The Tectonics, sedimentation and palaeoceanography of the North Atlantic Region

The Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region

Geological Society Special Publications Series Editor A. J. Fleet

GEOLOGICAL SOCIETY SPECIAL PUBLICATION NO. 90

The Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region EDITED BY

R. A. SCRUTTON Grant Institute, University of Edinburgh, UK

M. S. STOKER British Geological Survey, Edinburgh, UK

G. B. SHIMMIELD Grant Institute, University of Edinburgh, UK and

A. W. T U D H O P E Grant Institute, University of Edinburgh, UK

1995 Published by The Geological Society London

T H E G E O L O G I C A L SOCIETY The Society was founded in 1807 as The Geological Society of London and is the oldest geological society in the world. It received its Royal Charter in 1825 for the purpose of 'investigating the mineral structure of the Earth'. The Society is Britain's national society for geology with a membership of 7500. It has countrywide coverage and approximately 1000 members reside overseas. The Society is responsible for all aspects of the geological sciences including professional matters. The Society has its own publishing house which produces the Society's international journals, books and maps, and which acts as the European distributor for publications of the American Association of Petroleum Geologists, SEPM and the Geological Society of America. Fellowship is open to those holding a recognized honours degree in geology or cognate subject and who have at least two years' relevant postgraduate experience, or who have not less than six years' relevant experience in geology or a cognate subject. A Fellow who has not less than five years' relevant postgraduate experience in the practice of geology may apply for validation and, subject to approval, may be able to use the designatory letters C. Geol (Chartered Geologist). Further information about the Society is available from the Membership Manager, The Geological Society, Burlington House, Piccadilly, London WlV 0JU, UK. The Society is a Registered Charity No. 210161. Published by The Geological Society from: The Geological Society Publishing House Unit 7 Brassmill Enterprise Centre Brassmill Lane Bath BA1 3JN UK (Orders: Tel. 01225 445046 Fax 01225 442836) First published 1995 The publisher makes no representation, express or implied, with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility for any errors or omissions that may be made. 9 The Geological Society 1995. All rights reserved. No reproduction, copy or transmission of this publication may be made without prior written permission. No paragraph of this publication may be reproduced, copied or transmitted save with-the provisions of the Copyright Licensing Agency, 90 Tottenham Court Road, London WlP 9HE. Users registered with the Copyright Clearance Center, 27 Congress Street, Salem, MA 01970, USA: the item-fee code for this publication is 0305-8719/95 $07.00. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN 1-897799-27-6

Typeset by Bath Typesetting Ltd Bath, England Printed in Great Britain by Alden Press, Oxford

Distributors USA AAPG Bookstore PO Box 979 Tulsa Oklahoma 74101-0979 USA (Orders: Tel. (918) 584-2555 Fax (918) 584-0469) Australia Australian Mineral Foundation 63 Conyngham Street Glenside South Australia 5065 Australia (Orders: Tel. (08) 379-0444 Fax (08) 379-4634) India Affiliated East-West Press PVT Ltd G-l/16 Ansari Road New Delhi 110 002 India (Orders: Tel. (11) 327-9113 Fax (11) 326-0538) Japan Kanda Book Trading Co. Tanikawa Building 3-2 Kanda Surngadai Chiyoda-Ku Tokyo 101 Japan (Orders: Tel. (03) 3255-3497 Fax (03) 3255-3495)

Contents Preface

DRISCOLL, N. W., HOGG, J. R., CHRISTIE-BLICK, N. & KARNER, G. D. Extensional tectonics in the Jeanne d'Arc Basin, offshore Newfoundland: Implications for the timing of break-up between Grand Banks and Iberia

vii

1

SINCLAIR,I. K. Sequence stratigraphic response to Aptian-Albian rifting in conjugate margin basins: a comparison of the Jeanne d'Arc Basin, offshore Newfoundland, and the Porcupine Basin, offshore Ireland

29

EBDON, C. C., GRANGER, P. G., JOHNSON, H. & EVANS, A. M. Early Tertiary evolution and sequence stratigraphy of the Faeroe-Shetland Basin: implications for hydrocarbon prospectivity

51

BOILLOT,G., BESLIER,M. O., KRAwczYK, C. M., RAPPIN, D. & RESTON,T. J. The formation of passive margins: constraints from the crustal structure and segmentation of the deep Galicia margin, Spain

71

RESTON, T. J., KRAWCZYK,C. M. & HOFFMANN, H.-J. Detachment tectonics during Atlantic rifting: analysis and interpretation of the S reflection, the west Galicia margin

93

KIORBOE, L. & PETERSEN, S. m. Seismic investigation of the Faeroe basalts and their substratum

111

VANNESTE,K., HENRIET,J.-P., POSEWANG,J. & THEILEN,F. Seismic stratigraphy of the Bill Bailey and Lousy Bank area: implications for subsidence history

125

ANDERSEN, M. S. & BOLDREEL, L. O. Effect of Eocene-Miocene compression structures on bottom-water currents in the Faeroe-Rockall area

141

BOLDREEL, L. O. & ANDERSEN, M. S. The relationship between the distribution of Tertiary sediments, tectonic processes and deep-water circulation around the Faeroe Islands

145

STOKER, M. S. The influence of glacigenic sedimentation on slope-apron development on the continental margin off Northwest Britain

159

WAAGSTEIN, R. & HEILMANN-CLAUSEN,C. Petrography and biostratigraphy of Palaeogene volcaniclastic sediments dredged from the Faeroes shelf

179

JONES, E. J. W., CANOE, S. C. & SPATHOPOULOS,F. Evolution of a major oceanographic pathway: the equatorial Atlantic

199

ANDERSEN, M. S. & BOLDREEL, L. O. Tertiary compression structures in the Faeroe-Rockall area

215

HASLETT, S. K. Plio-Pleistocene radiolarian biostratigraphy and palaeoceanography of the North Atlantic

217

HUNT, J. B., FANNIN, N. G. T., HILL, P. G. & PEACOCK, J. D. The tephrochronology and radiocarbon dating of North Atlantic, late Quaternary sediments: an example from the St Kilda Basin

227

THOMSON, K. & HILLIS, R. R. Tertiary structuration and erosion of the Inner Moray Firth

249

WOLD, C. N. Palaeobathymetric reconstruction on a gridded database: the northern North Atlantic and southern Greenland-Iceland-Norwegian Sea

271

Index

303

Preface The stimuli for papers collected in this volume are the return of the Ocean Drilling Programme to the North Atlantic, and the exploration industry's advance into deep-waters off NW Europe. The 15 papers and two extended abstracts encompass the wide range of topics covered by the 'Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region'. Broad aspects of the plate tectonic evolution of the North Atlantic are presented in the papers by Driscoll et al. and Sinclair, who use sequence stratigraphic techniques to decipher the nature and timing of basin development and seafloor spreading between the conjugate margins of eastern Canada and western Europe. The crustal response to continental break-up and mode of lithospheric extension are discussed by Boillot et al. and Reston et al., based on advanced seismic processing of deep-seismic reflection data from the west European margin. The sedimentary response to North Atlantic rifting is dealt with in a series of regional papers focused on the Cenozoic development of the northwest European margin. Wold presents palaeobathymetric reconstructions of the North Atlantic between the Charlie Gibbs and Jan Mayen fracture zones. Ebdon et al. describe the early Tertiary evolution of the Faeroe-Shetland Basin; Boldreel & Andersen, Andersen & Boldreel (extended abstracts) and Vanneste et aL describe the interplay between tectonics, sedimentation and deep-water circulation in the Faeroe-Rockall area; Thomson et al. describe the erosional history of the Inner Moray Firth area of the North Sea. The remaining papers have a varied content. Kiorboe & Petersen present the results of a seismic investigation of the Faeroe basalts and underlying strata. On the adjacent shelf, the

nature and stratigraphy of the sediments overlying the basalts are described from dredge samples by Waagstein & Heilmann-Clausen. The relationship between continental separation and palaeoceanographic development in the equatorial Atlantic is discussed by Jones et al., whilst Haslett uses radiolarian biostratigraphy to interpret the Plio-Pleistocene palaeoceanographic record of the North Atlantic. The influence of glacigenic sedimentation on late Cenozoic slope-apron development on the continental margin off northwest Britain is described by Stoker; a summary of analogous deposits from the east Canadian, east Greenland and Barents Sea margins highlights the regional importance of glacigenic processes throughout the North Atlantic region. The use of tephrochronology in the correlation and dating of late Quaternary sediments in the North Atlantic is described by Hunt, who also draws attention to some of the problems involved in the use of this method. The editors are grateful to all of the people who helped with the organization and running of the meeting; to those who refereed papers; and to our respective institutions for secretarial, drafting and technical support. We are particularly grateful to the following oil companies who provided funds to cover the cost of the meeting: Amerada Hess, Amoco, BP, Chevron, Esso, Mobil, Phillips, Shell, Texaco and Unocal. Final processing of the manuscripts was undertaken by Angharad Hills of the Geological Society Publishing House. M. S. Stoker, R. A. Scrutton, G. B. Shimmield, A. W. Tudhope August 1994

Contents Preface

DRISCOLL, N. W., HOGG, J. R., CHRISTIE-BLICK, N. & KARNER, G. D. Extensional tectonics in the Jeanne d'Arc Basin, offshore Newfoundland: Implications for the timing of break-up between Grand Banks and Iberia

vii

1

SINCLAIR,I. K. Sequence stratigraphic response to Aptian-Albian rifting in conjugate margin basins: a comparison of the Jeanne d'Arc Basin, offshore Newfoundland, and the Porcupine Basin, offshore Ireland

29

EBDON, C. C., GRANGER, P. G., JOHNSON, H. & EVANS, A. M. Early Tertiary evolution and sequence stratigraphy of the Faeroe-Shetland Basin: implications for hydrocarbon prospectivity

51

BOILLOT,G., BESLIER,M. O., KRAwczYK, C. M., RAPPIN, D. & RESTON,T. J. The formation of passive margins: constraints from the crustal structure and segmentation of the deep Galicia margin, Spain

71

RESTON, T. J., KRAWCZYK,C. M. & HOFFMANN, H.-J. Detachment tectonics during Atlantic rifting: analysis and interpretation of the S reflection, the west Galicia margin

93

KIORBOE, L. & PETERSEN, S. m. Seismic investigation of the Faeroe basalts and their substratum

111

VANNESTE,K., HENRIET,J.-P., POSEWANG,J. & THEILEN,F. Seismic stratigraphy of the Bill Bailey and Lousy Bank area: implications for subsidence history

125

ANDERSEN, M. S. & BOLDREEL, L. O. Effect of Eocene-Miocene compression structures on bottom-water currents in the Faeroe-Rockall area

141

BOLDREEL, L. O. & ANDERSEN, M. S. The relationship between the distribution of Tertiary sediments, tectonic processes and deep-water circulation around the Faeroe Islands

145

STOKER, M. S. The influence of glacigenic sedimentation on slope-apron development on the continental margin off Northwest Britain

159

WAAGSTEIN, R. & HEILMANN-CLAUSEN,C. Petrography and biostratigraphy of Palaeogene volcaniclastic sediments dredged from the Faeroes shelf

179

JONES, E. J. W., CANOE, S. C. & SPATHOPOULOS,F. Evolution of a major oceanographic pathway: the equatorial Atlantic

199

ANDERSEN, M. S. & BOLDREEL, L. O. Tertiary compression structures in the Faeroe-Rockall area

215

HASLETT, S. K. Plio-Pleistocene radiolarian biostratigraphy and palaeoceanography of the North Atlantic

217

HUNT, J. B., FANNIN, N. G. T., HILL, P. G. & PEACOCK, J. D. The tephrochronology and radiocarbon dating of North Atlantic, late Quaternary sediments: an example from the St Kilda Basin

227

THOMSON, K. & HILLIS, R. R. Tertiary structuration and erosion of the Inner Moray Firth

249

WOLD, C. N. Palaeobathymetric reconstruction on a gridded database: the northern North Atlantic and southern Greenland-Iceland-Norwegian Sea

271

Index

303

Extensional tectonics in the Jeanne d'Arc Basin, offshore Newfoundland: implications for the timing of break-up between Grand Banks and Iberia N E A L W. D R I S C O L L , 1'2 J O H N R. H O G G , 3 N I C H O L A S & GARRY

C H R I S T I E - B L I C K 1'2

D. K A R N E R 1

1Lamont-Doherty Earth Observatory o f Columbia University, Palisades, New York, 10964, USA 2also Department of Geological Sciences, Columbia University 3petro-Canada Resources, Calgary, Alberta, T2P 3E3, Canada

Abstract: Using seismic reflection and exploratory well data from the Jeanne d'Arc basin,

offshore Newfoundland, we examined the link between unconformity generation and the onset of seafloor spreading between the central Grand Banks and Iberia. A prominent unconformity developed across the entire basin, previously interpreted as a 'break-up' unconformity, is reinterpreted as a late Barremian/early Aptian rift-onset unconformity on the basis of the stratal geometry and lithofacies. The rotation and divergence of seismic reflectors above this unconformity attest to differential subsidence documenting an episode of extension and block rotation within the basin at this time. Our seismic sequence analysis suggests that rifting and block rotation continued in the Jeanne d'Arc basin until at least late Aptian/early Albian time. The onset of seafloor spreading between the central Grand Banks and Iberia is uncertain because of limited marine magnetic and drilling data (ODP & DSDP), and the existence of the Cretaceous magnetic quiet zone along the margin. However, recent studies indicate that magnetic anomaly M0 (118 Ma) is not well resolved north of the Newfoundland Seamounts within the Newfoundland basin and is not present north of the Figueiro fracture zone along the conjugate Iberian margin. This suggests that seafloor spreading between the northern portion of the Newfoundland basin and the northern Iberian margin began after the early Aptian. Given that the cessation of rifting marks the onset of seafloor spreading our seismic sequence analysis indicates that the onset of seafloor spreading in the northern Newfoundland basin, north of the Newfoundland Seamounts, began after late Aptian time.

The sedimentary record along passive margins is punctuated by unconformities (Vail et al. 1977; Vail 1987). An unconformity, as defined by Mitchum (1977), is a surface separating older from younger strata, along which there is evidence of nondeposition or erosion (subaerial and/or submarine) with a significant hiatus indicated. Subsequently, Posamentier et al. (1988) and Van Wagoner et al. (1988) defined an unconformity as a surface separating older from younger strata, along which there is evidence of truncation by subaerial erosion (and possibly correlative submarine erosion) or subaerial exposure, with a hiatus indicated. This definition of unconformity is more restrictive than the definition used by Mitchum (1977), thereby limiting the usage of the term. In this study, we adhere to the more general definition of unconformity proposed by Mitchum (1977) because it is not always possible to discern

whether a submarine erosional or non-depositional surface is correlative with, or necessarily implies, subaerial exposure or erosion. Along many passive continental margins, the u n c o n f o r m i t y t h a t is a p p r o x i m a t e l y timeequivalent to the onset of seafloor spreading has been termed the break-up unconformity (Falvey 1974). Determining the onset of seafloor spreading on the basis of marine magnetic and drilling data at some passive margins is difficult owing to the existence of magnetic quiet zones and thick wedges of clastic sediment overlying basement, in many instances, the age of the break-up unconformity ascertained from seismic reflection and drilling data is used as a proxy for estimating the time at which rifting ceased and seafloor spreading began (Falvey 1974; Hubbard et al. 1985; Tankard & Welsink 1987; Boillot & Winterer 1988; Meador & Austin 1988; Meador et al. 1988; Austin et al. 1989; Tankard et al.

From Scrutton, R. A., Stoker, M. S., Shimmield, G. B. & Tudhope, A. W. (eds), 1995, The Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region, Geological Society Special Publication No. 90, pp. 1-28

2

N.W. DRISCOLL E T A L . 295 ~ 55oc~,

300...~~

31t5~

3111~

315~

J Orphan ~Knoll

50"

~5~ ~

Newfoundland

45 ~

~5~

Grand Banks

os .~l ~

O

o

Newfoundland 9 Seamounls f

o<

,_3 ~

__.,,

40~

295 ~

300~ -

3tl5~ ~

310 ~

315 ~

Fig. 1. Location map for the Jeanne d'Arc basin, offshore Newfoundland. The hatchered box denotes the study area. Note the location of Newfoundland Seamounts along the eastern margin of the Grand Banks. The arrows indicate the counterclockwise rotation of the extension direction from northwest-southeast to northeastsouthwest from the central Grand Banks to Orphan basin during the Cretaceous. 1989; Tucholke et aL 1989; Embry & Dixon, 1990). One of the primary difficulties with this approach is correctly identifying the break-up unconformity. In some instances, the most prominent unconformity identified along the margin is termed the break-up unconformity, regardless of the associated stratal configuration (Cloetingh et al. 1989). The mechanism of unconformity generation is then attributed to either thermal uplift or in-plane force variations resulting from the cessation of rifting and the onset of sea floor spreading (e.g., Falvey 1974; Cloetingh et al. 1989; Cathles & Hallam 1991). To avoid ambiguity, we employ the following criteria to identify the break-up unconformity. (1) Bedding within the sedimentary succession beneath the unconformity tends to diverge toward depocentres as a result of differential subsidence due to localized block rotation during rifting. In contrast, the sediments over-

lying the unconformity typically have greater spatial persistence and more uniform thickness reflecting the regional subsidence associated with the cooling and contraction of the lithosphere (Falvey 1974; Meador & Austin 1988; Embry & Dixon 1990; Karner et al. 1993). (2) Growth faults associated with expanded sedimentary sections on the down-thrown block normally occur beneath the unconformity. (3) Faulting and offset should diminish markedly across the break-up unconformity. (4) The subsidence rate generally decreases across the unconformity marking the transition from rift-related to purely thermal subsidence (Hegarty et al. 1988; Hiscott et al. 1990). (5) Igneous activity tends to be preferentially associated with the sedimentary section beneath the unconformity (Falvey 1974; Enachescu 1987, 1988; Chang et al. 1988; Embry & Dixon, 1990). Using seismic reflection and exploratory well data from the Jeanne d'Arc basin, offshore

EXTENSIONAL TECTONICS

3

Fig. 2. Map of the northern North Atlantic modified after Srivastava and Tapscott (1986) showing magnetic anomalies along both the Grand Banks and Iberian margins. Note that the location of magnetic anomaly M0 is not shown in the Newfoundland basin on this rendition. Subsequent magnetic studies suggest that anomaly M0 is located near and parallel to the 4000m contour along the central Grand Banks, shown in Fig. 1. (Abbreviations: NB, Newfoundland basin; FC, Flemish Cap; OK, Orphan Knoll). Newfoundland (Figs 1 and 2), we assess the link between unconformity generation and the onset of seafloor spreading by determining the temporal and spatial development of the unconformities that appear to be coeval with the onset of seafloor spreading around the Grand Banks region. The main objective of this paper is to investigate the origin of the late Barremian/early Aptian and late Aptian/early Albian unconformities and in so doing, determine if they are actually break-up unconformities as previously interpreted (Srivastava & Tapscott 1986; Keen et al. 1987; Tankard & Welsink 1987; Keen & deVoogd 1988; Hubbard 1988; Tankard et al. 1989; Cloetingh et al. 1989; Tucholke et al. 1989; Srivastava et al. 1990). In our interpretation, the

late Aptian unconformity related to the last phase of rifting is also the break-up unconformity marking the onset of seafloor spreading. The overlying sediments represent the thermal or post-rift subsidence following this last phase of rifting. New interpretation of the magnetic data along the Grand Banks (Enachescu 1988; Cande et al. 1989; Srivastava et al. 1990) and the conjugate Iberian margin (Whitmarsh et al. 1990) suggests that seafloor spreading between the northern portion of the Newfoundland basin and the northern Iberian margin began after the early Aptian. Consequently, we propose that the onset of seafloor spreading in the northern Newfoundland basin was concomitant with the late Aptian cessation of rifting.

4

N.W. DRISCOLL E T AL.

Onset of seafloor spreading between the Grand Banks and lberia In order to investigate the link between unconformity generation and the onset of seafloor spreading, it was necessary to examine and define the onset of seafloor spreading around the Grand Banks region. The southern margin of the Grand Banks is a transform margin delineated by the Newfoundland fracture zone, which separates the Grand Banks and the Nova Scotian margin (Figs 1 and 2; Enachescu 1988; Todd et al. 1988; Verhoef & Srivastava 1989; Welsink et al. 1989; Srivastava et al. 1990; McAlpine 1991). The Bonnition (Salar) basin, located along the western edge of the Newfoundland basin, and the Orphan basin delineate the eastern and northern boundaries of the Grand Banks, respectively (Figs 1 and 2; Tankard & Welsink 1987; Enachescu 1988; Keen & deVoogd 1988; Austin et al. 1989; Tucholke et al. 1989; Welsink et al. 1989). On the basis of limited marine magnetic anomaly data, multi-channel seismic reflection data, and DSDP, ODP and exploratory drilling results, it is generally agreed that seafloor spreading began within the Scotian basin south of the Newfoundland fracture zone in the midJurassic (c. 175-180 Ma, Bajocian-Bathonian, Haworth & Keen 1979; Royden & Keen 1980; Klitgord & Schouten 1986; Ebinger & Tucholke 1988). The timing of the onset of seafloor spreading between the Grand Banks and Iberia is less firmly established and remains controversial. Mauffret et al. (1989) proposed on the basis of marine magnetic data and the existence of landward-dipping crustal reflectors that seafloor spreading between the southern Grand Banks and the southern Iberia margin commenced in the early Tithonian (magnetic anomaly M21). According to Mauffret et al. (1989), a subsequent ridge crest jump trapped late Tithonian to early Hauterivian oceanic crust along the portion of the Iberian margin that underlies the sediments of the Tagus Abyssal Plain. In contrast, between the Grand Banks (bounded by the Newfoundland Ridge to the south and Flemish Cap to the north; Figs 1 and 2) and Iberia, it has been proposed by Srivastava & Tapscott (1986), Keen & deVoogd (1988), Klitgord et al. (1988), and Verhoef & Srivastava (1989), that seafloor spreading began in the late Neocomian (c. 125 Ma, late Hauterivian). Nevertheless, the oldest identified magnetic anomaly in the Newfoundland basin east of the Grand Banks is magnetic anomaly M0 (Fig. 1; Meador et al. 1988; Austin et al. 1989; Cande et al. 1989; Tucholke et al. 1989; Verhoef &

Srivastava 1989). Therefore, on the basis of magnetic, seismic reflection, and refraction data, it has been proposed that seafloor spreading between the central Grand Banks and Iberia commenced at magnetic anomaly M0 time (c. 118 Ma, early Aptian; Enachescu 1988; Meador et al. 1988; Austin et al. 1989; Tucholke et al. 1989). Recent studies indicate that magnetic anomaly M0 might not continue north to the Flemish Cap as previously proposed, but terminates in the vicinity of the Newfoundland Seamounts (Fig. 1; Enachescu 1988; Srivastava et al. 1990). Along the conjugate Iberian margin magnetic anomaly M0 is not observed north of the Figueiro fracture zone (Whitmarsh et al. 1990) indicating that the onset of seafloor spreading in this region occurred after the early Aptian (Fig. 2; Enachescu 1988; Srivastava et al. 1990; Whitmarsh et al. 1990). Farther north, between Flemish Cap and Goban Spur, spreading began by at least late Albian time (Montadert et al. 1979; Tankard et al. 1989; Ziegler 1989; Whitmarsh et al. 1990). Seafloor spreading between Orphan Knoll and Porcupine Bank commenced at or before magnetic anomaly 34 during Santonian time (c. 84 Ma, Verhoef & Srivastava 1989; Ziegler 1989; Srivastava et al. 1990). As extension propagated from the central North Atlantic northward around the Grand Banks, its orientation appears to have rotated from WNW (late Barremian/early Aptian) to WSW (late Albian/early Cenomanian) (Figs 1 and 2; Keen & deVoogd 1988; Tankard et al. 1989; Verhoef & Srivastava, 1989; Welsink et al. 1989). The counter-clockwise rotation of the seafloor spreading direction during the Cretaceous is consistent with palaeostress patterns inferred from igneous dyke swarms (McHone, 1988). Therefore, our seismic sequence analysis has focused on the upper Barremian to base Tertiary stratigraphy in the Jeanne d'Arc basin to determine the tectonic and stratigraphic response of the basin to the northward propagation of seafloor spreading and the counterclockwise rotation of the inferred regional stress field.

Geological setting of the Jeanne d'Arc basin The geometry and distribution of rift basins across the Grand Banks are controlled primarily by the Mesozoic reactivation of the pre-existing fabrics in the Avalon and Meguma terranes (Haworth & Keen 1979; Hubbard et al. 1985; Tankard and Welsink, 1987; Enachescu 1988). The Jeanne d'Arc basin is the largest rift basin in

EXTENSIONAL TECTONICS 50" ;8

;9

\

a ,

9

5

48"

67"

t=8"

t,7"30

67"30

L

~/B o n o vi s t (~,,, Plotforn

47"

~7 a

LEGEND , ~ Bolin Bounding Fault Major Basin Fault Reverse Fault WELL STATUS 0 Location 9 Oil well 0

tt6a30

Gas wall

'i" Oil and go'= well J~- Abandoned

~6 o

50 o

-

49 o

~8 a

_1 ~ 6 ~ t,7 9

Fig. 3. Portion of the Grand Banks relevant to this study showing the Jeanne d'Arc, Anson, and Flemish Pass basins. The multi-channel seismic reflection and exploratory well data used in this study are shown. the Grand Banks region, containing approximately 14km of syn-rift sediment (Keen et al. 1987; Tankard & Welsink 1987). The Murre and Mercury curvilinear border faults separate the southern and central portions of the Jeanne d'Arc basin from the Bonavista platform to the west, respectively (Figs 3 and 4). The border faults trend approximately NNE. Interpretation of seismic refraction data indicates that the continental crust beneath the Bonavista platform is approximately 35km thick and thins eastward ttoabout 15 km beneath the Flemish Pass and Orphan basins (Fig. 3; Keen & Barrett 1981; Tankard & Welsink 1987). Hanging-wall monoclines and antithetic faults delineate the eastern boundary between the 'funnel-shaped' Jeanne d'Arc basin and the relatively unextended Central High (Figs 3 and 4).

The stratigraphic succession preserved in the Jeanne d'Arc basin records the complex interplay between tectonics and eustasy that affected the space available to deposit sediments across the Grand Banks region (Fig. 5; Jansa & Wade 1975; Hubbard et al. 1985; Tankard & Welsink 1987; Grant et al. 1988; Tankard et al. 1989; Tucholke et al. 1989; McAlpine, 1991). The Jeanne d'Arc basin was established in the late Triassic. Earliest deposits are fluvial to lacustrine sediments (Eurydice Formation; Jansa & Wade 1975; Enachescu 1987, 1988; Tankard et al. 1989; McAlpine, 1991). Repeated marine incursions into the rift systems from the Tethys Sea led to the development of widespread evaporite deposits in the basin (Argo Formation) that overly the Eurydice Formation (Tucholke et al. 1989; McAlpine 1991). Regional

6

N.W. DRISCOLL E T AL.

Fig. 4. A generalized structure map illustrating the relationship between the transbasinal fault zones (transfer zones) and the Murre and Mercury border faults. From south to north, the transfer zones are referred to in the text as Egret, Ammonite, and Nautilus. The fault displacement and the relief of the intrabasinal highs associated with the transfer zones diminish basinward. Bold lines show the location of the multi-channel seismic reflection shown in Figs 6, 7, 8, and 10. thermal subsidence of the Grand Banks region following the late Triassic to early Jurassic rifting episode created space for the deposition of marine sandstone, shales, and carbonates (Tankard & Welsink 1987; Grant et al. 1988; McAlpine 1991). The Jeanne d'Are Formation, a coarse-grained sandstone to conglomeratic braided-fluvial deposit overlying the carbonates of the Rankin Formation, has been interpreted to record another rifting episode that affected the Grand Banks in the late Jurassic (Tankard & Welsink 1987; Grant et al. 1988; Tankard et al. 1989; McAlpine 1991). During the late Jurassic/ early Cretaceous, basin infilling by elastic sediments (Whiterose and Hibernia formations) with the occasional interspersed carbonate stringer (Fig. 5; e.g., B marker, Valanginian) kept pace with the overall thermal subsidence (Hubbard et al. 1985; Welsink & Tankard 1988). New interpretations of seismic reflection and exploratory well data suggest that extensional deformation within the Jeanne d'Are basin north of the Egret transfer zone continued until at least late Aptian/early Albian time resulting in the generation of the late Barremian/early

Aptian and late Aptian/early Albian unconformities above which were deposited the Avalon and Ben Nevis formations, respectively (Fig. 5; Driscoll et al. 1990; Driscoll 1992; Karner et al. 1993; Driscoll & Hogg 1995). The unconformity underlying the Avalon Formation has been interpreted to be of middle Barremian to late Barremian/early Aptian age on the basis of our recent biostratigraphie studies of the palynomorph assemblages using both cores and well cuttings. For the sake of simplicity, we here use the late Barremian age when referring to this unconformity. In a similar fashion, we use the term late Aptian when referring to the unconformity beneath the Ben Nevis Formation.

Structure and stratigraphy Multi-channel seismic reflection data were used to map the regional and local structure of the Jeanne d'Arc basin in order to assess the influence of the pre-existing border fault/transfer fault geometry on the late Barremian to base Tertiary stratigraphy (Fig. 3). Five major seismic sequence boundaries were identified within the

EXTENSIONAL TECTONICS

7

Fig. 5. Generalized lithology and lithostratigraphy for the Jeanne d'Arc basin. Note the relationship between the late Barremian, late Aptian, and late Albian unconformities and the lithostratigraphy (modified from McAlpine 1991).

8

N.W. DRISCOLL E T AL.

Fig. 6. Interpreted and uninterpreted seismic reflection profile NF79-114 illustrating the collapsed hanging-wall deformation due to late Barremian extension across the Murre border fault. The reflectors that downlap onto the late Barremian unconformity form a prograding wedge geometry that correlates with locally-derived upper Barremian-lower Aptian sandstones. The overlying reflectors that onlap the prograding wedge correlate with marine sandstones that are more regionally persistent throughout the basin. Location of seismic profile is shown in Fig. 4.

Jeanne d'Arc basin by applying the technique of seismic sequence stratigraphy (Vail 1987). Sequence stratigraphy is the study of repetitive, genetically related strata bounded by unconfor-

mities and their correlative conformities within a time-stratigraphic framework (Vail 1987; Van Wagoner et al. 1988; Christie-Blick & Driscoll 1995). Sequence stratigraphic concepts may be

EXTENSIONAL TECTONICS applied to predict depositional environments from the stratal patterns and acoustic character observed in seismic reflection data. Isopach maps were generated to analyse the spatial variations of the stratigraphic sequences. Multichannel stacking velocities and velocity information derived from exploratory well data were used to convert two-way travel time into sediment thickness. The depositional palaeoenvironments were estimated by correlating the seismic reflection data to the exploratory well data. The chronostratigraphy used in this study is based primarily on palynology and, where available, augmented by microfossil zonations (foraminifera). We used the Kent & Gradstein (1985) D N A G time-scale. The formations proposed by Grant et al. (1988) and McAlpine (1991) for the Jeanne d'Arc basin are identified on the basis of lithic characteristics and position within the stratigraphic succession and are not necessarily time-stratigraphic units. The timestratigraphic sequences in the Jeanne d'Arc basin referred to in this paper were identified by seismic sequence stratigraphy and palynology. The Murre and Mercury border faults strike primarily NNE and are segmented by transfer zones (Fig. 4). The transfer zones are highlyfaulted regions that form intra-basinal highs. We propose that the formation of these intrabasinal highs is a direct consequence of differential displacement on segmented border faults and the collapse of the hanging-wall block in three dimensions (Driscoll & Hogg 1995). Westnorthwest extension across the Grand Banks reactivated the Murre and Mercury border faults in a normal sense during late Barremian, early Aptian, and late Aptian time (Driscoll 1992; Driscoll & Hogg 1995). The structural and stratigraphic response of the Jeanne d'Arc basin to these extensional events is illustrated using seismic reflection dip lines across the border faults (Figs 6-8). The location of these seismic reflection profiles is shown in Fig. 4. Seismic reflection profile NF79-114 (Fig. 6) crossing the Murre border fault illustrates hanging-wall collapse in response to the late Barremian extensional event. The internal deformation of the hanging wall is accommodated by a complex pattern of synthetic and antithetic faults (Fig. 6). The observed horizontal component of displacement across the border fault, therefore, does not reflect the actual extension between the hanging wall and footwall (e.g., White et al. 1986). Consequently, caution should be employed when using the displacement of the hanging wall with respect to the footwall across a fault as a proxy to estimate the amount of

9

extension. A thick wedge of sediment is developed along the fault and thins basinward by downlap onto the pre-existing strata. Exploratory well Hibernia G-55 is located on the northern edge of this depocentre and sampled thick accumulations of coarse-grained clastic sediments (c. 1000 m; Fig. 4). High-angle crossstratification, and pebble lags observed within the cored interval of G-55 (c. 2243.2-2453.9m below the Kelly Bushing, KB) are interpreted to be indicative of fluvial or fan-delta sedimentation. The cored section from Hibernia G-55 shows no evidence of bioturbation, but it is uncertain whether the sediments accumulated in a subaerial or subaqueous environment. The subdued rift flank topography near the intersection of the Murre border fault and the Ammonite transfer zone might have allowed the fluvial drainage system access to the basin (Fig. 4). The geometry and location of this sedimentary wedge along the Murre border fault suggests that it is a fan delta (e.g., Leeder & Gawthorpe, 1987; McPherson et al. 1987). The overlying late Barremian/early Aptian sediments, which can be traced more regionally throughout the basin, onlap against the locally-developed sedimentary wedge (Fig. 6). Palynomorphs and sedimentary structures observed within the exploratory cores indicate that the more regionally deposited sediments are marine and were deposited in a middle to lower shoreface (Driscoll & Hogg 1995). The onlap surface observed in the seismic reflection profile separates regionally-derived sandstones above (e.g., Hibernia P-15) from the locally-derived sandstones that comprise the prograding wedge below (e.g., Hibernia G-55; Fig. 4). Many rift basins display a similar twostage stratigraphic response to extension: a local response restricted by the local basin architecture (i.e., border/transfer fault interaction), and a more regional response governed by the larger border fault offsets of the rift system and existing drainage networks (Leeder & Gawthorpe 1987; Frostick & Reid 1989; Morley et al. 1990; Lambiase 1991; Driscoll 1992; Driscoll & Hogg 1995). Farther north across the Nautilus transfer zone, the Flying Foam structure parallels the Mercury border fault and thus isolates the Mercury sub-basin from the rift basin proper (Fig. 4). The Flying Foam structure appears to be a rider block of Avalon terrane overlying the Mercury border fault. The faults associated with the rotation and internal deformation of this rider block are also oriented roughly paral!el to the Mercury border fault. Experimental sand models of hanging-wall deformation with a

10

N. W. DRISCOLL E T AL.

Fig. 7. Interpreted and uninterpreted seismic reflection profile (NF79-108) that crosses the curvilinear portion of the Mercury border fault and the Flying Foam structure. The divergence and rotation of the seismic reflectors indicate three episodes of differential subsidence beginning in the late Barremian and culminating in the late Aptian. Note the upper Aptian to upper Albian sediments are flat-lying with minimal evidence suggesting divergence. The location and penetration of exploratory well Flying Foam 1-13 is projected onto the seismic reflection profile. Location of seismic profile is shown in Fig. 4.

EXTENSIONAL TECTONICS

11

Fig. 8. Interpreted and uninterpreted seismic reflection profile (HM81-70) that crosses the Mercury fault and subbasin. Extensional reactivation of the Mercury border fault occurred during the late Barremian, early Aptian, and late Aptian as evidenced by the divergent and onlapping seismic reflectors. The sequences that onlap the border fault show no indications of the late Albian reactivation. (Abbreviations: LB, late Barremian unconformity; EA, early Aptian unconformity, LAp, late Aptian unconformity; LAb, late Albian unconformity, and BT, Base Tertiary unconformity). listric ramp-flat geometry produce similar deformational features during extension (Tankard et al. 1989; McClay & Scott 1991). Seismic reflection line 79-108 traverses this structure obliquely (Fig. 4), and thus crosses the Mercury border fault where the fault curves toward the Nautilus transfer zone. The seismic reflectors underlying the late Barremian unconformity appear parallel and are roughly concordant with the underlying Kimmeridgian unconformity (Tankard et al. 1989). The overlying seismic reflectors onlap onto the top of the parallel, concordant reflectors (Fig. 7). The dip of the onlapping seismic reflectors diminishes upsection and is indicative of differential subsidence and block rotation (Fig. 7). Note that the

divergent reflectors onlap onto a single surface, indicating that the increase in accommodation outpaced the input of sediment. Each phase of differential subsidence is recorded by a divergent stratigraphic package. Eustatic variations alone cannot explain the divergence and rotation of these stratigraphic packages. Consequently, we propose that the late Barremian, early Aptian, and late Aptian unconformities document episodes of active block rotation, and are therefore rift-onset unconformities. Conversely, reflectors in the upper Aptian to upper Albian interval are parallel with minimal signs of rotation suggesting that extensional deformation had ceased by late Aptian time. Minor deformation of the upper Barremian to upper Albian section is

12

N.W. DRISCOLL E T AL.

observed near the curved portion of the Mercury border fault. The small undulations (folds) observed in the seismic section (Fig. 7) are associated with the late Albian deformation (see below). Seismic reflection line HM81-70 also images the thick upper Barremian to upper Albian sediment preserved in the basin (Fig. 8). The divergence and onlap of seismic reflectors overlying the late Barremian, early Aptian, and late Aptian unconformities attest to an increase in accommodation at these times (Figs 7 and 8). The thickness of these sequences progressively decreases eastward away from the basin depocentre. Now we present a simple schematic (Fig. 9) to illustrate how the seismic reflectors can be used to infer the deformational history of a basin. Parallel reflectors (Fig. 9a) are indicative of uniform subsidence across a basin and are interpreted as the regional thermal subsidence of an earlier extensional event. In the case of the Jeanne d'Arc basin, this thermal subsidence phase is likely a consequence of the late Jurassic to early Cretaceous rifting episode (Tankard & Welsink 1987; Grant et al. 1988; McAlpine 1991). Divergence and rotation of seismic reflectors is indicative of differential subsidence and block rotation (Fig. 9b). Note that the divergent reflectors resulting from a phase of differential subsidence onlap onto a single surface (Fig. 9b). Each phase of differential subsidence is recorded by an onlapping stratigraphic package (Fig. 9c). In contrast to the increase of accommodation recorded by the divergent and rotated reflectors seen in Figs 6-9, the late Aptian to late Albian seismic reflectors overlying the Hibernia structure (Fig. 10) record a decrease or reduction of accommodation. Specifically, the seismic reflectors overlying the late Aptian unconformity do not onlap the Hibernia structural high (Fig. 10). The seismic reflectors do, however, onlap the structural high to the southeast. Thus, the uplift of the Hibernia structure must have occurred after the deposition of the upper Aptian to upper Albian sediments. The Hibernia structure is a consequence of reactivation of the transfer faults in a reverse sense. The reactivation uplifted the same crustal blocks that were involved in earlier extensional events. Erosional truncation registering this uplift is best developed along the intersection between the transfer zones and border faults. Given the observed stratal geometry (i.e., truncation and onlap; Fig. 10), the minimum amplitude of the relative topography due to the brittle deformation is 250 m. From the distribution of the deformational

Fig. 9. A simple schematic illustrating the divergence and rotation of the sediments in an extensional setting due to differential subsidence (block rotation). (A) Parallel reflectors are indicative of regional uniform subsidence interpreted to be due to thermal subsidence across the basin. (13) Divergence and rotation of seismic reflectors, indicating differential subsidence and block rotation. (C) Final stratal geometry of the basin demonstrating three rifting episodes characterized by stratigraphic packages bounded by unconformities. Post-rift prograding and onlapping stratigraphic packages subsequently fill the basin. structures, it is possible to define the trajectory of the in-plane force responsible for the late Albian unconformity. For example, the structures associated with the late Albian reactivation are observed only across the transfer zones (Fig. 10). Seismic reflection profiles crossing the border faults (e.g., Figs 6 and 8) do not show any evidence of reactivation structures. Therefore, the orientation of the late Albian compression was N N E parallel to the Jeanne d'Arc basin. The Hibernia structure is a direct result of

EXTENSIONAL TECTONICS

13

Fig. 10. Interpreted and urtinterpreted multi-channel seismic reflection profile across the Jeanne d'Arc basin (NF79-103). The profile crosses both the Nautilus transfer zone and the Ammonite transfer zone. The triangular ridge of unextended Avalon terrane (i.e., Nautilus transfer zone) extends into the basin due to the interaction of the two border faults (e.g., relay ramp). Re-activation of the Nautilus transfer zone by north-northeast in-plane compression uplifted and deformed the Hibernia structure in the late Albian. Note the seismic reflectors overlying the late Aptian unconformity onlap the structural high towards the southeast, but do not onlap onto the Hibernia structure. In fact, the reflectors overlying the Hibernia structure are truncated. (Abbreviations: LB, late Barremian unconformity; LAp, late Aptian unconformity; LAb, late Albian unconformity; BT, Base Tertiary unconformity)

14

N.W. DRISCOLL ET AL.

Fig. 11. Hibernia K-14 well is located just to the south of Hibernia P-15. See Figs 3 and 4 for exact location. Correlation of the unconformities determined on the basis of stratal relationships observed in the seismic reflection data to lithology and geophysical logs is shown. late Albian compression, and represents reactivated and uplifted crustal blocks across the Nautilus transfer zone (Fig. 10; Driscoll 1992; Karner et al. 1993). Possible candidates for this compression are the in-plane forces created by the cessation of rifting between Flemish Cap and Goban Spur and/or the change in the regional stress field associated with the onset of rifting in the Labrador Sea. It is important to note that the amount of shortening that caused the late Albian deformation is minor.

Correlation of seismic stratigraphy to lithology Downhole acoustic impedance contrasts were calculated using velocity and density data recorded at a 0.Sm sampling interval from the exploratory wells. A series of filtered synthetic seismograms, generated by convolving a zero phase wavelet with the downhole acoustic impedance contrasts, allowed the lithology to be correlated to the seismic stratigraphy. The majority of wells in the Jeanne d'Arc basin were

located to exploit structural traps (i.e., culminations), and in so doing, the missing sedimentary section due to erosion and non-deposition is maximized (e.g., Hibernia P-15 and Hibernia B27). L a t e Barremian unconformity

Correlation of the seismic stratigraphy to the lithology and geophysical logs from the exploratory wells indicates that the late Barremian unconformity identified on the basis of onlap in the seismic reflection data correlates with the base of a fine- to medium-grained, shallowmarine, calcareous/siliceous sandstone (Fig. 11). The base of this sandstone correlates with a sharp decrease upwards in natural radioactivity recorded in the gamma ray log. The low natural radioactivity signature on the gamma ray log is consistent with our interpretation that these sandstones are indicative of a winnowed shoreface depositional environment. The sandstones gradually become finer-grained upsection and pass into glauconitic siltstones. This trend is

EXTENSIONAL TECTONICS

15

Fig. 12. A cored interval from Hibernia K-14 recovered iron-stained shales beneath the late Barremian unconformity. See Fig. 11 for location of core with respect to the late Barremian unconformity.

reflected in the gamma ray log values as the natural radioactivity increases up section. In addition to fining-upward, the sandstones also become gradually finer-grained away from the border faults toward the basin depocentre. The cored interval (c. 2345-2439m beneath

KB) at Hibernia K-14 permitted a detailed examination of the preserved lithology and sedimentary structures below and above the late Barremian unconformity (Figs 12-14). Carbonaceous shales immediately beneath the unconformity are dark grey to black as well as iron-

16

N.W. DRISCOLL ET AL.

Fig. 13. A cored interval from Hibernia K-14 recovered dark to grey shales with abundant gastropods and bivalves beneath the late Barremian unconformity. See Fig. 11 for location of core with respect to the late Barremian unconformity.

stained, with sideritic nodules and contain abundant root structures, and bivalve and gastropod fossils (c. 2412-2413m; Figs 12 and 13). These iron-stained, carbonaceous shales are characterized by low-velocity zones in the sonic logs and can be easily correlated across the basin (Fig. 11; McAlpine 1991). These observations, together with palynomorph data, are taken to indicate a non-marine and lagoonal to restricted marine depositional environment. In contrast, sandstones overlying the unconformity are massive to low-angle cross-stratified (possibly with hummocky cross-stratification; Fig. 14). Bio-turbation is relatively common and for the most part restricted to the Skolithos ichnofacies (for example: Ophiomorphia, Diplocraterion, and Skolithos, with minor occurrences of

Thalassinoides, Astrosoma, and Zoophycos). These features are interpreted to indicate a relatively energetic shallow-marine shoreface environment. Therefore, the late Barremian unconformity, defined by stratal patterns in the seismic reflection data, correlates to a marine onlap surface that records an increase in accommodation across the unconformity. Late Aptian unconformity The late Albian erosional truncation along and across the western portions of the transfer zones diminishes eastward, and thus minimizes the hiatus associated with the unconformity. The more continuous stratigraphic section within the basin permitted correlation of the additional

EXTENSIONAL TECTONICS

17

Fig. 14. A cored interval from Hibernia K-14 recovered sandstone units above the late Barremian unconformity. Sedimentological, ichnological, and palynological evidence suggests that the sandstones were deposited in an open marine to restricted marine environment. See Fig. 11 for location of core with respect to the late Barremian unconformity. unconformities identified on the basis of reflector geometry observed in the seismic reflection data to the sampled lithology. Terra Nova well K-18 is located along the eastern portion of the Ammol~;te transfer (Figs 3 and 4). At this well, the late Barremian unconformity is overlain by a lower Aptian blocky glauconitic sandstone (c. 1790m beneath KB; Fig. 15). Continuing upsection the grain size diminishes and the sandstones grade into siltstones and shales with a concomitant increase in shell fragments (Fig. 15). The late Aptian unconformity correlates (c. 1710m beneath KB) with an abrupt facies

change from calcareous shale to sandstone (Fig. 15). The lower to upper Albian interval overlying the late Aptian unconformity consists of fine- to very fine-grained calcite-cemented, glauconitic sandstones with thin interbeds of sandy, fossiliferous limestones (Fig. 15). The Albian sequence (c. 1550-1710m beneath KB) displays an overall fining-upward trend, from fine-grained sandstones at the base through bioturbated, glauconitic siltstones in the middle to glauconitic shales toward the top (Fig. 15). Examination of cores recovered from the West Ben Nevis B-75 well, located 15 km NNE of

18

N.W. DRISCOLL ET AL.

Fig. i5. Exploratory well Terra Nova K-18 is located along the eastern portion Ammonite transfer zone. See Figs 3 and 4 for exact location. Correlation of the unconformities determined on the basis of stratal relationships observed in the seismic reflection data to lithology and geophysical logs is shown. Note the thick sandstone overlying the late Aptian unconformity.

Terra Nova K-18 (Figs 3 and 4), suggests that the sandstones were deposited in a shallow shelf environment during early Albian time with open to restricted marine conditions. The cored interval (c. 2004-2095m beneath KB) consists of bioturbated sandstones with interbedded shell debris (pelecypods, gastropods, and serpulid worm tubes). The bioclastic layers are interpreted to represent storm events. Ichnological studies of the West Ben Nevis B-75 core indicate that the bioturbation is predominantly in the Skolithos ichnofacies at the base of the Albian sequence and shifts to the Cruziana ichnofacies toward the top of the sequence (primarily:

Ophiomorphia, Planolities and Teichichnus). The transition from Skolithos to Cruziana ichnofacies implies an overall increase in palaeowater depth (Pemberton et al. 1984), which is consistent with the observed upward fining trend, a feature that is characteristic of the Albian sequence sampled throughout the Jeanne d'Arc basin.

Late Albian unconformity The late Albian unconformity identified on the basis of reflector geometry at Hibernia P-15 correlates with an abrupt upward facies change

EXTENSIONAL TECTONICS

19

Fig. 16. Exploratory well Hibernia P-15 is located on the southern flank of the Nautilus transfer zone. See Fig. 2 for exact location. Correlation of the unconformities determined on the basis of stratal relationships observed in the seismic reflection data to lithology and geophysical logs is shown. Note that the late Barremian unconformity correlates with the base of a blocky marine sandstone. from glauconitic, calcareous shale to shelly sandstone (Figs 10 and 16). Palynomorphs from both Hibernia P-15 and B-27 indicate that the oldest rocks overlying the unconformity are of late Albian age. Furthermore, at Hibernia 0-35 well (c. 2184-2200m beneath KB), close to the Nautilus transfer zone, upper Albian very finegrained, highly carbonaceous sandstones were recovered above the unconformity and are quite different to the lower Albian sandstones beneath the unconformity.

Lithostratigraphy and subsidence history The late Barremian to late Aptian and the late Aptian to late Albian sequences correspond

roughly to the Avalon and Ben Nevis formations in the Jeanne d'Arc basin (Fig. 17; Grant et al. 1988; McAlpine 1991). According to the subdivisions suggested by McAlpine (1991), the Avalon Formation can be subdivided into three units: (1) a basal unit characterized by varicoloured shale, (2) a middle unit composed predominantly of sandstone, and (3) an upper unit that is a coarsening-upward sandstonedominated unit. The Ben Nevis Formation is defined by McAlpine (1991) as the first finingupward succession overlying the Avalon Formation. Basinward both of these formations pass laterally into the Nautilus Shale (Fig. 6; McAlpine 1991). Formations are defined on the basis of lithic characteristics and stratigraphic

20

N.W. DRISCOLL E T AL.

Fig. 17. Lithostratigraphy of the Jeanne d'Arc basin for the late Jurassic to late Cretaceous and its relation to the tectonic history of the basin. The late Barremian and late Aptian unconformities are rift onset unconformities that correlate to onlap surfaces in the seismic reflection data. The late Albian unconformity documents compression-induced uplift across the basin. position and do not necessarily have timestratigraphic significance. In contrast, time surfaces determined on the basis of reflector geometry permit us to correlate time-equivalent coarse-grained proximal facies to the finegrained distal facies within the basin (Fig. 17). Our detailed seismic sequence analysis of the Jeanne d'Arc basin for the late Jurassic to late Cretaceous allows us to link the lithostratigraphy with the tectonic history of the basin (Fig. 17). The late Barremian and late Aptian unconformities are interpreted as rift onset unconformities, recording the formation of a physiographic hole, and they are characterized in seismic reflection data by well-developed onlap. An alternative interpretation, that the unconformities are primarily due to eustatic fluctuations during a time of more-or-less continuous block tilting, is not consistent with the absence of predicted lowstand deposits in the closed palaeobathymetric lows. The increase in accommodation associated with these extensional events is significantly greater than the sediment input to the basin as evidenced by the fact that the seismic reflectors overlying these unconformities all onlap onto a single surface

(Figs 7 and 8). The Avalon Formation, which correlates to one of these onlapping seismic packages, is a variable siliciclastic sequence and displays an overall upward-shoaling trend, that in places, is accompanied by an upward-coarsening trend in response to sediments infilling the physiographic hole (Fig. 17). Consequently, the extensional events are followed by periods of tectonic quiescence as recorded by these onlapping packages that shoal upward as the available space fills with sediment. Conversely, if the sedimentation kept pace with the differential subsidence, then a fanning pattern of seismic reflectors would develop with each successive layer of sediment onlapping the previous horizon at or near the same position. This is not observed. In addition, minimal variations in palaeowater-depth and facies distribution would occur if the sedimentation kept pace with the subsidence rate. The erosional truncation associated with the late Albian unconformity is best developed along the intersection between transfer zones (Nautilus and Ammonite) and border faults (Murre and Mercury; Figs 4 and 10). This truncation registers the reactivation and inver-

EXTENSIONAL TECTONICS sion of the same crustal blocks that were involved in the extension process. During the late Albian, palaeowater depths along the transfer zones increased toward the east away from the border faults, and thus the deformed sediments overlying the transfer zones in the eastern portion of the Jeanne d'Arc basin were not truncated. Consequently, the upper Albian sandstones, derived from the cannibalization of the underlying sequences, have a very limited aerial distribution in the basin (Fig. 17). The large magnitude of the post-Aptian thermal subsidence, a function of mantle extension, is not consistent with the minor late Barremian to late Aptian brittle crustal extension (Figs 6-8). In addition, because of the elapsed time since the late Triassic and inferred middle/late Jurassic extensional deformation (c. 60-200 Ma), it is difficult to explain the observed post-Aptian thermal subsidence by this earlier phase of extension in the Jeanne d'Arc basin. Therefore, we propose that the large magnitude of post-Aptian thermal subsidence associated with the rather minor late Barremian-late Aptian brittle deformation observed in the basin results from the Grand Banks region moving off the Newfoundland hotspot and the re-equilibration of the lithosphere to normal thicknesses beginning in the late Aptian (e.g., Newfoundland Seamounts, Fig. 1). The Grand Banks region was associated with volcanism, which we attribute to the Newfoundland hotspot, prior to the late Jurassic/early Cretaceous (Jansa & PePiper, 1988). Because lithospheric cooling rates are directly proportional to the lithospheric thickness, the close proximity of the hotspot modifies the equilibrium of the cooling extending lithosphere. Consequently, while the Grand Banks remained in close spatial proximity to the Newfoundland hotspot, its rate of subsidence was less than that for normal cooling lithosphere. While this mechanism accounts for diminished rates of lithospheric subsidence, it does not explain the accelerated rates and the large magnitude of subsidence subsequent to the late Barremian-late Aptian rifting. However, we note that seafloor spreading moved the Grand Banks away from the proximity of the hotspot after breakup, thus allowing extended lithosphere to re-equilibrate thermally to normal lithospheric thickness (c. 120km) with a concurrent increase in subsidence rates. Jansa & Pc-Piper (1988) dated basalts dredged from the Newfoundland Seamounts as 97 + 1.7 Ma indicating that the Grand Banks region had moved off the hotspot by at least late Albian time,' timing that is consistent with the increased thermal subsidence in the Jeanne d'Arc basin.

21

Conversely, if significant lower crustal and mantle extension occurred across the Grand Banks with only minor upper crustal deformation during the late Barremian to late Aptian time, then this style of extensional deformation could also account for the observed subsidence patterns across the Grand Banks. Such a distribution of extension implies the existence of a westward dipping intracrustal detachment that effectively thins the lower crust and generates thermal-type subsidence across the region. We require that this detachment has a ramp-fiat-ramp geometry such that it breached the surface close to the position of the continent-ocean boundary (Fig. 1; eastern edge of the Newfoundland basin). The 'fiat' component of the detachment occurred at mid-crustal depths across the Grand Banks region and ramped again beneath the North American continent. In this interpretation, the Grand Banks represents an upper plate margin and is separated from the lower plate margin (i.e., Iberia) by a westward dipping detachment (e.g., Enachescu 1987; Tankard & Welsink 1987; Keen & deVoogd 1988).

Isopach map Multi-channel stacking velocities and velocity information derived from well data were used to convert two-way travel time into sedimentary thickness in order to create the late Barremianlate Aptian isopach map. The isopach map illustrates the regional distribution and the overall northward-thickening of the upper Barremian to upper Aptian sediments in the Jeanne d'Arc basin (Fig. 18). Within portions of the highly faulted Nautilus transfer zone, only the late Barremian and late Albian bounding unconformities could be confidently identified. Identification of the late Aptian unconformity on the basis of stratal patterns in the highly faulted region proved to be difficult (Fig. 18). Thick accumulations of upper Barremianupper Aptian sediments are preferentially preserved along the border faults (e.g., Mercury K76 and Hibernia G-55; Fig. 18). The thickness and distribution of the syn-rift sediments, together with the marine magnetic anomalies, are consistent with our inferred WNW extension direction (Figs 2 and 18). Toward the south, a thick wedge of sediment was deposited near the intersection of the Egret border fault and the Egret transfer zone (Fig. 18). An erosional channel approximately 15 km wide can be traced away from the depocentre across the Murre subbasin (Fig. 18). The late Barremian depocentre near the intersection of the Egret border fault

22

N.W. DRISCOLL E T AL.

Fig. 18. Isopach map illustrating the regional distribution of upper Barremian to upper Aptian sediments. Note the influence of the border fault transfer zone geometry on the thickness of the sediments. The overall sedimcmt thickness increases across the transfer zones toward the north within the basin. and the Egret transfer zone, and the distribution and configuration of the 250 m contour toward the north within the basin suggest that the erosional channel was a major conduit for sediment to enter the southern Jeanne d'Arc basin at this time. A pronounced increase in sediment thickness is observed along the central segment of the Murre border fault just north of the Ammonite transfer zone (Fig. 18). The upper Barremianupper Aptian sediment thickness in this region exceeds 1500 m. Exploratory well Hibernia G-55 is located on the northern edge of this depocentre and sampled thick accumulations of coarse-grained clastic rocks (Fig. 6; c. 1000 m). The subdued rift flank topography near the intersection of the Murre border fault and the Ammonite transfer zone might have allowed

fluvial drainage systems access to the basin (Fig. 18). T h i c k accumulations of upper Barremian to upper Albian sediments (c. 2000 m, Mercury K76) are preserved in the Mercury sub-basin, which is located north of the Nautilus transfer and west of the Flying Foam structure. Seismic reflection profiles across and along the Bonavista platform indicate that there is little if any evidence for erosional down-cutting that might be associated with large drainage systems traversing the platform and supplying sediment to the Mercury sub-basin. This could be the consequence of the highly indurated metasedimentary rocks composing the Bonavista platform or the very low gradients across the platform precluding large amounts of erosional downcutting, or both.

EXTENSIONAL TECTONICS

Fig. 19. A generalized reconstruction of the North Atlantic during the late Barremian/early Aptian showing the boundary between continental and oceanic crust. The Newfoundland Seamounts (N.S.)/ Figueiro fracture zone separates regions undergoing continued continental extension toward the north from regions in the south where seafloor spreading had already begun. Gravity and magnetic lineations across the Grand Banks indicate that the eastern continuation of the Egret transfer zone in the Jeanne d'Arc basin (JD) is roughly coincident with the Newfoundland Seamounts offshore. Late Barremian to late Aptian extension Seismic reflection data indicate that the late Barremian-late Aptian and the late Aptian-late Albian sequences are faulted and offset. Within the transfer zones, the sequences are highly faulted and deformed. The fault displacement and relief of the intrabasinal highs associated with the transfer zones diminishes basinward. The deformation within the transfer zones appears to record the three-dimensional collapse of the hanging-wall blocks near the termination of border fault segments (Fig. 4). Growth faults and expanded sedimentary sections developed

23

across faults on the down-thrown blocks are observed only within the late Barremian-late Aptian sequence. From our analysis, large faultcontrolled accommodation generated during both late Barremian and late Aptian time is evidenced by the seismic reflection profiles, exploratory well data, and isopach map (Figs 6-18). Therefore, on the basis of these observations and the previously mentioned criteria necessary to identify a break-up unconformity, we conclude that the late Barremian and late Aptian unconformities observed in the Jeanne d'Arc basin are not 'break-up unconformities' as previously interpreted or modelled (Enachescu 1987; Tankard & Welsink 1987; Meador et al. 1988; Cloetingh et al. 1989; Kusznir & Egan 1989; Tankard et al. 1989; Tucholke et al. 1989). In fact, to the contrary, these unconformities are rift-onset unconformities documenting renewed phases of rifting. In our interpretation, rifting in the Jeanne d'Arc basin continued to at least late Aptian time with the overlying sediment representing the thermal phase of subsidence associated with this last phase of rifting on the Grand Banks. The thickness of upper Barremian to upper Aptian sediments in the Jeanne d'Arc basin systematically increases across the transfer zones from south to north within the discrete subbasins (Fig. 18). In conjunction with an increase in sediment thickness, there is also an increase in the number of upper Barremian to upper Albian rotated and divergent sedimentary packages observed within the northern sub-basins. Consequently, the structural depth to the late Barremian unconformity increases from south to north within the sub-basins (Tankard & Welsink 1987; Tankard et al. 1989). Correlation of the seismic reflection and well data indicates that the overall northward dip of the basin predominantly developed after the late Barremian (e.g., Tankard et al. 1989). Decompaction and backstripping of sediments from exploratory wells in the Jeanne d'Arc basin demonstrate a marked increase in the basement subsidence during the late Barremian (c. 120 Ma; Hiscott et al. 1990).

Implications for break-up As previously mentioned, new interpretations of the magnetic data north of the Newfoundland Seamounts suggest that seafloor spreading between the northern portion of the Newfoundland basin and the northern Iberian margin began after the early Aptian, timing indistinguishable from the timing of rift cessation in the Jeanne d'Arc basin determined by sequence

24

N.W. DRISCOLL E T AL. et al. (1989; Fig. 19). However, this is not the

Fig. 20. A generalized reconstruction of the North Atlantic during the late Aptian/early Albian showing the boundary between regions undergoing continental extension and regions undergoing seafloor spreading. Note that the northward propagation of seafloor spreading is predicted to occur abruptly, jumping from one transfer zone to the next as extension exceeds some threshold. analysis. Gravity and magnetic lineations across the Grand Banks indicate that the eastern continuation of the Egret transfer zone is roughly coincident with the trend of the Newfoundland Seamounts offshore (Welsink et al. 1989). The Egret transfer zone is an important structural boundary separating the southern and central Jeanne d'Arc basin (Fig. 4). Upper Barremian to upper Albian sediments are not preserved in the southern Jeanne d'Arc basin (Fig. 18). In fact, in the southern Jeanne d'Arc basin the late Barremian unconformity separates dipping and truncated upper Jurassic/lower Cretaceous sediments from the overlying upper Cretaceous sediments (e.g., Murre G-67). Consequently, we propose that the onset of seafloor spreading in the southern Newfoundland basin began by at least late Barremian to early Aptian time consistent with the predictions of Tucholke

case for the northern Newfoundland basin. The Newfoundland Seamounts and Figueiro fracture zone delineated a spatial and temporal boundary that separated seafloor spreading in the south from continental extension to the north. On the basis of the observed deformational structures and distribution of syn-rift sediment, together with the trend of the oldest marine magnetic anomalies (Fig. 2), we propose that the extension direction was predominantly west-northwest to west. The onset of seafloor spreading in the northern Newfoundland basin was concomitant with the late Aptian cessation of rifting in the Jeanne d'Arc basin (Fig. 20). Because the northward propagation of seafloor spreading in the region (Figs 19 and 20) took place during the Cretaceous quiet zone (c. 12084Ma), it is difficult to determine from the magnetic lineations whether the northward propagation of seafloor spreading occurred continuously (i.e., 'zipper' opening) or by a series of abrupt events where the onset of seafloor spreading jumped across transfer zones from one segment to the next. Nevertheless, the abrupt change in the late Barremian to late Aptian extensional deformation across the Egret transfer zone in the Jeanne d'Arc basin (Fig. 18) suggests that the northward propagation of seafloor spreading in this region occurred by a catastrophic jump from the southern Newfoundland basin to the northern Newfoundland basin in late Aptian times. We propose that the nature of how seafloor spreading propagates, be it gradual or catastrophic, is related to the proximity of the region to the pole of opening and to large offset transfer zones that can effectively displace the deformation away from the extending region. The propagation of seafloor spreading between the Kerguelen-Heard Plateau and Broken Ridge has also been interpreted to have occurred by a catastrophic jump across a transfer zone (Driscoll et al. 1989). An alternative interpretation of the marine magnetic data, suggesting that seafloor spreading between the central Grand Banks and lberia began at or before magnetic anomaly M0 time, requires that rifting and block rotation in the Jeanne d'Arc basin were contemporaneous with seafloor spreading. This interpretation would challenge the existing paradigm that the brittle deformation of extended continental lithosphere ceases when seafloor spreading begins (Falvey 1974; Cathles & Hallam 1991; Karner et al. 1993). It is difficult to understand in twodimensions how tensional forces would be maintained within the plate, given the existence of the nascent ridge crest and the attendant

EXTENSIONAL TECTONICS

compressional forces associated with ridge push. The northward propagation of seafloor spreading and the counter-clockwise rotation of the extension direction in the North Atlantic, however, could have generated a complex three-dimension stress field thereby reactivating the brittle deformation in regions already undergoing seafloor spreading. If this interpretation is correct, then we infer that a transfer zone separated regions that were undergoing rifting in the north (i.e., between Flemish Cap and Goban Spur) from regions where seafloor spreading had already commenced (i.e., between central Grand Banks and Iberia). In this scenario, rifting between Flemish Cap and Goban Spur influenced regions south of the transfer zone, thus reactivating crustal blocks in the Jeanne d'Arc basin even though seafloor spreading was occurring toward the east between the central Grand Banks and Iberia.

Conclusions In summary, the tectonic and stratigraphic evolution of the Jeanne d'Arc basin resulted from a number of rifting events beginning in the late Triassic and culminating in the late Aptian. The previously interpreted early Aptian breakup unconformity is actually a rift-onset unconformity documenting a late Barremian phase of extension. The rotation and divergence of the seismic reflectors above this unconformity attest to differential subsidence due to localized block rotation during rifting. In the rift basin, the late Barremian unconformity marks the transition from subaerial to submarine deposition. That is, the onlap patterns observed in the seismic reflection data, which define the unconformity, are actually recording a marine flooding surface within the basin. This increase in accommodation is further evidence of continued extension within the Jeanne d'Arc basin at this time. The unconformity related to the last phase of rifting is dated as late Aptian. In our interpretation, the unconformity related to the last phase of rifting is also recording the onset of seafloor spreading. The overlying sediments would then represent the thermal subsidence following this last phase of rifting. Recent studies indicate that magnetic anomaly M0 is not well resolved north of the Newfoundland Seamounts within the Newfoundland basin (Cande et al. 1988; Enachescu, 1988; Srivastava et aL 1990) and is not present north of the Figueiro fracture zone along the conjugate Iberian margin (Whitmarsh et al. 1990). This new interpretation of the magnetic data north of the Newfoundland Seamounts

25

suggests that seafloor spreading between the northern portion of the Newfoundland basin and the northern Iberian margin began after the early Aptian, timing that is indistinguishable from the cessation of rifting in the Jeanne d'Arc basin determined by sequence analysis. Consequently, we propose that the onset of seafloor spreading in the northern Newfoundland basin was concomitant with the late Aptian cessation of rifting in the Jeanne d'Arc basin. We would like to acknowledge discussions with M. Enachescu, D. McAlpine, B. Tucholke, S. Srivastava, J. Wheeler, D. Hunter and J. Weissel. J. Weissel, D. Hayes and G. Mountain critically read an earlier version of this manuscript and their comments are greatly appreciated. We also thank Petro-Canada Resources for granting us access to the seismic reflection and exploratory well data from the Jeanne d'Arc basin and surrounding region. This work was supported by National Science Foundation grant OCE 88-12509 and a research grant from Petro-Canada Resources. Lamont-Doherty Earth Observatory contribution no. 5256.

References AUSTIN, J. A., Jr., TUCHOLKE, B. E. & UCI-IUPI, E. 1989. Upper Triassic-Lower Jurassic salt basin southeast of the Grand Banks. Earth and Planetary Science Letters, 92, 357-370. BOILLOT, G. & WINTERER, E. L. 1988. Drilling on the Galicia margin: retrospect and prospect. In; BOILLOT, G., WINTERER, E. L., MEYER, A. W. ET AL. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 103, College Station, TX, 809-828. CANDE, S. C. ET AL. 1989. Magnetic lineations of the worM's ocean basins. AAPG Map series. CATHLES, L. i . • HALLAM, A. 1991. Stress-induced changes in plate density, Vail sequences, epeirogeny, and short-lived global sea level fluctuations. Tectonics, 10, 659-671. CHANG, H. K., KOWSMANN,R. O. & DE FIGUEIREDO, A. M. F. 1988. New concepts on the development of the east Brazilian marginal basins. Episodes, 11, 194-202. CHRISTIE-BLICK, N. & DR!SCOLL, N. W. 1995. Sequence Stratigraphy. Annual Review of Earth and Planetary Sciences, 23, 451-478. CLOETINGH, S., TANKARD, A. J., WELSINK, H. J. dk JENKINS, W. A. i . 1989. Tectonic linkage and Mesozoic of Eastern Canada, an explanation for Vail's coastal onlap curves: In: TANKARD, A. J. & BALKWlLL, H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins. AAPG Memoir, 46, 283-293. DRISCOLL, N. W. 1992. Tectonic and Depositional Processes Inferredfrom Stratal Relationships. PhD thesis, New York, Columbia University. - & HOGG, J. R. 1995. Stratigraphic Response to Basin Formation: Jeanne d'Arc Basin, Offshore

26

N . W . DRISCOLL ET AL.

Newfoundland. In: LAMBIASE,J. J. (ed.) Hydrocarbon Habitat in Rift Basins. Geological Society, London, Special Publication 80, 145-163. - - , KARNER, G. O. d~ CHRISTIE-BLICK, N. 1990. Stratigraphic Evidence from Jeanne d'Arc basin for the Timing of Break-up Between Grand Banks and Iberia. (Abst.) EOS, 71, 1563. & WEISSEL, J. K. 1989a. Unconformities: Are Variations of In-Plane Force a Viable Mechanism for Their Formation? (Abst.) EOS, 70, 1345. , & 1989b. Stratigraphic and tectonic evolution of Broken Ridge from seismic stratigraphy and Leg 121 drilling. In: PEIRCE, J., WEISSEL, J. K. E T A L . Proceedings of the Ocean Drilling Program, Initial Reports, 121A, 71-93, College Station, TX. EBINGER, C. J. & TUCHOLKE, B. E. 1988. Marine geology of Sohm Basin, Canadian Atlantic Margin. American Association of Petroleum Geologists Bulletin, 72, 1450-1468. EMBRY, A. F. & DIXON, J. 1990. The break-up unconformity of the Amerasia basin, Arctic Ocean: evidence from Arctic Canada. Geological Society of America Bulletin, 102, 1526-1534. ENACHESCU, M. E. 1987. Tectonic and structural framework of the northeast Newfoundland con, tinental margin. In: BEAUMONT, C. & TANKARD, A. J. (eds) Sedimentary Basins and Basin-forming Mechanisms, CSPG Memoir, 12, 117-146. 1988. Extended basement beneath the intracratonic rifted basins of the Grand Banks of Newfoundland. Canadian Journal of Exploration Geophysics, 24, 48-65. FALVEY, D. A. 1974. The development of continental margins in plate tectonic theory. APEA Journal, 14, 95-106. FROSTICK, L. E. & REID, I. 1989. IS structure the main control of river drainage and sedimentation in rifts? Journal of African Earth Sciences, 8, 165182. GRANT, A. C., JANSA, L. F., MCALPINE, K. D. & EDWARDS, A. 1988. Mesozoic~enozoic geology of the eastern margin of the Grand Banks and its relation to Galicia Bank. In: BOILLOT, G. WINTERER, E. L. ET AL. Proceedings of the Ocean Drilling Program, Scientific Results, 103, 787-808. HAWORTH, R. T. & KEEN, C. E. 1979. The Canadian Atlantic Margin, a passive continental margin encompassing an active past. Tectonophysics, 59, 83-126. HEGARTY, K. A., WEISSEL, J. K. & MUTTER, J. C. 1988. Subsidence history of Australia's southern margin: constraints on basin models: American Association of Petroleum Geologists Bulletin, 72, 615-633. HISCOTT, R. N., WILSON, R. C. L., GRADSTEIN,F. M., PUJALTE, V., GARCIA-MONDEJAR, J., BOUDREAU, R. R. d~ WISHART, H. A. 1990. Comparative stratigraphy and subsidence history of Mesozoic rift basins of North Atlantic. American Association of Petroleum Geologists Bulletin, 74, 60-76. HUBBARD, R. J. 1988. Age and significance of sequence boundaries on Jurassic and Early Cretaceous

rifted continental margins. American Association of Petroleum Geologists Bulletin, 72, 49-72. , PAPE, J. & ROBERTS, D. G. 1985. Depositional sequence mapping to illustrate the evolution of a passive continental margin. In: BERG, O. R. & WOOLVERTON, D. (eds) Seismic Stratigraphy I1." an integrated approach to hydrocarbon exploration. AAPG Memoir, 39, 93-115. JANSA, L. F. & PE-PIPER, G. 1988. Middle Jurassic to early Cretaceous igneous rocks along eastern North American Continental Margin. American Association of Petroleum Geologists Bulletin, 72, 347-366. & WADE, J. A. 1975. Geology of the Continental margin off Nova Scotia and Newfoundland. In: Offshore Geology of Eastern Canada. Geological Survey of Canada Paper, 7430, 51-105. KARNER, G. D., DRISCOLL, N. W. & WEISSEL, J. K. 1993. Response of the lithosphere to in-plane force variations. Earth and Planetary Science Letters, 114, 397-416. KEEN, C. E. & BARRETI~.,D. L. 1981. Thinned and subsided continental crust on the rifted margin of eastern Canada: crustal structure, thermal evolution and subsidence history. Geophysical Journal of the Royal Astronomical Society, 65, 443465. - • DE VOOGD, B. 1988. The continent-ocean boundary at the rifted margin off eastern Canada: new results from deep seismic reflection studies. Tectonics, 7, 107-124. - - , BOUTILIER, R., DE VOOGD, B., MUDFORD, B. & ENACHESCU, M. E. 1987. Crustal geometry and extensional models for the Grand Banks, Eastern Canada: constraints from deep seismic reflection data. In: BEAUMONT, C. & TANKARD, A. J. (eds) Sedimentary Basins and Basin-forming Mechanisms. Canadian Society of Petroleum Geologists Memoirs, 12, 101-115. KENT, D. V. & GRADSTEIN, F. M. 1985, A Cretaceous and Jurassic geochronology. Geological Society of America Bulletin, 96, 1419-1427. KLITGORD, K. D. & SCHOUTEN, H. 1986. Plate kinematics of the central Atlantic. In: VOGT, P. R. & TUCHOLKE, B. E. (eds) The western north Atlantic Region. The Geology of North America, 17, Geological Society of America, 351-378. , HUTCHINSON, D. R. & SCHOUTEN, H. 1988. U.S. Atlantic continental margin, structural and tectonic framework. In: The geology of North America, The Atlantic Continental Margin, Geological Society of America, 1-2, 19-55. KUSZNIR, N. J. & EGAN, S. S. 1989. Simple-shear and pure-shear models of extensional sedimentary basin formation: application to the Jeanne d'Arc basin, Grand Banks of Newfoundland. In: TANKARD, A. J. & BALKWILL,H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins, AAPG Memoir, 46, 305-322. LAMBIASE, J. J. 1991. A model for tectonic control of lacustrine stratigraphic sequences in continental rift basins. In: KATZ, B. J. (ed.) Lacustrine basin exploration, AAPG Memoir, 50, 265-276. LEEDER, M. & GAWTHORPE, R. 1987. Sedimentary

EXTENSIONAL TECTONICS models for extensional tilt-block/half-graben basins, In: COWARD, M. P., DEWEY, J. F. & HANCOCK, P. L. (eds) Continental Extensional Tectonics, Geological Society, London, Special Publication, 28, 139-152. MAUFFRET, A. & MONTADERT, L. 1987. Rift tectonics on the passive continental margin off Galicia (Spain). Marine and Petroleum Geology, 4, 49-70. --, MOUGENOT,D., MILES, P. R. & MALOD, J. A. 1989. Results from multichannel reflection profiling of the Tagus Abyssal Plain (Portugal)-Comparison with the Canadian Margin. In: TANKARD, A. J. & BALKWILL, H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins. AAPG Memoir, 46, 379-394. McALPINE, K. D. 1991. Mesozoic stratigraphy, sedimentary evolution, and petroleum potential of the Jeanne d'Arc basin, Grand Banks of Newfoundland. Geological Survey of Canada Paper 1-50. MCCLAY, K. R. & SCOTT, A. D. 1991. Experimental models of hangingwall deformation in ramp-flat listric extensional fault systems. Tectonophysics, 85-96. McHONE, J. G. 1988. Tectonic and paleostress patterns of Mesozoic intrusions in eastern North America. In: MANSPEIZER, W. (ed.) TriassicJurassic Rifting: Continental Breakup and the origin of the Atlantic Ocean and passive margins. Developments in Geotectonics, 22, 607-620. MCPHERSON, J. G., SHANMUGAM,G. & MOIOLA, R. J. 1987. Fan-deltas and braid-deltas: varieties of coarse-grained deltas. Geological Society of America Bulletin, 99, 331-340. MEADOR, K. J. & AUSTIN, J. A. 1988. Seismic comparison of the early stratigraphic evolution of conjugate passive continental margins: the Newfoundland/Flemish basin and the eastern Iberian Abyssal plain south of Galicia Bank. In: BOILLOT, G., WINTERE, E. L. ET AL., Proceedings of the ODP, Scientific. Results, 103, CoUege Station, TX (Ocean Drilling Program), 777-786. & DEAN, D. F. 1988. Shelf to basin correlations in the Newfoundland Basin. In." BALLY, A. W. (ed.) Atlas of Seismic Stratigraphy. AAPG Studies in Geology, 27, 2, 88-95. MITCHUM, R. M. Jr. 1977. Seismic stratigraphy and global changes of sea level, Part 11 Glossary of terms used in seismic stratigraphy. In: CLAYTON, C. E. (ed.) Seismic Stratigraphy--Application to Hydrocarbon Exploration. AAPG Memoir 26, 205-212. MONTADERT, L., DE CHARPAL, O., ROBERTS, O., GUENNOC, P. & SIBUET, J.-C. 1979. Northeast Atlantic passive continental margins: rifting and subsidence processes. In: TALWANI, M., HAY, W. & RYAN, W. B. F. (eds) Deep drilling results in the Atlantic Ocean; continental margins and paleDenvironment. AGU, Maurice Ewing Series, 3, 154186. MORLEY, C. K., NELSON, R. A. PATTON, T. L. & MUNN, S. G. 1990. Transfer zones in the East African rift system and their relevance to hydrocarbon exploration in rifts. American Association 188,

,

-

-

27

of Petroleum Geologists Bulletin, 74, 1234-1253. PEMBERTON, S. G., BROMLEY,G. G. & EKDALE, A. A. 1984. Ichnology the use of trace fossils in sedimentology and stratigraphy. SEPM Shortcourse notes. POSAMENTIER, H. W., JERVEY, M. T. & VAIL, P. R. 1988. Eustatic controls on clastic deposition I. In: WILGUS, C. K., HASTINGS, B. S., KENDALL,C. G., POSAMENTIER, H. W., ROSS, C. A. & VAN WAGONER, J. C. (eds) Sea level changes: An integrated approach. SEPM Special Publication, 42, 109-124. ROYDEN, L. & KEEN, C. E. 1980. Rifting processes and thermal evolution of the continental margin of eastern Canada determined from subsidence curves. Earth and Planetary Science Letters, 51, 343-361. SRIVASTAVA, S. P., SCHOUTEN, H., ROEST, W. R., KLITGORD, K. D., KOVACS,L. C., VERHOEF,J. & MACNAB, R. 1990. Iberian plate kinematics: a jumping plate boundary between Eurasia and Africa. Nature, 344, 756-759. , ROEST, W. R., KOVACS, L. C., OAKEY, G., LEVESQUE, S., VERHOEF, J. 8/; MACNAB, R. 1990. Motion of Iberia since the late Jurassic: results from detailed aeromagnetic measurements in the Newfoundland basin. Tectonophysics, 184, 229260. & TAPSCOTT, C. R. 1986. Plate kinematics of the North Atlantic. In: VOGT, P. R. & TUCHOLKE, B. E. (eds) The geology of North America, Volume M: The western North Atlantic, Geological Society of America, 379-404. TANKARD, A. J. & WELSINK, H. J. 1987. Extensional tectonics and stratigraphy of Hibernia oil field, Grand Banks, Newfoundland. AAPG Bulletin, 71, 1210-1232. & JENKINS, W. A. M. 1989. Structural styles and stratigraphy of the Jeanne d'Arc basin, Grand Banks of Newfoundland. In: TANKARD, A. J. & BALKWlLL, H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins. AAPG Memoir, 46, 265-282. TODD, B. J., REID, I & KEEN, C. E. 1988. Crustal structure across the southwest Newfoundland Transform Margin. Canadian Journal of Earth Sciences, 25, 744-759. TUCHOLKE, B. E., AUSTIN,J. A. Jr. & UcHuPI, E. 1989. Crustal structure and rift-drift evolution of the Newfoundland Basin: In: TANKARD, A. J. & BALKWILL, H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins, AAPG Memoir, 46, 247-264. UNDERHILL, J. R. 1991. Controls on late Jurassic seismic sequences, Inner Moray Firth, UK North Sea: a critical test of a key section of Exxon's original global cycle chart. Basin Research, 3, 7998. VAIL, P. R. 1987. Seismic stratigraphy interpretation utilizing sequence stratigraphy. In: BALLY, A. W. (ed.) Atlas of Seismic Stratigraphy, AAPG Studies in Geology, 27, 1, 1-10. , MITCHUM, R. M. Jr., TODD, R. G., WIDMIER, J. M., THOMPSON, S. III, SANGREE, J. l., BUBB, -

-

,

-

-

28

N . W . DRISCOLL ET AL.

J. N. & HATELID, W. G. 1977, Seismic stratigraphy and global changes of sea level. In: CLAYTON, C. E. (ed.) Seismic Stratigraphy---Application to Hydrocarbon Exploration, AAPG Memoir, 26, 49-212. VAN WAGONER, J. C., POSAMENTIER,H. W., MrrcHUM, R. M. Jr., VAIL, P. R., SARG, J. F. LOUTIT, T. S. & HARDENBOL,J. 1988. An overview of the fundamentals of sequence stratigraphy and key definitions. In: WILGUS, C. K., HASTINGS, B. S., KENDALL, C. G., POSAMENTIER, H. W., ROSS, C. A. & VAN WAGONER, J. C. (eds) Sea Level changes: An integrated approach. SEPM Special Publication, 42, 39-45. VERHOEF, J. • SRIVASTAVA,S. P. 1989. Correlation of sedimentary basins across the North Atlantic as obtained from gravity and magnetic data, and its relation to the early evolution of the North Atlantic. In: TANKARD, A.J. & BALKWILL,H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins, AAPG Memoir, 46, 131148. WELSINK, H. J. & TANKARD, A. J. 1988. Structural and stratigraphic framework of the Jeanne d'Arc

basin, Grandbanks. In: BALLY, A. W. (ed.) Atlas of Seismic Stratigraphy. AAPG Studies in Geology, 27, 2, 1 4 - 2 2 . - - - , SRIVASTAVA, S. P. & TANKARD, A. J. 1989. Basin architecture of the Newfoundland Margin and its relationship to 9ocean crust fabric during extension. In: TANKARD, A. J. &; BALKWILL, H. R. (r Extensional Tectonics and Stratigraphy of the North Atlantic Margins. AAPG Memoir, 46, 197-214. WHITE, N. J., JACKSON, J. A. & MCKENZIE, D. P. 1986. The relationship between the geometry of normal faults and that of the sedimentary layers in their hanging walls. Journal Structural Geology, 8, 897-909. WHITMARSH, R. B., MILES, P. R. & MAUFFRET, A. 1990. The ocean-continent boundary off the western continental margin of Ibeda - - I. Crustal structure at 40~ 30'N. Geophysical Journal International, 103, 509-531. ZIEGLER, P. A. 1989. Evolution of the North Atlantic--An overview. In: TANKARD, A. J. & BALKWILL, H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins, AAPG Memoir 46, 111-130.

Sequence stratigraphic response to Aptian-Albian rifting in conjugate margin basins: a comparison of the Jeanne d'Arc Basin, offshore Newfoundland, and the Porcupine Basin, offshore Ireland I A I N K. S I N C L A I R

University o f Aberdeen~Canada-Newfoundland Offshore Petroleum Board 5th floor TD Place, 140 Water Street, St John's, Newfoundland, Canada' A I C 6H6 Abstract. Analysis of well and seismic data indicates that both the Jeanne d'Arc Basin on the Grand Banks of Newfoundland and the Porcupine Basin on the continental shelf off western Ireland were undergoing extension characterized by subsidence and rotation of fault blocks during mid-Aptian to late Albian times. Stratigraphic sequences in these conjugate basins developed in response to synchronous rift tectonism. The base of each syn-rift sequence is defined by an angular unconformity which formed in response to regional uplift. This mid-Aptian rift-onset unconformity is buried beneath a depositional sequence characterized by retrogradational coastal sediments overlain by progradational coastal sediments in each basin. However, the lower retrogradational deposits dominate the syn-rift sequence in the Jeanne d'Arc Basin, while the upper progradational deposits dominate the syn-rift sequence in the northern Porcupine Basin. The relative dominance of sediments deposited along either transgressive or regressive coastlines was mainly determined by the local interplay of variable rates of rift-induced subsidence and sediment input. The initiation of seafloor spreading between the Flemish Cap/Orphan Knoll and Goban Spur/Porcupine Bank continental margins near the end of Albian time resulted in the establishment of regional thermal subsidence, rapid coastal transgression and fully marine deposition above a locally-developed sequence-bounding unconformity.

Stratigraphy and subsidence rates and styles have previously been compared to illustrate relationships between the opening of multiple segments of the North Atlantic Ocean and the formation of various Mesozoic rift basins on flanking continental margins (e.g. Pegrum & Mounteney 1978; Masson & Miles 1986; Ziegler 1988; H i s c o t t et al. 1990a). Deep seismic reflection profiles have also been used to compare structural styles across pairs of North Atlantic conjugate margin areas, such as the Flemish Cap and Goban Spur margins (Keen et al. 1989, fig. 1). This paper will focus on the midAptian to late Albian rift period that led to break-up and initiation of seafloor spreading between the Flemish Cap/Orphan Knoll and Goban Spur/Porcupine Bank conjugate margins. The purpose is to identify and compare stratigraphic responses to that rift period (i.e. depositional facies, facies stacking patterns, and stratigraphic sequence and parasequence boundaries) in the Jeanne d'Arc Basin on the Grand Banks of Newfoundland with those of the Porcupine Basin on the continental shelf off western Ireland (Fig. 1). Numerous oil and gas pools have been discovered in structured Aptian-Albian age sandstones in the Jeanne d'Arc Basin and oil shows have been encountered in

equivalent age sandstones in the Porcupine Basin. Definition of the relationships between mid-Cretaceous structuring and sedimentation are of potential value in field development planning, prospect generation and plate kinematic studies. The term 'sequence' is applied to genetically related packages of sediments bound by unconformities and their correlative conformities as defined by Mitchum (1977) and others. However, the development, extent and expression of the sequence boundaries studied herein are interpreted as having been controlled by large-scale, episodic tectonism. Therefore, the mid-Aptian to Upper Albian sequences of the Jeanne d'Arc and Porcupine Basins also correspond to the megasequence concept of Hubbard (1988).

Shared tectonic history Figure 2 illustrates the Mesozoic and Cenozoic lithostratigraphic units of the Jeanne d'Arc and Porcupine basins and an interpretation of three tectonic episodes which synchronously affected both basins. This interpretation is based on a combination of the author's own study of the two basins and published research (e.g. Jansa &

From Scrutton, R. A., Stoker, M. S., Shimmield, G. B. & Tudhope, A. W. (eds), 1995, The Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region, Geological Society Special Publication No. 90, pp. 29-49

29

30

I.K. SINCLAIR

Fig. 1. Earliest Cretaceous plate reconstruction modified from Masson & Miles (1986) showing the conjugate margin basins under comparison. FC, Flemish Cap; GS, Goban Spur; JD, Jeanne d'Arc Basin; OK, Orphan Knoll; PB, Porcupine Bank. Wade 1975; Wade 1978; Hubbard et al. 1985; Enachescu 1987; Hubbard 1988; Sinclair 1988; Tankard & Welsink 1988; Tankard et al. 1989; McAlpine 1990; Naylor & Shannon 1982; Croker & Shannon 1987; MacDonald et al. 1987; Tate & Dobson 1989; Croker & Klemperer 1989; Shannon 1991; Moore 1992). A comparison chart and brief discussion of published tectonic interpretations for the Jeanne d'Arc Basin is included in Sinclair (1993). The Mesozoic rift episode which initiated formation of the Jeanne d'Arc and Porcupine basins occurred during the Late Triassic into the earliest Jurassic. Syn-rift sedimentation during this initial extensional phase followed a common pattern in the two basins. Continental siliciclastic sediments were deposited in geographically isolated but tectonically linked NE-SW-trending

rift basins (Jansa & Wade 1975; Wade 1978; Croker & Shannon 1987; Ainsworth 1990). This was followed by precipitation of variable thicknesses of evaporitic minerals and finally b y deposition of marine carbonate beds (Fig. 2). Lithofacies deposited during mid-Jurassic times in the areas of the Jeanne d'Arc and Porcupine Basins are quite different from each other (Fig. 2) but these sediments share one trait. Mid-Jurassic sedimentation in both areas occurred beyond the limits of the earlier faultbound rift basins synchronous with a termination of significant basement faulting. This expansion of basins is interpreted to have been a response to the establishment of regional thermal subsidence (Hubbard et al. 1985; MacDonald et al. 1987). Initially broad uplift of the western Irish Shelf area following the first

APTIAN-ALBIAN RIFTING OF CONJUGATE ATLANTIC MARGINS

31

Fig. 2. Comparative lithostratigraphic chart for the Jeanne d'Arc and Porcupine Basins. Also shown are three successive and discrete rift stages (i.e. R1-R3) which caused synchronous structuring of basement and overlying sediments in both basins. Three lithostratigraphic units, the tops of which provide high-amplitude seismic reflections in the Jeanne d'Arc Basin are identified (i.e. Petrel Member, 'A' Marker member and 'B' Marker member). Hibernia Formation sandstones (H.) and Fortune Bay Formation shales (F.B.) are also identified. More complete lithostratigraphic nomenclature schemes for the Jeanne d'Arc Basin are provided in Sinclair (1988) and McAlpine (1990).

32

I.K. SINCLAIR

-48" 10~

-48'25'

-48"40'

-48"55'

-49' 10'

47"20',

-47"55'

47"20'

47"10'

~ 47"10'

46"55'

6"55'

Hibernia Bonavista Platform

N. Ben Nevis

A

46'40' Fig.

-

48"2~'

4-+

,

e . 4r.~

Fortune G-57

,~

46"2.=

/ 0

10

I

46"I0' I -49"I0'

-48"55'

-48"40'

-48:25'

I

-48"I0'

20 km. I

, -47"55'

46"11

Fig. 3. Jeanne d'Arc Basin isopach contour map of the mid-Aptian to Upper Albian syn-rift sequence (contours in hundreds of metres), modified from Sinclair (1993). Also shown are the locations of Figs 4, 7 & 10.

split of the Pangean supercontinent (i.e. between Africa and N. America) may account for both the extensive hiatus at the first Early Jurassic break-up unconformity (sensu Falvey 1974) as well as the continental character of overlying mid-Jurassic sediments in the Porcupine Basin

area. However, a number of authors have considered the structured and dominantly continental character of mid-Jurassic sediments in the Porcupine Basin to be evidence of a period of E-W-oriented rift extension (e.g. Croker & Klempcrer 1989; Shannon 1991).

APTIAN-ALBIAN RIFTING OF CONJUGATE ATLANTIC MARGINS Late Cimmerian rifting (Ziegler 1988) affected a wide area, spanning the Jeanne d'Arc and Porcupine Basins, beginning about the start of Tithonian times (Fig. 2; Sinclair et al. 1994). This rift period was characterized by rotation of N-S-trending fault blocks and accumulation of thick, coarse-grained siliciclastic sediments over wide areas of the Jeanne d'Arc Basin (Hubbard et al. 1985; Enachescu 1987) and locally over individual graben and half-graben blocks in the Porcupine Basin (Croker & Shannon 1987; MacDonald et al. 1987). The Late Cimmerian influx of siliciclastic sediments lasted into Early Valanginian times in the Jeanne d'Arc Basin (Sinclair 1988). Moore (1992) indicates that fanstyle deposition occurred over restricted areas of the Porcupine Basin synchronous with fault activity from latest Kimmeridgian/earliest Tithonian through Berriasian times, though he characterized this as a transitional stage between syn-rift and post-rift periods. A subsequent period of passive thermal sag began during the Valanginian in both basin areas. As a result Valanginian to Barremian marine carbonates and shales onlapped and blanketed the underlying structured strata (Fig. 2; Sinclair 1988; Moore 1992; Sinclair et al. 1994). Though subsidence in the Jeanne d'Arc and Porcupine Basins during the m i d - A p t i a n through Albian period has commonly been considered as passive (i.e. post-rift), active basement-involved faulting in both areas has been reported for this period (Hubbard et al. 1985; Sinclair 1988, 1993; Moore 1992; Foster & Robinson 1993). Furthermore, recognition of a mid-Aptian to Albian tectonic influence in the Porcupine Basin appears to be increasing. For example, the progradation of thick coastal sediments during the latter part of this depositional period has recently been attributed to a minor rift episode (Shannon 1991; Shannon et al. 1993). The Porcupine Basin area was locally affected by extension during the Senonian and by magmatism and inversion during Palaeocene to mid-Eocene time. These developments are considered to be in response to a combination of mechanisms including the Norwegian-Greenland Seas rift and Alpine compression (Tate & Dobson 1988; Ziegler 1988; Croker & Klemperer 1989; Moore & Shannon 1992). In contrast, the Jeanne d'Arc Basin was isolated from these stresses by extrusion of basaltic crust of the intervening North Atlantic Ocean and subsided passively from Cenomanian times t o the present.

-- 46*

33

._

50' N

Bonavista Platform (non-extending

'(L ~ ' ~ . - ~ ~ ' ~ ' - ~ i ~ ' ~ O* /I ~'~"~'~V~-'-'~'-_._ ~ = ~ . =

O

~

Skin

X RankinM-36~,~ I "'% 480 50' w

I

I 48Q 40' w

Fig. 4. Map of fault trends near the mid-Aptian unconformity over the Hibernia structure area. Intrafield faults are taken from a map of Mobil Oil Canada and partners as reproduced in C-NOPB (1986). Locations of seismic and wireline log sections (Figs 5 & 6, respectively) are also shown.

Jeanne d'Arc Basin Seismic, wireline log and core data are used in the three following sections to illustrate: (1) synrift subsidence during mid-Aptian to Late Albian times; (2) the location and character of the rift-onset unconformity bounding the sequence base and facies stacking patterns of immediately overlying sediments; and (3) facies stacking patterns of the sequence's uppermost sediments and the location and character of the break-up unconformity which marks the top of the syn-rift sequence. Rift-induced subsidence

~An isopach contour map of the mid-Aptian to Upper Albian sequence (Fig. 3) illustrates widespread and marked structuring of the Jeanne d'Arc Basin during this period. The Hibernia field, largest discovery in the basin to date (Arthur et al. 1982; Hurley et al. 1992), has sufficiently dense seismic and well coverage to reasonably constrain the age of this rift activity. The Hibernia structure is cut by numerous NW-SE-trending normal faults which formed in response to northeast-directed extension (Sinclair 1988). Two major extensional faults,

34

I.K. SINCLAIR

Fig. 5. Seismic section 83-2560 (PAREX 1983) and interpretation across the Hibernia structure. Sequence boundaries: 1, base Tertiary unconformity; 2, latest Albian to Cenomanian correlative conformity; 3, mid'-Aptian unconformity. Section location is given on Fig. 4. Vertical scale is two-way time. Nautilus and Rankin, bound the structure to the northeast and southwest (Fig. 4). Oblique-slip motion occurred along the western faulted margin of the Hibernia structure (Tankard & Welsink 1987) synchronously with, and in response to, the northeasterly-directed extension. Seismic data (Fig. 5, also fig. 8 of Hurley et al. 1992) show the mid-Aptian/Upper Albian sedimentary sequence to thicken southward in a step-like fashion toward the Rankin fault. One feature of particular interest is the B-27 graben located near the crest of the Hibernia structure (Fig. 4). Figure 5 demonstrates that this graben formed during the final fault-controlled growth of the Hibernia structure prior to the regional subsidence which characterized the Jeanne d'Arc Basin area during Late Cretaceous and Tertiary times. Correlation of delineation wells of the Hibernia oilfield illustrates the effects that seismicallyidentified faulting had on sedimentary sequences. Significant thickness variations are seen in the sedimentary sequence containing the oilbearing reservoir denoted 'R' and overlying

Upper Aptian to Upper Albian shales (Fig. 6), demonstrating their deposition during extensional fault growth. Wireline log correlations indicate that the NE-SW-oriented fault growth ended prior to deposition of deep marine marlstones and Petrel Member limestones which occur above the correlative conformity to a latest Albian (or Cenomanian) age unconformity. In fact, the Petrel Member increases in thickness from the B-27 graben location to the K-14 location, opposite to the trend seen in the underlying mid-Cretaceous sequence. This increasing thickness of the Petrel Member away from the Hibernia structure crest is attributed to regional draping of sediments across the existing structure rather than continued structure growth during the Late Cretaceous. However, the age of the base of the midCretaceous growth sequence over the Hibernia structure (i.e. unit 'R') is somewhat problematic, being either Barremian or Aptian (Hurley et al. 1992). An unconformity and subaerially deposited sediments are recognized in core taken across the base of unit 'R'. This unconformity

A P T I A N - A L B I A N R I F T I N G OF CONJUGATE ATLANTIC MARGINS

35

Fig. 6. Gamma ray (left) and sonic (right) logs correlation through the Hibernia K-14, B-27 and K-18 wells (location on Fig. 4). 'R' is used here to identify a retrogradational mid-Cretaceous sandstone which provides an oil-bearing reservoir across much of the Hibernia structure. Sequence boundaries 2 and 3 as for Fig. 5.

Fig. 7. Correlation of the basal sediments of the mid-Aptian to Upper Albian sequence between Hibernia B-27 and North Ben Nevis M-61 using gamma ray logs and facies interpretations of slabbed cores. The locations of core photos in Figs 8 & 9 are shown and well locations are provided on Fig. 3. Upright and inverted triangles represent transgressive and regressive half cycles, respectively.

36

I.K. SINCLAIR

has been correlated, herein, to a basin-wide mid-Aptian unconformity using seismic data and depositional facies analyses (see subsequent section). At odds with this correlation is the relatively common, though not uniform, occurrence of microfossils with Barremian to Early Aptian extinctions above the base 'R' unconformity in Hibernia field wells (e.g. Thompson 1987). Therefore, the base 'R' boundary is either a Barremian unconformity distinct from the mid-Aptian unconformity and seismic data are inadequate to distinguish the two or the microfossil record is locally unrepresentative. While the first possibility cannot be eliminated, the second possibility is preferred here as examples of microfossil reworking across the mid-Aptian unconformity have been identified in basin centre wells using petroleum industry biostratigraphic reports (Sinclair 1993). The difficulties inherent in microfossil dating of well-bore sediments overlying an angular unconformity dictate continuing uncertainty in age determination of basal sediments of the syn-rift sequence in a number of wells. Nevertheless, rift-induced growth of NW-SE-oriented normal faults is observed to have affected sedimentation from at least mid-Aptian times into the late Albian (Figs 5&6). Basal facies and sequence boundary Figure 7 illustrates a correlation of sediment facies adjacent to the mid-Aptian unconformity from Hibernia B-27 to North Ben Nevis M-61. Depositional facies shown for the B-27 and M61 wells are mostly consistent with interpretations of Hurley et al. (1992) and Harding (1988), respectively. Correlation of the mid-Aptian unconformity and the overlying marine ravinement, however, correspond to the interpretations of Charl du Toit (Chevron Canada, pers. comm. 1989) and Mobil Oil et al. (1988) for B-27 and Sinclair (1993) for M-61. Core photographs (Figs 8 & 9) are used to illustrate some of the depositional facies and facies contacts in sediments overlying the interpreted mid-Aptian unconformity. The mid-Aptian unconformity was cored in Hibernia B-27 where grey carbonaceous sandstones, containing plastically deformed sandstone clasts and layers, sit with angularity above relatively clean, oil-soaked, highly burrowed sandstones (Fig. 8a). The grey sandstones are interpreted to represent estuarine channel fill with abundant terrestrial plant debris and semiconsolidated clasts eroded from the channel banks. The underlying oil-bearing sandstones are considered to be remnants of partially

eroded regressive shoreface parasequence sets. Similar shoreface sandstones are more extensively preserved basinward, as at the North Ben Nevis M-61 wellsite (Fig. 7; Harding 1988). It is only through the incorporation of regional data sets, however, that the angular facies contact seen in Fig. 8a can be demonstrated to be a sequence-bounding unconformity. Regional seismic surveys have been used to correlate reflections from the Hibernia and North Ben Nevis structures to the eastern margin of the basin where gently dipping strata above the mid-Aptian unconformity onlap more steeply dipping strata below. Seismic data also show that progressively greater truncation of pre-midAptian unconformity sediments occurred on successively higher fault blocks on the basin's east flank (Fig. 10). Wireline logs from exploratory wells drilled on adjacent fault blocks in the same area provide the data necessary to establish the stratigraphic position and extent of erosion associated with the mid-Aptian unconformity. About 160m of section were eroded at the Archer K-19 wellsite relative to the nearby Amethyst F-20 wellsite (Fig. 11). This erosion is in addition to 490 m eroded at Amethyst F-20 relative to the Fortune G-57 location o n an adjacent fault block to the west (Figs 3 & 10). The estuarine sandstones of Hibernia B-27 are overlain by oxidized and rooted marsh shales (Charl du Toit, pers. comm. 1989; Hurley et al. 1992) providing clear evidence of subaerial deposition above the base 'R' unconformity. The red shales grade laterally and vertically to dark grey shales containing abundant mollusc shells (Fig. 9a). These invertebrate fossil assemblages are either in situ or transported only very short distances into low-energy, subaqueous (i.e. lagoonal) settings. Harding (1988) identified many genera of the diverse assemblage of gastropod and pelecypod fossils present in the dark grey and carbonaceous lagoonal shales cored in North Ben Nevis M-61. Conical, turreted gastropod shells (Fig. 9b) are among the fossils prevalent in these lagoonal shales. A high-energy lag deposit containing abundant bioclasts cannibalized from the underlying lagoonal shales is preserved at the contact between back-barrier facies and overlying marine sandstones in core from North Ben Nevis M-61 (Fig. 9c). The erosive contact separating marine and back-barrier facies is interpreted to be a marine ravinement diastem (sensu Nummedal & Swift 1987) formed by scouring at the high energy foreshore and upper shoreface environments and at the base of migrating tidal inlet channels. The ravinement and associated lag deposit have also been cored in a number of

37

Fig. 8. Three types of erosive contacts separating different sedimentary facies. (The coin diameter is 2 cm.) (a) Grey, carbonaceous estuarine channel fill overlying eroded shoreface sandstones. The angular facies contact is interpreted to correspond to the regionally-developed mid-Aptian unconformity. Hibernia B-27, 2598.2 m drill depth =2600.2 m log depth. (b) Gastropod-rich shell lag erosively overlying burrowed sandstones. The angular facies contact represents a time-transgressive marine ravinement diastem (sensu Nummedal & Swift 1987) separating marine and back-barrier environments. Hibernia B-27, 2582.2 m drill depth -2584.3 m log depth. (c) Storm sandstones containing transported shell debris and rounded intra-clasts of the underlying argillaceous fair weather beds. The sharp angular facies contact is a local storm-induced scour within a shoreface-to-offshore transition setting. Hibernia B-27, 2552.6m drill depth -2555.1 m log depth.

Fig. 9. (The coin diameter is 2cm). (a) Dark grey, bioclastic-rich lagoonal shale. Hibernia K-14, 2405.3m drill depth =2407.8 m log depth. (b) Plane view of conical, turreted gastropod commonly found in mid-Cretaceous lagoonal shales of the Jeanne d'Arc Basin. Hibernia K-14, 2414m drill depth =2416.5m log depth. (e) Highenergy lag (20 cm thick) comprising abundant bioclasts and a large siliceous cobblein a fine-grained sandstone matrix deposited above black, carbonaceous, bioclastic lagoonal shales. North Ben Nevis M-61, ravinement at 11.5 cm on rule, 3051.03 m drill depth =3048 m log depth. (d) Amalgamated cycles of fining-upward, fine-grained sandstone beds with basal shell lags deposited on a storm-dominated shoreface. North Ben Nevis M-61, 3048.793049.09 m drill depth =3046-3046.3 m log depth.

APTIAN-ALBIAN RIFTING OF CONJUGATE ATLANTIC MARGINS

39

Fig. 10. Seismic section 83-5040 (PAREX 1983) and interpretation across the eastern margin of the Jeanne d'Arc Basin. Sequence boundaries: 1, base Tertiary unconformity; 2A and 2B, latest Albian to Cenomanian unconformity and correlative conformity, respectively; 3, mid-Aptian unconformity; 4, mid-Valanginian conformity at base of 'B' Marker member; 5, Tithonian unconformity. Location given on Fig. 3. other wells including South Mara C-13 (Sinclair 1993) and Hibernia B-27. Abundant bioclasts, dominated by conical gastropods, are set in a fine-grained sandstone matrix above heavily burrowed sandstones in Hibernia B-27 (Fig. 8b). The erosional ravinement extends from the pen tip to c. 4 cm on the right side of Fig. 8b. The underlying burrowed sandstones, while not assigned to a specific depositional setting, are associated with identifiable back-barrier facies. Numerous amalgamated cycles of clean, finegrained sandstones with basal shell lags (Fig. 9d) are preserved above the marine ravinement and associated lag deposit of North Ben Nevis M-61 (Fig. 9c). The amalgamated cycles are interpreted as having been deposited on the shoreface of a barrier island/attached beach chain during periodic storms. The storms eroded sediments deposited on the middle shoreface during fair weather conditions, thereby preserving only stacked sandstones deposited during waning storm energy. Hiscott et al. (1990b) point to

evidence that winter storms and hurricane-force events had a strong.influence on marine deposition in the Jeanne d'Arc Basin during much of the Late Jurassic and Early Cretaceous. The amalgamated sandstones grade upsection into numerous thin interbeds of: heavily bioturbated, very argillaceous sandstones with little or no shell debris; alternating with clean, porous sandstones containing lags of relatively fine shell debris and rounded intra-clasts torn from the underlying burrowed sandstones. Sharp contacts at the base of the porous sandstone beds (Fig. 8c) are interpreted to represent basal scours of storm beds above partially preserved fair weather strata of the shoreface-to-offshore transition zone, as previously described in Hibernia area wells by Hurley et al. (1992). The interbedded high- and low-energy deposits grade upward into dominantly dark grey, variably silty shales encasing rare, thin sandstone and shell-rich layers. These are interpreted to represent an offshore setting below the storm

40

I . K . SINCLAIR

Fig. 11. Gamma ray (left) and sonic (right) logs cross-section using two datums, indicating erosion of 160 m of strata at the Archer K-19 wellsite relative to the Amethyst F-20 wellsite. MFS, maximum flooding surface at a condensed horizon. Sequence boundaries as for Fig. 10.

