radiation model by field m e a s u r e m e n t s and by measurements performed by the Laboratory of Radiation Research of the RIVM; field m e a s u r e m e n t s of solar irradiation and spectral a t t e n u a t i o n coefficients using an Optronic OL 752 spectroradiometer equipped with a submersible enclosure; m e a s u r e m e n t of intrinsic optical properties (absorption, scattering coefficients) of suspended material (algae and detritus) and dissolved organic m a t t e r for estimation of u n d e r w a t e r t r a n s m i s s i o n of UV-B radiation, development of a penetration model to estimate UV-B exposure of algae in different water types and to determine potentially sensitive water types.
972 B. Experimental studies of sensitivity and protection to natural ranges of UV-B radiation in natural phytoplankton assemblages and unialgal cultures. This included: assessment of dose-response relationships for representative marine and freshwater algae; analysis of reciprocity between dose rate and exposure time; experimental studies on causal relationships between algal sensitivity and relevant physiological mechanisms; development of a computer controlled dynamic light system (DLCS) to simulate natural light regimes in laboratory cultures; development of a special culture system for the assessment of action spectra and analysis of physiological reactions; assessment of a "general" action spectrum for the long-term effect of UV-B radiation on algal growth; integration of effect and exposure data into a general risk assessment of UV-B effects on algal production. 4.4
Results
Infrastructure and methodology built up during the project -
-
-
-
A spectrophotoradiometer (Optronics OL 752), purchased at the beginning of the project, was equipped with a submersible enclosure to allow for spectral u n d e r w a t e r m e a s u r e m e n t s to a depth of 12 meters. Expertise was accumulated using this equipment during several cruises (Dutch estuarine waters, open Atlantic waters). The spectroradiometer measurements also provided the possibility to establish an accurate dosimetry in experimental systems. Two specialised culturing facilities were developed and made operative: a computer controlled dynamic light system (see below) and a culturing device for measuring wavelength dependent biological effects (action spectra, see below). Special temperature controlled culture rooms were made operative for UV experiments. Detailed information was collected on changes in transmission of materials (perspex, cellulose acetate, polystyrene, glass filters) to be used in UV experiments, as a prerequisite for the establishment of an accurate dosimetry during long-term experiments. A method was developed to study in vivo DNA damage in individual phytoplankton cells, using immunofluorescent labelling. This has resulted in long-term cooperation projects with other research groups (see 4.7). An Image Analysis System was purchased to study DNA damage in natural p hy to p lan k ton assemblages. This system was also made operative for accurate cell size measurements. A Coulter Counter Multisizer system was purchased eventually to study in detail the UV-B induced changes in cell size and volume.
Radiation and penetration models A UV radiation model has been developed to estimate natural light regimes as a function of meteorological conditions, season and latitude. Verification of the radiation model was done using measurements performed by the Laboratory of Radiation Research of the RIVM as well as field data obtained with the Optronics OL 752 during several cruises. To allow analysis of under water UV exposure a separate model was developed to calculate the transmission of I . ~ under water. This model relates spectral transmission with intrinsic optical properties: absorption and scattering characteristics of suspended and dissolved substances
973 in various water types. Laboratory and field data on absorption properties of algal cells, detritus and humic acids were collected and included in the model. Through application of the model mean UV exposure levels could be calculated for Dutch aquatic systems. It was found that dissolved organic m a t t e r (humic acids) causes most of the a t t e n u a t i o n in freshwater and coastal regions. Here the UV-B attenuation coefficient is up to 20 times higher than the attenuation coefficient for visible light. Due to the very low levels of humic acids the attenuation of LW-B r a d i a t i o n in the ocean is relatively low. In clear ocean w a t e r s the UV-B a t t e n u a t i o n coefficient approaches the theoretical value for pure water i.e. two times the attenuation of visible light.
Dose-effect studies: interspecific differences in UV sensitivity Various marine and freshwater algae were tested for their sensitivity to chronic UV-B exposure using growth rate reduction as an overall indicator of UV-stress. Large differences in sensitivity were found between species. A comparison of a n u m b e r of freshwater species demonstrated a generally higher UV sensitivity of desmids and diatoms as compared to green algae. The prymnesiophyte Emiliania huxleyi, thought to be an important bloom forming alga in temperate and boreal marine waters, was found to be very sensitive to UV-B, more t h a n any other marine alga tested. Although marine pelagic diatoms differed considerably in their UV-sensitivity, a comparison with benthic diatoms, isolated from the Dutch Wadden Sea, revealed a significantly higher tolerance of the latter group. No general difference between pelagic diatoms and toxic dinoflagellates was found. Remarkably, a freshly isolated strain of a benthic diatom species did not have a higher UV-B tolerance t h a n a strain that had been cultured in the lab for several years. Reciprocity between UV dose rate and exposure time was tested in several algal species. The inhibition of growth rate was stronger at shorter exposure times (high dose rate) t h a n at long exposure times (low dose rate). In other words, damage of chronic UV exposures was not only a function of dose. These results further stress the importance of applying accurate dosimetry and careful experimental setup.
Studies of mechanisms behind growth inhibition In preliminary experiments photoinhibition did not explain the observed UV induced growth inhibition. Upon UV-B exposure, increases in cell dry weight were observed, while the light capturing capacity remained unchanged. This indicated t h a t the photosynthetic apparatus is not the primary target under chronic UV exposure. Also the effect of low UV-B levels on pigment content, cell size, protein and carbohydrate content, and u l t r a s t r u c t u r e were investigated. UV induced growth rate reduction was typically accompanied by increases in protein content, pigment content and mean cell volume. These trends were stronger at higher UV-B radiation levels. However, at very high UV-B doses, when complete growth inhibition occurred, no increase in cell components were measured. These results were found to be very consistent when testing different species and various UV culture systems (see also below). The increases in cell dry weight, protein and carbohydrate content, pigment content and mean cell size hint at a UV mediated delay in cell division. In this view, low UV-B levels would arrest DNA synthesis t h r o u g h the occurrence of DNA damage, but not the biosynthesis of cell components, thereby arresting the cell cycle at the end of the G1 phase. This was
974 later on confirmed by DNA content and DNA damage m e a s u r e m e n t s in the marine diatom CycloteUa sp. (see above).
UV-B Effect studies using computer controlled Dynamic Light Regime To realize an accurate and realistic dosimetry, a continuous culture systems with steering software was developed to simulate natural photosynthetic light and UV-B regimes simultaneously. The culture system was designed to allow exact quantification of UV-B exposures. High intensity light sources (PAR up to 2000 ~mol.m-2.s-1) provide a realistic UV/PAR ratio. By computer controlled angular slat displacement of Venetian blinds natural sine-curve-like intensity courses for both UV and the visible part can be simulated, as well as the light regime for a superimposed vertical mixing. The combination of rectangular algal cultures and light sources allows accurate dosimetry. In this culture system the effect of LW-B on g r o w t h rate and other physiological p a r a m e t e r s of the common green f r e s h w a t e r alga Selenastrum capricornutum was studied. During short term exposure experiments (transition from no UV-B to UV-B) only significant damage to the photosynthetic system could be detected. A reversed transition to no UV-B d e m o n s t r a t e d recovery of the system. However, although with a lower midday level, a more balanced photosynthetic activity, comparable with the conditions in the absence of I_W-B radiation, was observed after a prolonged exposure to UV-B radiation. The m e a s u r e d daily photosynthetic activity u n d e r chronic UV-B exposure poorly reflected the overall growth conditions. As shown in Figure 4.1, chronic UV-B exposure caused significant increases in mean cell dry weight. Also, mean cell volume and pigment and carbohydrate content increased as a result of UV-B exposure. These results, in combination with the p h o t o s y n t h e s i s m e a s u r e m e n t s , once more hint at UV-B induced cell cycle arrest due to DNA damage rather than inhibition of the photosynthetic apparatus. These results also e m p h a s i z e the incompatibility of acute and chronic UV effects" acute UV responses, such as those m e a s u r e d in short term oxygen evolution or 14C incorporation experiments, might well underestimate chronic effects m e a s u r e d after several generation times.
975
35-
30-
25Cn N (D C3 c- 2 0
I
--
._~
15-
I 10-
I P-sat No UV-B
I P-sat
I Posat
Low UV-B High UV-B
I P-lim No UV-B
I P-lim Low UV-B
treatment
Figure 4.1 Steady State levels of Selenastrum capricornutum cell dry weight as a result of UV-B exposure. P-sat = P-saturated; P-lim = P-limited. Open circles: start of light period; closed circles: end of light period; Vertical lines through data points: standard deviations
DNA damage and repair Very little is known about UV-B induced effects on microalgal DNA. This lack in knowledge is surprising, since l_~ effect studies on higher organised eukaryotic organisms, including man, are mainly focussed on structural changes in DNA. Different lesions can be induced by UV, of which thymine dimers are not only the most abundant photoproducts, but also typically formed as a result of UV-B exposure. A method was developed to detect this kind of cyclobutane dimer in single microalgal cells with a monoclonal antibody, following recent developments in skin cancer research. Detection and quantification of DNA damage is possible after fluorochrome labelling in combination with epifluorescence microscopy and flow-cytometry. A linear relationship was found between the applied UV-B dose and the amount of thymine dimers found in nuclear Cyclotella sp. D N A (Figure 4.2). Within a population of cells the amount of damaged cells, as well as the mean damage level in damaged cells increased with increasing UV-B dose. Kinetics of damage and repair have been studied at various low levels of UV radiation.
976 Simultaneous measurements of cellular DNA content showed a decrease in DNA synthesis during and after several hours of UV exposure. These results, in combination with the observed increases in cell size and cellular protein and pigment levels, indicate that the cell cycle is arrested at the end of the G1 phase. It was also found that the cell cycle can be resumed only after the damage is repaired. Additionally, a shift was observed towards G2 cells in populations of Cyclotella which had been exposed to I_W-B for several days. This seems to be due to the fact that cells with a higher DNA content contain more thymine dimers, requiring more time and energy for repair processes. Determination of species-specific differences in the induction of damage and repair may explain differences in UV-B tolerance and therefore be useful for UV risk assessment studies with respect to shifts in natural populations exposed to a structural increase in incident UV radiation.
Action spectra Action spectra determined so far were either based on the study of higher eukaryotic organisms or isolated cell organelles. Mostly monochromatic light instead of polychromatic light was used to establish these action spectra. For phytoplankton one general action spectrum has been described, based on short term photosynthetic rate measurements. Internationally, very little attention has been payed to the establishment of action spectra, probably because the description of an action spectrum is difficult and time consuming. A special culturing system was developed to study wavelength dependent chronic biological effects using whole organisms and polychromatic light. Using the marine diatom Cyclotella sp. as the test organism, tentative action spectra have been constructed for growth inhibition and DNA damage. Simultaneously wavelength dependent changes in physiological parameters were studied, such as protein and photosynthetic pigment content, and cell size. Both the DNA and growth inhibition action spectrum are very steep in the UV-B region of the spectrum and very much alike. Also, protein, pigment and cell size patterns showed high correlations with the action spectra. The results indicate that wavelength dependent UV-B effects on growth rate are strongly determined by DNA damage and thereby by changes in the cell cycle.
977 20 UV-B intensity: I W/mZ 1
~n
15
0
z
10 ~0
x~ z c~
0
I
I
I
2
3
UV-B dosis (KJ/m z)
Figure 4.2 The formation of thymine-dimers in Cyclotella sp. as a function of UV-B dose. DNA damage (Y-axis) is expressed as the mean damage level in damaged cells within a population of UV-B exposed cells UV-B effec t m o d e l To integrate the experimental results the development of a UV-B effect model has been started. The model structure is shown in Figure 4.3. To realize a global level the Atlantic Ocean has been selected as the model system. It has been chosen because of its pole to pole orientation, covering all climatic and latitudinal zones, both coastal and oceanic. Also, a substantial amount of literature data is available on the various subsystems. By the integration of a mathematical model with the effect study, an improved framework and structure is created for experimental research. The dynamic model is aimed at foodweb processes. P h y t o p l a n k t o n growth is r e g u l a t e d by n u t r i e n t s , t e m p e r a t u r e , PAR and UV-B light, and zooplankton grazing. A distinction will be made between different algal and zooplankton groups, each with their own specific parameters. In this way shifts in the food web are included as combined effects of ADT, ACO2 and AUV-B. Changes in primary and secondary production can be expected. The model will provide both science and politics with a tool to obtain an integrated picture of UV-B effects on aquatic ecosystems. Due to the technical complexity of the research and the r a t h e r unexplored field of UV-B effects on ecosystems, the development of the model (and the generation of experimental input data) will overrun the NRP I period. The expected uncertainties in the model will be quantified and reduced by means of a recently developed model analysis resulting in probability distributions for the parameters and variables.Using this model risk assessments will be made for different marine systems and UV scenarios.
978
temp
PAR
NSML
A i humic id
9
zoo
W........."..1. 4
sP
DET
/ N deep .~I~
Figure 4.3 Schematic structure of the UV-effect model. NSML = N u t r i e n t Surface Mixed Layer; Ndeep = Nutrient deep layer; PAR = Photosynthetic Active Radiation; temp = temperature; NPP = Net Primary Production; SP = Secondary Production; PHYT = Phytoplankton; ZOO = Zooplankton; DET = Detritus
4.5 C o n c l u s i o n s and p e r s p e c t i v e s of results o b t a i n e d d u r i n g N R P I Increased exposure to UV radiation can affect aquatic ecosystems. Phytoplankton produces about 30% of the biomass on Earth, and thereby fixes a large proportion of the greenhouse gas carbon dioxide. Changes in the structure of the algal community may seriously endanger species higher in the food chain. A study of the effects of UV on marine ecosystems is especially relevant because changes in biological activity in the ocean have repercussions on global biogeochemical cycles, not only of carbon but also of sulfur, another element that is directly related to climate change through dimethyl sulfide emission from ocean to atmosphere. During NRP I new insights were gained with respect to UV penetration in aquatic environments and UV stress phenomena after chronic exposure to low levels of LW-B. The results collected during this project will contribute to the knowledge of tolerance and selection. The results obtained during NRP I have revealed that many species show growth inhibition even at ambient UV-B levels. Benthic algae isolated from the Wadden
979 Sea are less sensitive to UV-B compared to pelagic species. This is most likely due to genetic adaptation to the ambient UV-B levels impinging on mud flats during spring and summer. In contrast, marine pelagic species of clear open waters, such as Emiliania huxleyi, are very sensitive to UV-B. This indicates t h a t pelagic algal c o m m u n i t i e s will be affected more strongly by increases in incident UV-B radiation. Clearly, UV-B is a n a t u r a l selective factor in plankton assemblages. Worrest (1983) already calculated t h a t a UV-B free environment would increase global p r i m a r y production with 12%. Since Emiliania huxleyi forms blooms in those a r e a s where ozone depletion has been recorded (northern p a r t s of the Atlantic Ocean), increases in incident UV-B radiation will strongly affect the performance of this organism in the field. As shown in Figure 4.4, which is a combination of experimental growth inhibition data and BED values derived from our radiation and penetration models, significant growth inhibition of Emiliania huxleyi due to UV-B will occur during summer at 52oN at the m e a n ozon level for this latitude: 50% growth inhibition at 5 m depth considering the DNA action s p e c t r u m (ED50(DNA) in Figure 4.4); 50% growth inhibition at 15 m w h e n considering the Plant action spectrum (ED50(PLANT) in Figure 4.4). The results have shown t h a t short term effect studies done so far to estimate UV-B effects on microalgae give a considerable underestimation of UV-B effects: our chronic exposure experiments have revealed t h a t short term photosynthesis m e a s u r e m e n t s are not r e p r e s e n t a t i v e for chronic effects such as growth r a t e reduction. This means t h a t the calculations of production decrease based on short t e r m e x p e r i m e n t s of UV effects on photosynthesis in Antarctic w a t e r s (12% reduction of primary productivity, as estimated by American scientists) have to be reevaluated using p a r a m e t e r s which take into account the chronic UV-B effects such as growth reduction. Our experiments also suggest that growth rate reduction is a more reliable and realistic p a r a m e t e r to be m e a s u r e d t h a n the production m e a s u r e m e n t s done so far: growth rate takes into account UV-B effects over several generations, and, moreover, growth rate is a r e s u l t a n t of all metabolic processes affected by UV-B, including DNA damage - certainly not only the photosynthetic process. UV-B effect modelling is in need of urgent adaptation with respect to this point. This will be realized in the second phase of NRP (NRP II).
980
E(~)~ 1
ED50(DNA)
-
ED50(PLANT)
Depth
lO
lOO
1000
I
,oi
(m)
Figure 4.4 Calculated IJW-B exposure for 15 J u n e at 52~ and ozone thickness of 320 DU. x-axis: biologically effective daily UV-B dose (J.m-2.day-1); Y-axis: water depth (m); DNA: biologically effective dose calculated with Setlow's DNA action spectrum; PLANT: biologically effective dose calculated with Caldwell's P l a n t action spectrum. ED50(DNA): depth at which 50% growth inhibition occurs for Emiliania huxleyi considering the DNA action spectrum; ED50(PLANT): depth at which 50% growth inhibition occurs for E. huxleyi considering the Plant action spectrum The results obtained during NRP I have indicated t h a t DNA damage and its subsequent effect on the cell cycle plays a major role in UV-B mediated growth rate reduction. This approach is new in aquatic UV-B effect studies and further underlines the importance of studying I.W-B effects over several generation times. DNA damage occurs at realistic UV-B levels in most tested species, and the ability to overcome DNA damage is now thought to be a primary determining factor of UV-B sensitivity in phytoplankton species. A substantial body of evidence is presented during the project, indicating t h a t UV-B affects the cell cycle of microalgae, causing an arrest in the G2 phase. This causes shifts in mean cell volume, pigmentation as well as the content of s t r u c t u r a l compounds and storage products. This will most certainly result in changes in the aquatic food web. Herbivore predators in aquatic environments will be affected by these changes because many herbivores rely on a defined food size spectrum, which is altered by cellular volume increases. Also sinking rates will be affected, resulting in a proportional increase in vertical carbon fluxes out of the euphotic zone. Although refinements have to be made with respect to validation, the development of the r a d i a t i o n and penetration models have provided useful tools for the
981 calculation of UV-B exposure to phytoplankton in situ. F u r t h e r m o r e , action spectra will furnish the assessment of biological effective dose rates in n a t u r a l waters. The action spectra for DNA damage and growth can be incorporated in the radiation models after necessary fine tuning operations planned for the future. The experimental results obtained during NRP I provide basic insight into the key p a r a m e t e r s to be included in the model dedicated to risk assessment for specific m a r i n e habitats. Insight into the exposure of algae to chronic n a t u r a l levels of UV-B and effects on the cellular level, to a great extent derived from the insights gained during NRP I, form a firm basis to establish first modelling on large scale. Within the coming years our UV-effect model can become operative (depending on NRP II funding). This simulation model is specifically focussed on the vulnerability of the entire Atlantic Ocean ecosystem for integrated climate change factors, including UV-B, t e m p e r a t u r e and carbon dioxide, and is so far unique within the international scientific community. 4.6 C o o p e r a t i o n w i t h o t h e r r e s e a r c h g r o u p s w i t h i n a n d o u t s i d e N R P
Within NRP framework: -
VU A m s t e r d a m The Netherlands (Dr. J. Rozema): cooperation with respect to DNA damage measurements and modelling. AZU The N e t h e r l a n d s (Prof. J.C. van der Leun, Dr. F.R. de Gruijl): DNA damage method and action spectra.
Outside NRP framework: -
-
Laboratory of Radiation Research (at RIVM), The Netherlands: UV/Ozone data. NIOZ, The Netherlands (Dr. M.J.W. Veldhuis): Flow cytometric DNA damage m e a s u r e m e n t s , Cooperation between NRP participant and NIOZ is being materialised in a joint Ph.D project dedicated to the effect of UV-B on oceanic picoplankters. This project (financed by NIOZ) will s t a r t in the a u t u m n of 1994. LU W a g e n i n g e n , The N e t h e r l a n d s (Dr. E. van Donk) UV-B effects on phytoplankon-zooplankton interactions. TNO, The Netherlands (Dr. L. Roza): DNA damage method development. RUG Plant Physiology Dept., The Netherlands (Dr. F. van Hasselt): General exchange of methods and equipment. University of Oldenburg (Prof. Dr. Th. Hiipner): Mesocosm experiments. University of Lund, Sweden (Dr. Nils Ekelund): DNA damage m e a s u r e m e n t s in mesocosm experiments.
4.7 E f f e c t s o f i n c r e a s e d UV-B r a d i a t i o n o n s t r u c t u r e a n d f u n c t i o n i n g o f a l g a l c o m m u n i t i e s in d i f f e r e n t c l i m a t i c zones: r i s k of a g l o b a l d e c r e a s e in s t r a t o s p h e r i c o z o n e
Assessment Among the innovating elements in this project are the prolonged exposures to enhanced UV-B radiation, extending over several generations of the organisms studied. This realistic experimental condition provides for the inclusion of processes such as repair of DNA damage, and adaptation. For the first time, attention is given to DNA-damage in algae, using techniques developed in Dutch research on skin cancer. Initial determinations were made of action spectra for some of the
982 effects of UV radiation on algae; these show a strong involvement of the UV-B wavelength range. This work is giving real contributions to the international efforts to predict the effects of increased UV-B irradiance on p h y t o p l a n k t o n and the potentially important consequences for the global atmospheric. 5.
INTEGRATION
Integration of the various topics is obviously desirable in a large national research programme. The interim evaluation by S&PA and HCG signalized t h a t in the area of h u m a n h e a l t h there were projects on m a l a r i a vectors and UV-B effects, but direct effects of climate on h u m a n health were not covered in the NRP. The latter should be included anyway in the integration efforts under Theme Integration. The immediate reason for the lack of projects on direct effects of climate on h u m a n h e a l t h was t h a t there were no project proposals in this area. The reason behind this is probably t h a t The Netherlands have a moderate climate, where extreme w e a t h e r conditions are rare; death by heat or cold is practically unknown, and is unlikely to occur with the changes now within the range of expectations. Research into such effects is apparently not considered urgent, or expected to be rewarding. The i n t e g r a t i o n desirable was achieved in a different direction. The groups i n v e s t i g a t i n g effects of UV-B r a d i a t i o n on health, t e r r e s t r i a l and a q u a t i c ecosystems joined forces, and s t a n d a r d i z e d equipment, m e a s u r e m e n t s and methods. Beyond the effects groups the contacts were extended to the groups m a k i n g m e a s u r e m e n t s of ambient UV radiation and those m a k i n g integration models. The specifications of the UV-measurements were carefully adjusted to the needs of the effects studies. And the UV-B effects groups provided input for a longitudinal integration model. This model was developed outside, but in close connection with the NRP, and ranged from production and release of ozone depleting chemicals through atmospheric change to penetration of UV-B radiation and future incidence of skin cancer in The Netherlands; it is the explicit intention to include also other effects of UV-B radiation as soon as sufficient q u a n t i t a t i v e information will be available (Slaper et a1.,1992). 6.
RELEVANCE FOR POLICY MAKING
Even while only the skin cancer effects are sufficiently quantified to be t a k e n into account in the integrated model mentioned, the results already have relevance for Dutch policy. The additional mortality from skin cancer, predicted on the basis of ozone depletions as expected during the coming decades well exceeds the limit of 1 d e a t h per million of the p o p u l a t i o n per y e a r ( H e a l t h Council of The Netherlands,1994). This limit of 1 per million per year is considered the "maximal tolerable risk" in Dutch policy with respect to other e n v i r o n m e n t a l factors influenced by h u m a n activities, such as chemicals and iozining radiation.
983 The ozone depletion t h a t will occur during the next 50 years in spite of successful international efforts to protect the ozone layer will, therefore, have at least one effect of relevance to Dutch policy. As skin cancer is the only effect for which such calculations can be made until now, this finding suggests t h a t quantification of other effects, on health as well as ecosystems, is highly desirable. Some of these effects may will be more important than skin cancer. The H e a l t h Council of The N e t h e r l a n d s is an advisory body to the Dutch Government in matters of health and of the environment. The report on ultraviolet radiation from sunlight was to some extent a byproduct of the research activities in the NRP. The report was prepared by a committee of 14 scientists; 8 of them were also involved in the NRP. The Minister of Housing, Physical Planning and the Environment recently wrote a letter to the Chairman of the Health Council (letter SNV/27694009, dd. July 7, 1994), recognizing the value of the lYV-report for Dutch environmental policy. Likewise, t h e D u t c h i n p u t to the new (UNEP, 1994) A s s e s s m e n t on Environmental Effects of Ozone Depletion results mainly from research sponsored by the NRP.
7.
OUTLOOK: FUTURE WORK
The conditions for research on effects of increasing UV-B radiation are unsteady. Granting agencies tend to refer to each other for the funding. This is made even easier by the fact t h a t the necessity of this research is contested from a variety of viewpoints: a. The problem of ozone depletion does not exist. b. The problem exists, but the biological impacts are negligible. c. The effects on h u m a n health may be easily prevented. d. The problem was already solved. Each of these statements will be briefly discussed.
The problem of ozone depletion does not exist There is a fierce publicity action to deny the problem of ozone depletion. It would be a fake problem, brought into the world by commercial interests. Moreover, the reasoning continues, even if ozone depletion would occur, it would not lead to more UV-radiation, and even if there would come more UV radiation, t h a t would be beneficial rather than harmful (Maduro and Schauerhammer, 1992). The message is pseudoscience, but for the media this message appears to have news value, more t h a n the points of view of regular science which are fairly well-known by now. The counteraction is partly successful, at least in inducing question m a r k s in public opinion. The potential danger is erosion of the public support for the efforts to protect the ozone layer.
The problem exists, but the biological impacts are negligible This viewpoint is around in many forms. I quote one example, written recently by a scientist working in a related field. He suggests t h a t the impacts of the expected increases of UV radiation cannot be very serious, because "an increase in yearly
984 effective UV of 15% at groundlevel is comparable with moving about 400 km nearer to the equator (e.g. Groningen-Maastricht) or moving to an altitude of 500 m compared to sealever'. It a p p e a r s useful to discuss this type of a r g u m e n t a t i o n , as it occurs more generally. The reasoning given can hardly apply to natural ecosystems. Seas and forests do not move appreciably. But even if the reasoning would be valid only for people, it would save a lot of work if it would be so easy to come to a conclusion on the health effects of increased UV-B radiation. The conclusion may appear plausible to a person who did move 400 km South. His individual risk for a particular health effect may have increased from 1.0 x 10-4 to 1.3 x 10-4, but he cannot possibly notice that. He may well be convinced t h a t there was no change at all. Multiplication of this zero-impression by the number of people in a population forms the basis for the suggestion that the risk of such moves is also insignificant for populations. The correct procedure would be, of course, to m u l t i p l y the estimated increase in individual risk by the number of people in the population. Such estimated risks are, however, not yet available for most of the effects under consideration. But this is no excuse for following a loose reasoning which cannot even produce a valid approximation. The risk of moves by entire populations cannot be checked directly, either, because such moves are practically impossible. The Dutch population could move to an altitude of 500 m only by going up into the air, and 400 km south only by invading Belgium and France. How can we know t h a t the risk of such moves would be small? The best indications in this area are produced by migration studies in epidemiology. These investigate how individual risks change by migration from, for instance, Western Europe to Israel or Australia. The results of such studies, together with m a n y other data, form the basis for the risk estimates made for h u m a n health. Only from such quantitative estimations it can be concluded w h e t h e r or not a certain change of UV-B radiation can have important effects. For m a n y effects, further investigations are needed before such a conclusion is possible, but some well-founded quantitive estimations have already been made for skin cancer. These have lead to the conclusion t h a t a 20 percent increase in the incidence of nonmelanoma skin cancer is to be expected after serveral decades as a consequence of ozone depletion (Madronich and de Gruijl, 1993; UNEP, 1994). As these cancers already have the highest incidence of all forms of cancer in several light-skinned populations, this will cause a significant aggravation of an existing public health problem. That can not be talked away with pseudo-quantitative suggestions.
T h e e f f e c t s o n h u m a n h e a l t h m a y be e a s i l y p r e v e n t e d For effects of increased UV-B irradiance on h u m a n health, there is a special consideration. The UV-doses received by people depend not only on the irradiance outdoors, but also on their own behaviour. Any increase in solar UV-B radiation m a y be compensated by staying indoors more of the time and, when outdoors, by keeping in the shade, wearing hats and more clothing and by using sunscreens. It is
985 easier, therefore, to limit undesirable effects by influencing the behaviour of people t h a n by policies r e q u i r i n g knowledge of the m a n y possible effects, such as melanoma and nonmelanoma skin cancer, cataracts, suppressions of the immune system, etc. This solution presupposes that it is feasible to influence the behaviour of people in the desired way. Experience until now gives no strong support to this assumption. Some success is reported only from Australia in recent years. That was achieved in an intensive campaign, launched under exceptionally favourable circumstances: the incidence of skin cancer in Australia is so high that almost everyone knows the problem in his own family. Branches of the Antarctic ozone hole passing over Southern Australia helped mobilize public attention and fear. Even there it is an open question if it will be possible to sustain this changed behaviour for long periods of time, t h a t is, for at least several decades. That would be necessary to influence some of the long-term effects, such as squamous cell carcinoma and cataract. In areas where light-skinned people live under less extreme conditions, it will be even more difficult to convince people t h a t they should change their behaviour. There is certainly reason for trying it, which was recently recommended by the Health Council of The Netherlands, and is now being carried out by the Dutch Cancer Campaign. There may also be reason to do research on how to influence peoples' behaviour more effectively. There is no reason, however, to think that this can be done instead of studying the effects of I.rV-B radiation. Epidemiological observations strongly suggest, for instance, t h a t working indoors, which appears one of the most efficient ways of reducing the UV-exposures, actually increases the risk of melanoma. That may have to do with the irregularity of the remaining exposures. In any case it indicates t h a t protection is not an easy overall-solution. J u s t as other preventive measures it requires knowledge, such as dose-effect relationships and action spectra of the various UV-B effects.
T h e p r o b l e m w a s a l r e a d y solved Prediction of effects of increased UV-B irradiance played an important role in the decision-making process leading to the international efforts to protect the ozone layer. The foundation was laid in the Vienna Convention for the Ozone Layer (1985). Limits to production and use of chemicals that damage stratospheric ozone were specified in the Montreal Protocol on Substances t h a t Deplete the Ozone Layer (1987), and made more stringent in the Amendments of London (1990) and Copenhagen (1992). Practically all producer nations joined this process. According to the latest Amendment, production of almost all ozone depleting chemicals will be stopped by 1996. This process cannot be speeded up any more, even if research would identify i m p o r t a n t new effects of increased UV-B radiation. This is reason for some policy-makers to decide t h a t this type of research does not deserve high priority anymore. This has led to a decrease of funding, especially in the USA. The reasoning given is convincing from the viewpoint of policy decisions on the Montreal Protocol. This is, however, a r a t h e r limited viewpoint. The banning of ozone depleting substances was a necessary first step. Even on the assumption of
986 full and worldwide compliance with the agreements reached, the chlorine loading of the atmosphere will continue to rise for several years, and the ozone layer will be reduced for at least half a century. And anything less than full compliance will make the situation worse. Policy makers, including those in the Dutch Ministry of Housing, Physical Planning and the Environment, are asking what will be the most important consequences, and wh at can be done to prevent or mitigate these. Giving answers to these questions requires much more knowledge than answering the initial question, of whether or not depletion of the ozone layer would have effects important enough to justify action to protect the layer. Organized science has expressed itself clearly on this point. The International Council of Scientific Unions has formed a Scientific Committee on Problems of the Environment (SCOPE). SCOPE paid special attention to the effects of damage to the ozone layer that will persist for a long time in spite of protective action. It produced two reports, "Effects of Increased Ultraviolet Radiation on Biological Systems" (1992) and "Effects of Increased Ultraviolet Radiation on Global Ecosystems" (1993). The potential impacts are considered of such importance that the scientific community is urged to increase its efforts: "This new stage of quantitative questions will require full utilization of existing research capacity, and a major expansion in this direction is essential" (SCOPE 1992, p. 25). The United Nations Environment Programme (UNEP) took a leading role in the international efforts to protect the ozone layer. It knows well what was achieved so far. But it does not support the notion that effects research fulfilled its function, and can have lower priority now. In its latest report on "Environmental Effects of Ozone Depletion" (1991), it calls for better funding of research on effects of increasing UV-B radiation: "Urgent problems are still waiting to be addressed" (UNEP, 1991, p.i).
Actual developments The various conflicting views about the urgency of research on effects of increased UV-B radiation play a role in all communities, and cause great variations in funding. Research efforts in the USA, which provided a large proportion of the information in earlier years, have decreased. The European Community, on the other hand, started late but increased its efforts in recent years in the research programme "Environment". Several European countries have national research programmes going. In The Netherlands, NRP-I has given a real boost, roughly doubling the research volume in this area. With the changing philosophy, NRP-II putting emphasis on systems rather than climate factors, the support may well decrease again. This would be a loss for the generation of much needed knowledge, and a pity of the investments in equipment and know-how during NRP-I. Other sources of funding on this scale are not available in The Netherlands.
987 8.
CONCLUSIONS
The projects on effects of increasing UV-B radiation in NRP-I have produced valuable contributions in their areas (section 3 of this assessment). With respect to the importance of increasing UV-B radiation the following conclusions appear justified: a. In the area of h u m a n health, UV-B radiation has many influences. The UV-B increases as expected even under the most favourable scenario of ozone depletion do have health consequences relevant to Dutch policy. Uncertainty about other potentially grave impacts calls for intensive studies. b. In t e r r e s t r i a l ecosystems, the importance of the UV-B irradiance is very different for the various subsystems. The NRP-researchers made progress in identifying those sub-systems where UV-B irradiance is important. c.In aquatic ecosystems the UV-B irradiance is likely to be an important factor. The effects of increased UV-B radiation also have feedback on the atmosphere. The projects on UV-B effects benefited from mutual interactions, with respect to equipment, m e a s u r e m e n t s as well as photobiological aspects. The interactions with the groups making measurements of ambient UV radiation and integration models for effects may be further intensified. The new philosophy for NRP-II, with its emphasis on systems rather than climate factors, may lead to a spreading of UV-B research over different Themes. In t h a t case, some form of interaction in the UV-B sphere should be maintained. It would be a real loss if this new philosophy would lead to cutting out the important study on effects of UV-B radiation on h u m a n health. In The Netherlands UV-B radiation probably has more serious impacts on h u m a n health t h a n any other climatic factor. 9.
REFERENCES
Blumthaler, M. and Ambach, W., 1990. Indication of increasing solar ultra-violet-B radiation flux in Alpine regions. Science 248: 206-208. Buma, A.G.J., Hannen, E.J. van, Veldhuis, M.J.W., Roza, L., and Gieskes W.W.C. Monitoring UV-B induced DNA damage and subsequent repair in a marine diatom u s i n g i m m u n o f l u o r e s c e n t t h y m i n e dimer detection. J. Phycol. submitted. Buma, A.G.J., Zemmelink, H.J., Sjollema, K.A., and Gieskes, W.W.C., 1994. Effect of UV-B on cell characteristics of the marine diatom Cyclotella sp. Proceedings First European Conference on Environmental UV Effects, Nov. 1993, Munich, Germany. (In press). Health Council of The Netherlands, 1994. "UV Straling uit Zonlicht", Report 1994/05. (The report will soon be available in English). Maduro R.A. and R. Schauerhammer, 1992. The Holes in the Ozone Scare, 21st Century Science Associates, Washington, D.C. Madronic, S. and de Gruijl, F.R. 1993. Skin cancer and UV radiation. Nature 366: 23.
988 Rozema, J., Lenssen, G.M. and Staaij, J.W.M. van der, 1990. The combined effect of increased CO2 and UV-B radiation on some agricultural and salt marsh species. In: J. Goudriaan, H. van Keulen and H.H. van Laar (eds). The greenhouse effect and primary productivity in European agro-ecosystems, Pudoc Wageningen, p. 68-71. Rozema, J., Staaij, J. van de, Costa, V., Torres Pereira, J., Broekman, R., Lenssen, G. and Stroetenga, M., 1991. A comparison of the growth, photosynthesis and transpiration of wheat and maize in response to enhanced ultraviolet-B radiation. In: Abrol., V.P. et al. (eds.). Impact of global climatic changes on photosynthesis and plant productivity. Asia publishing house, Sittingbourne, U.K., p. 163-174. Rozema, J., 1993. Plant responses to atmospheric CO2 enrichment: interactions with some soil and atmospheric conditions. Vegetatio 105/105: 173-190. Rozema, J., Lambers, H., van de Geijn, S.C. and Cambridge, M.L., 1993. CO2 and Biosphere. Kluwer, Dordrecht, p. 484. Rozema, J., Tosserams, M., van de Staaij, J., Magendans, E. Nelissen, H. and Brouwer, T., 1994. Impact of enhanced solar UV-B radiation on plants from terrestrial ecosystems. Proceeding EEG Workshop UV-B-Mfinchen, GSF. SCOPE, 1993. Effects of increased ultraviolet radiation on global ecosystems. Proceedings of a workshop arranged by the Scientific Committee on Problems of the Environment (SCOPE), Tramariglio. Sardinia. p. 47. Slaper et al., 1992. Ozone Depletion and Skin Cancer Incidence: An Integrated Modelling Approach. Report nr. 749292991, National Institute of Public Health, Bilthoven. S&PA and HCG Science Policy Associates, Inc. and Holland Consulting Group, 1992. Evaluation of the Technical Emphasis, Policy Relevance and Management Performance, p. 24. Staaij, J.W.M., Huysmans, R., Ernst, W.H.O. and Rozema, J., 1994. The effect of elevated UV-B (280-320 nm) radiation levels on Silene vulgaris: a comparison between a highland and a lowland population. Environmental Pollution (in press). Staaij, J.W.M. van der, Lenssen, G.M., Stroetenga, M. and Rozema, J., 1993. The combined effects of elevated CO2 levels and UV-B radiation on growth characteristcs of Elymus athericus. Vegetatio 104/105: 433-440. Steeneken, S.F., Buma, A.G.J. and Gieskes, W.W.C. Changes in transmission characteristics of polymethylmethacrylate and cellulose(III)acetate during exposure to ultraviolet light. Photochem. & Photobiol., submitted. Sontag, Y., Gurssen,J., de Gruijl, F.R., Van der Leun, J.C., Van Ulster, W.A. and Van Loveren H., 1994. Ultraviolet-radiation-induced impairment of the early initiating and the late effector phases of contact hypersensitivity to picrylchloride: regulation by different mechanisms. J. Invest. Dermatol. 132: 923-927. Tosserams, M. and Rozema, J., 1994. Effects of supplemental UV-B radiation (280-320 nm) on the dune grass land species Calamagrostis epigeios. Environmental Pollution (in press). UNEP, 1994. Assessment on Environmental Effects of Ozone Depletion. Van Hannen E.J., Buma, A.G.J., and Gieskes, W.W.C., 1993. Immunochemische detectie van schade in mariene microalgen geinduceerd door ultraviolet licht (UV-B): evaluatie van een nieuwe methode in ecosysteemonderzoek. NRP Report.
989 Van Liere, L. and Gieskes W.W.C., 1993. Effects of increased UV-B radiation on structure and function of algal communities in different climatic zones: risks of a global decrease in stratospheric ozone. Interim Report NRP Project 851054. Van Hannen E.J., Buma, A.G.J., Roza, L., Veldhuis, M., and Gieskes, W.W.C. Measurement of thymine dimers in phytoplankton on a single cell basis using a monoclonal antibody. Proceedings First European Conference on Environmental UV Effects, Nov. 1993, Munich, Germany (in press). Veen, A., 1992. Experimental analysis of phytoplankton production as affected by enhanced UV-B radiation. In: Annual Scientific Report 1991 of the National Institute of Public Health and Environmental protection, Bilthoven, p. 226-229. Veen, A., 1993. Effects of increased UV-B radiation on phytoplankton. A preliminary study and research programme. RIVM-report 731054001. Veen, A., 1994. Effekten op aquatische ecosystemen. In: Milieurapportage 1993 II. Integrale Rapportage Aantasting Ozonlaag en Blootstelling UV. RIVM, Bilthoven, p. 22-23. Veen, A., 1994. The effect of chronic and acute exposure to ultraviolet-B radiation on the daily oxygen production of Selenastrum capricornutum (Chlorophyceae): A continuous culture study using computer-controlled light regime. Proceedings First European Conference on Environmental UV Effects, Nov. 1993, Munich, Germany, (in press). Veen, A., 1994. Transmissie van UV straling onder water. In: Milieurapportage 1993 II. Integrale Rapportage Aantasting Ozonlaag en Blootstelling UV. RIVM, Bilthoven. p. 19. Yu, W., Teramura, A.H. and Sullivan, J.H. 1991. Model YMT-6 UV-B modulation system, manual of operation, final raport to US EPA, Corvallis, Oregon, USA.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
993
Effects of enhanced UV-B radiation on structure and function of phytoplankton communities A. Veen a and A.G.J. Buma b
aNational Institute of Public Health and Environmental Protection, Laboratory of Water and Drinking Water Research, P.O. Box 1, 3720 BA Bilthoven, The Netherlands
bState University of Groningen, Department of Marine Biology, P.O. Box 14, 9750 AA Haren, The Netherlands
Abstract In a research project aimed at identifying the primary targets and effects of UV-B radiation on microalgae, it was concluded that most research on this subject did not meet the criteria necessary for a quantitative assessment for the effects of ozone related UV-B enhancement. Therefore a new experimental approach was developed. Experiments on long-term effects were carried out under simulated natural light conditions. A reduced growth rate accompanied by an increased cellular biomass and size proved to be the main effect of UV-B exposure. By immunofluoresent labeling and flow-cytometry, specific UV-B induced damage could be detected in the DNA of the studied organisms.
A global decrease in ozone concentration has been detected over the last 15 years. This depletion, which is most distinct during the Antarctic spring (" ozone hole"), raises concern about the negative effects of the resulting increase in UV-B (290-315 nm) radiation on aquatic ecosystems. To assess the impact of enhanced UV-B radiation on the aquatic environment most research is logically focused on the phytoplankton compartment. By virtue of its capacity to store solar energy (photosynthesis) phytoplankton forms the first level of the aquatic trophic structure. Furthermore, because of its photosynthetic activity and formation of calcite (coccolithophores), marine phytoplankton is an essential link in global carbon dioxide cycles. Currently, 40% of fossil fuel carbon dioxide is assumed to be stored in the oceans. Dimethylsulfide (DMS) production from marine phytoplankton blooms is the major source of cloud-condensation nuclei over the oceans. The amount of nuclei determines the albedo of clouds and thus the earth's radiation budget. The role of phytoplankton raises two major questions: firstly, does enhanced UV-B radiation significantly affect climate feedback mechanisms and, secondly, what is the final impact on the
994 wind ~
% Io
D A
0.) r
=
A
9~
K
R
K
o~..4
~,,
,..........
,
,
,
time (h)
time (i)
Figure 1. Schematic presentation of natural light conditions as experienced by algal cells (solid line = PAR-, dotted line = UV-B radiation). Left graph shows surface irradiances. Center graph shows how algae travel through a light gradient (light intensities as % of surface irradiance [Io]). Right graph shows an example of the net effect of vertical transport on the fight regime experienced by algae. aquatic food chain. To answer both these questions, the primary targets and effects of enhanced UV-B radiation have to be defined. Research during the last decades has revealed overwhelming evidence for negative effects of enhanced UV-B radiation on algae. Effects on almost all metabolic processes have been described. However, the scientific merit of most findings is limited. The majority of UV-B effect studies on phytoplankton, both in the field and laboratory, has addressed short-term (hours) effects of acute exposure only. Furthermore, attention is directed mainly towards photosynthetic activity. Within a time frame of several hours neither DNA damage nor processes like induction of adaptation and repair mechanisms are likely to become manifest. As a matter of fact, the only prediction of the effects of enhanced UV-B radiation on Antarctic primary production is based on short-term photosynthesis experiments performed during the Antarctic spring [ 1]. Effects are rarely studied after a prolonged (several days) exposure period. Furthermore, most effects of UV-B radiation are measured in laboratory cultures under relatively high exposure rates. Both experimental time scales and appliedexposure levels hamper any extrapolation to field conditions. Nevertheless, field experiments indicate that even unaffected 86-
2;
g(440)=0.1 g(440)=0.05 g(44o)=o.o2 g(44o)=o
42
i
o
4
10
Chlorophyll-a (mg.m -3 ) Figure 2. Calculated relative attenuation of biologically effective UV-B radiation (Setlow DNA~3oo~) for marine systems. The relative attenuation is given as the ratio to the attenuation of DNA effective UV-B radiation (KDNA) to the attenuation of visible light (KpAR) for four concentrations of humic acids. Concentrations are given as the absorption coefficient at 440 nm (g44o). Most open oceans have a g44oof about 0.02 and a chlorophyll-a concentration of about 1 mg.mL
995 levels of UV-B radiation form a natural stress on phytoplankton communities. It was roughly estimated that if no UV-B radiation was incident at the earth's surface, phytoplankton primary production would increase by about 12% [2]. The assessment of UV-B effects for climate change studies is also complicated by the temporal variation of the damaging radiation and additional effects of UV-A and visible light in determining the final effect level. Under natural conditions algal cells are rarely stratified in the water column. Generally, algae are transported up and down the water column by wind-induced vertical mixing of surface waters (Fig. 1). By travelling through the underwater light gradient, algae will experience a more or less fluctuating light regime. Vertical transport of 10 metres might take from about 0.5 h to more then 10 hours. This dynamics will to a large extent determine the final effect level. As algae need visible light (PAR) for growth and repair of UV-B damage, the ratio of the attenuation of UV-B to PAR determines the effective exposure. Detritus and algae can decrease the ratio in attenuation of about 6 as determined by pure water to a minimum of about 2.5 (Fig. 2). However, the concentration of humic acids is the critical factor in determining UV-B transmission under water. Humic acids selectively absorb UV radiation.
cv wol3o5!
SM
SM
KG3
3530"7,= 25 exO
~D
~. 2 0 15~: 1 0 -
,~.~
-
|
~
.92" ~ . 8 4
UV12'
| |
',
a
~--~
10.64
v
IF
[
Figure 3. Schematic drawing of the culture system consisting of a fiat rectangular culture vessel (CV), two Venetian blinds (VB) controlled via stepper motors (SM), a highintensity PAR light source (HMI), a UV-B light source consisting of five fluorescent tubes (UV-12), a UV cut-off filter set (Schott WG305), a water-cooled blocking f'dter (Schott KG3) and a PC with interface (IF).
0.05
1.26
3.64
DNA eff(300) (KJ.day- 1)
Figure 4. Steady-state cellular dry weight of Selenastrum capricornutum in continuous culture. Open circles refer to dry weights at the start of the light period and closed circles to those at the end of the light period. Bars represent the standard deviations (n = 3). Culture conditions: L:D = 12"12 h, PAR~x = _ 550 larnol.m2.s 1, t = 18~ light regime is sinusoid as in Fig. 1.
We developed a special culture system, to simulate natural light dynamics as shown in Figs. 1 and 2. The system was characterized by a computer-controlled dynamic light system and the flat geometry of the culture vessel, enabling accurate dosimetry (Fig.3). Light intensities were regulated by angular slat displacements of the Venetian blinds. Continuous culture technique was applied to investigate long-term effects. The system and the quantification of the exposure levels have been described in full detail elsewhere [3]. In our experiments we aimed at long-term (several days) effects. Experiments performed so
996 1000
1000= t,.;
lOO
o
6~
100
O
o
10
~
1
.i
GI 0
20 40 D N A c o n t e n t (r.u.)
60
[--.
10
1 0
'
20
'
'~0
'
60
D N A c o n t e n t (r.u.)
Figure 5. Bivariate distribution of DNA content versus the amount of thymine dimers (both in relative units [r.u.]) in a population of Cyclotella sp. as detected by flow-cytometry. A: control, no UV-B exposure; B" exposed to 3 KJ.m 2 UV-B (Setlow DNAaf~3oo)). G 1 and G2 (cells with a double amout of DNA) refer to the different cell phases. far have given a conclusive picture of long-term effects of UV-B radiation. For several species tested we observed a decrease in growth rate coupled to an increase in cell weight and size after prolonged exposure ("steady-state" conditions) (Fig.4). This increase in cell size was caused by an increase in at least three major cell components (proteins, carbohydrates and pigments) [3]. These results do not match with the idea in which the photosystem is seen as the primary target. It does, however, fit a model in which DNA damage is the main target. Using a newly developed antibody labelling method [4], we were able to demonstrate the formation of thymine dimers in the DNA (Fig. 5). These dimers caused an arrest in the cell cycle (S phase) by inhibiting DNA replication. Through our recent research we arrived at a point at which we understand the primary targets and effects of enhanced UV-B radiation. Like damage to the photosystem, DNA damage is also a crucial factor. Obviously, estimations based on short-term effects will have to be reevaluated. Furthermore, changes in energy transfer to higher trophic levels are to be expected. Algal grazing by herbivorous predators (zooplankton) is directly dependent on cell size. Sinking rates might also be affected, resulting in a proportional increase in vertical carbon fluxes from the surface layers of the water. Fairly accurate methods are available to estimate UV-B transmission in different water types (Fig.2). Actually, only now can we direct research to collecting data for an accurate assessment of the effects of enhanced UV-B radiation on phytoplankton communities on a global scale.
References
1 Smith, R.C., B.B. Pr6zelin, K.S. Baker, R.R. Bidigare, N.P. Boucher, T. Coley, D. Karentz, S. Maclntyre, H.A. Matlick, D. Menzies, M. Ondrusek, Z. Wan and K.J. Waters. Science 255 (1992) 952. 2 Worrest, R.C. Physiol. Plant. 58 (1983) 428. 3 Veen, A., M. Reuvers and P. Ron~ak. J. Phycol. in press. 4 Buma, A.G.J., E.J. van Hannen, L. Roza, M.J.W. Veldhuis and W.W.C. Gieskes. J. Phycol. in press.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
997
I m p a c t of e n h a n c e d solar UV-B radiation on plants from terrestrial ecosystems Jelte Rozema, Marcel Tosserams and Erwin Magendans Department of Ecology and Ecotoxicology, Faculty of Biology, Vrije Universiteit, De Boelelaan 1087, 1081 HV Amsterdam, The Netherlands
Abstract The methodology of a UV-B supplementation and UV-B filtration system is described. The response to enhanced UV-B of terrestrial plant species ranges from very sensitive to tolerant or even a positive response; crop plant species tend to be more sensitive to elevated UV-B than wild plant species. Enhanced UV-B is expected to cause changes of both structure (competitive relationships, species composition) and functioning (plant-herbivore relationships, process of decomposition).
1. I N T R O D U C T I O N 1.1. Background Decrease in stratospheric ozone over the last decade has been revealed by satellite measurements. As a result a gradual increase in solar ultraviolet-B radiation occurs. Measurements of UV-B radiation in the Swiss Alps indicate an annual increase of UV-Bradiation of 1% since 1981 (Blumthaler and Ambach, 1990).
1.2. Solar UV-B and terrestrial ecosystems Solar UV-B radiation has important biological and ecological effects. Many crops and natural plant species demonstrate reduced growth, under enhanced UV-B radiation but there is a wide range of sensitivity to UV-B radiation between plant species and cultivars. A major part of research of UV-B effects relates to crop species and only a few UV-B effect studies are known of plant species of natural ecosystems (SCOPE 1992). Most studies of UV-B effects on plants have been conducted in greenhouse or in controlled environment cabinets with UV-B lamps. Relatively low levels of Photosynthetic Active Radiation (PAR), in controlled environment studies may prevent induction of photo-repair of UV-damage. This has led to an overestimation of growth reduction of plants exposed to enhanced levels of UV-B. Outdoor experimental UV-B supplementation systems, with natural levels of PAR provide more realistic responses of plants to enhanced UV-B.
1.3. Scientific aims Our research aimed at development and application of outdoor UV-B supplementation and UV-B filtrations systems and to assess effects of solar UV-B radiation to plant species of terrestrial ecosystems.
2. M E T H O D O L O G Y 2.1. UV-lamp systems In experimental studies with enhanced levels of UV-B irradiance UV-lamps are used that are pre-burnt and filtered with cellulose acetate as a cut off filter of wavelengths smaller than 280 nm. The cellulose acetate filters are renewed twice a week to avoid reduced transmission of 1IV-
998 cover UV-B lamps over control treatment that receive no UV-B irradiance, but where the UV-B levels are the same as in the treatment that receive no with enhanced UV-B (see Figure 1).
Fig. 1. Philips TL12/40 lamps are applied both in indoor and outdoor studies of UV-B effects on plants. The plots shown in the photograph, which are exposed to solar UV-B + UV-B supplied by the lamps, refer to monocultures and mixed cultures of the dune grassland species Calamagrostis epigejos and Holcus lanatus. Thus the effect of enhanced UV-B on competitive relationships between plant species is analysed. We use Philips TL 12/40 lamps both in indoor controlled environment studies and in outdoor studies in an experimental field and in natural ecosystems. Two outdoor unenclosed systems are applied: 1. UV-B Supplementation system and 2. UV-B filtration systems.
2.2. UV-B Supplementation system A lamp system to supplement ambient solar UV-B over experimental plots and over natural vegetation has been developed and applied (Fig. 1). At present this outdoor UV-B supplemenation system is applied in a "square wave" mode of enhanced UV-B radiation. This implies switching UV-B lamps on for a fixed period around midday, when natural solar UV-B is greatest. Currently an outdoor solar tracking UV-B supplementation is developed and installed. This involves continuous measurement of solar UV-B levels using appropriate UV-B sensors (Yu et al. 1991). This allows accurate simulation of enhanced UV-B throughout the day. There is evidence that plant responses to enchanced UV-B supplied in the solar tracking modulated mode differs both qualitatively and quantitatively from responses to UV-B supplied in the square wave mode. The development and testing of this solar tracking UV-B supplementation is in close cooperation with Dr. Andy McLeod, Institute of Terrestrial Ecology, Huntingdon, United Kingdom and Dr. Gaetano Zipoli, IATA-CNR, Florence, Italy.
999 UV-B lamps are mounted above the vegetation and are maintained at a constant height above the canopy. UV-B sensors will measure and track ambient solar UV-B radiation. Dual UV-B sensors are used beneath both treatment and control racks with UV-lamps, spaced such that one is unshaded from direct sunlight by lamps or frame. The unshaded sensor is selected by the control system preventing shading effects from interfering with feedback control of output of UV-lamps. The lamp output is adjusted to give a constant multiple of UV-B radiation above ambient UV-B radiation. Increases of UV-B radiation relative to ambient UV-B irradiance relating to 20% ozone depletion under clear sky are realized.
2.3. UV-filtration system Various types of solid plastic filters are installed above precultured plants in pots or above natural vegetation. These filters absorb either (a) little solar UV-B (and UV-A) (Acrylate filter), (b) all UV-B (Mylar polyester film on top of acrylate filters) and (c) all UV-A + UV-B (Lexan filter). The transmission spectra of these three filter types are given in Fig. 2. 1.0
0.20
0.8
0.15
"~ ~ 0.6
0,0 0.05
295
'
J
j
300
305
310
//,'t 315
320
0.4 .,4
...... A
9
0.2
....
L
M~ ..... no filter
0.0 10
300
320
340
360
380
400
Wavelength (nm) Fig. 2. Solar UV-irradiance (August 1992, 13.00 h), on a clear sky day, beneath the various filter types (A = acrylate, L = Lexan, M = Mylar), compared to full sunlight (no filter). Measurement with a Optronic spectroradiometer (OL 752, Optronics Laboratories, Florida, USA). The development and testing of low-cost solid plastic UV-B filters allow outdoor and ecosystem studies of reduction of ambient levels of solar UV-B radiation. In practice it will allow research of responses of plant from natural terrestrial ecosystems (Fig. 3).
1000
Fig. 3. Application of solid plastic UV-B filters in the study of reduction of ambient solar UV-B radiation on plant species of a dune grassland ecosystem. The height of the plastic UV-B filters can be adjusted according to seasonal development of the vegetation. At the background a rack of UV-lamps over the dune grassland vegetation is shown as a UV-B supplementation system.
3. R E S P O N S E S OF T E R R E S T R I A L PLANTS TO E N H A N C E D S O L A R UV-B IRRADIATION
Solar UV-B effect studies have been performed on a number of crop and native plant species both in controlled environment (greenhouse), unenclosed outdoor (experimental garden) and field studies in a dune grassland ecosystem (Tables 1 and 2). A wide range of sensitivity to enhanced UV-B measured in indoor controlled environment experiments, exists among crop species. Cultivars of crops such as soybean (Glycine max) vary greatly in sensitivity to UV-B. Of the natural plant species from Dutch coastal ecosystems: the salt marsh ecosystem and dune grassland ecosystem, no apparent sensitive plant species were found both in indoor and outdoor studies. It should be noticed that no natural plant species of the Leguminosae (a plant group with many plant species sensitive to UV-B) have been subjected to enhanced UV-B radiation levels in outdoor experiments. The woodland understorey species Alliaria petiolata indicates to be sensitive to enhanced UV-B. It may be that plant species occurring in open habitats (Calamagrostis epigejos, Spartina anglica) and being exposed to relatively high levels of ambient solar UV-B have evolved UV-B adaptation mechanisms. Other species occurring in the woodland understorey with reduced PAR and reduced ambient UV-B, seem to be relatively UV-B sensitive. This hypothesis is studied in current research.
1001 Table 1. Survey of qualitative plant responses to enhanced UV-B in indoor controlled environment studies. Daily biologically effective dose of UV-B relating to 20-50% ozone depletion under clear sky. Response to enhanced UV-B negative response very sensitive sensitive
no significant response
positive
Crop species Pisum sativum Phaseolus vulgaris Vicia faba Lycopersicon esculentum Cucumis sativus Triticum aestivum Zea mays
II II II II
Natural plant species Verbascum thapsus Calamagrostis epigejos Plantago lanceolata Alliaria petiolata Aster tripolium Elymus athericus Spartina anglica Holcus lanatus Silene vulgaris
II II II II II II II II II
The reader is referred to Rozema (1993) for more detailed information.
Table 2. Survey of qualitative plant responses to enhanced UV-B in outdoor studies. Daily biologically effective dose of UV-B relating to 20-50% ozone depletion under clear sky. Response to enhanced UV-B negative response no significant very sensitive sensitive response
positive
Crop species Vicia faba Triticum aestivum Zea mays
II II II
Natural plant species Calama grostis epigejos Plantago lanceolata Urtica dioica Holcus lanatus Verbascum thapsus Silene vulgaris
II II II II II II
The reader is referred to Rozema (1993) for more detailed information.
1002 4. E C O L O G I C A L E F F E C T S O F O Z O N E D E P L E T I O N ; I M P L I C A T I O N S ENVIRONMENTAL POLICY
FOR
1. Responses of terrestrial plant species to enhanced UV-B range from very sensitive to tolerant or even a positive response. This broad range type of response possibly reflects the intrinsic heterogeneity of the various terrestric ecosystems. Grassland ecosystems are much more exposed to full solar UV-B than under storey plants from forest ecosystems. 2. Adaptation to enhanced UV-B may consist of two types of mechanisms. a. UV-B tolerant plant species avoid enhanced UV-B radiation flux at sensitive metabolic sites by absorption of UV-B by epidermal structures and dissolved, UV-B absorbing compounds such as flavonoids. b. UV-B tolerant plant have effective UV-B repair mechanisms, in contrast with UV-B sensitive plant species. The nature and occurrence of these adaptation mechanisms has been analysed only for a few plant species. Understanding of UV-B repsonses of all groups of terrestrial plant species needs research priority. 3. Crop plant species tend to be more sensitive to enhanced UV-B than wild plant species. This may indicate that during the breeding procedure domesticating wild species to crop plants adaptation to (high) UV-B has been lost. Maybe a trade off exists between high productivity and sensitivity to UV-B. This is an intriguing hypothesis. It might indicate that plant species may loose (genetically determined) adaptations to solar UV-B. This implies that certain crops or certain cultivars of crops will be less productive with increasing solar UV-B. It is unknown whether selection for increased tolerance to enhanced UV-B will be successful and over what time period. 4. There are many uncertainties in the knowledge of UV-B effects on plant life in terrestrial ecosystems: effects of UV-B on reproductive biology are largely unknown, some of the reproductive features will be vulnerable to enhanced UV-B. Little is known of UV-B effects on "lower plants" like mosses ferns, fungi and algae. The effect of an enhanced UV-B flux is unknown in a "multiple stress" environment such as UV-B air pollution; UV-B x nutrient and water deficiency There are some indications that enhanced UV-B will cause increased occurrence of plant diseases by fungi, bacteria and viruses. 5. Consequences of enhanced solar UV-B to terrestrial ecosystems are expected to be substantial: there will be a shift to dominance of UV-B resistant plant species the altered biochemistry of plant species exposed to high UV-B fluxes may not only affect plant-herbivore relationships but also change the process of decomposition of leaf litter (Figure 4). -
5. N A T I O N A L AND I N T E R N A T I O N A L C O O P E R A T I O N National and International contacts. Between the different national UV-B research groups exchange of experimental results and methods occurs. Every six months a meeting is organized by one of the participating groups. During these meetings results are presented and discussed. Close collaboration between different groups exists on assessing the impact of UV-B on DNA Damage. Immunological
1003
TERRESTRIAL ECOSYSTE MS UV-B
PAR
y
NUTRIENTS
PLANT LEAFANGLE CANOPY ARCHIITECTURE LAI RGR BIOMASS PROD.
UV-B ABSORBING COMPOUNDS SECUNDARYMETABOLITES
PRIM. PROD.
UV-B
-I
DAMAGE PHOTOREPAIR
/
2
DEFENSEAGAINSTI HERBIVORY GRAZING SEC. PROD. DECOMPOSITION
LEAFLONGEVITY SPROUTING BUDBURST
COMPETITIVEABILITY
FIG.4. Conceptual model structure Impact of Enhanced Solar UV-B on Terrestrial Ecosystems: Physiology, Functioning and Dynamics. Ecosystem Functions: 1. Prim. Prod.= Primary productivity: 2. Sec. Prod. = Secundary Productivity: 3. Decomposition. Ecosystem structure: 4. Competitive ability. PAR = Photosynthetic Active Radiation; Pn = Net Photosynthesis; RGR = Relative Growth Rate: LAI = Leaf Area Index.
techniques used in skin cancer research are now also applied to assess DNA damage in marine algae and terrestrial plants. Contacts exist with research groups in Germany (prof. M. Tevini, Karlsruhe) and the United States (prof. A.H. Teramura and Dr. J. Sullivan, Univ. of Maryland, Washington), who are well known for their work on UV-B effects on terrestrial plants. Results were presented and discussed during visits to these research groups in 1992, 1993, 1994.
5.1. National cooperation 1. 2.
Dr. L. van Liere, Dr. A. Veen, RIVM, UV-B and aquatic Ecosystems. Dr. W. Gieskes, Dr. A. Buma, RUG, DNA-UV-B-damage, UV-B dosimetrie.
5.2. International cooperation 3. 4. 5.
Dr. A. McLeod, ITE, Huntingdon, U.K., Prof.Dr. M. Tevini, Dr. J. Ros, TU, Karlsruhe, Prof.Dr. A.H. Teramura, Un. Maryland & Hawai, Dr. J. Sullivan, USA,
5.3. International workshop UV-B and Biosphere The groups involved in this project and in the project on UV-B and algal communities (nr. 851054) will organize an international workshop on the influences of UV-B radiation on terrestrial and aquatic ecosystems. That will be held towards the end of 1995.
1004 Sponsors: the Dutch Ministry of Housing, Physical Planning and the Environment, the Dutch Organisation for Scientific Research, the Royal Dutch Academy of Sciences and the European Community.
6. R E F E R E N C E S
1
M. B lumthaler and W. Ambach, 1990. Indication of increasing solar ultra-violet-B radiation flux in Alpine regions. Science 248,206-208. 2 J. Rozema, Plant responses to atmospheric CO2 enrichment: interactions with some soil and atmospheric conditions. Vegetatio 105/105, 173-190 (1993). 3 J. Rozema, G.M. Lenssen and J.W.M. van der Staaij, The combined effect of increased CO2 and UV-B radiation on some agricultural and salt marsh species. In: J. Goudriaan, H. van Keulen and H.H. van Laar (eds). The greenhouse effect and primary productivity in European agro-ecosystems, Pudoc Wageningen, pp. 68-71. (1990). 4 J. Rozema, J. van de Staaij, V. Costa, J. Torres Pereira, R. Broekman, G. Lenssen, and M. Stroetenga, A comparison of the growth, photosynthesis and transpiration of wheat and maize in response to enhanced ultraviolet-B radiation. In: V.P. Abrol. et al. (eds.). Impact of global climatic changes on photosynthesis and plant productivity. Asia publishing house, Sittingbourne, U.K. pp. 163-174 (1991). 5 J. Rozema, H. Lambers, S.C. van de Geijn and M.L. Cambridge, CO2 and Biosphere. Kluwer, Dordrecht, pp. 484 (1993). 6 J. Rozema, M. Tosserams, J. van de Staaij, E. Magendans, H. Nelissen, and T. Brouwer, Impact of enhanced solar UV-B radiation on plants from terrestrial ecosystems. Proceeding EEG Workshop UV-B-Mtinchen, GSF (1994). 7 SCOPE, Effects of increased ultraviolet radiation on global ecosystems. Proceedings of a workshop arranged by the Scientific Committee on Problems of the Environment (SCOPE), Tramariglio. Sardinia. pp. 47 (1993). 8 J.W.M. van der Staaij, G.M. Lenssen, M. Stroetenga and J. Rozema, The combined effects of elevated CO2 levels and UV-B radiation on growth characteristcs of Elymus athericus. Vegetatio 104/105,433-440 (1993). 9 J.W.M. Staaij, R. Huysmans, W.H.O. Ernst and J. Rozema, The effect of elevated UV-B (280-320 nm) radiation levels on Silene vulgaris: a comparison between a highland and a lowland population. Environmental Pollution (1994.). 10 J.W.M. van de Staaij, Enhanced solar ultraviolet B radiation: consequences for plant growth. Ph.D. Thesis, Vrije Universiteit, Amsterdam, (1994) 130 pp. 11 M. Tosserams and J. Rozema, Effects of supplemental UV-B radiation (280-320 nm) on the dune grass land species Calamagrostis epigeios. Environmental Pollution (1994). 12 W. Yu, A.H. Teramura and J.H. Sullivan, Model YMT-6 UV-B modulation system, manual of operation, final raport to US EPA, Corvallis, Oregon, USA (1991).
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1005
U V B c a n a f f e c t t h e i m m u n e s y s t e m r e s u l t i n g in d e c r e a s e d r e s i s t a n c e to infections and tumors. J. Garssen a, W. Goettsch a, Y. Sonntag b, F. de Gruijl b and H. van Loveren ~. aLaboratory for Pathology, National Institute for Public Health and Environmental Protection (RIVM), P.O.Box 1, 3720 BA Bilthoven, The Netherlands bAcademic Hospital Utrecht, Department of Dermatology, P.O.Box 85500, 3508 GA Utrecht
Abstract As a result of depletion of the ozone layer by industrial waste compounds, all living organisms on the earth's surface may be exposed to increased amounts of UVB radiation. In man UVB radiation can cause deleterious effects on the skin and the eyes. During the last years it has become clear that UVB can affect the immune system also. Hence, UVB may affect the resistance against infectious diseases. It is demonstrated that UVB can inhibit the immune response against skin-associated infectious diseases. However, recently it became clear that UVB can also induce immunosuppression at other loci than the exposed skin. Hence, also the immunological resistance against non-skin associated infectious diseases and tumors can be affected. Recently we demonstrated that low doses of UVB can induce immunosuppression in rodents and man. This suppression was not restricted to the exposed skin. Finally we demonstrated that this immunosuppression leads to a significant suppression of the resistance against non-skin associated infections in the rat. It is noteworthy that the resistance against bacterial (Listeria monocytogenes), viral (cytomegalo virus) as well as parasitic (Trichinella spiralis) infections was inhibited and that this inhibition was correlated to a suppression of the cellular immune system. Because these data demonstrate that low doses of UVB can affect the immune system in man and rodents and because animal studies showed that this immunosuppression inhibits the resistance to infections it is worthwile to analyse the risk for increased UVB levels with respect to infectious diseases in man.
1. I N T R O D U C T I O N
A decrease of the atmospheric (stratospheric) ozone layer, induced by the emission of CFC's, may result in an increased exposure of humans to ultraviolet radiation. Especially the exposure to wavelenghts between 280 and 315 nm (=UVB) will be increased due to
1006 ozone depletion. Aside of a beneficial effect like vitamin D production ultraviolet radiation can cause deleterious effects on human health. Most studies regarding the toxic effects of ultraviolet radiation are restricted to deleterious effects on the skin and eyes. From experimental studies it can be concluded that UVB radiation causes tumors by at least two separate mechanisms: 1) by genotoxic activity of UVB radiation and 2) by affecting the immunologically mediated resistance to tumors. Both in laboratory animals and in humans evidence has been obtained that UVB radiation can affect the immune system. Hence it is reasonable to expect that immunological resistance to infections may also be altered by UVB exposure. It is known that resistance against infectious agents that enter the body via the skin, such as Leishmania, and Herpes simplex, can be affected by UVB irradiation. Whether resistance to infectious agents that enter the body via other routes is affected was as yet not known. Because UVB radiation can induce systemic immunosuppression, suppression of immunity against non-skin associated infections (diseases), may also be suspected. The main purpose of this research project was to determine whether UVB radiation affects the resistance to several systemic non-skin associated infectious diseases and skin tumors in man. For this purpose several infection models and a tumor model in rodents were used. Since such studies cannot be performed in human volunteers, (limited) noninvasive studies in man (blood, biopsies), rat and mouse were needed in order to compare the effects of UVB radiation on various components of the immune system (basal immune parameters). If data on susceptibility differences are obtained using non-invasive tests for basal immune parameters, effects of UVB radiation on resistance against infections in rodents may be extrapolated to man. Finally, epidemiological studies are needed to validate the extrapolated data.
2. R E S U L T S
In non-invasive studies dependency of immunological changes on UVB dose and species (rat, mice, man) was investigated. In all species tested macroscopic and microscopic effects of UVB exposure were studied and compared. These studies demonstrated that the skin of rodents is more sensitive to UVB exposure than that of humans and that the difference in sensitivity in primary skin effects might be due to skin thicknesses. The pathological parameters were dependent on UVB exposure (intensitity and duration dependent). Based on these initial studies the effects of UVB exposure on several basal specific and non-specific immune parameters were investigated. The applied UVB doses were suberythemal (no reddening). Investigations with respect to changes in the number and subtype of immune competent cells in the skin as well as in lymphoid organs were performed and demonstrated that certain subtypes of lymphoid cells migrate to the skin and other subsets leave the skin. Recently, we demonstrated that lymphoid cells with UVBinduced DNA damage were detectable in lymph nodes indicating that systemic effects can occur after exposure to low UVB doses. In addition to these descriptive studies studies on immune function were carried out also. A detailed study on the UVB-induced systemic suppression of a bi-phasic contact hypersensitivity response was carried out to ascertain whether suppression of the early phase would directly lead to the suppression of the late phase (the classical delayed reaction). This appeared not to be the case. Several immune parameters such as T cell responsiveness (e.g. delayed type hypersensitivity) and natural
1007 killer cell function were significantly suppressed by exposure to low UVB doses. This was true for humans and rodents. Differences in susceptibility with respect to these parameters were less clear than with respect to the primary skin effect studies. The effects of UVB exposure on resistance against infectious diseases were studied in rats. Three different models for non-skin associated infectious diseases and one for a skinassociated infectious disease were used. We demonstrated that UVB exposure (suberythemal doses) can inhibit the immunological resistance against Trichinella spiralis infections in rat. Exposure to suberythemal doses of UVB after oral Trichinella infection leads to higher a number of parasites in the carcasses, indicating that the resistance was impaired. Recently, we demonstrated that the specific lymphocyte response to Trichinella was significantly inhibited. This points to an effect of UVB radiation on the cellular (T cell mediated) immune system. Because Trichinella spiralis infections still occur, e.g. in Eastern Europe, effects of UVB on this animal model may be relevant. Additionally, we demonstrated that suberythemal doses of UVB inhibited the resistance against a intraveneously Listeria monocytogenes infection. Especially the specific T cell response to this pathogen was inhibited by UVB exposure. Listeria infections are still a problem for human health especially for humans with a depressed immune system such as transplantation patients, babies and eldery persons. As was found for the parasitic and bacterial infection model also resistance to systemic cytomegalovirus infections was inhibited by UVB exposure of rats. Cytomegalovirus infections are still a health problem in immunosuppressed humans such as transplantation patients. For this reason this infection model in rats is very relevant for the human situation. In sum, resistance against several non-skin-associated infectious diseases that also occur in humans, especially in humans with a suppressed immune system such as in transplantation patients, can be inhibited by UVB exposure. Finally, we demonstrated that UVB radiation also affects the resistance against Herpes simplex skin infections. In man this disease is suggested to be evoked by UVB: e.g. cold sores in skiers and sun bathers. This latter infection model appears particularly suited to compare effects of UVB on resistance in humans and rodents. In addition to these infection studies we demonstrated in an extensive time lapse study in hairless mice that prior to macroscopically observable UVB-induced skin tumors, alterations in composition of lymph node cells and skin immunocompetent cells ocurred. Mice with these immune alterations became incapable of rejecting UVB-induced skin tumors whereas normal (non-exposed) mice rejected these transplants almost immediately. The early immunological changes observed did not coincide with, but amply preceded the time by which the mice started to accept the tumor implants. These findings confirmed and timed the UVB-induced immune alterations in the course of the induction of skin cancer in the hairless mouse model. Using all the obtained data with respect to UVB-induced alterations in several immune parameters (systemic and local) in rodents and man, and UVB-induced alterations in resistance against infectious diseases in rodents, a basis for risk estimation of exposure to elevated UVB levels is formulated. Risk estimation until now was only restricted to skin cancer, and did not explicitely taken into account effects on the immune system. Now, there are several lines of evidence that immune alterations play an important role in the induction of skin cancer and probably also in severity and incidence of infectious diseases.
1008 3. C O N C L U S I O N
From all the experiments that were done we conclude that low UVB doses, i.e. suberythemal doses, can affect the immune system. It is remarkable that this immunosuppression is not restricted to the exposed site (skin). Even in the spleen and lymph nodes immune alterations have been found using several different tests. The effects found were not species specific although the primary skin effects in humans tended to be less severe than in rodents. From animal experiments we conclude that the UVB induced alterations of the immune system play a significant role in the decreased resistance against tumors and infections in rats and mice. With respect to the infection studies it is demonstrated that UVB radiation can affect the resistance against skin-associated infectious diseases, such as the well known cold sores (i.e. Herpes simplex infections) as well as the resistance against non-skin-associated infectious diseases (e.g. oral parasitic, intraveneous bacterial and viral infections). The data obtained, viz. regarding immune effects in man and rodents, and regarding the resistance against infectious diseases and tumors in rodents, form a basis for a better assessment of effects on the incidence and severity of infectious diseases and tumors in UVB-exposed populations. In future these data will be used for calculation of risk from an ozone depletion. Until now, resistance against infectious agents and tumors can only be studied in rodents, and therefor the effects in humans have to be extrapolated from the animal studies. Nevertheless, there are possibilities now to study effects of UVB radiation on the resistance against certain infectious diseases in humans. Some of these infectious diseases are even thought to be related to the induction of skin cancer (human papilloma virus infections). Finally, suitable epidemiological data are needed to improve the human studies and validate the extrapolation of animal studies to man.
4. REFERENCES
~
Y. Sontag, J. Garssen, F.R. de Gruyl, J.C. van der Leun, W.A. van Vloten and H. van Loveren, Photochem. Photobiol., 53 (1991) 24s. W. Goettsch, J. Garssen, F.R. de Gruyl and H. van Loveren, Change, 9 (1992) 3-5. W. Goettsch, J. Garssen, F.R. de Gruyl and H. van Loveren, Thymus, 21 (1993) 93114. J. Garssen, H. van der Vliet, A. de Klerk and H. van Loveren, The Toxicologist 1, (1993) 107. Y. Sontag. F.R. de Gruyl, J. Garssen, H. Van Loveren, M. Kripke and W.A. van Vloten, J. Invest. Dermatology, 160 (1993) 464. J. Garssen, W. Goettsch, A. Deyns, H. van der Vliet, F. Pierik, F.R. de Gruyl and H. van Loveren, J. Invest. Dermatology, 160 (1993) 473. Y. Sontag. F.R. de Gruyl, J. Garssen, H. Van Loveren, M. Kripke and W.A. Van Vloten, Photochem. Photobiol., 57 (1993) 75s. Y. Sontag, F.R. De Gruyl, W.A. Van Vloten, J. Garssen, H. Van Loveren and J.C. van der Leun, Exp. Dermatol., 3 (1994) 145. J. Garssen, W. Goettsch, A. de Klerk, M. Herremans, F.R. de Gruyl and H. van Loveren, Photochem. Photobiol., 59 (1994) 12s.
1009 10. 11. 12. 13. 14. 15. 16. 17.
W. Goettsch, J. Garssen, A. Deijns, F.R. de Gruyl and H. van Loveren, Photochem. Photobiol., 59 (1994) 12s. W. Goettsch, J. Garssen, F.R. de Gruyl and H. van Loveren, Toxicol. Lett., 72 (1994) 359-363. M. Norval, A. E1-Ghorr, J. Garssen and H. van Loveren, British J. Dermatol., 30 (1994) 693-700. W. Goettsch, J. Garssen, A. Deijns, F.R. de Gruijl and H. van Loveren, Env. Health Persp., 102 (1994), 298-301. M. Hurks, J. Garssen, H. van Loveren and B.J. Vermeer. General aspects of UVirradiation on the immune system. Plenum press, in press, 1994. W. Goettsch, J. Garssen, F.R. de Gruijl and H. van Loveren. In: E.G. Jung, M.F. Holick (eds.), Biological effects of Light, Walter de Gruyter, Berlin, 1994, 637-641. Y. Sontag, J. Garssen, F.R. de Gruijl, J.C. van der Leun, W.A. van Vloten and H. van Loveren, J. Invest. Dermatol., 102 (1994), 923-927. W. Goettsch, J. Garssen, F.R. de Gruijl and H. van Loveren, Photochem. Photobiol., 60 (1994), 373-379.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 1995 Elsevier Science B.V.
1013
ASSESSMENT REPORT ON NRP SUBTHEME "ENERGY DEMAND AND SUPPLY OPTIONS TO MITIGATE GREENHOUSE GAS EMISSIONS"
W.C. Turkenburg Department of Science, Technology and Society Utrecht University Padualaan 14 3584 CH Utrecht The Netherlands
With contributions by: E. van den Heuvel H. van Zeijts
BTG, Biomass Technology Group BV, Enschede CLM, Center for Agriculture and Environment, Utrecht
D. Gielen, T. Kram, J.R. Ybema
ECN, Netherlands Energy Research Foundation, Putten
M.J. Lexmond, S. Nonhebel
LLrW, Agricultural University of Wageningen
J. de Beer, K. Blok, J. Farla
RUU, University of Utrecht
M. Gillissen, E. Tellegen
VUA, Free University of Amsterdam
1014 CONTENTS
Abstract 1. 0
0
0
0
0
7.
Introduction E n e r g y efficiency i m p r o v e m e n t 2.1. The database ICARUS 2.2. The long term potential of improving the energy efficiency 2.3. Energy conservation and investment behaviour of firms Material efficiency i m p r o v e m e n t and waste m a n a g e m e n t 3.1. The production, flow and consumption of materials in the Netherlands 3.2. Organized thrift in waste disposal 3.3. Controlled anaerobic treatment of waste and waste water S w i t c h i n g to renewables 4.1. Harvesting the sun using ecosystems 4.2. Conversion routes for energy crops 4.3. Can Dutch agriculture provide clean energy? Decarbonization of fuels and flue gases, and storage of CO2 5.1. Carbon dioxide recovery based on coal gasification 5.2. Chemical absorption and other recovery techniques 5.3. Carbon Dioxide recovery in the manufacturing industry 5.4. Storage of carbon dioxide in aquifers 5.5. Conclusions Integrated assessments: strategies for CO2 emission reduction 6.1. CO2 emission reduction in the Netherlands energy system 6.2. Integrated emission reduction of CO2 and non-CO2 greenhouse gases 6.3. CO2 emission reduction in the Netherlands energy and materials system 6.4. Conclusions References
ABSTRACT The direct and indirect consumption is responsible for more t h a n half the anthropogenic emission of greenhouse gases, especially. It might well be t h a t within 50 - 100 years countries like The Netherlands should reduce their CO2 emission with 80% or more. In principle many options can be developed and applied to reduce the CO2 emission. Focused on The Netherlands, the following ones are investigated within the Dutch National Research P r o g r a m m e on Global Air Pollution and Climate Change, phase I: energy efficiency improvement; material efficiency improvement and waste management; a shift to renewable energy sources, especially biomass; and decarbonization of fuels and flue gases. The
1015 research projects and the results achieved so far are described in this report. Also the set-up and results of an integrated assessment of options to reduce greenhouse gas emissions, especially CO2, are described. From this a s s e s s m e n t one might conclude t h a t technically it seems feasible to reduce the CO2 emissions in The N e t h e r l a n d s with about 70% between 1990 and 2030. The associated marginal costs are calculated at 250-400 NLG or less per ton CO2 avoided. Options t h a t are identified as 'robust' to severe CO2 constraints are: a) the application of energy saving measures, like t h e r m a l insulation in the residential and the commercial sector; b) the use of selected renewable energy options, like geothermal and solar heat, biogas, hydropower, wood power production and energy from wind turbines; c) the removal of CO2 from power plants and other processes, e.g. combined with hydrogen production; d) the application of heat pumps, for space heating and hot water supply in the residential and the commercial sector; e) the use of fuel cells, like for automotive purposes in the transport sector; f) the replacement of fossil fuel use by electricity, like in electric heat pumps; g) the use of hydrogen, as substitute for natural gas in stationary applications, in fuel cells and as alternative automotive fuel in vehicles and aircrafts. The implementation of these options requires a strong R, D&D effort as well as policies and instruments to stimulate the introduction.
1. I N T R O D U C T I O N
The present way in which we produce and use energy carriers is responsible for more t h a n half the anthropogenic emission of greenhouse gases (GHGs) to the atmosphere. The predominant gas is carbon dioxide (CO2). Due to the use of fossil fuels to satisfy our energy needs, the emission of CO2 has increased from the preindustrial level of zero to about 6 Gigaton Carbon per year (GtC/y) at present. Unless major policy changes occur, this amount might increase further to 20 or 30 GtC/y, maybe even higher, at the end of the next century. To reduce the risk of a climate change to a level that can be sustained, it might be necessary to reduce these emissions to less than 3 GtC/yr. For countries like The Netherlands this might well imply a reduction with more than 80%. Following this approach the worldwide emissions of CO2 between the years 1990 and 2100 have to be reduced with an accumulated a m o u n t of about 1200 GtC compared to 'business-as-usual' (IPCC Working Group I, 1990; 1994). To reduce the emission of C O 2 and other greenhouse gases a wide range of options can be applied. Important options in the energy sector are: a. Improvement of the energy efficiency, leading to a reduction of the energy consumption per unit of product or unit of activity. Many technologies to improve the energy efficiency are readily available. New developments can enhance the potential of this option further. b. Improvement of the material efficiency. In modern societies a considerable part of the p r i m a r y energy consumption is connected with the production of materials and consumer goods. Therefore material efficiency improvement,
1016 including a better treatment and management of waste, can reduce the energy demand and the emission of greenhouse gases. Development and application of renewable energy sources. The energy from the sun can be harnessed directly or indirectly and be utilized to replace fossil energy sources. It requires a further development and i m p l e m e n t a t i o n of options like biomass energy production, hydropower, solar energy conversion systems and wind turbines. A shift in the use offuels, from resources with a high carbon content (like coal) to resources with a lower or even no carbon content (like n a t u r a l gas and u r a n i u m ) , although aspects like the availability of resources a n d the acceptability of the conversion technologies involved m i g h t r e s t r i c t its applicability and possible impact. Table 1.1 Potential contribution of options to reduce CO2 emissions compared to 'businessas-usual'. Accumulated figures are shown for the period 1990 to 2100, based on a v a r i e t y of recently published a s s e s s m e n t studies. It should be noted t h a t the figures are mutually dependent - energy and material efficiency improvement application of renewables nuclear fission nuclear fusion fuel shill from coal to natural gas - CO2 recovery and storage - afforestation -
-
-
-
e.
f.
300 - 600 200 - 600 100-300 0- 25 0 - 300 100-300 50- 100
GtC GtC GtC GtC GtC GtC GtC
Decarbonization of fuels and flue gases, e.g. by recovering carbon dioxide from energy conversion processes and storing it outside the atmosphere. This option has begun to receive attention only recently. Several research and development programmes have been set up to get a better u n d e r s t a n d i n g of this option. Afforestation, to remove carbon dioxide from the atmosphere. This option is already applied by some utilities to compensate for CO2 emissions due to electricity production.
As a l r e a d y mentioned, it might be necessary to prevent the emission of CO2 b e t w e e n 1990 and 2100 with an a c c u m u l a t e d a m o u n t of about 1200 GtC compared to 'business-as-usual'. As shown in Table 1.1, the above mentioned options are probably able to reach this goal, provided t h a t enough attention is given to their further development and that implementation barriers are removed. For some of these options the potential and prospects for The N e t h e r l a n d s are subject of investigation in the Dutch National Research Programme on Global Air Pollution and Climate Change, phase I (NRP-I). An overview of the r e s e a r c h projects, togethers with the name of the project coordinators, the project n u m b e r and the area in interest, is given in Table 1.2.
1017 In this report the results achieved so far are presented.The research is focused on possibilities to improve the efficiency of our energy consumption, to reduce the inefficient use of materials and the production of waste, to utilize biomass and organic waste as an energy source and to sequester and store CO2 from energy conversion processes. The results are used in a study of integrated strategies for The Netherlands to reduce the emission of CO2 and other greenhouse gases in a cost-effective way. In this study the potential and costs of about 500 energy demand and supply technologies to mitigate the emission of CO2 and other greenhouse gases are assessed and mutually compared using the MARKAL model. Focused on the year 2030 and assuming a continued growth of GDP, the results of this study suggest that in principle it is possible to reduce the net CO2 emission in The Netherlands with at least 70%, with marginal costs below 400 NLG per ton of CO2 avoided, depending on the assumptions made (see Section 6). Table 1.2 List of projects in the NRP subtheme "Energy demand and supply options to mitigate GHG emissions" Title
Project leader
Number
Programming study Theme Sustainable Solutions
C.J.H. Midden/ 850031 C. Daey Ouwens/ E.H. Lysen
Energy and material use scenarios for the limitation of emissions of C02 and other greenhouse gases
W. Kram
850028
The long-term potential for improving energy efficiency K. Blok
853107
Energy conservation and investment behaviour of firms
J.B. Opschoor/ K. Blok
853095
Study on the possibilities of anaerobic fermentation process aimed at the reduction of greenhouse gases
G. Zeeman
852091
Organizing prevention in waste management
E. Tellegen
852090
Harvesting the sun energy using agro ecosystems
J. Goudriaan
853117
Biomass conversion routes for agriculture crops in Europe
H.E.M. Stassen
853108
Sustainability of production and use of biomass for European energy supply (phase I)
E.E. Biewinga
853109
Sustainability of production and use of biomass for European energy supply (phase IIA)
E.E. Biewinga
854143
1018 Management of 4 projects of SOP CO2 removal and storage
P.W. Renaud
Definition and control of the boundary conditions of the system studies in the SOP CO2
L.J.M.J. Blomen 851049
CO 2 recovery from industrial processes
K. Blok
851047
Investigations regarding the storage of CO 2 in aquifers in The Netherlands
F.C. Dufour
851048
Evaluation study of SOP C02 removal and storage
K. Blok
851046
851043
2. E N E R G Y E F F I C I E N C Y I M P R O V E M E N T
An important option to reduce C02 emissions is improving the energy efficiency in conversion and end-use processes. Sometimes it is stated that in countries like The Netherlands the ultimate potential of this option might be limited to several ten percents. In the scientific literature, however, it is claimed this potential is much larger, theoretically more than 80% or even 90%. In the literature it is also indicate t h a t a l r e a d y now m a n y technologies can be applied to improve the energy efficiency in a cost-effective way. If this would be true, it raises the question why these technologies have not yet been implemented. As the a n s w e r to these questions is of major importance for energy policy m a k i n g as well as the formulation of policies to mitigate greenhouse gas emissions, the NRP has decided to sponsor research to the potential of energy efficiency improvement as function of time and to barriers that hinder the implementation of energy conservation technologies. 2.1 T h e d a t a b a s e I C A R U S To achieve a better and concrete u n d e r s t a n d i n g of the potential of energy efficiency improvement between 1990 and 2000 and between 1990 and 2015, the NRP-I is sponsoring the development of a database called ICARUS-3 on the potential and costs of technologies to improve the energy efficiency between 1990 and 2000/2015 in all sectors of the Dutch economic system. Special attention is given to the industry. In this research the present and future consumption of fuels and electricity (assuming a frozen energy efficiency) is divided in sub-sectors and broken down into major energy consuming processes and unit-operations (e.g. drying, pumping, heating and lighting). Next, an identification is made of (nearly) available techniques to improve the energy efficiency. Then, based on the current efficiency of energy use, an estimation is made of the potential of energy efficiency improvement techniques that can be applied. Also an estimation is made of the associated investment and operation and maintenance costs. The results of this study are contained in a Quattro compatible spreadsheet (Blok et al., 1993). To improve the energy efficiency, so far more than 400 options have been identified (De Beer et al., 1993a, 1993b; De Beer et al., 1994). The cost-effectiveness of all these options can be presented in a supply curve. The supply curve derived in this study is presented in Figure 2.1 one for the period 1990-2000 and one for the period
1019 1990-2015. The result indicates a technical potential for efficiency improvement of 36% and an economic potential of 29% for the period 1990-2000. The results vary for different sectors. The technical potential for the period 1990-2000 is found to be as high as 42% in the service sector but 'only' 10% in the basic metal industry. For the period 1990-2015 the results show a technical potential of 56% - which is an increase with more t h a n 50% compared to the period 1990-2000 - and an economic potential of 43%. These figures suggest t h a t there exists a considerable potential to improve the energy efficiency, even when all currently available measures are implemented. It should be noted t h a t for the y e a r 2015 also m e a s u r e s were considered t h a t are not readily available, but could come available if adequate actions are taken.
501
40
| | 30"1
2000 Projected primary energy demand in the year 2000:3510 PJ in the year 2015:4915 PJ
I
2015
I I /
!
/
20
r/~ o ~9
~ 9
0
.................................
..................
-10-
r.~ -20-
-30-4050 ' 0%
43% ~
,
10%
20%
,
,
30% 40% C u m u l a t i v e s a v i n g (%)
!
50%
60%
Figure 2.1 Supply curves of energy efficiency improvement measures for the periods 19902000 and 1990-2015. On the horizontal axis the c u m u l a t i v e i m p r o v e m e n t potential is given as percentage of the projected energy demand without efficiency improvements. The European Renaissance scenario is used with physical growth figures.Vertically the specific energy efficiency improvement costs are depicted. The calculations are based on a low energy price scenario, (EZ, 1994). The curves show the results for a discount rate of 5%
2.2 The long term potential of improving the energy efficiency Assessing the potential in an even longer term, technologies should be considered t h a t are now in a very early stage of development. This brings along problems in analyzing options and gathering information. Co-funded byNRP-I, two research projects addressed the long term potential to improve the energy efficiency, both focused on the industrial sector. The first research, also sponsored by SYRENE research programme of NOVEM, is directed at making a preliminary survey of technologies t h a t might reduce the
1020 end-use demand of industrial process on the longer term (Smit et al., 1994). The technology descriptions are based on accessible literature, supplemented with data provided by experts on a specific sector or technology. The descriptions are divided into two sections. The first one gives a description of the reference technology and of the new, energy efficient technology, together with information about state-ofthe-art, ongoing R&D, and applicability of the technology. The second section gives preliminary data about economic and energetic parameters. In Table 2.2 some results of this research are presented. It must be emphasized that the results are based on a limited literature research. A more thorough analysis of the energy efficiency improvement potential is topic of the second research. The second research focuses on the development of a methodology to make a more accurate analysis of the long term energy efficiency improvement potential. The methodology developed so far starts with the determination of the m i n i m u m energy requirement to perform a certain energy function and of the energy losses associated with performing the energy function with the current technology. The question posed is, whether these losses can be reduced without changing the current technology. And if not, can we imagine technologies t h a t can reduce the energy losses. After having compiled a list of potential efficiency improvement technologies, an assessment is made of a possible future development of these technologies. A list of parameters that determine these developments is filled out, based on a review of the literature and a consultation of experts. A study following this line of research has been conducted for the sector 'paper and board' industry (De Beer et al., 1993c). Furthermore, two studies are underway for the sectors 'iron and steel' and 'cement' (De Beer, 1994). Here we present some results of the studies for the paper and board industry and the iron and steel industry. At present the average primary energy demand to make paper out of wood pulp is about 10 GJ/ton paper. Theoretically this figure can be much smaller. The operation with the largest energy losses appeared to be steam generation (in a CHP-unit or boiler). Steam is mainly required for drying of the paper against steam heated driers. Elimination of these losses is only possible if paper is made without the addition of water, but this would have large negative effects on the characteristics of the product. Therefore five other technologies were selected t h a t have the opportunity to reduce the energy losses. An assessment of the potential development of these technologies, resulted in the selection of two t h a t are most promising: condensing belt drying and impulse drying. These technologies have the potential to reduce the specific steam demand by 60%. For the iron and steel industry the study has not been finished yet. The specific energy requirement of an efficient steel plant like Hoogovens in The Netherlands is 19.7 GJ/ton crude steel. Between 1990 and 2000 a reduction to 16.8 GJ/ton is technically feasible. Thermodynamically, the m i n i m u m energy r e q u i r e m e n t to reduce iron oxide is only 6.2 GJ/ton. However, an exergy analysis of an integrated steel mill has revealed that the room to improve this process is limited. Larger i m p r o v e m e n t s seems only feasible when other production routes are chosen. Several options are investigated: an increased share of secondary steel making, advanced iron making processes (e.g. plasma processes), direct steel m a k i n g (inbath melting of iron, ore-to-powder steel making), near shape casting, and the use of hydrogen as reductant. The largest efficiency improvement can be obtained by
1021 an increased share of secondary steel making. Taking into account an improved efficiency of electric arc furnaces and a h i g h e r energy d e m a n d for scrap bonification, a specific primary energy demand for secondary steel making of about 7 GJ/ton steel seems achievable in the longer term. For primary steel making, by a combination of new technologies, the specific primary energy demand might come down to about 12 GJ/ton steel (see also: Van Wijk et al., 1994).
1022
Table 2.3 Selection of long term energy efficiency improvement technologies. All figures are based on the industrial situation in The Netherlands Currents SEC (GJ/ton) 1)
Long-term technology
Long-term SEC (GJ/ton)
Process/ application
Current technologies
Milk powder
Two-stage drier
3.95
Condi-cyclone
2.3
Paper
Steam heated cylinders
9.9
Impulse drying
7.8
Ethylene
Naphtha steam cracking
61 3)
Selective steam cracking
55
Cryogenic distillation
61 3)
Membrane separation
56
Chlorine
Membrane electrolysis
9.9 electricity 1.2 steam
Improved membrane electrolysis
7.5 electricity steam
Bricks
Roller kiln
2.2
Tunnel kiln
1.8
Cement
Standard kiln (dry process)
3.4
Fluidized bed
2.7
Iron making
Blast furnace
17.6 (ton pig iron)
Converted blast furnace
10.6
Steel casting
Casting and rolling
1.64 (ton crude steel)
Strip casting
0.24
Aluminum
Hall-Heroult
51.1 electricity 16.3 fuel
AlCoA
33.1 electricity fuel
2)
Inert cathodes and 42.1 electricity anodes fuel
C r o s s-c u t t i n g technologies
Current technology
Best practice efficiency
Long-term improvements
Long-term efficiency
Electric motors
AC induction motor
95.5% (range 40-100 kW)
E.g. permanent magnets; soft magn. materials; moreefficient fan.
98%
Combined generation of heat and power
Gas turbine with waste heat boiler, or combined cycle
High temp. applications
Furnaces
1) 2) 3)
Tie -- 3 4 - 3 6 %
Energy demand: 200 PJ/year
Increase turbine inlet temperature; improvement of gas turbine cycle. Heat recovery with ceramic recuperators
He "- a b o u t
50%
Energy demand: 170 PJ/year
SEC = specific energy consumption; in this table normally p r i m a r y energy is meant. Savings on s t e a m d e m a n d are 60%, but as electricity d e m a n d increases the savings on p r i m a r y energy are smaller. Including use of fuel as feedstock (43 GJ/ton)
1023 2.3 E n e r g y c o n s e r v a t i o n and i n v e s t m e n t b e h a v i o u r of firms A p a r t from s t u d y i n g the p o t e n t i a l , the NRP-I also i n v e s t i g a t e s the implementation of technologies to improve the energy efficiency, focused on the i n v e s t m e n t behaviour of firms. The question is addressed why firms do not implement measures that, according to the database ICARUS, are economically profitable. The project consists of two parts. In the first part the differences between ICARUS and the observed i m p l e m e n t a t i o n behaviour of firms are analyzed in t e r m s of d e t e r m i n a n t s and barriers to the adoption of Energy Efficiency (EE) technologies. This study should result in an empirically validated implementation model. In the second part of the project the impact of different policy i n s t r u m e n t s on the investment behaviour of firms with respect to EEtechnologies is assessed. The study should provide a simulation model by which the effectiveness of policy instruments to accelerate the adoption of EE-technologies by firms can be assessed. The project was started with an analysis of the literature on investment decision m a k i n g and on the application of investment theories to energy conservation measures. From this study the most i m p o r t a n t theoretical d e t e r m i n a n t s and barriers to the adoption of energy conservation technologies were derived. In a world without uncertainty about future states of events and cash flows, with free and full information, with independence between technologies, and with unlimited access to capital markets, a profit maximizing firm would implement all available technologies t h a t have a positive net present value. However, if these premises are relaxed one can derive barriers that prevent firms from implementing EE-technologies. The potential barriers can be categorized in the following groups (Gillissen, 1994): a. economic barriers: i) low expected energy prices; ii) uncertainty due to expected fluctuations in energy prices; iii) low expected revenues due to low energy bill; iv) budgetary problems; v) too high required return on investment; b. physical~technology barriers: i) reduction in production quality; ii) bounded rationality; iii) "technology-lock"; iv) information gap; c. m a n a g e m e n t barriers: i) no specialized personnel; ii) no interest in energy conservation; iii) no high priority to conservation (high opportunity costs); iv) present technologies are not fully depreciated; v) lack of pressure. Potential determinants for the implementation of EE-technologies are for example the number of information sources, the size of the firm (as a proxy for economies of scale), the presence of an energy coordinator, the presence of an R&D department, external pressure and bilateral agreements. In order to understand why some firms do and others do not adopt EE-technologies that, according to ICARUS, are economically attractive, a conceptual model for the investment behaviour of firms was constructed. The model consists of three "modules". The first module describes the level of information of a firm. Variables are: the n u m b e r of information channels, the presence of an energy coordinator, the presence of an R&D department and the presence of an environmental care system. Other important variables are: firm size, the energy bill, the complexity of an EE-technology and the costs of this technology. The second module describes how firms evaluate the profitability of an investment. Variables influencing the profitability evaluation are the minimally required r e t u r n to investment, and a possible bias in perceived r e t u r n and risk through a low priority for energy
1024 conservation in comparison with "core business activities". It should be noted t h a t the profitability as perceived by the firm may differ from the profitability as calculated in ICARUS because of uncertainties and firm specific expectations. The third module describes the implementation stage. Rational behaviour theories predict t h a t a firm will implement a technology when it is evaluated as being profitable. In practice, however, there can be barriers t h a t prevent a profitable technology from being implemented, whereas non-economic influences can cause an unprofitable technology to be implemented. To validate the hypotheses made in this study and the conceptual implementation model, a survey was made of the investment behaviour of more t h a n 300 Dutch firms. The survey focused on the information and implementation of sector specific EE-technologies, and on variables related to the theoretical d e t e r m i n a n t s and barriers (such as firm size, size energy bill, and required r e t u r n on investment). From this survey, and starting from a set of more t h a n 100 possible influential factors, the most i m p o r t a n t d e t e r m i n a n t s and barriers were identified. Also indicators for the degree of information and implementation were constructed. Finally information was gathered concerning the influence of policy i n s t r u m e n t s and firm specific variables. In another p a r t of the investment behaviour study, the focus is on w h a t actually h a p p e n e d in the Dutch manufacturing industry with regard to energy efficiency improvement. For this study a number a technologies is selected from ICARUS. The actual implementation rate of these technologies in the recent past (19801993) is investigated in relation to economic developments and the previously mentioned potential barriers. The results obtained so far (see Table 2.4) suggest that energy is considered as one of the production factors, and t h a t investments to reduce the use of energy are made largely on the basis of an economic evaluation, taking into account physical and financial constraints. I m p o r t a n t d e t e r m i n a n t s are: firm size, r e t u r n on i n v e s t m e n t , the availability of capital, the possibility of early depreciation. B a r r i e r s t h a t prevail are: u n c e r t a i n t y due to fluctuations in energy prices, b u d g e t a r y problems, poor financial m a r k e t expectations, a lack of knowledge of EE-technologies and the complexity of those technologies. Variables t h a t hardly seem to influence the decision making are amongst others the "core business" argument, the size of the energy bill, and the presence of an R&D department. In other words, decisions on EE-investments do not differ basically from the decisions on "core business" investments.
1025 Table 2.4 D e t e r m i n a n t s a n d b a r r i e r s for the i n v e s t m e n t in EE-technologies, impact (preliminary result)
clustered to
Very important variables
Important variables
Unimportant variables
1. Return on investment 2. Contribution to total profit 3. Securing working conditions 4. Securing production quality 5. Availability of own financial sources 6. Additional required investment costs
1. Size of energy bill 2. Environmental image 3. Depreciation status of old equipment 4. Securing production flexibility 5. Market expectations 6. Availability external fin. sources 7. Uncertainty energy prices 8. Degree of competence
1. Internal opposition 2. Internal rules 3. External rules 4. Low energy prices 5. Distance to core business 6. Costs collection addit. info. 7. Additional time and effort 8. Uncertainty new technologies 9. Qualified personnel
As m e n t i o n e d before, the second p a r t of the project entails an e s t i m a t i o n of the effects of e n e r g y policies on the i m p l e m e n t a t i o n b e h a v i o u r of firms. For t h a t p u r p o s e p l a u s i b l e e n e r g y policy scenarios for the f u t u r e (1994-2015) w e r e constructed and e v a l u a t e d by an expert group. These scenarios consist of a set of economic a n d r e g u l a t o r y i n s t r u m e n t s , combined w i t h expectations r e g a r d i n g economic growth and energy prices. In the next phase of the study, the impact of these elements will be evaluated using the newly developed implementation model and the results of the survey concerning the i n v e s t m e n t behaviour of more t h a n 300 firms. The s t u d y will be focused on the i m p a c t of e n e r g y taxes, e n e r g y subsidies, covenants, a n d i n f o r m a t i o n policies to reduce the i n f o r m a t i o n gap. U l t i m a t e l y , t h e s e a n a l y s e s should r e s u l t s in a s i m u l a t i o n model by which the i m p a c t of different e n e r g y policies on the promotion of e n e r g y savings can be assessed. 3.
M A T E R I A L E F F I C I E N C Y I M P R O V E M E N T AND WASTE MANAGEMENT
While CO 2 is generally considered as an energy related problem, this depends on t h e point of view. For The N e t h e r l a n d s , i n d u s t r i a l m a t e r i a l s p r o d u c t i o n is responsible for a p p r o x i m a t e l y one t h i r d of the n a t i o n a l CO2 emissions (50-60 versus 160 Mt). This p a r t of the CO2 emissions can be influenced by changes in t h e m a t e r i a l s s y s t e m , especially by i m p r o v i n g the efficiency of m a t e r i a l use. Options to improve the m a t e r i a l efficiency are: good housekeeping; a d a p t i o n of p r o d u c t designs; s u b s t i t u t i o n of m a t e r i a l s ; recycling of products; m a t e r i a l recycling; and cascade use of materials. A reduced consumption of m a t e r i a l s and products as well as an improved m a n a g e m e n t of w a s t e s t r e a m s will lead to a reduced depletion r a t e of resources, a reduced consumption of the energy and a
1026 reduced emission of atmosphere.
CO2 and other greenhouse gases (especially CH4) to the
3.1 T h e p r o d u c t i o n , f l o w and c o n s u m p t i o n of m a t e r i a l s in T h e Netherlands To u n d e r s t a n d the potential of material efficiency improvement and its impact on CO2 emissions, within the NRP-I an overview has been made of the production, flow and consumption of materials in the Dutch economic system for the y e a r 1988 and for the year 2000. Also a detailed analysis was made of the associated C02 emissions (Blonk et al, 1991, 1993, 1994; Gielen 1994). In the overview all materials are categorized in 22 groups, all products in 8 groups. This division is not very accurate. However, due to constraints in the availability of statistic data and to limitations of the computer model MARKAL t h a t is used to calculate costeffective routes for GHG-emission reduction based on these data (see Section 6), it does not m a k e sense to strive for a higher accuracy at the moment. Case studies were made to the material composition of complex products, like p a i n t and cars (Mulder, 1993; Moll, 1993). Also possibilities to change the composition of products were investigated (Gielen, 1994). For several products, i.c. fertilizers and plastics, the potential of material efficiency improvement, without influencing the end function, was analyzed in depth (Worrell, 1994). For The Netherlands the technical potential to reduce the application of nitrogen fertilizers - without loss of services - is found to be nearly 45%. Its realization would lead to energy savings of about 40%. For the use of plastics in packaging these figures are 35% and 30% respectively. 3.2 Organized thrift in waste disposal One option to increase the efficiency of material use and to reduce the emission of GHG's is the reduction of the amount of waste that is produced and subsequently burned in incinerators or disposed of on landfills. One factor t h a t influences the a m o u n t of waste produced and the composition of the waste is the institutional setting of the handling and removal of waste. Consequently it might well be t h a t the a m o u n t of waste t h a t is burned or disposed of can be reduced by a change of the institutional structure. Within the NRP-I research is executed to explore the potential of waste reduction in The Netherlands and to design a structure for organized thrift. The project is divided into five phases: 1. Examination of the structure of the Dutch waste sector. 2. Examination of the waste sector in a number of other countries. 3. Comparative study of demand m a n a g e m e n t in other sectors of society. 4. Design of a waste structure aiming at waste reduction and e x a n t e evaluation of the effects on the size of waste streams. 5. Investigation of possible problems in the implementation of the designed structure. In the first phase of the project an inventory was made of the structure of the w a s t e sector in The Netherlands. Also its impact on the m a n a g e m e n t of waste s t r e a m s was investigated. It was found t h a t at some points the present structure does indeed stimulate the growth of waste streams. Also it was found t h a t there are a lot of organizations, both public and private, that try to influence the m a n n e r
1027 in which and the degree to which solid waste is removed and processed. The following categories of actors are distinguished: ( 1 ) w a s t e generators, ( 2 ) w a s t e collectors, (3) w a s t e processors, (4) waste disposers, (5) policy m a k e r s , (6) organizations t h a t provide d a t a and ideas to support policy m a k e r s , and (7) interest groups and umbrella-organizations. The study has shown a n u m b e r of structural impediments for waste reduction. They can be summarized as follows (De Jong and Wolsink, 1993): a. Organizations t h a t collect and provide data about size and composition of w a s t e s t r e a m s s o m e t i m e s have i n t e r e s t s in the outcome of the d a t a collection. This could hinder the provision of objective and reliable data. b. Public a u t h o r i t i e s t h a t are s i m u l t a n e o u s l y waste collectors and w a s t e processors can pass on the corporate risk of an incineration plant to residents by changing the tariffs for the collection of domestic waste. This opportunity m e a n s t h a t the former can accept such risks more readily t h a n private institutions, which stimulates overcapacity of incineration facilities. c. Owners of incineration capacity are able to force waste collectors and waste processors to deliver collected waste to them. Long-term contracts and political influence and extension of waste processing activities are among their means of influence on waste collectors. & I n c i n e r a t i o n p l a n t s t h a t are in accordance w i t h the p r e s e n t s t r i c t environmental standards require large investments. The financial necessity to use these plant at full capacity could hinder investments in waste prevention, recycling of waste or other waste reducing activities. e. The long write-off periods of incineration plants are an obstacle to waste minimization policies of governments. For public authorities t h a t are owners of such plants, waste minimization damages their own financial interests. f. The responsibility of public authorities dealing with waste in general is limited to either the local, regional, provincial, or national level. Private waste organizations often cannot be forced to implement the policy of a public authority because they operate at another scale. g. It is beneficial to private waste collectors not to sign long-term contracts with waste processors. In t h a t way they r e m a i n free to search for the cheapest processing option. This often results in discrepancies between predicted and actual amounts of waste in a certain area. This can lead to a too low supply to the incinerators which implies an increase of the tariffs for contractees. In this situation, private collectors who are not contractually tied can offer "extra" waste to processors who want to see their remaining capacity utilized. The price is agreed by negotiation and is far lower than that to the contractees. h. Private collectors have vested interests in the continuity of the existing waste structure and in the existence of uncertainty among public authorities about quantities of waste t h a t will be offered for incineration. Therefore they try to avoid common decision-making with public authorities. Waste producing enterprises are interested in: Continuity of the existing waste structure and uncertainty about quantities of waste t h a t will be offered for incineration. Overcapacity will lead to lower instead of higher tariffs for the processing of industrial waste. Avoidance of exchange of information t h a t could lead to a better insight is size, weight and composition of waste streams. More information and better planning of facilities will limit bargaining opportunities of enterprises and lead to higher processing tariffs. -
-
1028 Possible solutions to structural impediments are: 1. Changes in contracts between participants in the waste market. 2. Improved exchange of information, cooperation a n d decision-making between participants in the waste sector. 3. Revised division of tasks and competencies of authorities. 4. Creation of an independent institution that can force organizations within the waste sector to give correct information and is able to interpret and to report about these data. 5. Better use of available policy instruments by authorities. In the second phase of the project the waste sector in other countries is examined. Based on criteria derived from the first phase, the following states were selected: Denmark, France, Nordrhein-Westfalen (Germany), California (USA), New Jersey (USA) and Massachusetts (USA). A further selection will be made later on. New J e r s e y became the state of the first case study. The state was visited in May-June 1994. Interviews were held with key persons in the waste structure of this state and written documents were collected. The oral and written information is sorted out and will be published at a later date. Nowadays in m a n y different sectors of the Dutch society, and for quite a n u m b e r of reasons, the need is felt to limit the use of existing (semi) collective facilities. Ideas are formulated and organizational experiments are put in practice to limit the d e m a n d of these facilities. Possibly the ways in which other sectors of society p u t p r e v e n t i v e m e a s u r e s in practice, and in p a r t i c u l a r the o r g a n i z a t i o n a l structures they develop to do so, may be of interest in designing a waste structure t h a t aims at waste minimization. This is investigated in the third phase of the project. Case studies were made of demand m a n a g e m e n t in energy supply, housing, psychiatric hospitalization and social security in The Netherlands. The results will be published in a book (in Dutch) which is in preparation now. Work on phase 4 and 5 of the project has yet to be started. 3.3 C o n t r o l l e d a n a e r o b i c t r e a t m e n t o f w a s t e a n d w a s t e w a t e r
Biological degradation of waste and waste water containing degradable organic substances can generate large quantities of CI-I4 and/or CO2. The actual quantity and composition of the produced gas depends on the amount and the composition of waste produced but also on the way the waste is discharged, stored, and treated. Focused on waste water, aerobic degradation as conventionally applied in most t r e a t m e n t s y s t e m s leads to the production of m a i n l y CO2, w a t e r and a considerable amount of biomass or sludge. Most of the potential energy present in waste water, therefore, ends-up in biomass, the disposal of which is becoming a serious problem. Furthermore, aerobic treatment is rather energy consuming since this process depends on a more or less intensive aeration. An i n t e r e s t i n g alternative is formed by anaerobic t r e a t m e n t (Lettinga & Van Haandel, 1993). Anaerobic degradation occurs when no or insufficient oxygen is available for (complete) aerobic degradation. Under anaerobic conditions organic material can be completely degraded into CO2, CH4, water, a small a m o u n t of biomass, and traces of other components. The produced biogas contains the largest part of the potential energy present in waste water, and can be used as a fuel. By doing so, anaerobic degradation has the following advantages over aerobic degradation: 1)
1029 Production of a valuable fuel, the use of which can lead to a reduction of the a m o u n t of fossil fuel consumed, 2) no energy requirement for aeration, and 3) significantly less sludge production. On the other hand, if anaerobic degradation occurs in an uncontrolled way, CH4 can be emitted to the atmosphere where it enhances the greenhouse effect even more than CO2. Global atmospheric CH4 emission from the treatment, storage, and discharge of all waste water is roughly estimated to be about 26-40 Tg/y, thereby representing about 8-11% of the total anthropogenic CH4 emissions (Thorneloe, 1993). The contribution of the Food & Beverage (F&B) and the Pulp & Paper (P&P) industry, the main contributors to the organic pollution load, is estimated to be 12-18 Tg/y. The contribution of domestic waste water is estimated to be only about 2 Tg/y. In the NRP-I research is done to investigate the potential of applying controlled anaerobic processes for the t r e a t m e n t of both domestic and industrial (F&B and P&P) waste water in order to reduce the global atmospheric emission of both CH4 and CO2. In this project the present day emission of CH4 and CO2 from waste water t r e a t m e n t is estimated. Furthermore, the potential production and use of CH4 from waste water t r e a t m e n t and its impact on the emission of CH4 and CO2 emissions is analyzed (Lexmond and Zeeman, 1993). In the estimations the following cases are considered: 1) Complete anaerobic treatment: Calculated emissions are energy related CO2, and CH4 from anaerobic degradation. Three options are regarded: - All CH4 is flared. - CH4 is partly used for the maintenance of the treatment plant. The excess is flared. - CH4 is completely used for energy production. It is assumed t h a t a certain percentage of the produced CH4 is lost due to leakage and to amounts that are too small to collect. 2) Complete aerobic treatment: Calculated emissions are CO2 from the use of fossil fuels, and CH4 from uncontrolled anaerobic degradation also occurring in aerobic systems (Czepiel et al., 1993). CH4 produced in aerobic systems is assumed to be emitted to the atmosphere. 3) Current situation: Calculated emissions are the energy related CO2 emission and the CH4 emissions from controlled t r e a t m e n t s y s t e m s and from uncontrolled degradation in the environment. Relevant data concerning waste water (amount, composition, degraduability) and t r e a t m e n t systems (efficiency, sludge growth, energy demand, methane emission factors, frequency of operation) were collected from literature and queries. Models were developed to be able to estimate the CH4 and CO2 emissions (Lexmond and Zeeman, 1994). Due to scarcity of information, amongst others an assumption had to be made about the current use of different waste water t r e a t m e n t systems. As this figure has a strong influence on the final results, the a s s u m p t i o n is given in 3.1. Especially the percentage of uncontrolled aerobic and anaerobic degradation of discharged waste water in developing countries is an important parameter. It is assumed that a large part of the waste water is discharged on surface waters.
1030
Table 3.1 Assumptions concerning the t r e a t m e n t of waste water and the division in aerobic and anaerobic degradation a) treated (%)
u n t r e a t e d (%)
total
aerobic
anaerobic
total
aerobic
anaerobic
developing countries * domestic
10
70
30
90
75
25
* industrial
50
85
15
50
75
25
d e v e l o p e d countries * domestic * industrial
90 95
90 85
10 15
10 5
80 80
20 20
a) A s s u m p t i o n s are p a r t l y based on a query held among s t u d e n t s from the I n t e r n a t i o n a l I n s t i t u t e for H y d r a u l i c a n d E n v i r o n m e n t a l Engineering, The N e t h e r l a n d s , 1993
The results of the study are shown in Table 3.2. The estimated CH4 and C 0 2 emissions for F&B and P&P industry waste water and domestic sewage t r e a t m e n t are presented as calculated for the five cases described earlier. In the calculations two values for the loss of CH4 were used, 5% and 10%. As shown, the current CH4 emission from the investigated waste water streams is estimated at 5 Tg/y. This is considerably lower than the value of 14-20 Tg/y, estimated by Thorneloe (1993). Also it has been calculated that the present contribution of waste water t r e a t m e n t and disposal to the enhanced greenhouse effect is mainly d e t e r m i n e d by the emission of CH4 from uncontrolled disposal of domestic waste water in developing countries. So, not only from the point of view of h u m a n health, but also from the greenhouse point of view, it is of great importance t h a t in these countries waste w a t e r t r e a t m e n t is stimulated. The results show t h a t anaerobic t r e a t m e n t of waste w a t e r should be stimulated in order to reduce the emissions of greenhouse gases, provided t h a t the loss of CH4 is minimized and the produced CH4 is optimally used.
1031 Table 3.2 E s t i m a t e d emissions from industrial and domestic waste w a t e r t r e a t m e n t and disposal
CH 4 loss
i n d u s t r i a l e m i s s i o n (Tg/y)
d o m e s t i c e m i s s i o n (Tg/y)
5%
5%
10 %
10 %
CH4
C02
CH4
C02
CH4
CO 2
CH 4
C02
0.8
2.8
1.6
2.8
0.9
2.5
1.8
2.5
0.8
0.0
1.6
0.0
0.9
0.0
1.8
0.0
0.8
- 32.8
1.6
- 31.0
0.9
- 36.9
1.8
- 34.8
* aerobic
0.1
17.0
0.1
17.0
0.1
15.3
0.1
15.3
* current situation
1.2
12.1
1.3
12.1
3.6
3.9
3.6
3
Cases * only flaring * partial reuse * complete reuse
It should be noted t h a t the t r e a t m e n t method and the disposal of waste water are the most i m p o r t a n t p a r a m e t e r s for the e s t i m a t e d emissions in the c u r r e n t situation. For most countries very little data is available about these subjects. Within the calculations the most important p a r a m e t e r is the amount of anaerobic degradation in the case of untreated waste water in the developing countries. If we, for instance, would assume t h a t this amount is not 25% but 50%, the total CH4 emission would increase from 5 to about 11 Tg/y. Therefore more data should become available to be able to estimate the present emissions more accurately. In the next phase of the project, a similar study is planned for the t r e a t m e n t of sludge produced during waste w a t e r t r e a t m e n t . During aerobic w a s t e w a t e r t r e a t m e n t about 4 times as much sludge is produced as d u r i n g anaerobic t r e a t m e n t . This sludge has to be disposed of and treated. In The N e t h e r l a n d s sludge of aerobic installations is often digested, resulting in biogas which is normally used or flared. Anaerobic sludge contains less organic material and is therefore less suited for digestion. In order to obtain a complete picture of the emissions from waste w a t e r treatment, a comparison will be made of the most common sludge t r e a t m e n t systems and their related CH4 and CO2 emissions 4.
S W I T C H I N G TO R E N E W A B L E S
An important option to reduce C O 2 emissions is switching to renewable sources of energy. In 1990 the contribution of renewables to the world energy supply was about one-fifth. Several studies have indicated that it might be possible to increase this figure to one-quarter or one-third in the year 2025 and to about one-half in the second half of the next century (UNSEGED, 1992; World Bank, 1992; WEC Study Group on Renewable Energy Resources, 1993; World Energy Council, 1993; T.B. Johansson et al, 1993). In the coming decades, apart from hydro power and wind energy, especially the use of biomass to produce fuels and electricity seems attractive. In the longer term a major contribution is expected also from energy production by solar cells.
