Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup
The Geological Society of London Books Editorial Committee Chief Editor
BOB PANKHURST (UK) Society Books Editors
JOHN GREGORY (UK) JIM GRIFFITHS (UK) JOHN HOWE (UK) PHIL LEAT (UK) NICK ROBINS (UK) JONATHAN TURNER (UK) Society Books Advisors
MIKE BROWN (USA) ERIC BUFFETAUT (FRANCE )
RANDELL
RETO GIERE´ (GERMANY ) JON GLUYAS (UK) DOUG STEAD (CANADA ) STEPHENSON (THE NETHERLANDS )
Geological Society books refereeing procedures The Society makes every effort to ensure that the scientific and production quality of its books matches that of its journals. Since 1997, all book proposals have been refereed by specialist reviewers as well as by the Society’s Books Editorial Committee. If the referees identify weaknesses in the proposal, these must be addressed before the proposal is accepted. Once the book is accepted, the Society Book Editors ensure that the volume editors follow strict guidelines on refereeing and quality control. We insist that individual papers can only be accepted after satisfactory review by two independent referees. The questions on the review forms are similar to those for Journal of the Geological Society. The referees’ forms and comments must be available to the Society’s Book Editors on request. Although many of the books result from meetings, the editors are expected to commission papers that were not presented at the meeting to ensure that the book provides a balanced coverage of the subject. Being accepted for presentation at the meeting does not guarantee inclusion in the book. More information about submitting a proposal and producing a book for the Society can be found on its web site: www.geolsoc.org.uk.
It is recommended that reference to all or part of this book should be made in one of the following ways: KARNER , G. D., MANATSCHAL , G. & PINHEIRO , L. M. (eds) 2007. Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, London, Special Publications, 282. HOPPER , J. R., FUNCK , T. & TUCHOLKE , B. E. 2007. Structure of the Flemish Cap margin, Newfoundland: insights into mantle and crustal processes during continental breakup. In: KARNER , G. D., MANATSCHAL , G. & PINHEIRO , L. M. (eds) Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, London, Special Publications, 282, 47– 61.
GEOLOGICAL SOCIETY SPECIAL PUBLICATION NO. 282
Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup
EDITED BY
G. D. KARNER, ExxonMobil Upstream Research Company, Houston, USA
G. MANATSCHAL Universite´ Louis Pasteur, Strasbourg, France and
L. M. PINHEIRO Universidade de Aveiro, Aveiro, Portugal
2007 Published by The Geological Society London
THE GEOLOGICAL SOCIETY The Geological Society of London (GSL) was founded in 1807. It is the oldest national geological society in the world and the largest in Europe. It was incorporated under Royal Charter in 1825 and is Registered Charity 210161. The Society is the UK national learned and professional society for geology with a worldwide Fellowship (FGS) of over 9000. The Society has the power to confer Chartered status on suitably qualified Fellows, and about 2000 of the Fellowship carry the title (CGeol). Chartered Geologists may also obtain the equivalent European title, European Geologist (EurGeol). One fifth of the Society’s fellowship resides outside the UK. To find out more about the Society, log on to www.geolsoc.org.uk. The Geological Society Publishing House (Bath, UK) produces the Society’s international journals and books, and acts as European distributor for selected publications of the American Association of Petroleum Geologists (AAPG), the Indonesian Petroleum Association (IPA), the Geological Society of America (GSA), the Society for Sedimentary Geology (SEPM) and the Geologists’ Association (GA). Joint marketing agreements ensure that GSL Fellows may purchase these societies’ publications at a discount. The Society’s online bookshop (accessible from www.geolsoc.org.uk) offers secure book purchasing with your credit or debit card. To find out about joining the Society and benefiting from substantial discounts on publications of GSL and other societies worldwide, consult www.geolsoc.org.uk, or contact the Fellowship Department at: The Geological Society, Burlington House, Piccadilly, London W1J 0BG: Tel. þ44 (0)20 7434 9944; Fax þ44 (0)20 7439 8975; E-mail:
[email protected]. For information about the Society’s meetings, consult Events on www.geolsoc.org.uk. To find out more about the Society’s Corporate Affiliates Scheme, write to
[email protected]. Published by The Geological Society from: The Geological Society Publishing House, Unit 7, Brassmill Enterprise Centre, Brassmill Lane, Bath BA1 3JN, UK (Orders: Tel. þ44 (0)1225 445046, Fax þ44 (0)1225 442836) Online bookshop: www.geolsoc.org.uk/bookshop The publishers make 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. # The Geological Society of London 2007. All rights reserved. No reproduction, copy or transmission of this publication may be made without 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 W1P 9HE. Users registered with the Copyright Clearance Center, 27 Congress Street, Salem, MA 01970, USA: the item-fee code for this publication is 0305-8719/07/$15.00. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. ISBN: 978-1-86239-228-1 Typeset by Techset Composition Ltd, Salisbury, UK Printed by MPG Books Ltd, Bodmin, UK Distributors North America For trade and institutional orders: The Geological Society, c/o AIDC, 82 Winter Sport Lane, Williston, VT 05495, USA Orders: Tel þ 1 800-972-9892 Fax þ 1 802-864-7626 E-mail
[email protected] For individual and corporate orders: AAPG Bookstore, PO Box 979, Tulsa, OK 74101-0979, USA Orders: Tel þ 1 918-584-2555 Fax þ 1 918-560-2652 E-mail
[email protected] Website http://bookstore.aapg.org India Affiliated East-West Press Private Ltd, Marketing Division, G-1/16 Ansari Road, Darya Ganj, New Delhi 110 002, India Orders: Tel þ91 11 2327-9113/2326-4180 Fax þ91 11 2326-0538 E-mail
[email protected]
Contents KARNER , G. D., MANATSCHAL , G. & PINHEIRO , L. M. Imaging, mapping and modelling continental lithosphere extension and breakup: an introduction The Iberia– Newfoundland continental extensional system (geological and geophysical constraints) TUCHOLKE , B. E., SAWYER , D. S. & SIBUET , J.-C. Breakup of the Newfoundland–Iberia rift HOPPER , J. R., FUNCK , T. & TUCHOLKE , B. E. Structure of the Flemish Cap margin, Newfoundland: insights into mantle and crustal processes during continental breakup SIBUET , J.-C., SRIVASTAVA , S. P., ENACHESCU , M. & KARNER , G. D. Early Cretaceous motion of Flemish Cap with respect to North America: implications on the formation of Orphan Basin and SE Flemish Cap –Galicia Bank conjugate margins RESTON , T. J. The formation of non-volcanic rifted margins by the progressive extension of the lithosphere: the example of the West Iberian margin The Iberia– Newfoundland continental extensional system (dynamic modelling) HUISMANS , R. S. & BEAUMONT , C. Roles of lithospheric strain softening and heterogeneity in determining the geometry of rifts and continental margins BUROV , E. The role of gravitational instabilities, density structure and extension rate in the evolution of continental margins HARRY , D. L. & GRANDELL , S. A dynamic model of rifting between Galicia Bank and Flemish Cap during the opening of the North Atlantic Ocean The Iberia– Newfoundland continental extensional system (kinematic modelling) EGAN , S. S. & MEREDITH , D. J. A kinematic modelling approach to lithosphere deformation and basin formation: application to the Black Sea HEALY , D. & KUSZNIR , N. J. Early kinematic history of the Goban Spur rifted margin derived from a new model of continental breakup and sea-floor spreading initiation Observational characteristics of non-Atlantic extensional systems (offshore) GOODLIFFE , A. M. & TAYLOR , B. The boundary between continental rifting and sea-floor spreading in the Woodlark Basin, Papua New Guinea DIREEN , N. G., BORISSOVA , I., STAGG , H. M. J., COLWELL , J. B. & SYMONDS , P. A. Nature of the continent– ocean transition zone along the southern Australian continental margin: a comparison of the Naturaliste Plateau, SW Australia, and the central Great Australian Bight sectors Observational characteristics of non-Atlantic extensional systems (onshore) COCHRAN , J. R. & KARNER , G. D. Constraints on the deformation and rupturing of continental lithosphere of the Red Sea: the transition from rifting to drifting MANATSCHAL , G., MU¨ NTENER , O., LAVIER , L. L., MINSHULL , T. A. & PE´ RON -PINVIDIC , G. Observations from the Alpine Tethys and Iberia–Newfoundland margins pertinent to the interpretation of continental breakup ROBERTSON , A. H. F. Overview of tectonic settings related to the rifting and opening of Mesozoic ocean basins in the Eastern Tethys: Oman, Himalayas and Eastern Mediterranean regions Revisiting fundamental concepts of continental extension KUSZNIR , N. J. & KARNER , G. D. Continental lithospheric thinning and breakup in response to upwelling divergent mantle flow: application to the Woodlark, Newfoundland and Iberia margins
1
9 47 63
77
111 139 157
173 199
217 239
265 291
325
389
vi
CONTENTS
CHRISTIE -B LICK , N., ANDERS , M. H., WILLS , S., WALKER , C. D. & RENIK , B. Observations from the Basin and Range Province (western United States) pertinent to the interpretation of regional detachment faults DYKSTERHUIS , S., REY , P., MU¨ LLER , R. D. & MORESI , L. Effects of initial weakness on rift architecture MORESI , L., MU¨ HLHAUS , H.-B., LEMIALE , V. & MAY , D. Incompressible viscous formulations for deformation and yielding of the lithosphere Index
421
443 457 473
Imaging, mapping and modelling continental lithosphere extension and breakup: an introduction G. D. KARNER1, G. MANATSCHAL2 & L. M. PINHEIRO3 1
ExxonMobil Upstream Research Company, P.O. Box 2189, Houston, TX 77252-2189, USA 2
Ecole et Observatoire des Sciences de la Terre, Universite´ Louis Pasteur, 1 Rue Blessig, 67084 Strasbourg, France
3
Departamento Geocieˆncias and CESAM, Universidade de Aveiro, Campus de Santiago, 3810-193 Aveiro, Portugal Abstact: This Special Publication is a direct outcome of a small but dedicated group of researchers who met in Pontresina, Switzerland, to review and define the fundamental observations characterizing extensional systems and their application in guiding and constraining modelling efforts and results. The various summaries of the keynote addresses give an objective overview of the state of the art in modelling lithospheric extensional systems, both from the regional scale using dynamic models to individual basins using kinematic models with an emphasis on capturing the extensional history of the Iberia and Newfoundland margins. At the heart of all of these efforts is a simple question: Exactly what mechanisms allow the continental lithosphere to be thinned to the point of rupture? Related questions are: (1) Do crustal and mantle faults play a major role in this thinning process? If so, what is their geometry and does their importance and geometry change with time? (2) Are there other mechanisms of lithospheric and crustal thinning that cannot be imaged on seismic sections? (3) How is deformation accommodated in space and time? (4) What role do inherited mechanical, thermal and/or chemical heterogeneities play in controlling strain distribution and localization? (5) When, how and to what degree does magma production affect the distribution and localization of extension? And (6) what is the stratigraphic record of continental extension and how does it document the extension of the crust and thinning of the lithospheric mantle? The aim of this Special Publication is to address many of these fundamental questions concerning the extreme extension and thinning of continental lithosphere.
The Pontresina workshop – the source of the book During the week of 11–16 July 2004, in the Swiss Alps, 46 researchers working on various problems related to lithospheric extensional systems met in Pontresina, Switzerland. The aim of the workshop was to bring together observationalists and numerical modellers working on rifted continental margins. Its main objectives were to: (1) summarize the cutting-edge research related to extensional basin and lithospheric deformation studies (both from field observations and modelling), and recent advances and applications in basin modelling codes; (2) benchmark lithospheric deformation codes in terms of temperature history, brittle and ductile strain history, strain rates, and subsidence/uplift patterns in space and time; and (3) identify a set of field constraints (both onshore and offshore) of extensional systems (geological, geophysical, petrophysical and petrological data) that fundamentally constrain theoretical and conceptual models for the extensional deformation of the lithosphere. The workshop attracted
participants from Australia, Brazil, Canada, Egypt, France, Germany, India, Ireland, Norway, Portugal, Switzerland, the UK and the USA. The workshop was made possible with funding from the European Science Foundation, the UK National Environment Research Council, InterMARGINS and the US National Science Foundation. Key questions that need to be answered by academia relate to defining and testing the mechanisms responsible for thinning continental lithosphere. However, the key questions that need to be answered by the hydrocarbon industry for the exploration of deep-water rifted margins are: (1) Where is the location of the ocean –continent transition? (2) What are the structures, compositions and modes of deformation within the ocean–continent transition? And (3) what are the subsidence and heat-flow history of the ocean –continent transition and the zone of lithospheric thinning? There is a complete parallelism in the themes being asked and investigated by both groups. Consequently, a new period of industry–academic collaboration has begun, centred around these themes, and
From: KARNER , G. D., MANATSCHAL , G. & PINHEIRO , L. M. (eds) Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, London, Special Publications, 282, 1– 8. DOI: 10.1144/SP282.1 0305-8719/07/$15.00 # The Geological Society of London 2007.
2
G. D. KARNER ET AL.
in the spirit of this collaboration BP, Conoco – Phillips, ExxonMobil and Statoil have generously supported the colour publication of this book. We are very grateful to their managements and also for publication support from InterMARGINS.
The Pontresina experience – the purpose of the book With the advent of sophisticated finite-element schemes for modelling the brittle and plastic extensional deformation of the lithosphere, fully integrated thermal and mechanical geodynamic models simulating the large-scale lithospheric systems have been possible. To perform such modelling, a myriad of geological assumptions need to be made concerning the rheological zonation, composition, strain rate and strain-rate dependency, initial temperature structure, radiogenic element distribution and crustal thickness of the lithosphere. Likewise, kinematic models of extensional basin formation, while computationally simpler than dynamic models, are evolving to include the wide range of geological processes responsible for basin development. Comparing the results of kinematic and dynamic models is difficult, primarily because the model philosophies address very different spatial scales of the real-world problem. For the present modelling schemes, the prediction of syn- and post-rift subsidence and crustal structure allow a comparison to be made between predicted dynamic and kinematic heat-flow histories. New field observations and syntheses concerning the detailed development of the ultra-deep regions of passive margins and the ocean–continent transition (OCT) have produced new and exciting constraints for structural reconstructions of latestage extension, implications for magma production and intrusion, infiltration of synextensional magmas into the mantle, thermal constraints for the proximal and distal margin using radiometric techniques on exposed mantle rocks, and relationships between sediments and magmatic rocks within various extensional systems. This book focuses primarily on mapping and modelling extensional processes in deep magma-poor extensional systems using the Iberia –Newfoundland and other comparable examples for benchmarking of modelling results and modelling codes. Each chapter provides a reflective overview of a specific modelling technique or study area, highlighting the present state of knowledge and the first-order problems and challenges facing tomorrow’s researchers.
Chapter arrangement The book has been organized into six main sections and 17 Chapters. The first section – ‘The
Iberia –Newfoundland continental extensional system (geological and geophysical constraints)’ – presents new and exciting geological and geophysical data and observational constraints that are necessary to model the extensional history of the Iberian and Grand Banks/Newfoundland margins. This is followed immediately by the simulation or benchmarking experiments in the sections. ‘The Iberia– Newfoundland continental extensional system (dynamic modelling)’ and ‘The Iberia–Newfoundland continental extensional system (kinematic modelling)’. However, cutting-edge research on mapping and modelling extensional systems is happening across a wide front, not only in Iberia and Newfoundland. Many other passive continental margins are also being extensively surveyed and intensively investigated, such as the western and southern Australian margins, the Antarctic margin, the MARGINS Gulf of California focus site, the margins of the South Atlantic and back-arc basins such as the Woodlark Basin. Attempts to re-energize studies in interesting scientific (but logistically and politically difficult) regions, such as the northern Red Sea, have also been initiated. These research themes are highlighted in the next section – ‘Observational characteristics of non-Atlantic extensional systems (offshore)’. Perhaps one of the most dramatic realizations over the last decade has been the applicability of mapped Alpine extensional systems in supplying a fundamental insight into the boundary and initial conditions prior to Iberian margin extension and the implications for the changing role and range of geological processes in space and time responsible for the regional and local extensional deformation of the European and Adriatic continental lithospheres. Field mapping of exposed extensional systems, such as the collisional systems of the Mediterranean, offer equally important insights into the role, complexity and range of processes responsible for the thinning of continental lithosphere. These issues are the subject of the section – ‘Observational characteristics of non-Atlantic extensional systems (onshore)’. The final section – ‘Revisiting fundamental concepts of continental extension’ – contains a number of eclectic chapters that herald future directions and/ or a need to reflect on the current paradigm related to extensional systems. For example, do active lowangle normal faults really exist? Arguably an old question and, to some, a question that has already been answered by field observations and relationships, predicting theoretically low-angle faulting nevertheless remains elusive. Further, it is now clear that lithospheric extension and thinning cannot be uniform with depth given the exhumation of continental lithospheric mantle or the paucity of brittle deformation observed at many rifted continental margins. But how does this occur? These fundamental questions are addressed in this section.
CONTINENTAL LITHOSPHERE EXTENSION AND BREAKUP
Research essentials Despite numerous seismic reflection and refraction and ODP drilling cruises of the Iberian, Galician and Newfoundland margins (Leg 47B, Leg 103, Leg 149, Leg 173 and Leg 210), a physical understanding of the extension processes responsible for margin development remains rather incomplete. For example, we do not yet understand the driving forces of rift initiation, segmentation and continuation, nor what controls the switch from rifting to drifting. We do not know the controls on the loci of rifting or how extension partitions itself laterally or vertically within the lithosphere. We do not know the scale and role of low-angle normal faults and whether these faults are early or late extensional features. We do not know how synextensional sagging, i.e. subsidence in the absence of controlling normal faults, can occur early in the extensional system, or the feedback between sedimentation and the style of deformation. We do not understand how the crust can thin dramatically from 30–40 km onshore to about 10 km within 100 km of the coast. We do not understand the role of fluids, volatiles, and magma in facilitating rifting, or the mechanisms by which huge volumes of magma are generated very quickly over wide areas at many rifted margins. We do not know the heat budget associated with rifting or how it changes as a function of space and time. In short, we still do not know a lot about the fundamentals of extensional systems, in general, and the conjugate Iberian, Galician and Newfoundland margins, in particular. The conjugate aspect of continental margin studies is crucial as it represents the entire extensional system. Defining the first-order observations that necessarily constrain any modelling effort of extensional systems, dynamic or kinematic, are sometimes straightforward, such as: – What is the pre-extension crustal and lithosphere thicknesses? – What is the pre-event relief and environments of deposition? – What are the spatial and temporal scales of deformation? – What are the strain rates, regions of strain localization and modes of deformation? – What is the age of the onset of rifting and how can it be recognized? – What is the composition, facies and palaeobathymetry of synrift/post-rift sediments, subsidence history and the nature and age of unconformities/discontinuities? Defining other equally important constraints, however, is not so easy and their definition is either controversial or even contentious, as for example: – What is the distribution of mechanical, thermal and chemical heterogeneities?
3
– What is the age of breakup (i.e. when is normal ocean crust generated in a well-defined oceanic spreading centre)? – Then again, what exactly is breakup? – How many rift events are there and what is their timing? – Did these rift events pulse or was extension continuous? – What is the age, composition and volume of synextensional magmas? – How has magmatic underplating modified the crustal thickness of the region and when did it occur? Does it affect the rifting process? – Do volcanic (i.e. totally non-magmatic) margins really exist or is magma simply emplaced within the lower crust or lithospheric mantle? – What is the compositional variation of the lithospheric mantle, and how does it control the rheology, magma production history and flexural strength of the extending lithosphere? – What is the distribution of crustal radioactivity?
The Iberia – Newfoundland extension system: research controversies, contentions and unanticipated results Tucholke et al., Hopper et al., Sibuet et al. and Reston address many of these objectives while summarizing the geological history of the Iberian –Newfoundland conjugate margin. The key data sets are multichannel reflection and crustal refraction seismic data and drilling results, principally from the Ocean Drilling Program (ODP) and industry wells within the Jeanne d’Arc and Orphan Basins. As summarized by Reston, the Iberian –Newfoundland conjugate margins are characterized as non-volcanic margins exhibiting a deficit of synrift magmas, a zone of exhumed subcontinental mantle in the continent–ocean transition zone and an apparent extension discrepancy. The existence of basement ridges of serpentinized peridotite of continental affinity in the transition zones is extremely important. First, mantle exhumation is a direct indicator of depth-dependent extension. Second, these continental peridotites are both faulted and associated with magnetic anomalies presumably of M5 –M3 age. Apparently, the serpentinization process can create sea-floor spreading magnetic anomalies during continental mantle exhumation. This observation challenges the long-held view that the existence of correlatable magnetic anomalies unambiguously defines oceanic crust. The lack of synrift magmas is somewhat of a problem given the extreme extensional deformation that characterizes the Iberian and Newfoundland continental lithospheres. Reston suggests that the lack of melt may be explained by a combination
4
G. D. KARNER ET AL.
of heterogeneous extension of the lower lithosphere (i.e. depth-dependent extension) and a cool subcontinental geotherm. An infertile continental mantle due to earlier magma extraction events (e.g. during Hercynian orogenesis) may be another factor. So what now is the real meaning of breakup or even the ‘breakup unconformity’? As a part of the depth-dependent extension process, there are various geological forms of ‘breakup’: (1) the cessation of brittle deformation within the upper crust (although lower crustal and continental mantle thinning may continue to occur); (2) breaching of the lower crust, which is concomitant with mantle exhumation; and (3) the breaching of the continental mantle and thus the inception of oceanic crust generation (or the exposure of oceanic mantle). For the Jeanne d’Arc Basin, the crustal blocks last moved in the late Barremian–early Aptian. But, is this the time of oceanic crust generation? Tucholke et al. argue that the final breakup of the Newfoundland– Iberia rift and the beginning of ‘true’ sea-floor spreading occurs at the Aptian–Albian boundary (c. 112 Ma), which is some 15 Ma younger than the oldest proposed oceanic crust (anomaly M3). The Aptian–Albian boundary is represented by a strong, rift-wide reflection that is onlapped by post-rift sediments deposited as a stable sea-floor-spreading regime was established. Tucholke et al. have termed this unconformity the ‘Aptian event’. Tucholke et al. suggest that this unconformity was produced by a change of in-plane stress induced by the final breaching of the subcontinental mantle lithosphere. Although depth-dependent extension has been surmised by a number of authors based on the paucity of synrift brittle deformation on some margins, Reston questions the significance of this observation. In particular, Reston suggests that ‘extension discrepancies’, as he terms them, may simply be the result of incomplete recognition of the entire polyphase faulting history. In the absence of mappable preand synrift stratigraphic packages, it just is not possible to define the extensional history of the Iberian margin, polyphase or otherwise (because the stratigraphic ‘tape recorder’ was not particularly effective). For margins exemplified by well developed pre-, syn- and post-rift stratigraphies, the polyphase faulting history is known and is minor (e.g. Karner & Driscoll 1999). Hopper et al. concentrated on interpretations of seismic reflection and refraction data from the Flemish Cap margin off Newfoundland. The seismic data highlighted important asymmetries at a variety of scales that developed during the final stages of continental breakup and the onset of oceanic sea-floor spreading. In strong contrast to the conjugate Galicia Bank margin, Flemish Cap shows an abrupt necking of continental crust,
thinning from 30 km to 3 km thick over a distance of 80 km with a distinct absence of controlling normal faults. At the ocean–continent transition, oceanic crust averages only 3 km to 4 km thick and in places is less than 1.3 km thick, suggesting that there were large spatial and temporal variations in the available melt supply following continental breakup as ocean crust was being generated. There is little evidence for significant extensional deformation of the Flemish Cap, consistent with the hypothesis that it behaved as a microplate throughout the Mesozoic. This fact was used by Sibuet et al., who recognized that rifting between Iberia and North America involved two larger plates (Iberia and North America) and two smaller microplates (Flemish Cap and Galicia Bank), to infer that Flemish Cap rotated approximately 438 relative to Galicia Bank and Iberia and moved 200–300 km SE with respect to North America. The seismic data for the Newfoundland Basin highlight important asymmetries at a variety of scales that developed during the final stages of continental breakup and the onset of oceanic sea-floor spreading. In strong contrast to the conjugate Galicia Bank margin, Flemish Cap shows abrupt necking, thinning the continental crust from 30 to 3–5 km over a distance of 80 km with no clear evidence for horizontal detachments or normal faults within the continental crust. This remains a fundamental question– exactly by what process does the crust undergo dramatic thinning without obvious brittle deformation?
The modelling effort: dynamic models The Pontresina benchmarking exercise consisted of modelling a combined Iberian –Newfoundland transect, using ISE-1 and SCREECH-1 multichannel seismic reflection data, refraction inversions and tomography for crustal velocity structure and Moho topography, and ODP drilling constraints. In terms of results, some models dealt with the details but offered no insight into the larger picture, while other modelling efforts provided no details but gave important insights to understanding general crustal architecture. The chapters by Huismans & Beaumont, Burov and Harry & Grandell represent the dynamic modelling of the Iberia –Newfoundland transect, while Egan & Meredith and Healy & Kusznir are the custodians of the kinematic approach to basin modelling. A dynamic model is designed to model the extension of the continental lithosphere by allowing for the spontaneous location of the deformation in shear zones approximating faults as well as the accommodation of the deformation by ductile behaviour. This is in contrast to kinematic models where the structural deformation of the crust and mantle lithosphere deformation are prescribed, and the effects
CONTINENTAL LITHOSPHERE EXTENSION AND BREAKUP
of basement rebound, sediment loading, erosion, compaction, eustasy, and lateral and vertical heat flow are easily included in the basin models both during, and after, extension. In general, it is very difficult to compare kinematic and dynamic models because these models address different scales of the extensional system. Huismans & Beaumont used plane-strain thermo-mechanical finite-element modelling experiments to investigate the effects of frictional-plastic strain softening and inherited weakness on the style of lithospheric extension, comparing the model results to the Newfoundland –Iberia conjugate margin. The fundamental result is that coupling between the plastic –viscous layering, acting together with frictional-plastic strain softening localized on inherited weak heterogeneities, can explain the initially wide rifting and distributed rift basins that are later abandoned in favour of a narrow rift in which mantle lithosphere is exhumed to the surface. Lithosphere-scale models with a single weak seed exhibit a range of asymmetric and symmetric rifting modes. That is, pure shear-type extension can be transformed into simple shear-type extension as functions of rheological layering and/ or rifting velocity, thus disqualifying the concept of simple pure v. simple shear end-member models. Burov takes a very different approach by investigating Rayleigh –Taylor instabilities in the lithosphere triggered by abrupt density and viscosity contrasts and steep boundaries between cold deformed lithosphere and hot upwelling asthenosphere induced by lithospheric extension and thinning. Burov suggests that continental lithosphere necessarily must extend at relatively high strain rates compared with the average extension rates usually assumed. A problem? Perhaps, but then again, extensional strain rates of almost all rift systems are poorly known. Because the influence of Rayleigh – Taylor instabilities is strongly controlled by extension rate, density, rheology and thermal structure of the lithosphere, there is a clear need for an increased attention to this problem to better constrain strain rates. In a more focused modelling effort, Harry & Grandell used a finite-element model to simulate the Late Jurassic –Early Cretaceous rifting between the Flemish Cap and Galicia Bank continental margins. In particular, the style and timing of rifting can be explained as a consequence of the interaction of two pre-existing lithospheric weaknesses: (1) a crustal weakness attributed to structural fabrics in the Variscan front; and (2) a deep-seated weakness attributed to the thick crust beneath the central Variscan Orogen, which controls the location of mantle necking. Extension is initially diffuse, but accelerates and becomes more focused with time. Most interestingly, some
5
time after rifting begins, the locus of crustal extension shifts from the region of preweakened crust into the region of preweakened mantle, marking the end of subsidence in the Galicia Interior Basin and the onset of subsidence in the Flemish Cap and Galicia Bank marginal basins. The modelling results are interesting but the serendipitous assigning of the all-important crustal and mantle weaknesses preordains the results – more regional geological data are required to constrain and/or validate these assumptions.
The modelling effort: kinematic models While basin kinematic models may not have the pizzazz of dynamic models, they nevertheless are significantly more flexible when applied to real basins. Egan & Meredith, after providing the fundamental rationale for kinematic modelling of extension, apply such a model to the Black Sea and the Iberia –Newfoundland margin. Healy & Kusznir, working with a conceptually new kinematic approach to lithospheric thinning (explained fully in Kusznir & Karner later in this Special Publication), model the development of the Goban Spur margin in the eastern North Atlantic. While Healy & Kusznir model satisfactorily the observed bathymetry and the free air gravity anomaly, they do not predict the exhumation of continental lithospheric mantle, which has only been inferred at Goban Spur from seismic refraction and reflection studies. If serpentinized mantle does exist at this margin, then the possibility exists that the mantle exhumation at non-volcanic margins is melt suppression during sea-floor spreading initiation (rather than continental lithospheric mantle exhumation) analogous with mantle exhumation during ultra-slow sea-floor spreading at the Gakkel and SW Indian Ridges.
The knowledge base of other onshore and offshore extensional systems Five papers are dedicated to the study of other, nonAtlantic extensional systems. Perhaps one of the most important rift basins studied over the last decade is the Woodlark Basin of Papua New Guinea, which included ODP drilling, field mapping, seismic tomography and underway geophysical studies. Goodliffe & Taylor summarize the fundamental characteristics of this extensional system: west of the spreading tip, the southern margin is characterized by large fault blocks and a northern margin devoid of large offset normal faults but extreme subsidence, indicative of preferential lower crustal thinning. The margin asymmetry primarily reflects across-strike inherited differences in the prerift geology, morphology and rheology. ODP drilling
6
G. D. KARNER ET AL.
results identify an initial phase of synrift sag subsidence interrupted by periods of relatively minor normal faulting. The brittle deformation (low-angle fault) that characterizes the Moresby seamount is very young, less than 0.5 Ma. The 258–308 fault dip is explainable in terms of a low coefficient of friction, being formed or reactivated in serpentinized ophilitic basement. The progression from extension to spreading is characterized by a decrease in sedimentation as the conjugate margins are progressively thinned, subside below sea level, trap sediments in proximal basins and develop significant palaeo-water depths. There is no evidence for the outcrop of lower crust or continental mantle as the lateral balance to depth-dependent extension, which is a problem. While Goodliffe & Taylor’s preferred mode of crustal thinning is via lower crustal ductile flow, crustal flow nevertheless tends to minimize Moho topography, which is not the conclusion of the seismic tomography studies of Abers et al. (2002) and Ferris et al. (2006). A problem still exists as to the mechanisms that produce synrift sag subsidence. Direen et al. document the geological development of the ocean– continent transition of the Naturaliste Plateau and the Great Australian Bight of southern Australia using multichannel seismic reflection, gravity and magnetic data. The ocean – continent transition zone of these margins is geologically and geophysically complex, but interpretation of all the geophysical data and dredge hauls (principally in the Diamantina Zone) is consistent with the presence of a mixture of modified continental lower crust, volcanics related to breakup and exhumed continental mantle. The decision by the US National Science Foundation to ‘postpone indefinitely’ MARGINS research in the Red Sea region is a major setback in gaining access to sites crucial in mapping, modelling and understanding extensional processes. The Red Sea remains a fundamentally important region in studying cratonic rifting. As stated in the MARGINS Science Plans for Rupturing Continental Lithosphere, ‘The Red Sea and Gulf of Aden are the closest modern analogs to the rifting and rupturing of continental lithosphere that formed the vast majority of “Atlantictype” continental margins, and are the location where the processes that shaped the early development of rifted continental margins can be studied with the fewest tectonic complications. Nearly all of the passive continental margins of the Atlantic, Indian, Southern and Arctic oceans were formed by the nucleation of an oceanic spreading centre within a continental rift following an extended period of cratonic lithosphere extension’. Cochran & Karner revisit the geological setting of the northern Red Sea and underscore its absolute importance in understanding rift systems and the rift–drift transition.
As summarized by Cochran & Karner, the Red Sea basins formed by rifting of Precambrian continental lithosphere in the late Oligocene, leading to breakup and sea-floor spreading by approximately 5 Ma in the southern Red Sea and significant continental block rotation and the instalment of volcanic centres in the northern Red Sea. Comparing and contrasting the extension of the southern and northern Red Sea remains the definitive experiment to understand the development of Atlantic-type passive continental margins. For decades it has been simply assumed that the northern Red Sea had experienced insufficient extension to breach the continental lithosphere, but, in time, would develop into a spreading centre. However, as shown by Kusznir & Karner later in this Special Publication, rifts characterized by large offset fault systems, i.e. faults that generate synrift accommodation, tend not to proceed to breakup. Thus, Cochran & Karner explore the tantalizing possibility that the northern Red Sea region, with its interpreted rotated fault blocks, is rheologically stronger than the lithosphere of the southern Red Sea and will never evolve to sea-floor spreading. In marked contrast to the Woodlark Basin, Eastern Tethyan margins and even the Gulf of California margins (e.g. Persaud et al. 2003), the Alpine Tethys margin is characterized by a continent–ocean transition zone and exposed continental mantle in a style very similar to that of the Iberian margin (Manatschal et al.). The Tethys margin, now exposed in various nappes in the Alps, shows a change from initially regionally distributed and decoupled extension to later localized, coupled and asymmetric extension that results in thinning of the crust and exhumation of subcontinental mantle. The change in the mode of extension, together with the localization of deformation, reflects an evolution of the bulk rheology of the extending lithosphere. This is therefore a fundamental natural laboratory to study how the crust thins during extension. As underscored by the dynamic rifting model results, initial rifting of the Tethyan margin appears to be controlled by inherited heterogeneities (cf. chapters by Dysterhuis et al. and Harry & Grandell in this Special Publication) and mechanical localization processes, whereas the site of final extension and lithospheric rupture is controlled by serpentinization, magmatic and thermal weakening of the continental mantle. In the final paper of this section, Robertson compares and contrasts the geology and tectonics of the rifted margins of the Eastern Tethys with the Iberia –Newfoundland margin and the East Greenland margin. None of the Eastern Tethyan rifted margins show evidence of features characteristic of the non-volcanic rifted margins of the Atlantic (e.g. exhumation of continental mantle).
CONTINENTAL LITHOSPHERE EXTENSION AND BREAKUP
Combinations of geophysical studies and deep-sea drilling have in the past suggested that rifted margins fall into two end-members: volcanic-rifted margins (e.g. East Greenland) and non-volcanic rifted margins (e.g. Iberia–Newfoundland conjugate). Most of the Eastern Tethyan rifted margins are neither, and appear to be ‘intermediate’ between these end-member margin types, being characterized by pulsed rifting (extending over more than 50 Ma), limited rift volcanism and a narrow but well-developed continent– ocean transition zone. In terms of tectonic setting, there is little evidence to support previous suggestions that the Eastern Tethyan rifts are back-arc basins above either northward-dipping or southwarddipping subduction zones. All of the rifted margins of the Eastern Tethys are associated with rift-related volcanic rocks. The dominant controls of rifting are seen as the traction of rising asthenosphere on the base of the lithosphere, related deviatoric tensional stresses, inherited and thermally induced weaknesses in the crust, and slab pull.
Revisiting fundamental concepts of continental extension As stated earlier, this final section contains a number of eclectic chapters that may herald future directions and/or a need to revisit or reflect on supposedly well-established paradigms. Four papers comprise the last section of the book, Kusznir & Karner, Christie-Blick et al., Dyksterhuis et al. and Moresi et al. Although depth-dependent lithosphere thinning has been implicitly discussed in earlier chapters, Kusznir & Karner review the prime evidence for this process. Kusznir & Karner demonstrate that depth-dependent thinning characterizes the margins of the North and South Atlantic, Norway, South China Sea, Papua New Guinea, NW Australia and Vietnam. In short, depth-dependent lithosphere thinning appears to be the norm, rather than the exception, for both non-volcanic and volcanic extensional systems. However, depth-dependent thinning, if it is an extension process, may create a ‘space problem’ depending on how the extension is distributed with depth, i.e. the extension needs to be laterally balanced. For regions characterized by regionally significant synrift sags, such as for the Congo–Angola and NW Australian margins, an adjacent region of exposed lower continental crust and continental mantle needs to exist – clear for the Iberian margin but a problem, for example, for the Woodlark Basin. Recognizing or reconciling this extensional balance, in general, has been problematic. Consequently, Kusznir & Karner describe a fundamentally different form of kinematic extension model in which deformation and thinning of
7
continental lithosphere occurs in response to an upwelling divergent flow field. Such a model successfully predicts depth-dependent stretching of continental margin lithosphere and mantle exhumation for both non-volcanic and volcanic margins while maintaining a balance of extensional strain. This strain balance is achieved by redistributing lower crust and continental mantle away from the zone of active extension into the hinterland region, implying a broad region of crustal and continental thickening and broad hinterland uplift during and for some time following extension. Christie-Blick et al. report recent fieldwork that casts doubt on the interpretation of some well-studied regional detachment faults in the US Basin and Range Province. What has this to do with the Iberian – Newfoundland margin? Nothing, but it directly addresses the question of how crust thins during extension. Given that the Basin and Range examples were influential in the development of the active low-angle normal fault concept, it is somewhat disturbing to learn that the Mormon Peak detachment, for example, a structure that was one of the most influential in arguing for the importance of lowangle normal faults and for high extensional strain in regions not associated with metamorphic core complexes, can be reinterpreted as a series of rootless slide blocks on the basis of detachment characteristics and spatially variable kinematic indicators. Among issues addressed by Christie-Blick et al. are the manner in which brittle deformation relates to mylonitization, the degree to which faults are folded or tilted to lower dip and the role of unusual materials such as talc in reducing frictional coefficients. Christie-Blick et al. explicitly state that their work does not falsify the active low-angle normal fault paradigm nor do they deny the reality of extreme crustal extension – they ask for a critical re-examination of interpreted detachments elsewhere in the Basin and Range Province and in other extensional and passive margin settings. Dyksterhuis et al. investigate the primary factors controlling rift architecture. Key results are that assumed weaknesses fundamentally control rift mode rather than bulk rheology, initial lithospheric temperature structure or strain softening. This result is in complete agreement with the modelling results of Harry & Grandell. A secondary conclusion of Dyksterhuis et al. is that low-angle faults can indeed form as the result of rotation of initially highangle faults, i.e. the rolling hinge model of low-angle normal fault development. It is gratifying to know that the principles of isostasy in extensional systems remain very much intact. The final chapter by Moresi et al. presents an extremely important theoretical development that allows viscous formulations to include 2D and 3D approximations to elastic effects and plastic
8
G. D. KARNER ET AL.
failure – that is, a way to represent Mohr –Coulomb failure within the mathematical framework used by mantle-convection fluid dynamics codes. In the words of the authors ‘The algorithms described in this paper address a prominent problem in geodynamics: how to model the brittle deformation of the uppermost lithosphere that occurs when mantle convection deforms the continents at the same time as modelling the underlying fluid convection which drives the surface deformation’. In many ways, formulations of this type represent the future of complex, large strain approximations of lithospheric extensional deformation. As with any professional publication, the quality of the final product is a strong function of the dedication of reviewers within the earth science community. The volume editors would like to acknowledge the following people for their unselfish donation of time to critically review the various chapters of this book: G. Bertotti, R. Buck, E. Burov, N. Christie-Blick, J. Cochran, P. Cowie, N. Dawers, S. Deemer, N. Direen, S. Egan, M. Enachescu, T. Funck, A. Goodliffe, R. Govers, M. Pe´rez-Gussinye´, D. Harry, J. Hopper, G. Houseman, R. Huismans, N. Kusznir, L. Lavier, G. Lister, K. Louden, F. Martinez, T. Minshull, L. Montesi, L. Moresi, D. Mu¨ller, J. Nunn, W. Powell, D. Roberts, A. Robertson, D. Sawyer, D. Shillington, T. Reston,
M. Tischer, F. Tsikalas, B. Tucholke, H. Whitechurch, R. Whitmarsh, C. Wilson, P. Ziegler and last but certainly not least, our production editor, S. Oberst.
References A BERS , G. A., F ERRIS , A., C RAIG , M. ET AL . 2002. Mantle compensation of active metamorphic core complexes at Woodlark Rift in Papua-New Guinea. Nature, 418, 862– 865. F ERRIS , A., A BERS , G. A., Z ELT , B., T AYLOR , B. & R OECKER , S. 2006. Crustal structure across the transition from rifting to spreading: the Woodlark rift system of Papua New Guinea. Geophysical Journal International, 166, 622– 634. K ARNER , G. D. & D RISCOLL , N. W. 1999. Style, timing and distribution of tectonic deformation across the Exmouth Plateau, northwest Australia, determined from stratal architecture and quantitative basin modelling. In: MAC NIOCAILL , C. & RYAN , P. D. (eds) Continental Tectonics. Geological Society, London, Special Publications, 164, 271–311. P ERSAUD , P., S TOCK , J. M., S TECKLER , M. S. ET AL . 2003. Active deformation and shallow structure of the Wagner, Consag, and Delfin Basins, northern Gulf of California, Mexico. Journal of Geophysical Research, 108(B7), 2355, doi:10.1029/2002 JB001937.
Breakup of the Newfoundland– Iberia rift B. E. TUCHOLKE1, D. S. SAWYER2 & J.-C. SIBUET3 1
Department of Geology and Geophysics, Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA (e-mail:
[email protected]) 2
Department of Earth Science, Rice University, Houston, TX 77005, USA 3
Ifremer Centre de Brest, B.P. 70, 29280 Plouzane´, France
Abstract: The Newfoundland–Iberia rift is considered to be a type example of a non-volcanic rift. Key features of the conjugate margins are transition zones (TZs) that lie between clearly continental crust and presumed normal (Penrose-type) oceanic crust that appears up to 150– 180 km farther seaward. Basement ridges drilled in the Iberia TZ consist of exhumed, serpentinized peridotite of continental affinity, consistent with seismic refraction studies. Although the boundaries between continental crust and the TZs can be defined with relative confidence, there are major questions about the position and nature of the change from rifting to normal sea-floor spreading at the seaward edges of the TZs. Notably, drilling of presumed oceanic crust in the young M-series anomalies (,M5) has recovered serpentinized peridotite, and this basement experienced major extension up to approximately 15 million years after it was emplaced. In addition, existing interpretations place the ‘breakup unconformity’ (normally associated with the separation of continental crust and simultaneous formation of oceanic crust) near the Aptian–Albian boundary, which is also some 15 million years younger than the oldest proposed oceanic crust (anomaly M5–M3) in the rift. To investigate and potentially resolve these conflicts, we analysed the tectonic history and deep (pre-Cenomanian) stratigraphy of the rift using seismic reflection profiles and drilling results. Rifting occurred in two main phases (Late Triassic–earliest Jurassic and Late Jurassic– Early Cretaceous). The first phase formed continental rift basins without significant thinning of continental crust. The second phase led to continental breakup, with extension concentrated in three episodes that culminated near the end of Berriasian, Hauterivian and Aptian time. The first two episodes appear to correlate with separation of continental crust in the southern and northern parts of the rift, respectively, suggesting that the rift opened from south to north in a two-step process. The third episode persisted through Barremian and Aptian time. We suggest that during this period there was continued exhumation of subcontinental mantle lithosphere at the plate boundary, and that elevated in-plane tensile stress throughout the rift caused intraplate extension, primarily within the exhumed mantle. This rifting may have been interrupted for a time during the Barremian when melt was introduced from the southern edge of the rift by plume magmatism that formed the Southeast Newfoundland Ridge and J Anomaly Ridge, and the conjugate Madeira– Tore Rise. We propose that the rising asthenosphere breached the subcontinental mantle lithosphere in latest Aptian– earliest Albian time, initiating sea-floor spreading. This resulted in relaxation of in-plane tensile stress (i.e. a pulse of relative compression) that caused internal plate deformation and enhanced mass wasting. This ‘Aptian event’ produced a strong, rift-wide reflection that is unconformably onlapped by post-rift sediments that were deposited as a stable sea-floor-spreading regime was established. Although previously considered to be a breakup unconformity associated with separation of continental crust, the event instead marks the final separation of the subcontinental mantle lithosphere. Our analysis indicates that interpretation of tectonic events in a non-volcanic rift must consider the rheology of the full thickness of the continental lithosphere, in addition to spatial and temporal changes in extension that may occur from segment to segment along the rift.
The Newfoundland–Iberia rift (Fig. 1) is considered to be a type example of non-volcanic rifted margins, and because of its accessibility it has been extensively studied by geophysical means and by drilling over the past two decades. Most of these investigations have concentrated on the Iberia margin, but recent geophysical surveys and deep-sea drilling off Newfoundland also provide critical information
that contributes to understanding the evolution of the rift. A key feature observed on each margin is a transition zone (TZ) that lies between clearly continental crust and assumed normal oceanic crust that appears up to 150–180 km farther seaward in the vicinity of magnetic anomaly M3 (Figs 2 & 3). (Normal ocean crust is here considered to be crust that approximates the Penrose ophiolite
From: KARNER , G. D., MANATSCHAL , G. & PINHEIRO , L. M. (eds) Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, London, Special Publications, 282, 9– 46. DOI: 10.1144/SP282.2 0305-8719/07/$15.00 # The Geological Society of London 2007.
10
B. E. TUCHOLKE ET AL.
Fig. 1. North Atlantic bathymetric map (Smith & Sandwell 1997) showing features of the conjugate margins of the Newfoundland–Iberia rift. Boxes locate Figures 2 & 3. White dashed lines show Palaeozoic sutures and terrane boundaries in continental crust (after Keen et al. 1990; Silva et al. 2000) and black lines delineate major rift basins. Yellow line in the Tagus Abyssal Plain locates refraction profile of Pinheiro et al. (1992) (Fig. 4g, section I1). AM, Armorican Massif; BB, Bay of Biscay; CA, Collector magnetic anomaly; CIZ, Central Iberian Zone; CZ, Cantabrian Zone; F, Flemish Pass Basin; FC, Flemish Cap; GB, Galicia Bank; GI, Galicia Interior Basin; GRB, Gorringe Bank; GZ, Galicia-Tra´s-os-Montes Zone; H, Horseshoe Basin; IAA, Ibero-Armorican Arc that presumably connected to the northern European sutures prior to opening of the Bay of Biscay; IAP, Iberia Abyssal Plain; J, Jeanne d’Arc Basin; JAR, J Anomaly Ridge; L, Lusitanian Basin; MAR, Mid-Atlantic Ridge; MEG, Meguma Terrane; MTR, Madeira– Tore Rise; NNB, northern Newfoundland Basin; NS, Newfoundland Seamounts; O, Orphan Basin; OMZ, Ossa Morena Zone; P, Porto Basin; SB, Salar– Bonnition Basin; SENR, Southeast Newfoundland Ridge; SNB, southern Newfoundland Basin; SPZ, South Portuguese Zone; TAP, Tagus Abyssal Plain; TS, Tore Seamounts; W, Whale Basin; WZ, Western Asturian–Leonese Zone.
model (Penrose Conference Participants 1972) and consists of gabbro, sheeted-dyke and volcanic layers that accreted at a magmatic spreading centre). Basement ridges drilled in the TZ on the Iberia margin consist of serpentinized peridotite of apparent continental affinity (e.g. Whitmarsh & Wallace 2001), and seismic refraction studies show velocity structure that is consistent with the presence of serpentinized mantle (e.g. Chian et al. 1999; Dean et al. 2000). Thus, the TZ been referred to as the Zone of Exhumed Continental Mantle (ZECM) (Whitmarsh et al. 2001). Although the boundaries between continental crust and the TZ can be defined with relative confidence on the basis of basement morphology, fault patterns and velocity structure (Figs 2 & 3), major questions remain about the nature and position of
the seaward limit of the TZ and the first occurrence of normal oceanic crust. Notably, in basement identified as oceanic crust from seismic refraction studies, drilling at Site 1070 on the Iberia margin and at Site 1277 on the Newfoundland margin has recovered serpentinized peridotite (Whitmarsh et al. 1998; Tucholke et al. 2004). This puts into question the validity of identifying oceanic crust only on the basis of velocity structure. In addition to this uncertainty, several existing interpretations place the final breakup of the Newfoundland– Iberia rift near the Aptian –Albian boundary, about 112 Ma. This event has been interpreted to be marked by a breakup unconformity (e.g. Mauffret & Montadert 1987; Boillot et al. 1989c), which is conventionally considered to mark the separation of continental crust and
BREAKUP OF THE NEWFOUNDLAND– IBERIA RIFT
11
Fig. 2. Bathymetry of the Newfoundland Basin (Smith & Sandwell 1997) with basin features and profile locations. Identified magnetic anomalies are shown as black solid lines, from Srivastava et al. (2000) and Shillington et al. (2004). The approximate distribution of the high-amplitude magnetic J anomaly, which contains anomaly M0, is indicated by a heavier line. The dashed black line at the boundary between Aptian and Albian crust is based on interpolation between anomalies M0 and 34. Black dots locate ODP drill sites. The approximate seaward limit of certain continental crust is the white short-dot line (CL), and the white wide-dot line (TCL) shows the seaward limit of highly thinned continental crust interpreted from refraction data (Lau et al. 2006b; Van Avendonk et al. 2006). Multichannel seismic reflection lines illustrated in other figures are solid yellow lines, dotted yellow lines show the extent of crustal-thickness profiles (,10 km thickness where over continental crust) derived from seismic refraction data and illustrated in Figure 4g, and black/yellow lines show locations of SCREECH data not illustrated in this paper. Refraction line R94 is from Reid (1994), SCREECH2 is from Van Avendonk et al. (2006) and SCREECH1 is from Funck et al. (2003).
simultaneous formation of the first oceanic crust (Falvey 1974). The age of this postulated breakup is some 15 million years younger than the oldest magnetic anomaly (M5) that has been proposed as possible oceanic crust off Iberia (Russell & Whitmarsh 2003), and it is 9 million years younger than anomaly M0 which most investigators would argue is certain oceanic crust. There are three
possible ways to explain this paradox: (1) the breakup unconformity has been misidentified; (2) the identification has no meaning because continental breakup does not generate a consistent marker in the stratigraphic record; or (3) the breakup unconformity can represent an event other than simultaneous separation of continental crust and generation of the first normal oceanic crust.
12
B. E. TUCHOLKE ET AL.
Fig. 3. Bathymetry of the Iberia margin (Sibuet et al. 2004a) with basin features and locations of illustrated profiles. Explanation as in Figure 2. Magnetic anomalies are from Miles et al. (1996) and Russell & Whitmarsh (2003). Peridotite ridges identified from various sources are indicated by white lines and ‘R’ labels. Seismic refraction profile W96 is from Whitmarsh et al. (1996) and CAM127-IAM9 is from Dean et al. (2000).
In this paper, we review the rift history of the Newfoundland–Iberia margins with particular attention to the deep (pre-Cenomanian) stratigraphy of the rift. We focus on the issue of Late Jurassic – Early Cretaceous rift history, and in particular on development of the previously proposed breakup unconformity and the time that the first normal oceanic crust formed. From this analysis, we propose that the third explanation above is likely to be true in the Newfoundland–Iberia rift. We suggest that the first normal oceanic crust was not formed until the Aptian –Albian boundary, approximately 112 Ma, coincident with the formation of the previously proposed breakup unconformity. Up to that time new sea floor was emplaced largely by exhumation of subcontinental mantle lithosphere,
and the entire rift was under elevated in-plane tensile stress. This caused local extension throughout the rift, even though the last principal episode of extension in continental crust probably ended in Hauterivian –Barremian time. ‘Breakup’ is represented by the final separation of subcontinental mantle lithosphere, rise of the asthenosphere to shallow levels and release of large volumes of melt, and commencement of normal sea-floor spreading. Relaxation of plate-wide tensile stress at this time led to a pulse of relative compression within the plates, resulting in internal plate deformation (e.g. topographic amplification), enhanced mass wasting of serpentinite ridges in the deep basins, and proximal-margin erosion and basin flooding by turbidites. We suggest that this
BREAKUP OF THE NEWFOUNDLAND– IBERIA RIFT
deformation, the accumulation of coarse debris in the deep basins and the subsequent change to sedimentation in a low-stress environment combined to create a unique stratigraphic marker in seismic reflection records. According to our interpretation, this marker is not a breakup unconformity in the classical sense (i.e. simultaneous separation of continental crust and commencement of sea-floor spreading) but is the structural and stratigraphic expression of the final separation of the full thickness of continental lithosphere. Thus, we will refer to this feature as the ‘Aptian event’.
