Lakes on Mars

Elsevier Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK Copyright © 2010 Elsevier B.V. All rights reserved No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher. Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone: (+44) 1865 843830, fax: (+44) 1865 853333, E-mail: [email protected]. You may also complete your request online via the Elsevier homepage (http://elsevier.com), by selecting “Support & Contact” then “Copyright and Permission” and then “Obtaining Permissions.” Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN: 978-0-444-52854-4 For information on all Elsevier Publications visit our Web site at elsevierdirect.com Printed and bound in the Netherland 10 11

10 9 8 7 6 5 4 3 2 1

Working together to grow libraries in developing countries www.elsevier.com | www.bookaid.org | www.sabre.org

Contents Something special about Mars List of acronyms List of contributors Foreword

ix xi xiii XV

Searching for lakes on Mars: four decades of exploration Nathalie A. Cabral and Edmond A. Grin

2

1.1

Introduction

1

1.2

Conditions at the time of lake formation

3

1.3

The lake record of Mars

12

1.4

Time machines

18

Acquisition and history of water on Mars

31

Michael H. Carr and James W Head

3

2.1

Introduction

31

2.2

Acquisition and retention of water

32 34

2.3

Early geologic events

2.4

The Noachian era

36

2.5

Hesperian era

44

2.6

Amazonian era

52

2.7

Summary

58

Hydrologic provinces of Mars: physiographic controls on drainage and ponding

69

Rene A. De Hon

4

3.1

Introduction

69

3.2

Physiographic control

71

3.3

Hydrologic provinces

73

3.4

Discussion

84

Heated lakes on Mars

91

Horton E. Newsom

5

4.1

Introduction

91

4.2

Sources of water

93

4.3

Sources of heat

4.4

Discussion

100

4.5

Conclusions

102

Lakes in Valles Marineris

95

111

Baerbel K. Lucchitta 5.1

Introduction

111

5.2

The Mariner Era

111

Contents

VI

6

5.3

The Viking Era

115

5.4

The MGS Era

1 26

5.5

Odyssey and beyond

132

5.6

Discussion

144

5.7

Summary and conclusion

152

Episodic ponding and outburst flooding associated with chaotic terrains in Valles Marineris

163

Keith P Harrison and Mary G. Chapman

7

6.1

Introduction

1 63

6.2

Topographic constraints on lake setting

165

6.3

Morphological evidence of a VMD paleolake

166

6.4

Discussion

183

6.5

Conclusions

189

Evidence for ancient lakes in the Hellas region

1 95

Sharon A. Wilson, Jeffrey M. Moore, Alan D. Howard, and Don E. Wilhelms

8

7.1

Introduction

195

7.2

Regional geology

198 1 98

7.3

Valley networks and canyons as sources of lake water

7.4

Ancient lake shorelines: regional relations between topography, morphology, and mineralogy

200

7.5

Light-toned layered deposits

207

7.6

Discussion

213

7.7

Conclusions

215

Deltas and valley networks on Mars: implications for a global hydrosphere

223

Gaetano di Achille and Brian M. Hynek

9

8.1

Introduction

223

8.2

Deltas on Mars

225

8.3

Valley networks

235

8.4

A Test of the Martian ocean

238

8.5

Summary

243

The northern plains: A Martian oceanic basin?

249

Timothy J. Parke1; John A. Grant, and Brenda J. Franklin 9.1

Introduction

249

9.2

Coastlines and topography

252

9.3

Proposed "shorelines" and related landforms in the west

9.4

Deuteronilus Mensae/east Acidalia Region

254

Discussion

270

Contents

10

VII

The Western Elysium Planitia Paleolake

275

Matthew R. Balme, Colman J. Gallagher, David P Page, John B. Murray, Jan-Peter Muller, and Jung-Rack Kim

11

10.1

Introduction

275

10.2

Western Elysium Basin: general description

278

10.3

Landforms

285

10.4

Formation age of the Western Elysium Basin deposits

292

I 0.5

Discussion

293

I 0.6

Conclusions

301

The sedimentary record of modern and ancient dry lakes

307

Gian G. Ori

12

11.1

Introduction

307

11.2

Facies and sedimentary environments

308

I 1.3

Sabkhas as Mars analogs

318

Aqueous depositional settings in Holden crater, Mars

323

John A. Grant, Rossman P Irwin, III, and Sharon A. Wilson

13

12.1

Introduction

323

12.2

Geomorphic setting

326

12.3

Geologic history

327

I 2.4

Holden crater stratigraphy

328

12.5

Origin of stratigraphy in Holden crater

336

12.6

Discussion

341

Dynamics of declining lake habitat in changing climate

347

Nathalie A. Cabral, Edmond A. Grin, Guillermo Chong, Donat P. Hader, Edwin Minkley, Youngseob Yu, Cecilia Demergasso, John A. Gibson, and Darlene Lim I 3.1

Introduction

347

13.2

Environmental analogy to Mars

349

13.3

Methods

351

13.4

Results

353

13.5

Conclusion

362

Author Index

371

Subject Index

381

Something special about Mars

In the past 30 years, the myth about Mars has given way to the hard data collected by orbital and ground missions. The romance of the canals and oases of Schiaparelli and Lowell has faded with time, and today earthlings are taking their revenge on H. G. Wells’ Martians by invading the red planet and relentlessly poking its surface. But even if reality has replaced the imaginative visions of Mars from a century ago, the excitement of exploring this world has far from vanished. In fact, if we look closely, nothing has really changed: our investigations are still about water and life. Sure, the channels we find now are not artificial, but they comprise some of the evidence supporting the existence of ancient lakes, deltas, possibly an ocean. We have also uncovered vast reservoirs of underground water, giant volcanoes that seem to have erupted not so long ago, and small gullies that are a clear sign of some sort of activity in the past 7 years. But what makes Mars so special, and a place like no other in the solar system, can be found by searching deep into the human psyche. It’s about a postcard sunset over a hill, as imaged by a rover that landed 6 years ago on a giant impact crater basin; billions of marble-like spherules abandoned on a desolated plain and layered rocks sculpting a book of stone that tell tales of more clement times; a 24-hour day; night skies where Orion rises as it does on Earth during winter; four seasons punctuating a year; faint icy clouds passing in the sky; and dust devils and sandstorms and hills, volcanoes, deserts, dunes, mountains, canyons, and polar caps. There is no need to invent words to describe Mars. They have been in our vocabulary since the dawn of our existence since despite all the differences, Mars is for us the closest place to home in the solar system. Its frozen landscape has kept the record of a past not so dissimilar to ours. And that landscape might have preserved clues, long gone from our own planet, of how life originated. Mars is the keeper of our past. It also offers the promise of new beginnings as mankind’s first home away from home, our first step as an interplanetary civilization. The time might not be far off when oases and canals will again flourish on the surface of Mars, and beings will visit our blue planet in spaceships. But this time it won’t be science fiction and they will be human. Nathalie A. Cabrol

List of acronyms

AURA BP COSPAR CRISM CTX DEM DOAS ESA GES-DISC

NASA mission to study Earth’s ozone, air quality, and climate. Before Present Committee on Space Research Compact Reconnaissance Imaging Spectrometer for Mars Context imager (Mars Reconnaissance Orbiter) Digital Elevation Model Differential Optical Absorption Spectroscopy European Space Agency NASA’s Goddard Earth Sciences (GES) Data and Information Center (DISC) GIOVANNI GES-DISC Interactive Online Visualization ANd aNalysis Infrastructure GIS Geographic Information System GPS Global Positioning System GRS Gamma Ray Spectrometer (Mars Odyssey) HiRISE High Resolution Imaging Science Experiment (Mars Reconnaissance Orbiter) HLP High Lakes Project HRSC High Resolution Stereo Camera (Mars Express) IPCC Intergovernmental Panel on Climate Change IRTM Infrared Thermal Mapper (Viking mission) LHB Late Heavy Bombardment MARSIS Mars Advanced Radar for Subsurface Ionosphere Sounding (Mars Express) MAX-C Mars Astrobiology Explorer-Cacher (Potential Rover Mission to Mars) MEPAG Mars Exploration Program Analysis Group MER Mars Exploration Rover mission MEx Mars Express MGS Mars Global Surveyor MOC Mars Orbiter Camera (Mars Global Surveyor) MOLA Mars Orbiter Laser Altimeter (Mars Global Surveyor) MRO Mars Reconnaissance Orbiter MSL Mars Science Laboratory MSS Multispectral Scanner, (Landsat mission) NA Narrow Angle, MOC camera (Mars Global Surveyor) NASA National Aeronautics and Space Administration

xii

NIR NRC ODY OMEGA OMI PDS PREVCOM SHARAD SR-SAG TES THEMIS TOMS VIS WA

List of acronyms

Near Infrared (imagery, i.e., THEMIS NIR) National Research Council Mars Odyssey Observatoire pour la Minéralogie, l'Eau, les Glaces, et l'Activité. Ozone Monitoring Instrument, (EOS AURA spacecraft) Planetary Data System Preventing the Forward Contamination of Mars Shallow Subsurface Radar (Mars Reconnaissance) Special Regions -Science Analysis Group Thermal Emission Spectrometer (Mars Global Surveyor) Thermal Imaging System (Mars Odyssey) Total Ozone Mapping Spectrometer; measures back scattered radiances in the near UV Visible (imagery, i.e., THEMIS VIS) Wide angle, MOC camera (Mars Global Surveyor)

List of contributors

Gaetano di Achille, Research and Scientific Support Department, European Space Agency, ESA-ESTEC, Noordwijk, The Netherlands Matthew R. Balme, Department of Earth and Environmental Sciences, the Open University, Walton Hall, MK, UK; The Planetary Science Institute, Tucson, AZ, USA Nathalie A. Cabrol, SETI CSC/ NASA Ames, Space Science and Astrobiology Division, Moffett Field, CA, USA Michael H. Carr, U. S. Geological Survey, Menlo Park, CA, USA Mary G. Chapman, Planetary Science Institute, Tucson, AZ, USA Guillermo Chong, Centro de Investigación Cientifica y Tecnológica para Minería (CICITEM), Antofagasta, Chile Rene A. De Hon, Department of Geography, Texas State University, San Marcos, TX, USA Cecilia Demergasso, Centro de Biotecnología, Universidad Católica del Norte, Antofagasta, Chile Brenda J. Franklin, Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, USA Colman J. Gallagher, University College Dublin School of Geography, Planning and Environmental Policy, University College Dublin, Belfield, Dublin, Ireland John A. Gibson, Marine Research Laboratories, Tasmanian Aquaculture and Fisheries Institute, University of Tasmania, Hobart, Australia John A. Grant, Center for Earth and Planetary Studies, National Air and Space Museum, Smithsonian Institution, Washington, DC, USA Edmond A. Grin, SETI CSC/ NASA Ames, Space Science and Astrobiology Divi­ sion, Moffett Field, CA, USA Donat P. Häder, Department of Botanic, University of Erlangen, Erlangen, Germany Keith P. Harrison, Southwest Research Institute, Boulder, CO, USA James W. Head, Department of Geological Sciences, Brown University, Providence, RI, USA

xiv

List of contributors

Alan D. Howard, Department of Environmental Sciences, University of Virginia, Charlottesville, VA, USA Brian M. Hynek, Laboratory for Atmospheric and Space Physics, University of Colorado, Boulder, CO, USA; Department of Geological Sciences, University of Colorado, CO, USA Rossman P. Irwin, III, Center for Earth and Planetary Studies, National Air and Space Museum, Smithsonian Institution, Washington, DC, USA Jung-Rack Kim, Department of Space & Climate Physics, University College London, Mullard Space Science Laboratory, Dorking, RH, UK Darlene Lim, SETI CSC/ NASA Ames, Space Science and Astrobiology Division, Moffett Field, CA, USA Baerbel K. Lucchitta, U. S. Geological Survey, 2255 N. Gemini Dr. Flagstaff, AZ 86001, USA Edwin Minkley, Department of Biological Sciences, Carnegie Mellon University, Pittsburgh, PA, USA Jeffrey M. Moore, Space Sciences Division, NASA Ames Research Center, Moffett Field, CA, USA Jan-Peter Muller, Department of Space & Climate Physics, University College London, Mullard Space Science Laboratory, Dorking, RH, UK John B. Murray, Department of Earth and Environmental Sciences, the Open University, Walton Hall, MK, UK Horton E. Newsom, Institute of Meteoritics, University of New Mexico, Albuquerque, NM, USA Gian G. Ori, International Research School of Planetary Sciences Universita’ d’Annunzio Viale Pindaro, Pescara, Italy; Ibn Battuta Centre, Faculté des Sciences, Université Caddy Ayad, Boulevard de Safi Marrakech, Morocco David P. Page, Planetary and Space Science Research Institute, the Open University, Walton Hall, MK, UK Timothy J. Parker, Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, USA Don E. Wilhelms, Branch of Astrogeology, U.S. Geological Survey, Reston, VA, USA Sharon A. Wilson, Center for Earth and Planetary Studies, National Air and Space Museum, Smithsonian Institution, Washington, DC, USA Youngseob Yu, Department of Civil and Environmental Engineering, Carnegie Mellon University, Pittsburgh, PA, USA

Foreword

One of the privileges that goes with writing the Foreword to a science book is the opportunity thereby afforded to introduce some topics that rarely if ever get presented via the conventional modes of scientific discourse. These are topics in regard to which scientists tend to presume the truth of a particular position, generally without thinking about the basis for that presumption. Given that one can arguably define science as the organized questioning of all presumptions about the natural world, there is an obvious contradiction here. A book dealing with the controversial a topic of ancient lakes on Mars provides a great chance to discuss little-questioned scientific presumptions in the context of some exciting new discoveries. Ever since humans began to look inquisitively at the skies, they have been fascinated by the red, wandering speck of light that came to be known as the planet Mars. With the advent of the telescope, Mars became another world, but one that held fascinating similarities to our own. In 1794 the English astronomer Sir William Hershel announced the uncanny earthlike nature of Mars. Hershel’s telescopic obser­ vations revealed a Martian day that lasted about 24.6 hours, an axis of rotation tilted to approximately the same degree as that of Earth, and, most surprisingly, cyclic seasonal growth and decay of polar caps. By the late nineteenth century, greatly improved telescopes were viewing a planet that most astronomers then believed to have oceans, polar snows, arid landmasses, and clouds. It was not too difficult for many to extend the Earth analogy even further and speculate on Mars’ habitability for life. There were many problems for telescopic observations of Mars, including the relatively short periods during which the two planets came in close proximity to one another, and the distortions introduced by observing Mars through two planetary atmo­ spheres. It was during the favorable viewing conditions of 1877 and 1879 that Italian astronomer Giovanni Schiaparelli, during brief periods of “favorable seeing” (apparent clarity of view through the two atmospheres), identified Mars’ classic bright and dark areas, the latter being designated as “mare” (oceans and seas). Thus, the idea of water bodies on the surface of Mars was born during the period of telescopic observation. Moreover, it was Schiaparelli who gave the surface features of Mars the names from classical mythology that we continue to use today: Amazonis, Argyre, Chryse, Elysium, Hellas, Tempe, Tharsis, and many others. It was also Schiaparelli who confirmed Father Pietro Angelo Secchi’s observation of apparent lines on the planet’s surface, which he designated as “canali” (Italian for “channels”). Schiaparelli considered the “canali” to be natural channels that connected the various mare, in much the same way that the English Channel connects the North Sea to the Atlantic Ocean on Earth. The turn of the century, from nineteenth to twentieth, marks a key point in what became, over the next hundred years, oscillating changes in scientific thinking about

xvi

Foreword

Mars. These oscillations involved changes in the collective view held by the commu­ nity of those scientists who were most intensively involved in Mars study. In the public imagination, the high point in the oscillations was probably reached in 1906 or 1908, though for most scientists these years may be considered the low point. These years marked, respectively, the publication of Percival Lowell’s Mars and Its Canals and Mars as an Abode of Life. Percival Lowell was a wealthy Boston Brahmin, who invested his considerable personal fortune and used his many business connections to build an astronomical observatory in pursuit of his vision of Mars inhabited by intelligent creatures who constructed a global network of canals to irrigate their dying, arid planet. However, the defects of an individual, idiosyncratic vision readily become apparent when scrutinized through the skeptical lens with which many scientists view the extreme claims of individuals. Science is a collective and cumulative enterprise, and Lowell’s “canals” were soon discredited by the more advanced telescopic observations of the later twentieth century. Nevertheless, a general view of Mars as a somewhat earthlike planet, involving water transfers and even plant life, persisted to mid-century. It was a physicist-turned-historian/philosopher, Thomas Kuhn, whose 1962 book The Structure of Scientific Revolutions showed that scientific communities ascribe to certain ways of thinking, or worldviews, that predispose them to see things in a particular way, even blinding them to alternative possibilities. Kuhn introduced the terminology of “paradigms” to describe this kind of collective thinking in science, though the concept was applied to more broad historical periods of what Kuhn termed “normal science.” In Kuhn’s model of science, paradigms come into question when anomalous discoveries build up to such a degree that failure by the currently prevail­ ing view to explain them leads to a revolution in thinking, as in the case of the Copernican view of the solar system. The oscillations in scientific views about Mars were certainly much less grand in scale and intensity than what Kuhn was describing as “paradigms,” but they did involve prevailing views and the build up of anomalies in regard to those views, which I will here label as “conceptual positions.” The changes in conceptual positions about Mars occurred relatively rapidly because of the rapid pace of discoveries made during the later part of the twentieth century, when a succession of improved space­ craft observations replaced the problematic telescopic ones. Another difference from Kuhn’s model is that the conceptual positions for Mars oscillated between two end members in regard to the role of water on the planet’s surface. These are positions that can be termed “hydrophilic” (water loving, or water friendly) and “hydrophobic” (water fearing, or water avoiding). The viewpoint of Schiaparelli and his contemporaries in the late 1800s was hydrophilic in regard to Mars. The visible polar caps and their seasonal changes along with the presence of apparent water bodies suggested that water was cycled about on the planet in a manner that was only known for one other body in the universe, Earth. Regrettably, the hydrophilic vision of Mars became tainted by the work of Lowell, whose enthusiasm for the intelligent design of the “canals” led to a sham science in which the answer (intelligent life on Mars) came before the clear refinement of the observations necessary to both inspire and support that answer. The

Foreword

xvii

example of Lowell and his scientifically false position of dogmatism about intelligent life on Mars may even continue to fuel a scientifically false alternative philosophical position of skepticism, wherein the lack of certainty about particular phenomena is used to dismiss the general scientific beliefs that arise from the synthesis of many lines of evidence. Authentic science requires an attitude that is neither dogmatic nor skeptical. Instead, must scientist be a fallibilist, i.e., one who generally holds wellsupported scientific ideas (theories) to be probably true, but who also holds to a degree of uncertainty in, and thus a need to test further, any particular instance or conse­ quence of those ideas. As Mars science progressed through the twentieth century, the conceptual position of the community of Mars scientists gradually shifted toward the hydrophobic end member. Views subsequently began to oscillate as a succession of spacecrafts delivered improvements in spatial and spectral resolution that generated discoveries, thereby posing anomalies for whatever viewpoint was currently prevailing. On November 28, 1964, the Mariner 4 spacecraft returned 22 low-resolution flyby images of some cratered terrain, revealing a lunar-like surface that appeared to have been little modified over 4 billion years by any process other than the wind erosion of a cold, dry surface by an extremely thin atmosphere. This seemed to provide the hydrophobic antidote to the Lowell’s hydrophilic excesses. This conceptual position came under threat, starting in 1972, when a global dust storm cleared and the orbiting Mariner 9 spacecraft cameras produced imagery for the entire planet at low resolution, supplemented by a small sample of high-resolution (100-m) frames. Mariner 9 and subsequent, improved imagery from the Viking orbiter spacecraft revealed channels and valleys that were clearly formed by a flowing, free-surface fluid. The obvious candidate for this fluid, based on well-documented studies on Earth, was water. Nevertheless, this working hypothesis was contrary to the then-prevailing hydropho­ bic conceptual position. Mariner 9 and Viking spacecrafts provided new higher-resolution images to sti­ mulate Mars science through the discovery of many other water-related landforms, including the first images of the lacustrine features that are described in this book. The abundance of new imagery data also brought into play another kind of viewpoint that would contribute to the oscillation of hydrophobic and hydrophilic thinking. Before the spacecraft images of the 1970s and later, nearly all the thinking about Mars was from the methodological scientific view characteristic of physics. This view, built on the incredible success of experimental methods, involved first principles (established laws in physics and chemistry), combined with various initial conditions or constraints (including the presumed causes of the phenomena), to formulate models that predicted various expected outcomes, including the water-related features that might be observed on the surface of Mars. The actual observation of those features would then indicate the truth or falsehood of the claimed causes (such as the cycling of water). The new view, made possible by the abundance of imagery data, was geological. By geological I do not mean the assembly of facts about Earth (the “geo” part). I refer here to the mode of reasoning (the “logical” part). In contrast to the norm for physicists, geologists have long made their inferences in an inverse manner. The

xviii

Foreword

geologist starts by paying special attention to the outcomes of various processes. The natural outcomes of causative processes on Earth’s land surfaces include lakes, rivers, mountains, valleys, glaciers, hillslopes, and sand dunes. The geologist treats these features as signs or indicators of their associated formative processes. Thus, geologi­ cal reasoning involves a kind of semiosis, i.e., a system of signs, which is the continuous, back-and-forth mental interaction of the geological investigator with the signs or clues encountered in the natural world. The signs relate to one another causally, in a similar way that fossil fishes are related to the organisms that preceded their fossilization. The signs also trigger perceptions and ideas to those who are especially experienced in understanding their message, while still realizing the fallible character of any such interpretation. The coherence of numerous such observations leads to the formulation of hypotheses that account for the sequence of temporal and spatial associations of landforms recorded by the spacecraft images. The test of such an interpretation is not a controlled experiment, as would be the ideal in physics. Instead, it is the landscape itself that serves in the role of the “experiment,” and the geological investigator looks for consistency, coherence, and consilience in the overall relationships of the landscape features and their inferred causal processes. (Consili­ ence is the explanatory surprise generated when the working hypothesis for a general set of causes leads to finding unexpected phenomena that can be subsumed that same set of causes, thereby suggesting the likelihood of some connection in reality among all the phenomena and their causes.) The logic outlined in the previous paragraph will be familiar to readers of detective fiction, in which the master investigator assembles clues in such a way as to achieve a resolution of the crime, generally at the end of the story. However, an important difference is that for science the story never ends, which brings us back to the discussion of oscillations in conceptual positions about water on Mars. After the Viking era of the late 1970s, a succession of spacecraft failures and programmatic decisions led to a prolonged drought in new data bearing on the general problem of water on Mars. Though very interesting ideas had been generated about large-scale, ancient inundation of the northern plains of the planet and the presence of numerous smaller-scale paleolakes, mainly occupying impact craters, such wide­ spread aqueous activity still seemed incongruous with hydrophobic theories that held Mars to be, and generally to have been, extremely cold and dry throughout its history (at least since about 4 billion years ago). Through the 1980s and 1990s, a variety of nonaqueous hypotheses were posed to explain individual elements of the putative water-related assemblage of landforms. One extreme in this hydrophobic renaissance was the “White Mars” proposal that the planet had always been so cold and dry that water was never liquid on its surface. Instead, various fluid-related features on the surface of Mars were all inferred to derive from the phase transitions and flow properties of carbon dioxide. “White Mars” might explain the lack of evidence for carbonates on Mars and why some of the infrared spectral data generated by the Mars Global Surveyor in the late 1990s revealed unaltered mafic minerals over large areas of the planetary surface. However, the issue with the hypothesis was not the specific phenomenon that it (like other hypotheses) did explain, thereby appealing to hydrophobic purists. Instead, there was a holistic problem in that “White Mars”

Foreword

xix

would need to account for a whole assemblage of obviously interrelated landforms that can clearly develop from a nexus of interrelated and connected chains of causa­ tion involving the hydrological cycle on Earth. Is it physically reasonable to invoke such an extreme example of adaptive mimicry in a nonbiological system? The revived hydrophobic skepticism about Mars having had an abundant endow­ ment of water could not be allayed by the geological interpretation of planetary imagery. The discovery of young gullies on imagery generated by the very highresolution camera on Mars Global Surveyor simply led to more controversy. Non­ aqueous alternatives were considered to provide reasonable, hydrophobic alternatives to an assemblage of features that had nearly identical counterparts in cold/wet periglacial environments on Earth. Similarly, paleolacustrine features were alterna­ tively explained by pondings of lava. The hydrophobic proponents even managed to mix the differences in methodological viewpoints with their conceptual alternatives. Thus, there was sometimes a tendency to equate geological reasoning with a lookslike-a-duck-must-be-a-duck sort of triviality that would be an insult to the methodo­ logical positions argued for analogical reasoning in geology by such luminaries as Grove Karl Gilbert and T. C. Chamberlin. The most recent switch from hydrophobic to hydrophilic was not achieved until several discoveries were made in this millennium through the physical and chemical measurements by new instruments. Most telling in this regard was the application of nuclear physics, via neutron and gamma ray detections by the Mars Odyssey space­ craft. These data documented extensive water ice (later confirmed by the Phoenix lander) in near-surface soils at high latitudes. The radar instrument on Mars Recon­ naissance Orbiter revealed even more water ice, in this case constituting debriscovered glaciers that would have had to be replenished by cyclic water transfer through the atmosphere. Even more anomalous for the hydrophobic advocates was the hyperspectral detection of clay minerals and evaporites, the latter being documen­ ted, along with other aqueous alteration products, by the Mars Exploration Rover Mission. The chapters that follow in this book are all aligned with the current hydrophilic view of Mars, though some still retain hydrophobic vestiges, perhaps a legacy of the skepticism that crept into Mars science more than a century ago. While working as an instrumentation consultant to NASA during the early 1960s, James Lovelock became fascinated with the atmosphere of Mars. Data available at the time revealed the Martian atmosphere to be in chemical equilibrium, composed almost entirely of carbon dioxide, which could not be reactive under prevailing conditions of extreme cold and desiccation. By contrast, Earth’s atmosphere, with its abundant oxygen, was far from equilibrium. Moreover, the disequilibrium of Earth’s atmo­ sphere derived from the presence of a biosphere. Of course, a consequence of this line of thinking was Lovelock’s formulation of the “Gaia” hypothesis, which posits complex feedbacks whereby a biosphere is able to regulate the planetary environment in a way that sustains itself. Though Lovelock’s vision of a living planet has drawn considerable criticism from the science community, it has been inspirational to many in public, including many environmentalists. Lovelock’s criterion for the detection of a planetary biosphere involves identifying an atmosphere that is in disequilibrium. The recent detection of methane in the

xx

Foreword

Martian atmosphere now meets that criterion, though I am sure that Lovelock would be worried about the rather small quantities of the gas. Is methane a product of nonbiological processes, such as the hydrothermal alteration of olivine? Might it also be the product of a subsurface biosphere of methanogenic microbes? Such a biosphere could exist in the warm groundwater that likely underlies areas of ice-rich permafrost. Perhaps the methane produced by this biosphere, or by its past activity, was sealed into the ground ice at depth through long-term sequestration as gas hydrates that are now slowly destabilizing at local “hot spots,” thereby releasing the gas. Perhaps there was an even more diverse biosphere, possibly now in hibernation, that was temporarily extant on the Martian surface during episodes of enhanced volcanic activity. These episodes would have coincided with the periods of lake formation, described in the subsequent chapters. The ancient lake deposits would be where geologists would like to look for the fossil evidence of that biosphere. The image of seeing beneath or beyond the surface appearances of things to something that inspires awe and wonder is usually thought more to be the province of the arts or religion than of science. Though science tends to get expressed in writings about fact and theory, wonder and awe are, nevertheless, essential to the process of doing science. These feelings come to scientists in the zeal to understand and the inspiration to hypothesize. Scientists do not get to write about such things in their formal scientific papers, and you will not see poetic descriptions of exactly how the scientific process unfolded for the writers of the chapters that follow. However, I know most of these authors, and it is on their behalf that I can assure you that they experienced this same sense of wonder in the process of achieving the results that they report herein. Victor R. Baker Department of Hydrology and Water Resources The University of Arizona Tucson, AZ, USA

1 Searching for lakes on Mars: four decades of exploration

Nathalie A. Cabrol and Edmond A. Grin SETI CSC/ NASA Ames, Space Science and Astrobiology Division, Moffett Field, CA, USA

1.1

Introduction

Lakes are time capsules. On Earth, they are considered sentinels of climate change (Williamson et al., 2009) and may have played the same role on early Mars (e.g., Baker, 2001; Cabrol and Grin, 2002, 2005). Their basins capture the record of geological and environmental fluctuations over a wide range of temporal and spatial scales (Figure 1.1). Terrestrial lakes host a diversity of habitats where life’s adaptability can be pushed to the edge in often unstable environments (e.g., Bronmark and Hansson, 1998; Cabrol et al., 2009; Herbst, 2001; Hollibaugh et al., 2001; Oren, 2001; Price, 2000; Stivaletta et al., 2009; Warwick and Laybourn-Parry, 2008; Wynn-Williams and Edwards, 2000; see also Chapter 13). They preserve the evidence of ancient life as sedimentation rapidly entombs dead organisms and generates anoxic conditions favoring the formation of fossils (e.g., Beaty et al., 2005; Des Marais and Farmer, 1995; Des Marais et al., 2003, 2008; Farmer, 1999, 2003; Hoffman et al., 2008). This makes them prime candi­ dates for exploration. The existence of lakes on ancient Mars is now widely accepted but that was not always the case. The history of science shows that knowledge on any scientific question is shaped by the means of exploration and those means are molded by what we think the world is. Prior to MGS of the late 1990s, the relatively low resolution of orbital imagery made it difficult to confirm Martian paleolakes by direct observations, though their existence was inferred because valley networks had already been identified on Viking and Mariner 9 images. Interpretation rested on ambiguous morphological evidence at 200 m/pixel on average with only localized coverage at higher resolution. Yet, physical conditions measured at Mars and early modeling (Baker et al., 1991; Clifford, 1993; Haberle and Jakosky, 1990; Haberle et al., 1993; Jakosky and Phillips, 2001; Kasting, 1997; McKay and Davis, 1991; Sagan and Mullen, 1972; Sagan et al., 1973; Toon et al., 1980) supported the hypothesis that channels and valleys had been carved by water early in Martian history. Under such conditions, lakes could have formed in topographic lows unless water had completely evaporated and infiltrated before reaching basins. However, topographic uncertainties made for difficulty in identifying with confidence depressions that might have hosted paleolakes. The exception, of course, was the obvious basin Lakes on Mars. DOI: 10.1016/B978-0-444-52854-4.00001-5 © 2010 Elsevier B.V. All rights reserved.

2

Lakes on Mars

Figure 1.1 The existence of ancient lakes on Mars is now supported by converging evidence provided by both mineralogy and sedimentary records. Here, the Holden basin near the mouth of Uzboi Vallis has become one of the primary candidate sites for MSL (see Chapter 12). HiRISE image PSP_001666_1530, NASA/JPL/University of Arizona.

morphology created by impact craters, and this explains why much of the early Mars paleolacustrine research emphasized these features (Cabrol, 1991; Cabrol and Grin, 1995; De Hon, 1992; Forsythe, 1990; Forsythe and Blackwelder, 1998; Forsythe and Zimbelman, 1995; Goldspiel and Squyres, 1991; Grin and Cabrol, 1997; Kuzmin et al., 2000; Newsom et al., 1996; Scott et al., 1991, 1995). Although valley networks were identified, they appeared poorly integrated (see Section 1.2.2 and Chapters 2 and 8), and their origin by water was questioned. This set the stage for other hypotheses, such as lava (Carr, 1974; Greeley, 1973; Schonfeld, 1977), liquefaction of surface material (Nummedal, 1978), glaciers (Lucchitta, 1982), CO2 (Hoffman, 2000; Sagan et al., 1973), or the dissociation of clathrates (Baker et al., 1991; Kargel and Lunine, 1998; Max and Clifford, 2000; Mellon, 1996; Miller and Smythe, 1970; Milton, 1974; Musselwhite and Lunine, 1995; Peale et al., 1975; Yung and Pinto, 1978). Each of these hypotheses bears considerable limnological and astrobiological implications. As our understanding of Mars grows, especially with regards to the mineralogical nature of sedimentary deposits, the hypoth­ esis that valley networks and channels were formed by CO2 and/or clathrates seems less probable, although it cannot be completely dismissed (Kargel et al., 2000). It remains a plausible mechanism for the formation processes associated with outflow channels in the equatorial regions (Baker et al., 1991; Kargel et al., 2007; Komatsu et al., 2000). In a planet strongly influenced by volcanic processes for much of its history, it is not surprising that the study of high-resolution imagery has confirmed in greater detail

Searching for lakes on Mars: four decades of exploration

3

Figure 1.2 Elysium Planitia platy flows have been interpreted either as evidence for a frozen sea (see text and references therein) or lava flow. HiRISE image TRA_000854_1855, NASA/ JPL/University of Arizona.

the major role played by lava in the formation of many channels (Basilevskaya et al., 2009; Keszthelyi et al., 2006; Sakimoto and Zuber, 1998; and others). Lava lakes formed in basins and calderae during volcanic activity (Gregg and Lopes, 2007; Lang and Farrell, 2009; Mouginis-Mark and Robinson, 1992; Mouginis-Mark and Wilson, 1999) and, as it was the case in the Viking era (Milton, 1973), lava is still at the heart of the controversy when discussing the nature of the Martian channels, valleys, lakes, and ponds (Bleacher et al., 2010; Chapman, 1999; De Hon, 1992; Jaeger et al., 2008; McEwen et al., 2002; Scott and Chapman, 1991; and others). The putative paleolakes hosted by the Hellas and Elysium basins are the most recent examples of such ongoing debate (Chapters 6 and 7) and show the constraints and limitations of orbital explora­ tion even with high-resolution imagery and multispectral data (Figure 1.2). However, there are now converging morphological, geological, and mineralogical lines of evidence that reasonably support the existence of ancient standing bodies of water on Mars. These are the features discussed in this chapter and defined as lakes.

1.2

Conditions at the time of lake formation

Whether the early climate of Mars was much warmer and wetter in the Noachian compared to the later geological epochs is still the subject of ongoing debate. Many lakes do not require warm conditions to form (Chapter 4). On Earth, 60% of them are located at high latitudes in the northern hemisphere (ILEC web-based database), many inherited from the last deglaciation. But lakes require water to form, and water balance drives their evolution and duration. Therefore, uncertainties about climate and water on Mars weigh on our understanding of the Martian limnology. Current theories about the acquisition and evolution of water and early Martian climate are presented by Carr

4

Lakes on Mars

and Head in Chapter 2 (and references therein), thus are not developed here. The premise of our discussion is the following: The existence of standing bodies of water on Mars required that at some point in its history, possibly repeatedly, physicochem­ ical and environmental conditions allowing water to circulate and to pond were met. Any conditions allowing lake formation will be referred to generically as favorable conditions in the remainder of this chapter. In this section, we examine the collective mineralogical and morphological evidence, assess the range of plausible environments consistent with lake formation, and determine the significance of these environments in terms of relative water abundance and climate at the time of lake formation.

1.2.1 Clues from mineralogy Discrete ranges of physical environments can account for most of the temperaturedependent minerals and mineral assemblages that have recently been identified on Mars (Figure 1.3). Data suggest that, in addition to the largely mafic bedrock of the

Figure 1.3 Diverse mineralogy on a plateau near in Juventae Chasma, Valles Marineris 4.7°S, 296.4°E. The deposits contain opaline silica and iron sulfates, consistent with low-temperature, acidic aqueous alteration of basalt that could be the result of surface runoff and fluvial deposition during the Hesperian (Mustard et al., 2008). HiRISE image PSP_003579_1755, NASA/JPL/University of Arizona.

Searching for lakes on Mars: four decades of exploration

5

Martian surface, there are components of both (i) a mobile dust fraction and (ii) a chemically altered in situ bedrock (Bibring et al., 2006). Iron oxides (Christensen et al., 2000, 2001; Glotch and Rogers, 2007) and hydrated minerals have been detected in outcrops (Bibring et al., 2005, 2006; Mustard et al., 2008). Their strati­ graphic relationships suggests that mineralization of phyllosilicates essentially took place in the Noachian (Bibring et al., 2005; Mustard et al., 2008; Poulet et al., 2005). At Nili Fossae, carbonate-bearing rocks are present in layered exposures (Elhmann et al., 2008) associated with phyllosilicates and olivine-rich strata of comparable ages. The mineralogical sequence inferred from stratigraphy is consistent with alkaline to neutral waters during the early Noachian and acidic weathering during the Hesperian (Bibring et al., 2005, 2006; Mustard et al., 2008). The preservation of carbonates through the early Hesperian (Ehlmann et al., 2008) may indicate local to regional exceptions. The primary process of clay formation as hydrothermal processes through volcan­ ism and impact cratering may have included the weathering of Fe/Mg mafic minerals, as inferred from their dominance relative to plagioclase (Mustard et al., 2008). From an environmental standpoint, the presence of clays is consistent with, but does not necessarily require, sustained warm and wet surface conditions on early Mars (Catling, 2007). The existence of a volatile-rich subsurface during this early period is supported by the presence of lobate ejecta craters (Barlow et al., 2000 and others) and ancient water activity associated with hydrothermal activity (Bishop et al., 2008; Gulick, 1998; Neukum et al., 2004; Squyres et al., 2007), which provided a mechan­ ism for clays to form through interaction between the Martian crust and hydrothermal waters. It is also possible that the current locations for some of the deposits may not be relevant to the source areas for their contained minerals. The relevant depositional histories might include (i) fluvial erosion from distant source areas and sediment transport, (ii) in situ precipitation from solutions in lakes, or (iii) reactions of amor­ phous material in response to in situ conditions. Current deposits may be also evidence that conditions remained stable long enough for the minerals to remain inert in the particular environment where they were discovered. Alternatively, they may have been exposed to the surface only recently. Linking the recent discovery of clays and their geographical location with ancient surface conditions and global processes is, therefore, not straightforward. Their relation to an early warm and wet climate must be assessed together with morphological evidence on an individual basis. Sulfate deposits have been identified from orbit (Bibring et al., 2006; Gendrin et al., 2005) and from the ground by both MER rovers (Arvidson et al., 2006; Grotzinger et al., 2005; McLennan et al., 2005; Squyres et al., 2004; Tosca et al., 2005). Their abundance and distribution represent an arguable case for widespread evaporitic environments involving surface to near subsurface water processes extend­ ing into the Hesperian. Their formation is interpreted to be the result of sulfur and water release during volcanic activity, production of H2SO4 from oxidation of sulfur in the atmosphere, and precipitation (Bibring et al., 2006). Liquid water circulation at the surface could have been enabled by volcanism-induced greenhouse effects through sulfur dioxide climate feedback (Haveli et al., 2007), while increased regional

6

Lakes on Mars

heat flow might have facilitated subsurface water circulation. By raising temperature, sulfur dioxide climate feedback may have contributed to the wet deposition of sulfur through precipitation (e.g., Fujita et al., 2003). Climate models predict the likelihood of snow precipitation at the equator and in the highlands (Forget et al., 2006; Haberle et al., 2001), and these results are consistent with regional studies that combine geological evidence and modeling (Baker et al., 1991; Carr, 2003; Clifford and Parker, 2001; Gulick et al., 1997; Haberle et al., 2001; Moore et al., 1995) and with studies of runoff production rates that imply rainfall (Irwin et al., 2005).

1.2.2 Clues from morphology The proposition that lakes existed on ancient Mars finds key support from the physical connection of various paleolakes to channels and valley networks (Figure 1.4). Although, the contribution of volcanism and impact cratering to the Martian hydro­ graphic system is well-supported and generally accepted as evidence for local to regional water release, both are climate-independent. Their only causal relationship to climate is their potential to temporarily alter it. In contrast, valley networks and the deltaic termini of channels are central to the issue of climate on early Mars.

Valley networks The origin, morphometry, ages of (re)activation, and duration of valley networks have been the subject of numerous studies reviewed in Chapters 2 and 8 (and references therein), whose conclusions have significant implications for the formation of lakes and their duration. The immaturity of the Martian hydrographic system is a key observation (Aharonson et al., 2002; Baker and Partridge, 1986; Irwin et al., 2002).

Figure 1.4 Channel entering the Jezero impact crater (18.9°N, 77.5°E) in the Nili Fossae region to form a delta. Topographic correlation between delta and terraces in the basin argues for the presence of a body of water. THEMIS image V16660006, NASA/JPL/Arizona State University.

Searching for lakes on Mars: four decades of exploration

7

With Viking, valley networks were found to be diffuse (Pieri, 1980), with irregular tributary junction angles (Cabrol, 1991; 1993) and large undissected intervalley areas showing no compelling evidence that precipitation were involved (Carr, 1996; Malin and Carr, 1999). Instead, their formation was linked to headward extension due to basal sapping (Pieri, 1980). At high resolution, lower tributary orders are observed and networks appear more integrated. Drainage systems remain, however, immature if formed by runoff (Craddock and Howard, 2002; Grant, 2000; Gulick, 2001; Irwin et al., 2005) with possible exceptions (Hynek and Phillips, 2003). The lack of tributaries smaller than 100 m across have been attributed to either a lack of precipitation, or a lack of runoff due to high infiltration rates. It has also been suggested that non-fluvial erosive or depositional processes might have obliterated all valleys up to a maximum width (Hartmann and Neukum, 2001). High erosion rates (Craddock and Howard, 2002) and the occurrence of valley formation during heavy bombardment are consistent with both a greater supply of water and with warmer conditions, but these deviations from the present state did not have to be drastic (Chapter 2). An important clue may have been found by CRISM in the relative paucity of kaolinite (Mustard et al., 2008), which points toward limited hydrologic activity through space and time, and burial of deposits that have been shielded from further precipitation-related aqueous alterations. Hecht (2002) showed that flow would have been possible on Mars at nearly freezing temperature. Conditions supporting hydrologic and limnologic activity in a cold Mars environment have been investigated through field analogs in polar regions (Doran et al., 2010), and at high altitude (Cabrol et al., 2009; Chapter 13) using Terrestrial and Martian data. Ice-covered streams and lakes can survive over extended periods of times (Morgan and Head, 2009). Ice provides thermal insulation from outside temperature fluctuations and protection from evaporation (McKay et al., 2005; Wallace and Sagan, 1979). Subglacial rivers are sustained in non-turbulent flows, until ablation removes the ice or until flow stops. Eutectic solutions of single salts with modest freezing point depression such as NaCl could allow sporadic liquid water flow anywhere on Mars (e.g., Clark et al., 2005; Grimm and Stillman, 2008; Hecht, 2002; Hecht et al., 2008). In that respect, the Phoenix mission (Smith et al., 2009) has shown physical and thermodynamical evidence that such brine solutions can form and remain liquid mostly anywhere even under current conditions (Fairén et al., 2009; Hecht et al., 2008; Renno et al., 2009; Seinfeld and Pandis, 2006). Snow precipitation (Smith et al., 2009) was also observed for the first time on Mars by Phoenix.

Deltas Climatic and environmental conditions in the watershed affect discharge, erosion, transport, and deposition. Terminal basins are particularly sensitive to those changes and record them in their sedimentary deposits. Before MGS, large sediment accumu­ lation at the termini of channels and valley networks provided a rare opportunity to evaluate their origin and the climate conditions associated with their formation. Case studies concluded that many of the deposits were short-lived, possibly reactivated, alluvial fans (Moore and Howard, 2005). Other works at global to regional scale

8

Lakes on Mars

proposed the existence of deltas, some interpreted to be inherited from warm and wet conditions (De Hon, 1992; Di Achille et al., 2006; Fasset and Head, 2005; Irwin et al., 2005; Kleinhans et al., 2009; Malin and Edgett, 2003; Mangold and Ansan, 2006; Ori et al., 2000) and others from cold environments (Kraal et al., 2008, and Chapter 8). Mars has seasons and an atmosphere, thus a latitude and elevation-dependent climate much like Earth, accentuated by high-obliquity cycles. Therefore, a wide range of morphology is actually expected and may reflect geographical location and local to global environmental conditions at a given time. Comparison with Earth shows that vast and complex modern deltas can be con­ structed in cold environments. Those forming in cold continental to arctic conditions commonly have winter temperatures as low as –70°C, fluctuating between –40°C and –30°C in January and rising up to +10°C in July (Berezovskaya et al., 2005). This range is not too dissimilar to current-day Mars in the equatorial regions or to the temperatures in the intertropical belt during summer. Delta development is affected by climatic, hydrologic, and geologic factors and can last from tens to thousands of years. Regardless of climate conditions, then, the lifetime of a delta represents only a small fraction of the Noachian’s half-billion­ year duration. Over such vast period, the Martian deltaic record (not alluvial fans) appears limited to a few tens of landforms with relatively small areal development compared to their terrestrial counterparts (Chapter 8). Their original number, as that of valley networks, may have been affected by burial, impact gardening, and erosion but this remains speculative. If they are representative of the original record, this is a significant observation for the interpretation of the early Martian environment. This suggests either (i) short-lived favorable conditions or (ii) localized, accidental water releases. Stepped delta morphology recognized recently on Mars (Kraal et al., 2008) requires only decades to form and can result from a large single discharge (Figure 1.5). On the other hand, Gilbert deltas (Ori et al., 2000) can form over thousands of years. Their duration is discharge-dependent and may indicate favorable conditions either continuously or episodically present for a total of a few thousands of years.

1.2.3 Outstanding questions Ancient lake deposits are similar to time-lapse cameras integrating snapshots of conditions prevailing during their lifetime. In lakes dominated by overland flow from precipitation, this record captures, to a large extent, any global-scale changes in climatic conditions that occur while the lake is active. For limnologic systems essentially relying on groundwater, linking lake water volume, mineralogical deposits, and climate becomes a more complicated proposition.

Long groundwater flow paths This complexity is illustrated by lakes in the Central Andes (4300–6000 m elevation), which is one of the regions exposed to some of the most rapid and severe climate change on Earth (Giese et al., 2002; IPCC, 2001; Montecinos and Aceituno, 2003; Warren et al., 2006; see also Chapter 13). Temperature has increased between

Searching for lakes on Mars: four decades of exploration

9

Figure 1.5 This stepped delta at 8°S, 159°E has been interpreted as evidence of a single basinfilling event that lasted about 10 years (Kraal et al., 2008). CTX image P02_001644_17B, NASA/JPL/Arizona State University.

0.013°C/year and 0.02°C/year from 1932 to 1992 (Rosenblüth et al., 1995) while precipitation rates have fallen 50–75%, mainly between 1979 and 1988 (Legates and Willmott, 1990). Temperature increases are shifting the equilibrium line altitude of glaciers upward. As a result, accumulation areas are shrinking and/or the net accu­ mulation is becoming progressively smaller (Bradley et al., 2006). Lakes between 18°S and 24°S reached high stands in the past 12,500 to 11,000 years (Grosjean et al., 2001) following the glacial retreat at the end of Pleistocene, when they were supplied by glacial melt, 500 mm/year precipitation, groundwater, and hydrothermal activity. Isotopic studies (Aravena, 1995) show that the last two major aquifer recharges occurred between the last deglaciation and the onset of arid conditions with the beginning of the Holocene. Precipitation in the south Bolivian and northern Chilean altiplano for the past 8 years has fluctuated between 30 mm/year and 120 mm/year (Cabrol et al., 2009) with a water balance of –1500 mm/year (evaporation/precipita­ tion) for many lakes (Hock, 2008). However, the effect of climate and precipitation regimes is buffered in other local lakes dominated by long-distance groundwater flow paths from 12,500-year-old aquifers (Nester et al., 2007; Valero-Garcés et al., 2004).

10

Lakes on Mars

In their case, precipitation regimes lack relevance to understanding present fluctua­ tions. Such residence times are not uncommon for terrestrial aquifers (Kazemi et al., 2006), but more knowledge of prior Martian climates and subsurface aquifer para­ meters will be necessary before we can fully understand their impact on Mars.

Aquifer recharge On longer time scales, continuous flow with limited to zero recharge leads systems to shut down. Recharge ultimately comes from precipitation, either directly, as advocated for Mars (Craddock and Howard, 2002; Fassett and Head, 2008; Grant, 2000; Gulick, 2001; Hynek and Phillips, 2003; Irwin et al., 2005) or indirectly via groundwater from lakes, ponds, rivers, or other aquifers. On Earth, the most effective recharge takes place with sustained and regular precipitation under stable seasonal and inter-annual climate conditions (Seiler and Gat, 2007). Recharge is more limited in cold climates because of freezing and thawing of the subsurface. As melting proceeds from the surface down into the frozen soil layers, melt water is consumed by plants (Sugimoto et al., 2002). Alternatively, it is channeled by preferential subsurface flow paths, or contributes to surface runoff (Seiler and Gat, 2007). If the Martian hydrologic system reflects transient and episodic rainfall, then recharge may have been ineffective after the late heavy bombardment (LHB), see Chapter 2. Immature drainage networks could also reflect high rates of infiltration, and/or evaporation, or selective erosion of small valley orders (Harrison and Grimm, 2004, 2008, see also Section 2.4). Aquifer recharge, and surface and subsurface water flow models have been pro­ posed since Viking (Baker, 2001; Baker et al., 1991; Carr, 1996; Clifford and Parker, 2001; Hanna and Phillips, 2005; Mischna et al., 2003; Phillips et al., 2001; Russell and Head, 2003). Conclusions are affected by uncertainties over how representative the current hydrographic and aqueous sedimentary records are of the original record. Moreover, there are key unanswered questions in regard to the recharge itself, including (i) does the topographic catchment actually correspond to the groundwater catchment, and (ii) what may have been the stability of these systems over time. Basal polar melting was proposed as a mechanism for aquifer recharge at global scale (Clifford, 1993), and lithostatic loading from the rise of Tharsis and Elysium as a mechanism for large-scale redirection of groundwater flow paths in the equatorial region (Baker, 2001; Dohm et al., 2007; Solomon et al., 2005). But hydrographic and limnologic systems are also extremely sensitive to their local and regional environ­ ment, and geological conditions were met on early Mars to produce a regionally and locally compartmented subsurface, which is supported by modeling (Harrison and Grimm, 2009). On one hand, up to the end of the Noachian, the Martian crust was still being intensely bombarded, resulting in pronounced fracturing of the uppercrust. On the other hand, by-products of impact cratering, such as melt material and clays were producing seals and confining layers. Impact cratering provided an effective engine for the creation and destruction of surface and subsurface waterways, for establishing new topography, and for episodically redirecting flow by closing existing flow paths and opening new ones (Howard, 2007). Although local aquifers may have been

Searching for lakes on Mars: four decades of exploration

11

depleted, redirection of subsurface flow paths was a mechanism to tap into more distant underground water resources by creating new hydraulic heads, and those would have little to no surface expression. Earthquakes from volcanic and tectonic activity may have contributed to this process at local and regional scales.

1.2.4 Discussion Both precipitation and aquifer discharge find support in the data but neither provides an unequivocal or unique explanation for the structure of the Martian hydrographic and limnologic systems and their environment. Analogy to Earth shows that both processes may, in fact, have substantially contributed. The geochemistry and miner­ alogy of lake deposits provides an approach to the question of their relative contribu­ tion (Fassett and Head, 2008) but carries uncertainties. Long groundwater flow paths carry more solutes than water supplied by short-residence aquifers or by surface runoff from precipitation, but there are exceptions relevant to the Martian environment (Hancock et al., 2005). On Mars, if groundwater was dominant, mineralogical depos­ its in basins should have been, in principle, more influenced by the mineralogical properties of the crust compared to those supplied by runoff from precipitation (Fassett and Head, 2008). But residence time is only one among other factors affecting water properties. Amount of direct precipitation and runoff, evaporation, and infiltra­ tion rates are other critical variables, and those of early Mars are unknown. Solutes in a water column are regulated by the erosive power of the water, and by meteoric and groundwater chemistry. Volcanic activity and sulfate deposits are evidence of a widespread acidic environ­ ment toward the end of the Noachian lasting into the Hesperian (Bibring et al., 2006). If drainage systems relied on precipitation at that time, then surface runoff would likely have been sustained by acid rain and/or acid snow. Everything being equal, runoff had enhanced erosive power producing streams and basins rapidly overloaded with abnormally high amounts of solutes compared to those expected in more neutral conditions. Using orbital geochemistry and mineralogy data to determine whether water originated from precipitation or groundwater is, therefore, a difficult proposition. It is further complicated as watersheds respond rapidly and effectively to off-normal local and regional conditions. Discharge rate controls the concentration of solutes in streams. In this case, their amounts vary inversely with discharge (Johnson and East, 1982) and are reflected in the inter-annual lake sediment deposition. Even when discharge is the dominant control on concentration, the sources of ions can be from atmospheric loading and water–rock interaction. The contribution of groundwater to total discharge is also a major factor controlling the chemical balance of streams and lakes, thus their geochemical and mineralogical signatures. The presence of sulfate deposits in basins and outcrops shows that environmental acidic conditions were transferred to the chemistry of the basins. Clay minerals have been interpreted to be the result of long-term weathering of primary minerals by liquid water at neutral to alkaline pH (Chevrier et al., 2007). Their local association with carbonate is proposed as evidence of different

12

Lakes on Mars

environmental conditions compared to those responsible for the formation of wide­ spread sulfate deposits (Bibring et al., 2006). Neutral to alkaline conditions can be achieved in terminal basins in an acidic environment. On Earth, carbonates play a major role in the buffering capacity of pollution- or volcanically-induced acid rains (Hagar et al., 2000). In such watersheds, alkaline minerals dissolve to naturally impart alkalinity to surface water and neutralize acid components. Furthermore, the sulfate of the sulfuric acid component of acid rain can be retained in the watershed by reaction with iron and aluminum oxide residues left behind by dissolution of silicate minerals (Willander et al., 2007). These processes effectively minimize the impact of acid rain on the acidity of runoff (Krug and Frink, 1983). Kilham (1982) showed the role of acidic precipitation in the alkalinization of a basin primarily through increased rate of carbonate weathering over a 20-year period. In these conditions, greater water acidity and chemical erosive power characterize streams located in the upper sub-watersheds compared to downstream, which could provide an explanation for the relatively enlarged low-order tributaries observed on Mars. Over a period of time dependent on the duration of the system, the volume of available buffering material of the watershed area, and the thickness of the eroded layer from it, alkaline to neutral conditions can be achieved in the lake.

1.3 The lake record of Mars In addition to mineralogy and morphology, another clue about the conditions at the time of lake formation can be revealed by numbers. At Viking resolution, 179 putative impact crater lakes were identified by Cabrol and Grin (1999), and 210 open lake systems were cataloged by Fassett and Head (2008) from a global survey using the most recent datasets. By comparison, the Earth has an estimated 304 million active lakes and ponds covering a surface area of 4.6 � 106 km2 of the land area (Downing et al., 2006). Most are small, but 182,300 are 1 km2 or larger, enough for the basins to be detected by instruments such as THEMIS (Christensen et al., 2004) and HRSC (Neukum et al., 2004). For the smaller lakes, fine scale features such as shorelines would require the resolution of CTX or HiRISE to be detected. The lack of global image coverage for these instruments may explain in part the reduced number of Martian lakes compared to Earth. Yet, the apparent Martian record remains low and may reflect an originally low production or significant erasure due to impact garden­ ing and erosion. The development of terrestrial global lake databases and modeling provides insights into that question (Alsdorf et al., 2003; Hamilton et al., 1992; Herdendorf, 1984; Kalf, 2001; Lehner and Döll, 2004; Meybeck, 1995).

1.3.1 Quantitative approach to the question A recent GIS-based analysis at global scale by Downing et al. (2006) shows that regardless of their supply system, terrestrial lake-size distribution has an excellent fit to a size-frequency function of the form N(A) = αAβ, where N(A) is the number of lakes

Searching for lakes on Mars: four decades of exploration

13

Number of lakes of greater area, N(A)  A

whose area is equal or greater than A; α and β are fitted parameters, where α is the total number of lakes of one unit area in size and β is the logarithmic rate of decline of lakes with lake area (Downing et al., 2006; Lehner and Döll, 2004). Further, the equation fits regardless of climate and, by anchoring the canonical frequency distribution to the sizes of the world’s largest lakes, the number of lakes in the world can be approxi­ mated over any size range (Downing et al., 2006). This relationship relates to the fractal dimension of the landscape and the geomorphological pattern of depressions and overall reliefs (Hamilton et al., 1992) and is consistent with the stochastic model of Mandelbrot (1982). If Martian lakes were formed under circumstances comparable to those of Earth, then such law may be applicable and could allow the prediction of Martian lake-size distribution by a power function (Figure 1.6). Resolution is now available to identify lakes 1 km2 or larger on Mars. A global survey of their distribution was performed by Fassett and Head (2008) who cataloged all open lake basins to infer hydrology and climate in the Noachian. Although their study excluded closed basins because of the difficulty in ascertaining their origin, it is currently the most complete and represen­ tative dataset. The terrestrial inventory by Downing et al. (2006) includes all active terrestrial lakes and ponds over a period of 30 million years (the age of lake Baïkal the oldest on Earth). This timeframe also superposes the record of climate cycles, con­ tinental movement, major orogenesis, and erosion. Climate changes and erosion are common to Earth and Mars but their respective impact may have varied significantly in nature and magnitude due to distinct planetary environments, such as high obliquity on Mars (Laskar et al., 2004). The main process of

100,000

Earth (β = –1.06) Mars (β = –0.40)

10,000 1000 100 10 1 1

10

100 1000 10,000 100,000 Lake area (A, km2)

Figure 1.6 Comparison between the slope of the terrestrial lake-size frequency (–1.06, Downing et al., 2006) with that of the Martian paleolake distribution (–0.40) including the small classes of paleolakes mostly affected by erosion. To assess an original slope for Mars, this truncated part of the sample is removed. Because of differential erosion, the largest lakes may have preserved a slope closer to the original one: for A � 500 km2, β = –0.635.

14

Lakes on Mars

rock recycling on Earth is plate tectonics. On Mars, impact cratering dominated rock processing, and lake formation was contemporary with major fluctuations in crater production rates from the LHB in the Noachian up into the Hesperian (Hartmann and Neukum, 2001). Lake-size frequency on each planet thus may be affected by differences in their respective planetary environments, geological scales, and geological processes involved. Further, the Martian record is fossil and the product of both an unknown production rate, and alteration by 3.5 billion years of surface evolution. Although the history and magnitude of both records may be different, the existence of a natural law governing the number and size of lakes offers a tantalizing possibility to perform forensic limnology on Mars. Trying to reconstruct the total number of lakes ever produced by Mars is meaningless because erosion and resurfacing have partially obscured it. But, if Martian and terrestrial lake formation followed similar processes, the residual record may still show clues on how many might have been formed from this record. If a distribution law appears to hold for Martian lakes, this might imply that lakes were being quasi-continuously produced over a sustained period of time (or possibly a series of discontinuous, but sustained, period during favorable high-obli­ quity cycles) rather than triggered by a few isolated water releases.

Lake abundance and distribution Most of the variables to solve N(A) = αAβ on Mars are either uncertain or unknown but a number of reasonable hypotheses can be proposed. For instance, it can be postulated that small lakes were produced in greater number than large ones. For instance, on Earth, the terrestrial ratio of lakes of A (1 km2) to A (1000 km2) is 1/1863 (Downing et al., 2006). Moreover, as conditions supporting the formation of lakes stopped on Mars, small lakes were also more likely to be removed from the geological record than larger basins, altering the original ratio between small and large populations and resulting in a truncated record. This alteration is actually observed in the distribution of paleolakes by size in the most recent inventory by Fassett and Head (2008). Paleolakes smaller than 10 km2 are mostly absent between 30 N and 70 S and very few paleolakes smaller than 100 km2 remain. Truncation affects the accuracy and predictive nature of Pareto distributions (Hamil­ ton et al., 1992). Therefore, the deficit in small lakes leads to an underestimation of the total population. When applied to the complete distribution of Martian lakes as observed today, a power function fit of the relationship between lake number and size results in a rate of decline of –0.40 (Figure 1.6), which carries a low confidence interval of 77.58%. A high value can be obtained by arbitrarily fitting the Martian distribution with the terrestrial slope of –1.06 (Downing et al., 2006). Uncertainties over surface runoff duration and climate instability make this an improbable model for Mars but it helps confining boundaries for the unknown original distribution. Both slopes of power law are unsatisfactory, the first one because it includes truncated populations and the second because it arbitrarily fits an index that is likely not applicable to Mars. A more realistic approach is to propose that an original production rate, if any, should fall somewhere between those two values. Because large lakes are more resistive to erosion over time, their number may give a more valid

Searching for lakes on Mars: four decades of exploration

15

100

10

1 1

Number of lakes of greater area N(A) ≥ A

f(10S) = 620 A–0.635 f(20S) = 440 A–0.635 f(30S) = 755 A–0.635

10

100 1000 10,000 100,000 Lake area (A, km2)

(c) 1000 P1

P1: f(40S) = 4150 A–0.635 –0.635 P1: f(50S) = 2500 A –0.635 P2: f(40S) = 200 A P2: f(50S) = 60 A–0.635

100 P2(?) 10

1 1

10

100

1000 10,000 100,000

Lake area (A, km2)

Number of lakes of greater area N(A) ≥ A

(a) 1000

Number of lakes of greater area N(A) ≥ A

Number of lakes of greater area N(A) ≥ A

estimate of an original slope of power law, thus are the ones used for this estimation. Further, terrestrial lake sizes vary geographically due to climate factors, but they all have strong interregional similarities in slopes (Downing et al., 2006). By analogy, similar slope patterns at different latitudes on Mars could provide supportive evidence of an original slope that is representative at planet-scale. Over 59% of the population of the Martian open lakes are concentrated between 10°S and 30°S where the best distributions are preserved. The analysis of sizefrequency at these latitudes for repetitive linearity suggest statistically self-similar distributions on the segment of large lake populations least affected by truncation (A � 500 km2, n = 126/210). For these lakes, the number versus size distribution shows a good fit to an average slope of –0.635, and a confidence interval of 99.8% (Figure 1.7a). Inter-latitudinal similarities reinforce the possibility that this slope may

f(30N) = 2900 A–0.635 f(20N) = 1460 A–0.635 f(10N) = 1110 A–0.635 f(Eq.) = 198 A–0.635

(b) 1000

100

10

1 1

10 (d)

100 1000 10,000100,000 Lake area (A, km2) f(60S) = 287 A–0.635 f(70S) = 97 A–0.635

1000

100

10

1 1

10

100 1000 10,000 100,000 Lake area (A, km2)

Figure 1.7 (a-d) Paleolake size frequency on Mars by latitude with power law functions for A � 1 km2. The best-preserved distributions exhibit consistent slopes (β) between –0.62 and – 0.64 with an average value of –0.635. For 40S and 50S (d), P1 populations are those anchored on the largest lakes for each latitude. The P1 population for 40S is 4150 lakes and 2500 lakes for 50S. Negative deviations from fit are interpreted as removal from the geological record, which for both populations reached over 99%. P2 could represent a second production of smaller lakes but the statistical representativity of the sample is low. Similarities in the 40S and 50S distributions and deviations from fit strongly suggest that a similar event operated at both latitudes. This might be best explained by a climatic cause.

16

Lakes on Mars

be representative at planet scale and can be used to infer the lake population originally produced from the residual record. When a –0.635 slope is applied to the 30N-70S distributions, we find that nearly 12,000 lakes could have been produced from the residual record only (Figure 1.7a–1.7d). The robustness of this evaluation resides first in that distributions are anchored on the largest lakes less likely to be affected by erosion over time than smaller ones. Second, terrestrial lakes follow a Pareto distribution regardless of water supply, lake type, or size (Downing et al., 2006). Their total number is only affected by the rate of decline and the size of the largest lakes. While our dataset relies exclusively on open lake systems, none of the closed basins inventoried at global scale as of today (Cabrol and Grin, 1999) are larger than any open lakes, thus they should not significantly affect the result. There is, however, one exception. If the 720,000 km2 Elysium basin contained a lake rather than lava, then the relationship between size and frequency shows that the region between the equator and 10°N had the potential to produce four times more lakes than shown here.

Windows into past climate changes and resurfacing The fact that the size-frequency distribution of large lakes is consistent with a power law function is highly significant for the environment at the time of lake formation and supports the idea that conditions were sustained long enough to statistically establish a production rate. The distribution of lakes below 500 km2 can be interpreted in different ways. The first one is that small lakes followed the same power law as large lakes but experienced differential erosion due to their smaller size. This inter­ pretation fits the terrestrial model and is also consistent with a planet geologically active. It implies that the majority of small lakes have disappeared from the geological record. The second interpretation is that small lakes followed a different power law altogether, which would indicate changed conditions for lake formation. However, consistency in terrestrial slopes regardless of climate conditions and sizes suggests that the slope should have remained similar on Mars over time and makes this interpretation less likely. Here, more data, such as global coverage at CTX or HiRISE resolution would be needed to assess with greater accuracy the number of small lakes actually remaining. Finally, it is possible that small lakes did not follow any power law at all, which would suggest that they were associated with the random (accidental) filling of depressions. This interpretation might not be exclusive of the two previous and is likely to explain at least some of the lakes that were formed during local, catastrophic water release. The reason for the Martian lake production rate to be 60% that of Earth is unclear and may relate to planetary factors, climate stability, water availability and duration of surface runoff. Assuming that the Martian production rate was consistent across all lake sizes and over time (terrestrial model), we make hypothetical reconstructions of the original latitudinal populations by fitting the number of lakes to a size-frequency distribution of slope –0.635. Each distribution is anchored to the largest observed lake in each group (Figure 1.7). Figure 1.8a shows the reconstructed number of lakes per 10 degree latitudes and Figure 1.8b the deficit of lakes £ 100 km2. The deficit (expressed in

Searching for lakes on Mars: four decades of exploration

Total number of lakes

4500 3500

(a)

3000 2500 2000 1500 1000 500 0

Percent loss per class of size

17

100 90 80 70 60 50 40 30 20 10 0

30 20 10

0 –10 –20 –30 –40 –50 –60 –70

(b)

1 km2 10 km2 100 km2 30 20 10

0 –10 –20 –30 –40 –50 –60 –70 Latitude (degree)

Figure 1.8 (a) Fitted frequency per 10 degree latitude with a global distribution reaching nearly 12,000 paleolakes. This distribution may reflect several episodes of lake formation separated by time and climate conditions, thus it may not be interpreted directly. If paleolakes were only formed during the same type of obliquity cycles, then this distribution could reflect the climate influence in their relative spatial distribution. Because of the relationship between size and total number in the power law function, the concentration at 40–50°S appears as a solid result since the two largest paleolakes are observed at these latitudes; (b) Assuming similar slope for all latitudes, the loss of paleolakes as a function of size and latitude can be inferred as the ratio of paleolakes actually observed over those predicted by the fitted distributions.

percent) is the ratio between the lakes actually observed and the number predicted by the power law function. If small lakes followed a similar power law distribution, then over 90% of them have been erased from the geological record. In the original record, the Martian lake distribution is centered at 10°S (Fassett and Head, 2008). Our fitted distribution based on size and number rather than number only shows now a bimodal, hemispheric distribution with maximum concentration at 40°S and over 60% of the population located 30°S and higher (Figure 1.8a). Unknowns about the possible number of episodes of formation accumulated and time spans between those episodes do not allow a direct interpretation of this distribution, but it may still contain significant clues. The first one is that Martian lakes might not have been concentrated in the southern tropical regions as seen in the residual record but in the mid-latitudes in that hemisphere. This result is consistent with an abundance of morphologic evidence of mid-latitude glaciation and periglacial processes (e.g., Head et al., 2005; Holt et al., 2008). This may also be an indication of when obliquity favored their formation (Forget et al., 2006).

18

Lakes on Mars

Interpretation for the northern hemisphere is constrained by the erosion of the highlands and major resurfacing from Tharsis and Elysium. Concentration in this hemisphere is between 10°S and 20°S (Figure 1.8a). If interpreted directly, this could suggest that the northern hemisphere produced substantially fewer lakes compared to the south. However, the planetary dichotomy and resurfacing may have contributed to bias the record in the northern hemisphere. For all latitudes, the potential loss of a particular range of lake size is given by negative deviations to the fit. Most 1–10 km2 lakes may have been removed at all latitudes (Figures 1.7 and 1.8). In our reconstructed distributions, greatest deficit for larger lakes is between 40–50°S and 10–20°N with the most remarkable examples being the 40°S–50°S populations where both distributions have been reset (Figure 1.7d). The first production is associated with their largest lakes. Over 99% of this population is now missing, consistent with the geological activity at those latitudes where intense resurfacing is associated with subsurface ice and active layers. A second population is made of smaller size lakes. Their better preservation compared to larger lakes strongly suggests that we are observing at least two separate episodes of formation, with the smallest lakes being associated with the most recent event. Other important deviations from fit are observed in the intertropical belt (Figure 1.7a and 1.7b). They are best explained by erosion and resurfacing. Like in the southern hemisphere, positive deviations from fit may show several episodes of formation, in which case, the production of smaller lakes may be an indication that overall water availability had decreased, either through global changes at planetary scale, and/or that those lakes were formed during distinct climate configurations.

1.4 Time machines Combined, the results from morphology, mineralogy, and estimation of lake produc­ tion support the idea that favorable conditions did exist on Mars to produce a few large bodies of water, and many more, smaller lakes that formed by processes analogous to those occurring on Earth. The apparent low production rate for these lakes may indicate that these earth-like favorable conditions did not last. Over long geological time scales, lakes are presumably transient, fragile entities. It is, therefore, intuitive to think that Mars formed more lakes than still can be observed. Ultimately, knowing exactly how many is not as important as understanding the abundance underscored by uncovering a production rate in the residual record, and its significance for early Mars habitability. The exploration of lakes allows us to enter a time machine where the past can be deciphered. From the regular pulses of past cycles and their interruption by cata­ strophic events, some measure of the future can also be predicted. It is thus not surprising that ancient lakes have become primary targets for the surface exploration of Mars. With the MER mission, Opportunity was sent to Meridiani to explore hematite deposits. In the process, she was the first of the two rovers to encounter evidence of groundwater activity and surface ponding (Squyres et al., 2006).

Searching for lakes on Mars: four decades of exploration

19

Her extended traverse across Meridiani Planum and her 181-day incursion into Endurance crater unraveled an ancient environment of interdune playa in an already arid, acidic, and oxidizing environment (Grotzinger et al., 2006; Knoll et al., 2005). Spirit set off to explore Gusev crater with the hope of finding traces of an ancient lake fed by Ma’adim Vallis (Cabrol et al., 1996, 1998), itself carved from the release of an ancient lake (Irwin et al., 2002). With the data collected by MGS and ODY, it became obvious before Spirit departed Earth that an ancient lake would be now buried under a layer of lava of undetermined thickness (Cabrol et al., 2003) but the age of the crater basin could have made it possible for large impacts to have excavated some of the lake deposits. Thick lava flows have kept ancient lake deposits locked away from the rover, but on her journey, Spirit showed once again that exploration takes us always beyond the horizon and beyond our wildest expectations. On Sol 156, with her front wheels on the Columbia Hills and her rear wheels on the lava plains, she bridged two worlds: before and after water on Mars. It was a giant step of hundreds of millions of years under six small wheels (Figure 1.9). In the Columbia Hills, she entered a world closer to the time when lakes were still forming on Mars, but not close enough, though. However, on the other side of the summit, what Spirit discovered at Home Plate is a legacy of evidence that Mars was a habitable planet, a dynamic and violent world very much like early Earth, where encounters between magma and groundwater resulted in explosions that littered the basin with volcanic rocks and ash. They also produced hydrothermal springs (Morris et al., 2008; Rice et al., 2010; Ruff et al., 2008), possible abodes of life that may be revisited one day by rovers and/or humans with instruments adapted to their analysis. To last, these springs needed to be sustained by aquifers charged prior to that epoch (Grin and Cabrol, 1997). They were among the last possible surface oases before Mars became the world we know today. After six years of exploration, Spirit is now hibernating at Home Plate. A sunset on Gusev (Figure 1.10) is then a fitting image to close a chapter so that new ones can be opened.

Figure 1.9 Navcam panorama on June 11, 2004, Sol 156 at Gusev crater. After completing 3.4 km, Spirit arrived at the base of the Columbia Hills. The Plains Unit is to the east (right) where lava alteration was dominated by little water in a cold climate. To the west (left), are rocks that formed, or were altered, during an older epoch that still postdates the time when lakes in Gusev were formed. On the other side of the summit, Home Plate revealed evidence of volcanic/groundwater interaction and past habitability. MER, NASA/JPL-Caltech/Cornell.

20

Lakes on Mars

Figure 1.10 Sun setting on the rim of Gusev crater, May 19, 2005. Spirit was added to the panorama for artistic rendering. The panorama was acquired on Sol 489th, when the rover was well on her way to successfully completing the first hill ascent on another planet. MER, NASA/ JPL-Caltech/Cornell.

Acknowledgment The authors wish to thank Victor Baker and Keith Harrison for their in-depth reviews that helped improve the manuscript. The analysis comparing the terrestrial and Martian lake distributions used the data published in Table 1 by Fassett and Head (2008) and Table 2 by Downing et al. (2006), respectively with permission of the Icarus and Limnology and Oceanography journals.

References Aharonson, O., Zuber, M.T., Rothman, D.H., Schorghofer, N., Whipple, K.X., 2002. Drainage basins and channel incision on Mars. Proc. Natl. Acad. Sci. USA 99, 1780–1783. Alsdorf, D., Lettenmaier, D., Vörösmarty, C., 2003. The need for global, satellite-based observations of terrestrial surface waters. EOS 84 (269), 275–276. Aravena, R., 1995. Isotope hydrology and geochemistry of northern Chile groundwaters. Bull. Inst. Fr. Et. And. 24 (3), 495–503. Arvidson, R.E., Poulet, F., Morris, R.V., Bibring, J.-P., Bell, J.F., Squyres, S.W., et al., 2006. Nature and origin of the hematite-bearing plains of Terra Meridiani based on analyses of orbital and Mars Exploration rover data sets. J. Geophys. Res. 111 (E12), E12S08. Baker, V.R., 2001. Water and the Martian landscape. Nature 412, 228–236.

Searching for lakes on Mars: four decades of exploration

21

Baker, V.R., Partridge, J.B., 1986. Small Martian valleys: Pristine and degraded morphology. J. Geophys. Res. 91, 3561–3572. Baker, V.R., Strom, R.G., Gulick, V.C., Kargel, J.C., Komatsu, G., Kale, V.S., 1991. Ancient oceans, ice sheets and the hydrological cycle on Mars. Nature 352, 89–594. Barlow, N.G., Boyce, J.M., Costard, F.M., Craddock, R.A., Garvin, J.B., Sakimoto, S.E.H., et al., 2000. Standardizing the nomenclature of Martian impact crater ejecta morphologies. J. Geophys. Res. 105, 26733–26738. Basilevskaya, E.A., Neukum, G., the HRSC Co-Investigator team, 2009. The Olympus volcano on Mars: Geometry and characteristics of lava flows. Sol. Syst. Res. 40 (5), 375–383. Beaty, D.W., Clifford, S.M., Borg, L.E., Catling, D., Craddock, R.A., et al. 2005. Key science questions from the second conference on early Mars: Geologic, hydrologic, and climatic evolution and the implications for life. Astrobiology 5 (6), 663–689. Berezovskaya, S., Yang, D., Hinzman, L., 2005. Long-term annual water balance analysis of the Lena river. Global Plan Change 48, 84–95. Bibring, J.-P., Langevin, Y., Gendrin, A., Condet, B., Poulet, F., et al., 2005. Mars surface diversity as revealed by OMEGA/Mars Express observations. Science 307 (5715), 1576–1581. Bibring, J.-P., Langevin, Y., Mustard, J.F., Poulet, F., Arvidson, R., et al., 2006. Global mineralogical and aqueous Mars history derived from OMEGA/Mars Express data. Science 312 (5772), 400–404. Bishop, J.L., Noe Dobrae, E.Z., McKeown, N.K., Parente, M., Ehlmann, B.L., et al., 2008. Phyllosilicate diversity and past aqueous activity revealed at Mawrth Vallis, Mars. Science 321 (5890), 830–833. Bleacher, J.E., de Wet, A.P., Garry, W.B., Zimbelman, J.R., Trumble, M.E., 2010. Volcanic or fluvial: Comparison of an Ascraeus Mons, Mars, braided and sinuous channel with features of the 1859 Mauna Loa flow and Mare Imbrium flow, 41st Lun. Plan. Sci. Conf., Abstract No. 1612. Bradley, R.S., Vuille, M., Diaz, H.F., Vergara, W., 2006. Threats to water supplies in the tropical Andes. Science 312, 1755–1756. Bronmark, C., Hansson, L.-A., 1998. The Biology of Lakes and Ponds. Biology of Habitats. Oxford University Press, New York, 216 p. Cabrol, N.A., 1991. Etude des paleo-chenaux Martiens: caractéristiques morphologiques, distribution spatio-temporelle, mécanismes de genèse (PhD thesis), Paris I, PanthéonSorbonne University, France, 413 p. Cabrol, N.A., 1993. Martian channels: Statistics of spatial distribution, relative datation, and implications. Zeitschrift für Geom. 37.1, 57–76. Cabrol, N.A., E.A., Grin., 2005. Ancient and recent lakes on mars. In: Tokano, T. (Ed.), Water on Mars and Life. Springer, Berlin Heidelberg, pp. 235–259. Cabrol, N.A., Grin, E.A., 1995. A morphological view on potential niches for exobiology on Mars. Planet. Space Sci. 43 (1/2), 179–188. Cabrol, N.A., Grin, E.A., 1999. Distribution, classification, and ages of Martian impact crater lakes. Icarus 142, 160–172. Cabrol, N.A., Grin, E.A., 2001. The evolution of lacustrine environments on Mars: Is Mars only hydrologically dormant? Icarus 149, pp. 291–328. Cabrol, N.A., Grin, E.A., 2002. Overview on the formation of Paleolakes and Ponds in impact craters on Mars. Global Plan Change 35, 199–219. Cabrol, N.A., Grin, E.A., Carr, M.H., Sutter, B., Moore, J.M., et al. 2003. Exploring Gusev crater with MER A: Review of science objectives and testable hypotheses. JGR-Planets, Special Mars Exploration Rover (MER) Mission Issue, 8076, doi: 10.1029/2002JE002026. Cabrol, N.A., Grin, E.A., Dawidowicz, G., 1996. Ma’adim Vallis revisited through new topographic data. Icarus 123, 269–283.

22

Lakes on Mars

Cabrol, N.A., Grin, E.A., Landheim, A.R., 1998. Ma’adim Vallis Evolution: Geometry and models of Discharge Rate. Icarus 132, 362–377. Cabrol, N.A., Wynn-Williams, D.D., Crawford, D.A., Grin, E.A., 2001. Recent Aqueous environments in impact crater lakes on Mars: an Astrobiological perspective. 2nd Mars Polar Conference Special Issue. Icarus 154, 98–112. Cabrol, N.A., the HLP team, 2009 The high lakes project. J. Geophys. Res. Biogeosci., High Lakes Project, special issue, doi:10.1029/2008JG000818. Carr, M.H., 1974. The role of lava erosion in the formation of lunar rilles and Martian channels. Icarus 22, 1–23. Carr, M.H., 1996. Channels and valleys on Mars: Cold climate features formed as a result of a thickening cryosphere. Planet. Space Sci. 44, 1411–1423. Carr, M.H., Head, J.W., III, 2003. Basal melting of snow on early Mars: A possible origin for some valley networks. GRL 30 (24), 2245. Carr, M.H., 2006. The Surface of Mars. Cambridge University Press, Cambridge, 229 pp. Carr, M.H., 2008. The role of lava erosion in the formation of lunar rilles and Martian channels. Icarus 22, 1–23. Catling, D.C., 2007. Mars: Ancient fingerprints in the clay. Nature 448, 31–32. Chapman, M.G., 1999. Elysium basin lava flows: New interpretations based on MOC data. 30th Lun. Plan. Sci., Abstract 1279. Chevrier, V., Poulet, F., J.-P., Bibring, 2007. Early geochemical environment of Mars as determined from thermodynamics of phyllosilicates. Nature 448, 60–63. Christensen, P.R., et al., 2004. The Thermal Emission Imaging System (THEMIS) for the Mars 2001 Odyssey mission. Space Sci. Rev. 110, 85–130. Christensen, P.R., Bandfield, J.L., Clark, R.N., Edgett, K.S., Hamilton, V.E., et al. 2000. Detection of crystalline hematite mineralization on Mars by the Thermal Emission Spec­ trometer: Evidence for near-surface water. J. Geophys. Res. 105, 9623–9642. Christensen, P.R., Morris, R.V., Lane, M.D., Bandfield, J.L., Malin, M.C., 2001. Global mapping of Martian hematite mineral deposits: Remnants of water-driven processes on early Mars. J. Geophys. Res. 106, 23873–23885. Catling, D.C., Morris, R.V., McLennan, S.M., Gellert, R., Jolliff, B., 2005. Chemistry and mineralogy of outcrops at Meridiani Planum. Earth Plan. Sci. Lett. 240, 73–94. Clifford, S.M., 1993. A model for the hydrologic and climatic behavior of water on Mars. J. Geophys. Res. 98, 10937–11016. Clifford, S.M., Parker, T.J., 2001. The evolution of the Martian hydrosphere: Implications for the fate of a primordial ocean and the current state of the northern plains. Icarus 154, 40–79. Craddock, R.A., Howard, A.D., 2002. The case for rainfall on a warm, wet early Mars. J. Geophys. Res. 107 (E11), 5111. De Hon, R., 1992. Martian lake basins and lacustrine plains. Earth, Moon Planets 56, 95–122. Des Marais, D.J., et al., 2008. Focus Paper: The NASA Astrobiology Roadmap. Astrobiology 8 (4), 715–730, doi:10.1089/ast.2008.0819. Des Marais, D.J., Allamandola, L.J., Benner, S.A., Boss, A.P., Deamer, D., et al., 2003. The NASA Astrobiology Roadmap. Astrobiology 3, 219–236. Des Marais, D.J., Farmer, J.D., 1995. The search for extinct life: An exobiological strategy for Mars exploration. NASA Spec. Publ. 530 (56), 21–26. Di Achille, G., Ori, G., Reiss, D., Hauber, E., Gwinner, K., Michael, G., et al., 2006. A steep fan at Coprates Catena, Valles Marineris, Mars, as seen by HRSC data. Geophys. Res. Lett. 33, L07204. Dohm, J.M., Baker, V.R., Maruyama, S., Anderson, R.C., 2007. Traits and evolution of the tharsis superplume. In: Yuen, D.A., et al., (Eds.) Superplumes: Beyond Plate Tectonic. Springer Netherlands, Dordrecht, Part VI, pp. 532–536.

Searching for lakes on Mars: four decades of exploration

23

Doran, P.T., Lyons, W.B., McKnight, D.M., 2010. Life in Antarctic Deserts and other Cold Dry Environments. Astrobiological Analogues, Series. Cambridge University Press, Cambridge, UK, 306 p. Downing, J.A., Prairie, Y.T., Cole, J.J., Duarte, C.M., Tranvik, L.J., Striegl, R.G., et al., 2006. The global abundance and size distribution of lakes, ponds, and impoundments. Limn. Oceanogr. 51 (5), 2388–2397. Elhmann, B.L., Mustard, J.F., Murchie, S.L., Poulet, F., Bishop, J.L., et al., 2008. Orbital identification of carbonate-bearing rocks on Mars. Science 322 (5909), 1828–1832. Fairén, A.G., Davila, A.F., Gago-Duport, L., Amils, R., McKay, C.P., 2009. Stability against freezing of aqueous solutions on early Mars. Nature 459, 401–404. Farmer, J., 1999. Implementing a strategy to explore for martian life. In: Hiscox J.A. (Ed.), The Search for Life on Mars. British Interplanetary Society, London, England, pp. 58–65. Farmer, J.D., 2003. Exploring for a fossil record of extraterrestrial life, In: Briggs, D.E.G., Crowther, P.R. (Eds.) Paleobiology II. Blackwell Science, Oxford, U.K., pp. 8–13. Fasset, C., Head, J., 2005. Fluvial sedimentary deposits on Mars: ancient deltas in a crater lake in the Nili Fossae region. Geophys. Res. Lett. 32, L14201. Fassett, C.I., Head, J.W., 2008. Valley network-fed, open basin lakes on Mars: Distribution and implications for Noachian surface and subsurface hydrology. Icarus 198 (1), 37–56. Forget, F., Haberle, R.M., Montmessin, F., Levrard, B., Head, J.W., 2006. Formation of glaciers on Mars by atmospheric precipitation at high obliquity. Science 311 (5759), 368–371. Forsythe, R.D., 1990. A case for Martian salars and saline lakes during the Noachian, 21st Lun. Planet. Sci. Conf., pp. 379–380 (Abstract). Forsythe, R.D., Blackwelder, C.R., 1998. Highland crater basins as evaporative/sublimative discharge pumps for driving early Martian groundwater to atmosphere water transfers. In: Clifford, S.M., et al. (Eds.), Conference on Early Mars: Geologic and Hydrologic Evolu­ tion, Physical and Chemical Environments, and the Implications for Life. 34–35, LPI Contr. No. 916 (Abstract). Forsythe, R.D., Zimbelman, J.R., 1995. A case for ancient evaporite basins on Mars. J. Geophys. Res. 100, 5553–5563. Fujita, S.-I., Tatsuya, S., Kazuhide, M., 2003. Wet and dry deposition of sulfur associated with the eruption of Miyakejima volcano, Japan. J. Geophys. Res. 108 (D15), ACH 3-1 Gendrin, A., Mangold, N., Bibring, J.-P., Langevin, Y., Gondet, B., et al., 2005. Sulfates in Martian layered terrains: The OMEGA/Mars Express view. Science 307 (5715), 1587–1591. Giese, B., Urizar, C., Fuckar, S., 2002. Southern hemisphere origins of the 1976 climatic shift. Geophys. Res. Lett. 29 (2), 1014–1017. Glotch, T.D., Rogers, A.D., 2007. Evidence for aqueous deposition of hematite and sulfate-rich light-toned layered deposits in Aureum and Iani Chaos, Mars. J. Geophys. Res. 112, E06001. Goldspiel, J., Squyres, S., 1991. Ancient aqueous sedimentation on Mars. Icarus 89, 392–410. Grant, J.A., 2000. Valley formation in Margaritifer Sinus, Mars, by precipitation-recharged ground-water sapping. Geology 28, 223–226. Greeley, R., 1973. Mariner 9 photographs of small volcanic structures on Mars. Geology 1, 175–180. Gregg, K.P., Lopes, R.M., 2007. Volcanic depressions and lava lakes on Earth, Mars, Venus, Io, and Titan. Ice, Oceans, and Fire: Satellites of the Outer Solar System, Abstract No. 6054. Grimm, R.E., Stillman, D.E., 2008. Electrical properties of saline ices, salt hydrates, and icesilicate mixtures: Applications to solar system exploration, Science of Solar System Ices, Abstract No. 9062.

24

Lakes on Mars

Grin, E., Cabrol, N., 1997. Limnological analysis of Gusev Crater paleolake, Mars. Icarus 130, 461–474. Grosjean, M., Ammann, B., Camacho, M., Geyh, M.A., Kern, R., et al., 2001. Mid-Holocene (ca. 6000–3500 14C yr B.P.) climates in the Atacama Altiplano of the south Central Andes: humid or dry? Limnogeological, geomorphological, hydrological and vegetational evi­ dence for widespread mid-Holocene aridity. Workshop on Central Andes Paleoclimatol­ ogy, NSF, January 11–16, 2001, Tucson, AZ. Grotzinger, J.P., Arvidson, R.E., Bell, J.F. III, Calvin, W., Clark, B.C., et al., 2005. Stratigraphy and sedimentology of a dry to wet eolian depositional system, Burns Formation, Meridiani Planum, Mars. Earth Planet. Sci. Lett. 240, 11–72. Grotzinger, J., Bell, J. III, Herkenhoff, K., Johnson, J., Knoll, A., et al., 2006. Sedimentary textures formed by aqueous processes, Erebus crater, Meridiani Planum, Mars. Geology 34 (12), 1085–1088. Gulick, V.C., 1998. Magmatic intrusions and a hydrothermal origin for fluvial valleys on Mars. J. Geophys. Res. 103 (E8), 19365–19387. Gulick, V.C., 2001. Origin of the valley network on Mars: a hydrological perspective. Geo­ morphology 37 (3/4), 241–268. Gulick, V.C., Tyler, D., McKay, C.P., Haberle, R.M., 1997. Effects and lifetimes of oceaninduced CO2 pulses on Mars: implications for fluvial valley formation. Icarus 130, 68–86. Haberle, R.M., Jakosky, B.M., 1990. Sublimation and transport of water from the north residual polar cap on Mars. J. Geophys. Res. 95, 1423–1437. Haberle, R.M., Pollack, J.B., Barnes, J.R., Zurek, R.W., Leovy, C.B., Murphy, J.R., Lee, H., and Schaeffer, J., 1993. Mars atmospheric dynamics as simulated by the NASA/Ames general circulation model I: The zonal mean circulation. J. Geophys. Res. 98, 3093– 3124. Haberle, R.M., McKay, C.P., Cabrol, N.A., Grin, E.A., Schaeffer, J., Zent, A.P., et al., 2001. On the possibility of liquid water on present day Mars. J. Geophys. Res. 106, 23317–23326. Hagar, W.G., Crosby, B.A., Stallsmith, B.W., 2000. Comparing and assessing acid-rain sensi­ tive ponds. J. Hazard. Mat. 74 (1/2), 125–131. Hamilton, S.K., Melack, J.M., Goodchild, M.F., Lewis, W.M., Jr., 1992. Estimation of the fractal dimension of terrain from lake size distributions. In: Petts, G.E., Carling, P.A. (Eds.), Lowland Floodplain Rivers: A Geomorphological Perspective. Wiley, Chichester, West Sussex, 145–164. Hancock, P.J.A., Boulton, J., Humphreys, W.F., 2005. Aquifers and hyporheic zones: Toward an ecological understanding of groundwater. Hydrogeol. J. 13, 98–111. Hanna, J.C., Phillips, R.J., 2005. Hydrologic modeling of the Martian crust with application to the pressurization of aquifers. J. Geophys. Res. 110, E01004. Harrison, K.P., Grimm, R.E., 2004. Tharsis recharge: A source of groundwater for Martian outflow channels. Geophys. Res. Lett. 31, L14703–L14706. Harrison, K.P., Grimm, R.E., 2008. Multiple flooding events in Martian outflow channels. J. Geophys. Res. 113, E02002–E02012. Harrison, K.P., Grimm, R.E., 2009. Regionally compartmented groundwater flow on Mars. J. Geophys. Res. 114, E4, doi: 10.1029/2008JE003300. Hartmann, W.K., Neukum, G., 2001. Cratering chronology and the evolution of Mars. Space Sci. Rev. 96, 165–194. Haveli, I., Zuber, M.T., Schrag, D.P., 2007. A sulfur dioxide climate feedback on early Mars, 38th Lun. Plan. Sci. Conf., Abstract No. 1173. Head, J.W., Neukum, G., Jaumann, R., Hiesinger, H., Hauber, E., Carr, M., et al., 2005. Tropical to mid-latitude snow and ice accumulation, flow and glaciation on Mars. Nature 434, 346–351.

Searching for lakes on Mars: four decades of exploration

25

Hecht, M.H., 2002. Metastability of liquid water on Mars. Icarus 156 (2), 373–386. Hecht, M.H., Kounaves, S.P., Quinn, R.C., West, S.J., Young, S.M.M., et al., 2008. Detection of perchlorate and the soluble chemistry of Martian soil at the Phoenix lander site. Science 325 (5936), 64–67. Herbst, D.B., 2001. Gradients of salinity stress, environment stability and water chemistry as a template for defining habitat types and physiological strategies in inland salt waters. In: Melack, J.M., Jellison, R., Herbst, D.B., (Eds.), Saline Lakes, Developments in Hydrology, Kluwer Academic Publishers, USA, pp. 209–219. Herdendorf, C.E., 1984. Inventory of the morphometric and limnologic characteristics of the large lakes of the world. Tech. Bull. The Ohio State University, Sea Grant Program. Hock, A.N., 2008. Licancabur Volcano and Life in the Atacama: Environmental physics and analogies to Mars. Ph. D. Thesis, University of California, Los Angeles. Hoffman, N., 2000. White Mars: A new model for Mars’ surface and atmosphere based on CO2. Icarus 146, 326–342. Hofmann, B., Farmer, J., Blankenberg, F., Fallick, A., 2008. Subsurface filamentous fabrics: An evaluation of origins based on morphological and geochemical criteria, with implications for Exopaleontology. Astrobiology 8 (1), 87–117. Hollibaugh, J.T., Wong, P.S., Bano, N., Pak, S.K., Prager, E.M., Orrego, C., 2001. Stratification of microbial assemblages in mono lake, california, and response to a mixing event. In: Melack, J.M., Jellison, R., Herbst, D.B. (Eds.), Saline Lakes, Developments in Hydrology, Kluwer Academic Publishers, USA, pp. 45–60. Holt, J.W., Safaeinili, A., Plaut, J.J., Head, J.W., Phillips, R., et al., 2008. Radar sounding evidence for buried glaciers in the southern mid-latitudes of Mars. Science 322 (5905), 1235–1238. Howard, A.D., 2007. Simulating the development of Martian highland landscapes through the interaction of impact cratering, fluvial erosion, and variable hydrologic forcing. Geomor­ phology 91, 332–363. Hynek, B.M., Phillips, R.J., 2003. New data reveal mature, integrated drainage system on Mars indicative of past precipitation. Geology 31 (9), 757–760. IPCC, Climate Change, 2001, Impacts, Adaptation, and Vulnerability. Intergovernmental Panel on Climate Change (IPCC) Working Group II, MacCarthy, J.J. et al. (Eds.) Cambridge University Press, Cambridge, UK. (SENAMHI, 2001). International Lake Environment Committee, ILEC. World lakes database (online), http://www. ilec.or.jp/database/index/idx-lakes.html. Irwin, R.P. III, Howard, A.D., 2002. Drainage basin evolution in Noachian Terra Cimmeria, Mars. J. Geophys. Res. 107(E7), 5056–5079. Irwin, R.P., Howard, A.D., Craddock, R.A., Moore, J.F., 2005. An intense terminal epoch of widespread fluvial activity on early Mars. 2. Increased runoff and paleolake development. J. Geophys. Res. 110, E12S14. Irwin, R.P., Maxwell, T.A., Howard, A.D., Craddock, R.A., Leverington, D.W., 2002. A large paleolake basin at the head of Ma’adim Vallis, Mars. Science 296 (5576), 2209–2212. Jaeger, W.L., Kesztheli, L.P., McEwen, A.S., Titus, T.N., Dundas, C.M., Russel, P.S., 2008. Response to comment on “Athabasca Valles, Mars: A lava-draped channel system”. Nature 320 (5883), 1588. Jakosky, B.M., Phillips, R.J., 2001. Mars' volatile and climate history. Nature 412, 237–244. Johnson, F.A., East, J.W. 1982. East Cyclical relationships between river discharge and chemical concentration during flood events. J. Hydrol. 57, 93–106. Kalf, J., 2001. Limnology: Inland WaterEcosystems. Prentice Hall, Upper Saddle River, New Jersey.

26

Lakes on Mars

Kargel, J.S., Furfaro, R., Prieto-Ballesteros, O., Rodrigues, J.A.P., Montgomery, D., et al., 2007. Martian hydrology systained by thermally insulated gas and salt hydrates. Geology 11 (35), 975–978. Kargel, J.S., Lunine, J.I., 1998. Clathrate hydrates on earth and in the solar system. In: Schmitt, B., DeBergh, C., Festou, M. (Eds.), Solar System Ices. Kluwer Academic Press, Boston, pp. 97–117. Kargel, J.S., Tanaka, K.L., Baker, V.R., Komatsu, G., MacAyeal, D.R., 2000. Formation and dissociation of clathrate hydrates on Mars: Polar caps, northern plains, and highlands, 31st Lun. Plan. Sci. Conf., Abstract No. 1891. Kasting, J.F., 1997. Warming early Earth and Mars. Science 276, 1213–1215. Kazemi, G.A., Lehr, J.H., Perrochet, P., 2006. Groundwater Age. Wiley, Hoboken, New Jersey, 346 p. Keszthelyi, L., Self, S., Thordarson, T., 2006. Flood lavas on Earth, Io, and Mars. J. Geol. Soc. 163 (2), 253–264. Kilham, P., 1982. Acid precipitation: Its role in the alkalization of a lake in Michigan. Limnol. Oceanogr. 27 (5), 856–867. Kleinhans, M.G., van de Kasteele, H.E., Hauber, E., 2009. Palaeoflow reconstruction from fan delta morphology on Mars. Earth, Plan. Space Sci. doi: 10.1016. Knoll, A., Carr, M., Clark, B., Des Marais, D.J., Farmer, J.D., et al., 2005. An astrobiological perspective on Meridiani Planum. ESPL 240 (1), 179–189. Komatsu, G., Kargel, J.S., Baker, V.R., Strom, R.G., Ori, G.G., Mosangini, C., et al., 2000. A chaotic terrain formation hypothesis: explosive outgas and outlow by dissociation of clathrate on Mars. 31st Lun. Plan. Sci. Conf., Abstract No. 1434. Kraal, E.R., van Dijk, M., Posma, G., Kleinhans, M.G., 2008. Martian stepped-delta formation by rapid water release. Nature 451m, 973–976. Krug, E.C., Frink, C.R., 1983. Acid rain on acid soil: a new perspective. Science 221, 520–525. Kuzmin, R.O., Greeley, R., Landheim, R., Cabrol, N.A., Farmer, J.D., 2000. Geologic mapping of the Ma’adim Vallis-Gusev crater region of Mars. US. Geol. Surv.Misc. Inv. Ser. MTM, pp. 2256–2257, 1: 500K scale. Lang, N.P., Farrell, A.K., 2009. Volcanic eruption of the Apollimaris Patera summit caldera, Mars, GSA Annual Meeting, Portland, Paper No. 276-9. Laskar, J., Correia, A.C.M., Gastineau, M., Joutel, F., Levard, B., Robutel, P., 2004. Long term evolution and chaotic diffusion of the insolation quantities of Mars. Icarus 170 (2), 343–364. Legates, D.R., Willmott, C.J., 1990. Mean Seasonal and Spatial Variability Global Surface Air Temperature. Theor. Appl. Clim. 41, 11–21. Lehner, B., Döll, P., 2004. Development and validation of a global database of lakes, reservoirs, and wetlands. J. Hydrol. 296, 1–22. Lucchitta, B., 1982. Ice sculpture in the Martian outflow channels. J. Geophys. Res. 87, 9951–9973. Malin, M.C., Carr, M.H., 1999. Groundwater formation of Martian valleys. Nature 397, 589–591. Malin, M., Edgett, K., 2003. Evidence for persistent flow and aqueous sedimentation on early Mars. Science 302, 1931–1934. Mandelbrot, B., 1982. The Fractal Geometry of Nature. Freeman, San Francisco, 365 pp. Mangold, N., Ansan, V., 2006. Detailed study of an hydrological system of valleys, a delta and lakes in the Southwest Thaumasia region, Mars. Icarus 180, 75–87. Max, M.D., Clifford, S.M., 2000. The state, potential distribution, and biological implications of methane in the Martian crust. J. Geophys. Res. 105, 4165–4172.

Searching for lakes on Mars: four decades of exploration

27

McEwen, A.S., Keszthelyi, L.P., Milazzo, M., Burr, D.M., Christensen, P.R., et al., 2002. Athabasca Valles region: New insights from THEMIS. EOS Trans. Am. Geophys. Union 83 (47), Fall Meet. Suppl., Abstract P11B-09. McKay, C.P., Andersen, D.T., Pollard, W.H., Heldmann, J.L., Doran, P.T., Fritsen, C.H., et al., 2005. Polar lakes, streams, and springs as analogs for the hydrological cycle on mars. In: Tokano, T., (Ed.), Adv. Astrobiology and Biogeophys. Springer-Verlag, Berlin, Chapter 9, pp. 219–233. McKay, C.P., Davis, W.L., 1991. Duration of liquid water habitats on early Mars. Icarus 90, 214–221. McLennan, S.M., Bell, J.F. III., Calvin, W.M., Christensen, P.R., Clark, B.C., et al., 2005. Provenance and diagenesis of the evaporite-bearing Burns formation, Meridiani Planum, Mars. Earth Plan. Sci. Lett. 240 (1), 95–120. Mellon, M.T., 1996. Limits on the CO2 content of the Martian polar deposits. Icarus 124, 268–279. Meybeck, M., 1995. Global distribution of lakes. In: Lerman, A., Imboden, D.H., Gat, J.R. (Eds.), Physics and Chemistry of Lakes. Springer-Verlag, Berlin, 1–35. Miller, S.L., Smythe, W.D., 1970. Carbon dioxide clathrate in the Martian ice cap. Science 170, 531–533. Milton, D.J., 1973. Water and processes of degradation in the Martian landscape. J. Geophys. Res. 78, 4037–4047. Milton, D.J., 1974. Carbon Dioxide Hydrate and floods on Mars. Science 183, 654–656. Mischna, M.A., Richardson, M.I., Wilson, R.J., McCleese, D.J., 2003. On the orbital forcing of Martan water and CO2 cycles: a general circulation model study with simplified volatile schemes. J. Geophys. Res. 108 (E6), 5062. Montecinos, A., Aceituno, P., 2003. Seasonality of the ENSO-related rainfall variability in Central Chile and associated circulation anomalies. J. Clim. Am. Met. Soc. 16, 281–296. Moore, J.M., Clow, G.C., Davis, W.L., Gulick, V.C., Janke, D.R., McKay, C.P., et al., 1995. The circum-Chryse region as a possible example of a hydrologic cycle on Mars: Geologic observations and theoretical evaluation. J. Geophys. Res. 100 (E3), 5433–5447. Moore, J., Howard, A., 2005. Large alluvial fans on Mars. J. Geophys. Res. 110, E04005. Morgan, G.A., Head, J.W., 2009. Sinton crater, Mars: Evidence for impact into a plateau icefield and melting to produce valley networks at the Hesperian-Amazonian boundary. Icarus 202, 39–59. Morris, R.V., Ming, D.W., Gellert, R., Yen, A., Clark, B.C., et al., 2008. The hydrothermal system at Home Plate in Gusev crater, Mars: Formation of high silica material by acidsulfate alteration of basalt, 39th Lun. Plan. Sci. Conf., Abstract. Mouginis-Mark, P.J., Robinson, M.S., 1992. Evolution of the Olympus Mons caldera, Mars. Bull. Volc. 54 (5), 347–360. Mouginis-Mark, P.J., Wilson, L., 1999. New styles of Martian volcanic activity revealed by Mars orbiter camera, 5th Mars Conference, Abstract No. 6051. Musselwhite, D.S., Lunine, J.I., 1995. Alteration of volatile inventories by polar clathrate formation on Mars. J. Geophys. Res. 100, 23301–23306. Mustard, J.F., Murchie, S.L., Pelkey, S.M., Elhmann, B.L., Milliken, R.E., et al., 2008. Hydrated silicate minerals on Mars as observed by the Mars Reconnaissance Orbiter CRISM instrument. Nature 454, 305–309. Nester, P.L., Gayo, E., Latorre, C., Jordan, T.E., Blanco, N., 2007. Perennial stream discharge in the hyperarid Atacama desert of northern Chile during the latest Pleistocene. PNAS 104 (50), 19724–19729.

28

Lakes on Mars

Neukum, G., Jauman, R., Hoffmann, H., Hauber, E., Head, J.W., et al., 2004. Recent and episodic volcanic and glacial activity on Mars revealed by the high resolution stereo camera. Nature 432, 971–979. Newsom, H. E.1998. Hot rocks, wet rocks, and deep rocks: Eroded impact crater and channel deposits in Tiu Vallis. In Mars Surveyor 2001 Landing Site Workshop. V.C. Gulick, (Ed.), NASA Ames Research Center, Moffett Field, CA, 40–41. Newsom, H.E., Britelle, G.E., Hibbitts, C.A., Crossey, L.J., Kudo, A.M., 1996. Impact cratering and the formation of crater lakes on Mars. J. Geophys. Res. 101, 14951–14955. Nummedal, D., 1978. The role of liquefaction in channel development on Mars. NASA Tech. Memo. 79729, 257–259. Oren, A., 2001. The biogenic basis for the decrease in metabolic diversity at increasing salt concentrations: implications for the functioning of salt lake ecosystems. In: Melack, J.M., Jellison, R., Herbst, D.B., (Eds.), Saline Lakes, Developments in Hydrology, Kluwer Academic Publishers, USA, pp. 61–72. Ori, G., Marinangeli, L., Baliva, A., 2000. Terraces and Gilbert-type deltas in crater lakes in Ismenius and Memnonia (Mars). J. Geophys. Res. 105 (E7), 17629–17641. Peale, S., Schubert, J.G., Lingenfelter, R.E., 1975. Origin of Martian channels: Clathrates and water. Nature 187 (4173), 273–274. Phillips, R.J., Zuber, M.T., Solomon, S.C., Golombek, M.P., Jakosky, B.M., et al., 2001. Ancient geodynamics and global scale hydrology on Mars. Science 291, 2587–2591. Pieri, D.C., 1980. Martian valleys: Morphology, distribution, age, and origin. Science 210 (4472), 895–897. Poulet., F., Bibring, J.-P., Mustard, J.F., Gendrin, A., Mangold, N., et al., 2005. Phyllosilicates on Mars and Implications for early Martian climate. Nature 438, 623–627. Price, B.P., 2000. A habitat for psychrophiles in deep Antarctic ice. PNAS 97 (3), 1247–1251. Renno, N.O., Bos, B.J., Catling, D., Clark, B.C., Drube, L., Fischer, D., et al., 2009. Physical and thermodynamical evidence for liquid water on Mars. 40th Lun. Planet. Sci. Conf., Abstract No. 1440. Rice, M.S., Bell, J.F. III., Cloutis, E.A., Wang, A., Ruff, S.W., et al., 2010. Silica-rich deposits and hydrated minerals at Gusev crater, Mars: Vis-NIR spectral characterization and regional mapping. Icarus 205 (2), 375–395. Rosenblüth, B., Casassaand, G., Fuenzalida, H., 1995. Recent climate changes in Western Patagonia. Bull. Glacier Res. 13, 127–132. Ruff, S.W.J.D., Farmer, R.E., Arvidson, S.W., Squyres, P.R., 2008. Christensen and Athena Science Team, The nature and distribution of silica at Home Plate in Gusev Crater, Mars: Evidence for a hydrothermal system, Lunar Planet. Sci. Conf. 39, Abstract No. 2213. Russell, P.S., Head, J.W., 2003. Elysium–Utopia flows as mega-lahars: A model of dike intrusion, cryosphere cracking, and water-sediment release. J. Geophys. Res. 108, 5064. Sagan, C., Mullen, G., 1972. Earth and Mars: Evolution of atmospheres and surface tempera­ ture. Science 177 (4043), 52–56. Sagan, C., Toon, O.B., Gierasch, P.J., 1973. Climate change on Mars. Science 181, 1045–1049. Sakimoto, S.E.H., Zuber, M.T., 1998. Flow and convective cooling in lava tubes. J. Geophys. Res. Solid Earth 103 (B11), 27465–27487. Schonfeld, E., 1977. Martian volcanism, 8th Lun. Plan. Sci. Conf., 843–845. Scott, D.H., Chapman, M.G., 1991. Mars Elysium basin: Geologic/volumetric analyses of a young lake and exobiologic implications. Proc. Lun. Plan. Sci. 21, LPI 669–677. Scott, D.H., Dohm, J.M., Rice, J.W., 1995. Map of Mars showing channels and possible paleolake basins (1:30,000,0000). U.S. Geological Survey, Miscellanea Investigation Series, Map I-2461.

Searching for lakes on Mars: four decades of exploration

29

Scott, D.H., Rice, J.W. Jr., Dohm, J.M., 1991. Martian paleolakes and waterways: Exobiological implications. Orig. Life Evol. Biosph. 21 (3), 189–198. Seiler, K.-P., Gat, J.R., (Eds.) 2007. Recharge under different climate regimes. Book Series: Groundwater Recharge from Run-Off, Infiltration, and Percolation. Springer, Netherlands, vol. 55, pp. 159–186. Seinfeld, J.H., Pandis, S.N., 2006. Atmospheric Chemistry and Physics. Wiley and Sons, New York. Smith, P.H., Tamppari, L.K., Arvidson, R.E., Bass, D., Blaney, D., et al. 2009, H2O at the Phoenix landing site. Science 325 (5936), 58–61. Solomon, S.C., Aharonson, O., Aurnou, J.M., Banerdt, W.B., Carr, M.H., et al., 2005. New perspectives on ancient Mars. Science 307 (5713), 1214–1220. Squyres, S.W., Arvidson, R.E., Bell, J.F., III, Brückner, J., Cabrol, N.A., et al., 2004. The Opportunity rover’s Athena science investigation at Meridiani Planum, Mars. Science 306 (5702), 1698–1703. Squyres, S.W., Aharonson, O., Clark, B.C., Cohen, B.A., Crumpler, L., et al., 2007. Pyroclastic activity at Home Plate in Gusev crater, Mars. Science 316 (5825), 738–742. Squyres, S.W., Arvidson, R.E., Ruff, S., Gellert, R., Morris, R.V., et al., 2008. Detection of Silica-Rich Deposits on Mars. Science 320, 1063, doi: 10.1126/science.1155429. Squyres, S.W., Knoll, A.H., Arvidson, R.E., Clark, B.C., Grotzinger, J.P., Joliff, B.L., et al., 2006. Two years at Meridiani Planum: results from the Opportunity rover. Science 313 (5792), 1403–1407. Stivaletta, N., Barbieri, R., Picard, C., Bosco, M., 2009. Astrobiological significance of the sabkha life and environments of southern Tunisia. Plan. Space. Sci. 57 (5–6), 597–605. Sugimoto, A., Naito, D., Yanagisawa, N., Ichiyanagi, K., Kurita, N., Kubota, J., et al., 2002. Characteristics of soil moisture in permafrost observed in East Siberian taiga with stable isotopes of water. Hydrol. Proc. 17 (6), 1073–1092. Toon, O.B., Pollack, J.B., Ward, W., Burns, J.A., Bilski, K., 1980. The astronomical theory of climate change on Mars. Icarus 44, 552–607. Tosca, N.J., McLennan, S.M., Clark, B.C., Grotzinger, J.P., Hurowitz, J.A., Knoll, A.H., et al., 2005. Geochemical modeling of evaporation processes on Mars: Insight from the sedi­ mentary record at Meridiani Planum. ESPL 240 (1), 122–148. Valero-Garcés, B., Delgado-Huertas, A., Ratto, N., Navas, A., Edwards, L., 2004. Paleohydrol­ ogy of Andean saline lakes from sedimentological and isotopic records, northwestern Argentina. J. Paleolimn. 24 (3), 354–359. Wallace, D., Sagan, C., 1979. Evaporation of Ice in planetary atmospheres: Ice-covered rivers on Mars. Icarus 39, 385–400. Warren, R., Arnell, N., Nicholls, R., Levy, P., Price, J., 2006. Understanding the regional impacts of climate change, Report prepared for the Stern Review. Tyndall Centre Working Paper No. 90, Norwich: Tyndall Centre, 223 pages. Warwick, F.V., Laybourn-Parry, J., 2008. Polar Lakes and Rivers, Limnology of Arctic and Antarctic Aquatic Ecosystems. Oxford University Press, New York, 320 p. Willander, B.L., Skejlkvale, D.S., Vuorenmaa, Jeffries, J., Keller, B., Kopacek, J., Vesely, J., 2007. Dissolved organic carbo trends resulting from changes in atmospheric deposition chemistry. Nature 450, 537–540. Williamson, C.E., Saros, J.E., Schindler, D.W., 2009. Climate change: Sentinels of change. Science 323 (5916), 887–888. Wynn-Williams, D.D., Edwards, H.G.M., 2000. Antarctic ecosystems as model for extraterres­ trial surface habitats. Planet. Space Sci. 48, 1065–1075. Yung, Y., Pinto, L. 1978. Primitive atmosphere and implications for the formation of channels on Mars. Nature 273, 730–732.

2 Acquisition and history of water on Mars

Michael H. Carr† and James W. Head* † *

2.1

U. S. Geological Survey, Menlo Park, CA, USA Department of Geological Sciences, Brown University, Providence, RI, USA

Introduction

The purpose of this chapter is to summarize the geologic history of Mars and the role water has played in the evolution of the surface so that subsequent chapters on more specific topics can be viewed in a broader context. It focuses mainly on surficial processes such as erosion, sedimentation, and weathering, rather than on primary terrain-building processes such as impact, tectonism, and volcanism since surficial processes provide more information on surface conditions under which lakes could have formed. The role of liquid water in the evolution of Mars is puzzling. With a mean annual temperature of 215 K and a mean surface pressure of 6.1 mbar (Haberle et al., 2008) liquid water can exist at the surface only locally and temporarily under anomalous conditions. Yet geologic evidence for the widespread presence of liquid water is compelling, particularly for early Mars, and claims have also been made of present-day water activity. One of the outstanding unsolved problems of Martian geology is how conditions necessary for liquid water could have been so sustained at the surface on early Mars as to result in pervasive aqueous weathering and wide­ spread formation of valley networks and lakes. Martian surface features have been divided into three age groups—Noachian, Hesperian, and Amazonian—on the basis of intersection relations and the numbers of superimposed impact craters (Scott and Carr, 1978; Tanaka, 1986). Noachian terrains survive from the early heavy bombardment era. The era was named for the heavily cratered Noachis region, following the long-established terrestrial practice of naming eras after type localities. The rest of Mars’ history was divided into two eras, the Hesperian named for Hesperia Planum and the Amazonian named for the younger Amazonis Planum. From estimates of Martian cratering rates as a function of time (Ivanov, 2001), Hartmann and Neukum (2001) estimated that the Noachian era ended around 3.7 Gy ago and that the Hesperian era ended around 2.9–3.3 Gyr ago. The date of the end of the Noachian is unlikely to be grossly in error, but the date for the Hesperian–Amazonian boundary could incorporate significant errors. Dating younger (<1 Gyr) terrains, where small craters must be used, is even more uncertain because of the nonuniform distribution of secondary craters (McEwen et al., 2005) and the possibility of a long-term decline in impact rates (Quantin et al., 2004). Lakes on Mars. DOI: 10.1016/B978-0-444-52854-4.00002-7 © 2010 Published by Elsevier B.V.

32

Lakes on Mars

The geologic records of Mars and the Earth are very different. Most of the Martian geologic record, particularly that relating to liquid water, dates back to the Noachian and Hesperian, prior to 3 billion years ago, close to the middle of the terrestrial Archean. The record of liquid water from the Amazonian, which constitutes twothirds of Mars’ geologic history, is very sparse, although not absent, and restricted mainly to gullying of slopes, rare groundwater eruptions, and melting of ice. In contrast, most of the geologic record on Earth dates from after 3 billion years ago, the earlier record having been destroyed as a result of the much higher rates of geologic activity on Earth and surface conditions that enabled high rates of erosion and weathering.

2.2 Acquisition and retention of water Excess 182W in Martian meteorites indicates that Mars’ core formed remarkably quickly, within 20 million years of Solar system formation, and the 4.53 Gyr age of ALH84001 shows that at least some crust formed within a few tens of millions of years (Borg et al., 1997; Lee and Halliday, 1997). Rapid core formation and estimates of present average crustal thickness of several tens of kilometers (Zuber et al., 2000) place constraints on thermal evolution models. According to Hauck and Phillips (2002), heat flows would have peaked at 60–70 mWm–2 around 4.4 Gyr ago and then declined almost linearly to a present value of not more than 10–20 mWm–2, and possibly much lower, as suggested by the lack of flexure of the lithosphere under the present polar loads (Johnson et al., 2000; Phillips et al., 2008). According to the Hauck and Phillips model, by 4 Gyr ago over 70% and possibly considerably more of the crust would have accumulated. They also conclude that the mantle must have been wet and that delivery of water and other volatiles such as sulfur to the surface by volcanism during and subsequent to this early era could have affected surface environments. One of the more surprising results of the MGS mission was the discovery of large magnetic anomalies, mostly in the southern highlands (Acuna et al., 1999; Connerny et al., 1999). Anomalies are mostly absent around the large, easily recognizable impact basins. The simplest explanation is that pre-Noachian Mars had a magnetic field that left large anomalies that were subsequently destroyed in and around impact basins such as Hellas, Utopia, Argyre, and Isidis (Solomon et al., 2005 and references therein). Some of the anomalies in the southern uplands are striped, drawing comparisons with terrestrial seafloor features (Connerny et al., 1999), although there is no geomorphic evidence for plate tectonics. Nimmo (2000) alternatively suggested that the anomalies may be due to the presence of deep dike swarms. The amount of water acquired during accretion and subsequently outgassed and retained at the surface to participate in geologic processes is very uncertain. It depends, among other things, on the mix of meteorites and comets that accreted to form the planet, which has been estimated from modeling the mix of meteoritic materials required to reproduce the global composition of Mars inferred from the

Acquisition and history of water on Mars

33

chemistry of Martian meteorites (Dreibus and Wanke 1987) and from dynamical modeling of planet formation and isotopic studies (e.g., Lunine et al., 2003). Estimates of the amount of water originally accreted range up to an amount equivalent to a global layer of many tens of kilometers deep. But we just saw that the core formed very early, within no more than 20 million years of the start of accretion. During global differentiation to form the core water would have outgassed, possibly forming a steam atmosphere (Matsui and Abe, 1987), and would have reacted with metallic iron in the originally accreted material to form FeO and H, which would have outgassed (Dreibus and Waenke, 1987). The early atmosphere probably suffered a massive loss of hydrogen by hydrodynamic escape driven by extreme ultraviolet radiation from the early Sun (e.g., Pepin, 1994; Zahnle et al., 1988). The outflow of hydrogen to space would have carried other atmospheric gases with it, including CO2, N2, and most of the noble gases lighter than xenon (Pepin, 1994). The hydrodynamic phase was over within 200 MY of the start of accretion at which time the Sun’s output of extreme ultraviolet was no longer sufficient to drive the flow. A major uncertainty is the extent to which water incorporated into the planet during accretion was retained after these massive degassing and atmospheric losses. It has been argued that most of the early water was lost and that most of the present inventory of water on both Earth and Mars was delivered mainly by comets and carbonaceous chondrites late during heavy bombardment after the hydrodynamic phase was over (Chyba, 1990; Owen and Bar-Nun, 2000). Dreibus and Waenke (1987) argued against addition of such a late volatile-rich veneer for Mars because of the lack of excess siderophiles in Martian meteorites. However, there are other plausible explanations for the lack of a side­ rophile anomaly, such as a poorly mixed Martian mantle due to a lack of plate tectonics (Carr and Waenke, 1992). D/H enrichment of water in Martian meteorites provides some support for late cometary additions since comets have a higher D/H ratio than asteroids, the source of most of the original accreted materials (Baker et al., 2005); however, high D/H ratios also result from preferential loss of H from the upper atmosphere. After the hydrodynamic phase was over, the atmosphere would have been supple­ mented by further outgassing from the interior and depleted by various processes including erosion by large impacts (Melosh and Vickery, 1989) and losses by weath­ ering to form carbonates and other minerals. Losses from the upper atmosphere would have been largely restricted to hydrogen until the magnetic field turned off, which is estimated to have been around 4 Gyr ago (Connerny et al., 1999). After this time, impingement of the solar wind on the upper atmosphere would have resulted in enhanced losses of heavier species such as O and N as a result of ion pickup and sputtering. These losses continued for the rest of the planet’s history (Jakosky and Jones, 1997). Preferential loss of hydrogen over deuterium from the upper atmosphere can, in principle, be used to estimate the size of the water reservoir that was originally present in the Noachian and exchanging with the atmosphere ever since. Lammer et al. (2005), for example, estimate that 3.5 Ga ago this reservoir was the equiva­ lent of 35–115 m spread over the entire planet. Unfortunately we have no way of estimating the size of the reservoir, such as deep ice and groundwater, which is not

34

Lakes on Mars

equilibrating with the atmosphere. Nor do we know whether present loss rates of hydrogen are representative of the last 3.5 Ga. The average obliquity over the last 3.5 Ga is 40°, significantly higher than the current 25° (Laskar et al., 2004). At 40° obliquity, the water content of the atmosphere could have been higher than the present by a factor of 100 (Mellon and Jakosky, 1995), thereby leading to enhanced hydrogen losses from the upper atmosphere and the possibility of enriching a larger reservoir. In view of all the uncertainties outlined above, we must conclude that modeling of accretion and atmospheric evolution does not place strong constraints on the amount of water available for geologic processes. In an alternative approach, Carr (1986) estimated that a global equivalent of roughly 500 m of water was required to transport the material eroded away to form the outflow channels, but this figure also has large uncertainties.

2.3 Early geologic events How deep into the era of heavy bombardment can the geologic record be discerned from the surface topography is unknown. Part of the uncertainty stems from the cratering history: whether there was a late spike in basin formation around 3.9 Gyr ago (Tera et al., 1974) or a steady decline after accretion (Stöffler et al., 2006). Assuming a steady decline and using the Hartmann and Neukum (2001) estimates of the cratering rate in the late heavy bombardment period, Frey (2003) estimated that Hellas formed around 4.1 Gyr ago from the number of basin-like features super­ imposed on its rim. This number should, however, be viewed with considerable caution because of all the assumptions involved. Frey also suggested that Hellas be taken as the base of the Noachian and that the era from 4.55 to 4.1 Gyr ago be referred to as pre-Noachian. Possibly the earliest geologic event recorded in the topography of the surface is the formation of the global dichotomy (Carr, 2006; Nimmo and Tanaka 2005; Solomon et al., 2005). The dichotomy is expressed in three ways that do not coincide every­ where: as differences in elevations, as differences in crustal thickness, and as differ­ ences in crater densities. The dichotomy results in a bimodal distribution of elevations, with a difference of 5.5 km between the two hemispheres (Aharonson et al., 2001). Neumann et al. (2004) estimate that the thickness of the crust averages roughly 30 km north of the dichotomy boundary and roughly 60 km to the south. As expected, the differences in crater densities across the boundary may be only a superficial difference for a densely cratered surface that is present at depths below the present Hesperian– Amazonian surface north of the dichotomy as indicated by remnants of old craters that poke up through the younger plains and by vague circular outlines in both the Mars Orbiter Camera (MOC) images and Mars Orbiter Laser Altimeter (MOLA) data. The low-lying, heavily cratered Noachian surface, north of the dichotomy boundary, is simply covered by younger deposits. A distinction must also be made between the time of formation of the depression and the time of formation of the fill. The number of craters superimposed on the fill yields little information about the age of the

Acquisition and history of water on Mars

35

depression itself. From the geologic evidence the dichotomy could have formed at any time between the formation of the crust 4.5 Gyr ago and the formation of the oldest of the clearly superimposed impact basins, such as Utopia and Chryse, around 4.1 Gyr ago according to the Frey (2003) chronology. The mode of formation of the dichotomy is also uncertain. One possibility is that the dichotomy is the result of one or more large impacts (Andrews-Hanna et al., 2008; McGill and Squyres 1991; Wilhelms and Squyres 1984). The outline of the basin is roughly circular except in Tharsis where younger volcanics are superimposed on the boundary and in Chryse where there may be a younger superimposed basin. Zuber et al. (2000) and Neumann et al. (2004) expressed skepticism that the northern lowlands could be an impact scar because there is little evidence for extreme thinning of the crust as there is within Hellas and Isidis, nor is there a perceptible rim around the basin. They prefer an early internal origin, tied to global mantle convection (Wise et al., 1979; Zuber et al., 2000; Zhong and Zuber 2001; Solomon et al., 2005). However, the thicker crust and absence of a rim around the proposed impact basin may simply reflect an extremely old age, and Andrews-Hanna et al. (2008) have recently attempted to reconcile the geophysical data with an impact origin. If the basin formed very early, soon after formation of the crust, it would have experienced erosion, sedimentation, isostatic rebound, and volcanic filling for hundreds of millions of years, an era almost as long as the terrestrial Phanerozic, before a more complete geologic record emerged after the formation of the Hellas basin at the start of the Noachian. Surface conditions in this early era prior to the formation of Hellas are very uncertain. One certainty is that the surface was episodically disrupted by very large, basin-forming impact events. Formation of these large (>500 km diameter) craters and basins would have resulted in the ejection of large amounts of rock vapor and rock melt into and beyond the atmosphere, evaporated any oceans that might have been present, and raised the surface temperatures to several hundred kelvin (Segura et al., 2002; Sleep and Zahnle 1998). Despite the low solar luminosity, surface temperatures could have remained above freezing for years after each large impact event. Water that was injected into the atmosphere during the initial impact and during the subsequent warming of the surface and subsurface could rain out over years, the time depending on the size of the impact. Conditions during the long (possibly millions of years) periods between basin-forming events would have depended on the effects of smaller impacts and on the ability of the atmosphere to provide significant greenhouse warming during this era of low solar luminosity, which in turn would have depended on the thickness of the atmosphere and it composition, particularly the abundance of trace greenhouse gases such as CH4 and SO2. In summary, the geologic record of the pre-Noachian era extending from the time of formation of the planet 4.5 Gyr ago to the time of formation of Hellas estimated at around 4.1 Gyr ago is sparse. The planet differentiated into crust, mantle, and core within a few tens of millions of years of planet formation, the global dichotomy probably formed early, and the planet had a magnetic field. Large impact craters and basins that formed episodically would have had devastating environmental effects.