APTIAN-ALBIAN RIFTING OF CONJUGATE ATLANTIC MARGINS

13 ~ w ---~-

41

12 ~ w

50 Km

-53 ~ N 26/21-1, - ~ . . . ~ " " ' ~ ~ i

i

MPB

o 26/27

J FIG. 14

26/29-1

!/,-/

l

..............

/~.

/ 35/2-1

4 lO Km

I Fig. 12. Map of northern Porcupine area fault trends at the base of Upper Cretaceous chalk-dominated sediments. Abbreviations are: MPB, Main Porcupine Basin; NPB, Northern Porcupine Basin; ST, Slyne Trough. Basin outline simplified from map in Philips Petroleum's well history report for 35/8-2. Locations given for Figs 13-15. wave base with rare storm-induced turbidite deposition (Sinclair 1993). The bulk of midCretaceous strata occurring above the midAptian unconformity in the Jeanne d'Arc Basin were deposited under increasingly seaward palaeosettings.

Upper facies and sequence boundary The top of the mid-Aptian to Upper Albian sedimentary sequence penetrated by Amethyst F-20 and Archer K-19 displays a brief reversal of the dominant fining-upsection trend. This upward gradation from shale to siltstone, paralleling a decrease in gamma activity (Fig. 11), is interpreted to represent a minor and local phase of coastal regression in the Jeanne d'Arc Basin occurring towards the end of rift-induced faulting.

Peak levels of gamma activity occur over the offshore shale strata between the fining-upward and the coarsening-upward sediments of Amethyst F-20. These shales are interpreted to correspond to a maximum flooding surface (MFS) separating the lower, dominant retrogradational parasequence set from the upper subordinate progradational parasequence set (sensu Van Wagoner et al. 1988). There is little evidence of significant erosion at the overlying Cenomanian sequence-bounding unconformity (Fig. 11) which quickly passes basinward to a correlative conformity (Fig. 10). The local regression seen in the Amethyst F-20 area was followed by regional post-rift subsidence and flooding well beyond the rift basin's fault-limited margins, beginning in latest Albian or early Cenomanian times (Hubbard et al. 1985; Sinclair 1988).

42

I.K. SINCLAIR

Fig. 13. Seismic section BP Tie 1 and interpretation through the 26/28-1 discovery well of the Connemara field. Sequence-bounding unconformities: 1 + 2, base Tertiary and base Upper Cretaceous (too closely spaced to resolve at this scale); 3, mid-Aptian; 4, (mid ?) Valanginian; 5, Tithonian; 6, mid-Jurassic. SS sheet, upper sandstone sheet of the mid-Aptian to Albian sequence. Location shown on Fig. 12. This seismic section may obliquely cross a poorly imaged fault offsetting the base Tertiary near shotpoint 580 (see Croker & Klemperer 1989, fig. 5).

Porcupine Basin Seismic, wireline log and well cuttings data from a portion of the North Porcupine and Main Porcupine Basins (Fig. 12) are used in the following three sections to illustrate the sequence stratigraphic response to mid-Aptian/Late A1bian rifting on the continental shelf west of Ireland.

Rift-induced subsidence Discovery of oil in a tilted fault block on BP's Block 26/28 licence (MacDonald et al. 1987) led to the drilling of four delineation wells. This locally dense well control, tied to more widely spaced wells using a regional seismic grid, is sufficient to evaluate the structural history of the northern Porcupine Basin. The main oil-bearing reservoir of the 26/28 structure (i.e. Connemara

field; Murphy & Croker 1992) occurs within the widespread mid-Jurassic continental sandstones. A secondary reservoir flowed oil from the thin updip edge of a Tithonian sandstone wedge. Seismic data indicate that an eastward thickening wedge of Tithonian sediments was deposited during initial structuring of the Connemara area (i.e. between unconformities 4 and 5; Fig. 13). Figure 13 also illustrates marked variations in thickness of the mid-Aptian to Upper Albian sedimentary sequence (i.e. between unconformities 2 and 3) across the eastern faulted margin of the Connemara field. Additionally, correlations of wireline logs, well cuttings descriptions and biostratigraphic reports demonstrate that a significant increase in subsidence rates occurred over a portion of the Porcupine Basin following a period of slow onlap and burial of the Connemara area during Valanginian to Barremian times (Fig. 14; also fig. 4 subsidence curves

A P T I A N - A L B I A N R I F T I N G OF CONJUGATE ATLANTIC M A R G I N S

43

Fig. 14. Correlation of gamma ray (left) and sonic (right) logs through five wells on the eastern margin of the Porcupine Basin. Location on Fig. 12. Unconformity codes as for Fig. 13 but unconformities 1 and 2 are seen to be separated by a thin interval of Upper Cretaceous sandy limestones. No open-hole logs were run over this relatively shallow section of well 26/28-4.

Fig. 15. Seismic section BP Tie 3 and interpretation through 26/28-2 and 26/28-5 delineation wells. Location shown on Fig. 12 and unconformity codes as on Fig. 13.

44

I.K. SINCLAIR

Fig. 16. Correlation of the sedimentary facies at the base of the mid-Aptian to upper Albian sequence between 35/ 2-1 and 26/28-2 based on gamma ray (left) and sonic (right) log trends and cuttings descriptions. Depositional facies symbols as for Fig. 7. Well locations provided on Fig. 12. of Hiscott et al. 1990a). These patterns indicate that a separate period of rift-driven subsidence and extensional fault growth occurred during the mid-Aptian to Albian, millions of years after the Tithonian (to Berriasian?) period of structuring and coarse siliciclastic fan deposition. The fault pattern also changed from one rift period to the next, as the thickest Tithonian fan sequence has been penetrated by well 26/28-4 located on the high margin east of the Connemara field (Fig. 14). Therefore, the area of greatest Tithonian subsidence became part of a footwall block during mid-Aptian to Late Albian structuring. Unfortunately, the location of the fault which controlled formation of the Tithonian wedge in Block 26/28 is not clearly imaged on seismic data (Fig. 13).

Basal facies and sequence boundary Both well and seismic data indicate that the base of the mid-Aptian to upper Albian sequence of the Porcupine Basin is an intra-Cretaceous unconformity, separate and distinct from the more commonly recognized 'Base Cretaceous' unconformity (unconformities 3 and 4, respectively, on Figs 13-15). The two unconformities may locally converge on high blocks, however (i.e. area of wells 26/28-4 and 26/29-1; Fig. 14). BP's composite log for well 26/28-5 indicates that upper Aptian sediments sit non-sequentially

above Hauterivian and older marine shales (Fig. 14). This hiatus occurs at a significant bedding dip change such that beds dipping 4-6 ~ southwest overlie beds dipping 10-15 ~ southeast. This angular contact is, therefore, interpreted to correspond to the regional mid-Aptian unconformity reported by Croker & Shannon (1987). Additionally, this mid-Aptian sequence boundary ties eastward from the 26/28-5 wellsite to a high-amplitude dipping seismic reflector (c. 1.7 s on Fig. 15) identified as an Early Cretaceous unconformity by MacDonald et al. (1987). The depositional facies of sediments immediately overlying the mid-Aptian unconformity have been interpreted in the following two paragraphs on the basis of wireline log variations and the well operators' cuttings descriptions. Interpretation is limited to such a minimal data set as this horizon has not been cored, at least in the study area wells. The gamma ray curve from 26/28-2 displays a serrated pattern of gamma activity increasing upsection across an interval of very fine-grained, variably calcite-cemented sandstones and sandy limestones (Fig. 16). These patterns are attributed to episodic deposition of shoreface sandstones beneath transgressing seas, generating an irregular fining-upward trend. A highly radioactive shale layer, described as containing lignite fragments, occurs below similar calcareous and glauconitic sandstones in the

APTIAN-ALBIAN RIFTING OF CONJUGATE ATLANTIC MARGINS 35/2-1 well (Fig. 16). The carbonaceous content of the shales is consistent with deposition in a lagoonal setting, landward of the retrogradational shoreface sandstones and offshore shales. This interpreted pattern of retrogradational deposition is very similar to that recognized in core from the Jeanne d'Arc Basin (Figs 7-9), though it commonly occurs over a much thicker interval in the Jeanne d'Arc Basin (e.g. Fig. 11).

Upper fac&s and sequence boundary One of the most prominent features seen on seismic data from the northern Porcupine Basin is a set of clinoform reflectors which migrate southward within the mid-Aptian to late Albian sequence (Fig. 13). Well data (Fig. 14) and Croker & Klemperer (1989) show these clinoform reflectors to originate within open marine to pro-delta shales of Late Aptian to Albian age below an extensive layer of sandstones. A short core was recovered from the overlying sandstone sheet in well 26/28-2 (Fig. 14; 1188.31189.8 m). BP's well history report indicates that this cored interval comprises mollusc shellbearing, variably calcite-cemented, mediumgrained sandstones deposited within a series of prograding sub-tidal sand bars. These shoreface sandstones grade northward to delta plain facies overlying similar delta front sandstones in well 26/21-1 (Croker & Klemperer 1989). Gamma ray (Fig. 14) and lithology logs indicate that the contact between pro-delta shales and delta front sandstones is gradational, despite termination of clinoform reflectors at this facies contact. The apparent toplap of seismic reflectors is herein attributed to asymptotic flattening of bedding planes within the coastal sandstone sheet rather than to erosional truncation or non-deposition of sedimentary strata at the offshore/shoreface boundary. Flattening of bedding planes can cause seismic reflections to fall below the level of resolution within a progradational sheet of shoreface to delta plain sandstones, despite individual reflectors being discernible within a thick, ageequivalent interval of sloping offshore shales. The sloping shales and sub-horizontal sandstone sheet are, therefore, considered to be related lithofacies of a progradational parasequence set. The upper and lower boundaries to this parasequence set are described below. The pro-delta clinoforms appear to downlap on to a condensed section or MFS (Posamentier et al. 1988; Loutit et al. 1988) located within shales c. 150 ms or 200 m above the mid-Aptian unconformity at the Connemara discovery well (Figs 13 & 14). The interpreted MFS is

45

considered to separate the lower retrogradational parasequence set from the upper dominant progradational parasequence set within the main area of basin subsidence. Progradational pro-delta shales encountered on the eastern basin margin (e.g. in 26/29-1; Fig. 14) likely downlap directly on to the mid-Aptian sequence boundary as no basal retrogradational sediments are evident in this area. However, thin transgressive lag deposits may be locally preserved on this eastern flank between the midAptian unconformity and overlying regressive strata. There is little evidence that significant erosion occurred at the top of the Albian sandstone sheet, though non-deposition probably occurred in areas bypassed by sediments during regression of the coastline. The upper surface of the sandstone sheet was flooded by transgressing seas during the Late Cretaceous resulting in widespread limestone deposition. Flooding of the sandstone sheet top coincided with the end of differential subsidence between the faultcontrolled Connemara structure and the flanking margin area to the east (Fig. 4). The onset of regional subsidence even caused Upper Cretaceous chalks to overstep Carboniferous strata on previously uplifted basin margins, as seen in well 36/16-1 (Croker & Shannon 1987; Croker & Klemperer 1989). Thickness variations in Upper Cretaceous chalks indicate growth of the main east-west fault separating the North and Main Porcupine Basins (Fig. 12) during Senonian time. This Senonian subsidence pattern may be indicative of the onset of extension related to the Norwegian-Greenland Seas rift.

Discussion and conclusions Seismic and well data demonstrate that the Jeanne d'Arc and Porcupine Basins experienced rift-induced subsidence and fault-block rotation during mid-Aptian to Late Albian times. The sediments deposited during structuring form an unconformity-bound depositional sequence within each basin. The resultant sedimentary stacking pattern may also be referred to as a transgressive-regressive cycle in the style of Embry (1991). However, the lithostratigraphic record of this period is very different in detail for the two basins (Fig. 17). Both differences and similarities in the lithostratigraphy of the two basins form the basis for the following interpretations of variations in factors controlling deposition. A number of possibilities are presented where the data are inadequate to constrain the specific depositional variables.

46

I.K. SINCLAIR

Fig. 17. Lithostratigraphic chart showing comparison of mid-Aptian to upper Albian syn-rift sediments in the Jeanne d'Arc and Porcupine Basins. The presence of basal angular unconformities indicates that portions of each basin experienced uplift and erosion immediately preceding a midAptian onset of brittle failure and fault-controlled subsidence. Progradational parasequence sets, drilled in the Jeanne d'Arc Basin centre (Fig. 7), record a Barremian to early Aptian period of basin margin uplift. Erosion of sandrich basin margin areas generated the detritus which was deposited during northward regression of the Jeanne d'Arc Basin coastline. In contrast, erosional products of mid-Cretaceous uplift have not been recognized to date within the sedimentary column of the Porcupine Basin. This apparent lack of Barremian-age coastal deposits may indicate less uplift and erosion in the Porcupine Basin area relative to the Jeanne d'Arc Basin or the unroofing of shale-dominated facies unlikely to yield significant sand-sized detritus. Alternatively, it is possible that progradational coastal deposits are preserved below the mid-Aptian unconformity in the more central portion of the Main Porcupine Basin, south of the studied area. The basal retrogradational parasequence set (transgressive half-cycle) of coastal deposits is generally much thicker in the Jeanne d'Arc

Basin than in the Porcupine Basin. Intervals containing > 300m of retrogradational shoreface sandstones have been drilled in the Jeanne d'Arc Basin (Sinclair 1993), while the basal shoreface sandstones of 26/28-2 are only 40 m thick (Fig. 16). The following mechanism appears to have controlled deposition of the thick basal parasequence set in the Jeanne d'Arc Basinl Rapid fault-controlled subsidence was nearly matched by massive sediment input from the synchronously uplifted basin margins. The interplay of these two factors resulted in slow transgression of the coastline and high preservation rates for retrogradational sediments. In contrast, the relatively thin preservation of retrogradational shoreface sandstones within the down-faulted portion of the Porcupine Basin and the nil to minimal preservation of transgressive lag deposits on flanking margins (Fig. 14) indicates that the Porcupine Basin was rapidly transgressed in mid-Aptian times following a phase of regional uplift and erosion. This rapid transgression may be attributed initially to rapid subsidence rates for the down-faulted portion of the Porcupine Basin combined with low rates of sediment input from relatively low-relief basin margin areas during the initial phase of the syn-

APTIAN-ALBIAN RIFTING OF CONJUGATE ATLANTIC MARGINS rift episode. The low relief of the basin's northeastern margin (Fig. 13) also resulted in very rapid transgression once that area subsided below sea level and, consequently, low preservation potential for retrogradational deposits. The upper progradational parasequence (regressive half cycle) of coastal deposits is relatively thin (c. 75 m) and only locally developed in the Jeanne d'Arc Basin (Fig. 11). The continuous deepening of depositional environments from mid-Aptian into Cenomanian/Turonian times over most of the Jeanne d'Arc Basin (Fig. 10) indicates that rift-induced subsidence was generally followed without interruption by regional, thermal subsidence. As a result, no widespread phase of regression occurred in the Jeanne d'Arc Basin at the end of rifting. In contrast, a prominent package of progradational coastal sediments was deposited along the northern and eastern margins of the Porcupine Basin (see map of PK2 sediments in Shannon et al. 1993). This pattern may be driven by the arrival of large volumes of sandy detritus during the latter portion of the syn-rift episode, possibly in response to late uplift of clastic source areas marginal to the basin. Decreasing tectonic subsidence during the last phase of a syn-rift episode (Sinclair et al. 1994) may also account for such a late syn-rift regression of the coastline. Coastal environments of deposition rapidly migrated south and west across the relatively unstructured eastern flank of the Porcupine Basin (e.g. area of 26/29-1 and 26/28-4 wells), until they reached that portion of the basin which was downthrown along rift-induced faults (e.g. 26/28-1 area, Fig. 14). The large increase in accommodation space available west of the faulted margin resulted in formation of relatively steep seabed slopes (i.e. clinoform reflectors of Fig. 13) as the basin began infilling. As a further consequence, southward and westward regression of coastal environments was slowed and eventually halted along the northern and eastern margins of the Porcupine Basin where sediment input was insufficient to fill the large accommodation space made available by tectonism. The latest Albian to Early Cenomanian flooding of basin margins, termination of the preceding extensional phase, and the change to widespread marlstone and carbonate deposition in both the Jeanne d'Arc and Porcupine Basin areas closely matches the age of the first sediments deposited atop oceanic crust adjacent to the Goban Spur (i.e. DSDP 550; Masson et al. 1985). The above described features of Late Cretaceous deposition are, therefore, attributed to onset of thermal subsidence of the newly

47

formed Orphan Knoll/Flemish Cap and Porcupine Bank/Goban Spur continental margin areas, with the resultant effect being a significant rise in 'relative' sea level. In conclusion, it is seen that tectonicallycontrolled, unconformity-bound sequences can be matched and compared in conjugate margin basins flanking the North Atlantic Ocean. Synchronous rifting across continental plates is observed to effect depositional mechanisms including topography and rates of basin subsidence, margin uplift and sediment input over very wide areas. The lithostratigraphic response to the same rift stress, however, is clearly variable and depends on the local interplay of these depositional mechanisms. BP Exploration is gratefully acknowledged for their support of research on the Grand Banks area through their Les Adamson Memorial grant to the University of Aberdeen. I thank the Irish Department of Transport, Energy and Communications for their assistance in supplying well and seismic data from the Porcupine Basin. Thanks also to the University of Aberdeen and the Canada-Newfoundland Offshore Petroleum Board for their support of this work. Elvis Chippett and Bill Scott kindly digitized and plotted wireline log data and Mary Glynn typed the text.

References AINSWORTH, N. R. 1990. Uppermost Rhaetian to

lower Bajocian Ostrocoda from the Porcupine, Slyne, Erris and Donegal Basins, offshore west Ireland. Bulletin of the Geological Survey of Ireland, 4, 169-200. ARTHUR, K. R., COLE, D. R., HENDERSON,G. G. L. & KUSHNIR, D. W. 1982. Geology of the Hibernia discovery. In: HALBOUTY, M. T. (ed.) The Deliberate Search for the Subtle Trap. American Association of Petroleum Geologists Memoir 32, 181-195. C-NOPB (CANADA--NEWFOUNDLAND OFFSHORE PETROLEUMBOARD). 1986. Decision 86.01 Appli-

cation for approval Hibernia benefits and development plans. Released seismic surveys and reports are available for viewing at the offices of the Canada-Newfoundland Offshore Petroleum Board, Fifth Floor, TD Place, 140 Water Street, St. John's, Newfoundland. Requests for reproduction of such data may also be sent to the Exploration Manager at the same address. CROKER, P. F. & KLEMPERER, S. L. 1989. Structure and stratigraphy of the Porcupine Basin: relationships to deep crustal structure and the opening of the North Atlantic. In: TANKARD, A. J. & BALKWILL, H. R. (eds), Extensional Tectonics and Stratigraphy of the North Atlantic Margins. American Association of Petroleum Geologists, Memoir 46, 445-459. - d~ SHANNON, P. M. 1987. The evolution and hydrocarbon prospectivity of the Porcupine Ba-

48

I . K . SINCLAIR

sin, offshore Ireland. In: BROOKS, J. & GLENNIE, Newfoundland. In: VAN DER LINDEN, W. J. M. K. W. (eds) Petroleum Geology of North West & WADE, J. A. (eds) Offshore Geology of Eastern Europe. Graham & Trotman, London, 633-642. Canada. Geological Survey of Canada, Paper 74EMBRY, A. F. 1991. Mesozoic history of the Arctic 30, 2, 51-105. Islands. In: TRETTIN, n . P. (ed.) Geology of the KEEN, C., PEDDY, C., DE VOOGD, B. & MATTHEWS,D. Innuitian Orogen and Arctic Platform of Canada 1989. Conjugate margins of Canada and Europe: and Greenland. Geological Survey of Canada, 3, results from deep seismic reflection profiling. 371-433. Geology, 17, 173-176. ENACHESCU, M. E. 1987. Tectonic and structural LOUTIT, T. S., HARDENBOL, J., VAIL, P. R. & BAUM, framework of the northeast Newfoundland conG. R. 1988. Condensed sections: the key to age tinental margin. In: BEAUMONT, C. & TANKARD, dating and correlation of continental margin A. J. (eds) Sedimentary Basins and Basin-forming sequences. In: WILGUS, C. W., HASTINGS, B. S., Mechanisms. Canadian Society of Petroleum POSAMENTIER, H. W., VAN WAGONER, J., ROSS, Geologists Memoir, 12, 117-146 C. A. & KENDALL, C. G. St. C. (eds) Sea-level FALVEY, D. A. 1974. The development of continental Changes: An Integrated Approach. Society of margins in plate tectonic theory. Journal of Economic Paleontologists and Mineralogists, Australian Petroleum Exploration Association, Special Publication, 42, 183-213. 14, 95-106. MACDONALD, H., ALLAN, P. M. & LOVELL, J. P. B. FOSTER, D. G. & ROBINSON, A. G. 1993. Geological 1987. Geology of oil accumulation in block 26/28, history of the Flemish Pass Basin, offshore Porcupine Basin, offshore Ireland. In: BROOKS,J. Newfoundland. Bulletin of the American Associa& GLENNIE, K. (eds) Petroleum Geology of North tion of Petroleum Geologists, 77, 588-609. West Europe. Graham & Trotman, London, 643HARDING, S. 1988. Facies interpretation of the Ben 651. Nevis Formation in the North Ben Nevis M-61 MASSON, D. G. & MILES, P. R. 1986. Development well, Jeanne d'Arc Basin, G r a n d Banks, and hydrocarbon potential of Mesozoic sedimenNewfoundland. In: JAMES, D. P. & LECKIE, D. tary basins around margins of North Atlantic. A. (eds) Sequences, Sedimentology: Surface and Bulletin of the American Association of Petroleum Subsurface, Canadian Society of Petroleum GeolGeologists, 70, 721-729. ogists, Memoir, 15, 291-306. , MONTADERT, L. & SCRUTTON, R. m. 1985. HISCOTT, R. N., WILSON, R. C. L., HARDING, S. C., Regional Geology of the Goban Spur Continental PUJALTE, V. & KITSON, D. 1990b. Contrasts in Margin. In: lnitial Reports of the Deep Sea Early Cretaceous depositional environments of Drilling Project 80. US Government Printing marine sandbodies, G r a n d B a n k s - I b e r i a n Office, Washington, DC, 1115-1139. Corridor. Bulletin of Canadian Petroleum McALPINE, K. D. 1990. Mesozoic stratigraphy, Geologists, 38, 203-214. sedimentary evolution and petroleum potential , - - , GRADSTEIN, F. M., PUJALTE, V., of the Jeanne d'Arc Basin, Grand Banks of GARCIA-MONDEJAR, J., BOUDREAU, R. R. & Newfoundland. Geological Survey of Canada, WISHART, H. A. 1990a. Comparative stratigraphy Paper 89-17. and subsidence history of Mesozoic rift basins of MITCHUM, R. M. 1977. Seismic stratigraphy and North Atlantic. Bulletin of the American Associaglobal changes of sea level, Part I: Glossary of tion of Petroleum Geologists, 74, 60-76. terms used in seismic stratigraphy. In: PAYTON, HUBBARD, R. J. 1988. Age and significance of sequence C. E. (ed.) Seismic stratigraphy - Applications to boundaries on Jurassic and Early Cretaceous Hydrocarbon Exploration, American Association rifted continental margins. Bulletin of the Amerof Petroleum Geologists, Memoir, 26, 205-212. ican Association of Petroleum Geologists, 72, 49MOBIL OIL CANADA PROPERTIES, CHEVRON CANADA 72. RESOURCES LIMITED, GULF CANADA RESOURCES , PAPE, J. & ROBERTS, D. G. 1985. Depositional LIMITED AND PETRO-CANADA INC. 1988. Intersequence mapping as a technique to establish pretation of the Hibernia structure at the Hibernia tectonic and stratigraphic framework and evaluand Ben Nevis/Avalon horizons, C-NOPB Project ate hydrocarbon potential on a passive continenNo. 8627-M3-19E. Released seismic surveys and tal margin. In: BERG, O. R. & WOOLVERTON, D. reports are available for viewing at the offices of (eds) Seismic Stratigraphy 11." An Integrated the Canada-Newfoundland Offshore Pet:oleum Approach to Hydrocarbon Exploration. American Board, Fifth Floor, TD Place, 140 Water Street, Association of Petroleum Geologists, Memoir, 39, St. John's, Newfoundland. Requests for repro79-91. duction of such data may also be sent to the HURLEY, T. J., KREISA, R. D., TAYLOR, G. G. & Exploration Manager at the same address. YATES, W. R. L. 1992. The reservoir geology and MOORE, J. G. 1992. A syn-rift to post-rift transition geophysics of the Hibernia Field, offshore sequence in the Main Porcupine Basin, offshore Newfoundland. In: HALBOUTY, M. T. (ed.) Giant western Ireland. In: PARNELL, J. (ed.) Basins on Oil and Gas Fields of the Decade 1977-88, the Atlantic Seaboard." Petroleum Sedimentology American Association of Petroleum Geologists, and Basin Evolution. Geological Society, London, Memoir, 54, 35-54. Special Publication, 62, 333-349. JANSA, L. F. & WADE, J. A. 1975. Geology of the - - & SHANNON, P. M. 1992. Palaeocene-Eocene c o n t i n e n t a l margin off N o v a Scotia and deltaic sedimentation, Porcupine Basin, offshore

APTIAN-ALBIAN RIFTING OF CONJUGATE ATLANTIC MARGINS Ireland - a sequence stratigraphic approach. First Break, 10, 461-469. MURPHY, N. J. & CROKER, P. F. 1992. Many play concepts seen over wide area in Erris, Slyne troughs off Ireland. Oil and Gas Journal, 90, 9297. NAVLOR, D. & SHANNON, P. 1982. Porcupine Basin, In: Geology of Offshore Ireland and West Britain, Graham & Trotman, London, 65-75. NUMMEDAL, D. & SWIFT, D. J. P. 1987. Transgressive stratigraphy at sequence-bounding unconformities: some principles derived from Holocene and Cretaceous examples. In: NUMMEDAL, D., PILKEY, O. H. & HOWARD, J. D. (eds) Sea-level Fluctation and Coastal Evolution, Society of Economic Paleontologists and Mineralogists, Special Publication, 41, 241-260. PAREX 1983. Geophysical survey over the Jeanne d'Arc, Flemish Pass area, C-NOPB Project No. 8620-S14-8E. Released seismic surveys and reports are available for viewing at the offices of the C a n a d a - N e w f o u n d l a n d Offshore Petroleum Board, Fifth Floor, TD Place, 140 Water Street, St. John's, Newfoundland. Requests for reproduction of such data may also be sent to the Exploration Manager at the same address. PEGRUM, R. M. & MOUNTENEY, N. 1978. Rift basins flanking North Atlantic Ocean and their relation to North Sea area. Bulletin of the American Association of Petroleum Geologists, 62, 419-441. POSAMENTIER, H. W., JERVEY, M. T. & VAIL, P. R. 1988. Eustatic controls on clastic deposition I conceptual framework. In: WTLGUS, C. W., HASTINGS, B. S., POSAMENTIER, H. W., VAN WAGONER, J., ROSS, C. A. & KENDALL, C. G. St. C. (eds) Sea-level Changes - An Integrated Approach, Society of Economic Paleontologists and Mineralogists, Special Publication, 42, 109124. SHANNON, P. M. 1991. The development of Irish offshore sedimentary basins. Journal of the Geological Society, London, 148, 181-189. , MOORE, J. G., JACOB, A. W. B. & MAKRIS, J. 1993. Cretaceous and Tertiary basin development west of Ireland. In: PARKER, J. R. (ed.) Petroleum Geology of North West Europe: Proceedings of the 4th Conference, Geological Society, London, 1057-1066. SINCLAIR, I. K. 1988. Evolution of Mesozoic-Cenozoic sedimentary basins in the Grand Banks area of Newfoundland and comparison with Falvey's (1974) rift model. Bulletin of Canadian Petroleum Geologists, 36, 255-273. - 1993. Tectonism: the dominant factor in midCretaceous deposition in the Jeanne d'Arc Basin, Grand Banks. Marine Petroleum Geology, 10, 530-549.

- - ,

49

SHANNON, P. M., WILLIAMS, B. P. J., HARKER, S. D. & MOORE, J. G. 1994. Tectonic controls on sedimentary evolution of three North Atlantic borderland Mesozoic basins. Basin Research, 193-217. TANKARD, A. J. d~ WELSINK, H. J. 1987. Extensional tectonics and stratigraphy of Hibernia oilfield, Grand Banks, Newfoundland. Bulletin of the American Association of Petroleum Geologists, 71, 1210-1232. & 1988. Extensional tectonics and stratigraphy of the Mesozoic Grand Banks of Newfoundland. In: MANSPEIZER,W. (ed.) Triassic-Jurassic Rifting: Continental Breakup and the Origin of the Atlantic Ocean and Passive Margins, Part A, Developments in Geotectonics 22. Elsevier, Amsterdam, i29-165. & JENKINS, W. A. M. 1989. Structural styles and stratigraphy of the Jeanne d'Arc Basin, Grand Banks of Newfoundland. In: TANKARD, A. J. & BALKWILL, H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins. American Association of Petroleum Geologists, Memoir, 46, 265-282. TATE, M. P. & DOBSON, M. R. 1988. Syn- and post-rift igneous activity in the Porcupine Seabight Basin and adjacent continental margin W of Ireland. In: MORTON A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and Opening of the NE Atlantic, Geological Society, London, Special Publication, 39, 309-334. & 1989. Late Permian to early Mesozoic rifting and sedimentation offshore NW Ireland. Marine Petroleum Geologist, 6, 49-59. THOMPSON, L. B. 1987. Cretaceous transgressive and regressive events in the Avalon Basin, Grand Banks of Newfoundland. Cushman Foundationfor Foraminiferal Research, Special Publication, 24, 187-195. VAN WAGONER, J. C., POSAMENTIER, H. W., MITCHUM, R. M., VAIL, P. R., SARG, J. F., LOUTIT, T. S. HARDENBOL, J. 1988. An overview of sequence stratigraphy and key definitions. In: WILGUS, C. W., HASTINGS, B. S., POSAMENTIER, H. W., VANWAGONER, J., ROSS, C. A. & KENDALL,C. G. St. C. (eds) Sea Level Changes - An Integrated Approach, Society of Economic Paleontologists and Mineralogists, Special Publication, 42, 39-45. WADE, J. A. 1978. The Mesozoic-Cenozoic history of the northeastern margin of North America. Proceedings of the Tenth Annual Offshore Technology Conference, Houston, TX, USA, 8-11 May 1978, 3, 1849-1858. ZIEGLER, P. A. 1988. Evolution of the Arctic-North Atlantic and the Western Tethys. American Association of Petroleum Geologists, Memoir, 43. -

-

Early Tertiary evolution and sequence stratigraphy of the FaeroeShetland Basin: implications for hydrocarbon prospectivity C. C. E B D O N , 1'3 P. J. G R A N G E R , 1'4 H. D. J O H N S O N 2'5 & A. M. E V A N S 2

1Bp Exploration, 301 St Vincent Street, Glasgow, UK 2Shell UK, Shell-Mex House, Strand, London WC2R ODX, UK 3Present address: Amoco (UK) Exploration Company, West Gate, London W5 1XL, UK 4present address: Insight Training, Hillfoot House, The Green, WethersfieM, Essex 5Present address." Department of Geology, Imperial College, Prince Consort Road, London SW7 2BP, UK Abstract: Eight seismically resolvable packages have been identified in the Early Tertiary

succession of the Faeroe-Shetland Basin. These packages are calibrated with the microfaunal and microfloral biostratigraphy, thus allowing correlation between wells and comparison with other basins. Mapping of the facies distribution within each package has allowed a detailed understanding of the evolution of the southwestern part of the basin, including the Quad 204 acreage. The development of the area can be related to the plate

tectonic evolution of the North Atlantic and has highlighted implications for hydrocarbon exploration. In the earliest part of the Palaeocene (Sequences T10-T32) the position of the shelf/slope break was controlled by the underlying end Cretaceous fault induced topography. The depositional character of the submarine fans which accumulated in the basin is aggradational, collecting in basin floor deeps and progressively onlapping basin floor highs. Reservoir quality is initially controlled by depositional facies but is, essentially, excellent. Basinal muds, particularly regionally extensive maximum flooding surfaces, are believed to provide effective seals. Sequences T36-T40 are progradational in character, the position of the shelf/slope break advancing significantly beyond the positions of underlying Cretaceous fault scarps. This change in depositional character is related to thermal doming and rifting in the North Atlantic. The results of this uplift in the Faeroe-Shetland Basin also instigated a change in sediment provenance, volcanism and the deposition of extensive baseof-slope fan systems. Reservoir effectiveness remains excellent in these fan systems and potential seals are provided by downlapping, progradational shelf/slope systems. Sequences T45 and T50 represent flooding of the Palaeocene shelf systems as the thermal dome collapsed and rifting progressed. The overall control on Early Tertiary reservoir effectiveness in the Faeroe-Shetland Basin is depth of burial. In this part of the basin the Early Tertiary reservoir sands are generally buried at depths less than 2.5 km below sea-bed and reservoir quality remains excellent (porosity > 25% and permeability > 100 Md). The challenge to the oil industry is to understand the charge of this attractive hydrocarbon play. A Palaeocene h y d r o c a r b o n play has been recognized in the Faeroe-Shetland basin since the early 1970s. The earliest exploration for the play, commencing in 1978, focused on the northern part of the basin in Quads 205/206. The first hydrocarbon discovery of gas was made by Shell in 1980 (206/2-1). Other gas discoveries (e.g. 214/27-1) followed in the mid1980s, occurring in submarine fan reservoirs similar to those found in the Palaeocene of the North Sea. More recent efforts, by the BP/Shell partnership, have concentrated on the southern part of the basin, centred on Quad 204, more particularly on the twelfth round licences P556 (204/19 & 20), P557 (204/24) and P558 (204/22 &

23). This paper presents the results of a major

integrated review of the area culminating in 1992 and represents a benchmark technical view from this time. It incorporates a wide variety of geological and geophysical data and exploration techniques. Seismic and gravity data have been interpreted at both regional and prospect scales. Major improvements in seismic data quality, especially in the Tertiary section, were achieved by careful acquisition, processing and reprocessing. Improved seismic data quality has proved fundamental to regional sequence mapping and prospect definition. A sequence stratigraphic framework has been developed in order to understand the evolution of the Lower Tertiary interval. This is based on detailed reviews of the biostratigraphy, heavy mineral distribution and provenance, sedimentology and lithofacies of 15

ffrom Scrutton, R. A., Stoker, M. S., Shimmield, G. B. & Tudhope, A. W. (eds), 1995, The Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region, Geological Society Special Publication No. 90, pp. 51-69

51

52

C.C. EBDON E T AL.

Fig. 1. Location and Major Tectonic Elements. Shading indicates platform areas and basinal highs. BD, Brendans Dome Volcanic Centre; EVC, Erland Volcanic Centre; COR, Corona Ridge; CER, Central Ridge; VR, Victory Ridge; RR, Rona Ridge; CR, Clair Ridge; WR, Westray Ridge; SSH, Sula Sgeir High; WTR, Wyville-Thomson Ridge; SP, Shetland Platform; VT, Victory Transfer; UT, Unnamed Transfer; JF, Judd Fault; SSF, Shetland Spine Fault; FSB, Faeroe-Shetland Basin; VB, Victory Basin; CB, Clair Basin; WSB, West Shetland Basin; NRB, North Rona Basin; RT, Rockall Trough. wells within the area. These data have been integrated with the regional seismic sequence mapping to create a series of gross depositional environment (GDE) maps to illustrate the facies development and evolution of the area with time. This work provides the essence to understanding reservoir and seal presence within the playfairway. The objective of this paper is to summarize the Palaeocene evolution of the Faeroe-Shetland Basin and evaluate the implications for hydrocarbon prospectivity.

Regional tectonic setting The Faeroe-Shetland Basin is characterized by a dominant NE-SW tectonic grain inherited from the Early Palaeozoic Caledonian Orogeny (Fig. 1). This tectonic grain is clearly seen on gravity and magnetics data and is well documented (e.g. Ridd 1983; Bott 1984; Duindam & van Hoorn 1987). The basin is further compartmentalized by NW-SE-trending transfer elements which dissect this Caledonian grain.

HYDROCARBONS IN FAEROE-SHETLAND BASIN

53

Fig. 2. Tectonic elements of the study area. A-A' is the line of section in Figs 6 and 9; T, Transfer elements. The Shetland Platform provides the stable eastern margin of the basin, and joins with the Scottish Massif as a major area of cratonic basement. The eastern most boundary of the basin is limited by the Shetland Spine Fault and its along strike equivalents. On this eastern side of the basin a number of sub-basins are developed, separated from each other by N W SE-trending transfer zones. They are, from north to south, the Victory, Clair, West Shetland and North Rona Basins. All of these basins underwent significant Triassic extension and accumulated considerable thicknesses (up to

6km on seismic data) of Triassic clastic sediments. The Victory, Clair and West Shetland Basins underwent further extension during the Early Cretaceous. The eastern sub-basins are separated from the main Faeroe-Shetland Basin by the Rona Ridge and its along strike equivalents (the Clair Ridge and the Victory Ridge to the northeast). These positive features began to develop during Early Cretaceous extension and were enhanced during a phase of 'mid' Cretaceous (Coniacian-Santonian) extension. Over 4 km of Late Cretaceous deposits accumulated in parts of the Faeroe-

54

C.C. EBDON

ET AL.

Fig. 3. Palaeocene (Thanetian: Anomaly 26) plate reconstruction of the North Atlantic (modified from Knott et 1993). Dark shading, oceanic crust; medium shading, deep marine deposits; light shading, marine shelf deposits; horizontal shading, volcanic centres; NAM, North America; GB, Great Britain; IR, Ireland. Circle indicates area of thermal uplift. al.

Shetland Basin, most notably in the north in blocks 213/25, 214/21 and 214/22. Thicknesses are, however, variable due to the persistence of fault controlled, pre-rift, highs such as the Flett Ridge, the Central Ridge and the Westray Ridge (Fig. 1). The northwestern boundary of the basin is poorly mapped and poorly understood due to a lack of data and the existence of extensive volcanics. One significant NW-SE transfer element, and un-named transfer fault parallel to the Judd Fault, separates the Faeroe-Shetland Basin into northern and southern sub-basins which have had significantly different histories. It is the most southerly of these sub-basins, which includes the Quad 204 acreage, that forms the focus of this paper. The south and eastern limits of this southern basin were controlled by the Judd Fault and extension of the Rona Fault system.

The northwestern limit appears to be controlled by a major series of down-to-the-southeast faults (Fig. 2). The Judd Fault is thought to be a transfer fault, related to the Wyville-Thomson Ridge, a major Atlantic transform/transfer lineament. The down-to-the-basin faults and the associated transfer elements create a rectangular framework to the southern Palaeocene sub-basin. These faults are clearly very important in a consideration of both the depositional history of the Early Tertiary and subsequent tectonic movements in the region.

Early Tertiary plate tectonic setting During the Early Tertiary there was active rifting along the length of the North Atlantic. The rifting was superimposed by a major thermal dome or 'hot spot' centred over east

HYDROCARBONS IN FAEROE-SHETLAND BASIN

55

Fig. 5. Definition of sequences and sequence boundaries. MFS, maximum flooding surface.

Fig. 4. Early Terti.ary stratigraphy summary. Greenland. The extent of this 'hot spot' is estimated at over 2000 km in diameter (Fig. 3) and in the Faeroe-Shetland Basin uplift is estimated by Joppen & White (1990) to have been in the order of 1.5 km. A base Palaeocene unconformity at the margins of the North Sea and the West of Shetlands basins is related to this regional uplift as are the major Palaeocene

submarine fans of the North Sea Basin (Anderton 1993). The reactivation of NW-SE-trending structural elements is also suggested during the evolution of this 'hot spot' (Knott et al. 1993).

Stratigraphy of the Palaeocene playfairway Within the Early Tertiary of the North Sea a number of regionally extensive stratigraphic sequences are identified by BP. These are broadly equivalent to lithostratigraphic units and are from oldest to youngest, T10 (Ekofisk),

56

C.C. EBDON ET AL.

Fig. 6. Schematic chronostratigraphic summary. For key to facies see Fig. 10. T20 (Maureen), T30 (Lista), T40 (Forties), T45 (Sele), and T50 (Balder). These units are also identified in the Faeroe--Shetland Basin (Fig. 4). In the present study T30 is further subdivided into three packages (T32, T34 and T36) on seismic and well evidence, but not all can be mapped with confidence regionally. The older units (T10-T32) are all genetic stratigraphic sequences (sensu Galloway, 1989) and are picked on gamma ray maxima (Fig. 5) which are believed to indicate times of maximum flooding (maximum flooding surfaces). These would have been essentially isochronous across the entire area. Maximum flooding surfaces are represented on seismic data as downlap surfaces. The flooding surfaces have been constrained with the microfloral and microfaunal biostratigraphy (Fig. 4) and can, therefore, be calibrated and correlated with a large degree of confidence. Within the stratigraphy genuine sequence boundaries (sensu Vail et al. 1977) can also be recognized, being picked at the base of the lowstand systems tract. In the basinal settings, where sands are developed, the sequence boundary is readily picked at the base of the sharp based fan sands (e.g. T10 Fig. 5). If no sands are developed within the sequence, however, the sequence boundary cannot be picked with confidence (e.g. T20 Fig. 5). Since the maximum

flooding surfaces are most extensively developed, and are most confidently recognized on log data, sequences T10-T32 are defined as genetic stratigraphic sequences. The younger Palaeocene 'sequences', however, are generally developed in a more proximal setting. The base of Sequence T36 is a true sequence boundary in all wells in the area, T34 being truncated by T36 as seen in Fig. 6, and a downward shift in facies belts identified from T34 to T36. T40 is another lowstand system but is restricted to the basin centre. T40 is separated from T45 by a flooding surface and the base of T50 is a regional unconformity developed over most of the area during the T40 lowstand. The top of T50 is another regional maximum flooding surface. 'Sequences' T36, T40, T45 and T50 are, therefore, variably a true sequence (T36), a lowstand and transgressive system tract (T40), a highstand systems tract (T45) and a transgressive systems tract (T50).

Regional seismic mapping Considerable effort was made to map seismically all sequences identified. This task involved an iteration between well and seismic data, as neither dataset is optimal in terms of quality

HYDROCARBONS IN FAEROE-SHETLAND BASIN

57

Fig. 7. Geo-seismic section. and coverage. The seismic in the area is tied to the wells at several key horizons. Some of these sequence and interval boundaries were regionally mappable, whilst others are identifiable only on one or two seismic lines or in localized areas. The top T50 (Balder), top Cuillin (intra T36 fan), base T36, top T20, top T10 and Base Tertiary intervals are all mapped regionally. The horizons are interpreted across the entire available seismic dataset. Over the licence areas this includes a dataset of several thousand kilometres of block-specific lines. Although the coverage is relatively dense, the quality is generally very poor within the Palaeocene section except where lines have been re-processed specifically for the Palaeocene. The blockspecific nature of a majority of the seismic dataset has impaired the regional interpretation. However, in 1991, BP and Shell acquired a semiregional survey which proved useful in extending the regional stratigraphic knowledge and the definition of the extent of the southern Palaeocene sub-basin. All regional lines were constructed from small segments of block-specific data and semi-regional lines. A geoseismic interpretation of one of these lines is shown in Fig. 7. This line also shows the key stratigraphic intervals that were defined by the regional mapping and geological evaluation.

Early Tertiary structure The present-day Early Tertiary structure is controlled by several factors.

Underlying pre- Tertiary structure and lithology Much of the fault-block topography, characteristic of the Quad 204 area, had been blanketed by the end of the Late Cretaceous with a highly variable thickness of mudstones. Nevertheless, differential compaction of these mudstones around the topographic highs, resulted in basinal deeps overlying Late Cretaceous isopach 'thicks'. The depositional limits of the Early Tertiary were, therefore, controlled by underlying fault-block topography. In general the Early Tertiary basin is thickest where the Late Cretaceous isopach is at its thickest; however, in detail there are some complications. The T10 and T20 isopach 'thins' are offset from the Late Cretaceous 'thins'. Additionally some of the latest Late Cretaceous section is missing indicating that possible minor inversion and some erosion of the Late Cretaceous took place prior to the deposition of the oldest Palaeocene sediments.

The depositional style and geometry of the Early Tertiary basin-fill The basin filling style during the Early Tertiary can be simplified to a period of submarine fan aggradation followed by a period of progradation. The depositional shapes of an earlier sequence control, to a degree, shapes the overlying sequences, and these, in turn, have had an effect on the present-day structural attitude

58

C.C. EBDON ET AL.

of the horizons. The filling style and facies distributions of the Palaeocene sequences is discussed in detail below.

lntra-Palaeocene gravity and tectonic episodes The deposition of the Palaeocene in the Quad 204 area is believed to have been punctuated by gravity plate induced tectonic events. Folding and faulting observed at the Top T32 level are believed to be related to a large gravity slide, caused by an increase in sea-floor slopes associated with increased thermal uplift. Some localized thickening observed within specific intervals, close to major faults, may be related to minor tectonic movements. It is likely that these types of movements occurred during this period in response to minor plate reorganization associated with the opening of the North Atlantic basin. It is, however, difficult to distinguish these isopach variations from differential compaction effects over and beside the major basin-forming faults.

Late Palaeocene subsidence The dominant regional dip in the Quad 204 area is from southeast to northwest. This dip is caused by regional thermal subsidence in the Faeroe Channel area following Cretaceous rifting. Further west the regional dip switches direction, presumably marking the opposite side of the Early Tertiary basin. It is apparent that the Early Tertiary sub-basin described in this report is one half of a larger basin that is obscured by Late Tertiary volcanics.

Late Tert&ry structural inversion Two compressional events can be recognised during the Late Tertiary in this area. One is Early Oligocene and the other Late Miocene in age. Both are related to the onset of opening of the Norwegian-Greenland Sea (Knott et al. 1993). In Quad 204 these compressional events reactivated pre-existing basin forming faults and transfers, creating folding in the overlying Palaeocene sediments. The folding appears to be related to trans-pressional movements on the transfer faults that delineate the Palaeocene subbasin. Further southwest the Late Tertiary deformation becomes intense in the region of the Wyville-Thomson Ridge (Fig. 1). It appears as if this major transfer element has taken up much of the plate stress, protecting other areas from such major deformation.

Early Tertiary sequence stratigraphy and evolution A series of gross depositional environment maps have been constructed which represent an iteration of well data (wireline log, lithofacies, biofacies etc), seismic data (seismic facies, isopachs etc), structural setting and sediment provenance. These maps form the basis for understanding the evolution of the Early Tertiary in terms of reservoir and seal distribution. The sequences are discussed below from oldest to youngest.

Sequence TIO (Fig. 8a) Definition. Sequence T10 is a genetic stratigraphic sequence and is bounded by maximum flooding surfaces. Seismically its boundaries are represented by downlap surfaces, though these are either very subtle or indistinguishable over much of the area. The lower boundary occurs immediately above high amplitude, parallel reflectors characteristic of the more calcareous Cretaceous deposits of the Shetland Group. This maximum flooding surface is associated with earliest Palaeocene microfaunas and microfloras assigned to biozones MT1 (equivalent to P1 of Blow 1979) and PT1 respectively. Microfaunas are characterized by planktonic foraminifera and include Globigerina simplicissima, Globigerina aft. trivial& sensu Blow (1979), Globorotalia planocompressa, Globorotalia eobulloides and Globigerina trivialis. In terms of microflora the lower part of T10 is characterized by Senoniasphaera inornata, Spongodinium delitiense, Xenicodinium spp., Palaeoperidinium pyrophorum, Areoligera spp. and Palaeocystodinium cf. australinum. This lower flooding surface occurs immediately above the appearance of Cretaceous microfaunas and floras. The upper boundary occurs within biozones MT3 (associated with the extinctions of Globigerina pseudobulloides and Globigerina compressa), immediately above the top of MT2 (marked by the top occurrence of Globigerina trivial&), and within PT4 (immediately above the top of PT3, marked by the extinction of an un-named species of

Spiniferites). Description. Sequence T10 has been recognized in 12 wells in the immediate Quad 204 area. In the southwest the sequence is very thin (204/281; 12.5 m) or is condensed or absent (204/27a-1). It appears likely that the position of the Judd Fault complex controlled the position of the Palaeocene shelf/slope break and that a structu-

HYDROCARBONS IN FAEROE-SHETLAND BASIN

Fig. 8. Gross depositional environments, T10-T34. FI, footwall island; N/G, net/gross.

59

60

C.C. EBDON ET AL.

rally controlled shelf terrace was present in this area. This fault terrace collected only mud or was completely bypassed. It is possible that a footwall island complex existed during T10 times. In the southeast a sandy shelf is developed. These sands appear to have been derived from the Jurassic and Triassic sediments in the inverted sub-basins to the east. Further east, in 205/26-1, condensed carbonate shelf deposits are recorded during T10 times. The Judd Fault continues to control the position of the shelf/ slope break in this area as it does to the west. North of the Judd Fault complex two distinctive basinal thicks are recognized within T10. Two wells have been drilled in the eastern most of these thicks and both penetrated significant sand rich basin floor fans. On this basis basinal sands are interpreted to be present in these thicks and the edge of the fans ( < 10 m sand) is mapped on the 80m sequence isopach. High quality sands ( > 4 0 % N/G) have been proven where the sequence isopach exceeds 200 m, but in undrilled areas the limit of good sands is arbitrarily mapped where the sequence isopach exceeds 400m. These T10 thicks are slightly offset from the underlying Late Cretaceous thicks, possibly as a result of inversion during the Late Cretaceous and differential compaction over Cretaceous footwalls. The distribution of the thicks and the sands, in the basinal setting are, therefore, controlled by end Cretaceous structure and basin floor topography. A likely entry point to the basin for sediments in the easternmost T10 thick is the point where the Judd Fault complex intersects the Rona Fault complex in the northern part of block 204/29. This would have tapped the sandy shelf seen in the area and supplied coarse grained sediment to the basin floor fans observed. An entry point to the west, feeding sediment into the western isopach thick, is also interpreted to have existed but its position is less obvious. The presence of sand prone sediments in this fan is also considered to be of higher risk in view of the predominance of a mud prone shelf as a potential source area. Elsewhere, north of the Judd Fault complex, basinal muds were deposited (e.g. 204/23-1,204/19-1,205/16-1).

Interpretation. The basin floor thicks mapped on seismic data, and seen in well penetrations, are interpreted as lowstand fans the base of which marks a true sequence boundary. In the basin the deposition of the fans is terminated by the end T10 maximum flooding surface. In the basinal setting the sequence appears aggradational on seismic data, amplitude variations

possibly defining sands. Seismic data over the shelf area are poor and mapping of the shelf area is based almost exclusively on well data.

Sequence T20 (Fig. 8b) Definition. T20 is a genetic stratigraphic sequence, mapped between two maximum flooding (downlap) surfaces. The lower boundary is defined by the top of the underlying T10 sequence. The upper maximum flooding surface occurs within biozones MT4, characterized by an abundance of the radiolarian Cenodiscus lenticularis, and PT5, below the top of Isabeli-

dinium? viborgense. Description. Sequence T20 has been penetrated by 13 wells in the area and the distribution of facies within the sequence is similar to that of the preceding T10 Sequence. In the southwest T20, like T10, is thin and none of the wells in the southwest contains any sand within T20. These sediments may be locally derived from the adjacent shelf since Late Cretaceous and Late Jurassic mudstones subcrop the Palaeocene deposits. This is further supported by the presence of reworked palynomorphs of these ages which have been recovered from the sequence. The Judd Fault complex continues to control the position of the T20 shelf/slope break and there is no evidence that the shelf margin has advanced basinward since T10 times. This probably reflects the small amount of deposition in the area and/or the existing relief on the Judd Fault complex. In the southeast a sand prone shelf persists, possibly increasing in area relative to T10 times. The Judd/Rona Ridge Fault complex had less vertical relief in this area and this, coupled with the greater sediment supply, probably allowed the shelf edge to advance basinward from its T10 position. North of the T20 shelf edge the sequence is present in a deep water, basinal facies, but basin floor sands have only been penetrated by one well. Away from well control the edge of the T20 basin floor fans complex is mapped arbitarily on the 250 m isopach thick. This mapping indicates that the basin floor fans have partly onlapped the existing basin floor topography. The two distinct fans present during T 10 times appear to have coalesced in the unlicensed area north of 204/19. Sequence thicks, which are assumed to c()ntain higher quality sands (> 40% N/G), are mapped on the 400 m sequence isopach and are, as would be expected, offset from the T10 thicks. Sediment entry points into the basin were probably the same as those which existed during T10 though the possibility of input from the

HYDROCARBONS IN FAEROE-SHETLAND BASIN east, and northeast along the basin axis, cannot be discounted. Evidence for the derivation of sediments from a Mesozoic hinterland is given by the common occurrence of Late Jurassic and Cretaceous palynomorphs throughout the sequence.

Interpretation. The basin floor fans are characteristically very sharply based and have the 'box car' shape on wireline logs characteristic of lowstand fans. The basinal deposits continue to aggrade and fill the existing basin floor topography. Seismic quality over the shelf is poor, and once again mapping of the shelf is based almost exclusively on well data.

Sequence T32 (Fig. 8c) Definition. Sequence T32 is a genetic stratigraphic sequence. The upper boundary is defined by a gamma ray maximum within Biozone MT5 and below the top of Palaeocystodinium cf. australinum and consistent Palaeoperidinium

pyrophorum. Description. This sequence has been positively identified in 13 wells in the area and gross depositional environment mapping suggests significant changes in the position of the shelf edge and basin floor topography. The T32 shelf/slope break advanced significantly during T32 and the break of slope has been accurately mapped from seismic data. The position of the Judd Fault complex still, however, influences the position of the T32 shelf/slope break in the west. Further to the east the shelf edge has prograded significantly northwestwards from its position in T20 times, advancing beyond the position of well 205/16-1. This advance in the east is probably related to a combination of lower topographic relief over the existing footwalls and higher rates of sediment supply. In the west the shelf area continues to be mud dominated whilst in the east a sand dominated shelf persists. In 205/16-1 the presence of bryozoan sands provides excellent evidence of this sand prone shelf. In the area basinward of the T32 shelf > 10 m of sand has been encountered by all wells. Existing basin floor topography is believed to have been filled and basin floor fans onlap the base of the shelf slope. A basinal high may have still existed in the northern part of block 204/24 but its presence is speculative due to the paucity of good seismic evidence. The nature of the T32 sands is quite variable, possibly reflecting relative positions on the fan system or different controls on deposition. In the most basinal well (204/19-1) the sequence is particularly sand

61

prone (N/G 0.68) and appears on logs as blocky, probably amalgamated, sand packages with sharp bases and tops. Towards the T32 shelf edge the individual sands are more distinct, separated by mudstones. The sands continue to have sharp bases and tops but towards the top of the sequence the sands show a cleaning upwards motif. These differences are interpreted to represent central and more marginal fan positions respectively. Nearer the shelf again (e.g. 204/23-1) the sequence is predominantly mud prone and the only sands are thin and occur towards the top of the sequence.

Interpretation. From the evidence it is suggested that the initial sands of T32 were deposited as a series of lowstand fans. Sediment was probably fed into the basin through existing and, by now, well established entry points (e.g. Judd Fault/ Rona Ridge Fault intersection), though with the infilling of basin floor topography, axial sediment transport also appears likely. As relative sea level rose towards the end of T32, culminating in the end T32 maximum flooding surface, lowstand fan deposition was terminated. Evidence of this phase of fan abandonment is well documented on log data. It is suggested that some basinal sands may have been deposited as a result of slumping of the shelf edge due to slope instability. This is more likely to have happened towards the end of the sequence as relative sea level rise produced greater slope instability. Reworked fossils continue to indicate a Mesozoic provenance.

Sequence T34 (Fig. 8d) Definition. The base of T34 is defined by the top T32 maximum flooding surface. In wells the top of T34 is marked by a sharp change in facies suggested to define a type I sequence boundary. On seismic data this boundary is variably marked by onlap of T34 slope deposits or a downlap surface in both shelfal and basinal settings. In more proximal settings there is truncation or apparent truncation beneath this surface. Biostratigraphically Sequence T34 is confined to Biozones MT5 and PT7. Microfaunas are characterized by agglutinating foraminifera (including Bathysiphon spp., Rhizammina spp. and Spiroplectammina spectabilis) and diatoms whilst microfloras are characterized by Spini-ferites spp., Alisocysta margarita and Areoligera senonensis sensu Heilmann-Clausen (1985).

Description. Sequence T34 has been positively identified in seven wells and is probably pene-

62

C.C. EBDON E T AL.

Fig. 9. Sequence correlation. For line of section see Fig. 2. trated in five more where the sequence cannot be distinguished from those above it or below it. The position of the shelf edge is relatively well defined on seismic data and has prograded northwards relative to its position in T32 times. Proximal to the position of the shelf/slope break slope deposits the pattern of facies distribution on the T34 shelf area is similar to that which persisted throughout the preceding Palaeocene sequences. In the southwest the shelf is mud dominated (e.g. 204/27a-1,204/28-1 and 204/291) whereas in the southeast the shelf is sand dominated (e.g. 204/30-1), probably continuing to access Triassic deposits in the inverted basins to the east. In the basinal setting, beyond the base of slope, T34 has only been penetrated by two wells and in both of these the sequence is thin and mud prone. 204/23-1 is located on the T34 slope and contains a significant, coarse grained sandstone believed to represent channel fill deposits. This feeder channel, which is approximately on trend with the position of older transport corridors, probably feeds as yet undefined basin floor fans restricted to the western part of the sub-basin. In the southern part of the area the sequence

is absent due to erosion during the succeeding T36 lowstand. The erosional limit appears to have been partly structurally controlled by the southern limit of the Judd Fault Terrace which also influenced sedimentation during previous sequences. This erosional top to T34 is also manifested in the basin where the T34/T36 boundary is a Type I sequence boundary sensu Vail et al. (1977), occurring at the base of the T36 lowstand fan. Interpretation. The channel observed in 204/23-1, and the interpreted fans, are suggested to be lowstand deposits. The base of the channel represents a true sequence boundary. This fall in relative sea level is possibly related to uplift of the North Atlantic 'hot spot' and associated gravity sliding in the basin which produced structuring at end of T32 times. During T34 times there was an overall shift in depositional style in the basin from one of aggradation to one of progradation. This change is also associated with a significant change in the assemblages of heavy minerals and reworked palynomorphs between the end of T32 and the beginning of T36. Sequences T10-T32 are characterized by recycled Jurassic and Late Cretaceous palyno-

HYDROCARBONS IN FAEROE-SHETLAND BASIN

63

Fig. 10. Gross depositional environments, T36-T50. N/G, net/gross. morphs, the upward changes in the reworked assemblages suggesting a gradual unroofing of the provenance area. Heavy mineral suites are relatively depleted and are considered to have been derived from recycled sedimentary materi-

al. In Sequence T36, and younger Early Tertiary deposits, reworked palynomorphs are less abundant; Mesozoic forms are present in low numbers and occur in association with Palaeozoic (Carboniferous) palynomorphs. Heavy

64

C.C. EBDON ET AL.

mineral suites are characterized by abundant unstable minerals such as epidote, amphibole and pyroxene. These changes indicate derivation from much older sedimentary and first-cycle metamorphic material, and a major change in drainage patterns during the 'mid' Palaeocene. Due to the paucity of data from Sequence T34 it is not possible to say exactly when these changes occurred.

Sequence T36 (Fig. 10a & b) Definition. T36 is a true sequence (Vail et al. 1977), being bounded by unconformities or their correlative conformities. The sequence is well constrained biostratigraphically, occurring within biozones MT5 and PT7b. Fossil assemblages are similar to those recorded in the underlying T34 sequence. Microfloras are generally of low diversity and abundance due to the dilution effects of abundant terrestrially derived kerogen.

Description. The earliest deposits of T36 form onlapping and downlapping packages which are restricted to the basin centre (Fig. 7 & 10a). The oldest of these packages has been named the 'Cuillin' Package and it has been mapped with a considerable degree of confidence. The Cuillin Package is penetrated by two wells and in both its thickness exceeds 200m. In 204/19-1 the package has a net/gross of 0.89 and the sand has a blocky character typical of a basinal fan. The Cuillin Package is overlain by a package of similar seismic character and geometry which has been named the 'Kintail' Package. Both the proximal and distal limits of the Kintail Package can be mapped over the Quad 204 area (Fig. 10a). The Kintail Package has not been penetrated by any wells, its proximal edge only just being intersected by the northwestern parts of licences P556 and P557 and the majority of the unit lying in unlicensed waters. In view of its seismic geometry and character being so similar to that of the Cuillin Package the Kintail fan is also expected to be sand prone. During late T36 times a large prograding shelf system built across the area, significantly advancing the position of the shelf edge basinward (Fig. 10b). This prograding package is very obvious on seismic data (Fig. 7) and its proximal limit and the position of the edge of the shelf can be mapped with a fair degree of accuracy. The proximal limit of T36 extends south of the T34shelf edge and in terms of systems tracts this unit is thought to represent part of the lowstand systems tract (the prograding lowstand wedge) and, possibly, the transgressive systems tract. No attempt to break out the individual systems

tracts has been made in this review. The deposits of Late T36 are essentially silty and represent a period of high depositional rates. It is possible that submarine fan deposits were deposited during late T36 times, fed by established transport corridors or as a result of slope failure, but these cannot be mapped on the present data.

Interpretation. The deposition of Sequence T36 represents a major event in the Palaeocene history of the Faeroe-Shetland Basin. The change in sediment provenance appears to be associated with a major lowstand event which, during the early part of the sequence, limits deposition to basinward of the T34 shelf edge (Fig. 10a). This relationship indicates that the base of the sequence is a Type I sequence boundary (sensu Vail et al. 1977). The lowstand deposits of the T36 Cuillin and Kintail fans are suggested to have been derived from the shelf area to the south, though axial sediment transport from the northeast is also likely. The distal position of the Kintail fan, relative to the Cuillin fan, suggests that the downward shift in coastal onlap was pulsed. The basin floor deposits are characterized by kerogen typical of a proximal, shelfal setting thus indicating significant transport of shallow water sediments out into a deeper water setting. There is, however, no evidence of older reworked Palaeocene microfloras. Whilst there was significant transport of penecontemporaneous shelf sediments into the basin it is unclear if the old T34 shelf has been bypassed via a canyon system or whether it was subject to active erosion. In more proximal positions there was definitely erosion of the T34 shelf, the erosion cutting deeper in a proximal direction (south and east). This is reflected in well data with T36 deposits resting on progressively older Palaeocene deposits. Heavy mineral suites clearly show that fresh metamorphic basement is being eroded, providing further evidence of a significant lowstand event. This event is suggested to be associated with thermal uplift related to the evolution of the Iceland hotspot. Tuff horizons at the top of the Cuillin fan reflect volcanic activity associated with this thermal uplift.

Sequence T40 (Fig. 10c) Definition. On seismic the lower boundary is a type I unconformity and the upper boundary is a series of high amplitude refectors interpreted as coals, with a downlap surface above. Regional correlation indicates that T40 is characterized by species of Apectodinium (Biozone PT8),

HYDROCARBONS IN FAEROE-SHETLAND BASIN

A. augustum being restricted to Sequence T40, and agglutinated foraminifera (Biozone MT6).

65

Sequence T50 (Fig. 10d) Definition. The top of Sequence T50 equates to

Description. Sequence T40 has not been penetrated by any wells in the area of interest and its description is limited to seismic character. Where T40 is observed on seismic data it almost inevitably onlaps the T36 slope deposits and internally is characterized by prograding clinoforms. In terms of systems tracts this prograding and onlapping package is believed to represent the lowstand prograding wedge and, probably, the transgressive systems tract. The T40 shelf slope break is not seen on the current data set but it appears that little accommodation volume remained at the end of T40 times. The absence of the sequence in the licensed area is noted by the absence of Biozone PT8 in any of the wells in the Quad 204 area.

Interpretation. In the North Sea Sequence T40 includes the Forties Fan, a major lowstand fan system. In the Faeroe-Shetland Basin T40 is also a major lowstand deposit, being restricted to areas basinward of the T36 shelf edge. Since the entire sequence is accommodated basinward of the previous shelf edge the base of T40 represents a type I unconformity and, as such, lowstand fan deposits are expected to occur at the base of the Sequence. None have been observed on seismic data, probably due to the poor coverage of the unlicensed areas where the sequence occurs. Any basin floor fans present are predicted to be silty assuming derivation from T36 shelf sediments.

Sequence T45 (Fig. 10d) Definition. Sequence T45, like T40, has not been penetrated by a well section in the area of interest. Seismically the lower boundary is interpreted as a maximum flooding surface, being defined by a downlap surface. The upper boundary is characterized by high amplitude reflectors suggested to represent coals. Description and interpretation. The areal extent of Sequence T45 is similar to that of the preceding T40 Sequence, extending slightly more landward as the T36 slope was onlapped during a period of relative sea level highstand. No wells penetrate the sequence but from seismic and regional evidence it is suggested that the sequence is probably similar to the overlying T50 Sequence in terms of lithology, namely delta top/coastal plain sandstones possibly capped by coal.

the top Balder event which is a prominent regional marker throughout the North Sea and West Shetland area. It is interpreted to represent a maximum flooding event and the high amplitude reflectors which characterize the event in this area are indicative of coals. The lower boundary, over much of the area, is a regional unconformity which developed during the T40 lowstand event. Beneath the unconformity reflectors are truncated, with T50 onlapping this irregular surface. Biostratigraphically T50 is characterized by an abundance of terrestrially derived kerogen and the predominance of miospores including Caryapollenites simplex and Inapaturopollenites spp. Marine fossils are rare, but the rare records of Coscinodiscus sp. 1 (MT7) Deflandrea oebisfeldensis (PT10) and Ceratiopsis wardenensis (PT9) are of stratigraphic importance.

Description. The sequence is variable in both thickness and lithology, probably a reflection of available accommodation space. Shelf sands, muds and coastal plain deposits characterize the sequence and shallowing upwards trends can be seen on wireline logs. Microfloral assemblages are dominated by miospores and other terrestrially derived material, confirming a deltaic setting. Tufts (Balder Tufts) are also recorded, and are believed to be associated with the uplift of the N o r t h Atlantic hotspot (Jacque & Thouvenin 1975). T50 is invariably capped by a coal which is believed to represent a further flooding (maximum flooding surface) event as the basin again begins to subside in earliest Eocene times as the North Atlantic hotspot (Iceland) migrates northwards. Interpretation. Sequence T50 is suggested to have been deposited during a further period of relative sea-level highstand. The T50 highstand, however, flooded back over the earlier shelf systems which were exposed during T40-45 times making the T50 Sequence extremely extensive. Thin, fining upwards, backstepping parasequences are recognized on logs and are consistent with this interpretation (Vail & Wornardt 1990). In distal positions T50 conformably overlies deposits of T45 but in the licensed acreage over the area of interest T50 overlies T36 unconformably. In the south T50 is absent. It is possible that the sequence has been subsequently removed by erosion or, more likely, that inversion on the Judd Fault complex

66

.~

C.C. EBDON E T AL.

-1000

-1000

.1500

-1500

-20o0

[]

-2000

[]

[]

[]

Ja

-2500

-2500 []

10

E

[]

[]

-8OO0

.85O0

[]

III

0

III

5

[]

II

[]

[]

[]

[]

-3500

i!

iii

[]

[]

[]

Ill

is

Ill

[]

.8000

m

[]

t,-

K

[]

III

II

II

lllll

lllll

I

10

15 20 25 Mean Porosity ( % )

30

35

40

-4000

m[]

D~ i . ,nlli i

I

l i llllllJ

I I llnllJ

1

10

, i lllnl I

l l flirt, i

100

I ! IIUl

1000

10000

Mean Permeability (Md)

Fig. 11. Palaeocene reservoir quality, Faeroe-Shetland Basin. associated with thermal uplift provided a structural depositional limit to the Sequence.

Correlation and comparison with the North Sea Until recently, little detail on the Tertiary stratigraphy of the Faeroe-Shetland Basin had been published. A large number of papers on the geology of the Faeroe-Shetland Basin were published in the 1980s (Ridd 1983; Hitchen & Ritchie 1987; Mudge & Rashid 1987) but at that time few wells from the Tertiary depocentres had been released. A paper by Mitchell et al. (1993) undertakes a review of the Palaeogene sequence stratigraphy of the Faeroe Basin, covering a much larger area than the present review. They recognize nine sequences (three being sequence sets) over the same stratigraphic interval of this review (i.e. Base Tertiary to top Balder). Their analysis is based largely on the identification and mapping of systems tracts determined by seismic geometry. They note a major unconformity developed at the base of their Sequence 50, which correlates to the base of Sequence T36 as defined herein. The correlation, confirmed biostratigraphically, is further vindicated by their recognition of four sequences (10-40) preceding this unconformity, all characterized by the development of lowstand systems tracts. These sequences are probably approximately equivalent (within the variations of sequence definition) to sequences T10, T20, T32 and T34. Furthermore their identification of sequence sets is supported by the recognition that there are additional potential sequence boundaries within

the stratigraphy erected herein. On the current data set, however, these cannot be mapped regionally. Post dating the regional 'mid' Late Palaeocene unconformity Mitchell et al. note extensive and thick lowstand deposits, which probably correlate with the Cuillin and Kintail packages identified herein. Depositional style changed to one with a ramp profile. The five sequences identified by Mitchell et al. over this Late Palaeocene-Early Eocene interval (50-90) are probably equivalent to sequences T36 (early and late), T40, T45 and T50 of this paper. Rochow (1981) evaluates the seismic stratigraphy of the North Sea 'Palaeocene' with an emphasis on seismic data. Mudge and Copestake (1992a, b) discuss the lithostratigraphic evolution using a biostratigraphic template which allows direct correlation to the FaeroeShetland Basin. BP in-house studies indicate that the sequences identified west of Shetland also occur in the North Sea Basin. The relative consequences of individual events, particularly intra T30 events, however, are less marked, probably due to lesser effects of the tectonics associated with the uplift of the Iceland hotspot (Knott et al. 1993). Nevertheless, Sequences T10-T40 are characteristically lowstand deposits and Sequences T45 and T50 are highstand deposits.

Implications for hydrocarbon exploration Detailed gross depositional environment mapping of each sequence identified has allowed a detailed understanding of the distribution of potential reservoirs and seals to be developed.

HYDROCARBONS IN FAEROE-SHETLAND BASIN

67

Fig. 12. Schematic Early Tertiary evolution of the Quad 204 area. BASET, base Tertiary; RSL, relative sea level.

Reservoirs The primary reservoirs of the Palaeocene playfairway are the submarine fan sandstones developed throughout Sequences T10-T36 (and possibly T40). Reservoir effectiveness appears to be controlled by facies, the more massively bedded sandstones having the best reservoir characteristics. Porosities in excess of 25% and permeabilities in excess of 700 Md are normal in this facies. Throughout the entire Faeroe-Shetland Basin, however, depth of burial provides the overall control on reservoir quality (Fig. 11). In the northern sub-basin a number of gas discoveries have been made within the Palaeocene but flow rates have proved disappointing due to a reduction of permeability with depth. In the Quad 204 area, however, the Palaeocene reservoirs are buried at much shallower depths (generally < 2500 m SSB) and reservoir quality has not deteriorated at these depths of burial.

Sea& It is suggested that all basinal mudstones within the earlier part of the Palaeocene have the capacity to act as effective seals to these basin floor fans, in particular the regionally developed

flooding surfaces identified at the top of sequences T10, T20 and T32 (Fig. 9). No regionally developed flooding surfaces are developed in the section of interest post T32 (with the exception of top T50). The toes of the T36 progrades may, however, provide a top seal to the Cuillin and Kintail fans, but the effectiveness of this seal is unknown. It is suggested to carry a higher risk than the regionally developed maximum flooding surfaces of the earlier Palaeocene section.

Summary of key exploration risks The understanding of the distribution of reservoir within the playfairway is considered to be relatively low risk. Similarly the distribution of effective seal is also considered to be well understood. Traps within the playfairway can be purely structural (drape over Cretaceous structures, Tertiary slides or Oligocene inversion structures), purely stratigraphic or, more likely, a combination of the two. The shallower depth of burial of the play in this southern sub-basin makes it more attractive than the Palaeocene play in the northern sub-basin where the greater depth of burial has resulted in a deterioration

68

C.C. EBDON ET AL.

of reservoir quality. The challenge to the oil industry is to understand the potential charge of this attractive play. Discussion

This review of the Early Tertiary evolution and sequence stratigraphy of the Faeroe-Shetland Basin has assisted in the evaluation of the hydrocarbon prospectivity of the Palaeocene. Early Tertiary deposition in the area can broadly be subdivided into an early phase of basinal aggradation (sequences T10-T32), with the position of the shelf edge controlled by Cretaceous structure, a later phase of progradation (T36-T40) and flooding of these shelf systems (T45 & T50). During the aggradational phase (Fig. 12a) basin floor fans, probably lowstand fans derived from Mesozoic sediments, provide excellent potential reservoir whose quality is controlled by facies. The fans progressively in-fill the relict basin floor topography. Effective seal is provided by both the encasing basinal muds and, more significantly, by regionally extensive high gamma mudstones deposited during times of maximum flooding. At the end of T32 times major slumps and slides, suggested to have been initiated by the evolving North Atlantic 'hot spot', created structuring at top T32 level in the basin. Sequence T34 progrades basinward of the T32 shelf/slope break but there is still the potential for aggradational fans in the basin and the sequence is transitional between the aggradational and progradational phases (Fig. 12b). A major fall in relative sea-level at the end of T34 results in a change of sediment provenance, drainage pattern and depositional style. This event is related to a phase of uplift associated with the development of the North Atlantic hotspot. Extensive lowstand fans, which continue to exhibit excellent reservoir potential were deposited at the beginning of T~6 (the Cuillin and Kintail fans; Fig. 12c). Potential seals to these fans are the toes of the succeeding T36 shelf system. Thin tufts are also recorded within T36 and provide further evidence for the thermal uplift. The T40 Sequence is restricted to the basin centre and data on the sequence is limited to seismic evidence. The sequence represents a further basinward shift in coastal onlap with the potential of lowstand fans in the basin (Fig. 12d). There is a risk associated with both the reservoir and seal effectiveness in T40 with sediment probably derived from the silty T36 shelf system. A period of relative sea-level rise

resulted in the deposition of shallow marine and delta top deposits during T45 and T50. During T50 the exposed T36 and older shelf systems were transgressed (Fig. 12e). The use of a fully integrated approach has allowed a degree of comparison with both the Early Tertiary of the North Sea and the work of other authors in the Faeroe-Shetland Basin. Mudge and Copestake (1992a, b) recognize a number of biostratigraphically calibrated, regionally extensive high gamma shales in the Early Tertiary of the North Sea. BP in-house work uses these to define genetic stratigraphic sequences and allow detailed facies mapping similar to that presented herein. The major change in depositional style seen West of Shetland during Sequence T30 is not as evident in the North Sea, probably as a result of the greater accommodation space available and the effects of being more distant from the North Atlantic hotspot. The detailed sedimentological, mineralogical and biostratigraphic studies which contributed to this work provide data to allow comparison with and expand on the work of Mitchell et al. (1993). Permission to publish this paper has been granted by the British Petroleum Company plc and Shell UK, London. We would also like to thank D. Lynch and D. Lawrence for important contributions in the early part of the study and B. Mitchener and A. Fraser for constructive comments throughout. Analysis of heavy mineral assemblages was undertaken by A. Morton at the BGS, Keyworth and microfaunal analysis of recent wells by Simon Petroleum Technology, Aberdeen.

References ANDERTON, R. 1993. Sedimentation and basin evolu-

tion in the Palaeogene of the northern North Sea and Faeroe-Shetland Basins. In: PARKER, J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 31. BLOW, W. H. 1979. The Cainozoic Globigerinida. Elsevier, Amsterdam BOTT, M. H. P. 1984. Deep structure and origin of the Faeroe-Shetland Channel. In: SPENCER, A. M. ET AL. (eds) Petroleum Geology of the North European Margin. Graham & Trotman, London, 341-347. DUINDAM, P. & VAN HOORN, B. 1987. Structural

evolution of the west Shetland continental margin. In: BROOKS,J. & GLENNIE,K. W. (eds) Petroleum Geology of Northwest Europe. Graham & Trotman, London, 765-773. GALLOWAY, W. E. 1989. Genetic stratigraphic sequences in basin analysis I: Architecture and genesis of flooding surface bounded depositional units. American Association of Petroleum Geologists Bulletin, 73, 125-142.

HYDROCARBONS IN FAEROE-SHETLAND BASIN HEILMANN-CLAUSEN, C. 1985. Dinoflagellate stratigraphy of the uppermost Danian to Ypresian in the Viborg 1 borehole, central Jylland, Denmark. Undersogelse Danmarks Geologiske Series A, 7, 69 pp. HITCHEN, K. & RITCHIE, J. O. 1987. Geological review of the West Shetland area In: BROOKS, J. & GLENNIE, K. W. (eds.), Petroleum Geology of NW Europe Vol. 2, Graham & Trotman, London, 737-749. JACQU~, M. & THOUVENIN, J. 1975. Lower Tertiary tufts and volcanic activity in the North Sea. In: WOODLAND, A. W. (ed.) Petroleum and the Continental Shelf of North-west Europe. Elsevier, Barking, 455-465. JOPPEN, M. & WHITE, R. S. 1990. The structure and subsidence of Rockall trough from two-ship seismic experiments. Journal of Geophysical Research, 95, 19821-19837. KNOTT, S. D., BURCHELL, M. T., JOLLEY, E. J. & FRASER, A. J. 1993. Mesozoic to Cenozoic plate reconstructions of the North Atlantic and hydrocarbon plays of the Atlantic margins. In: PARKER, J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 953-974. MITCHELL, S. M., BEAMISH, G. W. J., WOOD, M. V., MALACEK, S. J., ARMENTROUT, J. m., DAMUTH, J. E. & OLSEN, H. C. 1993. Palaeogene sequence stratigraphic framework of the Faeroe Basin. In: PARKER, J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 1011-1023. MUDGE, D. C. & COPESTAKE, P. 1992a. A revised

Lower Palaeogene lithostratigraphy for the Outer Moray Firth, North Sea. Marine and Petroleum Geology, 9, 53-69. 1992b. Lower Palaeogene stratigraphy of the northern North Sea. Marine and Petroleum Geology, 9, 287-301. & RASHID, B. 1987. The Geology of the Faeroe Basin area. In: BROOKS, J. & GLENNIE, K. W. (eds). Petroleum Geology of Northwest Europe. Graham & Trotman, London, 751-763. RIDD, M. F. 1983. Aspects of the Tertiary geology of the Faeroe-Shetland Channel. In: BOTT, M. H. P., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) Structure and Development of the GreenlandScotland Ridge. Plenum Press, New York, 133158. ROCHOW, K. A. 1981. Seismic stratigraphy of the North Sea 'Palaeocene' deposits. In: ILLING, L. V. & HOBSON, G. D. (eds) Petroleum Geology of the Continental Shelf of Northwest Europe. Heyden, London, 255-266. VAIL, P. R. & WORNARDT, W. W. 1990. Well-log seismic sequence stratigraphy: an integrated tool for the '90's. In: ARMENTROUT,J. M. & PERKINS, B. F. eds. Sequence Stratigraphy as an Exploration Tool." concepts and practices in the Gulf Coast. Gulf Coast Section SEPM Foundation Eleventh Annual Research Conference, Program and Abstracts, 379-388. & 7 OTHERS 1977. Seismic stratigraphy and global changes in sea level. In: PAVTON, C. E. (ed.) Seismic Stratigraphy - - applications to hydrocarbon exploration. American Association of Petroleum Geologists Memoir, 26, 49-212. -

-

69

-

-

The formation of passive margins: constraints from the crustal structure and segmentation of the deep Galicia margin, Spain G. B O I L L O T , 1 M. O. B E S L I E R , 1 C. M. K R A W C Z Y K , 2 D. R A P P I N 3 & T. J. R E S T O N 2

10bservatoire Ocdanologique de Villefranche, Laboratoire de Gdodynamique Sous-Marine, B.P. 48, 06230 Villefranche-Sur-Mer, France 2 GEOMAR, Forschungszentrum ffir Marine Geowissenchaften, Christian-Albrechts Universitdt, Wischhofstrasse 1-3, 2300 Kiel, Germany 3 Ecole et Observatoire de Physique du Globe de Strasbourg, URA C N R S 323, ULP, 5 rue Descartes, 67084 Strasbourg Cedex; now at Elf Aquitaine Production, Centre Scientifique et Technique, 64018 PA U Cedex, France Abstract: The crustal structure of the Mesozoic deep Galicia margin and adjacent oceancontinent boundary (OCB) was investigated by seismic reflection (including pre-stack depth migration and attenuation of seismic waves with time). The seismic data were calibrated using numerous geological samples recovered by drilling and/or by diving with submersible. The N-S trending margin and OCB are divided in two distinct segments by NE-SW synrift transverse faults locally reactivated and inverted by Cenozoic tectonics. The transverse faulting and OCB segmentation result from crustal stretching probably in a NE-SW direction during the rifting stage of the margin in early Cretaceous times. The Cenozoic tectonics are related to Iberia-Eurasia convergence in Palaeogene times (Pyrenean event). In both segments of the deep margin, the seismic crust is made of four horizontal layers: (1) two sedimentary layers corresponding to post' and syn-rift sequences, where velocity ranges from 1.9 to 3.5kms 1, and where the Q factor is low, the two sedimentary layers being separated by a strong reflector marking the break-up unconformity; (2) a faulted layer, where velocity ranges from 4.0 to 5.2 km s-1, and where the Q factor is high. This layer corresponds to the margin tilted blocks, where continental basement and lithified pre-rift sediments were sampled; (3) the lower seismic crust, where the velocity (7 km s-1 and more) and the Q factor are the highest. This layer, probably made of partly serpentinized peridotite, is roofed by a strong S-S' seismic reflector, and resting on a scattering, poorly reflective Moho. A composite model, based both on analogue modelling of lithosphere stretching and on available structural data, accounts for the present structure of the margin and OCB. Stretching and thinning of the lithosphere are accommodated by boudinage of the brittle levels (upper crust and uppermost mantle) and by simple shear in the ductile levels (lower crust and upper lithospheric mantle). Two main conjugate shear zones may account for the observations and seismic data: one (SZ1), located in the lower ductile continental crust, is synthetic to the tilting sense of the margin crustal blocks; another (SZ2), located in the ductile mantle, accounts for the deformation of mantle terranes and their final unroofing and exposure at the continental rift axis (now the OCB). The S-S' reflector is interpreted as the seismic signature of the tectonic contact between crustal terranes and mantle rocks partly transformed into serpentinite by syn-rift hydrothermal activity. It is probably related to both shear zones SZ1 and SZ2. The seismic Moho is lower within the lithosphere, at the fresh-serpentinized peridotite boundary.

Passive continental margins are the scars of the break-up of continents. Their basement underw e n t stretching before seafloor s p r e a d i n g started, and contains crucial information about timing and pressure-temperature conditions of lithospheric deformation due to extensional tectonics. Unfortunately, as passive margins are also places where subsidence was important and rapid, in general the crust is covered by a

thick sedimentary layer which prevents observation and sampling of the basement. The West Galicia margin (Fig. 1) is exceptional in that it is a starved margin, covered only by a thin and discontinuous sedimentary layer. These conditions are favourable for imaging by seismic reflection the thinned continental crust and the crustal o c e a n - c o n t i n e n t b o u n d a r y (OCB), and also to sample the basement by

From Scrutton, R. A., Stoker, M. S., Shimmield, G. B. & Tudhope, A. W. (eds), 1995, The Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region, Geological Society Special Publication No. 90, pp. 71-91

71

72

G. BOILLOT E T AL.

Fig. 1. Location of the studied area. Magnetic anomalies Mo and 31-34 from Srivastava et al. (1990). PB, Palaeocene plate boundary after Grimaud et al. (1982). G, Galicia margin; lAP, Iberia Abyssal Plain. drilling or even by diving with the French submersible 'Nautile'. For that reason, the Galicia margin has been intensively studied for 20 years (see recent synthesis in Mauffret & Montadert 1987; Sibuet et al. 1987; Boillot et al. 1988b, 1989a). However, as the OCB and the main extensive structures of the margin are trending N-S, the seismic data were generally recovered along EW lines. For that reason, transverse structures were poorly imaged, although recognized in some places (Thommeret et al. 1988). To fill this gap, in 1990 we recorded (Lusigal cruise) several N-S seismic lines on the eastern, continental side of the OCB (Fig. 2, inset), so discovering that the West Galicia margin is actually made of two distinct segments separated by a major transverse structure. Moreover, a recent study of the Iberia Abyssal Plain (Beslier et al. 1993) revealed a segmented structure for the deep margin and OCB in the area located to the south of the Galicia margin (Fig. 2). In this paper we focus on the segmentation and the crustal structure of the deep Galicia margin and adjacent OCB.

Segmentation of the ocean-continent boundary (OCB) offshore Galicia The West Galicia margin results from lithosphere stretching and rifting during lower Cretaceous times, lasting from 140 to l l 4 M a (Boillot et al. 1987b, 1988c; time scale after Kent & Gradstein 1986). Accordingly, the M0 (118 Ma) magnetic anomaly is recognized offshore Portugal, while it is missing offshore Galicia (Fig. 1) where the margin is bounded by the Cretaceous quiet magnetic zone (Srivastava et al. 1990). The segmented

ocean-continent

boundary

The OCB to the west of Galicia is marked by a basement ridge made of serpentinized peridotite. The ultramafic basement was sampled in several locations by dredging (Boillot et al. 1980; Sibuet et al. 1987), by drilling (leg ODP 103, drill Site 637; Boillot et al. 1987b, 1988r or by diving with the Nautile (Boillot et al. 1988a). It separates two areas with different seismic and structural

FORMATION OF PASSIVE MARGINS

73

Fig. 2. Structural map of the West Iberia passive margin north of 40~ after Beslier e t al. (1993) (location on Fig. 1). J anomaly from Whitmarsh et al. (1990). Structural map of the Galicia margin after Thommeret et al. (1988) and Murillas et al. (1990). R1-R4, segments of the ridge bounding the oceanic and continental domains. Bathymetry after Lallemand et al. (1985). V, Vigo seamounts. DSDP Leg 47b and ODP Leg 103 sites, and dive sites (circled numbers) where peridotite was sampled are indicated. Inset: track map of multichaunel seismic lines used in the study of the deep Galicia margin.

"/4

13000

12o40 \

\

D I V E lC ,~.0 c,P

/ 42040

42040

D I V E 0 6 LLL

fig.4

A ~

R2

42020

42 ~ 20

i!

t

~% %

Fig. 11

D I V E 0 4 \ II

~~ ~ 13~

9

~

=

~

"

- - -

Fig. 8

9 q,.j~

~L~__-i~~

12~

Fig. 3. Depth to the break-up unconformity (BU) (in s twO. Ultramafic basement, faulted enigmatic terrane or syn-rift sediment are resting beneath BU, in different places. Drill and dive sites where samples of serpentinized peridotite were recovered are indicated. F and TF, Mesozoic and Cenozoic transverse faults, respectively.

FORMATION-OF

PASSIVE MARGINS

75

(1.) .*,a , ~

9r~ "~ "~ O o ~

*'~

~o~o

"~

~~ ~

O

76

G. BOILLOT ET AL.

characters. On the oceanic side, to the west, the relatively thin sedimentary layer is related to post-rift sediments of the margin ( l l 4 M a to Present). It overlies a diffractive basement interpreted as oceanic crust of the Cretaceous quiet magnetic zone. In fact, Albian oceanic basalt was observed and sampled on the northwestern slope of the peridotite ridge (PR) (Malod et al. 1993), and infered from refraction data to the west of the Galicia margin PR (Whitmarsh et al. 1993). On the eastern continental side the PR is bounded by a sedimentary basin, 10-20 km wide, infilled by syn- and postrift sediments (Mauffret & Montadert 1987). Hereabouts, syn-rift sediments overlap the flank of the PR, indicating that this part of the ridge is a syn-rift feature. Figure 3 shows the present morphology of the OCB at the bottom of post-rift sediments. The PR is divided into two segments, R1 and R2, by the transverse faults F and TF, both imaged on Fig. 4. Clearly, TF disturbs the lower part of the post-rift sequence, while it is sealed by the upper part. The oldest deformed sediments can be correlated with Paleogene strata, according to regional seismic stratigraphy calibrated by drilling (Mauffret & Montadert 1987). Thus, TF is a Cenozoic structure, related to Pyrenean tectonics, as are many other faults in that part of the Galicia margin (Boillot et al. 1979; Mougenot et al. 1984; Malod et al. 1993). Its seismic image is confused on Fig. 4 owing to a lot of diffraction events. It is clearer on the record of residual attenuation of seismic waves (Fig. 5A), suggesting a transpressional strike-slip fault or a reverse fault, as expected for structures related to Cenozoic plate convergence. On the contrary, F is sealed by the break-up unconformity. It is clearly a syn-rift normal fault which bounds the northern segment R 1 of the peridotite ridge (Fig. 4). However, its setting parallel to TF suggests that both faults are Mesozoic structures, one of them (TF) having been reactivated and inverted by Cenozoic tectonics. The thinned continental crust o f the margin

Further east, the upper crust of the Galicia margin consists of tilted blocks, 16 km across on average and bounded westward by N-S normal faults or sets of normal faults (Montadert et al. 1979; Mauffret & Montadert, 1987; Sibuet et al. 1987; Thommeret et al. 1988). The eastward block tilting involved the formation of halfgrabens infilled by syn- and post-rift sediments. The N-S extensional structures are cut and locally shifted by NE-SW transverse faults (Fig. 2; Thommeret et al. 1988), which are possibly

transfer faults related to the margin segmentation.

Crustal structure: data and methods Our interpretation of the crustal structure of the segmented margin is based on the following analyses. Seismic velocities f r o m reflection data

First we derived interval velocities from normal moveout velocities, using the Dix equation (Dix 1955). Because of its limitations this method was applied only to the portions of the seismic lines where the seafloor and the underlying layers are close to horizontal. The second, and more timeconsuming, approach adopted was the use of iterative pre-stack depth migration. This has, to date, been applied to two E-W profiles: GP102 and GP03, one from the southern region and one from the northern (Fig. 2, inset). The method is based on depth-focusing analysis (see Reston et al. this volume), and provides more meaningful estimates of velocity in the structures. Seismic refraction

The data recently published by Whitmarsh et al. (in press), related to two seismic refraction profiles fired across the OCB of Galicia along 42 ~10'N, was used. A m p l i t u d e attenuation o f recorded waves

This method consists of measuring the amplitude variations of seismic signals with time (see details in Rappin et al. in press). The results of a simple modelling of attenuation can be represented in two different images: (1) the distribution of the quality factor Q with time and shot location (Fig. 5A). The Q factor is one of the parameters used to perform the modelling of attenuation v. time. It is related to both absorption and scattering by interfaces and terrane heterogeneities. Its value decreases in particular where heterogeneities have a size close to the wavelength (c. 100m), for example in deformed zones; (2) the residual attenuation of seismic waves with time and shot location (Fig. 5B). This is the difference between the calculated and the measured curves of seismic waves attenuation. As it estimates the most coherent amplitude of the seismic wave it does not depend upon velocity estimation nor upon correlation of signal phases. This method provides an apparent reflectivity of terranes, and gives a more

F O R M A T I O N O F PASSIVE M A R G I N S

77

Table 1. Values of the Q factor, reflectivity and seismic velocities within layers 1-4 in the northern and southern

segments of the deep Galicia margin Seismic velocity (km s-1, t w o Northern segment* Layer 1: post-rift sediments Break-up unconformity Layer 2: syn-rift sediments Layer 3: pre-rift sediments and basement S-S' seismic reflector Layer 4: serpentinized peridotite Seismic Moho Layer 5: fresh peridotite

Reflectivity*

Northern segment

Southern segment

Northern segment

Southern segment

2.15-2.66 1.9-3.0 . . . 3.0-4.5 3.5

10-11 . -

4.0-4.5

0.4-0.6 -

0.25-0.35 -

4.0-5.7 .

4.0-5.1 .

600--650 .

6004575

0.2-0.25

0.15-0.35

7.0-7.8 . 8.1

800-950 . 1000

850-950

0.1-0.15 -

0.15-0.2 -

. -

Southern segmentf

Q factor*

.

.

90-200

1000

* Results of this study. Reflectivity values are calculated with respect to a value of 1 at the sea floor. f After Hoffmann & Reston (1992) for the layers overlying S; Horsfield (1992) and Whitmarsh et al. (1993) for layers 4 and 5, respectively.

Fig.

5. (A) Values o f the attenuation factor Q along a section of the seismic line LG03. 1-2, post- and syn-rift sediments; ET, enigmatic terranes, including continental basement and lithified sediments; 4, lower seismic crust, probably made o f serpentinized peridotite hereabouts; SRM, Scattering reflective Moho. Location on Fig. 3. (B) Values o f the residual attenuation of seismic waves along the same seismic line (compare with Figs 4 & 6). Location on Fig. 3.

78

G. BOILLOT ET AL.

Fig. 6. Comparison of the crustal structure in the southern (A) and in the northern (B) segments of the deep Galicia margin. 1-2, post- and syn-rift sediments; 3, enigmatic terrane; 4, probable serpentinized peridotite; BU, break-up unconformity; S, S', S and S' seismic reflectors interpreted in the paper as the tectonic contact between serpentinized peridotite and continental crust material. The seismic line LG03 is located on Figs 3 & 9. informative image of reflectors than the classical stack and post-stack migration method (cf. Figs 4 & 5A). The main result of the combination of these different methods was to show that the crustal structure is identical in the southern and northern segments of the Galicia margin, on each side of the Cenozoic transverse fault TF and associated deformed zone. In both cases the crust consists of four layers, characterized by seismic velocities and Q factors (Table 1). Between these layers a r e surfaces of large residual attenuation (reflectors), also similar to the north and to the south of TF (Figs 6 & 7).

Upper layers of the continental crust In this section crustal layers 1-3 are described and their probable geological nature from top to bottom considered.

Layer 1: the post-rift sedimentary sequence. The high absorption of energy (Q < 100) and the low seismic velocities (1.9-2.6kms-)1 measured within layer 1 are in good agreement with the physical properties of non- or poorlyconsolidated sediments. Layer 1 was drilled in several places during DSDP Leg 47b (Sibuet et

al. 1979) and ODP Leg 103 (Boillot et al. 1987b, 1988c). It is made of distal turbidites and pelagic sediments, deposited from l l 4 M a to Present, and imaged by reflectors of good continuity but variable amplitude (Mauffret & Montadert, 1988).

Layer 2: the syn-rift sedimentary sequence The physical properties of the layer 2 (Q = 100 ; seismic velocities ranging from 2.9-3.5kms -1) are those expected in buried, compacted and progressively lithified sediments. At the drill Site 639 (Leg ODP 103), layer 2 is made of coarse, siliciclastic turbidite and sandstone and by alternating clay and marl, lower Cretaceous in age (135-114Ma). The seismic facies of layer 2 ranges from chaotic to well layered, with a divergent, fan-like configuration related to the syn-rift tilting of underlying crustal blocks. The divergent structure, however, is poorly or not imaged on NS seismic lines (Figs 4 & 6).

The post-rift or break-up (BU) unconformity This is marked by a strong reflector and a high positive residual attenuation level located between layers 1 and 2. At drill Site 641 (Leg ODP

FORMATION OF PASSIVE MARGINS = L=|

o-

10

A

SF

20 30 40 50 60 70 80 90 100 110 4

5

6

7

8

9

10

11

12

13

14 sdH

0 10

$F

20 BU

30 40 50" 60" 70 80" 90 100

79

Layer 3 was sampled in the southern part of the margin, at dive Site 11 (Fig. 8). Here, the westernmost, deepest tilted crustal block is cropping out on the seafloor, allowing the Nautile to recover on a normal fault scarp several samples of granodiorite from the Hercynian upper crust. The continental basement is covered by Mesozoic pre-rift limestones and sandstones (Boillot et al. 1988a). However, ET may also include other terranes with similar physical properties, for example volcanic rocks, although the margin is devoid of significant magnetic anomaly, or Palaeozoic, poorly metamorphozed sediments as those recovered at another site on the margin by Mamet et al. (1991). In our opinion, a correct characterization of the enigmatic terrane remains an important target. It is crucial to verify its geological nature to constrain the interpretation of the underlying S reflector (see the next section). Enigmatic terranes are thickened within a synrift graben bounded by F to the north and TF to the south (Fig. 9). Here, layer 3 was preserved from erosion, or accumulated before the end of the rifting, confirming that TF was a syn-rift structure before its Cenozoic reactivation.

110 4

5

6

7

8

9

10

1"1

12

13

14 sdtt

Fig. 7. Attenuation curves of seismic waves through the southern (A) and northern 0B) segments of the Galicia margin. Attenuation rates and reflection patterns are identical on the two curves. SF, seafloor; BU, break-up unconformity; S-S', S and S' seismic reflectors; SRM, scattering reflective Moho; vertical scale, attenuation in decibels. 103) it corresponds with a level of coarse, calcareous turbidite deposited at the top of the syn-rift sequence. In many places the BU is also an erosional or non-depositional surface (Mauffret & Montadert 1987). L a y e r 3: f a u l t e d terrane or enigmatic terrane

(ET) Within layer 3 Q ranges from 600 to 675, and the seismic velocity from 4.0 to 5.7kms -1. The seismic image of ET shows diffractive acoustic basement locally covered by horizontal to gently dipping weak reflectors. Apparent dips in fault blocks range from 40 ~ to few degrees (Hoffmann & Reston 1992). On the N-S seismic line LG 03 (Fig. 4) dipping reflections in layer 3 are also apparent just to the south of TF, and may represent pre-rift sediments tilted by movement along TF.

S--S' seismic reflector, lower seismic crust and reflective Moho S - S ' seismic reflectors Crustal layer 3 rests on a strong seismic reflector, S (de Charpal et al. 1978; Montadert et al. 1979). S was recognized at first in the southern segment of the margin. Here it is either a single, strong reflector, or a sequence of elementary reflectors, horizontal to gently dipping (Hoffmann & Reston 1992). In general, S is located at depths ranging from 0.6 to 1.5s two-way time (twt) from the top of layer 3 (Mauffret & Montadert 1987). S' is a similar reflector recently recognized in the northern segment of the margin (Boillot et al. 1992). It is characterized by the same seismic signature as S (Fig. 7), and occurs at the same structural level (Fig. 6). More precisely, amplitude analyses show that S and S' have the same signature in the attenuation curves, i.e. a sharp and significantly high amplitude peak. These observations are in good agreement with rthe occurrence at depth of an abrupt geological interface between terranes very different in nature, rather than with a progressive geological transition, the expected signature of which being a reflection zone (and not a strong reflector), and a wide (and not a sharp) peak of small amplitude

80

G. BOILLOT E T AL.

9"~

o

r,.) ,_.;

.,.

"~

FORMATION OF PASSIVE MARGINS

13"00 "--'-1

1

81

12"40 T

42* 50

42 ~ 50

o ol

Fig. 4

42* 40

q

42* 40

Fig. 12 DIVE

06

"~ / 4?..* 2"50

30

o {L "

42* 20

E) 9

15" O0

t42 20

~ F 12"40

Fig. 9. Thickness of the layer 3 (enigmatic terrane ET) (in s twt). The velocity within layer 3 ranges from 4 to 5.7 km s-1. F and TF, Mesozoic and Cenozoic transverse faults, respectively. Location on Fig. 2.

in the attenuation curves. Moreover, the attenuation variations and parameters within terranes over- and underlying S are remarkably similar to those over- and underlying S' (Figs 5 & 6). The significance of S-S' and the geological nature of related terranes remains controversial (see the discussion in the next section and in Reston et al. this volume). However, recent multichannel seismic data allowed the tentative

connection of both S and S' to the top of the serpentinite terranes sampled on the two segments of the peridotite ridge bounding the margin (Boillot et al. 1992). For the southern segment, Fig. 10 shows a section of seismic line L G 07 from the top of the PR2, where ultramafic rocks were drilled, to the axis of the sedimentary basin bounding the ridge to the east. The roof of the serpentinite body forms a strong reflector dipping northeast. From south-

82

G. BOILLOT ET AL. west to northeast it is progressively deepening and covered by post-rift sediments, by syn-rift sediments and by enigmatic terranes at the north-eastern end of the seismic section. Here the reflector is at a depth of 9.8 s twt from the sea surface, i.e. at the same level as S on GP 105 where it disappears from the record, probably owing to Mesozoic deformation along TF (Fig. 11). In our opinion, the connection between S and the top of the peridotite ridge imaged at the same level and at few kilometres distance is highly probable. In the northern segment, the top of the PR1, sampled at dive Sites 6 and 10 (Fig. 2), is clearly connected with the S' reflector on seismic line GP 03 (Fig. 12). The lower seismic crust

Beneath S and S', layer 4 appears to have uniform physical properties. The seismic facies is layered, due to discontinuous horizontal or gently dipping reflectors. The Q factor is very high ( > 800). In the core of the PR2, the seismic velocity is 7.2kms -1 (Sibuet, 1992), while it ranges from 7.3 to 7 . 8 k m s -1 in the terranes underlying S to the east of PR2 (Horsfield 1992; Whitmarsh et al. 1993), suggesting these terranes to be directly connected at depth with the OCB ridge. These data are consistent with a serpentinized peridotite nature for layer 4. It implies that S is the petrological Moho (the boundary between mantle and crust derived terranes). The seismic M o h o

This is necessarily deeper within the lithosphere. In fact, layer 4 is underlain by a thin zone (0.20.8 s twt), in which a relatively high attenuation of seismic waves was measured (Q < 100), with systematic, scattered reflectivity. From the north (R1 segment of the margin) to the south (R2), the signature of this zone is constant and typical on amplitude attenuation curves obtained from seismic signal records (Figs 5B & 7). The weakness and the scattering character of the reflectivity at the bottom of layer 4 is different from other overlying seismic reflectors. The strong attenuation implies the presence of heterogeneities whose size is c l o s e to the wavelength of the signal (100-200 m), which induces a maximum of scattering (Herraiz &

Fig. 10. Section of the migrated LG07 MCS line, where the top of the peridotite ridge is imaged and can be connected at depth with the S seismic reflector. 1, 2, 3, 4, post-rift sediments, syn-rift sediments, enigmatic terranes and serpentinized peridotite, respectively. Location on Fig. 3.

FORMATION OF PASSIVE MARGINS

83

e,q

o o o

o

U

=

i,,,,,-i

o ~2

0

o

84

G. BOILLOT E T AL.

o

o

o

t~

~

~.~ ~'~

. ,,...~

~.,=~.~ ~,'~

08 m

~ ~9

m~ e.t.~

N~

.

FORMATION OF PASSIVE MARGINS

85

Espinosa, 1986). The thickness of the reflective horizon increases considerably beneath the PR, where serpentinized peridotite was unroofed and is now free of crustal cover. To the south of the margin it is located at the level where Recq et al. (1991) placed the seismic Moho from their refraction study. All these data are consistent with the idea that the scattering zone of reflectivity is actually the reflective Moho, located at the fresh-serpentinized peridotite boundary.

does the seafloor. From regional geological studies it is known that the uplift of the northern part of the margin, including Galicia Bank, resulted from Cenozoic tectonics (Boillot et al. 1979; Mougenot et al. 1984). Accordingly, the reflective Moho is regionally uplifted with the margin, thus accounting for the strong positive gravimetric anomaly located at the northwestern edge of the margin (Lalaut et al. 1981).