1032 In The Netherlands the current consumption of primary energy resources is about 2750 PJ. About 1% of it is obtained from renewables. The potential, however, is much higher. It is estimated that the exploration of renewable energy sources in The Netherlands that can replace about 700 P J of our fossil fuel consumption, with major contributions from solar cells, wind turbines and the production of fuels and electricity from biomass including organic waste (Van Wijk et al., 1994). There are however many uncertainties. Some of these uncertainties are addressed in a r e s e a r c h project i n i t i a t e d by the NRP-I, focused on the p o t e n t i a l and attractiveness of biomass energy production in The Netherlands. 4.1 H a r v e s t i n g the s u n u s i n g e c o s y s t e m s The agricultural production in Western Europe and the US has risen enormously in recent decade. As an example, in The Netherlands the wheat yields per hectare have doubled since 1950. This has led to a reduction in land area needed for food production, meaning that there is land that could be used to grow energy crops. The objective of energy plantations is a high sustainable yield at low cost with minimal adverse impact on the environment. The demands placed on biomass are different from those for regular agricultural crops. Mainly, an energy crop must produce a great deal of material, containing little moisture, and easy to harvest. In this context much attention is given to fast growing wood (e.g. poplar and willow) and some grassy species (e.g. reed and miscanthus). However, there is a great deal of vagueness about what the possible yields of such crops might be, whereas these yields are in fact a key factor in all evaluations of the attractiveness of biomass energy production. Therefore, within the NRP-I, a simple crop growth simulation model is developed to improve the yield estimates of (presently often unknown) energy crops (Nonhebel, 1994). As an example the calculation of the energy yield of a poplar plantation is given. In The Netherlands the poplars begin to sprout at the beginning of May, losing their leaves in October. The amount of solar radiation that is received during this period is 1.5 GJ/m2. Under optimal circumstances (sufficient water, etc.) 1.4 g of plant material is produced for every MJ of radiation. This would result in a biomass production of 21 ton per hectare. As the optimal circumstances seldom occur, in practice the production will be less. Also not all of the biomass can be harvested for energy production (leaves, young twigs, etc.). Therefore, in the year 2000 the annual yield might be 10-12 ton dry matter per hectare. It should be noted t h a t other studies have indicated yields of about 15 ton/ha. These studies were based on field experiments. In practice, yields will generally be lower. Also it should be noted that the growing conditions vary enormously throughout Europe. Thus great differences occur in the potential productivity between different regions. Therefore the simple simulation model will now be used to calculate the maximum yield of a variety of biomass crops in different European regions. This will give an insight into which crops can where best be grown with what productivity. 4.2 C o n v e r s i o n routes for e n e r g y crops There are several conversion routes for the transformation of energy crops into heat, fuels and/or electricity. Within the NRP-I ten conversion routes have been analyzed with respect to technical, financial and environmental characteristics (Van den Heuvel et al., 1994). The options considered include: cogeneration of heat
1033 and power with combustion technology; co-combustion in a pulverized coal fired power plant; electricity production with an integrated gasifier combined cycle plant of 50 MWe; combustion of gasified biomass in a conventional n a t u r a l gas fired power plant; fermentation of wheat to ethanol using wheat straw for combined h e a t and power generation; fermentation of sugar beet to ethanol; and RME production from rape seed followed by esterification. All these technologies are technically m a t u r e , with the exception of the integrated gasification combined cycle plant. The lowest electricity production costs are found for the large-scale gasification option and for co-combustion. The production costs are calculated at 0.15 NLG/kWh. This figure is much higher t h a n current electricity production costs with conventional power stations. In this study also the specific costs to avoid the emission of CO2 have been analyzed. The lowest figure is found for the BIG-CC plant of 50 MWe. The costs are calculated at about 75 NLG per ton CO2 avoided. Much higher costs are calculated for the investigated production of fuels. It is concluded t h a t future research should concentrate on large scale gasification of biomass and on cocombustion of biomass in existing power plants.
4.3 Can Dutch agriculture provide clean energy? By generating energy from vegeTable fibers (biomass) the emission of greenhouse gases can be reduced. As well as an ecological advantage, the cultivation of crops for the supply of energy could also improve the moderate to bad economical results of Dutch arable farms. Energy crops can also create new problems, such as an environmental burden due to their cultivation. So far research into the use of biomass as a source of energy has been focused mainly on its technical and economic feasibility. Therefore, within the NRP-I also research was set up to assess the ecological sustainability of the cultivation and use of energy crops (Zeijts et al., 1994). Can the Dutch agricultural sector provide clean energy? Elements of this question are: how harmful to the environment is the cultivation of energy crops; w h a t are the direct and indirect environmental effects of fitting energy crops into the cropping plan; w h a t indirect effects are to be expected at a regional level; how much energy is produced in the entire cultivation, transport end processing chain and w h a t effect does this have on the emission of greenhouse gases; and w h a t is the overall conclusion for the various crops with regard to sustainability. In this research the following effects of the cultivation and use of energy crops have been assessed: the emission of minerals; the emission of pesticides; the use of energy and the emission of greenhouse gases, the fixation of carbon from CO2; the use of by-products and waste products; desiccation; erosion; the contribution to n a t u r a l values; the contribution to scenic values; and the use of space. Nine crops and their processing chains have been studied: rape (to bio-diesel oil), sugar beet and w i n t e r w h e a t (to ethanol), and maize, hemp, reed, miscanthus, poplar and willow (to electricity and/or heat). First the direct effects of the cultivation of energy crops have been assessed. The results are compared with those of cultivation on fallow land and cultivation of cereal. The direct effects are found different for each energy crop, but in general the conclusions are as follows:
1034 - I f cereal is replaced by energy crops, the emission of minerals deteriorates because the cultivation of cereal leads to low mineral losses. If energy crops are grown on fallow land, the emission remains the same. -For most energy crops, the emission of pesticides reduces if they replace cereal. If they are grown on fallow land, there is no difference on average. - I f a n n u a l energy crops are grown on fallow land or replace cereal, the risk of erosion increases. -Cereals contribute considerably to n a t u r a l values; replacing them by energy crops leads to lower natural values. Compared with cultivation on fallow land, the average effect is neutral. -Fallow land and land on which cereal is grown require little water, so that the risk of desiccation increases if energy crops are grown there instead. Effects that are not directly related to the cultivation of energy crops are known as indirect effects. Indirect effects on environmental pollution and on n a t u r a l and scenic values are found at the level of the individual farm as well as at a regional and national level. A few examples are given here. At the level of the individual farm, perennial crops are likely to have a cleaning effect on the soil as regards persistent weeds and a number of soil pathogens and nematodes. However, the annual crops in the cropping plan will follow each other faster. It is expected t h a t the positive effect (clean soil) is only of interest during the first years after the cultivation of the perennial crop, and t h a t the negative effect (faster rotation of a n n u a l crops) will predominate. On balance, this means greater mineral and pesticide emissions and a greater risk of erosion. An indirect effect at a national level is related to the national mineral surplus. Energy crops will first be grown on fallow land, but will also replace cereal, which is normally used to produce animal feed. If the livestock population remains the same, replacement of cereal leads to an increase in the import of animal feed and/or the use of artificial fertilizers (for more intensive cultivation of animal feed). Negative effects on the national mineral surplus can be compensated by: 1) processing byproducts and waste products (such as b y p r o d u c t s from the fermentation of sugar beet and wheat) in animal feed, so t h a t imports will not increase as much; 2) exporting the waste product ash as a fertilizer as it contains almost all the phosphate from the biomass; 3) using animal m a n u r e r a t h e r t h a n fertilizers for energy crops. In this study the following net avoided emission of greenhouse gases are calculated for the various processing and conversion routes: - 2 to 3 tons of CO2-equivalent per hectare per year in the production of rape diesel oil by means of extraction; - 3 to 7 tons of CO2-equivalent per hectare per year in the production of ethanol by means of fermentation; -8 to 24 tons of CO2-equivalent per hectare per year in the production of electricity by means of burning and gasification. The differences are caused by differences in crops, regions, processing and conversion routes. Rape extraction produces the lowest score on the emission of greenhouse gasses, followed by fermentation of sugar beet and winter cereal. Crops that are burnt or gasified score the highest.
1035 Based on these results, and giving all evaluation criteria equal weight, the following order for the attractiveness of biomass energy production from an environmental point of view is derived: 1. winter wheat; 2. hemp and reed; 3. miscanthus, poplar and willow; 4. rape and maize; 5. sugar beet. Winter wheat scores very high. Its cultivation produces hardly any mineral losses, it has positive n a t u r a l and scenic values and it limits the risk of erosion and desiccation. A great disadvantage, however, is the low score on the emission of greenhouse gasses. Hemp and reed score relatively high. A great advantage is t h a t the net avoided emission of greenhouse gasses is high. A disadvantage of both crops is t h a t they use a great deal of water, which increases the risk of desiccation. Few pesticides are required for the cultivation of both hemp and reed. Reed p r e s u m a b l y uses m i n e r a l s very efficiently, but this effect is countered by the above mentioned negative effects of perennial crops on the emission of minerals and pesticides in the other crops in the cropping plan. Fitting hemp into the cropping plan has a positive indirect effect on the use of pesticides for the other crops. Miscanthus, poplar and willow have moderate scores as regards environmental impact. The net avoided emission of greenhouse gasses is fairly high. The mineral losses in cultivation are not likely to exceed the levels for nitrate in ground water, the use of pesticides is relatively small and erosion is prevented fairly effectively. On the other hand, however, there are negative effects of fitting perennial crops into the cropping plan. Because these crops use a relatively great a m o u n t of water, there is a risk of desiccation. Rape and maize have a slightly negative score. A great disadvantage of rape is the low score on the emission reduction of greenhouse gasses and the use of space. Rape does, however, contribute considerably to natural and scenic values; maize does not. Both crops have a fairly low score on the emission of minerals and pesticides and for the risk of desiccation. The main reason for the low score of rape in the field of pesticides is that rape has the same crop rotation problems as sugar beet. E x t r a rape or sugar beet in the cropping plan will lead to an increase in the need for pesticides. The cultivation of (extra) sugar beet has m a n y ecological disadvantages and no obvious a d v a n t a g e s . It scores low as regards pesticides (also because of the negative effects on soil health), erosion and scenic values. Cultivation of sugar beet also has a low score on minerals, desiccation, natural values and the use of space. An important disadvantage is that it contributes relatively little to decreasing the greenhouse effect. It should be noted t h a t miscanthus, reed, willow and poplar do not only score high from the point of view of environmental impact. They are also economically the most feasible, in view of the results of other studies. There are possibilities for these perennial crops especially at extensive farms with fallow land. Studies of the economic feasibility have shown that in The Netherlands energy from biomass can be profitable in the n e a r future, if the set-aside scheme is continued and an environmental tax on non-renewable energy or a subsidy per ton of CO2 avoided is introduced. Hemp is somewhat less feasible from an economic point of view. For
1036 the mainly intensive Dutch arable farming, however, hemp is interesting because it is easier to fit into the cropping plans. It could be grown on fallow land or it could replace cereal. Also from the point of view of environmental impact, hemp scores high. In other European countries energy from biomass is already being generated on a limited scale. If a European subsidy for avoided CO2 emissions or a E u r o p e a n environmental tax on energy is introduced, the cultivation of energy crops will sooner commence in these other countries t h a n in The Netherlands. In other E u r o p e a n countries ground prices are lower and more land will probably be set aside. Also, in m a n y European regions sustainable cultivation will be easier to realize than in The Netherlands. In the long term, energy from biomass will receive competition from two sides. Firstly, agricultural land will be claimed for other purposes, such as n a t u r e development. Secondly, there may be more cost-effective ways of reducing the emission of greenhouse gasses by then.
0
D E C A R B O N I Z A T I O N O F F U E L S AND F L U E GASES, AND S T O R A G E OF CO2
In the longer term, decarbonization of fuels and flue gas is the only possible GHG mitigation option allowing large-scale use of fossil sources. It can only be done practically in large scale conversion facilities. Till the beginning of the nineties, the potential of this option was hardly investigated. Therefore, in 1991 an explorative r e s e a r c h p r o g r a m was set up with a n u m b e r of research i n s t i t u t e s in The Netherlands. The objective of this program was to get a better u n d e r s t a n d i n g of Carbon Dioxide Removal (CDR) option and to investigate its potential for The Netherlands. The research was carried out in 1991 and 1992. The program was sponsored from various sources, the main ones being the Ministry of Housing, Physical P l a n n i n g and Environment, VROM and the Dutch National Research Programme on Global Air Pollution and Climate Change (NRP-I). The total budget amounted to about 1.5 million Netherlands guilders (NLG). Before the research program started, several publications on carbon dioxide recovery and storage in The Netherlands had already been issued, with special emphasis on carbon dioxide removal from ICGCC power plants and on storage of carbon dioxide in depleted n a t u r a l gas fields. In the new program the m a i n emphasis was on studying a range of techniques to recover carbon dioxide from gas s t r e a m s in detail. Some studies were devoted to the recovery of CO2 from industrial processes. One study explored the possibilities of CO2 storage in aquifers. A separate study was conducted to calculate the impact of CO2 removal on the conversion efficiency and costs of complete power plants. The findings have been presented in a final r e p o r t , published in 1993 (Blok, 1993). A s u m m a r y of this report is given here. Table 5.1 presents some conclusions. 5.1 CO2 r e c o v e r y b a s e d o n coal g a s i f i c a t i o n It was confirmed that carbon dioxide recovery based on coal gasification shows the smallest decrease in efficiency of electricity production. A detailed analysis carried out by the research institute of the electric utilities, KEMA, showed t h a t due to
1037
CO2 recovery the conversion efficiency of the power plant might be reduced from 42-43% (reference value) to about 36%. Two widely differing CO2 recovery technologies have been investigated: In the first approach, a water-gas shift reaction is applied after gasification of the coal, resulting in a fuel gas mainly consisting of hydrogen and carbon dioxide. For separation of hydrogen and carbon dioxide a number of options were studied: freezing out the CO2, membrane separation, hydrogen recovery, physical absorption and chemical absorption. The best option is physical absorption (using Selexol). By freezing out the C02 the required high degree of CO2 recovery (to less than about 120 g/kWh) can probably not be attained. When this limitation would not be set, freezing out the carbon dioxide should certainly be taken into consideration. The use of membranes at low temperatures is not attractive, mainly due to the high hydrogen loss. Combined with high-temperature gas clean up the application of membranes may become of interest, although the membranes with the required high H2/CO2 selectivity are still under development. Chemical absorption systems have a too high energy demand. Hydrogen recovery techniques are showing considerable hydrogen losses. The components for the favored shift/Selexol concept are commercially available. It is concluded that the technology is ready for demonstration. In the second approach, the ICGCC makes use of a gas turbine in which the fuel is combusted in a mixture of oxygen and recycled CO2. The combustion products (mainly CO 2 and water) are expanded through the turbine section. After cooling in a heat recovery steam generator and removal of the water, the CO2 is recycled to the gas turbine compressor. Part of the compressed CO2 is recycled to the gas turbine combustion chamber, the remainder is exported. As CO2 is the main working fluid in the gas turbine, the properties differ strongly from a conventional gas turbine. The results of an analysis focused on the integration of such a gas turbine in an ICGCC power plant are given in Table 5.1. The main bottleneck for the application of this scheme is the fact that a CO2-gasturbine is required which is not available at present. Its development could be costly. Starting such a development process is only justified if this recovery technology gives clear advantages above the ICGCC/shift/Selexol-process already mentioned. 5.2 C h e m i c a l a b s o r p t i o n and o t h e r r e c o v e r y t e c h n i q u e s As the future of coal gasification is still uncertain it is advisable to develop other CO2 recovery techniques as well. In the research program, a number of other options for CO2 recovery has been evaluated. It was found that in most cases chemical absorption, using amines, is the most attractive alternative.
1038 Table 5.1 The impact of Carbon Dioxide Removal on the performance of power plants, as calculated by KEMA (Koetsier et al., 1992) *) T y p e of plant +
method of recovery ICGCC + C02/O2 combustion (Texaco) ICGCC + C 0 2 / 0 2 combustion (Shell) ICGCC + shift & physical absorption (Texaco) Pulverized coal + chemical absorption (retrofit) Natural gas fired combined cycle + chemical absorption Reference coal fired power plant Reference natural gas fired power plant
Net conversion efficiency (%)
Specific C02 emission (g/kWh)
34.8
5
36.0
30
36.4
139
29.7
105
44.9
86
42-430
800-820
52
390
Also some preliminary costs analyses were made. It was found that the recovery costs might range from 60- 80 N L G per ton of C02 avoided, being the lowest for the ICGCC options. Electricity production costs in decarbonized power plants are estimated at 0.14 - 0.15 N L G / k W h for coal and 0.095 N L G / k W h for natural gas. For the recovery of CO 2 from the flue gas of a conventional coal fired power plant the use of gas separation membranes is more expensive than chemical absorption. This is mainly due to the high power requirements for the compression of the flue gases. When chemical absorption is applied, the use of gas absorption m e m b r a n e s is of interest. Gas absorption membranes are used in conjunction with chemical absorption liquids where the conventional absorption column is replaced by a m e m b r a n e contactor. This modification might reduce the loss in conversion efficiency of the power plant with approximately 0.5%. This i m p r o v e m e n t is mainly due to a reduction of the pressure drop over the absorber. Gas absorption m e m b r a n e systems, however, are still under development. Also in the case of a natural-gas fired combined cycle power plant the most costeffective option is chemical absorption. As shown in Table 5.1, it reduces the overall conversion efficiency from 52% (reference value) to approximately 45%. An alternative is a power plant based on a gas turbine using combustion in a CO2/O 2mixture.
Also a system based on methane reforming of natural gas (to a large extent similar to an ICGCC plant) was investigated. However, it showed a low conversion efficiency: about 37%.
5.3 CO2 recovery in manufacturing industry T w e n t y plants in manufacturing industry with the largest C02 emissions in The N e t h e r l a n d s together are responsible for about 20% of the total CO2 emission of
1039 The N e t h e r l a n d s . Main sectors are: refineries, and the iron and steel, petrochemical and fertilizer industries. The potential of CO2 recovery in these sectors has been investigated. Carbon dioxide recovery can be accomplished in refineries equipped with a residue gasification unit. Residue gasification is expected to be a good solution in the d e v e l o p m e n t t o w a r d s low sulfur oil products and deeper conversion. The gasification product is fed to a shift reactor in order to produce hydrogen for other refinery processes. The carbon dioxide that is co-produced can be recovered easily. In this way about one quarter of the CO2 emissions in future refineries can be avoided. Another attractive option is available in the fertilizer industry. In producing a m m o n i a , which is one of the m a i n feedstocks for fertilizer production, a p p r o x i m a t e l y 50% of the CO2 output of the fertilizer i n d u s t r y is already recovered. At present, the recovered CO2 is partly utilized. The remainder is vented to the a t m o s p h e r e . In this case CO2 recovery can simply be achieved by compressing this stream to transportation pressures. Both for the refineries and the fertilizer industry, the estimated mitigation costs are in the order of 20 NLG per ton of CO2 avoided. More costly options were identified in the iron and steel industry: recovery of CO2 from blast furnace gas; and in the petrochemical industry: the use of lowt e m p e r a t u r e waste h e a t (100 - 150 ~ for supplying the reboiler duty of a chemical absorption process. 5.4 Storage of c a r b o n dioxide According to one of the studies, CO 2 storage in deep aquifers is technically feasible. When injecting CO2 in an aquifer part of the water already present will be displaced. The main mechanisms for this displacement will be gravity segregation and viscous fingering. Extended simulations of the behavior of CO2 have been carried out for sample reservoirs. In one of these 15,000 tons of carbon dioxide per day could be injected during 8 years. After this period CO2 b r e a k t h r o u g h was observed at the spill point. The subsurface of The Netherlands contains a large number of aquifers t h a t are potentially suitable. Taking into account a number of constraints, a prudent estimate of the total storage capacity of these aquifers was made. The resulting figure was 1.2 gigaton CO2. The main chemical influence of carbon dioxide in aquifers is its effect on carbonate chemistry. The decrease of the pH due to the dissolution of carbon dioxide will cause solution of carbonates. This effect is so small that weakening of the porous structure is not expected. However, significant changes in permeability may occur. If the seal of the structural trap is a clay layer, drying of this layer could reduce its tightness. D e p a r t i n g from a CO2 delivery pressure of 110 bar, the injection costs are estimated to be 0.7 to 1.2 NLG per ton of CO2, depending on the storage depth. 5.5 C o n c l u s i o n s As far as C 0 2 recovery from power plants is concerned, options based on coal gasification with CO2 recovery t u r n out to be most energy-efficient. Of the remaining recovery options chemical absorption from flue gases, using amines seem most promising. A n u m b e r of recovery options based on m e m b r a n e technologies have been identified, but most of them still require considerable development. Storage of CO2 in aquifers seems to be technically feasible, but
1040 assuming very strict conditions the total storage capacity might be smaller than the storage capacity in exhausted natural gas fields. More than at present, attention should be paid to possibilities for CO2 recovery t h a t are not directly related with electricity production, e.g. in manufacturing industry. It is felt that storage of carbon dioxide will pose most problems. These problems are not only of a technical and environmental character. Also legal questions arise, e.g should storage of'waste' in the underground be allowed? 6.
I N T E G R A T E D A S S E S S M E N T S : S T R A T E G I E S F O R CO2 E M I S S I O N REDUCTION
Most GHG emissions in The Netherlands are related to the use of fossil energy carriers. An array of technological measures is available to reduce these emissions, ranging from fuel shifts and renewable energy sources to improved waste management or shifts in materials use. Beforehand, it is unclear how much CO2 reduction can be achieved and at what cost. The assessment of emission reduction options is complicated because technologies are linked in the energy system (i.e. the total of energy supply, conversion, distribution and use). If e.g. the electricity production becomes less CO2 intensive, electricity savings become less attractive for CO2 reduction. Sponsored by the NRP-I and the D e p a r t m e n t of Economic Affairs of the Government, the so-called EMS project was set up to assess the potential and cost-effectiveness of reduction options in an integrated approach, taking the whole energy system into account. EMS stands for Energy and Materials use Scenarios for the reduction of CO2 and other GHG emissions. Three main parts of the project are considered here: 1. Integrated assessment of CO2 emission reduction for The Netherlands energy system (Okken et al., 1993; 1994). 2. Integrated assessment of GHG emission reduction for The Netherlands energy system (Ybema and Okken, 1993). 3. Integrated assessment of CO2 emission reduction for The Netherlands energy and the materials system (Gielen and Okken, 1994). The i n s t r u m e n t t hat is used for this study is the MARKAL model (MARket ALlocation model). MARKAL is internationally used in IEA/ETSAP (International Energy Agency/Energy Technology Systems Analysis Programme) (Kram, 1993). This linear programming (LP) model can be used to develop integrated energy strategies while environmental targets are taken into account. The model contains a database with approximately 500 energy supply and demand technologies. Each technology is characterized by technical, financial and environmental parameters. The model is used to calculate the least-cost system configuration for the period 2000 - 2040, meeting exogenously defined national energy service demands and emission reduction targets. The outcomes strongly depends on the inputs in the database. These inputs were derived from many specific studies. Assumptions with respect to the costs, potentials, and performance of technologies and with respect to scenario characteristics such as fuel price paths and energy demand
1041 projections have been extensively reported. As an example, Table 6.1 gives some key data for a limited set of electricity generating technologies for the year 2030. 6.1 CO2 e m i s s i o n r e d u c t i o n in The N e t h e r l a n d s e n e r g y s y s t e m The database of energy technologies has been used to analyze the potential of cost-effective CO2 emission reduction. The analyses was done for a high (D) and a low (G) economic growth scenario, with and without nuclear energy. The scenarios are indicated by DK, DZ and GK, GZ respectively. First the base case emissions in these scenarios were assessed. Figure 6.1 shows the results (solid lines). The increase from 2000 to 2040 ranges from 3% to 32%. The increase is due to various mechanisms. On one hand the demand for energy services increases as well as the share of coal in the primary energy mix. On the other hand already in the base case significant efficiency improvements and enduse savings are taking place. Next, several CO2 emission reduction paths were studied, also shown in Figure 6.1 (dashed lines). The imposed constraints are such that the CO2 emissions follow linear paths from 2000 to 2030 and then stabilize. Reduction percentages imposed for the year 2030 include 0%, 20%, 40%, 50%, 60%, 70% and 80%, compared to the emission in 2000. It is found that, on the longer run, significant C02 reduction (up to 70%) can be achieved at marginal reduction costs below 400 NLG/t CO2, as shown in Figure 6.2. Many of the implemented CO2 removal technologies at present hardly or not available. Therefore the attainable emission reduction on short terms, e.g. in the year 2010, is much smaller. For this year also the marginal costs as function of reduction percentage will increase rapidly. It is concluded t h a t significant CO2 reduction are a m a t t e r of long-term planning and requires a high attention for technological innovation.
1042 Table 6.1 Key data for some selected electricity generating technologies in 2030, used in the MARKAL studies (NLG of 1992)
Investment Life cost (NLG/kWe) (year)
Technology
Potential Generation costs (GWe) (NLG/kWh)
ICGCC power plant
2400
25
no limit
0.090
ICGCC + CO 2 removal
3000
25
no limit
0.106
Light W a t e r Reactor
4775
25
12.0
0.101
PV systems on rooftops
2000
20
12.5
0.177
Wind turbines offshore
3000
20
12.5
0.115
Biomass power plant
3140
20
1.5
0.130
MCFC industrial cogeneration
2300
20
no limit
0.109
[Mt CO2/YEAR] 250 DZ DK 200
150
...................
GZ -~-GK CONSTANT
- ..........
"":-:'!i':!!)!i!)'::::: .......
"..............................................
20%
..... :::::..:::: ..... :::: ....... ::: ................................ ...... :::: ...... ".......... . .............................. ........ :: ........ " .............................
100
400/0
50o/. 600/0
......... "'.... " .............................
50
to%
9.......................
8 0 %
I
I
I
I
I
I
1990
2000
2010
2020
2030
2040
Figure 6.1 Base case scenarios for the emission of CO2 in The N e t h e r l a n d s and emission reduction constraints (see text)
1043 [DFL/t CO2] 500
l!
400
/i; I; '
/ o_..%_. OK / // p .--.u--..-"/ Gz
- /
......
~
......
iJ ,. , .,~. ~.~.,;,;,~,,.~ ....A ....
20O
...... ~.."'~'~""~" .... ~.:,.;':;" ~
100 -
. ......
0 ~''" . . . . 0 Figure
6.2
"-'-
-
U
..-
"~
' ' ' 2O 4O 6O c o 2 EMISSION REDUCTION [%]
'
80
Marginal costs of C02 emission reduction in The Netherlands in the year 2030 assuming scenario GK, GZ, DK and DZ, taking into account the emission of CO2 from the energy system only (Dfl = NLG) [Mt C O 2 / Y E A R ] 200]
150
-
~
RENEWABLES
~
H2 FROM NAT. GAS WITH CO2 REMOVAL
~
OTHERCO2REMOVAL MAINLY COAL
~
100
! E
FOSSIL FUEL SUBSTITUTION I CONVERSION SAVINGS END USE SAVINGS
50
0 constant 20 40 60 C O 2 E M I S S I O N R E D U C T I O N [%]
80
Figure 6.3 The allocation of CO2 emission reduction for the stand-alone energy system in the year 2030, starting from base case scenario DZ The impact of the required CO2 emission reduction on the configuration of the energy system is significant. On the energy demand and conversion side, further savings will be realized, e.g. through better insulation of buildings. On the energy supply side, the potential for renewable energy is limited due to climate and geographic conditions. CO2 recovery and storage plays a crucial role for achieving significant emission reduction at accepTable costs. At lower emission reduction levels, CO2 is removed from power plants. At higher emission reduction levels, syn-
1044 fuels are introduced and a "hydrogen economy" combined with C02 removal develops. In these analyses it is assumed t h a t within 15 years The N e t h e r l a n d s will have capacity for CO2 storage in depleted natural gas fields or aquifers. As The Netherlands already now rely to a large extent on natural gas, the potential for fuel substitution as CO2 mitigation option is limited. Figure 6.3 shows a breakdown of the CO 2 emission reduction in 2030 in the DZ scenario for different emission reduction targets. Most reduction is achieved at the supply side. The potential for demand side emission reduction is limited, as most energy conservation options were already implemented in the base case. As the model contains a whole array of reduction options, a detailed discussion of the attractiveness of each options is beyond the scope of this report. An overview of the a t t r a c t i v e n e s s of major options in each scenario is given in Table 6.2, assuming a reduction target of 60% in the year 2030. A general cost figure for each option, in NLG per ton CO2 avoided, is not available as this figure depends on scenario conditions. Cost figures presented in the l i t e r a t u r e should t h u s be considered with care, as scenario conditions determine their validity. The general picture from Table 6.2 shows the greatest cost-effective potential in electricity generation and in the residential and commercial sector. Shifts in the t r a n s p o r t a t i o n sector prove to be very costly, while the potential for shifts in the i n d u s t r y is limited (at least concerning energy related options in the industry, integrated chain m a n a g e m e n t shows a very different picture, see p a r a g r a p h 6.3). Conversion savings like CO2-free hydrogen and methanol production are allocated in Table 6.2 to final consumption. It should be noted that the potentials presented in Table 6.2 should be used only as an indication. Also it should be noted that these figures cannot be added straightforward, as reduction options show interaction (e.g. through the assumed limited capacity of CO2 storage, see Figure 6.3).
6.2 Integrated emission reduction of CO2 and non-CO2 greenhouse gases In a s e p a r a t e d study the cost-effectiveness of CO2 emission reduction has been compared with the reduction of non-CO2 GHGs (CH4, N20, CO and halocarbons). For this study an extended MARKAL database was used. In this study emissions of GHGs which occur outside The Netherlands but are related to the Dutch final energy use were also included. Also upstream GHG emissions were included; these are emissions from mining, processing and t r a n s p o r t of energy carriers. Such s y s t e m boundaries differ from the ones commonly used for national emission accounting, but they coincide with emission definitions in full fuel cycle analysis and life cycle analysis. The greenhouse effect of emissions of different GHGs were compared using the Global Warming Potential (GWP) concept. The incorporation of non-CO2 GHGs and upstream GHG emissions in the analysis appears to affect the cost-effectiveness of reduction options. Total u p s t r e a m CO2 emissions and non-CO2 GHG emissions account for 10-15 % of total energy-related GHG emissions. Upstream CO2 emissions and CH4 emissions are dominant. The impact of non-CO2 GHGs on the optimization of CO2 emission reduction strategies was analyzed assuming a penalty on GHG emissions. In the penalty concept, the GHG emissions are valued externally with a fixed sum per equivalent ton CO2. In the calculations all GHG mitigation options cheaper t h a n this penalty are assumed to be implemented.
1045
T a b l e 6.2 A t t r a c t i v e n e s s o f m a j o r r e d u c t i o n o p t i o n s i n d i f f e r e n t s c e n a r i o s i f t h e CO2 e m i s s i o n should be reduced with 60% in 2030 Option
Potential
Performance
(Mt/year)
DZ
DK
GZ
GK
300
225
200
220
++ ++ n.a. + -
+ + ++ + -
++
++
Marginal costs (NLG/t CO2) Electricity generation: - CHP - C02 removal - N u c l e a r 2) - Renewables - More natural gas
5-10 >50 >50 10-25 5-10
++ ++ n.a. ++ +
+ + ++ + -
Transportation: - Hydrogen - Methanol - Ethanol - Electric vehicles - RME
10-25 10-25 5-10 10-25 5-10
+ + -
+ + -
Residential & commercial: - Insulation - Hydrogen - Heat pumps - Efficient appliances
10-25 10-25 10-25 < 5
++ + ++ ++
++ + ++ ++
5-10 5-10 5-10 5-10 5-10 5-10
++ ++ ++ ++
+ ++ ++
Industry: - MoreCHP - More natural gas - Heat pumps - Hydrogen - CO2 r e m o v a l - Savings 1) ++ = + = = 2) n . a . =
1)
+ ++
++
++
++
+
+
+
+
++
++
++
++
achieves maximum potential achieves limited potential not applied o p t i o n is n o t a v a i l a b l e i n t h i s s c e n a r i o
T a b l e 6.3 s h o w s t h e c o n t r i b u t i o n of d i f f e r e n t g r o u p s of o p t i o n s to t h e e m i s s i o n r e d u c t i o n f o r t w o p e n a l t i e s (100 a n d 2 0 0 N L G / t CO2 eq.). T h e a n a l y s i s w a s d o n e for two approaches. In the 'only direct CO2' approach, the non-CO2 GHG emissions and the upstream emissions are neglected. In the 'all GHG' approach these emissions are included.
1046 At the investigated emission penalties, CO 2 removal at coal fired facilities appears to reduce less direct CO2 emissions than in the 'all GHG' approach. On the other hand, renewables play a more important role in the 'all GHG' approach. For most other options, such as end-use savings and efficiency improvements, the results are less sensitive to the inclusion of non-CO2 and upstream GHG emissions. Table 6.3 Contribution of options to the reduction of direct C02 emissions in The Netherlands in cost-optimal emission reduction strategies, assuming an 'all GHG' approach and an 'only direct CO2' approach (DZ scenario, 2030) p e n a l t y = 200 NLG/t CO2 equiv.
p e n a l t y = 100 NLG/t CO2 equiv.
reduction of reduction of reduction of all GHG direct CO2 all GHG emissions emissions emissions only
reduction of direct CO2 emissions only
(Mt C02 eq) (Mt CO2 eq) (Mt CO2 eq) (Mt CO2 eq) Savings on end-use
16.0
15.3
20.7
20.1
Savings in conversion
21.5
22.5
22.1
22.9
Fossil fuel substitution
10.7
9.1
0.0
0.0
CO 2 removal, coal-fired
27.5
33.9
30.9
35.1
C02 removal, natural gasfired
0.0
0.0
39.9
29.7
Renewables
8.4
5.8
18.7
14.9
84.2
86.7
132.4
122.7
Total r e d u c t i o n
The emission levels which result from an enforced emission penalty are shown in Figure 6.4 for the year 2030, indexed to the emission level in the reference case. Note t h a t the horizontal axis has a logarithmic scale. As expected, the level of direct CO2 emissions decreases with rising emission penalties. The gradual reduction is achieved by a mix of options, with prominent roles for energy saving, savings in conversion, CO2 removal and renewables. Upstream CO2 emissions show an initial increase, but they decrease at emission penalties above 200 NLG/t CO2 equivalent. The increase is caused by shifts towards more coal with CO2
1047 removal for power generation at emission penalties between 100 and 200 NLG/t CO2 equivalent. Coal production shows relatively high upstream CO2 emissions. The p a t h of the CH4 emissions is partly a result of specific CH4 a b a t e m e n t measures, such as technical measures at offshore gas production, and partly it is a result of changes in the fuel mix. At the lowest penalties (20 and 50 NLG/t CO2 eq.) the CH4 emissions is reduced by measures at gas production facilities and by a reduced coal consumption. The strong emission decrease at 100 NLG/t CO2 eq. is a result of a move away from certain coal types and imports of n a t u r a l gas which are linked with high production emission levels. The alternatives, surface-mined coal and n a t u r a l gas transported through high technical s t a n d a r d pipelines, have lower CH4 emission levels. Replacement of cast-iron n a t u r a l gas distribution networks is attractive at 200 NLG/t CO2 eq. The increase of CH4 emissions at 500 NLG/t CO2 eq. results from an increased consumption of n a t u r a l gas which is mainly used for hydrogen production. The emissions of halocarbons show a peak at 175 NLG/t CO2 eq., caused by an increased use of heat pumps. This is offset at higher penalty levels by improvements in cooling devices t h a t reduce halocarbon emissions. 140 _J m
>m _1
120
o
,~... . . . .
~. ~ ~
~w 1 0 0
,~
upstreamC 0 2
_
"r
~.~,
..... "" . . . . . . . . . ................
"..~. ~... ~ ...... \ ......
9....
",,
8O
\ ~
",
, ...... .-......h.alocarbons
x
..................
"""..~
,9 6 0 -
x
~_z. 40
~ 20 m
I
00
20
I
60 EMISSION
I
100 PENALTY
I
200 [DFL/I"CO2
I
I
600
1000
equivalent]
Figure 6.4 Indexed emissions in the year 2030 for various greenhouse gases at different emission penalties as calculated for the DZ scenario (Dfl = NLG) 6.3 CO2 e m i s s i o n r e d u c t i o n in T h e N e t h e r l a n d s e n e r g y a n d m a t e r i a l s system The emission of CO2 is generally considered as an energy related problem. Whether this is correct depends on the point of view. For The Netherlands, the industrial m a n u f a c t u r i n g of materials and products is responsible for approximately one third of the national CO2 emissions (50-60 Mt versus 160 Mt). This part of the CO2 e m i s s i o n s can be influenced by changes in the m a t e r i a l s s y s t e m . The environmental impacts of energy systems (energy production and consumption) and m a t e r i a l s systems (materials, products and waste materials) are closely related. Oil is used as feedstock for plastics, waste is incinerated for energy recovery. Wood can either be used as construction material or energy carrier or in a sequence of both applications. An integrated approach for both systems should enable the identification of ways to reduce CO2 emissions with lower costs. Therefore the existing energy system in
1048 the MARKAL model was extended to represent the materials system also. The new model describes the whole Dutch materials system, and it includes all processes "from cradle to grave". Figure 6.5 shows the materials system model structure. All material flows have been modeled t h a t are related to end-use of materials in products in The Netherlands (see also Section 3 of this report). A large effort was put into the characterization of 29 materials, 20 product groups, 30 waste m a t e r i a l s and some 200 processes which link the material flows. CO2 reduction options in the materials system include: - industrial energy savings; - CO2 removal from industrial plants and storage; - reduction of materials consumption (e.g. re-usable packaging); materials substitution; - biogenous fibre materials; - improved waste collection and separation systems; - waste recycling, cascading and energy recovery. -
, • &5
Removal separation
"~
7Disposal ,.']
Figure 6.5 Materials system model structure Figure 6.6 shows the CO2 emission from our energy system and from our material consumption as calculated for the base case scenario DZ (so high economic growth, no nuclear, no CO2 reduction policy). As shown, the emission of CO2 due to our material use is at present equal to approximately one third of the CO2 emissions from our energy system (with national boundaries). This notion is important, as large Dutch industrial CO2 emissions are generally dismissed as being related to exports. These results show, however, that these export-related CO2 emissions are offset by import-related CO2 emissions. Figure 6.6 also shows t h a t in the base case scenario DZ the emissions from the materials system (M) are stabilized in time, while the total emissions from the energy system (E) increase. On one hand, this stabilization is caused by improved efficiency and recycling; on the other h a n d dematerialization plays an important role.
1049 If options to change the materials system are taken into account when looking for cost-effective CO2 emission reduction strategies, significant savings in costs can be achieved. The long term marginal CO2 mitigation costs decrease by NLG 50100 per ton CO2 avoided, as costly reduction options in the energy system can be avoided. Figure 6.7 shows the structure of emission reduction options in the integrated energy and materials system. Comparing Figures 6.3 and 6.7 shows what type of CO2 emission reduction options in the energy system can be avoided at certain reduction targets. Generally speaking, energy savings in conversion processes and in end-use are reduced. The largest shift is however related to CO2 recovery and storage. The storage capacity is assumed to be limited. As at a certain cost more CO2 reduction can be achieved in the integrated energy and materials system without decarbonization, less CO2 storage per PJ is required. The consequence is that the limited storage capacity is used less effectively, but at lower costs. In the MARKAL simulations, the 'hydrogen economy' (hydrogen from natural gas with CO2 removal) is introduced later, while CO2 removal at coal fired power plants is still used at higher reduction targets. The simulation also shows t ha t substitution of materials in products is an attractive option in CO2 emission reduction strategies. Some options for reduced materials consumption (e.g. in packaging) are already included in the baseline, due to other environmental policies. [Mt CO2/YEAR] 250
ENERGY SYSTEM (E) MATERIALS SYSTEM (M)
200
.
.
.
.
.
.
.
.
.
150 100 50 0
!
2000
I
I
2010
]
I
2020
I
I
2030
I
I
2040
[YEAR]
Figure 6.6 Base c a s e CO2 emissions for The Netherlands energy system (E) and the materials system (M) as function as time in the DZ scenario
1050 [Mt CO2/YEAR] 200
150
• ~
WITH CO2-REMOVAL
~
OTHERCO2-REMOVAL
~
FOSSILFUELSUBSTITUTION
I
MATERIALSOPTIONS
RENEWABLES
H2FROMNAT.GAS MAINLYCOAL
100
,
I 1 I
50
constant 20 40 60 CO2 EMISSION REDUCTION [%]
END USE SAVINGS
80
Figure 6.7 Allocation of CO2 emission reduction in an integrated energy and materials system as calculated for the DZ scenario. The allocation is presented as function of the CO2 emission reduction target for the year 2030
INCREASE COMPARED TO BASE-CASE [%] [-7 CEMENT E~] STEEL PLASTIC WOOD ALUMINIUM
60 40 20 0 -20 -40
2000
2010
2020 [YEAR]
2030
Figure 6.8 Shifts in materials consumption in the DZ scenario as function of time, assuming a CO2 emission reduction target for the year 2030 of 60% Figure 6.8 quantifies the material substitution effects. The main shifts are in the construction and transportation areas. Traditional brick/concrete buildings are
1051 replaced by wooden skeleton buildings. The energy consumption per ton for brick and c e m e n t production is relatively low, compared to other m a t e r i a l s . The relatively high CO2 emission for traditional buildings is caused by the large amount of materials t h a t is required per house and because of inorganic CO2 emissions from c e m e n t production. In the t r a n s p o r t a t i o n sector, cars and trucks shift towards more aluminum and plastic is used instead of wooden pallets and crates. The fuel savings due to lighter constructions are in this area the main drive. The net result of materials substitution is a decrease in the use of cement, while the use of wood and a l u m i n u m increase after 2015. The use of steel and plastics remains constant. It should be noted t h a t for different products these results prove to be very sensitive to assumptions concerning assembly costs. Other shifts in the materials system occur in materials production as a result of shifts from one production technology to another. Shifts also occur as a result of changes in waste m a n a g e m e n t . For some materials recycling is favoured (e.g. plastics), while for others (elastomers, biogenous fibre materials) incineration seems the best solution. These shifts are not discussed in further detail in this report. 6.4 C o n c l u s i o n s The model calculations for The Netherlands 'stand alone' energy system indicate t h a t significant CO2 emission reductions are possible. Changes in the residential and commercial sectors (reduction is end-uses, high efficiency equipment such as heat pumps, etc.) and in electricity generation (fuel switching, cogeneration, etc.) appear more cost effective t h a n those in industry and transport. CO2 recovery and storage options are relatively cost effective, but are to be considered probably as t r a n s i e n t towards more sustainable configurations. Biomass and wind provide relatively cheap renewable energy, but have limited potential. In The Netherlands photovoltaic solar energy could serve as backstop technology only; it has a large potential but most probably also high costs, even in the year 2030. In The N e t h e r l a n d s it seems feasible to achieve drastic CO2 emission reductions (up to 70%) in the period 1990-2030 at marginal costs of NLG 250 till 400 per ton CO2 avoided. The following strategies are identified as 'robust' to severe CO2 constraints: a) the application of energy saving m e a s u r e s , like t h e r m a l i n s u l a t i o n in residential and commercial sector; b) the use of selected renewable energy options, like g e o t h e r m a l and solar heating, biogas, hydropower, wood power stations and wind turbines; c) the removal of CO2 from power plants and in other sectors, e.g. combined with hydrogen production; d) the application of heat pumps, for space heating and hot w a t e r supply in the residential and commercial sector; e) the use of fuel cells, like for automotive purposes in the transport sector; f) the replacement of fossil fuel use by electricity, like in electric heat pumps;; g) the use of hydrogen, as substitute for natural gas in stationary applications, in fuel cells and as alternative automotive fuel in vehicles and aircrai~s. The preferred measures are not significantly influenced by taking u p s t r e a m CO2 and other greenhouse gas emissions into account. Including u p s t r e a m CO2 and m e t h a n e emissions m a k e s a noticeable difference when assessing options to
1052 reduce GHG emissions from The Netherlands energy system. Nitrous oxide, carbon monoxide and halocarbons, however, are far less important. The manufacturing of materials and products is associated with considerable energy use. Today it constitutes approximately one third of The Netherlands CO2 emissions. Changes in material flows and material technologies can contribute importantly to cost effective CO2 mitigation strategies. Interaction between the materials system and the energy system at large are shown to be of importance and require further attention. The integrated assessment of energy and materials systems reveals more cost effective mitigation options than were found in the energy system alone, especially at higher reduction percentages. The reduction is achieved through shifts in material production and waste handling and through materials substitution in products. The impact on materials consumption seems most significant for cement (reduced), timber and aluminum (both increased). For steel and plastic, the net effect is balanced, but shifts between applications do Occur.
International trade issues complicate the design and implementation of integrated chain management policies, especially for open, trade oriented economies like The Netherlands. Therefore, extending the coverage of this type of studies to the European level is required. This would also provide valuable extra insights into the interactions between a broader array of energy system configurations and materials systems. 7.
REFERENCES
Blok, K. 1993. Final Report of the Integrated Research Program on Carbon Dioxide Recovery and Storage. Utrecht University. Blok, K., E. Worrell, R. Cuelenaere and W.C. Turkenburg, 1993. The costeffectiveness of CO2 emission reduction achieved by energy conservation. Energy Policy 21: 158-175. Blonk, T.J. and R. van Duin, 1992. Materiaalverbruik in Nederland (material use in the Netherland). Bureau B&G, Rotterdam. Blonk, T.J. and R. van Duin, 1994. Verkenning van de ontwikkelingen in materiaalverbruik tussen nu en 2000 (Analysis of developments in the use of materials between now and 2000). Bureau B&G, Rotterdam. Blonk, T.J., R. van Duin and P.A. Okken, 1991. CO2 en materialen (CO2 and materials). Netherlands Energy Research Foundation, Petten. Czepiel, P.M., P.M. Crill, and R.C. Harriss, 1993. Methane Emissions from Municipal Waste water Treatment Processes. Environmental Science and Technology 27, (hr. 12): 2472-2477. De Jong, P. and M. Wolsink, 1993. De structuur van de Nederlandse afvalsector (the structure of the Dutch waste sector). Onderzoeksreeks nr. 69. Interfacultaire Vakgroep Milieukunde, University of Amsterdam. De Beer J.G., 1994. Long term energy efficiency improvement in the industry. IIASA, Luxemburg, Austria (in press). De Beer, J.G., E. Worrell and K. Blok, 1993a. Energiebesparing in de Nederlandse industrie op de lunge termijn (Energy conservation in The Netherlands industrry in the long term). Utrecht University.
1053 De Beer, J.G., E. Worrell, K. Blok and R. Cuelenaere, 1993b. ICARUS-2.1; het potentieel van energiebesparing in Nederland tot 2000 en 2010 (ICARUS-2.1; the potential of energy conservation in The Netherlands till 2000 and 2010). Utrecht University. De Beer, J.G., E. Worrell and K. Blok, 1993c. Energy conservation in the paper and board industry. Utrecht University. De Beer, J.G., E. Worrell and K. Blok, 1994. ICARUS-3, potentials for energy efficiency improvement in The Netherlands till 2015. Utrecht University (in press). Gielen D.J. and P.A. Okken, 1994. Optimization of integrated energy and materials systems. Netherlands Energy Research Foundation, Petten, The Netherlands. Gillissen M. and H. Opschoor, 1994. Energy conservation and investment behaviour: an emperical analysis of influential factors and attitudes. Free University, Amsterdam, (in press). Houghton, G.J. Jenkins and J.J. Ephraums (eds.), 1990. Climate Change; the IPCC Scientific Assessment. IPCC Working Group I, Cambridge University Press, Great Britain. IPCC Working Group I, 1994. Report on Radiative Forcing and Global Warming Potentials (in press). Johansson, T.B., H. Kelly, A.K.N. Reddy and R.H. Williams (eds.), 1993. Renewable Energy, Sources for Fuels and Electricity. Island Press, Washington. Koetzier, H., S.P.N. van Rijen and H. Bresser, 1992. Calculations on Some Complete Systems for Electricity Production with CO2 recovery, Transport and Storage. KEMA, Arnhem, (In Dutch). Kram, T., 1993. National energy options for reducing CO2 emissions; Volume 1: the international connection. Report of ETSAP/Annex IV. Netherlands Energy Research Foundation, Petten. Lettinga, G. and A.C. Van Haandel, 1993. Anaerobic Digestion for Energy Production and Environmental Protection. In: T.B. Johansson et al. (eds.). 'Renewable Energy: Sources for Fuels and Electricity'. Island Press, Washington, USA, p. 817-839. Lexmond M.J. and G. Zeeman, 1994. Potential of controlled anaerobic waste water treatment in order to reduce the global emissions of methane and carbon dioxide. In: J. van Ham et al. (eds). Non-CO2 Greenhouse Gases. Kluwer Academic Publishers, p. 411-419. Moll, H.C., 1993. Energy counts and materials matter in models for sustainable development. Thesis, Groningen University. Mulder, H.A.J., 1993. Werkdocument metalen en verf (Working document metals and paint). IVM, Groningen University. Nonhebel, S., 1994. A simple method to estimate regionally average yields of biomass crops. Dept. of Theoretical Production Ecology, Wageningen Agricultural University, (in press). Okken, P.A., T. Kram, P. Lako, J.R. Ybema, J. van Doorn and D. Gerbers, 1992. Drastische CO2 reductie - hoe is het mogelijk? (drastic CO2 reduction - how can it be achieved?) Netherlands Energy Research Foundation, Petten. Okken, P.A., J.R. Ybema, T. Kram, P. Lako, D. Gerbers, 1994. Energy systems and CO2 constraints. Netherlands Energy Research Foundation, Petten. Smit, R., J. de Beer, E. Worell and K. Blok, 1994. Long term energy efficiency improvement: technology descriptions. Utrecht University.
1054 Thorneloe, S.A., 1993. Waste water treatment, from A.R. van Amstel (ed.). 'Methane and Nitrous Oxide, National Institute of Public Health and Environmental Protection (RIVM), Bilthoven, p. 115-130. UNSEGED, 1992. United Nations Solar Energy Group on Environment and Development. Solar Energy: a Strategy in Support of Environment and Development. United Nations, New York. Van Zeijts, H., E.B. Oosterveld and E.A. Timmerman, 1994. Kan de landbouw schone energie leveren? (Can Dutch agriculture provide clean energy?). Centrum voor Landbouw en Milieu, Utrecht. Van den Heuvel, E.J.M.T., H.E.M. Stassen and F.S. Feil, 1994. Conversieroutes voor energiegewassen (Conversion routes of energy crops). BTG Biomass Technology Group BV, Enschede. Van Wijk, A., B. Meuleman, J. de Beer, K. Blok, W. Turkenburg and E. Worrell, 1994. Sustainable Energy System; Technologies to Reduce the CO2 Emission. Utrecht University. WEC, 1993. World Energy Council, Energy for Tomorrow's World. St. Martin's Press, London, U.K. WEC Study Group on Renewable Energy Resources. 1993. Renewable Energy Resources: Opportunities and Constraints 1990-2020. World Energy Council, London, U.K. World Bank, 1992. Development and the Environment. World Development Report 1992, Oxford University Press, U.K. Worrell, E., 1993. Potentials for improved use of industrial energy and materials. Thesis, Utrecht University. Ybema J.R. and P.A. Okken, 1993. Full fuel chains and the basket of greenhouse gases. Netherlands Energy Research Foundation, Petten.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1055
D I S C U S S I O N O N N R P A S S E S S M E N T R E P O R T 'rENERGY S U P P L Y A N D E N E R G Y A N D M A T E R I A L S SAVINGS"
K. Blok (rapporteur) D e p a r t m e n t of Science, Technology and Society, Utrecht University, P a d u a l a a n 14, 3584 CH Utrecht, The Netherlands 1. I N T R O D U C T I O N
In a break-out sessions (session 20.I) at the conference, one of NRP's assessment reports "Energy demand and supply mitigation options" was discussed. The report was p r e s e n t e d by its author, prof. dr. Wim Turkenburg. On the basis of the presentation discussion took place. In this report the m a i n issues of the session will be discussed. T u r k e n b u r g ' s presentation will only be discussed briefly, as his complete assessment report is published elsewhere in these proceedings [1]. 2. R E S U L T S F R O M T H E N R P
T u r k e n b u r g started his presentation by referring to the recent emission scenarios of the IPCC. Although there have been changes with respect to the 1990 scenarios the conclusion is still valid t h a t considerable emission reductions with respect to 'business-as-usuar developments of emissions are necessary. He estimated t h a t for the next century m e a s u r e s have to be t a k e n t h a t lead to a total emission reduction of 1200 GtC (4400 Gtonnes of CO2). A range of options is available t h a t m a y contribute to such reductions. A n u m b e r of these options have been subject to study in the National Research Programme. A list of the subjects discussed is as follows: 9 Energy efficiency improvement. 9 Material efficiency improvement 9 Biomass as an energy source 9 Decarbonization of fuels and fuel gases 9 Integrated assessments of energy and material systems This presentation and discussion is m e a n t to react on the quality of the work and to determine what is missing (e.g. is the focus on CO2 right?). 3. D I S C U S S I O N E n e r g y efficiency i m p r o v e m e n t
Q. Up from the midst of the seventies there was a stabilization of energy consumption and on average an economic growth. Is it a simplification to say t h a t this difference is due to energy efficiency improvement? A. It is, but in the eighties energy efficiency improvement is responsible for about three quarters of the effect.
1056 Q. With respect to long-term efficiency improvement, also the relation between energy efficiency and s t r u c t u r a l change should be considered. Especially the rebound effect may be important: due to efficiency improvement energy services become cheaper and will be used more. A. This was not the subject of the study. It was shown t h a t in principle major efficiency improvements can be attained. Of course, also other effects, like the one mentioned, may play a role. Q. Will all the barriers for energy conservation be translated to cost categories in ICARUS? A. (by Kornelis Blok) This seems h a r d l y possible. But we try to introduce c h a r a c t e r i z a t i o n s of energy conservation m e a s u r e s (e.g. typical size of the i n v e s t m e n t ) t h a t can be used for calculation of, e.g., transaction costs u n d e r various circumstances. It should be noted t h a t transaction costs depend on the way energy efficiency improvement is stimulated. For instance, if standard setting is applied, the information phase for investments may become cheaper as the number of options to be considered is reduced.
Biomass energy Q. In the case of biomass energy production, is the energy required to produce fertilizer, for harvesting, etc. t a k e n into account? Does biomass result in net savings of primary fossil energy? A. This issue has been investigated. It turns out t h a t for all the biomass energy options considered, there is a positive net gain. The net gain is, in terms of GJ per h e c t a r e , h i g h e s t for production of wood, M i s c a n t h u s , etc. combined w i t h gasification. For production of transportation fuels, e.g. rapeseed oil, the net gain is smaller.
Integrative studies Q. With respect to the integrative studies, a ranking of options (amount of CO2 avoided, costs per ton of CO2) is desirable; this can not be obtained from the Markal results. A. This is true; but such figures can already be obtained from the analysis of individual options. Tom Kram, one of the authors of the integrative study, adds t h a t the advantage of an integrated model like Markal is that the mutual influence of various options is taken into account. Q. O t h e r studies, for instance the reports written by McKinsey, give different results for the costs per tonne of CO2 avoided. A. T h e r e m a y be various reasons for differences, and one should always be cautious in using the figures. But also the aim of the studies may differ. McKinsey carried out its analysis for the Netherlands energy distribution companies for the evaluation of concrete plans, and used a discount rate of 10 or 15%. The M a r k a l evaluations have a national-economic perspective and use real interest rates and depreciation over the lifetime. Q. There is a need for more extended overviews listing all the options with their costs. A. There are overviews, Markal is a way of integration, ICARUS gives an overview of demand-side saving potentials. It should be noted that thousands of data may be needed. For instance, in Markal, wind energy may be represented by only a few figures; however, the costs of wind energy at the coast will be very different from the costs of wind energy inside the country. So, for a good description one should
1057 have an extended database only for wind energy already. It seems impossible to cover all these data in one set. Future research Q. The winning of renewable energy may be economically more attractive in other countries in Europe (e.g. wind energy in Scotland, solar energy in Spain). Is this taken into account? A. This first phase of the NRP has a national perspective, although some studies were carried out, e.g. into wind energy potentials in European countries. It is agreed t h a t the international perspective should have much more attention in future studies. Other aspects are international exchange of energy-intensive materials, and of course b u r d e n s h a r i n g in international climate agreements. A better understanding of potentials and costs per country (determined on a common basis of a n a l y s i s ) is also i m p o r t a n t for e v a l u a t i n g the possibilities of j o i n t implementation. Within parts of NRP not reported here, some studies for other countries have been done. For instance, a study into CO2 emission reduction for China was conducted. Anyway, more international cooperation in this field is important. Q. Why are other greenhouse gases than CO2 not taken into account? A. The reason to pay so much attention to CO2 is t h a t it is the most important greenhouse gas, and this is still the case. In phase 2 of the programme there is room for research into emission reduction of other greenhouse gases. It should also be taken into account that there are many research programmes in the fields of energy and environment. In general the research in the NRP should have a clear added value for a better understanding of the climate problem and the response options.
REFERENCES 1. W.C. Turkenburg, Energy demand and supply mitigation options, Proc. of the International Conference on Climate Change Research, ....
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1061
Long-term energy and materials strategies for CO 2 reduction D.J. Gielen, T. Kram, P. Lako, J.R. Ybema; ECN - Policy Studies, P.O. Box 1, 1755 ZG Petten, The Netherlands
Abstract Long-term greenhouse gas emission reduction strategies for the Netherlands were studied, using a MARKAL energy model. The EMS study identifies longterm technological options for greenhouse gas emission reduction and assesses their cost-effectiveness, taking interactions between technologies into account. The project consisted of three parts: carbon dioxyde (CO 2) emission reduction in the energy system, integrated reduction of greenhouse gases from the energy system with consideration of upstream emissions and CO 2 reduction in the integrated energy and materials system. 500 energy technologies were assessed for their reduction potential. Significant emission reduction seems possible, but it takes an array of measures to achieve this goal at acceptable costs. Considering the full fuel chain and other greenhouse gases does not significantly affect the optimal solution. CO e emission reduction in the materials system proves to be a promising approach, resulting in significant CO 2 emission reduction at lower cost as for the stand-alone energy system.
1. I N T R O D U C T I O N
Reduction of anthropogenic greenhouse gas emissions is widely considered an essential measure to reduce the risk of enhanced global warming. The most important greenhouse gas (GHG) is carbon dioxide (CO2), but other gases like methane (CH 4) and nitrous oxyde (N20) are also relevant. Most GHG emissions in the Netherlands are related to the use of fossil energy carriers. An array of technological measures is available to reduce these emissions, ranging from fuel shifts and renewable energy sources to improved waste m a n a g e m e n t or shifts in materials use. Beforehand, it is unclear how much CO 2 reduction can be achieved and at what cost. The assessment of emission reduction options is complicated because technologies are linked in the energy system (i.e. the total of energy supply, conversion, distribution and use). If e.g. the electricity production becomes less CO 2 intensive, electricity savings become less attractive for CO 2 reduction. The goal of the EMS project (Energy and Materials use Scenarios for reduction of CO 2 and other GHGs) is to assess the potential and costeffectiveness of reduction options in an integrated approach, taking the whole energy system into account. Three main parts of the project are considered
1062 here: 1. Integrated assessment of CO 2 emission reduction in the Netherlands energy system [1,2], 2. Integrated assessment of GHG emission reduction in the Netherlands energy system [3], 3. Integrated assessment of CO 2 emission reduction in the energy and the materials system [4]. The instrument that is used for this study is the MARKAL (MARket ALlocation) model. MARKAL is internationally used in IEA/ETSAP (International Energy Agency/Energy Technology Systems Analysis Programme) [5]. This linear programming (LP) model can be used to develop integrated energy strategies, taking environmental restrictions into account (CO2, NOx, SO2). A MARKAL model for the Netherlands has previously been used to support governmental energy policy strategies and national energy technology research programmes. The model contains a database with approximately 500 energy supply and demand technologies. Each technology is characterised by technical, financial and environmental parameters. The model is used to calculate the least-cost system configuration for the period 2000 - 2040, meeting exogenously defined national energy service demands and emission reduction targets. Earlier studies with this tool concentrated on emission reduction for SO 2 and NO x (e.g. [6]). In this project, the model is extended to study the reduction of GHG emissions, especially the reduction of CO 2 emissions.
2. C O 2 E M I S S I O N R E D U C T I O N IN A S T A N D - A L O N E E N E R G Y S Y S T E M
An extended database of energy technologies is evaluated with respect to the cost-effective potential for CO e reduction. High (D) and low (G) economic growth scenarios with and without nuclear energy are studied (called DK, DZ and GK, GZ, respectively). Figure 1 shows the CO 2 emission from the energy system in the four base cases (lines). The increase from 2000 to 2030 ranges from 3% to 32%. The increase of baseline emissions of CO e is a result of various mechanisms. Energy demand increases and the share of coal in the primary energy mix grows, especially for electricity generation and methanol production. On the other hand the baseline includes significant efficiency improvements and enduse savings. Several CO e emission reduction paths were studied, also shown in figure 1 (dashed lines). The CO 2 constraints for the reduction cases follow linear paths from 2000 to 2030 and then stabilize. Reduction percentages imposed for the year 2030 include stabilisation (0%), 20%, 40%, 50%, 60%, 70% and 80%, compared to the emission in 2000.
1063 [Mt CO2/YEAR] 250 DZ
DK 200
GZ -~I-G K CONSTANT
150
100
50 -.. . . . . . . . . . . . . . . . . . . . . .
I
1990
I
2000
I
I
2010
I
2020
80%
I
2030
2040
Figure 1: B a s e case CO e e m i s s i o n s and emission reduction constraints [1].
In 2030, significant CO e reduction (up to 70%) can be achieved at m a r g i n a l reduction costs below 400 NLG/t CO e (see figure 2). However, as CO e removal technologies will h a r d l y be available in 2010, the a t t a i n a b l e emission reduction is m u c h smaller and m a r g i n a l costs increase rapidly in t h a t year, in p a r t i c u l a r in the D scenarios. S u b s t a n t i a l CO e reduction proves to be a m a t t e r of longt e r m planning.
[DFL/t 500
400
CO2]
-
DZ
_-
i
/
i 9
/ p -/
i
/,"
~ ----i--......~ ..... GK
,oo200
]/ -
O
0
. ,~ . . . . ~
..~-~_. .......... .,~.-w-..'7. A .... .....
I
I
I
20
40
60
002
EMISSION
REDUCTION
Figure 2" M a r g i n a l costs of CO 2 reduction 2030 [1].
[~
80
1064 The impact of CO 2 reduction on the energy system is very significant. On the energy supply side, the potential for renewable energy is limited due to climate and geographic conditions. Within 15 years it is expected that the Netherlands will have capacity for storage of CO 2 in depleted natural gas fields or aquifers allowing for CO e removal options. Such options have been considered for electricity generation and for synfuel production (e.g. hydrogen). CO e storage plays a crucial role for achieving significant emission reduction at acceptable costs. At lower emission reduction levels, CO 2 is removed from power plant flue gases. At higher emission reduction levels, synfuels are introduced and a "hydrogen economy" develops. As the Dutch energy system already depends on natural gas to a large extent, the potential for fuel substitution as CO 2 reduction option is limited. On the energy demand and conversion side, significant savings are still achievable, e.g. through better insulation of buildings. Figure 3 shows a breakdown of emission reduction in 2030 in the DZ scenario at increasing emission reduction targets. Most emission reduction can be achieved at the supply side. While the potential for demand side CO 2 reduction is limited, most of these options are already included in the baseline.
[Mt CO2/YEAR] 200
~
RENEWABLES
~
H2 FROMNAT.GAS WITHCO2 REMOVAL OTHERCO2 REMOVAL MAINLYCOAL
~
150
~ I
100
FOSSILFUELSUBSTITUTION I CONVERSIONSAVINGS
~ E N D USESAVINGS
50
constant 20 40 60 CO2 EMISSION REDUCTION [%]
80
Figure 3: Emission reduction allocation for the stand-alone energy system (DZ 2030) [1]
1065
A s t h e m o d e l c o n t a i n s a w h o l e a r r a y of r e d u c t i o n o p t i o n s , a d e t a i l e d d i s c u s s i o n of t h e a t t r a c t i v e n e s s of specific o p t i o n s is b e y o n d t h e s c o p e of t h i s paper.
Table 1 Attractiveness c a s e 2030).
of m a j o r r e d u c t i o n o p t i o n s i n d i f f e r e n t s c e n a r i o s ( 6 0 % r e d u c t i o n
Option
Potential [Mt/year]
M a r g i n a l c o s t s [ D F L / t C O 2]
Performancel DK GZ
GK
300
225
200
220
++ ++ n.a. + -
+ + ++ + -
++
++
DZ
Electricity generation CHP C02 removal Nuclear 2 Renewables More natural gas
5-10 >50 >50 10-25 5-10
++ ++ n.a. ++ +
+ + ++ + -
Transportation Hydrogen Methanol Ethanol Electric vehicles RME
10-25 10-25 5-10 10-25 5-10
+ + -
+ + -
Residential & commercial Insulation Hydrogen Heatpumps Efficient appliances
10-25 10-25 10-25 < 5
++ + ++ ++
++ + ++ ++
5-10 5-10 5-10 5-10 5-10 5-10
++ ++ ++ ++
+ ++ ++
Industry More CHP More natural gas Heatpumps Hydrogen CO 2 removal Savings ++ + n.a.
= = = =
achieves maximum potential achieves limited potential not applied o p t i o n is n o t a v a i l a b l e in t h i s s c e n a r i o
+ ++
++
++
+
+
+
+
++
++
++
++
1066
Table 1 shows an overview of reduction options in different scenarios. A general cost figure in DFL per tonne CO e for each option is not available, as the costs and the CO 2 reducing potential depend on the scenario conditions. For example the reducing potential for electricity generating or consuming technologies and for combined heat and power generation (CHP) are largely determined by reference technologies and load patterns. As these conditions vary between scenarios, the attractiveness varies accordingly. Cost figures per tonne CO 2 in the literature should thus be considered with care, as scenario conditions determine their validity. The general picture from table 1 shows the greatest cost-effective potential in electricity generation and in the residential and commercial sector. Shifts in the transportation sector prove to be very costly, while the potential for shifts in the industry is limited (at least concerning energy related options in the industry, integrated chain management shows a very different picture, see section 4). Conversion savings like e.g. COe-free hydrogen and methanol production are in table 1 allocated to final consumption. The potential in table 1 is only an indication; these figures cannot be added straightforward as reduction options show interaction (e.g. through limited CO 2 storage potential, see figure 3).
3. I N T E G R A T E D R E D U C T I O N OF G R E E N H O U S E G A S E S
The sensitivity of emission reduction results from consideration of non-CO e GHGs (CH 4, N20, CO and halocarbons), was studied with an extended MARKAL database. Emissions of GHGs which occur outside the Netherlands, but which are related to the Dutch final energy use were also included. The upstream GHG emissions include emissions from mining, processing and transport of energy carriers. Such system boundaries differ from the ones commonly used for national emission accounting, but they coincide with emission definitions in full fuel cycle analysis and life cycle analysis. The warming impacts of emissions of different GHGs were compared using the Global Warming Potential (GWP) concept. Incorporation of non-CO e GHGs and upstream GHG emissions in the analysis appears to affect the effectiveness of reduction options. Total upstream CO 2 emissions and non-CO 2 GHG emissions account for 10-15 % of total energyrelated GHG emissions. Upstream COe emissions and CH 4 emissions are dominant. The impact of other greenhouse gases on the optimisation was analysed, using a CO 2 "penalty". In the penalty concept, CO 2 emissions are valued externally with a fixed sum per tonne CO e. COe emissions are minimised again in a cost-effective way. Table 2 shows the contribution of groups of options to emission reduction in two approaches. In the 'only direct CO 2' approach the non-CO 2 GHG emissions and the upstream emissions have been neglected, while in the 'all GHG' approach these emissions were included. At two emission penalties (100 and 200 DFL/tCO2), CO 2 removal at coal-fired facilities appears to reduce less direct COe emissions than in the 'all GHG' approach. On the other hand, renewables play a more important role in the 'all GHG' approach. For most other options, such as end-use savings and efficiency
1067 improvements the results are less sensitive to the inclusion of non-CO 2 GHG and upstream GHG emissions.
Table 2 Contribution of options to reduction of direct CO 2 emissions in cost-optimal emission reduction strategies in 'all GHG' approach and in 'only direct CO 2' approach (DZ scenario, 2030).
Savings on end-use Savings in conversion Fossil fuel substitution CO 2 removal, coal-fired CO 2 removal natural gas-fired
100 DFL/tCO 2 penalty all GHGs only direct CO2 16.0 15.3 21.5 22.5
Renewables
Total Reduction
200 DFL/tCO 2 penalty all CHGs only direct CO2 20.7 20.1 22.1 22.9
10.7 27.5 0.0
9.1 33.9 0.0
0.0 30.9 39.9
0.0 35.1 29.7
8.4
5.8
18.7
14.9
84.2
86.7
132.4
122.7
140 ..J
uJ
>uJ J o
upstream
CO2
120 s~
~ "~ "~ ~" "~ " " ' " ~ ~ " " ' "
~ ~ ~ ,_ --',,Z,..X. ... " " ' " " ' " - . .
o
r 100 """ ".~.
W 0
" ~.d ~ ~,.,~,,,,
wZ rr
w
"~"...... "",,
.....'"
..'"" 9......
'~'.. \".....
...,"" ",
\
,,,,
80
".....
\\
\\
"'""-..
halocarbons .........................
w IT IT
oLL
60
"~
-
_oz'Z2040 ~ _
\
x\
'"",...
Q
O9 CO __ W
0
0
i 20
i 50 EMISSION
i 100 PENALTY
I 200
i 500
I 1000
[DFLII'CO2 equivalent]
Figure 4: Indexed emissions for various greenhouse gases at different emission penalties (DZ 2030).
The emission levels which resulted from the enforcement of penalties are shown in figure 4 for the year 2030, indexed to the emission level in the reference case. Note that the horizontal axis has a logarithmic scale. As expected the levels of direct CO 2 emissions decrease with rising emission
1068 penalties. The gradual reduction is achieved by a mix of options, with prominent roles for energy saving, savings in conversion, CO 2 removal and renewables. Upstream CO e emissions show an initial increase, but decrease at emission penalties above 200 DFL/t CO e. The increase is caused by shifts towards more coal with CO2 removal for power generation at emission penaties between 100 and 200 DFL/t CO 2. Coal production shows relatively high u p s t r e a m CO 2 emissions. The path of the CH 4 emissions is partly a result of specific CH 4 a b a t e m e n t measures, such as technical measures at offshore gas production, and the path is partly a result of changes in the fuel mix. At the lowest penalties (20 and 50 DFL/tCO 2) CH 4 emissions will be reduced by measures at gas production facilities and by a reduced coal consumption. The strong emission decrease at 100 DFL/tCO 2 is a result of a move away from certain coal types and n a t u r a l gas imports which are linked with high production emission levels. The alternatives, surface-mined coal and natural gas transported through high technical standard pipelines, have lower CH 4 emission levels. Replacement of cast-iron natural gas distribution networks is attractive at 200 DFL/tCO e. The increase of CH 4 emissions at 500 DFL/tCO 2 results from the increased consumption of natural gas which is mainly used for hydrogen production. The emissions of halocarbons show a peak at 175 DFL/tCO2, caused by an increased use of heatpumps. This is offset at higher penalty levels by improvements in cooling devices that reduce halocarbon emissions.
4. CO 2 E M I S S I O N R E D U C T I O N IN T H E I N T E G R A T E D E N E R G Y A N D MATERIALS SYSTEM
While CO 2 is generally considered as an energy related problem, this depends on the point of view. For the Netherlands, industrial materials production is responsible for approximately one third of the national CO 2 emissions (50-60 vs. 160 Mt). This part of the CO 2 emissions can be influenced by changes in the materials system. The environmental impacts of energy systems (energy production and consumption) and materials systems (materials, products and waste materials) are closely related. Oil is used as feedstock for plastics, waste is incinerated for energy recovery. Wood can either be used as construction material or energy carrier or in a sequence of both applications. An integrated approach for both systems should enable the identification of ways to reduce CO 2 emissions with lower costs. The existing energy system model was extended to represent the materials system. The model describes the whole Dutch materials system, and it includes all processes "from cradle to grave"; figure 5 shows the materials system model structure. All material flows are modeled that are related to end-use of materials in products in the Netherlands.
1069
1 Primary production
2 Recycling [
Material
3 Product Assembly
Product
4 Product Use
5 Removal & separation
6 Energy recovery
Waste material 7 Disposal
Figure 5: Materials system model structure.
A large effort was put into the characterisation of 29 materials, 20 product groups and 30 waste materials and some 200 processes which link the material flows. Appendix 1 shows the relation between materials and products. CO 2 reduction options in the materials system include: industrial energy savings; - CO e removal from industrial plants and storage; reduction of materials consumption (e.g. re-usable packaging); materials substitution; biogenous fibre materials; improved waste collection and separation systems; waste recycling, cascading and energy recovery. Figure 6 shows the model results for CO e emissions in the base-case (no CO 2 reduction). The materials system that is defined on the end-use principle is again responsible for approx, one third of the CO e emissions from the energy system (with national boundaries). This is important, as large Dutch industrial CO 2 emissions are generally dismissed as being related to exports. These results prove however t h a t these export-related CO 2 emissions are offset by importrelated CO 2 emissions. The emissions from the materials system (M) are stabilised in time, while the total emissions from the energy system (E) increase. On one hand, this stabilisation is caused by improved efficiency and recycling; on the other hand dematerialisation plays an important role. -
-
-
-
-
-
1070 [Mt CO2/YEAR] 250
ENERGY SYSTEM (E)
200
MATERIALS SYSTEM (M) ......
150
100
50
0
I
2000
I
I
2010
I
I
2020 [YEAR]
I
I
I
2030
I
2040
Figure 6: Base case CO 2 emissions for the energy system (E) and the materials system (M) (DZ).
[Mt CO2/YEAR] 200
RENEWABLES
150
~
WITH CO2-REMOVAL
~
OTHERCO2-REMOVAL
H2 FROMNAT.GAS
MAINLY COAL FOSSIL FUEL SUBSTITUTION
iiii!!ii:i!ii:ii:i:i!ii:i:i}ii!i!i:i:i:i!i::!ii!
i! m
100
i[
~TE~,~,s O~T,O~S
I ~176176
s~'~s
50
constant 20 40 60 CO2 EMISSION REDUCTION [%]
80
Figure 7: Emission reduction allocation for the integrated energy and materials system (DZ 2030).
1071
Considering emission reduction in the materials system results in significant cost reduction. The long term marginal CO 2 reduction costs decrease by NLG 50-100 as costly reduction options in the energy system can be avoided. Figure 7 shows the structure of emission reduction options in the integrated energy and materials system. Comparing figures 3 and 7 shows what type of CO 2 emission reduction options in the energy system can be avoided at certain reduction targets. Generally speaking, savings in conversion and end use are reduced. The largest shift is however related to CO 2 storage. The storage capacity is limited. As more CO 2 reduction can be achieved in the integrated energy and materials system at certain costs without storage, less storage per P J is required. The consequence is that the limited storage capacity can be used less effectively, but at lower costs. The 'hydrogen economy' (hydrogen from natural gas with CO 2 removal) is introduced later, while CO 2 removal at coal fired power plants is still used at higher reduction targets compared to the results for the stand-alone energy system.
INCREASE COMPARED TO BASE-CASE [%] CEMENT E'~ STEEL PLASTIC ~/~ WOOD ALUMINIUM
60 40 20
~20 -40
"
i
2000
i
i
2010
i
i
2020
I
I
i
~
20:30
[YEAR] Figure 8: Shifts in materials consumption due to CO 2 emission reduction (DZ -6O%).
As a result of CO 2 reduction, materials in products are substituted. Some options for reduced materials consumption (e.g. in packaging) are already included in the baseline; other environmental policies may cause such a shift.
1072 Figure 8 quantifies material substitution effects due to CO 2 reduction. The main shifts are in the construction and transportation areas. Traditional brick/concrete buildings are replaced by wooden skeleton buildings. The energy consumption per tonne for brick and cement production is relatively low, compared to other materials. The relatively high CO 2 emission for traditional buildings is caused by the large amount of materials that is required per house and because of inorganic CO 2 emissions from cement production. In the transportation sector, cars and trucks shift towards more aluminium and plastic is used instead of wooden pallets and crates. The fuel savings due to light weight constructions are in this area the main drive. The net result of materials substitution is a decrease in the use of cement, while the use of wood and aluminium increase after 2015. The use of steel and plastics remains constant. These results prove to be very sensitive to assumptions concerning assembly costs for different product options. The impact of 60% CO 2 reduction on product life cycle costs is generally below 10 %. If e.g. assembly costs for an aluminium car are 15% higher as for a steel car, the shift from steel to aluminium is not cost effective. The uncertainty range in future production costs is however in this order of magnitude. This problem occurs for most products. Other shifts in the materials system occur in materials production due to shifts from one production technology to another and occur also in waste management. For some materials recycling is favoured (e.g. plastics), while for others (elastomeres, biogenous fibre materials) incineration seems the best solution. These shifts are not discussed in further detail in this paper.
5. C O N C L U S I O N S
The model calculations for the stand alone energy system indicate that significant CO 2 emission reduction is possible. Changes in the residential and commercial sectors (conservation, high efficiency equipment such as heatpumps, etc.) and in electricity generation (fuel switching, cogeneration, etc.) appear more cost effective than those in industry and transport. Significant savings, mainly in the residential and commercial sectors, are still possible but will need to be supplemented by measures in the supply sectors to reach more ambitious targets. CO 2 removal and storage options are relatively cost effective, but are to be considered as transient towards more sustainable configurations only. Biomass and wind provide relatively cheap renewable energy, but have limited potential. Photovoltaic solar energy could serve as backstop technology only: large potential but high costs. The preferred measures are not significantly influenced by taking upstream CO2 and other greenhouse gases into account. Including upstream CO2 and m e t h a n e emissions makes a noticeable difference, but nitrous oxide, carbon monoxide and halocarbons are less important when assessing future Netherlands energy systems. Production of materials is associated with considerable energy use. Today it constitutes approximately one third of the Netherlands CO e emissions. Changes in material flows and material systems technologies appear important
1073 contributors to cost effective C02 reduction strategies. Model calculations indicate an increased use of aluminium and wood, while the use of cement decreases. This is caused by changes in the construction and transportation sectors. Interaction between the materials system and the energy system at large are shown to be of importance and require further attention. The integrated assessment of energy and materials systems reveals more cost effective options than were found in the energy system alone. Moreover, options within the materials system are often truly sustainable, an important feature supporting current long term policy strategies. International trade issues complicate the development and implementation of integrated chain management policies, especially for open, trade oriented economies like the Netherlands. Therefore extending the coverage of this type of studies to the European level is required. This would also provide valuable extra insights into the interactions between a broader array of energy system configurations and materials systems.
6. R E F E R E N C E S
1 P.A. Okken, T. Kram, P. Lako, J.R. Ybema, J. van Doorn, D. Gerbers: Drastische CO 2 reduktie - Hoe is het mogelijk. (Drastic CO 2 reduction) ECN-C--92-066. Petten, J a n u a r y 1993. 2 P.A. Okken, J.R. Ybema, T. Kram, P. Lako, D. Gerbers: Energy systems and CO 2 constraints. ECN-C-93-014. Petten, March 1994. 3 J.R. Ybema, P.A. Okken: Full fuel chains and the basket of greenhouse gases. ECN-C-93-050. Petten, December 1993. 4 D.J. Gielen, P.A. Okken: Optimisation of integrated energy and materials systems. ECN-C--94-010/011/012. Petten, June 1994. 5 T. Kram: National energy options for reducing CO 2 emissions, Volume 1: the international connection. A report of ETSAP/Annex IV. ECN-C--93-101. Petten, December 1993. 6 T. Kram et. al.: Koleninzetstudie (KIS). (Coal use study). ECN-C--91-072. Petten, November 1991.
1074
A P P E N D I X 1: M A T E R I A L REQUIREMENTS ALTERNATIVES
FOR
PRODUCT
3 O
O
roads Brick roads Concrete roads Tr..~.pic wood waterworks Steel waterworks Concrete waterworks Plastic waterworks Other build./infrastr. ....................................................................... .................................... ....................................................................... ....................................................................... Steel cars Aluminum cars Plastic cars .................................... ................................... ................................................................... .................................................................. tee tru s Aluminum trucks Plastic trucks Reference machinery
Max. plastic appliances Other objects at home Steel furniture Wooden furniture Other int. decoration .:.:.:.:.:.:+:.:.:.:.:.:.:.:.:.:.:.:.:.:.:+:.:.:.:.:.:.:.:.:.:.:.:.: :.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:. .................................... Steel cans Aluminum cans Glass multiple use bottle Paper/board packaging Plastic disp. packaging PET multi le use ack. One-way glass bottle Degradable plastic pack. Sanitary paper Wooden pallets Plastic pallets Other ind. packaging Plastic clothing Nat. org. clothing ........................................................... ::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: Agents Co_~.m..post N-fertilizer Paint Lubricating oil Detergentia Chlorine Na(OH)
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1075
Energy conservation and investment behaviour of firms Merlijn Gillissen and Hans Opschoor Ecological Economics Group, dept. of Spatial Economics, Vrije Universiteit Amsterdam, De Boelelaan 1105, 1081 HV Amsterdam, The Netherlands
Abstract This paper analyzes the determinants and barriers of energy conservation investment behaviour. A number of barriers were found in a literature survey. A three-phase investment model on the micro level was constructed. Hypotheses derived from the model were empirically tested by analyzing a survey of more than 300 Dutch Firms. Economic variables seem to determine investment behaviour to a large extent.
1. Background and problem description To reduce the emissions of Greenhouse gases (GHG), a reduction in CO2 emissions is necessary. Energy conservation (EC) is considered as one of the major strategies to achieve this. Industry (in its broad sense: agriculture, manufacturing, services) is one of the main users of energy and potentially an important energy conserver. A main objective of Dutch policy is to speed up the energy efficiency improvement from approx. 1% p.a. to 2.2% p.a. in the year 2000 (Nota Energiebesparing, 1989) to reduce CO2 emissions by 3-5% in 2000. It has been recognized (Blok 1991) that there exists large potentials for energy conservation in industry. Calculations with the data base ICARUS (see De Beer et aL 1993) show that the technical potential for energy conservation can be as much as 30% on average. Not all technologies are profitable. However, if one applies economic evaluation criteria, there still remains a profitable potential for energy conservation of about 20% (V.d.Werff and Opschoor 1992; Ayres 1994). Problems arise when one tries to apply the results of ICARUS to industry: large differences exist between what ICARUS indicates as profitable and what firms think is profitable. This study analyzes the differences between ICARUS's results and observed implementation behaviour of firms, in terms of determinants and barriers to the adoption of EC-technologies. The result is a theoretical implementation model which is empirically validated. In a second part of the study a set of realistic energy policy scenarios are constructed and these scenarios are applied to the implementation model of the first stage. The result here will be a simulation model that assesses the impacts of energy policy instruments on implementation behaviour and estimates how much the adoption process of EC-technologies can be accelerated and what the results are in terms of additional energy conservation.
1076 2. Methodology
A literature survey looked into investment decision in general and an application of investment theory to energy conservation and identified theoretical barriers that might arise. In this framework important theoretical determinants and barriers to energy conservation adoption have been derived and a conceptual model was constructed (section 3). Next, a survey among more than 300 Dutch firms was held. Its results were used to empirically validate or reject the hypotheses derived from the theoretical framework about the determinants of and barriers to the investment decision (section 4). One part of the survey focused on the information on and implementation of the six most applicable EC-technologies in a sector. Another part of the survey focused on variables related to the theoretical determinants and barriers. Thus, it is possible to estimate the impacts of the variables on investment behaviour (section 5). The second part of this study entails the estimation of the effects of additional energy policy on investment behaviour of firms. A set of plausible energy policy scenarios for the future (1994-2015) was constructed. These scenarios are used as an input for a simulation model that is currently built. For a schematic synthesis, see figure 1.
Technical potential o n sectoral level
l
Profitable potential o n sectoral level
Policy
l
Policy scenarios on: - energy taxes - financial incentives - direct regulation
- information c growth
S e c t o r a l model
of investment behaviour of farms for e n e r g y
I demand
t
Energy s a v e d by sector F i g u r e 1" Schematic representation
of project structure
3. Potential determinants and barriers
In a perfect world (e.g certain cash flows, free and full information, independence between technologies and unlimited access to capital markets), a profit maximizing firm would implement all available technologies that have a positive net present value. However, introducing imperfections lead to the existence of barriers that prevent firms from implementing EC-technologies. The potential barriers can be categorized in the follov'ing groups (see Gillissen, 1994a): a. e c o n o m i c barriers: i) low expected energy prices; ii) uncertainty due to
1077 expected fluctuations in energy prices; iii) low expected revenues due to low energy bill; iv) budgetary problems; v) too high required return on investment; b. physical/technology barriers: i) reduction in production quality; ii) bounded rationality; iii) "technology-lock"; iv) information gap; c. management barriers: i) no specialized personnel; ii) no interest in energy conservation by management; iii) no priority to conservation (high opportunity costs); iv) present technologies are not fully depreciated; v) lack of pressure. Potential determinants of energy conservation are, for example, firm size, the presence of an energy coordinator and R&D department. External pressure and bilateral agreements may also speed up the implementation process.
4. Modelling energy conservation implementation behaviour As a complement to ICARUS (where only EC-technologies are listed), our model provided detailed information to which extent the EC-technologies in are actually being implemented by firms. The model consists of three "modules". The first module analyses the information process of firms. Variables that describe the information capacity for energy conservation are: the number of information channels, the presence of an energy coordinator, R&D department or environmental care system. Other important variables that represent the importance of energy conservation technology information are: firm size, the energy bill, the complexity and costs of a EC-technology. Together, these variables serve as explanations for the level of information of a firm. Lack of information might lead to an information gap, which is a barrier to the adoption process. The second module analyses the economic evaluation process by firms. Technologies are judged on their expected profitability. The profitability as perceived by the firm might differ from the profitability as calculated in ICARUS, because of uncertainty and firm specific expectations about for instance energy prices. Other variables include possible biases in perceptions through a low priority for energy conservation in comparison with "core business activities". The implementation stage is analyzed in the third module. Rational behaviour theories predict that a firm will only implement technologies it considers to be profitable. However, there may be physical barriers that prevent a profitable technology from being implemented, whereas non-economic influences cause an unprofitable technology to be implemented. Possible barriers and positive influences were named above. Figure 2 shows the conceptual framework of the implementation process in firms.
5. Results of modelling energy conservation implementation The empirical modelling stage consisted of two steps. The first step empirically identified the most important determinants and barriers from a set of more that 100 possible influential factors (see Gillissen and Opschoor, 1994). Indicators of the degree of information and implementation were constructed and the influ-
1078
Conceptual model energyconservation
external:
=known V-] -unknown
Energy
teclmical
technical filter of a fmn
conservation
options based on ICTUS
l
potential ] of energy- I conservation measures of a firm
Influential factors on profitability-criteria: 1) size of project 2) marke~ition of a fLrm 3) non-core business ~ [non-profitable energy (~ profitability projects r - | criteria of ~.a firm T profitable energy projects
I i t i
information index
(;nergYprio~I/uctuati and ~. I I Qgovemmental~ policy S profitable I~~Pr~ /~. liquidity& solvaenergy investment" bility constraints options op~ons / marketexpectations and marketprospects actually implementeAprojects among which: energyprojects
noriti~j ~
' k
environ~, mental
awareness
xtemal ! pression ~
/ i'
governmental policy
I internal f a R&D c t O rfirm s:l) 2) planninghorizon 3) depreciationmethod
Figure 2: Conceptual framework of a 3-phase investment model ence of firm specific variables was assessed. The results suggest that energy in considered as one of the production factors, and that investments to reduce the use of energy i.e. by EC-technologies are made largely on a economic evaluation, taking into account the physical and financial constraints. Determinants are: firm size, return on investment, the availability of capital, the possibility of early depreciation. Barriers that prevail are: uncertainty due to fluctuations in energy prices, budgetary problems, poor financial market expectations, a lack of knowledge of EC-technologies and the complexity of those technologies. Variables that do not seem to influence the implementation decision are the "core business" argument, the size of energy bill and the presence of an energy coordinator or R&D department. Decisions on EC-investments do not basically differ from the decisions on "core business" investments (see table 1). The three phase investment model was estimated in the second stage (see Gillissen, 1994b). Preliminary results seem to confirm the results of the first step, with a few changes: the role of covenants stimulates the information and knowledge about EC-technologies. Also the expected positive role of the energy coordinator could sometimes be proven. Again, complex technologies were less known that simple measures.
1079 6. Policy simulation
The policy simulation part evaluates constructed energy policy scenarios on their contribution to energy savings. Scenarios consist a set of economic and regulatory instruments, combined with expectations regarding economic growth and energy prices. The instruments are constructed on the basis of actual and intended energy policy (VNEB, 1993); other scenarios line with "Scanning the future" and "Milieuverkenning 3". The advances in implementation, as a consequence of such a policy, will be calculated on a yearly basis up to the year 2015. Energy policies are then evaluated on their estimated contribution of additional energy savings. Instruments (control variables) that are analyzed include energy taxes, energy subsidies, the effectiveness of covenants, and information policy to reduce the information gap.
Table I: Determinants of EC-investment decision process
Important
variables
- Firm size -Information - Availability - depreciation
sources of capital moment
Less i m p o r t a n t
variables
- Size o f e n e r g y bill - Distance to core business - Low expected energy prices - competition
References
Ayres, R.U. (1994), "On economic disequilibruim and free lunch", Environmental and Resource Economics vol 4 (5) pp 435-454 Beer, J.G. de, Wees, M.T. van, Worrell, E, Blok, K, "ICARUS, the potential of energy efficiency improvement in the Netherlands up to 2000 and 2015", dept of Science Technonlogy and Society, report nr 94013, Utrecht, The Netherlands Blok, K (1991) "On the reduction of Carbon Dioxide emissions", Ph.D. thesis, University of Utrecht, Dept of Science Technology and Society, Utrecht, The Netherlands Gillissen, M. (1994a), "Energy conservation investments, a rational decision?", series research memorandum no 33, Vrije Universiteit Amsterdam, Faculty of Economics, The Netherlands Gillissen, M. (1994b), "Empirical modeling of energy conservation investment processes of firms", paper to be presented at the Summer Study of the ECEEE, june 6-10 1995, Cannes
1080 Gillissen M. and Opschoor, J.B. (1994), "Energy conservation investment behaviour, an empirical analysis of influential factors and attitudes", paper presented at the Regional Science Association, august 22-24 1994, Groningen, The Netherlands Nota Energiebesparing (1989): Energy Conservation Policy, Report of the Dutch Ministry of Economic Affairs, The Hague, The Netherlands VNEB (1993): Vervolg Nota Energiebesparing (Second Energy Conservation Policy), Report of the Ducth Ministry of Economic Affairs, The Hague, The Netherlands Werff, R.L. van der, Opschoor, J.B (1992). "De potentiele energiebesparing van Nederlandse Bedrijfstakken" in" Economische Statistische Berichten, vol 77 (3884), pp 1069-1072
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1081
Long T e r m E n e r g y Efficiency I m p r o v e m e n t Jeroen de Beer, E r n s t Worrell, Kornelis Blok
D e p a r t m e n t of Science, Technology and Society, Utrecht University, P a d u a l a a n 14, 3584 CH Utrecht, The Netherlands Phone: +31-30-537638; e-mail:
[email protected]
Abstract The opportunities for long term energy efficiency improvement in industry have been studied. Three studies are described. The first study was directed at making a preliminary survey of technologies that might reduce the end-use demand of industrial processes on the long term. The second study focused on the development of a methodology to make a more profound analysis of the long term potential. The third study describes a database for energy efficient technologies. It is concluded that, after technologies t h a t are currently technically feasible have been implemented, there still exists a considerable (new) potential for improvement.
1. INTRODUCTION A major option to reduce C O 2 emissions is efficiency improvement of energy end-use and energy conversion processes. Often it is stated that the ultimate potential of energy efficiency improvement in the western world is several ten percents. A large n u m b e r of studies showed that this potential is technically feasible within 10 years [see e.g. Lovins and Lovins, 1990; Pilavachi, 1993; ETSU, 1984; Giovannni and Pain, 1990]. However, it is also claimed t h a t in the longer r u n this potential is much larger, i.c. more t h a n 80% or even 90% [see e.g. Ayres, 1988; Jochem, 1991]. In this paper an overview is given of work performed to assess the potential of efficiency improvement on the long term. The results can be of use in R&D priority setting, developing CO 2 mitigation strategies, and long term energy infrastructure planning. Three studies addressed the opportunities of long term energy efficiency improvements. The approach and results of these studies will be described in this paper.
1082 2. S U R V E Y OF E N E R G Y E F F I C I E N T T E C H N O L O G I E S I N I N D U S T R Y
The first study is directed at making a preliminary survey of technologies t h a t might reduce the end-use demand of industrial process on the long term [Smit et al, 1994]. The technology descriptions are based on readily accessible literature, where necessary supplemented with data provided by experts on the specific sector or technology. The descriptions are divided into two sections. The first section gives a description of the reference technology and the new, energy efficient technology. Also information on state-of-the-art, ongoing R&D, and applicability of the technology is presented. The second section gives preliminary economic and energetic parameters. In table I some results of this research are presented. It must be emphasized that the results are based on a limited literature research, in some cases supplemented with consultation of experts. A more thorough analysis of the energy efficiency improvement potential is topic of the second study.
3. D E V E L O P M E N T A N D T E S T I N G OF A M E T H O D O L O G Y
The second study focuses on the development and testing of a methodology to m a k e an accurate analysis of the long term energy efficiency improvement potential. The methodology developed so far starts with the determination of the m i n i m u m energy requirement to perform a certain energy function and of the energy losses associated with performing the energy function with the current technology. The question posed is, can these losses be reduced without changing the current technology? And, if such is not the case, are technologies perceivable t h a t can reduce the energy losses? After having compiled a list of efficiency improvement technologies, an assessment of the possible technological development is made. A list of determinants of technological development is filled out on the basis of a literature review and consultation of experts. A study following this line of research has been conducted for the sector 'paper and board' industry. Furthermore, two studies are underway for the sectors 'iron and steel industry' and 'cement industry'. Here we will present some results of the studies for the paper and board industry and the iron and steel industry. Theoretically, the minimum energy demand for making paper out of wood pulp is very small (compared to a present average primary energy demand of about 10 GJ/ton paper). The operation with the largest energy losses is the steam generation (in a CHP-unit or boiler). Steam is mainly required for drying of the paper against steam heated driers. Elimination of these losses is only possible by making paper without the addition of water. However, this has a large negative effect on the product characteristics. Five technologies were selected t h a t have the opportunity to reduce the energy losses.
o m~
b~
b~
.,-i
o~ ~J
=.~ .~~
~
~-
m
L~
Z
L~
~
~
~ m
~
,
~--~
~
m
,--I
~
~
CO,
N
~ ~ ~
N
~~.~ ~ ~,~o~ ~
~I~ ,
~
~
|
o
o
~
~
~
o
r~
94 ~
~,'~1
0
9.~ ~ , ~
119
s~C
0
~ ~
,~
~1
~D ~
~.~ ~'~
II
.~
1083
..~
.1084 The third step of the methodology, assessment of the technological development, resulted in the selection of two technologies t h a t are most promising for reducing the energy consumption of a paper mill: condensing belt drying and impulse drying. R&D to these technologies is concentrated in Finland and the USA. Also in Italy, Sweden and Germany research is being conducted. These technologies have the ability to reduce the specific steam d e m a n d by 50-75% [Beer et al. 1993]. For the iron and steel industry only the first two steps and part of the third step of the methodology have been performed so far. The specific energy r e q u i r e m e n t of an efficient integrated steel plant (Hoogovens, the Netherlands) is 19.7 GJ/ton crude steel. On the short term a reduction to 16.8 GJ/ton is technically feasible. However, the thermodynamical minimum energy requirement for the reduction of iron oxide is only 6.2 GJ/ton. An exergy analysis of an integrated steel mill revealed that the room for improvement of the current process is limited. Larger improvements seem only feasible when another production route is chosen. Several technologies are outlined: an increased share of secondary steel making, advanced iron making processes (e.g. plasma processes), direct steel making (in-bath melting of iron, ore-to-powder steelmaking), near shape casting, and using hydrogen as reductant. The largest efficiency improvement is achievable with more secondary steel making. Taking into account an improved efficiency of electric arc furnaces and a higher energy demand for scrap benification a specific primary energy demand of about 7 GJ/ton steel seems possible in the long term. A combination of efficient technologies for primary steel making might reduce the specific p r i m a r y energy demand to about 12 GJ/ton steel [Beer, 1994].
4. A D A T A B A S E O N E N E R G Y E F F I C I E N T T E C H N O L O G I E S
A database (called ICARUS) on the potential and costs of energy efficiency improvement measures that can be applied in different sectors of the Dutch economy between 1990 and 2000 has been constructed [Beer et al, 1994]. D a t a were acquired using a sectoral, bottom-up approach. In figure 1 the results are summarized in a supply curve. The figure indicates a technical potential for efficiency improvement for the period 1990-2000 of 36%. If only those measures with a positive net present value are taken into consideration, the potential is 29% (we call this the economic potential). It is also possible to collect data on long term energy efficiency improvement technologies in a database like ICARUS. We have done this for technologies t h a t probably will be commercially available before the year 2015 [Beer et al, 1994]. The results are also summarized as a supply curve, see figure 1. Of course, on this longer term the uncertainty in the data is larger t h a n for the period 1990-2000.
1085
50t
2000
2015
40
30
Projected primary energy demand in the year 2000:3510 PJ in the year 2015:4915 PJ
~ 2o 100 ......................................................
,.........................................................
lO 20 3040 50 I" 0%
29% ,
10%
,
,
43% ~
20% 30% 40% Cumulative saving (%)
,
50%
60%
F i g u r e 1: Supply curves of energy efficiency improvement measures for the periods 1990-2000 and 1990-2015. On the horizontal axis the cumulative improvement potential is given as percentage of the projected energy demand without efficiency improvements. The European Renaissance scenario is used with growth figures based on physical growth. Vertically the specific energy efficiency improvement costs are depicted. Energy prices are taken from [EZ, 1994]. The low energy price scenario is used. A discount rate of 5% is used.
From figure 1 it can be seen t h a t for the period 1990-2015 the technical potential is 56%. This would m e a n a decrease in specific energy consumption in 2015 of 31% compared to 2000. The economic potential for the period 19902015 is 43%.
5. CONCLUSIONS It can be concluded t h a t the potential of energy efficiency i m p r o v e m e n t is not limited to the currently technically feasible technologies. In the steel m a k i n g process, for instance, the long t e r m potential is about three times as large as the short t e r m potential. An a s s e s s m e n t of the technological development by filling out a list of d e t e r m i n a n t s of this development, gives insight in the probability t h a t a technology will be commercialized and the time scale for this to happen. This information can be of aid in R&D-priority
1086 setting. For instance, all R&D to innovative drying techniques in paper making is currently concentrated in four or five countries, but not in the Netherlands. Therefore, national R&D-funds can better be applied for the development of other energy efficient technologies.
REFERENCES
Ayres, R.U. (1988) Energy Inefficiency in the U.S. Economy: A New Case for Conservation, Dept. of Engineering and Public Society, Carnegie Mellon University, Pittsburgh, 1988. Beer, J. de, E. Worrell and K. Blok (1993) Energy Conservation in the Paper and Board Industry on the Long Term, Department of Science, Technology and Society, Utrecht University, report 93006, Utrecht. Beer, J. de (1994) Long Term Energy Efficiency Improvements in the Iron and Steel Industry, Working paper prepared during a three months stay at IIASA, Laxenburg, Austria (draft). Beer, J.G. de,. M.T. van Wees, E. Worrell and K. Blok (1994) ICARUS-3; The Potential of Energy Efficiency Improvement in the Netherlands up to 2000 and 2015, Department of Science, Technology and Society, Utrecht University, report 94013, Utrecht. ETSU (1984) Energy Use and Energy Efficiency in UK Manufacturing Industry up to the year 2000, Energy Technology Support Unit, Energy Efficiency Office, Harwell, UK. EZ (1994) Ministry of Economic Affairs, Van wereldmarkt tot eindgebruiker, Energieprijzen voor de periode tot 2015 (From world market to end user, Energy prices for the period until 2015), beleidsstudies energie nr7. 1994. Giovanni, B. and D. Pain (1990) Scientific and Technical Arguments for the Optimal Use of Energy, Centre Universitaire d'Etude des Probl~mes de l'Energie, Universit~ de Gen~ve. Jochem, E. (1991) Long-term potentials of rational energy-use - the unknown possibilities reducing greenhouse gas emissions, Energy & Environment, 2, 1, pp. 31-44. Lovins (1990) Various report by A. Lovins and Lovins, for instance: The State of the Art: Lighting, Rocky Mountains Institute, Snowmass CO. Pilavachi, P.A. (ed.) (1993) Energy Efficiency in Process Technology, Proceedings of the International Conference, held in Athens, Greece 19-22 October 1992, Elsevier Science Publishers, London. Smit, R., J. de Beer, E. Worrell and K. Blok (1994) Long Term Industrial Energy Efficiency Improvement: Technology Descriptions, Department of Science, Technology and Society, Utrecht University, report 94076, Utrecht.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1087
Energy efficiency improvement in industrial sectors: international comparisons G.J.M. Phylipsen, E. Worrell and K. Blok
D e p a r t m e n t of Science, Technology and Society, Utrecht University, P a d u a l a a n 14, NL-3584 CH Utrecht, The Netherlands
Abstract. Eight major industrial processes areresponsible for over 50% of industrial energy consumption in most countries. The energy efficiency of these processes was determined in a number of countries, with appropriate corrections for structural differences between countries. It is shown t h a t considerable differences occur between countries, but t h a t m a n u f a c t u r i n g industry in Eastern Europe in general is less efficient t h a n in EU countries. In all cases efficiency is worse t h a n w h a t is technically and economically feasible. International comparisons provide information on energy efficiency differences, insight into technological differences between countries and into costs requirements for efficiency improvements. The comparisons can be used in international climate change negotiations and in the field of bilateral or multilateral cooperation.
1. I N T R O D U C T I O N
It is well-known that there are large differences in energy efficiency between countries. Up to now comparisons between countries are done on a national level, using aggregate measures, like the energy consumption per capita or energy consumption per unit of GDP [1]. Such comparisons can give a first impression of the differences between countries, they do not give much insight into the causes of differences or ways to reduce them. In this paper we will make a more detailed comparison between countries to show it can be used for international climate change policy making. First, we will discuss the results of a comparisons between countries for two main sectors. Subsequently we will evaluate the use of such comparisons for policy formulation.
1088 2. I N T E R N A T I O N A L E N E R G Y E F F I C I E N C Y C O M P A R I S O N S
The level of energy consumption in an economic sector is determined by three factors: the level of h u m a n activity, the mix of activities (the structure) and the energy efficiency within the sector (the energy consumption per unit of activity) [2]. All of these can be a subject of policy to reduce energy-related CO 2 emissions. Of the three, improving energy efficiency may be considered to be the most important option on the short term. For energy end-use activities two measures for (the inverse of) energy efficiency are used: 9 energy intensity: energy consumption per unit of value added; 9 specific energy consumption: energy consumption per physical unit of h u m a n activity (e.g. person-kin of transportation, tonnes of steel produced). In general the second measure gives a better insight in the technological characteristics of the use of energy. The first measure is also influenced by other factors, like feedstock and product prices. The second measure is not applicable to all sectors as not for all sectors a good physical indicator o f h u m a n activity can be defined.
2.1 Energy efficiency in heavy industry In the heavy industry the activity level can generally be measured i n tonnes of product, so energy efficiency is measured as Specific Energy Consumption (SEC). In an earlier analysis [3] we have identified t h e mix of feedstocks (e.g. primary or secondary feedstocks) and product mix as structural factorsl The potential for energy efficiency improvement is established by comparing the present SEC of a country with a 'best practice' SEC. Best practice SEC is here defined as the lowest SEC observed in a sector or plant in Europe in the reference y e a r (1988). In calculating the best practice SEC we take into account the structural effects mentioned before. It should be noted t h a t the energy efficiency improvement potential is time-dependent. On the basis of additional information [4..11] we extend the analysis to countries outside the European Union. The sectors we have studied are fossil fuel-based electricity production, refineries, iron & steel production, ammonia production, the paper & board industry, the cement industry and the chemical industry. Here we present the r e s u l t s for ammonia and steel production. 9
A m m o n i a production Ammonia can be produced by partial oxidation of oil residues and by steam reforming of natural gas. Steam reforming of natural gas is the more
1089 energy efficient process of the two. Eighty percent of the worlds ammonia production is produced by steam reforming of natural gas. The best practice S E C of 28 GJ/tonne ammonia is derived from the ICI-AMV steam reforming process (1988) [3]. Information for non-EU countries is retrieved from [4,6,8,-
11]. 9
Steel production
Steel production can be based on the Basic Oxygen Furnace (BOF) route or on the Electric Arc Furnace (EAF) route. The BOF route uses iron ore and scrap to produce primary steel, resulting in a higher quality of steel, but consuming more energy. The EAF route uses scrap only as feedstock for secondary steel production. Product type also influences the SEC. We distinguish slabs, hot rolled products and cold rolled products in the BOF route. The best practice values are based on the Hoogovens plant in the Netherlands (BOF route) and the Badische Stahlwerke plant in Germany (EAF route) [3]. Information for non-EU countries is also retrieved from [4,6,8,9,10,11].
The results are depicted in figures 1 and 2 for ammonia production and steel production respectively. From these pictures we see t h a t there are cases of developing countries and countries in transition t h a t are less efficient t h a n the European countries, but t h a t also in some cases there are no such differences.
sR1
GR
oll
CR POL
E3O E o20 (5
B
DK
F
O GR IR
I
L NL Country
P
E
UK
POL CR SR
Figure 1. Comparison of the specific energy consumption in ammonia production for various countries. The bars represent the present SEC, whereas the solid line represents the best practice (steam reforming of natural gas).
1090
35
30 -
I'
25"
Figure 2. Comparison of specific energy consumption in steel making. i represents present SEC and [::] represents best practice (with current feedstock and product mix). The solid lines indicates improvement potentials.
China
I
SR
POL CR
t5
++
-I
10
0 0%
i
i
:
i
20%
40%
60%
80%
100%
Share EAF
2.2 Discussion The methodology used heavily depends on data retrieved from international statistics. Production and energy statistics are main sources of information for studies like ours. Errors and deviations in these statistics will affect the reliability of the results. For non-OECD countries the accuracy of the statistics is in general less reliable than that of OECD countries. Improvement of available statistics, in an internationally harmonized way, is needed to improve the results of this type of analyses. There is also a strong need for the design of common methodologies to calculate energy efficiencies. The developed methodology needs to be tested for applicability in other sectors, t h a n the ones described in this paper.
4. A P P L I C A T I O N O F I N T E R N A T I O N A L E N E R G Y E F F I C I E N C Y COMPARISONS
In this section the applicability of international energy efficiency comparisons for the development of international climate policy will be discussed.
9Improvement of the knowledge on potentials and costs of energy efficiency improvement. For m a n y studies on potentials and costs of energy efficiency improvement carried out before, results are difficult to compare. International comparisons
1091 of energy efficiency, followed by a sector-by-sector comparison of costs can form the basis for a better understanding of the real differences in potentials and costs for energy efficiency improvement. 9I n t e r n a t i o n a l agreements on energy efficiency levels.
In international negotiations on CO 2 emission reduction several approaches (e.g. equal relative emission reduction) lead to objections from part of the countries involved. An alternative approach is to close agreements on energy efficiency levels by sector (taking into account structural differences) t h a t should be obtained by participating countries. International comparisons of energy efficiency are the first step of evaluating possibilities and effects of such agreements. 9I n t e r n a t i o n a l technological cooperation.
In international cooperation regarding C O 2 emission reduction (for instance in the field of 'joint implementation') international comparisons of energy efficiency can give an important contribution by steering the cooperation activities by indicating which sectors in which countries should have the highest priorities; furthermore, they can give an indication which type of transfer is most needed: investment capital, knowledge and education, licences, etc.
5. C O N C L U S I O N S
International comparisons of energy efficiency can be done for m a n y sectors, covering a considerable part of world energy demand. Development of common methodologies of measuring energy efficiency and improving the availability and quality of data is necessary. Taking into account the possible applications of international energy efficiency comparisons in international climate policy, such a task seems worthwhile undertaking.
6. R E F E R E N C E S
1. 2. 3.
See for example: The State of the Environment, OECD, Paris, 1991. L.J. Nilsson: "Energy Intensity Trends in 31 Industrial and Developing Countries 19501988" Energy, 18 (1993) pp. 309-322. L. Schipper, R.B. Meyers, R. Howarth, R. Steiner: Ener~zv Efficiency and Human Activity: past Trends, Future Prospects, Cambridge University Press: Cambridge USA; 1992. E. Worrell, R.F.A. Cuelenaere, K. Blok, W.C. Turkenburg: "Energy Consumption by
1092
4. 5. 6.
7. 8. 9. 10. 11.
Industrial Processes in the European Union", Energy (forthcoming). E. Bossenbroek: Energiegebruiken in Polen; trends en vergelijkingen met de EG/OECD, Dept. of Science, Technology & Society, Utrecht University, November 1993 J. Garcia del Valle, A. Torres: Outlook of Latin American Cement Industry, in: J. Sichis: Energy Efficiency in the Cement Industry, Elsevier Applied Science, London, 1990 S. Meyers, L. Schipper, J. salay, A. Gromadzinski, E. Hillie, P. Kaleta, M. Kumanwoski, J. Maron, J. Norwisz and S. Pasierb, Energy Use in Poland, 1970-1991: Sectoral Analysis and International Comparison, LBL, Berkeley, July 1993 M. Ross, L. Feng, "The Energy Efficiency of the Steel Industry in China", Energy 5 16, (1991), pp.833-848. Statistical information on the Czech Republic, Slovak Republic and Poland, 1990. Statistics on Energy in the Steel Industry (1990 Update), International Iron and Steel Institute, Brussels, 1990 Steel Statistical Yearbook 1992, International Iron and Steel Institute, Brussels, 1992 B. Vallance, East-West Comparisons of Energy Efficiency in Energy Intensive Industries, Symposium on Energy Efficiency and Economic Transition in Central and Eastern Europe, Paris, 25-28 May 1993
Abbreviations
B Br EU L CR D DK F E GR I IR NL POL SR UK
Belgium Brazil European Union Luxembourg Czech Republic Germany Denmark France Spain Greece Italy Ireland Netherlands Poland Slovak Republic United Kingdom
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1093
Carbon dioxide removal studies in the Netherlands K. Blok D e p a r t m e n t of Science, Technology and Society, Utrecht University, P a d u a l a a n 14, 3584 CH Utrecht, The Netherlands
Abstract
An explorative research programme o n C O 2 removal has been carried out in the Netherlands. The goal of this programme was to obtain a better understanding of the technical and economic feasibility of the recovery of CO 2 from flue gases and synthesis gases and the sustainable storage of CO 2 outside the atmosphere. As far as CO 2 recovery from power plants is concerned, options based on coal gasification with CO 2 recovery turn out to be most energyefficient. Of the remaining recovery options chemical absorption from flue gases, using amines, seems most promising. A number of recovery options based on membrane technologies have been identified, but most of them still require considerable development. In manufacturing industry, attractive options for CO 2 recovery are available in refineries equipped with residue gasification, and in the ammonia fertilizer industry. More costly options were identified in the iron and steel industry and in the petrochemical industry. CO 2 storage in aquifers is technically feasible. When injecting CO 2 in aquifers part of the water already present will be displaced. The main mechanisms for this displacement will be gravity segregation and viscous fingering, as was shown by simulation calculations. The Dutch subsurface contains a large number of aquifers that are potentially suitable for CO 2 storage.
1. I N T R O D U C T I O N
In the period of 1991 to 1993 an explorative research programme on CO 2 removal has been carried out by a number of companies and research institutes in the Netherlands. The goal of this programme was to obtain a better understanding of the technical and economic feasibility of the recovery of CO 2 from flue gases and synthesis gases and the sustainable storage of CO 2 outside the atmosphere. Before this research programme started a number of publications on
1094 carbon dioxide recovery and storage had already been issued in the Netherlands with special emphasis on carbon dioxide removal from IGCC power plants and on storage of carbon dioxide in depleted natural gas fields. In the new research programme the main emphasis was on studying a range of techniques to recover carbon dioxide from gas streams in detail. Some studies were devoted to the recovery of carbon dioxide from industrial processes. One study explored the possibilities of CO 2 storage in aquifers. A separate study was directed at the calculation of the efficiencies and costs of complete power plants in a comprehensive way. A more extended overview of the results is given in
[1]. The research programme was sponsored from various sources, the main ones being the Ministry of Housing, Physical Planning and Environment and the National Research Programme on Global Air Pollution and Climate Change. The total budget amounted to about 1.5 million Dutch guilders (1 Dfl = $0.6 - ECU 0.45).
2. C O 2 R E C O V E R Y B A S E D ON COAL G A S I F I C A T I O N
It was confirmed that carbon dioxide recovery based on coal gasification shows the smallest decrease in efficiency of electricity production. The detailed analysis carried out by the research institute of the electric utilities, KEMA, showed that efficiencies of over 36% can indeed be attained in combination with carbon dioxide recovery (table 1). Two widely differing CO 2 recovery technologies were studied. In the first case a shift reaction is applied after gasification resulting in a fuel gas mainly consisting of hydrogen and carbon dioxide. For separation of hydrogen and carbon dioxide a number of options were studied: freezing out the CO2, membrane separation, hydrogen recovery, physical absorption and chemical absorption. The best option is physical absorption (using selexol). By freezing out the CO 2 the required high degree of CO 2 recovery (to less than about 120 g/kWh) can not be attained. When this limitation would not be set, freezing out the carbon dioxide should certainly be taken into consideration. Using membranes at low temperatures is not attractive, mainly due to the high hydrogen loss. Combined with high-temperature gas clean up membrane applications may become of interest, although the membranes with the required high H2/CO 2 selectivities are still under development. Chemical absorption systems have a too high energy demand. Hydrogen recovery techniques showed considerable hydrogen losses. The components for the favoured shift/selexol concept are commercially available but were never applied in this combination. It is concluded t h a t the technology is ready for demonstration. The second IGCC approach makes use of a gas turbine in which the fuel is combusted in a mixture of oxygen and recycled CO 2. The
1095 Table 1. Some key results for complete power pla nt concepts as calculated by KEMA. P r e l i m i n a r y cost es tima te s ra nge from 60 - 80 Dfl per tonne of CO 2 avoided, being the lowest for the IGCC options. Electricity production costs are e s t i m a t e d to be 0.14 - 0.15 Dfl/kWh for the coal cases and 0.095 Dfl/kWh for the n a t u r a l gas case. Type of p l an t m e t h o d of recovery
Net conversion efficiency (%)
Specific CO 2 emission (g/kWh)
IGCC - CO2/O2-combustion (Texaco)
34.8
5
IGCC - CO2/O2-combustion (Shell)
36.0
30
IGCC - shift & physical absorption (Texaco)
36.4
139
Pulverized coal - chemical absorption (retrofit)
29.7
105
N a t u r a l gas fired combined cycle - chemical absorption
44.9
86
42 - 43
800 - 820
52
390
Reference coal fired power plant Reference n a t u r a l gas fired power p l a n t
combustion products (mainly CO 2 and water) are expanded t h r o u g h the t u r b i n e section. After cooling in a h e a t recovery s t e a m g e n e r a t o r and removal of the water, the CO 2 is recycled to the gas t u r b i n e compressor. P a r t of the compressed CO 2 is used in the gas turbine combustion chamber, the r e m a i n d e r is exported. As CO 2 is the m a i n working fluid in the gas turbine, the properties differ strongly from a conventional gas turbine. The results of i n t e g r a t i n g such a gas turbine in an IGCC p l a n t are given in table 1. The m a i n bottleneck for the application of this scheme is the fact t h a t such a C02-gas-turbine is not available at present. S t a r t i n g such a costly development process is only justified if it gives clear (cost or efficiency) a d v a n t a g e s above the IGCC/shift/selexolprocess m e n t i o n e d before.
3. CHEMICAL A B S O R P T I O N / O T H E R R E C O V E R Y T E C H N I Q U E S As the f u t u r e of coal gasification is still u n c e r t a i n it is advisable to develop other CO 2 recovery techniques as well. A n u m b e r of other options for CO 2 recovery was evaluated. In most cases chemical absorption, using amines, is the most attractive alternative.
1096 For the recovery of C O 2 from flue gas of conventional coal-fired power plants the use of gas separation membranes is more expensive t h a n chemical absorption. This is mainly due to the high power requirements for the compression of the flue gases. When chemical absorption is applied, the use of gas absorption membranes, is of interest. Gas absorption membranes are used in conjunction with chemical absorption liquids where the conventional absorption column is replaced by a m e m b r a n e contactor. This modification could increase the conversion efficiency of the power plant by approx. 0.5%. This improvement is mainly due to a reduction of the pressure drop over the absorber. Gas absorption membrane systems, however, are still under development. For natural-gas-fired combined cycle power plants the most costeffective option also is chemical absorption, with an overall conversion efficiency of approx. 45%. An alternative is a power plant based on a gas turbine using combustion in a CO2/O2-mixture. Also a system based on m e t h a n e reforming of natural gas (to a large extent similar to an IGCC plant) was investigated. However, it showed a low efficiency: about 37%.
4. C O 2 R E C O V E R Y IN M A N U F A C T U R I N G I N D U S T R Y
Twenty plants in manufacturing industry with the largest C O 2 emissions in the Netherlands together are responsible for about 20% of the total Dutch CO 2 emissions. Main sectors are refineries, the iron and steel, petrochemical and fertilizer industries. Carbon dioxide recovery can be accomplished in refineries equipped with a residue gasification unit. Residue gasification is expected to be a good solution in the development towards low sulfur oil products and deeper conversion. The gasification product is fed to a shift reactor in order to produce hydrogen for other refinery processes. The carbon dioxide that is co-produced can be recovered easily. In this way about one quarter of the CO 2 emissions in future refineries can be avoided. Another attractive option is available in the fertilizer industry. In producing ammonia, which is one of the main feedstocks for fertilizer production approx. 50% of the CO 2 output of the fertilizer industry is already recovered. At present part of this amount is utilized, the remainder is vented to the atmosphere. CO 2 recovery can be applied on this stream by just compressing it to transportation pressures. Both for the refineries and the fertilizer industry estimated mitigation costs are in the order of 20 Dfl per tonne of CO 2 avoided. More costly options were identified in the iron and steel industry: recovery of CO 2 from blast furnace gas; and in the petrochemical industry: the use of low-temperature waste heat (100 - 150 ~ for supplying the reboiler duty of a chemical absorption process.
1097 5. S T O R A G E O F C A R B O N D I O X I D E
According to one of the studies, CO 2 storage in aquifers is technically feasible. When injecting CO 2 in aquifers part of the water already present will be displaced. The main mechanisms for this displacement will be gravity segregation and viscous fingering. Extended simulations of the behaviour of CO 2 have been carried out for sample reservoirs; in one of these aquifers 15,000 tonnes of carbon dioxide per day can be injected during 8 years. After this period CO 2 breakthrough is observed at the spillpoint. The Dutch subsurface contains a large n u m b e r of aquifers t h a t are potentially suitable. Taking a number of constraints into account the total aquifer storage capacity for CO 2 a prudent estimate of the storage capacity of 1.2 Gtonne CO 2 is made. The main chemical effect of carbon dioxide in aquifers is its effect on carbonate chemistry. The decrease of the pH due to the dissolution of carbon dioxide will cause solution of carbonates. This effect is so small t h a t weakening of the porous structure is not expected. However significant changes in permeability may occur. If the seal of the structural trap is a clay layer, drying of this layer could reduce its tightness. The costs of injection (departing from a delivery pressure of 110 bar) are estimated to be 0.7 and 1.2 Dfl per tonne of CO 2 injected for aquifers above and below 1000 m depth respectively.
6. C O N C L U S I O N S As far a s CO 2 recovery from power plants is concerned, options based on coal gasification with CO 2 recovery turn out to be most energyefficient. Of the remaining recovery options chemical absorption from flue gases, using amines seem most promising. A number of recovery options based on membrane technologies have been identified, but most of them still require considerable development. More t h a n at present, attention should be paid to CO 2 recovery options outside the electricity production sectors, e.g. in manufacturing industry. Storage of CO 2 in aquifers seems to be technically feasible, but the total storage capacity, taking into account strict conditions, is limited. It is felt that with respect to underground storage of carbon dioxide, especially in aquifers, the largest uncertainties exist. F u r t h e r investigation of this option by reservoir simulations, laboratory experiments and field tests should have a high priority in further R&D planning.
1098 7. R E F E R E N C E
1 K. Blok: Final report of the Integrated Research Programme on Carbon Dioxide Removal and Storage", Department of Science, Technology and Society, Utrecht University, The Netherlands.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1099
STORAGE OF CARBON DIOXIDE IN AQUIFERS IN THE NETHERLANDS L.G.H. van der Meer, R. van der Straaten and J. Griffioen TNO Institute of Applied Geoscience EO. Box 6012, 2600 JA Delft, The Netherlands
Abstract This paper presents the results of a study about the technical feasibility of the underground storage of carbon dioxide (CO2) in aquifers. Special attention was paid to physical processes, limiting geological conditions and geochemical and environmental aspects. The CO 2 storage capacity of aquifers below the Dutch onshore is estimated based on these results. In addition, the long-term CO 2 storage potential of a hypothetical CO 2 storage reservoir is estimated. 1. INTRODUCTION The investigations were commissioned by the Dutch Ministry of Housing, Physical Planning and Environment and the Dutch National Research Programme on global air pollution and climatic change. This study of the technical feasibility, limiting geological conditions and consequences of carbon dioxide storage in aquifers was carried out as part of a programme entitled: A preliminary research programme for CO 2 removal and storage.