Present structural framework Continental crust The Newfoundland margin is dominated by a broad, shallow-water (,100 m) platform that forms the Grand Banks (Fig. 1). This platform is formed by thick continental crust (27 –35 km: Reid 1988; Reid & Keen 1990) that is deeply incised by large, deep rift basins. The seaward edge of the platform is a seaward-dipping basement hinge zone that runs mostly beneath the continental slope from the southern transform margin of the Grand Banks northwards to Flemish Cap. Along the southern Grand Banks, rift basins seaward of the hinge zone (Salar –Bonnition basins) contain an apparently thick sequence of evaporites (Austin et al. 1989) and they are marked at their eastern edge by a landward-dipping hinge in basement (Keen & de Voogd 1988; Austin et al. 1989). Flemish Cap, at the NE end of the Grand Banks, has thick (c. 35 km) continental crust (Funck et al. 2003) and it is separated physiographically from the Grand Banks by Flemish Pass (1100 m water depth), which overlies a deep rift basin in continental crust. A prominent east-dipping hinge zone in basement on the seaward side of Flemish Cap, together with the landward-dipping hinge zone along the margin to the south, is generally considered to be the seaward edge of clearly continental crust (CL, Fig. 2); these interpretations are based on crustal thickness and velocity structure, basement morphology, and the occurrence of evaporites (where present). Highly thinned (2–6 km) continental crust has been interpreted to extend up to about 60 km farther seaward (TCL, Fig. 2) based on refraction data (Lau et al. 2006b; Van Avendonk et al. 2006). The Iberia margin contrasts with the Newfoundland margin in that it has a much narrower continental shelf and limited development of rift basins in thick continental crust (Fig. 1). The principal rift basin in the shallow Iberia margin is the Lusitanian Basin, which is roughly
13
co-linear with the Porto Basin beneath the continental shelf and slope and with the deep-water Galicia Interior Basin to the north. Thinned continental crust underlies both the Galicia Interior Basin and Galicia Bank to the west (Boillot et al. 1989b; Murillas et al. 1990). Continental crust under Galicia Bank thins westward to zero thickness on the landward side of peridotite basement ridges (R1 –R3, Fig. 3). Based on geophysical and drilling results, the seaward edge of clearly continental crust runs SE along the southern margin of Galicia Bank and then southward in a poorly defined position to the southern limit of the rift (CL, Fig. 3). The seaward limit of continental crust is significantly younger in the northern part of the rift between Flemish Cap and Galicia Bank than in the southern part (Figs 2, 3 & 4g). Some interpretations suggest that this reflects relatively uniform northward propagation of rifting and crustal separation (e.g. Whitmarsh & Miles 1995). However, it is equally possible that rifting was a stepwise process. Although data are limited for the southern part of the rift, extension there appears to have caused separation of continental crust during the latest Jurassic–earliest Cretaceous (Fig. 4g, sections N1 and I2). This extension probably also affected the northern rift (Boillot et al. 1989c; Murillas et al. 1990), but Flemish Cap and Galicia Bank did not separate until near the Hauterivian–Barremian boundary approximately 16–19 million years later (Fig. 4g, sections N2, N3 and I3). Thus, breakup of the rift may have occurred in two successive and relatively discrete episodes.
Transitional crust On both sides of the rift south of the Flemish Cap– Galicia Bank conjugates, there is a wide (c. 150–180 km) TZ where basement has unusual geophysical properties and a complex origin. This zone extends from the outer limits of crust of certain continental origin seawards to about anomaly M3. The most extensive studies of the TZ are on the Iberia side of the rift, largely in the region of the southern Iberia Abyssal Plain (Fig. 3). In that area, refraction and multichannel seismic reflection studies document thin ‘crust’ (2–4 km) defined by low seismic velocities of approximately 4–4.5 km s21 (line IAM9, section I2 in Fig. 4g) (Whitmarsh et al. 1990; Discovery 215 Working Group 1998; Dean et al. 2000). Seismic velocities, velocity gradients and a general absence of a Moho reflection have been interpreted as representing intense serpentinization of mantle peridotites at the top of the basement and significantly reduced serpentinization with depth (c. 7.2 km s21 increasing to c. 7.9 km s21) (Discovery 215 Working Group 1998; Chian et al. 1999). Along the IAM9 line the upper
Fig. 4. Summary of geological features of the Newfoundland–Iberia rift v. time from the latest Jurassic through Early Cretaceous. The data are compiled from numerous sources, with principal sources discussed in the text. The light green bar marks the position of the ‘Aptian event’ at the Aptian– Albian boundary. (a) Timescale from Channell et al. (1995) and Gradstein et al. (1995). (b) Principal seismic and lithological sequences on the conjugate margins. Grey bars indicate biostratigraphic age constraints on the Aptian event (¼U reflection and orange reflection) at DSDP/ODP drill sites. (c) Principal rift events: red, magmatic, with significant magmatism as heavier lines; blue, tectonic. (d) Interpretation of in-plane tensile stress in the rift system; interpreted high stress in the Jurassic followed by decreasing stress up to Aptian time is from Tankard & Welsink (1987). (e) Rift phases defined from studies of extensional basins in the shallow margins of the rift and in Galicia Interior Basin. Albian extension in the northern Jeanne d’Arc Basin (grey shade) reflects detachment of basin fill associated with early extension of the Labrador Sea to the north of the Grand Banks. (f) Major unconformities;
14 B. E. TUCHOLKE ET AL.
Fig. 4. (Continued) all except the Galicia ‘breakup unconformity’ are based on data from extensional basins in the shallow margins of the rift and in Galicia Interior Basin (i.e. well and onshore outcrop data, and related seismic interpretations). (g) Representative thicknesses of velocity layers in basement v. age, derived from refraction profiles in the southern, central and northern parts of the rift. Note that these sections depict only layer thickness and not topography. Three sections are arranged from south to north (left to right) for each of the Newfoundland (N) and Iberia (I) margins. Sections N3 and I3 are conjugates, I2 is offset about 25 km south of N2 at anomaly M0, and N1 and I1 are approximately conjugate. Locations of sections are shown in Figures 2 and 3 (I1 is located in Fig. 1). Pale blue and pale green colours indicate thickness of interpreted oceanic and transitional ‘crust’, respectively. These have compressional wave velocities in the range of 3.5–7.2 km s21 and commonly exhibit steep velocity gradients. The oceanic layer is anomalously thick at the young end of section I2 because the profile intersects a seamount (see Fig. 3). Broken patterns at the young ends of sections N3 and I3 indicate where velocity structure is not well constrained. Dark blue is interpreted continental crust; continental crust thicknesses are shown only up to 10 km, although all sections except I1 extend onto thicker crust. Red indicates lithosphere with velocities from 7.2 up to 8.0 km s21 that typically has been interpreted as serpentinized mantle because velocities are considered to be too high to represent magmatic underplating; the velocity model in N2 did not detect 8 km s21 velocities where shown by white dotted lines, so in those locations the red area represents minimum thickness of the serpentinized layer. The transitional ‘crust’ above the red zones can be serpentinite and has specifically been interpreted as such on sections N2 and I2, and at the peridotite ridge on I3. Black dots show peridotite ridges sampled by drilling at ODP Sites 1277 (N2) and 637 (I3); open circles locate peridotite ridges interpreted from refraction data and also drilled at ODP Sites 1070, 897 and 899 (seaward to landward) approximately 50 km to the north. Basement ages are pinned to magnetic anomaly M0 (or in the case of I2, to both M0 and M5; in the case of N2, to both M0 and M3). Away from these anomalies the ages are calculated assuming that basement was emplaced at a rate of 9 km per million years (the average rate seaward of M0 to anomaly 34); ages landward of the seaward limits of continental crust have no meaning in this context. Data sources: N1, Reid (1994); N2, Van Avendonk et al. (2006); N3, Funck et al. (2003); I1, Pinheiro et al. (1992); I2, Dean et al. (2000); I3, Whitmarsh et al. (1996).
BREAKUP OF THE NEWFOUNDLAND– IBERIA RIFT 15
basement has a markedly non-reflective character that also has been interpreted as a layer of highly serpentinized mantle (Pickup et al. 1996). The Ocean Drilling Program (ODP) Leg 149–173 drilling transect, about 50 km to the north, drilled basement ridges that can be traced northward from the IAM9 line. Serpentinites were recovered from these ridges at Sites 897, 899 and 1070 (Figs 3 & 4g, open circles), consistent with seismic-velocity interpretations. The boundary between the serpentinized mantle and continental crust of Galicia Bank appears to run between Sites 899 and 1069, with some additional complexity just to the east at Sites 900, 1067 and 1068 where gabbros, amphibolites and serpentinites were recovered from basement. On the opposing margin the 565 km-long SCREECH3 (Studies of Continental Rifting and Extension off the Eastern Canadian sHelf Line 3) profile crosses the Grand Banks and reaches seaward to beyond anomaly M0 (Fig. 2). Based on combined seismic reflection and refraction data, Lau et al. (2006a, b) interpreted the inner part of the TZ there to be thinned continental crust (,3 km) over a thick (.5 km) layer of serpentinized mantle (7.6–7.9 km s21). Seaward to near anomaly M3, they proposed that an approximately 80 km wide zone with a thin (2 km) low-velocity basement and underlying 7.6–8.0 km s21 layer is probably exhumed and serpentinized mantle. The TZ in the southern part of the Newfoundland–Iberia rift has been much less studied. On the Iberia margin beneath the Tagus Abyssal Plain, Mauffret et al. (1989) inferred from multichannel seismic reflection and magneticanomaly studies that Tithonian–Hauterivian seafloor spreading formed the TZ, prior to a westward ridge jump at chron M10. However, a refraction profile shot over this area (Pinheiro et al. 1992) identified exceptionally thin ‘crust’ (2 km) that overlies mantle with low velocities of 7.6–7.9 km s21 (Fig. 4g, section I1), very much like the IAM9 basement farther north under the Iberia Abyssal Plain. Pinheiro et al. (1992) interpreted the Tagus basement to be formed by sea-floor spreading beginning at chron M11, but they noted that the low mantle velocities may indicate widespread serpentinization. In the conjugate Newfoundland TZ, Reid (1994) found mantle velocities beginning about 6 km below the basement surface just east of the Salar –Bonnition hinge. His results (Fig. 4g, section N1) show thin, low-velocity upper basement (4.5 km s21) and 7.2–7.7 km s21 velocities in the deeper basement, also similar to serpentinized mantle on the IAM9 profile. In the northern part of the rift, the TZ narrows abruptly between the Flemish Cap and Galicia Bank conjugates (Figs 2 & 3). The SCREECH2
16
B. E. TUCHOLKE ET AL.
reflection–refraction profile off Newfoundland is located where this narrowing occurs, conjugate to the ODP Leg 149 –173 drilling transect at the southern margin of Galicia Bank. Van Avendonk et al. (2006) interpreted approximately 60 km of the landward part of the TZ along the SCREECH2 profile as thinned, 2 –6 km thick continental crust (velocity 5.0–6.5 km s21); this overlies a 7.2– 8.0 km s21 layer that has laterally variable thickness (1–.4 km) and may be serpentinized mantle (Fig. 4g, section N2). An approximately 25 km wide zone at the outer edge of the TZ has basement velocities of 6.3–7.7 km s21 and was interpreted as exhumed, serpentinized mantle. Within the narrow TZ farther north, Todd & Reid (1989) observed a thin (2–3 km) 4.0–5.0 km s21 layer overlying probably serpentinized mantle (7.3 km s21) in one profile and overlying 8.0 km s21 mantle in a second profile. There is little, if any, TZ present still farther north at the position of the SCREECH1 and W96 conjugate profiles (Figs 2 & 3). Continental crust occupies the position where anomaly M3 would be expected on the Newfoundland side (Fig. 4g, section N3), but the anomaly could be present off Galicia Bank (Fig. 4g, section I3). In this part of the western Galicia margin, a layer of serpentinized peridotite underlies thin continental crust, a peridotite ridge, and a narrow zone of thin crust interpreted as oceanic (Whitmarsh et al. 1996). Two features are nearly universal among the seismic-refraction observations of the TZ. One is the presence of low-velocity (7.2–7.9 km s21), apparently serpentinized, mantle that is up to approximately 5 km thick. The second is the occurrence of a thin (c. 2–4 km) low-velocity layer that forms the overlying basement. Where thin continental crust is interpreted in the landward part of the TZ, it is likely that serpentinization of the underlying mantle occurred when the crust thinned sufficiently for fluids to penetrate there (Pe´rez-Gussinye´ & Reston 2001). The outer TZ appears to be exhumed mantle, based primarily on the combined drilling and seismic results around Galicia Bank (Fig. 4g, sections I2 and I3). The highly variable velocity structure of the thin upper ‘crustal’ layer in the outer TZ probably is caused by strong but heterogeneous serpentinization. The magnetic field over the TZ on both margins is characterized by low-amplitude magnetic anomalies. Srivastava et al. (2000) interpreted these as sea-floor-spreading anomalies and suggested that oceanic crust extends all the way from anomaly M3 to the edges of continental crust (to as old as anomaly M20, Tithonian, in the southern rift). However, considering the fact that the TZ appears to be serpentinized peridotite (with overlying thin continental crust in some parts of
the inner TZ) there may be two possible explanations for the source of the anomalies. One is that there are igneous intrusions in the peridotites and thinned continental crust. Detailed magnetic analyses on the Iberia margin (Whitmarsh & Miles 1995; Russell & Whitmarsh 2003) suggest that enough melt was emplaced at deep levels about the time of chron M5–M3 to form recognizable magnetic anomalies, and small amounts of melt emplaced at earlier times might help account for older anomalies. A second possibility is that the anomalies are generated by magnetization of the serpentinites. Magnetic studies of serpentinites cored on both margins show NRM intensities of 1–9 A m21, comparable to values in basalts, suggesting that they could generate significant magnetic anomalies (Zhao 1996; Shipboard Scientific Party 2004). In this case, the magnetization contrast between peridotite ridges and intervening sedimentary basins might explain at least part of the weak magnetic anomalies. Observed co-linearity of magnetic anomalies and peridotite ridges in the TZ is consistent with this idea (Fig. 3).
Oceanic crust Normal oceanic crust is considered by most investigators to have formed in the Newfoundland–Iberia rift beginning by about chron M3, but certainly no later than chron M0 (c. 121 Ma, earliest Aptian). There is less information on the velocity structure of this basement than that of the TZ landward of chron M3 and unfortunately there are no refraction profiles that extend seaward onto Albian crust, i.e. beyond the time when the previously defined ‘breakup unconformity’ is thought to have formed (Mauffret & Montadert 1987). The IAM9 profile seaward of anomaly M3 on the Iberia margin (Fig. 4g, section I2) shows crustal thickness and velocity that appear to be normal for oceanic crust, except at the western end where crust thickens significantly as the profile intersects a seamount. On the opposing margin (Fig. 2), the SCREECH3 profile seaward of anomaly M3 also has crustal thickness and velocity expected for normal ocean crust (Lau et al. 2006b). It is noteworthy, however, that the seaward ends of both the IAM9 and SCREECH3 profiles lie near seamounts and the northern edge of a zone that may have been influenced by plume magmatism (see below). In contrast to these profiles, the SCREECH2 profile (Fig. 4g, section N2) shows apparent oceanic crust that is only 2 –4 km thick above a 7.2–8.0 km s21 layer that has highly variable thickness and may be serpentinized mantle. Within approximately 80 km to either side of this line, refraction profiles over supposed oceanic crust at anomaly M0 (Srivastava et al. 2000) and
BREAKUP OF THE NEWFOUNDLAND– IBERIA RIFT
up to 50 km seaward of M0 (Todd & Reid 1989) also show a thin, low-velocity layer (2–2.5 km, 4.5–5.3 km s21) over probably serpentinized mantle (6.9–7.5 km s21) that is at least 4–5 km thick. To the north, the SCREECH1 profile (Figs 2 & 4 g, section N3) shows a large thickness (up to 5 km) of apparently serpentinized mantle with velocities of 7.2–7.9 km s21 (Funck et al. 2003). The overlying basement layer is very thin (1–3 km), and is interpreted as oceanic on the basis of velocity structure and reflection characteristics (Hopper et al. 2004). Normal-thickness oceanic crust is interpreted at the seaward end of the profile beyond anomaly M0, although this is not well constrained by the refraction data. The conjugate Galicia Bank profile (Fig. 4g, section I3) also shows a short section of apparent oceanic crust near its western end (Whitmarsh et al. 1996). A separate feature of the young M-series basement that suggests a magmatic component during its emplacement is the presence of the highamplitude J anomaly. This magnetic anomaly is thought to have formed in association with anomalous melt production and northward intrusion from the southern margin of the rift from about the time of chron M4 to a time younger than M0 (Tucholke & Ludwig 1982) (Fig. 4c; see the subsection on ‘Rift magmatism’ later). The anomaly amplitude is highest at the southern edge of the rift in the area of the Southeast Newfoundland Ridge (SENR), J Anomaly Ridge (JAR) and Madeira–Tore Rise (MTR) (Fig. 1); away from this area the high amplitude is confined to the zone around chrons M1 and M0 (Rabinowitz et al. 1978). Within the Newfoundland–Iberia rift the anomaly gradually loses its high amplitude north of the Newfoundland Seamounts and Tore Seamount on the Newfoundland and Iberia margins, respectively (bold lines, Figs 2 & 3). Farther north, anomaly M0 is identified with normal amplitudes between Flemish Cap and Galicia Bank, and it is also interpreted to be present in the Bay of Biscay (Sibuet et al. 2004b). If we assume that the amplitude of the J anomaly reflects the distribution of magma introduced from the southern edge of the rift, the effect reached to near the latitude of the ODP Leg 149– 173 drilling transect. Thus, the thickened crust in the young part of the IAM9 and SCREECH3 profiles might reflect this influence (e.g. Fig. 4g, section I2). The above data show that the structure of basement younger than approximately M3 is highly variable, ranging from thin crust with apparently limited serpentinization of underlying mantle, to thin with seemingly extensive serpentinization, to normal or thicker than normal with no clear serpentinization. Thus, in some places there appears to be a magmatic component while in others none is
17
evident. It is noteworthy that in the two locations where the presumed oceanic crust has been drilled, it is exhumed mantle. Basement at Site 1070, west of anomaly M3 on the Iberia margin (Figs 3 & 4g, section I2), is serpentinized peridotite with minor gabbro intrusions (Whitmarsh et al. 1998). Basement at Site 1277 near anomaly M1 off Newfoundland is also serpentinized peridotite beneath a thin interval of allochthonous basaltic, gabbroic and serpentinite debris (Figs 2 & 4g, section N2; see also Fig. 10 later) (Tucholke et al. 2004). It is well known that serpentinized peridotites can have compressional-wave velocities ranging from less than 5 km s21 up to normal mantle velocities of 8þ km s21, depending on the degree of serpentinization (e.g. Christensen 2004). Thus, in the absence of other constraints such as Poisson’s ratio or velocity anisotropy, and considering that serpentinites were recovered at Sites 1070 and 1277, it is possible that much of the basement layer that is interpreted from seismic velocity data to be normal oceanic crust in the area of the young M-series anomalies is variably serpentinized peridotite.
Development of the rift Pre-rift crust Continental crust in which the Newfoundland– Iberia rift developed was an assemblage of Precambrian– Palaeozoic rocks that accreted during the closing of the Palaeozoic Iapetus and Rheic oceans. Along the present Atlantic Canadian margin these rocks form the Appalachian Orogen, and from NW to SE they appear in three terranes comprising the Dunnage, Gander and Avalon zones (Fig. 1) (Williams & Hatcher 1982, 1983). The Avalon Zone occupies easternmost Newfoundland and most of the Grand Banks platform. Another zone (Meguma) lies south of the Avalon terrane at the southern edge of the Grand Banks, and a prominent magnetic anomaly (Collector Anomaly) marks its northern boundary. The first three terranes were accreted during the Taconian, Salinic and Acadian orogenies (Early–Middle Ordovician, Early Silurian and Devonian respectively) (van Staal et al. 1998; Waldron & van Staal 2001; Percival et al. 2004). The Meguma probably was emplaced during the Acadian orogeny, but it may have been reactivated during the Alleghanian orogeny (Carboniferous –Permian). Large granitic plutons were emplaced in the then-existing terranes during the Acadian orogeny. The Gander–Avalon and Avalon–Meguma boundaries are steep ductile shears and brittle faults that imply assembly by transcurrent motion rather than subduction/obduction (Keen et al. 1990). Original structural fabrics within the Avalon and Meguma terranes are well
18
B. E. TUCHOLKE ET AL.
defined by gravity and magnetic data, and in many places they appear to have controlled subsequent development of rift basins within the Grand Banks (Welsink et al. 1989). Continental basement of the Iberia margin was accreted by closure of Palaeozoic ocean basins and suturing during the Hercynian (Variscan) orogeny in approximately Middle Devonian –Carboniferous time, i.e. at the same time as the late Acadian and Alleghenian orogenies in North America. In Iberia, the accreted terranes have NW–SE orientations (Fig. 1); offshore, they are thought to have curved back to the NE and east to form the Ibero-Armorican Arc, continuous with approximately east– west sutures in NW France (Armorican Massif ), SW Britain and Germany (Ziegler 1982; Capdevila & Mougenot 1988; Silva et al. 2000). These terranes were intruded by large granitoid batholiths both during and following the Hercynian orogeny (Pinheiro et al. 1996). Tonalites also were intruded in the original continental crust of Galicia Bank (ODP Sites 1067) during the Hercynian and earlier (Rubenach 1999). Late Hercynian deformation included complex brittle faulting in generally NE–SW sinistral and conjugate NW–SE dextral shears (e.g. Ribeiro et al. 1990), and these features exerted significant control on the subsequent structure of the Mesozoic rifted margin (Wilson et al. 1989).
Mesozoic rifting Rifting between Newfoundland and Iberia occurred primarily in two phases. The initial phase occurred in the Late Triassic –earliest Jurassic, at which time rift basins formed over a broad region within the Grand Banks (e.g. Jeanne d’Arc, Whale and Horseshoe basins; Fig. 1) and the Iberia margin (Lusitanian Basin, possibly Porto and Galicia Interior basins: Murillas et al. 1990). The Grand Banks basins accumulated siliciclastic ‘red-bed’ sediments during Carnian– Norian time, and these were succeeded by evaporite deposits that reached into the earliest Jurassic (Hettangian– Sinemurian) in both the Grand Banks and Lusitanian basins (Jansa & Wade 1975; Wilson 1988; Rasmussen et al. 1998; Alves et al. 2003). During the Early and Middle Jurassic, the Grand Banks experienced epeirogenic subsidence without further rifting (Tankard & Welsink 1987) while an episode of faulting and basin subsidence occurred in the Lusitanian Basin south of the Nazare Fault (Rasmussen et al. 1998). A second phase of extension began in the Late Jurassic. During this phase, extension shifted from a wide-rift mode that encompassed future proximal margins, and it focused at distal margins where continental crust eventually separated (e.g. Manatschal & Bernoulli 1999). On the Grand Banks, extension
appears to have persisted with variable intensity from Oxfordian to Aptian time, with the most intense rifting occurring in the Kimmeridgian – Tithonian (Tankard & Welsink 1987). Unconformities that were developed in the Jeanne d’Arc Basin in late Barremian–early Aptian and late Aptian time (Fig. 4f ) have been interpreted by Driscoll et al. (1995) to have formed during the final rifting events on the main Grand Banks, immediately preceding the onset of sea-floor spreading. After this time, extension shifted to the northern margin of the Grand Banks (e.g. Orphan Basin) where it was associated with opening of the Labrador Sea (Enachescu 1988; Grant et al. 1988). On the Iberia margin, rifting in the Lusitanian Basin was limited largely to the Oxfordian–early Kimmeridgian, although there may have been later minor extension in Tithonian–Barremian time (Wilson 1988; Wilson et al. 1989; Rasmussen et al. 1998). In the Galicia Interior Basin to the north, Murillas et al. (1990) analysed seismic stratigraphy and ties to wells on the shallower Iberia margin. Their interpretations show that probable Tithonian– Berriasian shallow-water carbonates (seismic sequence 6, Fig. 4b) are highly faulted, while the overlying Valanginian–Hauterivian sequence 5 is only locally faulted, primarily towards the centre of the basin (see also Pe´rez-Gussinye´ et al. 2003). Thus, rifting in the Interior Basin appears to have occurred largely in Berriasian time, with some extension continuing into the Valanginian. On thinned continental crust at the deep western and southern margins of Galicia Bank, drilling and dredge/submersible sampling have recovered shallow-water Tithonian– (?)Berriasian carbonates and claystones of sequence 6, succeeded by deeper-water carbonates and clastics (Boillot et al. 1989c; Whitmarsh & Sawyer 1996; Whitmarsh & Wallace 2001). From these data the first rifting of Galicia Bank has been interpreted to date to the Berriasian (Fig. 4e). Rifting continued through the Valanginian and Hauterivian on the western margin of Galicia Bank, disrupting sequences 6 and 5 (Figs 5–7) (Mauffret & Montadert 1987). Some parts of the southern margin also show signs of extension during this period (Figs 8b & 9) but other parts were inactive (Fig. 8c). The last stage of rifting on the Iberia margin is generally interpreted to have occurred in the latest Aptian –earliest Albian (Mauffret & Montadert 1987; Boillot & Winterer 1988). Wilson (1988) discussed an Aptian unconformity in the Lusitanian Basin (Fig. 4f ) and suggested that it developed at the onset of sea-floor spreading. In contrast, Rasmussen et al. (1998) did not identify this event in the basin, although their figures show a local unconformity between Lower and Upper Cretaceous beds that may be equivalent. On the deep Iberia
Fig. 5. (a) Section of ISE 14 multichannel reflection profile across the SW margin of Galicia Bank. Location in Figure 3 (ISE 14E). The basement highs are continental fault blocks. Closed arrows mark the Aptian event between seismic sequences 3 and 4, and open arrows mark the unconformity between sequences 4 and 5. Note the level attitude of sequence 4. See Figure 4b for sequence ages and lithology. (b) Enlarged segment of the profile with interpreted structure and stratigraphic relations of the deep sedimentary sequences. Sequence 6 capping basement is probably Tithonian– Berriasian limestone and dolomite. Sequence 5 (Valanginian–Hauterivian) shows little divergence of sedimentary reflections and was faulted rapidly together with sequence 6 near the end of the Hauterivian. Sequence 4 (Barremian– Aptian) is horizontal and was not faulted or rotated. Note the unconformable onlap of sediments onto sequence 4.
BREAKUP OF THE NEWFOUNDLAND– IBERIA RIFT 19
20
B. E. TUCHOLKE ET AL.
Fig. 6. (a) Section of ISE 2 multichannel reflection profile across the western margin of Galicia Bank just south of the ODP Leg 103 drilling transect. Location in Figure 3. The basement ridges at the right are continental fault blocks above the S reflection, which is interpreted to be a detachment contact with underlying mantle (e.g. Reston et al. 1995); some of the distortion of S is probably an artifact of velocity pull-up in this time section. Closed arrows at top mark the Aptian event between seismic sequences 3 and 4, and open arrows mark the unconformity between sequences 4 and 5. Note the strong reflectivity of sequence 4; reflectivity is particularly strong near the peridotite ridge at left, which suggests that the sequence contains abundant serpentinite debris shed from the ridge. See Figure 4b for sequence ages and general lithology. (b) Enlarged segment of the profile with interpreted structure and stratigraphic relations of the deep sedimentary sequences. Rotated and diverging reflections in sequence 5 document extension in the Valanginian– Hauterivian (note that divergence of reflections would be enhanced in a depth section because of velocity increase with depth). Sequence 4 (Barremian–Aptian) was not faulted. Note the unconformable onlap of sediments onto sequence 4. ODP Site 640, 7 km to the north, penetrated about 65 m of deformed Hauterivian –Barremian sandy turbidites and marlstones that correspond to the top of sequence 5 and the base of sequence 4 (Boillot et al. 1987).
BREAKUP OF THE NEWFOUNDLAND– IBERIA RIFT
21
margin, a prominent, widespread reflection between Barremian–Aptian seismic sequence 4 and overlying Albian sediments has been interpreted as the ‘breakup unconformity’ (Groupe Galice 1979; Mauffret & Montadert 1987, 1988). This Aptian event (Figs 5–8) is the same as the ‘orange reflection’ at DSDP Site 398 on the southern part of Galicia Bank (Figs 4b & 9). The end-of-Aptian breakup that has been interpreted off Iberia is nearly synchronous with the development of the late Aptian ‘rift-onset’ unconformity in the Jeanne d’Arc Basin (Fig. 4f ) (Driscoll et al. 1995). In the offshore Newfoundland TZ, recent drilling at ODP Site 1276 dated a strong, basin-wide reflection as the same age (U reflection, Figs 4b & 10) (Tucholke et al. 2004). The seismic- and time-stratigraphic similarity of this horizon and the Aptian event off Iberia indicates that both originated from the same process. Together, these seismic markers document a major event during the evolution of the rift. The above interpretations that rifting continued throughout the entire Early Cretaceous into Aptian time are consistent with radiometric age dates on seafloor submersible/dredge samples and drill cores in the deep basin (Fig. 4c). Metagabbros that were emplaced in continental basement at ODP Sites 900 and 1067 during Hercynian time (Rubenach 1999; Manatschal et al. 2001) passed through the hornblende blocking temperature (c. 500 8C) at 161 + 1 Ma (Callovian) and through the plagioclase blocking temperature (c. 150 8C) at approximately 136–137 Ma (Berriasian) (Fe´raud et al. 1996); the latter age is interpreted to mark the time of exhumation (Manatschal et al. 2001). It is not known how much longer extension continued at this location, but it may have persisted through the Barremian or later (Whitmarsh & Wallace 2001). On peridotite ridges R1 and R2 bounding extended continental crust at the western margin of Galicia Bank (Fig. 3), granulites and sheared intrusive diorites and gabbros in the peridotites were cooled, either before or during exhumation, at ages ranging from 129.3 + 13.4 Ma to 117.7 + 1.8 Ma (approximately Hauterivian –middle Aptian) (Fe´raud et al. 1988; Boillot et al. 1989a; Fuegenschuh et al. 1998; Scha¨rer et al. 1995, 2000). To the south at ODP Site 1070, gabbro pegmatites in peridotites passed through the hornblende blocking temperature at 119 + 0.7 Ma (Aptian) and are interpreted to have been exhumed during the earliest Albian (110.3 + 1.1 Ma) as they passed through the plagioclase blocking temperature (Whitmarsh & Wallace 2001). Extension of basement at the SW margin of Galicia Bank during or following late Aptian time has also been interpreted at ODP Sites 897 and 899, both of which were drilled on the tops of
22
B. E. TUCHOLKE ET AL.
Fig. 7. (a) Section of GP 101 multichannel reflection profile over continental crust near ODP Sites 638, 639 and 641 on the western margin of Galicia Bank (from Winterer et al. 1988). Location in Figure 3. Site 639 is about 0.5 km south of the profile, and Sites 638 and 641 are about 1 km to the north. Closed and open arrows mark the tops of seismic sequences 4 and 5, respectively. See Figure 4b for sequence ages and lithology. (b) Enlarged section from the western part of the profile with interpreted structural and stratigraphic relations. In its upper part, sequence 5 (Valanginian–Hauterivian) shows divergence and rotation of sedimentary reflections, indicating that extension here occurred largely in the Hauterivian (note that divergence of reflections would be enhanced in a depth section because of velocity increase with depth). Sequence 4 (Barremian– Aptian) lacks these features and shows subhorizontal onlap at the basin edges; a chaotic mass at the base of the sequence may be a slump from the basement ridge to the east. Indicated ages were determined at the drill sites (Boillot & Winterer 1988). Hole 639D recovered Tithonian limestones, and Holes 639E and 639F recovered rhyolites that are interpreted to come from continental, possibly Palaeozoic basement (BMT). Hole 639A penetrated lower Valanginian marlstone and sandstone and bottomed in dolomite, which was also cored at the bottoms of Holes 639B and 639C. The dolomite is barren of microfossils but must date to the latest Jurassic –earliest Cretaceous. The S reflection (see Fig. 6) underlies the continental block on which these sites were drilled. (c) Enlarged section from the eastern part of profile in a with interpreted structural and stratigraphic relations (note that vertical exaggeration is slightly enhanced compared to profile in a). Reflections diverging towards the fault footwall in the upper part of sequence 5 suggest basin extension in late Valanginian– early Hauterivian time. Barremian–Aptian sequence 4 exhibits horizontal reflections that lap unconformably onto the basin perimeter, and it in turn is unconformably overlain by sequence 3.
BREAKUP OF THE NEWFOUNDLAND– IBERIA RIFT
23
peridotite ridges (Fig. 3) (Comas et al. 1996). At these sites, serpentinized peridotite olistostromes and debris flows were emplaced on the basement during early Aptian and early late Aptian time, respectively, before faulting and uplift formed the ridges.
Rift magmatism During most of the Late Jurassic–Early Cretaceous phase of extension, there was only minor magmatism in the rift (Fig. 4c). Tithonian–early Valanginian lamprophyre dykes and a small ultramafic intrusion are documented in Newfoundland, and recoveries of basalt and volcaniclastics both of this age and of late Barremian–Aptian age have been reported from wells in the South Whale Basin and southern Grand Banks (Jansa & Pe-Piper 1988; Sinclair 1988 and references therein). Dykes and sills of primarily Berriasian–Valanginian age are present in the southern Lusitanian Basin of Iberia (Montenat et al. 1988; Pinheiro et al. 1996). All these occurrences are volumetrically minor in the context of the larger rift and all are restricted to the southern part of the rift. However, the age of their emplacement is consistent with the timing of the most intense rifting in the latest Jurassic–Early Cretaceous (Fig. 4). By far the most extensive magmatism within the rift dates to the Barremian–early Aptian (approximately chron M4 to younger than M0; Fig. 4c) (Tucholke & Ludwig 1982). At this time the SENR and JAR formed on the North American side of the plate boundary, and the conjugate MTR and perhaps parts of Gorringe Bank (GRB; later deformed during the Cenozoic) were emplaced on the Iberia side (Fig. 1). Magmatism was centred at the southern edge of the rift in the position of the SENR –GRB, and at approximately chron M2– M0 (Barremian) it was channelled both southwards along the Mid-Atlantic Ridge axis to form the JAR –MTR (Tucholke & Ludwig 1982) and northwards into the Newfoundland –Iberia rift to form comparable basement edifices that reach toward the present positions of the Newfoundland Seamount and Tore Seamounts (Fig. 1) (Tucholke et al. 1989). At least at the southern margin of the rift, some of this magmatic crust was at and above sea level. Deep-sea drilling at DSDP Site 384 near anomaly M2 on the JAR recovered upper Barremian/lower Aptian –lower Albian reef deposits overlying tholeiitic basalt that most probably was subaerially weathered (Tucholke et al. 1979); the reef appears subsequently to have been exposed to meteoric waters, probably during Albian time, before it subsided to its present depth of 4100 m. Overall, the broad morphological expression of the SENR – JAR –MTR magmatic complex was very much like that of modern Iceland and the associated
Fig. 8. (a) Section of ISE 10 multichannel reflection profile along the southern margin of Galicia Bank. Location in Figure 3 (ISE 10E). The basement ridges are continental fault blocks, based on their continuity with interpreted continental crust at the ODP drill sites 30–40 km to the south. Closed arrows mark the Aptian event between seismic sequences 3 and 4, and open arrows mark the unconformity between sequences 4 and 5. Note the strong reflectivity and drape-and-fill character of sequence 4. See Figure 4b for sequence ages and lithology. (b) Enlarged segment of the profile with interpreted stratigraphic relations of the deepest sedimentary sequences. Sequence 5 (Valanginian – Hauterivian) in its lower part shows rotation and divergence of sedimentary reflections towards the fault footwall, indicating basin extension in the Valanginian. Note that sequence 4 (Barremian –Aptian) shows horizontal attitude and onlap at the basin edges and that sequence 3 laps unconformably onto sequence 4. (c) Enlarged segment of the profile in a with interpreted stratigraphic relations of the deepest sedimentary sequences. There is no apparent rotation or divergence of reflections in either sequence 4 or 5, indicating that extension in this basin ceased by about late Berriasian time. Note the chaotic debris immediately below the Aptian event on the right, indicating mass wasting from the adjacent basement ridge.
24 B. E. TUCHOLKE ET AL.
BREAKUP OF THE NEWFOUNDLAND– IBERIA RIFT
25
Fig. 9. (a) Section of GP 19 multichannel reflection profile across DSDP Site 398 on the southern margin of Galicia Bank. Location in Figure 3. (b) Interpretation of the profile, with seismic sequences as in Figure 4b. Sequence 5 is faulted but sequence 4 is basin-filling and relatively flat-lying. Note the strong reflectivity at the top of sequence 4. Ages determined in the lower part of the borehole are indicated.
Reykjanes and Kolbeinsey ridges (although at a somewhat smaller scale), and it appears likely that the magmatism was associated with a mantle plume. Either a thicker or more magnetized basaltic crust emanating from this plume probably accounts for the development of the magnetic J anomaly between M1 and M0, which has a high amplitude near the SENR–GRB but decays with distance both to the north (Figs 2 & 3) and south (Rabinowitz et al. 1978; Tucholke & Ludwig 1982). The younger age limit of this magmatic complex is not well constrained, but it appears that magmatism retreated rapidly towards the southern margin of the rift following chron M0 and died out by early Albian time. Post-rift magmatism in the Newfoundland Basin is observed in the form of diabase sills that were injected into sediments at and below the U reflection at ODP Site 1276 during late Albian –early Cenomanian time (105 –98 Ma; Hart & Blusztajn 2006). These sills are not clearly observed in seismic reflection records around the drill site
(Fig. 10), but probable sills at this stratigraphic level appear in profiles from the southern part of the basin (Fig. 11). The source of the magma is uncertain, but it may have been associated with formation of the Newfoundland Seamounts near the centre of the basin (Karner & Shillington 2005); a single age date indicates that one of these seamounts formed at 97.7 + 1.5 Ma (Sullivan & Keen 1977). Only minor magmatism of this age is documented on the Iberia margin (Fig. 4c), and the event seems primarily to have affected the Newfoundland margin.
Questions related to breakup The features of the Newfoundland–Iberia rift described above, and prevailing interpretations that the ‘breakup unconformity’ dates to the end of the Aptian, lead to several fundamental problems in understanding final continental separation and the initiation of sea-floor spreading. First, existing
Fig. 10. (a) Section of SCREECH2 multichannel seismic reflection profile (Shillington et al. 2004) from the seaward part of the TZ in the north-central Newfoundland Basin. Location in Figure 2 (SCREECH 2NW). Locations of magnetic anomalies M3 and M1 are indicated. ODP Site 1276 did not reach basement but penetrated 105 –98 Ma base sills that intrude uppermost Aptian –lowermost Albian sediments at the level of the U reflection, which is the time-stratigraphic equivalent of the ‘orange reflection’ and Aptian event on the Iberia margin (see Fig. 4). Note the basin-levelling character of U. The basement surface is poorly imaged beneath the U reflection because of strong reflectivity of U, because of low impedance contrast with the U-basement interval, or both. Drilling at Site 1277 recovered massive serpentinized peridotite, interpreted as basement, beneath approximately 57 m of mass-wasting debris consisting of basalt, gabbro and serpentinite (Tucholke et al. 2004). (b) Enlarged segment of the profile with interpreted structural and stratigraphic relations. If the identification of U is correct, the sediments of seismic sequence A and the underlying basement block were rotated during approximately latest Aptian time, prior to deposition of overlying, onlapping Albian sediments of sequence B. See Figure 4 for sequence ages and lithology.
26 B. E. TUCHOLKE ET AL.
BREAKUP OF THE NEWFOUNDLAND– IBERIA RIFT
27
Fig. 11. Section of Conrad 2510 multichannel reflection profile from the southern Newfoundland Basin (Tucholke et al. 1989). Location in Figure 2 (NB26). See Figure 4b for sequence ages and lithology. Sedimentary and basement structure below the U reflection (Aptian event) is poorly imaged because of the strong reflectivity of U, because of low impedance contrasts, or both. Short sections of high-amplitude reflections near the level of U probably are diabase sills, similar to those drilled at this stratigraphic level at ODP Site 1276 in the northern part of the basin (see Fig. 10).
magnetic analyses (e.g. Russell & Whitmarsh 2003) and refraction studies (Dean et al. 2000; Lau et al. 2006b; Van Avendonk et al. 2006) suggest that some form of oceanic crust was emplaced in the rift at least by the time of chron M3 (late Hauterivian –Barremian) (Figs 2 & 3). Thus, the Aptian event would appear not to be a ‘breakup unconformity’ in the conventional sense that it correlates to the final separation of continental crust and coincident onset of sea-floor spreading (Falvey 1974). Second, there is a conflict between previously described evidence for extension during Barremian–Aptian time and the corresponding structural and stratigraphic record as imaged in seismic reflection records both over continental crust of Galicia Bank and over the Newfoundland and Iberia TZs. Sediments of this age in seismic sequence 4 off Iberia and sequence A off Newfoundland (Fig. 4b) are mostly horizontal, and they rarely are faulted or show splayed reflections that we would expect if they were deposited over rotating fault blocks during extension (Figs 5–15).
Finally, in at least three cases, there was a significant lag between the time when upper mantle was first emplaced in the rift and a time when it was later extended. One instance is at peridotite ridges R1 and R2 at the western margin of Galicia Bank. Recent magnetic anomaly compilations (Miles et al. 1996) and identifications (Srivastava et al. 2000; Sibuet et al. 2004b) indicate that anomaly M0 lies approximately 50 km west of these ridges (Fig. 3). If we extrapolate basement age east of anomaly M0, assuming an accretion rate of 9 mm year21 (the average rate between anomaly M0 and anomaly 34 to the west), the age of the ridges is about 126–127 Ma (see section I3, Fig. 4g). Gabbro protoliths of chlorite rocks in these ridges were emplaced at 122.1 + 0.3 to 121.7 + 0.4 Ma (Scha¨rer et al. 1995, 2000) and subsequently sheared during exhumation. In the area of ODP Site 637, tectonized diorite dykelets in peridotite mylonites cooled below approximately 500 8C at about 122 Ma and below approximately 150 8C at 117.7 + 1.8 Ma as they were exhumed (Fe´raud et al. 1988; Boillot et al. 1989a). Thus, exhumation of the peridotites appears to post-date
28
B. E. TUCHOLKE ET AL.
initial emplacement of the basement by at least 4–8 million years. A similar interval between the time of basement emplacement and later extension is indicated by the study of Comas et al. (1996) at ODP sites 897 and 899 farther to the south. According to magnetic anomaly identifications (Fig. 3), basement ages there are approximately 125– 126 Ma, but faulting and uplift of the peridotite ridges did not occur until late Aptian time (c. 115 Ma) or later, a delay of at least 10 million years. An even larger time lag is observed at ODP Site 1070. Basement there was emplaced at about 124 Ma according to the identification of anomaly M1 (Fig. 3), and small gabbro pegmatites intruded the peridotites at essentially the same time (127 + 4 Ma; Beard et al. 2002). The gabbros cooled through approximately 600 8C at 119 + 0.7 Ma and through about 150 8C at 110.3 + 1.1 Ma (earliest Albian) (Whitmarsh & Wallace 2001). These small intrusions would have equilibrated quickly with the temperature of surrounding peridotites, so their long cooling history indicates that they were emplaced in hot rock at significant depth before they were later exhumed. Assuming a geothermal gradient of 25–30 8C km21 (continental, Whitmarsh & Wallace 2001) to approximately 90 8C km21 (oceanic, Expedition 309 Scientists 2005), the rocks were still at a depth of about 2–6 km at c. 110 Ma. The exhumed peridotite is covered by upper Aptian sediments (Whitmarsh et al. 1998), potentially as young as 112.2 Ma but still at least 0.8 million years older than the earliest possible exhumation suggested by the radiometric date (111.4 Ma). This inconsistency probably can be attributed to uncertainties in the absolute ages. However, unless the radiometric date is grossly in error, it appears that exhumation was rapid and that it post-dated the age of original basement emplacement by approximately 14 million years. We suggest that these discrepancies can best be explained by elevated plate-wide tensile stress and rifting that persisted throughout the Early Cretaceous until the end of Aptian time. Below, we examine this hypothesis by evaluating the deep structure and stratigraphy of the conjugate Newfoundland and Iberia margins.
Deep stratigraphy and structure of the rift Seismic sequences To examine the timing and character of rift events we analysed the deep stratigraphy and structure of Newfoundland and Iberia margins using existing seismic reflection profiles and correlations to DSDP and ODP drill sites (Figs 2 & 3). The Aptian event can be traced continuously throughout
the seismic grid except at the top of Galicia Bank where jump correlations have to be made on the basis of seismic character (e.g. Fig. 7a, right-hand side). Deeper seismic sequences in the Galicia Bank area commonly are in graben separated by basement ridges. Where they could not be traced continuously between basins these sequences were correlated on the basis of reflection character, relations to adjacent sequences (including development of bounding unconformities) and the limited control provided at drill sites (Figs 7 & 9). On the Iberia margin we used profiles from the Iberia Seismic Experiment (ISE) lines (Henning et al. 2004), CAM lines (Discovery 215 Working Group 1998; Chian et al. 1999), GP lines (e.g. Mauffret & Montadert 1987; Winterer et al. 1988), and the Lusigal 12 (Beslier 1996) and IAM9 lines (Pickup et al. 1996). On the Newfoundland margin we analysed profiles from R/V Conrad cruise 2510 (Tucholke et al. 1989), Lithoprobe (de Voogd & Keen 1987), ERABLE (Srivastava & Sibuet 1992) and R/V Ewing cruise 2007 (SCREECH program: Hopper et al. 2004; Shillington et al. 2004). The principal seismic sequences and their lithological correlations on the Iberia margin (Fig. 4b) were defined by Groupe Galice (1979) and later extended to include sequence 5 (Boillot et al. 1979, 1987). Mauffret & Montadert (1987) subdivided sequence 5 into 5a and 5b, and subsequently renamed 5a as 5 and 5b as 6 (Mauffret & Montadert 1988); this last classification is adopted in Figure 4. The age of the sequence 6–5 boundary is not well constrained; drilling at ODP Site 639 (Fig. 7) recovered lower Valanginian sediments over undated (possibly Berriasian) dolomites near the top of sequence 6 (Boillot & Winterer 1988), and these in turn overlie Tithonian shallow-water limestones. In Figure 4 we tentatively place the sequence boundary at the Berriasian–Valanginian boundary. The sequence 5 –4 boundary at nearby Site 638 dates to late Valanginian–Hauterivian, depending on the fossil group considered (Boillot et al. 1988). At Deep Sea Drilling Project (DSDP) Site 398 on the southern Galicia Bank (Fig. 9), the boundary is probably upper Hauterivian –lower Barremian (Sibuet & Ryan 1979). Although we place the 5–4 boundary at this latter level in Figure 4, it is possible that the boundary is somewhat diachronous and reflects locally different timing of extension events around Galicia Bank. The sequence 4– 3 boundary (Aptian event) is manifested as the ‘orange reflection’ in the area of Galicia Bank (Fig. 9) (Groupe Galice 1979), and the correlative sequence A –B boundary in the Newfoundland Basin is represented by the highamplitude U reflection (e.g. Figs 10 & 15) (Tucholke et al. 1989). Dates on the Aptian event consistently place it in the uppermost
BREAKUP OF THE NEWFOUNDLAND– IBERIA RIFT
Aptian –lowermost Albian where it has been drilled at Sites 398 and 641 on Galicia Bank (Sibuet & Ryan 1979; Boillot et al. 1987) and at Site 1276 on the Newfoundland margin (Tucholke et al. 2004).
Tithonian – Berriasian (sequence 6) Seismic sequence 6 (Figs 5–8) contains carbonates and mudstones deposited in water less than a few hundred metres deep on continental crust of the Iberia margin during the Tithonian and possibly Berriasian (Boillot & Winterer 1988; Whitmarsh & Wallace 2001). These sediments often are difficult to distinguish seismically from the underlying basement, probably because they include coarse, high-velocity debris from the underlying Palaeozoic basement as well as high-velocity dolomites (Boillot & Winterer 1988). Extensional faulting of this sequence occurred at two different times, as illustrated by contact relations with the overlying sequence 5. Over much of Galicia Interior Basin and the southern margin of Galicia Bank (Fig. 8), sequence 5 is not conformable to sequence 6 or to basement (Murillas et al. 1990), indicating that an episode of rifting faulted sequence 6 shortly before the Valanginian. Because the internal structure of sequence 6 is hard to detect, however, it is difficult to determine (e.g. by divergence of reflections) whether the extension began much earlier, for example in the Tithonian. In a few cases where internal reflections are observed (Figs 5b & 6b), they are mostly parallel. This suggests that faulting was confined to a short period in the Berriasian, and it is consistent with the development of unconformities in the Lusitanian and Porto basins (Fig. 4f ) that are ascribed to rifting. Age dates of 136–137 Ma on metagabbros at ODP Sites 900 and 1067 on the southern edge of the Galicia Bank crust show that basement there was being cooled and presumably exhumed near the end of Berriasian time (Fig. 4c) (Manatschal et al. 2001). In contrast to the above observations, sequence 5 is largely conformable to sequence 6 over the western, distal margin of Galicia Bank (e.g. Figs 5–7) and both sequences were faulted together. Numerous published reflection profiles show this relationship (e.g. Mauffret & Montadert 1987, 1988), indicating that most of the extension at the distal, western Galicia margin post-dates extension in the Interior Basin (see the next section). On the Newfoundland margin, time-stratigraphic equivalents of sequence 6 appear in the deep sedimentary basins of the Grand Banks (Grant et al. 1988; Tankard & Welsink 1988; Balkwill & Legall 1989) and extend seawards into basins beneath the continental slope. These rocks are mostly shales– sandstones with interbedded limestones, and they reflect rapid fill that kept pace with fault-controlled
29
basin subsidence. Interpretation of well and seismic data indicates that the most intense extension was in Late Jurassic time, decreasing thereafter into the Early Cretaceous (Fig. 4e). A widespread unconformity in the Jeanne d’Arc Basin (Tankard & Welsink 1988) suggests enhanced extension there at the Jurassic–Cretaceous boundary. The above data illustrate two important aspects of the rifting process. First, although there was probably rift-wide extension throughout this period, normal faulting and basin development were often amplified during relatively short pulses. Second, these pulses were not necessarily synchronous or co-located within the rift. However, they do tend to group at the end of the Jurassic into the Berriasian, suggesting a general extensional episode that culminated approximately in Berriasian time.