Discussion

Since it was discovered beneath the Armorican and Galica margins, S has been considered as a major feature of these passive margins, probably a key for understanding the rifting processes. It was tentatively interpreted as the brittle-ductile transition within the thinned continental crust (de Charpal et al. 1978; Montadert et al. 1979) or as a major syn-rift detachment fault rooted in the lower ductile crust or in the mantle (Wernicke & Burchfield 1982; Boillot et al. 1987; Le Pichon & Barbier 1987; Mauffret & Montadert 1987; Sibuet 1992; Hoffmann & Reston 1992). Another interpretation was recently proposed by Beslier & Brun (1991), who relate S to the development of two conjugate shear zones, located in the ductile crust and in the deeper ductile lithospheric mantle respectively. It is clear that a good understanding of the nature of S would improve our understanding of the processes of crustal extension leading to the creation of the margin. Sampling the rocks located at the level of, upon and under S is the most efficient way to progress in this discussion. It is of most interest to verify that at first the reflector corresponds to the boundary between crustal and mantle terranes, as proposed in this paper, and to further specify the kinematics of the deformation at the base of the crust and at the top of the mantle. Unfortunately, until now, available petro-structural data come only from mantle rocks recovered in the area where they were unroofed by syn-rift tectonics. In that area both crustal terranes and their contact with mantle material are missing. However, the petrology and fabric of the exposed rocks provide information on the deformation they underwent in the vicinity of the crust-mantle boundary before they were unroofed, at least if they were preserved from superficial erosion since they were exposed on the seafloor. By chance it seems that this was the case at drill Site 637 (Leg ODP 103). Petro-structural studies of cored samples constrained the timing and evolution of the peridotite up to its serpentinization, particularly the pressure-temperature conditions and the

The scattering reflective M o h o ( S R M )

The fresh serpentinized peridotite boundary is also the palaeohydrothermal front, i.e. the surface where syn-rift hydrothermal serpentinization of mantle rocks stopped after seafloor spreading started and margin lithosphere cooled. The depth of the reflective Moho (7-8km beneath the seafloor) is a clue for estimating the maximum depth of hydrothermal circulation during rifting within upper, brittle lithosphere covered by sea water. In areas where the continental crust thinned down to 7-8 km the syn-rift hydrothermal circulation probably reached the petrological Moho and the uppermost part of the mantle through the faulted and stretched basement of the margin. Boillot et al. (1989b) proposed naming this process 'undercrusting': it involves accretion, at the base of highly stretched continental crust, of a layer of serpentinite which belongs to the lower seismic crust by its physical properties, although made of mantle-derived rocks. In the case of the Galicia margin the hypothesis is supported by the occurence of layer 4 along the entire margin with the same physical properties, and its connection with the OCB ridge where serpentinized peridotite was sampled in several places. It is confirmed by the continuity of the SRM which underlies layer 4 in the deep margin and the PR as well. Note that beneath thick continental crust the Moho has a very different signature; around France for example, it is marked by a strong, continuous reflection at the bottom of scattering lower crust (Rappin 1992). In that case, it probably bounds terranes very different in nature, while the high attenuation and scattering reflectivity at the base of the layer 4 of the Galicia margin is in agreement with a transition zone containing both serpentinized and fresh peridotite. The SRM seems to be not or poorly affected by Cenozoic deformation beneath TF (Fig. 5). However, it shallows regularly northward as

The S seismic reflector

86

G. BOILLOT ET AL.

kinematics of the ductile deformation (Agrinier et al. 1988; Frraud et al. 1988; Girardeau et al. 1988; Beslier et al. 1990). The results show that the evolution of the rocks is compatible with a progressive uplift beneath a continental rift, and moreover that ductile simple shear played a major part in the stretching of the lithosphere before it broke up at l l4Ma. After partial melting under asthenospheric conditions the plagioclase-bearing peridotites experienced intense ductile deformation under lithospheric conditions within a normal shear zone gently dipping toward the continent. From these data it is concluded that a shear zone was actually drilled at ODP Site 637. The kinematics and timing of the deformation are compatible with models of passive margin formation involving simple shear, either along a single normal synrift detachment fault rooted in the mantle (Boillot et al. 1987a), or in a ductile shear zone within the mantle (Beslier & Brun, 1991; Brun & Beslier in press). The Galicia margin therefore provides a unique opportunity to study, in situ, the shearing of the upper mantle beneath a continental rift. However, the relationship between the drilled shear zone and the S reflector remains questionable (see the next section). In the previous section, it was stressed that the S' reflector and the surrounding terranes forming the northern segment of the Galicia margin have the same physical properties as S and layers 1-4 in the southern segment. In both cases, layer 4, roofed by S-S', is probably made, at least partly, of serpentinized peridotite sampled by drilling and by diving. Moreover, the crustal structure of the OCB and adjacent margin seems to be similar in the Iberia Abyssal Plain (IAP) further south. Here, the OCB is marked by a basement ridge (Fig. 2) with seismic characteristics and tectonic setting comparable to those of the Galicia margin PR. The deepest tilted block of the IAP margin is underlain by a strong seismic reflector S" similar to S and S' (Beslier et al. 1993), and S" is located at the level where Whitmarsh et al. (1990) placed the boundary between upper and lower seismic crust, with respective seismic velocities of 6.2 and 7 km s-1 or more. Such a similarity of the seismic images and seismic velocities with those recorded and measured on the Galicia margin strongly suggests that the terranes resting at the base of the IAP thinned continental crust are also made of serpentinized peridotite (Beslier et al. 1993; Whitmarsh et al. 1993). The crustal structure is thus very similar in the different segments of the west Iberia margin and adjacent OCB. The ubiquitous occurrence

of a strong reflector (S, S' or S") at the same structural level reinforces the interest to investigate it and surrounding terranes.

Fig. 13. Analogue modelling of the lithosphere stretching. Brittle continental crust (1) and brittle mantle (3) are modelled by sand, ductile continental crust (2) and ductile lithospheric upper mantle (4) by silicone putties, asthenosphere (5) by golden syrup. Senses of shear are inferred from the deformation of passive markers included in brittle and ductile layers. (A) Model 1, moderately stretched (shear zones in dark grey); (B) model 2, highly stretched. SZ1 and SZ2, shear zones discussed in the text. From Beslier & Brun (1991). Conceptual m o d e l f o r the f o r m a t i o n o f the Galicia passive margin

To build up this conceptual model, we started from two different sets of data. Firstly, analogue modelling of the lithosphere stretching was us~ed (description of models and their inferences for passive margins formation, including Galicia margin, are detailed in Beslier 1991; Beslier & Brun 1991; Brun & Beslier in press). This takes into account the differential mechanical behaviour of the main lithospheric rheological layers, and allows the study of the stretching mechanisms of the lithosphere and the related crustal thinning processes during a rifting episode (Faug~re & Brun, 1984; Vendeville et al. 1987; Allemand et al. 1989; Allemand & Brun 1991; Beslier & Brun 1991; Brun & Beslier in press). The rheological structure of the

FORMATION OF PASSIVE MARGINS lithosphere is simplified in the experiments by a brittle-ductile layered model, which is a good approximation of the structure of a stable continental lithosphere with a normal geothermal gradient (e.g. Ranalli & Murphy 1987; Davy & Cobbold 1991). The experimental analogue materials are dry sand for brittle behaviour and silicone putties for ductile behaviour. The four layers represent the upper brittle crust (sand), the lower ductile crust (silicone putty), the uppermost brittle mantle and the upper lithospheric ductile mantle. This tithospheric structure lays upon golden syrup which simulates the asthenospheric behaviour. The model is submitted to localized horizontal extension (see Beslier 1991; Davy & Cobbold 1991 for details on the method). Figure 13A shows a crosssection of a model at the end of the experiment. Brittle layers (upper crust and uppermost mantle) underwent boudinage, while ductile layers accommodated the boudinage by simple shear along conjugate normal shear zones. With progressive extension, the rupture of the brittle mantle is achieved in one of the necked zones, where the deformation is localized at depth. In the last stage of the rifting two main shear zones accommodate the stretching beneath the rift, the right part of which accounts for the west Galicia margin structure (Fig. 13A): one (SZ1), located in the lower ductile crust, acts with a top-tothe-west sense of shear, and accounts for the continentalward sense of blocks tilting sense in the upper brittle crust; the other one (SZ2), connecting the ductile crust and the ductile mantle in the ruptured zone of the brittle mantle, acts with an opposite sense, i.e. top-to-the-east. Another model (Fig. 13B), identical though extremely thinned, shows that the lower ductile crust can disappear at the rift axis, bringing the mantle directly in contact with the upper crustal blocks, or even with the syn-rift sediments as observed on the Galicia margin. Secondly, structural data from the Galicia margin have been used. We have already discussed the kinematics of the peridotite ductile deformation established from petro-structural studies of samples recovered on the PR: the mantle rocks have been deformed in a normal ductile shear zone dipping northeast, with a topto-the-east shear sense. Thus, the results of the analogue models are in good agreement with the available structural data from the margin. Two major shear zones are postulated, SZ1 in the ductile crust, and SZ2 rooted in the ductile lithospheric mantle (Figs 13A & 14A). Currently, only SZ2 has been sampled by drilling and diving on the peridotite ridge, and BoiUot et al. (1987a, b, 1988b, 1989b), developing the

87

initial model of Wernicke (1985), considered it as part of a detachment fault rooted in the mantle. However, the sense of blocks tilting on the margin also implies the existence of SZ1 (Faugrre & Brun 1984; Brun & Beslier in press). From this point in the discussion two interpretations are possible, although closely related. Reston et al. (this volume) propose S to be related to SZ 1, and the tectonic unroofing of PR to SZ2. Some of the current authors (G.B., M.O.B., D.R.) rather believe that the tectonic contact between upper crust material (the tilted blocks of the margin) and mantle material (the serpentinized peridotite) relates to both shear zones (Beslier & Brun 1991). This interpretation is supported by the highly stretched experimental model (Fig. 13B), which accounts for the contact between upper crust and upper mantle terrane. In that case, which corresponds to the final stage of the rifting, the lower crust is extremely thinned and even lost in the vicinity of the rift axis, at the places where the upper brittle mantle is broken (Fig. 13B). As a result, the upper, brittle crust lies either directly over mantle terranes deformed in SZ2 or ~on interbedded lenses of sheared lower crust initially belonging to SZ 1. Therefore, we suspect tectonic melange of various thickness and including sheets of lower crust to be related to the S reflector beneath the deeper crustal tilted blocks (Fig. 14B). Moreover, the serpentinization of mantle rocks beneath the very stretched crust at the rift axis does change the rheological behaviour of mantle terranes. It is thus possible that decollement of tilted blocks occurs at the top of the upper mantle in the final stage of the rifting (Beslier & Brun 1991). Figure 14C summarizes the crustal structure of the deep Galicia margin from this study. I n the deepest part of the margin, thin blocks of upper continental crust are scattered on 'undercrusted' serpentinite, S being the tectonic contact between the blocks and the serpentinized uppermost mantle. However, intercalation between 9these terranes of sheared lower continental crust is possible locally. Deeper is the fresh-serpentinized peridotite boundary, i.e. the actual seismic Moho. To the east, where the continental crust thickens, 'true' lower continental crust is expected in place of the serpentinite layer. The nature of the lateral transition between the two kinds of lower seismic crust (serpentinized peridotite in the most stretched area; lower continental crust elsewhere) remains a target for further clarification and discussion. Nevertheless, the disappearance of S beneath the tilted blocks of the eastern, upper part of the margin may be a consequence of the transition.

G. BOILLOT ET AL.

88

!_/

J

7

~

/

~---~--___~~,_~_..~

~-

~

>A 9

/

/

Brittle Crust

LOWer

SZ1

.

"'~CtileCrust ~

1- CONTINENTAL RIFT

~

" - 3Z2

UPPer Brit~s~. M U-- ~ - - "Mantle PPer Ductile M a n t l e " i

PR

Syn-rift sediments

B Serpentinisation of Smea,~t~g~nicSk~tii~176

PR

1oti

S reflector

. . . . . . . . .

~ ~ , - , / J t ,.~ , = .

Post-rift sediments ~

Serpentinite

M

"

~

UC M

I

2O km

I

LC

~" UM

Fig. 14. Conceptual model [(A) and (B)] and schematic cross-section [(C)] of the deep passive Galicia margin. The conceptual model is derived from the analogue modelling depicted in Fig. 13. It describes two representative stages of margin rifting: during continental rifting, and just before seafloor spreading starts (after Beslier& Brun 1991; Brun & Beslier in press). (C) is established along 42~

Plate kinematic implications o f margin segmentation

Beslier et al. (1993) suggested that the segmentation of the ridge marking the OCB results from discontinuous, northward propagation of continental break-up and opening of the North Atlantic in early Cretaceous time. Moreover, the SW-NE transverse fault offsetting R1 and R2 (Fig. 3) and R3-R4 (Fig. 2) segments can be interpreted as transfer faults indicating the direction of lithosphere stretching during rifting. In the southern segment of the Galicia margin, SW-NE is also the direction of shearing in mantle rocks recovered on the PR (Girardeau et al. 1988; Beslier et al. 1990) and the main direction of transverse faults (Thommeret et al. 1988). Thus, we suspect that the lithosphere underwent stretching along the present SW-NE direction during rifting, with consequences for the correct location and identification of the

conjugate margin in the Newfoundland Basin:

Summary and conclusions (1) The seismic crust of the deep Galicia Margin is made of four main layers characterized by their seismic facies, seismic velocities, and attenuation of reflected P waves and related reflectivity. Sediments (layers 1 and 2) are classically divided into syn- and post-rift sequences, separated by the post-rift or break-up (BU) unconformity. A reflector of high reflectivity emphasizes the BU. Within the sediments, the attenuation of seismic waves is high, the seismic image is layered, and the velocity ranges from 2.2 to 3.5 km s-1. Faulted layer 3 (enigmatic terrane) rests on the reflector S-S', another level of high reflectivity. Within ET the velocity ranges from 4 to 5.7 km s-1, and the attenuation of seismic waves is moderate and constant at places where no Cenozoic tectonics occurred.

FORMATION OF PASSIVE MARGINS The layer includes continental basement and lithified pre-rift sediments sampled from the thinned, upper crust of the margin. The lower seismic crust (layer 4) is characterized by low attenuation of seismic waves, high seismic velocities (7.0 k m s -1 or more), and rests on a scattering, poorly reflective Moho located at the level of the refractive Moho. Correlation with terranes sampled at the OCB shows that the main component of this layer is serpentinized peridotite, resulting from syn-rift hydrothermal alteration of the uppermost mantle. Layer 4 rests on fresh peridotite of the upper mantle at the seismic Moho. (2) The Galicia OCB is marked by a basement ridge made of serpentinized peridotite. It corresponds to the early Cretaceous rift axis, and results from tectonic unroofing of the mantle terranes in the latest stage of margin rifting. Mantle rocks were ductily sheared under lithospheric conditions before their serpentinization and brittle deformation, in agreement with lithosphere stretching models involving simple shear. Two syn-rift shear zones at least are necessary to account for the present margin structure: one (SZ1) located in the lower ductile crust, synthetic to the continentalward tilting of crustal blocks; another (SZ2) rooted in the ductile mantle, a part of which is now exposed beneath sediments at the OCB (PR). (3) The OCB and adjacent margin are divided into two segments by a transverse Mesozoic structure that was partly reactivated and inverted in Cenozoic times. The N E - S W orientation of this transverse structure suggests that the stretching of the lithosphere occurred in early Cretaceous time along the N E - S W direction. (4) The S-S' reflector is considered to be the seismic signature of the contact beween crustal material of the margin and underlying serpentinized peridotite. The actual nature of this contact constitutes an i m p o r t a n t target for further research. We propose that it is related to the extreme thinning of the lithosphere in the vicinity of the rift axis (now the deepest part of the margin), a place where the extension tends to localize in the latest stage of the rifting. Here, the top of the mantle deformed in SZ2 is put into contact either with the ductile crust deformed in SZ1 or with the base of the upper crust. Thus, we believe S-S' to be a major tectonic contact between crustal and mantle material, related to both shear zones SZ1 and SZ2. (5) We conclude that it is of most interest to sample by drilling the terranes located above, at the level of and beneath S-S'. From petro-structural studies of cored samples it is expected that the timing, kinematics and temperature-pressure conditions of the deformation will be investigated; and further, to constrain

89

better the mechanisms of lithosphere stretching and passive margins formation. We thank Captain J. C. Delmas and the crew of the O. V. Vessel 'Le Suroit'; J. Herv6ou and the technical team responsible for the acquisition of multichannel seismic reflection data during the Lusigal cruise (1990); IFREMER and GENAVIR for technical and financial support. The Institut Fran~ais du P&role, who provided MCS data (GP03 seismic line); CNRS-INSU (IST Program) for financial support of the MCS data processing; Institut de Physique du Globe de Strasbourg (France), and GEOMAR in Kiel (Germany), where MCS data were processed. R. Scrutton helped us to improve the English in the manuscript. Contribution no. 609 of the Groupe d'Etude de la Marge Continentale et de l'Oc6an (GEMCO), URA 718 du CNRS et de l'Universit6 P. et M. Curie (Observatoire Oc6anologique de Villefranche-SurMer).

References

AGRINIER, P., MEVEL, C. & GIRARDEAU, J. 1988. Hydrothermal alteration of the peridotites cored at the ocean-continent boundary of the Iberian margin: petrologic and stable isotope evidence. In: BOILLOT, G., WINTERER, E. L., er AL. (eds) Proceedings ODP, Science Results, 103, College Station, TX (Ocean Drilling Program), 225-233. ALLEMAND, P. & BRUN, J. P. 1991. Width of continental rifts and rheological layering of the lithosphere. Tectonophysics, 188, 63-69. --, DAVY, P. & VAN DEN DRIESSCHE, J. 1989. Sym&rie et asym&rie des rifts et m6canismes d'amincissement de la lithosph6re. Bulletin de la Soci~t~ G~ologique de France, 8, 445-451. BESLIER, M. O. 1991. Formation des marges passives et remont6e du manteau: mod61isation exp6rimentale et exemple de la marge de la Galice. M(m. Docum. Centre Arm. Et. Struct. Socles, Rennes, 45.

& BRUN, J. P. 1991. Boudinage de la lithosph6re et formation des marges passives. Comptes Rendus de l'Acad~mie des Sciences, Paris, 313, 951-958. - - , ASK, M. & BOILLOT, G. 1993. Oceancontinent boundary in the Iberia Abyssal Plain from multichannel seismic data. Tectonophysics, 218, 383-393. , GIRARDEAtJ, J. & BOILLOT, G. 1990. Kinematics of peridotite emplacement during North Atlantic continental rifting, Galicia, NW Spain. Tectonophysics, 184, 321-343. BOILLOT, G., BESLIER, M. O. & COMAS, M. 1992. Seismic image of undercrusted serpentinite beneath a rifted margin. Terra Nova, 4, 25-33. , GIRARDEAU,J. & KORNPROBST,J. 1988b. The rifting of the Galicia margin: crustal thinning and emplacement of mantle rocks on the seafloor. In: BOILLOT, G., WINTERER, E. L., ET AL. (eds) Proceedings ODP, Science Results, 103, College Station, TX (Ocean Drilling Program), 741-756. -

-

90 --,

G. BOILLOT E T AL.

FI~RAUD, G., RECQ, M. & GIRARDEAU, J. 1989b. 'Undercrusting' by serpentinite beneath rifted margins: the example of the west Galicia margin (Spain). Nature, 341, 523-525. --, MOUGENOT, D., GIRARDEAU,J. & WINTERER, E. L. 1989a. Rifting processes on the west Galicia margin, Spain. In: TANKARD, A. J. & BALKWILL, H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins, American Association of Petroleum Geologists Memoir, 46, 363377. --, AUXII~TRE,J. L., DUNAND, J. P., DUPEUBLE, P. A. & MAUFFRET, A. 1979. The northwestern Iberian Margin: a Cretaceous passive margin deformed during Eocene. In: TALWANI, M., HAYET, W. & RYAN, W. B. F. (eds) Deep Drilling Results in the Atlantic Ocean: Continental Margins and Paleoenvironment. Maurice Ewings Series 3, Washington DC, American Geophysics Union, 138-153. --, GRIMAUD, S., MAUFFRET, A., MOUGENOT, O., KORNPROBST, J., MERGOIL-DANIEL, J. & TORRENT, G. 1980. Ocean-continent boundary off the Iberian margin: a serpentinite diapir west of the Galicia Bank. Earth & Planetary Science Letters, 48, 23-34. --, COMAS, M. C., GIRARDEAU, J. ETAL. 1988a. Preliminary results of the Galinaute cruise: dives of the submersible Nautile on the western Galicia margin, Spain. In: BOILLOT, G., WINTERER, E. L., ET AL. Proceedings ODP, Science Results, 103, College Station, TX (Ocean Drilling Program), 37-51. --, RECQ, M., WlNTERER, E. L. ET AL. 1987a. Tectonic denudation of the upper mantle along passive margins: a model based on drilling results (ODP leg 103, western Galicia margin, Spain). Tectonophysics, 132, 335-342. --, WINTERER, E. L., MEYER, A. W. e~AL. 1987b. Proceedings, Initial Reports (Pt A), ODP, 103. College Station, TX (Ocean Drilling Program). er AL. 1988c. Proceedings ODP,-- Science Results, 103. College Station, TX (Ocean Drilling Program). BRUN, J. P. & BESLIER, M. O. (in press). Mantle exhumation at passive margin. Earth and Planetary Science Letters. DE CHARPAL, O., GUENNOC, P., MONTADERT, L. & ROBERTS, D. G. 1978. Rifting, crustal attenuation and subsidence in the Bay of Biscay. Nature, 275, 706-711. DAVY, P. & COBBOLD, P. 1991. Experiments on shortening of a 4-layer model of the continental lithosphere. Tectonophysics, 188, 1-25. DIX, C.H. 1955. Seismic velocities from surface measurements. Geophysics, 20, 68-86. FAUGERE, E. & BRUN, J. P. 1984. Modrlisation exprrimentale de la distension continentale. Comptes Rendus de l'Acaddmie des Sciences, Paris, 299, 365-370. FERAUD, G., GIRARDEAU, J., BESLIER, M. O. & BOILLOT, G. 1988. Datation 39Ar/40Ar de la mise en place des prridotites bordant la marge de la Galice (Espagne). Comptes Rendus de l'Acaddmie

des Sciences, Paris, 307, 49-55. GIRARDEAU, J., EVANS, C. A. & BESLIER, M. O. 1988. Structural analysis of plagioclase-bearing peridotites emplaced at the end of continental rifting: hole 637A, ODP leg 103 on the Galicia margin. In" BOILLOT, G., WINTERER, E. L., ET AL. (eds) Proceedings ODP, Science Results, 103, College Station, TX (Ocean Drilling Program), 209-223. GRIMAUD, S., BOILLOT,G., COLLETTE,B., MAUFFRET, A., MILES, P. R. & ROBERTS, D. B. 1982. Western extension of the Iberian-European plate boundary during the early Cenozoic (Pyrenean) convergence: a new model. Marine Geology, 45, 63-77. HERRAIZ, M. & ESPINOSA, A. F. 1986. Scattering and attenuation of high-frequency seismic waves: development of the theory of coda waves. Open File Report, 86-455, US Geological Survey, 1-92. HOFFMANN, H. J. & RESTON, T. J. 1992. The nature of the S reflector beneath the Galicia Bank rifted margin. Preliminary results from pre-stack depth migration. Geology, 20, 1091-1094. HORSEFIELD, S. J. 1992. Crustal structure across the continent-ocean boundary. PhD Thesis, University of Cambridge, UK. KENT, D. V. & GRADSTEIN, F. M. 1986. A Jurassic to Recent chronology in the western North Atlantic region. In: VOGT, P. R. & TUCHOLKE, B. E. (eds) Geology of North America, Geological Society of America, Boulder, CO, vol. M, 45-50. LALAUT, P., SIBUET, J. C. & WILLIAMS, C. A. 1981. Prrsentation d'une carte gravim&rique de l'Atlantique du nord-est. Comptes Rendus de l'Acad~mie des Sciences, Paris, D, 292, 597-600. LALLEMAND, S., MAZI~, J. P., MONTI, S. & SIBUET, J. C. 1985. Prrsentation d'une carte bathym&rique de l'Atlantique Nord-Est. Comptes Rendus de l'Acad~mie Sciences, Paris, 300, 145-149. LE PICHON, X. & BARBIER, F. 1987. Passive margin formation by low-angle faulting within the upper crust: the northern Bay of Biscay margin. Tectonics, 6, 133-150. MALOD, J. A., MURILLAS, J., KORNPROBST, J. & BOILLOT, G. 1993. Oceanic lithosphere at the edge of a Cenozoic active continental margin (northwest slope of Galicia Bank, Spain). Tectonophysics, 221. MAMET, B., COMAS, M. C. & BOILLOT, G. 1991. Late Palezoic basin on the west Galicia Atlantic margin. Geology, 19, 738-741. MAUFFRET, A. & MONTADERT, L. 1987. Rift tectonics on the passive continental margin off Galicia (Spain). Marine Petroleum Geology, 40, 49-70. & 1988. Seismic stratigraphy off Galicia. In." BOILLOT, G., WINTERER, E. L. ET A/~. (eds) Proceedings ODP, Science Results, 103. College Station, TX (Ocean Drilling Program), 13-30. MONTADERT, L., DE CHARPAL,O., ROBERTS, D. G., GUENNOC, P. & SIBUET, J. C. 1979. Northeast Atlantic passive continental margins: rifting and subsidence processes. In: TALWANI,M., HAY, W. & RYAN, W. B. F. (eds) Deep Drilling Results in the Atlantic Ocean: Continental Margins and Paleoenvironments. American Geophysical Union,

FORMATION OF PASSIVE MARGINS Maurice Ewing Series, 3, 154-186. MOUGENOT, D., KIDD, R. B., MAUFFRET, A., REGNAULD, n., ROTHWELL, R. G. & VANNEY, J. R. 1984. Geological interpretation of combined Sea-Beam, Gloria, and seismic data from Porto and Vigo Seamounts, Iberian continental margin. Marine Geophysics Research, 6, 329-363. MURILLAS, J., MOUGENOT, D., BOILLOT, G., COMAS, M. C., BANDA, E. & MAUFFRET, m. 1990. Structure and evolution of the Galicia interior basin (Atlantic western Iberian continental margin). Tectonophysics, 184, 297-319. RANALLI, G. & MURPHY, D. C. 1987. Rheological stratification of the lithosphere. Tectonophysics, 132, 281-295. RAPPIN, D. 1992. Apport des analyses d'amplitude et temps-fr6quence /t l'exploitation de donn6es de sismique profonde. Th6se de l'Universit6 Louis Pasteur de Strasbourg. , MARTHELOT, J. M., DE BAZELAIRE, E. & RAVAT, J. (in press). Analysis of the attenuation of amplitudes on records of the ECORS Pyrenees deep seismic profile. Geophysical Prospecting. RECQ, M., WHITMARSH, R. B. & SIBUET, J. C. 1991. Anatomy of a lherzolitic ridge, Galicia margin. Terra Abstract, 3, 122. RESTON, T. J', KRAWCZYK,C. M. & HOFFMANN, H. J. 1995. Detachment tectonics during Atlantic rifting: analysis and interpretation of the S reflector, the west Galicia margin. This volume. SmUET, J. C. 1992. New constraints on the formation of the non-volcanic continental Galicia-Flemish Cap conjugate margins. Journal of the Geological Society, London, 149, 829-840. , MAZE, J. P., AMORTILA, P. & LE PICHON, X. 1987. Physiography and structure of the western Iberian continental margin off Galicia from SeaBeam and seismic data. In: BOILLOT, G., WIN-

91

TERER, E. L., MEYER, A. W. Er AL. (eds) Proceedings Initial Reports (A). ODP, 103, 77-97. , RYAN, W. B. F. ETAL. 1979. Initial Reports of the Deep Sea Drilling Project, 47. US Government Printing Office, Washington, DC. SRIVASTAVA, S. P., ROEST, W. R., KOVACS, L. C., OAKEY, G., L/~VESQUE, S., VERHOEF, J. & MACNAS, R. 1990. Motion of Iberia since the Late Jurassic: Results from detailed aeromagnetic measurements in the Newfoundland Basin. Tectonophysics, 184, 229-260. THOMMERET, M., BOILLOT, G. & SIBUET, J. C. 1988. Structural map of the Galicia margin. In: BOILLOT, G., WINTERER, E. L. ET AL. (eds) Proceedings ODP, Science Results, 103. College Station, TX (Ocean Drilling Program), 31-36. VENDEVILLE, B., COBBOLD,P. R., DAVY, P., BRUN, J. P. & CHOUKROUNE,P. 1987. Physical models of extensional tectonics at various scales. In: COWARD, M. P., DEWEY,J. F. & HANCOCK,P. L. (eds) Continental extensional tectonics. Geological Society, London, Special Publication, 28, 95-107. WERNICKE, B. 1985. Uniform-sense normal simple shear of the continental lithosphere. Canadian Journal of Earth Sciences, 22, 108-125. & BURCHFIELD, B. C. 1982. Modes of extensional tectonics. Journal of Structural Geology, 4, 105-115. WHITMARSH, R. B., MILES, P. R. & MAUFFRET, A. 1990. The ocean-continent boundary off western continental margin of Iberia - I. Crustal structure at 40~ Geophysics Journal International, 509531. , PINHEIRO, L. M., MILES, P. R., RECQ, M. & SmUET, J. C. 1993. Thin crust at the western Iberia ocean-continent transition and ophiolites. Tectonics, 12, 1230-1239.

Detachment tectonics during Atlantic rifting: analysis and interpretation of the S reflection, the west Galicia margin T. J. R E S T O N ,

C. M. K R A W C Z Y K

& H.-J. H O F F M A N N

Geomar, Chr•tian Albrechts University, Kiel, Germany Abstract: Beneath the tilted fault blocks of the western Galicia rifted margin an unusually bright reflection, the S reflection, is observed. The waveform, polarity and amplitude of S indicate that it is a reflection from a seismic interface across which the acoustic impedance increases sharply. This result is consistent with its interpretation as a detachment fault juxtaposing a low velocity and density upper plate and a high velocity and density lower plate. The lower plate may represent partially serpentinized mantle material, brought into contact with pre-rift sediments and upper crustal basement by tectonic denudation. Pre-stack depth migration is applied to determine the true geometry of S, and its relationships with the overlying faults. It is found that S passes continuously beneath the upper crustal faults, which detach onto S. However, S does appear to be truncated westwards by east-dipping reflections associated with the peridotites exposed at the seafloor. We interpret these reflections as the continuation of the top-to-the-east extensional shear zone sampled within the peridotite, and suggest that S is either antithetic to a master mantle detachment, or that S is cut by a later mantle shear zone.

Rifted margins have been interpreted (e.g. Le Pichon & Sibuet 1981; Wernicke & Burchfiel 1982) in terms of various models for lithospheric extension, ranging from pure shear (McKenzie 1978) to simple shear (Wernicke 1981). Low angle detachment faulting is commonly associated with the simple shear model, but also features in composite pure shear/simple shear models (e.g. Lister et al. 1986), and may play an important role during continental break-up, as suggested by the inherent asymmetry of many passive margins (e.g. Lister et al. 1986; Wernicke

& Tilke 1988). Lister et al. (1986, 1991) thus introduced the concept of upper plate and lower plate margins: 'upper-plate' margins being characterized by relatively thick crust, widely-spaced faulting, relatively little crustal thinning and a generally abrupt transition to oceanic crust; 'lower-plate' margins in contrast being characterized by numerous small fault blocks, highly thinned crust over a wide region, and the presence of a detachment fault in the crust, dipping towards the ocean. The Galicia rifted margin (Fig. 1) exhibits

Fig. 1. Bathymetric map (in m) (adapted from Winterer et al. 1988) of the Galicia Banks continental margin. Portions of IFP profiles shown here are marked, as are the location of ODP boreholes from Leg 103.

From Scrutton, R. A., Stoker, M. S., Shimmield, G. B. & Tudhope, A. W. (eds), 1995, The Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region, Geological Society Special Publication No. 90, pp. 93-109

93

94

T.J. RESTON E T AL.

Fig. 2. Portion of profile GP12, showing tilted fault blocks beneath post-rift sequence. Block-bounding faults can be traced down to an underlying bright reflection, the S reflection, appearing here on the time section as a continuous but undulating single bright event. Note that no coherent reflections can be seen beneath S, where the Moho might be expected. many of the characteristics of a lower plate margin: the crust is extremely thin and the numerous high quality multichannel seismic profiles across the margin image numerous small fault blocks, tilted towards the continent by extensional faulting. The extensional faults consistently dip to the west and appear to detach downwards into the most fundamental structure of the Galicia Banks margin, the S reflector (Fig. 2). This is clearly imaged off Galicia as a single high amplitude reflection, locally continuous over > 20 km. However, the simple lower plate margin interpretation is complicated by the presence of a major landward-dipping shear zone within serpentinized peridotites exposed at the continent-ocean transition, suggesting that the margin may actually represent the upper plate to the master detachment fault (Boillot et al. 1988b). Thus, the Galicia margin does not fit simply into any one model for lithospheric extension. Indeed, the Galicia margin has been interpreted in terms of both pure shear and simple shear, and a variety of composite models (Fig. 3). Despite this, all interpretations have one aspect in common: all recognize that the S reflector, imaged as a bright, continuous reflection, is a critical structure, and the key to understanding the evolution of the margin. Consequently, the S reflector has been inter-

preted in a variety of different ways within the framework of the variety of different extensional models. For instance, de Charpal et al. (1978) interpreted the S reflector as representing an intracrustal transition from brittle faulting above to ductile flow beneath, in an essentially pure shear model for lithospheric extension (Fig. 3A). In contrast, Wernicke & Burchfiel (1982) suggested that the S reflector was a low-angle detachment fault separating the faulted rocks of the upper plate from a largely undeformed lower plate, in a simple shear model for lithospheric extension. Boillot et al. (1988a) interpreted the feature as an eastward dipping detachment fault (Fig. 3B), thinning the lithosphere by simple shear and exposing mantle rocks at the seafloor at the so-called Peridotite Ridge to the west of the margin, so that the detachment (S) effectively represents a boundary between crustal rocks in the upper plate and mantle in the footwall. Winterer et al. (1988) adopted a similar interpretation, except that in their model the detachment dipped to the west (Fig. 3C), and bore no relation to the (presumably diapiric) emplacement of the Peridotite Ridge. Sibuet (1992), seeking to explain a discrepancy between the amount of stretching measurable from faulting and that measurable from subsidence, adopted a composite model, akin to that proposed by Le Pichon & Barbier (1987) for

DETACHMENT TECTONICS OF GALICIA MARGIN

95

a) 1

0

seafloor reflection, unfiltered

--

--

6800

69'00

~

/

k/

k/'"

TIME [msec]

v

70'00

8000

8700

8800

b) 1

S reflection, unfiltered

~o -1

8500

8600

TIME [msec]

c) 1

<

seafl0or reflection, filtered

0

"

-1

|

6800

6900

TIME [msec]

7000

8000

8700

8800

d) Fig. 3. Cartoon representation of some of the models that have been proposed for the evolution of the Galicia margin (and its conjugate, Flemish Cap). (a) Pure shear~de Charpal et al (1978). (b) East-dipping simple shear--Boillot et al. (1988) (e) West-dipping simple shear--Winterer et al. (1988). (d) Combined model--Sibuet (1992). The S refector has played an important, but different, role in the different models proposed.

the similar north Biscay margin. In this model the S reflector is interpreted as a landwarddipping, simple shear detachment within the upper-middle crust, with the lower crust and mantle lithosphere below being considerably stretched by effective pure shear (Fig. 3D). Clearly, an understanding of the nature of the S reflector would further the understanding of this margin, in particular, and of the extension of the lithosphere, in general. The aim of this chapter is to describe preliminary results from analysis of the waveform, polarity and amplitude of S, and from pre-stack depth migration carried out on the S reflector at Geomar.

Polarity and waveform analysis The determination of the polarity, and the

1

-1

S reflection, filtered

8500

8600

TIME [msec]

Fig. 4. Comparison of the waveform of the seafloor refection and that of the S refector (both taken from a near-trace stack). The waveforms are similar prior to any other processing; the likeness is striking after filtering the data so that S and the seafloor reflection have a similar bandwidth. The comparison suggests that both S and the seafloor refections come from similar boundaries, namely a simple step-like increase in acoustic impedance.

modelling of the waveform of the S reflection, should constrain whether S is a reflection from a single interface, a thin layer or a more complex zone, and whether it represents an increase or decrease in acoustic impedance. Such analysis should, for instance, distinguish between a reflection from an igneous intrusion (thin layer of high acoustic impedance), a broad fault zone (layer of low acoustic impedance) and the crust mantle boundary (step-like or transitional/ complex increase in acoustic impedance). Pre-

96

T.J. RESTON E T AL.

Fig. 5. Apparent polarity section derived from complex trace analysis of a portion of GP12, indicated in Fig. 2. Both the seafloor and S are consistently positive polarity (black), except where the continuity of S is poor (see Fig. 6).

liminary results from such analysis are reported herein. As a first step, a 'near-trace stack' (a stack of the offsets between 300 and 1100 m) of the data, after applying normal moveout correction based on careful velocity analysis carried out using the complete range of offsets, was constructed. By stacking the data random noise is strongly attenuated; by making only a stack of the near-offsets, the waveform distortion before stacking associated with normal moveout stretch is negligible, and that on the stacked trace associated with inexact moveout correction is also minimized. The simplest way to examine the waveform of the S reflector is to compare it to that of the seafloor. This has been modelled as a step in acoustic impedance from water velocity and density to the higher velocities and densities of sediment (Warner 1990). It can be seen from Fig. 4 that the waveform of S is remarkably similar to that of the seafloor, implying that S is a similar sharp increase in acoustic impedance, that is the product of velocity and density. After suitably band-limiting both (limiting the band-

width of the waterbottom reflection to that of S), the similarity is even more striking: S and the seafloor reflection are effectively identical. The consistency of this result has been investigated using complex trace analysis (Taner & Sheriff 1977). A display of the apparent polarity of the deconvolved near-trace stack is shown in Fig. 5: positive polarities are black, negative polarities grey. It is clear that both the water-bottom reflection (everywhere) and the S reflection (mostly) are positive polarity events. The only places where S does not appear to be a positive polarity reflection corresponds to places where its amplitude is unusually low, as apparent on the envelope function of a true amplitude near-trace stack (Fig. 6). As the polarity of a weak reflection is affected by noise and should be discounted, the S reflection can be considered as consistently positive polarity. It is also apparent from the polarity function that there is no comparable negative polarity reflection beneath S, suggesting that S does not represent the top of a layer but rather a fundamental boundary. Similarly, there is no strong reflection beneath S on the envelope section (Fig. 6), and

DETACHMENT TECTONICS OF GALICIA MARGIN

97

Fig. 6, Envelope function, derived from complex trace analysis of true amplitude section of GPI2. S appears as a remarkably strong reflection: calibration using the water-bottom multiple gives a reflection coefficient for S of c. 0.2.

no reflection (Fig. 2) where the Moho might be expected at c. 10 s twt (R. B. Whitmarsh, pers. comm.). .

amplitude of the water-bottom reflection (in deep water) is given by: Ap = R x / d ,

Amplitude analysis It is apparent from the true amplitude envelope function (Fig. 6), that S is a remarkably strong reflection. As the amplitude of a reflection is a function of the change in physical properties across that interface, t h e quantification of the amplitude of S provides further constraints on the nature of this feature. Here preliminary results from our analysis of the amplitude of S are presented. A problem with determining absolute amplitudes from reflection data, is that only relative amplitudes are provided by the data. Thus, it is necessary somehow to calibrate the system before relative amplitudes can be translated into absolute values. The calibration method adopted here is that described by Warner (1990), and is based on the comparison between the water-bottom multiple and the water-bottom reflection. As Warner describes in detail, the

where R is the reflection coefficient of the seafloor, x is a calibration factor and d a factor related to geometrical spreading. Warner points out that the amplitude of the first water-bottom multiple is: Am = R2x/2d.

Thus, the value of R can simply be determined from the relative amplitude of the water-bottom reflection and its first multiple: R = 2 Am/Ap.

This analysis assumes that the energy is reflected at normal incidence, and that the seafloor is smooth and horizontal. Consequently we carried out our analysis on the near-trace stack described above (maximum angle of incidence at water bottom c. 7~ taking an average over the

98

T.J. RESTON E T AL.

600 CMPs where both the water-bottom reflection and the corresponding multiple were smoothest and flattest. The analysis also assumes that the amplitude of the seafloor wavelet and that of the multiple, are not affected by subsurface reflections. To check this the amplitudes of the water-bottom reflection and multiple for the first, second (largest) and third peaktroughs of the wavelet, and also the peak of the envelope function, were compared. Within error, all gave the same value, with the envelope function (as expected) showing the smallest standard deviation, giving a reflection coefficient for the water bottom of 0.23 +0.04. This is within the range of values expected for a soft water bottom (Warner 1990). Having calculated the amplitude of the water bottom, the amplitude of this was then compared with that of the S reflector to determine an estimate for the reflection coefficient of S, by determining its average amplitude over 6 km on GP12 where it is particularly continuous. We again used the near-trace stack of the data, simply comparing the maximum amplitude of the S reflector with that of the water bottom after spherical divergence correction, applied using a 2D velocity function based on our analysis of velocities. By using a stack the reflection coefficient obtained should be considered a minimum estimate, as the amplitude of S is likely to be reduced by imperfect moveout correction and stacking rather more than the water bottom reflection. By using a near-trace stack this problem is again minimized, while suppressing random noise that is likely to obscure the true amplitude of S. Future work will use forward modelling of shot gathers to constrain further the reflection coefficient of S. To ensure that the amplitudes measured were not affected by the digitization process the maximum amplitude of both S and the waterbottom reflection from the envelope function were chosen. To ensure that the effects of focusing and defocusing would not affect our results the total energy reflected over 6kin, a distance considerably wider than the Fresnel zone, was calculated: Raynaud (1988) points out that the total energy reflected from a reflector larger than the Fresnel zone is the same regardless of the shape of that reflector. The length of the reflection investigated corresponded to c. 1 wavelength of the apparent undulation of S, so all focusing effects should be averaged out. From the ratio of the energy reflected from S and the sea floor, an average amplitude for the S reflection was then calculated. This method was further applied to both the migrated and unmigrated data, to check that

diffractive energy was included. In both cases, before allowing for anelastic attenuation (Q) and for transmission losses, we found that the amplitude of the S reflection was approximately half that of the seafloor. After allowing for the effects of attenuation (Boillot et al. this volume) and transmission we estimate that S has a m i n i m u m reflection coefficient of close to 0.2. Work is underway to model the amplitude of S in more detail. Thus, it is concluded that the reflection coefficient of S is probably at least 0.2. This, coupled with the results of polarity and waveform analysis, indicates that S represents a major step-like increase in acoustic impedance. For instance, if the velocity just above S was 5.5kms -1, then, assuming that velocity and density increase together, the velocity beneath would be c. 6.7 km s-l. Although no clear wideangle reflection has been reported from S on recently acquired OBS data, the wide spacing of the receiver stations suggests that the wide-angle survey might simply have missed S (R. B. Whitmarsh, pers. comm.). Alternatively, the near-vertical and wide-angle data could be reconciled if S represents a step-up in velocity overlain by a zone of downward decreasing velocity, perhaps with a diffuse upper boundary. This would generate a positive polarity reflection on the near-vertical data, but would diminish the amplitude of S on wide-angle data. Clearly, this is a topic for further analysis, but such a model would be compatible with the interpretation of S developed below. The observation that S is best interpreted as a single interface rather than a thin layer indicates that S is unlikely to be a reflection from a thin igneous intrusion. It is also difficult to see how S could represent a reflection from a fault zone: in that case it might be expected to be a negative polarity reflection closely followed by a positive polarity event. This is not seen. However, it has been shown that many detachment faults are not broad zones, but rather are thin boundaries separating quite different lithologies. On a recent trip to the western USA it was pointed out to one of us (T.J.R.) that the bulk of the movement along the Dead Mountains detachment fault [over 50 km horizontal displacement (E. Frost, pers. comm.)] was accommodated along a fault zone only a few centimetres thick (Fig. 7). This fault separates virtually unmetamorphosed sedimentary and volcanic rocks in the upper plate from middle crustal rocks in the lower plate. The difference is even more pronounced in other detachment faults, such as the Chemehuevi detachment, which bring lower crustal rocks to the surface. The seismic contrast expected across

DETACHMENT TECTONICS OF GALICIA MARGIN

99

Fig. 7. Photograph of Dead Mountain detachment fault, showing upper plate rocks (dark) separated from lower plate rocks (light) by a thin (c. 2 cm) ledge of ultracataclasite (the detachment fault), on which the lens cap is resting. The seismic signature of such a detachment fault should be identical to that observed for S. Consequently, it is possible that S represents a reflection across a detachment fault. such major tectonic boundaries would be a sharp increase in acoustic impedance, much as observed for S. An alternative explanation for the observed amplitude, polarity and waveform of S is that it represents the crust-mantle boundary. Although the velocity structure reported from wide-angle data (R. B. Whitmarsh, pers. comm.) indicates that the velocities beneath S are less than expected for mantle rocks, it is possible that S might represent a boundary between crustal rocks and serpentinized mantle rocks, as suggested by Boillot et al. (1988b). The presence of serpentinized peridotite beneath S would imply that estimates of extension based on refraction or subsidence (Sibuet 1992) are likely to underestimate the total amount of extension, and also that the variation in apparent crustal thinning across the margin might represent variations in the degree and depth of serpentinization. In summary, our preliminary analyses~crf the polarity, waveform, and amplitude of S suggest that S represents a major steplike increase in acoustic impedance. This can best be interpreted as indicating that S is either a major detachment fault, separating an upper plate from a higher velocity and density lower plate, or that S

represents the boundary between the crust and serpentinized mantle. Both may be correct: it is possible that S is a detachment fault separating the crust (upper plate) from the mantle (subsequently serpentinized) of the lower plate. In this interpretation the crust-mantle boundary would locally be a tectonic boundary.

Relationship between S and overlying faults: pre-stack depth migration One problem in interpreting the S reflector as a major extensional detachment fault is that it appears in places to be cut and offset by the steeper faults that divide the overlying section into a series of tilted fault blocks (Fig. 8). Clearly, if these faults do cut and offset S, S must pre-date at least some of the extension of the upper crust. However, it is possible that the apparent offsetof the S reflector is an artefact resulting from the effects of velocity pull-up (Mauffret & Montadert 1988) and from the loss of continuity of S beneath the complex structure of the upper crust. It is vital to distinguish between these two explanations before the role of the S reflector in lithospheric extension can be

100

T.J. RESTON E T AL.

Fig. 8. (a) Time section of GP102 derived from standard processing (time migrated after DMO stack). The S reflector is clearly imaged beneath the tilted fault blocks, and appears to be offset. (b) Time section of GP102 derived from pre-stack depth migration (depth section shown in Fig. 11) and depth-time conversion to facilitate comparison with (a). Note that the continuity of S has been greatly increased and S can be traced as an undulating horizon beneath the fault blocks. Compare also with Fig. 2. assessed: if S is cut by the block-bounding faults it may not represent a detachment fault at all, or was only active during early extension, whereas if these steeper faults can be shown to detach into S it almost certainly must represent a detachment fault active during at least late rifting and continental break-up. To address this issue, we have applied iterative pre-stack depth migration to part of the existing seismic data where the apparent offset of S by the block-bounding faults is most marked, that is part of Institut Frangais du P&role (IFP) profile GP102. Depth migration not only provides a final section in depth (essential if the true geometry of structures is to be studied) but also incorporates the effects of raypath bending, particularly important in the presence of the rapid lateral velocity variations expected with tilted fault blocks and sediment-filled halfgrabens. Migration before stack avoids the smearing effects of C M P stacking (e.g. see Peddy et al. 1986), thus improving the image in regions

of complex geology. Pre-stack depth migration can therefore provide a more accurate and higher resolution depth section than otherwise possible. By performing the pre-stack depth migration iteratively, it is possible to extract accurate velocity information using Migpack T M software (Denelle et al. 1986), vital to provide an accurate depth migration. In so doing we construct a velocity model using focusing analysis (Fig. 9) of reflected and diffracted energy at closely spaced (500 m) intervals along the profile. It is necessary to build up the velocity model one layer at a time as the overburden velocity structure affects that determined for deeper levels. Thus, we first determine the water velocity (and test our geometry at the same time), then incorporate that velocity into the next iteration when we determine the velocity of the topmost sedimentary layer again through depth-focusing analysis. The procedure is repeated until the entire section has been analysed for velocity, resulting

DETACHMENT TECTONICS OF GALICIA MARGIN

.

vCLO~cITCTs

S reflector appears to pass almost straight beneath the blocks, without appreciable offset by the block-bounding faults. This indicates that much of the topography of S apparent on Fig. 8 can be related to velocity pull-up effects beneath the fault blocks. Furthermore, the continuity of S has been greatly increased by pre-stack depth migration (compare Fig. 8a & b), due to improved imaging beneath the complex overburden (Fig. 10). Rather than significantly offsetting S, the block-bounding faults appear to stop abruptly at S, which they intersect at c. 30 ~ (Fig. llb; see Hoffmann & Reston 1992). Therefore, we interpret S as a detachment fault active throughout the extension of the overlying section. The observations made from the depth section of GP102 can also be made from the time section of GP12 (Fig. 2). Here, the fault block structure is less pronounced and consequently the pull-up effects on S less severe. On this profile the faults that divide the upper crust into blocks can be clearly seen and traced downward as far as, but no further than, the underlying continuous (but undulating) S reflection. Hence it appears that here too, S can best be interpreted as a detachment fault. Work is underway to see if the topography of S on this profile is an effect of velocity pull-up, or whether it m a y partly represent the actual undulations of the reflecting surface. Variable topography is a feature of detachment fault systems and may reflect the variable unloading of the lower plate during tectonic denudation (e.g. Buck 1988). Thus, whether S is planar or not does not affect our interpretation that it represents a major detachment fault: the critical observation is that S is not offset. Indeed, we would expect some distortion of S due to the variable loading of this feature by the overlying fault blocks. The improved image within the blocks, and the detailed velocity model allow the identification of an early sedimentary sequence, tilted during block rotation (Fig. l lb). However, the base of this sequence can only be clearly identified within one block, so it is not possible to tell conclusively whether S cuts structurally deeper to the east or west. On the depth section S does appear to dip consistently to the west and does appear to approach the surface at the eastern end of the reprocessed part of GP102, as in the interpretation of Winterer et al. (1988). Thus, we believe that S probably is a westdipping detachment, but await confirmation of this from further work on this and other profiles. This interpretation is consistent with the sense of tilt of the overlying fault blocks (e.g. Beslier & Brun, 1991).

O••wardcon ~'

VELOCITY TOOHIGH

/

depth

~eth ~

true Apparent depth depth focusing........~ error

Fig. 9. Cartoon illustrating the method of depthfocusing analysis [adapted from Denelle et al. (1986)]. During simultaneous migration in both shot and receiver domains with the correct velocity function (a), maximum focusing occurs when the downward continuation operator reaches the actual depth of the reflector. If an incorrect migration velocity is used (b), then there is a discrepancy between the depth of maximum focusing and the depth to the reflector. Analysis of this discrepancy (the depth-focusing error) allows correction of the velocity model. in an optimized depth migration and a velocity model that is both detailed and geologically meaningful. The velocity model is not only essential to provide an optimum migration but is also useful in identifying the limits of poorly imaged units, such as sediments within strongly tilted fault blocks (Fig. 10). The principal disadvantage of the method is that detailed velocities can only be determined above bright reflections: focusing analysis on such reflections provides velocity information on the overlying but not the underlying section. The method is also very costly in terms of computer time. Results

The resulting depth section is shown at true scale in Fig. 1 la. The most immediate result is that the

101

102

T.J. RESTON E T AL.

Fig. 10. Simplified velocity model derived from iterative pre-stack depth migration and depth-focusing analysis (six full iterations). The basic fault block pattern is apparent.

Fig. 11. (a) IFP profile GP102 after iterative pre-stack depth migration. The S reflector can be seen to pass continuously beneath the block-bounding faults without being offset. The section is in depth, shown here with no vertical exaggeration. (b) Interpreted line-drawing. Note that the block-bounding faults are approximately planar and intersect S at c. 30~ Also note the complex reflectivity in the triangular zone between these block-bounding faults and S: this may represent a severely brecciated zone.

The intersection between the sedimentary sequence within the fault blocks and the faults implies an original fault dip of 60-70 ~ (Figs 11 b & 12a), assuming that the early sedimentary sequence was horizontal prior to faulting. As these faults currently dip at e. 35 ~ the upper

crustal fault blocks must have rotated by c. 30 ~ during the extension of the upper plate to S. This immediately poses a space problem (Fig. 12a): if the blocks are rotated back so that the top basement is horizontal, the base of the blocks forms a sawtooth pattern.

DETACHMENT TECTONICS OF GALICIA MARGIN

103

Fig. 12. (a) Simplified interpretation of GP102, and rigidly restored section, showing space problem: the base of the blocks form a sawtooth pattern. The simplest way to overcome this problem is to modify the base of the fault blocks during extension and tilting. In a domino model (b) a space problem arises as the blocks back-rotate. This can, however, be overcome if the base of the block is cataclastically modified (c). The lower plate to the detachment grinds the keels down, and the debris collects in the gaps opened up just above S and in the footwall to the block-bounding faults. The resulting fault block shape (outlined by hatched line) is similar to that observed [cf. (c) & (a)]. Another way to visualize this space problem is to consider the domino-style extension of an upper plate with a basal detachment, the S reflector (see also Axen 1988). In the domino model (Fig. 12b) extension is accommodated by the rigid-body rotation of the fault blocks; as these rotate their base tilts, with one corner rising and the other dropping, creating the classical keel problem of this model (Wernicke & Burchfiel 1982). One solution was proposed by Jackson et al. (1988), who suggested that the base of the fault block was thermally controlled (effectively related to the so-called brittle-ductile transition) and adjusted rapidly on a geological timescale to offsets along the overlying faults. However, this model cannot apply to the fault blocks imaged here as the S reflector occurs only a few kilometres beneath the seafloor, and was in places < 2 km deep prior to post-rift subsidence. This is well above the base of the shallow seismogenic zone, and well above the sort of transition envisaged by Jackson et al. (1988). Furthermore, the current base of the fault blocks (S) is a bright reflection which is not typical of a transition zone.

It is possible that the entire fault block deforms to accommodate slip on the basal detachment (e.g. Le Pichon & Chamot-Rooke 1991). However, the internal reflectors which can be recognized within the blocks do not appear deformed (nor do they on the Biscay margin, where a similar problem arises). Instead, we consider a more likely explanation to be that the base of the fault blocks has been cataclastically modified during progressive extension and rotation (Fig. 12c), and particularly the abrasion of the fault block keels (Axen 1988). As these rotate, we propose that the block keels were tectonically eroded and the resulting breccia scraped up by the gaps arising in the footwall to the steep faults (but above the detachment) as the blocks back-rotate. Recently, on an excursion to the Whipple M o u n t a i n region (led by Professor Eric Frost), one of us (T.J.R.) observed just such phenomena above the master detachment fault: as steep faults approached the gently-dipping detachment, the degree of brecciation in the footwall to those steep faults became intense, suggesting that the accommodation of block rotation was by such

104

T.J. RESTON E T AL.

Fig. 13. Model for formation of core complexes by flexural rotation of high-angle normal fault to low-angles (Buck 1988). Footwall to fault rises due to unloading, and the upper portion deforms flexurally beyond the angle at which the fault can remain active. New fault propagates up through the upper plate from depth, transferring a slice of upper plate to lower plate. As lower plate is further uplifted and deformed, the crustal slice is carried passively up on top of inactive portion of fault. Continued denudation results in a series of strongly rotated upper plate slices resting on top of a fault rotated to low angle. This may provide an alternative explanation for the S reflector, although the faults off Galicia appear to stop abruptly at S rather than sole into it, and the blocks are considerably less rotated than in the Buck model.

tectonic removal of the keel and cataclastic flow of the debris into the footwall to the blockbounding fault above the detachment. In contrast, the hanging wall to the steep faults was relatively intact, as was the lower plate to the detachment fault. This accords well with the model proposed here to explain the observed fault geometries. Both the removal of the keels to the fault blocks and the activity of S at low angle during the extension of the upper plate along steeper faults imply that the lower plate to S may have been significantly stronger than its upper plate. We suspect that the upper plate may have been substantially weakened by the passage of fluids up the detachment and into the overlying section: evidence for such fluids must await drilling. We are aware of at least two alternatives to the model proposed above. For instance, Lister & Davis (1989) proposed that the apparent

domino-style faulting observed above detachment faults could be interpreted as the upper half of fault blocks (listric or dominos), excised by a later through-going detachment fault. The principal problem with this model is difficulty in propagating a low-angle detachment through an already heavily faulted upper crust. The other model that has recently been used to explain the formation of low-angle detachment faults is that proposed by Buck (1988), and the related model of Wernicke & Axen (1988). In these models tectonic unloading of the footwall to a steep fault causes that fault to rotate either flexurally or in response to isostasy. If the flexural rigidity is low enough the denuded portion of the fault rapidly becomes sub-horizontal (Fig. 13). Although the sub-horizontal portion of the fault may carry slivers of the upper plate it is inactive, and the overlying fault slivers are carried as if on a conveyor belt away from the active portion of the fault. Thus, a key

DETACHMENT TECTONICS OF GALICIA MARGIN

105

Fig. 14. Portion of GP12 showing relationship between S and the Peridotite Ridge. S appears to dip gently to the west and be truncated by east-dipping reflections cutting down from the peridotite ridge. See Fig. 15 for interpretation. feature of these models is that the age of the detachment and of the faulting in the upper plate, increases towards the breakaway. The relevance of this model to the interpretation of the S reflector is obvious, implying that S might have only been active at high angles, and was rotated to its current low angle during unloading. However, the model described by Buck (1988), for instance, predicts that the upper plate to S should consist of thin crustal slivers rotated by the same amount as the detachment (60~ if S originated as a steep normal fault) and bound by faults which sole into the basal detachment. Instead, the fault blocks above S are rotated only by c. 30~ and are bound by approximately planar faults that intersect S at c. 30~ Thus, we consider that this model, either as formulated by Buck (1988) or by Wernicke & Axen (1988), may not apply to the Galicia margin. The test for these models, and that of Lister & Davis (1989), will be the dating of movement along the detachment and of the extension of the upper plate. However, such dating can only be achieved by drilling.

Relationship between S and the Peridotite Ridge The foregoing discussion describes the evidence that S is indeed a major detachment fault, active

during continental rifting. Although S has not itself yet been sampled, dredging, drilling and dives by submersible have revealed that the Peridotite Ridge is bounded to the east by an east-dipping, top-to-the-east extensional shear zone, formed during continental rifting (Boillot et al. 1988a). Thus, it is possible that S might represent the landward continuation of this detachment fault, as proposed by Boillot et al. (this volume). However, it has been pointed out that the sense of tilt of the fault blocks implies that movement along S was top-to-the-west, consistent with the gentle westerly dip of S apparent on the depth section of GP102, but conflicting with the observed top-to-the-east sense of shear observed on the eastern side of the Peridotite Ridge. To investigate the relationship between S and the Peridotite Ridge, and hence to resolve this paradox, we have paid particular attention to the processing of the data in the critical region just to the east of the ridge. Here we show our results from GP12: although the same observations can be made from the other profiles, they are somewhat less clear. Particular attention has been paid to velocity analysis in this critical region, with analyses being made before dip-moveout correction (DMO) every 200 common midpoints (CMPs), and after D M O every 50 CMPs (that is every 1.25 kin). As our velocity analysis is based to a

106

T.J. RESTON E T AL.

Fig. 15. Schematic illustration of relationship between S and Peridotite Ridge. (a) Sense of tilt of fault blocks above S suggests that S is a top-to-west shear zone, but sense of shear observed in peridotite is top-to-east. The seismic data (e.g. Fig. 14) show that S is truncated by east-dipping reflections cutting down from the Peridotite Ridge (PR). This may indicate (b) that S is conjugate to the 'mantle shear zone' (Beslier & Brun 1991), or that this post-dates and cuts S (e). Note that the exact position of the crust-mantle boundary (CMB) in (b) and (c) is unknown. large extent on constant-velocity stacks over 50 CMPs our complete analysis has investigated the optimum stacking velocity for every C M P in this region, i.e. every trace on the stack section. The resulting section is shown after migration in Fig. 14. The S reflector is clearly seen at the east end of the section, but becomes less coherent to the west, where it is intersected by a west-dipping normal fault. Beyond the fault, and continuous with S, are clear west-dipping reflections. These we consider to be the westernmost limit of the S reflector. Further west the section is dominated by clear

east-dipping reflections. Some of these appear to come off the Peridotite Ridge, others to come from underneath the ridge. Although the deeper portion of the data is typically obscured by migration smiles (Fig. 14) it is apparent that these east-dipping events truncate the westdipping S reflections. The east-dipping reflections continue off the bottom of the stack section at 11 s twt, so it is probable that the east-dipping structure continues to greater depths than is apparent on the migrated section. Because of these observations we consider it unlikely that the S reflector is continuous with the top-to-the-

DETACHMENT TECTONICS OF GALICIA MARGIN east shear zone sampled on the east face of the Ridge, but instead believe that S is truncated by this top-to-the-east shear zone (Fig. 14). We suggest that either S is antithetic to the mantle shear zone exposed on the ridge or that this shear zone cuts across an earlier west-dipping structure, of which S is a portion (Fig. 15). In one model (Fig. 15b) lithospheric extension may have been accommodated along conjugate crustal and mantle shear zones [the lithospheric boudinage model of Beslier & Brun (1991)]. A similar model has previously been proposed for lithospheric extension north of Scotland (Reston 1990). Alternatively, a major crustal detachment, active during continental rifting, may have been rendered inactive by a cross-cutting shear zone dipping in the opposite direction during the final phase of continental break-up (Fig. 15c). In either mode, the shear zone dipping to the east off the ridge is likely to be at least as fundamental a structure as the S reflector. Depending on the lithospheric level to which the S detachment cut, the lower plate to S might consist of either crustal or mantle material. If mantle material the velocity structure (Whitmarsh, pers. comm.) indicates that it must be serpentinized. If the lower plate to S consists of crustal material then the amplitude analysis and velocity structure implies that this must be high-grade lower-crustal rocks. Either way, S is clearly a major tectonic structure.

Conclusions and discussion In this paper we have presented preliminary results from our analysis of the existing reflection data. Our principal results are that: (1) S is a positive polarity reflection from a single interface; (2) S has a reflection coefficient of close to 0.2; (3) S appears to pass beneath the blockbounding faults of the upper crust without being offset; and (4) that S appears to terminate against east-dipping reflections associated with the Peridotite Ridge. From these results we interpret S as a detachment fault, separating the high velocity and density lower plate from the lower velocity and density crustal rocks of the upper plate: the lower plate to S might represent either lower crustal material or mantle peridotites, partially serpentinized. From the relationship between S and the overlying faults it appears that S was active as a low-angle detachment at least during the final stages of crustal extension, with the faults of the upper plate detaching onto S. However, the westward termination of S against east-dipping reflections cutting down from the Peridotite Ridge implies that S may be cut by a later east-dipping shear

107

zone, or that S was antithetic to a master eastdipping detachment. The identification of a major detachment beneath the Galicia margin has important implications for the lithospheric extension that led to the formation of the Atlantic. Firstly, it seems that low-angle detachments may play an important role in lithospheric extension. Secondly, the relationship between S and the interpreted mantle shear zone suggests that the simplest models of lithosphere penetrating detachments may not apply to this margin, and hence that a classification of the margin in terms of upper or lower plate is oversimplified: the bulk of the Galicia margin (e.g. the Galicia banks) may be in the lower plate to S, but may also be in the upper plate to a deeper mantle detachment, dipping landward. However, we consider it unlikely that this deeper detachment controlled the final opening of the Atlantic, as if the lithosphere continued to be pulled apart along an east-dipping shear zone the peridotitic lower plate to that shear zone should have been exposed on the western margin, not the eastern margin to the Atlantic. The transition from rifting to drifting requires further study, but it is clear that neither the lithospheric pure shear model, as it is generally perceived (symmetric arrays of steep normal faults overlying a region of pure shear), nor the lithospheric simple shear model (a lithospheric penetrating shear zone) appears fully applicable to the extension that led to the formation of this portion of the Atlantic. However, as with all seismic interpretations, the proof is in the drilling. Although our analysis leads us to the above conclusions, the only way to determine unequivocally the nature of S is to sample it. The only way to sample S in its original position (with both the upper plate and lower plate intact) is to drill it. Drilling S would not only determine the nature of this structure but also of analogous structures observed on other margins, such as Biscay (Le Pichon & Barbier 1987), and the Gulf of Lions (de Voogd et al. 1991). Furthermore, drilling will answer questions not readily solvable from the seismic data. For instance, dating of S would reveal the relative timing of movement along it and along the shear zone exposed on the east flank of the ridge, and hence constrain the nature of lithospheric extension still further. Drilling would reveal the importance of fluids during movement along the low-angle detachment. Drilling may also determine the importance of undercrusting by serpentinite (Boillot et al. this volume). The possibility that the lower plate to S contains serpentinized peridotite calls into question the estimate of crustal extension from subsidence

108

T.J. RESTON ET AL.

a n d r e f r a c t i o n d a t a . Q u a n t i f i c a t i o n o f the i m p o r t a n c e of serpentinite w o u l d have implications for the validity o f pre-rift reconstructions o f the Atlantic: these avoid the problems of d i a c h r o n o u s , p r o p a g a t i n g rifts (zipper tectonics) associated with pre-drift reconstructions, but are critically d e p e n d e n t on estimates o f crustal extension. Thus, drilling the S reflector, a n d its u p p e r a n d lower plates, can answer questions of global importance. We are grateful to Dirk Klaeschen for considerable technical help and advice during the pre-stack depth migration; Jacques Wannesson at I.F.P. for providing the data used in this study, and Gilbert Boillot for continuing discussion about the nature of S and the structure of this margin. T.J.R. would like to take the opportunity to thank Eric Frost, the trip leader, and the other participants on a recent Ocean Drilling Program, Tectonics Panel, field trip to the south of Las Vegas for much discussion about detachment systems.

References

AXEN, G. 1988. The geometry of planar domino-style normal faults above a dipping detachment. Journal of Structural Geology, 10, 405411. BESLIER, M . - O . & B R U N , J. P . 1991. Boudinage de la lithosphre et formation des marges passives. Comptes Rendus de l'Acad~mie des Sciences, Paris, 313, 951-958. BOILLOT, G., GIRARDEAU,J. & KORNPROBST,J. 1988a. Rifting of the Galicia margin: crustal thinning and emplacement of mantle rocks on the sea floor, In: BOILLOT, G. & WINTERER, E. L. (eds) Proceedings of the Ocean Drilling Program, Science Results, 103, College Station, TX, 741-756. , COMAS, M. GIRARDEAU, J. Er AL. 1988b. Preliminary results of the Galinaute cruise: dives of the submersible Nautile on the western Galicia margin, Spain, In: BOILLOT, G. & WINTERER, E. L. (eds) Proceedings of the Ocean Drilling Program, Science Results, 103, College Station, TX, 37-51 BUCK, W. R., 1988. Flexural rotation of normal faults. Tectonics, 7, 959-973. DE CHARPAL, O., GUENNOC, P., MONTADERT, L. & ROBERTS, D. 1978. Rifting, crustal attenuation and subsidence in the Bay of Biscay. Nature, 275, 706-711. DE VOOGD, B. ET AL. 1991. First deep reflection transect from the Gulf of Lions to Sardinia ( E C O R S - C R O P profiles in the w e s t e r n Mediterranean). AGU Geodynamics Series, 22, 265-274. DENELLE, E., DEZARD, Y. & RAOULT, J. 1986. 2-D prestack depth migration in the (S-G-W) domain. Extended abstract, 56th SEG Meeting, Houston. HOFFMANN, H. J. & RESTON, T. J. 1992. Nature of the S reflector beneath the Galicia Banks rifted margin. Preliminary results from prestack depth

migration. Geology, 20, 1091-1094. JACKSON, J., WHITE, N., GARFUNKEL, Z. & ANDERSON, H. 1988, Relations between normal fault geometry and vertical motions in extensional terranes, Journal of Structural Geology, 10, 155170. LE PICHON, X. & BARBIER, F. 1987. Passive margin formation by low angle faulting within the upper crust: the northern Bay of Biscay margin, Tectonics, 6, 133-150. - & CHAMOT-ROOKE, N. 1991. Extension of the continental crust, Controversies in Modern Geology. Academic Press, New York, 313-338. - & SIBUET,J. C. 1981. Passive margins: a model of formation, Journal of Geophysical Research, 86, 3708-3720. LISTER, G. & DAVIS, G. 1989. The origin of metamorphic core complexes and detachment faults formed during Tertiary continental extension in the northern Colorado River region, U.S.A. Journal of Structural Geology, 11, 65-94. - - , ETHERIDGE, M. & SYMONDS, P. 1986. Detachment faulting and the evolution of passive margins. Geology, 14, 246-250. -- & - 1991. Detachment models for the formation of passive continental margins. Tectonics, 10, 1038-1064. MAUFFRET, A. & MONTADERT, L. 1988. Seismic stratigraphy off Galicia. In: BOILLOT, G. & WINTERER, E. R. (eds) Proceedings of the Ocean Drilling Program, Science Results, 103, College Station, TX, 13-30. MCKENZlE, D. 1978. Some remarks on the development of sedimentary basins. Earth and Planetary Science Letters, 40, 25~42. PEDDY, C., BROWN, L. & KLEMPERER, S. 1986. Interpreting the deep structure of rifts with synthetic seismic sections. AGU Geodynamics Series, 13, 301-311. RAYNAUD, B. 1988. Diffraction modelling of 3-D lower crustal reflections. Geophysics Journal, 93, 149-161. RESTON, T. J., 1990. The lower crust and the extension of the lithosphere: kinematic analysis of BIRPS deep seismic data. Tectonics, 9, 1235-1248. SIBUET, J.-C. 1992. Formation of non-volcanic passive margins: a composite model applies to the conjugate Galicia and southeastern Flemish Cap margins. Geophysical Research Letters, 19, 769772. TANER, M. & SHERIFF, R. 1977. Application of amplitude, frequency, and other attributes to stratigraphic and hydrocarbon determination. American Association of Petroleum Geologists Memoir, 26, 301-327. WARNER, M. 1990. Absolute reflection coefficients from deep seismic reflections. Tectonophysics, 173, 15-23. WERNICKE, I . 1981. Low-angle normal faults in the Basin and Range: Nappe tectonics in an extending orogen. Nature, 291, 645-648. - & AXEN, G. 1988. On the role ofisostasy in the evolution of normal fault systems. Geology, 16, 848-851.

D E T A C H M E N T TECTONICS OF GALICIA M A R G I N -

-

-

-

& BURCHFIEL, B. C. 1982. Modes of extensional tectonics, Journal of Structural Geology, 4, 105-115. & TILKE, P. G. 1989. Extensional Tectonic framework of the U.S. Central Atlantic passive margin. American Association of Petroleum Geologists Memoir, 46, 7-21.

109

WINTERER, E. L., GEE, J. & VAN WAASBERGEN, J. 1988. The source area for lower Cretaceous clastic sediments of the Galicia margin: geology and tectonic and erosional history, In: BOILLOT, G. & WINTERER, E. (eds) Proceedings of the Ocean Drilling Program, Science Results, 103, College Station, TX, 697-732.

Seismic investigation of the Faeroe basalts and their substratum LISELOTTE

KIORBOE 1 & STEEN AGERLIN

PETERSEN 2

1 Geological Survey o f Denmark, Thoravej 8, DK-2400 Copenhagen N V , Denmark 2 Norsk Hydro, N-5020 Bergen, Norway

Abstract: Various seismic techniques have been used to investigate the subsurface in the area around the 2.2 km deep Lopra-1 well located on the Faeroe Islands. The techniques include vertical seismic profiling, walk-away vertical seismic profiling and refraction seismic profiling. Integration of these three different techniques, using a combined profiling method, makes it possible to overcome some of the problems arising from the large acoustic impedance contrasts and S-wave scattering which often exist in volcanic areas. The three seismic profiles show distinct compressional (P) and shear (S) energy arrivals. The direct S-wave is generated at basalt interfaces at or just below the sea bed and interferes in part with groundroll in the refraction recordings. The low noise activity before first arrival indicates the very high quality of the VSP data. The VSP and the walk-away VSP recordings show clear evidence of reflected P-waves, while reflected P-waves are more difficult to recognize on the surface recordings. The velocity through the basalt flows is shown to be dependent on the thickness of the basalt beds. Thick beds yield high velocities, while thin beds generate lower velocities. Velocity anisotropy is observed in the upper 700 m, probably resulting from cracks in the columnar basalt flows or from master joints. The seismic examination of the three profiles shows excellent interbasaltic reflections. A very strong reflector with a negative reflection coefficient is observed at a depth of c. 2.35 km and is interpreted to be the base of the basalt flows. Deeper reflections are interpreted as sills in the substratum of the basalt flows.

The Faeroe Islands are situated in the BritcrArctic Igneous Province. They consist of tholeiitic plateau basalts, extruded in the early Tertiary during the opening of the NorwegianGreenland Sea. More than 3 km of basalt lavas are exposed on the islands (Rasmussen & NoeNygaard 1970) and a further 2.2km of section has been encountered in the subsurface (Berthelsen et al. 1984). The Lopra-1 well was drilled for scientific purposes in 1981 on Suduroy, the southernmost island on the Faeroes, in order to investigate the substratum of the basalts (Berthelsen et al. 1984). However, the 2.2km deep well failed to penetrate the basalts and a well seismic survey was performed in 1988. This survey included a vertical seismic profile (VSP) and a walk-away VSP. In addition, a refraction seismic experiment was undertaken in 1989. This paper summarizes the interpretation of the three seismic profiles by a combined seismic profiling method. Earlier investigations of the subsurface of the Faeroe Islands include a refraction survey (Palmason 1965), the North Atlantic Seismic Profiling (NASP) project (Nielsen 1983) and a geochemical study of the basalts (Gari~py et al. 1983; Waagstein, 1988).

Geological setting The exposed lava-pile on the Faeroe Islands is divided into three series, Lower, Middle and Upper series, each being c. 1 km thick (Rasmussen & Noe-Nygaard 1970; Rasmussen 1991). The lava flows were extruded subaerially, and the Middle and Lower series are separated by a 10 m thick coal layer. The three basalt series are approximately conformable. The Lower basalt series has tentatively been dated by magnetostratigraphy as being of C26R to C25N age, and the two higher basalt series as C24R (Waagstein 1988); this is in agreement with biostratigraphic dates from intrabasaltic coals containing a terrestrial microflora of Late Paleocene age (Lund, 1983). The three basalt series give similar K - A r near-isochron apparent ages of c. 54 M a (Fitch et al. 1978). The igneous activity that led to the formation of the Lower basalt series has been suggested to have originated from kilometre-long eruption fissures with a N W - S E orientation (Rasmussen 1991). The exposed part of the Lower basalt series is characterized by flows with an average thickness of 20-30 m, and a maximum thickness of 70m. The basalt flows are often columnar, with columns typically being at right angles to

From Scrutton, R. A., Stoker, M. S., Shimmield, G. B. & Tudhope, A. W. (eds), 1995, The Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region, Geological Society Special Publication No. 90, pp. 111-122

111

112

L. KIORBOE & S. A. PETERSEN

Fig. 1. Map showing the location of the three seismic profiles. Profile 1 is a refraction profile, recorded at the Lopra-1 well site. Profile 2 is a VSP, with the source just east of the well site, and recorded in the Lopra-1 well. Profile 3 is a walk-away VSP recorded at a depth of 2 km and the source has an increasing distance to the northeast from the well-head. Geology in the vicinity of the well is partly from Hald & Waagstein (1984).

the layering. Between eruptions the basalt layers were eroded and volcaniclastic sediments and thin coal seams were deposited. The Lopra-1 well was drilled vertically in the Lower Basalt series, near the culmination of a dome structure on Suduroy, starting c. 750m below the top of the Lower basalt series and c. 100 m above the oldest exposed lava sequence. The basalt flows dip 3-7 ~ to the northeast, and the area is characterized by two sets of master joints (Fig. 1). The master joints strike N W - S E and N E - S W , are nearly vertical with only a slight offset and are heavily brecciated. A few joints, all striking NE, are intruded by dykes (Hald & Waagstein 1984). A total of 120 lava flows with an average thickness of 20m were drilled. One third of the lava flows are capped by a thin (0.5-2m) sediment layer. The upper part of each lava flow is very rich in vesicles due to the rapid cooling; thin lava flows are thus vesicular throughout, whereas thicker lava flows contain only few vesicles in the lower part (Hald & Waagstein 1984). Two dolerite intrusions were

observed during drilling. On re-opening the Lopra-1 well in 1983, gas was discovered at a pressure of 20 bars; the gas consisted of methane (72%) and nitrogen (27%). Isotopically, the methane is similar to gas thermally generated from organic matter, of marine (Jacobsen & Laier 1984) or lacustrine origin (Jacobsen & Laier pers. comm.). Oil film extracted from water samples from the well was analysed by gas chromatography; the gas chromatogram of the higher hydrocarbons and the pristane :phytane ratio are characteristic of a crude oil generated from organic matter of marine origin. Jacobsen & Laier (1984) state that an increasing gas:water ratio with depth indicates a source beneath the basalts.

A combined seismic method The combined seismic method employed here uses three different geometric arrangements: a surface seismic profile, a zero-offset vertical seismic profile (VSP) and a walk-away VSP. The surface seismic profile is recorded at the top of

SEISMIC STUDY OF F A E R O E BASALTS

113

Fig. 2. (Top) The three seismic profiles in accordance with the combined seismic method, coupled at their point of identical recording. The advantage of merging the three profiles is that reflections can ideally be followed from one profile to the next. (Bottom) Interpretation of the same data. The interpretations are also shown on the individual profiles on Figs 3-5.

114

Source type Source volume (1) Source position* Source depth Receiver type Receiver position* Receiver depth t

L. KIORBOE & S. A. PETERSEN Profile 1 refraction

Profile 2 VSP

Profile 3 walk-away VSP

Airgun array (3) 58.8 2.6-110km to the NE 18m b.m.s.1. Five three-component geophones 10 m to the north of wellhead 3m

Airgun array (4) 4.9 153m to the E 5 m b.m.s.1. One three-component geophone Along well track 220-2158m (RKB):~

Airgun array (4) 4.9 0.8-3.5 km to the NE 5 m b.m.s.l. One three-component geophone Along well track 220-2158m (RKB):~

* Position reference measured from the Lopra-1 well site. t depth reference 11.5 m above sea level = elevation of kelly bushing. RKB, referenced to kelly bushing.

Fig. 3. (Left) Profile 1 displayed between 0 and 40 km and a reduction time of: time~iistance/6.0 (s). (Right) Interpretation of the same profile. Note that the direct P- and S-wave becomes very weak or disappears at distances > 26 km. The shallowest P-reflection is interpreted as the base of the basalt. the well. Zero-offset registration o n the surface seismic profile is i d e n t i c a l to a z e r o - d e p t h registration on the zero-offset VSP. T h e walka w a y V S P is collected by m e a n s o f a g e o p h o n e

at a fixed level in t h e well, A zero-offset registration o n t h e w a l k - a w a y V S P is identical to a registration o n the zero-offset V S P at the same level. T h u s , the three different seismic

SEISMIC STUDY OF FAEROE BASALTS

115

profiles can be coupled at their points of identical registration. The advantage of combining the three profiles is that reflections ideally can be followed from one profile to the next (see also Kanestrom 1983). Figure 2 demonstrates that despite small gaps between profiles reflec.. tions can be followed. This combined seismic method has the potential to improve interpretation, especially in areas with acoustic impedance problems such as in volcanic terranes. Areas with high contrasts in acoustic impedances, such as the Lopra-1 area, will generate a large amount of S- and surface waves, which will suppress the P-wave reflections. However, the S-waves will show up in different parts of the three different seismic profiles. Thus, reflections can be identified on a segment of one of the seismic profiles, and correlated through the S-wave 'noise' to the next profile. In areas where the high acoustic impedance causes problems with the sea b o t t o m and interbed multiples, the VSP is particularly advantageous, because it is possible to remove multiples generated at surfaces between the shallowest and the deepest records during the processing of the VSP (VSP deconvolution techniques).

Data acquisition Three seismic profiles are presented in this paper; the locations of the profiles are shown on Fig. 1. For convenience profiles 1 and 3 are displayed on either side of profile 2, although in reality they both extend in a northeasterly direction from the well (see Fig. 2). The VSP data (profiles 2 and 3) were collected in June 1988 by the Geological Survey of Denmark in association with the contracting company Prakla-Seismos AG. The refraction data (profile 1) were collected by Faeroe Islands Natural History Museum, and the Institute of Solid Earth Physics, University of Bergen, Norway, in June 1989. Bad weather conditions during the collection of profile 1 necessitated lowering the airguns to a depth of 18 m.

Processing Profile 1 The five channels were stacked for each of the three components. Figure 3 displays the vertical component with a bandpass filter of 5-8 to 4550Hz, and with a reduction time of 6 k m s -1. Part of profile 1 is shown on Fig. 2 with the components rotated according to the direction

Fig. 4. (Left) Profile 2, the processed zero-offset vertical seismic profile (VSP) displayed in two-way travel time (two and between depths of 0.2 and 2.2 km. (Right) The interpreted profile. The P-reflection at a depth of 0.92 s twt is interpreted as the base of the basalts. of maximum polarization, displayed with an equal distance between the traces, and a bandpass filter of 5-8 Hz to 45-50 Hz.

Profile 2 The VSP has a processing route as follows: downgoing compressional wavefield suppression, deconvolution and pulseshaping (removal of interbed multiples), narrow band reject filtering to reject P and S diffractions and partly fault plane reflections. The VSP sections are displayed with a correction for spherical divergence. The VSP is displayed on Fig. 2 in oneway time and with a high-cut filter of 25 Hz. On Fig. 4 the VSP is displayed in two-way time (twt); no filter is applied. In addition, no special processing has been invoked in order to enhance the reflection pattern. A corridor stack of profile 2 is shown on Fig. 8 together with an estimated acoustic impedance log (Kiorboe & Petersen 1991).

116

L. KIORBOE & S. A. PETERSEN

Fig. 5. (Left) Profile 3, the processed walk-away VSP, recorded at a depth of 2 km in the Lopra-1 well, showing the in-line transverse component where Preflections are predominant (perpendicular to transmitted P). (Right) Interpretation of the same profile. The base of the basalts is interpreted as the second Preflection from top. Note the reflection-free interval below the base of the basalts.

Profile 3 The walk-away VSP data were subjected to only minor processing. Basically, the three component data were rotated according to the polarization of the first arrival (see Kanestrom et al. 1986; Petersen et al. 1987). The in-line transverse component is the component where P-reflections will show up (Fig. 5). On Fig. 2, the walk-away is displayed with an equal distance between the traces and a high-cut filter of 25 Hz.

Interpretation T r a n s m i t t e d events The three seismic profiles show distinct compressional (P) and shear (S) energy arrivals. The direct S-wave is generated at interfaces in the basalt succession at or just below the sea bed

Fig. 6. Raw data from the zero-offset VSP, profile 2, vertical component. The first arrivals exhibit strong near-surface reverberation which suddenly disappears in a high attenuation zone at 6-700 m depth.

and interferes partly with groundroll in the surface recording (profile 1). The low noise level before the first arrivals indicates the very high quality of the VSP data (profiles 2 and 3). The first arrivals in profile 2 exhibit strong near-surface reverberation which suddenly disappears due to a high attenuation zone at a depth of 6-700 m (see Fig. 6). High noise activity before first arrival, and in the same frequency domain as the first arrival, signifies the lower quality of the surface recording (profile 1; Fig. 3). In addition, strong water layer reverberations together with reverberations from the bubble pulse, decrease the data quality. The P interval velocities can be directly derived from the first arrivals of profile 2. The P-velocity throughout the basaltic pile seems rather constant, ranging from 4.8-6.2 km s-1 (see Fig. 7). On profile 1 (Fig. 3) the direct P-wave has an apparent velocity of 4.8 kms- 1 between 2.3 km and the cross-over distance at 14 km. The cross-over distance is characterized by strong polarization, as shown in Fig. 9. The first arrival P-wave continues with an apparent velocity of 5.2 km s-~ to a distance of 26 km where the first

SEISMIC STUDY OF FAEROE BASALTS

117

Fig. 7. A lithostratigraphic log of the Lopra-1 well is shown on the right. The basalt layers are shown locally with capping sediments. On the log of the occurrence of basaltic flow boundaries a high occurrence indicates thin beds, while low occurrence indicates thick beds. Interval P-velocities measured are from the first arrival in the VSP. The P-velocities correlate with the occurrence of basaltic flow boundaries; thick flows correspond to high velocities. The interval S-velocities are measured from the scattering direct S-wave. The Svelocities correlate with the P-waves, but Vp/Vs decreases slightly with depth. P-velocities in model 2, from forward modelling of profiles 2 and 3. arrival disappears. A refraction profile 700m long (not shown) crossing the Lopra-1 well from the profile-2 source (shooting to the WNW), shows an apparent P-velocity of 4.8 km s-1. The S-interval velocities on profile 2 are not easily estimated due to the scattering of the direct S-energrv; they typically range between 2.1 and 3 . 6 k m s - (Fig. 7). The apparent S-velocity 1 in profile 1 is 2.47 km s- with an offset between 2.6 and 6.2km (see Fig. 3). The apparent Svelocity increases to 2.53 km s-1 at a distance of 12.2km and to 3.1 k m s -1 at 26km. The 700mlong profile refraction profile, mentioned above, shows an apparent S-velocity of 1.9 km s-1 at an offset of 200m and 2 . 5 k m s -1 with an offset of 200-770 m from the profile-2 source. The seismic set-up around Lopra-1 makes it possible to measure the energy spread through the basalts in both the vertical and the horizontal directions for certain depths. The higher velocities measured on the zero-offset VSP compared to the lower refraction velocities on the surface recordings suggest a velocity anisotropy with higher vertical velocities than horizontal velocities for both P- and S-waves.

Reflected events Evidence of reflected P-waves is very clear on profile 2 (the VSP) in a broad corridor down to

Fig. 8. (Left) The P-velocities in models 1 and 2. (Right) An acoustic impedance log and a corridor stack derived from the zero-offset VSP. The events at 1.7-1.8 twt and 2.1-2.2 twt correlate to the 4.6-4.7 and 5.6-5.78 km events in the models.

1.0 s twt and downwards from 1.7 s twt (see Fig. 4). In between, remnants of P-S-converted reflections, perhaps from fault planes, suppress the primary reflections. Deeper reflectors hidden by the noise band at small distances (zero-offset) become visible in the time window for distances > 2.5 km on profile 3 (the walk-away VSP; Fig. 2). On the surface recording, profile 1, the reflected P-waves are more difficult to follow; the most prominent reflectors are shown on Fig. 3. All the interpreted events are shown on Fig. 2, and it is clear that the combined seismic method significantly improves confidence in the interpretation, particularly of the surface recordings (profile 1). Reflected S- to S-waves are evident on profile 1 (Fig. 3), but are not recognized on profiles 2 or 3, possibly because of the scattering of energy from the fault planes. However, reflected events with apparent velocities between P- and S-waves are observed from the direct P-waves on profile 2 (Fig. 10), and at the same depth as the observed

118

L. KIORBOE & S. A. PETERSEN

2,3-

5.0

~-'~--

E

5.0

0.40 -

~ J

-~

--

--

_~--

0.85 -

~

--~

17.5

20.0

-~52

20.0

22.5~

111rl i 1.15-

~.

1.30-

-o

17.5

1.00-

~'~ e-

15.0

__.~__~

1.5-

0.70-

10.0 12.5

15.o

/

1.0-

0.55 -

~--~--~

"E

125 5.6

0.25-

7.5 q

7.5 - - ~ J 10.0

2.3] ~

1.45 -

E

2.0-

'~

2.5-

1.60 -

1.75 -

3.0-

1.90-

22.5

2.05"

3.5

25.0 -4

-3

-2

-1

0

-75

-50

-25

log polarisation

declination

Refraction Profile-1

-75

-50

-25

0

-75

-50

-25

declination

declination

Refraction

VSP

Walkaway-VSP

Profile-1

Profile-2

Profile-3

0

Fig. 9. Log polarization and declination of the direct P-wave.

P-reflections. These 'reflected events' (Fig. 10) have an average velocity of 3.6 km s-1, which is much larger than that observed on the direct Swave (see Fig. 7). It is therefore assumed that the 'reflected events' shown on Fig. 10 are scattered P to S converted events. However, the coincidence of the initiation point of the reflected Pwaves and the 'reflected events' is remarkable, as is the high velocity of the reflected events. Reflected events with a velocity of c. 4 km s-1 are observed deeper in the section (Fig. 11). It is reasonable to assume that these events originate from interfaces which also give strong P-reflections, such as the downhole observation shown on Fig. 10. If so, and if the reflected events continue with the same velocity below TD, the depth to the interface can be determinated as the point of coincidence between the 'reflected event' and the reflected P-wave (Fig. 11).

Modelling The 2D seismic modelling package (Froyland

et al. 1988) developed by the Seismological Observatory, University of Bergen and Norsk Hydro Research Centre, was used to build the seismic models in the Lopra area. The model file contains information of seismic velocities, source and receiver positions. The same model could be used for the three seismic profiles by shifting the source and receiver

positions. A depth-dependent velocity distribution or a constant velocity was specified for each layer and horizontal layering assumed. The kinematic ray-tracing is based on initial value tracing, the so-called 'shooting' technique. The initial direction measured in angles and the increment are the basic shooting parameters. The initial direction for the first ray is specified, and then the 'shooting' of rays through the model continues until the receiver line is hit by the desired number of rays. The model respe~se is travel times. The velocities, or the depth to the interfaces, in the model were changed until the model output and the interpretation of the seismic profiles were identical. Information about the inclination (see Fig. 9) of the direct wave on the three profiles and the polarity of the direct wave on profile 1, was used as an additional constraint during modelling. The direct P-wave on the zero-offset VSP (profile 2) was used to build the velocity distribution of the upper 2.158m around the Lopra-1 well. Prominent reflections on profile 2 were chosen as interfaces in model 2. The model was then expanded to the east to include the walk-away VSP, profile 3. The sea-bottom topography and a thin sediment layer were incorporated into the model. The velocity distribution was modified to satisfy the declination of the direct P-wave on profile 3 together

SEISMIC STUDY OF FAEROE BASALTS

Fig. 10. Profile 2. Residual after removal of horizontal events (the P-reflections). The section shows reflected events (arrows) with apparent velocities between Pand S-waves, and with the same initial point as the Preflections.

with profile 2. The velocity distribution for the upper part of model 2 is shown on Fig. 7. The apparent P-velocity on profile 1 is 4.8 km s-1 to the cross-over distance of 14km. The first arrival P-wave continues with an apparent velocity of 5.2 km s-1. The velocities from model 2 are obviously too high to satisfy the data from profiles 3 and 1, and it was necessary to build a new velocity model with lower velocities in the upper part, model 1. The direct P-wave to 14 km was modelled by an 800m thick layer with an increasing velocity from 4.8-5.05 km s-1 (Fig. 8). To satisfy the declination of the P-wave on profile 3 a low-velocity zone of 5.0km s-1 was included between depths of 0.8 and 1.3 km. The first arrival P-wave continues to a distance of 26km in two layers between depths of 1.3 and 2.34 km and with increasing velocities from 5.15 to 5.45 km s-1. Both models 1 and 2 satisfy the walk-away VSP data (Fig. 12). Models 1 and 2 are identical at depths > 2.34km. Further modelling is based on the

119

Fig. 11. Profile 2. This VSP section is the same as that in Fig. 2. Reflected events with apparent velocities between P- and S-waves are observed below TD. These events have been extended to points of coincidence with prominent P-reflections. interpretation of the seismic reflections on profile 2 at depths of 1.75, 2.1 and 2.2s twt, and the assumption that the velocities only vary with depth, as the reflectors are approximately horizontal (see Fig. 4). Mode conversion from P to S was observed downhole (Fig. 10). Under the assumption that mode conversion takes place at the same interface as the P-reflections observed in the deeper section of profile 2, the interpretation of P- and S-reflections was extended to their point of convergence (Fig. 11). The time and depth to the reflectors are easily measured on Fig. 11, and were incorporated in the models. An average velocity from 5.0 to 5.35kms -1 was used to between 2.34 and 5.78 km depth, to fit the reflections on profiles 1 and 2. This interval appears as a low-velocity zone, in agreement with the interpretation of profile 1 where the first arrival P-wave is missing at distances > 26 km. Also the refraction seismic profiles from the North Atlantic Seismic Profiling (NASP) project (Nielsen 1983) lost the seismic energy with a

120

L. KIIDRBOE & S. A. P E T E R S E N

SEISMIC STUDY OF FAEROE BASALTS source receiver distance of similar amount which supports a low-velocity zone. Two prominent reflectors on profile 3 were also included in the models. The ray-tracing models and the model response are shown on Fig. 12.

Discussion The apparent high attenuation of the direct wave from 600-700 m is interpreted to be caused by fracturing in connection with the upper doleritic intrusion. A recent study by Pujol & Smithson (1991) suggests that wave propagation in volcanic rocks is not affected by attenuation to a greater extent than in sedimentary rocks. The interval velocities from the VSP of both the P- and S-waves are shown on Fig. 7, together with a lithostratigraphic log and a log of the occurrence of basalt-flows in the Lopra-1 well. From the figure it is clear that the P-velocities depend on the lithostratigraphy. Thin beds are associated with low P-velocities, whereas thick beds are associated with high P-velocities. This is in agreement with the observed distribution of vesicles. Thin lava flows are vesicular throughout and have low velocities, while the thicker lava flows only contain few vesicles in the lower part, are more massive in the upper part and exhibit higher velocities. The direct P-waves on profile 2 have travelled through the rock in a nearly vertical direction particularly at depths > 400m (Fig. 9). The apparent velocities from profile 1, on the other hand, resemble horizontal velocities. Comparing the two models it appears that the upper 0.8 km of the basalt beds exhibits a velocity anisotropy with c. 10% higher velocities in the vertical direction. This is interpreted as a result of the vertical fractures around basalt columns possibly in combination with the nearly vertical master joints. The joints and the fractures will act to reduce velocities in the horizontal direction. The base of the basalt is interpreted to be at a depth of 0.92 s twt, or 2.34 km. This is based on

121

the presence of a strong reflector with a negative reflection coefficient, the generally lower velocities below that reflector and the fact that the reflection free stratigraphic interval between 0.92 and 1.35 s twt (observed on profile 3) has a different seismic response compared to the reflection-rich downhole measurement. The rocks below the 0.92s twt level have been modelled using velocities of 5.0-5.35 km s-1. This represents rather high velocities for sedimentary rocks. The very strong reflectors below 1.35 s twt are interpreted as sills. Numerous sills and dykes may, however, raise the average velocity of an otherwise lower-velocity rock. In general, the S-velocities vary systematically with the P-velocities. There is, however, a more general increasing trend in the S-velocities, as the S-waves are more influenced by the closure of fractures downhole. The Vp/Vs (P-velocity/S-velocity) relationship decreases from 2.0 to 1.85 downhole on profile 2, while on profile 1 the Vp/Vs relationship decreases from 1.94-1.89, to 1.68 for distances of 6.2-14 and 2 6 k m respectively. The Vp/Vs relationship in rocks depends on several factors, including mineralogy, porosity, fractures and fluid types. In particular, an increasing quartz content will cause a decrease in lip and an increase in Vs. The porosity also affects Vp and Vs differently; Vs is affected much more by porosity and cracks. It is unlikely that a simple two-parameter (say Vp/Vs and lip) discriminant can be used between rock types, but once the lithology is defined, other variations can be determined. It is not easy, and perhaps not possible, to find corresponding P- and S-reflections. However, the S-reflection (full line) identified on Fig. 3 might originate from the same interface as the very strong P-reflection observed in the refraction profile, the zero offset VSP and the walk-away VSP. If so, the Sreflection from the corresponding reflector gives an average S-velocity of 3 . 2 k m s -1, and the corresponding Vp/Vsof 1.64. This is a rather low Vp/Vs ratio and indicates a change in mineralogy, porosity or pore content at that depth.

Fig. 12. (Top) The model response as travel-time curves. (Middle) Models 1 and 2 with rays. The velocities are shown on Fig. 8. The models are basically 1D except for the sea-bottom topography and a thin sediment layer. (Bottom) The model response shown on the three seismic profiles (profile 3 is shown twice). To the left is model 1 illustrating the wave propagation on profile 1. Only 17.5km is shown of the 40km-wide model; 0km is the receiver position at the Lopra-I well. The direct P-wave is modelled to 26km in accordance with the interpretation (see Fig. 3). In the middle is model 2 used on profile 2. The horizontal axis is extended in order to illustrate the wave-propagation, as the distance between the source and the well only is 153m. To the right is model 2 used on profile 3, between the Lopra-1 well at 0 km and the maximum offset of 3.5 km. To the far right is model 1, also used on profile 3. The average velocity of the two models between depths of 0 and 2.34km are similar as seen on the identical model response on the first arrival when used for profile 3.

122

L. KIORBOE & S. A. PETERSEN

Conclusions The different seismic techniques used around the Lopra-1 well provide an excellent opportunity to study the seismic behaviour of the plateau basalts. This combined seismic method can significantly improve the interpretation of surface seismic profiles in a region with 'volcanic problems'. The seismic experiments have revealed some important characteristics of the acoustic behaviour of the Faeroe platform. In similar investigations in the Colombia plateau basalt region with thick interflow sediments (Pujol et al. 1989), special attention was focused on near-surface reverberations, destructive interference and strong m o d e conversions in order to explain the difficulties of achieving a good surface seismic response. In the Faeroe case, with few and thin interbasaltic sedimentary beds, closely comparable phenomena are observed. However, structural features, such as master joints and cracks seem to play a more important role. The master joints and cracks are also t h o u g h t to be responsible for the observed velocity anisotropy in the upper 800 m. The downhole response from the basaltic stratigraphic sequence is as good as the response obtained in a normal sedimentary siliciclastic sequence. Only minor conversion or strong intrabed multiple activity are recognized downhole along the first-arrival corridor in the raw and processed zero-offset VSP sections. The main reason for the poor surface seismic response is related to the presence of widely distributed, steeply dipping joints and dykes, resulting in a large-scale scattering and conversion of the seismic energy. On profile 1 the small signal:bubble ratio does, however, also play an important role in the poor quality of the surface seismic data. The base of the basalt is interpreted to lie at a depth of 2.34 km in the area around the Lopra- 1 well, based on the reflectivity, the lower velocities and the different seismic response below that depth compared to the downhole response. The composition of the substratum of the basalts is, however, unknown. The very prominent reflectors at depths of 1.35-1.5, 1.7-1.8 and 2.1-2.2s twt are interpreted to originate from intruded sills. We thank the Geological Survey of Denmark and the Faeroe Islands Natural History Museum for permission to publish this paper. We thank Morten Sparre Andersen, Yngve Kristoffersen and colleagues at the Department of Solid Earth Physics, University of

Bergen, Norway; Norsk Hydro Research Centre, Bergen and the Geological Survey of Denmark for ideas and discussions. Helle Zettervall drafted the figures. This project was financially supported by the Faeroe Islands Natural History Museum.

References BERTHELSEN,O., NOE-NYGAARD,A. & RASMUSSEN,J. 1984. The Deep Drilling Project 1980-1981 in the Faeroe Islands. Foroya Frodskaparfelag, Torshavn, 9-10. FITCH, F. J., HOOKER, P. J., MILLER, J. A. & BRERETON, N. R. 1978. Glauconite dating of Palaeocene-Eocene rocks from East Kent and the time-scale of Palaeocene volcanism in the North Atlantic region. Journal of the Geological Society, London, 135, 499-512. FROYLAND, L. m., HELLE, H. B. & PAJCHEL, J. 1988. 2D Seismic Modelling Package, User's Manual, Version 2, 4.10.1988. Norsk Hydro, Norway. GARII~PY,C., LUDDEN,J. & BROOKS,C. 1983. Isotopic and trace element constraints on the genesis of the Faeroe lava pile. Earth & Planetary Science Letters, 63, 257-272. HALD, N. & WAAGSTEIN, R. 1984. Lithology and chemistry of a 2-km sequence of Lower Tertiary tholeiitic lavas drilled on Suduroy, Faeroe Islands (Lopra-1). In: BERTHELSEN,O., NOE-NYGAARD, A. & RASMUSSEN, J. (eds) The Deep Drilling Project 1980-1981 in the Faeroe Islands. Foroya Frtdskaparfelag, Ttrshavn, 15-38. JACOBSEN,O. S. & LAIER, T. 1984. Analysis of gas and water samples from the Vestmanna-1 and Lopra-1 wells, Faeroe Islands. In." BERTHELSEN,O., NOENYGAARD, A. & RASMUSSEN, J. (eds) The Deep Drilling Project 1980-1981 in the Faeroe Islands. Ftroya Frtdskaparfelag, Ttrshavn, 149-157. KANESTROM, R. 1983. Borehull-geofysik: VSP og Borehulls-gravimetri. Forl~esnings-notertil kursus i borehull-geofysik, Jordskjelvstasjonen. Inst. f. Fastjordsfysik, Universitet i Bergen, Norge. , PETERSEN, S. A. & NESSE, T. A. 1986. Threecomponent vertical seismic profiles: P and S wave velocity determinations. 48th meeting, The Casino, Ostend, Belgium, 3-6 June 1986, 28. KIORBOE, L. & PETERSEN, S. A. 1991. Seismic studies in the vicinity of the Lopra-1 well. In: European Association of Exploration Geophysicists, 52nd Meeting and Technical Exhibition, Bella Center Copenhagen, Denmark 28 May-1 June 1990, 166. LUND, J. 1983. Biostratigraphy of interbasaltic coals from the Faeroe Islands. In." BOTT, M. H. P., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) Structure and Development of the GreenlandScotland Ridge. Plenum Press, New York, 417423. NIELSEN, H. P. 1983. Geology and structure of the Faeroe Islands - a review. In: BOTT, M. H. P., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) Structure and Development of the GreenlandScotland Ridge. Plenum Press, New York, 77-90. PALMASON, G. 1965. Seismic refraction measurements

SEISMIC STUDY OF FAEROE BASALTS of the basalt lavas of the Faeroe Islands. Tectonophysics, 2, 475-482. PETERSEN, S. A., KANESTROM, R. & NESSE, T. A. 1987. Anvendelser av 3-komponent geofonsystemer i offset VSP. Tillreg til: KANESTROM, R. (ed.) 1983: Borehull-geofysik: VSP og Borehulls-gravimetri. Forlcesnings noter til kursus i borehullsgeofysik, Jordskjelvstasjonen. Inst. f. Fastjordsfysik, Universitet i Bergen, Norge. PUJOL, J. 8r SMITHSON, S. 1991. Seismic wave a t t e n u a t i o n in volcanic rocks f r o m VSP experiments. Geophysics, 56, 1441-1455. , FULLER, B. N. & SMITHSON, S. B. 1989. Interpretation of a vertical seismic profile con-

123

ducted in the Colombia Plateau basalts. Geophysics, 54, 1258-1266. RASMUSSEN, J. 1991. The Origin of the Faeroe Islands, in Text, Pictures and on Maps. Geological Survey of Denmark. - & NOE-NYGAARD, A. 1970: Geology of the Faeroe Islands. Danmarks Geologiske Undersogelse. First series, 25. WAAGSTEIN, R. 1988. Structure, composition and age of the Faeroe basalt plateau. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 225238.

Seismic stratigraphy of the Bill Bailey and Lousy Bank area: implications for subsidence history K. V A N N E S T E , 1 J.-P. H E N R I E T , 1'2 J. P O S E W A N G 3 & F. T H E I L E N 3

1Renard Centre of Marine Geology, University of Gent, Krijgslaan 281, B-9000 Gent, Belgium 2present address." IFREMER, D~partement des Gdosciences marines, 29280 Plouzand, France 3Institut far Geophysik, University of Kiel, Olshausenstrafle 40-60, D-2300 Kiel, Germany Abstract: The present study discusses some results of a reconnaissance seismic survey carried

out in 1988 in the area around the prominent Bill Bailey and Lousy Banks, located SW of the Faeroe Islands. The area is thought to be part of the Rockall-Faeroe microcontinent, which was flooded by Palaeocene plateau basalts. The seafloor topography is largely due to the present-day organization of the basalt surface. Important structural features exhibited by this basement are sub-surface diverging reflectors and a major fault. Sediment accumulation is confined to the basinal area where four unconformities were identified. The lower boundary corresponds to the well-known reflector R4 of the North Atlantic. The other unconformities are proposed to correlate with hiatuses at the beginning and end of the middle Miocene, and at the end of the Pliocene, respectively. The sediment units are briefly described by their upper and lower boundary, seismic facies and thickness variations. About the lithology, however, little is so far known. The sediments are deformed by intraformational faults and a diapir-like structure. Both deformational styles are probably related to temporary overpressurisation in fine-grained sediments, but resulting from different causes. Stratal geometric patterns indicate that initially the subsidence of the basalt surface was rather uniform, but became non-uniform after the Eocene/Oligocene boundary, differentiating the two banks from the surrounding basin.

In M a y - J u n e 1988 a reflection seismic survey was set up jointly by the Geophysical Institute of Kiel (Germany) and the Renard Centre of Marine Geology in Gent (Belgium). The aim was to investigate the sedimentary patterns on the flanks of the Iceland-Faeroe Ridge. The area southwest of the Faeroe Islands, marked by prominent seamounts, was involved in the investigations: six reconnaissance profiles were shot across the Bill Bailey and Lousy (or Outer Bailey) Banks. An airgun array totalling between 2.61 and 3.81 was used as acoustic source. The reflected signals were detected by a 24channel streamer with an active length of 600 m. Only the analogue monitor recordings and the common-offset plots are presented here. The Bill Bailey and Lousy Banks delimit the northern edge of the Rockall Trough. Together with the Faeroe Bank these banks link the Rockall Plateau in the SW with the Faeroe insular block in the NE. Their location is shown in Fig. 1, along with the available seismic lines. Seismic (Bott et al. 1971, 1974; Casten 1973) and isotopic (Hald & Waagstein 1983) evidence

shows that the Faeroe Island block may be underlain by continental crust which probably extends beneath the Faeroe, Bill Bailey and Lousy Banks (e.g. Roberts et al. 1983), thus adjoining the Rockall microcontinent in a southwestward direction. This so-called Rockall-Faeroe microcontinent became isolated by two distinct extensional phases: the area was split from the main European continent by an early to mid Cretaceous rifting episode (Roberts 1975; Roberts et al. 1983) which created the Rockall Trough and the Faeroe-Shefland Channel. And in late Palaeocene times the RockallFaeroe microcontinent broke away from Greenland, giving way to the opening of the northern North Atlantic. This second extensional phase was accompanied by the massive subaerial extrusion of plateau basalts which flooded large parts of the microcontinent within a few million years. The presence of a thick pile of Palaeocene plateau basalts is well established on the Faeroe Islands, on the n o r t h e r n R o c k a l l P l a t e a u (Laughton et al. 1972; Roberts 1975) and in the area bounding the northern Rockall Trough

From Scrutton, R. A., Stoker, M. S., Shimmield, G. B. & Tudhope, A. W. (eds), 1995, The Tectonics, Sedimentation and Palaeoceanographyof the North Atlantic Region, Geological Society Special Publication No. 90, pp. 125-139

125

126

K. VANNESTE E T A L .

Fig. 1. Generalized bathymetric map of the area, compiling available seismic grid, basaltic outcrop (from Waagstein 1988), plutonic centres (after Roberts et al. 1983), zone of dipping reflectors (Andersen 1988), oceancontinent boundary (from Smythe 1983), the southern lava front (from Wood et al. 1988), together with relevant DSDP sites. Mercator projection at 60~ N.

(e.g. Roberts et al. 1983; Wood et al. 1987). The edges of the microcontinent were identified on the northern margin of the Faeroe shelf by the recognition of oceanward dipping reflectors (Smythe 1983; Andersen 1988), but have not been detected along the northern flanks of the Bill Bailey and Lousy Banks up to now.