36

Lakes on Mars

However, the nature of the atmosphere, the surface inventory of volatiles, and surface conditions between large impact events are all unknown.

2.4 The Noachian era If the above chronology is correct, the Noachian era extends from approximately 4.1–3.7 Gyr ago, roughly coincident with the upper Hadean on Earth. Its most distinguishing features compared with later times are high rates of cratering, erosion, and valley formation, the accumulation of most of Tharsis, and surface conditions that enabled the widespread production of weathering products such as phyllosilicates. This is the era for which we have the best evidence for widespread water erosion. The density of visible craters larger than 100 km in diameter in Noachian terrains is roughly 2 � 10–6 km–2 (Strom et al., 1992), or 300 such-sized craters planet-wide, implying that one 100 km diameter crater formed every million years. The impacts would have ballistically distributed ejecta around the planet, caused hydrothermal activity around the impact sites, comminuted surface materials thereby enabling them to be moved by wind and water, and brecciated the near-surface materials thereby increasing their porosity, and so affecting groundwater movement and storage. Noachian craters with diameters between 500 and 1000 km would have deposited roughly 300 m of ejecta planet-wide (Segura et al., 2002). Hellas alone would have deposited 500 m. The coarser fraction from all these impacts would have formed bedded deposits with thicknesses depending on the size of the impact events and their proximity to the resulting craters. The fate of the finer ejecta was likely more complicated. Large areas of the Noachian terrain have an etched appearance (Greeley and Guest, 1987; Malin and Edgett, 2001; Scott and Tanaka 1986) as though parts of the surface had been formerly covered with easily erodible, horizon­ tally layered deposits that had been partly removed by the wind. Fine-grained impact ejecta are likely a significant component of these deposits, along with volcanic ash, as well as the products of weathering and erosion as discussed below. Correlations between gravity and topography suggest that the densely cratered terrain of the southern highlands has surface densities of 2500–3000 kg m–3 (McGovern et al., 2004), signifi­ cantly lower than the density of the Tharsis volcanics and Martian meteorites (3100–3300 kg m–3), and consistent with a crust that has been modified by impacts, erosion, and sedimentation. While volcanism likely occurred almost everywhere, Tharsis was particularly active, resulting in a volcanic pile roughly 5000 km across and 9 km high by the end of the Noachian (Phillips et al., 2001). Large impact basins and the northern basin may also contain significant amounts of Noachian volcanic fill that is buried by younger deposits. Almost everywhere else, the rates of volcanic resurfacing were low compared with the impact rate so that what is preserved in the morphology is an impact-cratered surface on which almost all traces of Noachian volcanic morphology have been destroyed. Despite the scarcity of geomorphic evidence for volcanism, most of the materials exposed in the cratered uplands are probably primarily volcanic rocks or volcanic rocks reworked by impacts. They are mainly basalts rich in low calcium pyroxene, with variable amounts of olivine (Bibring et al., 2006; Poulet et al.,

Acquisition and history of water on Mars

37

2007). The Columbia Hills (Squyres et al., 2006) may be typical of the cratered Noachian uplands in general. They comprise mostly basaltic rocks of various types, including pyroclastic flows and impact breccias. Many of the rocks have undergone aqueous alteration, suggestive of circulation of hydrothermal fluids. Detection of primary igneous minerals, particularly olivine, in much of the Noachian terrain (Bibring et al., 2006) may indicate limited weathering after the deposition of the uppermost layers. However, the widespread presence of hydrated silicates deeper in the section and in alluvial fans indicates widespread aqueous alteration prior to deposition of the upper olivine-rich units. Formation of Tharsis deformed the Martian lithosphere on a global scale to create a trough around the rise, an antipodal high, and gravity anomalies, as predicted by loading of a spherical elastic shell with the Tharsis topography (Phillips et al. (2001). That Tharsis was largely in place at the end of the Noachian is demonstrated by slope indicators such as valley networks and lava flows. Roughly 3 � 108 km3 of rock accumulated to form Tharsis, the equivalent of a global layer 2 km thick. If the magmas contained amounts of water similar to the Hawaiian basalts, the global equivalent of a layer of water 120 m deep would have been outgassed, together with significant amounts of sulfur. If all of Tharsis accumulated in the Noachian, the extrusion rate would have been 0.75 km3 year–1, roughly equivalent to the Hesperian extrusion rate estimated by Greeley and Schneid (1991) for the entire planet. For comparison, the extrusion rate for the Earth is 4 km3 year–1 (Crisp, 1984). Another possible site of large accumulation of Noachian volcanics is the northern basin, including Utopia.

2.4.1 Erosion rates The Noachian terrains are clearly more eroded than younger terrains. While Hesperian craters as small as a few kilometers across generally preserve all their primary impact features, even delicate textures on their ejecta, Noachian impact craters hundreds of kilometers across mostly have highly eroded rims and partly filled interiors. However, even though average Noachian erosion rates were 2–5 orders of magnitude higher than they were subsequently, they still appear to have been close to or well below terrestrial rates (Carr, 1992; Golombek and Bridges, 2000; Golombek et al., 2006). The number of fresh appearing craters with well preserved ejecta patterns on Noachian terrains is comparable to the number on Hesperian terrains, which suggests that high erosion rates persisted until the end of the Noachian and then rapidly declined (Craddock and Maxwell, 1993). Low average rates of erosion in the Noachian compared with the Earth are consistent with preservation of the planet’s larger features. The Noachian era is roughly equivalent to the time on Earth from the end of the Silurian to the present day. On Earth, during this time, continents assembled and disassembled, the presentday ocean basins opened, and numerous mountain chains formed and were eroded away. The fact that the Hellas basin is preserved gives an indication of the average Noachian erosion rates. The denudation rate for the continental United States is roughly 50 m 10�6 years (Judson and Ritter, 1964), or 20 km in 400 Myr, the

38

Lakes on Mars

estimated length of the Noachian. Clearly the rim of Hellas has not been eroded by 20 km. We do not know how much fill is in Hellas, but with a depth of over 9 km, it is unlikely to contain more than a few kilometers. If we assume the floor of the basin (2.5 � 106 km2) has 2 km of fill derived from the surrounding drainage area (17 � 10�6 km2), we derive a denudation rate of 0.75 m 106 years, almost 2 orders of magnitude lower than the US rates. Thus, the data from crater preservation and the paucity of filling within Hellas are consistent. Average Noachian erosion rates, while orders of magnitude greater than the rates for subsequent eras, still fell short of terrestrial rates.

2.4.2 Valley networks Valley networks provide compelling evidence of former conditions that enabled sustained flow of liquid water across the Martian surface. Much, but not all, of the Noachian terrain is dissected by valley networks. Most drain into local lows and are only up to a few hundred kilometers long, particularly in Cimmeria and Serinum where there is no strong regional slope. However, between Syrtis Major Planitia and Argyre several valleys thousands of kilometers long drain northwest down the long regional slope from the high ground around Hellas toward the Chryse–Acidalia low. Stream profiles are poorly graded and closely follow the regional slopes (Howard et al., 2005). There is little indication that formation of the presently identifiable valleys resulted in a general lowering and grading of the landscape as occurs with long-lived terrestrial rivers. The result is low basin concavities (Aharonson et al, 2001), poorly graded stream profiles, and poor correlation of basin circularity with elevation within the basins (Stepinski and O’Hara, 2003). Drainage densities vary considerably with location, up to the low end of the terrestrial range (Craddock and Howard, 2002; Hynek and Phillips, 2003). The apparent low drainage densities, amphitheater heads of tributaries, and rectangular cross-section suggested to many early workers that groundwater sapping had played a major role in the formation of many of the valleys (Baker, 1990; Carr and Clow, 1981; Gulick, 1998, 2001; Pieri 1980), although all acknowledged that precipitation and/or hydrothermal circulation were/was needed to recharge the groundwater system to enable sustained or episodic flow. Better imaging and altimetry now show that dense, area-filling networks are common throughout the Noachian terrains (Figure 2.1, see also Chapter 10). They indicate that precipitation followed by surface runoff, coupled with infiltration and groundwater seepage, must have occurred at least episodically in the Noachian (Carr, 2006; Howard et al., 2005; Hynek and Phillips 2003; Irwin and Howard, 2002). Major uncertainties are how persistent conditions necessary for precipitation and surface runoff were sustained and how such conditions were achieved. Many lows, such as craters having inlet and outlet valleys, indicate that lakes formerly occupied lows in the dissected terrains, as expected for a poorly graded landscape undergoing fluvial erosion (Cabrol and Grin, 1999, 2001, 2002, 2005; Fasset and Head, 2008) (Figure 2.2). Deltas or alluvial fans are commonly observed where valleys enter the lows. Particularly striking examples of deltas are in Eberlswalde crater, Holden Crater (Chapter 12), and in the Nili Fossae (Fassett and

Acquisition and history of water on Mars

39

Figure 2.1 The Warrego Valles at 42° S, 267° E. The dense drainage network strongly suggests precipitation and surface runoff (THEMIS).

Figure 2.2 Possible site of former lakes at 4° S, 111°E. Near the center of the figure, Tinto Vallis breaches the southern rim of the crater Palos. The northern rim is also breached, suggesting flow into the crater from the south and out to the north. Several other craters nearby have similarly breached rims. If the valleys are fluvial, water must have pooled in all the breached craters (THEMIS).

40

Lakes on Mars

Head, 2005). The dimensions of the channel remnants on the Eberswalde delta suggest that the discharges were comparable to terrestrial streams draining similarsized basins (Moore et al., 2003) and that the deltas and fans may have taken only decades to form (Jerolmack et al., 2004). Chlorine-rich deposits found in local lows within the Noachian uplands may be the result of evaporation of lakes (Osterloo et al., 2008). Some of the sulfate-rich deposits found in Meridiani may have been deposited in transient inter-dune lakes, and subsequently altered as a result of oscillations in the local groundwater table (Grotzinger et al., 2005; McLennan et al., 2005). The Mer­ idiani deposits are discussed more in detail below under the Hesperian. Howard et al. (2005) suggested that the more pristine valleys incised into the highland terrains are the result of a late Noachian to early Hesperian episode of incision. They make a distinction between the general degradation of the landscape and forma­ tion of the incised valley networks. They suggest that during most of the Noachian there was widespread fluvial erosion of crater rims and other high ground and partial infilling of lows such as craters, but that formation of the incised networks was fundamentally different. They were incised into a degraded landscape, but contributed little to that degradation. They form an immature drainage system in which individual valleys are poorly graded and basin development by erosion and alluviation barely occurred. Some support for the late incision model in which the more pristine, more easily detected valleys contribute little to the general landscape degradation is the observation that areas that appear only sparsely dissected, such as the region between Hellas and Argyre, are just as degraded as the highly dissected areas. Degradation of the Noachian landscape must have produced large amounts of erosional debris. The partial to nearly complete filling of Noachian craters of all sizes is a significant possible sink for erosional products. Malin and Edgett (2000) conclude that much of the crater fill consists of layered, sedimentary rocks. They also point out the common presence of layered rocks in intercrater areas, canyons, and areas of chaotic terrain. On the basis of their erodibility, the presence of steep scarps, and the lack of boulders at the bases of scarps, they conclude that most of these layered Noachian deposits are indurated, fine-grained sediments rather than coherent volcanic rocks. Unknown amounts of fill could also be hidden under younger deposits in the low-lying northern plains. One of the most striking characteristic of these sediments, irrespective of their age, is their rhythmic layering, which in many cases is remarkably regular (Malin and Edgett, 2000) (Figure 2.3). The layering could result from a variety of causes such as successive impacts and volcanic events or changes in the erosional regime as a result of climate changes. While all these three processes likely contributed to the sediments, the extreme regularity of some of the layering argues against volcanism and impacts as a primary cause, at least in these cases. The rhythmic depositional patterns suggest an astronomic cause such as changes in erosion rates brought about by climate changes, which in turn result from periodic changes in the orbital and rotational motions of the planet (Laskar et al., 2004). Many of the fluvial features found on post-Noachian terrains were formed by large floods. However, despite widespread dissection during the Noachian, large floods appear to have been rare. Ladon Vallis is one example. It may be part of a large ancient

Acquisition and history of water on Mars

41

Figure 2.3 Finely layered sediments with unconformities within Galle crater at 52.3° S, 329.9° E. The mode of deposition, whether by water or wind, is unknown, but the regular rhythmic layering suggests that deposition was modulated by changes induced by astronomic motions (MOC).

waterway extending from close to the rim of Argyre into the northern plains. It has been attributed to overflow of a lake in Argyre (Parker et al., 2000), although that interpretation has been challenged (Heisinger and Head, 2002). Another possible Noachian flood feature is Ma’adim Vallis, which Irwin et al. (2002) plausibly argue was formed in part by rapid drainage of a large lake upstream from the main valley. The widespread dissection of Noachian terrains coupled with surface runoff pat­ terns indicates at least episodic precipitation and temperature and pressure conditions that stabilized water at the surface. Nevertheless, there is considerable uncertainty as to how sustained such conditions were and whether there was ever a global hydrologic system in which precipitation, infiltration, runoff, and groundwater flow were in quasi-equilibrium with evaporation and sublimation from large bodies of water and ice. Despite considerable relief along the dichotomy boundary and around Hellas, large drainage basins analogous to the Mississippi and Amazon did not develop. Seemingly, the cumulative effects of erosion, alluviation, and stream capture were insufficient to result in integration of drainage over large areas and growth of large drainage basins before being destroyed by impacts. There are, for example, no significant valleys draining into Hellas from the north and west despite several kilo­ meters of relief and despite the area having experienced 300–400 My of erosion during the Noachian. Even if Hellas were filled with water to the –3.1 km level as suggested by Moore and Wilhelms (2001) (Figure 2.4, see also Chapter 7), there are still 5 km of relief from the rim crest down to the proposed sea level to enable

42

Lakes on Mars

Figure 2.4 Swirling textures on the floor of Hellas at 40° S, 52° E. The textures suggest plastic deformation, the result possibly of slumping of waterlogged sediments or interaction between ice and the sediments (HIRISE).

drainage. If the observed degradation of craters superimposed on the rim was due to fluvial erosion, then most of the drainage was likely local with the water accumulating in local lows to be lost by infiltration or evaporation. Such a scenario is also consistent with the apparent failure to transport large amounts of sediment from the Hellas drainage basin into the central depression. The sediment eroded from the highs must simply have accumulated in local lows. Whether there were ever oceans on Mars is one of the planet’s most controversial issues (Carr and Head, 2002; Clifford and Parker, 2001; Head et al., 1999; Parker et al., 1989, 1993, see also Chapters 9 and 10). Discussion has focused mainly on the possibility of post-Noachian oceans because they could have resulted from the large post-Noachian floods discussed later and because any evidence for oceans would be better preserved for the post-Noachian than that for the Noachian. However, the Noachian is the time for which we have the best evidence for conditions under which oceans might be present. Clifford and Parker (2001) argue from estimates of the global inventory of water and the thermal conditions implied by the valley networks that possibly one-third of the planet was covered by oceans during parts of the Noachian. Moore and Wilhelms (2001) identify two possible Noachian shorelines within Hellas (see Chapter 7), and Howard et al. (2005) proposed that the absence of valleys in the Noachian of northwest Arabia resulted from burial by sediments, along the periphery of a northern ocean. Despite these suggestions, the prospect for finding compelling geomorphic evidence of former Noachian oceans is poor, since such

Acquisition and history of water on Mars

43

evidence, if it ever existed, would be vulnerable to erasure by burial and erosion. Nevertheless, if during the Noachian, Mars had a large inventory of water and if ever warm condition prevailed, as is indicated by the valley networks, then bodies of water would have accumulated in lows such as the northern basin and Hellas.

2.4.3 Weathering A distinguishing feature of the Noachian as compared with later eras is the widespread presence of phyllosilicates, such as nontronite, Fe-rich chlorites, saponite, and mon­ tmorillonite (Bibring et al., 2006; Murchie et al., 2008), minerals that all form by the aqueous alteration of basalts (e.g., Zolotov and Mironenko 2008). Weathering to form these minerals probably also occurred in the pre-Noachian but the evidence has been largely destroyed. In some places the phyllosilicates appear to be excavated from below the surface or are overlain by unaltered, olivine-rich rocks. Mustard et al. (2007) show, for example, that olivine-rich rocks overlie phyllosilicates in the Nili Fossae and suggest that they formed from impact melts produced by the event that formed the Isidis basin. The relations suggest that prior to the uppermost Noachian, conditions were such that phyllosilicates could form, but conditions changed toward the end of the era such that rocks that formed at the end of the Noachian retain their primary mineralogy. The presence of phyllosilicates in Noachian terrains and their absence in younger terrains suggest that near the end of the Noachian, surface conditions changed from warm wet conditions under which hydrous weathering could occur, at least occasionally, to colder, drier conditions under which hydrous weathering was suppressed.

2.4.4 Noachian Climates The geomorphic evidence for lakes and rivers, the widespread presence of phyllosi­ licates in Noachian terrains, and the evidence for groundwater movement and surface water at Meridiani (Grozinger et al., 2005) all suggest at least episodic warm condi­ tions during and at the end of the Noachian. Greenhouse models indicate that it is very difficult to raise global temperatures sufficiently to allow widespread precipitation on early Mars with only a CO2–H2O atmosphere because of Mars’ distance from the sun, the expected low energy output of the Sun, and formation of CO2 clouds (Haberle, 1998; Kasting, 1991). In addition, we saw above that a thick CO2 atmosphere is difficult to sustain against impact erosion and weathering. Although some carbonate rocks have been detected from orbit (Ehlman et al., 2008), failure to detect widespread carbonate deposits (Bibring et al., 2006) argues against a thick (>1 bar) CO2 atmo­ sphere during and particularly at the end of the Noachian, when the most prominent valleys formed. If surface conditions on Mars were at least episodic such as to stabilize liquid water near the end of the Noachian, some mechanism other than, or in addition to, warming by a CO2-H2O greenhouse seems to be required. Possibilities include the presence of other greenhouse gases such as SO2 and CH4, or large-scale climatic perturbations resulting from large impacts or large volcanic events. Segura et al. (2002) suggest that large impacts would warm the surface and inject significant

44

Lakes on Mars

amounts of water into the atmosphere that could precipitate out over decades to form the valleys. They estimate, for example, that formation of 600, 1000, and 2500 km craters would result, respectively, in global precipitation of 2, 9, and 16 m of water. Some possible difficulties with the model are (i) the modest amounts of precipitation that result from these very large impacts, which are few in number; (ii) all the Noachian craters with the above sizes are highly eroded and must be much older than the valley networks that we observe, particularly the more pristine ones; and (iii) the two best preserved impact basins, the 200 km diameter Lowell and the 220 km diameter Lyot, are only minimally dissected by valley networks. Thus, while the mineralogic and geomorphic evidence for warm conditions near the end of the Noachian are convincing, how such conditions were achieved remains obscure. This discrepancy is one of the most puzzling aspects of Mars’ evolution.

2.5 Hesperian era The Hesperian era was initially invoked to distinguish old post-Noachian plains such as Hesperia Planum and Lunae Planum from younger plains such as those in Tharsis and Amazonis (Scott and Carr 1978). It was subsequently defined more quantitatively according to the number of superimposed craters (Scott and Tanaka 1986). The crater densities suggest that the period extends from the end of heavy bombardment around 3.7 Gyr ago to around 3 Gyr ago (Hartmann and Neukum, 2001), roughly coinciding with the lower Archean on Earth. The main characteristics of the Hesperian era are continued possibly episodic volcanism to form extensive lava plains, low rates of valley formation compared with the Noachian, formation of the canyons, formation of the largest outflow channels and their terminal lakes or seas, extremely low rates of erosion, a steep decline and possibly cessation of rock alteration to form phyllosili­ cates, and accumulation locally of sulfate-rich deposits, particularly in the western hemisphere (Figure 2.5). The steep decline in rates of erosion, weathering, and valley formation strongly suggests that surface conditions favorable to aqueous erosion and

Figure 2.5 Sulfate-rich layered deposits — gypsum (CaSO4.2H2O) overlying Kieserite (MgSO4.H2O) — in Juventae Chasma at 3° S, 297° E. Their origin is controversial. Alternate suggestions are that they are remains of the materials into which the canyon is cut, or that they were deposited by water or wind on the canyon floor after the canyon formed (HRSC).

Acquisition and history of water on Mars

45

weathering, seemingly common in the Noachian, were rare in the Hesperian. Thus, while the main era of local lake formation that accompanied formation of the valley networks was over by the start of the Hesperian, the main era of flooding and consequent formation of very large bodies of water was in the Hesperian. Hesperian volcanism is evident mainly in the form of ridged plains. In the western hemisphere Hesperian lava plains occur mainly around the eastern periphery of Tharsis. In the eastern hemisphere, they form Hesperia Planum, Syrtis Major Planum, Male Planum, and part of the floor of Hellas. Hesperian ridged plains, present in local lows throughout the cratered uplands of both hemispheres, may also have a significant volcanic component. Partly buried craters and subdued ridges in Vastitas Borealis suggest that the northern plains are underlain by Hesperian volcanics that are con­ tinuous with the volcanic ridged plains further south (Head et al., 2002) and Hesperian plains almost certainly underlie the younger Amazonian plains of central Tharsis and Elysium. The large Tharsis shields, including Olympus Mons, probably started to accumulate in the Hesperian, or even earlier despite the young ages of the present surfaces. Thus volcanism was widespread in the Hesperian, continuing at a rate of �1 km3 year–1, comparable to the Noachian (Greeley and Schneid, 1991). It resulted in resurfacing roughly 30% of the planet, if we assume that in central Tharsis and Elysium Hesperian volcanics underlie the younger Amazonian (Tanaka et al., 1986).

2.5.1 Valleys and Channels The rate of formation of valley networks declined precipitously at the end of the Noachian. Despite the decline, there are examples of Hesperian, and even Amazonian valley networks, particularly on volcanoes (e.g., Alba Patera and Ceraunius Tholus). A rare example of a heavily dissected Hesperian plain is that adjacent to southern Echus Chasma (Mangold et al., 2004). Within the uplands are numerous examples of valleys cutting, or having deposited sediments, upon what appears to be Hesperian plains in local lows. Thus, although a change in conditions resulted in the dramatic drop-off in valley formation at the end of the Noachian, conditions were occasionally such that fluvial erosion to form small valleys was enabled, at least locally. In contrast, most of the large outflow channels formed in the Hesperian, particu­ larly the upper Hesperian (Tanaka et al., 2005). The most important question con­ cerning outflow channels is whether or not they were carved by liquid water. Some outflow channels have features in common with lunar and venusian rilles, including abrupt beginnings, streamlined islands, inner channels, anastomosing reaches, and terraces (Leverington et al., 2004). Lava flows are clearly visible in some outflow channels (e.g., Marte Vallis, Athabasca Valles), and the source of some outflow channels (e.g., Cerberus Fosssae) are also sources of lava flows. Boulders, omnipre­ sent in the low-lying northern plains at the ends of the large Chryse channels, suggest lava flows at the surface rather than fluvial sediments (McEwen et al., 2007). Despite these observations, a fluvial origin for most of the large outflow channels seems secure. The lunar and venusian rilles are only a few kilometers across as compared with tens of kilometer widths of Kasei, Ares, Mangala, and others. Most of the rilles are simple in form and lack the rich array of landforms that are common to both the

46

Lakes on Mars

Martian outflow channels and large terrestrial flood features (Baker and Milton, 1974). In addition, the floors of several rubble-filled channel sources (e.g., Juventae Chasma and Aromatum Chaos) are at a much lower elevation than the outgoing channels and yet show no evidence of former lava lakes such as draping of the source depressions by lava. An aqueous origin is also supported by sulfate-rich deposits in several source depressions such as Juventae chasma (Gendrin et al., 2005) (Figure 2.6). Although some of the simpler, rille-like channels, such as Hrad Vallis, may be cut by lava, the following discussion will assume that the larger, more complex outflow channels such as Kasei, Tiu, Simud, Ares, Mangala, and Maja were cut by large floods of water. If this assumption is correct then large bodies of water must have been left in the lows at the ends of the channels when the floods were over. The abrupt start of outflow channels indicates that they are formed not by surface drainage immediately following precipitation but by the rapid release of large volumes of stored water. The storage medium could be an aquifer, or a lake, as with the Channeled Scablands of Eastern Washington, or ice, as with Icelandic jokulhlaups. All three possibilities may be represented on Mars: (i) several large channels that emerge from rubble-filled depressions south of the Chryse basin, and others elsewhere that start at graben, appear to have formed by eruptions of groundwater (Carr, 1979) (Figure 2.6); (ii) drainage of lakes in the Valles Marineris is suggested by eroded sections of Ganges, Eos, and Capri Chasmas and the mergers of Kasei Vallis with Echus chasma and of Maja Vallis with Juventae Chasma (Luchitta et al., 1992;

Figure 2.6 The lower reaches of Mangala Vallis at 18° S, 210° E. The channel starts abruptly at a graben and extends for over 1000 km to the north. The graben may have acted as a conduit allowing deep groundwater access to the surface, possibly in combination with injection of dikes (THEMIS).

Acquisition and history of water on Mars

47

McCauley, 1978; see also Chapters 5 and 6); and (iii) Chasma Boreale may have formed from meltwater from the north polar cap (Clifford, 1980; Fishbaugh and Head, 2002). Numerous estimates have been made of peak discharges and the volumes of water involved in the floods. The main difficulties are knowing the flood depth and how long the floods lasted. Channel depths can be measured but they give only an upper limit for the stream depth. Most estimates of peak discharges for the largest channels range from 107 to 108 m3 s–1 depending on the channel and the assumed depth (Baker, 1982; Leask et al., 2007; Robinson and Tanaka 1990). If the floods formed by water, then a groundwater source appears almost inescapable for channels such as Shalba­ tana, Tiu, Maja, and Ares that originate in chaos-filled depressions and for those such as Mangala and Athabasca that originate at graben. The sources are likely extensive aquifers trapped below a thick cryosphere. The discharges from such aquifers would be restricted by the dimensions of the aquifers, their permeability, the hydrostatic head, and the dimensions of the conduit to the surface (Carr, 1979; Manga, 2004). Andrews-Hanna and Phillips (2007), by modeling the eruption of groundwater from an overpressurized aquifer trapped below a kilometers thick cryosphere, estimated that for a typical Ares flood peak discharge ranged from 106 to 107 m3s–1 and that 103–104 km3 of water were erupted. The volume of Ares Vallis is roughly 8 � 104 km3, so many floods may have been needed to erode it, according to this model. The high discharges require that the aquifer be pressurized. This could result simply from the aquifer topography and supply of water from highs such as Tharsis and Elysium (Carr, 1979; Harrison and Grimm, 2005a, 2005b) or from tectonic pressurization, particularly for channels such as Mangala and Athabasca that start at faults (Hanna and Phillips, 2005). Emplacement of dikes may also have contributed to water release, by melting of ground ice and creating fractures that act as both horizontal and vertical conduits (Head et al., 2003). The apparent scarcity of groundwater eruptions to form large floods in the Noachian may have resulted from the lack of a thick cryosphere. Their repeated occurrence in the Hesperian may be another consequence of a change in surface conditions at the end of the Noachian that is implied by the decline in the formation of valley networks and hydrated weathering products. The change led to the growth of a thick cryosphere, thereby enabling the trapping of water and large groundwater eruptions. The decline in groundwater eruptions toward the end of the Hesperian could result from a variety of causes such as depletion of water below the cryosphere, growth of the cryosphere to engulf most of the high-porosity megaregolith, and declining tectonic and volcanic activities.

2.5.2 Valles Marineris The Valles Marineris present some of the most puzzling issues of Martian geology, including how and when they formed, the origin of their interior layered deposits, whether the canyons ever contained lakes, and if so how the lakes formed and dissipated (Chapters 5 and 6). The primarily structural origin by movement along faults radial to Tharsis was recognized early (Blasius et al., 1977; Sharp, 1973).

48

Lakes on Mars

NNE–SSW extension to form the rifts may have occurred over hundreds of millions of years, and thinned the crust under the canyons (Anderson and Grimm, 1998). Exten­ sion may also have been accompanied by dike intrusions (McKenzie and Nimmo 1999; Mège and Masson 1996). East of roughly 310 E structural control by Tharsis radials is much less obvious as the roughly E–W canyons merge with more northerly trending outflow channels. The age of the canyons is difficult to determine precisely. Since Tharsis appears to have largely been formed by the end of the Noachian, it is likely that the canyons started to form in the Upper Noachian, although we have no observational evidence. Side canyons and gullies on the canyon walls cut Hesperian plains and are themselves cut by faults. There is little if any evidence of alluvial fans on the canyon floors. The canyon floors were probably still subsiding when erosion occurred, and the fans that resulted were either eroded away or buried by younger deposits. In contrast, the landslides, the youngest features that cut the adjacent plains, accumulated on the canyon floor and are rarely cut by faults. Most of the landslides are Amazonian but some may be as old as upper Hesperian (Quantin et al., 2004). The floors of Coprates and Ganges are continuous eastward with the Upper Hesperian-aged Tiu and Simud Valles. These data collectively suggest that canyons started opening in the Noachian and that faulting, subsidence of the floor, and erosion of the walls continued through the upper Hesperian, after which faulting and subsidence was minor and widening was largely restricted to landslides. Mounds of layered sediments are widespread within the canyons, at elevations that range from under –3000 m in Melas to over 3000 m in west Candor. Most are rich in hydrated, mainly Mg and Ca, sulfates (Bibring et al., 2006; Gendrin et al., 2005). For the last 30 years the favored origin for the sediments is that they were deposited in intracanyon lakes (Komatsu et al., 1993; Lucchitta et al., 1992; McCauley, et al.,1978; Nedell et al., 1987; Weitz and Parker, 2000; see Chapters 5 and 6). Such an origin is consistent with the eastward merger of the canyon floors with large outflow channels, the fine layering of the sediments, superposition relations across the Ophir–Candor divide, the marked contrast in erosional styles between the sediments and the canyon walls, and the presence of sulfates. The only plausible shoreline so far identified within the canyons, however, is one in Coprates Chasma at an elevation of roughly –3500 m (Harrison, 2007), 6500 m below the top of the sediments in Candor. The lake hypothesis does not necessarily imply deep lakes. The sediments, together with evaporitic minerals, may have accumulated over many millions of years by repeated episodes of evaporation and/or sublimation following injections of water into the canyons as a result of climatic events, faulting, or other causes. As indicated above, outflow channels commonly start at faults, so it is not unreasonable to conclude that the huge faults that created the canyons could have been conduits that supplied groundwater for lakes within the canyons. If climatic conditions were similar to today’s, the lakes would have frozen and been hindered from draining away by a thick cryosphere. Even if the canyons did at times contain lakes, the origin of the layered deposits still remains puzzling. One possibility is that they are a mixture of subaqueous and subaerial deposits, the materials having been brought in by the wind and deposited in water when lakes were present and subaerially when lakes were not present. Such an origin is

Acquisition and history of water on Mars

49

consistent with the common presence of mounds of sediments in upland craters, mounds that resemble those in the canyon (Carr, 2006, figure 2.7, 2.9). Other sugges­ tions are that the canyon internal deposits are young pyroclastic deposits (Hauber et al., 2006), or products of sub-ice or subaqueous volcanism (Chapman and Tanaka, 2001). Malin and Edgett (2000) and Catling et al. (2006) suggest a very different origin for the sediments. They argue that the sediments do not postdate the canyons but are instead simply remnants of the Noachian–Hesperian materials into which the canyons are cut. By this hypothesis, the layering, the contrasting erosional styles, and the superposition relations are inherited from the original pre-canyon materials. Lakes may still have been present at times but they did not result in the deposition of kilometer-thick stacks of sediments, and the apparent young age of the sediments is an exposure age and not a depositional age.

2.5.3 Oceans If the outflow channels were formed by floods, as is likely, then large bodies of water must have been left at their termini, which are mostly in the northern plains. Evidence for such bodies of water remains equivocal. Several possible shorelines have been tentatively identified in and around the northern plains (Clifford and Parker, 2001; Parker et al., 1989, 1993; see Chapters 9 and 10) and Hellas (Moore and Wilhelms 2001; see Chapter 7) but they remain controversial. Supporting evidence for the presence of former bodies of water of Hesperian age in the northern plains are partly buried ridges and craters, interpreted as the result of burial by sediments carried by the large floods (Head et al., 2002). The burying unit, part of the Vastitas Borealis Formation (Tanaka et al., 2005), covers an area of 1–2 � 107 km2 and has a minimum volume of 3 � 106 km3 (Kreslavsky and Head, 2002). The boundary of the unit is roughly coincident with the Deuteronilus shoreline identified by Parker and cow­ orkers (Chapter 9). Its enclosed volume is more than adequate to account for even the largest flood volumes estimated by Andrews-Hanna and Phillips (2007) and Leask et al. (2007). Also supporting the former presence of bodies of water in the northern plains are numerous features that suggest that stagnant ice sheets could have been left behind when the bodies of water froze (Kargel et al., 1995, see also Chapter 10). Most of the features (e.g., thumbprint terrain, polygonal ground) are found around the edge of the Vastitas Borealis formation (Tanaka et al., 2005). Arguing against the former presence of large bodies of water in the northern plains are the lack of detection of evaporites (Bibring et al., 2005) and the presence of large boulders up to 2 m in diameter in low areas where fine-grained sediments would be expected by the flood hypothesis (McEwen et al., 2007).

2.5.4 Erosion and weathering Both orbital and surface observations (summarized in Golombek et al., 2006) indicate that average erosion rates dropped 2–5 orders of magnitude at the end of the Noachian. The low rates were sustained for the rest of the planet’s history. The rates of 0.02–0.03 nm year–1 estimated for the uppermost Noachian or Hesperian lava

50

Lakes on Mars

plains of Gusev and the Pathfinder landing site are many orders of magnitude below the lowest rates (104–105 nm year–1) for the Earth. However, despite the extremely low average rates, extensive post-Noachian erosion has occurred locally, causing some post-Noachian units, such as the Medusae Fossae Formation and the polar layered deposits, to be deeply eroded. The higher rates appear to occur mainly as a result of local events such as floods, or where rock properties are such that wind and sublimation are effective removal agents. In addition, steep slopes, particularly in midlatitude craters, are commonly gullied (see Section 2.6). The widespread detection from orbit of olivine (Putzig et al., 2005) on post-Noachian surfaces indicates persistently low weathering rates throughout much of Mars’ history (Hoefen et al., 2003), olivine being a mineral particularly susceptible to breakdown under moist conditions. Low weathering rates are also implied by alteration of the basalts in Gusev. The basaltic flows on the floor of Gusev crater have a crater retention age of 3.6 Gyr (Greeley et al., 2005) and although individual boulders analyzed by the Spirit rover cannot be dated they are likely also to be billions of years old. The rocks have a thin alteration rind in which S, Cl, and Br are enhanced, but the primary minerals olivine, plagioclase, and magnetite are retained. Chemical patterns in the soils indicate migration of soluble elements, thereby implicating liquid water. However, the alteration rinds and soil patterns are likely to be mainly the result of interactions at low water/rock ratios such as that might result at low rates from acid clouds or local melting of frost under present or higher obliquity conditions (Haskin et al., 2005).

2.5.5 Sulfates Abundant sulfates have been observed in the soils at all the landing sites so far visited; many of the rocks in the Columbia Hills have been pervasively altered by sulfate-rich fluids. Sulfates are a major component of the sediments at Meridiani, and thick sulfate deposits have been detected from orbit at several locations mainly in the western hemisphere, but also around the north pole. The sulfate-rich deposits sampled by Opportunity in Meridiani are part of a unit roughly 600 km across and several hundred meters thick, which overlies typical Noachian cratered terrain. It appears etched in orbital images (Arvidson et al., 2003). The Mars Exploration Rover (MER) science team interprets the composition of the deposits analyzed by the rovers as the result of a mixture of roughly equal parts of a sulfate end member and altered basalt that has been depleted of roughly 50% of its original Fe, Mg, and Ca. Jarosite, the only sulfate mineral detected by the rovers, has the same sulfur content as the hypothesized sulfate end member. Kieserite has been detected elsewhere in the etched unit by CRISM (Wiseman et al., 2007) as have phyllosilicates (Poulet et al., 2008). The MER science team interprets the section at Meridiani to result largely from eolian deposition of sandsized grains of the two end members to form dunes and sand sheets. Sedimentary structures indicative of aqueous deposition in the upper part of the section in Endurance crater suggest ephemeral, inter-dune playas, which are interpreted as acid because jarosite precipitates under very acid conditions. Mineral casts and incrustations together with variations in Cl and Br in the section probably result from groundwater oscillations (McLennan et al., 2005).

Acquisition and history of water on Mars

51

The sulfates at Meridiani appear to have formed elsewhere and been transported to Meridiani, mainly by the wind. The same may be true for sulfate deposits that occur at other locations, particularly those in the north pole. For the bulk of the sulfate deposits at Meridiani, the MER science team favors a playa-like source, similar to that inferred for the upper part of the Endurance section, but no extensive sources have been identified. Because of the large volume of the etched Meridiani unit (roughly 105 km3), a playa source, fed either by surface runoff or by groundwater, would imply the processing of large amounts of water. Evaporation of waters from the large floods appears to be ruled out because of timing and failure to detect evaporites in the northern plains. Another possibility is that the sulfates do not form by evapora­ tion but are instead primary weathering products. By this scenario, acid fogs or other forms of acidic precipitation form easily erodible, sulfate-rich weathering rinds that are eroded by the wind and ultimately accumulate in eolian sedimentary deposits. By this mechanism discrete bodies of water are not required. Much of the discussion above on Meridiani applies equally well to other sulfate-rich deposits such as those in Valles Marineris. We saw above that faulting could have caused groundwater eruptions into the canyons, where the water could have been contained. Evaporation of successive groundwater eruptions could have led to the accumulation of the thick sequences of sulfates observed, or the sulfates could have been brought in by the wind. Jarosite has not been detected in the canyons so the case for acid conditions is weaker than that at Meridiani, as would be expected if the groundwater was buffered by a reaction with basalt. Nevertheless, the evaporative origin of the sulfate deposits in the canyons is not proven. We have compelling evidence of movement of sulfur-rich particles by the wind in Meridiani and around the north pole, and the Valles Marineris deposits could similarly have been deposited by the wind. While the precise ages of the sulfate-rich deposits are uncertain, most (although not the north polar deposits) are upper Noachian or Hesperian in age. Where the deposits occur in Noachian terrain, as in Meridiani, they are at the top of the section. Although phyllosilicates are detected in Noachian terrains where craters have ejected materials from deeper in the section, sulfates are not. There appears to be a transition from a mainly phyllosilicate-producing era in the middle and lower Noachian to a sulfateproducing era in the upper Noachian and Hesperian. Bibring et al. (2006) suggested that the transition was due to massive eruptions of sulfur that accompanied the formation of Tharsis, an origin that may be at odds with the conclusions of Phillips et al. (2001) that Tharsis was largely built by the end of the Noachian. Another possibility is that the enhanced sulfur activity is the result of the eruptions that formed the widespread Hesperian lava plains. Yet another possibility is simply that sulfaterich deposits become more visible in transitioning from the Noachian to the Hesperian because as the pace of processes such as impacts, volcanism, and fluvial erosion slows, the results of evaporation and eolian activity become more evident. Thus, the planet underwent a major change in transitioning from the Noachian to the Hesperian. Rates of impact and erosion declined dramatically. The rate of valley formation also steeply declined although not to zero. Surface conditions changed such that the rate of weathering to produce phyllosilicates declined but sulfate-rich deposits became more evident. In contrast the rate of formation of large floods increased,

52

Lakes on Mars

which likely resulted in the episodic and temporary presence of large bodies of water, particularly in the northern plains. The Valles Marineris, which appear to have largely formed by the end of the Hesperian, may have episodically contained lakes that drained to the east to form outflow channels. Many of the changes suggest a climate change at the end of the Noachian and start of the growth of a thick cryosphere, although the magnitude of the change and its cause remain unclear.

2.6 Amazonian era The Amazonian Period extends from roughly 3 billion years ago, the middle of the terrestrial Archean, to the present, encompassing two-thirds of the planet’s history. Despite the long time represented by the period, only a modest amount of geologic activity occurred, compared with earlier periods, and the extremely low erosion and weathering rates that typified the upper Hesperian continued (Golombek et al., 2006). Partly as a consequence of the low rates of terrain building, the effects of some surficial processes such as those involving ice and wind are more evident than those for earlier eras and perhaps the most distinguishing feature of the Amazonian is the abundant evidence for the action of ice, particularly at mid-to-high latitudes. Processes driven by obliquity variations are also more evident for this era although such processes likely occurred throughout all of Martian history.

2.6.1 Volcanism Volcanic activity in the Amazonian was largely in, and peripheral to, Tharsis and Elysium, where the large shields continued to grow and lava plains continued to accumulate. However the eruption rate appears to have declined significantly. The eruption volumes estimated by Greeley and Schneid (1991) and the chronology of Hartmann and Neukum (2001) suggest that average eruption rates dropped from roughly 1 km3 year–1 in the Hesperian to roughly 0.1 km3 year–1 in the Amazonian. Most of the Amazonian volcanic plains are distinctively different from the earlier Hesperian plains. The earlier plains (e.g., the Lunae, Solis, Chryse, Hesperia, Syrtis Major, Hellas Plana) typically have numerous wrinkle ridges but few primary flow structures. In contrast, most of the Amazonian plains have few wrinkle ridges but numerous primary volcanic structures such as flow fronts, lava channels, and lines of skylights at the crests of lava ridges. Crater ages of tens of millions of years for volcanic surfaces in Tharsis and Elysium (Neukum et al., 2004; Berman and Hartmann, 2002) and crystallization ages as young as 150 MY from Martian meteorites (McSween, 2002) suggest that Mars is still episodically active, although at very low rates.

2.6.2 Ice Ice likely played a significant role in modifying the landscape throughout much of Mars’ history but its effects are most evident for the Amazonian. The possibility that extensive ice deposits were left in low areas after large Hesperian floods was

Acquisition and history of water on Mars

53

mentioned above. In addition, there are indications of pervasive near-surface ice at mid-to-high latitudes, widespread, ice-rich veneers cover most the surface also at mid­ to-high latitudes, and glaciation may have occurred locally. Also, much of the ice presently at the poles appears to have accumulated late during the Amazonian. At mid-to-high latitudes ice is unstable at the surface because summer daytime temperatures rise above the frost point. However, daily temperature fluctuations damp out rapidly with depth, and modeling suggests that water ice is stable a few tens of centimeters below the surface, the depth depending on the latitude and the thermal inertia of the materials overlying the ice (Farmer and Doms, 1979). As expected, at latitudes higher than 60°, neutron and gamma-ray spectrometer measurements detected large fractions of ice at depths of tens of centimeters below a dehydrated layer (Feldman et al., 2004), and the presence of an ice table centimeters below the surface was confirmed by the Phoenix lander. Comparably large fractions of ice are not detected by orbiter spectrometers at latitudes much lower than 60° although geologic indicators of ground ice, such as debris aprons, are present down to latitudes as low as 30°. The observations suggest that significant amounts of near-surface ice may be present down to latitudes as low as 30°, but at depths too deep to be detected by the spectrometers. The stability of ice at the surface is sensitive to the obliquity cycle. During periods of high obliquity ice tends to be driven from the poles to be deposited at lower latitudes (Jakosky and Carr, 1985; Mellon and Jakosky, 1995). The reverse occurs at low obliquities. During the current epoch, the obliquity oscillates between 15° and 35° about a mean of 24°, but Laskar et al., (2004) estimate that the average obliquity over geologic time is 40° and that there is a 63% probability that the obliquity reached 60° in the last 1 Gyr. At the current obliquity, ground ice should not be present at latitudes lower than 40° latitudes. Indicators of ice at latitudes as low as 30° may indicate that the ground ice has equilibrated with the more common higher obliquity conditions. Most of the terrain in the 30–55° latitude belts is covered with a thin (�10 m) veneer of material that forms a smooth surface where still intact and finely pitted surfaces were partly removed (Mustard et al., 2001). Head et al. (2003) suggested that it is an ice–dust mixture deposited during a recent era of higher obliquities 0.4–2 Myr ago and that it is now in the process of being removed. Much thicker, possibly ice-rich deposits occur preferentially on pole-facing slopes at midlatitudes (Carr, 2001). If thick enough, such deposits could flow to form glaciers. They have been invoked as a source of water that cut the gullies that commonly occur on steep slopes, as discussed below (Christensen, 2003). Lobate debris aprons adjacent to most steep slopes in the 30°–55° bands in both hemispheres (Squyres, 1979) are compelling indicators of the presence of ice (Figure 2.7). They typically have convex-upward surfaces, are roughly 500 m thick adjacent to the slope at their origin, and extend about 20 km away from the slope. Their radar properties are identical to those of the polar layered terrains, strongly suggesting large fractions of ice (Plaut et al., 2008). Numerous surface textures indicate flow away from the slopes, with the aprons commonly wrapping around obstructions or converging on gaps in obstacles to the flow. Similar features are not found at latitudes less than 30° where talus normally simply accumulates on slopes at the angle of repose. Mangold et al. (2002) suggested on the basis of experimental

54

Lakes on Mars

Figure 2.7 Ice-rich debris flow at 40° N, 25° E. Material shed from the cliff at the top of the image has flowed away from the cliff and through a gap in a barrier to the flow. Ground-penetrating radar indicates that the material is predominantly ice. Craters on the surface in various states of preservation suggest that the flow is tens to hundreds of millions of years old (THEMIS).

work that, if the debris flows are mixtures of rock and ice they must contain at least 28% ice. Lucchitta (1984) proposed that the ice was shed from the slopes at the head of the debris flows, which implies that the ground ice is pervasive in the 30°–55° latitude belts to depths of tens to hundreds of meters, consistent with flow of the nearsurface materials to produce a general softening of the terrain (Squyres and Carr 1986). In contrast, Head et al. (2003, 2006) and Dickson et al. (2008) have empha­ sized the role of glaciation, suggesting that many of the features observed in these latitude belts could be the result of glaciation caused by precipitation of ice during periods of high obliquity (Figure 2.8). Counts of all craters, irrespective of preserva­ tion, indicate ages of several hundred million years, whereas counts of small fresh craters give ages of a few million years (Mangold et al., 2003). The counts indicate that the debris flows began forming at least several hundred million years ago and that the superimposed craters have been episodically or continually undergoing degrada­ tion by sublimation, shear, and other processes ever since. Degradation rates are such that small (<0.5 km) craters are preserved for millions of years. Glaciers may have formed outside the 30°–55° latitude belts. On the northwest flanks of Olympus Mon and other large Tharsis volcanoes, several features, including lobate flows and fan-shaped formations with finely striated margins, strongly suggest that former glaciers modified the volcanic surfaces and left extensive moraines on the adjacent plains (Head and Marchant, 2003; Lucchitta, 1981; Shean et al., 2005). A glacial origin is supported by modeling studies of the atmosphere, which indicate that

Acquisition and history of water on Mars

55

Figure 2.8 Possible glacier near Hellas at 38° S, 104° E. Material has flowed from an alcove in the upper right of the image, south through one crater into another larger crater. Flow over such distances requires substantial fractions of ice. One issue is whether the ice was derived by precipitation that accumulated in the alcove or whether the ice was shed from the massifs around the alcove (HRSC).

the northwest volcano flanks are preferred sites for precipitation of ice during periods of high obliquity (Forget et al., 2006). While the geologic evidence for large fractions of near-surface ice at high latitudes (>60°) is compelling, and confirmed by direct observations, numerous issues remain. Geomorphic indicators of flow, such as lobate debris aprons, lineated valley fill, and concentric crater fill, suggest large (>30%) fractions of near-surface ice may also be present under a dehydrated layer down to latitudes as low as 30°. The thickness of the ice-rich layer is however undetermined. It could fill bedrock pores to substantial depths (hundreds of meters to kilometers) or be restricted to the interstices of the uppermost fragmental materials. Also unclear is when the ice accumulated. Crater counts indicate that ice-abetted flow has been occurring for at least several hundred million years. Some of the ground ice could have accumulated as early as in the late Hesperian, a conse­ quence of the large floods, or even earlier, the result of the changes in surface conditions at the end of the Noachian. Alternatively it may have accumulated entirely during the Amazonian as a consequence of deposition during obliquity highs.

2.6.3 Fluvial activity Although the main era of outflow channel formation was over by the end of the Hesperian, a few younger outflow channels have been identified, and more will likely

56

Lakes on Mars

be discovered as high-resolution imaging accumulates. The most prominent examples of young outflow channels are the Athabasca, Grjota, Rahway, and Marte Valles in southeast Elysium. These have crater ages that range from 2 to 140 Myr, according to Burr et al. (2002), and some cut plains with crater ages of 10 Myr (Berman and Hartmann, 2002). All the young outflow channels start at fault-created fissures. If formed by water, they imply that in places liquid water is present at depth, below the cryosphere, and can be released to the surface by tectonic activity, even in the present epoch. They also imply the occasional presence of young lakes. Very few demonstrably Amazonian valley networks have been identified. Unu­ sually young valley networks occur in Melas Chasma and to the west of the south end of Echus Chasma (Mangold et al., 2004) and in the crater Lyot (Dickson et al., 2009). While the units they dissect are late Hesperian (2.9–3.4 billion years old), the valleys could be Amazonian. Similarly, some of the valleys on densely dissected volcanoes such as Ceraunius Tholus and Hecates Tholus could be Amazonian. However, the most prominent unambiguously Amazonian valley networks are on Alba Patera. The origin of these valleys is unclear. Some form hierarchical networks that resemble those formed by terrestrial drainage systems, but interspersed among such networks are channels that are clearly formed by lava, so that the role of precipitation in forming these valleys remains obscure. If formed by precipitation, then one possibility is that they formed by melting of ice deposits that accumulated during periods of high obliquity (Forget et al., 2006). Gullies are by far the most common fluvial-like features that formed in the Amazonian (Figure 2.9). They typically consist of an upper theater-shaped alcove that tapers downslope to converge on one or more channels that extend further downslope to terminate in a debris fan (Malin and Edgett, 2000). They are mostly

Figure 2.9 Gullies a few meters across in the south-facing wall of Newton crater at 41° S, 192° E. The gullies cut though several ledges and extend almost up to the crater rim (MOC).

Acquisition and history of water on Mars

57

meters to tens of meters wide, hundreds of meters long, and are common on steep slopes in the 30–60 latitude belts, particularly in the south. They have a slight preference for pole-facing slopes, at least at midlatitudes (Balme et al., 2006; Bridges and Lackner, 2006). Their origin is controversial. Although initially attributed to groundwater seeps, this origin now seems unlikely given the probable thick cryo­ sphere during most of the Amazonian and the common presence of gullies at locations where groundwater is unlikely, as on slopes around mesas and central peaks and at crater rim crests. Dry mass-wasting may contribute to their formation but this also seems to be an unlikely primary cause since many of the gullies cut through bedrock ledges. Erosion by wind or ice appears ruled out by their morphology, and erosion by liquid or gaseous CO2 appears ruled out by stability relations (Stewart and Nimmo 2002). All the morphologic attributes are consistent with water erosion, and the broad consensus is that that is their cause. In the southern highlands at midlatitudes, where most of the gullies occur, average daily summer temperatures are in the 220–230 K range and surface pressures are below the triple point of water. While small amounts of liquid water might temporarily exist today under such conditions, particularly in the presence of salts, accumulation of sufficient liquid to erode gullies is unlikely, and although newly formed light-toned slope streaks starting at gullies have been attributed to liquid water (Malin et al., 2006), spectral data and closer examination have failed to find evidence that the recent bright deposits were deposited by water (McEwen et al., 2007). They may simply be dust avalanches. A plausible possibility is that the gullies result from the temporary presence of water produced by the melting of snow and ice deposited at midlatitudes during periods of high obliquity (Christensen, 2003; Costard et al., 2002; Lee et al., 2001). Such an origin is supported by modeling studies (Costard et al., 2002) and by observations of gullies emerging from beneath what appear to be ice deposits on steep slopes (Christensen, 2003). The age of the gullies cannot be accurately determined but they probably have been forming episodically, when obliquities were high throughout the 3-billion-year length of the Amazonian, and possibly longer (Schon et al., 2009). They appear fresh because of the extremely low erosion rates, but are unlikely to have been forming continuously since there is little evidence that they have caused sig­ nificant backwearing of crater walls and filling of the craters despite the long times over which they probably have been forming. Thus, fluvial activity during the last 3 billion years of Mars’ history has been minor and restricted mainly to rare ground­ water eruptions, very rare valley network formation of unknown causes, and the gullying of steep slopes, probably from melting of ice during high obliquities.

2.6.4 Poles The finely layered deposits at the poles provide the most complete record of geolo­ gically recent events on the planet. The deposits in the north form a mound roughly centered on the pole and reaching up to 3 km above the surrounding plains of Vastitas Borealis. Crater counts indicate that the average age of the surface is of the order of 105 years (Herkenhoff and Plaut, 2000). The deposits can be divided into two distinct units: (i) a basal, platey, low-albedo unit, up to 1 km thick, that rests directly on the

58

Lakes on Mars

much older, extensive fill of the Borealis basin and (ii) the overlying, finely layered deposits that constitute the bulk of the 3-km-high mound (Byrne and Murray, 2002). The layered deposits extend out to roughly the 80° latitude and are surrounded by a vast dune field that is in places rich in gypsum (Langevin et al., 2005). Radar sounding shows that the layers in the upper unit form four distinct packets and individual layers can be traced large distances across the entire cap both in the radar returns (Phillips et al., 2008) and in the images (Milkovitch and Head, 2005). The southern deposits are more complicated. A 3-km-high central mound extending roughly 5° from the pole is partly surrounded by thinner, older deposits that extend several degrees further out, where a much older layered unit, the Dorsa Argentea formation, is exposed. Crater counts on the central mound indicate an age of the order of 107 years. Herkenhoff and Plaut (2000) attribute the difference in ages between the two caps to differences in the persistence of the residual CO2 cap at the two poles. If composed mostly of water ice, the total volume of the water in the layered deposits is roughly equivalent to a 20-m-deep global layer, far short of the volume of water needed to cut the Hesperian flood channels. The layering has long been attributed to accumulations of dust and ice modulated by orbital and rotational motions (Murray et al., 1972) and this is still the prevailing theory. Phillips et al. (2008) suggest that the weakest radar reflectors detected by SHARAD could contain as little as 2% dust, the rest being ice; the strongest reflectors could contain as much as 30% dust. Variations in obliquity would affect deposition and removal of ice at the poles and the incidence of dust storms and hence the deposition of dust (Toon et al., 1980). While attempts have been made to correlate specific layers with recent obliquity variations (Milkovich and Head, 2005), the correlations will remain uncertain until samples are available for dating. Nevertheless, the layering appears to reflect geologically recent events. The absence of an older record is consistent with the interpretation that many features at midlatitudes result from removal of ice at high latitudes and deposition at lower latitudes during periods of high obliquity. Accumulation and removal of layered deposits at the poles probably have been occurring repeatedly throughout the history of the planet. At the north pole we have only a recent record, but a partial record of older polar events may be preserved in the south.

2.7 Summary Mars accumulated and differentiated into crust, mantle, and core within a few tens of millions of years of Solar System formation. The global inventory of near-surface water available to participate in a geologic process is poorly constrained because of large uncertainties in the amount of water originally accreted and subsequently lost during the first 0.5 Gyr of the planet’s history. The Noachian period (4.1–3.7 Gyr ago) was characterized by the presence of a magnetic field, high rates of cratering, erosion, and valley formation. Most of Tharsis formed and surface conditions were at least episodic such as to cause the widespread production of hydrous weathering products such as phyllosilicates. Erosion rates, though high compared with later epochs, fell

Acquisition and history of water on Mars

59

short of the lowest terrestrial rates. Water-worn valley networks are common, but the best preserved valleys form an immature system that has had only a modest effect in shaping the landscape. The record suggests that warm, wet conditions necessary for fluvial activity were met only occasionally, such as might occur if caused by large impacts or volcanic eruptions. A major change occurred at the end of the Noachian. The rates of impact, valley formation, weathering, and erosion dropped precipitously. On the other hand, volcanism continued at a relatively high rate throughout the Hesperian, resulting in the resurfacing of at least 30% of the planet. Large floods formed episodically, possibly leaving behind large bodies of water. The canyons formed. The observations suggest the change at the end of the Noachian suppressed most aqueous activity at the surface other than large floods, and resulted in the growth of a thick cryosphere. However, the presence of discrete sulfate-rich deposits and sulfate concentrations in soils suggests that water activity did not decline to zero. After the end of the Hesperian around 3 Gyr ago the pace of geologic activity slowed further. The rate of volcanism during the Amazonian was roughly a factor of 10 lower than that in the Hesperian and confined largely to Tharsis and Elysium. The main era of flooding was over, although small floods appear to have occurred episodically until geologically recent times. Canyon development was largely restricted to the formation of large landslides. Erosion and weathering rates remained extremely low. The most distinctive characteristic of the Amazonian is the formation of features that have been attributed to the presence, accumulation, and movement of ice. Included are the polar layered deposits, glacial deposits on volcanoes, ice-rich veneers at high latitudes, and a variety of landforms in the 30–55o latitude belts, including lobate debris aprons, lineated valley fill, and concentric crater fill. Most of the gullies on steep slopes also formed during this era. The rate of formation of the ice-related features and possibly the gullies probably varied as changes in obliquity affected the ice stability relations.

References Acuna, M.H., Connerny, J.E., Ness, N.F., Lin, R.P., Mitchell, D., et al., 1999. Global distribu­ tion of crustal magnetization discovered by the Mars Global surveyor MAG/ER experi­ ment. Science 284, 790–793. Aharonson, O., Zuber, M.T., Rothman, D.H., 2001. Statistics of Mars’ topography from the Mars Orbiter Laser Altimeter: Slopes, correlations, and physical models. J. Geophys. Res., 106 (E10), 23723–23735 Anderson, F.S., Grimm, R.E., 1998. Rift processes at the Valles Marineris Mars: Constraints from gravity on necking and rate-dependent strength evolution. J. Geophys. Res. 103, 11113–11124. Andrews-Hanna, J.C., Phillips, R.J., 2007. Hydrological modeling of outflow channels and chaos regions. J. Geophys. Res. 112, E08001, doi:10.1029/2006JE002881. Andrews-Hanna, J.C., Zuber, M.T., Banerdt, W.B., 2008. Global structure of the Martian dichotomy: An elliptical impact basin. LPSC XXXIX, Abstract 1980 Arvidson, R.E., Seelos, F.P., Deal, K.S., Koeppen, W.C., Snider, N.O., Kieniewicz, J.M., et al., 2003. Mantled and exhumed terrains in Terra Meridiani, Mars. J. Geophys. Res. 108, E12, doi:10.1029/2002JE001982.

60

Lakes on Mars

Baker, V.R., 1982. The Channels of Mars. Texas University Press, Austin. Baker, V.R., 1990. Spring sapping and valley network development. Geol. Soc. America Sp. Paper 252, 235–265 Baker, L., Franchi, I.A., Wright, I.P., 2005. The origins of Martian water: What we can learn from meteorites. In: Tokano, T., (ed.), Water on Mars and Life, Springer, Berlin, pp. 3–24. Baker, V.R., Milton, D.J., 1974. Erosion by catastrophic floods on Mars and Earth. Icarus 23, 27–41. Baker, V.R., Partridge, J., 1986. Small Martian valleys: Pristine and degraded morphology. J. Geophys. Res. 91, 3561–3572. Balme, M., et al., 2006. Orientation and distribution of recent gullies in the southern hemisphere of Mars: Observations from High Resolution Stereo Camera/Mars Express (HRSC/MEX) and Mars Orbiter Camera/Mars Global Surveyor (MOC/MGS) data. J. Geophys. Res. 111, E05001, doi:10.1029/2005JE002607. Berman, D.C., Hartmann, W.K., 2000. Recent fluvial, volcanic and tectonic activity on the Cerberus plains of Mars. Icarus 159, 1–17. Bibring, J., et al., 2005. Mars surface diversity as revealed by the Mars Express OMEGA observations. Science 307, 1576–1581, Bibring, J., et al., 2006. Global mineralogical and aqueous history derived from OMEGA/Mars Express data. Science 312, 400–404. Blasius, K.R., et al., 1977. Geology of Valles Marineris: First analysis of imaging from the Viking 1 primary mission, J. Geophys. Res. 82, 4067–4091. Borg, L.E., Nyquist, L.E., Tylor, L.A., Wiesmann, H., Shih, C.-Y., 1997. The age of Dar al Gani 476 and the differentiation history of the Martian meteorites inferred from their radiogenic isotopic systematics. Geochim et Cosmochim Acta 67, 3519–3536. Bridges, N.T., Lackner, C.N., 2006. Northern hemisphere Martian gullies and mantled terrain: Implications for near-surface water migration in Mars’ recent past. J. Geophys. Res. 111, E09014, doi:10.1029/2006JE002702. Burr, D.M., McEwen, A.S., Sakimoto, S.E., 2002. Recent aqueous floods from the Cerberus Fossae. Mars. Geophys. Res. Lett. 29, 1, doi:10.1029/2001Gl013345. Byrne, S., Murray, B.C., 2002. North polar stratigraphy and the paleo-erg of Mars. J. Geophys. Res. 107, E6, doi:10.1029/2001JE001615. Cabrol, N.A., Grin, E.A., 1999. Distribution, classification and ages of Martian impact crater lakes. Icarus 142, 160–172. Cabrol, N.A., Grin, E.A., 2001. The evolution of lacustrine environments on Mars: (Is Mars only Hydrologically Dormant?) Icarus 149, 291–328. Cabrol, N.A., Grin, E.A., 2002. Overview on the Formation of Paleolakes and Ponds in Impact Craters on Mars. Glob. Planet. Changes 35, 199–219. Cabrol, N.A., Grin, E.A., 2005. Ancient and recent lakes on mars. In: Tokano, T. (Ed.), Water on Mars and Life, Springer, Berlin, pp. 235–259. Carr, M.H., 1979. Formation of Martian flood features by release of water from confined aquifers. J. Geophys. Res. 84, 2995–3007. Carr, M.H., 1986. Mars: A water-rich planet? Icarus 68, 187–216. Carr, M.H., 1992. Post-Noachian erosion rates: Implications for Mars climate change. LPSC. XXIII, 205–206. Carr, M.H., 2001. Mars Global Surveyor observations of Martian fretted terrain. J. Geophys. Res. 106, 23571–23594. Carr, M.H., 2006. The Surface of Mars. Cambridge University Press, Cambridge, 307p. Carr, M.H., Clow, G.D., 1981. Martian channels and valleys: their characteristics, distribution and age. Icarus 48, 91–117.

Acquisition and history of water on Mars

61

Carr, M.H., Head, J.W., 2002. Oceans on Mars: An assessment of the observational evidence and possible fate. J. Geophys. Res 108, E5, doi:10.1029/2002JE001963. Carr, M.H., Wänke, H., 1992, Earth and Mars: Water inventories as clues to accretional histories. Icarus 98, 61–71. Catling, D.C., et al., 2006. Light-toned layered deposits in Juventae chasma, Mars. Icarus 181, 26–51. Chapman, M.G., Tanaka, K.L., 2001. Interior trough deposits on Mars: Subice volcanoes? J. Geophys. Res 106, 10087–10100. Christensen, P.R., 2003. Formation of recent Martian gullies through melting of extensive water-rich snow deposits. Nature 422, 45–48. Chyba, C.F., 1990. Impact delivery and erosion of planetary oceans in the early inner Solar System. Nature 343, 129–133. Clifford, S.M., 1980. Chasma Boreale (850N, 00W): Remnant of a Martian jokulhlaup? Bull. Amer. Astron. Soc. 12, 678. Clifford, S.M., Parker, T.J., 2001. The evolution of the martian hydrosphere: Implications for the fate of a primordial ocean and the current state of the northern plains. Icarus 154, 40–79. Connerny, J.E., et al., 1999. Magnetic lineations in the ancient crust of Mars. Science 284, 794–798. Costard, F., Forget, F., Mangold, N., Peulvast, J.P., 2002. Formation of recent Martian debris flows by melting of near-surface ground ice at high obliquity. Science 295, 110–113. Craddock, R.A., Howard, A.D., 2002. The case for rainfall on a warm, wet early Mars. J. Geophys. Res. 107, (E11), 5111, doi:10.1029/2001JE001505. Craddock, R.A., Maxwell, T.A., 1993. Geomorphic evolution of the Martian highlands through ancient fluvial processes. J. Geophys. Res. 98, 3453–3468. Crisp, J.A., 1984. Rates of magma emplacement and volcanic output. J. Volc. Geotherm Res. 20, 177. Dickson et al., 2009. Young Valley Networks on Mars: Persistent Flow of Water in Lyot Crater, a Distinctive Amazonian Impact Basin Microenvironment. Dickson, J.L., Head, J.W., Marchant, D.R., 2008. Late Amazonian glaciation at the dichotomy boundary on Mars: Evidence for glacial thickness maxima and multiple glacial phases. Geology 36, 411–414, doi:10.1130/G24382A.1. Dreibus, G., Wänke, H., 1987. Volatiles on Earth and Mars: A comparison. Icarus 71, 225–240. Ehlmann, B.L., et al., 2008. Orbital identification of carbonate-bearing rocks on Mars. Science 322, 1828–1832. Farmer, C.B., Doms, P.E., 1979. Global and seasonal water vapor on Mars and implications for permafrost. J. Geophys. Res. 84, 2881–2888. Fassett, C.I., Head, J.W., 2005. Fluvial sedimentary deposits on Mars: Ancient deltas in a crater lake in the Nili Fossae region. Geophys. Res. Lett 32, 14, doi:10.1029/2005GL023456. Fasset., C.I., Head, J.W., 2008. Open-basin lakes on Mars: Implications of valley networks lakes for the nature of martian hydrology. LPSC XXXIX, abstract 1139. Feldman, W.C., et al., 2004. The global distribution of near surface hydrogen on Mars. J. Geophys. Res. 109, E9, doi:10.1029/2003JE02160. Fishbaugh, K.E., Head, J.W., 2002. Chasma Boreale, Mars: Topographic characterization from Mars Orbiter Laser Altimeter data and implications for mechanisms of formation. J. Geophys. Res. 107, E3, doi:10.1029/2001JE001351. Forget, F., Haberle, R.M., Montmessin, F., Levrard, B., Head, J.W., 2006. Formation of glaciers on Mars by atmospheric precipitation at high obliquity. Science 311, 368–371. Frey, H.V., 2003. Buried impact basins and the earliest history of mars. LPSC XXXV, abstract 3104.

62

Lakes on Mars

Gendrin, A., et al., 2005. Sulfates in Martian layered terrains: THE OMEGA Mars Express view. Science 307, 1587–1591. Golombek, M.P., et al., 2006. Erosion rates at the Mars Exploration Rover landing sites and long-term climate change on Mars. J. Gophys. Res. 111, E12, doi:10.1029/2006JE002754. Golombek, M.P., Bridges, N.T., 2000. Erosion rates on Mars and implications for climate change: Constraints from the Pathfinder landing site. J. Geophys. Res. 105 (E1), 1841–1853. Greeley, R., et al., 2005. Fluid lava flows in Gusev crater, Mars. J. Geophys. Res. 110, E05008, doi:10.1029/2005JE002401. Greeley, R., Guest, J.E., 1987, Geologic map of the eastern equatorial region of Mars. U.S. Geological Survey, Misc. Inv. Map I-1802–B Greeley, R., Schneid, B.D., 1991. Magma generation on Mars: Amounts, rates and comparisons with Earth, Moon and Venus. Science 254, 996–998. Grotzinger, J.P., Arvidson, R.E., Bell, J.F., Calvin, W., Clark, B.C., Fike, D.A., et al., 2005. Stratigraphy, sedimentology and depositional environment of the Burns Formation, Mer­ idiani Planum, Mars. Earth Planet. Sci. Lett. 240, 11–72. Gulick, V.C., 1998. Magmatic intrusions and a hydrothermal origin for fluvial valleys on Mars. J.Geophys. Res. 103, 19365–19387. Gulick, V.C., 2001. Origin of the valley networks on Mars: a hydrologic perspective. Geomor­ phology 37, 241–268. Haberle, R.M., 1998. Early climate models. J. Geophys. Res. 103 (E12), 28467–28479. Haberle, R.M., et al., 2008. The effect of ground ice on the Martian CO2 cycle. Planet. Space Sci. 56, 251–255. Hanna, J.C., Phillips, R.J., 2005. Hydrologic modeling of the Martian crust with application to the pressurization of aquifers. J. Geophys. Res. 110, E01004, doi:10.1029/2004JE002330. Harrison, K.P., 2007. Ponding in central Valles Marineris due to late-stage Martian outflow channel activity. LPSC XXXVIII, Abstract 1806. Harrison, K.P., Grimm, R.E., 2005a. Groundwater-controlled valley networks and the decline of surface runoff on early Mars. J. Geophys. Res. 110, E12, doi:10.1029/2005JE002455. Harrison, K.P., Grimm, R.E., 2005b. Tharsis recharge and the Martian outflow channels: Observations and recent modeling. LPSC XXXVI, Abstract 1211. Hartmann, W.K., Neukum, G., 2001. Cratering chronology and the evolution of Mars. In: Kallen­ bach R., et al. (eds.), Chronology and Evolution of Mars, Kluwer, Dordrecht, pp. 165–194. Haskins, L.A., et al. 2005. Water alteration of rocks and soils on Mars and the Spirit rover site in Gusev crater. Nature 436, 66–69. Hauber, E., Gwinner, K., Gendrin, A., Fueten, F., Slesky, R., Pelkey, S., et al. 2006. An integrated study of interior layered deposits in Hebes Chasma, Valles Marineris, Mars, using MGS, MO and MEX data, LPSC XXXVII, Abstract 2022 Hauck, S.A., Phillips, R.J., 2002. Thermal and crustal evolution of Mars. J. Geophys. Res. 107, E7, doi:10.1029/2001JE001801. Head, J.W., Heisinger, H., Ivanov, M.A., Kreslavsky, M.A., Pratt, S., Thomson, B.J., 1999. Possible ancient oceans on Mars: Evidence from Mars Orbiter Laser Altimeter data. Science 286, 2134–2137. Head, J.W., Kreslavsky, M.A., Pratt, S., 2002. Northern lowlands of mars: Evidence for wide­ spread volcanic flooding and tectonic deformation in the Hesperian period. J. Geophys. Res. 107, E2, doi:19.1029/2000JE001445. Head, J.W., Marchant, D.R., 2003. Cold based mountain glaciers on Mars: Western Arsia Mons. Geology 31, 641–644. Head, J.W., Mustard., J.F., Kreslavsky, M.A., Millikan, R.E., Marchant, D.R., 2003. Recent ice ages on Mars. Nature 426, 797–802.

Acquisition and history of water on Mars

63

Head, J.W., Nahm, A.L., Marchant, D.R., Neukum, G., 2006. Modification of the dichotomy boundary on Mars by Amazonian mid-latitude regional glaciation. Geophys. Res. Lett. 33, L08S03, doi:10.1029/2005GL024360. Head, J.W., Wilson, L., Mitchell, K.L., 2003. Generation of recent massive water floods at Cerberus Fossae, Mars by dike emplacement, cryosphere cracking, and confined aquifer groundwater release. Geophys. Res. Lett 30 (11), 1577, doi:10.1029/2003GL017135. Heisinger, H., Head, J.W., 2002. Topography and morphology of the Argyre basin, Mars: Implications for it geologic and hydrologic history. Planet. Space Sci. 50, 939–981. Herkenhoff, K.E., Plaut, J.L., 2000. Surface ages and resurfacing rates on the polar layered deposits on Mars. Icarus 144, 243–253. Hoefen, T.M., Clark, R.N., Bandfield, J.L., Smith, M.D., Pearl, J.C., Christensen, P.R., 2003. Discovery of olivine in the Nili Fossae region of Mars. Science 302, 627–630. Howard, A.D., et al. 2005. A sedimentary platform in Margaritifer Sinus, Meridiani Planum and Arabia? LPSC XXXVI, Abstract 1545. Howard, A.D., Moore, J.M., Irwin, R.P., 2005. An intense terminal epoch of widespread fluvial activity on Mars: 1. valley network incision and associated deposite. J. Geophy. Res. 2110, E12, doi:10.1029/2005JE002459. Hynek, B.M., Phillips, R.J., 2003. New data reveal mature, integrated drainage systems on mars indicative of past precipitation. Geology 31, 757–760. Irwin, R.P., Howard, A.D., 2002. Drainage basin evolution in Noachian Terra Cimmeria, Mars. J. Geophys. Res. 107, E7, doi:10.1029/2001JE001818. Irwin, R.P., Maxwell, T.A., Craddock, R.A., Leverington, D.W., 2002. A large paleolake basin at the head of Ma’adim Vallis, Mars. Science 296, 2209–2212. Ivanov, A.B., 2001. Mars/Moon cratering rate ratio estimates. In: Kallenbach R., et al. (eds.), Chronology and Evolution of Mars, Kluwer, Dordrecht, pp. 97–104. Jakosky, B.M., Carr, M.H., 1985. Possible precipitation of ice at low latitudes of Mars during periods of high obliquity. Nature 315, 559–561. Jakosky, B.M., Jones, J.H., 1997. The history of Martian volatiles. Rev. Geophys. 35, 1–16. Jerolmack, D.J., Mohrig, D., Zuber, M.T., Byrne, S., 2004. A minimum time for the formation of Holden northeast fan, Mars. Geophys. Res. Lett. 31, L21701, doi:10.1029/ 2004GL021326. Johnson, C.L., et al., 2000. Lithosphere loading by the northern polar cap on Mars. Icarus 144, 313–328. Judson, S., Ritter, D.F., 1964. Rates of regional denudation in the United States. J. Geophys. Res. 69, 3395–3401. Kargel, J.S., Baker, V.R., Beget, J.E., Lockwood, J.E., Pewe, T.L., Shaw, J.S., et al., 1995. Evidence for ancient continental glaciation in the Martian northern plains. J. Geophys. Res. 100, 5351–5368. Kasting, J.F., 1991. CO2 condensation and the climate of early Mars. Icarus 94, 1–13. Komatsu, G., Geissler, P.E., Strom, R.G., Singer, R.B., 1993. Stratigraphy and erosional landforms of layered deposits in Valles Marineris. J. Geophys. Res. 98, 11105–11121. Kreslavsky, M.A., Head, J.W., 2002. Fate of outflow channel effluents in the northern lowlands of Mars: The Vastitas Borealis Formation as a sublimation residue from frozen, ponded bodies of water. J. Geophys. Res. 107, E12, doi:10.1029/2001JE001831. Lammer, H., et al., 2005. Atmospheric evolution and history of water on Mars. In: Tokano, T. ed., Water on Mars and Life, Springer, Berlin, pp. 25–43. Langevin, Y., et al., 2005. Sulfates in the north polar region of Mars detected by OMEGA/Mars Express. Science 307, 1584–1586.

64

Lakes on Mars

Laskar, J., et al., 2004. Long term evolution and chaotic diffusion of the insolation quantities of Mars. Icarus 170, 343–364. Leask, H.J., Wilson, L., Mitchell, K.L.K., 2007. Formation of Mangala outflow channel, Mars: Morphological development and water discharge and duration estimates. J. Geophys. Res. 112, E08003, doi:10.1029/2006JE002851. Lee, P., Cockell, C.S., Marinova, M.M., McKay, C.P., Rice, J.W., 2001. Snow and ice melt flow features on Devon Island, Nunavut, arctic Canada as possible analogs for recent slope flow features on Mars. LPSC XXXII, 1809. Lee, D.–C., Halliday, A.N., 1997. Core formation on Mars and differentiated asteroids. Nature 388, 854–857. Leverington, D.W., 2004. Volcalnic rilles, streamlined islands, and the origin of outflow channels on Mars. J. Geophys. Res. 109, E10, doi:10.1029/2004JE002311. Lucchitta, B.K., 1981. Mars and Earth: comparison of cold climate features. Icarus 45, 264–303. Lucchitta, B.K., 1984. Ice and debris in fretted terrrains. J. Geophys. Res. 89, B409–B418. Lucchitta, B.K., McEwen, A.S., Clow, G.D., Geissler, P.E., Singer, R.B., Schultz, R.A., et al., 1992. The canyon system on Mars. In: Kieffer, H.H. et al. (Eds.), Mars, University of Arizona Press, Tucson, pp. pp. 453–492. Lunine, L.I., Chambers, J., Morbidelli, A., Leshin, L.A., 2003. Origin of Martian water. Icarus 165, 1–8. Malin, M.C. et al., 2006. Present-Day Impact Cratering Rate and Contemporary Gully Activity on Mars. Science. 314, 1573–1577. Malin, M.C., Edgett, K.S., 2000. Sedimentary rocks of ealy Mars. Science 290, 1927–1937. Malin, M.C., Edgett, K.S., 2001. Mars Global Surveyor Mars Orbiter Camera: Interplanetary cruise through primary mission. J. Geophys. Res. 106, 23429–23570. Manga, M., 2004. Martian floods at Cerberus Fossae can be produced by groundwater dis­ charge. Geophys. Res. Lett. 31, L02702, doi:10.1029/2003GL018958. Mangold, N., 2003. Geomorphic analysis of lobate debris aprons on Mars at Mars Orbiter Camera scale: Evidence for ice sublimation initiated by fractures. J. Geophys. Res. 108, doi:10.1029/2002JE001885. Mangold, N., Allemand, P., Thomas, P., Duval, P., Geraud, Y., 2002. Experimental and theoretical deformation of ice–rock mixtures: Implications on rheology and ice content of Martian permafrost. Planet. Space. Sci. 50, 385–401. Mangold, N., Quantin, C., Ansan, V., Delacourt, C., Allemand, P., 2004. Evidence for precipitation on Mars from dendritic valleys in the Valles Marineris area. Science 305, 78–81. Matsui, T., Abe, Y., 1987. Evolutionary tracks of the terrestrial planets. Earth Moon Planets 39, 207–214. McEwen, A.S., et al., 2005. The rayed crater Zunil and interpretations of small impact craters on Mars. Icarus 176, 351–381. Mc Ewen, A.S., et al., 2007. A closer look at water-related activity on Mars. Science 317, 1706–1709. McCauley, J.F., 1978. Geologic map of the Coprates quadrangle of Mars. USGS Misc. Inv. Map I-897 McGill, G.E., Squyres, S.W., 1991. Origin of the Martian crustal dichotomy: Evaluating hypotheses. Icarus 386, 393. McGovern, P.J., Solomon, S.C., Smith, D.E., Zuber, M.T., Simons, M., et al., 2002. Correction to Localized gravity/topography admittances and correlation spectra on Mars: implications for regional and global evolution. J. Geophys. Res. 109, E07007. McKenzie, D., Nimmo, F., 1999. The generation of Martian floods by the melting of ground ice above dikes. Nature 397, 231–233.

Acquisition and history of water on Mars

65

McLennan, S.M., Bell, J.F., Calvin, W.M., Christensen, P.R., Clark, B.C., deSouza, P.A., et al., 2005. Evidence for groundwater involvement in the provenence and diagenesis of the evaporite-bearing Burns formation, Meridiani Planum. Earth Planet. Sci. Lett. 240, 95–121. McSween, H.Y., 2002. The rocks of Mars, from far and near. Meteorit. Planet. Sci. 37, 7–25. Mellon, M.T., Jakosky, B.M., 1995. The distribution and behavior of Martian ground ice during past and present epochs. J. Geophys. Res. 100, 11781–11799. Melosh, H.J., Vickery, A.M., 1989. Impact erosion of the primordial atmosphere of Mars. Nature 338, 487–489. Milkovich, S.M., Head, J.W., 2005. North polar cap of Mars: Polar layered deposit character­ ization and identification of a fundamental climate signal. J. Geophys. Res. 110, E05, doi:10.1029/2004JE002349. Moore, J.M., Howard, A.D., Dietrich, W.E., Schenk, P.M., 2003. Martian layered fluvial deposits: Implications for Noachian climate scenarios. Geophys. Res. Lett. 30 (24), 2292, doi:10.1029/2003GL019002. Moore, J.M., Wilhelms, D.E., 2001. Hellas as a possible site of ancient ice-covered lakes on Mars. Icarus 154, 258–276. Murchie, S., et al., 2008. First results from the Compact Reconnaissance Imaging Spectrometer for Mars (CRISM), LPSC XXXIX abstract 1472 Murray, B.C., Soderblom, L.A., Cutts, J.A., Sharp, R.A., Milton, D., 1972. A geologic frame­ work for the south polar region of Mars. Icarus 17, 328. Mustard, J.F., Cooper, C.D., Rifkin, M.K., 2001. Evidence for recent climate change on Mars from the identification of youthful near-surface ground ice. Nature 412, 4211–4414. Mège, D., Masson, P., 1996. Amounts of crustal stretching in Valles Marineris. Mars Planet. Space Sci. 44, 749–782. Nedell, S.S., Squyres, S.W., Anderson, D.W., 1987. Origin and evolution of the layered deposits in the Valles Marineris, Mars. Icarus 70, 409–441. Neukum, G., Jaumannn, R., Hoffmann, H., Hauber, E., Head, J.W., Basilevsky, A.T., et al., 2004. Recent and episodic volcanic and glacial activity on Mars revealed by the High Resolution Stereo Camera. Nature 432, 971–979. Neumann, G.A., Zuber, M.T., Wieczorek, M.A., McGovern, P.J., Lemoine, F.G., Smith, D.E., 2004. Crustal structure of Mars from gravity and topography. J. Geophys. Res. 109, E8, doi:10.1029/2004JE002262. Nimmo, F., 2000. Dike intrusions as a possible cause linear martian magnetic anomalies. Geology 28, 391–394. Nimmo, F., Tanaka, K., 2005. Early crustal evolution of Mars. Ann. Rev. Earth Planet. Sci. 33, 133–161. Osterloo, M.M., et al., 2008. Chloride bearing materials in the Souther Highlands of Mars. Science 319, 1651–1654. Owen, T.C., Bar-Nun, A., 2000. Volatile contributions from icy planetesimals. In: Canup, R.M. and Righter, K. Eds., Origin of the Earth and Moon, Univ. Arizona Press, Tucson, pp. 459–471. Parker, T.J., Clifford, S.M., Banerdt, W.B., 2000. Argyre Planitia and the Mars global hydro­ logic cycle. LPSC XXI, 2033. Parker, T.J., Gorsline, D.S., Saunders, R.S., Pieri, D., Schneeberger, D.M., 1993. Coastal geomorphology of the Martian northern plains. J. Geophys. Res. 98, 11061–11078. Parker, T.J., Saunders, R.S., Schneeberger, D.M., 1989. Transitional morphology in the west Deuteronilus Mensae region of Mars: Implications for modification of the lowland/upland boundary. Icarus 82, 111–145. Pepin, R.O.1994. Evolution of the Martian atmosphere. Icarus 111, 289–304.

66

Lakes on Mars

Phillips, R.J., et al., 2008. Mars north polar deposits: Stratigraphy, age, and geodynamical response. Science 320, 1182–1185. Phillips, R.J., Zuber, M.T., Solomon, S.C., Golombek, M.P., Jakosky, B.M., Banerdt, W.B., et al., 2001. Ancient geodynamics and global-scale hydrology on Mars. Science 291, 2587–2591. Pieri, D.C., 1980. Geomorphology of Martian valleys. NASA Tech. Memo. 81979, 1–160. Plaut, J.J., et al., 2008. Radar evidence for ice in lobate debris aprons in the mid-northern latitudes, LPSC XXXIX, Abstract 2290 Poulet, F., et al., 2007. Martian surface mineralogy from Observatoire pour la Mineralogie l’eau, les glaces and l’activite on board the Mars Express spacecraft (OMEGA/MEX): Global mineral maps. J. Geophys. Res. 112, E08S02, doi:10.1029/2006JE002840. Poulet, F., et al., 2007. Can the formation models of the Meridiani Planum outcrops be applied to the entire etched terrains of Terra Meridiani, 7th In. Mars Conference, abstract 3184. Poulet, F., et al., 2008. Phyllosilicates on Mars and implications for early Martian climates. Nature 438, 623–627. Putzig, N.E., Mellon, M.T., Kretke, K.A., Arvidson, R.E., 2005. Global thermal inertia and surface properties of Mars from the MGS mapping mission. Icarus 173, 325–341. Quantin, C., Allemand, P., Mangold, N., Delacourt, C., 2004. Ages of Valles Marineris (Mars) landslides and implications for canyon History. Icarus 172, 555–572. Robinson, M.S., Tanaka, K.L., 1990. Magnitude of a catastrophic flood event at Kasei Vallis, Mars. Geology 18, 902–905. Schon, S.C., Head, J.W., Fasset, C.I., 2009. Unique chronostratigraphic marker in depositional fan stratigraphy on Mars: Evidence for ca. 1.25 Ma gully activity and surficial meltwater origin. Geology 37, 207–210. Scott, D.H., Carr, M.H., 1978. Geologic Map of Mars. U.S. Geological Survey Misc. Inv. Map I-1083 Scott, D.H., Tanaka, K.L., 1986. Geologic map of the western equatorial region of Mars. USGS. Misc. Map I-1802–A Segura, T.L., Toon, O.B., Colaprete, A., Zahnle, K., 2002. Environmental effects of large impacts. Science 298, 1977–1980. Sharp, R.P., 1973. Mars troughed terrains. J. Geophys. Res. 78, 4063–4072. Shean, D.E., Head, J.W., Marchant, D.R., 2005. Origin and evolution of cold-based tropical mountain glacier on mars: The Pavonis Mons fan-shaped deposit. J. Geophys. Res. 110 (E5), 10, doi:1029/2004JR002360. Sleep, N.H., Zahnle, K., 1998. Refugia from asteroid impact on early Mars and the early Earth. J. Geophys. Res. 103 (E12), 28529–28544. Solomon, S.C., Ahoronson, O., Aurnou, J.M., Banerdt, W.B., Carr, M.H., Dombard, A.J., et al., 2005. New perspectives on ancient Mars. Science 307, 1214–1220. Squyres, S.W., 1979. The distribution of lobate debris aprons and similar flows on Mars. J. Geophys. Res. 84, 8087–8096. Squyres, S.W., et al., 2006. Rocks of the Columbia Hills. J. Geophys. Res. 111, E02, doi:1029/ 2005JE002562. Squyres, S.W., et al., 2006. Bedrock formation at Meridiani Planum. Nature 443 (7), E1–E2. Squyres, S.W., Carr, M.H., 1986. Geomorphic evidence for the distribution of ground ice on Mars. Science 231, 249–252. Stepinski, T.F., O’Hara, W.J., 2003. Vertical analysis of Martian drainage basins, LPSC XXXIV, abstract 166 Stewart, S.T., Nimmo, F., 2002. Surface runoff features on Mars: Testing of the carbon dioxide hypothesis. J. Geophys. Res. 107, E9, doi:10.1029/2000JE001465. Strom, R.G., Croft, S.K., Barlow, N.G., 1992. The Martian impact crater record. In: Kieffer, H. H., et al., (eds.), Mars. University of Arizona Press, Tucson.

Acquisition and history of water on Mars

67

Stőffler, D., Ryder, G., Ivanov, B.A., Artemieva, N.A., Cintala, M.J., Grieve, R.A., 2006. Cratering history and lunar chronology. Rev. Min. Geochem. 60, 519–596. Tanaka, K.L., 1986. The stratigraphy of Mars. Proc. 17th Lunar and Planet. Sci. Conf. J. Geophys. Res. 91, E139–E158. Tanaka, K.L., Skinner, J.A., Hare, T.M., 2005. Geologic map f the northern plains of Mars, U.S. Geological Survey Scientific Investigations Map 2888. Tera, F., Papanastassiou, D.A., Wasserburg, G.J., 1974. Isotopic evidence for a terminal lunar cataclysm. Earth Planet Sci. Lett. 22, 1–21. Toon, O.B., Pollack, J.B., Ward, W., Burns, J.A., Bilski, K., 1980.The astronomical theory of climate change on Mars. Icarus 44, 552–607. Weitz, C.M., Parker, T.J., 2000. New evidence that the Valles Marineris interior deposits formed in standing bodies of water, LPSC XXXI, Abstract 1693 Wilhelms, D.E., Squyres, S.W., 1984. The Martian hemispheric dichotomy may be due to a giant impact. Nature 309, 138–140. Wise, D.U., Golombek, M.P., McGill, G.E., 1979. Tectonic evolution of Mars. J. Geophys. Res. 84, 7934–7939. Wiseman, S.M., et al., 2007. Initial analyses of CRISM data over Meridiani Planum, LPSC XXXVIII, abstract 1945 Zahnle, K.J., Kasting, J.F., Pollack, J.B., 1988. Evolution of a steam atmosphere during Earth’s accretion. Icarus 74, 62–97. Zhong, S., Zuber, M.T., 2001. Degree-1 mantle convection and the crustal dichotomy on Mars. Earth Planet. Sci. Lett. 189, 75–84. Zolotov, M.Yu., Mironenko, M.V., 2008. Formation and fate of phyllosilicates on the surface of Mars: Geochemical modeling of aqueous weathering, LPSC XXXIX, Abstract 3365. Zuber, M.T., Solomon, S.C., Phillips, R.J., Smith, D.E., Tyler, L., Aharonson, O., et al., 2000. Internal structure and early thermal evolution of Mars from Mars Global Surveyor topo­ graphy and gravity. Science 287, 1788–1792.

3 Hydrologic provinces of Mars:

physiographic controls on drainage and ponding Rene A. De Hon Department of Geography, Texas State University, San Marcos, TX, USA

3.1

Introduction

This chapter examines the physiographic controls of planet-wide drainage and storage of water on the surface as lakes and ponds. Hydrogeomorphologic provinces are described as regional watersheds defined by topographic divides and terminal basins or low-elevation outlets. Most drainage systems on Mars appear relatively immature, although this view has been somewhat revised with high-resolution imagery. How­ ever, sufficient drainage and sedimentation developed to allow some originally closed basins to become integrated into open, through-flowing, drainage systems (Fassett and Head, 2008). Sparse dendritic drainage patterns viewed on Mariner 9 and Viking Orbiter images provided some evidence of possible rainfall on the surface of Mars, but many investiga­ tors of that time preferred groundwater sources for most of the channels (see Chapters 1 and 2 and references therein). More recently, MGS images provided a vast improvement in resolution and showed a highly dissected surface of more mature drainage patterns supporting the possibility of persistent rainfall (Craddock and Howard, 2002; Hynek and Phillips, 2003; Irwin et al., 2005). The probability of periods with a warmer, wetter Mars is now viewed as more likely (Pollack et al., 1987; Baker et al., 2000; Kargel, 2004). Baker (2009) envisions periodic, short climatic episodes of global warming brought about by volcanic release of stored carbon dioxide, methane, and water into the atmosphere. Vast catastrophic outflows would fill a northern ocean and support a brief period of wet Mars with a hydraulic cycle including the growth of glaciers and rainfall. As the possibility of a wetter Mars grows, so does the possibility of extensive glacial action. Abundant water stored as ice in a southern polar cap and at high elevations in the highlands provides an important source for melt water runoff (Baker et al., 1991; Head and Pratt, 2001; Head et al., 2003). As another possible mechanism for significant water at the surface, periodic climate change is proposed as a consequence of periodic shifts of Mars obliquity (Forget et al., 2006; Head et al., 2003). Polar obliquity ranging from 14° to 48° (Laskar and Robutel, 1993; Touma and Wilson, 1993) allowed glaciers to form at low latitudes. Lineated valley fill in fretted valleys, ridged surficial deposits, and alcoves Lakes on Mars. DOI: 10.1016/B978-0-444-52854-4.00003-9 © 2010 Elsevier B.V. All rights reserved.

70

Lakes on Mars

carved into the dichotomy boundary are cited as evidence of Amazonian mid-latitude regional glaciation (Dickson et al., 2008; Head et al., 2005; see also Chapter 2 and references therein). A warmer climate than present, possibly a northern ocean (see Chapters 8 and 9), glaciers in the highlands, and a southern polar ice cap set the stage for a hydrologic cycle (Baker et al., 2000). Evaporation from a Martian ocean provides atmospheric moisture for rain or snowfall in the highlands and supplies water to the southern ice cap. Rainfall and melt water runoff recharge the groundwater system and provide water to channel systems. Innumerable closed basins exist on the surface of Mars that would hold water as surface lakes and flooding of the northern lowlands would create an ocean (Chapter 9). Lakes are ephemeral in nature. Some dry out between occasional, episodic flood­ ing. Others are perennial—receiving inflow and maintaining a standing body of water throughout their lifetime. All lakes eventually cease to exist as their source is depleted; their catchment basin is filled with sediment, or through-flowing drainage is estab­ lished. Whether large or small, they require a source of water (see Chapter 1). Lakes are formed and maintained by the inflow of water into a closed basin. The source of water may be from channel inflow, direct precipitation, or groundwater discharge. Water in channels may originate by surface runoff of precipitation, runoff from snow pack or glacial melt, groundwater discharge, or spillage from a higher lake. Water­ sheds may include a multitude of sources that feed lakes at various intermediate levels as well as the lowest level of the terminal basin. A groundwater source, either by groundwater effluence or thawing of permafrost, is evidenced by large outflow channels associated by chaotic terrain. The chaos is created by the withdrawal of large amounts of groundwater and the subsequent collapse of the overlying terrain (Carr, 1979). Juvenate Chasma at the head of Maja Vallis (De Hon and Pani, 1992) and Echus Chasma at the head of Kasei Vallis (Tanaka and Chapman, 1992) are prime examples of channels that begin in chaos or collapsed terrain. Other large, well-formed channels such as Hrad Vallis, Granicus Valles, and Hebrus Valles are associated with fissures in volcanic terrains and may have been cut by water flow over prolonged periods of time (De Hon, 1992b; Mouginis-Mark, 1985). Some sinuous channels issuing from theater-headed valleys and characterized by short stubby tributaries such as Nirgal Vallis and Nanedi Vallis are probably formed by groundwater sapping (Baker, 1990; Laity and Malin, 1985). Compaction by the weight of overlying sediment in depositional centers or empla­ cement of volcanic material on saturated regolith may be expected to produce a pressure drive to force connate water to the surface as seeps and springs around the edge of sedimentary basins. Amazonian and Hesperian volcanoes could have supplied thermal energy for hot springs, geysers, and catastrophic discharge long after eruption (Farmer, 1996; Gulick, 1998; Schulze-Makuch et al., 2007). Ancient spring deposits may exist at scales too small to be resolved by available remote imaging. Large, flat-floored basins at the head of some large channel systems provide a ready source of catastrophic discharge by a rapidly developing spillway (De Hon and Pani, 1992). Although it was proposed that Mangala Valles were formed by a catastrophic groundwater discharge associated with Mangala Fossa (Zimbleman

Hydrologic provinces of Mars: physiographic controls on drainage and ponding

71

et al., 1992), Parker and Gorsline (1991b) and De Hon (1994) proposed that Mangala Valles were fed by a large surface impoundment south of the rill. Irwin et al. (2002) proposed that Ma’adim Vallis and Gusev crater were fed by a large paleolake. Water in the source basin may have been derived from rainfall runoff, or snow melt in the higher topography around the basin.

3.2

Physiographic control

The current Martian watershed configuration is the expression left by the valley networks that last survived the heavy bombardment and thus may only poorly reflect original conditions. Whether numerous fluvial features represent abundant rainfall on a warmer, wetter Mars than current conditions or whether streams were formed in spite of a cold, dry Mars in response to groundwater discharge, water was present. It was an important agent in shaping the surface of Mars through erosion and deposition. Once on the surface, movement of water was controlled by gravity and topography that controls the water collection area, or watershed. The watershed is the area that will supply water to a network of streams and neighboring watersheds are separated by physiographic divides. On a primitive surface, unmodified by previous flow, all drainage is consequent. Flow across the surface is controlled by initial topography, and water ponds in local surface irregularities. Initial channels are irregular, branching, and recombining in anastomosing patterns, and the formation of a network is subject to stream capture by mature channel alleys. Local ponding occurs until flowing water establishes subse­ quent channels that confine and direct any later flow in a more efficient drainage system. Water eventually collects in low-elevation, terminal basins. Scott et al. (1991, 1995) documented 15 possible lakes greater than 100,000 km2 surface area and depths greater than 1000 m. Smaller paleolakes or ponds abound along channels and down slope from channel and valley networks (Cabrol and Grin, 1999, 2001, 2002, 2005; De Hon, 1988). A great portion of the Mars surface slopes from a high-standing south polar region toward the northern lowland which is the ultimate trap for water and sediments moved by gravity. The constant-scale natural boundary map of Clark (2005) is constructed with the watershed divides as the outer edge of the map (Figure 3.1). This map emphasizes the planet-wide surface slopes toward a few terminal basins. All drainage is from the high elevations on periphery of the map toward the lowest elevations in Argyre and Hellas basins and the northern plains. Thus, Mars can be divided into three principle physiographic regions: the Northern Lowland Plains, the Central Highlands, and the Southern Plateau. The Northern Lowland Plains are essentially the northern third of the planet, north of the highland–lowland boundary scarp. Most of this region is below the zero elevation contour and is sparsely cratered. Relief is low within the northern lowlands except for the volcanic edifices which rise to elevations in excess of 20 km above planetary datum. The Central Highlands range from 0 to about 9 km elevation and consists mostly of cratered terrain and intervening plains. Broad

72

Lakes on Mars

Figure 3.1 Watershed map of Mars with constant-scale natural boundaries (Clark, 2005). The Mars globe is unzipped along major drainage divides. Divides form the outer edge of the map. Drainage is from edges of map to low regions of the map interior. The Hellas, Argyre, and Borealis basin floors stand out as terminal ponding for all Martian drainage.

troughs, canyons, large channels, and smaller integrated drainage systems are scat­ tered throughout the highlands. Flat-floored basins and breached craters connected by channels abound. The Southern Plateau region is composed of a moderate to lightly cratered, moderate elevation plateau surrounding and including the South Pole. Regional-scale topographic troughs are found in most of the provinces that contain significant highland terrains (De Hon, 1996). These troughs are long, shallow, broad topographic depressions of regional proportions that originate in highland terrains and extend to lower elevations, often terminating in lowland plains. Some troughs such as lower Mangalas Valles and Valles Marineris are structurally controlled. Others, such as upper Maja Valles, are the result of intersections of slopes of differing origins and orientations (De Hon, 1987). Dohm et al. (2001) identify a system of gigantic valleys northwest of Olympus Mons that may have fed flood discharge from the highlands into a northern ocean. In the highlands, drainage pathways are formed by coalescence of adjacent flat-floored topographic basins. The axes of some troughs are marked by channels and chaotic materials. Most troughs are floored by plains-forming material. Others are characterized by a patchy distribution of plains materials interspersed with furrowed highland terrains. Many highland troughs have served as physiographic controls to water that reached the surface as either precipitation or groundwater discharge. Although not carved by running water, troughs and their surrounding watersheds are natural funnels to direct surface runoff to lower elevations.

Hydrologic provinces of Mars: physiographic controls on drainage and ponding

73

Physiography controls the distribution, movement, and storage of water on the surface during times of high water tables (Coates, 1990). The location of groundwater discharge as seepage or catastrophic outflow provides some insight into groundwater conditions at the time of their formation. During the present period of extreme aridity, physiography may be only a minor guide to the position of the water table and to the movement of groundwater. Surface water infiltration occurs along channel floors and in areas of impoundment. Groundwater discharge occurs in head basins of outflow channels, chaotic terrain, polygonally grooved terrain, and perhaps along the walls of sinuous channels. Geologic controls that may influence the movement of groundwater include variations in permeability of the subsurface rocks; faults which provide paths for groundwater movement; and hydrothermal systems associated with intrusive magmas and volcanoes (Brackenridge et al., 1985), and with impact craters (Chapter 4) which create thermal drives and volatile sources. In addition, loading at sedimentary and volcanic depositional centers may create pressure drives that force fluids out of zones of high sediment or volcanic loading toward the surface. Features that may indicate past saturation of materials by water or ice include fluidized ejecta craters (Carr et al., 1977; Gault and Greeley, 1978; Johanson, 1979; Mouginis-Mark, 1979; Squyres, 1979), thumbprint terrain (Grizzaffi and Schultz, 1989), suspected pingos (Rossbacher and Judson, 1981), and polygonally fractured plains (McGill and Hills, 1992).

3.3

Hydrologic provinces

In order to effectively treat the occurrence of water on Mars, the surface is divided into eight hydrologic provinces (Table 3.1 and Figure 3.2). Each province is a regional watershed defined by an encompassing drainage divide and feeding to a low-elevation terminal basin. Nested and adjacent smaller watersheds occur within the larger regional provinces. Province boundaries are drawn on the crest of topographic divides as determined by MOLA topography (USGS, 2003). Hydrologic provinces are delineated by con­ tinuous, topographic divides that define regional-scale, topographic basins of internal Table 3.1 Summary of hydrologic provinces Collection area

Terminal basin

Basin center

Overflow

Southern Highland

Australe-Parva Planum Hellas Planitia Argyre Planitia Isidis Planita Aeolis Planum Amazonis Planitia Chryse Planitia Vastitas Borealis

65°S, 65°E 5°S, 200°E 50°S, 320°E 10°N, 155°E 5°N, 90°E 55°N, 180°E 50°N, 330° E 70°N, 330°E

Argyre or Aeolis Isidis (Unlikely) Chryse Borealis Borealis Borealis Borealis Lowest level

Central Highland

Northern Lowland

74

Lakes on Mars

0° 0°

° 60

24

0°

0°

0°

0°

12

12

30

24

° 60

30 0°

180°

0°

180°

km 8

60°

4

30° 0°

0

−30° −4 −60° −8 180°

240°

300°

0°

60°

120°

180°

Figure 3.2 Hydrogeologic provinces of Mars are delineated by continuous, topographic divides that define regional-scale, topographic basins of internal drainage. Most basins are floored by a low-elevation plain on which water or sediments originating at higher elevations accumulate. Most basins have an overflow outlet to adjacent, topographically lower basins. Large arrow show overflow outlets and smaller arrows show general drainage directions.

drainage. Most basins are floored by a low-elevation plain on which water or sedi­ ments originating at higher elevation accumulate. Low sections on the topographic divides provide low-relief, sill-like connections to at least one adjacent basin. Thus, most basins have an outlet to adjacent, topographically lower basins (Table 3.1). Where channels are present within a province, they conform to existing topography and follow hypothetical flow-lines toward the basin floor. Hydrologic provinces incorporate both smaller, closed basins as well as nested, open drainage basins. Major drainage pathways often consist of a series of connected basins in which water moved from one basin to the next. Within hydrologic provinces, physiographic features of either regional- or localscale provide additional controls to the movement of surface water or provide clues to the movement of groundwater. Features that are used to characterize hydrologic provinces include overall relief within basins; highest and lowest points on

Hydrologic provinces of Mars: physiographic controls on drainage and ponding

75

Table 3.2 Maximum relief in hydrologic provinces Province

Highest elevationa

Lowest elevationb

Overflowc

Southern Highland

Australe-Parva Planum Hellas Planitia Argyre Planitia

2.5 3.0 3.0

1.0 –8.0 –3.0

2.0

2.0

0.5

Central Highland

Isidis Planitia Aeolis Planitia Amazonis Planitia Chryse Planitia

3.0 13.5 18.0 18.0

–3.5 –3.0 –4.0 –3.5

–3.0

–3.0

–4.0

–3.5

Northern Lowland

Vastits Borealis

12.0

–5.0

—

a

Highest elevation on divide (elevations in km).

Elevation of terminal basin.

c Elevation required to overflow terminal basin to lower watershed.

b

topographic divides; and elevation of the basin floor (Table 3.2). Topographic controls that tend to concentrate surface flow include small, closed topographic basins within the provinces; large, topographic troughs (De Hon, 1996); and canyons, valleys, and channels of various types.

3.3.1 Southern Plateau provinces The hydrologic provinces associated with the Southern Plateau are those watersheds that have their higher reaches near the South Pole and drain into local shallow catchments or into Argyre (Figure 3.3) and Hellas Planum as final catchments. The south polar region is surrounded by a high plateau. At least half of the area south of 60°S feeds directly into either the Hellas or Argyre basin (Baker et al., 1991; Head and Pratt, 2001; Milkovich et al., 2002). The remaining terrain that forms the AustraleParva Province is nestled around the South Pole. This region may have been largely covered by ice fields during a wetter pluvial or glacial epoch. Very low relief of the region would not have allowed much storage of water, and the surrounding terrain would have allowed hydrologic continuity with the Argyre, Hellas, or Aeolis drai­ nage. Exchange of water between these watersheds by spillover of lakes or glacial ice is highly probable.

The Australe-Parva Province The Australe-Parva Province is composed of the watersheds of three very shallow plains-floored basins near the South Pole. Relief in the region is quite low. Parva Planum is centered at 75°S, 260°E. The surface slopes into the plain from the south pole and from the northern divide adjacent to Solis Planum near to the 30°S latitude. The Parva watershed is unique in processing significant southward directed drainage.

76

Lakes on Mars

0°

ARGYRE BASIN

−60

−70

CS

Hd −80

90W

Hd −70

PB

CA Api

−80

80

270W

SP Apl −80

(a)

180W

5 ARGYRE BASIN

DARWIN

AR

2

ITU

M

CH

O

NT

ES

4 MIRALDI

M

1

Hd/ HNu

3 PHILLIPS

VON KARMAN

Hd/ HNu

FONTANA SCHMIDT

Apl Api

(b)

Figure 3.3 Sketch map showing channels in the south polar region leading to Argyre Basin (Head and Pratt, 2001).

Two broad troughs feed southward to merge in the region of the Parva Planum lowland. Relief of the enclosing divide is generally only 1 km except in the north where the southern edge of Solis Planum is 4–5 km above the trough floors. Overflow from Parva Planum low area could have spilled into the Argyre or small lowland

Hydrologic provinces of Mars: physiographic controls on drainage and ponding

77

basins centered at 67°S, 200°E or at 70°S, 140°E. Beyond the southern polar cap material, the region consists primarily of Noachian hilly and cratered terrain with interspersed plains of undetermined origin. The low relief of the entire region does not allow for much surface storage except as an extended ice cap.

The Hellas Province The Hellas Province includes the Hellas Basin and the surrounding highland terrain that slopes into the basin. Hellas Planitia on the floor of the basin includes the lowest elevation on the planet, more than 5000 m below planetary mean. Maximum relief in the province is approximately 11,000 m. Nearly 80% of the region is within closed depression contours at the 3000 m level. Based on topographic, morphologic, and stratigraphic evidence, Moore and Wilhems (2001) proposed that a lake occupied the floor of Hellas (see also Chapter 7). Scarps and contacts mapped in Hellas are remarkably constant in elevation over thousands of kilometers. Channels are prominent on the south and northeast interior slopes of the basin, and two large troughs lead into the basin on the west side. Dao and Niger Valles are traced from the edge of Hadriaca Patera into Hellas, and Harmahkis Vallis extends from the southwest margin of Hesperian ridged plains material surrounding Tyrrhena Patera into Hellas. Two large topographic troughs extend from highlands west of Hellas Basin. The Hellas Basin is a deep, terminal basin without any apparent outlet to adjacent basins.

The Argyre Province The Argyre Province includes the Argyre impact basin, parts of Noachis Terra high­ land to the east, and Aonia Terra highland to the west. In the south, the province stretches all the way to the South Pole. Maximum relief is 6000 m. The highest point is found along the boundary with Solis Planum and the lowest elevation is slightly less than 3000 m below the planetary datum in the Argyre Basin. Argyre Planitia is approximately 1400–1600 km in diameter. The basin is surrounded by a mountainous rim. It lies on the southern edge of a broad topographic low referred to as the Chryse Trough by Saunders (1979). The floor of Argyre is covered with materials of probable sedimentary origin, and abundant enigmatic ridges on the floor have been attributed to fluvio-glacial origin (Kargel and Strom, 1992). Numerous valley systems of Noachian and Hesperian age are found in the northern part of the province. Doanus Vallis occurs south of Argyre Planitia. Two large branching-trough systems, floored by Hesperian plateau materials, run from high terrain southward west of Argyre Planum. These south-directed troughs merge with northward slope south of Argyre Basin and direct any surface flow into Argyre. Although no channel occupies the axis of the trough, small valley networks are found on the sides of the troughs directing flow and sediment toward the central axis (De Hon, 1996). A large channel system on the north flank of Argyre hints at the possibility that the Argyre Basin filled to overflowing and discharged water into the Chryse watershed through Uzbol Vallis (Parker, 1989).

78

Lakes on Mars

3.3.2 The Central Highland provinces The Central Highland provinces straddle the equator and direct surface runoff toward the northern plains. Isidis, Aeolis, Chryse, and Amazonis provinces are regions of rugged, highland catchments and low-relief, low-elevation terminal plains. Located between 30°N and 50°S latitude, these provinces are characterized by basin floors that lie below the planetary mean. All provinces in the Central Highlands include the boundary scarp that marks the divide between the Northern Lowland plains and the cratered Central Highlands. If Mars sustained an ocean, the lowest parts of these watersheds would have been flooded (Chapter 9).

The Isidis Province Isidis Province includes Isidis Planitia and surrounding highlands of Syrtis Major Planum to the west and Terra Tyrrhena to the south. Maximum relief is over 6000 m. Approximately one-fourth of the province, consisting of Isidis Planitia, is below the planetary mean elevation. Isidis Planitia displays possible shorelines along the contact with the highlands and concentrically whorled patterns in the center part of the basin (Scott et al., 1995). Surface water exchange with the Utopia Basin could have been possible across a low-relief sill between the two basins (Table 3.2). Isidis Planitia is floored with Amazonian and Hesperian plains materials. Syrtis Major Planum is composed mostly of Hesperian volcanic plains. The surrounding highlands are Noachian etched or dissected plateau materials. Channels are abundant in the highland portion of the region. Valley networks, single channels, and fretted channels indicate surface drainage from the highlands south of Isidis Planitia toward the basin floor.

The Aeolis Province The Aeolis Province includes Aeolis Planitia, the eastern half of the Elysium Rise, and the northern part of Terra Cimmeria of the Central Highlands. Very low to nonexistent sills on the east and west sides of the Planitia connect the Aeolis Basin to Elysium Planitia lowlands to the northwest and Amazonis Planitia lowlands to the east. Relief in the watershed varies from –3000 m in the south to 13,500 m at the tops of Elysium Mons in the north. Channels across the sills indicate that water was exchanged with both adjacent basins. The highest part of the watershed reaches southward to include Planum Chronium at 60°S, 140°E. The lowland part of the province is made up of various Amazonian plains materi­ als. Amazonian knobby material is found in the plains near the boundary scarp. The highland part of the province is composed of Noachian cratered or dissected plateau materials. A large paleolake within Aeolis Planitia is documented by Scott et al. (1995). Al-Qahira Vallis and Ma’adim Vallis are major channels in the highlands and a prominent deposit marks the inflow of Ma’adim into Gusev crater (Cabrol et al., 1996, 1998a, 1998b; Grin and Cabrol, 1997). A large paleolake basin is identified at the head of Ma’adim Vallis by Rossman et al. (2002). The catchment of Ma’adim Vallis may be quite extensive, and it may reach as much as 1500 km southward beyond the

Hydrologic provinces of Mars: physiographic controls on drainage and ponding

79

Figure 3.4 Closed basin at the head of Ma’adim Vallis. MOLA gray scale topography with probable source basin delineated.

head of the valley (Figure 3.4). Numerous small valleys and valley networks are concentrated along the boundary scarp and along the crest of the planetary divide at the northern extent of the province. The lowest part of the region (Aeolis Planitia) probably existed as a shallow, closed basin and a standing lake early in its history. The basin filled with sediment derived from the highlands. After filling, water and sediment discharged across the basin floor into either Utopia Planitia or Amazonis Planitia.

The Amazonis Province The Amazonis Province includes Amazonis Planitia, Olympus Mons, the western half of the Tharsis Ridge, and much of Terra Sirenum of the highlands. Maximum relief is in excess of 25 km from the floor of the basin to the top of Olympus Mons which is entirely within the boundaries of the province. Maximum relief from the basin floor to a high point along the basin divide at the Tharsis Volcanoes is 18 km. The Aeolis Province flanks this region to the west, the Australe-Parva Province to the south, the Chryse Province to the east, and the Borealis Province to the north. Low sills separate Amazonis from Arcadia and Aeolis provinces. Several small, closed basins occur within the highland portion of the province. Approximately one-fourth of the province is below the planetary mean and would have been flooded by a putative northern ocean. Scott et al. (1992, 1995) postulated a

80

Lakes on Mars

large paleolake basin in Amazonis Planitia based on remnant shoreline features. The lowland portion of Amazonis Basin may have received overflow from the Aeolis Basin and may have spilled over to feed water northward into the lowest portions of Vastitis Borealis. Surface materials are varied. Noachian cratered plateau materials make up approximately half of the highland portion of the region. Hesperian material is limited to channel material and sparsely-cratered Hesperian plateau material which floors large, shallow depressions. Plateau material is interpreted by Scott and Tanaka (1986) as mixed volcanic flows and sediment and by Parker and Gorsline (1991b) and De Hon (1992b) as probably lacustrine or fluvial sediments. Amazonian Tharsis and Olympus lavas make up most of the eastern part of the region. The lowland part of the region in Amazonis Basin, which is mapped as part of the Arcadia Formation and interpreted as lava flows (Scott and Tanaka, 1986), may be sedimentary materials. Mouginis-Mark (1993) proposed that the basal scarp of Olympus Mons is a wave-cut cliff. If the scarp is of marine origin, an Amazonian sea-level stand would have reached 2000–3000 m above the planetary datum. Such a level would have flooded much of Lunae Planum, Tempe Terra, Arabia Terra, and Xanthe Terra. The most prominent channel system is that of the Mangala Valles and Labou Vallis which extends at least 600 km from the boundary scarp. Mangala Vallis contains several intervalley basins which filled to overflow in a series of spillways (De Hon, 1994; Emrick and De Hon, 1999). The catchment for Mangala Valles may extend far beyond its reported source at the Memonia Fossae (Zimbelman et al., 1992). A large, closed basin centered on 40°S (Figure 3.5) could have fed water into the Mangala system (De Hon, 1994; Parker and Gorsline, 1991b). Short channels are common along the highland boundary scarp, and Noachian channels are found near the south­ ern boundary of the province. All surface flow is toward the Amazonis Planitia lowlands. No fluvial channels are identified on the Amazonian lavas of Olympus Mons.

The Chryse Province The Chryse Province is the largest of the equatorial hydrologic provinces and contains the most abundant water-related features. The province is bounded by the Tharsis Rise and Alba Patera to the west, cratered highlands of Aonia Terra and Noachis Terra in the south, and Terra Meridiani in the east. A low sill separates the Chryse floor from the lower Acidalia Planitia in the north. Elevations range from less than –3500 m elevation on the floor of the basin to 18,000 m elevation at the crest of Ascraeus Mons on the Tharsis Ridge. Other high points occur at the southwest corner of Solis Planum (13,000 m elevation) and the northern rim of Corprates Canyon (9000 m elevation), with a maximum occurring in the far southeast border along the west rim of the Hellas Basin (5000 m elevation). A large, shallow, closed depression in Solis Planum and Sinai Planum forms a hydrologic subprovince. Materials are varied within this province. Noachian cratered plateau material com­ prises most of the rugged highland region. Hesperian ridged plains material caps the

Hydrologic provinces of Mars: physiographic controls on drainage and ponding

81

Figure 3.5 Channels and ponds in Mangala and Labou Valles. MOLA gray scale topography with probable source basins and channels marked. Upper Mangala Vallis is not shown on this map as it is on the western floor of the lower source basin.

Lunae Plaum region and occurs on the floor of Chryse Planitia. Lowest Arcadia Formation occurs in northern Chryse Planitia. Volcanic materials of the Alba and Tharsis formations are found on the western edge of the province. Chryse Planitia, the lowest level of the drainage basin, is the natural ponding site for water shed from the higher elevations. The basin floor occupies a circular depression that has filled with sediments and perhaps volcanics to allow surface drainage to merge with Acidalia Planitia. The Chryse Basin is included in the putative northern ocean (Baker et al., 1991; Parker et al., 1989, 1993). Numerous local impoundments within the province have been documented (De Hon, 1992b; McCauley, 1978; Nedell et al., 1987; Scott et al., 1995). These local paleolakes, fed by outflow channels, were located in the northern section of the Kasei Vallis, the northern section of Maja Valles on Lunae Planum, parts of Xanthe Terra, the western edge of Chryse Planitia (De Hon, 1992b), the northern part of Ares Vallis (Baker et al., 1991), and the Ladon Vallis region (Boothroyd, 1982; Moore, 1982). Local ponding of drainage also occurred in the canyon system in Candor Chasma and Ophir Chasma (McCauley, 1978; Nedell et al., 1987) and in the Aurorae Basin in Eos Chasma at the eastern end of the canyon system (Scott et al., 1995). Surface drainage is prominent with a large number of channels draining toward Chryse Planitia. Outflow channels are abundant, and they are associated with the steep surface gradient north and east of the Valles Marineris canyon system. The east–west canyon

82

Lakes on Mars

system is over 2000 km long and is cut into a high-standing plateau that reaches up to 9000 m elevation. The canyon is open on the east end and provides a continuous, open trough which connects with the channels that empty onto Chyrse Planitia. Kasei Vallis and Maja Valles begin in box canyons on the north slope of the Valles Marineris plateau. Outflow channels associated with the east end of the canyon system and the southern flanks of the Chryse Basin include Simud, Tiu, and Ares Valles. Nirgal, Uzboi, and Ladon Valles occur in the highland terrain south of Eos Chasma. Chaotic terrain at the head of some outflow channels and adjacent to other channels attests to saturated surface materials and effluent seepage (Nummedal, 1978). Noachian and Hesperian network drainage is common in the highlands in the southern and eastern parts of the province. Plains materials of lacustrine or alluvial origins are common downslope from valley networks in the highlands. Basins, including Chryse Planitia, are floored by channel flood plain material and ridged plains material. Ridged plain materials were originally interpreted as volcanic because the ridges resemble those of the lunar mare material, but some evidence by Plescia and Golombek (1986) suggests that ridges may form in other types of materials and in basins such as Chryse Planitia which may, in fact, be filled with sedimentary materials.

3.3.3 The Northern Lowland—Borealis The Northern Lowland, consisting of the northern polar region, has only small, highland catchments, but they receive the spillover from the Central Highland basins. In the Borealis Province, Utopia and Acidalia sub-basins are lowland basins with very low relief and floor elevations below the planetary mean (Figure 3.6) and could have been completely inundated if an ocean occupied the Northern Lowlands (Baker et al., 1990, 1991; Head et al., 1999; see also Chapter 9). The Borealis Province includes all of the Northern Lowland and only a very small region of highlands west of the Isidis watershed. The overall relief is low except for Elysium and Alba Patera volcanic constructs. The lowest portion of the province is found in the North Polar Basin centered at 70°N, 327°E. This sub-basin exhibits low elevations and low relief. Another low, the Utopia sub-basin, is centered at 40°N, 110°E north of Isidis Basin. Most of the province consists of materials of the Amazonian age, Alba Patera Formation and mottled, grooved, and knobby materials of the Hesperian Vastitas Borealis Formation. Some faint traces of channels indicate surface flow from Chryse Planitia into Acidalia Planitia. The Acidalia Basin is fed primarily by overflow of the Chryse Basin, but as the lowest basin with very low divides, it could have received spillover from all adjacent basins during wet periods and have merged with the Amazonis and Utopia basins if an ocean had occupied the Northern hemisphere (Baker et al., 1990). Utopia Planitia encompasses a large part of the northern plains including much of Elysium and Utopia Planitiae. Surface materials include Elysium lavas, knobby plains, and smooth plains of Amazonian age, as well as ridged plains and grooved plains of Hesperian age. The lowest part of the basin, below the –5000 m contour, is in Utopia Planitia. Because the sill separating the lowest parts of Elysium and Utopia lowlands is less than 1 km in relief, the lowland plain of Elysium Planum west of Elysium Mons is

Hydrologic provinces of Mars: physiographic controls on drainage and ponding

83

Figure 3.6 The northern hemisphere of Mars (Carr and Head, 2003) showing north polar and Utopia sub-basins of the Borealis Province.

included as a sub-basin within this province. Maximum relief in the province is high because the Elysium Rise occurs on the eastern boundary of the region and reaches up to 14 km in elevation. Minor areas of the flanking highland terrains, eastern Arabia Terra on the west, a portion of Terra Cimmeria, and the western half of Elysium Rise are included within this province. Low sills separate Elysium Planitia from Isidis Planum to the southwest and from Aeolis Province to the southeast. In the far northern plains, Utopia is separated from the surrounding provinces of Acidalia and Arcadia by very low, broad topographic rises. Much of the province would have been flooded (Figure 3.7) by the northern ocean, Oceanus Borealis of Baker et al. (1991). Scott et al. (1995) propose four paleolakes within Utopia (identified as Utopia A, B, and C basins and Cebrena Basin) based on possible partial paleoshorelines. Identifiable surface drainage includes Augakuh Vallis and Huo Hsing Vallis in western Arabia Terra; Licus Vallis in northern Cimmeria; and the Herbus, Granicus,

84

Lakes on Mars 0°

CIR C

UM HR -C

HEAVILY CRATERED TERRAIN

FRET.

T FLOW CH

D

GRA

D

60°

FR ET TE D

GRAD.