2. DISPLACEMENT BEHAVIOUR In order to elucidate the dispersive character and the fluid flow mechanism of C O 2 in an aquifer system we have unravelled the individual mechanisms affecting the displacement process. In general, the dispersion or spreading out of CO 2 in an aquifer can be described at three different scales: pore-scale, stratum-scale, and reservoir scale. Each scale is characterized by a particular process. Although smaller scale processes are active in the dispersion process at reservoir-scale, they will play only a minor role. All the processes and/or effects are understood and well described in the literature. For further information, the reader is referred to publication of van der Meer 1. If CO 2 is injected into an aquifer, it will be able to displace the pore water in the aquifer to a large extent. The displacement process is determined by many individual mechanisms related to fluid properties and the specific conditions of the rock matrix. One of the most important parameters in this displacement process is the relative ability of the two fluids to flow in the porous medium. This property is referred to as the relative mobility of the fluid. When one fluid displaces another, the mobility ratio (M) of the displacement is defined as the mobility of the displacing fluid divided by the mobility of the displaced fluid. For average reservoir parameters, we calculated M=40 for an aquifer at 800 metres depth and M=13.2 for an aquifer at 1800 m depth. This means that CO 2 is 13 to 40 times as mobile as the formation water and because the CO 2 is pushing the water, it tends to by-pass the water.
1100
The effect of one fluid being displaced by another can be considered as a complex process. The large differences in the physical properties of the three main items (two fluids and the reservoir rock) for the adopted range of depths make it difficult to predict the results of their interaction in a displacement process. A CO2/water displacement process will be dominated by a gravity segregation effect. A layered permeability distribution i.e. a large kv/kh ratio will have a negative influence on the upwards migration of CO 2. The calculated mobility ratios for a process in which CO 2 displaces water enable us to predict substantial viscous fingering effects. The resulting areal sweep efficiency will be in the order of 25 to 60 %, whereas the vertical sweep efficiency will be very small (in the order of 2-25 %), due to the combined effects of gravity segregation and viscous fingering. With the exception of the permeability distribution, all other small and medium scale effects will have an insignificant influence on the displacement process.
a.
.038PVI
b.
.076PVI
c.
.114PVI
d.
.150PVI
e.
.190PVI
f.
.228PVI
Fig. 1. Results of numerical displacement simulations. Concentration distribution maps for increasing time slices. (PVI = Pore Volume Injected)
3. G E O C H E M I C A L A S P E C T S Two types of geochemical processes are associated with the injection of CO 2 in deep-seated aquifers. The first is enhanced dissolution of carbonate minerals due to an increase in the dissolved CO 2 in formation water. The amount of dissolution is almost independent of depth (and temperature) for depth below 750 m. The total groundwater composition is not greatly affected by this process. Effects on aquifer properties (permeability and porosity) are also small. The second process relates to the characteristics of electric double layers of clay minerals. The double layer thickness of (swelling) clay minerals depends on the di-electric constant of the fluid present. The change from water to CO 2 as pore fluid may lead to a decrease in double layer thickness for swelling clay minerals such as smectite. This may effect the aggregate structure of clay minerals. Unfortunately, no applicable information was available on this topic. Clay minerals with a swelling interlayer may shrink. The associated consequences for the permeability of the aquifer and the sealing characteristics of cap rock need to be investigated.
1101
4. L I M I T I N G A S P E C T S OF CO 2 I N J E C T I O N IN AN A Q U I F E R Much information about aspects limiting fluid injection in the subsurface was obtained from the practice of flooding with water when extracting oil. Flooding with water is the main fluid injection method. This information yielded two possible limiting aspects in respect to CO 2 storage in aquifers: well/formation damage and injection pressure. Laboratory and field studies indicate that almost every operation that has to do with drilling, completion, workover, production, particle induction and stimulation are a potential sources of damage to well injectivity. After evaluating all possible causes of well damage, we have concluded that well damage can have no direct limiting effect on CO 2 injection. All problems associated with well clogging or well damage are understood and technically solvable. The injection of fluids into an aquifer will result in an increase of the fluid pressure of the aquifer: this causes the grain pressure to decline. This shift in pressure regime can cause fracturing of the rock matrix, opening up existing faults and/or induction of microseismicity. These effects depend largely on the mechanical properties of the reservoir rock. If the average aquifer pressure exceeds the overburden pressure, there is a risk of absidence. 5. E N V I R O N M E N T A L A S P E C T S The major risks of the underground storage of C O 2 are suffocation, groundwater acidification and pollution, and damage by CO 2 blow outs or absidence of the earth's surface. If large amounts of CO 2 leak to the surface they will create blanket-like cloud of CO 2 that fills topographic depressions. Since this CO 2 will drive away all oxygen, any people or animals that enter theses areas may suffocate. Malfunctioning of the CO 2 injection system can be reduced by the use of appropriate materials and by intensive maintenance. A simple additional device, integrated in the pressure monitoring system, could shut off the failing subsystem from the rest of the system and limit the emission of CO 2 to minimal quantities. If large amounts of CO 2 escape the reservoir rock and invade the subsurface, the groundwater may be affected. Groundwater naturally contains CO 2. Escaped CO 2 could increase the natural CO 2 concentration of the groundwater. A tenfold increase of CO 2 concentration in the groundwater will decrease the pH number by 1. The risk of CO 2 escaping from a storage location can be reduced by introducing peripheral observation wells. As a result of manmade pressure changes in the subsurface the earth's surface may gradually sink or rise. A symptom of these changes is the occurrence of microseismicity. Several cases of sinking or subsidence are well known and have been extensively documented. The data on the occurrence of absidence is limited but it is understood that the same theories as for subsidence can be applied. Regular monitoring of the possible rise of the earth surface is recommended. 6. S U B S U R F A C E A S P E C T S The similarities between natural gas storage in aquifers and C O 2 storage in aquifers are obvious. The technical reservoir engineering knowledge gained in underground gas storage can be directly applied. In the following sections the subsurface aspects of CO 2 storage are discussed, using a hypothetical aquifer. We deal with subsurface aspects from the surface downwards. From the results of calculations it can be concluded that all the pipeline diameters we investigated (4.0-7.0 inch) are capable of delivering the CO 2 at the aquifer injection location. A smaller pipeline diameter or an increased injection flow rate will reduce the CO 2 delivery pressure at this location.
1103
lopment during these two time periods. A simulation model was constructed, representing a 30x30 km part of the subsurface. An injection period of 50 years followed by a shut-in period of 100 years was simulated. Figure 3. shows the results of this simulation run. The delta CO 2 distribution map shows only the upper part of the subsurface model. The observed CO 2 bubble diameter at the top of the storage location can be estimated as 16 km at the end of injection period and grows to 18 km during the shut-in period. CO2 movements are only active if there are large differences in pressure between the injected CO 2 bubble and the constant pressure boundary of the model. From the simulation results it can be concluded that CO 2 storage in a quasi-infinite aquifer is possible. It is however impossible to define a storage efficiency factor due the infinite nature of the storage location. From all simulation work performed it can be concluded that the suitability of aquifers depends entirely on their size, within the boundary conditions stipulated. Displacement process will be dominated by channelling, viscous fingering and gravity segregation.
Iil~~~f~ifBi[
lill!i[!l l i l | | | | B ! i | | | R l | | | | l l i | l l U [ i'1==Ill =lII [[ t',litIll I Ill I!*,IH~ I l l l / l l l l l l l l l l i l ~ l i l l l t M M l I R l l J i ~ l l l
llillWlililUlllli
g / ~ D i
II
i~!i~ii!iii!i
I'II
|gltlll: "....... I I B D I I ,
Ii!ilI|~
iii!iiiii!
'
i~!ili::ii!: i
'ii'i [ i;i~:~; .....I ,,,WMItI~,,,~ ,..
~_-.L~.::'..':'~-'--
.......
I
Ini~ Inlnaq ~
IqnE]nt~
~
[ i q l ~ I~lIIOl I l l . i l l [ l ~ I ~ l ~ I I
illIliI! [
-.02
-.06
-.10
~
............. :::::";'--
I31qnE3Iqai~qnImad ~lllll'.lPIll',l q iI:llmJnli~ IIqnE ~
-.~14
~ I l l l I h q '.iII I I: Ih q [i IIi Ie IlII [ l i I l l I i ~ I i q [
PlamlqnElninEIIqnE2qn~
[iI~Iiq [
qnE]qnIDmqnDqaiTb ~
i~l~llII ( l l l ~ l I~l~lll I l i q g I lilIleq [
qdcsqm~
qln~qmIE]qn~lqnl]nb ~
qkn n
~
~
]
[
[
~
~
-."18
Fig. 3. Map of the difference in CO 2 concentration between
150-year map and the 50-year map
7. CO 2 STORAGE CAPACITY IN THE NETHERLANDS The underground CO 2 storage capacity of the Permian to Quaternary aquifers of the Dutch onshore has been estimated from published data about the subsurface. First, an inventory was made of potentially suitable aquifers for CO 2 disposal (permeability > 50 mD, a depth below 800 m and covered by cap rock) and information was gathered on net reservoir thicknesses
1102
We investigated the sensitivity of aquifer parameters and the scale of the CO 2 injectivity in an aquifer. A computer program was written to compute the pressure at increasing drainage radius as function of the permeability and the skin factor. Analysis of the results clearly shows that the aquifer permeability and the well skin factor are the controlling parameters of a CO 2 aquifer storage operation. It was observed that in nearly all cases when the permeability is 0.025 bll-n2 there are large pressure gradients near the well bore. Clearly, the overall aquifer permeability will play a decisive role when selecting potential aquifers for CO 2 storage. An aquifer in the Netherlands was selected to investigate and estimate the technical reservoir aspects of CO 2 storage in aquifers. (Aquifer data: porosity Brussel sand 30 - 36 %, permeability .05 - .6 gm 2, thickness 50 m). From the outset it was assumed that 6 wells would inject 15 000 ton a day of CO 2. This, in combination with the domed shape of the aquifer under study, makes it possible to reduce the simulation model to one-sixth of its original aquifer size. A pie-slice segment, with an angle of 60 degrees, was selected. The results of the CO 2 storage simulation runs reveal that CO 2 will breakthrough at the spillpoint after a cumulative CO 2 injection of 5.921 x 109 Nm 3. The results clearly indicate that the CO 2 distribution is dominated by gravity segregation. If we compare the results of the theoretical storage volume calculation with the results of simulation, than only 4.3 % of the volume is used. Figure 2 is a graphical representation of the model selection procedure and shows in cross section the CO 2 distribution at breakthrough. A further parameter sensitivity study 2 has shown that the CO 2 storage efficiency of a predefined part of an aquifer is small. For practical purposes a CO 2 storage efficiency of 1 to 6 % can be used, depending on the vertical transmissibility of the potential reservoir. I..
4500 m Selected reservoir A
A'
750 mSS 800 ~ss--~.
o~ o
,,
[
.:. :
". "-..
A'
-. 800 mSS 775 mS.S \ 750mSS \'-.\
ii::2a ij
S . . . . . . .
.... :.: "".. !: :.. ' " : . ~77T: ........ ~,...... 9............ ~,.,..:.:.: .................................................. :"~......:,:::,:ii
E
o
0.84 0.0 COa concentration
Fig. 2. Selection of simulation model, and the simulation results.
All the above work and reported efficiency factors refer to predefined storage locations with a known maximum storage volume, i.e. a storage location within a geological trap and an outer storage boundary. However, large aquifers without a geological trap structure are known to exist. If we relax the trap constraint it will be essential to uphold the constraint that the aquifer will need a impermeable top layer to prevent any CO 2 from leaking out through the top of the aquifer. The omission of a trap and the presence of a top seal will require the size of the CO 2 bubble to be controlled during the active injection period as well as in the subsequent period of storage. We performed a limited simulation study to investigate the CO 2 bubble size deve-
1104
and porosities. Next, the percentage of the volume confined by traps was assessed by determining the area occupied by closed structures on available depth maps and extrapolating these data to the entire Dutch onshore. Finally, the storage capacity was calculated from the trapped pore volume, assuming a CO 2 occupation of 2% and a CO 2 reservoir density of 700 kg/m 3. The uncertainty introduced by extrapolation may be considerable. The Triassic structures in the study area, for example, are all related to salt tectonics. Similar structures do not occur elswhere in the Netherlands. In addition, we were not able to define stratigraphic traps (created by facies changes) or very large structures extending beyond the mapped area. This also forms a major uncertainty. The Permian aquifers, for example, are thought to be confined by large fault blocks below a thick package of Zechstein salt. These blocks are expected to form huge traps, but are not included in the storage estimate because they could not be defined. We indentified more than 100 traps in those parts of the Netherlands where suitable depth maps were available. Of these only 50 traps are potentially suited for CO 2 disposal. The remaining structures are either too shallow or do not contain appropriate aquifers. The pore volume in these 50 traps is about 15.7 km 3, of which 2.1 km 3 contains oil or gas. Extrapolation of these results to the entire Dutch onshore leads to a total trapped pore volume of about 35.7 km 3. This corresponds to a CO 2 storage capacity of approximately 0.50 Gt. Previous storage estimates were considerably more optimistic. Van Engelenburg & Blok 3 proposed a capacity of 40 to 82 Gt CO 2. Huurdeman 4 made an estimate of 2.5 to 10 Gt CO2. The discrepancy between these figures and ours can be readily explained by the use of different information and constraints. Van Engelenburg & B lok did not take into account the presence of trapping structures whereas Huurdeman assumed that the entire pore volume in a trap can be saturated with CO 2, an assumption that has to be revised in the light of the results of our simulation experiments.
8. CONCLUSIONS 1) C O 2 storage in aquifers is technically possible. The knowledge about the technology of CO 2 injection in aquifers is adequate, but there is a lack of reliable subsurface data. 2) The C O 2 w a t e r displacement will be dominated by gravity segregation, by channelling, and viscous fingering over the whole subsurface depth range investigated. 3) The C O 2 storage efficiency of a predefined part of an aquifer is small. For practical purposes a CO 2 storage efficiency of 1 to 6 % can be used, depending on the vertical transmissibility of the potential reservoir. 4) The storage capacity of traps on onshore aquifers in The Netherlands is estimated at 0.5 Gt CO 2. 5) The estimated CO 2 storage capacity of quasi-infinite aquifers in general is problematic. It can, however, be stated that they have a large potential. References
1. 2. 3. 4.
van der Meer, L.G.H., Griffioen, J., Geel, IGG-TNO-report OS 92-24-A, February 1992 van der Meer, L.G.H., Paper presented at the ICCDR-2, Kyoto Japan, Oct. 1994. Van Engelenburg, B., & Blok, K. (1991), Report nr. G-91006. Huurdeman, A.J.M. (1992) TNO-report 91-250.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1105
WASTE REDUCTION AND THE STRUCTURE OF THE DUTCH WASTE SECTOR. Paulien de Jong and Maarten Wolsink Department of Environmental Science University of Amsterdam Nieuwe Prinsengracht 130 1018 VZ Amsterdam The Netherlands Abstract Waste is a source of greenhouse gases. In general, waste producers are not encouraged sufficiently to limit waste production. Reduction of waste can not be achieved by just formulating waste policy. The organizational conditions under which removal and processing take place frustrate the achievement of waste reduction. The final objective of this research project is to design a structure for the waste sector which will contribute to the reduction of waste streams for incineration and landfilling. In this paper the results of the first two phases are reported: an analysis of the Dutch waste sector and the identification of key variables in the organizational structure of the waste sector. 1. I N T R O D U C T I O N In many ways waste is a source of greenhouse gases. Waste incineration directly leads to the production of carbon dioxide. Waste on landfill sites also produces CO2, but it is assumed that carbon containing waste decomposes while emitting methane. Since the global warming potential of methane is about 30 times that of carbon dioxide, carbon containing waste disposal in landfills might even be a greater problem than incineration. Furthermore, there is a loss of energy through the energy content of products and materials. Decomposition means the loss of energy previously used for refinement of raw materials and for processing these materials, as well as the energy used in the process of production of goods. In addition, a variety of non-carbonic waste is produced during production processes with sometimes high CO2 emission per kilogram. A reduction of waste which will either be dumped at landfills or incinerated, will be a contribution in controlling the greenhouse problem. 2. WASTE P O L I C Y AND T H E WASTE S E C T O R In Dutch waste policy and in research a great deal of attention has been paid to the formulation of targets and regulations and to the selection of policy-instruments. Nevertheless, the tendency of growing waste streams has not yet been stopped in the Netherlands. The expected waste streams for incineration are growing and a large amount of new incineration capacity is planned. The size and composition of waste streams is determined by several factors like demographic developments (population growth, composition of households), economic developments (prosperity) and technological trends (mixing of materials). The fact that the size and composition is also determined by the conditions under which the removal and processing of waste takes place is too often neglected. In many ways organizations (companies, public services, authorities) have conflicting interests. Their objectives often conflict with the policy goal of waste prevention. Neglecting the importance of organizational conditions is not an uncommon phenomenon. It is not restricted to the waste sector. A conflict of objectives also exists in the energy sector. Saving
1106 energy by means of a reduction of energy consumption is a policy goal, but generally it is not an interest of utilities that gain economic benefits from the supply of more energy. Due to the problem of conflicting objectives is the fact that policy ultimately has to be implemented within society by other actors than the policy formulating bodies. These actors have their own objectives and interests, which may differ from those of the policy makers. These actors try to find ways to seek the fulfilment of their own objectives and they have the capability to frustrate the seeking of fulfilment of others. Most of the actors have no interest in waste reduction or in altering the kind of waste that has to be processed. 3. A C T O R S IN THE W A S T E S E C T O R The best way to get insight in the structure of the waste sector and to understand the dynamics of it, is to adopt the idea that the organizations are linked together in an interorganizational network. Characteristic for such a network is the fact that the individual objectives of the organizations participating in the network are more important than the objective of the network itself (1). It seems as if they cooperate because they share interests and objectives, but more often the real reason for cooperation is mutual dependency (2). Among the public and private organizations which try to influence the circumstances under which the removal and processing of solid waste take place, seven groups can be distinguished. The categorisation is primarily based on the interests and activities that organizations have in common and secondarily on which phase in the material life cycle their activities have impact (3). 1. Waste generators 2. Waste collectors 3. Waste processors 4. Waste disposers 5. Policy makers 6. Research groups and consultancies 7. Interest groups and umbrella-organizations The first four categories of actors are those who are participating on the waste market. The waste sector is much broader: apart from the participants on the waste market other groups of actors are part of the waste sector. These are primarily governmental organizations that have to formulate waste policy. Secondly there are organizations that provide data and ideas to support the policy making process, like research groups and consultancies. Thirdly we distinguish organizations that either defend interests of groups of other actors, or try to influence policy (lobbying). After analysis of the relations between the actors of the seven categories we established several structural impediments for the reduction of waste. 4. I M P E D I M E N T S FOR PREVENTION IN THE WASTE SECTOR The governmental institutions in the Netherlands find themselves in a paradoxical situation. Municipalities invest in disposal plants from the perspective of environmental hygiene. They invest in incinerators which have to be built in accordance to strict environmental standards. The investments are large.and take a long period of time to write off. Therefore there is no interest in source-reduction which can be implemen-ted in the short term. To the contrary, it creates an interest in assurance o f long term waste supply.
1107 The long term of writing off causes short term risks when installed capacity is not fully utilized. Waste processors attempt to guarantee a sufficient flow by tying waste suppliers to long term contracts. As a rule, governmental institutions become tied, while private enterprises remain free to change their supply from one processor to another. Private organizations which are not contractually bound can offer "extra" waste, but then they can negotiate about the rates of the incineration. It is in the interest of the waste generating industrials or smaller enterprises that the above described situation of waste handling and management remains the same. Overcapacity leads to lower incineration rates. In general, it is in the interest of private organizations to keep authorities and policy makers in a situation of uncertainty about the amount, sort and composition of waste that will be released. More information and better planning of facilities will not only limit bargaining opportunities of enterprises and lead to higher processing tariffs. On the other hand exchange of information will give policymakers tools to formulate strict prevention goals in quantitative and qualitative sense. Policy and implementation is connected to a certain level of administration. In the meanwhile other, private organizations operate on a higher level. Those private organizations can not be forced to implement the policy of public authorities. The legal jurisdiction offers authorities a basis for power but it also restricts them. Public bodies are participating in the waste sector with different roles at the same time, which may sometimes cause entanglement: * A role as representative of the law (functions of control and issuing permits) * A role as a participant on waste market. * A role as policy makers, in which general public interests must be served. 5. S T R U C T U R A L E L E M E N T S IN T H E WASTE S E C T O R Based on the analysis of the Dutch waste sector and a rough inventory of twelve different countries, five structural elements determining barriers for waste prevention in the waste sector were indicated. These five variables have been used as criteria for the selection of three cases for a multiple case study (4). This is the next phase of the research, directed at the identification of elements in the organizational structure of waste sectors, which might be implemented in the Dutch waste sector. The five structural elements are: (5) 1. Scale of organizations which are handling waste. 2. Functional separation of tasks and responsibilities between actors 3. Activities directed at input and output attributed to different organizations. 4. The role of the authorities: utility-function vs. market participation. 5. Attribution of responsibilities for waste prevention to existing (or merged, or separated) organizations, or new organizations. 6. FURTHER RESEARCH PLANS The starting point of this research project is that size and composition of waste streams are partially determined by the organizational conditions under which the removal and processing take place. After a first examination of the structure of the waste sector in the Netherlands it became clear that there exist several aspects in the organizational structure of the waste sector that impedes the stimulation of waste minimalization incentives.
1108 To obtain ideas about improvements for (parts of) the structure of the waste sector that may lead to stimulation of waste reduction, a multiple case study has to be carried out. Therefore a broad inventory on significant structural conditions in twelve industrial societies was done. Out of this list of twelve, three cases were selected. Significant characteristics of the waste sectors of New Jersey (US), Nord-Rhein Westfalen (Germany) and Denmark will be investigated. The aim of these studies is to propose improvements for the structure of the Dutch waste sector. Another way to search for ideas for improvement can be found in comparison of the waste sector with other sectors in society. A literature study of this subject will be made. The final objective of this project is to design a structure for the waste sector which will contribute to the reduction of waste streams for incineration and landfilling. The last step in the research project will bring all results about possible improvements together and will have to result in a new concept for the structure of the waste sector. Ex ante evaluation has to be carried out to prospect the possible problems in the implementation of such a new designed structure. Also ex ante evaluation has to be done on the effects of altering (parts) of the existing structure in size and composition of waste streams.
7. REFERENCES: 1. Godfroy, Perspectieven op organisaties: organisatiepsychologie en -sociologie 2, Open Universiteit, Heerlen, 1990. 2. D. Marsh and R.A.W. Rhodes, Policy networks in British Government, Clarendon Press, Oxford, 1992. 3. P. de Jong and M. Wolsink, De structuur van de Nederlandse afvalsector, IVAM/UvA, Amsterdam, 1993. 4. R.K. Yin, Case study research; Design and methods, Sage, Beverly Hills, 1984. 5. P. de Jong, Verkenning afvalsituatie in: Belgie, Denemarken ..... Zwitserland; Internal report, IVAM/UvA, Amsterdam, 1994.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1109
Institutional barriers to waste reduction in Finland J. Hukkinen Maastricht School of Management, P.O. Box 1203, 6201 BE Maastricht, The Netherlands
Abstract Waste reduction, which is the top priority of waste management in industrialized countries, can significantly reduce greenhouse gas emissions. The aim of this study was to improve the capacity of Finnish decision makers to implement long-term waste management policies, such as waste reduction. Focused interviews were conducted in 1992 with 24 researchers, consultants, politicians, government officials, and entrepreneurs. The interviews are the empirical basis of cognitive maps, which represent the causal models underlying expert decisions. The analysis indicates that Finland's environmental institutions integrate conflicting policy interests and systematically prevent decision makers from taking long-term policy action. Administrative and procedural decoupling of conflicting interests characterizes the proposed reforms.
1. INTRODUCTION Waste reduction is the top priority of waste management policy in industrialized countries. Waste reduction can significantly reduce greenhouse gas emissions, because it entails a radical reduction in the material and energy intensity of industrial production, particularly in the burning of fossil fuels. But implementing long-term environmental policies, such as waste reduction, involves significant socio-political and institutional contingencies (1). This study set out to explore the institutional constraints of long-term waste management policy in Finland. The goal of the study was to develop strategies to improve the capacity of decision makers in Finnish waste management to control and adapt to uncertainties of the very longterm future. The main proposition of this study is that there exists an ironclad relationship between Finnish environmental institutions and the expert beliefs that uphold the institutions. Institutions such as laws, regulations, cultural traits, and habits are the rules that organizations follow in the social game. Institutions determine what individuals perceive to be possible to achieve, and these perceptions in turn shape social institutions (2). According to this study, far-sighted waste reduction policy is currently impossible in Finland, because the dominant environmental institutions nurture waste management experts' short-term operating assumptions, which in turn weave together the short-sighted institutional structure.
1110 2. COGNITIVE MAPPING OF EXPERT BELIEFS The perceptions of central decision makers and experts in Finnish waste management were investigated by means of a policy analytical approach known as cognitive mapping, which is based on the notion that it is not the empirically verified reality that determines our decisions, but rather what we perceive to be the reality (3-5). Analysis of the mental models upon which experts and decision makers in waste management base their decisions identifies many institutional constraints of policy making. Focused interviews were conducted in the summer of 1992 with 24 Finnish waste management experts and decision makers on problems that they perceive to become central in the country's waste management in the next 50 years. The interviewees, who were selected through snowball sampling (6), represent various interest groups in waste management, with 4 consultants, 5 researchers, 5 politicians, 6 government officials, and 4 entrepreneurs. The interviewees mentioned 282 different, causally related problem statements in the field of waste management. For analytical purposes the problem descriptions were coded as problem networks (7). Each interviewee's scenario of waste management problems can be represented as a problem network composed of nodes and links, where nodes are problem statements about future waste management and links the causal relationships between them as expressed by the interviewees. Since the interviewees' descriptions of future problems contain some elements in common, individual problem networks can be aggregated into "socially constructed" scenarios. Problem networks reveal mental constructs that an individual expert does not necessarily perceive. The circular network, or loop, is the most interesting one for policy planning, because it obscures the difference between cause and effect. Since the loops are made of statements that the interviewees perceive as problematic, they are unstable, positive feedback loops. Problems included in a loop keep reinforcing themselves (on feedback, see 8-9).
3. ENVIRONMENTAL CORPORATISM IN FINLAND The results of cognitive mapping can be summarized as follows: First, waste management experts typically describe future waste management problems as loops. Fourteen of the 17 loops that emerged in network aggregation by interest group were mentioned by individuals, and half of the 24 individuals mentioned loops. Loops are held together by a cognitive goal conflict between profit maximizing goals, which prioritize shortterm economic profitability, and sustainable goals, which aim at preservation of ecosystems over generations. Second, experts do not let the goal conflict interfere with their day-to-day decision making. All of the loops indicate that expert advise and decisions are guided by profit maximizing, short-term operating assumptions. Third, the loops indicate that the profit maximizing operating assumptions are institutionalized in the administrative, technological, economic, and political structures of the Finnish society. This phenomenon will in the following be referred to as environmental corporatism (on social corporatism, see 10). Its most prominent feature is the systemic integration of conflicting environmental policy interests, to the extent that open conflict resolution is impossible. Finally, the interviewees expect that profit maximizing operating assumptions will lead to troublesome consequences in the long run. The aggregated problem networks have 54 terminal problems, i.e., problems without
1111 outgoing causal links, 40 of which describe threats to the survival of Finnish waste management organizations, society, and ecosystem. Each of the 17 loops identified in the group-level aggregation of problem networks and the terminal problems that result from them support the results. Interviewee no. 1 (a government official) mentioned loop 3, which illustrates the results (Figure 1). The loop describes how Finland's semi-governmental hazardous waste treatment monopoly Ekokem is in a cycle of planning excess treatment capacity only to find the capacity inadequate when environmental regulators order more of the nation's hazardous wastes to be treated at the plant. The dual goals of short-term economic profitability of the plant and long-term ecological safety of waste treatment forces decision makers into a cognitive dilemma, in which they can but alternate their allegiance between the conflicting goals (the first result).
4 Conflict between waste reduction and disposal is polarized.
Environmental officials direct more wastes to Ekokem.
19 Ekokem is designed and redesigned to have excess disposal capacity.
I
~
Ekokemreceives so much waste that it cannot incinerate all of it.
Figure 1. The government officials' loop 3.
Loop 3 also shows interviewee 1 to believe that actual waste management policies will conform with profit maximizing operating assumptions (the second result). Ekokem is described as an automaton, which keeps on expanding as a result of continuous planning for excess capacity. Two features of environmental corporatism secure the operation of the automaton (the third result). First, the corporatist decision making system views waste management purely from a techno-economic point of view, which obfuscates the socioeconomic conflicts of interest between waste reduction and waste treatment. Second, implementors and regulators have intimate linkages in the administration of hazardous waste management - - the Ministry of the Environment is the top regulator of hazardous waste management but also owns a third of Ekokem - - which secures the flow of waste to Ekokem. Finally, the terminal problems emanating from loop 3 support the fourth result. According to interviewee 1, decision making may become systematically irrational, enterprises may lose all interest in sustainable waste management and focus on turning a profit regardless of means, and regulators may end up shifting waste from one environmental sector to another.
4. R E C O M M E N D A T I O N S The central principle of the following recommendations is the administrative and procedural separation of conflicting environmental policy interests. The objective is not to polarize environmental conflicts, but to resolve issues through existing conflict resolution
1112 mechanisms. Where they do not exist, they should be created. First, the close integration of implementation and regulation in Finnish waste management persuades regulators to compromise long-term ecological considerations for the sake of shortterm economics. Regulation should therefore be clearly separated from implementation in waste management. Second, the socio-economic conflict between different technological stages of waste management is a problem particularly in public waste management agencies, which are not just technical implementors but policy makers as well. Administrative separation of waste reduction, recycling, collection, and disposal in the public sector would increase the organizational autonomy of sustainable principles. Third, administrative separation of the technical steps of waste management would not remove the economic friction between them. It would just transform an intra-agency conflict into an inter-agency one. More attention should therefore be paid to political procedures for resolving such conflicts. Environmental impact assessment should be developed into a procedure that would promote scientifically enlightened political discourse on environmental policy (11). Policy choices in waste management would be made after comprehensive public criticism, much like the scientific community selects theories after competition between scientists. Finally, more neutral regulatory mechanisms, such as economic regulation, would dismantle some of the structures of environmental corporatism. The institution of environmental taxes, for example, would require political decisions, which would transfer negotiations from the closed corporatist arena to the open parliamentary one. This would force politicians to make the difficult choices between short-term economics and long-term ecological sustainability. What is more, it would allow the regulators to concentrate on what they do best, namely, monitor and evaluate the effects of regulation on environmental quality.
5. R E F E R E N C E S
1 Bolin, B. (1994). "Science and Policy Making," Ambio, Vol. 23, No. 1, pp. 25-29. 2 North, D.C. (1992). Institutions, Institutional Change and Economic Performance. Cambridge: Cambridge University Press. 3 J. Management Studies (1992). Special Issue on Cognitive Maps. Vol. 29, No. 3. 4 Berger, P.L. and T. Luckmann (1967). The Social Construction of Reality: A Treatise in the Sociology of Knowledge. New York: Doubleday Anchor Books. 5 Weick, K.E. (1969). The Social Psychology of Organizing. Reading: Addison-Wesley. 6 Goodman, L.A. (1961). "Snowball Sampling," The Annals of Mathematical Statistics, Vol. 32, No. 1, pp. 148-170. 7 Pearl, J. (1988). Probabilistic Reasoning in Intelligent Systems: Networks of Plausible Inference. San Mateo: Morgan Kaufmann. 8 von Bertalanffy, L. (1968). General System Theory: Foundations, Development, Applications. New York: George Braziller. 9 Prigogine, I. and I. Stengers (1984). Order Out of Chaos: Man's New Dialogue with Nature. Toronto: Bantam Books. 10 Pekkarinen, J., M. Pohjola, and B. Rowthom (eds.) (1992). Social Corporatism: A Superior Economic System? Oxford: Clarendon Press. 11 Taylor, S. (1984). Making Bureaucracies Think: The Environmental Impact Statement Strategy of Administrative Reform. Stanford: Stanford University Press.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1113
Energy Production on Farms - Sustainability of Energy Crops H. van Zeijts Centre for Agriculture and Environment, P.O. Box 10015, 3505 AA Utrecht, The Netherlands
Abstract
This article reflects the results of a study on sustainability of energy crops. Contribution to the reduction of the greenhouse effect and other environmental effects were investigated for the Netherlands. The study assumed that energy crops are grown on set aside or grain land. Generating electricity and/or heat from hemp, reed, miscanthus, poplar and willow have the best prospects. These crops are sustainable and may in the future be economically feasible. Ethanol from winter wheat has the most favourable environmental effects, but is economically not interesting. Liquid fuels from oil seed rape and sugar beet are not very sustainable.
1. I N T R O D U C T I O N Energy from arable crops may be interesting for both agriculture and the environment. Energy crops can provide new opportunities for agriculture; for the environment it can mean a reduction in the greenhouse effect. The Dutch Centre for Agriculture and Environment (CLM) has investigated the contribution of energy crops on limiting the greenhouse effect. The study was done for Dutch arable farming. CLM has investigated nine energy crops: - winter wheat and sugar beet, for the production of bio-ethanol to replace petrol; - oilseed rape, for the production of methyl-ester to replace diesel; - the annual crops hemp and silage maize and the perennials reed, miscanthus, poplar and willow, for the production of electricity and/or heat. Besides the contribution of these crops on reducing the greenhouse effect attention was also given to other criteria: the emission of nutrients and pesticides, contribution to aridity, erosion, utilisation of by-products and waste, use of space and the contribution to natural and scenic values.
2. E N E R G Y
CROPS AND GREENHOUSE
EFFECT
Presently biomass provides only 2% of the European energy requirement 1. Since the seventies, the role of biomass in the energy supply has regularly been discussed. In the Netherlands this discussion was initially unfavourable for energy crops, principally due to the low profit and the high cost price of Dutch arable products. The attention drawn by the greenhouse effect and the poor economic circumstances of Dutch arable farming has brought the cultivation of energy crops to the fore once again. Recent studies have shown that energy crops can indeed limit the greenhouse effect2.
1114 Energy crops have two effects: - avoiding emissions from fossile energy sources; - fixation of carbon from CO2 in biomass.
Avoiding emissions from fossile energy sources Energy from crops replaces fossile energy. Thus, the emission of CO2 caused by using fossile fuels is lowered. Carbon dioxide from the use of energy crops is part of a cycle: it was taken up during the growth of the energy crops. On the other hand cultivation, transport and processing also causes emission of greenhouse gases. The balance can be calculated in terms of the net avoided emission of CO2. Table 1 indicates the net avoided emission of CO2. Generating electricity by burning reed, hemp, miscanthus and poplar has a great effect on the net avoided emission of CO2 per ha. Producing transport fuels from winter wheat, sugar beet and oilseed rape scores far less favourable on the net avoided emissions.
Fixation of carbon from CO2 in biomass Energy crops temporarily fix CO2 from atmosphere in biomass. This also contributes to a reduction of an increase of the greenhouse effect. The fixed amount of CO2 correlates with the growth stadium of the crop. However, on a longer term the amount of fixed carbon from CO2 remains the same. On a long term scale, the contribution of fixed CO2 on limiting the greenhouse effect is therefore far less important than the contribution of the net avoided emissions from fossiele fuels. Table 1 shows that over a period of a hundred years the contribution of fixed CO2 is only a few percent of the contribution of the avoided emission from fossile fuels.
Table 1 Net avoided CO2 emission and CO2 fixation, for the Dutch central clay area 3 energy crop
reed hemp miscanthus willow maize poplar sugar beet winter wheat oilseed rape
type of energy
electricity, electricity, electricity, electricity, electricity, electricity,
50 MW powerplant 50 MW powerplant 50 MW powerplant 50 MW powerplant 50 MW powerplant 50 MW powerplant ethanol ethanol methyl-esther
net avoided CO2 emission (ton ha -1 yr-1) 20.3 18.0 15.3 14.6 13.2 10.7 6.9 3.1 3.0
CO2 fixation, (% of total net avoided CO2 over 100 yr)
1115 3. O T H E R E C O L O G I C A L E F F E C T S Basic assumption in the study is that energy crops are grown on the area that has been set aside and on part of the area that presently is used for grain production. The environmental consequences of this substitution are divided into direct and indirect effects. Here we just work out a few examples.
Direct effects Direct effects have to do with the following question: is the environmental burden of energy crops higher than those of fallow land or grain cropping? Examples for direct effects are: 9 The use of pesticides in winter wheat is relatively high. Hemp, reed, miscanthus, poplar and willow only need low quantities of pesticides. This means that substituting winter wheat by these crops leads to lower emissions of pesticides. 9 Winter wheat and crops on fallow land require little water. Substitution by energy crops leads to increased water use and therefore contributes to higher aridity of the land. This may have negative effects on nature and on agriculture itself. Indirect effects Indirect effects are effects on other crops at farm level and effects on a regional and national level: 9 At farm level, substitution of fallow and grain land by energy crops has effects on the emission of nutrients and pesticides and on erosion in the rest of the cropping pattern. On Dutch arable farms this concerns sugar beet and potato. In particular, fitting in perennial energy crops leads to intensification of the cropping pattern that may cause problems from an environmental point of view. 9 Fitting in energy crops has consequences for natural and scenic values at farm level as well as at a regional level. For example winter wheat and oil seed rape have great potential natural values and can contribute to natural values at farm and regional level in a positive way. 9 Growing energy crops also influences the Dutch animal breeding sector, because these crops compete with fodder crops for use of land. If arable farmers grow energy crops instead of grain, there will be less native grain on the market. As most of Dutch grain is processed into animal feed, this means that the imports of raw materials for animal feed will increase. The extra transport of raw materials for animal feed increases the CO2-emission and causes an extra disturbance of the Dutch national mineral balance.
4. C O N C L U S I O N S Table 2 summarises overall results on both ecological and economical sustainability of nine energy crops. Each of the used criteria for ecological sustainability (see w1) is given equal weight. Of course this choice is arbitrary: in practice, the ratios in each situation are different. Given unequal weight the ranking may slightly change. The ranking for economical sustainability is derived from results of other studies 2 4.
1116 Table 2 Ecological and economical sutainability of growing energy crops in the Netherlands energy crop
type of energy
ecological sustainability
economical sustainability
reed hemp miscanthus willow maize poplar sugar beet winter wheat oilseed rape
electricity and/or heat electricity and/or heat electricity and/or heat electricity and/or heat electricity and/or heat electricity and/or heat ethanol ethanol methyl-esther
0/+ 0/+ 0 0 -/0 0 + -/0
+ 0 + + 0 + -/0
+ : good long-term perspectives from an ecological and economical point of view - : low perspectives, compared to the other energy crops
From table 2 we can draw the following conclusions: 9 Liquid fuels from oilseed rape and sugar beet and electricity from maize score worst. 9 Production of ethanol from winter wheat scores highest in terms of ecological sustainability. But economic studies reveal that the costs per ton CO2 net avoided are rather high. 9 Generating electricity and/or heat from reed, hemp, miscanthus, willow and poplar can be a sustainable way to reduce the greenhouse effect. Generating electricity from crops will be profitable in the Netherlands in the near future, if the set-aside scheme of the European Union is continued and an environmental tax on energy or a subsidy per ton of avoided CO2 is introduced. But opportunities for energy crops are better in other European countries, due to the intensive Dutch arable farming and high land prices. The long-term perspectives for energy crops are uncertain. Therefore it is advisable to investigate at a European level what other future functions for the land may be supplanted by energy crops.
5. REFERENCES 1. Hall, D.O. (1991). 'Biomass energy'. In: Energy policy, yr. 19, nr. 8. p. 711-737. 2. Lysen, E.H., C. Daey Ouwens, M.J.G. van Onna, K. Blok, P.A. Okken and J. Goudriaan (1992). D e h a a l b a a r h e i d van de p r o d u k t i e van biomassa voor de N e d e r l a n d s e energiehuishouding. Netherlands Agency for Energy and the Environment, Apeldoorn. 3. Zeijts, H. van, E.B. Oosterveld and E.A. Timmerman (1994). Kan de landbouw schone energie leveren? - Onderzoek naar de duurzaamheid van energiegewassen. Centre for Agriculture and Environment, Utrecht. 4. Biomass Technology Group (1994). Conversieroutes voor e n e r g i e g e w a s s e n - Een overzicht van bestaande en mogelijke routes voor de produktie van elektriciteit en transportbrandstoffen. University of Twente, Enschede.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1117
A simple method to estimate regional yields of biomass crops S. Nonhebel, J. Goudriaan, R. Rabbinge, D e p a r t m e n t of T h e o r e t i c a l P r o d u c t i o n Ecology, A g r i c u l t u r a l U n i v e r s i t y Wageningen, PO BOX 430, 6700 AK Wageningen, The Netherlands.
Abstract The use of biomass crops as an energy source is frequently mentioned as an option to reduce CO 2 emissions. To evaluate the possibilities reliable yield e s t i m a t e s of biomass crops are required. In this p a p e r a simple m e t h o d is developed to estimate regional yields of various biomass crops, based on the linear relation between intercepted light and biomass production. The quality of the estimates was studied by using the method to estimate yields of several a g r i c u l t u r a l crops in two regions in The N e t h e r l a n d s . In general a deviation of less t h a n 10 % was found between actual and estimated average yield.
1. I N T R O D U C T I O N One of the options to diminish present CO 2 emissions is replacing a part of the fossil fuels by energy from crops. During their growth crops capture CO2, which is a g a i n released when the crops are used for e n e r g y supply. This m e a n s t h a t CO 2 is recycled. For a s s e s s m e n t of the possibilities of this energy source reliable estimates of the yields are required, since the expected yield of an energy crop determines the outcome of the evaluation. Candidate biomass crops (used for electricity production) like willow and m i s c a n t h u s are not yet grown on a large scale, so t h a t it is difficult to assess their yields. In current energy scenarios and in recent studies of possibilities of using biomass crops as an energy source (1,2,3), the yield estimates are based on a limited n u m b e r of field experiments and data are used for large areas (sometimes even globally). However, the yield of a crop is strongly determined by its growing conditions and large differences in yields between regions and years are observed. This
1118 implies t h a t yields obtained in one region in one year cannot easily be translated into yields in other regions or other years. For energy supply not the yield in one particular field is of interest but the amount of energy t hat can be produced in a region. Therefore yield estimates have to be on a regional rather t h a n on a farm scale. This regional yield cannot be determined in field experiments. Here a simple method is developed to estimate regional yields of biomass crops. The method is based on knowledge obtained in agricultural research.
2. METHOD Research on various agricultural crops has shown t hat a linear relation exists between the amount of solar radiation intercepted by the crop and the above ground biomass produced (4). The slope of the line is the so-called Light Use Efficiency (LUE). Under optimal conditions a LUE value of 1.4 g MJ -1 (global radiation) is found for most agricultural crops, and also for fast growing trees (4,5). This implies that the yield of a crop (Yp) can be calculated by: Yp=Iin t x LUE x HI
(1)
in which Iint is the radiation intercepted during the growing season and HI is the harvest index (fraction of the total biomass that is harvested). Data of both HI and Iint can be derived from literature (6). Yp is the production under optimal circumstances (the crop is growing with ample supply of water and nutrients and free from pests, diseases and weeds), it is a measure of what is potentially possible under given conditions In practice conditions are seldom optimal and and actual yields are generally lower than the calculated potential yield. To obtain actual yields, a correction is needed to account for suboptimal growing conditions. The ratio between actual and potential yield can be interpreted as a characteristic for the type of agriculture in a region. Here this ratio will be called the Yield Correction Factor (YCF). The value of the YCF can be determined by using the above described method for an agricultural crop from which yield data are available and divide actual obtained yield (Ya) by calculated potential yield (Yp): YCF = Ya Yp
(2)
1119
Estimates for regional crop yields under present growing conditions (Yr) can be obtained by using eq 3 Yr=Yp • YCF
(3)
3. RESULTS AND DISCUSSION 3.1. D e t e r m i n a t i o n Y C F f o r t w o d i f f e r e n t regions In The N e t h e r l a n d s potatoes are planted in April and the crop is harvested in September (7). The total amount of global radiation intercepted by the crop is about 1400 MJ m -2. Using eq 1 leads to a potato yield of 15.0 ton ha -1 (HI of a potato crop is 0.75). In 1992 the actual potato yield in Flevopolders (region 1, fig 1) was 10.6 ton ha "1 and in Veenkoloni~n (region 2, fig 1) it was 8.6 ton ha -1 (dry matter) (8). Applying eq. 2 results in a YCF for region 1 (YCF 1) of 0.71 and for region 2 (YCF 2) of 0.56.
S I
O0 km
Figure 1. Location of the regions mentioned in text.
1120
3.2. Regional yields of agricultural crops Since there are no data on average regional yields for biomass crops, the validation possibilities for the method are limited. The only available averages are those from the present agricultural crops. The method described above was used to estimate yields of three agricultural crops and the results were compared with actual average yields obtained in the two regions in 1992 (table 1).
Table 1 Comparison between actual yields (Ya) of agricultural crops and estimated yields (Yr) for two different regions, YCFI=0.71 and YCF2= 0.56. Deviation (dev in % of Ya) is also given. Yields (harvestable biomass) in ton dry matter ha -1. region 1 crop winter wheat sugar beet maize
region 2
Yr
Ya
dev.
Yr
Ya
dev.
7.6 15.1 14.6
7.3 15.4 15.4
4% 2% 5%
6.1 12.3 11.9
5.4 12.0 13.2
13% 3% 10%
The deviation between simulated and actual yields was not large, which shows that the method is a suitable tool for estimating crop yields.
3.3 Regional yields of biomass crops The application of the YCF for biomass crops assumes that knowledge on how to grow such a crop is comparable to that of an agricultural crop. However for the potential biomass crops this is not yet the case. So it is likely that the regional yields of these crops will be lower t han calculated here. F u r t h e r uncertainties exist with respect to values of intercepted radiation and harvest index of the biomass crops. This means that the yield data used for the biomass crops are only indicative. However, it can be concluded t h a t the potential biomass production of 'a' perennial biomass crop lies between 18 and
1121 22 ton ha -1. Present regional averages of these crops would be 13-15 ton ha -1 for the high yielding regions and 10-12 ton ha -1 for the low yielding regions in The Netherlands (table 2). Estimates of biomass crops yields given in literature for present conditions are 10-12 ton ha -1 (1, 9) which agree with values found here. The value of YCF is time and space dependent and must be determined for each region individually. It is likely that YCF values in other regions in Europe will be much lower due to less well developed agriculture which will result in lower yields. To improve yield estimates, detailed field experiments are required to obtain more information on light interception and harvest index of candidate biomass crops.
Table 2 The calculated potential yield (Yp) and the regional yields (harvested biomass, in ton (dry matter) ha -1) of three candidate biomass crops in two regions (Yrl, Yr2)" YCF1 =0.71 and YCF2= 0.56. Yields (harvested material) in ton (dry matter) ha -1.
Crop
harvested parts
Yp
Yr 1
Yr2
Miscanthus Poplar Willow
stems stems/branches stem/branches
21.9 18.0 19.6
15.3 12.6 13.7
12.3 10.0 11.0
4. CONCLUSIONS Based on the result that yields of agricultural crops are estimated with an inaccuracy of 10%, it is concluded that the estimation method described can be a useful tool in research on the possibilities of using biomass crops for energy supply. Estimated biomass crops yields are 10-15 ton ha -1 under present conditions in The Netherlands.
1122 Acknowledgement This work was funded by the Dutch National Research Program on Global Air Pollution and Climate Change Project nr 853117
5. REFERENCES 1 Hall, D.O., et a1.,1993. In: Johansson, T.B., Kelly, H., Reddy, A.K.N., Williams, R.H. (eds) Renewable Energy, Island Press, Washington, pp 593651. 2 0 k k e n , P.A., Ybema, J.R., Kram, T., Lako, P., Gerbers, D., 1994. Energy systems and CO 2 constraints. Netherlands Energy Research Foundation, 3 4 5 6 7 8 9
Petten, The Netherlands. Lysen E.H, et al., 1992. De haalbaarheid van de productie van biomassa voor de Nederlandse energie huishouding. Novem, Apeldoorn. (in Dutch) Monteith, J.L., 1977., Phil. Trans.R. Soc. Lond. 282, 277-294. Cannell, M.G.R. 1989. Scand. J. For. Res. 4, 459-490. Nonhebel, S. 1994.,A simple model to estimate regionally average yields of biomass crops, submitted to Biomass and Bioenergy Jong, J.A.,1985, De teelt van aardappelen,Drachten (in Dutch). PAGV, 1993. Kwantitatieve informatie voor de akkerbouw en de groenteteelt in de vollegrond, 1993-1994, PAGV, Lelystad (in Dutch). Christersson, L.et al, 1993. The Forest Chronicle 69, 687-693.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1123
Energy Accounting on Farms J.A.M. van Bergen Centre for Agriculture and Environment, P.O. Box 10015, 3505 AA Utrecht, the Netherlands
Abstract
This article describes the development of an energy accounting. This is a management tool to give farmers a clear understanding of their energy use and of the emission of greenhouse gases on their farm. Results are given of one year accounting on dairy farms and on pig farms. The results show large differences in energy use and in emissions of greenhouse gases between individual farms. These differences indicate that a substantial reduction of emission of greenhouse gasses is possible.
1. I N T R O D U C T I O N The Dutch Centre for Agriculture and Environment (CLM) has figured that agriculture in the Netherlands contributes 12 percent to the national emission of greenhouse gases (1). These emissions of carbondioxide, methane and nitrous oxide in Dutch agriculture have also been calculated per sector. Dairy farming has the major contribution with a total emission of 57 percent. Intensive livestock farming contributes 20 percent, greenhouse cultivation contributes 16 percent and field cultivation 7 percent (1). The high score of dairy farming and intensive livestock farming is partly caused by the high use of indirect energy. The importance of the use of indirect energy in animal production has been pointed out in literature earlier (for example 2,3). The CH4-emission by ruminant digestion of feed and the N20-emission by soil processes are the other causes of the high score of dairy farming. So far, no management tool is available to monitor the emissions at farm level. CLM has started a project to develop an accounting system to calculate the use of energy and the emission of greenhouse gases at individual farms. It should give farmers a clear understanding of their energy use and of the emission of greenhouse gases on their farm. The instrument is simply denominated as the Energy accounting. Because of the important contribution of dairy and intensive livestock farming to the greenhouse effect, the energy accounting is first being developed for these sectors.
2. F R A M E W O R K OF AN E N E R G Y ACCOUNTING SYSTEM Setting up an energy accounting system includes developing a registration form, a methodology to calculate emissions and a way to present farmers the results. In addition the project deals with advice on possible strategies and measures to reduce emissions of greenhouse gases. In this article we describe only the framework of the energy accounting.
1124
The energy accounting is based on administrative management data of farms, for example meter readings on energy use or energy bills, data on use of fertilizers and feed concentrates. These farm data are combined with standard factors on energy values and emissions of greenhouse gases. The calculations take place in six modules: - direct energy use and CO2-emission for both dairy and intensive livestock production; - indirect energy use and CO2-emission for both dairy and intensive livestock production; - CO2-emission by mineralization of peat for dairy production; - CH4-emission by feed fermentation for dairy production; CH4-emission by slurry storage for intensive livestock production; under Dutch conditions the CHa-emission by storage of slurry from cows is neglectable; - N20-emission by soil processes for dairy production. -
Table 1 Framework of the energy accounting Module
Basic data per farm
Calculations w i t h standard factors
Results MJ
Results CO2-emission
1. Direct energy use fuel/electricity
meter readings
MJ/I diesel, kWh
MJ
kg CO2
2. Indirect energy use
use of fertilizers, feed concentrates, tools and buildings
MJ/kg N, kg concentrate etc.
MJ
kg CO2
3. CO2-emissionby mineralization of peat
area peat s o i l , drainage
CO2-emission/ha
kg CO2
4. CH4-emissionby feed fermentation
number of c a t t l e , feed ration and level
CH4-emission/cow
kg CO2-equi.
5. CH4-emissionby slurry storage
slurry q u a n t i t y storage days
CH4-emission/ton
kg CO2-equi.
6. N20-emissionby soil processes
area, soil type, fertilization, grazing, drainage
N20-emission/ha
kg CO2-equi.
Totals per farm per product
MJ MJ
kg CO2-equi. kg CO2-equi.
Table 1 gives a schematic view of the framework of the energy accounting. The emissions of CH4 and N20 are converted to an emission of CO2-equivalents. Summation of the emission of the six modules results in a total emission per farm. For comparison of individual farms, this total is expressed in the form of an efficiency-figure, for example kg CO2-equivalent per 100 kg milk. The same applies to the total energy use, for dairy expressed as MJ per 100 kg milk.
1125
3.
F I R S T
R E S U L T S
The system mentioned aboved is now being tested in three study groups of dairy farmers, and four study groups of intensive livestock farmers. Table 2 shows the results of the first testing year of three dairy farms and two pig farms. The farms presented here are selected for the differences in their CO2-equivalent emissions.
Table 2 CO2-equivalent emissions (kg CO2/100 kg milk, kg CO2/100 kg growth) Direct energy Indirect energy
CO>min.
CH4
N20
Total
dairy farm 1 dairy farm 2 dairy farm 3
6 4 6
28 14 27
201 -
27 25 22
100 17 7
362 60 62
pig farm 1 pig farm 2
21 12
204 179
-
81 38
-
306 229
The first results give a good insight in the significance of the emissions and in the differences that were found between individual farms. Concerning the dairy farms, the following results were the most striking: for peat soils, CO2-emission from mineralization as well as N20-emission from soil processes (farm 1) have a large influence on total CO2-emission, compared to total CO2emission from clay (farm 2) and sandy soils (farm 3); the differences in indirect energy related CO2-emission are of much more importance than the differences in direct energy related CO2-emission; on farms with clay and sandy soils, the CO>emissions caused by the use of direct and indirect energy, the CHa-emission and the N20-emission are each of the same importance. The emissions per module vary not only between farms on different soils but also between farms on the same soil. In the study group with sandy soils the extreme values in total CO2emission were 62 and 86 kg CO2 per 100 kg milk. The results indicate that farmers can reduce the emissions of greenhouse gases by improving their efficiency of energy use, of fertilizer use and of concentrate and feed use. The results of the two pig farms show clearly that : the use of indirect energy in pig breeding is very important; the variation in CO2-equivalent emission between the two farms caused by use of direct and indirect energy is of the same magnitude than the variation caused by storage of slurry. The differences indicate that on pig farms there are possibilities to improve use of direct and indirect energy. Farmers can either make energy-saving investments in climate control or improve their general and feeding management. The calculated difference in methane emission is mainly caused by a difference in storage time: technical measures to reduce this emission are developed. -
1126
4. DISCUSSION The results have a provisional character. Changes in the methodology to calculate emissions are possible during the testing years. The purpose of this project is to create a management instrument for farmers. It is in discussion to which extend emissions should be part of this instrument that hardly can be influenced by farmers. Most discussion is about the mineralization of peat (4). Possible changes also depend on the availability of relevant farm management data and on changes in knowledge of emissions. An important source of new knowledge is the NRP research on emission of nitrous oxide from grassland. The variation in the results from individual farms show that farmers do have possibilities to reduce their energy use and emissions of greenhouse gases. These possibilities can lead to a substantial reduction of emission of greenhouse gases. In the second year of the project more research will be done on the contribution of advice on energy-saving investments and on better management practices to reduce the emission of greenhouse gases on farms.
5. R E F E R E N C E S
1 Bergen J.A.M. van and E.E. Biewinga (1992). Agriculture and Greenhouse effect, Survey to reduce the Emissions of Agriculture and Horticulture. (In Dutch with English summary), Centre for Agriculture and Environment, Utrecht. 2 Leach, G.A. (1976), Energy and Food Production. IPC Science and technology Press Limited, Guilford. 3 Pimentel, D. (1980) Handbook of Energy Utilization in Agriculture, CRC Press, Boca Raton. 4 Hanegraaf M.C. and E.E. Biewinga (1994). 'Use of Energy and Emission of Greenhouse Gases at individual dairy Farms'. In : Meststoffen 1994, p. 59-67.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1127
SPACE FOR BIOMASS An exploratory study of the spatial potential for the cultivation of biomass for energy in The Netherlands I. Steetskamp, A. Faaij, A. van Wijk Department of Science Technology and Society Utrecht University Padualaan 14, NL-3584 CH Utrecht, The Netherlands
Abstract The spatial and energy potential in The Netherlands for energy farming is assessed as well as for a number of biomass residues. The future supply of agricultural land is based on closures of farms. Various future claims for infrastructure and nature are taken into account. The net supply of land adds up to 100,000 - 185,000 in 2000 to 245,000 and a theoretical maximum of 465,000 ha in 2015. When this potential is used for energy crops like Miscanthus this land could contribute 20 37 PJ in 2000 and in 2015 62 - 117 PJ. Secondary yields of biomass can contribute a further 32 PJ in 2000, decreasing to approx. 24 PJ in 2015 This implies 2% of the Dutch energy demand in 2000, in 2015 about 3%, provided that energy farming is an economically feasible activity for farmers. 1. I N T R O D U C T I O N The role of biomass as a 'renewable' source of energy is once again the centre of attention for a variety of reasons. Technological developments make it possible to achieve a far higher yield from the conversion of biomass into electricity or fuel than in the past. Developments in the agricultural industry, such as the predicted shedding of agricultural land, also play a role. A study conducted by the Scientific Scientific Council for Government Policy (WRR), entitled 'Ground for choices', outlines a number of scenarios for agricultural use of the land in the European Union depending on the agricultural policy currently in force. Agricultural land is released in each of the described scenarios (1). This land could, however, be used for the cultivation of crops suitable for energy production (energy cultivation). The available surface area is a decisive factor in determining the energy potential of biomass. If agricultural land falls vacant in the Netherlands, there will be several sectors lining up to use it, given the high population density. This has resulted in the formulation of the following question: What is the spatial and energy potential of biomass production in the Netherlands in the long term (2000/2015), with or without other functions, seen together with other claims on the space? This exploratory study focuses primarily on land that is not used for other types of agriculture (any longer). Energy cultivation on this land is possible, provided it is economically viable for agricultural industry (alongside food production). If, however, the yield of energy crops increases in the future, competition with food production may become possible. The spatial potential would then be on a quite different scale.
1128 2. M E T H O D O L O G Y
& RESULTS
2.1 T h e s p a t i a l p o t e n t i a l : s u p p l y a n d d e m a n d for a g r i c u l t u r a l l a n d
The supply of and demand for agricultural land are based on calculations made in the LEI (Agricultural Economic Institute) study 'Regional Land Balances'(2). The base calculation in the study shows a total supply of 280,000 hectares in the period 1990-2000. In a high supply scenario this surface is 410,000 hectares. This LEI study assumes that land that falls vacant comes from closures of farms. An average closure percentage and an average farm hectarage, which incorporate upto-date developments in the sector, were used as a basis for estimating the total surface area of the land which will become vacant. A large part of the available land is grassland with a milk quota. This study assumes that grassland with a milk quota will be used for the same purpose after it goes on offer. We have also assumed that the claims of the intensive livestock farming industry and horticulture (under glass) will be honoured, entailing approximately 8,000 hectares until the year 2000. The demand for non-agricultural land can be divided into 'hard' claims and other claims. Hard claims on land are laid by housing, industry, traffic and military training grounds. Other claims come from forestry, nature and recreation. An analysis of each of these functions has been carried out. Each function was studied to ascertain the expected spatial development and how this translates into a claim on land. A full description of the applied methods is given in (3). Table 1 shows the spatial potential for energy cultivation in the year 2000 for the basic supply of 280,000 hectares and the variant with a higher supply of 410,000 hectares. The basic assumption is that the 'hard' claims will be honoured, and they have been deducted from the supply of agricultural land. This imposes an upper limit on the spatial potential.
Table 1
Spatial potential for energy farming on agricultural land in 2000, calculated as of 1990, in hectares (x 1,000).
Land supply until 2000 l
Hard agricultural claims2
Basic supply 280
105
Higher supply 410
200
Hard nonagricultural claims3
Other nonagricultural claims4
Spatial
potential 100 - 150
24
53
135- 185
The supply of land in the LEI study was based on 1989. This figure is translated to the period as of 1990. Claims from livestock farming (grassland with milk quota) and claims from intensive livestock farming and horticulture (under glass). Claims from housing, industry, traffic and military training grounds. For calculation, see table 9.5. Claims from forestry, nature and recreation.
1129
The spread of the spatial potential depends on whether the other, non-agricultural claims will be honoured wholly, partially or not at all in the future. If all other non-agricultural claims are honoured, the lower limit on spatial potential will be reached. At a supply of 280,000 hectares, at least 100,000 hectares could be available for energy cultivation in 2000, up to a maximum of 150,000 hectares. For the year 2000, this estimate of the spatial potential according to the basic supply seems the most realistic. One must not forget that this potential is based on calculations as of 1990. No part of this potential had been realized by 1994. For 2015, a linear extrapolation was made of the data in the LEI study for 2000. The different kinds of claims were then deducted. In 2015 between 245,000375,000 hectares could be available at a supply of 28,000 hectares per year. According to the author of the LEI study, the linear extrapolation of the data for 2000 produced a conservative estimate of the basic supply in 2015 (4). The LEI study is an approximation at micro level (supply on farm level), and assumes implicitly that the land market wishes of every farmer will be honoured. Developments at macro level (agricultural production ceilings, for example) were not included in the study. This would have made it impossible to meet each individual farmer's wishes. Consequently, the supply after the year 2000 could be considerably higher than in the base calculations, while the claims remain the same. On the other hand, there are other agricultural developments underway (essential reductions in emissions of environmentally harmful substances, biological farming) which could lead to more extensive use of the land, producing in turn a lower supply after 2000 than envisaged in the basic calculations.
x 1.000 ha 500 450
~M.~ 2ooo I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
400
[~ Ml'n. 2015 [ ~'lMax. 2015 1
350 300 X 250 200
X x x
.....................
150 i 00
x N
2~ base supply
higher supply
Figure 1. Spatial potential for energy cultivation on agricultural land in 2000 and 2015 for a basic supply (28,000 ha/yr) and a higher supply variant (41,000 ha/yr), as of 1990.
1130 The crop yields vary according to type of soil, which is why the spatial potential is subdivided according to soil type. Of all agricultural regions in the Netherlands, approximately 700,000 hectares can be counted as high-yield areas and 1.3 million hectares as low-yield areas. The ratio of productive and less productive land released for energy cultivation depends on the extent to which the claims are honoured. Nature claims concern mainly agricultural land of lower quality. At a spatial potential of 150,000 hectares in 2000 for the basic supply, the ratio between high-yield and low-yield land is 1:1.4, resulting in 63,000 hectares with a potential high yield and 87,000 hectares with a potential low yield.
2.2 Energy farming yields on agricultural land Taking the estimated spatial potential and the division according to type of soil as a point of departure, it is possible to calculate the energetic potential. This estimate is based on Miscanthus as the energy crop, because it produces a high net energy yield. The yield figures were derived from data based on the model calculations (5). For the current situation 12.3 tons dm/ha/yr in high-yield regions and 10.6 in low-yield regions is projected. For 2015, 20% higher yields can be expected due to developments in cultivation technology, crop improvement and increasing experience with harvesting methods and maintenance. Using this data as a basis, a calculation was made of the annual yields in tons of dry solids. The net energy yield was calculated by combining the calorific value of the crops (19 GJ/ton din) with the dry solid yield (gross energy potential) and deducting the energy costs of cultivation. For Miscanthus, this produces a net energy yield of 20 to 30 PJ per year for at a spatial potential of 100,000 to 150,000 hectares in the year 2000. At a potential of 245,000 to 375,000 hectares in 2015, a net energy yield of 62 to 95 PJ per year is possible. At a potential of 330,000 to 465,000 hectares, this would rise to between 83 and 117 PJ per year in 2015. 2.3 Energy yields as a secondary function In addition to yields from energy crops on agricultural land, biomass yields can also be generated by non-agricultural activities and by-products of regular agriculture (waste flows such as organic waste and sludge have not been included). Yields from wood produced by thinning activities for forestry and recreation are particularly significant: approximately 15 PJ in 2000. Straw can contribute more than 8 PJ per year on the basis of the current agricultural hectarage. In the future, this contribution will fall as the hectarage for food production decreases. A yield of approximately 32 PJ per year is possible from secondary activities up to the year 2000. After 2000, this figure will fall to approximately 24 PJ per year by the year 2015 due to smaller straw yields on the one hand, and a lower proportion of thinning wood in the total volume of wood cut on the other. These figures assume that all claims for forestry have been honoured. If this is not the case, forestry hectarage will be smaller and the energy yield lower as a result. Table 2 shows a brief summary of the secondary yields. There is no data available for a number of biomass flows, and these have not been included in the table. In the case of these secondary yields, it should be pointed out that these flows already have an alternative application. For example, straw is sold to livestock farmers and the bulb cultivation industry, turf is composted and reeds are used for roofing.
1131 Table 2
Overview of secondary yields of biomass
Function
Type of material
Yield (ton ds/ haJyr 1)
Hectarage 1993/2000 (x 1,000)
Gross energy yield 1993/2000 (PJ/yr)
Hectarage 2015 (x l,O00)
Gross energy yield 2015 (PJ/yr)
i,~i ~ i i
1 Forestry I
Thinning wood
2.0
447/460
15.8
4472-497
11.2-13.1
2 Nature
Cut sods
1.4
35
0.8
35
0.8
!
Reed
4.0
0.1
!: 3 Traffic
Verge grass
5.1
!i il il
!i Parks and Gardens
Residual wood
i, Agriculture
Straw
'i
i Total
0.1 37
2.6
37
2.6
+16
4.4
+16
44
"
!i Jl !i
3.7
149
8.3
753
4.2
695/708
32.0
621-671
23.3-25.2
~t
Including forestry designated for recreation, nature and military training grounds. Lower limit if none of the claims for forestry are met (and a consequent maximum spatial potential is achieved). Assuming that the hectarage of grain falls by 50% due to the increase of spatial potential of energy farming.
Table 3
An overview of spatial potential and total energetic potential in 2000 and 2015. Basic supply
:' Spatial potential 2000 (x 1,000 ha) Energy yield 2000 (PJ/yr)
Energy farming
100-150
Secondary yields Energy farming
Energy farming
Energy yield 2015 (PJ yr)
Energy farming I
135-185
i',
i
20-30
27-37 32
52-62
Spatial potential 2015 (x 1,000 ha)
Ii,
708
Secondary yields Total energy yield 2000 (PJ/yr)
i! Higher supply
!
245-375
59-69 330-465
670-620
Secondary yields 62-95
Secondary yields
i
83-117
25 -23
!I .....
:i Total energy yield 2015 (PJ/yr)
87-118
108-140
This figure assumes higher yields of dry solids (20% increase) than in 2000. Taking the same yields of dry solids as in 2000, the figures would read 50-84 PJ/yr for the basic supply and 67-94 PJ/yr for the higher supply.
1132 3. DISCUSSION AND CONCLUSION The spatial potential of biomass is made up of two components: energy cultivation on agricultural land and potential biomass yields on land with another function. The net supply of land adds up to 100,000 - 185,000 in 2000 to 245,000 and a theoretical maximum of 465,000 ha in 2015. When this potential is used for energy crops like Miscanthus this land could contribute 20 - 37 PJ in 2000 and in 2015 62 - 117 PJ. Secondary yields of biomass can contribute a further 32 PJ in 2000, decreasing to approx. 24 PJ in 2015. In the year 2000, the total potential contribution of these two flows of biomass can contribute approximately 2% to the primary energy demand (as estimated in the Follow-up Paper on Energy Conservation (6)). In 2015, this total can be about 3%. The expected growth in energy consumption has already been calculated into these percentages. A linear extrapolation is made up to 2015 for the potentially available land. It should be pointed out, however, that the WRR study entitled 'Ground for Choices' does support the theory that more land than estimated in the base calculations may become vacant. It may then become possible to achieve the results of the greater supply variant (410,000 hectares per year), which was conducted in the LEI study as a sensitivity analysis: a spatial potential in 2015 of between 330,000 and 465,000 hectares. Compared to the study 'The feasibility of biomass production for the energy system in the Netherlands'(7), the estimate of energy potential on agricultural land is clearly lower. The study calculated a yield of 140 PJ. The fact that the estimates in this study lag behind has a variety of causes. Firstly, the calculation of the spatial potential in this study according to the basic supply is significantly lower (245,000-375,000 hectares in 2015) than the maximum estimated long-term spatial potential of 500,000 hectares in the mentioned study. Secondly, this study has assumed lower yield figures and differentiated according to land quality. Thirdly, this study has deducted the energy costs of the cultivation from the potential. It does, however, employ a higher calorific value based on data from recent research material.
ACKNOWLEDGEMENTS This study was sponsored by the Netherlands Agency for Energy and the Environment (NOVEM) REFERENCES 1. Scientific Council for Government Policy (WRR), Ground for choices, four perspectives for rural areas in the European Union; Vol. 42; Sdu uitgeverij, Den Haag, 1992 2. F.H. Bethe, Regional ground balances. Exploration of demand and supply of land until the year 2000, report 83, Landbouw-Economisch Instituut, Den Haag, 1991. 3. I. Steetskamp, A. Faaij, A. Van Wijk, Space for biomass, An exploratory study of the spatial potential for the cultivation of biomass for energy in The Netherlands, Department of Science Technology and Society, Utrecht University (in Dutch), December 1994. 4. F.H. Bethe Personal communication, October 1994. 5. S. Nonhebel, Potential yields of woody crops, in: Fuelwood, perspectives for forestry and energy production, Lelystad, January 1994. 6. Ministry of Economic Affairs, Follow up paper energy conservation, Sdu, Den Haag, 1993. 7. E.H. Lyssen et. al., The feasibility of production of biomass for the Dutch energy system, Netherlands organization tbr energy and environment, 1992.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1133
CONVERSION ROUTES FOR ENERGY CROPS; INTEGRATING AGRICULTURAL AND ENVIRONMENTAL OPPORTUNITIES IN EUROPE Eric J.M.T. van den HEUVEL, BTG Biomass Technology Group B.V.
Introduction Energy crops are interesting for both the agricultural and energy sector. Several conversion routes exist for the transformation of these crops into electricity, heat and/or transport fuels.The major environmental advantage lies in avoided CO 2 emissions to the atmosphere in comparison to the use of fossil fuels. Those conversion routes are analyzed with respect to technical/financial and environmental characteristics.
Conversion routes Energy crops are considered as an alternative for food crops in the agricultural sector. After harvest, they can be converted into electricity and/or heat and into transport fuels. For energy crops with high cellulose contents, like poplar, willow and miscanthus, thermo-chemicai conversion routes which convert the crop into electricity are most viable. Important thermochemical conversion technologies are combustion, gasification and pyrolysis. For high oil-content or sugar-content energy crops physical-chemical (extraction) and biochemical (anaerobic digestion or fermentation) routes are available. Those routes produce respectively gaseous and liquid fuels that can be used in transport applications. Upgrading of the secondary products of the gasification and pyrolysis technology processes also can lead to the production of methanol or bio-diesel. In Figure 1 these routes are presented.
I
hot w i r e r
oils
up=
,din
Figure 1: Available conversion routes for energy crops
1134
The technical, financial and environmental aspects of energy crops based electricity generation and production of transport fuels has been investigated. Large scale power generation based on combustion technology and application of steam cycles is technically mature. The same holds for the use of biogas in gas engines. Also the production of ethanol from sugar and grain crops as well as the production of rapeseedmethylester (RME) is technically viable. Biomass gasification integrated with a combined cycle (gas turbine and steam turbine utilization), cocombustion of pulverized or gasified biomass in conventional large-scale coal or gas fired electricity plants and production of methanol through gasification are currently demonstrated and expected to become technically mature around the year 2000. Newer technologies like the use of pyrolytic oil in a combined cycle applications or the use of synthesis gas from biomass gasification as a fuel source for fuel cells still need further research and will certainly not be commercial before 2010.
Some options for conversion routes
Several conversion options have been analyzed through spreadsheet models in order to determine: o the specific production costs (FI/kWh or FI/I fuel); o the specific amount of avoided CO 2 (ton CO2/ha) generated; and o the specific costs of avoided CO 2 (FI/ton CO2), calculated as the difference between the annual financial returns and the conversion option costs divided by the amount of avoided CO 2. The options (numbers correspond to numbers in Figures 2 and 3) considered are: Conversion routes for electricity .qeneration: 1 Small-scale co-generation of heat and power with combustion technology (5 MWel/ 10 MWth); 2 Medium-scale electricity generation with combustion technology (50 MWel); 3 Co-combustion of pulverized miscanthus (max 10% on energy basis) in conventional powder coal electricity plant of 500 MWel; 4 Small-scale co-generation of heat and power with circulating fluidized bed gasification and gas turbine technology (5 MWel/lO MWth); 5 Medium-scale electricity generation with integrated circulating fluidized bed gasification and combined cycle utilization (50 MWel); 6 Gasification of energy crops and combustion of the resulting producer gas in coal fired conventional electricity plant; 7 Gasification of energy crops and combustion of the resulting producer gas in natural gas fired conventional electricity plant; Conversion routes for production of transport fuels: 8 Fermentation of wheat for ethanol production (100 million I/year), with combined heat and power generation through combustion based on wheat straw; 9 Fermentation of sugar beet for ethanol production (100 million I/year), with purchase of required heat and power; and 10 RME production based on oil-extraction and esterification of rapeseed (1 million I/year).
Results
All technologies mentioned above are technically mature, except for the integrated gasification combined cycle (gas and steam turbine) technology, which is still in a demonstration phase. The specific production costs figures for electricity and transport fuels and the specific amount of CO 2 avoided are presented in Figure 2. The specificproduction costs for electricity are lowest for the large scale gasification option and the co-combustion options in large scale coal or gas fired electricity plants.
1135
50-
--I
combustion
t ~ ]
gasification
t
~
t transport, fuels I
45-
180
40-
160
r : x~ vo
35-
,a
3oI
4~
25-
.O 4-1
20-
o
200
o
-140
r
-120
G) ::~
-IOO
tO
-8o
. . . .
15-
(D
I
.........
i
60 -40
10-
0
c~ r,/) r (~i
-20
1
2
3
4
5
6
7
8
I
i 9
(0 10
options (see text for expl.) I L~
electricity
~
transport, fuels !
Figure 2: Specific production costs of several conversion routes
At the moment, with current fossil fuel prices, none of these technologies are economically viable. The annual returns are lower than the annual costs. Utilization of sustainably grow energy crops for electricity generation results in avoided CO 2 emissions, because fossil fuel based electricity is replaced. For each option the amount of annually avoided CO 2 emission is calculated and from this the specific costs of avoided CO 2 emission are determined. The results are summarized in Figure 3. It can be concluded that thermochemical conversion routes, like combustion and gasfication processes, for electricity production have the highest potential for reducing CO2-emission (given as tonnes CO2/ha). This indicates that per ha of land used for energy crops through thermochemical conversion routes most CO 2 emission will be avoided. At the same time the specific costs for CO 2 emission are also much lower for electricity production as compared to the production of transportation fuels. It is therefore concluded that future research activities must concentrate on high-efficiency electricity production, from energy crops. The most promising routes seem to be large scale gasification and the co-combustion in existing power plants. This justifies research to further develop these technologies. The major problems with gasification lies in the required producer gas cleaning for gas turbine utilization and in the adaptation of gas turbines to low heating value gasses. For co-combustion it must be researched whether to co-combust producer gas or pulverized biomass. A study on the environmental impacts of the Netherlands Center for Agriculture and Environment (CLM) has shown that utilization of the energy crops miscanthus, poplar, willow, hemp and reed, under the Dutch agricultural system, will probably have the lowest negative impacts. Therefore, these crops will be considered in future research.
1136
! 0as'''~176 / i /trans ~
,07] [com us,,on I
2OO -180 -160
~" 35r
-140
04 30O O 25-
-120 O O -100
04
..........
~= 20cr
-80
:4i
1510-
-60
,-, o o
-40 20
0 1
2
0
-~ ..... 3
4
5
6
7
8
9
10
options (see text for expl.) 1~
quaniity ~
costs
!
Figure 3: Avoided CO 2 emission
References Johansson et al. (eds.), 1993, Renewable energy, sources for fuels and electricity, Island Press, Washington DC. USA. Lysen E.H. et al., 1992, The feasibility of the production of biomass for the Dutch energy sector, Novem, Utrecht, the Netherlands. Heuvel, E.J.M.T. van den, Stassen, H.E.M., 1994, Electricity from biomass, a comparison between combustion and gasification, Novem, Utrecht, the Netherlands. Siemons, R.V. 1993, Electricity generation through co-combustion of straw and grass residues in conventional power plants, BTG, Enschede, the Netherlands.
Acknowledgement This research has been carried out within the Framework of the National Research Programme on Global Air Pollution and Climate Change (NOP/MLK) of the Ministry of Housing, Urban Planning and Environment. The project has been co-financed by the Netherlands Company for energy and the Environment (Novem). The project has been executed in co-operation with the Centrum voor Landbouw en Milieu - which used information on technical and environmental aspects of the conversion routes for their study on the environmental impacts of energy crops cultivation - and the Department of Theoretical Production Ecology, Wageningen Agricultural University.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
Forests and wood consumption
1137
on the carbon balance.
Carbon emission reduction by use of wood products R. Sikkema (SBH a) and G.J. Nabuurs (IBN-DLO b) Institute for Forest and Forest Products (SBH), PO Box 253, NL-6700 AG Wageningen, The Netherlands.
Abstract Until now studies on the greenhouse effect paid much attention to carbon fixation by forests, while the entire CO2 cycle of forests and forest products remained underexposed. Utilization of wood products instead of energy-intensive materials (plastics/steel) and fossil fuels (coal) proves to play an important role as well. The effect of utilization is even greater t h a n that of fixation. In all, additional forests together with the multiple use of trees can contribute substantially to the reduction of CO2 emissions. The contribution can run from 5.3 ton CO2/ha/yr for a mixed forest of oak~eech to 18.9 ton CO2/ha/yr for energy plantations (poplar).
INTRODUCTION
The greenhouse effect is a problem acknowledged worldwide. The increasing concentration of greenhouse gasses in the atmosphere (carbon dioxide, methane, nitrogen oxide and others) may in time lead to an unwanted t e m p e r a t u r e rise. Forests and the use of wood contribute to the fight against the greenhouse effect in three ways: 1. Carbon fixation (during their growth trees convert CO2 into timber) 2. CO2 avoidance through substitution by wood of energy-intensive materials such as plastics, aluminium and steel. Processing timber uses relatively little energy (fossil fuels). After use it can easily be reused, e.g. as particle board. 3. CO2 avoidance by using timber instead of fossil fuels for generating energy. When recycling has become technically or economically impractical, wood may be used for energy purposes. The same is valid for timber from special energy plantations. Only the previously sequestered CO2 will be released when burning the woody material. This makes the use of timber 'CO2 neutral'. In 1989 the CO2-emissions in the Netherlands amounted to 183 million tons. These increase by 3.5 million tons annually. The Dutch government is striving for stabilization with respect to the 1989 emissions level. This corresponds to a reduction of 21 million tons CO2 by 1995.
1138 FIXATION OR AVOIDING? Recently a number of reports are published concerning the carbon sequestering potential of various forest types. These studies primarily examine carbon sequestration in biomass (the tree), soil and timber products. Findings suggest that long rotation forests provide a greater contribution than short rotation forests (Table A, No.l). Besides the average carbon sequestration by trees and in timber products, an important CO2 reduction effect is created through substitution for nonwood products (product substitution) and fossil fuels (fuel substitution). This CO2 avoidance was included in the NOVEMC-report 'Bossen en hout op de koolstofbalans'. Table A Potential contribution to CO2 reduction (in tons CO2/ha) of several forest types Oak/Beech
Spruce
Poplar 15 year
Poplar 5 year
1. CO2 fixation in biomass, humus and products (average) 2. CO2 avoidance through replacement of non-timber materials 3. CO2 avoidance through replacement of fossil fuels
432
394
104
-106
182
784
653
0
966
1289
1560
5788
Total CO2 reduction in 300 years
1580
2467
2317
5682
*)
*)Including processing energy of fertilization (0.44 ton CO2/ha/yr) Substituting wood for plastic, aluminium and steel leads to an important reduction in emissions. This is because the production of wood materials uses far less fossil energy than the mentioned alternatives. The effect is most evident in saw and packing timber products, such as frames, construction timber and pallets. Norway spruce is a forest type that produces relatively many saw logs and packaging timber products (Table A, No.2). The greatest contribution to CO2 reduction, however, results from substituting wood for coal in energy production. This is especially true for energy wood plantations (short rotation poplar), from which all wood produced, such as increment, is used as fuel (Table A, No.3). The effects ofboth product and fuel substitution are repeatable. During cultivation, harvest, use and renewed cultivation, no additional CO2 is released into the
1139 a t m o s p h e r e . T h e a v e r a g e s e q u e s t r a t i o n clearly h a s a once-time effect, b e c a u s e in t i m e t h e CO2 is r e l e a s e d again, e i t h e r t h r o u g h decay or t h r o u g h c o m b u s t i o n ( F i g u r e
A) F i g u r e A Total CO2 r e d u c t i o n effect* of N o r w a y spruce s t a n d h a v i n g a r o t a t i o n of 75 y e a r s (tons CO2/ha) C02-reduction ( tons C02/ha ) 2.500 [-
C02-avoidance through fuel substitution
~
C02-avoidance through material substitution
2.000 -
C02-sequestration
~
inproducts
-~
1.500
C02-sequestration in biomass and litter
-
1.000 -
500
0
0
75
150 Year
225
300
Source: Stichting Bos en Hout, Wageningen, The Netherlands
*) Carbon sequestration in biomass, litter and products is subject to fluctuations during every rotation. An average sequestration, which reaches a constant value of 394 tons CO2/ha after several rotations, was used for calculations in the text. Table B C o n t r i b u t i o n by forest type to 1995" goal for n a t i o n a l CO2-emission reduction Oak/Beech A n n u a l r e d u c t i o n (in tons CO2/ha) R e d u c t i o n p e r 100,000 h a (in 1000 tons CO2) C o n t r i b u t i o n to policy '95
Spruce
P o p l a r 15 year
Poplar 5 year
5
8
8
19
530
820
770
1890
2.5
3.9
3.7
9.0
*)National CO2-emissions were 183 million tons in 1989. Emissions increase 3.5 million tons annually. 1995 goal: Stabilization with respect to 1989 correspondends to a reduction of 21 million tons CO2.
1140
In the study CO2 balances of long rotation forests (oak/beech, 150 years) were compared with those of short rotation forests (poplar, 5 and 15 years), and supplemented with those of Norway Spruce, a wood species ideal for recycling. The contribution to CO2 emissions reduction (see Table B) proves to be substantial: 5 to 19 tonnes COJha/yr. The contribution to the Dutch reduction goal can run from 2.5% for a mixed forest of oak~eech to 9% for energy plantations of poplar.
FOREST TYPES
Calculations were made for four forest types (see Table C). An average tree-species specific increment was assumed. The average CO2 sequestration in the tree itself and in the upper soil layer (litter) of each forest type over a 300 year period was modelled. Sequestration in the stable humus was not considered, because this factor is greatly dependent on soil types and previous use of soil (e.g. agriculture). In the Netherlands the extra fixation by afforestation amounts to a small quantity. Table C Important characteristic figures of considered forest types Oak/Beech Rotation time (yr) N u m b e r of rotations in 300 yr Mean increment (m3/ha/yr) Density air dry (kg/m 3) Amount of carbon in dry m a t t e r weight (%)
150 2 5.4 700 50
Spruce
Poplar 15 year
75 4 11.5 460 50
15 20 15.9 450 50
Poplar 5 year 5 60 29.5 450 50
F I X A T I O N O F CO2 IN T R E E AND S O I L With each rotation, forests 'produce' wood that becomes available during thinning and during the final felling at the end of the cycle. Most of the wood ends up being used outside the forest as industrial wood. Another part of the felled trees remains in the forest as dead wood. This wood ends up in the litter layer and breaks down during decomposition to CO2 and water. Thus the fixed CO2 in wood is gradually released into the atmosphere. No wood is left behind in short rotation poplar forests. The entire above-ground biomass (the t r u n k including the branches, but excluding the leaves) of this forest is destined to serve as fuel for energy production. Thus, hardly any sequestration occurs in the litter layer. Removal of the t r u n k and branches also means the
1141 disappearance of nutrients. Fertilizing compensates for this effect. The processing energy of the fertilizer (0.44 ton CO2/ha/yr) such as lime and K20 is subtracted from the net sequestration. Total sequestration as a result has a negative value.
FIXATION OF CO2 IN P R O D U C T S Depending on the diameter and tree species, harvested forest products are assigned to be used as fuel wood, pulp wood, wood based panels, packing or sawn timber. The research model assumes optimal utilization of the available amount of raw material. This means sawing residues (bark, sawdust, chips, etc.) are assigned to the most durable uses, like chipboard and paper, or when not possible, to fuel. Fuel wood is delivered to the power station in chipped and dried form. Assumed is the possibility to allow the wind to dry the wood in the forest naturally (to a maximum moisture content of 15%). All timber products run through a so-called 'cascade model' (figure B). Where possible this calls for wood to be reused at the end of its technical life span. In this model packaging material and other timber products will be reused for chipboard or paper. Ultimately, when written off or replaced, particle board and waste paper are used for energy generation. Thus all sequestered carbon will again be released to the atmosphere as CO2. Figure B All wood and timberproducts end at the stage of energy-generation ~
/
~
~ .
f~ r~
C02
Ill ( ) llIV wood
\\X.."-"JJf
i
wasted
,---
i-z-
ood
L..2
wasted p a p e r / ~ (board) ~ L ~=~>"~
~
timber
{ ,=_~-, L._
!~
~j.v
JJ II II f J IIII I / /
recycling particle board paper(board)
1142
CO2 A V O I D A N C E T H R O U G H MATERIAL S U B S T I T U T I O N The first reduction of CQ-emission occurs when wood replaces non timber products. Consideration is made of the use of fossile fuels. An important factor is the energy applied during production and transportation of raw materials, semifinished products and finished products. Net energy applied is calculated in the model. This means t h a t r e m n a n t wood is utilized for drying other wood. Some production processes are therefore COJenergy neutral. It is assumed that mineral oil is the only fossil fuel used in this application.
CO2 A V O I D A N C E T H R O U G H F U E L S U B S T I T U T I O N Secondly, the application of energy wood, r e m n a n t wood, waste wood and waste paper for energy purposes is considered. Timber products (in contrast to fossil fuels) are CO2 neutral. Fuel wood does produce carbon dioxide emissions, but the emissions occur within a closed cycle. After all, wood originates and grows by extracting an equal amount of CO2 from the atmosphere. In this study, a comparison was made with coal, one of the most used fossil fuels for generating electricity in the Netherlands. This comparison is the most realistic, because electricity producers have serious plans to use wood in coal-fired power stations in the short term.
CONCLUSIONS Studies on CO2 reduction pay too much attention to sequestering of CO2 in biomass, soil and products. Thus, the total CO2 cycle of forests and forest products remains underexposed. Utilization of wood (multiple use of products and energy-generation) proves to play an important role. The influence on the CO2 balance (avoidance through product substitution and fuel substitution) is even greater than carbon sequestration by trees. Forests and the multiple use of wood (including energy-generation) can contribute substantially to the reduction of C02 emissions. The contribution to the Dutch government's policy (reduction of the annual emission by 21 million tons) can run from 2.5% for a mixed forest of oak and beech, to 9% for short cycle (5 years) poplar. These percentages are based upon an additional 100,000 ha of forest.
Notes aSBH:
bIBN/DLO: cNOVEM:
Stichting Bos en Hout, Wageningen, the Netherlands Instituut voor Bos- en Natuuronderzoek, Wageningen, the Netherlands Nederlandse Onderneming voor Energie en Milieu, Utrecht, the Netherlands
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1143
Potential of controlled anaerobic wastewater treatment in order to reduce the global emissions of methane and carbon dioxide Marjo J. Lexmond and Grietje Zeeman Department of Environmental Technology, Agricultural University of Wageningen P.O. Box 8129, 6700 EV Wageningen, the Netherlands
Abstract An estimation is made of the current global methane and carbon dioxide emissions from wastewater treatment and disposal. Furthermore, the potential of controlled anaerobic treatment to reduce these emissions is investigated. Considered wastewaters are: domestic wastewater and wastewater from the Food & Beverage and Pulp & Paper industry. The current methane emission is estimated to be about 5 Tg/y, and is mainly the result of uncontrolled degradation of untreated wastewater in developing countries. Carbon dioxide emission is estimated to be 1 5 Tg/y, which is mainly due to aerobic wastewater treatment in the developed countries. Anaerobic wastewater treatment, provided a minimization of the percentage methane loss and an optimal reuse of biogas, can significantly reduce the current emissions.
1. INTRODUCTION
Little is known about the quantities of the atmospheric CH4 emission from the treatment, storage, and discharge of domestic and industrial wastewaters. Thorneloe (1993) roughly estimated this to be about 26-40 Tg/y, thereby representing about 8-11% of the total anthropogenic CH4 emissions. However, large uncertainties concerning this estimation are recognized. In most CH4 emission reduction technologies the aim is to avoid uncontrolled anaerobic degradation and to promote aerobic degradation. However, by promoting aerobic treatment the CO2 emission due to fossil fuel consumption will strongly increase. Aerobic treatment, as conventionally applied in most wastewater treatment systems, is rather energy consuming since this process depends on a more or less intensive aeration. A very interesting alternative is formed by anaerobic treatment. Under anaerobic conditions organic material can be completely degraded into CO2, CH4, water, and a small amount of biomass. The produced biogas can be used as a fuel. By doing so, anaerobic degradation has the following advantages over aerobic degradation: 1) Production of a valuable fuel, the use of which can lead to a reduction of the amount of fossil fuel consumed, 2) no energy requirement for aeration, and 3) significantly less sludge production. On the other hand, if anaerobic degradation occurs in an uncontrolled way, CH4 can be emitted to the atmosphere where it can enhance the greenhouse effect.
1144
In our study we made an estimation of the present emissions of CH4 and CO2 from wastewater treatment. Furthermore, we estimated the potential production of CH 4 from anaerobic wastewater treatment and the possible reduction of the CO2 emission due to the use of this CH4. The treatment of the produced sludge is not yet taken into account.
2. ESTIMATION METHODS
In our estimations the following cases are considered: 1) Complete anaerobic treatment. Within this case, three options are regarded: * all CH4 is flared (FLARING) * CH4 is partly used, only for the maintenance of the wastewater treatment plants. The excess is flared (PARTIAL) * CH4 is completely used for energy production (COMPLETE) 2) Complete aerobic treatment (AEROBIC) 3) Current situation (CURRENT) In each case the calculated emissions are the energy related CO2 emission and the CH4 emissions from controlled treatment systems and from uncontrolled degradation in the environment. The world was divided into underdeveloped (UND) and developed (DEV) countries. Relevant data concerning the amount, composition, and degradability of the different types of wastewater were collected from literature and queries. The same method was applied for the information on the treatment systems (efficiency, sludge growth, energy demand, methane emission factors and frequency of operation). All data were assembled in QUATTRO-spreadsheets. Models were developed for the estimations of CH4 and CO2 emissions at different assumptions (Lexmond & Zeeman, 1994). Different percentages of CH4 loss were used: For aerobic systems all CH4 formed (Czepiel et aL, 1993) is emitted to the atmosphere, and for anaerobic systems a fixed percentage of loss (due to leaking, low concentrations, etc.) is assumed. Finally, the influences on the total global warming potential (GWP) due to wastewater treatment were calculated. The most important assumptions made were: 1) The presence of oxidizing compounds, such as nitrate, oxygen, and sulphate, in the wastewater (which can result in lower CH4 production), as well as possibly toxic compounds is ignored. 2) The influence of the temperature is ignored. 3) A certain percentage of CH4 produced within anaerobic treatment systems is lost (020%). 4) The energy requirement (in kWh per unit of organic material removed) of anaerobic treatment systems is about 25-30% of that of aerobic systems. We calculated our CO2 emissions based on an energy requirement of one third of that of the Dutch aerobic systems (CBS, 1992). 5) We used an average carbon dioxide emission factor (in m3/kWh) for the conversion of energy (viz. electricity) into CO2 emissions (Blok, 1994). 6) For the current situation of emissions we had to estimate the extent of the use of different treatment systems. Due to the scarcity of information about this, we had to assume these values. Because the large influence of these values on the results, we give
1145
the assumed values in table 1. Especially the percentage of uncontrolled aerobic and anaerobic degradation of discharged wastewater in underdeveloped countries, is a very important parameter in the model. We assume that a large part of the wastewater is discharged on surface waters. Disposal at sea is assumed not to result in significant CH4 emissions, disposal on large waters and on land is assumed to result in some CH4 emissions, and the disposal on small waters and lagoons will result in significant CH4 emissions. Furthermore, we assume that in underdeveloped countries average temperatures are higher and consequently the percentage of anaerobic degradation for untreated wastewater will be somewhat higher. Table 1. Assumptions concerning the treatment of wastewater and the division in aerobic and anaerobic degradation treated (%)
untreated (%)
total
aerobic
anaerobic
total
aerobic
anaerobic
underdeveloped countries * domestic * industrial
10 50
70 85
30 15
90 50
75 75
25 25
developed countries * domestic * industrial
90 95
90 85
10 15
10 5
80 80
20 20
3. RESULTS
AND
DISCUSSION
In figure 1 the CH4 and CO2 emissions and the resulting GWP for the treatment of the total amount of wastewater at the 5 different conditions are summarized, assuming a total CH4 loss percentage of 10% and a time horizon of 100 years. For the current situation a CH4 emission of 5 Tg/y is estimated from the investigated wastewater streams, which is considerably lower than the estimated 14-20 Tg/y by Thorneloe (1993).
.~
120
120
100"
L~
,_., 14' >,
Z
_o 03 W
"
T 0 N 0 ~
-40
z
__ 1o. 03
N
LU 0 0 0
-eO -8O
9100 ~"
12.
_H_o
60'
FLARING' PARTIAL C;OMPLETEAEROBIC 'CURRENT CH4
~
CO2
~
GWP total
Figure 1. CH4 and CO2 emissions and the resulting GWP (time horizon 100 years and 10% CH4 loss)in different cases
2-
98 0
"~ ._
"60
0
-40 Q.
?: -20
~
TOTAL 'DEV.IND, bEV.DOM~UND.IND.bND.DOM.
m cH.
~
co2
~
GwPtot,s
Figure 2. Current CH 4 and CO 2 emissions and the resulting GWP (time horizon 100 years) from the different sources
1146
Figure 2 presents the CH 4 and CO2 emissions and the resulting GWP from the 100" different sources in the current situation. 50 The amount of treated and untreated oJ wastewater, the applied treatment systO O 0 em, and the conditions at which the untreated wastewater is disposed of ~ -50. determine the current emissions. From figure 2 it can also be seen that the FLARING PARTIALCOMPLETEAEROBIC CURRENT total GWP of wastewater treatment and 0% LOSS ~ 7 ~ 5% LOSS r ~ 10% LOSS disposal in the current situation is mainly ~ - ~ 15% LOSS ~ 20% LOSS determined by CH4 emissions from domestic wastewater in underdeveloped counFigure 3. GWP due to wastewater treattries. The extent of anaerobic digestion of ment at different percentages of CH4 loss untreated wastewater in underdeveloped in the different cases countries is highly affecting the estimated GWP of the current state. If we, for instance, assume that the amount of uncontrolled anaerobic degradation is not 25 % but 50%, the total CH4 emission will increase from 5 to about 11 Tg/y. Unfortunately very little data are available on this subject for most countries. To be able to estimate the present emissions more accurately, more information is required.
Ji n
In figure 3 the effect of the percentage CH4 loss in anaerobic treatment systems is shown. It is clear that the percentage CH4 loss should be minimized. In the current situation this is not so important because only a small percentage of the wastewaters is actually treated anaerobically (see table 1). The current CO2 emission is about 15 Tg/y and is mainly the result of aerobic degradation of wastewater in the developed countries. So, not only from the human health point of view treatment of wastewater should be encouraged. Our results show that, providing a minimization of the CH4 loss and an optimal reuse of the produced CH4, anaerobic treatment should be stimulated in order to reduce the emissions of greenhouse gases.
5. REFERENCES
Blok, K.: 1994, Utrecht University, written communication, January 7. CBS: 1992, 'Water Quality Control, Part b: Purification of Wastewater 1990', Environmental Statistics, Voorburg/Heerlen, the Netherlands (in Dutch). Czepiel, P.M., Crill, P.M., and Harriss, R.C.: 1993, 'Methane Emissions from Municipal Wastewater Treatment Processes', Environmental Science and Technology 27 (12), pp.2472-2477. Lexmond, M.J. and Zeeman, G.: 1994, 'Potential of controlled anaerobic wastewater treatment in order to reduce the global emissions of methane and carbon dioxide', from Ham, J. van et aL (eds): 1994, 'Non-CO 2Greenhouse Gases, pp.411-419, Kluwer Academic Publishers, the Netherlands. Thorneloe, S.A.: 1993, 'Wastewater treatment', from Amstel, A.R. van (Ed.): 1993, 'Methane and Nitrous Oxide', pp. 115-130, National Institute of Public Health and Environmental Protection (RIVM), Bilthoven, the Netherlands.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1147
A S S E S S M E N T R E P O R T ON S U B T H E M E
"MOBILITY A N D M O T O R I S E D T R A N S P O R T I N R E L A T I O N TO S U S T A I N A B L E D E V E L O P M E N T "
C.A.J. Vlek Department of Psychology, University of Groningen Grote Kruisstraat 2/1 9712 TS Groningen The Netherlands
With contributions by: P. Kooreman, J. Rouwendal, L. van Staalduinen
LUW, Agricultural University of Wageningen
A.J. Rooijers, E.M. Steg
RUG, University of Groningen
I.M. de Boer, D. van Kreveld, P.G. Swanborn, G. Tertoolen, E.C.H. Verstraten P. Nijkamp, S.A. Rienstra, J.M. Vleugel
RUU, University of Utrecht VUA, Free University of Amsterdam
1148 Contents Abstract 1.
Introduction
2.
B e h a v i o u r a l d e t e r m i n a n t s of t r a v e l m o d e c h o i c e
3.
E f f e c t s of cost a n d / o r e n v i r o n m e n t a l f e e d b a c k on car u s e
4.
Problem awareness, measures
0
Car user differences susceptibility
willingness in
to c h a n g e , e v a l u a t i o n
environmental
pollution
of policy
and
6.
E c o n o m e t r i c a n a l y s i s of p r i v a t e car u s e by h o u s e h o l d s
7.
S u s t a i n a b l e t r a n s p o r t a n d traffic s y s t e m s for t h e 21st c e n t u r y
8.
General observations, conclusions and suggestions
9.
References
policy
ABSTRACT The seemingly irresistible growth of motorised transport and its environmental effects have led the NRP to also put mobility and transport (M&T) on its agenda. NRP questions are focused on psychosocial factors and mechanisms underlying the popularity of motorised transport, and on technical as well as behavioural measures and strategies to reduce global air pollution stemming from mobility and the use of motor vehicles. In Phase 1 of the programme, five NRP-funded M&T projects have been conducted. Together with one or two related projects, these will be briefly summarised and commented upon. General observations, conclusions and some suggestions will be provided at the end of this paper. 1.
INTRODUCTION
In western societies, the second half of the twentieth century is characterised by rapid expansion of h u m a n mobility and motorised transport. Motor-cars, vans and trucks have become available in large numbers. Many airlines are being operated through a multitude of airplanes. Although transport markets in the wealthiest countries are not even satisfied yet, the opening-up of Central and Eastern Europe and further "economisation" elsewhere in the world imply that motorised transport of persons and goods will further intensify, particularly through the air. "The private car," say Rouwendal, Van Staalduinen and Kooreman in their project
1149 s u m m a r y (this volume), "is an ambiguous symbol of western society in the late twentieth century. On the one hand it reflects its success in providing a high level of material wellbeing to a large majority of the population. On the other hand, it brings out the failure of the same society to solve the environmental problems evoked by its success." Some p e r t i n e n t data - quoted from (Vlek et al., 1992) are as follows. Over six h u n d r e d million cars, vans and trucks populate the world's roads of today: about 70% in the Western industrialized countries, and roughly 10% in Japan, 12% in the less industrialized countries and 8% in Eastern Europe and the Soviet Union (Low, 1990 and Bleviss et al., 1990). Worldwide some 35 million cars are being produced each year; a net total of about 20 million cars is added to the world's fleet of automobiles (Low, 1990, MacKenzie, 1990 and Walsh, 1990). Currently there are more than five million motor vehicles in the Netherlands and the projected figures for the year 2010 lie around 8 million (EZ & VWS, 1987 and VWS, 1989). A third of the Dutch population of 15 million now owns a car; this might well be 50% in 2010; there currently are about 130 motor vehicles per s q u a r e km. For the w e a l t h y w e s t e r n p a r t of G e r m a n y a figure of 70% car ownership (700 cars for 1000 inhabitants) is foreseen for the year 2010. Not only does the n u m b e r of available cars go up steadily, also the size and the engine capacity of the average car are increasing (EZ & VWS, 1987 and Lenz, 1990). If a third of all the world's population in the year 2010 would drive cars, there would be two billion motor vehicles altogether. At 50% and 70% worldwide car ownership in 2010 these total numbers of cars would be three and four billion, respectively, for the projected world population of six billion people. This seems unbelievable. But it does not seem unrealistic to expect a doubling of the n u m b e r of motor vehicles worldwide within the next decade; around 2030 there might well be two billion of such vehicles altogether (Bleviss et al., 1990, MacKenzie et al., 1990 and MacKay, 1990). By now, the negative external effects of all this are well known. Motor vehicles involve large amounts of direct and indirect (embodied) energy consumption and greenhouse gas emissions. Their use periodically contributes to urban smog, and it is dominantly responsible for urban noise and traffic accidents. The enormous volume of motor vehicles in m a n y countries has necessitated the construction of extensive systems of roads, motorways, parking places and traffic regulation. The growth in air t r a n s p o r t is involving more and bigger airports, more and more intensely used air routes and an increasing i n f r a s t r u c t u r e of various service industries, which have their own patterns of energy consumption and greenhouse gas emissions. All this has already significantly reduced the peace and quiet of natural and urban areas, and it is threatening basic environmental qualities. These developments are worrying policy makers and citizens alike. The quality of life in and around cities, the protection of our n a t u r a l environment, and the accessibility of i m p o r t a n t destinations together may well require t h a t the use of motor vehicles be significantly reduced in the years to come, so t h a t , e.g., motorized t r a n s p o r t is limited to serving only the essential needs of society. This would fit into the concept of sustainable development as applied to t~e area of h u m a n mobility and transportation: "(This) involves more than growth. It requires
1150
a change in the content of growth, to make it less material- and energy-intensive and more equitable in its impact" (WCED, 1987). NRP-questions concerning mobility and transport are aimed at obtaining a better view of the social and psychological determinants of travel mode choice, of the physical and technical options for environmentally less harmful mobility and transport, and of the relative effectiveness of various policy strategies for encouraging people to move around and transport their goods in socially and environmentally sustainable manners. In the following, five NRP-funded projects and related research are briefly presented and discussed. In commenting upon the various projects, the author has attempted to draw general conclusions about sustainable-transport policies and to pinpoint further research and policy questions. Table 1 provides an overview of the various research projects to be discussed. Table 1.1 List of projects in NRP subtheme "Mobility and motorised transport in relation to sustainable development" Title
Project leader
Number
Attitudes and behaviours toward the environment
D. van Kreveld
850013
A behaviour analysis of private car use by households
J. Rouwendal
852081
Environmentally relevant differences among L. Hendrickx car user groups and the effectiveness of policy measures
852092
P. Nijkamp Comparative analysis of options for sustainable transport and traffic systems in the 21st century
853102
Non-NRP project Problem awareness & behaviour change
2.
Steg et al.
BEHAVIOURAL DETERMINANTS OF TRAVEL MODE CHOICE
De Boer, Van Kreveld and Swanborn, (NRP project no. 850013) collected questionnaire responses from about 500 regular car drivers in the city of
1151 Hilversum, who first recorded their own transportation behaviour for a period of four days. Respondents filled in a comprehensive questionnaire containing items m e a s u r i n g their personal attitude, perceived social norms and the "opportunity structure" - t h e i r needs and m e a n s - for using a motor-car, p a r t i c u l a r l y for commuting. Questions about attitude salience, knowledge about pros and cons of car use and their personal image of using a car were also asked. It appeared t h a t travelling time is crucial for explaining car use preferences, while commuting distance and income explain a great deal of habitual t r a n s p o r t a t i o n behaviour. Personal attitudes add significantly to this, in contrast to perceived social norms which do not seem to be very influential. In a second survey, the authors studied the relationship between transportation behaviour and several personal variables such as feelings of alienation and powerlessness, self-enhancement with respect to environmental behaviour, and willingness to care for the environment under certain "pulling" or "pushing" policy measures. Completed questionnaires were r e t u r n e d by some 330 i n h a b i t a n t s of the city of Utrecht, half of them "always", the other half "sometimes" going to their work by car; within each group half of the respondents lived further away, the other half closer by t h a n 15 kms from their job location, a distance which (in The N e t h e r l a n d s ) conditions the possibility of considering to use a bicycle for commuting. It was revealed t h a t the habit of commuting by car goes along with a self-enhancing view of one's behaviour toward the environment CI am not polluting very much myself'). Respondents seemed willing to reduce their car use if they would be encouraged to do so on a voluntary basis. No correlations could be observed between personal feelings of alienation and powerlessness on the one hand, and habitual car use and willingness to reduce this, on the other. In view of the modest study results, the authors conclude that ingrained personal habits play a significant role in t r a n s p o r t a t i o n behaviour, which they consider to originate primarily in various external, i.e., sociostructural and organisational factors. C o m m e n t s o n D e B o e r et airs p r o j e c t The design of De Boer et al's two survey studies typically reflects the cognitivistic (as opposed to behaviouristic) psychological view t h a t h u m a n b e h a v i o u r is "reasoned" and emanates from beliefs and evaluations, social norms and perceived i n s t r u m e n t a l i t i e s for a c h i e v i n g p e r s o n a l goals. F o r a c u l t u r a l l y a n d socio-economically strongly embedded behaviour domain such as mobility and transport, this view may well be overpersonalised. That is, given the enormous a v a i l a b i l i t y of m e a n s and facilities, e v e r y d a y needs and desires and the socioeconomic system pressures related to the use of motor-cars, little additional variance in people's habits, preferences and choices with respect to using the car can be explained by variations in personal attitudes, perceived social norms and feelings of alienation and powerlessness. On the contrary, as the authors conclude, the massive use of cars has developed into a social and economic and (therefore) cultural habit for most people. And behaviouristic - not cognitivistic - psychologists well know t h a t deeply ingrained habits are a u t o m a t e d behaviour mechanisms which can only be modified via significant changes in the incentive structure of the physical and social environment which provokes such habits and p e r p e t u a t e s them.
1152 3.
EFFECTS CAR U S E
OF C O S T A N D / O R E N V I R O N M E N T A L
FEEDBACK
ON
In a carefully designed field experiment supervised by Van Kreveld, Tertoolen and V e r s t r a t e n (NRP project no. 850013-2), also in Utrecht, have e v a l u a t e d the effects on travel mode choice and frequency of car use, of providing feedback information on the financial costs of one's own car driving, of providing information about one's "own" environmental effects of car travel, and of providing both kinds of feedback simultaneously. Control groups of respondents received no information whatsoever; one group did (like the three experimental groups) and one did not s y s t e m a t i c a l l y record their own travel behaviour for a certain period of time. Altogether 350 people participated in the experiment; all of them were asked to use t h e i r car as little as possible during the experimental period. This approach, tailored to the individual, contrasts with current government approaches designed to address larger segments of the general public via mass media campaigns and general pricing measures. Results of both a pilot study and the main experiment revealed t h a t subjects did express changes in attitude as a result of feedback information. However, no significant behavioural effects (as recorded in transport diaries) could be observed in r e l a t i o n to feedback about e i t h e r cost and/or e n v i r o n m e n t a l effects. Respondents rated speed, comfort and independence to be the most i m p o r t a n t advantages of using a motor-car and they stated that neither the financial nor the environmental costs of car driving weighed heavily when they were travelling. The a u t h o r s conclude t h a t a t t e m p t s to influence car use b e h a v i o u r a r o u s e psychological resistance, often expressed t h r o u g h dissonance reduction and reactance (counter-behaviour). As a result of dissonance-enhancing information about e n v i r o n m e n t a l effects, intensive car users originally having a positive environmental attitude, may s t a r t thinking t h a t environmental pollution is not t h a t bad after all, and t h a t others bear a greater responsibility for environmental problems t h a n they themselves do. Information about the financial costs of car use especially leads to reactance. Car users seem to experience financial policy measures as a restriction of their individual freedom. They therefore tend to have a dim view of both such measures and the authorities contemplating to implement them. Comments on Tertoolen and Verstraten's project The field experiment by Tertoolen and colleagues is unusual both in its purpose and scope, and in the care with which it has been prepared and conducted. As the authors expected themselves, the intensive and personalised procedure followed should have had greater effects than any generalised mass media campaign urging car users to moderate their behaviour. The fact that, nevertheless, respondents in all t h r e e feedback conditions did not reduce the use of their car, gives little e n c o u r a g e m e n t for public authorities contemplating mass media a t t e m p t s at motorists' behaviour change. This conclusion is in line with the earlier one about the difficulty of changing frequently reinforced, habitual car use without modifying socio-economic system characteristics. Another conclusion worth noting is t h a t little progress on the way toward less rnotorised mobility can be expected from policy m a k e r s having too much respect for individual car users' freedom of travel mode choice. Individual behaviour change of everyone's own free will seems only plausible to the extent t h a t a general positive a t t i t u d e toward p r e s e r v i n g
1153 environmental qualities becomes central in most people's view of the world and their own lives.
4.
PROBLEM AWARENESS, WILLINGNESS E V A L U A T I O N OF POLICY M E A S U R E S
TO
CHANGE,
F u n d e d by the University of Groningen and the Ministry of Housing, Physical Planning and Environmental Affairs, Steg, Vlek and Rooijers (see project s u m m a r y and Steg et al., 1995) have carried out a related project on personal mobility and possible behaviour change. Steg et al. collected home interview responses from 539 r e g u l a r car drivers in and around the cities of A m s t e r d a m , Eindhoven and Groningen. A field-experimental design was followed in which respondents were categorised a priori by region (as indicated above), distance to their city centre (0-7, 7-15 and over 15 kms) and whether or not they had been presented with a brochure providing balanced information on the pros and cons of the massive use of motor-cars in The Netherlands. Respondents kept personal transport diaries for four days prior to the interview. The latter was conducted by a trained interviewer and contained item sets designed to assess respondents' problem awareness, their willingness to reduce car use and their evaluation of 17 different policy measures all regarding the use of private motor-cars. It appeared that, on average, massive car use was perceived to be "a problem" (not a small nor a particularly big one), but significantly less so in and around G r o n i n g e n t h a n in the more densely motorised areas of A m s t e r d a m and Eindhoven. Also, city dwellers are more problem-aware t h a n people living at 7 or more kms away from the city centre. The information brochure did not affect respondents' problem awareness very much, probably because the pros and cons of massive car use were quite well known already through regular media coverage over the years. With regard to behaviour change, less t h a n one third of all r e s p o n d e n t s declared to be willing to reduce their car use. In this respect Eindhoven stood out more positively than either Amsterdam or Groningen, as did city dwellers compared to people living beyond 7 kms from the city centre. Of seventeen actual or contemplated policy measures to reduce the use of private motor-cars, none was rated as significantly effective, Groningers proving to be even more sceptical t h a n inhabitants of the Amsterdam and Eindhoven regions. So-called push measures such as increasing fuel prices or parking rates, were evaluated as h a r d l y acceptable, while pull measures such as improving public t r a n s p o r t or bicycling facilities, were judged to be acceptable. A post hoc categorisation of the 539 respondents into a "low", "middle" and "high" problem awareness group, respectively, yielded the conclusion that the extent of problem awareness correlates significantly with people's willingness to reduce their own car use a n d t h e i r e v a l u a t i o n of the r e l a t i v e e f f e c t i v e n e s s (or r a t h e r : non-ineffectiveness) and acceptability of policy measures aimed at a reduction of p r i v a t e car use. The a u t h o r s conclude t h a t increasing collective problem awareness is a pre-requisite for getting motorists to change their behaviour for the common interest. Apart from this, clear policy goals and consistent government strategies are essential.
1154 C o m m e n t s o n S t e g et al.'s project This r e s e a r c h d e m o n s t r a t e s again t h a t using a private motor-car is highly attractive and i m p o r t a n t to most people. Nevertheless people t h r o u g h o u t The Netherlands perceive the collective disadvantages of car use to be a problem about which the government should do something. This problem awareness, which varies both between and within diverse regions of the c o u n t r y - as well as between sex and age groups of respondents - significantly covers with people's (limited) willingness to change t h e i r t r a n s p o r t behaviour and with t h e i r (sceptical) evaluation of various policy measures. According to the authors, increasing public problem a w a r e n e s s and providing feasible t r a n s p o r t a l t e r n a t i v e s would be essential ingredients of any serious government policy designed to reduce the intensity of car traffic. Official Dutch policy goals are: to save energy and diminish environmental pollution, to enhance the accessibility of important destinations and to keep cities worth while to live in or visit. Given t h a t policy m a k e r s currently know fairly well why and how to effectively influence individual people's t r a n s p o r t a t i o n b e h a v i o u r , the e s s e n t i a l q u e s t i o n now is: u n d e r w h a t environmental, social and/or economic conditions could the car driving population (top politicians included) be "moved" to vote for and accept restrictive policy m e a s u r e s aimed at reducing the use of motor-cars in order to preserve and promote vital collective goods and qualities?
5.
C A R U S E R D I F F E R E N C E S IN E N V I R O N M E N T A L P O L L U T I O N A N D POLICY SUSCEPTIBILITY
Building upon previous research by Rooijers (1990) on speed differences among different types of car drivers, Cavalini, Hendrickx and Rooijers (NRP project no. 852092) at the University of Groningen have analyzed and described systematic differences among distinct car user groups, with respect to their usual speed, fuel consumption and environmental effects. The latter were taken as depending upon various types of decisions or behaviour. Five categories were distinguished: car purchase, choice of car type, car use, timing and routing of trips, and driving behaviour. The idea here is that net environmental effects may be directly as well as indirectly related to current behaviour, so t h a t emission reductions could be a t t e m p t e d at various points in the above sequence. Through questionnaire, interview and field observation research, the authors were able to classify car drivers in terms of car possession, car characteristics, n u m b e r of kilometres driven annually, average occupation rate, type of (either or not congested) roads used, and driving speed and style. Three main car user groups could be identified, viz. private drivers, commuters and business drivers. The latter were subdivided into business drivers using their own car and those having a company car at their disposal. The first field study addressed four issues: the usefulness of the distinction among the four car user groups, the size of these groups in the total Dutch population, environmental parameter differences among the four groups, and the combination of user group and environmental p a r a m e t e r type so as to identify significant possibilities for emission reductions. Raw data were collected via mailed questionnaires returned by an apparently representative sample of 1150 respondents.
1155 It a p p e a r e d possible and useful to s e g m e n t the driver population in t e r m s of car u s e r group, as defined above. Forty-two percent of all car users drives for private r e a s o n s a n d r e p r e s e n t s 23% of all car kilometres driven. C o m m u t e r s r e p r e s e n t 37% of all drivers a n d cover 40% of all car kilometres. B u s i n e s s drivers w i t h private car and business drivers using a company car m a k e up 13% and 8% of the driver population, respectively, and t h e y each r e p r e s e n t 18% of all k i l o m e t r e s driven by car. W i t h respect to all p a r a m e t e r s studied t h e r e a p p e a r to be large differences among car user groups. Business drivers using a company car t u r n out to be the most energy-consuming and environment-polluting group, while private drivers are doing the (relatively) least h a r m to the environment. By focusing on p a r t i c u l a r c o m b i n a t i o n s of u s e r group and type of b e h a v i o u r or decision (see above), it s e e m s possible to achieve CO2 emission reductions of about 5%, per c o m b i n a t i o n , so t h a t the overall CO2 emission reduction p o t e n t i a l should be substantial. The second field study by Cavalini et al. was aimed at clarifying the potential of various policy strategies w h e n applied to different car user groups in relation to different t y p e s of b e h a v i o u r or decision (see above). Six categories of policy i n s t r u m e n t s were distinguished for inducing changes in environmental p a r a m e t e r s of car use. These are, respectively, physical a l t e r n a t i v e s a n d r e a r r a n g e m e n t s , regulation and enforcement, financial-economic stimulation, providing information a n d c o m m u n i c a t i o n , social m o d e l l i n g a n d s u p p o r t , a n d i n s t i t u t i o n a l a n d organisational change. The m a i n research questions were: 1. To w h a t extent are the different car user groups able to change various types of behaviour or decision regarding car use? 2. How sensitive are the various behaviours and decisions of the four user groups to the application of different policy instruments? 3. To significantly reduce CO2 emissions, which type of policy i n s t r u m e n t m a y best be applied to which group and in connection with which type of behaviour or decision concerning car use? To collect r a w data, first some 4000 roadside observations were m a d e of passing cars, whilst car type, license n u m b e r and speed were recorded. Then, car owners were identified via the national car registration s y s t e m and t h e y were invited by telephone to participate in the study. W h e n 50 confirmations for each user group (see above) h a d been obtained, personal interviews were conducted. The l a t t e r were focused on possible changes in car use and on the respondent's evaluation of various policy measures. Some m a i n results are the following. All car drivers indicated t h a t they would have less personal control over possible changes - like giving up their car, changing type of car or reducing car kilometrage to the extent t h a t these would have far-reaching consequences for t h e i r mobility a n d daily life. The m a j o r i t y of drivers could t a k e a s m a l l e r car, decrease t h e i r n u m b e r of "private" kilometres or drive more slowly and quietly. P r i v a t e drivers and c o m m u t e r s are generally more inclined to change their behaviour t h a n either group of b u s i n e s s drivers. For achieving reductions in h a r m f u l car emissions, communicative and educational m e a s u r e s seem to be less effective t h a n legal and financial m e a s u r e s as well as i n f r a s t r u c t u r a l a n d o r g a n i s a t i o n a l m e a s u r e s . Financial m e a s u r e s would be more effective in changing the behaviour of private drivers a n d c o m m u t e r s t h a n t h e y would be affecting the b e h a v i o u r of business
1156
drivers. Also, financial measures are most effective in making people to give up their car and to make them drive fewer "private" or "commuting" kilometres. Finally, the reasons given for not changing behaviour or decisions about car use, under any of the presented policy measures, reveal similarities with the reasons provided in relation to personal control, as mentioned above. C o m m e n t s o n C a v a l i n i et al.'s p r o j e c t The methodological approach ventured in this project proved to be successful in identifying distinctly different car user groups and demonstrating their differential energy consumption and environmental pollution, as well as their differing sensitivities to various policy measures. Moreover, a comparison of the two studies yields the conclusion that those who consume the most energy and produce the most harmful emissions, also are the least sensitive to current policy measures. Such information provides a useful basis for designing and targeting specific measures and strategies for reducing the harmful effects of mobility and motorised transport. The major policy conclusion from this research is t h a t m a n a g e m e n t policies for motorised mobility should be designed to fit the personal motives, habits and mobility needs of possible target groups of car users. Cavalini et al.'s respondents clearly signalled t h a t reducing the environmental effects of their mobility would also reduce the (perceived) "control" over their daily lives, if it would involve giving up their car or driving significantly less t h a n usual. Thus, f u n d a m e n t a l thought must be given to policies designed to compensate for this feared "loss of control" when the use of private motor vehicles is to be diminished for reasons of collective importance. This means that target groups should be investigated a priori, to determine the potential impact of contemplated policy measures on their daily lives and the extent of "cooperative power" that they could, or would, have. "Cooperative power" (i.e. sufficient- r e m a i n i n g - personal control) u n d e r changing conditions for mobility and t r a n s p o r t could be enhanced by decreasing people's structural needs and desires for motorised mobility, by helping people to better organize their daily or weekly travel patterns, and/or by providing a l t e r n a t i v e t r a n s p o r t modes known to be socially and e n v i r o n m e n t a l l y less harmful. 6.