Valanginian– Hauterivian (sequence 5) Seismic sequence 5, where drilled off Iberia, consists of limestones, marlstones, mudstones and turbidites that generally reflect deeper-water deposition than in the underlying sequence (Sibuet & Ryan 1979; Boillot et al. 1987). The sequence displays three kinds of configuration with respect to the underlying, faulted basement and/or sequence 6. By far the most common is a conformable relation in which internal reflections are parallel both to one another and to the underlying formation in rotated fault blocks (Fig. 5b). This configuration is widespread across the western margin of Galicia Bank. The general lack of reflections that diverge towards fault footwalls in sequence 5 indicates that normal faulting occurred during a geologically brief period of time, focused near the Hauterivian–Barremian boundary. This extensional episode may correlate with interpreted exhumation and cooling of granulites associated with peridotite ridges R1 and R2 at approximately 129–126 Ma (Fig. 4c) (Fuegenschuh et al. 1998). Less frequently, beds in sequence 5 thicken towards the footwall (Figs 6b, 7b & 8b). In the case of the profile in Figure 8b along the southern margin of Galicia Bank, weakly divergent reflections are present in the lower part of sequence 5 and relatively parallel reflections appear in the upper part. This suggests that the footwall comprising the Hobby High ridge, drilled at ODP Sites 900, 1067 and 1068 about 40 km to the south, was being exhumed during the Valanginian, but that extension slowed or ceased by Hauterivian time. This is consistent with the ages at which metagabbros drilled on Hobby High are interpreted to have been exhumed (Fig. 4c) (Manatschal et al. 2001). On parts of the western Galicia margin (Figs 6b & 7b, c) reflections diverging towards fault footwalls in
30
B. E. TUCHOLKE ET AL.
sequence 5 indicate that extension continued throughout the Valanginian–Hauterivian. Finally, in some places sequence 5 is represented by tectonically undeformed basin fill which demonstrates that extensional faulting ended no later than about Berriasian time (Fig. 8c). This is commonly observed in profiles across the Galicia Interior Basin (Murillas et al. 1990), although faulting of sequence 5 is apparent locally along the western margin of the basin (eastern margin of Galicia Bank) (e.g. Fig. 9). Where sequence 5 is faulted, it shows both diverging and parallel reflections, indicative of both prolonged and geologically brief extension at various locations. In the basins of the Grand Banks off Newfoundland, time-stratigraphic equivalents of sequence 5 have facies similar to sequence 6, noted above. In the Jeanne d’Arc Basin the sediments fine upwards, which Grant et al. (1988) interpreted as indicating peneplanation of the source area and, by inference, continued decline of crustal extension that would lead to elevation of the sources.
Significance of Hauterivian extension As noted above, most normal faulting that disrupts sequence 5 appears to have occurred during a relatively brief period of time, late during deposition of the sequence (i.e. in the late(?) Hauterivian and possibly into the Barremian, c. 127 –126 Ma). This was an interval associated with several significant events in the rift (Fig. 4). First, the voluminous magmatism that formed the SENR –JAR and the conjugate MTR began in the southern part of the rift about this time (Tucholke & Ludwig 1982). Perhaps not coincidentally, small volumes of magmatic products as well as other indications of magmatism also appeared farther north in the rift. Gabbro pegmatites were intruded into mantle at ODP Site 1070 (Beard et al. 2002), and within a few million years gabbros and diorites were intruded into the peridotite ridges farther north along the western margin of the Galicia Bank (Fe´raud et al. 1988; Boillot et al. 1989a; Scha¨rer et al. 1995, 2000). In addition, according to magnetic anomaly studies by Russell & Whitmarsh (2003), intrusive magnetic source bodies became more organized and linked at this time, beginning at about chron M5–M4. Results from refraction studies also suggest that ‘normal thickness oceanic crust’ developed by about 125 Ma on the Iberia margin (Fig. 4g, sections I2 and I3) (Whitmarsh et al. 1996; Dean et al. 2000) and on the Newfoundland margin (SCREECH3: Lau et al. 2006b), so this period is commonly interpreted to be the start of sea-floor spreading in the rift (e.g. Whitmarsh & Miles 1995; Russell & Whitmarsh 2003). However, accretion of
exceptionally thin oceanic crust up until about 117 Ma has been interpreted seaward of Flemish Cap, with possible normal thicknesses in younger crust (Fig. 4g, section N3) (Funck et al. 2003; Hopper et al. 2004). In the central Newfoundland Basin Van Avendonk et al. (2006) found no significant change in crustal thickness from the outer limit of continental crust (Fig. 2) seaward to the end of the refraction profile on approximately 114 Ma crust (Fig. 4g, section N2). Although Van Avendonk et al. (2006) interpreted most of this as thin oceanic crust, the recovery of peridotite basement at ODP Site 1277 seaward of anomaly M1 (Fig. 10) suggests that the basement may be dominantly exhumed mantle. A similar conclusion might be drawn for the Iberia side, where peridotite basement was drilled at ODP Site 1070 near anomaly M2 (open circle at youngest peridotite ridge: Fig. 4g, section I2) (Whitmarsh & Wallace 2001). From these observations, it seems that while a component of melt was introduced into the rift beginning at about chron M5 –M3, emplacement of basement at the plate boundary was still being accomplished primarily by exhumation of mantle. An increase in basement roughness is observed on the Newfoundland side of the rift at this time (Tucholke et al. 1989), and a similar change may have occurred on the Iberia side although it presently is not well documented by existing reflection data. It may be that the introduction of increased melt and heat facilitated longer-lived extension on normal faults, thus increasing the amplitude of the basement topography. This also could have temporarily reduced rift-wide tensile stress. However, the subsequent extension of mantle in existing ocean floor at ODP Sites 897, 899 and 1070 and at peridotite ridges R1 and R2 (Fig. 4c) shows that tensile stress was elevated in the Aptian during deposition of seismic sequence 4/A, discussed below.
Barremian– Aptian (sequence 4 and A) Where it has been cored at ODP drill sites off Iberia, seismic sequence 4 is dominated by muddy turbidites and debris flows derived largely from the adjacent continental margin (Sibuet & Ryan 1979; Boillot et al. 1987). These sediments accumulated at rates ranging from ,5 to c. 65 m per million years, with the highest rates in the Aptian at Site 398. Only the uppermost part of correlative sequence A has been cored in the Newfoundland Basin (Site 1276: Tucholke et al. 2004), but it has a similar facies consisting of sandy turbidites and mudstones. These sequences are characterized by several notable features in reflection profiles. Normally they are relatively flat-lying and their internal reflections are parallel to subparallel
BREAKUP OF THE NEWFOUNDLAND– IBERIA RIFT
(Figs 5, 6, 8 & 9). Where reflections diverge, their spacing usually increases toward basin centres rather than towards footwall fault blocks. Faulting and rotation of the sequences (e.g. Fig. 10) are not common but are observed in scattered locations, including the eastern margin of Galicia Bank (Pe´rezGussinye´ et al. 2003). Thus, most basins accumulated sediments in the absence of obvious extension. Where sequence 4 fills valleys between peridotite ridges on the Iberia margin, its reflectivity appears to be related to the basement composition. The rounded shape of peridotite ridge crests, as well as possible slumps and debris slides (Figs 12 & 14), suggests that the serpentinite was weak and mobile. In fact, drilling of peridotite ridges within the rift has consistently recovered serpentinite breccias and olistostromes that reflect mass wasting of the serpentinized basement (Whitmarsh et al. 1998; Tucholke et al. 2004). In many places mass wasting is manifested by exceptionally strong reflectivity of sequence 4 near the peridotite ridges and by weakening reflectivity with distance from the ridges (Figs 6 & 14). These deposits were probably emplaced by serpentine-rich turbidity currents and highly mobile debris flows. The decrease of reflectivity with distance from the ridges is most probably an effect of decreasing grain size. In other cases, mass wasting from both peridotite ridges and continental fault blocks is demonstrated by chaotic fill (Figs 7b & 12) and by instances where sequence 4 laps high onto the bounding fault blocks (Figs 6b & 8c). Where cored at ODP Sites 897 and 899 (see Fig. 12a), these materials are serpentine-rich olistostromes and breccias that were emplaced before faulting and uplift of the underlying peridotite ridges (Comas et al. 1996; Gibson et al. 1996). Thus, some ridge-flank deposits identified in reflection profiles may predate extension (e.g. Fig. 12b, right-hand side), although thickening of the deposits in valleys shows that mass wasting also occurred during or after the faulting. Elsewhere, the ridge-flank deposits clearly were emplaced following extension, and their configuration shows that they were derived largely from the local topography (Figs 6b & 8c). The above patterns are not well defined in sequence A in the Newfoundland Basin, although rounding of presumed serpentinite ridges and possible displacement of allochthonous blocks and debris are observed (Figs 10a & 13). Tucholke et al. (1989) observed rare instances where the Aptian event (U reflection) at the top of the sequence appears to truncate basement topography. A similar possible occurrence is shown on the left of Figure 13b. If these basement highs consisted of exceptionally weak serpentinite, it is conceivable that their exposed tops were eroded away by mass wasting or even by passing turbidity currents.
31
On a basin-wide scale, sequences 4 and A are widespread, laterally continuous and thick in locations away from protruding fault blocks that occur at the seaward limits of the sequences and at the edges of Galicia Bank. Sequence 4 is observed all along the northern part of the Iberia margin (and presumably along the southern margin as well, although it is not well mapped there), and sequence A extends some 600 km along the Newfoundland margin. The sequence thicknesses reach more than 1.5 km near the continental margins (Tucholke et al. 2004; Tucholke unpublished data). Thus, at a basin-wide scale the sequences appear to represent rapid deposition of sediments eroded from proximal-margin sources. A particularly noteworthy characteristic of the Aptian event (U reflection) at the top of sequence A in the Newfoundland Basin is its strong reflectivity (Tucholke et al. 1989). In the central and southern Newfoundland Basin the reflection is so strong that deeper structure is masked (Figs 11 & 15). Either the impedance of sequence A is so high that signal penetration is limited, the impedance contrast with basement is small (e.g. the shallow basement is strongly serpentinized), or both. In the northern part of the basin the masking is reduced, although deeper structure often is still difficult to detect (Fig. 10a). A major unconformity was eroded across the southern Grand Banks during the Early Cretaceous in response to deformation and doming associated with the Avalon uplift (e.g. Grant et al. 1988), and it is likely that coarse clastics shed from this region help to account for the strong reflectivity in at least the central to southern Newfoundland Basin. Off Iberia the Aptian event (orange reflection) is locally strong, but it generally has lower amplitude than off Newfoundland and it seldom masks underlying structure except where debris has been shed from peridotite ridges (Figs 6a & 14).
Albian – Cenomanian (sequence 3 and B) Seismic sequence 3 off Iberia and correlative sequence B off Newfoundland are Albian to Cenomanian in age (Fig. 4b). At Sites 398 and 641 off Iberia sequence 3 is largely laminated to massive claystone and mudstone (Sibuet & Ryan 1979). Sequence B off Newfoundland contains muddy to sandy turbidites at it base, changing upwards to mudstones, and then back to sandy turbidites in the Cenomanian at Site 1276 (Tucholke et al. 2004). Sedimentation rates at Sites 398 and 1276 varied from approximately 10 up to 100 m per million years and generally tapered off in the late Albian and Cenomanian, while rates at Site 641, although poorly determined, were at the low end of this range. A general trend towards
32
B. E. TUCHOLKE ET AL.
Fig. 12. (a) Section of ISE 10 multichannel reflection profile SW of Galicia Bank, just north of the ODP Leg 149–173 drilling transect. Location in Figure 3 (ISE 10W). Magnetic anomalies M1 and M3 are identified. The basement ridges are serpentinized peridotite where they were drilled to the south. Arrows mark the Aptian event between seismic sequences 3 and 4. Seismic sequence 5 predates emplacement of this basement and is not observed here. Note the very strong reflectivity of sequence 4, which probably has abundant serpentinite debris shed from the adjacent ridges. The reverse fault at the centre was caused by later Cenozoic compression within the plate (e.g. Boillot et al. 1979; Masson et al. 1994) and illustrates the weakness of fault zones in the serpentinized peridotite. (b) Enlarged segment of the profile with interpreted structure and stratigraphic relations. Note the chaotic character of sequence 4, which is interpreted to be mostly mass-wasting debris from the adjacent ridges, and the unconformable onlap of sediments above this sequence. Faults and an interpreted peridotite allochthon (shaded) are speculative, but they are consistent with the geological history of the adjacent ridge as indicated at Site 1070 (i.e. emplacement of the mantle in the rift and gabbro impregnation at about 124 Ma, followed by exhumation some 14 million years later; see text). The interval labelled ‘SP’ has somewhat lower reflectivity than the underlying basement and may be strongly serpentinized, much as has been postulated for poorly reflective upper basement observed on the IAM9 line to the south (Fig. 3) (Pickup et al. 1996).
BREAKUP OF THE NEWFOUNDLAND– IBERIA RIFT
33
finer-grained deposits in sequence 3/B compared to underlying sequence 4/A suggests reduced sediment supply from margin source areas. Seismic sequences 3 and B characteristically lap unconformably onto underlying sequences 4 and A, respectively (Figs 5–8 & 12 –15). This implies either a tectonic event or a sharp change in sedimentation pattern, or both, at the Aptian –Albian boundary.
Breakup of the rift Four puzzling features are observed in the Newfoundland –Iberia rift that require explanation: (1) the continued (or late) extension of peridotite basement at times up to 14 million years or more after it was emplaced at the plate boundary; (2) the lack of significant coeval extension manifested in the structural and stratigraphic relations of seismic sequence 4 off Iberia and sequence A off Newfoundland; (3) the sharp change in reflectivity marked by the Aptian event at the top of sequence 4/A; and (4) the seismic unconformity at this level. We propose that these can be explained by continued exhumation and rifting of subcontinental mantle lithosphere through Aptian time, followed by initiation of normal sea-floor spreading at the Aptian –Albian boundary, as follows. Stage A, late Aptian (Fig. 16a). Up to this time, sea floor was being emplaced primarily by exhumation of subcontinental mantle, and in-plane tensile stresses were elevated throughout the rift. The plates extended internally in local weak zones, probably where existing faults were highly serpentinized. Mass wasting and slumps contributed coarse, reflective, sometimes chaotic debris to sequence 4/A, particularly in the vicinity of peridotite ridges. These deposits were interbedded with turbidites derived from the proximal continental margins. Stage B, Aptian– Albian boundary (Fig. 16b). Normal sea-floor spreading began in latest Aptian –earliest Albian time. This introduction of heat and melt at the plate boundary caused in-plane tensile stress to drop sharply, resulting in a pulse of relative compression in the adjoining plates. The compression may have amplified topography in the deep basins and thus briefly enhanced mass wasting from basement highs. Although the common concave-upwards shape of the Aptianevent reflection between basement highs primarily reflects input of debris from adjacent ridges (e.g. Figs 6, 8 & 12), its curvature may have been augmented by the compression. The change in stress regime probably also caused differential vertical motions on the proximal continental margins, stimulating erosion and sediment transport to the
Fig. 13. (a) Section of SCREECH2 multichannel seismic reflection profile (Shillington et al. 2004) seaward of anomaly M0 in the northern Newfoundland Basin. Location in Figure 2 (SCREECH 2SE). The U reflection (Aptian event) marks the boundary between seismic sequences A and B, equivalent to sequences 4 and 3 on the Iberia margin (see Fig. 4b). Note the unusual, rounded form of the basement highs which suggests that they may be serpentinized peridotite that was subject to degradation by mass wasting. (b) Interpretation of part of the profile. Note the unconformable onlap of sediments onto the Aptian event (U reflection). The faults and possible allochthons (grey shade) are speculative, but are consistent with the idea that the basement is weak, serpentinized peridotite that failed easily; note the possible reverse fault in sequence A on the right. The low-reflectivity intervals marked ‘SP’ may be more completely serpentinized, as has been suggested for the IAM9 line off Iberia (Pickup et al. 1996).
34 B. E. TUCHOLKE ET AL.
Fig. 14. (a) Section of ISE 14 multichannel reflection profile across the SW margin of Galicia Bank. Location in Figure 3 (ISE 14W). The basement ridges at the right are probably continental fault blocks and the ridge on the left is peridotite ridge R3. Arrows mark the Aptian event between seismic sequences 3 and 4. See Figure 4b for sequence ages. Note the exceptionally strong reflectivity of sequence 4 and the Aptian event near the peridotite ridge, which we interpret to be caused by mass wasting of serpentinite debris from the ridge. (b) Enlarged segment of the profile with interpretation. Note the flat-lying nature of sequence 4, the chaotic bedding at the top of the sequence that indicates mass wasting and the unconformable onlap of sediments above the Aptian event.
BREAKUP OF THE NEWFOUNDLAND– IBERIA RIFT 35
36
B. E. TUCHOLKE ET AL.
Fig. 15. Section of ERABLE multichannel reflection profile from the centre of the Newfoundland Basin (Srivastava & Sibuet 1992). Location in Figure 2. See Figure 4b for sequence ages and lithology. The strong reflectivity of the Aptian event (U reflection) is a feature observed throughout most of the Newfoundland Basin (Tucholke et al. 1989). Deeper sedimentary and basement structure is poorly imaged because of the strong reflectivity of U, because of low impedance contrasts, or both. White arrows below U show apparent sedimentary reflections and grey arrows mark dipping, possibly sedimentary reflections that may indicate rotation of fault blocks. Note the unconformable onlap of sediments onto the U reflection.
deep basins. The Aptian event is represented by these deposits combined with debris eroded from basement highs. Stage C, Albian (Fig. 16c). Following the Aptian event, relaxation of stresses reduced longwavelength flexural topography, and sediment flux to the basins was attenuated. Slopes on basement highs also equilibrated, and there no longer was significant mass wasting from these sources. Finegrained turbidity currents spread widely across the basins and lapped unconformably onto the curved Aptian-event reflection in troughs between basement ridges. Deposition was largely conformable on the broader basin plains away from these ridges.
Discussion The relatively prolonged phase of extension that persisted from the Late Jurassic through to the Early Cretaceous appears to have been punctuated by three rifting episodes: one that initiated in the Tithonian and culminated near the Berriasian– Valanginian boundary; a second that peaked near the Hauterivian–Barremian boundary; and a third that is manifested largely in the Aptian. It is noteworthy that the timing of the first two episodes correlates relatively well with the times at which continental crust appears to have separated in the southern and northern parts of the rift. South of Galicia Bank and Flemish Cap (Fig. 4g, sections N1 and I2), extrapolated ages of the edges of
Fig. 16. Schematic representation (not to scale) of interpreted events through the time of the Aptian event. See text for discussion. Grey is continental crust, white is subcontinental mantle lithosphere and black is normal oceanic crust. (a) Latest Aptian emplacement of basement at the plate boundary, primarily by exhumation of subcontinental mantle lithosphere. Elevated in-plane tensile stress affects the entire rift system, and crests of fault blocks are locally uplifted by normal faulting. (b) At the end of Aptian time, introduction of heat and melt from the rising asthenosphere (As) relaxes in-plane tensile stress and results in relative compression of the adjacent plates, causing differential vertical motion (broad arrows) as discussed in the text. Mass wasting from basement highs together with turbidites and debris flows from the uplifted lower continental slope and continental rise creates a highly reflective stratigraphic marker (Aptian event). Relative compression may also cause reverse faulting, locally amplify short-wavelength topography
BREAKUP OF THE NEWFOUNDLAND– IBERIA RIFT
37
(a) LATE APTIAN: MANTLE EXHUMATION AND INTRA-PLATE EXTENSION Sediment input from shelf
Sea level Intra-plate normal faulting and uplift, mass wasting of serpentinites
Br
Rift-wide in-plane tensile stress
le i t t ile ct Du
(b) APTIAN-ALBIAN BOUNDARY: PLATE-BOUNDARY RELEASE, ONSET OF NORMAL SEA-FLOOR SPREADING Probably reduced sediment input from shelf
Sea level Possible topographic amplification and warping of inter-block strata, local reverse faulting, continued mass wasting Aptian event
Sediment input from uplifted lower slope and rise
Plate boundary release Pulse of relative compression
Melt As
(c) ALBIAN: SEA-FLOOR SPREADING Limited sediment input from shelf
Sea level Normal turbidite and pelagic sedimentation Aptian event
Normal sea-floor spreading Melt As
Fig. 16. (Continued) and warp the Aptian-event reflection. (c) Normal sea-floor spreading follows in Albian time. Sea-floor distortions caused by relative compression at the end of the Aptian are levelled out, major turbidity currents and mass wasting cease to be important, and hemipelagic sediments and flat-lying, fine-grained turbidites from distal sources cover the Aptian-event reflection. Rising eustatic sea level probably limited sediment input from the continental shelf.
38
B. E. TUCHOLKE ET AL.
continental crust fall near the Jurassic–Cretaceous boundary and thus within the envelope of the Tithonian–Berriasian rift episode. To the north, the edges of continental crust all are close to the Hauterivian–Barremian boundary (Fig. 4g, sections N2, N3 and I3), correlating with the peak of the second rift episode. If this correlation is valid, it suggests that the rift opened stepwise, south to north, during two principal extensional episodes rather than by continuous northward rift propagation. It is interesting to consider the reasons that these rift episodes may have been relatively short-lived. It has been proposed that once continental crust is thinned to approximately 6–10 km thick, fluids can penetrate to the mantle and facilitate serpentinization, thus weakening the lithosphere and localizing extension on large-scale detachments (Pe´rezGussinye´ & Reston 2001). If this process engenders further fluid penetration, serpentinization and fault weakening, extensional collapse make take place quite rapidly once the initial conditions for fluid access to the mantle are met. We suggest that this process may account for the rift episodes that mark the boundaries between seismic sequences 6, 5 and 4. The second rift episode correlates in time with the beginning of magmatism that formed the SENR–JAR and the conjugate MTR at the southern edge of the rift. This event began near the end of Hauterivian time and extended into the early Aptian (Fig. 4c). The high-amplitude J anomaly in the southern part of the rift and its northward decay (Figs 2 & 3) suggest that magma was introduced northward into the rift to near the latitude of southern Flemish Cap and Galicia Bank. This might explain the presence of the young M-series magnetic anomalies as well as apparently normal oceanic crust at the seaward ends of refraction lines IAM9 off Iberia (Fig. 4g) and SCREECH3 off Newfoundland; it might also account for thin oceanic crust overlying serpentinized mantle between Flemish Cap and Galicia Bank (Fig. 4g, sections N3 and I3). We have suggested, however, that elevated in-plane tensile stress continued throughout the rift (rift episode 3) until normal seafloor spreading began near the Aptian– Albian boundary. If this is correct, then it is likely that magma entering the rift was confined to the shallow lithosphere and that the deeper lithosphere retained its strength. Alternatively, the magmatic event may have temporarily relieved in-plane tensile stress during the Barremian, with high stress levels and mantle exhumation resuming in Aptian time. The clustering of observed extensional events in the later part of the Aptian (Fig. 4c) suggests that this may be the case. Refraction profiles that extend seaward across Aptian and Albian basement are needed in order to investigate these alternatives.
The fact that it is rare to see faulting and rotation in sequence 4/A suggests that Barremian–Aptian extension within the plates was limited and/or localized. However, Wilson et al. (2001) suggested several alternate explanations for why rotated and splayed reflections might not develop above fault blocks during extension; these include: (a) sediments too thin to be resolved seismically; (b) deposition on level, non-rotating fault surfaces; and (c) chaotic re-sedimentation that destroys layering. The first explanation is unlikely because it would require a drastic reduction in sediment supply during deposition of sequence 4/A, for which there is no evidence. Sedimentation rates at ODP Sites 638 and 641 (Fig. 7) were fairly constant or decreased from Valanginian through to Aptian time (Boillot et al. 1987), but rates at DSDP Site 398 (Fig. 9) increased from the Barremian into the Aptian (Sibuet & Ryan 1979). More importantly, sequence 4 clearly has thicknesses comparable to sequence 5 (e.g. Figs 5– 8). Considering the time spans of these sequences (Fig. 4), average sedimentation rates decreased in sequence 4, but only by about 40%. Deposition on a relatively level fault surface, such as might be produced by a rolling-hinge detachment, is a possibility for certain areas such as the part of Hobby High drilled at ODP Sites 900, 1067 and 1068 at the southern margin of Galicia Bank (Manatschal et al. 2001). However, this explanation is not applicable over most of the rest of Galicia Bank where faults have significant dip (e.g. Figs 5, 7 & 8). Chaotic redeposition appears to be a realistic possibility. The clearest evidence for extension during the Barremian–Aptian is confined to areas of exhumed mantle (Fig. 4c). Faulting and uplift of structurally weak serpentinite blocks could produce massive slumps, slides and debris flows that would obscure most structural information about the original faults. This kind of remobilization is suggested by the complex intrabasement structure in many parts of the exposed mantle (Figs 12 & 13), and it is also attested to, perhaps at a smaller scale, by the occurrence of serpentinite olistostromes and breccias on the peridotite ridges drilled at ODP Sites 397 and 399 (Comas et al. 1996; Gibson et al. 1996). As slopes stabilized following faulting and major mass failure at a given location, deposition would shift to more coherent patterns of relatively flat-lying, semi-layered to layered bedding (Figs 6 & 12 –14). This explanation, however, does not apply over most areas of continental crust, where sequence 4 overlies sequence 5 with little or no evidence of major mass wasting, faulting or rotation (Figs 5–9). Although some faults in continental crust probably soled out in serpentinized mantle, the continental
BREAKUP OF THE NEWFOUNDLAND– IBERIA RIFT
faults overall were probably stronger than those in the exhumed mantle. Thus, we suggest that during Barremian–Aptian time there was very limited extension in continental crust and that most intraplate extension occurred within areas of exhumed, serpentinized mantle (Fig. 16a). Why the proposed large amounts of heat and melt were introduced at the plate boundary to initiate normal sea-floor spreading at the end of Aptian time is uncertain, but there are several possibilities. Finite-element models of continental extension by Bowling & Harry (2001) suggest that asthenospheric upwelling and production of melt can accelerate rapidly very late in the history of rifting. They also note that the period of amagmatic rifting leading to this event can be protracted if the extension rate is low and if the lithosphere is homogeneous, or if the lithosphere is heterogeneous and zones of crustal and mantle weakness are offset from one another. In the Newfoundland –Iberia rift, both the heterogeneity of the prerift Palaeozoic crust (Fig. 1) and the observed spatial and temporal shifts in the locus of rifting suggest that the latter may be the case. Reston & Phipps Morgan (2004) proposed that the onset of sea-floor spreading might be triggered by invasion of melt and heat from an adjacent rift segment or mantle plume. In the present instance, this seems unlikely because no increased magmatism is known to have been present in adjacent spreading or rift segments at the end of Aptian time. In addition, the large magmatic event that created the SENR, JAR and MTR in Barremian–early Aptian time seems not to have resulted in normal sea-floor spreading, at least on a permanent basis. A further possibility is that the mantle lithosphere was chemically or thermally heterogeneous, and that sea-floor spreading began with the rise and decompression melting of melt-prone mantle. Such heterogeneity is well documented along the MidAtlantic Ridge (e.g. Bonatti et al. 1992) and the ultra-slow spreading Gakkel Ridge in the Arctic (Michael et al. 2003). However, the observed scale of heterogeneity is generally smaller (,c. 100– c. 200 km) than the length of the Newfoundland– Iberia rift (c. 800 km). Considering these observations, it seems likely that the proposed late-stage onset of normal seafloor spreading probably is an inherent feature of the rift evolution of non-volcanic margins (e.g. Bowling & Harry 2001). We suggest that the arrival of large volumes of melt at the rift axis, and thus the commencement of normal sea-floor spreading, reflects shoaling and decompression melting of the asthenosphere. Rigid, largely subcontinental mantle lithosphere would have been exhumed up until this time (with the possible exception of the aforementioned period of Barremian
39
magmatism), and this would have caused elevated in-plane tensile stress throughout the rift. Sharp reduction or release of the tensile stress (and thus relative compression) associated with introduction of heat and melt from the rising asthenosphere would be expected to produce a well defined event in the geological record. The character of the Aptian event is consistent with this postulated stress change. Relative compression may have amplified local basement topography and stimulated mass wasting, as suggested by chaotic deposits at the top of sequence 4 off Iberia (Figs 8c & 14b). It may also have enhanced the curvature of the Aptian-event reflection and even caused reverse faulting (Fig. 13b). Unfortunately, any reverse faulting that may have developed on the Iberia margin has been obscured by the effects of Cenozoic compression (e.g. Fig. 12) (Boillot et al. 1979; Masson et al. 1994). It is likely that the stress change was accompanied by long-wavelength differential vertical motions along the proximal margins (Fig. 16b). Numerical models indicate that release of in-plane tensile stress would cause uplift of the proximal rift basin (lower continental slope and continental rise), subsidence of the distal rift basin and the outer shelf, and uplift of the inner shelf and coastal plain (e.g. Braun & Beaumont 1989; Kooi & Cloetingh 1992). Details of positions and amplitudes of such vertical motions vary depending on flexural state and rheology of the lithosphere, but in general the models show that the lower slope and rise can be uplifted by several hundred metres. Thus, although most sediments eroded from the hinterland might not pass beyond the depressed outer shelf, the deeper proximal margin could provide a significant source of sediment to more distal basins. Eustacy may also have affected sediment input into the rift. Eustatic sea level was low during the Aptian (Haq et al. 1988), and it could have facilitated shallow-water erosion and sediment transport to the deep basins. By itself, such a eustatic effect is not a likely explanation for the Aptian-event reflection because numerous other eustatic lowstands had no comparable effect on the sedimentary record in the deep basins. Moreover, a eustatic effect would not explain the curvature of the reflection between basement ridges or the unconformable onlap of overlying sediments. The numerical models (e.g. Kooi & Cloetingh 1992) suggest that the long-wavelength vertical distortion of the lithosphere caused by relative compression would have decayed rapidly (Fig. 16c). With the introduction of a new and relatively stable stress regime, regional flux of coarse sediments in debris flows and turbidity currents would have been sharply curtailed. Slope stabilization
40
B. E. TUCHOLKE ET AL.
probably also resulted in a rapid reduction of mass wasting in local basins between basement ridges on the distal margins. Thus, finer-grained sediments were carried in turbidity currents following the Aptian event, and these dilute flows distributed sediments more widely and uniformly. This is consistent with available data from drill sites, which generally show reduced average grain size of sediments deposited during the Albian. These changes account for the sharp drop in reflectivity above sequence 4/A, and for onlap of beds onto the Aptian-event reflection (Figs 5 –8 & 12 –15).
Conclusions The Newfoundland–Iberia rift developed in structurally and lithologically heterogeneous continental crust that had previously been shortened and thickened in Palaeozoic suture zones during the closure of the Iapetus and Rheic oceans. These structures influenced, but did not dominate, the geometry of Mesozoic rifting. Initial rifting in the latest Triassic– earliest Jurassic created large, deep rift basins across a wide region, predominantly in the Grand Banks of Newfoundland and to a lesser extent on the Iberia margin, but this rifting failed to thin the continental crust substantially. A second, prolonged rift phase in the Late Jurassic – Early Cretaceous first thinned and separated the continental crust, and then exhumed subcontinental lithospheric mantle in wide transition zones on the conjugate margins. Emplacement of basement primarily by mantle exhumation appears to have continued until normal sea-floor spreading commenced near the Aptian –Albian boundary. The rift experienced elevated in-plane tensile stress throughout the Late Jurassic –Early Cretaceous rift phase, but most extension of continental crust was concentrated in two episodes: one culminating near the Berriasian –Valanginian boundary and one focused near the Hauterivian –Barremian boundary. In most places these rifting episodes occurred over geologically short intervals, as indicated by faulted sedimentary sequences that are rotated but commonly show only weakly divergent reflections towards fault footwalls. We suggest that this is a manifestation of crustal thinning to the point where fluids could penetrate to and serpentinize the mantle (Pe´rez-Gussinye´ & Reston 2001), which led to rapid collapse of the crustal blocks. The first rift episode primarily affected the Lusitanian and Galicia Interior basins, and it was probably associated with separation of continental crust and initial exhumation of lithospheric mantle in the transition zone south of the Flemish Cap – Galicia Bank conjugate margins. The second rift episode focused in continental crust between Flemish Cap and Galicia Bank, but
to a small extent it also affected the eastern Galicia margin. It also culminated in exhumation of lithospheric mantle and appears to have led to emplacement of thin and/or spatially variable occurrences of magmatic oceanic crust between Flemish Cap and Galicia Bank. Magnetic anomaly analyses and refraction studies along and south of the southern Flemish Cap–Galicia Bank margins suggest that some form of magmatic oceanic crust may have formed there during this rift episode, beginning by about the time of chron M3. However, drilling in this crust near anomaly M1 on both margins has yielded basement consisting of serpentinized peridotite with only minor occurrences of igneous rocks. This suggests that the interpreted oceanic crust seaward of anomaly M3 is not normal oceanic crust but may be largely exhumed mantle that is variably intruded by igneous rocks. Such basement could be difficult to distinguish from normal magmatic oceanic crust in magnetic and refraction data. A possible source of magma during this period is a mantle plume that constructed the SENR, JAR and MTR at the southern edge of the Newfoundland – Iberia rift from about chron M4 (Barremian) to a time somewhat younger than chron M0 (earliest Aptian). The high but northward-decaying amplitude of the J anomaly (c. M1 –M0) suggests that the influence of this plume temporarily extended northwards to near the southern margins of Flemish Cap and Galicia Bank. The ages of the two rift episodes appear to correlate with ages at the edges of continental crust in the southern and northern parts of the rift. This suggests that the rift opened in a stepwise, twophase process rather than by continuous northward propagation of rifting. Existing peridotite basement off Iberia (ODP Sites 897, 899 and 1070, and peridotite ridges R1 and R2) experienced strong normal faulting during a third rift episode in late Aptian time, and possibly throughout both the Barremian and Aptian. We suggest that this was caused by elevated, rift-wide in-plane tensile stress that persisted while subcontinental mantle lithosphere was being exhumed at the plate boundary. The normal faulting probably was focused at the weakest zones in the lithosphere and thus was concentrated in serpentinized, exhumed mantle rather than in adjacent continental crust. Although the high level of in-plane tensile stress may have been temporarily interrupted by the introduction of magma along the plate boundary during Barremian time, we do not consider this to mark the initiation of normal sea-floor spreading. We propose that sea-floor spreading began in the latest Aptian–earliest Albian. At this time the subcontinental mantle lithosphere was finally breached by the rising asthenosphere, and large amounts of
BREAKUP OF THE NEWFOUNDLAND– IBERIA RIFT
41
heat and melt were introduced at the rift axis. This resulted in relaxation of the in-plane tensile stress and thus relative compression throughout the rift. This event is recorded by a prominent, rift-wide sedimentary reflection (Aptian event) that is unconformably onlapped by post-rift sediments. We suggest that the above stress history explains the following characteristics of the Barremian –Aptian sedimentary sequence and the Aptian event.
produced widespread changes in stress state of the rift and is marked by prominent structural and stratigraphic features in the geological record. Thus, interpretation of continental breakup in a nonvolcanic rift must consider the rheology not only of the crust but also that of the subcontinental mantle lithosphere, in addition to spatial and temporal changes in extension that may occur from segment to segment along the rift.
† High reflectivity: flanks of basement ridges (particularly those consisting of serpentinized peridotite) were comparatively unstable up to the end of the Aptian, first because of footwall uplift associated with normal faulting, and then because of the pulse of relative compression associated with stress release at the plate boundary. This resulted in mass wasting, deposition of coarse, thickly bedded debris in basins, and high reflectivity of the Barremian–Aptian sequence. At the time of stress release, there may also have been enhanced input of coarse sediments from the proximal margins because of differential vertical motions there and because of a coincidental eustatic lowstand. † Change to low reflectivity above the Aptian event: sea-floor slopes stabilized in the new, lower-stress environment, thus curtailing mass wasting and turbidity currents. This probably reduced both average grain size and occurrence of thick graded beds, thus leading to lower reflectivity. † Concave-upward configuration of Barremian– Aptian sediments in distal, inter-ridge basins: this was created primarily by mass wasting from adjacent ridge sources, but it may have been enhanced by warping in response to relative plate compression at the end of Aptian time. † Unconformable onlap onto the Aptian-event reflection: following slope stabilization, turbidity currents that entered the basins were finergrained and thus spread more uniformly across basins, depositing flat-lying beds that lapped unconformably onto the Aptian-event reflection.
This research was supported by National Science Foundation grants OCE98-19053 and OCE03-26714 to Woods Hole Oceanographic Institution, grants OCE95-21517 and OCE99-11725 to Rice University, and the Ocean Drilling Program. The Ocean Drilling Program is sponsored by the US National Science Foundation and participating countries under management of Joint Oceanographic Institutions (JOI), Inc. B. Tucholke also acknowledges support from JOI/USSSP grant F001846 and the Henry Bryant Bigelow Chair in Oceanography at Woods Hole Oceanographic Institution. Our research has benefited from discussions with R. B. Whitmarsh and G. Manatschal, and with SCREECH program colleagues W. S. Holbrook, J. R. Hopper, H. Lau, K. E. Louden, D. J. Shillington and H. van Avendonk. We thank S. P. Srivastava for permission to use a segment of the ERABLE multichannel seismic data in our figures. We also thank K. E. Louden, R. C. L. Wilson, D. G. Roberts, G. Karner and L. Pinheiro for careful reviews that helped to improve the manuscript; in some cases their views differ from ours, and the interpretations made here remain our responsibility. This paper was submitted 10 May 2005 and revised 20 April 2006. Contribution No. 11,371 of Woods Hole Oceanographic Institution.
According to our interpretations, the historically identified ‘breakup unconformity’ (here, the Aptian event) in the Newfoundland–Iberia rift does not mark the separation of continental crust and simultaneous initiation of sea-floor spreading. Instead, this event records the final separation of subcontinental mantle lithosphere, the rise of the asthenosphere to shallow levels and emplacement of the first normal oceanic crust. This occurred some 14 million years after the late Hauterivian– early Barremian rupture of continental crust in the northern part of the rift, and even longer after the Tithonian– Berriasian separation of continental crust in the southern part of the rift. Each of these three events
References A LVES , T. M., M ANUPPELLA , G., G AWTHORPE , R. L., H UNT , D. W. & M ONTEIRO , J. H. 2003. The depositional evolution of diapir- and fault-bounded rift basins: Examples from the Lusitanian Basin of West Iberia. Sedimentary Geology, 162, 273– 303. A USTIN , J. A. J R ., T UCHOLKE , B. E. & U CHUPI , E. 1989. Upper Triassic–Lower Jurassic salt basin southeast of the Grand Banks. Earth and Planetary Science Letters, 92, 357– 370. B ALKWILL , H. R. & L EGALL , F. D. 1989. Whale Basin, Offshore Newfoundland: Extension and salt diapirism. In: T ANKARD , A. J. & B ALKWILL , H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins. AAPG Memoirs, 46, 233– 245. B EARD , J. S., F ULLAGAR , P. D. & S INHA , A. K. 2002. Gabbroic pegmatite intrusions, Iberia Abyssal Plain, ODP Leg 173, Site 1070: Magmatism during a transition from non-volcanic rifting to sea-floor spreading. Journal of Petrology, 43, 885– 905. B ESLIER , M.-O. 1996. Data Report. Seismic line LG12 in the Iberia Abyssal Plain. In: W HITMARSH , R. B., S AWYER , D. S., K LAUS , A. & M ASSON , D. G. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 149. Ocean Drilling Program, College Station, TX, 737– 739.
42
B. E. TUCHOLKE ET AL.
B OILLOT , G. & W INTERER , E. L. 1988. Drilling on the Galicia margin: Retrospect and prospect. In: B OILLOT , G., W INTERER , E. L. ET AL . Proceedings of the Ocean Drilling Program, Scientific Results, 103. Ocean Drilling Program, College Station, TX, 809– 828. B OILLOT , G., A UXIETRE , J.-L., D UNAND , J.-P., D UPEUBLE , P.-A. & M AUFFRET , A. 1979. The northwestern Iberian margin: A Cretaceous passive margin deformed during Eocene. In: T ALWANI , M., H AY , W. & R YAN , W. B. F. (eds) Deep Drilling Results in the Atlantic Ocean: Continental Margins and Paleoenvironment, Maurice Ewing Series, 3. American Geophysical Union, Washington, DC, 138– 153. B OILLOT , G., W INTERER , E. L. ET AL . 1987. Proceedings of the Ocean Drilling Program, Initial Reports, 103. Ocean Drilling Program, College Station, TX. B OILLOT , G., W INTERER , E. L. ET AL . 1988. Proceedings of the Ocean Drilling Program, Scientific Results, 103. Ocean Drilling Program, College Station, TX. B OILLOT , G., F ERAUD , G., R ECQ , M. & G IRARDEAU , J. 1989a. Undercrusting by serpentinite beneath rifted margins. Nature, 341, 523– 525. B OILLOT , G., G IRARDEAU , J. & K ORNPROBST , J. 1989b. Rifting of the west Galicia continental margin: A review. Bulletin de la Socie´te´ Ge´ologique de France, 8, 393– 400. B OILLOT , G., M OUGENOT , D., G IRARDEAU , J. & W INTERER , E. L. 1989c. Rifting processes on the west Galicia margin, Spain. In: T ANKARD , A. J. & B ALKWILL , H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins. AAPG Memoirs, 46, 363– 377. B ONATTI , E., P EYVE , A., K EPEZHINSKAS , P., K URENTSOVA , N., S EYLER , M., S KOLOTNEV , S. & U DINTSEV , G. 1992. Upper mantle heterogeneity below the Mid-Atlantic Ridge, 08– 158N. Journal of Geophysical Research, 97, 4461– 4476. B OWLING , J. C. & H ARRY , D. L. 2001. Geodynamic models of continental extension and the formation of non-volcanic rifted continental margins. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: A Comparison of Evidence From Land and Sea. Geological Society, London, Special Publications, 187, 511– 536. B RAUN , J. & B EAUMONT , C. 1989. A physical explanation of the relation between flank uplifts and the breakup unconformity at rifted continental margins. Geology, 17, 760–764. C APDEVILA , R. & M OUGENOT , D. 1988. Pre-Mesozoic basement of the western Iberian continental margin and its place in the Variscan Belt. In: B OILLOT , G., W INTERER , E. L. ET AL . Proceedings of the Ocean Drilling Program, Scientific Results, 103, Ocean Drilling Program, College Station, TX, 3 –12. C HANNELL , J. E. T., E RBA , E., N AKANISHI , M. & T AMAKI , K. 1995. Late Jurassic– Early Cretaceous time scales and oceanic magnetic anomaly block models. In: B ERGGREN , W. A., K ENT , D. V., A UBRY , M.-P. & H ARDENBOL , J. (eds) Geochronology, Time Scales and Global Stratigraphic Correlation. SEPM Special Publications, 54, 51– 63.
C HIAN , D., L OUDEN , K. E., M INSHULL , T. A. & W HITMARSH , R. B. 1999. Deep structure of the ocean–continent transition in the southern Iberia Abyssal Plain from seismic refraction profiles: Ocean Drilling Program (Legs 149 and 173) transect. Journal of Geophysical Research, 104, 7443–7462. C HRISTENSEN , N. I. 2004. Serpentinites, peridotites, and seismology. International Geology Review, 46, 795–816. C OMAS , M. C., S A´ NCHEZ -G O´ MEZ , M., C ORNEN , G. & DE K AENEL , E. 1996. Serpentinized peridotite breccia and olistostrome on basement highs of the Iberia Abyssal Plain: Implications for tectonic margin evolution. In: W HITMARSH , R. B., S AWYER , D. S., K LAUS , A. & M ASSON , D. G. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 149. Ocean Drilling Program, College Station, TX, 577– 591. DE V OOGD , B. & K EEN , C. E. 1987. Lithoprobe East: Results from reflection profiling of the continental margin; Grand Banks region. Geophysical Journal of the Royal Astronomical Society, 89, 195–200. D EAN , S. M., M INSHULL , T. A., W HITMARSH , R. B. & L OUDEN , K. E. 2000. Deep structure of the ocean– continental transition in the southern Iberia Abyssal Plain from seismic refraction profiles: The IAM-9 transect at 408200 N. Journal of Geophysical Research, 105, 5859–5885. DISCOVERY 215 WORKING GROUP . 1998. Deep structure in the vicinity of the ocean–continent transition zone under the southern Iberia Abyssal Plain. Geology, 26, 743–746. D RISCOLL , N. W., H OGG , J. R., C HRISTIE -B LICK , N. & K ARNER , G. D. 1995. Extensional tectonics in the Jeanne d’Arc Basin, offshore Newfoundland: Implications for the timing of breakup between Grand Banks and Iberia. In: S CRUTTON , R. A., S TOKER , M. S., S HIMMIELD , G. B. & T UDHOPE , A. W. (eds) The Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region. Geological Society, London, Special Publications, 90, 1– 28. E NACHESCU , M. E. 1988. Extended basement beneath the intracratonic rifted basins of the Grand Banks of Newfoundland. Canadian Journal of Exploration Geophysics, 24, 48– 65. EXPEDITION 309 SCIENTISTS . 2005. Superfast spreading rate crust 2: A complete in situ section of upper oceanic crust formed at a superfast spreading rate. IODP Preliminary Report, 309, doi:10.2204/ iodp.pr.309.2005. Integrated Ocean Drilling Program, College Station, TX. F ALVEY , D. A. 1974. The development of continental margins in plate tectonic theory. APEA Journal, 14, 95–106. F E´ RAUD , G., B ESLIER , M.-O. & C ORNEN , G. 1996. 40 Ar/39Ar dating of gabbros from the ocean/continent transition of the western Iberia margin: Preliminary results. In: W HITMARSH , R. B., S AWYER , D. S., K LAUS , A. & M ASSON , D. G. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 149. Ocean Drilling Program, College Station, TX, 489–495. F E´ RAUD , G., G IRARDEAU , J., B ESLIER , M.-O. & B OILLOT , G. 1988. Datation 40Ar/39Ar de la mise
BREAKUP OF THE NEWFOUNDLAND– IBERIA RIFT en place des pe´ridotites bordant la marge de la Galice (Espagne). Comptes Rendus de l’Academie des Sciences, Serie 2, Mecanique, Physique, Chimie, Sciences de l’Univers, Sciences de la Terre, 307, 49– 55. F UEGENSCHUH , B., F ROITZHEIM , N. & B OILLOT , G. 1998. Cooling history of granulite samples from the ocean– continent transition of the Galicia margin: Implications for rifting. Terra Nova, 10, 96–100. F UNCK , T. J., H OPPER , J. R., L ARSEN , H. C., L OUDEN , K. E., T UCHOLKE , B. E. & H OLBROOK , W. S. 2003. Crustal structure of the ocean–continent transition at Flemish Cap: Seismic refraction results. Journal of Geophysical Research, 108, doi:10.1029/ 2003JB002434. G IBSON , I. L., M ILLIKEN , K. L. & M ORGAN , J. K. 1996. Serpentinite– breccia landslide deposits generated during crustal extension at the Iberia margin. In: W HITMARSH , R. B., S AWYER , D. S., K LAUS , A. & M ASSON , D. G. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 149. Ocean Drilling Program, College Station, TX, 571–575. G RADSTEIN , F. M., A GTERBERG , F. P., O GG , J. G., H ARDENBOL , J., VAN V EEN , P., T HIERRY , J. & H UANG , Z. 1995. A Triassic, Jurassic and Cretaceous time scale. In: B ERGGREN , W. A., K ENT , D. V., A UBRY , M.-P. & H ARDENBOL , J. (eds) Geochronology, Time Scales and Global Stratigraphic Correlation. SEPM Special Publications, 54, 95–126. G RANT , A. C., J ANSA , L. F., M C A LPINE , K. D. & E DWARDS , A. 1988. Mesozoic–Cenozoic geology of the eastern margin of the Grand Banks and its relation to Galicia Bank. In: B OILLOT , G., W INTERER , E. L. ET AL . Proceedings of the Ocean Drilling Program, Scientific Results, 103. Ocean Drilling Program, College Station, TX, 787– 808. GROUPE GALICE . 1979. The continental margin off Galicia and Portugal: Acoustical stratigraphy, dredge stratigraphy, and structural evolution. In: S IBUET , J.-C., R YAN , W. B. F. ET AL . Initial Reports of the Deep Sea Drilling Project, 47, Part II. US Government Printing Office, Washington, DC, 633 –662. H AQ , B. U., H ARDENBOL , J. & V AIL , P. R. 1988. Mesozoic and Cenozoic chronostratigraphy and cycles of sea-level change. In: W ILGUS , C. K., H ASTINGS , B. S., R OSS , C. A., P OSAMENTIER , H. W., V AN W AGONER , J. & K ENDALL , C. (eds) Sea-level Changes; An Integrated Approach. SEPM Special Publications, 42, 72– 108. H ART , S. R. & B LUSZTAJN , J. 2006. Age and geochemistry of the mafic sills, ODP Site 1276, Newfoundland margin. Chemical Geology, 235, 222–237. H ENNING , A. T., S AWYER , D. S. & T EMPLETON , D. C. 2004. Exhumed upper mantle within the ocean– continent transition on the northern West Iberia margin: Evidence from prestack depth migration and total tectonic subsidence analyses. Journal of Geophysical Research, 109, B05103, doi:05110.01029/ 02003JB.002526. H OPPER , J. R., F UNCK , T. ET AL . 2004. Continental breakup and the onset of ultra-slow seafloor spreading off Flemish Cap on the Newfoundland rifted margin. Geology, 32, 93–96.
43
J ANSA , L. F. & P E -P IPER , G. 1988. Middle Jurassic to Early Cretaceous igneous rocks along eastern North American continental margin. AAPG Bulletin, 72, 347– 366. J ANSA , L. F. & W ADE , J. A. 1975. Geology of the continental margin off Nova Scotia and Newfoundland. In: VAN DER L INDEN , W. J. M. & W ADE , J. A. (eds) Offshore Geology of Eastern Canada, Volume 2, Regional Geology. Geological Survey of Canada Paper, 74–30, 51–105. K ARNER , G. D. & S HILLINGTON , D. J. 2005. Basalt sills of the U reflector, Newfoundland Basin: A serendipitous dating technique. Geology, 33, 985–988. K EEN , C. E. & DE V OOGD , B. 1988. The continent – ocean boundary at the rifted margin off eastern Canada: New results from deep seismic reflection studies. Tectonics, 7, 107–124. K EEN , C. E., L ONCAREVIC , B. D., R EID , I., W OODSIDE , J., H AWORTH , R. T. & W ILLIAMS , H. 1990. Tectonic and geophysical overview. In: K EEN , M. J. & W ILLIAMS , G. L. (eds) Geology of Canada. Geological Society of America, Boulder, CO, 32– 74. K OOI , H. & C LOETINGH , S. 1992. Lithospheric necking and regional isostasy at extensional basins 2. Stress-induced vertical motions and relative sea level changes. Journal of Geophysical Research, 97, 17573–17591. L AU , K. W. H., L OUDEN , K. E. ET AL . 2006a. Crustal structure across the Grand Banks–Newfoundland Basin continental margin (Part II) – Results from a seismic reflection profile. Geophysical Journal International, 167, 157 –170. L AU , K. W. H., L OUDEN , K. E., F UNCK , T., T UCHOLKE , B. E., H OLBROOK , W. S., H OPPER , J. R. & L ARSEN , H. C. 2006b. Crustal structure across the Grand Banks–Newfoundland Basin continental margin (Part I) – Results from a seismic refraction profile. Geophysical Journal International, 167, 127– 156. M ANATSCHAL , G. & B ERNOULLI , D. 1999. Architecture and tectonic evolution of nonvolcanic margins: Present-day Galicia and ancient Adria. Tectonics, 18, 1099– 1119. M ANATSCHAL , G., F ROITZHEIM , N., R UBENACH , M. & T URRIN , B. D. 2001. The role of detachment faulting in the formation of an ocean– continent transition: Insights from the Iberia Abyssal Plain. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: A Comparison of Evidence from Land and Sea. Geological Society, London, Special Publications, 187, 405– 428. M ASSON , D. G., C ARTWRIGHT , J. A., P INHEIRO , L. M., W HITMARSH , R. B., B ESLIER , M. & R OESER , H. 1994. Compressional deformation at the ocean– continent transition in the NE Atlantic. Journal of Geological Society London, 151, 607– 613. M AUFFRET , A. & M ONTADERT , L. 1987. Rift tectonics on the passive continental margin off Galicia (Spain). Marine and Petroleum Geology, 4, 49– 70. M AUFFRET , A. & M ONTADERT , L. 1988. Seismic stratigraphy off Galicia. In: B OILLOT , G., W INTERER , E. L. ET AL . Proceedings of the Ocean Drilling Program, Scientific Results, 103, Ocean Drilling Program, College Station, TX, 13–30.