270°

TE ET FR

S

SYRTIS MAJOR VOLCANICS

ANN EL

TE

R

MAMERS VALLES IN RA

E OU YS

MAWRTH VALLIS

90°

N.W. ELYSIUM CHANNELS 60°

THARSIS VOLCANICS

FRETTED TER

RA

ALBA PATERA VOLCANICS

ELYSIUM VOLCANICS IN MA RIE VALLIS 0°

AL QAHIRA VALLIS

RY UN D A BO L A GRADATION

MANGALA VALLES

MA’ADIM VALLIS

Figure 3.7 Shorelines of a putative ocean in the northern hemisphere. The northern plains are the lowest region on the planet and serve as the terminal basin for most of the watersheds.

Tinger, and Hrad Valles on the western and northern flanks of the Elysium Rise. Most channels on the western and northern flanks of Elysium Mons begin in fissures in the Elysium lavas or from grooved terrain on the northern margin of the lavas. Water may have been released from magma, from melting permafrost, from dewatering of volcaniclastic material, or expelled from sediments by loading of the surface with volcanic materials. Channels lead downslope to terminate in small, isolated topo­ graphic lows within Utopia or Elysium Planitiae.

3.4 Discussion Significant changes in position of divides that define provinces in the geologic past are those affected by the formation of the northern plains and boundary scarp, the formation of the Argyre and Hellas impact basins in the Noachian Period, and the

Hydrologic provinces of Mars: physiographic controls on drainage and ponding

85

construction of the large volcanoes in Amazonian time. It is difficult to speculate on the earliest configuration of the planet’s northern surface prior to the formation of the northern plains. Since that time, the surface configuration has apparently changed little other than the addition of volcanic constructs to the preexisting topography. Many of the provinces display unique occurrences of water-derived features. The highland portions of these provinces preserve ancient network drainage as well as younger outflow and sapping channels. The lowest elevation portions of the provinces typically exhibit only young plains-forming material and young channel systems. The lowest plains within a province are formed by materials that are often assigned to volcanic or mixed volcanic and sedimentary origins (Greeley and Guest, 1987; Scott and Tanaka, 1986) but many such plains may be fluvial or lacustrine in origin (Cabrol and Grin, 2001, 2002; De Hon, 1992b; Godspiel and Squyres, 1991). The polar ocean of Baker et al. (1991) and large paleolakes of Scott et al. (1995) are not mutually exclusive. Paleolakes may be remnants of a recessional ocean. Scott et al. (1995) document possible shoreline features that represent bodies of water that enclose areas that are significantly less than an all-encompassing polar ocean (see also Chapters 9 and 10). They also identify faint Hesperian- and Amazonian-age channels within the lowland plains that would have been submerged and would not have survived the presence of a polar ocean. However, these features could have formed after the ocean disappeared or even while it was shrinking. Of course, if the ocean shoreline coincided with the boundary scarp, much of the evidence may have been destroyed by subsequent modification of the scarp. Clear evidence for surface runoff exists on Mars. Wherever there is surface drainage, ponding is bound to occur. Craters provided a natural closed basins for the impoundment of water. On Earth, volcanic craters as small as maars to large collapse calderas exist as prominent lakes. Ancient impact scars in Canada are filled with water. On Mars many craters are breached by inflow channels and show deltas, alluvial fans, and layered materials attesting to sediment transport and deposition, and most likely standing bodies of water (Baker, 2009; Cabrol and Grin, 1999; Cabrol et al., 2001, 2002, 2005; Fassett and Head, 2008).

References Baker, V.R., 1990. Spring sapping and valley network development. Geol. Soc. Am. Spec. Pap. 252, 235–265. Baker, V.R., 2009. Megafloods and global paleoenvironmental change on Mars and Earth. Geol. Soc. Am. Spec. Pap. 453, 25–36. Baker, V.R., Strom, R.G., Croft, S.K., Gulick, V.C., Kargel, J.S., Komatsu, G., 1990. Ancient ocean-land-atmospheric interactions on Mars: Global model and geological evidence. Lunar Planet. Sci. Conf. XXI, 40–41. Baker, V.R., Strom, R.G., Dohm, J.M., Gulick, V.C., Kargel, J.S., Komatsu, G., et al., 2000. Mars, Oceanus Borealis, ancient glaciers and the MEGAOUTFLO hypothesis. Lunar Planet. Sci. XXXI, Abstract 1862 (CD-ROM).

86

Lakes on Mars

Baker, V.R., Strom, R.G., Gulick, V.C., Kargel, J.S., Komatsu, G., Kale, V.S., 1991. Ancient oceans, ice sheets, and the hydrological cycle on Mars. Nature 352, 589–594. Boothroyd, J.C., 1982. Ancient fluvial drainage systems: Margaritifer Sinus area, Mars. NASA TM-85127, 209–212. Brackenridge, G.R., Newsom, H.E., Baker, V.R., 1985. Ancient hot springs on Mars: Origins and paleosignificance of small martian valleys. Geology 13, 859–862. Cabrol, N.A., Grin, E.A., 1999. Distribution, classification and ages of martian impact crater lakes. Icarus 142, 160–172. Cabrol, N.A., Grin, E.A., 2001. The evolution of lacustrine environments on Mars: Is Mars only hydrologically dormant? Icarus 149, 291–328. Cabrol, N.A., Grin, E.A., 2002. Overview on the formation of paleolakes and ponds in impact craters on Mars. Global and Plan. Change 35, 199. Cabrol, N.A., Grin, E.A., 2005. Ancient and recent lakes on mars. In: Tokano, T. (Ed.), Water on Mars and Life. Springer, Berlin, pp. 235–259. Cabrol, N.A., Grin, E.A., Dawidowicz, G., 1996. Ma’adim Vallis revisited through new topographic data. Icarus 123, 269–283. Cabrol, N.A., Grin, E.A., Landheim, R., Kuzmin, R., Greeley, R., 1998a. Duration of the Ma’adim Vallis/Gusev crater hydrogeologic system, Mars. Icarus 133, 98–108. Cabrol, N.A., Grin, E.A., Landheim, R., 1998b. Ma’adim Vallis evolution: Geometry and models of discharge rate. Icarus 132, 362–377. Cabrol, N.A., Wynn-Williams, D.D., Crawford, D.A., Grin, E.A., 2001. Recent aqueous environments in impact crater lakes on Mars: an astrobiological perspective, 2nd Mars Polar Conference Special Issue. Icarus 154, 98–112. Carr, M.H., 1979. Formation of martian flood features by release of water from confined aquifers. J. Geophys. Res. 84, 2995–3007. Carr, M.H., Crumpler, L.S., Cutts, J.A., Greeley, R., Guest, J.E., Masursky, H., 1977. Martian impact craters and emplacement of ejecta by surface flow. J. Geophys. Res. 82, 4055–4065. Carr, M. H., and Head III, J. W. 2003. Oceans on Mars: An assessment of the observational evidence and possible fate. J. Geophys. Res., 108(E5), 5042–5070. Clark, C., 2005. The martian watershed: Geology, dichotomy, and paleohydrology on two world maps with constant scale natural boundaries. Lunar Planet Conf. XXXVI, Abstract 2189, Lunar and planetary Institute (CD-ROM). Coates, D.R., 1990. Geomorphic controls of groundwater hydrology. Geol. Soc. Am. Spec. Pap. 252, 341–356. Craddock, R.A., Howard, A.D., 2002. The case for rainfall on a warm, wet Mars. J. Geophys. Res. 107 (E11), 21-1–21–36. De Hon, R.A., 1987. Eastern Lunae Planum outflow complex: Analogy to overbank flooding. Lunar Planet. Sci. Conf. XVIII, 227–228. De Hon, R.A., 1988. Ephemeral martian lakes: Temporary ponding and local sedimentation. Lunar Planet. Sci. Conf. XIX, 261–262. De Hon, R.A., 1992a. Polygenic origin of Hrad Vallis regopn of Mars. Proceedings of Lunar and planetary Science, vol. 22 Lunar and Planetary Institute, Houston, TX, 45–51. De Hon, R.A., 1992b. Martian lake basins and lacustrine plains. Earth Moon Planets 56, 95–122. De Hon, R.A., 1994. Lacustrine sedimentation in lower Mangala Valles. Mars. Lunar Planet. Sci. Conf. XXV, 321–322. De Hon, R.A., 1996. Martian highland troughs: Regional controls of runoff and sedimentation. Lunar Planet. Sci. Conf. XXVII, 295–296.

Hydrologic provinces of Mars: physiographic controls on drainage and ponding

87

De Hon, R.A., Pani, E.A., 1992. Duration and rate of discharge, Maja Valles, Mars. J. Geophys. Res. Planets 98, 9129–9138. Dickson, J.L., Head, J.W., Marchant, D.R., 2008. Late Amazonian glaciation at the dichotomy boundary or Mars: Evidence for glacial thickness maxima and multiple glacial phases. Geology 36 (5), 411–414. Dohm, J.M., Anderson, R.C., Baker, V.R., Farris, J.C., Rudd, L.P., Hare, T.M., et al., 2001. Latent outflow for western Tharsis, Mars: Significant flood record exposed. J. Geophys. Res. 106 (E6), 12301–12314. Emrick, C.M., De Hon, R.A., 1999. Flood discharge through Labou Vallis, Mars. Lunar Planet. Sci. Conf. XXX, Abstract #1893, Lunar and Planetary Institute, Houston (CD-ROM). Farmer, J., 1996. Hydrothermal Systems on Mars: An Assessment of Present Evidence, in Evolution of Hydrothermal Ecosystems on Earth (and Mars?). John Wiley, New York, pp. 273–299. Fassett, C.I., Head, J.W., 2008. Valley network-fed, open basin lakes on Mars: Distribution and implications for Noachian surface and subsurface hydrology. Icarus 198 (1), 37–56. Forget, F., Haberle, R.M., Montmessin, F., Levrard, B., Head, J., 2006. Formation of glaciers on Mars by atmospheric precipitation at high obliquity. Science 311, 368–371. Gault, D.E., Greeley, R., 1978. Exploratory experiments of impact craters formed in viscousliquid targets: Analogs for martian rampart craters? Icarus 34, 486–495. Greeley, R., Guest, J.E., 1987. Geologic map of the eastern equatorial region of Mars. U.S.G.S. Misc. Geol. Inv. Map I–1802B. Grin, E.A., Cabrol, N.A., 1997. Limnologic analysis of Gusev crater paleolake, Mars. Icarus 130, 461–474. Grizzaffi, P., Schultz, P.H., 1989. Isidis Basin: Site of ancient volatile-rich debris layer. Icarus 77, 358–391. Gulick, V.C., 1998. Magmatic intrusions and a hydrothermal origin for fluvial valleys on Mars. J. Geophys. Res. 103, 19365–19387. Gulick, V.C., Baker, V.R., 1989. The role of hydrothermal circulation in the formation of fluvial valleys on Mars. Lunar Planet. Sci. Conf. XX, 369–370. Head, J.W., Hiesinger, H., Ivanov, M.A., Kreslavsky, M.A., Pratt, S., Thomson, B.J., 1999. Possible ancient oceans on Mars: Evidence from Mars Orbiter Laser Altimeter data. Science 286 (5447), 2134–2137. Head, J.W., Mustard, J.F., Kreslavsky, M.A., Milliken, R.E., Marchant, D.R., 2003. Recent ice ages on Mars. Nature 426, 797–802. Head, J., Nahm, A.L., Marchant, D.R., Neukum, G., 2005. Modification of the dichotomy boundary on Mars by Amazonian mid-latitude regional glaciation. Geophys. Res. Lett. 33, LO8S03. Head, J.W., Pratt, S., 2001. Extensive Hesprian-aged south polar ice cap on Mars: Evidence for massive melting and retreat, and lateral flow and ponding of melt water. J. Geophys. Res. 106, 12275–12299. Hynek, B.M., Phillips, R.J., 2003. New data reveal mature, integrated drainage systems on Mars indicative of past precipitation. Geology 31, 757–760. Irwin, R.P., Howard, A.D., Craddock, R.A., Moore, J.F., 2005. An intense terminal epoch of widespread fluvial activity on early Mars. 2. Increased runoff and paleolake development. J. Geophys. Res. 110, E12S14. Irwin, R.P. III, Maxwell, T.A., Howard, A.D., Craddock, R.A., Leverington, D.W., 2002. A Large paleolake basin at the head of Ma’adim Vallis, Mars. Science 296 (5576), 2209–2212.

88

Lakes on Mars

Johanson, L., 1979. The latitude dependence of martian splosh craters and its relationship to water. NASA TM-80339, 123–124. Kargel, J.S., 2004. Mars: A Warmer Wetter Planet. Springer-Praxis Books, New York, 557p. Kargel, J.S., Strom, R.G., 1992. Ancient glaciation on Mars. Geology 20, 3–7. Laity, J.E., Malin, M.C., 1985. Sapping processes and the development of theater-headed valley networks in the Colorado Plateau. Geol. Soc. Am. Bull. 96, 203–217. Laskar, J., Robutel, P., 1993. The chaotic obliquity of the planets. Nature 362, 608–612. McCauley, J., 1978. Geologic Map of the Coprates Quadrangle of Mars. U.S.G.S. Misc. Geol. Inv. Map I-897. McGill, G.E., Hills, L.S., 1992. Origin of martian polygons. J. Geophys. Res. 97, 2633–2647. Milkovich, S.M., Head, J.W., Pratt, S., 2002. Meltback of Hesperian-aged ice-rich deposits near the south pole of Mars: Evidence for drainage channels and lakes. J. Geophys. Res. 107 (E6), 10-1–10–13. Moore, H.J., 1982. Channel deposits on Mars. NASA TM-89127, 213–215. Moore, J.M., Wilhelms, D.E., 2001. Hellas a possible site of ancient ice-covered lakes on Mars. Icarus 154, 258–276. Mouginis-Mark, P.J., 1979. Martian fluidized crater morphology: Variations with crater size, latitude, and target material. J. Geophys. Res. 84, 8011–8022. Mouginis-Mark, P.J., 1985. Volcano-ground ice interactions in Elysium Planitia. Icarus 64, 265–284. Mouginis-Mark, P.J., 1993. The influence of oceans on martian volcanism. Lunar Planet. Sci. Conf. XXIV, 1021–1022. Nedell, S.S., Squyres, S.W., Andersen, D.W., 1987. Origin and evolution of the layered deposits in Valles Marineris, Mars. Icarus 70, 409–441. Nummedal, D., 1978. The role of liquefaction in channel development on Mars. NASA TM­ 79729, 257–259. Parker, T.J., 1989. Channels and valley networks associated with Argyre Planitia, Mars. Lunar Planet. Sci. Conf. XX, 826–827. Parker, T.J., Gorsline, D.S., 1991a. Where is the source for Uzboi Vallis, Mars? Lunar Planet. Sci. Conf. XXII, 1033–1034. Parker, T.J., Gorsline, D.S., 1991b. Formation of Mangala Valles, Mars, through catastrophic drainage of a large surface lake, Abs. Lunar Planet. Sci. Conf. XXII, 1031–1032. Parker, T.J., Gorsline, D.S., Saunders, R.S., Pieri, D.C., Schneeberger, D.M., 1993. Coastal geomorphology of the Martian Northern Plains. J. Geophys. Res. 98, 11061–11078. Parker, T.J., Saunders, R.S., Schneeberger, D.M., 1989. Transitional morphology in West Deuteronilus Mensae, Mars: Implications for modification of the lowland/upland bound­ ary. Icarus 82, 111–145. Plecsia, J.B., Golombek, M.P., 1987. Origin of planetary wrinkle ridges based on a study of terrestrial analogs. Geol. Soc. Am. Bull. 97, 1289–1299. Pollack, J.B., Kasting, J.F., Richardson, S.M., Poliakoff, K., 1987. The case for a warm, wet climate on early Mars. Icarus 71, 203–224. Rossbacher, L.A., Judson, S., 1981. Ground ice on Mars: Inventory, distribution and resulting landforms. Icarus 45, 39–59. Rossman, R.I. III, Maxwell, T.A., Howard, A.D., Craddock, R.A., Leverington, D.W., 2002. A large paleolake basin at the head of Ma’adim Vallis, Mars. Science 296 (5576), 2209–2212. Saunder, R.S., 1979. Geologic map of the Argyre Quadrangle. U. S. G. S. Misc. Inves. Map I-1134 (MC-26), Scale 1:5,000,000.

Hydrologic provinces of Mars: physiographic controls on drainage and ponding

89

Schulze-Makuch, D., Dohm, J.M., Fan, C., Fairén, A.G., Rodriquez, J.A.P., Baker, V.R., et al., 2007. Exploration of hydrothermal targets on Mars. Icarus 10 (189), 308–324. Scott, D.H., Chapman., M.G., Rice, J.W. Jr., Dohm, J.M., 1992. New evidence of lacustrine basins on Mars: Amazonis and Utopia Planitiae. Proc. Lunar Planet. Sci. Conf. 22nd, 53–62. Scott, D.H., Dohm, J.M., Rice, J.W. Jr., 1995. Map of Mars showing channels and possible paleolakes. U.S.G.S., Misc. Geol. Inv. Map I–2461. Scott, D.H., Rice, J.W., Jr., Dohm, J.M., 1991. Martian paleolakes and waterways: exobiologi­ cal implications. Origins of Life and Evolution of the Biosphere: Journal of International Society for the Origin of Life, Kluwer Academic Publishers, The Netherlands, pp. 189–198. Scott, D.H., Tanaka, K.L., 1986. Geologic map of the western equatorial region of Mars. U.S.G.S. Misc. Geol. Inv. Map I–1802A. Squyres, S.W., 1979. The distribution of lobate debris aprons and similar flows on Mars. J. Geophys. Res. 84, 8087–8096. Tanaka, K.L., Chapman, M.G., 1992. Kasei Valles, Mars: Interpretaion of canyon materials and flood sources. Proceedings of Lunar and Planetary Science, vol. 22; Conference, Houston, TX, Lunar and Planetary Institute, 73–83. Touma, J., Wilson, J., 1993. The chaotic obliquity of Mars. Science 259, 1294–1297. USGS, 2003. Color-Coded Contour Map of Mars, U.S.G.S. Misc. Inves. Map I-2782. Zimbelman, J.R., Craddock, R.A., Greeley, R., Kuzmin, R.O., 1992. Volatile history of Mangala Valles, Mars. J. Geophys. Res. 97 (E11), 18309–18317.

4 Heated lakes on Mars Horton E. Newsom Institute of Meteoritics, University of New Mexico, Albuquerque, NM, USA

4.1

Introduction

The sedimentary deposits in lakes are important for studying the nature of habitable environments for microbial life on Mars (Benison and Bowen, 2006; Cabrol et al., 2007; Mahaffy, 2008; McKay, 1997; Mormile et al., 2009) and represent high-priority landing sites for future rover missions, such as MSL to be launched in 2011. Geomorphic evidence for the past existence of liquid water lakes on Mars and their associated sedimentary deposits continues to grow (Cabrol and Grin, 1999; 2001; 2003; 2005; Cabrol et al., 1999; 2001; 2007; Di Achille et al., 2007; Malin and Edgett, 2000; Fassett and Head, 2005; Ehlmann et al., 2008a; Goldspiel and Squyres, 1991, 2000; Moore and Howard, 2005; Ori et al., 2009; Pondrelli et al., 2008). It includes deltaic structures (Figures 4.1 and 4.2), fluvial deposits, and channels leading into and out of depressions (see Chapter 8). Other types of evidence are harder to confirm, such as shoreline deposits, including wave-cut terraces, and ice-push deposits (Newsom et al., 2003). With HiRISE (30 cm/pixel resolution), features such as lake strandlines might have been identified (Di Achille et al., 2009). In cold conditions, ice dams can also result in the formation or enlargement of lakes (Newsom et al., 2003) that will be difficult to substantiate. The new orbital datasets can be used to address the nature of the connection between the Martian climate and the existence of lakes (Bell III et al., 2008). The early discovery of apparent large fluvial channels on Mars led to the concept of an early warm and wet Mars (Carr, 1996, see also Chapters 1 and 2). A warm climate, however, is not a necessary condition for the existence of lakes on Mars. A number of heat sources can result in the temporary presence of bodies of liquid water supplied by groundwater or water from snowmelt and ice due to impacts or volcanism (Abramov and Kring, 2005; Brakenridge et al., 1985; Gulick, 1998; McKay and Davis, 1991; McKay et al., 1985; Newsom, 1980; Newsom et al., 1996; Rathbun and Squyres, 2002). Figure 4.3 shows a concept of an impact crater with associated lake based on the theoretical studies discussed below. Once formed, lakes can persist for long periods under cold conditions, even in the absence of additional heat, due to the large enthalpy of the fusion of ice (McKay and Davis, 1991; Newsom et al., 1996). In this chapter, we examine the potential heat sources for Martian lakes, the implications of sedimentary and associated hydrothermal deposits for determining the history of climate on Mars, and for the search for deposits containing evidence of Lakes on Mars. DOI: 10.1016/B978-0-444-52854-4.00004-0 © 2010 Elsevier B.V. All rights reserved.

92

Lakes on Mars

Figure 4.1 An image of the fan-type delta in Jezero crater with numerous distributary channels, fed by a channel from the upper left. The delta is partly eroded with the channel deposits forming positive relief features characteristic of inverted terrain (Pain et al., 2007). CRISM data have been used to suggest that portions of the deltaic deposits may contain clay minerals (Elhmann et al., 2008a). Portion of CTX image PO2_001820_1984_XI_18N282W_070129 centered approximately at 18.47° N, 282.64° E longitude. Image credit: NASA/JPL/Malin Space Science Systems, image width 12 km.

microbial habitability (Cockell, 2006; Des Marais et al., 2008; Farmer, 2006; Farmer and Des Marais, 1999; Varnes et al., 2003). Finally, we discuss potential planetary protection concerns about the possible existence of current liquid lakes on Mars (Chyba et al., 2006) and NASA (Beaty et al., 2006).

Figure 4.2 This image shows a Gilbert-type layered deltaic deposit on the floor of a large impact crater. The material was transported into the crater, likely by running water from a valley on the upper left part of the image. The arcuate nature of the delta is consistent with deposition into a lake. HiRISE image PSP_008233_1920 centered at 11.7° N, 307.1° E, image width 6 km, credit: NASA/JPL/University of Arizona.

Heated lakes on Mars

93

Figure 4.3 Artist’s conception of an impact crater in the 20 km size range shown shortly after formation of the crater. The floor of the crater with a modest central uplift is covered by an impact melt bearing breccia deposit, which is covered by a lake. The lake is assumed to be heated by the impact melt. The vertical exaggeration is about 10 times. Image courtesy G. Fredrick (Nor’Webster Digital Imaging, [email protected]), modified by H. Newsom.

4.2

Sources of water

Lake types, lake deposits, and the evolution of climate are detailed in Chapters 1 and 2. In the following sections, we focus on the potential sources of water and heat that could have favored the formation of lakes throughout Martian history.

4.2.1 Climate and precipitation Mars is currently too cold to sustain lakes at the surface. Nevertheless, evidence for relatively abundant water early in the history of the planet suggests that lakes were common during more clement periods (Carr, 1996; Irwin et al., 2005, 2008; Moore et al., 2003; Newsom et al., 2009). Evidence that a relatively warmer climate existed at one time includes the existence of widespread valleys and channels, suggesting the role of precipitation or snowmelt. Irwin et al. (2005) note that generally sparse, immature valleys with poorly dissected interfluve areas were incised during one or more epochs of more intense fluvial activity around the Noachian/Hesperian transition (3.7 Ga). An additional argument for warmer periods on Mars comes from research into the effects of large obliquity changes for Mars that may have varied from the current 25° up to 60° (Laskar et al., 2004). The astronomically induced variations in insolation are significant (Lasue and Clifford, 2009)—resulting in kilometer-scale variations in the depth of the cryosphere at high latitude. However, warming Mars to temperatures above the freezing point, except locally, is still difficult, based on numerical climate models (Jakosky et al., 2005). Early in Mars history, the lower solar flux makes it even more difficult to have warm conditions above freezing over substantial areas of Mars (Gulick et al., 1997; Zahnle et al., 2008).

94

Lakes on Mars

A special case of climate change arises from the potential for moderate- to largersized impacts (>50 km diameter craters) to create greenhouse conditions for a tem­ porary warmer climate associated with abundant precipitation (Segura et al., 2002, 2008). In another special case, large volumes of surface water can also be supplied by local melting of ice, as may have occurred at the north polar cap during the last 20,000 years (Hovius et al., 2008).

4.2.2 Groundwater The role of groundwater and groundwater flow in the formation of lakes and channels on Mars is well recognized (Carr, 1996; Komatsu et al., 2009). Under cold climatic conditions, lakes could also have formed by large outflows of groundwater from chaos regions (Baker, 1982). The formation of impact craters can also lead to the formation of lakes if a groundwater table is penetrated, even in the presence of a thick permafrost layer or cryosphere (Barnhart et al., 2009; Newsom et al., 1996). Impact craters as small as 12 km diameter can penetrate a 6-km-thick cryosphere; this thickness is characteristic of the maximum cryosphere thickness at the poles during the late Noachian (Schwenzer et al., 2010). As the cryosphere was as thin as 2 km near the equator, most moderatesized craters probably had access to deep aquifers. The supply of groundwater to a lake depends on the depth of the basin, the thickness of the penetrated aquifer, and regional conditions that constrain the available supply of water (Newsom et al., 1996). The plausibility of supplying large volumes of groundwater to lakes and Martian outflow channels from aquifers has been addressed by groundwater flow calculations (Harrison and Grimm, 2008; Newsom et al., 1996). The largest uncertainty in these calculations is the regional permeability, which can vary over many orders of magnitude. Based on regional groundwater flow models, Harrison and Grimm (2008) derived a most likely permeability of 3.4 � 10–12 m2, which led to a flood volume of approximately 200 km3 over a 30-day period delivered to a circular point source from a thick regional aquifer. This volume would fill a 70-km-diameter impact crater to a depth of about 500 m in about 30 days with a continuing, but decreasing flow rate. For greater permeability, available water volumes could plausibly be much larger.

4.2.3 Melting of ice due to impacts As a result of a large impact event, melting of snow and ice deposits incorporated in the ejecta blanket and present in the subsurface below the transient cavity might lead to the formation of lakes. Experiments and modeling of impacts onto ice-rock mixtures suggest that ice is easily melted by impact due to its compressibility (Stewart et al., 2001). Recent observations of young impact craters on Mars have revealed deposits of material flowing down the walls of the craters (Mouginis-Mark, 2007; Tornabene et al., 2007). These deposits may consist of variable amounts of solidified impact melt and lahar-type mudflows produced by melting of ground ice. The formation of valley networks associated with large craters, such as Sinton Crater, may also imply melting of ice as a water source (Morgan and Head, 2009; Tanaka et al., 1998).

Heated lakes on Mars

4.3

95

Sources of heat

4.3.1 Heating due to impacts and volcanism Volcanism and impact cratering have the potential to lead to the formation of lakes and associated hydrothermal systems that could endure on Mars for many thousands of years, even under cold climatic conditions (Abramov and Kring, 2005; Gulick, 1998; Lanz and Saric, 2009; Newsom, 1980; Newsom, et al., 1996; Pope et al., 2006). The primary heat sources in large impact craters on Mars after their formation are (i) impact melts primarily generated by the shock wave, (ii) heat deposited in the basement below the transient cavity from passage of the shock wave, and (iii) additional heat in the central uplift from the uplifted geothermal gradient (Kieffer and Simonds, 1980; O'Keefe and Ahrens, 1994, 1999; Pierazzo et al., 2005). A number of factors can influence the amount of melt generated during an impact event. The amount of impact melt produced in Martian impact craters as a function of size has been estimated (Pope et al., 2006) but substantial uncertainties exist. Varia­ tions in the impact angle, composition of the impactor, composition of the planetary surface, nature of layering in the target, and amount of vaporization all contribute to uncertainties in impact melt production. The presence of volatile materials (e.g., CO2 and H2O) in the crust also decreases the amount of melt produced as compared to craters formed in crystalline rocks (Kieffer and Simonds, 1980). The discovery of extensive evaporite materials, for example, in Meridiani Planum, suggests that these effects can be locally important on Mars (McLennan and Grotzinger, 2008). The magnitude of the impact-induced shock heating depends on the transient crater diameter and the location relative to the transient cavity (Ivanov and Deutsch, 1999; Pierazzo et al., 2005). Numerical modeling indicates that this heating is highest under the transient cavity floor within the region where central peaks and central pits form. In Martian craters larger than 30 km, another source of heat is created by the uplift of warmer basement rocks (Abramov and Kring, 2004, 2005; Daubar and Kring, 2001; McCarville and Crossey, 1996; Newsom et al., 1996, 2001). The total heat content of the uplifted material is a function of the energy of the impact, the geothermal gradient, the amount of vertical displacement, and the volume of the uplifted material. The magnitude of heat available from the central uplifts in large craters on Mars is approximately equivalent to the amount of heat available from the impact melt (Abramov and Kring, 2005; Thorsos et al., 2001). This heating mechanism would have been more important early in Martian history when the geothermal gradient was hotter than at present (Guest and Smrekar, 2007; Zuber et al., 2000). As an example, large craters such as Gale and Gusev (at nearly 170 km in diameter) may have had as much as 20,000 km3 of impact melt, corresponding to a layer almost 900 m thick. The total heat available in these craters, assuming equal amounts from uplift and impact melt, ranges from 2 � 1022 to 2 � 1023 J (Newsom et al., 2001). In contrast, a small terrestrial crater like India’s Lonar Crater (1.8 km diameter), even though it has some limited evidence for shocked rock and impact melt (Fredriksson et al., 1973) and hydrothermal alteration beneath the crater (Hagerty and Newsom, 2003), will have only a small amount of heat available to keep a lake liquid under cold conditions.

96

Lakes on Mars

4.3.2 Hydrothermal systems The lifetime of heat sources in impact craters that is available to maintain lakes and associated hydrothermal systems depends on the size and cooling rate of the heat source and the permeability of the host rocks (Newsom et al., 1996). In small impact craters on Earth, the available heat can last for decades, as estimated for the 4-km-diameter Kärdla impact (Versh et al., 2006). Calculations of the possible duration of impact crater hydrothermal systems on Mars that can heat lakes include analytical studies (Daubar and Kring, 2001; Newsom, 1980; Newsom et al., 1996) and numerical modeling (Abramov and Kring, 2004, 2005; Ivanov and Deutsch, 1999; Rathbun and Squyres, 2002). A 30-km-diameter crater on Mars is estimated to have a lifetime of 6.7 � 104 years (Abramov and Kring, 2005), (Figure 4.4). For larger craters, Daubar and Kring (2001) estimated the lifetime of hydrothermal systems in terrestrial craters at 104–105 years for a 100-km-diameter crater and up to 106 years for a 180-km-diameter crater. For large impacts and basins, the ejecta blankets outside the transient cavity can also sustain high-temperature hydrothermal systems if water is available, as seen in the deposits at the Yaxcopoil 1 drill site in the Chicxulub impact crater (Newsom et al., 2010; Zürcher and Kring, 2007). The rate of cooling of rock, heated above ambient temperatures by volcanism or impact, is controlled by the permeability and the efficiency of convective heat transport, assuming the availability of water (Norton and Knight, 1977). Con­ ductive timescales for cooling of large impact melt sheets are thousands of years for a 100-m-thick sheet (Onorato et al., 1978). Convective systems can cool their heat sources faster than conduction, but surprisingly, geological evidence suggests that many terrestrial hydrothermal systems cool at close to the conductive rate because of low permeability (Cathles et al., 1997), a possibility that needs to be evaluated for Mars. On Earth, the sealing of hydrothermal systems by deposition of soluble minerals is commonly observed at terrestrial hot springs and mid-ocean hydrothermal systems, where horizontal fluid transport can occur for as much as 80 km (Davis et al., 1997). In Martian impact craters and volcanic areas, selfsealing by mineral deposition could also be an important mechanism in reducing the rate of convective heat losses from vapor or liquid transport. The observation of silica-rich deposits in the Columbia Hills of Mars suggests that dissolution and mineral deposition from aqueous fluids have occurred on Mars (Schmidt et al., 2008; Squyres et al., 2008; Yen et al., 2008). Theoretical models and terrestrial analog field studies can provide information on where to look for evidence for the existence of hydrothermal systems asso­ ciated with lakes in large craters on Mars. Terrestrial studies have found evidence of impact-generated hydrothermal processes in many craters (Naumov, 2005; Newsom et al., 1986; Osinski, 2005; Osinski et al., 2005; Petersen et al., 2007). Theoretical studies suggest that hydrothermal fluids concentration occurs at the edge of planar heat sources (Cathles et al., 1997) such as impact melt sheets, which is also seen in theoretical studies of Martian systems (Abramov and Kring, 2004; 2005; Ivanov and Deutsch, 1999; Rathbun and Squyres, 2002). In

Heated lakes on Mars

97 D = 30 km 5

Time = 25 years

Time = 1,000 years

5

10 Vertical distance (km)

300

8

30

250

30

250 200

200

50

50

6 150

150

4 100

100

2 0

a

b

150

5

Time = 10,000 years

150

5

Time = 100,000 years

Vertical distance (km)

10 8

30

30

50

50

100

100

200

6 4 2 0

c

d 2

4

6 8 10 12 Radial distance (km)

14

16

2

4

6 8 10 12 Radial distance (km)

14

16

Figure 4.4 Results of a numerical simulation of the hydrothermal system at a 30-km crater on early Mars (Abramov and Kring, 2005). The central peak of the crater is on the left side of each figure. Surface permeability k0 is 10–2 darcies. Solid arrows and dotted arrows indicate the water and steam fluxes, respectively. The lack of arrows in some regions indicates that fluxes are less than 2 orders of magnitude smaller than the maximum flux. Solid lines are isotherms, labeled in degrees Celsius. The length of the arrows scales logarithmically with the flux magnitude and the maximum value changes with each plot. (a) 25 years, maximum water flux = 6.66 � 10–5 kg s–1 m–2, maximum steam flux = 1.84 � 10–5 kg s–1 m–2; (b) 1000 years, maximum water flux = 1.50 � 10–5 kg s–1 m–2, max. steam flux = 2.33 � 10–7 kg s–1 m–2; (c) 10,000 years, maximum water flux = 6.01 � 10–6 kg s–1 m–2; (d) 100,000 years, maximum water flux = 8.49 � 10–7 kg s–1 m–2.

the case of a crater that penetrates a permafrost layer, the resulting hydrothermal system will be increasingly focused at the center of the structure as the system cools off and the frozen layer starts to close up (Barnhart et al., 2009). On Mars, evidence for hydrothermal processes associated with impact craters includes small valley networks on crater rims that could supply crater lakes (Brakenridge et al., 1985; Morgan and Head, 2009; Tanaka et al., 1998). The association of phyllo­ silicates with some impact crater materials, identified in CRISM data (e.g., Figure 4.5), could also be connected with hydrothermal processes (Marzo et al., 2008; Newsom, 2005; Poulet et al., 2005).

98

Lakes on Mars

–9.72

–9.72 Latitude –9.76

–91.64

–9.80 –9.84

–9.84

–9.84

–9.80

–91.72 E Longitude

–9.80

–9.80

–9.84

–91.80

–9.76

–9.76

–91.64

–9.72

–9.72

–91.76 –91.68 E Longitude

–91.72

–9.68

–9.68

–91.84

–91.80 –9.68

–91.68

–9.76

–91.76

–9.68

–91.84

Figure 4.5 CRISM visual image of a portion of the central peak of Oudemans Crater, and corresponding map showing the locations of hydroxylated silicates. Latitude: 9.8°S, 268.5° E. This impact crater is 124 km in diameter, and the CRISM spectrometer data suggest that Fe/Mg and Al phyllosilicates may be present in the outcrops of the peak. The uplifted central peaks of large impact craters could represent a heat source for impact crater lakes. On the CRISM map of hydroxylated silicates, red = BD2300 (Fe/Mg phyllosilicate), green = BD2210 (Al phyllosilicate or hydrated glass), and blue = BD1900 (hydrated sulfates, clays, glass, or water ice). The hydroxylated silicates are located in the outcrop areas of the central peak on the visual image. CRISM visual image FRT0000ABBA_07IF164S_TRU1, Image file FRT0000ABBA_07_IF164S_TRR2.LBL, CRISM image of hydrated minerals FRT0000ABBA/FRT0000ABBA_07_IF164L_HYD1. Image width �12 km.

4.3.3 Geothermal heat flux, solar energy, and the enthalpy of fusion of ice On Mars, the geothermal heat flux could have contributed energy to maintain lakes under cold conditions, especially early in Mars’ history when the solar flux was lower and the geothermal flux higher (McKay and Davis, 1991; Newsom et al., 1996). Calculations of the magnitude of the Martian heat flux are based on estimates of the amount of radioactive nuclides on Mars and constraints on the rheology of the crust (Guest and Smrekar, 2007; Schubert et al., 1992; Zuber et al., 2000). During the first half billion years of Mars’ history, the heat flow from the mantle may have been as much as five times greater (Schubert et al., 1992). Special cases of enhanced geothermal heat flux may occur as a result of impacts or volcanism. In large impacts, the uplift of the geothermal gradient can provide a source of heat, as discussed in the earlier section. Like on Earth, localized hot spots due to volcanism could also produce hot springs and hydrothermal deposits. As mentioned above, on Mars, evidence of hydrothermal activity in the Columbia Hills is inferred from the silica-rich soils observed by the Spirit rover (Schmidt et al., 2008; Squyres et al., 2008; Yen et al., 2008). Other areas of Mars also exhibit evidence of possible spring deposits and hydrothermal cone fields that could be an indication of enhanced

Heated lakes on Mars

99

heat flow allowing liquid water to create deposits at the surface (Allen and Oehler, 2008; Lanz and Saric, 2009). Solar energy also provides an external heat source for lakes or ice-covered lakes. The distance of Mars from the sun controls the solar flux, depending on the location in its annual orbit and the variations due to Milankovitch cycles (Schorghofer, 2008). Obliquity also affects the insolation and the stability of ice as a function of latitude (Mischna and Richardson, 2005), but the exact effects on lakes or ice-covered lakes has yet to be evaluated. In addition, the presence of wind-blown dust on the surface of an ice cover can also substantially reduce the effectiveness of solar heating, and heat deposited at the top of a layer of ice will be largely convected away by the atmosphere (McKay and Davis, 1991). The freezing of water at the base of an ice sheet due to the enthalpy of freezing is another source of heat that can sustain a liquid lake covered by ice in addition to the solar and geothermal heat flux (McKay et al., 1985; McKay and Davis, 1991; Newsom et al., 1996). Based on the Antarctic examples, the freezing of water at the base of the ice sheet may be coupled with ablation of the surface ice (McKay and Davis, 1991; McKay et al., 1985). If a supply of water is available from groundwater to replace the water lost by ablation, an equilibrium ice thickness on a liquid lake can be maintained (McKay and Davis, 1991). In this model, heat is only lost from the lake by conduction through the ice. Newsom et al. (1996) extended the numerical model to include the heat from impact melt sheets and to account for the difference between an equilibrium process and one in which the thickness of ice continuously increased (Figure 4.6). The models show that lakes can be maintained with an ice cover indefinitely if conditions are not too cold, and for tens of thousands of years, even under very cold conditions. The recent discoveries of widespread ice in the Martian subsurface with radar (Plaut et al., 500

Ice thickness (meters)

Ablation rate 1 cm/yr 400 300 200

Ablation rate 3 cm/yr

100

Ablation rate 10 cm/yr

0 0

5000

10000 Time (years)

15000

20000

Figure 4.6 Results of an analytical model for the thickness of ice as a function of time for a Martian crater lake containing an impact melt sheet 200 m thick (Newsom et al., 1996). The ice thickness is shown for three different ablation rates, which are assumed to be equal to the rate of formation of new ice averaged over an entire year.

100

Lakes on Mars

2010) and evidence for glacial ice and deposits in orbital images since this modeling was published (Head et al., 2006) suggest that this process may have occurred even recently in Martian history.

4.4 Discussion 4.4.1 Implications for climate The evidence for fluvial and lacustrine processes early in Martian history has been linked to a warm wet climate, primarily based on evidence for valley networks (Komatsu et al., 2009). The argument that valley networks are due only to ground­ water sapping and not surface runoff has been made less convincing as the THEMIS instrument has revealed a much more extensive record of early valley networks than previously noted from Viking orbiter data (Irwin et al., 2005, 2008). These valley networks represent good evidence for precipitation at least periodically in Mars history, suggesting a warmer climate. The existence of rhythmically layered sedimen­ tary deposits in impact craters is also best interpreted as evidence for long-term control of precipitation by climate (Lewis et al., 2008). There is now reason to believe that the existence of some lacustrine deposits could also be due to processes related to climate change connected with astronomical controls such as orbital and obliquity changes. In addition to the massive outflows of water from deep aquifers in chaotic regions, the stochastic processes of impact and volcanism may have also led to the formation of lakes and lake deposits. Even moderate-size impacts (50 km diameter) could have resulted in a temporary warm climate (Segura et al., 2002, 2008). Furthermore, the existence of heat sources due to impact, or locally enhanced geothermal heat flow associated with volcanism, could have led to the formation of lakes that left geomorphic structures and deposits that could be misinterpreted as evidence for a warmer climate. The possibility of nearsurface shallow aquifers that could supply recent gullies could also require a thermal contribution from geothermal energy (Costard et al., 2002; Mellon and Phillips, 2001). Clearly, each occurrence of lake deposits must be carefully evaluated for timing relative to associated volcanism and impact processes.

4.4.2 Implications for astrobiology and planetary protection Current research on Mars habitability is focused on identifying conditions where microbial life could evolve or flourish, as well as where to look for organic molecular evidence of prebiotic or biotic activity (Mahaffy, 2008). Heated lakes on Mars may represent habitable Petri dishes where life could have potentially evolved in situ or flourish (Cockell, 2006; Newsom, 1980; Newsom et al., 1996). Those associated with hydrothermal systems (Farmer, 2006; Varnes et al., 2003) could have provided many of the characteristics that could have favored the inception and evolution of life. These characteristics include an abundance of water, a source of nutrients, energy from hydrothermal inputs and solar flux, mineralogical substrates, and appropriate physical

Heated lakes on Mars

101

and chemical properties (Mottl et al., 2007). Lakes on Mars may have provided a habitat for organisms that evolved in another location. Raw materials for biotic and prebiotic activities and possibly living organisms could have been introduced into heated lakes from above, either within meteorites from the Earth or in impact ejecta from elsewhere on Mars (Chyba and Sagan, 1992; Nicholson, 2009). Lakes formed by impacts could have remained liquid for many thousands of years, even under cold conditions (Abramov and Kring, 2005; Newsom et al., 1996). Larger-diameter impact craters have more heat to power hydrothermal systems from impact processes, resulting in longer-lived lakes. Larger and deeper lakes will also take longer to freeze under cold conditions. By analogy to Earth, it is likely that their water supply included a connection to deep, long-lived aquifers (Clifford, 1993), which could also have contained microbial life and allowed the rapid colonization of a lake, even after the sterilization of the target rocks due to the impact (Parnell et al., 2010). In the unstable environment of Mars, they could have provided a relatively stable aqueous habitat compared to fluvial channels, which are thought to have formed over relatively short timescales (Irwin et al., 2008). A variety of mineralogical substrates that could be important for the evolution of life would also be abundant in Martian lakes and associated hydrothermal systems (Varnes et al., 2003; Ming et al., 2008). The formation of clays and other alteration minerals can provide a wide range of substrates that add to the potential for catalyzed biotic and prebiotic chemical activity (Hazen, 2005). Some inferences can be made about the possible chemistry of lake water on Mars depending on the water source and time of formation in Martian history. The discovery of phyllosilicates in ancient Martian terrains has led to the idea that the surface of Mars has steadily become more sulfur-rich and acidic through time (Bibring et al., 2006; Newsom, 2005; Poulet et al., 2005). Therefore, during the very early phyllosilicate era, the chemistry of surface water may have had a relatively neutral pH, while later in Martian history surface water would have become more acidic. However, the alteration materials in the relatively young basaltic Martian meteorites suggest that water in aquifers equili­ brated with basaltic rock in the deep crust, even at present, will have a neutral pH (McLennan and Grotzinger, 2008). In lakes dominated by neutral pH water, iron oxidation is rapid and aluminous clays can form. In contrast, based on some terrestrial lakes, the combination of thermal sources from thick impact melt sheets or igneous bodies and acidic water can result in the formation of concentrated mixtures of sulfuric and hydrochloric acids and even molten sulfur deposits (Delmelle and Bernard, 2000). Lakes are not only habitable environments, but lake-related deposits might be likely sites for the preservation of organic material. Doran and Wharton (1997) point out that high-energy fluvial or delta deposits are not ideal for the preservation of organic material. They suggest that lacustrine sand mounds similar to those found in Antarctic Dry Valley lakes may be a better place to look for preserved organics. Fine-grained lake sediments may represent good targets for organic preservation, along with chemical precipitates and evaporite minerals (Farmer and des Marais, 1999). Also, because lake deposits are surface manifestations, they are favorable for in situ investigations, such as those planned for MSL—in contrast to aqueous envir­ onments in deep aquifers or plutonic hydrothermal systems that are not accessible with current technology (Newsom et al., 1996, 2001).

102

Lakes on Mars

The possible existence of heated lakes on Mars at the present time or in the near future raises questions about planetary protection. The possibility of contamination by terrestrial organisms of a habitable “special region,” such as a heated lake on Mars, is of great concern to the planetary protection community (COSPAR, 2005; NASA, 2005; Rummel, 2009) and has led to several studies including the PREVCOM report (Chyba et al., 2006) and the SR-SAG report (Beaty et al., 2006) to evaluate the risks. Because the thermodynamic equilibrium conditions at the surface of Mars are cur­ rently too cold or too dry to sustain life (Beaty et al., 2006; Schorghofer et al., 2002; Tosca et al., 2008), the problem becomes one of identifying locations where local heating through volcanism or impact for example has occurred or might occur in the near future. The PREVCOM report (Chyba et al., 2006) concluded that the existence of favorable environments for microbial propagation almost anywhere on Mars, now or in the future, could not be ruled out. They argued that habitable environments could form due to a disequilibrium condition, for example, by a landslide or impact. Furthermore, they argued that habitable environments could also form anywhere at some distant future time, for example, when climate change due to obliquity changes might occur. However, the SR-SAG report concluded that extremely recent volcanic terrains and large impact craters that could produce a lake are characterized by distinctive geologic structures easily observable in any proposed landing site with currently available imagery (Beaty et al., 2006), and the potential for a large-enough impact to be of concern over the next hundred years was negligible. For example, no craters larger than 30 km in diameter that are young and fresh enough to retain the heat needed to exceed the temperature threshold for propagation have been identified on Mars to date (Barlow, 2003).

4.5 Conclusions Lakes can form on Mars from groundwater and surface runoff into basins formed by impacts or volcano-tectonic processes (Irwin et al., 2005). Large volcanic centers on Mars such as Tharsis and Elysium likely produced long-lived hydrothermal systems that would have had a major effect on driving groundwater circulation in the surrounding regions and possibly producing lakes and possibly oceans (see Baker et al., 1991). The formation of large impact craters can also result in hydrothermal systems that can lead to lakes (Abramov and Kring, 2005; Newsom, 1980; Newsom et al., 1996). Essentially all lakes that could occur on Mars can be heated by solar energy or the heat flow from the typical geothermal gradient, but these sources of energy are small. Greater heating can result from the freezing of bodies of water and from the transient heat in large impacts or volcanic fields. If the average temperature is below freezing, this heat can contribute to the maintenance of a permanently liquid, near-surface reservoir or extend the lifetime of a temporary lake created by a transient event such as an impact or release from a deep aquifer. The possible heat sources include the following: solar heat flux; enthalpy of fusion of ice during freezing of a lake, with continued groundwater recharge; normal geothermal heat flux; anomalous geothermal

Heated lakes on Mars

103

heat flux due to volcanism; uplifted warm basement in large impacts; and impact melt and shock-heated basement in large impacts. Many of these heat sources are often coupled. For example, the ablation of ice on the surface of a lake, combined with freezing at the base of the ice sheet and an influx of groundwater, can permanently maintain a liquid lake under an ice sheet as observed in Antarctica. Large impact craters involve melt sheets and uplifted warm basement. Even under cold climatic conditions, the heat sources associated with a large impact (>30 km diameter) could sustain a liquid lake for tens of thousands of years. The presence of extensive fluvial and lacustrine activity early in Martian history has long been interpreted as evidence for a warm climate. However, because of the different heating mechanisms discussed here, evidence for a lake or lake deposits does not necessarily imply a warm climate and surface runoff (Morgan and Head, 2009). Heated lakes may represent ideal sites to search for evidence of biotic or prebiotic organic processes on Mars. Lakes formed by these processes that still exist, if any, could potentially represent areas of concern from a planetary protection standpoint, but geological investigations should be able to identify any currently active sites with the available high-resolution imagery.

Acknowledgements This work was supported by NASA Planetary Geology and Geophysics Program grant NNX 08AL74G (PI, H. Newsom). Comments by Livio Tornabene on an early draft and reviews by two anonymous reviewers are appreciated. Nathalie Cabrol provided extensive editorial help throughout the preparation of the paper.

References Abramov, O., Kring, D.A., 2004. Numerical modeling of an impact-induced hydrothermal system at the Sudbury Crater. J. Geophys. Res. 109 (10), 16. doi:10.1029/2003JE002213. Abramov, O., Kring, D.A., 2005. Impact-induced hydrothermal activity on early Mars. J. Geophys. Res. 110 (20). doi:10.1029/2005je002453. Allen, C.C., Oehler, D.Z., 2008. A case for ancient springs in Arabia Terra, Mars. Astrobiology 8, 1093–1112. doi:10.1089/ast.2008.0239. Baker, V.R., 1982. The Channels of Mars. University of Texas Press, Austin, TX. 204 p. Baker, V.R., Strom, R.G., Gulick, V.C., Kargel, J.S., Komatsu, G., Kale, V.S., 1991. Ancient oceans, ice sheets and the hydrological cycle on Mars. Nature 352 (6336), 589–594. Barlow, N., 2003. Revision of the “catalog of large martian impact craters”, Proceedings of the Sixth International Conference on Mars, Abstract 3073. Barnhart, C.J., Nimmo, F., Travis, B.J., 2009. Geochemical and geomorphological effects of post-impact hydrothermal systems incorporating freezing, Proceedings of the Workshop on Modeling Martian Hydrous Environments, 13–14. Beaty, D., Buxbaum, K., Meyer, M., Barlow, N., Boynton, W., Clark, B., et al., 2006. Findings of the Mars special regions science analysis group. Astrobiology 6 (5), 677–732. doi:10.1089/ast.2006.6.677.

104

Lakes on Mars

Bell III, J.F., Glotch, T.D., Hamilton, V.E., McConnochie, T., McCord, T., McEwen, A., et al., 2008. Visible to near-IR multispectral orbital observations of Mars. In: Bell III J.F. (ed.), The Martian Surface – Composition, Mineralogy, and Physical Properties. Cambridge University Press, Cambridge, MA, pp. 169–192. Benison, K.C., Bowen, B.B., 2006. Acid saline lake systems give clues about past environments and the search for life on Mars. Icarus 183, 225–229. Bibring, J.P., Squyres, S.W., Arvidson, R.E., 2006. Merging views on Mars. Science 313 (5795), 1899–1901. doi:10.1126/science.1132311. Brakenridge, G.R., Newsom, H.E., Baker, V.R., 1985. Ancient hot springs on Mars; origins and paleoenvironment significance of small Martian valleys. Geology 13 (12), 859–862. doi:10.1130/0091-7613. Cabrol, N.A., Grin, E.A., 1999. Distribution, classification, and ages of martian impact crater lakes. Icarus 142 (1), 160–172. Cabrol, N.A. and Grin, E.A., 2001. The Evolution of Lacustrine Environments on Mars: Is Mars Only Hydrologically Dormant? Icarus 149, 291–328. Cabrol, N.A., Grin, E.A., 2003. Overview on the formation of paleolakes and ponds on Mars. Global Planet Change 35(3–4), 199–219. doi:10.1016/S0921-8181. Cabrol, N.A. and Grin, E.A., 2005. Ancient and Recent Lakes on Mars, In Water on Mars and Life, (e.d. T. Tokano) Springer, pp. 235–260. Cabrol, N.A., Grin, E.A., Newsom, H.E., Landheim, R., McKay, C.P., 1999. Hydrogeologic evolution of Gale Crater and its relevance to the exobiological exploration of Mars. Icarus 139 (2), 235–245. doi:10.1006/icar.1999.6099. Cabrol, N.A., Grin, E.A., Kiss, K.T., Ács, E., Grigorszky, I., Szabò, K., et al., 2007. Signatures of habitats and life in earth's high-altitude lakes: Clues to noachian aqueous environments on Mars. In: Mary Chapman, (ed.), The Geology of Mars. Chapter 14: Evidence from Earth-Based Analogs. Cambridge University Press, Cambridge, MA, pp. 349–370. Cabrol, N.A., Wynn-Williams, D.D., Crawford, D.A., and Grin, E.A., 2001. Recent aqueous environments in Martian impact craters; an astrobiological perspective. Icarus 154 (1), 98–112. doi:10.1006/icar.2001.6661. Carr, M.H., 1996. Water on Mars. Oxford University Press, New York. Cathles, L.M., Erendi, A.H.J., Barrie, T., Stein, H.J., Cathles, L.M., 1997. How long can a hydrothermal system be sustained by a single intrusive event? Econ. Geol. 92 (7–8), 766–771. Chyba, C., Sagan, C., 1992. Endogenous production, exogenous delivery and impact-shock synthesis of organic-molecules – an inventory for the origins of life. Nature 355 (6356), 125–132. doi:10.1038/355125a0. Chyba, C., Clifford, S., Delamere, A., Favero, M.S., Mathur, E.J., Niehoff, J.C., et al., 2006. Preventing the forward Contamination of Mars, Space Studies Board, National Research Council, The National Academies Press, Washington, DC. Clifford, S.M., 1993. A model for the hydrologic and climatic behavior of water on Mars. J. Geophys. Res. 98 (E6), 10973–911016. doi:10.1089/ast.2008.0759. Cockell, C.S., 2006. The origin and emergence of life under impact bombardment. Philos. Trans. R. Soc. B.Biol. Sci. 361 (1474), 1845–1855. doi:10.1098/rstb.2006.1908. Cockell C.S, Gronstal, A.L., Voytek, M.A., Kirshtein, J.D., Finster, K., Sanford, W.E., et al., 2009. Microbial abundance in the deep subsurface of the Chesapeake Bay impact crater: Relationship to lithology and impact processes. Geolo. Soc Am. Special Papers 458, 941–950. COSPAR, 2005. Cospar Planetary Protection Policy (20 October 2002; amended 24 March 2005), edited.

Heated lakes on Mars

105

Costard, F., Forget, F., Mangold, N., and Peulvast, J.P., 2002. Formation of recent Martian debris flows by melting of near-surface ground ice at high obliquity. Science 295 (5552), 110–113. doi:10.1126/science.1066698. Daubar, I.J., Kring, D.A., 2001. Impact-induced hydrothermal systems; heat sources and life­ times. Lunar Planet. Sci., XXXII, Abstract 1727. Davis, E.E., Fisher, A.T., Firth, J.V., Andersson, E.M., Aoike, K., Becker, K., et al., 1997. Introduction and summary; Hydrothermal circulation in the oceanic crust and its conse­ quences on the eastern flank of the Juan de Fuca Ridge, Proceedings of the Ocean Drilling Program, Part A: Initial Reports, 168, 7–21. Delmelle, P., Bernard, A., 2000. Volcanic lakes. In: Sigurdsson, H.H., Bruce, F., McNutt, S.R., Rymer, H., Stix, J., (Eds.), Encyclopedia of Volcanoes. Academic Press, New York, pp. 877–895, Des Marais, D.J., Borg, L.E., Beaty, D.W., 2008. Science priorities for Mars Sample Return. Astrobiology 8 (3) 489–535. Di Achille, G., Hynek, B.M., Searls, M.L., 2009. Positive identification of lake strandlines in Shalbatana Vallis, Mars. Geophys. Res. Lett. 36 (14), L14201. doi:10.1029/ 2009GL038854. Di Achille, G., Ori, G.G., Reiss, D., 2007. Evidence for late Hesperian lacustrine activity in Shalbatana Vallis, Mars. J. Geophys. Res. 112 (E7), 7007. doi:DOI 10.1029/ 2006JE002858. Doran, P.T., Wharton, Jr. R.A., 1997. Mars on Earth; Antarctic dry valley lakes and their deposits, Proceeding of the workshop on Early Mars: Geologic and Hydrologic Evolution, Physical and Chemical Environments, and the Implications for Life, LPI Contribution, Report: 916, abstract 24. Ehlmann, B.L., Mustard, J.F., Fassett, C.I., Schon, S.C., Head III. J.W., Des Marais, D.J., et al., 2008a. Clay minerals in delta deposits and organic preservation potential on Mars. Nature Geosci. 1 (6), 355. doi:10.1038/ngeo207. Farmer, J.D., 2006. Subsurface hydrothermal systems as havens for Martian life. Geochim. Cosmochim. Acta 70 (18), A166–A166. doi:10.1016/j.gca.2006.06.334. Farmer, J.D., Des Marais, D.J., 1994. The search for carbonates on Mars. Special Publication – Karst Waters Institute 1, 16–17. Farmer, J.D., Des Marais, D.J., 1999. Exploring for a record of ancient Martian life. J. Geophys. Res. 104 (E11), 26977–26995. Fassett, C.I., Head III. J.W., 2005. Fluvial sedimentary deposits on Mars: Ancient deltas in a crater lake in the Nili Fossae region. Geophys. Res. Lett. 32. doi:10.1029/ 2005GL023456. Fredriksson, K., Dube, A., Milton, D.J., Balasundaram, M.S., 1973. Lonar Lake, India; an impact crater in basalt. Science 180 (4088), 862–864. doi:10.1126/science.180.4088.862. Goldspiel, J.M., Squyres, S.W., 1991. Ancient aqueous sedimentation on Mars. Icarus 89 (2), 392–410. doi:10.1016/0019-1035(91)90186-W. Goldspiel, J.M., Squyres, S.W., 2000. Groundwater sapping and valley formation on Mars. Icarus 148 (1), 176–192. doi:10.1006/icar.2000.6465. Guest, A., Smrekar, S.E., 2007. New constraints on the thermal and volatile evolution of Mars. Phys. Earth Planet. Inter. 164 (3–4), 161–176. doi:10.1016/j.pepi.2007.06.010. Gulick, V.C., 1998. Magmatic intrusions and a hydrothermal origin for fluvial valleys on Mars. J. Geophys. Res. 103 (E8), 19,365–319,387 Gulick, V.C., Tyler, D., McKay, C.P., Haberle, R.M., 1997. Episodic ocean-induced CO2 greenhouse on Mars: Implications for fluvial valley formation. Icarus 130 (1), 68–86.

106

Lakes on Mars

Hagerty, J.J., Newsom, H.E., 2003. Hydrothermal alteration at the Lonar Lake impact structure, India; implications for impact cratering on Mars. Meteorit. Planet Sci. 38 (3), 365–381. Harrison, K.P., Grimm, R.E., 2008. Multiple flooding events in Martian outflow channels. J. Geophys. Res. 113 (E2), E2002. doi:10.1029/2007JE002951. Hazen, R.M., 2005. Genesis: Rocks, minerals, and the geochemical origin of life. Elements 1 (3), 135–137. doi:10.2113/gselements.1.3.135. Head, J.W., Marchantb, D.R., Agnew, M.C., Fassett, C.I., Kreslavsky, M.A., 2006. Exten­ sive valley glacier deposits in the northern mid-latitudes of Mars: Evidence for Late Amazonian obliquity-driven climate change. Earth Planet. Sci. Lett. 241, 663–671, doi:10.1016/j.epsl.2005.11.016. Hovius, N., Lea-Cox, A., Turowski, J.M., 2008. Recent volcano-ice interaction and outburst flooding in a Mars polar cap re-entrant. Icarus 197 (1), 24–38. doi:10.1016/j. icarus.2008.04.020. Irwin, R.P., Howard, A.D., Craddock, R.A., 2008. Fluvial valley networks on Mars. In: Rice, S., Roy, A.G., and Rhoads, B.L. (Eds.), River Confluences, Tributaries and the Fluvial Net­ work. John Wiley & sons, New York. Irwin, R.P., III, Howard, A.D., Craddock, R.A., Moore, J.M., 2005. An intense terminal epoch of widespread fluvial activity on early Mars; 2, Increased runoff and paleolake develop­ ment. J. Geophys. Res. 110 (E12), E12S15. doi:10.1029/2005je002460. Ivanov, B.A., Deutsch, A., 1999. Sudbury impact event; cratering mechanics and thermal history. Special Paper – Geol. Soc. Am. 339, 389–397. Jakosky, B.M., Haberle, R.M., Arvidson, R.E., 2005. The changing picture of volatiles and climate on Mars. Science 310 (5753), 1439–1440. doi:10.1126/science.1118031. Kieffer, S.W., Simonds, C.H., 1980. The role of volatiles and lithology in the impact cratering process. Rev. Geophys. Sp. Phys. 18 (1), 143–181. doi:10.1016/j.icarus.2004.04.005. Komatsu, G., Di Achille, G., Popa, C., Di Lorenzo, S., Rossi, A.P., Rodriguez, J.A.P., 2009. Paleolakes, paleofloods, and depressions in Aurorae and ophir Plana, Mars: Connectivity of surface and subsurface hydrological systems. Icarus 201 (2), 474–491. doi:10.1016/j. icarus. 2009.01.010. Lanz, J.K., Saric, M.B., 2009. Cone fields in SW Elysium Planitia: Hydrothermal venting on Mars? J. Geophys. Res. 114, 13. doi:E02008 10.1029/2008je003209. Laskar, J., Correia, A.C.M., Gastineau, M., Joutel, F., Levrard, B., Robutel, P., 2004. Long term evolution and chaotic diffusion of the insolation quantities of Mars. Icarus 170 (2), 343–364. Lasue, J., Clifford, S., 2009. The Impact of Astronomically Induced Insolation Variations on the Extent of the Martian Cryosphere. Proceedings of the Workshop on Modeling Martian Hydrous Environments, 13–14. Lewis, K.W., Aharonson, O., Grotzinger, J.P., Kirk, R.L., McEwen, A.S., Suer, T.-A., 2008. Quasi-periodic bedding in the sedimentary rock record of Mars. Science 322 (5907), 1532–1535. doi:10.1126/science.1161870. Mahaffy, P., 2008. Exploration of the habitability of Mars: Development of analytical protocols for measurement of organic carbon on the 2009 Mars science laboratory. Space Sci. Rev. 135 (1–4), 255–268. doi:10.1007/s11214-007-9223-1. Malin, M.C., Edgett, K.S., 2000. Sedimentary rocks of early Mars. Science 290 (5498), 1927–1937. doi:10.1126/science.290.5498.1927. Marzo, G.A., Davila, A.F., Fairen, A.G., Roush, T.L., Bishop, J.L., Dohm, J.M., et al., 2008. Evidence for relatively recent hydrothermal activity due to an impact within the syrtis major. Eos Trans. A.G.U. 89 (53), P53A–1438.

Heated lakes on Mars

107

McCarville, P., Crossey, L.J., 1996. Post-impact hydrothermal alteration of the Manson impact structure. Special Paper – Geol. Soc. Am. 302, 347–376. McKay, C.P., 1997. The search for life on Mars. Orig. Life Evol. Biosph. 27, 1–3, 263–289. McKay, C.P., Clow, G.D., Wharton, Jr., R.A., Squyres, S.W., 1985. Thickness of ice on perennially frozen lakes. Nature 313 (6003), 561–562. doi:10.1038/313561a0. McKay, C.P., Davis, W.L., 1991. Duration of liquid water habitats on early Mars. Icarus 90 (2), 214–221. doi:10.1016/0019-1035(91)90102-Y. McLennan, S.M., Grotzinger, J.P., 2008. The sedimentary rock cycle of Mars. In: Bell III, J.F. (Ed.), The Martian Surface – Composition, Mineralogy, and Physical Properties. Cam­ bridge Univer City Press, Cambridge. Mellon, M.T., Phillips, R.J., 2001. Recent gullies on Mars and the source of liquid water. J. Geophys. Res. 106, 1–15. doi:10.1029/2000je001424. Ming, D.W., Morris, R.V., Clark, B.C., 2008. Aqueous alteration on Mars. In: Bell III, J.F., (ed.), The Martian Surface – Composition, Mineralogy, and Physical Properties. Cambridge Press, Cambridge, pp. 519–540. Mischna, M.A., Richardson, M.I., 2005. A reanalysis of water abundances in the Martian atmosphere at high obliquity. Geophys. Res. Lett. 32 (3), 4. doi:10.1029/2004gl021865. Mormile, M.R., Hong, B.-y., Benison, K.C., 2009. Molecular analysis of the microbial com­ munities of Mars analog lakes in Western Australia. Astrobiology 9, 919–930. Moore, J.M., and Howard, A.D., 2005. Large alluvial fans on Mars, J. Geophys. Res. 110, E04005, doi:10.1029/2004JE002352. Moore, J.M., Howard, A.D., Dietrich, W.E., Schenk, P.M., 2003. Martian layered fluvial deposits; implications for Noachian climate scenarios. Geophys. Res. Lett. 30 (24), 19. doi:10.1029/2003GL019002. Morgan, G.A., Head, J.W., 2009. Sinton crater, Mars: Evidence for impact into a plateau icefield and melting to produce valley networks at the Hesperian-Amazonian boundary. Icarus 202 (1), 39–59. doi:10.1016/j.icarus.2009.02.025. Mottl, M.J., Glazer, B.T., Kaiser, R.I., Meech, K.J., 2007. Water and astrobiology. Chem. ErdeGeochem. 67(4), 253–282. doi:10.1016/j.chemer.2007.09.002. Mouginis-Mark, P., 2007. Geologic mapping of the Martian impact crater Tooting, Open File Report U S Geological Survey, Report: OF. National Aeronautics and Space Administration, 2005. Planetary protection provisions for robotic extraterrestrial missions, National Aeronautics and Space Administration, USA. Naumov, M.V., 2005. Principal features of impact-generated hydrothermal circulation systems: mineralogical and geochemical evidence. Geofluids 5 (3), 165–184. doi:10.1111/j.1468­ 8123.2005.00092.x. Newsom, H.E., 1980. Hydrothermal alteration of impact melt sheets with implications for Mars. Icarus 44 (1), 207–216. Newsom, H.E., 2005. Clays in the history of Mars. Nature 438 (7068), 570–571. doi:10.1038/ 438570a. Newsom, H.E., Barber, C.A., Hare, T.M., Schelble, R.T., Sutherland, V.A., Feldman, W.C., 2003. Paleolakes and impact basins in southern Arabia Terra, including Meridiani Planum; implications for the formation of hematite deposits on Mars. J. Geophys. Res. 108 (E12), 8075. doi:10.1029/2002JE001993. Newsom, H.E., Brittelle, G.E., Hibbitts, C.A., Crossey, L.J., Kudo, A.M., 1996. Impact crater lakes on Mars. J. Geophys. Res. 101 (E6), 14951–914955. Newsom, H.E., Graup, G., Sewards, T., Keil, K., 1986. Fluidization and hydrothermal alteration of the suevite deposit at the Ries Crater, West-Germany, and implications for Mars. J. Geophys. Res. 91 (B13), E239–E251.

108

Lakes on Mars

Newsom, H.E., Hagerty, J.J., Thorsos, I.E., 2001. Location and sampling of aqueous and hydrothermal deposits in Martian impact craters. Astrobiology 1 (1), 71–88. doi:10.1089/153110701750137459. Newsom, H.E., Lanza, N.L., Ollila, A.M., Wiseman, S.M., Roush, T.L., Marzo, G.A., et al., 2009. Inverted channel deposits on the floor of Miyamoto crater, Mars. Icarus 205, 64–72 Newsom, H.E., Salge, T., Nelson, M.J., Spilde, M.N., 2010. Discovery of andradite garnet and evidence for high temperature hydrothermal processes (>300°C) in the lower Yaxcopoil-1 impact-melt breccias. Lunar Planet. Sci. XLI, Abstract 1751. Nicholson, W.L., 2009. Ancient micronauts: interplanetary transport of microbes by cosmic impacts. Trends Microbiol. 17 (6), 243–250. doi:10.1016/j.tim.2009.03.004. Norton, D., Knight, J.E., 1977. Transport phenomena in hydrothermal systems; cooling plutons. Am. J. Sci. 277 (8), 937. O'Keefe, J.D., Ahrens, T.J., 1994. Impact-induced melting of planetary surfaces. Special Paper – Geol. Soc. Am. 293, 103–109. O'Keefe, J.D., Ahrens, T.J., 1999. Complex craters; relationship of stratigraphy and rings to impact conditions. J. Geophys. Res. 104 (E11), 27091–027104. doi:10.1029/ 1998je000596. Onorato, P.I.K., Uhlmann, D.R., Simonds, C.H., 1978. The thermal history of the Manicouagan impact melt sheet, Quebec. J. Geophys. Res. 83 (B6), 2789–2798. doi:10.1029/ JB083iB06p02789. Ori, G.G., di Achille, G., Pondrelli, M., 2009. Deltas on Mars. Lunar Planet. Sci. XL, Abstract 1579. Osinski, G.R., 2005. Hydrothermal activity associated with the Ries impact event, Germany. Geofluids 5 (3), 202–220. doi:10.1111/j.1468-8123.2005.00119. Osinski, G.R., Lee, P., Parnell, J., Spray, J.G., Baron, M.T., 2005. A case study of impactinduced hydrothermal activity: The Haughton impact structure, Devon Island, Canadian high arctic. Meteorit. Planet Sci. 40 (12), 1859–1877. Pain, C.F., Clarke, J.D.A., Thomas, M., 2007. Inversion of relief on Mars. Icarus 190 (2), 478–491. Parnell, J., Boyce, A., Thackerey, S., Muirhead, D., Lindgren, P., Mason, C., et al., 2010. Sulfur isotope signatures for rapid colonization of an impact crater by thermophilic microbes. Geology 38, 271–274. doi:10.1130/G30615.1. Petersen, M.T., Newsom, H.E., Nelson, M.J., Moore, D.M., 2007. Hydrothermal alteration in the Bosumtwi impact structure: Evidence from 2M(1)-muscovite, alteration veins, and fracture fillings. Meteorit. Planet Sci. 42 (4–5), 655–666. Pierazzo, E., Artemieva, N.A., and Ivanov, B.A., 2005. Starting conditions for hydrothermal systems underneath Martian craters; hydrocode modeling. Special Paper – Geol. Soc. Am. 384, 443–457. Plaut, J.J., Holt, J.W., Head, J.W. III, Gim, Y., Choudhary, P., Baker, D.M., et al., 2010. Thick Ice Deposits in Deuteronilus Mensae, Mars: Regional Distribution from Radar Sounding. Lunar Planet. Sci. XLI, Abstract 2454. Pondrelli, M., Rossi, A.P., Marinangeli, L., Hauber, E., Gwinner, K., Baliva, A., et al., 2008. Evolution and depositional environments of the Eberswalde fan delta, Mars. Icarus 197 (2), 429–451. doi:10.1016/j.icarus.2008.05.018. Pope, K.O., Kieffer, S.W., Ames, D.E., 2006. Impact melt sheet formation on Mars and its implication for hydrothermal systems and exobiology. Icarus 183 (1), 1–9. doi:10.1016/j. icarus.2006.01.012. Poulet, F., Bibring, J.P., Mustard, J.F., Gendrin, A., Mangold, N., Langevin, Y., et al., 2005. Phyllosilicates on Mars and implications for early Martian climate. Nature 438, 623–627. doi:10.1038/nature04274.

Heated lakes on Mars

109

Rathbun, J.A., Squyres, S.W., 2002. Hydrothermal systems associated with martian impact craters. Icarus 157 (2), 362–372. doi:10.1006/icar.2002.6838. Rummel, J.D., 2009. Special regions in Mars exploration: Problems and potential. Acta Astro­ naut. 64 (11–12), 1293–1297. doi:10.1016/j.actaastro.2009.01.006. Schmidt, M.E., Ruff, S.W., McCoy, T.J., Farrand, W.H., Johnson, J.R., Gellert, R., 2008. Hydrothermal origin of halogens at Home Plate, Gusev Crater. J. Geophys. Res. 113, E06S12, doi:10.1029/2007JE003027. Schorghofer, N., 2008. Temperature response of Mars to Milankovitch cycles. Geophys. Res. Lett. 35 (18), 4. doi:L18201 10.1029/2008gl034954. Schorghofer, N., Aharonson, O., Khatiwala, S., 2002. Slope streaks on Mars; correlations with surface properties and the potential role of water. Geophys. Res. Lett. 29 (23), 4. doi:10.1029/2002gl015889. Schubert, G., Solomon, S.C., Turcotte, D.L., Drake, M.J., Sleep, N.H., 1992. Origin and thermal evolution of Mars. In: Kieffer, H.H., et al. (Eds.), Mars. University of Arizona, Tucson, pp. 147–183 Schwenzer, S.P., Abramov, O., Allen, C.C., Clifford, S., Filiberto, J., Kring, D.A., et al., 2010. Exploring martian impact craters: what they can reveal about the subsurface and why they are important in the search for life. Lunar Planet. Sci. XLI, Abstract 1589. Segura, T.L., Toon, O.B., Colaprete, A., 2008. Modeling the environmental effects of moderate-sized impacts on Mars. J. Geophys. Res. 113 (E11), E11007. doi:10.1029/ 2008JE003147. Segura, T.L., Toon, O.B., Colaprete, A., Zahnle, K., 2002. Environmental effects of large impacts on Mars. Science 298 (5600), 1977–1980. doi:10.1126/science.1073586. Squyres, S.W., Arvidson, R.E., Ruff, S.W., Gellert, R., Morris, R.V., Ming, D.W., et al., 2008. Detection of silica-rich deposits on Mars. Science 320 (5879), 1063. doi:10.1126/ science.1155429. Stewart, S.T., O'Keefe, J.D., Ahrens, T.J., 2001. Impact cratering on ice-rich materials. Abstracts with Programs – GSA 33 (3), 66. Tanaka, K.L., Dohm, J.M., Lias, J.H., Hare, T.M., Clifford, S., Treiman, A., et al., 1998. Erosional valleys in the Thaumasia region of Mars; hydrothermal and seismic origins. J. Geophys. Res. 103 (E13), 31,407–431,41910.1016/S0032-0633(98)00141-X. Thorsos, I.E., Newsom, H.E., Davies, A.G., 2001. Availability of heat to drive hydrothermal systems in large Martian impact craters. Lunar Planet. Sci. XXXII, Abstract 2011. Tornabene, L.L., McEwen, A.S., Grant, J.A., Mouginis-Mark, P.J., Squyres, S.W., Wray, J.J., 2007. Evidence for the role of volatiles on Martian impact craters as revealed by HiRISE. Lunar Planet. Sci. XXXVIII, Abstract 2215. Tosca, N.J., Knoll, A.H., McLennan, S.M., 2008, Water activity and the challenge for life on early Mars. Science 320 (5880), 1204–1207. doi:10.1126/science.1155432. Varnes, E.S., Jakosky, B.M., McCollom, T.M., 2003. Biological potential of martian hydro­ thermal systems. Astrobiology 3 (2), 407–414. doi:10.1089/153110703769016479. Versh, E., Kirsimae, K., Joeleht, A., 2006. Development of potential ecological niches in impact-induced hydrothermal systems: The small-to-medium size impacts. Planet. Sp. Sci. 54 (15), 1567–1574. doi:10.1016/j.pss.2004.12.022. Yen, A.S., Morris, R.V., Clark, B.C., Gellert, R., Knudson, A.T., Squyres, S., et al., 2008. Hydrothermal processes at Gusev Crater: An evaluation of Paso Robles class soils. J. Geophys. Res. 113, E06S10, doi:10.1029/2007JE002978. Zahnle, K., Haberle, R.M., Catling, D.C., Kasting, J.F., 2008. Photochemical instability of the ancient Martian atmosphere. J. Geophys. Res. 113 (E11), 16. doi:E11004 10.1029/ 2008je003160.

110

Lakes on Mars

Zuber, M.T., Solomon, S.C., Phillips, R.J., Smith, D.E., Tyler, G.L., Aharonson, O., et al., 2000. Internal structure and early thermal evolution of Mars from Mars Global Surveyor topo­ graphy and gravity. Science 287 (5459), 1788–1793. doi:10.1126/science.287.5459.1788. Zürcher, L., Kring, D.A., 2004. Hydrothermal alteration in the core of the Yaxcopoil-1 borehole, Chicxulub impact structure, Mexico. Meteorit. Planet. Sci. 39 (7), 1199–1222.

5 Lakes in Valles Marineris Baerbel K. Lucchitta U. S. Geological Survey, 2255 N. Gemini Dr. Flagstaff, AZ 86001, USA

5.1

Introduction

The Valles Marineris is a large system of troughs extending just south of the Martian equator from about longitude 250° E to 320° E. They trend approximately N 75° W for a distance of about 4000 km, equivalent to nearly the entire width of the United States. In the west, the system includes the Noctis Labyrinthus network of troughs, in the center the linear depressions of Ius and Coprates Chasmata as well as the widened section containing Melas, Candor, and Ophir Chasmata, and in the east Eos/Capri and Ganges Chasmata. The latter merge with chaotic terrains and outflow channels (Figure 5.1). The troughs are interconnected and open toward the east except for the northern troughs, Echus and Juventae Chasmata, which open toward the north, and Hebes Chasma, which is entirely enclosed. Individual troughs are generally on the order of 50–100 km wide, but in the central part of multiple troughs the depression is about 600 km wide. Depths below the plateau surface of as much as 8 km are common, and locally depths of as much as 11 km are reached (Lucchitta et al., 1994; Peulvast et al., 2001). Mounds of interior layered deposits (ILD) occur within the troughs in many places (Figure 5.2). The troughs are locally paralleled by shallow grabens on the adjacent plateaus and by aligned pits, commonly called chain craters. The main alignments of the Valles Marineris and the grabens lie on a fracture system radial to the Tharsis/Syria rises (Anderson et al., 2001; Blasius et al., 1977; Carr, 1974; Dohm et al., 2008; Plescia and Saunders, 1982). The entire trough system was named Valles Marineris (Mariner valleys) in honor of the Mariner 9 mission, whereas some of the individual troughs were named after the classical dark markings, for instance Tithonius Lacus and Coprates. Coincidentally, the region called Lacus, meaning lake, is located where we now think lakes may indeed have existed.

5.2

The Mariner Era

When Mariner 4 in 1964 and Mariner 6/7 in 1969 flew by Mars, the news was disappointing. Returned images revealed a cratered landscape resembling that of the Moon and Mercury, except the craters were shallower; it appeared that the solar system was largely occupied by cratered bodies. However, Mariner 6 showed some intriguing landforms and caught a glimpse of disturbed terrain that did not fit the general pattern (Figure 5.3). Lakes on Mars. DOI: 10.1016/B978-0-444-52854-4.00005-2 © 2010 Published by Elsevier B.V.

Figure 5.1 Index map to the troughs and the location of figures. Unless otherwise noted, north is up in the images. Base is MOLA DEM shaded relief.

Figure 5.2 Geologic map of Valles Marineris. Modified after Lucchitta et al. (1994). Dark blue: well-defined stacks of ILD. Light blue: less welldefined and more irregular exposures of ILD. Green: chaotic material. Orange: landslides. Red: dark material. Greenish grey: wall material. Brown: wall-rock material on floor. Yellow, gray, greenish brown: various floor materials, which may contain small patches of ILD. Light purple: craters. Black: surrounding plateaus.

114

Lakes on Mars

Figure 5.3 Mariner 6 image 6n14, giving the first glimpse of a chaos on Mars.

Only in 1972, after Mariner 9 had orbited Mars for some time, did the planet reveal itself as a body resembling Earth in many ways. The equatorial troughs of the Valles Marineris and the chaotic terrains to its east were discovered and found to contain landslides, debris flows, ravines, and gullies resembling those formed by running water. McCauley et al. (1972) inferred artesian sapping and headward erosion for tributary canyons to the troughs, collapse due to withdrawal of material for the chaotic terrains, and drainage of loose material into the subsurface for the crater chains. They considered an origin of the troughs by collapse with erosional widening due to the action of water or ground ice or perhaps tectonic and magmatic processes. They did not see the interior layered deposits in the early pictures and no mention was made of lakes. Sharp (1973), in the first thorough study of the Valles Marineris, preferred the term troughs rather than canyons, because it did not imply fluvial processes in their formation. Arc-shaped reentrants were recognized as landslides with debris, anasto­ mosing ridges on the walls as created by dry mass movement or perhaps fluid erosion, and structurally controlled tributary canyons as formed by ground-ice sapping com­ bined with headward erosion. Because of the blunt ends of some troughs, Sharp (1973) preferred erosional widening and massive ground-ice collapse rather than rifting for the origin of the troughs. However, he was aware of the difficulties posed by such enormous volumes of segregated ice. Masson (1977), in contrast, proposed the formation of the troughs as an extensional graben system similar to the African rift valleys, followed by erosional widening. Sharp (1973) was the first to describe the ILD in Ganges and west Candor Chasmata (Figure 5.4). He recognized that they were different from layers in the trough walls, but he did not mention an origin as lake beds. Malin (1976) men­ tioned lakes but preferred that the layering was part of the ancient crust of Mars,

Lakes in Valles Marineris

115

Figure 5.4 Mariner 9 image DAS 09017619, showing Gangis Mensa. This Mensa and one in west Candor Chasma were the first ILD seen. Sharp (1973) commented on the rills, conjecturing that they might be caused by fluid seepage.

which was faulted and eroded to form the Valles Marineris. Thus, the ancient layers became exposed and exhumed from the walls. A detailed geologic map of the Coprates NW quadrangle (McCauley, 1978) led to the first thorough discussion of the hypothesis that the ILD are lacustrine sediments. The apparent horizontality, lateral continuity, and competence of individual layers suggested an origin as water-laid sediments in a low-energy environment. McCauley (1978) speculated that the lakes formed from melted ground ice that collected in depressions created by tectonic subsidence. The presence of ground ice inside the Valles Marineris was also suspected by Lucchitta (1978a) because lobes on wall rock and in tributary canyons resembled Antarctic mud flows and rock glaciers (Figure 5.5). Hanging valleys suggested ice-related processes and chain craters may have formed by collapse along structures by melting of ground ice due to increased heatflow. On the other hand, Lucchitta suggested that some curvilinear walls on the Valles Marineris were not necessarily due to erosional widening, but may have followed ancient crater scars.

5.3

The Viking Era

Viking image resolution covering the troughs is about 200–300 m/pixel for Capri/ Eos Chasmata and 30–80 m/pixel in most other places (Komatsu et al., 1993). Thus, the images revealed unprecedented detail on trough walls, trough floors, and the ILD.

116

Lakes on Mars

Figure 5.5 Comparison of Martian and terrestrial debris flows. From Lucchitta (1978). (a) Wright Valley, Antarctica. Vertical aerial photograph by the US Navy, January 1956. (b) South wall of Tithonium Chasma, north toward bottom. Mariner 9 frame DAS 10204959-B. Note similarity in chutes at head of flows (1), raised levees on lower parts of flows (2), and overlapping shallow lobes of deposits at base of flows (3).

5.3.1 The troughs The milestone paper by Blasius et al. (1977) strongly supported the hypothesis that the Valles Marineris are a large graben system. They identified cratered surfaces on the floor as down-dropped segments of plateau rock and recognized remarkably straight scarps truncating spurs and gullies near the canyon floor, making triangular facets (Figure 5.6). Adjustment of fault blocks under the influence of first north–south and then east–west extension would have formed the graben. If the sides of the blocks were near vertical, they would not require much extension to subside. Larger north– south tension would require strike slip faults at the blunt ends (Spencer and Fanale, 1990), for which no evidence is found.

Lakes in Valles Marineris

117

Figure 5.6 Fault scarp (arrows) at base of north wall of Coprates Chasma. Scarp cuts spurs and gullies of wall and forms triangular facets. Viking Orbiter image 608A54; �250 km wide. From Lucchitta et al. (1992).

Fractures associated with the Tharsis and Syria rises were generally thought to control the planimetric form of the Valles Marineris (Blasius et al., 1977; Carr, 1974; Plescia and Saunders, 1982), but details for graben formation varied. Frey (1979) compared the Valles Marineris to the African rift system and concluded that the thick Martian crust would have produced simpler fault trends and longer individual scarp lengths, such as seen in the Valles Marineris. A perceived bulge along the length of the Valles Marineris lead to the hypothesis that the graben formed as a keystone collapse at the crest of the bulge (Wise et al., 1979). Crater counts on the floor of Coprates Chasma agree with those on the Lunae Planum plateau and western Ophir Planum, confirming downward displacement by normal faulting (Schultz, 1991), and initiation of faulting and subsidence of the central troughs was considered early Late Hesperian by Schultz (1991). Early on, the walls eroded into spur-and-gully morphology, which make up 85–95% of the walls (Peulvast et al., 2001). The walls were thought to have volcanic layers near the top (Davis and Golombek, 1990; Lucchitta, 1978a; Scott and Carr, 1978; Scott and Tanaka, 1986;) underlain by massive brecciated units from the early, heavy bombardment on Mars (Carr, 1979, 1986). Komatsu et al. (1993) suggested that the deposition of the ILD was roughly contemporaneous with canyon enlargement by wall collapse and mass wasting, which closely followed the original graben formation. The walls were then cut by landslide scars that are devoid of spurs and gullies (Figure 5.7) (Lucchitta, 1978b, 1979). The changing erosional pattern may have been due to climate change or a change in physical environment, such as

118

Lakes on Mars

Figure 5.7 Landslides in Ophir Chasma (ls). Slide scars form three smooth-walled, talus-covered arcuate reentrants that are cut into wall rock (w) having spur-and-gully morphology. Note that landslide debris aprons spilled around the ILD mounds in foreground (il). Viking Orbiter image mosaic 13A09-12, by A. McEwen, �150 km wide. Vertical exaggeration about �2.

from a wet climate or subaqueous lake environment to dry troughs (Lucchitta, 1984; Lucchitta et al., 1992; Shaller 1991). The morphogenetic change on the walls from spurs and gullies to smooth talus in landslide reentrants may also be attributed to the loss of early ground ice at the Hesperian/Amazonian boundary; an early ice content in wall rock is confirmed by old rampart craters on the adjacent plateaus (Peulvast et al., 2001). However, some evidence points to the presence of later water or ice in the walls. For instance, rounded heads on Vshaped tributary canyons suggest sapping by water or ground ice (Sharp, 1973; Kochel and Piper, 1986; Laity and Malin, 1985). Tributary canyons occur in large numbers on the south wall of Ius Chasma but also as more solitary canyons elsewhere (Figure 5.8). They cut spur-and-gully morphology and some smooth talus slopes, but their age is mainly unknown. More evidence for water or ice in wall rock comes from Lucchitta (1987a), who observed features on late landslide deposits that implied mobility consistent with high water content. She found possible channels emerging from landslide debris and a debris flow that traveled 250 km from the source on a low gradient negotiating several bends; the debris included evidence for having contained blocks of ice, such as hills topped by depressions, reminiscent of glacial kettle holes on Earth. High ground-ice content is also supported by the observation that narrow ridges, apparently more resistant to erosion than the country rock, project into the troughs along the extension of faults and grabens (Figure 5.9) (Lucchitta, 1987a). These fault zones could have been depleted in ice due to higher heatflow, or they could have been more lithified than the surroundings owing to hydrothermal activity or other circulations of water (Lucchitta, 1987a).

Lakes in Valles Marineris

119

Figure 5.8 Tributary canyons on south wall of Ius Chasma. Canyons apparently developed along preexisting joint patterns. The blunt ends are commonly attributed to sapping processes, but small valleys above tributary heads (below image resolution) can be recognized even on Viking images. North toward upper right. Viking Orbiter image 645A59; �300 km wide.

Figure 5.9 Wall-rock spur separating Ophir Chasma above from Candor Chasma below. Spur is apparently more resistant to erosion than surroundings in spite of being topped by a graben system. Mars Express image HRSC h03334_0001, ID 1105; �120 km wide.

120

Lakes on Mars

5.3.2 The interior layered deposits Description, setting, and composition Viking images showed that the ILD are widely distributed in the trough system, but are most prevalent in the wider sections (Figure 5.2). Most of them occur in high mounds isolated from the walls (e.g., Ganges Mensa, Figure 5.4) or in benches lapping up against walls (Figure 5.10). Nedell et al. (1987) calculated that the volume of the ILD in the central trough is 1.29�105 km³. Lucchitta et al. (1994) found that the ILD are as thick as 9 km in the mesas and benches of Hebes, Ophir, and Candor Chasmata, where they may rise to less than 1 km below the rim (Chapman and Tanaka, 2001) but are much thinner in south Melas and Capri/Eos Chasmata. The void volume of the troughs is about six times that of the deposits, showing that, overall, the ILD are only minor components of the troughs (Lucchitta et al., 1994). Hebes, Ganges, and Juventae Chasmata have isolated ILD mounds that may have formed under closed-canyon conditions (Komatsu et al., 1993). The stratigraphy of the Ophir and Candor deposits are similar enough that they could be erosional remnants of a once larger deposit (Figure 5.10b). Ophir, Candor, Hebes, and locally Melas Chasmata have an upper layered unit and a lower light-toned, more massive fluted unit (Komatsu et al., 1993; Lucchitta et al., 1992; Lucchitta, 1999; Nedell et al., 1987). The fluting suggests that the light materials are highly susceptible to wind erosion (Lucchitta et al., 1992). Even though the horizontality and lateral continuity of the ILD were emphasized early on (Blasius et al., 1977; Komatsu et al., 1993; McCauley, 1978; Nedell et al., 1987; Peterson, 1981) they differ considerably in color, albedo, thickness, and competence, and Nedell et al. (1987) noticed that not all layers are horizontal. They proposed that the inclined layers may be depositional wedges or structurally tilted. In addition, Lucchitta (1990) found angular unconfor­ mities and disconformities within the ILD. A vexing problem has been the observation that many ILD occur as free-standing mesas and mounds, isolated from the walls by deep troughs or “moats” (Figure 5.10). Nedell et al. (1987) proposed that the moats formed after deposition of the ILD stacks, primarily by collapse and recession of the canyon walls due to removal of ground ice. Komatsu et al. (1993) considered eolian or water erosion, but conjectured that erosion by water was less likely because not much water would have been left for the erosion after the troughs were filled with sediments to near the brim. Lucchitta et al. (1992) proposed that the moats locally could be down-dropped subsidiary troughs (Lucchitta and Bertolini, 1990; Lucchitta et al., 1994). ILD occurring on the floors of the chasmata could be such down-dropped deposits (Lucchitta, 1990, 1999). ILD are generally older than landslides (Figure 5.7) but in west Candor Chasma low-lying ILD overlap land­ slides and are thus younger; they could represent a later stage of ILD emplacement, after the high ILD mounds were already partly eroded (Lucchitta, 1990). Early spectral investigations were based on three-point spectra of Viking color images, convolved to laboratory spectra of terrestrial rocks at the same spectral resolution. Geissler et al. (1993) noted that the ILD are relatively red and among the brightest materials in the troughs. They also detected a distinctly redder unit in west Candor Chasma on the flanks of Ceti Mensa, consistent with a higher abundance of

Lakes in Valles Marineris

121

Figure 5.10 ILD mesas. Note that mesas are free-standing mounds separated from wall rock by “moats.” (a) Candor Mensa in center, part of Baetis Mensa on right. Mars Express image HRSC h3334_0001, ID 1105. (b) Baetis Mensa in center, straddling Ophir Chasma in north and Candor Chasma in south. Note that the ILD unconformably overlap eroded remnants of wall rock, at arrow. Viking Orbiter image mosaic 13A09-12, by A. McEwen. (c) Hebes Mensa in the entirely enclosed Hebes Chasma. Excerpt of Viking-based MDIM 2.1. (d) The southern Mensa in Juventae Chasma. Note conspicuous, even layering in upper part. Looking east. Mars Express image HRSC h03334_0001, ID 1105.

bulk crystalline hematite in this region. The infrared spectrometer (ISM) instrument on Phobos II also indicated a local enrichment in ferric oxides or oxyhydroxides in this area perhaps due to secondary alteration from subaerial weathering of soils,

122

Lakes on Mars

hydrothermal alteration, or conversion of other previously formed ferric oxide miner­ als to hematite through the application of heat (Komatsu et al., 1993). As the ferric unit in west Candor Chasma occurred in topographic lows, Komatsu et al. (1993) inferred concentration of weathered out material. All of these observations were confirmed later (see Section 5.3). The ISM instrument also suggested that bright deposits within the troughs are hydrated (Mustard and Murchie, 2001) and possibly palagonitic (Erard et al., 1991; Murchie et al., 2000). Furthermore, Murchie et al. (1992) noted that the ILD of Ophir, Candor, and Melas Chasmata varied in composi­ tion and that the ILD differed from wall rock or the surrounding plateaus in pyroxene absorption features. The dark cap rocks on ILD mounds in Melas and Eos Chasmata seemed to be mafic with different amounts of incorporated water and ferric iron. One early idea was that the ILD were composed of carbonates (Croft, 1989b; McKay and Nedell, 1988). McEwen and Soderblom (1989) found bright layers in ILD in Ganges and Capri Chasmata, whose spectra permitted possible carbonate composi­ tion. Spencer and Fanale (1990) recognized that this process likely would have occurred only early in Martian history, necessitating that the ILD were ancient and exhumed later (Malin, 1976). However, using Mariner 6/7 infrared spectrometer data sets covering Ganges Chasma, McKay and Nedell (1988) found no evidence for carbonates or clay minerals in that area.

Origin of ILD in lakes Most researchers thought that the ILD were deposited in lakes (e.g., McCauley, 1978; Nedell et al., 1987) with water supplied by substantial subsurface aquifers (Carr, 1979; Sharp, 1973), subterranean piping (Croft, 1989a; Spencer and Croft, 1986), or sapping of groundwater in tributary canyons (Figure 5.8). McCauley (1978) suggested that shallow parallel grabens widened, deepened, and coalesced into troughs harboring lakes, which eventually overflowed to form the outflow channels. The detection of hydrated minerals on the ILD also supported sedimentation in lakes (Erard et al., 1991; Murchie et al., 1992). The most frequently cited issue was that no large channels empty into the central troughs. Therefore, Komatsu et al. (1993) proposed that sediments could have come off the walls, a viable proposition as the volume of eroded material from the walls is larger than that of the deposits inside (Lucchitta et al., 1994). Carr (1986) proposed that wall rock may have liquefied and mixed with a high proportion of clays, thus making the ILD easily erodible. In addition, lakes should have been capped quickly by an ice cover (Carr, 1983) and frozen solid to the depth of the permafrost layer, thus complicating the influx of sediments. The setting and age of purported lakes, however, varied widely. Blasius et al. (1977) thought that the ILD formed by widespread, apparently cyclical sedimentation within unconnected troughs hundreds of kilometers apart. They suggested that the cyclical sedimentation could have been caused by oscillations in Mars’ obliquity and orbital eccentricity cycles (Ward, 1974), perhaps modulating sedimentation of wind­ blown materials. Spencer and Fanale (1990) proposed that the layered deposits were emplaced in lakes that formed in ancient grabens and that they were exhumed later, similar to the idea of Malin (1976), who thought that the layered deposits were

Lakes in Valles Marineris

123

emplaced in crater lakes of Noachian age and exhumed after the chasmata formed. But many researchers noted that exhumation is not likely because the ILD are clearly superposed on eroded wall rock and have vastly different morphological character­ istics from the walls (Figure 5.10b) (Blasius et al., 1977; Chapman, 2002; Komatsu et al., 1993; Lucchitta et al., 1992; Nedell et al., 1987). Blasius et al. (1977) indeed inferred that substantial sedimentation postdated episodes of tectonism and erosion that produced the spurs and gullies on the trough walls. Another hypothesis was that ancient enclosed basins of Hesperian age, such as Hebes Chasma (Figure 5.10c), existed in the central Valles Marineris as well and were precursers to the current troughs. As the ILD water level at its maximum would have stood at a much higher elevation in the central troughs than in the peripheral troughs, water would have spilled out of them (Komatsu et al., 1993; Lucchitta et al., 1994; Schultz, 1998). Therefore Lucchitta and Bertolini (1990) proposed that ancestral basins were filled with ILD and later breached by renewed faulting. The large through-going grabens of Ius and Coprates Chasmata (Figure 5.1) may have been emplaced by such later faulting (Komatsu et al., 1993; Lucchitta et al., 1992, 1994; Schultz, 1998). Chapman and Smellie (2007) also noted that ILD are more common in elliptical-shaped troughs, whereas linear chasmata are noticeably more devoid of the deposits, suggesting a causal relationship. If the down-dropping of the ancestral basin floors were swift, deep troughs with spur-and-gully walls would have eventually filled with water and sediments. Alternatively, down dropping may have been concurrent with sedimentation, as seen in the fault troughs of the Basin and Range Province on Earth (Komatsu et al., 1993; Lucchitta, 1982; Nedell et al., 1987) and lakes would have remained shallow or were playas. If so, upon breaching of the ancestral basins, little water would have been left to spill out and form the outflow channels. The emplacement of ILD in the peripheral troughs may have been in enclosed basins as well (Komatsu et al., 1993; Lucchitta and Ferguson, 1983). Chaotic mounds on the floor of these troughs probably formed first. When the ancestral basins collapsed, water was released inside them and then spilled onto the adjacent plateaus, leaving channel markings in places (Komatsu et al., 1993; Lucchitta and Ferguson, 1983; Tanaka and Golombek, 1989; Witbeck et al., 1991) similar to those left by the later Elaver Vallis, which flowed out of a pit (Figure 5.11). The ILD, which mostly rest on top of the chaotic mounds (Komatsu et al., 1993; Lucchitta et al., 1992; Nedell et al., 1987; Witbeck et al., 1991), were deposited next inside these lakes. Subsequently, the ancestral basins were breached by headward erosion (Lucchitta and Ferguson, 1983) or structural adjustments, resulting in the current opening toward the outflow channels.

Origin by air fall An eolian or pyroclastic air fall origin with ash from Tharsis was considered by several authors, but generally dismissed as not yielding the required thickness of deposits. Also, similar thicknesses of eolian materials are not observed on the adjacent plateaus (Komatsu et al., 1993; Lucchitta et al., 1992; Peterson, 1981). In addition, the diversity of layers suggested that sedimentation by wind was not the dominant process, as wind deposits from global dust storms were thought to produce the same stratigraphy all over

124

Lakes on Mars

Figure 5.11 Channel markings on top of plateau south of Ganges Chasma. Water emerged from Ganges Cavus (GC), flowed through the nearby crater, over the plateau, and eventually emptied into Ganges Chasma. Ganges Chaos (GCh). ILD in Ganges Mensa (GM) overlaps chaos hills at small, white arrows. Excerpt of THEMIS day IR mosaic.

the troughs (Komatsu et al., 1993). If the ILD were eolian or ash fall deposits, the sediments had to be trapped. Nedell et al. (1987) proposed a trapping mechanism whereby sediments could have been deposited on an ice cover and foundered or that ice rose around the sediment mounds as diapirs because of Raleigh–Taylor instabilities.

Volcanic origin A volcanic origin for the ILD is still being debated because of dark layers within the ILD mounds and dark patches or caps on their surfaces (Komatsu et al., 1993; Lucchitta, 1990). If the mounds were subaerial volcanoes, no lakes would be needed inside the Valles Marineris. Lucchitta et al. (1994) proposed that the lower massive units within the ILD might be from mass wasting off the walls, and the upper, more resistant units might be volcanic. Because of the lack of source vents, Nedell et al. (1987) proposed that volcanic eruptions might have been in a lake, sub-ice, and deposits may have spread widely and covered vents. Croft (1990) thought that the ILD mound in Hebes (Figure 5.10c) might be a volcano emplaced into ice forming a table mountain (Van Bemmelen and Rutten, 1955). Lucchitta et al. (1994) suggested that the ILD mound in Ganges Chasma (Figures 5.4, 5.11) had table mountain characteristics, such as a massive basal

Lakes in Valles Marineris

125

unit possibly emplaced under ice, overlain by thinner and more resistant units, emplaced subaerially. Howard (1991) proposed that purported volcanoes in the chaotic terrain may have been intruded into ice laccolith that had formed in that area like gigantic open system pingoes derived from confined aquifers. Hydrated layers within the ILD (Geissler et al., 1993; Murchie et al., 1992) are also consistent with volcanic rocks erupted into lakes or hydrated by hydrothermal fluids. Dark patches lining the base of trough walls, associated with faults and locally overlying landslide deposits (Fig. 5.12), small depression-topped hills, and possible calderas, were thought to be young and of volcanic origin (Lucchitta, 1987b, 1990). Similar-looking dark material in the walls of the Valles Marineris has spectra most compatible with mafic volcanic glass (Geissler et al., 1990). Some very dark, raised spots on the east flank of Baetis Mensa could be necks that penetrated the ILD and are

Figure 5.12 Dark material lining base of north wall of Ius Chasma. Note that dark material (d) crops out along chasma-wall fault (f), which is also expressed within the landslide material. Dark patches (d) occur both within lower wall and on top of landslide debris. (li) Light mottled material is commonly associated with dark material. (r) Rugged, (g) grooved landslide. (l) Layered, (t) smooth-talus wall rock. Layered wall rock is edge of plateau lavas. Viking Orbiter image 66A15. From Lucchitta (1990).

126

Lakes on Mars

thus contemporaneous with or later than the ILD, supporting a volcanic origin (Lucchitta, 1990). If the ILD are volcanic, the preponderance of evidence points toward sub-ice or sub-water emplacement, followed by late stage subaerial mafic activity.

5.4 The MGS Era After MGS arrived at Mars in 1996, MOC resolution (2–10 m/pixel) showed unpre­ cedented detail of the Martian surface, which, in combination with TES and MOLA, led to major new discoveries.

5.4.1 The troughs McEwen et al. (1999) and Malin and Edgett (2000) recognized that the walls of the Valles Marineris were layered throughout, potentially from early volcanic flood lavas (Figure 5.12). This observation contrasted with earlier ones that recognized layers only near the top (Lucchitta, 1978a; Scott and Carr, 1978; Scott and Tanaka, 1986). The extensive volcanism throughout Mars’ first billion years could have maintained a thick atmosphere, making the presence of water in the Valles Marineris more likely (McEwen et al., 1999).

5.4.2 Layers planet-wide Malin and Edgett (2000) recognized fine-grained, bedded, indurated sediments in ancient craters, in the chaotic terrains, in the Valles Marineris, and in Meridiani Planum. Except for those in Meridiani, all were thought to have been emplaced by fluvial processes, which transported sediments to Noachian lakes in craters or craterlike depressions (Cabrol and Grin, 1999, 2001). The ILD in the Valles Marineris were thus thought to be erosional remnants of ancient layers, now exhumed from the walls. Malin and Edgett (2000) recognized layers in the Candor Chasma walls, supposedly emerging from these walls (Figure 5.13). However, Lucchitta (1999, 2001) interpreted the same outcrops as ILD plastered against the walls. In addition, Weitz et al. (2001) noted that the ILD overlap wall rock, and Schultz (2002) found the strength of wall rock to be consistent with igneous rock and that of the ILD consistent with weaker sandstone or siltstone. All these observations supported that the ILD are interior to the troughs, corroborating the earlier Viking observations. From TES data at about 3 km spatial resolution Christensen et al. (2001) recog­ nized an easily erodible basaltic sedimentary unit with about 10–15% crystalline gray hematite in Meridiani Planum, Aram Chaos, and scattered throughout the Valles Marineris (Figure 5.14), where the hematite-rich material is located in topographic lows and found only in association with the ILD. They confirmed red hematite in west Candor Chasma (Geissler et al., 1993) as possibly crystalline FeOOH polymorphs, including goethite. The authors settled on an origin as low-temperature precipitation of Fe-rich water in a lacustrine environment and subsequent alteration to gray

Lakes in Valles Marineris

127

Figure 5.13 ILD cropping out on chasma wall. From Lucchitta (2001). Wall on north side of central Candor Chasma. Wall-rock spur (dark arrows) is flanked by light outcrops of ILD (light arrows). The layers appear deposited on top of the wall because layer edges (white dashes) and the wall have the same slope. Furthermore, the ILD do not crop out in the spur. Also note the contrasting erosional morphologies of spur and ILD. MOC M17-00467.

hematite or perhaps direct precipitation of gray hematite from Fe-rich circulating fluids. They inferred that liquid water must have been stable at or near the surface probably for millions of years.

5.4.3 Lacustrine origin Weitz et al. (2001) favored a lacustrine origin of the ILD, deposited in individual, separate basins, because fine layering a few meters thick extends over several kilo­ meters distance and the morphologic properties are similar over great distances; yet the properties are not uniform for sequences in different chasmata. Also, all ILD top out at a level below that of the plateau surface. Many supposedly dark layers within the ILD were now recognized as benches onto which dark material was deposited. Weitz et al. (2001) also recognized possible fluvial channels dissecting the ILD mounds, perhaps eroded when the inferred lakes drained.

128

Lakes on Mars

3° S

50 km

285° E

8° S 290° E

Figure 5.14 Distribution of gray crystalline hematite in Candor and Ophir Chasmata. Warmer colors have a higher hematite index. Hematite-rich materials are associated with dark material located in topographic lows near the ILD. From Christensen et al. (2001). Copyright 2001 American Geophysical Union. Reproduced by permission of American Geophysical Union.

ILD occur not only in high mounds but also on the chasma floors. These outcrops may be down-faulted (Lucchitta, 1999), deeply eroded equivalents of ILD in the high mounds (Lucchitta, 1999) or, in west Candor Chasma, later deposits surrounding the mounds (Lucchitta, 1990). Another light-toned floor deposit, consisting of rounded blocks with upturned margins, occurs on the floor of western Melas and eastern Ius Chasmata (Figure 5.15). Dark dunes and other eolian deposits occur on top and in the inter-block areas. Skilling et al. (2002) and Weitz et al. (2003) interpreted the deposit as a landslide. Komatsu and Di Cencio (2002) noted the arc-shaped or concentric pattern within the rounded blocks and suggested that the deposit might be rich in evaporites. The recognition of late, low-lying light-toned deposits indicated that the ILD have a much more complex composition and history than previously thought.

5.4.4 Volcanic origin Detailed investigations of MOC images showed that many of the layers within ILD mounds were inclined (Figure 5.16). The “massive” unit (Komatsu et al., 1993;

Lakes in Valles Marineris

129

Figure 5.15 Rounded, blocky deposit on Melas Chasma floor. Note upturned edges and invasion of dunes into cracks and onto the deposit. MOC E13-01996.

Lucchitta, 1999; Lucchitta et al., 1992) was now recognized as composed of fine, dipslope layers with triangular facets (Figure 5.16) locally radiating from the ILD mounds (Beyer and McEwen, 2005b; Lucchitta and Chapman, 2002). Furthermore, the ILD showed unconformities (Weitz et al., 2001), possible cross-beds (Chapman and Tanaka, 2001), plunging anticlines and synclines in west Candor Chasma (Figure 5.17) (Malin and Edgett, 2001), and flow lobes (Lucchitta, 1990; Lucchitta and Chapman, 2002). These observations lead to a revival of the idea that the ILD might be volcanic constructs (Figure 5.18). Chapman and Tanaka (2001) and Chapman and Smellie (2007) proposed that subice volcanoes called tuyas could best explain the morphology of isolated ILD mounds in most chasmata, because of mound heights, flat-topped mesas, horizontal to steep dips, fine-grained materials, moats, rare possible volcanic vents and lava flows, spectral composition, and tuff-like weathering (Lucchitta, 1990). The lower massive layers of the ILD might be hyalotuffs and hyaloclastites, the tilted material flow collapse into water-forming lava deltas, and the dark flat layers on top subaerial lavas. With TES data, Christensen et al. (1998) indeed found such dark layers in Hebes Chasma to be mafic, consistent with Ca-rich clinopyroxene and in situ volcanic origin. Komatsu et al. (2004), comparing the ILD to tuyas of the Baikal rift system in Siberia, thought that the locally extensive layers of a few meters in thickness were favored by a low energy environment, consistent with slow settling in a lacustrine basin, perhaps distal facies of turbitidites derived from volcanic sheet flows. Palago­ nites in the ILD (Erard et al., 1991; Murchie et al., 2000) might have formed from

130

Lakes on Mars

Figure 5.16 Tilted layers on ILD. (a) ILD on south side of a mound in central Candor Chasma. White arrow points toward an exposure of tilted beds, elsewhere obscured by the weathered surface. Black arrows point toward triangular facets pointing uphill, indicating that bedding is tilted down toward the observer. This unit was mapped as “massive” on Viking images (Komatsu et al., 1993; Lucchitta, 1999). MOC ab105205. (b) Northern reentrant on Ceti Mensa in west Candor Chasma, looking southeast. Top layer extends from top right to bottom left over 20 km. Vertical exaggeration �1.5. THEMIS V11275002, draped over HRSC DEM.

volcanic heat and an ample supply of water. In Juventae Chasma, elongated ILD mounds could be tindars (sub-ice volcanic ridges), emplaced due to mantle plume activity, which also produced water driven out of the chasma as jökulhaups (Chapman et al., 2003). Here the water would have had to rise high enough to surmount the 3-km barrier that separates the Juventae trough from the Maja outflow channel. The jökulhaup idea was supported by the recognition of mega ripples in Maja Valles that strikingly resemble those found in Iceland. To form deep lakes elsewhere,

Lakes in Valles Marineris

131

Figure 5.17 Folds and knobs in southwestern west Candor Chasma. Plunging anticline at top, plunging syncline at bottom. Knobs (arrows) may reflect cementation by circulating fluids. MOC M04-01762.

Figure 5.18 Mound that could be a sub-ice volcano, in southwest central Candor Chasma. Example of tilted bedding outlined by dashes, angular unconformity shown by dots, separating inclined layers from dark cap. The mound is 4000 m high; picture width 18 km, looking southwest. Composite of Viking image mosaic and MOC images, draped on MOLA topography. Thin black horizontal lines are artifacts. Vertical exaggeration �1.5. From Lucchitta (2004).

132

Lakes on Mars

Chapman and Tanaka (2001) envisioned cold-based ice-plug dams near the exit of the peripheral troughs. Chapman and Tanaka (2001) ruled out a structural origin for the ILD mounds because no such structures are seen to extend into adjacent wall rock.

5.5 Odyssey and beyond A flurry of recent missions to Mars vastly improved our understanding of the red planet. High-resolution images from THEMIS in visible (VIS) and near infrared (NIR), MEx High Resolution Stereo Camera (HRSC), MRO’s HiRISE, and CTX cameras were obtained at lower sun illumination angles than images from previous missions and partly with stereographic viewing. The hyperspectral imagers OMEGA and CRISM gave brand-new insights into the mineralogic composition of the surface. GRS also provided information on the water content near the surface. In addition, the MER rovers Spirit and Opportunity gave us ground truth to an extent never before accomplished.

5.5.1 The troughs Beyer and McEwen (2005a) found that the trough-wall sequences of resistant basaltic layers (McEwen et al., 1999) are interrupted by hundreds of meters of mostly taluscovered weaker layers, perhaps composed of fluvial, eolian, or volcanic sedimentary material, of thin lava layers, welded tuff, or a combination of all of these. Some resistant layers may be of intrusive origin (Williams et al., 2003). Crater counts on the floor of the troughs gave a maximum age of 3.5 Ga (Quantin et al., 2004a) and counts on 56 individual landslides gave an age range of 3.5 Ga–50 Ma (Figure 5.19). Thus, the onset of sliding nearly coincided with the opening of the troughs. The similar morphology of all the landslides, regardless of their age, suggests that the dynamics of their emplacement has not changed within the last 3 Ga or so. Consequently, because the troughs were dry within the last 50 Ma, including the age of the last landslides, they were likely dry since close to their inception. McEwen (1989) had also suggested that the Martian landslides were subaerial because their mobility was one order of magnitude less than that of sub­ marine landslides on Earth. Furthermore, as the ancient trough floors are visible in several places, not much erosion or deposition has occurred in those places since trough inception (Quantin et al., 2004a). Most ILD predate the landslides, leading Quantin et al. (2004a) to propose the following sequence: formation of the troughs, deposition of the ILD in an aqueous environment most likely early in the trough’s history, loss of the water, emplacement of the landslides, and more recently only minor deposition and erosion mostly due to eolian activity. Quantin et al. (2004a), however, caution that locally ILD may have formed later than the earliest landslides. Olivine-enriched lava flows in benches of similar elevations, eroded out on the floors of Ganges and Eos Chasmata (Figure 5.20) (Christensen et al., 2003; Edwards et al., 2008), also support the idea that the troughs were dry soon after inception because olivine is unstable in aqueous conditions (Christensen et al., 2003).

Figure 5.19 Ages of landslides (squares) and trough floors (ovals). Small black numbers refer to listing of landslides in Table 1 of Quantin et al. (2004a). After Quantin et al. (2004a).

134

Lakes on Mars

Interpreted outcrop locations

313° E 2.5° N a

331.5° E

b

c

12.5° S

Figure 5.20 Olivine outcrops in Eos Chasma. (a) Interpreted outcrop locations (red) identified from the THEMIS multispectral 100 m spatial sampling mosaic. (b,c) Decorrelation stretch mosaics (bands 8, 7, and 5) showing details of outcrops. From Edwards et al. (2008). Copyright 2008 American Geophysical Union. Reproduced by permission of American Geophysical Union.

A quantitative study of the spur-and-gully walls of Coprates Chasma by Jernslet­ ten (2004) based on MOLA measurements and solar exposure also supports rela­ tively early dry conditions. The study showed a lack of asymmetry in the spur-and­ gully walls, suggesting that no ground ice was present in these walls since the time of their formation. Because all landslide scars are younger than the spurs and gullies, and the oldest landslides date back to the opening of the troughs around 3.5 Ga ago (Quantin et al., 2004a), Jernsletten (2004) contended that the spurs and gullies could have formed early enough to have developed in a subaqueous environment. In contrast, Quantin et al. (2004b) suggested that deeper parts of the rocks behind the walls may have been partly filled with water even during later landslide time. In a quantitative study of 45 landslides within the Valles Marineris, they found that the volume of the landslides on the floors was much less than that of material removed from the walls, leading them to propose that large porosities, probably filled by water, existed within the walls. A karstic or thermokarstic origin for Hebes and Ophir Chasmata, where the volume deficit is largest, would support the idea that the ancestral troughs formed by collapse. Water within trough walls would also support the hypothesis that the landslides were emplaced as wet debris flows (Lucchitta, 1987a; Quantin et al., 2004b).

Lakes in Valles Marineris

135

5.5.2 Valleys, channels, and lakes Even though some valleys were seen on Viking images above tributary canyons (Lucchitta and Ferguson, 1983), a proliferation of valleys on plateaus surrounding the Valles Marineris is now recognized. Valleys near Echus Chasma are incised into a weak, dark unit, perhaps ash or eolian materials, overlying the lavas of the Late Hesperian plateau; the valleys are buried at the north end by Early Amazonian lava flows, indicating late warm conditions (Figure 5.21) (Mangold et al., 2004, 2008a). The valleys are mature and connect to the heads of tributary canyons and thus the troughs, implying that lakes in the Valles Marineris may have formed from surface runoff. Mangold et al. (2008a) speculated that perhaps the early wet period extended well into the Hesperian, but that valley networks are not recognized on Hesperian fractured lavas, into which water readily infiltrated, whereas valleys on the abundant clays in the Noachian highlands (Bibring et al., 2005) were better preserved. Valleys with inverted relief on the plateau near Juventae Chasma (LeDeit et al., 2008a; Lucchitta, 2005; Weitz et al., 2008a, 2010) are associated with a finely layered light-toned deposit that is shown in Figure 5.22. Using HiRISE, CTX, and CRISM data, Bishop et al. (2008a, 2009) found a composition of hydrated silica for this deposit and inferred a hydrothermal origin. Milliken et al. (2008a, 2008b) compared these deposits to opal, chalcedony, and/or altered volcanic or impact glass; they also found some clay minerals or clay precursors. They interpreted the deposits to be from alteration of in situ or fluvially transported ash deposits. Weitz et al.(2008a, 2010) also

Figure 5.21 Small valley networks on plateau west of Echus Chasma. Note valleys emptying into chasma in shadow at lower right and into head of tributary canyon at bottom. CTX P04_002472_1812, illumination from left.

136

Lakes on Mars

Figure 5.22 Finely layered deposit on plateau west of Juventae Chasma. Illumination from lower left. MOC R09-03652.

preferred a fluvial origin, perhaps of reworked pyroclastic or altered basaltic material, for this as well as several other similar light-toned layers on the plateaus adjacent to the Valles Marineris. Weitz et al. (2008a, 2010) emphasized that the layers on top of the lava plateaus are morphologically and compositionally different from the ILD inside the troughs. They also recognized that some valleys occurred on through walls, indicating fluvial activity post trough opening. Highly organized valley networks are also found inside the chasmata on top of low-lying light-toned deposits in a marginal trough on the southwest side of Melas Chasma (Figure 5.23) (Mangold et al., 2004; Pelkey et al., 2003; Quantin et al., 2005). Outflow channels occur on the high plateaus near the eastern end of the troughs (Figure 5.11) (Coleman et al., 2007a). Water erupted from Ophir Cavus, Ganges Cavus, and related pit chains, flowed over the adjacent plateaus, and discharged into an ancestral Ganges Chasma, as did water near Eos and Capri Chasmata from a purported lake in eastern Coprates Chasma (Coleman et al., 2007b). At Ophir Cavus, an eruption level near an elevation of 2500 m above Martian datum suggested that the groundwater may have come from the Tharsis plateau and was channeled along highly permeable basalt fractures into the troughs (Harrison and Grimm, 2004), eventually spilling out of the cavi near the east end of the troughs into permanent or ephemeral ice-covered lakes during a volcanic-hydrologic climax in Hesperian time (Coleman et al., 2007a). Spillover channels at levels far above the bottom of the adjacent troughs were also recognized in Juventae Chasma (Catling et al., 2006; Chapman et al., 2003; Coleman and Baker, 2007), suggesting that water within Juventae once stood as much as 500 m above the current floor (Catling et al., 2006). Substantial lakes likely existed in the central and peripheral troughs to facilitate massive through flow and outflow of water into these high-level outflow channels.

Lakes in Valles Marineris

137

Figure 5.23 Valleys in marginal trough on south rim of west Melas Chasma. Channels (ch) flow toward lower right, into putative lake at right edge of image (Quantin et al., 2005). Also note dissected fans (f) and blocky deposits on trough floor (b, detail shown in Fig. 5.15), CTX P08_004252_1714.

The ILD mostly overlap chaos hills (Figure 5.11) that are associated with the peripheral troughs and early- or late-stage outflow (Harrison and Grimm, 2008). Therefore, the emplacement of ILD postdates the initial formation of the cavi, some pit chains, and the peripheral troughs and associated outflow channels. However, in a few places, such as Juventae Chasma (Chapman et al., 2003; Catling et al., 2006) and Aureum Chaos (Glotch and Rogers, 2007), light-toned deposits are found within chaos hills, suggesting that locally the chaos formation bracketed the emplacement of such deposits. Evidence supporting low-lying lakes inside the troughs has increased recently. Valleys of southwestern Melas Chasma empty into a well-defined depression, inter­ preted to have been a lake, 200 m deep (Figure 5.23) (Quantin et al., 2005). Some of these valleys also emptied into the region of rounded blocks (Figures 5.15 and 5.23, see also Section 5.4.3) (Pelkey et al., 2003; Weitz et al., 2003), possibly composed of

138

Lakes on Mars

evaporites (Komatsu and Di Cencio 2002). Harrison and Chapman (2008) found a depression in the floor of Coprates Chasma, extending from Melas through Coprates and into Eos and Capri Chasmata, for a distance of 1500 km. The depression is surrounded by scarps and benches, interpreted to be erosional ledges of bedrock. These, as well as the outlet, are roughly near the same contour line. The mean lake depth was calculated at 842 m. Evidence for another localized lake was discovered inside Coprates Catena to the south of Coprates Chasma, where a well-developed terraced fan occurs at the mouth of a tributary canyon sourced from a valley on the adjacent plateau (Weitz et al., 2006). The lake would have been about 1.5 km deep. Some nearby light-toned deposits on the trough floor could be evaporites. A low-level lake may also have existed in west Candor Chasma, where Lucchitta (1990) noticed benches on the walls that were reminiscent of strandlines (Fig. 5.24). The purported lake would have had a maximum depth on the order of 1000 m (Lucchitta, 2008a) and would have surrounded the high-standing ILD of Ceti Mensa, which reaches far above the putative strandline in elevation. This lake would have been relatively late, dating from a time when Ceti Mensa was already emplaced or eroded to a high-standing remnant. Some of these postulated lakes near the floors of the chasmata appear to postdate deposition and partial erosion of high ILD mounds. Thus, some ILD may have been emplaced after the troughs had reached their present configuration and older ILD were reduced to their current configuration. These lakes may reflect the last puddles remaining before all standing bodies of water in the Valles Marineris dried up.

Figure 5.24 Benches on southwest wall of west Candor Chasma. Benches (arrows) occur at the 1800–2000 m level, far below the top surface of nearby Ceti Mensa ILD. They may be the strandline of a former lake or reflect the former top surface of the ILD in this area. CTX P03_002195_1743. From Lucchitta (2008a).

Lakes in Valles Marineris

139

5.5.3 ILD, sulfates, and iron oxides A major advance in our understanding of Mars occurred when OMEGA returned detailed mineralogy of Mars. Bibring et al. (2005) found local concentrations of hydrated phyllosilicates in the ancient Martian highlands and hydrated sulfates, commonly associated with ferric oxide, in the Valles Marineris, Aram Chaos, and in the etched terrain (Arvidson et al., 2005) of Meridiani Planum near the MER landing site. The sulfates could have formed in stable lakes or seas, which might have existed at temperatures below 273 K because they were brines (Gendrin et al., 2005). On the other hand, sulfates might also have formed as altered volcanic ash or eolian deposits in the presence of sulfur-rich fluids in hydrothermal settings or from alteration of mafic materials by acidic rain and/or frost (Gendrin et al., 2005). Near-surface presentday water of about 3–4%, possibly recharged from higher obliquity cycles in the past, was identified near the Valles Marineris by the ODY Neutron Spectrometer (Feldman et al., 2005). At these low latitudes, this hydrogen enrichment can be attributed to hydrated minerals, such as sulfates, rather than subsurface ice (Feldman et al., 2005; Vaniman et al., 2004). Gendrin et al. (2005) noted that sulfates occur in light-toned deposits in Ius, Hebes, Capri, and Candor Chasmata, and in Aram Chaos. Many outcrops on bright, steep slopes have a signature consistent with kieserite (Gendrin et al., 2005; Mangold et al., 2008b). Polyhydrated sulfates are more common on less-bright and gentler slopes (Mangold et al., 2008b; Roach et al., 2008a). OMEGA data suggested that gypsum is present within the southern ILD mound in Juventae Chasma (Bibring et al., 2005; Bishop et al., 2008a, 2008b; Catling et al., 2006; Gendrin et al., 2005). However, Kuzmin et al. (2008) posited that the purported gypsum in Juventae Chasma could well be some other form of polyhydrates, and more recent investigations with CRISM found no gypsum in the area (Bishop et al., 2009). A thorough discussion of the location, characteristics, and geological implication of water-altered minerals in the Valles Marineris, including sulfates and iron oxides, is given by Chojnacki and Hynek (2008) (Figure 5.25). The absence of small craters on the ILD favors rapid erosion by eolian abrasion and deflation (Catling et al., 2006; Roach et al., 2008a, 2008b). In a quantitative study of slopes in Candor Chasma, Mangold et al. (2008b) found that kieserite is exposed mostly on slopes >5 degrees, also inferring rapid erosion. Kieserite requires tempera­ tures of 30°–50°C to form (Warren, 1999), which supports the supposition that kieserite was not formed near the surface (Mangold et al., 2008b). The current cold temperatures favor the stability of polyhydrated phases near the surface, and kieserite could be hydrated to polyhydrates on less-eroded slopes (Mangold et al., 2008b; Roach et al., 2008a, 2008b, 2009). The signature of pyroxenes is strongly depleted or absent in layered deposits, leading Mangold et al. (2008b) to suspect that most former pyroxene grains are altered to sulfates. CRISM VNIR spectra in west Candor Chasma showed that reddish materials may have the signature consistent with kieserite, whereas gray materials may have the signature of high-Ca pyroxene as well as monohydrated sulfates such as kieserite (Murchie et al., 2007). Furthermore, monohydrated sulfates commonly occur lower in

290° E

300° E

310° E

0° S

0° S

a

5° S 10° S 15° S

15° S

10° S

5° S

400 km

290° E

300° E

310° E

Figure 5.25 Location of water-altered minerals in Valles Marineris on THEMIS daytime IR base. Kieserite sulfate in red, polyhydrated sulfates in green, hematite in blue. From Chojnacki and Hynek (2008). Copyright 2008 American Geophysical Union. Reproduced by permission of American Geophysical Union.