ECONOMETRIC HOUSEHOLDS
ANALYSIS
OF
PRIVATE
CAR
USE
BY
At the Agricultural University of Wageningen, Rouwendal, Van Staalduinen and K o o r e m a n (NRP project no. 852081) have systematically looked into the dependence of car ownership and car use upon variations over time in the price of cars and of car fuel. Their aim was to find out the extent to which car users are actually sensitive to the "price mechanism", and in what respect (e.g., type of car, kind of trip or driving style) such sensitivity would be manifested in their behaviour. Knowing this is important for applying financial policy measures to reduce the volume of car traffic and/or the purchase and efficient use of smaller cars. On the basis of econometric analyses of some 3759 observations about 1379 motorists from the "private car panel" of the Central Bureau of Statistics (a continuous, time-variable sample of respondents), the authors have estimated various model parameters. This research is still in progress. Some preliminary results and conclusions are as follows.
1157 Automobile drivers do change their short-term "demand" for car kilometres in response to changes in fuel prices, especially when they do not receive an employer c o m p e n s a t i o n for automobile costs. Demand-price elasticities (behavioural sensitivities) appeared to be different for different age groups, male versus female car users, and for summer versus winter periods. Older, male and "winter" drivers appear to be less price-responsive t h a n younger, female or "summer" drivers. It also appeared t h a t higher income, greater commuting distance, being a company director, holiday driving and getting a car-use allowance, constitute circumstances under which car driving is intensified and less price-sensitive. The authors state that, although short-term demand-price elasticities are significant, the demand for cars and car kilometres has steadily grown over the past 15 years. This seems due to powerful other factors t h a n the variable costs of car driving. For instance, an increased general income level has made car ownership and car use relatively cheaper, women's increased participation in the labour market has enhanced their share of the car driving population, and backward developments in public transport have "forced" many people to equip themselves with private motor-cars. Comments
on Rouwendal
et al.'s project
In the past, demand-price elasticities for car ownership and car use have hardly been studied systematically in sufficient detail to u n d e r s t a n d which type of behaviour change occurs in response to certain price changes. Rouwendal and colleagues have demonstrated that fuel price changes affect particular kinds of car use (e.g., social-recreational trips) more than others, and that certain categories of people (e.g., middle-aged men) are less price-sensitive than others. In this respect Rouwendal et al.'s project goes nicely along with the work on distinguishing car user groups, conducted by Cavalini et al. (see above). Naturally, a fuel price change means different things to different people, to the extent that their "substitution behaviour" - what they can and will do instead of their higher-priced current car use - turns out to be different. Looking more closely (and perhaps prospectively i n s t e a d of retrospectively as m a n y econometrists do) into subjects' likely substitution behaviour may reveal their reasons for manifesting different patterns of reactions. Another problem in demand-price elasticity research lies in the distinction between short-term and long-term behavioural adaptations to price changes. A sudden change of price may yield a short-term behaviour change all right, but what happens on the longer term is often revealing of more fundamental driving forces u n d e r l y i n g a given category of behaviours, as the a u t h o r s themselves acknowledge. Factors discussed by the Dutch Physical P l a n n i n g Service (Allsop, 1993), for example, are the increased physical separation of living and working locations in the 1970s and 1980s, the growth in the labour market for women, the larger n u m b e r of one- and two-persons households emerging from "individualisation", and growing i m m i g r a t i o n from abroad. Such long-term developments and trends raise the question of the significance for car ownership and car travel of the variable costs of car driving as a factor by itself. With reference to Tertoolen and Verstraten's project (see above) we might say t h a t altering car ownership and car use via the "price mechanism" would require fairly drastic financial policy measures. The project by Steg et al. (see above) has revealed t h a t this would be unacceptable for most people, unless certain key conditions for feasible behaviour change would be fulfilled.
1158 7.
SUSTAINABLE TRANSPORT AND TRAFFIC SYSTEMS FOR THE 21ST C E N T U R Y
The mobility and transport research results so far may leave the reader with reserved feelings about the possibilities to control the growth of motorised traffic and reduce harmful emissions. One long-term policy strategy, therefore, could be to go more deeply into the structural determinants of mobility and attempt to adapt or to change social systems so as to reduce the inherent demand for mobility. An other policy strategy could be to acknowledge the need and the desire for greater mobility of persons and goods, and to design sophisticated transportation modes and systems whose environmental effects stay within ecological limits. Working towards the latter policy strategy, economists Nijkamp, Rienstra and Vleugel (NRP project no. 853102) at the Free University of Amsterdam have conducted a "comparative analysis of options for sustainable transport and traffic systems in the 21st century". Other research associates are at the Technical University of Delft, the University of Groningen, the Energy Research Centre in Petten, and University College London. The project has been conducted in two parts, one focused on exploring separate transport modes, the second directed at the construction and evaluation of diverse national transport scenarios. In their report of Part 1, after an analysis of various problems of transport in relation to environmental quality, the authors systematically describe current trends in transport demand and supply. They point at the quest for higher transport quality and discuss various factors underlying the growth in mobility, such as rising incomes, spatial spreading of homes and work places, population growth and individualisation (leading to more and smaller households). They also discuss several types of market failure yielding undesirable "externalities", and they indicate failures in government policy to manage societal demand for mobility. On the basis of interviews and workshops with international experts, the authors t h e n p r e s e n t a list of technical, economic, spatial, i n s t i t u t i o n a l and socio-psychological factors that would be important for future developments in t r a n s p o r t a t i o n . Subsequently, selected (new) t r a n s p o r t a t i o n modes are systematically evaluated against these various factors. The authors' analysis goes along the advanced automobile, the high-speed train, low-speed Maglev (magnetic levitation) systems for urban transport, the electronically guided vehicle, subterranean transport and liquid hydrogen aircraft. Conclusions from Part 1 are that there are many possibilities for future reductions of greenhouse gas emissions from transportation. Three general strategies for emission-reduction are: cleaner transport technology, changing the modal split between polluting and (relatively) clean transport modes, and reducing the demand for mobility. Focusing on the first and second strategies mentioned, the authors conclude that the most likely technologies seem to be improvements of the private car, the high-speed train and the use of telematics for increasing transport efficiency. It seems unlikely that Maglev high- and low-speed transport will be introduced at any large scale. The third general strategy, reducing mobility demand, has received less attention in this project.
1159 In the second stage of the project, two reference scenarios, a "regulatory" and a "market" scenario, have been constructed which reflect extreme profiles in a "spider model" comprising eight major dimensions. The latter are categorised in pairs as spatial, institutional, economic and social/psychological, respectively. Against this background an "expected" and a "desired" scenario were constructed on the basis of questionnaire responses from various Dutch t r a n s p o r t experts. Subsequently, these two scenarios were discussed by an international group of experts. In the "expected" scenario which comes close to the "market" scenario just mentioned, it is assumed that current trends will continue, and that therefore the private motor-car will remain the dominant transport mode. Its freedom of use would, however, be restricted by regulatory measures such as fuel price increases and higher parking rates. In the "desired" scenario more stringent policy measures to discourage the use of private cars would be introduced, together with policies aimed at developing higher-quality means of collective transport. The authors conclude t h a t the expected scenario is, of course, more plausible, but that it could not be called "sustainable" unless an environmentally much less harmful mode of private t r a n s p o r t would be developed t h a n seems technically feasible for quite some time. The "desired" scenario, on the other hand, would involve considerable social behaviour change, together with stricter government policies and fairly big investments in rather different infrastructure than the sort which is underlying the private car system. C o m m e n t s on N i j k a m p et al.'s project Although the scenario study has not yet been definitely reported, it m a y be concluded t h a t this project has been a useful exercise on the possibilities of sustainable mobility and transport. With a focus on The Netherlands and with i n p u t s from experts from E u r o p e a n countries facing s i m i l a r t r a n s p o r t developments, a multidisciplinary picture has been sketched of the main technical options and several distinct societal scenarios for mobility and transport in the early 21st century. It has become clear - once again, we might say - t h a t western industrial society is strongly tuned toward the free-market system and toward meeting the demand for individual transport by motor-car for any person at any time and in almost any place. Reducing the social, economic and environmental costs of this t r a n s p o r t system, which m a n y people find no longer sustainable, would seem to require principal decisions by government policy makers. It would also require far-reaching social a t t i t u d e and behaviour changes, which are conditioned by problem perception and the availability of behaviour alternatives in relation to mobility and t r a n s p o r t (see e.g., the projects by Tertoolen and Verstraten and by Steg et al., above). Methodologically, Nijkamp et al.'s project relies heavily on expert assessments and opinions (e.g., about the "desired" scenario). It was not designed to analyze the fundamental societal factors and individual motives underlying the increased demand for motorised transport. Nor was the project aimed at collecting attitude and behavioural data from diverse sectors and groups in society, so that an assessment could have been made of the social and economic viability of particular transport options in relation to specific behaviour changes. Finally, it appeared that the experts themselves, too, may be of different opinion when it comes to designing and recommending "sustainable" transport scenarios for the near future. One point of discussion, for example, was to what extent government could at all come to grips with the collective problems
1160
of private car use, given the degree of organization and the economic capabilities of the car industry. 8.
G E N E R A L OBSERVATIONS, CONCLUSIONS AND S U G G E S T I O N S
NRP-funded research on mobility and transport so far has primarily addressed the massive use of private cars. Research on freight transport on the roads has not been undertaken in phase 1 of the NRP, nor have any studies been started on air transportation. The private-car mobility research carried out has, on the one hand, been focused on the individual car user, to assess his or her motives, attitudes, behaviour and sensitivity to policy measures and/or feedback information. On the other hand, long-term physical and technical alternatives to the private motor-car have been explored and evaluated, under the premise that the demand for mobility is there and could not (or should not) be influenced. Both lines of investigation seem to underrate the importance of social and economic system factors underlying the demand for mobility and the need for motor vehicles. System factors strongly influence and shape individual motives and preferences to the extent t h a t individuals may be brought in a forced position to acquire and use a motor-car. Also, the fate of physical and technical alternatives for the private motor-car seems strongly dependent upon the nature of economic and social activities and upon the way in which social and economic interaction is organised. In this respect, there seems to be room for more fundamental studies into non-transport factors residing in various social and economic domains, whereby mobility and the need for motorised transport may be generated, or may be reduced. At the Dutch national level, the NRP research on mobility and t r a n s p o r t complements the research initiated and funded by the Advisory Service for Traffic and Transportation of the Ministry of Traffic and Waterways. This Service's programme of "anticipating research" for 1995 (VWS, 1995), for example, lists such topics as "green" transport scenarios, effects of changes in government administration, possibilities and consequences of improved transport informatics, electronic vehicle guidance, fast waterborne transport and improvement of government communication strategies. This government-directed research still largely evolves from the Second Structural Scheme on Traffic and Transportation (1988-1990) and it is designed to yield results potentially supporting current government policy. It is therefore still very much in line with Nijkamp et al.'s "market scenario" (see above) which reflects a societal as well as large-scale individual preference for privately organised mobility and transportation. Internationally, NRP research in phase 1 links up with the concerns and intentions expressed in various programmes, conferences and workshops. For example, in 1992 the European Commission published a "Green paper on the Impact of Transport on the Environment" (CEC, 1992); also, its Directorate-General XII funds several projects on "the integration of environmental concerns into transport policy" EC, 1994), perhaps a modest beginning but something that could fly-wheel itself up. The Human Dimensions of Global Environmental Change Programme (HDP, 1994) of the International Social Science Council has not yet identified a separate line of research on mobility and transport, although it has indicated "industrial transformation and energy use" to be a key area for research. More importantly, ISIRT, the International Scientific Initiatives on Road Traffic group
1161 since 1988, has conducted three international "round tables" on mobility and t ra ns po r t in relation to environmental problems. A s u m m a r y report entitled "Agenda for safe access to a stable environment" was prepared by Allsop (1993). His final conclusion - on behalf on the ISIRT steering committee - reads: "Radical changes in road traffic and its uses, w h e t h e r the changes be technical, institutional, behavioural, regulatory, financial or fiscal, are likely to be uncomfortable at least for some people in the short term. But they should be brought about because the alternative is to continue to put up with the many and severe adverse effects of road traffic in its present form, and thus fail to use it to the best advantage" (Allsop, 1993, p. 6). Another pertinent meeting, organised in September 1992 by the European Ministers of Transport (ECMT) Conference, was held at the OECD headquarters in Paris. On page 237 of the conference proceedings, titled "Transport policy and global warming" (ECMT, 1993), a summary and conclusions section is phrased in ten points. The first four of these "messages for ministers" read as follows. "(1) C u r r e n t trends are clearly inconsistent with the Rio (UNCED 2; Ch.V.) aspirations. (2) New technology can improve matters, but there is no complete technological fix immediately available. (3) Nevertheless it is possible to reduce transport's contribution to global warming; what is necessary is the political will to introduce the necessary measures. (4) Existing technology is not being put to best advantage because of the freedom of transport users to adapt their behaviour to convert potential environmental amelioration into more transport service. (5) Controlling this adaptive behaviour should begin immediately by seriously addressing the issues of reducing the specific power, performance and speed of vehicles" (ECMT, 1993, p. 237). In view of this assessment of NRPI-research and the wider conclusions mentioned above, mobility and transport remain on the NRP agenda. For phase 2 of the NRP, covering the period of 1995 through 2001, societal causes and solutions of potential climate problems will again also be sought in cleaning up and/or reducing mobility and t r a n s p o r t by motor vehicles. Relevant themes are: economic and social-cultural determinants of mobility and transport, options for limiting the need for mobility and transport, and sustainable mobility and transport policies and strategies for society. Thus more attention is being asked for mobility-generating developments and trends in society, such as, e.g., i n t e r n a t i o n a l tourism, development of the labour market and upscaling of the educational system. Also, a focus is being laid on social implementation and acceptance problems in relation to sustainable-transport options and strategies. Finally, mobility and transport are being linked to consumption patterns and lifestyles, in an attempt to clarify the potential effects on the quality of producers' and consumers' life of low-mobility activity patterns and collective- transport scenarios for society as a whole. 9.
REFERENCES
Allsop, R.E., 1993. "Agenda for safe access to a stable environment; issues for decision m a k e r s as identified at ISIRT Round Tables 1989-1991". International Association of Traffic and Safety Sciences (IATSS), Tokyo, 29 pp.
1162 Bleviss, D.L. and Walzer, P., 1990. Energy for motor vehicles. Scientific American, 263 (September): 55-61. CEC, 1992. Green paper on the Impact of Transport on the Environment. COM (92)46. Commission for the European Community, Brussels. EC, 1994. Project summaries; research on economic and societal aspects of environmental issues. European Commission, Directorate-General XII, Brussels. ECMT, 1993. Transport policy and global warming. Proceedings of a European Ministers of Transport seminar in Paris. ECMT Series 75 93 10 1, OECD Publications. EZ&VWS: Dutch Ministries of Environmental Affairs and Traffic and Waterways, 1987. Verkeer en Milieu (Traffic and Environment) Policy Document for Second Chamber of Parliament. Ministry of VROM, Division of Public Information, The Hague, 66 pp. EZ, Ministry of Environmental Affairs, 1991. Ruimtelijke verkenningen. (Spatial explorations). Yearbook of the National Physical Planning Service of The Netherlands (Rijksplanologische Dienst). Ministerie van VROM, The Hague. HDP, Human Dimensions of Global Environmental Change Programme, 1994. Work Plan 1994-1995. International Social Science Council, Paris. Lenz, K.-H., 1990. Motorization and trends in road traffic. In V~ig- och Trafikinstitutet LinkSping: Proceedings of "Road safety and traffic environment in Europe". September 1990, VTI-Report no. 362a. Lowe, M.D., 1990. Alternatives to the automobile: transport for livable cities. Paper 98, Worldwatch Institute, Washington D.C.,49 pp. MacKay, M., 1990. Towards a unified traffic science. IATSS Research: Journal of the International Association of Traffic and Safety Sciences 14: 19-26. MacKenzie,J.J. and Walsh, M.P., 1990. Driving forces; motor vehicle trends and their implications for global warming, energy strategies, and transportation planning. World Resources Institute, Washington D.C. Rooijers, A.J., 1990. Drivers' attitudes and beliefs towards speed limits and speeding on Dutch motorways. In V~ig- och Trafikinstitutet LinkSping: Proceedings of "Road safety and traffic environment in Europe", September 1990. VTI-Report no. 363a. Steg, L., Vlek, C.A.J. and Rooijers, T., 1995, in press. Gedragsverandering ter vermindering van het autogebruik: probleembesef, verminderingsbereidheid en beoordeling van beleidsmaatregelen. (Behaviour change for diminishing car use: problem awareness, willingness to change and evaluation of policy measures). In: F. Siero, E., van Schie, D. Daamen and A. Pruyn (Red.). Sociale psychologie en haar toepassingen. Deel IX, Eburon, Delft. Vlek, C.A.J. and Michon, J.A., 1992. Why we should and how we could reduce the use of motor vehicles in the near future. IATSS Research: Journal of the International Association of Traffic and Safety Sciences, 15: 82-93. VWS: Ministry of Traffic and Waterways: Advisory Service for Traffic and Transportation, 1995. Anticiperend onderzoek; projecten 1995. (Anticipating research; projects-1995). Directoraat-Generaal Rijkswaterstaat, AVV, Rotterdam. VWS: Dutch Ministry of Traffic and Waterways, 1989. Tweede Structuurschema Verkeer en Vervoer. Deel A: Beleidsvoornemen. (Second Structural Scheme on Traffic and Transportation: Part A: Policy Intentions). SDU, The Hague.
1163 Walsh, M.P., 1990. Global trends in motor vehicle use and emissions. Annual Review of Energy 15: 217-243. WCED: World C o m m i s s i o n on E n v i r o n m e n t and D e v e l o p m e n t , 1987. Brundtland-Report: Our common future. Oxford University Press, Oxford/New York.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1165
Discussion on the NRP assessment reports "Mobility and motorised transport which fit in sustaineble development" and "Culture, consumption and lifestyles" M.M. Berk Introduction In this session prof. Vlek gave an overview of the main results of the 11 projects within the subthemes "Consumption and lifestyles" and "Mobility and Transport" of NRP-phase I. Due to time restrictions their was limited time for discussion. During Vleks' presentation some correcting / additional remarks were made. These have been used for finalizing the subtheme reports. Here, only the main discussion items are reported as the main results of the projects are covered fully in the subtheme reports. General remarks Vlek started with presenting a few general notions of relevance for both studying Lifestyles and Consumption pattern and Mobility and Transport.
One general notion is the formula commonly used in environmental studies, according to which environmental impacts can be described as the product of Population(P), Economy (E) and Technology(T). Traditionally, policy efforts to restrict environmental impacts have focused on changing technologies used. Given the magnitude of environmental problems the question has been raised whether not also the level of economic activity- both production and consumption - needs to be reduced. Addressing the Population factor is being viewed as difficult for either social or political reasons. From the perspective of studies on consumption and lifestyles, in addition to the above mentioned factors, it is important to look at how culture (C) and institutions(I) interact with population, economy and technology. They influence population development, consumption needs and technological development. In studying consumption it is useful to take the so-called production-consumption cycle into account, in which, on the one hand, consumers in turn for wages deliver labour to the production process and at the same time influence production by buying products and services. Both at the production and consumption side there are environmental impacts due to use of land, materials and energy and the production of waste and pollution. Production and consumption are interrelated. Therefore, one should study the environmental effects of consumption and production in an integrated wayand not just focus on the consumption side only.
1166
In t h e context of the issue of Global E n v i r o n m e n t a l C h a n g e the s t u d y of t h e h u m a n dimensions is relevant for studying: the socio-economic impacts and risks of environmental change; t h e causal m e c h a n i s m s or causes b e h i n d h u m a n induced e n v i r o n m e n t a l changes an the h u m a n responses to these changes. Global E n v i r o n m e n t a l Changes generate large collective risks in situations which can be defined as social d i l e m m a s . In t h e s e s i t u a t i o n s it is not r a t i o n a l for individual actors to do w h a t m a y be rational to serve the interest of all. Needed for handling collective risks are: a clear description of the risks and its sources; a w a r e n e s s of the risks with major actors; an weighting of the risks against the cost and benefits of action a p e r c e i v e d n e e d for c h a n g e l e a d i n g to an a s s e s s m e n t of b e h a v i o r a l alternatives setting of risk limits and translated into behavioral objectives m e a s u r e s and i n s t r u m e n t s to change behaviour policy implementation and evaluation feedback on collective risk reduction and total benefits Lifestyles and Consumption patterns On t h e basis of the r e s e a r c h on Lifestyles and C o n s u m p t i o n p a t t e r n s w i t h i n NRP-I Vlek presented the following conclusions:
* * * * *
*
*
there is still a need for an operational definition of lifestyle, not j u s t in t e r m s of behaviour but also of attributes and values; r e s e a r c h into the possibilities for s u s t a i n a b l e b e h a v i o u r is m u c h m o r e p r o m i s i n g if n a t u r a l science, technological and socio-economic expertise is linked in joint research efforts and research planning; it m a y be wise for policy m a k e r s to a s s o c i a t e s t a t u s a n d p r e s t i g e to s u s t a i n a b l e c o n s u m p t i o n , b u t it s h o u l d be c o m b i n e d by o t h e r policy instruments; s u s t a i n a b l e lifestyles should be p r o m o t e d by positive, r e w a r d i n g a n d attractive policy strategies, emphasizing the "desirable" not the "undesirable"; in r e s e a r c h on s u s t a i n a b l e lifestyles the meso and micro-level should be a d d r e s s e d condordantely, because of the links b e t w e e n c o n s u m p t i o n a n d production; it seems t h a t m u c h is possible technically if there would be enough problem a w a r e n e s s . W i t h o u t sufficient problem a w a r e n e s s people will not accept the policy m e a s u r e s a n d b e h a v i o r a l c h a n g e s n e e d e d to m a k e use of t h e s e technological opportunities; problem a w a r e n e s s depends on "visible" e n v i r o n m e n t a l effects of household m e t a b o l i s m (giving feedback on e n v i r o n m e n t a l behaviour). Because of the absence of directly visible effects of the impacts of and behavioral response to m a n y global environmental changes this is a very i m p o r t a n t issue.
Discussion * The question was raised if it was important to distinguish different lifestyles as
1167 these were viewed to be only minor deviations of the overall abundant "western lifestyle'. It was replied by Aarts that social research gives insight into the social processes that constitute different lifestyles and how these may change or be changed. Moll of the University of Groningen responded that in practise lifestyles can be rather easily defined on the basis of two main dimension:, socio-economic status and level of education. Ester emphasized that from the (theoretical) sociological perspective lifestyles are much more complex than as defined in empirical research. They embrace also different values. The importance of distinguishing different lifestyles is t h a t they are an integrating explanatory variable for different sets of environmentally relevant behaviour. Vlek remarked that from a policy perspective it is only relevant to make distinctions between different lifestyles as far as these have implications for the use of policy strategies: when people with different lifestyles have to be approached in different ways. From a policy perspective the concept of lifestyles should not only be used in a descriptive way but also in a prescriptive way- exploring desirable sustainable lifestyles. It was noted that the distinction of different lifestyles can only be of any practical relevance if it is actually possible to clearly define and distinguish these. For that reason it was suggested to make only broad policy relevant distinctions. In response to the conclusions presented by Vlek van Kreveld of Utrecht University remarked that, as confirmed by the outcome of the NRP-research on private car use, environmental awareness itself is not sufficient to make people change their behaviour. To make people change their behaviour it is e.g. also necessary that people are offered real behavioral alternatives. Vlek acknowledged that with respect to global environmental change problem awareness is, indeed, only but the first prerequisite for changing h u m a n behaviour. However, he liked to stress the importance of problem awareness in response to presentations on technological options often ignoring the question how to bring about the development, implementation and social acceptance of these technologies.
Mobility and Transport By introduction Vlek stated that, presently, transport is responsible for about 20% of the global emissions of greenhouse gases and constitutes a fast growing source. Not only transport markets in the developed countries are not yet satisfied, also much growth is to be expected in eastern Europe and in industrializing developing countries, like in Asia. Besides its contribution to the emissions of greenhouse gases, transport is causing many other problems like air pollution, noise and space demands. An important question is whether all these problems can be solved by better technologies or that the demand for mobility and transport itself should be addressed. Both directions were researched within NRP-I.
1168 After giving an overview of the main result of the research projects Vlek presented the following policy oriented conclusion: * *
* *
* * *
motor car use is a very suitable and attractive mode of transport for citizens, companies and governments alike; car use is very individualistic, but it is socially and culturally regarded as an obvious thing to do. Restricting car use would provoke strong resistance. One of the important reasons for that resistance is that there is a lack of (socially) acceptable alternatives; neither personal nor environmental costs are dominant factors in determining car ownership and car use; private car drivers, commuters and business drivers differ systematically in their environmental impacts and their sensitivity to various policy measures. So for policy makers it is very useful to distinguish between these different target groups; F u t u r e options for more sustainable transport systems seem r a t h e r modest: improved car technology, high speed t r a i n s and more intensive use of telematics (e.g. regulating traffic). Changing the demand for mobility should get more research attention. There is little research into the underlying socio-economic system characteristics t h a t provoke the (growing) demand for mobility; The environmental impacts of, demand for and governmental policies related to air traffic have been neglected in the NRP sofar and need more attention.
Discussion * It was noted t h a t the results of the research seemed r a t h e r obvious and not very surprising. Were these outcomes not known already from previous research? According to Vlek this is not the case. The research did however confirm the hypothesis that it is very difficult to get people out of their car. Van Kreveld added t h a t often results of social sciences seem obvious in retrospect, but were not known or commonly accepted when the research started. By illustration Vlek r e m a r k e d t h a t m a n y measures t a k e n by the Dutch M i n i s t r y of T r a n s p o r t are based on a s s u m p t i o n s which differ from the conclusion presented here, like approaching car drivers indifferently, focusing policy on the effects of the price mechanism and persuading the public to drive less by public information campaigns without paying due a t t e n t i o n to structural causes of the growing demand for mobility and transport.
With respect to the reported lack of change in car driving behaviour it was r e m a r k e d that in historical perspective a profound decrease in the growth of car use can be detected. The general idea that things remain the same is not well founded. Vlek noted that, although such changes can be noticed, these are not enough in the light of the change needed to arrive at a sustainable transport system. While car use may not growth that fast any more, the same does not apply for e.g. air travel which is ever growing faster.
1169 Also with respect to car drivers the question was raised how r e l e v a n t knowledge of the differences in sensitivity of different car driver groups for policy measures is given the potential contribution of these kind of measures to the reduction of greenhouse gases. Cavalini from the University of Groningen r e m a r k e d t h a t their research clearly stated the relevance of differentiating between different driver groups. As an example he mentioned that while, relatively, business drivers pollute most, measures influencing private car use make a bigger contribution to limiting emissions as private car drivers are responsible for half the total milage driven, including many relatively polluting short trips for which car use is less necessary. It was asked w h e t h e r the research also took foreign efforts to cope with growing problems of t r a n s p o r t a t i o n and mobility into account, like the experiences with restricting car use in Singapore. Their experiences m a y indicate other ways to approach car use. Furthermore, it was noted t h a t incremental solutions to mobility and transportation problems may postpone real solutions, thereby making problems only worse in the end. Making the present system crash might open the road for more fundamental solutions. With respect to the research by Midden et. al. at the University of Eindhoven on the effectiveness of emotion-oriented c o m m u n i c a t i o n compared to cognitive-oriented communication strategies it was r e m a r k e d t h a t in the United States the use of fear arousal in the communication on global environmental issues eventually backlashes as the stated effects did not occur a n d were contradicted. So, fear which does not hold w o r k s contra-productive.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1173
Private car mobility. Problem awareness, willingness to change, and policy evaluation: a national interview study among Dutch car users Linda
Steg a, Charles Vlek a and Ton Rooijers b
Department of Psychology, University of Groningen, Grote Kruisstraat 2/1, 9712 TS Groningen, The Netherlands. a
b Traffic Research Centre, University of Groningen, P.O. Box 69, 9750 AB Groningen, The Netherlands.
Abstract This paper reports on a field study, based on personal interviews with 539 car users. Problem awareness appears to be an important condition for any attempts to make people voluntarily reduce car use. Problem awareness also is an prerequisite for the acceptance of policy measures aimed at reducing car use. Problem awareness is higher the more people are confronted with the problems of car use. The provision of information in a brochure did not influence respondents' problem awareness.
1. THEORETICAL BACKGROUND The social dilemma paradigm is a useful model to understand and to manage problems in which numerous individual benefits are running up against cumulative collective costs and risks, such as from car use [1]. In large scale social dilemmas it is attractive to continue to act in one's own interest. Individual contributions to collective costs and risks, as well as to their reduction, seem negligible. Moreover, most people are pessimistic about the cooperation of others. So, individuals tend not to feel responsible for collective problems. This makes individual contributions to collective solutions unlikely. Members of the public as well as policy makers will only contribute to resolving largescale social dilemmas if two conditions are fulfilled. First, people must perceive motorised traffic as a source of serious societal problems. This requires a clear and unambiguous description of the various negative consequences. Second, people have to balance the collective disadvantages against the personal advantages of car use, and they must be convinced that the problems need to be solved. Thus, p r o b l e m a w a r e n e s s is an important condition for any attempts to make people voluntarily reduce car use [2-3]. For this study, we hypothesised that the higher people's problem awareness, the more they are willing to reduce car use, and the more favourably they evaluate relevant policy measures. Furthermore, we expected that the more people are confronted with problems of car use (in densely populated areas, in city centres, or by reading information about these problems), the higher their problem awareness would be, the more they would be willing to reduce their car use, and the more favourably they would evaluate poficy measures.
1174
2. METHOD We studied problem awareness, possible behaviour change, and the evaluation of policy measures for reducing car use through in-depth interviews with 539 car users selected as living within 7 kilometres, between 7 and 15, and further than 15 kilometres away from the centre of Amsterdam, Eindhoven, and Groningen, three cities having rather different mobility profiles. The collective problems of car use are most visible in the Amsterdam region, because of the high traffic volume, while in the Groningen region traffic volume is low and a lot of problems are not visible yet. The Eindhoven region takes a middle position. Within each geographic condition, a few days before the interview two thirds of the respondents received systematically different amounts of prior information in a brochure about the most important societal problems of the massive use of cars and possible solutions for them. One third of the respondents received information about the present problem situation. Another third received information about the present and future problem situation. The remaining respondents received no information. Twenty people were interviewed in each (19 in one) research condition. Structured interviews were conducted at respondents' homes by trained interviewers. The questionnaire contained, amongst other things, several items measuring the key concepts of 'problem awareness', 'willingness to reduce car use', and 'evaluation of policy measures'. Prior to the interview, respondents were given a travel diary in which they recorded all movements on the Friday, Saturday, Sunday and Monday prior to the interview. Interviewers checked to what extent the respondents had actually studied the brochure.
3. RESULTS We will concentrate on subjects' problem awareness, their willingness to reduce car use, and their evaluation of policy measures. Only differences which are statistically significant at p < .05 will be reported. On average, the respondents perceive various collective consequences of car use as 'a problem'. The scores on 'problem awareness' could vary from -10 ('not a problem at all') to +10 ('a very big problem'). The mean score (M) was 3.1. As hypothesised, on average people living in Groningen (M = 2.5) do have a lower score on 'problem awareness' than people living in the Eindhoven (M = 3.5) and Amsterdam (M = 3.2) region. People living in or near the city centre (M = 3.6) do have a higher score on 'problem awareness' in comparison to people living outside the city centre (M = 2.8). No significant differences were found between the information conditions. Only 30% of the respondents appear to be actually willing to reduce their car use. People living in the Eindhoven region (38%) have a greater willingness to reduce their car use in comparison to respondents living in the regions of Amsterdam (25%) and Groningen (24%). Among the 'distance' groups, also, there is a significant difference in 'willingness to reduce car use'. Respondents living within 7 kilometres of the city centre (34%) are more willing to reduce car use than people living between 7 and 15 kilometres of the city centre (24%). No significant differences were found between the information conditions. Respondents were asked to evaluate the effectiveness and acceptability of 'push' and 'pull' measures. Push measures are directed at making car use less attractive, such as
1175 through higher fuel prices. Pull measures are aimed at improving the alternatives for car use, such as improving the quality of public transport. Scores could range from -10 ('not at all effective' or 'not at all acceptable') to +10 ('very effective' or 'very acceptable'). On average, people evaluate neither push measures (M = -4.3) nor pull measures (M = -3.7) as effective. Respondents evaluate pull measures as 'acceptable' (M = 4.4). Push measures were evaluated as 'not acceptable, nor unacceptable' (M = -0.1). Again, people living in the (quieter) Groningen region evaluate push measures as well as pull measures as less effective and less acceptable in comparison to the respondents living in the more populated regions of Eindhoven and Amsterdam (see table 1). There are also significant differences in the evaluation of policy measures among the distance groups. This only pertains to the evaluation of the acceptability of pull measures: respondents living within 7 kilometres of a city centre evaluate pull measures more favourably (M = 5.1), especially in comparison to respondents living between 7 and 15 kilometres of the city centre (M = 3.9).
Table 1 Evaluation of push measures and pull measures per region I
effectivity 'push' effectivity 'pull' acceptability 'push' acceptability 'pull'
Amsterdam -4.3 a -3.6 a -0.1 4.8 a
Eindhoven -3.9 a -3.0 b 0.4 a 4.5 a
Groningen -4.9 b -4.3 c -0.5 b
3.7 b
1 Means with unequal superscripts differ at p < 0.05.
The 539 respondents were divided into three equal groups, on the basis of their scores on the concept of 'problem awareness'. Table 2 shows that respondents with a higher 'problem awareness' are more willing to reduce their car use in comparison to people with a lower problem awareness. Moreover, respondents with a higher score on 'problem awareness' evaluate policy measures more favourably.
Table 2 Willingness to change and evaluation of push measures and pull measures for groups differing in problem awareness (all percentages and means differ at p < 0.05) problem awareness
low
middle
high
willing to reduce
18%
29%
39%
effectivity 'push' effectivity 'pull' acceptability 'push' acceptability 'pull'
-5.7 -4.7 - 1.7 3.5
-4.0 -3.6 -0.2 4.2
-3.2 -2.7 1.7 5.4
1176 4. D I S C U S S I O N On average people perceive car use as 'a problem'. However, most people are not willing to reduce car use. Respondents evaluate current Dutch push measures as well as pull measures as rather ineffective. They judge pull measures to be acceptable, while push measures are evaluated as 'acceptable nor unacceptable'. So, on average people believe that policy measures aimed at reducing car use are acceptable, but not very effective (or they think the measures are acceptable because they are not very effective). There are several explanations for the perceived ineffectiveness of policy measures. First, problem awareness may not be as high as to make people actually do something about it. Second, problem awareness by itself is not a sufficient condition for reducing car use. People also must have the impression that the collective problems c a n be solved, that their own contribution is useful, and that others will also contribute to the solution of the problems [3]. As hypothesised, there is a positive relationship between problem awareness, willingness to reduce car use, and the evaluation of policy measures. Heightening problem awareness, therefore, seems a useful strategy, provided there are sufficient feasible alternatives available to reduce car use. Our expectation that the more people are confronted with the problems of car use, the higher would be their problem awareness, is only party confirmed. Respondents living within 7 kilometres of a city centre do indeed have a higher problem awareness, are more willing to reduce car use, and evaluate policy measures more favourably. Moreover, respondents living in the quieter Groningen region have a lower score on problem awareness, are less willing to reduce car use, and evaluate policy measures less favourably. However, there were no differences between respondents who did or did not receive prior information. Maybe the information, which can regularly be read in the newspaper, was not new to the respondents. It is also possible that people perceive the information as unreliable, and deny or downplay the information as valid. Collective costs and risks of car use are difficult to control. Effective solution strategies require, besides problem awareness, clear policy objectives, and a forceful and consistent government policy, based on several different policy instruments.
5. REFERENCES
1 Ch. Vlek, L. Hendrickx and L. Steg, A social dilemmas analysis of motorised-transport problems and six general strategies for social behaviour change, in: ECMT: Transport policy and global warming, Paris: European Conference of Ministers of Transport (ECMT), OECD Publication Service, 1993, pp. 209-225. 2 D.M. Messick and M.B. Brewer, Solving social dilemma's: a review, in: L.Wheeler and O. Shever (eds.), Review of Personality and Social Psychology, 4 (1983), Beverly Hills, Calif.: Sage. 3 B. Klandermans, Persuasive communication: measures to overcome real-life social dilemmas, in: W.B.G. Liebrand, D.M. Messick and H.A.M. Wilke (eds.), Social dilemmas: theoretical issues and research findings, Oxford: Pergamon, 1992.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1177
A Behavioral Analysis of Private Car Use by Households Jan Rouwendal, Lanie van Staalduinen en Peter Kooreman Department of Household and Consumer Studies, Wageningen Agricultural University, P.O. Box 8060, 6700 DA Wageningen, The Netherlands. Abstract The relevance of econometric studies of ownership and use of private cars for environmental issues is sketched and some recent results are reviewed.
1 INTRODUCTION The private car has been in the centre of concern about environmental issues ever since it became a mass consumption good. Exhaustion of oil resources, air pollution, acid rain and the greenhouse effect are all related to ownership and use of private cars by a large number of households. The private car is an ambiguous symbol of western society in the late twentieth century. On the one hand it reflects its success in providing a high level of material well-being to the large majority of the population. On the other hand it brings out the failure of the same society to solve the environmental problems evoked by its successes. The widespread anxiousness about the future sustainability of western societies has given rise to much concern about the continued growth of automobile ownership and use. However, the popularity of the private car has increased over the years and should be expected to do so in the future. This apparent paradox has been interpreted by social scientists as a social dilemma: everybody knows how to improve society, but apparently nobody is willing to take the necessary actions himself. 2 AN E C O N O M I C A P P R O A C H Social scientists seem to agree on the proposition that human behavior with respect to automobile ownership and use can be fruitfully considered as being driven by the desire to reach some purposes, for instance being able to reach the work location fast and comfortably. The driver is then assumed to act deliberately on the basis of a consideration of the benefits and costs associated with the various alternatives available to him. This plausible vision may be termed 'rational,' although that term should be interpreted in a limited sense. It does not suppose, for instance, that de actor takes into account the effects of his behavior on other people in the same way as the effects that concern himself. Nor is it required that future consequences should be given the same weight as immediate consequences. It is precisely because of these characteristics of human decision making that widespread concern about environmental problems coexists with increasing popularity of the private car. Welfare economics has developed a recipe for this problem. The basic trick is to introduce a tax to be paid by the actor that has the same effect on his decision making
1178 as a proper calculation of all present and future effects of his behavior on society as a whole would have. The environmental costs that are neglected by the decision maker because they do not concern him, or do not concern him immediately, are in this way brought to his attention and the balance between individual and societal rationality is restored. With respect to the contribution of the private car to the greenhouse effect, this prescription would require an estimate of the costs of adding one additional unit of carbondioxide to the environment and charging them to each driver. Since the emission of greenhouse gases is closely related to the amount of fuel used, a fuel tax would be the appropriate policy instrument. Since such a tax has been introduced in all western countries, the required increase of this tax should not be expected to give rise to implementation problems. Moreover, the widespread concern for environmental problems may be expected to offer the necessary political support for such a measure. The real problem for following this strategy is the determination of the marginal environmental costs of the emission of greenhouse gases. The present state of knowledge only allows one to think about the effects of such gasses in general and imprecise terms. Nevertheless, it may be said that enough is known to justify a policy directed at a reduction of the further emission of such gasses. In order to see what policy efforts are required in order to reach the desired effect, it is necessary, among other things, to investigate the determinants of automobile ownership and use by households. 3 A DECISION CHAIN It is useful to distinguish a number of steps in decision making with respect to the private car: - The most elementary decision concerns ownership.. Should one buy one (or more) private car, or make use of public transport? If a car is purchased, what ~ should it be? A large number of brands and makes are available. Fuel type, cylinder volume and weight are important characteristics for the environmental aspects of automobile use. - Which use will be made of the car? How many kilometers should be driven for homework interactions, for business purposes, for social purposes and on holidays? - The driving ~ influences fuel use significantly. Car users who like to accelerate fast and drive at high speeds cause more environmental damage than others. - The decisions taken at all steps determine the emission of greenhouse, gases by automobiles. The various steps in this decision chain show a certain hierarchy in the sense that the ones made earlier are more basic. For instance, the type of car is only relevant if a car will be bought. It should be kept in mind, however, that the various steps should be considered as interrelated. For instance, decisions with respect to car ownership are made on the basis of, among other things, preferences with respect to car use for different purposes. One important consequence of the many facets of decision making with respect to automobile ownership and use is that the introduction of, for instance, a higher fuel price may be expected to have a number of different effects that may operate at different time scales and possibly in different directions. For instance, an increase in the fuel tax may have the immediate effect of a decrease in the number of kilometers driven for social purposes. When a new car is bought, fuel efficient makes will be bought more often. This results in lower costs per kilometer, which mitigates the immediate effect of -
1179 the higher. Moreover, the higher costs of mobility may induce people to consider the possibility of living in the neighbourhood of his work relation more intensely than he would have done otherwise. This may result in a shorter commute, which strengthens the original effect of the tax measure. 4 A REVIEW OF RESEARCH RESULTS Research on the various aspects of automobile ownership and use dates back to the early history of econometrics. The early studies concentrated on time series of numbers of automobiles owned or produced. Gradually research shifted towards the micro economic aspects of car ownership and use. In the eighties econometric models that enabled a researcher to study the ownership and use of one or more cars by individual households became available (see Mannering and Winston [1985]. For the Netherlands this type of model was introduced by De Jong [1990] who found substantially larger effects of changes in fuel prices on both the number of kilometers driven and the decision to own a car than were suggested by earlier studies: a change in variable costs of 1% would in the short run give rise to .65 % less kilometers driven, while in the long run the effect would increase to 1.11%. Since fuel costs are the major component of variable costs, this suggests that drivers are sensitive to changes in these prices. De Jong used cross section data and did not take into account differences between car types. His results were therefore not based on observed reactions to changes or differences in fuel costs per kilometer driven. In the period from 1980 to 1993 fuel prices changed significantly only in 1986 and 1991. Use of time series would therefore offer only limited opportunities for measuring the effects of fuel prices in a more direct way. De Jong's work provided a good starting point for the work that is currently being done at the Department of Household and Consumer Studies of Wageningen University. The aim of this research is to provide a more detailed picture of household behavior in the various parts of the decision tree. Although this work is still in progress, we can mention some preliminary results here. One part of the project is a more careful investigation of drivers reaction to the decrease in fuel prices occurring at the beginning of 1986. Monthly data concerning the year 1986 were analyzed by Van Staalduinen and Rouwendal [1994] who found gasoline price elasticities for the number of kilometers driven that are of the same order of magnitude as those found by De Jong. The short run sensitivity of the demand for automobile kilometers for changes in fuel prices is mainly due to the social motive, as commuting and business travel are usually harder to change. The demand for commuting kilometers is determined to a considerable extent by the choice of the residential and work locations. In Rouwendal and Rietveld [1994] a search model that explains these choices is developed and estimated. Empirical application a this model enables one to estimate the required compensation for an additional kilometer of commuting for various types of workers. Part of the required compensation consist of fuel prices. Preliminary estimates of the model confirm the existence of such a trade off, suggesting that the long run effects of higher fuel prices on commuting distance should not be ignored. Another study concerned the determinant of the driving style. Rouwendal [1994] regressed the fuel use per kilometer, as indicated by the main drivers, on characteristics of the car, characteristics of the driver and on monthly data about fuel prices and average temperature. Driver characteristics were included because of their presumed effect on driving style. A significant coefficient for the gasoline price indicates that
1180 drivers respond to changes in fuel prices by driving in a more or less fuel efficient way. Continuation of this line of research may be expected to contribute to a more detailed and coherent picture of the determinants of car ownership and the behavioral reactions to changes in fuel prices. 5 OUTLOOK In the recent past increases in the fuel tax have not been able to slow down the increasing popularity of the private car substantially. There are several reasons that explain this state of affairs. It may be reasonably expected that variable cost per kilometer is the crucial variable influencing driver's behavior. These costs are influenced by the price of crude oil and by the fuel efficiency of the motor, as well as by the fuel tax. Over the past 15 years there has been no significant overall increase in the variable costs, despite several increases in the fuel tax. Moreover, income growth and increased participation of women in the labor force have contributed significantly to the rising number of cars. It must be expected that these forces will still be effective in the near future. It is therefore important to consider the effect of fuel taxes within a broad framework that incorporates the major economic and social trends. In this way the study of the determinants of car ownership and use may be expected to contribute to a better insight into the possibilities and limitations of reducing the emission of greenhouse gases. 6 REFERENCES
de Jong, G.C. [1990] An Indirect Utility Model of Car Ownership and Private Car Use, European Economic Review, 34, 971-985. Goodwin, P.B. [1992] A Review of New Demand Elasticities with Special Reference to Short and Long Run Effects of Price Changes, Journal of Transport Economics and Policy, 26, 155-169. Mannering, F. and C. Winston [1995] A Dynamic Empirical Analysis of Household Vehicle Ownership and Utilization, Rand Journal of Economics, 16, 215-236. Oum, T.H., W.G. Waters and J.-S. Yong [1992] Concepts of Price Elasticities of Transport Demand and recent Empirical Estimates, Journal of Transport Economics and Policy, 26, 139-154. Rouwendal, J. [1994] An economic Analysis of Fuel Use per Kilometer by Private Cars, research paper, Wageningen Agricultural University. Rouwendal, J. and P. Rietveld [1994] A Structural Model of Commuting Distances and Spatial Job Search, research paper, Wageningen Agricultural University. Rouwendal, J. and L. van Staalduinen [1994] A Panel-Data Analysis of Short-Term Changes in Travel Demand, research paper, Wageningen Agricultural University.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1181
Differences among car user groups regarding CO2 emissions and sensitivity to policy measures P.M. CavalinP, L. Hendrickx ~ and A.J. Rooijers b aCenter for Energy and Environmental Studies, University of Groningen, P.O. Box 72, 9700 AB Groningen, The Netherlands bTraffic Research Centre, University of Groningen, P.O. Box 69, 9750 AB Haren, The Netherlands
Abstract Two field studies revealed large differences among various subgroups in the population of car drivers. Private drivers, commuters, and business drivers differed strongly with respect to current decisions and behaviour which affect CO2 emissions, and with respect to their sensitivity to various policy instruments. Several promising policy targets were identified" combinations of user groups and behaviours where substantial CO2 reduction may be achieved. The sensitivity of different car user groups to various policy measures showed whether and how desired behavioral changes may be realised.
1. I N T R O D U C T I O N Motorised traffic contributes to an important extent (app. 15%) to CO2 emissions. For the major part (app. 10%) passenger cars are responsible for these emissions. The number of cars has almost doubled between 1970 and 1992 from 2.8 to 5.3 million. The annual total number of kilometres driven by these cars has increased from 36 to 86 billion in this period. Decreasing (the negative effects of) car mobility has become a societal priority. Policy measures with regard to passenger car mobility, may be categorized according to: (1) the behaviour or decision they aim to alter, (2) the (sub)group of drivers they are directed at, and (3) the policy instrument used to achieve the desired change. 1
Type
of decisions
or behaviour
Decisions - car purchase - type of car - car use - timing and routing - driving behaviour
--- > ---> ---> --- > --->
System parameters affected car possession car park characteristics number of kilometres, occupation rate congestion, road use driving speed, driving style
1182
2 Type of car users - private drivers - commuters - business drivers with private car - business drivers with non-private car (company or leased cars) Type of policy instruments (as distinguished by Vlek and Michon, 1992) - physical alternatives and (re)arrangements - regulations and enforcement strategies - financial and economic strategies - information and communication strategies - social support strategies - institutional and organisational strategies Differences in current behaviours of the various user groups implicate that measures aimed at altering behaviour may have different potential effects upon various user groups. Moreover, users groups may differ in their sensitivity to various policy measures. For designing policies which will effectively reduce energy use and adverse emissions by passenger cars two steps, taken in two studies, are necessary.
Study 1" who and what? This study focuses on the various types of car users and types of decision or behaviour, distinguished above. The aim of this study is to identify combinations of behaviour type and car user type, where at least in principle substantial CO2 emission reductions are possible. Study 2: whether and how? Whether these CO2 reductions may actually be achieved depends on two factors: personal control and sensitivity to policy instruments. Personal control refers to the extent to which drivers are able to change their decisions and behaviour. Due to eg. infrastructural or organisational factors, the degree of personal control may differ among user groups. The extent to which drivers are sensitive to various types of policy instruments may also differ among the user groups. The aim of study 2 is to determine differences among user groups with regard to personal control over their decisions and behaviour, and their sensitivity to various policy instruments, in order to determine - for each of the 'behaviour and user group' combinations identified in study 1 whether behavioral changes are possible and how these changes may best be achieved.
2. METHOD In study 1 a large and representative sample of Dutch car drivers ( n = 1150) filled in a postal questionnaire in which information was collected about various types of decisions and behaviour, as indicated above. The sample was drawn from car registration files. In study 2 interviews were held with app. 50 representatives of each user group, in which the degree of personal control over decisions and behaviours and the respondent's sensitivity to different types of policy instruments were assessed. The sample was drawn by observing cars on motorways.
1183 3. R E S U L T S The results of study 1 indicate that user groups differ strongly with respect to almost every CO2 relevant decision or behaviour. For most behavioral parameters, the betweengroup differences have a similar pattern. Private drivers (42% of the total population of car drivers) score less negatively on all CO2 relevant parameters (except car age). They drive relatively light, old, and fuel-efficient cars. Private drivers have the lowest kilometrage. They report to drive more slowly and in a more energy-efficient way than the other groups. Hence, their driving style results in fewer CO2 emissions per kilometre driven~ Business drivers with non-private car (8% of the population) score most negatively on all parameters (except car age) and contribute disproportionably to CO2 emissions: on average, they have the highest kilometrage, they drive heavy cars with a low fuel efficiency, and their speed choice and driving style result in relatively high CO2 emissions per kilometre driven. The commuters (37%) and business drivers with private car (13%) fall in between with regard to all CO2 relevant parameters. On the basis of these results several promising combinations of user group and type of decision or behaviour with regard to the reduction of CO2 emissions, were identified (for details see Cavalini, Hendrickx, and Rooijers; 1993). The results of study 2 reveal that the amount of personal control drivers (perceive to) have varies for the different decisions and behaviours studied. Many respondents report that, even if they would be willing to do so, they would not be able to give up their car, to drive fewer commuting kilometres, and/or to drive fewer business kilometres. With regard to other behaviours (take a smaller car with the next purchase, decrease the number of private kilometres, drive on other times, drive more slowly, and drive more responsibly) the respondents report to have a considerable amount of control. In general, the drivers view that they have less personal control over decisions which, if altered, would have more far reaching consequences, and vice versa. Moreover, it was found that user groups differ in the amount of control they have over specific decisions or behaviours. For instance, business drivers think they have less freedom to give up their car than private drivers and commuters. Compared to the other groups, fewer private drivers could decrease the number of private kilometres. Only a minority of commuters could drive on other times, whereas both groups of business drivers may have more freedom to do so. Subjects were asked why it was not possible to change a certain type of decision or behaviour. Three sorts of reasons were given: 1) reasons referring to organisational circumstances or conditions, 2) from the subjects' point of view their behaviour is already 'optimal', and 3) anti-public transport reasons. With regard to the drivers' sensitivity to the various policy instruments, the results demonstrate the following. In general, private drivers and commuters are more inclined to change their behaviour than both groups of business drivers. All user groups are more willing to change behaviour which does not alter their mobility life style (eg. smaller car, drive slower). Drastic behavioral changes are less likely to occur (eg. give up car, drive less). On average, the drivers are less sensitive to communicative measures (education and information) than to legal, financial, infrastructural, and organisational measures.
1184 As expected, user groups differed with regard to their sensitivity to different policy instruments. For instance, private drivers and commuters are to a larger extent than the business drivers, willing to give up their car, to buy a smaller car, to drive fewer commuting kilometres and to drive on other times~ The former groups are more sensitive to infrastructural, organisational, legal, and financial measures than business drivers. The difference is largest for financial measures. Especially the business drivers with nonprivate car are not sensitive to this kind of measures. They are more sensitive to hffrastructural~ organisational, and legal measures~ Of all groups, private drivers are most sensitive to information measures~ Financial measures appear to be the best type of policy measures to induce people to give up their car and to decrease the number of private or commuting kilometreso Infrastructural and organisational measures may best be used to affect decisions about type of car~ drive on other times, and drive more slowly. Legal measures could best be utilised to diminish business kilometres, to induce drivers to drive on other times and to drive more slowly~ Information measures may best be used to change routing behaviour of drivers.
4. C O N C L U S I O N This research project has indicated that segmenting the total car users population into several user groups, and distinguishing various types of decisions or behaviour regarding car mobility, may enable policy makers to design more effective policy programs which intend to decrease (the negative effects of) car mobility. The first study revealed large differences among user groups with regard to current COz relevant behaviours. Combinations of behaviours and user groups, where in principle substantial COz reductiolm are possible, were identified. The second study of the project demonstrated that the user groups distinguished differ in the extent to which they are able to change various types of decision or behaviour. This study also showed which policy measures would be effective to induce the desired changes in different user groups.
5. REFERENCES Cavalini, P~ Hendrickx, Lo and Rooijers, A.J. (1993). Differences among car user groups regarding CO 2 emissions. IVEM-OR Noo 65, Rijksuniversiteit Groningen. Cavalini, P.M., Hendrickx, L. and Rooijers, A.J. (1995). Differences among car user groups regarding their sensitivity to policy measures. Rijksuniversiteit Groningen. In press. Vlek, C.A.J. and Michon, J.A. (1992). Why we should and how we could reduce the use of motorvehicles in tile near future. Journal of the hzternational Association of Traffic and Safety Sciences, 15, 2, 82-93.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1185
Choosing a means of transportation: Two inquiries into situational and personal determinants of moving behaviour
I.M. de Boer, D. van Kreveld, P.G. Swanborn Vakgroep Sociale en Organisatiepsychologie, University of Utrecht, Heidelberglaan 1, 3584 CS Utrecht, The Netherlands
Two survey research projects on situational and personal determinants of moving behaviour have been carried out. The Hilversum survey defines the problem of choosing a means of transport in relation to private mobility behaviour and the models and situational variables used for explaining this choice. We discuss the models which are normally applied to behaviourial science investigations in order to explain behaviour. Using various traditions we discuss attitude, social environment and the opportunity structure as separate complexes. These three are the determining factors in the basis model in which behaviour is the factor to be defined. Attitude is determined, at least according to the much used Ajzen and Fishbein model, by a person' s expectations for the consequences of his/her behaviour and the value he/she places on those consequences. The (pressure from the) social environment can, in turn, also be explained. We also discuss the discrepancy between reasoned action - on which most explanatory models are based - and habitual behaviour. Finally, we describe a few other psychological factors which can influence human behaviour, such as the salience and the strength of the attitude. More than 600 car drivers over a 4-day period monitored their movements in a journal and filled in a comprehensive questionnaire. The response was 78 resp. 79%. By means of this journal the most significant dependent variable was determined: relative car use (this is the number of movements made by car divided by the total number of movements). This same variable is also measured specifically for commuter travel and shopping trips. We determined, by means of the questionnaire, attitude(s), pressure applied by the social environment and aspects of the opportunity structure. Habitual behaviour was measured by means of a question about the means of transport used most within a household in general and for seven specific categories of mobility behaviour. The investigators tested different elements other than those specified in the Ajzen-Fishbein model, while measures of salience, knowledge questions and questions relating to the image of car use were included in the questionnaire as well. Firstly we examine the aspects of the opportunity structure, such as distance, time difference, income, stage of family development, because these are situated the most to the left in this model of causal variables. Furthermore, they are the most concrete and recognizable. The variance in car use explained by these variables was generally narrow, less than 10%. The variable playing the most important role here is always the time taken to travel by
1186 car versus the time taken to travel by an alternative means of transportation. Distance does not matter (provided that the time difference is taken into account) nor does income level. An exception to this is individual habitual-behaviour; here the structural variables accounted for 21% of the variance. For all dependent variables, the explained variance increases considerably when the attitude is added to the model (thus habitual behaviour reached 33% of the explained variance). Subsequently adding perceived pressure from the social environment hardly leads to a rise in the explained variance. We mention in the context of the Ajzen-Fishbein model, which specifies that the product sum of other people's opinions and the tendency to be influenced by them (a common procedure but one statistically not accountable), can as well be replaced by the generalized tendency to be influenced by what others think and say. The results proved disappointing with respect to the environmental salience: when attitude is already included in the model, no extra explained variance is recorded. The measure of environmental knowledge did not prove to be a useful instrument. The measures of misperception yielded clearly no data in the predicted direction. The measures of image in relation to cars, bicycles and public transport revealed interesting results. We also tried to assess the strength of attitudes, in the way described by Fazio et al., in order to increase the prediction of mobility behaviour. The strength of attitudes was individually determined under laboratory conditions. This was performed on only 55 participants in the Hilversum survey because of its time-consuming nature. It turned out to be impossible to determine their strength due mainly to too many errors made by the respondents, far more in fact than reported by Fazio et al. As we found this unsatisfactory we repeated the experiment on a sample of 34 respondents with an academic background. Here too, so many errors were made that it proved impossible to determine the strength of attitudes. The Utrecht survey was focused on the relationship between personality variables and mobility behaviour, especially commuting behaviour. The personality variables were: (1) if the person feels alienated or comfortable in society, (2)if the person feels powerless, especially concerning environmental problems, (3) if the person distorts reality, especially the self-enhancing illusion of one's own behaviour towards the environment (this means he/she perceives his/her own behaviour as being more environment-friendly than that of others), and (4) if the person expects he/she will behave in a more environmentally acceptable manner as the result of either voluntary or enforced changes. For the elements of mobility behaviour to be predicted we used the following: (1) individual habits when choosing a means of transport, (2) the intention to reduce car use, and (3) the tendency towards cooperative behaviour. The variables used were measured by means of a written questionnaire. The respondents were given questions and statements, partly taken from other research and partly developed for this particular study. All respondents were inhabitants of the city of Utrecht and were chosen at random from the telephone directory. Only people were chosen who had access to a private car and worked outside the home. The sample was compiled so that half of the respondents always went to work by car and the other half went to work, sometimes by car, and sometimes by an alternative means of transport. In both categories half of the respondents lived 15 kilometres at the most from work, which meant that it was possible for
1187 them to use a bike as an alternative means of transport. The other half lived more than 15 kilometres from work which meant that the most important alternative means of travel for them was public transport. The questionnaire was completed by 329 respondents (the response was 82%). The scale for measuring alienation proved to be only fairly reliable. The scale for measuring powerlessness was sufficiently reliable. Self-enhancing illusions were found to be present and reasonably reliable to measure. Voluntary or coerced behaviour change did not prove to be a reliable variable to measure. The analysis was, therefore, carried out with a number of separate items of the scale. The habit of commuting by car and the intention to reduce this practice was measured by posing direct questions. The scale for measuring the tendency towards cooperative behaviour proved insufficiently reliable. We therefore decided to use the two most important questions of it relating to driving speed. The habit of going to work by car proved to be predictable from a high self-enhancing illusion. The intention to reduce commuting by car can be predicted, though less reliably, if a person expects this to be achieved on a voluntary basis rather than as a forced measure. When the respondents thought of restraint, they thought primarily in terms of the probability of receiving a fine for exceeding the speed limit. The tendency towards cooperative behaviour can be predicted if one expects this to be achieved on a voluntary basis rather than by force; this also applies, but to a lesser extent, to a high self-enhancing illusion. The question of voluntariness or restraint applies here in particular to the intention to reduce car use for the benefit of the environment and in order to limit the number of fines. This latter variable, however, is conceptually related to cooperation. No correlation was established between the variables of alienation and powerlessness (that were at least reasonably reliable to measure) and the three aspects of mobility behaviour. Relationships were also found between the dependent variables and other data obtained from the respondents, but these connections were weak. All in all, the results from the Utrecht survey were negative if one believes that by using these measuring instruments, a clear increase in the predictability in the choice of travel mode for commuting would appear. In as far as the personality characteristics were reasonably and reliably measured, no relationship is shown to exist between them and mobility behaviour. This, however, can be seen as supporting the already established fact determined in the Hilversum survey, that deeply ingrained individual habits play a significant role in mobility behaviour. Although this particular behaviour proved to be correlated to a general environment attitude, still it is not considered to be the result of a configuration of personality characteristics. This behaviour originates primarily from a number of external factors. Finally, several policy recommendations are described, partly related to the results from both inquiries.
1188 Reference
I.M. de Boer, D. van Kreveld, P.G. Swanborn (1994) De keuze van een vervoermiddel. Twee onderzoeken naar situationele en persoonlijke determinanten van verplaatsingsgedrag. (Choosing a means of transportation: Two inquiries into situational and personal determinants of moving behaviour. With a summary and policy recommendations in English.) Utrecht: Vakgroep Sociale en Organisatiepsychologie, University of Utrecht.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1189
Changing Attitudes and Behaviour by Means of Providing Information. A Study on Private Car Use G. Tertoolen and E.C.H. Verstraten Section of Social and Organizational Psychology, University of Utrecht, Heidelberglaan 1, 3584 CS Utrecht, The Netherlands
Abstract In a field experiment we attempted to stimulate car users to come to a more selective use of their vehicle by means of providing information and feedback about different negative consequences of their car use. Attitude change was observed but the experimental treatments did not lead to behavioural changes. Attempts to influence car use arouse psychological resistance. Therefore, effects opposite to those intended occurred. We discuss the possible implications of the results for policymaking.
Introduction One of the emerging objectives of the Dutch environmental policy is to modify behaviour on a voluntary basis. As the car is a means of transportation that is rather damaging for the natural environment, one of the objectives of the present environmental policy in The Netherlands is to restrict private car use. 'Using a pricing policy' and 'influencing behaviour via communication and education' play an important part in the policy strategy. The goal is to achieve a social situation within which there is room for considerable structural changes and whereby traffic participants make a conscious choice between the different means of transportation. Implicitly it is assumed that the effects of the various measures will reinforce each other. In this paper an investigation is summarized on how private car use can be reduced by applying influence techniques based on behavioural science. We also attempt to gain an insight into the psychological resistance that is aroused when these influence techniques are applied to private car use (see for an extensive report: (1) and (2)).
Research Design Our study focuses on two research topics: how car use can be restricted by (1) emphasizing the negative collective environmental consequences or by (2) emphasizing the individual financial consequences. In a field experiment (N= 350) we attempted to stimulate car users to come to a more selective use of their vehicle by means of the following manipulations: providing information about the negative consequences of car use, self-monitoring of own transport behaviour and giving feedback on the negative consequences of personal car use. By means of a random procedure the respondents were assigned to five different conditions: three experimental and two control conditions. In the experimental conditions the respondents received information: in condition (1) about the environmental effects of car use, in condition (2) about the individual financial consequences of car use, and in condition (3) about both types of consequences of car use. Subsequently the subjects registered their own transport behaviour for eight weeks. Every two
1190
weeks they received feedback from the researcher's assistant about the consequences of their car use in a person-to-person talk. The content of the feedback referred to the particular kind of information received in the respective experimental conditions. The other respondents participated in the experiment without receiving any information or feedback about driving behaviour from a researcher's assistant (the control conditions). We asked all respondents if they were prepared to use the car as little as possible during the study period. In the experimental conditions, the respondents who gave a positive reply were requested to restrict their car use and thereby making a commitment to a research assistant. In all conditions, at the beginning and at the end of the experimental period questionnaires were filled in to measure the various attitudes with regard to car use and the environment. Results
The target group was chosen in such a way that it consisted of regular car users. They turned out to be more or less "attached" to using their vehicle. Speed, comfort and independence are mentioned as the most important advantages of the car (see figure 1). The respondents state that when they travel, neither the environment nor the costs are of much interest to them. Apart from the car, drivers make frequent use only of the bicycle; public transport is used sporadically by them. Attitudes play an important role in the perceived possibilities of the reduction of car use. However, these attitudes (including those related to the environment) appear to play hardly any role in the actual (reported) car use. In our study attitude change was observed but the experimental treatments did not lead to behavioural changes; i.e. no decline of car use was observed.
RAPIDITY INDEPENDENCE COMFORT HEALTH
COSTS ENVIRONMENT SAFETY
/
OTHERS
mm 0
5
lO
15
20
25
30
:35
%
F i g u r e I. M o s t transportation
important according
a s p e c t s in r e l a t i o n to to the p a r t i c i p a n t s .
1191 Information about the environment leads to a greater general concern about the environment but does not convince people to alter the way they use their car in order to create a cleaner environment. Unexpectedly, information about costs leads to less worry, not only about the environmental effects, but also about the financial consequences of car use. A combined environmental and costs information programme leads in many cases to results similar to those obtained with the control conditions; as in the pilot study, the effects often neutralize each other. Attempts to influence car use arouse psychological resistance. Information about the environment leads to dissonance reduction by means of attitude change. As a result of dissonance-enhancing information about the environment, car users who drive a lot, yet have a positive attitude towards the environment, start thinking that the environment is less important and point out that others are more responsible for the problems than themselves. They also become irritated with the behaviour of fellow road-users. Information about financial costs of the respondents car use leads to resistance as well. Car users experience financial measures as a restriction of their individual freedom and as a result they have a dim view of both the measures and the authorities responsible for implementing them. In addition, when car users react to these measures in a contrary manner, effects opposite to those intended often occur. The respondents who committed themselves to drive less, did in fact not keep their agreement. Instead they tended afterwards to displace the responsibility for environmental problems on to others. Those respondents who received information about the environment had a greater appreciation of environmental policy after the research. They probably have a greater understanding of the necessity for environmental protection and of the problems that can arise from a good environmental policy. Those respondents who received financial information had (slightly) less appreciation of environmental policy. By emphasizing how costly a car actually is, a reduction in appreciation of the policy was achieved. After all, it is the authorities who are responsible for the high costs of running a car. Our study received the lowest rating from the respondents who received financial information only. Discussion In our research some of the respondents were approached personally during a fairly long period and confronted with information specific to the individual about the effects on the environment of their car use. Such an intensive and personalized procedure should have more effect than a superficial, generalised attempt to influence via mass media, which the authorities often make use of. The environmental information we directed at individuals led to an increased general environmental awareness, but respondents did not become more aware of their own part in pollution. This result gives little encouragement for the authorities' publicity campaigns about the environment. With those respondents who were relatively well environmentally aware before the research and who used the car more than they judged, the information they received actually caused a reduction in environmental awareness. When the discrepancy between attitude (environmental awareness) and behaviour (car use) is pointed out then apparently people are more likely to alter their attitude than their behaviour. Even if the message is formulated so that the receivers cannot avoid the fact that it relates to their own individual behaviour, it still would not automatically lead to a change in behaviour. Just as disonnance theory forecasts, if attitude
1192 and behaviour are not in line, it is more likely that attitudes will change (which can mean that the environment becomes less important, as is shown in our research). Individually relevant information about the costs of running a car lead in our research to a higher estimate of the individual's car costs. However, the awareness that one's own car use has negative consequences (both individual financial and collective environmental consequences) was reduced as a result of the cost information. This is interpreted as a form of psychological reactance. Car drivers turn against the measures and those who implement the measures more when an attempt is made to reduce car use with financial methods than when environmental information is provided. We assumed that this reactance arose from a motivational state directed towards the re-establishment of free behaviour. If the car user holds the uncompromising view that he or she has a right to pollute for the very reason that he or she pays excise duty, reactance could be conceived as a form of protest as well. If taken from such an 'exchange' point of view other policy measures to restrict pollution (such as an appeal to change behaviour) could provoke irritation, since the person has already paid a compensation for the damaging behaviour. This would seriously harm the policy strategy "the polluter pays", which has the intended purpose to stimulate people to pollute less. Our research shows that in some cases the results of combined cost and environmental information are comparable with the results of the control group, i.e. respondents who did not receive any information at all. The environmental policy, as mentioned above, assumes that the effects of various measures will reinforce each other. Although it cannot be concluded on the basis of our research results that this assumption is essentially wrong, care is probably justified. Summing up it can be stated as a result of our reserch that little progress can be expected by requesting individual drivers to voluntarily reduce car use. The method of influence used in our research was based on some of the most powerful instruments available in psychology (giving individually directed feedback, self-registration and commitment). Nevertheless, there was no change in behaviour. Drivers will not leave their cars of their own free will, the car is too strongly linked to feelings of independence and convenience for that to happen. There are several positive attitudes, which are linked to various individual advantages where car use is concerned, whereas there are only limited negative attitudes, which are linked to the disadvantages of car use. With such a balance the dissonance theory forecasts that the negative attitudes will change in the direction of the most prominent attitudes. The environment is seen to be important, but we can not talk of a central attitude that is so important that car use is equivalent to it. Mass media publicity campaigns do not seem able to develop such a central environmental attitude on a large scale. References 1 Tertoolen, G. (1994). Uit eigen beweging...?.t Een veldexperiment over be~nvloedingspogingen van het autogebruik en de daardoor opgeroepen psychologische weerstanden (With a summary in English). Dissertation: Universiteit Utrecht. 2 Tertoolen, G. (1995). Free to Move... ?.t A field experiment on attempts to influence private car use and the psychological resistance it evokes. A policy oriented report. Utrecht: University of Utrecht, VSOP.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1193
Energy and environmental issues as choosing elements for selecting options in the transportation sector aimed at reducing CO 2 emissions" an application to the italian case D. Barbieri, A. Nucara, M. Pietrafesa and G. Rizzo Istituto di Ingegneria Civile ed Energetica, Facolt/l di Ingegneria, Universit/t di Reggio Calabria Via E. Cuzzocrea, 48, 89128 Reggio Calabria, Italy Abstract
A new transportation demand model is described showing a simple data-base structure. It only requires input data referring to the fleets and the engine characteristics of the transportation park. The main characteristic of the model is its expertise in analysing the effects of different policies oriented to the reduction of the pollution levels and to the energy savings in the transportation sector. The results are provided both in terms of energy consumption and quantities of pollutant released to the environment. The effects of different transportation scenarios can be easily analysed using a simple "electronic sheet" way of representation.
1. INTRODUCTION In this paper we will present a new transport demand model, showing a simple data-base structure. It is founded on generally available information about the structure of the transportation park and on the size and type of the used engines and allows the obtaining of a desegregated view of the system and the evaluation of the pollutant emissions. The model, built-up for the whole Italian sector, is easily applicable, with minor modifications, to any country for which the required input data is available. It is also suitable for analysis regarding smaller areas, even to a regional scale. The main feature of the model is to provide results both in terms of energy consumption and quantities of pollutants released in the environment, as effects of the assumed scenarios. Starting from the "zero" scenario, referring to the system when all the requirements of the Italian government' s rules are accomplished, some alternative options are analysed.
2. DESCRIPTION OF THE MODEL The structure of the model is essentially founded on the following four points: 1. the transportation demand of the analysed region is organised with respect to three components: the object of the transportation (people or goods), its spatial domain (urban or
1194 non-urban) and the way with which the transportation is accomplished. The units employed are the passengers per kilometre and per year (pkm/y) or the tons of goods per kilometre moved per year (tpk/y); 2. a distributive model assigns the total energy consumption to each fuel source, by means of the specific values of the consumptions, available in the literature. The units here utilised are the kilotons equivalent of oil (ktoe); 3. for each pollutant component released by the transportation means, the emission factor, intended as the quantity of pollutant released for unitary energy (t/ktoe) is calculated. These parameters link the energy consumption with the CO2 emissions [ 1-2]; 4. the quantities of the CO2 emissions are then computed for each component of the transportation demand and for each energy source. A description of the main features and potentialities of the model can be offered by analysing one of the tables that constitute the way of representation of the results obtained. Fig. 1 contains an example of these print outs: along with the modal distribution of the transportation demand, subdivided into the urban and extra-urban components. The figure reports the specific and total energy consumption by each component. The share of the energy demand covered by the fuel sources is also shown. The yearly increasing rate, for each modal component of the whole transportation sector, represents the most important parameter in order to characterize the scenarios. In any case, the numerical quantities reported in Fig. 1 are to be considered only as an example, since the main purpose of this paper is the presentation of the structure and the potentialities of the model. The model also allows evaluations of the emissions of the main environmental pollutants linked to the transportation sector: carbon monoxide (CO), nitrogen oxides (NOx), volatile organic compounds (VOC), including hydrocarbons and suspended particles (SP), in terms of yearly tons of quantities given off. The types of pollutants chosen depend on the kind of 1992
(ESTIMATED YALUES) SnerW rote by sources
Forms of transportetlon
Yearly~ncreaslng Accomplished M o d a l Specific Energy rate (%) demand split consumption consumptson 1989-1992 Mrd pKm % gep/pKm Kte~
r
emi.ians
Energy sources Gas~lne D,asel
Jet fuel
(ktr Energy sources
LPG
El energy CNG
0.073
0.011
Gasoline
Diesel
5269
1892
Jet fuel
LPG
El energy CNG
Total
Passengers urban traffic Cars
.3.14
166.85
75.12
52
8676.17
0.674
Motorcycles
3.89
35.51
15.99
Buses
0.56
15.13
6.81
21
745.74
I
18
272.36
Underground
0.56
2.60
1.17
10
26.03
Other collective means
056
2.00
0.90
15
30.05
Urban passengers total
3.02
222.10
100.00
3.14
860,48
0.242
571
62
245
245 48
1 0.024
7793
6T2
672 1 0.976
1
i
9750.86
55
5941
2188
571
62
8814
6150
4229
864
90
11332
Extraurban pass traffic Cars
70 9'3
35
Motorcycles
3,89
17,73
3,49
22
390.01
Buses. trams
4.05
76.35
15.02
15
1145.30
krcralts
8.51
7.32
1.44
313
2291.44
8
Railways
1.41
46,35
9,12
Extraurban pass total
321
508,23
100,00
12616,85
0,541
0,872
0,076
0.011
1
351 0.947
0.053
351 977
9"E
2065
370,77
0,25
16814,37
0,75
84 6501
5290
517 2065
864
517
600 90
15326
Fre4ght traff~c T rue,ks 451on
3,03
9,88
g0
Trucks > 5 ton
3,0,3
20.57
8,57
82
1686,92
I
1520
1520
Trucks. long vehicles
3.03
138.61
23,71
57 74
35
4851,46
1
4371
4"~71
Ships
0,95
34,98
14.57
6
209,87
I
189
9
199,81
0.25
Railway
1,39
22,20
9,25
Total freight traffic
2,56
240,08
100,00
Total
.~,00
970,40
2133,99
0,088
0,912
9082,05
Figure 1. Example of the output structure of the model.
169
1754
1923
189
45
279
324
169
7879
279
8326
1L~11
15306
2065
1435
898
151
32466
1195 analysis required [4]. Energy consumptions and pollutant emissions represent the selecting criteria in order to judge the effects produced by an assigned modal and structural distribution [5-6].
3. APPLICATION TO SOME TRANSPORTATION SCENARIOS Three scenarios have been assumed here in order to show the features of the method. 9 The "zero" scenario. This scenario has been identified as "zero" because it is considered the
reference point for the whole analysis [7]. It is characterised by the absence of specific interventions and therefore it appears as simply driven by the demand of mobility, for which an incrasing tendency is supposed. 9 The "modal split" scenario. This option takes into account the effects of some interventions that modify the modal split of transportation. From 1995 until 2020 the transportation demand is supposed to shift toward the public means, with rates of 20% for the urban passenger, 25% for the extra-urban passenger and 20% for the freight movement. Moreover, an increase of 15% in the use of bicycles in the urban context is also supposed. 9 The "car p o o l i n g " scenario. Within this alternative we suppose that the italian occupancy coefficient rises from 1.3 to 2.0 persons per car in the urban context and from 1.7 to 2.5 persons per car in the extra-urban context by the year 2020. The model provides a grafic representation of the compared effects of the assumed scenarios, but also details numeric results, within each scenario under analysis, referring to the other environmental pollutants. An analysis of the results provided by each of the previous mentioned scenarios is beyond the purposes of the present paper. Fig. 2 depicts the estimated trends of CO2 emissions of the whole Italian transportation sector from 1995 until 2020. Curves refer to the different scenarios. Table 1 contains the percentage of reduction for the considered emissions and for the energy consumption recovered by means of the "car pooling" scenario referred to the base case, that is the "zero" scenario.
~-, 55 50-II Zero Scenario o 45-r,r r~ .,..~
[] Modal Split Scenario 40--
D Car Pooling Scenario
r
9
~
35)f
.."
30 1992
,
~
I
l
I
I
I
1996
2000
2004
2008
2012
2016
2020 years
Figure 2. Estimated trends of CO 2 emissions for the whole italian transportation sector.
1196
Table 1 Reductions percentages obtained with the car pooling scenario with respect to the zero one. CO2 CO NOx VOC PS Final energy consumptions Urban
30.9
26.2
29.8
Extra-urban Total
19.3
29.8
31.2
23.1
21.3
19.4
23.3
19.6
18.9
26.9
23.5
8.7
17.1
17.5
19.7
4. CONCLUSIONS As it is possible to note, even within the summary here presented, the reliability of the approach strongly relies on the accuracy of the available data. Data on the car and truck fleet and on the freight movements are, as matter of fact, capable of affecting in a remarkable way the results obtained. This data, in fact, is employed as multiplier parameters by the algorithm of the model. On the other hand, the main assumptions of the methodology, especially concerning the evaluation of the emission factors, could introduce some simplifying features within the frame of approach. These considerations suggest a need for further attention when analysing the transportation sector and the complex links between energy consumption and environmental emissions. But the "electronic sheet" structure of the model and its capability of investigating different phenomena, such as pollutant releases and fuel use, make it a suitable tool in order to explore the effects of different scenarios referring to the policies to be selected in the transportation sector.
5. REFERENCES 1
2
3 4 5
6 7
S. Unnasch, C.B. Moyer, D.D. Lowell and M.D. Jackson, , Comparing the Impact of Different Transportation Fuels on the Greenhouse Effect, Acurex Corporation Environmental Systems Division, California, Mountain View, (1989). M.A. De Luchi, Emissions of Greenhouse Gases from the Use of Transportation Fuels and Electricity, Center for Transportation Research, Argonne National Laboratory, United States Department of Energy, Illinois, Argonne, (1991). Ministero dei Trasporti, Conto Nazionale dei Trasporti, Italy, Roma, (1992). G. Rizzo, Trasporto su Strada di Merci e Persone: Aspetti Tecnologici ed Ambientali, La Nuova Ecologia, Vol. 80, Italy, Milano, (1990). M. Fergusson and H. Claire, Atmosferic Emissions from the Use of Transport in the United Kingdom, Volume two: The Effect of Alternative Transport Policies, Earth Resources Research, United Kingdom, London, (1990). Peeters, The Netherlands Travelling Clean, Netherlands Agency for Energy and Environment, Netherlands, Utrecht, (1989). W. Leontieff and P. Costa, I1 Trasporto Merci e l'Economia Italiana: Scenari di Interazione al 2000 e al 2015, Piano Generale dei Trasporti, Italy, Roma, (1988).
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1197
C O M P A R A T I V E ANALYSIS OF OPTIONS FOR SUSTAINABLE T R A N S P O R T AND T R A F F I C SYSTEMS IN THE 21st CENTURY
Peter Nijkamp
Sytze Rienstra
Jaap Vleugel
Economic and Social Institute, Free University, De Boelelaan 1105, 1081 HV Amsterdam, The Netherlands. In association with: Marina van Geenhuizen Edith van der Heijden/ Ton Rooijers Richard Smokers Johan Visser
UCL, London VSC, Haren ECN, Petten OTB, Delft.
Abstract In this project on future sustainable transport altematives a two-step search process has been followed. First an analysis of critical success and failure factors of new technological options in passenger transport is made. These factors are found in the spatial, institutional, economic and social/psychological environment of the transport system. Next, systematically structured and expert based scenarios are constructed in order to achieve a sustainable transport system in the year 2030 in which possible, expected and desired developments in the distinct fields are analyzed. Finally some policy conclusions are drawn.
1.
INTRODUCTION
The current trends in almost all fields show a continuing shift in the modal split towards individual modes and a rising mobility rate. Therefore, the externalities (social costs) caused by transport are still rising, which makes it necessary to bend these trends in order to achieve a more sustainable transport system. In this project we have investigated which technical options are developed (or are being developed nowadays), which may reduce these externalities. The emphasis is here put on the resulting reduction of CO2-emissions and other greenhouse gases, since this provides a direct link to sustainable development. In the first phase (state o f the art) of the project, success and failure factors for the introduction of new technological options have been identified. In the second phase (scenarios for a sustainable transport system), several scenarios have been constructed in which these options have been filled in for the transport system. Finally, an assessment of policy choices has been made which may influence the future of transport.
1198 2.
RESEARCH STRATEGY
In the first phase, an extensive literature search has been carried out supplemented with an international workshop in order to identify the success and failure factors of several new options (new fuels, electric car, people mover, subterranean transport, telematics, HST, maglev and shuttles through vacuum tunnels), which may contribute to the reduction of externalities when they would replace current modes. In the second phase of the project reference scenarios, which describe contrasting future developments in the field of transport have been constructed by using the recently developed spider model. In this stage also a questionnaire has been sent to Dutch transport experts. The results of this survey have been presented to a second international workshop organized towards the end of the project. With this information 'expected' and 'desired' scenarios have been constructed, based on these expert opinions. This has also allowed us to asses the resulting environmental implications.
3.
PHASE 1" STATE OF THE ART
There are many critical success and failure factors which influence new technical options. The most important of them have been summarized in Figure 1. SUCCESS
ECONOMIC
SPATIAL
"
LEVEL OF MOBIUTY MODAL SPLIT EXTERNAL EFFECTS PER OPTION
Spatial organisation of living and working areas
"
1
Government support P r e s s u r e of n a t i o n a l I n d u s t r y
IN~IlITUTIONAL
SOCIAl./ PSYCHOLOGICAL
TECHNICAL
AND FAILURE FACTORS
Competitiveness Profitability Financing
TOTAL EXTERNAL EFFECTS OF TRANSPORT
Acceptance (society) Adoption (individual)
-
Direction of R & D Environmental criteria Technical Inertia
I
SUSTAINABLE MOBILITY
I
Figure 1. Success and failure factors of new technological options
It appears that especially modes and options which are compatible with existing modes have the best chance to succeed since they may use (temporarily) existing infrastructure and may easily be incorporated in prevailing institutions and existing transport systems. Therefore, the High Speed Train and improvements of the current car - and to a lesser extent also new fuels (and electricity) in cars - have the best chance of being introduced and accepted. In general however, a principal choice has to be made between an emphasis on individual or collective modes, since both modes require an entirely different spatial organisation, technical development and institutional environment. These also have a significant influence on important economic and social/psychological factors. For a more detailed analysis of these phenomenon and their solution strategies we refer to Nijkamp, Rienstra and Vleugel (1994).
1199
4.
PHASE 2: S C E N A R I O S F O R 2030
4.1
The methodology of the spider model Based on the phase 1 study and various scenario experiments developed by others, we have identified eight main factors, which influence the future transport system; these are to be found in four distinct scientific fields. These factors are presented in the so called spider model (see Figure 2). ~ c o m p a c t l ......
speelalisation _ "~." ~ ~ and concentratigjrl~ ; ~
" "
chalns~and'zones~\ in d i v i d ~ . i a ~ ~ soci,I
. - ~
I r:7
|
/
/
I
/
".. -
publlc
management
9
I
rolm~gement
profltTIble I~ansport 9 I
in~p~ subsidlsed transport
coor centralisation
Figure 3. The expert based scenarios Legend ....... expected developments ........ desired developments Scenarios can be constructed by connecting characteristic points on the distinct axes. In this way in principle thousands of scenarios may be constructed. On the basis of assumptions on the developments in the several axes the resulting transport system can be identified and described. We constructed two scenarios which form the inner and outer circle of the spider and are used to analyze the scenarios designed by means of expert opinions; therefore these are called reference scenarios. The regulatory scenario forms the outer circle of the spider. A compact and concentrated spatial development is combined with an emphasis on equity and regulatory measures. In this scenario a transport system occurs which is largely based on collective modes, while individual modes largely disappear. The market scenario on the other hand is formed by the inner circle. In this scenario diffuse spatial developments have been combined with market-oriented measures. In this way a transport system occurs in which individual modes are dominant.
4.2
The expected and desired scenario Both an expected and desired scenario have been constructed next by means of the questionnaire and the information gathered from the second international workshop. In the expected scenario (see Figure 2) it is expected that current trends will largely continue. Therefore, mainly improved versions of the current car will be dominant, while also measures which are common nowadays (reducing parking places, raising parking tariffs and fuel prices) will be introduced to a larger extent. Also road pricing may be introduced to some extent. It is expected that also other main trends in
1200 society and economy will largely continue. In the desired scenario an entirely other transport system is found. Measures which will be introduced to make the car more unattractive are introduced at a much larger scale, while collective modes are supported much more than expected. Also many trends in society are reversed in order to favour such a transport system. It is clear that the expected scenario will only lead to more sustainability, if a much larger improvement of the current car will occur than is now expected by technical experts. In the desired scenario much more sustainability may be achieved. In this scenario however, many more changes and measures are necessary, which will have a much larger impact on the life of individuals and the society at large. It is noteworthy that the latter phase generated many new insights into the feasibility and desirability of transport systems alternatives, in particular from the viewpoint of global environmental changes. More details can be found in a forthcoming publication (see Nijkamp, Rienstra and Vleugel, 1995).
5.
CONCLUSIONS
The current trends in transport are not expected to lead to a more sustainable transport system. Therefore, a change in the behaviour of individuals and a stricter government policy seem to be necessary. Several consistent and effective policy choices have to be made. The most important concern is the one between an emphasis on an individual or a collective transport system, because both systems require an entirely different policy in many fields, which have a profound impact on many other aspects in society, like individual lifestyles, the level of equity and individualisation, the choice of housing locations etc. Other issues related to this choice are the necessary reduction of the mobility growth, the investments in transport infrastructure, the direction of technical development, the way in which transport may be regulated etc. From the expert opinions it may be concluded that government policy alone may not be sufficient for achieving a more sustainable transport system, for example in the expected scenario the CO2-emissions will probably not be reduced in a sufficient way. The policy solution chosen by experts appears to favour a more collective transport system (as is shown in the desired scenario). It may be concluded that the road towards a (more) sustainable transport system will be very hard, but that with sufficient behaviourial changes and other changes in society such a (more) sustainable transport system seems to be socially and technically feasible.
REFERENCES -Nijkamp, P., S.A. Rienstra and J.M. Vleugel, 1994, Comparative analysis of options for sustainable transport and traffic systems in the 21st century, phase 1: state of the art, report of the Dutch National Research Programme on Global Air pollution and Climate Change, theme E: Sustainable solutions (policy research), ESI, Free University, Amsterdam. -Nijkamp, P., S.A. Rienstra and J.M. Vleugel, 1995, Comparative analysis of options
for sustainable transport and traffic systems in the 21st century, phase 2: Scenarios for a sustainable transport system in 2030, idem.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1201
ASSESSMENT REPORT ON SUBTHEME "CULTURE, CONSUMPTION AND LIFESTYLES IN RELATION TO SUSTAINABLE DEVELOPMENT" C.A.J. Vlek Department of Psychology, University of Groningen Grote Kruisstraat 2/1 9712 TS Groningen The Netherlands
With contributions by: B. Breemhaar, P. Ester
KUB, Catholic University Brabant, Tilburg
C. Wilke, M.A. Mentzel
RUL, University of Leiden
W. Biesiot, E.M. Kamminga, H.C. Moll, G. Slotegraaf, A.J.M. Schoot Uiterkamp, A.W.L. van der Veen, H. Wilting K. Blok, C. Vringer R.M. van Aarts, J. Goudsblom, F. Spier, C. Schmidt C.J.H. Midden, W.A. van Gool, A.L. Meijnders
RUG, University of Groningen
RUU, University of Utrecht UvA, University of Amsterdam
TUE, Eindhoven University of Technology
1202
Contents Abstract 1.
Introduction
2.
The e n v i r o n m e n t a l b u r d e n of h o u s e h o l d c o n s u m p t i o n
3.
The p r o d u c t i o n - c o n s u m p t i o n cycle
4.
Population, affluence and technology
5.
E c o n o m i s a t i o n a n d ecologisation
6.
Rise of t h e e n v i r o n m e n t a l p r o t e c t i o n m o v e m e n t
7.
S t a t u s seeking t h r o u g h m o d e r a t i o n
8.
CO2 emissions r e d u c t i o n via lifestyle c h a n g e s
9.
Low-energy, low-CO2 emissions scenarios
10. Lifestyles a n d domestic e n e r g y c o n s u m p t i o n 11. F e a r - a r o u s a l a n d a r g u m e n t a t i o n in risk c o m m u n i c a t i o n 12. Social welfare an e n v i r o n m e n t a l quality 13. T o w a r d s s u s t a i n a b l e h o u s e h o l d m e t a b o l i s m 14. G e n e r a l conclusions, suggestions a n d r e c o m m e n d a t i o n s 14.1 'Economisation' and environmental exploitation 14.2 Some policy recommendations 14.3 Lifestyles and behaviour change strategies 14.4 International research efforts 14.5 Further NRP intentions 14.6 Religions on consumption 15. R e f e r e n c e s
ABSTRACT Many people believe that 'sustainable solutions' to global air pollution and climate change should include significant changes in human consumption and lifestyles. Under this heading six different NRP projects have been conducted. This chapter gives a review and assessment of these projects, supplemented with discussions of related research. The paper starts with a general statement of the environmental
1203 problem of household metabolism as a key component of the socio-economic production- consumption cycle. It summarises and comments upon nine different (sub)projects. And it ends with general observations, conclusions and suggestions for r e s e a r c h and policy in relation to s u s t a i n a b l e c o n s u m p t i o n p a t t e r n s . C o n c e p t u a l problems, m u l t i d i s c i p l i n a r y perspectives and i n t e r n a t i o n a l implications are given special consideration. 1.
INTRODUCTION
Household consumption is at the beginning and at the end of industrial production. H u m a n needs and desires, habits and decisions, norms and rights in modern society materialise in an enormous 'household metabolism'. This involves the transformation, sooner or later, of many different kinds of energy, materials and products into various kinds of positive fulfilment, of course, b u t also in environment-polluting kinds of gaseous, liquid or solid waste. Through the direct and indirect use of fossil-fuel energy for household activities, including transport, and through the exhaust gases from landfills and waste incinerators, households contribute significantly to global air pollution and the risks of climate change. In this paper, brief reviews and commentaries are given of six different projects on 'culture, consumption and lifestyles', as conducted during Phase 1 (1990-1994) of the Dutch National Research Programme on Global Air Pollution and Climate Change (NRP). These will be supplemented with a few related but otherwise funded studies on household consumption vis-a-vis environmental resource use. Table 1.1 offers an overview of project codes, titles and principal investigators. For more extensive project descriptions and for full references to complete project reports, the reader is referred to the relevant project summaries elsewhere in this volume. Table 1.1 List of projects in the NRP subtheme "Culture, Consumption and Lifestyles" Title
Project leader
Number
Conditions for a moral code of moderation
J. Goudsblom
851038
Reduction of CO2 emissions by lifestyle changes
W. Biesiot/ H.C. Moll
852086
Analysis of the social significance of long-term lowenergy/low CO2 scenarios for The Netherlands
W. Biesiot/ H.C. Moll
852085
Toward a sustainable lifestyles
P. Ester/ C.J.M. Midden
853119
Cognitive vs emotion oriented information on sound C.J.M. Midden Environmental behaviour
852093
1204
Non-NRP-projects
Social welfare and environmental quality
M.A. Mentzel
Sustainable household metabolism
A.J.M. Schoot Uiterkamp
This section will be concluded by a general discussion of household consumption in view of sustainable development, followed by conclusions and suggestions for research and policy regarding household consumption and consumer lifestyles. 2.
T H E E N V I R O N M E N T A L B U R D E N OF H O U S E H O L D C O N S U M P T I O N
The throughput of energy, materials and products in households of varying size and style has grown impressively during the last fifty years. Some pertinent figures for The N e t h e r l a n d s are as follows. During the period of 1950-1990 the Dutch population has increased from 10 to 15 million inhabitants. In the same period, the percentage of Dutch land area (a total of about 34.000 square kilometres) used for buildings, roads and recreational facilities, increased from 8.4 to 16.1. Around 1950 there existed about 2 million household dwellings; this number had risen to 6 million in 1992. The average annual income, corrected for inflation, of heads of households in 1990 was twice as high as in 1950. Between 1965 and 1992 water consumption in Dutch households has increased from 100 to 135 liters per person per day; today about twice as much water is being used for bathing and showering and for textile washing t h a n 30 years ago. The ownership and use of motor vehicles - especially p a s s e n g e r cars, but also vans and lorries - has grown very strongly since the 1950s. In 1960 some 670,000 four-wheeled motor vehicles populated the Dutch roads and streets. In 1980 there were about 4 million and in 1990 about 6 million motor vehicles (Vlek et al., 1993). The Dutch fleet of motor vehicles is expected to approximate the figure of 10 million in 2010, an average of about 300 motor vehicles per square kilometre of land area. The number of airplane starts and landings at Schiphol Amsterdam Airport rose from about 90,000 in 1960 to some 235,000 in 1990. The Schiphol authorities expect (and stimulate) t h a t between 1990 and 2010 the n u m b e r of 'passenger movements' will triple from 16 to 50 million annually. From an international perspective it may suffice to quote Corson (1994) who - in a recent special issue of F u t u r e s - outlined various strategies for a sustainable future. The author starts his paper by describing current 'unsustainable trends': "Between 1950 and 1990, the world's h u m a n population more than doubled (from 2.6 billion to 5.3 billion), domestic livestock population grew 1.8-fold (from 2.3 billion to 4.1 billion), grain consumption rose 2.6-fold, water use nearly tripled, fish consumption grew 4.4-fold, and energy use quintupled. Over the same period, global consumption of wood and copper roughly doubled; steel production quadrupled;
1205 economic output nearly quintupled; industrial production grew sevenfold; aluminium output and the use of chemical fertilizers increased roughly 10-fold; world production of organic chemicals, major sources of air and water pollution, rose 20-fold; and global air travel, which causes significant atmospheric pollution, soared nearly 70-fold. On average, resource use per person nearly tripled between 1950 and 1990. This growth, coupled with a doubling of human population, resulted in roughly a sixfold increase in human impact on the global environment during the four decades. H u m a n activity is now altering the Earth's basic life-support systems and cycles, including the atmospheric system and the carbon, nitrogen, sulphur, biologic and hydrologic cycles" (Corson 1994, p. 206-207). Taking this together, we may conclude that the households sector since 1950 has developed as an environment-burdening consumer of energy, water and materials, of meat, fish and agricultural produce, and of motorised transport and land area, and it has become a major producer of diverse kinds of waste. Household metabolism, therefore, is an important focus for scientific research and for government policies aimed at reducing global air pollution and the risks of climate change. 3.
T H E P R O D U C T I O N - C O N S U M P T I O N CYCLE
To understand household metabolism, its driving forces and its potential for change toward sustainable conl~umption patterns, it is necessary to appreciate the interwovenness of house!lold consumption and industrial production. Figure 3.1 r e p r e s e n t s w h a t m a y be called the p r o d u c t i o n - c o n s u m p t i o n cycle, as institutionalised in a social, i.e., government-regulated market economy. The figure reflects the siJnple truth that consumers and producers need each other for different reasons, and that both parties need some government regulation for which the government :n turn needs them, again for different reasons. The relationships among consumers, producers and government are expressed in flows of money, products, labour, taxes and subsidies. Main system functions for consumers are feeding, clothing, housing, education and recreation. Major functions for producers are energy provision, industrial production, agriculture and stock-breeding, product distribution and commerce. Inputs from outside the socio-economic system are formed by various environmental resources such as energy, raw materials and land area. External outputs or derivatives occur in the form of various kinds of waste materials as well as transport.
1206
-" " "
/
land - ~ ~ ~ _ . ~ .
mobility~l
~
waste
~.
~
___
" " ,, ,,,
~
",%.
,
"~
,, "
energy ~
income
.~/
,,~,
~
7 // / / " / .
.
.
. '
......... " > \\\
i
9
I CONSUMER I . . . . . . . . . -, I
[ I
-'.,""'"'" ~~ housing feeding clothing education
\
N ~
\
i
.// products ~'- / ~ ~ ~ turnover----
~
~
I ~ /
/
land
raw materials
PRODUCER I : ~ " " ..... ]~--~-"---"~* waste
// "~ ~
energy
,,~l
/
/
recreation
"'''''.~ utilities industry agriculture transport
transport
commerce/ services
Figure 3.1 Social m a r k e t economy 4.
POPULATION, AFFLUENCE AND TECHNOLOGY
An other w a y of positioning household consumption is to follow E h r l i c h and Holdren's (Ehrlich et al., 1971) formula for e s t i m a t i n g the total e n v i r o n m e n t a l impact from h u m a n activities: I = P x A x T, or: environmental impact (I) equals the product of population size (P), the degree of affluence (A) per person, and the e n v i r o n m e n t a l d a m a g e from the technology used (T) to produce one u n i t of affluence. The formula implies t h a t there are three different fronts on which the battle for sustainable development is to be fought. The formula also reveals the s u b s t i t u t a b i l i t y of one component by the other. To illustrate, total e n v i r o n m e n t a l i m p a c t m i g h t r e m a i n constant under considerable population growth, as long as average personal affluence and/or the technological impact per unit of affluence is/are p r o p o r t i o n a l l y reduced. Also, while total e n v i r o n m e n t a l i m p a c t s t a y s c o n s t a n t , the degree of affluence per person m a y well increase significantly, provided t h a t the n u m b e r of people and/or technological impact per unit affluence is/are proportionally reduced. The 'IPAT-formula' enables one to explain and predict and to eventually m a n a g e the size and seriousness of environmental impact, for different geographic regions or countries of the world, to d e t e r m i n e the most i m p o r t a n t i m p a c t g r o w t h factor(s) a n d to d r a w conclusions on o p t i m a l environmental m a n a g e m e n t policies.
1207 Recently, Goodland, Daly and Kellenberg (1994) have systematically examined the potential for change in the three areas covered by the IPAT-formula: (1) limiting population growth, (2) limiting affluence and consumption growth, and (3) reducing the environmental impact of production and consumption technology. Like Corson (1994), these authors generally conclude on a number of policy priorities, which are different in character for high-income and low-income nations of the world. For example, high-income nations are advised to work on "transforming the culture of consumerism (..) into an ethics of sufficiency and environmental sustainability", and on "internalizing environmental costs in energy prices and accelerating the transition to renewable energy sources" (Goodman et al., 1994, p. 153). In contrast, the authors advise low-income nations to give priority to: "accelerating the transition towards population stability (..), supporting technologies which provide increased employment opportunities for unemployed and underemployed individuals (..), and improving efforts towards poverty alleviation .." (Goodman et al., 1994, p. 154). One general conclusion by Goodland et al. is: "Technological change and population stabilization cannot suffice to move the world towards an environmentally sustainable future. Instead, a reduction in per capita consumption in high-income nations and a decrease in environmental throughput are required" (Goodman et al., 1994, p. 154). Whether and how this could be accomplished is a major question underlying this review of recent Dutch research on household metabolism. Let us now look at the separate projects. 5.
ECONOMISATION AND ECOLOGISATION
Three interrelated NRP-funded projects on consumption and lifestyles (NRP project no. 851038) have been carried out at the University of Amsterdam. Together with sociologist-psychologist Goudsblom, researchers Schmidt, Spier and Aarts have looked into the possibilities for developing a 'morality of increasing moderation' among household consumers. All three projects started from the premise that environmental degradation, including global air pollution and climate change, is a problem of h u m a n civilisation (see also Thoenes, 1990 and Vermeersch, 1990), which strongly resides in ecologically unbridled economic growth. Reviewing available literature, Schmidt (NRP project no. 851038-1) - interacting with Goudsblom - critically analyses and describes the historical process of 'economisation' in western society. In his view, this has come to replace more traditional, often military ways to acquire, maintain and expand the wealth of nations. Through strong inherent growth tendencies economisation has also led to a decline of 'ecological regimes' for production and consumption. With this less violent, thus 'more civilised' economisation also came the translation of 'everything worthwhile' into equivalent monetary values. By implication, non-valuable goods and services tended to be seen as 'worthless'. Effective economisation required greater military self-control from powerful individuals, and it gave greater technical control over nature. Economisation also led to a significant increase in the division of labour, and to the development of industrialised cities. Thus, many people came to live at a distance from the natural environment and they lost sight of their natural living conditions. And gradually, with increasing affluence came a relaxation of traditional norms of frugality and thrift.
1208 The solution to 'economisation', according to Schmidt, lies in a g r a d u a l 'ecologisation' of society. This would amount to developing a strong environmental a w a r e n e s s and a t t e m p t i n g to keep h u m a n activities and h u m a n populations within the limits of the earth's carrying capacity. Such a 'next step in h u m a n civilisation' would require new forms of self-control and it would demand new kinds of intelligent control over the environmental effects of human action. Comments on Schmidt's project
Agriculturation, militarisation, economisation and ecologisation might well be distinguishable periods in h u m a n civilisation. It seems important to note t h a t these developmental stages occur(red) in correlation with a growing and spreading h u m a n population and with an ever more intensive exploitation of the earth's surface. Was economisation actually driven by war-weariness, or were military efforts only too supportive of economic expansion? Or was it population growth and increasing agricultural uncertainties that have led to economic industrialisation? Other socio-cultural analysts of environmental degradation point at the role of C h r i s t i a n i t y (White, 1967) in which m a n is positioned above, not among other living beings. Or they indicate the fundamental roles of science, technology and capital (Vermeersch, 1990) in creating a technologically violent 'here and now' consumer culture. In the latter, privacy, power and freedom are key values for the individual (Thoenes, 1990). Environmental economists (Opschoor, 1989), point at the conflict between short-term individual and long-term collective interests, which is inherent in a free-market economy; through the accumulation of the external costs of n u m e r o u s individual activities, society as a whole is burdened with collective costs and risks for which individual actors tend to feel little responsibility. Schmidt's analysis is i m p o r t a n t in so far as it draws our a t t e n t i o n to the drawbacks and the possible excesses of economisation as a civilisation stage. As a socio-cultural analysis of 'the environmental question', however, it needs to be supplemented by views from other social-science disciplines. 6.
RISE OF THE ENVIRONMENTAL PROTECTION MOVEMENT
What societal powers exist that are or could be pushing the needed 'ecologisation' deemed necessary by Schmidt? Spier (NRP project no. 851038-2), supported by Goudsblom, has explored the rise and the effectiveness of the environmental protection movement (EPM) in The Netherlands. Started in the early twentieth century as an elite form of appreciation and care for nature, the Dutch EPM now consists of various professional organisations of monitors, publicists, advisors and activists, working towards the protection of natural areas and warning against careless industrial pollution, the unlimited growth of motorised traffic and the inconsiderate use of open space for siting industries, house-building and/or road construction. EPM organisations, however, differ in effectiveness. Spier makes a distinction between the more traditional societies for the protection of birds and for the maintenance of natural monuments on the one hand, and the newer, more critically operating organisations for general environmental protection on the other. He notes that the, more successful, traditional organisations have always a t t e m p t e d to realise their relatively modest ambitions in a positive and prudent way, taking care to appeal to the general public and to be acceptable for policy makers. The more general EPM organisations, however, are pursuing their more
1209 ambitious programmes for social behaviour change and societal restructuring in a less discreet and more activistic m a n n e r , while e m p h a s i s i n g negative developments in environmental conditions. This may explain why many people don't like them and refuse to heed their advice. Spier concludes that "sounding the alarm is a necessary component of efforts to stimulate ecological awareness, (but) positively phrased campaigns to stimulate specific forms of moderation will in my view prove to be more successful than alarmist approaches, and should clearly be kept separated. In addition, the ability to exercise influence at the highest level of decision making (..) may be very helpful to spread forms of ecological moderation". C o m m e n t s o n Spier's p r o j e c t For environmental policy to be effective, there should be sufficient social understanding of the problem and acceptance of policy measures. This is often dependent upon the activities of pioneering organisations. Spier's project demonstrates that their public following and their political influence significantly depend - of course - on the relative attractiveness or painfulness of their message, but also upon the style and the way in which this message is being distributed. For environmental protection organisations, being stereotyped as a noisy minority preaching unachievable ideals, would be lethal. Alternatively, the stigma of the well-to-do land owner who wants to preserve natural peace and quiet for himself and his friends, would be similarly killing. The EPM would be well advised to think hard about the fundamental conditions that should be fulfilled for their messages to get t hr o u g h and to be transformed into actual policy measures and social behaviour changes. Using principles of commercial product marketing, taking account of basic social-behavioural mechanisms, and continually working on image-building might prove to be effective. The government, as the guardian of public environmental qualities, could assist here: by supporting diagnostic environmental research, by sharing the use of its distribution channels, and by being clear and consistent in its environmental policy goals. Currently, both the traditional and the 'modern' EPM organisations are dissatisfied with Dutch policy making - and more so with policy implementation - concerning environmental qualities and conditions. They agree with one a n o t h e r t h a t long-term e n v i r o n m e n t a l protection also is an economic necessity and t h a t various short-term economic priorities reflect the short-sightedness of government departments and industrial organisations. 7.
STATUS SEEKING THROUGH MODERATION
Obviously, the basic attitude and 'lifestyle' of the environmental protection movement is considerate for the environment, moderate in consumption, reserved towards industrial expansion and restrained in the use of motor vehicles. If such moderation in household consumption would appear to be necessary on any large scale, in order to limit greenhouse gas emissions and diminish the risks of climate change, how could it be accomplished? In her study on consumption and social stratification, Aarts in collaboration with Goudsblom (NRP project no. 851038-3) lists four general approaches. Economic policy, legislation and enforcement, and environmental information feedback combined with moral appeals are three of
1210 them. The fourth one involves encouraging spontaneous shifts in social behaviour through increasing the status and prestige derived from voluntary self-restraint in consumption. Aarts's project is focused on the latter strategy. Via in-depth interviews and questionnaire surveys she has investigated if and how moderate and self-controlled lifestyles among higher-income, higher-status groups (who have a choice) may 'trickle down' into environmentally less aware and more consumptive segments of society. "Affluence is an i m p o r t a n t condition and point of d e p a r t u r e for moderation," says Aarts, "as it liberates people from the struggle for subsistence and increases their opportunities to plan ahead." Moderation may enhance social prestige, because it reveals the 'distinctive' ability to restrain oneself, and it therefore may also be sought after by other population groups. Thus moderation and the 'new' social status it provides may trickle down into society as a whole and ultimately lead to the strong collective awareness of environmental processes and effects, which is needed to achieve an 'ecologisation' of society. The s t u d y by Aarts reveals t h a t among the higher-income groups, the better-educated are more sensitive to the environmental effects of consumption. They eat healthier food and less meat, buy more bottled instead of tetra-packed milk, save more fossil-fuel energy, and are more aversive to producing waste, than the average lesser-educated and/or lower-income group member. However, better-educated higher-income group members show little restraint in cultural and holiday travelling (often by car or airplane), which they see as being socially prestigious par excellence and as therefore putting the prestige from self-restraint into its shadow. C o m m e n t s o n Aarts's p r o j e c t This research capitalises on the importance of social comparison processes, in which individuals continually try to identify themselves among others, in terms of income, talent, education, power, and status. A basic principle of social comparison theory (Suls et al., 1997 and Masters et al., 1987) is t h a t people feel most comfortable when they are just a little better (off) - in various respects - than other people in their social environment. Thus, striving to reach a position of 'being slightly better (off)' is an almost daily activity for most everyone. And if 'being better (off)' is largely defined in terms of material consumption and possessions (Dittmar, 1992), then we have a powerful explanation for permanent consumption growth. Aart's thesis is that a re-definition of 'being better (off)' is necessary and seems feasible, w h e r e b y m a t e r i a l c o n s u m p t i o n m a y be reduced for the improvement of environmental quality.
One r a t h e r tricky and perhaps depressing property of the social processes described by Aarts is that new 'distinctive' and 'prestigious' forms of (moderate) consumption among high-status groups arise only when the old, to-be-rejected consumption behaviours are experienced as 'too common' and therefore no longer status-giving. This and other considerations lead Aarts to conclude t h a t other strategies such as economic measures, legislation and normative appeals should also be relied on, in order to bring about environmentally sustainable patterns of h u m a n consumption.
1211 8.
CO2 E M I S S I O N S R E D U C T I O N VIA L I F E S T Y L E C H A N G E S
Nicely complementary to the previously discussed projects by Schmidt, Aarts and Spier is the energy-technological study by Vringer, Wilting, Biesiot, Blok and Moll (NRP project no. 852086). This project team from the universities of Utrecht and Groningen has first developed and refined an input-output energy analysis (IOEA) methodology for determining the direct and indirect energy requirements of various household consumption patterns. In their first subproject, indirect energy use for consumption was determined by assessing the cumulative energy intensity, i.e., the total amount of energy needed for one financial unit of production, of 56 Dutch production sectors. This measure, of course, also includes the energy intensity of exported goods, but with the help of available export statistics it can be corrected for this. Also needed, then, are estimates of the cumulative energy intensity of imported goods, which constitute part of total domestic consumption. The primary aim of Vringer et al.'s project is to analyse if and how CO2 emissions can be reduced by changing lifestyles. Thus, in their second subproject, the authors attempted to identify and describe distinct 'lifestyles' in terms of systematically different patterns of household consumption in regard to their energy intensity and 'CO2 content'. Lifestyles are identified by correlating income, time budget and consumption variables with total household energy requirement. The latter was assessed by d e t e r m i n i n g the energy intensities of about 350 consumption categories and combining these using data from a recent national household expenditure survey. In a third subproject, the consequences of possible future technological developments on the energy intensities and the CO2 content of lifestyles are being modelled via a scenario approach. Some substantive results from the first two subprojects are as follows. Although the energy intensity per financial unit of production has significantly declined between 1969 and 1988 in most sectors of the Dutch economy, the embodied energy of total production has increased. This is because the growth of production has dominated the decline in energy intensity. Between 1969 and 1988 the embodied energy of Dutch exports has exceeded t h a t of imports; The Netherlands now is a net exporter of embodied energy of materials and products. This partly explains the rise in total Dutch energy consumption in this period. While m a n y energy conservation programmes are focused on reductions in direct energy consumption, several production sectors and the households have a higher indirect (embodied) than direct energy consumption. Therefore, reducing indirect energy consumption should get more attention in government energy policy. As far as households are concerned, their total average energy demand in 1990 was 240 gigajoules, of which 46% consisted of direct energy consumption (for heating, lighting and car fueling) and 54% reflects indirect energy use (as embodied in m a t e r i a l s , goods and services purchased). There appears to be a strong relationship between household expenditure and total energy demand, expenditure level being strongly correlated with net income. One additional factor is household size; one-person households use significantly less energy t h a n households consisting of two or more persons. Large differences in energy intensity were observed among different consumption categories, as well as among households
1212 having the same expenditure level. This is indicative of the fact t h a t lifestyle changes may result in significant reductions in energy use and CO2 emissions.
C o m m e n t s on Vringer et al.'s project The IOEA-methodology is an important development in energy analysis research. A tool like IOEA is indispensable if one wishes to assess the cumulative energy intensity of household consumption and to identify energy-relevant differences in lifestyle. For identifying and distinguishing lifestyles themselves, however, household-economic and demographic data, such as income and household size, m a y be greatly insufficient. Despite the fairly strong correlation between household income and total energy demand, there appear to be relatively energy-intensive low-income households as well as energy-thrifty high-income households. This is obviously due to different patterns of expenditure. Would these be explained by differences in 'lifestyle'? How could lifestyle be independently defined and assessed? And to what extent would the lifestyle of an household be d e t e r m i n e d by personal or family-cultural factors on the one hand, and by situational factors inherent in that household's physical and social environment on the other? At this point one would like to see research inputs from sociology, social psychology and cultural anthropology, where differences in lifestyle have been the subject of study for quite some time already. The lifestyle subproject on energy and CO2 emissions reduction is still under way (until mid-1995). By means of scenario construction possible changes in industrial practices and consumer behaviour will be modelled and evaluated with regard to their consequences and implications for fossil-fuel energy consumption.
9.
LOW-ENERGY, LOW-CO2 EMISSIONS SCENARIOS
Related r e s e a r c h is being conducted at the University of Groningen in a multidisciplinary project by Kamminga, Slotegraaf, Van der Veen and others (NRP project no. 852085), on the social significance, feasibility and acceptability of low-energy, low-CO2 emissions scenarios for The Netherlands. Here, recent macro-economic scenarios for the development of the Dutch economy in an international context formed the starting point (CPB, 1992). The investigators argued that prospective modelling by macro-economists insufficiently indicates the meaning and implications of the relevant scenarios for various social and economic groups in society, and that their acceptability as possible futures is empirically unclear. Also, the r e s e a r c h e r s wished to explicate the a s s u m p t i o n s and presuppositions underlying the scenarios and to inspect the way in which predictive elements of the scenarios - such as, e.g., employment rates or energy price levels - had been handled. As a beginning, the project team of sociologists, economists, psychologists and environmental scientists has critically evaluated the CPB scenarios. These had been published by their composers as an explorative means to shake up mental maps of policy makers and to provoke and guide public debates about the future of socio-economic life in The Netherlands. The CPB scenarios were designed on the basis of three different views on economic development, viz. the equilibrium perspective, the co-ordination perspective and the free market perspective. A next
1213 step in their elaboration was a comparative-strength analysis of seven different economic regions of the world. Finally, analyses were made of seven long-term trends, such as population growth, environmental qualities, world food supply and international co-operation. Eventually three different scenarios for the Dutch economy in international context emerged: (1) 'Balanced Growth', an optimistic scenario, (2) 'Global Shift', a pessimistic scenario, and (3) 'European Renaissance', a crisis-overcoming scenario. All three scenarios involved policy measures and expected effects with regard to energy, housing, agriculture, industry, transport and health care. None of the scenarios, however, clearly stood out as a 'low-energy, low-CO2 emissions scenario'. Because of this, because no formal scenario construction methodology had been followed, and because various assumptions and predictions that had been made, could not easily be validated, the University of Groningen team decided to develop its own policy scenario for a low-energy, low-CO2 emissions future for the Dutch social market-economy. After carefully studying available documents and interviewing relevant experts, the team has constructed an overall package of general energy-savings and emissions-reduction measures for The Netherlands as a whole, plus four subsets of sector-specific packages directed at industry, households, greenhouse horticulture and freight transport. The scenarios and subscenarios involve policy measures such as a general energy tax, energy-savings information campaigns, subsidies for energy-efficient technology, application of energy consumption standards and quota, subsidies for low-energy lighting equipment, promoting efficient transport management, discouraging air transport, and speed limitations for road traffic. In three subsequent empirical studies, the macro- and meso-economic significance and effects, the evaluation and assessment by meso-level social and economic actors and decision makers, and the evaluation and acceptance by micro-level (i.e., household) representatives were investigated by an economist, a sociologist and a social psychologist, respectively. Some substantive results are the following. Economically, the significance of a 'low-energy, low-CO2 emissions scenario' hinges upon: (1) its distributional effects in terms of income, employment and economic growth, (2) its structural effects in terms of new opportunities at the supply side of the Dutch economy, and (3) institutional changes necessary to support the restructuring and redistribution involved in a sustainable economic development. Sociologically, it appeared possible to specify the socio-political plausibility of major policy measures reasonably well, via a modelling of key meso-level actors' preferences and power positions in the overall political decision-making process. For example, it turned out to be 'quite probable' that a gradually increasing energy tax for 'small' consumers will actually be introduced, while the probability of significant car-use reduction measures was assessed to be a moderate 50% on the short term. Social-psychologically, it became clear that some 1200 Dutch household representatives are fairly well informed about the global greenhouse effect and judge it desirable that something be done about it. Also, on the average they evaluated a number of household energy-savings measures as reasonably effective and sufficiently acceptable in view of expected changes in quality-of-life. Women, higher-educated persons and non-motorists appeared to find
1214 mobility-directed energy-savings measures to be more acceptable t h a n men, lower-educated persons and regular car-drivers.
C o m m e n t s on K a m m i n g a et al.'s project "Macro-economists tend to see and contemplate things at a high level of aggregation. What certain future events and policy measures actually mean for the people and the organisations concerned, does insufficiently enter their functional range of vision, and so does the potential degree of social acceptability." This critical viewpoint has fruitfully stimulated the investigators to explore the essence, the meaning, the feasibility and the (differentiated) acceptability of socio-economic and energy-savings scenarios for The Netherlands. It is important to learn t h a t this macro-economic scenario construction by the Central Planning Bureau and associated institutes (CPB, 1992) was not based on formal concepts and an explicit methodology. It was a drawback for the project team to note that a significant 'low-energy, low-CO2 emissions scenario' was not available at the outset. But then it t u r n e d out to be highly instructive to go around energy documents and experts in an attempt to compose one's own package of feasible energy-savings measures. And it is worth-while to learn that 'social acceptability' has a different meaning for an economist (thinking about distributional effects), a sociologist (thinking about the preferences and political influence of meso-level actors) and a social psychologist (thinking about changes in quality-of-life and people's potential for adaptation). The multidisciplinary co-operation which has been established, needed its time to develop. A period of two years may be too short for operationalising the original research plan, getting the team to function effectively and to conduct the field research neccessary to test your hypotheses and underpin your conclusions. With a little more manoeuvring space this multidisciplinary project could have yielded even more useful and interesting results. For example, the separate evaluation of sector-specific packages of policy measures by meso- and by micro-level actors could have been extended from the households to all four sectors covered in the scenario design phase. Also, a further differentiation of socio-economic sectors could have been made, in order to obtain a more comprehensive picture of policy m e a s u r e s and their acceptability. The innovative thing about this project is its basic approach of exploring and describing lower-level social effects and responses related to energy-relevant conditions and policy measures, and of subsequently a s s e s s i n g t h e i r social acceptability in t e r m s of economic, sociological and social-psychological considerations. Such an approach m a y c o n s t i t u t e an important counterpart to macro-economic efforts at 'scanning the future'. 10. L I F E S T Y L E S AND D O M E S T I C E N E R G Y C O N S U M P T I O N A final consumption and lifestyles project is carried out at the universities of Tilburg and Eindhoven, by Breemhaar, Van Gool, Ester and Midden (NRP project no. 853119). Here, an exploration is made of the measurability of the concept of lifestyle which appears to be somewhat difficult to define. Also, an attempt is made to specify 'sustainable consumption patterns' with regard to household energy use. B r e e m h a a r et al. seek to define 'lifestyle' in terms of means-end chains, i.e., hypothetical strings of a particular consumer product, its perceived attributes, the
1215 consequences associated to the attributes and the basic (implicit)values t h a t are ultimately served when a consumer experiences those consequences. For example, a sports bike (a means) m a y be perceived as light, sturdy and dependable (its attributes), so t h a t one m a y reach a not-too-distant destination fast, without hassles and along a quiet route, while having some exercise (the consequences), all of which is valued for its goal-effectiveness, 'naturalness' and healthiness (the ends). Such cognitive means-end chains are assessed via in-depth interviews with consumers. A 'laddering technique' is used to lead r e s p o n d e n t s along the hypothetical links in a means-end chain. Such chains are likely to be different for different products. They may also be different for different domains of consumer behaviour, such as feeding, clothing and transportation. And means-end chains may be categorised into distinct groups which are characteristic of different groups of consumers. The authors' p r i m a r y research question reads: "Is it possible to group means-end chains concerning a p a r t i c u l a r behavioural domain with r e g a r d to energy consumption, and are the groups interpretable as lifestyles concerning energy consumption?" An answer to this question is being sought via consumer interviews on means-end chains regarding home-work commuting, home heating, living-room lighting and using a freezer, a washing machine and a washing-dryer. Through content analysis and cluster analysis, the investigators arrive at graphical r e p r e s e n t a t i o n s of adjectives describing attributes, consequences and values associated to particular consumer behaviours. Their project s u m m a r y elsewhere in this volume contains the example of home-work commuting, based on interviews with a small n u m b e r of respondents. Here, it appears t h a t 'motorists' could be clearly d i s t i n g u i s h e d from 'cyclists', and t h a t the general as well as the commuting-specific context variables were differentially clustered for these groups. Results for the other types of consumer behaviour are still being analysed. As 'lifestyles' may be strongly context-dependent, the researchers are also probing into the relationship between observed clusters of (personal) means-end chains a n d (more collective) clusters of context v a r i a b l e s for c e r t a i n groups of respondents and for given domains of consumer behaviour. They state t h a t "it is difficult to conclude whether or not the similarities in classification of respondents on the basis of their means-end chains and on the basis of context variables constitute a causal relationship." Another unresolved issue is the generality of clusters of consumption consequences and consumer values across different types of household consumption. For example, in what way and to what extent would the goal-effectiveness, 'naturalness' and healthiness of the bicycling commuter also show up in his or her means-end chains for living-room lighting, home heating and using electric household machinery? C o m m e n t s o n B r e e m h a a r et al.'s p r o j e c t
This research is methodologically explorative and it proves to be labour-intensive. D e t e r m i n i n g means-end chains for specific consumption behaviours requires a s e r i o u s a n d r e l a t i v e l y long i n t e r v i e w w i t h a t t e n t i v e r e s p o n d e n t s . Content-analysing recorded responses and cluster-analysing coded elements of means-end-chains demands sophisticated data analysis and careful interpretation of results. Someone's 'lifestyle' may emerge as a certain clustering of means-end chains across different types of a consumer's behaviour. A group of consumers
1216 sharing a particular lifestyle may show up, when it appears that their generalised means-end chains are sufficiently similar, in contradistinction from other groups of consumers who cherish other 'lifestyles'. This seems much to expect, and researchers m u s t have some luck to obtain the commonalities underlying the lifestyle concept. Too much differentiation of lifestyles with regard to types of consumer behaviour and/or with regard to subgroups of consumers, would weaken the use of any concept of lifestyle. Also, too much emphasis on cognitive elements such as perceived attributes, consequences and values, may detract from the policy relevance of the lifestyle concept ("de gustibus non est disputandum"). Finally, it m u s t eventually become clear to what extent lifestyles are person- or household-specific , and to what degree they depend upon characteristics of a consumer's physical and social context. The present project is still under way, until mid-1995. Since not all data have yet been analysed and a full report is not yet available, it is still unclear to what extent lifestyles provide an a p p r o p r i a t e c o n c e p t u a l i s a t i o n of domestic e n e r g y consumption. However, if they do, opportunities exist to alter energy-intensive lifestyles into more sustainable ones. The investigators are continuing their search for 'sustainable consumption' and are attempting to define this concept in terms of patterns of energy-extensive household behaviours. Eventually, such patterns will be presented to small consumer panels for their evaluation. 11. F E A R - A R O U S A L COMMUNICATION
AND
ARGUMENTATION
IN
RISK
For environmental policy in general and for climate policy in particular it is crucial t h a t the risks of global warming be presented such that respondents accept the need for remedial actions. Adequate diagnostic research on environmental change and climate processes is one condition for this. An other condition is effective communication of diagnostic results. To study the effects of problem information on energy-savings attitudes, Meijnders, Midden and Wilke at the universities of Eindhoven and Leiden (NRP project no. 852093) have performed a series of experimental studies. In the first experiment their goal was to observe the effects of fear-arousal and argument quality on people's attitudes toward purchasing 'a new type of energy-saving light bulbs'. Four experimental conditions were created by crossing a low- versus high-fear arousing problem-information variable with a weak versus strong purchase-argument variable. Low-fear information was given in a concise description of global warming; in the high-fear condition this information was supplemented with photographs illustrating potential w a r m i n g effects. W e a k versus strong a r g u m e n t quality was varied via selection of arguments previously rated for their 'convincingness'. On the average, the four groups of 19 subjects each (inhabitants of Eindhoven) reflected no overall (main) effects of fear level and of argument strength on their a t t i t u d e s towards purchasing the new light bulb. In the low-fear information condition, however, the average subject's attitude proved to be more favourable after strong arguments' presentation than after weak arguments. At the same time subjects, in a 'thought-listing' task, responded by giving more issue-relevant responses to the high-fear message t h a n to the low-fear message. These partly
1217 unexpected results are provisionally interpreted as a possible suppression, in the high-fear condition, of systematic information processing. The authors generally conclude t h a t "fear m a y have a positive effect on (people's) motivation to elaborate relevant information, but at high levels of fear, this positive effect may be overruled by a negative effect on information processing capacity". This project is being continued and therefore results of further experimentation are still to become available. C o m m e n t s on M e i j n d e r s et al.'s p r o j e c t In view of apparently serious problems of climate risk communication, one may wonder what conclusions and recommendations would emerge from the voluminous literature on fear arousal and information processing in the face of risk. So far in this project the impression is given that almost exclusive reference is made to the social-psychological literature, and not to the many chapters and articles on 'risk communication' t h a t have appeared since the mid-1980s, in several books on technological risk management, and in journals like Risk Analysis, J o u r n a l of Communication and Journal of Social Issues. Against the background of much of t h a t literature the question arises whether the sort of fear-arousal, the kind of argumentation and the type of 'action' used in this project's first experiment may hit the point hard enough. Methodologically, this study has been designed and conducted in a convincing manner, about which it is enjoyable to read. The theoretical basis of the project is interesting and important, but it could be broadened so as to incorporate sensitive elements from the multidisciplinary debate on technological-risk communication. For a policy-supporting research p r o g r a m m e like NRP the question is w h e t h e r theory-directed, high-quality e x p e r i m e n t a t i o n will indeed yield the sort of useful results the p r o g r a m m e committee is hoping for. Perhaps a more daring kind of field experimentation, based on a multidisciplinary effort to formulate problem information, select type of communication and design environment-protective actions, could provide the kind of conclusions and recommendations that would be both theoretically justified and practically useful.