Acoustic basement On all profiles (Figs 2-5) a strong acoustic basement reflector (labelled EE) can easily be identified. Mainly by virtue of its relatively rough topography and hyperbolic character - a reflection pattern commonly regarded as characteristic of basalt - it was interpreted as the top of the ubiquitous flood basalts. As can be readily observed on lines IF001/ IF002 (Fig. 2), the organization of two banks and an intervening basin is already manifest at the level of the acoustic basement. In the narrow

basin separating Bill Bailey Bank from Lousy Bank the acoustic basement appears at a depth of c. 2200 ms two-way travel time (TWT) below the bank crests, a pronounced downwarping which occurs without any major faulting. The pronounced basement relief is strongly reflected in the p r e s e n t - d a y seafloor t o p o g r a p h y , although over 1000ms of sediments have accumulated between the banks. The crests of the two banks bear nearly no sediments and are flattened, which - according to Roberts et al. (1983) - could be due to 'subaerial erosion of an original volcanic landform'. Underneath the basalt surface divergent reflectors can be observed beneath the northern and western flanks of Lousy Bank (line IF004, Fig. 3) and beneath the northern flank of Bill Bailey Bank. Due to the limited capacity of the acoustic source, however, these structures could not be very well elucidated. Roberts et al. (1983) observed similar 'outward'-dipping reflectors on

SEISMIC S T R A T I G R A P H Y OF BILL B A I L E Y A N D L O U S Y B A N K S

127

128

K. VANNESTE E T AL.

_o

O O

8

,,.._w

SEISMIC STRATIGRAPHY OF BILL BAILEY AND LOUSY BANKS their seismic lines across the banks. They interpreted them as volcanic layering. It is possible that these diverging reflectors correspond to the oceanward-dipping reflectors of Hinz (1981) and Mutter (1985) typical of volcanic passive margins. They could build the prolongation of the zone of oceanward-dipping reflectors identified north of the Faeroe Islands by Smythe (1983) and Andersen (1988), and mark the northwestern edge of the RockallFaeroe microcontinent. Another indication that the banks may be bordered to the northwest by oceanic crust can be found in the striking asymmetry of the banks on north-south sections (lines IF004 and IF006): the highest parts are located in the south and the flattened crests gently slope down towards the north (more probably to the northwest). It is also clear from line IF004 that north of Lousy Bank the sedimentary cover starts at a much lower level than south and east of this bank. If the ocean-continent transition really occurs northwest of the banks then the outer edges of the banks can be expected to have undergone enhanced subsidence by the thermal influence of the adjoining oceanic lithosphere. Centrally in the basin between the banks the basement is displaced over > 150 ms by a single fault (both lines IF001/IF002 and IF006). This fault strikes approximately WNW-ESE, downthrowing the basement towards the southwest. Bill Bailey Bank is thus located on the upfaulted side. Neither the mode of faulting nor the origin can be established due to the limited data. The sediments overlying the fault display a large deformational structure, which on the vertically exaggerated profiles shows up as a diapir, possibly originating in a movement of the fault. The broad convex-upward curvature of the basement appearing southward of the fault (line IF006, Fig. 4) may be the expression of an underlying intrusion centre. The location of this structure with respect to the above-mentioned fault suggests that there is no genetic relation between the two structures.

Sediment stratigraphy Introduction The sediment distribution is confined by the location of the banks: the flattened crests bear nearly no sediment, whereas the basin between the banks is filled with c. 1200 ms of sediments. Seismic stratigraphic interpretation allows identification of four regional unconformities in the basin between and southward of the

129

banks. They result in the definition of five depositional units which are briefly described below. The most important results of the seismic stratigraphic analysis (mean thicknesses, unit bounding unconformities and their correlations, geometrical relationships of internal reflectors along these unconformities, etc.) are summarized in the seismic stratigraphic correlation table of Fig. 6. Correlation with the sediments northward of the banks was not possible in this study.

Brief description of the depositional units The five depositional units that were recognized are successively described, from oldest to youngest. The following topics are discussed: (1) lower boundary; (2) upper boundary; (3) seismic facies; and (4) thickness variations.

Depositional unit 1 (1) The lower boundary of this unit, reflector EE, coincides with the top of the Palaeocene flood basalts already described above. Internal reflectors show onlap against the flanks of both banks. (2) The upper boundary of unit 1 is constituted by reflector R4, usually running parallel to the lava surface, except on line IF001/IF002 where R4 pinches out in westward direction against the crest of Lousy Bank. Internal reflectors are truncated by R4. Above the basement fault. observed on lines IF001/IF002 and IF006 it seems that R4 is interrupted by a large diapir piercing out of unit 1. Eastward of this presumed diapir - which we label diapir B - reflector R4 is no longer recognized. A second, though minor diapir (diapir A) occurs more southward, above the basement updoming. (3) Depositional unit 1 is characterized by a transparent seismic facies contrasting with the facies of the overlying units. Only sporadic internal reflectors of significant continuity can be recognized. (4) The depositional unit reaches its maximum thickness of c. 650ms where the basalt surface is at its greatest depth. The interval is apparently thicker in the eastern part of the basin, thins above the southern spur of Lousy Bank (200-300 m) and pinches out against the crest of this bank in the western part of profile IF001/ IF002. As will be discussed in the last section, these thickness variations indicate a different initial subsidence history for the two banks. The local thinning of the sediment interval above the presumed

130

K. VANNESTE ET AL. intrusion centre suggests that updoming of the lava surface started during the deposition of this oldest unit and after that went on for some time, as indicated by the rising of R4 along with the lava surface.

Depositional unit 2 (1) Reflector R4 marks the lower boundary of this unit. Stratal geometry is characterized by onlap of internal reflectors against R4, a relationship which is most prominently expressed around the basin margins. (2) Upper boundary is a reflector provisionally labelled BB3. This reflector shows evidence of a slight erosive truncation of internal reflectors towards the basin edges, the only place where the boundary could be recognized. Except for the eastern flank of Lousy Bank - where the stratal geometry is parallel (see line IF001/ IF002) - BB3 everywhere wedges out against R4. (3) The seismic facies of this second unit varies vertically as well as horizontally. In general it consists of discontinuous subparallel reflectors of medium to high amplitude and frequency. Where the unit reaches its maximum thickness the reflectivity decreases towards the lower boundary, so that the facies hardly differs from that of unit 1. In the zone southward of the banks depositional unit 2 (as well as the overlying units 3 and 4) is intersected by numerous subvertical faults with small offset (e. 10 ms). (4) The thickness of the sediment interval gradually increases from 200ms in the south to almost 400 ms between the banks in the north. This trend is interrupted at the place of the above-mentioned diapiric structures, where the interval thins. This thinning cannot be fully ascribed to differential compaction and is probably due to the initiation of activity of these structures during the time interval corresponding to the deposition of unit 2. Depositional unit 3 (1) The internal reflectors of unit 3 show distinct onlap against reflector BB3, especially at the basin margins and around the diapirs. Remarkably this onlap (as observed on an E-W oriented section, line IF001/IF002) occurs in a westward direction, as opposed to the observed onlap in the underlying unit which occurs in an eastward direction. A change in direction

of the sediment transport is therefore inferred to have taken place at the transition from unit 2 to unit 3. (2) The upper boundary of depositional unit 3 is constituted by a reflector labelled BB4. The boundary truncates the internal reflectors but can pinch out laterally against BB3 or R4. Underlying depositional units are thus not truncated. (3) The seismic facies of unit 3 does not differ markedly from that of the underlying and overlying units 2 and 4. Locally it can become chaotic to reflection free. (4) Unit 3 shows rather complex variations in thickness. The general trend is a thickening of the interval from < 100 ms in the north to 150-200ms in the south, i.e. the opposite trend of the underlying unit. On line IF005 (Fig. 5) the interval (before wedging out in westward direction against reflector R4) displays a remarkable thickening of c. 50 ms which is produced by the upward bulging of all internal reflectors as well as of the upper boundary BB4. The reflections here exhibit higher amplitudes and frequencies than normal and lap out in downward direction against R4. This bulge might have its origin in deposition under current-controlled conditions. Depositional unit 4 (1) The lower boundary of unit 4 is defined by reflector BB4. Stratal relationships with overlying internal reflectors are onlap (again best observed around the diapiric structures) and downlap (in westward direction) which was observed on line IF001/IF002. (2) The upper boundary BB5 is characterized by a distinct erosional truncation. The reflector laps out laterally against R4, precluding truncation of underlying units. Also, BB5 is the sole boundary remaining subhorizontal over the whole area. (3) The seismic facies is analogous to the facies of units 2 and 3: discontinuous subparallel reflectors of medium to high amplitude. (4) The thickness of the sediment interval varies strongly between 100 and 200 ms; in between the two diapirs the unit is only 25-50 ms thick. Depositional unit 5 (1) Reflector BB5 marks the lower boundary of unit 5. Laterally BB5 onlaps against R4. Internal reflectors run parallel to BB5. (2) At the top unit 5 is bounded by the

SEISMIC STRATIGRAPHY OF BILL BAILEY AND LOUSY BANKS

131

Fig. 5. Line IF005 (common offset plot). The blow-up is in fact a smaller-scale analogue monitor recording.

seafloor which generally parallels the internal reflectors. (3) The seismic facies of unit 5 is fairly uniform: continuous, parallel and subhorizontal reflectors with high amplitude. The high amplitude reflections contrast to the facies of the underlying unit. Furthermore, the reflectors are no longer interrupted by subvertical faulting. (4) The thickness of unit 5 is fairly constant and typically amounts to lOOms. The thickness decreases towards the basin margins and also above diapir B.

A concluding remark concerns the notable E-W asymmetry of the basin between the banks (see line IF001/IF002). At the level of the lava surface the basin is still relatively symmetrical. Above this surface this is no longer the case. Unit 1 pinches out in a westward direction against Lousy Bank, while in the east the interval is almost parallel to the lava surface forming the flank of Bill Bailey Bank. For the overlying unit the situation is different: unit 2 is largely parallel to R4 above the flank of Lousy Bank and laps out against the same boundary eastward of diapir B. This geometry suggests

132

K. VANNESTE E T AL.

D EP OS ITIONAL UNITS

5 (100m)

4 (150m)

SIGNIFICANT Geometrical relations

REFLECTORS

concordance truncation

BB5.0

IN

HIATUS

GEOLOGICAL TIME

- --

SCALE

&

BY

ORIGIN

JENKINS

(1986)

2-2.5 M a

:irculation effect lithification boundary

late Miocene late M.Miocene

B B 4.0

ANALYSIS

PEARSON

late Pliocene

-R1--

ordap t~uncation

PLACE

Miller & Tucholke (1983)

Present study

9 Ma . . . .

=irculation effect

3 (200m) onlap truncation

BB3.0---

R2

facies change

BB2.1

R 3

--

early M,Miocene -. .

-

2

.

.

.

onlap slight t~uncTtio"~ - -

1:, A

:irculation effect lithification boundary

Oligocene/Miocene boundary mid-Oligocene

(300m)

15 M a - -

30 M a

Eocene/Oligocene _ boundary

_

_

_

An(37? ) ~ . . . . .

sea level drop eircqtl atlnn

offoot

1 (600m)

concordance/onlap

EE

Paleocene

-

basalt/sediment boundary

Fig. 6. Seismic stratigraphic correlation table for the area around Bill Bailey and Lousy Banks.

that Lousy Bank acted as sediment source during deposition of unit 2. The transition from unit 2 to unit 3 is characterized by a change in direction of lapout: all overlying reflectors show onlap or downlap in a westward direction. This indicates a change in direction of the sediment transport from eastward to westward pointing towards Bill Bailey Bank as the new source for the deposited sediments.

Correlations The lack of well control in the area under investigation makes correlation of the identified unconformities very difficult to establish. Moreover, the sequence stratigraphic approach, although successful for passive continental margins, does not seem to apply in this deep basinal area. Therefore, the correlations have to rely on results published by other authors. Earlier stratigraphic studies were published by, amongst others, Roberts (1975) and Miller & Tucholke (1983) for the Rockall area, Roberts et al. (1983) for the area SW of the Faeroe Islands,

and Wood et al. (1987) for the small area south of Ymir Ridge. Where possible, some of these results were drawn into our interpretation. Additional useful information came from a hiatus analysis of all DSDP drillholes in the NE Atlantic by Pearson & Jenkins (1986). They filtered out all hiatuses having a local tectonic origin and came to the following results, representing regional hiatuses which are expected to occur throughout most of the NE Atlantic: (1) two hiatuses in the Middle Eocene which are difficult to recognize outside the Rockall area; (2) a well-known hiatus at the Eocene-Oligocene boundary (c. 37Ma); (3) a mid-Oligocene hiatus (e. 30 Ma) corresponding to the mid-Oligocene sea-level drop postulated by Vail et al. (1977); (4) a well-defined hiatus at the beginning of the Middle Miocene (15Ma), probably corresponding to reflector R2 of Miller & Tucholke (1983); (5) a widespread hiatus at the end of the Middle Miocene (9 Ma); (6) a hiatus at the end of the Pliocene (2-2.5 Ma). Our lowermost unconfon'aity is supposed here to correspond to the hiatus at the Eocene-

SEISMIC STRATIGRAPHY OF BILL BAILEY AND LOUSY BANKS Oligocene boundary. This unconformity marks the most prominent boundary. It is an outstanding onlap surface for overlying reflectors, and also corresponds to a striking facies change (e.g. line IF005; Fig. 5): below the seismic facies is largely transparent, while above the boundary numerous reflectors can be recognized. This characteristic suggests correlation with the well-known reflector R4 of the North Atlantic (Jones et al. 1970; Roberts 1975), which is supported by the analogy with a seismic line published by Roberts et al. (1983, fig. 3). This reflector was dated at the Eocene-Oligocene boundary (c. 37Ma) (Laughton et al. 1972; Roberts 1975) and can be traced throughout the Rockall area (Roberts 1975; Miller & Tucholke 1983), the eastern Reykjanes Basin (Ruddiman 1972), the Labrador Sea (Egloff & Johnson 1975), and perhaps even in the Bay of Biscay (Montadert & Roberts 1979). The younger unconformities were provisionally labelled BB3-BB5, corresponding to the number of the overlying unit. The uppermost unconformity, BB5, resembles the uppermost unconformity identified on seismic lines shot during the same survey on the northern flank of the Iceland-Faeroe Ridge. Interpretation parallel to the present study (Vanneste 1989) suggests that this reflector is of Pliocene age, as constrained by its depth at the intersection with D S D P drillhole 336. Although it was not recognized as a hiatus in drillhole 336 (Talwani et al. 1976), the reflector is likely to correlate with the hiatus at the end of the Pliocene common to most of the other DSDP drillholes in the Norwegian Sea. As presented in the correlation table of Fig. 6, unconformities BB3 and BB4 are proposed to correlate with the hiatuses at the beginning and end of the middle Miocene, respectively. While the other correlations are relatively firm, the dating of these two middlemost reflectors is less reliable and is primarily based upon comparison with the hiatus analysis of Pearson & Jenkins (1986), and upon the assumption (e.g. Miller & Tucholke 1983) that regional deep sea hiatuses in this area have their origin in a change of the bottom-current regime rather than corresponding to the sea level - induced hiatuses of Vail et al. (1977). Because of its location near to the Iceland-Faeroe Ridge and the Faeroe-Shetland Channel, the area has certainly been influenced by strong changes of the water circulation regime. Indeed, for two of the hiatuses a possible cause in a bottom current change is indicated in the literature. According to Miller & Tucholke (1983) the hiatus corresponding to reflector R4

133

has its origin in the opening of Fram Strait [dated at c. 38Ma by Talwani & Eldholm (1977)], which caused the first introduction of cold Arctic bottom water into the northern Atlantic. Berggren & Schnitker (1983) argue that the final subsidence of the Iceland-Faeroe Ridge beneath sea level in the middle Miocene is responsible for the hiatus corresponding to R2 (which is probably equivalent to BB3). Sediment

lithology

Until now little has been known about the lithology of the deposits around Bill Bailey and Lousy Bank, mainly because scientific drilling has not yet taken place in this area. A tentative correlation with the results reported by Talwani et al. (1976) for DSDP hole 352 (located on the southern rise of the Iceland-Faeroe Ridge) suggests that in the upper unit the lithology may be dominated by sandy mud and mud mixed with coarse fragments; unit 2, however, and by analogy units 3 and 4 (having a seismic facies analogous to that of unit 2), are not thought to correspond to the nannofossil ooze and chalk with cherty horizons of the Oligocene sediments in DSDP hole 352, but are believed to be rather terrigenous in origin. The transparent character and the observed diapirism in unit 1 point towards a homogeneous and fine-grained sediment texture.

Sediment deformations A mere glance at the seismic lines tells us that the sediments have experienced several kinds of deformation. In a zone southward from the banks, depositional units 2 to 4 are affected by a pattern of minor subvertical faults (best observed on line IF005; Fig. 5). Displacement amounts only to c. 10ms. Due to the rather transparent character of the underlying unit, it is not clear whether these faults continue downward into this unit or not, but, because reflector R4 seems not to be broken up in exactly the same places as the overlying reflectors, it is suggested here that the faults are intraformational. Such intraformational faults have not previously been described in the area. Similar stratigraphically bounded faults were described by Henriet et al. (1988) in Ypresian clay deposits of the Southern Bight of the North Sea. This deformational style is inferred to be related to the compaction history of the sediment rather than to have resulted directly from regional tectonic stresses. A temporary state of increased pore pressure, like that occurring in a sealed clay

134

K. VANNESTE E T A L .

body during burial, would produce these structures. Another striking deformational structure has developed above the large basement fault (lines IF006 and IF001/IF002). In the first instance we interpreted this structure as a diapir (diapir B) piercing out of depositional unit 1. An interpretation which is based chiefly upon the pronounced updoming shape of the structure as displayed on our seismic profiles, as well as upon the fact that reflector R4 is absent beneath this structure. However, other features commonly accompanying diapiric structures, such as rim synclines, are apparently absent. Another 'diapir' observed further southwards is only a minor feature. Apart from salt formations, which are not likely to be present in this area, diapirism is also known to occur in overpressurized fine-grained sediments. Bjorklund & Kellogg (1972) report Upper Eocene clay drilled on similar diapiric structures observed on the Voring Plateau, while Uenzelmann (1988) assumes diapirs in the Voring Plateau area to have developed from underconsolidated terrigenous mud or diatom ooze.

Judging from the thickness variations of overlying sediment units, diapir B started its activity- probably as a result of a movement of the underlying basement fault in response to some regional tectonic stress - at some time during deposition of unit 2, i.e. in Oligocene or early Miocene times.

Qualitative evaluation of the subsidence history One of the most remarkable features concerning the Palaeocene lavas is the pronounced relief presently exhibited by the top surface: the bank crests are now found at a water depth of 300500 ms, rising e. 2200 ms above the basin floor. In surrounding areas also the present depth distribution of the lavas is very variable. They are at surface on the Faeroe Islands, at shallow depth and under a thin cover of sediments on the NW Scottish continental margin and on most of the bathymetric highs north of Rockall Trough (Rosemary Bank, the Wyville-Thomson and Ymir Ridges), but at considerable depth in the intervening basins where the basement is deeply buried below Tertiary sediments. The voluminous extrusion of the flood basalts is generally accepted to have occurred predominantly subaerially, as indicated by: (1) the subaerial nature of the lava sequence on the Faeroe Islands (Noe-Nygaard & Rasmussen

1968); (2) the penetration of DSDP hole 117 on the west side of Rockall Bank into pre-Upper Palaeocene subaerial basalt (Laughton et al. 1972); (3) the steep lava scarp forming the southward limit of the thick lavas (Roberts et al. 1983; Wood et al. 1987), interpreted by Wood et al. (1988) to mark the position of a palaeoshoreline where subaerial lava flows were cooled rapidly by the sea water and stopped flowing. Two hypotheses can be considered concerning the origin of the present basement morphology. Roberts et al. (1983) suggested that the banks are the remnants of an original volcanic landform eroded subaerially. On the contrary, Wood et al. (1987, 1988) assume that all the lavas were extruded more or less horizontally, at or near sea level, implying radically different subsidence histories for areas lying only a small distance apart. The subsidence history of the bank area is estimated here qualitatively from sediment distribution and stratal geometry patterns. In particular, the E-W asymmetry of the basin between the banks at the level of the sediments indicates that the present basement topography did not exist as such immediately after outflow of the Palaeocene lavas. Maybe a less pronounced basement topography was present already before the lava top subsided below sea level. Arguments for this are slight onlap of pre-R4 sediments against the flank of Bill Bailey Bank (line IF006) and the flattened appearance of the bank crests, ascribed to subaerial erosion (Roberts et al. 1983). The initial subsidence of the basalt surface is thought to have occurred more or less uniformly. This is indicated by the fact that reflector R4 remains almost parallel to the basalt surface, with the exception of the eastern flank of Lousy Bank where R4 laps out against the bank crest. Since R4 is an outstanding onlap surface for all overlying reflectors, it is suggested that subsidence became non-uniform after generation of this surface, so that the organization into banks and basins became more prominent. Instead of explaining the observed stratal patterns by a shift from spatially uniform to differential subsidence, one could equally argue for a dramatic decrease of the sediment input after erosion of R4 under a steady but differential subsidence regime. This is, however, not supported by the accumulation rates inferred from the age correlations presented here for the unit bounding unconformities. The onlap of R4 against the crest of Lousy Bank implies that differential subsidence started firstly around the eastern edge of Lousy Bank (still in Eocene times) and only later around Bill

SEISMIC STRATIGRAPHY OF BILL BAILEY AND LOUSY BANKS

135

Fig. 7. (A) Simplified line drawing of profile IF001/IF002. 1-4 indicate the positions of four points along the profile for which the schematic total subsidence curves are shown in (B). Palaeobathymetry and erosion cannot be integrated at present. Bailey Bank and in the zone southward of the banks. Lousy Bank, subsiding more slowly, was thus isolated first. This suggestion - though apparently in contradiction with the present bathymetry (Bill Bailey Bank surmounts Lousy Bank by 200-300m) - is supported by the identifica-tion of Lousy Bank as the most important sediment source during deposition of unit 2, as interpreted from the observed geometry of this depositional unit. At the EoceneOligocene boundary Bill Bailey Bank also began differential subsidence while the rest of the basalt surface kept on subsiding at the same

pace. This is illustrated by the idealized total tectonic subsidence curves for four different points along line IF001/IF002 in Fig. 7. Stratal patterns further suggest that between deposition of units 2 and 3 (i.e. in early Miocene times) a shift took place in the direction of the sediment transport, and that Bill Bailey Bank at that time became the main source for the deposited sediments. This complication in the subsidence history could have been caused by the generation of the above described large basement fault, upthrowing Bill Bailey Bank. This fault is thought to have started its activity

136

K. VANNESTE E T A L .

/ UPPER CRUST -LOWERCRUST --_--

__

--

- - -

MANTLE

\

VERTICALLY VARYING EXTENSION SHALLOW DETACHMENT

DETACHMENT

=-~~-

-

WITH

~ 3

PRESENT BANK

D LATERALLY VARYING EXTENSION DURING CRETACEOUS RIFTING PHASE

OF OF

EROSION . EMPLACEMENT PALEOCENE FLOOD EASALTS AT BEGINNING PALEOCENE RIFTING PHASE

Fig. 8. (A) Some models which could account for differential subsidence between Ymir Ridge and its southern basin (from Wood et al. 1987; their preferred model being model 3). (B) Proposed evolutionary model for Bill Bailey Bank and Lousy Bank.

at some time during deposition of unit 2. The present bathymetry is also in favour of this hypothesis. It is not yet clear what mechanism caused these movements, but a phase of widespread compression has been inferred by Boldreel & Andersen (1993) for the NW margin of the Rockall-Faeroe area in the middle or late Miocene. Two aspects of this subsidence history remain to be solved: why was the subsidence differential and why did differential subsidence become important only after the Eocene-Oligocene boundary ? A possible explanation for the differential aspect of the subsidence could be the presence of magma chambers beneath the zones presently standing out in relief. However, apart from the Faeroe Bank Centre - identified from magnetic anomalies by Dobinson (1970), but not con-

firmed by gravity measurements according to Andersen & Boldreel (pers. comm.) - so far no intrusion centres have been reported beneath the banks. Extensional models have been proposed for the differential subsidence of Ymir Ridge and its southern basin by Wood et al. (1987) (Fig. 8A, their preferred model being model 3), but these models all imply large-scale faulting through the entire lava pile, which we did not observe in this area. But we can consider a slight modification of the model of laterally varying extension by assuming the presence of extensional faults in the underlying continental crust, which could be affected by the two extensional episodes preceding emplacement of the lava pile. The present bank zones would then represent more stable fragments, stretched to a lesser extent than the surrounding areas (Fig. 8B). Subsequent thermal disturbances, such as the

SEISMIC STRATIGRAPHY OF BILL BAILEY AND LOUSY BANKS

137

TERTIARY SUBSIDENCE CURVES FOR LAVA TOPS

Depth

(m)

~

E

I

[

M,OCENE

1 2

3

2--

1 2 3

.o

Ymir Ridge Basin south of Ymir Ridge NE Rockall Trough

.'o

;o

~'o

2'o

,'o

0

Time (Ma B. P.)

Fig. 9. Compilation of Tertiary subsidence curves for the lava tops of Ymir Ridge and the basin south of Ymir Ridge (from Wood et al. 1987), and of the NE Rockall Trough (from Wood et aL 1988).

massive extrusion of hot volcanics at the end of the Palaeocene, could have reactivated the subsidence, resulting in little subsidence for the bank areas and:enhanced subsidence in the intervening areas. It should be stressed, however, that this model is a tentative one, and has not yet been constrained by deep geophysical observations or calculations. The event that caused the shift from uniform to differential subsidence at the Eocene-Oligocene boundary is not well documented either. However, it is not an isolated observation. For the basin southward of Ymir Ridge, Wood et al. (1987) inferred a subsidence in two distinct phases as well. Their subsidence curves (Fig. 9) show a rapid increase in the subsidence rate in early Oligocene times. Together with the observation by Roberts (1975) of post-Early Oligocene subsidence events in the HattonRockall Basin and in the Porcupine Seabight, this strongly indicates that a tectonic event of regional importance occurred around this time. The cessation of spreading in the Labrador Sea was such an event that had major complications for the evolution of the northern North Atlantic: it resulted in the development of complementary fan-shaped spreading along two axes in the Greenland-Norwegian Sea, representing a reorientation of the direction of spreading between Greenland and Eurasia (Nunns 1983). This complex Oligocene spreading geometry is held

to be responsible for compression in the Wyville-Thomson area (Boldreel & Andersen 1993), and may also be a likely cause for the observed events.

Conclusions (1) The present banks are the expression of the morphology of the underlying volcanic substratum. On the flattened bank crests basaltic rocks are near surface, only covered by a thin veneer of sediments, while in the intervening basin the volcanic basement reaches a depth of more than 2000 ms and is buried below more than 1000 ms of sediments. The pronounced basement relief appears not to be fault controlled. (2) Intrabasement divergent reflectors beneath the northern and western flanks of Lousy Bank, and beneath the northern flank of Bill Bailey Bank are thought to build the prolongation of the zone of oceanward dipping reflectors identified north of the Faeroe Islands, and mark the northwestern edge of the Rockall-Faeroe microcontinent. The location of the ocean-continent transition near the northwest edges of the banks is also supported by the asymmetrical shapes of the banks as revealed on N-S sections. (3) Four unconformities were identified in the basin between and southward of the banks where sediment accumulation is confined. The lower reflector marks the most prominent boundary,

138

K. VANNESTE ET AL.

and is suggested to correspond to R4, a reflector which is k n o w n t h r o u g h o u t large parts of the N o r t h Atlantic and has been dated at the E o c e n e ~ ) l i g o c e n e boundary. The other unconformities were provisionally labelled BB3-BB5. They are proposed to correlate with hiatuses at the beginning (BB3) and at the end (BB4) of the middle Miocene, and at the end of the Pliocene (BB5). They are t h o u g h t to have their origin for the m a i n part in changes of the b o t t o m - c u r r e n t regime. (4) The sediments are affected by several kinds of deformation, including a pattern of intraforrnational faults with small offset and a diapir-!ike structure. Both deformational styles are probably related to an increase of the pore pressure in fine-grained sediments, but resulting from different causes. The overpressuring seems to be related to the compactional history o f the sediments in the case of the intraformational faults; in the other case it may have been induced by the m o v e m e n t of an underlying basement fault. (5) It is inferred from the stratal geometry that the investigated banks are not the eroded remnants of an original volcanic landscape, but have been formed by differential subsidence. The initial subsidence of the basalt surface is t h o u g h t to have occurred m o r e or less uniformly. The p r o n o u n c e d onlap of overlying reflectors against R4 suggests that subsidence became non-uniform after generation of this surface, at the E o c e n e - O l i g o c e n e boundary. A r o u n d Lousy B a n k differential subsidence started somewhat eariier. The mechanisms responsible for these observations are still poorly understood. Data presented here were collected during cruise No 146/3 of R.V. Poseidon. We would like to thank the captain and crew of R.V. Poseidon for their assistance during the data gathering. The survey was carried out within the project MFG 00664, which was funded by the German Ministry of Technology and Research (BMFT). The Eelgian participation into the research project was funded by the Fund for Joint Basic Research (FKFO). We wish to thank M. De Batist, as well as M. S. Andersen and L. O. Boldreel from the Danish Geological Survey for kindly reviewing the manuscript and for helpful suggestions. One of the authors (KoV.) is Research Assistant at the Belgian National Fund for Scientific Research (N.F.W.O.).

References ANDERSEN, M. S. 1988. Late Cretaceous and early Tertiary extension and volcanism around the Faeroe Islands. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 115-122. BERGGREN, W. m. & SCHNITKER, D. 1983. Cenozoic

marine environments in the North Atlantic and Norwegian-Greenland Sea. In: BOTT, M. H. P., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) Structure and Development of the GreenlandScotland Ridge. Plenum Press, New York, 495548. BJORKLUND, K. R. & KELLOGG, D. 1972. Five new Eocene Radio!arian species from the Norwegian Sea. Mieropaleontology, 18, 386-396. BOLOREEL, L. O. & ANDERSEN, M. S. 1993. Late Paleocene to Miocene compression in the FaeroeRockall area. In: PARKER, J. R. (ed.) The Petroleum Geology of NW Europe: Proceedings of the 4th Conference. The Geological Society, London, 1025-1034. BOTT, M. H. P., BROWITT, C. W. A. & STACEY,A. P. 1971. The deep structure of the Iceland-Faeroes Ridge. Marine Geophysical Researches, 1, 328351. --, SUNDERLAND,J., SMITH, P. J., CASTEN, U. SAXOV, S. 1974. Evidence for continental crust beneath the Faeroe Islands. Nature, 248, 202-204. CASTEN, U. 1973. The crust beneath the Faeroe Islands. Nature, 241, 83-84. DOBINSON, A. 1970. 1. The development of a marine seismic recording system. 2. A magnetic survey of the Faeroe Bank. PhD thesis, University of Durham, UK. EGLOFF, J. & JOHNSON, G. L. 1975. Morphology and structure of the southern Labrador Sea. Canadian Journal of Earth Sciences, 12, 2111-2133. HALO, N. & WAAGSTEIN, R. 1983. Silicic basalts from the Faeroe Islands: evidence of crustal contamination. In: BOTT, M. H. P., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) S t r u c t u r e a n d Development of the Greenland-Scotland Ridge. Plenum Press, New York, 343-349. HENRIET, J.-P, DE BATIST, M., VAN VAERENBERGH, W. & VERSCHUREN, M. 1988. Seismic facies and clay tectonic features of the Ypresian clay in the southern North Sea. Bulletin van de Belgische Vereniging voor Geologic, 97-3[4, 457-472. HINZ, K. 1981. A hypothesis on terrestrial catastrophes. Wedges of very thick oceanward dipping layers beneath passive continental margins - their origin and paleoenvironmental significance. Geologisches Jahrbuch, Reihe E: Geophysik, 22, 1741. JONES, E. J., EWING, M., EWING, J. I. & EITTREIM, S. L. 1970. Influences of Norwegian Sea overflow water on sedimentation in the northern North Atlantic and Labrador Sea. Journal of Geophysical Research, 75, 1655 1680. LAUGHTON, m. S., BERGGREN, W. A., ET AL. 1972. Initial Reports of the Deep Sea Drilling Project. 12, US Government Printing Office, Washington DC. MILLER, K. G. & TUCt-IOLKE,B. E. 1983. Development of Cenozoic abyssal circulation south of the Greenland-Scotland Ridge. In: BOTT, M. H. P., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) Structure and Development of the GreenlandScotland Ridge. Plenum Press, New York, 549589. MONTADERT, L. & ROBERTS, D. G. (eds) 1979. Initial

SEISMIC STRATIGRAPHY OF BILL BAILEY AND LOUSY BANKS

Reports of the Deep Sea Drilling Project, 48, US Government Printing Office, Washington DC. MUTTER, J. C. 1985. Seaward dipping reflectors and the continent-ocean boundary at passive continental margins. In: HUSEBYE, E. S., JOHNSON, G. L. & KRISTOFFERSEN, Y. (eds) Geophysics of the Polar Regions. Tectonophysics, 114, 117-131. NOE-NYGAARD, A. & RASMUSSEN, J. 1968. Petrology of a 3000 metre sequence of basaltic lavas in the Faeroe Islands. Lithos, 1, 268-304. NUNNS, A. G. 1983. Plate tectonic evolution of the Greenland-Scotland Ridge and surrounding regions. In: BOTT, M. H. P., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) Structure and Development of the Greenland-Scotland Ridge. Plenum Press, New York, 11-30. PEARSON, I. & JENKINS,D. G. 1986. Unconformities in the Cenozoic of the North-East Atlantic. In: SUMMERHAYES,C. P. & SHACKLETON,N. J. (eds) North Atlantic Palaeoceanography. Geological Society of London, Special Publication, 21, 79-86. ROBERTS, D. G. 1975. Marine geology of the Rockall Plateau and Trough. Philosophical Transactions of the Royal Society of London, 278A, 447-509. - - , BOTT, M. H. P. & URUSKI C. 1983. Structure and origin of the Wyville-Thomson Ridge. In: BOTT, M. H. P., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) Structure and Development of the Greenland-Scotland Ridge. Plenum Press, New York, 133-158. RUDDIMAN, W. F. 1972. Sediment redestribution on the Reykjanes Ridge: seismic evidence. Geological Society of America Bulletin, 86, 2039-2062. SMYTHE, D. K. 1983. Faeroe-Shet|and Escarpment and continental margin north of the Faeroes. In: BOTT, M. H. P., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) Structure and Development of the Greenland-Scotland Ridge. Plenum Press, New York, 109-119. TALWANI, M. & ELDHOLM, G. 1977. Evolution of the

139

Norwegian-Greenland Sea. Geological Society of America Bulletin, 88, 969-999. - & UDINTSEV,G., ETAL. 1976. Initial Reports of the Deep Sea Drilling Project, 38, US Government Printing Office, Washington DC. UENZELMANN, G. 1988. Sedimente des sfidlichen A'ufleren Vfring Plateaus - eine hochauflb'sende reflexionsseismische Untersuchung. Berichte aus dem Sonderforschungsbereich 313: Sedimentation im Europ~iischen Nordmeer, nr. 12, ChristianAlbrechts Universit~it zu Kiel. VAIL, P. R., MITCHUM, R. M., Jr & THOMPSON, S., III 1977. Seismic stratigraphy and global changes of sea level. Part 4: Global cycles of relative changes of sea level. In. PAYTON, C. E. (ed.) Seismic Stratigraphy - Applications to Hydrocarbon Exploration. The Association of American Petroleum Geologists Memoir, 26, Tulsa pp. 83-97. VANNESTE, K. 1989. Seismische stratigrafie en geodynamische evolutie van de Ijsland-Faeroe en Kolbeinsey Ruggen. Unpublished Lic. thesis, State University of Gent. WAAGSTEIN, R. 1988. Structure, composition and age of the Faeroe basalt plateau. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 225238. WOOD, M. V., HALL, J. & DOODY, J. J. 1988. Distribution of early Tertiary iavas in the NE Rockall Trough. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 283-292. & VAN HOORN, B. 1987. PostMesozoic differential subsidence in the north-east Rockall Trough related to volcanicity and sedimentation. In: BROOKS J. & GLENNIE, K. (eds) Petroleum Geology of North West Europe, 2, Graham & Trotman, London, 677-685.

Effect of Eocene-Miocene Compression Structures on Bottom-Water Currents in the Faeroe-RockaH Area MORTEN

SPARRE ANDERSEN

& LARS OLE BOLDREEL

Geological Survey o f Denmark, Thoravej 8, DK-2400 Copenhagen NV, Denmark

In the narrow basin between the Faeroe Bank and the Bill Bailey Bank, seismic data demonstrate the presence of a Neogene deep-water sedimentary succession. This succession is characterized by several erosional unconformities (Fig. 1). Based purely on the number of unconformities present, it is possible that they correspond to the third order cycles of sea level change (Haq et al. 1988). However, the topography of the surrounding area, the geometry of the sediment units and the internal reflector pattern suggest that deposition and erosion of

this succession is the result of bottom-water flow through the basin. A simple correlation between the observed unconformities and sea level fluctuations may therefore not be possible. Since Miocene times, cold dense water has formed the bottom waters in the Nordic Seas (the Arctic Ocean and the Greenland-Norway Sea). On their way out of the Nordic Seas the bottom waters pass the Greenland-Scotland Ridge. Meincke (1983) estimated that 30% of the dense bottom waters, formed in the Nordic Seas, flows south through the Faeroe channels

Fig. 1. Part of a migrated seismic section crossing from Faeroe Bank to Bill Bailey Bank. A number of unconformities are highlighted. The data was acquired by the Geological Survey of Denmark and Aarhus University using a 11 airgun array and a 150 m streamer. Vertical resolution is 6-10 m.

From Scrutton, R. A., Stoker, M. S., Shimmield, G. B. & Tudhope, A. W. (eds), 1995, The Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region, Geological Society Special Publication No. 90, pp. 141-143

141

142

M.S. ANDERSEN & L. O. BOLDREEL

LEGEND: .~.2.o" Depth contours Values In kin. Norma~ faults ...Jl--- Reverse fault Overflow

Faeroe Bank

Slgmundur Sear,qq

,

,

I.

Fig. 2. The proposed path for overflow of bottom water from the Faeroe Bank Channel to the basin between the Faeroe Bank and Bill Bailey Bank. The overflow path is superimposed on a map of the depth to the lower Tertiary volcanic rocks. This map is a good approximation of the topography of the Miocene seafloor.

(Faeroe-Shetland Channel and Faeroe Bank Channel). These channels form the deepest passage over the Greenland-Scotland Ridge, and the present threshold depth is c. 800 m. The Faeroe channels are particularly important for the outflow of cold water from the Nordic Seas because the oceanic basement on the F a e r o e - I c e l a n d - G r e e n l a n d Ridge, compared to the Greenland-Norway Sea, is rather shallow, typically < 500 m. This forces the deep cold water either to 'climb' over the ridge or to flow through the narrow but deeper Faeroe channels. The present shape of the Faeroe channels results from the complex interaction of different mechanisms (Boldreel & Andersen this volume). Following the onset of seafloor spreading in the Northeast Atlantic during the early Eocene, the Faeroe-Rockall Plateau has subsided. This is presumably due to thermal contraction of the lithosphere, as the spreading ridge moved away from the continental margin. Overall, the greatest subsidence is in the Faeroe-

Shetland Channel and the Rockall Trough. However, because of Palaeocene to Miocene compression, the Wyville-Thomson and Ymir Ridges now form a barrier to flow of deep water from the Faeroe-Shetland Channel to the Rockall Trough (Boldreel & Andersen 1993). Instead, the deep currents are forced into the Faeroe Bank Channel, which is a syncline between Wyville-Thomson Ridge and the Munkegrunnur Ridge (Boldreel & Andersen this volume). Although flow through the Faeroe Bank Channel is dominating in recent times, intermittent flow of cold bottom water over the WyvilleThomson Ridge has been reported (e.g. Meincke 1983), and the water mass flowing over the ridge continues south along the western side of the Rockall Trough. Overflow of cold bottom water over the Wyville-Thomson Ridge is the most likely explanation for the multiple erosional unconformities in the Miocene to Recent succession of the narrow basin between the Faeroe Bank and the

EOCENE-MIOCENE BOTTOM WATERS, FAEROE-ROCKALL AREA Bill Bailey Bank. In the Miocene the bottom topography channelled the bottom water to the west through the syncline between the WyvilleThomson and Ymir Ridges (Fig. 2). South of the Faeroe Bank the flow would pass the western part of the Ymir Ridge, but the Sigmundur Seamount acted as a barrier for flow further south. Until the seamount was buried in sediments, bottom water would instead flow north between the Faeroe Bank and the Bill Bailey Bank (Fig. 2). The many erosional unconformities are then evidence of significant variation in the rate of deep water flow through the narrow basin, probably resulting from variation in the gross flow through the Faeroe channels. During periods of high flow, the bottom water was dammed in the southern part of the FaeroeShetland Channel, thus allowing increased flow of bottom water over the Wyville-Thomson Ridge and through the basin between the Faeroe Bank and the Bill Bailey Bank. During the periods with the highest flow through the basin, currents would erode into the sediments deposited during periods with lower flow. It is likely that the sedimentary succession in the basin between the Faeroe Bank and the Bill Bailey Bank is an enhanced signal of variation in the bottom-water flow through the Faeroe channels. In turn, the flow through the Faeroe channels may reflect the production of cold

i43

dense bottom water in the Nordic Seas. Alternatively, vertical movements of the WyvilleThomson and Ymir Ridges may have caused fluctuations in the flow over the ridges. Although, Miocene tectonic activity in the area has been reported (Boldreel & Andersen 1993) we do not favour this alternative.

References BOLDREEL, L. O. & ANDERSEN, M. S. 1993. Late Paleocene to Miocene compression in the FaeroeRockall area. In: PARKER,J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference, Geological Society, London, 1025-1034. - &- 1995. The relationship between the distribution of Tertiary sediments and tectonic processes and deep water circulation around the Faeroe Islands. This volume. HAQ, B. U., HARDENBOL, J. & VAIL, P. R. 1988. Mesozoic and Cenozoic chronostratigraphy and eustatic cycles. In: WIL6US, C. H. ETAL. (eds) Sealevel Changes an Integrated Approach. Society of Economic Paleontologists and Mineralogists Special Publication, 42, 71-108. MEINCKE,J. 1983. The Modern current regime across the Greenland-Scotland Ridge. In: Boyr, M. H. P., SAXOV,S., TALWANI,M. & THIEDE, J. (eds) Structure and Development of the GreenlandScotland Ridge. Plenum Press, New York, 133158.

The relationship between the distribution of Tertiary sediments, tectonic processes and deep-water circulation around the Faeroe Islands L A R S OLE B O L D R E E L & M O R T E N

SPARRE

ANDERSEN

Geological Survey o f Denmark, Thoravej 8, DK-2400 Copenhagen N V , Denmark

Abstract: Compressional structures, which were initiated in the Eocene, and thermal subsidence played an important role in the distribution of the post-Palaeocene sediments around the Faeroe Islands. Additionally, the deep-water currents passing through the Faeroe-Shetland Channel and the Faeroe Bank Channel contributed to the distribution of the sediments from Oligocene-Miocene to Recent. The relationship is illustrated by the outcrop of four pre-Upper Pliocene sedimentary sequences below a thin cover of Pleistocene sediments and the surface of the upper Palaeocene basalts in the study area, which consists of three basins - the Norwegian Sea, the Faeroe-Shetland Channel and the Faeroe Bank Channel, separated by compressional ridges. The axial part of the Faeroe-Shetland Channel is characterized by a fairly fiat bottom topography caused by differential sedimentation and erosion by deep-water currents flowing against the topography. The flat channel floor narrows southwards to the tip of the Munkagrunnur Ridge. Due to tectonic structure, differential deposition and erosion, progressively older sediment units are exposed to the south. The channel floor stays narrow in the Faeroe Bank Channel. To the west of the Faeroe Channel Knoll, which is located within the central part of the deep Arctic bottom current, up to 2000 m of sediments are present. At the northern end of the Faeroe Bank Channel the basalt crops out as a result of non-deposition and erosion by the deep Arctic bottom current. The Faeroe Islands, found at the northern end of the Faeroe-Rockall Plateau, are located on the Faeroe Platform which is surrounded by the the Norwegian Sea, the Faeroe-Shetland Channel and the Faeroe Bank Channel. The shallow water areas are represented by fairly fiat bottom topography around the Faeroe Islands and at the Wyville-Thomson Ridge Complex and the Faeroe Bank (Fig. 1). The Faeroe-Shetland Channel and the Faeroe Bank Channel are located between these shallow water areas and the western part of the British Isles continental shelf and form a waterway between the Norwegian Sea and the Atlantic Ocean to the south of the Iceland-Faeroe Ridge. In most of the Faeroe-Rockall Plateau area a succession of early Tertiary volcanic rocks are present (Roberts et al. 1984b; Rasmussen & NoeNygfird, 1970). The basaltic succession, related to rifting and the formation of a passive volcanic continental margin, constitutes the substratum of the Eocene to Recent sediments at the same time as it covers the pre-Palaeocene geological succession. Following the formation of the passive volcanic continental margin during the Palaeocene subsidence and compression took place within the area. The deep-water current passing the Faeroe-Shetland Channel and the Faeroe Bank Channel had an important role in transporting the sediments in the Neogene. As a

result of these processes a cover of variable thickness of Eocene to Recent sediments has been deposited above the basalt. In this paper we discuss the distribution of four post-Palaeocene sedimentary units as a result of interaction between thermal subsidence, compressional phases, volcanic activity and the deep water circulation a r o u n d the Faeroe Islands. The study is based on interpretation of a data set consisting of 80 regional multichannel seismic profiles. Along the continental margin, west of the British Isles, Tertiary tectonic activity is mainly evidenced by inversion of Mesozoic basins under the continental slope (Earle et al. 1989). Additional evidence of Tertiary tectonic activity is found in uplift of significant parts of the continental shelf area, which is deduced by the present distribution of seafloor outcrops or subcrops under a thin Quaternary cover. Tertiary sediments on the shelf west of the British Isles are generally restricted. Thus, the distribution of Tertiary sediments along the margin is controlled both by subsidence in the FaeroeShetland Channel to the west and by material supply from uplifted source areas constituted by part of the shelf around the Shetland Isles to the east (Ziegler 1990; Anderton 1993; Knott et al. 1993). West of the British Isles Tertiary sediments

From Scrutton, R. A., Stoker, M. S., Shimmield, G. B. & Tudhope, A. W. (eds), 1995, The Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region, Geological Society Special Publication No. 90, pp. 145-158

145

146

L.O. BOLDREEL & M. S. ANDERSEN

Fig. 1. Bathymetry and the main morphological elements in the study area. FSC, Faeroe ShetlandChannel; FBC, Faeroe Bank Channel. and volcanic rocks are found in great thickness in the Faeroe-Shetland Channel and further west on the Faeroe-Rockall Plateau. The northern part of the Faeroe-Rockall Plateau is characterized by widespread lower Tertiary (Thanetian-Ypresian) volcanic rocks, which are mostly subaerial plateau basalt of tholeiitic character (Rasmussen & Noe-Nyg~rd, 1970; Roberts et al. 1983; Smythe 1983; Wood et al. 1987). The basaltic rocks are associated with the extensive volcanism which occurred in most of the area as a result of rifting and formation of the passive volcanic continental margin in the Northeast-Atlantic Ocean (Skogseid & Eldholm 1989). A number of primary volcanic escarpments are found on the Faeroe-Rockall Plateau either related to the extensive volcanism or to younger volcanic centres (Smythe 1983; Andersen 1988). These primary volcanic escarpments are formed where the basalt enters a watercovered area at the contemporary coastline 9(Smythe 1983). Hence, the primary volcanic escarpments play an important role in the reconstruction of the palaeocoastline and the

depth to the top of the escarpment can be used in the calculation of the Eocene to Recent subsidence. The height of the escarpment may indicate the depth of the contemporaneous basin. To the north and west of the Faeroe Islands the area is characterized by extreme extension accompanied by the formation of new oceanic crust since the Early Eocene. The continentocean transition between the Faeroe-Rockall Plateau and the North Atlantic province is represented by the occurrence of the seaward dipping reflector sequence, a succession of large wedges of basalt, which originate near the rift axes and presumably are mainly subaerially erupted (Smythe 1983; Roberts et al. 1984a; White 1988).

Structural elements of the surface of the basalt The structural elements and the outcrops of the surface of the basalt are seen on Fig. 2. The

DISTRIBUTION OF TERTIARY SEDIMENTS, FAEROE ISLANDS

147

Fig. 2. Structural elements of the surface of the basalt. EFH, East Faeroe High; FB, Faeroe Bank; FCK, Faeroe Channel Knoll; FCKE, Faeroe Channel Knoll Escarpment; FSE, Faeroe-Shetland Escarpment; FR, Fugloy Ridge; MR, Munkagrunnur Ridge; WTR, Wyville-Thomson Ridge; YR, Ymir Ridge; FP, Faeroe Platform. basalt generally outcrops on topographical highs due to structural uplift and erosion. However, the topographic level of the basalt within the outcrop area varies significantly, e.g. the basalt is subaerially exposed due to erosion on the Faeroe Islands whereas it outcrops, due to nondeposition caused by the deep-sea current, at the saddle between the Faeroe Bank and the Faeroe Platform. A number of compressional ridges and reverse faults are located in the area (Boldreel & Andersen 1993; Andersen & Boldreel, this volume). Among the ridges the Fugloy Ridge (Smythe 1983, fig. 5, which crosses the Fugloy Ridge and parallels our Fig. 5) and the Wyville-Thomson Ridge Complex (Boldreel & Andersen 1993) are the most impressive, followed by the Munkagrunnur Ridge. The East Faeroe High located between the Fugloy and Munkagrunnur Ridges is less pronounced (c. 600m elevation) and consists of two segments slightly offset to the right relative to each other (Fig. 8 is located at the intersection of the two segments). Furthermore, a compressional ridge

is located on the Faeroe Platform. In connection with the compressional ridges reverse faults offset the surface of the basalt, the Eocene sequence and the lower part of the Oligocene sequence (e.g. Figs 6 & 11; see also Boldreel & Andersen 1993; Andersen & Boldreel, this volume). The identification of faults can be rather difficult because of noise arising from the rough surface of the basalt, multiple problems arising from the hard seafloor and the fairly open seismic grid; thus, more faults can be present than shown on Fig. 2. The fault patterns identified are found in association with the structural uplifts and are related to the forming of the compressive ridges. The differential uplift takes place in three steps; Eocene (Figs 5, 7-9 & 13), Oligocene (Figs 5-7, 9 & 11) and during the mid Miocene (Figs 6-8 & 11), which is related to phases of seafloor spreading in the North Atlantic (Boldreel & Andersen 1993; Andersen & Boldreel, this volume). Not only do the above-mentioned compressional structures influence and compli-

148

L.O. BOLDREEL & M. S. ANDERSEN

cate the distribution of Eocene to Recent sediments (as shown later), they also significantly influence net subsidence, which is clearly demonstrated by investigation of the FaeroeShetland Escarpment. The Faeroe-Shetland Escarpment is located in the western part of the Faeroe-Shetland Channel (Talwani & Eldholm 1977). At the time of the extrusion of the plateau basalt to the west of the Faeroe-Shetland Escarpment the FaeroeShetland Channel to the east was submerged and thus the escarpment formed as a foreset breccia in response to quench cooling (Smythe 1983). Considering that the escarpment was presumably formed in very short time (we believe < 1Ma), the top break of the escarpment represents a latest Palaeocene coastline, and the present depth to the top break is a fairly good approximation of Eocene to Recent net subsidence. The seismic sections crossing the escarpment (e.g. Fig. 5) show that the escarpment and the basalt in front of it has been subjected to differential uplift. The total height of the escarpment (the height of the escarpment and the uplifted part of the basalt in front of the escarpment) varies between 0.4 and 2.6 s twoway-time (twt) along the escarpment (Fig. 2). Two local maxima are seen along the escarpment, at the locations where the escarpment aligns with the East Faeroe High and the Fugloy Ridge. This indicates that maximum differential uplift has affected the Faeroe-Shetland Escarpment where it approaches the two ridges. The thermal subsidence of the Faeroe-Shetland Channel since the end of the extrusion of the basalts is thus difficult to estimate as differential uplift has acted in the area. The net Eocene to Recent subsidence of the Faeroe-Shetland Channel, however, can be estimated rather accurately from the depth to the top break of the escarpment and is found to vary between 3000m at the southern part and 5200m in the northern part, but with considerable fluctuations as mentioned above. The Faeroe Channel Knoll (Roberts et al. 1983; Fig. 12) is a volcanic centre located in the southern part of the Faeroe Bank Channel. The volcanic centre is believed to be younger than the plateau basalt in this area and the crest of the volcanic centre outcrops because of the combined effect of the high topography and the location in the middle of a Neogene low deposition area. Northwest of the Faeroe Channel Knoll the Faeroe Channel Knoll Escarpment is located, which is c. 0.9 s twt high. This escarpment is related to the volcanism which resulted in the formation of the Faeroe Channel Knoll. At the front of the escarpment

hyaloclastite and foreset breccia indicates that the volcanic products entered into a submerged area (Fig. 12). The Faeroe Channel Knoll Escarpment indicates that during the late-stage volcanic activity part of the Faeroe Bank Channel to the west of the Faeroe Channel Knoll was already submerged. Thus, the Faeroe Channel Knoll Plateau forms a barrier at the entrance to the Faeroe Bank Channel. Based on the Faeroe Channel Knoll Escarpment we estimate that the net subsidence of the Faeroe Bank Channel, after formation of the Faeroe Channel Knoll, is between 1400 (at the outer sides of the escarpment) and 1800m (at the central part of the escarpment). However, the exact age of the Faeroe Channel Knoll is unknown, otherwise it would be possible to calculate a net subsidence rate representative of Eocene subsidence in the southern Faeroe Bank Channel.

Bathymetry and sediment units The base of the Faeroe-Shetland Channel and the Faeroe Bank Channel, where the deep Arctic bottom current (Meincke 1983) flows, is fairly flat in cross-section (Fig. 9) and narrows downstream from the northern part of the FaeroeShetland Channel towards the Munkagrunnur Ridge (Fig. 1). From thereon it widens somewhat, although it stays rather narrow in the Faeroe Bank Channel making a sharp turn entering the Atlantic Ocean at the saddle between the Faeroe Bank and the Faeroe Platform. The base of the channel system rises in the downstream direction, thus causing the deep-water currents to flow against the topography in the Faeroe-Shetland Channel and the Faeroe Bank Channel until the outlet of the Faeroe Bank Channel is reached. North of the Faeroe Platform a fairly flat base of a channel, but not as pronounced as in the FaeroeShetland Channel and the Faeroe Bank Channel, is found, which probably links up to the base of the Faeroe-Shetland Channel. The variation in thickness of the Tertiary sediments, deposited above the basalt surface, and the location of the outcropping basalt, is shown on Fig. 3. Outside the areas occupied by the main water currents the water depth is rather small and the thickness variation of the sediments reflects the topography of the surface of the basalt. The thickness of the Eocene to Recent sediments increases away from the areas where the basalt outcrops and two major depocentres are located in the Faeroe-Shetland Channel and the Faeroe Bank Channel. At the

DISTRIBUTION OF TERTIARY SEDIMENTS, FAEROE ISLANDS

149

Fig. 3. Thickness of the Tertiary sediments above the plateau basalt. The location of the seismicprofiles A-I (Figs 4-13) is indicated. eastern side of the Faeroe Platform the three compressional ridges (the Fugloy, the East Faeroe High and the Munkagrunnur, see Fig. 2) striking to the northeast are revealed by local minima in thickness of the sediments. Furthermore, the increase in sediment thickness is abrupt where the Fugloy Ridge parallels the Faeroe-Shetland Escarpment. In the Faeroe Bank Channel two minima in thickness are found: one at the Faeroe Channel Knoll Plateau and the other at the saddle between the Faeroe Bank and the Faeroe Platform. The Tertiary sediments can be divided into a number of sequences on the basis of multichannel seismic data and in this study we have divided the Eocene to Recent sediments into five seismic sequences separated by unconformities. These unconformities have been correlated to available well data from a limited number of released commercial wells along the western side of the British Isles continental margin. The seismic dating is in accordance with the dating proposed from dredge hauls on the eastern side of the Faeroe Platform (Waagstein & Clausen,

this volume). The areal distribution of the pre-Upper Pliocene sedimentary sequences is illustrated on Fig. 4. The Eocene sequence is the oldest deposited on the surface of the plateau basalt. On the continental slope north of the Faeroe Islands the Eocene sequence onlaps the basalt Seaward Dipping Reflector Sequence but, as the sequence is transgressed by the younger sequences, the Eocene sequence does not outcrop to the north of the Faeroe Islands. Further south in the Faeroe-Shetland Channel and the Faeroe Bank Channel the lower part of the Eocene sequence onlaps the basalt on the primary volcanic escarpments and at places where the surface of the basalt is rough, e.g. consisting of different flows, erosional escarpments (Figs 5-7 & 11-14). The Eocene sequence outcrops in the Faeroe-Shetland Channel (Fig. 4) in association with the Fugloy and Munkargrunnar Ridges, and at the eastern side of the Faeroe Platform where it is truncated due to differential uplift (Figs 6-9). In the southern part of the Faeroe-Shetland Channel the sequence

150

L.O. BOLDREEL & M. S. ANDERSEN

Fig. 4o Distribution of the pre-Upper Pliocene sediments above the surface of the basalt. Superimposed are the structural elements from the surface of the basalt (see legend to Fig. 3). Also imposed are important morphological elements on the seafloor (see legend to Fig. 1).

outcrops as a combined result of differential uplift and deep water erosion (Figs 8 & 9). In the Faeroe Bank Channel the sequence outcrops along the outer part of the basin (Fig. 4). The reflector pattern within the Eocene sequence is rather conformable but near the top of the sequence an internal unconformity is present in most of the study area (Figs 5-10 & 14) indicating later differential uplift. The Oligocene sequence onlaps the surface of the plateau basalt north of the Faeroe Islands, but due to the deposition of the younger sediments in a transgressive manner it does not outcrop here. However, the sequence outcrops further out in the Norwegian Sea in a narrow east-west striking area where the fairly flat base of a Recent channel is located (Fig. 4). In the Faeroe-Shetland Channel the Oligocene sequence outcrop (Fig. 4) along the eastern part of the Faeroe Platform (due to differential uplift and truncation; Figs 6 & 7), in association with the Fugloy Ridge and the East Faeroe High (caused by differential uplift; Figs 5 & 8), at the

Munkargrunnar Ridge and in the southern part of the Faeroe-Shetland Channel (due to uplift and erosion by deep-water currents; Figs 9 & 10). In the southern part of the Faeroe-Shetland Channel there are indications that deep-water currents were active during part of the Oligocene (Fig. 9) as the sequence is characterized by differential deposition. In the Faeroe Bank Channel the sequence onlaps part of the western side of the Faeroe Channel Knoll (Fig. 12). The sequence outcrops in the central part of the Faeroe Bank Channel where the fairly flat base of a recent deep Arctic bottom current is located (Fig. 4). In the Faeroe Bank Channel no indications of deep-Arctic bottom current activity is seen to affect the Oligocene sequence. The internal reflector pattern of the Oligocene sequence is rather conformable but an internal unconfbrmity is seen in most of the area in the upper part of the sequence (Figs 5-14) indicating later differential uplift. A pronounced uplift of the sequence is found at the Faeroe Platform (Fig. 8). It is expected that the sediments arising

DISTRIBUTION OF TERTIARY SEDIMENTS, FAEROE ISLANDS

151

Fig. 5. Section A crosses a part of the Faeroe-Shetland Escarpment where it is aligned with the Fugloy Ridge. The elevation of the escarpment is c. 2.3 s twt. The internal reflector configuration in the basalt and tile reflector terminations of the lower part of the Eocene sequence show that the escarpment is a primary volcanic escarpment which has later been subjected to differential uplift. Northwest of the uplifted Oligocene sequence the Miocene and Pliocene sequences 'build out' in a regressive manner. from the erosion of the sediments and basalt from the Eocene uplift have mixed with the sediments derived from other source areas, primarily Shetland. The Miocene sequence onlaps the plateau basalt but does not outcrop on the northern part of the Faeroe Platform. However, it outcrops in the Norwegian Sea on both sides of the Oligocene rocks (Fig. 4). In the Faeroe-Shetland Channel the sequence outcrops (Fig. 4) along the eastern side of the Faeroe Platform to the east of the Oligocene (Figs 6 & 7) and in conjunction with the East Faeroe High as a result of truncation and differential uplift. Otherwise the sequence outcrops (Fig. 4) in the eastern part of the Faeroe-Shetland Channel both towards the Faeroe Islands and in the central part, where differential deposition has prevailed. The upper part of the Miocene sequence builds out from the Faeroe Platform and the younger sediments are deposited further out in the basin (Figs 6 & 7). In the Faeroe Bank Channel the Miocene and Pliocene sequence are considered jointly as the limited thickness of the two sequences and the

lack of well ties in the basin makes it difficult to separate them. The combined sequence builds out from the flanks of the highs defining the Faeroe Bank Channel and outcrops basinwards of the older outcropping sequences (Figs 11-14). N o r t h of the Faeroe Islands the Pliocene sequence onlaps the plateau basalt and outcrops from the Faeroe Islands and towards the Norwegian Sea. The sequence outcrops (Fig. 4) in an area between the eastern part of the Faeroe Platform and the East Faeroe High and in the northern part of the Faeroe-Shetland Channel where differential deposition takes place. The Pliocene sequence builds out from the Faeroe Platform (Figs 6-8). The internal reflectors in the Miocene and Pliocene sequences are more continuous than in the Eocene and Oligocene sequences. In the b a s i n ' s b r o a d sediment mounds the internal reflectors are draped, indicating pelagic sedimentation (Figs 5 & 8). The upper part of the Miocene sequence onlaps the uplifted Oligocene sequence at the Faeroe Platform (Fig. 8) and in the upper part of the Miocene sequence an internal unconformity can

152

L . O . BOLDREEL & M. S. ANDERSEN

Figs 6 & 7. Sections B and C are located at the middle part of the Faeroe Platform. The surface of the older basalt benches are seen to continue below the younger benches. A reverse fault, which is found along the eastern part of the Faeroe Platform, is seen to offset the surface of the basalt (Fig. 6) and the same fault is located immediately outside the eastern part of the section shown in Fig. 7. The Eocene sequence onlaps the basalt benches and its lower part parallels the surface of the basalt. Three pulses of uplift (Eocene, Oligocene and mid-Miocene) are seen, and the Eocene, Oligocene and the lower part of the Miocene sequence are truncated by erosion, whereas the mid-Miocene and Pliocene sequences show an outbuilding of the sedimentation from the Faeroe Islands. The post-Pliocene deposits show that contour currents have been active and the unit fills out the depressions. The high-resolution seismic profiles were collected by the Geological Survey of Denmark and the Department of Geology, University of Aarhus, Denmark.

DISTRIBUTION OF TERTIARY SEDIMENTS, FAEROE ISLANDS

153

Fig. 8. Section D crosses part of the East Faeroe High at the intersection of two ridge segments. Differential uplift has taken place and a pronounced uplifted Oligocenesequence is seen. The Miocene and Pliocene sequence builds out from the Faeroe Islands and fills out the lows associated with the uplifted Oligocene sequence. be distinguished (Figs 5 & 6) showing later differential uplift. In the areas subjected to uplift, and in the central part of the basins (where the deep Arctic bottom current is located), the late Pliocene to Recent sediments are generally missing or the cover is too thin to be recognized from the seismic profiles (Figs 5, 8-10 & 12-14). Outside these areas the cover has a variable thickness (Figs 5-14). In summary, we deal with three basins in the study area. In the Norwegian Sea the sediments outcrop in a transgressive manner with the youngest pre-Upper Pliocene sequences outcropping closest to the outcropping basalt. In the Faeroe-Shetland Channel two trends are seen: (1) the sequences outcrop regressively in an easterly direction away from the Faeroe Islands, thus the oldest sequences outcrop closest to the outcropping basalt; (2) in the outer part of the area the outcropping sequences become progressively older in a southerly direction. In the Faeroe Bank Channel the sequences outcrop in a regressive manner, although an Oligocene outcrop is found in the central part of the basin.

Evolution of the area around the Faeroe Islands At the end of the late Palaeocene volcanic activity the Faeroe Islands formed a high area

relative to the Faeroe-Shetland Channel and the Norwegian Sea. At the time corresponding to the onset of seafloor spreading the culmination of the high coincided more or less with the spreading axis to the west of the Faeroe Islands. This we conclude on the basis of the inferred extrusive geometry of the Seaward Dipping Reflector Sequence (e.g. Smythe 1983; Palmason 1981). The Faeroe-Shetland Escarpment shows the location of the contemporaneous coastline east of the Faeroe Islands (Fig. 4). Following the rifting, subsidence was initiated in the Faeroe-Shetland and Faeroe Bank Channels and sediments were deposited in these basins. Pronounced differential uplift, initiated during the Eocene (Boldreel & Andersen 1993; Andersen & Boldreel, this volume), was associated with reverse faulting and the formation of anticlines. The compressional deformation also affected the Faeroe-Shetland Escarpment (Fig. 2). In the Faeroe Bank Channel the Faeroe Channel Knoll was formed during volcanic activity which is younger (presumably earliest Eocene) than the plateau basalt found there. The escarpment at the northwestern part of the volcano indicates the location of the contemporaneous coastline. The topography of the study area was further enhanced by the tectonic processes and sediments were transported into the basins from the positive areas both on the Faeroe area and the Shetland platform. Addi-

1~4

1, 13 R13T.i-)RFF.I, ,~ M

R AN13FRRFN

Figs 9 & 1O. Sections E and F cross the western part of the Faeroe-Shetland Channel and part of the fairly flat base of the channel (Section E). The uplifted part of the basalt constitutes the northern part of the Munkagrunnur Ridge. Uplift of the surface of the basalt and the overlying sediments took place during the Eocene and Oligocene. The deep-Arctic bottom current has eroded the Eocene and differential deposition has been active since the Oligocene.

tional pronounced differential uplifts took place during the Oligocene and the mid Miocene, most likely caused by the tectonic activity in the North Atlantic (Boldreel & Andersen 1993; Andersen & Boldreel, this volume). On the northern part of the Faeroe Platform and the continental slope towards the Norwegian Sea the Tertiary sediment transgresses the surface of the basalt. This pattern is mainly due

to thermal subsidence of the Norwegian Sea. Only the Oligocene and younger sequences outcrop in the Norwegian Sea. On the slope, at depths < 1000m, Miocene and a thin belt of Oligocene sediments outcrop (Fig. 4). The upper limit of the Oligocene and Miocene outcrops coincides with the fiat bottom of a recent deep water channel. Since the Miocene the deep-water current has been active and differential deposi-

DISTRIBUTION OF TERTIARY SEDIMENTS, FAEROE ISLANDS

155

Fig. 11. Section G is a cross section of the Faeroe Bank Channel with a narrow fairly flat base shaped by the deepsea current. The uplifted basalt outcrop to the northeast on the Faeroe Platform and to the southwest on the Wyville-Thomson Ridge Complex, and reverse faults, offset the surface of the basalt and affect the Eocene to mid-Miocene sequences. In the middle of the section the surface of the basalt forms a high, the Faeroe Channel Knoll plateau, which is younger than the plateau basalt. The sequences outcrop in a regressive manner and the deep-sea current passing the channel causes differential deposition in the middle part of the channel.

tion has taken place, and it seems that part of the deep Arctic bottom current flows at the base of the channel. In the Faeroe-Shetland Channel the combined effect of differential uplift and thermal subsidence shaped the basin. The sequences outcrop sequentially away from the Faeroe Platform in a regressive manner (Fig. 4). The Eocene, Oligocene and the lower Miocene sequences show erosional truncation and the erosional unconformity forms a base level above which the younger sediments bypass before being deposited in the low between the East Faeroe High and Faeroe Platform. The present deep-Arctic bottom current located in the central part of the Faeroe-Shetland Channel, flows in a channel with a fairly fiat base. The channel narrows to the south at the same time as the current flows against the topography (Fig. 1). Due to a funnel effect in a southerly direction the velocity of the current increases to the south and successively older sequences outcrop (Fig. 4). In the southernmost part of the Faeroe-Shetland Channel the deepArctic bottom current played an important role in deposition since at least the Miocene. Evidence for bottom-current activity in Miocene-Pleistocene times are also seen as numerous internal unconformities in the Pliocene and Miocene sequences to the south of the WyvilleThomson Ridge, showing that a significant part of the deep-Arctic bottom water flowed from the

Faeroe-Shetland Channel across the WyvilleThomson Ridge Complex and out into the N E Atlantic Ocean during the Miocene (Andersen & Boldreel, this volume). The shaping of the Faeroe Bank Channel took place as a combination of compressional tectonics and thermal subsidence. A pronounced deepening of the basin took place during the Eocene and a thick succession of sediments is found to the northwest of the Faeroe Channel Knoll (Figs 12 & 13). The sequences outcrop in a regressive way (Fig. 4). No indications of deepArctic bottom current activity in the Faeroe Bank Channel is seen in the Oligocene sequence. During Miocene-Pliocene times differential deposition took place in the central part of the Faeroe Bank Channel as the result of deep-water current activity, and therefore the Oligocene sequence outcrops here. The combined effect of the uplift of the Wyville-Thomson Ridge Complex (Boldreel & Andersen 1993), during the mid Miocene, and continued thermal subsidence of the Faeroe Bank Channel caused the Faeroe Channel Knoll to subside to a level where the deep-Arctic bottom current was channelled into the Faeroe Bank Channel. At the outlet of the Faeroe Bank Channel, at the saddle between the Faeroe Bank and the Faeroe Platform, the deep-Arctic bottom current makes a sharp turn (Fig. 4). On the saddle the basalt outcrops due to differential deposition. Southwest of the saddle the seafloor drops

156

Figs

L . O . BOLDREEL & M. S. ANDERSEN

12 & 13. Sections H1 and H2 cross the Faeroe Bank Channel. The basalt and the Eocene-Oligocene sediments are uplifted at the outer parts of the Faeroe Bank Channel and the basalt outcrops at the Faeroe Platform and the Faeroe Bank. At the entrance to the Faeroe Bank Channel the Faeroe Channel Knoll is located and the basalt is exposed at the central part of this volcano. The primary volcanic escarpment at the northwestern part of the volcano shows the location of the palaeocoast line at the time of the eruption. The internal reflector configuration of the basalt and the Eocene sequence shows that the volcano is younger than the plateau basalt. The Faeroe Channel Knoll divides the Faeroe Bank Channel into two parts: (1) a deep western part with a large amount of Eocene sediments and (2) a more shallow basin to the east. The Miocene and Pliocene sequences build out from the Faeroe Platform and the Faeroe Bank towards the basin but are absent in the central part, most likely because of non-deposition where the base of the deep-Arctic bottom current is fairly flat. Signs of contour currents are seen at the outer parts of the channel and around the Faeroe Channel Knoll.

DISTRIBUTION OF TERTIARY SEDIMENTS, FAEROE ISLANDS

157

Fig. 14. Section I is located at the outlet of the Faeroe Bank Channel crossing the fairly fiat base shaped by the deep-sea current. The surface of the basalt rises towards the saddle between the Faeroe Bank and the southwestern part of the Faeroe Platform at the same time as the thickness of the sequences diminishes. It seems that the upper part of the Eocene sequence builds out from the Faeroe Bank whereas the remaining sequences build out from the Faeroe Platform. The deep-sea current causes non deposition and transports the sediments out from the Faeroe Bank Channel passing the saddle between the Faeroe Bank and the Faeroe Platform. The highresolution seismic profile was acquired by the Geological Survey of Denmark and the University of Aarhus, Denmark.

a n d p r o n o u n c e d erosion (an erosional cut of 600 m) is f o u n d at the seafloor to the south o f the saddle (Boldreel et al. 1992). We wish to thank Western Geophysical Company for allowing us to use part of their data for illustrations in this paper.

References ANDERSEN, M. S. 1988. Late Cretaceous and early Tertiary extension and subsidence around the Faeroe Islands. In: PARSON, L. M. & MORTON, A. C. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic, Geological Society, London, Special Publication, 39, 115-122 • BOLDREEL, L. O. 1995. Effect of EoceneMiocene compression structures on bottom water currents in the Faeroe-Rockall area. This volume. - & - 1995. Tertiary compression structures in the Faeroe-Rockall Area. This volume. ANDERTON, R. 1993. Sedimentation and basin evolution in the Paleogene of the northern North Sea and the Faeroe-Shetland basins. In: PARKER,J. R. (ed.) Petroleum Geology of N W Europe: Proceed-

-

ings of the 4th Conference, Geological Society, London, 31-32. BOLDREEL, L. O. & ANDERSEN, M. S. 1993. Late Paleocene to Miocene compression in the FaeroeRockall Area. In: PARKER, J. R. (ed.) Petroleum Geology of N W Europe: Proceedings of the 4th Conference, Geological Society, London, 10251034. --, KIeRBOE, L. & Hov, T. 1992. New high-resolution marine seismic data in the Faeroe-RockaU area. Applied Geophysics, 29, 77. EARLE, M. M., JANKOWSKI,E. J. & VANN, I. R. 1989. Structural and stratigraphic evolution of the Faeroe-Shetland Channel and Northern Rockall Trough. American Association of Petroleum Geologists, Memoir, 46, 461-469. KNOTT, D. S., BURCHELL, M. T., JOLLEY, E. J. & FRASER, A. J. 1993. Mesozoic to Cenozoic plate reconstructions of the North Atlantic and the tectonostratigraphic history of the UKCS Western Margin. In: PARKER, J. R. (ed.) Petroleum

Geology of N W Europe: Proceedings of the 4th Conference, Geological Society, London, 953974. MEINCKE, J. 1983. The modern current regime across the Greenland-Scotland Ridge. In: Borr, M. H.

158

L.O. BOLDREEL & M. S. ANDERSEN

P., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) Structure and Development of the GreenlandScotland Ridge. NATO Conference Series, Series IV, 8, 637-650. PALMASON, G. 1981. A continuum model of crustal generation in Iceland; kinematic aspects. Journal of Geophysics, 47, 7-18. RASMUSSEN,J . & NOE-NYGAARD,A. 1970. Geology of the Faeroe Islands. Danmarks Geologiske Undersogelse, 1, serie 25. ROBERTS, D. G., BOTT, M. H. P. & URUSKI, C. 1983. Structure and origin of the Wyville-Thomson Ridge. In: BOTT, M. H. P., SAxov, S., TALWANI, i . & THIEDE, J. (eds) Structure and Development of the Greenland-Scotland Ridge, NATO Conference Series, Series IV, Marine Sciences, 133158. , MORTON, A. C. & BACKMAN,J. 1984a. Late Paleocene-Eocene volcanic events in the northern Atlantic Ocean. In: ROBERTS, D. G., SCHNrrKER, D. ET AL. Deep Sea Drilling Program, Initial Repots, 81, 913-923. & SCH~TKER, D. ET AL. 1984b. Deep Sea Drilling Program, Initial Reports, 81. SKOGSEID,J. & ELDHOLM,O. 1989. Voting Continental Margin: Seismic interpretation, stratigraphy and vertical movements. In: ELDHOLM, O., THIEDE, J., TAYLOR, E. ET ~Z. (eds) Proceedings ODP, Scientific results, 104, 993-1030. SM~CrHE, D. K. 1983. Faeroe--Shetland Escarpment

and continental margin north of the Faeroes. In: Boar, M. H. P., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) Structure and Development of the Greenland-Scotland Ridge NATO Conference Series, Series IV, Marine Science, 109-119 TALWANI, M. & ELDHOLM, O. 1977. Evolution of the Norwegian-Greenland Sea. Geological Society of America, 88, 969-999. WAA~STErN, R. 1988. Structure, composition & age of the Faeroe basalt plateau. In: PARSON, L. M. & MORTON, A. C. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 3-13. - & CLAUSEN, C. H. 1995. Petrography and biostratigraphy of Palaeogene volcaniclastic sediments dredged from the Faeroes shelf. This volume. WHITE,R. S. 1988. A hot-spot model for early Tertiary volcanism in the N Atlantic. In: PARSON,L. i . & MORTON, A. C. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 3-13. WOOD, M. W., HALL, J. & VAN HOORN, B. 1987. PostMesozoic differential subsidence in the north-east Rockall Trough related to volcanicity and sedimentation. In: BROOKS,J. & GLENNIE, K. W. (eds), Petroleum Geology of North West Europe, 677-687. ZIEGLER, P. A. 1990. Geological Atlas of Western and Central Europe. Geological Society, London.

The influence of glacigenic sedimentation on slope-apron development on the continental margin off Northwest Britain M. S. S T O K E R

British Geological Survey, Murchison House, West Mains Road, Edinburgh E H 9 3 L A , U K

Abstract: The Hebrides and West Shetland shelves were extensively glaciated on several occasions during the mid- to late Pleistocene, with grounded ice locally reaching the shelf edge and depositing vast amounts of sediment directly on to the adjacent slopes. On the Hebrides Slope, the discrete, slope-front depocentre of the Sula Sgeir Fan has accumulated a thick (200 m) succession of interbedded debris-flow diamictons, turbidite sands and muds. This contrasts with a 'between fan' area to the south, where a thinner (< 100m) sediment drape of glaciomarine ice-rafted and suspended sediment, deposited partly under the influence of alongslope currents, is preserved. On the West Shetland Slope, the slope-apron may consist of several coalescing wedges of sediment which form a laterally more extensive, albeit thinner (up to 100m) succession of mass flow deposits. On both slopes the glacigenic deposits can be separated into discrete mass-flow packages bounded by thin, occasionally slope-wide, prograding clinoforms. Accumulation of these packages was episodic and related to specific rapid phases of downslope sedimentation, concomitant with maximum glaciation and glacio-eustatic lowstand. The thin clinoform units reflect intervals of reduced sediment supply to the slope. The 'between fan' area of the Hebrides Slope has been a region of relatively persistent sediment bypass. Analagous deposits preserved on other mid to high-latitude continental margins, such as the East Greenland, Barents Sea and East Canadian margins, highlight the regional importance of glacigenic processes to late Cenozoic slope-apron development in the North Atlantic region. In particular, they demonstrate the influence that climate can have on the stratigraphical and sedimentological record.

Slope-aprons form the transition zone between the shelf and basin-floor, flanking both small shelf basins as well as deep-ocean basins, where they include the continental rise (Stow 1985). They are important transient storage regions for the transfer of terrigenous material shed from the continents to the deep sea, and probably account for the greatest volume of sediment preserved on continental margins, yet they are among the least understood of marine sedimentary environments. In particular, mechanisms of sediment transfer across the shelfbreak, slope sedimentation processes and depositional architecture, remain to be investigated. Submarine slope evolution in front of continental ice sheets is additionally complicated by ice-margin processes. Glacio-eustatic lowering of sea level may result in ice sheets extending out to the shelf edge and depositing vast amounts of sediment directly on to the slope. The high sedimentation rates may result in slope-aprons ranging in style from constructional, with slope progradation, to destructional, with slope failure. The impact of glaciation on slope-apron development is particularly evident in the North Atlantic region, where the repeated growth and

decay of late Cenozoic ice-sheets has directly affected the sedimentary evolution of the mid- to high-latitude slopes. In an attempt to demonstrate the importance of glacigenic processes to slope-apron development this paper presents a summary of the nature and style of glacially-influenced slope sedimentation-on the continental margin off Northwest Britain. This is an area of complex bathymetry which flanks the oceanic Iceland and Norwegian Basins (Fig. 1). Morphologically, it can be divided into an inner margin comprising the Hebrides and West Shetland shelves, an outer margin including the Rockall Plateau, Faeroe Shelf and intervening banks, which together form relatively shallow platform areas separated by the deeper water Rockall Trough and Faeroe-Shetland Channel. The record of ice-rafting from the Rockall Plateau and Hebrides Slope indicates that the continental margin has been accumulating glacially-derived sediment since the late Pliocene, c. 2.48 Ma (Shackleton et al. 1984; Stoker et al. 1994). Although the glacial influence remained predominantly distal in character on the outer margin, there was a change in

From Scrutton, R. A., Stoker, M. S., Shimmield, G. B. & Tudhope, A. W. (eds), 1995, The Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region, Geological Society Special Publication No. 90, pp. 159-177

159

160

M.S. STOKER

Fig. 1. Bathymetric setting of continental margin off NW Britain (after Roberts

et al. 1977) (contours in metres).

Box shows location of the study area which is expanded in Figs 2 & 4. sedimentation style on the inner margin during the early mid-Pleistocene as major ice sheets crossed the Hebrides and West Shetland shelves. Seismic sequences mapped across the edge of these shelves suggest that ice-marginal and proglacial processes have both contributed to the development of the adjacent Hebrides and West Shetland slope-aprons. It is the glaciallyinfluenced development of these slope-aprons which forms the focus of this paper. The extensive geophysical and geological database collected by the British Geological Survey (BGS), as part of its reconnaissance mapping programme on the inner margin, provides the basis for this overview of the large-scale geometry and principal components of these slope-aprons. These data include highresolution seismic profiles, comprising 2 x 40 cu. in. airgun, 1 kJ sparker and deep-tow boomer, together with borehole and shallow core material. Comparisons with glaciated margins bordering the Norwegian-Greenland Sea, and off eastern Canada, provide the basis for a preliminary review of glacigenic slope sedimen-

tation throughout the North Atlantic region. The object of the review is to identify some of the variables that may influence sedimentation patterns on mid- to high-latitude continental margins.

Physiography and geological setting There is considerable variation in physiography along the length of the inner margin (Fig. 2). The Hebrides Slope changes in strike northwards from north- and northwest-trending to northeast-trending. This is marked by a distinct bulge on the margin which owes its origin to a buried Tertiary volcanic centre on the outer shelf (Evans et al. 1989). The strike of the West Shetland Slope is more uniformly northeasttrending. The shelf break is located in water depths ranging from 120 to 250 m, although it may be as deep as 500m on the WyvilleThomson Ridge. West of Lewis (Fig. 2), the shelf break occurs at c. 200m, although the maximum change in slope angle occurs at

SLOPE-APRON DEVELOPMENT OFF NW BRITAIN

Fig. 2, Detailed bathymetry of the study area based on Roberts et al. (1977) and well-navigated BGS Atlas Deso 20 echo-sounder profiles (contours in metres). Location of shelf break delineated from BGS regional offshore mapping, WTR, Wyville-Thomson Ridge; FBC, Faeroe Bank Channel; GE, Geikie Escarpment; AD, Anton Dohrn Seamount; HT, Hebrides Terrace Seamount.

161

c. 600m, at the Geikie Escarpment. This is probably an Oligocene erosion surface, up to 230m high and traced for 215km (Evans et al. 1989). On the Hebrides Slope average slope angles range from < 1 to 4 ~ occasionally reaching up to 7~ on the upper slope, although the Geikie Escarpment displays extreme slope angles of up to 26~ (Strachan & Evans, 1991). Slope angles on the West Shetland Slope are typically < 1~ on the upper slope, increasing to 1.5-2 ~ in the mid-slope area, before decreasing towards the basin floor (Stoker et al. 1993). On both slopes, down to c. 450 m water depth, the sea bed has been intensely scoured by icebergs. Below this depth the sea bed is relatively smooth and featureless due to a combination of contouritic and mass-flow processes (Richards et al. 1987; Stoker 1990; Stoker et al. 1991; Damuth & Olsen 1993). Although the slopes are locally cut by gullies and channels (Kenyon 1987), the lack of submarine canyons is striking. This contrasts markedly with the intensely dissected continental margin west of Ireland and southwest of Britain (Roberts et al. 1977). The underlying structure of the inner margin consists of a Neogene sediment wedge which unconformably overlies tilted, eroded and locally deformed Palaeogene strata (Fig. 3). On the outer shelf and slope the Neogene wedge can be divided into a laterally impersistent and eroded Miocene section (locally including upper Oligocene) unconformably overlain by a PlioPleistocene slope-apron. However, in the deepwater basins of the Rockall Trough and the Faeroe-Shetland Channel, the Plio-Pleistocene appears to be conformable on the Miocene. Where the Miocene is absent, Plio-Pleistocene strata rest directly on the Palaeogene. In contrast, on the Geikie Escarpment lower Oligocene rocks locally crop out at the sea bed (Jones et al. 1986). The profiles in Fig. 3 indicate that the character of the Plio-Pleistocene slope-apron varies greatly in geometry and thickness along the strike of the inner margin. They essentially reflect three main depositional settings on the slope, the extent of which is aptly demonstrated by the isopachyte map (Fig. 4). These depositional settings are represented by: (1) the discrete Barra and Sula Sgeir slope-front fans; (2) the 'between-fan' area, including the Geikie Escarpment; and (3) the more laterally persistent, outer shelf/upper slope wedge west of Shetland. The latter may comprise several coalescing depocentres, less well defined than in the Hebridean region, although two significant accumulations, here informally termed the Rona and Foula

162

M.S. STOKER

Fig. 3. Geological framework of inner continental margin. Inset map shows location of profiles. wedges, have been tentatively identified (R. Holmes, pers. comm.) The total Plio-Pleistocene sediment thickness on the slopes exceeds 600m on the fans, although more typically the slope-apron is between 200-400m thick. In the 'between-fan' area, a much-reduced sediment accumulation, < 200m thick, is preserved. Significantly, the main depocentres are located in front of reentrants 'at the shelf edge which connect with glacially-eroded, transverse troughs and basins on the adjacent shelves (Fig. 4) (Holmes & Stoker 1990; Stoker et al. 1993). These reentrants may have been repeatedly utilized to channel glacial ice across the shelf. The 'between-fan' area has no similar across-shelf links. Whereas the shelfward thinning of the Plio-Pleistocene is partly a function of erosion, the basinward thinning into the Rockall Trough and Faeroe-Shetland Channel to locally < 100 m

is a primary depositional characteristic. Thus, the slope-apron has been a major depocentre throughout Plio-Pleistocene times.

Temporal framework On the Hebrides Slope the Plio-Pleistocene can be divided into two main depositional sequences: (1) the essentially pre-glacial, slope-front to basin-plain sediments of the Lower Macleod sequence; and (2) the slope-front, glacigenic deposits of the Upper MacLeod sequence (Fig. 5) (Stoker et al. 1993). The latter correlates with a complex shelf glacigenic succession which rests in erosive unconformity on the pre-glacial sequence. BGS borehole 88/7,7A penetrated the entire Plio-Pleistocene succession preserved in the 'between-fan' area on the upper Hebrides Slope, landward of the Geikie Escarpment (see Fig. 9). Magneto-stratigraphical and biostrati-

SLOPE-APRON DEVELOPMENT OFF NW BRITAIN

Fig. 4. Plio-Pleistocene isopachyte map highlighting mare slope-front depocentres, and their relation to sediment transport pathways (after Hall 1991 and Stoker et al. 1993) on the adjacent shelves. Map also shows location of seismic reflection profiles and interpreted data illustrated in Figs 7 & 9-11. RT, RockaU Trough; WTR, Wyville-Thomson Ridge; FSC, Faeroe-Shetland Channel.

163

graphical data (Stoker et al. 1994) proved the Lower MacLeod sequence (67.82-89.25 m below sea bed) to comprise Pliocene to early midPleistocene shallow-marine sands, overlain by mid- to late Pleistocene glaciomarine muds of the Upper MacLeod sequence (0.0-67.82m below sea bed) including a 0.5 m thick veneer of early Holocene sands (Fig. 6). Sporadic dropstones in the older sequence reflect a distal glacial influence since the late Pliocene. However, a significant increase in sediment input to the upper slope appears to coincide with the onset of direct glaciation of the Hebrides Shelf. Biostratigraphical data from borehole 88/ 7,7A suggest that this occurred at c. 0.44Ma, with two main phases of sedimentation correlating with the Anglian and Devensian glacial stages, although the age of the deposits between 22.0 and 34.2m depth in the borehole remains uncertain (Stoker et al. 1994). On the adjacent Hebrides Shelf evidence of formerly extensive glaciations is preserved in the form of subglacial meltwater channels and large morainal ridges preserved at the shelf edge (Fig. 6) (Stoker, 1990; Stoker & Holmes, 1991; Stoker et al. 1993). Regional seismostratigraphical relations indicate that these pre-date the last Scottish ice sheet (late Devensian), the extent of which appears to have been more restricted and less influential on slope-apron development in this area. This is manifest by the preservation of late Devensian distal glaciomarine deposits which form part of a slope-front veneer, the MacAulay and Gwaelo sequences, draping t h e Upper MacLeod sequence (Figs 5, 7 and 8) (Stoker et al. 1989, 1993). The coincidence between widespread shelf glaciation and slope progradation was demonstrated by Stoker (1990) on the northern part of the Hebrides Shelf, where the glacigenic slope deposits of the Sula Sgeir Fan appear to be intimately related to a submarine end-moraine at the shelf edge (Fig. 7). Reflections within the Upper MacLeod sequence are partly truncated by, and partly interbedded with, the mounded shelf deposits. The regional stratigraphical evidence suggests that the glacigenic fan deposits may be a product of at least two glaciations (Stoker et al. 1993). On the West Shetland Slope the Plio-Pleistocene succession largely comprises the Morrison sequence which has been tentatively divided into two units: (1) the pre-glacial unit 1, together with the underlying Sinclair sequence; and (2) the glacigenic unit 2 (Fig. 5) (Stoker et al. 1993). This subdiyision is inferred to be broadly equivalent to that on the Hebrides Slope, although the boundary separating pre-glacial

164

M . S . STOKER

Fig. 5. Cross-sections showing the generalized Plio-Pleistocene stratigraphic framework of the shelves and slopes west of the Hebrides and Shetland (after Stoker et al. 1993). Inset map shows location of sections.

SLOPE-APRON DEVELOPMENT OFF NW BRITAIN

165

Fig. 6. Temporal framework of Quaternary deposits on the Hebrides and West Shetland shelves referenced to BGS borehole 88/7,7A, on the upper Hebrides Slope (based on Stoker & Holmes 1991 and Stoker et al., in press). and glacial sections remains poorly defined on the West Shetland Slope. In the Faeroe-Shetland Channel the undifferentiated Plio-Pleistocene basin-floor deposits comprise the FaeroeShetland Channel sequence. Relict morainal ridges are preserved on the outer shelf (Fig. 6), but regional seismostratigraphical relations indicate that they are younger than those preserved at the edge of the northern Hebrides Shelf, probably correlating with the more restricted late Devensian ridges (Stoker & Holmes 1991; Stoker et al. 1993). Thus, the late Devensian component of the West Shetland slope-apron may be more substantial than on the Hebrides Slope. However, the older glacigenic sequences on the northern Hebrides

Shelf can be traced at depth on to the West Shetland Shelf and into the slope-apron, suggesting a multi-glacial component to the slopefront deposits.

The glacigenic slope-apron The seismic and lithological characteristics of the glacigenic slope deposits are summarizd in this section by reference to four detailed study areas: (1) the Sula Sgeir Fan and (2) the 'between-fan' area on the Hebrides Slope, and the upper- to mid-slope and lower slope/basin floor areas west of Shetland, in the vicinity of (3) Foula and (4) Rona wedges (Fig. 4), respectively.

166

M . S . STOKER 6 O

"8

"8 6

8

~o .~ ~oQ ,...,

O

o

~

~

o

,'~

~

.

0

2~ .,,.~ . ,...~ ~,~

SLOPE-APRON DEVELOPMENT OFF NW BRITAIN

167 i

~

~

~.~ ~ . ~

~ o

t:

o

~ n : ~ ~,-~

~,.~ i=l ,,~ o

~,

~ " ~

o

~ 9

0

~

-~ ~.,~ ~ "~ 0 .~

,'-~ 0 ~~ ~,.~.~ 1~o

~

0 ~. 0

~ ~, ~-~ ~

+.~ *'A~ 0

~.9

~

-~ 8 ~ - ~

~~ .~ .~

~o~O~ 9

0

I=l I=l

168

M.S. STOKER

Hebrides Slope Sula Sgeir Fan. The glacigenic component (Upper MacLeod sequence) is up to 200 m thick and is characterized on airgun profiles by a prograding, oblique-parallel reflection configuration of limited downslope and alongslope extent (Figs 7 & 8a) (Stoker 1990). The bulk of the succession consists of three distinct packages, 60-70 m thick, of hummocky, structureless-to-chaotic seismic facies. These are separated by two, thin, slope-wide, prograding clinoforms which downlap on to the basin-plain deposits of the Rockall Trough. Each package consists of an amalgamated sequence of lensoid bodies (Fig. 8b), which range from 3 to 20 m thick and up to 12 km in downslope length and 7.5 km alongslope. They appear to have an overall lobate form, and reflectors separating lenses are of low amplitude and discontinuous. The tops and bases are irregular but it is uncertain to what extent these reflect erosional or depositional contacts. At the distal end of the fan, discrete lenses overlying Miocene to early Pleistocene deep-water sediment waves may have eroded some of the wave crests, although most of the crests appear to have remained intact (Fig. 8c). The clinoforms separating the hummocky, structureless to chaotic seismic packages are of high amplitude and are conformable with substrate irregularities. This has resulted in a draped, planar to slightly irregular reflection character, the form of which is strongly controlled by the surface morphology of the underlying lensoid bodies. The surface of the uppermost package is notably irregular; this is expressed in the morphology of the present slope veneer (MacAulay sequence) and sea bed (Figs 7 & 8). Sediment cores from the upper slope tested the upper structureless-to-chaotic package, and recovered up to 4.75m of diamicton with occasional, interbedded, sands and muds (Stoker 1990). The diamicton is predominantly nonbioturbated, matrix-supported and massive, although thin sandy laminae may impart a crude stratification which is locally deformed. The sands vary from 0.3-0.7m thick and are generally massive, although the tops and bases are graded with inverse grading at the base and normal grading at the top. Sporadic, outsized gravel clasts occur within and at the tops of these beds. The muds form beds 0.1-0.4 m thick which are bioturbated and contain scattered matrix-supported clasts. Shell fragments occur disseminated throughout the deposits. Although palaeontological data are sparse, a cold to arctic

depositional environment is envisaged (Stoker 1990). The morphology, acoustic texture, external dimensions and thickness of the hummocky, structureless-to-chaotic seismic facies, together with the lithological data, led Stoker (1990) to interpret the Sula Sgeir Fan glacigenic succession as comprising large-scale mass-transport deposits; predominantly debris-flow diamictons with subordinate turbidite sands and muds. Such characteristics have previously been extensively documented for debris flows and associated mass-wasting deposits (Hampton 1972; Embley 1976, 1980; Jacobi 1976; Embley & Jacobi 1977; Nardin et al. 1979; Damuth 1980; Damuth & Embley 1981; Almagor & Wiseman 1982; Lowe 1982; Chough 1984; Simm & Kidd 1984). Accumulation of these deposits was episodic and related to specific rapid phases of downslope sedimentation and/or resedimentation, most probably concomitant with icemarginal deposition on the Hebrides Slope. The palaeontological data are consistent with this interpretation. The nature of the clinoforms remains less clear. Their lateral continuity and conformable character are indicative of lower-energy sedimentation, and marine and glaciomarine muds are inferred. Hemipelagites, contourites and dropstone muds have been sampled from the present slope veneer (MacAulay sequence) (Stoker et al. 1989), which may be a thicker accumulation of such deposits. These strata probably represent phases of reduced sediment input to the slope. "Between-fan' area. On seismic profiles the glacigenic deposits of the Upper MacLeod sequence display an acoustically layered, laterally continuous, reflection character (Fig. 9). The amplitude of the reflections is variable, with a weakly layered lower section overlain by a strongly layered upper section, although there is some loss of reflection towards the sea bed. This sequence forms a sheet-drape on the underlying pre-glacial strata (Lower MacLeod sequence), although localized erosional furrows are observed (Fig. 9). Similar, albeit more extensive erosion occurs within the glacigenic sequence at about the level of the change in intensity of the acoustic layering. This also broadly correlates with the contact between Anglian and Devensian deposits; the erosion surface representing a hiatus of up to 0.15 Ma (Fig. 6) (Stoker et al. 1994). Despite the hiatus and variation in amplitude of the acoustic signature, there is no discernible change in lithology, which is characterized by

SLOPE-APRON DEVELOPMENT OFF NW BRITAIN

169

Fig. 9. Sparker profile 84/06-17 and interpreted line drawing across the 'between-fan' area of the Hebrides Slope, showing the relations between the pre-glacial and glacial deposits (after Stoker et al. 1994). Boxed area shows detail in area of BGS borehole 88/7,7A. Profile located in Fig. 4. muds and sandy muds as proved by borehole 88/ 7,7A (Leslie 1992, 1993; Stoker et al. 1994). The muds are predominantly massive through bioturbational activity and contain abundant dropstones scattered throughout the sequence. In some intervals, however, the muds are laminated and graded; laminations range from 0.5-2.0 cm thick, whereas normal and reverse grading may occur over a distance of several metres. Occasional intraformational erosion surfaces are preserved in the laminated beds, and concentrations of dropstones along the bases of some laminae may further indicate erosion of fines during net deposition. The draped reflection configuration is indicative of low-energy deposition (Sangree & Widmier 1977). This is consistent with the lithological data interpreted as a distal glaciomarine sediment record, derived predominantly from a combination of ice-rafting and sediment plumes. In contrast to the adjacent fans, there is

no direct evidence of downslope movement of sediment in the 'between-fan' area. Although processes such as turbidity currents and storm reworking cannot be discounted, it is possible that the sediment plumes were principally derived from the ice margin in front of the fans, and deflected alongslope by weak bottom currents (Leslie 1992, 1993; Stoker et al. 1994). The present current regime in the N o r t h Atlantic, along with reconstructions for the last glacial interval, imply that sediment will have been mainly derived from the south (Barra Fan) and transported northwards (Mclntyre & Kipp 1976). Such alongslope currents may have been responsible for the graded units several metres thick. In comparison to the fans, this particular part of the slope appears to have been predominantly an area of sediment bypass. Nevertheless, slopeapron development in this area has clearly intensified through proglacial sedimentation

170

M.S. STOKER

Fig. 10. Interpreted line drawings of deep-tow boomer profiles on the West Shetland Slope. (a) Revised correlation between uppermost debris flow package and BGS borehole 85/1. Inset shows schematic correlation between profiles and the relative proximity of the borehole. (b) Contrasting styles of acoustically layered deposits. See text for details. Study area located in Fig. 4. Permission to utilize commercial lines BPAR-89-8 and -13 courtesy of BP. processes relative to the pre-glacial depositional regime; the 67.82 m thick sediment record for the interval dated between the present day and 0.44Ma contrasts with the 21.43 m thick record for the interval from 0.44-4.2 M a (Stoker et al.

1994). In detail, sediment accumulation rates of 560 and 170 mm ka -1 have been estimated for the Anglian and Devensian sections, respectively, of the glacigenic sequence, in comparison to the much lower 5 - 2 5 m m k a - 1 for the Pliocene to

SLOPE-APRON DEVELOPMENT OFF NW BRITAIN

171

Fig. 11. Deep-tow boomer profile 85/07-11 from the south-west end of the Faeroe-Shetland Channel showing the detailed seismic characteristics of the structureless lensoid bodies (debris flows), their relation to the acoustically well-layered deposits, and the location of BGS core 60-06/42 (from Stoker et al. 1991, reproduced with permission of Elsevier Science Publishers BV). AUL and MOI, the MacAulay and Morrison sequences, respectively. Inset (top left) illustrates schematic seismic interpretation of core site stratigraphy, the stippled units A and C probably representing separate debris flows. Profile is located in Fig. 4.

early mid-Pleistocene sections (Stoker et al. 1994). Clearly, the greatest input of sediment on to the upper slope in this 'between-fan' area can be directly attributed to glaciation of the adjacent Hebrides Shelf. West Shetland Slope Upper to mid-slope area, Foula wedge. Holmes (1991) described a series of composite debris flow packages preserved on the Foula wedge (Fig. 4). The base of these packages appears to be erosive, cutting into and trun-cating acoustically layered deposits; their top is characteristically draped by reflective strata (Fig. 10). The packages have a variable thickness, commonly ranging from 10 to 60 m, and internally comprise numerous smaller lensoid bodies. The internal complexity is partly manifested by the locally irregular sea bed. BGS borehole 85/1 tested the uppermost debris flow package which is sandwiched be-

tween acoustically layered deposits (Fig. 10a). According to Holmes (1991), the borehole penetrated a 3.5 m thick veneer of glaciomarine muds (MacAulay sequence) overlying c. 10.5m of lightly overconsolidated diamicton (Morrison sequence). The diamicton was divided into two units; an upper, debris-flow unit c. 6.5m thick with shelf-derived faunal remains, mud rip-up clasts and occasional fissuring, and a lower, sandier unit correlated with the underlying acoustically layered strata, inferred to include glaciomarine sediments and older debris flows. The contact between the two units was located at c. 10m below sea bed, coinciding with a reversal in shear strength within the diamicton sequence. Unfortunately, Holmes's (1991) seismic data (BGS profile 79/14-35 in Fig. 10a) do not resolve the subdivision of the diamicton, as the borehole is located c. 1.65 km south of the seismic profile, thereby preventing direct correlation. A recent commercial seismic profile, BPAR89-13, col-

172

M.S. STOKER

lected within 7 0 m of the borehole imaged a thicker debris-flow package closer to the borehole site, comprising at least two separate flows. The underlying acoustically-layered deposits appear to lie deeper within the section at this location, below the base of the borehole (Fig. 10a). An alternative interpretation of the diamicton subdivision reported by Holmes (1991) is that two separate flows have been sampled, consistent with the composite nature of the debris-flow package as shown by both profiles in Fig. 10a. The nature of the acoustically-layered strata is partly revealed by commercial data which image the uppermost part of the slope apron. Figure 10b illustrates two styles of acoustically-layered deposition; an onlapping facies which commonly infills the hollows created by the hummocky tops of debris flows, and a more uniform, draped facies. The onlapping facies is suggestive of current-controlled sedimentation, such as that related to turbidity or contour-current activity, whereas the draped facies is more typical of lower-energy hemipelagic settling from suspension (Sangree & Widmier 1977). Lower slope to basin floor, Rona wedge. Despite the downslope thinning of the slope-apron, debris flows are present on the lower slope and the basin floor, particularly in the southwest part of the region where they still form c. 50% of the lower-slope/basin-floor succession. Recently, Stoker et al. (1991) described partly exposed debris-flow deposits from the distal part of the Rona wedge, at the southwest end of the Faeroe-Shetland Channel. A seismic profile parallel to the axis of the channel shows seismic facies and stratigraphical relations indicative of episodic, mass-flow deposition on the lower slope/basin floor amidst a general background of relatively quiescent marine and glaciomarine sedimentation (Fig. 11). The mass-flow deposits (Morrison sequence) are preserved in discrete, mounded packages up to 50 m thick; these are traceable for up to 19 km along the profile. The partly exposed package displays a hummocky top and a planar to slightly irregular base; partial truncation of the underlying acoustically layered deposits implies some erosion. The internal structure of the package is illustrated at its southwest end, where a number of small lenses c. 5 m thick and 1 km in profile length occur in stacked association. Although acoustic penetration is severely attenuated as the package thickens to the northeast, discontinuous reflectors are identified near the sea bed. These relationships confirm that the packages comprise an amalgamated sequence of

numerous smaller debris flows. The debris-flow packages are interbedded and generally draped by acoustically layered deposits, although there is some indication of onlapping pinch-out. In Fig. 11, the upper debris-flow package is gradually buried by the overlying layered sediments (MacAulay sequence) which onlap to the northeast. Exposure of the debris flows may be due to bottom currents which are active in the channel (Akhurst 1991). At the southwest end of the profile, an older package of debris flows is buried and draped by the layered sediments, although there is some evidence of onlap in the surface undulations of the package (Fig. 11). BGS vibrocore 60/06-42 sampled two separate debris flows (units A and C in Fig. 11) from the uppermost part of the package, where they appear to interdigitate with acoustically-layered strata near the sea bed. The debris flows consist of very soft to soft, non-bioturbated, massive diamicton containing a shelf-derived arctic fauna and flora (Stoker et al. 1991). The gravel content of the diamicton can be up to 30%. Two beds of bioturbated, glaciomarine and marine muds and sands at the top of the core (units B and D) correlate with the acoustically-layered strata; unit B may be partly preserved between the two flows (Fig. 11). Upward-coarsening sands in unit D are consistent with bottomcurrent activity in the channel, and the layered sediments, in general, appear to comprise an interbedded sequence of contourites, thin distal turbidites and glaciomarine hemipelagites (Stoker et al. 1989, 1991; Akhurst, 1991). Summary

On both the Hebrides and West Shetland Slopes, the glacigenic deposits can be separated into discrete mass-flow packages bounded by thin, slope-wide, prograding clinoforms. Accumulation of the mass-flow packages was episodic and related to specific rapid phases of downslope sedimentation concomitant with maximum glaciation and glacio-eustatic lowstand. One significant observation is the apparent lack of slope-channel systems associated with the downslope transfer of these sediments. This may partly reflect the rapid style of sedimentation. The thin clinoform units probably reflect intervals of reduced sediment supply to the slope. This may represent proglacial sedimentation during times of ice-sheet retreat from the shelf-edge, possibly in response to interstadial or interglacial rise in sea level (Damuth & Olsen 1993). Alternatively, it may reflect more loca-

SLOPE-APRON DEVELOPMENT OFF NW BRITAIN

173

lized conditions within a glacial phase, such as a highly lobate ice margin or a less extensive shelf glaciation, resulting in areas of the slope being relatively starved of sediment even during a glacio-eustatic lowstand. The latter suggestion is supported by core data from the present slope veneer (MacAulay sequence) of the Sula Sgeir Fan which recovered late Devensian glaeiomarine muds (Stoker et al. 1989). The 'between-fan' area of the Hebrides Slope has been a region of relatively persistent sediment bypass, with alongslope currents possibly having an important influence on the nature of the preserved sequence. Throughout the glacial period, ice-marginal and proglacial sedimentary processes have contributed to a largely constructional phase of shelf-margin development, the shelf edge having prograded locally in excess of 5 kin.