44
B. E. TUCHOLKE ET AL.
M AUFFRET , A., M OUGENOT , D., M ILES , P. R. & M ALOD , J. A. 1989. Results from multichannel reflection profiling of the Tagus abyssal plain (Portugal) – comparison with the Canadian margin. In: T ANKARD , A. J. & B ALKWILL , H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins. AAPG Memoirs, 46, 379 –393. M ICHAEL , P. J., L ANGMUIR , C. H. ET AL . 2003. Magmatic and amagmatic sea-floor generation at the ultraslow-spreading Gakkel ridge, Arctic Ocean. Nature, 423, 956 –961. M ILES , P. R., V ERHOEF , J. & M ACNAB , R. 1996. Compilation of magnetic anomaly chart west of Iberia. In: W HITMARSH , R. B., S AWYER , D. S., K LAUS , A. & M ASSON , D. G. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 149. Ocean Drilling Program, College Station, TX, 659– 663. M ONTENAT , C., G UERY , F., J AMET , M. & B ERTHOU , P. Y. 1988. Mesozoic evolution of the Lusitanian Basin: Comparison with the adjacent margin. In: B OILLOT , G., W INTERER , E. L. ET AL . Proceedings of the Ocean Drilling Program, Scientific Results, 103. Ocean Drilling Program, College Station, TX, 757–775. M URILLAS , J., M OUGENOT , D., B OILLOT , G., C OMAS , M. C., B ANDA , E. & M AUFFRET , A. 1990. Structure and evolution of the Galicia Interior Basin (Atlantic western Iberian continental margin). Tectonophysics, 184, 297–319. PENROSE CONFERENCE PARTICIPANTS . 1972. Penrose field conference on ophiolites. Geotimes, 17, 24–25. P ERCIVAL , J. A., B LEEKER , W., C OOK , G. A., R IVERS , T., R OSS , G. & VAN S TAAL , C. 2004. PanLITHOPROBE Workshop IV: Intra-orogen correlations and comparative orogenic anatomy. Geoscience Canada, 31, 23– 39. P E´ REZ -G USSINYE´ , M., R ANERO , C. R., R ESTON , T. J. & S AWYER , D. 2003. Mechanisms of extension at nonvolcanic margins: Evidence from the Galicia Interior Basin, west of Iberia. Journal of Geophysical Research, 108, EPM 6-1– EPM 6-19, doi:10.1029/ 2001JB000901. P E´ REZ -G USSINYE´ , M. & R ESTON , T. J. 2001. Rheological evolution during extension at nonvolcanic rifted margins: Onset of serpentinization and development of detachments leading to continental breakup. Journal of Geophysical Research, 106, 3961–3975. P ICKUP , S. L. B., W HITMARSH , R. B., F OWLER , C. M. R. & R ESTON , T. J. 1996. Insight into the nature of the ocean– continent transition off West Iberia from a deep multichannel seismic reflection profile. Geology, 24, 1079– 1082. P INHEIRO , L. M., W HITMARSH , R. B. & M ILES , P. R. 1992. The ocean–continent boundary off the western continental margin of Iberia – II. Crustal structure in the Tagus Abyssal Plain. Geophysical Journal International, 109, 106–124. P INHEIRO , L. M., W ILSON , R. C. L., P ENA DOS R EIS , R., W HITMARSH , R. B. & R IBEIRO , A. 1996. The western Iberia margin: A geophysical and geological overview. In: W HITMARSH , R. B., S AWYER , D. S., K LAUS , A. & M ASSON , D. G. (eds) Proceedings of the Ocean
Drilling Program, Scientific Results, 149. Ocean Drilling Program, College Station, TX, 1– 23. R ABINOWITZ , P. D., C ANDE , S. C. & H AYES , D. E. 1978. Grand Banks and J-Anomaly Ridge. Science, 202, 71–73. R ASMUSSEN , E. S., L OMHOLT , S., A NDERSEN , C. & V EJBÆK , O. V. 1998. Aspects of the structural evolution of the Lusitanian Basin in Portugal and the shelf and slope area offshore Portugal. Tectonophysics, 300, 199– 225. R EID , I. 1988. Crustal structure beneath the southern Grand Banks: Seismic-refraction results and their implications. Canadian Journal of Earth Sciences, 25, 760–772. R EID , I. 1994. Crustal structure of a nonvolcanic rifted margin east of Newfoundland. Journal of Geophysical Research, 99, 15161–15180. R EID , I. D. & K EEN , C. E. 1990. Deep crustal structure beneath a rifted basin: Results from seismic refraction measurements across the Jeanne d’Arc Basin, offshore eastern Canada. Canadian Journal of Earth Sciences, 27, 1462– 1471. R ESTON , T. J. & P HIPPS M ORGAN , J. 2004. Continental geotherm and the evolution of rifted margins. Geology, 32, 133– 136. R ESTON , T. J., K RAWCZYK , C. M. & H OFFMANN , H.-J. 1995. Detachment tectonics during Atlantic rifting: analysis and interpretation of the S reflection, the west Galicia margin. In: S CRUTTON , R. A., S TOKER , M. S., S HIMMIELD , G. B. & T UDHOPE , A. W. (eds) The Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region. Geological Society, London, Special Publications, 90, 93–109. R IBEIRO , A., Q UESADA , C. & D ALLMEYER , R. D. 1990. Geodynamic evolution of the Iberian Massif. In: D ALLMEYER , R. D. & M ARTINEZ G ARCIA , E. (eds) Pre-Mesozoic Geology of the Iberian Peninsula. Springer, Berlin, 399–409. R UBENACH , M. 1999. Microstructures, metamorphism, and tectonic significance of metabasic rocks, ocean– continent transition, Iberia Abyssal Plain. The Specialist Group in Tectonics and Structural Geology Field Conference, Halls Gap, Victoria, Australia, February 14–19, 1999, Abstract Volume, JESSELL , M. (ed.) Geological Society of Australia, Abstract Series, 53, 231–232. R USSELL , S. M. & W HITMARSH , R. B. 2003. Magmatism at the west Iberia non-volcanic rifted continental margin: Evidence from analyses of magnetic anomalies. Geophysical Journal International, 154, 706–730. S CHA¨ RER , U., G IRARDEAU , J., C ORNEN , G. & B OILLOT , G. 2000. 138–121 Ma asthenospheric magmatism prior to continental breakup in the North Atlantic and geodynamic implications. Earth and Planetary Science Letters, 181, 555–572. S CHA¨ RER , U., K ORNPROBST , J., B ESLIER , M.-O., B OILLOT , G. & G IRARDEAU , J. 1995. Gabbro and related rock emplacement beneath rifting continental crust: U–Pb geochronological and geochemical constraints for the Galicia passive margin (Spain). Earth and Planetary Science Letters, 130, 187– 200.
BREAKUP OF THE NEWFOUNDLAND– IBERIA RIFT S HILLINGTON , D. J., H OLBROOK , W. S., T UCHOLKE , B. E., H OPPER , J. R., L OUDEN , K. E., L ARSEN , H. C. & VAN A VENDONK , H. 2004. Data report: Marine geophysical data on the Newfoundland nonvolcanic rifted margin along SCREECH Transect 2. In: T UCHOLKE , B. E., S IBUET , J.-C., K LAUS , A. ET AL . Proceedings of the Ocean Drilling Program, Initial Reports, 210 [CD-ROM]. Ocean Drilling Program, College Station, TX, 1–36. SHIPBOARD SCIENTIFIC PARTY . 2004. Site 1277. In: T UCHOLKE , B. E., S IBUET , J.-C., K LAUS , A. ET AL . Proceedings of the Ocean Drilling Program, Initial Reports, 210 [CD-ROM]. Ocean Drilling Program, College Station, TX, 1–39. S IBUET , J.-C. & R YAN , W. B. F. 1979. Site 398: Evolution of the west Iberian passive continental margin in the framework of the early evolution of the North Atlantic Ocean. In: S IBUET , J.-C., R YAN , W. B. F. & K LAUS , A. Initial Reports of the Deep Sea Drilling Project, 47, Part II. US Government Printing Office, Washington, DC, 761–775. S IBUET , J.-C., M ONTI , S., L OUBRIEU , B., M AZE´ , J.-P. & S RIVASTAVA , S. 2004a. Carte bathyme´trique de l’Atlantique nord-est et du golfe de Gascogne: Implications cine´matiques. Bulletin de la Socie´te´ Ge´ologique de France, 175, 429–442. S IBUET , J.-C., S RIVASTAVA , S. P. & S PAKMAN , W. 2004b. Pyrenean orogeny and plate kinematics. Journal of Geophysical Research, 109, B08104, doi:10.1029/2003JB002514. S ILVA , E. A., M IRANDA , J. M., L UIS , J. F. & G ALDEANO , A. 2000. Correlation between the Palezoic structures from West Iberian and Grand Banks margins using inversion of magnetic anomalies. Tectonophysics, 321, 57–71. S INCLAIR , I. K. 1988. Evolution of Mesozoic–Cenozoic sedimentary basins in the Grand Banks area of Newfoundland and comparison with Falvey’s (1974) rift model. Bulletin of Canadian Petroleum Geology, 36, 255–273. S MITH , W. H. F. & S ANDWELL , D. T. 1997. Global sea floor topography from satellite altimetry and ship depth soundings. Science, 277, 1956–1962. S RIVASTAVA , S. P. & S IBUET , J.-C. 1992. A Joint AGC and IFREMER Geophysical Cruise to the Newfoundland and Orphan Basins. Cruise Report, CSS Hudson 92–22, Mission ERABLE. Geological Survey of Canada, Dartmouth, Nova Scotia, Canada, and Institut Francais de Recherche pour L’Exploitation de Mer, Centre de Brest, France. S RIVASTAVA , S. P., S IBUET , J.-C., C ANDE , S., R OEST , W. R. & R EID , I. R. 2000. Magnetic evidence for slow seafloor spreading during the formation of the Newfoundland and Iberian margins. Earth and Planetary Science Letters, 182, 61–76. S ULLIVAN , K. D. & K EEN , C. E. 1977. Newfoundland Seamounts: Petrology and geochemistry. In: B ARAGAR , W. R. A., C OLEMAN , L. C. & H ALL , J. M. (eds) Volcanic Regimes in Canada. Geological Association of Canada Special Paper, 16, 461–476. T ANKARD , A. J. & W ELSINK , H. J. 1987. Extensional tectonics and stratigraphy of Hibernia oil field, Grand Banks, Newfoundland. AAPG Bulletin, 71, 1210–1232.
45
T ANKARD , A. J. & W ELSINK , H. J. 1988. Extensional tectonics, structural styles and stratigraphy of the Mesozoic Grand Banks of Newfoundland. In: M ANSPEIZER , W. (ed) Triassic– Jurassic Rifting and the Opening of the Atlantic Ocean. Elsevier, Amsterdam, 129 –165. T ODD , B. J. & R EID , I. 1989. The continent – ocean boundary south of Flemish Cap: Constraints from seismic refraction and gravity. Canadian Journal of Earth Sciences, 26, 1392–1407. T UCHOLKE , B. E., V OGT , P. R. ET AL . 1979. Site 384: The Cretaceous– Tertiary boundary, Aptian reefs, and the J-Anomaly Ridge. In: T UCHOLKE , B. E., V OGT , P. R. ET AL . Initial Reports of the Deep Sea Drilling Project, 43. US Government Printing Office, Washington, DC, 107– 154. T UCHOLKE , B. E. & L UDWIG , W. J. 1982. Structure and origin of the J Anomaly Ridge, western North Atlantic Ocean. Journal of Geophysical Research, 87, 9389– 9407. T UCHOLKE , B. E., A USTIN , J. A. & U CHUPI , E. 1989. Crustal structure and rift-drift evolution of the Newfoundland basin. In: T ANKARD , A. J. & B ALKWILL , H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins. AAPG Memoirs, 46, 247–263. T UCHOLKE , B. E., S IBUET , J.-C., K LAUS , A. ET AL . 2004. Proceedings of the Ocean Drilling Program, Initial Reports, 210 [CD-ROM]. Ocean Drilling Program, College Station, TX. V AN A VENDONK , H. J. A., H OLBROOK , W. S. ET AL . 2006. Seismic velocity structure of the rifted margin of the eastern Grand Banks of Newfoundland, Canada. Journal of Geophysical Research, 111, B11404, doi:10.1029/2005JB004156. VAN S TAAL , C. R., D EWEY , J. F., M AC N IOCAILL , C. & M C K ERROW , S. 1998. The Cambrian –Silurian tectonic evolution of the northern Appalachians: History of a complex, southwest Pacific-type segment of Iapetus. In: B LUNDELL , D. J. & S COTT , A. C. (eds) Lyell: The Past is the Key to the Present. Geological Society, London, Special Publications, 143, 199–242. W ALDRON , J. W. F. & VAN S TAAL , C. R. 2001. Taconic Orogeny and the accretion of the Dashwoods block: A peri-Laurentian microcontinent in the Iapetus Ocean. Geology, 29, 811– 814. W ELSINK , H. J., S RIVASTAVA , S. P. & T ANKARD , A. J. 1989. Basin architecture of the Newfoundland continental margin and its relationship to ocean crust fabric during extension. In: T ANKARD , A. J. & B ALKWILL , H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins. AAPG Memoirs, 46, 197– 213. W HITMARSH , R. B. & M ILES , P. R. 1995. Models of the development of the West Iberia rifted continental margin at 408300 N deduced from surface and deep-tow magnetic anomalies. Journal of Geophysical Research, 100, 3789–3806. W HITMARSH , R. B. & S AWYER , D. S. 1996. The ocean/ continent transition beneath the Iberia Abyssal Plain and continental-rifting to sea-floor-spreading processes. In: W HITMARSH , R. B., S AWYER , D. S., K LAUS , A. & M ASSON , D. G. (eds) Proceedings of
46
B. E. TUCHOLKE ET AL.
the Ocean Drilling Program, Scientific Results, 149. Ocean Drilling Program, College Station, TX, 713– 733. W HITMARSH , R. B. & W ALLACE , P. J. 2001. The rift-to-drift development of the west Iberia nonvolcanic continental margin: A summary and review of the contribution of Ocean Drilling Program Leg 173. In: B ESLIER , M.-O., W HITMARSH , R. B., W ALLACE , P. J. & G IRARDEAU , J. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 173. Ocean Drilling Program, College Station, TX, 1–36. W HITMARSH , R. B., B ESLIER , M.-O. ET AL . 1998. Proceedings of the Ocean Drilling Program, Initial Reports, 173. Ocean Drilling Program, College Station, TX. W HITMARSH , R. B., M ANATSCHAL , G. & M INSHULL , T. A. 2001. Evolution of magma-poor continental margins from rifting to sea-floor spreading. Nature, 413, 150–154. W HITMARSH , R. B., M ILES , P. R. & M AUFFRET , A. 1990. The ocean–continent boundary off the western continental margin of Iberia – I. Crustal structure at 408300 N. Geophysical Journal International, 103, 509– 531. W HITMARSH , R. B., W HITE , R. S., H ORSEFIELD , S. J., S IBUET , J.-C., R ECQ , M. & L OUVEL , V. 1996. The ocean– continent boundary off the western continental margin of Iberia: Crustal structure west of Galicia Bank. Journal of Geophysical Research, 101, 28,291– 28,314. W ILLIAMS , H. & H ATCHER , R. D. J R . 1982. Suspect terranes and accretionary history of the Appalachian Orogen. Geology, 10, 530– 536. W ILLIAMS , H. & H ATCHER , R. D. J R . 1983. Appalachian suspect terranes. In: H ATCHER , R. D. JR , W ILLIAMS , H. & Z IETZ , I. (eds) Contributions to the Tectonics and Geophysics of Mountain Chains. Geological
Society of America Memoir, 159. Geological Society of America, Boulder, CO, 33– 53. W ILSON , R. C. L. 1988. Mesozoic development of the Lusitanian Basin, Portugal. Revista de la Sociedad Ge´olo´gica de Espana, 1, 393–407. W ILSON , R. C. L., H ISCOTT , R. N., W ILLIS , M. G. & G RADSTEIN , F. M. 1989. The Lusitanian Basin of west-central Portugal: Mesozoic and Tertiary tectonic, stratigraphic, and subsidence history. In: T ANKARD , A. J. & B ALKWILL , H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins. AAPG Memoirs, 46, 265– 282. W ILSON , R. C. L., M ANATSCHAL , G. & W ISE , S. 2001. Rifting along non-volcanic passive margins: Stratigraphic and seismic evidence from the Mesozoic successions of the Alps and western Iberia. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: A Comparison of Evidence From Land and Sea. Geological Society, London, Special Publications, 187, 429– 452. W INTERER , E. L., G EE , J. S. & W AASBERGEN , R. J. V. 1988. The source area for Lower Cretaceous clastic sediments of the Galicia margin: Geology and tectonic and erosional history. In: B OILLOT , G., W INTERER , E. L. ET AL . Proceedings of the Ocean Drilling Program, Scientific Results, 103. Ocean Drilling Program, College Station, TX, 697–732. Z HAO , X. 1996. Magnetic signatures of peridotite rocks from Sites 897 and 899 and their implications. In: W HITMARSH , R. B., S AWYER , D. S., K LAUS , A. & M ASSON , D. G. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 149. Ocean Drilling Program, College Station, TX, 431–446. Z IEGLER , P. A. 1982. Geological Atlas of Western and Central Europe. Elsevier, Amsterdam.
Structure of the Flemish Cap margin, Newfoundland: insights into mantle and crustal processes during continental breakup J. R. HOPPER1, T. FUNCK2 & B. E. TUCHOLKE3 1
Department of Geology and Geophysics, Texas A&M University, College Station, TX 77843, USA (e-mail:
[email protected]) 2
Geological Survey of Denmark and Greenland, Øster Voldgade 10, DK-1350 Copenhagen K, Denmark 3
Department of Geology and Geophysics, Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA
Abstract: Seismic reflection and refraction data from the Flemish Cap margin off Newfoundland reveal the large-scale structure of a magma-starved rifted margin. There is little evidence for significant extensional deformation of the Flemish Cap, consistent with the hypothesis that it behaved as a microplate throughout the Mesozoic. The seismic data highlight important asymmetries at a variety of scales that developed during the final stages of continental breakup and the onset of oceanic sea-floor spreading. In strong contrast to the conjugate Galicia Bank margin, Flemish Cap shows: (1) an abrupt necking profile in continental crust, thinning from 30 km thick to 3 km thick over a distance of 80 km, and a narrow, less than 20 km-wide, zone of extremely thin continental crust; (2) no clear evidence for horizontal detachment structures beneath continental crust similar to the ‘S’ reflection; and (3) evidence for at least a 60 km-wide zone of anomalously thin oceanic crust that began accreting to the margin shortly after continental crustal separation. The oceanic crust averages only 3 –4 km thick and in places is as thin as 1.3 km thick, although seismic layer 3 is missing where this occurs. The data suggest that there are large spatial and temporal variations in the available melt supply following continental breakup as oceanic sea-floor spreading becomes established. In addition, wide-angle data show that anomalously slow mantle P-wave velocities appear approximately where continental crust has thinned to 6– 8 km thick, indicating that low-degree serpentinization begins where the entire crust has become embrittled.
Continental extension leading to lithospheric rupture and the creation of new oceanic sea-floor spreading systems has long been a cornerstone of global plate tectonics. Because rifted margins contain the geological record of key processes involved with extension and breakup, investigations of these margins are central to advancing our understanding of lithospheric rupture. Thinning of the crust and lithospheric mantle is the most obvious manifestation of continental extension. Two key consequences of thinning are brittle failure of the upper crust and decompression melting of asthenospheric mantle as it nears the surface. Despite the seeming simplicity of these two effects, considerable controversy over basic aspects of how they work has resulted from enigmatic observations from a number of areas. For the brittle behaviour of the crust, the possible role of low-angle normal faulting and subhorizontal detachments during extension is still debated (e.g. Lavier et al. 1999; Lavier & Manatschal 2006). In the case of mantle melting, there is no firm consensus on the primary causes of the extreme range of magmatic productivity during the initial stages of
sea-floor spreading (e.g. Boutilier & Keen 1999; Nielsen & Hopper 2004). The rifted passive margins around the North Atlantic are a unique natural laboratory where many of these questions can be studied. Magnetic anomalies permit the reconstruction of conjugate margin pairs with a reasonable degree of certainty (Srivastava & Tapscott 1986). This is critical as many rift systems exhibit asymmetry that can only be quantified by characterizing the entire rift system. In addition, the North Atlantic margins are themselves extremely diverse. Figure 1 shows the regional setting with measured oceanic crustal thicknesses marked along several of the margins. At one end are the margins along Greenland, Norway and the British Isles where 15 –30 kmthick oceanic crust accreted to the rifted continental margin (Holbrook et al. 2001). At the other extreme are the margins of SW Greenland and Iberia, where 0 –5 km-thick oceanic crust is observed (Chian & Louden 1994; Minshull et al. 2001). The lack of volcanism along these latter margins has proven particularly useful for studying
From: KARNER , G. D., MANATSCHAL , G. & PINHEIRO , L. M. (eds) Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, London, Special Publications, 282, 47–61. DOI: 10.1144/SP282.3 0305-8719/07/$15.00 # The Geological Society of London 2007.
48
J. R. HOPPER ET AL.
Fig. 1. Regional overview of the North Atlantic Ocean and rifted margins. The map is shaded relief of topography and bathymetry with shallow-water areas highlighted in darker shades. Bold numbers indicate the thickness of igneous crust produced immediately after continental crustal separation (values taken from Fowler et al. 1989; Chian & Louden 1994; Chian et al. 1995; Smallwood et al. 1999; Holbrook et al. 2001; Whitmarsh et al. 2001; Funck et al. 2003, 2004; Hopper et al. 2004). CGFZ, Charlie Gibbs Fracture Zone; FC, Flemish Cap; GIB, Galicia Interior Basin; GM, western Galicia margin; IAP, Iberia Abyssal Plain; MAR, Mid-Atlantic Ridge; NF, Newfoundland; RP, Rockall Plateau; UK, United Kingdom.
mechanical deformation of the upper crust because the extensional structures are accessible to seismic imaging techniques. On the Galicia Bank margin off Iberia, the strong regional ‘S’ reflection is observed that is now believed to be a detachment surface at the crust-mantle boundary (Reston et al. 1996). The detachment is thus a major rheological boundary that probably formed as a heavily serpentinized ductile shear zone during the final stages of extension (e.g. Reston et al. 1996; Pe´rez-Gussinye´ & Reston 2001). The uniform sense of shear and rotation indicated by fault blocks above ‘S’ has led to various hypotheses about possible asymmetry during extension and breakup. Farther south along Iberia, a broad, up to 100 km-wide, zone of exhumed serpentinized mantle was exposed at the sea floor (e.g. Whitmarsh
et al. 2001), reminiscent of Hess’ (1962) original hypothesis for the formation of oceanic basins. In this so-called transition zone, the mantle is mechanically unroofed by brittle extension and interacts with sea water to form a thin layer of serpentinite with crust-like seismic velocities (Dean et al. 2000). Initially, this exhumation is thought to be largely amagmatic, but as extension progresses more melt is eventually produced. In the Alps, where these melt products can be sampled and analysed, there is a progression from transitional mid-ocean ridge basalt (T-MORB) towards the continent to normal mid-ocean ridge basalt (N-MORB) towards the ocean (Desmurs et al. 2002; Schaltegger et al. 2002). Desmurs et al. (2002) suggest that this indicates progressive thinning of continental lithospheric mantle, which is eventually
STRUCTURE OF THE FLEMISH CAP MARGIN
replaced by asthenospheric ‘oceanic’ mantle as a magmatic sea-floor spreading system becomes established. It is becoming increasingly clear that resolving the nature and origin of ‘transition zone’ crust is central to understanding final lithospheric rupture along magma starved margins. In this contribution, results from a recent seismic reflection and refraction survey conducted off the coast of Newfoundland are summarized. The data shown are published in Funck et al. (2003) and Hopper et al. (2004, 2006). We do not attempt to replicate all of those results here, but focus on key summary figures and portions of the data set that are particularly relevant for the numerical modelling exercises described in this volume.
Regional tectonic setting and survey description The extensional history of the region between Iberia and Newfoundland spans nearly 100 million years, beginning with an early phase of rifting in the Late Triassic –Early Jurassic that formed the major basins within the Grand Banks and off Portugal (e.g. Tankard & Welsink 1987; Wilson 1988;
49
Tucholke et al. 1989; Rasmussen et al. 1998; Wilson et al. 2001). A 40– 50 million year period of quiescence was followed by a second major phase of extension that ultimately led to complete lithospheric rupture and the establishment of magmatic sea-floor spreading in the North Atlantic some time between Barremian and Aptian time (127 –112 Ma using the timescale of Gradstein et al. 1994). Details regarding the full history and development of the region are summarized in Tucholke et al. (2007) and Tucholke & Whitmarsh (2007). The SCREECH survey (Studies of Continental Rifting and Extension off the Eastern Canadian sHelf) was conducted in 2000 to better understand the large-scale structure of the Newfoundland margin at positions conjugate to the major geophysical surveys and drilling transects off the Iberian margin (Funck et al. 2003; Hopper et al. 2004, 2006; Lau et al. 2006a, b; Shillington et al. 2004, 2006; Van Avendonk et al. 2006). Multichannel seismic reflection and wide-angle seismic refraction data were acquired along three primary transects. Each transect began on the shelf where continental crustal thinning should be minimal and continued seaward well past magnetic anomaly M0 identified over oceanic crust (Figs 2 & 3). Crustal thickness,
Fig. 2. Expanded view of the region offshore Newfoundland showing the SCREECH survey. Open circles are OBS locations along the three major transects, labelled Lines 1, 2 & 3. The thick black line is magnetic anomaly M0 as picked by Srivastava et al. (2000). FC, Flemish Cap. Bathymetric contours are spaced every 500 m.
50
J. R. HOPPER ET AL.
Fig. 3. Regional magnetic anomaly map based on the data of Verhoef et al. (1992). Seismic transects and the anomaly M0 pick from Figure 2 are also shown. Bathymetric contours are at 1000 m intervals, and the bathymetry is shaded with illumination from the NW.
seismic P-wave velocity and reflection imaging provide critical constraints on the large-scale structure of the full rift system between North America and Europe. Here, the primary focus is on the northernmost transect across the Flemish Cap (Line 1 in Fig. 2), a block of Appalachian continental crust that is conjugate to the northernmost Iberian peninsula where the Galicia Interior Basin (GIB) and Galicia Bank margins are located. The profile is also conjugate to the Ocean Drilling Program (ODP) Leg 103 drilling transect as well as a number of seismic reflection and refraction profiles described in Reston et al. (1996), Zelt et al. (2003) and Henning et al. (2004). Data from SCREECH Line 1 are summarized in Figures 4 & 5. The oldest rift-related sediments found by ODP Leg 103 are Tithonian in age (c. 142 Ma Boillot et al. 1987; Enachescu 1987) and, within the GIB, the oldest sediments that can be identified with confidence are also Tithonian (Murillas et al. 1990; Pe´rez-Gussinye´ et al. 2003). Murillas et al. (1990)
described a seismic reflection sequence beneath the Tithonian unit that they proposed is either Late Triassic–Middle Jurassic or older Palaeozoic sediments. In addition, the Flemish Cap graben along the SW part of the cap contains evaporites that are presumably Triassic (Enachescu 1987). However, available data suggest that the Late Triassic phase of extension may have only minimally affected the region this far north. In considering the timing of rifting and the duration of extension along the conjugate profiles, however, it should be noted that drilling and sampling along the Flemish Cap side of the margin is essentially non-existent. An additional point regarding the conjugate margin is that continental crust terminates just seaward of the ‘S’ reflection at a single peridotite ridge beyond which magmatically produced oceanic crust is thought to exist. This is unlike the area farther south along Iberia where a broad, approximately 100 km-wide, zone of exhumed continental mantle and multiple peridotite ridges were exposed at the sea floor prior to magmatic sea-floor
Fig. 4. Seismic velocity model along the SCREECH 1 transect derived from wide-angle data. See text and Funck et al. (2003) for details. Top panel is the model with no vertical exaggeration. Bottom panel is exaggerated 3:1 to show detail. Labelled contours are P-wave velocity.
STRUCTURE OF THE FLEMISH CAP MARGIN 51
Fig. 5. (a) Pre-stack depth migrated data from Hopper et al. (2004) shown from 200 to 320 km along the SCREECH 1 transect. No vertical exaggeration. (b) same as (a) but with seismic velocity model from Figure 4 superimposed. White lines are velocity contours at 0.2 km s21 interval. Greyed area towards the end of the profile indicates no ray coverage in the wide-angle data.
52
J. R. HOPPER ET AL.
spreading (Pickup et al. 1996; Dean et al. 2000; Whitmarsh et al. 2001). However, seismic refraction surveys along Galicia Bank terminate less than 40 km seaward of the peridotite ridge (Whitmarsh et al. 1996), so precisely where magmatically produced oceanic crust begins is not well constrained.
Initial condition and crustal thinning Flemish Cap itself is 30 km thick and appears to be unaffected by regional extension (Fig. 4). Large, sediment-filled rift basins are absent in the reflection images and the wide-angle data show no evidence for sediment cover (see Funck et al. 2003; Hopper et al. 2006). The basement has uniform P-wave velocities between 6.0 and 6.7 km s21, consistent with that typical of Appalachian continental crust. King et al. (1985) correlate Flemish Cap to the Avalon Terrane onshore Newfoundland, which has a seismic velocity structure very similar to that measured along the central part of the cap (Fig. 4) (Hall et al. 1998; Funck et al. 2003). Rock samples from the cap are primarily granodiorites and granites (King et al. 1985) and the crust onshore has whole-crust bulk Poisson’s ratios of 0.23–0.24, indicating that quartzo-feldspathic rocks dominate the crustal composition (Hughes et al. 1994; Hall et al. 1998). One important constraint for modelling exercises is to establish probable compositions for the lower continental crust. In the velocity model of Funck et al. (2003), there are no refracted arrivals from the lower-most layer in the crust. The velocity is thus constrained primarily by PmP phases (Moho reflections), which are clear on most of the Ocean Bottom Seismometers (OBSs) across the Flemish Cap. Despite the inherent ambiguity in determining velocity– depth profiles based only on reflections, lower crustal velocities in access of 6.75 km s21 are not supported by the data. This effectively rules out the possibility of significant gabbroic intrusions in the lower crust. At pressures appropriate for 25–30 km depth, gabbro has a P-wave velocity of more than 6.85 km s21 (Christensen & Mooney 1995). The Flemish Cap data are most consistent with measured velocities for intermediate composition rocks such as diabase (see the compilation in Christensen & Mooney 1995). More mafic rocks such as amphibolites, anorthositic granulites and mafic granulites are also possible, but require temperatures in access of 650 8C. No heatflow data exist on Flemish Cap, but such high temperatures seem unreasonable in this setting. Given the similarity of the velocity structure with onshore observations, where Poisson’s ratio provides additional constraints, and the difficulty with reconciling mafic compositions with the available data, it
is very likely that Flemish Cap is dominated by relatively weak crustal mineralogies. Towards the Flemish Pass basin at the NW end of the transect, reduced seismic velocities are observed in the upper crust. Although the shape of this unit is suggestive of a sedimentary basin, the lack of stratigraphic layering in the reflection section argues against its origin as a rift basin associated with Triassic or later rifting, which should be well preserved. In addition, the seismic velocities are much higher than observed in other Mesozoic basins around Newfoundland. Funck et al. (2003) suggested that this unit most probably corresponds to Precambrian and Palaeozoic sediments and volcanics associated with the Avalon Terrane of the Appalachians. The lack of significant basin formation indicates that Flemish Cap is an intact block of continental lithosphere and thus can be considered as a microplate. This is consistent with previous work on reconstructions of the region (Srivastava et al. 2000; Sibuet et al. 2007). Behaviour as a microplate implies that Flemish Cap must be a strong block of lithosphere that did not deform internally in response to regional extensional stresses responsible for lithospheric rupture. Given the likelihood that the crust is dominated by weaker lithologies, it is inferred that the strong part of the lithosphere must rest primarily in the mantle. This places a qualitative constraint on the initial thermal condition of the lithosphere at this location. In particular, a strong upper mantle indicates reduced temperatures and thermal gradients and implies a cool initial condition (Brace & Kohlstedt 1980). This probably precludes the possibility of significant lower crustal flow in response to crustal buoyancy forces as this requires relatively warm lithosphere even for very weak crustal mineralogies (e.g. Hopper & Buck 1998). Lower crustal flow, it should be noted, is distinct from ductile stretching of the lower and middle crust during extension. The seaward transition from relatively undeformed, 30 km-thick, continental crust to extremely thinned, less than 5 km-thick crust occurs over a narrow interval of 80 km. Moderately thinned continental crust is marked by a single sedimentary basin at 180 km (Figs 4 & 6). Although the basin is clearly observed in the both the wide-angle and normal incidence data, the reflection image quality is poor due to the incomplete removal of multiple energy. However, weak sedimentary reflections appear to fan seaward, indicating a halfgraben bounded by a landward-dipping fault. The crust beneath this basin is 20 km thick, but it thins abruptly seaward to a thickness of less than 5 km over only 40 –50 km. The abruptly necking zone of crust from 150 to 240 km is particularly interesting. Figure 7 shows
Fig. 6. Seismic reflection section in two-way travel time, where continental crust thins abruptly from depths 170 to 215 km. Top panel is uninterpreted data, bottom panel includes an interpretation as discussed in the text. Seaward of 215 km, the data are time-converted from the pre-stack depth migration from Hopper et al. (2004). The full depth migration is also shown in Figure 5.
STRUCTURE OF THE FLEMISH CAP MARGIN 53
54
J. R. HOPPER ET AL.
Thinning factor
5.00
4.00
3.00
UC
MC
LC
2.00
1.00 150
160
170
180
190
200
210
Distance along profile (km)
Fig. 7. Thinning factors of different seismically defined layers as a function of distance along the profile. The thinning factor is simply a reference thickness divided by the actual crustal thickness at a location. Reference thickness is taken from 110 km, which is approximately where each layer is thickest across the Flemish Cap on the SCREECH profile. UC, upper crust (Vp ¼ 6.0– 6.2 km s21); MC, middle crust (Vp ¼ 6.3–6.4 km s21); LC, lower crust (Vp ¼ 6.6– 6.7 km s21). See text for further details.
the thinning factors of different crustal layers as defined by the seismic refraction data. The upper crust has velocities from 6.0 to 6.2 km s21, middle crust has velocities from 6.3 to 6.4 km s21, and lower crust has velocities from 6.6 to 6.7 km s21 (Funck et al. 2003). From 150 to 210 km, the upper crust shows extensional thinning and strain concentrated in a local area defined by the rift basin at 160 –180 km. Little thinning of the upper crust is observed elsewhere over this interval. In contrast, the middle and lower crust thin uniformly as is expected for layers that have undergone ductile stretching by bulk pure shear. At least locally, these observations seem to require that deformation of the upper crust was detached from deformation deeper in the crust. The overall amount of extension in the upper crust seems to balance with the amount of extension indicated in the middle and lower crust over this interval despite a small lateral offset in the location where strain was focused. The amount of stretching in the middle and lower crust increases rapidly from 200 km seaward and one or both these layers may pinch out. The details of this pinch-out, however, are difficult to resolve. In addition, the use of seismic velocities to define original upper, middle or lower crust under such extreme thinning is questionable because decompression, porosity changes and mineral alteration from circulating fluids will alter the seismic velocity as deeper material is brought closer to the surface. Thus, we cannot attach significance to layer thickness changes defined solely by seismic velocities beyond 210 km. In the zone from 210 to 240 km, the middle and lower crust appear to be very thin, and the upper crust thins significantly. The presence of small rift
basins indicates brittle failure of the upper crust, but specific interpretation of the fault pattern is difficult because of the lack of reflections from fault surfaces. One possible interpretation of this section is shown in Figure 6. An important feature of the data is an interval of east-dipping reflections primarily in the lower crust that is interpreted to be Palaeozoic structure formed during the assembly of the Avalon terrane (Hopper et al. 2006). These reflections continue through and below reflection and refraction Moho, although the reflection Moho is discontinuous through the cross-over. It is possible that the Palaeozoic structure was reactivated by normal faulting and may have acted as a detachment surface at which overlying high-angle normal faults terminate. The small basins centred at 216 and 227 km appear to enclose an interval with high velocities of 6.0–6.1 km s21. On the basis of velocity alone, this might be considered to represent upper continental crust. However, the sections show horizontal layering that is unconformable to the basin margins, suggesting a possible sedimentary origin. It is possible that they are limestones and dolomites, which can have velocities of 6 km s21 or more (e.g. Press 1966). In addition, it is known from drilling at ODP site 639 that Upper Jurassic–Lower Cretaceous dolomite is present on the conjugate Galicia margin (Boillot et al. 1987). If the interpretation that much of the seismic upper crust is in fact sedimentary, then faulting probably formed the basins early during the second Newfoundland–Iberia rift phase. Faulting during the earlier, Late Triassic – Early Jurassic, rifting is unlikely because there is no indication of evaporites, which probably would be mobile and would distort the sedimentary fill. Deposition of evaporites was widespread in the basins formed on the Grand Banks and Iberia margins during the first rift phase (Tankard & Welsink 1987; Wilson et al. 1989). The origin of the basement highs at 230–240 km is uncertain. Hopper et al. (2006) present arguments for and against various hypotheses of the section as either continental blocks or as exhumed continental mantle. Figure 6 shows one possible interpretation where the blocks are continental crust bounded by east-dipping normal faults that formed during the early phase of extension that led to the abrupt thinning along the margin. Erosion of the tops of the blocks would explain their rounded tops and apparent coarse debris shed into the adjacent basins. The fault labelled 1.1 would post-date the main faulting associated with these blocks, thus forming the basins centred at 216 and 222 km probably during the Jurassic (see above). The orange faults within the sedimentary fill are related to extension that post-dates the main crustal thinning phase of extension.
STRUCTURE OF THE FLEMISH CAP MARGIN
Although it is not possible to rule out the possibility that the blocks from 230 to 240 km could be exhumed continental mantle, major large-scale asymmetry between Flemish Cap and Galicia is clearly supported by the data. The abrupt necking profile and narrow region of basin formation is in strong contrast to the conjugate, which shows a broad, 300 km-wide region of attenuated continental crust. Figure 8 shows the profile described here reconstructed to a composite profile of the Galicia Bank and Galicia Interior Basin based on data from Pe´rez –Gussinye´ et al. (2003), Reston et al. (1996), Gonza´lez et al. (1999) and Whitmarsh et al. (1996). Importantly, the asymmetry observed is independent of the precise placement of the seaward termination of continental crust along Flemish Cap profile. Moving the boundary 10 km landward and interpreting the blocks as exhumed upper mantle does not fundamentally alter the conclusion of asymmetric margin formation. The question of symmetry v. asymmetry on conjugate rifted margins has received a great deal of attention (e.g. Lister et al. 1986; Dunbar & Sawyer 1989b; Sibuet 1992; Louden & Chian 1999). Early models suggested that asymmetry requires strong crust– mantle decoupling, whereby crustal strain is displaced from mantle lithospheric strain, eventually resulting in asymmetric conjugate rifted margin pairs (e.g. Braun & Beaumont 1989; Dunbar & Sawyer 1989a). Such a mode of decoupling, however, may be difficult to achieve for realistic geological conditions (Hopper & Buck 1998). Hopper & Buck (1996) proposed instead that asymmetric margins form when wide-rift mode extension breaks along one edge. Both these possibilities, decoupling and wide-rift mode extension, probably require warm lithospheric conditions (Buck 1991), which seems at odds with the arguments presented above for cool, stronger lithosphere. Recently, a new class of dynamical models of extension have shown that symmetry and asymmetry may instead be controlled by strain-dependent
55
rheologies (Huismans & Beaumont 2002, 2003; Lavier & Manatschal 2006). In particular, the results of Huismans & Beaumont (2003) suggest that strongly asymmetric rifting occurs when: (1) extension is slow; (2) there is significant strain softening of the dominant rheology; and (3) crust is coupled to the lithospheric mantle. Their conclusions are broadly consistent with data available from the Flemish Cap.
Mantle serpentinization It has become increasing clear that penetration of sea water into the upper mantle can result in large-scale serpentinization along divergent plate boundaries. This results in fundamental but still poorly understood changes in the rheology of the upper mantle that probably have a significant impact on mantle and crustal deformation during extension. In addition to having rheological effects, the serpentinization process is strongly exothermic and results in significant volume expansion and density reduction (e.g. Schroeder et al. 2002). Thus, it may play an important role in the thermal structure and evolution of divergent plates boundaries as well as having an impact on the dynamic evolution through buoyancy changes. One of the most important results of research on the Iberia margin is the recognition of the important role that mantle serpentinization plays in the final stages of non-volcanic margin development (e.g. Manatschal 2004). In particular, crustal separation is not always coincident with final lithospheric rupture. Instead, a broad zone of exhumed continental mantle forms and there is an absence of extrusive magmatic rocks and oceanic crust. A key result of the SCREECH Line 1 is that, with the possible exception of the crustal block at 230–240 km, there is no convincing evidence of mantle exhumation occurring in a similar environment. In previous work, Funck et al. (2003) and Hopper et al. (2004,
Fig. 8. Plot of Flemish Cap data summary from this work reconstructed to conjugate data sets. Data were spliced together at the seaward-most point where continental crust can be identified in each case and the data shifted so that the sea floor is at the same depth. Conjugate data are based on profiles published in Reston et al. (1996), Whitmarsh et al. (1996), Gonza´lez et al. (1999) and Pe´rez-Gussinye´ et al. (2003).
56
J. R. HOPPER ET AL.
2006) presented a number of arguments demonstrating that some type of oceanic crust exists seaward of 240 km. OBS 19, just on the eastern side of the basement blocks at 230 –240 km, shows clear evidence for a rapid seaward velocity change that is best explained by anomalously thin oceanic crust overlying serpentinized mantle beginning at this location (Figs 4 & 5) (Funck et al. 2003). Thus, it appears that a combination of extrusive magmatism and mantle exhumation characterized extension shortly after continental crustal separation along this transect. Nevertheless, mantle serpentinization is clearly indicated along the profile, and it must have played an important role in the late stages of margin development that led to full lithospheric rupture. Well-defined and easily identifiable upper mantle Pn phases from beneath the thin magmatic crust along the profile consistently show phase velocities of 7.6 km s21. This is significantly less than the 8 km s21 measured below Flemish Cap (Funck et al. 2003), and it suggests a maximum of approximately 15– 20% serpentinization of olivine peridotite immediately beneath the thin crust (100 –150 Mpa: e.g. Christensen 2004). Amplitude modelling further constrains the velocity gradient in the mantle and shows that velocities of 8 km s21 are reached by about 5 km below the base of the crust. The velocity data provide convincing evidence that serpentinization becomes important when the crust thins to less than 6–8 km thick. Figure 9 shows the mantle P-wave velocity just below Moho as a function of crustal thickness along the SCREECH Line 1. Assuming that reduced mantle velocities indicate the onset of serpentinization, this process appears to initiate at a crustal thickness of approximately 8 km. Simple thermal models of lithospheric extension show that this is about the point at which the entire crust becomes brittle (Pe´rez-Gussinye´ & Reston 2001). In the absence of significant melt production, extension can then only be accommodated by brittle failure, which
Vp (km/s)
8.25
8.00
7.75 onset of serpentinization
7.50 0
3
6
9
12
15
Crustal Thickness (km)
Fig. 9. Mantle P-wave velocity just below the Moho as a function of thickness of the overlying crust. See the text for explanation.
provides pathways for water to penetrate into the mantle and to facilitate serpentinization. This idea is generally consistent with the interpretation of the seismic reflection section of SCREECH transect, given that major faults bounding the basins at 210–230 km most probably cut deep into the crust and possibly into the upper mantle. These faults could provide the conduits necessary for fluids to circulate deep into the lithosphere.
Magma production and sea-floor spreading The last important observation the SCREECH data provide is information on the onset of melting and the generation of magmatically produced oceanic crust, which marks final lithospheric rupture. Historically, amagmatic margins have been studied to understand brittle and mechanical deformation of the crust. The absence of melt in the system has important consequences. From an observational point of view, extensional structures that result from thinning and stretching are easier to image and sample as they are not covered up by volcanic rocks. From a modelling point of view, extensional stress cannot be relieved by magmatic intrusion and dyke injection so only mechanical deformation of the lithosphere needs to be considered up to the time that volumetrically significant melt is produced. The SCREECH Line 1 results are fully consistent with an absence of significant decompression melting prior to breakup of the continental crust, despite extreme stretching and thinning. Across Flemish Cap and along the abruptly thinning crust farther seaward, seismic velocities measured in the lower crust are low, consistent with an absence of underplating or gabbroic intrusion into lower crust. Crustal velocities of more than 6.8 km s21 are not observed anywhere in the section until 240 km, where thin magmatic crust is interpreted. In addition, the reflection data show little evidence for significant sill intrusion into the few sedimentary basins that are observed. To the extent that melts may have formed and infiltrated the upper mantle, there was insufficient production to generate extrusive flows or geophysically resolvable intrusive layers within the crust. Melting and magmatism, when it did appear, produced anomalously thin, 3–4 km thick, oceanic crust, indicating a magma-starved crustal accretion system. Such crust is observed in modern settings along ultra-slow spreading ridges like the Gakkel Ridge where crust is only 4 km thick (Coakley & Cochran 1998; Jokat et al. 2003). This is further consistent with Srivastava’s estimate that extension at the time of opening was approximately 14 mm
STRUCTURE OF THE FLEMISH CAP MARGIN
year21 (full rate), which is also comparable to Gakkel Ridge spreading rates. Crust with closer to average oceanic thickness did not appear until about 295 km on SCREECH Line 1, some 20–25 km seaward of anomaly M0 as picked by Srivastava et al. (2000) (Fig. 5). Unfortunately, velocities and crustal thickness on this part of the refraction line are not well constrained (Funck et al. 2003), and the line does not extend far enough to confirm that the crust continues seawards with normal oceanic thicknesses. While it is tempting to conclude that extension rate was the controlling factor in suppressing melting in the Iberia –Newfoundland rift, it is not clear that this is the only important factor. Reston & Phipps-Morgan (2004) proposed that anomalously cool upper mantle may be important in limiting or suppressing melt production. A compilation of continental geotherms by Ro¨hm et al. (2000) shows that the potential temperature of the upper mantle may have a natural variation of several hundred degrees, and the possibility that the Newfoundland Iberia rift was underlain by an asthenospheric cool spot is one that must be considered. It is clear that large variations in melt supply are indicated once magmatism was established in the rift. At 273– 290 km, oceanic layer 3 is absent, and the crust thins to as little as 1.5 km thick in part of this zone. Hopper et al. (2004) attributed this to a cessation of magmatism that resulted in the formation of an oceanic core complex in which the lower crust was mechanically removed. They further proposed that magmatism was episodically more robust, in order to explain apparent volcanic stratigraphy that is observed in the upper crust. Such volcanic stratigraphy is also observed on data 10 km to the north of Line 1 (Hopper et al. 2006). This kind of variation is not unique to the Flemish Cap –Galicia transect. There are significant disparities in the amount of melt introduced throughout the rift following breakup of the continental crust, although a common denominator appears to be that all areas exhibit significant exhumation and serpentinization of mantle (see Tucholke et al. 2007). Chemical heterogeneities may have also played a key role in the temporal and spatial variability of magmatism. In a magma-starved rift, small pockets of anomalously fertile mantle could have a large effect on local areas. For example, passively imbedded mantle inhomogeneities have been proposed to explain isolated magmatic centres along the ultra-slow Gakkel Ridge (Cochran et al. 2003). The available data indicate that large spatial and temporal variations in melt supply may be important for understanding the detailed structure of non-volcanic margins during the extensional phase following complete separation of continental crust
57
and prior to the establishment of a robust sea-floor spreading system. A complex interplay of brittle faulting and mantle exhumation, melting, and intrusion and extrusion of melt controls the detailed structure of this crust. Geophysical data alone cannot constrain these processes and further progress in this area will require basement sampling and geochemical analyses.
Conclusions The seismic reflection and refraction data from the Flemish Cap margin, together with seismic surveys over the conjugate Galicia Bank margin and Galicia Interior Basin, provide first-order constraints for numerical modelling of lithospheric rupture in a nonvolcanic rift. In particular, the data show important differences in margin development and document asymmetries on a variety of scales. Well-defined crustal and mantle arrivals provide detailed constraints on the crustal thickness across the entire transect. Flemish Cap itself has crustal seismic velocities consistent with velocities measured in Appalachian crust from the Avalon terrane throughout the Newfoundland region. Such crust is probably dominated by quartzo-feldspathic lithologies (Hughes et al. 1994). There is a conspicuous absence of significant rift basins in the continental block comprising Flemish Cap. Rift basins appear on the eastern margin of the block, but they are small compared to other rift basins on the western side of the block and within the Grand Banks. The continental crust thins rapidly on the eastern margin of Flemish Cap, from 30 to approximately 3 km thick, over a distance of about 80 km. Mantle velocities beneath Flemish Cap are 8.0 km21 and thus do not indicate anything unusual about the composition of the upper mantle. Reduced seismic velocities are observed under thinned continental crust farther seaward, beginning where the crust is approximately 8 km thick. Reflection images suggest that the embrittlement of the entire crust may have occurred at this point, providing pathways for fluids to penetrate into the mantle and facilitate serpentinization. However, with the possible exception of one 10 km-wide basement block at the edge of the thinned continental crust, there is no evidence that mantle was extensively exhumed as is observed off the Iberia margin. Nevertheless, serpentinization of the upper 5 km of mantle is indicated by anomalously low seismic velocities. Serpentinization is an exothermic reaction that results in a significant volume increase and density reduction. In addition, the rheology of serpentinite is considerably different than peridotite. Thus, there clearly are important, but as yet unexplored, implications of this process
58
J. R. HOPPER ET AL.
for understanding the dynamic evolution of rifted margins during the final stages of lithospheric rupture. We do not observe high seismic velocities in the lower crust that would be expected from underplating or gabbroic intrusion as a result of decompression melting during rifting of the continental crust. There is no evidence for melt production until after the breakup of the continental crust, at which point thin (3–4 km) magmatic crust with oceanic velocity structure was produced. This crust shows a strong variation in thickness and it consistently overlies serpentinized mantle several kilometres thick, indicating a complex phase of mantle extension with variable magmatism before a more magmatically robust sea-floor spreading system was established. The timing of rifting events on the Flemish Cap margin is uncertain because of a lack of drilling and sampling. It seems likely that the first phase of Newfoundland–Iberia extension (Late Triassic – Early Jurassic) did not significantly affect the crust between Flemish Cap and Galicia Bank. The oldest sediments drilled on Galicia Bank are shallow-water Tithonian carbonates and mudstones, which suggests that rifting of the crust there began during the second, Late Jurassic –Early Cretaceous, phase of extension. Possible high-velocity sediments in the small rift basins at the seaward edge of thinned Flemish Cap crust may be the counterparts of these Tithonian carbonates at the western edge of the rift. If it is assumed that rifting of the eastern Flemish Cap margin began in the Tithonian, then continental crustal extension followed by the formation of thin magmatic crust combined with mantle extension and exhumation occupied an interval of at least approximately 25 million years until apparently normal ocean crust was formed seaward of anomaly M0. There are several parameters critical for numerical modelling that the presently available data do not constrain very well. There is no information on the lithospheric thickness of the region, and we can only speculate on the thermal structure of the lithosphere prior to rupture. However, the data are consistent with Flemish Cap behaving as a microplate, implying that is a relatively strong block of lithosphere. Thus, a relatively cool geotherm and thick lithospheric mantle is indicated. The SCREECH survey was a joint project sponsored by the US National Science Foundation (NSF), the Danish National Research Foundation, and the Natural Science and Engineering Research Council of Canada. We thank the captains and crews of the R/V Maurice Ewing and R/V Oceanus for a successful experiment. Depth imaging was supported by European Union grant HPRI-CT1999-00037 to GEOMAR (now Leibniz Insititute for
Marine Science) in Kiel, Germany. This work has benefited from discussions with our SCREECH colleagues K. Louden, D. Shillington, H. van Avendonk, W. S. Holbrook, H. C. Larsen, H. Lau & S. Deemer. Reviews from S. Deemer, R. Huismans, G. Manatschal & M. Steckler helped to improve the manuscript. B. Tucholke acknowledges support by NSF grants 9819053 and 0326714, and by the Bigelow Chair in Oceanography at Woods Hole Oceanographic Institution.