12. SOCIAL W E L F A R E AN E N V I R O N M E N T A L QUALITY "The currently dominant idea of material welfare is at odds with a lifestyle t h a t does justice to basic h u m a n values. M e a s u r e m e n t of welfare needs to attach importance to a good environment." This dual thesis forms the starting point of a non-NRP project which fits into the debate on sustainable consumption and lifestyles, conducted by Mentzel at the University of Leiden (see Table 1.1.). The author critically reviews the concept of social welfare as used by economic policy makers and he puts this in contrast with perceived well-being and quality-of-life as experienced by individuals. 'Economic' welfare is expressed in terms of ownership and consumption of material goods and of access to high-energy activities such as in transport. Aggregate economic welfare is quantified into a country's gross national income (GNI), and economic growth in terms of GNI is believed to be crucial. It is becoming clear that material economic growth is damaging basic environmental qualities and ultimately threatens the earth's life support systems. Therefore, particularly in the industrialised consumer societies of the northern hemisphere, a search for a 'sustainable lifestyle' is necessary. This, says Mentzel,
1218 should cover the main spheres of life: at home, in the shopping-mall, in the workplace, in transport and traffic, and in recreational activities. To delineate what is needed, a reconceptualisation of what we mean by 'the good life' is required, as well as empirical research yielding people's own conceptions and dimensions of welfare and quality-of-life. Various empirical studies on perceived well-being and quality-of-life have been performed which reveal basic dimensions of perceived welfare. For example, an important empirical dimension appears to be the capacity to control and consciously direct one's own living conditions. 'Having', 'loving' and 'being' are useful central labels for characterising essential conditions for human development and existence. Good health is highly valued in present-day society, while societal improvements are being sought in better interpersonal relationships and a higher quality of the natural environment. The author keenly puts his finger on the role of national and international institutions by which socio-economic developments towards sustainable consumption and production patterns are to be promoted and co-ordinated. Two questions are sensitive here: how to arrive at a just (re-)distribution of welfare, and how to increase the socially perceived, and thus (also) the political value of nature and its resources. C o m m e n t s on Mentzel's project
The problem of unsustainable economic growth and ecologically unbridled consumption (see also Schmidt's, Spier's en Aarts's projects above) necessitates a fundamental reconsideration of classical notions of welfare and quality-of-life. To conduct this debate in a fair manner, it seems useful to keep in mind that the currently dominant concept of economic (i.e., material) welfare is rooted in people's natural motivation to be safeguarded against poverty, discomforts and diseases. It is the 'overshoot' of an economic system originally designed to meet such essential h u m a n desires, which has put many (though by far not all) of us up with a consumer society where personal satisfaction, power and prestige have become strongly associated to material possessions and consumption. The critical goal of sustainable development, therefore, should be a set of economic (i.e. material) conditions which could be considered 'sufficient' and 'fair'. In a shortlist of recommendations for a sustainable lifestyle, Thoenes (1990) indicates the necessity of, among other: a guaranteed satisfaction of basic needs, the creation of a basic social equality for everyone, and expansion of possibilities for energy- and material-extensive behaviours. Official present-day economic views of welfare are not as materialistic as Mentzel seems to suggest. The Organisation of Economic Co-operation and Development (OECD, 1982), for example, considers productivity, employment rate, purchasing power, balance of payments, government deficit and rate of inflation as basic economic indicators. The OECD recommends, however, that governments also pay attention to such dimensions as health, education, work conditions, social life, and the quality of public environmental goods such as air, water and natural areas. In a systematic review of social indicators research, Henderson (1994) searches for new indicators of wealth and progress and for changes in the meaning of 'development'. For example, for some time already the city of Jacksonville in Florida, U.S.A. evaluates its 'progress' in terms of nine categories of indicators, viz. 'the economy', public safety, health, education, natural environment, mobility (transport), government/politics, social environment and culture/recreation.
1219 According to Henderson, a clarification of the confusion of m e a n s (e.g. material consumption, economic growth) with essential e n d s of human development is badly needed. It would seem that this could best take place in a public debate among policy makers, supported by relevant specialists from social philosophy, economics, sociology, psychology and cultural antropology. 13. T O W A R D S S U S T A I N A B L E H O U S E H O L D M E T A B O L I S M
A final project deserving attention is funded by N.W.O., the Netherlands Organisation for Scientific Research. In 'HOMES: HOusehold Metabolism Effectively Sustainable', Schoot Uiterkamp at the University of Groningen co-ordinates a multidisciplinary group of environmental scientists, economists, spatial scientists, social psychologists and administrative scientists. Since early 1994 these investigators first of all attempt to diagnose and explain developments and trends in household consumption between 1950 en 1990 and as far into the future as 2030. Because household consumption encompasses a multitude of goods, services and activities and therefore must be delineated, the project's focus is on housing and transportation, home heating and lighting, and durable household equipment, whereby a distinction is made between strategic (mostly: purchasing) decisons and the operational use of electricity, water and different fossil fuels for daily activities. Secondly, the project group is determining the environmental impacts of household consumption and assessing its (un)sustainability, both in terms of descriptive variables such as various kinds of resource use and waste materials and in terms of subjective judgements collected from household representatives. Thirdly, the HOMES team will systematically analyse and describe possible technical as well as behavioural options and strategies for changing household consumption such that it may be considered 'sustainable' in the long run. To this end, specific technical options and behaviours will be considered, and consumption patterns and lifestyles will be designed and evaluated in collaboration with consumer groups and individuals. Also, various different policy strategies for encouraging households to adopt sustainable consumption patterns will be described and evaluated for their potential effectiveness. By doing all this in a multidisciplinary fashion, the HOMES team aspires to cover and integrate the physico-chemical and the technical aspects and possibilities of household consumption as well as the social and behavioural components and opportunities for sustainability. The project as a whole is to be concluded in 1998. Comments on Schoot Uiterkamp's project 'HOMES' is a problem-oriented, multidisciplinary endeavour to assess and understand household metabolism and to indicate ways and means for modifying this into a sustainable direction. Such an approach is explicitly stimulated by N.W.O. (see above) which - in its priority research programme on 'Sustainability and environmental quality' (1993-1998) funds altogether three such pluralistic projects (the other two deal with five major metal flows through the economy and with international river basin management, respectively). Considering the social and economic opportunities and motives for household consumption, looking into its relation to demographic developments and to physical infrastructure and government policies, and charting its various environmental effects, requires wide-ranging exploration and assessment as well strong co-ordination and overall
1220 modelling. Furthermore, household metabolism and industrial metabolism are strongly interwoven (see Section 3 above). Hence both a diagnosis of current consumption and the design of sustainable metabolism would sooner or later have to cover both the households and various relevant production sectors of the economy. The la tt er perspective is already taken in K a m m i n g a et al.'s NRP-project on 'low-energy, low-CO2 emissions scenarios' (see Section 8 above), and it is also adopted in a newly started NRP phase II project on emissions reductions via lifestyle changes by Biesiot (Groningen), Blok (Utrecht) and others, which links up directly with the environmental-science subproject of HOMES. Whether the broad-ranging and ambitious HOMES-project will succeed is a matter of prudent delineation, effective co-operation, personal enthusiasm and some luck in designing and conducting data collection and overall modelling of results. A scientifically hazardous approach like this, however, seems badly required for u n d e r s t a n d i n g costly and harmful developments in society and for designing sustainable patterns of social and economic behaviour. 14. GE N E R A L CO N C L US I O N S, RECOMMENDATIONS
SUGGESTIONS
AND
14.1 'Economisation' and e n v i r o n m e n t a l e x p l o i t a t i o n The NRP-research on consumption and lifestyles conducted so far has been a mixture of social-science, technological and environmental-science studies. These investigations have yielded important data and conclusions about societal motivations, developments and trends about consumption and lifestyles. By virtue of this, an interesting and useful picture of 'household metabolism' is emerging. During several decades now, strong increases in consumer purchasing power, in technological potentialities and in the market supply of a great variety of products, services and facilities, have met with social-cultural changes in individual and social motives of consumers. 'Economisation' has gradually led consumers, who are always partly driven by producers and advertisers, to adopt or aspire a prevailing lifestyle of high-quality material possessions and facilities, and of fast, short-term consumptive behaviours, whereby some basic goal or sense of life is easily obscured. For a long time the economic system of western industrialised countries has been truly successful in combatting human poverty, ignorance, discomfort and diseases. In recent times, however, it seems to be overshooting its original goals and to be developing into an energy- and material-intensive monster which gradually eats up its own existential conditions and seriously diminishes various qualities-of-life. The 'modern consumer lifestyle' inherent to this system is increasingly expansive, mobile and environmentally harmful. Many technical options seem to be available for increasing energy- and materials-efficiency and for reducing the amount and variety of household waste. But the social implementation of these, as well as the possible occurrence of 'unsaving' substitution behaviours and further consumption growth, deduct from the environmental effectiveness of technology. Therefore, behaviour change and particularly 'moderation' are becoming the key words for policy makers who - on behalf of society as a whole - are trying to steer away from unsustainable household consumption.
1221
14.2 Some policy recommendations In their co-ordinated project summary Goudsblom, Aarts, Schmidt and Spier (NRP project no. 851038, see Sections 4-6 above) present a number of useful conclusions a n d policy r e c o m m e n d a t i o n s . For example, as ' i m p o r t a n t obstacles to ecologisation' Goudsblom et al. mention: the strong social pressures to produce, inherent to industrial market regimes; the constantly rising productivity of labour as a result of competition; the fact that economic growth also is to create, or at least maintain, sufficient job opportunities; and the boosting effect on consumption of the s t a t u s h i e r a r c h y in industrial m a r k e t regimes. After listing various 'facilitating conditions for ecologisation', these authors also provide a n u m b e r of policy recommendations, for example: to use and exploit the s t a t u s motive in environmental policies, for instance, by associating social prestige to energy- and material-extensive behaviours; to stimulate further research into fossil-fuel energy savings techniques and their social implementation; to utilise the m a r k e t mechanism through levies, taxes and subsidies for an 'ecologisation' of production, commerce and consumption; and to develop specific campaigns with regard to car-driving, holiday air travel, meat consumption and other energy-intensive social behaviours. Goudsblom et al.'s conclusions and recommendations fit in with the fifth policy direction: 'institutional and cultural change', that emerged from a multidisciplinary and multi-party series of specialists' workshops conducted by Klabbers (Nijmegen) and Vellinga (Amsterdam); see (Klabbers et al., 1994). This strategic policy option came out of intense deliberations as one that might be inevitable to select if it would appear t h a t 'no regrets', 'least regrets', 'acceleration' and 'technological innovation', the other four policy directions, are not effective enough to secure a sustainable development of society. Some focal policy actions under 'institutional and cultural change' would be: to initiate a debate on improving the quality of society; to intensify the care for residential environments; to encourage consumers to select higher-quality food products, to buy local products and to follow local cuisine; to promote active participation in cultural activities; and to use trendsetters as examples of behavioural change. Such a view also links up with recent ideas about a 'greening of the economy', about which a key author remarks t h a t "we should be aiming to maximise the welfare obtained from economic activity while minimising the volume of matter and energy which flows through the economy" (Jacobs, 1991, p. 114).
14.3 Lifestyles and behaviour change strategies In none of the NRP-studies conducted so far has the concept of lifestyle been given a theoretically convincing and methodologically sound definition. Therefore the notion of lifestyle is up for further improvement and operationalisation to support continuing research aimed at delineating sustainable lifestyles. It would seem that such research should be interdisciplinary in nature; consumption patterns might be defined as 'lifestyles' to the extent t h a t they meet certain sociological, economic-psychological and ecological criteria. Candidate variables for these are family background, education, income, energy consumption, amounts of waste, degree of mobility, major life goals, habits and attitudes toward energy-savings, and appreciation of nature and natural living conditions. In a t h r i v i n g consumer society, changes in lifestyle or in prevailing social behaviours, in order to achieve energy savings and emission reductions, are hard to
1222 explain and to bring about. It is an underestimated problem that such changes need to rest upon a sufficient awareness of environmental problems, that they cannot occur without the availability of feasible behaviour alternatives, that policy instruments for inducing behaviour changes may, if wrongly selected and tuned-in, be ineffective or even counter-productive, and that the subject of the desired behaviour changes to whom the policy instruments are applied, wants to have an idea of 'what the future will bring'. The social and behavioural sciences have much in store to clarify this problem and to support the design of effective policy instruments. A major kind of conceptual tools are models for analysing and explaining consumer behaviour. Attitude- intention-behaviour models (Ajzen, 1991), o r i e n t a t i o n - p u r c h a s e - u s e - d i s c a r d models (Van Raaij, 1994), and motivation-opportunity-ability models (Oelander et al., 1994) have proven to be suitable means for coming to grips with consumer behaviour and its potential for modification. 14.4 I n t e r n a t i o n a l r e s e a r c h efforts Household consumption, its stimulation by and its implications for various production sectors, and its gross environmental impact in terms of resource use, land exploitation and waste, is a topic of increasing international interest. For example, Oelander and colleagues in Denmark, with the support of the Danish government, are conducting a multidisciplinary project on ' u n d e r s t a n d i n g consumer behaviour as a prerequisite for environmental protection' (Oelander, 1994). At the International Institute for Applied Systems Analysis in Laxenburg, Austria, Nakicenovic and colleagues are conducting and forming a network around their 'Environmentally Compatible Energy Strategies Project' (Nakicenovic et al., 1994 and Grfibler, 1991).So do Schipper and colleagues at Lawrence Livermore Laboratories in Berkeley, California (Schipper et al., 1989 and 1992). Also in the U.S.A., Stern [see Stern, 1994 for a review] has long investigated the psychological determinants of energy- and material-intensive behaviours and possible strategies for achieving environment-saving behaviour.
In The Netherlands, the Netherlands Energy Research Foundation in Petten has organised and published the results of several national workshops on 'lifestyle and energy consumption' (Perrels 1993 and 1994), where various motives, types of behaviour and strategies for behaviour change have been critically discussed. Again, it appeared that technical options are to be supplemented with behavioural options, and that the acceptability of any behaviour changes significantly depends upon their feasibility and their (perceived) environmental effectiveness. ECN-editor Perrels also recommends a multidisciplinary a t t e m p t at better defining the concept of lifestyle, and to carefully consider what different types of actors in society (e.g., consumers, producers, retailers, utility companies and government policy makers) actually do and could do to stimulate energy- and material-extensive behaviour patterns. Like the Dutch National Institute for Public Health and Environmental Protection RIVM (1991) in its National Environmental Survey 1990-2010, Perrels (1994, p. 73) also concludes that, in order to arrive at sustainable household metabolism, our cherished concept of economic growth may have to be differently filled-in (i.e., rather more qualitatively than quantitatively) and that international re-distribution of economic potential and wealth would be important for realising world-wide sustainability. Vivid stimulation of international comparative studies on consumption and lifestyles is
1223 to be expected from the H u m a n Dimensions of Global E n v i r o n m e n t a l Change programme (HDP), initiated by the International Social Science Council in Paris. In HDP's Work Plan for 1994-1995 (HDP, 1994) household metabolism is not m e n t i o n e d as such, but energy consumption, household r e s o u r c e use and individuals' attitudes and behaviours towards the environment are somehow incorporated in several 'major research areas', such as 'industrial transformation and energy use', 'demographic and social dimensions of resource use' and 'public attitudes, perceptions, behaviour and knowledge'. One problem with the HDP research programming so far, however, seems to be the predominance of general explorative questions as contrasted with specific research hypotheses about reasonably delineated (potential) policy issues. Another problem, it would seem, is the r u d i m e n t a r y development of an interdisciplinary perspective on global e n v i r o n m e n t a l change, whereby component research tasks might be usefully allocated in a multidisciplinary fashion. 14.5 F u r t h e r N R P i n t e n t i o n s In the second phase on the NRP (1995-2001), the problem of energy- and m a t e r i a l - i n t e n s i v e household consumption and the search for s u s t a i n a b l e consumer lifestyles remain high on the programme's agenda. Study topics are, for instance: the relationship between (total) household metabolism, population development and trends in household formation and household activity patterns; the identification and explanation of 'unsustainable' lifestyles; and methods and i n s t r u m e n t s for designing and implementing 'sustainable' consumption patterns. Investigations concerning household consumption and lifestyles will be deliberately linked with studies on climate-problem awareness and with research on personal mobility and the diverse use of motor vehicles (see the review chapter on mobility and transport, elsewhere in this volume). Also, the NRP committee will promote international co-operation and exchange of ideas and research findings, as a way to improve i n t e r n a t i o n a l u n d e r s t a n d i n g and policy making regarding household metabolism. 14.6 R e l i g i o n s on c o n s u m p t i o n To conclude this review chapter on culture, consumption and lifestyles in view of global environmental change, it may be appropriate to cite Durning (1992) who after documenting, characterising and criticising western-industrial consumption styles much like Vermeersch (1990) does - provides a t a b u l a r overview of ideological statements on h u m a n consumption and wealth, as derived from nine major world religions. The Buddhists, for example, profess that "who in this world transcends his selfish desires, his worries drop from his shoulders as dew-drops from a lotus flower". The Hindi like to say: "He who is fully free of desires and without craving .. reaches peace". The Christians cherish their biblical quote: "It is easier for a camel to go through the eye of a needle than it is for a rich man to enter the kingdom of God", while the Muslims repeat after their prophet Mohammed: "Poverty is my pride". But perhaps the most applicable statement in view of the present review comes from the Confucianists: "Excess and want are equally bad". Would 'the middle way' be truly sustainable?
1224 15. R E F E R E N C E S
Ajzen, I., 1991. The theory of planned behavior. Organizational Behavior and Human Decision Processes 50: 179-211. Corson, W.H., 1994. Changing course: an outline of strategies for a sustainable future. Futures 26: 206-223. CPB: Centraal PlanBureau, 1992. Nederland in Drievoud. Een scenariostudie van de Nederlandse economie 1990-2015. SDU Uitgeverij Den Haag. Dittmar, H., 1992. The social psychology of material possessions. Harvester Wheatsheaf, Hemel Hempstead, U.K.; St. Martin's Press, New York. Durning, A.T., 1992. How much is enough? W.W. Norton Company, New York~ondon. (In Dutch: Hoeveel is genoeg? De konsumptiemaatschappij en de toekomst van de aarde. Pauli Publishing, Worldwatch Institute Europe, Berlaar Belgium). Ehrlich, P.R. and Holdren, J.P., 1971. Impact of population growth. Science 171: 1212-1217. Goodland, R., Daly, H. and Kellenberg, J., 1994. Burden sharing in the transition to environmental sustainability. Futures 26: 146-155. Grfibler, A., 1991. Energy in the 21st century: from resource to environmental and lifestyle constraints. Entropie 164/165: 29-33. HDP-committee, 1994. H u m a n dimensions of global environmental change programme. HDP Work Plan 1994-1995. Occasional Paper no. 6. Paris: International Social Science Council at UNESCO. Henderson, H., 1994. Paths to sustainable development; the role of social indicators. Futures 26(2): 125-137. Jacobs, M., 1991. The green economy: sustainable development and the politics of the future. Pluto Press, London. Klabbers, J., Vellinga, P., Swart, R., Van Ulden, A. and Janssen, R., 1994. Policy options addressing the greenhouse effect. NRP, Bilthoven, The Netherlands. Masters, J.C. and Smith, W.P., (eds), 1987. Social comparison, social justice and relative deprivation: theoretical and policy perspectives. Volume 4. Hillsdale (N.J.), Erlbaum. Nakicenovic, N., Nordhaus, W.D., Richels, R. and Toth, F.L., Eds, 1994. Integrative assessment of mitigation, impacts and adaptation to climate change. International Institute for Applied Systems Analysis Laxenburg (Austria). OECD, 1982. The OECD list of social indicators. Organisation for Economic Cooperation and Development Parijs. Oelander. F. and Thogerson, J., 1994. Understanding of consumer behaviour as a prerequisite for environmental protection. Keynote address presented at 23rd International Congress of Applied Psychology.Aarhus School of Business, Aarhus, Denmark. Opschoor, H., 1989. Na ons geen zondvloed. Voorwaarden voor d u u r z a a m milieugebruik. Kok/Agora, Kampen. Perrels, A.H., 1994. Slotbeschouwing. In: De Paauw, K.F.B., A.H. Perrels and A.F.M. van Veenendaal, (Red.): Leefstijl en energie; van intentie naar actie? Petten: Energie Centrum Nederland, Rapport ECN-C-94-068. Perrels, A.H. (Red.), 1993. Leefstijl en energie" waar moet dat heen, hoe zal dat gaan.. Een interdisciplinaire kruisbestuiving. Rapport ECN-C-93-049, Energie Centrum Nederland, Petten.
1225 RIVM, 1991. Nationale Milieuverkenningen 2: 1990-2010.RijksInstituut voor Volksgezondheid en Milieuhygiene, Bilthoven en Samson/Tjeenk Willink, Alphen a/d Rijn. (in Dutch). Schipper, L., Bartlett, S., Hawk, D. and Vine, E., 1989. Linking life-styles and energy use: a matter of time? Annual Review of Energy 14: 273-320. Schipper, L. and Meyers, S., 1992. Energy efficiency and human activity; past trends, future prospects. Cambridge University Press, USA. Stern, P.C., 1992. Psychological dimensions of global environmental change. Annual Review of Psychology 43: 269-302. Suls, J.M. and Miller, R.L., Eds, 1977. Social comparison processes: theoretical and empirical perspectives. Hemisphere Publishers, Washington D.C.. Thoenes, P., 1990. Milieu en consumptie: blijft meer steeds beter? In Commissie Lange Termijn Milieubeleid: Het milieu: denkbeelden voor de 21ste eeuw. Kerkebosch, Zeist. Van Raaij, W.F., 1994. Consumentengedrag en milieu. In: Midden, C.J.H. and G.C. Bartels (Red.). Maatschappelijke aspecten van het milieuvraagstuk Bohn Stafleu Van Loghum, Houten. Vermeersch, E., 1990. Weg van het WTK-complex: onze toekomstige samenleving. In Commissie Lange Termijn Milieubeleid: Het milieu: denkbeelden voor de 21ste eeuw. Kerkebosch, Zeist. Vlek, Ch., Hendrickx, L and Steg, L., 1993. A social dilemmas analysis of motorised-transport problems and six general strategies for social behaviour change. In ECMT: Transport policy and global warming. European Conference of Ministers of Transport, OECD Publication Service Paris, p. 209-225. White, L., 1967. The historical roots of our ecologic crisis. Science 155: 1203-1207.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1229
NRP-project LIFESTYLE Reduction of CO2 emissions by lifestyle changes Vringer, K., Wilting, H.C., Biesiot, W., Blok, K. & Moll, H.C. ~* poster presentation International Conference on Climate Change Research, Evaluation and Policy Implications 8 December 1994, Maastricht, The Netherlands
Abstract The aim of the Lifestyle project is to analyse the C O 2 emission reduction potential of lifestyle change. The analysis is carried out by examining the direct and the indirect energy contents of the average Dutch household consumption. An overview of the past developments of Dutch sector energy intensities is produced and its consequences for the average household energy requirement are studied. Also differences in energy requirement related to differences in lifestyle are assessed. Calculations of the Dutch household expenditure survey has resulted in an overview of the energy requirement per income and spending subcategory. The correlations between some relevant household factors are determined and discussed.
Introduction The Lifestyle project succeeds preliminary studies about the direct and indirect energy contents of an average household consumption pattern. The aim of the project is to analyse if and how COz emissions can be reduced by changing lifestyles or by changes within lifestyles. Six research stages are discerned. First, - t o serve this goal- it is necessary to enlarge the scope of the methodology to calculate the energy content of consumption patterns and to improve the quality of data (research stages A1 - A3). Next differences in CO2 emission related to differences in lifestyle and possibilities of lifestyle changes are to be assessed and evaluated on their potential to reduce the CO2 emission (research stages B1 - B3). The six research stages are: A1. An improvement of the input output energy analysis methodology (including CO2 emissions) by correction of possible biases and by an assessment of its scope by application on several generic lifestyles.
# Harry Wilting, Wouter Biesiot and Henk Moll, Centre for Energy and Environmental Studies (IVEM) University of Groningen, Nijenborgh 4, 9747 AG Groningen, The Netherlands * Kees Vringer, Kornelis Blok, Department of Science, Technology and Society Utrecht University, Padualaan 14, NL-3584 CH Utrecht, The Netherlands
1230 A2. A3. B1. B2. B3.
An assessment of structural trends with regard to the energy intensities of economic sectors from a time series of Dutch energy consumption data. Further and deeper analysis of consumption and construction of a database concerning the energy and CO2 content of consumption activities. Identification of different lifestyles by correlating income, time budget and consumption with energy requirement and trend analysis on these lifestyles. Description of the lifestyles in terms of financial and energy/CO2 costs and trend analysis on these costs. Assessment of the effects of possible technological developments on energy intensities and on energy and CO2 content of lifestyles by a scenario approach.
This paper discusses the results of two subprojects: Energy consumption in relation to economic activities, 1969- 1988, addressing the research stages A1 and A2, and The direct and indirect energy requirements of households in the Netherlands, addressing the research stages B1 and B2.
Energy consumption in relation to economic activities, 1969 - 1988
1
Economic activities, production and consumption, are closely related. Production, in fact, occurs on behalf of consumption (exports included). Therefore, the total energy use of an economy can be attributed to the consumption sectors. So, the indirect energy requirements of households, as a consequence of the purchase of goods and services, are not only determined by the consumption patterns of the households, but also by the cumulative energy intensities of the production sectors. The cumulative energy intensity gives for each sector the total amount of energy, direct and indirect, that is needed for one financial unit of production of that sector. We aim to obtain an overview of the developments of the cumulative energy intensities for the Dutch production sectors over a period of twenty years (1969-1988). Besides, we attempt to determine the historic trends of the embodied energy of imports and exports and the indirect energy requirements of households. The cumulative energy intensities of the production sectors are calculated by using inputoutput energy analysis, which makes use of economic input-output tables. These tables, published by the Netherlands Central Bureau of Statistics annually, describe the transactions in an economy in financial terms. To gain insight in the development in the energy consumption of the whole production system, we calculated the cumulative energy intensity of the total production of the economy. Using the energy intensities, we calculated the energy flows in the economy, especially the embodied energy of the imports and the exports, and the indirect energy requirements of the households. The energy data required are taken from the Dutch Energy Statistics.
Results In the period 1969-1988, the cumulative energy intensity decreased for 40 of the 56 Dutch production sectors. 31 sectors showed a decrease by more than 10%. This points to an energy efficiency increase for these sectors. The direct and the cumulative energy intensity of total production decreased both by about 20% in the period 1973-1988.
1231 The energy flows are determined by the energy intensities. Figure 1 shows the embodied energy of the 2500 imports versus the embodied = _ energy of the exports during the = = = 2000 = = period 1969-1988. Since 1971 the embodied energy of the exports has increased more than the embodied energy of the imports. In 1988 the embodied energy of _ ~_ the exports was 28% higher than 500 I the embodied energy of the !/-/ / I / imports. 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 Figure 2 shows the direct and year indirect energy consumption of the Figure 1 Embodiedenergy of imports, for production and households during the period consumption, and exports (PJ). 1969-1988. In this period, in energy c o n s u m p t i o n (P J) which the number of households 2500 [] indirect [] direct increased with about 50% the total energy consumption of the 2000 households grew with about 30%. In 1988, the total energy --- Z -1500 consumption of the households was about the same as the total z z 1000 energy consumption in 1973. The xx • energy consumption per household decreased by about 10%, partly 500 caused by a decline of the number of persons per household. Changes in the indirect energy 69 70 71 72 73 74 75 76 77 7B 79 80 81 82 83 65 86 B7 88 year consumption of the households are Figure 2 Directand indirect energy consumption households caused by changes in the energy in 1969-1988. intensities of the production sectors and volume and structure changes in the consumption patterns of the households. Figure 3 shows the progression of the indirect energy requirements of the households due to the changes mentioned above with regard to the base year 1969. The progress in the indirect energy requirements mainly resulted from volume changes and the improvement of the energy efficiency of the economic sectors. Changes in the structure of the consumption pattern, i.e. shifts in the purchases from production sectors to other production sectors, hardly affect the indirect energy consumption of the households. 3000
embodied energy (P J)
imports (production) [~ exports
'0001 00 !IIi"
[~ imports (consumption)
F
II-!
fli:
t lll
Conclusions Many energy conservation programs consider only direct energy consumption. Several production sectors and the households have a higher indirect energy consumption than
1232 direct energy consumption. Therefore, the indirect energy consumption should get more 160 attention in energy policy. We have found a significant negative correlation between 140 energy prices and energy intensities. A negative correlation 120 means that an increase in energy >~ '--x-- x ""k'J x "~O prices coincides with a decline in 100 -o- - o - _ D . o _ o _ e _ a _ , 3 _ a"'~-"~.--o--o--o--o-'~176 energy intensities. The strongest "x ~ -x ~ -x correlation is between the energy intensities and the energy prices 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 three years before. This is year consistent with the fact that the Figure 3 Indices indirect energy consumption households due to changes in volume and structure o f the consumption, due energy intensities of some energy sectors started to to changes in energy intensities of the production sectors, and in intensive decrease some years after the first total. oil crisis. Sectors need some time to react to energy price changes.According to goods and services, the Netherlands has become a net exporter of energy. This means that the CO2-emission caused by the use of fossil energy in the Netherlands on behalf of other countries is higher than the CO2emission in other countries on behalf of the Netherlands. The Dutch production system has concentrated more on exports. 180
index indirect energy consumption 0969 = 100) - - total
-~ volume
-o- structure --~ intensity
+
The direct and indirect energy requirements of households in the Netherlands 2
One way of reducing C O 2 emissions is to reduce direct and indirect household energy requirements by influencing the household consumption pattern. A household not only uses direct energy in the form of gas, electricity and petrol, but it also uses indirect energy embodied in consumer goods such as food, furniture and services. Before discussing the ways in which the household consumption pattern should be influenced, one needs to have quantitative information about these energy requirements. We aim to obtain an overview of the total energy requirement of households and the energy requirement per consumption category. Also we attempt to quantify the relation of household expenditure, net household income and number of household members to the total energy requirement of households. To obtain an overview of the cumulative energy requirement of Dutch households, we analysed the total consumption package for its cumulative energy requirement. The energy intensities (in MJ per Dutch guilder (Dfl)) of about 350 basic consumption categories are determined, using a hybrid energy analysis method. The energy requirement of Dutch households is calculated by combining the 350 energy intensities with data from the Netherlands Household Expenditure Survey of 1990. This
1233
survey gives the expenditure of 2767 representative households in the Netherlands in 1990. The result is an overview of the total energy requirement of Dutch households. Results The total average energy demand per household in the Netherlands in 1990 was 240 GJ, of which 54% was indirect. Table 1 gives the average energy requirement and energy intensity
of
the
Dutch
households,
Energy requirement (GJ) (% of total) Total
240
100
Indirectenergy requirement
Energy
intensity (MJ/Dfl) 6.3
130
54
3.5
Food
41
17
5.6
categories.
Household effects
19
8
5.5
Figure 4 shows the relation between the total energy requirement and the household expenditure. We give the 10, 25, 50 (median), 75 and 90
Clothing&footwear
3 5 2
2.7
Hygiene
8 12 5
4.1
Education&recreation
24
10
3.0
percentile
Transport& communication
11
5
2.8
110
46
45.0
28 60
12
25
46.5
22
9
22.4
aggregated into 11 main consumption
lines
in
this
figure
to
demonstrate the variance of the energy requirement for the spending subcategories. The 10 percentile line represents the levels for which 10% of the households of the corresponding spending subcategory requires less e n e r g y given by this line.
than
the
level
House
9
Medicalcare
Directenergy requirement Electricity
Heating
Petrol
4
1.4
3.4
57.8
Table 1 Total energy requirement and energy intensity of an average Dutch household in 1990 per main category.
Figure 4 shows - as expected - that the energy requirement increases with household expenditure. But also substantial variance within the spending subcategories is observed: e.g. 10% of the households use 22% less energy than the energy requirement of an average household with the same expenditure. 700 go pete, The relation between energy 75 pete. requirement and net household ..(125%) 50 I)erc. income shows also an ~50025 pete. increasing relationship of the E ..... (100"/,,) net household income and the 10 I~rc, ~400(BIPI.) energy requirement. But, the (TB*/.) variance is larger than the >'300variance shown in figure 5 (9 2oobecause of differences between o income and expenditure. 100In Figure 5 we plot the total energy requirement versus the ~ i'o ~o ~o do ~o ~o #o ~o ~;o 1oo net income for various household expenditure (Dfl x 1000) household sizes to investigate a ..o"
9"
Figure 4 expenditure,
(112"/o)
Total household energy requirement versus household
possible dependence of these factors
apart
from
the
1234 500
dependence related to differences in net income. Figure 5 demonstrates that only a significant difference in energy requirement, independent of the net household income, is observed between one-person households and several-person households (approx. 45 GJ).
450-
.-.400~350-
."5 .... .'.-'-..............2
e-
~30o-
"_~
250-
~200r 150-
o
o
"" 100-
500
0
io
2o
3o
4o
so
6o
~o
80
~
1oo
total net income (Dfl. x I000)
Figure 5 T o t a l h o u s e h o l d e n e r g y r e q u i r e m e n t versus net household income for 1 to 4 household members.
Conclusions Because the indirect energy requirement amounts at least 54% of the total requirement of households, further research is needed into the indirect household energy requirement. Future energy policy must pay attention to the indirect energy requirement of households. The strong relation between income and total energy requirement suggests that, with further increases in income levels, the average household energy requirement will probably rise as well. However, the large differences between the energy intensities of the various consumption categories indicate that the total household energy requirement can be reduced by a change of our consumption patterns.
References Wilting H.C., Biesiot , W., Moll, H.C., Economische activiteiten vanuit energetisch perspectief" Veranderingen in Nederland in de periode 1969-1988, IVEM onderzoeksrapport no. 72, juli 1994, Groningen. Vringer K., Blok, K., The direct and indirect energy requirement of households in The Netherlands, NW&S, 1993, Utrecht
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1235
Life styles and domestic energy consumption: a pilot study B. Breemhaar 1, W. van Gool 2, P. Ester ~, C. Midden2
1Institute for Social Policy Research and Consultancy, Tilburg, The Netherlands
2Eindhoven University of Technology, Dept. of Philosophy and Social Sciences, The Netherlands
Abstract The contribution of households to C O 2 production is still increasing. To alter patterns of energy consumption for example with respect to commuter traffic, using the freezer, and warming the house, changing life styles related to domestic energy consumption is considered. In our study, we have operationalized life style as m e a n s - e n d chains, that link perceived benefits of a particular behavior to basic values that people pursue. In this paper, preliminary results are presented of the study that is aimed at empirically establishing the feasibility of the concept of life style in relation to domestic energy consumption.
1. INTRODUCTION Domestic consumption of energy contributes considerably to CO2 production in the Netherlands. The proportion of CO 2 produced by households has increased strongly over the last 30 years, a tendency which is unlikely to change in the future without policy modifications. To promote sustainable development, a substantial reduction is required in domestic energy consumption. A promising approach to alter patterns of energy consumption may be to change life styles related to domestic energy consumption. Various authors [1, 2, 3] suggested that individuals' separate consumption behaviors constitute a more or less coherent pattern. It has been supposed that a person's particular acts of consumption are guided by basic values [4]. Values are consumers' mental representations of important end states they are trying to achieve in their lives [5]. These values guide individual choices with respect to consumption, causing characteristic differences or similarities between individuals. If this conception is correct, its implication would be that the focus of change must be patterns of energy consumption, rather than separate consumption behaviors. Attempts to change these energy consumpti-
1236 on patterns may eventuelly lead to change in particular energy consumption behaviors that make up the complete pattern [6]. The concept of coherent patterns of consumption within individuals is consistent with the concept of life styles, put forward by Weber [7]. According to Weber, individual tastes and preferences in behavior conform to socially determined structm'es. He maintained that groups of citizens can be distinguished by their socio-economic and status position. That is, persons sharing a similar status position, enjoy equal prestige in society, and are characterized by a common life style. Weber defines life style as a collection of explicit and expressive modes of behavior or behavioral preferences, in which consumption of material goods plays a dominant role. 1.1 An alternative conceptualization of life style In this study, the concept of life style is operationalized as so called means-end chains. This operationalization stems from economic psychological theories applied to marketing [8]. It is used to link products to individuals, according to the product's attributes that the individual considers valuable. According to the conception of meansend chains, people consider a particular product attractive, because it has attributes that implicate particular desirable consequences. In turn, these consequences are desired, because they serve to realize basic values which an individual considers of vital importance to pursue. Individuals can be classified according to differences and similarities in their means-end chains. This innovative operationalization of life style as means-end chains meets several of the criticisms with respect to the traditional concepts of life style. First, it is a domain specific approach: it recognizes that a person's life style may differ between different behavioral domains. Second, it does not limit itself to observable behaviors, but also includes a person's attitudes, socially determined normative beliefs, and basic values. To end this theoretical part, we formulate the research question of this study: is it possible to describe domestic energy consumption adequately in terms of life styles?
2. METHOD In order to trace means-end chains, we interviewed thirty-four consumers about their household energy behavior regarding six behavioral domains, namely commuter traffic, heating the house, lighting the living room, using the freezer, using the washingmachine, and using the washing-dryer. Every respondent has been interviewed on three of the six behavioral domains, according to a predetermined scheme. In the first part of each interview, we determined the common household context of the respondent, namely household situation, residential situation, and employment situation. Subsequently, we determined the context variables that were specific to a particular domain of energy consumption. For example, in the case of commuter traffic, the distance to work, the means of transport, and receipt of allowance were recorded. This part of the interview was ended by asking for the perceived benefits of the behavioral domain. In the second part of each interview, we used the laddering depth-interview method, that is used in consumer behavior research to trace means-end chains of products [8]. In our study, we have replaced products by perceived benefits of the behavioral domain. This implies that the attribute level in the means-end chain will be skipped. The laddering interview consists of determining why the most important perceived benefits are so important to the respondent. At best, each means-end chains ends at value level. In Figure 1, an example of one of our interviews on using the freezer
1237 is depicted. The percived benefit of using the freezer is the functional consequence
always food in stock. Through a number of steps, this means-end chain ended on the value level gives me a feeling of hospitality. GIVES A FRRTJNG OF HOSPITALITY
PEOPLE GET THE IDEA THEY ARE REAILY WELCOME
PRESENT S O n G
TO UNEXPECTED G ~
ANTICIPATE IN UNEXPECTED SITUATIONS
ALWAYS FOOD IN STOCK
Figure 1. Example of a means-end chain on using the freezer
2.1 Analysis The initial task of the analysis is to make a content-analysis of all elements from the laddering interviews [8]. All responses were divided as functional consequences, psychosocial consequences, instrumental values, and terminal values. This process resulted in a codebook of thirty-eight codes. Next, all individual ladders were rewritten in these number codes. In this case, we have made our data suited for the data-analysis. In order to detect groups of individuals with common characteristics, we have made use of a tandem use of correspondence analysis and cluster analysis [9]. This results in a graphical representation, which contains both the consequences and values, and the clusters of respondents.
2.2 Research questions To answer our research question whether domestic energy consumption can be described adequately in terms of life styles, we have made the following operationalization: 1. Is it possible to group means-end chains concerning a particular behavioral domain with regard to energy consumption, and are the groups interpretable as life styles concerning energy consumption? 2. Is it possible to group common context variables and context variables that are specific to a particular domain of energy consumption behavior? 3. Do groups of means-end chains regarding a particular behavioral domain of energy consumption overlap with groups of common context variables and with groups of context variables that are specific to that behavioral domain?
1238 3. PRELIMINARY RESULTS Currently, we merely have preliminary results at our disposal. We have enlarged our sample, but we have not assimilated it in the present results. Due to lack of space, we will only discuss the results of one of the six behavioral domains, namely commuter traffic. In Figure 2, the results of the data-analysis on the means-end chains on commuter traffic is depicted. It shows both the consequenes and values (the name codes), and the clusters of respondents (the numbers). The closer a respondent is situated near a name code, the more that name code applies to that respondent. In the case of commuter traffic, the cluster analysis resulted in three clusters. We can see that two large clusters (n=7 and n=6) and one very small cluster (n=l) were formed. Respondents in the first cluster, which were mainly cyclists, emphasized in the laddering interviews the healthy fresh air, physical movement, saving money, feeling pleasantly and at ease, and a better environment. Respondents in the second cluster, which were mainly motorists, emphasized saving time, ambition, being independent of external facors, freedom, functionality, and safety. The respondent in the third cluster, who travels by train, emphasized the atmosphere and the possibility to relax and dream.
~/ III
/
33
heze
better e n ~ t healthier 21 24 sa.ve money \ clean fresh ai~ 34 20 move~en t pleasuzable health health/ 7 save e ~ g y ol
2 ~, zng
Figure 2. Clusters of the means-end chains on commuter traffic
When considering the common context, three clusters were formed, mainly on the basis of the respondents' education, working situation, income, age, size of the house, and the presence or absence of school attending children. When considering the specific context, one large cluster and three small clusters were formed, mainly on the basis of distance to work, means of transport, the amount of travel-expenses, and receiving of a pay for expenses.
1239 When looking at the overlap between the clusters of the means-end chains, the common context, and the specific context, we notice that two groups of respondents are clustered together every time. This means that they show a strong correspondence in their means-end chains, and that they have a common and specific context that is very identical. Due to lack of space, we will not go into the content of these groups with an overlap in the means-end chains and context.
4. DISCUSSION In this article, we only discussed the results of the analysis on commuter traffic. The other behavioral domains - warming the house, lighting the living room, using the freezer, the washing-machine, and the washing-dryer - were left out of consideration in this contribution. In each domain, we found different clusters of similar means-end chains, and different clusters of similar demographic and relevant contextual variables, which partly overlap with clusters based on means-end chains. So, the preliminary results of this study offer indications that different groups of individuals can be distinguished with respect to a single domain of domestic energy consumption, based on the consequences attached by each individual to that particular behavior and the basic values he or she attains by that behavior. However, as yet no definite conclusions can be drawn about the relationship between on the one hand the consequences and values of a particular energy behavior, and on the other hand the relevant context of their living circumstances. It is difficult to conclude whether or not the similarities in classification of respondents on the basis of their means-end chains and on the basis of context variables constitute a causal relationship. Another unresolved issue concerns the relationship between structures of consequences and values with respect to one domain of domestic energy consumption and another. For example, are a person's positively valued consequences and related values with respect to his or her mode of commuting related to positively valued consequences and values with respect to heating the house, or are they unrelated? Further analyses of the data will have to provide answers to these questions. First, additional data will be collected in order to obtain information about means-end structures with respect to each domain of domestic energy consumption from approximately 35 respondents. This will provide a sound basis for further establishment of the reliability and validity of clusters of consequences and associated values regarding various domains of domestic energy consumption. Further, by means of discriminant analysis, we will explore the relationship between clusters of valued consequences and associated values, and clusters of general (demographic) and domain specific context variables. That is, we will examine how respondents clustered in a particular group on the basis of appreciated consequences and values differ from respondents clustered in a second group, with respect to general and domain-specific context variables.
1240 5. REFERENCES A. Mitchell, The Nine American Lifestyles, New York, Warner Books, 1984. H. Ganzeboom, Leefstijlen, in Jaarboek '90-'91 Nederlandse Vereniging van Marktonderzoekers, Haarlem, De Vrieseborch, 1989. P.A.F. de Bruijn and R.J.T. Custers, Voorwaarden voor Consumptieverandering, 's Gravenhage, SWOKA, 1993. M.J. Rokeach, The Nature of Human Values, New York, The Free Press, 1973. J.P. Peter and J.C. Olson, Consumer Behavior and Marketing Strategy, 3rd edition, Hollywood, Irwin, 1993. C.A.J. Vlek, Leefstijlen, gedragsverandering en energiebesparing: een conceptuele en methodologische beschouwing, in K. de Paauw, A. Perrels, en A. van Veenendaal (eds.), Leefstijl en energie: van intentie naar actie?, Petten, ECN2, 1994. M. Weber, Wirtschaft und Gesellschaft, Tfibingen, 1972. T.J. Reynolds and J. Gutman, Laddering Theory, Method, Analysis, and Interpretation, J. of Advertising Research, 28 (feb/mar) (1988), 11-31. P.E. Green, C.M. Schaffer and K.M. Patterson, A reduced space approach to the clustering of categorical data in market segmentation, J. of the Market Research Society, 30 (3) (1988), 267-288.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1241
NRP-project SCAN (SCenario ANalysis) Analysis of the social significance, acceptability and feasibility of longterm low energy/low CO2 scenarios for The Netherlands Kamminga, K.J., Slotegraaf, G., van der Veen, H.C.J. & Moll, H.C. poster presentation International Conference on Climate Change Research, Evaluation and Policy Implications 8 December 1994, Maastricht, The Netherlands
Abstract Long term low energy/low CO2 scenarios are developed for an interdisciplinary research about the significance, acceptability and feasibility of such scenarios. The social psychological research - to measure the acceptability - is directed at three determinants of cooperative behaviour in a social dilemma situation: knowledge, trust and morality. On the basis of this triad, research variables have been formulated and a research model has been developed. The sociological r e s e a r c h - to measure the feasibility - concerns an assessment of resistances, blockades, interests, and conditions for cooperation of the involved organisations with regard to the (package of) measures and the expected effects of the measures. The economical research- to measure the significance - will be mainly directed to acquire qualitative insight into effects of sectoral measures and packages of measures at a macroeconomic level.
Introduction
1
The aim of the research in the SCAN-project is to supplement general long-term lowenergy/low-CO2 scenarios and to clarify these in terms of their social significance, acceptability and feasibility. This research is conducted by an interdisciplinary team (psychology, sociology, economic and environmental sciences). The following research questions in the SCAN-project are discerned: A. 1. What is the expected quality of life within the variants of these scenarios? A.2. What is the acceptability of the. proposed scenario variants? A.3. What is the feasibility of the prospective elements of the proposed scenario variants?
Cor Kamminga and Henk Moll, Centre for Energy and Environmental Studies (IVEM), Nijenborgh 4, 9747 AG Groningen; Goos Slotegraaf, Institute for Social and Organisational Psychology (S&O), Grote Kruisstraat 2/1, 9712 TS Groningen; Henk van der Veen, Department of Sociology, Grote Rozenstraat 31, 9712 TG Groningen. All authors are working at the Groningen University of The Netherlands.
1242 Firstly, a new scenario is devised, i.e. packages of measures aiming at a (substantial) reduction of the energy consumption and the CO2-emission by The Netherlands. This scenario has been revised and evaluated by experts with regard to energy conservation and CO2-emission reduction, and to the behaviourial and economic effects of the considered measures. The scenario consists of measures derived from different relevant categories, i.e. technical measures, regulatory measures, measures providing financial-economic incentives, educative and communicative measures and measures aiming at organisational and institutional change. It is supposed, that an effective scenario should cover all these categories. The scenario measures are directed at four sectors: industry, greenhouse-horticulture, freight transport and (household) consumption. About 80% of the total energy use of The Netherlands is consumed by these sectors. This scenario is the starting point for the research, concerning the social significance, acceptability and feasibility of low energy/low CO2 scenarios. The relationships with these issues and the different disciplines are presented in figure 1. The results of these scenarios are predicted with help of the
Package of Measures SIGNIFICANCE economy environmental studies
economic and environmental
analysis. The results are described in terms of reduction Soc~a~ Psychology of the total Dutch energy consumption, of reduction of Results the Dutch CO2 emission, and of the changes of relevant Energy economic parameters e.g. C02 employment for industry, greenhouse horticulture and freight transportation. These ACCEPTABILITY results will demonstrate the Social Psychology environmental and economic Sociology significance of these scenFigure 1 Integrationof the 3 SCAN lines of research. arios. The issue of acceptability of the scenarios is primarily elaborated by social-psychological research by a postal survey among a large sample (several thousands) of households. Also the significance of individual and economic effects is measured for the households by this research. The issue of the feasibility of these scenarios is examined by a sociological field research. Political and institutional actors are interviewed about their interests, opinion, position, resources and influence. In this way major resistances and barriers concerning the acceptance of CO2 emission reducing measures will be determined and the feasibility of these measures will be estimated. FEASIBILITY
Soc,ology
1243 The survey results about the acceptability of CO2 emission reduction measures by households may influence also the feasibility of these measures. Political and institutional actors will be confronted with these acceptability judgments of households.
The social psychological line of research The social psychological line of research, aims at understanding and predicting the acceptability of policy measures at the individual level. For this purpose, specific policy measures aiming at the energy saving behaviour of individuals and households have been selected. This kind of behaviour can be characterised as a social dilemma; long term benefits can only be achieved by the cooperative behaviour of others.
Theoretical background Dawes (1980) 2, who elaborated on the social dilemma paradigm, earlier introduced by Hardin (1968) 3, argues that the three most important determinants of cooperative behaviour in a social dilemma situation, are best described by the psychological constructs I knowledge, trust and morality. ! On the basis of this triad, research variables have been : ._ , lull formulated and a preliminary i ' model has been developed. The model can be characterised, as demonstrated in figure 2, by the two 'routes', by which the acceptability is li ! UNCERTAINTY influenced. The first route ~dlllHIIIIIIIIIIIIIIIIIIIIIIIIIIIIIimllllWIHIIIIIIHIIIIIIIIIIIIIIl~lll concerns the causation of the problem and the uncertainty about environmental processes. Knowledge, problem awarei , ` " ~ ~ ~ ~ ] ness, responsibility for causing the problem and the perceived solvability of the problem are 1 1 the key variables. The other route affects the solutions proposed to tackle the problem and the uncertainty about the (cooperative) behaviour of Social psychological determinants of the acceptability of policy measures others. The impact of the pro@ Stimulus , , Predictor 9 Dep. var. posed solutions (or policy Figure 2 Model and variables for the social-psychological measures) on the individual research approach to assess the acceptability of energy reduction 'quality of life', the amount of measures and scenarios.
IIIllllill lllllllllllllJ tJ
_ll l U UNCERTAINTY nl lP-
11111
L
llff
r
1244 trust in others, the responsibility for solving the problem and perceived effectiveness of the policy measures are the key variables here.
Methodology In a large scale postal survey research amongst 3000 households, a package of future policy measures presented is presented as a short scenario. The model parameters are supposed to have a predicting value in relation to the acceptability of the policy measures in the scenarios. This will be tested by means of a multiple regression model. Furthermore, differences between certain groups will be inspected. Next to the variables mentioned above, attention will be paid to individual and group differences, regarding income level, education, age, gender, type of household, car use and ownership, etc. Final results can be expected in the spring of 1995.
The sociological line of research The research objective of the sociological research line is to determine the short term and long term social political feasibility of extensive and controInvolved Organizations versial energy saving measures on four social sectors each having large I I energy saving potentials. Quantitative Data Qualitative Data The defined package of measures for the sector Households will be completely covered and for Descriptive Statistics Content Analysis Acceptability Statistic the other sectors some Policy Decision Making Models measures are selected according to controversiality, effect size, concreteness and the Organizational Acceptability Resistance and Blockades number of involved orgCollective Acceptability interests predicted Outcome of Policy Process Conditions for Cooperation anisations. For each Power Distribution of Organizations Effects and Side Effects Impact of Policy Position Charge sector the acceptability and feasibility of the introduction of an energy I tax will be examined. Acceptability of Energy Saving Measure For the sector GlassFeasibility of Energy Saving Measure house Industry the estabFigure 3 Sociologicalapproach to study the feasibility and accepta- lishment of usage quota bility of energy use reducing measures and scenarios. Measure
1245 of natural gas and for the sector Freight Transport of the realisation of the Betuwespoorlijn are analysed specifically.
Methodology The methodological framework is demonstrated in figure 3. The first stage of the research consists of the specification of the measures and the determination of the organisations who participate in the political decision making process concerning that measure. In the second stage qualitative data as well as quantitative data will be gathered in expert interviews for each involved organisation. In the third stage the data will be analysed. Content analysis of the qualitative data will result in resistances, blockades, interests, conditions for cooperation of the involved organisation with regard to the (package of) measures and the expected effects and side effects of the measures. The quantitative data will be analysed using several statistics and political decision making models. This will result in estimates about the acceptability, predictions about the outcome of the decision making process, in an overview of the power distribution of the involved organisations and an assessment of the impact of policy position change of key organisations. The final results are judgments of acceptability and feasibility of the (package of) energy saving measures. The other (not selected) measures will be dealt with in a merely qualitative way. Though the survey is not yet completed, it seems that most energy saving measures will be supported at the institutional level. The notion appears to exist that energy saving measures are necessary and inevitable. These initial results make the question about the political feasibility of the measures very salient.
The economic line of research
The main research question of the economical research line is about the economic significance of the considered low energy/low CO2-scenarios, i.e. packages of measures concerning the sectors Glasshouse Industry, Industry and Freight Transport in the Netherlands. Before this main question can be answered, another question arises: How, from an economic point of view, has the present situation arisen concerning energy use and CO2emission in the Netherlands? In answering these research questions two phases can be distinguished. The first phase has already been finished. These results will be applied in phase 2. In the first phase attention was paid to the description of the relations between the two basic economic elements 'production' and 'consumption'. In doing so emphasis was laid upon a qualitative approach. Within this framework the following question was answered: How, from an economic point of view, has the present situation arisen concerning energy use and CO2-emissions in the Netherlands? The accent was on establishing the connection between the economic development during the 2 0 th century and the economic sector structure. The importance of this phase lies in the fact that it generates a diagnosis by
1246 which possible economic bottle-necks can be demonstrated to get to a low-energy/low CO: future. Supplemented with elementary quantitative data on energy-use, CO2emissions and economic performance, it also offers a first general insight into possible ways a measure or a package of measures might work out in social-economic terms during the time needed to get from the present situation to a low-energy/low CO2 society (the transition period). This in turn offers the opportunity to recognise possible socialeconomic problems in time and to think of strategies to avoid them.
Conceptual framework In phase 2 the main research question will be answered based on the general concept shown in figure 4. In answering this question interviews will be held with both general economic experts and experts from the relevant sectors. Concerning the economic significance two angles can be distinguished. Firstly, attention will be directed towards the meaning of a sectoral package of measures for the economic sector structure. In this way it is tried to generECONOMICAL
ENVIRONMENTAL
I
REQUIREMENTS - innovation - strenghtening sectorstructure
REQUIREMENTS - energy saving - reduction CO2 emission
I ECONOMICAL
I I .
I
INSTITUTIONAL CHANGES
I I
IOot::UoOellem~
CHANCES
enwronmental I production sector
I MEASURES RES iI
v ECONOMIC TRANSITION
" ALTERING SECTORSTRUCTURE
iore eveI I ':n ntoflI
II
ate some qualitative insight into effects of sectoral measures and packages of measures at a macro-economic level. Next to that, the significance of the packages at a sectoral level will be examined. In that connection more quantitative research is carried out.
Figure 4 Conceptual framework regarding environmental and economic requirements.
References This paper is partly based on the interim report of the SCAN-project of July 1994. A final report will be produced in the spring 1995. 2.
Dawes, R.M. (1980). Social Dilemmas. Annual Review of Psychology, 31, 169-193
3.
Hardin,G. (1968). The tragedy of the commons. Science, 162, 1243-1248
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1247
Toward a morality of increasing moderation W. Aarts, C. Schmidt and F. Spier Amsterdam School for Social science Research, Oude Hoogstraat 24, 1012 CE Amsterdam, The Netherlands
Abstract This paper sumlnarizes some provisional conclusions of three interrelated historicalsociological studies on 'economization' and 'ecologization'. Special attention is paid to the role status aspirations play in these processes. The focus is on ecologization, its conditions and obstacles.
1. E C O N O M I Z A T I O N AND E C O L O G I Z A T I O N AS CIVILIZING PROCESSES Current environmental problems are to a considerable extent caused by ecologically unbridled economic growth. Central problem of the first study is what social driving forces are behind this type of growth and to what extent they obstruct the control of environmental degradation. The answer to this question requires the elaboration of two theoretical concepts: 'economization' and 'ecologization'. Economization refers to a long-term social process in the course of which a growing number of societies turned into relatively peacefully competing regimes for generating wealth. For the societies involved this meant that 'economic' ways of doing and thinking gradually penetrated more and more spheres of life including that of the state. Economization might be considered a civilizing process for two related reasons. First, 'economic' activities were increasingly looked upon as more 'civilized' and prestigious than the extraction of surplus under the threat of violence, predation or war. Second, the process of economization brought about a growing social pressure towards self-control as well as an increasing control over nature. In the course of time, however, the resulting increase in affluence led to a relaxation of standards of frugality in the sphere of consumption. Economization implied an enormous increase in the division of labour. This meant in practice that a growing number of people were living and working in cities where they were not immediately confronted with the ecological effects of their activities. They could even cherish the illusion not to be dependent on nature any more. In reality, however, the increasing control over nature that made the urban-industrial way of life possible implied a growing, though less directly felt, dependence on the environment. The illusion of being released from ecological constraints explains the long-term short-sightedness of 'economized' societies with respect to the ecological effects of unbridled intensive growth. The term 'ecologization' refers to a re-awakening to these effects - the development of
1248 what came to be called 'environmental awareness' - as well as to attempts to keep the nature of human activity and the numbers of the human species within constraints considered 'ecologically acceptable'. In a way, the process can be looked upon as a continuation of economization because ecologization implies striving for optimum welfare within ecological constraints. In the twentieth century, social pressures toward more 'ecological self-control' have increased considerably. The second study deals with an important aspect of this long-term development.
2.THE RISE AND EFFECTIVENESS OF NON-GOVERNMENTAL ENVIRONMENTAL ORGANIZATIONS IN THE NETHERLANDS
From the beginning of this century, private organizations made efforts to protect specific parts of the Dutch landscape, flora and fauna, such as the Organization for the Protection of Birds and, most notably, the Organization for the Conservation of Nature Monuments. They focused on limited goals, the conservation of specific sites and/or of certain biological species. They were largely made up by members of the higher classes, whose rather effective political lobbying was mostly done in a discreet way. At the same time, they promoted their goals publicly by trying to get attached to it high status and prestige, as the name 'nature monuments' already suggests. This image-tbrming strategy can be summarized by saying that they sought to project a positive, 'high culture' image. Nature was beautiful, and should consequently be preserved. This motivated many people to associate with their cause. Although today the leadership of Nature Monuments expresses discontent with the current situation, the organization has been highly successful in terms of its original goals. The idea of protected areas has ahnost completely been accepted by the Dutch public (which explains why they are so easily overlooked). Such sites have steadily grown in size and numbers. In the 1980s, membership sharply increased and by 1994, its paying following is the largest of all ecological organizations in the Netherlands. By contrast, many sections of the ecological movement that came up in the 1960s had very wide-ranging aims, which included major changes in consumption as well as incisive societal change. Their campaigns were characterized by a rather informal code of conduct. Such activists tended to sound the alarln and projected an image of their goal which up to today is seen by many as an abhorrent example (the 'goat's woollen socks' image). For instance, the organization Environmental Defense (Milieudefensie) continually prophesied doom and gloom if its advice would not be heeded. Yet, their positively phrased 'Action Plan Sustainable Netherlands ' attracted a great deal of attention at home as well as abroad. This leads to the conclusion that those organizations which addressed tar-reaching issues like personal general ecological awareness and moderation chose a rather ineffective strategy to attract followers to their cause. By contrast, their not so spectacular predecessors reached their less ambitious goals by a rather effective strategy. Although sounding the alarm is a necessary component of efforts to stimulate ecological awareness, positively phrased campaigns to stimulate specific forms of moderation are likely to be more successful than alarmist approaches, and should clearly be kept separated. In addition, the ability to exercise influence at the highest level of decision making, including
1249 getting public support of highly-placed citizens, not only verbal but also in practice, may be helpful to spread forms of ecological moderation. The third study deals with clues for these and other forms of moderation in consumption, especially in the Netherlands.
3. CONSUMPTION AND STRATIFICATION The striving for the maintenance and improvement of social status is one of the primary driving forces underlying the continuing increase in consumption. The same drive, however, may also lead to an increasing moderation of consumption. In search of feasible solutions for environmental degradation, the third study focuses on the counterforces to the growth of consumption. Broadly speaking, sociological research reveals a positive relationship between power, wealth and prestige on the one hand, and the quantity of consumption on the other. In addition, a 'trickle down-effect' has been frequently observed. Patterns of consumption and consumer goods that were initially reserved for the members of privileged groups spread out to society at large. Ii1 this way holidays by air, cars and eating meat every day trickled down as did less tangible elements such as sensitivity to art and nature. High status, however, does not always coincide with conspicuous consumption. Historical-sociological research indicates that the members of privileged groups have time and again imposed restrictions on each other and on themselves. For example, in situations of rivalry between groups with economic power on the one hand and groups that possess principally cultural power on the other, the latter frequently tend to distinguish themselves by consumption that bears witness to self-control, tact and good taste. Moreover, whenever consumer goods become more widespread, their status-conferring character diminishes and from that moment on moderation might become prestigious. Closer analysis of research into the development of smoking and eating habits since the Second World War demonstrates that status has played an important part in pushing back smoking and eating fat food in industrial societies. The spread of nonsmoking and health food are typical examples of the effectiveness of the trickle down-effect. Interviews with members of high-status groups who practise forms of restraint which are not (yet) common indicate that they meet with growing social esteem. However, various sorts of environmentally friendly restraint do not seem to add much to social prestige. This may change, though, as a result of extensive attention in the media and the efforts of government and industry. People's preferences for moderation in different areas are interrelated. They are part of a more general status-related morality in which striving for self-control, responsibility and quality are at the centre. In most cases environmental concerns appear to be not the main motivation. Analyses of secondary resources demonstrate, for instance, that the educated hardly refrain from consumption that causes excessive emissions of carbon dioxide. Their environmental concern in this respect is still mainly symbolic though communicatively significant. But then, the anxiety about the greenhouse effect is relatively recent and still controversial. The fact that educated people are over-represented among the members of environmental organizations and buyers of environmentally more friendly products shows at least their willingness to do something for the environment. So there seems to be a potential for change here. However, as the report shows, the problem remains that the political-economic regime,
1250 which created the conditions for the beginning ecologization of society, continues to be permeated with strong social pressures obstructing that very same process. Nevertheless, some support has been found for the hypothesis that under specific conditions an appeal to status may be effective in strengthening ecological regimes.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1251
C l i m a t e C h a n g e , L i v i n g E n v i r o n m e n t and W a y s o f Life M. J~irvel~i and M. Wilenius Research Institute for Social Sciences, University of Tampere, P.O. Box 607, SF-33101 Tampere, Finland
Abstract
Our empirical material based on interviews with influental actors in environmental policy in Finland suggested that possible future climatic changes illustrates the greatest single environmental threat on a global scale. The influential actors did not hesitate to consider as an issue of high certainty a kind of man-induced climate change. In mapping out social resources among actors to tackle climatic risks we have utilizised a teleological reasoning of rational action as an ideal model.
1. Introduction
From the point of view of global social and political regulation, the most complex and challenging issue in present-day ecological policies can be seen the question of climate change. Furthermore, within scientific world, there is a widening consensus about necessity to carry out social science research to contribute, on the side of natural and technical sciences, our knowledge about global environmental issues like climatic changes (see Morrisette & Plantinga 1991, Buttel & Taylor 1992,). In this presentation we will draw attention to the issue of climate change as a special case of social and environmental conflict in late modernity. In our complex societies, experts seem to gain ever more influence over issues like social regulation as the problems themselves grow more compicated (Fores et al. 1991, pp. 83-84, Parsons 1958, 34). In the field of environmental protection, the task to create rational modes of thinking and political action strategies is easily left to few highranking experts (Sundqvist 1992). The present part of our research focuses on socially influential groups that have a significantly important status in determining the interests, knowledge and morality in the definition of problems in environmental policies. Our empirical research sample includes various environmentally influential experts found in major industrial companies in Finland, politicians active in environment policy, experts in public administration and in the field of science, and journalists interested in environmental issues. We have also interviewed some eminent civil activists. The empirical material consists of sociological theme interviews. Rather than outlining different viewpoints of interest, our research focuses on the idea and knowledge resources
1252 that project different ways of thinking. The following figure illustrates the various social dimensions and structures embedded in the handling of the issue: The social resources of climate change politics
Theoretical Teleologies
f life
economicalworldview/value ; ~ ; e n d e s X ecological worldview / ,,l., \ sdentiflc/ i I ~ ~ tTaditJonal/mythical knowledaeT" "/
interpretaUgnof reality ~.1/ -
~
~
and sden~lflc
knowledge
fields of "conflict
I elites
> <
I lay people
There are three basic cognitive layers to be defined: The first layer consists epistemological question which goes whether the climate change is a real phenomenon or not. The experts we interviewed were very affirmative on this issue as we shall see later. The second layer deals with societal objectives. What should form the policy basis? Is the goal set in prevent dooming climatic changes or should we orientate rather to adjust to changes and mitigate the effects where possible. Here we found out two clearly differing emphasis on issue: the one which pointed towards reorientating the present public policy and the other which questioned the whole structure of that policy. The third layer points to the measures implemented to fulfill the objectives. Here we could also identify even more radically diverging concepts of action: the predominating discourse which adhere to technical rationality and the counter discourse which points our consumeristic life-styles. Before going more into substantial implications of this model we like to explore the social risk profile of climate change.
2. Climate change as a future risk
Over the past twenty years, there has been a lot of discussion concerning man-made reinforcement of the natural greenhouse effect. Statistics show that since the 1950's, the atmospheric concentration of carbon dioxide (CO2), the main proponent of the greenhouse effect, has risen sharply (Kanninen 1992, p. 33). Furthermore, at global level, statistics demonstrate some increase of tropospheric average temperature. Yet, due to the lack of sufficient long-term and homogenous observational data, scientists have been unable to pick up the "signal" of greenhouse warming from the amplitude of "natural climate variability". This all means that, according to the distinguished Intergovernmental Panel of Climate Change, it may take at least another decade to detect the intensified greenhouse effect from the observational data (IPCC 1990; Stehr & yon Stroh 1993)
1253 However, our empirical material based on interviews with influental actors in environmental policy in Finland suggested that climate change based on acceleration of the greenhouse effect illustrates the greatest single environmental threat on a global scale. On the basis of our sample, we were able to form a three-level risk pyramid which illustrates schematically the order of global environmental risks as reflected by actors.
Environmental risk pyramid in global perspective
TOP RISKS
ozone depletion ~eshwaterand sea polluUon
nuclear energy 2 N \ D RATE the spreadof acidrain ~RISKS toxic waters and chemicals
death of
organic soil
t desertiflcatlon
pollutive production forest destruction
extermination of Indigenous
peoples poverty
imbelanced distribution power Instrumentalizatlon of nature
overconsumptlon
of energy
biased ways
of
of life
resistance overc0nsumption to change of natural resources world economy system areal integration processes
3.Teleological structures in the reasoning of influential actors As we have pointed out earlier, in mapping out social resources among actors to tackle climatic risks we have utilizised a teleological reasoning of rational action as an ideal model. In sociological terms, the interpretation of reality and the suggested goals for societal development, together with appropriate means, form the agenda for social action strategies as in the case of the climate change. As far as climate change problematique is considered, following predominant structures of rationality among experts were detected in our study:
1254
Teleological rationality of predominant discourse as expressed by actors" Interpretation of Reality
Objectives
Measures
Climate change is a fact Environmental policy should be based on this Scientific evidence inadequate
Adjust to changes Prevent harmful effects through effective policy
Save energy Promote low CO2 energy production (nuclear power, new technology)
Counter discourse: Opposed to above rationality. Replacement proposal: Interpretation of Reality
Objective
Measures
Climate change is a fact but no reason to ignore other environmental problems
Prevent changes through complete reform of policy principles
Abandon consumptioncentered lifestyle Application of new and renewable energy sources
References 1 Morrisette Peter & Plantinga Andrew 1991: Global Warming: A Policy Rewiew. Policy Studies Journal, vol 19, No. 2, pp. 163-173. 2 Buttel Fredrick H. & Taylor Peter J. 1992: Environmental Sosiology and Global Environmental Change: A Critical Assessment. Society and natural Resources, vol 5, pp. 211230. 3 J~irvel~i Marja & Wilenius Markku 1993: Climate Change, Living Environment and Ways of Life. University of Tampere, Research Institute for Social Sciences, Working Papers 9/1993. 4 Fores Michael & Glover Ian & Lawrence Peter 1991: Professionalism and Rationality: a Study in Misapprehension, Sociology, Vol. 25, No. 1. 5 Parsons Talcott 1958: Essays in Sociological Theory (Revised Edition) Glencoe: Free Press. 6 Sundqvist, G6ran 1991: Vetenskapen och milj6problemen (Science and the Dilemma of Environment - in Swedish), Monograph from the Department of Sociology, University of Goethenburg. 7 Kanninen, Markku (ed.) 1992: Muuttuva ilmakeh~i. Ilmasto, luonto ja ihminen. (Changing Athmosphere. Climate, Nature & Human Being.) Helsinki: Valtion painatuskeskus. 8 IPCC 1990: Climate Change - The IPCC Scientific Assessment. Report prepared for IPCC (Intergovernmental Panel on Climate Change) by working group 1. Houghton J.T. & Jenkins G.J. & Ephraums J.J. (eds.) Cambridge: Cambridge University Press. 9 Stehr Niko & Stroh Hans von 1993: Climate Change, the Social Construct of Climate and Climate Policy. Hamburg: Max-Planck-Institut.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
WELFARE QUESTION
AND
ITS
RELATION
TO
THE
1255
ENVIRONMENTAL
Maarten A. Mentzel Leiden Institute for Social Scientific Research (LISWO), University of Leiden, Wassenaarseweg 52, 2333 AK Leiden, The Netherlands
Abstract The article argues that the currently dominant idea of material welfare is at odds with a lifestyle that does justice to basic h u m a n values. Measurement of welfare needs to attach importance to a good environment. With regard to the subjective counterpart: social research on the experience of welfare - the quality of life - in various Western countries shows clearly that intangible values are very important in the lives of most people. In a process of globalization of economy and cultural supplies, it will be important to strengthen democratic concern with the environmental question. The legitimation of political authority in (inter)national negotiations will depend on the attention paid to welfare and quality of life as a public affair.
1. A C H A N G I N G W E L F A R E C O N C E P T In welfare theory the main stream of research is directed towards the way a society provides for individual and collective needs. Welfare, defined by the dictionary as 'a state of prosperity', varies with both time and culture. The dominant interpretation of 'welfare' in Western democracies is economic welfare as a component of total welfare. Indispensable for a new orientation is first of all a model that can illuminate and quantify prevailing ideas of economic or material welfare. Concepts like well-being, quality of life and a 'good life' are used to refer to the welfare experienced subjectively by individuals. In the context of the rise of Modernity one can understand how this material interpretation of welfare came to dominate. In the West people have come to emphasize active processing and use of natural resources. A second element is the concept of welfare during recent decades. Increase in national income is used not only to measure economic growth, but also to indicate rising welfare. Although welfare and national income are related, the level of welfare is obviously not synonymous with national income. After all, economic growth has been achieved to a great extent at the expense of the quality of the environment. The literature supporting this proposition leads to the conclusion that the old concept of 'growth' is no longer appropriate. Population growth and increa-
1256 sed material welfare threaten the environment. Environmentally sustainable economic growth (1991) is one of the m a n y studies that conclude clearly: 'The world has reached limits'. 1 In the early seventies several researchers advocated a new concept of welfare, in which external effects of production processes and consumption patterns that were not priced would be given weight. 2 In the meantime we know for sure t h a t the striving for more and more - the idea of growth - underlies the domin a n t lifestyle in the West. The Declaration of Rio (1992) proposes using a new measure for growth. 3 During the UNCED it was recognized that the capacity of the environment is limited and that a re-orientation in the Western world with regard to material lifestyles and consumption patterns is necessary. Moreover it was affirmed that under current consumption and production patterns in the North a j u s t distribution of the global environmental space is not possible. Western culture, with its consumption and production patterns, has spread around the world. Since the Club of Rome's Limits to growth (1970) much attention has been paid to this, but how people experience welfare has been insufficiently examined. Not until the eighties did the discussion get started. What does new research have to say about this? The following two sections treat both the idea of a new, 'sustainable' lifestyle and its limits.
2. L I F E S T Y L E Dominating the current image of h u m a n s - as shown by John Passmore in
Man's responsibility for nature (1974) - is the anthropocentric perspective on the environmental question. In a relation towards nature in which h u m a n s are central, the main preoccupation will be how to manage the environment. If this does not change, science, technology and the capitalist economy will lead to the self-destruction of the existing order. Therefore, individuals as well as governments are starting to realize t h a t there are natural limits to the expansion of the science-technology-capital system. 4 In contrast, in a more ecocentric perspective on the environmental question the emphasis will be on cooperation with nature. W h a t recent developments can we identify that exhibit this new mentality and lifestyle? ~ Obviously, what is happening in the sphere of standards and values is only part of a host of developments - in population, economics, politics, science, technology, physical p l a n n i n g - which together must change in order for there to be a sustainable future. By this we mean a future in which distributional considerations (including inter-temporal distribution) and welfare maximization insure the preservation of the environment. A sustainable lifestyle finds expression in the main spheres of life, at home, at the work place, in the traffic system, in leisure activities and travel. Principles of a sustainable lifestyle may include: - attention for the value of intangible aspects of life - happiness is related to the development of one's talents - acceptance of self-imposed limitations
1257 - shorter working hours - sharing of paid work, household and n u r t u r i n g activities by both men and women. 6 These considerations have repercussions at a more fundamental level as well. A lasting change in lifestyle or mentality requires a change in the framework of collective meanings under which people live, too. If these collective meanings including the value assigned to the environment - do not change, then superficial changes in lifestyles will not achieve the desired effect. One approach to designing an image of h u m a n s that is consistent with a sustainable lifestyle, is the conception of fundamental h u m a n functional capabilities. M a r t h a N u s s b a u m has drawn up a list of capabilities which are basic to h u m a n life. A minimal theory of the good can be designed that is consistent with this list of basic capabilities. This list should be further examined, for instance by redefining the 'good life' or the 'quality of life'. 7 Moreover, empirical research is needed on the usefulness of this list as a starting point, as we will see in the next section.
3. M E A S U R E M E N T
OF WELFARE
'The' lifestyle in Western society includes both a set of consumer activities and a set of preferences. These two sets need to be described, thereby separating the economic from the socio-psychological aspects and also adjusting the sets to the various social groups in society. By comparing different countries in the West, a coherent and empirically sound view of various Western lifestyles can be presented which together make up Western culture. (For a world-wide overview, see for instance the World Development Report 1992.) The crucial point is to identify the connection between well-being and environmental space. This method of identifying components of well-being and measuring them by use of indicators or by surveys can bring together new data on Western societies. Four recent studies have been done along these lines: (1) A Swedish study, based on surveys in 1968, 1974 and 1981, explores questions of poverty and inequality. Searching for what causes people to experience a sense of well-being, the findings of this study emphasize people's capacity to satisfy their needs or, more generally, 'to control and consciously direct [their] living conditions'. The redistributive function of the state is emphasized: 'a redistributive model of social policy should cover the basic needs of all citizens', s (2) Another Swedish study of the quality of life is a comparative research based on interviews held in the seventies in Denmark, Sweden, Finland and Norway. 9 Having, Loving and Being are the labels used for central necessary conditions of h u m a n development and existence. 'Having' refers to material conditions. The indicators measure economic resources (income and wealth); housing conditions; employment; working
1258 conditions; health; education. 'Loving' can be assessed by measuring attachments and contacts in the local community; attachments to family and kin; active patterns of friendship; attachments and contacts with fellow members in associations and organizations; relationships with workmates. And 'Being' may be characterized as personal growth as opposed to alienation. The indicators measure the extent to which a person can participate in decisions and activities influencing his life; opportunities for leisure-time activities; opportunities for meaningful work, and opportunities to enjoy nature, either through contemplation or through activities such as walking, gardening, and fishing. (3) Also from Scandinavia is the 'well-being' index drawn up in the nineties in Norway. The central question is: What makes life worth living? The results show that the following factors are decisive: social relations; good health; a clean environment and scenic experiences; and meaningful work. In the fifth place is material possessions. 10 Most important for the improvement of society are the first and the third factors mentioned, interpersonal relationships and the state of nature. These have deteriorated during the last 30 years. (4) A last example of empirical research on how people experience welfare is a calculation of the consumption level per country. This method is based on the premise that the contours of an ecological society cannot be sketched by making a sum of individual consumption patterns. In The Netherlands, Milieudefensie (associated with Friends of the Earth) believes that the two most important goals of ecological change are to reduce consumption of natural resources and to lower the expectations of the material side of the 'good life'. Research along these lines will be an essential supplement to the seminal findings of Partha Dasgupta. These focus on the conditions in which people live and die in rural communities of poor countries. 11 Common to the four studies outlined above is the questioning of economic growth and the search for a shift away from economic growth; it is precisely the emphasis on economic growth that undermines the intangible values which are so important in the lives of most people.
4. W E L F A R E AS A P U B L I C A F F A I R
However convincing a new framework may be, translating ideas into reality requires political decision making. At least two points deserve attention: the relation between national and international decision making and the relation between short-term and longterm policy. Questions relevant to the national level: What instruments can be used? Has politics a role in it? Can politics manage this process of change, or must the public opinion first change? But this last approach, however important it is, has
1259 been tried for many years without much effect. Therefore many leading researchers have argued for a paradigm shift towards a reduction in the consumption culture. Instead of just talking about the necessity of economic measures, it is preferable to shift to positive incentives for austerity. And at the international level one of the important questions is: how to invest institutions with sufficient authority to present an appealing vision of sustainable welfare - welfare that both includes distributional considerations and covers the value of nature - and to incorporate this vision into international and supranational decision making processes. Above all it is clear that responsibility as a political category deserves much attention. In the first case a view towards the future of national democracies is needed; in the second case a view of the globalization process deserves attention. The role of collective and political action and the role of restricting demand deserve an analysis, in so far as sustainable welfare in relation to quality of life is involved. In short, research in the fields of both political and welfare theory together with social inquiries help to clarify the seeming contradiction between the maximization of the quality of life and the quality of nature. 12
5. L I T E R A T U R E This paper is part of the SWPA project, by the author in cooperation with J.W. de Beus (University of Amsterdam) and H. Verbruggen (Free University of Amsterdam).
E. Allardt 1993, "Having, Loving, Being: an Alternative to the Swedish Model of Welfare Research", in: Nussbaum & Sen, pp. 88-94. P. Dasgupta 1993, An Inquiry into Well-being and Destitution. Oxford: Clarendon Press. A. Dobson & P. Lucardie (Eds.) 1993, The Politics of Nature. Explorations in Green Political Theory. London: Routledge. R.B. Douglass, G.M. Mara & H.S. Richardson (Eds.) 1990, Liberalism and the Good. London: Routledge. Dutch Committee for Long-Term Environmental Policy (DCLEP) (Eds.) 1994, The Environment: Towards a Sustainable Future. Dordrecht etc.: Kluwer. R. Erikson 1993, "Descriptions of Inequality: the Swedish Approach to Welfare Research", in: Nussbaum & Sen, pp. 67-83. Environmental Resources Limited 1993, The Best of Both Worlds: Sustainability and Quality Life Styles in the 21st Century. The Hague: Ministry of the Environment. R. Goodland, H. Daly, S. E1 Serafy & B. von Droste (Eds.) 1991, Environmentally Sustainable Economic Development: Building on Brundtland. Paris: Unesco. D. Hareide 1991, Det Gode Norge. Oslo: Gyldendal Norsk Forlag. - 1994, "Has the Quality of Life Inproved in Western Europe?" Ms. F. Hirsch 1977, Social Limits to Growth. London: Routledge & Kegan Paul. R. Hueting 1980, New Scarcity and Economic Growth. Amsterdam: North-
1260 Holland Publishing Co. M. Mentzel & P.B. Lehning 1994, "A Political Basis for a Sustainable Society", in: DCLEP, pp. 443-462. L. Milbrath 1993, "Redefining the Good Life in a Sustainable Society", in: Environmental Values 2, pp. 261-269. M. Nussbaum 1990, "Aristotelian Social Democracy", in: Douglas et al., pp. 203252. T. O'Riordan (Ed.) 1995, Environmental Science for Environmental Management. Harlow: Longman. T. Scitovsky 1976, The Joyless Economy: an Inquiry into Human Satisfaction and Consumer Dissatisfaction. New York: O.U.P. World Bank 1992, World Development Report. Development and the Environment. Oxford: O.U.P.
6. R E F E R E N C E S
1. Goodland, Daly, E1 Faleh & von Droste 1991. 2. Hueting 1980; Hirsch 1977; Scitovsky 1976. 3. Cf. O'Riordan (Ed.) 1995, Ch. 1, 'The Global Environmental Debate', pp. 20f. 4. Dutch Committee for Long-Term Environmental Policy (DCLEP) 1994, p. 11. 5. Dobson & Lucardie (Eds.) 1993. 6. Environmental Resources Limited 1993. 7. Nussbaum 1990. 8. Erikson 1993. 9. Allardt 1993. 10. Hareide 1994. 11. Dasgupta 1993. 12. Cf. Mentzel & Lehning 1994.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1261
A S S E S S M E N T R E P O R T ON N R P S U B T H E M E
?NATIONAL INSTRUMENTS FOR C L I M A T E C H A N G E P O L I C Y ~'
H. Verbruggen Institute for Environmental Studies Free University Amsterdam De Boelelaan 1115 1081 HV Amsterdam The Netherlands
With contributions by: A. Nentjes
RUG, University of Groningen
J.G. Backhaus
RULI, University of Limburg, Maastricht
H.M.A. Jansen
VUA, Free University Amsterdam, Institute for Environmental Studies (IVM)
1262
Contents Abstract 1.
Introduction
2.
T r a d a b l e e m i s s i o n rights
3.
T h e feasibility of an ecological tax reform in The N e t h e r l a n d s
4.
Climate fund
5.
Applied general equilibrium model
6.
Evaluation
7.
References
ABSTRACT Economic i n s t r u m e n t s for environmental protection feature in textbooks for their superior performance in t e r m s of effectiveness and both static and dynamic efficiency, especially in cases characterized by a large n u m b e r of polluters with large differences in abatement cost. These i n s t r u m e n t s are thus pre-eminently suited to be included in climate change policies. However, this proves to be very difficult as yet. The projects u n d e r this research sub-theme share a common interest in the complexities of implementing various types of economic instruments. The first project deals with the design of a E u r o p e a n s y s t e m of t r a d a b l e emission rights. The second project is on the feasibility of ecological tax reform, with special reference to The Netherlands. The C l i m a t e F u n d project, the third project, aims at i n v e s t i g a t i n g w h e t h e r an international fund for side payments is an effective and efficient tool to reduce CO2 emissions. To t h a t purpose, the world is divided into 9 regions. Finally, the fourth project uses an applied general equilibrium model to analyze the effectiveness and especially the income distributional effects of differently designed CO2 charges. The projects yield interesting results from which policymakers can benefit. There still are, however, blind spots. These and further research questions are identified. 1.
INTRODUCTION
The projects under the research theme national and international instruments for greenhouse policy share a common interest in the complexities of implementing various types of economic instruments. It is commonly known t h a t there appears to be a wide gap between, on the one hand, the r a t h e r abstract textbook version of a particular economic i n s t r u m e n t and its implementation in practice on the other hand. This is so, because especially the category of economic i n s t r u m e n t s change
1263 the system of relative prices most directly. This means, firstly, t h a t these instruments have an immediate and clear-cut impact on the distribution of income within and among countries and on international competitiveness. To overcome the inherent societal resistance and increase the political feasibility of economic i n s t r u m e n t s , resort to tailor-made solutions has to be made. Secondly, for economic instruments to realize their presumed superior performance in terms of effectiveness and static and dynamic efficiency, they have to operate in wellfunctioning markets. And thirdly, economic instruments cannot stand alone, but have to be embedded in an institutional context and a supporting framework of regulations. If these conditions are not met and if no account is taken of these implementation difficulties, economic instruments will either not be used or their operation will be so distorted that they lose much of their imputed attractiveness. In fact, a circular research process can be ascribed to all projects under this research theme, be it with different accents, namely: 1. Exploration of one or more prototypes of economic instruments. 2. Thorough investigation of implementation problems. 3. Design of improved and feasible instruments. The major objective of the research is to come up with better designed instruments and/or complementary and supportive conditions in order to make the use of economic instruments more feasible, efficient and effective. A p r e l i m i n a r y a s s e s s m e n t of the results reveals that, in general terms, the research projects have been successful. There are interesting results from which policymakers can benefit. Thus, the project on tradeable carbon emission permits focuses on the design of a feasible system on EU-scale and examines to t h a t end the characteristics of the permit, its initial distribution and m a r k e t allocation, the time path by which the number of permits available is reduced and monitoring an enforcement of the system. Attention is also paid to the economic consequences of such a system for industry and consumers and to specific problems such as the existence of entry barriers and other market imperfections that might t h w a r t the functioning of the system. The project on the feasibility of an ecological tax reform in The Netherlands indicates that the feasibility of ecological taxes is determined by the tax design, the taxing authority and by the constitutional, institutional and fiscal s t r u c t u r e s in which these taxes are embedded. An i m p o r t a n t r e c o m m e n d a t i o n is t h a t ecological taxes can best be imposed by a taxation authority t h a t is directly related to the environmental good i.e. tax base, at hand. The project on the socio-economic impact of economic instruments makes use of a computable general equilibrium model of the Dutch economy. With the help of this empirically-based model, the income distributional effects of economic instruments to reduce CO2 emissions are thoroughly investigated. It appears t h a t changes in the design of these instruments might substantially mitigate these effects. Special attention is paid to a recently hotly debated issue, namely to what extent a shift of taxes from labour to environmental goods is conducive to employment generation. The final results of this analysis are not yet available. In addition, the project "Socio-economic Aspects of the Greenhouse Effect: Climate Fund" studied the impact of international capital transfers on the efficiency and efficacy of emission a b a t e m e n t . After p u t t i n g considerable effort in identifying the i n t e r n a t i o n a l distribution of the costs and benefits of such policies, and including them in an integrated assessment model, the study shows that there is not much room for
1264 international capital transfer, as emission abatement is economically rational only to a limited extent. This assessment report is organized as follows. In the Sections 2-5, the following individual research projects will be discussed. Table 1.1 List of projects in the NRP subtheme "National i n s t r u m e n t s for greenhouse policy" Title
Project leader
Number
Tradeable CO2 -emission permits
A. Nentjes
851052
The feasibility of an ecological tax reform in The Netherlands
J.G. Backhaus
851051
Socio-economic aspects of the greenhouse effect: Climate Fund
H.M.A. Jansen
851055
Socio-economic aspects of the greenhouse effect: Applied General Equilibrium Model
H.M.A. J a n s e n
851061
Per project, a description will be given of the objectives, progress, present state of affairs, (preliminary) research results and a (preliminary) evaluation. The evaluating Section 6 tries to assess the overall results of the research theme on national and international instruments for greenhouse policy. This assessment will be structured along the 3 different accents indicated in the introduction. On 4 November, 1994, a workshop was organized among the project coordinators and researchers to discuss progress and results, as well as to investigate the possibilities for closer cooperation. 2.
TRADEABLE EMISSION RIGHTS
The project consists of three parts: a. design of a E u r o p e a n system of tradable rights and investigation of the feasibility of the system; b. investigation of side effects, in particular entry barriers; c. investigation of effects of international coordination of environmental policy. Parts (a) and (b) are now nearly finished, part (c) will be the main topic of the last year of the project. It appears t h a t a system of tradeable rights can indeed be designed. Monitoring and enforcement, issues that in many cases are problematic, can in the case of
1265
CO2 easily be solved by introducing the system there where the energy comes first in the market, i.e. producers and importers. The initial distribution can be achieved through auctioning or grandfathering. This choice has implication for the creation of entry barriers. A micro-economic approach was used to analyze side effects such as entry barriers. G r a n d f a t h e r i n g of permits creates such barriers for newcomers in a sector. Auctioning does less so, but auctioning reduces the acceptability of the system by the target group. G r a n d f a t h e r i n g provides the receiving instances actually with a capital transfer (the value of the permits). By using the so-called limit price model, the somewhat surprising conclusion was reached t h a t transaction costs will raise the entry barriers. Another interesting conclusion, derived from the deep purse theory, is that imperfect capital m a r k e t s may raise the entry barriers, in particular in the case of grandfathering. The third part of the project is still underway. A second-best two countries model is being developed. The second-best approach is attributable to the question how revenues are being channelled back to the economy. Due to differences in CO2 damage functions and differences in reduction cost functions, cooperative behaviour can come about through applying side payments. A very preliminary and tentative conclusion t h a t seems to follow from the model is that Joint Implementation may not be an optimal instrument. The project is an interesting one, and in particular the investigation of entry barriers, a subject t h a t so far did not receive much attention in the literature, is innovative. 3.
T H E F E A S I B I L I T Y OF A N E C O L O G I C A L TAX R E F O R M NETHERLANDS
IN T H E
The aim of the project is to describe and analyze feasible ecotaxes and e n v i r o n m e n t a l charges; to give advice to policy makers on this issue; and to provide policy m a k e r s with policy alternatives with respect to ecotaxes. The methodology of the project is based on the economic theory of environmental policy, the theory of public finance, fiscal federalism theory, club theory and transaction costs theory. Three types of taxes can be distinguished: Pigouvian taxes, revenue raising taxes and earmarked taxes. It seems that earmarked taxes have a larger public acceptability t h a n other taxes. For ecotaxes to be both ecologically effective and acceptable, it is necessary that they are being levied by a tax a u t h o r i t y t h a t is closely connected to the ecological circumstances. This means t h a t ecological taxes can best be imposed by a taxation authority t h a t is directly related to (the m a n a g e m e n t of) the environmental good, t h a t is the tax base, at hand. It is therefore recommended that institutional framework with socalled ecological taxing units is created. Water authorities are an example of such units. For use by policy makers a checklist was drawn up so as to make sure t h a t no i m p o r t a n t aspects are overlooked when introducing an e a r m a r k e d tax, a revenue raising tax or a Pigouvian tax. The problem is that the present taxing authorities tend to be political units, which do not coincide with ecological taxing units. This means t h a t institutional and constitutional changes are required. The conclusion is that an ecological tax reform in The N e t h e r l a n d s is feasible, if each type of charge t h a t aims at ecological improvement is being designed carefully. In bringing about the start of such a tax and institutional reform, local authorities (like municipalities, water authorities
1266 and provinces) can play the important role of providing an example for the introduction of ecological taxing by higher authorities. 4.
CLIMATE FUND
The Climate Fund project (IES) aims at investigating whether an international fund for side payments is an effective and efficient tool to reduce CO2 emissions. First, an extensive literature review was made with respect to damage due to the enhanced greenhouse effect and with respect to costs of CO2 reduction. Where possible, regional differences in damages and costs were identified. Then a model was constructed (Climate FUND: Climate Framework for U n c e r t a i n t y , Distribution and Negotiation) in which 9 regions in the world are distinguished. The main emphasis in the model is on damages. In particular the intangible damages seem to grow over the next century. The spatial distribution of damages is unevenly distributed over the regions, with South and South East Asia and Africa as the main victims followed by Centrally Planned Asia, Latin America and the Middle East. A probabilistic analysis with the model shows that best guesses (on which policy is mainly based) of tangible damages are lower than expected values, under optimistic assumptions on the available knowledge, but much more so under pessimistic assumptions. An analysis was carried out on the effectiveness and the efficiency of an international fund, by comparing optimal reduction strategies in cooperative and non-cooperative games, with international side payments. The results are preliminary but indicate that international prisoners' dilemmas don't seem too sharp and that the opposite situation, where regions cooperate out of free will and self-interest, is more common. There are at least two possible explanations for this counter-intuitive result. First, countries that would benefit most from emission reduction have the least capital available to make cost-effective transfers. Moreover, these countries generally attach a low, or no priority to climate change. Second, the rich countries of the North are also hesitant to invest in abatement options in other countries, because then they lose the so-called secondary benefits, that go with CO2 reductions, i.e. the local impact of sulphate aerosols and conventional air pollution. The project had good seminal effects: articles were published and will be published in the international literature. The project researcher could, from the project results, contribute to the IPCC Working Group III 1995 report in Section 6 (The Social Costs of Climate Change) and 10 (Integrated Assessment) as one of the lead authors. The Climate Fund project is taken up in the Energy Modelling Forum 14: Integrated Assessment of Climate Change.
5.
APPLIED GENERAL EQUILIBRIUM MODEL
The aim of the project is to construct a general equilibrium model to supplement the macro-economic results of the Central Planning Bureau (CPB) calculations
1267 with more detailed information at the sectoral level, in particular with respect to distributional effects (60 firms and 44 households types). For this purpose, an existing general equilibrium model, developed by Keller at the Central Planning Bureau of Statistics, has been modified so as to be applicable to CO2 charges. The classification of energy inputs has been subdivided to allow for changes in the fuel mix by individual firms (and households). Another change in the model is an iterative computing lemma for adaptation by the target groups to relatively large changes in taxes (including CO2 charges). Since the model has a linear structure, it can originally only deal with small changes. In reality, the marginal d e m a n d functions of the target groups are non-linear. To avoid large linearisation errors, a relatively large tax impulse is subdivided in a series of small steps. After each step, the marginal adaptation behaviour of the target group is calculated, using the underlying non-linear relation. The model modifications and the updating of data as well as the gathering of additional data is now finalized. This has been a major effort. Four energy charge c u m rebate scenario have been analyzed. In the labour tax scenario a 50% charge on all energy carriers is implemented and the revenues are used to lower employers' contributions to social security. As a variant, the effects of a 100% charge were also calculated. Another scenario is the household scenario where only households are subject to the charge. In these two v a r i a n t s the revenues are recycled as in the labour tax scenario. The fourth variant is the 50% charge with lump sum recycling of the revenues. The main results are: 1. Energy saving in households is ca. 6% in the 50% scenarios and 9% in the 100% variant. Energy saving of firms varies widely between sectors (between 0 and 30% in the labour tax scenario) and is for all sectors together ca. 16% in the 50% scenarios and 25% in the 100% scenario. 2. There is no indication t h a t a sizeable double dividend (both lower energy use and higher employment) can be reaped. A small dividend can be realized in the household scenario. 3. In all scenarios where revenues were recycled as in the labour tax scenario, the effect on income distribution is increased inequality. Moreover the income differences between workers and non-workers are r a t h e r large. But in the lump sum scenario these effects are the other way around. In the evaluation of the project, it appears t h a t politically significant empirical results were derived: 1. Effects on the income distribution are significant in the chosen scenarios. As the income distribution is politically sensitive, more scenarios should be developed to investigate if this effect can be mitigated. 2. There are large differences in the energy savings and the economic impacts between the various sectors. These differences are more highly desegregated than in CPB's studies. 3. Energy savings are larger than calculated in the CPB studies. 4. No decisive conclusions can be derived on the existence of a significant double dividend.
1268 6.
EVALUATION
At the time of writing, all four projects are in the final stage. This makes it difficult to m a k e a definite judgement. The project on tradable emission rights (UG) has very interesting conclusions, in particular with respect to entry barriers, a subject t h a t has so far received little attention. It is reassuring t h a t tradable emission rights will not face major problems with respect to monitoring and enforcement. The work on international aspects is not yet m a t u r e for judgement, but the plans look promising. The project on the feasibility of an ecotax in The Netherlands (UL) is very much focused on practical aspects. This is commendable, because so far most literature on ecotaxes tends to be r a t h e r theoretical and to overlook such practical aspects. Although the results of the project are indeed very interesting, (to a certain extent) they seem to be of somewhat less relevance to the CO2 issue, at least for the n e a r future. The recommendation to introduce first ecotaxes at small scales (by lower authorities) is difficult to extend to CO2 taxes. However, in the final phase of this project, attention is especially directed toward the use of ecotaxes for climate change issues. And as long as world-wide coordinated policies are not feasible, these insights might be useful. The Climate F u n d project (IES) has yielded a lot more results t h a n could be foreseen at the start. The modelling effort for 9 regions in the world is of necessity based on rough assumptions. The stochastic analysis teaches us t h a t the existence of uncertainty leads to the necessity to take measures earlier, not later, t h a n one would do on the basis of information of the most probable effects alone; t h a t result m a y look familiar to statistically trained economists, but politicians often use uncertainty as an argument to delay action. The information gathered in Climate Fund can be used in last phase of the UG project t h a t also t r e a t s i n t e r n a t i o n a l effects. Right now, we have in The N e t h e r l a n d s the IMAGE model, t h a t is strong on physical effects but needs fortification at the economic side, and the Climate Fund model, that is basically an economic model and weak on the physical side. Integration of the two might result in a model that is strong on both sides. It is somewhat puzzling and contrary to the intuition, that Climate Fund yields the - only preliminary - results that an international fund for side payments does not lead to considerable improvements of CO2 reduction and lowering of reduction costs. This has to be investigated further. In the light of the above-mentioned preliminary result, it is understandable t h a t so far little attention was paid to practical, political issues of how such a fund should be institutionalized, which mechanisms can be applied, which side p a y m e n t s criteria could be used etc. This might be a follow-up study, if indeed it can be concluded that a fund mechanism is effective and efficient. The analyses carried out with the applied general equilibrium model of The N e t h e r l a n d s (IES) indeed show t h a t substantial energy savings are possible, but t h a t the income distributional effects are worrisome from a political point of view.
1269 The analyses also show that no substantial positive employment effects are to be expected from an ecological tax shift, at least not in a small open economy as The Netherlands. Taken together, the four projects have led to a substantial improvement of information, aimed at application in the decision making on CO2 reduction. A further extension of this information can be expected in the year to come. Finally, at the November workshop, mentioned in the Introduction, the following blind spots and research questions were identified: 1. There was a general feeling among the project coordinators and researchers, based on their research experience, that the practical problems and complexities of implementation of economic instruments still constitute the major impediment to their application. These problems and complexities are closely related to the following issues, which are all in need of further research. 2. The very limited social, and hence, political acceptance of economic instruments both at the national and international level. It was even feared that the present tendency to minimise the distributional effects and the negative competitive effects on industry may seriously paralyse economic instruments. The instruments then become ineffective in promoting the internalisation of environmental costs, fostering the restructuring of industry, s t i m u l a t i n g e n v i r o n m e n t a l l y - s o u n d technologies and/or achieving environmental policy objectives. 3. The imperfect operation of market forces, i.e. the existence of market failures in various respects. This is especially relevant for energy markets. These markets are characterized by government intervention, and a mixture of oligopolistic supply structures (with often a regulated distinction between production and distribution of energy) and diffuse and mobile demand structures. Moreover, a number of decisions have a long gestation period: a coal fire power plant or a nuclear energy plant last for 40 years or ore; the same holds for distribution networks. 4. The impact of uncertainty surrounding climate change on the decision making process and the design and implementation of environmental policy instruments, economic instruments in particular. 5. The influence of economic policy instruments on (environmentally-sound) technological development is not sufficiently analyzed and modelled. 6. Too often, research is carried out on one specific environmental policy instrument, whereas in reality a mix of instruments (direct regulation, communication, economic instruments) is applied. Research should rather be undertaken with respect to the optimal mix of policy instruments. 7.
REFERENCES
Jansen, H. and R. Dellink, 1994. Applied general equilibrium model. (Interimreport). Free University Amsterdam. Koutstaal, P.R., (eds.) 1992. Verhandelbare CO2 emissierechten in Nederland en de EG. ECOF, Groningen. Paulus, A.T.G., J.G. Backhaus and G. Meijer 1994. The feasebility of ecological taxation in The Netherlands. (Concept report). University of Limburg.
1270 Tol, R.S.J., T. van der Burg, H.M.A. Jansen and H. Verbruggen, 1995. The Climate fund; Some notions on the socio-economic impacts of greenhouse gas emissions reductions in an international context. (Concept-report). Free University of Amsterdam.
1271
NRPWS A S S E S S M E N T R E P O R T ON NATIONAL AND I N T E R N A T I O N A L I N S T R U M E N T S F O R G R E E N H O U S E POLICY T r a n s c r i p t of d i s c u s s i o n 1. Mr. Michaelowa (HWWA institute for Economic Research, Hamburg) informs to which project included the case study on joint implementation in the cement industry. Mr. Verbruggen answered that this was the case in the "Ecotax study" study of the University of Limburg. This was however only a very limited part of this study and it was therefore not payed separate attention to in his presentation. 2. Mr. Metz (Ministry of VROM) asked Mr. Bruggink was m e a n t with the term
global embedding. Mr. Bruggink answered t h a t he introduced the term. He m e a n t t h a t in light of the increasing dependency of global m a r k e t s and the strong i n t e r n a t i o n a l competition as a result of the opening-up of domestic economies, the relation between environmental policies and the problems related to international mobility of capital and trade balance deficits become more and more important. Mr. Metz asks w e t h e r this calls for i n t e r n a t i o n a l a g r e e m e n t s , f.i. fixed standards? Mr. Bruggink replies that this could be an outcome, but that in m a n y cases only a few companies are involved and t h a t it might be sufficient to take their reactions into account when designing the policies. 3. Mr. Midden (Eindhoven University of Technology) asks mr. Verbruggen for clarification of the link between certainty and the choice of policy. Mr. V e r b r u g g e n clarifies the issue by example: As long as the problem is unclear, soft instrument are preferred over economic instruments. 4. Mr. Metz s t a t e s t h a t the conclusion of tax reform project (University of Limburg) t h a t e a r m a r k e d taxes by environmental agencies are to be preferred, is r a t h e r opposing to the current practise. He asks w h a t can be learned from this without radically changing course? The researchers emphasise t h a t they only looked at the acceptability of the new taxes by the public. This acceptability is g r e a t e r w h e n a direct link between paying and receiving is introduced. 5. Mr. Metz asks how come t h a t in the Climate F u n d project capital transfers between the North and the South are of negligible importance? This result opposes other studies, and is also contra-intuitive.
1272 Mr. Tol (Institute for Environmental Studies, Amsterdam) answers that in the Climate Fund model each region has an own incentive in reducing emissions. This incentive follows from their intertemporal utility optimizing behaviour, and ultimately is based on the presumed utility functions. Utility functions are presumed to be equal among regions, implying that developed countries incentives to invest in emission abatement are not stronger than developing countries incentives. 7. Mr. Michaelowa (HWWA institute for Economic Research, Hamburg) asked whether in the deforestation project land reform could be a useful instruments to avoid deforestation. Researchers (Leiden University) confirm this.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1275
TRADEABLE CARBON PERMITS
P.R. Koutstaal
Department of Economics and Public Finance, Faculty of Law, University of Groningen, P.O. Box 716, 9700 AS Groningen, The Netherlands
Abstract The research project on tradeable carbon permits has focused on three
elements. First of all, the practical implications of designing a system of tradeable emission permits for reducing CO2 has been studied. In the second part, the consequences of introducing a system of tradeable carbon permits for entry barriers have been considered. Last, the institutional requirements and welfare effects of coordination of CO2 abatement in a second-best world have been examined.
1. Designing a system of tradeable carbon permits
One of the projects has been concerned with the study of a policy instrument for reducing CO2 emissions: tradeable carbondioxide emission permits. A tradeable emission permit is basically quite similar to the classical permit, which allows a polluter a certain amount of emissions. The only difference is that the permit is tradeable. Because of this feature of tradeability, total costs of emission reduction will be as low as possible, which is not necessarily the case when the permits are not tradeable. A source which can only abate it's CO2 emissions at high costs will have the opportunity to buy permits from a source which has low abatement costs. Both parties will be better off and aggregate abatement costs will fall. Another attractive feature of tradeable emission permits is that it can be easily combined with other policy initiatives like Joint Implementation. Suppose a system of tradeable carbon permits is operating in the European Union. A firm from the
1276 European
Union
which
acquired
emission
reduction
credits
through
Joint
Implementation with a country or firm in say Africa would be credited with an equal number of CO2 permits. Consequently, it can emit more or sell the permits. The study has focused on the practical issues related to the design of a workable system of tradeable emission permits for reducing carbondioxide emissions. Up till now, research has predominantly concentrated on the efficiency advantages of economic instruments like taxes and tradeable emission permits. Less attention has been paid to the feasibility of implementing such an instrument. However, this issue should not be neglected. Indeed, an instrument might theoretically be very attractive because it is cost minimising and because it will realise the policy target, but if it cannot be implemented it is of small practical value. Specific points which have been addressed are monitoring, enforcement, administration and acceptability of the instrument of tradeable carbon permits. The main conclusion is that a feasible system of tradeable emission carbon permits can be implemented in the European Union. Monitoring and enforcement need not be more complicated than when a carbon tax would be used to reduce CO2 emissions. Contrary to popular believe, emissions need not necessarily be monitored at the end of the pipe for a system of tradeable emission permits to work satisfactory. Instead, producers and importers of fossil fuels can be obliged to hand over carbon permits for the amount of carbon contained in the fuels they bring on to the market. This significantly diminishes the number of firms which have to be monitored. Instead of having to monitor not only every industrial source of CO2 but also all houses and cars, which is clearly impossible, the number of firms which have to be monitored would in the Netherlands be about 50 producers and importers of fossil fuel. A carbon reduction policy can lead to large costs and expenditures for especially energy intensive industries. The revenues which these firms have to pay when a tax is levied on carbon or when carbon permits are sold will be huge, reducing their competitiveness. This burden can be alleviated when the permits are handed out for free to firms (which is called grandfathering). This will reduce resistance of industry against a carbon reduction policy. Firms can hand over the permits they received (or bought from other firms) to their suppliers of fossil fuels. They can in turn hand them over to the monitoring authorities as described above.
1277 2.
Entry Barriers In addition to the study of the practical problems associated with tradeable
emission permits, two other fields have been investigated. First, the possibility that the introduction of a system of tradeable emission permits might create entry barriers has been examined. It has been argued that grandfathering permits to the existing firms will put new firms at a disadvantage because they have to buy the permits they need. It is important to understand that grandfathering permits does not reduce the production costs of a firm as compared with selling the permits. Even though the permits are received for free, they do have opportunity costs. If a firm could more profitably sell it's permits instead of using them for production, it would do so. However, established firms do not need to make the expenditure on the permits which new firms have to make. Under certain conditions (when capital markets do not work perfectly), this can constitute an entry barrier for the new firms. This will reduce industry dynamics, which in turn can reduce R&D effort, for example R&D on energy efficiency measures. It is difficult to estimate the size of this potential entry barrier, but our guess is that it will be small.
3. International coordination of COz abatement Secondly, a closer look has been taken at the issue of international coordination of carbon reduction policies. The greenhouse problem is a truly worldwide problem and therefore need to be addressed at an international level. This poses additional policy problems as compared with policies which only need to be implemented within a single country because countries will have to coordinate their efforts. A complicating factor is that countries already tax fossil fuels for different reasons. Both carbon taxes and tradeable carbon permits will interact with these taxes on fossil fuels. The question which arises is how to combine a carbon tax with existing taxes on fossil fuels. This problem is analyzed in a theoretical second-best two country model in which the governments of both countries have as their objective to reduce environmental damage and to raise revenue. Special attention is paid to the role of sidepayments in policy agreements. The results show that agreements will differ
1278 considerable between a second-best and a first-best world as regards who pays whom and the consequences for pollution. An interesting result is that allowing for sidepayment can under certain conditions increase pollution compared with an agreement on coordinated abatement without sidepayments, both in a second-best and in a first-best world. In one variant studied, countries have exogenously set emission targets instead of including an environmental damage function in their utility function. This corresponds more closely with the idea of Joint Implementation (JI), a policy initiative described in the Framework Convention on Climate Change. It is analyzed
what the specific institutional
arrangements
necessary for a well
functioning cooperated abatement policy are. In an efficient JI agreement, countries should also agree on how existing taxes on fossil fuels may change. If this is omitted, countries will change their current taxes as they will see fit and as a result JI will not be cost-efficient.
4.
Conclusions
It is possible to design a system of tradeable carbon permits for the European Union in which monitoring and enforcement will perform at the same level as with a carbon tax. Moreover, tradeable permits are both efficient and effective. Grandfathering permits can increase entry barriers for new firms who want to enter a market, but this problem appears to be small. In designing an international agreement, attention should be paid to the consequences existing taxes on fossil fuels have for the role of sidepayments and the institutions needed for realising an efficient agreement.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1279
The feasibility of ecological taxation A.T.G. Paulus a aUniversity of Limburg, Faculty of Economics and Business Administration, P.O. Box 616, 6200 MD Maastricht, The Netherlands
Abstract From an analysis of the possibilities and complexities of ecological taxation, conducted within the context of the first NRP (research project 851051E), it follows that the feasibility of ecological taxes is determined by their design, the level at which they are implemented, the taxing authority by which they are imposed and by the constitutional, institutional and fiscal framework in which they are embedded.
1. E C O L O G I C A L TAXATION: POSSIBILITIES AND COMPLEXITIES Generally, ecological or environmental taxes can be defined as types of levies (i.e. compulsory contributions or forms of payments) which conform with one or more of the following characteristics: In order to incorporate negative external environmental effects into the decision calculus of the externality generator, taxes are aimed at accomplishing tax avoiding changes in the behaviour of those who are confronted with these taxes. In order to achieve a particular environmental goal, taxes are aimed at raising an amount of revenues that suffices to fund specific environmental policy measures. There is an intended or unintended relationship between the tax and the nature and size of activities or elements that are relevant from an environmental point of view.
Being defined as such, ecological taxes can be included in a broader category of ecologically relevant taxes. This category also includes taxes initially introduced for non environmental reasons but with a direct or indirect (un)desirable impact on the environment. Following this description of ecologically relevant taxes, it can be shown that governments have several options to introduce ecological taxation. More specifically, options include the supplementation of the existing tax system with new ecological taxes (e.g. earmarked taxes or regulatory charges) or the re-orientation or replacement of existing taxes and charges by ecological ones. The latter option is also known as ecological tax reform. To introduce these options, however, governments face important complexities. These are related, among others, to the ecological effectiveness and the administrative costs of these taxes, the linkage between the tax point and the point of pollution, the informational needs required to set the proper tax rates, their fiscal and steering effects and to the earmarking of ecological tax revenues. Further complexities arise with the concor-
1280 dance of ecological taxation with existing national and international tax and legal systems. Additionally, complexities are associated with the distributional incidence and the national and international economic effects of ecological taxes.
2. THE DESIGN AND IMPOSITION OF FEASIBLE ECOLOGICAL TAXES Given these complexities and possibilities, which differ per type of ecological tax, it follows that the use of taxes for ecological purposes comprises several complicated problems. These problems can be solved only if the ecological tax as a policy instrument is placed within a general policy framework. This can be achieved by using an ecological tax in the form of a regulatory charge as an instrument of environmental policy. In this case it is important to formulate and search for an accommodative policy that eases the making of behavioural changes. Outside the existing tax system, also new earmarked environmental taxes can be introduced. In this case it is important to find a proper relationship between the taxes, tax liable individuals and the provision of an ecologically relevant service. In addition, the possibility exists to reform the entire tax system in order to stimulate an optimal use of natural resources. If ecological taxes of the ordinary revenue raising type are used, this in fact opens a political discussion on the entire tax system. From this it follows that ecological taxes are complicated, yet that in principle these taxes could be feasible when properly designed and when coherent with certain preconditions. Besides being properly designed and coherent with certain preconditions, ecological taxes, in order to be feasible, acceptable and effective from an ecological point of view, have to be imposed on a level and by a taxing authority or decision making unit that is closely related to relevant ecological problems and circumstances. In addition, ecological taxes have to be imposed on a level which minimizes undesirable economic effects in terms of competitiveness and costs of control, information and transactions. In order to be feasible, ecological taxes also have to be embedded within a constitutional, institutional and fiscal framework (whether governmental or club like) in which there is a close relationship between those who pay and benefit from the taxes in question. Ecological tax units, i.e. taxing units which are closely related to relevant ecological circumstances, are the ideal representatives of such decision making arrangements. Since it can be expected that existing taxing authorities tend to be political units which do most of the times not fit the description of being clearly related to relevant ecological circumstances, this calls for a dynamic process in which a proper sized level for ecological taxation is gradually created via different types of co-ordination, co-operation and environmental diplomacy.
3. ECOLOGICAL TAXATION AND GLOBAL CLIMATE CHANGE The dynamic process in which a proper sized level for ecological taxation is gradually created via different types of co-ordination, co-operation and environmental diplomacy seems especially important for global environmental problems such as, for instance, the global climate change problem. Since single countries will probably face a competitive disadvantage once unilaterally introducing taxes on greenhouse gas (G.H.G.) emissions, the call for a multilateral or international introduction of these taxes can be
1281
expected to increase in the future. Within the context of the dynamic process described above, the analysis conducted within the framework of the above mentioned research project shows that taxes that are intended to be used with regard to the problem of global climate change are preferably imposed and introduced at a surveyable level, i.e. at the level of a particular country or a small group of countries. The imposition of taxes at this level can then serve as an important starting point and role model for GHG taxes that are intended to be introduced within an international greenhouse gas taxing arrangement. In this sense, the introduction of ecological taxes by single countries plays an important intermediate role in the dynamic process described above. Within this stepwise process, GHG taxes can first be introduced by single countries. Via co-operation, co-ordination and environmental diplomacy it is then possible for GHG taxes to be embedded within an international climate change agreement, which enhances the possibilities for these taxes to be imposed by an authority and at a level which is clearly related to relevant ecological circumstances.
4. E C O L O G I C A L TAXATION: THE NETHERLANDS In the Netherlands, the existing governing and financial framework for ecological taxation forces particular restrictions upon the level and authority by which these taxes can be imposed, upon the design and purpose of these taxes and upon the possibilities to differentiate the imposition of ecological taxes throughout the country. An analysis of the actual use of ecological taxation in the Netherlands shows that ecological taxation in this country is gradually developing, via a stepwise process, from a system in which different sectoral environmental levies are introduced within a relatively unco-ordinated manner towards a more structured system in which environmental levies are introduced on the basis of relevant ecological circumstances. It also follows that ecological taxation in the Netherlands has now entered the phase in which it is tried to accomplish a structured financing system for environmental policy. At the same time, more room is created for the introduction of ecologically relevant taxes and tax elements that are not primarily intended to be used for fiscal purposes. At this point in time, the greater part of the ecologically relevant levies that have been introduced in this country share the characteristic of being relatively small earmarked types of taxes that have relatively low tax rates. Raising a proper amount of tax revenues that can be used to fund relevant environmental expenditures is the main purpose of the greater part of environmental levies in this country. In the Netherlands, only the imposition of ecologically relevant levies by water authorities closely reflects the idea of ecological tax units. Generally, however, there are few opportunities in this country to connect existing taxes and levies to relevant ecological circumstances. An analysis of these opportunities within the context of the above mentioned research project shows that such opportunities are mainly provided by the tax systems of municipalities, provinces and water authorities. With regard to the feasibility of additional ecological taxation in the Netherlands this implies that the possibilities and opportunities for these sub national governments to further use and introduce ecologically relevant taxes should be enlarged in the future.
1282 5. THE FEASIBILITY OF ECOLOGICAL TAXATION: CONCLUSIONS From the analysis of the feasibility of ecological taxation in the above mentioned research project it can be concluded that sub national governments can play a significant role in providing starting points for the introduction, imposition and use of ecologically relevant taxes which are intended to be introduced within a broader national, transnational or international context. More specifically, it can be concluded that sub national government bodies fulfil an important exemplifying, connecting and starting role within a dynamic process in which these taxes are gradually embedded within their proper institutional, fiscal and constitutional structures. The latter, as this research project shows, is a precondition for feasible ecological taxes, i.e. for (alternative) courses of policy action in which thoroughly and carefully designed ecological taxes are embedded within institutional, constitutional and fiscal structures that allow and facilitate the ecological effectiveness and acceptability of these taxes.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
Socio-economic Fund
a s p e c t s of t h e g r e e n h o u s e
1283
effect: C l i m a t e
R.S.J. Tol, T. van der Burg, H.M.A. Jansen and H. Verbruggen a
aInstitute for Environmental Studies, Vrije Universiteit, De Boelelaan 1115, 1081 HV Amsterdam, The Netherlands
Abstract The project S o c i o - e c o n o m i c aspects o f the g r e e n h o u s e effect: C l i m a t e f u n d studies the impact of international capital transfers on the efficiency and efficacy of greenhouse gas emission reduction. The absolute costs of emission abatement is substantially lower in less developed countries. The associated reduction of the damage due to conventional air pollution is higher in the richer countries in both absolute and relative terms. The costs of climatic change are relatively higher (but absolute lower) in the developing countries. Prime impacts are on agriculture (in the developing world) and hum an health (highly valued in the developed world). Costs of emission reduction and climatic change are joined in a nine region, quasi-Ramsey, integrated climate-economy model, called F U N D . The first calculations with this model show that the (hardly known) dynamics of climate change and the great uncertainties play a critical role, that free riding behaviour need not be as prominent a problem as is generally believed, and that international capital transfers do not seem to substantially influence the optimal emission control, as the regions most interested in climate change do not have much capital to transfer. Negotiated emission caps are likely to alter this conclusion.