Lakes in Valles Marineris

141

Figure 5.26 CRISM-derived sulfates (right) superposed on HiRISE image (left), in southwestern Candor Chasma. In this area polyhydrated sulfate (yellow) correlates with discrete bright layers, whereas monohydrated sulfate (kieserite) (blue) correlates with darker material. Some of the dark material forms a distinct layer, but other dark material appears to be debris and eolian sediment on an eroded surface. CRISM targeted observation HRL000033B7, HiRISE image PSP_001641_1735, red band. From Murchie et al. (2007).

the section and polyhydrated sulfates higher in the section (Bishop et al., 2009; Murchie et al., 2009). Bishop et al. (2009) additionally noted, for some ILD in Juventae Chasma, that monohydrated units bearing kieserite tend to be located above such units bearing szomolnokite. In west Candor Chasma, polyhydrated sul­ fates correlate with discrete bright, commonly cliff-forming layers, whereas mono­ hydrated sulfates (kieserite) correlate with darker, more friable layers forming debris modified by eolian processes (Murchie et al., 2007, 2009) (Figure 5.26). This obser­ vation contrasts sharply with those made in most other locations, where kieserite-rich layers tend to form cliffs, whereas polyhydrated layers tend to form slopes (Chojnacki and Hynek, 2008; Mangold et al., 2008b; Roach et al., 2008a, 2008b, 2009). The sulfates occur at discrete horizons and/or layers or groups of layers, which lead Murchie et al. (2007) to propose variations during formation of layers and migrating of diagenetic fluids along particular strata. Okubo and McEwen (2007) indeed found a network of haloed fractures within the ILD in west Candor Chasma, which could be part of an old system of groundwater migration. Cemented knobs in westernmost Candor Chasma (Figure 5.17) might also be caused by mobile groundwater (Chan et al., 2010; Okubo et al., 2008; Treiman, 2008). Roach et al. (2009) thought alteration along specific layers less likely for east Candor Chasma, because the layers of alternating composition are only on the order of meters thick. Chojnacki and Hynek (2008) saw such great diversity among the physical characteristics of the sulfates that they proposed multiple origins for their provenance and formation. Iron oxide outcrops, seen with TES and OMEGA, are systematically located in the vicinity of bright ILD and occur both on sulfate-rich scarps and at the base of these scarps (Chojnacki and Hynek, 2008; Le Deit et al., 2008b; Mangold et al., 2008b; Murchie et al., 2009; Weitz et al., 2008b). Coarse-grained, gray crystalline hematite also occurs in relatively smooth, dark mantles superposed on or adjacent to the ILD (Weitz et al., 2008b), and scattered iron oxide outcrops correlate with dunes or thin

142

Lakes on Mars

veneers of sand (Le Deit et al., 2008b). However many morphologically similar dark mantles do not have detectable hematite, and iron oxide dunes are mostly found in areas where pyroxene-rich dunes are absent (Mangold et al., 2008b). The iron oxides are probably present in the rock as sand-sized grains or larger concretions, as in Meridiani Planum (Mangold et al., 2008b). Le Deit et al. (2008b) inferred that the ferric oxides would have precipitated within the ILD from an Fe-enriched solution, and red crystalline hematite later became diagenetically altered to form secondary gray crystalline hematite, perhaps 1 Ga later (Glotch and Rogers, 2007). Erosion by wind and mass wasting then produced a hematite lag at the surface and concentrations at the base of scarps. Kieserite and polyhydrated sulfates also occur in the etched terrain of Meridiani Planum, whose upper layers may be present at the Opportunity MER landing site (Arvidson et al., 2005), giving us potential ground truth. Silicates and sulfate minerals dominate the outcrop geochemistry, but hematite makes up as much as 11% of the rock by weight. The MER site includes large-scale cross-bedded dunes (Figure 5.27) below and festoon cross-bedding above (Grotzinger et al., 2005; McLennan et al., 2005), which indicates likely deposition in a playa-like inter-dune setting (Squyres et al., 2004; Squyres and Knoll, 2005). The evaporites are probably reworked from a siliciclastic-rich evaporite deposit elsewhere and transported in (Grotzinger et al., 2005; Squyres and Knoll, 2005). The hematite concretions make a lag deposit at the surface, which was responsible for the detection of gray crystalline hematite with the TES instrument (Christensen et al., 2001, 2003).

Figure 5.27 “Burns Cliff” in “Endurance crater,” MER Opportunity landing area in Meridiani Planum. Note the cross-bedded layers (arrow) below the more evenly bedded cliff.

Lakes in Valles Marineris

143

The spectral and morphologic similarity of the units in Meridiani Planum with ILD in Valles Marineris and light-toned deposits in Aureum, Iani, and Aram Chaos suggests a similar aqueous origin (Catling et al., 2006; Glotch and Rogers, 2007; Weitz et al., 2008b). Glotch and Rogers (2007) furthermore thought that an aqueous origin is more likely than an impact or volcanic origin for the sulfates, because they are predominantly found in depressions. Le Deit et al. (2008b) suggested that the ILD may have been deposited as evaporites from precipitated sulfates or, alterna­ tively, be composed primarily of sedimentary or pyroclastic material altered to sulfates by percolation of groundwater. Red hematite could have been produced by abrading or crushing of gray hematite concretions. Mangold et al. (2008b) thought that the ILD mounds were likely composed of a fine-grained sulfate-rich matrix with hematite sand-size concretions. Roach et al. (2008a, 2008b, 2009) preferred an origin in quiescent water because of layer uniformity and extent, but eolian or volcanic ash deposits could also result in such layers under orbital-driven climate change (Head et al., 2003). Roach et al. (2009) further thought that the varying compositions in the layer stacks could result from exposure of metastable cyclic original evaporite sequences or perhaps from more recent hydration of kieserite to polyhydrates under periods of high obliquity. Catling et al. (2006) favored that the ILD are Noachian features exhumed from the walls, originating as salt deposits mixed with sand and dust and laid down in ancient lakes. Or perhaps, they came from polar-like volcanic sulfate aerosols co-precipitated with icy particles under high obliquity. Fueten et al. (2008) proposed that the ILD were deposited syntectonically in ancestral ponds or playas and that groundwater may have altered some, but not all, layered deposits. Their analysis is based on the observation that many, but not all, ILD carry the sulfate signatures. Also in favor of deposition in lakes are Murchie et al. (2009). They proposed trapping of eolian dust, sand, and ash when groundwater discharge evaporated (Andrews-Hanna et al., 2007), leaving hypersaline conditions in which the sulfates precipitated. HRSC, HiRISE, and CTX images confirmed the existence of outwardly dipping, inclined layers on ILD mounds (Fueten et al., 2006, 2008; Gaddis et al., 2006; Hauber et al., 2006; Lucchitta, 2004; Mangold et al., 2008b; Roach et al., 2008a; Zegers et al., 2006). In places the sulfate-rich inclined layers are truncated with angular unconfor­ mities (Lucchitta, 2004) by a dark cap (Figure 5.18) containing pyroxenes (Mangold et al., 2008b) and perhaps emplaced by a volcanic event of local derivation (Fueten et al., 2008). The inclined layers and free-standing ILD mounds or benches were attributed to faulting associated with the final opening of the troughs by Fueten et al. (2006, 2008). Murchie et al. (2007), observing the same mineral content in high mesas and on the floor of west Candor Chasma, concurred that perhaps faulting (Fueten et al., 2008; Lucchitta, 1999) cut and dropped the beds of the high mesas. Chojnacki and Hynek (2008) thought that the radially outward dips of �10 degrees, but in places as high as �20 degrees (Fueten et al., 2006, 2008; Hauber et al., 2006; Zegers et al., 2006), are consistent with a low-energy draping of subaerial or deepwater material, perhaps by Tharsis-related volcanic air fall (Hynek et al., 2003). In this scenario, the sulfates formed by aqueous alteration of volcanic rocks and ash either during or after formation of the ILD. Some dike-like features observed locally within ILD mounds

144

Lakes on Mars

Figure 5.28 Dike-like dark band (arrow) cross-cuts bedding (near right edge of image) at base of Ceti Mensa in west Candor Chasma. HiRISE PSP_001641_1735, red band.

(Figure 5.28) also support the fact that volcanism may have been involved in the formation or alteration of the ILD. Alternatively, evidence for shear deformation on the south side of Ceti Mensa in west Candor Chasma (Figure 5.29) lead Mangold et al. (2008b) to propose that the inclined layers may have come from gravity sliding or tectonic folding set in motion by soft-sediment deformation (Lucchitta, 2008b). Kieserite, like most sulfates and salts, has mechanical behavior of soft material under ambient or relatively low temperatures. Beyer et al. (2000) modeled the behavior of salt diapirs and Brand et al. (2008) the effects of gypsum diapirs. Milliken et al. (2007) proposed that the folds we now see in west Candor Chasma (Figure 5.17) could have been formed by diapirism with hundreds of meters of overburden stripped away. Soft-sediment deformation before and after significant erosion may be important in understanding why the ILD now stand as isolated mounds in the middle of many chasmata.

5.6 Discussion 5.6.1 Morphology and composition The increasingly better resolution of images and refinements in topographic data from recent missions made it apparent that the ILD within the Valles Marineris are not nearly as uniform as previously thought. Some appear massive but at high resolution turn out to be finely layered (Figure 5.16) (Lucchitta and Chapman, 2002) while others are

Lakes in Valles Marineris

145

Figure 5.29 Shear horizon on south side of Ceti Mensa (arrows). Also note the highly disturbed and broken-up terrain in right of picture, where a sliding mass appears completely disrupted. CTX T01_000836_1739.

cross-bedded (Figure 5.30) (Chapman and Tanaka, 2001) and have numerous disconfor­ mities (Weitz et al., 2001) or major angular unconformities (Figure 5.18) (Lucchitta, 2004). Some beds are competent, some weak, and some cyclical (Roach et al., 2008a). Their thickness and lateral continuity vary widely. Most, but not all beds, are light-toned. Many ILD occur in free-standing mesas or mounds separated from the walls by “moats”

Figure 5.30 Cross-beds (arrows) on uppermost layers in Ceti Mensa. HiRISE PSP_002841_1740, red band.

146

Lakes on Mars

(Figure 5.10) (Nedell et al., 1987). In many places the mounds have flat-lying beds in the center but beds that conform to the local slope on the outside (Figure 5.16) (Chojnacki and Hynek, 2008; Fueten et al., 2006, 2008; Gaddis et al., 2006; Hauber et al., 2008; Lucchitta and Chapman, 2002; Roach et al., 2008a). All of these characteristics have to be accounted for if the ILD were deposited in lakes. In addition, most ILD mounds contain hydrated sulfates (Bibring et al., 2005; Gendrin et al., 2005; Mangold et al., 2008b) and ferric oxides in the form of crystalline grey hematite (Christensen et al., 2001; Weitz et al., 2008b) and fine-grained iron oxides (Geissler et al., 1993) mostly at the base of ILD scarps (Figure 5.25) (Le Deit et al., 2008b; Mangold et al., 2008b). CRISM data show that different types of hydrated sulfates are specific to individual layers within the ILD, suggesting different circumstances during deposition or later alteration along selected beds (Murchie et al., 2007, 2009). The prevalence of hydrated sulfates and ferric oxides indicates the involvement of water either during or after emplacement of the ILD.

5.6.2 Are the ILD exhumed from the walls? The hypothesis that the ILD are ancient Noachian lake deposits later exhumed from the Valles Marineris walls (Malin, 1976) was revived after the MGS mission showed that layered terrain was ubiquitous on Mars and had mostly formed in ancient craters (Malin and Edgett, 2000, 2001). Light-toned patches are indeed seen in several places in the walls (Figure 5.13), and therefore a number of authors favored Malin and Edgett’s contention (Bishop et al., 2008a; Catling et al., 2006; Milliken et al., 2007; Montgomery and Gillespie, 2005; Montgomery et al., 2009). However, these light patches are probably pasted on ILD (Lucchitta, 1999, 2001). On the other hand, the light patches may be outcrops of light-toned layers within wall rock that are equivalent to light-toned layered deposits found on top of the plateaus near the Valles Marineris walls (Figure 5.22) (Le Deit et al., 2008a; Lucchitta, 2005; Weitz et al., 2010). These surface layers are different from the ILD in being flat-lying, finely layered (Figure 5.22) (Lucchitta, 2005), and having different mineralogy from the ILD in the troughs (Bishop et al., 2008a; Le Deit et al., 2008a; Mangold et al., 2008b; Milliken et al., 2008a, 2008b; Murchie et al., 2009; Weitz et al., 2008a, 2010). A light-toned layered deposit near Junventae Chasma, on the floor of a depression cut deeply into wall rock (Weitz et al., 2008a, 2010), makes it plausible that such light layers are present throughout the wall rock sequence. Beyer and McEwen’s (2005a) observation that wall material has different competencies also supports the notion that softer material, perhaps such light-toned sedimentary layers, is present in the walls. Early on, the hypothesis that the ILD are exhumed from the walls had been rejected by most researchers (Blasius et al., 1977; Lucchitta et al., 1992; Nedell et al., 1987) because the ILD in many places are clearly superposed on walls that were already eroded into spurs and gullies (Figure 5.10b), because the morphology, competence, and sus­ ceptibility to erosion are vastly different for the ILD than for wall rock (Schultz, 2002), and because ILD lap up onto the walls (Figure 5.13) (Hamelin et al., 2008; Lucchitta, 1999; Lucchitta, 2008a; 2001; Weitz et al., 2001). In addition, Okubo et al. (2008), studying small structures within west Candor Chasma on HiRISE images, concluded

Lakes in Valles Marineris

147

that the ILD at this location postdate formation of the trough because the ILD beds dip into the center of the basin and small structures within the ILD do not reflect the major structures of the trough-bounding faults. Bishop et al. (2009) and Weitz et al. (2010), on the basis of morphology and composition, also confirmed that the ILD were deposited inside the troughs rather than having been eroded out of the walls.

5.6.3 Are the ILD pyroclastic or eolian? An origin for the ILD as pyroclastic deposits has been contemplated by many research­ ers (Chapman, 2002; Chapman and Tanaka, 2001; Hynek et al., 2003; Lucchitta, 1999; Weitz et al., 2001). As many of the ILD within the Valles Marineris are composed of hydrated sulfates with a component of ferric oxides, their compositions are similar to those of other light-toned layered deposits in the depressions of Aram, Aureum, and Iani Chaos (Glotch and Rogers, 2007). If all of these deposits were of volcanic, impact, or eolian origin, they should be found outside their depositional basins as well (Glotch and Rogers, 2007) but they are not. Furthermore, all of these hydrated sulfates and ferric oxides resemble those at the Meridiani MER landing site, where ground truth is available (Arvidson et al., 2005). The beds at Meridiani were deposited in a wet environment showing signs of water activity (Grotzinger et al., 2005; McLennan et al., 2005; Squyres et al., 2005). Even though the beds at Meridiani are largely eolian deposits, apparently reworked from lakebeds or playas nearby (McLennan et al., 2005), water or rising groundwater levels are thought to be responsible for the diagenetic alterations associated with the deposits (Squyres and Knoll, 2005). All of these observa­ tions argue against an airfall pyroclastic origin for the ILD (Chojnacki and Hynek, 2008; Hauber et al., 2006; Hynek et al., 2003; Zegers et al., 2006). However, the ILD could be pyroclastic if the tuffs were trapped by water inside the depressions and then altered by acidic or hydrothermal fluids (Gendrin et al., 2005; Murchie et al., 2009); dry tuffs outside the depressions could have blown away. The same scenario would apply to eolian deposits. A trapping mechanism on ice through which sediments might have foundered was already advocated by Nedell et al. (1987). Overall, even if the ILD were pyroclastic or eolian, a lake would be required to trap the sediments.

5.6.4 Are the ILD sub-ice volcanoes? Chapman and Tanaka (2001), Chapman et al. (2003), Komatsu et al. (2004) and Chapman and Smellie (2007) had proposed that the ILD mounds are subaqueous and/ or sub-ice volcanoes, called tuyas because many ILD mounds are surrounded by out­ wardly dipping layers forming dip slopes (Fueten et al., 2006, 2008; Chojnacki and Hynek, 2008; Lucchitta and Chapman, 2002), and late dark unconformable deposits (Figure 5.18) having mafic spectral signatures (Murchie et al., 2007) are ubiquitous on top of the ILD mounds and on the chasma floors (Lucchitta, 1987b, 1990, 2004; Mangold et al., 2008b). Furthermore, putative dikes have been identified (Figure 5.28) (Lucchitta, 2008b). The tuya hypothesis also explains the tuff-like weathering of many ILD into yardangs and the unconformable dark mafic caps on mesas and mounds. In addition, the hypothesis explains why many of the ILD occur as freestanding mounds,

148

Lakes on Mars

separated from wall rock. General arguments against volcanoes inside the Valles Marineris are the lack of clearly identifiable volcanic craters, calderas, and volcanic flows, even though a few possible flows have been identified (Lucchitta, 1990; Mangold et al., 2008c). Also, the similar light-toned hydrated sulfates in Meridiani and within Aram, Aureum, and Iani Chaos (Glotch and Rogers, 2007) do not show any evidence for a tuya origin. Another consideration is that none of the ILD mounds extends above the surrounding plateau elevations (Weitz et al., 2001) even though some come close (Chapman and Tanaka, 2001). If the ILD mounds were ancient volcanoes, one might expect that at least some would extend above this level. Overall, a tuya origin is debatable considering the presence of similar light-toned deposits in many nonvolcanic areas on Mars, but it cannot be ruled out yet because the ILD inside the Valles Marineris have a vastly different geomorphic setting compared to light-toned deposits elsewhere.

5.6.5 Mounds and moats How did the free-standing mounds acquire this shape if they were not volcanoes? Nedell et al. (1987) proposed that the original troughs were much smaller and that later loss of ice-rich walls enlarged the troughs, leaving the central lake beds as isolated mesas. Fueten et al. (2008) thought that subsidiary grabens between the ILD and trough walls later enlarged the troughs. However, remnants of ILD within wall reentrants (Figure 5.13) (Lucchitta, 1999, 2001) would imply that the original ILD spread across the current extent of the troughs (Chojnacki and Hynek, 2008). If so, the removal may be due to erosion. Weitz et al. (2008b) suggested that hydrated sulfates erode readily, as shown at the Opportunity MER landing site, where at least one meter is removed. Le Deit et al. (2008b), Mangold et al. (2008b), and Chojnacki and Hynek (2008) noted that kieserite and polyhydrated sulfates are soft, fine-grained materials that are easily eroded by wind. The concentration of iron oxides at the base of scarps (Le Deit et al., 2008b; Mangold et al., 2008b; Weitz et al., 2008b) also suggests mobilization by erosion and mass wasting. Catling et al. (2006) calculated that the mounds in Juventae Chasma were eroded at a rate of 1 km/Ga, and Bishop et al. (2009) hypothesized that two of the mounds in this chasma were once contin­ uous, but are now separated by erosion. The scarcity of impact craters on the ILD further confirms deep erosion. Dark mafic layers resting on ILD with major angular unconformities (Chapman and Tanaka, 2001; Fueten et al., 2006, 2008; Lucchitta, 2004) indicate deep erosion before the mafic layers were emplaced. Thus, wind erosion may have removed much of the original lake beds. Perhaps wind was funneled along the structurally weakened seam between wall rock and layered deposits, leaving the ILD as erosional remnants forming isolated mounds.

5.6.6 Inclined layers If the ILD were emplaced as flat-lying lake beds rather than free-standing volcanoes, how did they acquire the peripherally dipping beds? Eolian air fall over early eroded ILD remnants is one possibility, as is air fall from volcanic eruptions (Chojnacki and Hynek, 2008; Hauber et al., 2006; Hynek et al., 2003; Zegers et al., 2006). Sulfate-rich

Lakes in Valles Marineris

149

aerosols in association with snow or ice during high-obliquity cycles (Catling et al., 2006) might also have produced draping beds. However, any air fall mechanism has the problem that ILD are confined to basins and do not occur on surrounding plateaus. Structural tilting was advanced by Fueten et al. (2006, 2008) but such deformation does not readily explain the peripherally dipping beds and the observations that the structures affecting the ILD do not extend into adjacent wall rock (Chapman and Tanaka, 2001). Rossi et al. (2008) proposed the idea that the ILD are overlapping spring mounds, but spring mounds are usually composed of carbonates and not sulfates. The inclined layers could be due to gravity tectonics; soft-sediment deforma­ tion within evaporites and perhaps halite (Gendrin et al., 2005) may have induced surficial layers to slide (Lucchitta, 2008b; Mangold et al., 2008b). As intense softsediment deformation, perhaps due to diapiric activity (Beyer et al., 2000; Brand et al., 2008; Milliken et al., 2007), is evident in westernmost (Figure 5.17) and southern west Candor Chasma (Mangold et al., 2008b) (Figure 5.29), gravity tectonics becomes a plausible mechanism to have deformed formerly flat-lying beds into inclined layers.

5.6.7 Derivation of water by surface runoff It has become well established that small dendritic valley networks existed on the plateaus adjacent to the Valles Marineris. They have been identified near Echus Chasma (Figure 5.21) (Mangold et al., 2004, 2008a), Juventae Chasma (Bishop et al., 2008a; Catling et al., 2006; Le Deit et al., 2008a; Lucchitta, 2005), Ganges Chasma, and elsewhere (Weitz et al., 2008a, 2010). Those near Echus Chasma are Late Hesperian in age and thought to be contemporaneous with trough tributary canyons because they flow into them (Mangold et al., 2004, 2008a). Because the tributary canyons are graded toward the troughs and therefore postdate their opening, these valleys would be contemporaneous with the troughs or younger. Weitz et al.’s (2010) observation of valleys on trough walls also supports that fluvial activity persisted for some time after opening of the chasmata. Mangold et al. (2008a) thought that the valleys on top of the plateaus date back to a time when precipitation was possible, either from an extension of an earlier wet and warm climate or from periodic increase in atmospheric vapor pressure from impact processes, obliquity variations, degassing from volcanism, or outflow channel eruptions. Overall, it is apparent that water flowed on the surface in the vicinity of the troughs during and after their formation and that runoff may have contributed to lakes and lake beds in the Valles Marineris.

5.6.8 Aquifers, channels, and lakes Other evidence for water in the vicinity of the troughs comes from the peripheral troughs and the outflow channels. Coleman et al. (2007a, 2007b) found ample evidence for overflow channels on the plateaus at the eastern end of Candor and Coprates Chasmata (Figure 5.11) (Lucchitta and Ferguson, 1983). These channels flowed over the high plateaus and were erupted from pits and pit chains or they came

150

Lakes on Mars

from a supposedly dammed Coprates Chasma; some flowed into an ancestral Ganges Chasma. Only water erupting with large volumes and considerable force could have formed these pits and channels. A similar overflow situation is present in Juventae Chasma (Catling et al., 2006; Chapman et al., 2003; Coleman et al., 2007c) where a lake at least 500 m deep spilled over an outlet barrier. Coleman et al. (2007a) thought that the water for the peripheral eruptions was recharged from Tharsis (Harrison and Grimm, 2004) and that a potentiometric surface above an absolute elevation of 2500 m was required to cause the water eruptions near the eastern end of the troughs. The water within Candor Chasma could indeed have stood at this elevation; Mangold et al. (2008b) found that this elevation was the level of the uppermost hydrated sulfates in Candor Mensa. Recharge from subterranean aquifers is also advocated by AndrewsHanna et al. (2007) and Murchie et al. (2009). The above considerations favor the hypothesis that the central and peripheral troughs contained relatively deep lakes that permitted through-flowing drainage during outflow channel time. Some of the per­ ipheral chaos date back to the early collapse and effluence of the troughs, but others are later than the initial outflow channels; apparently they formed from subsidiary outbursts of water (Coleman et al., 2007a, 2007b; Rodriguez et al., 2006). With a few exceptions in Juventae Chasma and Aureum Chaos (Catling et al., 2006; Chapman et al., 2003; Glotch and Rogers, 2007), the emplacement of the ILD is even later because they generally overlie chaos hills (Witbeck et al., 1991).

5.6.9 Lakes and ancestral basins Some evidence, however, argues against the existence of deep lakes within the Valles Marineris. No major channels are seen to flow into the central Valles Marineris, which could have brought in the massive sediments required to fill the lakes and form the ILD. There are few deltas (Komatsu et al., 2004) even though deltas are common elsewhere on Mars (see Chapters 8 and 12), where channels emptied into crater lakes (Cabrol and Grin, 1999, 2001; Malin and Edgett, 2003; Pondrelli et al., 2005, 2008), and the ILD are not as flat-lying and evenly bedded as those apparently deposited in lakes such as Holden and Eberswalde craters (Pondrelli et al., 2005, 2008). However, one fan deposit in a minor trough parallel to Coprates Chasma has been reported (Weitz et al., 2006), and a sublacustrine fan complex supposedly existed in a sub­ sidiary trough on the southwest side of Melas Chasma (Metz et al., 2009). Under current climatic conditions, deep lakes would have frozen eventually to the level of the base of the cryosphere, making sedimentary processes more difficult (Carr, 1983). Moreover, deep lakes could not have existed in the central troughs under their present configuration; the water would have spilled out through Coprates Chasma, unless temporary ice and debris dams were in place (Chapman and Tanaka, 2001; Coleman et al., 2007b). For this reason, Lucchitta et al. (1994) and Schultz (1998) argued for a two-stage development of the troughs. Ancestral, structurally controlled, enclosed collapse basins similar to Hebes Chasma formed initially (Figure 5.10c) (Quantin et al., 2004b). These would have been ancestral Ophir, Candor, and parts of Melas Chasmata. These ancient basins probably harbored lakes and were eventually filled with ILD. The basins were later cut by the major through-going grabens of Ius and

Lakes in Valles Marineris

151

Coprates Chasmata, which contain only minor light-toned floor deposits. If floods ensued after the breach, they would have been small, because the central troughs were mostly filled with deposits at that time.

5.6.10 Shallow lakes The initial collapse of the central troughs could have been precipitous, releasing massive amounts of water, possibly from ground ice, into ancestral enclosed basins (Sharp, 1973). However, in analogy with processes operating on Earth, it is more likely that these major structures subsided or collapsed slowly and that sedimentary influx kept pace with the sinking of the trough floors throughout a considerable amount of time (Fueten et al., 2008; Lucchitta, 1982; Mangold et al., 2008b). The lakes may have been playas (Lucchitta, 1982). Thus, evaporites within the ILD could have come from direct precipitation in brines (Gendrin et al., 2005) that were perhaps interlayered or mixed with sand and dust (Catling et al., 2006). CRISM data reveal a great variety in the type and amount of hydrated sulfates within individual beds of the ILD (Murchie et al., 2007, 2009). Possible clastic fractions within the layered sediments may have come from fluvial influx as well as volcanic ash, reworked volcanic rocks, or from eolian materials trapped in shallow or temporary lakes. Some of the clastic material was probably diagenetically altered to hydrated sulfates and iron oxides by varying water tables, similar to the processes that once affected the MER Opportunity landing site. Assuming the ILD are lake beds, a question remains: Why are the ILD within the Valles Marineris so similar in composition to the hydrated sulfate and ferric oxides within the chaos basins and Meridiani Planum, yet have very different morphologic characteristics, occur in isolated mounds, and have sloping layers on their sides? One possible explanation may relate to the different setting during the emplacement of the ILD. The deposits within crater lakes, the chaos, and Meridiani Planum were emplaced in relatively shallow depressions, resulting in deposits of moderate thick­ nesses. Within the deep troughs of the Valles Mariners the ILD are kilometers thick, allowing for the accumulation of a greater variety of sedimentary materials during possibly changing climates and varying water tables. Once the troughs were filled with primary sulfates or diagenetically altered sediments, erosion likely exposed the beds and left mounds of great height and weight, fostering internal and external softsediment deformation. Alternatively, if the mounds are tuyas, whose primary compo­ sition has been intensely altered by hydrothermal activity or brines, the different characteristics are readily explained.

5.6.11 Late lakes It is now becoming increasingly evident that shallow lakes existed within the Valles Marineris at a time when some of the older ILD mounds were already formed and then eroded. Some of the light-toned floor deposits (Lucchitta, 1999) may be attributed to this origin. A late lake in west Candor Chasma may have stood at a surface level of around 1800–2000 m, far below the top of the already eroded Ceti Mensa (Figure 5.24) (Lucchitta, 1987a, 2008a). Possible late lakes were identified on the

152

Lakes on Mars

south side of west Melas Chasma (Figure 5.23) (Quantin et al., 2005, Metz et al., 2009), in a 1500-km-long depression in Coprates Chasma (Harrison and Chapman, 2008), and in Coprates Catena (Weitz et al., 2006). Water in these late lakes may have come from the trough walls; Quantin et al. (2004b) established that the volume of landslide deposits is much less than that of the reentrants whence the landslides came, indicating that material behind the walls was very porous or karstic and likely filled with water. This water may have also contributed to the efficiency of the landslides. On the other hand, Jernsletten (2004) conjectured that no ice was present near the trough-wall surfaces since the end of spur-and-gully erosion and the inception of landslide emplacement; he found no asymmetry on spur-and-gully walls regardless of variations in insolation (asymmetry would be expected if water or ice were present in the walls).

5.6.12 Age of lakes and ILD Quantin et al. (2004a, 2004b), in studying the age and morphology of landslides within the Valles Marineris, established that the environment of their emplacement has not changed since near 3.5 Ga ago and that the Valles Marineris were likely mostly dry since that time. This notion is further supported by the observations that some of the original Valles Marineris floors date back to around 3.5 Ga ago and show little erosion (Quantin et al., 2004a) and that olivine is preserved on ancient trough floors, suggest­ ing that little water-based alteration has taken place since the exposure of these minerals (Figure 5.20) (Christensen et al., 2003; Edwards et al., 2008). Nearly all landslides were emplaced after the ILD mounds were formed and partly eroded (Figure 5.10b), implying that most ILD mounds also date back to the early history of the troughs. As all landslides are developed at the expense of spurs and gullies (Lucchitta 1979; Quantin et al., 2004b), spurs and gullies also formed around that time, perhaps in a subaqueous environment and concomitant with emplacement of the ILD (Jernsletten, 2004; Komatsu et al., 2004). However, possibly some ILD on trough floors in west Candor (Lucchitta, 1987a, 2008a), Melas (Weitz et al., 2003), and Ius Chasmata (Roach et al., 2008c) are later than the high ILD mounds of the central troughs, but even these likely date back to early in the troughs’ history (Quantin et al., 2005). Overall, the opening of the troughs, formation of major lakes, and emplace­ ment of the main body of the ILD began in early Hesperian and likely was over before the end of that period.

5.7 Summary and conclusion The troughs opened about 3.5 Ga ago. Initial collapse along structural planes of weakness formed ancestral troughs, littered with chaotic hills, which filled rapidly with water from ground ice, subterranean aquifers, or nearby valley networks. Ances­ tral outflow channels spilled out of these troughs and flowed across the adjacent plateaus. Sediments may have filled the troughs gradually, as the basins sank, or were deposited into deep lakes after rapid basin collapse. The sediments could have been

Lakes in Valles Marineris

153

influx from valley networks or could have been eolian or ashfall deposits that became trapped in the lakes. The water in the lakes was highly acidic; thus, sediments could have been deposited directly as evaporites or they were later selectively altered to evaporites by hydrothermal activity or the acidic water. Iron oxide concretions formed within these sediments. An alternative interpretation suggests that the ILD did not come from sediments, but formed as sub-ice or sub-water volcanoes to form tuya mounds. Similar alteration by acidic waters could have converted the volcanic rocks and ash to sulfates. Most of the ILD were emplaced early in the troughs’ history. Subsequently, more water erupted from peripheral troughs, leading to additional chaos formation and to the breaching of the ancestral basins. Grabens, including the through-going graben system of Coprates and Ius Chasmata, may have formed at that time and aided in the breaching. The major lakes within the Valles Marineris dried up, except for a few late shallow lakes on the trough floors that were locally fed by tributary canyons or lingering valleys. Spur-and-gully erosion, which carved the early walls, apparently also stopped. Vigorous wind erosion reduced the friable, already altered sediments to isolated mounds and iron oxide concretions weathered out to form lag deposits mostly at the base of scarps. Alternatively, the volcanoes became free-standing edifices. Inclined beds on the mounds could have come from draping by eolian sand, pyroclastic fallout, or occasional atmospheric ice-charged dust fallout during Martian high-obliquity excursions. The inclined layers could have also come from downfaulting or gravity sliding off the sides of the eroded evaporite-charged ILD mounds. Landslides fell into the newly created voids and their scars were blanketed by smooth talus. Minor, continuing volcanic activity, much of it along faults at the base of the graben walls, spewed mafic ash onto the eroded ILD-mound surfaces and onto the trough floors, where it was extensively reworked by wind. Irrespective of whether the above scenario applies or the detailed history diverged from this proposition, it is nearly inevitable that lakes or playas with varying levels of surface or ground water existed at some time inside the Valles Marineris. It is difficult to envision the formation of or alteration to hydrated evaporite and iron oxide deposits without some aqueous activity. Ongoing investigations, especially with hyperspectral data, may eventually shed more light on the origin of the ILD and perhaps clarify the still ongoing debate over the extent to which the original ILD were of sedimentary, volcanic, or mixed provenance.

Acknowledgments I thank the mission teams for making their data available. The images were obtained from Mars Global Surveyor, NASA/JPL/Malin Space Science System; Odyssey, NASA/JPL/Arizona State University; Mars Express, ESA/DLR/FU Berlin (G. Neukum); and Mars Reconnaissance Orbiter, NASA/JPL/University of California. I also thank Phil Christensen, Cathy Quantin, Chris Edwards, Matt Chojnacki, and Scott Murchie for giving permission to reproduce Figures 5.14, 5.19, 5.20, 5.25, and 5.26. Mary Chapman, Chris Okubo, Ken Tanaka, and Cathy Weitz gave insightful reviews that significantly improved the manuscript. The work was supported by the Planetary Geology and Geophysics Program of NASA.

154

Lakes on Mars

References Anderson, R.C., Dohm, J.M., Golombek, M.P., Haldemann, A., Franklin, B.J., Tanaka, K.L., et al., 2001. Significant centers of tectonic activity through time for the western hemisphere of Mars. J. Geophys. Res. 106, 20563–20585. Andrews-Hanna, J., Phillips, R., Zuber, M., 2007. Meridiani Planum and the global 600 hydrology of Mars. Nature 446, 163–168, doi:10.1038/nature05594. Arvidson, R.E., Poulet, F., Bibring., J.-P., Wolff, M., Gendrin, A., Morris, R.V., et al., 2005. Spectral reflectance and morphologic correlations in eastern Terra Meridiani, Mars. Science 307, 1591–1594. Beyer, R.A., McEwen, A.S., 2005a. Layering stratigraphy of eastern Coprates and northern Capri Chasmata, Mars. Icarus 179, 1–23, doi:10,1016/j.icarus.2005.06.014. Beyer, R.A., McEwen, A.S. 2005b. Constraints on the origin of fine layers in Ganges Mensa and Hebes Mensa, Mars, 36th Lunar Planet. Sci. Conf., Abstract 1070. Beyer, R.A., Melosh., H.J., McEwen, A.S., Lorentz, R.D., 2000. Salt diapirs in Candor Chasma, Mars? 31st Lunar Planet. Sci. Conf. Abstract 2022. Bibring, J.-P., Langevin, Y., Gendrin, A., Gondet, B., Poulet, F., Berthé, M., et al., and the OMEGA team, 2005. Mars surface diversity as revealed by the OMEGA/Mars Express observations. Science 307 (5715), 1576–1581. Bishop, J.L., Noe Dobrea, E., Murchie, S.L., Weitz, C.M., Calvin, W.M., Roach, L.A., et al., and the MRO CRISM Team 2008b, Sulfates and mafic minerals in Juventae Chasma as seen by CRISM in coordination with OMEGA, HiRISE and Context images. Seventh Int. Conf. on Mars, Abstract 3350. Bishop, J.L., Parente, M., Weitz, C.M., Noe Dobrea, E., Calvin, W.M., Milliken, R.E., et al., and the CRISM Team, 2008a. Characterization of light-toned sulfate and hydrated silica layers at Juventae Chasma using CRISM, OMEGA, HiRISE and Context images, 39th Lunar Planet. Sci. Conf., Abstract 2334. Bishop, J.L., Parente, M., Weitz, C.M., Noe Dobrea, E.Z., Roach, L.H., Murchie, S.L., et al., 2009. Mineralogy of Juventae Chasma: Sulfates in the light-toned mounds, mafic minerals in the bedrock, and hydrated silica and hydroxylated ferric sulfate on the plateau. J. Geophys. Res. 114, E00D09, doi:10.1029/2009JE003352. Blasius, K.R., Cutts, J.A., Guest, J.E., Masursky, H., 1977. Geology of the Valles Marineris: first analysis of imaging from the Viking I Orbiter primary mission. J. Geophys. Res. 82, 4067–4091. Brand, H.E.A., Middleton, C.A., Grindrod, P.M., Fortes, A.D., Wood, I.G., Vocadlo, L., 2008. Modeling of gypsum and ice diapirs in the Martian crust. 39th Lunar Planet. Sci. Conf., Abstract 1641. Cabrol, N.A., Grin, E.A., 1999. Distribution, classification and ages of Martian impact crater lakes. Icarus 142, 160–172. Cabrol, N.A., Grin, E.A., 2001. The evolution of lacustrine environments on Mars: (Is Mars only Hydrologically Dormant?) Icarus 149, 291–328. Carr, M.H., 1974. Tectonism and volcanism of the Tharsis region of Mars. J. Geophys. Res. 79, 3943–3949. Carr, M.H., 1979. Formation of Martian flood features by release of water from confined aquifers. J. Geophys. Res. 84, 2995–3007. Carr, M.H., 1983. Stability of streams and lakes on Mars. Icarus 56, 476–495. Carr, M.H., 1986. Mars: a water-rich planet? Icarus 56, 187–216. Catling, D.C., Wood, S.E., Leovy, C., Montgomery, D.R., Greenberg, H.M., Glein, C.R., et al., 2006. Light-toned layered deposits in Juventae Chasma, Mars. Icarus 181, 26–51.

Lakes in Valles Marineris

155

Chan, M.A., Ormö, J., Murchie, S., Okubo, C.H., Komatsu, G., Wray, J.J., et al., and the HiRISE Team, 2010. Geomorphic knobs of Candor Chasma, Mars: New Mars Reconnais­ sance Orbiter data and comparisons to terrestrial analogs. Icarus 205, 138–153. Chapman, M.G., 2002. Layered, massive, and thin sediments on Mars: possible late Noachian to late Amazonian tephra? In: Smellie, J.L., & Chapman, M.G. (Eds.), Volcano-Ice Interactions on Earth and Mars. Special Publications, Geological Society, London, 202, 273–203. Chapman, M.G., Gudmundsson, M.T., Russell, A.J., Hare, T.M., 2003. Possible Juventae Chasma subice volcanic eruptions and Maja Valles ice outburst floods on Mars: Implica­ tions of Mars Gobal Surveyor crater densities, geomorphology, and topography. J. Geo­ phys. Res. 108 (E10), 5113, doi:10.1029/2002JE002009. Chapman, M.G., Smellie, J.L., 2007. Mars interior layered deposits and terrestrial sub-ice volcanoes compared: observations and interpretations of similar geomorphic characteris­ tics. In: M.G. Chapman, (Ed.), Geology of Mars: Evidence from Earth-based Analogs. Cambridge University Press, UK, 178–210. Chapman, M.G., Tanaka, K.L., 2001. Interior trough deposits on Mars: Subice volcanoes? J. Geophys. Res. 106 (E5), 10087–10100. Chojnacki, M., Hynek, B.M., 2008. Geological context of water-altered minerals in Valles Marineris, Mars. J. Geophys. Res. 113, E12005, doi:10.1029/2007JE003070. Christensen, P.R., Anderson, D.L., Chase, S.C., Clancy, R.T., et al., 1998. Results from the Mars Global Surveyor Thermal Emission Spectrometer. Science 279 (5357), 1692–1698. Christensen, P.R., Bandfield, J.L., Bell, J.F. III, et al., 2003. Morphology and composition of the surface of Mars: Mars Odyssey THEMIS results. Science 300, 2056–2061. Christensen, P.R., Morris, R.V., Lane, M.D., Bandfield, J.L., Malin, M.C., 2001. Global mapping of Martian hematite mineral deposits: Remnants of water-driven processes on early Mars. J. Geophys. Res. 106 (E10), 23873–23885. Coleman, N.M., Baker, V., 2007. Evidence that a paleolake overflowed the rim of Juventae Chasma, Mars. 38th Lunar Planet. Sci. Conf., Abstract 1046. Coleman, N.M., Dinwiddie, C.L., Baker, V.R., 2007b. Evidence that floodwaters filled and overflowed Capri Chasma, Mars. Geophys. Res. Lett. 3, L07201, doi:10.1029/ 2006GL028872. Coleman, N.M., Dinwiddie, C.L., Casteel, K., 2007a. High outflow channels on Mars indicate Hesperian recharge at low latitudes and the presence of Canyon Lakes. Icarus 189, 344–361. Croft, S.K., 1989a. Mars’ canyon systems: Problems of origin, 4th Intl. Conf. Mars, Jan. 10–13, Tucson, Arizona, 88–89. Croft, S.K., 1989b. Spelunking on Mars: The carbonate-tectonic hypothesis for the origin of Valles Marineris, Mars, MEVTV Workshop on Tectonic Features on Mars, LPI Tech. Rept. 89-06, 21–23. Croft, S.K., 1990. Geologic map of the Hebes Chasma quadrangle, VM500K 00077. NASA Tech. Memo. 4210, 539–541. Davis, P.A., Golombek, M.P., 1990. Discontinuities in the shallow Martian crust. J. Geophy. Res. 95 (B9), 14231–14248. Dohm, J.M., Anderson, R.C., Baker, V.R., Tanaka, K.L., Hare, T.M., Boynton, W.V., 2008. New evidence for a magmatic influence on the origin of Valles Marineris, 39th Lunar Planet. Sci. Conf., Abstract 1948. Edwards, C.S., Christensen, P.R., Hamilton, V.E., 2008. Evidence for extensive olivine-rich basalt bedrock outcrops in Ganges and Eos chasmas, Mars. J. Geophys. Res. 113, E11003, doi:10.1029/2008JE003091.

156

Lakes on Mars

Erard, S., Bibring, J.-P., Mustard, J., Forni, O., Head, J.W., Hurtrez, S., et al., 1991. Spatial variation in composition of the Valles Marineris and Isidis Planitia Regions of Mars derived from ISM data. Proceedings 21st Lunar Planetary Science Conference, Houston, TX, 437–455. Feldman, W.C., Prettyman, T.H., Maurice, S., Nelli, S., Elphic, R., Funsten, H.O., et al., 2005. Topographic control of hydrogen deposits at low latitudes to midlatitudes of Mars. J. Geophys. Res. 110, E11009, doi:10.1029/2005JE002452. Frey, H.V., 1979. Martian canyons and African rifts: Structural comparisons and implications. Icarus 37, 142–155. Fueten, F., Stesky, R., MacKinnon, P., Hauber, E., Gwinner, K., Scholten, F., et al., 2006. A structural study of an interior layered deposit in southwestern Candor Chasma, Valles Marineris, Mars, using high resolution stereo camera data from Mars Express. Geophys. Res. Lett. 33, L07202, doi:10.1029/2005GL025035. Fueten, F., Stesky, R., MacKinnon, P., Hauber, E., Zegers, T., Gwinner, K., et al., 2008. Stratigraphy and structure of interior layered deposits in west Candor Chasma, Mars, from High Resolution Stereo Camera (HRSC) stereo imagery and derived elevations. J. Geophys. Res. 113, E10008, doi:10.1029/2007JE003053. Gaddis, L.R., Skinner, J., Hare, T., Kirk, R., Titus, T., Weller, L., et al., 2006. Morphology and morphometry of Ceti Mensa, west Candor Chasma, Mars, 37th Lunar Planet. Sci. Conf., Abstract 2076. Geissler, P.E., Singer, R.B., Komatsu, G., Murchie, S., Mustard, J., 1993. An unusual spectral unit in west Candor Chasma: evidence for aqueous or hydrothermal alteration in the Martian canyons. Icarus 106, 380–391. Geissler, P.E., Singer, R.B., Lucchitta, B.K., 1990. Dark materials in Valles Marineris: indica­ tions of the style of volcanism and magmatism on Mars. J. Geophys. Res. 95, 14399–14413. Gendrin, A., Mangold, N., Bibring, J.-P., Langevin, Y., Gondet, B., Poulet, F., et al., 2005. Sulfates in Martian Layered Terrains: The OMEGA/Mars express view. Science 307, 1587–1591, 10.1126/science.1109087. Glotch, T.D., Rogers, A.D., 2007. Evidence for aqueous deposition of hematite – and sulfaterich light-toned layered deposits in Aureum and Iani Chaos, Mars. J. Geophys. Res. 112, E06001, doi:10/1029/2006JE002863. Grotzinger, J.P., Arvidson, R.E., Bell, J.J., Calvin, I.I.I.W., Clark, B.C., Fike, D.A., et al., 2005. Stratigraphy, sedimentology and depositional environment of the Burns Formation, Mer­ idian Planum, Mars. Earth Planet. Sci. Lett. 240, 11–72. Hamelin, N., Racher, H., Fueten, F., Stesky, R., MacKinnon, P., Hauber, E., et al., 2008. Structural analysis of an interior layered deposit in northern Coprates Chasma, Mars. 39th Lunar Planet. Sci. Conf., Abstract 1424 Harrison, K.P., Chapman, M.G., 2008. Evidence for ponding and catastrophic floods in central Valles Marineris, Mars. Icarus 198, 351–364. Harrison, K.P., Grimm, R.E., 2004. Tharsis recharge: a source of groundwater for Martian outflow channels. Geophys. Res. Lett. 31, doi:10.1029/2004GL020502. Harrison, K.P., Grimm, R.E., 2008. Multiple flooding events in Martian outflow channels. J. Geophys. Res. 113, doi:10.1029/2007JE002951. Hauber, E., Gwinner, K., Gendrin, A., Fueten, F., Stesky, R., Pelkey, S., et al., 2008. Hebes Chasma, Mars: slopes and stratigraphy of interior layered deposits. 39th Lunar Planet. Sci. Conf.Abstract 2375 Head, J.W., Mustard, J.F., Kreslavsky, M.A., Milliken, R.E., Marchant, D.R., 2003. Recent ice ages on Mars. Nature 246, 797–802.

Lakes in Valles Marineris

157

Howard, A.D., 1991. Role of groundwater in formation of Martian channels. Reports Planet. Geol. Geophys. Program-1990, NASA Tech. Memo., 4300, 117–119. Hynek, B.M., Phillips, J., Arvidson, R.E., 2003. Explosive volcanism in the Tharsis region: global evidence in the Martian geologic record. J. Geophys. Res. 108 (E9), doi:10,1029/ 2003JE002062. Jernsletten, J.A., 2004. A topographic test for the existence of ground ice in the walls of Coprates Chasma, Mars. J. Geophys. Res. 109, E12004, doi:10.1029/2004JE002272. Kochel, R.C., Piper, J., 1986. Morphology of large valleys on Hawaii: evidence for ground water sapping and comparisons with Martian valleys, Proceedings of Lunar Planetary Science 17, J. Geophys. Res., Suppl. 91, E175–E192. Komatsu, G., Di Cencio, A., 2002. The origin of light-color units on the floor of Valles Marineris, Mars, 33rd Lunar Planet. Sci. Conf., Abstract 1184. Komatsu, G., Geissler, P.E., Strom, R.G., Singer, R.B., 1993. Stratigraphy and erosional land­ forms of layered deposits in Valles Marineris, Mars. J. Geophys. Res. 98, 11105–11121. Komatsu, G., Ori, G.G., Ciarcelluti, P., Litasov, Y., 2004. Interior layered deposits of Valles Marineris, Mars: analogous subice volcanism related to Baikal rifting, Southern Siberia. Planet. Space Sci. 52, 167–187. Kuzmin, R.O., Mironenko, M.V., Evdokimova, N.A., 2008. Mars: spectral and thermodynamic restriction on the gypsum existence in the Juventae Chasma, 39th Lunar Planet. Sci. Conf., Abstract 1575. Laity, J.E., Malin, M.C., 1985. Sapping processes and the development of theater-headed valley networks on the Colorado Plateau. Geol. Soc. Am. Bull. 96, 203–217. Le Deit, L., Bourgeois, O., Le Mouelic, S., Mege, D., Combe, J.-Ph., Sotin, C., et al., 2008a. Light-toned layers on plateaus above Valles Marineris (Mars), 39th Lunar Planet. Sci. Conf., Abstract 1740. Le Deit, L., Le Mouelic, S., Beougeois, O., Combe, J.-P., Mege, D., Sotin, C., et al., 2008b. Ferric oxides in East Candor Chasma, Valles Marineris (Mars) inferred from analysis of OMEGA/Mars express data; identification and geological interpretation. J. Geophys. Res. 113, E07001, doi:10.1029/2007JE002950. Lucchitta, B.K., 1978a. Morphology of chasma walls, Mars. U.S. Geol. Survey J. Res. 6 (5), 651–662. Lucchitta, B.K., 1978b. A large landslide on Mars. Geol. Soc. Am. Bull. 89 (11), 1601–1609. Lucchitta, B.K., 1979. Landslides in Valles Marineris, Mars. J. Geophys. Res. 84 (B14), 8097–8113. Lucchitta, B.K., 1982. Lakes or playas in Valles Marineris (abs). NASA Tech. Mem. 85127, 233–234. Lucchitta, B.K., 1984. A late climatic change on Mars (abs.), 15th Lunar Planet. Sci. Conf., 493–494. Lucchitta, B.K., 1987a. Valles Marineris, Mars: wet debris flows and ground ice. Icarus 72, 411–429. Lucchitta, B.K., 1987b. Recent mafic volcanism on Mars. Science 235 (4788), 565–567. Lucchitta, B.K., 1990. Young volcanic deposits in the Valles Marineris, Mars. Icarus 86, 476–509. Lucchitta, B.K., 1999. Geologic map of Ophir and Central Candor Chasmata of Mars (MTM ­ 05072), U.S. Geol. Survey Misc. Inv. Ser. Map I-2568. Lucchitta, B.K., 2001. MOC images confirm layered deposits formed within Valles Marineris, Mars, 32nd Lunar Planet. Sci. Conf., Abstract 1359. Lucchitta, B.K., 2004. A volcano composed of light-colored layered deposits on the floor of Valles Marineris, 35th Lunar Planet. Sci. Conf., Abstract 1881.

158

Lakes on Mars

Lucchitta, B.K., 2005. Light layer and sinuous ridges on plateau near Juventae Chasma, Mars, 36th Lunar Planet. Sci. Conf., Abstract 1500. Lucchitta, B.K. 2008a, HiRISE images of layered deposits in west Candor Chasma, Mars (I): wall rock relations, enigmatic ridges, and possible dikes. 39th Lunar Planet. Sci. Conf., Abstract 2169. Lucchitta, B.K., 2008b. HiRISE images of layered deposits in west Candor Chasma, Mars (II): unconformities and possible gravity tectonics. 39th Lunar Planet. Sci. Conf., Abstract 2317. Lucchitta, B.K., Bertolini, L.M., 1990. Interior structures of Valles Marineris, Mars, 20th Lunar Planet. Sci. Conf. 590–591. Lucchitta, B.K., Chapman, M.G., 2002. Are the Valles Marineris giant volcano-tectonic depres­ sions? 33th Lunar Planet. Sci. Conf., Abstract 1689. Lucchitta, B.K., Ferguson, H.M., 1983. Chryse basin channels: low gradients and ponded flow. J. Geophys. Res. 88, (Suppl. 13), Proceedings of Lunar Planetary Science Conference, pt. 2, A553–A568. Lucchitta, B.K., Isbell, N.K., Howington-Kraus, A., 1994. Topography of Valles Marineris: implications for erosional and structural history. J. Geophys. Res. 99, 3783–3798. Lucchitta, B.K., McEwen, A.S., Clow, G.D., Geissler, P.E., Singer, R.B., Schultz, R.A., et al., 1992. The canyon system on Mars. In: Kieffer, H.H (Ed.), Mars. University of Arizona Press: Tucson, pp. 453–492. Malin, M.C., 1976. Nature and origin of the intercrater plains on Mars, Ph.D. Thesis, California Institute of Technology, Pasadena, California. Malin, C.M., Edgett, K.S., 2000. Sedimentary rocks of early Mars. Science 290, 1927–1937. Malin, M.C., Edgett, K.S., 2001. Mars Global Surveyor Mars Orbiter Camera: Interplanetary cruise through primary mission. J. Geophys. Res. 106 (E10), 23429–23570. Malin, M.C., Edgett, K.S., 2003. Evidence for persistent flow and aqueous sedimentation on early Mars. Science 302, 1931–1934. Mangold, N., Ansan, V., Masson, M.Ph., Quantin, C., Neukum, G., 2008a. Geomorphic study of fluvial landforms on the northern Valles Marineris plateau, Mars. J. Geophys. Res. 113, E08009, doi:10.1029/2007JE002985. Mangold, N., Gendrin., A., Gondet, B., LeMouelic, S., Quantin, C., Ansan, V., et al., 2008b. Spectral and geological study of the sulfate-rich region of west Candor Chasma, Mars. Icarus, 194, 519–543. Mangold, N., Gendrin, A., Gondet, B., LeMouelic, S., Quantin, C., Bibring, J.-P., et al., 2008c. Mineralogy of west Tithonium Chasma-Noctis Labyrinthus, Mars: putative volcanics on Noctis canyon floors, 39th Lunar Planet Sci. Conf., Abstract 1592. Mangold, N., Quantin, C., Ansan, V., Delacourt, C., Allemand, P., 2004. Evidence for pre­ cipitation on Mars from dendritic valleys in the Valles Marineris area. Science 305, 78–81. Masson, P., 1977. Structure pattern analysis of the Noctis Labyrinthus, Valles Marineris region of Mars. Icarus 30, 49–62. McCauley, J.F., 1978. Geologic map of the Coprates quadrangle of Mars. U. S. Geol. Surv. Misc. Invest. Ser. Map I-897. McCauley, J.F., Carr, M.H., Cutts, J.A., Hartmann, W.K., Masursky, H., Milton, D.J., et al., 1972. Preliminary Mariner 9 report on the geology of Mars. Icarus 17, 289–327. McEwen, A.S., 1989. Mobility of large rock avalanches: evidence from Valles Marineris, Mars. Geology 17, 1111–1114. McEwen, A.S., Malin, M.C., Carr, M.H., Hartmann, W.K., 1999. Voluminous volcanism on early Mars revealed in Valles Marineris. Nature 397, 584–586. McEwen, A.S., Soderblom, L.A., 1989. Mars color/albedo and bedrock geology, Fourth Intern. Conf. Mars, Jan. 10–13, Tucson, Arizona, Abstract Book, 138–139.

Lakes in Valles Marineris

159

McKay, C.P., Nedell, S.S., 1988. Are there carbonate deposits in Valles Marineris, Mars? Icarus 73, 142–148. McLennan, S.M., Bell, J.F. III, Calvin, W.M., et al., 2005. Provenance and diagenesis of the evaporitebearing Burns formation, Meridiani Planum, Mars. Earth Planet. Sci. Lett. 240, 95–121. Metz, J.M., Grotzinger, J.P., Mohrig, D., Milliken, R., Prather, B., Pirmez, C., et al., 2009. Sublacustrine depositional fans in southwest Melas Chasma. J. Geophys. Res. 114, E10002, doi:10.1029/2009JE003365. Milliken, R.E., Grotzinger, J.P., Beyer, R.A., Murchie, S., McEwen, A., and the CRISM Science Team, 2007. Evidence for salt tectonics in Valles Marineris, Mars, 7th Intern. Conf. Mars Extended Abstract 3383. Milliken, R.E., Swayze, G., Arvidson., R., Bishop, J., Clark, R., Ehlmann, B., et al., and the CRISM Science Team, 2008a. Spectral evidence for sedimentary silica on Mars, 39th Lunar Planet. Sci. Conf., Abstract 2025. Milliken, R.E., Swayze, G.A., Arvidson, R.E., Bishop, J.L., Clark, R.N., Ehlmann, B.L., et al., 2008b. Opaline silica in young deposits on Mars. Geology 36 (11), 847–850, doi:845 810.1130/G24967A.24961. Montgomery, D.R., Gillespie, A., 2005. Formation of Martian outflow channels by catastrophic dewatering of evaporate deposits. Geology 33, 625–628, doi:10.1130/G21270.1. Montgomery, D.R., Jackson, M.P.A., Schreiber, B.C., Gillespie, A.R., Adams, J.B., 2009. Continental-scale salt tectonics on Mars and the origin of Valles Marineris and associated outflow channels. Geol. Soc. Am. Bull. 121, 177–133, doi:10,1130/B26307.1. Murchie, S.L., Erard, S., Mustard, J.F., Bibring, J., Langevin, Y., Head, J.W., et al., 1992. The geology of the interior deposits of Valles Marineris from Viking images and ISM imaging spectroscopy. Proceedings of the 23rd Lunar Planetary Science Conference, Houston, TX, pp. 945–946. Murchie, S., Kirkland, L., Erard, S., Mustard, J., Robinson, M., 2000. Near-infrared spectral variations of Martian surface materials from ISM imaging spectrometer data. Icarus 147, 444–471. Murchie, S., Milliken, R., Roach, L., Bibring, J.-P., Mustard, J., Pelkey, S., et al., and the CRISM Science Team, 2007. CRISM mapping of layered deposits in western Candor Chasma. Seventh International Conference on Mars, Extended Abstract 3238. Murchie, S., Roach, L., Seelos, F., Milliken, R., Mustard, J., Arvidson, R., et al., 2009. Evidence for the origin of layered deposits in Candor Chasma, Mars, from mineral composition and hydrologic modeling. J. Geophys. Res. 114, E00D05, doi:10.1029/2009JE003343. Mustard, J.F., Murchie, S.L., 2001. Unusual spectral properties observed in Valles Marineris. 32nd Lunar Planet. Sci. Conf. Abstract 2194. Nedell, S.S., Squyres, S.W., Anderson, D.W., 1987. Origin and evolution of the layered deposits in the Valles Marineris. Icarus 70, 409–441. Okubo, C.H., Lewis, K.W., McEwen, A.S., Kirk, R.L., 2008. Relative age of interior layered deposits in southwest Candor Chasma based on high-resolution structural mapping. J. Geophys. Res. 113, E12002, doi:10.1029/2008JE003181. Okubo, C.H., McEwen, A.S., 2007. Fracture controlled paleo-fluid flow in Candor Chasma, Mars. Science 315, 983–985, doi:10.1126/science.1136855. Pelkey, S.M., Jakosky, B.M., Christensen, P.R., 2003. Surficial properties in Melas Chasma, Mars, from Mars odyssey THEMIS data. Icarus 165, 68–89. Peterson, C., 1981. A secondary origin for the central plateau of Hebes Chasma, Proceedings of Lunar and Planetary Science Conference, 12th, v. 2, Pergamon Press, 1459–1471. Peulvast, J.-P., Mège, D., Chiciak, J., Costard, F., Masson, P.L., 2001. Morphology, evolution and tectonics of Valles Marineris wallslopes (Mars). Geomorphology 37, 329–352.

160

Lakes on Mars

Plescia, J.B., Saunders, R.S., 1982. Tectonic history of the Tharsis region of Mars. J. Geophys. Res. 87, 9775–9791. Pondrelli, M., Baliva, A., Di Lorenzo, S., Marinangeli, L., Rossi, A.P., 2005. Complex evolu­ tion of paleolacustrine systems on Mars: an example from the Holden crater. J. Geophys. Res. 110, E04016, doi:10.1029/2004JE002335. Pondrelli, M., Rossi, A.P., Marinangeli, L., Hauber, E., Gwinner, K., Baliva, A., et al., 2008. Evolution and depositional environments of the Eberswalde fan delta, Mars. Icarus 197, 429–451. Quantin, C., Allemand, P., Delacourt, C., 2004b. Morphology and geometry of Valles Marineris landslides. Planet. Space Sci. 52, 1011–1022. Quantin, C., Allemand, P., Mangold, N., Delacourt, C., 2004a. Ages of Valles Marineris (Mars) landslides and implications for canyon history. Icarus 172, 555–572. Quantin, C., Allemand, P., Mangold, N., Dromat, G., Delacourt, C., 2005. Fluvial and lacustrine activity on layered deposits in Melas Chasma, Valles Marineris, Mars, J. Geophys. Res. 110, E12S19. Roach, L.H., Mustard, J.F., Murchie, S.L., Aharonson, O., Lowenstein, T., Weitz, C.M., et al., and the CRISM Science Team, 2008b. Sulfate mineral stratigraphy in Valles Marineris interior layered deposits, 39th Lunar Planet. Sci. Conf., Abstract 1891. Roach, L.H., Mustard, J.F., Murchie, S.L., Bibring, J.-P., Arvidson, R.E., Bishop, J.L., et al., and the CRISM Science Team, 2008a. Constraints on the rate of sulfate phase changes in Valles Marineris Interior Layered Deposits, 39th Lunar Planet. Sci. Conf., Abstract 1823. Roach, L.H., Mustard, J.F., Murchie, S.L., Bibring, J.-P., Forget, F., Lewis, K.W., et al., 2009. Testing evidence of recent hydration state change in sulfates on Mars. J. Geophys. Res. 114, E00D02, doi:10.1029/2008JE003245. Roach, L.H., Mustard, J.F., Murchie, S.L., Milliken, R.E., Crowley, J.K., Bishop, J.L., et al., 2008c. Sulfates and other hydrated minerals in Ius Chasma, Valles Marineris and implica­ tions for water geochemistry, AGU Fall 2008, Abstract P43B–1394. Rodriguez, J.A., Kargel, J., Crown, D.A., Bleamaster, L.F. III, Tanaka, K.L., Baker, V., et al., 2006. Headward growth of chasmata by volatile outbursts, collapse, and drainage: evidence from Ganges Chaos, Mars. Geophys. Res. Lett. 33, doi:10.1029/2006GL026275, 2006.. Rossi, A.P., Neukum, G., Pondrelli, M., van Gasselt, S., Zegers, T., Hauber, E., et al., 2008. Large-scale spring deposits on Mars? J. Geophys. Res. 113, E08016, doi:10.1029/ 2007JE003062. Schultz, R.A., 1991. Structural development of Coprates Chasma and western Ophir Planum, Valles Marineris rift, Mars. J. Geophys. Res. 96, 22777–22792. Schultz, R.A., 1998. Multiple-process origin of Valles Marineris basins and troughs, Mars. Planet. Space Sci. 46, 827–834. Schultz, R.A., 2002. Stability of rock slopes in Valles Marineris, Mars Geophys. Res. Lett. 29 (19), 1932, doi:10.1029/2002GL015728. Scott, D.H., Carr, M.H., 1978. Geologic Map of Mars, scale 1:25,000,000. U. S. Geol Survey Misc. Inv. Series Map I-1083. Scott, D.H., Tanaka, K.L., 1986. Geologic map of the western equatorial region of Mars, scale 1:15,000,000, U. S. Geol. Survey Misc. Inv. Series, Map I–1802–A. Shaller, P.J., 1991. Analysis and implications of large Martian and terrestrial landslides, Ph.D. Thesis, 586 pp., California Institute of Technology, Pasadena. Sharp, R.P., 1973. Mars: troughed terrain. J. Geophys. Res. 78, 4063–4072. Skilling, I.P., Chapman, M.G., Lucchitta, B.K., 2002. Young, blocky flows in East Ius/West Melas and West Candor Chasmata, Mars: debris avalanche deposits derived from interior layered deposit (ILD) mounds? 33rd Lunar Planet. Sci. Conf., Abstract 1361.

Lakes in Valles Marineris

161

Spencer, J.R., Croft, S.K., 1986. Valles Marineris as karst: feasibility and implications for Martian atmospheric evolution (abs.), in Reports of Planet. Geol. Geophys. Program, 1985, NASA Technical Memorandum 88383, 193–195. Spencer, J.R., Fanale, F.P., 1990. New models for the origin of Valles Marineris closed depressions. J.Geophys. Res. 95, 14301–14313. Squyres, S.W., et al., 2004. In situ evidence for an ancient aqueous environment at Meridiani Planum, Mars. Science 306, 1709–1713. Squyres, S.W., Knoll, A.H., 2005. Sedimentary rocks at Meridiani Planum: origin, diagenesis, and implications for life on Mars. Earth Planet. Sci. Lett. 240, 1–10. Tanaka, K.L., Golombek, M.P., 1989. Martian tension fractures and the formation of grabens and collapse features at Valles Marineris, Proceedings of Lunar Planetary Science Con­ ference 19th, 383–396. Treiman, A.D., 2008. Ancient groundwater flow in the Valles Marineris on Mars inferred from fault trace ridges. Nature Geosci. 1, 181–183. Van Bemmelen, R.W., Rutten, M.G., 1955. Table Mountains of Northern Iceland. E. J. Brill, Leiden, Netherlands, 217 pp. Vaniman, D.T., Chipera, S.J., Bish, D.L., et al., 2004. Sulfate salts, regolith interactions, and water storage in equatorial Martian regolith, 35th Lunar Planet. Sci. Conf., Abstract 1690. Ward, W.R., 1974. Climatic variations on Mars. I. Astronomical theory of insolation. J. Geophys. Res. 79, 3375–3386. Warren, J., 1999. Evaporites. Blackwell Sciences, London, 438 pp. Weitz, C.M., Irwin, R.P. III, Chuang, F.C., Bourke, M.C., Crown, D.A., 2006. Formation of a terraced fan deposit in Coprates Catena, Mars. Icarus 184, 436–451. Weitz, C.M., Lane, M.D., Staid, M., Noe Dobrea, E., 2008b. Gray hematite distribution and formation in Ophir and Candor Chasmata. J. Geophys. Res. 113, E02016, doi:10.1029/ 2007JE002930. Weitz, C.M., Milliken, R.E., Grant, J.A., McEwen, A.S., Williams, R.M.E., Bishop, J.L., 2008a. Light-toned strata and inverted channels adjacent to Juventae and Ganges Chasmata, Mars. Geophys. Res. Lett. 35, L19202, doi:10.1029/2008GL035317.. Weitz, C.M., Milliken, R.E., Grant, J.A., McEwen, A.S., Williams, R.M.E., Bishop, J.L., et al., 2010. Mars Reconnaissance Orbiter observations of light-toned layered deposits and associated fluvial landforms on the plateaus adjacent to Valles Marineris. Icarus 205, 73–102. Weitz, C.M., Parker, T.J., Anderson, F.S., Grant, J.A., 2001. The interior layered deposits of Valles Marineris: Layering, erosional processes, and age relationships, 32nd Lunar and Planet. Sci. Conf., Abstract 1629. Weitz, C.M., Parker, T.J., Bulmer, M.H., Anderson, F.S., Grant, J.A., 2003. Geology of the Melas Chasma landing site for the Mars Exploration Rover mission. J. Geophys. Res. 108 (E12), 8082, doi:10.1029/2002JE002022. Williams, J.-P., Paige, D.A., Manning, C.E., 2003. Layering in the wall rock of Valles Mar­ ineris: intrusive and extrusive magmatism. Geophys. Res. Lett. 30, doi:10.1029/ 2003GL017662. Wise, D.U., Golombek, M.P., McGill, G.E., 1979. Tharsis province of Mars: Geologic sequence, geometry, and a deformation mechanism. Icarus 38, 456–472. Witbeck, N.E., Tanaka, K.L., Scott, D.H., 1991. Geologic map of the Valles Marineris region, Mars, Scale 1:2,000,000, U. S. Geol. Survey Misc. Inv. Series, Map I-2010. Zegers, T.E., Dabekaussen, W., Hauber, E., Gwinner, K., Scholten, F., Fueten, F., et al. , 2006. 3D structural analysis of Ophir Chasma based on HRSC image data and stereo-derived DTM, 37th Lunar Planet. Sci. Conf., Abstract 1605.

6 Episodic ponding and outburst

flooding associated with chaotic terrains in Valles Marineris Keith P. Harrison† and Mary G. Chapman* † *

6.1

Southwest Research Institute, Boulder, CO, USA Planetary Science Institute, Tuscon, AZ, USA

Introduction

The Valles Marineris canyons are among the deepest of topographic depressions on Mars. As such, they may be expected to have served as sinks for a variety of mobile materials, including water, regardless of their particular formation history. Certainly, the likelihood of ancient occurrences of ponding in other topographic depressions, such as impact craters, is widely recognized (Cabrol and Grin, 1999, 2001, 2005; Irwin et al., 2002; Malin and Edgett, 2003). However, identification of lacustrine activity in the Valles Marineris has generally lagged comparable discoveries at other locations, and in some cases has been less conclusive. This is partly due to the greater obscurity of some canyon floors in older, lower-resolution datasets, and because the formation history of the canyons is extremely complex, involving a wide range of processes that have (over billions of years) widened, mantled, and fractured the canyon walls and floors (for an overview, see Lucchitta et al., 1992). In Chapter 5, Lucchitta presents an analysis of those Valles Marineris formation and modification processes relevant to paleolakes. Recent data suggest that ponding is evident at three scales. At the smallest (few 10 s km), a putative lake in southwest Melas Chasma appears to have formed by the local ponding of surface runoff, as indicated by influent valley networks (Quantin et al., 2005; Metz et al., 2009). At an intermediate scale (1000 km) is the relatively shallow lake in the Valles Marineris canyon system described in this chapter. At the largest scale (possibly the entire canyon system) are deep lakes inferred from interior layered deposits (ILDs), which form large rounded mounds and mesas on the canyon floors described in Chapter 5. Much controversy remains regarding the formation of these deposits. A lacustrine interpretation has remained viable since its introduction by McCauley (1978), although it is complicated by the significant dipping of some beds and differences in the layering between deposits that would (according to the current topography of the canyons) have formed in the same body of water (see Chapter 5). The interpretation of ILD formation is not critical to the putative Valles Marineris lake considered here. Considerably more important are chaotic terrains, which may constitute the primary source of water for certain putative ancient lakes in, and near, Lakes on Mars. DOI: 10.1016/B978-0-444-52854-4.00006-4 © 2010 Elsevier B.V. All rights reserved.

164

Lakes on Mars

the Valles Marineris (Di Achille et al., 2007; Ori and Mosangini, 1998). Chaotic terrains are large (10 s–100 s km) regions of fractured and subsided materials found, in several cases, at the termini of Valles Marineris canyons. The largest observed fluvial channels in the solar system appear to emanate from Martian chaotic terrains, prompt­ ing initial observers to conclude that these terrains were formed by the catastrophic expulsion of large volumes of groundwater from a pressurized aquifer, or by melting of ground ice due to magnetic intrusions (Komatsu et al., 2000; Masursky et al., 1977; Max and Clifford, 2000; Saunders, 1979; Sharp, 1973). Water discharged in this way may have flooded the neighboring plateaus, carving the large outflow channels. Carr (1996) estimated the volume of water employed in the construction of these channels to be a few billion km3, although large uncertainties in the sediment carrying ability of the floodwaters permit even greater estimates (Harrison and Grimm, 2004). The aforementioned observation that some chaotic terrains are topographically connected to Valles Marineris canyons suggests, at least superficially, that groundwater discharge at these terrains flooded both the canyons and the outflow channels. Such floods undoubtedly led to ponding, either locally or at points along the drainage path. Topography data collected by MOLA (Smith et al., 2001) on board MGS indicate that almost all channel head regions lie well above the chaotic depressions from which they emanate (elevation differences range from 100s to 1000s m). Although the depressed topography of chaotic terrains undoubtedly arose, in part, due to collapse of fractured crustal blocks, it appears in some cases to have had a prior tectonic or impact cratering origin. Even in wholly collapse-related depressions, the final stages of discharge through chaotic terrains would likely have produced ponding. Coleman et al. (2007) used elevation relationships to infer ponding in eastern Capri Chasma, northeast of the lake described here. Other studies have interpreted observed morphological features to be lacustrine. Three of these studies involve ponding in outflow channel source regions: Hydraotes Chaos (Ori and Mosangini, 1998), Aram Chaos (Glotch and Christensen, 2005), and Juventae Chasma (Chapman et al., 2003; Coleman and Baker, 2007). At least one study has proposed ponding in an outflow channel: Di Achille et al. (2007) identified shoreline terraces and a Gilbert-type delta deposit along Shalbatana Vallis, later confirmed in high resolution imagery (Di Achille et al., 2009). It is also worth noting that Gorgonum Chaos, a terrain far west of the Tharsis rise, may also have held bodies of standing water (Howard and Moore 2004). As an addition to the aforementioned investigations, and as a case study of Martian lacustrine processes, we present evidence (previously described by Harrison and Chapman, 2008) of what may be discharge-related ponding in a Valles Marineris depression (VMD) that includes the deepest parts of the canyon system. Evidence for ponding in the VMD consists primarily of bench morphologies observed along canyon walls, and fluvial features that suggest overflow at the eastern end of the depression. Benches are found over a wide area of the canyon, encompassing central and eastern Melas Chasma, the entire length of Coprates Chasma, and parts of Capri and Eos Chaotes (Figure 6.1). An important part of their lacustrine interpretation is their topographic relationship to the depression thought to have been occupied by the paleolake. Thus, before characterizing bench morphology in detail, we describe the approach used to constrain the boundaries of the depression.

Episodic ponding and outburst flooding associated with chaotic terrains in Valles Marineris

a

165

−6 −8

Latitude

−10 −12 −14 −16 100 km

−18

−5.7

285

6.9 km

290

295

300 East longitude

b

305

310

14 6e 6h 6f

13a 6g

13b

6i 13c

3b 13d

6a, 7b 6b 6c 6k 12c 6j

7a 4a

6d

3a 8a

5a 8c

12d

9, 10

Figure 6.1 (a) Location of putative paleolake in the central Valles Marineris. Black contour defines the maximum ponding elevation, with overflow point indicated by “OP.” Yellow and orange marks indicate possible lacustrine bench features belonging to groups ‘A’ and ‘B’ respectively (see text). The underlying image is a THEMIS daytime IR mosaic, with gridded MOLA data shown in color. (b) The maximum ponding level contour provides context for other figures in this work. Dots correspond to the centers of figure images.

6.2

Topographic constraints on lake setting

Gridded MOLA elevation data can be used to delineate depressions and determine their overflow points independent of photogeologic interpretation. To do so, we must assume that the present topography is a reasonable representation of its state at the time of interest. The VMD lake, as suggested below, appears to have formed in the late Hesperian to early Amazonian, and therefore postdates the formation of large features such as the dichotomy boundary, the Tharsis rise (Anderson et al., 2001; Dohm et al., 2001; Phillips et al., 2001), and large impact basins. Furthermore, formation of the large circum-Chryse chaotic terrains and the Valles Marineris can­ yons themselves were probably largely complete prior to VMD ponding (Lucchitta et al., 1994 and Chapter 5). It thus seems reasonable to use present day topography for paleolake considerations. The VMD was first identified topographically by Smith et al. (1999), who noted that a gently sloping, 1 km high barrier would impede the flow of water from Melas Chasma to the circum-Chryse outflow channels. Our analysis attempts a more detailed delineation of the VMD and its overflow point. We used gridded MOLA elevation

166

Lakes on Mars

data at the maximum available resolution of 128 pixels/degree, or approximately 0.5 km/pixel at the low latitudes of the VMD. Since the VMD is 1800 km long and (on average) about 150 km wide, this resolution is sufficient to provide an accurate representation of the area inundated by the putative paleolake. This area was com­ puted by simulating incremental filling of the VMD with groundwater discharge assumed to have emanated from the deepest chaotic terrain in the depression, in Capri Chasma. Our algorithm takes as a starting inundation area all points that lie beneath a predetermined, low initial lake surface elevation. It then removes those points not in contiguous contact with the source location in Capri Chasma. Incremen­ tally greater surface elevations are then fed into successive iterations of the algorithm, resulting in incrementally larger inundated areas. This process continues until the execution of one further iteration that causes a large increase in the ponded area due to the inundation of neighboring depressions (i.e., chaos and channel regions east of the VMD). The surface elevation (–3560 m relative to the Martian datum) of the inun­ dated area just before this final iteration serves as a maximum ponding level for the VMD paleolake. Furthermore, the overflow point of the VMD can be identified as that location along the VMD boundary from which the large addition to the ponded area springs in the final iteration of our algorithm (Figure 6.1). The overflow level is significantly lower than canyon rim elevations (Figure 6.2) and yields an average ponding depth of 842 m, small compared to the canyon depth of 5–10 km. None­ theless, the maximum volume of water the lake could have held is 110,000 km3, similar to that of the large paleolake proposed for the Ma’adim Vallis drainage basin (Irwin et al., 2002), and on the same order as the Caspian and Black Seas.

6.3 Morphological evidence of a VMD paleolake Within the VMD, we detect two distinct groups of low topographic benches obser­ vable in high-resolution data sets from the MGS MOC camera (Malin and Edgett, 2001), THEMIS (Christensen et al., 2004), HRSC (Neukum and Jaumann, 2004), and the MRO CTX instrument (Malin et al., 2007). We also identify large-scale fluvial features located exclusively at the topographically derived overflow point in Eos Chasma. We proceed with a detailed description of bench and overflow morphologies, followed by a discussion of more ambiguous features on the VMD floor.

6.3.1 Bench morphology We identify bench features distributed across the VMD at two distinct topographic levels. An upper, more prominent group (which we define as group ‘A’) is observed several hundreds of meters above the adjacent depression floor, while a more local group (group ‘B’) is observed at lower elevations. A key collective property of the observed benches is that they both lie within (and for group ‘A’, close to) the area that would be inundated by a lake filled to the maximum ponding level of the depression (–3560 m).

Episodic ponding and outburst flooding associated with chaotic terrains in Valles Marineris

F

167

N

S

200 km 6 4

Elevation [km]

S

2

N

0 –2 –4 F

–6

290

295

300 305 Longitude [deg]

310

Figure 6.2 Elevation profiles along the floor (“F”) of the putative paleolake in the VMD and along plateau surfaces immediately to the north (“N”) and south (“S”). In the lower panel, plus signs mark individual MOLA shot elevations of bench features marked in yellow and orange in Figure 6.1. The upper panel is a THEMIS daytime mosaic.

Group ‘A’ benches This topographically higher system of benches is observed along the north and south walls of the VMD, and around massifs and chaotic blocks on its floor (yellow marks in Figure 6.1). The benches locally form a basal terrace beneath the easily recogniz­ able cliffs and spur-and-gully outcrops of wallrock around the VMD, and on wallrock inselbergs within the VMD, and is not observable in regions of the canyon west or east of the depression area. A relatively flat esplanade of resistant surface material overlain by talus materials separates the top of the benches from the adjacent wallrock outcrops above it. Bench morphology is similar to that observed in putative lacustrine settings elsewhere on Mars, including nearby Hydraotes Chaos (Ori and Mosangini, 1998) and Shalbatana Vallis (Di Achille et al., 2007, 2009). Bench widths vary from 10s m to 10s km, and heights vary from a few 100 m to about 1500 m relative to the adjacent depression floor (Figures 6.2 and 6.3). In some cases, multiple benches occur at different elevations locally. Benches are generally arcuate in planform, with surfaces that are level to concave. They have a smoother texture than the scarps they abut and

168

Lakes on Mars

a

Elevation, km

–3.5 A

C

B

F D

E

H

G

–4.0

–4.5

–5.0

0

20

40 Distance, km

60

80

b

–3.5 A

B

C

Elevation, km

D –4.0

–4.5

–5.0

0

20

40 Distance, km

60

80

Figure 6.3 HRSC images and topographic profiles; (a) north wall to center of west Capri Chasma, image (from orbit 0394_0009) is merged with topography; (b) north wall of east Coprates Chasma (orbit 0438_0000). Letters indicate points at which the profiles cross bench landforms. Note the similar elevations of these points.

Episodic ponding and outburst flooding associated with chaotic terrains in Valles Marineris

169

appear to be layered. The benches occur most commonly in east Coprates Chasma and Capri Chasma, where they are also most pristine. Canyon walls are considerably higher in the western part of the VMD (Figure 6.2), suggesting that mass wasting has degraded the lower walls and floor to a greater extent there. Locally, the widest bench in group ‘A’ skirts the SE-facing wall of Capri Chasma (Figure 6.4a). Talus deposits and an apparent alluvial fan (Figure 6.4b) cover much of the bench, yielding a gentle slope toward the canyon interior. However, where the bench is not overlain by these structures, it is remarkably level. A MOLA track along the bench front (track 5, Figure 6.4c) demonstrates a uniform absolute elevation of –3510 m. Further south in Capri Chasma is a set of benches along the northern edge of a massif in the chasma floor (Figure 6.5). Again, the benches have quite uniform

b

a

5

4 2 h

3

nc

Be

1

h

nc

Be

10 km c

–2500 1

Elevation [m]

–3000

2

4

3

5

–3500 –4000 –4500 –5000 20

40

60

80

Distance [km]

Figure 6.4 (a) Bench along SE-facing wall of Capri Chasma (THEMIS daytime mosaic). (b) A network of divergent valleys exits a tributary canyon in the Capri Chasma wall and traverses a fan-shaped mound on the bench (see black box in (a) for context; THEMIS VIS V06344002 and V16865001). (c) Elevation profiles of numbered MOLA tracks in (a). Arrows indicate bench location. Track 4 covers the length of the alluvial fan in (b). Track 5 demonstrates the uniform elevation of the bench front close to the maximum ponding elevation of the region (–3560 m).

170

Lakes on Mars

a

Bench

Bench

Bench

2

1

4

3

5

5 km b –3000

2

Elevation [m]

4 5

–3500

3 –4000 1 –4500 –5000 0

2

4

8 6 Distance [km]

10

12

14

Figure 6.5 (a) Benches along the north-facing scarps of mesas in Capri Chasma (THEMIS VIS mosaic). Narrow reentrants between benches appear to be related to amphitheater-shaped source regions cut into the mesas. (b) Elevation profiles of numbered MOLA tracks in (a). Arrow indicates bench location.

elevations. They are separated by narrow re-entrants that appear to be related to amphitheater-shaped source regions cut into the central massif, and which appear to have formed after bench emplacement. Further examples of group ‘A’ benches in the VMD are shown in Figure 6.6. These features occupy a 500 km stretch of the north wall of central Coprates Chasma, and a region further west in Melas and western Coprates Chasmata. Images taken by the CTX camera reveal horizontal layering within the upper caprock of the bench (Figure 6.7a). These layers follow the same trend as layers within the resistant cliffs and also spur-and­ gully outcrops of surrounding Valles Marineris wallrock. Points and re-entrants in the bench are mimicked by overlying layers in the wallrock proper (Figure 6.7b). These relations may suggest that group ‘A’ benches represent a sequence of layers in the wall rock separated from neighboring layers by less-resistant materials that tend to form talus deposits. This interpretation is uncertain, however, because a set of constructional layers

Episodic ponding and outburst flooding associated with chaotic terrains in Valles Marineris

a

b

171

c

e

d

h

f g

i

j

k

Figure 6.6 Examples of group ‘A’ benches (arrows; see Figure 6.1 for context). Scale bar in each panel represents 2 km. (THEMIS VIS: (a) V15168001, (b) V16079001, (c) V16990002, (d) V16915001, (e) V15755002, (f) V05683002, (g) V14432002, (h) V16878002, (i) V14357002, (j) V04659002, (k) V18862002.)

laid down by lacustrine processes would be expected to duplicate the pattern of prominences and re-entrants of the underlying canyon wall, and to be approximately horizontal, again mimicking wall layering. A possible advantage of the wall outcrop interpretation is that local benches with multiple elevations could be interpreted as outcrops of separate layers, implying that bench elevations would not be expected to conform to a single value. However, with a standard deviation of 310 m (around a mean of –4100 m), group ‘A’ bench elevations (Figure 6.2) are remarkably similar, especially

172 a

Lakes on Mars b

Figure 6.7 Horizontal layering in wallrock (“W”) and nearby group ‘A’ bench morphology (“A”); (a) north wall of west Capri Chasma (P03_002115_1651_XN_14S052W); (b) north wall of east Coprates Chasma (P07_003869_1655_XN_14S058W).

considering their E-W spread of 1280 km. Indeed, this standard deviation is smaller than the corresponding values of 1700 m and 560 m for the Clifford and Parker (2001) Contact 1 and 2 shorelines (see also Chapter 9), although these contacts have consider­ ably greater length and may have been warped due to the effects of polar wander (Perron et al., 2007). The standard deviation is also considerably smaller than those of the north and south plateaus adjacent to the VMD, which are 1850 m and 1220 m respectively (using the traces of Figure 6.2). Finally, bench elevations vary less than paleolake floor elevations: the standard deviation of all gridded MOLA data falling within the maximum ponding level is 450 m. We note, however, that sediment settling from a VMD paleolake would likely have smoothed the VMD floor to some degree. The lake hypothesis therefore eschews large standard deviations in floor elevation, and the moderate value of 450 m supports this view.

Group ‘B’ benches This system of benches, observed in Melas, Capri, and Coprates Chasmata (orange marks in Figure 6.1) lies below group ‘A’ benches and therefore also below the maximum putative lake level of the VMD. Their mean elevation is –4500 + 140 m. They have soft sinuous edges where they surround blocks of material and islands of wallrock within the VMD (Figures 6.8a and 6.8c). They reach heights of at most 200 m above the depression floor and are sometimes not high enough to be properly distinguished by MOLA or HRSC topography. Layering is not observed within these benches. In many areas, the benches are overlapped by eolian dunes (Figure 6.8c). Some group ‘B’ benches have marginal fractures suggesting collapse of the bench front (Figures 6.8a and 6.8c). Similar fissures have been observed along Gorgonum Chaos benches (Figure 6.8b; Howard and Moore, 2004) and around knobs in the northern plains (Tanaka, 1997; Tanaka et al., 2003). While the Gorgonum Chaos

Episodic ponding and outburst flooding associated with chaotic terrains in Valles Marineris

a

173

b

c

Figure 6.8 Examples of group ‘B’ benches (“B”). Fractures (arrows) in: (a) A low bench in east Coprates Chasma (THEMIS VIS V18113004); (b) At a smaller scale in Gorgonum Chaos (MOC M2101910; Howard and Moore, 2004); (c) A low bench around mesa in southeast Coprates Chasma (“a” denotes overlying aeolian dunes; CTX image P03_002115_1651_XN_14S052W).

fissures are only about 10 m across, compared to approximately 250 m for the Coprates and Capri features, their similar morphologies may suggest that, like in Gorgonum Chaos, a rotation and collapse of the bench margin occurred in response to lowering of the neighboring floor.

6.3.2 Indications of lake overflow Perhaps the strongest evidence for VMD ponding is the coincidence of the MOLA-derived overflow point in Eos Chasma with a set of fluvial features

174

Lakes on Mars

a

b

Figure 6.9 Region where overflow from VMD is likely to have occurred. (a) THEMIS daytime IR mosaic and (b) shaded MOLA topography. White line indicates maximum ponding elevation contour. The overflow point is associated with flow features, some of which are shown in greater detail in Figure 6.10 Labeled boxes in (a) show position of Figure 6.10 images.

(Figures 6.9 and 6.10). Carr (1996) noted that these “sculpted, streamlined” forms could be attributed to catastrophic release of ponded water in an easterly direction, a view supported more recently by Greeley et al. (2003). Large-scale fluvial features within the maximum ponding contour appear to be limited to this region. The orientation of the features suggests a convergence of flow from a relatively wide area in the west toward a narrow eastward exit point. Features appear to be predomi­ nantly erosive, suggesting an energetic flow and are analogous in appearance to terrestrial catastrophic flood bars (e.g., Baker 2009; Carling et al., 2002). MOLA profiles near the overflow point suggest a channel depth of about 100 m, implying that prior lake levels could have exceeded the nominal –3560 m ponding level by at least this amount. The general lack of well-defined bench forms above the ponding level suggests that higher lake levels attained over the history of VMD ponding were either too brief for bench formation or that freezing inhibited shoreline development. The path of overflowing water as inferred from morphology and topography appears to lead directly to the channel connecting Eos Chasma with the next large chaotic terrain, Aurorae Chaos (Figure 6.11). Aurorae Chaos lies in a relatively

Episodic ponding and outburst flooding associated with chaotic terrains in Valles Marineris

a

b

175

c

d

Figure 6.10 Flow features in the vicinity of the VMD overflow point. (a) Scour marks in the immediate neighborhood of the overflow point (“OP”). (b) Fluvial erosion upstream of the overflow point. (c) Similar features downstream of the overflow point, including braided channels (bottom right). (d) Flow toward the overflow point (to right of image, not shown) appears to have been parted by an obstacle (far left). Scale bars in all images correspond to 5 km. Image contexts are shown in Figure 6.9. ((a) THEMIS VIS V18587004, (b) THEMIS daytime IR I04359002, (c) VIS V03535003, (d) VIS V17389003.)

shallow depression, so limited ponding would be required before the flow continued northward toward Hydraotes Chaos. The floor of Hydraotes forms a significant depression and contains the lowest point of the circum-Chryse chaotic terrains. Indeed, the VMD and Hydraotes are two of the lowest regions at low latitudes on Mars (Candor Chasma is a third). The resulting likelihood of ponding in Hydraotes is supported by benches encircling chaos blocks (Ori and Mosangini, 1998). Bench morphology is similar to that observed in Capri and Eos Chasmata. Flow northward out of Hydraotes Chaos would produce some shallow ponding in Simud Valles and then continue to the northern plains (Figure 6.11). Thus, only two of the large chaos regions linking Coprates Chasma to the northern plains are capable of ponding water to significant depths, and the same two regions (and no others along this flow route) have been found, thus far, to exhibit benches along massif and/or canyon walls. The action of lacustrine processes may have strongly influenced this relationship. Further evidence of VMD overflow is a smooth deposit near the entrance of Aurorae Chaos designated by Witbeck et al. (1991) as chasma channel material including “material transported and deposited by flowing water.” Flow downstream

176

Lakes on Mars

Figure 6.11 Simulated inundation of chaotic terrains and channels downstream of the VMD. The colored region indicates the area that would be inundated if groundwater discharge at the eastern end of Coprates Chasma (bottom left) was sufficient to overtop the maximum ponding level of the VMD and eventually reach the northern plains (top right). Color denotes water depth (see color bar). The deepest ponding occurs in the VMD and Hydraotes Chaos. Gridded MOLA data were used to calculate the inundation area and depth, and to produce the shaded background image.

of the VMD overflow point likely spread laterally in this region and deposited some of its sediment load before continuing on to Aurorae Chaos.

6.3.3 Additional evidence of ponding A further set of observations appears to support the interpretation of ponding in the VMD. These features (described below) may not be independently sufficient indica­ tors of ponding, but taken with the most compelling evidence already described, they appear to be significant. Specifically, we discuss intra-lake massifs, floor morpholo­ gies, and landslides.

Episodic ponding and outburst flooding associated with chaotic terrains in Valles Marineris

177

Massif morphology A survey of massifs on the VMD floor suggests an elevation-dependent dichotomy in peak morphology. We surveyed all massifs covered by THEMIS (VIS) images, including blocks of chaotic terrain in Eos and Capri Chasmata and more isolated massifs in Melas and Coprates Chasmata. Each massif was rated on a continuous, qualitatively based scale from 0 to 1, with 0 corresponding to flat-topped or mesa-like morphologies and 1 corresponding to steep, rugged peaks dominated by spur-and­ gulley morphology similar to that observed on canyon walls. The results, grouped into 11 bins (Figure 6.12), show that massifs with peaks above the maximum ponding elevation of the VMD are predominantly sharply peaked. The few mesa-like b

70 60

high massifs low massifs

50 Frequency

–2000

–2500 Massif peak elevation (m)

a

40 30 20

–3000

–3500

–4000

–4500

10

–5000

0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Massif peak index

c

0

1

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Massif peak index

1

d

2 km

2 km

Figure 6.12 (a) Morphology of massif peaks in the VMD expressed as a qualitative index ranging from 0 for flat-topped peak morphologies to 1 for sharply crested, rugged peak morphologies. “Low massifs” are those with peak elevations below the maximum ponding level of the VMD; “high massifs” have peaks above this level. (b) Peak elevations of Melas and Coprates Chasma massifs plotted against the peak morphology index show a high concentration of sharply crested massifs above the maximum ponding elevation of the VMD paleolake (dashed line), and a predominantly subdued morphology below this elevation. In (c) and (d) are examples of massifs with peak morphology indices of 0 and 1 respectively (THEMIS VIS V16478004 and V06881001).

178

Lakes on Mars

morphologies in this group appear to be limited to chaotic blocks which preserve the flat topography of the original plateau from which they were detached. Massifs with peaks below the maximum ponding level are more varied but are skewed toward flatter morphologies. Low massifs with peaked morphologies seldom exhibit rugged spurs and gullies but consist more commonly of simple ridges grading smoothly into the canyon floor. A χ2 test indicates that the probability of the sub- and super-ponding level populations being the same is less than 10−4, confirming that the observed dichotomy is statistically significant. It is possible that the dichotomy is due to differences in material strength between layers of bedrock making up the massifs, but the elevation of the dichotomy close to the maximum ponding level of the VMD suggests a lacustrine origin. We envisage that massifs with peaks below the ponding level may once have exhibited sharply created morphologies like their taller counter­ parts, but submergence during ponding resulted in a smoothing of this morphology. We note that a dichotomy in massif scarp (as opposed to peak) characteristics was also observed, with tall massifs exhibiting spur-and-gully morphology along the upper parts of their walls, with smoother aprons below (e.g., Figure 6.12b). This dichotomy may be due to the formation of talus deposits, but its consistent appearance at the maximum ponding level of the VMD region, and its often sharp delineation, may suggest erosion through shoreline or channel flow processes.

Floor morphology On a regional scale, the VMD floor is flat relative to the neighboring canyon plateaus (Figure 6.2), suggesting that it has been substantially modified since its original formation as a graben. Alternatively, regional changes in plateau elevation following graben formation were not imparted to the canyon floor. However, the abundance of processes able to modify the floor, including mass wasting, infill from eolian and airfall ash materials, and chaotic terrain processes, supports the former hypothesis. Lacustrine activity may have contributed to the current floor morphology. A pattern of polygonally arranged ridges is found on the floor of eastern Coprates adjacent to the bench of Figure 6.8a, and to the east and north in the lowest parts of the Capri Chasma floor (Figure 6.13). Ridge morphology appears to resemble that of Hesperian volcanic plains wrinkle ridges, but an order of magnitude more closely spaced (few km vs. few 10s km). MOLA tracks crossing the Capri Chasma ridges do suggest, in some cases, an offset in elevation on either side of each ridge as expected for movement along thrust faults, the most widely accepted mechanism for wrinkle ridge formation (Golombek et al., 1991; Watters, 1991). Possibly, the desiccation of a volatile-rich floor deposit caused the floor deposit to sink, forcing it to occupy a more confined space (canyon width decreases with depth). This compression may have led to polygonally arranged thrust faults. Ridges are found in large, smooth areas of the canyon floor devoid of chaotic blocks that might otherwise have interrupted the lowering and compression process. The marginal fractures on group ‘B’ benches may support the floor lowering hypothesis: they are found in Capri Chasma and west Eos Chasma adjacent to ridged areas, and their morphology suggests rotation and collapse of the bench margin in response to lowering of the neighboring floor. We note

Episodic ponding and outburst flooding associated with chaotic terrains in Valles Marineris

179

a

b

c

d

Figure 6.13 Polygonally arranged ridges on Coprates and Capri Chasmata floors. Note the association of ridges in (b) and (c) with marginal fractures in benches at the top of each image. In (d), ridges are associated with cones. ((a) MOC NA e0200985, (b) THEMIS VIS V18113004, (c) V19648002, (d) V18625001.)

that because wrinkle ridges on Mars are most commonly found in volcanic settings, it is possible that volcanic processes formed the VMD floor ridges. Although no obvious volcanic source is observed for such a large intra-canyon deposit, a more detailed study of volcanic processes in the region might help to resolve this issue. Finally, we observe another feature associated with the Coprates and Capri floor ridges, namely small cones, a few 100 m–1 km in diameter, frequently (but not exclusively) with summit pits (Figure 6.13d). Several candidate cone types exist, including cinder cones, tuff cones, rootless cones, pingos, and mud volcanoes. Mud volcanism may be a reasonable explanation in the context of a compressional envir­ onment (Komar, 1991), and similar cones have been attributed to a mud volcano origin in Acidalia Planitia, north of the Chryse chaotic terrains (Farrand et al., 2005; Tanaka et al., 2003). However, once again, we note that a volcanic interpretation of these features (as small cinder cones or tuff cones) is plausible.

Landslides Much controversy has surrounded the wallrock landslides of the Valles Marineris. These features involve relatively large volumes of material (10–1–104 km3; McEwen, 1989) and, as with large terrestrial landslides, their formation appears to have required coefficients of friction much lower than that of dry rocky material. A number of

180

Lakes on Mars

Figure 6.14 An example of a long runout landslide deposit within the boundary of the VMD paleolake. The multi-lobed debris apron exhibits longitudinal lineations, lateral levees, and lobate flow fronts (mosaic of THEMIS daytime IR images I01264001, I06095001, and I07929021).

processes have been invoked to explain this issue (e.g., Lucchitta, 1979, McEwen, 1989), and although no strong consensus has been reached, processes involving a lubricating fluid seem most plausible (Legros, 2002). Some of the best examples of long, sinuous landslide debris aprons are found in the VMD (e.g., Figure 6.14). A study by Quantin et al. (2004) suggests that landslide ages vary significantly, ranging from only 10s of Ma to about 3 Ga. Although visual inspection does, indeed, suggest some range in ages, the quantitative constraints of Quantin et al. (2004) are potentially problematic because areas used for crater counting may be too small to take advantage of reliable crater populations, and may not properly represent the landslide deposit. The median counting area is only 60 km2; most (77%) counting areas are less than 200 km2. Furthermore, most (70%) of the counting areas cover only 20% or less of their respective total landslide area (the median value is 11%). Because of these small counting areas, the maximum crater diameter contributing to a given landslide age estimate has a median value of only 360 m (and a mean of 560 m), below the roughly 1 km minimum thought to be necessary to avoid secondary cratering issues (McEwen and Bierhaus, 2006). Even if small (likely secondary) craters do turn out to be reliable for age dating (e.g., Hartmann, 2007), they are most reliable for younger surfaces only: many small craters on older surfaces are likely to have been modified or completely obliterated by erosive processes. Impor­ tantly, the effect of this removal is to yield artificially young ages. The preservation of ridges and grooves on some landslide deposits suggest, however, that modification

Episodic ponding and outburst flooding associated with chaotic terrains in Valles Marineris

181

may have been minimal. To test this idea, we used elevation data from individual MOLA tracks to analyze longitudinal ridges on two large landslides in eastern Melas Chasma. We obtained amplitudes (crest to trough) of 50–200 m and wavelengths of 100 m to about 2 km (commensurate with the measurements of Lucchitta, 1979). Conversely, small craters (diameters from a few 10s m–500 m,) are likely to have had depths (when fresh) from only a few meters to about 50 m (McEwen et al., 2005; Pike and Wilhelms 1978). It is therefore possible that a significant fraction of small craters, especially those near the lower end of the above range, have been severely degraded, or even obliterated, while leaving ridges and grooves intact. Finally, Valles Marineris landslides have a wide spatial distribution, implying that regional hetero­ geneity in secondary crater populations (e.g., McEwen et al., 2005) could yield spurious age variations among widely separated counting areas, especially when these areas are small. The above concerns regarding landslide age raise once more the possibility, originally inferred from compelling morphological evidence (Shaller et al., 1989), that long runout debris aprons may have formed subaqueously either during those epochs when (possibly closed) precursor canyons of the Valles Marineris contained lakes, or more recently when the VMD paleolake was present. Submarine landslides are the only terrestrial mass-wasting features that involve deposit volumes and runouts of a similar scale to their Martian counterparts (e.g., Lipman et al., 1988). They also exhibit similar morphologies (such as ridges and troughs, slump-related head scarps, and marginal levees; e.g., McMurtry et al., 1999; Shaller et al., 1989), although we note that in rare cases terrestrial landslides also exhibit such features (e.g., the Sher­ man landslide, Shreve, 1966). In both submarine and Martian environments, the failure of relatively steep wallrock abutting a gentle slope results in two distinctive morphologies: large faulted blocks where the initial failure occurs, and long runout debris aprons where fluidized material overruns the gentle slope. The ratio of final drop height to runout length is a very rough measure of landslide mobility (smaller ratios imply higher mobility). Legros (2002) demonstrated that the value of this ratio for Martian landslides lies at the upper end of the range for submarine landslides, and at the lower end of the range for subaerial volcanic slides. Legros (2002) also argues that a potentially more useful measure of mobility is the ratio of runout length to deposit volume: this measure also shows some overlap between Martian landslides and both subaerial and submarine terrestrial slides, although only the latter extend into the Martian volume range. Interestingly, a plot of Martian landslide runout elevations (Figure 6.15) shows that most landslides with long-runout, low friction-coefficient debris aprons lie beneath the maximum ponding level of the VMD. Conversely, most landslides without debris aprons lie above this elevation. Importantly, there are no exceptions to this dichotomy within the entire southern Valles Marineris canyon (Ius to Eos Chasmata) containing the VMD. Alternative interpretations of this pattern may, however, be possible. For instance, long runout debris aprons might be expected to form in only the widest canyons. However, landslides appear to change from slump block to debris apron morphology close to the initial failure scar, so that at least some apron-like morphol­ ogy should be expected in many of the narrower locations where wholly blocky

182

Lakes on Mars

0

landslides without debris aprons

landslides with debris aprons

Floor elevation (m)

–1

–2

–3

VMD ponding elevation

–4

–5

–6 0

10

20

30 40 Landslide ID

50

60

Figure 6.15 Canyon floor elevations immediately adjacent to Valles Marineris landslide deposits both with and without long runout debris aprons. Landslide ID corresponds to the numbering of Quantin et al. (2004). Of the five landslides with debris aprons lying above the ponding elevation of the VMD, the leftmost four are in a separate, closed canyon (Tithonium Chasma), and the fifth is on the low plateau connecting Melas and Candor Chasmata. The two landslides that do not have debris aprons and which lie below the ponding elevation are also in a separate, closed canyon (Hebes Chasma).

landslides are instead observed. If the formation of debris aprons depended on drop height, then perhaps they would be expected in only the deepest, and therefore the widest, canyons. However, there is no obvious distinction between drop height for debris apron landslides and blocky landslides: drop height data for the landslides of Figure 6.15 yield almost identical means and standard deviations (6.7 + 1.2 km and 6.3 + 1.3 km for debris apron and blocky slides, respectively). Perhaps debris aprons occur only where the runout slope is relatively smooth and gentle, conditions that may be prevalent within the VMD because of prior lacustrine deposition. However, even under this scenario, the likely requirement of some interstitial fluid to allow long runouts still applies (Legros, 2002). It is not clear how such a fluid could be main­ tained in the canyon walls over the �3 Ga period of landsliding suggested by Quantin et al. (2004). Furthermore, the presence of unaltered olivine in wall outcrops may suggest a predominantly dry history (Christensen et al., 2003). However, this latter inference appears uncertain when considering observations of unaltered olivine in regularly wetted outcrops along the Grand Canyon’s Colorado river (Hamblin, 1994). It seems more plausible that most landsliding occurred not long after the Valles

Episodic ponding and outburst flooding associated with chaotic terrains in Valles Marineris

183

Marineris canyons opened, when groundwater would likely be present in the freshly exposed wallrock, and ponded bodies of water would be in place due to groundwater discharge at active chaotic terrains or along the same fractures responsible for canyon formation. Indeed, rapidly changing water levels in ponded bodies (due to ground­ water discharge contributions or lake outburst losses) may themselves have served as triggers for landsliding, as observed in some terrestrial cases (e.g., the Vaiont land­ slide, Hendron and Patton, 1985; Müller, 1964). A possible impediment to water-related processes in landsliding is the need to explain debris apron landslide morphology observed elsewhere in the Valles Mari­ neris. Our suggestion is that subaqueous formation may turn out to be a viable explanation for all such morphologies. Other paleolakes in the Valles Marineris have already been identified (Lucchitta, 1987; Quantin et al., 2005; Weitz et al., 2006) and added to the list of lakes suggested for nearby chaotic terrains and outflow channels listed above.

6.4

Discussion

The primary evidence for ponding in the VMD consists of bench forms and overflow morphology. Group ‘A’ bench elevations occupy a narrow range that approximately matches the maximum ponding level of the VMD, and bench morphology is flat relative to neighboring wallrock. We are thus led to suggest that the bench enclosed a putative paleolake that enhanced the erosional face of the cliffs. The sinuous, locally collapsed edge of Group ‘B’ benches, together with their lack of layering, suggests that these benches are not part of the Valles Marineris wallrock and are instead morphologically analogous to terrestrial lacustrine terraces. Thus, while the two bench groups likely formed under different conditions, they both appear to have been influenced by ponded water in the VMD. Scours and channels at the topographically derived overflow point of the VMD strongly support the ponding hypothesis. We explore VMD ponding further by constructing a possible history of lacustrine activity consistent with stratigraphic and morphological constraints.

6.4.1 History of lacustrine activity Temporal constraints Our hypothesis is that groundwater discharge at Capri and Eos chaotic terrains, or at fractures along the boundaries of the VMD, ponded before being released into outflow channels and onward into the large northern plains basins. Chaotic terrains are traditionally thought to be sources of catastrophic groundwater discharge, although problems regarding high required aquifer permeabilities remain (Harrison and Grimm, 2008). Large-scale fractures, as observed at Mangala and Athabasca Valles, are also likely sources of groundwater discharge. Recent inferences of past groundwater flow in fractures in the Valles Marineris region (Chapman et al., 2009; Davatzes and Gulick

184

Lakes on Mars

2007; Okubo and McEwen 2007; Okubo et al., 2009; Treiman 2008) suggest this may have been a source of water for the circum-Chryse outflow channels also. The VMD would have ponded following what Lucchitta et al. (1994) and Schultz (1998) suggest was a secondary stage of Valles Marineris trough development that opened the long, narrow, graben-controlled canyons of Ius and Coprates Chasmata. Ponding would thus have contributed to the episode of “late” lake activity described in Chapter 5 by Lucchitta. However, our suggested VMD paleolake age of late Hesper­ ian to early Amazonian differs somewhat from the age of late lake formation sug­ gested by Lucchitta (Chapter 5). The latter age depends primarily on crater counts of Ganges Chasma landslides and canyon floors (Quantin et al., 2004), and so may not be representative of the remainder of the Valles Marineris. Furthermore, the Ganges chasma floor ages estimated by Quantin et al. (2004), and the remaining two age measurements of 3 Ga or older estimated by these workers for the Valles Marineris, all conflict with younger ages estimated for the same areas by other mappers (Rotto and Tanaka 1995; Witbeck et al., 1991). A consequence of these age issues is that no clear difference between landsliding and late lake formation ages is readily apparent. Because few landslides underlie other morphologies (rare examples include overlap with interior layered deposits), it may be safe to assume that they are among the youngest features in the Valles Marineris. However, the question of precisely how young they are, and how close in age they are to other features (such as late lake floors), remains uncertain. Topographic considerations also elucidate the timing of lacustrine activity. Capri Mensa, an interior deposit separating Capri and eastern Eos Chasmata, must have been in place prior to ponding. This is because the paleolake overflow point consists of channel features incised into the interior deposit: indeed, the deposit likely con­ tributed significantly to the damming of groundwater discharge. Finally, the large channel and chaos structures downstream of Eos Chasma must have been in place prior to ponding. If Eos Chasma was closed at its eastern margin during ponding, water levels would not have been limited by the observed overflow point, and shore­ line features would likely have formed along eastern Eos Chasma walls and massifs. Thus, the ponding of the VMD likely attended the late stages of groundwater discharge in Capri and Eos Chasmata. The appearance of benches at multiple eleva­ tions suggests that paleolake levels changed episodically, but remained at each eleva­ tion long enough for shorelines to develop. Such behavior is expected if groundwater discharge occurred episodically and produced varying quantities of water, a likely hydrological scenario (Andrews-Hanna and Phillips, 2007; Harrison and Grimm, 2008). Those ponding episodes related to the lower, more pristine benches in Capri, eastern Coprates, and western Eos Chasmata, likely postdate the damming of the entire VMD, an event which produced the more degraded benches observed at higher elevations and as far west as Melas Chasma. Late-stage ponding may have recurred repeatedly into the early Amazonian (depending on the supply of groundwater discharge) without necessarily producing large overflow events and attendant cata­ strophic flooding. Finally, we estimate how long it may have taken to fill the VMD paleolake, and how long the lake may have lasted. Assuming that groundwater discharge at chaotic

Episodic ponding and outburst flooding associated with chaotic terrains in Valles Marineris

185

terrains was the principle source of water for the paleolake, work by Harrison and Grimm (2008) can be used to estimate the time needed to deliver 110,000 km3 of water to the VMD. Harrison and Grimm (2008) and Andrews-Hanna and Phillips (2007) suggest that groundwater discharge would be episodic, with each discharge event ceasing due to refreezing of discharge pathways, and the next event beginning due to a resurgence of aquifer groundwater pressures. We estimate conservatively that the VMD could have been filled to its maximum ponding level in a minimum of about 200 years, with water produced by about 20 discharge events (Harrison and Grimm, 2008). This duration could be reduced if water was also found to have come from spring sources or runoff. Although we find no clear evidence of precipitation as a major contributor, it has been implicated in the formation of fluvial morphologies elsewhere in the Valles Marineris and its adjacent plateau (Mangold et al., 2004; Quantin et al., 2005; Weitz et al., 2008) and so cannot be ruled out. Our time estimate could be reduced still further if groundwater discharge took advantage of the higher effective permeabilities afforded by large-scale fracture flow. Assuming, again conservatively, that ambient atmospheric conditions were akin to those of today, we also suggest that major episodes of outburst flooding must have reached completion within a few thousand years of lake formation: this is the approximate timescale for complete freezing of the lake (Kreslavsky and Head, 2002). It is quite possible, however, that the sublimation of ice from a frozen VMD lake eventually cleared the way for a later set of discharge events and a repeat of the ponding, and perhaps overflow, processes. For this reason, it is not possible to conclude that all the putative lacustrine features within the VMD formed within a single period of ponding lasting only a few thousand years. Instead, multiple stages of ponding, separated by millions of years or more, may have produced independent sets of lacustrine features. Despite this complexity, it is probably still reasonable to expect the topographically lower set of benches (group ‘B’) to have formed after the group ‘A’ benches.

Style of lacustrine processes Group ‘A’ benches appear to be composed of a series of wallrock-derived layers of variable strength (the layers were likely initially exposed by collapse of the Valles Marineris floor; Tanaka and Golombek, 1989). As suggested above, their common height close to the maximum ponding elevation of the VMD and their flat morphology relative to neighboring wallrock suggest that a hydrostatic agent contributed to erosion of weaker layers, enhancing terrace formation at the observed elevation. Wave action on Mars may have been relatively weak (Kraal et al., 2006), and erosion by ice-ridge pile up may be more likely (Barnhart et al., 2005). Identification of ridges of material pushed ashore by a moving ice-plug would support the latter process, while the former process may have been enhanced in the VMD by the long fetch of the inundated area, the likely high winds through the canyons, and water currents attending episodes of emptying and refilling of the VMD. Group ‘B’ benches appear to be constructional, based largely on their pattern of broad, convex projections and narrow, sharp re-entrants (Howard, 1994, 1995; e.g.,

186

Lakes on Mars

Figure 6.8). Thus, we favor the process of bench formation invoked by Howard and Moore (2004) for the Gorgonum chaos benches that so closely resemble eastern VMD benches, namely the compression of sediments by an overlying floating ice cover. In this scenario, sediments may be deposited originally through fluvial, lacustrine, or airfall processes. In the case of the Valles Marineris, an important contributor is likely to be mass wasting in the form of landsliding and talus accumulation. Indeed, the lobate form of some benches suggests initial emplacement by small landslides, followed by significant deformation or degradation. An alternative (or perhaps additional) formation process for group ‘B’ benches is the desiccation of a volatile-rich layer and associated floor collapse. Although the benches appear to be constructional, collapse is compatible with marginal bench fractures and nearby reticulated wrinkle ridge patterns. However, as discussed above, we cannot rule out a volcanic origin for these ridges. Finally, we note that the lateral distribution of VMD benches is skewed toward the east, an observation that may be explained by the maximum ponding level of the region. One of the shallowest points of a putative lake filled to this level is approxi­ mately halfway along Coprates Chasma (at approximately 301º E; Figures 6.1 and 6.2). Depths here range from 100 to 200 m, compared to a maximum lake depth (excluding crater floors) of about 1650 m. If lake water was provided predominantly by groundwater discharge at Capri and Eos chaotic terrains, ponding would have extended westward to Melas Chasma and western Coprates Chasma only if water levels came within 200 m of the maximum ponding elevation. The obstruction (which appears to be a landslide deposit spanning the width of the canyon) is estimated by Quantin et al. (2004) to be no more than a few 100 Ma old, suggesting emplacement some time after lacustrine activity. However, the concerns with landslide age-dating described above suggest that the deposit may be considerably older. (If the landslide is, indeed, young, and it accounts for most of the topographic rise, then it cannot have caused the skewed distribution of bench features, but if the pre-landslide surface was higher than the surrounding canyon floor, then it may nonetheless have limited lacustrine activity to the west.)

Outflow channel floods As suggested by groundwater modeling studies (Carr 1979; Hanna and Phillips 2005; Harrison and Grimm 2008; Manga 2004), the direct production of outflow channel floods by groundwater discharge tests the limits of acceptable aquifer permeabilities. As already stated, most chaotic terrains lie in topographic depres­ sions (e.g., Harrison and Grimm, 2008), and any groundwater discharged to the surface after the formation of these depressions must have ponded locally. As suggested by Chapman and Tanaka (2002), the presence of the outflow channels themselves also suggests ponding: catastrophic terrestrial floods usually form due to the failure of natural or man-made dams. It is therefore unsurprising that initial suggestions of various dam failure mechanisms for Martian channel flood produc­ tion (Masursky et al., 1977; McCauley et al., 1972) are today experiencing renewed consideration (e.g., Chapman and Tanaka 2002; Chapman et al., 2003;

Episodic ponding and outburst flooding associated with chaotic terrains in Valles Marineris

187

Payne and Farmer 2001; Rice and Edgett 1996). We suggest that the initial analogy between the Martian outflow channels and the Channeled Scablands in Washington State (Baker and Milton, 1974) can be made complete by including a similar outburst delivery mechanism attributed to the latter floods, namely the collapse (or floatation) of a melting ice dam. The substantial body of literature on terrestrial glacial outbursts offers a valuable opportunity to test this mechanism quantitatively in the Martian context.

Alternative water sources The present day maximum ponding elevation of the VMD corresponds roughly to the –3760 + 560 m elevation of the Contact 2 northern plains shoreline (Clifford and Parker, 2001; see also Chapter 9), suggesting that the Valles Marineris paleolake might once have been part of a much larger northern ocean. This hypothesis is also supported by the similarity of group ‘B’ bench morphology with proposed wavecut terraces in the northern plains (Parker et al., 1993). Although water levels would initially need to exceed –3560 m in order to flood Melas and Coprates Chasma, this value is well within the error of the northern ocean shoreline elevation. As ocean levels later fell, water would have been trapped within the VMD, with breakout eventually occurring at the observed overflow point due to dam collapse. The draw­ back of this hypothesis is the lack (thus far) of evidence supporting initial flow of water into, rather than out of, the Valles Marineris, and its inundation of other canyons exposed to the northern plains (such as Ganges Chasma and Hydaspis Chaos).

6.4.2 Non-lacustrine formation hypotheses Although the evidence for ponding appears to be compelling, potential problems with a lacustrine interpretation ought to be addressed. We therefore consider the strength of alternative formation hypotheses, and arguments concerning the mineralogy of the region.

Faulting The formation of the Valles Marineris involved substantial contributions from tectonic processes (Blasius et al., 1977; Plescia and Saunders 1982; Schultz 1991). The possibility thus exists that the VMD benches are derived from motion along normal faults. Collapse related to tension fracturing may provide a plausible origin for group ‘A’ benches (Tanaka and Golombek, 1989). However, in many cases the margins of these benches (and more frequently those of group ‘B’) are not linear as expected for fault scarps, but are arcuate, sometimes completely ringing massifs. Second, the orientation of many benches bears no spatial relation­ ship to visible fault boundaries such as the linear Coprates Chasma walls (those benches that appear along linear canyon walls cannot unambiguously be attributed a tectonic origin). Finally, the approximately uniform elevation of the benches (in contrast to their significantly undulating regional context) is not easily explained by a tectonic collapse origin.

188

Lakes on Mars

Backwasting of resistant beds As suggested above, group ‘A’ benches appear superficially to be composed of backwasted resistant cliffs cropping out from the lowest levels of the plateau wallrock. Once again, though, their flat morphology relative to neighboring wallrock, and their height approximately matching that of the maximum ponding level of the VMD, suggest that they enclosed a putative paleolake that enhanced the erosional face of the cliffs. Group ‘B’ benches appear to be constructional by virtue of their soft, sinuous edges and narrow re-entrants, making a backwasting origin unlikely. Additionally, deposits of removed material in the form of talus cones and landslides are absent from the floors immediately in front of group ‘B’ bench margins. Fluvial features are also absent from group ‘A’ and ‘B’ bench scarps. A small number of fluvial features are on the relatively flat upper surfaces of group ‘A’ benches, but this indicates removal of material from the canyon wall behind the bench, rather than from the bench itself.

6.4.3 Mineralogy Possible evidence against ponding in the VMD lies in the cursory conclusion that no mineralogic support is forthcoming. Thus far, almost no evidence of water-related minerals such as sulfates and phyllosilicates has been found within the boundary provided by the maximum ponding level (Bibring et al., 2006; Mangold et al., 2007a). Also, olivine (which is unstable in aqueous conditions) is found in high concentrations in isolated basaltic outcroppings in some Chryse Planitia channels, although not within the VMD (Christensen et al., 2003; Edwards et al., 2007). The absence of phyllosilicates is not unexpected. These minerals are thought to have formed predominantly in the Noachian before outflow channel formation and the proposed surface ponding (Bibring et al., 2006). Although Noachian layering is exposed in the Valles Marineris walls, the presence (or absence) of phyllosilicates in these layers is likely a consequence of their Noachian origin prior to canyon forma­ tion. If, instead, phyllosilicate formation was still occurring in parts of the planet during VMD ponding, the absence of these minerals in the VMD, together with their requirement of a long-term source of water (Mangold et al., 2007b) may merely limit VMD ponding to short periods of time, in agreement with the episodic nature of lake formation and drainage described above. Sulfates, on the other hand, are present in parts of the Valles Marineris but are closely associated with freshly exhumed scarps in interior layered deposits. Their absence from gently sloping and dust-mantled paleolake floors in the VMD thus seems reasonable. Ultimately, though, alternative processes such as groundwater alteration or ice-magma interactions may explain both the sulfates and their correla­ tion with interior layered deposits (Chapman and Tanaka 2002; Gendrin et al., 2005). In this case, the lack of sulfate production during subsequent lake formation may be the result of unfavorable mineralogic and thermal conditions. Finally, olivine-rich basalts in Chryse Planitia appear to conflict with the occur­ rence of surface flooding: outcrops occur in channel walls and occasionally on channel floors. Perhaps transient floods are not sufficiently long-lived to produce

Episodic ponding and outburst flooding associated with chaotic terrains in Valles Marineris

189

the weathering necessary to destabilize olivine. Furthermore, terrestrial flows lasting significant periods have been observed to leave olivine basalts unaltered (e.g., Grand Canyon olivine basalts at Lava Falls; Hamblin, 1994). Nonetheless, no olivine-rich outcrops have yet been identified in circum-Chryse chaotic terrain depressions that likely held standing bodies of water (Edwards et al., 2007).

6.5

Conclusions

The observation of numerous bench structures along the walls of Melas, Coprates, Capri and Eos Chasmata suggests the past presence of a body of standing water, which was partially released, perhaps episodically, into outflow channels and thence onto the northern plains. Coincidence of the MOLA-derived basin overflow point with channel erosion strongly supports the ponding hypothesis. If the maximum water levels are assumed to be limited by the present day maximum ponding level, and if the elevation of the canyon floor has not changed significantly since the early Hesperian, then the average depth of the completely filled paleolake is estimated to be 840 m, with a total volume of 110,000 km3. Although the most likely source of ponded water was groundwater discharge at chaotic terrain in Capri and Eos Chasmata, or at largescale fractures bounding the Valles Marineris canyons, an alternative supply from a northern plains ocean, which has a putative shoreline at a similar elevation to the Valles Marineris paleolake, is possible. Finally, the likelihood of ponding in the VMD (and in other depressions associated with chaotic terrains) supports the idea that failure of ice- or sediment-dammed lakes may be primarily responsible for outflow channel flooding. Direct production of channel floods by groundwater discharge remains difficult to explain, and the buffer­ ing effect provided by surface ponding may circumvent this problem.

References Anderson, R.C., Dohm, J.M., Golombek, M.P., Haldemann, A.F.C., Franklin, B.J., Tanaka, K.L., et al., 2001. Primary centers and secondary concentrations of tectonic activity through time in the western hemisphere of Mars. J. Geophys. Res. 106 (E9), 20563–20586. Andrews-Hanna, J.C., Phillips, R.J., 2007. Hydrological modeling of outflow channels and chaos regions on Mars. J. Geophys. Res. 112, doi:10.1029/2006JE002881. Baker, V.R., 2009. Channeled scabland morphology. In: Burr, D., Carling, P., Baker, V. (Eds.), Megaflooding on Earth and Mars. Cambridge University Press, New York. Baker, V.R., Milton, D.J., 1974. Erosion by catastrophic floods on Mars and Earth. Icarus 23, 27–41. Barnhart, C.J., Tulaczyk, S., Asphaug, E., Kraal, E.R., Moore, J. 2005. Ice-ridge pile up and the genesis of Martian “shorelines”, 36th Lun. Plan. Sci. Conf., Abstract #1560. Bibring, J.-P., Langevin, Y., Mustard, J.F., Poulet, F., Arvidson, R., Gendrin, A., et al., and the OMEGA team, 2006. Global mineralogical and aqueous Mars history derived from OMEGA/Mars Express data. Science 312, 400–404.