Comparison with other North Atlantic glaciated margins Submarine slope evolution in front of continental ice sheets is a feature of other mid- to highlatitude continental margins in the North Atlantic region. FOr comparison with the continental margin off Northwest Britain, slopes from both the Northeast and Northwest Atlantic (Fig. 12) are summarized below. In contrast to the margin off Northwest Britain, these more typically encompass continental slope and rise settings, with a much greater bathymetric range. Northeast

Fig. 12. Maps showing bathymetric setting of (a) Norwegian-Greenland Sea (after Elverhoi 1992), and (b) east Canadian and adjacent margins (after Piper et al. 1990), and geographic locations referred to in the text (contours in metres). NS, Nova Scotia; LS, Labrador Shelf.

Atlantic

In the area of the Norwegian-Greenland Sea (Fig. 12a) similar, albeit larger, fans are present along the East Greenland continental margin, such as the Scoresby Fan, as well as fan complexes in the Barents Sea-Svalbard area, such as the Bear Island Trough Mouth Fan and the Isfjorden Fan (Elverhoi 1992). These fans represent major sediment build-outs in front of large, glacially-eroded troughs on the adjacent shelves. The fan at the mouth of the Bear Island Trough has accumulated up to 1000 m of glaeigenie sediments beneath the outer shelf and upper slope (Vorren et al. 1989, 1991). Shelf progradation by several kilometres has accompanied the growth of these fans. These fan complexes display the same acoustic character as the deposits northwest of Britain, and have been similarly interpreted as debris lobes laid down through submarine resedimentation processes during glacial intervals. In common with the northwest British area, the

174

M.S. STOKER

lobes are separated by more continuous reflecting horizons interpreted to represent quiescent phases with less slump activity. Although channels or canyons are not reported from the Istfjorden or Bear Island Trough Mouth Fans themselves, they do occur outside of the latter. This supports the idea that rapid deposition and frequent slumping may suppress channel formation on this type of fan (Solheim et al. 1991). Any pre-existing gully systems in such areas may simply be inundated and buried due to the high rate of sediment input. Northwest Atlantic

In the Northwest Atlantic, glacigenic sediments form an important component of the slope succession off eastern Canada, although the slope-front architecture appears more variable. On the Scotian and Labrador Slopes (Fig. 12b) glacigenic debris flows occur on the upper slope passing downslope into acoustically well-stratified proglacial deposits (Josenhans et al. 1987; Piper 1988). Similar transitions at the Northeast Atlantic locations occur more typically in lower slope and rise settings. The Scotian and Labrador Slopes are further characterized by highly dissected upper slopes, which may be connected to subglacially eroded troughs on the adjacent shelves (Piper et al. 1990; Hesse 1992). These multiple source outlets are envisaged to have delivered large amounts of proglacial sediment to the slope in suspension, although much of this sediment may have bypassed the slope via the slope valley-canyon system, to be ultimately deposited in the adjacent deep-water basins (Piper & Sparkes 1987; Hesse 1992). On the dissected parts of the slope, continuous sedimentation appears to have been restricted to levees and inter-valley areas (Piper 1988). Locally, high sedimentation rates may have led to oversteepening and subsequent slope failure, resulting in downslope resedimentation and further accumulation of debris flows. In contrast, the northeast Newfoundland Slope is not dissected by canyon systems and the glacigenic slope succession is entirely composed of debris-flow deposits interbedded with thin, acoustically stratified, hemipelagic sediments (Aksu & Hiscott 1992). The debris flows are most common on the middle and lower slopes, where they display seismic characteristics and slope depositional architecture (stacked, prograding) comparable to the Northeast Atlantic examples, including the slope-apron off Northwest Britain. A single piston core recovered gravelly sandy mud consistent with a massflow origin. According to Aksu & Hiscott

(1992), the slope sediments were derived from an ice-marginal, line source, situated at the edge of the adjacent shelf.

Discussion and conclusions The repeated growth and decay of northern hemisphere ice sheets, over at least the last 3 Ma years, has directly affected the sedimentary evolution of North Atlantic, mid- to highlatitude slopes. However, this brief review has demonstrated marked regional differences in the nature and style of slope sedimentation. Huge volumes of sediment released from the ice margin on to the Scotian and Labrador Slopes (eastern Canada) have, in part, been transferred via the upper-slope valley-canyon systems into the ocean basins. Slope oversteepening has resulted in destructive phases of slope development, with progradation mostly confined to lower slope and rise settings. In contrast, the slopes off Northwest Britain, the Barents Sea (Norway) and Northeast Newfoundland (eastern Canada) preserve large thicknesses of massflow deposits on major, progradational, slopeaprons. These slopes appear to be predominantly constructional along their length. The reasons for these differences in depositional architecture remain unclear. Hesse (1992) has suggested that such differences may be related to factors such as the sediment delivery system, the characteristics of the glacial and subglacial drainage system, distance from the ice margin, relative sea level, and the amount of sediment available to the system. Piper (1988) has particularly emphasized the role of meltwater discharge, which will clearly influence the amount of glacial detritus supplied by the ice margin. The productivity of basal melt water largely reflects the state of the grounded ice, whether warm or cold based (Powell 1984; Eyles et al. 1985). Deposition rates in marine environments are much greater around thawing ice margins than cold ones (Eyles et al. 1985), although it appears that the greatest volumes of sediment are associated with ice sheets which have alternate areas of melting and freezing (Boulton 1974; Clapperton 1975; Mills 1978; Eyles 1979; Shaw 1979). The physical state of the ice sheets is thus likely to have had a major control on the supply of sediment to the slope. Presumably, the nature of the bedrock geology will also partly determine sediment availability. The contrasting physiography of dissected and non-dissected slopes may be related to the number of melt-water outlet points (Hesse 1992). The canyoned Labrador Slope appears to have been fed by multiple point-source outlets

SLOPE-APRON DEVELOPMENT OFF NW BRITAIN (Hesse, 1992), whereas the ice margin supplying detritus to the northeast Newfoundland Slope is described as a single line source (Aksu & Hiscott 1992). The significance of these observations requires some clarification. Off Northwest Britain sediment delivered to the slope-aprons of the Sula Sgeir and Barra Fans was transferred along discrete sediment pathways feeding on to the Hebrides Slope. However, the West Shetland Slope is also a non-dissected slope, yet it was fed from several outlets, resulting in coalescence of depocentres (Fig. 4). It is interesting to note that the canyoned margin to the west of Ireland and southwest of Britain is inherently steeper (Roberts et al. 1977), and flanks basinal areas which have undergone greater tectonic subsidence, than that of the study area off Northwest Britain. Perhaps original slope angle and tectonic setting also need to be considered. Additionally, margins which have experienced repeated deglaciation events, such as those described in this paper, are more likely to accumulate large volumes of sediment in preference to areas such as West Greenland (Hesse et al. 1990) where the ice-cap probably did not vanish during Plio-Quaternary interglacial stages. The resulting starved West Greenland margin contrasts with, for example, the oversupplied Labrador margin producing a markedly asymmetric basin-fill on the flanks of the Labrador Basin (Hesse, 1992). Further investigations are clearly required at both local and regional scales in order to better compare and contrast North Atlantic glaciated margins. The result would clearly improve general models of continental slope development. From the above discussion, the impact of glaciation is significant for two main reasons: (1) the variability in slope sedimentary evolution emphasizes the importance of local and regional geological controls, amidst the broader background of global change; and (2) slope evolution in front of continental ice sheets is expected to be more climatically controlled, the effect of which on stratigraphy and sedimentation is often overlooked in favour of tectonics and eustasy. For example, condensed sequences or intervals of reduced sedimentation are all too often related to highstands of sea level. The data from the Hebrides Slope, particularly the clinoforms on the Sula Sgeir Fan, suggest that such sequences are equally capable of being generated during lowstands; a response to more localized conditions which, on glaciated margins, are climatically influenced. Mid- to high-latitude slopes may preserve the most complete record of late Cenozoic glaciation on North Atlantic margins. Potentially, this

175

record could provide an understanding of the frequency and intensity of long-term climate change which is crucial in order to evaluate and integrate climatic effects with tectonic and eustatic controls on continental margin development. I would like to thank D. Evans and D. A. Ardus for their comments on an early draft of the manuscript, and E. J. Gillespie for draughting the figures. BP are thanked for allowing access to data, acquired by GEOTEAM UK Ltd, from block 205/9-A; however, I accept all responsibility for the interpretation of lines BPAR-89-8 and -13. Elsevier Science Publishers BV are gratefully acknowledged for allowing the reproduction of Fig. 11. This paper is published with permission of the Director, British Geological Survey, NERC.

References AKHURST, M. A. 1991. Aspects of late Quaternary sedimentation in the Faeroe-Shetland Channel, northwest UK Continental Margin. British Geological Survey Technical Report WB/91/2.

AKSU, A. E. & HlSCOTT, R. N. 1992. Shingled Quaternary debris flow lenses on the north-east Newfoundland Slope. Sedimentology, 39, 193206. ALMAGOR, G. & WISEMAN, G. 1982. Submarine slumping and mass movements on the continental slope of Israel. In: SAXOV, S. & NIEUWENHUIS, J. K. (eds) Marine Slides and Other Mass Movements. Plenum, New York, 95-128. BOULTON,G. S. 1974. Processes and patterns of glacial erosion. In: COATES,D. R. (ed.) Glacial Geomorphology, State University of New York, Binghampton, NY, 41-87. CHOUGH, S. K. 1984. Fine-grained turbidites and mass-flow deposits in the Ulleung (Tsushima) Back-arc Basin, East Sea (Sea of Japan). In: STOW, D. A. V. & PIPER, D. J. W. (eds), FineGrained Sediments: Deep-water processes and facies. Geological Society, London, Special Pub-

lication 15, 185-196. CLAPPERTON, C. M. 1975. The debris content of surging glaciers in~Svalbard and Iceland. Journal of Glaciology, 14, 395-406. DAMUTH, J. E. 1980. Use of high frequency (3.5 KHz12KHz) echograms in the study of near-bottom sedimentation processes in the deep sea: A review. Marine Geology, 38, 51-75. & EMBLEY, R. W. 1981. Mass-transport processes on Amazon Cone: western equatorial Atlantic. American Association of Petroleum Geologists Bulletin, 65, 629-643. - & OLSON, H. C. 1993. Preliminary observations of Neogene-Quaternary depositional processes in the Faeroe-Shetland Channel revealed by high-resolution seismic facies analysis. In: PARKER, J. R. (ed.) Proceedings of the 4th

176

M.S. STOKER

Conference on the Petroleum Geology of North West Europe. Geological Society, London, 10351046. ELVERHOI, A. 1992. A presentation of the ESF PONAM Programme: a study of the last 5 million years of the Norwegian-Greenland Sea and its margins. Oceanology International 92, Abstract Volume, Brighton, UK. EMBLEY, R. W. 1976. New evidence for the occurrence of debris flow deposits in the deep sea. Geology, 4, 371-374. 1980. The role of mass transport in the distribution and character of deep ocean sediments with special reference to the North Atlantic. Marine Geology, 38, 23-50. & JACOm, R. D. 1977. Distribution and morphology of large sediment slides and slumps on Atlantic continental margins. Marine Geotechnology, 2, 205-228. EVANS, D., ABRAHAM,D. A. & HITCHEN,K. 1989. The Geikie igneous centre, West of Lewis: its structure and influence on Tertiary geology. Scottish Journal of Geology, 25, 339-352. EYLES, C. H., EYLES, N. & MIALL, A. D. 1985. Models of glaciomarine sedimentation and their application to the interpretation of ancient glacial sequences, Palaeogeography, Palaeoclimatology, Palaeoecology, 51, 15-84. EYLES, N. 1979, Facies of supraglacial sedimentation on Icelandic and Alpine temperate glaciers. Canadian Journal of Earth Sciences, 16, 13411361. HALL, A. M. 1991. Pre-Quaternary landscape evolution in the Scottish Highlands. Transactions of the Royal Society of Edinburgh, 82, 1-26. HAMPTON, M. A. 1972. The role of subaqueous debris flow in generating turbidity currents. Journal of Sedimentary Petrology, 42, 775-793. HESSE, R. 1992. Continental slope sedimentation adjacent to an ice margin. 1. Seismic facies of Labrador Slope. Geo-Marine Letters, 12, 189-199. , RAKOFSKY, A. & ENOUGH, S. K. 1990. The central Labrador Sea: Facies and dispersal patterns of elastic sediment in a small ocean basin. Marine and Petroleum Geology, 7, 13-28. HOLMES, R. 1991. Foula (60~ 04~ Quaternary Geology. British Geological Survey, 1:250000 Offshore Map Series. • STOKER, M. S. t990. Pleistocene glaciomarine facies on the continental shelf and slope, north-west of Scotland. In: 13th International Sedimentological Congress, Abstract Volume, 230. JACOBI, R. D. 1976. Sediment slides on the northwestern continental margin of Africa. Marine Geology, 22, 157-173. JONES, E. J. W., PERRY, R. G. & WILD, J. L. 1986. Geology of the Hebridean margin of the Rockall Trough. Proceedings of the Royal Society of Edinburgh, 88B, 27-51. JOSENHANS,W. H., BARRIE, J. V. & KIELY, L. k. 1987. Mass wasting along the Labrador Shelf margin: submersible operations. Geo-Marine Letters, 7, 199-205. KENYON, N. H. 1987. Mass-wasting features on the -

-

continental slope of Northwest Europe. Marine Geology, 74, 57-77. LESLIE, A. B. 1992. A sedimentological study of Tertiary and Quaternary sediments in borehole 88/7,7A, Hebrides Slope, northern Rockall Trough region. British Geological Survey Techni-

cal Report, WB/92/16. 1993. Shallow Plio-Pleistocene contourites on the Hebrides Slope, northwest U.K. continental margin. Sedimentary Geology, 82, 61-78. LOWE, D. R. 1982. Sediment gravity flows: ii. Depositional models with special reference to the deposits of high-density turbidity currents. Journal of Sedimentary Petrology, 52, 279-297. MClNTYRE, A. & KIPP, N. G. 1976. Glacial North Atlantic 18,000 years ago: a CLIMAP reconstruction. Memoir of the Geological Society of America, 145, 43-75. MILLS, H. H. 1978. Some characteristics of glacial sediments on Mount Rainier, Washington. Journal of Sedimentary Petrology, 48, 1345-1356. NARDIN, T. R., HEIN, F. J., GORSLINE, D. S. & EDWARDS, B. D. 1979. A review of mass movement processes, sediment and acoustic characteristics, and contrasts in slope and base-of-slope systems versus canyon-fan-basin floor systems.

-

-

Society of Economic Paleontologists and Mineralogists, Special Publication 27, 61-73. PIPER, D. J. W. 1988. Glaciomarine sedimentation on the continental slope off eastern Canada. Geoscience Canada, 15, 23-28. & SPARKES, R. 1987. Proglacial sediment instability features on the Scotian Slope at 63~ Marine Geology, 76, 15-31. - - , MUDIE, P. J., FADER, G. B., JOSENHANS, H. W., MACLEAN, B. & VILKS, G. 1990. Quaternary Geology. In: KEEN, M. J. & WILLIAMS, G. L. (eds) Geology of the Continental Margin of Eastern Canada. Geological Survey of Canada, Geology of Canada, 2, 475~07. POWELL, R. D. 1984. Glacimarine processes and inductive lithofacies modelling of ice shelf and tidewater glacier sediments based on Quaternary examples. Marine Geology, 57, 1-52. RICHARDS, P. C., RITCHIE, J. D. & THOMSON, A. R. 1987. Evolution of deep-water climbing dunes in the Rockall Trough - implications for overflow currents across the Wyville-Thomson Ridge in the (?)Late Miocene. Marine Geology, 76, 177183. ROBERTS, D. G., HUNTER, P. M. & LAUGHTON, A. S. 1977. Bathymetry of the northeast Atlantic, sheet -

-

2: continental margin around the British lsles. Institute of Oceanographic Sciences. SANGREE, J. B. & WIDMIER, J. M. 1977. Seismic stratigraphy and global changes of sea level, part 9: Seismic interpretation of elastic depositional facies, ln." PAYTON, C. (ed.) Seismic stratigraphy applications to Hydrocarbon Exploration. American Association of Petroleum Geologists, Memoir, 26, 165-184. SHACKLETON, N. J., er AL. 1984. Oxygen isotope calibration of the onset of ice-rafting and history of glaciation in the North Atlantic region. Nature,

SLOPE-APRON DEVELOPMENT OFF NW BRITAIN 3117, 620-623. SHAW, J. 1979. Genesis of the Sveg fills and Rogen moraines of Central Sweden: a model of basal melt-out. Boreas, 8, 409-426. SIMM, R, W. & KIDD, R. B. 1984. Submarine debris flow deposits detected by long-range side-scan sonar 1000 Kilometres from source. Geo-Marine Letters, 3, 13-16. SOLHEIM, A., ELVERHOI, A., ANDERSEN, E. S. & JAHRE, H. 1991. Marine geological/geophysical cruise on the western Svalbard margin 1990 cruise report. Norsk Polarinstitutt Rapportserie, 69. STOKER, M. S. 1990. Glacially-influenced sedimentation on the Hebridean slope, northwestern United Kingdom continental margin, ln." DOWDESWELL, J. A. & SCOURSE, J. D. (eds) Glacimarine Environments: Processes and Sediments, Geological Society, London, Special Publication, 53, 349362. & HOLMES, R. 1991. Submarine end-moraines as indicators of Pleistocene ice-limits off northwest Britain. Journal of the Geological Society, London, 148, 431-434. , HARLAND, R. & GRAHAM, D. K. 1991. Glacially influenced basin plain sedimentation in the southern Faeroe-Shetland Channel, northwest United Kingdom continental margin. Marine Geology, 100, 185--199. - - , HITCHEN, K. & GRAHAM, C. C. 1993. United Kingdom offshore regional report: the geology of -

-

177

the Hebrides and West Shetland shelves, and adjacent deep-water areas. HMSO for th~ British Geological Survey, London. , HARLAND, R., MORTON, A. C. & GRAHAM, D. K. !989. Late Quaternary stratigraphy of the northern Rockall Trough and Faeroe-Shetland Channel, northeast Atlantic Ocean. Journal of Quaternary Science, 4, 211-222. , LESLIE, A. B., SCOTT, W. D. eT AL. 1994. A record of late Cenozoic stratigraphy, sedimentation and climate change from the Hebrides Slope, NE Atlantic Ocean. Journal of the Geological Society, London, 151, 235-249. STOW, D. A. V. 1985. Deep-sea elastics: where are we and where are we going? In." BRENCHLEY,P. J. & WILLIAMS, B. P. J. (eds) Sedimentology: Recent Developments and Applied Aspects, Geological Society, London, Special Publication, 18, 67-93. STRACHAN, P. & EVANS, D. 1991. A local deep-water sediment failure on the NW slope of the UK. Scottish Journal of Geology, 27, 107-111. VORREN, T. O., LEBESBYE, E., ANDREASSEN, K. & LARSEN, K.-B. 1989. Glacigenic sediments on a passive continental margin as exemplified by the Barents Sea. Marine Geology, 85, 251-272. , RICHARDSEN, G., KNUTSEN, S. M. & HENRIKSEN, E. 1991. Cenozoic erosion and sedimentation in the western Barents Sea. Marine and Petroleum Geology, 8, 317-340.

Petrography and biostratigraphy of Palaeogene volcaniclastic sediments dredged from the F aeroes shelf REGIN WAAGSTEIN 1 & CLAUS HEILMANN-CLAUSEN

2

1 Geological Survey o f Denmark, Thoravej 8, DK-2400 Kobenhavn N V , D e n m a r k 2 Department o f Earth Sciences, University o f Aarhus, DK-8000 Aarhus C, D e n m a r k

Abstract: A large number of dredge hauls made on the outer shelf and slope east of the Faeroe Islands have yielded Lower Tertiary sedimentary rocks (both in situ and glacially transported). The rocks have been examined petrographically and biostratigraphically (dinoflagellates), and they are all considered to be derived from the Faeroe shelf. They can be classified according to sediment type and chronostratigraphy in the following groups: (1) Lower Eocene basaltic tufts (dominantly non-marine); (2) Lower and Middle Eocene tuffaceous limestones and phosphatic sediments; and (3) Lower and possibly Upper, Oligocene fine-grained feldspathic volcanic sandstones. Two dredge hauls of local till near the shelf edge west of the Faeroe Islands gave coarse to fine grained volcanic sandstones of Oligocene age. The basaltic tufts of the eastern shelf bear witness to a major phase of explosive volcanism succeeding the extrusion of the Faeroe lava plateau. Whole-rock chemical analyses suggest that the explosive magmas had a tholeiitic composition rich in iron and titanium. Some of these eruptions are likely to be represented among the tufts of the Balder Formation in the North Sea. After the deposition of the tuffaceous sediments in a marine setting during the Early and Middle Eocene the central parts of the Faeroe basalt plateau were probably uplifted in the Bartonian or Priabonian. As a result of erosion of the basalt plateau and overlying sediments volcanic sandstones (often rich in reworked dinoflagellates) were deposited in a marine environment during the Early Oligocene. A detailed stratigraphic framework has been established for the southeastern part of the Faeroe Basin on the basis of numerous seismic reflection profiles and exploration wells made by the oil industry over the last two decades (Ridd 1981; Earle et al. 1989). In contrast, little geophysical work and no drilling has been carried out as yet in the northwestern part of the basin where vigorous volcanic activity resulted in the formation of the Faeroe basalt plateau in the Late Palaeocene. This paper presents the first general description of the Palaeogene sediments above the basalts forming part of the Faeroe shelf and is based on dredged in situ and glacially-transported rocks. The sedimentary rocks are described petrographically and their biostratigraphy and age are established on the basis of dinoflagellates. These results are used to elucidate and discuss the geological events which took place after the formation of the Faeroe basalt plateau.

Geological setting The Faeroe Island shelf has a triangular shape bounded by the Norwegian Sea to the north, the Faeroe-Shetland Channel to the southeast and the Faeroe Bank Channel and the Iceland Basin to the southwest (Fig. 1). This elevated area is also called the Faeroe Block and is underlain by

thick continental crust (Bott et al. 1974). Most of the shelf is formed by the Lower Tertiary basalt platform of the Faeroe Islands. However, the outer southeastern shelf contains up to c. 2 km of Tertiary sediments overlying the basalts (Nielsen et al. 1979; Ridd 1981; Boldreel & A n d e r s e n , this volume). A n o t h e r smaller sedimentary wedge progrades from the Faeroe Islands into the narrow part of the Faeroe Bank Channel to the southwest (Stride et al. 1967). The basalts dip gently away from the Faeroe Islands beneath the sediments, usually at an angle of 2-10 ~ (Nielsen et al. 1979), The s e d i m e n t a r y shelf areas are characterized morphologically by the presence of banks and marginal and transverse erosional channels typical of former glaciated shelves (Holtedahl 1970). The Tertiary sediments of the eastern shelf form part of the Faeroe Basin which has its depocentre in the Faeroe-Shetland Channel and consists of Palaeogene and Neogene strata separated by a major Oligocene unconformity (Ridd 1981). In most places Quaternary deposits unconformably overlie the Tertiary.

Dredge hauls The present study is based on sedimentary rocks recovered in 21 dredge hauls made around the Faeroe Islands (Fig. 1; Table 1). Individual rock

From Scrutton, R. A., Stoker, M. S., Shimmield, G. B. & Tudhope, A. W. (eds), 1995, The Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region, Geological Society Special Publication No. 90, pp. 179-197

179

180

R. WAAGSTEIN & C. HEILMANN-CLAUSEN

Fig. 1. Map of the southern part of the Faeroe shelf showing the position of dredge hauls with sediments of local origin. fragments are identified by the dredge number followed by a hyphen and a running number. The samples include all sedimentary rock clasts > 5 c m of assumed Faeroe shelf provenance which were recovered during an early study of the distribution of glacial erratics around the Faeroe Islands (Waagstein & Rasmussen 1975; Waagstein 1977). One sample (87-1) was recovered accidentally in the net of a Faeroese trawler. The remaining samples examined are from several dredge hauls which were made in 1987 in an attempt to sample supposed sediment outcrops identified east of the Faeroe Islands by shallow seismics (Nielsen et al. 1979; Fig. 2). Rubble of exposed bedrock has been recovered in at least five, possibly six, dredge hauls (Table 1). Five were made within the marginal channel west of Sandoy Bank along a seismic reflection profile (Fig. 2). The length of the hauls varies from 0.7 to 2.8 km and the sediments were recovered together with abundant glacial erratics mainly consisting of basalt. However, the number of lithologically similar sediment fragments strongly suggest that they originate from

seafloor exposures near the dredge path. Two pieces of sediment from dredge 145 east of the island of Suduroy are likewise considered nearly in situ because they are extremely soft and thus unlikely to have survived any glacial transport. The sediments from the remaining 15 dredges are probably all glacially transported although glacial striae are rarely observed. In these dredges the sediments are only represented by one or a few fragments, and fragments from the same dredge are usually lithologically different. Calcareous fragments often show abundant dissolution pits on surfaces exposed to seafloor weathering, while Pholas borings are observed in a few limestones. Two fragments (65-6 and 87-1) have a regular ovoid shape suggesting that they are concretions derived from a softer sediment. Most of the glacially transported rocks are from the shelf or uppermost slope at water depths <400m; their subrounded shape, the almost complete absence of rock types other than basalt and the above sediments suggest that they originate from lodgement tills deposited by a Faeroese ice sheet. Five dredges are from depths

PALAEOGENE VOLCANICLASTICS, FAEROES SHELF

183

Table 2. (A). Chemical composition of group A basaltic tufts No. SiO2 TiO2 A1203 Fe203 FeO MnO MgO CaO Na20 K20 P205 Loi Total CaCO3 Sc V Cr Co Ni Cu Zn Rb Sr Y Zr Nb Ba La Ce Nd

3-14

3-31

4-9

4-78

4-125

5-20

5-33

18.95 1.36 6.68 5.38 1.25 0.69 1.67 31.12 1.31 0.90 1.50 28.41 99.22 55.5

18.38 0.71 4.72 9.46

23.84 2.15 8.10 2.13 6.11 1.89 3.36 24.78 2.03 0.64 0.43 20.51 95.97 42.3

21.31 2.02 8.11 4.90 1.59 1.03 1.23 30.04 1.57 0.55 1.28 25.51 99.14 50.7

22.58 1.94 8.15 6.45 2.88 0.60 1.31 27.88 1.66 0.55 0.32 25.33 99.65 44.6

21.63 1.13 6.85 4.54 2.12 1.35 2.46 30.13 1.46 0.77 1.29 25.48 99.21 52.5

8.11 0.72 3.39 3.18 1.93 2.23 0.85 42.64 0.41 0.05 0.84 35.30 99.65 77.1

45 189 60 24 23 75 67 17 251 12 91 13 46 4 20 18

1.84 1.65 32.44 1.22 0.94 1.71 26.40 99.47 56.2 13 92 26 24 27 23 61 24 198 74 73 8 60 35 32 23

50 363 75 28 38 119 82 6 149 34 175 20 41 14 43 30

49 256 70 20 27 76 84 9 213 25 142 18 68 12 32 23

Clasts: Mean (mm) Max. (mm)

0.15 0.70

0.1 0.5

0.8 2.5

0.6 2.0

No.

71-9

145-1

145-2

154-7

23.49 2.20 8.53 12.09 1.24 0.94 1.54 23.51 1.63 0.71 1.31 21.80 98.99 39.3

39.71 4.22 14.80 19.63 1.50 0.11 1.60 0.77 1.75 1.28 0.35 14.27 99.99 1.4

40.71 4.14 17.45 18.24 2.57 0.07 2.29 0.45 1.91 1.02 0.23 11.05 100.13 1.4

12.10 0.97 4.34 6.13 20.79 1.62 0.69 19.62 0.76 0.28 2.05 30.01 99.37 29.1

SiO2 TiO2 A1203 Fe203 FeO MnO MgO CaO Na20 K/O P205 Loi Total CaCO3 Sc V Cr Co Ni Cu Zn 9

52 270 64 40 42 134 108

51 353 122 78 60 226 239

48 332 81 74 67 179 157

16 124 27 85 15, 31 54

50 225 63 29 32 86 87 6 176 24 141 14 45 10 36 29

50 255 41 22 20 62 70 16 196 106 120 15 36 34 76 42

0.4 2.5

0.3 1.2

157-3

157-4

20.53 1.90 7.48 0.00 10.11 0.83 2.30 28.52 1.39 0.32 0.47 25.68 99.54 48.0 29 256 60 31 38 111 76

21.34 1.86 7.68 7.84 0.00 0.83 0.98 29.66 1.46 0.73 0.31 21.40 94.08 50.0 23 230 58 19 29 115 48

49 94 16 16 26 51 40 1 150 26 48 6 21 15 22 13

5-34

5-41

17.12 1.58 6.53 4.35 1.36 1.01 1.33 33.99 1.73 0.50 0.69 29.56 99.75 60.6

24.63 1.77 8.53 7.59 1.35 0.87 2.31 26.63 1.80 0.74 0.63 22.73 99.58 43.9

52 223 56 23 27 77 77 7 174 26 116 14 39 10 31 21

51 263 77 32 39 94 76 5 146 25 147 21 35 11 38 26

0.1 1.0

0.8 5.0

0.3 0.8

157-10

157-11

161-1

24.04 1.96 8.05 8.70 2.43 1.17 2.52 24.61 1.61 0.62 0.36 18.83 94.90 42.1

21.28 2.08 8.06 4.00 2.57 1.00 1.23 31.14 1.50 0.61 0.60 25.22 99.29 53.0

47.61 1.91 14,24 12.72 4.54 0.02 4.54 1.77 2.81 3.33 0.26 5.89 99.64 3.4

26 235 67 28 33 111 67

28 268 77 26 31 111 106

44 363 73 62 54 151 130

184

R. WAAGSTEIN & C. HEILMANN-CLAUSEN

Table 2. (continued) No.

71-9

145-1

145-2

154-7

157-3

Rb Sr Y Zr Nb Ba La Ce Nd

12 212 43 156 17 77 19 52 34

15 67 75 373 39 31 14 72 50

14 85 52 361 43 33 21 72 51

4 168 41 91 10 58 19 45 25

7 347 22 127 14 88 14 31 23

157-4

157-10

8 156 27 135 13 73 13 28 22

20 183 27 134 15 98 14 30 21

157-11

161-1

8 265 33 141 15 61 14 36 25

69 102 33 243 28 81 34 91 49

Clasts: Mean (mm) max. (mm)

0.4 2.5

0.7 3.0

2.0

0.1 0.35

1.2 2.5

2.0 6.0

0.3 1.5

1.0 5.0

0.2 1.0

Major elements are reported in weight percent and trace elements in parts per million. The samples are analysed by XRF at the Geological Survey of Greenland (majors) and at the University of Copenhagen (traces). CaCO3 is determined by titration at the Geological Survey of Denmark.

of 500-700m on the northwest slope of the Faeroe-Shetland Channel and only recovered c. 50% basalt and minor volcaniclastics, the remaining rocks being mainly metamorphic rocks and siliciclastic sediments of assumed Scottish provenance. All the rocks in the latter five dredges are probably ice-rafted (Waagstein & Rasmussen 1975) but the few volcaniclastics resemble those recovered from sub-sea exposures and lodgement tills in more shallow water and are, therefore, assumed to originate from the Faeroe shelf.

Petrography and age of sediments All the volcaniclastic sediments from the dredges of glacial erratics and a representative set of sediments from the dredged exposures have been described macroscopically and most of them have also been examined in thin section. A subset of samples has been analysed biostratigraphically using dinoflagellates. The dinoflagellate assemblages have been compared with zones originally established in Denmark (Heilmann-Clausen 1988) and northern Germany (K/Sthe 1990) (the zones are shown in Fig. 4). The zones have been applied to the North Sea by one of us (C. H.-C.) and also seem applicable to the present material. Comparisons were also made with other studies, in particular with information from the Voring Plateau

(Manum et al. 1989). The dinoflagellates of analysed samples are shown in Table 4 and the interpreted ages are shown in Fig. 4. On the basis of their petrography and age we have classified the sediments into three main groups, A, B and C, which are described in detail below. We present geochemical data on some of the sediments and minerals.

A. Post lava plateau basaltic tufts Description. The tufts are yellowish brown, yellowish grey or greenish grey in colour. Most are carbonate cemented, hard (Fig. 3e), while a few are carbonate poor and friable (Fig. 3a). The tufts range in grain size from very coarse to very fine grained, and they are moderately to well sorted, sometimes showing a crude bedding. Clasts > 2 m m in diameter often show some rounding (Fig. 3a), whereas clasts < 0.2mm are generally angular or irregular in shape (Fig. 3e). The large majority of tufts are non-marine (fluvial or paralic) with spores, pollen and less commonly plant fragments. Calcified wood occurs in tufts 4-9 and 4-78, and lignite in 1588. In addition, 6-8 cm long pieces of lignite have been recovered in dredge hauls 157 and 161. Tufts deposited in sea water occur only in dredge 161 and they are mainly fine grained, yellowish grey and carbonate free. They contain dinoflagellate cysts (161-1 and 161-2) and are often bioturbated. Two very fine-grained tufts from

PALAEOGENE VOLCANICLASTICS, FAEROES SHELF

185

Table 3. Average composition of basaltic tufts from the Faeroe shelf recalculated calcite and water-free, with comparisons Basaltic tuff Main type Faeroe shelf No. of analyses

Basaltic tuff Balder Fm. Norw. well 30/2-1

15

Basaltic tuff Balder Fm. UK well 16/7a-2

14

Basalt 35-9 Upper Fm. Faeroes

5

1

Mean

s.d.

Mean

s.d.

Mean

s.d.

Per cent SiO2 TiO2 A1203 Fe203 MnO MgO CaO Na20 K20 P205

45.83 3.94 16.70 19.78 2.24 3.64 1.85 3.12 1.41 1.47

3.05 0.71 1.29 3.72 2.48 1.27 2.21 0.71 0.75 1.22

42.38 4.29 16.84 20.92 0.30 7.72 3.87 2.19 0.50 0.98

2.08 0.40 0.73 3.43 0.03 0.95 2.10 0.44 0.09 0.55

55.71 5.19 20.65 6.25 0.07 1.64 4.61 4.28 0.97 0.61

2.29 1.42 1.15 2.62 0.01 0.17 1.25 0.33 0.41 0.16

46.81 3.33 13.97 14.82 0.21 6.13 10.77 2.40 0.18 0.32

Loi*

22.10

7.44

9.59

0.46

9.74

1.76

1.74

p.p.m. V Cr Ni Sr Y Zr Nb La Ce Nd

500 129 72 392 62 301 35 29 77 54

88 28 15 209 23 59 7 13 15 6

146 43 226 61 269 30

26 18 60 16 39 8

495 132 63 371 73 371 43 45 83 63

92 30 9 67 26 56 4 13 37 20

374 185 89 294 37 196 23 60 36

* Loss on ignition (not recalculated). Mean and standard deviation (s.d.) of the basaltic tufts from the Faeroe shelf are based on analyses normalized to 100% without loss on ignition after subtraction of CO2 a s C a C O 3. Trace elements are similarly normalized. Source of data: Faeroe shelf tufts are analysed by XRF at Geological Survey of Greenland (majors) and Department of Earth Sciences, Leeds University (traces). Norwegian well data are from Maim et al. (1984) (calcite-free basis). U K well data are from Morton & Knox (1990). The basalt 35-9 is sample no. 9 from profile 35 in Rasmussen & Noe-Nygaard (1969), which is taken from the third highest flow in Tindur on Sandoy, Faeroe Islands. The new analysis is made by XRF at Leeds University. the same dredge contain spores and pollen but no dinoflagellate cysts a n d are p r o b a b l y nonm a r i n e (161-14 a n d 161-17). The volcanic ash m a k i n g up the tufts is d o m i n a n t l y vitric. T h e glass is c o m p l e t e l y replaced, m a i n l y by palagonite, but also by clays, zeolites a n d calcite. Altered clasts of o p a q u e basalt glass (tachylite) a n d of very finegrained glass-rich basalt are also present in small but variable a m o u n t s . The gas vesicles o f the vitric clasts are filled with clay, calcite or zeolites, which m a y m a k e t h e m indiscernible from the s u r r o u n d i n g altered glass. The a b u n d a n c e of f o r m e r vesicles varies widely in individual clasts but seems generally to be < 50%.

P l a g i o c l a s e occurs as p h e n o c r y s t s in the various types o f clasts and as isolated crystal fragments in the matrix forming up to a few per cent of the v o l u m e of the tufts. The phenocrysts a n d fragments fall mainly within the compositional range An76 to Ans0, but in m a n y samples a few fragments of andesine a n d / o r oligoclase are also seen, whereas m o r e calcic plagioclase (An81-82) has only been f o u n d in one sample. O t h e r minerals rarely observed as phenocrysts or small clasts include ilmenite, titanomagnetite, quartz, a m p h i b o l e and clinopyroxene. Fresh olivine has never been observed a n d clay or calcite p s e u d o m o r p h s after olivine seem to be uncommon.

PALEOCENE (part)

.~

OLIGOCENE

EOCENE

UPPER

MIDDLE

LOWER

UPPER

LOWER

MIOCENE UPPER

(part)

LOWER

0

~

e

7

7

'i! t'.iiiil;..... ~ 161-1 m

m 65-7

8-8

~ ........

RW ?

~--,

152-1

.....

L

=

I

o 65-6

...... m

=

=

87-1 8-6

76-12 = 5-32

~ ' " o 170-5

+RW

.....

, !

I o 154-2

RW

I

o

I

m

21-3

o 155-2

.....................

m

RW

I

o 156-8

RW

I

o155-12

RW ~

I I o 155-3 RW ~

I

I

o 20-116 +RW

I

o

+RW

21-2

I

o 71-11

I

o 5-36 o

I 154-1 I

....................................................................

+RW

.................

RW" . . . . m .

n 154-12 +RW..... ~

m

Fig. 4. I n t e r p r e t e d age o f the s e d i m e n t s b a s e d o n d i n o f l a g e l l a t e s . S o l i d lines i n d i c a t e the m o s t likely ages. R W , moderate reworking, + RW, stronger reworking.

PALAEOGENE VOLCANICLASTICS, FAEROES SHELF

187

Table 4. Dinoflagellates observed in dredged sediments from the Faeroe shelf. (Samples are listed in order of presumed age)

~

~

-

,

&~

.~ . ~. . . .

, ~

-

~

~1

I

I

1 2 3 4 5 6 7 8 9

XX

I

x

11

OI

O

~1 Xl

O

l

X

l

l

l

l

I

I

l

X X l X X ~ X l I

I I

~1111

I

I

I

~1~11

"~

x

~1111

I

I ~--- r

I I

[ [

I

I

I I

I

I

[

I

~IXO

I ..~

I

10 I

I

12

I

I

13

I

I

I

14

I

I

I

15

I

I

16

x

"~ 9~

I [

*~1 ~

17

x

~

I

o

x

x

I

I

I

I

I

20

I

I

I

I

I

21

I

I

I

I

I

I

22

I

I

I

I

I

I

23

x

x

x

x

x

X I

Xl

I

I

OI X

18

I

9-.z I

I

24 x t

IX

9~-.~ - ~ x

I

II

I

I

I

I

II

I

I

I

I

I

I

I

25 26

I .~ x

27

I ~

I

28

I

I I

I

30

Xll O

19

.~

29

I

X[II

X

XI~I I

X o

I

I

-~

I

~

31

I

III

I

I I I

32

I

III

I

I

I I

33

~

III

I

I I

I

34 35

x

I

I I I

I

I

I

I

37 I

38

I

I

I

I

I

I

39

I

I

I

I

I

40

I

I

i

I

41

I X

I

I

42

I

I

I

I

I

43

o

I

I

I

I

I

x

o

~.~

xx

~1

II .~1~

X

~

x

II1~1

I

I

I

x

I I ~ l l

L

I

I

x~

L

44 I [

x

x

I

I

I

x x

I

49

I

I t~

50

I

~ 1 1

II

I

I

I ~ 1

II

I

I

I

I

I

I ~ 1

Xl

I

t

I

I

I

~[

I

.~

I

I

It~

IX

~ 1 ~

I

;~

IIX

II

~111

I

I

I1~

II

IIII

I

I

46 48

IX

I

45

I I

~ 1 1

IIX

36

I

x

x.~

I

I

I

x x l

47

51 52

x

53 54 55

I

I

I

I

57

I

I

I

58

I

I

I

59

I x

56

Apectodinium homomorphum-quinquelatum g r o u p Areoligera cf. senonensis Cordosphaeridiu~ exilirnurum Cordosphaeridiumfibrospinosum Cordosphaeridium gracilis Diphyes colligerum Glaphyrocysta ordinata Hystrichosphaeridium tubiferum Lejeunecysta tenella Microdinium cf. ornatum i n H e i l m a n n - C l a u s e n ( 1 9 8 5 ) Microdinium s p . 1 i n C h a t e a u n e u f ( 1 9 8 0 ) Oligosphaeridium complex Paralecaniella indentata Thalassiphora delicata Trivalvadiniumformosum aft. Isabelidinium? viborgense Cribroperidinium tenuitabulatum Homotrybliurnfloripes-plectilum g r o u p Systematophora placacantha Deflandrea denticulata s u b s p , minor Eisenackia scrobiculata Fibrocysta cf. axialis Glaphyrocysta divaricata Hystrichokolpoma cincture Thalassiphora pelagica Wetzeliella lunaris Cordosphaeridium inodes Deflandrea phosphoritica Eatonicysta ursulae Selenopemphix aft. armata Wetzeliellas p . i n d e t . Pentadinium cf. taeniagerum i n D e C o n i n c k ( 1 9 8 5 ) Rottnestia borussica A . cf. diktyoplokus ( r e d u c e d p l a t f o r m s ) Areosphaeridium diktyoplokus sensu stricto Cerebrocysta bartonensis Diphyesficusoides Glaphyrocysta cf. spinetum Glaphyrocysta texta Nannoceratopsis s p . Operculodiniurn cf. nanaconulum Samlandia chlamydophora Turbiosphaera cf. magnifica Wetzeliellia articulata Achilleodinium biformoides Areosphaeridium arcuatum Cerebrocysta bartonensis i n K f t h e ( 1 9 9 0 ) Lentinia serrata Lingulodinium machaerophorum M . cf. aspinatum i n H e i l m a n n - C l a u s e n & Costa (1990) Charlesdowniea coleothrypta Deflandrea cf. heterophlycta Dracodinium pachydermum Hystrichosphaeropsis s p . 1 i n H e i l m a r m - C l a u s e n & C o s t a W. articulata - W. ovalis i n D e C o n i n c k ( 1 9 7 7 ) Chiropteridium cf. dispersum i n E a t o n ( 1 9 7 6 ) Cordosphaeridiumfuniculatum g r o u p Corrudinium incompositum Hystrichosphaeropsis s p . 1 i n M a n u m et al. ( 1 9 8 9 )

(1990)

188

R. WAAGSTEIN

& C. HEILMANN-CLAUSEN

4. continued

Table

''

~ & ~

<~--

~

~ 1 1 1

I

I

~ f l l ~ 1

I I

I

,

I

0 0 0 0

I

H i l l

I

I I

I

I

I

I

I

I

I

" ~ 1 ~ ~ 1 1 1 1 1 ~ 1 1 1 1

I

I

I

I o

I I "~

O O I I 1 ]

~ 0 ~ 0 0 ~ 1 1 1 1 ~ l l l l

I .-o I

l

~111~ I

~ l l l l o1~11 ~ l l l l ~ 1 ~

I

I

I

I

I I

t I

I I

I o

I

I

~1Ol Oll

I I

I I

I I I I

I ;~ o

I I

I

I

I

I I

I I I

I I ~ I I

No

I I

o J

I I

I

I I

I I I

I I I I I I

I I

t~O

I

104

I

105 106 107 108 109 110

o I

I ~r f I I I I I I

60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103

I I I f tJ.~ I

I t r I I o

111 112 113 114 115 116 117 118

A. diktyoplokus ( s h o r t processes) Dracodinium of. simile Phthanoperidinium geminatum Phthanoperidinium sp. indet. Pyxidinopsis densepunctata WilsonidiniumechinosuturatuminHeilmann-Clausen&Costa(1 Achomosphaera alcicornu G e n . et sp. in p r e p . ('Costacysta buccina') Hystrichokolpoma rigaudiae Tectatodinium pelliturn Adnatosphaeridiurn vittatum Cordosphaeridium cantharellum Dapsilidinium pseudocolligerum Distatodinium paradoxum in E a t o n (1976) Dracodinium cf. varielongitudum Homotryblium tenuispinosum Melitasphaeridium asterium Membranophoridium aspinatum Phthanoperidinium comatum Phthanoperidinium levimurum Reticulatosphaera actinocoronata Reticulatosphaera sp. n o v . ('pseudoursulae') Thalassiphora fenestrata Areosphaeridium capricornum Gerdiocysta conopeum Hemiplacophora semilunifera Riculacysta perforata Wetzeliella gochtii Areosphaeridium arcuatum o r A. pectiniforane Pentadinium lophophorum Svalbardella cooksoniae Areoligera semicirculata Distatodinium paradoxum Distatodinium tenerum Distatodinium virgatum Pentadinium laticinctum subsp, imaginatum Phthanoperidinium cf. amoenum tVetzeliella symmetrica Caligodiniurn sp. ( g r a n u l a r ) Chiropteridium lobospinosum Charlesdowniea columna Glaphyrocysta cf. microfenestrata Charlesdowniea variabilis Cordosphaeridium of. minimum Glaphyrocysta exuberans Hystrichokolpoma globula Palaeocystodinium golzowense Chiropteridium mespilanum Dracodinium solidum Rhombodinium draco Wetzeliella ovalis sensu stricto Caligodinium amiculum Achilleodinium sp. 1 in H e l l m a n n - C l a u s e n & C o s t a (1990) Distatodinium bijff~i Cyclopsiella elliptica Homotryblium pallidum Deflandrea spinulosa Homotryblium vallum Polysphaeridium congregatum

S y m b o l s in t h e table: o, c o m m o n o r a b u n d a n t ; 1-9, c o u n t e d s p e c i m e n s ; x, n o n - q u a n t i f i e d r e c o r d ; ?, u n c e r t a i n identification; c, cf., atypical. possibly w i t h i n c i r c u m s c r i p t i o n o f the t a x o n ; a, aft., a r e c o r d c o m p a r e d with, b u t different f r o m , t a x o n .

PALAEOGENE VOLCANICLASTICS, FAEROES SHELF The major- and trace-element composition of 18 tuff samples is shown in Table 2. The composition is strongly affected by secondary alteration. In order to try to compensate for the varying amounts of calcite cement we have subtracted all CO2 as CaCO3 and recalculated the major elements to 100% without loss on ignition (Loi). We have also normalized the trace elements by the same factor. The recalculated analyses show a fairly narrow range of concentrations for most elements and only their mean and standard deviation (s.d.) are therefore presented in Table 3. Three samples are not included in the mean because of anomalous high normalized concentrations of Fe203 (52.8%) or Y (180-240p.p.m.), but otherwise they are not much different. The recalculated compositions are clearly affected by the alteration of the glass but give some indication of the original composition of the volcanic ashes as discussed in a later section.

Stratigraphy. The basaltic tufts have been sampled nearly in situ in dredge hauls 157, 158 and 161 west of Sandoy Bank, and presumably also in 145 further south. Dredge haul 145 seems to have crossed the boundary between the lava plateau and the overlying sediments, and the tufts recovered probably represent the oldest sediments post-dating the Faeroe basalt plateau. The other three dredge hauls are made between 4.5 and 7km seaward of the basalt border representing increasingly higher stratigraphic levels in the following order: 158, 157 and 161 (Fig. 2). The basalt flows underlying the tufts presumably form the youngest preserved part of the Faeroe basalt plateau as judged by the gentle dome-like structure of the plateau (Waagstein 1988). Dinoflagellates have been found only in two out of 16 samples examined and they are both from dredge 161, representing the youngest tufts sampled nearly in situ. Sample 161-1 is fairly rich in dinoflagellates, whereas sample 161-2 is richer in inertinite, land-plant debris and various pollen types. Both samples include common Paralecaniella indentata, a questionable dinoflagellate cyst which is often (but not only) represented in marginally marine sediments. For sample 161-1 an Early Eocene (middle Ypresian) age is safely indicated on the basis of the presence of Hystrichosphaeridium tubiferum (common), Eatonicysta ursulae subs. ursulae (common), Wetzeliella sp. and Deflandrea phosphoritica (D. solidum-C, coleothrypta Zone interval). Sample 161-2 is probably earliest Eocene or latest Palaeocene, based on several specimens of Hystrichosphaeridium tubiferum

189

and common representatives of the Apectodinium homornorphum group, combined with the absence of Eatonicysta ursulae. Microdinium sp. 1 in Chateauneuf (1980) (two specimens) has a little known range. It is hitherto occasionally recorded from Middle Eocene-Lower Oligocene strata. The remaining 14 basaltic tufts examined contain only spores and pollen and other matter of terrestrial origin and are probably nonmarine. Because of their petrographical similarities we believe that all the tufts form a single stratigraphical unit overlying the Faeroe basalt plateau. The two uppermost basalt formations of the Faeroe Islands are formed within magnetochron C24R (Waagstein 1988) which straddles the Palaeocene-Eocene boundary (Aubry et al. 1988). The basaltic tufts are thus presumably not older than the latest Palaeocene, and assuming that the two above-mentioned marine tufts are from near the top of the unit (Fig. 2) they cannot be younger than middle Ypresian. The unit has a total estimated thickness of 0.35-0.40 s. two-way travel time (twt) in Fig. 2, equivalent to c. 400 m.

B. Lower and Middle Eocene tuffaceous limestones and phosphatic sediments Description. The group B sediments are represented by only 19 rocks and they are all glacial erratics. However, they make up 43% of the total number of volcaniclastics from glacial deposits. The tuffaceous sediments are matrix supported with a turbid recrystallized calcium carbonate matrix averaging between 0.005 and 0.1 mm in grain size, often with some zeolites (Fig. 3c). The sediments contain dinoflagellates, i.e. are marine. Most of the samples show evidence of bioturbation and many carry pellets. The pellets are usually made up of collophane, a yellowish isotropic phosphate, which may comprise 20-30% of the rock (4-64, 5-12 and 65-6; Fig. 3d). The group B sediments contain variable amounts of altered vitric volcanic ash with a maximum grain size of between 0.1 to 1.3 ram. In sample 5-3 the ash forms a > 8cm thick, graded bed but regular bedding is not observed in any other sample, probably because of intense bioturbation. Sample 4-137 is cut by a thin vein of pyritized basalt. The largest vitric clasts are often strongly vesicular with an irregular indented outline while small ones are mostly bubble-wall fragments (Fig. 3c & d). The glass is usually replaced by brownish alteration products like palagonite, less often by a colourless gel, zeolites or calcite. The tuffaceous sediments contain sparse

190

R. WAAGSTEIN & C. HEILMANN-CLAUSEN

phenocrysts and fragments of plagioclase. The plagioclase is mainly labradorite and andesine, although it ranges in composition from bytownite to oligoclase. Plagioclase of andesine composition seems to be more common than in group A basaltic tufts, as do phenocrysts or fragments of iron-titanium oxides. Tuffaceous limestone 4-140 is rather unusual in containing in addition a few fragments/phenocrysts of amphibole, biotite, aegirine and ruffle. Judged by the mineral content, the volcanic component of the group B sediments is dominantly basaltic, but eruptions of more evolved magmas of andesitic or rhyolitic composition also seem to be represented in several samples, and at least one (4-140) shows evidence of highly alkaline volcanism. The vesicularity and shape of the vitric clasts suggest that the ashes have been produced by both phreatomagmatic and magmatic eruptions.

Stratigraphy. Almost half of the tuffaceous sediments have been examined palynologically. In addition to dinoflagellates the organic matter in palynological preparations includes diverse terrestrially derived particles (pollen, spores, wood tissue and cuticle fragments). There are usually equal amounts of marine and nonmarine organic particles, or a dominance of non-marine matter. Only two samples (76-12 and 8-6) are dominated by dinoflagellates. The group B sediments are interpreted to be of Early Eocene (Ypresian) and early Middle Eocene (Lutetian) age (Fig. 4). The two oldest samples (5-12 and 3-16) are of a slightly older or similar age as the youngest group A basaltic tuff (161-1). For the tuff-rich limestone 3-16 an Early Eocene age is safely indicated. The presence of Wetzeliella lunaris (common), Hystrichosphaeridium tubiferum, Apectodinium homomorphum (several specimens) and Hystrichokolpoma cinctum, combined with the lack of Eatonicysta ursulae, thus suggests an early or possibly middle Ypresian age (the W. meckelfeldensis-D, varielongitudum Zone interval). The tuffaceous, phosphatic pelletstone 5-12 has a suggested Early Eocene age based on Areoligera cf. senonensis. The species-poor assemblage lacks more age-diagnostic forms. However, a single Microdinium sp. 1 in Chateauneuf (1980) is very similar to the specimens in sample 161-2 of group A. The remaining seven samples, with the possible exception of 5-32 range in age from middle? Ypresian to Lutetian (Fig. 4). The youngest sample is probably 5-32, for which a late Lutetian to possibly earliest Bartonian age is indicated by the presence of an undescribed

form ('Costacysta buccina') known from the North Sea and Denmark, and by Areosphaeridium arcuatum, which is common. Diphyes cf. ficusoides (transitional to D. colligerum) is previously known from the upper Lutetian in Denmark. No exposures of the tuffaceous sediments have been proven as yet. However, their age suggests that they should be present in the lower eastern slope of the marginal channel west of Sandoy, possibly beneath a thin cover of surficial sediments (Fig. 2). Only dredge 156 has been made so far in this part of the channel. It recovered abundant Recent corals and a few rocks. However, the rocks are presumably all glacial erratics including a single piece of feldspathic sandstone similar to the sandstone from dredge 155 (see below). The estimated thickness of tuffaceous sediments which could be accommodated in this place is c. 250 m (0.25 s. twt).

C. Oligocene feldspathic and lithic volcanic sandstones Description.This last group of marine sediments bears evidence of the erosion of the Faeroe basalt plateau. Some of the sandstones have been sampled nearly in situ from the sub-sea exposures on the western slope of Sandoy Bank (dredges 154 and 155), while the rest are from dredges of glacial erratics. The sandstones form 32~ of the total number of volcaniclastic erratics in dredges 154 and 155. The lithic sandstones are represented only by the two glacial erratics 20-117 and 21-2, both from near the shelf edge west of Suduroy. They are medium to coarse grained carbonate-cemented sandstones with rounded clasts of altered fine-grained basalt lava (Fig. 3b). The basalt seems to have been fairly rich in olivine, like many flows from the two uppermost basalt formations in the Faeroe Islands. Most of the feldspathic sandstones are fine grained, well sorted, with a median grain size of the order of 0.1 mm. Many sandstones sampled nearly in situ west of Sandoy Bank are poorly cemented, whereas almost all glacial erratics are indurated with carbonate. The sandstones show signs of bioturbation. Abundant Palaeophycus tubularis burrows, made by a suspension feeder, occur in very fine-grained sandstones from dredge 155 indicating a fairly energy- and oxygen-rich environment (R. Bromley, pers. comm.). The sand fraction consists of rounded clasts of microcrystalline clay and of more or less angular

PALAEOGENE VOLCANICLASTICS, FAEROES SHELF

== 0~

,,=, ,~

o o ~r 1' 6~2

_{

eo,, =..,, =-i.== o-9r=1~6o2o~ o- . = O'N

SOUT. WESTERN ROCKALL PLATEAU

~ " = Ii =-== =

(based o~ Uo.o~

~'OE

Keene 1984)

(based on Soper & Costa

S

1976,

Tuffaceous limestones / / /'""

Volcanic phase 3 (minor air fall tufts)

!,:: ---ji NP11

i

VORING PLATEAU (based On ODP leg 104)

Kap Dalton Formation, lower part sediments) (shallow marine

/'

2d

(lower part)

"

f - ].- - E

. ""

.

.

.

-4 2c

Basaltic tufts Volcanic phase 2 (air fall tufts)

9

Qui . . . .

t interval (marginal marine sediments)

'E ' .

.

.

Upper basalt series with interbedded tufts (dipping reflectors) I : I 2 E ! E ] 2 E

2b

(= main ash marker in the Balder Form.)

Blosseville Group (basalts)

ash -17

.

Upper and middle basalt formations

Volcanic phase 1 (mainly basalts)

Lower basalt series

negative series

il 7~ - t-,l,

_i '~

7

--. NP 8 (upper part)

positive series

--

2a

NP 9

Danish ash phasesof Beggi d 1918)

/

' ._F_-I__I_: . . . . .

- - "

NORTH SEA BASIN tuff phases of Knox (1983, 1984)

Soper et al. 1976)

I ~ C24N

CENTRALEAST GREENLAND

FAEROEBLOCK

191

i

I

.... OseO ..... T

Lower basalt formation (upper part)

i

I

Fig. 5. Correlation of major basalt and tuff phases in various parts of the North Atlantic region at the Palaeocene-Eocene boundary. An internationally agreed standard Palaeocene-Eocene boundary is not yet available and we have arbitrarily defined this boundary as the base of the calcareous nannoplankton NP10 Zone, following a common practice. The calibration between the magnetostratigraphy and calcareous nannoplankton zonation is from Aubry et aL (1988). The correlation is based on our interpretation of published bio- and magnetostratigraphic data. The dinoflagellate Apectodinium augustum, shown in the left column, has only been recorded with certainty in the North Sea Basin and on the Vering Plateau. Records from the Rockall Plateau (Costa & Downie 1979; Brown & Downie 1984) are not typical according to the authors. The acme of the genus Apectodinium refers to the North Sea Basin. Stratigraphic control in the areas concerned is generally limited, except in the southwestern Rockall Plateau. The correlations are therefore tentative. In central East Greenland the dinoflageUate Apectodinium homomorphum occurs in the lowermost part of the Blosseville Group (Soper et al. 1976b). Palaeomagnetic studies of the Blosseville Group (Tarling et aL 1988) locate the extrusion of basalt lavas to within the magnetochron C24R. In the Voring Plateau the stratigraphic position of the lower basalt series is based on the occurrence of typical Apectodinium augustum figured in Boulter & Manum (1989). This is perhaps in conflict with the magnetostratigraphy. The palaeomagnetic polarity of the lower series is tentatively interpreted originally to have been normal (Schrnharting & Abrahamsen 1989). It is thus possible that the lower basalt series should instead be placed within magnetochron C25N although such a low position would seem to conflict with the range of Apectodinium augustum as known from the North Sea Basin.

fragments of plagioclase, F e - T i oxides, zeolites and clinopyroxene (Fig. 3f). Foraminifera, calcispheres and echinoid spines occur in several rocks, while pellets are less c o m m o n . The feldspar has an average composition o f about An60-7o close to that in Faeroe basalts (cf. Hald & W a a g s t ein 1984; Waagstein & Hald 1984). The pyroxene is an augite, but this mineral m a y be missing due to complete replacement by

calcite. Although the clay clasts form roughly h a l f o f the sand particles their origin is uncertain. They are sub-rounded and consist o f yellowish or brownish microcrystalline clay identified as smectite with the microprobe. This is a typical alteration product o f basaltic glass, but the clasts very rarely show any internal structures, suggesting the former presence o f microlites or vesicles. A few of the clay clasts

192

R. WAAGSTEIN & C. HEILMANN-CLAUSEN

may show a greenish colour and they consist of glauconite rather than smectite.

Stratigraphy. Fourteen group C sediments have been examined palynologically. They contain dinoflagellate assemblages which may be relatively diverse. In addition they contain variable, often considerable, amounts of diverse terrestrially-derived organic particles. Most of the samples are of an Early Oligocene (Rupelian) age (Fig. 4). Sample 170-5 is possibly slightly older. An earliest Oligocene or possibly latest Eocene age is indicated by Thalassiphora fenestrata (range: Zone D12nb-D13) and an undescribed species of Reticulatosphaera, R. "pseudoursulae'. The latter form has previously been recorded in Zone D14na in the North Sea by the second author, but was absent in Zone D12 and below. This sample has a more abundant calcite matrix than the other feldspathic sandstones. The sandstones from dredge 155 and 156 are all of early Rupelian age. Samples 154-1 and 154-12 are of late Rupelian or early Chattian age in accordance with the higher stratigraphic position of dredge 154. Dredge 154-2 is more likely early Rupelian in age but could have been sampled more westerly along the almost 3 km long dredge path than the other two. The three sandstones examined from dredges 20 and 21 west of Suduroy are probably all Rupelian in age. The majority of the sandstones east of the Faeroe basalt platform contains reworked dinoflagellates of Early and Middle Eocene age in addition to the indigenous assemblage (Table 4). Two sandstones from west of the platform contain reworked dinoflagellates of Mid-Late Eocene age (20-116, 21-2). The reworked dinoflagellates may be abundant irrespective of the age of the sandstone or its geographical position. It is thus possible that the aberrant latest Ypresian or Lutetian age tentatively assigned to sandstone 65-6 (Fig. 4) may be erroneous. The sample contains very few dinoflagellates (Table 4), and the age is based on a single specimen of Diphyes ficusoides, which could be reworked. The Oligocene sandstones have a total thickness of c. 250m (0.25 s. twt) east of the Faeroe Islands (Fig. 2). Discus~on

Origin of basaltic tufts The basaltic tufts forming the lowermost sedimentary unit on top of the Faeroe basalt plateau are dominantly non-marine and more or less

reworked. Evidence of reworking includes: (1) clear signs of rounding of the vitric clasts in many coarse-grained tufts; (2) a wide range of plagioclase compositions in individual rocks suggesting a mixture of ashes from magmas of different compositions; and (3) lack of graded bedding. The basaltic tufts are assumed to originate from phreatomagmatic eruptions, which are extremely explosive due to the interaction of magma and ground water or shallow surface water. The interpretation of eruption type is based on the high abundance of former sideromelane glass (now palagonite) within the tufts, the generally low but variable vesicularity of the glass and the abundance of small blocky vitric clasts (cf. Fisher & Schmincke 1984; Heiken & Wohletz 1985). The original chemical composition of the volcanic ash has been greatly modified by alteration and cementation. Differences in cementation may be partly accounted for by subtraction of CO2 as CaCO3 (Table 3). The effect of the alteration of the glass on composition is more difficult to estimate. However, some elements seem to be almost immobile during alteration of basaltic glass and they are concentrated together in the palagonite or smectite replacing the glass due to selective loss of the major elements Ca, Mg, Si and A1. In a recent study of a sample of fresh and altered glass dredged from the Mid-Atlantic Ridge Nb, Zr, Ce, Ti, Fe and Mn seem immobile and are passively concentrated in the replacing smectite by a factor ofc. 1.81 relative to the fresh glass on a water-free basis (Bienvenu et al. 1990). A comparison of the average composition of Faeroe shelf tufts (Table 3) with the above smectite suggests that the tufts have experienced less extreme losses of the above major elements and thus a smaller passive concentration of immobile elements (ignoring Mn because of its substitution for Ca in the calcite matrix). A concentration factor for immobile elements of c. 1.33 is probably more realistic, which gives 3.0% TiO2, 14.9% Fe203, 226 p.p.m. Zr, 26 p.p.m. Nb and 58 p.p.m. Ce on average in the original volcanic ashes. This is fairly similar to the concentrations found in some of the most ironand titanium-rich dykes and flows in the upper basalt formation of the Faeroe Islands (Hald & Waagstein 1991; Table 3). The abundance of immobile elements thus strongly suggests that the tufts dredged from the Faeroe shelf originate from Fe-Ti-type tholeiitic magmas. The sodic composition of the plagioclase phenocrysts in the tufts lends further support to this suggestion. Because of the large total thickness of the

PALAEOGENE VOLCANICLASTICS, FAEROES SHELF tufts and the presence of vitric clasts > 5mm in diameter, the major part of the tuff must originate from eruption sites within the Faeroe block itself. It is thus possible that some sites are represented by dykes within the present Faeroe Islands. The shift in volcanic style from effusion of basalt flows to phreatomagmatic eruptions bears witness to an increase in magma-water interactions. The most likely cause of this change is that the top of the basalt plateau came close to sea level resulting in poor drainage, formation of shallow lakes and perhaps marine incursions. Usually the base of a basalt plateau subsides during build-up in response to the removal of magma from below and addition of solid basalt above. The inferred relative rise of sea level may thus simply be due, at least in part, to a slowing down of lava production in the final stage of volcanism.

Correlation with North Sea tufts The tufts on top of the Faeroe basalt plateau testify to a major phase of explosive volcanism, evidence of which is likely to be preserved in the sedimentary record of neighbouring basinal areas. Volcanic ashes are found at several levels in the Palaeocene and Eocene of the North Sea, and the ash falls seem to have culminated in the pyroclastic subphase 2b of Knox & Morton (1983) within the Balder Formation, i.e. in the earliest Ypresian (Fig. 5). The phase 2b tufts are almost exclusively basaltic and although they are usually strongly altered, fresh glass is preserved in 2b ashes within the mo-clay (diatomite) of the Fur Formation in Denmark, confirming that the basaltic magmas are of high Fe-Ti tholeiitic type (Pedersen et al. 1975; Morton & Evans 1988). The total thickness of the 2b tufts increases northwards from c. 1.5m in northernmost Germany to > 8 m east of the Shetland Islands. The general thickness trends and the major and trace element chemistry of the tufts suggest that the eruption sites were located on the protoGreenland-Scotland Ridge (Knox & Morton 1988; Morton & Knox 1990). Smythe (1983) tentatively correlates the above North Sea ash marker with the coal-bearing sequence and overlying tufts and agglomerates between the lower and middle basalt formations of the Faeroe Islands on the basis of seismic and palaeontological evidence. However, the pyroclastic rocks overlying the Faeroese coal sequence are distinctly richer in MgO and generally poorer in TiO2 than the magmas considered to have produced the 2b tufts (cf. Waagstein & Hald 1984). Although Morton et.

193

al. (1988) claim that geochemical studies in progress on Balder Formation tufts from commercial boreholes west of Shetland confirm Smythe's correlation, they seem to consider the phase 2b tufts and the middle and upper formations of the Faeroes to be broadly contemporaneous, all being formed during the magnetochron 24R. The pollen and spore assemblage in the Faeroe coals (Lund 1983, 1989) is closely comparable to that of the Lower Basalt Series in the Voring Plateau, whereas the assemblage in the North Sea ash marker is largely different (Boulter & Manum 1989). According to Morton & Knox (1990) the best geochemical correspondence of the Balder tufts is with some high Fe-Ti dykes in East Greenland. The new data presented here on tufts from the Faeroe shelf shows that they resemble the Balder Formation tufts both geochemically and in terms of phenocryst contents (Table 3; Maim et al. 1984; Morton & Knox 1990). The geochemical correspondence is especially convincing for concentrations and inter-element ratios of the so-called immobile elements, suggesting a common source of magma. We therefore consider the Faeroe shelf tufts to represent proximal deposits of the phase 2b tufts of the North Sea. Because of reworking, the present thickness of the Faeroe shelf tufts does not directly reflect the thickness of pyroclastics primarily formed on top of the basalt plateau, although they presumably have been much thicker here than in the North Sea area located far beyond the plateau. The correlation with the phase 2b tufts of the North Sea allows us to refer most of the Faeroe shelf tufts to the NP10 chronozone (Knox 1984) in the lowermost part of the Eocene. The top of the tuff sequence east of the islands is probably located in t h e lower part of the NP11 chronozone according to the present study (Fig. 4), and the observed reworking of the tufts may thus, to some extent, have taken place penecontemporaneously with the volcanic activity forming the original pyroclastic deposits. Correlation with seaward-dipping reflector sequences A wedge of seaward-dipping reflectors occurs along the northern margin of the Faeroe Block. The reflectors are interpreted as basalt flows formed subaeriaUy during the initial stage of opening between the Faeroes and Greenland (Smythe 1983). Seaward-dipping reflector sequences are characteristic of volcanic continental margins and are found in many areas

194

R. WAAGSTEIN & C. HEILMANN-CLAUSEN

flanking the North East Atlantic (White et al. 1987). Their volcanic nature has been proven by drilling into the Voting Plateau and Southwest Rockall Plateau margins northeast and southwest of the Faeroe Block, respectively. The seaward-dipping flows are assumed to have been erupted from the active rift located further seaward and their present dip is explained as a result of rapid subsidence of the oceanic lithosphere forming continuously along the rift. In the Voring Plateau margin ODP hole 642E penetrated 900 m into the volcanic basement. A 770 m thick upper series of subaetial basalt flows forms the dipping reflector sequence and is underlain by an at least 140 m thick lower series of andesitic and basalto-andesitic flows partly derived from fusion of continental crust (Viereck et al. 1989). Dinoflagellates occurring in sediments interbedded between the flows of the lower series in Hole 642E include Apectodinium augustum (Boulter & Manum 1989). A. augusturn has hitherto been reported only within the zone of that name in the North Sea Basin and it is most probable that the occurrence in the Voting margin indicates a similar age (Fig. 5). The zone in the North Sea Basin is restricted to the lower part of the Sele Formation and equivalent onshore formations from Denmark, i.e. the uppermost Palaeocene, below the Balder ashes. The oldest post-basaltic sediments recovered from the Voring margin are from Hole 642D (Manum et al. 1989). The dinoflagellate assemblage includes, in particular, Eatonicysta ursulae and Charlesdowniea coleothrypta, which indicate a late Ypresian age, distinctly younger than the Balder tufts. According to these age indicators below and above, the dipping reflector sequence in the Voring Plateau margin may, therefore, be coeval with the Balder tufts. A correspondence in age with the Balder tufts is supported by the presence of tufts interbedded between the flows of the upper series. The tufts in the upper half of the series seem, in general, to originate from magmas richer in Ti-Fe than the flows (Vierick et al. 1989), and they resemble the Balder tufts in both major and trace element chemistry (Viereck et al. 1988; Morton & Knox 1990). The oldest dipping reflectors north of the Faeroes are probably of a similar age to those off the Varing Plateau, both being formed during the initial opening of the Norwegian Sea, although the opening may possibly have been slightly diachronous due to northward propagation of the rift as suggested by Larsen (1988). Accepting the above interpretations, they are therefore also similar in age to the Balder Formation tufts in the North Sea and their

suggested proximal deposits on the Faeroe Block. The shift of volcanism in the interior parts of the Faeroe Block, from quiet effusion of lava flows to violent phreatomagmatic eruptions of basaltic ashes, very probably took place approximately at the same time as the onset of seafloor spreading along the northern margin of the block and the formation of the dipping reflector sequence (cf. Waagstein 1988). This synchronism may be due to a reduction of the accumulations of flows on the block itself by northward channelling of magma to the newly formed rift. The reduced accumulation rate might have caused a relative sea-level rise and increased magma-water interactions leading to the explosive volcanism as discussed earlier. Eocene transgression and continued volcanic activity The group A basaltic tufts, which are nearly all non-marine, are overlain by the group B tuffaceous limestones and minor phosphatic sediments, all deposited in the sea. The transgression occurred in Ypresian, probably near the NP10-NP11 boundary, although the exact time is uncertain because the precise age of the oldest marine samples is ill defined (Fig. 4). The group B sediments have only been found as glacial erratics and only on the shelf and slope east and southeast of the Faeroe Islands; they probably form a narrow continuous band of subcrops along the eastern margin of the basalt platform (Fig. 2). Marine Eocene sediments have probably once covered a larger part of the basalt platform judged by the presence of reworked dinoflagellates in the overlying Oligocene sandstones, as discussed below. However, terrestrial organic material is common or abundant in most of the limestones suggesting the proximity of land. The Eocene sea was at least bounded to the northwest by new land on the adjoining part of the Iceland-Faeroe Ridge, where an increased thickness of seaward-dipping reflectors were formed during the early subaerial phase of opening between the Faeroe Islands and Greenland (Smythe 1983). The East Greenland basalt plateau became partly submerged at about the same time. At Kap Dalton, the first marine incursion is recorded c. 300 m below the top of basalts in a shale belonging to the Wetzeliella meckelfeldensis Zone (Soper & Costa 1976; Soper et al. 1976a) at about the NP10-NP11 boundary (Fig. 5). Shallow marine sediments, which may possibly be assigned to the Dracodinium varielongitudum Zone, overlie the basalts with no

PALAEOGENE VOLCANICLASTICS, FAEROES SHELF marked discordance (Soper et al. 1976a). The D. varielongitudum Zone spans the boundary between the nanoplankton zones N P l l and NP12. The ubiquitous volcanic component of supposed airfall origin in the sediments shows that vigorous volcanic activity continued within the region throughout Ypresian and Lutetian time. The presence of a thin basalt vein in a group B limestone (4-137) suggests that the submerged part of the Faeroe Block itself was still volcanically active. However, the mixed character of the volcanic ashes bear evidence not only of basaltic but also acid and minor highly alkaline volcanism suggesting that most of the ashes erupted from localized volcanic centres. Except for a steep gravity low, northwest of the Faeroe Islands, which may possibly be interpreted as a small granitic pluton (Fleischer et al. 1974), no such centres have been identified on the Faeroe Block. However, potential sources exist to the west and north of the Faeroe Islands, in the then nearby East Greenland area and also to the south and east of the islands. In those areas several igneous centres have been identified, some of which were probably active in Eocene time (Brooks & Nielsen 1982; Roberts et al. 1983; Hitchen 1992). Tectonism and erosion o f the basalt plateau The palynological datings of the dredged sediments suggest the presence of an unconformity encompassing all, or almost all, of the Bartonian and Priabonian (Fig. 4). The overlying Lower Oligocene feldspathic and lithic sandstones have a mineral composition which shows that they are derived from erosion of basalt. The source of this basalt is probably the present Faeroe Islands and the surrounding platform which must, therefore, have been partly above sea level when the sandstones were deposited. The common presence of reworked Early and Middle Eocene dinoflagellates in the Oligocene sandstones east of the basalt platform suggests that the Eocene group B limestones partly covered the basalt and that they were simultaneously eroded. The western platform margin may have been covered by somewhat younger Eocene sediments judged by the occurrence of reworked Middle-Late Eocene dinoflagellates in Oligocene sandstones from here. Thus, there is evidence of a major regression in the later part of the Eocene followed by extensive erosion of the central part of the basalt plateau and overlying Eocene sediments in the Early Oligocene. The magnitude of these events can hardly be explained by eustatic sea-level

195

changes alone, but must be the result of local tectonic processes. Major uplift thus seems to have occurred in Bartonian-Priabonian time (c. 43-37 Ma ago). The tectonism may be due to compression causing differential uplift of the basalt plateau (Boldreel & Andersen 1993), possibly related to a phase of plate reorganization in the Norwegian Sea. Also at Kap Dalton, in central East Greenland, Lower Oligocene sediments seem to overlie Eocene strata unconformably (Soper & Costa 1976).

Conclusions (1) A major explosive volcanic phase of high Fe-Ti tholeiitic composition succeeded the formation of the Faeroe basalt plateau at the onset of seafloor spreading in the Norwegian Sea and is probably the main source of the Balder Formation tufts in the North Sea. (2) The Faeroe Block became at least partially submerged in the earliest Eocene and tuffaceous limestones and phosphatic sediments were deposited over the block during Early and Middle Eocene time. Volcanic ashes of variable compositions were erupted concurrently from central volcanoes. The Faeroe Block was uplifted again in the Middle or Late Eocene (BartonianPriabonian times), possibly due to compression. (3) In the Early Oligocene the Faeroe shelf subsided and feldspathic and lithic volcanic sandstones derived from erosion of the Faeroe basalt plateau and overlying Eocene sediments were deposited.

References AUBRY, M.-P., BERGGREN, W. A., KENT, D. V., FLYNN, J. J., KLITGORD, K. D., OBRADOVICH, J. D. & PROTHERO, D. W. 1988. Paleogene geochronology; an integrated approach. Paleoceanography, 3, 707-742. BIENVENU, P., BOUGAULT,H., JORON, J. L., TREUIL, M. & DMITRIEV, L. 1990. MORB alteration: Rare-earth element/non-rare-earth hygromagmaphile element fractionation. Chemical Geology, 82, 1-14. BOLDREEL, L. O. • ANDERSEN, M. S. 1993. Late Paleocene to Miocene compression in the FaeroeRockall area. In: PARKER,J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 1025-1034. ,- M. S. 1995. The relationship between the distribution of Tertiary sediments and tectonic processes and deep water circulation around the Faeroe Islands. This volume.

196

R. WAAGSTEIN & C. HEILMANN-CLAUSEN

BOTT, M. H. P., SUNDERLAND, J., SMITH, P. J., CASTES, U. & SAXOV, S. 1974. Evidence for continental crust beneath the Faeroe Islands. Nature, 248, 202-204. BOULTER, M. C. & MANUM, S. B. 1989. The BritoArctic igneous Province Flora around the Palaeocene-Eocene boundary. In: ELDHOLM, O., THIEDE, J., TAYLOR, E., ET AL. (eds) Proceedings Ocean Drilling Program, Scientific Results, 104, 663-680. BROOKS, C. K. & NIELSEN, T. F. n . 1982. The E Greenland continental margin: a transition between oceanic and continental magmatism. Journal of the Geological Society, London, 139, 265275. BROWN, S. & DOWNIE, C. 1984. Dinoflagellate cyst biostratigraphy of late Paleocene and Early Eocene sediments from Holes 552, 553A, and 555, Leg 81, Deep Sea Drilling Project (Rockall Plateau). In: ROBERTS, D. G., SCHNITKER,n., ET AL. (eds) Initial Reports of the Deep Sea Drilling Project, 81, 565-579. B~,zGILD, O. B. 1918. Den vulkanske Aske i Moleret samt en Oversigt over Danmarks afldre Tertia~rbja~rgarter. Danmarks Geologiske Undersogelse, H R~ekke, 3 3 . CHATEAUNEUF, J.-J. 1980. Palynostratigraphie et paleoclimatologie de l'Eocene superieur et de l'Oligocene du Bassin de Paris. Memoire du B.R.G.M., 116. COSTA, L. I. & DOWNIE, C. 1979. Cenozoic dinocyst stratigraphy of Sites 403 to 406 (Rockall Plateau), IPOD, Leg 48. In: MONTADERT, L., ROBERTS, D. G., ETM~. (eds) Initial Reports of the Deep Sea Drilling Project, 48, 513-529. & MANUM, S. B. 1988. The description of the interregional zonation of the Paleogene (D 1 - D 15) and the Miocene (D 16 - D 20). In: VINKEN, R. (ed.) The Northwest European Tertiary Basin. Geologische Jahrbuch, Reihe A, 100, 321-330. EARLE, M. M., JANKOWSKI,E. J. & VANS, I. R. 1989. Structural and stratigraphic evolution of the Faeroe-Shetland Channel and northern Rockall Trough. In: TANKARD, A. J. & BALKWlLL,H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins. American Association of Petroleum Geologists Memoir, 46, 461-469. FISHER, R . V . & SCHMINCKE,H.-U. 1984. Pyroclastic Rocks. Springer Verlag, Berlin. FLEISCHER, O., HOLZKAMM, F., VOLLBRECHT, K. & VOPPEL, O. 1974. Die Struktur des Island-FiirrerRtickens aus geophysikalischen Messungen. Deutsches Hydrographische Zeitschrift, 27, 97113. HALD, N. & WAAGSTEIN, R. 1984. Lithology and chemistry of a 2-km sequence of Lower Tertiary tholeiitic lavas drilled on Suduroy, Faeroe Islands (Lopra-1). In: BERTHELSEN, O., NOE-NYGAARD, A. & RASMUSSEN, J. (eds) The Deep Drilling Project 1980-1981 in the Faeroe Islands. Foroya Frodskaparfelag, Tfrshavn, 15-38. & 1991. The dykes and sills of the Early Tertiary Faeroe Island basalt plateau. Transactions of the Royal Society, Edinburgh, 82, 373-

388. HEIKEN, G. & WOHLETZ, K. 1985. Volcanic Ash. University of California Press, Berkeley. HEILMANN-CLAUSEN, C. 1985. Dinoflagellate stratigraphy of the uppermost Danian to Ypresian in the Viborg 1 borehole, central Jylland, Denmark. Danmarks Geologiske Undersogelse, Serie A, 7. 1988. Denmark, Paleogene dinoflagellates. In: VINKEN, R. (ed.) The Northwest European Tertiary Basin. Geologische Jahrbuch, Reihe A, 100, 339-343. HITCHES, K. 1992. The crustal characteristics, volcanic and sedimentary history of the Rockall continental margin. British Geological Survey Technical Report, WB/92/7. HOLTEDAHL, O. 1970. On the morphology of the West Greenland shelf with general remarks on the 'marginal channel' problem. Marine Geology, 8, 155-172. KNOX, R. W. O'B. 1984. Nannoplankton zonation and the Paleocene/Eocene boundary beds of NW Europe: an indirect correlation by means of volcanic ash layers. Journal of the Geological Society, London, 141, 993-999. & MORTON, A. C. 1983. Stratigraphical distribution of Early Palaeogene pyroclastic deposits in the North Sea basin. Proceedings of the Yorkshire Geological Society, 44, 355-363. - &- 1988. The record of early Tertiary N Atlantic volcanism in sediments of the North Sea Basin. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 407-419. KOTHE, A. 1990. Palaeogene dinoflagellates from Northwest Germany - biostratigraphy and paleoenvironment. Geologische Jahrbuch, Reihe A, 118, 1-111. LARSEN, H. C. 1988. A multiple and propagating rift model of the NE Atlantic. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 157158. LUND, J. 1983. Biostratigraphy of interbasaltic coals from the Faeroe Islands. In: BOTT, M. H. P., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) Structure and Development of the GreenlandScotland Ridge. Plenum Press, New York, 417423. 1989. A late Paleocene non-marine microflora from the interbasaltic coals of the Faeroe Islands, North Atlantic. Bulletin of the Geological Society of Denmark, 37, 181-203. MALM, O. A., CHRISTENSEN, O. B., FURNES, H., L~rLIE, R., RUSELATTEN, H. & ~STBY, K. L. 1984. The Lower Tertiary Balder Formation: an organogenic and tuffaceous deposit in the North Sea region. In: SPENCER, A. M., e r A~.. (eds) Petroleum Geology of the North European Margin. Graham & Trotman, London, 149-170. MANUM, S. B., BOULTER,M. C., GUNNARSDOTTIR,H., RANGNES, K. & SCHOLZE, A. 1989. Eocene to Miocene palynology of the Norwegian Sea (ODP

PALAEOGENE VOLCANICLASTICS, FAEROES SHELF Leg 104). In: ELDHOLM, O., THIEDE, J., TAYLOR, E., ET AL. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 104, 611-662. MORTON, A. C. & EVANS,J. A. 1988. Geochemistry of basaltic ash beds from the Fur Formation, Island of Fur, Denmark. Bulletin of the Geological Society of Denmark, 37, 1-9. - & KEENE, J. B. 1984. Paleogene pyroclastic volcanism in the southwest Rockall Plateau. In: ROBERTS, D. G., SCHNITKER, D., e r AL. (eds) Initial Reports of the Deep Sea Drilling Project, 81, 633-643. - & KNOX, R. W. O'B. 1990. Geochemistry of late Palaeocene and early Eocene tephras from the North Sea Basin. Journal of the Geological Society, London, 147, 425-437. , EVANS, D., HARLAND, R., KING, C. & RITCHIE, D. K. 1988. Volcanic ash in a cored borehole W of the Shetland Islands. Evidence for Selandian (late Palaeocene) volcanism in the Faeroes region. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 263-269. NIELSEN, P. H., WAAGSTEIN, R., RASMUSSEN, J. & LARSEN, B. 1979. Marine seismic investigation of the shelf around the Faeroe Islands. Frodskaparrit, T6rshavn, 27, 102-113. PEDERSEN, A. K., ENGELL, J. & RONSBO, J. G. 1975. Early Tertiary volcanism in the Skagerak: New chemical evidence from ash-layers in the mo-clay of northern Denmark. Lithos, 8, 255-268. RASMUSSEN,J. & NOE-NYGAARD,A. 1969. Beskrivelse til geologisk kort over F~eroerne. Danmarks Geologiske Undersogelse, I Rtekke, 24. RIDD, M. F. 1981. Petroleum geology west of the Shetlands. In: ILLINa, L. V. & HOBSON, G. D. (eds) Petroleum Geology of the Continental Shelf of North-west Europe. Heyden, London, 414-425. ROBERTS, D. G., BOTT, M. H. P. & URUSrd, C. 1983. Structure and origin of the Wyville--Thomson Ridge. In: BOTT, M. H. P., SAXOV, S., TALWANI, i . & THIEDE, J. (eds) Structure and Development of the Greenland-Scotland Ridge. Plenum Press, New York, 133-158. SCHt)NHARTING, G. & ABRAHAMSEN,N. 1989. Paleomagnetism of the volcanic sequence in Hole 642E, ODP Leg 104, Voting Plateau, and correlation with Early Tertiary basalts in the North Atlantic. In: ELDHOLM, O., THIEDE, J., TAYLOR, E., ETAL. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 104, 911-920. SMYTrIE, D. K. 1983. Faeroe--Shetland escarpment and continental margin north of the Faeroes. In: BOTT, M. H. P., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) Structure and Development of the Greenland-Scotland Ridge. Plenum Press, New York, 109-119. SOPER, N. J. & COSTA, L. I. 1976. Palynological evidence for the age of Tertiary basalts and postbasaltic sediments at Kap Dalton, central East Greenland. Rapport, Grenlands Geologiske Undersogelse, 80, 123-127. --, DOWNIE, C., HIGGINS, A. C. & COSTA, L. I.

197

1976a. Biostratigraphic ages of Tertiary basalts on the East Greenland continental margin and their relationship to plate separation in the Northeast Atlantic. Earth and Planetary Science Letters, 32, 149-157. --, HIGGINS, A. C., DOWNIE, C., MATTHEWS, D. W. & BROWN, P. E. 1976b. Late Cretaceousearly Tertiary stratigraphy of the Kangerdlugssuaq area, east Greenland, and the age of opening of the north-east Atlantic. Journal of the Geological Society, London, 132, 85-104. STRIDE, A. H., BELDERSON, R. H., CURRAY, J. R. & MOORE, D. G. 1967. Geophysical evidence on the origin of the Faeroe Bank Channel - I. Continuous reflection profiles. Deep-Sea Research, 14, 1-6. T ARLING,D . H . , H A I L W O O D , E . A . & LOVLIE,R. 1988. A palaeomagnetic study of lower Tertiary lavas in E Greenland and comparison with other lower Tertiary observations in northern Atlantic. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 215-224. VIERECK, L. G., HERTOGEN, J., PARSON, L. M., MORTON, A. C., LOVE, D. & GIBSON, I. L. 1989. Chemical stratigraphy and petrology of the Voring Plateau tholeiitic lavas and interlayered volcaniclastic sediments at ODP Hole 642E. In: ELDHOLM, O., THIEDE, J., TAYLOR, E., ET AL. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 104, 367-396. , TAYLOR, P. N., PARSON, L. M., MORTON, A. C., HERTOGEN, J., GIBSON, I. L. & the ODP Leg 104 Scientific Party 1988. Origin of the Palaeogene Voring Plateau volcanic sequence. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 69-83. WAAGSTEIN, R. 1977. The geology of the Faeroe Plateau. PhD thesis, Kobenhavns Universitet. - 1988. Structure, composition and age of the Faeroe basalt plateau. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 225-238. - & HALD, N. 1984. Structure and petrography of a 660m lava sequence from the Vestmanna-1 drill hole, lower and middle basalt series, Faeroe Islands. In: BERTHELSEN, O., NOE-NYGAARD, A. & RASMUSSEN, J. (eds) The deep drilling project 1980-1981 in the Faeroe Islands. Foroya Frodskaparfelag, T6rshavn, 39-65. - & RASMUSSEN, J. 1975. Glacial erratics from the sea floor south-east of the Faeroe Islands and the limit of glaciation. Frodskaparrit, T6rshavn, 23, 101-119. WHITE, R. S., SPENCE, G. D., FOWLER, S. R., MCKENZIE, D. P., WESTBROOK, G. K. & BOWEN, A. N. 1987. Magmatism at rifted continental margins. Nature, 330, 439-444.

Evolution of a major oceanographic pathway: the equatorial atlantic E. J.,W. J O N E S , 1 S. C. C A N D E 2 & F. S P A T H O P O U L O S

1

1 Department of Geological Sciences, University College London, Gower Street, London WCIE 6BT, UK 2 Lamont-Doherty Earth Observatory, Palisades, New York 10964, USA

Abstract: The history of continental separation in the equatorial Atlantic is important to our understanding of the events which have led to the establishment of the present patterns of water circulation. Orientations of oceanic basement-lineaments determined from bathymetric, seismic, magnetic and satellite altimetry data, and the distribution of seismic reflectors in deep-water sediments indicate that during its early opening stages the Atlantic was bounded to the south by the Guinea Fracture Zone. Using stage poles obtained from South Atlantic spreading patterns, basement ages and the palaeobathymetry of the equatorial region have been derived. The proximity of magnetic anomaly M0 to the present continental slopes suggests that the deep-water basins began to form in the Aptian. During the early stages of basin development water circulation was greatly restricted by fracture zone ridges, leading to the formation of thick sequences of carbonaceous shales. Outflow of dense, saline water from the equatorial basins may have been an important factor in controlling deposition along the Atlantic margins, contributing to the development of unconformities within the Cretaceous sedimentary record. By Santonian time the equatorial rift had reached a width of c. 1200km, water depths close to the continental margins exceeded 5000 m and the transfer of surface water between the North and South Atlantic was well established. The conjugate Sierra Leone and Ceara Rises were built up during the late Cretaceous and existed as separate features by Early Oligocene time. Both the Romanche and Vema Fracture Zones have acted as important conduits for the transfer of bottom-water from the western to the eastern equatorial basins, with seismic profiles providing evidence for vigorous bottom water flow during the Eocene and later Tertiary. In the Sierra Leone Basin the circulation of bottom water may have reversed during the late Tertiary as a result of the movement of the eastern portion of the Romanche Fracture Zone north of the equator.

The equatorial region of the Atlantic from the latitude of the Cape Verde Islands to 10~ forms a major route for the transfer of some of the principal water masses of the oceans between the northern and southern hemisphere (Fig. 1). It is through this wide gap between South America and Africa that N o r t h Atlantic Deep Water moves southwards and Antarctic Intermediate Water and Antarctic Bottom Water pass into the N o r t h Atlantic (Warren 1981). There is also a complex interchange of wind-driven surface waters through the equatorial current system (Pickard & Emery 1990). The rates at which surface, intermediate and near-bottom waters are transferred across the equatorial divide are, of course, closely governed by bottom topography within the oceanic rift. Since the South Atlantic did not begin to open until the early Cretaceous, when an extensive seaway already existed to the north (Rabinowitz & LaBrecque 1979), it is clear that the circulation pattern of

the Atlantic must have undergone profound changes over the past 140 Ma. Several reconstructions of the opening in the transitional region between the N o r t h and South Atlantic have been proposed. In an early paper Le Pichon & Fox (1971) suggested that the late Jurassic Atlantic seaway extended as far south as the present equator, a marine connection between the North and South Atlantic developing through the G u l f of Guinea in the Lower Cretaceous. Other studies have indicated that the Jurassic North Atlantic terminated at the Guinea Fracture Zone c. 1000 km further north, the Sierra Leone Basin in the east and the Ceara and Demerara Basins in the west forming during the Lower Cretaceous (Mascle et al. 1986; Jones 1987). Such models are broadly consistent with palaeontological evidence from the continental margins and the deep-water areas, which indicate that seaways began to develop between South America and Africa in the early Cretac-

From Scrutton, R. A., Stoker, M. S., Shimmield, G. B. & Tudhope, A. W. (eds), 1995, The Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region, Geological Society Special Publication No. 90, pp. 199-213

199

200

E.J.W. JONES E T A L .

Fig. 1. Bathymetry of the equatorial Atlantic [isobaths in metres; simplified from Canadian Hydrographic Survey, (1984)]. Oceanic fracture zones are labelled as follows: F, Fifteen Twenty North; V, Vema; G, Guinea; D, Doldrums; SL, Sierra Leone; ST, Strakhov; SP, St Patti; R, Romanche; C, Chain; A, Ascension. Other features are: CV, Cape Verde Islands; K, Kane Gap; GP, Guinea Plateau; SLR, Sierra Leone Rise; ICR, Ivory Coast Rise; DP, Demerara Plateau; AC, Amazon Cone; CR, Ceara Rise. Locations of reflection profiles discussed in the text are labelled a-i. Selected deep-sea drilling sites are also shown. Arrows indicate the present flow paths of Antarctic Bottom Water.

Fig. 2. Trends of basement lineaments derived from bathymetric, magnetic, seismic and radar altimetry data. The flow lines are based on stage poles derived from South Atlantic spreading patterns (Cande et al. 1988). For annotations of bottom features see Fig. 1.

eous, with strong N - S oceanic connections being established during the late Cretaceous (Reyment & Tait 1972; Kennedy & Cooper 1975; Premoli Silva & Boersma 1977). Unlike most of the Atlantic area, details of the opening history near the equator have not been

unravelled by dating oceanic magnetic anomalies. Except near fracture zones, anomalies at these low magnetic latitudes are generally of small amplitude and cannot be reliably tied to the geomagnetic timescale. However, in recent years, our knowledge of the surface morphology

EVOLUTION OF EQUATORIAL ATLANTIC of the equatorial basement has greatly improved as a result of satellite altimetry and a more extensive bathymetric coverage through the use of swath bathymetry and GLORIA, as well as the conventional echo-sounder. Furthermore, surface-ship gravity and magnetic data have been used to trace basement features beneath regions covered by thick sediments. This new information on the structural fabric forms an important constraint on models of Atlantic development. In this paper we show that the initial opening of the equatorial region south of the Guinea Fracture Zone is closely related to the early development of the South Atlantic. We present reconstructions of the opening based on stage poles derived from fracture zone trends and well-dated magnetic lineations south of 10~ (Cande et al. 1988). We demonstrate that the flow lines deduced from the pattern of opening in the South Atlantic are largely consistent with recently mapped basement lineaments near the Equator and with a tentative seismic stratigraphy inferred from deep-sea drilling results. Using basement ages derived from the South Atlantic opening we determine th e palaeobathymetry of the main equatorial basins. Finally, we discuss the implications of the basin configurations in relation to the palaeoceanography of the Atlantic region.

Bathymetry and basement trends The equatorial Atlantic is structurally one of the most complex regions in the oceans, being dominated by closely-spaced transform faults that cumulatively offset the crest of the MidAtlantic Ridge by some 3800 km in a left lateral sense between 16~ and 3~ (Fig. 1; Gorini 1981). The largest offset occurs across the 840km long Romanche Transform (Heezen et al. 1964; Belderson et al. 1984; Searle et al. 1994). Other fracture zones indicated in Fig. 1 include the Vema ( l l ~ the Strakhov (also known as the Four North), St Paul, Chain and Ascension. Depths in the main fracture zone valleys exceed 4000 m, reaching > 7000 m in the Romanche Fracture Zone. Many portions of the flanking transverse ridges are shallower than 1500m. Several fracture zones, which include the St Paul, Romanche and Chain, can be traced to the continental margins where they are associated with changes in shelf width and the edges of deep sedimentary basins. Two prominent marginal platforms, the Guinea Plateau off

201

Africa and the Demerara Plateau off Brazil, are bounded by fracture zones. Between the Mid-Atlantic Ridge and the continental margins lie a number of abyssal plains; the Gambia, Sierra Leone and Guinea in the east and the Demerara, Ceara and Pernambuco Abyssal Plains in the west. The region also contains two aseismic elevations - the Sierra Leone Rise and the Ceara Rise. Kumar & Embley (1977) have argued that these are conjugate volcanic features formed during a period of extensive igneous activity at a time of plate reorganization in the late Creta-ceous. Drilling has not yet reached the basement of the Sierra Leone Rise. At site 354 on the Ceara Rise (Fig. 1) sediments of early Maastrichtian age resting on basaltic basement have been recovered (Supko et al. 1977). The Sierra Leone Rise terminates at the Guinea Fracture Zone and is separated from the African margin by a deep channel known as Kane Gap (K; Fig. 1). Along its western and northern sides the Ceara Rise has been partly inundated by the sediments of the Amazon Cone. The deepest route for bottom water flowing northwards presently lies between the Rise and the foothills of the Mid-Atlantic Ridge. The orientations of linear basement features derived from bathymetric compilations, satellite radar altimetry, magnetic and seismic data collected between 22~ and 10~ are shown in Fig. 2. The pattern of basement lineaments associated with the opening of the South Atlantic can be followed for more than 1000 km north of the Equator. Lineaments run south of west near the Sierra Leone Rise and can be traced to the Guinea Fracture Zone at c. 10~ Further north they are oriented 1020 ~ north of west. The change in bathymetric trends indicates that the Guinea Fracture Zone separates regions which opened about different poles of rotation (Jones 1987). Also shown in Fig. 2 are flow lines deduced from stage poles derived from South Atlantic fracture zone trends and magnetic anomalies 34 and younger by Cande et al. (1988) and a pole at 53.47 ~ N, 34.18 ~ W to describe the motion between chron M0 and chron 34, which is little different from that used by Pindell et al. (1988) (52.08~ 34.03~ to obtain closure in the equatorial region. South of the Guinea Fracture Zone there is a close correspondence between the flow lines and basement lineaments. Further north discrepancies in trend become marked, as can be seen near 30~ W. The Vema Fracture Zone at l l ~ (V; Fig. 1) is clearly associated with the opening of the South Atlantic.

202

E.J.W. JONES E T A L .

Deep-water seismic stratigraphy and the opening of the equatorial region The importance of the Guinea Fracture Zone as a plate boundary during the early opening of the Atlantic is evident from the depositional pattern 9 f late Mesozoic sediments as recorded on seismic reflection profiles. North of the Guinea Fracture Zone four reflectors (labelled 1-4 in Fig. 3) have been traced over a wide area of the Gambia Basin. The deepest reflector (4) lies at the level where basaltic rocks were penetrated at drilling site 367 south of the Cape Verde Islands (Fig. 1; Lancelot et al. 1977b). Reflector 3 forms the upper boundary of a sequence consisting of marls, limestones and chert which are mostly O x f o r d i a n / K i m m e r i d g i a n - N e o c o m i a n in age but appear to include an early Aptian sequence. Reflector 2 corresponds to the top of a thick formation of late Aptian-early Turonian black shales. Reflector 1 correlates with a zone of Eocene cherts. Reflector 1 can be traced over a wide area of the equatorial Atlantic (Fig. 3). Eocene cherts at the level of this reflector have been confirmed by drilling at site 660 just north of Kane Gap (Ruddiman et al. 1988) and at site 13 in the Sierra Leone Basin (Maxwell et al. 1970) (Fig. 1). The seismic evidence suggests that Cretaceous black shale sequences are extensively developed in the eastern and western basins of the equatorial region. Reflector 2 can be traced into the Sierra Leone and Guinea Basins and is tentatively identified in the Demerara and Ceara Basins. The presence of Cretaceous black shales in the latter region is indicated by drilling results at site 144 on the northern side of the Demerara Plateau where S e n o n i a n - T u r o n i a n carbonaceous shales deposited in a poorly oxygenated environment were recovered (Fig. 1; Hayes et al. 1972). Carbonaceous, pyritic shales of Campanian age have been sampled at site 24 near the foot of the Brazilian continental rise (Fig. 1; Maxwell et al. 1970). Reflector 3 is not recorded south of the Guinea Fracture Zone, implying that the equatorial basins did not begin to develop until Fig. 3. Seismic reflection profiles correlated with drilling results at DSDP Site 367 in the Gambia Basin. Two-way reflection time below the sea floor is given in seconds. Locations of the profiles are indicated in Fig. 1 as follows: DSDP Site 367 (Lancelot et al. 1977b). (a) Southern Gambia Basin (Jones 1987); ~) West of Sierra Leone Rise (Jones 1987); (e) Sierra Leone Basin (Jones 1987); (d) Guinea Basin (Delteil et al. 1974); (e) Demerara Abyssal Plain (Neprochnov et al. 1977).

EVOLUTION OF EQUATORIAL ATLANTIC

I

i

[MO ~-~"~--~ ~ . '

I

~,~.~

2~

I

=-'~"~'~'~" ~ - - ' ~ - - ~

I

'~-----~ ~

/

/

,~.-/'~

A'O(3

~( J

7"

203

I

I

0 ~

km --,-

20~

1000 -

10 ~

|

_

A

m

e

r

i

c

34

a

~w

'

4b ~

5

~

1

oo

13

o~

lo~

Fig. 4. Synthetic flow lines and positions of magnetic anomalies 5, 13, 21, 25, 31, 34 and M0 deduced from stage poles computed from South Atlantic fracture zone trends and magnetic anomalies. GFZ, Guinea Fracture Zone; SLR, Sierra Leone Rise; CR, Ceara Rise. Aptian time at the earliest (Jones 1987). An Aptian opening would account for the magnetically quiet zone between the transforms in the Sierra Leone Basin, since seafloor spreading would have occurred during a period lacking geomagnetic reversals. Reflector 3 is also not recorded beneath the Demerara, Ceara and Pernambuco Abyssal Plains off South America. The equatorial basins are thus younger than those of the South Atlantic where the Valanginian magnetic anomaly M l l (c. 134Ma) has been identified (Rabinowitz & La Brecque 1979). If South America and Africa had each acted as single plates during the early Cretaceous opening then the pole of rotation based on the trends of the Falkland and Agulhas Fracture Zones would lie close to 3~ S, 45~ a position that requires considerable crustal shortening in the equatorial region. Apart from local regions of transpression near the intersections of the fracture zones with the continental margins, the equatorial area appears to have been largely under extension at this time (Machens 1973; Campos et al. 1974). Widespread normal faulting is seen on reflection profiles (Delteil et al. 1974; Mascle et al. 1986). The lack of major compressional structures indicates the development of short-lived plate boundaries within Africa and South America during the early Cretaceous. These probably include the Benue Rift (Burke & Dewey 1974) and the Tiete and Piquiri lineaments which extend through the Parana Basin of South America (Unternehr et

al. 1988; Ntirnberg & Miiller 1991). The close

correspondence between the deep-water equatorial basement trends and the flow lines associated with the South Atlantic spreading (Fig. 2) indicates that the major movements along these continental lineaments had ceased by late Aptian time.

Equatorial Atlantic reconstructions and palaeobathymetry In view of the close correspondence between the orientation of basement lineaments and the synthetic flow lines in Fig. 2, we have computed a series of reconstructions of the equatorial Atlantic based on stage poles derived from fracture zone directions and dated magnetic anomalies in the South Atlantic. Figure 4 shows the positions of anomalies 5, 13, 21, 25, 31, 34 and M0. Anomaly M0 is situated within 40 km of the ocean--continent boundary off Sierra Leone and the Guinea Plateau as determined from seismic, gravity and magnetic data (Jones & Mgbatogu 1982), suggesting that seafloor spreading began at about Chron M0. The anomaly tracks along the eastern portion of the Guinea Fracture Zone and also lies close to the edge of the Demerara Plateau. M0 is located some 600km landward of the foot of both the Niger and Amazon cones which have been largely built up during the late Tertiary (Machens 1973; Damuth & Kumar 1975). The

204

E.J.W. JONES E T AL.

!

1000km

J

\

AF

SA

E.Apt. Fig. 5. Pre-drift reconstruction of Africa (AF) and South America (SA) in the equatorial region (Early Aptian: pre-Chron M 0 - 119Ma). Approximate water depths in the Atlantic to the north of the join of the Guinea Plateau (GP) and Demerara Plateau (DP) are shown by isobaths in kilometres. present deep-water basins appear to have been formed between Chron M0 and Chron 31. Anomaly 25 lies on the seaward margins of the Sierra Leone and Ceara Rises, which is consistent with the late Cretaceous age suggested by Kumar & Embley (1977) for these features. A reconstruction of the region for the early Aptian is shown in Fig. 5. The Guinea and Demerara Plateaus were juxtaposed, a geometry also suggested by other studies (Mascle et al. 1986). The present deep-water embayment in the continental margin northwest of the Demerara Plateau (Fig. 1) probably formed part of the Jurassic rift. Also shown in Fig. 5 are approximate palaeodepths, determined from the subsidence curves of Sclater et al. (1977), with a correction (Crough 1983) for the thickness of the pre-Aptian sediments drilled at site 367 and recorded on surrounding seismic profiles in the Gambia Basin. Except in the region of the Cape Verde Islands, where evidence of shallow water conditions is found in the early Cretaceous sediments of Maio (Stillman et al. 1982), depths in the region floored by late Jurassic oceanic basement exceeded 5000 m.

An early Albian reconstruction is shown in Fig. 6. A narrow oceanic rift extended southwards into the Gulf of Guinea. In the north this separated the Guinea from the Demerara Plateau, and the southern Gambia Basin from the embayment northwest of the Demerara Plateau. The central equatorial region consisted of a series of shallow ( < 4000 m) oceanic basins which may have evolved quite separately because of the presence of numerous fracture zones, the transverse ridges perhaps providing sporadically emergent routes between Africa and South America. Freshwater fish and ostracod faunas indicate that there were strong nonmarine links between South America and Africa during the Albian (Kr6mmelbein 1966; Neufville 1973). By Santonian time the oceanic rift had reached a width of c. 1200 m, with water depths exceeding 4000m close to the continental margins of West Africa and Brazil (Fig. 7). The development of the Sierra Leone and Ceara Rises towards the end of the Cretaceous produced a large elevated area in the central equatorial region. Faunal studies of sediments

EVOLUTION OF E Q U A T O R I A L ATLANTIC

-- ,, @ ,

~ow~i~)/ i! I'~

EQI

GP

1000km I

././"

AF eJ

205

e

Fig. 6. Early Albian reconstruction of the equatorial Atlantic. Isobaths are given in kilometres. The position of the Equator is taken from the maps of Smith & Briden (1977). The arrows show the inferred movement of dense Equatorial Outflow Water (EOW) which is discussed in the text. GP, Guinea Plateau; DP, Demerara Plateau; V, Vema Fracture Zone; R, Romanche Fracture Zone.

;/J

|

--

s~~

~

,

\

\

~

~

~

.

,

,

~

~.,,,,,~.-,--~,~

' / ~

Fig. 7. Santonian reconstruction of the equatorial Atlantic. For notations, see caption to Fig. 6.

,

206

E . J . W . JONES E T AL.

