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
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

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




-~

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

.~ 5= o_o

B

600"

on-

3~

A

400" ~UJ ,
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