1.
INTRODUCTION
The project S o c i o - e c o n o m i c aspects o f the g r e e n h o u s e effect: C l i m a t e f u n d is centred around the question: What is the impact of international capital transfers on the efficiency and efficacy of greenhouse gas emission reduction measures? The reason for asking this question is clear: Those countries which are willing and able to pay for greenhouse gas emissions are not necessarily the same as those who can establish this in an efficient and effective manner. The simple fact that enlarging the number of reduction options by allowing international capital transfers lowers, at least not raises, reduction costs is an important motivation. Moreover, climatic change is a global problem, and can only be controlled by international action. This is true because the impact of most countries on the climate is small, individual measures could well disadvantage domestic competitiveness without reducing emissions, and the
1284 risk of free rider behaviour is high. International capital transfers might convince countries to participate in internationally concerted climatic change a b a t e m e n t action. Capital transfers avoid many of the problems associated with joint implementation and emission permits in that emission targets and baseline paths need not be explicitly assessed. A n u m b e r of matters need to be discussed in order to investigate the impact of international capital transfers on greenhouse gas emissions. First, the international distribution of the costs and benefits of emission reduction need to be studied (since emission reduction is assumed to be based on a cost-benefit analysis). As little attention has been paid in the literature to the benefits of emission a b a t e m e n t (benefits stand for avoided damage costs of climatic change) and more to the costs, the focus is on the benefit side. This m a t t e r is dealt with in Section 2. Section 3 discusses the second phase of the study: The design of an integrated climate economy model, the Climate Framework for Uncertainty, Negotiation and Distribution, which is used to calculate the optimal emission control for nine major world regions with and without capital transfers. The results of this model, FUND, are presented in Section 4. Section 5 concludes the paper. A number of smaller activities took place along the main line of research, j u s t described. Their results are here briefly presented where appropriate. The subjects covered are the social rate of discount, the dynamics of climate change damage costs, and the impact of uncertainties on optimal emission control.
@
T H E C O S T S AND B E N E F I T S O F E M I S S I O N R E D U C T I O N IN AN INTERNATIONAL PERSPECTIVE
The brief discussion on the costs of greenhouse gas emission reduction in this section draws on Tol (1993a, 1994a). The discussion on the costs of climate change and the benefits of the emission reduction draws on Dorland et al. (1994), J a n s e n (1993), Pearce et al. (1995) and Tol (1993a, 1994a,c,d, 1995a). 2.1.
T h e C o s t s of E m i s s i o n R e d u c t i o n Many studies on the costs of reducing the emissions of carbon dioxide by the burning of fossil fuels suggest that the costs are modest or even negative for an emission cut of about 25%. This is caused by the present inefficiency of the energy m a r k e t and the cheap alternatives (primarily energy switching and saving) available. This goes for most countries, developed, developing and transitional alike. Cost differences between countries seem to be mainly induced by differences in absolute levels of economic prosperity. Knowledge on poorer countries is substantially more limited than knowledge on richer countries, however. The secondary benefits of emission control, i.e., reductions of conventional air pollution, are much higher in the richer countries in both relative and absolute terms. Generally, the costs per tonne of carbon sequestered through afforestation and slowing deforestation is lower in tropical, poorer countries than in extratropical, richer countries because of the lower prices per hectare and higher CO 2 uptake per hectare in the tropics.
1285 2.2.
T h e D a m a g e Costs of C l i m a t e C h a n g e Table 1 contains the estimated socio-economic costs of 2xC02 climatic change for the present day economy for nine damage categories and nine world regions. Non-market impacts are assessed using (approximate) willingness to pay and willingness to accept compensation methods. Table 1 shows that the impact of climate change on human mortality is the largest category. The sensitive issue what value to place on a statistical life is thus very prominent in the climate debate. Table 1 also shows that the poorer regions, which contributed very little to the past built up of the atmospheric concentration of greenhouse gases, are substantially more vulnerable to climatic change.
Table 1 Total damage costs of 2xC02 climate change damage category
(1095)
region
(1095)
(%GP)
coastal defence dryland loss wetland loss species loss agriculture amenity life/morbidity migration natural hazards
9.5 9.8 18.8 22.0 14.5 38.0 188.0 13.8 1.4
OECD-America OECD-Europe OECD-Pacific Eastern Europe and former USSR Middle East Latin America South Asia China Africa
74.0 56.5 59.0
(1.5) (1.3) (2.8)
-7.9 1.3 31.0 53.6 18.0 30.3
(-0.3) (4.1) (4.3) (8.6) (5.2) (8.7)
total
315.7
World
315.7
(1.9)
2.3.
F r o m B e n c h m a r k D a m a g e Costs to Benefits of A b a t e m e n t The previous section discussed the best guess impact of 2xCO 2 climate change on the present day economy. This is not what a decision maker would be interested in. First of all, 2xC02 is quite an arbitrary benchmark; it will certainly not affect the present day economy. In Tol (1994f, 1995a,b), I argue that climate change damage is dynamic and these dynamics do matter. Six types of dynamics are distinguished: Non-equilibrium climate change socioeconomic vulnerability, damage valuation, damage accumulation, learning, and higher-order impacts. Graph 1 displays the optimal carbon emission control according to Nordhaus' D I C E model in the base case (U), in case the intangible damage is allowed to influence only the utility (U'), and in case the intangible damage is assumed to grow linearly with per capita income (U"). Second, decision makers are not interested in best guess but in expected damage. The mean damage is much larger than the best guess damage because most uncertainties are positively skewed, most damage functions are convex, and the uncertainties cascade through many levels. Figure 2 displays some
1286 numerical examples of best guess damages and expected damages under limited (optimistic) and profound (pessimistic) uncertainties. 0.45
25000
0.4-
~o.3~ I
0.3
~
0.25
20000
J
t 15000 10000
0.2 ~.0.15
U
0.1 ~
5000
U'
O, 1990
............................. 1995 2035 2075 2115 2155 2195 2235 2275
0.05
2010
Figure 1. Optimal emission control in D I C E for three different welfare functions.
t
2030
2050
2070
2090
yea"
yeor
Figure 2. Best guess versus expected damage according to the damage module of F U N D .
THE CLIMATE FRAMEWORK N E G O T I A T I O N AND D I S T R I B U T I O N
FOR
UNCERTAINTY,
Here we discuss the set-up of the central tool: The F U N D model. This model is first described by Tol (1994a) and subsequently improved in Tol (1994b,c,d,e,g). Here version 1.4 is discussed. A further revision and a userfriendly version are to be presented early 1995 (Tol, 1995c). F U N D is a nineregion, quasi-Ramsey closed-loop, integrated climate-economy model for the period 1990-2100. For a general discussion of the current generation of integrated models the reader is referred to Weyant et al. (1995). Figure 3 displays a flow diagram. Most boxes have very simple representations in the model. F U N D is capable of calculating optimal greenhouse gas emission control under certainty and uncertainty, for cooperative and non-cooperative games, and with and without interregional capital transfers. Instruments include fiscal and regulatory measures, and afforestation. The discount rate is one of the most crucial parameters in climatic change cost-benefit analysis but general rules for its value cannot readily be derived (van der Burg, 1993).
4.
O P T I M A L G R E E N H O U S E GAS EMISSION C O N T R O L
The amount of emission control according to F U N D is rather high in all the model's optimisation settings. One reason is that version 1.4 only considers emission reduction in the period 1990-2000. Under uncertainty and in the cooperative game, the optimal reduction is higher. As opposed to the literature, where free riding is seen as a major hurdle to arrive at an internationally concerted emission abatement strategy, F U N D also points at the opposite
1287 behaviour: In some cases it is economically rational to abate more if other regions do the same (Sen's assurance game). As a result of non-linear feedbacks, the marginal damage function is not monotone. The reduction is sensitive to the model parameters, such as the climate sensitivity, the costs of climate change and the costs of emission abatement. In the OECD regions, the secondary benefits, i.e., the reduction in the costs of conventional air pollution, is a very important motivation for emission reduction. Interregional capital transfers do not seem to play a very critical role in emission reduction. Possible explanations are that secondary benefits are transfered as well, capital transfers are costly, and the regions that benefit most from emission reduction have the least to transfer. non-CO2 emissions
t
T economiCgrowth
climate
T t
I
i j
J
impact
l
population growth
J
economy
1 population
agriculture
T aeei acei
emission abatement
J
t
decision makers
f
welfare function
Figure 3. Flow diagram of FUND. 5.
CONCLUSION
Despite a wide divergence in the regional costs and benefits of greenhouse gas emission control, and despite substantial optimal emission abatement, interregional capital transfers are not found to play a critical role in the emission reduction game. The prime reason appears to be that the regions which have a potential incentive to transfer capital, i.e., the OECD, is more interested in combatting conventional air pollution than climatic change. This situation is likely to change if fixed (instead of optimal) emission trajectories are negotiated, a topic to be studied in the near future. 6.
REFERENCES
Burg, T. van der (1993), On the Social Rate of Discount, Institute for Environmental Studies W93/03, Vrije Universiteit, Amsterdam.
1288 Dorland, C., R. Hoevenagel, H.M.A. Jansen and R.S.J. Tol (1994), The Dutch Coal Fuel Cycle, Institute for Environmental Studies, Vrije Universiteit, Amsterdam (in preparation). Jansen, H.M.A. (1993), 'Are We Underestimating, When Valuing the Benefits of Greenhouse Gas Emission Reduction?' in Y. Kaya, N. Naki6enovi6, W.D. Nordhaus and F.L. Toth (eds.), Costs, Impacts, and Benefits of COs Mitigation, International Institute for Applied Systems Research, Laxenburg. Pearce, D.W, A.N. Achanta, W.R Cline, S. Fankhauser, R. Pachuari, R.S.J. Tol and P. Vellinga (1995), 'The Social Costs of Climate Change: Greenhouse Damage and the Benefits of Control', in IPCC WGIII Second Assessment Report (in preparation). Tol, R.S.J. (1993a), The Climate F u n d - Survey of Literature on Costs and Benefits, Institute for Environmental Studies W93/01, Vrije Universiteit, Amsterdam. Tol, R.S.J. (1994a), The Climate F u n d - Modelling Costs and Benefits, Institute for Environmental Studies W93/17, Vrije Universiteit, Amsterdam. Tol, R.S.J. (1994b), The Climate F u n d - Interregional Capital Transfers, Institute for Environmental Studies W94/02, Vrije Universiteit, Amsterdam. Tol, R.S.J. (1994c), The Climate F u n d - Optimal Greenhouse Gas Emission Abatement, Institute for Environmental Studies W94/08, Vrije Universiteit, Amsterdam. Tol, R.S.J. (1994d), The Climate Fund -Sensitivity, Uncertainty and Robustness Analyses, Institute for Environmental Studies, Vrije Universiteit, Amsterdam (in preparation). Tol, R.S.J. (1994e), The Climate Fund, Institute for Environmental Studies, Vrije Universiteit, Amsterdam (in preparation). Tol, R.S.J. (1994f), 'The Damage Costs of Climate C h a n g e - A Note on Tangibles and Intangibles, Applied to DICE', Energy Policy, 22 (5), 436-438. Tol, R.S.J. (1994g), The Climate FUND, versions 1.41 and 1.42, report to Energy Modeling Forum - Integrated Assessment of Global Climate Change. Tol, R.S.J. (1995a), 'The Damage Costs of Climate C h a n g e - Towards More Comprehensive Calculations', Environmental and Resource Economics (forthcoming). Tol, R.S.J. (1995b), The Damage Costs of Climate Change - Towards a Dynamic Representation, Institute for Environmental Studies, Vrije Universiteit, Amsterdam (in preparation). Tol, R.S.J. (1995c), The Climate Framework for Uncertainty, Negotiation and Distribution - Description and Application of an Integrated Climate-Economy Model, Institute for Environmental Studies, Vrije Universiteit, Amsterdam (in preparation). Weyant, J., W.R. Cline, 0. Davidson, S. Fankhauser, T. Parson, R. Richels, P.R. Shukla and R.S.J. Tol (1995), 'Integrated Assessment' in IPCC WGIII Second Assessment Report (in preparation).
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1289
S o c i o - e c o n o m i c a s p e c t s of the g r e e n h o u s e effect:
Applied general equilibrium model R. Dellink, F. Groot, H. Jansen and H. Verbruggen Institute for Environmental Studies, Vrije Universiteit, De Boelelaan 1115, 1081 HV Amsterdam
Abstract
To assess the economic consequences of environmental taxation a general equilibrium model is applied. The model contains 60 firm sectors and 44 household groups, which makes it especially suitable to analyse the sectoral and distributional effects of environmental taxes. These sectoral effects are rather large and diverse in comparison to the macro-economic consequences. After a short overview of the relevant literature, the original model and the model adaptations are described. These model adaptations include an iterative procedure to avoid substantial linearisation errors when large impulses are simulated. Possible model simulations are identified and the working programme is presented.
1.
Introduction
The economic models of the Central Planning Bureau have been used to estimate effects of a CO2 charge. The CPB models yield macro-economic consequences, but are not sufficiently capable to analyse distributional consequences. Disaggregated general equilibrium models, assuming an optimizing behaviour of the target groups, are better suited to analyse the long term tax incidence. In this sense, an applied general equilibrium model is complementary to the CPB models. In this project (NOP 851061E), the so-called Keller model is being modified and applied to assess the sectoral economic consequences of a energy/CO2-tax on both firms and households. 2.
O v e r v i e w of the literature
The literature on the economic consequences of environmental taxation may be divided into two categories. First, there is a theoretical approach, which uses analytical general equilibrium models to analyse the effects of increased environmental concern on macro-economic variables like production growth, employment and environmental quality in a second-best framework. If labour is immobile and goods, capital and natural resources can be freely traded, the burden of the environmental tax is transfered completely to labour, and is thus an implicit labour tax, which is less efficient than an explicit labour tax. Consequently, employment declines and environmental quality improves mainly due to lower production. If on the contrary capital is the fixed input
1290 and involuntary employment exists at fixed wages, the fixed factor bears part of the burden of the environmental tax. In this case, employment may rise in combination with higher environmental quality (see Bovenberg and van der Ploeg, 1993). In addition, Bovenberg and de Mooij (1993) argue that part of the burden of an environmental tax in combination with lower labour taxes may be transferred to non-labour income, thereby decreasing unemployment at the cost of greater inequality in income distribution. Secondly, there are several empirical models. These models can be sub-divided into three different approaches. The first sub-category, the aggregated general equilibrium long-term world models, is the focus of an other NOP-project carried out at the IVM (The Climate Fund, see Tol, 1993) and will not be discussed here. A second subcategory are the applied general equilibrium models (e.g Jorgenson and Wilcoxen, 1993; OECD, 1994). These models show on average an slightly negative effect of environmental taxation on the economy. A last sub-category that can be identified contains the macro-economic disequilibrium models. The Dutch Central Planning Bureau used a macro-econometric model to investigate the economic consequences of regulating energy levies (CPB, 1992). Their results are rather negative: large economic damage with relatively small environmental benefit. This results hinges on the (exogenous) movement of industries outside the taxed area. Other macro-economic studies, like the HERMES-study for the European Union (e.g. European Commission, 1993) don't take industry-movements into account and find slightly positive results for both production growth and employment. 3.
The original model
The original Keller model is a comparative static general equilibrium model, which distinguishes between different production sectors and different household groups. It is based on optimising behaviour of the target groups, for whom demand equations are specified. The public sector is sub-divided into three sectors, i.e. a household sector for public consumption, the public services firm and a so-called "fisc" to deal with all tax payments. One of the household sectors in the model is called "rest of the world", including all international transactions (see Keller, 1980). Capital may be specified as mobile or immobile. Furthermore, households may be rationed (see Cornielje, 1990). For every good a market exists in which the price mechanism ensures that demand is equal to supply. There are three types of goods in the model. First, there are the primary inputs, which are provided by the households (including the public sector and rest of the world). These primary inputs, a.o. labour services and net imports, are demanded by firms. Firms provide the other two types of goods, intermediate goods and final output. These two types of goods only differ with respect to the consuming sector: intermediate goods are demanded by firms and final output is demanded by the households. The original purpose of the model is to analyse the incidence of small changes in the taxing structure. The equations in the model are marginal and linear. This means that the marginal behaviour of the agents is specified, using a locally defined linear approximation of the underlying global non-linear micro-economic relations.
1291 4.
An updating procedure
As said before, the original model is only locally defined, since the equations are linear approximations of global non-linear relations. The results of simulations wth the linear model therefor only hold exactly for infinitesimally small impulses. However, environmental taxes are likely to be of substantial magnitude. The results thus obtained are biased, or in other words, a linearisation error is made. To reduce this linearisation error a specific updating procedure is required. The large total tax impulse is broken down into several small steps and the consumer and producer elasticities are re-evaluated after each step. In this way, the advantages of the linear model are retained, while the re-evaluation of the parameters on basis of the global non-linear model reduces the linearisation error substantially (see Bovenberg and Keller, 1984 for technical details). The graph shows an example where the linearisation error is brought down from 0-1 to 0-2 when a two-step procedure is applied (i.e. one update). Figure 1.
Linearisation errors in a demand equation
i
.
i i i
! ! !
.
i i
i i
i i
po
5.
0
p~
Data adaptations
The original data set which was available needed adaptations on several points. First, the old data set was for 1981. Unfortunately the most recent disaggregated household data available at this moment are on 1988, so a full new dataset for 1988 has been constructed. Energy inputs for both households and finns were disaggregated, to allow for more detailed analysis of the economic consequences of energy and CO2 levies. The new dataset contains 60 firms, 44 private households, 88 intermediate goods and 7 primary inputs. The input-output table has been transformed into so-called total accounts, i.e. a sector times sector matrix. As said before, each firm uses primary inputs and intermediate goods to produce one and only one homogeneous output. In an analogous manner household data (including the public sector and rest of the world) are transformed from a 46 x 95 matrix into a 46 x 67 total account.
1292
6. Model simulations The (adapted) model is capable of analysing various simulation alternatives. Besides the usual choices concerning height of the tax (e.g. 50 percent), base of the tax (based on energy-content vs. based on CO2-content) and destiny of the tax (e.g. lower VAT, lower labour taxes, an international fund, etcetera), our sectoral model can cope with differential taxes, where some energy-users are faced with lower or no environmental taxes (tax exemption for energy-intensive export-oriented industries is the most likely example). Furthermore, the model is capable of analysing the effects of specific policies, like an abolishment of energy price differentiation between sectors. In coordination with the State University of Groningen, possible simulations on tradeable emissions permits will be identified.
7. Working programme In the next months the proposed simulations will be run with the new calibrated model. We expect that these simulations will show the diverse and large sectoral effects of environmental taxes, which are are not reflected in the traditional policy-oriented macro-economic models. The final report will be completed Spring 1995.
8. References Bovenberg, A.L. and W.J. Keller, 1984 - 'Non-linearities in applied general equilibrium models', Economic Letters 14, pp. 53-59. Bovenberg, A.L. and R.A. de Mooij, 1993 - 'Environmental Policy in a Small Open Economy with Distortionary Labor Taxes: a general equilibrium analysis', OCFEB research memorandum 9304, Rotterdam. Bovenberg, A.L. and F. van der Ploeg, 1993 - 'Optimal Taxation, Public Goods and Environmental Policy with Involuntary Unemployment', CentER discussion paper 9377, Tilburg. Cornielje, O.J.C., 1990 - Rationing and capital mobili(y in applied general equilibrium models, VU University Press, Amsterdam. CPB, 1992 - 'Economische gevolgen op lange termijn van heffingen op energie', CPB working document 43, The Hague. European Commission, 1993 - Taxation, employment and environment: fiscal reform for reducing unemployment, Brussels. Jorgenson, D. and P. Wilcoxen, 1993 - 'Reducing U.S. carbon dioxide emissions: an assessment of different instruments', Journal of Policy Modeling 15, pp. 491-520. Keller, WJ., 1980, - Tax incidence: a general equilibrium approach, North-Holland, Amsterdam. OECD, 1994 - 'The OECD GREEN model: an updated overview', OECD Development Centre technical paper 97, Paris. Tol, R.SJ., 1994 - 'The climate fund', VU/IES mimeo, Amsterdam.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1293
ASSESSMENT REPORT ON NRP SUBTHEME "INTERNATIONAL INSTRUMENTS FOR C L I M A T E C H A N G E P O L I C Y ~'
J.J.C. Bruggink Netherlands Energy Research Foundation (ECN) Policy Studies P.O. Box 1 1755 ZG Petten The Netherlands
With contributions by: J.C. Jansen P. van Beek, H. Folmer, Z.X. Zhang K. Blok, D. Phylipsen, E. Worrell J. Gupta, G. Junne, R. van der Wurff
ECN, Netherlands Energy Research Foundation, Petten LUW, Agricultural University of Wageningen RUU, University of Utrecht UvA, University of Amsterdam
1294
Contents Abstract Introduction 1.1 International climate change policy in the Netherlands 1.2 North-South relations as research priority 1.3 Structure of the assessment
0
International policies to address the greenhouse effect 2.1 An application of the theory of international regime formation 2.2 National case studies on climate change policies 2.3 Strategic choices for international negotiations 2.4 Evaluation of project results
0
Local actors and global tree cover policies 3.1 Integrating micro-oriented, site-specific studies with macro-oriented political studies 3.2 National case studies on deforestation processes 3.3 Emphasis on transition towards sustainable agricultural development 3.4 Evaluation of project results
0
Strategies and instruments to promote energy efficiency in d e v e l o p i n g countries 4.1 Survey of experiences and strategies for efficiency improvement 4.2 Regional case studies on industrial energy conservation policies 4.3 Policy priorities for developing countries 4.4 Evaluation of project results
0
Compatibility of CO2-emission reduction targets with long-term e c o n o m i c development in china 5.1 Computable general equilibrium modelling and power capacity planning combined 5.2 Analysis of the Chinese energy system 5.3 Merits of different approaches to CO2-emission reduction cost estimates 5.4 Evaluation of project results
0
0
e
8.
Evaluation of guidelines for sharing of international CO2-emission budgets 6.1 An international, statistical comparison of industrial energy efficiencies 6.2 Establishing a basis for emission reduction agreements 6.3 Comparative analysis of efficiencies in electricity production and industrial sectors 6.4 Evaluation of project results General conclusions 7.1 Programme effectiveness 7.2 Programme quality References
1295 ABSTRACT The projects implemented in the Dutch National Research Programme on Global Air Polluttion and Climate Change are organised in several t h e m e s and subthemes. Within the theme on Sustainable Solutions five projects are grouped under the heading International Instruments for Climate Change Policy. These five projects deal primarily with issues concerning the position of developing countries in the debate on limiting global CO2-emissions. They cover a broad spectrum of topics: international negotiation strategies, tropical deforestation, industrial energy conservation, national energy scenarios, emission guidelines. This contribution presents an overview of the objectives, methodologies and results of the projects and includes a critical evaluation of the potential relevance of the work for policy makers. 1.
INTRODUCTION
1.1 International climate change policy in The Netherlands International climate change policy arrived on the Dutch political agenda in the late 1980's, particularly after the establishment of the Intergovernmental Panel on Climate C h a n g e (IPCC) and the Toronto Conference on the C h a n g i n g Atmosphere, where a 20% reduction t a r g e t for CO2 by the y e a r 2005 was recommended. Although no firm national commitments were considered at t h a t time, the Ministry of Environment opted in favour of precautionary measures in case of international consensus. International environmental diplomacy should pave the way for a leap forwards in domestic environmental policies. The 1989 Noordwijk Conference on Atmospheric Pollution and Climate Change was intended to secure such international consensus for a precautionary approach. The belief that precautionary measures were only possible in case of international consensus prevailed until the publication of the first National Environmental Policy Plan in 1989, when climate change policies began to influence domestic environmental policies in a major way. Political parties began to view climate change as an election issue and the emphasis shifted from international diplomacy to domestic policy formulation. The national goal of stabilization by 2000 was reformulated and a new "plus" version of the National Environmental Policy Plan was adopted in 1990 with a 3%-5% reduction target by 2000. Interministerial debates on actual domestic policies, particularly regarding energy efficiency, replaced the earlier concentration on international environmental diplomacy. At the time of the 1992 Rio Conference on Development and Environment, the Netherlands had already established targets exceeding the stabilization goal of the Framework Convention on Climate Change. The Ministry of Economic Affairs, responsible for energy policies, had established an ambitious programme for promoting energy efficiency and discussions on the feasibility of unilateral domestic tax measures caused considerable controversy among policy makers. After 1990 the earlier optimism on i n t e r n a t i o n a l consensus decreased.and domestic i m p l e m e n t a t i o n bottlenecks required close attention. Gradually the implications of moving ahead out of line with international climate change policy p r o g r e s s i o n became clear. Insistence on the 5% domestic t a r g e t became unfashionable, when reaching the 3% target appeared difficult. Moreover, attention
1296 shii~ed away from the political complexities of reaching international consensus on reduction targets among industrial nations to the role and position of developing nations in the climate change debate. The second National Environmental Policy Plan appeared in 1993, but it did not introduce any new domestic measures. Instead, it showed renewed interest for the role of international environmental diplomacy. The potential role of international i m p l e m e n t a t i o n i n s t r u m e n t s r a t h e r t h a n specific targets for reduction now obtained attention. The deliberations of the Internationale Negotiation Committee (INC) in the wake of the Rio Framework Convention formed an important focus for the debate on Joint Implementation (JI) and other mechanisms for financial and technological transfer. The 1994 Groningen Conference on Joint Implementation is symptomatic for this renewed international interest. In the European context the N e t h e r l a n d s also increased the diplomatic pressure to reach consensus on imposing energy taxes for climate change purposes. The effort was unsuccessful and a unilateral national tax for small-scale energy users may be imposed by 1996. 1.2 N o r t h - S o u t h r e l a t i o n s as r e s e a r c h p r i o r i t y The F i r s t Phase of the Dutch National Research P r o g r a m m e on Global Air Pollution and Climate Change (NRP-I) was initiated in 1991 to stimulate scientific research on climate change. NRP-I covers five different research themes. Out of a total budget of Dfl 63 million Dfl 9 million were earmarked for Theme "Sustainable Solutions". The projects considered in this section were grouped as a subtheme under the heading "international instruments for climate change policy".
From the start of NRP-I there has been increasing awareness of the importance of N o r t h - S o u t h relations in solving climate change problems. The P r o g r a m m i n g Group for Theme Sustainable Solutions commissioned a s e p a r a t e s t u d y on establishing priorities for policy research on climate change and developing countries. This emphasis on North-South relations is also apparent from the list of projects committed in the area of international climate change policies. Of the five projects approved and implemented four concern primarily developing countries. The success of the Framework Convention on Climate Change depends in m a n y ways on the evolution of North-South relations during its i m p l e m e n t a t i o n . Research on international climate change policies should not just focus on the environmental obligations of the North, but also on the development rights of the South and the mechanisms for financial and technological transfers from the North to the South. To establish an effective international climate change policy regime the views and preoccupations of the different national interest groups in developing and industrial countries are of eminent importance in order to link priority issues strategically. This is the purpose of the first project on international negotiation strategies. To prepare possible strategies for solutions three types of questions are of eminent importance for any kind of transfer mechanism between developing and industrial nations: how to tackle deforestation taking into due consideration the interests of the third world actors involved; how to reach an energy-efficient path for optimal industrial growth in the South, and how to limit the potentially disastrous global w a r m i n g effects of a coal-dominated energy development road for the world's most populous countries. These three questions
1297 are addressed in three separate projects. Finally, an international comparison between energy efficiencies seemed useful to estimate the scope for improvements and potential pitfalls in interpreting energy efficiency figures across nations. This is the topic of the fifth project in this subtheme. To provide a sound empirical base for possible solutions, the five NRP-1 projects funded to a n s w e r these questions all depended on a case study approach. Moreover, existing working relations between Dutch researchers and their counterparts in the South were used as far as possible. 1.3 S t r u c t u r e o f t h e a s s e s s m e n t Given the diversity of topics in this subtheme and the lack of common ground between projects, the evaluation concentrates on the results of individual projects. Each project is covered in a separate paragraph. A list of the projects evaluated is presented in table 1.1 below. Only the concluding section of the a s s e s s m e n t concerns general observations on programme effectiveness and scientific quality of the entire sub-theme and does not concern individual projects. Each project paragraph is divided into a number of sections. The first section deals with factual information on the stated goals and methodological approach of the project, while the last section deals with a normative evaluation of the project results in terms of scientific quality, effectiveness in relation to stated goals, relevance to policy m a k e r s and i n t e r n a t i o n a l embedding. The intermediate sections describe the project results. References to project publications are given in a final bibliography. An i m p o r t a n t caveat for this assessment concerns the fact, t h a t at the time of writing this assessment insufficient material was available to evaluate projects thoroughly and definitely. In most cases final reports were not available.
Table 1.1 List of projects in the NRP subtheme "International Instruments for Climate Change Policy" Title
Project leader
Number
International policies to address the greenhouse effect
G.C.A. Junne
853103
Local actors and global tree cover policies
W.T. de Groot
851056
Strategies and instruments to promote energy efficiency developing countries
J.C. J a n s e n
853101
Compatibility of CO2-emission reduction targets with long-term economic development in China
H. Folmer
852064
An international, statistical comparison of industrial energy efficiencies
K. Blok
852084
1298 2.
I N T E R N A T I O N A L P O L I C I E S TO A D D R E S S THE G R E E N H O U S E EFFECT
2.1 An application of the theory of international regime formation This project is a cooperative effort of the Department of International Relations and Public International Law of the University of Amsterdam (project leader: prof. dr. Gerd Junne; main researcher: drs. Richard van der Wurff) and the Institute for Environmental Studies of the Free University (main researcher: Joyeeta Gupta LL.M.). The project aims to provide an in-depth analysis of the political feasibility of different instruments and mechanisms to encourage developing countries to adopt national programmes for the limitation of their greenhouse gas emissions. The research provides background information on the position of important participants in international climate change policy negotiations. The project approach is based on the theory of international regimes. Regime theory analyses conditions for the establishment of international regimes (stable sets of principles, norms, rules and procedures, that guide i n t e r n a t i o n a l cooperation). The approach is characterised by emphasis on the importance of domestic conditions within the countries concerned (perceived interests of different actors) and the catalytic role of issue-linkages. In particular, linking climate change issues with issues of development and industrialization may be important for international regime formation. The empirical material for the study is based on seven national case studies: four of which are developing nations (Indonesia, India, Kenya and Brazil) and three of which are industrial nations (Germany, UK, USA). 2.2 N a t i o n a l case studies on e x i s t i n g climate c h a n g e policies It appears, that developing countries have a dual perception of climate change negotiations. In the light of their historical experience they fear that industrial countries will try to negotiate a reduction in their emissions regardless of the consequences for their economic development. On the other hand they hope, that they will be allowed a fair share of emission allowances in terms of per capita equity. Wavering between scenarios of Angst and Hope, the South views the issue of climate change as a test case for Northern sincerity in global partnership. Several factors are threatening to alienate developing countries in this respect. First, they believe that industrial nations are making an artificial and unnecessary distinction between global e n v i r o n m e n t a l costs and benefits and local environmental costs and benefits. Moreover, the focus on cost-effectiveness in terms of incremental costs may lead to the externalisation of important social and local costs and consequently to non-optimal solutions from their point of view. Secondly, given their vulnerability to climate change the one-sided emphasis on mitigation options with the exclusion of adaptation measures prevalent in negotiation discussions appears unjustified. Thirdly, the extended discussion on Joint Implementation provoked by primarily industrial nations and the poor record of industrial nations so far in making available new and additional funds, make them sceptical about the seriousness with which the North is t a k i n g up responsibilities. Finally, developing countries receive contradictory messages from related international regimes. Such inconsistencies make them worried about the real motives of industrial nations.
1299 From the case studies, it is also clear, t h a t developing countries show little domestic support for or interest in climate change actions per se. The ratification of the FCCG is not viewed as a binding constraint on any kind of domestic policy, but a purely intellectually and morally correct thing to do in international relations. It reflects rhetorical p a t e r n a l i s m in foreign policy r a t h e r t h a n domestic commitment or social consensus. Industrial countries differ in their approach towards international climate change policies, ranging for a legalistic approach in the UK (we do what we are obliged to do, but nothing more) via a commercial approach in the USA (climate change will offer industry business opportunities) to a more development oriented approach in G e r m a n y ( e m p h a s i s i n g the d e v e l o p m e n t cooperation aspects a n d J o i n t Implementation). They share the same liberal economic outlook and a preference for involving m a r k e t forces. This explains their emphasis on cost-effective solutions.
2.3 Strategic choices for international negotiations From a strategic point of view the study concludes, t h a t three elements are necessary for successful negotiations. In the first place, global priorities such as climate change should be matched with local priorities within developing countries. Without such matching international climate change policies will never become a priority for developing countries. Secondly, industrial countries should demonstrate conclusively, t h a t they are making explicit sacrifices to bring emissions down. The evidence for such sacrifices so far is not considered very convincing. Finally, more efforts should be devoted to making existing regimes consistent and to establish issue-linkages. With these strategic elements in mind, a number of recommendations regarding existing i n s t r u m e n t s of international climate change policies follow. The Global E n v i r o n m e n t a l Facility should focus much more on capacity building and promoting appropriate domestic institutions. This avoids the counterproductive focus on global versus local costs and benefits and the sensitive and unpopular debate on incremental costs. Joint Implementation should be pursued under the concept of dual commitments; thus demonstrating the willingness of industrial countries to m a k e domestic sacrifices. Development cooperation policies, t h a t have an indirect effect on climate change policies such as population policies, p o v e r t y a b a t e m e n t , m a s s t r a n s i t s y s t e m s or energy policies should be strengthened, not weakened as part of environmental policy. In terms of strategic alliances agreements on the basis of similar problems and solutions between small groups of nations should receive more attention. Examples are regional groups of big emitters, nations affected by desertification or deforestations or island and coastal nations threatened by sea-level rise. In international negotiations nations can follow three types of progressively more complex and pervasive strategies for consensus building, which are t e r m e d respectively Pragmatic Dialogue, Pragmatic Synergy and Cultural Synergy. The Pragmatic Dialogue strategy is based on an issue-specific approach with emphasis on no regret options. It is a strategy, that does not a t t e m p t to change existing power relations and economic conditions. The underlying reasons why different actors support the same policy are unimportant and may even be very different.
1300 The goal is to adopt simple and easy options. The Pragmatic Synergy strategy follows a least regret approach and requires some changes in existing power relations and economic conditions. Acceleration measures with respect to related policies and technological innovations are actively pursued. The Cultural Synergy strategy requires much more fundamental changes in attitudes and goals and is only feasible through shared ideologies and perceptions. It requires cross-cultural understanding and institutional changes. It uses the very diversity of people to enhance problem solving by combined action and it requires the abolition of parochial, ethnocentric negotiation attitudes, that are presently often apparent. 2.4 E v a l u a t i o n of project results Regime theory has been widely used in analyzing international environmental policies. In addition, climate change policies are obviously linked with development issues in general. This makes regime theory with its emphasis on domestic factors and issue-linkages a convincing point of departure. However, the connection between the theoretical concepts of regime formation and the empirical evidence gathered in the case studies is rather weak. Although the terminology is shared and a definite attempt has been made to follow an outline according to concepts from regime theory, the strength of the case studies lies in providing a descriptive overview of domestic situations rather than in answering specific empirical questions generated by the theory or providing insight in potentially important issue-linkages.
The case studies on developing countries are very instructive for those with an overly optimistic view on continuing progress in the field of international climate change policies. As to positive linkages with development in general the views expressed in the case studies can be characterized as pessimistic, in particularly concerning the clash between the vested interest of the North in any type of aid commitment and the growing needs of the South. These case studies provide important lessons in this respect. Regime theory makes a distinction between the substantive and the procedural elements of regime formation. It also classifies different types of linkages. Since the project intents to come up with practical recommendations regarding climate change negotiations, one would expect more recommendations regarding the importance of substantive versus procedural elements and the relative weight of the different types of linkages. In fact, many concerns of developing nations appear to be addressed to the procedural rather than the substantive elements of a climate change regime; with organizational linkages rather than material linkages as is clear from the intense debates on the Global Environmental Facility and Joint Implementation. One bottleneck for successful negotiations is insufficient synergy between the goals and actions of the Ministries of Environment and International Cooperation in industrial nations, in particular regarding the relative importance of issues and linkages. Domestic actors in industrial countries can have conflicting views on such questions. Moreover, worries of developing countries of having additional environmental strings attached to aid funds and of simple rerouting of aid funds through climate change oriented channels are not without foundation and should be treated in the case studies.
1301 The project provides an interesting theoretical approach to problems of international negotiation. The project reports demonstrate considerable awareness of the issues at stake in actual negotiations, both in preparation of and in following up the FCCC through the INC process and GEF-related dialogues. The case studies on developing countries provide an objective and well-organized assessment of the domestic background for climate change politics in the developing world, although the connection between these case studies and the theoretical framework can be considerably improved. The case studies on industrial countries are lacking in focus with respect to the major goal of the project. Overall, the recommendations on specific instruments make sense. 3.
LOCAL ACTORS AND GLOBAL TREE COVER P O L I C I E S
3.1 I n t e g r a t i n g m i c r o - o r i e n t e d , s i t e - s p e c i f i c s t u d i e s w i t h m a c r o - o r i e n t e d political studies The project is implemented by the Centre of Environmental Science of Leiden University (project leader prof. dr. Wouter T. de Groot; main researcher drs. Evelien M. Kamminga). Country specialists were involved in the case studies. The project objectives are twofold: to assess and integrate existing scattered knowledge concerning protecting and restoring tree cover through understanding local people's motivation and to compile this information in a format, which makes it useful for global climate change policy making.
The project approach is based on the principles of the action-in-context theory. This is a method of social science research of environmental problems developed by the Centre of Environmental Science. The method starts with an analysis of the available options and motivational factors of primary actors, who are directly involved in the problem. In the next step it looks at the secondary actors, which influence the availability of options and the strengths of motivational factors for primary actors; thus establishing a network of power relations between primary and secondary actors. Following this procedure at higher levels of action finally generates a so-called actors field for the problem of deforestation. The action-incontext method tries to integrate the strength of purely micro-oriented local deforestation studies with the insights from macro-oriented political-economic studies. The empirical material for the study is based on three regional case studies" the Cagayan Valley region in the Philippines, the Southern forest region in Cameroon and the North-eastern Amazon region in Ecuador. The case studies are based on literature research and short field missions. 3.2. N a t i o n a l c a s e s t u d i e s on d e f o r e s t a t i o n p r o c e s s e s The three case studies are located on different continents and demonstrate the diversity of deforestation in practice. The deforestation process in the Sierra Madre region of Luzon in the Philippines (Cagayan Valley) has operated in the past through an interactive mechanism of corporate loggers opening up the forest with logging roads and local people moving in as migrant farmers with little interest to improve their land holdings. Provincial politicians played a facilitating role in the sense that the existing patronage system encouraged the buying of votes and the
1302 use of forest resources for immediate cash. The present situation is however much more diffused. Logging permits are cancelled, road construction is under scrutiny, protection funds are available and tenure is obtainable. The future is now dependent upon the evolving practice of small-scale logging for the regional furniture industry and the pace of transition from slash-and-burn agriculture towards more sustainable forms. Although the process of deforestation in the Southern forest region of Cameroon is also started by corporate loggers, no migrant farmers follow in their tracks because of lack of population pressure. Timber operations however tend to be very wasteful and are not kept in check by any countervailing forces, because the local elite are involved through small-scale licenses and are generally inclined to sell their regulatory power. Although NGO's and the World Bank are gradually moving in as an intermediary force which could keep the logging companies better in check, the t h r e a t of timber certification, which would discriminate against timber harvested under unsustainable conditions, is stimulating intensified operations in the short term. In contrast to the Philippines and Cameroon deforestation in the North-Eastern Province of Ecuador has not started because of corporate loggers, but because of corporate oil exploration and exploitation. Farmers stimulated by the government move in along access roads; first starting subsistence farming, but gradually moving towards cash crops on additional land and finally reverting to extensive cattle ranching on large plots. Although the rate of colonization has slowed down recently because of environmental pressures on oil companies and opposition from indigenous people, the future is highly uncertain given the lack of legislative action to protect forest resources. 3.3 E m p h a s i s on t r a n s i t i o n t o w a r d s s u s t a i n a b l e agricultural d e v e l o p m e n t Deforestation studies tend to be of two types: the micro-oriented, site-specific studies, which usually follow an ecological or anthropological approach and the macro-oriented, aggregate studies, based on statistical analysis of correlation between data on forest cover and data on population, GDP, etc. Both types of studies are not particularly helpful in implementing solutions to stop deforestation. The first type of study tends to focus on the descriptive and physical aspects of the problem, thus addressing the proximate causes of deforestation. Such studies do not consider forest protection policies in a consistent national framework. The second type of studies tends to focus on the underlying forces driving deforestation. Statistical studies however leave too many open questions regarding the direction of causality and explanatory factors on the macro level such as population growth or foreign debt can not be viewed as directly instrumental in terms of forestry protection purposes. The present study tries to avoid this dual dilemma by avoiding broad statistical inferences, yet derive recommendations on the macro level from detailed case studies.
The major conclusion is, that forest protection policies should primarily be aimed at influencing the decisions of migrant farmers towards sustainability involving intensification of agricultural practices. Such policies are most important at the national level. International policies should be targeted towards enhancing national capabilities and motivations towards this goal. National policies in the area of land
1303 use and agriculture such as extension services, credit schemes, tenure regulation and product m a r k e t stabilization seem more important in this respect t h a n forest policies per se. From this perspective developments in the international timber t r a d e a p p e a r less crucial t h a n is often suggested in the literature. In t e r m s of popular slogans fair aid, for instance through a Global Forest Fund, is considered better t h a n fair trade. With regard to population policies it is important not only to slow the rural exodus from agricultural areas towards city and forest fringes, but also to shift the "modal split" between these two choices away from forest fringes. Instead of a project by project approach in forestry sector funding as exemplified by World B a n k practices the creation of a Global Forest F u n d is preferred, from which p a y m e n t s are made on the basis of actual national achievements in forest protection. This is a choice for broad programmatic support and output-oriented financing on the national level r a t h e r t h a n narrow input-oriented financing on a b a n k a b l e project base. Such an a p p r o a c h also avoids the s u b s t a n t i a l controversies regarding sovereignty and compliance and simplifies performance measurements, since relatively objective remote sensing data could be used.
3.4 E v a l u a t i o n of project results From an analytical point of view the appeal of the action-in-context perspective as an i n t e r m e d i a t e and integrating approach between the site-specific case study perspective and the aggregate political-economic perspective is evident. Yet the analytical concepts such as available options, motivational factors, action fields etc. are not used in a systematic and interconnected way in the case studies; the theory seems to offer no more than a way of classifying information about different actors. Conclusions from the case studies do not depend in a major way on the actual use of action-in-context theory. The regional case studies provide an excellent overview of how and why local, national and international actors are involved in the process of deforestation in totally different parts of the world. They illustrate well how little generalization is possible on the proximate causes of deforestation and the motivations of local actors. Policy m a k e r s in the field of climate change are already w a r y of the complexities of slowing down deforestation or stimulating reforestation in terms of the national sovereignty issues at stake. This study will tend to reinforce their views with arguments from the local level. The study results point out, t h a t a smooth transition of farming systems to more intensified and p e r m a n e n t land-use is the most important prerequisite for forest conservation on the local level. On the national level, attempts should be made to delink the historical connection between population pressure, economic growth and the process of deforestation. These results are useful from a diagnostic perspective in the sense, t h a t they stress the importance of an integrated approach, where actions on each level of policy making and between different fields of policy making reinforce each other. But the step between analytical diagnosis and practical r e m e d y is m u c h h a r d e r to conceive t h a n suggested. In a way the focus of i m p l e m e n t a t i o n bottlenecks is shifted from forestry and timber issues towards land-use and agricultural issues, which pose their own set of constraints. The report is optimistic about the potential merits of a Global Forest Fund. Although this recommendation appears to be unrelated to the actual case studies
1304 and the action-in-context theory, it contains interesting material for policy makers in the climate change debate, since it addresses the important issues of sovereignty and principles of fund financing and disbursement, that pervade all discussions about financial and technological transfers to the third world. A strong preference for output-oriented national financing through annual transfers is indicated (based on hectares of non-degraded tropical forest). Input-oriented project financing through a long-term capital fund is considered less attractive. There is the implicit assumption, that national governments are somehow better equipped to produce ultimate results than the forestry experts from international development banks with their emphasis on bankable projects. Given the right incentive this might be the case, but it is uncertain if the limited contributions of a Global Forest Fund will be able to change relatively powerless, corrupted and short-term oriented governments as exemplified in the case studies in a fundamental way.
4.
S T R A T E G I E S AND I N S T R U M E N T S TO P R O M O T E EFFICIENCY IN DEVELOPING COUNTRIES
ENERGY
4.1 Survey of experiences and strategies for efficiency i m p r o v e m e n t The project concerns a cooperative effort by ECN-Policy Studies (project leader: drs. J.C. Jansen) and three institutes in developing countries: Tata Energy Research Institute (TERI, New Delhi; main researcher: Sharmila Barathan), E n v i r o n n e m e n t et D~veloppement du Tiers-Monde (ENDA, Dakar; main researcher: dr. Souleymane Diallo) and Instituto de Economia Industrial, Universidade Federal do Rio de Janeiro (IEI/UFRJ, Rio de Janeiro; main researcher: Prof. dr. Joao Lizardo R.H. de Araujo). Regional and sectoral specialists from these four research centres were involved in the case studies. The project objectives are threefold. The first goal is to provide a detailed analysis of the relation between industrialization, energy use and conservation efforts in the three major developing regions of the world. The major regional bottlenecks for energy efficiency improvements in the manufacturing sector are identified. Finally, strategies and actions to be followed for effective transfer of resources and technology from the North to the South with respect to industrial energy saving technology are recommended. The project approach is not based on a specific theoretical perspective from the field of economic or political science. It follows a pragmatic approach, in which the participating regional institutes collect specified data on industrial development and energy consumption for a limited number of nations and sectors, comment on the evolution of energy policies and recommend strategies and actions based on their regional observations. The leading institute provides background materials and a synthesis report. The case studies focus on India, Bangladesh, Thailand and South-Korea in Asia, on Tunisia, Senegal, Cameroon and Zimbabwe in Africa and on Brazil, Bolivia, Costa Rica and Mexico in Latin America. In addition, the studies concentrate on the energy-intensive subsectors of iron and steel, aluminum, chemicals, cement, paper and board.
1305
4.2 R e g i o n a l c a s e studies on industrial e n e r g y c o n s e r v a t i o n policies As could be expected the regional case studies show, t h a t industrial energy conservation experiences in different countries and in different subsectors have been quite divergent and to a great extent related to the general economic and technological conditions prevalent. The case studies contain numerous countryand sector-specific details, that cannot be summarised easily. The African countries surveyed show few success stories except perhaps Tunisia. The Latin American experience is more diverse. Several countries such as Brazil and Mexico show substantial improvement in terms of specific energy use per physical unit, but the impacts on energy demand are cancelled by a pronounced s t r u c t u r a l shift towards energy-intensive industries. In Asia the outlook for continued i m p r o v e m e n t in industrial energy conservation is most promising. Countries like Korea and Thailand have large industrial energy conservation programmes operating with related legislative and institutional arrangements. The two industrial giants China and India are also making progress. In particular, the meteoric industrialization of China has fortunately not been m a t c h e d by an equally strong growth of industrial energy demand. Looking at specific energy intensity differences by sector, it is generally true, that developing countries are not performing as well as developed countries. The differences are not the same across subsectors: they tend to be relatively small (<10%) for industries such as aluminum, but they tend to be relatively large for industries such as pulp and paper (>30%). These differences are a function of the scale of operations, the vintage of capital stock and the amount and quality of different feedstocks and products. Apart from such physical factors economic factors such as distorted energy prices and protected product markets account for lower energy efficiency levels.
4.3 P o l i c y priorities for d e v e l o p i n g c o u n t r i e s Three types of policy action in the area of strengthening industrial conservation efforts can be distinguished: capacity building, financial incentives, and regulatory m e a s u r e s . Capacity building concerns the setting-up of a mission-oriented, coordinating agency which acts as a intermediate between industry and other parties and which can effectively administer information campaigns and training programmes. Such an agency should also facilitate the delivery of audit services through their own channels or through involvement of specialised engineering firms. Financial incentives should foremost address problems in reaching full-cost, market-oriented energy pricing. Credit programmes have proven less successful in the past, but their performance could possible ameliorate with proper design. R e g u l a t o r y m e a s u r e s should concentrate on establishing a comprehensive legislative structure for an energy sector, that can operate relatively independent of pressing social and financial concerns of the government and in conformity with l o n g - r u n m a r g i n a l cost pricing and acceptable e n v i r o n m e n t a l s t a n d a r d s . Regulatory action in the form of product labelling requirements and equipment efficiency standards are useful only, if compliance can be effectively enforced. 4.4 E v a l u a t i o n of project results The development of industrial energy demand in a global perspective is based on statistical analysis of structural shifts in energy demand over extended periods of time. This structural analysis is applied to projected aggregate energy demand in
1306 order to derive the global share of industrial energy demand in the future. In principle, this is an interesting approach with solid roots in traditional development theory. The analysis however treats energy efficiency improvement on the aggregate level, not on the sector-specific level. This makes the separation of a structural component (due to changes in activity levels between sectors) and an efficiency component (due to changes in energy efficiency levels between sectors) impossible. Although this would create substantial problems of data availability and collection, one would like to see the top-down approach of the study complemented by a bottom-up approach in this respect. The regional reports follow a uniform set-up and are well-conceived in terms of making the energy role and position of the diverse and fragmented industrial sector in widely different parts of the world accessible. As can be expected, industrial energy conservation opportunities and bottlenecks are very different not only for the three major global regions, but also for the different countries, that have been analyzed in detail within those regions. However, the success of the project in terms of describing this diversity clearly in the regional reports at the same time makes the overall project objective of coming up with general conclusions regarding industrial energy conservation policies in the third world more debatable. The project has produced a wealth of empirical material on industrial energy use in developing countries. Unfortunately, it appears that the conclusions regarding the difficulties of energy conservation policy in developing countries are not specific to energy as a major factor of production. Developing countries are often inefficient in the use of all types of resources, not just energy, and many studies have been devoted to the generic causes of this phenomenon. One would expect more awareness of those general problems of efficiency and the policies usually recommended to alleviate them demonstrated in the study. The special place of energy as a factor of production within this general framework could than be made more clear. Although the instrument of joint implementation became widely discussed only after the start of the project, the issue was considered of sufficient importance to merit separate attention. A case study for the cement sector, which is of central importance in many developing countries, was used as basic material for a paper on joint implementation in this sector. Such an approach appears very relevant for actual policy making, since it makes the opportunities and limitations of joint implementation much more transparent on the level, where decisions about pilot projects must be made. Negotiation discussions on joint implementation usually take place on a highly abstract level, while the real implementation problems can only be understood on a much more practical level, where the interests and motivations of the private sector agents ultimately involved in financial and technological transfers are clear. The project has succeeded in providing a thorough analysis of industrialization and energy use in the third world including the potential regional bottlenecks. There is however at this time a lack of conclusions regarding the consequences of this analysis for financial and technological transfers from the industrialized North to the developing South. The policy relevance of the study for those involved in climate change related decision making is therefor not completely clear. An
1307 i m p o r t a n t exception concerns the case study on the cement industry, which affords an actor-oriented view of the potentials and bottlenecks for Joint Implementation in the industrial sector. The international embedding of the work is well taken care off, because of the direct involvement of three leading research institutes from developing countries.
5.
COMPATIBILITY OF CO2-EMISSION REDUCTION TARGETS WITH LONG-TERM ECONOMIC DEVELOPMENT IN CHINA
5.1 C o m p u t a b l e g e n e r a l e q u i l i b r i u m m o d e l l i n g a n d p o w e r c a p a c i t y p l a n n i n g combined This project concerns Ph.D. research at Wageningen Agricultural University (Promoters prof.dr. Henk Folmer and prof.dr. Paul van Beek; promovendus ZhongXiang Zhang). The project objectives are to develop a CO2-emission trend analysis for China and to make a cost-effectiveness analysis of reduction options in the Chinese energy system. The project approach is based on developing two new national models for China: a Computable General Equilibrium (CGE)-Model and a power sector model based on linear programming. These models are to be used for analyzing the stated research questions. It should be noted, that at the time of this assessment insufficient materials were available to evaluate the project properly. Submitted papers are of an introductory nature, describing in a general way the state-of-the-art in economic modelling and instruments for climate change policy and the structure of the models to be applied. 5.2 Analysis of the Chinese energy system In the debate about limiting global CO2-emissions from energy sources, the future contribution of China occupies a central role. As the world's most populous and coal dependent region it is evidently of pivotal importance in any international greenhouse abatement strategy. The general description of the present situation and future prospects provided in this project underlines the daunting nature of the challenge; particularly since the alternatives to coal power in the field of hydro and nuclear create their own share of environmental problems. Although energy efficiency improvement in China has been more impressive t h a n in other developing nations in the past decade, the energy intensity remains high both in economic and in physical terms. This is partially accounted for by the high share of industry in the national product and the high share of energy-intensive heavy subsectors within industry. Moreover, there are considerable m e a s u r e m e n t problems when it comes to economic comparisons because of difficulties of establishing relevant purchasing power parity conversion rates. Nevertheless very substantial improvements are likely, particularly when high industrialisation rates lead to fast replacement of old and small-scale processing equipment by modern, large scale, integrated plants. 5.3 M e r i t s of d i f f e r e n t a p p r o a c h e s to C O 2 - e m i s s i o n r e d u c t i o n c o s t estimates The project results available so far are limited to the methodological choices made in the project. Empirical results are not yet available. Available cost estimates in the literature are based on ad-hoc, technology-specific comparisons, dynamic optimization models of a sectoral nature, input-output or macroeconomic models
1308 or general equilibrium models. It is concluded, that general equilibrium modelling is to be preferred, because effective CO2-emission reduction policies require nonmarginal shifts in the prices of production factors and consumer goods and services. The ultimate effects of the resulting structural changes cannot be captured adequately in an ad-hoc, sectoral or partial equilibrium framework. However, since such models cannot capture the technical details required for specific energy policy choices, a hybrid approach linking a general equilibrium model with a dynamic sectoral optimization model would be even better. The essential features of such a hybrid model construction for China are described.
5.4 E v a l u a t i o n of project results To develop two new advanced models inclusive of estimating empirical parameters and actual application for a huge and diverse economy such as China in the course of one Ph.D. thesis is indeed a very substantial challenge. The development and empirical use of just the CGE-model would really require a substantial team effort as would the development of a detailed power sector model. Therefore, the results of the project must be viewed primarily as a first and innovative effort to explore the potential of a hybrid model for climate change policy purposes. Although CGE-models are in principle attractive to track the ultimate effects of climate change policy, the empirical requirements appear to be considerable in practice. Results tend to be very sensitive for chosen parameter values. So far, the major lessons to be learned from CGE-models are of a general nature: indicating the potential complexities of predicting the effects of applying specific economic instruments in the long run, but of limited relevance to actual policy making. Although the required investments in power plants to sustain present economic growth rates in China are very high, it is unlikely that the capital requirement and electricity price differentials related to different power scenario's will be large enough to cause economy-wide effects, that are traceable within the already large margins of uncertainty attached to CGE-runs. The danger of spurious results is large. Moreover, to reach the original objective of estimating the cost-effectiveness of power plant options, a first order estimate based on a stand-alone power supply model would already be sufficient and interesting and would be helpful anyway in testing the performance of such a model.
6.
EVALUATION OF GUIDELINES FOR INTERNATIONAL CO2-EMISSION BUDGETS
SHARING
OF
6.1 A n i n t e r n a t i o n a l , s t a t i s t i c a l c o m p a r i s o n of i n d u s t r i a l e n e r g y efficiencies The project is implemented by the Department of Science, Technology and Society of Utrecht University (Project leader: dr. Kornelis Blok; main researchers" dr. Ernst Worrell and drs. Dian Phylipsen). The project objective is to evaluate potential guidelines for the determination of emission limits for CO2 per country taking into account the present economic structure and the per sector level of energy efficiency. The project is limited to the determination of the comparative energy efficiency in the power and industrial sectors of primarily developed countries. The methodological approach is based on the collection and analysis of
1309 s t a n d a r d statistical data on national energy consumption per sector. Actual activities are not aimed at the evaluation of guidelines for target setting in i n t e r n a t i o n a l negotiations, but at providing relevant data concerning energy conservation potentials.
6.2 Establishing a basis for emission reduction a g r e e m e n t s The discussion on an acceptable basis for emission reduction agreements has so far concentrated on an equitable formula for setting targets. Such a target-based discussion focuses on the ultimate goals of an agreement r a t h e r t h a n the means by which such goals could be reached. Discussions along these lines are very dependent on basic assumptions on acceptable criteria for a j u s t s h a r i n g of obligations. An alternative to this approach is to try to reach agreements on international efficiency standards, which when accepted would automatically lead to emission reductions. In a way such an approach emphasizes the common goal of efficiency, which perhaps would cause less controversy in i n t e r n a t i o n a l negotiations. Such efficiency standards would require more and more detailed i n f o r m a t i o n about the economic s t r u c t u r e and sectoral energy efficiency performance. The latter approach towards emission reduction agreements forms the rationale behind the results of this project. 6.3 C o m p a r a t i v e a n a l y s i s of e f f i c i e n c i e s in e l e c t r i c i t y p r o d u c t i o n and industrial sectors A choice was made to analyze energy-intensive sectors, that are characterised by a relatively homogenous product and can be adequately described in terms of specific energy consumption (average energy use per physical unit of production). These sectors are electricity production, refineries, iron & steel, ammonia, pulp & paper, cement and petrochemicals. The electricity study indicates, t h a t the potential for raising energy efficiency with commercially available technology is very high; from a purely technical perspective savings of 30-35% on a worldwide scale are feasible. This figure refers to efficiency improvements only; when fuel mix changes are included, the effect on CO2-emission reduction would even be larger. The studies on the manufacturing sector concentrate on Europe. Taking the best performer among countries in specific industrial subsectors as indicative of best available performance, it appears that substantial conservation potential exists and t h a t the differences between countries are often large. For instance in Spain improvement potentials for subsectors range from 1% in the case of ammonia to 42% in the case of iron&steel. In the pulp & paper industry potentials range from 13% in the Netherlands to 55% in Denmark. Average conservation potentials per subsector in the E u r o p e a n Union range from 13%-27%. Countries in E a s t e r n Europe are often, but not always less efficient. 6.4 Evaluation of project results The available reports provide excellent material on the n a t u r e and level of efficiency differences between countries and the problems of generating efficiency figures that are sufficiently comparable for policy purposes. From a methodological perspective these are important contributions. There are however several i m p o r t a n t assumptions hidden in the idea, t h a t comparative efficiency figures based on statistical evidence are useful for climate change policy. Most important is the assumption, that lower specific efficiencies
1310 represent opportunities for cost-effective measures, that countries are keen to take up. The available statistical figures however do not account for generic causes for efficiency differentials such as the scale of operation, capacity utilization, plant vintages in manufacturing or the structure and level of energy pricing. The influence of such factors may be much more important than is often clear from statistical figures. They may therefor be a misleading guide for costeffectiveness. Moreover, when analyzing industries in detail it is often very difficult to separate the structural effects of differences in feedstocks and product qualities from efficiency effects related to process type. Although avoiding controversial issues of equity evaluation in climate change negotiations, such an approach is bound to raise equally complicated questions of efficiency measurement. In fact, the conclusions of an approach based on efficiency standards instead of targets would lead to heavier burdens for poor and inefficient nations. It would tend to put additional pressure on Spain instead of Germany and thus put the equity issue in the limelight rather than avoiding it. Nevertheless, from a methodological point of view, the results provide important background information for a dialogue on guidelines. 7.
G E N E R A L CONCLUSIONS
7.1 Programme effectiveness If anything, the studies showed decisively, that the questions asked were the right ones. The future of international climate change policies will depend in large measure on the position and actions taken by developing nations as far as deforestation and industrialization is concerned. At the same time, it is clear that climate change is not at all a priority issue in developing countries. The studies underline the impressive problems facing the reconciliation of development and climate change goals at the domestic level, while at the same time indicating, that the scope for improvement is large and dependent upon international action. In terms of problem formulation and orientation this subtheme must be evaluated very positively. The orientation of the work has been primarily diagnostic, describing the causes and consequences of anthropogenic CO2-emissions in analytical terms. Although the studies also include some policy-oriented work (global forest fund, joint implementation in industry), this aspect is relatively weakly represented. In this sense, the direct policy relevance of the work is less than originally intended. Although this is partially a result of the rapidly changing international policy environment (the NRP-I programme dates from before the UNCED Rio Conference and the establishment of the FCCC), this can not be viewed as the only reason. Two other reasons must be mentioned in addition: the gap between scientific approaches and policy relevance remains wide and relatively unexplored and the lack of tools, that incorporate policy instruments as an integral part of the methodological approach.
7.2 Programme quality It appears difficult to find the right kind of balance between theoretical rigour and empirical quality. Where a specific theoretical approach is proposed, there is often a limited influence on the empirical data gathering and analysis beyond a purely
1311 organizational impact (action-in-context theory and regime theory). Sometimes there is a lack of empirical material (China) or theoretical backbone (industrial energy conservation). The work on guidelines has perhaps found the best balance between attention for methodology and emphasis on empirical data. The strong dependence on a case study approach in most projects is very fortunate. The studies are excellent in drawing attention to the great diversity between developing countries and the dangers of coming up with universal solutions at the global level. From a scientific point of view this leaves little scope for drawing general conclusions. From a policy point of view however the studies may be viewed as supportive of the tendency to move away from grand solutions towards a phased approach in which pilot projects play a major initiating role for exploring the potential of instruments such as joint implementation. A case study approach clearly helps in bridging the widening gap between analytical problem descriptions of a general nature and the ad-hoc project-by-project approach emerging in today's international climate policy practice. Particularly in the case of international climate change policies it appears imperative, t h a t serious attempts are made to embed policy research internationally; both in terms of inputs and in terms of outputs. In this respect, this subtheme generally has succeeded well in diverse ways. Sometimes by linking up with other on-going international projects; sometimes by actually involving third-world scientists; sometimes by organizing seminal workshops. In addition, several projects have already published part of the results in the international literature. 8.
REFERENCES
International policies to address the greenhouse effect; G.C.A. Junne / UVA (Project 853103) Gupta, J., 1993. Interviews with climate change negotiators. IPAGE-project Background Report I. Gupta, J., 1994. Case study on Kenya" climate change politics. IPAGE-project Background Report III. Gupta, J., 1994. Case study on India: climate change politics. IPAGE-project Background Report IV. Gupta, J. 1994. Case study on Indonesia: climate change politics. IPAGE-project Background Report V. Gupta, J., 1994. Case study on Brazil: climate change politics. IPAGE-project Background Report VI. Gupta, J., 1994. Chapter 6: domestic background of positions of major developing countries. IPAGE-project Working Paper 2. Gupta, J. and R. van der Wurff, 1993. Seminar 1: Political aspects: conditions of successful regime formation. IPAGE-project Background Report II. Gupta, J., G. Junne and R. van der Wurff, 1993. Determinants of regime formation. IPAGE-project Working Paper 1.
1312 Gupta, J., R. van der Wurff, G. Junne, M. HisschemSller and P. Vellinga, 1994. International policies to address the greenhouse effect: an evaluation of international mechanisms to encourage developing country participation in global greenhouse gas strategies. IPAGE-project Working Paper 3 (draft final report). Junne, G., 1994. Climate negotiations: links to other issues. Change 20: 1-2. Local actors and global tree cover policies; W.T. de Groot/RULE and E.M. Kamminga / R U L E (Project 851056)
Kamminga, E.M., 1993. The social context of rain-forest destruction. Change 1 4 : 8-9. Kamminga, E.M., 1994. Deforestation in context: the North-eastern Amazon region in Ecuador. Case Study Report (draft). Kamminga, E.M. and W.T. de Groot, 1994. Forest, people, government: a policyoriented analysis of the social dynamics of tropical deforestation. Main Report (in prep.). Kamminga, E.M. and G. M. van den Top, 1994. Deforestation in context: the Cagayan Valley region in the Philippines. Case Study Report (draft;). Toornstra, F.H., G.A. Persoon and A. Youmbi, 1994. Deforestation in context: the Southern Forest region in Cameroon. Case Study Report (draft). Strategies and instruments to promote energy efficiency in developing countries; J.C. J a n s e n / E C N (Project 853101)
Barathan, S., P. Bhandari and P. Dadhich, 1994. Strategies and instruments to promote energy efficiency in Asia. Regional Project Study on Asia. Buskens, V. and J. Jansen, 1994. Industrial energy demand and CO2-emissions in developing countries in global perspective. Project Working Paper 1, Report no. ECN-C-94-039, Petten. Buskens, V. and F. Diepstraten, 1994. Energy efficiency in selected industries of the manufacturing sector. Project Working Paper 3, Report no. ECN-94-040. Diallo, S., 1994. Efficacit~ ~nerg~tique des industries manufacturi~res en Afrique, Regional Project Study, Dakar. Diepstraten, F., 1994. Industrial energy efficiency in developing countries: sector studies on iron and steel, aluminium and pulp and paper. Project Working Paper 5, Report no. ECN-94-042. Jansen, J., 1994. Industrial energy efficiency in developing countries: how can it be enhanced? Change 21: 18-19. Jansen, J. and V. Buskens, 1993. Energy outlook for China and India in global perspective. Paper presented at the Energy and Environment in India workshop, oktober 1993 Enschede. Jansen, J. and V. Buskens, 1994. The role of sustainable energy issues in development cooperation. Resources, Conservation and Recycling vol. 12. Jansen, J. and F. van der Vleuten, 1994. Joint implementation in the cement industry. Paper presented at the International Conference on Joint Implementation.
1313 Kant, A. 1995. The effectiveness of industrial energy conservation programmes in IEA countries. Project Working Paper 2, Report no. ECN-C-95-32. Lizardo, J., R.H. de Araujo, A, de Oliveira, E.A. Guimaes, R. Tolpan and W. Cateb, 1994. Industrial energy efficiency in Latin America. Regional Project Study on Latin America. Van der Vleuten, F. 1995. Energy and environment in the global cement industry. Project Working Paper 4, Report no. ECN-C-94-035.
Compatibility of C02 emission reduction targets with long term economic development in China; H. Folmer/LUW (Project 852064) Zhang, Z.X., H. Folmer and P. van Beek, 1994. Compatibility of CO2-emission targets with long-term economic development in China. Change 21: 15-18. Zhang, Z.X., 1994. Setting targets and the choice of policy instruments for limiting CO2-emissions. Energy and Environment 5: no.4. Zhang, Z.X., 1994. Analysis of the Chinese energy system: implications for future CO2-emissions. Journal of Environment and Pollution, vol.4: nos 3/4. Zhang, Z.X., 1994. Economic approaches to cost estimates for limiting CO2emissions. Journal of Environment and Pollution 5: no.1.
Evaluation of guidelines for sharing of international C02-emission budgets K. Blok / R UU(Project 852084) Blok, K. G.J.M. Phylipsen, A.P.C. Faaij and E. Worrell, 1994. Energy efficiencies of industrial processes and electricity production in European and Non-European countries. Proceedings of the 1st International Conference on Joint Implementation. Faaij, A., K. Blok, E. Worrell, 1994. Worldwide comparison of efficiency and carbon dioxide emissions of public electricity generation. Martin, N., E. Worrell, L. Schipper and K. Blok, 1994. International comparisons of energy efficiency. Workshop Proceedings, March 1994, Utrecht. Worrell, E., R. Culenaere, K. Blok, and W. Turkenburg,1994. Energy consumption by industrial processes in the European Union. Energy 19: (11).
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1317
Tropical Forest Policies for the Global Climate Wouter T. de Groot and Evelien M. Kamminga Programme Environment and Development, Centre of Environmental Science, Leiden University, P.O. Box 9518, 2300 RA Leiden, The Netherlands
Abstract This paper summarizes the approach and findings of the NRP project 'Local Actors and Global Tree Cover Policies' (Kamminga and De Groot, 1995; Toornstra, Persoon and Youmbi, 1995; Kamminga and Van den Top, 1995; Kamminga, 1995). The aim of this project has been to identify the most effective and efficient options for global climate policies focusing on the tropical forest. Tropical deforestation is a process with very complex and variable causes. In the project's conclusions, therefore, much care has been given to arrive at a coherent image of what really counts most in the myriad of factors, actors, policy levels and policy options.
1. TROPICAL FOREST PROTECTION: THE COST-EFFECTIVE NO-REGRET OPTION FOR THE GLOBAL CLIMATE Rainforest is the climax vegetation of the humid tropics, representing a substantial part of the world's biomass, - in global climate terms, locked-up carbon dioxide that may contribute to either climate stabilization or deterioration. Due to its three-dimensional structure and stability, the rainforest is the global 'h0tspot' of biodiversity and evolutionary processes. Below this global level, the tropical forest provides livelihoods for millions of local people, added to which are the regulation of soil and water balances on a wider regional scale and contributions to national exports. In somewhat more detail, the global climate function of the tropical forest is that: [] forest conversion (e.g., burning) adds to carbon dioxide emissions [] forest regeneration and planting subtract from carbon dioxide emissions [] forest use for firewood, having a net emission of zero, substitutes for fossil fuels. Roughly, the prevention of forest conversion decreases current carbon dioxide emissions by 1 to 2 Gt carbon per year for several decades, which is a fair share of the current emission excess of 3 Gt/year. Moreover, the upkeep of the forest as a firewood resource permanently prevents the emission of approximately 1 Gt/year from fossil fuels. The prevention of carbon dioxide emissions through the forest route tends to be ten times less expensive (per tonne of carbon) than other, more technological means, and the protection of existing forests tends to much more efficient than planting new forest.
1318 On top of that, tropical forest protection contributes to the support of the other (biodiversity, local, regional and national) forest functions. In short, tropical forest protection is one of the most important options for climate policies, and a no-regret option as well.
2. THE STUDY'S APPROACH Many types of studies are performed with respect to the tropical forest. Ecological and anthropological research focus on the forest itself and on its peoples. Forest management and land evaluation studies yield physical prescriptions for sustainable use. These studies do not provide the key insight necessary for forest protection policy design, however, which is why deforestation actually takes place. Of the studies aiming to elucidate the social causality of deforestation, the statistical approaches are the predominant type; Brown and Pearce (1994) are a characteristic collection. These studies correlate data on forest cover and deforestation rates with data on population, GNP, national debt, roads density and so on. Overviewing these studies, they appear to have run into a dead-end street, due to both (internal) problems of inference and (external) problems of relevance. 9 Internally, there are problems of data reliability and statistical method, such as the purely inductive data manipulation. Moreover, the structure of the multiple regression formulas does not reflect causal relationships. This aggravates the well-known causality problem of statistical findings; if, for instance, poverty appears to correlate with deforestation, is poverty then a cause or an effect of forest loss? Many authors of this type of studies recommend to shift towards more 'micro-economic' approaches, focusing on the choices of the relevant actors. 9 Externally, the statistical studies suffer from a problem of relevance. If, for instance, a general correlation is established between forest cover and population density, that is, if it appears that it is difficult to have dense forests and dense population at the same time, what does that then mean for policy making? The crux of forest protection policies is to have people, economic growth and forest, or (borrowing from the energy field) to de-link population and GNP from the forest's fate. In other words, relevance lies not in the rather trivial general 'law', but in the reasons why countries deviate from it. The solution here is the same as for the internal problems: a research focus seeking for actual causal mechanisms rather than statistical correlations. The approach adopted for the NRP project is of the actor-oriented, causal, 'microeconomic' type. More specifically, it has applied the 'Action-in-Context' framework, a methodology designed especially for the causal explanation of the social actors and factors driving environmental problems. A characteristic element is the so-called 'actors field', connecting actors to actors by way of the influences that one actor has on the options and motivations of an other actor. For instance, migrant farmers may be the 'primary actors' that directly influence the forest by their slash-and-burn agriculture. Large land-owners somewhere else may be important 'secondary actors', however, because they take away one option of the farmers, namely, to settle on these lands that they might well prefer over the forest. In such actors fields, farmers, loggers, regional politicians, government agencies and global actors may be discovered interplaying, willingly or unintended, towards forest destruction. As usual in actor-oriented approaches, the research has used case studies as the
1319 primary data source. Care has been taken, however to interconnect the case studies thoroughly with the international literature. The case study areas, in the Philippines, Ecuador and Cameroon, cover significantly different situations in the three tropical continents.
3. THE CASE STUDIES
In the preceding decades, the Sierra Madre region in North Luzon, the Philippines, has been deforested through an interactive mechanism of corporate loggers and local people. The loggers opened up the forest and migrant farmers settled along the logging roads, enhancing in their turn the cut-and-run behaviour of the loggers, who could not envisage their concessions to ever survive the people's intrusion. At the same time, the farmers were not inclined towards investing in the sustainablity of their land use, because they were settled illegaly and hence were uncertain of the long-term benefits to ever materialize. In the background, provincial politicians played a key role in this process. Caught as they were in the patronage system of politics, they were obliged to buy votes rather than win them through a political programme, and becoming involved in the forest plunder was the only ready source of cash. Under international and national pressure, this uni-directional deforestation machine has dissolved into a much more diffuse situation. Logging permits have been cancelled. New roads are put under EIA scrutiny. Nature protection funds have been released to save the last remaining primary forest. Migrant farmers can receive forms of tenure. Local communities may enter into licenced sustainable forestry of secondary forest stands. The future will depend on two major factors: the regulation of small-scale logging for the regional furniture industry and, even more pivotal, the transition of farming from space-consuming and unsustainable slash-and-burn to permanent agriculture. This transition, in turn, depends on subtle cost-benefit balances in which factors such as markets, tenure security, agricultural extension and forest protection play a role and through these, many secondary actors and factors up to the national and higher levels. Roughly, agricultural policies will be more decisive than forestry policies. The Southern forest region of Cameroon is characterized by large-scale logging operations of European firms. The population pressure areas of Cameroon lying far away, the intrusion of migrant farmers is not (yet) prevalent, and the national meat demand being met by wildlife and cattle from elsewhere, no ranching pressure does (as yet) exist. The timber operations are generally wasteful, e.g., skimming the forest for the few most valuable species but leaving much destruction in their wake. This practice is enhanced by small-scale licences issued to local elite members without sustainability restrictions. Moreover, reflecting the general moral crisis of Cameroonian society, many public officials are quite willing to sell the regulatory power of their public office for private benefit, thus depriving the logging companies of any pressure that could balance their natural inclination toward short-term profit maximisation. Yet, in spite of the forest destruction at the small scale and in spite of the undenyable decline in forest quality (including a dramatic poaching of wildlife along the logging roads), the logging has had little effect on the general tree cover. If left to themselves, the logging companies tend to become locked in conflicts
1320 with the local forest dwellers, over issues such as valuable tree species, wildlife poaching and employment opportunities. Practice shows, however, that both parties are responsive to mediating action of NGOs that try to work out new ways of partnership between forest, forest cultures and sustainable logging. The World Bank and other global actors have had a hand in the adoption of a new and better national forestry law. The World Bank's style of operation has contributed, however, to the existing association of global forest policies with disrepect for national sovereignty. The current discussions about a possible certification of'sustainable timber' have caused an expectation of declining prices for timber logged in the way prevalent in Cameroon. This has triggered logging companies to intensify their operations. The North-Eastern Province in Ecuador, part of the Amazone basin, offers a picture of deforestation in full action. The physical backbone of the process is formed by roads constructed by oil companies operating deep inside the forest. Settlement of migrant farmers along these axes is sponsored by the government, that perceives colonization as a low-cost solution to evade structural reform elsewhere. Settlers receive large plots (up to 50 ha), with tenure offered if it is deforested in sufficient time and degree (an example of a 'perverse incentive'). No attention has been paid until recently to the rights of indigenous forest peoples. The leading idea of the farmers and their supporting agencies is that the farmer starts out with subsistence crops on the new clearing, then expands the clearing to plant cash crops, and then invests the profits in cattle, finally to graze on the fully deforested 50 hectares. Cattle keeping is instrumental to much forest loss in Latin America, because it is a very extensive type of land use and also because cattle can walk to its own market, thus enabling farmers to venture far from the roads. Yet, cattle is favoured for its profitability (often subsidized) and cultural reasons. Recently, the rate of colonization has slowed down somewhat, because of environmental pressures on the oil companies, the struggle for the rights of indigenous peoples and other external reasons. No real political will to protect and manage the forest has been developed yet, however, and the future of the forest thus continues to depend on incidental events, external pressures and market shifts. The case studies have been analyzed looking for similarities and focusing on general topics in the international literature, such as the roles of roads and tenure. Comparisons have also been made with other countries; one difference between Ecuador and Brazil, for instance is that in Brazil, the national elite is actively involved in the forest, thus creating a much larger pressure on farmers to move on, making way for large-scale ranchers.
4. SPECIAL TOPIC: A GLOBAL FOREST FUND The Action-in-Context methodology enables causal insights, qualitative or quantitative, in the 'vertical' linkages between the local and global levels. Besides these linkages, the global level has its own, 'horizontal' system-level characteristics (e.g., global equity) that require independent attention. This section focuses on these, especially on the modalities of a possible Global Forest Fund. A Global Forest Fund is a mechanism of international transfer of funds from all
1321 nations to the nations where the forest can be protected most efficiently, which is the nations that still have it. Agencies such as the World Bank usually envisage this transfer to fund forest projects, i.e., forest management promises and the degree to which countries comply to these. This idea lacks a clear-cut economic rationality and arouses reluctancies on the part of the receiving nations, who see it as impinging on their sovereignty. These dilemmas are overcome by financing what really counts, that is, not the input but the output, in other words, not the projects but the forest itself. Thus, the Global Fund pays out yearly disbursements for hectares of forest, easily monitorable by means of remote sensing, without there being anything to comply with. Effective and efficient disbursements lie in the order of magnitude of $ 1 0 per hectare of forest per year on the average, variable for different forest regions. This implies a yearly throughput of the Fund of approximately $ 1 5 billion per year to cover the whole of the tropical forest. This is a full order of magnitude less than the transfers expected to be necessary for carbon dioxide abatement by non-forest means. On the financing side of the Fund, several ethical/political logics apply to distribute the financing over the nations. An example mixture of these rules results in a financing obligation of the Netherlands of 0.1% of the country's GNP; it will be lower for most other nations.
5. CONCLUSIONS The conclusions have yet to be refined in the final report of the project. The overall image of the main conclusions is as follows. 9 The heart of the tropical forest matter is to influence the choice of migrant farmers at the forest fringe to either intensify on the land they have, or to exhaust the land and then move on again. There is a good scope for policies influencing both sides of the farmers' decision balance. Policy options for discouraging forest intrusion consist of, for instance, the hand-over of tenure to tribal people, the cancellation of state-induced transmigration and roads, physical protection of forest reserves, the protection of sustainable logging concessions and the creation of forest product markets. Policy options encouraging agricultural intensification are agro-ecological research and extension, input subsidies, credit schemes, tenure regulation, market creation and stabilization, and so on. 9 The hot policy issue of roads construction can be seen in the same light. Through-forest roads encourage forest intrusion, but roads that connect farmers to markets can encourage agricultural intensification, hence work to protect the forest. 9 National policies and markets are much more influential than the global system level (e.g., the international timber trade). Consequently, global policies should primarily aim to strengthen national capabilities and motivations to move to the types of policies mentioned above. 9 One option for the strengthening of national motivations is a Global Forest Fund, supporting the tropical forest through global financial transfer. If designed properly, such a fund can be economically consistent, politically acceptable, and effective.
1322 Some more detailed conclusions, partly following from the above and partly of a more additional nature, are as follows. [] On timber certification: Since only a small percentage of the trees cut in the tropical forest are traded internationally, the international timber trade is relatively un-important for the future of the forest. Hence, a 'green label' for sustainable timber will be relatively un-important. Its initial effect up till now has been negative. [] On commercial logging: A better approach of the logging companies is to help them to find new ways of partnership with the forest and the forest peoples, combined with improved on-the-ground surveillance. [] On forestry-sector policies: Sectoral forestry policies tend to be less influential on the forest than policies emanating from other sectors. [] On population policies: Working on a very long time scale, population control policies are unrelated to saving the tropical forest. [] On general land and agricultural policies: National policies for land reform and general agricultural intensification will decrease the number of people squeezed out of rural areas. These policies therefore have a general effect of alleviating the pressure on the forest. The far majority of the rural migrants moves to the cities, however, not to the forest. This 'modal split' between cities and forest is more directly decisive over the forest's fate. [] On global forest funding: A proper design of a Global Forest Fund moves away from the current 'banking paradigm' that rewards compliance with forest protection promises. Instead, the financing should be related to. actually existing forest. This avoids sovereignty problems, compliance problems and measurement problems; disbursements simply follow remote-sensing data. Some of these conclusions contradict a number of well-kown environmental slogans. The relative un-importance of the timber trade compared to the Forest Fund implies that for the tropical forest, fair aid (the Fund) is better than fair trade. For the same reason, the slogan of "A better environment begins with yourself" does not hold true for the forest (as is does in most other cases). Green labels and consumer guilt feelings cut little ice. The funds and energies of consumers are better spent on NGOs involved in on-the-spot forest protection and on governments to take their global responsibilities.
6. REFERENCES - Brown, K. and D.W. Pearce, The Causes of Tropical Deforestation, UCL Press, London, 1994 - De Groot, W.T. and E.M. Kamminga, Forest, People, Government; A policy-oriented analysis of the social dynamics of tropical deforestation, Main report of the project 'Local Actors and Global Treecover Policies', Centre of Environmental Science, Leiden University, 1995 -Kamminga, E.M. and G.M. Van den Top, Deforestation in Context: The Cagayan Valley Region in the Philippines, Centre of Environmental Science, Leiden University, 1995 - Kamminga, E.M., Deforestation in Context: The North-Eastern Amazon Region in Ecuador, Centre of Environmental Science, Leiden University, 1995 - Toornstra, F.H., G.A. Persoon and A. Youmbi, Deforestation in Context: The Southern Forest Region in Cameroon, Centre of Environmental Science, Leiden University, 1995
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1323
International policies to address the greenhouse effect: Encouraging developing country participation in global greenhouse control strategies J. Gupta a, R. v.d. Wurff b, M. Hischenmoller~, G. Junne b and P. Vellinga a alnstitute for Environmental Studies, Vrije Universiteit, Amsterdam, The Netherlands bDepartment of International Relations and Public International Law, University of Amsterdam, The Netherlands.
Abstract This article outlines in brief the conditions under which developing country governments are likely to feel motivated to take real action in addressing the greenhouse gas problem and the international mechanisms that are likely to succeed.
1. INTRODUCTION Although the industrialised countries (ICs) are mainly responsible for past and present greenhouse gas (GHG) emissions, it is expected that developing country (DC) emissions will increase in the future. ICs emphasize that the future emissions of DCs might render their GHG reduction efforts negligible. They do this, either out of a genuine concern or to divert attention from themselves. DCs, on the other hand, perceive climate change as one event in a sequence of problems in North-South relations. Assuming that climate change ultimately calls for the global stabilization of emissions, they are negotiating with the hope (the 'hope scenario') that they will be allowed to emit a 'fair' share on the basis of per capita equity. They are, however, afraid that extrapolation of historical trends and realpolitik might instead imply that IC governments will try and prevent the growth of their emissions and, hence, development (the 'angst scenario'). They see climate change as a test case for Northern sincerity in global partnership. This paper presents a few highlights of our research which focuses on the conditions under which DC actors will take real action to address the climate change problem. It analyzes how the international mechanisms can be tailored towards that end, while keeping the perspectives of ICs in mind. The implementation of the Framework Convention on Climate Change (FCCC) calls for a reduction of GHG emissions in the North and a reduction of the growth of the emissions in the South. In order to achieve the latter, the FCCC recommends the transfer of appropriate technologies, through a funding mechanism (Global Environment Facility) and through the
1324 market (Joint Implementation). ICs have agreed to provide "new and additional" funds to finance the "agreed full incremental costs" (article 4.3) of DCs in implementing national commitments.
2. CASE STUDIES OF DEVELOPING COUNTRIES There are differences in the way climate change is perceived in the four developing countries (India, Indonesia, Kenya and Brazil) studied. However, this article focuses on the similarities underlying the bottlenecks in international cooperation: 1) DCs perceive climate change as symptomatic of the malaise of the international economic and political order, past and present. It is seen in terms of a global inequitable economic order, adverse (non-liberal) terms of trade, etc. 2) Although ICs invited DCs to cooperate on climate change, DCs are alienated by the way ICs conceptualize the problem of climate change. This is because: a) DCs perceive ICs as making an artificial distinction between global (read: Western) and local problems and benefits. For example, desertification is not treated as a climate priority but rather as a separate regional issue. b) DCs perceive that the focus on cost-effective measures in relation to incremental costs in the FCCC and GEF leads to the externalization of social and local costs. This could lead to lop-sided development and social unrest. c) DCs are especially vulnerable to climate change. Hence, they want assistance with adaptation measures in addition to GHG limitation measures. The IC preference for funding only limitation measures is perceived as negative. d) DCs perceive ICs as not taking their responsibilities seriously as they are looking for ways of exporting sacrifices through mechanisms like Joint Implementation and as there are very limited "new and additional" funds. e) DC actors perceive that they receive contradictory messages from related international regimes such as the forest/timber regimes. These messages are inconsistent if environmental objectives are important, but consistent in that ICs use different regimes to promote their own economic interest at the cost of DCs and the environment. 3) The policies of DC governments are more a reaction to international developments than a reflection of societal consensus. Hence, the emanating policies are likely to be symbolic in the hope of meeting both foreign and domestic obligations. 4) Despite the dismal picture painted above there are several policies in climate related fields in these countries and there is considerable space for policy measures that are also likely to address the climate change problem.
3. CASE STUDIES OF INDUSTRIALISED COUNTRIES As in the DCs, critical differences in the way climate change is perceived in the ICs exist. These differences are expressed in the approach adopted by to deal with climate change issues, ranging from a legalistic approach in the UK (we do what we are obliged to do; but nothing more), via a commercial approach in the USA (climate change will offer industry
1325 business opportunities) to a more development oriented approach in Germany (emphasising the development cooperation aspect of JI). However, we focus here on the similarities underlying the bottlenecks: 1) All three ICs emphasize the need to involve DCs in global climate policies because of the perceived need a) for cost effective solutions, b) to ensure that emission increases in the South do not offset decreases in the North, and because of c) perceived opportunities for mutual economic benefits of such cooperation. 2) ICs tend to focus on the role of DCs that are or will become large GHG emitters. They state that there is a need to stop ideological discussions in multilateral fora and start to work pragmatically. 3) ICs perceive DCs as uninterested in environmental issues and interested only in the funding. Although inter-linkages between climate change and other environmental issues are recognized in principle, ICs want to make sure that money spent by them on climate change is really used to reduce GHG emissions. There seems to be consensus that DCs have to pay their share of the costs of climate policies. 4) IC actors are losing interest in climate change and development cooperation partly because of the economic recession. It is unclear whether the countries analyzed will realise their diverging emission reduction targets. 5) ICs perceive that a liberal economic ideology should underlie international climate change and development cooperation policies, in which market forces and trade liberalisation are emphasized. Most actors agree that there is a need to involve industry in addressing the climate change problem. These aspects shape IC positions on the various instruments. They favour JI because of its perceived cost effectiveness and its potential to involve DCs. They also support the GEF and its focus on 'global' as opposed to 'local' benefits. They disapprove of stricter rules for technology transfer, greater financial burden for ICs through a new cost-sharing protocol, and global emissions trading in practice.
4. CONCLUSION If lCs want DCs committed to the problem, they need to modify their strategy by:
1) Matching local priorities with global priorities: If DC actors are to be convinced of the need for climate relevant policies, then such policies have to be linked with their priorities. Climate change affects local as well as global priorities and linking the global priorities to the local ones is possible. However, the dominant international initiatives are artificially differentiating between global and local problems and benefits, thereby sending the message that climate policy is not a priority for the South. This is counter-productive. 2) Making explicit sacrifices: DCs believe that ICs have caused the problem and should take measures to address it. However, it now appears to them that ICs are not serious about the problem. So why should they be serious? 3) Addressing linkages - making regimes consistent: Most issue-linkages (associations with other issues) made by DC actors are in relation to other international regimes that are perceived as inconsistent with the FCCC regime. These include forest related regimes, trade in natural resource regimes, etc. An effort to synchronizing these regimes is necessary.
1326 The above policies are also in the interest of ICs and their criteria, because ICs should" 1) Take real economic cost-effectiveness into account: We believe that it is possible to generate cost-effective measures, a) by reducing the inconsistencies between different regimes, b) by identifying solutions to those global problems that are also relevant for local problems as the first priority options, (which would in a business as usual approach not be undertaken in DCs because of financial and institutional bottlenecks); c) by funding global benefits but not at social costs. 2) Consider political cost-effectiveness: ICs should take not only economic but also political cost-effectiveness of climate change policies into account. Short term viable policies, that are in contradiction with DC priorities and positions, even if implemented, might turn out to be political costly in the future (reducing opportunities for consensus formation). On the other hand, policies that are politically efficient may bring larger economic returns in the future. 3) Show the political courage necessary for sustainable development: The ICs are reluctant to fund additional global measures because economic concerns dominate over environmental concerns. This approach is not consistent with a global sustainable development approach which requires political courage, as well as societal support. Reconunendations for instruments: 1. As there is a decline in the concern for climate change and the issue has become politicised, related fields/fora need to be used to address the climate change issue. 2. Different but related regimes such as on forestry need to be harmonised to ensure that their message and effect is consistent and, hence, cost-effective. 3. Imperfect market conditions leading to the dumping of outdated technology needs to be corrected through some means (regulations, guidelines, etc.). 4. Joint Implementation would be acceptable to DCs if the fear that it is neo-colonialistic (the angst scenario) is addressed. 5. DC actors feel marginalised by the international trends in trade, forcing them to denudate their natural resource base while reducing their financial surpluses. 6. The international debt problem requires serious re-thinking. It is, in combination with the trade regime, one of the major causes of environmental degradation in DCs, as it reduces the surpluses that could be invested in environmental measures. 7. Innovative ways of raising money, building on human nature, such as using lotteries to generate funds for climate change need to be considered. 8. The decline in public interest is counter-productive. There is a continual need for global public awareness programmes. 9. NGOs need to stimulate discussions on global/local sustainable development to make the link between development and environmental problems explicit.
In the final analysis, where the commitment at the level of domestic actors is limited, national policy tends to become rhetorical; where national commitment is limited, the foreign policy becomes symbolic; and where foreign policy is symbolic, the international regime becomes a farce.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1327
Practical aspects of Joint Implementation Axel Michaelowa HWWA-Institute for Economic Research, Neuer Jungfernstieg 21, 20347 Hamburg, Germany
Abstract Article 4, 2a of the UN Framework Convention on Climate Change states the possibility of joint policies of different countries to achieve national greenhouse gas reduction commitments ("Joint Implementation"). The cost of reducing greenhouse gas emissions can be reduced drastically if industrialized countries shift abatement activities to developing countries as marginal cost of reduction is much higher in the former countries. In this way economic efficiency of abatement measures can be raised to the point where marginal cost is equal all over the world. At the Conference of the Parties in Berlin in March 1995, criteria for Joint Implementation are to be established. The paper discusses possible forms of Joint Implementation and develops criteria.
1. JOINT IMPLEMENTATION AS AN INSTRUMENT TO ACHIEVE EMISSION REDUCTION AT REDUCED COSTS It is improbable that there will be an international agreement on efficient global instruments to abate greenhouse gas emissions such as an international carbon tax or tradeable emission rights. Therefore, globally inefficient national policies will prevail that use fiscal or regulatory instruments. To achieve higher efficiency, national policies can be extended by Joint Implementation. Emission reductions achieved through projects abroad are credited to the national goal. Domestic emitters will face reduced taxes or regulation if they prove a reduction abroad. Joint Implementation is a very flexible instrument and does not address difficult distributional issues like taxes or tradeable permits do. It also does not depend on the existence of a global agreement; only the domestic incentives are a necessary condition. Private firms, non-governmental organizations (NGOs) and individuals can take part in Joint Implementation projects if there are sufficient incentives. The level of incentives depends on the amount of tax relief and regulatory concessions granted. The prevailing huge differences in marginal costs of reduction can be reduced via Joint Implementation; this leads to a rise of global abatement efficiency. Moreover, Joint Implementation can be decisive in helping developing countries to attain a development path which is less emission-intensive than the business-as-usual case. These countries could avoid high sunk costs in this way. Some caveats apply, though. Transaction costs, market failures and intransparencies can seriously hamper the efficiency of Joint Implementation. The calculation of reductions achieved through projects depends on a realistic reference scenario ("baseline"). Verification and evaluation of emission reduction seem rather difficult. The necessary precaution should not lead to inappropriate regulation.
1328 2. INSTITUTIONAL DESIGN OF JOINT IMPLEMENTATION The Framework Convention on Climate Change states in Art. 4,2a that OECD countries and countries in transition are allowed to implement abatement measures and measures to protect greenhouse gas sinks jointly with other contracting parties. Art 4, 2b allows cooperation to achieve national abatement goals. The first Conference of the Parties shall develop criteria for this cooperation (Art. 4,2d). In the sessions of the Intergovernmental Negotiating Committee in 1993 and 1994 it became obvious that there is no international consensus concerning Joint Implementation. Especially some of the developing countries fear that the industrialized countries will use Joint Implementation as "cheap buy out". As this argument is not based on economic, but only on political reasons, it will not be discussed in detail. Other criticisms are more valid and will be mentioned below. Nevertheless, the political pressure of the opponents has led to the postponement of crediting reductions of Joint Implementation projects until a pilot phase of several years to test the concept has been completed. Some of the following possibilities to institutionalize a Joint Implementation regime should be tried in the pilot phase: 2.1. Multilateral approach: international fund Countries wishing Joint Implementation credits contribute to an international fund whereas other countries offer projects. The fund management selects suitable, efficient projects. Contributors receive a credit proportional to their share of the total project portfolio. The experience of multilateral institutions (e.g. the Global Environment Facility) could be used and project risks would be pooled. On the other hand the incentive to minimize transaction costs would be reduced as the difference between the stipulated price per ton carbon reduced and the actual marginal costs of reduction accrues to the fund. Moreover, costs of bureaucracy tend to rise. Therefore, this approach should not be pursued further. Some of its advantages can be utilized by installing an international database where investors and project hosts are matched. Access to the database should be free, whereas a fee could be levied for successful matching. 2.2. Bilateral contracts Governments can organize Joint Implementation projects on a national, regional or local level and have them funded by other governments. Especially the local level seems appropriate as cities often have strong ties to twin cities abroad. On the other hand the investing government can hire private firms or NGOs as subcontractors. Moreover, governments can act as facilitators by concluding framework contracts to recognize projects of private firms and NGOs acting independently if a government agency gets a certain amount of information from the project participants. The existence of investment protection contracts can facilitate Joint Implementation contracts. 2.3. Contracts between private firms Private firms are interested in investing into Joint Implementation projects if they can get tax relief or reduced regulation in their home country and the project-related costs are lower than the tax or costs of regulation. Furthermore, such "green" investments are valuable for marketing or lobbying purposes at home and for market entry abroad. Hosts of Joint Implementation projects receive new technology and additional capital at no cost. Governmental recognition of the projects is essential. Private projects will not require costly
1329 bureaucracies and should keep transaction costs low. Additional monetary and technological transfers into developing countries can be generated in this way. 2.4. Contracts with participation of NGOs
NGOs are transferring huge amounts of money into developing countries. Therefore, they should be able to invest and participate in Joint Implementation projects. The projects must be recognized by the governments as indicated above and an investing NGO should be granted an amount equal to the tax relief granted for a private firm as incentive. NGOs are able to implement projects in fields which are not attractive for private or governmental activity, e.g. grass-roots projects with many participants and low budget projects, which do not fit into government priorities, e.g. with information-sharing and educational content. NGOs should be able to integrate large parts of the population in the developing countries into abatement efforts especially in the primary sector whereas in industrial countries they can promote measures in the household sector. They can contribute to behavioural changes. Many NGOs, though, are against Joint Implementation and therefore will be reluctant to take part in projects.
3. CRITERIA FOR JOINT IMPLEMENTATION 3.1. Eligible greenhouse gases
All greenhouse gases should be eligible for Joint Implementation projects. In the beginning, though, only those gases shall be considered which are mentioned in the Framework Convention. New scientific results which lead to changes of emission factors or global warming potentials should only apply for new projects, not for ongoing projects as the investment decision cannot be altered any more. Only where obviously negative effects on climate are recognized as a result of new scientific knowledge, the project should be discontinued. 3.2. Acceptable measures - both reduction and sink enhancement
The most promising category of emission reduction projects is the raising of efficiency in electricity generation on the base of fossil fuels. Further options are substitution of fuels, reduction of energy use in production processes and household use. Another option is the reduction of agriculture-related emissions through change of feedstocks. Sink enhancement, e.g. afforestation or restoration of wetlands should also be accepted, if it can be sustained for a long time. 3.3. Baseline scenarios
The most important question concerning Joint Implementation is the calculation of the reduction achieved. One therefore has to calculate the emission level occuring if the project had not taken place - the baseline scenario. The indirect effects of the projects, e.g. price changes, have to be considered if the baseline shall be realistic. Obviously, developing an aggregated baseline scenario is a complex and expensive matter. As the baseline problem occurs whenever any national reduction goal has to be decided, it is not only a problem of Joint Implementation. A project-related baseline developed with an internationally accepted methodology should be sufficient to evaluate impacts of Joint Implementation.
1330 3.4. Discount rates As Joint Implementation projects have very different durations and the national goals are standardized on certain years there has to be some discount factor for the valuation of reductions at different points of time. It should be positive because there is a distinct probability that damage caused by climate change could rise sharply after certain thresholds of greenhouse gas concentrations have been passed. As abatement now lowers the probability that these thresholds will be passed, it is more valuable than abatement in the future. 3.5. Crediting Crediting of emission reductions should only be possible after independent verification of the reduction and should be in full. Any partial crediting raises marginal reduction costs and is therefore inefficient. Moral hazard, i.e. the tendency to declare unrealistically high reductions, should be contained by strict verification. 3.6. Reduction of leakage By reducing the difference between marginal reduction costs the tendency to relocate production facilities with high emissions abroad ("leakage") if greenhouse gas taxes will be reduced for firms investing in Joint Implementation. The same applies for trade-related effects. The comparative advantage of countries with no climate policy measures is reduced as well as the climate policy-related cost of domestic firms. 3.7. Incentives for transfer and development of technology The autonomous technology transfer into developing countries is likely to be reduced because of the new, relatively strict GATT rules on intellectual property rights. Joint Implementation can reverse this trend by alleviating the transfer of latest technology, e.g. in the case of renewable energy use, as this is the only way to achieve high emission reductions. Because of secondary effects like training of technicians the indigenous technology capacity of developing countries should rise. Economies of scale should become relevant in the case Joint Implementation leads to a rise in application of new, hitherto expensive technology.
3.8. Verification Verification procedures should be compulsory. Any project is to be certified by independent accountants. A national verification board can check the quality of any verification. Standards for verification should be harmonized internationally. As the accuracy of verification is still rather low, the development of more reliable measuring methods should be encouraged and funded. Again, verification is not only necessary for Joint Implementation but also for the control whether national goals have really been reached. 3.9. Sanctions If project participants have intendedly overstated the quantity of emission reduction, they should be excluded from the mechanism of Joint Implementation forever. If countries are involved directly, dispute settlement similar to the new GATT procedures should be applied. It should consist of several layers and include an independent panel and appellative procedures. Finally, countries should be allowed to use retaliatory measures if their counterpart does not comply with the recommendations of the panel.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1331
Industrial energy efficiency in developing countries" present situation and scope for new initiatives J.L. de
Arat~jo 1, S.
Barathan 2, S. Diallo 3, F.M.J.A. Diepstraten 4, J.C. Jansen 4, A.D. Kant 4
1 Instituto de Economia Industrial, UFRJ, Av. Pasteur 250, 22290-240 Rio de Janeiro, Brazil 2 Tata Energy Research Institute, TERI, Habitat Place, Lodi Road, New Delhi 110003, India 3 ENDA Programme Energie, B.P. 3370, Dakar, Senegal 4 Netherlands Energy Research Foundation, ECN, P.O. Box 1, 1755 ZG Petten, The Netherlands Abstract The manufacturing sector in developing countries accounts for a steadily increasing share of world energy consumption and global greenhouse gas emissions. This paper is based on a study on policies to stimulate improvement of energy efficiency in the industrial sector in developing countries. The paper highlights developments in respect of the efficiency of industrial energy use in Africa, Asia and Latin America. The paper begins to outline some salient features of energy and technology use in four energy-intensive industries. Subsequently, recent policy initiatives and institutional development in support of rational use of energy (RUE) in industry are considered. The paper concludes with national policy priorities in developing countries on industrial energy conservation and opportunities for international cooperation.
1. INTRODUCTION This paper gives an overview of some highlights of the research project entitled: Strategies and Instruments to Promote Energy Efficiency in Developing Countries. The
project falls into the theme called 'Sustainable Solutions' of the Dutch National Research Programme on Global Air Pollution and Climate Change (NOP/MLK). The project is financed by the NOP/MLK and by ECN and has been carried out jointly by ECN, ENDA, IEI and TERI. The study sets out: 9 to make an assessment of policy strategies and instruments implemented in developing countries to promote more efficient use of energy in the industrial sector; 9 to put forward recommendations on effective external assistance to improve the institutional framework and to enhance local capabilities to design and implement
1332 measures and programmes to effectively stimulate efficient use of energy in the industrial sector of developing countries. Although NOP/MLK is a national programme, the solution for global environmental problems is strongly dependent on the extent to which countries, especially but not only in the developing world, succeed in managing the crucial problems of: 9 demographic developments 9 eradication of poverty 9 making economic development more environmentally compatible (with implications for both life style and production patterns). These problems are of global relevance, but solutions have to be found at regional and national level. The Rio summit showed that these problems should be treated in a framework of international cooperation as wealth is distributed very unevenly across the world. Sustainable solutions to the greenhouse effect necessarily have to address the energy issue. A growing energy demand results in a rapidly growing contribution to global CO2 emissions [1]. The industrial energy demand in the developing world is already quite large and is bound to rise fast. In fact, the industrial sector is the most important energy end-use sector, accounting for almost one half of the final demand for commercial energy in the developing world. Typically, energy intensity of production in developing-country manufacturing plants is substantially higher than in their counterparts in OECD member states. It can be inferred that developing-country industry encompasses vast energy conservation potential. Besides, compared to other major final demand sectors - notably the residential and transportation sectors - actors in the industrial sector, if operating under competitive market conditions, tend to be more sensitive to market signals and government interventions. The collaborative endeavour has adopted a diagnostic approach. Based on a survey of literature and existing knowledge with the participating institutes of their respective local situation, patterns of industrial energy use and policy instruments deployed to stimulate industrial energy efficiency are reviewed for three developing regions, i.e. Africa, Asia (excluding Japan), and Latin America. The emphasis is on four country cases per region and four industrial subsectors per country. The decision was prompted by the size and the heterogeneity of the industrial sector, the vastness of the developing world at large, constraints of data availability and resource limitations for conducting this study. Special attention has been given to a selection of energy-intensive industrial subsectors, i.e., cement, iron and steel, aluminium, and pulp and paper. As for the geographical coverage of the developing countries the emphasis is put on: 9 for Africa: Cameroon, Senegal, Tunisia, and Zimbabwe; 9 for Asia: Bangladesh, (Republic of) Korea, India, and Thailand; 9 for Latin America: Bolivia, Brazil, Costa Rica, and Mexico. In making this geographical choice it has been attempted to select a sample of four countries per region with a fair extent of representativity for the respective development regions in terms of size (large and small countries), industrial development phase, and geographical dispersion within the region.
1333 2. PRESENT SITUATION
2.1. Sector findings Introduction This section describes the results of a comparison between technology and typical levels of specific energy consumption prevalent in selected manufacturing industries of developing countries and industrialized nations. This comparison and the information on which it has been based have been described extensively in the regional reports [2-4], the in-depth report on cement manufacturing [5] and energy consumption and production technology in industrialized countries [6]. Iron and steel Specific energy consumption (SEC, i.e average energy use per unit of physical output) in steel production plants in developing countries takes on a wide range of values. For some developing-country plants it is comparable with SEC values prevalent in industrialized countries, whereas in others it is more than double. Part of this variation will undoubtedly be caused by difference in technology and operational practices. Furthermore, differences in energy carriers used, differences in the output structure withun the industrial subsector among other local circumstances play a role. In developing countries, a major ongoing development is the upsurge in electric steel making from scrap. In part, new electric furnaces replace antiquated open-hearth furnaces, while on the other hand these contribute to capacity expansion. Dry coke quenching can be used instead of the traditional quenching with water. It improves coke quality, and reduces energy consumption (by recovery of heat) and dust emissions. Yet in a situation of large technological backlog, this would not be a technology to start with to improve energy efficiency. Several simpler and more cost-effective measures would be possible. Furthermore, continuous casting equipment is installed in an increasing number of production facilities. These technologies have proved to be reliable in various business environments prevalent in both industrialized and developing countries. The new production routes for steel still in the development stage are unsuitable for technology transfer projects, simply because they are not ready for production yet. Significant R&D-efforts are necessary to commercialize these concepts. Aluminium For aluminium production the difference in specific energy consumption between developing and industrialized countries is fairly small. Regional differences in SEC are as large as 7%, although differences between specific countries or plants can be 20% or more. Energy consumption in aluminium production depends strongly on the resource used. Aluminium can be produced from pretreated ore (alumina) in which the metal oxide is already freed from the ore matrix. In this case 'only' the electrolysis from alumina to aluminium has to take place. If the starting point is the raw ore, treatment is necessary, accompanied by significant environmental effects. Energy conservation options in developing countries will be similar to those in the industrialized world on many occasions. This is mainly so because of the predominance of transnational companies in the aluminium industry, also in developing countries. Developments with significant
1334 impacts on energy consumption like the use of inert electrodes or the development of entirely new processes have not progressed far enough for wide scale implementation.
Cement Differences in specific energy consumption between developing countries and OECD member states can be as large as 30%. Some clear options exist to improve this situation. The most important energy conservation measure is undoubtedly to blend clinker with a higher proportion of additives for cement making. In many developing countries additives are sparsely used. Energy conservation is roughly proportional to the amount of additives. Use of additives may not require special equipment. Rather information is needed about operational practice and characteristics of blended cements. A second important action is the change from the wet process to the dry process. This avoids the evaporation of water from the mix, which saves a significant amount of heat. In countries where national companies active in the cement sector are technologically backward, outside assistance both financially and technically - may be needed. Furthermore, the use of a so-called precalciner has proved to improve energy efficiency and to be cost-effective. Energy consumption data indicate elevated electricity consumption in developing regions. In OECD countries, significant improvement has been achieved in efficiency of grinding and mixing techniques. In many cases these technologies may be eligible for transfer to facilities which use older, less efficient equipment. The production of cement is also very suitable to use low-quality fuels such as waste products (e.g. tyres) and (cheap) coal. An advantage of the cement production is that residual material from such fuels can be used as additive. Pulp and paper The subsector 'manufacture of paper and paper products' poses special difficulties when comparing energy efficiency and technology based on figures on a national level. Quality of the output (different types of paper), type of resources used, level of integration of process steps (pulping, paper making), and the extent of waste utilization severely limit the drawing of conclusions on the technology employed based on energy consumption data only. On the energy supply side, installation of efficient solid fuel (wood residues/coal) boilers or cogeneration units seems possible at least in a number of cases in the investigated countries. Reduction of steam requirements by installation of improved pressing techniques, coveting of evaporation area and insulation of process equipment may further reduce specific energy consumption. The development of small-scale chemical recovery units (to be used in chemical pulping process) would be beneficial to numerous small-scale paper manufacturing facilities. Conclusions There is ample scope for energy conservation in developing countries in iron and steel, cement production and most likely also in the pulp and paper, while developing-country energy conservation potential in aluminium production appears limited. Much of the typically high SEC values in the South, as compared to the North, can be explained by dated technology, smaller scale, differences in intra-industry output patterns, as well as differences in quality of raw material and fuel inputs. Although part of the underlying factors may prove hard to change. Yet special attention on the part of policy makers is in order. Table 1 summarizes areas in which a role can be played by external assistance.
1335 Table 1 Technology transfer possibilities in studied sectors. Sector
Technology transfer possibilities
Cement
information and training on raising the share of additives - (conversion to) dry process - precalciners - efficient grinding equipment
Iron and steel
- electric furnaces continuous casting - dry coke quenching
-
Aluminium
data point at relatively small differences in SEC between the OECD area and the developing world
Pulp and paper
-
2.2.
Regional
dual fuel boilers (e.g. biomass/coal) small scale chemical recovery units efficient pressing equipment insulation measures cogeneration equipment and legislative and regulative experiences
issues
Africa
With the notable exception of Tunisia, the countries in the African region still focus almost exclusively on energy supply policy issues. The limited number of energy efficiency measures taken are quite recent and appear hesitant. These largely concern surveys of current efficiency levels and the scope for improvement. To date, only a few countries have to date ventured beyond the reconnaissance stage and then only marginally. Effective institutional settings for the formulation of longer-term energy policies/strategies are conspicuously absent. Most regional economic cooperation organisations, such as SADC, ECOWAS and ECCAS, do not appear to give much specific attention to energy demand management issues, while even regional cooperation on energy supply has still made moderate advances. In some African countries, mainly in energy exporting countries, a line ministry is responsible for energy policy issues. In several other countries energy policy making is entrusted to energy directorates within a ministry or to national (inter-ministerial) energy councils. However, energy supply security and energy tariff issues are the dominant preoccupations. Besides, severe budget restrictions prevail on policy implementation. In most African countries energy prices are still below economic costs. Hence, industrialists lack the proper incentives to investigate possibilities for improvement of energy efficiency. Furthermore, given the generally poor information infrastructure and the lack of capable energy service companies, transaction costs to African entrepreneurs for obtaining reliable information on feasible, low-investment-cost options to improve energy productivity are high. As a result, little awareness exists of the vast energy conservation potential that can be realised - even with the existing, largely dated capital stock.
1336 Moreover, presently available capabilities to cash in on energy savings at low investment costs from upgrading operational practices and energy monitoring procedures are limited at both management and workfloor level. Present conditions in the product markets of most African industrialists do not encourage much effort on their part to invest in energy efficiency improvement. Even if energy prices are raised to a level reflecting full economic costs, poor market prospects may inhibit other necessary outlays to cater to short-term product demand. On the continent, international cooperation is lacking and domestic markets are small. In addition, reliability of public energy supply facilities is poor, occasioning both costly production discontinuities and part-load production as well as the need for captive power systems. This unfavourable business environment severely hampers efforts to achieve higher energy efficiency. Asia The pursuit of industrialisation policy for economic development created a pressure to increase commercial energy consumption in Asia. Along with increasing rates of urbanisation, the switch from non-commercial to commercial energy use became more pronounced. The enactment of energy conservation legislation and the establishment of implementing agencies has been made in Indonesia, Malaysia, Philippines, Thailand, Pakistan, and Sri Lanka. These countries have developed programmes to conduct regular training courses, seminars and the establishment of pilot projects to demonstrate the viability of energy conservation measures. Although oil taxes were levied in net oil importing countries, the conservation programmes and measures had limited access to these funds other than for their administrative requirements. Amongst the countries studied in Asia, the Republic of Korea seemed most impressive in pursuing a national energy conservation programme. The 'Rationalization of Energy Utilization Act" placed energy conservation as a very high priority on the government agenda. The energy conservation strategy falls into four broad categories: information dissemination and regulation, financial incentives, research, development and demonstration (RD&D), and structural changes. In addition, the financial assistance made possible through the setting up of the Petroleum Business Fund has been instrumental in making the national effort a relative success. As for India, there has been some improvement over the years in the efficiency of use of commercial energy in several sectors of the economy. However, these efficiency improvements have been inadequate to make a visible impact on the pattern of growth of demand for commercial energy. It is, however, hoped that the comprehensive National Energy Efficiency Programme (NEEP) in the Eighth Plan, which is to coordinate and organize, existing and, new efforts/activities on energy conservation in the different sectors of the economy for achieving targeted energy savings may prove successful. In Thailand, more recently, with the privatisation of the electricity sector, the promulgation of the Energy Conservation Law and the implementation of the Demand Side Management Programme, the strategy to conserve energy has gained further strength and potential. As for Bangladesh, the Energy Monitoring and Conservation Centre conducts energy audits and holds training programmes and seminars. The success of this unit has been reasonable given the limited resources.
1337 In Asia, a number of barriers were found to the implementation of energy conservation policies. The key barriers were: - inadequate energy pricing policy lack of information lack of innovative financial incentives - socio-cultural barriers. The importance and dynamism of small and medium-scale industries is one of the distinguishing features of the manufacturing sector in Asia. Also, in parts of India and China that are undergoing fundamental reforms in the manufacturing sector, this is the most dynamic subsector. Because of these developments, streamlined information dissemination to this sector may prove very effective. Governments of many energy-importing Asian countries -including notably China and India - have revised energy pricing policies rather drastically, bringing energy prices charged to industrial users more in line with long-term marginal cost of energy. In Korea, energy prices have played a significant role in determining the energy intensity in the manufacturing sector. The high energy intensity in manufacturing sector during the 1980s has been the result of a combination of lower energy price with lower capital investment for efficiency improvement. Another example of the crucial role of energy pricing is that analysis indicated that the DSM programme has limitations in the market without rationalizing the energy price in Korea [7]. In Thailand, it is observed that the energy prices were still low, representing only a small fraction of their total cost. In Bangladesh, the price of natural gas has been unrealistically low which did not provide an incentive for conserving energy. A point that warrants mentioning is that with the exception of the newly industrialised countries of East Asia, the DCs continue to be trapped in a technology-import spiral: imported technology into developing countries seldom reaches its design capacity and its performance is more unstable and decreases much faster over its operational life than is usually the case in industrialised countries [8]. Latin
America
After the second world war, Latin America had a long period of continued economic growth. Import substitution policies induced industrialization and urbanization, changing the economic and social situation of the region. These policies led to economic development, but they brought in severe economic distortions as well. Trade balances ran a deficit most of the time, the tax regime was unable to finance public expenditures and government policies were strongly oriented to foster industrialization and urbanization, at the expense of rural activities. The outcome of these imbalances was inflation, economic instability and profound social inequalities. Despite large inter-country disparities the economic situation of all Latin American countries deteriorated rapidly in the 1980's. The lost decade slowed down the industrialization process, and reduced the per capita GDP. More recently, economic reforms, relaxed debt obligations and opening markets present a changed picture that warrants a cautious optimism Latin America has a mixed energy profile. The region has large energy resources, exporting a substantial surplus of oil (and, increasingly, Colombian coal as well). There is no reason to believe that a major disruption in energy supply can occur in the future, although a large change in oil price could be extremely harmful to non-exporting countries. The intensity of energy consumption per unit of GDP showed a worrisome trend in the
1338 1980's: energy intensity either increased or remained constant, save in the Brazilian case where it tended to decrease slightly. It is important to remark that this occurred together with stagnant per capita energy consumption. Undoubtedly, pricing policy played an important role in this process. Domestic availability of low cost fuels induced energy prices in the region to be kept at a relatively low level. More seriously, political or macroeconomic considerations led to price distortions through cross-subsidies or readjustments below inflation rates, hampering the emergence of energy-efficient behaviour. From the environmental point of view, the Latin American picture is relatively favourable thus far. Latin America largely uses hydropower for electricity generation and biomass for fuelling industry and households. Transportation however is a major consumer of oil and oil products. This situation places regional emissions per unit of GDP below industrial countries' emission levels. Nevertheless, there are signs that GDP growth is likely to boost inefficient use of fuels in the region, if no appropriate policies are enforced. In this circumstance, emissions will have a substantial increase, rapidly deteriorating the environment. The manufacturing sector has grown relatively fast in Latin America until the 1970's. The debt crisis dramatically changed that trend: the share of industry in the regional GDP dropped from 37.1% in 1980 to 31.3% in 1991. This process has substantially changed the structure of value added, since distinct industries have reacted differently to the economic crisis. Globally, the region has moved towards more energy-intensive industries (chemicals and basic metals), taking advantage of its large natural resource base to generate a trade balance surplus that could pay for its external debt. Traditional industries such as food and textiles showed a reduced share in the regional industrial output. Industry is the main energy consumer in Latin America. The economic crisis cut the rate of growth of energy consumption in the manufacturing sector; but as soon as the economy recovers, this rate will increase once again, as the Brazilian and Mexican experiences indicate. Average industrial energy intensity is similar in large and small countries of the region, irrespective of large differences among countries regarding the structure of their manufacturing output. Generally, the energy crisis, particularly the second oil shock, has greatly increased Latin American government's awareness of the important role of policies for the rational use of energy. In a first step, policies were set up at sector level; by the late 1980's, agencies were established for energy policy design and implementation. Education, dissemination of technical information and auditing are basic components of the regional approach to energy conservation; financial support and tax incentives have been less successful. More recently, the regional movement towards the elimination of energy subsidies is likely to have a very substantial impact on the pattern of energy use in industry.
1339 3. POLICY OPPORTUNITIES
3.1. Policy priorities in developing countries From the cases studied, we may derive a few conclusions of a general character. Important opportunities exist to improve energy efficiency in the industrial sector. However, effective use of these opportunities requires consistent, lasting government policies. Government action is required in a multifold way, which would adjust the classical intervention patterns while adding important new dimensions. In fact, significant success has been achieved in some cases by government agencies acting as negotiators between parties that would remain isolated otherwise, and effectively establishing a network of firms and laboratories. Also significantly, other actions failed when such a network could not be established. The role of government in fostering the rational use of energy may cover a broad range of activities and should present a set of features.
First, the traditional functions of government as regulator remain in force, particularly in the energy sector. In this respect, special attention should be given to pricing policy. Price distortions have been a common feature in the cases studied, leading to inefficient energy use patterns. Such distortions often originated from ad-hoc policies to fight inflation and appease dissatisfaction by holding public prices down; in other cases, they have reflected particular policies. In any case, their impact has been negative from the point of view of the financial health of the energy sector and public finance on the one hand and rational energy use on the other. Another task of major importance is the setting-up of a mission-oriented, coordinating agency for the rational use of energy, suitably empowered and endowed. Other traditional objects for regulation concern externalities of diverse sorts, among which environmental impacts closely touch on the theme of this conference. Second, government policy should be consistent with long-term projects for economic development and with a drive towards overall efficiency. As important, policy should be perceived as sound, consistent and durable to gain credibility and effectiveness. Credibility is not an easy condition to achieve, but its absence has been the underlying factor in several programme failures. In particular, energy efficiency should be approached in the context of economic efficiency. Third, traditional government functions like education and strengthening of research institutions are important elements in a strategy for enhancing energy efficiency in the economy. By themselves, however, they may fail to achieve their goals. Government has thus to take on a fourth function: that of acting as a negotiator and mediator between parties, for effective interaction and networking and the setting up of realistic goals. This is of particular importance for information diffusion, technological capacitation and design of efficiency standards. A fifth traditional mode of government intervention, i.e. credit incentives to conservation projects, has been less successful for reasons of inadequate design. This does not preclude it as a valid mode of action. However, considerable care has to be exercised in its design so that it may achieve its goals. Furthermore, in many developing countries the scarcity of public funds suggests that these actions be taken with international support.