190

Lakes on Mars

Blasius, K.R., Cutts, J.A., Guest, J.E., Masursky, H., 1977. Geology of the Valles Marineris: first analysis of imaging from the Viking I Orbiter primary mission. J. Geophys. Res. 82, 4067–4091. Cabrol, N.A., Grin, E.A., 1999. Distribution, classification, and ages of Martian impact crater lakes. Icarus 142, 160–172. Cabrol, N.A., Grin, E.A., 2001. The evolution of lacustrine environments on Mars: (Is Mars only Hydrologically Dormant?) Icarus 149, 291–328. Cabrol, N.A., Grin, E.A., 2005. Ancient and recent lakes on mars. In: Tokano, T. (Ed.) Water on Mars and Life. Springer, Berlin, 235–259. Carling, P.A., Kirkbride, A.D., Parnacho, S., Borodavko, P.S., Berger, G.W. 2002. Late qua­ ternary catastrophic flooding in the altai mountains of south-central siberia: A synoptic overview and an introduction to flood deposit sedimentology. In: Flood and Megaflood Processes and Deposits: Recent and Ancient Examples. Int. Assoc. of Sedimentologists, Special Publication 32, pp. 17–35. Carr, M.H., 1979. Formation of Martian flood features by release of water from confined aquifers. J. Geophys. Res. 84, 2995–3007. Carr, M.H., 1996. Water on Mars. Oxford University Press, New York. Chapman, M.G., Gudmundsson, M.T., Russell, A.J., Hare, T.M., 2003. Possible Juventae Chasma subice volcanic eruptions and Maja Valles ice outburst floods on Mars: implica­ tions of Mars Global Surveyor crater densities, geomorphology, and topography. J. Geophys. Res. 108, 20. doi:10.1029/2002JE002009. Chapman, M.G., Neukum, G., Dumke, A., Michaels, G., van Gasselt, S., Kneissl, T., et al., 2009. Evidence of late-stage fluvial outflow in Echus Chasma, Mars, 40th Lun. Plan. Sci. Conf., Abstract #1374. Chapman, M.G., Tanaka, K.L., 2002. Related magma-ice interactions: Possible origins of chasmata, chaos, and surface materials in Xanthe, Margaritifer, and Meridiani Terrae, Mars. Icarus 155, 324–339. Christensen, P.R., et al., 2003. Morphology and composition of the surface of Mars: Mars Odyssey THEMIS results. Science 300, 2056–2061. Christensen, P.R., Jakosky, B.M., Kieffer, H.H., Malin, M.C., McSween, H.Y., Nealson, K., et al., 2004. The Thermal Emission Imaging System (THEMIS) for the Mars 2001 odyssey mission. Space Sci. Rev. 110, 85–130. Clifford, S.M., Parker, T.J., 2001. The evolution of the Martian hydrosphere: implications for the fate of a primordial ocean and the current state of the northern plains. Icarus 154, 40–79. Coleman, N.M., Baker, V. 2007. Evidence that a paleolake overflowed the rim of Juventae Chasma, Mars, 38th Lun. Plan. Sci. Conf., Abstract #1046. Coleman, N.M., Dinwiddie, C.L., Baker, V.R., 2007. Evidence that floodwaters filled and overflowed Capri Chasma, Mars. Geophys. Res. Lett. 34, 5. doi:10.1029/2006GL028872. Davatzes, A.K., Gulick, V.C. 2007. Evidence for tectonically controlled hydrothermal fluid flow in relay zones on Mars from early HiRISE images, 38th Lun. Plan. Sci., Abstract #1788. Di Achille, G., Hynek, B.M., Searls, M.L., 2009. Positive identification of lake strandlines in Shalbatana Vallis, Mars. Geophys. Res. Lett. 36, 5. doi:10.1029/2009GL038854. Di Achille, G., Ori, G.G., Reiss, D., 2007. Evidence for late Hesperian lacustrine activity in Shalbatana Vallis, Mars. J. Geophys. Res. 112, 15. doi:10.1029/2006JE002858. Dohm, J.M., Ferris, J.C., Baker, V.R., Anderson, R.C., Hare, T.M., Strom, R.G., et al., 2001. Ancient drainage basin of the Tharsis region, Mars: potential source for outflow channel systems and putative oceans or paleolakes. J. Geophys. Res. 106, 32943–32958.

Episodic ponding and outburst flooding associated with chaotic terrains in Valles Marineris

191

Edwards, C.S., Christensen, P.R., Hamilton, V.E. 2007. Evidence for extensive olivine-rich basalt bedrock outcrops in Ganges and Eos chasma on Mars, 7th Int. Conf. Mars, Abstract #3280. Farrand, W.H., Gaddis, L.R., Keszthelyi, L., 2005. Pitted cones and domes on Mars: observa­ tions in Acidalia Planitia and Cydonia Mensae using MOC, THEMIS, and TES data. J. Geophys. Res. 110, 14. doi:10.1029/2004JE002297. Gendrin, A., Mangold, N., Bibring, J.-P., Langevin, Y., Gondet, B., Poulet, F., et al., 2005. Sulfates in Martian layered terrains: the OMEGA/Mars Express view. Science 307, 1587–1591. Glotch, T.D., Christensen, P.R., 2005. Geologic and mineralogic mapping of Aram Chaos: evidence for a water-rich history. J. Geophys. Res. 110, 21. doi:10.1029/2004JE002389. Golombek, M.P., Plescia, J.B., Franklin, B.J., 1991. Faulting and folding in the formation of planetary wrinkle ridges, Proc. of 21st Lun. Plan. Sci. Conf., 679–693. Greeley, R., Kuzmin, R.O., Nelson, D.M., Farmer, J.D., 2003. Eos Chasma, Mars: Regional setting for a potential landing site for astrobiology. J. Geophys. Res. 108, 13. doi:10.1029/ 2002JE002014, 2003. Hamblin, K.W., 1994. Late Cenozoic Lava Dams in the Western Grand Canyon. GSA Memoir, Colorado, 183, 139 pp. Hanna, J.C., Phillips, R.J., 2005. Hydrological modeling of the Martian crust with application to the pressurization of aquifers. J. Geophys. Res. 110, 13. doi:10.1029/2004JE002330. Harrison, K.P., Chapman, M.G., 2008. Evidence for ponding and catastrophic floods in central Valles Marineris, Mars. Icarus 198, 351–364. Harrison, K.P., Grimm, R.E., 2004. Tharsis recharge: a source of groundwater for Martian outflow channels. Geophys. Res. Lett. 31, doi:10.1029/2004GL020502. Harrison, K.P., Grimm, R.E., 2008. Multiple flooding events in Martian outflow channels. J. Geophys. Res. 113, doi:10.1029/2007JE002951. Hartmann, W.K., 2007. Martian cratering 9: toward resolution of the controversy about small craters. Icarus 189, 274–278. Hendron, A.J., Patton, F.D., 1985. The Vaiont slide, a geotechnical analysis based on new geologic observations of the failure surface. US Army Corps of Engineers Technical Report GL-85-5 (2 volumes). Howard, A.D., 1994. Rockslopes. In: Abrahams, A.D. and Parsons, A.J. (Eds.), Geomorphol­ ogy of Desert Environments. Chapman & Hall, London, 123–172. Howard, A.D., 1995. Simulation modeling and statistical classification of escarpment plan­ forms. Geomorphology 12, 187–214. Howard, A.D., Moore, J.M., 2004. Scarp-bounded benches in Gorgonum Chaos, Mars: formed beneath an in ice-covered lake? Geophys. Res. Lett. 31, doi:10.1029/2003GL018925. Irwin, R.P., Maxwell, T.A., Howard, A.D., Craddock, R.A., Leverington, D.W., 2002. A large paleolake basin at the head of Ma’adim Vallis, Mars. Science 296 (5576), 2209–2212. Komar, P.D. 1991. Mud volcanoes on Mars? Reports Plan. Geol. Geophys. Prog., N92–10728 01–91, 539–541. Komatsu, G., Kargel, J.S., Baker, V.R., Strom, V.R., Ori, G.G., Mosangini, C., et al. 2000. A chaotic terrain hypothesis: explosive outgas and outflow by dissociation of clathrate on Mars, 31st Lun. Plan. Sci. Conf., Abstract #1434. Kraal, E.R., Asphaug, E., Moore, J.M., Lorenz, R.D., 2006. Quantitative geomorphic modeling of Martian bedrock shorelines. J. Geophys. Res. 111, doi:10.1029/2005JE002567. Kreslavsky, M.A., Head, J.W., 2002. Fate of outflow channel effluents in the northern lowlands of Mars: the Vastitas Borealis Formation as a sublimation residue from frozen ponded bodies of water. J. Geophys. Res. 107, doi:10.1029/2001JE001831.

192

Lakes on Mars

Legros, F., 2002. The mobility of long-runout landslides. Eng. Geol. 63, 301–331. Lipman, P.W., Normark, W.R., Moore, J.G., Wilson, J.B., Gutmacher, C.B., 1988. The giant submarine Alika debris slide, Mauna Loa, Hawaii. J. Geophys. Res. 93 (B5), 4279–4299. Lucchitta, B.K., 1979. Landslides in Valles Marineris, Mars. J. Geophys. Res. 84, 8097–8113. Lucchitta, B.K., 1987. Valles Marineris, Mars: wet debris flows and ground ice. Icarus 72, 411–429. Lucchitta, B.K., Isbell, N.K., Howington-Kraus, A., 1994. Topography of Valles Marineris: implications for erosional and structural history. J. Geophys. Res. 99, 3783–3798. Lucchitta, B.K., McEwen, A.S., Clow, G.D., Geissler, P.E., Singer, R.B., Schultz, R.A., et al., 1992. The canyon system on mars. In: Keiffer, H.H. Jakosky, B.M. Snyder, C.W. and Matthews, M.S. (Eds.), Mars. University of Arizona Press, Tuscon, Arizona, 453–492. Malin, M.C., Edgett, K.S., 2001. Mars global surveyor mars orbiter camera: interplanetary cruise through primary mission. J. Geophys. Res. 106, 23429–23570. Malin, M.C., Edgett, K.S., 2003. Evidence for persistent flow and aqueous sedimentation on early Mars. Science 302, 1931–1934. Malin, M.C., Bell, J.F., Cantor, B.A., Caplinger, M.A., Calvin, W.M., Clancy, R.T., Edgett, K.S., Edwards, L., Haberle, R.M., James, P.B., Lee, S.W., Ravine, M.A., Thomas, P.C., Wolff, M.J., 2007. Context Camera Investigation on board the Mars Reconnaissance Orbiter, J. Geophys. Res. 112, 25. doi: 10.1029/2006JE002808. Manga, M., 2004. Martian floods at Cerberus Fossae can be produced by groundwater dis­ charge. Geophys. Res. Lett. 31, doi:10.1029/2003GL018958.. Mangold, N., Gendrin, A., Quantin, C., Bibring, J.-P., Gondet, B., Langevin, Y., et al., and the OMEGA and HRSC Co- Investigator Teams, 2007a. An overview of the sulfates detected in the equatorial regions by the OMEGA/MEX spectrometer, 7th Int. Mars Conf., Abstract #3141. Mangold, N., Poulet, F., Mustard, J.F., Bibring, J.P., Gondet, B., Langevin, Y., et al., 2007b. Mineralogy of the Nili Fossae region with OMEGA/Mars Express data: 2. Aqueous alteration of the crust. J. Geophys. Res. 112, doi:10.1029/2006JE002835. Mangold, N., Quantin, C., Ansan, V., Delacourt, C., Allemand, P., 2004. Evidence for precipitation on Mars from dendritic valleys in the Valles Marineris area. Science 305, 78–81. Masursky, H., Boyce, J.M., Dial, A.L. Jr., Schaber, G.G., Strobell, M.E., 1977. Classification and time of formation of Martian channels based on Viking data. J. Geophys. Res. 82, 4016–4038. Max, M.D., Clifford, S.M., 2000. The initiation of Martian outflow channels through the catastrophic decomposition of methane hydrate, 31st Lunar Planet. Sci. Conf. Abs. Lunar and Planetary Institute, March 15–18, Houston, Texas, LPSC 31st CD, #2094. McCauley, J.F., 1978. Geologic map of the Coprates quadrangle of Mars, U.S. Geol. Surv. Misc. Invest. Ser. Map I-897. McCauley, J.F., Carr, M.H., Cutts, J.A., Hartman, W.K., Masursky, H., Milton, D.J., et al., 1972. Preliminary Mariner 9 report of the geology of Mars. Icarus 17, 289–327. McEwen, A.S., 1989. Mobility of large rock avalanches: evidence from Valles Marineris, Mars. Geology 17, 1111–1114. McEwen, A.S., 2007. Mars Reconnaissance Orbiter’s High Resolution Imaging Science Experi­ ment (HiRISE). J. Geophys. Res. 112, 14 colleagues.doi:10.1029/2005JE002605. McEwen, A.S., Bierhaus, E.B., 2006. The importance of secondary cratering to age constraints on planetary surfaces. Ann. Rev. Earth Plan. Sci. 34, 535–567. McEwen, A.S., Preblich, P.B., Turtle, E., Artemieva, N., Golembek, M., Hurst, M., et al., 2005. The rayed crater Zunil and interpretations of small impact craters on Mars. Icarus 176, 351–381. McMurtry, G.M., Herrero-Bervera, E., Cremer, M.D., Smith, J.R., Resig, J., Sherman, C., et al., 1999. Stratigraphic constraints on the timing and emplacement of the Alika 2 giant Hawaiian submarine landslide. J. Volcan. Geotherm. Res. 94, 35–58.

Episodic ponding and outburst flooding associated with chaotic terrains in Valles Marineris

193

Metz, J.M., Grotzinger, J.P., Mohrig, D., Milliken, R., Prather, B., Pirmez, C., et al., 2009. Sublacustrine depositional fans in southwest Melas Chasma. J. Geophys. Res. 114, doi:10.1029/2009JE003365. Müller, L., 1964. The rock slide in the Vaiont valley. Felsmech. Ingenieurgeol. 2, 148–212. Neukum, G., Jaumann, R., 2004. HRSC: the high resolution stereo camera of mars express. In: Wilson, A. (Ed.), Mars Express: The Scientific Payload. ESA SP-1240, ESA Publications Division, Netherlands. Okubo, C.H., McEwen, A.S., 2007. Fracture-controlled paleo-fluid flow in Candor Chasma, Mars. Science 315, 983–985. Okubo, C.H., Schultz, R.A., Chapman, M.A., and the HiRISE Team, 2009. Deformation band clusters on Mars and implications for subsurface fluid flow. GSA Bull. 121, 474–482. Ori, G.G., Mosangini, C., 1998. Complex depositional systems in Hydraotes Chaos, Mars: an example of sedimentary process interactions in the Martian hydrologic cycle. J. Geophys. Res. 103, 22713–22723. Parker, T.J., Gorsline, D.S., Saunders, R.S., Pieri, D.C., Schneeberger, D.M., 1993. Coastal geomorphology of the Martian northern plains. J. Geophys. Res. 98, 11061–11078. Payne, M.C., Farmer, J.D., 2001. Volcano-ice interactions and the exploration of extant Martian life, AGU Fall Meeting, Abstract #P22B–0549. Perron, J.T., Mitrovica, J.X., Manga, M., Matsuyama, I., Richards, M.A., 2007. Evidence for an ancient Martian ocean in the topography of deformed shorelines. Nature 447, 840–843. Phillips, R.J., Zuber, M.T., Solomon, S.C., Golombek, M.P., Jakosky, B.M., Banerdt, W.B., et al., 2001. Ancient geodynamics and global-scale hydrology on Mars. Science 291, 2587–2591. Pike, R.J., Wilhelms, D.E., 1978. Secondary-impact craters on the Moon: topographic form and geologic process. 9th Lun. Plan. Sci.Conf., IX, 907–909. Plescia, J.B., Saunders, R.S., 1982. Tectonic history of the Tharsis region, Mars. J. Geophys. Res. 87, 9775–9791. Quantin, C., Allemand, P., Mangold, N., Delacourt, C., 2004. Ages of Valles Marineris (Mars) landslide and implications for canyon history. Icarus 172, 555–572. Quantin, C., Allemand, P., Mangold, N., Dromart, G., Delacourt, C., 2005. Fluvial and lacus­ trine activity on layered deposits in Melas Chasma, Valles Marineris, Mars. J. Geophys. Res. 110, doi:10.1029/2005JE002440. Rice, J.W., Edgett, K.S., 1996. Catastrophic flood sediments in Chryse Basin, Mars, and Quincy Basin, Washington: Application of sandar facies model. J. Geophys. Res. 102, 4185–4200. Rotto, S., Tanaka, K.L. 1995. Geologic/Geomorphic map of the Chryse Planitia region of Mars, U.S. Geol. Surv. Misc. Invest. Ser. Map I-2441, scale 1:5,000,000. Saunders, R.S., 1979. Geologic map of the Margaritifer Sinus quadrangle of Mars, U.S. Geol. Surv. Misc. Invest. Ser. Map I-1144, 1: 5,000,000 scale. Schultz, R.A., 1991. Structural development of Coprates Chasma and western Ophir Planum, Valles Marineris rift, Mars. J. Geophys. Res. 96, 22777–22792. Schultz, R.A., 1998. Multiple-process origin of Valles Marineris basins and troughs, Mars. Planet. Space Sci. 46, 827–834. Shaller, P.J., Murray, B.C., Albee, A.L. 1989. Subaqueous landslides on Mars?, 20th Lun. Plan. Sci. Conf., 990–991. Sharp, R.P., 1973. Mars–Troughed terrain. J. Geophys. Res. 78 (20), 4063–4072. Shreve, R., 1966. Sherman landslide, Alaska. Science 154, 1639–1643. Smith, D.E., et al., 1999. The global topography of Mars and implications for surface evolution. Science 284, 1495–1503.

194

Lakes on Mars

Smith, D.E., et al., 2001. Mars Orbiter Laser Altimeter: experiment summary after the first year of global mapping of Mars. J. Geophys. Res. 106, 23689–23722. Tanaka, K.L., 1997. Sedimentary history and mass flow structures of Chryse and Acidalia Planitiae, Mars. J. Geophys. Res. 102, 4131–4149. Tanaka, K.L., Golombek, M.P. 1989. Martian tension fractures and the formation of grabens and collapse features at Valles Marineris, Proceedings of 19th Lunar And Planetary Science Conference, 383–396. Tanaka, K.L., Skinner, J.A., Hare, T.M., Joyal, T., Wenker, A., 2003. Resurfacing history of the northern plains of Mars based on geologic mapping of Mars Global Surveyor data. J. Geophys. Res. 108 (E4), doi:10.1029/2002JE001908. Treiman, A.H., 2008. Ancient groundwater flow in the Valles Marineris on Mars inferred from fault trace ridges. Nature Geosci. 1, 181–183. Watters, T.R., 1991. Origin of Periodically Spaced Wrinkle Ridges on the Tharsis Plateau of Mars. J. Geophys. Res. 96, 15599–15616. Weitz, C.M., Irwin, R.P., Chuang, F.C., Bourke, M.C., Crown, D.A., 2006. Formation of a terraced fan deposit in Coprates Catena, Mars. Icarus 184, 436–451. Weitz, C.M., Milliken, R.E., Grant, J.A., McEwen, A.S., Williams, R.M.E., Bishop, J.L., 2008. Light-toned strata and inverted channels adjacent to Juventae and Ganges chasmata, Mars. Geophys. Res. Lett. 35, doi:10.1029/2008GL035317. Witbeck, N.E., Tanaka, K.L., Scott, D.H. 1991. The geologic map of the Valles Marineris region, Mars, U.S. Geol. Surv. Misc. Invest. Ser. Map I-2010, scale 1:2,000,000.

7 Evidence for ancient lakes in the Hellas region

Sharon A. Wilson†, Jeffrey M. Moore*, Alan D. Howardþ, and Don E. Wilhelms§ †

Center for Earth and Planetary Studies, National Air and Space Museum, Smithsonian Institution, Washington, DC, USA * Space Sciences Division, NASA Ames Research Center, Moffett Field, CA, USA þ Department of Environmental Sciences, University of Virginia, Charlottesville, VA, USA § Branch of Astrogeology, U.S. Geological Survey, Reston, VA, USA.

7.1

Introduction

The geologic and climatic history of Mars is preserved in the ancient highland terrain that covers roughly 60% of the planet. The Hellas region in particular, due to its size, age, and diversity of landforms, provides insight into a long record of dynamic and complex geomorphic processes on Mars that likely varied in response to fluctuations in climate (e.g., Laskar et al., 2004). Although most climate models for early Mars predict cold and dry conditions wherein liquid water was not stable at the surface, characteristic morphologies of landforms dating back to the Noachian and Early Hesperian examined in this chapter suggest that water played a role in their formation and modification. As defined here, the greater Hellas region encompasses the area from 20°S to 65°S between 40°E and 110°E (Figure 7.1). The unmodified term Hellas refers to the part of the region that is enclosed by the topographic rim of the Hellas impact structure, and therefore includes both uplifted highlands terrain and an infilled inner basin, Hellas Planitia. Hellas Planitia contains the lowest point on Mars (Figure 7.1), so with roughly 9 km of relief, Hellas is the deepest and broadest enclosed depression on the planet (Leonard and Tanaka, 2001; Smith et al., 1999). Telescopic investigations of the Hellas region have been documented since the nineteenth century, where observers, misinterpreting seasonal albedo changes, thought Hellas was a raised landmass, island, or ice cap. Very little was known about the region until 1969, when the Mariner 7 fly-by imaged the surface, and scientists determined that Hellas is a depression rather than a raised plateau (Murray et al., 1971). The early studies—and some recent ones—were hampered by seasonal frost, water-ice clouds, and dust storms. Lakes on Mars. DOI: 10.1016/B978-0-444-52854-4.00007-6 © 2010 Elsevier B.V. All rights reserved.

196

Lakes on Mars

Figure 7.1 Present-day topographic view of the Hellas region from 128 pixel/degree Mars Orbiter Laser Altimeter (MOLA) data. DV, NV, HV, RV, AC, and HC represent Dao Vallis, Niger Vallis, Harmakhis Vallis, Reull Vallis, Alpheus Colles, and Hellas Chaos, respectively. Layered deposits discussed in text are in craters Terby (T), Millochau (M), A and B, Niesten (N), and informally named SW crater (S). Blue lines from preliminary map (produced by co-author Howard) reconstructing valley networks extant when widespread runoff ceased near Noachian– Hesperian boundary (Section 7.3). Valleys digitized in commercial program, Surfer®, on base maps of THEMIS IR day and night image mosaics (256 pixel per degree sampling) and contour map based on MOLA Precision Experiment Data Records (PEDR) shot data gridded to a 1-km2 DEM using natural neighbor interpolation. Valleys visible in THEMIS data were identified and digitized on contour base map and additional valleys were defined by aligned V- or U-shaped contours with a consistent upslope trend. Valleys were interpolated through minor depressions and across small-to-moderate post-Noachian impact craters (identified by fresh rims and ejecta and paucity of crater infilling). Because the purpose of mapping was to delineate extent of valley networks rather than absolute drainage density, only a representative sample of small headwater tributaries was mapped. Preliminary mapping does not attempt to date relative ages of different parts of valley networks using crater counting, but most valleys, with the exception of Dao and Harmakhis Valles, probably date to the Late Noachian to Early Hesperian (Howard et al., 2005; Fassett and Head, 2008). This includes valleys incised into Amphrites and Peneus Paterae deposits along southern margin of Hellas; these valleys presumably date to shortly after the 3.6–3.8 By inferred age of the volcanoes (Williams et al., 2008). White box, Figure 3a. MOLA shaded-relief map in Lambert Conformal projection of region from 20°E to 130°E between 0° and 70°S. Scale along axes in kilometers relative to projection center at 75°E and 35°S; north to top.

In this chapter, we summarize the geomorphic settings of the Hellas region (Figure 7.2) from previously published observations, which were analyzed and mapped by numerous authors using orbital datasets from several instruments and missions over the past four decades. Images from Mariner 9 and Viking and data from the Viking

Evidence for ancient lakes in the Hellas region

197

Figure 7.2 Geomorphic map of Hellas region after Moore and Wilhelms (2001) with updated unit contacts based on thermal properties from THEMIS day IR base map, available CTX and HiRISE images, and recent studies of western Hellas (Moore and Wilhelms, 2007), eastern Hellas (Bleamaster and Crown, 2010), and volcanic provinces around Hellas (Williams et al., 2008). Units include ancient highland terrain (h), upper (m2) and lower (m1) mantling material, volcanic material (v) from Malea Planum in the south and Hesperia Planum in the east, reticulate terrain (r), fan deposit (f), plains material (p), channel material (c), plateau material (pl), and honeycomb terrain (ht). Contours indicate potential stands of water (or ice-covered lakes) discussed in text. Many craters with floors at or below +600 m contour contain interior layered deposits (white dots, updated from Moore and Howard, 2005) and possibly hosted isolated lakes in Noachian Period. MOLA shaded-relief map in Lambert Conformal projection, covers region from 20°E to 130°E between 0° and 70°S. Scale along axes in kilometers relative to projection center at 75°E and 35°S; north to top.

IRTM were utilized since the early 1970s, before MGS began returning datasets from MOC and MOLA in 1997. The primary science phase of the ODY mission began returning IR and VIS images from THEMIS in 2001, and images from OMEGA have provided spectral information since 2003. HiRISE, CTX, CRISM, and SHARAD on MRO have offered unprecedented resolution, coverage, and insight into the Hellas region since 2006. Many observations are consistent with water-related processes and a lacus­ trine environment in the Hellas region during the Noachian and Hesperian. This chapter focuses on several compelling lines of evidence. Our summary of valley networks and canyons emphasizes new data describing the nature and extent of Hellas-wide valleys that were mapped using reconstructed topography (Section 3.0). We discuss potential stands of water based on relations among topography, morphology, and mineralogy of

198

Lakes on Mars

landforms around the basin (Section 4.0). Finally, we examine the nature and origin of sedimentary layered deposits in the Hellas region, particularly in Terby crater (Section 5.0). Evidence supporting the sustained role of liquid water and/or ice in the Hellas region has significant implications for the past climate and habitability of Mars, including the potential for the formation and preservation of biogenetic material.

7.2 Regional geology The impact that formed the Hellas basin in the Early Noachian about 4 billion years ago (e.g., Werner, 2008) widely distributed materials that contributed to the ancient highlands and rugged rim. Recognized as some of the oldest terrains on Mars (e.g., Tanaka et al., 1992), these include Noachis Terra to the west, Terra Sabaea to the northwest, Tyrrhena Terra to the northeast, and Promethei Terra to the east and southeast (Figure 7.1). The topography north and west of Hellas are characterized by heavily cratered terrain with elevations ranging from –1 km to over 5 km, and Hellespontus Montes, a system of mountainous scarps forming the western basin rim (Figure 7.1). The volcanic provinces northeast, south, and southwest of Hellas contain the six oldest central vent volcanoes on Mars, which were active from the Late Noachian to Early Hesperian (Figures 7.1 and 7.2) (e.g., Crown and Greeley, 1993; Crown et al., 1992; Greeley and Crown, 1990; Gregg et al., 1998; Peterson, 1977; Potter, 1976; Tanaka and Leonard, 1995; Williams et al., 2008). The distribution of layered deposits in craters north of Hellas Planitia (Figure 7.2) (e.g., Malin and Edgett, 2000) and large alluvial fans (Moore and Howard, 2005a) provides morphological evidence that a complex history of deposition and erosion characterized the Late Noachian to Early Hesperian, coeval with a proposed period of enhanced precipitation at the Noachian–Hesperian boundary. Recent studies of the topography, morphology, stratigraphy, and mineralogy have proposed that Hellas may have at times hosted glaciers, ice-covered lakes, or a basin-wide sea during the Noachian and Hesperian (e.g., Malin and Edgett, 2000; Moore and Wilhelms, 2001). Geologic activity during the Amazonian was greatly reduced (see Chapter 2) and dominated by aeolian and mass-wasting processes. The morphology of viscous-flow features and lobate debris aprons on highland massifs and the interior walls of craters and valleys provides evidence for episodes of ice-rich deposition (e.g., Mangold, 2003; Pierce and Crown, 2003; Squyres, 1979). Radar properties from SHARAD of a typical large lobate debris apron in eastern Hellas are consistent with massive water ice, supporting previous interpretations of debris-covered accumulations of water-ice or buried glaciers (Holt et al., 2008).

7.3 Valley networks and canyons as sources of lake water The ancient, densely cratered highland terrain and intercrater plains of Mars are believed to be modified by water, as evidenced by ubiquitous valley networks and larger channels (e.g., Baker, 1982; Carr, 1981), see Chapter 8. The origin of the valley

Evidence for ancient lakes in the Hellas region

199

networks, the relative roles of groundwater sapping and runoff (e.g., Pieri, 1980; Sharp and Malin, 1975), and the climate in which they formed have been heavily debated since they were observed by Mariner 9 (McCauley et al., 1972) and are detailed in Chapter 2. Analyses of the valley networks within the cold-and-dry paradigm of early Mars resulted in interpretations of an origin via groundwater sapping driven by hydrothermal processes (e.g., Clifford, 1993; Goldspiel and Squyres, 2000; Gulick, 1998) or, alternatively, via cold-climate processes and the transient flow of liquid water at the surface (Gaidos and Marion, 2003). Others have proposed a generally arid early climate punctuated with intermittent surface wettings (Stepinski and Stepinski, 2005). Recent observations of the physical characteristics of valley networks and the extensive volume of sediment redistributed by them indicate widespread fluvial erosion, leading to an emerging consensus that the early climate could sustain a hydrologic cycle of precipitation and runoff during the Late Noachian to Early Hesperian (e.g., Craddock and Howard, 2002; Grant, 2000; Howard et al., 2005). The entire periphery of Hellas—its rim and surrounding highlands—is incised by fluvial channels and valley networks that transported material into low-lying Hellas Planitia (Figure 7.1) (e.g., Carr and Chung, 1997; Greeley and Guest, 1987; Leonard and Tanaka, 2001; Tanaka and Leonard, 1995). Along the eastern rim, Hesperian-aged Reull, Dao, Niger, and Harmakhis Valles extend through the cratered highlands and plains toward Hellas Planitia (Figure 7.1) (e.g., Cabrol, 1991; Mest and Crown, 2001, 2002, 2003). These canyons, once considered outflow channels formed by cata­ strophic outbursts of either subsurface or surface water or ice released by volcanism (e.g., Crown et al., 1992; Mars Channel Working Group, 1983; Price, 1998; Squyres et al., 1987), are now thought to be more consistent with an early fluvial stage followed by collapse and sapping. The collapse and sapping processes, supported by a volatile-rich substrate along the broad depositional shelf, may have varied with time and might represent a transition from a water- to ice-dominated surface environ­ ment (Crown et al., 2005). One difficulty with interpreting the original flow paths of valley networks is recognizing and then adjusting for the topographic effects of post-flow modification from processes such as mass wasting, aeolian deposition and erosion, and cratering. The preliminary map of the valley networks in the Hellas region reproduced in Figure 7.1 attempts to reconstruct the extant valley networks at the time immediately following the cessation of widespread runoff near the Noachian–Hesperian boundary (e.g., Howard et al., 2005; Irwin et al., 2005, 2008). The map demonstrates that Hellas received drainage from a wide area on its rim and the surrounding uplands, which could help maintain, at least episodically, lakes within Hellas Planitia. Valley networks in the highlands north of Hellas are well defined and typically incised the terrain 100–300 m relative to adjacent plains surfaces. The networks become less distinct and sparser below about 0 m elevation (relative to the MOLA datum), and only a few, shallowly-incised valleys can be recognized below –1000 m. Broad, shallow valleys can be digitized in MOLA topography down to about the –6000 m level. These broad valleys are often extensions of the more deeply incised valleys on the uplands and are indicated on the valley network map (Figure 1). Due to

200

Lakes on Mars

the steep slopes of Hellas, if valley incision had resulted from precipitation and drainage into a lakeless Hellas, the deepest incision should be on the interior rim because the rate of fluvial erosion is a strong positive function of local gradient (e.g., Howard et al., 2005). Because that is not the case, there are several alternative possibilities: (1) rocks on the interior crater rim are systematically more resistant than those in the uplands, (2) precipitation, if any, was most intense on the highlands and the climate was sufficiently arid to prevent significant runoff into the basin interior, (3) valleys at low elevations within the basin were preferentially modified by postflow air-fall mantling or lacustrine processes, or (4) much of the valley incision occurred when lake levels in Hellas were near or above the –1000 m level. If that was the case, then some of the lower valleys (which are shallow, broad, but steep) might have formed underwater, analogous to terrestrial continental-shelf valleys. Many Noachian valley networks that may have existed in the Hesperia and Amenthes Planum regions northeast of Hellas and the Malea Planum region south of Hellas have largely been obliterated by Hesperian volcanic resurfacing. In many locales below 60°S, thick mantling by dust and/or ice obscures visible evidence of Noachian valley networks. Thinner, latitude-dependent mantling is present (e.g., Mustard et al., 2001; Soderblom et al., 1973) but does not inhibit recognition of valleys in contour maps below 40°S. The volcanic edifices of Hadriaca, Amphrites, and Tyrrhena Paterae have been densely dissected, probably because they include locally thick, easily eroded pyroclastic deposits (e.g., Greeley and Crown, 1990; Gulick and Baker, 1990). Preferential collection of snow on volcanic edifices and presumably snowmelt and runoff (e.g., Fassett and Head, 2006, 2007) may have also contributed to this erosion.

7.4 Ancient lake shorelines: regional relations between topography, morphology, and mineralogy The broad physiography of Hellas is defined by the regional slope toward the center of the basin. Hellas Planitia, which extends approximately 2100 km (46º–95ºE) by 1000 km (29ºS–55ºS), acts as a natural sediment trap for material eroded from the surrounding rim and highlands (Figure 7.1). Magmatic intrusions into potentially volatile-rich, easily eroded terrain, now occupied by Malea Planum to the south and Hesperia Planum to the northeast, may have triggered fluvial and/or debris flows that deposited material into central Hellas Planitia (Figure 7.1) (Howard and Moore, 2009; Tanaka et al., 2002). Hellas Planitia exhibits relatively uniform intermediate albedo (Christensen, 1988) and intermediate thermal signature (e.g., Putzig et al., 2005), consistent with material that is no larger than medium-sized sand or crust-bonded dust (Moore and Edgett, 1993), or with a surface that has few blocks within a matrix of fines (e.g., Mellon et al., 2000). The nature of the floor has been attributed to a combination of volcanic, sedimentary, and aeolian materials that were redistributed by fluvial processes and modified by wind (Crown et al., 1992; Greeley and Guest, 1987; Leonard and Tanaka, 2001; Moore and Edgett, 1993; Potter, 1976; Price, 1998;

Evidence for ancient lakes in the Hellas region

201

Tanaka and Leonard, 1995). Others propose glacial origins (Kargal and Strom, 1992). We propose that origins in stands of water or ice-covered water are more consistent with the topography, morphology, and mineralogy of the layered deposits observed in the inner periphery of Hellas. An argument for ancient lakes is drawn from regular association of landforms and deposits with contours of constant elevations that extend for thousands of kilometers around Hellas (as mapped from a digital elevation model generated from MOLA data). Well-developed, inward-facing scarps, morphological boundaries, preserved cratered terrain, and types of layered deposits occur repeatedly at the same elevations (Figure 7.2). Moore and Wilhelms (2001) suggest that Hellas contained ice-covered lakes with stands or “shorelines” at –6900 m, –5800 m, –4500 m, and –3100 m. Crown et al. (2005) propose –1800 m as the maximum paleolake high stand, and Wilson et al. (2007a) suggest the entire region may have been submerged to +600 m, corresponding to the highest well-developed, inward-facing scarp around Hellas. The deepest place in Hellas, and on Mars, is in the northwestern interior of Hellas Planitia (Figure 7.1). The depression is enclosed by the –6900 m contour, which corresponds to the lowest shoreline proposed by Moore and Wilhelms (2001) (Figure 7.2). The floor of the depression exposes highly deformed “taffy-pull” or “honeycomb” terrain (unit ht, Figure 7.2), which is characterized by low thermal inertia and a generally planar surface of individual, light- and dark-toned banded “cells” roughly 5–10 km across (Figure 7.3) (Leonard and Tanaka, 2001; Moore and Wilhelms, 2001, 2007). Various origins for unit ht have been proposed including kettle or ice stagnation (Kargel and Strom, 1992), soft-sediment deformation from the weight of lake-remnant ice blocks (Moore and Wilhelms, 2001), ductile defor­ mation analogous to salt diapirs or crustal doming (e.g., Mangold and Allemand, 2003), crustal-convection cells involving mud or igneous rocks (Kite et al., 2009; Mangold and Allemand, 2003), and concentric aeolian dunes (Tanaka and Leonard, 1995). Unit ht is remarkably analogous in size and morphology to the diaperically distorted evaporitic sequence exposed in the Great Kavir basin in central Iran (Jackson et al., 1990). Recent CTX and HiRISE data suggest that unit ht is more widespread than previously mapped and is stratigraphically below the adjacent Alpheus Colles (Figure 7.3), which is a Hesperian-aged, thick, rough, hummocky, ridged, and layered plateau that dominates central Hellas Planitia (unit pl, Figures 7.1–7.3). The topography of western Alpheus Colles is characterized by irregular, elongate, 1-km-high ridges spaced 100–200 km apart and separated by smooth or knobby depressions, multi-kilometer plateaus, or parallel ridges (Figure 7.3) (Howard and Moore, 2009). The ridges trend northeast/southwest parallel to the western Hellas rim, a structural grain also observed in the adjacent honeycomb terrain (unit ht) and nearby reticulate terrain (unit r). The complex topography of this area has been attributed to selective sublimation of ice-rich deposits (e.g., Tanaka and Leonard, 1995) or, alter­ natively, to deformation by a series of large, subaqueous mega-slides from the north­ west or southeast Hellas rims that may be analogous to terrestrial submarine landslides (Howard and Moore, 2009). Hellas Chaos (Figure 7.1) may also result from subaqu­ eous landslides or from loss of ground ice (Moore and Wilhelms, 2001). Another

202

Lakes on Mars

Figure 7.3 (a) Western floor of Hellas Planitia showing detail of honeycomb terrain (ht), reticulate terrain (r), central plateau (pl), and plains materials (p) from subframe of THEMIS day IR mosaic (see Figure 7.1). Solid white (–6900 m), solid black (–5800 m), dashed white (–3100 m), and dashed black (–1800 m) contours from MOLA data are approximate. Unit ht is in deepest depression on Mars (Figure 7.2) and underlies adjacent Alpheus Colles (AC) plateau materials. Ridges on western margin of AC oriented northeast/southwest, similar to structural grain in units ht and r (possibly deformed by subaqueous mega-slides from the northwest or southeast rims (Howard and Moore, 2009)). Locations of Figures 3b, 3c and 6a indicated. Image 1000 km across; north to top. (b) Example of knobs on floors of depressions between ridges on AC (unit pl), oriented northeast/southwest. Knobs are typically rounded and often exhibit complex patterns of linear depressions that may represent fractures or deformed layers (Howard and Moore, 2009). Subframe of CTX image P19_008295_1387. Image 6 km across (see 3a); north to top. (c) Relatively planar honeycomb terrain consists of individual, light- and darktoned banded cells 5–10 km across. Subframe of CTX image P13_006278_1392. Image 8 km wide (see 3a); north to upper left.

hypothesis is that this landscape is shaped by glacial drumlins (Kargel and Strom, 1992). The most credible of the putative ancient paleolake margins is observed at the –5800 m contour based on the consistency of landforms and contacts of deposits around the basin at this elevation (Figure 7.2). Although obscured in the south by

Evidence for ancient lakes in the Hellas region

203

Figure 7.4 (a) Detail of northwestern Hellas Planitia (see Figure 7.2) showing relation between morphologic boundaries (dashed lines) and proposed shorelines at –6900 m, –5800 m, –4500 m, –3100 m, and –1800 m (white lines). Units include lower mantling unit (m1), plains materials (p), honeycomb terrain (ht), fan deposit (f), and central plateau (pl). Small fan-like forms (black dotted lines) and associated channels (arrows) along margin of m1 and p could be fluvial fans or lacustrine in origin. The large fan-like form (unit f) exhibits layering in high-resolution images, but its origin remains unresolved. North to top. (b) Contact between m1, characterized by differentially eroded layered material, and overlying p unit. Subframe of map projected CTX image P19_008387_1489, see (a) for context. Image 25.5 km wide; north to top.

volcanic material (unit v), the –5800 m contour correlates with the contact between the plains unit (unit p) and the lower mantling unit (unit m1) along the majority of the basin periphery (Figures 7.2 and 7.4). The morphological change in the lower reaches of Harmakhis and Dao Valles (unit c) in eastern Hellas (Figures 7.2 and 7.5), and the margin of a discontinuous annular band of reticulate material (unit r) in western Hellas, also lies along this contour (Figures 7.2, 7.3, and 7.6). The Hesperian-aged unit p has a thick, layered, smooth to wavy surface character­ ized by wrinkle ridges (Figures 7.4 and 7.5). In western Hellas, unit p exhibits high thermal inertia values from THEMIS and TES that may represent a combination of coarse sand, strongly-crusted fines, abundant rocks, and/or scattered bedrock expo­ sures (Putzig et al., 2005). Like the rest of Hellas Planitia, this surface is sparsely cratered, an observation that implies a recent history of deposition, exhumation, and/ or erosion (e.g., Greeley and Guest, 1987). Unit p is interpreted to be fluvial sediment (Leonard and Tanaka, 2001) or sediment deposited into a lake or lakes (Bleamaster and Crown, 2010; Moore and Wilhelms, 2001). Small fan-like forms along the margin of units m1 and p in western Hellas could be fluvial in origin or analogous to subaqueous fans (Figure 7.4). Along the 1000-m-long scarp bounding the termini of Dao and Harmakhis Valles in eastern Hellas, the channels change from deeply incised, prominent, sharp, and exposed above –5800 m to muted, partially buried, and diffuse below (Figures 7.2 and 7.5).

204

Lakes on Mars

Figure 7.5 (a) Morphology of Dao and Harmakhis Valles in eastern Hellas changes from deeply incised, prominent, sharp, and exposed above –5800 m to muted, partially buried, and diffuse below, possibly analogous to terrestrial marine channels. Morphologic units correspond to plains materials (p), lower mantling unit (m1), channels (c), and volcanic materials (v). MOLA shaded relief 1000 km across (Figures 7.1 and 7.2 give context). White dot shows location of (b). (b) Example of layered deposits along and above scarp at –5800 m contour and termini of Dao and Harmakhis Valles from subframe of HiRISE PSP_007925_1385 near 40.9S, 84.2E. Image 6.5 km across (50 cm/pixel resolution); north to top.

Moore and Wilhelms (2001) suggest these valleys are similar to terrestrial marine channels that are underwater extensions of large rivers, in agreement with the inter­ pretation by Kargel and Strom (1992) of a channel flowing into a lake. Specific fluvio­ lacustrine depositional landforms such as deltas, however, have yet to be identified. Crown et al. (2005) identified layered outcrops along and above the scarp associated with the –5800 m contour and the termini of Dao and Harmakhis Valles, providing further evidence for the transition between the rim and the basin interior deposits (Figure 7.5). The mostly flat-lying layers exposed in isolated outcrops were likely an extensive, continuous deposit that covered much of the eastern Hellas rim (Bleamaster and Crown, 2010). The nature of the sedimentary layers at the termini of Dao and Harmakhis Valles is consistent with a lacustrine, fluvial, or rhythmic aeolian origin (Crown et al., 2005), but Bleamaster and Crown (2010) favor a water-laid origin or deposition in a Hellaswide lake. The –5800 m contour corresponds to the edge of the reticulate terrain along the western periphery of Hellas Planitia (Figures 7.2 and 7.3). This unit is characterized by rectangular or elliptical depressions bounded by sharp, interconnected ridges generally tangential or normal to the basin rim (Figure 7.6) (Leonard and Tanaka, 2001; Moore and Wilhelms, 2001, 2007). Individual depressions enclosed by the sharp interconnecting ridges are typically 1–2 km across, ridge crests are rounded to sharp, and the floors of the cells are often flat, though some are crossed by smaller quasi-parallel ridges. Reticulate terrain imaged by HiRISE indicates an abundance of

Evidence for ancient lakes in the Hellas region

205

Figure 7.6 (a) Reticulate terrain (unit r) occurs in band adjacent to plains material (unit p) in western Hellas Planitia along –5800 m contour and has a northeast/southwest structural grain also seen in honeycomb terrain (ht) and central plateau (pl) (see Figures 7.2 and 7.3a). Subframe of CTX image P16_007188_1374_XI_42S310W near 41.8°S, 49.4°E. Image 25 km across; north to top. Box shows location of HiRISE image in (b). (b) View of reticulate terrain in subframe of HiRISE image PSP_007188_1380 RED, image 10 km across (50 m/pixel). Image center located  41.8°S, 49.4°E. Box shows location of (c). (c) Layered nature and blocky texture of the reticulate terrain.

>1-m-scale blocks (Figure 7.6), contrasting with the generally smooth meter-scale surface of unit ht (Figure 7.3). It is difficult to discern, even at HiRISE resolution, whether the blocks are primary features (deposited as meter-scale blocks) or second­ ary (resulting from the break-up of indurated, fine-grained material). Moore and Wilhelms (2001) proposed that the sharp ridges may have formed by material forced up or falling into spaces between intermittently jostling ice blocks along the periphery of an ice-covered lake margin. Isolated surfaces in the reticulate terrain in western Hellas near 42°S, 48°E have a mineralogy of nearly 80% silica-rich phases (amorphous silica, phyllosilicates, or zeolites), which appears to correlate with intermediate-toned layered material that is being exposed by differential erosion (Bandfield, 2008). Hydrated silica could form in a variety of ways including acid weathering of volcanic ash or lava flows, precipita­ tion from hydrothermal discharge, or some combination of chemical precipitation and

206

Lakes on Mars

detrital sedimentation in a fluvial or lacustrine environment (Milliken et al., 2008; Weitz et al., 2008, 2009). Bandfield (2008) favors the last of these, and hydrated silica absorption features were confirmed with CRISM data at this location, providing further evidence for an aqueous origin (Bandfield et al., 2009). The zone from –5800 m to between –3100 m and –1800 m roughly corresponds to the lower mantling unit (m1), which is characterized by a discontinuous, smooth, and layered surface that is dissected by channels (Figures 7.2 and 7.4) (e.g., Moore and Wilhelms, 2001; Tanaka and Leonard, 1995). Crown et al. (2009) describe unit m1 in western Hellas as softened and buried and note a paucity of large (>50 km in diameter) craters relative to higher terrain, indicating either that young geologic materials were emplaced in Hellas or that erosion was preferential at lower elevations. Moore and Wilhelms (2001) describe the upper margin of unit m1 as in contact with the “landform-defined” rim along the northern and western rims of Hellas. They suggested an origin of lake sediment for unit m1, although erosion of soft volcanic tephra erupted from the volcanic edifices Hadriaca, Amphitrites, and Tyrrhena Paterae cannot be ruled out. Bleamaster and Crown (2010) attribute the material on the broad depositional shelf surrounding the canyons in eastern Hellas to be fluvial or lacustrine sedimentary material of possible volcanic origin derived from the surrounding ter­ rains, extending previously mapped boundaries of this mantling unit. Scarps and/or abrupt changes in texture coincide with the –4500 m and –3100 m contours proposed as potential shorelines by Moore and Wilhelms (2001). The position of these two putative stands is especially intriguing based on their correlation with high and low topographic features in Terby, a large Noachian crater described in Section 5.0 (Figures 7.1 and 7.7). The –4500 m and –3100 m contours correlate with the elevation of the flat crater floor in Terby and partway up its crater rim, respectively (Figure 7.7). The –2100 m contour and the maximum proposed high stand of water at –1800 m favored by Crown et al. (2005) based on evidence for collapse in Dao and Harmakhis Valles, the change in morphology along Reull Vallis, preservation of highland craters, and the channelized shelf in eastern Hellas represent a contour range that corresponds to the highest layered bench in Terby (Figure 7.7). If lacustrine deposition occurred in Terby, the lake in Hellas would have extended to at least –1800 m and would have been 5.5 km deep relative to the lowest part on the modern basin floor. The contact between the lower (unit m1) and upper (unit m2) mantling units is gradational. The upper contact of unit m2 corresponds to the highest inward-facing scarp associated with the +600 m contour (Figure 7.2). Mest et al. (2008) and Crown et al. (2009) describe this zone in western Hellas between the heavily cratered high­ lands (unit h) and the Hellas rim materials as the Terra Sabaea plains (found at the elevation range of –1800 m to 500 m). Unit m2 is characterized by smooth, low-relief plains, remnants of highland terrain consisting mainly of rims of large impact craters, and more abundant moderate to large impact craters than on m1, though fewer than on the adjacent highlands (unit h). Many of the craters and inter-crater plains within unit m2 that are at or below the +600 m contour have evidence of pits and layers (Korteniemi et al., 2005a; Moore and Howard, 2005b) that are analogous to features in Terby crater (Figure 7.2) (Wilson et al., 2007a). As described in Section 7.5,

Evidence for ancient lakes in the Hellas region

207

Figure 7.7 Noachian-aged Terby crater contains a suite of landforms including 2.5 km-thick sequences of layered deposits that are exposed in western (W), central (C), and eastern (E) benches extending from the northern rim (Figure 7.1 gives context). Location of viscous flow (VF) features, fan deposit (FD), trough floors (TF), moat deposit (MD), crater floor (CF), depressions (d), avalanche deposit (a), and scoured cap rock (*) noted for reference. The –4500 m, –3100 m, –2100 m, and –1800 m (dashed) contours may represent stands of water or ice-covered lakes. THEMIS daytime IR mosaic; north to top.

craters with sedimentary deposits consistent with these stands of water and/or ice that occur below the +600 m contour, but have rims that are higher than their interior deposits, may have hosted isolated lakes.

7.5

Light-toned layered deposits

The coverage and resolution of narrow-angle MOC images first revealed ubiquitous light-toned layered deposits throughout the highland terrain of Mars in a latitude belt between 30ºN and 30ºS, with a cluster in northern Hellas (Malin and Edgett, 2000). The “tone” of the material refers to the relative brightness (i.e., visible albedo) as seen in grayscale images. Layered deposits on Mars and in the Hellas region are most commonly associated with crater interiors, where material is typically exposed in the

208

Lakes on Mars

floors and walls of pits within the interior deposits, or in mounds and mesas on the crater floors, as in Terby (e.g., Figure 7.7). Layered deposits also occur on the lower slopes of the Hellas rim (Figure 7.4) and basin floor (Figures 7.3–7.5) and are exposed by scarps bounding irregular depressions in both crater floor deposits and within the intercrater plains, suggesting emplacement of sedimentary deposits across the region (Crown et al., 2009). The predominantly fine-grained, repetitive, indurated, laterally continuous, scarp-forming nature of the deposits is consistent with sedimen­ tary rocks.

7.5.1 A case study of crater Terby Terby is the largest (165 km-diameter) Noachian-aged (Leonard and Tanaka, 2001) crater in Hellas Planitia (Figures 7.1 and 7.7), and hosts the thickest and best-exposed sequence of sedimentary layers in Hellas. The size, age, and diversity of landforms in Terby are a testament to spatially and temporally varying geomorphic processes that are intimately associated with Hellas and are suggestive of an ancient lacustrine environment (Wilson et al., 2007a). Banked along the northern rim and exposed by north-trending troughs, up to 2.5 km-thick layered benches in Terby are accompa­ nied by a suite of landforms interpreted to be influenced by water and/or ice, including channels, fans, avalanche deposits, oblong depressions, striated caprock, grooved surfaces, viscous flow features, sinuous ridges, and arcuate scarps (Figures 7.7 and 7.8). The gently sloping (1° to the south) surfaces of the layered benches in Terby are bound by erosional scarps at their southern ends, abruptly terminating in the moat-like depression near the center of the crater (Figure 7.7). The original deposi­ tional geometry of the layered deposits, however, may have steepened by 3° near their southern terminus and subsequently eroded to its present form (Wilson et al., 2007a). Light-toned layers in Terby are also exposed on the floor of the troughs in between the layered benches, along scarps in the alluvial fan deposit, and in the flat crater floor (Figure 7.8). Distinct subunits are discernable within the western layered bench in Terby (Figure 7.7), a sequence that appears to correlate with the central and presumably eastern bench (Wilson et al., 2007a) (Figure 7.9). The oldest unit at the base of the sequence (LD1) is characterized by laterally continuous, thin (1–25 m) intermediatetoned layers interbedded with thicker light-toned beds (typically 80 m, ranges from 10 to over 150 m). Stratigraphically above, Subunit 2 (LD2) exhibits irregular, nonhorizontal, discontinuous, and possibly folded beds of varying thicknesses (down to sub-meter layers or laminations) as well as parallel bedding. The Subunit 3 (LD3) resembles the basal unit in its characteristic regular, laterally continuous beds having relatively constant thicknesses with hints of small-scale layering or lamina­ tions at HiRISE scale. The benches are capped by Subunit 4 (LD4), a 100 m-thick unit of light-toned layers between distinctive dark-toned, rubbly layers containing large, resistant, meter-scale clasts rather than indurated beds that break down along widely (multi-meter) spaced fractures as seen in the basal unit. Although the regularity of the bedding in the majority of the Terby sequence is not as rhythmic as the meter­

Evidence for ancient lakes in the Hellas region

209

Figure 7.8 THEMIS VIS mosaic of western (W) and central (C) layered benches capped by flat mesa tops (MT) in Terby crater showing approximate locations of –4500 m, –3100 m, and –2100 m contours from MOLA data (Figure 7.7 gives context). Layered deposits (LD) exposed along scarps of benches, along the north-facing scarp of the moat deposit (MD), and as isolated, layered mounds and ridges on the trough floors (TF) between layered benches. Bowl-like depression (d) in center of western bench contains light-toned concentric layers. Avalanche deposit (a), scoured caprock (*), and acruate ridges on TF (black arrow). North to top. Figure modified from Wilson et al. (2007a).

scale sedimentary layered deposits in Arabia Terra (Lewis et al., 2008), the tonal variations in the bedding likely reflect changes in grain size or degree of cementation, which may have been influenced by climate fluctuations controlled by quasi-cyclical orbital variations. The highly jointed and irregular nature of LD2 could be related to large-scale aeolian dunes, dune-like structures related to pyroclastic surge deposits (Fisher and Schmincke, 1984), deformation of solid rock, and diapiric intrusions related to ductile evaporate beds within the sequence or soft-sediment deformation (Wilson et al., 2007a). Deformation of soft sediment is common in deep-water basins subject to turbidity currents, shallow-water marine environments, deltas, and river floodplains (Allen, 1984). On Earth, potential triggers for soft-sediment deformation include mass

210

Lakes on Mars

movements (Reading, 1996), impact cratering (Alvarez et al., 1998), instabilities related to slope, sediment burial or load (e.g., Sarkar et al., 1982), tectonic activity (e.g., Alfaro et al., 1997), glacial activity (e.g., Iverson, 1999; Van der Wateren, 1995), or wave action and tsunamis (e.g., Dalrymple, 1979; Rossetti et al., 2000). All of these processes are plausible given the regional context and geologic activity of the Hellas region during the presumed time of deformation and diagenesis. Soft-sediment defor­ mation is also a mechanism proposed for the layered “cells” within unit ht in Hellas Planitia (Figure 7.3). The spectral signature of the layered benches in Terby crater correlating with a few OMEGA pixels indicates the presence of hydrated phyllosilicate minerals (Ansan

Figure 7.9 (Continued)

Evidence for ancient lakes in the Hellas region

211

et al., 2005; Bibring et al., 2006). Although research is ongoing to constrain the mineralogy from higher-resolution CRISM data in Terby, targets of the western layered bench (FRT 0000622B and 000059DF) indicate that the entire stack of light-toned layers is hydrated, and the youngest subunit at the top of the sequence (LD4) specifically correlates with iron (Fe) or magnesium (Mg) phyllosilicates (Wilson et al., 2007b). Murchie et al. (2009) also identified Fe/Mg smectities in the alluvial fan deposit on the northern rim of Terby (Figure 7.7). The presence of hydrated minerals and Fe/Mg phyllosilicates is consistent with, but not limited to, sediment deposited in a lacustrine environment.

7.5.2 Similar layered deposits around Hellas Most of the craters north and west of Hellas contain pits and/or light-toned layers (Figure 7.2) that are thickest on the Hellas-distal side of their rims (Korteniemi et al., 2005a; Moore and Howard, 2005b). Layered deposits in Terby, Millochau (Mest and Crown, 2005, 2006), “A” and “B” craters of Korteniemi et al. (2005b), Neisten, and the younger crater to the southwest (a.k.a. “SW crater”) (Figure 7.1) are similar in tone, stratigraphy, mineralogy, asymmetrical distribution, and original depositional geometry (Figure 7.10) (Wilson et al., 2007a). Layered deposits at comparable elevations, therefore, likely share a common age and mechanism of emplacement that is consistent with a lacustrine

Figure 7.9 Layered deposits in western bench of Terby crater (hollow arrows point downslope). (a) Detail of 2.5 km-thick layered sequence exposed on western bench in HiRISE image PSP_001662_1520 (see Figure 7.8). Dashed lines show contacts between subunits (see Section 7.5.1). Boxes show location of (b), (c), and (e). (b) LD1 exhibits thicker, light-toned (LT) beds interbedded with thin, intermediate-toned layers (black arrows) that weather along meter-spaced joints to produce boulders, which are transported (e.g., white arrow) and accumulate downslope. Finer layers or laminations hinted at HiRISE resolution (28 cm/pixel). Subframe of HiRISE image PSP_001662_1520; north to top. (c) Contact between LD3 and LD4 (dashed line) marked by dark-toned, boulder-rich layer that weathers to meter-scale boulders. LD3 characterized by thicker light-toned beds interbedded with thin, indurated, intermediate-toned layers (black arrows). Subframe of HiRISE image PSP_001662_1520; north to top. (d) Subframe of HiRISE image PSP_001596_1525 (left) showing sub-meter scale layering within LD1 in the western bench that indicates deposition by suspension (see Figure 8). Inset (right) highlights differences in tone (visible albedo) of layers, possibly indicating differences in mineralogy. North to left. (e) Contact (dashed line) between LD1 (lower left) and LD2 (upper right) in western bench. Predominantly light-toned, LD2 is characterized by irregular, non-horizontal, discontinuous (folded?) beds of varying thicknesses (arrows) as well as horizontal bedding. Subframe of HiRISE image PSP_1596_1525; north to top. (f) Distorted, non-horizontal, densely jointed nature of bedding in LD2 also observed on eastern face of layered mesa (see Figure 8). Well-defined contacts between LD1, LD2, and LD3 (dashed lines) likely represent an unconformity and/or change in depositional environment. Subframe of HiRISE image ESP_013872_1520. Image is 1.7 km wide; north to top. Figure modified from Wilson et al. (2007a).

212

Lakes on Mars

Figure 7.10 Craters with layered deposits around Hellas (located in Figure 7.1). (a) Crater Niesten and younger “SW” crater from part of high resolution stereo camera (HRSC) rectified nadir image H0554-0000 (centered 57.6°E and 28.2°S). Deposits less eroded than Terby (Figure 7.7) but exhibit similar layered deposits (LD) banked against north crater walls, smooth crater floor (CF) in the south and enigmatic depressions (d). Their northern parts slope gently southward from crater rims, but steepen near crater centers. (b) Informally-designated craters A (lower right) and B (upper left) of Korteniemi et al. (2005b) from subframe of HSRC image H0389. LD in crater B are little eroded and extend higher near northeast crater wall. Similar to Millochau in (c), fluvial annulus or bajada deposited after LD, possibly also burying and eroding layered-deposit exposures. (c) THEMIS daytime IR mosaic of Millochau crater, 81.1ºE, 21.1ºS, 725 km northeast of Terby. Sloping and steepening as in Niesten and SW crater (a). Pits surrounding three sides of the LD presumably postdate the adjacent fluvial bajada-like surface. North to top.

environment, although an aeolian origin cannot be entirely ruled out. If lacustrine in origin, layered deposits on the floor, on the walls, and in craters in Hellas Planitia (e.g., Terby crater) and at low elevations along its rim (e.g., Niesten and SW crater) would have been directly related to the proposed lacustrine environment with paleolakes extending to –1800 m. Layered deposits in craters at higher elevations (e.g., Millochau and craters A and B) were likely influenced by the regional physiographic setting and climate, but may have been closed systems that hosted isolated lakes. The detailed morphologic, topographic, and stratigraphic studies of layered depos­ its in Terby and similar deposits in Noachian craters at comparable and higher elevations, including Millochau, Niesten, SW crater, and craters A and B, yield a consistent picture of deposition and erosion (Figure 7.10). As much as 2.5 km of hydrated, light-toned layered material accumulated in craters in Hellas Planitia and in craters to the north and west. Beds within the layered sequences are laterally con­ tinuous over tens of kilometers, maintain relatively consistent thicknesses (individual beds range from 1 to 10 m thick), and lack evidence of channel deposits and crossbeds. Unconformities occur at both a regional and local scale, but do not appear to involve tectonic deformation or steeply-dipping bedding. Ease of erosion implies that most layers are composed of weakly-indurated sediment that is sand-sized or finer, but some beds are more strongly indurated and weather to blocky fragments, and a few beds at the top of the sequence may be composed of rubbly, meter-scale debris. The distribution of the layered deposits is generally thickest on the northern side (distal from central Hellas) of the crater basins; in Terby, Niesten, Millochau, and SW crater,

Evidence for ancient lakes in the Hellas region

213

the surface of the layered deposits typically slope gently to the south (Figure 7.10). Erosional exposures in Terby, Niesten, and SW crater indicate that the uppermost several hundred meters of layers conform to the overlying surfaces and like Mill­ ochau, steepened at their southern end by about 2–3°. The fine-grained nature, lateral continuity, and thickness of the beds are indicative of deposition from suspension in a fluid (air or water) rather than from shear flows (e.g., fluvial bedload, mass flows, or glacial till) (Wilson et al., 2007a). The widespread, preferential deposition of light-toned layers was followed by deposition of a relatively dark-toned deposit. The terraces in the upper, northern surfaces of the dark-toned deposits in Terby, Niesten, and SW crater may correlate with changes in lake levels. Subsequent erosion of the layered deposits at lower elevations (Terby, Niesten, SW craters) resulted in flat southern crater floors and moat-like depressions, possibly due to the involvement of ice and/or lacustrine processes (Figures 7 and 7.10) (Wilson et al., 2007a). Craters at higher elevations (Millochau and craters A and B) were subject to fluvial erosion resulting in dissection of the inner crater rim and the formation of an annulus or bajada around the layered deposits (Figure 7.10). This period of fluvial erosion and deposition at higher eleva­ tions may be contemporaneous with the general Mid- to Late-Noachian fluvial crater degradation (e.g., Craddock et al., 1997; Craddock and Howard, 2002; Irwin and Howard, 2002) or to the terminal episode of fluvial activity within the highlands. After this period of extensive erosion, a regional meter-scale blanket of dark-toned sediment was emplaced and subsequently indurated during the Hesperian, protecting the expo­ sures of layered sediments from further erosion. Sparse channels and depression-floor deposits suggest a period of not more than modest precipitation and local lacustrine activity including the formation of alluvial fans in Terby (Figure 7.7) and other craters north of Hellas (Moore and Howard, 2005a). The relative timing of the mantling and the late fluvial incision is uncertain. The late history of this region is characterized by aeolian erosion and local development of ice-related landforms such as the viscous flow features in the crater on the northwestern rim of Terby (Figure 7.7).

7.6

Discussion

Though there is consensus that the original Hellas basin was formed by a giant impact, the origin of the materials filling the basin and the depositional mechanisms involved still invokes some controversy. We infer that the major deposits in Hellas Planitia (units m1, p, r, ht, and pl) that are layered, laterally continuous, and composed of indurated, fine-grained material are most consistent with a hypothesis that they were deposited from suspension in a fluid (air or water), rather than from shear flows (e.g., fluvial bedload, mass flows, or glacial till). Fine-grained material can be produced in various ways on Mars, including physical and chemical weathering, pyroclastic eruptions, impact cratering, communication during transport, and chemical precipita­ tion. The enormous volume of material incorporated into the layered deposits in northern and western Hellas, however, favors sediment sources from a large region.

214

Lakes on Mars

Production, transport, and deposition of ash from explosive volcanism, redistribution of fine ejecta from large impacts, and aeolian redistribution of fine sediment abraded and/or deflated from the global regolith are all possible origins. Erosion of a deep, fine-grained global soil layer produced by chemical weathering is not a likely source of sediment for the layered deposits in the Hellas region as indicated by their mineralogy (Poulet et al., 2005). Several alternate non-lacustrine hypotheses are discussed below. The nature of the circumferential layered deposits in Hellas indicates that several potential depositional mechanisms can be readily eliminated. Mass wasting processes (e.g., debris flows and landslides) would not result in the regular, parallel bedding but would likely consist of coarse debris and contorted bedding (though as reported this may explain some of the material in the center of the basin such as honeycomb terrain). Glacial deposits (Kargel and Strom, 1992) would deliver a wide range of sediment grain sizes rather than well-sorted fines, and there is no evidence for glacial scour in Hellas or associated glacial features in the highlands. Although lava flows can produce deposits with constant-elevation contacts, the fine-grained nature, erodibility of the layers, and observed slope of the peripheral deposits toward the center of the basin disfavor origins such as volcanic lava flows or intrusions. Fine-grained, layered ignimbrite deposits (pumice-dominated, gravity-driven pyroclastic flow deposits with subordinate ash) are plausible given their susceptibility to erosion and the history of explosive volcanism in eastern Hellas, but the thick, well-bedded layering is difficult to reconcile given the distance from potential sources (Tyrrhena Patera and Syrtis Major). Moreover, pyroclastic deposits do not typically exhibit a strong, constant elevation control on their limits over thousands of kilometers. A fluvial (deltaic) origin was proposed for the layered deposits in Terby crater based on the morphology, which was reported to be reminiscent of the transition from topset to foreset beds (Ansan et al., 2005). However, coarse-grained material and lenticular channels that are associated with this process have not yet been identified, and the continuity of the layers is not typical of alluvial fan deposition that is characterized by frequent avulsions. Moreover, there is no evidence for a source valley(s) capable of supplying requisite sediment to Terby, which abuts rugged topography to the north. Niesten and SW craters have similar deposits and their northern rims are unbroken and higher than the adjacent plains. Deposition in a basin-filling aeolian sand erg is not likely due to the regularity of the layers, the thermal properties of the deposits that indicate material finer than sand, and the paucity of cross-bedding. Tanaka and Leonard (1995) favor an aeolian sand origin for the basin deposits (e.g., honeycomb terrain, unit ht and Alpheus Colles plateau, unit pl) to account for the observed volume of material, based on the assumption that the material available for non-aeolian deposition originated from the 2300 km diameter landform-defined basin rim. Because the available material in the basin was not sufficient to account for volume in the interior deposits, they assumed that the majority of the deposit originated beyond the rim of Hellas, requiring aeolian transport. Subsequent topographic data from MOLA, however, revealed a larger (5000 km across) “drainage” basin, dissected by numerous channels (Figure 7.1), which could have potentially transported the requisite sediment.

Evidence for ancient lakes in the Hellas region

215

Several hypotheses have been forwarded to explain the origin of the honeycomb (unit ht) and reticulate (unit r) terrains, and the adjacent complex topography of Alpheus Colles (unit pl) in the western Hellas interior. Some processes, such as ductile deformation from salt diapirs or lower crustal tectonism (e.g., Mangold and Allemand, 2003), could have occurred independently of any lacustrine environment. Other proposed mechanisms, such as ductile deformation, are potentially genetic to the lacustrine scenario (Howard and Moore, 2009; Moore and Wilhelms, 2001). Additional coverage and detail from CTX, HiRISE and mineralogical data from CRISM may help decipher the nature and origin of these deposits. Despite subsequent modification and erosion of the layered deposits in craters around Hellas in the form of troughs and pits, the upper surface of the deposits likely approximates their original depth of accumulation. The gentle gradients of the upper, Hellas-distal portions of the layered deposits in Terby, Niesten, Millochau, and SW crater suggest a strong gravitational control on layer deposition (Figures 7.7 and 7.10). Accumulation as dunes or as direct air-fall deposition of aeolian dust, volcanic ash, or globally redistributed impact ejecta is unlikely to result in planar, nearly level deposi­ tional surfaces. Moreover, the aforementioned depositional mechanisms would not be expected to produce the observed break in slope along the surface of accumulation seen in Niesten, Millochau, and SW crater (Figure 7.10). The apparent gravitational control on layered deposit morphology and hydrated mineral signature of the layered deposits in Terby and Niesten craters suggest that the layers were deposited, or at least modified, by lacustrine processes, but a rhythmic air-fall or loess-like origin for the layers cannot be definitively ruled out (Wilson et al., 2007a). If lakes were present in Hellas during air-fall deposition events, the water would act as a natural sediment trap, whereas on dry land, deposition might not occur or might be subject to later wind redeposition or removal. The lacustrine environment might encou­ rage diagenetic cementation of sediments, protecting them from subsequent aeolian deflation. If a deeper-water environment persisted in Hellas, submerged craters would also be sediment traps, encouraging deposition and discouraging erosion because of protection from waves and currents (including density currents). Longshore drift and density flows could also deliver sediment to craters that were only shallowly submerged, as are similarly seen in local basins of terrestrial oceans (e.g., Lamb et al., 2005).

7.7

Conclusions

The Hellas impact basin is one of the oldest recognizable features on Mars, and Hellas Planitia and the surroundings highlands preserve a long and complex period of Martian history that is critical to the overall understanding of the geology and past climate of the planet. Although the nature of the climate on early Mars remains unresolved, topographic, morphologic, stratigraphic, and mineralogic observations in Hellas are consistent with aqueous processes and a lacustrine environment during the Noachian and Hesperian. The broad physiography of Hellas is defined by the regional slope toward the center of the basin, which serves as a natural trap for material eroded from the

216

Lakes on Mars

surrounding highlands. Ubiquitous valley networks and channels, formed by water, dissect the entire periphery of Hellas. Reconstruction of the extant valley networks and channels in the Hellas region at the cessation of widespread runoff near the Noachian–Hesperian boundary demonstrates that the basin received drainage from a wide area on its rim and surrounding uplands, which we believe could have supported lakes in Hellas. The valleys that dissect the highland paterae in eastern Hellas Planitia extend to at least –6000 m elevation, attesting to fluvial processes that were active into the Hesperian. The layered nature of the materials in Hellas is consistent with a lacustrine depositional origin, and their relation to contours of constant elevation around the basin provides additional support for deposition into stands of water. Well-developed inward-facing scarps, abrupt changes in morphology, zones of differentially cratered terrain, and characteristic sequences of layered deposits that occur at comparable elevations around Hellas provide evidence for paleolake shor­ elines (Crown et al., 2005; Moore and Wilhelms, 2001). The lake levels in Hellas likely fluctuated, reaching at least –1800 m in elevation (Crown et al., 2005), but decreased through time to a late-stage lake in the Early Hesperian with a final shore­ line around –6000 m. Smaller lakes may have persisted into the beginning of the Amazonian. A terminal lake margin around –6000 m in elevation is consistent with the best preserved suites of landforms and unit contacts around the basin at –5800 m, the most compelling of the shorelines proposed by Moore and Wilhelms (2001). In eastern Hellas, the –5800 m contour marks the morphological boundary in the lower reaches of Harmakhis and Dao Valles that is analogous to terrestrial marine channels above –5800 m, and underwater extensions of such large rivers below. The nature of layered deposits along the depositional shelf at this elevation is consistent with a water-laid origin or deposition into a Hellas-wide lake (Bleamaster and Crown, 2010; Crown et al., 2005). The morphology and mineralogy of the extensive band of reticulate material along the –5800 m contour in western Hellas are suggestive of an aqueous environment, perhaps marking the margin of an ice-covered lake (Moore and Wilhelms, 2001). Finally, the abrupt change in texture marking the contact between the plains unit and lower mantling unit, both interpreted to be fluvial or lake sediment, is associated with this contour where not obscured by subsequent volcanic deposits (e.g., Bleamaster and Crown, 2010; Crown et al., 2009; Moore and Wilhelms, 2001). Although the origin of the deposits in Hellas Planitia is complex and unresolved, the thermal properties and morphology of the landforms in the center of the basin provide additional evidence that is consistent with a lacustrine environment. The dominant intermediate thermal signature of Hellas Planitia is interpreted to be material smaller than medium sand-sized grains or crust-bonded dust, which is most consistent with sedimentation in stands of water given the regional context (Moore and Edgett, 1993). The northeast/southwest structural trend of the ridges in western Alpheus Colles and the similar small-scale structural grain observed in the adjacent honeycomb and nearby reticulate terrains may be the result of deformation from a series of large subaqueous mega-slides (Howard and Moore, 2009). Additional supportive evidence for materials in Hellas consistent with a lacustrine origin comes from the nature of layered deposits in circum-Hellas craters (Figure 7.2).

Evidence for ancient lakes in the Hellas region

217

Up to 2.5 km-thick sequences of laterally continuous beds composed of hydrated sediment finer than sand were deposited by suspension from a fluid (air or water). The deposits, generally thickest on the Hellas-distal side of their respective basins, are similar in tone, layering properties, asymmetrical distribution, and original deposi­ tional geometry, suggesting that layered deposits at comparable elevations share a common age and emplacement mechanism. The gentle gradients and nearly level depositional surfaces on the upper Hellas-distal portions of the deposits in Terby, Niesten, Millochau, and SW crater suggest a strong gravitational control during deposition of the layers. This observation in combination with the hydrated signature of the sediment suggests that the layered units were deposited, or atleast modified, by lacustrine processes. If the layered deposits in Hellas and at low elevations along its rim (e.g., Terby, Niesten, and SW crater) were deposited in a lacustrine environment, the associated paleolake would have extended to at least –1800 m in elevation (Figure 7.2). Submersed craters would act as natural sediment traps and the lacustrine environment might encourage diagenetic cementation of trapped sediment, protecting it from later aeolian deflation. Similar layered deposits in craters at higher elevations (Millochau, and craters A and B) may have been deposited in an early, Hellas-wide sea, but likely hosted closed systems with isolated lakes during the Noachian to Early Hesperian. The presence of a large, deep, Hellas-wide lake(s) with local geothermal heat sources could have provided a relatively stable, long-lived aqueous environment for life compared to smaller, and presumably ephemeral, crater lakes elsewhere on Mars. Further studies of this region, including the synthesis of mineralogical and morphological data and in situ analyses, should help validate or refute the existence of large, ancient lakes in Hellas.

References Alfaro, P., Moretti, M., Soria, J., 1997. Soft-sediment deformation structures induced by earth­ quakes (seismites) in Pliocene lacustrine deposits (Guadix-Baza Basin, central Betic Cordillera). Eclogae Geol. Helv. 90 (3), 531–540. Allen, J.R.L., 1984. Sedimentary structures: their character and physical basis, unabridged onevolume edition. In: Developments in Sedimentology, 30. Elsevier, Amsterdam, Oxford, New York, Tokyo, 663p. Alvarez, W., Staley, E., O’Connor, D., Chan, M.A., 1998. Synsedimentary deformation in the Jurassic of southeastern Utah – a case of impact shaking? Geology 26 (7), 579–582. Ansan, V., Mangold, N., Lucas, A., Gendrin, A., LeMouelic, S., Poulet, F., et al., and the OMEGA co-Investigator Team, 2005. Analysis of layered deposits in Terby crater (Hellas region, Mars) using multiple datasets MOC, THEMIS, and OMEGA/MEX data. Lunar Planet. Sci., XXXVI (CD-ROM), Abstract 1324. Baker, V.R., 1982. The Channels of Mars. University of Texas Press, Austin. Bandfield, J.L., 2008. High-silica deposits of an aqueous origin in western Hellas basin, Mars. Geophys. Res. Lett. 35, L12205, doi: 10.1029/2008GL033807. Bandfield, J.L., Mayorga, L.C., Edwards, C.S., Glotch, T.C., 2009. An integrated analysis of Martian surface compositions using near infrared through thermal infrared spectroscopic data. Lunar and Planet. Sci., XXXX (CD-ROM), Abstract 1072.

218

Lakes on Mars

Bibring, J.P., et al., 2006. Global mineralogical and aqueous Mars history derived from OMEGA/Mars Express data. Science 312, 400–404. Bleamaster, L.F. III, Crown, D.A., 2010, Geologic map of MTM -40277, -45277, -40272, and -45272 quadrangles, eastern Hellas Planitia region of Mars: U.S. Geological Survey Scientific Investigations Map 3096 [http://pubs.usgs.gov/sim/3096/]. Cabrol, N.A., 1991. Etude des paleo-chenaux martiens: caractéristiques morphologiques, distribution spatio-temporelle, mécanismes de genèse. Université de Paris I Panthéon­ Sorbonne/Observatoire de Paris-Meudon, Ph-D Thesis, 413 pp. Carr, M.H., 1981. The Surface of Mars. Yale University Press, New Haven, Conn. Carr, M.H., Chung, F.C., 1997. Martian drainage densities. J. Geophys. Res. 102, 9145–9152. Christensen, P.R., 1988. Global albedo variations on Mars: Implications for active aeolian transport, deposition and erosion. J. Geophys. Res. 93, 7611–7624. Clifford, S.M., 1993. A model for the hydrologic and climatic behavior of water on Mars. J. Geophys. Res. 98, 10973–11016. Craddock, R.A., Howard, A.D., 2002. The case for precipitation on a warm, wet early Mars. J. Geophys. Res. 107, E11, doi:10.1029/2001JE001505. Craddock, R.A., Maxwell, T.A., Howard, A.D., 1997. Crater morphometry and modification in the Sinus Sabaeus and Margaritifer Sinus regions of Mars. J. Geophys. Res. 102, 13321–13340. Crown, D.A., Bleamaster, L.F. III, Mest, S.C., 2005. Styles and timing of volatile-driven activity in the eastern Hellas region of Mars. J. Geophys. Res. 110, E112S122, doi:110.1029/2005JE002496. Crown, D.A., Bleamaster, L.F., Mest, S.C., Mustard, J.F., 2009. Geologic mapping of the NW rim of Hellas basin, Mars. Lunar and Planet. Sci. XL (CD-ROM), Abstract 1705. Crown, D.A., Greeley, R., 1993. Volcanic geology of hadriaca patera and the eastern Hellas region of Mars. J. Geophys. Res. 98, 3431–3451. Crown, D.A., Greeley, R., 2007. Geologic map of the MTM –30262 and –30267 quadrangles, Hadriaca Patera Region of Mars: U.S. Geological Survey Miscellaneous Investigation Series Map I–2936, scale 1:1,000,000. Crown, D.A., Price, K.H., Greeley, R., 1992. Geologic evolution of the east rim of the Hellas basin, Mars. Icarus 100, 1–25. Dalrymple, R.W., 1979. Wave-induced liquefaction: a modern example from the Bay of Fundy. Sedimentology 26 (6), 835–844. Fassett, C.I., Head, J.W. III, 2006. Valleys on hecates tholus, Mars: origin by basal melting of summit snowpack. Planet. Space Sci. 54, 370–378. Fassett, C.I., Head, J.W., 2007. Valley formation on Martian volcanoes in the Hesperian: evidence for melting of summit snowpack, caldera lake formation, drainage and erosion on Ceraunius Tholus. Icarus 189, 118–135. Fassett, C.I., Head, J.W., 2008. The timing of Martian valley network activity: constraints from buffered crater counting. Icarus 195, 61–89. Fisher, R.V., Schmincke, H.-U., 1984. Pyroclastic Rocks. Springer-Verlag, Berlin, Heidelberg, New York, Tokyo, 472 p. Gaidos, E., Marion, G., 2003. Geological and geochemical legacy of a cold early Mars. J. Geophys. Res. 108 (E6), 5055, doi:10.1029/2002JE002000. Goldspiel, J.M., Squyres, S.W., 2000. Groundwater sapping and valley formation on Mars. Icarus 148, 176–192. Grant, J.A., 2000. Valley formation in Margaritifer Sinus, Mars, by precipitation-recharged ground-water sapping. Geology 28, 223–226.

Evidence for ancient lakes in the Hellas region

219

Greeley, R., Crown, D.A., 1990. Volcanic geology of tyrrhena patera, Mars. J. Geophys. Res. 95, 7133–7149. Greeley, R., Guest, J.E., 1987. Geologic map of the eastern equatorial region of Mars. U.S. Geological Survey Miscellaneous Investigation Series I–1802–B, scale 1:15,000,000. Gregg, T.K.P., Crown, D.A., Greeley, R., 1998. Geologic map of the Tyrrhena Patera region of Mars: U. S. Geological Survey Miscellaneous Investigations Series I–2556, scale 1:500,000. Gulick, V.C., 1998. Magmatic intrusions and a hydrothermal origin for fluvial valleys on Mars. J. Geophys. Res. 103, 19365–19387. Gulick, V.C., Baker, V.R., 1990. Origin and evolution of valleys on Martian volcanoes. J. Geophys. Res. 95, 14325–14344. Holt, J.W., Safaeinili, A., Plaut, J.J., Head, J.W., Phillips, R.J., Seu, R., et al., 2008. Radar sounding evidence for buried glaciers in the southern mid-latitudes of Mars. Science, 322 (5905), 1235–1238, doi: 10.1126/science.1164246. Howard, A.D., Moore, J.M., 2009. Subaqueous mega-slides on the floor of Hellas. Lunar Planet. Sci. Conf. XXXX, Abstract 1376. Howard, A.D., Moore, J.M., Irwin, R.P. III, 2005. An intense terminal epoch of widespread fluvial activity on early Mars: 1. Valley network incision and associated deposits. J. Geophys. Res. 110, E12S14, doi:10.1029/2005JE02459. Irwin, R.P. III, Howard, A.D., 2002. Drainage basin evolution in Noachian Terra Cimmeria, Mars. J. Geophys. Res. 107 (E7), 1–23, 10.1029/2001JE001818. Irwin, R.P. III, Howard, A.D., Craddock, R.A., 2008. Fluvial valley networks on mars. In: S.P. Rice, et al. (Eds.), River Confluences, Tributaries and the Fluvial Network. John Wiley & Sons, Chichester, pp. 419–450. Irwin, R.P. III, Howard, A.D., Craddock, R.A., Moore, J.M., 2005. An intense terminal epoch of widespread fluvial activity on early Mars: 2. Increased runoff and paleolake development. J. Geophys. Res. 110, E12S15, doi:10.1029/2005JE002460. Iverson, N.R., 1999. Coupling between a glacier and a soft bed: II. Model results. J. Glaciol. 45 (149), 41–53. Jackson, M.P.A., Cornelius, R.R., Craig, C.H., Gansser, A., Stocklin, J., Talbot, C.J., 1990. Salt diapirs of the Great Kavir, central Iran. Geol. Soc. Am. Mem. 177, 139 pp. Kargel, J.S., Strom, R.G., 1992. Ancient glaciation on Mars. Geology 20, 3–7. Kite, E.S., Manga, M., Perron, J.T., 2009. Evidence for past kilometer-scale overturn(s) in deformed, layered terrain near the deepest point on mars, Lunar and Planet. Sci., XXXX (CD-ROM), Abstract 1248. Korteniemi, J., Kostama, V.-P., Törmänen, T., Aittola, M., Öhman, T., Lahtela, H., et al., 2005b. Complex geology of two large impact craters in Tyrrhena Terra, Mars: detailed analysis using MEX HRSC camera data. J. Geophys. Res. 110 (E12), E12S18, doi:10.1029/ 2005JE002427. Korteniemi, J., Lahtela, H., Raitala, J., Neukum, G., and the HRSC Co-Investigator Team, , and 2005a. Anomalous depressions on the circum-Hellas crater floors as seen in the first year MEX HRSC images. Lunar Planet. Sci. XXXVI (CD-ROM), Abstract 1669. Lamb, M.P., Toniolo, H., Parker, G., 2005. Trapping of sustained turbidity currents by intraslope minibasins. Sedimentology 53, 147–160. Laskar, J., Correia, A.C.M., Gastineau, M., Joutel, F., Levrard, B., Robutel, P., 2004. Long term evolution and chaotic diffusion of the insolation quantities of Mars. Icarus 170, 343–364. Leonard, G.J., Tanaka, K.L., 2001. Geologic map of the Hellas region of Mars, U.S Geological Survey Miscellanous Investigation Series, I–2694, scale 1:5,000,000.

220

Lakes on Mars

Lewis, K.W., Aharonson, O., Grotzinger, J.P., Kirk, R.L., McEwen, A.S., Suer, T.-A., 2008. Quasi-periodic bedding in the sedimentary rock record of Mars. Science 322, 1532–1535, doi:10.1126/science.1161870. Malin, M.C., Edgett, K.S., 2000. Sedimentary rocks of early Mars. Science 290 (5498), 1927–1937. Malin, M.C., Edgett, K.S., 2001. The Mars global surveyor Mars orbiter camera: interplanetary cruise through primary mission. J. Geophys. Res. 106, 23429–23570. Mangold, N., 2003. Geomorphic analysis of lobate debris aprons on Mars at Mars Orbiter Camera scale: evidence for ice sublimation initiated by fractures. J. Geophys. Res. 108, 13, doi:10.1029/2002JE001885. Mangold, N., Allemand, P. (2003), Ductile deformation in Hellas floor: salt diapirs or crustal domes? Sixth International Conf. on Mars, Abstract 3047. Mars Channel Working Group, 1983. Channels and valleys on Mars. Geol. Soc. Am. Bull. 94, 1035–1054. McCauley, J.F., Carr, M.H., Cutts, J.A., Hartmann, W.K., Masursky, H., Milton, D.J., et al., 1972. Preliminary Mariner 9 report on the geology of Mars. Icarus 17, 289–327. Mellon, M.T., Jakosky, B.M., Kieffer, H.H., Christensen, P.R., 2000. High-resolution thermal inertia mapping from the Mars global surveyor thermal emission spectrometer. Icarus 148, 437–455. Mest, S.C., Crown, D.A., 2001. Geology of the Reull Vallis region, Mars. Icarus 153, 89–110. Mest, S.C., Crown, D.A., 2002. Geologic map of the MTM –40252 and –40257 quadrangles, Reull Vallis region, Mars, U.S Geological Survey Miscellaneous Investigation Series, I–2694, scale 1:500,000. Mest, S.C., Crown, D.A., 2003. Geologic map of MTM –45252 and –45257 quadrangles, Reull Vallis region of Mars: U.S Geological Survey Miscellanous Investigation Series, I–2694, scale 1:500,000. Mest, S.C., Crown, D.A., 2005. Millochau crater, Mars: infilling and erosion of an ancient highland impact crater. Icarus 175, 335–359. Mest, S.C., Crown, D.A., 2006. Geologic map of MTM –20272 and –25272 quadrangles, Tyrrhena Terra region of Mars: U.S Geological Survey Miscellaneous Investigation Series, I–2934, scale 1:1,004,000. Mest, S.C., Crown, D.A., Bleamaster, L.F., Mustard, J.F., 2008. Morphology, morphometry, and distribution of craters in NW Hellas, Mars. Lunar and Planet. Sci. XXXIX (CD-ROM), Abstract 1704. Milliken, R., Swayze, G., Arvidson, R., Bishop, J., Clark, R., Ehlmann, B., et al., 2008. Opaline silica in young deposits on Mars. Geology 36, 847–850. Moore, J.M., Edgett, K.S., 1993. Hellas Planitia, Mars: site of net dust erosion and implications for the nature of basin floor deposits. Geophys. Res. Lett. 20, 1599–1602. Moore, J.M., Howard, A.D., 2005a. Large alluvial fans on Mars. J. Geophys. Res. 110 (E04005), 24, doi:10.1029/2004JE002352. Moore, J.M., Howard, A.D., 2005b. Layered deposits and pitted terrain in the circum Hellas region, Lunar Planet. Sci. XXXVI (CD-ROM), Abstract 1512. Moore, J.M., Wilhelms, D.E., 2001. Hellas as a possible site of ancient ice-covered lakes on Mars. Icarus 154 (2), 258–276. Moore, J.M., Wilhelms, D.E., 2007. Geologic map of part of western Hellas Planitia, Mars, U.S Geological Survey Miscellaneous Investigation Series, I–2953. Murchie, S.L., Mustard, J.F., Ehlmann, B.L., Milliken, R.E., Bishop, J.L., McKeown, N.K., et al., 2009. A synthesis of Martian aqueous mineralogy after one Mars year of observations

Evidence for ancient lakes in the Hellas region

221

from the Mars Reconnaissance Orbiter. J. Geophys., Res., 114, E00D06, doi:10.1029/ 2009JE003342. Murray, B.C., Soderblom, L.A., Sharp, R.P., Cutts, J.A., 1971. The surface of Mars: 1. Cratered terrains. J. Geophys. Res. 76, 313–330. Mustard, J.F., Cooper, C.D., Rifkin, M.K., 2001. Evidence for recent climate change on Mars from the identification of youthful near-surface ground ice. Nature 412, 4111–4114. Peterson, J.E., 1977. Geologic map of the Noachis Quadrangle of Mars, U. S. Geological Survey Miscellaneous Investigation Series, I–910 (MC–27). Pierce, T.L., Crown, D.A., 2003. Morphologic and topographic analyses of debris aprons in the eastern Hellas region, Mars. Icarus 13, 46–65. Pieri, D.C., 1980. Martian valleys: morphology, distribution, age and origin. Science 210, 895–897. Potter, D.B., 1976. Geologic map of the Hellas quadrangle of Mars: U.S. Geological Survey Miscellaneous Investigation Series I–941, scale 1:5,000,000. Poulet, F., Bibring, J.-P., Mustard, J.F., Gendrin, A., Mangold, N., Langevin, Y., et al., 2005. Phyllosilicates on Mars and implications for early Martian climate. Nature 438, 623–627, doi:10.1038/nature04274. Price, K.H., 1998. Geologic map of the Dao, Harmakhis, and Reull Valles Region of Mars, U.S. Geological Survey Miscellanous Investigation Series, I–2557, scale 1:1,000,000. Putzig, N.E., Mellon, M.T., Kretke, K.A., Arvidson, R.E., 2005. Global thermal inertia and surface properties of Mars from the MGS mapping mission. Icarus 173, 325–341. Reading, H.G. (Ed.), 1996. Sedimentary Environments and Facies, 3rd ed. Blackwell Scientific Publications, Oxford, 687pp. Rossetti, D.D.F., Góes, A.M., Truckenbrodt, W., Anaisse, J., 2000. Tsunami-induced large-scale scour-and-fill structures in Late Albian to Cenomanian deposits of the Grajau Basin, northern Brazil. Sedimentology 47 (2), 309–323. Sarkar, S., Chanfa, S.K., Bhattacharya, A., 1982. Soft-sediment deformation fabric in the Precambrian Bhander oolite, central India. J. Sediment. Petrol. 52 (1), 95–107. Sharp, R.P., Malin, M.C., 1975. Channels on Mars. Geol. Soc Am. Bull. 86, 593–609. Smith, D.E., et al., 1999. The global topography of Mars and implications for surface evolution. Science 184, 1421–1576. Soderblom, L.A., et al., 1973. Latitudinal distribution of a debris mantle on the Martian surface. J. Geophys. Res. 78 (20), 4117–4122. Squyres, S.W., 1979. The distribution of lobate debris aprons and similar flows on Mars. J. Geophys. Res. 84 (B14), 8087–8096. Squyres, S.W., Wilhelms, D.E., Moosman, A.C., 1987. Large-scale volcano-ground ice inter­ actions on Mars. Icarus 70, 385–408. Stepinski, T.F., Stepinski, A.P., 2005. Morphology of drainage basins as an indicator of climate on early Mars. J. Geophys. Res. 110 (E12S12), 10, doi:10.1029/2005JE002448. Tanaka, K.L., Kargel, J.S., MacKinnon, D.J., Hare, T.M., Hoffman, N., 2002. Catastrophic erosion of Hellas basin rim on Mars induced by magmatic intrusion into volatile-rich rocks. Geophys. Res. Lett. 29, 371–374. Tanaka, K.L., Leonard, G.J., 1995. Geology and landscape of the Hellas region of Mars. J. Geophys. Res. 100 (E3), 5407–5432. Tanaka, K.L., Scott, D.H., Greeley, R., 1992. Global stratigraphy. In: Keiffer, H.H., Jakosky, B. M., Snyder, C.W., and Matthews, M.S. (eds), Mars. University of Arizona Press, Tuscon, pp. 345–382. Van der Wateren, F.M., 1995. Structural geology and sedimentology of push moraines – Processes of soft sediment deformation in a glacial environment and the distribution of glaciotectonic styles. Mededelingen – Rijks Geologische Dienst 54, 1–168.

222

Lakes on Mars

Weitz, C.M., Milliken, R.E., Grant, J.A., McEwen, A.S., Williams, R.M.S., Bishop, J.L., 2008. Light-toned strata and inverted channels adjacent to Juventae and Ganges Chasmata, Mars. Geophys. Res. Lett. 35 (L19202), 5, doi:19210.11029/12008GL035317. Weitz, C.M., Milliken, R.E., Grant, J.A., McEwen, A.S., Williams, R.M.E., Bishop, J.L., et al., 2009. Mars Reconnaissance Orbiter observations of light-toned layered deposits and associated fluvial landforms on the plains adjacent to Valles Marineris. Icarus, 205 (1), 73–102, doi: 10.1016/j.icarus.2009.04.017. Werner, S.C., 2008. The early Martian evolution – Constraints from basin forming ages. Icarus 195, 45–60. Williams, D.A., Greeley, R., Fergason, R.L., Kuzmin, R., McCord, T., Combe, J.-P., et al., and the HRSC Co-Investigator Team, 2008. The Circum-Hellas Volcanic Province, Mars: Overview. Planet. Space Sci., doi:10.1016/j.pss.2008.08.010. Wilson, S.A., Doe Nobrea, E., Howard, A.D., Moore, J.M., Cohen, B., 2007a. Terby crater: MSL landing site candidate, presentation at the Second Mars Science Laboratory Landing Site Workshop. Old Town Pasadena, CA, October 23–25. Wilson, S.A., Howard, A.D., Moore, J.M., Grant, J.A., 2007b. The geomorphic and strati­ graphic analysis of crater Terby and layered deposits north of Hellas basin, Mars. J. Geophys. Res. 112 (E08009), 39, doi:10.1029/2006JE002830.

8 Deltas and valley networks on Mars: implications for a global hydrosphere

Gaetano di Achille‡ and Brian M. Hynek†,* †

Laboratory for Atmospheric and Space Physics, University of Colorado, Boulder, CO, USA * Department of Geological Sciences, University of Colorado, CO, USA ‡ Present address: Research and Scientific Support Department, European Space Agency, ESA-ESTEC, Noordwijk, The Netherlands

8.1

Introduction

During the last decade of Mars exploration, the progressive increase of both the resolution and coverage of available imagery has provided evidence for a remark­ able variety of deltaic deposits (Cabrol and Grin, 1999; De Hon, 1992; Di Achille et al., 2006a, 2006b, 2007; Fassett and Head, 2005; Hauber et al., 2009; Irwin et al., 2005; Malin and Edgett, 2003; Ori et al., 2000a), confirming the occurrence of a complex and vast hydrological system during the early geological history of Mars (e.g., Baker, 2001; Baker et al., 1991; Carr, 1996). Deltaic depositional systems are the most prominent evidence suggesting the existence of long-lasting standing bodies of water on Mars. Due to this genetic link, Martian deltas are fundamental for the reconstruction of the paleohydrological cycles and paleoclimates of Mars, testifying to the occurrence of past climatic conditions that created hydrological settings quite different from those of modern Mars. Furthermore, deltas might be key to understanding potentially habitable periods in Mars history. In fact, their typical rapid burial capabilities make the deltaic deposits ideal environments for the preservation of organic material, and consequently, they are high-priority targets for future landed missions to Mars (e.g., Cabrol et al., 2001; Ehlmann et al., 2008; Ori et al., 2000b). Ancient terrains of Mars are covered with additional signatures of past water. Branching channel systems on Mars, known as “valley networks,” have long been viewed as some of the best evidence that water flowed across the surface (Carr, 1996; Milton, 1973). Valley networks were noted as evidence for ancient precipitation and surface runoff, implying complex river systems (Masursky, 1973; Milton, 1973; Sharp and Malin, 1975); however, further scrutiny of the data revealed immature drainage systems consisting of widely spaced, U-shaped valleys that sometimes had alcove-like heads (e.g., Carr, 1995; Pieri, 1980). Groundwater flow was thought to be responsible Lakes on Mars. DOI: 10.1016/B978-0-444-52854-4.00008-8 © 2010 Elsevier B.V. All rights reserved.

224

Lakes on Mars

for these characteristics, and this process was hypothesized to dominate Martian valley formation (Carr, 1995; Squyres and Kasting, 1994). Recent data, including topography and high-resolution imagery, have shed new light on valley network formation. It is now believed that most valley networks were formed by precipitation and surface runoff instead of groundwater processes (e.g., Craddock and Howard, 2002; Hynek and Phillips, 2003; Hynek et al., 2010). These authors note the extensive erosion that is seen globally as well as complex valley networks with their heads at topographic divides. These observations, among others, imply that climatic conditions were different on early Mars than at present, since the modern ∼6 mbar atmospheric pressure and sub-freezing temperatures pre­ vent liquid water from being stable at the surface. A number of these valley networks end at closed or open basins (Cabrol and Grin, 1999, 2001, 2005; Fassett and Head, 2008; Forsythe and Blackwelder, 1998), and their sedimentary deposits have been preserved until the present. Some of these include local paleolake deposits (often thinly layered, friable materials) and also fan-shaped deposits. Several questions about the formation and evolution of Martian deltaic deposits are still unresolved. First of all, it is unclear whether fan-shaped features found at the valley termini represent the unequivocal expression of deltaic deposition into standing bodies of water or the results of alternative mechanisms, such as alluvial transport, mass-wasting, eolian erosion, or glacial and volcanic processes, which do not neces­ sarily imply (if at all) the occurrence of significant ponding of water within the receiving basins (e.g., Leverington and Maxwell, 2004; Malin and Edgett, 2003; Moore and Howard, 2005). Second, the time required for the formation of the deposits, and thus their paleoclimatic and paleoenvironmental interpretations, is con­ troversial. Current estimates for the duration of these depositional systems range from years to hundreds of thousands of years (e.g., Kleinhans, 2005; Moore et al., 2003). Therefore, despite their significant paleoclimatic implications, Martian deltas cannot be uniquely used to assess whether they formed during extended epochs of clement climatic conditions or during limited and episodic climatic optima produced by regional factors, such as impact craters, volcanism, or tectonics and resultant hydro­ thermal activities (Cabrol et al., 1997; Gulick, 1998; Newsom et al., 1996; Segura et al., 2002). Additionally, deltas are a relatively rare landform across the planet (a few tens of deposits are currently known), and especially when compared to the large number of geomorphologic settings potentially favorable for their formation (e.g., the hundreds of basins and craters dissected by channels and valleys and interpreted to be possible paleolakes) (Cabrol and Grin, 1999, 2005; Cabrol et al., 2001; Fassett and Head, 2008; Howard, 2007). Finally, the effects of the reduced Martian gravity on the physics of the deltaic sedimentary processes and on the morphology/stratigraphy of the resultant deposits are unclear and poorly investigated. Here, we (i) briefly retrace the history of the detection of Martian deltas and valley networks, (ii) review the issues related to the interpretations of the fluvial systems and associated terminal deposits, and (iii) present the database of currently known deltas and valleys. The latter are used to implement a new test of the Martian ocean hypothesis (Baker et al., 1991; Parker et al., chapter 9, this book; Parker et al., 1989, 1993). Previous validations of the Martian ocean notion have primarily focused on topographic tests

Deltas and valley networks on Mars: implications for a global hydrosphere

225

of the putative ocean shorelines (e.g., Carr and Head, 2003; Head et al., 1999), while Dohm et al. (2009) have recently used global elemental distributions derived from GRS to test the possibility of paleooceans on Mars from a geochemical point of view. Here, we argue that the global distribution of ancient deltaic deposits and valley networks supports the occurrence of a global Martian hydrosphere; integrating valleys, deltas, and an ocean during the early history of the planet.

8.2

Deltas on Mars

8.2.1 Two decades of observations The presence of possible deltas on Mars and a global classification of Martian deposits and paleolakes were first proposed based on Viking images (Cabrol and Grin, 1999; De Hon, 1992; Ori et al., 2000a). The deposits were mainly identified as a few kilometers wide, single-lobe, fan-shaped features formed at the mouths of valleys opening into impact craters and showing the typical morphology of terres­ trial fan deltas, including Gilbert-type deltas (Figure 8.1a) (Gilbert, 1885). With MGS and the first meter-scale images from MOC (Malin et al., 1992), the first depositional feature showing a complex distributary pattern, mimicking that of a terrestrial river delta, was discovered (Figure 8.1b) (Malin and Edgett, 2003). This feature, referred to as the Eberswalde delta, is generally considered to be the best evidence for persistent water flow on Mars (e.g., Moore et al., 2003) and is one of the four finalist candidate landing sites for MSL. In recent years, all the previously suggested deposits have been imaged at high resolution and a few previously unreported deposits have been identified (Figures 8.11c, d and 8.2) (Di Achille et al., 2006a, 2006b, 2007; Fassett and Head, 2005; Hauber et al., 2009; Irwin et al., 2005; Pondrelli et al., 2005; Weitz et al., 2006). In particular, high-resolution topography (down to 75 m per pixel) provided by HRSC (Neukum et al., 2004) enabled the first detailed morphometric analyses of deltaic sedimentary deposits (e.g., Di Achille et al., 2006a, 2007, Pondrelli et al., 2008), allowing wellconstrained paleohydrologic reconstruction of possible lacustrine environments. The sub-meter resolution of HiRISE (McEwen et al., 2007) and the hyper-spectral capabilities of CRISM (Murchie et al., 2007) led to the detailed facies analysis of deltaic deposits and discovery of clay minerals and unambiguous strandlines within previously suggested deltas (Di Achille et al., 2009; Ehlmann et al., 2008; Pondrelli et al., 2008). Collectively, all the above observations have shown that deltaic depositional systems represented an important geologic setting on early Mars.

8.2.2 Delta database and global distribution Using THEMIS visible data, Irwin et al. (2005) re-examined all the previously reported putative Martian deltas based on Viking images (Cabrol and Grin, 1999), and many of them were not confirmed (see also footnote 22 in Malin and Edgett, 2003). Therefore, the database considered here is based on the catalog from Irwin

226

Lakes on Mars B Apex

A

–2200

Elevation (m)

a

–2500 –2800

Distance (km)

10

5

N

b

0

–3100

N

5 km

F

A

B F

5 km N

c

N

d

A

5 km B –2000

A Apex Elevation (m)

–2200

–2500

B

5 km

Distance (km) –2900

0

5

10

15

Figure 8.1 Morphology and morphometry of Martian deltaic deposits: (a) HRSC nadir image showing a single lobate depositional feature in Shalbatana Vallis (#9 in Table 8.1). The MOLA profile shows the characteristic distal step (F—delta front, M1). (b) Multilobate deposit within the Eberswalde crater (#28 in Table 8.1), showing evidence for vertical and lateral accretion and extensive channel avulsion over the putative delta plain; F marks the delta front (CTX). (c) HRSC nadir view of a complex fan-shaped feature at the mouth of Tyras Vallis (#16 in Table 8.1, M1). The HRSC longitudinal profile reveals an intermediate flat terrace (M2 morphometry) possibly formed due to wave erosion. (d) CTX image of a M3 feature in Coprates Catena. The deposit is characterized by a series of concentric steps along its surface. A channel entrenches only the proximal (pE) part of the fan.

et al. (2005), expanded to a total of 46 deposits (see Table 8.1) after the identification of several unreported deposits that we observed during a new survey using a THEMIS infrared global mosaic as a basemap and all the available HRSC nadir images (12.5–25 m/pixel of resolution) at the time of this writing. We conservatively ruled

Deltas and valley networks on Mars: implications for a global hydrosphere

N

a

227

N

b

F F

5 km

5 km N

c

N

d F

F

5 km

5 km

Figure 8.2 Morphology and morphometry of Martian deltaic deposits: (a) CTX image showing the depositional feature at the mouth of Sabrina Vallis (#7 in Table 8.1, M1), F marks the delta front; (b) multilobate deposit within the Jezero crater (#3 in Table 8.1), showing typical characteristics of terrestrial river-dominated deltas—evidence for vertical and lateral accretion, and channel avulsion over the putative delta plain; F marks the delta front (CTX); (c) CTX view of a single lobate fan-shaped feature at the mouth of an unnamed valley along the dichotomy boundary (#4 in Table 8.1, M1pE); (d) CTX image of deposit #5 in Table 8.1, M1pE. This deposit is found in the same basin of the deposit #4 (Figure 8.2c). Their delta fronts share a common elevation with respect to the MOLA datum (see Table 8.1), suggesting they formed during the same water highstand.

out all the deposits lacking clear evidence for distinct delta front and/or internal layering, and also highly eroded deposits. This resulted in the exclusion of the deposits #10, #22, and #27 in Irwin et al. (2005) as well as the deposit found in the Gusev crater at the mouth of Ma’adim Valley (e.g., Cabrol et al., 1996). Finally, given the progressive increase of available high-resolution images, our current database is by no means definitive and it may be expanded by further discoveries.

Table 8.1 Database of Martian deltaic deposits #

Longitiude (E)

Latitude

Fronta (m)

Hydrology

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

−159.27 77.61 77.6 132.7 132.85 142.62 −46.8 −45 −43.45 −44.15 −147.21 137 137.3 121.6 −15.76 −49.8 12.3 18.1 −52.7 −149.47 −155.2 −146.58 −58.6 −34.5 25.7 27.2

−8.6 18.45 18.4 −3.65 −5.15 −3.7 11.9 12 3.1 4.27 −5 −5.37 −5.8 2.1 8.95 8.5 30.9 34.3 12 −6.25 −15.65 −7.96 5.13 −26.8 29.5 28.1

−2289.05 −2528.3 −2466.31 −2482.92 −2451.71 −2533.87 −2548.89 −2648.71 −2792.24 −2722.2 −2533.55 −2753.11 −2600 −2218 −2800 −2437.96 −2624 −2421.71 −1010 −1712 89.41 −1000 −950 −2100 −2175 −2387.45

C Ew Ew Lw Lw C Ew Ew C C Ew C C C C Lw C C Lw C C C C C C C

Avulsion

Y Y

Y Y Y

Layering

Morphometry

Basinb

References

Y Y Y Y Y Y Y Y Y

M3 M1pE M1pE M1pE M1pE M1 M1 M1pE M1pE M1pE M1pE M1pE M1 M1 M3 M2pE M1 M1pE M1pE M1 M2 M3 M1E M1 M1 M1

O O O O O O O O O O O O O O C C C C C C C C C C C C

Ori et al. (2000a) Fassett and Head (2005) Fassett and Head (2005) Irwin et al. (2004) Irwin et al. (2004) This study Hauber et al. (2009) Hauber et al. (2009) Di Achille et al. (2007) Di Achille et al. (2007) This study Cabrol and Grin (2001) Cabrol and Grin (2001) Irwin et al. (2005) This study Di Achille et al. (2006a) Grant and Schultz (1993) Cabrol and Grin (1999) Cabrol and Grin (1999) Ori et al. (2000) Ori et al. (2000) Cabrol and Grin (2001) Cabrol and Grin (2001) Grant and Parker (2002) McGill (2002) McGill (2002)

Y

Y Y Y

Y

Moderate

Y Y

27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 a b

26.3 −33.7 83.1 −103.4 −6.35 11.6 −60.3 −53.75 −48 −49.4 −51.7 −51.3 149.73 148.8 148.48 144.6 140.3 17.5 141.14 −63

35.5 −23.8 −27.9 −39.2 −19.08 27.9 −15 −9.88 8.5 9.8 2.3 11.4 −9.2 −9.68 −7.7 −10.43 −5.62 33.9 −6.53 −8.31

−2434 −1248.99 −2233.94 2224.56 −306 −4150 777 1283.6 −1784.05 −1653.58 −2154.89 −2078 −1509 −423.91 −1682.99 −702.15 −2327 −2865 −2257.7 2237

C Ew Ew C C C Lw C Lw C C C Lw C C C Lw C C C

Y Y Y

Y Y

Moderate Moderate Y

Y Y Y

Moderate

Y Y Y Y

Moderate Y Moderate

MOLA elevation values of the delta fronts used for the ocean test (see Section 8.4.1). “O” indicates open basins (as defined in the Section 8.4.2); “C” for closed basins.

Y Y Y Y

M3pE M1 M1pE M2 M1pE M1 M3pE M3 M1pE M1 M3 M1 M1 M1 M1 M1 M1pE M1pE M3pE M2pE

O C C C C C C C C C C C C C C C C O C C

McGill (2002) Malin and Edgett (2003) Howard and Moore (2004) Mangold and Ansan (2005) Irwin et al. (2005) Irwin et al. (2005) Di Achille et al. (2006b) Di Achille and Komatsu (2008) Harrison and Grimm (2005) Hauber et al. (2009) Hauber et al. (2009) This study This study This study This study This study This study Cabrol and Grin (1999) This study This study

230

Lakes on Mars

Classification We consider the possibility of an end-member classification for the deposits as a way of introducing the database (Di Achille et al., 2008), trying to characterize key factors like the hydrologic characteristics of inflowing channels and the morphometric and sedimentary structures of the deltas.

Hydrology Based on the main characteristics of the inflowing channels (e.g., morphology, length, depth, and stream order), we divided the deposits into three main categories: C, Lw, and Ew sourced deposits, in order to indicate features found at the mouth of single and steep canyons (C), small and partially integrated watersheds (Lw), and extended watersheds with many tributaries (Ew), respectively.

Internal structure Stratigraphy and sedimentary dynamics of terminal deposits are strongly influenced by the hydrological characteristics of the channels. Therefore, we characterized deposits as a function of the presence of visible internal layering and/or geometries (L) and evidence for channel avulsion or multilobate deposition (A).

Morphometry From a morphometric point of view, we divided deposits into three main types: (M1) with a single distinct steep front in the distal part, likely corresponding to a traditional delta front (Figures 8.1a and 8.2); (M2) with a single intermediate terrace or slope break (Figure 8.1c); (M3) with radial profile characterized by multiple steps and breaks in slope (Figure 8.1d). Additionally, we considered the entrenchment on the putative delta plain as a key morphodynamic parameter. In particular, we distin­ guished between the entrenchment of the entire deposits from the apex down to the floor of the receiving basin (E) and a partial dissection limited only to the proximal or distal part of the deposits (pE). Table 8.1 summarizes the observational evidence. Only seven features are found at the mouth of extended and well-integrated watersheds (Ew), eight formed at the termini of limited hydrological basins (Lw), whereas 31 deposits are located at the mouth of steep canyons (C) rarely exceeding 100 km in length and lacking tributaries. Eleven deposits show significant evidence for channel diversion on the putative delta plains and possible multilobate deposition (A). Six additional deposits could have been affected by channel avulsion and multilobate depositional patterns; however, the majority (29) of the investigated features is not characterized by the parameter A. In general, a strong correlation between the Ew and A classification is evident. This is also in agreement with the characteristics of fluvial-dominated terrestrial deltas. Twenty-eight features show internal layers and structures (L), whereas the remaining 18 do not present unambiguous evidence for bedding, either because they are covered by eolian deposits or are relatively pristine and lacking of erosional windows. Morphological and topo­ graphic analyses, using both MOLA (Smith et al., 2001) and HRSC digital elevation models, revealed that the common morphometry is represented by the type M1 with 36

Deltas and valley networks on Mars: implications for a global hydrosphere

231

examples, of which 17 are not entrenched and 18 are partially dissected (pE), whereas only one of the deposits (#23, Table 8.1) shows clear evidence for entrenchment and also for a profile rejuvenation within the feeder channel. Finally, four deposits (of which two are pE: #16 and #46, Table 8.1) belong to the M2 category and six to the M3 (of which three are pE: #27, #34, and #45, Table 8.1). Based on the above parameters, the survey contains rather significant morphody­ namical variability among the known deposits. This reveals a major problem in identifying unambiguous end-member deposits for a classification in agreement with terrestrial studies, which emphasizes the difficulties in obtaining consistent data on a global spectrum of deltas (e.g., Syvitski and Saito, 2007). Nevertheless, the majority of the Martian deposits show a basic morphometry (M1) and sedimentology (simple stratigraphic patterns and limited or absent diversion of channels, L and A) and are found in association with short and steep canyons (C). On the other hand, there are only a few deposits characterized by a well-developed stratigraphy and switching of channels with possible multilobate depositional patterns, and they are found exclusively at the mouth of well-developed and regionally extended valley systems presumably sustained by persistent surface runoff discharges (Ew). Also M2 and M3 morpho-types are mainly found in association with C valleys. However, the presence of steps and terraces along the longitudinal stretch, in addition to visible layers and partial entrenchment (L and pE), makes these deposits unique. In particular, deposit #16 (prototype of the category M2 and shown in Figure 8.1c) consists of a sequence, up to 1000-m-thick, of fine layers and presents a large intermediate terrace along its slope. M3 deposits with multiple concentric steps and terraces represent the most unusual and unambiguous features (Figure 8.1d). The origin of these stepped fans is unclear (Di Achille et al., 2006b; Kraal et al., 2008; Weitz et al., 2007); however, the possibility that M3 deposits are highly eroded remnants of M1-type features cannot be discounted. Exposures of the internal layering and of stacked stratigraphic sequences of M1-type deposits could also mimic steps and terraces on the exhumed surfaces (Di Achille et al., 2006b). Finally, the almost total lack (excluding #23) of entirely entrenched deposits (E) is also apparent as is a significant occurrence (21 out of 46) of partially dissected deposits (pE).

8.2.3 Alternative hypotheses In addition to a sedimentary origin into lacustrine environments, alternative mechan­ isms should be evaluated to explain the formation of the fan-shaped Martian deposits. Indeed, multiple formation mechanisms might have been at work on early Mars, given its diverse geological settings and likely regional to global climate fluxes through time. Below we review the previous hypotheses based on the observational evidence and also in comparison with terrestrial analogues.

Alluvial vs. deltaic deposition Alluvial and deltaic deposits can share general morphologic characteristics as a result of similar transport and depositional processes (e.g., Figures 8.2 and 8.3). For

232

Lakes on Mars A

1400

Elevation (m)

1200 1000 800 600

Apex

Distance (km) 0

4

B

8

B

12

toe

N

5 km A

Figure 8.3 Example of a Martian alluvial fan (CTX): the deposit shows evidence for multiple distributary channels on its proximal part. The MOLA topographic profile shows a relatively constant slope and a distal termination concordant to the basin floor (compare with the profiles in Figure 8.1a and 8.1c).

example, both types of deposits could have distributary channels on their proximal part. However, these deposits are quite different in terms of depositional environments and resultant stratigraphic and depositional facies. Moreover, they have a different paleoenvironmental implication: alluvial fans form when sediments from a stream spread at the entrance of a receiving basin not occupied by water, whereas deltas form at the mouth of a river opening into a standing body of water. This distinction is quite straightforward in the field and efforts have been made to link morphology to formation mechanisms (Williams et al., 2006), but difficulties remain in unique interpretations from remote sensing data. Additionally, for the Martian case, the preservation status of the deposits can make the differentiation of deltas from alluvial fans, and vice versa, more difficult. Nevertheless, both large (km-scale) and smaller (sub-kilometer scale) alluvial fans have been investigated on Mars (e.g., Moore and Howard, 2005; Williams

Deltas and valley networks on Mars: implications for a global hydrosphere

233

and Malin, 2008), and they represent a distinct category of deposits with respect to the possible Martian deltas. This is especially clear after the first investigations of Martian fan-shaped deposits made by using HRSC topography (Di Achille et al., 2006a, 2006b, 2007; Hauber et al., 2009; Pondrelli et al., 2008). The latter topographic dataset, in fact, enabled the detailed morphometric analysis of the deposits, allowing the detection of breaks in slope and clear terminal steps (up to a few hundreds of meters high) along the longitudinal stretch of the deposits (Figure 8.1a, c). This characteristic is typical of deltaic deposits and is the result of the adjustment of the sedimentary deposition (through vertical aggradation and horizontal progradation) to the water base level within the receiving basins. Contrarily, alluvial fans typically have a constant or concave slope with a longitudinal gradient steeper than deltas (Blair and McPherson, 1994a, 1994b). Finally, they show a gradual downslope gradation toward the floor of the receiving basins with their distal part and toe typically concordant with the elevation of the basin floor (Figure 8.3).

A volcanic origin Putative Martian fluvial and lacustrine assemblages have been also suggested to be formed through igneous processes involving the flow and ponding of lava (Leverington and Maxwell, 2004; Leverington, 2006). Terraces along the slopes of impact craters and channels previously identified as inlets and outlets of candidate lacustrine basins were interpreted instead to be the result of the accu­ mulation of volcanic materials in topographic lows and of volcanic rilles, respec­ tively (Leverington and Maxwell, 2004; Leverington, 2006). Although these interpretations are potentially viable for the origin of channels and the infilling of craters, their potential for the formation of delta-like deposits appears less applicable. On Earth, delta-like deposits (known as lava deltas) are formed when lava flows enter the sea. Lateral and vertical coalescences of subaqueous elongate tongues can gradually produce a thick deltaic body with a flat subaerial portion and a coastal steep front characterized by basinward-dipping foreset­ bedded volcanic sedimentary sequences, connecting to the sea floor (Moore et al., 1973). Thus, the presence of a standing body of water would also be required in order to sustain a volcanic origin for the deposits. Moreover, except for the deposit in Rahe crater on the northern flank of Ceraunius Tholus (Fassett and Head, 2007), which is not considered a fluvial delta in the present study because of a lacking of sufficient evidence, none of the known possible Martian deltas are found close to volcanic districts or in relation with lava tubes or channels.

The glacial hypothesis On Earth, deltas can be produced in glacial basins and form in association with glaciers in two main settings: (i) as glacier-fed deltas, which do not differ much from regular deltas, apart from the fact that they receive sediment from glacial meltwater streams after transportation across an intervening land surface; and

234

Lakes on Mars

(ii) as ice-contact deltas (also referred to as kame deltas), which are built out directly from glacier margins (Benn and Evans, 2003). Among glaciolacustrine depositional environments, grounding-line fans (also referred to as subaqueous fans) and Gilbert-type deltas are the most common sedimentary associations. Grounding-line fans form where water from subglacial meltwater channels open directly into deep water and sediment load is deposited rapidly as a result of the sudden drop in stream velocity. Moreover, if sediment supplies are high and glacier margins remain stationary for enough time to let them aggradate, ground­ ing-line fans can build up to the water surface to form deltas (Powell, 1990). The discrimination between glaciolacustrine or regular deltas is not trivial, since ice-contact deltas have no diagnostic sedimentary characteristics (e.g., Benn and Evans, 2003). Nevertheless, contextual morphology (e.g., moraines, lineated val­ ley flows, and other glacial and periglacial features) and the regional geology can help to bolster claims for a glacial origin. Although poorly investigated at present, the above mentioned processes could have occurred also on Mars (Figure 8.4) and a proglacial origin of some of the known deposits cannot be discounted (Di Achille and Ori, 2008). Such a possibility, however, would not alter the main paleoclimatic significance of the potential deposits, since it still implies that they formed at glacier margins opening into standing bodies of water.

N 2 km

a

b

a

b

Figure 8.4 Candidate fan deposit for a glaciolacustrine origin (#44 in Table 8.1, M1pE). Although it is difficult to place constraints on the onset of the glacial activity, the latter could have accompanied its formation as a regular delta. MOC images show a feeder channel with a lineated valley floor (inset a), interpreted to be evidence for glacial flow and transport, and a delta surface characterized by the presence of large boulders over its proximal portion (inset b). Transport of such large (up to a few tens of meters) sedimentary fractions seem unlikely from water-driven processes.

Deltas and valley networks on Mars: implications for a global hydrosphere

8.3

235

Valley networks

All of the fan delta deposits on Mars identified to date occur at the termination of a system of valleys. The carving of these valley networks provided the erosional power to entrain the sediment that composes the terminal fan deposits. Understanding the formation of valley networks on Mars, including their timing, magnitude and duration of water flow, and associated implications for climate, has direct bearing on under­ standing the formation of their associated fan delta deposits. Most valley networks have no terminal deposits. Even systems thousands of kilometers long and with many tributaries have no remaining evidence of the amount of sediment eroded and moved downslope, likely due to their antiquity (most are >3.7 Ga old), and thus in many cases, the mouths of the systems and their deposits have been resurfaced by erosion or deposition of younger geologic materials. Still, the valleys associated with terminal fan delta deposits have bearing on the origin of these deposits. Moreover, as we argue below, a global study of valley network distribution provides constraints on the base level of a global Martian hydrosphere, possibly integrating valleys, deltas, lakes, and an ocean during their formation.