/// (~ ",~,,,

~/f~

i

1000km

j

.....71

I

, ~k

I~IIIIi,

',1,1,,,

III1tll

Ill

i,,,,,,~,1~\,1,,, i,

i

1111111

itll i

~

~.~

..........

,,,,~ .... , , , , , ~ . . . . . . . ~ ~ . . .

sA

,,,,,,

,I~ ,J/t

,f

.... %1.

...........

\~

lit/

.............. ......... " .....

L. Pal. Fig. 8. Late Palaeocene reconstruction of the equatorial Atlantic. SLR, Sierra Leone Rise; CR, Ceara Rise. For other notations, see caption to Fig. 6.

~

i ~i,';"'~f'~ 'f'~' ' ' ~ ' ',,: .......

\ \ 5 .... " . . . .

(~

' .......... 'i~..... ,;.

//~J/

~

i

1000km

I

I([ \

,......... ~...................... ,,,~

,

~

~

AF

////................................... Fig. 9. Early Oligocene reconstruction of the equatorial Atlantic. SLR, Sierra Leone Rise; CR, Ceara Rise. For other notations, see caption to Fig. 6 recovered at site 354 on the C e a r a Rise indicate water depths < 1000m in the M a a s t r i c h t i a n (Supko et al. 1977). P a l a e o b a t h y m e t r y for the late Palaeocene is s h o w n in Fig. 8. Drilling at sites 354 and 366 (Fig. 1) has shown that sedimentation on both features t o o k place in

a well-oxygenated e n v i r o n m e n t at this time (Supko et al. 1977; Lancelot et al. 1977a). S h a l l o w - w a t e r corals a n d associated f a u n a s dredged f r o m the Sierra L e o n e Rise reveal that parts of the feature r e m a i n e d close to sea level until the late Eocene, with large areas of the Rise

EVOLUTION OF EOUATORIAL ATLANTIC

Fig. 10. Reflection profile recorded over the continental rise in the Gambia Basin at position f in Fig. 1; 1 s of two-way reflection time (twt) is indicated. Reflectors 1 and 2 can be traced to DSDP 367 where they have been shown to be of Eocene and Cretaceous age, respectively (see Fig. 3). The undulations in Reflector 2 may be due to the movement of dense, saline water flowing northwards from constricted basins close to the Equator. lying a t depths < 3000m (Jones & Goddard 1979). By early Oligocene time the Sierra Leone and Ceara Rises existed as separate features and the equatorial seaway had expanded to 2000 km (Fig. 9). Wide, deep-water basins were present, although there is evidence from dredge samples that many parts of the fracture zone ridges were close to sea level (Bonatti 1978; Udintsev et al. 1990.)

Discussion

Palaeoceanography and the early opening of the equatorial Atlantic During the early separation of South America from equatorial Africa the deep-water areas

207

were confined to a system of narrow basins bounded by the transverse ridges of large-offset fracture zones such as the Romanche (Fig. 6). The occurrence of T u r o n i a n carbonaceous shales at sites 24 and 144 off South America (Fig. 1), and our tentative identification of black shale sequences on seismic reflection profiles (Fig. 3), suggest that deposits laid down under poorly oxygenated conditions are widespread in the equatorial deep-water basins. The many fracture zone ridges probably played an important role in restricting water transfer between these early basins. It was only after the Turonian that strong marine connections between the North and South Atlantic were established, with the influx of Tethyan planktonic foraminiferal faunas south of the equator recorded in the late Cretaceous sections at several deep-sea drilling sites (Premoli Silva & Boersma 1977). The geometry of reflectors within the Cretaceous sequences of the African continental margin north of the Guinea Fracture Zone suggests that sedimentation was governed in part by bottom currents. One example of current activity is shown in Fig. 10 where the Cretaceous Reflector 2 beneath the continental rise off Senegal is distorted by giant ripples which are similar in character to those recorded on well known sedimentary drift deposits, such as the Blake-Bahama Outer Ridge in the Western Atlantic. As there was no direct route through the Atlantic to high latitudes at this time (Smith & Briden 1977) it is unlikely that such current activity is related to dense polar water. We suggest that a potential source for the bottom water is the region of narrow oceanic basins between West Africa and Brazil (Figs 6 & 7). Being near the palaeo-equator, conditions were favourable for the formation of hot, highly saline water in areas restricted by shallow fracture zone sills. Outflow into the main Atlantic basin would tend to flow northwards as a deep geostrophic current along the continental margin of Africa. In Fig. 6 this is referred to as the 'Equatorial Outflow Water'. Close to southern Iberia the equatorial outflow may have merged with a saline overflow from Tethys to circulate northwards and then westwards into the western Atlantic. The effects of this density-driven current on sedimentation north of the Guinea Fracture Zone would perhaps have been similar to those resulting from the recent outflow of Mediterranean water, which has caused erosion and influenced deposition rates and sedimentary bedforms to the west and north of the Straits of Gibraltar (Heezen & Johnson 1969). Coming at the end of a period of deposition of black shales, the development

208

E.J.W. JONES ET AL.

Fig. 11. Reflection profile recorded across the southern margin of the Gambia Basin (i in Fig. 1). Total reflection time in seconds is shown. Reflector 1 is of Eocene age (see Fig. 3). Variations in the thickness of the sediments above Reflector 1 are probably due to non-uniform deposition and erosion in the path of Antarctic Bottom Water which flows eastwards after emerging from the Vema Fracture Zone (Fig. 1). AP: Abyssal Plain. of giant ripples on Reflector 2 may record the initiation of vigorous ocean circulation following a long period of bottom water stagnation. Equatorial outflow water may also have reached the Cape and Argentine Basins in the southern South Atlantic (McCoy & Zimmerman 1977). Furthermore, the movement of equatorial outflow water may have been sufficiently intense to erode parts of the continental slope. Unconformities within the deep-water Cretaceous sections off South America and Africa have been reported by Supko & Perch-Nielsen (1977) and Von Rad et al. (1982). Tertiary palaeoceanography a n d the m o v e m e n t o f A n t a r c t i c B o t t o m Water

At the present time Antarctic Bottom Water (AABW) enters the equatorial region along the eastern side of South America (Fig. 1). Nearbottom potential temperatures indicate that a component of AABW flows into the Romanche Fracture Zone, the remainder moving north-

wards over the Ceara Abyssal Plain and to the east of the Ceara Rise and Amazon Cone (Metcalf et al. 1964; Worthington & Wright 1970). Further north, part of the AABW passes into the Vema Fracture Zone and part flows along the flank of the Mid-Atlantic Ridge towards the main basins of the Western Atlantic (Fig. 1; McCartney et al. 1991). The Romanche component is deflected southwards into the Gulf of Guinea across a deep saddle in the ridge topography of the St Paul Fracture Zone near 14~ although Heezen et al. (1964) indicate that part moves into the Sierra Leone Basin where it continues northwards along the continental margin of West Africa. A photograph taken during sediment coring reveals bottom current activity in Kane Gap (Hobart et al. 1975), but recent oceanographic measurements point to little net transfer of bottom water between the Sierra Leone and Gambia Basins (McCartney et al. 1991), so AABW may turn southwards along the flank of the Sierra Leone Rise. The Vema component

EVOLUTION OF EQUATORIAL ATLANTIC

209

Fig. 1~. ReflectiOn profile recorded in the Sierra Leone Basin at position g in Fig. 1; 0.5 s twt is shown. Reflector 1 is of E6~ene age (see Fig. 3). The undulations on this reflector are probably due to deposition in the path of Antarctic BottOm Water, of AABW has been documented from direct current measurements and temperature-salinity data. Easter|y bottom currents in excess of 10cms -1 have been measured in the valley of the Vema Fracture Zone (Vangriesheim 1980; Eittreim et al. 1983). McCartney et al. (1991) have shown that AABW continues to flow eastwards along the southern edge of the Gambia Basin and then no~hwards off the Senega! margin before passing around the Cape Verde Rise. There is abundant evidence from sedimentary b e d f o ~ S that AABW has influenced recent sedimentation. Kumar & Embley (1977) have described areas of large sediment waves in the region where the AABW passes east of the Ceara Rise towards the Vema Fracture Zone. Evidence of recent current activity in the eastern Gambia

Basin has been reported by Ruddiman et al. (1988). Jacobi & Hayes (1982) and Rossi et al. (1992) have mapped giant sediment waves in the Sierra Leone and Gambia Basins. It is clear from the configuration of the widespread Eocene reflector (1; Fig. 3) and reflectors within the overlying sediments that such bottom current activity is not confined to the recent past. A large drift structure in the path of the AABW at the southern margin of the Gambia Basin is shown in Fig. 11. At this location there are marked variations in the thickness of the sediments above reflector 1, in a succession which has been uplifted above the abyssal plain turbidites by faulting. The wavelike form of Reflector 1 and small drift structures in the sediments above (Figs 12 & 13, respectively) indicate significant Eocene and

210

E.J.W. JONES E T AL.

Fig. 13. Parts of a short reflection profile recorded over the northwestern part of the Sierra Leone Abyssal Plain (h in Fig. 1); 0.5 s twt is indicated. Reflector 1 is of Eocene age (see Fig. 3). The undulations within the sedimentary section above Reflector 1 are probably due to irregular deposition and perhaps erosion in the path of Antarctic Bottom Water. Reflector 4 is the top of the seismic basement.

post-Eocene b o t t o m current activity in the Sierra Leone Basin. Sarnthein and Faug6res (1993) have recorded sediment waves on Reflector 1 in the southeastern Gambia Basin. There is also strong evidence to demonstrate that the present low net transport of bottom water through Kane Gap between the Sierra Leone Rise and the African margin has not persisted through the Tertiary. The seismic profile in Fig. 14 reveals that the thick sediments normally recorded above the Eocene reflector (1) are absent across much of the deep channel, with the strongly reflective Eocene chert section either

exposed or lying within a few metres of the seafloor. The paucity of post-Eocene sediments may be a result of the vigorous southward flow of AABW which entered the Sierra Leone Basin from the Gambia Basin, after first passing through the Vema Fracture Zone, or the northward m o v e m e n t of A A B W which flowed t h r o u g h the Sierra Leone Basin from the Romanche Fracture Zone. The Ivory Coast Rise in the southern part of the Sierra Leone Basin (Fig. 1) is underlain by thick sediments which appear to have been transported to the area by southward-flowing

EVOLUTION OF EQUATORIAL ATLANTIC

211

Fig. 14. Reflection profile across Kane Gap between the Sierra Leone Rise and the Guinea Plateau (K in Fig. 1). Total reflection time in seconds is shown. The relatively transparent section labelled A can be correlated with sediments deposited above the Eocene Reflector 1 in the Gambia and Sierra Leone Basins. The lack of this sediment cover near B indicates strong bottom water flow through this region during post-Eocene time.

Fig. 15. Inferred movement of Antarctic bottom water through the equatorial Atlantic during the Early Oligocene. The position of the Equator is taken from Smith & Briden (1977). SLR, Sierra Leone Rise; CR, Ceara Rise. For other notations see Fig. 6.

bottom currents (Emery et al. 1975; Jacobi & Hayes 1982). One possible source of sediments is the Guinea Plateau, which was extensively eroded in the early Tertiary (Mascle et al. 1986), p r o b a b l y as a result of a greater interchange of surface water between the North and South Atlantic. However, the present bottom water appears to flow northwards along the Liberia and Sierra Leone margins (Fig. 1). An explanation for the reversal of flow can be found by considering the relation between the

palaeo-equator and the exit points of A A B W from the Vema and Romanche Fracture Zones. During the Early Oligocene the Vema Fracture Zone was still situated north of the equator so any component of A A B W reaching the transform valley would have followed a course similar to that at the present time (Fig. 15). On emerging from the main fracture zone, the core of A A B W would be deflected to the right and continue into the Gambia Basin. The Romanche Transform, on the other hand, lay some 1100 km

E . J . W . JONES ET AL.

212

south o f the equator. A A B W passing from the eastern end o f the fracture zone w o u l d be deflected to the left to f o r m a clockwise gyre in the Sierra Leone Basin. Since w a t e r depths in the vicinity o f K a n e G a p were c. 800 m less than at present (Jones & G o d d a r d 1979) little cold b o t t o m w a t e r w o u l d have p a s s e d into the G a m b i a Basin f r o m the south. T h e present flow p a t t e r n of A A B W w o u l d p r o b a b l y not have been initiated until late M i o c e n e - P l i o c e n e time w h e n the eastern end of the R o m a n c h e Transf o r m passed n o r t h o f the E q u a t o r . This region m a y be one of the few areas o f the o c e a n where a reversal of b o t t o m - w a t e r flow o c c u r r e d as a result o f large-scale plate movements. Such a reversal m i g h t be d e t e c t a b l e by e x a m i n i n g palaeo-current directions in the N e o g e n e section by drilling near the eastern end of the R o m a n c h e F r a c t u r e Zone. We thank the many oceanographic groups in Europe and the USA who kindly made available the geophysical data used in this study. We are also indebted to J. Callomon, D. T. Donovan, W. Haxby, J. Mascle, N. Morton, P. F. Rawson, E. Robinson and M. Sarnthein for their help. Financial support for the work was provided by the Natural Environment Research Council (Grant GR3/4674).

References BELDERSON, R. n., JONES, E. J. W., GORINI, M. A. & KENYON, N. H. 1984. A long-range side-scan sonar (GLORIA) survey of the Romanche active transform in the equatorial Atlantic. Marine Geology, 56, 65-78. BONATTI, E. 1978. Vertical tectonism in oceanic fracture zones. Earth and Planetary Science Letters, 37, 369-379. BURKE, K. & DEWEY, J. F. 1974. Two plates in Africa during the Cretaceous? Nature, 249, 313-316. CAMPOS, C. W. M., PONTE, F. C. & MIURA, K. 1974. Geology of the Brazilian continental margin. In: BURK, C. A. & DRAKE, C. L. (eds) The Geology of Continental Margins, Springer-Verlag, Berlin, 447-461. CANADIAN HYDROGRAPHIC SERVICE 1984. General Bathymetric Chart of the Oceans. Sheets 5.08 and 5.12 (scale 1:10 million). CANDE, S. C., LA BRECQUE, J. L. & HAXBY, W. F. 1988. Hate kinematics of the South Atlantic: Chron C34 to Present. Journal of Geophysical Research, 93, 13 479-13 492. CROUGH, S. T. 1983. The correction for sediment loading on the sea floor. Journal of Geophysical Research, 88, 6449-6454. DAMUTH, J. E. & KUMAR, N. 1975. Amazon Cone: Morphology, sediments, age, and growth pattern.

Bulletin of the Geological Society of America, 86, 863-878. DELTEIL, J.-R., VALERY, P., MONTADERT, L., FON-

DEUR, C., PATRIAT, P. & MASCLE, J. 1974, Continental margin in the northern part of the Gulf of Guinea. In: BURK, C. A & DRAKE, C. L. (eds) The Geology of Continental Margins. Springer-Verlag, Berlin, 297-3! !, EIITREIM, S. L., BISCAYE,P. ~. & JACOBS, S. S. 1983. Bottom-water observations in the Vema Fracture Zone. Journal of Geophysical Research, 88, 26092614. EMERY, K. O., UCHUPI, E., PHILLIPS, J. D., BOWIN, C. O. & MASCLE, J. R. 1975. Continenta! margin off Western Africa: Angola to Sierra Leone.

Bulletin of the American Association of Petroleum Geologists, 59, 2209-2265. GORINI, M. A. 1981. The tectonic fabric of the equatorial Atlantic and adjoining continent~! margins: Gulf of Guinea to Northeastern Brazil, In: Serie Projeto Remac, 9, Petrobras, ~ o d e Janeiro, 11-117. HAYES, D. E., PIMM, A. C., BENSON, W. E., ET AL. 1972. Sites 143 and 144. Initial Reports of the Deep Sea Drilling Project, 14, 283-338. HEEZEN, B. C. 8/: JOHNSON, G. L. 1969. Mediterranean undercurrent and microtopography west of Gibraltar. Bulletin of the Oceanographic Institute of Monaco, 67, 1367-1393. , BUNCE, E. T., HERSEY, J. B. & THARP, M. 1964. Chain and Romanche fracture zones. DeepSea Research, 11, 11-33. HOBART, M. A., BUNCE, E. T. & SCLATER,J. G. 1975. Bottom water flow through the Kane Gap, Sierra Leone Rise, Atlantic Ocean. Journal of Geophysical Research, 80, 5083-5088. JACOBI, R. D. & HAYES, D. E. 1982. Bathymetry, microphysiography and reflectivity characteristics of the West African margin betweeen Sierra Leone and Mauritania. In: YON ~ , U., HINZ, K., SARNTHEIN, M. & SEIBOLD, E. (eds) The

Geology of the Northwest African Continental Margin. Springer-Verlag, Berlin, 183-212. JONES, E. J. W. 1987. Fracture zones in the equatorial Atlantic and the breakup of western Pangea. Geology, 15, 533-536. - & GODDARD, D. A. 1979. Deep-sea phosphorite of Tertiary age from Annan Seamount, eastern equatorial Atlantic. Deep-Sea Research, 26, 13631379. - & MGBATOGU, C. C. S. 1982. The structure and evolution of the West African continental margin off Guin6 Bissau, Guinre, and Sierra Leone. In: SCRUTrON, R. A. & TALWANI,M. (eds) The Ocean Floor. John Wiley, Chichester, 165-202. KENNEDY, W. J. & COOPER, M. 1975. Cretaceous ammonite distributions and the opening of the South Atlantic. Journal of the Geological Society of London, 131, 283-288. KROMMELBEIN,K. 1966. Preliminary remarks on some marine Cretaceous ostracodes from Northeastern Brazil and West Africa, Proceedings of the Second West African Micropaleontogical Colloquium, Ibadan, 119-123. KUMAR, N. & EMBLEY, R. W. 1977. Evolution and origin of Ceara Rise: An aseismic rise in the western equatorial Atlantic. Bulletin of the Geolo,

EVOLUTION OF EQUATORIAL ATLANTIC

gical Society of America, 88, 683-694. LANCELOT, Y., SEIBOLD, E., CEPEK, P., ETAL. 1977a. Site 366: Sierra Leone Rise. Initial Reports of the Deep Sea Drilling Project, 41, 21-161. , --, ET aL. 1977b. Site 367: Cape Verde Basin. Initial Reports of the Deep Sea Drilling Project, 41, 163-232. LE PICHON, X. & Fox, P. J. 1971. Marginal offsets, fracture zones, and the early opening of the North Atlantic. Journal of Geophysical Research, 76, 6294-6308. MACHENS, E. 1973. The geologic history of the marginal basins along the north shore of the Gulf of Guinea. In: NAIRN, A. E. M. & STEHLI, F. G. (eds) The Ocean Basins and Margins; Vol. 1. The South Atlantic. Plenum Press, New York, 351390. MASCLE, J., MARINHO, M. O. & WANNESSON,J. 1986. The structure of the Guinean continental margin: implications for the connection between the Central and South Atlantic oceans. Geologische Rundschau, 75, 57-70. MAXWELL, A. E., VON HERZEN, R. P., ANDREWS, J. E., ET aL. 1970. Site 13. Initial Reports of the Deep Sea Drilling Project, 3, 27-70. MCCARTNEY, M. S., BENNETT, S. L. & WOODGATEJONES, M. E. 1991. Eastward flow through the Mid-Atlantic Ridge at 1I ~ and its influence on the abyss of the Eastern Basin. Journal of Physical Oceanography, 21, 1089-1121. McCoY, F. W. & ZIMMERMAN,H. B. 1977. A history of sediment lithofacies in the South Atlantic Ocean. Initial Reports of the Deep Sea Drilling Project, 39, 1047-1079. METCALF, W. G., HEEZEN, B. C. (~ STALCUP, M. C. 1964. The sill depth of the Mid-Atlantic Ridge in the equatorial region. Deep-Sea Research, 11, 1-10. NEPROCHNOV, Y. P., MERKLIN, L. R. & SUPKO, P. R. 1977. Underway geophysical measurements, Leg 39. Initial Reports of the Deep Sea Drilling Project, 39, 971-1043. NEUFVILLE, E. M. H. 1973. Upper CretaceousPalaeogene Ostracoda from the South Atlantic. Publications from the Palaeontological Institution of the University of Uppsala, Special Volume, 1. NORNBERG, D. & MI~LLER, R. D. 1991. The tectonic evolution of the South Atlantic from late Jurassic to Present. Tectonophysics, 191, 27-53. PICKARD, G. L. & EMERY, W. J. 1990. Descriptive Physical Oceanography. Pergamon Press, Oxford. PINDELL, J. L., CANDE, S. C., PITMAN, W. C. III, ROWLEY, D. B., DEWEY, J. F., LABRECQUE, J. & HAXBY, W. 1988. A plate-kinematic framework for models of Caribbean evolution. Tectonophysics, 155, 121-138. PREMOLI SILVA, I. & BOERSMA, A. 1977. Cretaceous planktonic foraminifers - DSDP Leg 39 (South Atlantic). Initial Reports of the Deep Sea Drilling Project, 39, 615-641. RABINOWlTZ, P. D. & LA BRECQUE, J. 1979. The Mesozoic South Atlantic Ocean and evolution of its continental margins. Journal of Geophysical Research, 84, 5973-6002.

213

REYMENT, R. A. & TAIT, E. A. 1972. Biostratigraphical dating of the early history of the South Atlantic Ocean. Philosophical Transactions of the Royal Sociey of London, B264, 55-95. RossI, S., WESTALL, F. & MASCLE, J. 1992. The geomorphology of the Southwest Guinea Margin: Tectonic, volcanic, mass movement and bottom current influences. Marine Geology, 105, 225-240. RUDDIMAN, W., SARNTHEIN, M. & SHIPBOARDPARTY 1988. Proceedings of the Ocean Drilling Program: Part A, 108 (Sections 1 & 2). SARWrHEIN, M. & FAUGERES, J. C. 1993. Radiolarian contourites record Eocene AABW circulation in the equatorial East Atlantic. Sedimentary Geology, 82, 145-155. SCLATER, J. G., HELLINGER, S. & TAPSCOTr, C. 1977. The paleobathymetry of the Atlantic Ocean from the Jurassic to the Present. Journal of Geology, 85, 509-552. SEARLE, R. C., THOMAS, M. V. & JONES, E. J. W. 1994 Morphology and tectonics of the Romanche Transform and its environs. Marine Geophysical Researches. (in press). SMITH, A. G. & BRIDEN, J. C. 1977. Mesozoic and Cenozoic Paleocontinental Maps. Cambridge University Press, Cambridge. STILLMAN, C. J., FURNES, n., LEBAS, M. J., ROBERTSON, A. H. F. • ZIELONKA, J. 1982. The geological history of MaiD, Cape Verde Islands. Journal of the Geological Society, 139, 347-361. SUPKO, P. R. & PERCH-NIELSEN, K. 1977. General synthesis of central and South Atlantic drilling results, Leg 39. Initial Reports of the Deep Sea Drilling Project, 39, 1099-1131. --, ET AL. 1977. Site 354: Ceara Rise. Initial Reports of the Deep Sea Drilling Project, 39, 45-99. UDINTSEV, G. B., KURENTSOVA,N. A., PRONINA, N. V., SMIRNOVA, S. B. & USHAKOVA, M. G. 1990. Finds of continental rocks and sediments of anomalous age in the equatorial segment of the Mid-Atlantic Ridge. Transactions (Doklady) of

the USSR Academy of Sciences (Earth Science Section), 312, 111-114. UNTERNEHR, P., CURIE, D., OLIVET, J. L., GOSLIN, J. & BEUZART, P. 1988 South Atlantic fits and intraplate boundaries in Africa and South America. Tectonophysics, 155, 169-179. VANGRIESHEIM,A. 1980 Antarctic Bottom Water flow through the Vema Fracture Zone. Oceanologica Acta, 3, 199-207. VON RAD, U., HINZ, K., SARNTHEIN,K. & SEIBOLD,E. 1982. The Geology of the Northwest African Continental Margin. Springer-Verlag, Berlin. WARREN, B. A. 1981. Deep water circulation in the world ocean. In: WARREN, B. A. & WUNSCH, C. (eds) Evolution of Physical Oceanography, MIT Press, Cambridge, Mass. WORTHINGTON, L. V. & WRIGHT, W. R. 1970. North

Atlantic Ocean Atlas of Potential Temperature and Salinity in the Deep Water, Including Temperature, Salinity and Oxygen Profiles from the "Erika Dan' Cruise of 1962. Vol. 2, Woods Hole Oceanographic Institution, USA.

Tertiary compression structures in the Faeroe-Rockall area MORTEN

SPARRE ANDERSEN

& L A R S OLE BOLDREEL

Geological Survey of Denmark, Thoravej 8, DK-2400 Copenhagen NV, Denmark

Three phases of Tertiary compressional deformation have been demonstrated in the FaeroeRockall Plateau (Boldreel & Andersen 1993). Compressional structures are mainly evident on the northern part of the Faeroe-Rockall Plateau and near the northern and western margins of the plateau (Fig. 1). The first deformation phase followed extrusion of the Faeroe plateau basalts and affect the oldest of the sediments above the basalts. The age of these sediments is presumed to be earliest

Eocene, and the authors believe this deformation phase followed immediately after the final break-up between the Faeroes and Greenland. The major ridges south of the Faeroe Islands (Wyville-Thomson, Ymir and Munkegrunnur) are evidence of this deformation phase. During this phase the Wyville-Thomson and Ymir Ridges formed as ramp anticlines on a northdipping fault system (Boldreel & Andersen 1993). Renewed compression in the Oligocene was located in the same area, and these authors

Fig. 1. Map of the most important Palaeocene-Miocene compressional structures on the northern part of the Faeroe-Rockall Plateau. Some Mesozoic and older structures in Scotland and on the continental shelf north and west of Scotland are also shown. Possible deformation associated with Miocene ridge-push is also indicated.

From Scrutton, R. A., Stoker, M. S., Shimmield, G. B. & Tudhope, A. W. (eds), 1995, The Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region, Geological Society Special Publication No. 90, pp. 215-216

215

216

M.S. ANDERSEN & L. O. BOLDREEL

believe that during this phase inversion of the West Lewis Basin was linked to deformation of the Wyville-Thomson and Ymir Ridges (Earle et al. 1989). The latest and most widespread compression phase occurred in the Miocene. This resulted in the NW-SE trending compression structures along the continental margin north, west and southwest of the Faeroes. Evidence of this phase is also seen in the Faeroe-Shetland Channel. It is likely that pre-existing structural elements controlled the actual location and orientation of some, if not all, of the structures we have observed. For instance, the proposed linkage between inversion of the West Lewis Basin and deformation of the Wyville-Thomson Ridge suggests that the fault system which was responsible for the formation of the WyvilleThomson and Ymir Ridges may be associated to a Mesozoic transfer zone in NW Europe (Ziegler 1990). The overall pattern of Tertiary deformation observed on the Faeroe-Rockall Plateau is apparently the consequence of roughly N-S to NW-SE compressional stresses in the area. During the first phase, the Wyville-Thomson and Ymir Ridges formed as almost pure compressional features, indicating a north to northeast stress orientation. In the Oligocene and Miocene deformation in the Faeroe-Shetland Channel, the West Lewis Basin and the Wyville-Thomson Ridge appear to be the result of simple shear along c o n j u g a t e d shear

zones. Miocene deformation near the continental margin was the result of simple compression. The overall distribution and chronology of compression structures suggests that the deformation pattern seen on the Faeroe-Rockall Plateau was the result of ridge push associated with seafloor spreading in the NE Atlantic. During the Eocene and the Oligocene the active spreading ridge north of the Faeroes was the Agir Ridge. This ridge was oblique relative to the movement between Greenland and Europe, and associated stress was responsible for deformation in the area around Wyville-Thomson Ridge. The Miocene structures at the continental margin thus were the result of ridge push from the Reykjanes Ridge.

References BOLDREEL, L. O. & ANDERSEN, M. S, 1993. Late Paleocene to Miocene compression in the Faero~Rockall area. In: PARKER,J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference, Geological Society, London, 1025-1034. EARLE, M. M., JANKOVSKI,E. J. & VANN,I. R. 1989. Structural and stratigraphic evolution of the Faeroe-Shetland Channel and Northern Rockall Trough. American Association of Petroleum Geologists Memoir, 46, 461-469. ZIEGLER, P. A. 1990. Geological Atlas of Western and Central Europe, Geological Society Publishing House, London.

Pliocene-Pleistocene radiolarian biostratigraphy and palaeoceanography of the North Atlantic S I M O N K. H A S L E T T

Faculty o f Applied Sciences, Bath College o f Higher Education, Newton Park, Newton St Loe, Bath B A 2 9BN, U K

Abstract:The Plio-Pleistocene radiolarian record of the North Atlantic differs considerably from that of the Pacific and Indian Oceans. The standard cosmopolitan low-latitude radiolarian zonal scheme can be applied to the tropical Atlantic, but a separate zonation has been developed for the endemic radiolarian faunas of the high latitude North Atlantic. A new (preliminary) radiolarian zonation for the Plio-Pleistocene is offered here for middle latitudes, consisting of five zones defined on Last Appearance Datum levels only: Cycladophora davisiana zone (0-0.5Ma), Stylatractus universus zone (0.5-1.75Ma), Antarctissa whitei zone (1.75-3Ma), Sphaeropyle langii zone (3-4.75Ma) and the Stichocorys peregrina zone (4.75Ma to Late Miocene). This new zonation, being geographically transitional, contains elements of both the previous high- and low-latitude zonations. Radiolaria also possess great potential for palaeoceanographical analysis in the North Atlantic. The lecognition of subarctic, boreal, subtropical and tropical radiolarian assemblages m~y prove useful in tracking water masses at times of climatic and oceanographic cnange. Radiolarian species characteristic of Indo-Pacific upwelling areas, e.g. Pterocanium auritum Nigrini & Caulet, Lamprocyrtis nigriniae (Caulet), Acrosphaera murrayana (Ha~ckel), Pterocorys minythorax (Nigrini) and Lithostrobus hexagonalis Haeckel, have been found in the eastern tropical North Atlantic (ODP Hole 658C), and it may be possible, through future research, to develop an Upwelling Radiolarian Index for the Atlantic. This would be capable of documenting upwelling histories and help to interpret ocean-atmosphere interaction in the eastern tropical Atlantic during Plio-Pleistocene time.

Polycystine radiolaria are marine planktonic protozoa that secrete a siliceous, opaline test. Their geological record spans the whole of Phanerozoic time, which potentially makes them one of the most biostratigraphicaUy important microfossil groups. At present, radiolaria are found throughout the world's oceans, although their greatest abundances occur in high-productivity upwelling systems such as the Peru and Californian Currents, and along the Oman Margin. Up until the Miocene-Pliocene boundary, and the closure of the Atlantic-Pacific gateway through the uplift of the Panamanian Block, radiolarian faunas in the Atlantic were very similar to those occurring in the Pacific. However, during Plio-Pleistocene time, advanced faunal provincialism has resulted in a number of differences between the Atlantic (the North Atlantic in particular) and the rest of the world's oceans. For this reason this paper will focus on North Atlantic Plio-Pleistocene radiolarian stratigraphical and palaeoceanographical records, and in particular the development of a new (preliminary) Plio-Pleistocene radiolarian zonal

scheme for the North Atlantic middle latitudes. This review is largely based on the results of various North Atlantic Legs of the Deep Sea Drilling Project (DSDP) and Ocean Drilling Program (ODP), from sites where radiolaria have been studied. These include Riedel & Hays (1969), Cita et al. (1970), Riedel & Sanfilippo (1970), Benson (1972), Petrushevskaya & Kozlova (1972), Foreman (1973), Riedel & Sanfilippo (1973), Sanfilippo & Riedel (1973), Bj6rklund (1976), Dzinoridze et al. (1978), Goll (1978), Johnson (1978), Weaver & Dinkleman (1978), Ling (1979), Sanfilippo & Riedel (1979), Westberg et al. (1980), Westberg-Smith & Riedel (1984), Westberg-Smith et al. (1986), Goll & Bj6rklund (1989) and Lazarus & Pallant (1989).

Biostratigraphy The widely-used Plio-Pleistocene radiolarian zonal scheme (Sanfilippo et al. 1985) is based largely on material from Pacific and Indian Ocean low-latitude DSDP sites and is not particularly well suited for use in the Atlantic.

From Scrutton, R. A., Stoker, M. S., Shimmield, G. B. & Tudhope, A. W. (eds), 1995, The Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region, Geological Society Special Publication No. 90, pp. 217-225

217

218

S.K. HASLETT

MAGNETIC I POLARITY SANFILIPPO ET AL.I 1ff85 Omy

(n

B.invaginata

z

Collosphaera tuberosa

r

Amphirhopalum ypsilon

RADIOLARIAN

ZONES

BJORKLUND. 1976 WESTBERG-SMITH GOLL & GOLL & THIS PAPER BJs 1980 & RIEDEL, 1 9 8 4 BJORKLUND 1989

Cycladophora davisiana

Cycladophora davisiana Cycladophora davisiana

Anthocyrtidium angulare

Stylatractus Universus

UNZONED

Pterocanium prismatium Antarctissa whitei Spongaster ? tetras UNZONED

Spongaster pentas

Antarctissa whitei Sphaeropyle langii

Pseudodictyo phimus gracilipes tetracanthus Sphaeropyle langii

Antarctissa whitei

Stichocorys peregrina

Stichocorys peregrina

Liriospyris cricus Tessarastrum thiedei

Fig. 1. North Atlantic radiolarian zonal schemes.

Stichocorys peregrina

PLIO-PLEISTOCENE RADIOLARIAN BIOSTRATIGRAPHY AND PALAEOCEANOGRAPHY This is primarily due to provinciality, for example, the lineage Spongaster berminghami (Campbell & Clark)-Spongaster pentas Riedel & Sanfilippo-Spongaster tetras Ehrenberg is complicated in the Atlantic, because whereas S. berminghami and S. pentas become extinct by the mid-Pliocene in the Pacific and Indian Oceans they have both been found living in the Atlantic. Nevertheless, most cosmopolitan radiolarian events are recorded in the Atlantic, particularly first occurrences, although there is some time delay as new species enter the Atlantic from the Indo-Pacific around the Cape of Good Hope (Casey & McMillen 1977). Therefore, until further research is undertaken on tropical Atlantic Sites, and a new low-latitude Atlantic zonal scheme devised, the scheme of Sanfilippo et al. (1985) should be adequate for coarse biostratigraphical work (Fig. 1). In middle and high latitudes the cosmopolitan zonal scheme cannot be applied. Here radiolarian faunas are often sparse, affected by dissolution and diluted with non-biogenic material, which hinders the study of radiolaria. Also, in high-latitude areas, such as the Labrador Sea (Lazarus & Pallant 1989) and the Norwegian Sea (Goll & Bj6rklund 1989), the radiolarian faunas appear to be endemic, which will ultimately result in the erection of separate zonal schemes for each sea. Goll & Bj6rklund (1989), using ODP Leg 104 material, have refined their previous schemes (Bj6rklund 1976; Goll & Bj6rklund 1980) and produced a detailed Norwegian Sea zonal scheme (Fig. 1). Considerably less attention has been given to middle latitude sites, and where radiolarian studies have been made they have focused on palaeoceanographical questions (WestbergSmith et al. 1986). However, Westberg-Smith & Riedel (1984) did attempt a zonation spanning the Middle Miocene to mid-Pliocene which, up until the Miocene-Pliocene boundary, closely resembles the scheme of Sanfilippo et al. (1985). The Early Pliocene was divided into two zones, the earliest Pliocene represented by the Stiehoeorys peregrina zone, with its top not defined by the First Appearance Datum (FAD) of S. pentas as in Sanfilippo et al. (1985), but by the Last Appearance Datum (LAD) of Stichoeorys peregrina (Riedel). This difference is because Westberg-Smith & Riedel (1984) found the LAD of S. peregrina to occur much earlier in the Atlantic than in either the Pacific or Indian Oceans. The remainder of the Early Pliocene is represented by the Sphaeropyle langii zone, which was first defined by Foreman (1975) for the North Pacific. Westberg-Smith & Riedel (1984) left the Late Pliocene and Pleistocene unzoned (Fig. 1).

219

N e w zonation The new zonation offered here (Fig. 1) is intended for use in middle-latitude North Atlantic sites, and attempts to divide the PlioPleistocene into five roughly equal, easily recognizable zones. Because of its geographically transitional nature, the new zonation incorporates elements from both high- and low-latitude zonations, and because this scheme is intended for use by commercial biostratigraphers as well as academic researchers, all zones are defined by LAD levels. This is a preliminary zonation and may be refined or changed substantially according to the results of future research, particularly with the recent JOIDES Resolution drilling in the North Atlantic. Cycladophora davisiana Partial range zone Bj6rklund (1976), Goll & Bj6rklund (1980), emended Goll & Bj6rklund (1989), emended here. Definition. From the LAD of Stylatraetus universus Hays to the Holocene. Age. e. 0--0.5 Ma (Holocene-Pleistocene). Remarks. Bj6rklund (1976) and Goll & Bj6rklund (1980) defined this zone as a partial range zone from LAD of Antaretissa whitei Bj6rklund to the Holocene, and have subsequently redefined it (Goll & Bj6rklund 1989) as a total range zone in the Norwegian Sea. The emended definition given here only applies to middle latitude sites. Stylatractus universus Partial range zone Defined here. Definition. From the LAD of A. whitei to the LAD of S. universus. Age. c. 0.5-1.75 Ma (Pleistocene). Remarks. The LAD of S. universus is a conspicuous event and a number of zones have been named after it. Chen (1975) defined a Stylatractus universus zone from the Antarctic, which is similar to the Psi zone of Hays & Opdyke (1967). Johnson et al. (1989, and references therein) also define a Stylatractus universus zone from the tropical Indian Ocean. In the Atlantic both Westberg-Smith & Riedel (1984) and Westberg-Smith et al. (1986) recognized the LAD of S. universus in DSDP Legs 81 and 94, respectively, therefore making it an important datum level for the North Atlantic with potential for interocean correlation. The position of the top of the Stylatractus universus zone in Fig. 1 is based on dates from the Pacific (Johnson & Knoll 1975) and Indian Oceans (Johnson et al. 1989), and is unlikely to be accurate.

220

S.K. HASLETT

Antarctissa whitei Partial range zone Bjrrklund (1976), Goll & Bjrrklund (1980), emended Goll & Bjrrklund (1989), emended here. Definition. From the LAD of Sphaeropyle langii Dreyer to the LAD of A. whitei. Age. c. 1.75-3 Ma (Late Pliocene). Remarks. Originally defined as a total range zone (Bjrrklund 1976; Goll & Bjrrklund 1980), and subsequently emended to a partial range zone with the base defined by the LAD of Liriospyris cricus Westberg-Smith & Riedel and the top of the F A D of Pseudodictyophimus gracilipes tetracanthus (Popofsky). Within this zone, as defined here, the FAD of Cycladophora davisiana Ehrenberg occurs at c. 2.6 Ma (Goll & Bjrrklund 1989), and may be used to divide this zone into two subzones. Furthermore, the C. davisiana FAD event may be used in southernmost middle latitude sites to recognize the Antarctissa whitei zone, where the predominantly high latitude species A. whitei may be rare or absent. Sphaeropyle langii Partial range zone Foreman (1975), emended here. Definition. From the LAD of S. peregrina to the LAD of S. langii. Age. c. 3--4.75 Ma (Pliocene). Remarks. In the North Pacific, Foreman (1975) defined the top of this zone as the LAD of S. peregrina and the base of the F A D of S. langii. However, as already mentioned, the LAD of S. peregrina in the North Atlantic occurs earlier than in the Pacific (Westberg-Smith & Riedel 1984) and cannot be used to identify the top of the zone here. In its place, the LAD of S. peregrina defines the base of the Sphaeropyle langii zone, with the evolutionary transition between S. langii and Sphaeropyle robusta defining the top. Furthermore, the FAD of S. langii occurs earlier in the Atlantic than in the Pacific. Stichocorys peregrina Total range zone Riedel & Sanfilippo (1970), emended Riedel & Sanfilippo (1978), Sanfilippo et al. (1985), non Johnson et al. (1989), emended here. Definition. From the FAD to the LAD of S. peregrina. Age. c. 4.75Ma (Early Pliocene) to Late Miocene. Remarks. Sanfilippo et al. (1985) defined the top of this zone by the S. berminghami to S. pentas evolutionary transition; however, for reasons already explained, this criterion cannot be employed in the Atlantic. Furthermore, the LAD of S. peregrina occurs earlier in the

Atlantic than in the Pacific and Indian Oceans, and consequently it is used here to define the top of this zone. The FAD of A. whitei occurs near the top of this zone, roughly coincident with the Miocene-Pliocene boundary. Also, the F A D and LAD of L. cricus, which was first described from a middle latitude site (Westberg-Smith & Riedel 1984), occurs within this zone at c. 5.5 and 5.2Ma, respectively, and may be used to divide this zone into three subzones.

Palaeoceanography Few radiolarian palaeoceanographical studies have been made in the North Atlantic; Westberg-Smith et al. (1986) attempted such a study for the entire Late Plio-Pleistocene interval of DSDP Sites 607 and 609, based on approximately five samples per core; however, their 'results indicate that analysis of much more closely spaced samples (so that several consecutive samples have similar assemblages) will be necessary for effective investigation of the palaeoenvironmental signals in the radiolarian record' (Westberg-Smith et al. 1986, 770-771). Radiolaria are potentially very useful in high resolution studies of short time intervals, as Molina-Cruz (1991) recently illustrated in a Holocene palaeoceanographical study of the northern Iceland Sea. Molina-Cruz concluded that 'fluctuations in radiolarian abundance, and the first occurrence of each species inhabiting the Iceland Sea at present' (in particular C. davisiana, Amphimelissa setosa (Cleve), Lithomitra lineata (Ehrenberg), Acrobotrys borealis (Cleve) and Stylodictya validispina (Jrrgensen)), 'coincide with changes in oceanographic conditions that occurred during the Holocene' (Molina-Cruz 1991, 303). Other research which may have important palaeoceanographical implications concerns radiolarian thanatocenoses and the recognition of a radiolarian upwelling assemblage in the eastern tropical North Atlantic.

Thanatocenoses The distribution of radiolaria in surface sediments of the present-day North Atlantic is shown by Goll & Bjrrklund (1971, fig. 1). There are three areas of the North Atlantic where radiolaria are well preserved in surface sediments; the tropical Atlantic south of 15~ N, the Caribbean and north of 45 ~ N. Therefore, there is a large area north of 15~ N, south of 45~ and east of 60 ~ W, where radiolaria are absent from, or poorly preserved in, surface sediments.

PLIO-PLEISTOCENE RADIOLARIAN BIOSTRATIGRAPHY AND PALAEOCEANOGRAPHY Subarctic

221

Boreal

7O ,60

-30

Sub tropical

u

1

60

30

0

Tropical

o.

.

Fig. 2. Percentage distribution of subarctic, boreal, subtropical and tropical radiolarian assemblages in the North Atlantic (after Matul 1990).

222

S.K. HASLETT

Matul (1989, 1990) recognized four distinct radiolarian assemblages in surface sediments throughout the North Atlantic (Fig. 2), which may prove useful in palaeoceanographical analysis of Plio-Pleistocene deposits, and in particular they may enable the tracking of water mass movement and current development through time.

Upwelling

centrations of this assemblage (up to 25%) occur where SSTs are c. 15-18~

Recently, Nigrini & Caulet (1992) described Neogene-Holocene radiolarian assemblages which characterize zones of upwelling in the Indian and Pacific Oceans, which have subsequently been used to interpret the Pleistocene upwelling history of the Somalian Gyre (Caulet et al. 1992). Haslett (unpublished data) has encountered elements of the upwelling assemblage in the eastern tropical Atlantic (ODP Hole 658C) (Fig. 3), including Pterocanium auritum Nigrini & Caulet, Acrosphaera murrayana (Haeckel), Pterocorys rainy-thorax (Nigrini), Lithostrobus hexagonalis Haeckel and Lamprocyrtis nigriniae (Caulet). In addition, Anthocyrtidium zanguebaricum (Ehrenberg) and Phormospyris scaphipes (Haeckel), found to be restricted to the eastern Atlantic by GoU & Bj6rklund (1971), are abundant, suggesting that their restricted distribution is possibly related to upwelling off the West African coast. In the Early Pleistocene of the eastern tropical Pacific, Haslett (1992) found A. zanguebaricum and P. scaphipes, and Hexacontium enthacanthum (J6rgensen) (also common at Site 658), to be more abundant at glacial maximas, when upwelling is expected to be at its most intense, than during interglacials. Furthermore, the high-latitude species Lamprocyrtis gamphonycha (Jrrgensen), C. davisiana and S. osculosa (Nigfini & Moore 1979; Lombari & Boden 1985) are common, suggesting that the waters being upwelled at Site 658 are possibly either Arctic or Antarctic in origin. The recognition of a tropical Atlantic upwdling assemblage could be extremely useful in developing an Atlantic UpwelIing Radiolarian Index (URI) along similar lines to the URI constructed by Caulet et al. (1992) for the Somalian Gyre. The Atlantic URI could then be applied to high resolution studies of sedimentary sequences below Atlantic upwelling systems, possibly revealing fluctuating patterns in upwelling intensity, ocean-atmosphere interaction and climatic change.

Tropical assemblages. Characterized by Ellipsoxiphus attractus Haec-kel, Tetrapyle quadriloba Haeckel, ?Lithocircus ?reticulata (Ehrenberg), Didymocyrtis tetrathalmus (Haeckel), Dictyocoryne profunda (Ehrenberg) and S. tetras. This

I would like to thank G. Eglington and his team at Bristol University for providing ODP 658C material; P. Judge for drafting Fig. l; C. Dunn for processing 658C samples; A. Etchells for developing and printing the radiolarian photomicrographs in Fig. 3; and my colleagues B. Funnell and K. Kennington for many hours of discussion.

Subarctic assemblage. Characterized by Spongopyle osculosa Dreyer, Phorticium clevei (Jrrgensen), A. borealis, Pseudodictyophimus gracilipes (Bailey), C. davisiana, Siphocampe arachnea (Ehrenberg) and Spongotrochus glacialis Popofsky. These species comprise > 50% of the entire radiolarian assemblage in areas where sea-surface temperatures (SSTs) are 8~ i.e. north of the northern polar front where this assemblage is coincident with the subarctic water mass of the Labrador Current.

Boreal assemblage. Characterized by Lithelius spiralis Haeckel, A. setosa, Botryostrobus eupora (Ehrenberg), S. validispina, Spongodiscus resurgens Ehrenberg, Stylochlamydium venustum (Bailey) and Stylatractus pyriformis (Bailey). This assemblage dominates radiolarian faunas where the SST is > 8~ Maximum concentrations (75%) occur where the SST ranges between 11 and 13~ however, with a further increase in SST the assemblage decreases to c. 50% where the SST is c. 21-22~

Subtropical assemblages. Characterized by Theocorythium trachelium dianae Nigrini, Lithomelissa thoracites Haeckel, Actinomma medianum Nigrini, Tricolocapsa papillosa mediterranea Haeckel, Spongocore puella Haeckel, Eucyrtidium acuminatum (Ehrenberg), Axoprunum stauraxonium Haeckel, Lamprocyclas maritalis Haeckel, Hymeniastrum euclydis Haeckel and Acrosphaera spinosa (Haeckel). Maximum con-

assemblage char-acterizes faunas south of 20~ where the SST is >24~ At the thermal equator, where the SST is c. 27~ the tropical assemblage comprises >70% of the entire radiolarian fauna.

References BENSON, R. N. 1972. Radiolaria, Leg 12, Deep Sea Drilling Project. Initial Reports of the Deep Sea Drilling Project, 12, 1085-1113.

Fig. 3. Pleistocene radiolaria from ODP Hole 658C (sample C-3) in the eastern tropical Atlantic. (a) Acrosphaera murrayana (Haeckel); (b) Spongopyle osculosa Dreyer; (e) Hexacontium laevigatum (J6rgensen); (d) Cycladophora davisiana Ehrenberg; (e) Phormospyris scaphipes (Haeckel); (f) Pterocorys minythorax (Nigrini); (g) Lamprocyclas maritalis Haeckel ventricosa Nigrini; (h) Anthocyrtidium zanguebaricum (Ehrenberg); (i) Lamprocyrtis nigriniae (Caulet); (j) Lamprocyrtis gamphonycha (J6rgensen); (k) Pterocanium auritum Nigrini & Caulet; (1) Lithostrobus hexagonalis Haeckel. All figures are x 250.

224

S . K . HASLETT

BJORKLUND, K. R. 1976. Radiolaria from the Norwegian Sea, Leg 38 of the Deep Sea Drilling Project. Initial Reports of the Deep Sea Drilling Project, 38, 1

1

0

1

-

1

1

6

8

.

CASEY, R. E. & MCMILLEN, K. J. 1977. Cenozoic radiolarians of the Atlantic Basin and margins. In: SWAIN, F. M. (ed.) Stratigraphic Micropaleontology of Atlantic Basin and Borderlands. 521-524. CAULET, J.-P., VENEC-PEYRI~,M. T., VERGNAUD-GRAZZINI, C. & NIGRINI, C. 1992. Variation of South Somalian upwelling during the last 160 Ka radiolarian and foraminifera records in core MD 85674. In: SUMMERHAYES,C. P., PRELL, W. L. & EMEIS, K. C. (eds) Upwelling Systems: Evolution Since the Early Miocene. Geological Society, London, Special Publication, 64, 379-390. CHEN,P. H. 1975. Antarctic radiolaria. Initial Reports of the Deep Sea Drilling Project, 28, 437-513. CITA, M. B., NIGRINI, C. A. & GARTNER, S. 1970. Biostratigraphy Leg 2. Initial Reports of the Deep Sea Drilling Project, 2, 391-411. DZINORIDZE,R. N., JOUSE,A. P., KOROLEVA-GOLIKOVA, G. S., KOZLOVA,G. E., NAGAEVA,G. S., PETRUSHEVSKAYA, M. G. & STRELNtKOVA, N. I. 1978. Diatom and radiolarian Cenozoic stratigraphy, Norwegian Basin; DSDP Leg 38. Initial Reports of the Deep Sea Drilling Project, Supplement to volumes 38, 39 & 40, 289-427. FOREMAN, H. P. 1973. Radiolaria of Leg 10 with systematics and ranges for the families Amphipyndacidae, Artostrobiidae, and Theoperidae.

lnitial Reports of the Deep Sea Drilling Project, 10, 407-474. 1975. Radiolaria from the North Pacific Deep Sea Drilling Project, Leg 32. Initial Reports of the Deep Sea Drilling Project, 32, 579-676. GOLL, R. M. 1978. Five Trissocyclid radiolaria from Site 338. Initial Reports of the Deep Sea Drilling Project, Supplement to volumes 38, 39, 40 & 41, 177-191. 8r BJt)RKLUND, K. R. 1976. Radiolaria in surface sediments of the North Atlantic Ocean. Micropaleontology, 17, 434-454. & - 1980. The evolution of Eucoronis fridtjofnanseni n. sp. and its application to the Neogene biostratigraphy of the NorwegianGreenland Sea. Micropaleontology, 26, 356-371. & ~ 1989. A new radiolarian biostratigraphy for the Neogene of the Norwegian Sea: ODP Leg 104. Proceedings of the Ocean Drilling Project, Scientific Results, 104, 697-737. HASLETT, S. K. 1992. Early Pleistocene glacial-interglacial radiolarian assemblages from the eastern equatorial Pacific. Journal of Plankton Research, 14, 1553-1563. HAYS, J. D. & OPDYKE, N. D. 1967. Antarctic radiolaria, magnetic reversals and climate change. Science, 158, 1001-1011. JOHNSON, D. A. 1978. Cenozoic radiolaria from the eastern tropical Atlantic, DSDP Leg 41. Initial Reports of the Deep Sea Drilling Project, 41, 763789. & KNOLL, A. H. 1975. Absolute ages of -

-

-

-

-

-

Quaternary radiolarian datum levels in the Equatorial Pacific, Quaternary Research, 5, 99110. , SCHNEIDER, D. A., NIGRINI, C. A., CAULET, J.-P. & KENT, D. V. 1989. Plio-Pleistocene radiolarian events and magnetostratigraphic calibrations for the tropical Indian Ocean. Marine Micropaleontology, 14, 33-66. LAZARUS, D. & PALLANT, A. 1989. Oligocene and Neogene radiolarians from the Labrador Sea, ODP Leg 105. Proceedings of the Ocean Drilling Program, Scientific Results, 105, 349-380. LING, H. Y. 1979. Radiolarians from the West flank of Reykjanes Ridge, Leg 49 of the Deep Sea Drilling Project. Initial Reports of the Deep Sea Drilling Project, 49, 583-588. LOMBAm, G. & BODEN, G. 1985. Modern radiolarian global distributions. Cushman Foundation for Foraminiferal Research, Special Publication, 16A, 1-125. MATUL, A. G. 1989. The distribution of radiolarians in the surface layer of the North Atlantic bottom sediments. Oceanology, 29, 740-745. 1990. Radiolaria thanatocenoses in the surface layer of the North Atlantic sediments as a reflection of natural environmental conditions. Oceanology, 30, 76-79. MOLINA-CRUZ, A. 1991. Holocene palaeo-oceanography of the northern Iceland Sea, indicated by radiolaria and sponge spicules. Journal of Quaternary Science, 6, 303-312. NIGRINI, C. & MOORE, T. C. 1979. A guide to modern radiolaria. Cushman Foundation for Foraminiferal Research, Special Publication, 16, 1-342. PETRUSHEVSKAYA, U. G. & KOZLOVA, G. E. 1972. Radiolaria: Leg 14, Deep Sea Drilling Project. Initial Reports of the Deep Sea Drilling Project, 14, 495-648. RIEDEL,W. R. & HAYS,J. D. 1969. Cenozoic radiolaria from Leg 1. lnitial Reports of the Deep Sea Drilling Project, 1, 40ff402. & SANFILIPPO,A. 1970. Radiolaria Leg 4 Deep Sea Drilling Project. Initial Reports of the Deep Sea Drilling Project, 4, 503-575. - & - 1973. Cenozoic radiolaria from a Caribbean, Deep Sea Drilling Project Leg 15. lnitial Reports of the Deep Sea Drilling Project, 15, 705-751. &- 1978. Stratigraphy and evolution of tropical Cenozoic radiolarians. Micropaleontology, 24, 61-96. SANFILIPPO, A. & RIEDEL, W. R. 1973. Cenozoic radiolaria (exclusive of theoperids, artostrobiids and amphipyndacids) from the Gulf of Mexico, DSDP Leg 10. Initial Reports of the Deep Sea Drilling Project, 10, 475-611. & - 1979. Radiolaria from the northeastern Atlantic Ocean, DSDP Leg 48. Initial Reports of the Deep Sea Drilling Project, 48, 493511. --, WESTBERc-SMITH,M. J. & RIEDEL,W. R. 1985. Cenozoic radiolaria. In: BOLLI, H. M., P~RCHNIELSEN, K. & SAtrNDERS, J. B. (eds) Plankton Stratigraphy. Cambridge University Press, Cam-

-

-

-

-

-

-

-

PLIO-PLEISTOCENE RADIOLARIAN BIOSTRATIGRAPHY AND PALAEOCEANOGRAPHY bridge, 631-712. WEAVER, F. M. & DINKLEMAN,M. G. 1978. Cenozoic radiolaria from the Blake Plateau and the Blake Bahama Basin, DSDP Leg 44. Initial Reports of the Deep Sea Drilling Project, 44, 865-886. WESTBERG, M. J., SANEILIPPO, A. & RIEDEL, W. R. 1980. Radiolarians from the Moroccan Basin, Deep Sea Drilling Project Leg 50. Initial Reports of the Deep Sea Drilling Project, 50, 429-434.

225

WESTBERG-SMITH, M. J. & RIEDEL, W. R. 1984. Radiolarians from a western margin of the Rockall Plateau: Deep Sea Drilling Project Leg 81. Initial Reports of the Deep Sea Drilling Project, 81, 479-501. - - , TWAY, L. E. & RIEDEL, W. R. 1986. Radiolarians from the North Atlantic Ocean, Deep Sea Drilling Project Leg 94. Initial Reports of the Deep Sea Drilling Project, 94, 763-777.

Note added in proof Since submission of this review, Haslett (1994) has identified a number of new Pliocene--Pleistocene radiolarian biodatums for the mid-latitude North Atlantic at DSDP Site 609. Each biodatum was dated with reference to palaeomagnetic stratigraphy. The recognition of these biodatums will enable North Atlantic radiolarian biostratigraphic schemes to be refined further. HASLETT, S. K. 1994. Plio-Pleistocene radiolarian biostratigraphy and palaeoceanography of the mid-latitude North Atlantic (DSDP Site 609). Geological Magazine, 131, 57-66.

The tephrochronology and radiocarbon dating of North Atlantic, Late-Quaternary sediments: an example from the St. Kilda Basin J O H N B. H U N T , 1 N I G E L G. T. F A N N I N , 3 P E T E R G. H I L L , 2 & J. D O U G L A S P E A C O C K 3'4

1Department of Geography & Geology, Cheltenham & Gloucester College of Higher Education, Francis Close Hall, Swindon Road, Cheltenham, GL50 3AZ, UK 2Department of Geology & Geophysics, University of Edinburgh, The Grant Institute, West Mains Road, Edinburgh EH9 3JW, UK 3British Geological Survey, Murchison House, Kings Buildings, West Mains Road, Edinburgh EH9 3LA, UK 418 McLaren Road, Edinburgh EH9 2BN, UK Abstract: A sequence of disseminated basaltic tephras of Icelandic provenance has been investigated in sediments of Late Quaternary age recovered from the St Kilda Basin, on the Scottish continental shelf. The tephras were deposited from gradually melting rafted pack ice, transported on an anti-clockwise surface current originating to the north of Iceland. The

presence of these ice-rafted tephras extends the zone of this current activity well beyond its previously documented western limit, demonstrating current impingement on the UK continental shelf. The evidence of ice-rafting, together with the biostratigraphy and a series of AMS 14C dates, confirm that this deposition occurred during the Younger Dryas chronozone. Electron probe microanalysis (EPMA) of glass-shard geochemistry is used to relate the St Kilda tephras to tephras found in marine and terrestrial deposits throughout the North Atlantic area, and to possible volcanic centres in Iceland. The joint role of tephrochronology and radiocarbon dating is discussed in relation to the comparative reliability of marine and terrestrial timescales. Problems with the chronology of the terrestrial equivalents of these tephras in Northern Iceland are highlighted.

This paper examines the application o f tephrochronology in studies of the Late Glacial stratigraphy of sediments on the northwest U K continental shelf. British Geological Survey vibrocore 57/-09/46 was selected for this investigation as earlier studies (Selby 1989; Austin 1991) demonstrated high sedimentation rates for the Late Quaternary deposits. The record of Late Quaternary environmental change reveals a period of intense and rapid climatic fluctuations in the North Atlantic area, the forcing mechanisms of which have yet to be fully understood. A reliable time frame is vital to the study of these changes, whose age and rate of change must be pinned down. The rapidity of the changes, possibly on a decadal timescale (Lehman & Keigwin 1992), can be so great that high precision and accuracy are required to define a satisfactory timescale. Radiometric 14C dating methods, which are commonly applied to marine and terrestrial deposits in the age range 20 ka BP to the present, are known to suffer from accuracy problems arising from uncertainties surrounding the

variability of the palaeo-atmospheric carbon content. This has been highlighted by the mismatch between 14C and U - T h ages from Barbados corals (Bard et al. 1990) for the last 20ka, and by a 14C 'age-plateau' in Swiss lake sediments (Ammann & Lotter 1989). The comparison of 14C ages obtained from marine and terrestrial deposits is also made difficult as a result of the longer residence time of marine carbon, prior to its incorporation in the fossil record. A correction factor must be applied, as marine and terrestrial deposits of the same age will give differing radiocarbon dates, the marine deposits appearing to be older. Uncertainty over the accuracy and universal applicability of this correction factor has resulted in the commonly held assumption that greater accuracy and reliability can be obtained from the ~4C dating of terrestrial deposits. Tephrochronology, the use of isochronous volcanic ash (tephra) horizons in the correlation and dating of sediment sequences (Thorarinsson 1944), has been applied to Icelandic tephras in the Tertiary and Pleistocene sediments of the

From Scrutton, R. A., Stoker, M. S., Shimmield, G. B. & Tudhope, A. W. (eds), 1995, The Tectonics,Sedimentationand Palaeoceanographyof the North Atlantic Region, Geological Society Special Publication No. 90, pp. 227-248

227

228

J.B. HUNT E T A L .

Fig. 1. A palaeoenvironmental interpretation for the Late Quaternary deep-water sequences of the northern Rockall Trough and the Faeroe-Shetland Channel. The Younger Dryas climatic event formed a short-lived cold phase which interrupted the gradual warming which started after c. 18 ka aP. The indication of sea-iceis consistent with the ice-rafting of tephra onto the UK continental shelf (after Stoker et al. 1989; Boulton et al. 1991). North Atlantic. In recent years, Icelandic tephras have been seen to offer the exciting potential of correlating Late-Quaternary terrestrial deposits with the record of the ocean basins and the continental shelf deposits of the North Atlantic area (see Mangerud et al. 1984). Such tephra isochrons can enable the synchroneity (or lack thereof) of climatic signals to be assessed over considerable longitudinal and latitudinal ranges. The conjunction of dating by the ~4C method, with the relative dating and correlation potential of tephrochronology promises considerable advances in our understanding of the rates of climatic change, as recorded in the Late Quaternary and Holocene of the North Atlantic area. This paper aims to provide a tephrochronological framework for future investigations into the chronology of Late Glacial climate change, as recorded in the high sedimentation rate sequences of the western UK continental margin.

Climatostratigraphic template The expansion of the Late Quaternary ice sheets of the northern hemisphere culminated in the last glacial maximum (LGM) between 20 and 18ka BP. Subsequent to this, atmospheric

warming led to the eventual disappearance of the Laurentide and Eurasian ice sheets. In the amphi-Atlantic area this warming trend was interrupted (Fig. 1) by a period of climatic deterioration (the Younger Dryas, 11 000-10 000 radiocarbon years BP) during which tundra flora replaced the north European forests, polar planktonic species replaced temperate species in the North Atlantic (Ruddiman & Mclntyre 1981), and glaciers and ice sheets either hesitated in their retreat or actually re-formed and readvanced. Although the tempo of Quaternary climate change has been largely controlled by Milankovitch orbital cyclicity, the Younger Dryas cool period does not fit with the known orbital periodicities (Berger 1990) and seems to have been affected by some other mechanisms. These mechanisms, both for the inception and termination of the Younger Dryas, have recently been the subject of considerable interest (Broecker et al. 1988, 1989; Broecker & Denton 1989; Fairbanks 1990; Lehman & Keigwin 1992), and indeed Broecker & Denton (1989) have stated that 'the explanation of the origin of this brief but intensely cold event is a major challenge to climate theorists.' Various hypotheses have been put forward to explain this climatic event, ranging from a major influx of tabular icebergs from a disintegrating

LATE QUATERNARY TEPHRAS OF THE ST KILDA BASIN Arctic ice shelf (Mercer 1969; Ruddiman & Mclntyre 1981), to shifts in orographic winds in response to retreating ice sheets (Boyle & Keigwin 1987), to a turn off of North Atlantic deep water (NADW) as a result of Laurentide melt water diversions between the Mississippi and St Lawrence (Broecker et al. 1988). Recent climate theories call for rapid reorganizations of the surface and deep-ocean circulation of the North Atlantic to explain the Younger Dryas event. Although the mechanisms remain uncertain, it is increasingly apparent that complex oceanic circulation patterns are likely to be pivotal in solving the problems of the Younger Dryas. It is known (Ruddiman & Mclntyre 1976, 1981) that the Younger Dryas corresponds to a period in which the polar front extended further south than it had done so since the Last Glacial Maximum (c. 18 ka aP). The Younger Dryas event is the most recent of the Earth's major climatic shifts. As the terrestrial and marine evidence is so relatively fresh, its study has been of great importance for climate modelling, and a full understanding of the nature of the event is necessary for the climate models to be tested. It is not within the scope of this paper to discuss the relative merits of the causal hypotheses, but to note that these important problems can only be solved by the provision of a reliable chronological framework.

Tephrochronology Tephrochronology is founded on the assumption that discrete eruptions of different volcanoes produce tephras that can be identified, distinguished and correlated on the basis of layer thickness, particle size and shape, colour, stratigraphical relationships, and mineral assemblages. In practice, particularly with distal tephras, this is usually achieved by electron microprobe analysis of the major and minor element geochemistry of glass shards whose chemistry is generally representative of the contents of the magma chamber. Although both rhyolitic and basaltic Icelandic tephras can be characterized the more evolved rhyolites usually present less ambiguity than some of the more similar basaltic tephras. In terms of their geographical distribution both rhyolitic and basaltic tephras of Icelandic origin have been found as far afield as Germany (Hunt 1992; Merkt et al. 1993), Finland (Salmi 1948) and Sweden (Persson 1966, 1971). Geochemical analysis

The geochemistry of the tephras discussed in this

229

paper was determined by electron microprobe analysis, the most reliable and effective procedure for the analysis of grains which are of small size and/or low abundance (Kittleman 1979; Westgate & Gorton 1981; Hunt & Hill 1993, 1994). The vitric shards were set in an Araldite resin on a frosted glass-slide and ground to an approximate thickness of 75#m, and then successively polished with diamond pastes of 6, 1 and 0.25#m grade. A carbon coat was evaporated on to the samples and a colloidal graphite paint applied between the individual grain mounts and the edge of the slide, thereby ensuring good electrical earthed contact via the sample holder.

Table 1. Electron microprobe data from a secondary standard (andradite garnet) and an obsidian of proven homogeneity

Mean

Standard deviation

Range

35.25 1.73 30.22 0.41 32.14 99.34

0.38 0.02 0.15 0.05 0.30 0.60

2.31 0.22 0.47 0.14 0.92 2.24

Lipari Obsidian SiO2 73.53 A1203T 12.87 FeO* 1.51 Na20 4.06 K20 4.99 Total (%) 97.85

0.36 0.24 0.06 0.06 0.09 0.61

0.99 0.75 0.17 0.19 0.22 1.80

Andradite SiO2 A1203 Fe203r* MnO CaO Total (%) n=ll

n=6

The data were gathered during the same run as the data for the St Kilda tephras, such data are necessary to quantify the stability of the microprobe during the acquisition of tephrochronologically important data. *In andradite garnet, the iron is largely ferrous and is expressed as total Fe203; in the obsidian the iron is mainly ferric, and hence expressed as total FeO.

The samples were analysed on a dual spectrometer Cambridge Instruments Microscan V, using wavelength dispersive spectrometry (WDS) with an accelerating voltage of 20 kV, a beam current of 15 nA (measured by Faraday cup), a 10s peak count per element and a defocused (5-10#m) beam. A mixture of pure metals, oxides and simple silicates were used as standards. Corrections were made for counter dead time, atomic number effects, fluorescence

230

J.B. HUNT E T A L . and absorption using a ZAF procedure described by Sweatman & Long (1969). Sodium mobility and associated over-estimation of high abundance elements were minimized by selecting a low beam current consistent with a precision of 1% or less, as determined by counting statistics on a homogeneous obsidian (Hunt & Hill 1993). A current of 15 nA proved a suitable compromise between analytical precision and accuracy of result. Beam stability and current drift were monitored by repeated analysis of an andradite secondary standard and an obsidian of known composition (Table 1).

The St Kilda Basin

Fig. 2. Location of vibrocore 57/-09/46 and the bathymetry of the St Kilda Basin in relation to Late Devensian terminal moraines. The solid arrow indicates the influx of rafted ice into the area of the Scottish continental shelf (modified from Peacock et al. 1992).

The Hebridean continental shelf of NW Scotland has remained a relatively stable margin and records a Quaternary succession characterized by glacial and glaciomarine sedimentation (Stoker 1988). Recent work has suggested that a series of morainal banks to the west of the Outer Hebrides record the maximum offshore extent of a Late Devensian (Dimlington Stadial) ice sheet (Selby 1989). A unique sedimentary basin is found in this area of the shelf (Figs 2 & 3), and

Fig. 3. The location of the North Atlantic counter-clockwise surface gyre, as delimited by the presence of icerafted tephras (NAAZ1) of Younger Dryas age (after Kvamme et al. 1989). Areas discussed in the text are: A, The St Kilda Basin and the Minch borehole; B, the Faroe-Shetland Channel tephras; C, the Skagi peninsula, northern Iceland; D, Fnjoskadalur, central north Iceland; E, the Alesund area, western Norway.

LATE QUATERNARY TEPHRAS OF THE ST KILDA BASIN

231

Fig. 4. Stratigraphy of vibrocore 57/-09/46. The Younger Dryas-Holocene (unit 1-unit 2) boundary is indicated by an increase benthic foraminifera and a drop in faunal dominance. The ~4C dates may indicate a possible hiatus at about 2.10 m depth, though this is not supported by stratigraphical evidence. The tephra is restricted to unit 1. (Modified from Peacock et al. 1992.) takes its name from the nearby island of St Kilda. This basin approaches 40 km across and lies between the morainal banks to the west and the limit of an undulating rock platform extending west from the Outer Hebrides. Seismostratigraphic (Selby, 1989) and sedimentological (Selby 1989; Peacock et al. 1992) evidence suggest that the basin formed a protected lowenergy depositional environment in which sediments post-dating the glacial maximum were deposited from suspension. The basin was surveyed and sampled by the British Geological Survey during its regional offshore mapping programme, from which several studies have already appeared (Selby 1989; Austin 1991; Peacock et al. 1992). The tephrochronology and 14C dating discussed in this paper have been established from BGS vibrocore 57/-09/46 (57~ 8~ 156 m water depth) which penetrated the upper 6 m of the succession.

Stratigraphy of vibrocore 57/-09/46 The lithostratigraphical and biostratigraphical details in this paper, together with a significant amount of the palaeoenvironmental interpreta-

tions, are taken largely from the works of Peacock et al. (1992) and Austin (1991), and are summarized in Fig. 4. L i t h o s t r atigr aph y

There are two lithological units within the core which are separated by a gradational boundary between 0.80 and 0.85 m. The lower unit (unit 1) is formed of dark grey, poorly- to well-sorted sandy silty clays and sandy-clayey silts. A gradual decrease in fines is apparent from 80% at 4.50m, to c. 50% at 1.05m. However, the sand content appears to decrease above this point. Sulphide blebs occur between 1.5 and 2.0m, and bioturbation is indicated by monosulphide mottling and by hollow and infilled tubes up to 2.5mm long and 0.5mm across. Faint layering and horizontal bioturbation are apparent on X-ray radiographs, and shell fragments are also present. Pyritization of diatoms and microfauna is common, along with sandsized amorphous pyrite grains. The coarse fraction ( > 500#m) contains well-rounded to very-angular clasts of sandstone, quartzite, quartz amphibole and rare limestone. Shards of vesicular and platy glass are also present,

232

J.B. HUNT ET AL.

Fig. 5. Occurrence, size, and age of the sand-sized (> 500 #m) tephra shards in vibrocore 57/-09/46. The curves termed p2, p3 and p4 (also pl) are taken from Selby (1989). No pattern is visible in the distribution of the four geochemical types.

between depths of c. 1-3m. Generally, the coarse fraction rarely contains grains of more than a few millimetres in size, although occasional clasts up to 20 mm have been found. The gradational boundary between the upper and lower units is shelly in its uppermost part and shows indications of bioturbation.

Biostratigraphy Unit 1 is characterized by a dominantly cold water Arctic fauna. From the base of the core to a depth of 1.05 m the molluscan fauna is sparse, of low diversity and is dominated by Nuculoma

belloti and N. tenuis. The foraminifera are dominated by Cassidulina reniforme and Elphidium excavatum, both of which are strongly associated with Arctic conditions (Peacock et al. 1992). Nonion labradoricum, a deep- and/or cold-water indicator is abundant in the middle two-thirds of the unit and reaches a maximum of 25% at a depth of 1.75m. Cold conditions are also suggested by the ostracod fauna and by the dinoflagellate cysts. Foraminifera, such as Islandiella spp. and E. excavatum are distributed so as to suggest that the conditions were coldest at the onset of deposition in unit 1. Although unit 2 has been divided into four

LATE QUATERNARY TEPHRAS OF THE ST KILDA BASIN

233

Fig. 6. Geochemical fields for the three principal areas of Late Quaternary volcanism in the North Atlantic region and selected tephra data. (After Jakobssen 1979; Larsen 1982; Dugmore 1989.) faunal subunits, an overall warming trend is apparent from the top of unit 1. Initially sparse, the molluscan fauna between 0.82 and 0.75m contains several boreal taxa, which become more abundant higher in the core. The benthic foraminifera are initially dominated by C. reniforme and E. excavatum, although these are less abundant than in unit 1. Warming is further indicated by an increase in the number of taxa, including C. laevigata, C. lobatulus and Spiroplectammina wrightii. An increase in faunal diversity is accompanied by a decrease in faunal dominance. Furthermore, from a depth of 0.72-0.32m there is a marked increase in the percentage of boreal foraminiferal taxa, most notably S. wrightii, which constitutes over 20% of the benthic taxa and continues to increase through to the sediment surface. At a depth of 0.2-0.3m, warmer, temperate conditions are also indicated by an ostracod fauna which consists largely of boreal or boreal-lusitanian species. The biostratigraphy and lithostratigraphy (Fig. 4) clearly indicate that unit 1 belongs to the Younger Dryas chronozone, and unit 2 to the Holocene. This conclusion is verified by the evidence of an adjacent core (57/-09/89) discussed by Peacock et al. (1992), in which both the Dimlington Stadial and the Windermere Interstadial are also indicated, both of which precede the Younger Dryas itself.

Tephrochronology of vibrocore 571-09/46 Late Quaternary tephras were first identified in the St Kilda Basin by Selby (1989), although Strong (1987) also reported the presence of large ( > 1 mm), but unidentified, vesicular vitric fragments. Low concentrations of tephra are present in core 57/-09/46 over a depth range of 3.41.0 m. Both basic (brown) and acidic (colourless) shards are present, and these have been grouped into 3-4 broad peaks (Peacock et al. 1992) (Fig. 5). The acidic shards are reported to be as large as 500#m across, although they are typically much smaller, and the basic shards are commonly up to 200 #m across and may approach 1500-1800#m. Strong (1987) identified shards exceeding 2 mm in length, confirmed by the data presented here. The geochemical data in this paper have been obtained from the analysis of the sand-sized basaltic tephras which were hand-picked from the coarse (> 500 #m) fractions of core 57/-09/ 46. The number of shards of this size available for analysis is dearly less than would be the case for the smaller size fractions in which the peak of a Gaussian curve (Fisher & Schmincke 1984) for the tephra size distribution would more probably occur. However, the shards are large in terms of microprobe requirements and multiple analyses have been obtained from each shard, thereby improving the confidence with which

234

J . B . H U N T E T AL.

LATE QUATERNARY TEPHRAS OF THE ST KILDA BASIN

Fig. 8. Size-distance relationships for subaerial fallout of tephra (after Fisher 1964 and Walker 1971). The St Kilda tephras fall beyond this maximum range, suggesting that the tephras were ice-rafted. each population can be defined. This can be justified on the basis that smaller shards are merely the more fragmented constituents of initially larger masses. The geochemical analyses of the St Kilda tephras indicate an Icelandic origin (Fig. 6) and defne three (four?) distinct basaltic populations (Fig. 7a~l), each originating from a single explosive eruption. The basaltic shards fall into two associations: tholeiites characterized by low FeO t (total iron; FeO + Fe203) and TiO2 content with relatively high A1203; and transitional alkali basalt with high FeO t and TiO2, and low A1203. All these tephras are restricted to lithological unit 1. The four populations are discussed below. (1) St Kilda 1 (STK-1). This population was found between 3.0 and 1.3 m depth. A total of nine sand-sized shards (1000-2000 #m) were removed, from which it was possible to obtain 44 individual analyses. The chemistry is typically tholeiitic and is

235

characterized by low TiO2 and high CaO and MgO contents. (2) St Kilda 2 (STK-2). A total of four shards (800-1200#m) were removed, between depths of 3.0 and 1.75m. From these it was possible to obtain 18 individual analyses. The chemistry again is typically tholeiitic, although characterized by high TiO2 and low A1203. (3) St Kiida 3 (STK-3). Five shards (5001800 #m) were found in the coarse fraction between 3 and 1.7 m. From these a total of 36 individual analyses were possible. The shards belong to a more evolved transitional alkali basalt association and are characterized by a high MgO and CaO cmaent and by low TiO2. (4) St Kilda 4 (STK-4). Only one analysed shard (1400#m) belonging to this population was found in the coarse fraction, at a depth of 3.0m. A total of nine individual analyses were obtained. This tephra is described separately as it appears as a discrete population on certain geochemical plots, especially with respect to TiO2. It is not certain whether ~his represents a fourth tephra event or whether it may in fact be part of the STK-2 group. Over half of these tephra shards are highly vesicular and pyrite infilling is commonly seen in polished sections (Fig. 9a). This process of intravesicular pyritization is consistent with the occurrence of pyrite in unit 1 and is probably indicative of the open network of cavities within the shards. The presence of pyrite presents another reason for the use of the electron microprobe, as techniques requiring shard digestion would generate anomalously high Fe and S values.

Tephra transportation The means by which the tephra was transported to the St Kilda basin has been the subject of

Fig. 7. Harker-type variation diagrams for the NAAZ1 (Kvamme et al. 1989) and St Kilda tephras. (a) TiO2/ A1203 plot of the St Kilda data, showing four distinct fields. The shaded areas represent the l~r error fields as calculated from the raw data using equation (1). (b) CaO v. FeO plot of St Kilda data; the distinction between STK-3 and STK-4 is less obvious. (e) (Na20 + K20) v. SiO2 plot of St Kilda data; here there is overlap between STK-4 and STK-2. The solid line marks the division between the tholeiitic and transitional alkali basalt fields. (d) (FeO/MgO) v. SiO2 plot of St Kilda data; again, four distinct fields are present. (e) TiO2/A1203 plot of the NAAZ1 data. Three distinct fields are present, if-h) Further plots of NAAZ1 data for comparison with (b-d). Three fields are present. The NAAZ1 data appear to have a much greater spread. This may be due in part to the larger data sets and/or geographical differences in the geochemistry of the tephra deposits (arising from changing wind directions during thb course of the eruption). Alternatively, this may also reflect lower analytical precision. Correlation between the St Kilda and NAAZ1 tephras is supported by these plots, although there is no equivalent for STK-4.

236

J.B. HUNT E T AL.

Table 2. Similarity coefficients for the St Kilda and other tephras, derivedfrom equation (1)

Torfadalsvatn North Atlantic Ash Zone 1 Tv-3 Tv-2

St Kilda Basin

Sk6gar Grimsv6tn

1Thol. 1 1Thol.2

1Tab. 1 STK-1 STK-2 STK-3 STK-4

Tv-3 Tv-2

1.00 0.77

1.00

1 Thol. 1 1 Thol. 2 1 Tab. 1 STK-1 STK-2 STK-3 STK-4

0.90 0.75 0.69 0.95 0.74 0.71 0.66

0.84 0.95 0.86 0.84 0.94 0.83 0.88

1.00 0.83 0.75 0.95 0.82 0.73 0.77

1.00 0.89 0.83 0.98 0.86 0.92

1.00 0.76 0.89 0.95 0.94

1.00 0.82 0.73 0.77

1.00 0.84 0.93

1.00 0.92

1.00

Sk6gar Grimsv6tn

0.69 0 . 8 5 0.75 0 . 7 3 0.91 0.81

0.89 0.94

0.97 0.90

0.75 0.83

0.87 0.94

0.96 0.88

0.93 0.94

1.00 0.90

1.00

The coefficients are obtained from mean analyses. These mean values have not been normalized (see Hunt & Hill, 1993), but each oxide is expressed as a percentage of the total oxides. For the purpose of calculating the similarity coefficient, oxides totaling < 1.0% have been disregarded as they are inherently of lower precision. Published data have been used as follows: Torfadalsvatn, Bj6rck et al. (1992); NAAZ1, Kvamme et al. (1989); Sk6gar, Nor~dahl & Hafli~ason (1992); Grimsv6tn, the average tephra geochemistry from the 1983, 1934, 1922 and 1903 eruptions (Gr6nvold & J6hannesson 1984). Values of s.c. > 0.9 are generally indicative of likel.y correlation. some discussion. Pumice rafting has been cited as a mechanism by which volcaniclastic material can be transported around the North Atlantic (Binns, 1972). This was observed following the 1947 eruption of Hekla (Thorarinsson 1967, plate XIV B). The collision of floating pumice fragments can cause a rain of abraded fragments to sink to the ocean floor. However, as evidence of abrasion commonly seen on such fragments (Fisher & Schmincke 1984) is absent from the St Kilda shards, pumice rafting is thought to be unlikely in this case. The possibility that the tephras are a result of direct airfall must be considered. Ram & Gayley (1991) report the presence of 300#m shards in the Greenland ice sheet, some 1500 km from the nearest volcanic province (Iceland). Until convincing SEM photomicrographs and reliable geochemical data are presented for these grains these interesting conclusions must be viewed with some uncertainty. Therefore, until a reappraisal of the size-distance relationships for distal pyroclastic-fall deposits is undertaken, the evidence presented by Fisher (1964) and Walker (1971) must be accepted (Fig. 8). From these works it can easily be demonstrated that the sand-sized shards of the St Kilda tephras lie well beyond their theoretical maximum transport distance. From these arguments it would appear that ice-rafting is the only plausible mechanism by which the large tephra shards could have been deposited in the St Kilda Basin, some 1000km

from their source. This conclusion is consistent with earlier work on the North Atlantic Ash Zone 1 (NAAZ1) (Ruddiman & Glover 1982; Kvamme et al. 1989) which demonstrated the existence of a counter-clockwise surface gyre during the Younger Dryas. The evidence for this gyre was provided by the gradual decrease in tephra abundance which could only have originated through the gradual melting of ash laden pack ice as it drifted from the north of Iceland, through the Denmark Strait and into the North Atlantic (Fig. 3). The evidence of the St Kilda tephras therefore suggests that the surface gyre extended further north and east than has previously been demonstrated. This correlation to the NAAZ1 is supported by the geochemistry of the basaltic tephras presented here (see Appendix), and, to a lesser extent, by the less stringent data for both basic and acidic shards presented by Selby (1989).

Correlation techniques S i m i l a r i t y coefficients

In North Atlantic tephrochronological studies geochemical correlations have been made by graphical (Harker and triangular plots) comparison alone. Although this method is adequate in the most part, it lacks the rigorous approach to correlation that is possible through the application of similarity coefficients (Borchardt et al.

LATE QUATERNARY TEPHRAS OF THE ST KILDA BASIN 1972; Hunt & Hill 1993) and discriminant function analysis (Stokes et al. 1992). For this reason we have attempted to apply the similarity coefficient (s.c.) to the Late Glacial tephras which are candidates for correlation with the St Kilda tephras (Table. 2). The similarity coefficient is given by: d(A.B) = n Y~(i: 1) Ri n

(1)

where: d(A.B) = d(B.A) is the similarity coefficient for comparison between sample A and sample B, i is the element number, n is the number of elements in the calculation, Ri XiA/XiB (if XiB > Xig ) o r Xt~/Xig (if XiA > XiB), XiA is the concentration of element i in sample A and Xm is the concentration of element i in sample B [from Borchardt et al. (1972)]. This technique was first applied in the study of the North American Bishop Ash, on the basis of data from neutron activation analysis (Borchardt et al. 1972), but has since been applied to electron microprobe data from many other American tephras (e.g. Beg~t et al. 1992). The numerical values of the similarity coefficient range from s.c. = 1, for absolutely identical samples, to s.c. = 0.6 for dissimilar samples. Borchardt et al. (1972) found that analyses of tephras from the same layer provided similarity coefficients > s.c. = 0.8, whereas Beg~t et al. (1992) state that 'similarity coefficients > = 0.95 are taken as indicative of geochemical identity and (positive) correlation, whereas s.c. = 0.900.94 may indicate a different tephra from the same volcano, and s.c. < 0.90 indicates that the tephras are geochemically dissimilar and unrelated.' The difficulty in applying the similarity coefficient to the Icelandic tephras is twofold. Firstly, the statistical variability between tephras from different eruptions of the same, and of different, volcanoes has not been previously investigated; secondly, the published data, to which correlation must be made, have been obtained from different analytical centres, thus introducing the problems of interlaboratory correlation in addition to the tephra correlation itself. For this reason, the similarity coefficients which correspond to tephra correlation, to correlation to volcano but not eruption, and to non-correlation, are not yet certain. Nevertheless, by judicious application of the similarity coefficient method, with additional stratigraphical and geographical evidence, it is possible to suggest correlation between many of the Late Glacial tephras, with reasonable confidence. =

The CIPW

237

norm

Ambiguities that may arise through the correlation of tephras by means of similarity coefficients can often be clarified by calculation of the CIPW norm (Table 3), which can be considered to simulate equilibrium crystallization for igneous rocks (Cross et al. 1903; Cox et al. 1979) and is ideally suited to unhydrated aphyric material such as glassy tephras. The actual correlations are discussed below but it is worthwhile here to illustrate the point with reference to the St Kilda tephra (STK-4). In terms of the similarity coefficient this tephra appears to correlate with two of the Kvamme et al. (1989) tephras, 1 Thol. 2 (s.c. = 0.92) and 1 Tab 1 (s.c. = 0.94). This problem of dual correlation reveals some of the limitations of the similarity coefficient which may be resolved by the CIPW calculation. STK-4 contains some normative olivine and approximates more closely to the marginally olivine normative 1 Thol. 2, as opposed to 1 Tab 1, which could crystallize 2.2% olivine. The differences are more readily seen with respect to the crystallization of hypersthene, c. 19% of which could be crystallized by both STK-4 and 1 Thol. 2, compared to 14% by 1 Tab 1. From this evidence the statistical similarity between STK-4 and 1 Tab 1 can be disregarded. By employing both of these techniques correlations of Atlantic tephras are discussed below.

Amphi-Atlantic tephra correlations In establishing tephra correlations material is considered from cores sited on Iceland, the Iceland plateau, Norway, the Norwegian Sea, the North Atlantic deep ocean, and the U K continental shelf. Suggested correlations are presented in Tables 2 & 4. North Atlantic Ash Zone 1 (NAAZ1)

Although tephras have been identified in the North Atlantic from as early as the 1940s (Bramlette & Bradley 1941), adequate geochemistry has only recently been published (Kvamme et al. 1989). Four geochemical populations (one acidic, three basic) have been identified in the ice-rafted NAAZ1. Two of the basic populations are basaltic tholeiites and one a transitional alkali basalt (Fig. 7g). These have been referred to as 1 Thol. 1, 1 Thol. 2 and 1 Tab 1, respectively (Kvamme et al. 1989). Three of the St Kilda tephras can be correlated to these components of NAAZ1 (Table 2). These are STK-1 with 1 Thol. 1 (s.c. = 0.95), STK-2 with 1

238

J.B. HUNT ET AL.

Table 3. CIPW norm valuesfor tephras in Table 2

Quartz Orthoclase Albite STK-I STK-2 STK-3 STK-4 Sk6gar Grimsvftn 1 Thol. 1 1 Thol. 2 1 Tab. 1

0.00 0.03 0.00 0.00 0.00 2.54 0.00 0.00 0.00

1.07 2.12 4.13 2.75 4.34 2.55 1.08 2.05 4.17

18.98 24.23 27.35 25.27 27.09 21.42 18.69 23.55 25.53

HyperMg Anorthite Diopside sthene Olivine Magnetite Ilmenite number 27.05 22.13 17.59 20.93 19.75 25.46 29.42 23.37 20.77

27.93 24.71 25.73 22.29 24.04 20.71 22.90 22.50 21.45

16.80 18.58 6.42 18.04 7.91 18.60 20.59 19.49 14.30

2.66 0.00 6.22 1.01 5.03 0.00 1.85 0.83 2.20

2.56 3.02 3.37 3.29 3.11 3.08 2.56 3.07 3.22

2.95 5.17 9.19 6.40 8.74 5.65 2.93 5.15 8.35

0.632 0.552 0.535 0.521 0.567 0.520 0.636 0.556 0.551

Phosphorus has been excluded from this calculation as not all published works have attempted to determine/ quote a value. Although the norm values cannot be used to determine degree of correlation between tephras in their own right they can be helpful in resolving problems arising from dual correlation by the similarity coefficient. Data from the same sources as in Table 2 Table 4. Suggested correlations between geographically separate tephras

Region St Kilda North Atlantic

Western Norway Torfadalsvatn

Sk6gar

Y T

STK-1 STK-3 STK-2 STK-4

1 Thol. 1 1 Tab. 1 1 Thol. 2

Vedde Ash

Thol. 2 (s.c. = 0.98) and STK-3 with 1 Tab 1 (s.c. = 0.95). On the basis of the geochemistry we are able to support the correlation of the St Kilda tephras with those of the NAAZ1, as was tentatively suggested earlier on the grounds of shard size and geographical distribution. The earlier work of Selby (1989) failed to identify the three basaltic components, although an additional peak of acidic shards (up to 500#m across) was correlated to the (1 R h y 1) shards found by Kvamme et al. (1989) in the NAAZ1. The tephra peak in NAAZ1 has been AMS dated to 10400 Be (Broecker et al. 1988), superseding the age estimates of c. 9800 BP suggested by Duplessy et al. (1981) and Ruddiman & Mclntyre (1981). The S k 6 g a r tephra a n d the Vedde ash

In central north Iceland (Fig. 3) the Sk6gar tephra has been found in abundance within the deltaic and glaciolacustrine sediments of a former ice-damned lake in the valley, Fnjoskadalur (Nor~dahl & Hafli~ason 1992). The Sk6gar tephra is composed of two geochemical populations, a basalt-basaltic andesite (STP-1)

Tv-3 Tv-2 Tv-1

Norwegian Sea

North Rockall ?

STP-1

?

and a rhyolite (STP-2). Comparison of the basaltic member of STP-1 with St Kilda tephra STK-3 shows that they are the same tephra (s.c. = 0.96). This correlation is also supported by the CIPW norms, particularly in respect of olivine, hypersthene, orthoclase and albite (Table 3). Although the Sk6gar tephra has not been directly dated there is strong evidence (Nor~dahl & Hafli~ason 1992) to support a correlation between the bimodal Vedde ash (Mangerud et al. 1984) and both STP-1 and STP-2. The Vedde ash has been found both in the middle of Younger Dryas lacustrine and sublittoral sediments in western Norway and has been 14C dated to 10 600-4-60 BP (Mangerud et al. 1984). This has become a widely accepted date for the age of the Vedde ash, and for the mid Younger Dryas itself. Correlation between the Vedde-Sk6gar-St Kilda (STK-3) therefore provides an approximate age for the tephra bearing sediments in core 57/-09/46, which is in near agreement with the 10 400 BP age from the NAAZ1 (see above). Both components of the Vedde ash have been found to the north of Iceland (Sejrup et al.

LATE QUATERNARY TEPHRAS OF THE ST KILDA BASIN 1989), from the eastern Norwegian Sea (Mangerud et al. 1984; Karpuz & Jansen 1992), from the northern North Sea (Long & Morton 1987) and from the Faeroe-Shetland Channel and northern Rockall Trough (Stoker et al. 1989). T h e N o r t h M i n c h ( 7 8 / 4 ) tephra

The British Geological Survey borehole 78/4 in the N o r t h Minch 9 k m SE of Stornoway penetrated a total of 57.77m of Quaternary sediments (Graham et al. 1990). Peaks containing both acidic and basic tephras were located at depths of 22.75 and 24.50m in the borehole, where biostratigraphy was considered indicative of the Younger Dryas. The two peaks are assumed to represent the same eruption, though the large interval between the peaks is not adequately explained. Although the geochemical data are presented in a manner which makes direct comparison with data from other sources impossible, the conclusion that the tephra is equivalent to the Vedde ash (and hence 1 Tab 1 and STK-3) is probably correct as this is the only contemporaneous bimodal tephra. AMS 14C dates have been obtained which bracket the lower of the two peaks. These ages are 10860+120 BP at a depth of 25.74-26.00m (Graham et al. 1990) and 10755+70 BP at a depth of 24.25-24.50 m (this study). T h e T o r f a d a l s v a t n tephras

The study of the tephrochronology of Torfadalsvatn (Bjrrck et al. 1992), one of the numerous lakes on the Skagi peninsular of northern Iceland (Fig. 3), probably constitutes the most significant advance in the Late Glacial and early Holocene tephrochronology of the North Atlantic area since Mangerud et al. (1984) first demonstrated that distal Icelandic tephras could be used to correlate marine and terrestrial sediments in the Younger Dryas. To date, studies of marine tephras have been unable to resolve discrete layers of tephra. The importance of the Torfadalsvatn study is derived from the high sedimentation rate within the lake which has resulted in the tephras forming discrete and identifiable layers. This has therefore enabled the eruptive, and hence stratigraphical, sequence of the tephras to be identified for the first time. Bjrrck et al. (1992) have identified five important tephra layers which seem to span the Younger Dryas and early to mid Holocene. Three of these are of relevance to this paper. These are discussed below, and correlations are again discussed with reference to the similarity coefficients found in Table 2.

239

(1) Tv-1. This is the oldest tephra and forms a 1.0cm thick band at a depth of 11.13m below the lake bed. Compositionally the tephra is a tholeiitic basalt characterized by relatively high TiO2 and AlzO3. On the basis of similarity coefficients Tv-1 can be correlated to the St Kilda tephra STK-2 (s.c. = 0.94) and to the Kvamme et al. (1989) 1 Thol. 2 (s.c. = 0.95). Bjrrck et al. (1992) date this tephra to 10 700010 800 BP on the basis of interpolation between AMS dated horizons. (2) Tv-2. This occurs as a 1.0-1.5cm thick layer of greyish black tephra at a depth of l l.05m. Compositionally it is bimodal, consisting of a rhyolitic and a basalticintermediate transitional alkali basalt component. From the published data it is not possible to calculate a similarity coefficient, but Bjrrck et al. (1992) confidently assign this tephra to the bimodal Vedde. The basaltic population can therefore be linked with STP-1 (Nor~dahl & Hafli~ason 1992), and with the St Kilda tephra STK-3. This is therefore dated to 10 600 BP on the basis of Mangerud et al. (1984), as discussed earlier. (3) Tv-3. An additional tholeiitic basalt, 0.3 cm thick, is found at a depth of 10.67m. It is characterized by low TiO2 and high MgO and CaO. The tephra may be correlated to the Kvamme et al. (1989) 1 Thol. 1 (s.c. = 0.95) and to the St Kilda tephra STK-2 (s.c. = 0.94). Bjrrck et al. (1992) date this tephra to c. 9200 BP, again on the basis of interpolation. This dating may be somewhat problematic, as will be discussed later.

T h e source volcanoes

Determination of the volcanic source of the Younger Dryas tephras is not as straightforward as is often the case with younger tephras. The Iceland of Younger Dryas times, and in particular the neo-volcanic zone, was covered in the main part by an extensive ice sheet (Ingolfsson 1991), and as a consequence proximal in situ tephras are virtually unknown. It is therefore not possible to trace tephra isopachs as they thicken towards a volcanic centre. To ascribe a source for these tephras it is therefore necessary to rely on geochemical characteristics of both tephra and volcano. Fortunately, extensive studies of Icelandic rock suites of Holocene age has demonstrated that geochemical trends can be used to characterize different volcanic centres/areas (Imsland 1978;

240

J.B. HUNT E T AL.

Fig. 9. (a) Reflected light photomicrograph of a polished section of one of the St Kilda tephras (1 Thol. 1), illustrating the degree of intravesicular pyritization. The pyrite forms the brighter areas of the shard. (b & e) SEM photomicrographs of large (c. 1 mm), blocky tephra shards from a depth of 2.2 m. (d) SEM photomicrograph of a more vesicular shard from the same core. Note the elongation of the vesicles formed by flow of the melt during the eruption. Jakobsson 1972, 1979). Using Jakobsson's (1979) whole-rock geochemical data, obtained from lavas, it is possible to apply the similarity coefficient (Table 5) to suggest correlations between the St Kilda tephras and source volcanoes. (1) STK-1. The volcano Veidiv6tn (Fig. 10) has been suggested as a possible source for 1 Thol. 1, the NAAZ1 equivalent of this tephra (Kvamme et al. 1989). This suggestion is strongly supported by the geochemical correspondence between STK-1 and Veidiv6tn (s.c. = 0.97) as presented in Table 5. The Younger Dryas age of STK1 is in accord with the earlier of the two periods of activity of the southern part of the Veidiv6tn system, in the interval 11000--6 500(?) BP (Vilmundardottir 1977).

Table 5. Similarity coefficients for the St Kilda tephras

and their possible volcanic sources

St Kilda tephras

Volcanic centre

STK- 1

Veidivotn Grimsvrtn Hekla Katla

0.97 0.83 0.76 0.75

STK-2

STK-3

0.81 0.97 0.91 0.88

0.72 0.86 0.92 0.97

STK-4 0.77 0.92 0.93 0.92

* Source geochemistry from Jakobsson (1979). The V e i d i v r t n system lacks a central volcano (Jakobsson 1979) which could, like G r i m s v r t n today, have formed a N u n a t a k in the extensive Younger Dryas ice sheet. Veidivrtn would therefore have been covered by a considerable thickness of

LATE QUATERNARY TEPHRAS OF THE ST KILDA BASIN

241

Fig. 10. The volcanic zones of Iceland with the location of the main volcanic centres which may have contributed to the Younger Dryas tephra record (after Imsland 1978). ice and a tephra producing eruption would have necessitated an extremely explosive phase in order to puncture the ice surface. If the STK-1 tephra was produced by the first eruption of this system after a period of quiescence, then such an eruption may indeed have been highly explosive. In addition, the water beneath the glacier may have entered an opening fissure system, resulting in enhanced magma volatile content and the initiation of an explosive phreato-magmatic eruption. (2) STK-2. The 1 Thol. 2 tephra of the NAAZ1 to which STK-2 is correlated has been ascribed to an eruption of the Grimsv6tn volcano (Kvamme et al. 1989) in the centre of the Vatnajokull ice sheet (Fig. 10). This correlation is supported by the analyses of STK-2 which show a good correspondence with the Grirnsv6tn data (s.c. = 0.97). (3) STK-3. The data from this tephra support correlation to the Katla volcano (s.c. = 0.97), and are in agreement with Mangerud et al. (1984), who correlate the Vedde ash to Katla. Kvamme et al. (1989), however, raised the possibility that Hekla could be a source of the NAAZ1 tephra, 1 Tab 1 [and hence the basic part of the Vedde ash, STK-3, and the Nor~dahl & HafliOason

(1992) Skrgar, STP-1 tephra]. The correlation between Hekla and STK-3 is relatively strong (s.c. = 0.92), but does not match that with Katla which therefore appears the prime candidate for STK-3, in agreement with the conclusions of NorOdahl & Hatti~ason (1992). (4) STK-4. As discussed earlier, there is some uncertainty over the nature of this tephra and this is further illustrated by the high s.c. values for three separate volcanoes. Further members of this population must be analysed before any firm conclusions can be made.

Radiocarbon dating and the age of the tephras Calibration of the 14C timescale over the past 30000 years, using mass spectrometric U - T h ages from Barbados corals, has shown that 14C ages are systematically younger than might at first sight be expected (Bard et al. 1990). This is supported by the evidence from Swiss lake sediments (Amman & Lotter 1989). Additional evidence for the greater antiquity of Late Quaternary events has recently appeared from a study of the Summit core from the Greenland

242

J.B. HUNT E T AL.

Table 6. Radiocarbon ( A M S ) dates from vibrocore 57/-09/46 and Borehole 78/4 Laboratory number

Species

Depth in core (m)

Vibrocore 57[-09/46 OXA-2786 OXA-2787 TO-3127 TO-3128 OxA-1324 OXA-2788 TO-3126

Acanthocardia echinata Nuculoma belotti Nuculoma belotti Nuculoma tenuis Buccinum terraenovae Nuculoma belotti Nuculana pernula

0.47-0.51 1.05-1.30 2.06-2.09 2.30-2.33 4.50 4.80-5.00 5.65-5.68

BGS borehole 78/4 TO-3129 Portlandia arctica

Conventional age Adjusted age (~4Ca B1,+ ltr) (14Ca BP4- lcr) Source

10 380 4-100 10 580 4- 100 10 610 4- 70 10 970 4- 70 11 680 4- 240 11420 4- 120 11 400 4- 70

24.25-24.50 11 160 -4-70

9975 4-110 10 175 4-110 10 205 4- 80 10 565 4- 80 11275 4- 250 11015 4-130 10 995 4- 80

Peacock et al. (1992) Peacock et al. (1992) This paper This paper Hedges et al. (1988) Peacock et al. (1992) This paper

10 755 • 80

This paper

613C are as follows: TO-3127, 613C = +2.8%0; TO-3128 613C = + 1.9%o; TO-3126 613C = -0.1%o; TO-3129 613C = + 0.7%0. Information on the Oxford dates can be seen in Peacock et al. 1992. The adjusted age is based on an apparent age of 405 +40 years for sea water (Harkness, 1983). The OXA-1324 date is from Hodges (1988).

ice sheet (Johnsen et al. 1992). Preliminary absolute ages from this core, determined by counting of annual variations, suggests that the Younger Dryas climatic event was initiated at 127004-100 ae (calendar) and terminated at 11 550 BP (calendar). Despite these problems regarding the accuracy of the 14C timescale its value can be maintained provided that 14C dates are internally consistent. We believe that the most reliable way of monitoring this consistency is by the 14C dating of carefully and consistently selected organic material associated with geographically extensive isochronous tephras, such as were deposited in the North Atlantic during the Younger Dryas. For the most part the amphi-Atlantic tephra correlations suggested earlier conform to the 14C dates which have been obtained from surrounding lacustrine, deep-marine and continental-shelf sediments (Table 6). A problem arises, however, with the dating of Tv-3 (Bjrrck et al. 1992). Stratigraphically Tv-3 is the youngest of the three Torfadalsvatn tephras discussed here, and is correlated with the Kvamme et al. (1989) 1 Thol. 1 tephra, which in turn can be correlated to the St. Kilda Basin. Based on interpolation between an AMS date of 9180+210 BP obtained 7cm above the tephra, and an AMS date of 9890 + 290 m, from 19 cm below the tephra, Tv-3 is dated to c. 9200 Be. Bj6rck et al. (1992) recognized that this was markedly different from the age of 10 200-11 000 BP for 1 Thol. 1 in the NAAZ1 (Kvamme et al. 1989). Agreeing that Tv-3 and 1 Thol. 1 are the same tephra they believe 'that this age difference

well illustrates the problem of obtaining a reliable 14C chronology in marine sediments with low time resolution'. In some cases marine 14C dates are certainly problematical; however, this conclusion is not supported by the evidence of the St Kilda tephras and the associated AMS dates. These, in fact, suggest two alternatives. (1) That it is the terrestrial high resolution age of 9200 BP (Bjrrck et al. 1992) which should be considered unreliable. Given that the errors on the Torfadalsvatn dates (+ 210 and + 290 years) are large, and that it was only possible to date bulk sediments, such unreliability would not be all together surprising. (2) That despite the high similarity coefficients between Tv-3 and 1 Thol. 1, these tephras are not the same, and that Tv-3 constitutes a previously unrecognized tephra. This alternative is considered less likely as the presence of 1 Thol. 1 in marine cores to the north of Torfadalsvatn (see Kvamme et al. 1989), suggests that 1 Thol. 1 should be present in Torfadalsvatn itself. The evidence of the St Kilda tephras is not ideal. Insufficient contiguous high resolution subsamples were available to constrain geochemically the actual disseminated peaks (Fig. 5) found by Selby (1989). The tephras have therefore not been resolved into a stratigraphical sequence of eruptive events, as was possible at Torfadalsvatn (Bjfrck et al. 1992). This is unfortunate as the sediments in the St Kilda Basin offer the potential for a higher time resolution than was seen at Torfadalsvatn, and closely spaced tephra samples could have been used to examine bioturbation profiles. However,

LATE QUATERNARY TEPHRAS OF THE ST KILDA BASIN

243

Fig. 11. Adjusted 14Cages (AMS) v. depth in vibrocore 57/-09/46. The vertical bars on the 14Cdates are standard deviations at 1 s. Horizontal bars represent depth ranges of each sample. A possible non-sequence exists between 2.09 and 2.33m depth, as dates TO-3127 and TO-3128 give greater than expected age differences for a small difference in depth. There is, however, no stratigraphical evidence to verify this possibility. it has been possible to demonstrate that the different geochemical populations (1 Thol. 1, 1 Thol. 2, 1 Tab 1 and STK-4) are all restricted to unit 1 (the Younger Dryas) in 57/-09/46. Whilst it is possible to introduce old material into younger sediments, by sediment reworking, the biostratigraphical evidence in this vibrocore suggests that the introduction of younger (Holocene) tephras into older deposits > 2m below the Younger Dryas-Holocene boundary is highly unlikely, if not impossible. Furthermore, the actual presence of large shards of the STK-1/ 1 Thol. 1 population in the St Kilda Basin indicates a Younger Dryas age for the eruption, as this is the last period in which ice rafts could have reached this southern latitude. From the presence of nine sand-sized shards in unit 1 we are forced to conclude that STK-1 (and hence 1 Thol. 1 and Tv-3) is a Younger Dryas tephra. This is supported by three new AMS dates presented here (Figs 5 & 9; Table 6), and by those of Peacock et al. (1992). The

reliability of these dates appears to be high, as indicated by their relatively low standard deviations, and by the interpolated age for Selby's (1989) acidic peak, p2. On the assumptions that the position of the peak is well constrained, and that, being a rhyolitic population, it correlates to the Vedde ash, the age corresponds remarkably well (Fig. 11) with the 106004-60 aP age, as determined for the Vedde by Mangerud et al. (1984).

Conclusions Late Quaternary tephras from the St Kilda Basin have been analysed by electron microprobe. These have been correlated, on the basis of their geochemistry, to marine tephras (NAAZ1 tephras: 1 Thol. 1, 1 Thol. 2 and 1 Tab. 1) from the North Iceland Plateau, the Norwegian Sea, and the North Atlantic, and to terrestrial tephras from Iceland and Norway. In establishing these correlations similarity coefficients (s.c.)