1340 The above description has touched upon a few policy priorities directly linked to government functions, as perceived by this study. However, they do not exhaust the list of needs for action. These include, besides the above: 9 Improving awareness of decision makers and the general public. There is much scope for making the general public aware of the importance of behavioural change toward energy conservation and good equipment maintenance. General education and targeted mass media information are major modes to sensibilize the population. This is relevant to energy conservation in industrial plant and office facilities, as well as in many other areas. On the other hand, policy makers must be aware of the importance of Rational Use of Energy policies in the first place. 9 Bridging the technical information gap. In particular in developing countries, reliable information on technical options to improve industrial energy efficiency is in short supply, either about substantial changes requiring large investments, such as change in production process or major retrofits, or actions like "good housekeeping" and minor retrofit measures such as improving insulation, burners, heat recovery and introduction of process control equipment. This implies the need for extensive training activities, audits and general technical assistance to industry. To this end, existing educational and research institutions should be strengthened; energy services companies (ESCOs) may be of invaluable help in auditing and other technical assistance to industry. 9 Improving the availability of energy-efficient equipment. Another major hurdle to technical improvements is unavailability of energy-efficient devices in developingcountry markets. This may relate to lack of demand from industrial users. Less energyefficient plant equipment may be preferred because of ignorance, or because it is cheaper or robust (withstanding poor maintenance, bad energy quality or other unfavourable conditions). Other reasons might be high transportation costs (bad roads, small markets, poor port handling) and import protection of the country concerned. 9 Improving technological capabilities. Introducing improved hardware is not sufficient to improve general productivity and energy efficiency. In developing countries, rated capacities of plant equipment may never be achieved, while equipment performance tends to deteriorate much quicker than in industrialised countries. On the one hand, this relates to harsh operating conditions. On the other, to low technological capabilities of both management and workers. In transferring technologies to the developing world, much more emphasis on the soft side of technology use, and on long-term issues, needs to be put. This suggests that technology transfer schemes (e.g. BOT and BOOT) should be designed with much greater care than has been the norm. 9 Improving access to finance. Economically and socially justified investments for energy conservation in developing countries face a scarcity of financial resources. Several lines of action can be pursued: capital market development, improving business climate for private investment (e.g. franchising, joint-ventures, foreign investments) and introducing special credit lines to investments in rational use of energy in industry. The latter line of action needs much prudence and care to avoid diversion to other uses, insufficient or ineffective use, and biases in favour of large, energy-intensive industrial facilities. In targeting energy-intensive industry, a trade-off has to be made between better energy conservation results and introducing biases in favour of big energy users. 9 The items above point to the importance of networking. It has been found (e.g. in Brazil) that effective interaction of government agencies, teaching and research
1341 institutions, user as well as equipment producer firms may bridge several gaps through synergy and induce efficiency-seeking behaviour. 3.2
Institutional
implications
To implement energy conservation programmes, government interventions as well as energy utility actions are necessary. However, it is of crucial importance that the general policy context is consistent with the programme aims. Macroeconomic, trade and industrial policies should be sound and coherent, and perceived as such. Past experience suggests that, generally, interventions of a closed command-and-control type are less effective in achieving energy efficiency goals than those obtaining voluntary commitment of targeted energy users. Past experience also shows that RUE programmes designed to cope with an emergency have often proven inadequate for lack of adequate data and a legislative structure. A clear legislation differentiates between the main energy sources and provides clear responsibilities, rights and obligations to the important actors: government, industry and trade. In developing countries, laws on basic energy end-use data collection in industry and standards and regulations for energy using equipment have proven to be useful for the design of effective energy policy making and interventions. The cases also suggest the importance of a national coordinating agency for RUE activities. Irrespective of the exact institutional framing of an energy policy agency (at arm's length of the ministry or as a part of the ministry), this should have resources and power adequate to the task; it should also take into account the position of other actors in the energy field, and use them as much as is adequate; these include, among others: ESCOs; utility-based DSM agencies; private industry associations; E-cells (energy management and conservation cells within industrial companies). A national energy conservation agency or institute (ECA) can perform, among others, the following functions in the field of industrial energy conservation: - training target groups: ESCOs, E-cells, DSM agencies, industrial bodies, ministries initiating, monitoring and evaluating industrial audits, using ESCOs and DSM agencies -demonstration programmes to disseminate feasible energy-efficient options and technologies with a large replication potential gather data and identify plants with best energy-efficiency performance that may be widely replicated - encourage private-public partnerships to achieve conservation goals set at branch and plant levels, based on info from abroad and national 'best practices programmes'; standards setting and equipment labelling can also follow this approach - information campaigns (in cooperation with branch associations and ESCOs) - educational campaigns, in cooperation with the education ministry technical management of special credit line programmes (in cooperation with appropriate financial institutions). ECAs have to be government-sponsored as many activities cannot be completely commercialised. On the other hand, complete or partial cost recovery for activities with a commercial character should be given serious consideration, as the willingness of clients to pay gives important feedback information on the societal usefulness of the institute's activities.
1342 Nonetheless, the capacity to pay (for instance of small-scale enterprises) should also be taken into account. A source for government funding could be an indirect tax on energy, notably on fossil fuels. Such a tax should be economy-wide but moderate, so that the competitiveness of national firms in energy-intensive branches vis-a-vis foreign producers will not be seriously affected.
3.3. Opportunities for international cooperation Analysis of the opportunities for national policies shows that international cooperation has an important role to play in developing countries. Government commitment to a sound, consistent policy to improve energy as well as economic efficiency is a necessary condition for success, but may be insufficient without external help. This is particularly the case in poorer countries, or those with an incipient industrial basis, but in most or all cases cooperation will be beneficial. We see the following opportunities for useful international cooperation: Strengthening the capabilities of energy policy agencies, particularly at the national level. In most developing countries, energy policy is poorly coordinated. Programmes for the rational use of energy are often ad-hoc and lack continuity; agencies have to use staff that has little formation in the required analytical skills. Furthermore, information gathering on energy use is most often insufficient to design and target specific actions. The situation is particularly critical in Africa, but in many Asian and Latin American countries there is also ample scope to improve technical capabilities of agency staff to design, implement, monitor and evaluate integrated energy strategies. Assistance could take the form of training the technical staff at such agencies, either on the job or through traineeships abroad; interchange of experiences with well-developed institutes would be another mode of cooperation. Support for this could come through well-designed bilateral arrangements or through multilateral organizations like regional development banks or regional energy organizations. Strengthening national regulatory agencies. Globally, important new developments are taking place in the energy supply industry, with consequences on energy regulation. Several aspects of a RUE policy also have implications for energy regulation: demand-side management, stimulus to co-generation and to independent generation can be rational from a societal point of view, but require important and delicate changes in the way industry is regulated. Here, cooperation might be more effective through experience interchange, given the wide variation in national contexts. Strengthening other relevant agents, such as research institutions, E-cells and ESCOs. These agents have important functions in a consistent RUE policy; the former are important components of training programmes and the development of technological capability. The latter are crucial for a vigorous programme of auditing and technical assistance to industries. Cooperation could be done through participation in national training programmes, and assistance to national networks. Facilitating the access to information on energy-efficient technology. This is an important component of several lines of cooperation. Up to date knowledge of efficient options requires systematic access to information not readily available in developing countries (and perhaps elsewhere as well). This line of action might take e.g. the form of a multilaterally managed data base on existing technologies.
1343
Assistance to design and fund credit lines of adequate design to foster industrial energy efficiency. We have seen that financial incentives may fail for poor design or lack of funds. Cooperation should aim at a three-pronged goal: to identify viable projects in need of funding, to design effective financial stimuli and to devise ways to fund the actions. Facilitating transfer of energy-efficient technology. As argued above, this should go beyond the simple process of equipment acquisition, installation and operation to include 'soft' aspects of technical know-how, aiming at an effective capacitation of the receiving end. Given the two-sided nature of such exchanges, and since firms who own transferable technology have little interest in going beyond the usual arrangements, cooperation of international organisms would be important in facilitating this process. International networking. A careful survey of the short list above indicates that an important role for international co-operation will be to stimulate the interchange of experiences, and more broadly to foster networking at both local and international levels. Information interchange could be done through the creation of a world-wide information centre; other possible interchange channels are international seminars and networks of institutions. Each has its own virtues and limitations. Possibly, a world-wide information centre will be adequate for managing a data base on efficient technology, but fail in other respects. Seminars are useful for exchanging views but have little continuity. Networks may be the best bet for ensuring continuity and depth; but they are slow to build, and easy to break: they require continuing attention and support. In our view, effective co-operation will require thoughtful use of all three channels. 3.4. Concluding remark This study has corroborated the presumption that a vast energy conservation potential exists in developing-country industry. Some highly effective energy efficiency actions have been identified, while warranted policy measures and institutional frameworks have been outlined. The study outcomes suggest that, given a conducive general policy framework, cooperation between North and South in expediting industrial energy efficiency programmes might well rank among the most effective lines of action to implement the Framework Convention on Climate Change and Agenda 21.
References
[1] [2]
[3] [41
V.W. Buskens, J.C. Jansen, Industrial energy demand and C02 emissions in developing countries in global perspective, ECN-C--94-039 S. Diallo, Efficacit~ ~nerg~tique des industries manufacturiOres en Afrique, ENDA Programme Energie, Dakar, Senegal J.L. de Arafijo, A. de Oliveira, E.A. Guimaraes, R. Tolipan, Regional Project Study on Latin America, final report, IEI, Rio de Janeiro, Brazil S. Barathan, P. Bhandari, P. Daclhich, Regional Project Study on Asia, final report, TERI, New Delhi, India
1344 [5] [6] [7] [8]
F. van der Vleuten, Cement in development, Energy and Environment, ECN-C--94035 V.W. Buskens, F.M.J.A. Diepstraten, Energy efficiency in selected industries of the manufacturing sector, ECN-C--94-040 J.D. Kim, Priorit&ation of energy efficiency technologies and strategies to improve effectiveness of DSM policy, Korean Energy Economics Institute, Korea, 1993 AEI, Climate Change in Asia and Brazil: The Role of Technology Transfer, New Delhi, India, 1994
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1345
MACROECONOMIC ANALYSIS OF COz EMISSION LIMITS FOR CHINA ZhongXiang Zhang a, Henk Folmer a , and Paul van Beek b a Department of General Economics, Landbouwuniversiteit, P.O. Box 8130, 6706 KN Wageningen, The Netherlands
b Department of Mathematics, Wageningen, The Netherlands
Landbouwuniversiteit,
P.O. Box 8130, 6703 HA
Abstract Using a newly developed time-recursive dynamic CGE model for energy and environmental policy analysis of the Chinese economy, a business-as-usual scenario is first developed assuming no specific policy intervention to limit the growth rate of CO2 emissions. Counterfactual policy simulation is then carried out to compute the macroeconomic implications of a carbon tax to limit the Chinese energy-related CO2 emissions.
1. INTRODUCTION As a developing country, China is currently undergoing significant transformation. This has led to spectacular growth of the Chinese economy, with an annual growth rate of about 9% for GNP during the period 1980-90. In the meantime, energy consumption rose from 602.75 Mtce (million tons of coal equivalent) in 1980 to 987.03 Mtce in 1990. The corresponding CO2 emissions grow from 358.60 MtC (million tons of carbon) to 586.87 MtC during the same period (Zhang, 1994a). This means that China ranks second in global CO2 emissions when Soviet emissions are split among the newly independent republics. Assuming a business-as-usual scenario, China's contribution to global CO2 emissions is estimated to rise from 11% in 1990 to 28% in 2100 (Manne, 1992). Thus, advocates of controlling CO2 emissions call for substantial efforts in China. Indeed, given China as the world's most populous country and largest coal producer and consumer, her coal-dominated energy structure and her energy-and-carbon-intensive economy, her economic development and her efforts to limit CO2 emissions are of great influence on future global CO2 emissions. This study aims to explore the evolution of the Chinese energy-related CO2 emissions in the absence of a carbon limit and to analyze the economy-wide impacts if a carbon tax is imposed to achieve the predefined carbon limits. In doing so, we have chosen a computable general equilibrium (CGE) approach. This choice has been motivated by the wide recognition of CGE approach as an appropriate tool for such a purpose (Zhang, 1994b).
2. GENERAL FEATURES OF THE CGE MODEL OF THE CHINESE ECONOMY The CGE model for energy and environmental policy of the Chinese economy is of a time-recursive dynamic structure. It operates by simulating the operation of markets for factors, products and foreign exchange, with equations specifying supply and demand behaviour across all markets. Moreover, since focus is olaced on auant-
1346 model pays particular attention to modelling the energy sector and its linkages to the rest of the economy. In our CGE model, energy use is disaggregated into coal, oil, natural gas and electricity. Along with capital, labor and intermediate inputs, the four energy inputs are viewed as the basic inputs into the nested constant elasticity of substitution-Leontief production function for each producing sector. Our model includes ten producing sectors and four types of agents (producers, households, the government, and foreigners). It is made up of the following blocks: production and factors, prices, income, expenditures, investment and capital accumulation, foreign trade, energy and environment, welfare measures, and market clearing conditions and macroeconomic balances. The model allows endogenous substitution among energy inputs and alternative allocation of resources as well as endogenous determination of foreign trade and household consumption in the Chinese economy for coping with the environmental restrictions. Thus, the CGE model makes it possible to analyze the Chinese economy-energy-environment system interactions simultaneously, both at the sectoral level and at the macroeconomic level. The equilibrium solution for a given year produces a wealth of detailed information, including market clearing prices, GNP, productivity levels by industry, investment by industry, final consumption levels by commodity, employment by industry, imports and exports by commodity, energy consumption in physical terms, and CO2 emissions. See Zhang (1995) for a detailed description of the model.
3. THE BUSINESS-AS-USUAL (BaU) SCENARIO 1
The BaU scenario assumes no policy intervention to limit the rate of C O 2 emissions, but does allow for anticipated changes in demographic, economic, industrial and technological developments and environmental policies not directly aimed at carbon abatement. To develop the BaU scenario using the time-recursive dynamic CGE model involves two steps. The first step is to make a set of underlying baseline assumptions about how the exogenous variables in the model would evolve over the period till 2010. This involves updating time-dependent variables and revising certain parameters over time to reflect the world and Chinese economic development as well as the changes in tastes or technology. The second step is to use these assumptions to construct the BaU projections about the endogenous variables mentioned earlier. Table 1 shows some selected results, while a detailed representation of the CGE simulation results for China will appear in Zhang (1995).
Table 1 Selected CGE simulation results for China (abbreviations are explained in text) 1990
GNP (Billion Yuan in 1990 prices) Population (millions) Energy consumption (Mtce) Elasticity of energy consumption Elasticity of electricity consumption Annual rate of energy conservation (%) CO2 emissions (MtC) CO2 emissions per capita (tC)
13763 1093 987.0 0.56 0.84 3.6 586.9 0.51
2010 BaU
CAS
61946.7 1413 2515.6 0.61 1.01 2.8 1527.5 1.08
61301.0 1413 2047.1 0.48 0.93 3.7 1242.6 0.88
1347 4. CARBON ABATEMENT: C O U N T E R F A C T U A L P O L I C Y SIMULATION It has been argued that a carbon tax is an effective means of providing economic incentives for limiting CO2 emissions (Zhang, 1994c). For this reason, our CGE model incorporates a carbon tax as an economic instrument to serve such a purpose. In counterfactual policy simulation, we impose a carbon tax of 300 Yuan per ton of carbon to achieve 20% cut in CO2 emissions in 2010 relative to the BaU scenario. The carbon tax is equivalent to 25 in 2010 US$. Table 1 gives some selected results (in absolute values) of the carbon abatement simulation labelled as the CAS, while Table 2 reports main economic effects of the carbon tax relative to the BaU. For a representation of alternative CGE policy simulation results of sectoral detail for China, see Zhang (1995).
Table 2 Main macroeconomic indexes in 2010 Percentage derivations relative to BaU (-: declines) GNP Exports Imports Private consumption Energy consumption CO2 emissions
- 1.042 - 2.778 - 0.675 - 0.858 - 18.624 - 18.746
The results in Table 2 can be briefly explained as follows. The unilateral imposition of the carbon tax to cut the Chinese energy consumption and hence CO2 emissions will reduce the international competitiveness of the Chinese products. As a result, China will face a decline of export volumes. Such a decline also means a loss of GNP, given that the increasing exports are one of the driving forces behind China's booming economy. Moreover, the combination of a decline in export volume plus a rise in export prices makes import volumes fall somewhat less than exports. With private consumption constituting part of the final demand (GNP), this will therefore lead to that private consumption needs not to fall as much as GNP.
5. C O N C L U D I N G R E M A R K S From the preceding analysis we draw the following main conclusions. First, driven by the threat of further degradation of the environment and the harmful economic effects of energy shortages, China has made and will continue to make great efforts directed at energy conservation and enhanced energy efficiency whether or not the global climate change issue requires special action on China's part. This can be reflected by the fact that, with an annual energy saving rate of 2.8% during the period 1990-2010, China would achieve an annual economic growth rate of 7.8%, at the same time the Chinese CO2 emissions per capita in 2010 are still controlled below the current world average (1.14 tC in 1990)under the BaU path. Moreover, it is conceivable that the Chinese government is to take a broad range of measures to further slow down the growth of her per capita emissions when curbing global CO2 emissions requires special action on China's part. Second, compared with some global studies that treat China as a separate region, we analyze the impacts of a less restrictive carbon emission scenario in which
1348 China's CO2 emissions in 2010 are allowed to be 110% higher than her 1990 level. Even in this case, China would face a GNP lose of 1% relative to the BaU. This supports the general findings that China would be one of the regions hardest hit by carbon constraint (Manne, 1992). Finally, in our calculations, the imposition of a carbon tax (25 in 2010 US$) would cut CO2 emissions by 285 MtC in 2010. This tax is much smaller than those reported for the industrialized countries to achieve the same amount of carbon cutback. This suggests that international action should consider, among others, joint implementation as an useful means of reducing global CO2 emissions effectively. This mechanism will not only help China alleviate the suffering from the future possible carbon limits, also act to reduce the pressure put on the industrialized countries for yet stringent measures to stabilize global CO2 emissions.
6. NOTE 1The results reported here are only preliminary and are subject to changes. This also applies to the results of counterfactual policy simulation. For a detailed representation of the final results, see Zhang (1995).
7. REFERENCES Manne, A.S. (1992), Global 2100: Alternative Scenarios for Reducing Carbon Emissions, OECD Department of Economics and Statistics, Working Papers No. 111, Paris. Zhang, Z.X. (1994a), Analysis of the Chinese Energy System: Implications for Future CO2 Emissions, Int. J. of Environment and Pollution, Vol. 4, Nos. 3/4. Zhang, Z.X. (1994b), Economic Approaches to Cost Estimates for Limiting CO2 Emissions, Quantitative and Technical Economics, Vol. 11, No. 12 (in Chinese); Forthcoming in Int. J. of Environment and Pollution, Vol. 5, No. 1, 1995. Zhang, Z.X. (1994c), Setting Targets and the Choice of Policy Instruments for Limiting CO2 Emissions, Energy & Environment, Vol. 5, No. 4. Also published in Wageningen Economic Papers No. 1994-2, Department of General Economics, LUW, Wageningen, The Netherlands. Zhang, Z.X. (1995), Integrated Economy-Energy-Environment Policy Analysis: A Case Study for P.R. China, Final Report to the Dutch National Research Programme on Global Air Pollution and Climate Change, Department of General Economics, LUW, Wageningen, The Netherlands (forthcoming).
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1351
A S S E S S M E N T R E P O R T ON N R P S U B T H E M E
?RISK ANALYSIS"
W. Biesiot and L. Hendrickx University of Groningen Center for Energy and Environmental Studies P.O. Box 72 9700 AB Groningen The Netherlands
With contributions by: J. van Ham A.A. Olsthoorn
TNO, Institute for Environmental Sciences Delft VUA, Free University of Amsterdam
1352 Contents Abstract 1. Q
0
0
Introduction The risks of non-linear climate c h a n g e s 2.1 Summary of findings 2.2 Discussion of results S o c i o - e c o n o m i c a s p e c t s of e x t r e m e e v e n t s 3.1 Summary of findings 3.2 Discussion of results C h a r a c t e r i z i n g the risks 4.1 Summary of findings 4.2 Discussion of results
5.
General discussion
6.
References
ABSTRACT This report presents an overview and assessment of the three research projects carried out under NRP funding that concern risk-related topics: (1) The risks of nonlinear climate changes, (2) Socio-economic and policy aspects of changes in incidence and intensity of extreme (weather) events, and (3) Characterizing the risks: a comparative analysis of the risks of global warming and of relevant policy strategies. 1.
INTRODUCTION
The great degree of uncertainty in the dynamics of the climate system, in the (quality and quantity of the) effects that may be expected, and the long time delays between causes and effects imply that the greenhouse problem can be regarded as a risk problem. The consequences of climate change may be great, and the probability of its occurrence is hardly known. Risk analysis is therefore regarded under the NRP program to be an important means for the evaluation and assessment of the possible (positive and negative) consequences of climate change. This not only implies the necessity of research into the risks attached to global change itself and the relationships with the presently existing natural and man-made risks, but also research into to risks attached to the social and policy response. This is pertinent to the Dutch situation, where Dutch climate policy is based upon the precautionary principle, proceeding from a risk approach to the formulation of environmental quality objectives. These are provisionally based on
1353 the indicators and environmental quality objectives formulated by the RIVM and the Stockholm E n v i r o n m e n t a l Institute. The continued development of such indicators and risk limits requires regionalisation, attention to extreme values and non-linear effects, and analysis of multi-stress situations. Such a line of research should be related to the limited analyzing power of current climate models and thus to the relevance of additional i n f o r m a t i o n and to the p r o p a g a t i o n (and accumulation) of uncertainties of various kinds. The programming m e m o r a n d u m "Integration" of the Dutch NRP on Global Air Pollution and Climate Change thus concludes t h a t research is needed in the analysis of risks associated with the various effects of climate change, and also in the comparative analysis of the risks of different policy strategies on national, European and global levels. This should result in a comparison of the risks and costs associated with strategies directed at adaptation or prevention or directed at temporisation (too early or too late) in the adoption of measures. Specifically, the memorandum proposed the following studies: 1. Research (in close collaboration with others NRP bodies) into regional indicators of global change. 2. A systematic inventory of extreme climatic conditions and of non-linear events that have been neglected thus far. 3. A systematic inventory and an assessment of the risks and costs associated with various abatement/adaptation/temporisation options. 4. A coordinating study into the development of an integrated risk approach which should enable the support and evaluation of the setting of standards. The selection process t h a t followed the publication of the m e m o r a n d u m has resulted in the funding of the three projects described in this assessment report. They cover only a part of the requirements formulated above, due to a limited supply of project proposals and to limitations in the NRP funding. Table 1.1 List of Projects in the NRP subtheme "Risk Analysis" Title
Project leader
Number
The risks of non-linear climate changes
J. van Ham
853118
Socio-economic and policy aspects of changes in incidence and intensity of extreme (weather) events
P. Vellinga
853137
Characterizing the risks: a comparative analysis of the risks of global warming and of relevant policy strategies
L. Hendrickx
853114
A common characteristic of the projects described here is a relatively late start in the first phase of the NRP-program, namely only in the summer and fall of 1993. Not surprisingly, none of the projects has been completed at the time of writing of
1354 the a s s e s s m e n t report. Two of the projects (number 2 and 3) are scheduled to continue into mid 1995. A s s e s s m e n t of such studies is not an easy job, especially because the issues involved are complex and because the studies under focus are in general first-of-akind projects. Developing a suitable research methodology tends thus to be an iterative process. Producing interim documents for the NRP evaluation process m a y very well not be the first priority of the teams involved, nor apt to generate the most useful ingredients for an assessment. Special procedural attention should be given to the second project as it has been financed in two steps: after a preliminary study the results have been evaluated by the NRP programming committee in order to judge the wisdom of continuation with the other phases. This evaluation is not laid down in written documents, but has resulted in modifications of the original research plan, putting more emphasis on the description of the possible socio-economic aspects t h a n on scenario development.
2.
THE RISKS OF NON-LINEAR CLIMATE CHANGES
2.1 S u m m a r y of f i n d i n g s This project aims to provide a survey of non-linear mechanisms of global change t h a t operate in addition to those already incorporated in the usual GCMs. In this survey special attention is paid to the relevance for the climate in North-west Europe. The project is conducted within the TNO-organization (project leader" dr. J. v a n Ham, SCMO-TNO), and covers less t h a n one m a n year research in a period of about 20 months in 1993 and 1994 (Van Ham, 1994). S t a t e - o f - t h e - a r t Coupled General Circulation Models (CGCMs) describe the behaviour of the atmosphere and the oceans as well as the interaction between both. The r e s u l t s r e p r e s e n t several aspects of p r e s e n t and p a s t climate r e a s o n a b l y well, a l t h o u g h s y s t e m a t i c deviations from observations exist. Consequently errors occur in predicted distribution patterns of e.g. t e m p e r a t u r e , wind and precipitation. The spatial resolution is relatively low, so the performance in the prediction of regional climatic features is rather low too. The addition of extra feed-back mechanisms to such models makes only sense as far as significant changes will occur within the context of CGCM outcomes t h a t exhibit various degrees of accuracy and resolution. The project has produced a list of mechanisms t h a t may provide feedbacks with a risk for non-monotonous or discontinuous behaviour of the climate system. Included are the following: * Snow-ice albedo feedback, which m a y lead to a positive feedback on a timescale of a few centuries, * Cloud feedback is an issue not well covered in GCMs due to the complex role of clouds in the radiative balance. Nevertheless the sensitivity of GCMs for cloud feedback is known to be relatively high. Satellite m e a s u r e m e n t s have revealed t h a t at present the globally averaged net cloud forcing is negative. The interactions between aerosol and cloud feedbacks and their coupling with other feedback mechanisms is only partially known and will require an integrated approach that is needed before these mechanisms can be incorporated into or coupled with GCMs.
1355 *
*
* *
*
Biogeochemical feedbacks form an important but complex issue. With regard to phosphorous compounds it turns out t h a t the increased availability of p h o s p h a t e in surface and coastal w a t e r s m a y promote the p r i m a r y production. In the carbon budget opposing trends are present, although most of them tend to provide a positive feedback to the CO2 accumulation in the atmosphere. The overall nitrogen budget seems to be neutral, although major uncertainties concerning e.g. nitrous and nitric oxide have been identified. The net result of changes in the sulphur budget are uncertain. The dominant trends here will be determined by the developments in the anthropogenic sources of carbon and nitrogen. Interaction between greenhouse and ozone effects can occur due to the fact t h a t an e n h a n c e d g r e e n h o u s e effect r e s u l t s in cooling of the lower stratosphere, thereby favouring conditions under which the ozone hole is formed in spring. Also, loss of ozone reinforces stratospheric cooling, so here a positive feedback loop is available. Combined effects of global change processes on forests constitute an additional stress factor for forests and natural vegetations of a yet unknown magnitude. Glaciological mechanisms could result in effects that become pronounced only at timescales of centuries. On a shorter timescale the formation of melting ponds in the Greenland ice-sheet may result in a positive feedback through albedo loss. An enhanced run-off could in principle affect the salinity in the North Atlantic ocean (see also next point). Modifications of ocean circulation patterns are a major concern as e.g. future climate changes in North Western Europe may be caused by a collapse of the Warm Gulf Stream. The so-called thermohaline circulation patterns t u r n out to be strongly influenced by changes in the influx of freshwater. The overall effect on the climate in North Western Europe may be moderate: a global warming of about 3oC could be compensated by a lowering of on the average 4oC in The Netherlands during winter days with advection from sea. Of course also other effects on weather conditions will occur (rainfall, wind patterns). The overall changes could occur on a timescale of centuries.
2.2 D i s c u s s i o n o f r e s u l t s The final reports of this project are not yet available. This hinders the final assessment. This is especially true for a major step in the project proposal: the confrontation of the project results with the opinion of experts via a dedicated workshop, to be held in 1994. The uncertainties associated with the various non-linear feedback mechanisms are high. A full a s s e s s m e n t of the risks (timing, magnitude, consequences) is impossible at the present state of knowledge. Following the results of this project, the most important issue concerns the interaction of greenhouse and ozone effects. In the following decades a gradual decline in ozone depleting substances is foreseen, while at the same time the concentration of greenhouse gases may reach levels t h a t are critical for the process of stratospheric cooling with its associated negative consequences.
1356
3.
SOCIO-ECONOMIC ASPECTS OF EXTREME EVENTS
3.1 S u m m a r y of findings The enhanced greenhouse effect is not only expected to raise average global t e m p e r a t u r e s , it may also affect the frequency and intensity of extreme w e a t h e r events, such as windstorms, tropical cyclones, heat or cold waves, and periods of extreme rainfall or drought. These events may have profound consequences, in terms of loss of life, material damages, and/or social-economic disruption. It has been suggested t h a t in m a n y parts of the world the negative consequences of global w a r m i n g associated to increases in extreme weather events will be more serious t h a t those resulting from a change in average temperatures. Thus, the following questions need to be answered: (1) Will the accumulation of greenhouse gasses result in changes in the incidence and severity of extreme weather events? (2) If so, w h a t will be the (direct and indirect) impacts of an increase in w e a t h e r related n a t u r a l disasters? (3) Which strategies can be developed by relevant actors (e.g., governments, insurance companies) to avoid climate change or to ameliorate its consequences? The aim of the project is to generate information required to a n s w e r such questions. The project is conducted at the I n s t i t u t e for Environm e n t a l Studies (IES) of the Free University in Amsterdam (NL) with prof.dr. P. Vellinga as the project leader. Parts of the project are carried out by the Environmental Change Unit (ECU) of the University of Oxford (UK). The total n u m b e r of m a n years spend in this project amounts to Section 3.2. The project consists of four main research activities: (1) organization of an introductory workshop; (2) development of weather scenarios; (3) a s s e s s m e n t of socio-economic impacts; and (4) assessment of possible response options. 1. In june 1993, a two-day workshop was organized, the main aim of which was to identify and demarcate the specific research questions and issues to be addressed in the remainder of project. A comprehensive list of research issues was composed, centring around five key topics: (1) w e a t h e r scenarios, (2) vulnerable areas, (3) exposure scenarios, (4) risk scenarios, and (5) policy scenarios (for details, see Olsthoorn and Tol, 1994). It was decided to focus on two combinations of w e a t h e r events and t a r g e t regions: "storms in NWeurope' and "cyclones in the SW-Pacific". 2. A weather scenario is: "a plausible future for weather patterns with reference to extreme weather events". Methodologies for developing weather scenarios are discussed in Downing et al. (1993). Weather scenarios may be based on General Circulation Models (GCM's) or they m a y be e x t r a p o l a t e d from observed w e a t h e r statistics. In the latter case, so-called stochastic w e a t h e r generators (SWG's) are used. Basically, SWG's are models t h a t g e n e r a t e w e a t h e r patterns with the same statistical features as the observed weather; in the case of climate change studies, SWG-models with "perturbed parameters" may be used. Because of their expertise in SWG's, ECU Oxford was asked to develop the weather scenarios. The work on w e a t h e r scenario development still continues. It is clear, however, t h a t for both the events studied (storms and cyclones) scenario development turned out to be more difficult than expected. One reason for this is that the resolution of current climate models is insufficient for predicting (regional) changes in extreme weather events. For instance, a review of recent GCM-studies (in Tol et al. 1994) reveals that the evidence with regard to the effect of CO2-doubling on the frequency and intensity of tropical cyclones is
1357
,
highly conflicting. While some models predict an increase in the destructive potential of cyclones of up to 50%, others predict a decrease of cyclones under 2 x CO2; yet other investigators conclude t h a t outcomes strongly depend on the model p a r a m e t e r s chosen. Analysis of historical d a t a on cyclone prevalence in the SW-Pacific also failed to result in clear conclusions about existing trends, due to limitations in the available data sets. Tol et al. (1994) conclude that: "neither theory nor observations are conclusive on how the incidence and intensity of tropical cyclones will change under a doubling of CO2". Consequently, scenario construction is to a large extent based on "educated guesswork". With respect to "windstorms in europe", a similar conclusion applies (Olsthoorn and Tol, 1994). Assessment of socio-economic impacts and assessment of relevant response options both belong to the second phase of the project. Since the reports available to date all refer to the first phase, information on the nature and content of these activities is limited. With regard to point (4), the results of a preliminary analysis of response options, available to (re)insurance companies, is presented in Olsthoorn and Tol (1994). Most of the response options available to insurance companies merely aim to restrict the potential financial consequences for the company (e.g. raise premiums or deductibles, hedge risks t h r o u g h r e i n s u r a n c e , or exclude certain risks or areas). Some response options, however, are more f u n d a m e n t a l in the sense t h a t they aim to decrease the amount of damage resulting from climate change (by promoting reduced exposure through, e.g., education, premium incentives or contract conditions), or to prevent climate change itself (e.g. by investing in energy saving measures or by lobbying for climate change prevention).
3.2 D i s c u s s i o n of r e s u l t s The list of research issues, which resulted from the introductory workshop, makes clear that answering the three questions indicated in the introduction will require a v a s t r e s e a r c h effort. In view of this, the project team's decision to focus on methodological aspects and two case studies (storms in Europe, cyclones in SW/Asia) appears to be a wise one. The currently available project output addresses the first part of the project and, as a consequence, mainly pertains to meteorological issues (i.c. weather scenarios). This is a s o m e w h a t u n f o r t u n a t e situation, for two reasons. The first is that, although the work on w e a t h e r scenarios has not been completed, it is clear t h a t the outcome will not be entirely satisfying. The current state-of-the-art in climate (change) modelling does not allow u n a m b i g u o u s conclusions with respect to regional changes in extreme w e a t h e r events. And at least for one of the events studied (cyclones), scenarios can neither be reliably extrapolated from existing data sets (which, in our view, would be tricky anyhow since global w a r m i n g m a y change existing t r e n d s or may have non-linear effects, see section 2). As a consequence, the weather scenarios to be used in the remainder of the project are chosen r a t h e r arbitrarily; they are not firmly rooted in meteorological theory or analyses. In view of the lack of relevant knowledge and data, this seems to be the only viable course of action, but it may cause difficulties later on in the course of the project; for instance, choosing between the response options (to be) identified in step 4 will probably require information on the conditional likelihoods of weather scenarios (conditional on emission scenarios). W h e t h e r such estimates can be generated in a reliable way seems doubtful.
1358 That the existing project output mainly deals with meteorological issues "tends to conceal the mainly socio-economic nature of the study" (Olsthoorn and Tol, 1994). We agree with these authors that the quintessence of the project is to develop a methodology for assessing and quantifying the social and economical impacts of extreme weather events. The emphasis on meteorological issues is probably temporary and due to the premature timing of the assessment (see section 1). Nevertheless, the fact that the current output hardly addresses the impact assessment methodology precludes an evaluation of the project as a whole. Despite the difficulties encountered with regard to weather scenario development, the project may yield a valuable contribution to NRP-program in general and to decision-oriented approaches in particular (see e.g. section 4), if the project team succeeds in developing an appropriate methodology for impact assessment and quantification. Whether this will be the case cannot be decided on the basis of the currently available material.
4.
CHARACTERIZING THE RISKS
4.1 S u m m a r y of findings The purpose of this project is to examine whether or not a decision-analytic approach yields an effective and integrated assessment methodology for ordering and combining the research results about global warming. The focus of the project is on representing the global warming as a decision problem for the Dutch (public) policy makers. The research is conducted at the Center for Energy and Environmental Studies of the University of Groningen, with dr. L. Hendrickx acting as the project leader. The projects has started in September 1993, will continue until mid 1995 and entails 3 man years of research. Research presently conducted under the NRP program is primarily problemcentred (or 'diagnostic'), i.e., it is aiming at identifying and quantifying the risks associated to the (anthropogenic) greenhouse effect itself. Some therapeutic), i.e., they aim at the identification of possible and relevant policy measures and at the assessment of their societal impacts. The present project aims at the development of a methodology that relates and combines conclusions from 'diagnostic' oriented research with those from 'therapeutic' oriented research. In other words, this project is directed at the development of a methodology for ordering the various (geophysical, ecological, technical, socio-economical and political) elements of the global warming problem in a coherent and meaningful decision analysis. Integrated assessment has to do with surveying and ordering a particular state of knowledge about a particular issue. However, it is a misconception to consider integrated assessment to be simply a matter of collecting and ordering all available knowledge in all its details. The challenge of integrated assessment is primarily one of information management and the development of representations at an adequate high level of aggregation. The issue is to make the analysis as simple as possible but no simpler (Morgan and Henrion, 1990). It is useful to distinguish integrated assessment in a pure scientific context from that in a policy context. In the latter case the additional - and possibly even more important - challenge exists in casting the problem from the perspective of policy makers. Decision-analytical approaches might fit the purpose of such integrated assessment as they explicitly incorporate an analysis of the uncertainties involved. Uncertainties are attached
1359 to almost every step in the global warming causal chain (driving forces - h u m a n activities - greenhouse gases and aerosols - enhanced greenhouse effect t e m p e r a t u r e increase - secondary climate changes - socio-economic impacts). Much global climate research is aimed at the reduction or even elimination of these uncertainties, although the relevance for public policy is not always clear. Insights from decision theory might facilitate the assessment problem at hand as a policy oriented decision-analytic tool could (a) result in a t r a n s p a r e n t and intelligible synoptic view which spans all significant aspects of the global warming problem, (b) assist policy making by representing the problem such that it links up with the decision perspective of policy makers, and (c) explicitly include uncertainty in the analysis for, amongst other things, identifying critical knowledge gaps or performing sensitivity analyses. The DEMOS model (DEcision MOdeling System) is the product of over a decade of research on uncertainty analysis and on tools for integrated assessment at the Department of Engineering and Public Policy of Carnegie Mellon University, USA. This model is currently in use for their Global Changes Integrated Assessment Program which aims at integrating the many disparate pieces of science relevant to the global warming problem. DEMOS is a software environment for creating, analyzing and communicating complex models involving uncertainties for risk and policy analysis. These models are described through graphical influence diagrams, much like the tools used in system dynamics. Large models can conveniently be organized into a hierarchy of more manageable submodels, each with its own influence diagram. DEMOS contains built-in sensitivity and uncertainty tools for the explicit t r e a t m e n t and propagation of uncertainty. It is flexible enough to include traditional approaches like subjective expected utility (SEU) analyses and multi-attribute utility (MAUT) analyses. This model has been made available for use in the current IVEM project, and functions as the primary tool for the characterization of the global warming risks. The present state of affairs in this project is that an initial conceptualization of the global warming problem has been constructed and is being filled with quantitative information (data, values, relationships). The current representation involves a first and provisional conceptualization that should be adjusted in a subsequent iterative process of modification and refinement in the research period until mid 1995. The results of the first phase of the NRP program can thus be used in this process. The IPCC scenarios are used for the representation of possible different economic, social and environmental developments. For the formulation of Dutch policy options the results of the NRP project "Development of policy options for dealing with the greenhouse effect on sustainable development" have been used. Greenhouse damage appears to be one out of five socio-economic impact categories. The other categories are: a b a t e m e n t costs, a d a p t a t i o n costs, international relations and first mover. The greenhouse damage category could be subdivided using the Cline categorization as a point of departure. The end result will be a (multi-attribute) tree of impacts with the five impact categories mentioned before as the main branches. The impacts at the end of branches should provide operational measures for evaluating and comparing Dutch policy options. Within this context it is an option to relate the set of expected Dutch socio-
1360 economic impacts to the OECD list of socio-economic indicators as proxies for welfare indicators. 4.2 D i s c u s s i o n of r e s u l t s A comparative analysis of the risks of global warming and of relevant policy strategies concerning that problem is central to the current NRP program. One of the approaches chosen concerns decision analysis. This could provide a hierarchical representation of the policy problem at a high level of aggregation, thereby avoiding irrelevant reliance on complex models and massive handling of information by actors not fit for that purpose. The present project capitalizes on the insights gained in this field of research by the cooperation with Carnegie Mellon University. The use of the DEMOS model greatly facilitates the development of the required methodology. At present a first representation of the global warming problem in terms of hierarchical influence diagrams is being finalized. Results of the first phase of the NRP program should be made available in order to refine and detail that version. Tuning with the current research on the (meta) IMAGE model and with projects t h a t generate policy options in cooperation with public policy officials seems necessary for the next phase in this research project.
5.
GENERAL DISCUSSION
None of the projects discussed in this assessment report has entered their last stage - hat of producing the final documents. That hinders a full assessment. The overview of extreme weather events project concerns a study of only moderate manpower, and has produced a list of possible categories of extreme weather events that may pose serious and non-linear risks to society and nature. The final assessment of these results will be produced after a workshop with various experts that will be conducted in 1994. The assessment of the socio-economic aspects of extreme weather events project has thus far mainly addressed meteorological issues. The results cast doubts upon the feasibility of generating reliable extreme weather scenarios. This requires the rest of the project to focus and concentrate on the development of a methodology for the assessment and quantification of the social and economical impacts of extreme weather events. The project concerning the characterization of the risks of global warming and of relevant policy strategies has thus far resulted in the selection of an appropriate p r o g r a m m i n g environment, and in the preliminary formulation of a set of hierarchical influence diagrams describing the global problem from a Dutch policy perspective. This implies that the first steps are made in the formulation of a suitable methodology, that may be of great importance for other lines of research (to be) pursued under the NRP programme.
1361 6.
REFERENCES
Downing, T.E., Lohman, E.J.A., Maunder, W.J., Oltshoorn, A.A., Tol, R.S.J., Tiedemann, H., 1993. Socio-economic and policy aspects of changes in incidence and intensity of extreme (weather) events. Position paper of the Amsterdam Workshop, June 1993. Institute for Environmental Studies, Free University, Amsterdam. Morgan, M. and Henrion, M., 1990. Uncertainty: a guide to dealing with uncertainty in quantitative risk and policy analysis. University Press, Cambridge,. Oltshoorn, X.A. and Tol, R.S.J., 1994. Socio-economic and policy aspects of chanes in incidence and intensity of extreme (weather) events. Report of the first phase. Institute for Environmental Studies, Free University, Amsterdam. Tol, R.S.J., Dorland, C. and Oltshoorn, X.A., 1994. Tropical cyclones in the south-west pacific; past and future occurance and intensity. Institute for Environmental Studies, Free University, Amsterdam. Van Ham, J., 1994. Risks of non-linear climate change, summary of project findings. TNO-Milieuwetenschappen, Delia. Van Lenthe, J., Hendrickx, L. and Vlek, C.A.J., 1994. Integrated assessment of the global warming problem: Adecision-analytical appraoch. Interim report. November 1994, Center for Energy and Environmental Studies (IVEM), University of Groningen.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1363
D I S C U S S I O N ON NRP'S A S S E S S M E N T R E P O R T ON RISK ANALYSIS Chair: W. Biesiot Rapporteur: J. van Lenthe The chairman starts with providing an overview of the backgrounds and the major objectives of the "Risk Analysis" subtheme of the Dutch NRP as described in the programming memorandum Integration of the Dutch NRP. He explains t h a t in the current meeting project leaders will get a 10-15 minute opportunity to describe main results. These project presentations will be followed by a discussion which should focus on the question whether the research projects have contributed to the realization of the goal of the risk analysis subtheme: ... a systematic and integrated analysis of the risks of climate change at (inter)national levels is necessary for the evaluation and support of environmental quality standards and the assessment of policy options.
The three research projects carried out under the risk analysis subtheme are (1) the risks of non-linear climate changes, (2) socio-economic aspects of extreme events, and (3) characterizing the risks: a comparative analysis of the risks of global w a r m i n g and of relevant policy strategies. For information concerning the content of the projects, see the separate project papers and the A s s e s s m e n t report concerning N R P funded risk research projects. Some comments during the discussion were related to problems having their origins in the relatively late start of the research projects (summer and fall 1993). Two of the projects (number 2 and 3) are scheduled to continue until mid 1995. One of the members at the meeting pointed out that it might still be a little bit early for risk analysis. After all, we know very little about the probabilities and the m a g n i t u d e of the greenhouse damages. This might also be the reason t h a t risk analysis and risk management are further developed in areas other than the global warming area. Some project-related comments: - The project on the risk of non-linear climate changes revealed t h a t the state-oft h e - a r t knowledge of the climate system does not allow a full q u a n t i t a t i v e assessment of the risks. There still appear to be large uncertainties concerning timing, magnitude, and consequences of the non-linear climate changes. - The title of the second project suggests more t h a n is covered. The focus appears to be on (re)assurance aspects of storms and cyclones and there seems to be relatively little attention to socio-economic impacts. One speaker expressed his doubt on the apparently assumed relation between t e m p e r a t u r e increase and extreme weather events. - The decision-analytical modelling (Demos influence-diagram modelling) proposed in the third p r o j e c t - c h a r a c t e r i z i n g the risks- was considered a promising approach for comparing the risk of global warming with the risks associated to
1364 mitigative policy options. An important recommendation was to increase the interaction with other projects, notably, projects on the IMAGE model and projects on generating policy options. The conclusion with regard to the main goal was that the three research projects cover only part of the requirements from the programming memorandum. A full systematic and integrated assessment of all the risks is not yet realized. However, the meeting agrees about the relevance of risk analysis for handling the global warming problem. A final remark stated that current and future risk-analysis research projects might benefit from increased mutual interactions and from interactions with other NRP projects.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1367
Integrated assessment of the global warming problem A decision-analytical approach J. van Lenthe a, L. Hendrickx a, and C.A.J. Vlek b aCenter for Energy and Environmental Studies, University of Groningen, Nijenborgh 4, 9747 AG Groningen, The Netherlands bDepartment of Psychology, University of Groningen, Grote Kruisstraat 2/1, 9712 TS Groningen, The Netherlands
Abstract The current project is aimed at developing a policy-oriented methodology for the integrated assessment of the global warming problem. Decision analysis in general and influence diagrams in particular appear to constitute an appropriate integrated assessment methodology. The influence-diagram approach is illustrated at a preliminary integrated modeling of the global warming problem. In next stages of the research, attention will be shifted from the methodology of integrated assessment to the contents of integrated models.
1. AN INTEGRATED ASSESSMENT METHODOLOGY The Netherlands initiated a National Research Program (NRP) on Global Air Pollution and Climate Change. An important objective of the Dutch NRP is to produce policy-relevant information. The first phase (NRP-I) identified "integration" as one of five central research themes. The multi-disciplinary character of the global warming problem asks for an integrated assessment methodology for ordering and combining the various physical, ecological, economical, and sociological results. As a result, the second phase (NRP-II) shows a growing concern for integrated assessment issues. The current two-year research project "Characterizing the risks: a comparative analysis of the risks of global wanning and of relevant policy strategies", which started in September 1993, comes under the integrated assessment part of the Dutch NRP. Its main objective is to examine whether a decision-analytical approach yields an effective integrated assessment methodology. From a policy angle, the global warming problem boils down to a choice between policy options that might bring on different (types of) risks. Therefore, the project concentrates on modeling the global warming problem as decision problem from a Dutch policy point of view. For an interim report describing the first year results, see Van Lenthe, Hendrickx, and Vlek (1994). 1.1 Decision analysis - - W h y ? Decision analysis has been developed to address problems w h i c h - just like the global warming problem - - are characterized by a set of response options, unceaain events, and
1368 possible consequences (e.g., Clemen, 1990; Von Winterfeldt, & Edwards, 1986). There are three central reasons why a decision-analytical approach might fit the purpose of a policyoriented integrated assessment methodology. A first reason to opt for a decision-analytical approach is that integrated assessment is more than laying out scientific facts. It is also a question of information management and the development of problem representations at an adequately high level of aggregation. Decision analysis includes appropriate tools (a) for ordering and combining data and (b) for evaluating and analyzing the resulting integrated models. A second reason to consider a decisionanalytical approach is that decision analysis is adequate method for framing the global warming problem from a policy point. From a policy angle, the problem is essentially a decision problem in which the ultimate decision relates to the choice between policy options. A third reason for choosing a decision-analytical approach is its explicit incorporation and analysis of uncertainty. The global warming problem is a complex problem in which uncertainties play a critical role. Sensitivity and uncertainty analysis, which can be used to identify important knowledge gaps, belong to the standard equipment of decision analysis. 1.2 Influence diagrams m Why?
The decision analysis toolbox includes ~ Radiative various techniques for structuring and analyzing forcingR complex problems. In the current project, it was decided to use influence diagrams which are a relatively recent development. Influence Warming diagrams allow a compact graphical represenperunitR tation of complex problems. Different problem elements show up as different nodes which are linked with arrows to show mutual influences. Feedback multiplier The example of Figure 1 involves the Temperature increase submodel from the preliminary modeling in Van Lenthe, et al. (1994). There Figure 1 Influence diagram example. are three important motives for choosing an influence-diagram approach. A first advantage is that influence diagrams can be considered at different levels of specification. At a low level of specification m e.g., postulating relationships without specifying their numerical form m the influence diagrams are intuitive and compact enough for communicating the general structure of complex problems in an intelligible way. At a high level of specification, the same influence diagrams may contain a wealth of information and data for quantitative analysis. In this way, influence diagrams bridge the gap between qualitative formulation and quantitative modeling. A second advantage is that influence diagrams are not only suitable for decision modeling, but for any formal description of a set of relationships. Thus, the same modeling language can used for policy-oriented representations as well as scientific-oriented representations which might be quite different things. Scientists are usually interested in fundamental processes, whereas policy makers usually find more benefit in 'broad outline' information presented in a way that is effective for policy development. However, major decisions can not without information about the environment and policy-decision models should therefore include (high-aggregation-level versions of) scientific representations. A third motive for choosing a influence diagram approach is that we can build upon the research on integrated assessment and uncertainty analysis at the Department of Engineering
1369 and Public Policy of Carnegie Mellon University (Morgan, & Henrion, 1990). On of their products is Demos (Decision modeling system), a software environment for creating, analyzing, and communicating complex hierarchical influence diagram models for risk and policy analysis. Researchers at Carnegy Mellon are currently using Demos in the Global Climate Change Integrated Assessment Program (Dowlatabadi, & Morgan, 1993) which proved to be an indispensable source of inspiration for the current project.
~riving forces) ~
1
Annual "N enhouse gas }
missionsj /
1
Atmospheric~'~ ncentratio~) 2. MODELING THE GLOBAL WARMING PROBLEM
1
In its first year, the project forcing concentrated on illustrating the scenarios decision-analytical approach. 1 The modeling has resulted in a ~Temperature~ increase j rough outline of the problem usalchainJ structure in terms of hierarchical I influence diagrams. At some Dutch places, the qualitative structure reenhouse ] # Dutch ~ ' ~ ~ . Dutch .~ damages ~,~ has been filled with quantitative li~y9o"ti-ns ]--~soc~o-econom=cl . u,~/ ~ impacts data. However, both structure and contents should be conFigure 2 Top-level Figure 3 Top-level influence sidered as provisional. influence diagram diagram of the policy decision The modeling discriminates of the basic causal representation. between (a) a basic causal chain chain. from a scientific point of view and (b) a decision representation from a policy point of view. Figure 2 shows the top-level influence diagram for the basic causal chain. Each node represents a submodel that contains its own influence diagram (cf. Figure 1). Figure 3 shows the top-level influence diagram for the policy decision representation. The models should be considered as complementary rather than separate. The policy decision model is the most comprehensive representation as it includes a high-level version of the scientific basic causal chain. Global emission Dutch socioscentufos economic impacts Figure 3 and 4 illustrate that the G1 G2 G3 G4 global warming problem involves Dutch Greenhouse damages considerations at both a global and policy a national level. Dutch greenhouse Abatement costs options damages are only one category of AdOrationcosts socio-economic impacts. Other Internationalrelations categories include: abatement costs, First mover adaptation costs, international relations, and first mover effects (see Figure 4). The ultimate goal is to Figure 4 Each combination of Dutch policy options evaluate and compare Dutch policy and Global emission scenarios results in a particular options on these socio-economic set of Dutch socio-economic impacts.
Global ) emission 9
.
1
1
1370 impact categories. The dual character (global as well as national) of the enhanced greenhouse problem is a complicating factor for comparing the risks of global warming with the risks associated to mitigative policy options. For example, whereas Dutch policy efforts might bring on substantive abatement costs, they will usually have only a small direct effect on reducing global warming and thus on reducing Dutch greenhouse damages. The benefits of Dutch policy efforts should be mainly located in other socio-economic impacts (e.g., international relations and first mover effects).
3. MAIN CONCLUSION AND FUTURE RESEARCH The current search for an integrated assessment methodology appeared to be faidy successful. Decision analysis in general and influence diagrams in particular constitute an appropriate policy-oriented approach for the integrated assessment of the global warming problem. In its first year, the project concentrated (a) on identifying an appropriate decisionanalytical approach and (b) on illustrating the selected approach at a first and provisional integrated modeling. In the second year, attention will be shifted from the methodology of integrated assessment to the resulting integrated models. Both their structure and their contents will be central issues for further research. Emphasis will be on the policy decision representation and its link with the basic causal chain. Finally, the integrated models will be subjected to evaluations in which sensitivity and uncertainty analyses play a central role. The current two-year project is expected to result in an appropriate integrated assessment methodology and an integrated modeling of the global wanning problem from a Dutch policy point of view. The influence-diagram approach is anticipated to (a) contribute to a transparent and intelligible synoptic view that spans all significant elements of the global warming problem, (b) to facilitate policy development by representing the global warming problem as a policy decision problem, and (c) to support the identification of critical knowledge gaps through uncertainty and sensitivity analyses. The influence diagram approach and the integrated models are useful points of departure for future research in which they should put through an iterative process of modification and refinement. Data that will come available at the end of NRP-I and during NRP-II could be assimilated in the iterative integrated modeling.
4. REFERENCES
Clemen, R.T. (1990) Making hard decisions. An introduction to decision analysis. Boston: PWS-Kent publishing company. Dowlatabadi, H., & Morgan, M.G. (1993) A model framework for integrated studies of the climate problem, Energy Policy, 21(3), 209-221. Morgan, M.G., & Henrion, M. (1990) Uncertainty - A guide to deal with uncertainty in quantitative risk and policy analysis. Cambridge: Cambridge university press. Van Lenthe. J., Hendrickx, L., & Vlek, C.A.J. (1994) Integrated assessment of the global warming problem ~ A decision-analytical approach. Center for energy and environmental studies (IVEM), University of Groningen. Von Winterfeldt, D., & Edwards, W. (1986). Decision analysis and behavioral research. Cambridge: Cambridge university press.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1371
Clouds, aerosols and biogeochemical cycles" risks of non-linear climate change J. van Ham a, R.J. van Beers a, P.J.H. Builtjes a, G.P. KSnnen b, j. Oerlemans c, M.G.M. Roemer a a
TNO Institute of Environmental Sciences, P.O. Box 6011, NL-2600 JA Delft
KNMI, Royal Netherlands Meteorological Institute, P.O. Box 201, NL-3730 AE De Bilt
b
c Institute for Meteorological and Atmospheric Research, University of Utrecht, P.O. Box 80005, NL-3508 TA Utrecht
Abstract In this paper part of an investigation is described into risks for climate change which are presently not adequately covered in General Circulation Models. The investigation included the interaction with biogeochemical cycles, the effects of clouds and aerosols, ice flow instability, albedo instability and modified ocean circulation. In this paper our results for clouds and aerosols and for biogeochemical cycles are reported.
1. I N T R O D U C T I O N In the concept of climate change as a result of the enhanced greenhouse effect it is generally assumed that the radiative forcings from increased concentrations of greenhouse gases (GHG) will result in a proportional or quasi-linear global warming. Though correlations of this kind are known from palaeoclimate research, the variability of the climate seems to prevent the direct proof of a causal relation between recent greenhouse gas concentrations and temperatures observations. In order to resolve the issue the use of General Circulation Models (GCMs), though still inadequate at present, is indispensable. Around the world some 10 leading GCMs exist which have been the subject of evaluation and intercomparison in a number of studies (AMS, 1994; Cess et al, 1993; Boer et al, 1992; Gates et al, 1992; Randall et al, 1992; Cess et al, 1990; Cubasch and Cess, 1990). GCMs represent several aspects of present climate reasonably well, but systematic deviation of observations exist for upper troposphere temperatures and tropical lower stratosphere temperatures. Consequent errors in zonal wind distribution produce differences in precipitation patterns. Also, the reliability of simulated polar climates remains low. The use of GCMs is
1372 further limited because of their course spatial resolution; this leads to a reduced performance in the prediction of regional climates. A discussion on the causes of their weak points in simulating present and past climates shows t h a t the depiction of clouds is a major w e a k n e s s of GCMs. Uncertainties extend to cloud formation, in particular as a function of altitude, and to cloud properties, while the sensitivity of the models for these variables is high. A second element which is virtually absent in GCMs are the feedbacks from natural biogeochemical cycles. All these cycles are influenced by man in a n u m b e r of ways. Apparently, not all relevant processes have been included in the GCMs. T h a t situation constitutes a risk, since it cannot be ruled out t h a t a missing process could cause or trigger a non-linear climate change. In addition, if global t e m p e r a t u r e s start to rise, we do not know too well which responses we will see. So we might be surprised by non-linear changes. We report here the result of an investigation aiming to identify processes which potentially could provide a risk for such a non-linear change.
2.
METHODOLOGY
In order to collect as m a n y entries as possible with a relevance for the topic of interest the investigation was directed to all major areas of research with respect to climate and climate change. The state-of-the-art of the General Circulation Models stands at the base of our investigation. The following areas which either are inadequately covered in GCMs or are outside the focus of the GCM-based climate research have been selected for a separate screening on potential nonlinear effects: 9 global change and biogeochemical cycles 9 clouds and aerosols 9 ice flow instability 9 albedo instability 9 ocean circulation patterns In the present paper we will, s t a r t i n g from the body of knowledge which is absorbed within GCMs, communicate our estimates for non-linearities in climate change resulting from the effects of clouds and aerosols, and the effects of global change processes on biogeochemical cycles. Our estimates on albedo and ice-flow instability and on modified ocean circulation will be published elsewhere (Van H a m et al, 1995a). Full details of this work will be in our final report for the National Research Programme on Global Air Pollution and Climate Change (Van H a m et al, 1995b).
3.
C L O U D S AND A E R O S O L S
The general u n c e r t a i n t y on the m a g n i t u d e and even the sign of some of the different feedback mechanisms through cloud formation, and its interaction with aerosols constitutes a risk for major positive feedback. Clouds exert an overall negative forcing ( R a m a n a t h a n et al, 1989; Harrison, et al, 1990; A r d a n u y et al,
1373 1991), but the net effect varies with altitude and cloud type: clouds in the upper troposphere (cirrus-type) tend to a net positive forcing while lower clouds (cumulus, stratus and stratocumulus) exert a negative forcing. The uncertainty in the balance between low, medium and high clouds is reflected in an uncertainty in the trend of the overall effect of GHG increase on global mean temperatures. Most GCMs predict a decrease in the fraction cloud a m o u n t (average 4_+2.5% decrease). Without further specification of cloud types, however, it is not clear w h e t h e r this decrease acts as a positive or negative feedback. Since cloud formation could take a path different from the one which is outlined in most GCMs the situation is even more uncertain. Instead of the generally predicted global w a r m i n g between 1.5 and 4.5 ~ with a central value of 2.5 ~ for a doubling of the CO2-concentration (IPCC, 1990; 1992; 1994), the greenhouse gas forcing could also express itself as an increase in average cloud coverage, the effects of which (food production, hydrological aspects, recreation, etc) have not been very well quantified. In this respect it is crucial to monitor any trends in cloud cover worldwide. U n f o r t u n a t e l y , we do not have a long h i s t o r y of reliable cloud coverage observations (London, et al, 1991; Warren, et al, 1986; 1988). In the period 19521981 observations from ships have shown that over the latitudes 20~176 of the tropical oceans cirrus and cumulonimbus cloud t y p e s have increased while cumulus and stratus types decreased or remained nearly constant. The net effect in this area has been estimated to come down to an increase in greenhouse forcing. Reliable databases of cloud observations over land includes data which date back as far as 1971. The significance of these analysis is still of minor value, due to the relatively short period and the spatial limitation of the study. Due to the uncertainties in this area it is not possible to quantify a risk.
4.
B I O G E O C H E M I C A L CYCLES AND G L O B A L C H A N G E
Apart from the enhanced greenhouse effect several more processes of global change are now apparent: stratospheric ozone depletion, land use change, deforestation (causing in some areas erosion and desertification), acidification and changes in water resource management. All these processes are connected in several ways to the n a t u r a l biogeochemical cycles: m a n k i n d is increasing the turnover of these cycles through emissions to air, surface water, soils and the m a r i n e environment; in addition, global change processes have a p p a r e n t consequences for the n a t u r a l sources and sinks of most cycles. Some of these interactions constitute additional risks, which are outlined below.
4.1. G r e e n h o u s e i n d u c e d s t r a t o s p h e r i c o z o n e d e p l e t i o n Stratospheric cooling, due to the enhanced greenhouse effect, could result in conditions which favour the formation of an Arctic ozone hole (Austin et al, 1992). The report of Austin et al suggests that a doubling of CO2 could already have a substantial effect. The probability depends on the relative timing of the growth rate of total greenhouse gas concentrations and the atmospheric loss rate of ozone depleting substances (ODS). The risk could be enhanced by volcanic eruptions,
1374 which f u r t h e r the heterogeneous chemical reactions leading to ozone loss. As a result of the phase-out of ODS the period of highest risk probably falls within the forthcoming 3 to 5 decades. The effect would be an increase in UV-B levels in the n o r t h e r n h e m i s p h e r e and could also cause a change in the s t r a t o s p h e r i c circulation patterns. The exact significance of such a change is unclear at present.
4.2. F e e d b a c k s t h r o u g h modification of natural e m i s s i o n s In table 1 the impact of global change processes has been e s t i m a t e d for major source strengths of n a t u r a l emissions in a n u m b e r of cycles.
Table 1. S u m m a r y of feedbacks t h r o u g h biogeochemical cycles (biogenic emissions only); the scores are based on the expected effect on the atmospheric concentration of each component. Aspect of global change
CO2
CH4 NMHC
CO
DMS
CH3C1 CH3Br
N20
NO
Global w a r m i n g 9 SST 9 Air
+ 0
+ -
+ -
+ -
+ 0/-
+ 0/-
? ?
0 -
Soil h u m i d i t y reduction
+
0
0
0
0
0
0
0
UV-B increase (OH; p r i m a r y production)
+
.
Acidification
+
0
0/-
-
0
0
0/+
0
0
0/+
0/+
+
+
+
+
Eutrophication Fertilization
.
.
.
.
0
Land use change and W a t e r resource management
+
-
-
-
0
0
-
0
Overall effect
+
0/-
0/-
0/-
+/-
+/-
0
0
+ positive feedback: the indicated aspect of global change results in increased emission of the respective GHG - negative feedback 0 neutral: no effect expected or two counteracting effects ? no estimate made
1375 It is seen that most of the above-mentioned processes of global change result in additional CO2-emission or reduce the global CO2-absorption capacity, thus providing a positive feedback to radiative forcing. The supposed positive feedback of methane through the melting of permafrost in a warmer climate (Mellilo, J.M., 1990) is expected to be partly compensated for by gradual drying and afforestation of former tundra areas (Ihle, 1993). The accumulated effect of stress on forest ecosystems, in combination with deforestation, could reduce the total forest area in the world as well as forest vitality. The resulting decrease of emissions of volatile organic components (VOC) might reduce tropospheric ozone formation worldwide, thus providing a negative feedback for global warming. It should be remembered, however, that the effect of a decrease of VOC-emissions from forests could be counterbalanced by an increase in man-made VOC-emissions.
5. C O N C L U S I O N S From the body of information in the respective literature it can be derived that clouds are the major uncertainty with respect to future climate and global warming: clouds could counteract global warming, but ~ should be realized that in doing that a more cloudy world results. Present global change processes nearly all tend to contribute to either increased emissions of CO2 or decreases in CO2-sinks. They do not seem to provide major increases in the source strengths of other greenhouse gases. There is a certain risk for loss of stratospheric ozone as a result of the enhanced greenhouse effect.
6.
REFERENCES
AMS, Preprints of "Sixht Conference on Climate variations, J a n u a r y 23-28, 1994. Nashville, Tennessee. Amer. Met. Society, Boston, Ma, USA 1994 Ardanuy, P.E., L.L. Stowe, A. Gruber, and M. Weiss, Shortwave, longwave and net cloud-radiative forcing as determined from Nimbus 7 observations. J. Geophys. Res., 96, 18537-18549 (1991) Austin, J., N. Butchart and K.P. Shine, Possibility of an Arctic ozone hole in a doubled-CO2 climate. Nature 360, 221 (1992) Boer, G.J. et al, Some results from an intercomparison of the climates simulated by 14 atmospheric General Circulation Models. J. Geophys. Res. 93, 8305-8314 (1992) Cess, R.D. et al, Uncertainties in carbon dioxide radiative forcing in atmospheric General Circulation Models. Science 253, 1252-1255 (1993) Cess, R.D. et al, Intercomparison and interpretation of climate feedback processes in 19 atmospheric General Circulation Models. J. Geophys. Res., 95, 1660116615 (1990) Cubasch, U. and R.D. Cess, Processes and modelling, in: Climate Change, the scientific assessment, J.T. Houghton, G.J. Jenkins and J.J. Ephraums (eds), UNEP/WMO. Cambridge University Press, Cambridge, UK 1990
1376 Gates, W.L., J.F.B. Mitchell, G.J. Boer, U. Cubasch and V.P. Meleshko, Climate modelling, climate prediction and model validation, in: Climate change 1992, The supplementory report to the IPCC Scientific Assessment, J.T. Houghton, B.A. Callander and S.K. Varney (eds.), UNEP/WMO. Cambridge University Press, Cambridge UK, 1992 Ham, J. van, R.J. van Beers, P.J.H. Builtjes, G.P. KSnnen, J. Oerlemans and M.G.M. Roemer (1995a), Albedo and Ice-flow instability and modified ocean circulation: risks of non-linear climate change. Proc. of the 10th World Clean Air Congress, Espoo, Finland, 28 May-2 June 1995. Ham, J. van, R.J. van Beers, P.J.H. Builtjes, G.P. KSnnen, J. Oerlemans and M.G.M. Roemer (1995b), Risks of non-linear climate change. Report TNO MW 95/003, 1995 Harrison, E.F., P. Minnis, B.R. Barkstrom, V. Ramanathan, R.D. Cess and G.G. Gibson, Seasonal variation of cloud radiative forcing derived from the Earth Radiation Budget Experiment. J. Geophys. Res., 95, 18687-18703 (1990) Ihle, F., Methaanflux uit toendra's in een veranderd klimaat (Methane flux from tundras in a changed climate). TNO-report IMW- R93/193 (1993) IPCC, Climate Change, the Scientific Assessment. WMO/UNEP, Cambridge University Press, Cambridge 1990 IPCC, Climate Change 1992, the supplementary report to the IPCC Scientific Assessment. WMO/UNEP, Cambridge University Press, Cambridge 1992 IPCC, Radiative forcing of climate change, the 1994 report of the Scientific Assessment Working Group of IPCC; Summary for policymakers. WMO~SNEP, 1994 London, J., S.G. Warren and C.J. Hahn, Thirty-year trend of observed greenhouse clouds over the tropical oceans. Adv. Space Res. 11 (3), 345-349 (1991) Mellilo, J.M., T.V. Callaghan, F.I. Woodward, E. Salati and S.K. Sinha, Effects on ecosystems. In: IPCC, 1990 Ramanathan, V., R.D. Cess, E.F. Harrison, P. Minnis, B.R. Barkstrom, E. Ahmad and D. Hartmann, Cloud-radiative forcing and climate: results from the Earth Radiative Budget Experiment. Science 243, 57-63 (1989) Randall, D.A., et al, Intercomparison and interpretation of surface energy fluxes in atmospheric General Circulation Models. J. Geophys. Res., 97, 3711-3724 (1992) Warren, S.G., C.J. Hahn, J. London, R.M. Chervin and R.L. Jenne, Global distribution of total cloud cover and cloud type amounts over land. NCAR Technical Note 273 and DOE Technical Report No. ER/60085-H1, US Department of Energy, Carbon Dioxide Research Division, Washington D.C. 20545, (1986) Warren, S.G., C.J. Hahn, J. London, R.M. Chervin and R.L. Jenne, Global distribution of total cloud cover and cloud type amounts over the oceans. NCAR Technical Note and DOE Technical Report No. ER/60085-H2, US Department of Energy, Carbon Dioxide Research Division, Washington, D.C. 20545, (1988)
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1377
S o c i o - e c o n o m i c a n d p o l i c y a s p e c t s of c h a n g e s in i n c i d e n c e a n d i n t e n s i t y of e x t r e m e w e a t h e r e v e n t s . P r e l i m i n a r y r e s u l t s .
Dorland, C., W.J. Maunder, A.A. Olsthoorn, R.S.J. Tol, P.E. van der Werff and P. VeUinga
Institute for Environmental Studies, Vrije Universiteit, De Boelelaan 1115, 1081 HV Amsterdam, The Netherlands
Abstract Climate change results in an alteration of spatial and temporal patterns of climate hazards. The trend in weather related disaster seems upward. Various socio-economic sectors are affected by these changes, e.g. the disaster reduction institutions and the insurance industry. We report about an ongoing project addressing the vulnerabilities of sectors affected and policy options in various sectors, notably "Storms over NW-Europe", "the insurance sector" (both as a sector impacted by change and as a mechanism to cope with risk) and "cyclones in the South Pacific".
1.
Introduction
The upward trend in climate-disaster statistics has aroused interest in climate change issues in various sectors affected, among them the the insurance sector. The goal of the study is to contribute to the evaluation of this aspect of climate change, particularly the socio-economic aspects. The study was outlined at a workshop in which scientists from climatology, insurance and economics participated (Olsthoorn & Tol, 1993). The methodology is to construct scenarios of extreme weather events (storms, droughts etc.), their impacts and policy responses. Insurance is a focal point. This paper reports about the ongoing work at IES-Vrije Universiteit, Amsterdam: the work done by the collaborating institutes (Environmental Change Unit, University of Oxford, Aristotelian University of Thessaloniki) will be published elsewhere. Sofar the IES-work, reported here, addressed the three topics briefly discussed below.
2.
Storm over Northwest Europe" Daria modelled
In the winter of 94/95 Northwest Europe experienced an unusual period of seven severe storms. One of them, Daria, hit the Netherlands on the 25th and 26th of January 1990. This storm, the strongest since 1944 in the Netherlands, caused the death of 25 people and, at its height, brought traffic to a virtual standstill while the railway system was blocked completely due to many broken high-tension cables (Dorland et al, 1994). It caused widespread damage estimated at about Dfl 2.6 billion, of which Dfl 1.5 billion was covered by the insurance industry.
1378 An extensive literature search was carried out to identify the policy responses to this disaster. It appeared that only in a few sectors (Dutch Railway, National Forestry) the events prompted reactions. Five years later hardly any societal traces are left. Apparently, a developed society is not vulnerable to such event (Albada-Bertrand, 1993); at a societal level an incidental damage of billions of guilders can be coped with (at its current incidence). However, studying the events will give clues about what will happen if climate change results in more intense and frequent storms. Unfortunately, much of the information gathered about the storm impact proved anecdotical and not suitable for modelling, except for damage data from the insurance industry. This industry provided information which allowed construction of a storm-damage model seeking to predict damages done by hypothetical storms. Such model is constructed by specifying a relation between spatially disaggregated (2-digit postal code area) data about stock-at-risk and corresponding windspeed data, and subsequent statistically estimating of the coefficients using the information about the actual damages done by storms. Solar, only Daria could be modeled. The Centre of Insurance Statistics (CVS) provided data of storm-related claims (from privates, and not businesses and agricultural objects). Windfield data were obtained from the Dutch Met Office. Other variables included are the number of houses and the average income of households in the postal code areas. These variables together reflect the stockat-risk in each area. Damage = 1.810-'PH+1.901v3+54250AI-11.9316 Damage = Insured damage including deductibles (MDfl) PH = Number of private houses AI --Average income of households (thousand) v = Maximal windgust (m.s-1) R z = 0.84
The damage is put proportional to the third-power windspeed for physical reasons (force is proportional to windspeed to the third power). The model will be improved by incorporating other storms in the statistical analysis.
3.
A weather insurance simulation model.
The recent large losses due to hurricanes, floods and storms not only aroused the attention of the climate change community but also of the (re)insurers. The larger part of the increased damage is, however, ascribed to the increase of socio-economic vulnerability and short-sighted price setting. Responses are now vividly discussed within the insurance sector. Three response types can be distinguished: the product can be adjusted by raising premia and restricting cover (increase of deductibles and rejecting low-probability high risks), the financial buffer-capacity can be enlarged (closer cooperation with banks and governments), and the vulnerability can be reduced (enforcement of building codes, land use zoning). The first type shifts the risk back to the insured, the latter types redefine the roles the various stake-holders play in risk management. The analyses suggest that the insurance sector is likely to be able to adjust to climate change without disproportional damage, provided that changes are not too fast or drastic and responses are clever.
1379 In order to be able to quantitavely study the insurance sector in a changing climate, a a numerical dynamic model of the weather insurance market was constructed. This model, the Weather Insurance Simulation Model (WISM.IO) (Tol et al, 1994), is used in a preliminary study the implications of hypothetical storm scenarios on the insurance sector. WISM is a model of the insurance market driven by a double stochastic process (PoissonPareto). The Poisson-Pareto process generates a random number of storms in a year and the corresponding random intensities (windspeeds) of these. One thousand sequences of a hunderd years are typically computed in order to calculate the means and variances of the model's output. Average frequencies and intensities are determined by factors constituting constant risk, gradual increases, slow cycles (representing the decadal climate variability), and cycles plus a trend. Again, storm damage is assumed to be the third power of storm intensity. Storm inflicted damage is fed into a highly stylised, behavioural model of the main actors on the insurance market, viz., the insured, the insurers and the reinsurers. The actors' behaviour is based on their perception of the distribution of storm number, intensity and damage, which is in turn based on past observations. That is, the actors observe the random (!) figures of the previous years and base their expectations of the current year thereon. The (re)insurers are assumed to maximise their profits (premia minus expected losses minus risk factor), the insured to minimise their losses (premia plus expected losses minus expected claims plus risk factor). The (re)insurance market can assume several forms ranging from a regulated monopsony/monopoly to a free market. The average and standard deviation of main parameters, such as deductables, premia, (re)insurance cover, rentability and chance on insolvency, are calculated. The first results indicate that the market form is largely irrelevant to the outcomes. In case storm frequency increases, the (re)insurers can and do shift the risk to the insured. The position of the (re)insurers is only mildly affected, whereas the insured face a substantial increase in expenditures. Long cycles in storm frequency have a similar impact. The picture is different when climate change changes storm intensity. The results indicate that increases in storm intensity affect all parties alike. Insurance cover and premia rise, but profits fall, and expenditures and insolvency chances rise. However, it is again the insured who suffer most. Moreover, important instruments to limit the reinsurers' risk have, solar, been left out of the analysis. In conclusion, the weather insurance model confirms the qualitative notions described above: The insurance sector is not likely to be a disproportionate loser under climate change.
C y c l o n e s in t h e S o u t h - P a c i f i c . Apparently, wealthy countries such as the Netherlands can cope with severe storms; an obvious proposition is that developing countries (DCs) subject to tropical cyclones, more fierce than North-Atlantic cyclones, do not. The conditions necessary for tropical cyclone genesis prevail in the Tropical part of the Atlantic Ocean (except at the southern hemisphere), in the Indian Ocean and in the Pacific, by coincidence the area were most of the small island states are located. Tropical cyclones over Fiji is the subject of a case study adressing vulnerability of DCs to cyclones. Fiji (population about 750,000) is a small
1380 island state exhibiting typical characteristics of these kind of countries: a less diversied economy, tourism being a major sector. Agriculture, including a large subsistence sector, provides for most (80%) of the employment. Fiji is member of the Alliance Of Small Island States (AOSIS), which, as a representative of countries particularly impacted by climatic change, is involved in the international negotiations on climate change policy. The statistics of cyclone incidence and and intensities in the South Pacific do not allow firm conclusions about trends (Tol et al, 1993); comprehensive statistics are available only since 1970. Since then 238 cyclones were recorded. Since 1980, the islands of the Fiji archipelapo were struck by major cyclones in 1983, 1985 and in 1993. Seven, 27 and 23 people killed was reported. Damage was similar in the tree events: in the order of F$ 100 million. The government respectivily spent F$ 3.2 million (1985) and F$ 10 million (1990) on food relief in reaction of subsistence households losing crops. This amount is to compare with the government's capital expenditures which were in 1990 F$ 79 million. The Fijian tourism industry faced a 200%300% increase of insurance premiums. Fiji started to develope pine forestry around 1980. In 1990 the export of timber accounted for 2.5% of the country's total export. However, cyclones appear to be the limiting factor for developing this sector; in 1990 10% of the stands had to be written off due to cyclone damage.
4
Conclusions
A comparison of the impacts of storms confirms the view that vulnerability is largely a function of development. Both in terms of casualties and damage done small, less developed, island states suffer much more in comparison. However, empirical data do not confirm the view that weather disasters hamper macro-economic development (AlbalaBertrand, 1993). The economic impact is distributional. In case of a changing storm regime insurance industry will suffer when it cannot adapt swiftly enough, while Fijian agriculture, forestry and tourism are likely to be restricted in development potential.
5
References
Albada-Bertrand, J.M. (1993), The political economy of large natural disasters. ClarendonPress, Oxford Dorland, C., A.A. Olsthoorn and R.S.J. Tol, (1994), The 1990 Winterstorms in the Netherlands, Institute for Environmental Studies, VU, Amsterdam (forthcoming) Olsthoom, A.A. and R.S.I. Tol, (1993), Socio-economic and policy aspects of changes in the incidence and intensity of extreme weather events. Report workshop 24 -25 June 1993, Amsterdam, IvM-VU Amsterdam Tol, R.S.I. and A.A. Olsthoorn, (1994), Climate change, wind storms and the insurance industry. Some notions from a weather insurance simulation model, part I, Institute for Environmental Studies, Vrije Universiteit W-94/15, Amsterdam Tol, R.S.J., C. Dorland and A.A. Olsthoorn, (1994), Statistical evidence on genuine and observational trend in tropical cyclone frequency in the Southwest Pacific, Submitted to Geophysical Research Letters
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1381
Probability of climatic change; identification of key questions Wieger Fransen Royal Netherlands Meteorological Institute KNMI, P.O. Box 201, 3730 AE De Bilt, The Netherlands; fransen @knmi.nl Abstract Addressing the question what the probability is of an anthropogenically induced change in the climate, leads to a number of other, underlying questions. These questions, which deal with the characteristics of climate, of climatic change, and of probabilistic statements on climatic change, should be adressed first. The long-term objective of the underlying study, i.e., a quantitative assessment of the risks and opportunities of the predicted climatic change, sets the context against which these questions should be answered. In addition, this context induces extra questions, i.e., about the characteristics of risk.
1. I N T R O D U C T I O N The probability of climatic change, which is due to the anthropogenic enhancement of the Greenhouse Effect, is questioned increasingly in the press. Due to this, interest by both politicians and managers of private enterprises in the predicted climatic change is fading. Policymakers working in the field of climatic change need this interest because they need support for measures aimed at mitigating the risks of, and at adapting to, the predicted global warming, both on a national and an international level. So, they look for ways to renew this interest. The renewal sought for can perhaps be accomplished by founding climate policy on a risk-analysis of the global warming issue. If so, it may also be possible to compare the risks of climatic change with other risks for society. For such a risk-analysis, first of all quantitative probabilistic statements on the expected change are needed. However, information on climatic change is currently almost without exception presented in terms of uncertainty. The project 'Probability of Climatic Change' aims at identifying ways which may result in the probabilistic statements required. This paper identifies the key questions which should be addressed first. 2. U N R A V E L L I N G THE ISSUE
In this section will be shown that there are many possible ways to come to probabilistic statements on climatic change. The issue has in this paper been reduced to linking four categories of information: information on climate, on climatic change, on the foundations of probability, and on risks. Within each category several key questions are presented. The answers to these questions -nine in total- present options which may be focused on when assessing the risks of a change in the climate. 2.1. Characteristics of probabilistic statements For most climatologists, including myself, this category of information, which deals with the foundations of probabilty, is remote from their daily activities. For this reason, some more background information will be presented on the topic than will be the case for the other three categories.
All probability ideas, which can be traced back to the ancient Egyptians and Greeks, have contributed to the foundations of probability. Three interpretations of probability prevail currently:
1382 1) The interpretation adhered to most often is the frequentist view, or frequency interpretation, of probability which is based on a notion of randomness and repeated experiments modelled by the sample space. 2) The subjective view of probability describes the strength of belief of an individual concerning the occurrence of events. Strength of belief is determined through a process of introspection and manifests itself through overt choice or betting behaviour, 3) Logical probability presents an objective assessment of the degree to which an evidence statement (inductively) supports a hypothesis statement. A gap may be expected between the foundations of probability, to which these three interpretations belong, and applied statistics; what kind of probability model can be used for which interpretation. Knowledge of the different interpretations of probability, though, can guide the selection of families of probability models (not necessarily numerical ones) so as to better reflect the indeterminate, uncertain, or chance phenomena being treated. Knowledge of the different interpretations may also clarify a choice among the divergent, conflicting statistical methodologies now current. One should realize that different methodologic schools rely on different concepts of modelling probability, albeit this difference is obscured by common agreement on the mathematical structure of probability. Regarding the development of probability models for uncertain events, three different concepts are identified: 1) A model cannot be developed (Neyman-Pearsonians postulate that a class of uncertain phenomena, i.e., the 'unknown parameter', cannot be given a probability model), 2) A choice of models may be developed (Bayesians, personalists and subjectivists insist upon giving the unknown parameter an overly precise numerical probability model, but allow great freedom in the subjectively based choice of the model), 3) Only one, unique model can be developed (structuralists, fiducialists and maximum entropists carry the modelling process one step further by claiming to provide objective, rational grounds for the selection of a unique numerical probability model to describe the unknown parameter). A domain contains both events whose occurrences are of interest to a reasoner and a setting identified by the reasoner as informative about the occurrence of events and as relevant to achieving its goals. In some fashion, the reasoner decides that it can perhaps identify which of the events are probable, or which events are more probable than other events, or even assign a numerical probability to each event. Implicit in this process is an initial determination as to what provides the evaluative basis for the probability concept being invoked (e.g., what climatic records and theory can we use to calculate the probability of a climatic change in the next century). The evaluative basis largely fixes the meaning of the probability concept, which must have meaning extending beyond its evaluative basis if it is to serve a role other than that of data summarization. Question 1: What evaluative basis should we choose for the probabilistic statements on climatic change sought for? Possible evaluative bases are: 1) Past occurrences of other events of the same type (the palaeo-analogue method), 2) Experiments generating the events (output from simple climate models, e.g., autoregression models, or complex climate models, e.g., coupled-GCMs), 3) The strength of belief of an expert concerning the events (surveys of expert opinion or statements by individual experts), 4) The inductive relation between a formally presented amount of information and the event (due to the complexity of the climatic systm this method is not usually applied and, if so, often patronizingly called 'hand-waving'). 1 and 2 belong to the frequency interpretation of probability, 3 belongs to the subjective interpretation of probability, and 4 belongs to the logical interpretation of probability.
1383 Once the reasoner has adopted a concept of probability supported by a domain of application, he then wishes to move this empirical relational system into a formal mathematical domain so as better to determine the implications of the position. The events of interest in the domain are represented either by sets or by propositions. It is generally not possible to enumerate all possible events (complex systems occasionally surprise us by behaving in unforeseen ways) and therefore the sample space is at best a list of practical possibilities. The recognition that probabilistic reasoning must confront a wide range of domains and levels of information, knowledge, belief, and empirical regularity can lead us to an acceptance of an hierarchy of increasingly precise mathematical concepts of probability. This hierarchy has been little explored, as almost all of the effort has been devoted to numerical probability. That numerical probability may be inadequate to the full range of uses of probabilistic reasoning is suggested by the following observations: 1) For some categories of empirical phenomena (e.g., climate) there is no obvious stability of relative frequency for all events of interest. 2) An ensemble of events may lack information; the resulting indeterminacy should be respected and not be obscured by applying dubious hypotheses (e.g., "If you know nothing about the parameter, then adopt a uniform maximum entropy for it"). 3) Self-knowledge of individuals is intrinsically limited, and attempts to force belief or conviction to fit the mold of a particular 'rational' theory can only yield results of unknown value. Question 2: What precision should the probabilistic statements have? An attempt to accommodate to the preceding observations leads to the following hierarchy of concepts: 1) Possibly, the globe will warm by between 1 and 3 ~ in 2050, 2) Probably, the globe will warm by between 1 and 3 ~ in 2050, 3) That the globe will warm by between 1 and 3 ~ in 2050 is at least as probable as that the globe will warm by between 0 and 1 ~ in 2050, 4) That the globe will warm by between 1 and 3 ~ in 2050 has a probability of between 4 out of 10 and 8 out of 10, 5) That the globe will warm by between 1 and 3 ~ in 2050 has a probability of 6 out of 10. Conditional versions of each of the foregoing concepts are also available and will in reality be the versions dealt with. An example of a conditional version of the foregoing concept is established when the following phrase is put before each concept: If atmospheric greenhouse concentrations continue to increase according to the IPCC IS92a scenario, then .... (for this, see question 5). The probability concepts just introduced must then be given structure through a set of axioms and definitions of significant terms (e.g., independence, expectation). While it is the role of interpretation to co-ordinate the mathematical, axiomatically constrained concept with the domain of events of interest to the reasoner, this coordination is typically idealized and not itself a working basis for probabilistic reasoning. Statistics is the discipline that supplies the working basis for numerical probability with a frequentist interpretation. Statistics is also of value in supplying the basis for numerical probability in the subjective setting. Little is yet known about the practical issues connected either with formal concepts of probability other than the numerical one or with the logical interpretation of probability. 2.2. Characteristics of climate Question 3: Which climatic variables are of most importance when assessing the risks of a change in the climate? Four categories of variables could be identified, of which two comprised variables which were considered to be strongly related. These are:
1384 1) Precipitation (intensity, surplus, i.e., precipitation minus evapotranspiration), 2) Temperature (extremes, averages, freezing days or 'tropical' days), 3) Cloud coverage and irradiance (of both short-wave and long-wave radiation), 4) Storms, tidal amplitude and sea-level. It should be added that risks may also be due to changes in several variables, which do not necessarily have to be of climatic origin, occurring at the same time, which are not significant by themselves but which are significant when occurring in ensembles. This is called multi-stress. Question 4" What statistics should be used? Climate can be defined as 'the characteristics of weather seen over longer periods'. But depending on the statistical processing of weather information, one and the same climate could be presented in different forms. Eight categories of statistical representation were identified. These are: 1) Extreme values, 2) Averages, 3) Trends, 4) Variability, 5) Spatial and temporal correlation, 6) Run events, 7) Distribution, 8) Timing.
2.3. Characteristics of climatic change Question 5" What reference should we take if we talk about climatic change? When discussing climatic changes, it is implicitly assumed that the climatic issue of interest shows a temporal evolution or trend. However, climate does not change due to time, but because processes influencing climate directly or indirectly change in time -the internal climatic variability is disregarded at this point. Processes influencing climate directly are changes in land-use influencing the hydrological cycle (e.g., changes in runoff due to deforestation influencing in turn rain patterns and groundwater levels) and temperature (e.g., via albedo changes and by creating so-called islands of urban heat). The direct effect is primarily regional, i.e., only extending to neighbouring areas. Processes changing climate indirectly via perturbation of the Earth's radiative balance are absorption and emission of long-wave or short-wave radiation by gasses, reflection and absorption of short-wave radiation by the Earth's surface and scattering of short-wave radiation by particles in the air, e.g., aerosols. The indirect effect is primarily global. Two references were identified: 1) Changes in the concentrations of atmospheric constituents influencing the radiative balance (affecting climate globally, with regional variation due to uneven distribution of emissions of some short-living constituents like, for example, sulfur dioxide and soot), 2) Changes in land use (affecting the climate globally via changes in emissions and albedo. Land-use changes have a regional influence because they may influence the local heat-balance by the heat produced directly, e.g., by cities, and by changes in the albedo which influences the amount of incoming solar radiation used for surface warming. Changes in land use may also influence the hydrological cycle regionally). Question 6: What geographical scale should we consider if we discuss climatic change? It can be observed that when people or the press discuss climatic change, global or hemispheric values are presented if any. The same can be said of many scientific reports and of the policymakers summaries, for instance of the IPCC. In general it can be said that 'the global climate' is an empty concept; there is no such thing as a global climate. This is made clear by the following example, which at the same time links the question of geographical scale and risks of climatic change. Four coupled models predict
1385 that the global temperature will have increased by between 1.3 and 2.3 ~ at the time of CO2 doubling. All four models show local hearings of up to 2.5 "C, one of up to 5 ~ one of up to 6 ~ and one of up to 7 ~ Two models also show regions which will cool, one of these even up to -6 ~ Two geographical scales of interest for the underlying study were identified: 1) Regional (a region with a specific climate as defined by a climatic classification system, e.g., the K6ppen System or the Holdridge Classification. Such a region may be very large indeed) 2) Local (an area within a climatic region). Question 7: What temporal resolution should we strive for when studying climatic change? If one is interested in studying climatic changes in the past then the temporal resolution of the records tells us what changes can be resolved. The Nyquist-theorem indicates that the sampling frequency should be at least twice as high as the highest frequency that should be resolved. So, if we have measurements with 50 year intervals than changes over 100 years can be resolved. If we, on the other hand, are interested in seasonal changes in the future, then we need model output -if we wish to rely on model output, that is- with monthly resolution. Climatic risks are often associated with changes in agricultural production, ff so, even higher resolutions are required. The cultivation of rice is, for example, highly dependent on maximum daily temperatures. Prediction of sealevel rise due to thermal expansion of the water allows climatic information with a lower temporal resolution. Taking future societal risks as boundary condition for the question of temporal resolution, the following hierarchy is identified: 1) Annual values, 2) Seasonal values (by definition seasonal values should be climatological homogeneous, this implies that the number and location of seasons should be chosen appropriately), 3) Values of Julian days, 4) Day-rime or night-time values. 2.4. Characteristics of risks in the context of climatic change It should be noted first that climatic change will also have beneficial effects. Indeed, as risk-assessments on the global warming issue have not yet been done, it cannot be said in advance that the risks will be smaller or larger than the opportunities.
Question 8: What kind of risks are we talking about. One may think of: 1) Damage to, or loss of, ecosystems (leading to an increase of the 'natural debt'), 2) Direct economical loss (which may also be expressed by changes in discount rate), 3) Damage to the physical and mental health of people, 4) Political instability due to indirect socio-economic effects (climatic change may indirectly lead to migration of large groups of people. It may also lead to tension between neighbouring countries if, for example, the agricultural production increases in the first country and decreases in the second), 5) Food production (including agricultural production and fishery). Question 9: Whose risks are we talking about. From an anthropogenic point of view can be defended that the answer on this question is dependent on the people who, or institutions which, decide over, or are in power in, a specific area: the decisive bodies. The following hierarchy is proposed: 1) Bodies operating intercontinentally (e.g., UN, OESO, and OPEC), 2) Bodies operating continentally or transboundary (e.g., EU, USA, and BENELUX), 3) Bodies operating nationally (e.g., nations and states), 4) Bodies operating locally (e.g., cities, towns, and municipalities), 5) Individuals, families, households, offices, shops, communities, et cetera.
1386 Ideally, the bodies one to four should comprise of politicians only. The politicians in turn should ideally represent the interest of institutions like NGO's, industries, trade associations, organisations, et cetera, as promised to all individual voters before the elections. In reality one may often have the impression that in certain regions some institutions have more power than the politicians in charge. Another point which should not be overlooked is that the climatic risks for a specific country may partly depent on the climatic risks of another country with which is has a physical, economical, or some other relationship (for this, see question 8). 3. D I S C U S S I O N AND C O N C L U S I O N Several options within nine categories representing four categories og information have in this paper been presented. These options may be focused on when assessing the risks of a change in the climate. However, when it is decided that the focus should be on one or more of the options given, this may effectively exclude that other relevant information can be extracted from the study results. For instance if the focus is on local risks of a temperature increase on the agricultural sector, one may not reveal information on the risks of changes in precipitation patterns for the ecosystems within a nation. Shortly, starting with a decision in one category will have consequences for decisions to be taken in the other categories. So, first of all decisions should be taken on the value of the different approaches that can be considered. The value of different possible approaches will be elaborated on in the course of the project and has, thus, not been discussed in this paper. The assessment of quantitative probabilities on the risks and opportunities of the predicted climatic change should be started with an analysis of the problem. According to this study, the problem analysis should consist of addressing at least nine keynote questions which deal with four different types of information. It is acknowledged here that even more fundamental approaches to the different types of information may be required. For instance, addressing the question which climatic variable should be studied, as has been done in this paper, may by some be interpreted as that changes in all climatic variables are predictable. From the work of Lorenz and others follows that if we are also interested in the chronological order in which these changes take place the answer should be 'no'. Addressing the question 'whose risks' implies that there is a common perception about risks. This seems improbable and should thus be accounted for when pursuing the long-term objective mentioned. The differences in character between the key questions identified leads to the conclusion that a risk-assessment of climatic change will be the result of a common effort by many specialist, among which experts from both the natural and the social sciences, policymakers, and politicians. This also stresses the need for NRP projects aimed at communication. 4. REFERENCE AND A C K N O W L E D G E M E N T S The results presented have been partly based on information provided by participants of the conference. These were actively asked to write their comments on the poster presented by the author; a so-called 'interactive poster'. The persons who did so are kindly acknowledged. I am indebted to Albert Klein Tank from KNMI for critical notes on a draft of this paper. Section 2.1. is an adaptation of an article by Terrence Fine in the Encyclopedia of statistical sciences (Eds Kotz, Johnson and Read), Vol. 3, pp 175-184, WileyInterscience (1983). The project 'Probabilities of climatic change' has been commissioned by the Ministry for the Environment of The Netherlands.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1387
T h e role of fear a n d threat in c o m m u n i c a t i n g risk scenarios a n d the n e e d for actions: E f f e c t of fear on i n f o r m a t i o n p r o c e s s i n g
A.L. Meijnders", C.J.H. Midden", H.A.M.
Wilke b
" Eindhoven University of Technology, Faculty of Philosophy and Social Sciences, Department of Psychology and Linguistics, P.O. Box 513, 5600 MB Eindhoven, The Netherlands
b Leiden University, Faculty of Social Sciences, Department of Social and Organizational Psychology, P.O. Box 9555, 2300 RB Leiden, The Netherlands
Abstract
This paper is about the for actions. The results fear of environmental environmentally sound
role of fear and threat in communicating risk scenarios and the need of our first experiment are discussed, in which we examined whether risks increases the tendency to carefully process information on behaviour.
1. INTRODUCTION
For many people large scale environmental risks such as global warming are hardly more than an abstract topic. The process of global warming is invisible, complex and distant. In addition daily newspapers inform us continuously about the many uncertainties concerning the nature, magnitude and time-scale of negative consequences. The policy approach to this situation of uncertainty has been the introduction of the precautionary principle, which holds that policy actions have to be undertaken in spite of existing uncertainties. Although this may be a wise strategy of risk management, it unavoidably raises questions about the justification of controversial measures with strong impacts on society. Non-justified policy measures are likely to be rejected, especially when large groups are affected. From that perspective it is crucial for environmental policy, that the possible threats of global warming can be presented in such a way, that citizens as consumers or entrepreneurs become convinced of the need for actions. Two elements are important in achieving this goal. Firstly it is a task of science to reduce uncertainty by diagnostic research. Secondly it is essential that risk scenarios are communicated in an effective way which overcomes the tendency to downgrade these risks because of their lack of perceived significance. Recent research findings indicate that the impact of information campaigns (e.g. the Dutch campaign on climate change) is limited. Information often is insufficiently elaborated and therefore attitudinal changes do not come about, let alone behavioural changes [1]. Thinking in terms of the dual-process theories of persuasion, this may be due to a lack of
1388 motivation to process such information [2-3]. Attempts to overcome these motivational difficulties have resulted in a higher emphasis on emotional factors in communicative programmes. However, little is known about the effects of emotion oriented communications, and how these effects come about. The purpose of the present research project is to increase our understanding of the role of emotional factors in communicating risk scenarios and the need for actions. Our special interest is in fear appeals, because there is a rich and promising theoretical and empirical literature on the effects of fear appeals in the field of health education (see [4], for an overview of the theoretical literature; for an example of an empirical study, see [5]). A first step in increasing the effectiveness of environmental information, is to stimulate people to elaborate this information. In our first experiment we therefore examined whether fear of environmental risks increases the tendency to elaborate information on environmentally sound behaviour.
2. M E T H O D
2.1. Design
To examine whether fear of environmental risks increases the tendency to elaborate information on environmentally sound behaviour, the following variables were manipulated in a 2 x 2 between-subjects factorial design: Fear level (low or high) and argument quality (weak or strong). The manipulation of argument quality is assumed to be an effective way of locating differences in message processing [2]. The idea is that only when a persuasive message is carefully processed, the arguments presented in the message will have an impact on attitudes towards the message topic. This implies, that the effect of argument quality on attitudes can be considered an indication of the degree to which a message is elaborated. Other widely employed indicators of message processing are: The number of issue-relevant cognitive responses generated during message exposure and the number of message arguments recalled afterwards. The underlying idea is, that the more a message on a certain issue is elaborated, the more issue-relevant thoughts will be generated during message exposure, and the more arguments presented in the message will be recalled. 2.2. Procedure
Subjects were 76 inhabitants of Eindhoven, the Netherlands, who were assigned randomly to the experimental conditions. Subjects received all experimental instructions, manipulations and measures by means of computers. The experimental procedure was as follows. First, subjects were exposed to either a slightly or a highly frightening message on the greenhouse effect. Next, they received either a weak or a strong persuasive message arguing for the use of a new type of energy saving light bulbs. After having read these messages, subjects completed a questionnaire. The most important measures in this questionnaire were manipulation check measures, and measures of the dependent variables, i.e. cognitive responses, attitudes towards using the new light bulb, and recall of arguments. After having completed the questionnaire, subjects were debriefed and then dismissed. 2.3. Stimulus materials
The message on the greenhouse effect presented the manipulation of fear level. The slightly frightening version of the message described the process of global warming and its
1389
possible negative consequences, whereas in the highly frightening version in addition five black and white photographs of the possible negative consequences of global warming (e.g. floods) were shown. These photographs were impoverished by means of a computer to such a degree, that risk imagination was tickled without providing extra information. The persuasive message on energy saving bulbs presented the manipulation of argument quality. This message consisted of a description of a new (fictitious) type of energy saving light bulbs and four arguments in favour of purchasing and using this new type of bulbs. In the weak version of the message four weak arguments were presented, whereas in the strong version of the message four strong arguments were presented. These arguments were selected from a large pool of arguments that were pretested in a pilot study on 8 subjects.
3. RESULTS
3.1. M a n i p u l a t i o n c h e c k s
To check on the success of the manipulation of fear level, subjects were asked to rate on four 7-point scales (anchored at 1 = not at all and 7 = extremely) the extent to which they thought the message on global warming they were previously exposed to, was frightening. Ratings on these four items, which were correlated with one another (correlations ranged from 0.47 to 0.67), were averaged to create a composite measure of frightfulness (Alpha = 0.85). Next, the composite measure was analysed in a 2 (low versus high fear) x 2 (weak versus strong arguments) between-subjects ANOVA. This analysis yielded a significant effect of fear level, F (1,72) = 7.06, p < 0.01. Subjects in the high fear condition rated the message on the greenhouse effect as significantly more frightening (M = 5.56) than subjects in the low fear condition (M = 4.80). To check on the success of the manipulation of argument quality, subjects were asked to rate the strength of each of the arguments presented to them on a 7-point scale (anchored at 1 = not at all and 7 = extremely). Judgments of the four arguments were averaged to create a composite measure of argument quality. Next, the composite measure was analysed in a 2 (low versus high fear) x 2 (weak versus strong arguments) between-subjects ANOVA. This analysis yielded a significant effect of argument quality, F (1, 74) = 24.18, p < 0.0001. The strong arguments received significantly higher ratings of strength (M = 5.58) than the weak arguments (M = 4.25). 3.2. Effects on a t t i t u d e s
To assess subjects' attitudes towards using the new energy saving bulb, they were asked to rate on four 7-point scales (ranging from 1 = not at all to 7 = extremely) the extent to which they thought the bulb was suitable for usage in their own households. Ratings on these four items, which were correlated with one another (correlations ranged from 0.52 to 0.79), were averaged to create a composite measure of attitude (Alpha = 0.89). The composite measure of attitude was then analysed in a 2 (low versus high fear) x 2 (weak versus strong arguments) between-subjects ANOVA. This analysis yielded a significant interaction-effect of fear level and argument quality, F (2, 75) = 4.48, p < 0.038. Figure 1 shows that argument quality had no effect in the high fear condition, but in the low fear condition strong arguments had more effect on attitudes than weak arguments.
1390 Attitudes 7
-
6
-
strong arguments
weak arguments
strong
5 -
4 -
3 -
0
/
low fear
high
Figure 1" attitudes towards the new energy saving bulb level and argument quality
fear
as a function
of fear
3.3. Effects on cognitive responses To assess subjects' cognitive responses to the persuasive message, they were requested to complete a thought-listing task. Subjects were asked to write down all the thoughts that came to mind while reading the persuasive message on the energy saving bulb. For this purpose, subjects were provided with a form containing numbered boxes, and they were instructed to write down only one thought per box. issue-relevant cognitive responses 5-
4-
3 -
2-
1-
low fear Figure 2: issue-relevant cognitive responses,
high f e a r as a function
of fear
level
The thoughts listed by the subjects were categorized by two independent judges, who rated the relevance of the responses. Agreement between the judges was 97 %. Mean scores
1391
for the two judges were analysed. As can be seen in Figure 2, subjects in the high fear condition generated significantly more issue-relevant cognitive responses, M = 4.32, than subjects in the low fear condition, M = 3.40, F (1,75) = 4.03, p < 0.049. 3.4. Effects on a r g u m e n t recall
To assess subjects' recall of the arguments that were presented in the persuasive message, they were requested to write down everything they remembered about the persuasive message on a blank sheet of paper. Two independent judges rated the number of correctly remembered arguments. Agreement between the judges was 90 %. Mean scores for the two judges were analysed in a 2 (low versus high fear) x 2 (weak versus strong arguments) between-subjects ANOVA. No significant main effect of fear level was found, F < 1, n.s..
4. DISCUSSION
Does fear of environmental risks increase the tendency to process information on environmentally sound behaviour? The results of the present study indicate, that this question cannot be answered with a simple yes or no. On the one hand it was found that significantly more issue-relevant cognitive responses were reported in the high fear condition, than in the low fear condition. On the other hand it was found that argument quality affected attitudes in the low fear condition, but had no effect on attitudes in the high fear condition. The extensive literature on the effects of fear on information processing may help us to interpret these findings. Recently, the results of two studies on the effects of fear on information processing were published, which showed that fear may interfere with systematic processing of irrelevant information, i.e. information which is unrelated to the threat [6-7]. For example, in one of the experiments reported by Baron and colleagues, it appeared that fear of a dental treatment interfered with systematic processing of information on sailes taxes. In another study, Baron and colleagues found that fear of a dental treatment facilitated systematic processing of information on fluoridated water, which suggests that fear may stimulate systematic processing of relevant information, i.e. information that is related to the threat. In the present study we also found indications that low levels of fear may facilitate systematic processing of relevant information, for we found that when fear of the greenhouse effect was low, information on energy saving bulbs was systematically processed. On the other hand however, we found that when fear of the greenhouse effect was high, information on energy saving bulbs was not elaborated. This latter finding suggests that high levels of fear may interfere with systematic information processing. According to the protection motivation theory of Rogers, fear motivates people to seek protection from the threat they are exposed to [9-10]. This protection seeking process most likely involves mental activity. High levels of fear presumably induce stronger motivation to seek protection, and hence more mental activity, than low levels of fear. This might explain why in our experiment subjects in the high fear condition generated significantly more cognitive responses, than subjects in the low fear condition. At the same time it explains why subjects in the high fear condition made no distinction between weak and strong arguments, whereas subjects in the low fear condition did differentiate between weak and strong arguments. Subjects in the high fear condition invested so much mental capacity in dealing with the threat, leaving insufficient capacity for elaboration of the message on energy saving bulbs. In other words, our hypothesis is, that fear may have a positive effect on motivation
1392 to elaborate relevant information, but at high levels of fear, this positive effect may be overruled by a negative effect on information processing capacity. An alternative explanation of the results presented in this paper is, that the pictures presented to the subjects in the high fear condition made great demands on their information processing capacity, not because they aroused fear, but because cognitive capacity was needed to interpret them. Although it is too early to formulate clear recommendations on how to deal with emotions in persuasive communications, the literature as well as the results of our experiment suggest, that the role of fear in the persuasion process is far from simple. In several communicative programmes emotional appeals have been applied to convince people to take account of the environmental consequences of their behaviour. However, the outcomes of our research project so far indicate, that emotional appeals should be applied only under carefully specified conditions. In the remaining two years of the project we hope to learn more about the role of fear and threat in communicating risk scenarios and the need for policy measures and behavioural changes.
5. R E F E R E N C E S
1
Staats, H.J., & Midden, C.J.H. (1991). Voorlichting over het broeikaseffect: Evaluatie van de eerste fase van de klimaatcampagne (E&M/R-91/27). Leiden: Rijksuniversiteit Leiden, Werkgroep Energie- en Milieu-onderzoek. 2 Petty, R.E., & Cacioppo, J.T. (1981). Attitudes and persuasion: Classic and contempora~ approaches. Dubuque, IA: Wm. C. Brown Company Publishers. 3 Chaiken, S. (1980). Heuristic versus systematic information processing and the use of source versus message cues in persuasion. Journal of Personality and Social Psychology, 9, 752-766. 4 Eagly, A.H., & Chaiken, S. (1993). The psychology of attitudes. Fort Worth: Hartcourt Brace Jovanovich College Publishers. 5 Rippetoe, P.A., & Rogers, R.W. (1987). Effects of components of protection-motivation theory on adaptive and maladaptive coping with a health threat. Journal of Personality and Social Psychology, 52, 596-604. 6 Baron, R., Inman, M., Kao, C ~F., & Logan, H. (1992). Negative emotion and superficial social processing. Motivation and Emotion, 16, 323-346. 7 Wilder, D.A., & Shapiro, P. (1989). Effects of anxiety on impression formation in a group context: An anxiety-assimilation hypothesis. Journal of Experimental Social Psychology, 25, 481-499. 8 Baron, R., Logan, H., Lilly, J., Inman, M., & Brennan, M. (1994). Negative emotion and message processing. Journal of Experimental Social Psychology, 30, 181-201. 9 Rogers, R.W. (1975). A protection motivation theory of fear appeals and attitude change. Journal of Psychology, 91, 93-114. 10 Rogers, R.W. (1983). Cognitive and physiological processes in fear appeals and attitude change: A revised theory of protection motivation. In J.T. Cacioppo & R.E. Petty (Eds.), Social psychophysiology: A sourcebook. New York: Guilford.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1395
O V E R V I E W OF I M A G E 2.0: A N I N T E G R A T E D M O D E L O F C L I M A T E
CHANGE AND THE GLOBAL ENVIRONMENT J. Alcamo, C. Battjes, G. J. Van den Born, A. F. Bouwman, B. J. de Haan, K. Klein Goldewijk, O. Klepper, G.J.J. Kreileman, M. Krol, R. Leemans, J. G. van Minnen, J.G.J. Olivier, H.J.M. de Vries, A.M.C. Toet, R.A. van den Wijngaart, H.J. van der Woerd, G. Zuidema National Institute of Public Health and Environmental Protection (RIVM); P.O. Box 1; 3720 BA Bilthoven; The Netherlands 1. INTRODUCTION This purpose of this paper is to present a brief overview of the IMAGE 2.0 model, a multi-disciplinary, integrated model designed to simulate the dynamics of the global society-biosphere-climate system (Alcamo et a l . , 1994a). The paper emphasizes the scientific aspects of the model, while another paper in this volume emphasizes its policy aspects (Alcamo, et al., 1995). The objectives of IMAGE 2.0 are to investigate linkages and feedbacks in the global system, and to evaluate consequences of climate policies. Dynamic calculations are performed to year 2100, with a spatial scale ranging from grid (0.50 x 0.50 latitude-longitude) to world political regions, depending on the sub-model. A total of 13 submodels make up IMAGE 2.0, and they are organized into three fully linked subsystems: Energy-Industry, Terrestrial Environment, and Atmosphere-Ocean. The fully linked model has been tested against data from 1970 to 1990, and after calibration can reproduce the following observed trends: regional energy consumption and energy-related emissions, terrestrial flux of carbon dioxide and emissions of greenhouse gases, concentrations of greenhouse gases in the atmosphere, and transformation of land cover. The model can also simulate current zonal average surface and vertical temperatures. 2.
ENERGY INDUSTRY SUBSYSTEM The purpose of the Energy-Industry subsystem of the IMAGE 2.0 model is to develop global scenarios of regional emissions of greenhouse gases based on future estimates of regional energy use and industrial production (de Vries et a l . , 1994). The subsystem consists of four linked models. The Energy-Economy model computes regional energy consumption with a special emphasis on final energy consumption in end-use sectors, based on economic activity levels and energy conservation potential. The Industrial Production and Consumption model estimates future levels of industrial output in sectors that are important emitters of greenhouse gases. These two models are complemented by two other models that compute the complete range of emissions of greenhouse gases and ozone precursors, based on emission factors per compound and per activity. A wide range of emission control strategies can be specified by adjusting technological and economic variables in the subsystem of models. For a baseline "Conventional Wisdom" scenario*, global CO2 emissions from energy and industry increase from 6.1 Pg C a-1 in 1990 to 16.7 Pg C a-1 in 2050,
1396 and to 24.2 Pg C a-1 in the year 2100 (de Vries et al., 1994). Global emissions of CH4, N20, and CO sharply increase because of increased emissions from industrial processes. The uncertainty of baseline CO2 estimates was investigated in a detailed mathematical analysis of parameter uncertainty of the model. Based on this analysis, the 90% confidence interval of computed CO2 emissions (year 2050, Conventional Wisdom scenario) was estimated to be 11.3 to 24.4 Pg C a-1 (de Vries et al., 1994). 3.
THE TERRESTRIAL ENVIRONMENT SUBSYSTEM The purpose of the Terrestrial Environment subsystem is to simulate changes in global land cover on a grid-scale based on climatic and economic factors. This subsystem also estimates the fluxes of C02 and other greenhouse gases between the biosphere and atmosphere. The subsystem consists of five linked models, covering agricultural demand, terrestrial vegetation, land cover, land use emissions, and terrestrial carbon flux (Figure 1). The Land Cover model simulates land cover transformations on a global grid by reconciling the regional demand for land with the local potential for land (Zuidema et al., 1994). The regional demand for land comes from demands for cropland and rangeland computed by the Agricultural Demand model (Zuidema et al., 1994), and for fuelwood from the Energy Economy model (not part of the Terrestrial Environment subsystem). The potential for land is estimated by the Terrestrial Vegetation model which computes potential crop yield and potential natural vegetation based on climate and other environmental factors (Leemans and van den Born, 1994). Once land cover and its conversion rate are computed, this information is used by the Land Use Emissions model to compute the flux of methane and other greenhouse gases from land use activity (Kreileman, and Bouwman, 1994), and by the Terrestrial Carbon model to compute the flux of CO2 from biomass burning, soil respiration, and plant productivity (Klein Goldewijk, et al., 1994). Model calculations have been tested against data from 1970-90 for crop production, vegetation cover, and country-scale deforestation rates. For a baseline Conventional Wisdom scenario, it was found that there could be very large differences in future trends in land cover changes between regions (Zuidema et al., 1994). However, deforestation rates in all regions rapidly diminish after the middle of next century because demands for land stabilize or because forests are depleted. For the same scenario, the model computes that the terrestrial biosphere acts as a strong carbon sink in the 21st century because of forestation of abandoned agricultural land in the Northern Hemisphere and global feedbacks to vegetation (Alcamo, et al., 1994b).
4.
THE ATMOSPHERE-OCEAN SUBSYSTEM The purpose of the Atmosphere-Ocean subsystem of IMAGE 2.0 is to compute dynamic changes in greenhouse gases and resulting changes in global temperature and precipitation patterns. The basic idea of this subsystem is to compute transient changes in climate, and to do it in a way that is computationally efficient. This is a necessary condition for an integrated model. Zonal average changes in climate are computed, and these are downscaled to a global grid using results from general circulation models. The four components of the subsystem are: atmospheric composition (Krol and van der Woerd, 1994), atmospheric climate,
1397 ocean climate, and ocean biosphere/chemistry (de H a a n et al., 1994). The model has been tested against field data of atmospheric chemistry, long-term climate patterns, and data from general circulation models. For t h e b a s e l i n e C o n v e n t i o n a l Wisdom scenario, the a t m o s p h e r i c CO2 concentration increases up to 777 ppmv by the end of the next century. By comparison, CH4 stabilizes in the atmosphere because of stabilized emissions of carbon monoxide, a precursor compound of CH4. Under this scenario, the average surface t e m p e r a t u r e of the world's oceans increases about 10C from 1990 to 2100. For air temperatures, the average increase in this period is 2.50C, and ranges from 3 to 50C in the northern latitudes. 5.
S O M E S C E N A R I O S OF C L I M A T E C H A N G E A N D T H E G L O B A L ENVIRONMENT IMAGE 2.0 has been used to compute a range of comprehensive scenarios about climate change and the global environment (Alcamo eta[., 1994b). Above we have described results from the baseline Conventional Wisdom scenario. Here we briefly describe results from two other types of scenarios.
Biofuel Crops and No Biofuels Scenarios. The Biofuel Crops scenario assumes t h a t the usage of biofuels increases from its present low levels to 74 EJ/yr in year 2050, and 208 EJ/yr in 2100. It is further assumed t h a t 40% of total biofuel demand will be delivered by energy crops, with the r e m a i n d e r coming from crop residues and other sources. The No Biofuels scenario is used as a b e n c h m a r k to study the sensitivity of the global system to biofuel use. These scenarios confirm t h a t biofuel use can be a successful strategy for lowering CO2 emissions. However, they also show t h a t emissions of other i m p o r t a n t compounds, such as carbon monoxide (CO), could increase. Hence it is i m p o r t a n t to examine the impact of biofuels on the full range of greenhouse gases, r a t h e r than only CO2. Another finding of the scenarios is that land needed for energy crops could compete with land needed for food crops in Africa and Asia. Moreover, the conversion of n a t u r a l vegetation to energy cropland could m a r k e d l y reduce the uptake of CO2 by the terrestrial biosphere (Alcamo et al., 1995).
Ocean Realignment The Ocean R e a l i g n m e n t scenario investigates the possible global effects of a surprising change in n a t u r a l driving forces, namely, the slowing down of ocean circulation. This scenario aims to look at the possible consequences of a low probability occurence. The assumed changes in ocean circulation lead to a slowing of the n o r t h w a r d t r a n s p o r t of heat in the Atlantic, and a t e m p o r a r y cooling of ocean and air t e m p e r a t u r e in the N o r t h e r n Hemisphere. E v e n t u a l l y surface t e m p e r a t u r e s in the North increase because of the build-up of greenhouse gases, but the increase between 1990 and 2100 is much lower (1.50C) t h a n in the baseline scenario (3 to 5 0C). This smaller increase in t e m p e r a t u r e reduces carbon u p t a k e in the n o r t h e r n biosphere, as compared to the baseline. Consequently, under this scenario, the CO2 level in the atmosphere is 90 ppm higher in the year 2100 t h a n in the baseline. Results of the scenario illustrate t h a t an unexpected, low probability event can both enhance the build-up of greenhouse gases, and at
1398 the same time cause a temporary cooling of surface air temperatures in the Northern Hemisphere. 6.
CONCLUSIONS In summing up, the main innovation of the IMAGE 2.0 model is its presentation of a g e o g r a p h i c a l l y - d e t a i l e d , global, and dynamic view of the linked society-biosphere-climate system. With respect to society, the model represents in some detail the relation between economic and demographic trends and the generation of greenhouse gas emissions. Regarding the biosphere, it is a first attempt to simulate in geographic detail the transformation of land cover as it is affected by climatic, demographic, and economic factors. And with respect to the climate system, it dynamically couples emissions from society and the biosphere with processes in the atmosphere and ocean. Because of its components and spatial resolution, the model is particularly well-suited to investigate both scientific and policy oriented questions. In particular, it has the potential to provide new insight into the linkages and feedbacks of the global-biosphere-climate system. 7. A C K N O W L E D G E M E N T S The IMAGE Project is supported by the Dutch Ministry of Housing, Physical Planning and Environment (VROM), and the Dutch National Program on Global Air Pollution and Climate Change (NRP). This paper was partly funded under NRP contracts 853129, 853130, 853131, and 853132. REFERENCES
Alcamo, J., G.J.J. Kreileman, M. Krol, and G. Zuidema: 1994a, Modeling the global society-biosphere-climate system, Part 1: model description and testing. Water, Air, Soil Pollution, 76(1-2): 1-36. Alcamo, J., G.J. van den Born, A.F. Bouwman, B.J. de Haan, K. Klein Goldewijk, J. Krabec, O. Klepper, R. Leemans, J.G.J. Olivier, A.M.C. Toet, de Vries, H . J . M . , a n d H. v.d. Woerd: 1994b, Modeling t h e global society-biosphere-climate system, part 2: computed scenarios. Water, Air, Soil Pollution, 76(1-2): 37-78 Alcamo, J. Krol, and R. Leemans: 1995, Stabilizing greenhouse gases: global and regional consequences, this volume. de Haan, B.J., M. Jonas, O. Klepper, J. Krabec, M.S. Krol, K. Olendrzynski: 1994, An atmosphere-ocean model for integrated assessment of global change. Water, Air, Soil Pollution, 76(1-2): 283-318. Klein Goldewijk, K., J.G. van Minnen, G.J.J. Kreileman, M. Vloedbeld, and R. Leemans: 1994, Simulating the carbon flux between the t e r r e s t r i a l environment and the atmosphere. Water, Air, Soil Pollution, 76(1-2): 199-230. Kreileman, G.J.J. and A.F. Bouwman: 1994, Computing land use emissions of greenhouse gases. Water, Air, Soil Pollution, 76(1-2): 231-258. Krol, M.S. and H.J. van der Woerd: 1994, Atmospheric composition calculations for evaluation of climate scenarios. Water, Air, Soil Pollution, 76(1-2): 259-282.
1399 Leemans, R. and G.J. van den Born, G.J.: 1994, Determining the potential distribution of natural vegetation, crops, and agricultural productivity. Water, Air, Soil Pollution, 76(1-2): 133-162. de Vries, H.J.M., J.G.J. Olivier, R.A. van den Wijngaart, G.J.J. Kreileman, and A.M.C. Toet: 1994, A model for calculating regional energy use, industrial production and greenhouse gas emissions for evaluating global climate scenarios. Water, Air, Soil Pollution, 76(1-2): 79-132. Zuidema, G., G.J. van den Born, J. Alcamo, and G.J.J. Kreileman: 1994, Simulating changes in global land cover as affected by economic and climatic factors. Water, Air, Soil Pollution, 76(1-2): 163-198. Endnote
*The baseline scenario is based on the Conventional Wisdom scenario documented in: Alcamo, J., van den Born, G.J., Bouwman, A.F., de Haan, B., Klein Goldewijk, K., Klepper, O., Leemans, R., Olivier, J.A., de Vries, B., van der Woerd, H. and van den Wijngaard, R., 1994b. Modeling the global society-biosphere-climate system, Part 2: computed scenarios. Water, Air and Soil Pollution, 76: 37-78. This scenario takes population and economic growth assumptions from the intermediate emissions scenario (IS92a) of the IPCC (1992). The population assumptions correspond to median estimates of the U.N. F u r t h e r assumptions of the Conventional Wisdom scenario are given in Alcamo, et al., Ibid. ENERGY.INDUSTRYSYSTEM
Output "missions Gasesl ATMOSPHERE.OCEANSYSTEM
Input Economic Technological Demographic
ntrolPolicies
il Energy L_.J Energy
Climate
TERRESTRIAL ENVIRONMENTSYSTEM ::: ::
LFeedbacks
II
AYnosphedc~ Zonal Composition Atmospheric
:~@~,'~i" ~ $'' :::~:''~ :"~ ' u162 ~ ~ ~ b-%..::
:~ ~:::::: ~ .~::
!il l~..".:!i I~ ::ii~i~i##~f.:!~ii~i#.'~,~.~"..;%iLand ~..*.,:: Jss ~i!
iI!."~nd!iiiii!!i[~ ~i!i|~!i~f~= i'miss i!ili!~!ioi~ns ii Cover ~!
~.~.-.'~!i~l . !~:,
Biosphere
M
dChemLsW
"Greenhouse Ozone-related gases
- Atmosph.C~c.
t
"i~!:
7
9GreenhouseGases~,. "Ozone-related | _gases __.j
Ocean
Cllmate
EChangedClimate ~ Emi~sior~
I
7
"Greenhouse GasesI
Ehe. LandUse Pattern ii!
| _j
..~
:.:
Figure 1. Box diagram of IMAGE 2.0 model. Each box represents a submodel of IMAGE 2.0.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1401
Uncertainty management in integrated modelling, the IMAGE case Jeroen P. van der Sluijs Department of Science Technology and Society, Utrecht University, Padualaan NL-3584 CH Utrecht, The Netherlands (E-mail:
[email protected])
14,
Abstract Integrated assessment models of global environmental problems play an increasingly important role in decision making. This use demands a good insight regarding the reliability of these models. In this paper we analyze uncertainty management in the IMAGE-project (Integrated Model to Assess the Greenhouse Effect). We use a classification scheme comprising type and source of uncertainty. Our analysis shows reliability analysis as main area for improvement. We briefly review a recently developed methodology, NUSAP, that systematically addresses the strength of data in terms of spread, reliability and scientific status (pedigree) of information. This approach is being tested through interviews with model builders.