8.3.1 Valley network database: an updated view The combination of newly-acquired global datasets has allowed an unprecedented characterization of valleys across Mars. Global THEMIS daytime IR data (100 m/ pixel) and local higher resolution visible images and topography from MOLA (∼500 m/pixel) allow near-complete identification and characterization of valley net­ works on a global basis. A new map of global valley networks using these data has been produced (Figure 8.5a, b) (Hynek et al., 2010). Our mapping reveals approximately four times more valleys than previously identified, totalling a summed length ∼2 times greater than work from the Viking imagery (e.g., Carr and Chuang, 1997; Carr, 1995). Many smaller tributaries are seen that were not evident with Viking data, as well as structure within the valleys, including braided channels and terraces indicative of sustained flow. The newly-mapped drainage densities of networks show closely spaced valleys and much greater complexity in each network (Figure 8.6). All of these observations are consistent with valley formation by an active hydrologic cycle and resultant fluvial erosion. These results are in agreement with other recent work showing that precipitation and surface runoff must have been important processes in valley network formation (e.g., Craddock and Howard, 2002; Fassett and Head, 2008; Harri­ son and Grimm, 2005; Hynek and Phillips, 2001, 2003; Irwin and Howard, 2002; Irwin et al., 2005). However, in several places sapping appears to have been an important formation mechanism, particularly with the younger valley networks. As previously understood (e.g., Carr, 1995), most of the valleys occur on ancient terrains (84% lie entirely on Noachian terrain) (Figure 8.5a). These valleys occur over most of the ancient crust and are constrained in elevation with a rough Gaussian distribution centered around 1500 m relative to the Martian datum. Hesperian and Early Amazonian valleys also show signs of formation by precipitation and surface runoff, although at much decreased levels. These networks are mostly localized on

236

Lakes on Mars

60°S 30°S

0°

30°N

60°N

a

180°

150°W 120°W 90°W

60°W

30°W

0°

30°E

60°E

90°E

120°E 150°E

180°

60°S

30°S

0°

30°N

60°N

b

MOLA Elevation (km)

3 Terra sirenum Tharsis Tempa rise terra c 2

Chryse planitia

Arabia terra deuteronilus mensae

Nepenthes/Aeolis mensae

1 0 –1 –2

S1–S2 S1–S2

–3 –4 –5 –180

–120

Figure 8.5 (Continued)

–60

0 Longitude (degrees)

60

120

180

Deltas and valley networks on Mars: implications for a global hydrosphere

a

237

b

Figure 8.6 Comparison of valleys mapped by Carr (1995) with Viking data (a) to those identified by Hynek et al. (2009) in THEMIS and MOLA data (b) centered near 3°S, 5°E; both on a THEMIS basemap. Much greater maturity is evident in the new data suggesting a more substantial role of precipitation and surface runoff than previously thought.

volcanoes of these ages and the Hellas Basin (Figure 8.5a). The youngest valleys on the planet are confined to Amazonian volcanic constructs (Gulick and Baker, 1990), particularly in the Elysium and Tharsis regions, and most likely had a hydrothermal origin. Areas on Mars that have higher valley density correlate with paleolake basins noted by Fassett and Head (2008). This supports the idea that many valleys and associated deposits were generally formed by long-lived surface runoff from precipi­ tation; however, the Late Hesperian and younger periods show valley formation that Figure 8.5 (a) The new global map of Martian valley networks (black lines) from Hynek et al. (2010) draped on the terrains dated to the three major geological epochs of Mars (major craters in yellow) (Scott and Tanaka, 1986; Greeley and Guest, 1987); (b) MOLA color-coded shaded relief map of Mars with superimposed valley networks (black lines) from Hynek et al. (2010) and deltaic deposits topographically connected to the northern plains (red squares) and in closed basins (green triangles and blue diamonds). Closed-basins deltas indicated by blue diamonds have not been used to define the equipotential surfaces S1 and S2. Light blue marks the reconstructed ocean–land boundaries; white contours indicate the equipotential surfaces S1 and S2. The latter are shown as a single contour because they are too close to each other (S1 not shown in Hellas and Argyre for simplicity). (c) Plot of the deltas’ elevation as a function of longitude: S1–S2 (red line) indicates the inferred ocean levels. Gray dots represent elevation values of the “Arabia shoreline” extracted from the contact as demarcated in Clifford and Parker (2001). The average value from the distribution is about –2499 m; the black dashed line represents the linear trend line of the values. S1–S2 are consistent with the previous observational evidence at (i) Terra Sirenum (which also matches the terminations of valley networks, like for example Mangala Valles), (ii) in the northern part of Tempe Terra, (iii) along the circum-Chryse Planitia region (where also matches the openings of large outflow channels into the northern lowlands, e.g., Kasei and Maja Valles, Shalbatana Vallis, the complex system of Tiu/Simud Valles, and Ares Vallis), (iv) within the northern Arabia Terra and the fretted terrain regions of Deuteronilus Mensae, and finally (v) across the crustal dichotomy along the Nepenthes and Aeolis Mensae regions and surrounding Gale Crater.

238

Lakes on Mars

was more consistent with a sapping mechanism (Harrison and Grimm, 2005). In summary, we hypothesize that early Mars was characterized by a climate that could have supported an active hydrological cycle, which produced the majority of valley networks. Close association with a number of fan delta deposits suggests that these features were formed under the same climatic conditions, and thus most, if not all, had contributions from flowing water at the Martian surface.

8.4 A Test of the Martian ocean Evidence for complex valley network systems on Mars suggests that the planet was likely warmer and wetter during part of its early history. Sustained discharges debouch­ ing into the northern lowlands could have led to water accumulation in the northern lowlands and the formation of oceans (e.g., Baker, 2001; Baker et al., 1991). The theory of an ancient ocean has been repeatedly proposed and challenged over the last two decades. It was first proposed from Viking images (Baker et al., 1991; Parker et al., 1989) and subsequently expanded with the detection of a number of possible paleoshor­ elines inferred from morphological and topographic observations within the margins of the northern lowlands (e.g., Clifford and Parker, 2001; Parker et al., 1993). A complex hydrological cycle associated with transient oceans has been suggested (e.g., Baker et al., 1991). This would have implied the formation of large standing bodies of waters and episodic and cyclic climate changes during the early history of Mars (Baker, 2001). In such a case, Mars should have been characterized by the presence of a global hydrosphere, integrating precipitation, groundwater reservoirs, regions of ice accumula­ tion, and diffuse surface runoff toward an ocean with the formation of deltas along the coastal areas, as well as in smaller basins and paleolakes across the planet. This theory provided the foundation for the test implemented below. By using the global distribution of deltaic deposits (one of the most typical indicators of a longlived standing body of water on Earth) and valley networks in conjunction with topography from MOLA, a direct geological approach for the detection of a possible Martian global hydrosphere, including an ocean, and of equipotential surfaces reflect­ ing its past margins is possible.

8.4.1 Data and methods Marine deltas are among the most typical coastal landforms on Earth (e.g., Milliman and Meade, 1983). They exhibit a large diversity of morphologies as a result of a combination of several factors including climate, geology, river characteristics, basin bathymetry, and wave and tide regimes. Nevertheless, they share the characteristic of being formed approximately at the same elevation all over the planet, that is, the mean global sea level. Analyses of terrestrial deltas and their correlation across the planet have given clues on worldwide trends and changes of the mean sea level during the geological history of the Earth (e.g., Stanley and Warne, 1994). Using terrestrial analogy, Martian deltas could provide a reliable approach to test the existence of a putative Martian ocean and characterize its levels at a global scale,

Deltas and valley networks on Mars: implications for a global hydrosphere

239

Elevation (km)

assuming that (i) the present topography of Mars is close to that at the time of the existence of the putative ocean, and that (ii) the deltaic deposits formed during the same epoch. These assumptions are supported by the fact that the major topographic features of Mars should have been largely in place by the end of the Noachian (∼3.7 Ga) (Phillips et al., 2001) and that fans and deltas likely formed during an epoch of intense and terminal fluvial activity during the Noachian–Early Hesperian (3.7–3.4 Ga) (Irwin et al., 2005). We used the 46 candidate deltas reported in Table 8.1 in combination with the global topography from MOLA and all available imagery. In particular, by using both the global MOLA gridded dataset and the individual altimetry profiles (∼75 m spot size along orbital tracks every 300 m; see Figure 8.7), we extracted the elevation values of the apex (maximum water level), delta front (mean highstand), and the basin floor (minimum water level) for each delta (Figures 8.5b and 8.7). These values were used as proxies for the maximum water level excursion (apex elevation–floor eleva­ tion) and of the main highstands during the formation of the sedimentary deposits. These morphometric indicators have been plotted as a function of the longitude (Figure 8.5c) to detect equipotential surfaces indicative of possible ancient oceans’

A –2

Max B

Apex

Front

–2.2 –2.4 Distance (km)

–2.6 0

3

Min 6

9

B

N

A

1 km

Figure 8.7 CTX image of a sedimentary deposit in Nepenthes Mensae (#25, in Table 8.1). Dots indicate the location of the MOLA-shot measurements. The topographic profile AB shows the morphometry of the deposit and a schematic of the method used for the extraction of the elevation values for each delta. Error bars define the maximum water level excursion; the black square corresponds to the elevation of the delta front (mean water highstand).

240

Lakes on Mars

coastlines. This plot should also reflect the elevation of any emergent groundwater table saturating the planet’s crust and ponding on the surface across the planet (Clifford and Parker, 2001). Additionally, we extracted the termini elevations of all the Martian valleys, as mapped in Hynek et al. (2010), in order to compare them with the distribution of deltas. If a planet-wide standing body of water was present during the formation of the deltas, then it should have determined abrupt channel termina­ tions all over the planet at approximately the same elevation indicated by the analysis of the sedimentary deposits. Finally, our test does not take into account possible isostastic, tectonic, and subsidence movements that may have occurred after the disappearance of the putative oceans and displaced the contacts from their original positions. The latter assumptions were also implicit in all previous tests for the Martian oceans (e.g., Carr and Head, 2003; Head et al., 1999; Perron et al., 2007).

8.4.2 Results and discussion At a global scale the deposits appear preferentially located within a few hundreds of kilometers of distance upslope from the crustal dichotomy of the planet (apart from #29 and #30, Table 8.1) and show a regional correlation with the distribution of valley networks (Figure 8.5b). Clusters of deposits are concentrated within the Memnonia, Arabia Terra, Elysium Planitia, and especially Xanthe Terra regions. Figure 8.4b shows a plot of all delta fronts (mean inferred water level) as a function of longitude with error bars indicating the inferred maximum water level excursion for each delta. The highstand levels of all the deltaic deposits show a mean value of about –1747 m with a standard deviation of 1320 m. However, 17 of the deltas are not contained in a local basin or were formed in basins or outflow channels that connect to the northern lowlands (∼33% of total, red squares in Figure 8.5b, c, marked with “O” in Table 8.1). Collectively, these delta front elevations approximate an equipotential surface (here­ after referred to as S1) at the mean elevation of –2540 m with a standard deviation of 177 m (Figure 8.1a, c). Finally, the second closest approximation of a surface of equal gravitational potential (hereafter referred to as S2) is found at –2501 m with a standard deviation of 193 m by considering an additional nine deltas formed in closed basins (green triangles in Figure 8.5a, b) totaling ∼54% of the current global database. Although the distribution and elevation of all the 46 sedimentary deposits appear scattered and therefore not consistent with ocean coastlines or hydrostatic crustal ponding (Figure 8.5c), the deposits topographically connected to the northern lowlands (the site putatively occupied by the ocean) define the closest approximation of an equipotential surface (S1) as hypothesized by Parker et al. (1989) and others. A contour traced at this elevation effectively outlines a complete closure within and along the margins of the northern lowlands (Figures 8.5b and 8.8), encompassing the boundary of the basin within which the deposits could have formed. The standard deviation of the level inferred from the analysis of the deltas is considerably smaller if spread across the entire length of the global equipotential surface and remarkably smaller (up to one order of magnitude) than the dispersion values previously obtained for the elevations of Contact 1 (Arabia shoreline, Figure 8.5c) and Contact 2 (Deuteronilus shoreline, see also Chapter 9), which have been estimated to deviate in elevation across the planet up

Deltas and valley networks on Mars: implications for a global hydrosphere

a

180°

b

241

0°

S2

Argyre S1-S2

S2

S1-S2 90° 270°

270°

90° Hellas

S1-S2 S2 0°

Arabia terra 180°

Figure 8.8 (a) North polar stereographic projection (positive east longitude) from MOLA data showing the enclosures (black contours) determined by S1 and S2 within and along the northern lowlands. For simplicity S2 is shown as a separate contour only upslope of the S1–S2 boundaries where it determines the occurrence of several possible paleolakes, mostly concentrated in the gradational zone of Arabia Terra. (b) South polar stereographic projection from MOLA data showing the S2 surface (black contours) and resultant basins in Argyre and Hellas Planitia (S1 not shown for simplicity).

to 2.5 km and 0.7 km, respectively (Clifford and Parker, 2001; Perron et al., 2007). Moreover, the S1 level follows large portions of the “Arabia shoreline” previously identified from geomorphologic and topographic analyses (Figure 8.5b, c) (Clifford and Parker, 2001; Parker et al., 1993). In addition, the level of the inferred S1 surface is consistent with the average value of the “Arabia shoreline” (–2499 m, compare also the trend line of this contact with the S1 surface in Figure 8.5c). The second closest approximation of a surface of constant geopotential (S2) also traces a complete closure along the margins of the northern lowlands (Figures 8.5b, c and 8.8). However, since this has been inferred by considering deltas formed in basins not directly connected to the northern lowlands (green triangles in Figure 8.5b, c), it also implies that water should have been present at this base level all over the planet’s surface. Indeed, the S2 level is remarkably close to the –2550 m level suggested by theoretical calculations for the global distribution of water during the Noachian (Clifford and Parker, 2001). The latter value was determined from thermophysical properties of Mars with the assumption that water was saturating the crust and ponding in hydrostatic equilibrium on the surface of the planet, determining the occurrence of several standing bodies of water, including also those in the Hellas and Argyre Basins (Clifford and Parker, 2001, see also Chapter 7). The analysis of these 25 sedimentary deposits (∼54% of total deltas) provides supporting evidence of this thermal-hydraulic reconstruction of the Martian hydrosphere for the Noachian and is consistent with the existence of oceans and/or large seas in the northern hemisphere, Argyre, and Hellas Basins along with several groundwater-fed paleolakes contem­ porarily occurring within a few hundreds of kilometers wide region upslope from the

242

Lakes on Mars

S2 ocean boundaries and the crustal dichotomy (Figures 8.5b and 8.8a). Particularly, the paleolakes would have been almost entirely concentrated in the gradational zone of Arabia Terra, a province where a concentration of craters with extensive exposures of eroded layered sedimentary deposits (e.g., Malin and Edgett, 2000) and other distinguishable spectral and elemental properties (including an elevated hydrogen content) have been reported and interpreted to be the results of a past volatile-rich history (e.g., Dohm et al., 2007). Both the equipotential surfaces S1 and S2 are also generally consistent with the distribution and terminations of the Martian valley networks (Figure 8.9) and open

MOLA Elevation (meters)

a

8000

LEGEND 7000

Deltas Noachian valleys 6000

5000

4000

3000

2000

1000

0 –1000 –2000 S1-S2 –3000 –4000 –5000 Terra sirenum Valles marineris –6000

–180 –150

–100 –50 0 50 Longitude (degrees) 18,000 16,000

Aeoli/Nephentes mensae 100 150 180

b

LEGEND Deltas Amazonian valleys Hesperian valleys

14,000 MOLA Elevation (meters)

S1-S2

12,000 10,000 8000 6000 4000 2000 0 –2000 S1-S2

–4000 –6000

Terra sirenum

–180

–150

Valles marineris –100 –50 0 50 Longitude (degrees)

S1-S2 Aeoli/Nephentes mensae 100 150 180

Figure 8.9 (a) Elevation values of Noachian valley termini and of deltas as a function of longitude. The dashed black line shows the S1–S2 boundaries: a few Noachian valleys are found below these levels. (b) Hesperian and Amazonian valley termini and delta elevations as a function of longitude. Valleys that occur along the eastern walls of Valles Marineris and Hellas Basin show more widely distributed terminations below S1–S2.

Deltas and valley networks on Mars: implications for a global hydrosphere

243

basins (see Figure 8.2a in Fassett and Head, 2008) at a global scale: (i) the majority of the valleys terminate at higher elevation with respect to the inferred levels, (ii) a particularly good match between S1 and valley terminations is found at Terra Sirenum and Aeolis/Nepenthes Mensae regions, (iii) a small amount of Noachian valleys appears located at lower elevations, and finally (iv) Hesperian (∼3.7–3.0 Ga) and Amazonian (younger than 3 Ga) valleys show terminations below the S1 level and mostly occur along the same longitude intervals, corresponding to the eastern walls of Valles Marineris and Hellas Basin (Figure 8.9). The above line of evidence is consistent with a progressive retreat of the ocean after its maximum extension during the Noachian. Volumetric analysis of the inferred levels indicates that the possible ocean volumes could have oscillated from 0.9 to 1.4 � 108 km3 (the total volume of oceans on Earth is about 1.4 � 109 km3). The minimum estimate has been obtained by considering water at the lowest reconstructed elevation only within the northern lowlands, whereas the maximum value reflects the maximum inferred ocean level in the northern plains and standing bodies of water in other low-lying basins like Hellas and Argyre, as well as paleolakes ponding at the same base level across the planet. Approximately onethird (32–37%) of the planet surface should have been covered by water ponding and saturating the crust in hydrostatic equilibrium (equivalent to a 480–630 m deep global ocean). All these estimates are compatible with the total amount of water predicted by Clifford and Parker (2001) for the Noachian. Finally, given the lack of a statistically significant number of fluvial deposits below the inferred levels, our analysis cannot give clues regarding any progressively retreating ocean levels. On the other hand, the distribution of the deltas is not consistent with the occurrence of higher shorelines, such as the “Meridiani shoreline” suggested as the highest possible boundary for a Martian ocean (Clifford and Parker, 2001).

8.5

Summary

During the last decade of Mars exploration, the increase of resolution and coverage of remote sensing datasets has shed new light on the global hydrology of the planet: (i) a new global map of valley networks (Hynek et al., 2009) shows approximately four times more valleys as previously identified based on Viking imagery, and (ii) global surveys of deltaic deposits (Cabrol and Grin, 1999; Di Achille et al., 2008; Irwin et al., 2005) reveal a significant number of candidates, suggesting that they were an impor­ tant geological component during the early history of the planet. Assuming all the latter deposits were formed as deltas, they can be used, along with the distribution of Martian valleys and global MOLA topography, to test the occurrence of a global hydrosphere on Mars. Such analysis delineates two main planet-wide equipotential surfaces that encompass complete topographic enclosures within and along the mar­ gins of the northern lowlands (Figures 8.5b, c and 8.8). The inferred levels could represent two paleoshorelines of a past ocean covering the northern hemisphere of Mars during its early history. These boundaries are generally consistent with the

244

Lakes on Mars

“Arabia shoreline” previously suggested from geomorphologic and topographic observations and also with the global distribution and age of valley networks (Figure 8.9). Additionally, the reconstructed ocean boundaries are in agreement with theoretical predictions from hydrologic and thermophysical properties of Mars and with the elevation values that would minimize the deviation of the “Arabia shoreline” if it was caused by motions related to true polar wander. These results support the theory of an ancient ocean of Mars lending further credence to the interpretation of Martian fan-shaped terminal deposits as deltaic bodies. Given the key assumptions of this analysis present topography of Mars similar to that at the time of the occurrence of the putative ocean (Phillips et al., 2001), and formation of deltas during the same epoch of fluvial activity (Irwin et al., 2005), the inferred boundaries likely reflect the maximum extension of a global hydrosphere near the Noachian–Hesperian boundary. This hypothesis is also supported by the overall age and distribution of ancient valley networks, which were mostly formed by precipitation and surface runoff and are located upslope from the inferred shorelines (Figure 8.9a, b) and by the thermo-hydraulic predictions for the potential volume of Noachian water. The reconstructed scenario implies that climatic conditions allowed the occurrence of a hydrosphere presenting similarities with that of Earth, and integrating valleys, deltas, lakes, an ocean encircling the planet, and possibly a number of suitable niches for life to exist.

Acknowledgments Special thanks to Monica Hoke and Michael Beach for their contributions to the drafting of the valley networks database. We appreciate the thoughtful reviews by C. I. Fassett (Brown University) and an anonymous reviewer. This work was funded by NASA Mars Data Analysis Program Grant #NNX06AE08G.

References Baker, V.R., 2001. Water and the Martian landscape. Nature 412, 228–236.

Baker, V.R., Strom, R.G., Gulick, V.C., Kargel, J.S., Komatsu, G., 1991. Ancient oceans, ice

sheets and the hydrological cycle on Mars. Nature 352, 589–594. Benn, D., Evans, D., 2003. Glaciers and Glaciation. Oxford University Press, New York. Blair, T.C., McPherson, J.G., 1994a. Alluvial fans and their natural distinction from rivers based on morphology, hydraulic processes, sedimentary processes and facies assemblages. J. Sedim. Res., A 64 (3), 450–489. Blair, T.C., McPherson, J.G. 1994b. Alluvial fan processes and forms. In: A.D. Abrahams and A.J. Parsons (Eds.), Geomorphology of Desert Environments, Chapman and Hall, London, pp. 354–402. Cabrol, N.A., Grin, E.A., 1999. Distribution, classification and ages of Martian impact crater lakes. Icarus 14 (2), 160–172. Cabrol, N.A., Grin, E.A., 2001. The evolution of lacustrine environments on early Mars: Is Mars only hydrologically dormant? Icarus 149, 291–328.

Deltas and valley networks on Mars: implications for a global hydrosphere

245

Cabrol, N.A., Grin, E.A., 2005. Ancient and recent lakes on Mars. In: Tetsuya T. (Ed.), Water on Mars and Life. Springer, Heidelberg, 235–259. Cabrol, N.A., Grin, E.A., Dawidowicz, G., 1996. Ma’adim Vallis revisited through new topographic data: evidence for an ancient intravalley lake. Icarus 123 (2), 269–283. Cabrol, N.A., Grin, E.A., Dawidowicz, G., 1997. A model of outflow generation by hydro­ thermal underpressure drainage in volcano-tectonic environment, Shalbatana Vallis (Mars). Icarus 125 (2), 455–464, doi:10.1006/icar.1996.5625. Cabrol, N.A., Wynn-Williams, D.D., Crawford, D.A., Grin, E.A., 2001. Recent aqueous environments in impact crater lakes on Mars: an astrobiological perspective. Icarus 154, 98–112. Carr, M.H., 1995. The Martian drainage system and the origin of valley networks and fretted channels. J. Geophys. Res. 100, 7479–7507. Carr, M.H., 1996. Water on Mars. Oxford University Press, New York. Carr, M.H., Chuang, F.C., 1997. Martian drainage densities. J. Geophys. Res. 102 (E4), 9145–9152. Carr, M.H., Head, J.W.I.I.I., 2003. Oceans on Mars: an assessment of the observational evidence and possible fate. J. Geophys. Res. 108 (E5), 5042. Clifford, S.M., Parker, T.J., 2001. The evolution of the Martian hydrosphere: implications for the fate of a primordial ocean and the current state of the northern plains. Icarus 154, 40–79. Craddock, R.A., Howard, A.D., 2002. The case for rainfall on a warm, wet early Mars. J. Geophys. Res. 107, 5111–5147, doi:10.1029/2001JE001505. De Hon, R.A., 1992. Martian lake basins and lacustrine plains. Earth Moon Planets 56, 95–122. Di Achille, G., Hynek, B.M., Searls, M.L., 2009. Positive identification of lake strandlines in Shalbatana Vallis, Mars. Geophys. Res. Lett. 36. L14201, doi:10.1029/2009GL038854. Di Achille, G., Komatsu, G. 2008. Using HRSC and HiRISE for the study of Martian deposi­ tional environments: an example from Ophir Planum. In: Proc. Lunar Planet. Sci. Conf. 39th, Abstract #1608. Di Achille, G., Marinangeli, L., Ori, G.G., Hauber, E., Gwinner, K., Reiss, D., et al., 2006a. Geological evolution of the Tyras Vallis paleolacustrine system, Mars. J. Geophys. Res. 111, 19, E04003, doi:10.1029/2005JE002561. Di Achille, G., Ori, G.G. 2008. Complex intermontaine glacial systems in Arabia Terra, Mars: Evidence for an Amazonian proglacial lake with associated glacilacustrine deposits, In: Proc. Lunar Planet. Sci. Conf. 39th, Abstract #2096. Di Achille, G., Ori, G.G., Hynek, B.M., Hauber, E., 2008. Global distribution of putative Martian deltas in the light of HRSC and HiRISE instruments: open issues and hydrological inferences. In: R.A. Craddock (Ed.), Second Workshop on Mars Valley Networks, Smith­ sonian Institution, Washington, DC, pp. 16–19. Di Achille, G., Ori, G.G., Reiss, D., 2007. Evidence for late Hesperian lacustrine activity in Shalbatana Vallis, Mars. J. Geophys. Res. 112, E07007, doi:10.1029/2006JE002858. Di Achille, G., Ori, G.G., Reiss, D., Hauber, E., Gwinner, K., Michael, G., et al., 2006b. A steep fan at Coprates Catena, Valles Marineris, Mars, as seen by HRSC data. Geophys. Res. Lett. 33, L07204, doi:10.1029/2005GL025435. Dohm, et al.., 2007. Possible ancient giant basin and related water enrichment in the Arabia Terra province, Mars. Icarus 190, 74–92. Dohm, J.M., et al., 2009. GRS evidence and the possibility of paleooceans on Mars, Planet. Space Sci., doi:10.1016/j.pss.2008.10.008. Ehlmann, B.L., Mustard, J.F., Fassett, C.I., Schon, S.C., Head, J.W. III, Des Marais, D.J., et al., 2008. Clay minerals in delta deposits and organic preservation potential on Mars. Nat. Geosci. 1, 355–358.

246

Lakes on Mars

Fassett, C.I., Head, J.W. III, 2005. Fluvial sedimentary deposits on Mars: ancient deltas in a crater lake in the Nili Fossae region. Geophys. Res. Lett. 32, L14201, doi:10.1029/ 2005GL023456. Fassett, C.I., Head, J.W., 2007. Valley formation on Martian volcanoes in the Hesperian: evidence for melting of summit snowpack, Caldera lake formation, drainage and erosion on Ceraunius Tholus. Icarus 189, 118–135. Fassett, C.I., Head, J.W., 2008. Valley network-fed, open-basin lakes on Mars: distribution and implications for Noachian surface and subsurface hydrology. Icarus 198, 37–56. Forsythe, R.D., Blackwelder, C.R., 1998. Closed drainage crater basins of the Martian highlands: constraints on the early Martian hydrologic cycle. J. Geophys. Res. 103 (E13), 31421–31431. Gilbert, G.K., 1885. The topographic features of lake shores. U.S. Geol. Surv. Annu. Rep., 5, 75–123. Grant, J.A., Schultz, P.H., 1993. Degradation of selected terrestrial and Martian impact craters. J. Geophys. Res. 98, 11025–11042. Grant, J.A., Parker, T.J., 2002. Drainage evolution in the Margaritifer Sinus region, Mars. J. Geophys. Res., 107(E9), 5066, doi:10.1029/2001JE001678. Greeley, R., Guest, J.E., 1987. Geologic map of the eastern equatorial region of Mars, U.S. Geol. Surv. Misc. Invest. Ser. Map, I-1802–B. Gulick, V.C., 1998. Magmatic intrusions and a hydrothermal origin for fluvial valleys on Mars. J. Geophys. Res. 103, 19365–19388. Gulick, V.C., Baker, V.R., 1990. Origin and evolution of valleys on Martian volcanoes. J. Geophys. Res. 95 (B9), 14325–14344. Harrison, K.P., Grimm, R.E., 2005. Groundwater-controlled valley networks and the decline of surface runoff on early Mars. J. Geophys. Res. 110, E12S16, doi:10.1029/2005JE002455. Hauber, E., Gwinner, K., Kleinhans, M., Reiss, D., di Achille, G., Ori, G.-G., et al., 2009. Sedimentary deposits in Xanthe Terra: implications for the ancient climate on Mars. Planetary Space Sci., in press, 944–957, doi:10.1016/j.pss.2008.06.009. Head, J.W., Hiesinger, H., Ivanov, M.A., Krelavshky, M.A., Pratt, S., Thomson, B.J., 1999. Possible ancient oceans on Mars: evidence from Mars orbiter laser altimeter data. Science 286, 2134–2137. Howard, A.D., 2007. Simulating the development of Martian highland landscapes through the interaction of impact cratering, fluvial erosion, and variable hydrologic forcing. Geomor­ phology 91 (3–4), 332–364, DOI: 10.1016/j.geomorph.2007.04.017. Howard, A.D., Moore, J.M. 2004. Changing style of erosion during the Noachian-Hesperian transition and a possible climatic optimum. Proc. Lunar Planet. Sci. Conf. 35th, Abstract 1192. Hynek, B.M., Beach, M., Hoke, M.R.T., 2010. Updated global map of Martian valley networks and implications for hydrologic processes. J. Geophys. Res., in review, doi:10.1029/ 2009JE003548. Hynek, B.M., Phillips, R.J., 2001. Evidence for extensive denudation of the Martian highlands. Geology 29, 407–410. Hynek, B.M., Phillips, R.J., 2003. New data reveal mature, integrated drainage systems on Mars indicative of past precipitation. Geology 31, 757–760. Irwin, R.P. III, Howard, A.D., 2002. Drainage basin evolution in Noachian Terra Cimmeria, Mars. J. Geophys. Res. 107 (E7), 5056, doi:10.1029/2001JE001818. Irwin, R.P. III, Howard, A.D., Craddock, R.A., Moore, J.M., 2005. An intense terminal epoch of widespread fluvial activity on early Mars: 2. Increased runoff and paleolake development. J. Geophys. Res. 110, E12S15, doi:10.1029/2005JE002460. Irwin, R.P. III, Watters, T.R., Howard, A.D., Zimbelman, J.R., 2004. Sedimentary resurfacing and fretted terrain development along the crustal dichotomy boundary, Aeolis Mensae, Mars. J. Geophys. Res. 109, E09011, doi:10.1029/2004JE002248.

Deltas and valley networks on Mars: implications for a global hydrosphere

247

Kleinhans, M.G., 2005. Flow discharge and sediment transport models for estimating a mini­ mum time scale of hydrological activity and channel and delta formation on Mars. J. Geophys. Res. 110, 23, E12003, doi: 10.1029/2005JE002521. Kraal, E.R., van Dijk, M., Postma, G., Kleinhans, M.G., 2008. Martian stepped-delta formation by rapid water release. Nature 451, 973–976, doi:10.1038/nature06615. Leverington, D.W., 2006. Volcanic processes as alternative mechanisms of landform develop­ ment at a candidate crater-lake site near Tyrrhena Patera, Mars. J. Geophys. Res. 111, E11002, doi:10.1029/2004JE002382. Leverington, D.W., Maxwell, T.A., 2004. An igneous origin for features of a candidate craterlake system in western Memnonia, Mars. J. Geophys. Res. 109, E06006, doi:10.1029/ 2004JE002237. Malin, M.C., et al., 1992. The Mars observer camera. J. Geophys. Res. 97 (E5), 7699–7718. Malin, M.C., Edgett, K.S., 2000. Sedimentary rocks on Mars. Science 290 (5498), 1927–1937. Malin, M.C., Edgett, K.S., 2003. Evidence for persistent flow and aqueous sedimentation on Mars. Science 302, 1931–1934, doi:10.1126/science.1090544. Mangold, N., Ansan, V., 2005. Detailed study of an hydrological system of valleys, delta and lakes in the Southwest Thaumasia region, Mars. Icarus 180, 75–87. Masursky, H., 1973. An overview of geologic results from Mariner 9. J. Geophys. Res. 78, 4009–4038. McEwen, A.S., et al., 2007. MRO’s High Resolution Imaging Science Experiment (HiRISE). J. Geophys. Res. 112, E05S02. McGill, G.E. 2002. Geologic map transecting the highland/lowland boundary zone, Arabia Terra, Mars: Quadrangles 30332, 35332, 40332, and 45332, scale 1:1M, U.S. Geol. Surv. Geol. Invest. Ser., Map I-2746. Milliman, J.D., Meade, R.H., 1983. World-wide delivery of river sediment to the oceans. J. Geol. 91, 1–21. Milton, D.J., 1973. Water and processes of degradation in the Martian landscape. J. Geophys. Res. 78, 4037–4047. Moore, J.M., Howard, A.D., 2005. Large alluvial fans on Mars. J. Geophys. Res. 110, E04005, doi:10.1029/2004JE002352. Moore, J.M., Howard, A.D., Dietrich, W.E., Schenk, P.M., 2003. Martian layered fluvial deposits: implications for Noachian climate scenarios. Geophys. Res. Lett. 30, 2292–2295, doi: 10.1029/2003GRL019002. Moore, J.G., Phillips, R.L., Grigg, R.W., Peterson, D.W., Swanson, D.A., 1973. Flow of lava into the sea, 1969–1971, Kilauea Volcano, Hawaii. Geol. Soc. Am. Bull. 84 (2), 537–546. Murchie, S., et al., 2007. Compact Reconnaissance Imaging Spectrometer for Mars (CRISM) on Mars Reconnaissance Orbiter (MRO). J. Geophys. Res. 112, E05S03, doi:10.1029/ 2006JE002682. Neukum, G., Jaumann, R., Co-Investigator, H.R.S.C., Experiment, T. 2004. HRSC—The High Resolution Stereo Camera of Mars Express. Eur. Space Agency Spec. Publ., SP-1240, 17–35. Newsom, H., Brittelle, G., Hibbitts, C., Crossey, L., Kudo, A., 1996. Impact crater lakes on Mars. J. Geophys. Res. 101 (E6), 14951–14955. Ori, G.G., Marinangeli, L., Baliva, A., 2000a. Terraces and Gilbert-type deltas in crater lakes in Ismenius Lacus and Memnonia (Mars). J. Geophys. Res. 105 (E7), 17629–17642. Ori, G.G., Marinangeli, L., Komatsu, G., 2000b. Martian lacustrine environments and their geological constraints on drilling operations for exobiological research. Planet. Space Sci. 48, 1027–1034. Parker, T.J., Gorsline, D.S., Saunders, R.S., Pieri, D.C., Schneeberger, D.M., 1993. Coastal geomorphology of the Martian northern plains. J. Geophys. Res. 98, 11061–11078.

248

Lakes on Mars

Parker, T.J., Saunders, R.S., Schneeberger, D.M., 1989. Transitional morphology in the west Deuteronilus Mensae region of Mars: implications for modification of the lowland/upland boundary. Icarus 82, 111–145. Perron, J.T., Mitrovica, J.X., Manga, M., Matsuyama, I., Richards, M.A., 2007. Evidence for an ancient Martian ocean in the topography of deformed shorelines. Nature 447, 840–843. Phillips, R.J., et al., 2001. Ancient geodynamics and global-scale hydrology on Mars. Science 291, 2587–2591. Pieri, D.C., 1980. Martian valleys: morphology, distribution, age, and origin. Science 210, 895–897. Pondrelli, M., Baliva, A., Di Lorenzo, S., Marinangeli, L., Rossi, A.P., 2005. Complex evolu­ tion of paleolacustrine systems on Mars: an example from the Holden crater. J. Geophys. Res. 110, E04016, doi:10.1029/2004JE002335. Pondrelli, M., Rossi, A.P., Marinangeli, L., Hauber, E., Gwinner, K., Baliva, A., et al., 2008. Evolution and depositional environments of the Eberswalde fan delta, Mars. Icarus 197 (2), 429–451, doi: 10.1016/j.icarus.2008.05.018. Powell, R.D., 1990. Glacimarine processes at grounding-line fans and their growth to icecontact deltas. In: J.A. Dowdeswell and J.D. Scourse Eds., Glacimarine Environments: Processes and Products. vol. 53. Geological Society, London, Specical Publication, pp. 53–73. Scott, D.H., Tanaka, K.L. 1986. Geologic map of the western equatorial region of Mars. U.S. Geol. Surv. Misc. Invest. Ser. Map, I-1802–A. Segura, T.L., Toon, O.B., Colaprete, A., Zahnle, K., 2002. Environmental effects of large impacts on Mars. Science 298, 1977–1980, doi:10.1126/science.1073586. Sharp, R.P., Malin, M.C., 1975. Channels on Mars. Geol. Soc. Am. Bull. 86, 593–609. Smith, D.E., et al., 2001. Mars orbiter laser altimeter: experiment summary after the first year of global mapping of Mars. J. Geophys. Res. 106 (E10), 23689–23722. Squyres, S.W., Kasting, J.F., 1994. Early Mars: how warm and how wet? Science 265, 744–749. Stanley, D.G., Warne, G.W., 1994. Worldwide initiation of Holocene marine deltas by decelera­ tion of sea-level rise. Science 265, 228–231. Syvitski, J.P.M., Saito, Y., 2007. Morphodynamics of deltas under the influence of humans. Glob. Planet. Changes 57, 261–282. Weitz, C.M., Irwin, R.P., Chuang, F.C., Bourke, M.C., Crown, D.A., 2006. Formation of a terraced fan deposit in Coprates Catena, Mars. Icarus 184, pp. 436–451. Williams, R.M.E., Zimbelman, J.R., Johnston, A.K., 2006. Aspects of alluvial fan shape indicative of formation process: a case study in southwestern California with application to Mojave Crater fans on Mars. Geophys. Res. Lett. 33, L10201, doi:10.1029/ 2005GL025618. Williams, R. M. E., Malin, M.C. 2008. Sub-kilometer fans in Mojave crater, Mars. Icarus, 198, 365–383. doi:10.1016/j.icarus.2008.07.013.

9 The northern plains: A Martian oceanic basin?

Timothy J. Parker†, John A. Grant* and Brenda J. Franklin† †

Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, USA * Center for Earth and Planetary Studies, National Air and Space Museum, Smithsonian Institution, Washington, DC, USA

9.1

Introduction

The question of whether oceans were present on ancient Mars has been addressed in peer-reviewed publications beginning during the mid-1980s, based primarily on orbiter images of the northern plains. This work was based on data acquired by spacecraft preceding the MGS mission of the late 1990s, chiefly Viking Orbiter (Baker et al., 1991; Lucchitta et al., 1986; Clifford and Parker, 2001; Edgett and Parker, 1997; Jons, 1990; Parker et al., 1989, 1993; Parker and Currey, 2001; Scott et al., 1992, 1995). These publications indicated specific locations where proposed shorelines could be evaluated with high-resolution cameras onboard MGS, ODY, MEx, and MRO. Multiple, “nested” plains unit contacts, terraces, curvilinear ridges, and other features were identified in several locations around the northern plains. After con­ sidering the alternate genetic processes that had been suggested for their origin, these features were interpreted as most likely related to highstands of an ancient northern plains ocean by Parker et al. (1989, 1993). Two of the contacts, one associated with the “gradational boundary” (Rossbacher, 1985) and a second one plainward of it and described as an “interior plains boundary,” were traced to almost complete closure around the northern plains (Contacts 1 and 2, in Parker et al., 1989). Several others were traced laterally for hundreds of kilometers within a few key regions that appear to be sub-parallel to the two global contacts (either between or to either side of them) and a number of others were traced only locally in the highest resolution Viking Orbiter mosaics. The large number of these features necessitated a mapping and naming scheme. Clifford and Parker (2001) suggested a scheme similar to that used by terrestrial geomorphologists for shorelines in Pleistocene Lake Bonneville basin and used local classical albedo feature names from where the proposed shoreline feature was first identified. Thus, “Contact 1” was renamed “the Arabia shoreline” and “Contact 2” was renamed “the Deuteronilus shoreline.” However, the decision to apply the term “shoreline” to these features was unfortu­ nate, as shoreline is a decidedly genetic term. Selection of a nongenetic but still Lakes on Mars. DOI: 10.1016/B978-0-444-52854-4.00009-X © 2010 Published by Elsevier B.V.

250

Lakes on Mars

precise and descriptive term for these features is not straightforward, because they are not typically simple demarcations separating mappable geologic units and so are not strictly “contacts,” but can be expressed as a subtle association of a number of landform types with or without distinct boundaries between them. Referring to them as “contacts” implies that the materials or textures (really geomorphic units) to either side are distinct from one another, which is not always the case. While the larger of the Martian features do show distinct textural differences between surfaces on either side (such as the contact separating the well-known mottled plains from the thumbprint terrain in eastern Acidalia Planitia), the less-extensive or smaller-scale features, for the most part, do not. In addition, even the best expressed of these features displays numerous breaks in the available global mosaics (initially, Viking Orbiter, but now including THEMIS daytime IR). Terrestrial paleolake shorelines used as analogs for the earlier work can often be expressed as nothing more than narrow terraces cut into preexisting lithologic units, such as an outcrop of basalt, or geomorphic units, such as an alluvial fan, and are often mapped as colored lines or symbols within these units. The lithology and morphology of the preexisting surface is not necessarily altered or obscured, although at very high resolution, modification of the surface during eustatic changes and burial by sediment deposition can usually be identified. This may require aerial photographic image scales or even field observations at ground level to verify their coastal origin. Lacustrine materials or landforms can be recognized at smaller scales, provided the lake was present long enough or the rates of erosion and deposition were high enough to produce changes visible and recognizable from orbit. In the interest of objectivity then, this discussion will refer to the Martian features as “levels.” Since MGS began mapping Mars in the late 1990s, other investigators began offering independent tests of the ocean hypothesis, either by directly addressing the question based on results from MGS and later missions (Clifford and Parker, 2001; Head et al., 1998, 1999; Malin and Edgett, 1999) or by considering the hypothesis as one possible explanation for the discovery of coarse crystalline hematite in Meridiani Planum (Christensen et al., 2000). Early work by members of the MOLA team (Head et al., 1998, 1999) would seem to support the ocean hypothesis, at least for the Deuteronilus Level, which defines an equipotential surface reasonably well, but with two notable exceptions. However, they found that the Arabia Level exhibits a much poorer fit to the MOLA topography, deviating from the horizontal by more than a kilometer at some localities. Head et al., (1998, 1999) also identified broad topo­ graphic terraces in Utopia that have no obvious geomorphic expression in Viking Orbiter images, and thus constitute an MGS discovery, which they interpret as additional shorelines below the Deuteronilus Level. Malin and Edgett (1999) targeted the MOC to image sites that had been proposed as shorelines prior to MGS and were unable to equate their observations to the hypothesis that a former ocean had produced them. The work by Di Achille and Hynek using valley networks and deltaic deposit elevation detailed in Chapter 8 is the most recent test of the ocean hypothesis and appears to be consistent with the presence of oceans or large seas in the northern hemisphere.

The northern plains: A Martian oceanic basin?

251

MOC on MGS gave us the first very high-resolution images ever taken of Mars— truly comparable to aerial photos in resolving fine detail (Malin and Edgett, 2001). MOC imaged only a few percent of Mars’ surface however, so regional coverage of interesting targets often required acquisition of mosaics of overlapping images. Prior to the arrival of ODY, Viking Orbiter images still provided the best regional context maps for the MOC Narrow Angle images. ODY arrived at Mars in late 2001 and began its primary mission in early 2002. Of interest to geomorphology and geology investigations, Odyssey’s camera has two imaging modes—thermal infrared (100 m/pixel, taken during the day and night por­ tions of the orbit) and visible (VIS, 18 m/pixel daytime)—(Christensen et al., 2004). The daytime IR images have been compiled by the Odyssey team into a global mosaic with nearly complete coverage of the planet. The latest version of this map is being compiled at the full resolution of the daytime IR images, which is greater than twice the average resolution provided by the previous global Viking Orbiter mosaics (still being finalized at the time of this writing). Any remaining gaps in the daytime IR mosaic can still be filled with Viking mosaic data, where needed. THEMIS VIS coverage, while not complete for Mars, is nonetheless extensive enough to enable compilation of mosaics of regions of interest around the northern plains at 18 m/pixel. Some of these areas are not yet covered by higher-resolution images (MOC, CTX, or HiRISE); thus, the THEMIS VIS images provide moder­ ately high-resolution coverage in these areas. A benefit to be considered of THEMIS VIS over the higher-resolution CTX and HiRISE images is that the later afternoon orbit of Odyssey enhances photometric shading and often highlights subtle topographic features such as putative shorelines and plains textures asso­ ciated with them. MOLA topography is orders of magnitude higher in resolution than the Mariner 9 and Viking topography that was used for the early work on the ocean hypothesis. All the MOLA data have been compiled into global gridded image products that are available from the PDS. In addition, the individual data points can be quickly accessed and map projected with the gridded product to determine the quality of the elevation value averaged for a particular grid cell or to make higher resolution measurements of elevation at points of interest. It is now possible to compare shore­ line positions with topography to look for elevation changes that might indicate systematic variations due to “neotectonic” adjustments in the topography or other tectonic changes to support or contradict the shoreline interpretation for the origin of these features. Significant improvements over Viking data have been realized from the Orbiter missions since the 1990s and the acquisition of new data continues. The image data sets provide high-resolution views that when mosaiced and combined with the MOLA gridded topography, enable the compilation of regional and global maps of the proposed shorelines. To produce these maps, GIS (Arcmap) and image processing software (ISIS3, Global Mapper, and Photoshop) are being used to generate an imagebased map of the planet’s northern hemisphere (incorporating mainly CTX images) that is georeferenced to the MOLA gridded topography basemap provided by the USGS, Flagstaff (Figure 9.1).

252

Lakes on Mars

Figure 9.1 Shaded relief topographic map of the northern hemisphere of Mars. Grayscale overlays of primary study areas along the northern plains margin. GIS polar.

The follow the water theme underscores all our current and foreseeable future plans for spacecraft missions to Mars, both orbiters and landers. As a result, it makes sense to take stock of the image data we now have available from past and current missions and assess what these data say about the published predictions of an ancient ocean on Mars. This chapter describes recent and ongoing work to utilize the new image and topography data to revisit sites along the lowland/upland boundary where landforms were identified in Viking Orbiter images (Parker et al., 1989, 1993 Parker (1994), Parker and Currey (2001), and Clifford and Parker (2001)) and interpreted to be ocean shorelines. These localities now have extensive high-resolution image coverage from MOC, THEMIS VIS, HRSC, CTX, and HiRISE. Here, we will focus on the west Deuteronilus Mensae region that was first described by Parker et al. (1989) as well as on part of the Arabia Level that was expressed as a prominent albedo boundary in low-resolution Viking Orbiter data that have since been shown to exhibit terraces and other aspects reminiscent of coastal landforms in the recent very high-resolution images (Figure 9.2).

9.2 Coastlines and topography Standing water forms an equipotential surface that intersects topography at a fixed elevation around the margin of a depression. Simply stated, modern lakes and ocean shorelines on Earth are level and possess attributes modulated by storms and tides

The northern plains: A Martian oceanic basin? 20°0'0''N

0°0'0''

253 20°0'0''E

50°0'0''N

40°0'0''N

40°0'0''N

30°0'0''N

0°0'0''

Figure 9.2 CTX images georeferenced to East Acidalia Planitia and west Deuteronilus Mensae. Mapped levels indicated in color over grayscale images. From topography highest (oldest?) to lowest (youngest?): green—Arabia Level; red—Ismenius Level; blue—Deuteronilus Level; and cyan—Acidalia Level. Subset of GIS polar stereographic map of northern hemisphere (USGS).

(although Mars would have weaker solar and no lunar tides). In contrast, abandoned shorelines on Earth are seldom level, though they often indicate a planar surface that has been tilted, faulted, or warped due to structural changes, isostatic rebound, or loading. All these effects have been identified in terrestrial paleolakes (e.g., Pleisto­ cene Lake Bonneville, Utah, Currey, 1980). Ancient shorelines on Mars should not be precisely level for the same reasons, although it should be possible to recognize systematic changes in elevation along the shoreline due to these effects. Coastal landforms are often subtle or are difficult to distinguish from other features, such as stratigraphic terraces and fault scarps, and often extend over vast areas, requiring both high-resolution imaging and regional coverage. For example, on Earth, Landsat MSS data (� 80 m/pixel) reveal only the very largest coastal forms in Lake Bonneville, but provide good regional coverage of the basin. Aerial photographs do much better with the detection and recognition of the smaller features, but even meter-scale aerial photos cannot be relied on for “closure” of a mapped shoreline, for at least two reasons: first, regional coverage is lacking, so placement into context requires large mosaics, and second, post-shoreline degradation by geologic processes subsequent to the formation of the feature can subdue it through burial or erosion. At very high resolution, it is important to consider subsequent, possibly unrelated

254

Lakes on Mars

processes that might dominate the character of a surface and understand their effects on preservation of preexisting landforms.

9.3 Proposed “shorelines” and related landforms in the west Deuteronilus Mensae/east Acidalia Region Parker et al. (1989) focused on the west Deuteronilus Mensae region for two reasons: first, the features of interest had been identified on the sloping surface of the southern highlands where the topography dipped gradually downward to the north into the northern lowland plains west of Deuteronilus Mensae. At typical Viking Orbiter image scales, the fretted terrain of west Deuteronilus Mensae appears incised into the sloping margin of the highlands, resulting in scattered mesas and plateaus that seem to record the former sloping surface. However, a rare, very high-resolution swath of Viking images (7–10 m/pixel) of the wall of one of the westernmost fretted valleys of the Mensae (north Mamers Valles) shows that the prominent boundary features visible on the sloping surface to the west can be traced into the fretted terrain, suggesting either erosional exposure of stratigraphic contacts by the fretted terrain or emplacement of the contacts after it. Parker et al. (1989) derived crater counts for the major geomorphic units associated with the three most prominent levels and found them to be progressively younger toward the plains’ interior and lower elevations. The favored interpretation was that these materials onlap the preexisting sloping margin and fretted terrain, such that the emplacement of each subsequent plains unit was less extensive than the previous one. The earlier publications (Parker, 1994; Parker and Currey, 2001; Parker et al., 1989, 1993) described the general geomorphology of west Deuteronilus Mensae separately from the detailed associations of local features in northern Mamers Valles because of the very limited high-resolution coverage by the Viking Orbiters. Today, however, there is a wealth of high-resolution image data available in this area, so that we can now describe the levels and their associated landforms in order of their appearance, from topographically highest level (the Arabia Level) to lower levels (e.g., the Deuteronilus Level), and characterize the landform associations visible in high-resolution images that are available across the region. Parker et al. (1989) identified seven levels in the Mamers Valles region (Figure 9.3). Based on the appearance of these features at Viking image scales and their inferred topographic and age relationships, they were interpreted to be most comparable to wave-eroded shorelines in terrestrial paleolakes (lacking terrestrial-scale oceanic tides). With MOLA topography, these levels were found to indeed approximate level surfaces, or planar surfaces that appear to have been tilted (locally, to the west). However, the higher-resolution image data, beginning with MOC, show boundary morphology that does not appear to support a shoreline interpretation involving wave erosion and longshore sediment transport, an argument first made by Malin and Edgett (1999). At a very large image scale, many of the best preserved of these features at the lower elevations instead exhibit lobate flow-front morphology suggestive of low-viscosity lava or debris flows advancing up the sloping highland

The northern plains: A Martian oceanic basin?

255

Figure 9.3 Terraces and plains contacts identified based on Viking Orbiter images and interpreted to be shorelines (adapted from Parker et al., 1989). Labeled with type locality names used in this text.

margin from the northern plains, although the measurable thickness of this material is often negligible at the scale of the gridded MOLA topography data in this region. This morphology suggests the material encroaching onto the sloping margin is still present. The Arabia Level does exhibit morphologies at high resolution that are reminiscent of terrestrial paleolake strandlines, however, although unusual accumulations of some­ times relatively thick material at the Arabia Level are present in a number of areas. All of the levels described and any material deposits associated with them remain elevated with respect to the northern plains interior, even just a few kilometers from the contact defining a particular level, and thus still appear to require withdrawal of vast amounts of some fluid from the northern basin after their emplacement. Finally, while this chapter focuses on the details of the proposed shorelines and landforms associated with them in and around Deuteronilus Mensae, it is important to note that many of these features have been identified elsewhere around the Martian dichotomy boundary. There is no “other side” of these features somewhere north of Deuteronilus Mensae outside the study area. If these do indeed prove to be shorelines, they are shorelines of an ocean, not a paleolake or sea.

9.3.1 The Arabia level Across eastern Acidalia and into west Deuteronilus Mensae, the Arabia Level is most easily recognized as a sharp albedo contact separating dark plains surfaces from

256

Lakes on Mars

lighter, rougher surfaces of the highlands to the south. This contrast is evident in all images from the days of Viking through the present day and is most pronounced in the THEMIS daytime and nighttime IR data. In daytime IR images, the surface plainward of the Arabia Level is dark, whereas in the nighttime IR images, it is brighter than the highlands surface to the south. This suggests that it comprises lithified or indurated material that is slow to change in temperature, relative to the highland materials to the south. Depending on local slopes, the Arabia Level can have two morphological expres­ sions. On gentle slopes, it is a thin, relatively dark mantle unit onlapping the highland margin. On steep slopes, it shows one or more narrow terraces that appear approxi­ mately level over great distances around elevated topography, such as fretted valley walls and around crater rims and ejecta deposits. Locally, these terraces are found within several tens of meters of one another in elevation (Figures 9.4–9.7). In Mamers Valles, the Arabia Level is often expressed as a sharp break in slope below an escarpment, with two or three broad, subdued swales or terraces on the slope between this break and the tops of the debris aprons. These features are reminiscent of terrestrial paleo-lake shore morphology viewed at high resolution (Figure 9.8). In southeast Cydonia, multiple terraces along the inner rims of degraded craters along the Arabia Level are strikingly similar in appearance to strandlines in terrestrial paleolakes.

Figure 9.4 Eroded craters at the Arabia Level at the “Gradational Boundary” between NW Arabia Terra and Cydonia Mensae. THEMIS daytime IR, MOLA topography (contour interval 100 m).

The northern plains: A Martian oceanic basin?

257

Figure 9.5 Eroded craters at Arabia Level (contd.). CTX image. Arrows indicate contact between dark “plains” surface and lighter elevated “highlands” surface, first identified in Viking Orbiter images. This contact appears to lie at an elevation of approximately –3800 m locally and consists of an onlapping of dark material over the brighter material.

Figure 9.6 Closeup of west crater wall. CTX image. Note multiple terraces at base of west wall, defining Arabia Level locally.

258

Lakes on Mars

Figure 9.7 Closeup of crater wall (contd.). HiRISE image. Detail of two broad terraces and one higher, narrow terrace near base of crater wall.

Figure 9.8 Oblique aerial photo of Tule Valley, Utah, showing terraces cut into bedrock and alluvium in Lake Bonneville. Provo Level is the prominent terrace across center of frame.

The northern plains: A Martian oceanic basin?

259

Much of the Arabia Level between the fretted terrains of Deuteronilus Mensae and southeastern Cydonia Mensae was covered only by Viking Orbiter images with resolutions greater than 100 m/pixel. In this region, the Arabia Level is defined by a prominent albedo contact between dark plains surfaces and relatively bright highland surfaces. But even at the low resolution of the Viking images, it is apparent that erosion of the highland margin has occurred at this contact. The primary indication that the Arabia Level is an erosional feature is shown in Figure 9.4, which is an enlargement of part of the area along the Arabia Level from Figure 9.2. MOLA topography confirms the visual impression that four large craters along this contact have been eroded, not just buried, because hundreds of meters of rim material is missing from all four craters and their floors are all at the same elevation. Indeed, the contoured MOLA topography appears to follow the albedo contact exactly across the scene in Figure 9.4. Topography above an elevation of about –3900 m is degraded and that below –3900 m is filled in with dark material, resulting in a �100-km-wide topographic terrace that can be traced laterally for several hundred kilometers at this elevation. In the past few years, enough very high-resolution image coverage has accumu­ lated on this part of the Arabia Level that continuous mosaics can be assembled. Highresolution regional mosaics are essential in establishing the regional continuity of what are often very subtle terraces and plains boundaries at the Arabia Level and determining where the albedo contact identified in Viking images actually falls with respect to these terraces and other topographic features. Figure 9.5 is an enlargement of part of the eroded remnant of rim material between two of these craters, which shows how the Arabia Level was identified around the eroded topography. The highest of several identifiable terraces exhibits a sharp albedo boundary that appears to consist of a thin mantle of dark material onlapping the topography below about –3800 m elevation, locally. Similar terraces can be traced around the western edge of these craters to the wall shown in Figures 9.6 and 9.7, which include 6 m/pixel images from CTX and an HiRISE image at 25 cm/pixel. Figure 9.8 is an oblique aerial view of shorelines in the Lake Bonneville basin. The Provo Shoreline, here representing both erosional and depositional processes by wave activity, is similar in scale to the broader terraces depicted in Figures 9.6 and 9.7. While small valley networks are relatively less abundant in the study area com­ pared to other areas along the highland margin, they are present. In nearly all instances where they are found near the Arabia Level, they lie exclusively on the highlands side of the albedo contact and often terminate abruptly at the contact. At least one large valley network lies plainward of the contact at 42°N, 3.5°E. This valley appears draped by the material comprising the plains and is indistinct at either end, suggesting that it was buried or otherwise subdued when the plains were emplaced, but not completely obscured. In addition, there is at least one modest-scale outflow channel in the region. This is a broad, shallowly incised braided system that alternates from being sharply defined, where it crosses topographic ridges or encounters other obstacles, to being indistinct, where it crosses depressions. This channel first appears in an intercrater plains region west of the crater Focas (34°N, 11°E). It bifurcates southwest of the Semeykin Crater (40°N, 8°E), with one branch terminating at the Arabia Level on

260

Lakes on Mars

the northwest side of the crater (after first breaching its southwest rim and filling the crater) and the other 80 km west of the crater. Landforms and their associations with the Arabia Level, described in Parker et al. (1989) and Parker (1994) using the very high-resolution Viking images of northern Mamers Valles, can now be readily identified throughout the fretted terrain in west Deuteronilus Mensae. The strip of Viking images here had a gap where the Arabia Level intersects the wall of northern Mamers Valles. Parker et al. (1989) inferred that a sharp break in slope near the top of the wall and south of the point of intersection between the Arabia Level and the valley was the continuation of the Arabia Level southward where it was following topography and “veeing” up-valley. As this break was traced southward, the depth at which it appeared below the top of the canyon wall seemed to increase, again suggesting that the break followed the topography as the surrounding terrain climbed to the south. High-resolution post-Viking images verify that the Arabia Level does intersect this valley wall, and several others fretted valleys and mesas to the east, where this break in slope begins (Figure 9.9) until they become lost beneath the ejecta from the Lyot Crater. Detailed measurements of MOLA and HRSC-derived elevations of the Arabia Level at these valley intersections do indeed suggest that it defines an approximately horizontal surface.

Figure 9.9 Point where Arabia Level on plateau surface (albedo contact from upper left to center of frame) intersects fret valley wall at 45.4°N, 16°E. Note that fret valley wall north of (below) this level is rounded, whereas wall south of (topographically higher) this level is more sharply defined. The Arabia Level is inferred to continue south, veeing up-valley as the regional topography rises to the south.

The northern plains: A Martian oceanic basin?

261

Parker et al. (1989) attempted to derive quantitative topographic profiles across northern Mamers Valles in order to convey the impression that the fretted terrain has been inundated by the material that formed the levels, but the fact that it was not completely obscured by them, suggesting partial burial by some material associated with a fluid responsible for the mapped levels, and that fluid has subsequently receded. Viking-era topography was too coarse for making these measurements, so photoclinometry was used to generate five profiles across the valley. Comparing the results of this method with measurements taken from the MOLA and HRSC DEMs reveals that the depths of the fretted valley were consistently underestimated using photoclinometry by 10 or 20% in the northern end of the valley (adjacent to the Deuteronilus Level) and as much as 100% in the south. The profile shapes appear to be correct, however. In Mamers Valles and in west Deuteronilus Mensae as far as the Arabia Level can be traced to the Lyot Crater, debris aprons are only found in association with fretted valley walls and mesa scarps south of where the Arabia Level intersects the walls. This relationship, first described by Parker et al. (1989), appears to hold east of the Lyot Crater as well, though the mesas are smaller and more scattered, so the details are more difficult to ascertain. West of the fretted terrain, the plains interior to these terraces are often polygonally fractured at a scale of a few tens of meters, suggesting a desiccated or frozen sedimentary deposit or a cooled lava lake surface (Figure 9.5). Locally, this surface is on order a few tens of meters lower than the Arabia albedo feature on adjacent slopes. Stepped massifs (Parker, 1994; Parker and Currey, 2001; Parker et al., 1993) are scattered over the plains interior to the Arabia Level. Many of these exhibit prominent terraces or flat-topped aprons around them, particularly in Cydonia Mensae, where they are common. Stepped massifs typically exhibit one prominent terrace that can be attributed to the Deuteronilus Level having been preserved at that location, particu­ larly if it is near enough to make a good correlation based on elevation and morphol­ ogy. A smaller number exhibit steps that can probably be correlated with either the Ismenius Level or the Acidalia Level, but these are not as pronounced as those at the Deuteronilus Level. In northern Mamers Valles, the Arabia Level intersects the fretted valley wall at an elevation of approximately –3600 m. The elevation of the Arabia Level decreases to the west until it is at approximately –3800 m southeast of Cydonia Mensae. There also appears to be a slope to the surface defined by the Arabia Level that is downward to the northwest. Similar boundary morphology identified in the Acidalia Mensae region (46°N, 334°E) is approximately 600 m lower in elevation than at the point of inter­ section at Mamers Valles. Similarly, the prominent break in slope below the scarp in Mamers Valles south of this point (43°N in Mamers Valles) is approximately 300 m higher than at the point of intersection with the fretted valley wall.

9.3.2 The Ismenius level This feature, previously identified as “bench 4” (Parker et al., 1989), is expressed in the very high-resolution Viking Orbiter images of northern Mamers Valles as a local

262

Lakes on Mars

scarp within the plains midway between the Arabia and Deuteronilus Levels. It locally intersects the wall of the fretted valley and vees up-valley, appearing to conform to local topography. It was also identified in high-resolution (�50 to 75 m/pixel) Viking Orbiter images west of Mamers Valles as a band of non-branching fluvial rills that were also traced across mesas in the fretted terrain southwest of the Lyot crater. Parker et al. (1989) compared the rills to swash rills on a beach, morphologically, but at a large scale. This level was simply labeled bench 4 by Parker et al. (1989) because it was only recognized locally near northern Mamers Valles. Today’s extensive highresolution orbiter images show the break in slope to comprise a discontinuous, faint terrace and scarp within the plains unit between the Arabia and Deuteronilus Levels (termed “Intermediate Plains” by Parker et al., 1989). The rills appear to terminate at their downslope ends at about the elevation of this terrace, where both features can be identified (Figure 9.10). The rills are only found immediately above the elevation of this terrace—they never cut terrain at lower elevations, even when they are found in direct proximity to those lower surfaces. Because this terrace and the rills associated with it do appear to define an approximately level surface that can be traced laterally for more than 1000 km, it was given a local geographic name, Ismenius Level. In medium-resolution visible light images, the plains surfaces on either side of the Ismenius Level appear similar, that is, they appear to be smoother and darker than the surfaces above the Arabia Level and below the Deuteronilus Level. In THEMIS daytime and nighttime IR, there is a pronounced contrast between the two surfaces at the Ismenius Level, with the daytime IR brightness suggesting that the higher plains

Figure 9.10 Ismenius Level terrace and rills at 46.4°N, 18.7°E on edge of fretted terrain mesa in west Deuteronilus Mensae.

The northern plains: A Martian oceanic basin?

263

surface above the Ismenius Level has a higher thermal inertia than the surface below it. At visible image resolution above about 50 m/pixel, the two surfaces also appear distinct, with the surface below the Ismenius Level exhibiting polygonal patterned ground on a scale of tens to a few hundreds of meters across, but the surface above appearing comparatively smooth (Parker et al., 1989). The plains surface between the Deuteronilus and Ismenius Levels contains four landforms that were either not described in detail or not recognized as significant in the earlier, Viking Orbiter-based studies. These are small mounds or knobs, lobate mounds, platy-fractured plains surfaces, and dark streaks trailing from many of the small knobs. The small knobs have been described since Mariner 9 data became available in the early 1970s. When they occurred in abundance, the surface was often mapped as “knobby terrain.” The small knobs on the plains west of Mamers Valles are typically 1–2 km across and are rather scattered, although they can be found in clusters or concentrations in places. Most notably, they tend to be more numerous, and larger, just above the Deuteronilus Level and smaller and more scattered toward the Ismenius Level (Figure 9.11). Many of the small knobs are sharply defined, but a number of them feather into the surrounding plains and appear as though draped by plains materials that settled onto preexisting highs. Similar small knobs seem to be lacking on the plains surface between the Ismenius and Arabia Levels.

Figure 9.11 Small knobs on plains surface between Deuteronilus Level (upper left and upper right) and Ismenius Level (band of rills across lower right of frame) at 44.3°N, 8.5°. Note that knobs decrease in size from Deuteronilus Level to Ismenius Level (from lower to higher elevation). Most knobs have sharp outlines, but some smaller knobs appear draped by plains material.

264

Lakes on Mars

Figure 9.12 Lobate or concentric mound associated with platy/fractured plains texture adjacent to the Deuteronilus Level, 45.5°N, 7.9°E.

Lobate mounds can appear similar to the small knobs at regional scales, such that they were not recognized as morphologically distinct in the earlier Viking-based studies of the region (Figure 9.12). They tend to occur on the plains surface just upslope from the Deuteronilus Level and within several kilometers of it. A few of these lie in contact with the Deuteronilus Level and appear to be overlain or cut by it. Most have a small peak in the center a few hundred meters across, and multiple lobes or concentric ridges and troughs that suggest effusion and flow of relatively dark material from them and thinly overlapping the plains. The lobate mounds tend to be a few to several kilometers across. In daytime IR images, they appear bright relative to the surrounding plains surface, suggesting they comprise fine or unconsolidated material with lower thermal inertia compared to the surrounding plains. Much of the plains surface between the Deuteronilus and Ismenius Levels exhibits the polygonally patterned ground first described by Parker et al. (1989) for this region. The abundance of very high-resolution data available today has revealed a subtle platy-fractured appearance that is similar to, although not as pronounced or large scale, the platy flow material described since Viking days in southern Elysium Planitia. In a few cases, these plates and their associated fractures appear in proximity to the lobate mounds with an expression that suggests the fractures may provide pathways for flows that may be precursor lobate mounds (Figure 9.13). In several places in west Deuteronilus Mensae, a number of dark streaks extend eastward from many of the knobs on these plains. Most of these are found between 45.5°N, 7.6°E and 48°N, 17.8°E with the greatest concentration of them at 46.7°N and

The northern plains: A Martian oceanic basin?

265

Figure 9.13 Fractured plains (lighter unit) above Deuteronilus Level (contact with dark material at left and top of scene), showing flow emanating from fracture at 44.9°N, 9.1°E.

13.4°E. At Viking image scales, they appeared similar to wind streaks and their confinement to the plains between the Deuteronilus and Ismenius Levels went unno­ ticed as anything significant. With very high-resolution regional coverage from CTX, they are revealed to be not wind related at all, but more similar to gores in the wellknown platy flow crusts in the southern Elysium region where topographic obstacles have torn the crust as it moved past them. The streaks in the west Deuteronilus examples never occur below the Deuteronilus Level and fade out before reaching the Ismenius Level (Figures 9.14–9.16). The Ismenius Level intersects northern Mamers Valles at an elevation of approximately –3700 m. As with the Arabia Level, this surface also appears to be tilted westward, from about –3700 m southwest of the Lyot Crater to –3800 m northeast of the Bamberg Crater.

9.3.3 Mamers Valles levels 2, 3, 5, and 7 In addition to the Ismenius Level, three other terrace-like features were identified in the Viking very high-resolution strip of images of Mamers Valles. These terraces, labeled “bench 2, 3, and 5,” from topographically lowest to highest, can still only be traced for a few tens of kilometers along the fretted valley wall and onto the plains surface to the east and west with confidence. For consistency, these were relabeled Mamers Valles Levels 2, 3, and 5 (the Deuteronilus Level would be Mamers Valles Level 1 and the Arabia Level Mamers Valles Level 6). These levels lie at elevations of approximately –3800, –3750, and –3675 m, respectively. An additional, similar

266

Lakes on Mars

Figure 9.14 Dark streaks across plains and plateau surfaces between Deuteronilus and Ismenius Levels. Center of frame, 46.9°N, 12.8°E.

Figure 9.15 Dark streaks associated with three knobs at 46.8°N, 11.8°E. Note that streaks and bright plains between them have a fractured platy appearance similar to ice floes or lava. These streaks are cut by the Deuteronilus Level at right center of frame, partially obscured by a lobate mound above the right center.

The northern plains: A Martian oceanic basin?

267

Figure 9.16 Closeup of dark streak at 46.7°N, 12.4°E.

feature has been identified upon examining the current very high-resolution orbiter images. This feature lies at an elevation approximately 150–200 m above that of the Arabia Level and was given the label Mamers Valles level 7.

9.3.4 The Deuteronilus level This level is probably the best known of the proposed shoreline features. In the east Acidalia and west Deuteronilus region, it separates the thumbprint terrain from the intermediate plains of Parker et al. (1989) and was called either the “interior plains boundary” or “Contact 2.” The Deuteronilus Level is arguably the most pronounced or best preserved of the features interpreted by Parker et al. (1989, 1993), Parker (1994), and Parker and Currey (2001) as coastal in origin. It has been traced around the northern plains, where it does appear to define an approximately horizontal surface (Head et al., 1998). Where it embays topography, it often exhibits a lobate, even arcuate, plan form (Figure 9.17 and 9.18). This led Parker et al. (1989) to compare the feature with lacustrine beach ridges lobate flow fronts with relief at the margins, although on fretted valley walls it is expressed as a topographic bench (e.g., in Mamers Valles where it was first identified, Figure 9.17). The lobate fronts suggest a flow direction from the plains interior and up the flanks of the highlands, to about an elevation of –3850 to –3900 m in Mamers Valles. The plains interior to the Deuteronilus Level is the classic “thumbprint terrain” identified in Viking Orbiter images over 30 years ago (Guest et al., 1977). The thumbprint appearance of this surface is due to a system of bright, low-relief conical

268

Lakes on Mars

Figure 9.17 Northern Mamers Valles fretted valley wall (Mosaic of HRSC and CTX images 7–10 m images), 46.5°N, 14.2°E.

mounds—many with shallow summit pits or craters—contrasted against a relatively dark, smooth plains background. Where these mounds are found near the Deuteronilus Level—the sharp geomorphic contact separating the thumbprint terrain from the intermediate plains (Parker et al. (1989))—they often lie in chains or arcs that parallel the lobate appearance of the contact, thereby giving it this thumbprint appearance. In THEMIS Daytime IR images, the thumbprint terrain may appear somewhat lighter than the plains between the Deuteronilus and Ismenius Levels, though this difference is not substantial. Most of the brightness in IR seems to be due to the abundance of low cones or mounds, which are also relatively bright in visible images. The Deuteronilus Level ranges in elevation from –3800 m southwest of Lyot to about –4000 m at 0°E. A long lobe of thumbprint terrain bounded by the Deuteronilus Level vees into the fretted terrain by about 250 km at 17.5°E. The elevation difference between the northern and southern expression of this “level” is approximately 100 m, with the tilt downward toward the interior of the northern plains.

9.3.5 The Acidalia level This level is very similar in expression to the Deuteronilus Level, although the flow fronts and terraces are in general subtler in this region. Yet, it may be more pro­ nounced and exhibit textures that strongly resemble lava flows to the west in Cydonia Mensae. The Acidalia Level typically lies about 50 m lower in elevation than the Deuteronilus Level and separates the thumbprint terrain from the mottled plains

The northern plains: A Martian oceanic basin?

269

Figure 9.18 Oblique aerial photo of barrier beach ridges in the Bonneville basin, Cedar Valley, Utah. These wave-built ridge forms, and the bright playa surfaces between them, were interpreted as comparable in appearance and scale to those imaged by Viking in Mamers Valles. The lake was toward the top of the frame, past the beach ridges in the foreground. At this scale, these features might be interpreted to suggest that the plains to the north onlap the surface to the south (bottom of scene), similar to the appearance of the Deuteronilus Level on Mars, Figure 17. Scene width �2 km.

identified in Viking Orbiter images by Carr (1981) and Whitbeck and Underwood (1984). The Acidalia Level has been mapped in east Acidalia over about 500 km in a southwesterly direction from about 48°N, 11.3°E. Across this distance, it maintains an elevation of about –4050 m. The mottled plains surface appears relatively dark in daytime IR images, comparable to that of the plains between the Ismenius and Arabia Levels. In west Deuteronilus, the Acidalia Level is at an elevation of –3950 to – 4000 m. Perhaps the most-imaged example of both the Acidalia and Deuteronilus Levels is a 200-m-high mesa east of Cydonia Mensae at 44.7°N, 352.8°E, which was first imaged at moderate resolution early in the Viking mission. Very high-resolution images have been acquired of this feature since, beginning with transects by MOC, and now THEMIS VIS, CTX, HiRISE, and HRSC, with CTX providing a continuous mosaic of the entire feature at 6 m/pixel (Figure 9.19). The mesa lies about 250 km plainward and 200–300 m lower in elevation than the contiguous Deuteronilus and Acidalia Levels along the sloping highland margin, suggesting a plainward tilt of these levels similar to that measured at Acidalia Mensae for the Arabia Level. The Deuteronilus Level encroaches onto the top edges of this mesa, nearly inundating it completely. The Acidalia Level comes partway up the flanks of the mesa on all sides.

270

Lakes on Mars

Figure 9.19 200 m high mesa in east Acidalia Planitia (44.7°N, 352.8°E). The Deuteronilus Level encroaches partway onto the upper surface of the mesa. The Acidalia Level lies partway up the flanks of the mesa about 50 m below the Deuteronilus Level.

The bright, conical mounds characteristic of the thumbrint terrain surface between the two levels is evident on the flanks of the mesa and the mottled plains surface below the Acidalia Level comprises the plains around the mesa. At the Acidalia Level on the south side of the mesa, the conical mounds appear surrounded by plains material associated with the Acidalia Level without being buried by it, suggesting that the plains material is thin and the level was not high enough to cover the mounds. In at least one location along the Acidalia Level, a series of mounds appear to be bisected at the contact, suggesting erosion of preexisting mounds as the plains associated with the Acidalia Level were emplaced.

9.4 Discussion Based on the highest-resolution image data available during the 1980s and the 1990s, Parker et al. (1989, 1993) hypothesized that the best explanation for the geomorphic features and contacts seen along the highland margin was erosion and deposition of material onto preexisting terrain at a series of shorelines around the planet’s northern lowland plains. These shorelines were inferred to indicate paleoclimate conditions that allowed liquid water to remain stable at the surface long enough for wind-driven waves to produce the features through wave refraction and longshore sediment transport. While the Arabia Level does exhibit terracing in the most recent very

The northern plains: A Martian oceanic basin?

271

high-resolution images that is reminiscent of strandlines in terrestrial paleolakes, most of the other mapped levels in this region do not. Instead, boundary morphology at the prominent Deuteronilus Level exhibits lobate flow fronts and textures that resemble low-viscosity lava or debris flows. Still, the MOLA-based topography does verify the earlier impression that the contacts are elevated by hundreds of meters to kilometers with respect to the northern plains interior, which does suggest that millions of cubic kilometers of material receded after the levels were emplaced at the margins of the plains. Parker et al. (1989) inferred marine conditions in a cooling climate based on the following observations: Starting at the Arabia Level and working plainward, plains textures transition from “smooth plains” between the Arabia and Ismenius Levels, to small-scale polygonally patterned ground between the Ismenius and Deuteronilus Levels, to thumbprint terrain (with bright conical hills interpreted as pingos) between the Deuteronilus and Acidalia Levels, to mottled plains below the Acidalia Level. The reasoning was that smooth plains could indicate cold climate conditions had not ensued until the shoreline had receded and plains had been desiccated; small-scale polygons were ice-wedge polygons formed by thermal cycling in a cold climate with water or ice present in the near surface; “one-off” pingos formed in a permanently cold climate after shoreline recession as near-surface groundwater froze. This scenario may have been plausible based on the limited resolution data from Viking, but other morphologies described in this chapter need also be considered in formulating a testable hypothesis based on the most recent, very high-resolution image data. These newly discovered landforms might also be consistent with a cooling marine environment, but may suggest that the initial conditions were never more clement than terrestrial arctic environments. In other words, if Mars had oceans, the observed morphologies may suggest that they were covered with thick ice, and the flow front morphology seen along the Deuteronilus Level may indicate ice-shoving due to a short-lived transgressive event caused by channel activity elsewhere into the northern plains. However, for the morphology to be preserved at the surface at these latitudes today, the ice cover would either need to be dirty or itself be mantled with other material, such as eolian debris, impact ejecta, or lava flows—all of which suggest the long-term presence of subsurface ice and water in the northern plains. Indeed, four of the newly identified landforms at the Deuteronilus and Ismenius Levels might be best understood as related to brief disruptions of a thick debris- and ice-covered ocean, that is, the fluvial rilles above the Ismenius Level, and the lobate mounds, dark streaks, and platy flow-textured plains between the two levels. For the sake of argument, let us assume that an ice-and debris-covered ocean is present at about the elevation of the Ismenius Level, but that it is gradually receding due to loss via sublimation and redistribution elsewhere on Mars. The ice cover is frozen to the substrate at the edges, but floating as the topography drops off toward the northern plains interior. The fluvial rills could have formed due to a catastrophic disruption of this ice cover—perhaps due to an impact or landslide into the ocean. The dark streaks might also have formed at this time, their orientations with respect to the small knobs and bends in the streaks themselves indicating the direction of motion of the ice cover over underlying topography. In this scenario, the dark streaks represent tears in the ice

272

Lakes on Mars

cover as the ice was pushed past the immovable knobs. Subsequent refreezing and sublimation of the water from the debris/ice cover at the Ismenius Level have permanently preserved the tears as pseudomorphs of the earlier setting. When the ocean had receded to about the Deuteronilus Level, floods into the northern plains triggered a minor transgression to produce the lobate flow fronts and mounds at the Deuteronilus Level. The mounds resemble mud volcanoes on Earth and may have formed as water and sediment under pressure broke through the debris/ice cover that had once again frozen to the substrate at the margins. Following cessation of the floods, the disrupted cover refroze, this time producing pingos (and thumbprint terrain) as the debris/ice cover froze.

References Baker, V.R., Strom, R.G., Gulick, V.C., Kargel, J.S., Komatsu, G., Kale, V.S., 1991. Ancient oceans, ice sheets and the hydrological cycle on Mars. Nature 352, 589–594. Carr, M.H., 1981. The Surface of Mars. Yale University Press, New Haven, CT, 226p. Christensen, P.R., Bandfield, J.L., Clark, R.N., Edgett, K.S., Hamilton, V.E., Hoefen, T., et al., 2000. Detection of crystalline hematite mineralization on Mars by the thermal emission spectrometer: evidence for near-surface water. J. Geophys. Res. 105 (E4), 9623–9642. Christensen, P.R., Jakosky, B.M., Kieffer, H.H., Malin, M.C., McSween, H.Y. Jr., Nealson, K., et al., 2004. The Thermal Emission Imaging System (THEMIS) for the Mars 2001 Odyssey Mission. Space Sci. Rev. 110, 85–130. Clifford, S.M., Parker, T.J., 2001. The evolution of the Martian hydrosphere: implications for the fate of a primordial ocean and the current state of the northern plains. Icarus 154, 40–79. Currey, D.R., 1980. Coastal geomorphology of Great Salt Lake and vicinity. Utah: Geol. Mineral Surv. Bull. 116, 69–82. Edgett, K.S., Parker, T.J., 1997. Water on early Mars: possible subaqueous sedimentary deposits covering ancient cratered terrain in western Arabia and Sinus Meridiani. Geophys. Res. Lett. 24, 2897–2900. Guest, J.E., Butterworth, P.S., Greeley, R., 1977. Geological observations in the Cyonia region from Viking. J. Geophys. Res. 82, 4111–4120. Head, J.W.I.I.I., Hiesinger, H., Ivanov, M.A., Kreslavsky, M.A., Pratt, S., Thomson, B.J., 1999. Possible ancient oceans on Mars: evidence from Mars orbiter laser altimeter data. Science 286 (5447), 2134–2137. Head, J.W., I.I.I., Kreslavsky, M., Hiesinger, H., Ivanov, M., Smith, D.E., Zuber, M.T., 1998. Oceans in the past history of Mars: tests for their presence using Mars Orbiter Laser Altimeter (MOLA) data. Geophys. Res. Lett. 24, 4401–4404. Jons, J.-P., 1990. Das relief des Mars: versuch einer zusammenfassenden ubersicht. Geol. Rundsch. 79, 131–164. Lucchitta, B.K., Ferguson, H.H.M., Summers, C., 1986. Sedimentary deposits in the northern lowland plains, Mars. J. Geophys. Res. 91 (Suppl.), E166–E174. Malin, M.C., Edgett, K.S., 1999. Oceans or seas in the Martian northern lowlands: high

resolution imaging tests of proposed coastlines. Geophys. Res. Lett. 26, 3049–3052.

Malin, M.C., Edgett, K.S., 2001. Mars global surveyor Mars orbiter camera: interplanetary

cruise through primary mission. J. Geophys. Res. 106 (E10), 23429–23570.

The northern plains: A Martian oceanic basin?

273

Parker, T.J., 1994. Martian Paleolakes and Oceans. PhD dissertation, University of Southern California, 200p. Parker, T.J., Currey, D.R., 2001. Extraterrestrial coastal geomorphology. Geomorphology 37, 303–328. Parker, T.J., Gorsline, D.S., Saunders, R.S., Pieri, D.C., Schneeberger, D.M., 1993. Coastal geomorphology of the Martian northern plains. J. Geophys. Res. 98, 11061–11078. Parker, T.J., Saunders, R.S., Schneeberger, D.M., 1989. Transitional morphology in West Deuteronilus Mensae, Mars: implications for modification of the lowland/upland bound­ ary. Icarus 82, 111–145. Rossbacker, L.A., 1985. Ground ice models for the distribution and evolution of curvilinear landforms on mars. In: Woldenberg M. (Ed.), Models in Geomorphology. Allen and Unwin, Boston, MA, pp. 343–372. Scott, D.H., Chapman, M.G., Rice, J.W. Jr., Dohm, J.M., 1992. New evidence of lacustrine basins on Mars: Amazonis and Utopia Planitae, Proceedings of Lunar and Planetary Science, 22, 53–62. Scott, D.H., Dohm, J.M., Rice, J.W. Jr., 1995. Map of Mars showing channels and possible paleolake basins. 1:30,000,000, U.S. Geological Survey Miscellaneous Investigations Series, Map I-2561. Whitbeck, N.E., Underwood, J.R. Jr., 1984. Geologic Map of the Mare Acidalium Quadrangle, Mars (revised). Atlas of Mars, 1:5,000,000 Geologic Series, USGS Map I–1614.

10 The Western Elysium Planitia Paleolake

Matthew R. Balme†,*, Colman J. Gallagherþ,

David P. Page§, John B. Murray†, Jan-Peter Muller**

and Jung-Rack Kim**

†

Department of Earth and Environmental Sciences, the Open University, Walton Hall, MK, UK * The Planetary Science Institute, Tucson, AZ, USA þ University College Dublin School of Geography, Planning and Environmental Policy, University College Dublin, Belfield, Dublin, Ireland § Planetary and Space Science Research Institute, the Open University, Walton Hall, MK, UK ** Department of Space & Climate Physics, University College London, Mullard Space Science Laboratory, Dorking, RH, UK

10.1

Introduction

10.1.1 Regional description The Elysium region of Mars (Figure 10.1) lies at the foot of the dichotomy boundary, the broad scarp that marks the transition between the old, heavily cratered southern highlands and the younger northern plains. The region is dominated by three large volcanic edifices atop a broad volcanic rise; compressional (wrinkle ridges) and extensional (fractures) tectonic features; and broad, flat plains. The Cerberus Fossae, a series of kilometer-wide fractures, extend for over 1000 km to the southeast of the Elysium rise. Originating at one of the westernmost fractures is Athabasca Valles, a fluvial channel carved by catastrophic floods that erupted from the Cerberus Fossae (Burr et al., 2002a, 2002b). The channel is over 10 km wide, contains kilometer-scale streamlined islands and extends more than 300 km to the southwest where it terminates in a large, flat-floored basin. Several authors have suggested that these deposits represent a debris-covered, frozen paleolake or sea (Chapman et al., 1990; Scott et al., 1991; Scott and Chapman, 1995). The floor of this basin, greater than 150,000 km2 in areal extent (and hereafter referred to as the “Western Elysium Basin”) is covered by distinctive “platy-ridged” terrain (Plate 10.1), composed of a jumble of fractured plates and sinuous ridges that appear similar in form to terrestrial pack-ice (Brackenridge, 1993; Murray et al., 2005; Rice et al., 2002). Others suggest that the platy deposits represent flood lavas (Hartmann and Berman, 2000; Jaeger et al., 2007; Keszthelyi et al., 2000, 2004a, 2004b; McEwen et al., 1998; Plescia, 1990, 2003) and Lakes on Mars. DOI: 10.1016/B978-0-444-52854-4.00010-6 © 2010 Elsevier B.V. All rights reserved.

276

Lakes on Mars

Figure 10.1 MOLA hill-shaded relief map of the Elysium region of Mars. The study area for this research is shown by the white box and is equivalent to the extent of Figure 10.4. The same background image is used in all following figures except where otherwise stated. Image credit: NASA/JPL/MOLA Science team.

that any original fluvial deposits have long since been covered by a thin lava crust. Such terrain is not restricted to this basin and can be traced continuously for more than 1000 km to the west, and in a huge area over 2000 km to the east (this larger area is

The Western Elysium Planitia Paleolake

277

Plate 10.1 Oblique view of platy terrain in Western Elysium Basin. The impact crater in the middle of the image is 3 km in diameter. This demonstrates that some plates in this image are greater than 10 km across. This southward-looking oblique view uses an HSRC color image from Mars Express orbit h0032. Image credit ESA/DLR/FUB.

hereafter referred to as the Cerberus Plains). In this chapter, we concentrate on the Western Elysium Basin and describe the regional topography, the morphology of the landforms that compose the complex, and the geological evidence in support of the interpretation that this is a paleolake basin. We also compare the main arguments for the alternative interpretation, that this basin is filled with lava. At the end of the chapter, a context figure (Figure 10.13) is provided that gives the location of all the other figures.

10.1.2 Historic context In the earliest map of this region, made using Mariner 9 data (Scott and Allingham, 1976), the Cerberus Plains were interpreted to comprise lava flows mantled by aeolian materials. As higher-resolution Viking Orbiter data became available and the region was studied in greater detail, Plescia (1990) proposed that the Cerberus Plains were covered by geologically recent flood lavas deposits, identified several possible source vents, and made the tentative suggestion that the Cerberus Fossae fractures might themselves be the source for some of these lavas. The alternate interpretation, that the deposits were of fluvial origin (although probably interspersed with lava flows), was also first developed based on mapping from Viking Orbiter data (Scott and Chapman, 1995; Tanaka et al., 1992). Chapman et al. (1990) and Scott et al. (1991) proposed that the basin formed part of a regional paleolake that had been fed by outflow channels from the discontinuity boundary to the south. Brackenridge (1993) went further and suggested that the platy terrain—so obvious in contemporary imaging data, but difficult to discern in Viking images—represented grounded “ice floes,” and that the Cerberus Plains were the site of a geologically recent, inland sea.

278

Lakes on Mars

With the arrival of the MGS, ODY and MEx missions at Mars, data from MOLA, HRSC, THEMIS, and meter-resolution imaging from MOC Narrow Angle (NA) became available for new studies of the region. MOLA and MOC NA data confirmed that Athabasca Valles is a geologically recent, flood-carved, fluvial channel and that its source region is the Cerberus Fossae (Burr et al., 2002a; 2002b). However, McEwen et al. (1998), Keszthelyi et al. (2000; 2004a; 2004b), and Plescia (2003) concluded on the basis of visual similarities to terrestrial lava flow surfaces that the platy terrains on the channel and basin floors were flood lavas, not paleolake deposits. Plescia (2003) suggested that the terminus of Athabasca Valles was buried by these flood lavas and that the Cerberus Plains fluvial systems were reoccupied by later lava flows. In contrast, Rice et al. (2002) and Murray et al. (2005) found evidence for either a sea-ice or ice-rich fluvial (perhaps hyperconcentrated or debris flow) emplacement process for the platy and ridged terrain. Most recently, HiRISE data with spatial resolution up to 25 cm/pixel have permitted the morphology of the landforms in the region to be explored at a meter scale. Jaeger et al. (2007) concluded that small pitted mounds and cones (referred to by them as “Ring Mound Landforms”) in Athabasca Valles are volcanic rootless cones and that the platy terrain in the Western Elysium Basin must also therefore be lava. However, these conclusions have been challenged on stratigraphic grounds (Page, 2008), based on MOC NA and HiRISE observations that show these mounds and other associated landforms superposing small impact craters. In addition, Page et al. (2009) document large (108 years) age disparities between platy and polygonal ground, demonstrating that parts of the surface have evolved since emplacement. No clear consensus as to the origin of these terrains has been reached, however, and it remains as intriguing and controversial a region today as it did in the days of the Viking Orbiter investigations.

10.2 Western Elysium Basin: general description 10.2.1 Basin floor material The terrain of the Western Elysium Basin comprises two main surface morphologies: raft-like plates with a rough, rubbly texture and smoother inter-plates that have a characteristic, decameter-scale, convex-up, polygonal pattern (Figure 10.2c,10.2d). A variety of other landforms exist within the mapped contact but these two surface textures dominate the basin floor. At a small scale, the plate and inter-plate regions are particularly obvious in THEMIS infrared data (e.g., Figure 10.3), indicating marked differences in their thermal properties. The plates range from a few tens of meters to over 50 km in size, show signs of breakup, rotation and drift, and, in many cases, their present pattern can be reassembled into an unbroken, original form (Murray et al., 2005). Where plates have drifted against topographic obstacles, straight and curved “lanes,” many up to several kilometers in length, have formed downstream within the plates themselves (see Fig. 1 of Murray et al., 2005). “Rubble piles” have formed on the upstream side of many of these obstacles (see Figure 10.2 of Murray et al., 2005). Multiple episodes of fracturing can be seen in some plates (Balme et al., 2007),

The Western Elysium Planitia Paleolake

279

suggesting fracturing, healing, and re-fracturing. Where plates have collided, their margins have deformed into rubbly ridges (Figure 10.2e, 10.2f). Similar ridges also occur within plates and sometimes cross over from plates into inter-plate regions. Often these ridges have a distinct saw-tooth, square, or sinusoidal form, similar in some respects to “finger rafting” in terrestrial pack-ice (Balme et al., 2007). Where plates have been heavily broken up and “jostled,” they can be entirely rimmed by ridges as a result of plate-to-plate collisions. Taken together, these observations can only be interpreted to indicate formation of a fractured crust on a mobile substrate, leading to the two main hypotheses for formation on this scale: debris-laden pack-ice or plates of solid, cooled lava floating on a molten interior. HiRISE images (Figure 10.2) reveal that the surfaces of the plates contain many meter-sized clasts set in a networked (cell size about 1 m), ripple-textured background. Given the texture and abundance of larger aeolian features in this region, we interpret the surface to be aeolian material infilling a rough, bouldery substrate. The inter-plate regions appear to be mainly free of drift materials, perhaps explaining why they have higher thermal inertia than the plates (i.e., the plates are brighter in daytime THEMIS infrared images in Figure 10.2 and darker in nighttime images). The thermal inertia (based on maps from the MGS TES, Mellon et al., 2000) of the basin floor as a whole ranges from <100 to 300 thermal inertia units (Putzig, 2006), consistent with a substrate composed of sand-grade material, but nearly an order of magnitude lower than might be expected for bedrock (Presley and Christensen, 1997). It could be argued, however, that any bedrock signal might be hidden by aeolian material. The plates are extremely flat at the 10 m, 100 m, and kilometer scale but, due to the many large clasts (which we estimate to cover about 10–20% of the surface area), are rough at the meter scale, confirming Earth-based radar roughness measurements of this region (Harmon et al., 1992). The inter-plates (Figure 10.2c, 10.2d) display a smooth, “polyconvex” polygonal pattern, generally with cell sizes of less than 10 m. Very few clasts can be seen on the inter-plate surfaces, although linear and curvilinear ridges and grooves and sinuous channels are sometimes seen. Polygon network patterns range from almost circular to rectilinear, and smaller polygonal patterns are often superposed upon a larger, background texture with wavelength of a few tens of meters. Different styles of polyconvex terrains often exist alongside one another (Figure 10.2e) and the network pattern is sometimes controlled by the proximity of plate margins or ridges. Ridges can extend across both plates and inter-plates although, in other examples, ridges are truncated by plate margins. Together, these terrains form a distinctive assemblage of “platy/polyconvex” sur­ faces that are bounded by a clear morphological contact. They contrast strongly, in terms of both morphology and number of impact craters (i.e., surface age), with the surrounding volcanic plains associated with the Olympus Mons volcano to the north.

10.2.2 Mapping the extents of Western Elysium Basin materials We have mapped the extents of the platy/polyconvex morphologies in the area shown by the white box in Figure 10.1 using HSRC, THEMIS, MOC NA, and HiRISE data (Figure 10.4). East of 148.2° E there is essentially complete coverage of

280

Figure 10.2 (Continued)

Lakes on Mars

The Western Elysium Planitia Paleolake

281

HRSC data at <20 m/pixel resolution, but west of this THEMIS VIS (sparse cover­ age at >18 m/pixel) and THEMIS IR (full coverage at 100 m/pixel) data were used, and the mapped geomorphic contact can therefore be considered less reliable. MOC NA and HiRISE data have very limited spatial extent but very high resolution and were used to investigate contacts and morphology in detail. Our mapping shows that platy terrain in Athabasca Valles (first described in this location by Rice et al., 2002) is continuous with the platy terrain on the basin floor, in contrast to suggestions that basin floor materials superpose channel materials (Plescia, 1990). Of the erosional overspill channels that branch south from Athabasca, the largest leads to a channel, about 50 km across and poorly defined topographically (black arrows in Figure 10.3), which is marked by “jostled” kilometer-scale plates. The channel can be traced continuously for nearly 500 km into the main Western Elysium Basin. Small overspill channels from Athabasca Valles merge into this southern channel, but are not overlain by it—again in contrast to a previous mapping (Plescia, 2003). We have found no vent structures in the Western Elysium Basin: the circular feature at 153° E, 5.5° N, tentatively identified as a vent by Plescia (1990), is instead a bright, inter-plate area formed at a slight topographic break where the southern channel enters the main Western Elysium Basin. We therefore suggest that the platy terrain that forms the floor of the basin was entirely sourced from the Cerberus Fossae and transported down Athabasca Valles and the southern channel. We have found no continuous internal flow margins within the Western Elysium Basin, although there are several topographic discontinuities within the basin that are

Figure 10.2 HiRISE observations of surface textures in the Western Elysium Basin. (a) HiRISE image PSP_007685_1850 (center) covers kilometer-scale plates and inter-plate terrain on the basin floor. The background images are HRSC nadir images from Mars Express orbits h2154 and h2143. (b) Platy terrain surface at full resolution. Inset is a two times magnification view that shows the rough surface to comprise clasts set in network-patterned background. (c) Interplate “polyconvex” surface texture. A faint 30–40 m wavelength, convex-up, polygonal pattern is overlain by a more sharply defined 5–10 m wavelength, convex-up, slightly reticulate pattern. These discontinuous mounds and troughs form the majority of the inter-plate surface. Very few clasts are seen, and the polygonal boundaries are marked by gentle troughs, rather than by fractures. (d) Another example of polyconvex terrain. In this example, a single polygonal pattern with wavelength of about 10–20 m dominates. A finer polygonal pattern is perhaps faintly visible in places. (e) Ridges and polyconvex material. Plates can be heavily deformed by ridges both at their margins and within the plates themselves. The topmost part of the image shows the texture of the plate to be dominated by these ridges. Small, sinuous ridges can also be seen within the inter-plate polyconvex terrain and, as seen here, often mark a transition in style of polyconvex terrain. Sinuous fractures or channels (just left of image center, trending north– south) can sometimes be seen within the polyconvex material. (f) Ridges surrounding plates. Where many small plates jostle together, their margins are often ridged as shown here. Image credits NASA/JPL/UofA and ESA/DLR/FUB. North is up and illumination from the left in this and all subsequent images unless otherwise stated.

282

Lakes on Mars

150 °E

155 °E

5 °N

5 °N

150 °E

155 °E

Figure 10.3 THEMIS daytime infrared image mosaic of platy terrain in the Western Elysium Basin. Lighter tones indicate warmer temperatures of surface materials. The contrast in surface temperature reflects the thermophysical properties of the materials that compose the terrain. Athabasca Valles, marked by white arrows, and the broad southern channel, marked by black arrows, can be clearly seen. The jostled plates in the southern channel can be traced into the central part of the main basin. Illumination is from the right. Image credit NASA/JPL/ASU.

draped by platy/polyconvex material and several local flow margins near the edges of the basin where flow has followed two different paths before reforming. We therefore find no evidence for the platy terrain having formed from multiple flows or smaller flows from local vents. The fact that the platy terrain was sourced from Athabasca Valles is reinforced by mapping of rubble piles formed behind topographic obstacles and lanes within the Western Elysium Basin that show a consistent trend away from the termination of these channels (Figure 10.4).

10.2.3 Topography and depth of platy terrain Western Elysium Basin is an almost closed basin, drained only by two spillways—one in the southeast named Lethe Vallis and another in the southwest—that are marked by down-cut, erosional channels. The basin floor has an astonishingly low slope: the top of the western spillway is only 15–20 m below the terminus of Athabasca Valles, over 450 km distant (a regional slope of <0.005°), and the deepest point in the basin is only about 20–30 m below the height of the margins. The contact of the platy/polygonal

The Western Elysium Planitia Paleolake

145 °E

283

150 °E

155 °E

10 °N

10 °N

5 °N

5 °N

145 °E

150 °E

155 °E

Figure 10.4 Map of the extents of Western Elysium Basin materials. The mapped extents of the platy/polyconvex textures are shown in pale gray. The dark arrows indicate flow directions of mobile plates based on observations of rubble-piles built up behind topographic obstacles. The black dashed lines represent down-cut, erosional channels. To the east of the main basin, subbasins are labeled 1 through 4. See Figure 10.13 for the full extents of the mapped platy/ polyconvex terrains and for extents of all other figures presented in the chapter.

terrain, mapped only from morphology, closely follows an equipotential surface (Figure 10.5) and is between –2710 m and –2700 m below Mars datum for much of the Western Elysium Basin. Similarly, the mapped extent of the deposits closely correlates almost exactly with an area defined by MOLA grid-to-grid slopes of <0.05°. Murray et al. (2005) suggested that the surface of the platy terrain has lowered by several tens of meters since formation, inferring that fluid drained from beneath the platy surface or was lost by sublimation. This is in agreement with the overall mapping that shows the basin margins to be a few tens of meters higher than the basin center and with observations of plates that have been grounded on obstacles (Figure 10.6a). In places, fractures up to 80 km long form parallel to the “shore” of the platy terrain and show movement away from the margin, sometimes leaving large platy fragments attached to the margin in a configuration redolent of land-fast ice. These are analogous to “tide cracks” at sea-ice margins on the Earth and also suggest that fluid retreated from beneath the platy crust before complete solidification (Figure 10.6b). Murray et al. (2005) estimated that the maximum depth of fluid in the basin was between 31 and 53 m. They measured the diameters of 15 craters that had been

284

Lakes on Mars 145 °E

150 °E

5 °N

5 °N X

150 °E

145 °E

Y

X

Y’

X’

Figure 10.5 Topography of the Western Elysium Basin from MOLA gridded data. The mapped contact from figure 10.4 is shown by the dotted black line. XX’ and YY’ represent topographic profiles across the basin as shown below the main figure. The tone of the shading represents the topographic height of the MOLA gridded data. White: –3700 m to –3710 m; pale gray: –3710 m to –3720 m, mid gray: –3720 m to –3730 m; and dark gray: <–3730 m. The profiles and gridded data show that the floor of the basin is extremely flat and that the margins of the basin closely follow an equipotential surface. For much of the main basin the mapped contact is close to the –3700 contour, but at the southwest of the main basin the contact is closer to the –3710 m contour. Note: the MOLA data are gridded at 128 pixels per degree (a grid spacing of 500 m) but this near the equator, the original MOLA tracks are in places separated by several kilometers.

submerged and applied the rim height to diameter expression H = 0.04D0.31 (for craters <7 km in diameter and where H is rim height and D is crater diameter; Garvin et al., 2003). We have applied the same method to 50 additional submerged or breached craters and 12 craters that were not breached by the flow. The average rim height of inundated craters is 43 m and that of unbreached craters is 60 m. The

The Western Elysium Planitia Paleolake

285

Figure 10.6 Evidence for retreat of fluid from beneath the plates. (a) A plate (upper left) has overridden the rim of an impact crater and come to rest over it. The rim of the crater is draped by polyconvex material. This suggests that the entire crater was once totally within the flow. MOC NA image E2300722. Image credit NASA/JPL/MSSS. (b) A plate has withdrawn from a “shoreline,” suggesting lowering of the base level of the fluid. Part of HRSC nadir image from Mars Express orbit h2154. Image credit ESA/DLR/FUB.

smallest diameter unbreached crater has an inferred rim height of 41 m and the largest submerged crater has a rim height of 65 m. Given that the original topography might not have been horizontal, these results give only a first order estimate, but show that the local maximum depth of fluid in the basin was between 40 and 65 m. Murray et al. (2005) found several examples of inundated craters forming closed basins. This demonstrates that post-depositional lowering of the surface of the flow occurred even where no outflow was possible, suggesting a loss of volume of the flow in these small, closed basins, rather than flow withdrawal.

10.3

Landforms

10.3.1 Constructional landforms In addition to the distinctive rubbly plates, the polyconvex intra-plates, and the heavily ridged terrains, other distinctive landforms such as pitted mounds and several types of patterned ground are present within the Western Elysium Basin. Pitted mounds (e.g. Figure 10.7) range in morphology from simple mounds or cones to compound structures comprising central cones or mounds (often with a summit pit) surrounded by a moat defined by ring fractures or steeply tilted and fractured edges. The moat fractures often show a polygonal pattern. Cones-within-cones and multiple cones within single moats are sometimes seen. Pitted mounds in the Western Elysium Basin are most common within the erosional channels such as Athabasca Valles, Lethe

286

Lakes on Mars

Figure 10.7 Pitted mounds. (a) Pitted mounds with moats bounded by polygonal fractures and ridges. Chains of mounds, multiple central mounds formed within the same moat, and concentric mounds-within-mounds are all common. Part of HiRISE image PSP_002226_1900. Image credit NASA/JPL/UofA. (b) Here, the mounds form a more random pattern and do not show moats. Also, mounds without apparent pits can be seen, perhaps arguing for an intrusive, rather than explosive, origin. Occasionally, “wakes” can be seen behind mounds. Part of HiRISE image PSP_002648_1880. Image credit NASA/JPL/UofA.

Vallis, and the western spillway, but also occur within the main basin. They tend to form near the margins of the flows but, in Lethe Vallis for example, the mounds are seen both on the channel floor and high on streamlined islands within the channel. Locally, the mounds can form in chains, groups, or can be approximately randomly distributed. Such mounds in Athabasca Valles and across the Cerberus Plains have been interpreted to be volcanic rootless cones (Jaeger et al., 2007; Keszthelyi et al., 2004a; Lanagan et al., 2001), ice-cored mounds or pingos (Burr et al., 2005; Page and Murray, 2006; Rice et al., 2002), or even cryo-volcanic cones (Hoffman and Tanaka, 2002). Polygonally patterned ground is common in the Western Elysium Basin. In addi­ tion to the ubiquitous polyconvex terrain, positive-relief margin (i.e. low-centred) polygons occur (Figure 10.8). These are common on the banks of the erosional channels and on the surfaces of the streamlined islands in the Athabasca Valles, but are also occasionally found on the channel floors. Polygonal cells are generally less than 10 m in diameter with 4–5 sides. Radial and circumferential associations with pits and topographic features are common, similar to features described in the eastern part of the Cerberus Plains (Page, 2007). These positive relief polygonal patterned grounds closely resemble orthogonal networks that form in ice-cemented soils on Earth (Lachenbruch, 1962). The high spatial resolution of HiRISE has recently revealed the presence of sorted patterned ground (Figure 10.9) in the Western Elysium Basin. These landforms

The Western Elysium Planitia Paleolake

287

Figure 10.8 Polygonal patterned ground with positive relief margins. (a) Polygonal ridges with a wavelength of 10–20 m on the ejecta blanket of a crater in the Athabasca Valles main channel. Some of the ridges appear to cross-cut lobate boundaries, suggesting a postdepositional origin. Part of HiRISE image PSP_006696_1895. Image credit NASA/JPL/UofA. (b) Polygonal ridges on the topmost surface of streamlined island in the Athabasca Valles. The distinctive radial/concentric patterns seen just above the center of the image are similar to orthogonal networks seen in frozen soils on Earth. Part of HiRISE image PSP_1540_1890. Image credit NASA/JPL/UofA.

Figure 10.9 Sorted stone circles on the banks of Lethe Vallis. (a) More than 200 sorted stone circles with flat interiors, cell-sizes of about 15–20 m and wide, clastic margins have been identified on the northern (shown here) and southern banks of Lethe Vallis. They form within the mapped contact of the main Western Elysium Basin deposits, note the polyconvex terrain within the small crater just to the north, and provide strong evidence for formation in an ice-rich, particulate substrate. (b) Closeup view of the central part of Figure 10.8a. Part of HiRISE Image PSP_004072_1845. Image credit NASA/JPL/UofA.

suggest that the substrate in which they formed was particulate and ice-rich, with some degree of freeze–thaw activity (Balme et al., 2009). There is no precedent or known mechanism for the sorting of clastic material or self-organization of such patterned ground by volcanic processes, although sorting can occur on “icy” volcanic surfaces

288

Lakes on Mars

on Earth (Noguchi et al., 1987). However, this relies on there being both a source of particulates (e.g., from severe weathering of the upper parts of the lava flow) and a source of ice. It is unlikely that either of these effects could occur for very young lava flows near the Martian equator unless they were later inundated with fluvially deposited materials.

10.3.2 Erosional landforms There is substantial evidence in the Western Elysium Basin for erosion by fluid flow. For example, the western spillway drains into a broad channel that extends northwest for over 450 km, before terminating in another basin approximately 10,000 km2 in area. The floor of this basin has the same platy/polygonal surface morphology seen in the main Western Elysium Basin and has been filled through a single breach in a low ridge at the end of the feeder channel. The sides of the breach and the topographic obstacles in the channel show deep notches, demonstrating the erosional power of the fluid flow within the channel, even when more than 1500 km away from its source. Similar features are seen both on the upstream sides of topographical obstacles within Athabasca Valles, many hundreds of kilometers closer to the source, and in terrestrial cataclysmic flood terrains such as the channeled scablands of the northwest United States. The eastern outflow from the Western Elysium Basin is through an interconnected complex of subbasins and spillways (Figure 10.4), including Lethe Vallis. In contrast to earlier studies (Keszthelyi et al., 2004a), HRSC images allow the Lethe Vallis to be linked with the termination of Athabasca Valles; the two channels form part of the same system. At the termination of Lethe Vallis and a smaller, unnamed channel to the North are delta-like, distributary forms (Figure 10.10). These debouch into basins 1 and 3, which were probably already filled from a higher southern Athabasca channel spillover. Lethe Vallis bifurcates as it enters basin 3, with each branch of the channel ending in a similar, delta-like landform. The northern branch of the channel clearly became abandoned before the flooding ceased, as there is now a fluvial hanging valley (figure 10.10b) on the wall of the deeper, eastern-flowing channel. The presence in Lethe Vallis of a distributary hanging valley, abandoned due to avulsion into the eastward-flowing channel, with both the hanging valley and the avulsion fronted by a terminal distributary channel network, suggests that this system originally terminated in standing water. This is supported by the observation that similar distributary forms are not seen in basins fed from a single source. We suggest that the formation of the hanging valley was driven by base-level lowering, perhaps due to the overspill of water at the margin of basin 3 and its rapid drainage. Evidence for headward erosion can be seen in many of the erosional channels. In Lethe Vallis (Figure 10.11) a series of cuspate scarps and streamlined islands, over 1.5 km in width, mark a topographic discontinuity in the channel-long profile and are similar in plan form to terrestrial flood cataract systems (e.g., Dry Falls nr. Spokane, WA, USA; Baker and Milton, 1974). In Lethe Vallis, both rubbly (similar to the surface texture of the plates in the main Western Elysium Basin) and polyconvex terrain occur within the channel. Rubbly material is also found on the top surfaces of

The Western Elysium Planitia Paleolake

289

Figure 10.10 Termination of Lethe Vallis. (a) Lethe Vallis, an erosional channel linking the main Western Elysium Basin to the lower basins in the complex, terminates in two delta-like distributary landforms. The mapped extent of the Western Elysium Basin material is shown by the black line. The black arrow indicates the location of the sorted stone circles shown in Figure 10.9 and the white box shows the location of Figure 10.10b. HRSC nadir image from Mars Express orbit h2099. Image credit ESA/DLR/FUB. (b) Closeup of the terminal bifurcation in Lethe Vallis. The northern branch has become abandoned, leaving a fluvial hanging valley. CTX image P16_007118_1836_XI_03N204W. Image credit NASA/JPL/MSSS.

streamlined islands and high on the channel banks. The polyconvex texture is found in all settings within the mapped contact. This suggests that the fluid that emplaced (or solidified to form) the rubbly material was at some point filling the channel, but between the platy material and at flow margins (i.e., extending all the way to the “high-stand”) was the material that had (or went on to have) the polyconvex texture. The rubbly texture, however, is not seen on erosional features such as dry cataracts and there is no evidence of plates “settling” over these breaks in slopes. Instead they have a well-defined, pristine morphology. Similar cataracts are found in the small channel that connects the eastern Western Elysium Basin to basin 1 (Figure 10.11d), in the channel connecting basin 1 to basin 3, at the western spillway from the Western Elysium Basin, and in the northwest-trending channel into which it feeds. A different suite of erosional landforms occurs near the source of Athabasca Valles at the margins of an irregular pit formed in polyconvex material (Figure 10.12a). Here, there appears to be evidence for postdepositional erosion. The polyconvex material, similar in form and setting to other polyconvex terrain throughout the region, has been eroded back, with scarp retreat leaving cuspate and spur-and-gully forms in the pit walls. As described by Balme and Gallagher (2009), this morphology appears to be a result of sapping or undermining by loss of a fluid, as shown by the small tributary networks draining from the scarp margin toward the fracture (Figure 10.12b). Metersized clasts visible at the base of the scarp appear to have eroded out of the polyconvex material. There is also evidence of gully formation and deep incision

290

Lakes on Mars

Figure 10.11 Cataracts in Western Elysium Basin channels. (a) Lethe Vallis shows evidence of fluvial erosion, including streamlined islands, down-cut channels, and relict cataracts. The white box indicates position of Figure 10.12b. Part of HiRISE image PSP_006762_1840. Image credit NASA/JPL/UofA. (b) Up to three possible channel systems appear to have been active in this part of Lethe Vallis, each headed by one or more horseshoe-shaped erosional scarps. Note the streamlined forms below the cataracts. The white box shows the position of Figure 10.12c. (c) Closeup view of the channel floor textures. A polyconvex surface texture can be seen above the cataract and a positive-relief polygonized texture below. (d) A similar cataract in the small erosional channel leading from the eastern side of the main Western Elysium Basin to the lower basin 1. Part of HiRISE image PSP_008265_1860. Image credit NASA/JPL/UofA.

into the polyconvex layer that indicate overland flow. Figure 10.12c shows branching gullies that perhaps link up with the polygonal sculpture on the surface. These gullies cross-cut both the polygonal sculpture and the fractures that have formed parallel to the scarp edge. As such, they represent evidence of postdepositional erosion of the polyconvex material, and the drainage speaks of a volatile-rich—not rocky—source.

The Western Elysium Planitia Paleolake

291

Figure 10.12 Postdepositional erosion near the source of Athabasca Valles. (a) Two low-relief surfaces are separated by a crenulated front of spur-and-gully forms and cuspate niches. (b) Retrogressive niches backed by a higher surface (lower left) dominated by polygonal fractures and fronted by a dendritic channel system. (c) Branching, backwasting gullies (arrowed “X”) formed by the exploitation of down-slope-oriented cracks. The gullies grade, via the headwalls of retrogressive niches, onto a lower polygonized surface rich in pitted mound landforms, and perhaps what appear to be the scars of collapsed pitted mounds. (d) Crack-controlled retrogressive niches with pitted floors lacking surface channels. The backwasting that forms niches appears to have been controlled by the preexisting polyconvex network. We interpret the transition from pitted ground at the base of the scarp to a dendritic channel network to suggest a transition from subsurface sapping of thawed ground ice to surface channelized flow. Part of HiRISE image PSP_009280_1905. Image credit NASA/JPL/UofA.

292

Lakes on Mars 140 °E

145 °E

150 °E

155 °E

10 °N

10 °N

5 °N

5 °N

0°

0° 140 °E

145 °E

150 °E

155 °E

Figure 10.13 Context image for figures shown in this chapter. The extent of figures shown in the chapter are represented by white boxes, except where their small size at this scale would make this impossible, in which case the center of the image is located by a white arrow. The gray shaded area is the mapped extent of the Western Elysium Planitia floor materials (including the western outflow and basin).

10.4 Formation age of the Western Elysium Basin deposits A formation age for the Western Elysium Basin floor materials of <10 Ma has been derived using impact crater counting (Murray et al., 2005). Athabasca Valles itself has been determined to have been resurfaced <20 Ma (Berman and Hartmann, 2002) or 2 to 8 Ma (Burr et al., 2002a). However, these ages could be misleading for several reasons. First, large numbers of secondary impact craters from the large, recent impact crater Zunil are found across the Western Elysium Basin (McEwen et al., 2005; Preblich et al., 2007). Although the presence of secondaries probably does not invalidate the crater-counting method per se (Hartmann, 2007), Zunil formed very recently (perhaps as recently as few million years; McEwen et al., 2005), and its secondaries clearly do cover the Western Elysium Basin material, so their presence might dominate crater numbers locally. Furthermore, although in the north the basin materials superpose older, volcanic plains, in the south the basin materials have been exhumed from beneath the winderoded Medusa Fossae Formation (Burr et al., 2002a; Malin and Edgett, 2001). The Medusa Fossae Formation is composed of friable, low-density, or ice-rich material (Watters et al., 2007) unconformably overlying the equatorial topographic discontinuity. The deposition age of these materials is debated, with evidence for ongoing deposition, mobility, or modification throughout the Amazonian (Head and Kreslavsky, 2004; Hynek et al., 2002; Tanaka et al., 2005), but the fact that it at least partly overlies the southern platy terrain shows that the reported crater-retention age of the platy terrain is possibly too young: if the Medusa Fossae Formation formerly covered the basin, this

The Western Elysium Planitia Paleolake

293

might have prevented craters from accumulating, and the entire region might be older than its crater-retention ages suggests.

10.5

Discussion

10.5.1 Did the Western Elysium Basin contain a paleolake? The relict cataracts, down-cut erosional channels, streamlined islands in the channels, fluvial distributaries, hanging valley, eroded topographic obstacles (even when many hundreds of kilometers downstream from source), and the topographic and spatial arrangement of interconnected basins, spillways, and channels, all provide strong evidence for the Western Elysium Basin having once contained a standing body of water. Whether this was a transient lake, with any lacustrine deposits overlain by later flood lavas, or whether the surfaces we see today are the product of these floods, depends upon the interpretation placed upon the platy terrain, polyconvex inter-plates, and other landforms that make up this assemblage. Although transitional scenarios involving a combination of fluvial and volcanic processes might have occurred (see Section 10.5.3), we can summarize the volcanic and fluvial hypotheses using the following two end-member models: • Volcanic model—One or more periods of fluvial flooding carved the erosional landforms and channels. One or more subsequent episodes of volcanic activity reoccupied the fluvial landscape, filled in the basins to the same or slightly higher level than the fluvial episodes, and left the observed platy and polyconvex morphologies on the basin floor. Besides burial (and perhaps weathering) the surfaces visible now are considered primary. • Fluvial model—One or more fluvial episodes were responsible for both carving the erosional landforms and deposition of the platy and polyconvex basin floor material. The surfaces visible now are a combination of primary flood deposits and secondary landforms that have evolved over time due to the initial presence of ice in the regolith.

The volcanic hypothesis implies a landscape that, erosion and burial apart, has remained essentially unchanged since formation (for it must be solid rock). The aqueous hypothesis, however, allows for a dynamic landscape that has “evolved” over time; ice in the regolith potentially responding to sublimation, melting, thermal contraction, cryoturbation, and deflation to modify the landscape in a number of constructive and destructive ways. Using topography, morphology, stratigraphy, and nonimaging remote sensing, we now consider which of these ‘static’ or ‘dynamic’ landscapes accords best with observation. We also discuss theoretical considerations for the persistence of ice in the Martian subsurface.

Topography and regional mapping First, the deposits mapped in the Western Elysium Basin appear to all have been sourced from the Athabasca Valles or the parallel, southern channel. The regional slope across the Western Elysium Basin is extremely low (about 10 m over a 500 km distance; Figure 10.5), about the same as variations in the geoidal ocean surface on Earth (Reigber et al.,

294

Lakes on Mars

2007). Given the lower gravity and the non-Newtonian rheology of lavas, such a low slope argues strongly against volcanic deposition—the slope of even very flat terrestrial continental flood basalt surfaces being around 0.1° (Keszthelyi et al., 2006)—unless the rheology of the lava was essentially the same as that of water. Second, there are several examples of topographically closed impact craters in the Western Elysium Basin that have been completely inundated by the flow (Murray et al., 2005) and show evidence for substantial (i.e., at least 50% of the estimated original depth) postdepositional lowering of the surface. Whilst such a considerable volume loss would be expected if the flow was of water (i.e., by infiltration into the regolith, sublimation of ice, and deflation of liberated sediments), this is not the case for solidified lava flows. Third, all large terrestrial basaltic volcanic flow fields studied to date are inflated compound pahoehoe flows (Hon et al., 1994; Self et al., 1998), in which small breakouts of lava from a parent flow form individual lobes that fill in local topo­ graphic lows and then “inflate” as more lava enters. Flow lobes can inflate to tens of meters thick and can merge laterally to form sheet flow lobes that can be several kilometers wide (Self et al., 1996). Such flows are emplaced slowly (i.e., the advance of the flow front is slow, even though the transport of lava packets from source to front might be rapid) and are thermally efficient due to the insulating upper crust. The transport of lava for distances of up to several thousand kilometers is therefore possible. The distance from the Cerberus Fossae source to the western basin is a minimum of 1250 km. Individual plates within the Western Elysium Basin are tens of kilometers across. Thus, for these plates to be mobile on the surface, the flow must have both been >100 km in width, and continuous with no internal flow margins. If the Western Elysium Basin platy terrain formed as an inflated pahoehoe flow, then it must essentially have formed as a single, enormous sheet lobe. Moreover, for plates to drape over impact craters as observed, the flow must have inflated over and around these obstacles, and then have completely encased these craters within the mass of the flow, before then settling down onto it. This is unlikely, because local topographic obstacles in inflating lavas tend to form inflation or “lava-rise” pits (Walker, 1991), rather than being overridden by the flow. Again, rather than being consistent with terrestrial analogs for inflating lavas, we suggest that only if lava acted in the same way as water, and that the entire Western Elysium Basin was deposited as a single, turbulent flow, could the observed topography be observed. This is the exact model proposed by Jaeger et al. (2008), who cite the absence of lava tubes, lava inflation features, and compound flow lobes in Athabasca Valles as evidence against a slow inflation model, and in which they suggest that the entire complex was emplaced as a turbulent lava flow in a time span of only a few weeks. Such a lava flow would cool extremely quickly, however, due to the rapid entrainment of crust into the flow and due to enhanced radiative cooling from an increased fraction of exposed hot core. For example, in the simple model of Keszthelyi et al. (2004a), the rate at which lava crust is entrained into the flow is proportional to the heat lost due to this “ice-cube effect.” Similarly, the heat loss by radiation is inversely proportional to the fraction of the flow covered by a crust. Given that these two effects dominate the

The Western Elysium Planitia Paleolake

295

heat loss of the flow, any turbulence that would disrupt the crust would cause the flow to rapidly solidify. Fourth, we note that although neither the slow, sheet-lobe inflation model nor the fast, turbulent emplacement model appear to be appropriate, an alternative is sug­ gested by Keszthelyi (2007) and Keszthelyi et al. (2000, 2004a) who propose that the Western Elysium Basin platy flows are more like lava flows from the Icelandic Laki 1783–1784 eruptions. In these eruptions, a surge in the lava flux into an inflating pahoehoe flow allowed the crust to break up into platelike rafts, some up to several hundred meters in diameter. Although the platelike surfaces of the Laki flow are similar in some ways to the Western Elysium Basin deposits, obvious lobate and digitate flow-margins and distinctive breakouts from which lava has emerged can all be seen in the Laki flow field. These are not seen in the Western Elysium Basin and, in many cases, plates can be seen touching “shorelines” and obstacles and not bounded by a margin of pahoehoe-like textures.

Morphology Because we only see planetary surfaces in two dimensions, the third dimension of depth that relates geomorphology to geology is, in almost all cases, unavailable to us. However, in the absence of in situ confirmation of lithology or process, morphologic arguments leave us open to the problems of form convergence, where different causes or processes may result in strikingly similar visual effects. For example, the pitted mounds in the Western Elysium Basin have morphologies similar to both periglacial (pingos) and volcanic (rootless cones) landforms. As another example, it might be argued that the presence of clasts and rubbly ridges demonstrate that the platy rafts cannot be dustcovered ice floes and instead must be rubbly pahoehoe lava flow surfaces (Keszthelyi, 2007), but this again is a morphological argument from analogy that is incomplete for two reasons. First, debris flows on Earth can entrain very large clasts (Decaulne et al., 2005; Iverson, 1997) and it is likely that the low Martian gravity would make transport of boulder-sized clasts by turbulent fluvial floods even more likely—that Athabasca Valles itself has been substantially down-cut demonstrates there is no shortage of clastic material available for transport—so flood deposits could be just as rough as a lava-flow surface. Second, it is essentially unknown what surface morphologies would result from a frozen Martian catastrophic flood. A Martian flood would have simultaneously been freezing and boiling where open water was exposed to the thin atmosphere—perhaps creating a slurry of ice crystals and debris (Bargery et al., 2008)—and entraining vast amounts of debris due to the highly energetic turbulent flow. Furthermore, it is unknown to what extent the original surface morphology would have been altered by volatile loss due to sublimation in the thin Martian atmosphere. In short, we have no good terrestrial analogs with which to compare the observations: whilst the remnants of frozen cata­ strophic mudflows or debris flows might be smooth, flat, and have sublimed quickly away, they might also be as rubbly, blocky, and contain similar abundances of ridged and lobate forms as do lava flows. Despite the limitations of purely morphological arguments, we suggest that HiRISE images do reveal at least two landforms that support the existence of a

296

Lakes on Mars

permafrost or periglacial landscape in the Western Elysium Basin: first, the sorted stone circles shown in Figure 10.9 demonstrate that the substrate was ice-rich, particulate, and that there was some degree of freeze–thaw (Balme et al., 2009); second, the gullies, cuspate scarps, and tributary channels shown in Figure 10.12 demonstrate postdepositional retreat of the polyconvex terrain. The morphology is consistent with both retrogressive badlands-type thermokarst and block failures asso­ ciated with secondary thermoabrasion and thermoerosion of permafrost by fluvial undercutting (French, 1996), but is difficult to explain from a volcanic perspective. This implies that these polygonal terrains, at least, are not lava flows.

Stratigraphy We need not rely on morphology alone in a planetary geological context, we can also constrain origin stratigraphically. This line of reasoning has the advantage over morphologic interpretation in that we simply replace the question “what is this” (with an unknown number of possible answers) with “what happened first?” and “what happened next?” so as to derive a geological history of the landform or land­ scape studied. We explore two ways in which stratigraphy can constrain process: the overall pattern of deposition of platy terrain and its interaction with topography, especially in Lethe Vallis, and evidence of stratigraphic superposition from highresolution images across the wider Cerberus Plains. First, Lethe Vallis (Plate 10.2) appears to follow a path defined by the margins of ancient lava flows. However, these flow margins are not formed in the platy terrain— instead, platy rafts can be seen to overlie these flows at the mouth of Lethe Vallis (Plate 10.2) and also within basin 2. Platy textures, pitted mounds, and polyconvex terrain all occur on Lethe’s banks and on streamlined islands in the channel. Erosional forms such as the cataract in Figure 10.11 appear pristine and are not overlain with textures that could be said to represent lavas. Thus, in a unit bounded by a single contact, platy terrain is the topmost texture in one area but pristine erosional channel floor is the topmost texture just a few kilometers downstream. This does not fit with a model in which fluvial flooding carved the channels and later lava floods formed the platy material. If two episodes of flooding occurred, one on either side of an episode of flood lavas, the stratigraphy might be explained, but the model would be incon­ sistent with the wider view—we have not observed fluvial erosion of the plates themselves, and the floor of Lethe Vallis is itself occupied in places by platy material, which would likely have been destroyed by later fluvial floods. The stratigraphy is, however, explained if the process that emplaced the platy rafts also carved the erosional channels. For example, plates of icy material and flood debris could become stuck at the entrance to Lethe Vallis while still allowing flow through the channel, in the same way that “arching” of pack-ice in channels (Sodhi, 1977) occurs on Earth. Eventually the rafts would ground on the underlying topography as the basin drained, as appears to be the case here. Similar associations between platy terrain and erosional forms are found in the other channels in the Western Elysium Basin. Second, because surface land form is often the first expression of underlying geology (Rhoads and Thorn, 1996), every feature we see at the surface relates to

Plate 10.2 Lethe Vallis. (a) MOLA colorized hill-shade of Lethe Vallis. The color ramp has been stretched to bring out the topography, with red representing –2690 m (and anything higher) and dark blue –2745 m (and anything lower). The thin dark line represents the mapped contact and the heavy dashed lines represent ancient flow-margins that are overlain by the platy terrain. The white box shows the location of Plate 10.2b. The discontinuities in the Lethe channel are probably due to interpolation errors in the gridded data. (b) Plates overlying ancient flowmargins and the edge of the Lethe channel. HRSC nadir image from Mars Express orbit h2143. Image credit ESA/DLR/FUB.