244

J.B. HUNT E T AL.

are found to have been particularly useful, as has been the calculation of the CIPW norm, a technique not previously employed by tephrochronologists. A l t h o u g h w o r k on equivalent tephras in Iceland (Bj6rck et al. 1992) has suggested that one of the tephras (Tv-3, and hence STK-1)is of Holocene age, the evidence from vibrocore 57/09/46 indicates that the St Kilda tephras are undoubtedly all of Younger Dryas age. This conclusion is supported by a series of seven AMS dates obtained from the tephra-bearing unit 1. We therefore suggest that it is unwise to assume that terrestrial ~4C dates are perforce more reliable than well controlled dates obtained from high resolution cores in marine sediments. As much reliance may be placed upon tephrochronology in the dating of marine and terrestrial sequences it is important that numerical ages are firmly established, and that incorrect ages are not perpetuated. F r o m a consideration of particle size, shard morphology and on the basis of geographical comparisons with geochemically equivalent tephras, it is concluded that the tephras were transported to the St Kilda Basin aboard gradually melting pack-ice, rafted on a counterclockwise surface gyre which operated during

the Younger Dryas. This evidence demonstrated that the magnitude of this gyre (northern and eastern extent) is greater than was previously supposed. It is possible that this surface gyre, by impinging on the U K continental shelf, could have provided a mechanism whereby variations in meltwater discharge from the Laurentide ice sheet may have impacted upon the climate of the British Isles. Finally, in order that optimum information is obtained both with respect to tephrochronology and to the use of tephras for bioturbational modelling, we strongly believe that analysis (including identification, detailed stratigraphical location, geochemical characterization, regional correlation and dating) of tephra to be one of the major objectives in the subsampling of high resolution sequences of Younger Dryas and early Holocene age, on the NE Atlantic margin. J.B.H. acknowledges the receipt of a NERC research studentship and the support of his project supervisors (N.G.T.F., P.G.H. and R. Thompson). The SEM photomicrograph was taken at the University of Edinburgh, Science Faculty SEM Facility, with the assistance of John Findlay. The obsidian used as a glass standard was kindly provided by S. Sparks. N.G.T.F. publishes with the permission of the director of the British Geological Survey.

Appendix WDS geochemical analyses of the St Kilda tephras, performed on a Cambridge Instruments Microscan V, under the following conditions: an accelerating voltage of 20 kV, a beam current of 15 nA (measured by Faraday cup), a 10 s peak count per element, and a defocused (5-10#m) beam. A mixture of pure metals, oxides and simple silicates were used as standards. Corrections were made for counter dead time, atomic number effects, fluorescence and absorption using a ZAF procedure described by Sweatman & Long (1969).

St Kilda tephra, STK-1 SiO2

TiO2

A1203

FeOT

MnO

MgO

CaO

Na20

K20

P205

Total

49.58 50.28 47.98 47.22 14.48 48.61 47.76 47.81 47.53 47.85 48.26 48.56 48.56 48.33 48.43 48.36 48.33 48.26

1.57 1.56 1.65 1.48 1.53 1.46 1.62 1.56 1.46 1.44 1.49 1.51 1.43 1.65 1.56 1.54 1.59 1.47

13.36 13.45 13.92 13.35 13.68 13.66 13.81 13.76 13.64 13.41 13.70 13.90 13.77 13.75 13.63 13.60 13.79 13.91

11.22 11.78 11.34 11.57 11.42 11.63 11.47 12.76 11.81 11.21 11.59 11.58 11.50 11.39 11.49 11.31 11.31 11.51

0.24 0.22 0.18 0.28 0.22 0.25 0.20 0.26 0.27 0.22 0.19 0.26 0.25 0.27 0.21 0.20 0.23 0.24

7.47 7.28 7.80 7.27 7.59 7.52 7.40 8.07 7.64 7.45 7.35 7.27 7.08 7.44 7.64 7.48 7.51 7.38

12.49 2.13 12.28 2.14 12.20 2.07 11.88 2.37 12.10 2.20 12.05 2.02 12.04 2.30 12.02 2.35 12.25 2.33 11.88 2.52 11.92 2.09 11.90 2.14 12.14 2.04 11.99 2.20 12.17 2.08 1 1 . 7 9 2.11 12.21 2.08 11.96 2.10

0.18 0.17 0.09 0.19 0.17 0.18 0.20 0.19 0.23 0.23 0.19 0.18 0.20 0.19 0.19 0.15 0.16 0.26

0,00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.03 0.00 0.08 0.10 0.10 0.00 0.03 0.00 0.00

98.23 99.16 97.33 95.61 97.39 97.36 96.80 98.78 97.16 96.25 96.79 97.39 97.07 97.31 97.39 96.58 97.21 97.08

LATE Q U A T E R N A R Y TEPHRAS OF THE ST KILDA BASIN 48.53 49.04 47.88 47.90 49.00 48.22 47.84 48.26 48.17 48.54 48.22 47.36 47.60 47.70 48.19 49.80 49.68 49.46 48.78 48.95 49.34 49.60 48.19 50.12 49.33

1.45 1.55 1.47 1.54 1.54 1.50 1.55 1.55 1.54 1.60 1.56 1.72 1.77 1.48 1.54 1.50 1.37 1.45 1.46 1.50 1.44 1.37 1.40 1.43 1.44

13.59 13.96 13.58 13.56 13.63 13.68 13.81 13.59 13.76 14.06 13.73 14.41 !4.22 14.82 14.60 13.69 13.66 13.52 13.56 13.89 13.79 13.68 14.31 13.90 13.78

11.72 11.64 11.39 11.37 11.64 11.56 11.27 11.55 11.49 11.38 11.49 11.05 11.10 10.72 10.88 10.60 10.80 10.53 10.73 10.78 10.94 10.70 10.72 10.58 10.71

0.25 0.20 0.22 0.20 0.23 0.20 0.22 0.20 0.21 0.22 0.26 0.22 0.17 0.20 0.21 0.23 0.25 0.19 0.19 0.24 0.21 0.21 0.26 0.23 0.22

7.44 7.34 7.23 7.45 7.46 7.80 7.68 7.36 7.51 7.33 7.35 8.18 7.98 8.11 8.26 7.76 7.55 7.54 7.65 7.80 7.63 7.69 7.78 7.83 7.69

12.03 12.05 11.98 12.13 11.77 12.03 12.03 11.51 12.06 12.11 11.97 11.66 11.73 11.98 11.87 11.77 11.92 12.16 12.17 12.43 12.09 12.27 11.77 12.23 12.09

245

2.13 2.11 2.15 2.18 2.24 2.18 2.17 2.05 2.01 2.03 2.13 2.14 2.22 2.24 2.01 2.20 2.36 2.20 2.30 2.30 2.27 2.26 2.42 2.38 2.29

0.22 0.21 0.20 0.17 0.23 0.19 0.18 0.16 0.21 0.18 0.18 0.16 0.18 0.14 0.17 0.18 0.16 0.15 0.16 0.17 0.17 0.15 0.21 0.17 0.17

0.10 0.00 0.68 0.16 0.47 0.00 0.00 0.42 0.00 0.00 0.00 0.10 0.00 0.00 0.16 0.00 0.00 0.00 0.11 0.00 0.22 0.00 0.05 0.00 0.04

97.45 98.10 96.78 96.65 98.22 97.36 96.77 96.64 96.98 97.43 96.88 96.91 96.97 97.39 97.87 97.74 97.75 97.20 97.11 98.05 98.09 97.94 97.12 98.76 97.75

43 analyses, 9 shards.

St IOlda tephra, STK-2 SiO2

TiO2

A1203

FeO T

MnO

MgO

CaO

Na20

K20

P205

Total

48.90 49.25 49.81 49.32 48.96 49.56 50.08 49.23 49.46 50.40 50.10 49.73 49.21 49.86 49.66 49.06 49.37

2.70 2.87 2.74 2.77 2.83 2.60 2.66 2.50 2.65 2.73 2.67 2.64 2.66 2.67 2.54 2.79 2.66

12.57 13.27 12.88 12.90 13.30 12.94 12.93 13.09 13.06 13.32 13.54 12.89 13.07 13.21 13.24 13.35 13.30

13.94 13.91 13.98 13.94 13.68 13.55 13.62 13.68 13.63 13.36 13.16 13.62 13.36 13.38 13.10 13.82 13.46

0.33 0.29 0.23 0.28 0.22 0.26 0.25 0.28 0.25 0.28 0.23 0.22 0.23 0.24 0.20 0.26 0.23

6.01 5.59 6.11 5.90 6.15 5.96 6.23 5.93 6.07 6.00 5.97 5.78 5.83 5.89 5.89 5.79 5.84

9.97 9.82 10.31 10.03 10.50 10.50 10.62 10.09 10.43 10.36 10.22 10.72 10.51 10.45 10.46 10.43 10.44

219 217 279 215 219 271 200 211 215 269 309 208 208 279 217 211 214

0.42 0.35 0.34 0.37 0.34 0.36 0.29 0.35 0.34 0.38 0.40 0.33 0.38 0.37 0.34 0.37 0.35

0.32 0.11 0.16 0.20 0.00 0.00 0.00 0.11 0.03 0.00 0.00 0.43 0.00 0.11 0.00 0.00 0.00

98.04 98.32 99.36 98.57 98.97 98.43 99.57 98.07 98.76 99.52 99.37 99.04 97.93 98.97 98.31 98.71 98.51

17 analyses, 4 shards

St Kilda tephra, STK-3 SiO2

TiO2

A1203

FeO T

MnO

MgO

CaO

Na20

K20

P205

Total

45.83 45.98 45.80 45.45 46.25 45.89 45.64

4.74 4.55 4.62 4.60 4.72 4.75 4.66

12.56 12.52 12.40 12.62 12.10 12.49 12.40

14.71 14.46 14.47 14.56 14.70 14.81 14.65

0.28 0.30 0.23 0.28 0.28 0.28 0.24

5.04 5.13 4.96 4.94 5.03 5.09 4.88

9.63 9.60 9.55 9.50 9.66 9.64 9.79

3.10 3.08 3.03 3.05 3.15 3.24 3.30

0.80 0.75 0.70 0.71 0.69 0.69 0.76

0.16 0.16 0.41 0.26 0.00 0.62 0.46

96.84 96.54 96.16 95.98 96.58 97.50 96.77

J.B. H U N T ET AL.

246 47.49 46.80 46.95 46.75 47.26 47.58 46.43 47.22 46.61 47.01 47.25 46.62 46.85 47.86 46.97 47.12 47.32 46.21 47.31 47.06 45.65 47.28 47.76 47.32 47.00 46.96 46.02 47.73 46.80

4.48 4.64 5.04 4.59 4.71 4.68 5.06 4.76 4.82 4.75 4.75 4.71 4.60 4.77 4.46 4.92 4.91 4.86 4.68 4.74 4.57 4.51 4.84 4.92 4.71 4.59 4.72 4.85 4.72

12.48 12.60 12.34 12.91 12.83 12.23 12.48 12.48 12.30 12.52 12.31 12.57 12.88 12.60 12.67 12.46 12.71 11.99 12.41 12.51 12.78 12.64 13.06 12.51 12.75 12.53 12.30 12.91 12.58

15.07 14.78 14.75 15.37 15.00 14.41 15.21 14.98 14.59 14.91 15.24 15.29 14.72 14.50 14.85 14.78 15.00 14.94 14.98 14.92 14.89 14.62 15.03 15.09 14.91 14.96 14.96 14.93 14.95

0.27 0.27 0.29 0.36 0.27 0.26 0.25 0.23 0.27 0.28 0.28 0.24 0.24 0.26 0.25 0.26 0.24 0.25 0.27 0.25 0.27 0.27 0.29 0.26 0.28 0.19 0.21 0.30 0.23

4.99 5.12 5.09 4.70 5.08 5.02 4.86 5.09 5.14 5.01 5.17 5.12 5.11 5.10 5.35 5.06 5.07 4.97 5.06 5.11 5.30 5.00 4.99 4.97 5.07 4.96 5.18 4.88 5.01

9.64 9.28 9.12 9.31 9.25 9.34 9.25 9.69 9.35 9.36 9.58 9.65 9.47 9.64 9.48 9.41 9.99 9.78 9.58 9.62 9.51 9.19 9.53 9.73 9.49 9.73 9.57 9.86 9.72

3.18 3.32 3.10 3.10 3.15 3.08 3.26 3.13 3.20 3.17 3.12 3.30 3.09 2.95 3.17 3.23 3.09 3.13 3.11 3.13 3.04 3.19 3.11 3.19 3.13 3.17 3.21 3.20 3.19

0.78 0.79 0.70 0.69 0.72 0.78 0.68 0.67 0.68 0.72 0.69 0.76 0.65 0.66 0.72 0.72 0.72 0.76 0.71 0.71 0.77 0.63 0.73 0.85 0.75 0.66 0.72 0.67 0.69

0.59 0.21 0.11 0.11 0.00 0.00 0.48 0.69 0.27 0.27 0.11 0.11 0.21 0.96 0.53 0.00 0.64 0.80 0.43 0.42 0.48 0.59 0.96 0.75 0.69 0.21 0.58 0.16 0.32

98.97 97.82 97.50 97.89 98.27 97.38 97.96 98.95 97.22 98.00 98.50 98.36 97.81 99.30 98.46 97.94 99.70 97.69 98.53 98.48 97.26 97.91 100.30 99.60 98.77 97.97 97.48 99.19 98.21

36 analyses, 5 shards

St Kiida tephra, STK-4 SiO2

TiO2

A1203

FeO T

MnO

MgO

CaO

Na20

K20

P205

Total

48.76 48.85 49.54 49.40 48.14 49.49 48.68 48.66 48.74

3.16 3.31 3.22 3.31 3.19 3.17 3.24 3.21 3.38

12.20 12.56 12.64 12.39 12.76 12.92 12.55 12.73 12.42

14.21 14.26 14.23 14.53 14.08 14.50 14.47 14.35 14.32

0.21 0.27 0.30 0.31 0.22 0.27 0.25 0.24 0.28

5.34 5.20 5.18 5.26 5.53 5.30 5.34 5.42 5.34

9.18 9.10 9.22 9.36 9.42 9.26 8.98 9.31 9.21

2.76 2.91 2.82 2.17 2.78 2.69 2.91 2.92 2.88

0.45 0.41 0.43 0.47 0.48 0.45 0.48 0.48 0.43

0.27 0.32 0.16 0.16 0.05 1.29 0.32 0.32 0.32

96.55 97.21 97.74 98.38 96.64 99.34 97.21 97.65 97.33

9 analyses, 2 shards

References AMMAN, B. & LOTTER, A. F. 1989. Late-Glacial radiocarbon and palynostratigraphy on the Swiss Plateau. Boreas, 18, 109-126. AUSTIN, W. E. N. 1991. Late Quaternary benthonic

foraminiferal stratigraphy of the western U.K. continental shelf. PhD thesis, University College of North Wales, UK. BARD, E., HAMELIN, B., FAIRBANKS, R. G. & ZINDLER, A. 1990. Calibration of the ~4C timescale over the past 30,000 years using mass spectrometric U - T h ages from Barbados corals. Nature, 345, 405-410. BEGET, J., MASON, O. & ANDERSON, P. 1992. Age,

extent and climatic significance of the c. 3400 nP Aniakchak tephra, western Alaska, USA. The Holocene, 2, 51-56. BINNS, R. E. 1972. Composition and derivation of pumice on post glacial strandlines in northern Europe and the western arctic. Geological Society of America Bulletin, 83, 2303-2324. BJt3RCK, S., INGOLFSSON, O., HAFLISASON,H., HALLSDOTTIR, M. & ANDERSON, N. J. 1992. Lake Torfadalsvatn: a high resolution record of the North Atlantic ash zone I and the last glacialinterglacial environmental changes in Iceland. Boreas, 21, 15-22.

LATE QUATERNARY TEPHRAS OF THE ST K I L D A BASIN BORCHARDT, G., ARUSCAVAGE,P. & MILLARD, H., Jr 1972. Correlation of the Bishop ash, a Pleistocene marker bed, using instrumental neutron activation analysis. Journal of Sedimentary Petrology, 42, 301-306. BOULTON, G. S., PEACOCK, J. D. & SUTHERLAND, D. G. 1991. Quaternary. In: CRAIG, G. Y. (ed.), Geology of Scotland, The Geological Society, London, 503-543. BOYLE, E. A. & KEIGWIN, L. 1987. North Atlantic thermohaline circulation during the past 20,000 years, linked to high latitude surface temperature. Nature, 330, 35-40. BRAMLETTE, M. N. & BRADLEY,W. H. 1941. Geology and Biology of North Atlantic deep-sea cores between Newfoundland and Ireland. United States Geological Survey Professional Paper, 196A, 1-34. BROECKER, W. S. & DENTON, G. H. 1989. The role of ocean-atmosphere reorganizations in glacial cycles. Geochimica et Cosmochimica Acta, 53, 2465-2501. - - , ANDREE, M., WOLELI, W., OESCHGER, H., BONANI, G., KENNETT, J. & PETEET, D. 1988. The chronology of the last deglaciation: Implications to the cause of the Younger Dryas event. Paleoceanography, 3, 1-19. - - , KENNETT, J. P., TELLER, J., TRUMBORE, S., BONANI, G. & WOLELI, W. 1989. The routing of the Laurentide ice-sheet meltwater during the Younger Dryas cold event. Nature, Cox, K. G., BELL, J. D. & PANKHURST,R. J. 1979. The Interpretation of Igneous Rocks. George Allen & Unwin, London. CROSS, W., IDDINGS, J. P., PIRSSON, L. V. & WASHINGTON, H. S. 1903. Quantitative Classification of Igneous Rocks. University of Chicago Press, Chicago. DUGMORE, A. J. 1989. Icelandic volcanic ash in Scotland. Scottish Geographical Magazine, 105, 168-172. DUPLESSY, J. C., DELIBRIAS,G., TURON, J. L., PUJOL, C. & DOPRAT, J. 1981. Deglacial warming of the northeastern Atlantic Ocean: Correlation with the palaeoclimatic evolution of the European continent. Palaeogeography, Palaeoclimatology, Palaeoecology, 35, 121-144. FAIRBANKS, R. G. 1990. The age and origin of the 'Younger Dryas climatic event' in Greenland ice cores. Paleoeeanography, 5, 637-642. FISHER, R. V. 1964. Maximum size, median diameter, and sorting of tephra. Journal of Geophysical Research, 69, 341-355. - & SCHMINCKE,H.-U. 1984. Pyroclastic Rocks. Springer-Verlag, Berlin. GRAHAM, D. K., HARLAND, R., GREGORY, D. M., LONG, D. ~; MORTON, A. C. 1990. The biostratigraphy and chronostratigraphy of BGS Borehole 78/4, North Minch. Scottish Journal of Geology, 26, 65-75. GRONVOLD, K. & JOHANNESSON,H. 1984. Eruption in Grimsvftn 1983; course of events and chemical studies of the tephra. J6"kull, 34, 1-10. HARKNESS, D, D. 1983. The extent of natural 14C

247

deficiency in the coastal environment of the United Kingdom. In: MOOK, W. G. & WATERBOLK, H. T. (eds) Proceedings of the First Symposium on 14C and Archaology. PACT 8, Council of Europe, Strasbourg. HEDGES, R. E. M., HOUSLEY, R. A., LAW, I. A., PERRY, C. & HENDY, E. 1988. Radiocarbon dates from the Oxford AMS System: Archaeometry Datehst 9. Archaeometry, 30, 291-305. HUNT, J. B. 1992. The Saksunarvatn Tephra: A Reassessment of the Distribution and Importance of an Early Holocene Isochron. In: GEIRSDOTTIR, A., NORDDAHL, H. & HELGADOTTIR, G. (eds) Abstracts: 20th Nordic Geological Meeting, Reykjavik 1992. The Icelandic Geoscience Society and the Faculty of Science, University of Iceland, Reykjavik, 133. - t~r HILL, P. G. 1993. Tephra geochemistry: A discussion of some persistent analytical problems. The Holocene, 3, 271-278 - & - 1994. Geochemical data in tephrochronology: A reply to Bennett. The Holocene, 4, 436-438. IMSLAND, P. 1978. The Petrology of Iceland: Some general remarks. Nordic Volcanological Institute, 7 8 0 8 , 26 INGOLFSSON, O. 1991. A review of the Late Weichselian and early Holocene glacial and environmental history of Iceland. In: MAIZELS, J. K. & CASELDINE, C. J. (eds) Environmental Changes in Iceland, Past and Present. Kluwer, Dordrecht, 13-29. JAKOBSSON, S. P. 1972. Chemistry and distribution of recent basaltic rocks in Iceland. Lithos, 5, 365386. - 1979. Petrology of recent basalts of the eastern volcanic zone, Iceland. Acta Naturalia Islandica, 26, 103. JOHNSEN, S. J., CLAUSEN, H. B., DANSGAARD, W., ET AL. 1992. Irregular glacial interstadials recorded in a new Greenland ice core. Nature, 359, 311-313. KARPUZ, N. C. & JANSEN, E. 1992. A high resolution diatom record of the last deglaciation from the S.E. Norwegian Sea: Documentation of rapid climatic changes. Paleoceanography, 7, 499-520. KITTLEMAN, L. R. 1979. Tephrochronology by microprobe glass analysis. In: SHEETS, P. D. & GRAYSON, D. K. (eds) Volcanic Activity and Human Ecology. Academic Press, London, 49-82. KVAMME, T., MANGERUD, J., FURNES, H. & RUDDIMAN, W. F. 1989. Geochemistry of Pleistocene ash zones in cores from the North Atlantic. Norsk Geologisk Tiddskrift, 69, 251-272. LARSEN, G. 1982. Gjoskutimatel Jokuldals og nagrennis. In: THORARINSDOTrlR, H., OSKARRSON, O. H., STEINTHORSSON,S. & EINARSSON,Th. (eds) Eldur er i NorSri. Reykjavik, S6guf61ag, 5165. LEHMAN, S. J. & KEIGWIN, L. D. 1992. Sudden changes in North Atlantic circulation during the last deglaciation. Nature, 356, 757-762. LONG, D. & MORTON, A. C. 1987. An ash fall within the Loch Lomond Stadial. Journal of Quaternary Science, 2, 97-101.

248

J.B. HUNT ET AL.

MANGERUD, J., LIE, S. E., FURNES, H., KRISTIANSEN, I. L. & LOMO, L. 1994. A Younger Dryas ash bed in western Norway, and its possible correlations with tephra in cores from the Norwegian Sea and the North Atlantic. Quaternary Research, 21, 85104. MERCER, J. H. 1969. The Allerod oscillation: A European climatic anomaly? Arctic and Alpine Research, 6, 227-236. MERKT, J., MULLER, H., KNABE, W., MI]LLER, P. & WEISER, T. 1993. The early Holocene Saksunarvatn tephra found in Lake sediments in NW Germany. Boreas, 22, 93-100. NORODAHL, H. & HAFLIOASON, H. 1992. The Sk6gar tephra, a Younger Dryas marker in North Iceland. Boreas, 21, 23-41. PEACOCK, J. O., AUSTIN, W. E. N., SELBY, I., GRAHAM, D. K., HARLAND, R. & WILKINSON, I. P. 1992. Late Devensian and Flandrian palaeoenvironmental changes on the Scottish continental shelf west of the Outer Hebrides. Journal of Quaternary Science, 7, 145-161. PERSSON, C. 1966. Ffrs6k till tefrokronologisk datering av N~tgra Svenske Torvmossar. Geoliska F6reningens i Stockholm F6rhandlingar, 89, 181197. 1971. Tephrochronological investigation of peat deposits in Scandinavia and on the Faroe Islands. Sveriges Gelogiska Unders6kning, 65, 241-266. RAM, M. & GAYLEY, R. I. 1991. Long range transport of volcanic ash to the Greenland ice sheet. Nature, 349, 401-404. RUDDIMAN, W. F. & GLOVER, L. K. 1982. Mixing of volcanic ash zones in subpolar North Atlantic Sediments. In: SCRUTrON, R. A. & TALWANI, M. (eds) The Ocean Floor, John Wiley & Sons, Chichester. - & MClNTYRE, A. 1981. The North Atlantic Ocean during the last deglaciation. Palaeogeography, Palaeoclimatology, Palaeoecology, 35, 145214. SALMI, M. 1948. The Hekla ashfalls in Finland.

Comptes Rendus de la Soci~t~ Gdologie de Finlande, 21, 87-96

SEJRUP, H. P., SJOHOLM, J., FURNES, H., BEYER, I., EIDE, E., JANSEN., E. & MANGERUD, J. 1989. Quaternary tephrochronology on the Iceland

Plateau, north of Iceland. Journal of Quaternary Science, 4, 109-114. SELBY, I. 1989. Quaternary geology of the Hebridean continental margin. PhD thesis, Nottingham University, UK. STOKER, M. S. 1988. Pleistocene ice-proximal glaciomarine sediments in boreholes from the Hebrides Shelf and Wyville-Thomson Ridge, NW U K Continental Shelf. Scottish Journal of Geology, 24, 249-262. , HARLAND, R., MORTON, A. C. & GRAHAM, D. K. 1989. Late Quaternary stratigraphy of the North Rockall Trough and Faeroe-Shetland Channel. Journal of Quaternary Science, 4, 211222. STOKES, S., LOWE, D. J. & FROGGATT, P. C. 1992. Discriminant function analysis and correlation of Late Quaternary rhyolitic tephra deposits from Taupo and Okataina volcanoes, New Zealand, using glass shard major element composition. Quaternary International, 13/14, 103-117. STRONG, G. E. 1987. Petrographical notes on marine core specimens from glacial deposits from U K continental shelf localities south of St Kilda.

British Geological Survey Report No. SRG/87/6 (unpublished). SWEATMAN,T. R. • LONG, J. V. P. 1969. Quantitative electron-probe microanalysis of rock-forming minerals. Journal of Petrology, 10, 332-379. THORARINSSON, S. 1944. Tefrokronologiska studier p~ Island. Geogafisker Annaler Stockholm, 26, 1-127. - 1967. The Tephra-Fall from Hekla on March 29th, 1947. In: EINARSSON,Th., KJARTANSSON,G. & THORARINSSON,S. (eds), The Eruption of Hekla 1947-1948. Societas Scientatis Islandica, II, 3, 168. VILMUNDARDOTTIR,E. 1977. Tungnfirhraun. Jardfraediskyrsla. National Energy Authority, OS ROD 7702 (mimeogr.), 156. WALKER, G. P. L. 1971. Grain-size characteristics of pyroclastic deposits. Journal of Geology, 79, 696714 WESTGATE, J. A & GORTON, M. P. 1981. Correlation techniques in tephra studies. In: SELF, S. & SPARKS, R. S. J. (eds), Tephra Studies, Dordrecht, Reidel, 73-94.

Tertiary structuration and erosion of the Inner Moray Firth K. T H O M S O N 1 & R. R. H I L L I S 2

1Department o f Geology and Geophysics, University of Edinburgh, Grant Institute, Kings Buildings, West Mains Road, Edinburgh EH9 3JW, UK Present address: Department o f Earth Sciences, University o f Oxford, Parks Road, Oxford 0)(1 3PK, UK 2Department o f Geology and Geophysics, University o f Adelaide, GPO Box 498, Adelaide, South Australia 5001, Australia Abstract: Seismic profiles and field data show that the Inner Moray Firth (IMF) experienced

significant structural modification during Early Tertiary times with the development of inversion, strike-slip and extensional oblique-slip geometries as well as uplift and erosion at a mid-late Danian unconformity. Seismic reflection profiles across the IMF also show progressively older stratigraphic subcrop towards the west. Analysis of sonic velocities and vitrinite reflectance demonstrate that up to 1.5 km of basin fill has been removed from the IMF. The height of the sequences above maximum burial depth (apparent erosion) is at a maximum in the northwestern part of the basin, where inversion geometries are found, and decreases to zero in the Outer Moray Firth. However, if post-erosional burial is taken into account, the actual amount of erosion during Early Tertiary exhumation (total erosion) is shown to be more evenly distributed, and of greater magnitude throughout the IMF. Incorporation of the effects of Tertiary erosion into analysis of basin development requires much greater post-rift burial than if Tertiary erosion is ignored. It seems most likely that the Early Tertiary deformation of the IMF occurred in response to NE Atlantic (Thulean) and Alpine events. Late Cretaceous-Early Tertiary times saw the change of the predominant stress field of North West Europe from one of extension to one of NW-SE compression (Ziegler, P. A. 1987, Tectonophysics, 137, 1-5 & 389-420). The present-day stress field of Western Europe is described by a NW-SE direction of maximum principal stress (MOiler, B. et al. 1992, Journal of Geophysical Research, B97, 11783-11803). The present-day-, and by inference palaeo-, intraplate stress field of Western Europe can be attributed to plate-driving forces acting on the boundaries of the Eurasian plate. On average, the orientation of present-day maximum stress in western Europe is subparallel to the direction of relative plate motion between Africa and Europe (Moiler, B. et al. 1992, Journal of Geophysical Research, B97, 11 783-11 803). A combination of stresses associated with Alpine collision between Europe and Africa, and those associated with opening of the North Atlantic are considered responsible for the Late Cretaceous-Early Tertiary NW-SE compressional stress field of North West Europe.

This paper addresses the effect of the changing Mesozoic-Cenozoic stress field on the development of the Inner Moray Firth (IMF) Basin. Most discussions of the evolution of the I M F have not addressed the Tertiary evolution of the I M F in any detail (McQuillin et al. 1982; Barr, 1985; Andrews & Brown, 1987; Bird et al. 1987; Frostick et al. 1988, Andrews et al. 1990, Roberts et al. 1990). Tertiary extensional, strike-slip and compressional tectonism within the I M F are described in this paper. Furthermore, compaction and vitrinite reflectance data that suggest Early Tertiary erosion of c. 1 km occurred throughout the IMF. Not only is this period of tectonism and erosion intrinsically

important to understanding the evolution of the basin, but proper allowance for Tertiary erosion significantly changes the picture of earlier, Mesozoic basin development.

Tertiary structural reactivation Basin inversion Most discussions of the evolution of the I M F ignore the existence of compressional structures (McQuillin et al. 1982; Barr, 1985; Andrews Brown, 1987; B i r d et al. 1987; Frostick et al. 1988, Andrews et al. 1990; Roberts et al. 1990). However, Underhill (1991 a, b) and Thomson &

From Scrutton, R. A., Stoker, M. S., Shimmield, G. B. & Tudhope, A. W. (eds), 1995, The Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region, Geological Society Special Publication No. 90, pp. 249-269

249

250

K. THOMSON & R. R. HILLIS

.~

~

59N

inversion and the associated faults cut through to sea bed of Cretaceous age the inversion must have occurred after Cretaceous times.

E x t e n s i o n a l reactivation

//v//2-~_

4W

3W

2W

1W

Fig. 1. Map of the Inner Moray Firth showing the major structural elements discussed in the text and the locations of the figures. PG, Portgower; LP, Lothbeg Point; BR, Brora.

Underhill (1993) have documented the presence of inversion structures close to the intersection of the Great Glen and Wick faults (Figs 1 & 2). The inversion structures consist of a hangingwall anticline and an associated 'short-cut' fault. The anticline formed as a result of N W - S E compression, consistent with the Tertiary stress regime (England 1988), expelling the basin-fill up the main fault. As compressional stresses increased, buttressing against the footwall resulted in the development of the 'short-cut' fault. As Cretaceous sediments are affected by the

There are two distinct styles of Tertiary extensional faulting (Figs 3 & 4). While Tertiary extensional faulting is widespread, many faults are typical of those due to differential compaction (Fig. 3; Prosser, 1991; Thomson & Underhill, in press). Hillis et al. (1994) have shown that these structures were c. 1 km more deeply buried during latest Cretaceous-earliest Tertiary times than they are at present. Hence, differential compaction across half-graben-bounding faults such as that illustrated in Fig. 3, may have been significant. A second population of extensional faults (Fig. 4) form the Sinclair Horst (Fig. 1). Isopach maps demonstrate that these faults are neither syn-rift nor compaction related. They displace syn-rift and thermal subsidence sediments by a similar amount. These faults must have been active in post-Cretaceous times, but there is no evidence for an earlier history. Hence, these faults probably formed as a new population of post-Cretaceous faults. The post-Cretaceous age of the faults is consistent with the age of the inversion structures in the northwest. The (re)activated extensional structures trend W S W ENE, an orientation suitable for dextral oblique

Fig. 2. Seismic line showing an inversion-related hanging wall anticline and associated short-cut fault in the region of the Wick-Great Glen Fault intersection. BCRT, base Cretaceous.

TERTIARY STRUCTURATION AND EROSION, INNER MORAY FIRTH

251

Fig. 3. Seismic lines showing a syn-sedimentary halfgraben-bounding fault with evidence of post-Cretaceous activity and compaction related geometries. BCRT, base Cretaceous; TTR, top Triassic. extension in the N W - S E compressive, N E - S W extensional Tertiary stress regime. Strike'slip reactivation

The most recent significant phase of movement on the onshore Great Glen Fault (GGF) appears to have occurred during Tertiary times (Holgate 1969; Speight & Mitchell 1979; Rogers et al. 1989). Offshore, seismic reflection profiles show that the G G F consists of several strands (Fig. 5), displaying geometries indicative of strike-slip motion (Naylor et al. 1986; Harding 1990). Isopach maps of the offshore area suggest strikeslip movement was probably small and dextral (Thomson & Underhill 1993), consistent with evidence from onshore (Holgate 1969; Speight & Mitchell 1979; Rogers et al. 1989). Movement probably occurred during Early Tertiary times as reflection profiles show the G G F displacing Cretaceous sediments in the offshore IMF in a manner consistent with that observed onshore. Strike-slip reactivation probably affected the Helmsdale Fault during Early Tertiary times (Thomson & Underhill 1993). In the Sutherland Terrace, the area between the Great Glen and Helmsdale Faults, numerous folds can be found on both a seismic and field scale, plunging SE, at high angles (50~ to the Helmsdale Fault (Fig. 6). The orientation of these folds is inconsistent

Fig. 4. Post-Cretaceous-Early Tertiary extensional faults forming the Sinclair Horst. The faults show no evidence of syn-sedimentary or compaction related geometries. BCRT, base Cretaceous; TTR, top Triassic: MOX, middle Oxfordian.

252

K. THOMSON & R. R. HILLIS

Fig. 5. Seismic lines across the Great Glen Fault. The helicoidal and flower structure geometries shown are indicative of strike-slip movement, probably during the Early Tertiary. Abbreviations as in Fig. 4. with their origin as 'classical' inversion folds. These folds may relate to Tertiary dextral and sinistral strike-slip motion on the GGF and the Helmsdale Fault, respectively (Thomson & Underhill 1993). It is suggested that the opposing slip senses on the faults 'pushed' the Sutherland Terrace into a smaller area to the northeast. This resulted in a local compressive stress orientation of N E - S W and the generation of N W - S E trending folds (Fig. 6).

Regional tilting Both seismic profiles (Fig. 7) and geological maps of the sea bed show that the stratigraphy

at sea bed in the IMF becomes progressively younger eastward. Tertiary units are only present at the Outer Moray F i r t h - I M F boundary at c. 2 ~ (Fig. 7). Seismic evidence combined with compaction analysis (Hillis et al. 1994) and vitrinite reflectance data show that the pre-Cenozoic units are no longer at their maximum burial depth and that the outcrop pattern is due to uplift and erosion during the mid-late Danian, followed by gentle burial during the Tertiary, the magnitude of which increased to the west. The evidence for regional Early Tertiary erosion in the I M F is discussed in the following section.

TERTIARY STRUCTURATION AND EROSION, INNER MORAY FIRTH

253

Fig. 6. (a) Seismic line showing the large-scale folding in the Sutherland Terrace; abbreviations as in Fig. 4. (b) Contoured stereographic projection of the orientation of fold axes from the onshore exposures. Note that fold axes trend NW-SE. (e) Opposing slip senses on the Helmsdale and Great Glen Faults resulted in local NE-SW compression and the generation of NW-SE trending folds.

Evidence for Tertiary erosion Analyses of sediment compaction, vitrinite reflectance and apatite fission tracks all suggest that the IMF was subject to regional erosion of c. 1 km in Early Tertiary times. The compaction and vitrinite reflectance data are presented in this section of the paper. S e d i m e n t compaction

The approach to compaction-based analysis of erosion magnitude follows that of Hillis et al. (1994) where it is discussed in more detail. However, in this study a further 11 wells (some of which remain commercial-in-confidence) have

been added to those studied by Hillis et al. (1994). Since depth-controlled compaction of sediments is largely irreversible, exhumed formations will be overcompacted with respect to their present burial depth (e.g. Magara 1976; Lang 1978; Bulat & Stoker 1987). The amount of erosion of such overcompacted sediments above their maximum burial depth is given by the displacement, along the depth axis, of the observed compaction trend from the normal (undisturbed) trend (Fig. 8). In this study sonic velocity is taken to represent compaction state. Stratigraphically-equivalent units, which exhibit a vertically- and laterally-consistent relationship between depth and compaction, are

254

K. T H O M S O N & R. R. HILLIS

Fig. 7. Seismic reflection profiles showing tilted, eastward-dipping pre-rift, syn-rift and thermal subsidence deposits. Note that progressively younger units crop out at, or near, sea bed from west to east. TCHK, top Chalk; BCHK, base Chalk; other abbreviations as in Fig. 4.

TERTIARY STRUCTURATION

AND EROSION, INNER MORAY FIRTH

SONIC S L O W N E S S 80 1 2 0,,

40

i

(ps/ft) 160

i

i

E

s

500

SJ

"1" I--- 1 0 0 0 13LU r~ 1500

A

,

1D

~

SONIC SLOWNESS 80 120 i

w

B.J

500 E "lF- 1 0 0 0 n ILl

A v

2000

J

v

x...rt 1 |

40

1500

J

B

A s

,,P s s

255

(ps/ft) 160

C

D

!

E -.-~

_-_-. ~ .-% .-% -_~ ~176

1

I,

. . L - - 1.

-_-.~. -~176 s -~176 %-.-~176 ~ ~176

S

/

S S

-i:i: | 1 ! ! 1 9

! ! 1 !

E

Fig. 8. Porosity evaluation in a Chalk sequence during burial (A), subsequent uplift (B), and post-uplift burial (CE). The apparent erosion is the amount of erosion (EA) not reversed by subsequent subsidence. It can be measured by the displacement along the depth axis of the porosity-depth (or in this case sonic slowness~lepth) relation of the uplifted sequence (B or C) from that of a reference or normally compacted sequence (A, D or E).

required for such analysis. Sonic log data indicate that, in the IMF, only the Chalk and the Kimmeridge Clay exhibit a reasonably consistent increase in velocity with depth (Fig. 9). The absence of significant bulk lateral facies variation within the Chalk and Kimmeridge Clay suggests that depth--compaction relations should also be laterally consistent. The Chalk was divided into the Hidra, Plenus Marl, Hod and Tor Formations of Deegan, & Scull (1977) on the basis of sonic and gammaray log character. The Hidra and Plenus Marl Formations show variable shale content and were not used in this study. Although the sonic character of the Hod and Tor Formations is internally consistent, a log discontinuity marks their mutual boundary and the two formations were analysed separately. The Kimmeridge Clay

Formation was defined following Deegan & Scull (1977). The tops and bases of the Kimmeridge Clay, Hod and Tor Formations were picked from sonic and gamma-ray logs (Fig. 9). The mean slowness (reciprocal of velocity) of the resultant intervals was determined from sonic log data. Table 1 lists the mean slowness and depth to formation mid-point (below sea bed) for the Kimmeridge Clay, Hod and Tor Formations in the wells used in this study (confidential data have been excluded). The well with the lowest velocity (highest slowness) for its burial depth, with allowance for the mean slowness-depth gradient (as determined by Hillis et al. 1994), was taken as the reference well from which apparent erosion for the other wells was calculated. Well 13/30-3 is the reference well for

o~t 0

0

d ~_

oOt

o

O0

~,,

rr W

OOl

W

~

o

~

t-tll

o7

rr

-~ ~

o

~

o

~

O,J

II

OOl

o ~

"~

rr

~

~ 0

Z

..4

W

~r-

0

o ~ .~

~: w

<

~v o~

w

zO

..j

/

w

d

~

Z...~

0

r~

~3

e~

n

0 o

OO-i 00

>

-~

N

C~

W

~

cc

~

w ~

l ~ , , . , ~ - - . i . l ~ , ~ , , ~

0

0 0 0

~

N

~

o 0

o

(W) NZcF:IO

o

o~.~

o

o ~

0

TERTIARY STRUCTURATION AND EROSION, INNER MORAY FIRTH

257

Table 1. Mean slowness,formation mid-point and apparent erosion data for the Inner Moray Firth KimmeridgeClay Fm Well

Mean FormationApparent slowness midpoint erosion ~s/ft) (m bsb) (m)

11/25-1 11/30-1 11/30-2 11/30-3 11/30-4 11/30-5 12/21-2 12/21-3 12/21-5 12/22-2 12/23-2 12/24-2 12/27-1 12/27-2 12~28-1 12129-1 13111-1 13q2-1 13 q2-2 13q3-1 13114-1 13q7-1 13'18-1 13q9-1 13q9-3 13122-4 13 '27-1 13 t28-2 13 '28-3 13 '29-1 13 '29-2 13 ~29-3 13 t30-1 13 '30-2 13 ~30-3

88.2 106.6 97.1 99.7 100.9 99.7 95.5 88.5 93.4 98.3 94.8 79.3 123.8 119.7 91.5 91.3 72.5 80.2 95.2 96.6

1400 1333 1159 1204 1159 1210 1280 1516 1406 855 1083 2128 744 1044 1419 1461 2760 2352 1377 1333

1099 577 1056 926 934 921 986 972 926 1296 1204 654 617 449 975 937 239 402 897 894

90.0 129.6

1560 1054

895 105

86.1 85.6 88.9 87.0 83.6

1716 2241 2115 2241 2500

849 340 361 297 146

73.3 80.8 78.8

2932 2430 2798

41 303 0

Hod Fm

Tor Fm

Mean FormationApparent slowness midpoint erosion (#s/ft) (m bsb) (m)

Mean Formation Apparent slowness midpoint erosion (#s/rt) (m bsb) (m)

80.6

669

693

75.9

445

694

83.0 79.7 88.0 86.2 86.0 73.7 72.6 75.9 71.8 70.7 71.7 65.6 72.8 68.4 70.9 74.1

715 825 710 651 691 942 1129 948 1205 1183 1254 1293 1264 1518 1478 1533

582 561 455 561 527 601 443 537 389 440 342 465 304 164 138 0

81.0 82.4 84.0 96.8 82.0 71.7 67.0 68.7 63.1 64.8 64.1 65.8 64.5 62.8 61.7 68.5

499 580 476 478 634 802 863 735 970 959 995 1028 1038 1316 1236 1277

539 430 503 261 385 419 451 545 420 399 375 311 325 80 182 0

Average slowness-depth gradients are: -3.121 x 10-2, --6.138x 10 -2 and -3.833 • Hod and Tor Formations, respectively, bsb below sea bed; Fm, formation. all three units studied. Apparent erosion (EA) is defined as the displacement on the depth axis of the mean slowness-depth gradient between the mean slowness value of the reference well and the well under consideration. Apparent erosion was determined numerically using the simple equation:

EA=

1 - d u + dR, m ( A t u - AtR)

where m is the mean gradient of the slownessdepth relation, A t o and AtR are the mean slownesses of the well under consideration and the reference well, respectively, and du and dR

10 -2

for the Kimmeridge Clay,

are the depths of the formation mid-points (below sea bed) of the well under consideration and the reference well respectively. The resultant apparent erosion values are given in Table 1 and have been contoured in Fig. 10. If the reference well is above its maximum burial depth then the apparent erosion magnitudes determined in this study will be consist e n t l y u n d e r e s t i m a t e d by the a m o u n t o f apparent erosion in the vicinity of the reference well (Fig. 8). Apparent erosion values from the Kimmeridge Clay, H o d and Tor F o r m a t i o n s were plotted against each other in order to check their consistency (Fig. 11; cf. Bulat & Stoker

258

K. THOMSON & R. R. HILLIS ~^oL,

4~

3~N

the shallower formation (Table 2). The t-statistic of the coefficients of correlation were calculated and tested against the one-tailed Student's tdistribution in order to determine whether the coefficients of correlation were significant (e.g. Till 1974). There is < 0.05% chance that the coefficient of correlation between apparent erosion values determined from the Hod and Tor Formations comes from a population of coefficients with a mean value of zero. This probability is < 0.5% in the case of the apparent erosion values from the Kimmeridge Clay and Tor Formations and 5~ from the Kimmeridge Clay and Hod Formations (Table 2). It is unlikely that a sedimentological and/or diagenetic mechanism could account for similar amounts of overcompaction in the carbonate and clastic formations analysed. In the absence of an alternative explanation , the fact that the correlation between the results of apparent erosion from the three formations is statistically significant supports the argument that burial at depth beyond that currently observed is responsible for overcompaction in the formations analysed.

2:W

Vitrinite reflectance

4~V

3~W

2:W

4*W

$*W

2~/

1~N

Fig 10. Maps of apparent erosion (in metres), for the Inner Moray Firth based on sonic slowness in the; (a) Tor Formation (b) Hod Formation; and (e) Kimmeridge Clay Formation. Locations of data are shown except for confidential data. 1987; Hillis et al. 1994). Least-squares, best-fit, linear relations between the apparent erosion values and associated coefficients of correlation were determined by regression of apparent erosion values from the deeper formation on

Pearson & Watkins (1983) state that for the majority of wells in the North Sea the interpolated vitrinite reflectance (Ro%) value for the sea-bed is 0.2%. Consequently, this value may be taken as the depositional vitrinite reflectance value for the area and if higher values occur at sea-bed, uplift and erosion may have occurred. The vitrinite reflectance trend method, as used for southern offshore N o r w a y (Jensen & Schmidt 1993) and for the Irish Sea (Hardman et al. 1993; Naylor et al. 1993), relies on the fact that vitrinite reflectance increases irreversibly with time and temperature. At deposition, surface vitrinite reflectance values are believed to be 0.2%. Intersection of the vitrinite reflectance/ depth trend with the 0.2% value should be at zero burial depth in non-uplifted wells and at negative burial depths if uplift and erosion has occurred (Fig. 12). The amount of section removed from the location of a well is equal to the negative burial depth at the surface vitrinite reflectance value of 0.2~ For the quantification of Neogene uplift in southern offshore Norway, the vitrinite reflectance estimates of Jensen & Schmidt (1993) compared favourably with their estimates derived from compaction based studies similar to those previously described. This section applies the same methodology to the estimation of uplift and erosion from the vitrinite reflectance data of the ten offshore

TERTIARY STRUCTURATION AND EROSION, INNER MORAY FIRTH 1000- HOD - 84.678 + 0.91193TOR C,C. = 0.949 N = 17

259

J

800o.. o Q.U)

600

9


nI~B. I/,~i

o E~

9

400

Vltrlnite reflectance (Ro%):

200~

0

2;0

0

400

600

600

1000

1 -

1000

TOR FM APPARENT EROSION (METRES)

J:

KIM = 49.561 + 0.94731HOD C.C. 0,535 N := 12

800

>.~ 5= o_o

B

600"

on-

3~

A

400" ~UJ ,<

200"

200

0

400

800

800

10'00

HOD FM APPARENT EROSION (METRES)

1000-

~

800

~

600

~.~ ~

4o0

~.,~ :0;5:9295 § ,,948To.

/

200"

J 0

/

i R 200

,.. 400

600

800

1000

TOR FM APPARENT EROSION (METRES)

Fig. 11. Cross-plots of apparent erosion (metres) derived from; (a) Hod and Tor Formations, (b) Kimmeridge Clay Formation and Hod Formations and (e) Kimmeridge Clay and Tor Formations. The least-squares, best-fit, linear relations between the apparent erosion values are plotted and quoted on each plot. The number of data points (N) and the coefficient of correlation between apparent erosion values (C.C.) are also quoted.

Fig. 12. Determination of apparent uplift magnitude from vitrinite reflectance data. Burial depth-vitrinite reflectance relations regress to linear equations. For wells at their maximum burial depth (A) the trend line crosses the surface deposition value (0.02%) at the surface. Uplifted wells (B) have trend lines which intersect the surface depositional value at negative burial depths. The magnitude of erosion is equal to the magnitude of negative burial depth at the surface depositional value.

wells contained within Prajoga (1990) and four onshore localities given in Scotchman (1991). In order for the vitrinite reflectance trend method to be applied to the Inner Moray Firth it must first be established that vitrinite reflectance increases with depth. In practice, current burial depth was plotted against log10 vitrinite reflectance and a best-fit, least-squares, linear regression of burial-depth against log]0 vitrinite reflectance was produced (Fig. 13). For each well, the a m o u n t of section removed was calculated using the equation of the general form: E = - ( m log]00.2 + c) where E is the amount of section missing; rn is the gradient of the burial depth/logl0 vitrinite reflectance relation; c is the intercept of the burial depth/vitrinite reflectance relation on the

260

K. THOMSON & R. R. HILLIS

o

t

I

!

I

|

9

o

0

,

k~ 0

9

9

Ill

2

,JD

!

!

,

.G 12.

e

CI

'~ ..~

TERTIARY STRUCTURATION AND EROSION, INNER MORAY FIRTH

261

Table 2. Correlation between apparent erosion results (EA)from the Kimmeridge Clay, Hod and Tot Formations Formations

No of wells with both formations

Hod/Tor Kimmeridge/Tor Kimmeridge/Hod

17 12 12

cc between EA values from the two formations

t-Statistic

0.849 0.734 0.535

4.07 3.50 2.0

Confidence level at which Ho can be rejected 95.00% Y Y Y

99.50% Y Y N

99.95% Y N N

Ho is the null hypothesis that the correlation coefficient (cc) is from a population the mean of which is zero.

Table 3. Regression data for burial depth v. vitrinite reflectance for wells in the Inner Moray Firth Well

11/30-2 11/30-6 12/21-1 12.22-1 12/23-1 12/23-2 12/26-2 12/27-1 12/28-1 12/30-1

Gradient of burial depth/loglo vitrinite reflectance relation

Intercept on burial depth axis of the burial depth/log10 vitrinite reflectance relation

Correlation coefficient for burial depth/log~0 vitrinite reflectance regression

5943 5354 6413 5566 4746 4238 3284 5073 6305 5825

- 3424 - 2890 -4175 - 3814 - 3004 - 2606 - 2423 - 3581 -4478 - 3698

0.918 0.802 0.871 0.933 0.962 0.955 0.912 0.951 0.881 0.824

Uplift and erosion estimate (m) 730 850 310 80 310 360 - 130 - 40 - 70 370

The estimates of erosion derived from the burial depth-vitrinite reflectance data are also shown.

burial-depth axis. The uplift and erosion estimates are shown in Table 3. For the onshore localities the approach is slightly different. As all the onshore data comes from the surface measurements it is impossible to produce a burial depth-vitrinite reflectance relation. Consequently, the burial depth-vitrinite reflectance relation for the combined data from wells 11/30-2 and 11/30-6 (Table 4) was used to calculate the maximum burial depth for the average vitrinite reflectance value at each locality. The two wells were chosen as they are the closest to the onshore localities. As the data points are currently at the surface, the maximum burial depth equals the a m o u n t of section removed and was calculated using the equation: E = (m lOgl0Ro(mean)) + c where E is the amount of section missing; rn is the gradient of the burial depth/log10 vitrinite reflectance relation; Ro(mean) is the mean vitrinite reflectance value for the outcrop locality; c is the intercept of the burial depth/vitrinite reflectance relation on the burial-depth axis. The data,

equation details and erosion estimates are shown in Table 4. The results contained in Tables 3 and 4 suggest that the I M F experienced uplift and erosion. However, compared to the results from sediment compaction the results are less internally consistent. For the offshore data a wide range of uplift and erosion estimates were produced, ranging between - 1 3 0 and 850m. However, the majority of the results have the same order of magnitude as the compaction results. Wells 11/30-2 and 11/30-6 give results directly comparable with those derived from compaction, with values around 700-900m. Four other wells (12/21-1, 12/23-1, 12/23-2 and 12/30-1) yield lower values around 300m, but still with the same order of magnitude as results from compaction analysis. The remaining four wells (12/22-1, 12/26-1, 12/27-1 and 12/28-1)are substantially different from the compaction evidence with values near zero and even negative and it appears that for these wells at least the method breaks down. For the onshore data the results are consistent with both the compaction and apatite fission track analyses and conse-

262

K. THOMSON & R. R. HILLIS

Table 4. Onshore vitrinite reflectance data and erosion estimates derived from it Location

Vitrinite reflectance values (Ro) (%)

Average vitrinite reflectance (Ro) (%)

Apparent erosion estimate (m)

Ballintore/Ethie Brora Lothbeg Point Portgower

0.38 0.26 0.35 0.36 0.43 0.41 0.45, 0.49

0.38 0.35 0.41 0.47

840 620 1030 1390

Apparent erosion estimated using the burial depth-vitrinite reflectance relation for the combined data of wells 11/ 30-2 and 11/30-6 (m = 5.963, c = 3343).

quently some confidence could possibly be placed in them. On balance it appears that the vitrinite reflectance trend method, although more erratic in this study than for Jensen & Schmidt (1993), provides some supportive evidence for uplift and erosion in the Inner Moray Firth with a magnitude of hundreds of metres.

Implications of Tertiary erosion Uplift, erosion, apparent erosion and m a x i m u m burial depth

England & Molnar (1990) highlighted the distinction between three different measures of 'uplift': uplift of the surface with respect to the geoid; uplift of rocks with respect to the surface (i.e. erosion); and the displacement of rocks with respect to the geoid. Of these three quantities only surface uplift is associated with work done against gravity, hence this is the relevant value for studies of tectonic controls on uplift (England & Molnar 1990). Previous authors (e.g. Bulat & Stoker 1987; Hillis 1991) have used the term 'apparent uplift' to describe the value determined from studies of compaction. The term 'apparent erosion', as used by Hillis et al. (1994), is now preferred in view of the distinctions highlighted by England & Molnar (1990). Apparent erosion is equal to the height above maximum burial depth (with respect to sea bedland surface) of a given stratigraphic unit (Fig. 8). It is not necessarily the same as the amount of erosion that occurred when the unit was exhumed. If renewed burial follows erosion the magnitude of apparent erosion is reduced by the amount of that subsequent burial. Although overburden weight caused by burial following erosion does not cause further compaction until a unit exceeds its greatest burial depth, the apparent erosion becomes smaller as the unit approaches its maximum burial depth (Fig. 8; Hillis 1991; Hillis et al. 1994). Consequently, the

,

j/ .-

J J

$~

2"W

1"W

Fig. 14. Total erosion (in metres) map for the Inner Moray Firth based on apparent erosion values from the Kimmeridge Clay Formation. Total erosion (E) equals the apparent erosion (EA) plus any post erosional burial (BE). actual amount of erosion that occurred during exhumation (total erosion) is equal to apparent erosion plus any post-erosional burial. Total erosion magnitudes for the IMF, assuming midlate Danian uplift and based on apparent erosion values for the Kimmeridge Clay Formation and Tertiary sediment thickness are shown in Fig. 14. Since only regional uplift of the Earth's surface with respect to the geoid (as opposed to erosion or the uplift of rocks with respect to the geoid) demonstrates work done against gravity, it is necessary to prove regional surface uplift to invoke isostatic forces requiring uplift (England & Molnar 1990). Since marine, or near-marine, sediments were deposited before and after the mid-late Danian period of erosion, it seems reasonable to infer that erosion was associated with uplift of rocks with respect to the geoid (which here approximates sufficiently well to sea level), as opposed to base-level change. Erosion, hence the inferred uplift of rock, in the I M F occurred over some 104km 2, and the

TERTIARY STRUCTURATION AND EROSION, I N N E R MORAY FIRTH

263

AGE (Ma) 200

300

100

I

0

0

I

',~ l ~ %.. ~

I

I DECOMPACTEDBUR'~ I I (WITHOUTALLOWANCE

/ FORTERTIARY EROSION) I

\ I DECOMPACTEDBURIALI 9'~"..'~ I (WITHALLOWANCEFOR I " "~T".l L TERTIARYEROSlON,

I

I

#

ii

ul i-i (a)

12/27-1

AGE ( M a ) 200

300 0

I

100

~

0

I

"~

I

]OeSERVEDI

A

E "-'1 a 1,1.1

Ill. ~

%. "%.

m

D E ~ P A C T E D BURIALI ~ (WlmOUTALLOWANCE I

~ V

I FORTERTIARYEROSION)I

IAI

~2 O .J w

TERTIARYEROSION) I

m -r-

~

-~'~

1

[3 w

(b)

13/29-1

Fig. 15. Present observed (solid) and decompacted burial histories for the IMF basement in wells (a) 12/27-1 and (b) 13/29-1. The dotted burial plot is corrected for compactional effects without consideration of Tertiary erosion, whereas the dashed plots for both wells have been corrected for compaction with allowance for Tertiary erosion. In both wells the maximum burial depth (B) equals present burial depth (Bp) plus apparent erosion (EA), i.e. B = EA + Be. In 12/27-1, there was no post-erosional burial, apparent erosion (EA) equals the erosion that occurred at the time the rocks were exhumed. However, in well 13/29-1 erosion at the time the rocks were exhumed equals apparent erosion (EA) plus the post-erosional burial (BE), i.e. E = EA + BE.

264

K. THOMSON & R. R. HILLIS

material eroded was not deposited within the IMF. Hence, regional uplift of the surface in mid-late Danian times is inferred.

Implications for burial history The erosion inferred from sonic velocities has a profound effect on the reconstruction of the burial history of the IMF. Firstly, the methodology for sediment decompaction, the procedure by which the original thickness of a stratigraphic unit is calculated (Sclater & Christie 1980), needs to be modified. In areas such as the IMF, where sediments are not at their maximum burial depth, this procedure is more complex than in areas where sediments are at their maximum burial depth. Firstly, a normal porosity-depth relationship will not simply be an average one for the area; it should be determined only from wells at their maximum burial depth. Secondly, a unit must be added to the observed stratigraphy in order to allow for the effect of the eroded sediments. Units deposited prior to the erosion will decompact more (increase in thickness more) than if allowance is not made for the eroded sediments (Fig. 15). Furthermore, if the sequence is not at its maximum burial depth then the units deposited after erosion (in this case the Tertiary) will not cause further compaction of the older units (in this case the pre-Tertiary). In the eastern IMF pre-Tertiary units acted as compaction basement during Tertiary burial associated with deposition of the post-erosional sequence (Fig. 15b). The occurrence of erosion in the IMF has an even more profound effect on the modelling of pre-erosional burial history of the basin than the implications for the sediment decompaction procedure. As the pre-erosion units were at their maximum burial depth prior to exhumation, they must have attained their maximum burial depth before the mid-late Danian. Since apparent erosion is the height above maximum burial depth, the actual magnitude of burial prior to mid-late Danian times equals the observed burial plus apparent erosion. Consequently, late Cretaceous-early Tertiary burial must have been of greater magnitude and more rapid than the preserved stratigraphy suggests (Fig. 15).

Erosion and inversion Compression in the IMF was restricted to the Wick-Great Glen Faults intersection and in the Sutherland Terrace areas. As in the East Midlands Shelf (Green 1989) and the Western Approaches Trough (Hillis 1991) the areas of

crustal compression are coincident with local erosional maxima. However, erosion in the IMF is regional, hence t h e isostatic response to crustal compression and thickening cannot be the sole driving force of uplift. The final section of this paper discusses possible alternative driving mechanisms for the regional uplift.

Driving mechanisms of Tertiary uplift and erosion As discussed above, the regional Early Tertiary erosion in the IMF is considered to be indicative of regional surface uplift. Hence, it is necessary to invoke an isostatic force requiring uplift (England & Molnar 1990). This section discusses possible sources of uplift. If uplifted crust is eroded, isostatic rebound resulting from gravitational unloading will amplify the initial (tectonic) uplift (UT) according to the relation: U E ~- U T -

rm r m -- Ts)

where UE is the total magnitude of erosion, rm is the density of the mantle (c. 3.3 g cm- 3) and rs is the density of the eroded sediment (c. 2.2 g cm-3). Hence, the regional erosion of 1 km in the IMF requires a tectonic uplift of only 1/3 km. It is also important to note that, as discussed above and illustrated by Fig. 16, erosion was preceded by burial to the depth at which the observed compaction-vitrinite characteristics were attained. Any complete model of this period of basin development should also account for this burial prior to erosion.

Compression and lithospheric thickening In the inverted parts of the I M F (around the intersection of the Wick and Great Glen Faults) where there is evidence of Early Tertiary crustal compression, and hence thickening, uplift is readily explained, indeed required, in terms of the isostatic response to crustal-lithospheric thickening (Chadwick 1985; Murrell 1986). In areas where there is no evidence of crustal compression and thickening the driving force for uplift, is more enigmatic.

Thermal uplift associated with North Atlantic rifting Tectonism and uplift in the I M F was synchronous with opening of the North Atlantic between Greenland and Norway/NW Britain.

TERTIARY STRUCTURATION AND EROSION, INNER MORAY FIRTH The North Atlantic margins in this area are volcanic margins and witness tiffing over a region of anomalously hot asthenosphere (now the Iceland Hotspot). The effects of such rifting extensive intrusive and extrusive igneous activity, thickening of the passive continental margin, inhibition of margin subsidence, e t c . have been well documented (Bott 1988; White 1989, 1992; White & McKenzie 1989). Despite the synchroneity, it is not considered likely that regional uplift in the IMF was driven by the buoyancy effects of the North Atlantic thermal plume. The IMF lies beyond the British Tertiary Igneous Province of western Scotland, and hence probably beyond the buoyancy effects associated with the North Atlantic thermal plume (White & McKenzie 1989). Furthermore, uplift due to the effects of the North Atlantic thermal plume could not generate the observed subsidence/burial prior to uplift. Flexural uplift of the IMF in response to the uplift of the area of the igneous province is not considered a likely mechanism because stretched continental crust has very little flexural rigidity (Barton & Wood 1984; Fowler & McKenzie 1989). Furthermore, flexural uplift due to the North Atlantic rift dome could not generate the observed subsidence/burial prior to uplift. -

l n t r a p l a t e stress

The predominant stress field of North West Europe changed from one of extension to one of compression during Late Cretaceous-Early Tertiary times (Ziegler 1987a, b). Intraplate compressional stresses essentially act to amplify existing deflections of the lithosphere and, for example, basin centres are deepened while flanks are uplifted (Cloetingh et al. 1985; Kooi & Cloetingh 1989; Cloetingh et al. 1990). Hence, if the IMF is seen as a flank to the North Sea Rift System intraplate compression could generate tectonic uplift. However, this mechanism is only likely to generate up to 50 m of uplift (Cloetingh et al. 1985; Kooi & Cloetingh 1989; Cloetingh et al. 1990). Furthermore, the IMF is in itself a significant basin, being part of the North Sea Triple Rift System of the Viking GrabenCentral Graben-Moray Firth (e.g. Ziegler 1990, enclosures 37--43). Hence, it is unlikely that the entire IMF would have been uplifted in response to intraplate compressional stresses sensu Cloetingh et al. (1985). The Mesozoic depocentre of the IMF would be expected to subside in response to intraplate compression and its flanks would be expected to be uplifted. Such a pattern does not account for the observed regional uplift of the IMF.

265

M a n t l e lithospheric c o m p r e s s i o n

Regional uplift of a basin which only shows localized inversion (as seen in the IMF) has also been described in the East Midlands Shelf (Green 1989), and in the Western Approaches Trough, offshore SW UK-NW France (Hillis 1991). In order to account for regional uplift of areas where there is only localized evidence of crustal compression and thickening (inversion), Hillis (1992) proposed a decoupled, two-layer model of lithospheric compression (Fig. 16). This model may account for the regional uplift of the IMF. It is clear from the inversion structures in the IMF that the (upper) crust was under compression in Early Tertiary times. Hillis's (1992) model assumes that the entire lithosphere is under compression, but that compression and thickening in the lower lithosphere (possibly equivalent to the mantle part of the lithosphere) is decoupled, and laterally displaced, from that in the upper lithosphere (Fig. 16). Such a twolayer model is directly analogous to the commonly invoked two-layer response of the lithosphere to extension (e.g. Royden & Keen 1980; Hellinger & Sclater 1983; Coward 1986; Kusznir et al. 1987; White & McKenzie 1988). Indeed, the model illustrated in Fig. 16 is based on Kusznir et al.'s (1987) model. Reverse movement on the pre-existing, weak crustal detachments, extension on which had formed the basins, caused the widely described basin inversion (cf. Beach 1987; Kusznir et al. 1987). Thickening of the mantle lithosphere without accompanying thickening of the crust is invoked to account for the regional uplift (Fig. 16). Submersion of cold, dense mantle lithosphere into the surrounding asthenosphere would have caused an initial, isostatically-driven subsidence. Subsequent warming of the lithosphere would have caused uplift. Hence, thickening of the mantle lithosphere without thickening of the overlying crust (which is concentrated in the inversion axes) can account for the initial subsidence then uplift of uninverted areas inferred from compaction and vitrinite data (Fig. 16). The amount of mantle lithospheric thickening required to produce 1/3 km of tectonic uplift can be easily determined from the curves of Sandiford Powell (1990), which are paramaterized in terms of thickening factors, f (reciprocal of extension factor, 13). In the absence of crustal thickening a mantle lithospheric thickening factor (fmi) of 1.1 will generate the inferred subsidence then uplift. It is proposed that mantle lithospheric compression and thickening is

266

K. THOMSON & R. R. HILLIS

HALF GRABEN

iii!iiiii

/

RAMP BASIN

/

FLAT

ii!iiiiiiiiiiii~!iliiii!i!iiiii~!iiiiiiiiii!!~iiiiii!~iii!~!ii!!i!

RAMP

MANTLE LITHOSPHERE

INVERTED INVERTED HALF RAMP GRABEN BASIN

SYN-COMPRESSIONAL REGIONAL SUBSIDENCE

B

CRUSTAL ~ THICKENING--P

MINOR THERMAL UPLIFT

D

z rr 7'0

1.4

w~vc~
1.2

~u.

1.0

41

MANTLE L I T H O S P H E R I C ~ t-]~GKE~ING

POST COMPRESSIONAL THERMAL UPLIFT

fml fl

'~ ..- .,<""" "" """ ' - " - . , , .

~-'~ "5"""

~ ""~ "~ ~'

A20 E

o ~-

~

1.0 = _.~176176176

.....

..

balanced by crustal compression and thickening in inverted areas both within the I M F and indeed throughout North West Europe. In inverted areas crustal thickening alone would lead to an initial uplift (no prior subsidence) and relatively minor subsequent thermal uplift (Fig. 16). In the inverted areas, where pre-erosional units crop out at sea bed, there is no evidence of burial immediately prior to erosion. Regional uplift and erosion in the I M F occurred during a period of c. 5 Ma. during the early Palaeocene. If thermal re-equilibration, following mantle lithospheric thickening, drove the uplift, lithospheric warming must occur more rapidly following compression than does cooling following lithospheric extension and thinning (which has a thermal time constant of 5 0 ~ 0 Ma). The thermal boundary layer which separates the lithosphere from the asthenosphere may become unstable, and detach, if submersed in hot asthenosphere following lithospheric thickening (Houseman et al. 1981). Such detachment may cause rapid warming of the overlying mantle part of the lithosphere directly juxtaposed to convecting asthenosphere. The implications of the two-layer compressional model are more fully discussed in Hillis (1992) and although it requires special pleading for the thermodynamics, it remains the only mechanism which can account for all the features seen in the I M F if the assumption of Danian uplift and erosion is correct.

. . . . .

Conclusions

Fig. 17. Two-layer lithospheric compression and thickening. (A) Existing sedimentary basins such as the IMF were formed above ramps in the crustal detachment during Mesozoic extension (cf. Beach 1987). (B) Compression reactivates the crustal detachment causing basin inversion. Mantle lithospheric thickening is decoupled and laterally displaced from that in the crust and causes initial subsidence; the subsidence prior to uplift phase of the IMF. (C) The lithosphere thermally re-equilibrates to its pre-compressional level, uplifting the region of mantle lithospheric thickening; the regional uplift phase of the IMF. (D) Balanced distribution of lithospheric thickening similar to that illustrated schematically in (A-C). The crustal thickening factor (f~- solid line) is the ratio of thickness of the deformed crust to its initial thickness (Sandiford & Powell 1990). Similarly, fmi is mantle lithospheric thickening (dashed) andji is whole lithospheric thickening (dotted). 0E) Surface elevation changes associated with the distribution of lithospheric thickening shown in (D), assuming local isostasy and no surface loading (Sandiford & Powell 1990). Syncompressional elevation changes are shown by the solid lines and post-compressional, thermal uplift is shown by the dashed line.

(1) The I M F experienced extensional, strikeslip and compressional reactivation during Danian times. The style of reactivation is consistent with regional oblique dextral extension in a N W - S E compressive, N E SW extensional stress field. Compressional reactivation (basin inversion) was restricted to the vicinity of the intersection of the Great Glen and Wick Faults. (2) Sediment compaction and vitrinite reflectance data suggest that the I M F underwent regional Danian erosion of c. 1 km. Although there is a local maxima of erosion in the inverted area, erosion was regional. The regional erosion is indicative of regional surface uplift and hence requires a tectonic source. (3) Regional erosion was preceded by regional subsidence to depths at which the rocks acquired their anomalous (with respect to their present burial depth) compaction and reflectance characteristics.

TERTIARY STRUCTURATION AND EROSION, INNER MORAY FIRTH

(4) The pattern of uplift can be broadly explained by a decoupled, two-layer model of lithospheric compression. Uplift in the inverted areas was driven by crustal compression while mantle lithospheric compression drove subsidence then regional uplift in the areas where there is no evidence of crustal c o m p r e s s i o n and thickening. (5) Structural reactivation and uplift and erosion in the IMF was contemporaneous with both Alpine and North Atlantic 'Thulean' events. Plate boundary forces associated with these events probably combined to produce the intraplate stress field of the I M F that was responsible for structural reactivation and erosion. K.T. acknowledges a Shell/Esso Studentship. R.R.H. thanks Shell UK Expro, the Royal Society (Marine Sciences Grant) and the Carnegie Trust for the Universities of Scotland for financial support. K.T. acknowledges the support of Shell UK Expro, Esso, Unocal, Arco, Mobil, Halliburton, Shelf Exploration, British Petroleum and Geco through the release of data. The British Geological Survey (BGS) are also thanked for extensive logistical support. Paul Green and Richard Bray at Geotrack International are thanked for their interest and support. Thanks to John Underhill, Roger Scrutton, Jon Turner, John Dixon and Sarah Prosser of Edinburgh University for comments on the work, to Larry Wakefield, Griff Cordey and Ceri Powell at Shell UK Expro, and Iain Scotchman at Enterprise Oil.

References ANDREWS, I. J. & BROWN, S. 1987. Stratigraphic evolution of the Jurassic, Moray Firth In: BROOKS, J. & GLENNIE, K. (eds) Petroleum Geology of North West Europe. Graham & Trotman, London, 785-795. , LONG, D., RICHARDS,P. C., THOMSON,A. R., BROWN, S., CHESHER, J. A. & MCCORMAC, M. 1990. The Geology of the Moray Firth. British Geological Survey, United Kingdom Offshore Report, HMSO, London. BARR, D. 1985. 3-D Palinspastic reconstruction of normal faults in the Inner Moray Firth: implications for extensional basin development. Earth and Planetary Science Letters, 75, 191-203. BARTON, P. & WOOD, R. 1984. Tectonic evolution of the North Sea basin: Crustal stretching and subsidence. Geophysical Journal of the Royal Astronomical Society, 79, 987-1022. BEACH, A. 1987. A regional model for linked tectonics in north-west Europe. In: BROOKS,J. & GLENNIE, K. (eds) Petroleum Geology of North West Europe. Graham & Trotman, London, 43-48. BIRD, T. J., BELL, A., GIaaS, A. D. & NICHOLSON,J. 1987. Aspects of strike-slip tectonics in the Inner

267

Moray Firth basin, offshore Scotland. Norsk Geologisk Tidsskrift, 67, 353-369. BOTT, M. H. P. 1988. A new look at the causes and consequences of the Icelandic hot-spot In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 15-23. BULAT, J. & STOKER, S. J. 1987. Uplift determination from interval velocity studies. In: BROOKS, J. & GLENNIE, K. (eds) Petroleum Geology of North West Europe. Graham & Trotman, London, 293305. CHADWICK, R. A. 1985. Permian, Mesozoic and Cenozoic structural evolution of England and Wales in relation to the principles of extension and inversion tectonics. In: WHITTAKER, A. (ed.) Atlas of Onshore Sedimentary Basins in England and Wales: Post-Carboniferous Tectonics and Stratigraphy. Blackie, Glasgow, 9-25. CLOETINGH, S., MCQUEEN, H. & LAMBECK,K. 1985. On a tectonic mechanism for regional sea level variations. Earth and Planetary Science Letters, 75, 157-166. , GRADSTEIN,F. M., KooI, H., GRANT,A. C. & KAMINSKI, M. 1990. Plate reorganisation: a cause of rapid late Neogene subsidence and sedimentation around the North Atlantic? Journal of the Geological Society of London, 147, 495-506. COWARD, M. P. (1986) Heterogeneous stretching, simple shear and basin development. Earth and Planetary Science Letters, 80, 325-336. DEEGAN, C. E. • SCULL, B. J. 1977. A proposed standard lithostratigraphic nomenclature for the central and northern North Sea. Report of the Institute of Geological Sciences 77/25; Bulletin of the Norwegian Petroleum Directorate 1. ENGLAND, P. & MOLNAR, P. 1990. Surface uplift, uplift of rocks, and exhumation of rocks. Geology, 18, 1173-1177. ENGLAND, R. W. 1988. The early Tertiary stress regime in NW Britain: evidence from the patterns of volcanic activity. In: MORTON, A. C. & PARSONS, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 381389. FOWLER, S. & MCKENZIE, D. 1989. Gravity studies of the Rockall and Exmouth Plateaux using Seasat altimetry. Basin Research, 2, 27-34. FROSTICK, L., REID, I., JARVIS, J. t~ EARDLEY, H. 1988. Triassic sediments of the Inner Moray Firth, Scotland: early rift deposits. Journal of the Geological Society of London, 145, 235-248. GREEN, P. F. 1989. Thermal and tectonic history of the East Midlands shelf (onshore UK) and surrounding regions assessed by apatite fission track analysis. Journal of the Geological Society of London, 146, 755-733. HARDING, T. P. 1990. Identification of Wrench Faults Using Subsurface Data: Criteria and Pitfalls. Bulletin of the American Association of Petroleum Geologists, 74, 1590-1609. HARDMAN, M., BUCHANAN, P., HERRINGTON, P. &

268

K. THOMSON & R. R. HILLIS

CARR, A. 1993. Geochemical modelling of the East Irish Sea Basin: its influence on predicting hydrocarbon type and quality. In: PARKER, J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 809-821. HELLINGER, S. J. & SCLATER, J. G. 1983. Some comments on two-layer extensional models for the evolution of sedimentary basins. Journal of Geophysical Research, B88, 8251-8269. HILLIS, R. R. 1991. Chalk porosity and Tertiary uplift, Western Approaches Trough, SW UK and NW French continental shelves. Journal of the Geological Society of London, 148, 669-679. 1992. A two-layer lithospheric compressional model for the Tertiary uplift of the southern United Kingdom. Geophysical Research Letters, 19, 573-576. , THOMSON, K. & UNDERHILL, R. R. 1994. Quantification of Tertiary erosion in the Inner Moray Firth by sonic velocity data from the Chalk and Kimmeridge Clay. Marine and Petroleum Geology. HOLGATE, N. 1969. Palaeozoic and Tertiary transcurrent movements on the Great Glen Fault. Scottish Journal of Geology, 5, 97-139. HOUSEMAN, G. A., MCKENZIE, D. P. & MOLNAR, P. 1981. Convective instability of a thickened boundary layer and its relevance for the thermal evolution of continental convergence belts. Journal of Geophysical Research, B86, 6115-6 132. JENSEN, L. N. & SCHMIDT, B. J. 1993. Neogene uplift and erosion offshore Norway; magnitude and consequences for hydrocarbon exploration in the Farsund Basin In: SPENCER,A. M. (ed.) Generation, accumulation and production of Europe's Hydrocarbons III. European Association of Petroleum Geologists, Special Publication, 3. KooI, H. & CLOETINGH, S. 1989. Intraplate Stresses and the Tectono-Stratigraphic Evolution of the Central North Sea. In: TANKARD, A. J. & BALKWILL, H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins. American Association of Petroleum Geologists, Memoir, 46, 541-558. KUSZNIR, N. J., KARNER, G. D. & EGAN, S. 1987. Geometric, thermal and isostatic consequences of detachments in continental lithosphere extension and basin formation. In: BEAUMONT, C. & TANKARD, A. J. (eds) Sedimentary Basins and Basin-Forming Mechanisms. Canadian Society of Petroleum Geologists Memoir, 12, 185-203. LANG, J. R. 1978. The determination of prior depth of burial (uplift and erosion) using interval transit time. Society of Profession Well Log Analysts Nineteenth Annual Logging Symposium, June 1316, Paper B. MAGARA, K. 1976. Thickness of removed sedimentary rocks, paleopore pressure, and paleotemperature, southwestern part of Western Canada Basin. Bulletin of the American Association of Petroleum Geologists, 60, 554-565. McQUILLIN, R., DONATO, J. A. & TULSTRUP, J. 1982. Development of basins in the Inner Moray Firth -

-

and the North Sea by crustal extension and dextral displacement of the Great Glen Fault. Earth and Planetary Science Letters, 60, 127-139. MOLLER, B., ZOBACK,M. L., FUCHS, K., ETAL. 1992. Regional patterns of tectonic stress in Europe. Journal of Geophysical Research, B97, 1178311803. MURRELL, S. A. F. 1986. Mechanics of tectogenesis in plate collision zones In: COWARD, M. P. & RIES, A. C. (eds) Collision Tectonics. Geological Society, London, Special Publication, 19, 95-111. NAYLOR, M. A., HAUGHEY, N., CLAYTON, G. & GRAHAM, J. R. 1993. The Kish Bank Basin, offshore Ireland. In: PARKER, J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 845-855. , MANDL, G. & SIJPESTEIJN, C. H. K. 1986. Fault geometries in basement-induced wrench faulting under different initial stress states. Journal of Structural Geology, 8, 737-752. PEARSON, M. J. & WATKINS, D. 1983. Organofacies and early maturation effects in the Upper Jurassic sediments from the Inner Moray Firth Basin, North Sea. In: BROOKS, J. (ed.) Petroleum Geochemistry and Exploration of Europe. Geological Society, London, Special Publication, 12, 161-173. PRAJOGA, J. 1990. Thermal and Maturation History of the Inner Moray Firth Basin. MSc Thesis, Royal Holloway and Bedford New College, University of London, UK. PROSSER, S. D. 1991. PhD Thesis, University of Keele, UK. ROBERTS, A. M., BADLEY, M. E., PRICE, J. D. & HUCK, I. W. 1990. The structural evolution of a transtensional basin: Inner Moray Firth, NE Scotland. Journal of the Geological Society of London, 147, 87-103. ROGERS, D. A., MARSHALL, J. E. A. & ASTIN, T. R. 1989. Devonian and later movements on the Great Glen fault system, Scotland. Journal of the Geological Society of London, 146, 369-372. ROYDEN, L. & KEEN, C. E. 1980. Rifting processes and thermal evolution of the continental margin of eastern Canada determined from subsidence curves. Earth and Planetary Science Letters, 51, 343-361. SANDIFORD, M. & POWELL, R. 1990. Some isostatic and thermal consequences of the vertical strain geometry in convergent orogens. Earth and Planetary Science Letters, 98, 154-165 SCLATER, J. G. • CHRISTIE,P. m. F. 1980. Continental stretching: an explanation of the post-mid-Cretaceous subsidence of the Central North Sea Basin. Journal of Geophysical Research, 85, 3711-3739. SPEIGHT, J. M. & MITCHELL, J. G. 1979. The PermoCarboniferous dyke-swarm of northern Argyll and its bearing on dextral displacement on the Great Glen Fault. Journal of the Geological Society of London, 136, 3-11. THOMSON, K. & UNDERHILL, J. R. 1993. Controls on the development and evolution of structural styles in the Inner Moray Firth Basin. In: PARKER,J. R.

TERTIARY S T R U C T U R A T I O N A N D EROSION, INNER M O R A Y F I R T H (ed.) Petroleum Geology of North West Europe. Proceedings of the 4th Conference. Geological Society, London, 1167-1178. TILL, R. 1974. Statistical Methods for the Earth Scientist. Macmillan Press, London. UNDERHILL, J. R. 1991a. Implications of MesozoicRecent basin development in the western Inner Moray Firth, UK. Marine and Petroleum Geology, 8, 359-369. 1991b. Controls on Late Jurassic seismic sequences, Inner Moray Firth, UK North Sea: a critical test of a key segment of Exxons original global cycle chart. Basin Research, 3, 79-98. WHITE, N. & MACKENZIE, D. P. 1988. Formation of the 'steer's head' geometry of sedimentary basins by differential stretching of crust and mantle. Geology, 16, 250-253. WHITE, R. S. 1989. Initiation of the Iceland Plume and Opening of the North Atlantic. In: TANKARD, A. J. & BALKWILL, H. R. (eds) Extensional

269

tectonics and stratigraphy of the North Atlantic Margins. American Association of Petroleum Geologists, Memoir, 46, 149-154. 1992. Crustal structure and magmatism of the North Atlantic continental margins. Journal of the Geological Society of London, 149, 841-854. & MCKENZIE, D. P. 1989. Magrnatism at rift zones: the generation of volcanic continental margins and flood basalts. Journal of Geophysical Research, B94, 7685-7729. ZIEGLER, P. A. 1987a. Late Cretaceous and Cenozoic intra-plate compressional deformations in the Alpine foreland - a geodynamic model. Tectonophysics, 137, 389420. 1987b. Compressional intra-plate deformations in the Alpine foreland - an introduction. Tectonophysics, 137, 1-5. - 1990. Geological Atlas of Western and Central Europe, 2rid ed. Shell Internationale Petroleum Maapascappij B.V., Netherlands. -

-

-

-

-

-

Palaeobathymetric reconstruction on a gridded database: the northern North Atlantic and southern Greenland-lceland-Norwegian sea C. N. W O L D

G E O M A R Research Center for Marine Geoscience, Christian Albrechts University, Wischhof Strasse 1-3, D-24148 Kiel, Germany Present address: National Center for Atmospheric Research, Interdisciplinary Climate Systems Section, PO Box 3000, Boulder, Colorado 80307, USA Abstract: A computer program combining backstripping with plate tectonics has been used to reconstruct the palaeobathymetry of the region between the Charlie Gibbs and Jan Mayen Fracture Zones from the East Greenland margin to the European shelf for the past 50 Ma. Two different models were tested, one assuming that continental extension and rifting opened the Rockall Trough and Faeroe-Shetland Channel during the Cretaceous and that the area has subsided ever since. The second model assumes that the Rockall-Faeroe region rifted in the Cretaceous and was uniformly uplifted in the late Palaeocene due to heating of the lithosphere by the mantle plume of the Iceland Hotspot. The alternative assumptions about the thermal age of the lithosphere produce very different results with implications for the palaeoceanography and sedimentation in the region. The isostatic model presented here predicts crustal thicknesses along the Iceland-Faeroe Ridge and in the Rockall-Faeroe region that are in close agreement with estimates using seismic and gravity techniques.

This work represents part of a project to understand the development of the North Atlantic and its adjacent seas using mass-balance to reconstruct the palaeogeography of the region (Hay et al. 1989). The ultimate goal is to reconstruct the configuration of the ocean basins and the adjoining continents through the Mesozoic and Cenozoic, and to determine the history of erosion and sedimentation throughout the region. The study area extends from the Charlie Gibbs Fracture Zone to the Jan Mayen Fracture Zone and from the eastern margin of Greenland to the edge of the European continental shelf (Fig. 1). The terminology followed in this paper is that of Coachman & Aagaard (1974), and the region between Greenland, Jan Mayen and Spitsbergen is referred to as the Greenland Sea, the region between Greenland, Jan Mayen and Iceland as the Iceland Sea, and the remainder as the Norwegian Sea; the entire ocean area between the Greenland-Scotland Ridge and the Fram Strait is referred to as the GreenlandIceland-Norwegian (GIN) Sea. Backstripping involves the removal of layers of sediment that are younger than the age of the reconstruction from a stratigraphic column, restoration of the underlying sediment to its unloaded condition (decompaction), incorporation of the sea level at that time, removal of the effects of thermal subsidence and isostatic

adjustment. Plate tectonic reconstruction involves movement of different parts of the area with respect to each other using rotations on the surface of a sphere. The palaeobathymetric reconstruction program (BalPal) requires that certain data be input as boundary conditions to run the model: (1) stratigraphic-lithologic sections; (2) thermal age of the lithosphere; (3) a eustatic sea level curve; (4) definition of continent-ocean and plate boundaries; and (5) rotation parameters for plate tectonic reconstructions. BalPal is a collection of computer programs for reconstruction of palaeogeography. The software is written in the C-programming language and implemented on the U N I X operating system. BalPal works with gridded data and a 1 x 1~ latitudelongitude grid was chosen for the study area. The area contains a total of 668 grid cells (Fig. 2). To reconstruct p a l a e o b a t h y m e t r y BalPal makes a series of operations on each grid cell: (1) young sediments are removed from the top of grid cell columns; (2) the remaining sediment thickness is unloaded (decompacted), based on empirical equations for the porosity v. depth relationship of different sediment lithologies; (3) the effects of thermal subsidence are removed by changing the mantle density, based on lithosphere subsidence curves; (4) the thickness of the overlying water column is changed according to

From Scrutton, R. A., Stoker, M. S., Shimmield, G. B. & Tudhope, A. W. (eds), 1995, The Tectonics, Sedimentation and Palaeoeeanographyof the North Atlantic Region, Geological Society Special Publication No. 90, pp. 271-302

271

272

C.N. WOLD

Fig. 1. The study area between the Charlie Gibbs Fracture Zone (CGFZ), the Jan Mayen Fracture Zone (JMFZ), Greenland and Europe. Bathymetric contours are every 250m. The map was constructed from the ETOPO5 (1986) data set using 10 x 10' average elevations. Longitude is negative towards the west and latitude is positive towards the north. Geographic features mentioned in the text are: ADS, Anton Dohrn Seamount; AR, Aegir Ridge; BBB, Bill Bailey's Bank; BD, Bjorn Drift; BI, British Isles; DS, Denmark Strait; EB, Edoras Bank; ED, Eirik Drift; ES, Eriador Seamount; FB, Faeroe Bank; FBC, Faeroe Bank Channel; FD, Faeroe Drift; FI, Faeroe Shelf and Islands; FSC, Faeroe-Shetland Channel; FSE, Faeroe-Shetland Escarpment; GBB, George Bligh Bank; GD, Gardar Drift; GLD, Gloria Drift; HAD, Hatton Drift; HAT, Hatton Bank; HTS, Hebrides Terrace Seamount; IB, Irminger Basin; ID, Isengard Drift; IFR, Iceland-Faeroe Ridge; IP, Iceland Plateau; IR, Ireland; JMR, Jan Mayen Ridge; KR, Kolbeinsey Ridge; LB, Lousy Bank; NB, Norway Basin; PA, Porcupine Abyssal Plain; RB, Rosemary Bank; RCK, RockaU Bank; RP = RockaU Plateau, RR = Reykjanes Ridge, RT = Rockall Trough, SC, Scotland; SD, Snorri Drift; SI, Shetland Islands; SIB, South Iceland Basin; SS, Scoresby Sund; WTR, WyviUe-Thomson Ridge.

the eustatic sea level curve; and finally, (5) the elevation of each grid cell is adjusted assuming Airy-type isostatic compensation.

Boundary conditions Stratigraphic-lithologic sections The stratigraphy and lithology of the sedimentary rocks in each grid cell were derived from reports of the Deep Sea Drilling Project (DSDP) and Ocean Drilling Program (ODP), from other literature sources, and from single-channel

reflection seismic profiles distributed by the Marine Geophysics Section of the NOAA (US National Oceanographic and Atmospheric Administration), National Geophysical Data Center (Boulder, Colorado). The locations of DSDP and ODP Sites and of the reflection seismic profiles used in the compilation are shown in Fig. 3.

Thermal age of the lithosphere The term 'thermal age' refers to the time when the lithosphere was heated. The hypothesis is

PALAEOBATHYMETRY OF NORTH ATLANTIC & S GIN SEA

273

Fig. 2. The 1 x 1~ grid applied to the area for compilation of the stratigraphic database. Contour lines are shown for reference at 500 m intervals below sea level. There are 668 grid cells in the database. that on being heated the lithosphere expands, becoming less dense resulting in elevation of the land or seafloor due to isostatic equilibrium. Starting from the time of maximum heating, the 'thermal age', the lithosphere contracts by cooling, becoming progressively more dense resulting in subsidence. The thermal ages of the lithosphere in the grid cells were calculated from plate tectonic modelling using gridded plates at 1 Ma time increments. The crustal age for each grid cell of ocean crust was determined from the age of seafloor magnetic anomalies (Fig. 4) and the time when grid cells on adjacent plates overlapped (see discussion below). The thermal ages of continental crust are taken to be the time of rifting or thermal uplift indicated by other geologic evidence. Eustatic sea-level curve

The eustatic sea-level curve of Haq et al. (1987) serves as the basis for determining relative sea

level at times in the past. Hay et al. (1989) argued that although the timing and relative magnitude of the sea-level fluctuations may be correct, the absolute magnitudes of sea level change given by Haq et al. (1987) for the Cenozoic are too large. Hay et al. (1989) suggested that the correct magnitude is probably on the order of half that given by Haq et al. (1987). For the model results presented here Hay et al. (1989) are followed and half of the amplitude of the Haq et al. (1987) sea-level curve is used. The full-amplitude and half-amplitude sea-level curves after Haq et al. (1987) were averaged over 1 Ma intervals (Fig. 5).

Definition o f continent-ocean a n d plate boundaries The continent-ocean boundary. The least welldefined aspect of plate tectonic models for the North Atlantic is the position of the continent-

274

C.N. WOLD

Fig. 3. The distribution of single channel seismic reflection profiles and DSDP/ODP sites used in the stratigraphic-lithologic compilation. DSDP and ODP drill sites are shown as solid black circles and the seismic profiles are shown as solid lines. Coastlines and the 500, 1000, 2000, 3000, 4000 and 5000 m bathymetric contours are shown for reference. ocean boundary (COB; Fig. 4). Fracture zones and seafloor magnetic anomalies constrain the seafloor spreading history of lithospheric plates (Pitman & Talwani 1972), but the plate configuration at the initiation of seafloor spreading depends on the fit of conjugate COBs. In some areas, as along the margin of the Bay of Biscay, the COB is well defined in reflection seismic sections (Montadert et al. 1979). In others, the COB is a gradual transition between continental and ocean crust (e.g. Roberts et al. 1984). Elsewhere, from interpretation of aeromagnetic data off the northeast Newfoundland margin and off Goban Spur, Srivastava et al. (1988a) concluded that a large amplitude magnetic anomaly between continental and oceanic crust corresponds closely to the location of the COB. Differential stretching during the formation of

a passive margin may make definition of the COB difficult. As suggested by Vink (1982) and modelled by Courtillot (1982), a rift propagating through a continent may result in differential stretching. The idea that the rift propagated from the south to north between Greenland and Eurasia is supported by the decreasing distance between the COB and anomaly 24 towards the north. This was apparently also the case between Flemish Cap and Charlie Gibbs Fracture Zone, where the COB decreases in distance to anomaly 34 from south to north on both sides of the Atlantic (Scrutton 1985; Srivastava et al. 1988b). Prior to, and during, the initiation of seafloor spreading in the northern North Atlantic Ocean and G I N Sea there was much volcanic activity. This activity is documented by the widespread occurrence of seaward-dipping reflectors (Hinz 1981; Mutter et al. 1982), now known to be

PALAEOBATHYMETRY OF NORTH ATLANTIC & S GIN SEA

275

Fig. 4. Present tectonic features digitized for plate tectonic modelling of the southern Labrador Sea, northern North Atlantic and southern GIN Sea. The seafloor magnetic anomalies are shown as thin solid lines, fracture zones are heavy solid lines, ridges are dashed and Present COBs are shown at 1~ resolution. Longitude is negative towards the west and latitude is positive towards the north. Tectonic features are labelled in the diagram: JMFZ, Jan Mayen Fracture Zone; AR, extinct Aegir Ridge; KR, Kolbeinsey Ridge; COB, continent-ocean boundary; RR, Reykjanes Ridge; CGFZ, Charlie Gibbs Fracture Zone; LS, Labrador Sea extinct spreading center; and L, the Leif;, M, Minna; S, Snorri; H, Hudson fracture zones in the Labrador Sea.

subaerial tholeiitic flood basalts within a region known as the Tertiary Igneous Province (White 1992). The volcanic rocks were emplaced approximately 57Ma within a timespan of c. 3 Ma (Eldholm 1991). They cover a region almost 3000km in diameter that was centred over the Iceland Hotspot (White 1992) at that time. Lava flows and sills of the Tertiary Igneous Province can be found on Disco Island (Davis Strait, western Greenland margin), on the eastern Greenland margin from the southern tip of Greenland to the Jan Mayen Fracture Zone, and along the continental margin of Europe from Hatton Bank to the Voring Plateau. The flood basalt volcanism was succeeded by a phase of pyroclastic volcanism that resulted in widespread ash deposits recognizable throughout the NE Atlantic region (Roberts et al. 1984). There have been two main hypotheses for the emplacement of seaward-dipping reflectors. Hinz (1981) recognized seaward-dipping reflectors in seismic reflection profiles from a number of passive margins around the world and

suggested that they are emplaced during the late rifting stages on thinned continental lithosphere. However, Mutter et al. (1982) proposed that seaward-dipping reflectors are emplaced during the early stages of seafloor spreading on oceanic lithosphere. Both models assume that the layered basalt sequences were deposited either subaerially or in shallow water, as has been confirmed by drilling on DSDP Leg 38 (Talwani et aL 1976) and ODP Leg 104 (Eldholm et al. 1987) on the outer V~ring Plateau. Independent of the model for emplacement of seaward-dipping reflectors, their location is closely associated with the transition from continental to oceanic crust. Skogseid & Eldholm (1987) interpreted the COB off the Voring Plateau Escarpment to lie near the seaward edge of the base of the seaward-dipping reflector sequence. North of the Faeroe Islands, Smythe (1983) interpreted the COB to be located near the landward end of the oldest seaward-dipping reflectors off the Faeroe-Shetland Escarpment. On the western margin of Rockall Plateau

276

C.N. WOLD

I

,

I

,

I

,

I

,

200

9

.o 9 ~

A

,,'~ 9

9

E

,,--I

9

150

,

J' o~

o

I ~

9

~~176

~

~

.

.~176 9

..

.."

.

9 o~ 9~ ~

~

200

~176

~ ~

9 .

o

.

.

~

..

~ .

..

:-.

.

U

... :..

,

~

9 .

m G)

I .o

v

,,.

:y

|

....

150

9

9

.:

100

100

50

50 G)

tr

~

-50

-#

-50

i

i

i

i

l

i

10

20

30

40

50

60

Age

I

I

(Ma)

LI

Eoc

j

j-E

-L

r - - ' l Pal

E

I

Fig. 5. Sea-level curve digitized from that published by Haq et al. (1987). The age of the curve was converted to the Berggren et al. (1985) Cenozoic timescale. The curves shown in this diagram were averaged to 1 Ma intervals and plotted at the mid-point of each interval. The dotted curve is the full amplitude of the averaged eustatic curve and the solid curve is half of the full amplitude. The timescale (Berggren et al. 1985) is shown on the bottom of the diagram and present sea level is marked by the horizontal dashed line.

(western margin of Hatton Bank) the emplacement of the seaward-dipping reflector sequences occurred within the interval between Anomalies 25 and 24B (Roberts et al. 1984). Along the western margin of Rockall Plateau the COB is more difficult to locate and may occur as a transition zone beneath the seaward-dipping reflectors (Roberts et al. 1984). It has also been interpreted to lie near the 2000 m bathymetric contour including Edoras Bank (Roberts et al. 1981; Srivastava & Tapscott 1986; Morgan et al. 1989). The location of the COB on the southwest margin of Rockall Bank is from Roberts et al. (1981); it is assumed to terminate to the south at the Charlie Gibbs Fracture Zone. The exact location of the COB along the southeastern Greenland margin is not well known and there have been several proposed locations (Featherstone et al. 1977; Surlyk et al. 1981; Larsen 1984, 1990). The COB chosen by Larsen (1984, 1990) lies closer to Greenland than those previously proposed. It was chosen as approximating the fit between Greenland and Europe just prior to the initiation of seafloor spreading. Along the eastern Greenland margin seaward-dipping reflec-

tors were emplaced on extended continental crust (Roberts et al. 1984). The Greenland-Scotland Ridge began to form as a result of excessive volcanism from the Iceland Hotspot at the same time that seafloor spreading started between Greenland and the Faeroe Islands. The Faeroe Islands are assumed to be underlain by continental crust (Bott et al. 1976); the continent-ocean boundary occurs to the west between the Faeroe Block and the oceanic Iceland-Faeroe Ridge, and also north of the Faeroes. Bott (1983, 1985) used a model for the thermal subsidence of sea floor (Parsons & Sclater 1977) and Thiede & Eldholm (1983) used a thermal subsidence model for aseismic ridges (Detrick et al. 1977) to reconstruct the subsidence history of the G r e e n l a n d - S c o t l a n d Ridge. Weber (1990) used an Airy-type isostatic model to reconstruct the elevation of the Greenl a n d - S c o t l a n d Ridge. All of these authors concluded that the Greenland-Scotland Ridge was formed > 1000 m above sea level and that parts of it began to subside below sea level in the Iceland and the G r e e n l a n d - S c o t l a n d Ridge.

PALAEOBATHYMETRY OF NORTH ATLANTIC & S GIN SEA Eocene. Thiede (1980) and Thiede & Eldholm (1983) argued that the Iceland-Faeroe Ridge was emergent until the Middle Miocene. The subsidence of the Greenland-Scotland Ridge has played an important role in the palaeoceanography of the Atlantic, controlling the exchange of water between the North Atlantic Ocean and the GIN Sea. It has generally been assumed that as the Ridge subsided below sea level dense water from the GIN Sea spilled over it and formed abyssal currents flowing south. These bottom currents are thought to have eroded sediment and initiated the accumulation of sediment drifts. The present spreading centre in eastern Iceland has only been active for the last 3-4 Ma (Saemundsson 1974; Palmason 1974). Prior to that time spreading occurred along the western rift axis (Talwani & Eldholm 1977). In their reconstruction of the North Atlantic, Talwani & Eldholm (1977) modelled the formation of Iceland between Anomaly 7 and Anomaly 5 (25-10Ma), citing estimates of the maximum age of Icelandic rocks as 20 Ma (Dagley et al. 1967; Moorbath et al. 1968). Vogt et al. (1980) interpreted basement steps (palaeo-shelf edge) on the southeast (Kristjansson 1976) and southwest (Egloff & Johnson 1979) margins of Iceland as reflecting abrupt increases in mantle plume discharge and basalt magrnatism. Vogt et al. (1980) concluded that the insular platform of Iceland began to form c. 25Ma ago (Anomaly 7). Extension along the European continental margin. Opinions differ on whether the crust underlying the Rockall Trough is oceanic (Roberts 1975; Roberts et al. 1983) or continental (Talwani & Eldholm 1977). Roberts et al. (1988) interpreted seismic refraction and wide-angle reflection profiles to indicate that the crust is continental and has been stretched to one fifth of its initial thickness. Extension probably began in the late Early Cretaceous (Montadert et al. 1979; Roberts et al. 1981) and was complete by the late Cretaceous (Roberts 1975; Roberts et al. 1981, 1984). The stretching is thought to have propagated from the south towards the north with the youngest stretching in the More Basin (Hanisch 1984). Hanisch concluded that rifting was Aptian or younger and ended between Anomaly 32-31 (72-69 Ma). Kristoffersen (1978) assumed that the Rockall Trough was formed by seafloor spreading during the Cretaceous Normal Quiet interval (c. 11884 Ma) and that spreading ended near the end of the interval around Anomaly 34 time. Roberts et al. (1981) also concluded that spreading had

277

occurred in the Rockall Trough between the Aptian and Maastrichtian (124-65 Ma). Crustal extension in the Hatton-Rockall Basin may have occurred between Anomalies 32 and 31 (Hanisch 1984) and was probably complete prior to the initiation of seafloor spreading west of Hatton Bank.

R o t a t i o n p a r a m e t e r s f o r plate tectonic reconstructions Previous plate tectonic reconstructions of the North Atlantic. Bullard et al. (1965) presented a quantitative method for rotating digitized outlines of continents together to find the best pre-rift fit, using the Atlantic as an illustration. Pitman & Talwani (1972) showed that sea floor magnetic anomalies could be used to reconstruct the positions of lithospheric plates with respect to one another in the past. Talwani & Eldholm (1977) provided a detailed spreading history for the GIN Sea based on the then newly identified magnetic anomalies, fracture zones and the fit of the conjugate continent-ocean boundaries (COB). They concluded that the initiation of seafloor spreading occurred between Anomalies 25 and 24. Le Pichon et al. (1977) reconstructed the pre-rift fit of the continents around the North Atlantic using 3000 m isobaths along the older margins (Africa-North America) and the 2000m isobaths between younger conjugate margins (Greenland-Eurasia). They noted that the pre-rift fit of conjugate continental margins could be further constrained by taking preexisting lineaments into account (Ramsay 1969; Arthaud &Matte 1975). Kristoffersen (1978) and Srivastava (1978) treated the Rockall Plateau as part of Greenland during the early rifting (Creta-ceous) of the North Atlantic to explain the opening of the Rockall Trough. Most plate tectonic reconstructions of the northern North Atlantic Ocean and the GIN Sea have used a three-plate model consisting of Greenland, Eurasia and the Jan Mayen microplate (Talwani & Eldholm 1977; Nunns 1982, 1983; Unternehr 1982; Bott 1985). The Jan Mayen plate was originally added to the basic two-plate configuration of Greenland and Eurasia to account for the fan-shaped sea floor magnetic anomalies around the extinct Aegir Ridge in the Norway Basin. The Jan Mayen plate adds significant complexity to the development of this region. Unternehr (1982) showed detailed reconstructions of the position of Jan Mayen Ridge based on seafloor spreading

278

C.N. WOLD

around the Kolbeinsey and Aegir ridges. These reconstructions were modified by Nunns (1983) who assumed that spreading along the Greenland-Scotland Ridge was analogous to the spreading history along the Reykjanes Ridge. Nunns believed that there were initially two spreading centres, one along the Reykjanes Ridge and the other along the Aegir Ridge, separated by a transform fault north of the Greenland-Scotland Ridge. At about Anomaly 20 time, Jan Mayen began to separate from Greenland and there was spreading to both the west and east of Jan Mayen Ridge. Then at Anomaly 7 time, spreading stopped along Aegir Ridge and shifted entirely to the Kolbeinsey Ridge, where it continues today. The width of the rift valley of the Aegir Ridge is greater than those of active slow-spreading centres, this may reflect ultra-slow spreading along the Aegir Ridge just before it became extinct (Vogt 1986). Bott (1985) followed the same chronology of events in the spreading history north of the Greenland-Scotland Ridge as outlined by Nunns (1983), but modified the rotation of Jan Mayen with respect to Greenland based on the magnetic anomaly map of Nunns et al. (1983). The Cretaceous history of the North Atlantic south of the Charlie Gibbs Fracture Zone was equally complex. Srivastava et al. (1988b) revised the earlier reconstruction parameters of Srivastava & Tapscott (1986) for Anomaly 34. They also identified Anomaly K in the Cretaceous Quiet Zone south of Flemish Cap, and north and south of the Bay of Biscay, and showed that it represents a triple junction that existed during the separation of the Grand Banks, Europe and Iberia. They gave rotation parameters for Anomalies K and M-0 for Eurasia relative to North America, showing the development of the Porcupine Plate from the triple junction. Roest & Srivastava (1989) published reconstruction parameters for magnetic Anomalies 25 and older, based on newly acquired geophysical data. Their rotation parameters resulted in a fit of Greenland to North America 100 km further south than the reconstruction of Srivastava & Tapscott (1986). Miiller & Roest (1992) used Geosat/Seasat altimetry data to identify new fracture zones and extend old ones closer to continental margins in the North Atlantic. Verhoef et al. (1989, 1990) described a method of gridded plate tectonic reconstruction illustrating the technique with the NE Atlantic. The boundaries of the plates along a mid-ocean rift were defined by seafloor magnetic anomalies and the plates rotated together so that the conjugate boundaries coincided. They used this method to

display present bathymetry on a palinspastically corrected database. R o t a t i o n s f o r a new plate tectonic reconstruction o f the northern North A t l a n t i c

In order to find the best fit between magnetic anomalies and the along-strike position of the plates with respect to one another different published rotation models were compared with newly digitized information. Rotations from the literature were used when they provided the best fits, otherwise new rotations were calculated to fit conjugate anomalies. The Charlie Gibbs and Jan Mayen Fracture Zones, and the Hudson, Snorri, Minna and Leif Fracture Zones, constrain the reconstructions. The magnetic anomalies, fracture zones and ridges (Fig. 4) were digitized from Cande et al. (1989), Nunns et al. (1983) and Bott (1985). The ages of the anomalies are from the timescale of Kent & Gradstein (1986) and are given in Table 1 along with the rotation parameters. The rotations given in Table 1 were recalculated so that all are expressed as total reconstruction poles. A total reconstruction pole is a single rotation starting at the present and going backwards in time to reconstruct the position of a plate at some specific time in the past (Cox & Hart 1986). The reference frame for the reconstructions presented here is based on palaeomagnetic data for the North American plate. The palaeolatitude total reconstruction poles for North America (Table 2) have been derived from the polar wander curve calculated from palaeomagnetic poles by Harrison & Lindh (1982). The best fits between Greenland and Eurasia for Anomalies 5, 13 and 24 use the total reconstruction poles and rotations of Srivastava (1985) and Srivastava & Tapscott (1986). The fit between Greenland and Eurasia for Anomaly 7 is from Bott (1985). The fit of Greenland to Europe uses the closure rotation from Talwani & Eldholm (1977). The published rotations used in the literature to fit Anomalies 6, 20 and 21 between Greenland and Eurasia (Talwani & Eldholm 1977; Nunns 1983; Bott 1985; Srivastava 1985; Srivastava & Tapscott 1986; Rowley & Lottes 1988) were not satisfactory. The new total reconstruction poles and rotation angles for these anomalies are presented in Table 1. The youngest seafloor magnetic anomaly in the southern Labrador Sea is Anomaly 20 (Cande et al. 1989) which formed c. 45 Ma. The age of the extinct spreading centre in the southern Labrador Sea is not known exactly but seafloor spreading is thought to have

PALAEOBATHYMETRY OF NORTH ATLANTIC & S GIN SEA

279

Table 1. Total reconstruction poles for the Greenland, Europe and Jan Mayen Plates Chron

Age (Ma)

Latitude

Longitude

Angle

Source

3.19 4.42 2.03 3.12 4.92

1, 2, 3, 4 5 5 5 6 6

Greenland relative to North America 20 21 24 25 31

35.00 44.66 48.75 55.14 58.64 69.00

- 62.10 - 61.71 - 5.36 - 24.48 -43.94

119.37 110.51 18.33 42.75 34.69

7 13

25.50 35.29

Jan Mayen relative to Eurasia 64.9 - 12.3

8.0

7 7

20

44.66

Jan Mayen relative to Greenland 77.61 - 3.3

- 32.82

5

5 6 7 13 20 21 24 fit

8.92 19.35 25.50 35.29 44.66 48.75 55.14 57.50

Eurasia relative to Greenland 68.00 137.00 68.94 135.30 67.05 128.95 68.00 129.90 48.92 134.88 69.44 137.86 46.78 126.85 41.70 124.50

-

3 5~ 8 3 5 5 3 9

2.50 5.09 5.85 7.78 8.10 11.15 10.50 10.15

1, Vogt & Avery (1974); 2, Emery & Uchupi (1984); 3, Srivastava & Tapscott (1986); Srivastava & Arthur (1989); 5, Wold (this paper); 6, Roest & Srivastava (1989); 7, Nunns (1983); 8, Bott (1985); 9, Talwani & Eldholm (1977).