References B OILLOT , G., W INTERER , E. L. ET AL . (eds). 1987. Proceedings of the Ocean Drilling Program, Initial Reports, 103. Ocean Drilling Program, College Station, TX. B OUTILIER , R. R. & K EEN , C. E. 1999. Small-scale convection and divergent plate boundaries. Journal of Geophysical Research, 104, 7389–7403. B RACE , W. F. & K OHLSTEDT , D. L. 1980. Limits on lithospheric stress imposed by laboratory experiments. Journal of Geophysical Research, 85, 6248– 6252. B RAUN , J. & B EAUMONT , C. 1989. Dynamical models of the role of crustal shear zones in asymmetric continental extension, Earth and Planetary Science Letters, 93, 405–423. B UCK , W. R. 1991. Modes of continental lithospheric extension. Journal of Geophysical Research, 96, 20,161–20,178. C HIAN , D. & L OUDEN , K. E. 1994. The continent –ocean transition across the southwest Greenland margin. Journal of Geophysical Research, 99, 9117– 9135. C HIAN , D., L OUDEN , K. E. & R EID , I. 1995. Crustal structure of the Labrador Sea conjugate margin and implications for the formation of nonvolcanic continental margins. Journal of Geophysical Research, 100, 24,239– 24,253. C HRISTENSEN , N. I. 2004. Serpentinites, peridotites, and seismology. International Geology Review, 46, 795–816. C HRISTENSEN , N. I. & M OONEY , W. D. 1995. Seismic velocity structure and composition of the continental crust: A global view. Journal of Geophysical Research, 100, 9761–9788. COAKLEY, B. J. & C OCHRAN , J. C. 1998. Gravity evidence of very thin crust at the Gakkel Ridge (Arctic Ocean). Earth and Planetary Science Letters, 162, 81–95. C OCHRAN , J. C., K URRAS , G. J., E DWARDS , M. H. & C OAKLEY , B. J. 2003. The Gakkel Ridge: Bathymetry, gravity anomalies, and crustal accretion at extremely slow spreading rates. Journal of Geophysical Research, 108, B2, 2116, doi:10.1029/2002JB001830. D EAN , S. M., M INSHULL , T. A., W HITMARSH , R. B. & L OUDEN , K. E. 2000. Deep structure of the ocean– continent transition in the southern Iberia Abyssal Plain from seismic refraction profiles: The IAM-9 transect at 408200 N. Journal of Geophysical Research, 105, 5859–5885. D ESMURS , L., M U¨ NTENER , O. & M ANATSCHAL , G. 2002. Onset of magmatic accretion within a magmapoor rifted margin: a case study from the Platta ocean–continent transition, eastern Switzerland. Contributions to Mineralogy and Petrology, 144, 365–382.
STRUCTURE OF THE FLEMISH CAP MARGIN D UNBAR , J. A. & S AWYER , D. S. 1989a. How preexisting weaknesses control the style of continental breakup. Journal of Geophysical Research, 94, 7278–7292. D UNBAR , J. A. & S AWYER , D. S. 1989b. Patterns of continental extension along the conjugate margins of the central and North Atlantic Oceans and Labrador Sea. Tectonics, 8, 1059–1077. E NACHESCU , M. E. 1987. Tectonic and structural framework of the northeast Newfoundland continental margin. In: B EAUMONT , C. & T ANKARD , A. J. (eds) Sedimentary Basins and Basin-forming Mechanisms. Canadian Society of Petroleum Geologists Memoir, 12, 117–146. F OWLER , S. R., W HITE , R. S., S PENCE , G. D. & W ESTBROOK , G. K. 1989. The Hatton Bank continental margin – II. Deep structure from two-ship expanding spread profiles. Geophysical Journal International, 96, 295–309. F UNCK , T., H OPPER , J. R., L ARSEN , H. C., L OUDEN , K. E., T UCHOLKE , B. E. & H OLBROOK , W. S. 2003. Crustal structure of the ocean–continent transition at Flemish Cap: Seismic refraction results. Journal of Geophysical Research, 108, B11, 2531, doi:10.1029/2003JB002434. F UNCK , T., J ACKSON , H. R., L OUDEN , K. E., D EHLER , S. A. & W U , Y. 2004. Crustal structure of the northern Nova Scotia rifted continental margin (eastern Canada). Journal of Geophysical Research, 109, B09102, doi:10.1029/2004JB003008. G ONZA´ LEZ , A., C O´ RDOBA , D. & V ALES , D. 1999. Seismic crustal structure of Galicia continental margin, NW Iberian Peninsula. Geophysical Research Letters, 26, 1061–1064. G RADSTEIN , F. M., A GTERBERG , F. P., O GG , J. G., H ARDENBOL , J., VAN V EEN , P., T HIERRY , J. & H UANG , Z. 1994. A Mesozoic time scale. Journal of Geophysical Research, 99, 24051–24074. H ALL , J., M ARILLIER , F. & D EHLER , S. 1998. Geophysical studies of the structure of the Appalachian orogen in the Atlantic borderlands of Canada. Canadian Journal of Earth Sciences, 35, 1205–1221. H ENNING , A. T., S AWYER , D. S. & T EMPLETON , D. C. 2004. Exhumed upper mantle within the ocean– continent transition on the northern West Iberia margin: Evidence from prestack depth migration and total tectonic subsidence analyses. Journal of Geophysical Research, 109, B05103, doi:10.1029/ 2003JB002526. H ESS , H. H. 1962. History of ocean basins. In: E NGEL , A. E. J., J AMES , H. L. & L EONARD , B. F. (eds) Petrologic Studies: a Volume in Honor of A. F. Buddington. Geological Society of America, Boulder, CO, 599–620. H OLBROOK , W. S., L ARSEN , H. C. ET AL . 2001. Mantle thermal structure and active upwelling during continental breakup in the North Atlantic. Earth and Planetary Science Letters, 190, 251–266. H OPPER , J. R. & B UCK , W. R. 1996. The effect of lower crustal flow on continental extension and passive margin formation. Journal of Geophysical Research, 101, 20175–20194. H OPPER , J. R. & B UCK , W. R. 1998. Styles of extensional decoupling. Geology, 26, 699–702.
59
H OPPER , J. R., F UNCK , T. ET AL . 2004. Continental breakup and the onset of ultraslow seafloor spreading off Flemish Cap on the Newfoundland rifted margin. Geology, 32, 93– 96, doi:10.1130/G19694.1. H OPPER , J. R., F UNCK , T., T UCHOLKE , B. E., L OUDEN , K. E., H OLBROOK , W. S. & L ARSEN , H. C. 2006. A deep seismic investigation of the Flemish Cap margin: Implications for the origin of deep reflectivity and evidence for asymmetric breakup between Newfoundland and Iberia. Geophysical Journal International, 164, 501–515, doi:10.1111/j.1365-246X. 2006.02800.x. H UGHES , S., H ALL , J. & L UETGERT , J. H. 1994. The seismic velocity structure of the Newfoundland Appalachian orogen. Journal of Geophysical Research, 99, 13633–13653. H UISMANS , R. S. & B EAUMONT , C. 2002. Asymmetric lithospheric extension: The role of frictional– plastic strain softening inferred from numerical experiments. Geology, 30, 211– 214. H UISMANS , R. S. & B EAUMONT , C. 2003. Symmetric and asymmetric lithospheric extension: Relative effects of frictional–plastic and viscous strain softening. Journal of Geophysical Research, 108, B10, 2496 doi: 10.1029/2002JB002026. J OKAT , W., R ITZMANN , O., S CHMIDT -A URSCH , M. C., D RACHEV , S., G AUGER , S. & S NOW , J. 2003. Geophysical evidence for reduced melt production on the Arctic ultraslow Gakkel mid-ocean ridge. Nature, 423, 962–965. K ING , L. H., F ADER , G. B., P OOLE , W. H. & W ANLESS , R. K. 1985. Geological setting and age of the Flemish Cap granodiorite, east of the Grand Banks of Newfoundland. Canadian Journal of Earth Sciences, 22, 1286– 1298. L AU , K. W., L OUDEN , K. E. ET AL . 2006a. Crustal structure across the Grand Banks–Newfoundland Basin continental margin (Part 2) – results from a seismic reflection profile. Geophysical Journal International, 167, 157–170, doi:10.1111/j.1365246X.2006. 02989.X L AU , K. W., L OUDEN , K. E., F UNCK , T., T UCHOLKE , B. E., H OLBROOK , W. S., H OPPER , J. R. & L ARSEN , H. C. 2006b. Crustal structure across the Grand Banks–Newfoundland Basin continental margin (Part 1) – Results from a seismic refraction profile. Geophysical Journal International, 167, 127 –156, doi:10.1111/j.1365-246X.2006.02988.X L AVIER , L. L. & M ANATSCHAL , G. 2006. A mechanism to thin the continental lithosphere at magma-poor margins. Nature, 440, 324–328, doi:10.1038/nature04068. L AVIER , L. L., B UCK , W. R. & P OLIAKOV , A. N. B. 1999. Self-consistent rolling-hinge model for the evolution of large-offset low-angle normal faults. Geology, 27, 1127–1130. L ISTER , G. S., E THERIDGE , M. A. & S YMONDS , P. A. 1986. Detachment faulting and the evolution of passive continental margins. Geology, 14, 246– 250. L OUDEN , K. E. & C HIAN , D. 1999. The deep structure of non-volcanic rifted continental margins. Philosophical Transactions of the Royal Society of London, A357, 767– 804. M ANATSCHAL , G. 2004. New models for evolution of magma-poor rifted margins based on a review of
60
J. R. HOPPER ET AL.
data and concepts from West Iberia and the Alps. International Journal of Earth Sciences, 93, 432–466. M INSHULL , T. A., D EAN , S. M., W HITE , R. S. & W HITMARSH , R. B. 2001. Anomalous melt production after continental breakup in the southern Iberia Abyssal Plain. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Nonvolcanic Rifting of Continental Margins: A Comparison of Evidence from Land and Sea. Geological Society, London, Special Publications, 187, 537–550. M URILLAS , J., M OUGENOT , D., B OILLOT , G., C OMAS , M. C., B ANDA , E. & M AUFFRET , A. 1990. Structure and evolution of the Galicia Interior Basin (Atlantic western Iberian continental margin). Tectonophysics, 184, 297–319. N IELSEN , T. K. & H OPPER , J. R. 2004. From rift to drift: Mantle melting during continental breakup. Geochemistry Geophysics Geosystems, 5, Q07003, doi: 10.1029/2003GC000662. P E´ REZ -G USSINYE´ , M. & R ESTON , T. J. 2001. Rheological evolution during extension at nonvolcanic rifted margins: Onset of serpentinization and development of detachments leading to continental breakup. Journal of Geophysical Research, 106, 3961– 3975. P E´ REZ -G USSINYE´ , M., R ANERO , S. R., R ESTON , T. J. & S AWYER , D. 2003. Structure and mechanisms of extension at the Galicia Interior Basin, west of Iberia. Journal of Geophysical Research, 108, 2245, doi:10.1029/2001JB000901. P ICKUP , S. L. B., W HITMARSH , R. B., F OWLER , C. M. R. & R ESTON , T. J. 1996. Insight into the nature of the ocean– continent transition off west Iberia from a deep multichannel seismic reflection profile. Geology, 24, 1079– 1082. P RESS , F. 1966. Seismic velocities. In: S YDNEY , J. & C LARK , P. (eds) Handbook of Physical Constants, rev. edn. Geological Society of America Memoir, 97, 195– 218. R ASMUSSEN , E. S., L OMHOLT , S., A NDERSEN , C. & V EJBÆK , O. V. 1998. Aspects of the structural evolution of the Lusitanian Basin in Portugal and the shelf and slope area offshore Portugal. Tectonophysics, 300, 199–225. R ESTON , T. J. & P HIPPS M ORGAN , J. 2004. Continental geotherm and the evolution of rifted margins. Geology, 32, 133–136. R ESTON , T. J., K RAWCZYK , C. M. & K LAESCHEN , D. 1996. The S reflector west of Galicia (Spain): Evidence from prestack depth migration for detachment faulting during continental breakup. Journal of Geophysical Research, 101, 8075–8091. R O¨ HM , A. H. E., S NIEDER , R., G OES , S. & T RAMPERT , J. 2000. Thermal structure of continental upper mantle inferred from S-wave velocity and surface heat flow. Earth and Planetary Science Letters, 181, 395–407. S CHALTEGGER , U., D ESMURS , L., M ANATSCHAL , G., M U¨ NTENER , O., M EIER , M., F RANK , M. & B ERNOULLI , D. 2002. The transition from rifting to seafloor spreading within a magma-poor rifted margin: field and isotopic constraints. Terra Nova, 14, 156–162. S CHROEDER , T., J OHN , B. & F ROST , B. R. 2002. Geologic implications of seawater circulation through peridotite exposed at slow-spreading mid-ocean ridges. Geology, 30, 367 –370.
S HILLINGTON , D., H OLBROOK , W. S., T UCHOLKE , B. E., H OPPER , J. R., L OUDEN , K. E., L ARSEN , H. C. & VAN A VENDONK , H. 2004. Data report: marine geophysical data on the Newfoundland non-volcanic rifted Margin around SCREECH Transect 2. In: T UCHOLKE , B. E., S IBUET , J.-C. ET AL . Proceedings of the Ocean Drilling Program, Initial Results, 210. Ocean Drilling Program, College Station, TX. S HILLINGTON , D. J., H OLBROOK , W. S. ET AL . 2006. Evidence for asymmetric nonvolcanic rifting and slow incipient sea-floor spreading from seismic reflection data on the Newfoundland margin, Journal of Geophysical Research, 111, B09402, doi:10.1029/ 2005JB003981. S IBUET , J. C. 1992. Formation of non-volcanic passive margins: A composite model applied to the conjugate Galicia and southeastern Flemish Cap margins. Geophysical Research Letters, 19, 769– 773. S IBUET , J. C., S RIVASTAVA , S. P., E NACHESCU , M. & K ARNER , G. D. 2007. Early Cretaceous motion of Flemish Cap with respect to North America: implications on the formation of Orphan Basin and SE Flemish Cap– Galicia Bank conjugate margins. In: K ARNER , G. D., M ANATSCHAL , G. & P INHEIRO , L. M. (eds) Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, London, Special Publications, 282, 63–76. S MALLWOOD , J. R., S TAPLES , R. K., R ICHARDSON , K. R., W HITE , R. S. & THE FIRE WORKING GROUP . 1999. Structure and development of the crust above the Iceland mantle plume: from continental rift to oceanic spreading center. Journal of Geophysical Research, 104, 22,885– 22,902. S RIVASTAVA , S. P. & T APSCOTT , C. R. 1986. Plate Kinematics of the North Atlantic. In: V OGT , P. R. & T UCHOLKE , B. E. (eds) The Geology of North America. Vol. M. The Western North Atlantic Region. Geological Society of America, Boulder, CO, 379–404. S RIVASTAVA , S. P., S IBUET , J.-C., C ANDE , S., R OEST , W. R. & R EID , I. D. 2000. Magnetic evidence for slow seafloor spreading during the formation of the Newfoundland and Iberian margins. Earth and Planetary Science Letters, 182, 61–76. T ANKARD , A. J. & W ELSINK , H. J. 1987. Extensional tectonics and stratigraphy of Hibernia oil field, Grand Banks, Newfoundland. AAPG Bulletin, 71, 1210–1232. T UCHOLKE , B. E. & W HITMARSH , R. B. 2007. The Newfoundland–Iberia conjugate rifted margins. In: B ALLY , A. W. & R OBERTS , D. G. (eds) Principles of Phanerozoic Regional Geology. Elsevier. (in press) T UCHOLKE , B. E., A USTIN , J. A. JR & U CHUPI , E. 1989. Crustal structure and rift-drift evolution of the Newfoundland Basin. In: T ANKARD , A. J. & B ALKWILL , H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins. AAPG Memoir, 46, 247–263. T UCHOLKE , B. E., S AWYER , D. S. & S IBUET , J.-C. 2007. Breakup of the Newfoundland– Iberia rift. In: K ARNER , G. D., M ANATSCHAL , G. & P INHEIRO , L. M. (eds) Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, London, Special Publications, 282, 9 –45.
STRUCTURE OF THE FLEMISH CAP MARGIN T UCHOLKE , B. E., S IBUET , J. -C. ET AL . 2004. Leg 210. In: Proceedings of the Ocean Drilling Program, Initial Results, 210. Ocean Drilling Program, College Station, TX. V AN A VENDONK , H., H OLBROOK , W. S. ET AL . 2006. Seismic velocity structure of the rifted margin of the eastern Grand Banks, of Newfoundland, Canada. Journal of Geophysical Research, 111, B11404, doi:10.1029/2005JB04156. V ERHOEF , J., M ACNAB , R. & S RIVASTAVA , S. P. 1992. New magnetic anomaly maps of the Arctic and North Atlantic oceans. Proceedings of the International Geological Congress, 29, 974. W HITMARSH , R. B., M ANATSCHAL , G. & M INSHULL , T. A. 2001. Evolution of magma-poor continental margins from rifting to seafloor spreading. Nature, 413, 150– 154. W HITMARSH , R. B., W HITE , R. S., H ORSEFIEL , S. J., S IBUET , J.-C., R ECQ , M. & L OUVEL , V. 1996. The ocean–continent boundary off the western continental margin of Iberia: Crustal structure west of Galicia Bank. Journal of Geophysical Research, 101, 28,291–28,314.
61
W ILSON , R. C. L. 1988. Mesozoic development of the Lusitanian Basin, Portugal. Revista de la Sociedad Geolo´gica de Espan˜a, 1, 393– 407. W ILSON , R. C. L., H ISCOTT , R. N., W ILLIS , M. G. & G RADSTEIN , F. M. 1989. The Lusitanian Basin of west-central Portugal: Mesozoic and Tertiary tectonic, stratigraphic, and subsidence history. In: T ANKARD , A. J. & B ALKWILL , H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins. AAPG Memoir, 46, 265– 282. W ILSON , R. C. L., M ANATSCHAL , G. & W ISE , S. 2001, Rifting along non-volcanic passive margins: stratigraphic and seismic evidence from the Mesozoic successions of the Alps and western Iberia. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: Evidence from Land and Sea. Geological Society, London, Special Publications, 187, 429–452. Z ELT , C. A., S AIN , K., N AUMENKO , J. V. & S AWYER , D. S. 2003. Assessment of crustal velocity models using seismic refraction and reflection tomography. Geophysical Journal International, 153, 609– 626.
Early Cretaceous motion of Flemish Cap with respect to North America: implications on the formation of Orphan Basin and SE Flemish Cap –Galicia Bank conjugate margins J.-C. SIBUET1, S. P. SRIVASTAVA2, M. ENACHESCU3 & G. D. KARNER4,5 1
Ifremer Centre de Brest, B.P. 70, 29280 Plouzane´ Cedex, France (e-mail:
[email protected])
2
Natural Resources Canada, Geological Survey of Canada, Bedford Institute of Oceanography, P.O. Box 1006, Dartmouth, N.S., Canada B2Y 4A2 3
Memorial University of Newfoundland, St John’s, Newfoundland, Canada 4
Lamont-Doherty Earth Observatory, Palisades, NY 10964, USA
5
Present address: ExxonMobil Upstream Research Company, Mail Stop URC-URC-S169A, P.O. Box 2189, Houston, TX 77252-2189, USA Abstract: Bouguer gravity anomalies together with deep seismic reflection and magnetic data on both sides of the North Atlantic are used to locate the hinge zones of the Flemish Cap and Galicia Bank within the Iberian and North American plates, regions across which there were abrupt changes in lithospheric extension. The characteristic shape and alignment of these hinge zones suggest that they were conjugate features generated during chrons M25– M0 (Late Jurassic– Early Aptian) around a distally located Euler pole of rotation. Rifting between Iberia and North America involved these two larger plates and the two smaller microplates – the Flemish Cap and Galicia Bank microplates. The motion of the microplates, which were adjacent to Eurasia, was much more complex than those of the larger plates. The motion between the microplates from chron M25 or older to chron M0 was complicated by the fact that they remained attached to each other for most of the time when regions to the south were rifting apart. As a result, continental regions landward of these segments were subjected to extension that created the Orphan and Flemish Pass basins on the North American side and the Galicia Interior Basin on the Iberian side. By comparing the hinge zones delineated off Galicia Bank and Flemish Cap using the Bouguer anomalies, we were able to infer that Flemish Cap rotated approximately 438 relative to Galicia Bank and Iberia, and moved 200 –300 km SE with respect to North America. Such motions of Flemish Cap and Galicia Bank agree remarkably well with extensional episodes deduced from industry multichannel seismic reflection data acquired in the Orphan Basin. Normal fault orientations identified in the West Orphan Basin trend N0208 and are approximately perpendicular to the flow lines of our proposed Flemish Cap–North American motion during the M25–M0 period, which provides an independent constraint on our proposed kinematic model. Therefore, extensional events affected not only the Galicia Bank–Flemish Cap conjugate margins but also the Galicia Interior and Orphan Basins, and need to be taken into account in any assessment of the geological development of the Iberian and North American continental margins.
During the InterMARGINS Workshop meeting ‘Modelling the Extensional Deformation of the Lithosphere’ (IMEDL) held in Pontresina in 2004, a benchmark exercise was proposed to model the crustal and mantle rifting processes between the western Galicia Bank and the SE Flemish Cap conjugate margins where high-quality seismic reflection and refraction data exist (SCREECH Line 1: Funck et al. 2003; Hopper et al. 2004) and ISE Line 17 (Pe´rez-Gussinye´ et al. 2003; Henning et al. 2004). During the meeting it became apparent that additional boundary conditions and constraints were needed to correctly characterize the known extension across
the Iberian–Newfoundland margin. The aim of this paper is to provide some of these conditions and constraints, namely a knowledge of the initial positions of plates and microplates before their separation by sea-floor spreading processes. The positions of Flemish Cap and Galicia Bank are well constrained at chron M0 (Fig. 1a) because of the identification of correlatable sea-floor magnetic anomalies in both the Bay of Biscay (Sibuet et al. 2004b) and in the North Atlantic (Srivastava et al. 2000). The resulting Euler pole position was constrained by the geometry of the triple junction that had remained at the mouth of the Bay of
From: KARNER , G. D., MANATSCHAL , G. & PINHEIRO , L. M. (eds) Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, London, Special Publications, 282, 63–76. DOI: 10.1144/SP282.4 0305-8719/07/$15.00 # The Geological Society of London 2007.
64
J.-C. SIBUET ET AL.
Fig. 1. (a) Reconstruction of the North Atlantic between Africa (AF), Iberia (IB) and North America (NA) at chron M0 with respect to Eurasia (EU) as given by Srivastava et al. (2000) (IB– EU 43.858N, 5.838W, 244.768; AF– EU 43.618N, 6.938W, 240.898; NA –EU 69.678N, 154.268E, 23.178). M0 picks west of IB (crosses) and east of
EARLY CRETACEOUS ROTATION OF FLEMISH CAP
Biscay during this period and by palaeomagnetic declination data on the stable Iberia plate. However, similar constraints are not available for earlier time periods. Based on palaeomagnetic data, Srivastava & Verhoef (1992) derived the positions of the Eurasian (EU) and Iberian (IB) plates relative to North America (NA) for chron M25. For consistency with previous workers in this region, we have adopted the Kent & Gradstein (1986) timescale in this paper. The time of chron M25 (156 Ma, Late Jurassic) thus corresponds to the beginning of the Late Jurassic–Early Cretaceous rifting phase. Although M25 magnetic lineations have not been identified between the EU, IB and NA plates, for convenience we will use ‘M25’ either as the time of chron M25 or as the beginning of the Late Jurassic– Early Cretaceous phase of extension. When the positions of the three plates (EU, IB and NA) are reconstructed there is significant overlap. To remove the overlap the plates need to be translated, rotated and unstretched by an amount equal to the amount of overlap between them. Thus, the reconfigured plates at chron M25 time suggest that the Flemish Cap (FC) would lie NW of its present position. Similarly, the Galicia Bank (GB) would lie eastward of its present position at this same time. In deriving the relative positions of the Flemish Cap and Galicia Bank, Srivastava & Verhoef (1992) assumed that an equal amount of extension took place within all the basins lying significantly landward of the hinge zones. The detailed seismic reflection and refraction data recently acquired in this region suggest that such an assumption is not valid. To derive more accurate restored positions of these basins and plates, we have combined an interpretation of Bouguer gravity imaging with interpretations of commercial seismic reflection data from the various adjacent basins.
Quantifying the amount of extension between the Eurasian, North American and Iberian plates at M25 – M0 time Figure 1b shows the amount of crustal extension calculated during the M25–M0 period (156 – 118 Ma, Late Jurassic –Early Aptian) using parameters
65
from Srivastava & Verhoef (1992) for chron M25 and from Srivastava et al. (2000) for chron M0. As the M25–M0 stage poles summarizing the rotation for EU –NA, IB –EU and IB –NA are distant from our study area, the amount of extension between each plate pair will not vary significantly along strike, at least within the region shown in Figure 1b. This gives an estimated extension of 350 km between the IB and EU plates, 130 km between the EU and NA plates, and 450 km between the IB and NA plates. In this paper we use the expression ‘transitional crust’ in the traditional sense of the expression; namely, as the transition between distal thinned continental crust and the first clearly recognized oceanic crust, even though we appreciate that the transitional crust between the margins of the Iberia Abyssal Plain and the Newfoundland Basin consists mostly of exhumed serpentinized continental mantle with some decompression melting products (e.g. Tucholke et al. 2007). Inversion of magnetic data (Sibuet et al. 2007) shows that magnetic sources are N–S-trending horizontal bodies located within the highly serpentinized upper crust rather than in the lower crust (cf. Russell & Whitmarsh 2003) and that the serpentinization process is able to produce magnetic lineations in an analogous way as that for the oceanic crust. Thus, within transition zones, sequences of magnetic anomalies can provide constraints concerning the timing of emplacement of crust and continental mantle within an extensional regime but not on the nature of that material. During the M25–M0 period, extension between the EU and IB plates resulted in the formation of the northern and southern Bay of Biscay continental margins, the Armorican Basin transitional crust and the subsequent formation of oceanic crust (or possibly transitional crust) in the Bay of Biscay during M3–M0 period (Montadert et al. 1979; Sibuet et al. 2004b). During this same period, extension also resulted in fault reactivation in the Paris Basin, the Celtic Sea and the English Channel (Ziegler 1988), south of the Ebro Variscan massif (Verge´s & Garcı´a-Senz 2001) in the Iberia, Catalan and Basque–Cantabrian basins (Casas 1993; Salas et al. 2001), in the Organya` Basin (southern Pyrenees)
Fig. 1. (Continued ) NA (white circles) are from Srivastava et al. (2000); north of IB (crosses) and south of EU (white circles) from Sibuet et al. (2004b). A, B and C are conjugate points of the fossil triple junction trajectories at this time. FC, Flemish Cap; FPB, Flemish Pass Basin; GB, Grand Banks; GS, Goban Spur; GB, Galicia Bank; JAB, Jeanne d’Arc Basin; PSB, Porcupine–Seabight Basin; WB, Whale Basin. (b) The calculated amount of extension between M25 and M0 (Late Jurassic–Early Aptian) is shown in grey on the M0 reconstruction. The parameters used are those from Srivastava & Verhoef (1992) for M25 and from Srivastava et al. (2000) for M0. M0 picks west and north of IB (crosses and open dots, respectively) are from Srivastava et al. (2000) and Sibuet et al. (2004b). Large transparent arrows show the M25–M0 directions of plate motions. Hachured lines show continental shelf basins and continental margins where extension occurred during that period. Small black arrows indicate rotation of Galicia Bank and Flemish Cap during this period. BCb, Basque–Cantabrian basins; Cb, Catalan basins; Ib, Iberian Basins; Mb, Maule´on Arzacq basins; NPF, North Pyrenean Fault; Ob, Organya` Basin; Pb, Parentis Basin.
66
J.-C. SIBUET ET AL.
(Bera`stegui et al. 1990; Verge´s & Garcı´a-Senz 2001), and north of the Pyrenees in the Parentis, Maule´on and Arzacq basins (Bois et al. 1997; Masse 1997; Winnock 1971) (Fig. 1b). Between the EU and NA plates, continental rifting was still ongoing at chron M0. Transitional or oceanic crusts were not formed during this period but extension occurred on their continental margins, in the Porcupine and Seabight basins, and in the Rockall Trough and on the adjacent shelf areas. Between the IB and NA plates, extension gave rise to the development of conjugate margins, to the formation of a transitional crust, and then to oceanic crust since chrons M5–M3 (Russell & Whitmarsh 2003) or more probably since the Aptian–Albian time as implied from drilling results at sites 1067 (Whitmarsh & Sawyer 1996) and 1277 (Shipboard Scientific Party 2004) at time of chrons M2 and M1, respectively (Sibuet et al. 2007; Tucholke et al. 2007). For example, extension was recorded not only in the Salar and Carson basins located on the continental slope of the Grand Banks but also across the Grand Banks, in the Flemish Pass and in the Jeanne d’Arc, Horseshoe and Whale basins (Fig. 1) from Late Jurassic to Early Cretaceous. In general, this extension was significantly smaller than the earlier Late Triassic–Early Jurassic extensional phase in the region. Between the northern Newfoundland Basin (NNFB) and Iberia Abyssal Plain (IAP) conjugate margins, a transitional crust about 300 km wide was formed as evidenced from wide-angle seismic refraction and drilling data both in the IAP (Chian et al. 1999; Dean et al. 2000; Whitmarsh et al. 2001) and in the NNFB (Lau et al. 2003; Shipboard Scientific Party 2004). Perhaps even a wider transitional domain exists along the conjugate Tagus Abyssal Plain and southern Newfoundland Basin margins, but not to such a large extent between Galicia Bank and the SE Flemish Cap (Whitmarsh et al. 1996). This is owing to the simultaneous relative motion of Flemish Cap and Galicia Bank with respect to their adjacent NA and IB plates (Fig. 1b), as previously suggested by Le Pichon et al. (1977), Srivastava & Verhoef (1992), Srivastava et al. (2000) and Sibuet et al. (2004b). In this paper, we use new geophysical constraints in order to quantify the FC– NA motion, which has occurred since Late Jurassic, and compare the resulting motions with the direction of structural trends and the amount of extension compiled by Enachescu et al. (2004a– c) from industry seismic reflection data collected in the Orphan Basin located NW of FC.
North American and Iberian hinge zones Le Pichon & Sibuet (1981) demonstrated that if the thinning mechanism of continental margins is
basically pure shear, then the location of the plate boundary before stretching approximately follows a present-day isobath at a distance equal to the width of unstretched crust; for example, the 3000 m-isobath on the northern Biscay margin. In other words, if a is the width of the unstretched continental crust on a plate, which has been thinned during extension, then the isobath defining the prestretched position of the plate will be located at a distance a from the edge of the undeformed continental crust and will correspond to the maximum gradient in the Bouguer gravity anomaly (Fig. 2). We call these lines or boundaries ‘hinge zones’ in Figure 2, across which abrupt changes in lithospheric extension occur. Thus, if the extension mechanism is pure shear, the reconstruction of the hinge zones would correspond approximately to the restoration in space of the two portions of initial continental crusts before extension. The locations of these boundaries across the margins of NA and IB plates were deciphered from the Bouguer gravity anomalies determined from the 11 minute global free-air gravity grid of the world’s ocean generated by Sandwell & Smith (1994) and the 11 minute GEBCO topographic grid (IOC, IHO and BODC 2003) using a sediment –water contrast of 2200–1030 kg m23 (Karner et al. 1997). To decide what density contrast to use, for example 2200 v. 2800 kg m23, is not straightforward. This is because the edge-effect free-air gravity anomaly of a margin is partitioned between the gravity effects across the hinge zones and across the shelf-break, as shown by Karner & Watts (1982). These effects can play important roles in determining the location of hinge zones, depending on whether the margin is sediment rich or starved. Using 2200, 2400 and 2800 kg m23 as the reduction density made little difference in the location of the maximum gradients in the Bouguer anomaly maps across the NA and IB margins. That is, the gravity effect of the hinge zone is the dominant contribution to the gravity. Thus, we use a sediment –water reduction density of 2200– 1030 kg m23 to show our reconstruction of the NA and IB margins at chron M0. On each side of the IB–NA M0 boundary, the Bouguer values are significantly different, implying that the two crusts are of different thicknesses. This is confirmed by the depths of unloaded basement, which is 500–800 m deeper west of Galicia Bank than east of Flemish Cap, and by different mean values of heat flow data collected on both sides of the ocean (Louden et al. 1997; Louden & Lau 2004). Beneath the NA and IB continental slopes, the hinge zones roughly follow the 280 mGal isolines (white continuous lines in Fig. 3). The existence of such a crustal gravity gradient is the consequence of the oceanward thinning of the continental crust.
EARLY CRETACEOUS ROTATION OF FLEMISH CAP
67
location of the initial rupture
continental crust before extension a
HZ
HZ
thinned continental crust
undeformed continental crust
mantle a
200
thinned continental crust
undeformed continental crust
transitional or oceanic crust
crustal Bouguer anomaly
0 –200
HZ
HZ
mGal
Fig. 2. Pure shear model of continental crust extension. HZ, hinge zone; a, width of the deformed continental crust. The maximum Bouguer gravity anomaly gradient occurs at the vertical of the hinge zone located at distance a from the undeformed continental crust. Fitting the hinge zones identified on the Bouguer gravity map brings the extended plates to their prerift positions.
However, the trends of the hinge zones do not necessarily follow the bathymetric trends and gradients because bathymetric contours are influenced by local post-rift sedimentary processes. For example, along the IAP –NNFB conjugate margin transect, deep bathymetric trends (within the green colour of Fig. 3 at 448N; 138W) are 458 oblique to the gravity trends. Concerning the segment of margin located immediately to the south, which corresponds to the Estramadura Spur and the bathymetric ridge located between the IAP and the Tagus Abyssal Plain, the hinge zone is clearly defined by the Bouguer gravity anomalies. It is for this reason that plate tectonic reconstructions based solely on the fit of bathymetric trends between IB and NA are not convincing. Interestingly, south of the Galicia Bank and Flemish Cap, the hinge zones on the NA and IB plates have similar geometries. We suggest that they are conjugate features whose positions can be described by a distally located pole of rotation,
primarily because the distance between these zones is the same from south of Galicia Bank and Flemish Cap to the Newfoundland –Gibraltar fracture zone. Based on the Bouguer gravity and bathymetric data, flow lines can be inferred between the NA and IB conjugate segments (thick black lines in Fig. 3). Consequently, if we assume that the beginning of the rifting phase between IB and NA plates started in the Late Jurassic (close to M25 time), then the distance between the two hinge zones in the M0 reconstruction would correspond approximately to the M25–M0 displacement between the NA and IB plates. South of Galicia Bank and Flemish Cap, the M0 trend, as defined by the M0 picks (Fig. 3), is approximately linear, while the trends of the hinge zones are not. Such a modification of geometry occurred some time close to the onset of the emplacement of transitional crust. As commonly observed, the hinge-zone boundary evolves from an irregular continental rupture to a rather linear mid-oceanic rift system.
68
J.-C. SIBUET ET AL. 22°W
20°W
18°W
16°W
14°W
12°W
10°W
8°W
RT PB
52°N
IRELAND
OK
WOB
PSB 50°N
EOB GS
JAB FPB
GRAND BANKS
48°N
FC
M25-M0 FC/NA
BB
M3
NNFB
M0
GB
M20
46°N
GIB
M25-M0 GB/EU
M4 M11
IAPM17
M17 M15 M11
Ne
M4 M3
ou wf
SNFB nd la nd -G
M0
TAP
44°N
M11 M17 M20
42°N
ib ra lt a rF Z Reconstruction at chron M0 (118 Ma, Early Aptian)
IBERIA 40°N
38°N mGal –200
–150
–100
–50
0
50
100
150
Fig. 3. Plate reconstruction of the North Atlantic and Bay of Biscay (BB) at chron M0 (118 Ma) keeping Eurasia (EU) fixed. Bouguer gravity anomalies are in colour. The 200 m-interval bathymetric contours shown in black are extracted from the bathymetric map of the NE Atlantic (Sibuet et al. 2004a) and from Sandwell & Smith (1994) for the rest of the map. Magnetic anomaly picks of the M sequence between Iberia and North America are from Srivastava et al. (1988, 2000) and Sibuet et al. (2004b). Black dots are IPOD and ODP holes. FC, Flemish Cap; FPB, Flemish Pass Basin; GB, Galicia Bank; GS, Goban Spur; IAP, Iberian Abyssal Plain; JAB, Jeanne d’Arc Basin;
EARLY CRETACEOUS ROTATION OF FLEMISH CAP
Thus, for the same conjugate segment, the distance between chron M0 and the two hinge zones could vary along strike. For example, this distance is constant south of the IAP–NNFB segment but is different to the north of this segment. In conclusion, south of Galicia Bank and Flemish Cap, our proposed IB–NA M25–M0 kinematic pattern, based primarily on Bouguer gravity anomalies, seems to be coherent.
Flemish Cap and Galicia Bank motions The relative motion between the GB and FC conjugate segments is a more complex problem as it also involves additional interactions with the neighbouring larger plates. While the NA and IB plates were being rifted apart in the south, the FC and GB plates remained attached to each other. This required the crust landward of these features to undergo stretching, forming extensive and deep sedimentary basins in these regions, namely the Orphan and Galicia Interior basins. Ultimately, the NA and IB blocks were rifted apart, resulting in the formation of transitional and oceanic crusts between them. From the Bouguer gravity anomaly (Fig. 3) and using the same criteria as before, we have delineated the locations of the main hinge zones associated with the Galicia Interior and Orphan basins. Additional hinge zones have been delineated along the continental margins west of Galicia Bank and east of Flemish Cap. However, the hinge zone located on the eastern side of Galicia Bank cannot be clearly defined because the Moho gradient is too small (Pe´rezGussinye´ et al. 2003). In general, the motions between various plates, which for simplicity can be regarded as microplates, can only be inferred from geological constraints because there are no oceanic magnetic anomalies between these plates to constrain their motion. However, by using four independent geophysical constraints, we can uniquely define the FC– NA motion during the M25–M0 period. These constraints are as follows. (1) Because the Galicia Interior Basin seems to have approximately the same width along its whole length, we assume that the Galicia Bank is acting as a microplate that did not rotate with respect to IB (or that the GB– IB pole of rotation was distally located) during Late Jurassic –Valanginian extension (Murillas et al. 1990). Assuming that the Galicia Interior Basin was stretched between the Iberia continental
69
shelf and Galicia Bank sensu stricto, we have calculated 105 km of crustal extension (yellow arrows in Fig. 3) from the geometry of the crust as seen in seismic reflection and refraction data (Murillas et al. 1990; Pe´rez-Gussinye´ et al. 2003). (2) Because the hinge zones east of the SE Flemish Cap and west of Galicia Bank (white lines in Fig. 3) have approximately the same strike length, but make a 438 angle between them, it suggests that the FC plate has rotated by this amount relative to GB (or IB). Prior to Late Jurassic–Early Cretaceous rifting, the hinge zones were parallel, similar to the hinge zones between NA and IB to the south. (3) Such a motion of the FC and GB plates would require that the total extent of the crust formed or thinned between them should be about the same as between the hinge zones to the south (440 km). This would suggest the FC rotated by 438 relative to the GB (or IB) and moved by 335 km (440 –105 km) with respect to the GB. (4) The position of Flemish Cap at chron M25 is constrained in the north–south direction by the geomorphologies and geometries of the Orphan Basin and the Orphan Knoll. Thus, the resulting motion of FC with respect to NA during the M25–M0 period can be described by the following parameters of rotation (pole at 46.178N, 49.098W, 2438). The location of FC with respect to NA can thus be uniquely defined at the time of chrons M0 and M25. The consequences of such a motion concerning the timing, amount and direction of extension in the Orphan Basin can be determined since chron M25 and compared with independent geological constraints.
Reconstruction of Flemish Cap with respect to North America at chron M25 time Figure 3 shows the initial position of Flemish Cap (dashed black lines) with respect to North America at chron M25 and its M25– M0 motion (yellow arrows). At chron M25, Flemish Cap was located SW of Orphan Knoll and between two large, east –west negative magnetic anomalies (Fig. 4), which are not connected to the well-known curved magnetic anomalies located farther west on the continental platform. In addition, the northern
Fig. 3. (Continued) NNFB and SNFB, north and south Newfoundland basins; GIB, Galicia Interior Basin; EOB and WOB, East and West Orphan basins; OK, Orphan Knoll; PB, Porcupine Bank; PSB, Porcupine –Seabight Basin; RT, Rockall Trough; TAP, Tagus Abyssal Plain. Explanation of poles of rotation and other symbols as in Figure 1. Solid pink lines are boundaries of basins. The large continuous white lines represent major gradients of the Bouguer gravity and the large continuous black lines the flow lines between IB and EU from Late Jurassic to Early Aptian (approximately M25–M0) deduced from the morphology and crustal Bouguer anomalies. Yellow arrows give the FC– NA and GB– IB motions.
70
J.-C. SIBUET ET AL.
22°W
20°W
18°W
16°W
14°W
12°W
10°W
8°W
RT PB
52°N
IRELAND
OK
WOB
PSB 50°N
EOB GS
JAB FPB
GRAND BANKS
48°N
FC
M25-M0 FC/NA
BB
M3
NNFB
M0
GB
M20
M25-M0 GB/EU
M4 M11
46°N
GIB
IAPM17
M17 M15 M11
44°N
Ne
M4 M3
u fo w
SNFB M0 TAP nd la nd -G ib ra lta rF Z Reconstruction
M11 M17 M20
42°N
IBERIA 40°N
at chron M0 (118 Ma, Early Aptian)
38°N nT
–250 –200 –150 –100 –50
0
50
100
150
200
250
300
Fig. 4. Plate reconstruction of the North Atlantic and Bay of Biscay at chron M0 (118 Ma) with EU kept fixed. Shown in colour are the magnetic anomalies extracted from Verhoef et al. (1996). The 200 m-bathymetric contours are from Sandwell & Smith (1994) and Sibuet et al. (2004a). Explanation of poles of rotation and symbols as in Figures 1 & 2.
EARLY CRETACEOUS ROTATION OF FLEMISH CAP
M25–M0 flow line closely matches the EU– NA plate boundary (Figs 3 & 4). It is thus apparent that our interpreted position of Flemish Cap at chron M25 does not violate geophysical data presented in recent studies (e.g. Enachescu et al. 2004a–c; Skogseid et al. 2004). Furthermore, the position of FC at M25 time and its motion with respect to NA during the M25–M0 period compares favourably with geological data obtained by the oil industry in the Orphan Basin, as described below. Another major implication is that FC is bounded to the NE and SE by major shear zones during its motion with respect to NA and EU. Thus, for example, the Flemish Pass Basin and the Beothuk Ridge (Fig. 5) should have been formed in a transtensional regime.
Formation of the Orphan Basin by the SE motion of Flemish Cap The compilation of over 30 000 km of seismic data acquired by Geophysical Seismic Incorporated (GSI) in the Orphan Basin shows the presence of two central rift zones (Enachescu et al. 2004a–c; Skogseid et al. 2004) characterized by thin, stretched continental crust without obvious volcanic intrusions or high velocity layers (7.2 km s21) at the base of the crust (Chian et al. 2001) (Fig. 5a). Seismic and stratigraphic relationships indicate that the first episode of rifting in this basin could be as old as Late Triassic–Early Jurassic, based on the rifting age established from seismic correlations and well data in the Jeanne d’Arc Basin (e.g. Enachescu et al. 2004a, c), creating a deep trough (East Orphan Basin) immediately landward of Flemish Cap–Orphan Knoll ridge. A second welldated rifting phase occurred from Late Jurassic – Early Cretaceous creating the West Orphan Basin (Fig. 5a), located NW of the East Orphan Basin, but also reactivating the sedimentary features of the East Orphan Basin (Enachescu et al. 2004a–c; Skogseid et al. 2004). Further to the west, a third rifting phase occurred during Late Cretaceous. The Orphan Basin was thus created during three rifting phases sequentially from east to west. Figure 5b shows an interpreted seismic profile, which traverses the entire Orphan Basin from Flemish Cap to the Bonavista platform. From east to west, it shows that the East Orphan Basin is limited to the west by the White Sail bounding fault (Fig. 5b). Above the Palaeozoic basement, Early Jurassic, Triassic or older sediments were deformed during the first Late Triassic –Early Jurassic rifting phase. Here the structural trends are oriented north –south. The second Late Jurassic – Early Cretaceous rifting phase is also recorded in the East Orphan Basin, but to a lesser degree
71
compared with rifting further to the west. This second rifting phase took place essentially in the eastern part of the West Orphan Basin and the oldest sediments are Early Cretaceous in age. The strike of tilted fault blocks shows a N0208 orientation. The third rifting phase occurred during the Late Cretaceous in the westernmost part of the West Orphan Basin. In the West Orphan Basin the basement corresponds to the top of prerift sedimentary units, which might be Palaeozoic, as indicated in Figure 5b. The trends of the main basement ridges and troughs (Enachescu et al. 2004a, c) are summarized in Figure 6, which shows that ridges correspond to N0208 linear tilted faults blocks (e.g. Chian et al. 2001). Relative positive magnetic anomalies are associated with the crest of these tilted blocks (Chian et al. 2001), the trend of which is perpendicular to the flow lines of our proposed FC–NA M25–M0 motion, suggesting that the West Orphan Basin was created by the FC –NA rotation and not by the NE motion of Orphan Knoll. In fact, Orphan Knoll is a portion of the Flemish Cap– Porcupine Bank arch feature existing before Late Jurassic, which was left behind at the beginning of FC rotation with respect to NA. One of the consequences of the Early Cretaceous FC microplate motion with respect to EU and NA is that shear zones would have developed on the NE and SW sides of Flemish Cap. A series of basement ridges, elongated and aligned in the NW–SE direction and located east of the Cumberland Ridge (Fig. 5), might be the topographic expression of this FC– NA motion. Similarly, a preliminary inspection of old industrial seismic reflection lines collected across the continental margin SE of Orphan Knoll (Fig. 5) shows elongated basement ridges parallel to the margins, which might be interpreted as the fossil trace of the FC–EU shear zone. For the M0 reconstruction, the East Orphan Basin is adjacent to the Porcupine–Seabight Basin and the West Orphan Basin is adjacent to Rockall Trough (Figs 3 & 6). Two phases of rifting are recorded in the Porcupine–Seabight Basin and are dated as Early Jurassic and Late Jurassic–Early Cretaceous (Dore´ et al. 1997; Tate et al. 1993), identical to that of the East Orphan Basin. For the Rockall Trough, the rifting phase is considered to be Late Jurassic–Early Cretaceous (Dore´ et al. 1997; Shannon et al. 1999), as for the eastern part of the West Orphan Basin. For the Labrador Sea Basin, located NW of Orphan Basin, the rifting phase is Late Cretaceous (Enachescu et al. 2004a), as is the rifting age for the westernmost part of the West Orphan Basin. The width of the Rockall Trough, being approximately the same width as the West Orphan Basin (perpendicularly to the M25–M0 flow lines) during the Late Jurassic–Early Cretaceous rifting
72
J.-C. SIBUET ET AL.
56°W
54°W
52°W
50°W
48°W
46°W
44°W
42°W 52°N
CGFZ
Orphan Knoll DSDP 111
J-57
W Orphan B. H-28
A
Newfoundland
B
C-99
Bonavista Platform Bo na vis t
lt au rF e v Do
50°N
E Orphan B .
White Sail F. L-11
Jeanne d'Arc B.
Flemish Pass B.
48°N I-78
M-71
. yF
Cumberland Ridge St. John's
r cu er M
aF .
E-63
G-70
O-23
Flemish Cap
C-60
Central Ridge Murre F.
46°N
So ut h
Wh ale
B.
ODP 1276 ODP 1277
Carso n B. Sala r B.
B. oe sh e rs Ho
Whale B.
Newfoundland Beothuk Ridge Basin
Ne wfo u nd land
44°N Seam ount s
a 42°N
A (west)
B (east)
Bonavista Platform C - 99 100
Pz Pz Pc Pc
v BtoanFa
av is
Pc Pc
ul t Fa
Early Early Cretaceous Cretaceous
ult
15 km
Pc Pc
600 km
Pz Pz
LateCretaceous Cretaceous Late
Fa Saaili l F hithe WW ite S
). m (k th p e D
400
Pz Pz
i st a
10
300
Tertiary Tertiary
Bon
5
200
Jurassic Jurassic Pz Pz Pc Pc l
Triassic or or older Pz Triassic Pz Pc Pc
Pc Pc
phh a ann KKnn o ol l l l F Fa ul t
0
Flemish Cap/ Orphan Knoll Ridge 500
East Orphan Basin
Pz Pz
older
OOr rp
0
West Orphan Basin
West
1:10 VE 0
10 20 30 40
b
Kilometers
East
SCHEMATIC GEOLOGICAL CROSS-SECTION ACROSS ORPHAN BASIN
Fig. 5. (a) Structural map of the Orphan Basin established from Geophysical Seismic Incorporated (GSI) seismic reflection data (modified from Enachescu et al. 2004a– c). Green-coloured features and thin black lines in the Orphan Basin are basement ridges and axes of depressions, respectively. Distribution and shapes of other basins (light–dark yellow) from Enachescu et al. (2004c). Basement platform in brown. CGFZ, Charlie-Gibbs Fracture Zone. Black dots are well locations with their names. Dashed lines are oceanic fracture zones lying in the prolongation of continental or thinned continental crust features. (b) Schematic geological cross-section of the Orphan Basin from the Flemish Cap to the Bonavista platform showing that basins become younger westward (Enachescu et al. 2004a– c). The location of the cross-section is defined in the upper panel.
EARLY CRETACEOUS ROTATION OF FLEMISH CAP
22°W
20°W
18°W
16°W
14°W
12°W
73
10°W
8°W
RT 52°N
WOB
PB
OK
PSB
EOB
Bo
White Sail F.
IRELAND 50°N
vis na F. ta
GS
rcu
Me
Jeanne d'Arc B.
ry
GRAND BANKS Murre F.
F.
FP 48°N
FC
M25-M0 FC/NA
B.
sesh
Ca rs
on
Hor
BB
. oe B
GB
NNFB
46°N
M25-M0 GB/EU
IAP SNFB
44°N
IBERIA
Fig. 6. Plate reconstruction of the North Atlantic Ocean and Bay of Biscay at chron M0 (118 Ma) with EU kept fixed. Explanation of poles of rotations and symbols are as in Figures 1 and 2. Basement trends (thick black lines are lows and thin black lines are highs) in the East (EOB) and West (WOB) Orphan basins from Enachescu et al. (2004a –c). Note that the basement trends are approximately perpendicular to the FC–NA motion. Thick dashed line shows the position of Flemish Cap rotated back to chron M25 time.
phase, suggests that the Porcupine–Seabight and East Orphan basins opened as a single system and remained adjacent to each other during the M25– M0 rifting period. The FC displacement along the flow lines during the M25–M0 period varies from 200 km in the south to 300 km along the EU boundary. The distance between our M25 position of Flemish Cap and the Bonavista platform further to the NW is 200– 250 km. If the region located between the M25 position of Flemish Cap and the Bonavista platform corresponds to the initial non-deformed continental crust, the formation of Orphan and Flemish Pass
basins corresponds to a mean stretching factor of approximately 2, which is similar to the 2–2.5 mean stretching factor given by J. Skogseid (pers. comm., 2005) for the entire Orphan Basin. The Late Triassic –Early Jurassic plate boundary was located along the adjacent Porcupine–Seabight and East Orphan basins. This boundary extended northward through a system of basins connected to the North Sea basins (Dore´ et al. 1997; Shannon et al. 1999). To the south, it extended through the Flemish Pass, Jeanne d’Arc, Horseshoe and Whale basins. During the Late Jurassic –Early Cretaceous rifting phase, this plate boundary
74
J.-C. SIBUET ET AL.
extended from the Vøring Basin, Møre Basin, Faeroe– Shetland Basin, Rockall Trough, the entire Orphan Basin, Flemish Pass, Jeanne d’Arc, Carson and Salar basins, and then mostly between the Iberia margin and the Grands Banks. Thus, the FC–NA motion was controlled by the propagation of rifting through the EU and NA plates, explaining the unexpected diachronous structure of the crust within the Orphan Basin.