298

Lakes on Mars

the substrate in such a way as to give us clues about its origin. Given the problems of form convergence, a campaign has been under way for the past four years to determine the stratigraphic affinities of the landforms in the wider Cerberus Plains region, relative to impact craters (Page, 2007, 2008; Page and Murray, 2006; Page et al., 2009). To date, every landform investigated—the polygonal sculpture, pitted mounds, subsidence pits, and surface ridging—has shown evidence of secondary origins consistent with an intrusive, periglacial origin (i.e., ice-wedge polygonization, pingos, and thermokarst and solifluction) and argues strongly against a primary (volcanic) origin for these landforms and the landscape they form within. Jaeger et al. (2007) demonstrate that the sites of pitted mound formation have moved or “evolved” over time, the wakes behind the mounds also severed where plates have rafted apart, indicating that the mound formation must therefore be primary. There is no precedent for such mobile pingo genesis in the terrestrial environment, so the substrate that the mounds are formed in may also have a postdepositional component, as appears to have been confirmed by recent observations of large difference in crater retention ages between platy and polyconvex terrain in Athabasca Valles (Page et al., 2009).

Nonimaging remote sensing and survival of ice In addition to imaging studies, remote sensing spectrometry and radar instruments can provide information about the composition of the Western Elysium Basin deposits. GRS data show the abundance of hydrogen in the upper 50–100 cm of the regolith (Boynton et al., 2002; Feldman et al., 2004) and are generally used as a proxy measurement for water-ice. These data show that there is less than 5% water-ice in this topmost layer in the Western Elysium Basin region. Radar studies using the ESA MARSIS (Picardi et al., 2004) and NASA SHARAD (Seu et al., 2007) instru­ ments have found evidence for a radar transparent layer in both the Western Elysium Basin and across the wider Cerberus Plains (Safaeinelli et al., 2007), but find little evidence consistent with a buried ice-rich layer (Boisson et al., 2008), except in a few local areas (Orosei et al., 2008). Taken together, these data suggest that if the Western Elysium Basin was originally filled by water-rich material then most of this drained away as the basin emptied, or sublimed away later. The fact that the floors of the eastern and western spillways are only a few meters above the main basin floor suggests that the basin drained significantly during, or very soon after, the event that filled it. This in turn suggests that any ice-rich layer that might have been left in the main basin would have been originally only a few meters in depth, even though the basin was likely filled to a few tens of meters’ depth at its high stand. Whether the platy morphologies currently observed are consistent with debriscovered ice-rich material is debated. The survival of dust-covered ice under Martian conditions is strongly dependent on model parameters (Smoluchowski, 1968) and no model is able to reproduce the natural variability of real geological materials (Clifford, 1998). It is therefore difficult to say to what depths debris-covered ice, or ice-rich sediments, will become desiccated as a function of time. The fact that the Medusa Fossae Formation overlies the southern margin of the platy/polygonal terrain suggests

The Western Elysium Planitia Paleolake

299

also that these deposits might have been protected from degradation by overlying material for some unknown time. Whether there is any ice remaining is also debated: we note that the results of Boisson et al. (2008) appear to be as consistent with a substrate with 20% ice by volume as with a “dry” substrate and the data presented by Orosei et al. (2008) also suggest local regions in which ice might still be present.

10.5.2 A nonvolcanic interpretation for Western Elysium planitia Although there is still much work to be done to understand the details of the processes that shaped this region, the morphologic and stratigraphic evidence presented here do not appear consistent with a lava flow origin for the platy/polygonal terrain. We suggest that permafrost and periglacial processes provide a plausible alternative to volcanism and that a range of observations support this interpretation. We further suggest that the source of water (and therefore ice) for these processes was a transient lake that filled the basin. The observed erosional landforms and down-cut channels were formed by the draining of this basin. In this model, the main basin was inundated a small number of times (perhaps only once) and for a geologically short time period. The plates seen today formed as competent, mobile rafts of ice and debris, perhaps with debris material welded to the bottom of the mobile ice. Withdrawal of the underlying fluid left the plates grounded on either the original topography or on newly deposited sediments. Sublimation then left these areas covered by debris or lag deposits (and aeolian material) with little or no ice remaining in the uppermost tens of centimeters today. The fact that the original depth of the ice-rich layer might only have been a few meters could explain why later sublimation of ice from the substrate has not been destructive of morphology and allowed large blocks of material that were once on or within the plates to now be seen. The positive-relief polygons and polyconvex terrain represent postdepositional thermal contraction cracking in the ice-rich substrate, and the sorted stones circles, back-wasted polygonal layers, and pitted mounds demonstrate that there was sustained freeze–thaw activity in the deposits, although this might have occurred substantially after the original fluvial floods (indeed, there is every indication that this was the case; Page et al., 2009). The absence of morphologies consistent with periglacial or permafrost processes immediately outside of the mapped “high stand” reinforces the interpretation that the catastrophic flood deposits provided the source of ice for these landforms. If the ice in the regolith was deposited from the atmosphere, as is likely in mid to high Martian latitudes (Head et al., 2003a; Kreslavsky and Head, 2002; Mustard et al., 2001), it would be expected to drape the topography, rather than infill the lowest areas as is observed here.

10.5.3 Other hypotheses The two models described in detail here (fluvial then periglacial vs. fluvial then volcanic) represent the best articulated range of possible formation mechanisms for

300

Lakes on Mars

the platy/polygonal terrains. We also briefly consider two other hypotheses: (1) a purely volcanic “lava acting as an erosive agent in the same way as water” model and (2) a “combination” model of multiple episodes of volcanic and fluvial action. The purely volcanic hypothesis has the advantage of simplicity, but struggles to explain how the widespread erosional and constructional landforms described above could have formed. Although lava erosion is implicated in carving lunar rilles (Carr, 1974), it is unlikely that avulsions, notches, and cataracts could have been created hundreds of kilometers downstream from the source without the lava being fully turbulent. As described in Section 10.5.1.1, turbulent fluid lava would cool and solidify extremely quickly. Thus, the arguments given against lava infilling the preexisting fluvial landscape apply even more so to this hypothesis. In addition, the stratigraphic and morphological evidence discussed above also argue against such flood volcanism. A model based on a combination of processes seems stronger in that the platy terrain might be explained by slowly emplaced lavas and that landforms such as the polygonal terrain, pitted cones and mounds, sorted circles, and the degradation features at the head of Athabasca Valles might be explained by later fluvial/periglacial processes that covered much of the finer-scale morphology of the earlier lavas. However, this model also suffers from topographic arguments against slowly inflating lavas (Section 10.5.1.1) and is complex in that it requires at least three specific and separate fluvial–lava–fluvial flood events. It also does not fit the stratigraphic argu­ ments outlined in Section 10.5.1.3. We therefore suggest that the mechanism by which the Western Elysium Basin terrain was emplaced is most similar to the fluvial endmember model described in Section 10.5.1.

10.5.4 The wider implications of a western Elysium Basin paleolake The water that formed the paleolake emerged from the Cerberus Fossae. Several models have been proposed to explain this, including groundwater discharge from an aquifer breached by fracturing (Manga, 2004; Plescia, 2003) and dike emplacement into a cryosphere (Head et al., 2003b). Given that these models suggest that liquid water from the deep subsurface was emplaced onto the surface of the Western Elysium Basin, several authors have proposed that the deposits of the Western Elysium Basin would be an ideal target for the search for relict Martian life (Chapman et al., 1990; Murray et al., 2005; Rice 1997). We suggest that the latest data support the fluvial interpretation of the Western Elysium Basin and thus, if the challenge of landing upon and traversing terrain of significant meter-scale roughness can be overcome, this area would be a prime candidate for future in situ investigation. Thermal contraction polygons, fluvial erosion features, and the deflation of a surface by sublimation are all suggestive of the presence of ice in the regolith. Sorted stone circles, pingos, and permafrost degradation point not only to the presence of ice but also to a periglacial environment that has undergone freeze–thaw cycles in the near-surface. As discussed by Balme et al. (2009) and Balme and Gallagher (2009) this suggests a warmer regional climate in the geologically recent past than that currently considered and forms part of a growing body of evidence, including observations of recent gullies

The Western Elysium Planitia Paleolake

301

(Malin and Edgett, 2000), that supports the recent melting of ice into liquid water at the Martian surface (Costard et al., 2002; Page, 2007). An assemblage of landforms, surface textures, and topographic and stratigraphic signatures similar to those described here are present throughout the Cerberus Plains region, although not confined to such a well-delimited basin. These deposits cover an area of 107 km2 (Tanaka et al., 2005) and, being geologically recent, have been posited as the source for multiple ejections of the young Martian meteorites (e.g., Tornabene et al., 2006). The discussion as to whether these deposits represent recent voluminous flood lavas, or the products of aqueous floods, is therefore wider than the provenance of the Western Elysium Basin platy material: as a type area for Martian surface dating and impact crater chronology (Hartmann and Neukum, 2001), the nature of these materials is central to Mars science.

10.6

Conclusions

We have found neither continuous, internal flow margins nor evidence for vent structures in the main Western Elysium Basin. Platy terrain in Athabasca Valles is continuous with platy terrain on the basin floor. We therefore suggest a single source (the Cerberus Fossae) for the Western Elysium Basin platy material. The topographic and mapping data support a fluvial, rather than volcanic, empla­ cement mechanism for platy terrain in the Western Elysium Basin, based on our current knowledge of terrestrial flood lavas. We note that preliminary interpretations of radar data do not show the basin to currently contain a deep, widespread layer of ice, but (i) do reveal possible evidence for local, remnant, ice-rich regions and (ii) are inconclusive in terms of discriminating between the surface being ice-free or contain­ ing ice at the 10–20% by volume level. The distinctive landforms of the Western Elysium Basin are consistent with formation in an ice-rich, particulate substrate. The sorted patterned ground and the back-wasting of polygonized terrain are indicative of an ice-rich regolith modified by freeze–thaw or melt processes. These morphological assertions are consistent with stratigraphic observations in the wider Cerberus Plains region that show landforms superposing impact craters. Taken together, these observations provide evidence for a recent, warmer climate and a recent, low-latitude, periglacial landscape across the Elysium Planitia. The low-latitude, geologically young, lacustrine deposits in Western Elysium Basin provide an important target for future in situ astrobiological investigations.

References Baker, V.R., Milton, D.J., 1974. Erosion by catastrophic floods on Mars and Earth. Icarus 23 (1), 27–41. Balme, M.R., Gallagher, C., 2009. An equatorial periglacial landscape on Mars. Earth Planet. Sci. Lett.

302

Lakes on Mars

Balme, M.R., Gallagher, C., 2009. An equatorial periglacial landscape on Mars. Earth Planet. Sci. Lett. 285, 1–15. Balme, M.R., Gallagher, C., Page, D.P., Murray, J.B., Muller, J.-P., 2009. Sorted stone circles in Elysium Planitia, Mars: implications for recent Martian climate. Icarus 200 (1), 30–38. Balme, M.R., Murray, J.B., Ackley, S.F., Muller, J.-P., Kim, J.-R., 2007. Morphological evidence for a sea-ice origin for Elysium Planitia platy terrain, Lunar Plan. Sci. Conf. XXVIII (CDROM), Abstr. # 2202. Bargery, A.S., Lans, S.J., Wilson, L., Gilbert, J.S., 2008. Laboratory simulation of the release of subsurface water and brine on to the surface of Mars and icy satellites, Lunar Plan. Sci. Conf. XXIX (CDROM), Abstr. # 1702. Berman, D.C., Hartmann, W.K., 2002. Recent fluvial, volcanic and tectonic activity on the Cerberus Plains of Mars. Icarus 159, 1–17. Boisson, J., Heggy, E., Clifford, S.M., Frigeri, A., Plaut, J.J., Picardi, G., 2008. Exploring Athabasca subsurface geoelectrical properties using MARSIS RADAR data: hypothesis on volcanic or fluvial origin of the local morphology, Lunar Plan. Sci. Conf. XXIX (CDROM), Abstr. # 1819. Boynton, W.V., Feldman, W.C., Squyres, S.W., Prettyman, T.H., Brückner, J., Evans, L.G., et al., 2002. Distribution of hydrogen in the near surface of Mars: evidence for subsurface ice deposits. Science 297 (5578), 81–85. Brackenridge, G.R., 1993. Modern shelf ice, equatorial Aeolis quadrangle, Mars, Lunar Plan. Sci. Conf., XXIV, Abstr. # 175. Burr, D.M., Grier, J.A., McEwen, A.S., Keszthelyi, L.P., 2002a. Repeated aqueous flooding from the Cerberus Fossae: evidence for very recently extant, deep groundwater on Mars. Icarus 159 (1), 53–73. Burr, D.M., McEwen, A.S., Sakimoto, S.E.H., 2002b. Recent aqueous floods from the Cerberus Fossae, Mars. Geophys. Res. Lett. 29, 1013, doi:10.1029/2001GL013345.. Burr, D.M., Soare, R.J.,, Wan Bun Tseung, J.-M., Emery, J.P., 2005. Young (late Amazonian), nearsurface, ground ice features near the equator, Athabasca Valles, Mars. Icarus 178 (1), 56–73. Carr, M.H., 1974. The role of lava erosion in the formation of lunar rilles and Martian channels. Icarus 22 (1), 1–23. Chapman, M.G., Scott, D.H., Tanaka, K.L., 1990. Elysium Basin, Mars: implications of a deep intermittent lake system, Lunar Plan. Sci. Conf. XXI, 180–181. Clifford, S.M. (1998), Mars: the effect of stratigraphic variations in regolith diffusive properties on the evolution and vertical distribution of equatorial ground ice., paper presented at Lunar Plan. Sci. Conf. XXIX (CDROM), Abstr. # 1922. Costard, F., Forget, F., Mangold, N., Peulvast, J.P., 2002. Formation of recent Martian debris flows by melting of near-surface ground ice at high obliquity. Science 295, 110–113. Decaulne, A., Sæmundsson, P., Petursson, O., 2005. Debris flow triggered by rapid snowmelt: a case study in the Gleidarhjalli area, northwestern Iceland. Geogr. Ann. A 87 (4), 487–500. Feldman, W.C., Prettyman, T.H., Maurice, S., Plaut, J.J., Bish, D.L., Vaniman, D.T., et al., 2004. The global distribution of near-surface hydrogen on Mars. J. Geophys. Res. 109, E9, doi:10.1029/2003JE002160. French, H.M., 1996. The Periglacial Environment. , 2nd edn. Addison Wesley Longman, Harlow. Garvin, J.B., Sakimoto, S.E.H., Frawley, J.J., 2003. Craters on Mars: global geometric proper­ ties from gridded MOLA topography, Sixth international conference on Mars, Abstr. #3277. Harmon, J.K., Sulzer, M.P., Perillat, P.J., Chandler, J.F., 1992. NOTE: Mars radar mapping: strong backscatter from the Elysium basin and outflow channel. Icarus 95, 153–156.

The Western Elysium Planitia Paleolake

303

Hartmann, W.K., 2007. Martian cratering 9: toward resolution of the controversy about small craters. Icarus 189 (1), 274–278. Hartmann, W.K., Berman, D.C., 2000. Elysium Planitia lava flows: crater count chronology and geological implications. J. Geophys. Res. 105 (E6), 15011–15025. Hartmann, W.K., Neukum, G., 2001. Cratering chronology and the evolution of Mars. Space Sci. Rev. 96 (1–4), 165–194. Head, J.W., Kreslavsky, M.A., 2004. Medusae Fossae Formation: ice-rich airborne dust depos­ ited during periods of high obliquity? Lunar Plan. Sci. Conf. XXXV (CDROM), Abstr. # 1635. Head, J.W., Mustard, J.F., Kreslavsky, M.A., Milliken, R.E., Marchant, D.R., 2003a. Recent ice ages on Mars. Nature 426, 797–802. Head, J.W., Wilson, L., Mitchell, K.L., 2003b. Generation of recent massive water floods at Cerberus Fossa, Mars by dike emplacement, cryospheric cracking and confined aquifer groundwater release. Geophys. Res. Lett. 30, 11, doi:10.1029/2003GL017135. Hoffman, N., Tanaka, K.L., 2002. Co-existing “flood” and “volcanic” morphologies in Elysium as evidence for cold CO2 or warm H2O outbursts, Lunar Plan. Sci. Conf. XXXIII (CDROM) Abstr # 1505. Hon, K., Kauahikaua, J., Denliner, R., McKay, K., 1994. Emplacement and inflation of pahoehoe sheet flows: observations and measurements of active floes on Kilhauea Vol­ cano, Hawaii. Geol. Soc. Am. Bull. 106, 351–383. Hynek, B.M., Arvidson, R., Phillips, R.J., 2002. Geologic setting and origin of Terra Meridiani hematite deposits on Mars. J. Geophys. Res. 107, E10, doi:10.1029/2002JE001891. Iverson, R.M., 1997. The physics of debris flows. Rev. Geophys. 35 (3), 245–296. Jaeger, W.L., Keszthelyi, L., McEwen, A.S., and the HiRISE team, 2008. Emplacement of Athabasca Vallis flood lavas, Lunar Plan. Sci. Conf. XXIX (CDROM), Abstr. # 1836. Jaeger, W.L., Keszthelyi, L.P., McEwen, A.S., Dundas, C.M., Russell, P.S., 2007. Athabasca Valles, Mars: a lava-draped channel system. Science 317 (5845), 1709–1711. Keszthelyi, L., 2007. Comparison of flood lavas on Earth and Mars. In: M.G. Chapman (Ed.), The Geology Of Mars: Evidence From Earth-Based Analogs. Cambridge University Press, Cambridge, pp. 126–150. Keszthelyi, L., Thordarson, T., McEwen, A.S., Haack, H., Guilbard, M.-N., Self, S., et al., 2004a. Icelandic analogs to Martian flood lavas. Geochem. Geophys. Geosys. 5, Q11014, doi:10.1029/2004GC000758. Keszthelyi, L., McEwen, A.S., Thordarson, T., 2000. Terrestrial analogs and thermal models for Martian flood lavas. J. Geophys. Res. 105 (E6), 15027–15050. Keszthelyi, L., Self, S., Thordarson, T., 2006. Flood lavas on Earth, Io and Mars. J. Geol. Soc. London 163, 253–264. Keszthelyi, L., Burr, D.M., McEwen, A.S., 2004b. Geomorphologic/thermophysical mapping of the Athabasca region, Mars, using THEMIS infrared imaging, Lunar Plan. Sci. Conf. XXV (CDROM), Abstr #1657. Kreslavsky, M.A., Head, J.W., 2002. Mars: nature and evolution of young latitude-dependent water-ice-rich mantle. J. Geophys. Res. 29, 15, 10.1029/2002GL015392. Lachenbruch, A.H., 1962. Mechanics of thermal contraction cracks and ice-wedge polygons in permafrost, GSA Special Papers, 70, 1–69. Lanagan, P.D., McEwen, A.S., Keszthelyi, L., Thordarson, T., 2001. Rootless cones on Mars indicating the presence of shallow equatorial ground ice in recent times. Geophys. Res. Lett. 28 (12), 2365–2367. Malin, M.C., Edgett, K.S., 2000. Evidence for recent groundwater seepage and surface runoff on Mars. Science 288, 2330–2335.

304

Lakes on Mars

Malin, M.C., Edgett, K.S., 2001. Mars global surveyor Mars orbiter camera: interplanetary cruise through primary mission. J. Geophys. Res. 106, 23429–23570. Manga, M., 2004. Martian floods at Cerberus Fossae can be produced by groundwater. Geophys. Res. Lett. 31, L0202702, doi:10.1029/2003GL018958. McEwen, A.S., Edgett, K.S., Malin, M.C., Keszthelyi, L., Lanagan, P.D., 1998. Mars Global Surveyor camera tests the Elysium Basin controversy: it’s lava, not lake sediments, Geol. Soc. Am. Abstr. Prog., 30 (7). McEwen, A.S., Preblich, B.S., Turtle, E.P., Artemieva, N.A., Golombek, M., Hurst, M., et al., 2005. The rayed crater Zunil and interpretations of small impact craters on Mars. Icarus 176, 351–381. Mellon, M.T., Jakosky, B.M., Kieffer, H.H., Christensen, P.R., 2000. High-resolution thermal inertia mapping from the mars global surveyor thermal emission spectrometer. Icarus 148, 437–455. Murray, J.B., Muller, J.-P., Neukum, G., Werner, S.C., van Gasselt, S., Hauber, E., et al., and the HRSC CoI team, 2005. Evidence from the Mars express high resolution stereo camera for a frozen sea close to Mars’ equator. Nature 434, 352–355. Mustard, J.F., Cooper, C.D., Rifkin, M.K., 2001. Evidence for recent climate change from the identification of youthful, near-surface ice. Nature 412, 411–414. Noguchi, Y., Tabuchi, H., Hasegawa, H., 1987. Physical factors controlling the formation of patterned ground on Haleakala, Maui. Geogr. Ann. A 69A (2), 329–342. Orosei, R., Cartacci, M., Cicchetti, A., Federico, C., Flamini, E., Frigeri, A., et al., 2008. Radar subsurface sounding over the putative frozen sea in Cerberus Pallas, Mars, Lunar Plan. Sci. Conf. XXIX (CDROM), Abstr. # 1866. Page, D.P., 2007. Recent low-latitude freeze–thaw on Mars. Icarus 189 (1), 83–117. Page, D.P., 2008. Comment on “Athabasca Valles, Mars: a lava-draped channel system”. Science 320 (5883), 1588b, (doi:10.1126/science1154849). Page, D.P., Balme, M.R., Grady, M., 2009. Dating climate change on Mars. Icarus 203 (2), 376–389. Page, D.P., Murray, J.B., 2006. Stratigraphical and morphological evidence for pingo genesis in the Cerberus plains. Icarus 183 (1), 46–54. Picardi, G., Biccari, D., Seu, R., Plaut, J.J., Johnson, W.T.K., Jordan, R.L., et al., 2004. MARSIS: mars advanced radar for subsurface and ionosphere sounding. In: A. Wilson (Ed.), MARS EXPRESS: The Scientific Payload. ESA Publications Division, Noordwijk, pp. 51–69. Plescia, J.B., 1990. Recent flood lavas in the Elysium region of Mars. Icarus 88, 465–490. Plescia, J.B., 2003. Cerberus Fossae, Elysium, Mars: a source for lava and water. Icarus 164, 79–95. Preblich, B.S., McEwen, A.S., Studer, D.M., 2007. Mapping rays and secondary craters from the Martian crater Zunil. J. Geophys. Res. 112, E05006, doi:10.1029/2006JE002817. Presley, M.A., Christensen, P., 1997. The effects of bulk density and particle size sorting on the thermal conductivity of particulate materials under Martian atmospheric pressures. J. Geophys. Res. 102 (E4), 9221–9229. Putzig, N.E. (2006), Thermal inertia and surface heterogeneity on Mars, PhD thesis, University of Colorado, Boulder. Reigber, C., Schwintzer, P., Stubenvoll, R., Schmidt, R., Flechtner, F., Meyer, U., et al., 2007. A High Resolution Global Gravity Field Model Combining CHAMP and GRACE Satellite Mission and Surface Data: EIGEN-CG01C, GeoForschungsZentrum, Potsdam, Scientific Technical Report STR 06/07. Rhoads, B.L., Thorn, C.E., 1996. Observation in geomorphology. In: B.L. Rhoads, and C.E. Thorn (Eds)., The Scientific Nature of Geomorphology, Proceedings of the 27th Binghamton Geomorphology Symposium. John Wiley, New York, pp. 21–56.

The Western Elysium Planitia Paleolake

305

Rice, J.W., 1997. Elysium Basin, Mars: an intermittent or perennial lake from Noachian to Amazonian time, Workshop on Early Mars, Abstr. # 3052, Lunar and Planetary Insititute, Houston. Rice, J.W., Parker, T.J., Russell, A.J., Knudsen, O., 2002. Morphology of fresh outflow channel deposits on Mars, Lunar Plan. Sci. Conf. XXXIII (CDROM), Abstr. # 2026. Safaeinelli, A., Plaut, J.J., Phillips, R.J., Seu, R., Gim, Y., Biccari, D., et al., 2007. Radiotransparent deposits in the Elysium region of Mars as observed by MARSIS and SHARAD RADAR, Seventh International Conference on Mars, Abstr. # 3206. Scott, D.H., Allingham, J.W., 1976. Geologic map of the Elysium quadrangle of Mars, U.S. Geol. Surv. Misc. Invest. Ser., MAP I-935 (MC–15). Scott, D.H., Chapman, M.G., 1995. Geologic and topographic maps of the Elysium palaeolake basin, Mars, U.S. Geol. Surv. Misc. Invest. Ser., MAP I–2397. Scott, D.H., Rice, J.W., Dohm, J.M., 1991. Palaeolakes and lacustrine basins on Mars, Lunar Plan. Sci. Conf. XXII, 1203–1204. Self, S., Keszthelyi, L., Thordarson, T., 1998. The importance of pahoehoe. An. Rev. Earth. Plan. Sci. 26, 81–110. Self, S., Thordarson, T., Keszthelyi, L., Walker, G.P.L., Hon, K., Murphy, M.T., et al., 1996. A new model for the emplacement of Columbia River basalts as large, inflated pahoehoe lava flow fields. Geophys. Res. Lett. 23 (19), 2689–2692. Seu, R., Phillips, R.J., Biccari, D., Orosei, R., Masdea, A., Picardi, G., et al., 2007. SHARAD sounding radar on the Mars Reconnaissance Orbiter. J. Geophys. Res. 112, E05S05, doi:10.1029/2006JE002745. Smoluchowski,R., 1968. Mars: retention of Ice. Science 159, 1348–1350. Sodhi, D.S., 1977. Ice Arching and the Drift of Pack Ice through Restricted Channels, CRREL Report 77-18, Hanover. Tanaka, K.L., Chapman, M.G., Scott, D.H., 1992. Geologic map of the Elysium region of Mars, U.S. Geol. Surv. Misc. Invest. Ser., MAP I–2147. Tanaka, K.L., Skinner, J.A., Hare, T.M., 2005. Geologic map of the northern plains of Mars, U.S. Geol. Surv. Misc. Invest. Ser., Map I–2888. Tornabene, L.L., Moersch, J., McSween, H.Y., McEwen, A.S., Piatek, J.L., Milam, K.A., et al., 2006. Identification of large (2–10 km) rayed craters on Mars in THEMIS thermal infrared images: implications for possible Martian meteorite source regions. J. Geophys. Res. 111, E10006, doi:10.1029/2005JE002600. Walker, G.P.L., 1991. Structure, and origin by injection of lava under surface crust, of tumuli, “lava rises”, “lava-rise pits”, and “lava-inflation clefts” in Hawaii. Bull. Volc. 53 (7), 546–558. Watters, T.R., Campbell, B.A., Carter, L., Leuschen, C.J., Plaut, J.J., Picardi, G., et al., 2007. Radar sounding of the medusae fossae formation Mars: equatorial Ice or Dry, low-density deposits? Science 318, 1125–1128.

11 The sedimentary record of

modern and ancient dry lakes Gian G. Ori

International Research School of Planetary Sciences Universita’ d’Annunzio Viale Pindaro, Pescara, Italy; Ibn Battuta Centre, Faculte´ des Sciences, Universite´ Caddy Ayad, Boulevard de Safi Marrakech, Morocco

11.1

Introduction

The existence of dry lake deposits on Mars has been proposed since the Viking era. It appears to be supported now by mineralogy and high-resolution imagery from the most recent missions. Combined, morphology and mineralogy provide convincing lines of evidence that the planet experienced a hydrological cycle involving both ground and surface water. As climate changed,evaporitic environments were likely to have dominated before conditions close to the present were emplaced. The large family of dry lakes includes a variety of morphologies, types of surfaces, facies, and sedimentary structures where the evaporation of water plays a major role. To describe them, the term sabkha is largely used in the literature, mostly for describing case histories from Northern Africa and the Middle East. This term describes flat areas of clay and sand deposits, which are often encrusted by salt. The term also applies to both inland and coastal sabkhas. The processes are, to a certain extent, common to both settings, as far as the most inner part of the coastal sabkhas is concerned. Playa is a Spanish term for sabkha and is used in North America where most of the playas, unlike sabkhas, are controlled by active tectonics. North American playas are filled with water and closed basins bound by mountain chains or ridges, whereas sabkhas form mostly in cratonic areas and they may also be located in shallow depressions. Most Sabkhas in the Middle East are located on the coast of the Persian Gulf and reflect shoreline sedimentation. The term sabkha is also largely used for continental dry lakes. In South America the sabkhas of the Atacama Desert in Chile are called salars. This latter term is somehow different from the monotonous American and African examples because in many instances they show permanent or quasi-permanent standing bodies of water. They are very shallow (less than a meter) and the main depositional processes are evaporitic. The intricate terminology reflects the different geographical areas where each type of lake occurs and is not a mere terminological issue as different names are based on different geographical and geological settings. Therefore, under the broad term “dry lakes” a large variety of environments are included (Figure 11.1). They obviously also share a large number of features and processes that distinguish them from other lacustrine basins and arid settings. Lakes on Mars. DOI: 10.1016/B978-0-444-52854-4.00011-8 © 2010 Elsevier B.V. All rights reserved.

308

Lakes on Mars

Alluvial fan Barchan dune

Dry lake

Terminal fan

Figure 11.1 The main sub-environments of a dry lake system.

Dry lakes are flat areas of siliciclastic deposits (clay to sand) encrusted or covered by salts deposited by the evaporation of water. Surface water can be totally absent or can be present intermittently and in rare instances it can be present for long periods of time (Warren, 1999). The terminology also reflects environmental and climatic conditions but processes, facies, and structures are similar and belong to the same sedimentary facies model. The sedimentation at the margin of dry lakes is dominated by mechanical deposi­ tion of siliciclastic detritus that form the background for chemical or biochemical deposition of mostly sulfates and carbonates (Drake et al., 1994). Toward the center the evaporitic deposition becomes more important. The central facies may consist of dry evaporitic surfaces fed by groundwaters or by saline lakes (Figure 11.2). These lakes can be ephemeral or permanent (at the scale of tens of years). These character­ istics make the sabkha good analogs to several sulfate deposits on Mars that may have been deposited in lacustrine environments. In this chapter, we use the term playa as synonymous to dry lake and equivalent to continental sabkha. Sabkhas can be marine and formed on coastal areas such as those of the Persian Gulf, or continental like the dry lakes spread over the dry lands. The processes in the marine and continental sabkha are somewhat similar apart from the mechanical influence of the wave or tidal actions and the chemical influence of the salt content of the seawater. However, in the context of lakes on Mars, we will discuss examples from continental sabkhas that are probably the most relevant terrestrial analogs.

11.2 Facies and sedimentary environments Dry lakes are complex environments subjected to numerous processes. We will refer to the area where the direct salt sedimentation occurs as saline lakes. These are the major, but not the only, components of their environments. Dry lakes form in flat areas where the most active depositional process is the evaporation of water. This evapora­ tion bears to the deposition of the salts present in the water and to the formation of evaporitic deposits. The deposition of these salts may occur on the surface (sedimen­ tary interface) from a body of water or within the desiccated sediments due to the groundwater. In perennial or ephemeral saline bodies of water, salts may form at the

The sedimentary record of modern and ancient dry lakes

309

Figure 11.2 Top—The Boujoudur sebkha in Western Sahara. The dry lake fills a depression deep enough to intersect the groundwater. The saline lake occupies a large part of the central area and mud flats are present mostly in the northern margin. The south margin is dominated by a large terminal fan (image from Google Earth); Bottom—The Columbus crater high-albedo material could be the remnants of evaporitic deposits accumulated in a playa.

water–sediment interface producing brine-pool floors and at the water–atmosphere interface (Braitsch, 1971; Hardie, 1991). Evaporitic deposits in arid lands are highly variable. They occur in continental settings and range from saline soils up to bodies of standing, over-concentrated water (brine) in salinas (Figure 11.3). Therefore, there is a continuum of processes and morphologies from soil containing nodular carbonates (caliche) or gypsum crusts (gypscrate) to sabkhas. The sabkha’s environment is a sink for both evaporitic and siliciclastic sediments. In addition to the lacustrine-related processes, fluvial and eolian activity plays an important role in providing sediment and in modeling the environment that need to be considered in the formation and evolution of sabkhas. These terrigenous deposits directly affect their margins. The fine-grained portion of their deposits may be transported by winds into the central part of the dry lakes. However, the evaporitic process provides the bulk of the sediments. Surface evaporation has a major role in the accumulation of salt deposits, with contribution from groundwater as well. Another important process affecting sabkhas is the wind deflation degrading the salty surface and creating a flat basin floor. Continental sabkhas form in the lowest areas of arid enclosed basins where the surface is horizontal and vegetation absent. Inside the basin, several environments occur along with the evaporitic deposits proper: alluvial fans, sand flats, mud flats,

310

Lakes on Mars

a

b

Figure 11.3 (a) Exposed evaporite deposits in a saline lake in the Atacama of Chile; (b) Halite deposits at the surface of Chott el Djerid (Southern Tunisia) extensively affected by polygonal ridges.

saline lakes (ephemeral or quasi-perennial), ephemeral streams and their terminal fans, shoreline features, and spring deposits or mounds. Alluvial fans border the basins when it is located in tectonically active areas such as in the American Southwest but are less frequent in the cratonic setting of the Sahara and Arabian Peninsula. Coarser deposits occur in the apron of the alluvial fan because arid alluvial fan chiefly consists of mass flow deposits that are not able to move over large distances. During large floods, sheet floods from the alluvial fans may affect the margins of the dry lakes forming sand flats. These areas are monotonous horizontal surfaces made up of sand with very shallow channels incised by localized erosion of the unchanneled floods. Moreover, because these deposits dry quickly, the wind action is able to rework the deposits efficiently, leveling the entire surface. Sand flat can also be the product of the eolian degradation of terminal fans. These features represent the termini of ephemeral streams and consist of distributary channels that distally vanish into mud flats. Shorelines are subtle in the driest sabkhas because they are rarely inundated and wave action cannot build appreciable fronts and terraces. In tectonically active basins, lacustrine facies show a sharp boundary with the surrounding desert environments, in many cases as alluvial fan aprons. Instead, in flat and stable settings they may form gradual boundaries that are represented by ephemeral waterlogged zones. Water does not percolate into the subsurface due to the clay content of the bordering mud flats and it sustains halophilic vegetation. This vegetation also controls eolian erosion during the dry periods by stabilizing the shoreline. Eolian environments are major compo­ nents in arid lands and they play a special role in the dry lakes systems. Barchans and

The sedimentary record of modern and ancient dry lakes

311

transverse dunes as well as sand sheets may border the playas. Sometimes, in places, they can be flooded during major water highstand events, producing interdune playas, possibly resembling those identified at Meridiani on Mars (e.g., Grotzinger et al., 2005). Mud flats are actually dominated by evaporitic sedimentation such as those active in saline lakes that occupy the hydrological lows of basins. The transition from sand flat to mud flat is gradual and consists in an increasing mud content. The difference is that in the mud flats the salt minerals are deposited in a substratum made up of clay and carbonates.

11.2.1 Saline lakes and mud flats Evaporitic deposition in saline lakes is controlled by the presence of surface runoff or by the intersection of the water table with the surface. Some saline lakes are perma­ nent or quasi-permanent and, in such case, subsurface water is high enough to flood the depression at the center of the basin. The deposition of evaporitic minerals occurs when concentrated brines form due to evaporation (Figure 11.3). The balance between evaporation and the influx of fresher waters controls the brine formation and evolution. The sedimentation occurs directly from the brine and in some cases up to the complete evaporation of the water. Evaporitic sedimentation can also take place when the water table is deep and cannot directly influence the sedimentation in the sabkha. The capillarity zone above the water table is enlarged by evaporation in arid climate. This process, called evaporative pumping, brings capillarity water near the surface and evaporates the water before it reaches the interface. Evaporitic minerals are therefore deposited within the sediments such as isolated nodules and subsurface crust (Warren, 1999). This process also occurs in arid and semi-arid soils. Due to the two different mechanism of evaporitic deposition (from a surficial brine or from evaporative pumping), evaporitic minerals form irrespective of the salinities of the water that feeds the basins. However, the brine types and the mineral assem­ blages that precipitate are dependent on the composition of the solute carried on by the water supply (Sonnenfeld, 1984). The continuous evaporation and evapo-transpiration (evaporative pumping) occur­ ring in the central part and hydrological lows of saline lakes cause a concentration gradient of groundwater toward the basin’s center. This concentration produces first a saturation in calcium and magnesium carbonates with the consequent deposition of carbonates such as caliche, calcite cemented layers, and then diffuse micrite in the host sediment. These deposits are typical of alluvial fan to sand flats at the basin margins, whereas in the most central part carbonate deposition exhibits travertines, pisolites, and calcite cements. These evaporitic carbonate are mostly deposited below the sedimentary interface, which has important implications for the quest of carbonates on Mars since in most of the arid and semi-arid environments carbonate precipitation occurs in the near subsurface. Brine deposition and evaporative pumping, in addition to sheet floods from runoff and wave actions, continuously compete in the formation and destruction of

312

Lakes on Mars

sedimentary structures. Surficial processes produce features such as stratification, wave and current ripples that are destroyed by the displacement and intrusion of nodular evaporites from the subsurface (Schreiber, 1988). The evaporation at the surface of pond accumulates evaporitic crusts at the sedimentary interface. This process is fast and the loss of water is complete. As a result, the crust includes metastable or highly soluble salts and can reach several tens of centimeter in thickness. The evaporite can be structureless or display sedimentary structures such as horizontal lamination and symmetrical cross-lamination. Structures are usually produced by wave action and are better preserved when they are draped and covered by fine-grained sediments (Figure 11.4). Siliciclastic sediments are brought into the saline lake through eolian and fluvial processes. The composition of the crust depends on a diversity of factors including the chemical composition of the water which in turn depend on the geology of the catchment basin. Because most of the calcium carbonate has already been deposited in the marginal areas, crusts can be rich in halite, sodium sulfate, and carbonate. Gypsum may form in this environment and in the drier areas of the mud flat. Deposition is dominated by three stages: flooding, evaporitive concentration, and desiccation (Hardie 1985). During flooding, water flows into the

Figure 11.4 Top: Pleistocene deposits of Chott el Rarsha (Southern Tunisia) showing a saline lake facies of symmetrical (wave) ripple stratification in halite and gypsum crystals embedded in clay; Bottom: Wave ripple at the shore of a saline lake, Chott el Jerid, Southern Tunisia.

The sedimentary record of modern and ancient dry lakes

313

basin from the surrounding alluvial fans and ephemeral streams. Water is brackish and may dissolve part of the minerals deposited during early evaporitic episodes. Mud transported by flood water is deposited in the basin. During this stage, mud flats grow with siliciclastic deposition that are then affected by interstitial evaporitic cementation and the formation of nodules and cornerstones. These subsurface processes linked to the presence of an aquifer combine in the formation of the standing body of water. The newly formed saline lake starts to evaporate, salt deposition occurs and involves vadose and phreatic processes. The brackish water is quickly transformed into a brine that begins to crystallize salts within the water column and deposits them at the bottom of the lake. With evaporation, the saline lake separates into ponds scattered across the lake surface. Following this stage, desiccation is complete but evaporation continues to affect the system by triggering evaporitic pumping that results in the formation of groundwater brines. The now dry pan surface corrugates break the saline crust in polygons and form pressure ridges (Figure 11.3). Early diagenetic processes disrupt the near subsurface deposits. During the evaporation phase and the formation of the salt crusts, microorganisms bloom in the receding waters (Figure 11.5). Microbial concentrations may produce very large communities forming extensive mats at the bottom of the lakes (Figure 11.6). In the early stages, the ecosystem is composed of a variety of halophilic populations (e.g., Stivaletta et al., 2009). With increasing brine concentrations, only microorganisms with osmotic capabilities can survive the salt concentration and Archea dominates (Grant et al., 1998). These microbial organisms can be entrapped inside growing crystals. Cyanobacteria form microbial mats on the floor of the lakes and ponds where water often assumes a red color (Figure 11.5) due to the presence of

Figure 11.5 Red pond waters produced by microbial activity in Chott el Rarsha (Southern Tunisia).

314

a

Lakes on Mars

b

Figure 11.6 (a) Microbial mat in a salar in the Atacama desert of Chile; (b) Newly formed biofilm over an active microbial community (Atacama desert).

halophilic Archea. An eukaryotic algae, Dunaliella, is present in dry lakes world-wide and can result in bright red coloration. Depositional cycles result in a complexity of facies patterns and produce different lithologies, differential erosion, and overgrowth of salt structures at the interface and within the sediment. Moreover, if flood periodicity is long enough, the surface may undergo sustained eolian erosion resulting in the flattening of the corrugated floor of the pan and mobilization of large amounts of clastic evaporitic sediment. The pre­ servation of sedimentary rocks is therefore a compromise between these sedimento­ logical factors and the production of mud intervals interbedded with salt layers. Sedimentary structures produced by waves can be present in both lithologies and structures suggesting dissolution of former salt accumulation. Additionally, eolian deflation may produce subtle discontinuity and sharp erosional features that still can be easily removed by the subsequent flooding event.

11.2.2 Mineral assemblages Mineral assemblages depend on the chemical composition of the water. The properties of the water are also critical and changes in eH and pH may alter the ability of some minerals to crystallize. The mineralogical system of the saline lakes, and more generally of playas, is extremely complex (Eugster and Hardie, 1978) and mineralo­ gical signatures may help us recognize analog environments on Mars. Evaporation in standing bodies of water creates concentrated brines at the surface where saline minerals nucleate and grow. Therefore, the brine concentrates in the upper part of the body of water and sinks toward the less-concentrated underlying water. This produces convection cells in the water that replaces the sinking brine at the surface, and result in its dissolution as it sinks into more diluted waters. This mechanism is repeated until the brine concentration has become homogeneous throughout the entire water column. At this stage, the salt minerals start to precipitate on the lake floor. This mechanism occurs in permanent or quasi-permanent lakes where there is enough time for the brine to concentrate. In ephemeral lakes, this

The sedimentary record of modern and ancient dry lakes

315

mechanism can be interrupted by the physical evaporation of the body of water (Hardie and Eugster, 1970). In geological terms, lakes that survive long enough to reach homogeneity in brine concentration can be considered permanent. Ephemeral playa lakes display more complex sedimentation patterns. Flood waters are enriched in salt content by the dissolution of the less soluble minerals previously deposited. As a result, brine sinkage and concentration are more efficient but strongly affected by evaporation that increases concentration. The mineralogical composition of evaporites depends on the composition of both surface and groundwater. Calcium carbonate is a major evaporitic mineral that tends to accumulate early in the evaporitic processes. Gypsum may accumulate in a subse­ quent phase followed by halite, sodium sulfate, and carbonate. Water rich in calcium and sodium will also deposit calcium carbonate, pirossite, gaylussite, and trona but sulfate-rich water will deposit glauberite, epsomite, and mirabilite. The mineralogy of arid lacustrine environments is a discipline in itself that goes beyond the focus of this chapter. One of the best-studied evaporitic systems is the Chott el Jerid in Southern Tunisia (Swezey, 1996). The term Chott is another synonym for sabkha. Chott el Jerid is underlain by a thick accumulation of Cretaceous and Tertiary (Mio-Pliocene) sedi­ ments that contain gypsum-rich ancient marine evaporites. A thin Quaternary deposit lies on top of the accumulation. This deposit consists of a wide variety of evaporites and eolian materials. The most common minerals are gypsum, halite, and also carnal­ lite (KMgCl3·6H2O). Evaporite formation is an ongoing process in Chott el Jerid. The present mean annual precipitation ranges between 80 and 140 mm and the evaporation rate is about 1500 mm, generating a strong negative water balance. A large quantity of new evaporite minerals is formed after major flooding events. An example of such events was the 1990 flood that inundated the entire basin (Bryant, 1999; Bryant et al., 1994). As water evaporation continues, sequential crystallization of different types of minerals occurs, a process that is normally controlled by chemical changes in the brine. This process results in the zonation of evaporite minerals that can be documented with remote sensing techniques (Drake et al., 1994). The zonation of evaporites is a dynamic process in space and time due to repeated crust destruction and re-formation (Bryant et al., 1994). The waters entering the basin generally have the same bulk chemistry, suggesting that they underwent the same general chemical evolution. In this case, the waters flowed through the ancient marine evaporite deposits of the region (Bryant et al., 1994). Therefore, the evaporite mineral assem­ blage of Chott el Jerid is generally consistent with that derived from the evaporation of seawater.

11.2.3 Ephemeral streams and terminal fans Playa lakes occupy the center of hydrological basins and consequently they concen­ trate both the surface and the subsurface waters. Both types of hydrological supply have a profound impact on lacustrine sedimentation. Surface water is mostly supplied by ephemeral streams (Figs. 11.1 and 11.2). These river types are usually short headed with small basins. They are episodically active with large floods, high peak discharge,

316

Lakes on Mars

and upper-flow regimes. Evaporation and infiltration play a role in depleting the stream along its course to produce a massive loss of water. Consequently, the rivers lose their capacity to transport and the channels diverge and become shallower and broader. Downstream the channels may totally fade out, spreading the remaining water over large sand and mud flats. These distributary systems were investigated for the first time by Mukerij (1976) who documented the arid regions of India. The terminal fans display a large variability of morphologies and facies (Friend, 1978; Tooth, 2000). For example, in the Sahara, and other deserts i.e. in Australia, the terminal fans differ if they debouch in mud flats or in saline lakes. The former consists of channels that gradually lose the channeled morphologies of the systems and fades out into large flat subaerial zones. The latter debouche directly into the saline lake portions of the sabkhas and do not develop long and broadening channels. Instead, they form a number of well-defined channels creating a better-defined fan-shape in plan view. The terminal fans contribute to the development of sand and mud flats and also to the supply of the central saline lakes with mud. They are major elements in the formation of the terrestrial sabkhas. They are common on Earth and have now been recognized at high resolution on Mars (see Chapter 12).

11.2.4 Spring and spring mounds One of the recharge mechanisms for dry lakes is the injection of water from the subsurface through springs, which results in the formation of spring mounds. Recognizing spring mounds on Mars would be an important step in the location of ancient lakes and would provide evidence for a sustained hydrological cycle. Mounds also provide a potential mean for biota to migrate from the subsurface to the surface and vice-versa. Recently, spring mounds have been suggested to be a major feature on the surface of Mars and spring mounds are present in a few dry lakes on Earth (Rossi et al., 2007). They are abundant in the southeastern part of Chott el Jerid (Ori et al., 2001; Roberts and Mitchell, 1987). Their morphology includes ridges, towers, pinnacles, and conical hills with or without central cones (Figure 11.7). Their heights reach can range from less than 1 m up to over 30 m and their diameter can reach up to 500 m. Most consist of sand and silt size particles cemented by varying amounts of tufa, travertine, and gypsum (Stivaletta and Barbieri, 2009). They are fed by spring water from the main aquifer—the Complex Terminal—the most important bed being the Upper Senonian (late Cretaceous) lime­ stones. Radiocarbon analyses give ages ranging from 30,000 to 3,500 years for this aquifer (Stivaletta and Barbieri, 2009) implying that recharges occurred during the pluvial periods of the Quaternary. Current recharging from sporadic rainfall is relatively limited.

11.2.5 Environmental and climatic changes Like many other extreme environments, dry lakes are extremely sensitive to the changes of the variables in their systems (Williamson et al., 2009, see also Chapter 13). Subtle changes in the volume of water or in the water table may affect its chemical composition and may produce major changes in the lacustrine environment. Climate, however, remains the most significant agent of change.

The sedimentary record of modern and ancient dry lakes

317

Figure 11.7 A mound spring in the salt flat of the Chott el Jerid (Southern Tunisia). The mound is about 10 m high and is covered by palm trees and shrubs due to the presence of water.

Climatic changes leave their signature at different scales. On Earth, these differ­ ences can be detected and analyzed, whereas it is extremely difficult to analyze or even detect these stages on Mars. High-quality 3D images and close inspection by rovers may provide data for the identification and analysis of the signature of climatic changes in lacustrine deposits. Minor climatic changes may produce modifications in the water supply and chemical composition of the waters. Sabkha environments may increase or reduce in size, or even disappear. These changes impact lacustrine facies by modifying the depositional patterns. During wet periods, increased water volume results in an enlargement of the basin size and the inundation of mud flats and sand flats. Wet periods are also associated with a progradation of terminal deltaic facies as the water volume, discharge, and flow competency are increased. If the basin’s enlargement is only due to groundwater rise, then the terminal fan system will recede along with the transgressive shoreline. Minor climatic changes impact the facies pattern but do not modify the nature of the lacustrine basins. On the other hand, major climatic changes that bring signifi­ cantly larger amount of water profoundly change the environment, stopping the evaporitic sedimentation and forming perennial lakes. Paleoterraces and ridges can be currently observed in many dry lake areas (Figure 11.8). They are the remnants of older shorelines formed during wet (pluvial) climatic conditions (e.g., last deglaciation � 18,000 year BP). One of the first examples to be recognized was Lake Bonneville in Utah (Gilbert, 1980). Many others have been identified today in the Sahara (Ori et al., 2007), Central Asia (Komatsu et al., 2001), and elsewhere. This clearly indicates that it is a common and regular pattern for the playas to become large perennial lakes (megalakes) during wetter and colder periods.

318

Lakes on Mars

a

b

Figure 11.8 (a) Terrace at the margin of the Jerid—Fajal megalake in Southern Tunisia; (b) Outcrops of the terrace—The lower part consists of wave-reworked gravels, with the cross showing relatively well-sorted, horizontal stratification, with a strata laterally continuous and well defined.

The margins of these megalakes show wave-reworked terraces (Richards and VitaFinzi, 1982, Figure 11.8), Gilbert-type deltas, and shorelines with low-angle cross bedding and wave ripples. These marginal facies of fine-grained sediment, mostly mud, transition into the basins. In many cases, these deposits are rich in organic matter and their stratigraphic columns include alternating mudstone and evaporitic deposits (Figure 11.9). By contrast, sabkhas do not show clear shoreline deposits. Their facies transition is smooth and very gradual with no major morphological expression. The drying shallow water basins of the sabkhas do not leave a clear and long-lasting signature, except when bands of evaporitic materials are preserved and are large enough to be captured from orbital imagery. However, these bands usually have fragile fabrics and subtle compositional changes and do not resist subsequent flooding events.

11.3 Sabkhas as Mars analogs The presence of abundant sulfate minerals has been confirmed by a number of recent Mars missions from both orbit and surface (e.g., Christensen et al., 2004, Gendrin et al., 2005) as well as exposures of carbonates (Ehlmann et al., 2008). Sulfate deposits appear to be extensively present in several areas and in various geological settings. It is not the aim of this chapter to provide an extensive review of the occurrence of sulfate on Mars. However, these deposits are of interest in their possible connection, at least for some of them, with sabkha analog environments on Mars.

The sedimentary record of modern and ancient dry lakes

319

Figure 11.9 Pleistocene sabkha deposits showing two horizontal evaporitic layers marked with crosses with an intervening large-scale cross-bedded set of detrital gypsum. The crossbedded unit is an eolian deposit formed probably during a climate change that produced increased arid conditions. Since then, the sabkha has completely dried out and the evaporitic sediment has been reworked by wind into barchan dunes.

On Earth, the evaporation of dry lakes is the largest producer of sulfate deposits and it can be reasonably argued here, considering what is currently known about Mars’ past climate and overall environmental conditions (Gasse, 2000; Nash 1976; Petit-Maire, 1986), that the same mechanism was at the origin of some of the sulfate deposits observed on Mars today (Gendrin et al., 2005). In some cases, deposits are contained in basins while in others they seem to spread over large open areas, consistent with several distinct depocenters separated by low-relief ridges. The existence of dry lakes on Mars consistent with the morphological and miner­ alogical record (e.g., De Hon 1992; Forsythe and Blackwelder, 1998; Cabrol and Grin, 1999, 2001, 2005; Irwin et al., 2005; Ori et al., 2000, Fassett and Head 2005) suggests that the planet experienced a large-scale hydrological cycle involving both ground and surface water. A potential candidate of sabkha on Mars is the Columbus crater in Terra Sirenum (Figure 11.1), where data support the presence of evaporitic sulfates and clays (Wray et al., 2009). Deposits are interbedded and form several meter-thick units of clays and evaporates, which is consistent with successions associated with hydro­ logical cycling induced by climate change. The high-albedo units are interpreted as evaporites whereas low-albedo units are interpreted as clays. The clays could be the result of either deposition in a permanent lacustrine environment or deposition in a mud flat. In the former case, the clays would represent wetter and colder conditions and in the latter case clays would represent dryer and warmer climatic conditions. Other features, like spring mounds (Figure 11.8), could have analogs on Mars (Carlton and Oehler, 2008) but more data will be necessary to be

320

Lakes on Mars

conclusive. In particular, while morphologic, geologic, and mineralogic evidence have already been recognized at megascale, facies, laminations, and stratigraphic succes­ sion still lack to definitely link these features to ancient lakes, whether perennial or ephemeral. This evidence will be brought on by future ground missions.

References Braitsch, O., 1971. Salt Deposits; Their Origin and Composition. Springer Verlag, New York, 237 pp. Bryant, R.G., 1999. Application of AVHRR to monitoring a climatically sensitive playa. Case study: Chott el Jerid, southern Tunisia. Earth Surface Processes and Landforms, 24, 283–302. Bryant, R.G., Drake, N.A., Millington, A.C., Sellwood, B.W., 1994. The chemical evolution of the brines of Chott el Jerid, southern Tunisia, after an exceptional rainfall event in January 1990. In: Sedimentology and Geochemistry of Modern and Ancient Saline Lakes, Tulsa, OK, SEPM Special Publication No. 50, pp. 3–12. Cabrol, N.A., Grin, E.A., 1999. Distribution, classification and ages of Martian impact crater lakes. Icarus, 142, 160–172. Cabrol, N.A., Grin, E.A., 2001. The evolution of lacustrine environments on early Mars: Is Mars only hydrologically dormant? Icarus, 149, 291–328. Cabrol, N.A., Grin, E.A., 2005. Ancient and recent lakes on mars. In: Tokano, T. (Ed.), Water on Mars and Life, Springer, Berlin, pp. 235–259. Carlton, C.C., Oehler, D.Z., 2008. A case for ancient Springs in Arabia Terra, Mars. Astrobiology, 8, 1093–1112. Christensen, P.R., Wyatt, M.B., Glotch, T.D., Rogers, A.D., Anwar, S., Arvidson, R.E., et al., 2004. Mineralogy at Meridiani Planum from the Mini-TES experiment on the opportunity Rover. Science, 306, 1733–1739. De Hon, R.A., 1992. Martian Lake basins and lacustrine plains. Earth Moon and Planets, 56, 95–122. Drake, N.A., Bryant, R.G., Millington, A.C., Townshend, J.R.G., 1994. Playa sedimentology and geomorphology: Mixture modeling applied to Landsat thematic mapper data of Chott el Jerid, Tunisia. In: Sedimentology and Geochemistry of Modern and Ancient Saline Lakes, Tulsa, OK, SEPM Special Publication No. 50, pp. 125–131. Ehlmann, B.L., Mustard, J.F., Murchie, S.L., Poulet, F., Bishop, J.L., Brown, A.J., et al., 2008. Orbital identification of Carbonate-Bearing Rocks on Mars. Science 322 (5909), 1828–1832, DOI: 10.1126/science.1164759. Eugster, H.P., Hardie, L.A., 1978. Saline lakes. In: Lerman, A. (Eds.), Lakes: Chemistry, Geology, Physics. Springer, New York, pp. 237–293. Fassett, C.I., Head, J.W. III, 2005. Fluvial sedimentary deposits on Mars: Ancient deltas in a crater lake in the Nili Fossae region. Geophys. Res. Lett. 32, L14201, doi:10.1029/ 2005GL023456. Forsythe, R.D., Blackwelder, C.R., 1998. Closed drainage crater basins of the Martian highlands: Constraints on the early Martian hydrologic cycle. J. Geophys. Res. 103 (E13), 31, 421–431. Friend, P.F., 1978. Distinctive features of some ancient river systems. In: Miall, A.D. (Ed.), Fluvial Sedimentology. vol. 5, Canadian Society of Petroleum Geologists Memoir, Alberta, CAN, pp. 531–542.

The sedimentary record of modern and ancient dry lakes

321

Gasse, F., 2000. Hydrological changes in he African tropics since the last glacial maximum. Quatern. Sci. Rev., 19, 189–211. Gendrin, A., Mangold, N., Bibring, J.-P., Langevin, Y., Gondet, B., Poulet, F., et al., 2005. Sulfates in Martian layered terrains: the OMEGA/Mars Express view. Science, 307, 1587–1591. Gilbert, G.K., 1980. Lake Bonneville, US Geological Servey, Monograph Series, 1, 457 pp. Grant, W.D., Gemmel, R.T., McGenity, T.J., 1998. Halophiles. In: Horikoshi, K.,Grant, W.D. (Eds.), Extremophiles: Microbial Life in Extreme Environments. Wiley-Liss, Inc, New York, pp. 93–132. Grotzinger, J.P., Arvidson, R.E., Bell, J.F. III, Calvin, W., Clark, B.C., Fike, D.A., et al., 2005. Stratigraphy and sedimentology of a dry to wet eolian depositional system, Burns forma­ tion, Meridiani Planum, Mars. Earth. Planet. Sci. Lett. 240 (1), 11–72. Hardie, L.A., 1985. Evaporites: marine or non-marine?. Am. Jour. Sci., 285, 667–672. Hardie, L.A., 1991. On the significante of evaporites. Annu. Rev. Earth Planet. Sci., 290, 43–106. Hardie, L.A., Eugster, H.P., 1970. The evolution of closed basin brine. Geol. Soc. Am. Spec. Publ., 3, 273–290. Irwin III, R.P., Howard, A.D., Craddock, R.A., Moore, J.M., 2005. An intense terminal epoch of widespread fluvial activity on early Mars: 2. Increased runoff and paleolake development. Journal of Geophysical Research E: Planets, 110, 1–38. Komatsu, G., Brantingham, P.J., Olsen, W.J., Baker, V.R., 2001. Paleoshoreline geomorphology of Boon Tsagaan Nuur, Tsagaan Nuur and Orog Nuur: the Valley of Lakes, Mongolia. Geomorphology, 39/3–4, 83–98. Mukerij, A.B., 1976. Terminal fans on inland streams in Sutlej-Yamuna Plain, India, Zeit. Fur Geomorph., 20, 190-204. Nash, D.J., 1976. Greoundwater as geomorphological agent in dryland. In: Thomas, D.S.G. (Ed.), Arid Zone Geomorphology: Process, Form and Change in Dryland. Wiley, Chichester, pp. 321–348. Ori, G.G., Di Achille, G., Komatsu, G., Marinangeli, L., Rossi, A.P., 2007. River morphologies and plaeodrainage of western africa (sahara and sahel) during humid climatic conditions. In: Nichols, G., Williams, E., Paola, C. (Eds.), Processes, Environments and Basins, International Association Sedimentologists, Special publication. vol. 38. Blackwell pub­ lishing, Oxford, pp. 519–523. Ori, G.G., Komatsu, G., Marinangeli, L., 2001. Exploring mars surface and its Terrestrial analogues: filed trip guide book Chott el rahars and Chott el Jerid, Alenia Spazio, Tozeur (Tunisia), 25–27 September 2001, Alenia, Spazio, 95 pp. Ori, G.G., Marinangeli, L., Komatsu, G., 2000. Martian Paleolacustrine environments and their geological constrains on drilling operations for exobiological research. Planet. Space Sci., 48, 1027–1043. Petit-Maire, N., 1986. Paleoclimates in the Sahara of Mali a multidisciplinary study. Episodes, 9, 7–16. Richards, G.W., Vita-Finzi, C., 1982. Marine deposit 35,000–25,000 years old in the Chott Jerid, Southern Tunisia. Nature, 295, 54–55. Roberts, C.R., Mitchell, C.W., 1987. Spring mounds in southern tunisia. In: Desert Sediments: Ancient and Modern, Geological Society Special Publication No. 35. Geological Society, London, pp. 321–334. Rossi, A.P., Neukum, G., Pondrelli, M., Zegers, T., Mason, P., Hauber, E., et al. 2007. The case for large-scale spring depositson Mars: light-toned deposits in crater bulges, VallesMar­ ineris and chaos. Lunar and Planetary Science Conference, 38, 1549.

322

Lakes on Mars

Schreiber B.C. (Ed.), 1988. Evaporites and Hydrocarbons. Columbia University Press, New York, 235 pp. Sonnenfeld, P., 1984. Brines and Evaporites. Academic Press, New York, 385 pp. Stivaletta, N., Barbieri, R., 2009. Endolithic microorganisms from spring mount evaporite deposits (Southern Tunisia). J. Arid Environ., 73, 33–39. Stivaletta, N., Barbieri, R., Picard, C., Bosco, M., 2009. Astrobiological significance of the sabkha life and environments of southern Tunisia, In: Flamini, E., Ori, G.G., Osinski, F., Di Pippo, S. (Eds.), Mars Analogues, Planetary Space Science Special Issue. Swezey, C.S., 1996. Structural controls on Quaternary depocentres within the Chotts Trough region of southern Tunisia. J. Afr. Earth Sci. 22 (3), 335–347. Tooth, S., 2000. Process, form and change in dryland rivers: a review of recent research. Earth Sci. Rev., 51, 67–107. Warren, J.K., 1999. Evaporites: Heir Evolution and Economics. Blackwell, Oxford, 345 pp. Williamson, C.E., Dodds, W., Kratz, T.K., Palmer, M.A., 2009. Lakes and streams as sentinels of environmental change in terrestrial and atmospheric processes. Front. Ecol. Environ., 6, 247–254. Wray, J.J., Milliken, R.E., Swayze, G.A., Dundas, C.M., Bishop, J.L., Murchie, S.L., et al., 2009. Columbus Crater and other possible plaeolakes in Terra Sirenum, Mars, Lunar and Planetary Science Conference, 40, 1896.

12 Aqueous depositional settings in Holden crater, Mars

John A. Grant, Rossman P. Irwin, III, and Sharon A. Wilson Center for Earth and Planetary Studies, National Air and Space Museum, Smithsonian Institution, Washington, DC, USA

12.1

Introduction

Holden crater in southwestern Margaritifer Terra (approximately 26°S, 326°E; Figure 12.1) likely contained two distinct lakes of differing character and sources during the Late Noachian on Mars. The first lake may have included distal alluvial deposits, was longer-lived, and formed when drainage from within the crater ponded on the crater floor. The second lake was shorter-lived and created when water impounded in Uzboi Vallis breached Holden’s rim and rapidly drained into the crater. Megabreccia deposits in the crater walls expose what may be even older aqueously deposited sediments. Holden crater is 154 km in diameter and was formed near the transition from the Noachian to the Hesperian Epoch (Grant et al., 2008; Pondrelli et al., 2005; Scott and Tanaka, 1986). The crater interrupts the middle reaches of the Uzboi–Ladon–Morava (ULM) meso-scale outflow system that dominates the ancient drainage system in southwest Margaritifer Terra. The formation of Holden crater blocked Uzboi Vallis, with the resultant crater rim reaching some 900 m above the valley floor. The eventual breaching of Holden’s rim demonstrates that the ULM system was active at least episodically until at least the Late Noachian. The juxtaposition of the crater and valley set the stage for formation of multiple lakes of varying extent and duration within both. Both Holden crater and Uzboi Vallis (Figure 12.1) preserve bedded sedimentary materials. In Holden, the deposits are best exposed across the southwestern floor, but images from the eastern and south central crater floor and vicinity of the central peak of the crater reveal layered outcrops there as well (Figure 12.2). The layered deposits in Uzboi Vallis are less well mapped, because most high-resolution data targets the lower reaches just above the entry breach in Holden’s rim. Hence, much of Uzboi Vallis is less well characterized and will not be discussed further. Photogeologic interpretation of images ranging from about 1 to >100 m/pixel resolutions suggested that light-toned, layered sediments in Holden crater are lacus­ trine (Grant and Parker, 2002; Irwin et al., 2005a; Malin and Edgett, 2000; Parker, 1985; Pondrelli et al., 2005) or airfall deposits (Malin and Edgett, 2000), perhaps with late glacial activity (Pondrelli et al., 2005). A conclusive statement regarding which Lakes on Mars. DOI: 10.1016/B978-0-444-52854-4.00012-X © 2010 Elsevier B.V. All rights reserved.

324

Lakes on Mars

Figure 12.1 (A) Regional map of Margaritifer Terra and surroundings near the eastern end of Valles Marineris showing place names mentioned throughout the text, including the ULM meso-scale outflow system. The ULM system alternately incises and fills as it crosses between and into ancient multi-ringed impact basins, respectively. (B) The �150 km diameter Holden crater that includes the breach from Uzboi Vallis in the southwest rim, a variety of sedimentary beds related to multiple lakes that probably occupied much of the crater floor, and several large fractures crossing the crater floor. Lakes likely occurred in Holden crater and Uzboi Vallis during the Late Noachian Epoch. (A) derived from topographic data from the Mars Orbiter Laser Altimeter (MOLA) over a subset of THEMIS daytime infrared mosaic for the region (230 m/pixel; black areas are gores in coverage). Black box in (A) shows location of (B). (B) derived from thermal inertia data over a subset of the THEMIS daytime infrared mosaic for Holden crater (Fergason et al., 2006). Black box in (B) shows the approximate location of Figure 12.2.

process was responsible for the stratigraphy was precluded prior to the availability of data from MRO because diagnostic characteristics could not be resolved from the available datasets that preceded that mission. More recent work by Grant et al. (2008), based on interpretation of images from HiRISE (McEwen et al., 2007) at �26–52 cm/pixel scales together with data from CRISM (Murchie et al., 2007) details the submeter sedimentary stratigraphy exposed

Aqueous depositional settings in Holden crater, Mars

325

Figure 12.2 MOC narrow angle and THEMIS VIS mosaic of southwestern Holden crater, showing the extensive outcrops of light-toned layered materials (see Figure 12.1B for context). Locations identified as Holden breach fan deposits refer to deposits associated with breaching of the crater rim rather than coalescing deposits flanking lower portions of the south and west wall. Individual HiRISE images are overlain, 6 km wide, and their locations are outlined in black. The approximate locations of Figures 12.3, 12.4, 12.6–12.9, and 12.11 are shown for reference. HiRISE images PSP_003433_1540 is located �40 km to the northeast and crosses a portion of the crater central peak complex and reveals layers and an eroded bench. HiRISE image PSP_002444_1535 is located �70 km to the east-northeast near the eastern edge of the crater floor and exposes what appear to be beds of the upper member of the lower unit. HiRISE image ESP_011542_1530 is located �20 km to the east on the south central floor of the crater and crosses a portion of the floor fracture system where it exposes apparent beds of at least the upper member of the lower unit. See text for discussion.

in Holden and confirms that most of it is likely related to deposition in a distal alluvial or lacustrine setting. Broadly comparable stratigraphy on the floor of Uzboi Vallis provides evidence for water impoundment that breached the 900-m-high rim of Holden and was the source of a second, shorter-lived lacustrine system in the crater. The Late Noachian stratigraphy within Holden crater and Uzboi Vallis records an important chapter in the aqueous history of Mars, when conditions conducive to widespread valley formation and an active hydrologic cycle were ending (e.g., Carr, 2006; Fassett and Head, 2008; Grant, 2000; Grant and Parker, 2002; Grant et al., 2009; Howard et al., 2005). As such, Holden deposits may preserve evidence of habitable conditions at the end of an early wetter period on Mars (Bibring et al., 2006).

326

Lakes on Mars

12.2 Geomorphic setting The ULM system (Figure 12.1) apparently emerges from Argyre basin and traverses northward along the southwestern flank of the Chryse trough in Margaritifer Terra (Baker, 1982; Phillips et al., 2001; Saunders, 1979). However, the formation of Hale crater on the northern margin of Argyre partly buried what appears to be a direct connection with the ULM, complicating the interpretation that the system headed farther to the south. Nevertheless, the ULM system is at nearly full width where it emerges to the north of Hale, and there are few major tributaries along this reach, implying that the system was fed from overflow of Argyre basin. Several large valley systems heading even farther to the south likely fed the ULM drainage (Parker, 1985, 1994), resulting in a total watershed potentially covering more than �11�106 km2, or about 9% of Mars (Banerdt, 2000; Phillips et al., 2001). The ULM system is characterized by deeply incised trunk segments of 15–20 km width, separated by depositional plains that partially fill the Early Noachian Holden and Ladon multiringed impact basins whose inner ring massif diameters are approximately 300 km and 470 km, respectively (Frey et al., 2003; Saunders, 1979; Schultz et al., 1982). From the incised margin of Argyre, the ULM outflow system descends more than 1800 m in a series of longitudinal steps from south to north. Although any outlet from Argyre is partly obscured by Hale crater and its ejecta, a segment is visible to the north of the crater and south of Bond crater, which also disrupts the ULM. The north rim of Bond crater is at 120 m elevation and is well above the confluence of the Nirgal Vallis tributary with Uzboi Vallis, more than halfway between Argyre and Holden basins at an elevation of around –760 m. Uzboi Vallis then drops to approximately –1275 m to the south of Holden crater before descending (prior to interruption by the Holden crater impact) to the floor of the multi-ringed Holden basin at an elevation of –1700 m to –1800 m (Irwin and Grant, 2009). The ULM system continued into Ladon basin to the northeast, dropping to –2000 m along Ladon Valles (Figure 12.1). The system finally drops an additional 480 m through Morava Valles to the floor of Margaritifer basin, where some of this water may have been stored and later released to contribute to the incision of Ares Vallis (Carr, 1979; Carr and Clow, 1981; Grant and Parker, 2002; Grant et al., 2009; Rotto and Tanaka, 1995). The floor of Ladon basin (Figure 12.1) is relatively flat (maximum relief of 0.27 km over 350 km), and preserves evidence of extensive, finely layered deposits as seen in HiRISE images, further suggesting that infilling may have occurred in standing water and was associated with multiple depositional events along the ULM drainage system. Holden crater crosscuts the inner massif ring of the older Holden basin, and the floor of the crater lies at –2300 m, making it the lowest surface of any appreciable extent for approximately 700 km in any direction. Formation of Holden crater inter­ rupted the ULM outflow system (Grant et al., 2008; Pondrelli et al., 2005), but eventually the rim was breached by Uzboi Vallis, and the crater floor became the terminal basin for Uzboi and Nirgal Valles (Grant and Parker, 2002; Grant et al., 2008; Irwin and Grant, 2009; Irwin et al., 2005a; Pondrelli et al., 2005). Although there is no outlet from Holden crater, a region of collapse outside and to the east of the crater and

Aqueous depositional settings in Holden crater, Mars

327

slightly above its floor appears to post-date formation of the crater and may have provided a late source of water to Ladon Valles. Therefore, although the bulk of the through-flowing drainage occurred during the Late Noachian Epoch, some segments may have seen late discharge in the Early Hesperian Epoch (Grant et al., 2008; Rotto and Tanaka, 1995) that was contemporaneous with late fluvial activity elsewhere, at least in Margaritifer Terra (Grant et al., 2009). The crater walls within Holden are partly fringed by coalescing alluvial fans (Figure 12.2) that form a broad bajada along much of the western side of the crater (Moore and Howard, 2005). The Late Noachian fluvial erosion and deposition that was responsible for these fans formed deep alcoves and gullies primarily in the higherstanding western half of the crater rim (Irwin and Grant, 2009). The less dissected eastern rim is �1 km lower than the southern and western segments, and the greater extent of fans to the west may reflect larger collection areas associated with basins extending to higher elevations and/or an orographic enhancement of precipitation and runoff that created the fans. Moreover, the lack of dissection of alluvial fan surfaces (Moore and Howard, 2005) suggests that direct precipitation onto the crater floor, as opposed to the high-standing walls, was limited at least during terminal stages of activity. The only sizeable inlet to Holden crater, where Uzboi Vallis breaches the southwestern wall, post-dates much of the deposition that occurred on the crater floor. Hence, deposition of the sediments predating the Uzboi breach was associated with erosion of materials from the crater walls and did not include appreciable waterborne sediment derived and transported from outside the crater.

12.3

Geologic history

Geological mapping in Margaritifer Terra constrains the general timing of activity along the ULM outflow system with respect to regional geomorphic events (Grant, 1987, 2000; Grant and Parker, 2002; Grant et al., 2008, 2009; Parker, 1985, 1994; Pondrelli et al., 2005; Saunders, 1979; Scott and Tanaka, 1986). This mapping includes crater counting and assessment of crosscutting relationships to establish the relative age of various features. As summarized from Grant (1987) and Grant et al. (2009), the Early Noachian, degraded Holden basin and Ladon basin multi-ringed impact structures (Schultz and Glicken, 1979; Schultz et al., 1982) are the oldest features crossed by the ULM system. These basins imparted considerable structural and topographic influence on the course of the ULM drainage, with incised segments turning to become radial to depositional basin centers (Figure 12.1). Ladon Valles is also diverted from its radial entry into Ladon basin by the topography of the inner basin ring, and both inner and outer rings of these basins influenced valley network paths. Formation of these ancient impact basins was followed by evolution of the diverse cratered upland surface that included three general resurfacing events hypothesized to have occurred between the Early and Late Noachian Epochs. Surfaces created during all three resurfacing events are incised by the ULM system and its tributaries. By contrast, a fourth more localized resurfacing event of Early-to-Mid-Hesperian

328

Lakes on Mars

age is hypothesized, emplacing materials that embay most ULM tributaries (Grant, 1987). Hence, major incision of the ULM system and its tributaries occurred between the Middle and Late Noachian Epochs and had largely ended by the Early Hesperian (Irwin and Grant, 2009). It is possible that the ULM system could have been formed by waxing and waning discharge during a single prolonged event, but several erosional and depositional forms imply contributions from multiple large discharge events (Irwin and Grant, 2009). The Holden and Ladon basins have anabranching outlet valleys, which likely required filling and overflowing the large basins. The hanging relationships between side channels and the main stem suggests that multiple overflow points remained active until the central one was incised deeply enough to confine the entire flow. In support of this interpretation, a number of possible terraces are found along Uzboi Vallis, and at least five distinct terraces occur along Ladon Valles (Boothroyd, 1983; Grant, 1987; Grant and Parker, 2002; Parker, 1985). Discharge estimates along Ladon Valles, based on the elevation of terraces combined with channel cross-section and gradient, vary between 150,000 m3 s−1 and 450,000 m3 s−1 (Grant and Parker, 2002), which is 5–10 times higher than terrestrial discharge rates from the Mississippi River (Komar, 1979) and on the lower end of discharge rates from the Channeled Scabland (Baker, 1982; Baker and Nummedal, 1978). Such numbers require significant discharge associated with the evolution of the system.

12.4 Holden crater stratigraphy Holden crater is partially filled by an impressive sequence of more than 150 m of finely bedded, sedimentary deposits that are best exposed in the southwestern portion of the crater but are also observed in the central and eastern portions of the crater (Figure 12.2). HiRISE images reveal that bedded sedimentary rocks overlie megabreccia in the walls (Figure 12.3) and occur in two distinct stratigraphic units (Grant et al., 2008). The bedded rocks consist of a lower light-toned unit displaying mostly flat-lying, meter- to submeter-scale layers that are variably continuous, phyllosilicate-bearing, and asso­ ciated with alluvial fans (Moore and Howard, 2005). The lower unit is overlain by an upper darker-toned unit that is associated with the Uzboi Vallis breach. The upper unit exhibits distinct alluvial morphology including forms radial to the Uzboi breach that grade to more flat-lying beds further out into the basin (Grant et al., 2008). The sedimentary beds comprising the lower and upper units exposed in Holden record two intervals of ponding within the crater (Grant et al., 2008). The first of these was related to water derived from the crater walls and/or groundwater, resulting in deposition of the lower unit. The second related to flooding when water impounded in Uzboi Vallis overtopped and breached the crater rim (Figure 12.1).

12.4.1 Megabreccia in Holden crater HiRISE images of outcrops forming the southern wall of Holden crater reveal large blocks within a finer matrix, which are features that correspond to impact megabreccia

Aqueous depositional settings in Holden crater, Mars

329

Figure 12.3 One of several outcrops of megabreccia exposed in the southern wall of Holden crater near 26ºS, 325.3ºE (see Figure 12.2 for context). Although most blocks are darker than the surrounding wall materials, some (not in this image) are lighter toned and/or stand in positive relief (Grant et al., 2008). HiRISE image PSP_001666_1530_RED, 26 cm/pixel scale.

(Figure 12.3; Grant et al., 2008). Thermal inertias on the surfaces with exposed blocks approach 1200 J m−2 K−1 s−1/2 and are consistent with exposures of bedrock (Fergason et al., 2006) rather than unconsolidated talus material eroded from the wall. Megabreccia blocks are variably rounded, poorly sorted, and chaotically arranged. Indivi­ dual blocks are up to 50 m across and occur as distinct clasts that are typically relatively dark-toned, but they are occasionally light-toned relative to the matrix. Most blocks appear easily eroded in comparison to adjacent wall materials (Figure 12.3), but some are more resistant and stand in positive relief. On Earth, impact megabreccias of comparable scale occur in the walls of Popigai crater, Russia (Vishnevsky and Montanari, 1999) and are impact-fragmented target rocks (Grieve et al., 1977) that are buried beneath younger crater-filling deposits (Melosh, 1989). Although the scale of the megabreccia blocks in Holden is small relative to the pixel scale of data from the CRISM instrument on the MRO spacecraft, phyllosilicates appear to occur on or within portions of the crater wall rocks (Milliken et al., 2007) that may correspond to regions of megabreccia. Thus, it is possible that some of the megabreccia exposed in Holden’s walls may be derived from ancient ULM deposits because the impact that formed Holden crater interrupted the ULM system and excavated materials from the pre-existing distal Uzboi Vallis or Holden basin. Although no obvious evidence of sedimentary layering is observed, if derived from Holden basin, these blocks and surrounding matrix may represent the oldest (either intact or derivative of) water-transported materials exposed in Holden crater. It is also possible that any phyllosilicates in the megabreccia were formed as hot fluids circu­ lated through the rocks following formation of Holden crater (Newsom et al., 1996), or they may be derived from post-depositional weathering of the surface while the alluvial fans were forming.

330

Lakes on Mars

12.4.2 Lower unit stratigraphy The oldest bedded sedimentary rocks in Holden crater form a lower unit that overlies the crater walls and floor and is associated with the fringing bajada of coalesced alluvial fans. The lower unit rocks are incised by channels associated with the Uzboi Vallis rim breach and thus predate the flood deposits. Because the crater has no other inlet or outlet channel, the lower unit sediments must have been emplaced by internally derived drainage from the crater walls via the alluvial fan distributary channels (Malin and Edgett, 2000; Moore and Howard, 2005). The sedimentary rocks comprising the lower unit (Figure 12.4) can be divided into three members on the basis of varying albedo, phyllosilicate content, bed thickness, and lateral extent of individual beds (Grant et al., 2008). Detailed mapping by Pondrelli et al. (2005) reveals that the lower unit rocks reach a common upper elevation of –1960 m in the southwestern portion of the crater. The lower member of the lower unit is relatively dark-toned, and outcrops are limited to the deepest sections of the exposed sequence; the full thickness of the lower

A

nt le ab ista Fri Res

C

Dark-toned aeolian material

Uzboi breach Upper unit Dark toned, crudely layered 1 – 100 m blocks of lower unit Lower unit, upper member Capped by thin, dark polygonal layer Lighter in tone, thinner beds Lens of dark-toned coarse material Lower unit, middle member Lighter in tone, fine-grained, horizontal, ~1 m-thick beds

75 m

B

Lower unit, lower member Fine-grained, horizontal, ~1 m-thick beds

Basal impact materials Poorly sorted blocks up to few tens of meters in diameter Intermediate- to light-toned blocks frequently in negative relief

Unit contact relationships 100 m

Erosional unconfirmity Depositional contact

Figure 12.4 (A) Upper and middle members of the lower unit emplaced in Holden crater. (B) Middle and lower members (above and below dashed line, respectively) of the lower unit, with the lower member displaying the strongest phyllosilicate signatures. Lower unit beds are unconformably overlain by an upper unit (capping deposits in (A)) associated with flooding from Uzboi Vallis via a breach in Holden crater’s rim. (C) Idealized stratigraphic section for Holden crater. Sedimentary section in (C) is �150 m thick but is a minimum estimate due to the unknown thickness of the lower member of the lower unit. Stratigraphic positions of units in (A) and (B) are indicated in (C) by connecting lines. North is up in (A) and down in (B). HiRISE image PSP_001468_1535_RED (see Figure 12.2 for context) with an image scale of 26 cm/ pixel was used for (A) and (B). Figure reproduced from Grant et al. (2008).

Aqueous depositional settings in Holden crater, Mars

331

member is unknown, because the underlying floor of the crater is not exposed. Nevertheless, the strata appear to be horizontal or nearly so, and individual meterscale beds are of constant thickness and can be traced for hundreds of meters (Grant et al., 2008). Outcrops are typically characterized by extensive conjugate fractures. Rocks comprising the middle member appear flat-lying and conformable with the lower member rocks, but they are lighter-toned. Individual middle member beds vary in their expression (Figure 12.4), as the meter-scale beds of uniform thickness are traceable for hundreds of meters in some locations, but they appear more massive and difficult to trace laterally where the surface expression of some outcrops has been modified by erosion. Hence, the extent of individual beds is difficult to establish, despite the greater extent of outcrops relative to the lower member. Like the lower member, middle member beds are characterized by extensive, conjugate fractures. The ease with which erosion masks bed expression in the middle member and the expres­ sion of both lower and middle member beds as sloping rather than vertical outcrops implies that the constituent rocks are relatively weak (Grant et al., 2008). Upper member rocks are distinct from the lower and middle members in several ways. First, individual beds are often less than a meter in thickness and can be traced for kilometers (Figure 12.4). Beds are comparably light-toned to those in the middle member, but they are often separated by lensoidal, meter-scale accumulations of darker material that sometimes occurs as discrete blocks (Figure 12.4). Unlike the middle and lower members, outcrops of the upper member are often expressed as cliffs and are likely composed of rocks that are more resistant to erosion than those lower in the section. Although fracturing of the upper member appears less extensive than in middle and lower members, its upper surface is capped by a thin (�1 m thick), dark-toned layer that is often disrupted by 4–5 m-diameter polygonal fractures (Figure 12.4) that sometimes extend meters into the subsurface, based on their expression in outcrops. Like the middle and lower members, upper member beds exhibit a fairly constant thickness and appear conformable with underlying middle member strata. There is a single location, however, where middle member beds may be truncated by overlying upper member layers. Upper member beds occur near the top of the sequence exposed in the crater and outcrop more widely than the underlying middle and lower members. Hence, the distribution of outcrops may better reflect the actual extent of the lower unit. Although outcrops are relatively rare outside of the southwestern portion of the crater, lighttoned layers are visible near fractures on the southern crater floor and the eastern crater wall (Figure 12.2) and appear similar to, at least, the upper member beds. In addition, images from the center of the crater (Figure 12.5) indicate widespread finescale layers that may be analogous to the upper member or could be related to other upper unit beds. In either case, the beds are easily deflated, suggesting that sediments comprising the beds are finer toward the basin center (Figure 12.5). The occurrence of these widely dispersed materials suggests that the lower unit occurs across much of the crater floor. CRISM data indicate that phyllosilicates occur in the lower unit, and their signatures generally decrease from the lower member to the upper member (Milliken et al., 2007, 2008). Spectral signatures of the phyllosilicates are consistent with Fe/Mg-bearing

332

Lakes on Mars

Figure 12.5 HiRISE image from the vicinity of the central peak of Holden crater near 25.9ºS, 326ºE. (A) Fine-scale bedding is apparent and (B) is susceptible to deflation, suggesting that the beds are composed of fairly uniform, fine-grained sediments. Positive-relief features (B) are likely inverted topography created by deflation of fine-grained material from around the coarse deposits associated with formation of small impact craters. HiRISE image PSP_003433_1540, 26 cm/pixel scale. North is toward the top of image.

smectites, but they could also represent a mixture of smectite and chlorite (Milliken et al., 2007). If smectite, the materials may be saponite or nontronite. Absolute clay abundance is difficult to constrain, but the detections by the CRISM instrument likely require at least �5% by weight in all three members (Grant et al., 2008; Milliken et al., 2007). A meter-scale DEM derived from HiRISE stereo pairs of a portion of the lower unit beds in Holden crater provides insight into their orientation. The DEM can be used to help determine whether the beds are flat-lying or possess an eastward dip that might suggest a depositional association with fans to the west. For example, upper member beds appear relatively continuous and flat-lying by comparison with the scale of distributaries on the adjacent bajada to the west which are also similar to what is observed in eroded fan fronts to the south (Figure 12.6). The DEM confirms that upper member beds, which are the easiest to trace, can rise or fall �5 m in elevation over distances of hundreds of meters (Figure 12.7). In addition, it is possible that the

Aqueous depositional settings in Holden crater, Mars

333

Figure 12.6 (A) Relict distributary channels crossing from west-to-east across the bajada fringing the western wall of the crater near 26.6ºS, 352.2ºE (see Figure 12.2 for context). In contrast to upper member beds, which are fairly flat-lying and can be traced for kilometers, the spacing of distributaries (white arrows) is measured in hundreds of meters. HiRISE images PSP_001468_1535_RED. (B) Alluvium and underlying upper member beds of the lower unit exposed in an eroded fan front at the base of the southern crater wall near 27ºS, 325.5ºE (see Figure 12.2 for context). Bedding in the fan alluvium (black arrow) varies in orientation over tens to hundreds of meters. HiRISE images PSP_001468_1535_RED and PSP_003077_1530_RED (A) and (B), respectively, and both have a pixel scale of 26 cm with north toward the top of the images.

Figure 12.7 DEM of the lower unit exposed in SW Holden crater (see Figure 12.2 for context) from HiRISE images PSP_001468_1535_RED and PSP_002154_1468_RED with 26 cm/pixel scale. Some beds are not horizontal and cross �5 m of relief over hundreds of meters. Elevations decline from yellow to blue, green, and then red contours from east to west at this escarpment. Contour interval is 5 m, and north is toward the top. DEM provided by O. Aharonson.

334

Lakes on Mars

beds decrease slightly in elevation from west to east, implying a dip of a degree or two toward the crater center. Such variability in bed orientation may be real and provide insight into the responsible depositional process (e.g., alluvial), or it may relate to inherent limitations of the DEM and/or deformation of the beds after deposition. A series of large fractures crosses the crater floor to the east of the DEM (Figure 12.2) and cuts deposits associated with both the upper and lower units. Examination of coarser-resolution MOLA data indicates that surfaces adjacent to these fractures rise tens of meters above surfaces �5–10 km away. Although HiRISE image coverage of the fractures does not permit correlation of detected beds across the fractures, which might confirm faulting, surfaces adjacent to the fractures appear otherwise similar to those farther away. This characteristic implies that the relief bounding the fractures formed without significant differential erosion and is the result of uplift. The presence of fractures and apparent uplift relatively near the beds viewed in the DEM does not rule out the possibility that the current bed orientation reflects the depositional process. It seems just as likely, however, that the beds were originally flat and were subsequently slightly deformed.

12.4.3 Upper unit stratigraphy The expression of the upper unit differs considerably from that of all three lower unit members and likely reflects differences in depositional setting. The upper unit uncon­ formably overlies the lower unit, and the contact between the two units crosses considerable relief where the lower unit has been eroded into a series of broad troughs and channels approximately radial to the Uzboi Vallis breach. Numerous lower unit beds are truncated at the contact with the upper unit at these locations (Figure 12.8), and at some sites, large blocks of the lower unit may have slumped and are entrained within the upper unit. Large bedforms that incorporate meter-scale blocks

Figure 12.8 Angular unconformity (white arrow) between the upper and lower units in SW Holden crater due to scour during emplacement of the upper unit. The upper unit entrains large blocks of the lower unit (black arrows), which are often deposited in the lee of flow obstacles. HiRISE image PSP_001468_1535 (see Figure 12.2 for context) with north toward the left.

Aqueous depositional settings in Holden crater, Mars

335

Figure 12.9 Bedforms on the upper unit that are distributed radially to the Uzboi breach (crests are orthogonal to breach) in the wall of Holden crater. The bedforms incorporate meter-scale blocks (inset from location of small black arrow), likely requiring an energetic process for their formation. HiRISE image PSP_003644_1530_RED (see Figure 12.2 for context). North is toward the bottom of the image.

(Figure 12.9) mark the surface of the upper unit in some locations. These bedforms are located on a positive-relief deposit rather than in a confined channel, rise meters above adjacent surfaces (Grant and Parker, 2002), are separated from one another by tens of meters, and are oriented radially to the Uzboi Vallis breach (crests are orthogonal to the breach). Erosion and transport of blocks tens of meters across and deposition of large-scale bedforms that incorporate meter-scale rocks require that the process responsible for emplacing the upper unit was highly energetic. Upper unit deposits (Figures 12.4, 12.8, and 12.9) are topographically confined below –2060 m (Pondrelli et al., 2005) and are locally tens of meters thick. Closer to the Uzboi Vallis breach in the southwestern crater wall, individual beds can be traced for tens of meters before truncating one another. Layers higher in the section and farther from the Uzboi breach are more continuous laterally and display less obvious truncating relationships. In at least one location, the upper unit forms a large fan delta that is exposed in cross-section by a large fracture (Holden breach fan deposits in Figure 12.2). In that location, the fan consists of a package of crudely bedded material containing numerous meter-scale blocks and rising tens of meters above the level of surrounding surfaces. The fan delta is capped by more continuously bedded and finergrained (block-free) materials that imply a transition to a less energetic depositional setting over time. Because these flat-lying beds are well above the level of the surrounding surface, they cannot be associated with late deposition from fans to the west and south, and were associated with the late stages of emplacement of the upper unit. The spectral signatures from CRISM and THEMIS data of the upper unit indicate abundant mafic materials and only minor abundances of phyllosilicates (Glotch, 2006; Milliken et al., 2007).

336

Lakes on Mars

12.5 Origin of stratigraphy in Holden crater 12.5.1 Origin of lower unit stratigraphy Various processes have been proposed to account for the origin of the lower unit deposits in Holden crater. These processes include one or multiple lacustrine episodes on the crater floor (e.g., Grant and Parker, 2002; Grant et al., 2008; Irwin et al., 2005a; Malin and Edgett, 2000; Moore and Howard, 2005; Parker, 1985; Pondrelli et al., 2005), air-fall (Malin and Edgett, 2000), or glacial origin (Pondrelli et al., 2005). Submeter-scale images afforded by HiRISE, however, reveal a series of characteristics associated with the lower unit that point to emplacement in a distal alluvial or lacustrine setting during the Late Noachian to Early Hesperian Epoch. The thin, laterally continuous bedding that is confined below –1960 m within the crater is most consistent with a water-lain origin for all three members of the lower unit. Wind-blown traction deposits, loess, tephra, or other air-fall mantles would drape a wider range of topography, and associated master bed sets would not be as thinly bedded or restricted in elevation as is observed. Furthermore, the upper member beds are often separated by lensoidal accumulations of clasts that are too large for eolian transport. There are no nearby, large volcanic constructs or deposits that would point toward a primary volcanic origin for the deposits (Scott and Tanaka, 1986). Compositional spectral data indicating incorporation of phyllo­ silicates supports this contention (Glotch, 2006; Milliken et al., 2007). The gen­ erally block-poor nature of the deposits, parallel and conformable bounding surfaces of thin, laterally continuous, individual strata and encompassing members, and elevation confinement argue against their emplacement as impact ejecta, although impacts could contribute to some of the clasts distinguishing beds of the upper member. While non-aqueous depositional processes for the lower unit appear improbable, distinguishing between a distal alluvial versus lacustrine depositional environment is challenging. This is true of the Holden deposits as well as some terrestrial strata where deposition in relatively low-energy settings can yield bedding that is similar in appearance (McCormick and Grotzinger, 1993; Winston, 1978). The upper member of the lower unit incorporates some large clasts and may be most consistent with an alluvial origin. However, the relatively close spacing of relict fan distributaries on the bajada along the western crater wall implies that greater lateral variability of alluvial bedding should be observed if the lower unit is an alluvial deposit (Figure 12.6). In addition, the lower unit is exposed beneath coarser alluvium in eroding fan fronts, and constituent beds are uniformly thick and continuous over much greater length scales than are observed in the overlying, sometimes steeply dipping (>2°) fan materials. Bedding in the fan materials varies over tens to hundreds of meters, whereas the underlying upper member beds are nearly flat-lying and continuous along kilometer length scales. Collectively, these observations coupled with the broad distribution (of at least the upper member) and restriction of the lower unit strata below a common elevation favor a lacustrine origin, which could include a playa setting especially for upper portions of the section.

Aqueous depositional settings in Holden crater, Mars

337

Variability in the elevation of individual lower unit strata relative to DEM contours (Figure 12.7) may reflect deformation related to later fracturing of the crater floor materials, but the gentle eastward dip of the beds may occur at a scale larger than the deformation associated with the fractures. If true, then it is uncertain whether the broad similarity between the slope of the lower bajada to the west and the lower unit may be related to deposition in a distal alluvial system. A decrease in slope might be expected at the toe of the bajada (Blair and McPherson, 1994) however, and its apparent absence could also be consistent with deposition in a lake, if the adjacent fans were prograding and there was insufficient wave energy for redistributing sedi­ ments along the shoreline. The interpretation most consistent with the suite of observations is one where the lower unit members record the transition from lacustrine to distal alluvial facies up-section. An early lacustrine setting responsible for the lower and middle members could have been partially sustained by groundwater, given the low elevation of the crater floor compared to surrounding surfaces, as well as precipitation induced surface runoff from crater walls. Early stripping of debris from the crater walls extended fans onto the crater floor and sourced the thicker-bedded lower and middle members that were deposited in a lake. Later, the expanding alluvial system fringing the crater walls may have prograded across the lacustrine beds and deposited the upper member. Occasional pulses of channelized alluvium could account for the lenses of blocky material in the upper member (Howard et al., 2007), which is consistent with this scenario. The polygonally fractured material capping the lower unit may represent a terminal playa phase and suggests that the surface hydrologic system generally decayed with time at the end of the Noachian and into the Early Hesperian. Deposition of the lower unit occurred before the Uzboi Vallis rim breach was incised and requires that all incorporated sediments were derived from within Holden crater. Hence, observed phyllosilicate spectral signatures (Milliken et al., 2007) must also have been derived locally; either eroded from crater walls or formed in situ in a lacustrine setting. If the phyllosilicates result from in situ weathering in a lake, then their decreasing abundance up-section likely records changing environmental condi­ tions that may be consistent with a decaying hydrologic system. Alternatively, the phyllosilicate-bearing sediments may be derived from the crater walls or bajada, perhaps from materials altered during and after the impact as hot fluids circulated through the rocks (Newsom et al., 1996), from subaerial rock weathering following precipitation, or from deposits excavated from the older Holden basin. For phyllosi­ licates derived from the crater walls, decreasing abundance up-section likely reflects changing degradational processes on the walls and/or erosional exposure of more resistant (phyllosilicate-poor) materials.

12.5.2 Origin of upper unit stratigraphy The upper unit postdates the lighter-toned, layered lower unit deposits in the south­ western portion of the crater (Grant and Parker, 2002; Irwin et al., 2005a; Moore and Howard, 2005; Parker, 1985; Pondrelli et al., 2005). Many of the fan-like deposits associated with the upper unit are approximately radial to the rim breach created by

338

Lakes on Mars

Uzboi Vallis (Figure 12.2), and there is an increase in bed continuity and inferred decrease in grain size away from the breach. Hence, emplacement of the upper unit was probably associated with high-magnitude discharge out of Uzboi that was capable of eroding into the underlying lower unit and facilitating movement of large blocks of eroded lower unit strata that are usually found in the lee of flow obstructions (Figure 12.8). The result was a boulder-rich deposit tens of meters thick that unconformably overlies the lower unit and is locally capped by large bedforms incorporating meterscale blocks (Figure 12.9). Upper unit deposits drape over antecedent relief, including locally elevated surfaces, but they are confined to elevations below –2060 m (Grant et al., 2008; Pondrelli et al., 2005). Additional evidence suggests that discharge out of Uzboi Vallis into the crater fed a late-stage lake achieving a depth of at least 50 m (Grant and Parker, 2002; Grant et al., 2008; Irwin et al., 2005a), but this lake was geologically short-lived. Topographic confinement, association with rim breaches, incorporation of large blocks of locally eroded and transported material, and bedforms comprised of metersized rocks are consistent with high-energy aqueous depositional settings. The absence of large-scale cross beds, which should be resolvable and are expected for eolian deposits, together with topographic confinement of the upper unit rules out deposition by the wind. Volcanic, impact, or other processes also cannot account for observed characteristics. Most likely, the upper unit is the result of alluvial and subsequent lacustrine deposition. The large fans (Figure 12.2) and numerous truncat­ ing beds indicate that alluvial processes were responsible for at least the proximal portion of the upper unit. The transition to more flat-lying beds distally and upsection, however, implies that a lacustrine system was also present and was likely created as discharge from the rim breach ponded on the crater floor. Fine-scale beds noted toward the center of the crater (Figure 12.5) occur below – 2060 m and appear to be conformable with a series of layers and a bench exposed at approximately –2060 m along the western bajada and central peak (Figure 12.10). The beds could represent deposits from the lacustrine phase associated with upper unit deposition or could be exposures of the upper member of the lower unit. By contrast, the eroded bench could represent a paleo-strandline (Figure 12.10) and would be consistent with occurrence of a lacustrine phase associated with the upper unit. Topography indicates that the impoundment in Uzboi Vallis, responsible for over­ topping of Holden’s rim at an elevation of –350 m, would have extended back to the northern rim of Bond crater (which rises 450–500 m above the level of the breach at Holden) and represented approximately 4000 km3 of water. Such a volume was more than sufficient to flood Holden to the –2060 m inferred from landforms and mapping (Pondrelli et al., 2005). The result of flooding would be a lake some 200–250 m deep and consistent with the observed transition from truncating to more parallel and continuous bedding away and up section from the breach. Although the source of water for impoundment in Uzboi Vallis is uncertain, the rim of Bond crater, which appears to predate the impoundment, rises to an elevation of �120 m and would have precluded any back drainage into Argyre. Hence, it is unlikely that water was contributed to Holden crater from sources in Argyre basin or even farther to the south. The Noachian-aged Nirgal Vallis (Reiss et al., 2004) is a

Aqueous depositional settings in Holden crater, Mars

339

Figure 12.10 HiRISE image of a portion of the central peak in Holden crater near 25.9ºS, 326ºE revealing an abrupt break in slope and local relief at an elevation of –2060 m (white dashed line). This elevation closely corresponds with topographic confinement of the upper unit and possibly represents a paleo-strand line from a late stage lake associated with discharge into the crater from Uzboi Vallis. HiRISE image PSP_003433_1540_RED; 26 cm/pixel scale. North is toward the top of the image.

major tributary to Uzboi Vallis and is a likely contributor to the impounded water. However, Nirgal enters Uzboi about 400 m lower than the level of Holden’s rim, so there was considerable flooding along at least the lower reaches of Nirgal Vallis prior to breaching of the rim of Holden crater. Nirgal Vallis had an estimated minimum discharge of 4800 m3 s−1 (Irwin et al., 2005b), which could have filled Uzboi Vallis to the point that it overtopped Holden’s rim in 25–30 years or less, if flow was constant. The likelihood that water was also contributed from other sources might have decreased the time required for flooding Uzboi Vallis. By analogy with the Earth, discharge rates from multiple sources likely varied considerably, and losses due to infiltration and evaporation undoubtedly increased the period necessary to flood Uzboi. Nevertheless, it appears that estimates of discharge into Uzboi Vallis from Nirgal Vallis alone could contribute the volume of water needed to overtop Holden’s rim over a geologically short period. Water impounded in Uzboi Vallis initially overtopped Holden’s rim at multiple locations along the western side of the crater at an elevation of about –350 m, but eventually consolidated at a single breach that is deeply incised into the rim farther to the east (Figure 12.11). As water first overtopped the rim, drainage down the interior wall created sizeable gullies and channels around pre-existing relief as it flowed out onto the crater floor and scoured the older lower unit. It is worth noting that erosion of the lower unit by discharge from Uzboi Vallis is responsible for creating many of the outcrops that remain visible today. As drainage continued and became focused in the main breach at the eastern end of the original spillway, crosscutting relationships indicate that channelized flows from

340

Lakes on Mars

Figure 12.11 HiRISE images showing the channels incised into Holden’s rim when water impounded in Uzboi Valles drained into the crater. (A) Drainage initially occurred via multiple active western channels. HiRISE image PSP_004277_1530_RED near 26.8ºS, 324.9ºE (see Figure 12.2 for context). (B) Drainage eventually converged on a single dominant eastern channel that deeply incised the rim. Collapse along the wall of the dominant breach may have periodically interrupted drainage into the crater. A prominent slump (bottom right) that is not incised confirms that this process continued after drainage through the breach had ended. HiRISE image PSP_003710_1530_RED near 27ºS, 325ºE (see Figure 12.2 for context). Both HiRISE images in (A) and (B) have 26 cm/pixel scale; north is indicated and is the same in (A) and (B).

the western side of the spillway gave way to preferential erosion of the lower unit to the east and below the bajada (Figure 12.2). Relatively early discharge scoured the long trough extending north from the breach that is bounded by the most extensive visible outcrops of the lower unit in the crater. As drainage through the rim breach continued, discharge shifted eastward, incised, and deposited a series of low fans (Figure 12.2). The isolated nature of some of the fans, such as the largest one visible in Figure 12.2, suggests that discharge may have been pulsed, perhaps related to irregular rates of downcutting or blocking and opening of the breach due to wall collapse and reincision (Figure 12.11). Initial ponding of early drainage on the crater floor may have limited more widespread scour of the lower unit materials. Several observations indicate that the discharge and ponding associated with emplacement of the upper unit occurred over a geologically brief interval of time. First, a small segment of Nirgal Vallis incises a short segment of the drained floor of Uzboi Vallis (Figure 12.12), but it does not continue into Holden as expected if there was a lengthy period of discharge following the breach and drainage of impounded water. In addition, the fans and deposits comprising the upper unit exhibit mafic

Aqueous depositional settings in Holden crater, Mars

341

Figure 12.12 THEMIS daytime IR image I08503003 of the floor of Uzboi Vallis just below (north of) the confluence with Nirgal Vallis. White arrows denote a shallow and locally sediment-choked channel associated with Nirgal that incises the floor of Uzboi Vallis for a short distance. The incised channel does not extend across the entire floor of Uzboi Vallis and into Holden crater, which may be due to shallow residual ponding south of Holden ejecta emplaced within lower Uzboi Vallis, and a discharge that was not adequately sustained to fully incise the ejecta. North is toward the top of image.

compositions and a paucity of phyllosilicate signatures (Glotch, 2006; Milliken et al., 2007), consistent with rocks derived from the Uzboi Vallis breach that had limited subsequent contact with water. Although the absolute lifetime of a 200–250 m-deep lake in Holden resulting from discharge out of Uzboi remains speculative, the crater is a closed basin and must have lost all water via infiltration and evaporation. Infiltration is difficult to constrain, but comparison with terrestrial evaporation rates of about a meter per year (Idso, 1981; Kohler et al., 1959) suggests that a 200–300 m-deep lake might persist for hundreds of years.

12.6

Discussion

The upper and lower units in Holden crater record deposition during two very different phases of lacustrine and/or distal alluvial activity during the Late Noachian Epoch. These units mark the end of a long history of aqueous erosion and deposition in southern Margaritifer Terra that likely dates back to the Early Noachian Epoch. Aqueous deposits emplaced during the Middle or Early Noachian Epochs predate Holden crater, but they may also be exposed as megabreccia in the walls of crater.

342

Lakes on Mars

The megabreccia could include large clasts of sedimentary materials deposited in Holden basin or lower Uzboi Vallis, prior to formation of Holden crater during the Late Noachian or Early Hesperian Epoch. If present, such materials may be altered by the impact (Melosh, 1989) or subsequent hydrothermal alteration (Newsom et al., 1996), but they could retain sedimentologic and geochemical information relating to conditions when they were originally deposited. Deposition of the lower unit occurred in a closed basin over an extended period in what was most likely a lacustrine setting initially, but which probably evolved to a more distal alluvial setting over time. Although any contributions from groundwater to this early aqueous phase in the crater remain speculative, the low regional elevation of the crater floor suggests that they could have been important. A transition to a distal alluvial setting could reflect decreasing water and sediment shed from crater walls and across the fringing bajada as well as decreased contributions from groundwater, perhaps accentuated by the increasing level of fill in the crater. Polygonal fractures capping the lower unit strata are consistent with occurrence of a terminal playa stage and decreasing abundance of water over time. Hence, it is possible that the lower unit records the transition from relatively wet conditions during much of the Noachian and into the Early Hesperian to mostly drier conditions that have persisted since that time. Lower unit strata also provide clear depositional context for phyllosilicates in an alluvial/lacustrine environment on Mars. The lower unit stratigraphy and incorporated phyllosilicates suggest emplacement in a quiescent aqueous setting that requires widespread, stable, wet, and clement conditions unlike those of modern Mars. Because Holden was a closed basin during deposition of the lower unit, the associa­ tion between the lower unit and fringing bajada indicates that at least some of the water contributing to transport and deposition was derived from precipitation. Such precipitation may have been orographically enhanced, as supported by the greater extent of the bajada fringing the higher, western wall of the crater, but it must in part reflect a larger, likely global atmospheric circulation. There is nothing special about Holden’s elevation or location that would limit rain or snowfall to its confines, and the surroundings have widespread valley networks and alluvial deposits (e.g., Grant, 1987; Moore and Howard, 2005; Moore et al., 2003). By contrast, the upper unit reflects deposition in a very high-energy aqueous setting that was not very long-lived. Water impounded in Uzboi Vallis eventually overtopped the crater rim and drained into the crater. The result was the relatively dark-toned and variable alluvial (grading to lacustrine) deposits comprising the upper unit. Although the upper unit was deposited in a high-energy setting and over a short period, water cascading down and through the crater wall during incision of the Uzboi Vallis rim breach was responsible for exposing the outcrops of the lower unit that remain visible today. In the absence of an event like the one responsible for emplace­ ment of the upper unit, the lower unit in Holden might go largely or completely undetected. Because Holden is generally similar in morphology to many other large craters on Mars, it is possible that deposits similar to those of the lower unit are more widespread than previously recognized. The quiescent aqueous setting indicated by the lower unit rocks is more likely to have preserved geochemical or lithological signatures consistent with any habitable

Aqueous depositional settings in Holden crater, Mars

343

environments than the high energy, shorter-lived setting responsible for the upper unit. It is possible that phyllosilicates in the lower unit may preserve subtle biogenic signatures or even adsorbed organics if they were present (Knoll and Grotzinger, 2006) and deposited in a low energy aqueous setting, as is inferred. As such, the materials comprising the lower unit may hold the key to understanding the potential habitability of surface water environments on early Mars. By contrast, the higherenergy and shorter-lived environment recorded by the upper unit suggests an envir­ onment where indicators of habitability were less likely to be preserved (Knoll et al., 2005). Access to stratigraphy exposed in Holden crater represents a unique opportunity to evaluate a potentially habitable setting on Mars and is a prime factor in the crater’s inclusion among four final sites under consideration as the landing site for the 2011 Mars Science Laboratory rover. It is possible that the lower unit beds in Holden correlate with the fluvial delta and inferred lacustrine deposits in Eberswalde crater to the north (Figure 12.1; Malin and Edgett, 2003; Moore et al., 2003). The ejecta from Holden superposes the rim of Eberswalde crater (Moore et al., 2003), but underlies the delta and lacustrine deposits, thereby indicating their deposition post-dates the for­ mation of Holden crater. If the deposits in these two craters formed at the same time, it would lend support to the interpretation that the Eberswalde deposits are the result of relatively long-lived formative flows derived from at least regional-scale precipitation (Malin and Edgett, 2003; Moore et al., 2003). Since the Eberswalde delta deposits include multiple lobes and may record multiple depositional episodes (Malin and Edgett, 2003; Moore et al., 2003), it is possible that some correlate with the upper unit rim breach deposits in Holden. If not, impoundment of water in Uzboi Vallis, responsible for the upper unit breach deposits in Holden, may have included con­ tributions from more distant sources and/or groundwater-derived discharge from Nirgal Vallis rather than regional or broader-scale precipitation over Holden and Eberswalde craters.

Acknowledgments We thank the people at the University of Arizona, Ball Aerospace, the Jet Propulsion Laboratory, and Lockheed Martin that built and operate the HiRISE camera and the Mars Reconnaissance Orbiter Spacecraft. Reviews by Alan Howard and Nadine Barlow improved the manuscript. Work supported by NASA.

References Baker, V.R., 1982. The Channels of Mars. University of Texas Press, Austin, TX, 198 pp. Baker, V.R., Nummedal, D., 1978. The Channeled Scabland: a guide to the geomorphology of the Columbia Basin, Washington. Comparative Planetary Geology Field Conference, Columbia Basin, 5–8 June 1978, NASA, Washington, DC, 186 pp.

344

Lakes on Mars

Banerdt, W.B. 2000. Surface drainage patterns on Mars from MOLA topography, Eos Trans. AGU, 81, 48, Fall Meet. Suppl., Abstract P52C–04. Bibring, J.-P., Langevin, Y., Mustard, J.F., Poulet, F., Arvidson, R., Gendrin, A., et al., 2006. Global mineralogical and aqueous Mars history derived from OMEGA/Mars Express data. Science 21, 400–404, DOI: 10.1126/science.1122659. Blair, T.C., McPherson, J.G., 1994. Alluvial fan processes and forms. In: Abrahams, A.D., Parsons, A.J. (Eds.), Geomorphology of Desert Environments. Chapman and Hall, London, pp. 354–402. Boothroyd, J.C., 1983. Fluvial drainage systems in the Ladon Basin area: Margaritifer Sinus area, Mars. Geol. Soc. Am. Abstr. Programs 15, 530. Carr, M.H., 1979. Formation of Martian flood features by the release of water from confined aquifers. J. Geophys. Res. 84, 2995–3007. Carr, M.H., 2006. The Surface of Mars. Cambridge University Press, Cambridge, 307 pp. Carr, M.H., Clow, G.D., 1981. Martian channels and valleys: their characteristics, distribution, and age. Icarus 48 (1), 91–117, doi:10.1016/0019-1035(81)90156-1. Fassett, C.I., Head, J.W., 2008. The timing of Martian valley network activity: constraints from buffered crater counting. Icarus 195, 61–89, doi:10.1016/j.icarus.2007.12.009. Fergason, R.L., Christensen, P.R., Kieffer, H.H., 2006. High resolution thermal inertia derived from THEMIS: Thermal model and applications. J. Geophys. Res. 111, E12004, doi:10.1029/2006JE002735. Frey, H.V., Frey, E.L., Hartmann, W.K., Tanaka, K.L.T. 2003. Evidence for buried “preNoachian” crust pre-dating the oldest observed surface units on Mars, Lunar Planet. Sci. Conf. XXXIV, Abstract 1848. Glotch, T.D., 2006. Olivine and pyroxene-rich deposits in Holden crater, Mars, Eos Trans. AGU, 87, Fall Meet. Suppl., Abstract P22A–04. Grant, J.A., 1987. The geomorphic evolution of Eastern Margaritifer Sinus, Mars. Adv. Planet. Geol. NASA Tech. Memo. 89871, 1–268. Grant, J.A., 2000. Valley formation in Margaritifer Sinus, Mars, by precipitation-recharged ground-water sapping. Geology 28, 223–226. Grant, J.A., Irwin, R.P. III, Grotzinger, J.P., Milliken, R.E., Tornabene, L.L., McEwen, A.S., et al., 2008. HiRISE imaging of impact megabreccia and sub-meter aqueous strata in Holden Crater, Mars. Geology 36, 195–198, doi: 10.1130/G24340A. Grant, J.A., Parker, T.J., 2002. Drainage evolution of the Margaritifer Sinus region, Mars. J. Geophys. Res. 107, doi: 10.1029/2001JE001678. Grant, J.A., Wilson, S.A., Fortezzo, C.M., Clark, D.A., 2009. Geologic map of MTM -20012 and -25012 quadrangles, Margaritifer Terra region of Mars, U. S. Geol. Sur., Scientific Investigations Map 3041, scale 1:1,000,000. Grieve, R.A.F., Dence, M.R., Robertson, P.B., 1977. Cratering processes: as interpreted from the occurrence of impact melts. In: Roddy, D.J., et al. (Eds.), Impact and Explosion Cratering. Pergamon Press, New York, pp. 791–814. Howard, A.D., Moore, J.M., Irwin, R.P. III, 2005. An intense terminal epoch of widespread fluvial activity on early Mars: 1. Valley incision and associated deposits. J. Geophys. Res. 110, E12S14, doi: 10.1029/2005JE002459. Howard, A.D., Moore, J.M., Irwin, R.P. III, Dietrich, W.E., 2007. Boulder transport across the Eberswalde delta, Lunar Planet. Sci. Conf. XXXVIII, Abstract 1168. Idso, S.B., 1981. Relative rates of evaporative water losses from open and vegetation covered

water bodies, Water Resource Research Bull., Am. Water Res. Assoc., 17, pp. 46–48

Irwin, R.P. III, Craddock, R.A., Howard, A.D., 2005b. Interior channels in Martian valley

networks: Discharge and runoff production. Geology 33, 489–492.

Aqueous depositional settings in Holden crater, Mars

345

Irwin, R.P. III, Grant, J.A., 2009. Large basin overflow floods on Mars, in megafloods on Earth and Mars. In: Burr, D.M., Baker, V.R., and Carling, P.A. (Eds.) Cambridge University Press, Chapter 11. Irwin, R.P. III, Howard, A.D., Craddock, R.A., Moore, J.M., 2005a. An intense terminal epoch of widespread fluvial activity on early Mars: 2. Increased runoff and paleolake develop­ ment. J. Geophys. Res. 110, E12S14, doi:10.1029/2005JE002460. Knoll, A.H., Carr, M.H., Clark, B.C., Des Marais, D.J., Farmer, J.D., Fischer, W.W., et al., 2005. An astrobiological perspective on Meridiani Planum. Earth. Planet. Sci. Lett. 240, 179–189, doi: 10.1016/j.epsl.2005.09.045. Knoll, A.H., Grotzinger, J.P., 2006. Water on Mars and the prospect of Martian life. Elements 2, 171–175. Kohler, M.A., Nordenson, T.J., Baker, D.R., 1959. Evaporation maps for the United States, U.S. Weather Bureau, Technical Paper 37. Komar, P.D., 1979. Comparisons of the hydraulics of water flows in Martian outflow channels with flows of similar scale on Earth. Icarus 37, 156–181. Malin, M.C., Edgett, K.S., 2000. Evidence for recent groundwater seepage and surface runoff on Mars. Science 288 (5475), 2230–2335. Malin, M.C., Edgett, K.S., 2003. Evidence for persistent flow and aqueous sedimentation on Mars. Science 302, 1931–1934, doi:10.1126/science.1090544. McCormick, D.S., Grotzinger, J.P., 1993. Distinction of marine from alluvial facies in the early Proterozoic (ca. 1.9 Ga) Burnside Formation, Kilohigok Basin, N.W.T. J. Sediment. Petrol. 63, 398–419. McEwen, A.S., Eliason, E.M., Bergstrom, J.W., Bridges, N.T., Hansen, C.J., Delamere, W.A., et al., 2007. Mars Reconnaissance Orbiter’s High Resolution Imaging Science Experiment (HiRISE). J. Geophys. Res. 112, E05S02, doi:10.1029/2005JE002605. Melosh, H.J., 1989. Impact Cratering. Oxford University Press, New York, 245 pp. Milliken, R.E., Grotzinger, J., Grant, J.A., Arvidson, R., Murchie, S., 2008. Understanding sedimentary sources and sinks on Mars from orbit, Geol. Soc. Am. Prog. with Abs., Abstract 134–1. Milliken, R.E., Grotzinger, J., Grant, J., Murchie, S., The CRISM Science Team, 2007. Clay minerals in Holden crater as observed by MRO CRISM. 7th International Mars Conf., Abstract 3282. Moore, J.M., Howard, A.D., 2005. Large alluvial fans on Mars. J. Geophys. Res. 110, E04005, doi:10.1029/2005JE002352. Moore, J.M., Howard, A.D., Dietrich, W.E., Schenck, P.M., 2003. Martian layered fluvial deposits: implications for Noachian climate scenarios. Geophys. Res. Letts. 30, doi:10.1029/2003GL019002. Murchie, S., Arvidson, R., Bedini, P., Beisser, K., Bibring, J.-P., Bishop, J., et al., 2007. Compact Reconnaissance Imaging Spectrometer for Mars (CRISM) on Mars Reconnais­ sance Orbiter (MRO). J. Geophys. Res. 112, E05S03, doi:10.1029/2006JE002682. Newsom, H., Brittelle, G., Hibbitts, C., Crossey, L., Kudo, A., 1996. Impact crater lakes on Mars. J. Geophys. Res. 101 (E6), 14951–14955. Parker, T.J., 1985. Geomorphology and geology of the southwestern Margaritifer Sinus-northern Argyre region of Mars, California State University, M.S. Thesis, Los Angeles, California. Parker, T.J., 1994. Martian paleolakes and oceans, University of Southern California, Ph.D. Thesis, Los Angeles, California. Phillips, R.J., Zuber, M.T., Solomon, S.C., Golombek, M.P., Jakosky, B.M., Banerdt, W.B., et al., 2001. Ancient geodynamics and global-scale hydrology on Mars. Science 291, 2587–2591.

346

Lakes on Mars

Pondrelli, M., Baliva, A., Di Lorenzo, S., Marinangeli, L., Rossi, A.P., 2005. Complex evolu­ tion of paleolacustrine systems on Mars: An example from the Holden crater. J. Geophys. Res. 110, E04016, doi:10.1029/2004JE002335. Reiss, D., van Gasselt, S., Neukum, G., Jaumann, R., 2004. Absolute dune ages and implica­ tions for the time of formation of gullies in Nirgal Vallis, Mars. J. Geophys. Res. 109, doi: 10.1029/2004JE002251. Rotto, S., Tanaka, K.L., 1995. Geologic/Geomorphic map of the Chryse Planitia region of Mars, U.S. Geol. Surv. Misc. Invest. Ser. Map I-2441, scale 1:5M. Saunders, S.R., 1979 Geologic map of the Margaritifer Sinus quadrangle of Mars, U.S. Geol. Surv. Misc. Invest. Ser. Map I–1144, scale 1:5M. Schultz, P.H., Glicken, H., 1979. Impact crater and basin control of igneous processes on Mars. J. Geophys. Res. 84, 8033–8047. Schultz, P.H., Schultz, R.A., Rogers, J., 1982. The structure and evolution of ancient impact basins on Mars. J. Geophys. Res. 87, 9803–9820. Scott, D.H., Tanaka, K.L., 1986, Geologic map of the western equatorial region of Mars, U.S. Geol. Surv. Misc. Invest. Ser. Map I–1802–A. Vishnevsky, S., Montanari, A., 1999. The Popigai impact crater (Arctic Siberia, Russia): Geology, petrology, geochemistry, and geochronology of glass-bearing impactites. In: Dressler, B.O., Sharpton, V.L. (Eds.), Large Meteorite Impacts and Planetary Evolution II: Geological Society of America Special Paper 339, 19–60. Winston, D., 1978. Fluvial systems of the Precambrian Belt Supergroup, Montana and Idaho, U.S.A. In: Miall, A.D. (Ed.), Fluvial Sedimentology: Canadian Society of Petroleum Geologists Memoir 5, 343–359.

13 Dynamics of declining lake

habitat in changing climate Nathalie A. Cabrol*, Edmond A. Grin*, Guillermo

Chong‡, Donat P. Ha¨der§, Edwin Minkleyk,

Youngseob Yu¶, Cecilia Demergasso**,

John A. Gibson††, and Darlene Lim*

*

SETI CSC/ NASA Ames, Space Science and Astrobiology Division, Moffett Field, CA, USA ‡ Centro de Investigacio´n Cientifica y Tecnolo´gica para Minerı´a (CICITEM), Antofagasta, Chile § Department of Botanic, University of Erlangen, Erlangen, Germany k Department of Biological Sciences, Carnegie Mellon University, Pittsburgh, PA, USA ¶ Department of Civil and Environmental Engineering, Carnegie Mellon University, Pittsburgh, PA, USA ** Centro de Biotecnologı´a, Universidad Cato´lica del Norte, Antofagasta, Chile †† Marine Research Laboratories, Tasmanian Aquaculture and Fisheries Institute, University of Tasmania, Hobart, Australia

13.1

Introduction

As discussed in previous chapters, identifying traces of water activity and ancient aqueous sedimentary basin deposits was an essential step in documenting whether early Mars was habitable. Evidence for a wetter history was provided repeatedly during the past 40 years of exploration by orbiters and more recently from the ground by the MER mission. Data collected by the Opportunity rover are consistent with episodic ponds in Meridiani Planum that may have reoccurred over a prolonged period of time from groundwater upwelling and evaporation (e.g., Anderson and Bell, 2008; Andrews-Hanna et al., 2007; Ruff et al., 2008; Squyres and Knoll, 2005, 2006; Squyres et al., 2004). In Gusev crater, the oldest units explored by the Spirit rover in the Columbia Hills and Inner Basin show soils containing up to 50% inferred iron sulfates and altered rocks (e.g., Arvidson et al., 2006; Squyres and Knoll, 2005; Wang et al., 2008; Yen et al., 2008). At Home Plate, scoria, outcrops showing lapilli and layered ash deposits (e.g., Arvidson et al., 2008; Cabrol et al., 2008; Lewis et al., 2008), and soils and nodules highly enriched in silica support the hypothesis Lakes on Mars. DOI: 10.1016/B978-0-444-52854-4.00013-1 © 2010 Elsevier B.V. All rights reserved.