Table 2. Total reconstruction poles for the North American palaeolatitude reference frame Age (Ma) 20 30 40 50 60

Latitude 0 0 0 0 0

Longitude 61.1 67.7 75.4 88.2 93.8

Angle 4.1 5.3 6.6 6.9 10.0

These values were derived from Harrison & Lindh (1982) apparent polar wander curve for North America.

c o n t i n u e d until c. 3 5 M a (Vogt & Avery 1974; E m e r y & U c h u p i 1984; Srivastava & Tapscott 1986; Srivastava & A r t h u r 1989). The fits of A n o m a l i e s 25 a n d 31 in the southern L a b r a d o r Sea are f r o m Roest & Srivastava (1989). The best fits for Anomalies 20, 21 and 24 were d e t e r m i n e d f r o m digitization of the C a n d e et al. (1989) m a p a n d are presented in Table 1. N u n n s (1983) reconstructed the position of the J a n M a y e n M i c r o p l a t e with respect to Eurasia a n d G r e e n l a n d since the time o f initial seafloor spreading. A g o o d fit o f A n o m a l y 20 on

the Jan M a y e n and E u r a s i a n plates was not achieved with the N u n n s (1983) rotation, but A n o m a l y 13 did fit satisfactorily using his rotation. The age o f initial rifting between Jan M a y e n a n d Eurasia is the same as that in the rest of the n o r t h e r n N o r t h Atlantic (just prior to A n o m a l y 24B). Based on the fit of conjugate a n o m a l i e s a b o u t the A e g i r a n d R e y k j a n e s Ridges, Jan M a y e n is m o d e l l e d here as having begun to separate f r o m G r e e n l a n d at A n o m a l y 20 time with c o n t e m p o r a n e o u s spreading to the west and east of the Jan M a y e n Ridge. Seafloor spreading between Rockall Plateau and southeast G r e e n l a n d began prior to A n o m aly 24 ( R o b e r t s et al. 1979; Srivastava & Tapscott 1986). The plate tectonic reconstructions presented here begin just after that time, at 50 Ma.

Reconstructing palaeobathymetry Previous p a l a e o b a t h y m e t r i c reconstructions Sclater et al. (1977) reconstructed the sedimentfree b a t h y m e t r y o f the N o r t h Atlantic. T h e y calculated d e p t h along m a g n e t i c anomalies at times in the past using the t h e r m a l subsidence

280

C.N. WOLD

curves of Parsons & Sclater (1977) without correction for sediment loading. They also assumed symmetric spreading about the midocean ridges. Sclater et al. (1985) reconstructed Neogene palaeobathymetry for the global ocean basins at DSDP sites. Using the equations of Parsons & Sclater (1977) they calculated the sediment-free water depths of the sites (TWO. They referred to the present water depths (with sediment cover) by the term, Tw2. To calculate the sediment-free water depth they developed an isostatic model assuming a compensation depth at the base of the lithosphere. They described the mass equivalent relationship between the sedimented and sediment-free water depths as: TwlPw + TLPL + YPm = Tw2Pw + TLPL + TsPs

(1) where pw is the density of sea water, TL is the thickness of the lithosphere, PL is the density of the lithosphere, y is the thickness of the asthenosphere which accounts for isostatic equilibrium and Pm is the density of the mantle. The sediment thickness (Ts) and sediment density (ps) were taken from DSDP drill site reports. They described the depth equivalent relationship between the two columns as: Tw1+ TL + y =

Tw2 + Ts + TL.

(2)

Solving equations (1) and (2) for Twl they derived:

Tw~ = Tw2 + Ts [(Ps--Pm)/(Pw --Pm)]-

(3)

Here, Tw~ is independent of lithosphere thickness and can be used to calculate the depth of the ocean crust with or without a sediment load. From equation (3) they calculated the sedimentfree water depths (Twl) and the difference between the depth predicted by the Parsons & Sclater (1977) curve and that predicted by their isostatic model. They referred to the difference between these two values as the offset (depth anomaly). This was assumed to be a constant value for calculation of palaeobathymetry. Sclater et al. (1985) chose sites that they thought lie on ocean crust that had subsided without being influenced by thermal swells (hotspots) and reconstructed the palaeobathymetry for those sites at 8, 16 and 22 Ma. Although they used age-depth information from the drill sites to estimate total sediment thickness at times in the past, they did not decompact the older sediment to restore it to its thickness before

loading by younger sediment. Their method also did not include estimates of changes in the water depth due to sea level change. Their reconstructed palaeobathymetric maps, however, were calculated for sediment-free ocean crust. Tucholke & McCoy (1986) published palaeogeographic maps of the North Atlantic showing coastlines, major sediment lithologies, major plate boundaries, palaeobathymetry and palaeocirculation patterns. They used an age-depth equation for the Atlantic based on DSDP drill sites (Tucholke & Vogt 1979) and showed palaeobathymetry for sediment-free oceanic lithosphere. Shaw (1989) used the methods described by Hay et al. (1989) to backstrip sediment from the Gulf of Mexico and eastern US margin. His model included sediment decompaction and no palinspastic correction was needed because seafloor spreading occurred earlier than his reconstruction ages. S u b s i d e n c e o f oceanic lithosphere

Young oceanic lithosphere is elevated along mid-ocean ridges and subsides to greater depths with increasing age and distance from the ridge. Menard (1969) noted the correlation between the depth and age of oceanic lithosphere. Sclater & Francheteau (1970) and Sclater et al. (1971) described the empirical relationship between the depth of ocean crust and its age, where increasing depth is proportional to the square root of age. Davis & Lister (1974) analysed the data of Sclater et al. (1971) and found a linear relationship between increasing depth and the square root of age (t 1/2) for ocean crust from 0 to 80 Ma. The equations expressing this relationship have the form: d(t) = a + b t 1/2, where d(t) is the depth of the ocean crust at time t (in Ma), a is the initial depth of the ocean crust at the ridge crest (in m), and b is a constant describing the time-dependent subsidence rate. Parsons & Sclater (1977) calculated two equations for the depth of oceanic lithosphere as a function of age for the northwest Pacific and western North Atlantic. For sediment-free lithosphere < 70 Ma depth was given as: d(t) = 2500 + 350 t 1/2,

(4)

where the ridge depth at t = 0 is assumed to be 2500 m. For lithosphere older than 20 Ma, they calculated: d(t) = 6400 - 3 2 0 0 e (-t/62"8),

(5)

PALAEOBATHYMETRY OF NORTH ATLANTIC & S GIN SEA where 6400 m was assumed to be the maximum depth to which sediment-free ocean crust could subside. Heestand & Crough (1981) concluded that the data Parsons & Sclater (1977) had used to calculate the age-depth relationship was affected by hotspots. Hotspots are thought to represent plumes of hotter mantle material that rise from the lower mantle to the base of the lithosphere. The surface expression of a hotspot is a broad region of elevated topography c.1 km high at its centre and covering a radius of c. 500km (Heestand & Crough 1981) from the centre of the hotspot. Reheating of lithosphere by hotspots could account for the flattening of the Parsons & Sclater's (1977) curve [equation (5)]. Heestand & Crough (1981) modelled age-depth for oceanic lithosphere they thought to be unaffected by hotspots and calculated depth as: d(t) = 2700 + 295 t ~/2,

(6)

for seafloor between the ages of 0-80 Ma, where the ridge depth at t = 0 is assumed to be 2700 m. Hotspots have either remained fixed (Morgan 1983) or move slowly with respect to the mantle (e.g. Jurdy & Stefanick 1991). Lithospheric plates move more rapidly, so that as a plate moves over a hotspot a linear feature is produced. This feature may be elevated topography, a volcanic ridge or a seamount chain. Schroeder (1984) calculated the age-depth relationship for lithosphere in the Pacific Ocean basin for regions > 800km from known hotspots and their tracks. He found the following relationship to be true for lithosphere within the 0-80 Ma age range: d(t) = 2846 + 298 t 1/2.

(7)

Lithosphere in the Pacific Ocean basin that is older than 80 Ma has been reheated by hotspots (Schroeder 1984) and was found to be shallower than the depth predicted by equation (4). Like Parsons & Sclater (1977), Schroeder (1984) found the age--depth relationship for the older lithosphere to be best modelled by an exponential decay curve: d(t) = 6400 - 3 1 1 6 e(-t/54"9).

(8)

The t 1/2 age--depth relationship for young oceanic lithosphere was also found to be true in a major back-arc basin where Park et al. (1990) analysed age and bathymetric data for the Philippine Basin. There the age of the lithosphere ranges from 0 to 60 Ma and the sedimentfree depth is approximated as:

d(t) = 3222 + 366 t 1/2.

281 (9)

The similarity of the coefficients, 350 and 366, in equations (4) and (9) for the western North Atlantic, northwest Pacific and Philippine Sea back-arc basin indicate that subsidence curves for these regions have approximately the same shape. However, the depth to basement in the Philippine Basin is c. 800 m greater than in major ocean basins. The data used to calculate the coefficients for young lithosphere unaffected by hotspots [equations (6 and 7)] come from the Pacific Ocean Basin, which is characterized by higher spreading rates than the Atlantic. The coefficients, 295 and 298, in equations (6) and (7), are less than the 350 of Parsons & Sclater (1977), indicating that oceanic lithosphere in the Pacific subsides less rapidly than Atlantic Ocean lithosphere. Detrick et al. (1977) studied results from DSDP drill sites to model the subsidence of aseismic ridges. Aseismic ridges occur throughout the world's ocean basins and are defined as linear volcanic ridges free of earthquake activity (Laughton et al. 1970). Detrick et al. (1977) found that the Ninetyeast Ridge and southeast Mascarene Plateau (Indian Ocean), Rio Grande Rise and Walvis Ridge (South Atlantic Ocean), and the Chagos-Laccadive Ridge (eastern Central Pacific Ocean) all formed close to sea level and subsided at rates comparable to normal ocean crust. From analysis of DSDP Site 336 (Talwani et al. 1976), however, Detrick et al. (1977) found it to be anomalously shallow for its age and suggested that the Iceland-Faeroe Ridge was above sea level near Site 336 for the first 15Ma after its formation. They used empirical curves from Sclater et al. (1971) for depth v. age of oceanic lithosphere in the East Pacific to model the subsidence of the aseismic ridges.

G r i d d e d p l a t e tectonic reconstructions: m o d e l l i n g the age o f oceanic lithosphere

The main objective of using a gridded plate tectonic model for reconstructing palaeobathymetry is to be able to calculate lithospheric ages and rotation parameters that are consistent with one another. Seafloor magnetic anomalies were digitized from maps published by Cande et al. (1989), Nunns et al. (1983) and Bott (1985), and are shown in Fig. 4. Although it is conceivable that the lithospheric ages for the grid cells could be calculated from interpolation using a gridding algorithm, this would only be a first approximation. When the plates were rotated

282

C.N. WOLD t=O

A.

o,.

/

/l

, - - ~L, II

"F I

I'__/k__ C.

'

L

I

rt" /L

,

I

l---~--'iI I

I

I

i

......

I. __ .L__L__~'

' I

___i___i___i__.i___i i i i

2 i

i

2

I I

I

i

,

,

1---t-

I

',

',

:

I

--$---1---1-

2

__~__~

i

,

--1---1----

- -"

:

......

. . . . . . . . .

D.

t=2

I i

I i

i

--~--~--~-'" '~ ', '', '

:-

i J'---'---'---'---'' 'I 'I I

I

I

t

'I

'i

i

--,~--,~--,~ . . . . . .

I'___L__L__L__'~

Fig. 6. Illustration of the gridded plate tectonic method for determining the age of the lithosphere on a spherical grid. (A) The present (t = 0) plate configuration from 60 to 63~ latitude along a section of the Reykjanes Ridge, the ridge is shown as a heavy solid line in the centre of the diagram and magnetic lineations are thin lines parallel to the ridge. The ages of the magnetic lineations are given in Ma. The Greenland plate is on the left (solid grid) and the Eurasian plate is on the right (dashed grid). (B) The present grid of the Eurasian plate rotated to its position at 2 Ma relative to a fixed Greenland plate. (C) The three grid cells that were removed from the 2 Ma reconstruction. At the resolution of the grid the oceanic lithosphere in these three grid cells is 2 Ma (D) The palinspasticaUy reconstructed plate configuration at 2 Ma.

together there would invariably be gross overlaps and gaps. Verhoef et al. (1989, 1990) described a method of gridded plate tectonic reconstruction in the northeast Atlantic. The boundaries of the plates along a mid-ocean rift were defined by seafloor magnetic anomalies and the plates rotated together so that the conjugate boundaries coincided. In their method, Verhoef et al. (1990) defined the plate boundaries along a given anomaly and applied rectangular grids to the plates in their present positions. One plate was kept fixed and the polygon representing the other plate is rotated to the fixed plate. After

r o t a t i o n only the grid on the fixed plate remained rectangular. The rotated grids were distorted, so they then applied a new grid to data on the rotated plates. They used an interpolation algorithm to transfer the digitized data from the initial grid to the new one. There are two important differences between the method used by Verhoef et al. (1989, 1990) and the method described here. Firstly, the new method palinspastically reconstructs plate positions continuously along spreading centres. Positions can be interpolated to any time within the interval defined by the rotation model; it is not restricted to the specific times represented by the magnetic anomalies. Secondly, the new method does not use a rectangular grid in Cartesian coordinates, but a spherical grid that remains fixed on each plate. To set up the initial conditions for the model the region that is to be reconstructed is first divided into a grid. The most readily available grid on a sphere is the present grid of latitude and longitude. On a spherical grid each cell should have the same dimensions, and a 1 • 1~ grid was chosen for this study. Although equidimensional in terms of angular measure, the 1 • 1~ grid 'squares' are not equidimensional in terms of linear measure. At the latitude of the northern North Atlantic and southern G I N Sea they are trapezoids that are about half as wide in the E - W direction as they are high in the N-S direction. The area of a 1 • 1~ grid 'square' decreases from 7718 km 2 at the southern side of the study area (Lat. 51.5~ to 3949km 2 at the northern side (Lat. 71.5~ The boundaries of oceanic lithosphere for each plate are defined at different times by present and extinct spreading centres, transform faults, plate convergence zones and the continent-ocean boundary. The boundaries are located along the present latitude-longitude grid and magnetic lineations are digitized to assign initial ages to the grid cells. Magnetic lineations that are discontinuous are interpolated from existing data. The last step in setting up the initial conditions is to find the Euler poles and rotation angles (Cox & Hart 1986) that can be used to fit conjugate magnetic lineations. During seafloor spreading the lateral position of the plates relative to one another is constrained such that any transform motion can only occur along fracture zones. The changing plate boundaries along spreading centres are reconstructed from: (1) age of magnetic lineations; (2) age of adjacent grid cells that lie along the current vector of plate motion; and (3) the amount of overlap with grid cells on the adjacent plate. The age of a grid cell formed

PALAEOBATHYMETRY OF NORTH ATLANTIC & S GIN SEA

283

Fig. 7. Thermal age of oceanic and continental lithosphere in the study area. The age of oceanic lithosphere ( < 60 Ma) was calculated using the gridded plate tectonic method outlined in the text and in Fig. 6. The contour interval from 0 to 60 Ma is 5 Ma, and from 60 to 140 Ma the contour interval is 20 Ma Except for the South Labrador Sea, the regions in this figure that are older than 60 Ma are assumed to be underlain by continental crust. The age isochrons in the Rockall-Faeroe region were estimated assuming that extension in the Rockall Trough, Faeroe-Shetland Channel and Hatton-Rockall Basin occurred during the Cretaceous (Hanisch 1984). This is referred to as the 'Cretaceous rift' model and approximates the time when the main phase of continental extension or rifting probably occurred.

by seafloor spreading is calculated starting with the present plate configuration and working backwards in time. The plate boundaries are redefined along spreading centres with each timestep. Figure 6 illustrates the method for determining the age of the lithosphere on a spherical grid. The present (t = 0) plate configuration from 60 to 6 3 ~ latitude along a section of the Reykjanes Ridge is shown in Fig. 6A. The ridge is shown as a heavy solid line in the centre of the diagram and magnetic lineations are thin lines parallel to the ridge. The ages of the magnetic lineations, given in millions of years BP, increase symmetrically away from the Reykjanes Ridge and the ridge is the youngest tectonic feature on the map. There are also two lithospheric plates

shown as 1 x 1~ grids in Fig. 6A. The boundary between the plates is defined by the Reykjanes Ridge. The Greenland plate is on the left (solid grid) and the Eurasian plate is on the right (dashed grid). Figure 6B shows the grid of the Eurasian plate rotated to its position at 2 M a relative to a fixed Greenland plate. Three pairs of grid cells overlap along the spreading centre, but only three cells can be removed from the grids to make a palinspastic reconstruction of the gridded Greenland and Eurasian plates. The cells that must be removed depend on the age of the magnetic lineations that pass through them. Comparing the age of magnetic lineations in pairs of adjacent grid cells of the two plates, it is apparent (Fig. 6A) that the cells containing the largest proportion of young seafloor are the

284

C.N. WOLD

Eurasian grid cell from 60 to 61 ~ N, the Greenland grid cell from 61 to 62~ and the Eurasian grid cell from 62 to 63~ latitude. The age of these three grid cells is then 2Ma, i.e. at the resolution of the grid these cells were formed by sea floor spreading 2 Ma, as shown in Fig. 6C. In the 2 Ma reconstruction the grid cell along the spreading centre between 62 and 6 3 ~ is part of the Greenland plate and those between 60 and 62~ would be part of the Eurasian plate, as shown in Fig. 6D. This procedure, which cannot be automated readily, must be performed for all plate boundaries every 1 Ma. The gridded plate tectonic model was used to estimate the age of ocean crust younger than 60 Ma (Fig. 7). Regions landward of the 60 Ma isochron are underlain by continental crust, except in the Labrador Sea. The age isochrons in the Rockall-Faeroe region represent a model for the timing of Cretaceous rifting in the Rockall Trough, Faeroe-Shetland Channel and Hatton-Rockall Basin.

0

-1

""~'~I -2 "E" --~._. ,-o. ~-3 a .~ t~ -4

LS:

S=[(D/6.02)

SS:

S = [ 1 -(0.574 e D/3J )] x 100%

SH:

S=[1-(0.707e~

o.18s ] x 1 0 0 %

-5 "

Backstripping layers of sediment The term 'backstripping' was introduced by Steckler & Watts (1978) to describe the process whereby layers are progressively removed from a column of sedimentary rock and the new water depths to the top of the remaining sedimentary rock or to basement are calculated. This technique was developed to study different styles of subsidence on passive margins and to separate the effect of thermal subsidence due to upwelling of hot asthenospheric material during rifting from that due to sediment and water loading. In this paper backstripping consists of the following steps: (1) removal of sediment younger than the age of the reconstruction; (2) restoration of the remaining sediment column to its thickness prior to loading and compaction ('decompaction'); (3) removal of the effect of thermal subsidence; (4) change of sea level to its position at the time relative to present sea level; and (5) bringing the entire area into isostatic equilibrium.

Removal of sediment younger than the age of the reconstruction The minimum thickness of sediment that existed at times in the past is estimated by removing sediment younger than the reconstruction age and restoring the older sediment layers to their condition prior to loading. This process is called 'decompaction' and is based on estimates of the increasing solidity of sediment layers with depth of burial. In the NE Atlantic there are both regional unconformities and significant sediment

0

i

!

|

i

20

40

60

80

.|

100

Solidity (%) Fig. 8. The three compaction curves used in this study expressed in terms of the increase of solidity with increasing burial depth for shales (dotted line), limestones (dashed line) and sandstones (solid line). The equations used to calculate solidity (S) as a function of depth (D) are given for limestone (LS), sandstone (SS) and shales (SH). Solidity is expressed as a per cent and depth is given in kilometres below the sediment surface. The equation of shales was derived from Huang & Gradstein (1990). The equation for sandstones was modified from Sclater & Christie (1980), Baldwin & Butler (1985) and Huang & Gradstein (1990). The equation for limestones was modified from Baldwin & Butler (1985) and Huang & Gradstein (1990).

drifts which complicate this process. The drifts are current-controlled accumulations of sediment indicating erosion and redistribution of sediment within the basin. A more realistic model would include estimates of the thickness of sediment eroded where unconformities are observed today, but this is beyond the scope of the present study.

Decompaction of the remaining sediment column Decompaction requires knowledge of how the porosity and its converse, the solidity, of the sediment vary with burial depth and lithology. The solidity of a layer of sediment at any given depth can be estimated from empirical equations

PALAEOBATHYMETRY OF NORTH ATLANTIC & S GIN SEA

I

-2000

,

I

*

285

I

-3000

v

E

-4000

r

GI -5000

d(t) = 6400 - 3200 e (-t/62.8) -6000

0

I

I

I

50

100

150

200

Age (Ma) Fig. 9. Thermal subsidence curves for oceanic lithosphere plotted using equations (4) and (5) shown on the diagram. The subsidence equations are from Parsons & Sclater (1977). Depth is given in metres below a constant sea level. calculated for different lithologies. Empirical equations (Sclater & Christie 1980; Baldwin & Butler 1985; Huang & Gradstein 1990) that describe the compaction of sediment at increasing burial depth were compared. From analysis of the compaction curves it was concluded that sediments that compact like shale (clay, claystone, muds, mudstone and shale) and sands are best modelled with exponential decay equations (Fig. 8). The increase of solidity with increasing burial depth for sediments that compact like limestone (unlithified calcareous ooze, semilithified chalk and lithified limestone) are modelled as a power-law function (Fig. 8). The pore space in sediment below sea level is assumed to be filled with seawater (pw = 1027 kg m-3). The average solidity of a sediment layer is assumed to be equal to the solidity calculated for the depth (D) at the mid-point of that layer. When younger layers of sediment are removed from the top of the column, D will decrease because of the isostatic response to unloading and the solidity (S) will also decrease. The result is a new sediment surface at a level intermediate between the original surface and the original depth of base of the younger sediment that was

removed. Prior to unloading a given sediment layer has thickness, T and solidity S. After unloading the new solidity (S') of the layer is calculated from the solidity-depth relation and the decompacted thickness (T') of the layer is: T' = T ( S / S ' ) .

(10)

Equation (10) is used to calculate the decompacted thickness of each remaining sediment layer after stripping off the younger layers of sediment. Decompacted thicknesses are estimated from the top of the column towards the bottom. This is because the thickness of each layer increases after decompaction, requiring that the depth to the middle of the next layer be increased by the amount the intervening sediment was decompacted. After decompaction the total sediment thickness is the sum of all the decompacted layer thicknesses.

Removal of the effect of thermal subsidence Hay et al. (1989) suggested using the isostatic balance of a column of rock and water above a horizontal isobaric surface to compensate for the effects of thermal subsidence. In the model

286

C.N. WOLD

described here this isobaric surface will be called the compensation depth and is taken to be 100 km below present sea level. Isostatic equilibrium across a region is maintaine d by requiring that the total mass (M) of each column be equal at the compensation depth. Columns contain a layer of mantle and crust; they may also have layers of sediment and a layer of water. A 'layer' is a 3D entity, and can be mantle, crust, sediment or water. The height of a column (H 0 varies and is equal to the sum of the thickness of mantle (TM), crust (Tc), sediment (Ts) and water (Tw) layer on the top of the column above the compensation depth: H I = T M + T c + T s + Tw.

(11)

For any given column the total mass of all the layers is equal to a constant value, M: M = TMPM + T c p e +

TsPs + T w P w

(12)

where the thickness of each layer can be considered numerically equivalent to volume for a column with a unit area of 1 m 2. Any column that is in isostatic equilibrium has a mass equal to M regardless of whether the column is above or below sea level. A reference value of M is calculated using values from Hay et al. (1989) assuming that a 6.5 km thick layer of sediment-free 2 0 0 M a ocean crust (Pc = 2750kg m-3) lies at a depth of 6268m beneath the ocean surface (pw -- 1027 kg m-3). If the thickness of the water layer (Tw) is calculated from equation (5) for 200 Ma ocean crust (Fig. 9) with no sediment (Ts = 0) and the mantle density (pM) is assumed to be 3300 kg m -3, then: M = (87232m 3 x 3300kgm -3) + (6500m 3 x 2750 kg m -3) + (6268 m 3 x 1027 kg m -3) M = 3.1218 x 107kg. To simplify the calculations and eliminate the problem of uncertainty whether crust at a particular site is oceanic or continental, the same density (Pc = 2750 kg m -a) is assumed for both. This is reasonable because young ocean crust includes many cracks and spaces that reduce its density below that of basalt, and old ocean crust contains much light altered material as a result of submarine weathering. The subsidence of lithosphere due to cooling with increasing age (thermal subsidence) is modelled here by changing the density of the upper mantle. Both the thickness of the crust and its density are assumed to remain constant through time. This is an obvious oversimplifica-

tion, but adding the complexities of increasing crustal thickness and changing its density with age would not appreciably affect the palaeobathymetric reconstructions. The depth from present sea level to the top of the crust is calculated using equations (4) and (5) for a sediment-free ocean crust with a thickness of 6500 m and density of 2750 kg m -3. As discussed above, Sclater et al. (1985) suggested that equation (4) be used for crust younger than 70 Ma and equation (5) for older crust. However, the depth v. age curves predicted by these two equations intersect at 26.4Ma (Fig. 5). Switching from equation (4) to (5) at 70Ma would result in a sudden decrease in depth from the older crust. Because depth must increase smoothly with age in this paper equation (4) is used for crust < 26.4 Ma and equation (5) for all older crust. The equation to calculate mantle density was derived from equations (11) and (12), where the unknown terms were PM and TM: PM = (M -- T c P c -- T w P w ) / ( H I -- T c -- Tw). (13) The water depth (Tw) was calculated as described above and the average thickness of ocean crust (Tc = 6500m) was used. Mantle density is thus modelled only as a function of age, i.e. all columns with the same age crust have the same mantle density. Isostatic equilibrium is achieved assuming Airy-type isostatic compensation and is implemented by calculating the present thickness of the crust in each column. This thickness is kept constant through time. When the stratigraphic column is compiled for each grid cell the following values are either known or have been estimated: M, HI, Tw, Pw, Ts, Ps, Pc, PM, leaving TM and Tc unknown. Because there are two equations [equations (11) and (12)] and two unknowns, we can solve for either of the unknown values. From equations (11) and (12) the thickness of the crust is: T r = [ M --PM (HI (Pc -- PM).

-- Ts -- Tw)

-- TsPs

--

TwPw]/

(14)

Before any backstripping can be performed the thickness of the crust in each column must be calculated. The present thickness of the crust (Fig. 10) was estimated using equation (14). The model presented here predicts anomalously thick crust over the entire study area from the Charlie Gibbs to Jan Mayen Fracture Zones. Normal ocean crustal thicknesses (6-8 km) occur only in

PALAEOBATHYMETRY OF NORTH ATLANTIC & S GIN SEA

287

Fig. 10. Present thickness of crust in the study area estimated from equation (13). The region is assumed to be in isostatic equilibrium, the mantle density was calculated from the age of the crust and the thickness of the crust was calculated from the mantle density, total thickness and mass of sediment and the present water thickness or elevation above sea level.

the southeast comer of the study area in the Porcupine Abyssal Plain. Crustal thicknesses in Fig. 10 agree well with published estimates along the Iceland-Scotland Ridge and in the RockallFaeroe region. From gravity-density modelling, Bott & Gunnarsson (1980) estimated up to 30 km of anomalously thick crust in the centre of the Iceland-Faeroe Ridge, with the crust thickening to c. 35 km under the Faeroe Block. The thickness of the crust predicted by equation (14) at the centre of the Iceland-Faeroe Ridge is slightly less (26 km). Assuming a two-layer crust, Weber (1990) also used a gravity-density model to estimate the crustal thickness across the Iceland-Faeroe Ridge. His estimate was app r o x i m a t e l y the same as t h a t of Bott & Gunnarsson (1980). Crustal thicknesses estimated with equation (14) show a local maximum under the Faeroe Block of 30 km. Based on wide-angle ocean bottom seismometer and multichannel seismic refraction pro-

files, Morgan et al. (1989) constructed a velocity model across the western margin of H a t t o n Bank. Their maximum estimated crustal thickness was 2 4 k i n a n d t h a t estimated from equation (14) in Fig. 10 is 22 km. In the northern Rockall Trough, Roberts et al. (1988) used ocean-bottom seismometers and wideangle seismic reflection measurements to model seismic velocities. They indicated that the crust under Rockall Bank was c. 3 0 k m thick, and 2 0 k m thick in the northern Rockall Trough. This agrees well with the estimated crustal thicknesses in Fig. 10 where the maximum thickness under Rockall Bank is 3 0 k m and thicknesses in the North Rockall Trough are c. 18 km. Change o f sea level to its position at the time o f the reconstruction. If a column is below sea level and covered by sea water, then a change in sea level is taken into account by changing the height of the column (HI). If the top of a column was initially above

288

C.N. WOLD

Fig. 11. Present bathymetry of the northern North Atlantic and southern GIN Sea plotted from ETOPO5 (1986) elevations averaged to a 1 x 1~ latitude-longitude grid.

sea level and the change of sea-level does not submerge it the height of the column is not affected. A sea-level curve averaged for 1 Ma intervals relative to present eustatic sea-level is one of the parameters input to the model (Fig. 5). The age scale of the Haq et al. (1987) eustatic curve was adjusted to conform to the Berggren et al. (1985) timescale that was used to assign ages to sediments compiled for palaeobathymetric reconstructions. The eustatic curve was then averaged along 1 Ma time intervals. Hay et al. (1989) suggested that the Haq et al. (1987) sea level curve is too high in the Palaeogene. The problem of overestimation of the magnitude of sea level changes has been recognized by Haq (1989), who suggested that it may be the result of failing to take isostatic effects into account. Accordingly, the amplitude of the Haq et al. (1987) sea-level curve was reduced by a factor of one-half (after Hay et al. 1989) to approximate a more realistic estimate of the relative sea level in the past.

Bringing the area into isostatic equilibrium

The last step to complete the backstripping process and reconstruct the palaeobathymetry of a single column is to balance the column isostatically assuming Airy-type equilibrium. After removing the younger sediment and decompacting what remained, a new total sediment thickness (Ts) and sediment density (Ps) were calculated for the column. An agedependent mantle density (PM) was calculated from modelling the thermal subsidence. The thickness and density of the crust (Tc and Pc) are assumed to remain constant. The present thickness of the crust has been calculated from the observed sediment thickness and elevation of its surface. The total mass of the column (M) is also a constant value (3.1218 • 107kg) and the height of the column (HI) has been adjusted to account for relative sea level change. An isostatically-balanced elevation is now calculated from equations (11) and (12). The unknown parameters are Tw and TM. Equations

PALAEOBATHYMETRY OF NORTH ATLANTIC & S GIN SEA (11) and (12) can be solved for Tw and the result of the calculation will tell us if the column was above or below sea level after isostatic balance and what the reconstructed water depth or elevation was: Tw = [M --PM (HI -- Tc - Ts) - T c P c -- TsPs]/ (Pw --PM) (15) The technique for backstripping developed here is based on the assumption that the thermal history of a region is better known than the palaeobathymetric history. The thermal history is derived using equations (4) and (5) from Parsons & Sclater (1977), solved in conjunction with equations (11) and (12), to calculate the age-dependent mantle density. The fundamental equations for backstripping [equations (11) and (12)] are robust and can be solved to calculate mantle density, crustal thickness or elevation. The model could also be used to calculate the thermal history under the assumption that palaeobathymetry is known. This would be useful for predicting past heat flow based on palaeobathymetry.

Palaeobathymetric reconstructions of the northern North Atlantic and southern GIN sea There are two sets of reconstructions of the northern North Atlantic and southern G I N Sea (Figs 12a-16a & 12b-16b). The first set assuming the 'normal' thermal history or 'Cretaceous rifting' model, and the second set assuming the 'Palaeocene reheating' thermal history model. The first model uses an older thermal age for the lithosphere in the RockaU-Faeroe region, assuming that rifling or extension in the Rockall Trough, Faeroe-Shetland Channel and H a t t o n Rockall Basin occurred during the Cretaceous (Hanisch 1984). The second model assumes the same Cretaceous rifting as the first model, but it also assumes that the entire region was uniformly reheated at 60 Ma during an episode of early Tertiary volcanism that lasted c. 3 Ma near the Palaeocene-Eocene boundary (Eldholm 1991). The reconstructions are shown on a gridded database corrected for seafloor spreading. Palaeolatitude (Table 2) was calculated relative to a North American reference frame using the apparent polar wander curve of Harrison & Lindh (1982). Both sets of reconstructions will be described starting from the present and going back to 50 Ma. The general bathymetric features of the region are seen in the 1 x 1~ grid (Fig. 11) but features

289

smaller than the resolution of the 1 x 1~ grid (Fig. 1) are not resolved in Fig. 11. The 1 • 1~ resolution was chosen to conform to the density of seismic lines in this region and the stratigraphic data used in this study were compiled on the 1 x 1~ grid. Features seen in Fig. 1 that are not resolved in Fig. 11 include the WyviUe-Thomson Ridge, Faeroe Bank Channel, Faeroe Bank, Bill Bailey's Bank, Rosemary Bank, George Bligh Bank, A n t o n D o h r n Seamount, Hebrides Terrace Seamount, Edoras Bank, Eriador Seamount and Eirik Ridge. Lousy Bank is a relatively small-scale feature that can be seen on both Figs 1 & 11. In Fig. 1 the Faeroe-Shetland Channel (just north of the Wyville-Thomson Ridge) is shown having a depth > 750m, but at the coarser resolution of the 1 • 1~ grid (Fig. 11) the channel appears to be blocked by a ridge slightly < than 500 m deep. The sill from the FaeroeShetland Channel into the South Iceland Basin lies in the Faeroe Bank Channel between Faeroe Bank and the Faeroe Shelf on which the Faeroe Islands are located. The minimum depth in the Faeroe Bank Channel is c. 850m. At the 10 x 10' resolution of Fig. 1 the Faeroe Bank Channel appears to be only slightly > 500 m deep, and at the 1 • 1~ resolution of Fig. 11 the Faeroe Bank Channel cannot be discerned. The area where it is located appears to be between 250 and 500 m deep. The sill depth over the W N W - E S E trending Wyville-Thomson Ridge is 650m. It is correctly portrayed as between 500 and 750 m in Fig. 1, but Fig. 11 shows a N-S trending ridge between the Faeroe Shelf and Scotland with a sill depth slightly < 500m. The general morphology of the Denmark Strait and IcelandFaeroe Ridge is relatively well-preserved in the coarser grid, where the sill depths of c. 600 and 500 m, respectively, are seen in both Figs 1 & 11. The shelf margin of Iceland that is well defined by the 250m bathymetric contour in Fig. 1 is lost in Fig. 11. The sinuous shape of the Jan Mayen Ridge seen in Fig. 1 is also obscured in Fig. 11. The entire irregularity of the seafloor seen in Fig. 1 is smoothed out when plotted at the coarser grid resolution of Fig. 11. At 1 • 1~ grid resolution (Fig. 11) the depth of the Faeroe-Shetland Channel is c. 500 m and the Faeroe Islands appear to be c. 100 m below sea level. Rockall Bank is below sea level with a minimum depth of c. 200m. Hatton Bank is deeper with a depth < 1000m. Rockall and Hatton Banks are separated from Lousy Bank by a shallow ridge to the north of Rockall Bank that is c. 1200m deep. The maximum depth of Rockall Trough is c. 2800 m and the bathymetry drops off sharply south of Rockall Trough to

290

C . N . WOLD

Fig. 12. (a) Cretaceous rifling model for reconstructed early Late Miocene (10Ma) palaeobathymetry. (b) Palaeocene reheating model for reconstructed early Late Miocene (10 Ma) palaeobathymetry. For (a) and (b) the boundary of the reconstructed region is marked by the edge of the grey shading and contour lines. Contours at the edges of the reconstructed region do not reflect true bathymetry but are an artifact of the contouring algorithm. The present coastlines of Greenland and the British Isles are shown for reference and depths are given in kilometres below sea level. White areas within the reconstructed region are modelled to be at or above sea level at that time. Sea level is 8 m lower than at present (Fig. 5).

PALAEOBATHYMETRY OF NORTH ATLANTIC & S GIN SEA

291

Fig. 13. (a) Cretaceous rifting model for reconstructed Early Miocene (20 Ma) palaeobathymetry. (b). Palaeocene reheating model for reconstructed Early Miocene (20 Ma) palaeobathymetry. Sea level is 42 m higher than at present (Fig. 5). See caption to Fig. 12 for more detail..

292

C . N . WOLD

Fig. 14. (a) Cretaceous rifting model for reconstructed mid-Oligocene (30 Ma) palaeobathymetry. (b) Palaeocene reheating model for reconstructed mid Oligocene (30 Ma) palaeobathymetry. Sea level is 14m higher than at present (Fig. 5). See caption to Fig. 12 for more detail.

PALAEOBATHYMETRY OF NORTH ATLANTIC & S GIN SEA

293

Fig. 15. (a) Cretaceous rifting model for reconstructed early Late Eocene (40Ma) palaeobathymetry. (b) Palaeocene reheating model for reconstructed early Late Eocene (40 Ma) palaeobathymetry. Sea level is 56m higher than at present (Fig. 5). See caption to Fig. 12 for more detail.

294

C.N. WOLD

Fig. 16. (a) Cretaceous rifting model for reconstructed early Middle Eocene (50 Ma) palaeobathymetry. (b) Palaeocene reheating model for reconstructed early Middle Eocene (50 Ma) palaeobathymetry. Sea level is 106 m higher than at present (Fig. 5). See caption to Fig. 12 for more detail.

PALAEOBATHYMETRY OF NORTH ATLANTIC & S GIN SEA depths > 4000m. The South Iceland Basin is c. 3000m deep and the Iceland-Faeroe Ridge has a sill depth of 500 m. The shallow platform around Iceland is delineated by the 250m bathymetric contour and is oriented along the axis of the Greenland-Scotland Ridge. Reykjanes Ridge presently has an average depth of c. 1500m. The Irminger Basin has approximately the same depth as the South Iceland Basin with an average depth of c. 3000 m. There is a broad shelf along East Greenland and bathymetric contours drop off sharply into the Irminger Basin. The shelf break along East Greenland is deep and lies > 500m below sea level. On the 1 • 1~ average bathymetric map (Fig. 11) the Denmark Strait also has a sill depth of c. 500m. The Iceland Plateau is generally shallow with an average depth of c. 1200 m. The southwest-northeast trending Kolbeinsey Ridge separates the Iceland Plateau into two basins to the west and east of it. The basin between Greenland and the Kolbeinsey Ridge is c. 1500 m deep and the basin between the Kolbeinsey and Jan Mayen Ridges is close to 2000m deep. The Jan Mayen Ridge is apparent in Fig. 11 a s a north-south linear feature separating the Iceland Plateau from the Norway Basin. The Norway Basin including the extinct Aegir Ridge is the deepest basin in the study region with an average depth of c. 3200 m. The general morphology of the region in the early Late Miocene, assuming the Cretaceous rifting model (10Ma; Fig. 12a), was similar to present. The Faeroe-Shetland Channel was slightly shallower than at Present with an average depth < 500m. The Faeroe Islands appear to be c. 100 m below sea level. The depth of Rockall and Hatton Banks was approximately the same as at Present but the connection between Rockall and Lousy Banks was shallower with a depth of c. 900 m. The bathymetry in the Rockall Trough was symmetrical and slightly shallower than at Present, with a maximum depth of c. 2600 m. The abyssal plain to the south of Rockall Trough was still relatively deep, reaching c. 4000m. The South Iceland Basin was slightly shallower than at Present and was 2700m deep. The IcelandFaeroe Ridge was also shallower with a sill depth < 500m. The shallow platform around Iceland was oriented more along the axis of Reykjanes Ridge. Reykjanes Ridge was deeper than at Present with an average depth of c. 1700 m. The Irminger Basin and South Iceland Basin were still approximately the same depth (2700m). There was still a broad shelf along East Greenland but the shelf break appeared slightly shallower at 500 m below sea level. In the

295

10Ma reconstruction (Fig. 12a) the Denmark Strait was approximately as deep as it is at Present and was thus the deepest connection between the eastern North Atlantic and the GIN Sea at that time. The Iceland Plateau was much reduced in area due to rapid seafloor spreading the interval from 10 Ma to Present. The shallow basin seen in Fig. 11, north of Scoresby Sund between Greenland and the Kolbeinsey Ridge at Present had not yet been formed in the 10Ma reconstruction (Fig. 12a). At 10Ma a basin south of Scoresby Sund between Greenland and the Kolbeinsey Ridge, that is not well defined in present bathymetric maps, is seen in Fig. 12a. Thus, the Iceland Plateau appears to have two different basins between Greenland and the Kolbeinsey Ridge, the younger basin to the north and the older basin to the south of Scoresby Sund. The Kolbeinsey Ridge is visible in the 10Ma reconstruction (Fig. 12a) as a bathymetric high at c. 20~ longitude. The linear trend of the Jan Mayen Ridge was more pronounced at 10Ma as it formed a boundary between the Iceland Plateau and the Norway Basin. The Norway Basin was still the deepest basin in the study region with an average depth of c. 3000 m. Differences between the Cretaceous rifting model (Fig. 12a) and the Palaeocene reheating model (Fig. 12b) for palaeobathymetry in the early Late Miocene (10Ma) occur only in the Rockall-Faeroe region. The general morphology is similar to the Cretaceous rifting model (Fig. 12a) with the following exceptions: (1) the Faeroe-Shetland Channel is slightly shallower and narrower with the northwest Scottish coastline further seaward; (2) the Faeroe Islands still appear to be c. 100 m below sea level; (3) Rockall Bank is shallower and Hatton Bank is less well defined bathymetrically; (4) Rockall, Hatton and Lousy Banks now appear to form a single platform with a depth of c. 900 m; (5) Rockall Trough is also shallower with a maximum depth of about 2500 m. The palaeobathymetry reconstructed with the Cretaceous rifting model in the Early Miocene (20Ma; Fig. 13a) is similar to the early Late Miocene reconstruction (10Ma). The FaeroeShetland Channel was shallower than at 10Ma with an average depth of c. 350m. The Faeroe Islands were apparently below sea level. The depth of Rockall and Hatton Banks was approximately the same as at 10Ma and the Hatton-Rockall Basin was shallower. The shape of Rockall Trough was approximately the same as at 10Ma and the maximum depth was also the same (2600m). The abyssal plain to the south of Rockall Trough was as deep as it was

296

C.N. WOLD

10 Ma (Fig. 12a). The South Iceland Basin was shallower at 20 Ma (2500m) than at 10 Ma. The Iceland-Faeroe Ridge was shallower with a sill depth of 250 m. The subareal platform around Iceland was narrower and still oriented southwest-northeast along the axis of the Reykjanes Ridge. The Reykjanes Ridge was deeper than at 10 Ma with an average depth of c. 2000 m. The vertical relief between the Reykjanes Ridge and the Irminger and South Iceland Basins to the northwest and southeast of it, respectively, was steadily decreasing from Present to 20 Ma. The Irminger and South Iceland Basins were both shallower with approximately the same depth (2600m). The shelf break off East Greenland was much shallower in the 20 Ma reconstruction (Fig. 13a) at c. 300m water depth. When the Irminger Basin was younger (20 Ma) the Denmark Strait was narrower but may actually have been deeper at 20 Ma than it was at 10 Ma where the sill depth has increased from 500 to c. 600 m. The basin south of Scoresby Sund and west of Kolbeinsey Ridge is depicted as two smaller basins in the 2 0 M a reconstruction (Fig. 13a). The Kolbeinsey Ridge extended to the northwest from Iceland at 2 0 M a and may have been stationary adjacent to the continental margin of East Greenland as the Jan Mayen Ridge was separating from Greenland. Sediment input from Greenland to the basin between the Kolbeinsey and Jan Mayen Ridges may have spilled over into the much deeper Norway Basin. It was still the deepest basin in the study region with an average depth of c. 3000 m. Palaeobathymetric features of the Palaeocene reheating model (Fig. 13b) that are different from the Cretaceous rifting model (Fig. 13a) for the Early Miocene (20Ma) are as follows: (1) the Faeroe-Shetland Channel was significantly narrower and shallower with a sill depth of c. 250m.; (2) the region around the Faeroe Islands was above sea level (Fig. 13b); (3) Rockall Bank was also subaerially exposed and Hatton Bank was c. 700 m below sea level; (4) Rockall Trough was shallower in the Palaeocene reheating model with a maximum depth of < 2500 m. P a l a e o b a t h y m e t r y reconstructed with the Cretaceous rifting model for the middle Oligocene (30 Ma; Fig. 14a) looks very different from younger reconstructions. The Faeroe--Shetland Channel was almost non-existent and appeared as a shallow bank with an average depth < 250m. The Faeroe Islands were above sea level on the eastern end of a subaerial ridge extending over the central portion of the Greenl a n d - S c o t l a n d Ridge. Rockall and H a t t o n Banks were both shallower with an average

depth of c. 500m. The shape of the Rockall Trough was similar to younger reconstructions, but it was slightly shallower (2500m). The abyssal plain to the south of Rockall Trough remained deep at c. 4000 m below sea level. The South Iceland Basin was shallower than at 20 Ma and was c. 2200 m deep. Iceland was an indistinct feature because it was part of the subaerial ridge extending along the GreenlandScotland Ridge from the eastern margin of the Denmark Strait to the Faeroe Islands. Reykjanes Ridge was approximately as deep as in the 20 Ma reconstruction (Fig. 13a) with an average depth of c. 2000 m. The vertical relief between Reykjanes Ridge and the Irminger and South Iceland Basins was only c. 500m. The Irminger Basin was shallower with a depth of c. 2200 m. The shelf break along the East Greenland margin was very shallow in the 30Ma reconstruction (Fig. 14a) and was coincident with the eastern Greenland palaeoshoreline. The Denmark Strait was more constricted than in younger reconstructions and as shallow as the Faeroe-Shetland Channel (200m). Kolbeinsey Ridge was oriented approximately north-south at 24~ longitude. There was subaerial seafloor spreading along most of the Kolbeinsey Ridge at 30 Ma and two distinct basins to the west and east of it. Jan Mayen Ridge was also a n o r t h south bathymetric rise located at 20~ (Fig. 14a). The Norway Basin was shallower in the middle Oligocene with a depth of c. 2700 m. In the middle Oligocene (30 Ma; Fig. 14b) the thermal model based on Palaeocene reheating started to have a large impact on palaeobathymetric reconstructions. Features observed on the palaeobathymetric reconstruction, assuming the Palaeocene reheating model (Fig. 14b) that are not seen on palaeobathymetry, assuming the Cretaceous rifting model (Fig. 14a) are as follows: (1) the Faeroe-Shetland Channel did not exist at 30 Ma, but was part of a subaerial ridge extending from the eastern margin of the Denmark Strait to Scotland; (2) Rockall Bank had significant subaerial exposure in the middle Oligocene (Fig. 14b) and could have been a local source of detrital sediment deposited in Rockall Trough; (3) Hatton Bank was shallower (Fig. 14b) and was c. 500m below sea level; (4) Rockall Trough was shallower in the Palaeocene reheating model (Fig. 14b) with a maximum depth of c. 2200 m. The Faeroe-Shetland Channel was at or above sea level on the eastern end of the subaerial Greenland-Scotland Ridge in the Late Eocene (40 Ma; Fig. 15a) based on palaeobathymerry reconstructed with the Cretaceous rifting model. The narrow width of the ridge across the

PALAEOBATHYMETRY OF NORTH ATLANTIC & S GIN SEA site of the future Faeroe-Shetland Channel does not rule out the possibility of a shallow surface water connection in the Late Eocene. Rockall Bank was shallower with a depth < 250 m below sea level. Hatton Bank was c. 500 m below sea level. Rockall Trough was also slightly shallower with a depth of c. 2200 m. The abyssal plain to the south of Rockall Trough was slightly shallower with a depth of c. 3700 m. The South Iceland and Irminger Basins and the Reykjanes Ridge were almost indistinguishable from one another as bathymetric features in the Late Eocene. The ocean basin between Rockall Plateau and Greenland was c. 1800 m deep with only c. 250 m difference in elevation between the central spreading centre and adjacent ocean basins. The shelf break along the East Greenland margin was probably coincident with the palaeoshoreline along eastern Greenland. The Denmark Strait did not exist in the Late Eocene (Fig. 15a) and was almost certainly above sea level. The proto-Kolbeinsey Ridge was a slight bulge on the northern margin of the GreenlandScotland Ridge at 23~ longitude. The basin to the west of Kolbeinsey Ridge was a shallow platform in the Late Eocene that would have allowed fine-grained sediment to be transported across this basin and into the Norway Basin. In the Late Eocene the Jan Mayen Ridge was rifting from Greenland. The Norway Basin was slightly shallower in the Late Eocene with a depth of c. 2400 m. Palaeobathymetry reconstructed for the Late Eocene (40 Ma; Fig. 15b) using the Palaeocene reheating model indicates that the GreenlandScotland Ridge was > 200 km wide along its entire length and was above sea level along its entire length. Rockall Bank was well above sea level and must have supplied more sediment into the Rockall Trough and the South Iceland Basin than in younger times. Hatton Bank was shallower in the Late Eocene with a water depth of c. 250m and Rockall Trough was also shallower with a maximum depth of c. 2000 m. The site of the future Faeroe-Shetland Channel was probably above sea level on the eastern end of the Greenland-Scotland Ridge in the early Middle Eocene (50 Ma; Fig. 16a) assuming the validity of the Cretaceous rifting model for reconstructed palaeobathymetry. Surface water connections between the Norwegian Sea and North Atlantic across this ridge were unlikely in the early Middle Eocene. Rockall, Hatton and Lousy Banks were at or slightly above sea level in the early Middle Eocene (Fig. 16a). Rockall Trough was also shallower with a depth of c. 2100m. The abyssal plain to the south of Rockall Trough was slightly shallower with a

297

depth of c. 3500m. The Reykjanes Ridge was not a distinct bathymetric feature at the 1 x 1~ resolution of the early Middle Eocene reconstruction (Fig. 16a). The young ocean basin that existed between the Rockall Plateau and Greenland was only c. 1600 m deep in the early Middle Eocene. Rifting between the Jan Mayen Ridge and Greenland had not started in the early Middle Eocene and the Kolbeinsey Ridge did not yet exist. The Norway Basin appears in the reconstruction to have been isolated and relatively shallow in the early Middle Eocene with a depth of c, 2000 m. The Greenland-Scotland Ridge was a broad topographic feature in the early Middle Eocene (50 Ma; Fig. 16b) that may have had elevations of > 1000m assuming that the Palaeocene reheating model for reconstructed palaeobathymetry is correct. The Palaeocene reheating model also indicates that Rockall Bank was a significant topographic feature connected to the Greenland-Scotland Ridge by a land bridge. Hatton Bank may have been above sea level and separated from Rockall Bank by a shallow basin. Rockall Trough was also shallower with a maximum depth of c. 1700 m. If the Palaeocene reheating model is correct then Rockall Trough and the South Iceland Basin would have received more sediment from erosion of the surrounding land areas than at any time since the early Middle Eocene.

Summary and conclusions The reconstructions are not intended to model palaeogeography for any time prior to the early Middle Eocene (50 Ma). Palaeobathymetric reconstructions of the Palaeocene and older times will require additional modelling techniques not discussed here. The most important factor for modelling the pre-Eocene palaeogeography of the region is the effect of reheating and crustal intrusion and underplating that resulted from the late Palaeocene--early Eocene volcanism centred around the Iceland Hotspot. The two sets of maps show very different palaeogeographies but there are no direct geologic data that can be used to verify one or other of the models. The reconstructed bathymetry is dependent primarily on the initial boundary conditions of the model: present elevation, present thickness and age of sediment and the assumed thermal age of the lithosphere. The Palaeocene reheating model conforms to the most generally accepted view of the thermal evolution of the area. The assumption of a thermal reheating event, where the thermal age of the lithosphere was essentially reset to zero at

298

C.N. WOLD

60 Ma, is not unreasonable considering that a region almost 3000 km in diameter was heated by the mantle plume forming the Iceland Hotspot in the early Tertiary (White et al. 1987; White 1988; Parson et al. 1988; White 1989; Eldholm 1991; White 1992) that was centred only hundreds of kilometres to the west of the Faeroe-Shetland Channel at that time. According to this model, the Faeroe-Shetland Channel was rifted in the Cretaceous (Hanisch 1984) and uplifted by reheating in the Late Palaeocene. The Palaeocene was also the time of emplacement of seamounts in the Rockall Trough (e.g. Anton Dohrn) and around the Rockall Bank. The palaeobathymetric reconstructions (Figs 12-16b) based on the Palaeocene reheating model indicate that the Iceland-Scotland Ridge was at, or above, sea level until between 30 and 25 Ma. If one were to assume that there was no reheating event c. 60Ma, then the FaeroeShetland Channel would have begun to subside below sea level prior to the Denmark Strait, at c. 35 Ma. The late Eocene-early Oligocene age of Feni Drift in the Rockall Trough (Masson & Kidd 1987) favours the model assuming Cretaceous rifting without significant subsequent thermal uplift of the Faeroe--Shetland Channel. The model assuming Palaeocene reheating shows the Greenland-Scotland Ridge north of the Rockall Trough to be above sea level during most of the Oligocene. Even assuming that a Faeroe-Shetland Channel too narrow to be resolved by the model provided a marine connection, it seems unlikely that it would have been deep enough to allow the passage of large quantities of dense outflow. A combination of the two thermal models may best explain the initiation of Feni Drift together with the later initiation of the Bjorn and Gardar Drifts in the South Iceland Basin in the early Middle Miocene (Wold 1992). The initiation of early Middle Miocene drift formation in the South Iceland Basin could be explained by dense water formation on shallow shelf areas on the eastern segment of the Iceland-Faeroe Ridge (indicated by the Palaeocene reheating model). Since the Palaeocene reheating model with uniform reheating of the same magnitude across the entire Rockall-Faeroe region appears to be incorrect, palaeobathymetric reconstructions and the timing of initiation of drift sedimentation indicate that the amount of thermal uplift caused by the mantle plume of the Iceland Hotspot decreased radially from its centre. There was probably uplift along the eastern segment of the Iceland-Faeroe Ridge and no uplift in the Faeroe-Shetland Channel in the late

Palaeocene-early Eocene. The model assuming Palaeocene reheating of the lithosphere indicates significantly shallower bathymetry than the Cretaceous rifting model. In the middle Oligocene (30 Ma) reconstructions based on the Palaeocene reheating model, the Greenland-Scotland Ridge is above sea level except through a shallow Denmark Strait. Rockall Bank is also shown above sea level. The Cretaceous rifting model, however, indicates that many of these areas were submerged at that time. It is apparent from the two sets of reconstructions (Figs 12a-16a & 12b-16b) that the model is sensitive to the input thermal age of the lithosphere. This has important implications for the reconstruction of palaeobathymetry and palaeoceanography of the region. The reconstructions presented here are the first t h a t combine a backstripping model, including decompaction of sediment with a gridded plate tectonic model. Palaeobathymetry, or elevation of a stratigraphic column, is calculated based on the age of the lithosphere or last major reheating event. The fundamental equations for backstripping [equations (11) and (12)] are robust and can be solved to calculate mantle density, crustal thickness or elevation. The model could also be used to determine the thermal history if the palaeobathymetry were known. This would be useful for predicting past heat flow based on palaeobathymetry. I thank William W. Hay for stimulating discussions and critical review of the manuscript and Jrrn Thiede for his comments and suggestions. I am indebted to Pompeyo Bajar for his help in the initial development of BalPal and to Richard J. Wold for his early support of the development of the software. This work was supported in part by Deutsche Forschungsgemeinschaft Grant Du 129/5. All of the diagrams in this manuscript were plotted using the GMT (P. Wessel & W. H. F. Smith, 1991, EOS, 72, 441-446) software package.

References ARTHAUD, F. & MATTE, P. 1975. Late Hercynian wrench faults in southwestern Europe. Tectonophysics, 25, 139-171. BALDWIN, B. & BUTLER, C. O. 1985. Compaction curves. American Association of Petroleum Geologists Bulletin, 69, 622~526. BERGGREN, W. m., KENT, D. V., FLYNN, J. J. & VAN COUVERINr, J. A. 1985. Cenozoic geochronology. Geological Society of America Bulletin, 96, 14061418. BOTT, M. H. P. 1983. The crust beneath the IcelandFaeroe Ridge. In: BOTT, M. H. P., SAXOV, S., TALWANI, M. & THmDE, J. (eds) NATO Con-

PALAEOBATHYMETRY OF NORTH ATLANTIC & S GIN SEA

ference Series, Series IV." Marine Sciences, Vol. 8, Structure and Development of the GreenlandScotland Ridge: New Methods and Concepts. Plenum Press, New York, 63-76. 1985. Plate tectonic evolution of the Icelandic transverse ridge and adjacent regions. Journal of Geophysical Research, 90, 9953-9960. & GU~NARSSON, K. 1980. Crustal structure of the Iceland-Phaeroe Ridge. Journal of Geophysics, 47, 221-227. , NIELSEN, P. H. & SUTrtERLAND, J. 1976. Converted P-waves originating at the continental margin between the Iceland-Phaeroe Ridge and the Phaeroe Bank. Journal of the Royal Astronomical Society, 44, 229-238. BULLARD, E. C., EVERETT, J. E. & SMITH, A. G. 1965. The fit of the continents around the Atlantic. In: BLACKETT, P. M. S., BULLARD, E. C. & RUNCORN, K. S. (eds) A Symposium on Continental Drift, Philosophical Transactions of the Royal Society of London, Series A, 258, 41-51. CANDE, S. C., LA BRECQUE, J. L., LARSON, R. L., PITMAN, W. C., III, GOLOVCHENKO,X. & HAXBY, W. F. 1989. Magnetic lineations of the world's ocean basins, map. American Association of Petroleum Geologists, Tulsa, Oklahoma, USA. COACHMAN, L. K. & AAGAARD, K. 1974. Physical oceanography of Arctic and Subarctic seas. In: HERMAN, Y. (ed.) Marine Geology and Oceanography of the Arctic Seas. Springer-Verlag, Berlin, -

-

-

-

1 - 7 2 .

COURTILLOT, V. 1982. Propagating rifts and continental breakup. Tectonics, 1, 239-250. Cox, A. & HART, R. B. 1986. Plate Tectonics: How It Works. Blackwell Scientific, Palo Alto, CA, USA. DAGLEY, P., WILSON, R. L., ADE-HALL, J. M., ETAL. 1967. Geomagnetic polarity zones for Icelandic lavas. Nature, 216, 25-29. DAVIS, E. E. & LISTER, C. R. B. 1974. Fundamentals of ridge crest topography. Earth and Planetary Science Letters, 13, 405-413. DETRICK, R. S., SCLATER, J. G. & THIEDE, J. 1977. Subsidence of aseismic ridges. Earth and Planetary Science Letters, 34, 185-196. EGLOFE, J. & JOHNSON, G. L. 1979. Erosional and depositional structures of the southwest Iceland insular margin: Thirteen geophysical profiles. In: WATKINS, J. S., MONTADERT, L. & DICKERSON,P. W. (eds) Geological and Geophysical Investigations of Continental Margins. American Association of Petroleum Geologists Memoir, 29, 43-64. ELDHOLM, O. 1991. Magmatic-tectonic evolution of a volcanic rifted margin. Marine Geology, 102, 4361. - - , THIEDE, J., TAYLOR, E., ETAL. 1987. Proceedings of the Ocean Drilling Program, Part A Initial Reports, 104. Ocean Drilling Program, College Station, TX. EMERY, K. O. & UCHUPI, E. 1984. The Geology of the Atlantic Ocean. Springer-Verlag, New York. ETOPO5 1986. Relief Map of the earth's surface. LOS, 67, 121. FEATHERSTONE, P. S., BOTT, M. H. P. & PEACOCK, J. H. 1977. Structure of the continental margin of

299

south-eastern Greenland. Geophysical Journal of the Royal Astronomical Society, 48, 15-27. HANISCH, J. 1984. The Cretaceous opening of the northeast Atlantic. Tectonophysics, 101, 1-23. HAQ, B. U. 1989. Sequence stratigraphy, sea level change and significance for the deep sea. In: MACDONALD, D. I. M. (ed.) Sedimentation, Tectonics and Eustasy: Sea Level Change at Active Margins. International Association of Sedimentologists, Special Publication, 12, 3-39. - - , HARDENBOL, J. & VAIL, P. R. 1987. Chronology of fluctuating sea levels since the Triassic. Science, 235, 1156-1167. HARRISON, C. G. A. & LINDH, T. 1982. A polar wandering curve for North America during the Mesozoic and Cenozoic. Journal of Geophysical Research, 87, 1903-1920. HAY, W. W., SHAW, C. A. & WOLD, C. N. 1989. Massbalanced paleogeographic reconstructions. Geologische Rundschau, 78, 207-242. HEESTAND, R. L. • CROUGH, S. T. 1981. The effects of hot spots on the oceanic age-depth relation. Journal of Geophysical Research, 86, 6107-6114. HINZ, K. 1981. A hypothesis on terrestrial catastrophies: Wedges of very thick oceanward dipping layers beneath passive margins - Their origin and paleoenvironmental significance. Geologisches Jahrbuch, Reihe E, 22, 3-28. HUANG, Z. & GRADSTEIN, F. M. 1990. Depth-porosity relationship from deep sea sediments. Scientific Drilling, 1, 157-162. JURDY, D. M. & STEEANICK, M. 1991. Are plumes fixed? Or do they go with the flow? LOS, 72 445. KENT, D. V. & GRADSTEIN, F. M. 1986. A Jurassic to Recent chronology. In: VOGT, P. R. & TUCHOLKE, B. E. (eds) The Geology of North America, Vol. M: The Western North Atlantic Region. Geological Society of America, Boulder, Colorado, 45-50. KRISTJANSSON, L. 1976. A marine magnetic survey off southern Iceland. Marine Geophysical Researches, 2, 315-326. KRISTOFFERSEN, Y. 1978. Sea-floor spreading and the early opening of the North Atlantic. Earth and Planetary Science Letters, 38, 273-290. LARSEN, H. C. 1984. Geology of the East Greenland shelf. In: SPENCER,A. M. (ed.) Petroleum Geology of the North European Margin. Graham & Trotman, London, 329-339. 1990. The East Greenland shelf. In: GRANTZ, A., JOHNSON, L. & SWEENY, J. F. (eds) The Geology of North America, Vol. L: The Arctic Ocean Region. Geological Society of America, Boulder, Colorado, 185-210. LAUGHTON, A. S., MATTHEWS,D. W. & FISHER, R. L. 1970. The structure of the Indian Ocean. In: MAXWELL, A. E. (ed.) The Sea, Vol. 4, Part 2: New Concepts of the Sea Floor Evolution. John Wiley & Sons, New York, 543-586. LE PICHON, X., SIBUET,J.-C. & FRANCHETAU,J. 1977. The fit of the continents around the North Atlantic Ocean. Tectonophysics, 38, 169-209. MASSON, D. G. & KIDD, R. B. 1987. Revised Tertiary seismic stratigraphy of the southern Rockall Trough. In: RUDDIMAN, W. F., KIDD, R. B.,

-

-

300

C.N. WOLD

THOMAS, E., Er AL. (eds) Initial Reports of the Deep Sea Drilling Project, 94. US Government Printing Office, Washington, DC, 1117-1126. MENARD, H. W. 1969. Elevation and subsidence of oceanic crust. Earth and Planetary Science Letters, 6, 275-284. MONTADERT, L., ROBERTS, O. G., DE CHARPAL, O. t~ GUENNOC, P. 1979. Rifting and subsidence of the northern continental margin of the Bay of Biscay. In: MONTADERT, L., ROBERTS, O. G., ETAL. (eds) Initial Reports of the Deep Sea Drilling Project, 48. US Government Printing Office, Washington, DC, 1025-1060. MOORBATH, S., SIGURDSSON,H. & GOODWIN, R. 1968. K - A t ages of the oldest exposed rocks in Iceland. Earth and Planetary Science Letters, 4, 197-205. MORGAN, J. V., BARTON, P. J. & WHITE, R. S. 1989. The Hatton Bank continental margin - I I I . Structure from wide-angle OBS and multichannel seismic refraction profiles. Geophysical Journal International, 98, 367-384. MORGAN, W. J. 1983. Hotspot tracks and the early rifting of the Atlantic. Tectonophysics, 94, 123139. M~LLER, R. D. & ROEST, W. R. 1992. Fracture zones in the North Atlantic from combined Geosat and Seasat data. Journal of Geophysical Research, 97, 3337-3350. MUTTER, J. C., TALWANI, M. & STOFFA, P. L. 1982. Origin of seaward-dipping reflectors in oceanic crust off the Norwegian margin by 'subaerial seafloor spreading'. Geology, 10, 353-357. NUNNS, A. G. 1982. The structure and evolution of the Jan Mayen Ridge and the surrounding regions. In: WATKINS, J. S. & DRAKE, C. L. Studies in Continental Marine Geology. American Association of Petroleum Geologists Memoir, 34, 193208. 1983. Plate tectonic evolution of the Greenland-Scotland Ridge and surrounding regions. In: BOTT, M. H. P., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) NA TO Conference Series, Series IV." Marine Sciences, Iiol. 8, Structure and Development of the Greenland-Scotland Ridge: New Methods and Concepts. Plenum Press, New York, 11-30. , TALWANI,M., LORENTZEN,G. R., ETAL. 1983. Magnetic anomalies over Iceland and surrounding areas. In: BOTT, M. H. P., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) NATO Conference Series, Series IV." Marine Sciences, Vol. 8, Structure and Development of the GreenlandScotland Ridge: New Methods and Concepts. Plenum Press, New York, 661-678. PALMASON,G. 1974. The insular margin of Iceland. In: BURK, C. A. & DRAKE, C. L. (eds) The Geology of Continental Margins. Springer-Verlag, New York, 375-379. PARK, C.-H., TAMAKI, K. & KOBAYASHI, K. 1990. Age-depth correlation of the Philippine Sea backarc basins and other marginal basins in the world. Tectonophysics, 181, 351-371. PARSON, L. M. & ODP Leg 104 Scientific Party 1988. Dipping reflector styles in the NE Atlantic Ocean.

In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 57-68. PARSONS, B. • SCLATER,J. G. 1977. An analysis of the variation of ocean floor bathymetry and heat flow with age. Journal of Geophysical Research, 82, 803-827. PITMAN, W. C., I I I & TALWANI, M. 1972. Sea floor spreading in the North Atlantic. Geological Society of America Bulletin, 83, 619-646. RAMSAY, J. G. 1969. The measurement of strain and displacement in orogenic belts. In: Time and Place in Orogeny. Geological Society, London, Special Publication, 3, 43-79. ROBERTS, D. G. 1975. Tectonic and stratigraphic evolution of the Rockall Plateau and Trough. In: WOODLAND, A. W. (ed.) Petroleum and the Continental Shelf of North-west Europe, Vol. 1, Geology. Applied Science Publishers, Barking, Essex, UK, 77-89. , BOTT, M. H. P. 8s URUSKI, C. 1983. Structure and origin of the Wyville-Thomson Ridge. In: BOTT, M. H. P., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) NA TO Conference Series, Series IV." Marine Sciences, Iiol. 8, Structure and Development of the Greenland-Scotland Ridge. Plenum Press, New York, 133-158. , MASSON, D. G. & MILES, P. R. 1981. Age and structure of the southern Rockall Trough: New evidence. Earth and Planetary Science Letters, 52, 115-128. , MONTADERT, L. &; SEARLE, R. C. 1979. The western Rockall Plateau: Stratigraphy and structural evolution. In: MONTADERT, L., ROBERTS, D. G., ET AL. (eds) Initial Reports of the Deep Sea Drilling Project, 48. US Government Printing Office, Washington, DC, 1061-1088. , GINZBERG, A., NUNN, K. & McQUILLIN, R. 1988. The structure of the Rockall Trough from seismic refraction and wide-angle reflection measurements. Nature, 332, 632-635. , BACKMAN,J., MORTON, A. C., MURRAY,J. W. & KEENE, J. B. 1984. Evolution of volcanic rifted margins: Synthesis of Leg 81 results on the west margin of the Rockall Plateau. In: ROBERTS, D. G., SCHNITKER, D., ET AL. (eds) Initial Reports of the Deep Sea Drilling Project, 81. US Government Printing Office, Washington, DC, 883-911. ROEST, W. R. & SRIVASTAVA, S. P. 1989. Seafloor spreading in the Labrador Sea: A new reconstruction. Geology, 17, 1000-1003. ROWLEY, D. B. & LOTTES, A. L. 1988. Plate-kinematic reconstructions of the North Atlantic and Arctic: Late Jurassic to Present. Tectonophysics, 155, 73120. SAEMUNDSSON, K. 1974. Evolution of the axial rifting zone in northern Iceland and the Tjornes Fracture Zone. Geological Society of America Bulletin, 8 5 495-504. SCHROEDER, W. 1984. The empirical age--depth relation and depth anomalies in the Pacific Ocean Basin. Journal of Geophysical Research, $9(B12), 9873-9883.

PALAEOBATHYMETRY OF NORTH ATLANTIC & S GIN SEA SCLATER, J. G. & CHRISTIE, P. A. F. 1980. Continental stretching: An explanation of the post midCretaceous subsidence of the central North Sea Basin. Journal of Geophysical Research, 8 5 , 37113739. - & FRANCHETAU, J. 1970. The implications of terrestrial heat flow observations on current tectonic and geochemical models of the crust and upper mantle of the earth. Geophysical Journal of the Royal Astronomical Society, 20, 509-542. , ANDERSON, R. N. & BELL, M. L. 1971. The elevation of ridges and the evolution of the central Eastern Pacific. Journal of Geophysical Research, 76, 7883-7915. , HELLINGER, S. & TAPSCOTT, C. 1977. The palaeobathymetry of the Atlantic Ocean from the Jurassic to the Present. The Journal of Geology, 85, 509-552. , MEINKE, L., BENNETT, A. & MURPHY, C. 1985. The depth of the ocean through the Neogene. In: KENNETT, J. P. (ed.) The Miocene Ocean. Geological Society of America Memoir, 163, 1-19. SCRUTTON, R. A. 1985. Modelling of magnetic and gravity anomalies at Goban Spur, Northeast Atlantic. In: DE GRACIANSKY, P. C., POAG, C. W., ET AL. (eds) Initial Reports of the Deep Sea Drilling Project, 8 0 . US Government Printing Office, Washington, DC, 1141-1151. SHAW, C. A. 1989. Mass balanced paleogeographic modeling: examples from the Western North Atlantic Ocean and Gulf of Mexico. PhD Thesis, University of Colorado, Boulder, USA. SKOGSEID, J. & ELDHOLM, O. 1987. Early Cenozoic crust at the Norwegian continental margin and the conjugate Jan Mayen Ridge. Journal of Geophysical Research, 92, 11 471-11491. SMYTHE, D. K. 1983. Faeroe-Shetland escarpment and continental margin north of the Faeroes. In: BOTT, M. H. P., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) NA TO Conference Series, Series IV: Marine Sciences, Vol. 8, Structure and Development of the Greenland-Scotland Ridge: New Methods and Concepts. Plenum Press, New York, 109-120. SRWASTAVA,S. P. 1978. Evolution of the Labrador Sea and its bearing on the early evolution of the North Atlantic. Geophysical Journal of the Royal Astronomical Society, 52, 313-357. 1985. Evolution of the Eurasian Basin and its implications to the motion of Greenland along the Nares Strait. Tectonophysics, 114, 29-53. & ARTHUR, M. A. 1989. Tectonic evolution of the Labrador Sea and Baffin Bay: Constraints imposed by regional geophysics and drilling results from Leg 105. In: SRIVASTAVA, S. P., ARTHUR, M. A., CLEMENT,B., ETAL. Proceedings of the Ocean Drilling Program, Scientific Results, 105. Ocean Drilling Program, College Station, TX, 989-1009. & TAPSCOTT, C. R. 1986. Plate kinematics of the North Atlantic. In: VOGT, P. R. & TUCHOLKE, B. E. (eds) The Geology of North America, Vol. M:

301

The Western North Atlantic Region. Geological Society of America, Boulder, Colorado, 379-404. , VERHOEF, J. & MACNAB, R. 1988a Results from a detailed aeromagnetic survey across the northeast Newfoundland margin, Part I: Spreading anomalies and relationship between magnetic anomalies and the ocean--continent boundary. Marine and Petroleum Geology, 5, 306-323. & 1988b. Results from a detailed aeromagnetic survey across the northeast Newfoundland margin, Part II: Early opening of the North Atlantic between the British Isles and Newfoundland. Marine and Petroleum Geology, 5, 324-337. STECKLER, M. S. & WATTS, A. B. 1978. Subsidence of the Atlantic-type continental margin off New York. Earth and Planetary Science Letters, 41, 1-13 SURLYK, F., CLEMMENSEN, L. B. & LARSEN, H. C. 1981. Post-Paleozoic evolution of the east Greenland continental margin. In: Geology of the North Atlantic Borderlands. Canadian Society of Petroleum Geologists Memoir, 7, 611~i45. TALWANI, M. & ELDHOLM, O. 1977. Evolution of the Norwegian-Greenland Sea. Geological Society of America Bulletin, 88, 969-999. , UDINTSEV, G., ET AL. 1976. Initial Reports of the Deep Sea Drilling Project, 38. US Government Printing Office, Washington, DC. THIEDE, J. 1980. Palaeo-oceanography, margin stratigraphy and palaeophysiography of the Tertiary North Atlantic and Newfoundland--Greenland Seas. Philosophical Transactions of the Royal Society of London, Series A, 294, 177-185. & ELDHOLM, O 1983. Speculations about the paleodepth of the Greenland-Scotland Ridge during the Late Mesozoic and Cenozoic times. In: BoTr, M. H. P., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) NATO Conference Series, Series IV." Marine Geology, Vol. 8, Structure and Development of the Greenland-Scotland Ridge. Plenum Press, New York, 445-456. TUCHOLKE, B. E. & McCoY, F. W. 1986. Paleogeographic and paleobathymetric evolution of the North Atlantic Ocean. In: VOGT, P. R. & TUCHOLKE, B. E. (eds) The Geology of North America, Vol. M." The Western North Atlantic Region. Geological Society of America, Boulder, Colorado, 589-602. - ~; VOGT, P. R. 1979. Western North Atlantic: Sedimentary evolution and aspects of tectonic history. In: TUCHOLKE, B. E., VOGT, P. R., ET AL. (eds) Initial Reports of the Deep Sea Drilling Project, 43. US Government Printing Office, Washington, DC, 791-825. UNTERNEHR, P. 1982. Etude structurale et cin~,matique de la mer de Norvdge et du Groenland. Evolution du microcontinent de Jan Mayen. Thrse 36me Cycle, Universite de Bretagne Occidentale, Brest, France. VERHOEE, J., ROEST, W. R. & SRIVASTAVA,S. P. 1989. Plate reconstructions and gridded data: A new tool in deciphering correlations across oceans. LOS, 70, 614-618.

302 - - ,

C . N . WOLD

Usow, K. H. & ROEST, W. R. 1990. A new method for plate reconstructions: T h e use of gridded data. Computers & Geosciences. 16, 5174. VINK, G. E. 1982. Continental rifting and the implication for the plate tectonic reconstructions. Journal of Geophysical Research, 87, 10 677-10 688. VOGT, P. R. 1986. Geophysical and geochemical signatures and plate tectonics. In: HURDLE, B. G. (ed.) The Nordic Seas. Springer-Verlag, New York, 413-662. & AVERY, O. E. 1974. Detailed magnetic surveys in the Northeast Atlantic and Labrador Sea. Journal of Geophysical Research, 79, 99499959. - - , JOHNSON, G. L. & KRISTJANSSON, L. 1980. Morphology and magnetic anomalies north of Iceland. Journal of Geophysics, 47, 67-80. WEBER, J. R. 1990. The structures of the Alpha Ridge, Arctic Ocean and Iceland-Faeroe Ridge, North Atlantic: Comparisons and implications for the evolution of the Canada Basin. Marine Geology, 93, 43-68.

WESSEL, P. & SMITH, W. H. F. 1991. Free software helps map and display data. LOS, 72, 441--446. WHITE, R. S. 1988 A hot-spot model for early Tertiary volcanism in the N Atlantic. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 3-13. 1989. Initiation of the Iceland Plume and Opening of the North Atlantic. In: TANKARD, A. J. & BALKWlI~L, H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins. American Association of Petroleum Geologists Memoir, 46, 149-154. 1992. Crustal structure and magmatism of North Atlantic continental margins. Journal of the Geological Society, London, 149, 841-854. - - , SPENCE, G. D., FOWLER, S. R., MCKENZIE, D. P., WESTBROOK,G. K. & BOWEN,A. N. 1987. Magmatism at rifted continental margins. Nature, 330, 439-444. WOLD, C. N. 1992. palaeobathymetry and sediment accumulation in the northern North Atlantic and southern Greenland-Iceland-Norwegian Sea. PhD Thesis, Christian Albrechts University, Kiel, FRG.

Index References to figures and tables are printed in italics. acoustic impedance analysis 95-7 Aegir Ridge 272, 278 aegirine 190 Amazon Cone 200, 201,203 amphibole 189, 190 amplitude of seismic signal Galicia margin studies 76-8, 97-9 analogue modelling lithospheric stretching 86-7 Antarctic Bottom Water (AABW) 208-12 Antarctissa whitei 220 Anton Dohrn Seamount 161, 272 apparent uplift/erosion 262 Arctic Bottom Current 154-5, 155 Argo Formation 5, 7 Ascension Fracture Zone 200, 201 aseismic ridges 281 Atlantic Ocean (Equatorial) bathymetry 201 palaeoceanography Cretaceous 207-8 Jurassic-Cretaceous 199-201 Tertiary 208-12 plate reconstructions 203-7 seismic stratigraphy 201-3 Atlantic Ocean (North) continent-ocean boundary 273-7 palaeobathymetric reconstruction 271-2 backstripping methods isostasy effects 288-9 removal of sediment 284 sea level effects 288 sediment decompaction 285 thermal subsidence effects 286-7 boundary conditions used lithosphere thermal age 272-3 plate boundaries 273-7 plate reconstruction 277-9 sea level 273 stratigraphy and lithology 272 lithosphere age reconstruction 281-4 lithosphere subsidence effects 280-1 previous research methods 279-80 results of reconstruction 289-97 plate setting 277-9 Avalon Formation 6, 7, 19, 20 backstripping 271,280, 284 application to N Atlantic reconstruction 284-9 Balder event 65 Balder Formation 191, 193 Balder Tufts 65 Banff fault 250, 254 Banquereau Formation 7 Barra Fan 161,163 basalt volcanism 275 Faeroe Islands 125, 134

Faeroe-RockaU Plateau 145--6 basaltic tuff Faeroe Island Shelf 182, 184-5, 189, 192-3 correlation with North Sea 193-4 Bear Island Trough Mouth Fan 173 Ben Nevis Formation 6, 7, 19 Bill Bailey Bank basement 126-9 sediment deformation 133-4 seismic stratigraphy 129-32 setting 125, 272 subsidence history 134-7 biostratigraphy Atlantic Ocean (North) radiolaria 217-20 Faeroe Island Shelf sediments 186, 187-8, 189, 190, 192 St Kilda Basin 232-3 biotite 190 bioturbation 16, 192 Biscay, Bay of 274 Bjorn Drift 272 Blosseville Group 191 Bonnition basin 4 break-up unconformity 1, 2, 78 burial depth studies 258-61

14C dating uncertainties 227, 241-3 cap rocks, Faeroe-Shetland basin 67 Catalina Formation 7 Ceara Abyssal Plain 200, 201 Ceara Rise 200, 201 Central Ridge 54 Chain Fracture Zone 200, 201 Chalk Group 255 Charlie Gibbs Fracture Zone 276 Chemehuevi detachment 98 CIPW norm 237 Clair basin 53 Clair Ridge 53 clinopyroxene 182, 189, 192 coal, Faeroe-Shetland basin 65 collophane 189 compaction analysis, Moray Firth (Inner) 253-8 compression events in Eocene Faeroes 153, 155 Norwegian Sea 195 in lithosphere 265-6 structures in Faeroe-Rockall 142, 215-16 continent--ocean boundary Galicia margin segmentation 72-6 structural analysis methods 76-8 results lower layers 79-82 seismic Moho 82-5

304

INDEX

upper layers 78-9 results discussed 85-8 GIN Sea 273-7, 287 Cretaceous Atlantic Ocean (Equatorial) circulation 207-8 reconstruction 204, 205 sediments 202 Atlantic Ocean (North) extension 277 Jeanne d'Arc basin 6, 7, 33, 45-7 deformation 11-12 unconformities 14-19 lithosphere rifting 72 Porcupine basin 33, 45-7 seafloor spreading 4, 23-5 Cretaceous rifting model North Atlantic/GIN Sea reconstruction 298-97 Cromer Knoll Group 256 crustal studies Galicia margin ocean-continent boundary segmentation 72-6 structural analysis methods 76-8 results lower layers 79-82 seismic Moho 82-5 upper layers 78-9 results discussed 85-8 GIN Sea 287 continent-ocean boundary recognition 273-7 see also lithosphere Cuillin Package 64 currents, Faeroe-Rockall area 141-3 Cycladophora davisiana 219 Dawson Canyon Formation 7 Dead Mountains detachment fault 98 decompaction effects 271,280, 284, 285 Demerara Abyssal Plain 200, 201 Demerara Plateau 200, 201 Denmark Strait 272, 289, 296, 297 detachment faults 93, 104-5 Devensian NW Britain slope-aprons 163, 165 diatoms, Palaeocene 60 dinoflagellates Faeroe Island Shelf dredge haul 186, 187-8, 189, 190, 192 St Kilda Basin 232-3 Voring Plateau 194 domino fault block model 103 Downing Formation 7 dredge haul studies Faeroe Island Shelf methods of sampling 179-80 results of analysis 181-4 sample ages 184-92 results discussed 192-5 Dunrobin Group 256 East Faeroe High 147, 149 Eastern Shoals Formation 7

Edoras Bank 272 Eirik Drift 272 Eocene Atlantic Ocean (Equatorial) circulation 209 reconstruction 206 sediments 202 Atlantic Ocean (North) reconstruction 293, 294, 297 Bill Bailey/Lousy Banks subsidence 134, 137 Faeroe Bank Channel 155 Faeroe Island Shelf dredge 189, 189-90 regression evidence 195 transgression evidence 194-5 Faeroe Plateau and Channel sediments 149-50, 153 Faeroe-RockaU Plateau compression 215 Faeroe-Shetland basin, stratigraphy 65-6 GIN Sea reconstruction 293, 294, 297 Eocene-Oligocene boundary Lousy/Bill Bailey Bank subsidence 135, 137 unconformity 132-3 Eriador Seamount 272 erosion studies Moray Firth (Inner) compaction analysis 253-8 driving mechanisms 264--6 uplift analysis 261-4 vitrinite reflectance data 258-61 Eurydice Formation 5, 7 facies analysis Jeanne d'Arc basin 36-41 Porcupine basin 44-5 Faeroe Bank 125, 272 Faeroe Bank Channel 142, 145, 148, 161, 272, 289 Eocene 149-50, 153, 155 Miocene 151 Oligocene 150 Recent 153 Faeroe Block continental crust thickness 276, 287 Palaeocene-Eocene boundary 191 reflector sequence 194 see also Faeroe Island Shelf Faeroe Channel Knoll 147, 148, 153, 155 Faeroe Channel Knoll Escarpment 147, 148 Faeroe Channel Knoll Plateau 149 Faeroe Drift 272 Faeroe Islands basalt lavas 125, 134, 146-8, 215 seismic study of basalt 112-22 setting 111-12 Faeroe Islands Platform basaltic basement 146-8 sediment stratigraphy Eocene 149-50 Miocene 151 Oligocene 150-1 Pliocene 151 Recent 153 sediment thickness 148-9 setting 145-6 Tertiary evolutionary history 153-7 Faeroe Islands Shelf 160, 272

INDEX dredge haul study 179-95 setting 179 see also Faeroe Block Faeroe Ridge 289 Faeroe-Rockall Plateau 142, 215-16 Faeroe~Shetland Basin correlation with North Sea 66-7 hydrocarbon prospectivity 51, 67-8 plate tectonic environment 54-5 seismic mapping 56-7 sequence stratigraphy 55-6, 58-66 setting 52-4 structure 57-8 Faeroe~Shetland Channel 142, 145, 148, 155, 160, 161, 272, 289 dredge haul 184 evolution Cretaceous 125 Eocene 149-50 Miocene 151 Oligocene 150 Pliocene 151, 165 Recent 153 palaeobathymetric reconstruction 289, 295, 296, 297 Faeroe-Shetland Escarpment 147, 148, 153, 272 seaward dipping reflector 275 fans see slope-aprons feldspathic sandstone 182, 190-2 Fladen Group 256 Flett Ridge 54 flood basalts 275 foraminifera Atlantic Ocean (Equatorial) 207 Faeroe Island Shelf 182, 192 Faeroe-Shetland Basin 58, 61, 65 St Kilda Basin 232-3 Fortune Bay Formation 7 Foula Wedge 161-2, 163, 172 Four North Fracture Zone 200, 201 Fugloy Ridge 147, 149, 150 Fur Formation 193 Galicia margin crustal structure enigmatic terrane (ET) 79 method of analysis 76-8 results 78-85 results discussed 85-6 summary of characteristics 88-9 modelling formation 86-8 OCB segmentation 72-6 S reflector study amplitude 97-9 analysis of polarity and waveform 95-7 relationships with faults 99 results 101-5 relationships with peridotite ridge 105-7 significance of 94-5 summary of features 107-8 Gambia Abyssal Plain 200, 201 gamma ray logs Jeanne d'Arc basin 40 Moray Firth (Inner) 256

305

Porcupine basin 43 Gardar Drift 272 gas, Faeroe Islands 112 Geikie Escarpment 161,162, 163, 164 geochemistry Faeroe Island Shelf dredge haul 183-4, 185, 189 St Kilda Basin tephrochronology study methods 229-30 results 233-5, 244-6 George Bligh Bank 272 GIN (Greenland-Iceland-Norwegian) Sea 271 palaeobathymetric reconstruction 271-2 backstripping effects 284-9 boundary conditions used 272-9 lithosphere age reconstruction 281-4 lithosphere subsidence effects 280-1 previous research methods 277-9 results of reconstruction 289-97 see also Greenland Sea; Iceland Sea, Norwegian Sea glacigenic sedimentation see ice-sheet sedimentation glauconite 182, 192 Gloria Drift 272 Goban Spur 274 Grand Banks-Iberia seafloor spreading 4 Great Glen fault 250, 251, 253, 254 Greenland (East) continent-ocean boundary 276 marginal fans 173 Palaeocene-Eocene boundary 191 Greenland Sea 271 see GIN Sea Greenland-Norway Sea see Nordic Seas Greenland-Scotland Ridge 276-7, 278, 297 Grimsvrtn 241 Guinea Fracture Zone 201,202 Guinea Plateau 200, 201 Halibut Horst 250 Hatton Bank 272, 287, 289, 295, 296, 297 Hatton Drift 272 Hatton-Rockall Basin 277 Heather Formation 256 heavy mineralogy, Faeroe-Shetland basin 63-4 Hebrides Shelf 160, 161, 163 Hebrides slope-apron lithology 168-71 setting 160-2 time of formation 162-3 Hebrides Terrace Seamount 161, 272 Helmsdale fault 250, 251, 253, 254 Hibernia Formation 6, 7 Hidra Formation 255, 256 Hod Formation 255, 256, 257 hot spots 54-5, 62, 65, 265, 275, 281 hydrocarbons, Faeroe Islands 112 Iberia Abyssal Plain 72 Iberia~3rand Banks, seafloor spreading 4 ice-rafting 159, 184, 236 ice-sheet sedimentation slope-aprons 159 NE Atlantic 173-4

306

INDEX

NW Atlantic 174 NW Britain 160-5, 168-72 Iceland, tephra formation in 238-9, 240-1 Iceland Hotspot 265 Iceland Plateau 272 Iceland Sea 271 radiolarian abundance 220 see also GIN Sea Iceland-Faeroe Ridge 272, 276, 277, 281,287, 295, 296 ichnofacies 16 ilmenite 182, 189 inertinite 189 Inner Moray Firth see Moray Firth (Inner) intraplate stress 265 Irminger Basin 272, 295, 296, 297 Iroquois Formation 7 Isengard Drift 272 Ist]orden Fan 173 isopach map, Jeanne d'Arc basin 22, 32 Ivory Coast Rise 200, 210 Jan Mayen microplate 277-8 Jan Mayen Ridge 272, 289, 295, 296, 297 Jeanne d'Arc basin correlation with Porcupine basin 45-7 deformational history 12-14 extension 23 facies analysis 36--41 initiation 30 lithostratigraphy 7, 31 sedimentary thicknesses 21-2 sequence and lithology correlation 14-19 sequence stratigraphy 8-12 setting 4-6 subsidence history 19-21, 33-6 Jeanne d'Arc Formation 6, 7 Judd fault 54 Jurassic Atlantic Ocean (Equatorial), sediments 202 Jeanne d'Arc basin sedimentation 6, 7 Scotian basin 4 Kane Gap 200, 201,208, 211 Kap Dalton Formation 191 Katla 241 kerogen, Faeroe-Shetland basin 64, 65 Kimmeridge Clay Formation 255, 256, 257 Kintail Package 64 Kolbeinsey Ridge 272 278, 295, 296, 297 Labrador Slope 174, 175 last glacial maximum (LGM) 228 lithic volcanic sandstone, Faeroe Island Shelf 182, 190-2 lithosphere age calculation 281-4 compression 265-6 extension modelling 93-4 stretching 86-7 subsidence 280-1,286-7 thermal age 272-3

thickening, role in uplift/erosion of 264 Lopra- 1 well 112 Lossiemouth fault 250, 254 Lousy Bank (Outer Bill Bailey Bank) 272, 297 basement 126-9 sediment deformation 133-4 seismic stratigraphy 129-32 setting 125 subsidence history 134-7 low angle detachment faults 93, 104-5 magrnatism, Porcupine basin 33 mantle lithosphere compression 265-6 Marl Formation 255 Mascarene Plateau 281 Mercury border fault 5, 9 Mexico, Gulf of 280 microflora, Faeroe--Shetland basin 58, 61-2, 64, 65 Miocene Arctic Bottom Current 155 Atlantic Ocean (Equatorial), sediments 202 Faeroe Bank Channel 142, 155 Faeroe Plateau and Channel sediments 142-3, 151, 154 Faeroe-RockaU Plateau compression 216 GIN Sea/N Atlantic reconstruction 290, 291, 295-6 Lousy/Bill Bailey Bank unconformity 133, 135, 137 slope-apron sediments 161 Wyville-Thomson Ridge 142, 155 Ymir Ridge 142 modelling Cretaceous rifting model North Atlantic/GIN Sea reconstructions 289-97 lithosphere extension 93-4 lithosphere stretching 86-7 Palaeocene reheating model North Atlantic/GIN Sea reconstructions 289-97 seismic 118-20 Moho scattering reflective 85 seismic 82-5 molluscs, St Kilda Basin 232, 233 Montrose group 256 Moray Firth (Inner) basin inversion 249-50 erosion history driving mechanisms 264-6 Sediment compaction data 253-8 uplift analysis 261-4 vitrinite reflectance data 258-61 extensional reactivation 250-1 regional tilting 252 strike-slip reactivation 251-2 Munkegrunnur Ridge 147, 149, 150, 215 Murre border fault 5, 9 Nautilus fault 34 Nautilus Formation 7 Newfoundland offshore 274 Newfoundland Basin 88 Newfoundland Slope 174, 175 see also Jeanne d'Arc basin

INDEX Niger Cone 203 nontronite 182 Nordic Seas currents 141-2 North Atlantic see Atlantic Ocean (North) North Atlantic Ash Zone 237-8 North Minch tephra 239 North Sea Palaeocene-Eocene boundary 191 rift system 265 tuff correlations 193-4 Norway Basin 272 Norwegian Sea 153, 154, 271 Eocene plate setting 195 Miocene 151 Oligocene 150, 154 Plio-Pleistocene radiolaria 218, 219 see also GIN Sea Norwegian~3reenland Sea 173-4 ocean-continent boundary Galicia margin study crustal structure methods of analysis 76-8 results 78-85 results discussed 85-6 modelling formation 86-8 segmentation 72-6 summary of characters 88-9 GIN Sea 273-7, 287 oceanography Atlantic Ocean (Equatorial) Cretaceous 207-8 Tertiary 208-12 oil in Faeroe Islands 112 Oligocene Atlantic Ocean (Equatorial) circulation 211 reconstruction 206, 207 Atlantic Ocean (North) 292, 296 Bill Bailey/Lousy Banks 137 Faeroe Bank Channel 155 Faeroe Island Shelf dredge 190-2, 195 Faeroe Plateau and Channel sediments 150-1, 154 Faeroe-Rockall Plateau compression 215 GIN Sea 292, 296 NW Britain erosion features 161 slope-apron sediments 161 Orphan basin 4 ostracods, St Kilda Basin 232-3 Outer Bill Bailey Bank see Lousy Bank P waves Faeroe basalt study reflected 117-18 transmitted 116-17 Pacific Ocean lithosphere 281 palaeobathymetry of GIN Sea and North Atlantic reconstruction methods 271-2 backstripping processes 284-9 boundary conditions used 272-9 lithosphere age effects 281-4 lithosphere subsidence effects 280-1

previous research methods 279-80 reconstruction results 289-97 Palaeocene Atlantic Ocean (Equatorial) reconstruction 206 sediments 202 Atlantic Ocean (North) volcanism 289 Bill Bailey/Lousy Banks basalt lavas 134 Faeroe Islands coastline 148 shelf dredge haul 189 volcanism 153 Faeroe-Shetland basin hydrocarbon prospectivity 51 plate setting 54 sequence interpretation 58-65 stratigraphy 55-6 Rockall-Faeroe-Greenland breakup 125 Palaeocene reheating model North Atlantic/GIN Sea reconstruction 289-97 Palaeocene-Eocene boundary 191 palaeoenvironment analysis Jeanne d'Arc Formation 14-19 Palaeophycus tubularis 192 palagonite 182, 185, 190 passive margins characteristics 1-2, 71 see Galicia margin Peridotite Ridge 94, 105-7 Pernambuco Abyssal Plain 200, 201 phosphoritic pelletstone 182, 189, 190 plagioclase 182, 185, 190, 192 plant fossils 189, 190 plate tectonics characteristics of plate margins 93 Eocene plate reorganization 195 reconstructions for North Atlantic gridding technique 281-4 past 277-8 present 278-9 Pleistocene radiolaria 223 see also Quaternary Plenus Formation 255 Plio-Pleistocene NW Britain slope-aprons 161, 162, 163-5 radiolarian zones 218, 219-20 Pliocene Faeroe Bank Channel 155 Faeroe Plateau and Channel sediments 151-2 unconformity 133 polarity analysis 95-7 pollen analysis 189, 190 Porcupine Abyssal Plain 272 Porcupine basin correlation with Jeanne d'Arc basin 45-7 facies analysis 44-5 initiation 30 lithostratigraphy 31 subsidence 42-4 porosity 285 pre-stack depth migration method 100-1 results 101-5

307

308

INDEX

pyroclastics 275 Q factor 77 Quaternary climatic fluctuations 227, 228, 242 Devensian slope-aprons 163, 165 Faeroe Island Shelf sediments 179 St Kilda tephrochronology 233-5 correlation 236-41 transport 235-6 Younger Dryas 228-9, 239, 240, 242 radiolaria Faeroe-Shetland basin 60 latitude effects on 217-19 North Atlantic zones 219-20 palaeoceanography 220-2 Rankin fault 34 Rankin Formation 6, 7 reflectivity, seismic 77 reflector sequence, Faeroe Island Shelf 149, 153, 194 reheating model, Palaeocene 289-97 reservoir character, Faeroe-Shetland basin 67 Reykjanes Ridge 272, 295, 296 ridge push 216 rifted margins 93 rifting Cimmerian Jeanne d'Arc basin 33 Porcupine basin 33 Palaeocene-Eocene North Atlantic 264-5 Triassic Jeanne d'Arc basin 30 Porcupine basin 30 rifting model, Cretaceous 289-97 Rio Grande Rise 281 Rockall Bank 272, 276, 287, 289, 295, 296, 297 Rockall Plateau 125, 160, 272 Palaeocene-Eocene boundary 191 reflector sequence 194 seaward dipping reflector 275-6 Rockall Trough 125, 160, 161, 272, 277, 287, 295, 296, 297 Rockall-Faeroe microcontinent 125 Romanche Fracture Zone 200, 201 Romanche Transform 200, 201 Rona Ridge 53 Rona Wedge 161-2, 163, 172 Rosemary Bank 272 rutile 190 S reflector analysis of amplitude analysis 97-9 polarity waveform analysis 95-7 relationships with faults 99 peridotite ridge 105-7 significance of 94-5 summary of features 107-8

S waves Faeroe basalt study reflected 117-18 transmitted 116-17 St Kilda Basin core sample analysis 231 biostratigraphy 232-3 lithostratigraphy 231-2 tephrochronology 233-5 core sample dating 241-3 core sample interpretation tephra correlation 236-41 tephra transport 235--6 setting 230-1 St Paul Fracture Zone 200, 201 Salar basin 4 scattering reflective Moho (SRM) 85 Scoresby Fan 173 Scoresby Sund 272 Scotian Slope 174 sea level curves, reconstruction of 273, 288 seafloor spreading Grand Banks-Iberia 4, 23-5 relation to unconformities 3 seal rocks, Faeroe-Shetland basin 67 seaward dipping reflectors 274-6 Faeroe Island Shelf 149, 153, 194 segmentation, OCB 72-6, 88 seismic methods Faeroe basalt study data acquisition 112-15 data processing 115-16 interpretation 116-18 modelling 118-20 results discussed 121-2 seismic Moho 82-5 seismic refraction 76 seismic reflectors (S reflectors) Galicia margin 79-82, 85-6 seismic sections Bill Bailey Bank 127, 128 Faeroe Bank Channel 155, 156, 157 Faeroe Bank-Bill Bailey Bank 141 Faeroe Island Shelf 181 Faeroe Plateau 151, 152 Faeroe-Shetland Channel 154 Galicia margin 75, 80, 82, 83, 84 Jeanne d'Arc basin 8, 10, 11, 13, 39 Lousy Bank 127 Moray Firth (Inner) 250, 251, 252, 253, 254 Porcupine basin 43 seismic signal amplitude analysis 76-8, 97-9 seismic velocities 76, 77 sequence stratigraphy Faeroe-Shetland basin 58-66 Jeanne d'Arc basin 8-12, 14-19, 36-41 Porcupine basin 44-5 sideromelane 182 Sierra Leone Abyssal Plain 200, 201 Sierra Leone Rise 200, 201 Sigmundur Seamount 142, 143 similarity coefficients 236-7 Sinclair Horst 250 Sk6gar tephra 238-9

INDEX slope-aprons introduction 159 NE Atlantic 173-4 NW Atlantic 174 NW Britain 160-5, 168-72 smectite 182, 192 Snorri Drift 272 solidity 284, 285 sonic logs Jeanne d'Arc basin 40 Moray Firth (Inner) 256 Porcupine basin 43 South Iceland Basin 272, 295, 296, 297 Spain see Galicia margin Sphaeropyle langii 220 Stichocorys peregrina 220 Strakhov Fracture Zone 200, 201 Stylatractus universus 219 Sula Sgeir Fan 161,162, 163, 166, 167, 168, 175 surface seismic profile 112, 115 Sutherland Terrace 251-2 tachylite 182, 185 tephrochronology applications in St Kilda Basin 233-5 tephra correlation 236-41 tephra transport 235-6 principles 227-8, 229 terrane analysis 79 Tertiary Faeroe channels 148-51 Faeroe-Rockall Plateau basalts 145-6 sediments 145-6 Great Glen fault 251 Jeanne d'Arc basin 7 Moray Forth (Inner) erosion 253-61 structures 250 Porcupine basin 33 Tertiary Igneous Province 275 thanatocenoses, radiolarian 220-2 thermal age 272-3 thermal uplift, role in erosion of 264 titanaugite 189 Tor Formation 255, 256, 257 Torfadalsvatn tephras 239 trace element analysis Faeroe Island Shelf dredge haul 183-4, 185, 189 trace fossils 16, 192 Triassic Jeanne d'Arc basin 5, 7, 30 tuff, basaltic Faeroe Island Shelf 182, 184-5, 189, 192-3 correlation with North Sea 193-4 tuffaceous limestone, Faeroe Island Shelf 182, 189-90 turbidites 78 unconformities defined 1

recognition at passive margins 1-2 Jeanne d'Arc basin studies Albian 18-19 Aptian 16-18 Barremian 14-16 Tertiary East Greenland 195 Faeroe Plateau 155, 195 Faeroe-Rockall 141 uplift studies, Moray Firth (Inner) 261-4 Upwelling Radiolarian Index (URI) 222 upwelling zones 222 Vedde ash 238-9 Veidiv6tn 240-1 Vema Fracture Zone 200, 201,209 vertical seismic profile (VSP) 112, 115, 116 Victory basin 53 Victory Ridge 53 Vigo seamounts 73 vitrinite reflectance 252, 258-61 volcaniclastics Faeroe Island Shelf dredge study basaltic tufts 184-9, 192-3 lithic volcanic sandstones 190-2 tuffaceous limestones 189-90 volcanism, basaltic 275 Faeroe Islands 125, 134 Faeroe Island Shelf 182, 184-5, 189, 192-3 Faeroe-Rockall Plateau 145-6 Voring Plateau 275 Palaeocene-Eocene boundary 191 reflector sequence 194 Voyager Formation 7 walk-away vertical seismic profile 112, 114, 116 Walvis Ridge 281 waveform analysis, Galicia margin 95-7 West Galicia margin see Galicia margin West Lewis Basin 216 West Shetland basin 53 West Shetland Shelf 160 161 West Shetland slope-apron lithology 171-2 setting 160-2 time of formation 163-5 Westray Ridge 54 Whipple Mountain detachment 103 Whiterose Formation 6, 7 Wick fault 250, 254 WyviUe-Thomson Ridge 147, 160, 161, 272, 289 formation 142, 155, 215, 216 Ymir Ridge 142, 147, 215, 216 Younger Dryas 228-9, 239, 240, 242 zeolite 182, 185, 192

309

The Tectonics, Sedimentation and Palaeoceanographyof the North Atlantic Region edited by R.A. Scrutton, M.S. Stoker, G.B. Shimmield

and A.W. Tudhope The North Atlantic region is an excellent natural laboratory in which to study the tectonics, sedimentation and palaeoceanography of an evolving oceanic rift basin. Sandwiched between the active research communities of North America and Europe, and with its margins targeted for hydrocarbon exploration, it is not surprising that a remarkable level of understanding has been reached of the interplay between these three disciplines. Yet there are still important questions to be addressed - by the active geophysical programmes on the Mid-Atlantic Ridge system and the passive margins, by ongoing Ocean Drilling Program work and by hydrocarbon exploration in frontier areas in more hostile North Atlantic waters. Just one topic that illustrates what the North Atlantic has to offer ~ as a natural laboratory is the research into oceanic gateways, such as the North Atlantic-Arctic Gateway and the Greenland-Scotland Ridge. These features are created by the tectonics of the basin, the sedimentary record documents the history of their development and the palaeoceanography was strongly influenced by the circulation patterns permitted through the gateways. This volume is aimed at a very wide audience. Although there is material in this book of interest to almost all geoscientists working in the North Atlantic region, there is a focus of papers on the basin margins, and on the NW European margin in particular, covering aspects from Mesozoic rifting to Quaternary sedimentation. Papers on the evolution of the Grand Banks and Iberian passive margins, and sedimentation over the Iceland-Scotland Ridge and in the equatorial Atlantic gateway all relate strongly to the Ocean Drilling Program, whilst of interest to the oil industry will be a number of papers on shelf basins, such as the Jeanne D'Arc and the Moray Firth. •

320 pages

•

195 illustrations

•

17 chapters

•

index

Cover illustration: Palaeocene reheating model for reconstructed mid-Oligocene (30 Ma) palaeobathymetry (see p.292).

ISBN

1-897799-27-6