Conclusions The main conclusions of this study are as follows. † Using Bouguer gravity anomalies, we have been able to define the hinge zones of the North American and Iberian plates at chron M25 time. From chron M25 to chron M0 time, a single far-field Euler pole of rotation can describe the relative motion between these plates, which includes regions north of the Newfoundland–Gibraltar Fracture Zone and south of Flemish Cap and Galicia Bank. Rifting between Iberia and North America involved not only these two large plates but also two small microplates (the Flemish Cap and Galicia Bank microplates). † The motion of these microplates, which were adjacent to Eurasia, was much more complex than for the larger plates. The motion between the Flemish Cap and Galicia Bank microplates from chron M25 time or older to chron M0 time was complicated owing to the fact that they remained attached to each other for most of the period when regions to the south were rifting apart, as suggested by Srivastava et al. (2000) and others (e.g. Le Pichon et al. 1977). As a result continental regions landward of these segments were subjected to extensional thinning that created the Orphan and Flemish Pass basins on the North American side and the Galicia Interior Basin on the Iberian side, as well as shear transtensional regimes NE and SW of Flemish Cap. † By delineating hinge zones using Bouguer gravity anomalies at the edges of the conjugate margins, we were able to infer that Flemish Cap rotated 438 relative to Galicia Bank and Iberia, and moved 200– 300km SE with respect to North America. † During the M25–M0 period, most of the Orphan and Flemish Pass basins were initiated. The flow lines describing the translational motion of Flemish Cap are orthogonal to the linear trends of the tilted fault blocks, which were mapped by the oil industry from a compilation of seismic reflection data in this region, giving independent support for the location of the FC–NA pole of rotation.
† New seismic reflection data collected by the oil industry in the Orphan Basin show two central rift zones characterized by highly stretched continental crust without volcanic intrusions at the base of crust. Seismic stratal and stratigraphic relationships indicate that the first episode of rifting could be as old as the Late Triassic – Early Jurassic in the East Orphan Basin. A second rifting phase occurred from Late Jurassic to Early Cretaceous giving rise to the West Orphan Basin but also reactivating earlier structures within the East Orphan Basin. A third phase of rifting occurred in the western part of the West Orphan Basin during the Late Cretaceous. † In our M0 reconstruction, the East and West Orphan basins are adjacent to the Porcupine – Seabight Basin and Rockall Trough, respectively. Two phases of rifting are suggested in the Porcupine–Seabight Basin: Late Triassic – Early Jurassic and Late Jurassic–Early Cretaceous, the same as for the East Orphan Basin. For the Rockall Trough, the rifting phase is Late Jurassic –Early Cretaceous, as it is for the West Orphan Basin. The width of the Rockall Trough, similar to the width of the West Orphan basin, and with structural trends perpendicular to the M25–M0 flow lines imply that the Porcupine–Seabight and East Orphan basins opened simultaneously as a single Basin and remained juxtaposed during the M25–M0 motion of FC– NA. Finally, the Late Cretaceous phase of extension recorded in the westernmost part of the West Orphan Basin is also observed in the south Labrador Basin, which is in the northwestward prolongation of the West Orphan Basin.
The GMT software package was used to make some of the figures (Wessel & Smith 1991). We acknowledge numerous fruitful discussions with W. Roest, J. Skogseid and P. Werner, as well as very helpful and constructive reviews provided by K. Louden, B. Tucholke and R. Whitmarsh. Geophysical Seismic Incorporated (GSI) of Calgary is acknowledged for the permission to use their Orphan Basin seismic data.
References B ERA` STEGUI , X., G ARCIA , J. M. & L OSANTOS , M. 1990. Structure and sedimentary evolution of the Organya` basin (Central South Pyrenean Unit) during the lower Cretaceous. Bulletin de la Socie´te´ Ge´ologique de France, 8, 251 –264. B OIS , C., P INET , B. & G ARIEL , O. 1997. The sedimentary cover along the ECORS Bay of Biscay deep seismic reflection profile. A comparison between the Parentis Basin and other European rifts. Bulletin de la Socie´te´ Ge´ologique de France, 171, 143– 165.
EARLY CRETACEOUS ROTATION OF FLEMISH CAP C ASAS , A. M. 1993. Tectonic inversion and basement thrusting in the Cameros Massif (Northern Spain). Geodinamica Acta, 6, 202–216. C HIAN , D., L OUDEN , K. E., M INSHULL , T. A. & W HITMARSH , R. B. 1999. Deep structure of the ocean–continent transition in the southern Iberia Abyssal Plain from seismic refraction profiles: 1. Ocean Drilling Program (Legs 149 and 173) transect. Journal of Geophysical Research, 104, B4, 7443–7462, doi:10.1029/1999JB900004. C HIAN , D., R EID , I. D. & J ACKSON , H. R. 2001. Crustal structure beneath Orphan Basin and implications for nonvolcanic continental rifting. Journal of Geophysical Research, 106, B6, 10,923–10,940. D EAN , S. M., M INSHULL , T. A., W HITMARSH , R. B. & L OUDEN , K. E. 2000. Deep structure of the ocean– continent transition in the southern Iberia Abyssal Plain from seismic refraction profiles: The IAM-9 transect at 408200 N. Journal of Geophysical Research, 105, B3, 5859– 5886, doi:10.1029/1999JB900301. D ORE´ , A. G., L UNDIN , E. R., F ICHLER , C. & O LESEN , O. 1997. Patterns of basement structure and reactivation along the NE Atlantic margin. Journal of the Geological Society, London, 154, 85– 92. E NACHESCU , M., H OGG , J. & M EYER , K. 2004a. East Orphan Basin, offshore Newfoundland and Labrador: A deep water super extended rift with potential petroleum system. In: CSPG Annual Convention, expanded abstract. E NACHESCU , M. ET AL . 2004b. Orphan Basin, offshore Newfoundland, Canada: Structural and tectonic framework, petroleum systems and exploration potential. In: 74th SEG Annual Meeting and Exposition, expanded abstract. E NACHESCU , M., M EYER , K. & H OGG , J. 2004c. East Orphan Basin: Structural setting and evolution with seismic and potential fields arguments. In: CSEG Annual Convention, expanded abstract. F UNCK , T., H OPPER , J. R., L ARSEN , H. C., L OUDEN , K. E., T UCHOLKE , B. E. & H OLBROOK , W. S. 2003. Crustal structure of the ocean–continent transition at Flemish Cap: Seismic refraction results. Journal of Geophysical Research, 108, B11, 2531, doi:10.1029/2003JB002434. H ENNING , A. T., S AWYER , D. S. & T EMPLETON , D. C. 2004. Exhumed upper mantle within the ocean–continent transition on the northern West Iberia margin: Evidence from prestack depth migration and total tectonic subsidence analyses. Journal of Geophysical Research, 109, B05103, doi:10.1029/2003JB002526. H OPPER , J. R., F UNCK , T. & T UCHOLKE , B. E. 2004. Continental breakup and the onset of ultraslow sea floor spreading off Flemish Cap on the Newfoundland rifted margin. Geology, 32, 93–96. IOC, IHO AND BODC. 2003. Centenary Edition of the GEBCO Digital Atlas. GEBCO Digital Atlas (GDA). British Oceanographic Data Centre, Liverpool. K ARNER , G. D. & W ATTS , A. B. 1982. On isostasy at Atlantic-type continental margins. Journal of Geophysical Research, 87, 2923– 2948. K ARNER , G. D., D RISCOLL , N. W., M C G INNIS , J. P., B RUMBAUGH , W. D. & C AMERON , N. R. 1997. Tectonic significance of syn-rift sediment packages across the Gabon-Cabinda continental margin. Marine and Petroleum Geology, 14, 973–1000.
75
K ENT , D. V. & G RADSTEIN , F. M. 1986. A Jurassic to Recent chronology. In: V OGT , P. R. & T UCHOLKE , B. E. (eds) The Geology of North America, Volume M, The Western North Atlantic Region. Geological Society of America, Boulder, CO, 45– 50. L AU , K. W. H., L OUDEN , K. E., F UNCK , T. ET AL . 2003. Seismic velocity model across the eastern Newfoundland (Grand Banks) continental margin. In: 38th Annual Meeting of the Geological Society of America, Abstracts with Programs, Northeastern Section. L E P ICHON , X. & S IBUET , J.-C. 1981. Passive margins: a model of formation. Journal of Geophysical Research, 86, 3708–3710. L E P ICHON , X., S IBUET , J.-C. & F RANCHETEAU , J. 1977. The fit of continent around the North Atlantic Ocean. Tectonophysics, 38, 169–209. L OUDEN , K. E. & L AU , H. 2004. Anomalous heat flow and basement depth in the Newfoundland Basin ocean–continent transition compared with the Iberia Abyssal Plain conjugate. Eos transactions, AGU, 85, 461. Washington, DC. L OUDEN , K. E., S IBUET , J.-C. & H ARMEGNIES , F. 1997. Variations in heat flow across the ocean–continent transition in the Iberia Abyssal Plain. Earth and Planetary Science Letters, 151, 233– 254. M ASSE , P. 1997. The early Cretaceous Parentis basin (France), a basin associated with a wrench fault. Bulletin de la Socie´te´ Ge´ologique de France, 171, 177–185. M ONTADERT , L., DE C HARPAL , O., R OBERTS , D., G UENNOC , P. & S IBUET , J.-C. 1979. Northeast Atlantic passive continental margins: Rifting and subsidence processes. In: T ALWANI , M., H AY , W. & R YAN , W. B. F. (eds) Deep Drilling Results in the Atlantic Ocean: Continental Margins and Palaeoenvironments. American Geophysical Union, Washington, DC, 154– 186. M URILLAS , J., M OUGENOT , D., B OILLOT , G., C OMAS , M. C., B ANDA , E. & M AUFFRET , A. 1990. Structure and evolution of the Galicia Interior basin (Atlantic western Iberian continental margin). Tectonophysics, 184, 297–319. P E´ REZ -G USSINYE´ , M., R ANERO , C. R. & R ESTON , T. J. 2003. Mechanisms of extension at nonvolcanic margins: Evidence from the Galicia Interior basin, west of Iberia. Journal of Geophysical Research, 108, B5, 2245, doi:10.1029/2001JB000901. R USSELL , S. M. & W HITMARSH , R. B. 2003. Magmatism at the west Iberia non-volcanic rifted continental margin: Evidence from analyses of magnetic anomalies. Geophysical Journal International, 154, 706– 730. S ALAS , R., G UIMERA , J., M AS , R., M ARTIN -C LOSAS , C., M ELE´ NDEZ , A. & A LONSO , A. 2001. Evolution of the Mesozoic Central Iberian Rift system and its Cainozoic inversion. In: Z IEGLER , P. A., C AVAZZA , W., R OBERTSON , A. H. F. & C RASQUIN -S OLEAU , S. (eds) Peri-Tethyan Rift/Wrench Basins and Passive Margins. Me´moires du Muse´um National d’Histoire Naturelle, Paris, 145– 185. S ANDWELL , D. T. & S MITH , W. H. F. 1994. New global marine gravity map/grid based on stacked ERS1, Geosat and Topex altimetry. Eos, 75, 321. S HANNON , P. M., J ACOB , A. W. B., O’ REILLY , B. M., H AUSER , F., R EADMAN , P. W. & M AKRIS , J. 1999.
76
J.-C. SIBUET ET AL.
Structural setting, geological development and basin modelling in the Rockall Trough. In: F LEET , A. J. & B OLDY , S. A. R. (eds) Petroleum Geology of NW Europe: Proceedings of the 5th Conference. Geological Society, London, 421–431. SHIPBOARD SCIENTIFIC PARTY . 2004. Leg 210 summary. In: T UCHOLKE , B. E., S IBUET , J.-C. & K LAUS , A. (eds) Proceedings of the Ocean Drilling Program, Initial Reports, 210. Ocean Drilling Program, College Station, TX, 1 –78. S IBUET , J.-C., M ONTI , S., L OUBRIEU , B., M AZE´ , J.-P. & S RIVASTAVA , S. 2004a. Carte bathyme´trique de l’Atlantique nord-est et du golfe de Gascogne: implications cine´matiques. Bulletin de la Socie´te´ Ge´ologique de France, 175, 429– 442. S IBUET , J.-C., S RIVASTAVA , S. & S PAKMAN , W. 2004b. Pyrenean orogeny and plate kinematics. Journal of Geophysical Research, B08104, doi:10.1029/2003 JB002514. S IBUET , J.-C., S RIVASTAVA , S. P. & M ANATSCHAL , G. 2007. Exhumed mantle forming transitional crust in the Newfoundland–Iberia rift and associated magnetic anomalies. Journal of Geophysical Research. (in press) S KOGSEID , J., B ARNWELL , A., A ARSETH , E. S., A LSGAARD , P. C., B RISEID , H. C. & Z WACH , C. 2004. Orphan basin: Multiple failed rifting during early opening of the north Atlantic. Eos, Transactions of the American Geophysical Union, 85, 17, Joint Assembly Supplement, Abstract T41B-03. S RIVASTAVA , S. & V ERHOEF , J. 1992. Evolution of Mesozoic sedimentary basins around the north Central Atlantic: a preliminary plate kinematic solution. In: P ARNELL , J. (ed.) Basins on the Atlantic Seaboard: Petroleum Geology, Sedimentology and Basin Evolution. Geological Society, London, Special Publications, 397–420. S RIVASTAVA , S., S IBUET , J.-C., C ANDE , S., R OEST , W. R. & R EID , I. R. 2000. Magnetic evidence for slow sea-floor spreading during the formation of the Newfoundland and Iberian margins. Earth and Planetary Science Letters, 182, 61– 76. S RIVASTAVA , S., V ERHOEF , J. & M ACNAB , R. 1988. Results from a detailed aeromagnetic survey across the northeast Newfoundland margin, Part 1: Spreading anomalies and relationship between magnetic anomalies and the ocean– continent boundary. Marine and Petroleum Geology, 5, 306–323.
T ATE , M., W HITE , N. & C ONROY , J. J. 1993. Lithospheric extension and magmatism in the Porcupine Basin west of Ireland. Journal of Geophysical Research, 98, B8, 13,905–13,924. T UCHOLKE , B. E., S AWYER , D. S. & S IBUET , J.-C. 2007. Breakup of the Newfoundland–Iberia rift. In: K ARNER , G. D., M ANATSCHAL , G. & P INHEIRO , L. M. (eds) Imaging Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, London, Special Publications, 282, 9–45. V ERGE´ S , J. & G ARCI´ A -S ENZ , J. 2001. Mesozoic evolution and Cainozoic inversion of the Pyrenean rift. In: Z IEGLER , P. A., C AVAZZA , W., R OBERTSON , A. H. F. & C RASQUIN -S OLEAU , S. (eds) Peri-Tethyan Rift/Wrench Basins and Passive Margins. Me´moires du Muse´um National d’Histoire Naturelle, Paris, 187–212. V ERHOEF , J., R OEST , W. R., M ACNAB , R. & A RKANI -A HMED , J. Members of the Project Team. 1996. Magnetic Anomalies of the Arctic and North Atlantic Oceans and Adjacent Land Areas. Geological Survey of Canada, Open File, 3125. W ESSEL , P. & S MITH , W. M. F. 1991. Free software helps map and display data. Eos, 72, 441– 446. W HITMARSH , R. B. & S AWYER , D. S. 1996. The ocean/ continent transition between the Iberia Abyssal Plain and continental-rifting to sea floor spreading processes. In: W HITMARSH , R. B., S AWYER , D. S. & K LAUS , A. (eds) Proceedings of the Ocean Drilling Program, Scientific Results. Ocean Drilling Program, College Station, TX, 713– 733. W HITMARSH , R. B., M ANATSCHAL , G. & M INSHULL , T. A. 2001. Evolution of magma-poor continental margins from rifting to sea floor spreading. Nature, 413, 150– 154. W HITMARSH , R. B., W HITE , R. S., H ORSEFIELD , S. J., S IBUET , J.-C., R ECQ , M. & L OUVEL , V. 1996. The ocean–continent boundary off the western continental margin of Iberia – III. Crustal structure west of Galicia Bank. Journal of Geophysical Research, 101, 28,291–28,313. W INNOCK , E. 1971. Ge´ologie succincte du bassin d’Aquitaine. In: D EBYSER , J., L E P ICHON , X. & M ONTADERT , L. (eds) Histoire Structurale du Golfe de Gascogne. Technip, Paris, 1 –30. Z IEGLER , P. A. 1988. Evolution of the Arctic– North Atlantic and the Western Tethys. AAPG Memoirs, 43. AAPG, Tulsa, OK.
The formation of non-volcanic rifted margins by the progressive extension of the lithosphere: the example of the West Iberian margin T. J. RESTON1,2 1
School of Geography, Earth and Environmental Sciences, University of Birmingham, Birmingham B15 2TT, UK 2
Previous address: IFM-GEOMAR, Wischhofstrasse 1-3, D24148 Kiel, Germany (e-mail:
[email protected]) Abstract: Non-volcanic margins such as the West Iberian margin exhibit certain characteristics, such as a deficit of synrift igneous rock, a zone of exhumed subcontinental mantle in the continent –ocean transition and an apparent extension discrepancy. These observations can be explained as a consequence of the progressive extension of the lithosphere above relatively cool mantle. The evolving rheological stratification of the lithosphere controls the style of extension at different lithospheric levels at different times; extension is probably heterogeneous at all stages, with lower crustal and upper mantle boudinage controlling the patterns of thinning and mantle upwelling early in the rift history, and complete crustal embrittlement and mantle serpentinization controlling the formation of late-stage detachment faults. Extension in the brittle crust is via multiple phases of faulting, with a general focusing of extension towards the incipient ocean. The lack of melt is explained by a combination of heterogeneous extension of the lower lithosphere and a cool subcontinental geotherm. The extension discrepancy may in places be controlled by depth-dependent stretching of the crust through lower crustal boudinage, but may also simply be the result of incomplete recognition of the entire polyphase faulting history. The latter seems to be the case for West Iberia. Evidence for all these processes can be found at the West Iberian rifted margins as well as those preserved and partially exposed in the Alps.
Rifted margins include two main end-members: those termed ‘volcanic rifted margins – VRMs’ where magmatism is much more voluminous than predicted by passive asthenospheric upwelling (e.g. White & McKenzie 1989), and those where magmatism is consistent or even less than the same predictions. The latter are termed ‘non-volcanic rifted margins – NVRMs’ to emphasize the contrast with the VRMs: the name does not exclude the presence of minor amounts of magmatic activity. The NVRMs are typified by the North Biscay (Le Pichon & Barbier 1987), South Australian (Sayers et al. 2001) and the West Iberian margins (Boillot & Winterer 1988; Krawczyk et al. 1996; Whitmarsh et al. 1996), which share a number of common characteristics (Fig. 1): † low –moderate sediment accumulation rates; † extreme crustal thinning, increasing towards the ocean; † rotated fault blocks – however, at the featheredge of the continent the amount of extension that can be inferred from the geometry of these faults is less than that indicated by the crustal thinning and/or observed: this is the so-called extension discrepancy;
† presence, in places, of a detachment fault at the base of the fault blocks; † little evidence for voluminous synrift magmatism and in general less magmatism than predicted by standard extension models; † the presence of a broad zone of partially serpentinized mantle (Boillot & Winterer 1988; Krawczyk et al. 1996; Pickup et al. 1996; Whitmarsh et al. 1996), both occurring beneath the highly thinned and faulted continental crust of the continental rise region, and as a zone of exhumed continental mantle (ZECM: Whitmarsh et al. 2001), now largely buried by post-rift sediments. I suggest that such margins are the logical result of progressive extension of cool, sediment-starved continental lithosphere away from hotspots. I focus first on the expected evolution before considering the application of this to the West Iberian margin in particular. The key factors are the occurrence of multiple phases of faulting and the rheological evolution of the lithosphere and the way this controls the four-dimensional (4D) distribution of deformation within the lithosphere and the serpentinization of the uppermost mantle.
From: KARNER , G. D., MANATSCHAL , G. & PINHEIRO , L. M. (eds) Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, London, Special Publications, 282, 77–110. DOI: 10.1144/SP282.5 0305-8719/07/$15.00 # The Geological Society of London 2007.
897
1070 898
(ODP Legs 149 and 173)
continental crust blocks within serpentinized mantle? mounded layer, Vp ~ 4.6 km/s - serpentinite breccia partially serpentinized peridotite - Vp 5 – 8 km/s
continental crust (inc. prerift)
50 km
901
1065
1067 900 1068
1069
TWT [s] 11
7
6 TWT [s] 10
Fig. 1. Cross-sections at approximately true scale of typical non-volcanic margins, showing the presence of a wide zone of unroofed mantle rocks and the presence of local detachments. Fault block structure is invariably incompletely resolved.
sediments
oceanic crust
SW Greenland margin (Chian & Louden 1994) - no vertical exageration
Ligurian Margin (Alps) Froitzheim + Manatschal 1996
11
7
899
S Australia margin (Sayers et al. 2001)
W. Iberia
78 T. J. RESTON
RIFTED MARGINS BY PROGRESSIVE EXTENSION
Terminology In this paper, I will discuss various aspects of margin evolution, including results from seismic imaging, from wide-angle studies of crustal structure, numerical, thermal and rheological modelling, and structural geology. As the terminology between the different fields is commonly different, it is worth defining some of the terms to be used: it is important that the terminology used is appropriate, not misleading and not too specific for general application. In particular, the description of the different rheologies of the lithosphere can be a source of confusion: whereas structural geologists commonly refer to brittle and plastic deformation (Rutter 1986), meaning failure by faulting (cracking plus sliding) and failure by intracrystalline plastic creep, modellers commonly refer to the same as ‘plastic’ and ‘viscous’, respectively, referring to the stress –strain relationship of faults (no deformation until the yield stress reached) and the ‘creeping’ deep viscous crust. The term ‘plastic’ is used in two contradictory ways by different groups. The use of the term ‘viscous’ for deep crustal deformation may also be a simplification as it implies that the exponent in the stress–strain rate equation is 1, which is not the case for most flow laws. The term ‘flow’ also carries too much baggage in terms of large-scale channel flow to be used to describe rheology, and the term ‘ductile’ also has its critics (e.g. Rutter 1986) who point out that ductile means distributed and includes brittle deformation by cataclastic flow. In this paper I will refer to failure along faults (cracking and sliding according to a Mohr–Coulomb criteria) as brittle, and to deformation by temperature- and lithologydependent intracrystalline plastic creep as ‘creeping’. Some authors use the term ‘stretching factor’ only to imply deformation by pure shear (e.g. Davis & Kusznir 2004). This I consider unhelpful as it may be difficult to determine accurately the exact mode of deformation, whereas it is generally simpler to measure crustal thinning and/or extension. Others describe whole crustal stretching factors, upper and lower crustal stretching factors, and mantle stretching factors. I will use the term ‘extension factor’ to describe the degree of thinning and extension of the whole lithosphere, and, where specified, of the various portions of that lithosphere, regardless of whether this occurs by ‘pure’ pure shear or by some more realistic mixed mode extension. Finally, as will become clear in the paper, the term continent–ocean transition (COT) both describes the process and the resulting structure, and should be preferred to other terms (ocean continent transition – OCT; continent–ocean boundary – COB; ZECM – zone of exhumed continental
79
mantle) which, respectively, have the wrong sense of transition, imply a sharp boundary or which may be too specific for general use.
West Iberia margin The West Iberian margin is the best-studied nonvolcanic rifted margin, having been the subject of part of a Deep Sea Drilling Project (DSDP) leg, three dedicated Ocean Drilling Program (ODP) legs, as well as numerous seismic experiments and submersible diving. Two main segments have been studied (Fig. 2); the Deep Galicia margin and the southern Iberia Abyssal Plain (IAP) margin (Fig. 1).
Deep Galicia margin The northernmost segment of the West Iberian rifted margin is the Deep Galicia margin located to the west of the Galicia Bank (Fig. 2) and conjugate to the Flemish Cap margin (Fig. 3). Comparison between the two conjugate margins shows that whereas crustal thickness reaches ‘normal values’ within approximately 100 km of the COT at the Flemish Cap margin, at the west Galicia margin, the Galicia Bank 100 km east of the COT is still characterized by relatively thin crust (15– 20 km) and is flanked by the Galicia Interior Basin to the east and the Deep Galicia margin to the west. The latter is characterized by well-defined tilted fault blocks of continental material that, west of approximately 128150 W (ODP Site 639), decrease abruptly in size and become underlain by the bright S reflection (Fig. 3). Above S, the amount of extension that can be deduced from the geometry of these fault blocks is considerably less that that implied by the degree of subsidence (Sibuet 1992) and of crustal thinning (Reston 2005). S is thought to be both a detachment fault and a tectonic CMB, forming the boundary between the tilted blocks and the underlying partially serpentinized mantle (Reston et al. 1996). Mantle rocks crop out at the sea floor 25 km to the west of the last sampled continental fault block (other probable continental blocks lie between), and can be traced along strike as a partially buried ‘peridotite ridge’. Synrift igneous rocks have not been sampled within the main drilling transect here, although possible synrift lower crustal gabbros have been sampled by submersible further to the north. Drilling during ODP Leg 103, as well as sampling by submersible and by dredging, has shown that the tilted fault blocks are capped by a deep-water clastic sequence, rather than true prerift sediments (e.g. shallow-water carbonates). The implications of this stratigraphy are discussed at the end of this paper.
80
T. J. RESTON
Fig. 2. Map of the West Iberian margin showing the location of the DGM and IAP margins, of ODP holes and of profiles discussed in the text.
join made at peridotite ridge Flemish Cap margin - crust thinning oceanward from ~30 km thick
10 km
12°15'W
serpentinized mantle (base gradational)
Galicia Bank margin - crust thinning west from ~15 km thick CMB
CMB
20 km
W
2400 shotpoint
2600
2800
ODP 639
E
Dive 11
6
6
8
8
S CMB
10
10
serpentinized mantle km 5 km base of postrift section
km early generations of faults
latest generation of faults, including S
base of late synrift (Aptian ?)
base of early synrift - Valanginian - Barremian?
Fig. 3. Deep Galicia margin as part of conjugate pair with Flemish Cap margin. Top: generalized crustal sections at 2:1 vertical exaggeration (after Funck et al. 2003 and Zelt et al. 2003) showing a contrasting crustal thickness of Flemish cap and Galicia Bank. The S reflector in bold may continue as CMB on the Flemish Cap side. Below: Prestack depth migration of profile GP101 shown at no vertical exaggeration (location approximately marked by box), showing latest fault blocks and tentative identification of earlier faults (short dashes). ODP Site 639 straddles 128150 W.
RIFTED MARGINS BY PROGRESSIVE EXTENSION
Iberia Abyssal Plain The southern Iberian Abyssal Plain (IAP) margin has been the subject of two ODP drilling legs (149 and 173) as well as numerous seismic experiments. The basement geology across the feather edge of continental crust of the IAP is well constrained by the presence of several ODP boreholes (Figs 1, 2 & 4). Moving from the continent towards the ocean, Sites 901 and 1065 sampled the top of two tilted fault blocks and recovered shallow-water clastics, believed to indicate the presence of continental basement underneath. Sites 1067, 900 and 1068 sampled the same block and recovered, respectively, anorthosites, amphibolites and tonalites (1067), sheared mafic granulites (900) and serpentinized peridotites (1068). These are interpreted to represent lower crustal rocks (1067 and 900) and the subcrustal mantle (1068). The H detachment (Krawczyk et al. 1996) appears to intersect top basement near Site 900, so that the CMB here might be a tectonic boundary similar to that inferred at the West Galicia margin further north in the region of S. Site 1069 sampled a fault slice that appears to be resting on the same peridotitic basement sampled at Site 1068. As continental sediments were recovered,
81
it is thought that this fault slice is allochthonous (see below). Further to the west still, Site 899 recovered serpentinite breccias intercalated with sediments, perhaps indicating the presence of a nearby serpentinite high of some form (see below), and 897 recovered highly sheared partially serpentinized continental mantle peridotites. Finally, Site 1070, located just 15 km east of the J anomaly (unequivocal oceanic crust), recovered serpentinized mantle rocks and a lesser amount of gabbro. Although the main focus of the ODP drilling was the basement type, the south IAP drilling transect (legs 149 and 173) does constrain certain aspects of rift history. Tithonian clastics tilted within the fault blocks at sites 1065 and 901 were deposited in enclosed, relatively shallow and anoxic basins, close to vegetated land. These can be interpreted as the earliest synrift units, deposited before substantial extension had occurred (Whitmarsh & Wallace 2001). Later synrift units have, however, not been sampled, but the timing of crustal extension is constrained by Ar/Ar cooling data for samples from sites 1067 and 900, indicating that these deep crustal rocks were exhumed from more than 18 km (0.6 GPa) to approximately 4 km depth by 137 Ma (Wilson et al. 2001). If this
Fig. 4. Prestack depth migration of critical portion of profile LG12 (Krawczyk et al. 1996) and a structural interpretation. Note that early synrift sediments sampled at Sites 901 and 1065 are not obviously present in the intervening blocks, the presence of structures within the basement, interpreted here as segments of earlier faults, and the kink in the fault scarp west of Site 901, interpreted as two intersecting fault planes. H is a probable detachment fault at the top of serpentinized mantle. The location of the seismic on the transect is shown at the top: note that the top basement does not rise (and hence crust does not thicken) appreciably for over 50 km to the east.
82
T. J. RESTON
degree of exhumation applied to the whole crust, then in approximately 10 Ma between the onset of rifting in the Tithonian and the unroofing of these rocks, the crust would have been thinned by a factor of approximately 5. The final phase of crustal extension that produced the current geometries observed probably occurred shortly after this. Between the first oceanic crust, marked by oceanic magnetic anomalies, and the highly extended continental crust is a 120 km-wide zone of unroofed mantle, in which a series of basement highs were drilled (1068, 899, 897 and 1070). The sediment atop the highs represents a minimum age for the unroofing of that particular portion of peridotitic basement and is characterized by mixtures of serpentine breccias and mass flow deposits of Aptian–Barremian age. Intriguingly, the oldest of these deposits corresponds to the age of the first spreading anomaly (M3) observed further to the west, which can be taken as the onset of sea-floor spreading (base of the Barremian at 127 Ma according to Gradstein et al. 1994). The mass flow deposits may be related to the breakup process (Tucholke et al. 2007). To summarize, in the southern Iberia Abyssal Plain the evidence points towards major rifting and crustal extension occurring over perhaps 10 million years between the Tithonian and the Valanginian, followed by a phase of mantle unroofing over a further 10 million years before the onset of sea-floor spreading. As with the Galicia Bank margin, extension appears to have migrated oceanwards, but culminating in the unroofing of a far wider zone of mantle rocks. Although the overall time between the onset of rifting and of sea-floor spreading is similar at the two margin segments, approximately half of that time at the IAP margin only involved the unroofing and subsequent extension of mantle rocks. Hence, the rift duration leading to complete crustal separation may have only been about 10 million years. Both these margin segments thus exhibit tilted fault blocks, detachment faulting, serpentinization beneath the highly attenuated crust and a zone of mantle exposed at the sea floor during rifting. In this paper I will discuss how such margins might have formed, describing the structures that might be expected in each case, then comparing them with the observations from West Iberia.
Thermal and rheological evolution The thermal structure of the lithosphere is a major control on the rheology of the lithosphere, and hence an important consideration during progressive extension. A common misconception is that rifting leads to heating, and hence to thermal weakening of the rocks above a developing rift. Although rapid pure shear stretching (McKenzie 1978) leads to the steepening of the geotherm and increased
heat flow as hot rocks towards the base of the lithosphere are brought towards the surface, no actual rock gets hotter during uniform stretching at sediment-starved margins. Hot rocks are simply brought nearer the surface. Instead, adiabatic decompression means that even for instantaneous stretching the temperature of an individual rock actually drops marginally. For more realistic finite duration stretching (Jarvis & McKenzie 1980), some cooling occurs during rifting, leading to a cooling of those deep lithospheric rocks that have been transported towards the surface by the process of thinning, and consequently a less pronounced steepening of the geotherm.
Embrittlement Only where simple shear or strongly heterogeneous stretching occurs or where the margin is blanketed by thick sequences of thermally insulating sediment (e.g. Allen & Allen 1990) is some heating possible. For instance, some heating occurs in the hanging wall of a low-angle shear zone as heat from the exhumed footwall is transferred across the shear zone (e.g. Voorhoeve & Houseman 1988), but heterogeneous stretching is local in nature and is balanced by excessive cooling elsewhere (e.g. of the footwall in simple shear models). A rheological effect of extension more immediate than cooling is the pressure reduction at depth that results from lithospheric thinning. Decreasing the pressure leads to a reduction in the resistance to frictional sliding and fracturing, i.e. to brittle failure. Any simultaneous decrease in the temperature (due to cooling during slow rifting) serves to increase the resistance to creeping (e.g. dislocation creep). Consequently, rocks which once preferentially deformed by creep start to deform by brittle failure during the rifting process (Pe´rez-Gussinye´ & Reston 2001). For lithospheric parameters suitable for the West Iberia margin, Pe´rez-Gussinye´ & Reston (2001) showed that as extension factors increased, the upper creeping layer thinned, and the upper brittle –creeping transition shallowed until the entire upper and mid-crust became brittle at extension factors slightly below 3 (Fig. 5). If the spacing of faults is controlled by the thickness of the brittle crust (e.g. Scholz 1990), this implies that the spacing between new faults should decrease until extension factors approach 3. However, once the upper creeping layer has become entirely brittle, the spacing of new faults may increase again as they cut from the surface down to the deeper creeping layer. Note, however, that although the thickness of the crust prior to rifting may be reasonably well constrained, the boundary between the upper and lower crust is less well resolved.
RIFTED MARGINS BY PROGRESSIVE EXTENSION
0 10
0
yield stress (MPa) 200 400 600
(b) 0
middle and lower crust creep
WQ
Depth [km]
(a)
20 Depth [km]
AN 30 40 50
strong mantle lid
0
OL
120
10
m
110
r the geo
80
100
20
peridotites toward base lithosphere (here below ~ 80 km) are very weak and may be able to flow
200
400
OL
30
t = 6.5 m.y. β = 2.8
40
70
90
yield stress (MPa) 200 400 600
Middle crust brittle. Only lowermost crust creeps 10 strong mantle 20 lid
(c) 60
0
600
800
0
yield stress (MPa) 200 400 600 Whole crust now brittle OL t = 8.1 m.y. β = 3.6
30
T (°C)
original base of lithosphere
Fig. 5. The 1D rheological evolution of the lithosphere during pure shear, based on a model appropriate for west Iberia (see Pe´rez-Gussinye´ & Reston 2001 and references therein for the flow laws used: WQ, wet quartz; AN, anorthosite; OL, olivine) for a strain rate of 5 10215 s21. Base original lithosphere is a bold dashed line – lower lithosphere is very weak and may flow. (a) Starting model (geotherm shown by a dashed line), crust contains creeping zones within the mid crust and lower crust; (b) as extension proceeds, the mid-crust creeping zone becomes entirely brittle; (c) lower crust now also entirely brittle, faults cut down into the mantle. Note that lower crustal flow is not required for the entire crust to become brittle.
Furthermore, Pe´rez-Gussinye´ & Reston (2001) showed that for a West Iberia model, the entire crust should become brittle at extension factors of about 4 (Fig. 6), depending on rift duration and the rheology of the lower crust. Complete crustal embrittlement has two important consequences for lithospheric extension. The first is the disappearance of any weak decoupling zone within the middle or lower crust, allowing tighter coupling between the crust and the mantle. The second is the serpentinization of the uppermost mantle.
Coupling and lithospheric-scale simple shear The initial lithospheric strength profile in Figure 2 shows creeping zones both at the base of the upper crust and at the base of the lower crust. In these zones, strength decreases downwards so if the zones are thick enough, their lower portions may represent significant weak zones within the lithosphere.
83
The pattern of decoupling predicted is somewhat more complex than has been modelled in most studies that include strain weakening (e.g. Huismans & Beaumont 2003). In particular, by having a strong lower crust bound above and below by weak zones, the upper crust, lower crust and uppermost mantle may to some extent deform independently (Coward 1986; Reston 1990). As the overall resistance to shearing in these weak zones increases with distance between the loci of upper crustal and mantle extension, large lateral offsets of these loci can be ruled out (Braun & Beaumont 1987), but local decoupling and differences in the loci and style of deformation is likely. In particular, the relatively strong top layer of the lower crust may deform initially by boudinage, leading to a degree of depth-dependent stretching within the crust, discussed more below. Low strain lozenges of lower crust have been observed within deep crustal high strain zones (e.g. see Reston 1990), and inferred from seismic reflection data (Reston 1988). However, during progressive extension, the mid-crustal weak zone gradually disappears and the top of the lower crust becomes ever more tightly coupled to the upper crust. Complete coupling between the upper and lower crust is likely before the entire mid-crust becomes brittle at about 150% extension, and the location of the lower crustal boudins and shear zones may well control the development of second-generation faults at this time (see below). This might limit further depth-dependent stretching within the crust. Further extension will also couple the crust to the mantle, as the base crustal creeping zone also becomes brittle. At this stage, the entire upper lithosphere should be strongly coupled and resemble the coupled models of Huismans & Beaumont (2003), which (depending on strain rate and strain weakening) can develop strong asymmetries that resemble lithospheric simple shear.
Serpentinization The second effect of complete crustal embrittlement is the formation of mantle serpentinites. Serpentinized mantle forming both top basement within the COT (e.g. Pickup et al. 1996) and underlying the feather edge of the crust (Boillot et al. 1989) are important characteristics of NVRMs. The presence of serpentinites beneath the thinned continental crust implies that water passes down through the crust to react with the underlying mantle, requiring (O’Reilly et al. 1996; Pe´rez-Gussinye´ & Reston 2001) that the entire crust is brittle: fluid flow in sufficient volumes only occurs along faults and fractures. Thus, only when the entire crust has become brittle (extension .300%) should serpentinization first occur by the passage of water down
84
T. J. RESTON
Rift duration [Ma]
7 βb
30
6
0
5
4
1
3 10
IAP
Alps
IAP
DGM
Labrador Sea
20
2
3
4
5
6
2 7 8
2 3 9 10 11
β Fig. 6. Plot in rift duration v. extension factor space of crustal embrittlement curve (bb – bold line), maximum thickness of the serpentinizing zone (thin solid contours – brittle mantle thickness between the CMB and the 500 8C isotherm assuming pure shear) and melt thickness (dashed contours after Bown & White 1995). See Pe´rez-Gussinye´ & Reston (2001) for details. DGM, Deep Galicia margin; IAP, southern Iberia Abyssal Plain margin; Lab, Labrador margins; Alps, rifted margins exposed in Alps. Note the maximum extension factor (b) plotted is 12. For melt thickness at b ¼ 50 see Figure 8.
through the brittle crust along active faults (Pe´rezGussinye´ & Reston 2001). This prediction agrees well with the thickness of the crust at the landward limit of the underlying partially serpentinized peridotites at the West Iberian margin (Pe´rez-Gussinye´ & Reston 2001). Serpentinization leads to a dramatic drop in the strength of the uppermost mantle, leading to the development of serpentinite detachment systems, a different style of decoupling, at the base of the crust. Unlike the decoupling by the creeping zones discussed above, the continuity of brittle faulting passing down from the crust into the mantle may also lead to continuity of structure even if parts of the brittle fault zone are weakened by serpentinization. Further extension is thus likely to still resemble asymmetric simple shear, leading to the complete separation of the crust along the brittle faults and the final unroofing of a wide region of mantle. The structural dip and sense of shear deduced for both the S and H detachments (Krawczyk et al. 1996; Reston et al. 1996; Manatschal et al. 2001) is consistent with these
structures contributing to the unroofing of mantle rocks observed oceanward of each. A similar detachment (P) has been observed beneath the Porcupine Basin (west of Ireland), where it cuts from the eastern flank of the basin, beneath the centre of the basin (where the crust beneath the post-rift appears to be only about 3 km thick) and cuts to depth under the western flank of the basin (Reston et al. 2004). This is of interest as it is an asymmetric structure, implying that equivalent mirror image structures of S and H need not lie under the conjugate Newfoundland margins. In summary (Fig. 7), thinning results in rheological changes within the crust, with the lower crust becoming progressively less prone to creep and more brittle. The entire crust becomes brittle at extension factors in excess of about 4, when water can reach the mantle, leading to serpentinization. The weak hydrated mantle rocks formed at the base of the fault blocks will allow the formation of detachments near the base of the crust, perhaps accompanied by a change from pure shear to simple shear and the unroofing of mantle rocks.
RIFTED MARGINS BY PROGRESSIVE EXTENSION
85
Depth (km)
TIME / CONTINUING EXTENSION mid-crust brittle
0
entire crust brittle - seawater down to mantle
crustal separation
t
detachmen
10
Mantle exhumation
20 30 brittle crust
plastic crust
Mantle
Serpentinized peridotite
top lower crust
upper brittle-plastic transition
Detachment at top serpentinized mantle (e.g. H or S)
Fig. 7. Cartoon showing the effect of rheological evolution during progressive extension. Note the gradual embrittlement of the crust, the changing depth of detachment of the main faults (dashed line) from mid to deep crust, and the development of serpentinite detachments leading to final crustal separation once the entire crust has become brittle. Modified from Pe´rez-Gussinye´ & Reston (2001).
An important question to address when considering rifted margins is why some are characterized by voluminous magmatism and others by very little, if any. From studies of mid-ocean ridges it has been inferred (McKenzie & Bickle 1988) that the ‘normal’ temperature of the asthenosphere had it had been brought instantaneously (adiabatically) to the surface would have been approximately 1300 8C (the potential temperature, Tp). During rifting, the upwelling asthenosphere is likely to have undergone partial melting when its temperature exceeded the mantle solidus. When this occurred would depend on three main factors: the initial thickness of the lithosphere, the rate and amount of asthenospheric upwelling, and the initial temperature of the asthenosphere. Various authors (e.g. Bown & White 1995; Pe´rez-Gussinye´ et al. 2001) have calculated in 1D models the point at which the geotherm would intersect the solidus (the onset of melting) and also the melt thickness generated. Although using different parameterizations, these studies find that, depending on the rift duration, significant melting is not expected during the evolution of a margin similar to West Iberia until b . 5, where b is the extension factor (Fig. 6) and that total melt thickness is unlikely to exceed 4 km until after continental breakup (Fig. 8). This is, however, considerably more than is thought to have occurred at the West Iberia margin (e.g. Minshull et al. 2001; Reston & Phipps Morgan 2004) (Fig. 8). The same applies to the Alpine margins and other less wellconstrained examples (Reston & Phipps Morgan 2004). Similar results are generally obtained in
those 2D modelling studies where melt production has been explicitly included, e.g. Bowling & Harry (2001) showed that approximately 5 km of melt are expected before crustal separation can take place for a mantle potential temperature of 1300 8C. Thus, there is a general observation that at NVRMs, the amount of melt observed is less than that predicted by pure shear models for the
6
Melt thickness (km)
Melting and mantle unroofing
5
13
00
4 12
50
3
12
2
00
1
Alps
IAP 0 0
10
20
30
Rift duration (m. a.) Fig. 8. Melt thickness expected for various rift durations at an extension factor of 50 for various asthenospheric potential temperatures (after Minshull et al. 2001 using parameterization of Bown & White 1995). The thickness observed at the IAP (Minshull et al. 2001) and Alpine margins (Mu¨ntener & Herrmann 2001) is consistent with cool sublithospheric mantle.
86
T. J. RESTON
extension factor derived from crustal thinning. This, I term the melt discrepancy.
Temperature (°C) 400
600
800
1000
1200
1400
1600
Cool subcontinental geotherm Geotherms from seismic and heat flow Global: Archaean craton, stable platform, tectonic province
200
Eastern North America average western U.S. average
300
Kaapvaal kimberlite geotherms
unheated
Depth (km)
100
“plume-heated” at base lithosphere
mantle adiabats 400 1300
1100
Potential temperature Tp (°C)
1200
One solution is that the paradigm of a 1300 8C asthenosphere may be incorrect (Reston & Phipps Morgan 2004). The thickness of melt at volcanic margins is far greater, and is generally attributed to a Tp greater than 1300 8C in the upwelling mantle (e.g. a plume): could the lack of magmatism at nonvolcanic margins imply that elsewhere Tp beneath the lithosphere is less than 1300 8C? If the mantle beneath the continental lithosphere is even slightly cooler (Tp c. 1250 or 1200 8C, for instance), the volume of melt produced at effectively infinite extension factors by partial melting of dry peridotite is considerably reduced (Fig. 8), although small amounts of wet melting will take place. Some variation in sublithospheric temperature is, of course, to be expected if the asthenosphere is convecting (Lenardic & Moresi 2000) – convection is, after all, driven by temperature variations. Surprisingly little is known about the temperature structure beneath the base of the continental lithosphere, an issue complicated by the different definitions of where this base lies. In the following discussion, we concentrate on evidence of the temperature structure beneath the thermal lithosphere, i.e. in the presumably convecting region beneath the outer thermal boundary layer (Fig. 9). The first evidence comes from cratonic regions (Reston & Phipps Morgan 2004) via geothermobarometry on xenoliths brought up from the base of the lithosphere by kimberlites (e.g. Kohler & Brey 1990). The xenoliths occur in two families – sheared xenoliths heated at the base of the lithosphere by the mantle plume that initiated kimberlite magmatism and unsheared xenoliths from the entire lithospheric column. The former exhibit a Tp range between approximately 1200 and 1400 8C at a largely constant depth of about 200 km, indicating both the depth of the base of the lithosphere and the temperature range between unheated and plume-heated xenoliths there. In contrast, the unsheared xenoliths show that away from plume heating, the potential temperature at the base of the lithosphere was generally equal to or below a potential temperature of 1200 8C, and thus that the convecting subcratonic mantle has a Tp well below 1300 8C. Unless this differs from other subcontinental convecting mantle, the implication may be that this is also relatively cool. More generally, the temperature beneath the subcontinental lithosphere can be deduced from the depth to and thickness of the mantle transition zone (e.g. Gossler & Kind 1996), from seismic
Fig. 9. Subcontinental geotherms derived xenoliths in kimberlite pipes (cratonic) and from seismic þ heat flow (global and North American averages). Considerable variation in the seismic/heat flow geotherms above about 100– 200 km represents the varying temperature structure of the continental lithosphere, beneath which the geotherms are subparallel to mantle adiabats at potential temperatures of between 1000 and 1400 8C. Allowing for the error in such seismic geotherms, it is still clear that many sublithospheric geotherms from a variety of tectonic settings must lie well below the 1300 adiabat. Xenoliths within Kaapvaal kimberlites show two families: those following the steady-state geotherm down to a potential temperature of 1200 8C, and those which were heated to a higher potential temperature by the impact of the plume on the base of the lithosphere. These imply that the potential temperature at the top of the convecting interior was 1200 8C away from plumes. Other kimberlite provinces are similar or slightly cooler still. See Reston & Phipps Morgan (2004) for a complete discussion of these geotherms.
velocities (e.g. Ro¨hm et al. 2000; Goes & van der Lee 2002) and waveforms (Cammarano et al. 2005), and from studies of heat flow (e.g. Jaupart & Mareschal 1999). All of these suggest considerable variability in the potential temperature of the convecting mantle, but with the general indication that it may be considerably cooler beneath the continents than beneath the oceans (Fig. 9). The differences are sufficient to explain dramatic variations in melt volumes during rifting (Fig. 8). Differences in temperatures of the convecting interior immediately beneath the the continental and oceanic lithosphere
RIFTED MARGINS BY PROGRESSIVE EXTENSION
are an integral part of some models of mantle convection (e.g. Lenardic & Moresi 2000). Others have suggested that the oceanic ridge system may be supplied by a relatively thin layer of warm asthenosphere derived from mantle plumes and ponding beneath the thinner oceanic lithosphere (Phipps Morgan et al. 1995), a suggestion supported by the studies of transition zone thickness and seismic waveforms cited above.
Melt trapping within the lithosphere It is, of course, possible that the lack of observed magmatism at NVRMs is not owing to a lack of melting, but rather that considerable volumes of melt are trapped and solidify in the mantle lithosphere, a suggestion also made for those portions of slow-spreading mid-ocean ridges characterized by thin magmatic crust (Cannat 1996). A recent seismic experiment (Lizarralde et al. 2004) showed that beneath slow-spreading crust the velocity structure of the uppermost mantle was consistent with the presence of numerous deep gabbroic intrusions with an integrated thickness of about 1.5 km. A similar thickness trapped at depth at rifted margins could help explain the melt discrepancy. The recognition of intrusions in the mantle at rifted margins will be complicated by the difficulty not only in distinguishing between gabbros and serpentinites, but also by any intrinsic velocity contrast between exhumed subcontinental lithospheric mantle and its oceanic equivalent. Exhumed mantle rocks in the Alps do show evidence both of intrusion by gabbroic bodies and impregnation by basaltic melt (Mu¨ntener & Herrmann 2001); these are included in the estimates of melt thickness in the Alps (Fig. 8). However, even a small amount of melting may focus extension by weakening the thinning lithosphere through which they are percolating (Lavier & Manatschal 2003). The observation at the Galicia Bank margin of minor synrift gabbros (Scha¨rer et al. 1995) that were strongly sheared during further extension is compatible with deformation at depth being focused where partial melt is present. If melt lenses do act to focus deformation, this should increase the degree of exposure of igneous rocks (brought up in the footwall of the fault) and makes the scarcity of such rocks still more problematic.