348

Lakes on Mars

that water and magma interacted at Gusev and generated explosive volcanism (Squyres et al., 2007) with hydrothermal activity (McAdam et al., 2008; Rice et al., 2008; Schmidt et al., 2008; Squyres et al., 2008; Wang et al., 2008; Yen et al., 2008). In the Northern polar region of Mars, the identification of perchlorate salts by the Phoenix lander has raised the question of potentially limiting factors for habitability (Smith et al., 2008) but does not rule out the possibility of life on Mars, whether past or present, as the role of perchlorates on habitability and life is complex. For instance, while highly oxidizing perchlorates can be toxic, perchlorate salts are also formed on Earth, such as in the Atacama Desert in Chile, where microbial life is present including species using them as a source of energy (Logan et al., 2001). Early Mars had water, carbon, energy sources, a dynamic geology, and magnetic fields. If life developed, the loss of most of the atmosphere within 600 million years of the solidification of the planet’s crust was a major setback. This loss was accompanied by the dissipation of the magnetic field (Connerney et al., 1999) and rapid climate change as evidenced by the geological record (e.g., Baker, 2001, 2004; Beaty et al., 2005; Bibring et al., 2005; Carr, 1996; Clifford, 1993; Craddock and Howard, 2002; Haberle, 1998; Jakosky and Phillips, 2001; Jakosky, 1999; Kasting, 1991; Malin and Edgett, 2000a, 2000b; Phillips et al., 2001 and many more). Without magnetic field, increased UV and other deadly radiation would have reached the surface. A thinning atmosphere would also have generated colder temperature with larger daily and seasonal amplitudes, increasing the potential for inverse relationships between UV radiation and temperatures. On Earth, studies have demonstrated that these conditions are detrimental for DNA repair, especially in a cold environment (Folt et al., 1999; Grad et al., 2001; Leech et al., 2005; Williamson et al., 2002; Zagarese et al., 1997). By comparison with Earth’s evolution, such change had the potential to trigger widespread extinction for species that could not adapt. As a result, it is broadly accepted that microorganisms, if any, already living underground or those that were able to migrate into the subsurface had the best chance of survival (Allen et al., 2005; Brack et al., 1998; Des Marais and Farmer, 1995; Farmer, 1995, 2000; Farmer and Des Marais, 1999; Jakosky et al., 2007; McKinley et al., 2000; Schulze-Makuch et al., 2008; Tokano, 2005). Assuming life evolved in aquatic niches such as lakes and ponds during the early part of the Martian history, the Noachian/Hesperian transition was a period of increased environmental stress that ended with the loss of surface liquid water. Their physical environment included high-daily amplitude, low average temperatures, thin atmosphere, high solar irradiance, ice, reduced yearly precipitation, enhanced evaporation, and volcanic and hydrothermal environment. How long those bodies of water remained habitable as their water column shrank and how fast their chemical and physical properties changed are key to understanding whether their ecosystems, if any, had a chance to survive and possibly to adapt to more sheltered habitats offering UV protection, a source of moisture, and energy (Cabrol et al., 2007a). High altitude lakes (4200–6000 m) in the Central Andes of Chile and Bolivia may provide important clues about the physical and chemical transformations experienced by Martian lake habitats. Their environment is close to that of Mars 3.5 billion years ago. These lakes are evaporating rapidly due to a climate change initiated at the last

Dynamics of declining lake habitat in changing climate

349

deglaciation (e.g., Grosjean et al., 2001; Messerli et al., 1993). Many were formed and/or reached their peak volume under precipitation regimes of 400–700 mm/year between 18,000 and 14,000 BP and started to recede at the beginning of the Holocene (Servant and Fontes, 1978; Sylvestre et al., 1999; Grosjean et al., 2001; Fornari et al., 2001; and many others). Their decline is accelerated today by global warming. The Central Andes have lost 50% of their precipitation in the last 50 years. Most of the South American glaciers from Bolivia, Chile, and Argentina up to 25oS latitude have drastically reduced their volume with glacial retreat reaching seven-folds in the past 13 years (Bradley et al., 2006). In this context, several of these lakes have been monitored by the High Lakes Project (HLP) since 2002. They are being used as proxies to early Martian lakes to infer the impact of rapid climate change and interannual variability on their physical and chemical properties and their overall habitability. The scientific objectives of HLP are detailed in Cabrol et al., 2007b, 2009 and can be summarized as follows: (i) to explore and characterize the habitability potential of high-altitude lakes in an envir­ onment presenting analogies to early Mars; (ii) to assess the impact of low ozone/high solar irradiance in nonpolar aquatic environments; (iii) to document and map poorly known ecosystems before they disappear; and (iv) to quantify the impact of climate change on planetary lake environments and ecosystems. Among the sites surveyed by HLP, the summit lake of the Licancabur volcano (designated hereafter as the Licancabur lake) is where data were collected most consistently over the years and they give us an insight into what could have happened in Martian lakes under a comparable set of environmental stresses 3.5 billion years ago. The results of the study at Licancabur are the focus of the following sections. We discuss the analogy with Mars in Section 13.2, in particular that of the Licancabur area. Data collection and analytical methods are presented in Section 13.3. Section 13.4 shows the results from the meteorological and geophysical data (Sections 13.4.1 and 13.4.2) and lake chemistry (Section 13.4.3) over the years. Lake biodiversity, including zooplankton and microbial community, is detailed in Sections 13.4.4 and 13.4.5, respectively. We conclude on the timing of physicochemical and biological cycles and its importance in the survival of the ecosystem.

13.2

Environmental analogy to Mars

The Licancabur volcano (22°83′S and 67°88′W) towers at 5970 m on the border of Chile and Bolivia. Constructed largely in Late Pleistocene from basaltic andesitic flows, it is considered potentially active (de Silva and Francis, 1991; Marinovic and Lahsen, 1984). It is known as one of the two highest volcanic lakes on Earth and is nestled about 50 m inside the summit crater at 5916 m (Figure 13.1). The lake is permanent. Ojos del Salado (27°07′S/28°32′W, 6891 m) located to the south of Licancabur in the Chilean Andes hosts a lake on its eastern flank at 6390 m. It is generated by snowmelt; however, its depth and current status (perennial, episodic) are unknown as data still lacks.

350

Lakes on Mars

a

b

Figure 13.1 (a) The Licancabur volcano from the Chilean altiplano, and (b) the 100-m long lake nestled in the summit crater (November, 2006). After 4 years of steady decline, the lake was back at its maximum level since 2002. Licancabur (marked by a dot on the satellite image to the right) is bound to the west by the Atacama Desert in Chile, which maintains a high aridity most of the year. Between January and March, humid air originating from Argentina to the east brings 80% of the scarce yearly precipitation. Photo credit: 2006 HLP expedition, NASA NAI/ SETI CSC/NASA Ames (satellite image from NASA).

The environmental analogy of the Licancabur region with early Mars is multifold. The aridification of the Central Andes coincided with the end of the last deglaciation and the beginning of the Holocene and is accelerating today as a result of climate change (Bradley et al., 2006). Similarly to most of the neighboring lakes, the water level of the Licancabur lake has fluctuated over the past 20 years. In an environment of aridity and evaporation, its geographical location (latitude and altitude) results in intense solar irradiance with UV flux double that of present-day Mars at the equator and UVB only half that of the red planet (Hock, 2008), low average total ozone, and a low (480 mb) atmospheric pressure (Cabrol et al., 2007a, 2000b, 2009, 2010). The yearly temperature extremes range from –40°C to +9°C with daily averages of –12°C in winter and –5°C in summer. Ice covers the surface of the lake starting from April (austral fall), reaching a thickness of 80 cm by August and trapping 1350 m3 of the lake 9000 m3. Typically, thawing occurs in September (austral spring), but negative

Dynamics of declining lake habitat in changing climate

351

night temperatures regularly result in the formation of a thin film of ice (2–5 mm) that thaws by mid-morning in spring and summer. Low average relative humidity (33%) from April through December, scarce snow precipitation (30–100 mm/year), and high evaporation (500–1000 mm/year) generate a strong negative water balance. Because of its extreme geophysical environment, rapid climate change, isolation, and a mostly uncharted ecosystem at the start of the study in 2002, the Licancabur is both a representative of an end-member class of terrestrial lakes (Cabrol et al., 2009) and a meaningful analog to early Martian lakes.

13.3

Methods

The monitoring of the Licancabur lake used a synergetic approach including geophy­ sical, climatologic, limnologic, and biological techniques. The environment was investigated with survey stations, including meteorological dataloggers and UV dosimeters. HLP maintained two main stations: one on the shore of the lake at 5916 m and one at the base of the volcano (4340 m) to characterize the influence of altitude on environmental factors (Hock, 2008). Meteorology was surveyed with two HOBO 12-channel stations continuously logging air temperature, relative humidity, and wind speed. Solar irradiance was monitored with ELDONET dosimeters (Häder et al., 2007). These ground stations are three-channel waveband dosimeters with filter functions covering the UVB (280–315 nm), UVA (315–400 nm), and photosynthetic active radiation, (PAR, 400–700 nm). The dosimeters measure the irradiances in the defined wavelength band. The filter function matches the selected wavelengths. The UVB filter has a peak wavelength at 296 nm, a half-band of 30 nm, a tenth width of 55 nm, maximal transmission of 24%, and a diameter of 25 mm. The UVA and PAR filters are broad­ band filters 20 mm in diameter. PAR covers the visible range (93% transmission, 50% average 450–675 nm, 92% transmission at 722 nm, and 84% transmission at 531 nm). The Eldonets also measured air temperature, which provided redundancy and control for the results obtained with the HOBO dataloggers. A Solar Light Company (SLC) PMA 1122 UV sensor was used to collect data between 260 and 270 nm (+2 nm). The SLC instrument was deployed for local and regional point measurements in 2006. As a result of positive readings on Licancabur, a UV dosimeter dedicated to these short UV wavelengths was deployed for yearly data integration in 2007 on a neighboring volcano and had similar specification as the SLC sensor. Test data were collected at 4500 m before deployment at higher altitude (see Section 13.4.2) with results consistent with those obtained in 2006 at similar elevation at the foot of the Licancabur volcano. The characterization of the geophysical environment was completed by the analy­ sis of the NASA TOMS and Aura-OMI data. Aura-OMI global 0.25°0.25° gridded products and raw data were used to compile ozone total column data based on the TOMS algorithm (OMTO3e) and on the DOAS algorithm (OMDOAO3e). The radiative cloud fraction data were also acquired from OMI at similar resolution. Erythemal UV was obtained from our Eldonet stations on the ground and OMI data.

352

Lakes on Mars

The bathymetric and surface water temperature maps of the lake were completed using a 52-cm-long remote-controlled boat equipped with a GPS, a sonar, a thermal mapper, and an onboard Eagle Fish Elite 480 computer. This bathymetric system was especially designed by the HLP team in response to the need for lightweight data acquisition equipment when climbing at high altitude. It produced the first bathy­ metric and surface water temperature maps of Licancabur, which contain 3693 and 1881 measurement points for depth and temperature, respectively, and are accurate for latitude and longitude to GPS precision (+6 m). Data were acquired in a grid pattern. Occasional deviations from the pattern were due to wind-related wavelets. The margins of the lake, as shown in the maps, correspond to boat reachability and not necessarily shore contour, the difference being usually less than a meter. The photo and video documentation of the lake’s habitat was completed by a team of three divers on O2 CODE rebreathers (Morris et al., 2007) during two dive sessions averaging 30 minutes each on November 19 and 20, 2006. The first dive was performed to complete a reconnaissance of the lake’s habitat and the second to collect sedimentologic and biological samples along a transect covering representative depths. The comprehensive biological investigation of the lake started in 2006 with reconnaissance sampling of a mostly unknown ecosystem. Zooplankton was collected through standard plankton netting techniques (horizontal tows) using 10–50 µm nets. Sediment samples were analyzed for microbial community composition. They were collected in 2006 by the HLP diving team over the depth profile of the lake. The GPS unit of the bathymetric system was used by the divers to co-register the location of bottom sediment samples on the maps. Results from the preliminary investigation of phytoplankton were detailed in Acs et al. (2003), Kiss et al. (2004), and Cabrol et al. (2007b) and are not discussed here. DNA was extracted from seven of the sediment samples (20 cm, 50 cm, 2 m, 2.5 m, 3.0 m, 3.5 m, and 4.0 m) with a PowerSoil DNA Purification Kit (MoBio, Carlsbad, CA). Polymerase Chain Reaction (PCR) analyses of the extracted DNA with 16S rDNA primers BAC27F and BAC1492R were used to detect the presence of members of the domain Bacteria. PCR analyses with primers ARC25F and ARC915R allowed the detection of Archaea. Presumptive sulfate-reducing bacteria (SRB) were detected with the group-specific primer SRB385F and bacterial primer BAC805R. Denaturing gradient gel electrophoresis (DGGE) analysis of Bacteria utilized DNA extracted from the same seven sediment samples and either 16S rDNA primers BAC341F-GC and BAC534R for bacteria (Muyzer et al., 1993) or ARC340F-GC and ARC519R for Archaea. Bacterial community compositions were derived from sequences of 16S rDNA PCR amplicons prepared using Taq DNA polymerase and primers BAC27F (5′-AGAGTTTGATCCTGGCTCAG-3′) (Dojka et al., 1998) and BAC805R (5′-GACTACCAGGGTATCTAATCC-3′) (Takai and Horikoshi, 1997) and cloned into the pCR2.1-TOPO vector (Invitrogen, Carlsbad, CA). PCR-amplified inserts were screened for correct size and then their sequences determined by SeqWright (Houston, TX). Tentative taxonomic identifi­ cations were assigned using the Classifier of the Ribosomal Database Project (RDP) II (http://rdp.cme.msu.edu/).

Dynamics of declining lake habitat in changing climate

13.4

353

Results

13.4.1 Climate Meteorological conditions and overall climate evolution at Licancabur were found to be directly related to the yearly reach of the so-called altiplanic winter, a climate phenomenon occurring during austral spring and summer which brings the major part of the yearly precipitation. They also relate to the altitude, the proximity to the Atacama desert (40 km), and to a lesser extent, to winter weather systems from the Pacific ocean. Total yearly precipitation was £100 mm/year on average with exceptional years (extremes: 30–195 mm/year). Months with no measurable precipitation were common. Most precipitation occurred during austral spring and summer when humid air originating from Argentina reaches the area (Figure 13.2). The location of Licancabur 70W 20S

60W 70W

60W 70W

60W 70W

60W 20S 800 720 640 560 480 400 320 240 160 80 0

30S 20S

30S 20S 800 720 640 560 480 400 320 240 160 80 0

30S 20S

30S 20S 800 720 640 560 480 400 320 240 160 80 0 30S 20S

30S 20S

800 720 640 560 480 400 320 240 160 80 0 30S 70W

30S 60W 70W

60W 70W

60W 70W

60W

Figure 13.2 Role of the altiplanic winter in total yearly precipitation accumulation between 2003 and 2006. The Licancabur area is shown by a black dot on the maps. Credit data: The data shown on the map were acquired using GIOVANNI as part of the NASA’s GES-DISC.

354

Lakes on Mars

at the western fringe of the altiplanic winter’s zone of influence and near the Atacama makes its yearly fluctuations key to the lake’ survival. Between 2002 and 2007, 80% of total yearly precipitation was from this event alone. In order to assess the lake’s water balance, evaporation was modeled following Hamon’s method (1961). Methods and results are detailed in Hock (2008) and predict a 500–1000 mm/year negative water balance, consistent with results from other, lower, lakes studied in the region (Corripio and Purves, 2002). Substantial evaporation was confirmed by yearly observation but the trend is not monotonical. In 2006, after several years of consecutive decline, the water level rose back higher than its 2002 level following the most abundant precipitation accumulation in 5 years.

13.4.2 Physical environment Maximum UV flux (UVA + UVB) at summer solstice is 116 W/m2, or 160% that of sea level (Cabrol et al., 2009; Hock, 2008). The 2003–2004 survey gave daily peak values (averaged every 10 minutes) of PAR and UVA of 460 W/m2 and 86 W/m2, respectively. Higher values up to 512 W/m2 PAR were recorded several days in a row in 2006 during point-measurements with the SLC instrument. In 2007, with no station recording on Licancabur, over 700 W/m2 (result averaged every minute) was recorded on Simbad, a neighboring 5930 m-high volcano. Averaged UVB flux at solar noon peaked at  4 W/m2. The daily integrated erythemally dose is a factor of two greater compared to sea level values of 180–360 mW/m2 obtained for the same latitude by OMI. Shorter wavelength radiation between 260 and 270 nm was also detected by the SLC sensor both on the lakeshore and regionally. Maximum values of 9.6–14.6 mW/m2 were reached between noon and 1:00 pm local time (Figure 13.3). Because of the spectral characteristics of the PMA 1122 SLC sensor, UVB bleeding cannot be completely ruled out. The presence of abundant short UV was further confirmed in November 2009 when a different dosimeter was deployed in the altiplano at 4300 m elevation and recorded peak values of 10 mW/m2 (Cabrol et al., 2010). High solar irradiance is combined with low (0.25 yearly average over 3 years) radiative cloud fraction and year-round low total ozone concentration due to low latitude and high altitude (Figure 13.4a). The 2004–2007 TOMS OMI data show a yearly total ozone average of 248 DU and a mean yearly minimum of 232 DU. Each year values decline to ozone hole threshold (£220 DU or 40% ozone depletion) for 30–35 days. The lowest value recorded was 205 DU on March 10, 2006. The daily ozone fluctuations remain on average close to +3%, with major and minor oscillations (Figure 13.4b) generating a stable and permanently depleted ozone environment. The largest oscillation in 3 years was observed between June 1–3, when the mean total ozone concentration varied from 247 to 223 DU in 24 hours and came back to 247 DU the following day. Within fluctuation cycles, ozone lows are followed by rapid and complete recovery within a few days. These specific patterns are consistent with models of ozone mini-hole formation (James et al., 1997; Newman et al., 1988) described outside the polar regions as the result of the correlation between local short-term fluctuations of total ozone and tropospheric weather systems (Stenke and Grewe, 2004).

Dynamics of declining lake habitat in changing climate 600

Summer solstice Fall equinox Winter solstice Spring equinox

500

355

100

PAR

80

UVA

Summer solstice Fall equinox Winter solstice Spring equinox

Watt/m2

Watt/m2

400 300

60

40 200 20

100 0 00:00 3.5 3

06:00

12:00 Time

18:00

Summer solstice Fall equinox Winter solstice Spring equinox

0 00:00

24:00

06:00

12:00 Time

18:00

24:00

15

UVB

4500 m: 09/11/2007 4500 m: 08/11/2007 5870 m: 28/11/2006

260–270 nm

2.5

mW/m2

Watt/m2

10

2 1.5

5

1 0.5 0 00:00

06:00

12:00 Time

18:00

24:00

0 00:00

06:00

12:00 Time

18:00

24:00

Figure 13.3 Typical solar flux at Licancabur (data are averaged every 10 minutes). Observations (2006 and 2007) show measurable short wavelength flux at the surface (260–270 nm). The ecological effects of these wavelengths are not yet characterized.

Deviations from the mean show annual cycles: one high from June to December (winter to spring) and one low from January to May (summer to fall). The timing of the low cycles is anti-correlated with the formation of the ozone hole in Antarctica that culminates in October–November (e.g., Lubin et al., 1992; Pitari et al., 1992; Schoeberl et al., 1989; Solomon, 1990; and many others) while the lowest values at Licancabur are reached in May–June. The beginning of each cycle is correlated with winter and summer solstices though exceptions were observed. For instance, significant perturba­ tions in the ozone and increased UV radiation (+12.5% erythermal UV) were noted in September 2005, when the total ozone concentration went from a normal seasonal average of 245 to 215 DU. This period corresponded to the largest extent of the Antarctic ozone hole which was one of the deepest and largest recorded so far (British Antarctic Survey (BAS) Ozone Bulletins, September 23 and October 13, 2005, unpublished). While Licancabur is not considered in a region normally affected by the Antarctic ozone hole, previous studies have suggested that in some instances, its effect was

356

Lakes on Mars

Total column ozone (Dobson Unit, DU)

260

Daily ozone fluctuation (%)

270

200

Licancabur area Minimum ozone

250 240 230 220 210 a

10

0

–10 Oct 04

b Apr 05

Oct 05

Apr 06

Oct 06

Apr 07

Oct 07

Figure 13.4 (a) Total ozone concentration (daily minimum) at Licancabur between October 2004 and December 2007 (OMI data). Yearly cycles with lows in April–June reach below the threshold of ozone hole (£220 DU); (b) Daily ozone variations with minor oscillations showing recovery from deviations within a few days. Minor oscillations gradually increase in amplitude until ozone reaches its minimum. Maximum oscillations appear generally associated with periods of minimum ozone concentration.

perceived as high as Antofagasta, Chile (23oS) (e.g., Jaque et al., 1994). In its bulletin, BAS noted that the edge of the ozone hole reached South America in several instances between August 30 and September 27. Each episode was well correlated with fluctua­ tions in the ozone and UV data in the Licancabur area within 24 hours (Cabrol et al., 2009). This combination of year-round low ozone concentration with small amplitude variations and high solar irradiance makes Licancabur a distinct environment for life compared to other bodies of water, including those directly within ozone hole range. Up to 65% variation in ozone concentration has been measured in survey stations in Antarctica and 55% in high latitudes, i.e., Punta Arenas, Chile (e.g., Casiccia et al., 2003; Chubachi, 1997; Frederick and Lubin, 1994; Galtier et al., 1994; Herman et al., 1993). These stations also measure a much lower minimum ozone (95–195 DU) than

Dynamics of declining lake habitat in changing climate

357

Licancabur during these events. However, they have substantially higher total ozone values for a large part of the year (339–399 DU) and lower solar angle resulting in lower solar irradiance (e.g., Jung et al., 2006; Stanhill and Cohen, 1997), i.e., UVB  1.4 µW/cm2 at the McMurdo stations. As a result, while ozone hole events in Antarctica subject ecosystems to seasonal crises, Licancabur’s severe ozone depletion is a permanent factor; UVB is consistently, and substantially higher than in these higher latitude and lower altitude environments and short UV wavelength are detected on the ground.

13.4.3 Lake chemistry Habitability is also impacted by water chemistry, and interannual variability may increase stress on microorganisms. The Licancabur lake’s mean total dissolved solid content (TDS = 1096 mg/L) results in clear waters. Measurements of solar irradiance as a function of depth yielded the optical depth (κ) as a function of wavelength, with κUVB = 0.53 m, κUVA = 1.10 m, and κPAR = 10.1 m (Hock, 2008) implying that beyond the UVB cutoff region, life is sheltered from radiation damaging DNA replication and transcription processes (e.g., Bothwell et al., 1994; Häder, 1993; Vinebrook and Leavitt, 1996; Williamson et al., 2002), while photosynthesis remains possible. Despite intense solar UV irradiance, zooplankton and microbial organisms, including periphyton attached to rocks, are observed in abundance above the UVB cutoff limit, many of them dwelling at the surface of the lake at solar noon. Each year, water samples were collected mid-November during field deployment. Like other small shallow lakes (Douglas and Smol, 1994), Licancabur’s chemistry is highly susceptible to atmospheric changes, such as decreased precipitation. For example, ionic concentrations increased from 2002 to 2005 (in ppm: TDS = 1050– 1429; SO4 = 589–1010; Ca = 230–298; Cl = 58–184; Na = 67.5–99; Mg = 39.6–54.9) following several years of scarce winter precipitation. Surface pH also shows interannual variations (pH 8.5–7.6, 2002–2005) and variations with depth. The pH of waters shallower than 2 m ranges from 6.5 to 7.0, while waters deeper than 2 m vary from pH 7.5 to 8.1 suggesting a stratification in the water column. The increased salinity levels in the shallowest waters may drive the decrease in pH. Phosphate – (PO3– 4 ) and nitrite (NO2) plus nitrate (NO3) concentrations are at or below detection limits (0.02 ppm, 0.01 ppm, respectively). Given the site’s isolation, N and P cycling is likely very tightly controlled, with their biologically available forms being rapidly taken up by primary producers. Geothermal waters are a source of phosphorus inputs (Pringle, 1991). As such, the low concentrations of these biologically important nutrients put into question whether or not the lake continues to be, or ever was, internally supplied. Its content in Cl– + SO4 and pH would classify it as a quiescent lake (Varekamp et al., 2000); however light, but steady, degassing was observed for the first time in 2006 in the shallow waters on the northeast shore. The surface water temperature map (Figure 13.5) shows a co-location between the highest temperatures and this degassing site. HLP did not return to Licancabur in 2007 and the nature of the degassing remains to be investigated and its impact on life evaluated. The 2.44 mg/L mean dissolved organic carbon (DOC) concentration of the surface waters is low and comparable to values from high Arctic lakes (Lim and Douglas,

358

Lakes on Mars

North

Temperature (°C)

14.2

Lake surface

0 Meters 0

30

North 0

a

κUVB 0.53 m

b

κUVA

a Depth (m)

b

Meters 0

1.10 m

5.2 30

Figure 13.5 The surface water temperature and bathymetric maps of Licancabur were generated by HLP in November 2006 between 11:00 am and 1:30 pm local time and constitute the first geophysical maps of the lake. The surface water temperature structure suggests a two-cell convection system with warmer waters near shore and cooler waters sinking at the center of the lake. The black dots on the temperature map indicate the location of the samples collected at 50 cm and 4 m depth discussed in the text in Section 13.4.5. The image of the water column to the right shows the evolution of copepod density (red microorganisms) as a function of depth and UVA and UVB optical depths. Credit photo: 2004 HLP expedition.

2003; Michelutti et al., 2002). In these regions, they have been attributed to runoff characteristics related to the lack of vegetation in the catchment areas, a common factor with Licancabur. DOC concentrations have been identified as one of the key controlling factors in determining the level of underwater penetration of UV into aquatic systems (e.g., Vincent and Pienitz, 1997). As such the DOC concentration in the Licancabur surface water has a direct effect on the high UV penetration. Since DOC strongly absorbs the more biologically damaging, shorter UV wavelengths (Scully and Lean, 1994), the low concentration may bring little protection to the microbial organisms dwelling in the upper waters.

Dynamics of declining lake habitat in changing climate

359

Understanding the impact of the physical and chemical environment on life’s distribution in the water column required a three-dimensional vision of the lake. Licancabur’s bathymetry was therefore completed in November 2006 (Figure 13.5). The underwater topography showed a 5.2 m deep, 5000 m2 crater lake. While no previous measurements were available, ancient shoreline levels in the caldera could indicate a lake possibly 10–12 m deep at its maximum. The average surface water temperature is +4.3°C with significantly warmer areas (+14°C) co-located with the site of recently observed degassing.

13.4.4 Zooplankton In such an environment, early geographical and archeological expeditions (Leach, 1986; Rudolft, 1955) mentioned the presence of a scarce zooplankton community in the lake. By contrast, the plankton net collection and the HLP diving expedition of 2006 documented a dense population of metazoan zooplankton species, including at least two species of calanoid copepod, two species of ostracod, three species of cladoceran, a rotifer, and the larvae of a chironomid (Cricotopus sp.). The latter is the highest occurrence for this midge with the previous record being 5600 m in the Himalayas (Saether and Willassen, 1987). Not all are identified to species level (Cabrol et al., 2009). In most cases, the species identified are widely distributed over South America, particularly in the altiplano and high Andes (Williams, 1995). The distribution of at least one species, the rotifer Notholca walterkostei reaches the sub-Antarctic islands and Antarctica (Izaguirre et al., 2003). The cladoceran Daph­ niopsis chilensis was first described from Licancabur (Hann, 1986) and has not been reported from any other location. Whether this species is truly endemic to the lake is uncertain; many regional lakes have yet to be investigated and other unidentified members of this genus have been recorded from nearby bodies of water (Williams, 1995). Most species are pigmented, indicating adaptation to high-UV. Swarms of the orange-pigmented copepod Boeckella sp. form near the sediment across the lake; however, their behavior varies with depth. At the time of the observation (midNovember, close to local noon) copepods not dwelling at the surface or near the UVB cutoff region clustered almost exclusively between rocks or near rocks. Below 1 m, most were observed migrating in the water column or lying on the lakebed sediment. Whether this relates to a UV protection strategy is unclear. A late stage copepodite or adult was found frozen within the ice cover and resumed swimming immediately on the melting of the ice after being frozen for 9 months. This shows a possible over-wintering strategy in adults (diapause) similar to that of some Antarctic copepods (Atkinson, 1998) with likely breeding soon after ice thaws. Copepods, cladocera, and rotifers occur in Antarctic lakes, where they can survive with permanent ice covers over 5 m thick as well as in small pools that freeze over during winter (Gibson, 2007). There, all species have adapted their lifestyle to survive winter, either as an over-wintering egg or at a stage that does not develop during winter. The combination of high-UV and year-round low atmospheric ozone, as observed at Licancabur, is unprecedented, and the occurrence of a complex

360

Lakes on Mars

community suggests that these species are either UV resistant as a result of their pigmentation, can easily adapt by positioning themselves in the water column to minimize exposure (Figure 13.5), and/or have adapted their biological cycles to short favorable seasonal windows.

13.4.5 Microbial community The role played by microorganisms in the recycling of lake nutrients was investigated through the analysis of samples from the sediment–water interface. Results show that sulfur cycling plays an important role in the lake biogeochemistry. The samples were collected along a gradient of increasing depth in 10-cm intervals from depths of 10–50 cm along the north shore and below that by divers in 0.5-m intervals to a depth of 4 m. The sediments fall into two classes: above 2 m depth they consist mostly of tan-colored mm-sized sand grains with rusty fines and a pH of 6.5–7.0, and below 2 m they are made of a fine black material with a strong hydrogen sulfide odor and a pH of 7.5–8.0. The 2 m depth appears to be an important transition in the lake’s structure as sediment type and pH change and total organic carbon increases more than 300-fold, from 0.04 to 0.05% in the shallow sediments to 13.7–18.9% in the 2.5–4 m sediments. An exception to this pattern was the plate counts of total aerobic heterotrophic bacteria that were patchy (Figure 13.6a). The temperature response indicates most aerobic heterotrophs are psychrotolerant although the 4-m sample probably includes psychro­ philes. These data suggest that biomass from the water column accumulates in the deepest parts of the lake and nutrients from that biomass are recycled by microorgan­ isms using sulfate as the terminal electron acceptor.

Aerobic heterotrophic bacteria Total organic carbon 5 10 colony forming unit/mL slurry (% g c/g dry sediment) 0 5 10 15 20 0 10 20

Depth (m)

a 0.01 0.1 0.2 0.3 0.4 0.5 1.0 1.5A 1.5B 2.0A 2.0B 2.5 3.0 3.5 4.0

Bacterial community composition (% total sample population)

b

Sample Depth (m) 0.5 4m

UVB Cut-Off Region UVA Cut-Off Region

c

0.5 m

T=4°C T=17°C T=25°C T=33°C

0

5 10 20 6 Aerobic heterotrophic bacteria 105 colony forming unit/mL slurry

7 pH

8

Proteobacteria 31.9 Bacteroidetes 29.2 Unclassified 12.5 Actinobacteria 4.2 Acidobacteria 4.2 Verrucomicrobia 2.8 Firmicutes 2.8 Nitrospira 2.8 Cyanobacteria 1.4 Gemmatimonadetes 1.4 Planctomycetes 1.4 TM7 1.4 Spirochaetes

34.8 42.0 17.4

4.3

1.4

4.0 m

Figure 13.6 Analysis of surface sediment samples collected at incremental depths in the Licancabur lake—(a) Left: enumeration of total aerobic heterotrophic bacteria on R2A agar plated at either 4°C and 17°C or 4°C, 17°C, 25°C, and 33°C; Right: total organic carbon in dried sediment and pH of sediment slurries. (b) Bacterial community composition of the sediment sample from a lake depth of 50 cm; percentages are based on a total of 72 clones out of 96 sequenced after eliminating chimeras. (c) Bacterial community composition of the 4 m sediment sample (based on 69 clones out of 96 sequenced).

Dynamics of declining lake habitat in changing climate

361

Bacteria, including SRB, were detected in all sediment samples, whereas Archaea were only present at depths  2.5 m. DGGE analysis showed substantial differences among the patterns of bacteria in the 20 cm, 50 cm, and the 2–4 m samples with the latter having minor sample-to-sample differences. Consistent with the two distinct geochemical classes of lake sediments, there were significant differences between the bacterial community compositions for the 50-cm (UVB cutoff region) and 4-m samples (Figures 13.6b, c). Both were dominated by members of just two phyla, the Bacteroidetes and the Proteobacteria (61% at 50 cm and 77% at 4 m), consistent with results reported in recent studies of two altiplanic lakes (Huasco and Salar de Ascotán, Chile) (Demergasso et al., 2004; Dorador, 2007) located between 2350 and 3900 m elevation and lake Quinghai, an athalassohaline lake located in the Tibetan Plateau of Northwest China at 3200 m (Dong et al., 2006). The Licancabur samples also con­ tained a small contribution from a third phylum, the Firmicutes (2.8% and 4.3%, respectively). Beyond these, the 50-cm sample contained members of eight additional phyla while there was only one other phylum represented in the 4-m sediment. Both sediments presented a significant fraction of cloned sequences currently unclassified at the phylum level (12.5% and 17.4%, respectively). The vast majority is also unclassified at the genus level (74% and 68%, respectively). The narrow range of diversity at 4 m is even more apparent when analyzed at the genus-species level. Thirteen of the 69 clones (19%) are in the genus Thiobacillus (class Betaproteobacteria) and, using a criterion of 95% sequence similarity, are the same species. Another 5 clones (7% of the total) are the same species in the genus Thiomicrospira (class Gammaproteobacteria) and 7 of the 29 clones in the phylum Bacteriodetes (10% of the clone library), though unclassified, are predicted to be the same species. Thus, 36% of the bacteria in the 4-m sediment sample are potentially represented by just 3 species of bacteria. Given the presence of sulfides in the 4-m sample, it is surprising that only 3 out of 69 clones (4%) are predicted to be in the class Deltaproteobacteria and thus candi­ dates for SRB. Since sulfate reduction consumes protons, the elevated pH in the 4-m sample is also consistent with sulfate reduction being an important component of the metabolic activity. Archaeal species in the 4-m sample (present by PCR and DGGE, but not yet quantified) could be methanogens that alternatively oxidize carbon com­ pounds linked (directly or via syntrophy) to sulfate as electron acceptor (Boetius et al., 2000; Orphan et al., 2001) since methane formation seems unlikely in the presence of 500 mg/L sulfate (Winfrey and Zeikus, 1977). Regardless, 26% of the clones in this sample are Thiobacillus or Thiomicrospira species and thus members of the genera of sulfur-oxidizing chemoautotrophic bacteria. Thiobacilli could be growing heterotro­ phically on sediment organic carbon and using sulfur oxidation only for energy generation (Perez and Matin, 1980) and thus increasing their relative numbers. The apparent disparity in sulfur oxidizing and reducing bacteria may reflect a seasonal variation in the lake’s sulfur cycle (Bak and Pfennig, 1991; Holmer and Storkholm, 2001). The summit lake does not freeze in its deepest part during the winter, and the ice cover maintains a water temperature between 3°C and 6.2°C; therefore, growth of SRB may continue throughout winter producing high levels of sulfides. The spring thaw brings oxygenated water and a bloom of sulfur-oxidizing bacteria. A further

362

Lakes on Mars

explanation for the high proportion of sulfur-oxidizing bacteria would be an input of hydrogen sulfide from the volcano, which could be related to the recent degassing on the northeastern shore, an hypothesis that remains to be tested.

13.5 Conclusion With discrepancies inherent to the study of terrestrial analogs, i.e., the presence of zooplankton, the overall environmental analogy of Licancabur with Mars at the Noachian/Hesperian transition period makes this Andean lake a unique analog to early Martian lakes and a window into the likely transformation they experienced during the transition period from a wetter to a drier Mars. From a climate standpoint, the demonstrated interannual variability in water input, related water level fluctuations, and the resulting changes in physicochemical properties of the water column of the lake show that the decline was unlikely to be monotonical over time on Mars, and therefore probably challenging for putative life to adapt to. For instance, in 2005, a particularly dry year at Licancabur with only 30 mm precipitation, the lake lost 55% of its total volume and about 2 m of water column (as estimated from the 2006 bathymetry and comparison between 2005 and 2006 photographs). In 2006, the lake level was higher than in 2002, its highest in all 6 years surveyed. The multi-year survey shows that physicochemical changes associated with such variability are significant. There are currently no data to show how this environmental roller coaster impacts biodiversity at Licancabur at the level of the overall population density or that of the species. However, the relatively low diversity shown in the samples of the microbial community are consistent with biodiversity loss and selection of specific species highly capable of mutating and adapting rapidly as shown by the compactness of their cladograms (Cabrol et al., 2009). The low atmospheric pressure and the low cloud cover combine to generate strong daily and yearly temperature variations. Cloud passage results in rapid and severe fluctuations of both temperature and UV flux such as those shown in Figure 13.7. Inverse relationships between UV and temperature were observed during those fluc­ tuations, with summer daily temperature gradients between 3.75 and 6.3°C per hour, and significant dip and recovery patterns up to 10°C within 1 minute in some cases. This specific set of stresses would have been common on Mars as the atmosphere was thinning. Previous studies have documented that it impairs life’s ability to repair DNA and, as a response, organisms were observed using the water column to shelter themselves (e.g., Williamson et al., 2002; Zagarese et al., 1997). This survival strategy is efficient as long as the water column is deep enough (e.g., deeper than the UVB cutoff region) but becomes a dead-end as soon as the water level becomes shallower than this limit. While this behavior was noted at Licancabur for some of the zooplankton species, other pigmented species have been observed swimming at the surface at solar noon following their food sources (e.g., phytoplankton). Also, as shown in the Andes, as lakes evaporate, brine concentration tends to add solid

Dynamics of declining lake habitat in changing climate

120

363

8 UVA

7

100 6

UVA (W/m2)

80 5 4

60

3 40 2 20

UVB (W/m2) and temperature (°C)

Temp UVB

1 January 22, 2007

0 0 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00 Local time

Figure 13.7 Occurrences of inverse relationships between UV radiation and temperature.

particulates in the water, which helps UV scattering and might temporarily mitigate the effect of a reducing water column. However, survival requires species having specific osmotic characteristics in order to adapt to increased salinity (see also Chapter 11). From a physical perspective, data suggest that the timing of key environmental cycles could be critical to the ecosystem’s survival. For instance, minimum ozone occurs at the end of the austral fall (May–June) when typical daily UV flux peak nears its lowest values. This is also the time when the lake freezes. The ice cover mitigates UV radiation and water temperature fluctuations, making the lake a safer and more stable habitat. Thawing occurs in September when ozone concentration reaches its maximum yearly value. These observations suggest that the combination of extreme atmospheric and environmental factors is what governs habitability and life rather than their sum. Yet even minimum UV radiation and maximum ozone concentration at Licancabur must be regarded as some of the most severe on our planet. They are further combined with low, yearly average temperatures including high daily variability and sharp fluctuations due to the thin atmosphere, ice, a rapidly changing climate, and fast chemical changes making Licancabur an end-member environment, possibly one of the closest terrestrial lake habitat to those of early Mars.

364

Lakes on Mars

Acknowledgment The authors thank SERNAP and the Direction and Park Rangers of the Reserva Eduardo Avaroa in Bolivia, the Universidad Catolica del Norte in Antofagasta, Chile, and NASA Ames EERRB for their support to this project. The High Lakes Project (HLP) was funded through the NASA Astrobiology Institute (NAI)/SETI team grant No. NNA04CC05A. Other financial support to the HLP has been provided over the years by the NASA Ames Directorate Discretionary Funds, Planetary Biology Internship programs, the Lewis and Clark Foundation, National Geographic Research Grant, Wings WorlQuest, AquaLung, and Specialized. Some of the data used in this study were acquired using the GES-DISC Interactive Online Visualization ANd aNalysis Infrastructure (Giovanni) as part of the NASA’s Goddard Earth Sciences (GES) Data and Information Services Center (DISC). We are also particularly grateful to the Biological Sciences Division, British Antarctic Survey, Cambridge (UK) for allowing the use of their UV database for the Rother Research Station, Antarctica.

References Acs, E., et al., 2003. Similarities and dissimilarities in biodiversity of three high-altitude mountain lakes (Andes, Bolivia). 6th Hungarian Ecological Congress. In: Dombos, M., Lakner, G. (Eds.), St. Stephan University, Publishers, Godollo, 305. Allen, M., Miller, C., Webster, C., Wennbe, P. 2005. Exploration for Habitability and Life in the Solar System and Beyond: an Earth Analogue, White paper, 2 pages. Anderson, R.B., Bell, J.F.2008. Stratigraphy of phyllosillicates and sulfates in northern Meridiani planum, Mars. AGU, Fall Meeting, Abstract No. P43B–1393. Andrews-Hanna, J.C., Phillips, R.J., Zuber, M.T., 2007. Meridiani planum and the global hydrology of Mars. Nature 446, 163–166. Arvidson, R.E., Ruff, S.W., Morris, R.V., Ming, D.W., Crumpler, L.S., Yen, A.S., et al., 2008. Spirit Mars Rover Mission to the Columbia Hills, Gusev Crater: Mission overview and selected results from the Cumberland Ridge to Home Plate. J. Geophys. Res. 113, E12S33, doi:10.1029/2008JE003183. Arvidson, R.E., Squyres, S.W., Anderson, R.C., Bell, J.F. III, Blaney, D., Brückner, J., et al., 2006. Overview of the Spirit Mars Exploration Rover Mission to Gusev Crater: Landing site to Backstay Rock in the Columbia Hills. J. Geophys. Res. 111, E2, E02S01 10.1029/ 2005JE002499. Atkinson, A., 1998. Life cycle strategies of epipelagic copepods in the Southern Ocean. J. Mar. Syst. 15 (1–4), 289–311. Bak, F., Pfennig, N., 1991. Microbial sulfate reduction in littoral sediment of Lake Constance. FEMS Microbiol. Lett. 85, 31–42. Baker, V.R., 2001. Water and the Martian landscape. Nature 412, 228–236. Baker, V.R., 2004. A brief geological history of water on mars. In: Seckbach, J., (Ed.) Cellular Origin, Life in Extrem Habitats and Astrobiology. Springer Netherlands Book Series, Dordrecht, The Netherlands, vol. 6. pp. 621–631. Beaty, D.W., Clifford, S.M., Borg, L.E., Catling, D., et al., 2005. Key science questions from the second conference on early Mars: Geologic, hydrologic, climate evolution, and the implications for life. Astrobiology 5, 663–689. Bibring, J.P., Langevin, Y., Gendrin, A., Gondet, B., Poulet, F., Berthé, M., et al., 2005. Mars surface diversity as revealed by the OMEGA/Mars express observations. Science 307 (5715), 1576–1581.

Dynamics of declining lake habitat in changing climate

365

Boetius, A.K., et al, 2000. Microscopic identification of a microbial consortium apparently mediating anaerobic methane oxidation above marine gas hydrate. Nature 407, 623–626. Bothwell, M.L., Sherbot, D.M.J., Pollock, C.M., 1994. Ecosystem response to solar ultravioletB radiation: Influence of trophic level interactions. Science 265, 97–100. Brack, A., Clancy, P., Fitton, B., Hoffmann, B., Horneck, G., Kurat, G., et al., 1998. Search for life on Mars. Biol. Sci. Space 12 (2), 119–123. Bradley, R.S., Vuille, M., Diaz, H.F., Vergara, W., 2006. Threats to water supplies in the tropical Andes. Science 312, 1755–1756. Cabrol, N.A., et al., 2009. The high-lakes project: science overview. J. Geophys. Res. Bio­ geosci. 114, 1–21. Cabrol, N.A., Andersen, D.T., Stoker, C.R., Lee, P., McKay, C.P., Wettergreen, D.S., 2010. Other analogs to mars: high altitude, subsurface, desert, and polar environments. In: Doran, P. (Ed.), Life in Antarctic Deserts and Other Cold Dry Environments: Astrobiological Analogues. Cambridge University Press, New York. Cabrol, N.A., Grin, E.A., Hock, A.N., 2007a. Mitigation of environmental extremes as a possible indicator of extended habitat sustainability for lakes on early Mars. Proc. SPIE Conf., 6694–36, 1, San Diego, 15 pages. Cabrol, N.A., Grin, E.A., Kiss, K.T., Ács, E., Grigorszky, I., Szabò, K., et al., 2007b. Signatures of habitats and life in Earth’s high-altitude lakes: clues to Noachian aqueous environments on mars. In: Chapman, M. (Ed.), The Geology of Mars, Chapter 14: Evidence from EarthBased Analogs. Cambridge University Press, New York, pp. 349–370. Cabrol, N.A., Herkenhoff, K.E., Greeley, R., Grin, E.A., Schröder, C., d‘Uston, C., et al., 2008. Soil sedimentology at Gusev Crater from Columbia Memorial Station to Winter Haven. J. Geophys. Res. 113, E06S05, doi:10.1029/2007JE002953. Carr, M.H., 1996. Water on Mars. Oxford University Press, New York. Casiccia, C., Kirchhoff, V.W.J.H., Torres, A., 2003. Simultaneous measurements of ozone and ultraviolet radiation: Spring 2000, Punta Arenas, Chile. Atm. Environ. 37/3, 383–389. Chubachi, S., 1997. Annual variation of total ozone at Syowa station, Antarctica. J. Geophys. Res. 102, 1349–1354. Clifford, S.M., 1993. A model for the hydrologic and climatic behavior of water on Mars. J. Geophys. Res. Planets 98, 10973–11016. Connerney, J.E.P., Acuña, M.H., Wasilewski, P.J., Ness, N.F., Rème, H., Mazelle, C., et al., 1999. Magnetic lineations in the ancient crust of Mars. Science 284 (5415), 794–798. Corripio, J., Purves, R.2002. Energy balance of high altitude glacierised basins in the Central Andes: Climatic and hydrological aspects. 27th EGS General Assembly, Abstract No. 4402. Craddock, R.A., Howard, A.D., 2002. The case for rainfall on a warm, wet early Mars. J. Geophys. Res. 107 (E11), 5111, doi:10.1029/2001JE001505. de Silva, S., Francis, P., 1991. Volcanoes of the Central Andes. Springer-Verlag, Berlin. Demergasso, C., et al., 2004. Distribution of prokaryotic genetic diversity in athalassohaline lakes of the Atacama Desert, Northern Chile. FEMS Microbiol. Ecol. 48, 57–69. Des Marais, D.J., Farmer, J.D., 1995. The search for extinct life. In: Meyer, M., Kerridge, J. (Eds.), An Exobiological Strategy for Mars Exploration, NASA Ames Research Center, California, USA, pp. 21–25. Dojka, M.A., Hugenholtz, P., Haack, S.K., Pace, N.R., 1998. Microbial diversity in a hydrocarbon- and chlorinated-solvent-contaminated aquifer undergoing intrinsic bioremediation. Appl. Environ. Microbiol. 64 (10), 3869–3877. Dong, H., et al, 2006. Microbial diversity in sediments of saline Qinghai Lake, China: linking geochemical controls to microbial ecology. Microbio.l Ecol. 51, 65–82.

366

Lakes on Mars

Dorador, C., 2007. Microbial communities in high altitude altiplanic wetlands in northern Chile: phylogeny, diversity and function. PhD thesis. Max Planck Institut für Evolutions biologie. Christian-Albrechts-Universität, Kiel, 170 pp. Douglas, M.S.V., Smol, J.P., 1994. Limnology of high Arctic ponds (Cape Herschel, Ellesmere Island, N.W.T.). Arch. Hydrobiol. 131, 401–434. Farmer, J.D., 1995. Mars Exopaleontology. Palaios 10 (3), 197–198. Farmer, J.D., 2000. Strategies for the astrobiological exploration of Mars, Concepts and approaches for Mars exploration, part 1, Lunar and Planetary Institute Contribution No. 1062, 110–112. Farmer, D.J., Des Marais, D.J., 1999. Exploring for a record of ancient Martian life. J. Geophys. Res. Planets 104 (E11), 26977–26995. Folt, C.L., Chen, C.Y., Moore, M.V., Burnaford, J., 1999. Synergism and antagonism among multiple stressors. Limnol. Oceanogr. 44, 864–877. Fornari, M., Risacher, F., Feraud, G., 2001. Dating of paleolakes in the central Altiplano of Bolivia. Palaeogeogr. Palaeoclimatol. Palaeoecol. 172 (3/4), 269–282. Frederick, J.E., Lubin, D., 1994. Solar ultraviolet irradiance at Palmer Station, Antarctica, Ultraviolet Radiation in Antarctica: Measurements and Biological Effects, Antarctic Research Series, vol. 62, pp. 43–52. Galtier, C., He, G., Yang, S., Lubin, D., 1994. Role of clouds and ozone on spectral ultravioletB radiation and biologically active dose over Antarctica. Ultraviolet Radiation in Antarc­ tica: Measurements and Biological Effects, Antarctic Research Series, vol. 62, pp. 83–91. Gibson, J.A.E., et al, 2007. Biogeographical trends in antarctic lakes. In: Bergstrom, D.M., Convey, P., Huiskes, A.H.L. (Eds.), Trends in Antarctic Terrestrial and Limnetic Ecosys­ tems: Antarctica as a Global Indicator. Springer, Dordrecht, pp. 71–98. Grad, G., Williamson, C.E., Karapelou, D.M., 2001. Zooplankton survival and reproduction responses to damaging UV radiation: A test of reciprocity and photoenzymatic repair. Limnol. Oceanogr. 46, 584–591. Grosjean, M., Ammann, B., Camacho, M., Geyh, M.A., Kern, R., et al., 2001. Mid-Holocene (ca. 6000–3500 14C yr B.P.) climates in the Atacama Altiplano of the south Central Andes: humid or dry?—Limnogeological, geomorphological, hydrological and vegetational evi­ dence for widespread mid-Holocene aridity. Workshop on Central Andes Paleoclimatol­ ogy, NSF, 11–16 January 2001, Tucson AZ. Haberle, R.M., 1998. Early Mars climate models. J. Geophys. Res. 103 (28), 467–28479. Häder, D.-P., 1993. Risks of enhanced solar ultraviolet radiation for aquatic ecosystems. Prog. Phycol. Res. 9, 1–45. Häder, D.-P., Lebert, M., Schuster, M., del Ciampo, L., Heibling, E.W., McKenzie, R., 2007. ELDONET- A decade of monitoring solar radiation on five continents. Photochem. Photobiol. 83, 1–10. Hamon, W.R., 1961. Estimating potential evapotranspiration. J. Hydraulics Div., In: Proceedings of the American Society of Civil Engineers, 87, 107–120. Hann, B.J., 1986. Revision of the genus Daphniopsis Sars, 1903 (Cladocera: Daphniidae) and a description of Daphniopsis chilensis, new species, from South America. J. Crust. Biol. 6, 246–263. Herman, J.R., McPeters, R., Larko, D., 1993. Ozone depletion at northern and southern latitudes derived from January 1979 to December 1991 Total Ozone Mapping Spectrometer data. J. Geophys. Res. 98, 12783–12793. Hock, A.N.2008. Licancabur volcano and life in the Atacama: environmental physics and analogies to Mars. Ph. D. Thesis, University of California, Los Angeles. Holmer, M., Storkholm, P., 2001. Sulphate reduction and sulphur cycling in lake sediments: a review. Freshw. Biol. 46, 431–451.

Dynamics of declining lake habitat in changing climate

367

Izaguirre, I., Allende, L., Marinone, M.C., 2003. Comparative study of the planktonic commu­ nities of three lakes of contrasting trophic status at Hope Bay (Antarctic Peninsula). J. Plankton Res. 25, 1079–1097. Jakosky, B.M., 1999. Martian stable isotopes: Volatile evolution, climate change, and exobio­ logical implications. Orig. Life Evol. Biosph. 29, 47–57. Jakosky, B.M., Phillips, R.J., 2001. Mars' volatile and climate history. Nature 412, 237–244. Jakosky, B., Westall, F., Brack, A., 2007. Planets and Life: The Emerging Science of Astrobiology. In: Sullivan, W., Baross, J. (Eds.), Cambridge University Press, New York, pp. 357–387. James, P.M., Peters, D., Greisinger, K.M., 1997. A study of ozone mini-hole formation using a tracer advection model driven by barotropic dynamics. Meteorol. Atmos. Phys. 64, 107–121. Jaque, F., Tocho, J.O., Da Silva, L.F., Bertuccelli, G., Crino, E., et al., 1994. Ground-based ultraviolet radiation measurements during springtime in the Southern Hemisphere. Euro­ phys. Lett. 28 (4), 289–293. Jung, Y., Kim, J., Cho, H., Lee, B., 2006. Recent changes in solar irradiance and infrared irradiance related with temperature and cloudiness at the King Sejong Station, Antarctica. Am. Geophys. Union, Fall Meeting, Abstract #A33B1006J. Kasting, J.F., 1991. CO2 condensation and the climate of early Mars. Icarus 94 (1), 1–13. Kiss, K.T., Acs, E., Boris, G., Cabrol, N., Grigorsky, I., et al., 2004. Habitats extrêmes pour les communautés de diatomées dans les lacs de haute altitude (Laguna Blanca et lac de cratère du volcan Licancabur, Bolivie). 23ème Colloque de l’ADLaF. Orléans, France. Leach, J.W.P., 1986. Andean high altitude expedition. Underwater Technol. 12 (1), 27–31. Leech, D.M., Padeletti, A.T., Williamson, C.E., 2005. Zooplankton behavioral responses to solar UV radiation vary within and among lakes. J. Plankton Res. 27, 461–471. Lewis, K.W., Aharonson, O., Grotzinger, J.P., Squyres, S.W., Bell, J.F., Crumpler, L.S., et al., 2008. Structure and stratigraphy of Home Plate from the Spirit Mars Exploration Rover. J. Geophys. Res. 113, E12S36, doi:10.1029/2007JE003025. Lim, D.S.S., Douglas, M.S.V., 2003. Limnological characteristics of 22 lakes and ponds in the Haughton crater region of Devon Island, Nunavut, Canadian High Arctic. Arct. Antarct. Alp. Res. 35, 509–519. Logan, B.E., Zhang, H., Mulvaney, P., Milner, M.G., Head, I.M., Unz, R.F., 2001. Kinectics of perchlorate- and chlorate respiring bacteria. Appl. Environ. Microbiol. 67 (6), 2499–2506. Lubin, D., Mitchell, B.G., Frederick, J.E., Alberts, A.D., Booth, C.R., Lucas, T., et al., 1992. A contribution toward understanding the biospherical significance of Antarctic ozone deple­ tion. J. Geophys. Res. 97, 7817–7828. Malin, M.C., Edgett, K.S., 2000a. Evidence for recent groundwater seepage and surface runoff on Mars. Science 288 (5475), 2330–2335. Malin, M.C., Edgett, K.S., 2000b. Sedimentary rocks of early Mars. Science 290 (5498), 1927–1937. Marinovic, N., Lahsen, A.1984. Hoja Calama, Region de Antofagasta. Carta geologica de Chile No. 58, 1:250,000. Serv. Nac. Geol. Min. ISSN 0716-0194. McAdam, A.C., Zolotov, M.Y., Mironenko, M.V., Sharp, T.G., 2008. Formation of silica by low-temperature acid alteration of Martian rocks: Physical-chemical constraints. J. Geophys. Res. 113, E08003, doi:10.1029/2007JE003056. McKinley, J.P., Stevens, T.O., Westall, F., 2000. Microfossils and paleoenvironments in deep subsurface basalt samples. Geomicrobiol. J. 17 (1), 43–54. Messerli, B., Grosjean, M., Bonani, G., Bürgi, A., Geyh, M.A., Graf, K., et al., 1993. Climate Change and Dynamics of Natural Resources in the Altiplano of Northern Chile during Late Glacial and Holocene Time. First Synthesis. Mt. Res. Dev. 13 (2), 117–127.

368

Lakes on Mars

Michelutti, N., Douglas, M.S.V., Lean, D.R.S., Smol, J.P., 2002. Physical and chemical limnology of 34 ultra-oligotrophic lakes and ponds near Wynniatt Bay, Victoria Island, Arctic Canada. Hydrobiologia 482, 1–13. Morris, R.L., Berthold, R., Cabrol, N., 2007. Diving at extreme altitude: Dive planning and execution during the 2006 High-Lakes science expedition. Proceedings of the American Academy of Underwater Sciences, 2007 Diving for Science Symposium, Miami, FL. Muyzer, G., deWaal, E.C., Uitterlinden, A.G., 1993. Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified genes coding for 16S rRNA. Appl. Environ. Microbiol. 59, 695–700. Newman, P.A., Lait, L.R., Schoeberl, M.R., 1988. The morphology and meteorology of south­ ern hemisphere spring total ozone mini-holes. Geophys. Res. Lett. 15, 923–926. Orphan, V.J., et al., 2001. Comparative analysis of methane-oxidizing archaea and sufate­ reducing bacteria in anoxic marine sediments. Appl. Environ. Microbiol. 67, 1922–1934. Perez, R.C., Matin, A., 1980. Growth of Thiobacillus novellas on mixed substrates (mixo­ trophic growth). J. Bacteriol. 142, 633–638. Phillips, R.J., Zuber, M.T., Solomon, S.C., Golombek, M.P., Jakosky, B.M., Banerdt, W.B., et al., 2001. Ancient geodynamics and global-scale hydrology on Mars. Sciencexpress March 15, 4. Pitari, G., Visconti, G., Verdecchia, M., 1992. Global ozone depletion and the Antarctic ozone hole. J. Geophys. Res. 97, 8075–8082. Pringle, C.M., 1991. Geothermally modified waters surface at La Selva biological station, Costa Rica: Volcanic processes introduce chemical discontinuities into Lowland tropical streams. Biotropica 23, 523–529. Rice, M.S., Wang, J.F., Cloutis, E.A., 2008. High-silica rocks and soils at Gusev crater, Mars: Distribution, spectra, and implications for past hydrothermal activity. AGU, Fall Meeting, abstract No. P53A–1437. Rudolft, W.E., 1955. Licancabur: Mountain of the Atacamenos. Geogr. Rev. 45 (2), 157–171. Ruff, S.W., Christensen, P.R., Glotch, T.D., Blaney, D.L., Moersch, J.E., Wyatt, M.B., 2008. The mineralogy of gusev crater and meridiani planum derived from the miniature thermal emission spectrometers on spirit and opportunity rovers. In: Bell, III, J.F. (Ed.), The Martian Surface: Composition, Mineralogy, and Physical Properties. Cambridge University Press, New York, pp. 315–338. Saether, O.A., Willassen, E., 1987. Four new species of Diamesa Meige, 1835 (Diptera, Chironimidae) from the glaciers of Nepal. Entomol. Scand. 29, 189–203. Schmidt, M.E., Ruff, S.W., McCoy, T.J., Farrand, W.H., Johnson, J.R., Gellert, R., et al., 2008. Hydrothermal origin of halogens at Home Plate, Gusev Crater. J. Geophys. Res. 113, E06S12, doi:10.1029/2007JE003027. Schoeberl, M.R., Stolarski, R.S., Krueger, A.J., 1989. The 1988 Antarctic ozone depletion: comparison with previous year depletions. Geophys. Res. Lett. 16, 377–380. Schulze-Makuch, D., Fairén, A.G., Davilla, A.F., 2008. The case for life on Mars. Int. J. Astrobiol. 7 (2), 117–141. Scully, N.M., Lean, D.R.S., 1994. The attenuation of ultraviolet radiation in temperate lakes. Arch. Hydrobiol. Beih. Ergebn. Limnol. 44, 135–144. Servant, M., Fontes, J.C., 1978. Les lacs quaternaires des hauts plateaux des Andes boliviennes. Premières interprétations paléoclimatiques. Cah. ORSTOM, Série Géologie 10 (1), 9–23. Smith, P.H., Tamppari, L., Arvidson, R.E., Bass, D., Blaney, D., Boynton, W., et al., 2008. Introduction to special section on the Phoenix mission: landing site characterization experiments, mission overviews, and expected science. J. Geophys. Res. 113, E00A18, doi:10.1029/2008JE003083.

Dynamics of declining lake habitat in changing climate

369

Solomon, S., 1990. Progress towards a quantitative understanding of Antarctic ozone depletion. Nature 347, 347–354. Squyres, S.W., Aharonson, O., Arvidson, R.E., Bell, J.F. III, Christensen, P.R., Clark, B.C., et al., 2006. Bedrock formation at Meridiani planum. Nature 433. Squyres, S.W., Aharonson, O., Clark, B.C., Cohen, B.A., Crumpler, L., de Souza, P.A., et al., 2007. Pyroclastic activity at home plate in Gusev Crater, Mars. Science 316 (5825), 738–742. Squyres, S.W., Arvidson, R.E., Ruff, S., Gellert, R., Morris, R.V., Ming, D.W., et al., 2008. Detection of silica-rich deposits on Mars. Science 320 (5879), 1063–1067. Squyres, S.S., Grotzinger, J.P., Arvidson, R.E., Bell, J.F., III, Christensen, P.R., Clark, B.C., et al., 2004. In situ evidence for an ancient aqueous environment on Mars. Science 306, 1709–1714. Squyres, S.W., Knoll, A.H.2005. Sedimentary Geology at Meridiani Planum, Mars. Special Issue of the Earth and Planetary Science Letters, 240, 189 p. Stanhill, G., Cohen, S., 1997. Recent changes in Solar irradiance in Antarctica. J. Clim. 10(8), 2078–2086. Stenke, A., Grewe, V., 2004. Impact of dynamically induced ozone mini-hole events on PSC formation and chemical ozone destruction. Adv. Space Res. 33/7, 1062–1067. Sylvestre, F., Servant, M., Servant-Vildary, S., Causse, C., Fournier, C., 1999. Lake-level chronology on the southern Bolivian Altiplano (18 degrees -23 degrees S) during lateglacial time and the early Holocene. Quat. Res. 51, 54–66. Takai, K., Horikoshi, K., 1997. Rapid detection and quantification of members of the Archaeal community by quantitative PCR using Fluorogenic Probes. Appl. Environ. Microbiol. 66 (11), 5066–5072. Tokano, T. (Ed.), 2005. Water on Mars and Life, Adv. Astrobiol. Biogeophys. Springer, Berlin Heidelberg. Varekamp, J.C., Pasternack, G.B., Rowe, G.L., 2000. Volcanic lake systematics II. Chemical constraints. J. Volcanol. Geotherm. Res. 97, 161–180. Vincent, W.F., Pienitz, R.P., 1997. Sensitivity of high-latitude freshwater ecosystems to global change: temperature and solar ultraviolet radiation. Geosci. Can. 23, 231–236. Vinebrook, R.R., Leavitt, P.R., 1996. Effects of ultraviolet radiation on periphyton in an alpine lake. Limnol. Oceanogr. 41 (5), 1035–1040. Wang, A., Bell, J.F., Li, R., Johnson, J.R., Farrand, W.H., Cloutis, E.A., et al., 2008. Light-toned salty soils and coexisting Si-rich species discovered by the Mars exploration Rover spirit in Columbia Hills. J. Geophys. Res. 113, E12S40, doi:10.1029/2008JE003126. Williams, W.D., et al., 1995. Invertebrates in salt lakes of the Bolivian Altiplano. Int. J. Salt Lake Res. 4, 65–77. Williamson, G., Grad, H., De Lange, J., Gilroy, S., 2002. Temperature-dependent ultraviolet responses in zooplankton: Implications of climate change.. Limnol. Oceanogr. 47(6), 1844–1848. Winfrey, M.R., Zeikus, J.G., 1977. Effect of sulfate on carbon and electron flow during microbial methanogenesis in freshwater sediments. Appl. Environ. Microbiol. 33, 275–281. Yen, A.S., Morris, R.V., Clark, B.C., Gellert, R., Knudson, A.T., Squyres, S., et al., 2008. Hydrothermal processes at Gusev Crater: An evaluation of Paso Robles class soils. J. Geophys. Res. 113 (E06S10), doi:10.1029/2007JE002978. Zagarese, H.E., Feldman, M., Williamson, C.E., 1997. UVB-induced damage and photoreacti­ vation in three species of Boeckella (Copepoda, Calanoida). J. Plankton Res. 19, 357–367.

idx1

Author Index

A Abe, Y., 7

Abramov, O., 91, 95–96, 97f, 101–102

Aceituno, P., 8–10

Acs, E., 352

Acuna, M.H., 32

Aharonson, O., 6–7, 34–35, 38, 333f

Ahrens, T.J., 95

Alfaro, P., 209–210

Allemand, P., 201, 215

Allen, C.C., 98–99

Allen, J.R.L., 209–210

Allen, M., 348

Allingham, J.W., 277

Alsdorf, D., 12

Alvarez, W., 209–210

Anderson, D.L., 47–48

Anderson, R.B., 347–348

Anderson, R.C., 111, 165, 347–348

Andrews-Hanna, J., 143, 149–150

Andrews-Hanna, J.C., 35, 47, 49, 184–185,

347–348

Ansan, V., 7–8, 210–211, 214, 228t

Aravena, R., 8–10

Arvidson, R.E., 5–6, 50, 139, 142, 147,

347–348

Atkinson, A., 359

B Baker, L., 32–33

Baker, V., 164

Baker, V.R., 1–2, 5–7, 10, 38, 45, 47, 69–70,

75, 81–83, 85, 94, 102, 173–174,

186–187, 198–200, 223–225, 235–238,

249, 288–289, 326, 328, 348

Bak, F., 361–362

Balme, M., 56–57

Balme, M.R., 275–306

Bandfield, J.L., 205–206

Banerdt, W.B., 326

Barbieri, R., 316

Bargery, A.S., 295

Bar-Nun, A., 32–33 Barlow, N., 102

Barlow, N.G., 5

Barnhart, C.J., 94, 96–97, 185

Basilevskaya, E.A., 2–3 Beaty, D., 91–92, 102

Beaty, D.W., 1, 348

Bell, J.F., 347–348 Benison, K.C., 91

Benn, D., 233–234 Berezovskaya, S., 8

Berman, D.C., 52, 55–56, 275–277, 292–293 Bernard, A., 101

Bertolini, L.M., 120, 123

Beyer, R.A., 128–129, 132, 144, 146,

148–149

Bibring, J., 36–37, 43–44, 48–49, 51

Bibring, J.P., 4–6, 11–12, 101, 135–136, 139,

146, 188, 210–211, 325, 348

Bierhaus, E.B., 180–181 Bishop, J., 149

Bishop, J.L., 5, 135–136, 139–141, 146–149 Blackwelder, C.R., 1–2, 224, 319

Blair, T.C., 232–233, 337

Blasius, K.R., 43, 111, 116–117, 120,

122–123, 146–147, 187

Bleacher, J.E., 2–3 Bleamaster, L.F. III., 197f, 203–204, 206, 216

Boetius, A.K., 361–362 Boisson, J., 298–299 Boothroyd, J.C., 80–81, 328

Borg, L.E., 32

Bothwell, M.L., 357

Bowen, B.B., 91

Boynton, W.V., 298

Brack, A., 348

Brackenridge, G.R., 73, 275–277 Bradley, R.S., 8–10, 348–351

372

Brakenridge, G.R., 91, 96–97

Brand, H.E.A., 144, 148–149

Bridges, N.T., 37, 56–57

Bronmark, C., 1

Bryant, R.G., 315

Burr, D.M.., 55–56, 275–278, 292–293

Byrne, S., 57–58

C Cabrol, N.A., 1–30, 38, 71, 78–79, 85, 91,

126, 150–151, 163, 199, 223–227, 228t,

243–244, 319, 347–370

Carling, P.A., 173–174 Carlton, C.C., 319–320 Carr, M.H., 2–7, 10, 31–68, 70, 73, 83f, 91,

93–94, 111, 117–118, 122, 126, 150–151,

163–164, 173–174, 186–187, 198–199,

223–225, 235–238, 237f, 239–240,

268–269, 300, 325–326, 348

Casiccia, C., 356–357 Cathles, L.M., 96–97 Catling, D.C., 5, 49, 136–137, 139, 143, 146,

148–151

Chan, M.A., 139–141 Chapman, M.G., 2–3, 48–49, 70, 120,

122–123, 128–132, 136–138, 144–152,

163–195, 275–277, 300

Chevrier, V., 11–12 Chojnacki, M., 139–141, 140f, 141–149 Christensen, P.R., 4–5, 12, 53, 57, 128f,

126–127, 129–132, 142, 146, 152, 164, 166,

181–183, 188, 200–201, 250–251, 318

Chuang, F.C., 235

Chubachi, S., 356–357 Chung, F.C., 199

Chyba, C., 91–92, 100–102 Chyba, C.F., 32–33 Clark, B.C., 7

Clark, C., 71–72, 72f Clifford, S., 93

Clifford, S.M., 1–2, 5–6, 10, 42–43, 45–47,

49, 164, 170–172, 187, 198–199, 237f,

238–243, 249–250, 252, 298–299, 348

Clow, G.D., 38, 326

Coates, D.R., 73

Cockell, C.S., 91–92, 100–101 Cohen, S., 356–357 Coleman, N.M., 136, 149–151, 164

Connerney, J.E.P., 348

Author Index

Connerny, J.E., 32–33 Corripio, J., 354

Costard, F., 57, 100, 300–301

Craddock, R.A., 7, 10, 37–38, 69, 198–199,

213, 224, 235, 348

Crisp, J.A., 37

Croft, S.K., 122, 124–125 Crossey, L.J., 95

Crown, D.A., 197f, 198–201, 203–204,

206–208, 211–212, 216

Currey, D.R., 249, 252–254, 261, 267

D Dalrymple, R.W., 209–210

Daubar, I.J., 95–96

Davatzes, A.K., 183–184

Davis, E.E., 96

Davis, P.A., 117–118

Davis, W.L., 1–2, 91, 98–100

Decaulne, A., 295

De Hon, R., 1–3, 7–8

De Hon, R.A., 69–90, 223, 225, 319

Delmelle, P., 101

Demergasso, C., 347–370, 361

de Silva, S., 349

Des Marais, D.J., 1, 91–92, 101, 348

Deutsch, A., 95–97

Di Achille, G., 7–8, 91, 163–164, 167–169,

223–249, 250

Di Cencio, A., 128, 137–138 Dickson, J.L., 53–54, 56, 69–70 Dohm, J.M., 10, 72, 111, 165, 224–225,

241–242

Dojka, M.A., 352

Döll, P., 12–13 Doms, P.E., 53

Dong, H., 361

Dorador, C., 361

Doran, P.T., 7

Douglas, M.S.V., 357–358 Downing, J.A., 12–13, 13f, 14–16 Drake, N.A., 308, 315

Dreibus, G., 32–33 E East, J.W., 11

Edgett, K., 7–8 Edgett, K.S., 36, 40, 49, 56–57, 126,

128–129, 146, 150–151, 163, 166,

Author Index

186–187, 198, 200–201, 207–208, 216,

223–227, 228t, 241–242, 249–251,

254–255, 292–293, 300–301, 323–324,

330, 336, 343, 348

Edwards, C.S., 132, 134f, 152, 188–189

Edwards, H.G.M., 1

Ehlmann, B.L., 4–5, 91, 223, 225, 318

Emrick, C.M., 80

Erard, S., 120–122, 129–132

Eugster, H.P., 314–315

Evans, D., 233–234

F Fairen, A.G., 7

Fanale, F.P., 116, 122–123

Farmer, C.B., 53

Farmer, D.J., 1, 91–92, 101, 348

Farmer, J., 1, 70

Farmer, J.D., 1, 91–92, 100–101, 186–187,

348

Farrand, W.H., 179

Farrell, A.K., 2–3

Fasset, C., 7–8, 38–40, 223, 4–5, 228t, 319

Fasset, C.I., 10–14, 17, 38–40, 69, 195, 200,

224, 233, 235–238, 242–243, 325

Feldman, W.C., 53, 139, 298

Fergason, R.L., 324f, 328–329

Ferguson, H.H.M., 123, 135–136,

149–150

Fishbaugh, K.E., 45–47

Fisher, R.V., 209–210

Folt, C.L., 348

Fontes, J.C., 348–349

Forget, F., 5–6, 17, 54–56, 69–70

Fornari, M., 348–349

Forsythe, R.D., 1–2, 224, 319

Francis, P., 349

Franklin, B.J., 249–275

Frederick, J.E., 356–357

Fredriksson, K., 95

French, H.M., 295–296

Frey, E.L., 34–35, 117, 326

Frink, C.R., 11–12

Fueten, F., 143–149, 151

Fujita, S.-I., 5–6

G Gaddis, L.R., 143–146 Gaidos, E., 198–199

373

Gallagher, C., 289–290, 300–301 Galtier, C., 356–357 Garvin, J.B., 283–285 Gasse, F., 319

Gat, J.R., 10

Gault, D.E., 73

Geissler, P.E., 120–122, 124–127, 146

Gendrin, A., 5–6, 45, 139, 146–149, 151, 188,

318–319

Gibson, J.A.E., 347–370 Giese, B., 8–10 Gilbert, G.K., 317

Gilbert, J.S., 8, 225

Gillespie, A., 146

Glicken, H., 327–328 Glotch, T.D., 4–5, 137, 141–143, 147–150, 164, 335–336, 340–341 Goldspiel, J.M., 1–2, 91, 198–199 Golombek, M.P., 37, 49–50, 52, 82, 117–118,

123, 178–179, 185, 187

Gorsline, D.S., 80

Grant, J., 324–328 Grant, J.A., 7, 10, 198–199, 228t, 249–275, 323–347 Grant, W.D., 313–314 Greeley, R., 2, 36–37, 44–45, 50, 52, 73, 85,

173–174, 198–201, 203

Gregg, K.P., 2–3 Gregg, T.K.P., 198

Grewe, V., 354

Grieve, R.A.F., 328–329 Grimm, R.E., 7, 10, 47–48, 94, 136–137, 149–150, 163–164, 183–187, 228t, 235–238 Grin, E., 1–2, 19

Grin, E.A., 1–30, 38–40, 71, 78–79, 85, 91,

126, 150–151, 163, 223–227, 228t,

243–244, 319, 347–370

Grizzaffi, P., 73

Grosjean, M., 8–10, 348–349 Grotzinger, J., 18–19 Grotzinger, J.P., 5–6, 38–40, 95, 101, 142,

147, 310–311, 336, 342–343

Guest, A., 95, 98

Guest, J.E., 36, 85, 95, 98, 199–201, 203,

267–268

Gulick, V.C., 5–7, 10, 38, 70, 91,

93, 95, 183–184, 198–200, 224,

235–238

374

H Haberle, R.M., 1–2, 5–6, 31, 43–44,

348

Häder, D.-P., 351, 357

Hagar, W.G., 11–12 Hagerty, J.J., 95

Halliday, A.N., 32

Hamblin, K.W., 181–183, 188–189 Hamelin, N., 146–147 Hamilton, S.K., 12–14 Hancock, P.J.A., 11

Hann, B.J., 359

Hanna, J.C., 10, 35, 47, 49, 143, 149–150,

184–187, 347–348

Hansson, L.-A., 1

Hardie, L.A., 308–309, 312–315 Harmon, J.K., 279

Harrison, K.P., 10, 47–48, 94, 136–138, 149–152, 163–195, 228t, 235–238 Hauber, E., 48–49, 143–149, 223, 225, 228t,

232–233

Hauck, S.A., 32

Haveli, I., 5–6 Hazen, R.M., 101

Head, J., 7–8, 48–49 Head, J.W., 7, 10–14, 17, 31–68, 69–70, 75,

76f, 82, 83f, 85, 94, 96–97, 99–100, 103,

143, 185, 195, 200, 224–225, 228t, 233,

235–240, 242–243, 292–293, 299–300,

325

Head, J.W. III., 200, 223–225, 228t, 239–240,

250, 267, 319

Hecht, M.H., 7

Heisinger, H., 40–41 Hendron, A.J., 183

Herbst, D.B., 1

Herdendorf, C.E., 12

Herkenhoff, K.E., 57–58 Herman, J.R., 356–357 Hills, L.S., 73

Hock, A.N., 8–10, 350–351, 354, 357

Hoffman, N., 1–2, 285–286 Hollibaugh, J.T., 1

Holmer, M., 361–362 Holt, J.W., 17, 198

Horikoshi, K., 352

Hovius, N., 94

Howard, A., 7–8

Author Index

Howard, A.D., 7, 10–11, 38, 40, 42–43, 69, 91,

124–125, 164, 172–173, 173f, 185–186,

195–223, 224, 228t, 232–233, 235, 325,

327–328, 330, 336–338, 342, 348

Hynek, B.M., 7, 10, 38, 69, 139–141, 140f,

139–141, 143–149, 223–249, 292–293

I

Idso, S.B., 340–341 Irwin, R.P., 5–6, 35, 7–8, 10, 19, 38, 40–41,

69–71, 93, 100–102, 163, 165–166, 199,

213, 223, 225–227, 235, 319

Irwin, R.P. III., 323–347

Ivanov, A.B., 31

Ivanov, B.A., 95–97

Ivanov, M.A., 96–97

Iverson, N.R., 209–210

Iverson, R.M., 295

Izaguirre, I., 359

J Jackson, M.P.A., 201

Jaeger, W.L., 2–3, 275–278, 285–286,

294–298

Jakosky, B., 348

Jakosky, B.M., 1–2, 33–34, 53, 93, 348

James, P.M., 31–68

Jaque, F., 355–356

Jaumann, R., 166

Jernsletten, J.A., 134, 151–152

Jerolmack, D.J., 38–40

Johanson, L., 73

Johnson, C.L., 32

Johnson, F.A., 11

Jones, J.H., 33

Jons, J.-P., 249

Judson., 37–38, 73

Jung, Y., 356–357

K Kalf, J., 12

Kargel, J.S., 2, 49, 69, 77, 201–204, 214

Kasting, J.F., 1–2, 43–44, 223–224, 348

Kazemi, G.A., 8–10

Keszthelyi, L., 2–3, 275–278, 285–286, 288,

293–294

Keszthelyi, L.P., 295

Kieffer, S.W., 95

Author Index

Kilham, P., 11–12 Kim, J.-R., 275–306 Kiss, K.T., 352

Kite, E.S., 201

Kleinhans, M.G., 7–8 Knight, J.E., 96

Knoll, A., 18–19 Knoll, A.H., 142, 342–343, 347–348 Kochel, R.C., 117–118 Kohler, M.A., 326

Komar, P.D., 179, 328

Komatsu, G., 2, 48, 94, 100, 115, 117–118,

120–125, 128–129, 130f, 129–132,

137–138, 147–148, 150–152, 163–164,

228t, 317

Korteniemi, J., 206–207, 211–212, 212f Kraal, E., 7–8, 116f, 185, 231

Kreslavsky, M.A., 49, 185, 292–293, 299

Kring, D.A., 91, 95–97,

97f, 101

Krug, E.C., 11–12 Kuzmin, R.O., 1–2 L Lachenbruch, A.H., 286

Lackner, C.N., 56–57 Lahsen, A., 349

Laity, J.E., 70, 117–118 Lamb, M.P., 215

Lammer, H., 33–34 Lanagan, P.D., 285–286 Langevin, Y., 57–58 Lang, N.P., 2–3 Lanz, J.K., 95, 98–99 Laskar, J., 13–14, 33–34, 40, 53, 69–70, 93,

195

Lasue, J., 93

Laybourn-Parry, J., 1

Leach, J.W.P., 359

Lean, D.R.S., 357–358 Leask, H.J., 47, 49

Leavitt, P.R., 357

Le Deit, L., 141–143, 146, 148–149 Leech, D.M., 348

Lee, D.–C., 32

Lee, P., 57

Legates, D.R., 8–10 Legros, F., 179–183

375

Lehner, B., 12–13 Leonard, G.J., 195, 198–206, 208, 214

Leverington, D.W., 45, 224, 233

Lewis, K.W., 100, 208–209, 347–348 Lipman, P.W., 181

Logan, B.E., 348

Lopes, R.M., 2–3 Lubin, D., 355–357 Lucchitta, B., 2

Lucchitta, B.K., ., 48, 53–55, 163, 165,

179–181, 183–184, 249

Lunine, J.I., 2

Lunine, L.I., 32–33 M Mahaffy, P., 91, 100–101 Malin, C.M., 126, 146

Malin, M., 7–8 Malin, M.C., 6–7, 36, 40, 49, 56–57, 70,

114–115, 117–118, 122–123, 128–129,

146, 150–151, 163, 166, 198–199,

207–208, 223–227, 228t, 233, 241–242,

250–251, 254–255, 292–293, 300–301,

323–324, 330, 336, 343, 348

Mandelbrot, B., 12–13 Manga, M., 47, 186–187, 300

Mangold, N., 7–8, 45, 53–54, 56, 135–136,

139–144, 146–151, 184–185, 188, 198,

201, 215, 228t

Marchant, D.R., 54–55 Marinovic, N., 349

Marion, G., 198–199 Marzo, G.A., 96–97 Masson, P., 47–48, 114

Masursky, H., 163–164, 186–187, 223–224 Matin, A., 361–362 Matsui, T., 32–33 Max, M.D., 2, 163–164 Maxwell, T.A., 37, 233

McAdam, A.C., 347–348 McCarville, P., 95

McCauley, J., 80–81 McCauley, J.F., 45–48, 111–115, 120, 122,

163, 186–187, 198–199

McEwen, A., 118f, 121f McEwen, A.S., 2–3, 31, 45, 49, 57, 122, 126,

128–129, 132, 139–141, 146, 166,

376

179–181, 183–184, 225, 275–278, 292–293, 324–325 McGill, G.E., 35, 73, 228t

McGovern, P.J., 36

McKay, C.P., 1–2, 7, 91, 98–100, 122

McKenzie, D., 47–48 McKinley, J.P., 348

McLennan, S.M., 5–6, 38–40, 50, 95, 101,

142, 147

McMurtry, G.M., 181

McPherson, J.G., 232–233, 337

McSween, H.Y., 52

Meade, R.H., 238

Mege, D., 47–48 Mellon, M.T., 2, 33–34, 53, 100, 200–201,

279

Melosh, H.J., 33, 328–329, 342

Messerli, B., 348–349 Mest, S.C., 199, 206–207, 211–212 Metz, J.M., 150–152, 163

Meybeck, M., 12

Michelutti, N., 79

Milkovich, S.M., 58, 75

Miller, C., 2

Milliken, R., 206

Milliken, R.E., 135–136, 144, 146,

148–149, 328–329, 331–332, 335–337,

340–341

Milliman, J.D., 238

Milton, D.J., 2–3, 45, 186–187, 223–224, 288–289 Mironenko, M.V., 43

Mischna, M.A., 10, 99

Mitchell, C.W., 316

Montanari, A., 328–329 Montecinos, A., 8–10 Montgomery, D.R., 146

Moore, H.J., 80–81 Moore, J., 7–8 Moore, J.G., 233

Moore, J.M., 5–6, 38–43, 49, 77,

91, 93, 164, 172–173, 173f,

185–186, 195–223, 224–225, 228t,

232–233, 327–328, 330, 336–338,

342–343

Morgan, G.A., 7, 94, 96–97, 103

Mormile, M.R., 91

Morris, R.L., 352

Morris, R.V., 19

Author Index

Mosangini, C., 163–164, 167–169, 174–175

Mottl, M.J., 100–101

Mouginis-Mark, P., 94

Mouginis-Mark, P.J., 2–3, 70, 73, 80

Mukerij, A.B., 315–316

Mullen, G., 1–2

Muller, J.-P., 275–306

Müller, L., 181–183

Murchie, S., 43, 120–122, 129–132, 139–141,

141f, 141–144, 146–151, 225, 324–325

Murchie, S.L., 120–122, 124–125, 210–211

Murray, B.C., 57–58, 195

Murray, J.B., 275–306

Musselwhite, D.S., 2

Mustard, J.F., 4f, 4–5, 7, 43, 53, 120–122,

200, 299

Muyzer, G., 352

N Nash, D.J., 319

Naumov, M.V., 96–97

Nedell, S.S., 48, 80–81, 120, 122–125,

146–147

Nester, P.L., 8–10

Neukum, G., 5, 7, 12–14, 31, 34, 44, 52, 166,

225, 301

Neumann, G.A., 34–35

Newman, P.A., 354

Newsom, H., 224, 328–329, 337, 342

Newsom, H.E., ., 1–2

Nicholson, W.L., 100–101

Nimmo, F., 32, 34–35, 47–48, 56–57

Noguchi, Y., 286–288

Norton, D., 96

Nummedal, D., 2, 81–82, 328

O Oehler, D.Z., 98–99, 319–320

O’Hara, W.J., 38

O’Keefe, J.D., 95

Okubo, C.H., 139–141, 146–147, 183–184

Onorato, P.I.K., 96

Oren, A., 1

Ori, G.-G., 7–8, 91, 163–164, 167–169,

174–175, 223, 225, 228t, 233–234,

307–323 Orosei, R., 298–299 Orphan, V.J., 361–362 Osinski, G.R., 96–97

Author Index

Osterloo, M.M., 38–40 Owen, T.C., 32–33 P Page, D.P., 275–306 Pain, C.F., 92f Pandis, S.N., 7

Pani, E.A., 70–71 Parker, T.J., 5–6, 10, 40–43, 48–49, 70–71,

77, 80–81, 170–172, 187, 224–225, 228t,

237f, 238–243, 249–275, 249–250, 252,

254, 255f, 260–264, 267–268, 270–271,

323–328, 334–338

Parnell, J., 101

Partridge, J.B., 6–7 Patton, F.D., 183

Payne, M.C., 186–187 Peale, S., 2

Pelkey, S.M., 135–138 Pepin, R.O., 32–33 Perez, R.C., 361–362 Perron, J.T., 170–172, 239–241 Petersen, M.T., 96–97 Peterson, C., 120, 123–124 Peterson, J.E., 198

Petit-Maire, N., 319

Peulvast, J.-P., 111

Pfennig, N., 361–362 Phillips, R.J., 1–2, 7, 10, 32, 36–38, 47, 49,

51, 57–58, 69, 100, 165, 184–187, 224,

235, 238–239, 244, 326, 348

Picardi, G., 298

Pienitz, R.P., 357–358 Pierazzo, E., 95

Pierce, T.L., 198

Pieri, D.C., 6–7, 38, 198–199, 223–224 Pike, R.J., 180–181 Pinto, L., 2

Piper, J., 117–118 Pitari, G., 355–356 Plaut, J.J., 53–54, 99–100 Plaut, J.L., 57–58 Pondrelli, M., 150–151, 225, 232–233,

323–324, 326–328, 330, 335, 325, 337–338

Pope, K.O., 95

Potter, D.B., 198, 200–201 Poulet, F., 4–5, 36–37, 50, 96–97, 101,

213–214

Powell, R.D., 233–234

377

Pratt, S., 69, 75, 69, 76f

Preblich, B.S., 292–293

Presley, M.A., 279

Price, B.P., 1

Price, K.H., 199–201

Pringle, C.M., 357

Purves, R., 354

Putzig, N.E., 50, 200–201, 203, 279

Q Quantin, C., 31, 48, 132, 133f, 134–136, 137f,

137–138, 150–152, 163, 180, 182f,

181–186

R Rathbun, J.A., 91, 96–97 Reigber, C., 293–294 Reiss, D., 338–339 Renno, N.O., 7

Rhoads, B.L., 296–298 Rice, J.W., 186–187, 275–278, 281–282,

285–286, 300

Rice, M.S., 19, 347–348 Richards, G.W., 318

Richardson, M.I., 99

Ritter, D.F., 37–38 Roach, L.H., 139–146, 152

Roberts, C.R., 316

Robinson, M.S., 2–3, 47

Robutel, P., 69–70 Rodriguez, J.A., 149–150 Rogers, A.D., 4–5, 137, 141–143, 147,

147–150

Rosenblüth, B., 8–10 Rossbacher, L.A., 73, 249

Rossetti, D.D.F., 209–210 Rossi, A.P., 148–149, 316, 319–320 Rossman, R.I. III., 78–79 Rotto, S., 184, 326–327 Rudolft, W.E., 359

Ruff, S.W.J.D., 19, 347–348 Rummel, J.D., 102

Russel, P.S., 10

Rutten, M.G., 124–125 S Saether, O.A., 359

Safaeinelli, A., 298

Sagan, C., 1–2, 7, 100–101, 347–370

378

Saito, Y., 231

Sakimoto, S.E.H., 2–3 Saric, M.B., 95, 98–99 Sarkar, S., 209–210 Saunders, R.S., 77, 111, 117, 163–164, 187

Saunders, S.R., 326–328 Schmidt, M.E., 96, 98–99, 347–348 Schmincke, H.-U., 209–210 Schneid, B.D., 37, 44–45, 52

Schoeberl, M.R., 355–356 Schon, S.C., 57

Schonfeld, E., 2

Schorghofer, N., 99, 102

Schreiber, B.C., 311–312 Schubert, G., 98

Schultz, P.H., 73, 228t, 326–328 Schultz, R.A., 117–118, 123, 126, 146–147,

150–151, 184, 187

Schulze-Makuch, D., 70, 348

Schwenzer, S.P., 94

Scott, D.H., 2–3, 31, 36, 44, 71, 78–83, 85,

117–118, 126, 237f, 249, 275–277, 323,

327–328, 336

Scully, N.M., 357–358 Segura, T.L., 35–36, 43–44, 94, 100, 224

Seiler, K.-P., 10

Seinfeld, J.H., 7

Self, S., 294

Servant, M., 348–349 Seu, R., 298

Shaller, P.J., 117–118, 181

Sharp, R.P., 47–48, 114–115, 128f, 117–118,

122, 151, 163–164, 198–199, 223–224

Shean, D.E., 54–55 Sherman, C., 181

Shreve, R., 181

Simonds, C.H., 95

Skilling, I.P., 128

Sleep, N.H., 35

Smellie, J.L., 123, 129–132, 147–148 Smith, D.E., 164–166, 195, 230–231 Smith, P.H., 7, 348

Smol, J.P., 357

Smrekar, S.E., 95, 98

Smythe, W.D., 2

Soderblom, L.A., 122, 200

Sodhi, D.S., 296

Solomon, S., 355–356 Solomon, S.C., 10, 32, 34–35

Author Index

Spencer, J.R., 116, 122–123 Squyres, S.S., 347–348 Squyres, S.W., 1–2, 5–6, 18–19, 35–37,

53–54, 73, 85, 91, 96–99, 142, 147,

198–199, 223–224, 347–348

Stanhill, G., 356–357 Stanley, D.G., 238

Stenke, A., 354

Stepinski, A.P., 198–199 Stepinski, T.F., 38, 198–199 Stewart-Mukhopadhyay, S.T., 94

Stewart, S.T., 56–57, 94

Stillman, D.E., 7

Stivaletta, N., 1, 313–314, 316

Stöffler, D., 34

Storkholm, P., 361–362 Strom, R.G., 36, 77, 200–204, 214

Sugimoto, A., 10

Swezey, C.S., 315

Sylvestre, F., 348–349 Syvitski, J.P.M., 231

T Takai, K., 352

Tanaka, K., 34–35 Tanaka, K.L., 31, 36, 44–45, 47–49, 70, 80,

85, 94, 96–97, 120, 123, 126, 128–132,

144–148, 150–151, 172–173, 179,

184–188, 195, 198–206, 208, 214, 237f,

277, 285–286, 292–293, 301, 323,

326–327, 336

Tera, F., 34

Thorn, C.E., 296–298 Thorsos, I.E., 95

Tokano, T., 348

Toon, O.B., 1–2, 58

Tooth, S., 315–316 Tornabene, L.L., 94, 301

Tosca, N.J., 5–6, 102

Touma, J., 69–70 Treiman, A.D., 139–141 Treiman, A.H., 183–184 U Underwood, Jr. J.R., 268–269 V

Valero-Garcés, B., 8–10 Van Bemmelen, R.W., 124–125

Author Index

Van der Wateren, F.M., 209–210

Vaniman, D.T., 139

Varekamp, J.C., 357

Varnes, E.S., 91–92, 100–101

Versh, E., 96

Vickery, A.M., 33

Vincent, W.F., 357–358

Vinebrook, R.R., 357

Vishnevsky, S., 328–329

Vita-Finzi, C., 318

W Walker, G.P.L., 294

Wallace, D., 7

Wang, A., 347–348

Wänke, H., 32–33

Ward, W.R., 122–123

Warne, G.W., 238

Warren, J., 139

Warren, J.K., 307–308, 311

Warren, R., 8–10

Warwick, F.V., 1

Watters, T.R., 178–179, 292–293

Weitz, C.M., 48, 126–129, 135–138,

141–152, 183–185, 205–206, 225, 231

Werner, S.C., 198

Wharton, Jr. R.A., 101

Whitbeck, N.E., 268–269

Wilhelms, D.E., 35, 41–43, 49, 180–181,

195–223 Willander, B.L., 11–12

Willassen, E., 359

379

Williams, D.A., 195, 197f, 198

Williams, J.-P., 132

Williams, R.M.E., 231–233

Williams, W.D., 359

Williamson, C.E., 1, 316

Williamson, G., 308, 357, 362–363

Willmott, C.J., 8–10

Wilson, J., 69–70

Wilson, L., 2–3

Wilson, S.A., 195–223, 323–347

Winfrey, M.R., 361–362

Winston, D., 336

Wise, D.U., 35, 117

Wiseman, S.M., 50

Witbeck, N.E., 123, 149–150, 175–176, 184

Wray, J.J., 319

Wynn-Williams, D.D., 1

Y Yen, A.S., 96, 98–99, 347–348

Yung, Y., 2

Z Zagarese, H.E., 348, 362–363

Zahnle, K.J., 32–33, 35, 93

Zegers, T.E., 143–144, 147–149

Zeikus, J.G., 361–362

Zhong, S., 35

Zimbelman, J.R., 1–2, 80

Zolotov, M.Y., 43

Zuber, M.T., 2–3, 32, 35, 95, 98

Zürcher, L., 96

idx1

Subject Index

A Ablation 7, 99–100, 99f, 103

Accretion 32–34, 226f, 227f

Acidalia Planitia 80–82, 179, 253f, 270f

Acidic

acid rain 11–12

acid snow 315

acid weathering 205–206

environment 11–12

Aeolian (eolian)

activity 51, 132, 309

deflation 215–217, 314

transport 214, 336

Aeolis Mensae 237f

Airfall 147, 178, 185–186, 323–324

Alcoves 55f, 56–57, 69–70, 223–224, 327

Alkaline

alkalinity 11–12

alkalinization 11–12

environment 11–12

Alluvial

bedding 336

deposits 323, 342

fan distributaries 330

fans 7–8, 36–40, 48, 85, 198, 213, 231–233,

309–310, 312–313, 327–330 Alpheus Colles (AC) 196f, 201–202, 202f, 214–216 Altiplanic

lakes 361

winter 353–354

Amazonian 31, 34–35, 44–45, 48, 52, 58–59

Amenthes Planum 200

Amphitrites Paterae 196f, 200, 206

Analog

high altitude 7, 349

polar 7

Ancestral basins 123, 152–153

Aquatic niches 348

Aquifer

over-pressurized 47

permeability 47, 184–187

recharge 8–10

short-residence water in 11

Arabia

level 250, 252, 253f, 254–255,

262–271

shoreline 237f, 240–241, 243–244, 249

Arabia Terra 80, 82–84, 208–209, 237f, 240,

241f, 241–242, 256f

Aram Chaos 126–127, 139, 143, 164

Ares Vallis 47, 80–81, 237f, 326

Argyre basin 75, 76f, 77, 241–242, 326,

338–339

Astrobiology 100

Atacama desert 307, 314f, 348, 350f,

353–354

Athabasca Vallis 275–278, 281–282, 282f,

281–283, 285–286, 276f

Atmosphere 5–8, 32–36, 43–44, 54–55, 69, 99,

126, 295, 299, 308–309, 348

Aureum Chaos 137, 149–150

Aurorae Chaos 174–176

B Backwasting 188, 291f

Bacterial communities 352, 360f, 361

Baetis Mensa 121f, 125–126

Bajada 114, 213, 327, 326, 332–334, 333f,

336–340, 342

Bamberg crater 264–265

Barchans 308f, 310–311, 319f

Basal polar melting 10

Basin

382

Basin (Continued)

closed 13, 16, 69–70, 73–74, 79f,

79–80, 85, 123, 151, 237f, 240,

282–285, 307, 309–310, 326, 342

impact 32, 34–37, 43–44, 77, 84–85, 165,

326–328, 339f

terminal 7–8, 11–12, 69–73, 73t, 75t, 77,

84f, 326–327

Bathymetry 238, 359

Benches

“bench 4” 261–262

group A 167, 172–173, 185, 187–188

group B 172, 178–179, 183, 185–186, 188

Biochemical deposition 308

Biodiversity 349

Biology 100

Bond crater 326, 338–339 Brines 139, 151, 314–315

concentration 313–315,

C Calderae 2–3 Candor Chasma 80–81, 114–115, 115f, 116f, 121f, 120–122, 126, 127f, 126–129, 130f, 131f, 138f, 138–141, 141f, 143–144, 144f, 146–152, 174–175, 182f Canyon 40, 42f, 44, 47–49, 51, 58–59, 71–72,

74–75, 80–82, 111–118, 119f, 120, 122,

135–136, 135f, 137–138, 149, 163–170,

169f, 170–172, 175, 177–179, 181–183,

182f, 184–189, 195–198, 206, 230–231,

260

Capillarity zone 311

Capri Chasma 45–47, 122, 136–138, 164–166, 168f, 167–170, 169f, 170f, 172f, 177–179, 179f Carbonaceous chondrites 32–33

Carbonate 4–5, 11–12, 33, 43–44, 122,

148–149, 308–309, 311–312, 315, 318

nodular carbonates (caliche) 309

Cataracts 288–289, 290f, 293, 300

dry 288–289

Catchment

groundwater 10

topographic 10

Central Andes 8–10, 348–351 Ceraunius Tholus 45, 56, 233

Cerberus fossae 275–278, 281–282, 294, 300

Subject Index

Ceti Mensa 120–122, 130f, 138f, 137–138, 144, 144f, 145f, 145f, 151–152 Channels

avulsion 226f, 226f, 228t, 230–231, 288, 300

braided 175f, 235, 259–260

lenticular 214

outflow 2, 34, 44–49, 51–52, 55–56, 70, 73,

80–82, 94, 111, 122–123, 129–132, 136–

137, 149–150, 163–166, 183–184, 186,

188, 199, 237f, 277

Chaotic terrain 40, 70, 73, 81–82,

111–114, 122, 126, 163–195

Chlorine 38–40, 43

Chryse Planitia 73t, 75t, 80–82, 188–189, 237f

Clathrates 2

Clay 5, 10–12, 92f, 98f, 101, 122, 135–136, 225,

307–308, 310–311, 312f, 319, 331–332 Climate change 1, 8–10, 13–14, 16, 40, 51–52,

69–70, 94, 100, 102, 117–118, 143, 238,

319f, 319, 348–351

cold 10, 19f, 198–199, 271

dry 363

stability 16

variability 316

warm 91, 100, 103, 149

wet 5, 100, 117–118

Closed basins see Basin, closed

CO2 2, 32–33, 43–44, 56–58, 95

Columbia Hills 19, 19f, 36–37, 50, 96, 98–99,

347–348

Columbus crater 309f, 319

Contact

Contact 1 170–172, 240–241, 249

Contact 2 187, 240–241, 249, 267

nested plains unit contacts 249

Coprates Chasma 48, 111, 117–118, 117f, 123,

136–138, 149–152, 164, 168f, 167–172,

172f, 172–173, 173f, 175, 176f, 177–178,

177f, 184, 186–187

Crater

concentratic crater fill 55, 58–59, 241–242

ejecta 5, 73

impact 1–2, 5–6, 6f, 10–14, 31, 35–37, 73,

91, 92f, 93f, 94–97, 98f, 100–102, 148,

163–164, 196f, 206–207, 209–210, 213–

214, 224–225, 233, 276f, 278–279, 285f,

292–294, 296–298, 301, 326, 332f

Subject Index

lakes 12, 96–97, 98f, 99f, 122–123, 150–151, 359

Cratered Uplands 36–37, 44–45, 327–328

Crust

crustal dichotomy 237f, 240–242 crustal thickness 32, 34–35 Cryosphere 47–48, 51–52, 55–59, 93–94, 150–151, 300

Cryoturbation 293

Cryo-volcanic cones 285–286

Cydonia mensae 256f, 259, 261, 268–270

D Dao Vallis 196f Debris aprons 53–55, 58–59, 118f, 179–183, 182f, 181–183, 198, 256, 261

fans 56–57

lobate debris aprons 53–55, 58–59, 198

Deflation 139, 215, 293–294, 300–301, 309,

314, 332f

Degassing 32–33, 149, 357, 359, 361–362

Deltas

deltaic deposits 92f, 92f, 223–225, 226f,

227f, 228t, 231, 237f, 238–240,

243–244, 250

deltaic facies 317

deltaic record 8

Eberswalde 38–40, 150–151, 225, 226f,

343

Gilbert 8

glacier-fed 233–234

ice-contact 233–234

kame 233–234

plain 226f, 227f, 230–231

stepped 8, 9f

Detritus, Siliciclastic 308

Deuteronilus

level 250, 253f, 254, 261–263, 263f, 264f, 264, 265f, 264–267, 266f, 267–269, 269f, 269–270, 270f, 270–272 mensae 237f, 252, 253f, 254

shoreline 49, 240–241, 249

Diagenetic alterations 147

Diapirs

gypsum 144

salt 144, 201, 215

383

Dichotomy 18, 34–36, 41–42, 69–70, 165,

177–178, 181–183, 227f, 237f, 240–242,

255, 275–277

Differentiation 32–33, 231–232

Dikes 46f, 47, 147–148

Discharge 7–8, 11, 38–40, 47, 70–73, 77, 79,

136, 143, 163–166, 176f, 181–187,

205–206, 231, 238, 300, 315–317,

326–328, 337–338, 339f, 339–341,

341f, 343

Dissolved organic carbon (DOC) 357–358

Distributary patterns 225

Distribution

bimodal, hemispheric 17, 34–35 lake size-frequency 12–13, 16–17 non uniform distribution of secondary craters 31

Pareto 14, 16

Dosimeters (UV) 351

Drainage

basins 41–42, 73–74

divide 72f, 73

integration of 41–42

Dunes 50, 129f, 128, 141–142, 172–173, 173f, 201, 209–210, 215, 310–311, 319f Dust

storms 35–36, 123–124, 195

wind-blown 99

E Eberswalde crater 150–151, 226f, 343

Echus Chasma 45, 56, 70, 135–136, 149

Elaver Vallis 123

Elysium

Basin 2–3, 16, 275–277, 277f, 275–278, 285–286, 288, 292–293, 300

paleolake 300

Planitia 3f, 78, 82–84, 263–264

Enthalpy 91, 98, 102–103

Environment 1, 3–8, 10–11, 16, 18–19, 32,

35–36, 91, 101, 114–115, 117–118,

126–127, 129–132, 134, 147, 152, 179,

181, 195–199, 205–206, 208–211, 211f,

211–212, 215–217, 223, 225, 231–234,

271, 296–298, 300–301, 307–308, 311,

319, 336–338, 348–349, 351, 356–357

arid 311

384

Eos Chasma 80–82, 115, 120–122, 132, 134f,

166, 173–175, 178–179, 181–184, 189

Equator 2, 5–6, 8, 10, 16, 78, 80, 94, 111–114,

284f, 286–288, 292–293

Equipotential surfaces 237f, 238–244, 250,

252–253, 282–283, 284f

Erosion rate 7, 37, 40, 49–50, 57–59

Evaporation, rate of 315, 340–341

Evaporative pumping 311–312

Evaporites

evaporitic environment 5–6

siliciclastic-rich 142

zonation of 315

Exploration 1–30, 50, 223, 243–244

F Finger rafting 278–279

Fluvial activity 55, 58–59, 93, 135–136, 149,

238–239, 326–327

Focas crater 259–260

Foreset 214, 233

Fretted terrain 84f, 237f, 254, 259–262, 262f,

268

G Ganges

Chasma 111, 122, 124f, 136, 149–150,

184, 187

Mensa (GM) 120, 124f

Geochemistry 11, 142

Geology 31, 47–48, 198, 215, 233–234, 238,

251, 295–298, 312

geological record 14, 15f, 16–17, 348

Geopotential 241–242

Geothermal

gradient 95, 98–99, 102–103

heat (sources) 100, 102–103, 217

Glacial melt 8–10, 70, 233–234 processes 17, 299–300 Glaciation 17, 52–54, 69–70

Glaciers 2, 8–10, 53–55, 69–70, 115, 198,

233–234, 348–349

Gorgonum Chaos 164, 172–173, 173f,

172–173, 185–186

Grabens 45–47, 46f, 111, 117–118, 122–123,

148, 150–153, 178, 184

Subject Index

Greenhouse

conditions 94

gas 35, 43–44

Ground ice 47, 53–55, 94, 111–115, 117–118,

120, 134, 151–153, 163–164, 201–202,

291f

Grounding-line fans 233–234 Groundwater

circulation 102

eruptions 31–32, 47, 51, 57

flow paths 8, 10–11

seepage 38

storage 36, 45–47, 69, 73, 75

Gullies 48, 53, 56–57, 56f, 56–59, 100,

111–114, 116, 117f, 117–118, 122–123,

134, 146–147, 152, 177–178, 289–290,

291f, 295–296, 300–301, 327, 339

Gusev crater 19, 50, 70–71, 225–227,

347–348

Gypsum crusts (gypscrates) 309

H Habitability

habitable planet 19

habitats 1, 100–101, 347–370

Hadriaca Patera 77

Hale crater 326

Halites 148–149, 310f, 312, 315

Harmakhis Vallis 196f

Heat

convective heat loss 96

convective heat transport 96

flow 5–6, 32, 98–100, 102–103

Hebes Chasma 111, 121f, 123, 129–132,

150–151, 182f

Hellas

basin 35, 37–38, 71–72, 77, 198, 213–214,

235–238, 241–242, 242f, 242–243

chaos 196f, 201–202

planitia 73t, 75t, 77, 195, 198–201, 202f,

203f, 203–205, 205f, 208–216, 241f

Hellespontus Montes 198

Hematite, crystalline 120–122, 141–142, 250

Hesperian ridged plains 44–45, 77, 80–81

Hesperia Planum 31, 44–45, 197f, 200–201

Highlands 5–6, 18, 32, 36, 40, 57, 69–72, 73t,

75t, 77–78, 82–83, 85, 135–136, 139,

Subject Index

195, 197f, 198–201, 206–208, 213–216, 254–256, 257f, 259–260, 267, 269–271, 275–277 Highstand 227f, 239, 239f, 240, 249, 310–311

Holden crater 38–40, 323–347

Home Plate 19, 19f, 347–348

Honeycomb terrain (ht) 197f, 201–202, 202f,

203f, 205f, 214

Huasco 361

Hydraotes Chaos 164, 167–169, 174–175,

176f Hydrated

layers 53

minerals 4–5, 98f, 122, 139, 209–210

phyllosilicates 139, 210–211

polyhydrated sulfates 139–141, 140f, 141f,

142, 148

silica 36–37, 135–136, 205–206 Hydrodynamic escape (atmosphere) 32–33 Hydrographic system immaturity 6–7

maturity 6

Hydrological provinces

Aeolis 78–79, 82–83

Amazonis 78

Argyre 77

Australe-Parva 75t, 75, 79

Central Highland 78, 82

Chryse 79–80

Hellas 77

Isidis 78

northern lowland 73t, 75t, 78, 82

Southern Plateau 75

Hydrologic cycle 70–71, 235, 325

Hydrology 13, 228t, 230, 243–244

Hydrosphere, Global 223–249

Hydrothermal

activity 5, 8–10, 36, 98–99, 117–118, 151–153, 347–348

alteration 95, 120–122, 342

circulation 38

deposits 91–92, 98–99

fluids 36–37, 96–97, 124–125, 147

processes 5, 96–97, 198–199

springs 19

systems 73, 95–96, 100–102

385

I

Iani Chaos 147–148 Ice

table 53

cap 70, 75–77, 70

-contact delta 233–234

-cored mounds 285–286

-covered lakes 99, 136, 197f, 198, 201,

207f

dam 91, 186–187

floes 266f, 277, 295

movement of 58–59

stability 58–59

wedges 271, 296–298

Impact crater

impact breccia 36–37

impact cratering 5–6, 10–11, 13–14, 95,

164, 209–210, 213–214

impact melt 43, 93f, 94–97, 99–100, 99f,

101–103

Impoundment 70–71, 73, 80–81, 85, 324–325,

338–339, 343

Infiltration 7, 10–11, 38, 41–42, 73, 294,

315–316, 339–341

Interior layered deposits (Valles Marineris)

47–48, 111–114, 120, 163, 184, 188,

197f

Intertropical belt 8, 18

Iron oxides 4–5, 139

Ismenius

Ismenius level 253f, 261, 265–267, 266f,

267–268, 271–272

Ius Chasma 117–118, 119f, 125f, 128, 152,

152–153

J Jökulhaups 129–132 Juventae Chasma 4f, 42f, 45–47, 111, 112f,

120, 129–132, 135–136, 136f, 136–137,

139–141, 148–150, 164

K Kieserite 42f, 50, 139–141, 140f, 141f,

142–144, 148

Knobs 131f, 139–141, 172–173, 202f, 263,

263f, 264–265, 266f, 271–272

386

L Ladon Vallis 40–41, 80–81 Lakes

abundance 14

crater 12, 96–97, 98f, 122–123, 150–151, 217

declining 347–370

deposits 8, 11, 19, 93, 100–101, 103, 146,

224, 278

dry 307–323

duration 224

ecosystem 313–314, 348–349, 351–352,

356–357 ephemeral 314–315 formation 3, 12–14, 16, 17f, 44, 184–185, 188

habitat 347–370

heated 91–111

high altitude 348–349

interdune lakes, playas 18–19, 43, 310–311

intra-canyon 178–179

inventory 13–14, 32–33, 35–36, 42–43,

58–59

open lake systems 12, 16

paleolakes 1–3, 6, 13f, 14, 17f, 71, 80–83,

85, 163, 183, 211–212, 224–225, 238,

241f, 241–243, 252–256, 270–271

perennial 317

record (residual) 12

saline 308, 309f, 308–310, 310f, 311,

315–316

spillover 75, 82, 136, 288

strandline 91, 254–255

volume 8, 45–47, 94, 122, 134, 149–152,

163–166, 244, 316–317 Landslides 48, 58–59, 113f, 111–114, 120,

132, 133f, 134, 151–153, 176, 179–183,

182f, 184–186, 188, 201–202, 214

Late Heavy Bombardment 10, 34

Lava

flow 3f, 19, 37, 45, 80, 129–132, 135–136,

205–206, 214, 233, 268–269, 271, 277–

278, 286–288, 294–296, 299

inflating 294, 300

lakes 2–3, 45

-rise pits 294

low viscosity 254–255, 270–271

tubes 233, 294–295

Subject Index

Lethe Vallis 282–283, 285–286, 287f, 288–289, 289f, 290f, 296, 297f Licancabur

lake 349–351, 357, 360f

volcano 349, 350f, 351

Life adaptation 1, 19, 359–360 survival potential 313–314, 348, 359–360, 362–363

Limnology 3–4, 14

Lineated valley fill 55, 58–59, 69–70

Lobate debris aprons 53–55, 58–59, 198

Lunae Planum 44, 80–81, 117–118

Lyot crater 260–262, 264–265

M Ma’adim Vallis 19, 40–41, 70–71, 78–79, 79f, 84f, 165–166 Magma 19, 84f, 188, 347–348 Magnetic anomalies 32

field 32–33, 35–36, 58–59, 348

Maja Vallis 45–47, 70

Malea Planum 197f, 200–201

Mamers Valles 84f, 254–256, 260–265, 267,

268f, 269f Mangala Vallis 46f, 80, 81f Mantle 32–33, 35–36, 58–59, 98, 129–132, 256, 259

Margaritifer

basin 326

Terra 323, 324f, 326–328, 341

Mars 1–30, 31–68, 69–90, 91–111, 111–114,

114f, 114–115, 117–118, 119f, 121f,

122–123, 126, 132, 139, 146–148, 150–

151, 163, 167–169, 174–175, 179, 185,

195–199, 201, 202f, 207–208, 213–215,

217, 223–249, 249–251, 252f, 252–253,

269f, 271–272, 275–277, 276f, 278, 281f,

282–283, 285f, 289f, 297f, 301, 308,

310–311, 314, 316–318, 323–347, 347–

351, 362–363

Mass flow deposits 309–310 Mass wasting (processes) 56–57, 117–118,

124–125, 141–142, 148, 167–169, 178,

181, 185–186, 198–199, 214, 224

Megabreccia 323, 328, 341–342

Subject Index

Melas Chasma 56, 120, 117, 135–136, 137f,

137–138, 150–152, 163–166, 180–181,

184, 186

Melting 10, 40, 31–32, 47, 50, 56–57, 83–84,

94, 115, 164, 186–187, 293, 300–301, 359

Meridiani Planum 10, 18–19, 95, 126–127,

139, 141–143, 142f, 151, 250, 347–348

Microorganisms 313–314, 348, 357, 358f, 360

Millochau crater 212f Mineral

deposits 254

igneous 36–37

mineralogical sequence 4–5

Moats 121f, 120, 129–132, 144–146, 148,

285–286, 286f

Morava Valles 326

Morphology 1–2, 6–8, 12, 18, 36–37, 56–57,

117–118, 118f, 129–132, 144, 146–147,

152, 164, 166–169, 172f, 174–175,

177–178, 181–183, 185, 187–188,

195–198, 200, 214–216, 224–225, 226f,

227f, 230–234, 250, 254–256, 261,

270–271, 275–283, 285–286, 288–290,

293, 295–296, 299–300, 316, 328, 342

Morphometry 6–7, 226f, 227f, 228t, 230–231, 239f Mounds lobate 263–264, 271–272 pitted 278, 285–286, 286f, 291f, 295–299 Mud

flats 309f, 309–313, 315–317

stone 318

Multilobate deposition 230–231 N Nepenthes Mensae 239f, 242–243 Niesten crater 215

Niger Vallis 196f Nili Fossae 4–5, 6f, 38–40, 43

Nirgal Vallis 70, 326, 338–341, 341f, 343

Noachian 3–5, 8, 10–11, 13–14, 31–36,

44–45, 47–52, 55, 58–59, 75–82, 84–85,

93–94, 122–123, 126, 135–136, 143,

146, 188, 195, 196f, 197f, 195–200, 206,

207f, 208, 212–213, 235–239, 241–242,

242f, 242–244, 323, 324f, 325–328,

336–342, 348

387

Noctis Labyrinthus 111

Northern Plains (lowlands) 18, 35, 40, 45,

71–72, 73t, 75t, 78, 82–83 –83, 84f, 237f,

238, 240–241, 241f, 241–243, 254,

270–271

O Obliquity cycles

high obliquity cycles 7–8, 14

low obliquity cycles 53

Ocean oceanic basin 249–275 paleoocean 224–225 S2 ocean boundaries 241–242 Olivine 4–5, 36–37, 43, 50, 132, 134f, 152,

181–183, 188–189

Olympus Mons 44–45, 72, 79–80, 279

Ophir Cavus 136

Chasma 80–81, 111, 118f, 119f, 121f, 128f,

134

Opportunity 7–8, 18–19, 50, 132, 142, 142f,

148, 151, 186–187, 347–348

Outburst flooding 163–195 Outflow 2, 32–34, 44–49, 51–52, 55–56,

69–70, 73, 80–82, 85, 94, 100, 111,

122–123, 129–132, 136–137, 149–150,

152–153, 163–166, 183–184, 186, 188,

199, 237f, 259–260, 277, 283–285, 288,

292f, 323, 324f, 326–328

Outgassing 33

Ozone depletion 354, 356–357 P Pack-ice 275–279, 296

Periglacial environment 300–301 Permafrost 70, 83–84, 94, 96–97, 122,

295–296, 299–301

Permeability 47, 73, 94, 96, 97f

Phyllosilicates 4–5, 36, 43–44, 50–52, 58–59,

96–97, 98f, 101, 139, 188, 205–206,

210–211, 328–330, 330f, 331–332,

335–337, 340–343

Physiographic control 69–90 Pingos 73, 77, 271–272, 285–286, 295–298,

300–301

Planetary protection 91–92, 100, 103

388

Platy

flow textured plains 271–272

fractured plains 263, 264f, 265f

rafts 295–296

-ridged terrain 275–278

surface morphology 288

textures 296

Playa

deposits 142, 147, 309f

inter-dune 50, 310–311

Polar layered deposits 49–50, 58–59

Polyconvex terrain 279, 281f, 283f, 286, 287f,

288–290, 295–299

Polygonal patterned ground 262–263, 286,

287f Polygonization 296–298 Ponds episodic ponding 163–195 Potentiometric surface 149–150 Power-law function 15f, 16, 17f, 16–17 Precipitation mineral 5, 8, 11–12, 100–101, 205–206, 315

rainfall 5–6, 10, 70

snow 5–7, 350–351

of sulfur 18, 50, 101

Production (rate)

crater production rate 13–14

lake production rate 16

melt 95

runoff production rate 5–6

Progradation 232–233, 317

Pyroclastic flows 36–37

Pyroxene 36–37, 120–122, 129–132, 139–144

R Rahe 233

Rate of decline 12–14, 16

Regolith 70, 213–214, 293–294, 298–301

Resurfacing

volcanic 36–37, 200

Reticulate terrain 197f, 201–202, 202f,

204–205, 205f, 205–206, 216

Reull Vallis 196f, 206

Rilles

fluvial 271–272

lunar 300

volcanic 233

Subject Index

Rootless cones 179, 278, 285–286, 295

Rubbly material 288–289

Runoff 4f, 5–7, 10–12, 14, 16, 38, 39f, 41–42,

51, 69–72, 78, 85, 100, 102–103,

135–136, 149, 163, 184–185, 196f,

198–200, 215–216, 223–224, 231, 235,

235–238, 237f, 235–238, 244, 311–312,

327, 337, 357–358

S Sabkhas

coastal 307

continental 308–310

Salar de Ascotän 361

Salt

perchlorate 348

Sand

deposits 307

flats 309–310 –311, 317

Sapping processes 119f, 199

Sea

frozen 3f

inland 277

Sediment

accumulation 7–8

deposits 11, 203, 210–211, 250

indurated, fine-grained 40

layered sediments, (rhythmically) 41f, 48,

213, 323–324 mounds of 48–49 sedimentary record 307–323 soft sediment deformation 144, 148–149, 201, 209–210 –210 traps 215–217

Semeykin crater 259–260

Shalbatana Vallis 164, 167–169, 226f, 237f

Shock heated basement 102–103

Shorelines wave-eroded 254–255

Silicate hydrated 36–37

Siliciclastic deposits 307–308

Simud Valles 48, 175, 237f

Size-frequency

lake size-frequency distribution 12–13, 13f, 13–14, 15f Solar flux 93, 98–101, 355f Solar irradiance 348–351, 354, 356–357 Solifluction 296–298

Subject Index

Solutes 11

South Polar Cap 69, 75–77

Spirit 19, 19f, 20f, 98–99, 132, 347–348

Springs

deposits 70, 98–99, 309–310 mounds 148–149, 316, 319–320 Stone circles 287f, 289f, 295–296, 300–301 Strandline 91, 138f, 137–138, 225, 254–256, 270–271, 338

Stratigraphy

stratigraphic column 318

stratigraphic relationships 4–5

Streamlined islands 45, 275–277, 285–286,

288–289, 290f, 293, 296

Sublimation 41–42, 48–50, 53–54, 185,

201–202, 271–272, 272, 282–283,

293–295, 299–301

Subsurface flow paths 10–11, see also

Watershed, waterways

Sulfate deposits 5–6, 11–12, 50–51, 308,

318–319 Sulfides 361–362 Suspension 211f, 212–214, 216–217 SW crater (Hellas) 196f, 211–212, 212f, 211–217

Syrtis Major 38, 44–45, 52, 78, 84f, 214

T Taffy-pull terrain 201

Temperature 4–5, 4f, 5–10, 31, 35, 41–44, 53,

57, 93, 96, 102, 126–127, 139, 144, 224,

255–256, 282f, 348, 350–352, 357, 358f,

359–360

Terby crater 195–198, 206–207, 207f, 209f, 210–211, 211f, 211–212, 214

Terraces Lacustrine 183

Terra Sabaea 198, 206–207

Terra Sirenum 79, 237f, 242–243, 319

Tharsis 10, 18, 35–37, 44–45, 47–48, 51–52,

54–55, 58–59, 79–81, 102, 111, 117,

123–124, 136, 143–144, 149–150,

164–165, 235–238

Thawing 10, 70, 350–351, 363

Thermal

contraction 293, 299–301

signature 200–201, 216

389

thermocontraction 293, 299–301 thermoerosion 295–296 Thermokarst 134, 295–298 Thumbprint terrain 49, 73, 249–250, 267–269,

271–272

Tindars 129–132 Tithonius Lacus 111

Topset 214

Total dissolved solids 357

Transient cavity 94–96 Transport 5, 7–8, 34, 41–42, 51, 85, 92f, 96,

126, 135–136, 142, 175–176, 199, 211f,

213–214, 224, 231–234, 234f, 254–255,

270–271, 281–282, 294–295, 309,

312–313, 315–316, 327–329, 334–336,

338, 342

Tributaries 6–7, 11–12, 38, 70, 92f, 111–115,

117–118, 119f, 122, 135–136, 135f,

137–138, 149, 152–153, 169f, 196f, 225,

230–231, 232f, 231–232, 235, 288, 289f,

289–290, 293, 295–296, 309–310,

315–316, 326–328, 330, 332–334, 333f,

336, 338–339

Troughs 71–72, 74–77, 111, 112f, 111–116,

120–124, 126, 129–132, 135–138,

143–144, 146–153, 181, 208, 215, 264,

281f, 334–335

Truncation

truncated distribution 13f, 14–15, 331

Tuffs hyalotuffs 129–132 Turbulent flow 294–295 Tuyas 129–132, 147–148, 151

Tyrrhena Patera 77, 200, 206, 214

U Ultraviolet radiation (UV)

PAR (photosynthetic active radiation) 351,

354

penetration 357–358

short wavelength 355f

UVA 351, 354, 355f, 358f, 360f, 363f

UVB 350–351, 354, 355f, 356–357, 358f,

359, 360f, 361–363, 363f

Uzboi–Ladon–Morava (ULM) outflow channel system 323, 324f, 326–329

390

V Valles Marineris Depression (VMD) 164–166, 183–188 Valleys dendritic valley networks 149 hanging 115, 288, 289f, 293 integration of valley networks 2, 6–8, 11–12, 224–225, 235, 238, 243–244

termini 224, 242f

“veeing” up 260

Vastitas Borealis Formation 49, 82 Volcanism mud volcanoes 179, 271–272 subaqueous volcanism 48–49 volcanic activity 2–3, 5–6, 11, 47, 52, 152–153, 216, 293 volcanic ash 36, 139, 143, 151, 205–206, 215 volcanoes 45, 54–56, 58–59, 70, 73, 79, 84–85, 124–125, 129–132, 147–149, 152–153, 179, 195, 198, 235–238, 271–272 W Water abundance 3–4 acidic 101, 152–153 acquisition and history of 31–68 alkaline 4–5, 11–12 balance (negative) 315, 350–351, 354

Subject Index

capillarity 311 catastrophic release of 173–174 column 11, 312–315, 348, 357, 358f, 359–360 estimates amount of 32–33, 42–43, 47, 163–164, 243, 339

hydrothermal 5

inventory of water (global, surface,

underground) 32–33, 42–43 liquid 5–7, 11–12, 31–32, 38, 43–45, 50, 55–57, 91, 98–99, 126–127, 195–199, 224, 270–271, 300–301, 348 neutral 4–5 storage 69, 73, 75–77 Watershed waterways 10–11, 40–41, Subsurface flow paths Wave ripples 312f, 318 Weathering hydrous 43, 58–59

pervasive aqueous 31

Wind deflation 309 erosion 120, 148, 152–153 X Xanthe Terra 80–81, 240 Z Zooplankton 349, 352, 357, 359