1. I N T R O D U C T I O N Facilitated by developments in computer technology, "integrated modelling" emerges in the mid eighties as a new approach to interface science and policy concerning complex environmental issues. For instance, in the 1986 RIVM annual report vice director general of the ministry of VROM Kees Zoeteman argues that RIVM can give optimal shape to its role as interface between science, policy and monitoring by means of integrated modelling. To give an other example: the RAINS model (Regional Acidification INformation and Simulation), developed in the eighties at the International Institute for Applied System Analysis (IIASA), was used in the international acid deposition negotiations and became an annex of the SO2-protocol. For the climate problem there are at least 16 different integrated assessment models in active use or under active development (Weyant, 1994). In this paper we focus on the IMAGE model, developed at RIVM, which started as a pioneer in this field. IMAGE has been used for scenario calculations in the influential policy document "Zorgen voor Morgen" (Concern for Tomorrow) (Langeweg, 1988) and for the development of emission scenario's for IPCC working groups II and III, the latter in combination with the Atmospheric Stabilization Framework of the US Environmental Protection Agency (Swart, 1994a). The revised version of the model, IMAGE 2, is currently being used in all three working groups of IPCC. For the future it is likely that integrated models will be used to support the negotiations during the United Nations Conference of Parties to the Climate Convention in Berlin, March, 1995, and follow up (Swart, 1994a; Alcamo, 1994b).
1402 The inherent uncertain character of the knowledge on future climate and the poor scientific understanding of the geosphere biosphere system combined with the high decision stakes associated with policy choices supported by these models, demand a good insight in and a high awareness of the quality, the reliability and the limitations of the model. According to Swart (1994a), the uncertainties in the knowledge of climate change are so large that climate scientists sometimes claim that an integrated approach is not meaningful in the case of climate change. The model builders themselves oppose this view by stating that "while there is a great uncertainty regarding our future, we have a certain responsibility to take our best scientific understanding, and use that to develop reasonable policies" (Alcamo, 1994b). Such controversies imply that the question of uncertainty
management is becoming increasingly important. In this study we investigate for the IMAGE case how scientific uncertainties are being managed and we analyze the scope of the current practice of uncertainty management, using a two dimensional classification scheme comprising type and source of uncertainty.
2. U N C E R T A I N T Y M A N A G E M E N T IN THE IMAGE PROJECT
On the basis of literature study and interviews with the model builders, we have made an inventory of the way questions of uncertainty and quality have been and are being addressed in the IMAGE project. The first version of IMAGE was developed in the period 1985-1990 (De Boois and Rotmans, 1986; Rotmans 1990). In this period climate change started being signalized as a policy issue. IMAGE was even used to put the issue on the policy agenda and the model was developed despite initial lack of interest in such a model of policy makers (Rotmans, 1994; Swart, 1994b). IMAGE is designed as a deterministic model. The treatment of uncertainty has not been explicitly considered in the design of the model (Weyant, 1994). This might be because the issue of uncertainty management was less urgent at that time than it is nowadays. Despite these circumstances, sensitivity analysis and uncertainty analysis have been carried out from the very beginning. In the beginning, IMAGE suffered a lot of criticism from scientists, who thought that the approach was far too simplified (Rotmans, 1994). The scientific status of IMAGE has increased significantly since the IMAGE team started to publish their work in high impact scientific journals (e.g. Science Citation Index, 1991) such as Climatic Change (Rotmans, 1994). In 1992 Joe Alcamo became project leader of IMAGE, bringing the experience from the development of the RAINS model. Since then several changes have become visible: a complete revision of the model; improved communication with policy makers and scientists and improved exposure to peer review (e.g. the complete IMAGE 2.0 model was published as a special issue of the journal "Water Air and Soil Pollution" (Alcamo, 1994a)). Thorough sensitivity analyses and uncertainty analyses have been carried out for several sub-models of IMAGE 2 (e.g. Krol and van der Woerd, 1994). The technique used is Latin Hypercube Sampling, which is an advanced Monte Carlo based tool to map the relative contribution of specified uncertainty sources to the spread in model output and to assess the propagation of uncertainties through the model. A special software package for this purpose, UNCSAM (UNCertainty analysis by Monte Carlo SAMpling techniques) was developed (Janssen et al., 1994). The complete inexactness-uncertainty of the model has not been computed yet. A major problem in determining total inexactness uncertainty is
1403 the identification of the spread and distribution functions of all input data and model parameters (Rotmans, 1994). Variables that are highly uncertain (such as population growth) are managed by making them scenario variables. The completeness and the quality of the IMAGE 2 model have been addressed by two "international review meetings" (Hordijk, 1993). An other strategy to improve quality is to include the best science available (Alcamo, 1994b). This means that the model is often being revised in dialogue with scientists. This is also reflected by the attempts to improve the integration of IMAGE and the NRP-research (National Research Program on Global Air Pollution and Climate Change) (Berk, 1993). Other sources of error, however, such as numerical artefacts in model calculations and errors due to numeral approximation are not assessed. According to Hordijk (1994), who agrees that this is an omission, even the NRP is not interested in the mathematical tour de force required to assess these errors.
3. A C L A S S I F I C A T I O N S C H E M E F O R U N C E R T A I N T Y According to its source, scientific uncertainty can be classified as: (1) data uncertainties that arise from the quality or appropriateness of the data used as inputs to models; (2) modelling uncertainties that arise from: (a) incomplete understanding of the modelled phenomena or (b) numeral approximations used in mathematical representation; and (3) completeness uncertainties covering all omissions due to lack of knowledge (Vesely and Rasmuson, 1984). Funtowicz and Ravetz (1990) have given a classification of types of uncertainty: (i) inexactness (significant digits/error bars); (ii) unreliability; (iii) border with ignorance. The combination of both classifications produces a two dimensional classification scheme defining areas to be addressed in uncertainty management in integrated models. We have applied this scheme to the findings of the previous section to indicate the scope of the current practice of uncertainty management in IMAGE (Table 1). The table shows that the best covered area are inexactness-uncertainties in input data and parameters. The reliability of the input data and model structure are not systematically addressed. Other omissions are that possible numerical artefacts in model calculations and errors due to numeral approximation are not assessed. This area can be addressed by a
Table 1 Two dimensional classification scheme for uncertainty, showing the scope of uncertainty analysis efforts on IMAGE input data ""-,,•urce type ~
parameters
model structure relations
inexactness systematicallyaddressed for some sub models unreliability (passively via peer review of publications)
systematicallyaddressed not addressed for some sub models
ignorance
(passively via "include (passively via "include best science available") best science available")
(passively via "include best science available")
(passively via peer review of publications)
model completeness Explicitly addressed by advisory board
Explicitly addressed Explicitly addressed by advisory board by advisory board; (passively via peer review of publications) (passively via "include best science available")
1404 mathematical tour de force which lies beyond the scope of this paper. Uncertainty resulting from ignorance is not addressed too. Ignorance and completeness uncertainties are the most difficult to address. In fact they can only be addressed indirectly via quality control procedures, such as peer review, of the production process of the scientific information used in the model.
4. NUSAP AND THE PEDIGREE MATRIX
Funtowicz and Ravetz (1990) have designed an innovative tool to address the issue of uncertainty and reliability: the NUSAP (Numeral Unit Spread Assessment Pedigree) notational scheme for scientific information. NUSAP is designed to act as a heuristic for good scientific practice and as a system for expressing and communicating uncertainties. It consists of five qualifiers: Numeral, Unit, Spread, Assessment and Pedigree. The last three qualifiers address the various aspects of uncertainty. The spread qualifier conveys an indication on the inexactness of the numeral and unit places. The assessment qualifier should express a judgement on the reliability of the three previous qualifiers, it is a measure for the strength of the data. Pedigree conveys an evaluative account of the production process of the information, and can be seen as a measure for scientific status of the associated knowledge. Because many aspects are relevant in evaluating the production process, a matrix is used to represent pedigree. Depending on its application a pedigree matrix consists of a set of suitable evaluation criteria (e.g. peer acceptance), and defines modes of these criteria (e.g. low, high) which are coded hierarchically. A pedigree matrix for research is given in Table 2. For instance the CO2-fertilization parameter in the IMAGE 1 model was based on a theoretically based model, experimental data, faced medium peer acceptance and its value was subject to competing schools, yielding a research pedigree (3,4,2,2). The Second Law in thermodynamics has a pedigree (4,4,4,4). The pedigree matrix enables to compare strength of data in terms of the applied criteria and brings to light the weak parts of the model. This also helps in priority setting for model improvement.
Table 2 The pedigree matrix for research as designed by Funtowicz and Ravetz (1990) Code
Theoretical Structure
Data-input
Peer acceptance
Colleague consensus
4 3 2 1 0
Established theory Theor. based model Computational model Statistical processing Definitions
Experimental data Historic/field data Calculated data Educated guesses Uneducated guesses
Total High Medium Low None
All but cranks All but rebels Competing schools Embryonic field No opinion
Funtowicz and Ravetz also proposed a pedigree matrix for environmental models (Table 3). When tested in our interviews, the hierarchy in the columns of this pedigree matrix proved to be controversial: according to Alcamo (1994b), there is no 'good' or 'bad' in
1405 model structure. His alternative ranking of the modes for data input and testing is given between brackets in Table 3. With respect to model structure it is obvious that a black box model has a lower scientific status (if any at all) than a model that is completely governed by established physical laws and has a high process detail. However, if a very simple meta model is derived from, and secured by, a complex model with high process detail, the simple meta model can be equally good to model the process. For these cases we propose to apply the pedigree matrix to the mother model while taking into account the consequences of the simplifications in the meta model. It is not surprising that model builders use other criteria to evaluate model quality than Funtowicz and Ravetz do. They have different critical roles (compare Clark and Majone, 1985). The model builder has to fulfil the needs of policy makers without compromising too much the scientific credibility of the model. This results in usefulness as the main quality criterium (Mermet and Hordijk, 1989; Swart, 1994a).
Table 3 The pedigree matrix for environmental models as designed by Funtowicz and Ravetz (1990). Between brackets: alternative ranking codes, attributed by a model-builder. Code
Model structure
Data input
4 3 2 1 0
Comprehensive Finite-element approximation Transfer function Statistical processing Definitions
Review Historic/field Experimental Calculated Expert guess
Testing (1) (1) (1) (1) (0)
Corroboration Comparison Uncertainty analysis Sensitivity analysis None
(2) (1) (1) (1) (0)
5. C O N C L U S I O N S The analysis of uncertainty management in the IMAGE project shows as main area of improvement the assessment of reliability of the input data and model structure. The NUSAP notational scheme for scientific information as designed by Funtowicz and Ravetz (1990) is a tool to systematically address the issues of reliability and quality. Model builders use other evaluation criteria for model quality than Funtowicz and Ravetz. These differences can be understood from their differing critical role. They are however no obstacle to using the NUSAP methodology as a guiding checklist to identify weak parts of the model in terms of the applied criteria.
6. F U R T H E R R E S E A R C H
Further research is needed to identify suitable sets of evaluation criteria for pedigree matrices accommodating different critical roles. An other important area for research is the development of a sensible way to represent and communicate the information on uncertainty, reliability and limitations of a model in a form comprehensible to the users and minimizing ambiguity regarding its interpretation.
1406 7. R E F E R E N C E S
J. Alcamo (ed), IMAGE 2.0: Integrated Modeling of Global Climate Change, in: Water, Air, and Soil Pollution, 76, Nos. 1/2, 1994a. J. Alcamo, personal communication (interview) 21 november 1994b. M. Berk, Integratie van NOP-onderzoek en het IMAGE model, Report of a NRP Studyday, Utrecht, 22 January 1993. W.C. Clark and G. Majone, The Critical Appraisal of Scientific Inquiries with Policy Implications, in: Science Technology and Human Values, 10 (3), 6-19, 1985. H. de Boois and J. Rotmans, Overzichtsmodel van de CO2-Problematiek, in: Berichten uit het RIVM 1985, Bilthoven, 1986. S.O. Funtowicz and J.R. Ravetz, Uncertainty and Quality in Science for Policy, Kluwer, Dordrecht, 1990. L. Hordijk (ed.), Report International Review Meeting IMAGE 2.0 11-13 January 1993 Amsterdam, NRP report 00-09, Bilthoven, 1993. L. Hordijk, personal communication (interview), 15 november 1994. P.H.M. Janssen, P.S.C. Heuberger and R. Sanders, UNCSAM: a Tool for Automating Sensitivity and Uncertainty Analysis, in: Environmental Software, 9, 1-11, 1994. M.S. Krol and H.J. van der Woerd, Uncertainty Analysis for the Computation of Greenhouse Gas Concentrations in IMAGE, in: J. Grasman and G. van Straten (eds), Predictability and Nonlinear Modelling in Natural Sciences and Economics, Kluwer, Dordrecht, 1994. F. Langeweg, Zorgen voor Morgen, Nationale Milieuverkenningen 1985-2010, Samson H.D. Tjeenk Willink, Alphen aan den Rijn, 1988. L. Mermet and L. Hordijk, On Getting Simulation Models Used in International Negotiations: a Debriefing Exercise, in: F. Mautner-Markhof (ed), Processes of International Negotiations, Westview Press, Boulder CO, 205-237, 1989. J. Rotmans, IMAGE An Integrated Model to Assess the Greenhouse Effect, (Thesis), Rijksuniversiteit Limburg, 1990. J. Rotmans, personal communication (interview), 9 november 1994. Science Citation Index, Journal Citation Reports, A Bibliometric Analysis of Science Journals in the ISI database, 1991. R. Swart, Climate Change: Managing the Risks, (Thesis), Free University Amsterdam, 1994a. R. Swart, personal communication (interview), 23 november 1994b. W.E. Vesely and D.M. Rasmuson, Uncertainties in Nuclear Probabilistic Risk Analyses, in: Risk Analysis, 4, 313-322, 1984. J.P. Weyant, Integrated Assessment of Climate Change: A Overview and Comparison of Modeling Approaches, paper for the writing team 6/7 of working III Intergovernmental Panel on Climate Change Lead Authors Meeting, Geneva, September 8-10 1994. B.C.J. Zoeteman, Verslag van de Directeur van Hoofdsector III: Chemie en Fysica, in: Berichten uit het RIVM 1986, Bilthoven, 1987.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1407
Linking IMAGE 2 and WORLD SCAN George Gelauff, Ben Geurts, Arden Gielen, Adri den Ouden a Joe Alcamo and Reyer Gerlagh ~ a Central Planning Bureau (CPB), P.O. Box 80510, 2508 GM The Hague, The Netherlands b National Institute of Public Health and Environmental Protection (RIVM), E O. Box 1, 3720 BA Bilthoven, The Netherlands
Abstract This paper presents the links between the climate model IMAGE 2 and the economic model WORLD SCAN, which are set up to obtain an integrated scenario instrument for comprehensive and consistent climate-economy scenarios. The links are made with respect to energy (in WORLD SCAN) and agriculture (in IMAGE 2), thus providing a consistent linkage with feedbacks running both ways.
1
INTRODUCTION
In March 1994, the CPB and the RIVM started the joint NRP-1 project 853139 "Linkage of WORLD SCAN and IMAGE-2 models". Purpose of the project was to devise an analytical tool for integrated comprehensive scenarios ofworld economic development, greenhouse gas emissions and climate change. It is a precondition for the follow-up project - which is in the running for NRP-2 funding - in which "greenhouse policies" are analyzed on the basis of two reference scenarios for world economic development, bringing to the fore an assessment of fundamental uncertainties regarding crucial economic developments. The two reference scenarios to be developed differ according to the rate of convergence of the world's regions, in terms of economic, political and social developments. Questions with respect to the lifestyles, the openness of economies, and technological catching up by less developed regions will be addressed in this framework. Greenhouse policies to be analyzed include regional and interregional tax measures and regulations to curb emissions of CO 2. This paper first provides an overview of both models in section 2, after which section 3 digresses on the adjustments in both models. Section 4 pulls together the threads and hints on the type of analysis to be made in the NRP-2 project, after which it concludes the paper.
1408 2.
THE TWO MODELS
As models always paint a simplified picture of the real world, W O R L D SCAN and I M A G E both provide a consistent treatment of a sub-system. The former is an elaborate world economic model remaining relatively blind for environmental feedbacks. The latter describes climate change processes in a consistent manner, while treating economics as an exogenous input. 2.1
WORLD SCAN 1 The CPB WORLD SCAN is a theory-based, multi-sector multi-region long-term world economic model, in which three basic paradigms of economic development are present. These paradigms coincide roughly with the Neo-Classical, the Keynesian and the Schumpeterian views. As most models in the climate discussions are Neo-Classical of nature, e.g. G R E E N 2, we will focus here on how WORLD SCAN deviates from the NeoClassical perspective. The central extension of the model in a Keynesian direction is the role of investment behaviour. The explicit investment decision is based on expectations about the future and is driven by unstable "animal spirits". Schumpeterian tensions on markets are modelled by letting stocks determine market outcomes and non-wage incomes. Moreover, returns on investment are uncertain due to the uncertainty with respect to future economic and technological developments. A characteristic feature of the model is that it exhibits short-term disturbances, which have their impact on longterm developments. It is the analysis of these crucial issues which is left out in more NeoClassical economic analyses of greenhouse policies. 2.2
IMAGE 2.03 The RIVM IMAGE 2 model is a multi-disciplinary integrated model designed to simulate the dynamics of the global society-biosphere-climate system. The objectives of the model are to investigate linkages and feedbacks in the system, and to evaluate consequences of climate policies. Dynamic calculations are performed to the year 2100, with a spatial scale ranging from grid (0.5 ~ x 0.5 ~ latitude-longitude) to world regional level, depending on the sub-model. The model consists of three fully linked sub-systems: Energy-Industry, Terrestrial Environment, and Atmosphere-Ocean. The fully linked model has been tested against data from 1970 to 1990, and after calibration c a n reproduce the following observed trends: regional energy consumption and energy-related emissions, terrestrial flux of CO 2 and emissions of other greenhouse gases, concentrations of greenhouse gases in the atmosphere, and transformation of land cover. The model can also simulate long term zonal average surface and vertical temperatures.
3.
THE LINKS
The two overlapping areas of the models are agriculture and energy. The first is affected directly by climate change and has feedbacks to economics, while for the second the major feedbacks run from economics to climate. The links are therefore embedded in the agriculture module of IMAGE and a new energy system in WORLD SCAN.
1409
3.1
Integrating Energy in WORLD SCAN In WORLD SCAN we have taken the same end-use approach as in IMAGE 2, by translating all demand for the energy services heat and electricity into demand for primary energy carriers. In contrast to IMAGE, throughout nested constant-elasticity-ofsubstitution (CES) functions are used, which allow cost-driven substitution between and within bundles of items. This runs via two channels: intermediate demand for energy services stemming from production and final demand stemming from consumption. Intermediate demand for energy is illustrated in the input-output matrix presented in table 1 below, where shaded areas indicate deliveries. We have created two intermediate sectors - electricity and other intermediates - which use raw materials and primary energy and deliver the energy services to the other sectors. As can be seen immediately, this results in a simple matrix with no intra-sectoral deliveries. The intermediate demand for primary energy is split into oil, natural gas, coal, and biomass.
Table 1
The Input-output Matrix of WORLD SCAN primary energy
raw materials
electricity
other intermediates
other sectors
primary energy
raw materials electricity
other intermediates other sectors
Next to intermediate demand, there is final demand for energy, which is divided into consumption and net exports, where net exports are the result of both consumption and intermediate demand for energy in other regions. Consumption demand is divided into energy and non-energy, with energy divided among heat and electricity. Heat for consumption consists of the fossil fuels, fuelwood (in LDC's) and commercial biomass. In figure i below, the nested structure of demand is presented. The trees illustrate the structure of demand for consumption and production. Every aggregate branches off according to a CES function. Left-side items in italics refer to links in the energy chain. All demand is fed back fully into markets for energy, products, labour and financial capital, and the resulting energy use is delivered to IMAGE 2 as input in the Energy-Industry model. Markets for energy are - for now - modelled with perfect competition regimes, with stocks playing a role for fossil fuels. Primary energy supply is modelled as production sectors with decreasing returns to scale and exogenous technological progress, reflecting the finite character of the resource base 4.
1410 An outline of demand structures in WORLD SCAN
Figure 1
Low-skilled L a b o u r - High-skilled Labour
LABOUR~ VALUE ADDED
Capital NON- PRIMARY~PRODUCTI ON E1 e c t r i ci ty Other Intermediates--
Low-skilled L a b o u r - High-skilled Labour
LABOUR ~ VALUE ADDED 1
Capital OTHER INTERMEDIATES
Raw Materials Coal N a tural G a s Oil Commercial Biomass--
NON- PRIMARY HEAT - -
capital Coal Natural Oil Uranium
ELECTRICITY Gas
HEAT--
Agriculture and F o o d s t u f f s - Consumer Goods Capital Goods Sheltered Sector Goods
NON-ENERGY~coNSUMPTION
E1 ec tri ci ty Coal Natural Oil
ENERGY Gas
HEAT--
Commercial Biomass-Fuelwood
3.2
Agriculture in IMAGE Transformations of global land cover are strongly related to changing land use for agricultural products, which require croplands, pasture land, rangelands, and managed forests. The purpose of the Agricultural Economic Demand (AED) Model is to estimate demand for agricultural products. It is developed to be included in the IMAGE 2.1 model within the Terrestrial Environment sub-system. The main driving forces are population growth and income (Gross Domestic Product). The AED Model covers physical as well as economic aspects of agricultural production and consumption. The conversion of economic data into physical data and vice versa is done in the valuation sub-module. WORLD SCAN computes regional consumption, production and trade of aggregated agricultural products. First, economic
1411 flows (in constant prices) are converted in the AED Model to physical flows, expressed as the demand for land (km2). Secondly, shares for twelve different food products (seven crops and five animal products) are computed. These are derived from utility maximization, given total available land for production. For the next period, allocated land in combination with quality indices for production are used to alter technical coefficients in the production functions of WORLD SCAN.
Figure 2
The linked agricultural module in IMAGE 2
Linked Agricultural Module in Image 2 WorldScan (Agricultural Sector) in $1990 US Cons
~
Trade
Production
] I ] I
I
[
Image 2 I (Food products) ] in kcal I ICons, I I Trade
I I
,L
Jr
Prod,]
I
I
[
(Allocation) in km2
II
Production
Im
(Aggre.gation) Indices
Quality of Land
4.
~
]
I
e2
[ I
I !
I
CONCLUDING REMARKS
With the instrument developed in NRP-1 we will develop two reference scenarios in the NRP-2 project. Economic, political and social characteristics of the reference scenarios have their effect for instance on the composition of food bundles determined in IMAGE and on the consumption structure in WORLD SCAN. Furthermore, economic development differs widely between the two scenarios letting growth, trade and sectoral structures be very different. This affects the energy inputs for IMAGE in a strong way, leading to characteristic scenario profiles. Subsequently, extensive policy analyses can be performed in the two reference scenario worlds.
1412 Concludingly, we may state that the project has resulted in a remarkable and powerful tool for integrated economy-climate scenario analysis. Especially the combination of the original economic features of WORLD SCAN with the accepted and well-founded climate analyses of IMAGE contributes to the distinct qualities of the intrument. When NRP-2 funding for the scenario project comes through, we are convinced that new economy-climate scenarios can be developed which can shape policy discussions on the interaction between climate change and economic development.
WORLD SCAN was developed as analytical tool for the scenario study "Scanning the Future" published by Sdu Publishers, Plantijnstraat, The Hague in 1992. We refer to CPB internal notes IV/94/30 - IV/94/32 for a detailed description of WORLD SCAN. GREEN is the GeneRal Equilibrium ENvironmental model of the OECD, which has played a major role in the first discussions on greenhouse policies. See for instance Working Papers No's 116 and 143 of the OECD Economics Department for a description of the model and its performance in the inter-comparison projects of the OECD and the Stanford University Energy Modelling Forum. We refer to J. Alcamo (Editor) "IMAGE 2.0 - Integrated Modeling of Global Climate Change", Kluwer Academic Publishers, Dordrecht, 1994. Technological progress is negative in these sectors, acting as a proxy for depletion. A resource depletion model is not included but will be, if relevant.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1413
Modelling land use dynamics by integrating biophysical and human dimensions (CLUE) Costa Rica 1973-1984. A. Veldkamp and L.O. Fresco Agricultural University Wageningen, Department of Agronomy, P.O. Box 341, 6700 AH Wageningen Phone:+31 (0)8370 82513/83074, Fax:+31 (0)8370 84575. E-mail: Tom Veldkamp@
[email protected] or Louise Fresco@
[email protected]
A b s t r a c t
As a pilot study potential biophysical and human land use drivers in Costa Rica were evaluated using multi-variate statistical methods in a nested scale analysis. The reconstructed land use drivers and their quantified effects on land use were applied within a dynamic framework CLUE (Conversion of Land Use and its Effects) to model land use dynamics in Costa Rica from 1973 to 1984. Our pilot study demonstrates that a land use/cover system can be described as a scale-dependent hierarchical system and that its dynamics can be satisfactorily modelled as functions of biophysical and human drivers.
1. Introduction Most land cover modification and conversion is now driven by human use, rather than natural change [1,2]. In general, land use is viewed to be constrained by biophysical factors such as soil, climate, relief and vegetation. On the other hand, human activities that make use of or change land attributes are considered as the proximate sources of land use/cover change. Interpretations of how such land use/cover driving forces act and interact is still controversial, especially with respect to the assessment of the relative importance of the different forces and factors underlying land use decisions in specific cases [2]. There is also an increasing need in global change research for more realistic and integrated modelling of land cover conversions as a result of land use changes. There are several problems related to such modelling: Dynamics of natural vegetation vs. agricultural use. Spatial scales, global vs. regional. Temporal scales, rapid agricultural changes vs. relatively slow climate changes. Idenfication of relevant land use drivers, socio/economic vs. biophysical. In order to investigate these problems, we selected Costa Rica with its diverse environment and rapidly growing population, as a first pilot study to attempt modelling land use related drivers and their temporal and spatial variability. Potential land use drivers for Costa Rica were identified and combined in a geo-referenced census [3-6] and biophysical [7,8] data set (Figs. 1). -
-
-
1414
Altitudes
Relief
Soil Drainage 9
.
=~iiiiiilNiiiiiiiiiiiiiiiiiiiiiiii!iii, ....9 9 ,~.,
i!"
.:
..
9 9
: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : :
....................
iiiiiii~:,::ii!iiii!i~ .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ........ ....
:
Urban Population 1973 i!iii! ..... i
::.. .. ::::
Urban Population 1984 :.::::::=::::
i .......
.............
::::::::::::::::::::::::::::::::::::::::::::::::::
~:.
::::::::::::::::::::::::::::::::;:::::::::::::::::....:::::::: i.;~[iiiiiiiiii!i!i!iiii?i!iiii!i!i!iiii!fii!ii!i!iiiiliiiiiiiiiii ........................................................................ .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . . . .. .................................................................................................................................................................................................................................. .
................................................................... :::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: ...................................................................... "!!iiiiiiiiiiili!i?!iiiiili!iiiiiiiiiii!iiiiiiiiig!iii!!?!iiiiiiii!i!iiii::
""
iiiiiiiiii?i!gliii:':i!ifi[!ii[iiglg!i!~ii!igiiiii~ii?g!ii[i!iiiiii[i!iiii..
iiii)!iiii1111i? ...... :s~!!iiiiiitiiiii;i~iiiitit;i;iii;iii;i;i;i;iiiii: 9
::i:::iiii!i~i?i::
i i :
......
iiiiii}!iiiiii}i
: "~'i! ~
':!!!iiii!i!iiiiiiiii!ii!!?!ili?iiiiii?!i!iii!i
i!ii!iii!i,.
9 :
....
::ili!i!!!iiiii!i!ii~:,::!iiiiiiiiiiii!ii!!i~izii!::
9
~'~',~,~:~
:~!ii!iii !iiiiiiiiii!iiiiiiiiiiii ::::::::::::::::::::::::::::::::::::
9
9 ..... ~!!!ii!iiii[iiiiiiiiiiii!i!ili ........................ .
.
:::::::::::::::::::::::: ........................... :::::::::::::::::::::::::: :::::::::::::::::::::: .......................
::::::::::::::::::::::::::..................
9
.
9 9149 9149 . . . . . . . . .9. . . . . . . . . .
::::::.::::::::::::::::::. ::::::::::::::::::::::::::
" : i ! i. i i.i i i. i i . . . . . . : ' i i i i i : i !
........ .,
..
Rural Population 1973
Rural Population 1984
...... :::::: ......
,ii!i,i!i!!}i?iii!!iliiiiiiiiiiiiiiii!ii i,,,,,: ,:i . . . .~. i. i. i. .i .i .i.i.i.i.!.i. i. !. .!.i.i.i.i.i.i.i. i. i. i. .i .i .i .i.~. .i .i .?.i. .. i. i. ~. ...i.i. i. i. i. .i .!.i.i.i.i.i.i.i. i. i. ~. .,.i i i i ~
;i;iiiii!iiiiiji!!iiiiiiiiii,, gi!iiiiiii,iiiiiiiiiLi ,,
9
i!ili!iii!ililiiiiiiiiii}}iiilff!!!!!!!!~iii!ili!iiiiiiii!~~iii~.: :
.......~.:i~,~:~'!%!~iii!iiiiiiiiiii?,iii.. 9
============================''"'""'"""'""
:
9 9
.
: .
:
~
m
~
:
:
~
:
:
: : : : : : : : : : : : : : : : : : : : : : l
l l l l l l
l
[ [ [ [ l l ~ [
[ [
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. 9
9
.
.
.
:
: : : i i ! g i i ! i l i. . . . . . . . ........ .
.
::iiiiiiii
~
....
.:.
i: ! :
....
:"
. : . ! !'"i~.i!i~;.i'~r:':".:~.~?~!~i!i?~!~i!~'.:!iii~
: :
:
ii
. . . . 9
.
: ::!!!ii!ii[i!tiiiiii!iii,, !!ii:!!i!i~iiiiiiiii~iii~
:
.....: :
.
i
"~.~'"
~::iiiii!!~iI ..i!ii~::! . ::::
Fig. 1" Some biophysical characteristics of Costa Rica: Relief (mountainous flat [black]), Altitude (sea level [white] to > 3000 m [black]), Soil (poorly drained [white] to extremely well drained [black]). Some human characteristics of Costa Rica: Rural population 1973 (low [white] to high population density [black]) and Urban population 1984 (low [white] to high population density [black]).
[white] to drainage and 1984 1973 and
1415
2. Results and discussion Potential drivers were evaluated applying multi-variate statistical methods in a nested scale analysis, using six artificial aggregation scales [9]. Spatial distributions of potential biophysical (annual temperature, relief/precipitation and soil) and human (urban and rural population) drivers were statistically related to the distribution of pastures, arable lands, permanent crops, natural and secondary vegetation, for 0.1 ~ (6') grid units and five artificially aggregated spatial scales. Multiple regression models describing land use/cover variability have changing model fits and varying contributions of biophysical and human factors, indicating a considerable spatial scale dependence of land use/cover and its drivers. The observation that for both years each land use/cover type has its own specific scale dependencies, suggests a rather stable scale dependent system. In Costa Rica two land use/cover trends between 1973 and 1984 can be discerned: 1) intensification in the urbanized Central Valley and its surroundings where agriculture is extended to steeper and less favourable soils due to a high population density; 2) land use expansion in remoter areas, where the extension of arable land and pastures increased at the cost of natural vegetation (mainly forest). This deforestation was not driven by land shortage. Most changes in agricultural land use from 1973 to 1984 seem to be driven by changes in population density and to be limited to specific biophysical environments [9]. To support future efforts to model land use/cover dynamics by its drivers, a dynamic framework to simulate Conversion of Land Use and its Effects (CLUE) was developed [ 10]. CLUE attempts to simulate land use conversion and change in space and time as a result of interacting biophysical and human drivers. The CLUE framework was applied for Costa Rica to construct a first multi-scale dynamic land use model (CLUE-CR) simulating changes in natural vegetation, pastures, arable land, permanent crops and a rest group simutaneously. Initial simualtions (for example Fig. 2, pastures and range land changes simulated for 10 years) demonstrate a plausible and realistic land use dynamics as steered and controlled by population and biophysical conditions.
3. Conclusions The project resulted in the development of concepts for handling the highly dynamic features of land use change and its drivers for a small country (Costa Rica) at different spatial scales [9]. An analysis of Costa Rican land use/cover system distribution and their dynamics at six different spatial scales demonstrated that the human/biophysical dimensions of land use/cover systems are scale dependent. Each land cover has its own specific set of human and biophysical scale related drivers. CLUE-CR simulations (Fig. 2) suggest that the integrated approach of CLUE to model land use/cover dynamics as a function of its drivers [10], will contribute to more realistic simulations of land use/cover changes. It can thus be finally concluded that a land use/cover system can be described as a scaledependent hierarchical system and that its dynamics can be satisfactorily modelled as being driven by biophysical and human factors.
1416
T=T o
T=T o+2
yrs
T=T o+4
....... . . . l u t ~ . ~
yrs
:
::
iiii~
ii::iilF
iii::
....
T = T o+8 yrs
T = To + 6 yrs
T = T o+ 10yrs ::::::::
!!ii !i::
..
::.. ....
Fig. 2: Initial CLUE-CR simulation (10 years) of pasture/rang land dynamics, controlled by population and biophysical drivers in 1973-1984. ( <5% cover [white] to >70% [black])
4. References
10
B.L. Turner II, R.H. Moss and D.L. Skole (eds.), Relating Land use and Global land cover change: a proposal for an IGBP-HDP Core project. IGBP report no 24 and HDP report no 5, 65 pp., 1993 R.A. Houghton, D.S. Lefkowitz and D.L. Skole, Changes in the landscape of Latin America between 1850 and 1985. I. Progressive loss of forests. Forest Ecology and Management (1991) 38:143-172. DGEC, Censo Agropecuario 1973. San Jos6, Costa Rica, 1976 DGEC, Censo National de poblaci6n 1973, San Jos6, Costa Rica, 1976 DGEC, Censo Agropecuario 1984. San Jos6, Costa Rica, 1987 DGEC, Censo National de poblaci6n 1984, San Jos6, Costa Rica, 1987 L.D. G6mez, (ed.) Vegetaci6n y clima de Costa Rica, San Jos6, Costa Rica. vol. 1 pp 328. and vol. 2 pp 118. 1985. H. Nuhn, Atlas preliminar de Costa Rica. San Jos6, Costa Rica 50 pp, 1978. A. Veldkamp and L.O. Fresco, Reconstructing land use drivers and their spatial scale dependence for Costa Rica (1973 and 1984). (submitted) A. Veldkamp and L.O. Fresco, CLUE: a conceptual model to study Conversion of Land Use and its Effects. Ecological Modelling. (in press)
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 1995 Elsevier Science B.V.
1419
The Finnish Research Programme on Climate Change SILMU P. Heikinheimo and M. Kanninen The Finnish Research Programme on Climate Change/SILMU, The Academy of Finland, P.O. Box 57, FIN-00551 Helsinki, Finland Abstract SILMU, which runs from 1990 to 1995, aims at studying climate change and its impacts. It also seeks to provide information to Finnish policy makers on adaptation and mitigation. The topics range from air chemisrty to sociology, and the total number of projects is 74. Interim evaluation of the programme was carded out in 1992. During the second half of SILMU, 10 % of the total budget (total: 14 Million ECU) has been devoted to programme integration.
1. INTRODUCTION The need for reliable information on the complex aspects of climate change and its related effects became increasingly evident in mid 1980's both among scientists and policy makers. Global climate change can have a substantial impact on the world's ecosystems and on human welfare and the development of rational and effective strategies for dealing with this problem called for a strong multidisciplinary research effort. Shortly after the first IPCC panel in 1988, the Finnish Ministry of Environment set up a committee to analyze current knowledge on climatic changes and to define research needs related to the issue. The committee concluded by proposing that a large-scale mulfidisciplinary national research programme on climate change should be initiated. The following goals were set in 1989 for the Finnish Research Programme on Climate Change (SILMU): 9 9 9 9
To To To To
increase our knowledge of climate change, its causes, mechanisms and consequences strengthen the research on climate change in Finland increase the participation of Finnish researchers in international research programmes prepare and disseminate information for policy makers on adaptation and mitigation.
The six year programme started in 1990, and its funding was allocated in the state budget to the Academy of Finland. The Finnish Acidification Project had just ended, and positive experiences from this multidisciplinary programme partly paved the way to SILMU.
1420 2. PROGRAMME ADMINISTRATION AND FUNDING The SILMU programme is led by a directing committee (DC) in which all the relevant ministries are represented. The programme is coordinated and the programme office is located at the Academy of Finland. The decisions concerning scientific planning and funding of projects are carried out by a scientific committee (SC), which is a joint committee of the Academy of Finland and The Delegation of Finnish Academies of Science and Letters. This committee also acts as The Finnish National IGBP Committee. SILMU has four subprogrammes: Atmosphere, Waters, Terrestrial Ecosystems and Human Aspects / Integration. Each subprogramme has its subpogramme leader and coordinator. The programme secretariat consists of project manager and 1-2 secretaries. Total financing for 1990-1995 is approximately 85 Million Finnish marks (14 Million ECU / 17 Million US $ ) and the funding is channeled through the Academy of Finland. During 1990, 40 research projects were started. Funding was granted for periods of 1-3 years. The ministries in DC were actively involved in formulating the goals of the programme and also in the selection process of the research projects. Additional projects were started during 1991-1994 and several of the initial projects were given funding for the period 1993-95 after scientific reviews. Total number of research projects funded through SILMU in 1990-95 is 74.
3. INTERIM EVALUATION OF SILMU IN 1992 The purpose of the interim evaluation was to provide a critical assessment of the ongoing research and to offer suggestions for a possible redirection in the research agenda for the second half of the SILMU programme (Hordijk et al., 1992). The Central Board of the Academy of Finland listed the topics for evaluation. The First Progress Report of SILMU had just been published (Kanninen and Anttila, 1992) as well as an integrating book "Muuttuva Ilmakeh~i" (The Changing Atmosphere) which was geared to the public and to the decision makers (Kanninen, 1992). Many of the research projects had only preliminary results because the programme was so young at that point. Mainly therefore, individual projects were not evaluated but the Evaluation Committee focused on the evaluation of the programme as a whole and on the evaluation of each subprogramme. Interim evaluation provided SILMU with several concrete suggestions some of which have been put into practice. Fostering of integration and cooperation were the two main messages in the evaluation report and they have been given more emphasis. The recommendations of the interim evaluation guided funding decisions for the last three years of SILMU. In additon, three new projects on economic impacts and policy questions as well as five integrating projects were started. The programme organization has been simplified. Workshops within subprogrammes as well as a national SILMU meeting in 1994 have helped in enhancing scientific cooperation between different research groups.
1421 4. PROGRAMME INTEGRATION
The work in SILMU's 74 projects takes place in seven universities and eleven research institutions throughout Finland and the topics range from air chemistry to sociology. The fact that the subprogrammes operated fairly separately from each other during the first years of the programme, maybe delayed the development of contacts between research groups in separate research areas/subprogrammes. Spontaneous cooperation and integration has, nevertheless, been taking place among scientists especially in research areas with cooperative traditions. For the second half of SILMU i.e. 1993-95, the subprogramme Human Aspects was renamed Integration & Human Aspects and 10 % of total funding of the SILMU programme was allocated to the integration of research results. As described earlier, the recommendations presented in the interim evaluation have been partly guiding programme integration. In 9 9 9 9 9
1994, five integrative research projects were started. scenarios development integrated assessment of the effects of climatic change on waters an analysis of climate change impacts on forests and forestry the GIS-system development the influence of the forest industries and the use of forest products on carbon balance
Four international SILMU conferences or workshops will be organized in Finland during 1995. The topics are as following: Climate Change, Biodiversity and Boreal Forest Ecosystems (Aug. 1995), Past Present and Future Climate (Aug. 1995), Nothern Peatlands and Climatic Change (Oct. 1995), Effects of Climate Change on waters on Boreal Zone (Dec. 1995). These meetings will serve programme integration, and the main results will be published as a series of books on climate change in boreal ecosystems. The planning process of the conferences has helped Finnish scientists identify some gaps in the current knowledge which has led to some new research efforts.
5. DISSEMINATION OF INFORMATION AS PART OF INTEGRATION Reporting in SILMU is taking place in many forms. Scientific publications in international journals, SILMU-progress reports (Kanninen and Anttila, 1992; Kanninen and Heikinheimo, 1994), final report and workshop proceedings (Carter et al., 1993; Kanninen, 1993; Heino, 1994) as well books aimed at the general public like "The Changing Atmosphere" all have their target groups. SILMU has used Internet as a means for delivering information since April 1994. The feedback in the net has been positive from the scientists. Another electronic approach has been the production of an educational multimedia on global change. It has been well received among scientists and teaching professionals, and the generation of ideas for a multimedia for SILMU's final reporting is in progress.
1422 Bridging the gap between policy makers and scientists has been a continuous effort in SILMU. Funding ministries were heavily involved in the first place in formulating the goals of the programme. Their active role has decreased after the majority of funding decisions was made. In 1991, when the negotiations of the Framework Convention on Climate Change were in process, a seminar for scientists and policy makers was set up to exchange information and to facilitate the formulation of Finnish policies on climate change (Anttila, 1991). Another meeting of similar type will be held in early 1995 in order to feed new research results to the policy making process. Towards the end of SILMU, the scientists will acquire more integrated knowledge for the presentation of policy-related assessments. It is expected that the interpretation of the information to the language of policy makers will be one of the most challenging integration tasks.
5. REFERENCES
Anttila, P. 1991. Ilmastonmuutos ja Suomi - kohti kansallista toimintastrategiaa. Suomen Akatemian Julkaisuja 4/91. Carter, T., Holopainen, E., Kanninen, M. (eds.). 1993. Techniques for developing climatic scenarios for Finland. Report of an International Workshop held in Espoo (Hanasaari), Finland, 2-4 June 1993. Publications of the Academy of Finland 2/93. ISBN 951-37-1282-6. Heino, R. (ed.). 1994. Climate variations in Europe. Proceedings of the European Workshop on Climate Variations held in Kirkkonummi (Majvik), Finland, 15-18 May, 1994. Publications of the Academy of Finland 3/94. ISBN 951-37-1484-5. Hordijk, L., Lange, M. Pietil~i, S. Routti, J. 1992. Interim Evaluation of The Finnish Research Programme on Climate Change. Publications of the Academy of Finland 6/92. ISBN 951-37-0991-4. Kanninen, M. (ed.). 1992. Muuttuva Ilmakeha. Ilmasto, luonto ja ihminen. ("The Changing Atmosphere"). VAPK kustannus. 163 p. ISBN 951-37-0832-2. Kanninen, M. (ed.). 1993. Carbon balance of world's forested ecosystems: towards a global assessment. Report of the IPCC Workshop held in Joensuu, Finland. 11-15 May 1992. Publications of the Academy of Finland 3/93. ISBN 951-37-1365-2. Kanninen, M. & Anttila, P. (eds.). 1992. The Finnish Research Programme on Climate Change. Progress Report. Publications of the Academy of Finland 3/92. ISBN 951-370846-2 (Out of print). Kanninen, M. & Heikinheimo, P. (eds.). 1994. The Finnish Research Programme on Climate Change. Second Progress Report. Publications of the Academy of Finland 1/94. ISBN 951-37-1413-6.
Climate Change Research: Evaluation and Policy Implications S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk, (Eds.) 9 Elsevier Science B.V. All rights reserved.
1423
CLIMATE CHANGE RESEARCH IN BULGARIA Antoaneta I otova, Ekaterina Koleva - National Institute of Meteorology and Hydrology, Bulgarian Academy of Sciences Blvd. Tsarigragdsko chausaee 66, Sofia 1784, Bulgaria INTRODUCTION Climate is traditionally one of the main fields or research interest and objects for study in Bulgaria so many investigations on its genesis and specific features are c a m e d out in the pant and present. Recently climate change research appem~ to be the most actual topic and it in in the centre of climatic studies. A major part of these studies are realized at National Institute of Meteorology and Hydrology (NIMH) b e c a m e of its essential role in collection and analysis of the basic climatic data for the cotmtry. In the paper brief description or the climate change research at NIMH is presented and the obtained results are summarized. DESCRIPTION OF CLIMATE CHANGE RESEARCH The two main types of study - general and concrete - are covered but in different extent. The general studies - greenhouse gas (GHG) emmsiona estimation, environmental pollution as climate forcing factor, etc. - are undertaken by researchers at NIMH in the last few years while concrete studies on the climate variability are carried out in the c o m e of last decades, Most of the concrete research examine s i r ~ e climate elements - air temperature, precipitation, sun-shine duration; some of studies are on complexes or climate elements like air temperature - humidity, and on climatic phenomena drought, extreme and unfavourable events, etc. The attention is f o a m e d mostly on air temperature and precipitation as well as to drought phenomenon considering their part in the climate s.vstem and their importance for present climate change processes both at local and regional (Balkan Peninsula) scale. METHODS In order to e ~ i n e climate variability in Bulgaria time series of the basic climate elements are treated using the following methods: smoothing by weighted (9 or 10) moving averages, rifting by polynoms (of 7th to 9th order), integral difference curves, Spearman and Mann-Kendal rank tests. The first two methods are more subjective while the second two methods provide more objective results. Both types of methods could be used for short and long-term trend analysis. It could be mentioned also the introduced variability indexes D - for temperature [Koleva, Iotova- 1994], and P for precipittiiion time series analysis [Koleva, Iotova: 1992] which make easier the task for trends identification in these series. SUMMARY OF THE OBTAINED RESULTS Before s u r n m a r i z ~ the obtained results it has to be pointed out the following specific features of the climate variation pattern in Bulgaria- the difference between the fiat and mountain parts of the country as well as seasonal differences. These features are determined by the country's very complex orography conditions: two relatively large fiats, few considerable for the small territory mountains (the Balkan Mountains, Vitosha, Rila, Pirin, the Rhodopes) and many closed or semiclosed plains. That is why the climate variability is usually studied separately for the flat and mountain parts as seasonal and annual characteristics of the climatic d e m e n t s are processed. The most slmmmrized results on the air temperature variations in Bulgaria are presented in [Koleva, Iotova. 1994 (under press)] where it is concluded that the Spearman rank test results in negative trend in July and positive - in January for the most stations, with the 95% significance level. The annual temperature shows positive trend in North Bulgaria and negative - in South Bulgaria. Similar results for the annual temperature are obtained by the Mann-Kendal test [Koleva, Iotova. 1994]. The
1424
last ones are compared with the corresponding figures for Bucharest, Prague and Northern Adriatic showing positive trends, as well as for Warsaw and Thessaloniki - showing negative trend, as "all these trends are statistically significant (95% level of confidence). So the global tendency for warming can reveal as temperature increase in someplace but, at the same time - temperature decrease in other one. It is interesting to note the good agreement between the results for air temperature and these for the sun-shine duration [Koleva, Iotova: 1992]: there is positive trend in the winter and negative - in the summer, but the trends in the sun-shine duration time series are not statistically significant. As it is mentioned above, the orography conditions in Bulgaria determine considerable differences between the climatic elements in tile flat and mountain parts of the country. It is true especially for precipitation so separate examination of the precipitation variability for the two orography types is carried out [Koleva: 1994; Koleva, Iotova: 1992]. Both statistical methods - smoothed curves and Spearmar~Mann-Kendal tests result in positive trend for January, and negative - for July and annual precipitation in the flat parts of Bulgaria. The spectral analysis is applied for the annual precipitation finding 4 and 6-7 year oscillations but not 1 l-year ones so that to be associated with the sunspot activity. In respect to the mountain parts the same methods result in similar pictures: the smoothed curves show a decreasitlg trend in recent two-three decades, and the Spearmml/Mann-KendaU tests find negative trends with 95% significance for the half of mountain time series. Drought phenomenon appears to be of great importance for Bulgaria, especially in the flat part of North Bulgaria (Danube Plain), where the continental type of climate occurs. In the last years considerable attention is paid in respect to drought and some results are already available [Koleva: 1994 (under press)]. It is round a tendency for dryness and frequent droughts in Danube Plain as a total lack of precipitation can occur in any month but the probabilities are very low in June and May. UsiI~ the Budyko's dryness ratio it is obtained that Danube Plain climate is insufficient moist sub-humid. On the other part, the variability of average precipitation from year to year expressed by the variability index, which ranks the years of the 20th century from driest to wettest one, shows ttmt the 1980-s is the driest decade in this century. In recent years drier than normal conditions persist in the examined region. Especially dry is May-September season. NATIONAL CLIMATE PROGRAMME In accordance with the WMO' World Climate Programme it is developed National Climate Programme (,NCP) ot the Republic ot Bulgaria in order to unite all efforts in this regard. Beside o1' NIMH as initiator and co-ordinator, some other Institutes of Bulgarian Academy ot Sciences Institute of Forestry., Institute of Geography - and Higher Institute of Forestry, Nikola Pushkarov Institute for Soil Science and Agroecology, etc. are involved in NCP. The primary objectives or NCP are the following creation of computerized climate data bank; research on possible climate change in Bulgaria as a reflection of global climate change; analysis and optimization of the weather and climate use as natural resources; improvement of the monitoring system rot climate and related environmental elements observation and analysis. The NCP's particular objectives include: forwarding to Bulgarian Government information and analyses on climate, climate change and related topics: providing the public and governmental bodies with data and research results which they need and require. The above objectives are realized through 9 concrete projects without special financing for them. Bulgarian NCP has to be actualized now - to do evaluation of what is done, what are the main results, and what new to be included in the light of the last international and national priorities. Bulgaria signed the Framework Convention on Climate Change (Rio de Janeiro, 1992), which is in force already, but it is still not ratified by Bulgarian Parliament. Everything necessary for the ratification is prepared by the Ministry of Environment and simultaneously corresponding activities are undertaken to develop Bulgarian National Programme for Actions in Response to Climate Change. In this way it will be met the Convention's and other international agreements' requirements.
1427
A n n e x 1: List of participants
R.M. v a n A a l s t N a t i o n a l Inst. of Public H e a l t h a n d E n v i r o n m e n t a l Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE N E T H E R L A N D S Tel: + 31 30 742884 Fax: + 31 30 287531 e-mail:
[email protected]
W.L.E. A a r t s University of A m s t e r d a m A m s t e r d a m School for Social Science Research Oude H o o g s t r a a t 24 1012 CE A m s t e r d a m THE N E T H E R L A N D S Tel: + 31 20 5252244 Fax: + 31 20 5252446
K.F. Albrecht Technische Universit~it D r e s d e n Inst. for Allgemeine Okologie u n d Umweltschutz Pienner Strasse 8 01737 T h a r a n d t GERMANY Tel: + 49 35203 37331, app. 309 Fax: + 49 35203 37495
J. Alcamo Na t i o n a l Inst. for Public H e a l t h a n d E n v i r o n m e n t a l Protection (RIVM) P.O. Box 1 3720 BA Bitlhoven THE N E T H E R L A N D S Tel: + 31 30 743487 Fax: + 31 30 250740 e-mail:
[email protected]
F.J.M. A l k e m a d e N a t i o n a l I n s t i t u t e of Public H e a l t h a n d E n v i r o n m e n t a l Protection P.O. Box 1 3720 BA Bilthoven THE N E T H E R L A N D S Tel: + 31 30 742331
A.R. v a n A m s t e l N a t i o n a l I n s t i t u t e of Public H e a l t h a n d E n v i r o n m e n t a l Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE N E T H E R L A N D S Tel: + 31 30 743780 Fax: + 31 30 293651 e-mail: an d r e .v a n ,
[email protected]
J.L.R.H. de Aratijo U n i v e r s i d a d e F e d e r a l de Rio de Janeiro Inst. de E c onom i a I n d u s t r i a l Av. P a s t e u r , 250, s a l a 11 Praia Vermelha Rio de J a n e i r o , R J 22290-270 BRAZIL
M.A. Arickx Limburgs U n i v e r s i t a i r C e n t r u m Universitaire Campus 3500 H a s s e l t BELGIUM Tel: +32 11 268695 e-mail:
[email protected]
1428
N.E.M. Asselman Utrecht University Dept. of Physical Geography Heidelberglaan 2 3584 CS Utrecht THE NETHERLANDS Tel: + 31 30 532167 Fax: + 31 30 540604 e-mail: n.
[email protected]
S. Azzali DLO - The Winand Staring Centre P.O. Box 125 6700 AC Wageningen THE NETHERLANDS Tel: + 31 8370 74323 Fax: + 31 8370 24812
A.P.M. Baede Royal Netherlands Meteorological Institute (KNMI) P.O. Box 201 3730 AE De Bilt THE NETHERLANDS Tel: + 31 30 206446 Fax: + 31 30 210407 e-mail:
[email protected]
D.C.E. Bakker Netherlands Institute for Sea Research (NIOZ) P.O. Box 59 1790 AB Den Burg (Texel) THE NETHERLANDS Tel: + 31 2220 69439 Fax: + 31 2220 19674 e-mail:
[email protected]
H.K. Bakker E r a s m u s University Rotterdam DVM P.O. Box 1738 3000 DR Rotterdam THE NETHERLANDS Tel: + 31 10 4087469 Fax: + 31 10 4363388 e-mail:
[email protected]
S. B a r a t h a n Tata Energy Research Institute 9, Jor Bagh New Delhi 110 003 INDIA
D.B. Barbieri Istituto di Ingegneria Civile ed Energetica Universit~ degli Studi di Reggio Calabria Via E. Cuzzocrea, 48 89128 Reggio Calabria ITALY Tel: + 39 965 875202 Fax: + 39 965 875254
N.H. Batjes International Soil Reference and Information (ISRIC) P.O. Box 353 6700 AJ Wageningen THE NETHERLANDS Tel: + 31 8370 71732 Fax: + 31 8370 24460 e-mail:
[email protected]
J.J. Battjes National Inst. of Public Health and Environmental Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743915 Fax: + 31 30 250740
J.P. Beck National Institute of Public Health and Environmental Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 742362 Fax: + 31 30 287531
1429
J.J. B e e r s m a Royal Netherlands Meteorological Institute (KNMI) P.O. Box 201 3730 AE De Bilt THE NETHERLANDS Tel: + 31 30 206589 Fax: + 31 30 210407
J.J.M. Berdowski T N O - MW P.O. Box 6011 2600 JA Delft THE NETHERLANDS Tel: + 31 15 696237 Fax: + 31 15 616812 e-mail:
[email protected]
F. Berendse Wageningen Agricultural University Dept. of Terrestrial Ecology and Nature Conservation Bornsesteeg 69 6708 PD Wageningen THE NETHERLANDS Tel: + 31 8370 84973 Fax: + 31 8370 84845
P.A.H.M. Berendsen NWO ESR P.O. Box 93120 2509 AC Den Haag THE NETHERLANDS Tel: + 31 70 3440813 Fax: + 31 70 3471623
H.J.A. Berendsen Utrecht University Dept. of Physical Geography Heidelberglaan 2 3508 TC Utrecht THE NETHERLANDS Tel: + 31 30 531369 Fax: + 31 30 540604 e-mail:
[email protected]
J.A.M. van Bergen Centre for Agriculture and Environment P.O. Box 10015 3505 AA Utrecht THE NETHERLANDS Tel: + 31 30 441301 Fax: + 31 30 441318
M.M. Berk National Institute of Public Health and Environmental Protection (RIVM) Global Change Dept. (pb 47) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743723 Fax: + 31 30 250740
C. Bernabo Science and Policy Associates Inc. The West Tower, Suite 400 1333 H Street, N.W. Washington, D.C. 20005 USA Tel: + 1 202 7891201 Fax: + 1 202 7891206
W. Biesiot University of Groningen Center for Energy and Environmental Studies (IVEM) Nijenborgh 4 9737 AG Groningen THE NETHERLANDS Tel: + 31 50 634611 Fax: + 31 50 637168
K. Blok University Utrecht Dept. NW&S P a d u a l a a n 14 3584 CH Utrecht THE NETHERLANDS Tel: + 31 30 537600 Fax: + 31 30 537601
1430
Y. de Boer Ministry of Housing, Spatial Planning and the Environment (VROM) DGM/LE P.O. Box 30945 2500 GX Den Haag THE NETHERLANDS Tel: + 31 70 3394690 Fax: + 31 70 3391310
C. Boix Fayos University of Amsterdam Landscape and Environmental Research Group Nieuwe Prinsengracht 130 1018 VZ A m s te r d a m THE NETHERLANDS Tel: + 31 20 5257451 Fax: + 31 20 5257431
H. de Boois Netherlands Organization for Scientific Research (NWO) P.O. Box 93138 2509 AC Den Haag THE NETHERLANDS Tel: + 31 70 3440752 Fax: + 31 70 3852045 e-mail: booisC-~wo.nl
P.A. Boot Ministry of Economic Affairs P.O. Box 20101 2500 EC Den Haag THE NETHERLANDS Tel: + 31 70 3796376 Fax: + 31 70 3797423
G.J. van den Born National Institute of Public Health and Environmental Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743782 Fax: + 31 30 250740 e-mail: gert-j
[email protected]
W. Bouten University of Amsterdam Lab. of Physical Geography and Soil Science Nieuwe Prinsengracht 130 1018 VZ A m s te r d a m THE NETHERLANDS Tel: + 31 20 5257451 / 12 Fax: + 31 20 5257431 e-mail:
[email protected]
A.F. B ouwm a n National Institute of Public Health and Environmental Protection P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743635 Fax: + 31 30 250740 e-mail:
[email protected]
B. Breemhaar Catholic University Brabant Inst. for Social Policy Research and Consultancy (IVA) P.O. Box 90153 5000 LE Tilburg THE NETHERLANDS Tel: + 31 13 662958 Fax: + 31 13 662959
A.H.M. Bresser National Institute of Public Health and Environmental Protection P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743756 Fax: + 31 30 252066
E.M. Bridges International Soil Reference and Information (ISRIC) P.O. Box 353 6700 AJ Wageningen THE NETHERLANDS Tel: + 31 8370 71711 Fax: + 31 8370 24460 e-mail:
[email protected]
1431
J. Bril A B - DLO P.O. Box 129 9750 AC H a r e n THE NETHERLANDS Tel: + 31 50 337777 Fax: + 31 50 337291 e-mail:
[email protected]
H.M. ten Brink Netherlands Energy Research Foundation (ECN) P.O. Box 1 1755 ZG Petten THE NETHERLANDS Tel: + 31 2246 4568 / 4148 Fax: + 31 2246 3488
R.A. Bruel NWO ESR P.O. Box 93120 2509 AC Den Haag THE NETHERLANDS Tel: + 31 70 3440857 Fax: + 31 70 3471623
J.J.C. Bruggink ECN Policy Studies P.O. Box 1 1755 ZG Petten THE NETHERLANDS Tel: + 31 2246 4321 Fax: + 31 2246 3338 e-mail:
[email protected]
H.R. de Bruin Agricultural University Wageningen Duivendaal 2 6701 AP Wageningen THE NETHERLANDS Tel: + 31 8370 83981 Fax: + 31 8370 82811
T.A. Buishand Royal Netherlands Meteorological Institute (KNMI) P.O. Box 201 3730 AE De Bilt THE NETHERLANDS Tel: + 31 30 206450 Fax: + 31 30 210407 e-mail:
[email protected]
A.G.J. BUMA University Groningen Marine Biology K e r k l a a n 30 975O AA HAREN THE NETHERLANDS Tel: 31-50-632393 Fax: 31-50-635205
G.J.H. Burgers KNMI P.O. Box 201 3730 AE De Bilt THE NETHERLANDS Tel: + 31 30 206682 Fax: + 31 30 210407 e-mail:
[email protected]
D. Carson Director Hadley Center for Climate Prediction and Research Meteorological Office Bracknell, Berkshire UNITED KINGDOM Tel: + 44 344 856611 Fax: + 44 344 854898
H. Cattle Hadley Centre for Climate Prediction and Research Meteorological Office London Road, Bracknell Berkshire UNITED KINGDOM Tel: + 44 344 856209 Fax: + 44 344 854898
1432
R. Coenen Kernforschungszenturm Karlsruhe GmbH Abteilung ftir Angewandte Systemanalyse Postfach 36 40 76021 Karlsruhe GERMANY Tel: + 49 07247 822509 Fax: + 49 07247 82406
S. Cohen Environmental Adaptation Research Group Climate and Research Research Directorate, Atm. Env. Service Downsview, Ontario M3H 5T4 CANADA Tel: + 1 416 7394389 Fax: + 1 416 7394297
L.A. Conrads University Utrecht Inst. for Marine and Atmospheric Research Princetonplein 5 3584 CC Utrecht THE NETHERLANDS Tel: + 31 30 533274/5 Fax: + 31 30 543163
D.W. Cornland Lund University Environmental and Energy Systems Studies Gerdagatan 13 S-223 62 Lund SWEDEN Tel: + 46 46104684 Fax: + 46 46108644 e-mail: deborah.cornland@milj o.lth.se
J.C. van Dam Van der Horstlaan 9 2 6 4 1 R T Pijnacker THE NETHERLANDS Tel: + 31 1 7 3 6 9 3 8 8 4
A. van Dasselaar Wageningen Agricultural University Dept. of Soil Science and Plant Nutrition P.O. Box 8005 6700 EC Wageningen THE NETHERLANDS Tel: + 31 8370 83195 Fax: + 31 8370 83766
R. Dellink Free University Amsterdam Institute for Environmental Studies De Boelelaan 1115 1081 HV A m s t e r d a m THE NETHERLANDS Tel: + 31 20 4449555 Fax: + 31 20 4449553
H.A.C. Denier van der Gon Wageningen Agricultural University Soil Science and Geology P.O. Box 37 6700 AA Wageningen THE NETHERLANDS Tel: + 31 8370 82422 Fax: + 31 8370 82419 e-mail:
[email protected]
T. Denne Ministry for the Environment P.O. Box 10362 Wellington NEW ZEALAND Tel: + 6 4 4 4 1 3 0490 Fax: + 64 4 471 0195
F. Diepstraten ECN Policy Studies P.O. Box 1 1755 ZG Petten THE NETHERLANDS Tel: + 31 2246 4424 / 4347 Fax: + 31 2246 3338
1433
P. Dijkstra Research Inst. for Agrobiology and Soil Fertility (AB-DLO) P.O. Box 14 6700 AC Wageningen THE NETHERLANDS Tel: + 31 8370 75941 Fax: + 31 8370 23110 e-mail:
[email protected]
B.O.M. Dirks Wageningen Agricultural University Dept. of Theoretical Production Ecology P.O. Box 430 6700 AK Wageningen THE NETHERLANDS Tel: + 31 8370 84769 Fax: + 31 8370 84892 e-mail:
[email protected]
H. van Dop University Utrecht IMAU Princetonplein 5 3584 CC Utrecht THE NETHERLANDS Tel: + 31 30 533154 Fax: + 31 30 543163 e-mail:
[email protected]
J.W.M. van der Drift University of Amsterdam Dept. of Physical Geography Nieuwe Prinsengracht 130 1018 VZ A m s t e r d a m THE NETHERLANDS Tel: + 31 20 5257398 Fax: + 31 20 5257431
J. Dronkers National Institute for Coastal and Marine Management (RIKZ) P.O. Box 20907 2500 EX Den Haag THE NETHERLANDS Tel: + 31 70 3745172 Fax: + 31 70 3282059
J.H. Duyzer T N O - MW P.O. Box 6011 2600 JA Delft THE NETHERLANDS Tel: + 31 15 696263 Fax: + 31 15 616812 e-mail:
[email protected]
M. van Eijk KNMI G. T r o m p l a a n 28 1243 LA s'Graveland THE NETHERLANDS Tel: + 31 35 561835 e-mail:
[email protected]
R. Eisma Energy Research Foundation P.O. Box 1 1755 ZG PETTEN THE NETHERLANDS Tel: 31-2246-4203 Fax: 31-2246-3468 e-mail: e i s m a ~ s . r u u . n l
A.J. Elshout N.V. KEMA P.O. Box 9035 6800 ET Arnhem THE NETHERLANDS Tel: + 31 85 562381 Fax: + 31 85 515022
B.J. Ens Institute for Forestry and Nature Research (IBN-DLO) P.O. Box 167 1790 AD Den Burg (Texel) THE NETHERLANDS Tel: + 31 2220 69750 Fax: + 31 2220 19235 e-mail:
[email protected]
1434
P. Ester Catholic University Brabant IVA P.O. Box 90153 5000 LE Tilburg THE NETHERLANDS Tel: + 31 13 662011 Fax: + 31 13 662959 e-mail:
[email protected]
F.A. Eybergen EC, ENRICH Office, DGXII/JRC Rue de la Loi 200 B-1040 Brussels BELGIUM Tel: + 3 2 2 2 9 5 9 1 7 7 Fax: + 32 2 2950146
H.G. van F a a s s e n AB - DLO P.O. Box 129 9750 AC Ha r e n THE NETHERLANDS Tel: + 31 50 337777 Fax: + 31 50 337291 e-mail:
[email protected]
T.R.F. Feitsma University Utrecht IMAU Princetonplein 5 3584 CC Utrecht THE NETHERLANDS Tel: + 31 30 533394 Fax: + 31 30 543163 e-mail: feitsmaC~ys.ruu.nl
S. Flaim International Energy Agency (IEA) 2, rue Andre-Pascal 75016 Paris Cedex FRANCE Tel: + 33 1 4524 9966 Fax: + 33 1 4524 9004
W. Fransen Royal Netherlands Meteorological Institute (KNMI) Section Ozone and Climate Scenarios P.O. Box 201 3730 AE De Bilt THE NETHERLANDS Tel: + 31 30 206675 Fax: + 31 30 210407 e-mail:
[email protected]
L.O. Fresco Agricultural University Wageningen Dept. of Agronomy H a a r we g 333 6709 RZ Wageningen THE NETHERLANDS Tel: + 31 8370 82513 Fax: + 31 8370 84575 e-mail:
[email protected]
A.P. Gaasbeek Shell Nederland B.V. Environmental Affairs P.O. Box 1222 3000 BE Rotterdam THE NETHERLANDS Tel: + 31 10 4696594 Fax: + 31 10 4696605
A. Garcia-Alvarez CSIC Serrano 115 dpdo 28006 Madrid SPAIN Tel: + 34 91 5625020
B.C.M. Gatersleben RUG Soc. en Organisatie Psychologie Grote Kruisstraat 2/1 9712 TS Groningen THE NETHERLANDS Tel: + 31 50 636433 e-mail:
[email protected]
1435
S.C. van de Geijn Research Inst. for Agrobiology and Soil Fertility (AB-DLO) Head of Dept. P l a n t Physiology P.O. Box 14 6700 AA Wageningen THE NETHERLANDS Tel: + 31 8370 75850 Fax: + 31 8370 23110
G.M.M. Gelauff CPB Van Stolkweg 14 2585 JR Den Haag THE NETHERLANDS Tel: + 31 70 3383441 Fax: + 31 70 3383350 e-mail:
[email protected]
R. Gerlagh National Institute of Public Health and Environmental Protection (RIVM) MTV P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743610 Fax: + 31 30 250740 e-mail:
[email protected]
B.M.E. Geurts CPB Van Stolkweg 14 2585 J R Den Haag THE NETHERLANDS Tel: + 31 70 3383324 Fax: + 31 70 3383350 e-mail:
[email protected]
A.M. Gielen CPB Van Stolkweg 14 2585 J R Den Haag THE NETHERLANDS Tel: + 31 70 3383329 Fax: + 31 70 3383350 e-mail:
[email protected]
A.M. de Gier KNAW Klimaatcommissie P.O. Box 19121 1000 GC A m s t e r d a m THE NETHERLANDS Tel: + 31 20 5510862 / 727 Fax: + 31 20 6204941
A. Gijswijt University of A m s t e r d a m SISWO Plantage Muidergracht 4 1018 TV A m s t e r d a m THE NETHERLANDS Tel: + 31 20 5270626 Fax: + 31 20 6229430 e-mail:
[email protected]
M. Gillissen Free University A m s te r d a m De Boelelaan 1105 1081 HV A m s t e r d a m THE NETHERLANDS Tel: + 31 20 4446095 Fax: + 31 20 4446005 e-mail:
[email protected]
J.H. van Ginkel Research Inst. for Agrobiology and Soil Fertility (AB-DLO) P.O. BOx 14 6700 AA Wageningen THE NETHERLANDS Tel: + 31 8370 75848 Fax: + 31 8370 23110 e-mail: in%
[email protected]
M. Glantz National Center for Atmospheric Research Environmental and Societal Impacts Group P.O. Box 3000 Boulder, CO 80307-3000 USA Tel: + 1 303 497819 Fax: + 1 303 497825
1436
W.A.C. van Gool Eindhoven University of Technology Dept. of Philosophy and Social Sciences P.O. Box 513 5600 MB Eindhoven THE NETHERLANDS Tel: + 31 40 472756 Fax: + 31 40 449875 e-mail:
[email protected]
W.J. Goossen Greenpeace Keizersgracht 174 1016 DW AMSTERDAM THE NETHERLANDS Tel: 31-20-6261877
A. Gorissen Research Inst. for Agrobiology and Soil Fertility (AB-DLO) P.O. Box 14 6700 AA Wageningen THE NETHERLANDS Tel: + 31 8370 75846 Fax: + 31 8370 23110 e-mail: in%
[email protected]
J. Goudriaan Agricultural University Wageningen Bornsesteeg 65 6708 PD Wageningen THE NETHERLANDS Tel: + 31 8370 83987 Fax: + 31 8370 84892 e-mail:
[email protected]
F.K. de G r a a f NWO GB-BOA p/a Geerdinkhof 422 1103 RE A m s t e r d a m THE NETHERLANDS Tel: + 31 20 4447180 Fax: + 31 20 4447123
D. Grimes Atmospheric Environment Service Env. Canada, Director Policy, Program & International Branch 4905 Dufferin Street Downsview, Ontario, M3H 5T4 CANADA Tel: + 1 416 7394344 Fax: + 1 416 7394380
M. Groen MGM&C P.O. Box 5578 2000 GN H a a r l e m THE NETHERLANDS Tel: + 31 23 424656 Fax: + 31 23 312481
W.T. de Groot Leiden University Centre of Environmental Science P.O. Box 9518 2311 RA Leiden THE NETHERLANDS Tel: + 31 71 277487 Fax: + 31 71 277496
R. Guicherit TNO Inst. of Environmental Sciences (IMW) P.O. Box 6011 2600 J A Delft THE NETHERLANDS Tel: + 31 15 696187 Fax: + 31 15 616812
J. Gupta Free University Amsterdam Institute for Environmental Studies De Boelelaan 1115 1081 HV A m s te r d a m THE NETHERLANDS Tel: + 31 20 4449548 Fax: + 31 20 4449553
1437
J. van H a m TNO I n s t i t u t e for E n v i r o n m e n t a l Sciences P.O. Box 6011 2600 J A Delft THE NETHERLANDS Tel: + 31 15 696877 Fax: + 31 15 613186 e-mail:
[email protected]
J.P. Heederik TNO Inst. of Applied Geoscience Dept. of Geo-Energy P.O. Box 6012 2600 JA Delft THE NETHERLANDS Tel: + 31 15 697197 Fax: + 31 15 564800
G.J. Heij National I n s t i t u t e of Public H e a l t h and Environmental Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743108 Fax: + 31 30 251932
P.S. Heikinheimo The Finnish Research P r o g r a m m e on Climate Change P.O.Box 57 FIN-00551 H E L S I N K I FINLAND Tel: 358-0-77488338 Fax: 358-0-7748299 e-mail:
[email protected]
G.W. Heil Resource Analysis Z u i d e r s t r a a t 110 2 6 1 1 S J Delft THE NETHERLANDS Tel: + 31 15 122622 Fax: + 31 15 124892 e-mail:
[email protected]
H.J. Heipieper Wageningen Agricultural University Division of Industrial Microbiology P.O. Box 8129 6700 EV Wageningen THE NETHERLANDS Tel: + 31 8370 84412 Fax: + 31 8370 84978 e-mail:
[email protected].
W. Helder N e t h e r l a n d s I n s t i t u t e for Sea Research (NIOZ) P.O. Box 59 1790 AB Den Burg (Texel) THE NETHERLANDS Tel: + 31 2220 69443 Fax: + 31 2220 19674
L. Hendrickx University of Groningen Center for Energy and Environmental Studies (IVEM) Nijenborgh 4 9747 AG Groningen THE NETHERLANDS Tel: + 31 50 634608 Fax: + 31 50 637168 e-mail:
[email protected]
A.P. Hensen N e t h e r l a n d s Energy Research Foundation (ECN) P.O. Box 1 1755 ZG P e t t e n THE NETHERLANDS Tel: + 31 2246 4203 Fax: + 31 2246 3488 e-mail:
[email protected]
E.J.M.T. van den Heuvel Biomass Technology Group B.V. P.O. Box 217 7500 AE Enschede THE NETHERLANDS Tel: + 31 53 894489 / 2897 Fax: + 31 53 893116 e-mail:
[email protected]
1438
J.E. van Hinte Free University Amsterdam Geomarien Centrum De Boelelaan 1085 1081 HV A m s t e r d a m THE NETHERLANDS Tel: + 31 20 4447309 Fax: + 31 20 6764811
M. HisschemSller Free University Amsterdam Institute for Environmental Studies De Boelelaan 1115 1081 HV Amsterdam THE NETHERLANDS Tel: + 31 20 4449523 Fax: + 31 20 4449553
J. Hoekstra KEMA KES P.O. Box 9035 6800 ET Arnhem THE NETHERLANDS Tel: + 31 85 562383 Fax: + 31 85 515022 e-mail:
[email protected]
H.H. Hoff German Advisory Council on Global Change Columbusstr 27515 Bremerhaven GERMANY Tel: + 49 471 4831701 Fax: + 49 471 4831218 e-mail: hhhofi~awi-bremerhaven.de
P. Hofschreuder Wageningen Agricultural University (LUW) Dept. of Air Quality P.O. Box 8129 6700 EV Wageningen THE NETHERLANDS Tel: + 31 8370 82104 Fax: + 31 8370 84457 e-mail:
A.A.M. Holtslag KNMI and University of Utrecht P.O. BOx 201 3730 AE De Bilt THE NETHERLANDS Tel: + 31 30 206458 Fax: + 31 30 210407 e-mail:
[email protected]
P.J.C. Honkoop Netherlands Institute for Sea Research (NIOZ) P.O. Box 59 1790 AB Den Burg (Texel) THE NETHERLANDS Tel: + 31 2220 69492 Fax: + 31 2220 19674
L. Hordijk Agricultural University Wageningen Center for Environmental and Climate Studies P.O. Box 9101 6700 HB Wageningen THE NETHERLANDS Tel: + 31 8370 84812 / 84919 Fax: + 31 8370 84839
E.J. Houwing Utrecht University Dept. Physical Geography, IMAU P o s t b u s 80.005 3508 TA Utrecht THE NE T H E R I AND S Tel: + 31 30 535735 Fax: + 31 30 540604 e-mail: e.houwing@fi~.ruu.nl
J.I. Hukkinen Maastricht School of Management P.O. Box 1203 6201 BE Maastricht THE NETHERLANDS Tel: + 31 43 618318 Fax: + 31 43 618330 e-mail:
[email protected]
1439
F. Ihle National Inst. of Public H e a l t h and E n v i r o n m e n t a l Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743724 Fax: + 31 30 250740 e-mail:
[email protected]
A.M. Iotova National Inst. of Meteorology and Hydrology Blvd. Tsarigradsko chaussee 66 1784 Sofia BULGARIA Tel: + 359 2 722271 (ext. 388) Fax: + 359 2 884494 / 880380
C.M.J. Jacobs Agricultural University Wageningen Dept. of Meteorology Duivendaal 2 6701 AP W a g e n i n g e n THE NETHERLANDS Tel: + 31 8370 82942 Fax: + 31 8370 82811 e-mail: cor.j
[email protected]
H. J a n s e n Free University A m s t e r d a m Inst. For E n v i r o n m e n t a l Studies De Boelelaan 1115 1081 HV A m s t e r d a m THE NETHERLANDS Tel: + 31 20 4449555 Fax: + 31 20 4449553
J.C. J a n s e n ECN Policy Studies P.O. Box 1 1755 ZG P e t t e n THE NETHERLANDS Tel: + 31 2246 4437 / 4347 Fax: + 31 2246 3338
M.A. J a n s s e n National I n s t i t u t e of Public H e a l t h and Environmental Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743320 Fax: + 31 30 252973
L.H.J.M. J a n s s e n N a t i o n a l I n s t i t u t e of Public H e a l t h and E n v i r o n m e n t a l Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 742771 Fax: + 31 30 287531
T.B. J o h a n s s o n Lund University Environmental and E n e r g y Systems Studies G e r d a g a t a n 13 S-22362 L u n d SWEDEN Fax: + 46 46 108644
P.J. J o h n s t o n Free University A m s t e r d a m Fac. der A a r d w e t e n s c h a p p e n De Boelelaan 1085 1081 HV A m s t e r d a m THE NETHERLANDS Tel: + 31 20 4447374 Fax: + 31 20 6462457 e-mail:
[email protected]
J.J. de J o n g Ministry of Economic Affairs P.O. Box 20101 2500 EC Den H a a g THE NETHERLANDS Tel: + 31 70 3796414 Fax: + 31 70 3797423
1440
P.T. de Jong University of Amsterdam Interfaculty Dept. of Environmental Science Nieuwe Prinsengracht 130 1018 VZ Amsterdam THE NETHERLANDS Tel: + 31 20 5255186 Fax: + 31 20 5256272 e-mail:
[email protected]
N.N. Joosten Research Inst. for Agrobiology and Soil Fertility (AB-DLO) P.O. Box 14 6700 AA Wageningen THE NETHERLANDS Tel: + 31 8370 75700 Fax: + 31 8370 23110 e-mail:
[email protected]
G.C.A. J u n n e University of Amsterdam Dept. of International Relations Oudezijds Achterburgwal 237 1012 DL Amsterdam THE NETHERLANDS Tel: + 31 20 5252163 Fax: + 31 20 5252086
P. Kabat The Winand Staring Centre (SC-DLO) P.O. Box 125 6700 AC Wageningen THE NETHERLANDS
E.N. Kairu Kenyatta University Dept. of Geography P.O. Box 43844 Nairobi KENYA Tel: + 2542 714682 Fax: + 2542 810759
A.D. Kant ECN Policy Studies P.O. Box 1 1755 ZG Petten THE NETHERLANDS Tel: + 31 2246 4442 / 4347 Fax: + 31 2246 3338
H. Kelder Royal Netherlands Meteorological Institute (KNMI) P.O. Box 201 3730 AE De Bilt THE NETHERLANDS Tel: + 31 30 206472 Fax: + 31 30 210407 e-mail:
[email protected]
S. Kelecsenyi Ministry for Environment and Regional Policy F6 u. 44-50 H- 1011 Budapest HUNGARY Tel: + 36 1 2013737 Fax: + 36 1 2011771
S.W.M. Kengen Wageningen Agricultural University Dept. of Microbiology Hesselink van Suchtelenweg 4 6703 CT Wageningen THE NETHERLANDS Tel: + 31 8370 83101 Fax: + 31 8370 83829 e-mail:
[email protected]
W.M. Kieskamp Netherlands Energy Research Foundation (ECN) P.O. Box 1 1755 ZG Petten THE NETHERLANDS Tel: + 31 2246 4644 Fax: + 31 2246 3488 e-mail:
[email protected]
1441
W. K l a a s s e n U n iv e r s ity of Groningen Dept. of Physical G e o g r a p h y K e r k l a a n 30 9751 N N H a r e n THE N E T H E R L A N D S Tel: + 31 50 636141 Fax: + 31 50 635205
J.H.G. K l a b b e r s KMPC BV Oostervelden 59 6681 B e m m e l THE N E T H E R L A N D S Tel: + 3 1 8 8 1 1 6 2 4 5 5 Fax: + 31 8811 62455 e-mail: j k l a b b @ a n t e n n a . n l
M. Kleber Universit~it H o h e n h e i m Inst. fOr B o d e n k u n d e u n d S t a n d o r t s l e h r e (310) 70593 S t u t t g a r t GERMANY Tel: + 49 711 4593668 Fax: + 49 711 4593117 e-mail:
[email protected]
E.G.M. Kleijn Centre of E n v i r o n m e n t a l Science Section S u b s t a n c e s a n d Products P.O. Box 9518 2300 RA Leiden THE N E T H E R L A N D S Tel: + 31 71 277480 Fax: + 31 71 277496 e-mail:
[email protected]
K. Klein Goldewijk N a t i o n a l I n s t i t u t e of Public H e a l t h and E n v i r o n m e n t a l Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE N E T H E R L A N D S Tel: + 31 30 743833 Fax: + 31 30 250740 e-mail:
[email protected]
H. Klein G u n n e w i e k AB - DLO P.O. Box 129 9750 AC H a r e n THE N E T H E R L A N D S Tel: + 31 50 337777 Fax: + 31 50 337291 e-mail: p o s t k a m e r @ a b . a g r o . n l
A.M.G Klein T a n k Royal N e t h e r l a n d s Meteorological I n s t i t u t e (KNMI) P.O. Box 201 3730 AE De Bilt THE N E T H E R L A N D S Tel: + 31 30 206589 Fax: + 31 210407 e-mail: k l e i n t a n @ k n m i . n l
M. Kok P r o g r a m m e Office N R P p/a N a t i o n a l I n s t i t u t e of Public H e a l t h and E n v i r o n m e n t a l Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE N E T H E R L A N D S Tel: + 31 30 743529 Fax: + 31 30 250740
G.P. K S n n e n Royal N e t h e r l a n d s Meteorological I n s t i t u t e (KNMI) P.O. Box 201 3730 AE De Bilt THE N E T H E R L A N D S Tel: + 3 1 3 0 2 0 6 4 5 1 Fax: + 31 30 210407
J. v a n der Kooij N.V. S E P U t r e c h t s e w e g 310 6800 AN A r n h e m THE N E T H E R L A N D S Tel: + 31 85 721473 Fax: + 31 85 430858
1442
E.A. Koster University Utrecht Vakgr. Fysische Geografie p/a g. van P r i ns t e r e r l a a n 35 1272 GB Huizen THE NETHERLANDS Tel: + 31 30 532749 Fax: + 31 30 540604
H.W. K6ster National Institute of Public Health and Environmental Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 3 1 3 0 7 4 3 3 2 0 Fax: + 31 30 252973
L. Koster Shell Nederland BV Dept. Environmental Affairs Hofplein 20 3032 AC Rotterdam THE NETHERLANDS Tel: + 31 10 4696035 Fax: + 31 10 4696605
P.R. Koutstaal University of Groningen Fac. of Law Dept. of Economics P.O. Box 716 9700 AS Groningen THE NETHERLANDS Tel: + 31 50 635263 Fax: + 31 50 635603 e-mail:
[email protected]
C.G.F. de Kovel Wageningen Agricultural University Dept. of Terrestrial Ecology and Nature Conservation Bornsesteeg 69 6708 PD Wageningen THE NETHERLANDS Tel: + 31 8370 83528 Fax: + 31 8370 84845 e-mail:
T. Kram Netherlands Energy Research Foundation (ECN) Dept. Policy Studies P.O. Box 1 1755 ZG Petten THE NETHERLANDS Tel: + 31 2246 4427 Fax: + 31 2246 3338 e-mail:
[email protected]
K. K r a m e r Inst. for Forestry and Nature Research (IBN-DLO) P.O. Box 23 6700 AA Wageningen THE NETHERLANDS Tel: + 31 8370 77899 Fax: + 31 8370 24988 e-mail:
[email protected]
G.J.J. Kreileman Resource Analysis p/a RIVM P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743554 Fax: + 31 30 250740 e-mail:
[email protected]
D. van Kreveld University Utrecht VSOP p/a Walenburg 11 3904 J M Veenendaal THE NETHERLANDS Tel: + 31 8385 11383 Fax: + 31 30 537584 e-mail:
[email protected]
S. van Kreveld Free University Amsterdam Fac. Ea r th Sciences De Boelelaan 1085 1081 HV A m s te r d a m THE NETHERLANDS Tel: + 31 20 4447251 Fax: + 31 20 6462457 e-mail:
[email protected]
1443
C. Kroeze National Institute of Public Health and Environmental Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743899 Fax: + 31 30 293651 e-mail:
[email protected]
M.S. Krol National Institute of Public Health and Environmental Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743835 Fax: + 31 30 250740 e-mail:
[email protected]
F. Kuik Royal Netherlands Meteorological Institute (KNMI) P.O. Box 201 3730 AE De Bilt THE NETHERLANDS Tel: + 31 30 206482 Fax: + 31 30 210407 e-mail:
[email protected]
P.J. K u ik m a n Research Inst. for Agrobiology and Soil Fertility (AB-DLO) P.O. Box 14 6700 AA Wageningen THE NETHERLANDS Tel: + 31 8370 75700 Fax: + 31 8370 23110 e-mail:
[email protected]
L.C. Kuiper University of Am s t e r da m Dept. Landscape and Environmental Research Group Nieuwe Prinsengracht 130 1018 VZ A m s t e r d a m THE NETHERLANDS Tel: + 31 20 5257458 Fax: + 31 20 5257431 e-mail:
[email protected]
P. Kuoppam~iki ETLA, Research Inst. of the Finnish Economy Ltinnrotinkatu 4B 00120 Helsinki FINLAND Tel: + 358 0 60990 Fax: + 358 0 601753 e-mail:
[email protected]
F.J.P.M. Kwa a d University of A m s t e r d a m Dept. FBL Nieuwe Prinsengracht 130 1018 VZ A m s t e r d a m THE NETHERLANDS Tel: + 31 20 5257447 Fax: + 31 20 5257431
J. Kwadijk Utrecht University Dept. of Physical Geography P.O. Box 80.115 3508 TC Utrecht THE NETHERLANDS Tel: + 31 30 532758 Fax: + 31 30 540604 e-mail:
[email protected]
A.C.A.P. van L a m m e r e n Royal Netherlands Meteorological Institute (KNMI) P.O. Box 201 3730 AE De Bilt THE NETHERLANDS Tel: + 31 30 206911 Fax: + 31 30 210407 e-mail:
[email protected]
G. Landrieu Inst. National de l'Environnement Industriel et des Risques (INERIS) Parc Technologique ALATA B.P. 2 60550 Verneuil-en-Halatte FRANCE Tel: + 33 44556392 Fax: + 33 44556699
1444
C.A. Langeveld Wageningen Agricultural University Dept. of Theoretical Production Ecology P.O. Box 430 6700 AK Wageningen THE NETHERLANDS Tel: + 31 8370 82140 / 41 Fax: + 31 8370 84892 e-mail:
[email protected]
H.J.M. Lankreijer University of Groningen Dept. of Physical Geography Kerklaan 30 9751 NN Haren THE NETHERLANDS Tel: + 31 50 634795 Fax: + 31 50 635205
W. Lass Philipps-Universiti~t Marburg Scientific Global Change, Advisory Council of the German Government (WBGU) Am Plan 2 350037 Marburg GERMANY Tel: + 49 6421 283166/72 Fax: + 49 6421 284852
A.W.F. van der Lee Milieudienst Regio Eindhoven P.O. Box 435 5600 AK EINDHOVEN THE NETHERLANDS Tel: + 31 40 386138 Fax: + 31 40 450195
Wen-Chin Lee Industrial Technology Research Centre Energy and Resources Lab. Bldg. 64, 195-6 Section 4 Chung Hsing Rd. Chutung, Hsinchu, 310 TAIWAN, R.O.C. Tel: + 886 35 917703 Fax: + 886 35 820376
P.A. Leffelaar Wageningen Agricultural University Dept. Theoretical Production Ecology Bornsesteeg 65 6708 PD Wageningen THE NETHERLANDS Tel: + 31 8370 83918 Fax: + 31 8370 84892 e-mail:
[email protected]
J. van Lenthe University of Groningen Center for Energy and Environmental Studies (IVEM) Nijenborgh 4 9747 AG Groningen THE NETHERLANDS Tel: + 31 50 636845 Fax: + 31 50 637168 e-mail: j.van.lenthe@~n.rug.nl
J.C. van der Leun University Hospital Utrecht (AZU) Dept. of Dermatology Heidelberglaan 100 3584 CX Utrecht THE NETHERLANDS Tel: + 31 30 507386 Fax: + 31 30 518328
M.J. Lexmond Agricultural University Wageningen (LUW) Dept. Environmental Technology P.O. Box 8129 6700 EV Wageningen THE NETHERLANDS Tel: + 31 8370 82023 Fax: + 31 8370 82108
J.M. Libre Elf Atochem Company 4, Cours Michelet Cedex 42 - 92091 Paris la D~fense 10 FRANCE Tel: + 33 1 49007887 Fax: + 33 1 49007021
1445
R. v a n Lier Royal N e t h e r l a n d s Meteorological I n s t i t u t e (KNMI) P.O. Box 201 3730 AE De Bilt THE N E T H E R L A N D S Tel: + 31 30 206379 Fax: + 31 30 210407 e-mail:
[email protected]
L. v a n Liere N at i o n a l I n s t i t u t e of Public H e a l t h a n d E n v i r o n m e n t a l Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE N E T H E R L A N D S Tel: + 31 30 743720 Fax: + 31 30 252066 e-mail:
[email protected]
T. L o u t e r s RIKZ P.O. Box 20907 25OO EX D E N HAAG THE N E T H E R L A N D S Tel: 31-70-3744803
H. van Loveren N at i o n a l I n s t i t u t e of Public H e a l t h a n d E n v i r o n m e n t a l Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE N E T H E R L A N D S Tel: + 31 30 742476 Fax: + 31 30 285283
P. M a h r e n h o l z F e d e r a l E n v i r o n m e n t a l Agency Bismarckplatz 1 14191 B e r l i n GERMANY Tel: + 49 30 89032840 Fax: + 49 30 89032285
M. Mandel Israel Meteorological I n s t i t u t e 61/33 E l n e k a v e TEL AVIV ISRAEL Tel: 03-399246 Fax: 03-9682124
E. Manrique-Reol I n t e r m i n i s t e r i a l Commission for R e s e a r c h a n d T e c h n o l o g y - S G P N I+D Rosario Pino, 14-16 28020 M a d r i d SPAIN Tel: + 34 1 3 3 6 0 4 1 8 / 3 9 4 1 7 7 1 Fax: + 34 1 3941774
W.J.M. M a r t e n s N at i o n a l I n s t i t u t e of Public H e a l t h a n d E n v i r o n m e n t a l Protection p/a F r a n k e n s t r a a t 197 6224 GP M a a s t r i c h t THE N E T H E R L A N D S Tel: + 31 43 883555 Fax: + 31 43 211889 e-mail: p . m a r t e n s @ m a t h . r u u l i m b u r g . n l
M. Mazzini U n i v e r s i t y of Pisa DCMN Via Diotisalvi, 2 56126 P i s a ITALY Tel: + 39 50 585258 Fax: + 39 50 585265 e-mail: ccii.unips.it
L.G.H. v a n der Meer TNO Inst. of Applied Geoscience Dept. of Geo-Energy P.O. Box 6012 2600 J A Delft THE N E T H E R L A N D S Tel: + 31 15 697197 Fax: + 31 15 564800
1446
G. Meijer University of Limburg Fac. of Economics P.O. Box 616 6200 MD Maastricht THE NETHERLANDS Tel: + 31 43 883763 Fax: + 31 43 258440 e-mail: g.meij
[email protected]
A.L. Meijnders Eindhoven University of Technology Fac. of Philoosphy and Social Sciences P.O. Box 513 5600 MB Eindhoven THE NETHERLANDS Tel: + 31 40 474210 Fax: + 31 40 449875 e-mail:
[email protected]
M.A. Mentzel University of Leiden Leiden Inst. for Social Scientific Research (LISWO) W a s s e n a a r s e w e g 52 2333 AK Leiden THE NETHERLANDS Tel: + 31 71 273845 Fax: + 31 71 273788
B. Metz Ministry of Housing, Spatial Planning and the Environment (VROM) Chairman Steering Body NRP P.O. Box 30945 2500 GX Den Haag THE NETHERLANDS Tel: + 31 70 3394383 Fax: + 31 70 3391311
L.A. Meyer Ministry of Housing, Spatial Planning and the Environment (VROM) DGM/LE P.O. Box 30945 2500 GX Den Haag THE NETHERLANDS Tel: + 31 70 3394407 Fax: + 31 70 3391310
A. Michaelowa HWWA Institute for Economic Research Neuer Jungfernstieg 21 20347 Hamburg GERMANY Tel: + 49 403562479 Fax: + 49 40351900
H. Middelkoop Utrecht University Dept. of Physical Geography P.O. Box 80.115 3508 TC Utrecht THE NETHERLANDS Tel: + 31 30 532749 Fax: + 31 30 540604 e-mail:
[email protected]
C.J.H. Midden Eindhoven University of Technology Fac. of Philosophy and Social Sciences P.O. Box 513 5600 MB Eindhoven THE NETHERLANDS Tel: + 31 40 473446 Fax: + 31 40 449875
J.G. van Minnen National Inst. of Public Health and Environmental Protection (RIVM) Global Change Department P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743915 Fax: + 31 30 250740 e-mail:
[email protected]
G.M.J. Mohren Institute for Forestry and Nature Research (IBN-DLO) P.O. Box 23 6700 AA Wageningen THE NETHERLANDS Tel: +31 8370 77891 Fax: + 31 8370 24988 e-mail:
[email protected]
1447
H.C. Moll University of Groningen IVEM Nijenborgh 4 9747 AG Groningen THE NETHERLANDS Tel: + 31 50 634607 Fax: + 31 50 637168 e-mail:
[email protected]
W.G. Mook Netherlands Institute for Sea Research (NIOZ) P.O. Box 59 1790 AB Den Burg (Texel) THE NETHERLANDS Tel: + 31 2220 69366 Fax: + 31 2220 19674
G.J. N a b u u r s Institute for Forestry and Nature Research (IBN-DLO) P.O. Box 23 6700 AA Wageningen THE NETHERLANDS Tel: + 31 8370 77700 Fax: + 31 8370 24988 e-mail: in%"
[email protected]
C. Nemes Ministry for Environment and Regional Policy Global Environment Office Programme Officer F5 u. 44-50 H-1011 Budapest HUNGARY Tel: + 36 1 2014091 Fax: + 36 1 2014091
J.P. Nieveen Wageningen Agricultural University Dept. of Meteorology Duivendaal 2 6701 AP Wageningen THE NETHERLANDS Tel: + 31 8370 82940 Fax: + 31 8370 82811 e-mail: j
[email protected]
S. Nonhebel Wageningen Agricultural University Theoretical Production Ecology P.O. Box 430 6700 AK Wageningen THE NETHERLANDS Tel: + 31 8370 84770 Fax: + 31 8370 84892 e-mail:
[email protected]
L.P. Norberg Statens Naturv~rdsverk 17185 Solna SWEDEN Tel: + 46 8 7991195 Fax: + 46 8 7991253 e-mail:
[email protected]
W.C. Oechel San Diego State University Dept. of Biology San Diego, CA 92182-0057 USA Tel: + 1 619 5946613 Fax: + 1 619 5947831
J.G.J. Olivier National Institute of Public Health and Environmental Protection (RIVM) P.O. Box 1 3720 BA Bilthovem THE NETHERLANDS Tel: + 31 30 743035 Fax: + 31 30 293651 e-mail:
[email protected]
A.A. Olsthoorn Free University A m s t e r d a m Institute for Environmental Studies De Boelelaan 1115 1081 HV A m s t e r d a m THE NETHERLANDS Tel: + 31 20 4449509 Fax: + 31 20 4449553
1448
J. Oonk TNO-ME P.O. Box 342 7300 AH Apeldoorn THE NETHERLANDS Tel: +31 55 493416 Fax: + 31 55 493287
W.A. Oost Royal Netherlands Meteorological Institute (KNMI) P.O. Box 201 3730 AE De Bilt THE NETHERLANDS Tel: + 31 30 206670 Fax: + 31 30 210407 e-mail:
[email protected]
J.B. Opschoor Free University Amsterdam Dept. of Environmental and Urban Economics P.O. Box 7161 1007 MC A m s t e r d a m THE NETHERLANDS Tel: + 31 20 4446092 Fax: + 31 20 4446005
R. Osinga Netherlands Institute for Sea Research (NIOZ) P.O. Box 59 1790 AB Den Burg THE NETHERLANDS Tel: + 31 2220 69573 Fax: + 31 2220 19674 e-mail:
[email protected]
L. Otto Netherlands Institute for Sea Research (NIOZ) T. Backerlaan 7 3984 P J Odijk THE NETHERLANDS Tel: + 31 3405 72081
A.H. Ovaa Agricultural University Wageningen Center for Environment and Climate Studies P.O. Box 9101 6700 HB Wageningen THE NETHERLANDS Tel: + 31 8370 82247 Fax: + 31 8370 84839
B. P a r m e t RIZA P.O. Box 9072 6800 ED Arnhem THE NETHERLANDS Tel: + 31 85 688574 Fax: + 31 85 688678
M. Parry University College London Dept. of Geography 26 Bedford Way London WC1H 0AP UNITED KINGDOM Tel: + 44 71 3807579 Fax: + 44 71 9160379
A.T.G. P a u l u s University of Limburg Fac. of Economics P.O. Box 616 6200 MD Maastricht THE NETHERLANDS Tel: + 31 43 883763 Fax: + 31 43 258440 e-mail:
[email protected]
M Pietrafesa Istituto di Ingegneria Civile ed Energetica Universit~ degli Studi di Reggio Calabria Via E. Cuzzocrea, 48 89128 Reggio Calabria ITALY Tel: + 39 965 875202 Fax: + 39 965 875254
1449
G.R. P i tt a ESS Division Dept. of Science and Technology New Mehrauli Road New D e l h i - 110 016 INDIA
J. van der Pligt University of A m s te r d a m Faculty of Psychology Dept. of Social Psychology Roetersstraat 15 1018 WB A m s t e r d a m THE NETHERLANDS Tel: + 31 20 5256891/0 Fax: + 31 20 6391896 e-mail:
[email protected]
S. Postle H a m m o n d Science and Policy Associates Inc. The West Tower, Suite 400 1333 H Street, N.W. Washington, D.C. 20005 USA Tel: + 1 202 7891201 Fax: + 1 202 7891206
H. Postma KNAW Klimaatcommissie p/a Schilderend 132 1791 BK Den Burg THE NETHERLANDS Tel: + 31 2220 13354
J. P o s t m a Staring Centre Marijkeweg 11 6500 C Wageningen THE NETHERLANDS Tel: 31-8370-78911
J. Q u a k e r n a a t TNO-ME Dept. of Environmetnal Technology P.O. Box 342 7300 AH Apeldoorn THE NETHERLANDS Tel: + 31 55 493935 Fax: + 31 55 493410
M. R a a k RIZA P.O. Box 9072 6800 ED Arnhem THE NETHERLANDS Tel: + 31 85 688574 Fax: + 31 85 688678
M. Redclift University of London Wye College Dept. of Environmental Sciences Wye, Kent TN25 5AH UNITED KINGDOM Tel: + 44 233 812401 Fax: + 44 233 813187
D. van Regemorter Catholic University Leuven Center for Economic Studies N a a m s e s t r a a t 69 B-3000 Leuven BELGIUM Tel: + 32 16 326812 Fax: + 32 16 326796
H.A.J.M. Reinen National Institute of Public Health and Environmental Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743828 Fax: + 31 30 291604 e-mail:
[email protected]
1450
S.A. Rienstra Free University Amsterdam Dept. Regional Economics De Boelelaan 1105 1081 HV A m s t e r d a m THE NETHERLANDS Tel: + 31 20 4446096 Fax: + 31 20 4446005 e-mail:
[email protected]
W. Roeleveld Free University Amsterdam Fac. Aardwetenschappen De Boelelaan 1085 1081 HV A m s te r d a m THE NETHERLANDS Tel: + 31 20 4447355 Fax: + 31 20 6462457 e-mail:
[email protected]
M. Roemer MW-TNO P.O. Box 6011 2624 ZK DELFT THE NETHERLANDS Tel: 31-15-696037 Fax: 31-15-616812 e-mail:
[email protected]
R.P. Roetter DLO - The Winand Staring Centre Dept. Land Evaluation Methods P.O. Box 125 6700 AC Wageningen THE NETHERLANDS Tel: + 31 8370 74229 Fax: + 31 8370 24812 e-mail:
[email protected]
R.S.A.R. van Rompaey Pr og r a m m e Office NRP p/a National Institute of Public Health and Environmental Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743781 Fax: + 31 30 251932 e-mail:
[email protected]
J. Rouwendal Wageningen Agricultural University P.O. Box 8060 6700 DA Wageningen THE NETHERLANDS Tel: + 31 8370 84730 Fax: + 31 8370 82593
W.P.M. de Ruijter University Utrecht IMAU Princetonplein 5 3584 CC Utrecht THE NETHERLANDS Tel: + 31 30 533275 Fax: + 31 30 543163
M.G. Schaap University of Amsterdam Lab. of Physical Geography and Soil Science Nieuwe Prinsengracht 130 1018 VZ A m s te r d a m THE NETHERLANDS Tel: + 31 20 5257450 Fax: + 31 20 5257431 e-mail:
[email protected]
C. Schmidt University of Amsterdam A m s t e r d a m School for Social Science Research Oude Hoogstraat 24 1012 CE A m s t e r d a m THE NETHERLANDS Tel: + 31 20 5252262 Fax: + 31 20 5252446
I. Schmidt Projekttr~iger des BMFT ftir Arbeit, Umwelt und Gesundheit Stidstr. 125 D-53175 Bonn GERMANY Tel: + 49 228 3821224 Fax: + 49 228 3821256 e-mail:
[email protected]
1451
T. Schneider Programme Director NRP p/a National Institute of Public Health and Environmental Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 742979 Fax: + 31 30 251932
B. Schoenmaeckers MGM&C P.O. Box 5578 2000 GN H a a r le m THE NETHERLANDS Tel: + 31 23 424656 Fax: + 31 23 312481
E. Schols DSM Research P.O. Box 18 6160 MD Geleen THE NETHERLANDS Tel: + 31 46 761344 Fax: + 31 46 760700
C.J.E. Schuurmans Royal Netherlands Meteorological Institute (KNMI) P.O. Box 1 3730 AE De Bilt THE NETHERLANDS Tel: + 31 30 206445 Fax: + 31 30 210407
R. Segers Wageningen Agricultural University Dept. of Theoretical Production Ecology Bornsesteeg 65 6708 PD Wageningen THE NETHERLANDS Tel: + 31 8370 82141 Fax: + 31 8370 84892 e-mail: Reinoud. Segers@staff. tpe.wau.nl
R. Sikkema Institute for Forest and Forest Products P.O. Box 253 6700 AG Wageningen THE NETHERLANDS Tel: + 31 8370 24666 Fax: + 31 8370 10247
V.J. Simpson Dept. of Environment B246, Romney House 43 M a r s h a m Street London, S W l P 3PY UNITED KINGDOM Tel: + 44 71 2768297 Fax: + 44 71 2768509
R.S. Singh Wageningen Agricultural University Dept. of Meteorology Duivendaal 2 6701 AP Wageningen THE NETHERLANDS Tel: + 31 8370 82942 Fax: + 31 8370 82811 e-mail:
[email protected]
J. Slanina Netherlands Energy Research Foundation (ECN) P.O. Box 1 1755 ZG Petten THE NETHERLANDS Tel: + 31 2246 4236 Fax: + 31 2246 3488
H. Slaper National Institute of Public Health and Environmental Protection (RIVM) LSO P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743488 Fax: + 31 30 291604 e-mail:
[email protected]
1452
G. Slotegraaf University of Groningen Inst. for Social and Organisational Physchology Grote K r u i s s t r a a t 2/1 9712 TS Groningen THE NETHERLANDS Tel: + 31 50 636482 Fax: + 31 50 636304 e-mail:
[email protected]
J.P. van der Sluijs University Utrecht Dept. of Science Technology and Society P a d u l a a n 14 3584 CH Utrecht THE NETHERLANDS Tel: + 31 30 537631 Fax: + 31 30 537601 e-mail:
[email protected]
C.J. Smit IBN-DLO P.O.Box 167 1790 AD Den Burg THE NETHERLANDS Tel: 2220-69712 Fax: 2220-19235
H. Spoelstra KEMA BV P.O. Box 9035 6845 ED Arnhem THE NETHERLANDS Tel: + 31 85 563006 Fax: + 31 85 515022
L. Srivastava Tata Energy Research Institute 9, Jor Bagh New Delhi 110 003 INDIA
O. van Steenis Programme Office NRP p/a National Institute of Public Health and Environmental Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 742970 Fax: + 31 30 251932
E.M. Steg RUG Social and Organizational Psychology Grote K r u i s s t r a a t 2/1 9712 TS Groningen THE NETHERLANDS Tel: + 31 50 636431 Fax: + 31 50 636304 e-mail:
[email protected]
A. Sterl Royal Netherlands Meteorological Institute (KNMI) P.O. Box 201 3730 AE De Bilt THE NETHERLANDS Tel: + 31 30 206349 Fax: + 31 30 210407 e-mail:
[email protected]
W.A.S. Stibbe N.V. SEP Utrechtseweg 310 6800 AN Arnhem THE NETHERLANDS Tel: + 31 85 721425 Fax: + 31 85 430858
M.A. Stritt University of Neuch~tel Inst. for Economic and Regional Research Pierre-~-Mazel 7 Neuchatel, 2000 SWITZERLAND Tel: + 41 38 211340 Fax: + 41 38 211085 e-mail:
[email protected]
1453
L.C.P.M. Stuyt Winand Staring Center (SC-DLO) P.O. Box 125 6700 AC Wageningen THE NETHERLANDS Tel: + 31 8370 74298 Fax: + 31 8370 24812 e-mail:
[email protected]
R.J. Swart National Institute of Public Health and Environmental Protection P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743026 Fax: + 31 30 250740 e-mail:
[email protected]
W. T a k k e n Wageningen Agricultural University Dept. Entomology P.O. Box 8031 6700 EH Wageningen THE NETHERLANDS Tel: + 31 8370 84652 Fax: + 31 8370 84821 e-mail:
[email protected]
G. Tertoollen University Utrecht VSOP Heidelberglaan 1 3584 CS Utrecht THE NETHERLANDS Tel: + 31 30 534196 Fax: + 31 30 537585
J.H.J. Terwindt University Utrecht Dept. of Geography P.O. Box 80.115 3508 TC Utrecht THE NETHERLANDS Tel: + 31 30 532740
H. Thorgeisson Agricultural Research Institute Dep. Environmental Research Keldnaholt IS-112 REYKJAVIK
J.T. Tirkkonen University of Tampere Dept. of Regional Studies P.O. Box 607 SF-33101 T a m p e r e FINLAND Tel: + 358 31 2157194 Fax: + 358 31 2157311 e-mail:
[email protected]
A.M.C. Toet National Institute of Public Health and Environmental Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 744036 Fax: + 31 30 250740 e-mail:
[email protected]
R.S.J. Tol Free University A m s t e r d a m Inst. for Environmental Studies De Boelelaan 1115 1081 HV A m s t e r d a m THE NETHERLANDS Tel: + 31 20 4449555 Fax: + 31 20 4449553 e-mail:
[email protected]
M.W.A. Tosserams Free University A m s te r d a m De Boelelaan 1087 1081 HV A m s t e r d a m THE NETHERLANDS Tel: + 31 20 4447061 Fax: + 31 20 4447123
ICELAND Tel: 354-1-873230 Fax: 354-1-87604 e-mail:
[email protected]
1454
W. T u i n s t r a National Institute of Public Health and Environmental Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 744036 Fax: + 31 30 250740 e-mail:
[email protected]
W.C. Turkenburg University Utrecht Dept. of Science Technology and Society P a d u l a a n 14 3584 CH Utrecht THE NETHERLANDS Tel: + 31 30 537625 Fax: + 31 30 537601
A.P. van Ulden Royal Netherlands Meteorological Institute (KNMI) P.O. Box 201 3730 AE De Bilt THE NETHERLANDS Tel: + 31 30 206911 Fax: + 31 30 210407
M. Unsworth Oregon State University Center for Analysis of Environmental Change Weniger Hall 283 Corvallis, OR 97331-6511 USA Tel: + 1 503 7371744 Fax: + 1 503 7373399
P.C.F. van der Vaart University Utrecht Inst. for Marine and Atmospheric Research Utrecht (IMAU) Princetonplein 5 3584 CC Utrecht THE NETHERLANDS Tel: + 31 30 533394 Fax: + 31 30 513163 e-mail:
[email protected]
M. Vanderstraeten Federal Office for Scientific, Technical and Cultural Affairs Wetenschapsstraat 8 1040 BRUSSEL BELGIUM Tel: + 32 2 2383610 Fax: + 32 2 2305912
J.F. van de Vate IAEA P.O. Box 100 A- 1400 VIENNA AUSTRIA Tel: 43-1-23602787 Fax: 43-1-234564
A. Veen National Inst. of Public Health and Environmental Protection (RIVM) Dept. LWD P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743458 Fax: + 31 30 252066
D. Veenhuysen Wageningen Agricultural University (LUW) Dept. of Air Quality P.O. Box 8129 6700 EV Wageningen THE NETHERLANDS Tel: + 31 8370 82684 Fax: + 31 8370 84457
A. Veldkamp Wageningen Agricultural University Dept. of Agronomy P.O. Box 341 6700 AH Wageningen THE NETHERLANDS Tel: + 31 8370 83074 Fax: + 31 8370 84575 e-mail:
[email protected]
1455
P. Vellinga Free University A m s t e rd a m Institute for Environmental Studies De Boelelaan 1115 1081 HV A m s t e r d a m THE NETHERLANDS Tel: + 31 20 4449555 Fax: + 31 20 4449553
G.L. Velthof Wageningen Agricultural University NMI, Dept. of Soil Science & Plant Nutrition P.O. Box 8005 6700 EC Wageningen THE NETHERLANDS Tel: + 31 8370 85049 Fax: + 31 8370 83766 e-mail:
A.C. Veltkamp Netherlands Energy Research Foundation (ECN) P.O. Box 1 1755 ZG Petten THE NETHERLANDS Tel: + 31 2246 4251 Fax: + 31 2246 3488
H. Verbruggen Free University A m s te r d a m Institute for Environmental Studies De Boelelaan 1115 1081 HV A m s t e r d a m THE NETHERLANDS Tel: + 31 20 4449555 Fax: + 31 20 4449553
P.S.J. Verburg Wageningen Agricultural University Dept. of Soil Science and Geology P.O. Box 37 6700 AA Wageningen THE NETHERLANDS Tel: + 31 8370 84043 Fax: + 31 8370 82419 e-mail:
[email protected]
J.W. Verkaik Wageningen Agricultural University Dept. of Meteorology Duivendaal 2 6701 AP Wageningen THE NETHERLANDS Tel: + 31 8370 84113 Fax: + 31 8370 82811 e-mail: job.verkaikC-~sers.met.wau.nl
A.T. Vermeulen Netherlands Energy Research Foundation (ECN) P.O. Box 1 1755 ZG Petten THE NETHERLANDS Tel: + 31 2246 4194 Fax: + 31 2246 3488 e-mail:
[email protected]
E.C.H. Verstraten University Utrecht Vreuchdenberchdreef 63 3562 HL Utrecht THE NETHERLANDS Tel: + 3 1 3 0 6 2 5 1 0 8 Fax: + 31 30 537584
P.J.F. de Vink National Institute of Public Health and Environmental Protection P.O.Box 1 3720 BA BILTHOVEN THE NETHERLANDS Tel: 31-30-742979 e-mail:
[email protected]
F. Visser G a n s b r o e k s t r a a t 31 2870 Ruisbroek BELGIUM Tel: + 32 3 8867125 Fax: + 32 3 8863038
1456
A.J. Visser Free University A m s t e r d a m De Boelelaan 1087 1081 HV A m s t e r d a m THE NETHERLANDS Tel: + 31 20 4447061 Fax: + 31 20 4447123
C.A.J. Vlek University of Groningen Dept. of Psychology Grote K r u i s s t r a a t 2/1 9712 TS Groningen THE NETHERLANDS Tel: + 31 50 636443 Fax: + 31 50 636304 e-mail:
[email protected]
J.M. Vleugel Free University A m s t e r d a m Dept. Spatial Economics De Boelelaan 1105 1081 HV A m s t e r d a m THE NETHERLANDS Tel: + 31 20 4446096 Fax: + 31 20 4446005 e-marl:
[email protected]
A.J.H. van Vliet National Inst. of Public H e a l t h and Environmental Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743170 Fax: + 31 30 250740 e-mail:
[email protected]
A. Vollering KNAW P.O. Box 19121 1000 GC A m s t e r d a m THE NETHERLANDS Tel: + 31 20 5510712 Fax: + 31 20 6204941
M. Vonk Programme Office NRP p/a National Institute of Public H e a l t h and Environmental Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743211 Fax: + 31 30 251932 e-mail:
[email protected]
M.E.J.P. Vosbeek KEMA P.O. Box 9035 6800 ET A r n h e m THE NETHERLANDS Tel: + 31 85 562383 Fax: + 31 85 515022 e-mail:
[email protected]
L.P.M. de Vrees National Institute for Coastal and Marine M a n a g e m e n t Coastal Zone M a n a g e m e n t Centre P.O. Box 20907 2500 EX The Hague THE NETHERLANDS Tel: + 31 70 3745169 Fax: + 31 70 3744959 e-mail:
[email protected]
H.J.M. de Vries N a t i o n a l I n s t i t u t e of Public H e a l t h and Environmental Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743320 Fax: + 31 30 252973
F.H. Walsteijn Utrecht University Institute for Marine and Atmospheric Research Princetonplein 5 3584 CC Utrecht THE NETHERLANDS Tel: + 31 30 533169 Fax: + 31 30 543163
1457
M. van Weele University Utrecht INMAU Princetonplein 5 3584 CC UTRECHT THE NETHERLANDS Tel: 31-30-533271 Fax: 31-30-543163 e-mail:
[email protected]
J.B. Weenink Ministry of Housing, Spatial Planning and the Environment (VROM) DGM/LE P.O. Box 30945 2500 GX Den Haag THE NETHERLANDS Tel: + 31 70 3394690 Fax: + 31 70 3391310
J.J. van de Wege Wageningen Agricultural University Dept. Entomology P.O. Box 8031 6700 EH Wageningen THE NETHERLANDS Tel: + 31 8370 85061 Fax: + 31 8370 84812 e-mail: j
[email protected]
P. Westbroek Leiden University Gorlaeus Laboratories P.O. Box 9502 2300 RA Leiden THE NETHERLANDS Tel: + 31 71 274722 Fax: + 31 71 274537
J. Wieringa Wageningen Agricultural University (LUW) Dept. of Meteorology Duivendaal 2 6701 AP Wageningen THE NETHERLANDS Tel: + 31 8370 83981 Fax: + 31 8370 82811
T. Wigley UCAR Office for Interdisciplinary Ea r th Studies P.O. Box 3000 Boulder, CO 80303 USA Tel: + 1 303 4972690 Fax: + 1 303 4972699
H.P.J. de Wilde Netherlands Institute for Sea Research (NIOZ) P.O. Box 59 1790 AB Den Burg (Texel) THE NETHERLANDS Tel: + 31 2220 69444 Fax: + 31 2220 19674 e-mail:
[email protected]
P. de Wildt Royal Netherlands Meteorological Institute (KNMI) P.O. Box 201 3730 AE De Bilt THE NETHERLANDS Tel: + 31 30 206465 Fax: + 31 30 210407
M.J. Wilenius University of Tampere Research Institute for Social Sciences P.O. Box 33 00014 Helsinki FINLAND Tel: + 358 01912079 Fax: + 358 01912124 e-mail:
[email protected]
H.C. Willers I M A G - DLO P.O. Box 43 6700 AA Wageningen THE NETHERLANDS Tel: + 31 8370 76591 Fax: + 31 8370 25679 e-mail: h.c.willers@imag, agro.nl
1458
R. de Winter-Sorkina National Institute of Public Health and Environmental Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 742331 Fax: + 31 30 287531 e-mail:
[email protected]
J. Wolf Wageningen Agricultural University Dept. TPE P.O. Box 430 6700 AK Wageningen THE NETHERLANDS Tel: + 31 8370 84769 Fax: + 31 8370 84892 e-mail:
[email protected]
W.J. Wolff DLO- I n s t i t u u t voor Bos- en Natuuronderzoek P.O. Box 23 6700 AA WAGENINGEN THE NETHERLANDS Tel: + 31 3434 55250 Fax: + 31 3434 55288 e-mail: w.j.wolff@ibn, agro.nl
R.J.W. van der Wurff University of Amsterdam Dept. of International Relations Oudezijds Achterburgwal 237 1012 DL Amsterdam THE NETHERLANDS Tel: + 31 20 5252160 Fax: + 31 20 5252086 e-mail:
[email protected]
G.P. Wyers Netherlands Energy Research Foundation (ECN) P.O. Box 1 1755 ZG Petten THE NETHERLANDS Tel: + 31 2246 4155 Fax: + 31 2246 3488
J.P. van Ypersele Inst. d'Astronomie et de G~ophysique George Lemaitre U.C.L. 2, Chemin de Cyclotron B-1348 Louvain-la-Neuve BELGIUM Tel: + 32 10473296 Fax: + 32 10474722 e-mail:
[email protected]
G. Zeeman Agricultural University Wageningen Dept. of Environmental Technology P.O. Box 8129 6700 EV Wageningen THE NETHERLANDS Tel: + 31 8370 84804 Fax: + 31 8370 84802
M.H. Zemankovics Hungarian Meteorological Service Agrometeorological Research Station H-8360 Keszthely HUNGARY Tel: + 36 83 312856 Fax: + 36 83 315105
Z.X. Zhang University of Wageningen Dept. of General Economics P.O. Box 8130 6706 KN Wageningen THE NETHERLANDS Tel: + 31 8370 84637 Fax: + 31 8370 84763 e-mail: zhang.zhongxiang@alg, shhk.wau.nl
W.M. Zijlstra BMRO / VNO&NCW P.O. Box 93093 2509 AB Den Haag THE NETHERLANDS Tel: + 31 70 3497481 Fax: + 31 70 3850707
1459
G. Zuidema National Institute of Public Health and Environmental Protection Global Change Dept. P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743554 Fax: + 31 30 250740 e-mail:
[email protected]
M. van Zwetselaar National Institute of Public Health and Environmental Protection P.O.Box 1 3720 BA BILTHOVEN THE NETHERLANDS Tel: + 31 30 744018 e-mail:
[email protected]
S. Zwerver Programme Secretary NRP p/a National Institute of Public Health and Environmental Protection (RIVM) P.O. Box 1 3720 BA Bilthoven THE NETHERLANDS Tel: + 31 30 743211 Fax: + 31 30 251932 e-mail:
[email protected]
1461
A n n e x 2: A c r o n y m s Organizations/programmes GEIA IGAC IPCC IUPAC NRP
Global Emission Inventories Activity International Global Atmospheric Chemistry Programme Intergovernmental Panel on Climate Change International Union on Pure and Applied Chemistry National Research Programme on Global Air Pollution and Global Change IGBP-GAIM International Biosphere Geosphere Programme - Global Analysis, Interpretation, and Modelling IGBP-GCTE International Biosphere Geosphere Programme - Global Change and Terrestrial Ecosystems FAO Food and Agriculture Organization OECD Organization for Economic Co-operation and Development
Institutes CABO-DLO ECN IB-DLO IBN-DLO IFP IIASA IRRI ISRIC KEMA KNMI LUW-TPE NIOZ NMI NOAA RIVM TNO- MW TNO - ME
Centre for Agrobiological Research Netherlands Energy Research Foundation Institute for Soil Fertility Research Institute for Forestry and Nature Research Institute Francais du Petrol (France) International Institute for Applied System Analysis (Austria) International Rice Research Institute (Puerto Rica) International Soil Reference and Information Centre KEMA Environmental Services Royal Netherlands Meteorological Institute Wageningen Agriculture University - Theoretical Production Ecology Netherlands Institute for Sea Research Institute for Soil Fertility Research National Oceanographic and Atmospheric Administration (USA) National Institute of Public Health and Environmental Protection N e t h e r l a n d s Organization for Applied Scientific Research Environmental Sciences N e t h e r l a n d s Organization for Applied Scientific Research Environmental and Energy technology
Models / databases IMAGE EDGAR WISE
Integrated Model to Assess the Greenhouse Effect Emission Database for Global Atmospheric Research World Inventory of Soil Emission Potentials
1463
A n n e x 3: U n i t s peta tera giga mega kilo micro nano
P = 1015 T = 1012 G = 109 M = 106 k = 103 =m = 10-3 = ~t = 10-6 = n = 10-9
pptv ppbv ppmv
= p a r t s p e r t r i l l i o n (1012) b y v o l u m e = p a r t s p e r b i l l i o n (109) b y v o l u m e = p a r t s p e r m i l l i o n (106) b y v o l u m e
ton kton Mton 1000 kton
= metric ton = 1000 kg = 1 0 0 0 t o n = 109 g = 1012g= 1Tg =lTg
ha year
= 10000 m2 =y
PDB pmC
= Pee Dee Belemniet = percentage modern carbon
mil]i
= = = = =