Depth-dependent stretching (DDS) The volume of melt produced might also be reduced if the extension of the lithosphere is heterogeneous (Minshull et al. 2001), i.e. if depth-dependent
87
stretching (DDS) takes place so that crustal thinning exceeds the total lithospheric thinning beneath the rift centre. More specifically, crustal separation may occur before complete lithospheric separation. The key in suppressing melt within the COT is thus to extend the crust in a narrower zone than the subcrustal lithosphere, perhaps particularly the deep lithosphere (see below). This is a very different sort of DDS to that proposed as an explanation for the extension discrepancy, where it has been argued that different crustal levels are extended by different amounts in different places. Apart from local heterogeneities associated with faults and localized shear zones, significant and systematic lithospheric-scale DDS may require the decoupling of the deeper lithosphere from the shallower structures. I suggest that lithospheric-scale DDS to suppress melting may occur through a combination of several processes (Fig. 10): by flow of the deepest lithosphere (e.g. see models of Huismans & Beaumont 2003), which is consistently weak during rifting (Fig. 5), by displacement along large structures cutting through the entire mantle lid (Fig. 5), leading to a lateral displacement of the locus of mantle and crustal extension (Reston 1993), by shear along the mid and lower crustal weak zones (e.g. Reston 1988, 1990, 1993), and by displacements along late-stage serpentinite detachments (Pe´rez-Gussinye´ & Reston 2001). The development of serpentine detachments (Fig. 7) occurs late in the rifting process and would appear to provide a mechanism for widening the region of mantle extension and focusing that of the overlying crust (leading to complete crustal separation as the blocks are pulled apart). However, serpentinization is not predicted to occur until extension factors exceed about 4, by which time the asthenosphere should have upwelled to within about 30 km of the surface and, for a Tp of c. 1300 8C, melting should have just started (Pe´rez-Gussinye´ et al. 2001). It is hard to see how the development of serpentine detachments can prevent further asthenospheric upwelling while still allowing the unroofing of a 120 km-wide expanse of mantle rocks, as observed west of Iberia (Fig. 1). If DDS is allowed by decoupling within the crust, it is likely to be most pronounced relatively early in the extension history, before the crustal weak zones disappear through crustal embrittlement and the crust becomes tightly coupled to the mantle (Figs 5 & 7). Such a decoupling mechanism has been inferred from seismic reflection data (Reston 1990, 1993) and is compatible with numerical modelling (e.g. Braun & Beaumont 1987), although doubts remain about the efficiency of the process over large length scales (Braun & Beaumont 1987). Reston (1990, 1993) proposed that this mechanism
88
T. J. RESTON
Fig. 10. Depth-dependent stretching controlled by shear in the lower crust and shear zones in the uppermost mantle from linedrawings of unmigrated seismic sections from offshore UK, and from numerical modelling. (a) DRUM profile showing mantle reflection dipping to a depth of approximately 75 km (after depth migration). This structure may be an extensional shear, detaching upwards in a creeping layer at the base of the crust (Reston 1993). (b) North Sea line NSDP-84-01, showing symmetric outward-dipping mantle reflections, detaching upwards at the base of the crust (Klemperer 1988; Reston 1993). In both cases, lower crustal reflectivity may be related to shearing as strain is transferred between the crust and the mantle (Reston 1990). (c) Numerical model of decoupled lithospheric extension, including viscous strain softening (Huismans & Beaumont 2003), showing highly sheared creeping layer at the base of the crust (only one crustal lithology modelled), and boudinage of strong mantle lid, together with shearing of deeper mantle. Dashed lines represent interpreted high-strain zones in the mantle for comparison with the seismic. (d) More extended coupled lithospheric model (also with viscous strain softening) showing how mantle shear zones both broaden the region of mantle extension over focused crustal thinning and allow the incipient unroofing of rocks from the lower lithosphere at the sea floor. Displacement of lower lithospheric mantle to the centre of the rift should suppress melt generation to some extent.
RIFTED MARGINS BY PROGRESSIVE EXTENSION
might be coupled with the localization of deformation within the uppermost mantle lithosphere: a dipping band of seismic reflections are observed on the DRUM profile from the base of the crust down to a depth of approximately 75 km (Reston 1993) (Fig. 6), roughly the thickness of the strong mantle lid (Fig. 5). Symmetric extensional mantle shear zones under the rift flanks and dipping away from the rift centre, such as those reported beneath the North Sea (Reston 1993) (Fig. 10), might then allow the region of mantle extension to be more widely distributed than that in the crust, and may further allow the subcrustal lithosphere to be displaced towards the rift centre. Such outward-dipping mantle structures are predicted to develop according to some numerical models (Harry & Sawyer 1992; Huismans & Beaumont 2003) and would help explain the onlap of the post-rift sequences onto the prerift of the basin margins (Reston 1993). However, numerical modelling is somewhat equivocal on whether the zone of mantle extension is likely to be more or less focused than that of the crust. The models of Huismans & Beaumont (2003) with an (unrealistically) weak decoupling lower crust show that crustal extension is more distributed and less focused than mantle extension; their coupled models where the lower crust is (unrealistically) strong do predict more localized crustal than mantle extension (Fig. 10). The behaviour of a model containing initially moderately decoupling zones in the mid and lower crust that gradually couple deformation more strongly with ongoing extension is unclear, although may show a gradual increase in the focusing of crustal deformation as discussed above. Perhaps the most promising form of heterogeneous extension throughout the rifting process is the idea that the lowermost lithosphere (initially beneath c. 80 km depth – see below) is stretched over a wider region than the stronger upper lithosphere (e.g. Minshull et al. 2001) (Fig. 10d). Unlike decoupling within the crust (early extension only) and movement along serpentinite detachments (late extension only), displacement of the lower lithosphere towards the rift zone may
89
continue through most of the rifting process (Huismans & Beaumont 2003). The presence of subcontinental lithospheric mantle (e.g. Manatschal et al. 2001) directly beneath the post-rift succession implies that some form of depth-dependent stretching took place during rifting. Similar subcontinental mantle has been described in the Alps, where the rocks appear to come from a progressively deeper and hotter source moving oceanwards: although the mantle directly in contact with the crust is preserved (e.g. in the Malenco area where it has a T at the start of rifting of c. 600 8C: Mu¨ntener et al. 2000), Desmurs et al. (2001) report that the prerifting temperature of the peridotites sampled across the Err– Platta COT increase oceanward from 850 + 50 to 1050 8C. At the West Iberian margin, the trend is less obvious with temperatures generally in the range 1000– 1250 8C across the entire COT (Table 1, Fig. 1). Taking the thermal model of Pe´rez-Gussinye´ et al. (2001) this would place the exhumed mantle at original depths of 90– 120 km, indicating that DDS is probably important within the lower half of the lithosphere. Within the COT the lack of mantle samples that were at temperatures of between 600 (slightly above Moho temperature) and 800 8C before rifting implies that DDS does not systematically lead to the displacement of the upper two-thirds of the mantle lithosphere levels towards the COT. Mantle levels initially above about 80 km (Fig. 5) may be too strong to undergo systematic large lateral displacement towards the developing rift, and may instead fail by a boudinage process, bound by landward-dipping shear zones in the mantle and above by a decoupling lower crust (Fig. 10). The presence of inferred landwarddipping high-temperature normal shear fabrics in the peridotites sampled west of Iberia (e.g. Beslier et al. 1996) and of landward-dipping structures in the COT (Boillot & Froitzheim 2001; Reston et al. 2004) is entirely consistent with the development of such outward-dipping mantle shear zones. It seems likely that this process together with early and late contributions from crustal decoupling
Table 1. Summary of prerift/synrift temperatures of mantle peridotites sampled in the COT of the West Iberian margin. All results are from the IAP transect except for Site 637 from the DGM transect. Sites ordered from ocean (left) to continent (right). Note that there is no obvious landward progression to cooler temperatures (Opx, orthopyroxene; Cpx, clinopyroxene) Site
1070
897/899
T (8C) Opx: 800 –1250 880 –1000 shearing 1170 – 1230 Cpx: ,900 8C crystallization Ref. He´bert et al. (2001) Beslier et al. (1996), Seifert & Brunotte Cornen et al. (1996) (1996)
637 (DGM) 1000 – 1250 Beslier et al. (1996)
1068 Opx: 1200 –1280 Cpx: ,900 8C He´bert et al. (2001)
90
T. J. RESTON
and mantle serpentinization, respectively, may contribute to DDS of the lithosphere and hence to the reduction of melt generation. Although numerical models of lithospheric extension shows that the region of mantle extension may be wider than that of the crust (e.g. Braun & Beaumont 1987; Reston 1993; Huismans & Beaumont 2003), these studies have not yet been combined with melt generation. However, we can estimate the required heterogeneous stretching by assuming that melt generation is directly related to the rate and amount of asthenospheric upwelling. To generate only about 1–2 km melt at the IAP margin and less than 1 km at the Alpine margins, the upwelling should only be equivalent to that occurring during pure shear by a extension factor of approximately 9+2 (Fig. 6), implying that across the COT about 15 km thickness of (mantle) lithosphere should still be present. Minshull et al. (2001) came to a similar conclusion and pointed out that, although a degree of heterogeneous stretching could easily explain the lack of melt within a narrow region of exhumed continental mantle, it is hard to see how this could explain a paucity of melt in a zone of exhumed continental mantle with a width of up to 170 km as any layer of lithospheric mantle originally exhumed within the COT would be continually stretched and extended with ongoing crustal separation. Thus, the amount of synrift igneous rock in the COT may simply be underestimated as it is trapped at depth, or be reduced either by a cooler subcontinental geotherm or by depth-dependent stretching. All of these hypotheses have their merits and all are likely to contribute to the apparent melt discrepancy at rifted margins. There is ample evidence that the continental sublithospheric mantle may be cooler than the asthenosphere beneath the oceans; there is good evidence of depthdependent stretching of the lower lithosphere and for the development of serpentine detachments during late-stage rifting, and some melt is likely to be trapped at depth beneath rifted margins as it is beneath slow-spreading oceanic crust.
Faulting and the extension discrepancy Some (but not all) non-volcanic rifted margins are characterized by well-defined rotated fault blocks. (Similar blocks may be present at volcanic margins but are buried beneath thick sequences of lavas which obscure the underlying image.) Where fault blocks are well imaged (White 1990) the amount of extension can be determined from the geometry of the blocks, specifically from the amount that they have been rotated (e.g. Barr 1987). In effect the presence of the prerift sedimentary units can be used to determine the amount of
the extension, but accurate determination of what is prerift is essential. However, at some margins thinning appears to occur without significant faulting (e.g. the Exmouth Plateau margin: Driscoll & Karner 1998) and at others the amount of extension that can be inferred from the geometry of the faults at the feather edge of the continent is less than that indicated by the crustal thinning observed (e.g. Sibuet 1992). This is the so-called extension discrepancy. The rolling hinge model (Buck 1988) for the development of fault slices (which may well be applicable to late-stage faulting during the formation of the West Iberia margin: e.g. Reston et al. 1996) does not alone solve the extension discrepancy as it should still be possible to reconstruct the geometry of the hanging wall from the various fault slices if the prerift is present. Only if these slices were cutting through a previously thinned section should the reconstructed section fail to restore to full crustal thickness. This can be seen with restoration of depth sections from the West Iberia margin (see below). The West Iberia margin exhibits an apparent extension discrepancy. The structure of the south Iberia Abyssal Plain segment of this margin is well imaged by the prestack depth migration of a high-quality reflection profile LG12 (Krawczyk et al. 1996) (Fig. 4) and further constrained by seismic refraction data (Chian et al. 1999; Dean et al. 2000). The most obvious features of the seismic data are the L fault, a large partly listric fault that appears to flatten at a depth of about 14 km, and the H detachment, which can be traced westwards down from the 1065 high, beneath a couple of wedge-shaped fault blocks and back to the top of the site 900 high. At the DGM, the seismic signature (polarity, waveform and amplitude: Reston 1996) of the S detachment fault indicates that it forms a sharp boundary between the crustal rocks and underlying higher velocity partially serpentinized peridotites (reflection coefficient 0.2), a result confirmed by wide-angle data (Zelt et al. 2003). Similarly, the H reflector shows an almost identical waveform to the sea-floor reflection, itself shown by comparison with its multiple to be a reflection from a single interface; the reflection coeffcient of H is estimated to be 0.15 (Krawczyk 1995). Between L and H, another fault (here labelled K) can be traced to depth from near the west flank of the 1065 fault block. To the west of H, a reflection that marks the continuation of top basement underlies the 1069 high; it is thus thought that the 1069 fault slice is allochthonous and is separated from the underlying basement (serpentinized peridotite?) by a low-angle fault (F: Krawczyk et al. 1996; or HHD, Whitmarsh et al. 2001) that follows top basement for much of its length.
RIFTED MARGINS BY PROGRESSIVE EXTENSION
A restoration of the depth section of LG12 illustrates the extension discrepancy. This restoration differs from those of Manatschal et al. (2001) and of Whitmarsh et al. (2001) in several ways. First, I have restored movement along the fault K, ignored previously. Second, I have taken the CMB within the limits of Whitmarsh et al. (2000), constrained by wide-angle data. It is important to take the CMB and not the deeper Moho as the
91
serpentinites between the two is petrologically part of the mantle not the crust. Third, I have made the restoration in two phases (Fig. 11), indicating the oceanward migration of the location of the extension. By initially removing the effect of fault F, which clearly cuts across and hence postdates the H detachment, I have also ‘undeformed’ the wedge-shaped fault blocks directly above the H detachment and in doing so restored H to a
21.4 km 1065 brittle crust + seds
movement along older 901 faults not restored
1069
new fault existing fault
old ?
CMB interpretations drillsite
H
fault old faults
K
old fault L
1067 900 K H
CMB
L
5 km 1068
H, K and L cut through entire brittle crust and into mantle. Passage of fluid along faults leads to mantle serpentinization. Movement along H,K,L results in offset Moho and back-rotated blocks. 901 1065 1069
? H
1067 900
H
L K
H
1068
F
L
H, K, L rotated to low angles, become inactive. New fault F cuts down through H and to depth to west.
900
1067
F
1068
37.4 km 901 1065
37.4 km
L H 1069 block c. 10 km to west
K
H CMB CMB
L
CMB
Final section. Movement along F caused flexural back rotation of H and of F due to footwall unloading. Possibly enhanced by serpentinization (volume increase) of mantle beneath H and F and by further faults to west. βf (stretching factor) during these last phases of faulting (H, K, L and then F): 37.4 km / 21.4 km = 1.75.
Fig. 11. Two-stage restoration of LG12 based on the structural interpretation shown in Figure 4. The restoration illustrates several key points: the fact that movement along the latest faults cannot explain the extreme thinning of the crust, the possible occurrence of polyphase faulting, the development of faults cutting across the entire crust, the development of serpentinite detachments, and the unroofing of mantle and deep crustal rocks. See text for discussion.
92
T. J. RESTON
consistent, if largely low-angle, westward dip. Despite these differences, the general result is similar to previous restorations, showing that the 1069 block was originally contiguous with the blocks overlying the H detachment and that the crust was already quite thin before the latest phase of faulting. Finally, I emphasize that this is only a partial restoration which aligns intrabasement reflections that may correspond to earlier faults (see below). The wedge-shaped fault blocks above H, and the resultant oceanward steepening of the bounding faults on the restored section, is consistent with a rolling hinge model, but is not evidence for the application of this model to more than the breakup of the crustal block above H. Although it is quite possible that the other faults marked in bold in Figure 11 were active sequentially (e.g. in a rolling hinge model with the locus of deformation migrating oceanward), the lack of cross-cutting relationships between the faults prevents confirmation of this hypothesis, so I have restored the movement along these faults in one go. The restored section has a crustal thickness of less than 6 km. Even using the deeper Moho of Manatschal et al. (2001), the crust was only about 8 km thick prior to the development of the latest faults. Clearly, the crust has not been solely thinned from a thickness of approximately 30 km by movement along these faults. The question is how the extra thinning has occurred. Proposed solutions to the extension discrepancy include lithospheric- (or
at least crustal-) scale simple shear, depth-dependent stretching and polyphase faulting.
Simple shear If the lithosphere was pulled apart by simple shear along a low-angle detachment (Wernicke 1985), the failure to recognize this detachment at either the upper or lower plate margins would result in a large underestimate of the amount of extension accommodated by faulting. At an upper plate margin, such a detachment may form a tectonic Moho and will not project up to a breakaway, making the recognition of the detachment particularly difficult: at a lower plate margin, the detachment fault should cut up through the crust towards a breakaway, and so should be easier to identify. As a result, attempts to explain the extension discrepancy by simple shear generally invoke such an invisible detachment to thin the crust beneath the upper plate. Apart from the rheological unlikelihood of cutting through the entire rheologically stratified crust with a through-going detachment at an early stage of rifting (e.g. Huismans & Beaumont 2003), another problem is that not all margins can be upper plate margins (the ‘upper plate paradox’: Driscoll & Karner 1998). The simple shear idea might, however, still apply if the detachment on the lower plate was simply not identified as such, so that the extension on this margin is also underestimated (Fig. 12). For instance, if it were to be cut by later faults
upper plate
lower plate ent
lt
fau
hm tac
upper crust
de
Moho
Moho
lower crust
mantle
upper plate
successor faults cut detachment + further thin lower plate. Lower crust exhumed and faulted
upper crust upper plate over-thinned by removal of lower crust
lower plate upper crust lower crust
Fig. 12. Cartoon illustrating how movement along a crustal-scale shear zone can explain an apparent extension discrepancy on the upper plate and on the lower plate if the nature of the detachment surface is not recognized, for instance if it is cut by later normal faults (note that owing to the embrittlement of the crust these can cut across lower crustal rocks). Only if the prerift is recognized and dated on both conjugate margins can such a possibility be excluded.
RIFTED MARGINS BY PROGRESSIVE EXTENSION
(e.g. successor faults, similar to those observed cutting detachment surfaces at mid-ocean ridges) the detachment might well not be recognized and the total amount of extension measured would correspond to that along the later, steep faults. This is, of course, an example of polyphase faulting (see below). The key here, as always, is to accurately identify the prerift sequences on both sides of the ocean and to distinguish these from those ‘prefaulting’ sequences deposited prior to the development to the visible faults.
Depth-dependent stretching in the crust Simple shear is just one form of depth-dependent stretching (DDS), which if occurring within the crust provides a possible explanation for the extension discrepancy. (It is not, however, the only solution, so DDS should not be used as synonymous with the extension discrepancy.) In DDS, the total amount of extension undergone by all the lithospheric levels must be the same at a regional scale to avoid space problems, but the lateral distribution of this extension may be different. Various forms of DDS has already been discussed above as a possible mechanism to explain the small amount of igneous activity associated with the development of NVRMs. It is also clear that some form of DDS is necessary to explain the unroofing of large expanses of subcontinental lithospheric mantle within the COT of some margins (West Iberia and the Alpine margins: Manatshal et al. 2001). However, as discussed above, the petrology of these mantle rocks indicates that they came from what was originally the lower part of the lithospheric mantle and as such do not provide evidence for DDS at higher lithospheric levels. This is important as only DDS within the crust can be invoked to explain the extension discrepancy, by thinning the crust more than observed from the faults imaged. In effect, explaining the extension discrepancy by DDS requires that the lower crust has somehow been displaced away from beneath the little faulted brittle upper crust, which may even be placed directly atop lithospheric mantle. This might in principal thin the crust by a factor of 3, assuming everything beneath the initial brittle upper crustal layer (c. 10 km thick, see Fig. 5) is removed without thinning the upper crust at all. Such a displacement might occur away from (Brun & Beslier 1996) or towards the margin (Driscoll & Karner 1998). The latter would seem more likely to lead to an inverse discrepancy as it would serve to maintain crustal thickness where the upper crust has been highly extended, and even completely separated, unless final breakup occurs where the lower crust has been removed, in which
93
case the conjugate margin should show the inverse discrepancy as an apparent ‘lower plate margin’. In contrast, any postulated displacement of the lower crust toward the continent would have to be accompanied by opposite displacements of the mantle lithosphere toward the COT, where it is exhumed. Apart from crustal-scale simple shear, several mechanisms of crustal DDS are conceivable, including lower crustal flow and lower crustal boudinage (Fig. 13). Lower crustal flow is problematic as it has been shown that, at the start of rifting, the lower crust beneath a cool rift is too viscous to flow over large distances (Kusznir & Matthews 1988; Hopper & Buck 1998; McKenzie & Jackson 2002), although still capable of small-scale adjustment in response to the tilting of the fault blocks. Second, as discussed above, the rheology of the crust is not fixed, so that during progressive extension the crust becomes ever more tightly coupled to the mantle until, at large extension factors, even originally creeping rocks become brittle. The entire crust may be brittle by an extension factor of about 4– 5 (Pe´rez-Gussinye´ & Reston 2001), so that the lower crust should not be able to flow anywhere. Far from lower crustal flow being favoured at high extension factors as suggested by analogue models (e.g. Brun & Beslier 1996), a decoupled deformation within the lower crust must become progressively more difficult. The fundamental flaw of most analogue modelling is that the rheology of materials is fixed and cannot evolve in response to pressure, temperature, strain rate and fluid pressures. This hinders meaningful simulation of rift evolution at extension factors greater than about 1.5 when rheological changes start to become important. The implication is that at NVRMs, such as West Iberia, the extension discrepancy can neither be studied by such modelling nor be solved by lower crustal flow. More promising is the necking or boudinage of the strong core of the lower crust between weak decoupling zones above and below. A degree of such boudinage has been inferred from deep seismic reflection images (Reston 1988). Completely removing the lower crust beneath sections of the developing margin would require a far more extreme boudinage, and it is hard to see how the resultant differential subsidence and shear on the base of the brittle layer would not induce extensive faulting. Such boudinage is thus unlikely to explain those places where upper crustal extension appears negligible although the entire crust is strongly thinned (Driscoll & Karner 1998; Davis & Kusznir 2004), but may help to explain places where the amount of upper crustal thinning is substantial but nevertheless smaller than whole
94
T. J. RESTON
(a)
ED
IED COT
UC
? creeping LC CL Mantle
Mantle
lower crust displaced oceanward (b)
IED ED COT UC
Mantle
CL creeping LC Mantle
lower crust displaced landward (c)
IED
ED
IED
ED COT
UC CLC L Mantle strong
Mantle LC boudinage
Fig. 13. Models for DDS within the crust that may give rise to an extension discrepancy. ED, extension discrepancy; IED, inverse extension discrepancy. (a) Lower crustal flow or displacement towards the COT (after Driscoll & Karner 1998). This should give an inverse discrepancy near the COT, unless the lower crust is itself subsequently faulted and misinterpreted as upper crust. (b) Lower crustal flow away from the COT (Brun & Beslier 1996). (c) Lower crustal boudinage along shear zones (Reston 1988). This should give local extension discrepancies between the boudins and inverse discrepancies above the boudins.
crustal thinning. If the locus of lower crustal boudinage influences the site of eventual breakup, a moderate extension discrepancy might thus result near the COT. Davis & Kusznir (2004) proposed that depthdependent stretching could develop in response to a divergent flow field associated with the onset of sea-floor spreading which would, in effect, shear the lithosphere and thus allow extension of the lower lithosphere over a wider zone than that of the upper crust (Fig. 14). This idea is attractive in that it provides a mechanism for extremely large-scale DDS linked to processes associated with the onset of sea-floor spreading, i.e. which may be restricted to margins. However, such a divergent flow field can only explain the extension
discrepancy if it leads to DDS of the different crustal levels. If, for instance, the lower crust is sheared away from the little extended upper crust, the displacement relative to the upper crust of the lower crust should be in the same direction as that of the underlying mantle, unless the kinematics of the deformation field are very complex. As the subcontinental mantle appears to be displaced towards the COT at NVRMs, the lower crust should also be (e.g. Driscoll & Karner 1998), which as discussed above may not provide an explanation for the extension discrepancy. Furthermore, it is clear that an active divergent upwelling can only affect the lower crust once it has affected the entire subcrustal lithosphere. The presence of mantle rocks within the COT dominantly from the lower half of the
RIFTED MARGINS BY PROGRESSIVE EXTENSION 0
100
200
300
400
km
0 40 km
original mantle lithosphere 100 km
original asthenosphere
passive rifting upper crust 0
100
200
lower crust 300
400
km
0
100 km
original mantle lithosphere active rifting
original asthenosphere
Fig. 14. Half of symmetric depth-dependent stretching models for the formation of rifted margins resulting from divergent upwelling, both passive and active (Kusznir et al. 2005). Although these models each predict some of the features discussed in the text, neither can duplicate all of the observations.
subcrustal lithosphere may mean that DDS has only affected the lower half of the mantle lithosphere and not the strongest mantle lid beneath the Moho. If this is the case, it cannot then have affected the lower crust and cannot help resolve the extension discrepancy. Despite these inherent problems, Davis & Kusznir (2004) and Kusznir et al. (2005) proposed that depth-dependent stretching explained the extension discrepancy at a variety of margins. Davis & Kusznir (2004) reproduced a small amount of depth-dependent stretching, but the amount was far less (maximum bwhole crust 2 bupper crust 0.2) than they had inferred at rifted margins (bwhole crust 2 bupper crust 1.8) to explain the extension discrepancy. Kusznir et al. (2005), explored the effect of active v. passive upwelling on the margin structure, including the unroofing of subcontinental mantle in the COT. The passive model (Fig. 14, top) predicts the unroofing of a zone of subcontinental mantle approximately 50 km wide (somewhat less than observed west of Iberia, but comparable to other NVRMs – Fig. 1). The model also predicts that a complete lithospheric section should be unroofed in the COT, with mantle lithosphere from all depth and temperature ranges. This is not observed: as mentioned above and discussed more below, the mantle rocks unroofed appear to come from the lower lithosphere. Also note that the thinning of the upper continental crust occurs rapidly at about 300 km and exceeds the total crustal thinning further oceanwards. Far from explaining the extension discrepancy, this model predicts an inverse discrepancy, which has not been noted. The active upwelling model in contrast (Fig. 14, bottom) does predict
95
less upper crustal than whole crustal thinning between about 220 and 270 km, but not sufficient to explain anything more than a very slight extension discrepancy, and beyond 280 km the upper crust appears to be thinned more than the whole crust. Also note that this model can only explain the unroofing of a very narrow zone of subcontinental mantle (at depth of 305–320 km). There are other problems with crustal DDS as an explanation for the extension discrepancy. First, if the upper crust near the COT is extended far less than the whole crust because the LC has somehow been removed, where has the missing lower crust gone? Another way of looking at this is that if the upper crust near the COT is less extended than the deep crust, then somewhere else the upper crust must be extended considerably more than the lower crust. Such an inverse discrepancy is, however, never described, although required by models of DDS (e.g. Kusznir et al. 2005). I term this the ‘paradox of the missing lower crust’. As discussed above, if the lower crust has been excised and the upper crust placed directly against lithospheric mantle, it might in theory be possible to thin the crust considerably without significant faulting. The amount of crustal thinning that can occur without extending the brittle layer is simply dependent on the relative thickness of the whole crust and the brittle upper layer. In a warm lithospheric model with relative thick crust, appropriate for instance for the Woodlark Basin the upper brittle layer might be thinner and the whole crust thicker than shown in Figure 5, potentially allowing more crustal DDS, although whether this would actually occur is another matter altogether. However, for West Iberia, crustal DDS that removes everything beneath the approximately 10 km-thick brittle upper layer can only thin the crust by a factor of 3. Even after restoring the extension along the block-bounding faults West of Iberia (Fig. 11), the crustal section is considerably thinner than 10 km, requiring at least a supplementary explanation other than DDS beneath the brittle upper crust. A further problem of applying crustal DDS to the West Iberia margin is that, although some authors have suggested otherwise (e.g. Nagel & Buck 2004), there is no evidence that the lower crust is generally missing towards the COT of the West Iberia margin where the extension discrepancy is most pronounced. (The same holds for the Alps, where deep crustal prerift gabbros have been identified towards the COT, although their thickness apparently varies considerably through boudinage (Manatschal et al. 2001; Mu¨ntener & Herrmann 2001)). Although the entire crust is brittle, this does not exclude the presence of the lower crust as explained above. No unequivocal
96
T. J. RESTON
lower crust has been sampled along the drilling transect at the DGM, but basement rocks have only been sampled in one place there (granodiorites were recovered from the top of a fault block during Galinaute Dive 11), and the gabbros described by Scha¨rer et al. (1995) slightly further to the north have been interpreted as lower crustal. At the IAP margin, the anorthosites, mafic granulites and tonalities sampled at sites 1067 and 900 are typical lower crustal lithologies and record temperatures and pressures appropriate for a lower crustal origin (e.g. Manatschal et al. 2001), yet on the restored section are at a depth of only about 5 km – they were within 5 km of the surface prior to the latest phase(s) of faulting. Given that the brittle upper crust at the start of rifting had a thickness of about 10 km (Fig. 3), it is hard to see how such a fault slice could sample rocks from the deep crust without previously thinning the crust by brittle faulting, i.e. requiring polyphase faulting. As noted above, DDS can only realistically thin the crust without faulting beneath the brittle upper crustal layer, yet the restored thickness of the deep margins West of Iberia is less than the likely original thickness of this layer. I conclude that, although there is undoubtedly both an apparent extension discrepancy and lithospheric-scale DDS at many margins including West Iberia, it remains difficult to explain the extension discrepancy by DDS alone at the West Iberian margin. This does not imply, however, that it cannot be important at other margins, subject to the issues discussed above.
Polyphase faulting Even after restoring the extension along the blockbounding faults West of Iberia, the crustal section is considerably thinner than 10 km. This is not simply a function of pushing the toe of a large fault block underneath the restored section, as for instance at the IAP margin (Fig. 4), depth to top basement (and hence presumably crustal thickness) does not increase significantly for the next 80 km moving landward: there is no suitable large fault block. One possible solution is the occurrence of unrecognized polyphase faulting. Only if the prerift configuration can be recognized and distinguished from a prefaulting sequence (which may predate only the latest faults) and shown to be continuous at the top of the fault blocks on both sides of the rift –ocean can such a scheme be ruled out: without excellent stratigraphic control, an extension discrepancy cannot be proven to exist. A problem at most margins is that the sedimentary sequences are only sampled at the margin edge, i.e. in shallow water. Correlations with the units in the deep margin are partly carried out in terms of reflection
character, partly through jump correlations across large fault blocks/grabens and partly through the apparent relationship of units to faulting. This may well prove that certain units are prefaulting (or post-faulting) on a fault block scale, but does not imply that they are prerift or post-rift on a margin scale. Polyphase faulting is perhaps to be expected as the crust is progressively more and more extended. Simple mechanical considerations such as the Mohr–Coulomb criteria for shear failure (internal friction coefficient 0.85; cohesion, e.g. Scholz 1990) indicate that normal faults should form at angles of approximately 658 (Fig. 3). Once a fault has formed, the development of weak fault gouge (friction coefficient of perhaps 0.6) and the loss of cohesion along the fault may mean that the fault can stay active to angles of perhaps 358, allowing extension along faults to be accommodated by progressive back-rotation as in the domino model (Fig. 15). Beyond that in classical fault mechanics we would expect the fault to lock-up and a new fault to develop. The observation that most normal faulting earthquakes occur at dips between about 708 and 308 (e.g. Jackson 1987) supports this conceptual model. A rotation from about 658 to 358 for a simple domino model corresponds to an extension factor of about 1.6 (e.g. Barr 1987). Allowing the faults to initiate at a steeper angle of 708 and remain active to 308 increases the extension factor at which the faults should become inactive to 1.9. At higher values, we might expect faults to lock-up. These calculations are assuming that the fault block rotates rigidly, whereas it is quite probable that the block will internally deform, adding another component, the thinning of the block, that is not controlled by domino rotation. Thus, for the properties we might expect, faults should lock-up close to a extension factor of 2, and extension beyond such extension factors will require either the development of a new fault, or movement along an existing fault at low angle, or some other mechanism other than large-scale pure shear. Assuming that the former occurs, at extension factors beyond 2, movement will be largely accommodated by movement along a second generation of faults, which will rotate both the crustal blocks and the faults further (Proffett 1977). Furthermore, when these faults themselves become strongly rotated, they too should lock-up – at a extension factor of between 3 and 4. Beyond this a third generation of faults may be required, and so on. As discussed above, the spacing of a second generation of faults is likely to be less than that of the first phase, so that the second generation of faults is likely to cut the pre-existing blocks in a variety of places
RIFTED MARGINS BY PROGRESSIVE EXTENSION
97
Fig. 15. Basic domino model from brittle extension, showing fault-block rotation and synrift sedimentary geometries. Although such models are an oversimplification they do make some useful predictions and reveal some interesting characteristics.
(Figs 15 & 16), although perhaps the most likely is where the faulted crust is weakest, i.e. close to the base of the half-graben. In two dimensions, the second generation of faults may dip in the same way as the first direction, or in the opposite direction (Figs 16 & 17). In both cases, the predicted structure is very complicated, and may include: † old fault segments isolated within basement that may become increasingly hard to recognize; † local deep basins may develop, bounded by steep-sided highs. Such basins may be starved of external sediment input, but may collect the products of mass-wasting of highly fractured basement blocks and exposed early synrift sediments; † faulted exposures of synrift sediments; these are likely to be subject to reworking and masswasting, leading to the partial destruction of early synrift geometries.
Other predicted features when the second generation of faults dip in the same direction (Fig. 16) include: † fault surfaces intersecting at approximately 308; † old faults continue to dip in original direction (but at ever lower angles) until b3; a combination of the new and old fault surfaces may be interpreted as a single phase of faulting (partly concave up, partly convex-up), dipping at approximately 208; † old faults rotated to low angles and even to dip opposite to transport direction; † latest faults intersecting the original surface at approximately 908; † steep top basement surfaces and earliest synrift sediments–both of which are likely to be poorly imaged; † preferential reworking of the thicker portions of synrift sediment wedges, which are likely to occur in the footwall to the new faults.
98
T. J. RESTON synrift sediment
β ~ 2 - original faults lock-up, new faults develop
reworking of early synrift
mass-wasting of fractured basement
reworking of early synrift
β ~ 2.6 - early fault blocks dismembered. Rugged topography leads to mass-wasting / reworking of early synrift and isolation of small basins reworking of early synrift
mass-wasting of fractured basement reworking of early synrift
β ~ 4 - Early faults rotated to dip in opposite direction; original top basement and earliest synrift very steep (not imaged?); mass-wasting / reworking will have continued
Fig. 16. Domino model showing effects of further extension, with faults locking up and new faults cutting across. The resulting sedimentary and structural geometries are complicated even in such a simplistic model.
synrift sediment
β ~ 2 - original faults lock-up, new faults develop, dipping in opposite direction
mass-wasting of fractured basement
mass-wasting of fractured basement reworking of early synrift
reworking of early synrift
β ~ 2.6 - early fault blocks dismembered. Rugged topography leads to mass-wasting / reworking of early synrift and of fractured basement and isolation of small basins mass-wasting of fractured basement
reworking of early synrift
mass-wasting of fractured basement
reworking of early synrift
β ~ 4 - Original top basement ~ horizontal but widely-spaced; early synrift forms upward thickening fans; early faults rotated to steep dips (not imageable in basement). Mass-wasting / reworking will have continued
Fig. 17. Domino model in which second-generation faults dip in the opposite direction to initial faults. A whole new set of complications can result.
RIFTED MARGINS BY PROGRESSIVE EXTENSION
Other predicted features when the second generation of faults dip in the opposite direction (Fig. 17) include:
99
Extension discrepancy at the West Iberian margin As noted above, there is an extension discrepancy at the IAP margin (Fig. 4) as the restoration (Fig. 11) of the extension along the block-bounding faults gives a crust that was only about 6 km thick (or c. 8 km thick if Manatschal’s CMB is taken). Two other relevant observations can be made from the restored section. The first is that the restoration successfully places the lithologies drilled in an apparently logical structural sequence, with sites 1069, 1067 and 900 providing an almost vertical section through the thinned crust. However, as discussed above, the thinning of the crust from more than 30 to about 7 km cannot have been by ductile stretching, but must involve faulting not fully resolved. The mafic granulites drilled at site 900, however, occur perhaps 2 km above the Manatschal CMB but directly atop the Whitmarsh CMB, in which they must sample lowermost crust directly above a mantle transition zone. Furthermore, the presence of lower crustal rocks so close to the surface and to upper crustal sediments means that we cannot simply consider these fault blocks as slices through the upper crust transferred from the hanging wall to the footwall during extension along a rolling hinge model (Buck 1988). Similarly, the presence of deep crustal rocks in these small fault blocks is hard to explain using models in which the crust was simply thinned by the removal of the toe of one or more large fault blocks out to the east. Such an interpretation also conflicts with the inference that the crust to the east of Site 901 is not appreciably thicker for a distance of 50 km (Fig. 4). Second, after the restoration, weak basement reflections become aligned and appear to be associated with residual offset of the Whitmarsh CMB. I interpret this as an anastomosing network of earlier faults, evidence that more that one phase of faulting controlled the evolution of this margin. Unfortunately, uncertainities in the interpretation and in the direction of extension mean that it is
† different generations of faulting intersecting at approximately 908; † later faults intersecting the top basement at about 308, and rotating this to a lower angle; † apparently inverted wedges of early sediments, and the reversal of thickening directions; † preferential reworking of the thinner portions of synrift sediment wedges, which are likely to occur in the footwall to the new faults. The latter two effects may make identification of the synrift particularly difficult. These structures are, however, far simpler than those likely to occur in the real world, when polyphase faulting is likely to be 3D, is unlikely to be regularly spaced and subparallel, and may have variable polarity. Furthermore, the sketches shown in Figures 16 & 17 only show two stages of faulting, corresponding to a extension factor of, at most, 4. At the feather edge of the continents, crustal extension factors may in contrast exceed 10 as the crust is thinned to a few kilometres or less. In such cases, several other phases of faulting are to be expected, as observed in a simplified cross-section through part of the Betics (Booth-Rea et al. 2003) (Fig. 18) that has undergone post-orogenic extension. Here, a section of crust initially 20 km thick has been extended and thinned to 1 km by at least seven phases of normal faults. In the process, deep ductile crust has moved entirely into the brittle field and younger faults have cut and offset older faults. Although such structures may be identified after detailing mapping at outcrop scale, on seismic sections the recognition of polyphase faulting will be far more difficult on a seismic image, especially a time (i.e. distorted) section, and particularly when the margin is sediment starved, or when the early synrift is extensively reworked during later faulting, and the sedimentary geometries observed are likely to give a very incomplete picture of the rift evolution.
1 km
1 km
1
2
3
4
5
6
7
phase of faulting
Fig. 18. Simplified cross-section from the Betics (Spain), showing extensional fault geometries produced during thinning of this section from more than 20 down to 1 km (after Booth-Rea et al. 2003). Lithological information left off for clarity. How many of these faults would be recognized on a seismic section?
6.9
LC
5.2
6.1
3.7
8.0
6.7
4.2 4.9 6.1
6.9
5.2
7.8
5.8
5.2 5.2 5.7 6.7
100
110
Distance [km]
120
4 km streamer
120
1 km streamer
130
130
5.3 6.0 6.2 6.7
6.1
5.0
140
140
5.0
LC
Central Iberian terrane
7.8
4.0
8.0
150
multiples
6.9
MC
UC
East
20
15
10
5
0
0
12 14 16 18 20
14
16
18
20
Depth [km]
Fig. 19. Velocity section (numbers P-wave velocity in km s21) and seismic depth section of profile ISE17 across the Galicia Interior Basin (Pe´rez-Gussinye´ et al. 2003) at true scale. Note that this is a depth migration not a time migration so multiples do not come in at twice the travel time of the primary. Key here is the presence of late east-dipping faults (bold, solid) cutting and offsetting a gently east-dipping top basement (depth of 90– 110 km). Note also that beneath this east-dipping surface, high-velocity (lower crustal? – very light shading) rocks approach top basement. This can be explained if the gently east-dipping surface (dashed) is an earlier fault that has partially exhumed deep crustal rocks. Other gently east-dipping reflections are similarly interpreted (dashed lines).
22
10
12
8
6
10
8
6
4
90
100
5.0 5.2 5.6 6.0
4.9
7.0
2
80
70
90
6.9
6.4 6.7
4.7 5.6
Ossa Morena terrane
3.15 3.25
6.7
4
80
70
4.13.5 4.3 5.0
mantle
2.9 3.3
7.8
1.8
1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 5.0 5.2 5.4 5.6 5.8 6.0 6.2 6.4
2-D VELOCITY STRUCTURE (OBS) + MCS INTERPRETATION
6.1
UC
5.2
km/s 1.5
2
0
25
20
15
10
5
0
West
Depth [km]
Depth [km]
100 T. J. RESTON
RIFTED MARGINS BY PROGRESSIVE EXTENSION
not possible to meaningfully restore the section further. That the CMB was already offset by faults when the crust had been thinned to approximately 6 km is completely consistent with the expected rheological evolution of the crust, which should become brittle at an extension factor of about 4. Crustal embrittlement may lead to mantle serpentinization, allowing faults such as H and F to remain active at low angles. Continued movement along these resulted in the dismemberment of the crust, final crustal separation and the exposure of the mantle forming the footwall to these faults as top basement. The restored section of LG12 thus requires some mechanism to have prethinned the crust, and provides evidence that this may have involved movement along an earlier phase of faults. There is independent evidence for at least two phases of faulting here. Formation microscanner data (Whitmarsh & Wallace 2001) show that the sediments drilled at Site 1065 were first tilted 158 to the SE during early rifting and then, after a pause of several million years, tilted to the east, consistent with movement along the latest westdipping faults. It is thus not surprising that a 2D restoration along the west-dipping faults does not explain all the observed thinning as these only controlled the latest phase of polyphase 3D extension.
The West Galicia margin Evidence for polyphase faulting is clearer at the West Galicia margin, where it is constrained by stratigraphic control, by the velocity structure and by the seismic image. This segment comprises the Galicia Interior Basin in the east, the Galicia Bank and to the west the Deep Galicia Margin (DGM). The structure of the GIB is constrained by a coincident reflection and wide-angle profile ISE17 (Pe´rez-Gussinye´ et al. 2003; Reston 2005) (Fig. 19). The wide-angle model shows a pronounced velocity layering beneath the Galicia Bank to the west; these layers can be traced to the east beneath the GIB, but appear to show that the locus of upper crustal and lower crustal extension are laterally offset, with lower crustal velocities at shallow depths (upper crust largely absent) at the west flank of the basin, and in contrast an absence of lower crustal material beneath the middle and west flank of the basement. The latest phase of faults defines the current fault blocks. However, although these faults dip to the east, top basement (marked by velocities well above 5 km s21) does not show consistent dips: in places it dips steeply to the west (intersecting the faults at c. 908), elsewhere it dips to the east. Both can be explained if the latest faults cut across an earlier set of east-dipping faults – where top basement dips to the east, it is the trace of an earlier
101
fault; where it dips steeply to the west, it is prerift top-basement that has been rotated to the west by two phases of east-dipping faults. This interpretation is illustrated by the restoration (Fig. 20), which clearly shows that two phases of faulting restores the crust that has acted in a brittle manner to approximately 24 km thickness and the entire crust to approximately 30 km. The restoration also illustrates how two-phases of faulting can explain the presence of high-velocity lower crust at top basement – this occurs where top basement is east-dipping and corresponds to the footwall of a major east-dipping normal fault. In this case, large east-dipping faults have lead to a lateral offset of the locus of upper crustal and lower crustal extension in a localized and specific form of depth-dependent stretching, which is only revealed by the presence of coincident reflection profiles and wide-angle data. To the west of the Galicia Bank is the DGM where Davis & Kusznir (2004) ruled out significant extension along unrecognized faults developed after the formation of the fault blocks imaged. However, they did not consider the possibility of earlier unrecognized faulting, for which available data provide ample evidence. The DGM has been the subject of one ODP leg, two submersible sampling campaigns and several high-quality seismic datasets. From these, a general stratigraphic framework has been constructed (Fig. 21) (Reston 2005). The salient points are that the seismostratigraphic unit 5a (terminology of Mauffret & Montadert 1988) tilted within fault blocks consists of deep-water deposits, and represents sediment deposited after the onset of rifting (as deep water) but before the latest phase of faulting (as tilted within the fault blocks). The synrift nature of 5a is supported by variations in its thickness. First, along an individual profile it appears to thicken across successive fault blocks to a maximum before being thinner in the next block. Second, its thickness also varies considerably between profiles. These characteristics imply that it was deposited in basins controlled by faults and that these basins did not simply run north –south in response to east –west extension but had components of extension oblique to the east –west profiles. The synrift nature of unit 5a implies that at least two phases of faulting occurred at this margin, one before or during deposition of unit 5a and one after. Furthermore, seismic unit 5a is diachronous, younging to the west across the drilling transect: sediment tilted within the blocks at site 640 are the same age as seismic unit 4a (post-faulting) further east at site 641 (Reston 2005). This diachroneity might be explained by an oceanward migration of extension during the second phase of faulting, or by an
102
T. J. RESTON "crestal collapse" graben accommodating flexure of the rollover margin
A: Present-day: simplified seismic interpretation (below postrift section) early synrift reworked
? ?
plastic deformation vertical grey bars - velocity boundary between upper and mid-crust from wide-angle data black-white dashed line - predicted boundary between upper and mid-crust from crustal restoration.
Moho
B: latest phase of faulting restored: end Valanginian ? mass-wasting
mass-wasting
"core complex". Mid-crustal rocks exhumed
"crestal collapse" graben accommodating flexure of the rollover margin ?
10 km
D: early faulting restored: pre-rift configuration (Tithonian) faults steeper away from suture 50
E: pre-rift restoration, including removal of 20% bulk stretching
initial low-angle due to reactivation of Variscan suture? 45
34
35 35¡
upper crust
0
58
2
55
45
45
35¡
4 6
mid crust
8 10 12 14 16 18 20 22 24 26
Depth [km]
Crustal blocks at depth shown rigid but probably deformed plastically during early phases of extension. Note however depth from which blocks ere exhumed.
Including a component of internal block stretching steepens initial faults considerably and increases depth of exhumation
original Moho ?
28
Fig. 20. Two-stage restoration of the section shown in Figure 19. Top: present situation. The bars show the location of the velocity boundary between the upper and lower crust, the through-going dashed line is an initially horizontal marker (see bottom). Note the close match between the two, illustrating that the restoration can explain the presence of high-velocity rocks near the top basement in the centre of the section (Fig. 6) by the exhumation of deep crustal rocks. Middle: restoring the first phase of extension brings gently east-dipping structures into alignment. Bottom; further restoration along these yields a tight reconstruction. Note only the portions of the crust that deformed by brittle faulting are shown; deeper levels deformed by plastic– ductile creep throughout evolution.
additional phase of faulting to the west of the 639– 641 high. If the former were the case, it should be possible to partially restore the section as one phase of migrating faulting (Fig. 22). Restoring the latest phase of movement along the blocks to the west of Site 639 produces a close fitting section less than 5 km thick: the quality of the fit suggests that faulting may have migrated oceanards for these faults, although it may equally well have occurred simultaneously (jump from Fig. 22a to 22d). However, the less than 5 km-thick section is considerably thinner than the Site 639 fault block – attempting to restore the section further by sliding the section west of 639 level with the top of block 639 produces a large misfit at the base of the blocks (Fig. 22e). Either the CMB is offset in a way suggesting that the restoration has gone too far, or the blocks must be very strongly distorted.
Removing further subseismic extension associated with the latest faults (Marrett & Allmendinger 1991) does not help unless it is both extreme and restricted to the small blocks west of Site 639. A solution to this misfit is to thicken the section west of 639 further along a set of intermediate age faults before attempting to restore movement along the older fault I at the west flank of the 639 block (Fig. 22f) (Reston 2005). Reston (2005) pointed out that this further restoration is only illustrative as the 3D geometry and sense of movement along the earlier faults is poorly constrained, but Figure 22 shows that it is feasible and apparently necessary. This interpretation is completely consistent with the diachroneity of seismostratigraphic unit 5a; west of 639 the latest block-bounding faults represent a later phase of faulting than that forming the west flank of the 639 block itself. Invoking an additional
RIFTED MARGINS BY PROGRESSIVE EXTENSION
112
latest faulting ?? Barremian- Aptian 5a sandstone sandstone ? 5a uncertain age
Aptian
rift durations
Sites 639 ,638 and 641
Site 640
Gal 11
breakup ?
WDGM
121 Barremian 127 Hauterivian 132 Valanginian 137 Berriasian 144 EDGM Tithonian + GIB Ma basement
deep-water limestone: Tithonian Valanginian?
5
granodiorite
B
103
latest fa
ulting d iachron ous synrift blocks rotated within is diach ronousfault
641: Aptian-Barremian 4a turbididtes, post local faulting 638: 250 m.sandstone Barremian , Hauterivian, Valanginian 5a 639: Valanginian marls, sandstone.
rift onset ? 5b 639: Dolomites - shallow water; Tithonian limestones -sandstones
pre-rift
B ? 639: rhyolites
seafloor 6 8
6
Depth
S
P
10
shot
2200
2300
2400
km
CMB? 2500
2600
2700
2800
Site 639
2900
Site Site 638 641
Dive Site 11
*
seafloor
Depth
km 10
serpentinized mantle
GP11
8
S
S
6
6
Depth 8
S
S
10
S5
S4
8
S6
km
10
GP101 6 8
seafloor 6
P
?
S
S
10
shot
Depth
6
km
8 10
1500
1600
8
km
serpentinized mantle
GP12
Depth
S
10
CMB 1700
Site 640
1900
2000
2100
2200
2300
seafloor 6
?
S
8
S
S
10
GP102 5 km
ODP Leg 103 sites
unit 4a (post- local faulting, pre-breakup)
post-breakup (unit 3) unit 4b (latest syn-faulting)
unit 5 (early synrift)
basement
Fig. 21. Summary of drilling (ODP Leg 103) and diving results across the Deep Galicia margin (DGM). Bottom: basic structure of the margin (in depth, modified after Reston et al. 1996) where the sampling took place. Note the division of the margin into seismostratigraphic units based on relationship to local faulting. Unit 5a: deep-water sediments rotated within the fault blocks and hence interpreted as early synrift – note that this sequence is not constant thickness, and thus not prerift; 4b: syntectonic wedges (nowhere sampled), deposited during motion on the adjacent fault; 4a: infill of half-graben topography after movement on a local fault but probably before breakup. Top: sampling results, defined in terms of the seismo-stratigraphic units: note how unit 5a is clearly diachronous – at Site 640 it is of the same age as the overlying unit 4a further east (Site 641). From this we can deduce at least three phases of faulting: one predating unit 5a across the entire section; one post-dating unit 5a in the east but predating it in the west; and one post-dating unit 5a in the west. Timescale (Gradstein et al. 1994; Pe´rez-Gussinye´ et al. 2003) shows duration of faulting in the Galicia Interior Basin (GIB), and the east and west DGM.
phase of faulting west of 639 also explains the dramatic drop in the size of the fault blocks west of 639. Finally, the latest fault blocks cut down to one of the earlier array of faults that formed a tectonic CMB and thus became a weak serpentine detachment (S): the last two phases of faulting at the DGM accompanied the serpentinization of the uppermost mantle (crust entirely brittle once thinned to 7 km) and the development of the S detachment.
The restoration with two phases of faulting is, however, still only a partial restoration. Movement occured along the faults that controlled the deposition of the deep-water units drilled at Site 639, and 639 has still yet to be restored; as noted above the geometry of these faults is not known and they may run oblique to the margin and the profiles. It is worth noting that the restored section has a similar thickness to the thinnest crust under
104
T. J. RESTON
Dive Gal-11
(a) 5 km
Proj. 640
638
IV
(b)
serpentinized mantle
base of postrift section
III base of late synrift (Aptian ?)
(c) base of early synrift - Valanginian - Barremian ? latest faults
(d)
II
S
earlier faults - Hauterivian?
I-IV
I
timing of deposition of latest synrift wedges
how much earliest synrift ?
F
(e)
serpentinizing mantle
130 % extension removed
space problem - light grey blocks need to be restored further before pushing back
Fig. 22. Partial restoration (modified from that of Reston 2005) of GP101 (Fig. 3), carried out sequentially (youngest faults to the west) Even restoring this section assuming an oceanward progression of faulting in a rolling hinge model it is clear that S was active at low angle. Misfit in E means that the section west of 639 needs to be further thickened before restoring movement along fault I. Reston (2005) suggested that movement along the interpreted earlier faults (black dashed) could solve this problem (F).
the GIB, where both the structural geometry and velocity structure require two phases of faulting. Thus, at the DGM, the stratigraphic and structural evidence indicates two phases of faulting (one for the entire DGM section, one just west of the 639 block) since the deposition of the early synrift section. The double restoration gives a crustal thickness of approximately 7 km, implying that all the faulting occurred after the onset of mantle serpentinization and was associated with
the development of the S reflector. Further landward (e.g. below the 639 high), the S reflector is not present. Within the GIB, the velocity structure and seismic image are consistent with two, but not with one, phase of faulting; the double restoration increases the thickness from about 8 to about 30 km. The rift history of the Galicia margin can be summarized. Rifting started in the Tithonian – Berriasian and two phases of faulting led to the formation of the Galicia Interior Basin and a similar
Fig. 23. Simplified 2D cartoon illustrating the key steps in the progressive extension of the lithosphere leading to continental break-up. (a) Early rift conditions up to the initiation of a second generation of faults (dashed). Initial decoupling zones at the base of the upper and midcrust (grey) thin, and the upper crust becomes increasingly coupled to the lower crust in the centre of the rift. Extension of the strong core of the lower crust and of the strong mantle lid (cross-hatching) is by necking/boudinage, the latter controlled by outward dipping shear zones that displace deep mantle rocks towards the rift axis. (b) An extension factor of 4 in the rift axis. The entire crust has now become brittle, mantle serpentinization (dark grey) has started and a weak detachment develops at the base of the crust. A third generation of faults cut down to this detachment (bold). (c) Crustal separation is imminent and will expose the underlying mainly deep lithospheric mantle. (d) After crustal separation, continued unroofing of the lithospheric mantle may require displacement along landward-dipping shear zones of such mantle from beneath the edges of the rift towards the COT. Complications owing to late igneous activity, 3D nature of extension and order of movement along faults (e.g. rolling hinge) are not shown for reasons of clarity.
RIFTED MARGINS BY PROGRESSIVE EXTENSION
Fig. 23.
105
106
T. J. RESTON
deep basin at the future DGM by the Hauterivian. Rifting now stopped within the GIB and on the west flank of the Galicia Bank, and focused instead within the deep basin to the west. Two more phases of faulting accompanied by the serpentinization of the uppermost mantle and the development of the S detachment here led to final crustal separation and the unroofing of mantle peridotites within the COT. Sea-floor spreading probably started in the Aptian.
Closing remarks As the preceding discussions have shown, I suggest that during progressive extension leading to continental breakup at a cool non-volcanic margin, such as west Iberia, lithospheric extension evolves through a number of main steps (Fig. 23). It is this evolution that distinguishes continental margins from continental rifts, which are subject to far smaller degrees of extension. Particularly important are polyphase faulting and the rheological evolution of the lithosphere (with implications for DDS, for crustal embrittlement, for mantle serpentinization and for final crustal separation along detachment faults rooting in serpentinized mantle). † Extension of rheologically layered lithosphere: faulting in the upper crust, and boudinage of both the lower crust (some component of crustal DDS) and of the uppermost mantle. Extension may focus where lower crustal and mantle boudinage are in phase. During this phase, DDS of the whole lithosphere may start through the displacement of the deepest lithospheric mantle towards the basin antiroot along outward-dipping shear zones. Symmetry at this stage is likely to be controlled by strain rate, and by heterogeneities in initial structure. † At an extension factor of 2, the original normal faults should lock-up and new faults develop (Fig. 23a). A failure to recognize such polyphase faulting may explain the extension discrepancy. At about the same time, or shortly after, the lower crust becomes increasingly coupled to the upper crust. When this coupling is complete, lower crustal boudinage is replaced by faulting, and DDS within the crust will be limited to local asymmetries associated with the largest offset faults. Coupling between the crust and mantle should increase, and displacement of lowermost lithosphere continues towards axis of rift. † Extension focuses strongly towards the site of eventual breakup as the crust thins to about one-quarter of its original thickness (Fig. 23b). At this stage, the second generation of faults lock-up and a third phase of faulting starts. This is accompanied by whole crustal
embrittlement and complete coupling of the crust to the mantle. The new faults may thus cut through entire crust into the mantle leading to serpentinization and the development of (asymmetric?) serpentine detachments (Fig. 23c). The serpentinization is effective in weakening the strong mantle lid. † Crustal separation and unroofing of deep subcontinental mantle oceanward of the last crustal blocks. Although melt generation might be expected during this stage, it is partially suppressed by a combination of mantle DDS and cool subcontinental mantle; continued unroofing of mantle may be controlled by successive landward-dipping shear zones (Boillot & Froitzheim 2001), which provide a mechanism for the displacement of deep sublithospheric mantel into the widening COT. † The onset of sea-floor spreading is probably controlled by either the final necking of the deep lithospheric mantle within the COT or by the influx of warm asthenosphere. In both cases sea-floor spreading is likely to substantially post-date crustal separation, as indicated by the presence of a broad zone of unroofed mantle within the continent–ocean transition (COT) at several margins (Fig. 1). At the West Iberia margin and at the preserved Alpine margins, we see evidence for all of the above processes. The apparent discrepancy between extension determined from crustal thinning or subsidence and that measurable from faults may be explained if extension is accommodated along multiple generations of faults that are incompletely resolved and which probably accommodated variable extension directions. As the amount and duration of extension increases towards the COT, so does the number of fault generations required: as not all of these can be fully imaged, let alone interpreted, an apparent extension discrepancy results. At the West Galicia margin, multiple phases of faulting are evidenced by the presence of deep-water sediments tilted within the fault blocks (and hence predating the faults that bound those blocks) and supported by structural reconstructions. At the IAP margin to the south, structural reconstructions and the tilting history of fault blocks also document the importance of polyphase faulting. This does not mean that at the West Iberian margin no crustal DDS occurred (e.g. through lower crustal necking and boudinage), just that it is not required to explain the extension discrepancy. Indeed, by analogy with the Alps and the basins offshore the UK, lower crustal boudinage probably did occur during the early evolution of the Iberian margin (Manatschal et al. 2001). Nor does it mean
RIFTED MARGINS BY PROGRESSIVE EXTENSION
that at other margins crustal DDS (including large-scale lower crustal flow) may not be required, but does suggest that polyphase faulting has to be rigorously ruled out before invoking these. At both the West Iberian and Alpine margins, the distribution and thickness of serpentinized mantle beneath the feather edge of the continental crust matches well the predictions of a simple numerical model for the embrittlement of the crust. The thickness of the partially serpentinized zone also matches the predictions of the model, particularly if a transition from pure shear to simple shear occurs when serpentinization starts (Pe´rez-Gussinye´ & Reston 2001). Detachment faults that appear to have been active at low angles, and which indicate a change from crustal-scale pure shear to crustal-scale simple shear, are observed and appear to follow the upper portion of the serpentinized mantle over much of their length. The presence of subcontinental lithospheric mantle unroofed within the COT implies heterogeneous stretching (DDS), which may be controlled by landward-dipping mantle shear zones similar to those imaged offshore the UK and beneath the West Iberian margin. Such DDS may have contributed to the reduction in melting at depth: again by analogy with the extended basins of NW Europe, this heterogeneous mantle extension probably started early during the rift history, and continued until final lithospheric separation and the onset of sea-floor spreading. Finally, the lack of significant igneous activity within the COT may be related to the presence of cool sublithospheric mantle in this region prior to the onset of seafloor spreading. Although there is undoubtedly both an apparent extension discrepancy and lithospheric-scale DDS at many margins, it remains difficult to explain such a discrepancy by DDS alone. At the West Iberian margin, polyphase faulting did occur and may suffice to explain the observed margin structure. However, other margins such as the Exmouth Plateau margin (Driscoll & Karner 1998) remain problematic. Here, the upper crust appears little extended, yet the margin has subsided several kilometres. Some other, as yet unknown, process may be important. This work has been supported by DFG grants Re 873-1, 873-3, 873-6, 873-7 and 873– 8, for which I am very grateful. Contributions by J. Hoffman, D. Klaeschen, C. Ranero, C. Krawczyk, J. Phipps Morgan and, in particular, M. Pe´rez-Gussinye´ made the work possible, although not all of these may agree with every conclusion reached. I would like to thank the organizers of the IMEDL workshop for organizing a superb meeting and especially G. Manatschal and O. Mu¨ntener for wonderful field excursions. Reviews by D. Sawyer, T. Minshull and N. Kusznir helped tighten the arguments.
107
References A LLEN , P. A. & A LLEN , J. R. 1990. Basin Analysis; Principle and Applications. Blackwell Scientific, Oxford. B ARR , D. 1987. Lithospheric stretching, detached normal faulting and footwall uplift. In: C OWARD , M. P., D EWEY , J. F. & H ANCOCK , P. L. (eds) Continental Extensional Tectonics. Geological Society, London, Special Publications, 28, 75– 94. B ESLIER , M.-O., C ORNEN , G. & G IRARDEAU , J. 1996. Tectono-metamorphic evolution of peridotites from the ocean–continent transition of the Iberia Abyssal Plain Margin. Proceedings of the Ocean Drilling Program, Scientific Results, 149. Ocean Drilling Program, College Station, TX, 397– 412. B OILLOT , G. & F ROITZHEIM , N. 2001. Non-volcanic rifted margins, continental breakup and the onset of seafloor spreading: some outstanding questions. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: A Comparison of Evidence From Land and Sea. Geological Society, London, Special Publications, 187, 9– 30. B OILLOT , G. & W INTERER , E. 1988. Drilling on the Galicia margin: Retrospect and prospect. Proceedings of the Ocean Drilling Program, Scientific Results, 103. Ocean Drilling Program, College Station, TX, 809– 828. B OILLOT , G., F ERAUD , G., R ECQ , M. & G IRARDEAU , J. 1989. ‘Undercrusting’ by serpentinite beneath rifted margins: The examples of the west Galicia margin (Spain). Nature, 431, 523 –525. B OOTH -R EA , G., A ZANON , J. M. & G ARCIA -D UENAS , V. 2003. Extensional tectonics in the northeastern Betics (SE Spain): case study of extension in a multilayered upper crust with contrasting rheologies. Journal of Structural Geology, 26, 2039–2058. B OWLING , J. C. & H ARRY , D. L. 2001. Geodynamic models of continental extension and the formation of nonvolcanic rifted continental margins. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: A Comparison of Evidence From Land and Sea. Geological Society, London, Special Publications, 187, 511–536. B OWN , J. W. & W HITE , R. S. 1995. Effect of finite extension rate on melt generation at rifted continental margins. Journal of Geophysical Research, 100, 18,011–18,029. B RAUN , J. & B EAUMONT , C.1987. Styles of continental rifting: Results from dynamic models of lithosphere extension. In: B EAUMONT , C. & T ANKARD , A. J. (eds) Sedimentary Basins and Basin-Forming Mechanisms. Canadian Society of Petroleum Geology Memoir, 12, 241–258. B RUN , J. P. & B ESLIER , M. O. 1996. Mantle exhumation at passive margins, Earth and Planetary Science Letters, 142, 161– 173. B UCK , W. R., 1988. Flexural rotation of normal faults. Tectonics, 7, 959– 973. C AMMARANO , F., R OMANOWICZ , B., S TIXRUDE , L. & L ITHGOW -B ERTELLONI , C. 2005. Towards Inverting Seismic Waveform Data for Temperature and Composition of the Earth’s Upper Mantle. AGU Fall Meeting
108
T. J. RESTON
Supplement, DI41A-1255. American Geophysical Union, Washington, D.C. C ANNAT , M. 1996. How thick is the magmatic crust at slow-spreading ridges? Journal of Geophysical Research, 101, 2847– 2857. C HIAN , D., L OUDEN , K. E., M INSHULL , T. A. & WHITMARSH, R. B. 1999. Deep structure of the ocean– continent transition in the southern Iberia abyssal plain from seismic refraction profiles; Ocean Drilling Program (legs 149 and 173) transect. Journal of Geophysical Research, 104, 7443–7462. C ORNEN , G., B ESLIER , M.-O. & G IRARDEAU , J. 1996. Petrologic characteristics of the ultramafic rocks from the ocean/continent transition in the Iberia Abyssal Plain. Proceedings of the Ocean Drilling Program, Scientific Results, 149. Ocean Drilling Program, College Station, TX, 377– 395. C OWARD , M. 1986. Heterogeneous stretching, simple shear and basin development. Earth and Planetary Science Letters, 80, 325–336. D AVIS , M. & K USZNIR , N. J. 2004. Depth-dependent lithospheric stretching at rifted continental margins. In: K ARNER , G. D. (ed.) Proceedings of NSF Rifted Margins Theoretical Institute. Columbia University Press, New York, 92– 136. D EAN , S. M., M INSHULL , T. A., W HITMARSH , R. B. & L OUDEN , K. E. 2000. Deep structure of the ocean– continent transition in the southern Iberia abyssal plain from seismic refraction profiles; the IAM-9 transect at 40 degrees 200 N. Journal of Geophysical Research, 105, 5859– 5885. D ESMURS , L., M ANATSCHAL , G. & B ERNOULLI , D. 2001. The Steinmann Trinity revisited: mantle exhumation and magmatism along a ocean–continent transition: the Platta nappe, eastern Switzerland. In: W ILSON , R. C. L., W HITMARSH R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: A Comparison of Evidence from Land and Sea. Geological Society, London, Special Publications, 187, 235– 266. D RISCOLL , N. W. & K ARNER , G. D. 1998. Lower crustal extension across the Northern Carnarvon basin, Australia: Evidence for an eastward dipping detachment. Journal of Geophysical Research, 103, 4975–4991. F UNCK , T. J., H OPPER , J. R., L ARSEN , H. C., L OUDEN , K. E., T UCHOLKE , B. E. & H OLBROOK , W. S. 2003. Crustal structure of the ocean-continent transition at Flemish Cap: Seismic refraction results. Journal of Geophysical Research, 108, B11, 2531, doi:10.1029/2003JB002434. G OES , S. & VAN DER L EE , S. 2002. Thermal structure of the North American uppermost mantle inferred from seismic tomography. Journal of Geophysical Research, 107, B3, ETG2, doi:10.1029/2000JB000049. G OSSLER , J. & K IND , R. 1996, Seismic evidence for very deep roots of continents. Earth and Planetary Science Letters, 138, 1 –13. G RADSTEIN , F. M., A GTERBERG , F. P., O GG , J. G., H ARDENBOL , J., VAN V EEN , P., T HIERRY , J. & H UANG , Z. 1994. A Mesozoic time scale. Journal of Geophysical Research, 99, 24,051–24,074. H ARRY , D. L. & S AWYER , D. S. 1992. A dynamic model for extension in the Baltimore Canyon trough. Tectonics, 11, 420– 436.
H E´ BERT , R., G UEDDARI , K., L AFLE` CHE , M. R., B ESLIER , M.-O. & G ARDIEN , V. 2001. Petrology and geochemistry of exhumed peridotites and gabbros at non-volcanic margins: ODP Leg 173 West Iberia ocean–continent transition zone. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds), Nonvolcanic Rifting of Continental Margins: A Comparison of Evidence from Land and Sea. Geological Society, London, Special Publication, 187, 161–189. H OPPER , J. R. & B UCK , W. R. 1998. Styles of extensional decoupling. Geology, 26, 699–720. H UISMANS , R. S. & B EAUMONT , C. 2003. Symmetric and asymmetric lithospheric extension: Relative effects of frictional–plastic and viscous strain softening. Journal Geophysical Research, 108, B10, 2496, doi:10.1029/ 2002JB002026. J ACKSON , J. A. 1987. Active normal faulting and crustal extension. In: C OWARD , M. P., D EWEY , J. F. & H ANCOCK , P. L. (eds) Continental Extensional Tectonics. Geological Society, London, Special Publications, 28, 3 –17. J ARVIS , G. T. & M C K ENZIE , D. P. 1980. Sedimentary basin formation with finite extension rates. Earth and Planetary Science Letters, 48, 42– 52. J AUPART , C. & M ARESCHAL , J.-C. 1999. The thermal structure and thickness of continental roots. Lithos, 48, 93–114. K LEMPERER , S. L. 1988. Crustal thinning and nature of extension in the northern North Sea from deep seismic reflection profiling. Tectonics, 7, 803–821. K OHLER , T. P. & B REY , G. P. 1990. Calcium exchange between olivine and clinopyroxene calibrated as a geothermobarometer for natural peridotites from 2 to 60 kb with applications. Geochimica et Cosmochimica Acta, 54, 2375– 2388. K RAWCZYK , C. M. 1995. Detachment tectonics during continental rifting off the West Iberia margin: Seismic reflection and drilling constraints. PhD dissertation, University of Kiel. K RAWCZYK , C. M., R ESTON , T. J., B ESLIER , M. O. & B OILLOT , G. 1996. Evidence for detachment tectonics on the Iberia Abyssal Plain rifted margin. Proceedings of the Ocean Drilling Program, Scientific Results, 149. Ocean Drilling Program, College Station, TX, 603–615. K USZNIR , N. J. & M ATTHEWS , D. H. 1988. Deep seismic reflections and the deformational mechanics of the continental lithosphere. Journal of Petrology, Special Lithosphere Issue, 63– 87. K USZNIR , N. J., H UNSDALE , R., R OBERTS , A. M. & I SIMM, TEAM . 2005. Norwegian margin depthdependent stretching. In: D ORE´ , A. G. & V INING , B. A. (eds) Petroleum Geology: North-West Europe and Global Perspectives – Proceedings of the 6th Petroleum Geology Conference. Petroleum Geology Conferences. Geological Society, London, 767–783. L AVIER , L. L. & M ANATSCHAL , G. 2003. From upward to downward concave faulting in magma-poor rifted margins: the role of serpentinization and magmatism at the transition from rifting to seafloor spreading. Geophysical Research Abstracts, 5, 11448. L ENARDIC , A. & M ORESI , L. 2000. A new class of equilibrium geotherms in the deep thermal lithosphere of the continents. Earth and Planetary Science Letters, 176, 331– 338.
RIFTED MARGINS BY PROGRESSIVE EXTENSION L E P ICHON , X. & B ARBIER , F. 1987. Passive margin formation by low-angle faulting within the upper crust: The northern Bay of Biscay margin. Tectonics, 6, 133–150. L IZARRALDE , D., G AHERTY , J. B., C OLLINS , J. A., H IRTH , G. & K IM , S. D. 2004. Spreading-rate dependence of melt extraction at mid-ocean ridges from mantle seismic refraction data. Nature, 432, 744–747. M ANATSCHAL , G., F ROITZHEIM , N., R UBENACH , M. & T URRIN , B. D. 2001. The role of detachment faulting in the formation of an ocean-continent transition: insights from the Iberian Abyssal Plain. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: A Comparison of Evidence From Land and Sea. Geological Society, London, Special Publications, 187, 405–428. M ARRETT , R. & A LLMENDINGER , R. 1992. Amount of extension on ‘small’ faults: an example from the Viking graben. Geology, 20, 47–50. M AUFFRET , A. & M ONTARDET , L. 1988. Seismic stratigraphy off Galicia (Spain). Proceedings of the Ocean Drilling Program, Scientific Results, 103, 13–20. M C K ENZIE , D. P. 1978. Some remarks on the development of sedimentary basins. Earth and Planetary Science Letters, 40, 25–32. M C K ENZIE , D. P. & B ICKLE , M. J. 1988, The volume and composition of melt generated by extension of the lithosphere. Journal of Petrology, 29, 625– 679. M C K ENZIE , D. & J ACKSON , J. 2002. Conditions for flow in the continental crust. Tectonics, 21, 6. M INSHULL , T. A., D EAN , S. M., W HITE , R. S. & W HITMARSH , R. B. 2001. Anomalous melt production after continental breakup in the southern Iberia Abyssal Plain. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: A Comparison of Evidence From Land and Sea. Geological Society, London, Special Publications, 187, 537– 550. M U¨ NTENER , O. & H ERRMANN , J. 2001. The role of the lower crust and continental upper mantle during formation of non-volcanic passive margins: evidence from the Alps. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Nonvolcanic Rifting of Continental Margins: A Comparison of Evidence From Land and Sea. Geological Society, London, Special Publications, 187, 267– 288. M U¨ NTENER , O., H ERRMANN , J. & T ROMMSDORF , V. 2000. Coooling history and exhumation of lowercrustal granulite and upper mantle (Malenco, Eastern Central Alps). Journal of Petrology, 41, 175–200. N AGEL , T. J. & B UCK , W. R. 2004. Symmetric alternative to asymmetric rifting models. Geology, 32, 937– 940. O’R EILLY , B. M., H AUSER , F., J ACOB , A. W. B. & S HANNON , P. M. 1996. The lithosphere below the Rockall Trough: wide-angle seismic evidence for extensive serpentinisation. Tectonophysics, 255, 1–23. P E´ REZ -G USSINYE´ , M. & R ESTON , T. J. 2001. Rheological evolution during extension at passive non-volcanic margins: onset of serpentinization and development of detachments to continental breakup. Journal of Geophysical Research, 106, 3961– 3975.
109
P E´ REZ -G USSINYE´ , M., R ANERO , C. R., R ESTON , T. J. & S AWYER , D. 2003. Mechanisms of extension at nonvolcanic margins: Evidence from the Galicia Interior Basin, west of Iberia. Journal of Geophysical Research, 108, EPM 6-1–EPM 6–19, doi:10.1029/ 2001JB000901. P E´ REZ -G USSINYE´ , M., R ESTON , T. J. & P HIPPS M ORGAN , J. 2001, Rheological and magmatic evolution during extension at passive nonvolcanic margins: the effect of initial lithospheric structure. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: A Comparison of Evidence From Land and Sea. Geological Society, London, Special Publications, 187, 551–576. P HIPPS M ORGAN , J., M ORGAN , W. J., Z HANG , Y.-S. & S MITH , W. H. F. 1995. Observational hints for a plume-fed, suboceanic asthenosphere and its role in mantle convection. Journal of Geophysical Research, 100, 12,753– 12,767. P ICKUP , S., W HITMARSH , R. B., F OWLER , C. M. R. & R ESTON , T. J. 1996. Insight into the nature of the ocean-continent transition off West Iberia from a deep multichannel seismic reflection profile. Geology, 24, 1079–1082. P ROFFETT , J. M. 1977. Cenozoic geology of the Yerington dsitrict, Nevada, and implications for the nature of Basin and Range faulting. Bulletin of the Geological Society of America, 88, 247–266. R ESTON , T. J. 1988. Evidence for extensional shears zones in the lower crust. Tectonics, 7, 929 –945. R ESTON , T. J. 1990. The lower crust and the extension of the continental lithosphere: kinematic analysis of BIRPS deep seismic data. Tectonics, 9, 1235–1248. R ESTON , T. J. 1993. Evidence for extensional shear zones in the mantle offshore Britain, and their implications for the extension of the continental lithosphere. Tectonics, 12, 492– 506. R ESTON , T. J. 1996. The S reflector west of Galicia: The seismic signature of a detachment fault. Geophysical Journal International, 127, 230–244. R ESTON , T. J. 2005. Polyphase faulting during the development of the west Galicia rifted margin. Earth and Planetary Science Letters, 237, 561– 576. R ESTON , T. J. & P HIPPS M ORGAN , J. 2004. The continental geotherm and the evolution of rifted margins. Geology, 32, 133– 136. R ESTON , T. J., G AW , V., P ENNELL , J., K LAESCHEN , D., S TUBENRAUCH , A. & W ALKER , I. 2004. Extreme crustal thinning in the south Porcupine Basin and the nature of the Porcupine Median High: Implications for the formation of non-volcanic rifted margins. Journal of the Geological Society, London, 161, 783– 798. R ESTON , T. J., K RAWCZYK , C. M. & K LAESCHEN , D. 1996. The S reflector west of Galicia (Spain); evidence from prestack depth migration for detachment faulting during continental breakup. Journal of Geophysical Research, 101, 8075–8091. R O¨ HM , A. H. E., S NIEDER , R., G OES , S. & T RAMPERT , J. 2000. Thermal structure of continental upper mantle inferred from S-wave velocity and surface heat flow. Earth and Planetary Science Letters, 181, 395– 407.
110
T. J. RESTON
R UTTER , E. 1986. On the nomenclature of modes of failure transitions in rocks. Tectonophysics, 122, 381– 387. S AYERS , J., S YMONDS , P. A., D IREEN , N. G. & B ERNADEL , G. 2001. Nature of the continent – ocean transition on the non-volcanic rifted margin of the central Great Australian Bight. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: A Comparison of Evidence From Land and Sea. Geological Society, London, Special Publications, 187, 51–76. S CHA¨ RER , U., K ORNPROBST , J., B ESLIER , M.-O., B OILLOT , G. & G IRARDEAU , J. 1995. Gabbro and related rock emplacement beneath rifting continental crust; U –Pb geochronological and geochemical constraints for the Galicia passive margin (Spain). Earth and Planetary Science Letters. 130, 187–200. S CHOLZ , C. H. 1990. The Mechanics of Earthquakes and Faulting. Cambridge University Press, Cambridge. S EIFERT , K. & B RUNOTTE , D. 1996. Geochemistry of serpentinized mantle peridotite from site 897 in the Iberia Abyssal Plain. Proceedings of the Ocean Drilling Program, Scientific Results, 149. Ocean Drilling Program, College Station, TX, 413– 421. S IBUET , 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. T UCHOLKE , B. E., S AWYER , D. S. & S IBUET , J. -C. 2007. Breakup of the Newfoundland– Iberia rift. Continental. In: K ARNER , G. D., M ANATSCHAL , G. & P INHEIRO , L. M. (eds) Imaging, Mapping and Modelling Lithosphere Extension and Breaks. Geological Society, London, Special Publications, 282, 9– 45. V OORHOEVE , H. & H OUSEMAN , G. 1988, The thermal evolution of lithosphere extending on a low-angle detachment zone. Basin Research. 1, 9– 45. W ERNICKE , B. 1985. Uniform-sense normal simple shear of the continental lithosphere. Canadian Journal of Earth Sciences, 22, 331–339. W HITE , N. 1990. Does the uniform stretching model work in the North Sea? In: B LUNDELL , D. J. &
G IBBS , A. D. (eds) Tectonic Evolution of the North Sea Rifts. International Lithosphere Program, 181, 217–239. W HITE , R. S. & M C K ENZIE , D. P. 1989. Magmatism at rift zones; the generation of volcanic continental margins and flood basalts. Journal of Geophysical Research, 94, 7685– 7729. W HITMARSH , R. B. & W ALLACE , P. J. 2001. The rift-to-drift development of the west Iberia nonvolcanic continental margin: a summary and review of the contribution of Ocean Drilling Program Leg 173. In: B ESLIER , M.-O., W HITMARSH , R. B., W ALLACE , P. J. & G IRARDEAU , J. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 173. Ocean Drilling Program, College Station, TX, 1– 36. W HITMARSH , R. B., D EAN , S. M., M INSHULL , T. A. & T OMPKINS , M. 2000. Tectonic implications of exposure of lower continental crust beneath the Iberia Abyssal Plain, northeast Atlantic Ocean: geophysical evidence. Tectonics, 19, 919–942. W HITMARSH , R. B., M ANATSCHAL , G. & M INSHULL , T. A. 2001. Evolution of magma-poor continental margins from rifting to seafloor spreading. Nature, 413, 150–154. W HITMARSH , R. B., W HITE , R. S., H ORSEFIELD , S. J., S IBUET , J., R ECQ , M. & L OUVEL , V. 1996. The ocean–continent boundary off the western continental margin of Iberia: Crustal structure west of Galicia Bank. Journal of Geophysical Research, 101, 28291–28314. W ILSON , R. C. L., M ANATSCHAL , G. & W ISE , S. 2001. Rifting along nonvolcanic passive margins: stratigraphic and seismic evidence from the Mesozoic of the Alps and Western Iberia. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: A Comparison of Evidence From Land and Sea. Geological Society, London, Special Publications, 187, 429–452. Z ELT , C. A., S AIN , K., N AUMENKO , J. V. & S AWYER , D. S. 2003. Assessment of crustal velocity models using seismic refraction and reflection tomography. Geophysical Journal International, 153, 609–626.
Roles of lithospheric strain softening and heterogeneity in determining the geometry of rifts and continental margins R. S. HUISMANS1,2 & C. BEAUMONT1 1
Geodynamics Group, Department of Oceanography, Dalhousie University, Halifax (Nova Scotia), Canada, B3H 4J1
2
Current address: Department of Earth Sciences, Bergen University, 5006 Bergen, Norway (e-mail:
[email protected]) Abstract: Plane strain thermo-mechanical finite-element model experiments are used to investigate the effects of frictional–plastic strain softening and inherited weakness on the style of lithospheric extension. The model results are compared with the Newfoundland– Iberia conjugate rifted margins with the goal of understanding the lithospheric properties that controlled their evolution during rifting. Our proposition is that coupling between the plastic–viscous layering, acting together with frictional– plastic strain softening localized on inherited weak heterogeneities, can explain the initial wide rift and distributed rift basins that are later abandoned in favour of a narrow rift in which mantle lithosphere is exhumed to the surface. The models comprise uniform composition viscous and plastic layers in which focused deformation is nucleated on either a single weak ‘seed’ or a statistical white noise distribution of inherited strain. Strain softening of frictional–plastic layers acts as a positive feedback mechanism that creates localized shear zones from the inherited weak heterogeneities. The sensitivity of deformation to the choice of softening parameters and the type of inherited noise is examined in cases where the deeper part of the crust is either weak or strong. Lithosphere-scale models with a single weak seed exhibit a range of asymmetric and symmetric rifting modes that are mostly determined by the feedback between two primary controls, coupling between the plastic and viscous layers and strain softening. Decreasing and increasing the rifting velocity can change the mode, and asymmetry is strongest in models with low rifting velocities and a strong lower crust. Analysis of equivalent simple-bonded plastic– viscous two-layer models using the minimum rate of dissipation principle demonstrates that the mode selected depends on the division of the dissipation between the layers. Criteria developed on minimizing the total dissipation show how mode selection changes with increasing viscosity, or rifting velocity, from the: asymmetric plug or half-graben (AP) mode; through the symmetric plug or graben (PS) mode, to the distributed pure shear (PS) mode. Numerical models confirm these results. Models with statistical white-noise-inherited strain have similar modes to those with a single seed. In addition, modes with multiple sets of shear zones develop in the plastic layer for a range of intermediate parameter combinations. We believe that distributed noise in combination with a weak lower crust and slow extension can produce model results in accord with general features of the Newfoundland– Iberia conjugate margins; an initially distributed wide rift mode, followed by a late-stage narrow rift with a significant component of mantle exhumation.
Despite the large number of studies of continental rifts and rifted continental margins, many of which have produced high-quality data (Keen et al. 1987a, b, 1989; Mutter et al. 1989; Boillot et al. 1992; Sibuet 1992; Brun & Beslier 1996; Louden & Chian 1993; Dean et al. 2000; Funck et al. 2003, 2004; Hopper et al. 2004), we still lack a unified understanding of the mechanical and thermal processes that control their extensional geometry. This problem is compounded by the wide range of styles that include both non-volcanic and volcanic types, in which the latter exhibits direct evidence of voluminous mafic extrusives, and/or indirect evidence of intruded and underplated mafic magmas. In addition, the relative importance of active v. passive rifting, that is whether rifting is driven by active mantle upwelling
or whether mantle upwelling is a passive response to lithospheric extension (Salveson 1978; Sengor & Burke 1978; Turcotte & Emmerman 1983; Ruppel 1995; Huismans et al. 2001) remains to be determined. Important contributions to the problem of passive rifting leading to non-volcanic rifted margins defined possible end-member styles of lithospheric extension and their associated geometries, pure shear (McKenzie 1978), simple shear (Wernicke Burchfield 1982; Wernicke 1985) and combinations of these styles, for example Lister et al. (1986), referred to here as compound models (Fig. 1a–c). These kinematic models subsequently served as templates for the interpretation of observations. A primary focus in these interpretations has been whether rifting is symmetric, corresponding to the pure shear template, or
From: KARNER , G. D., MANATSCHAL , G. & PINHEIRO , L. M. (eds) Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, London, Special Publications, 282, 111– 138. DOI: 10.1144/SP282.6 0305-8719/07/$15.00 # The Geological Society of London 2007.
112
R. S. HUISMANS & C. BEAUMONT
Narrow Rift Mode (b) Asymmetric ‘Simple Shear’ Models
(a) Symmetric ‘Pure Shear’ Models Crust
Crust
Mantle
Mantle
Asthenosphere
Asthenosphere
(c) Compound Models Crust Mantle Asthenosphere
(d)
Wide Rift Mode
Crust Mantle
Asthenosphere
Fig. 1. End-member styles of rifting; symmetric, asymmetric and compound narrow models, and the wide rift mode. What controls these models of lithosphere extension; wide v. narrow, symmetric v. asymmetric? Can mode transitions be expected between asymmetric and symmetric modes, and between wide and narrow rifting?
asymmetric as predicted by the simple shear template (Fig. 1a, b). More generally, our lack of understanding of passive rifting at non-volcanic continental rifts and rifted margins can be expressed in terms of the following questions. † Can the geometrical styles of natural continental rifts and rifted margins be classified according to modes, such as the narrow, wide and core complex rift modes proposed by Buck (1991) (Fig. 1 a –d). † If so, how do the kinematic template styles of extension listed above, and their associated symmetries and asymmetries (Fig. 1), relate to the modes proposed by Buck (1991) (Fig. 1)? † If natural rifts can indeed be classified according to modes of extension, what controls the selection of these modes? † Conversely, if natural rifts exhibit various geometries that defy any simple mode classification, what controls this variability? Numerical modelling has the potential to provide at least partial answers to these questions. Model experiments can be designed to test the
role of a range of factors and processes on the styles of lithospheric extension. If the models are reasonable, albeit simplified, representations of their natural counterparts, the model results can indicate which processes are most likely to be important and how processes compete and tradeoff in different circumstances. We have chosen to focus on aspects of the rifting problem that we believe were important in the development of the Newfoundland–Iberia conjugate margin system and we address this specific application at the end of the paper. More generally, observations confirm that faulting and deformation on shear zones are important during rifting. At the small scale, deformation on faults and shear zones is inherently asymmetric. However, the question of the role of shear zones on the large-scale asymmetry of passive margins is still unanswered. This is partly a result of the lack of good constraints on the structure of the continental mantle lithosphere and its evolution during extension. Nevertheless, data do point to asymmetric geometries of conjugate margins (Boillot et al. 1992; Sibuet 1992; Louden & Chian 1999), and, for example, large-scale low-angle detachments and features
GEOMETRY OF RIFTS AND RIFTED MARGINS
interpreted to be high-angle shear zones that extend into the upper mantle (Brun & Gutscher 1992; Whitmarsh et al. 2001). There is some indication of patterns, or modes, in margin geometries. The significant variability in crustal-scale geometry that commonly occurs over quite short distances along strike on many margins poses an additional problem (see, for instance, the Iberia –Newfoundland margin system: Keen et al. 1987a, b; Torne et al. 1994; Funck et al. 2003; Dean et al. 2000; Pe´rez-Gussinye´ et al. 2003). Either this strike variability, or margin segmentation, is an inherent and perhaps predictable consequence of the mechanics of rifting in three dimensions, or it represents an effect of inherited inhomogeneous lithospheric properties, ‘noise’, in the system. For these reasons our general focus in the models described here is on processes that create shear zones and lead to mode selection, and on the effects of noise. We first assume that the lithosphere can be represented by a prototype uniform-layered model comprising frictional –plastic and viscous layers. We then consider the effects of different factors or processes on the style of extension of this model lithosphere. This approach is designed to answer the following particular questions. † What is the effect of strain-softening of the frictional –plastic parts of the system on the geometry during extension and does this feedback into mode selection? † Can the mode selection be explained by a general governing principle, specifically the minimum rate of dissipation of energy? † What is the effect of inherited heterogeneity, ‘noise’, in the model crustal properties on the style of extension? Does mode selection still occur in the presence of noise, or does noise lead to incoherent extension that is merely an expression of the noise? In section on ‘Thermo-mechanical models of lithosphere extension’ it is shown that strainsoftening leads to mode selection, and to asymmetric and symmetric styles of lithospheric extension. By analysing simple two-layer models we then demonstrate in the section on ‘Analysis of simple models using the principle of minimum energy dissipation’ that the minimum dissipation principle can be used as a guide to mode selection, at least in the numerical models. The section on ‘systems with statistical heterogeneity’ contains a preliminary treatment of the effects of distributed noise in the system on mode selection, and on how noise can lead to wide and narrow styles of lithospheric extension. Lastly, in ‘implications for the Iberia –Newfoundland conjugate margin system’ we ask whether what has been learned from the numerical models can be applied to the
113
Newfoundland –Iberia conjugate margins to determine the controls on the tectonic style of this system. These sections are prefaced by brief reviews of processes that result in strain-dependent strength of rocks, the role and effect of inherited heterogeneity, and the two ways, deterministic and statistical, that this heterogeneity is introduced into the models. The section on ‘Thermo-mechanical numerical model’ describes the numerical models.
Strain-dependent strength of rocks and its representation in the numerical models The strength of faults and shear zones in the crust and mantle lithosphere on geological timescales is a controversial subject. They are interpreted to be either much weaker (Che´ry et al. 2001; Provost & Houston 2003) or of similar strength (Scholz 2000) to adjacent relatively non-deforming regions. Here, we accept that faults may be very weak and outline mechanisms that produce weak faults and shear zones. We then describe a simple parametric representation of these processes that we use in the numerical models. The frictional, or brittle, strength of dry rocks is represented to a first approximation by the Coulomb yield criterion for incompressible deformation in plane strain: ðJ20 Þ1=2 ¼ p sin f þ c cos f
ð1Þ
where (J20 )1/2 is the second invariant of the at yield deviatoric stress tensor, c is the cohesion, f is the internal angle of friction and p is the dynamic pressure (mean stress). Several mechanisms may contribute to strain softening, that is a reduction of the strength of the rock (deviatoric stress at yield, equation 1), with increasing strain but their relative and absolute importance is poorly constrained. At the scale of the crust, the brittle strength is dominated by the pressuredependent term (equation 1). Consequently, cohesion loss results only in a limited reduction of the strength of the brittle layer in the order of 2–10% depending on the ambient dynamical pressure and the depth of the brittle–ductile transition in the crust (Fig. 2a). Other mechanisms, such as formation of gouge and mineral transformations which reduce f with increasing deformation, can potentially result in much larger frictional strength reduction. For example, experimental work (Bos & Spiers 2002) on quartz–feldspar aggregates and their analogues has shown that mineral transformations in granitic rocks in the presence of water may produce mica-rich shear zones with very low effective internal angles of friction (f ¼ 38–108), resulting in a strength reduction of the order of 50–80% (Fig. 2b). Similar deformationassisted mineral transformation mechanisms have
114
R. S. HUISMANS & C. BEAUMONT
(a) Cohesion loss 0
C = 20 MPa
φ = 30o
Depth (km)
C=0 10
Brittle-ductile transition 20
30 0
100
200
300
Yield Strength (MPa)
(b) Mineral transformations 0
Depth (km)
φdry = 30
o
10
20
φ = 2–12
30 0
100
o
200
300
Yield Strength (MPa)
(c) Pore fluid pressure weakening 0
φeff(dry) = 30 Depth (km)
o
10
φ
eff
(hydro.) = 18
o
20
φeff(~lithostatic) = 1
o
30 0
100
200
300
Yield Strength (MPa)
Fig. 2. Summary of the causes of weakening in frictional-plastic rocks: (a) cohesion loss results in limited strength reduction; (b) mineral transitions can produce phyllosilicates that have very low internal angles of friction; and (c) high pore fluid pressures, pf, can reduce brittle strength very significantly. feff 308 (pf ¼ 0, dry strength), feff 188 (pf ¼ hydrostatic), feff 18 (pf ¼ lithostatic).
been documented in mafic lower crustal and upper mantle lithosphere rocks in shear zones in the Ivrea zone, which formed during extension (Handy & Stuenitz 2002). Transient and long-term fluid pressure variations may be significant factors in weakening brittle rocks (Sibson 1990; Rice 1992; Ridley 1993; Streit 1997; Ingebritsen & Manning 1999; Connolly & Podladchikov 2000). Pore fluid pressure, pf, acts to reduce the dynamic pressure and p in equation 1 is replaced by the effective pressure, p 2 pf. It can be seen that the rock strength can become very small when pf p. An approximate way to consider the effect of pore fluid pressures is to introduce an effective internal angle of friction feff such that p . sinfeff ¼ (p – pf) sinf ¼ p(1 – l)sinf, where l ¼ pf/p is the pore pressure ratio. (Note that this definition differs from the Hubbert – Rubey pore pressure ratio for which l ¼ pf/poverburden, where poverburden is the weight of the overburden.) For example, using this definition of feff, rocks with f ¼ 308 have feff 188 in crust with hydrostatic pore fluid pressure. When the full range of pore fluid pressure variations that may occur during tectonic deformation is taken into account feff may vary from dry, feff 308, through the hydrostatic value to strongly overpressured, feff 08 (Fig. 2c). We use this effective internal angle of friction approach and adopt reference initial values of feff ¼ 308, 158 and 78 in order to investigate the sensitivity of the model results to a range of pore fluid pressure regimes. We also include a simplified parametric form of strain-dependent weakening in the numerical model experiments by linearly decreasing the internal angle of friction with the second invariant of the deviatoric strain, (I20 )1/2. In most of our models strain softening occurs in the range 0.5 , (I20 )1/2, 1.5. By using this approach we avoid the details of the physical mechanisms of strain softening, which are not well known in general but we can determine the sensitivity of the models to a range of possible strain-softening mechanisms. Strain-rate dependent weakening has also been used (Behn et al. 2002) to represent the rate dependence of frictional strength, which is observed in laboratory experiments. This approach characterizes the frictional behaviour of faults on seismic timescales (Scholz 1990). In contrast, our modelling addresses fault strength variations on geological timescales; therefore, we focus on the strain dependence of material properties.
Role of inheritance and heterogeneity We believe that the style of deformation of the lithosphere during rifting reflects the interplay
GEOMETRY OF RIFTS AND RIFTED MARGINS
of extension of pristine lithosphere with the superimposed effects of inherited structural heterogeneities, which can include both anomalously weak and strong regions. Models of extension of pristine lithosphere will display deformation modes that reflect the continuum nature of the system. Uniform pure shear is one such mode. However, the observed styles of rifts and rifted margins are highly variable, suggesting that inherited heterogeneities, ‘noise’ in the system, have a strong influence on its deformation. We cannot argue that the lithosphere is pristine. Therefore, it is critical to understand whether the noise overprints the inherent modes, or whether the underlying modes persist despite the noise. The noise is even more important if it contributes to the positive feedback effect of strain weakening, as discussed in the previous section. A small inherited weak heterogeneity, which is initially the focus of mildly enhanced strain, can become the seed for major localized deformation through this feedback mechanism. We consider two types of interaction between inherited ‘noise’ and strain softening. In sections on ‘Thermo-mechanical models of lithosphere extension’ and ‘Analysis of simple models using the principle of minimum energy dissipation’ we focus on the behaviour of deterministic model systems that include a single weak spot, or seed. This seed is taken to represent the weakest single zone of inherited damage that controls the strain localization process and, therefore, dominates over all other preexisting heterogeneities. A sensitivity analysis enabled us to conclude that weak seeds need to have a critical size and are most effective in localizing deformation when placed in the strongest part of the system, the upper mantle lithosphere in our models, thereby causing the greatest strength contrast with the surrounding material. From this analysis, we conclude that the system will preferentially select this type of seed over others, and we therefore focus on seeds placed in the upper mantle lithosphere. The approach that employs a single deterministic weak seed fails to consider that the weakest point may be ignored in favour of a distribution of inherited weak heterogeneities that link together to form even weaker shear zones. In the sections on ‘Systems with statistical heterogeneity’ and ‘Implication for the Iberia–Newfoundland conjugate margin system’ we consider the behaviour of systems characterized by a statistical distribution of inherited weaknesses that both compete and co-operate by linking to form shear zones. In this type of system, the continuum deformation activates the noise to create nascent-linked shear zones. Multiple shears may develop, evolve and compete. These two types of initial weakness, a single deterministic location or a statistical distribution, can be interpreted to be equivalent to natural lithospheric systems that are either dominated by one major zone of weakness or an
115
evenly distributed background level of ubiquitous damage in the form of minor faults or shear zones that have not developed into penetrative shears at the crust or lithospheric scale.
Thermo-mechanical numerical model Model description We use an Arbitrary Lagrangian–Eulerian (ALE) finite-element method for the solution of thermomechanically coupled, plane-strain, incompressible viscous–plastic creeping flows (Fullsack 1995; Willett 1999; Huismans & Beaumont 2003). The model is used to investigate extension of a layered lithosphere with frictional–plastic and thermally activated power-law viscous rheologies (Fig. 3). For details on the numerical method see Fullsack (1995), Willett (1999) and Huismans & Beaumont (2003). When the state of stress is below the frictional – plastic yield the flow is viscous. We use both linear Newtonian viscous materials (see later sections in this paper) and temperature-dependent non-linear power-law rheologies (see later sections in this paper) based on laboratory measurements on ‘wet’ quartzite (Gleason & Tullis, 1995) and ‘dry’ olivine (Karato & Wu 1993), where the latter also includes the pressure dependence of viscosity. The effective viscosity, h, in the power-law model is of the general form: Q þ Vp 0 h ¼ A1=n ðI_ 2 Þð1nÞ=2n exp ð2Þ nRT where (I˙20 )1/2 is the second invariant of the deviatoric strain rate tensor (121˙ij0 1˙ij0 )1/2, n is the power-law exponent, A is a scaling factor, Q is the activation energy, V is the activation volume (which makes the viscosity dependent on pressure, p), T is the absolute temperature and R is the universal gas constant. A, n, Q and V are derived from the laboratory experiments, and the parameter values are listed in Table 1. The reference parameter values for wet quartz listed in Table 1 lead to a weak viscous lower crust. Models of this type are described as weak crust models. In other models crustal viscosity, h(wet quartz), is everywhere increased by a factor of 100 to achieve a crust that is totally in the frictional –plastic regime. Models of this type are termed strong crust models. The viscosity scaling represents a simple technique that creates either strong frictional lower crust or moderately weak viscous lower crust without recourse to additional flow laws, each with its own uncertainties. The scaling can either be interpreted as a measure of the uncertainty in the flow properties of rocks
116
R. S. HUISMANS & C. BEAUMONT
Thermo-Mechanical Model Setup Strain softening Crust: Wet quartz Frictional plastic ρ0 = 2800 kg/m3 V
x,z = 0 km 35 km
Weak seed: 12 x 10 km von Mises plastic ρ0 = 3300 kg/m3 o σy = 107 Pa TMoho = 550 C
7
φ(ε)
1
0.5 ε 1.5
Strong lower crust
V 0
Weak lower crust
85 km 120 km
Moho
o
φ =1 φ =7
Vb
Lithosphere and sublithospheric mantle Dry olivine Frictional plastic (see right) 3 ρ0 = 3300 kg/m
Wet quartz
Depth
Vb
o
1330 C
o
Isothermal mantle, Temperature, Ta = 1330 C
o
1330 C x = 1200 km
Dry olivine
120 0
z = 600 km
o
250
0
250
Stress [MPa]
Fig. 3. Model geometry showing crust and mantle lithosphere layer thicknesses, the weak seed, their corresponding properties and the velocity boundary conditions, V, with Vb chosen to achieve a mass balance. Extension is driven by velocity boundary conditions and seeded by a small plastic weak region. Velocity boundary conditions are chosen to achieve a mass balance in the system. The model has a free top surface and the other boundaries have zero tangential stress (free slip). Whether materials deform plastically or viscously depends on the ambient conditions. At yield, flow is plastic. Below yield, deformation is viscous. Sedimentation and erosion are not accounted for in the model apart from surface smoothing resulting in small amounts of surface diffusion. Eulerian grid dimensions in horizontal and vertical dimension, nx, ny are, respectively, 401 and 151. The initial temperature field is laterally uniform, and increases with depth from the surface, T0 ¼ 0 8C, to base of crust, Tm ¼ 550 8C, following a geotherm for uniform crustal heat production, A ¼ 0.9 mW/m23 and a basal heat flux, qm ¼ 19.5 mW/m22. The temperature increases linearly with depth in the mantle lithosphere and the sublithospheric mantle is isothermal at Ta ¼ 1330 8C. The boundary conditions are specified basal temperature, 1330 8C, and insulated, q ¼ 0 mW/m22, lateral boundaries. Thermal conductivity k ¼ 2.25 W m21 8C and thermal diffusivity, k ¼ k/rcp ¼ 1 1026 m2 s21. Densities of crust and mantle, respectively, at 0 8C are r0 ¼ rc(T0) ¼ 2800 kg/m23 and r0 ¼ rm(T0) ¼ 3300 kg/m23 and depend on temperature with a thermal expansivity aT ¼ 3.1 10258C21, r ¼ r0 [1 2 aT(T –T0)]. On the right: initial (solid lines) and strain-softened (dashed lines) friction angle, f, 78 ! 18 representative strength envelopes of strong and weak crust models when V ¼ 0.3 cm year21. Frictional-plastic strain-softening behaviour is shown at the top.
where flow is dominated by quartz (e.g. ‘wet’ or ‘dry’) or to represent strong lower crust dominated by minerals that deform in the frictional regime for the conditions (temperature, strain rate) chosen for the model experiments. The plastic (frictional or brittle) deformation is modelled with a pressure-dependent DruckerPrager yield criterion which, with suitable adjustment of constants, is equivalent to the Coulomb yield criterion for incompressible deformation in plane strain. Yielding occurs when: ðJ20 Þ1=2 ¼ p sin feff þ c cos feff
ð3Þ
where (J02)1/2 ¼ (12 sij0 sij0 )1/2 is the second invariant of the deviatoric stress, c is the cohesion, feff is the effective internal angle of friction, defined earlier. With appropriate choices of c and feff this yield criterion can approximate frictional sliding in rocks.
Plastic flow is incompressible. Strain softening is introduced by a linear decrease of feff (1) over a range of strain where 1 represents the second strain invariant, (I20 )1/2 (Fig. 3, Tables 1 and 2). We have tested the dependence of the model results on mesh resolution and on the range of strain over which weakening occurs. Increasing the mesh resolution leads to narrower shear zones and to earlier localization because strain accumulates at a higher rate. Moderate changes in resolution do not, however, alter the main character of any particular model result. Also, the models do not show a strong sensitivity to the range of strain over which softening occur. The tests indicate that the major control on the model behaviour are the threshold at which strain softening starts and the amount of softening that occurs. The strong positive feedback between softening and strain accumulation implies that once initiated strain softening occurs rapidly.
GEOMETRY OF RIFTS AND RIFTED MARGINS
117
Table 1. Parameters for the lithosphere-scale thermo-mechanical models. Power-law creep parameters wet quartz (Gleason & Tullis 1995) and dry olivine (Karato & Wu 1993) Parameter Rheological parameters Angle of internal friction Cohesion Power-law exponent wet quartz Activation energy wet quartz Initial constant wet quartz Activation volume wet quartz Power-law exponent dry olivine Activation energy dry olivine Initial constant dry olivine Activation volume dry olivine Universal gas constant Thermal parameters Diffusivity Heat production crust Moho temperature Asthenosphere temperature Thermal expansion coefficient Surface temperature Densities Crustal density at 273 ºC Mantle lithosphere density at 273 ºC Sublithospheric mantle density at 273 ºC Dimensions and boundary condition Base of crust Base of mantle lithosphere Base of upper mantle Extension velocity
Symbol
Value
f(e) amount of softening c nqt Qqt Aqt Vqt nol Qol Aol Vqt R
20 106 Pa 4.0 223 103 J/mol 1.10 10228 Pa2n s21 0 m3/mol21 3.5 540 103 J mol21 2.4168 10215 Pa2n s21 25 1026 m3 mol21 8.3144 J mol21 8C21
k A Tm Ta aT T0
1 1026 m2 s21 0.9 1026 W m23 550 8C 1330 8C 3.1 1025 8C21 0 8C
rc(T0) rm(T0) rm(T0)
2800 kg m23 3300 kg m23 3300 kg m23
V
35 km 125 km 600 km 0.06, 0.3, 0.5, 10 cm year21 full rate
In addition to solving the equilibrium equations for plastic–viscous flows, we also solve the thermal evolution in two dimensions. The mechanical and thermal systems are coupled though the temperature dependence of viscosity and density, and are solved successively during each time step. Initial conditions and other thermal properties are given in Figure 3 and Table 1.
18 –78 and 28 – 158
Thermo-mechanical models of lithosphere extension We start with an overview of a more comprehensive set of models from Huismans & Beaumont (2002, 2003). The models selected here have frictionalplastic strain softening, and deformation is nucleated by deterministic noise in the form of a
Table 2. Parameters for two-layer models Parameter Angle of internal friction Cohesion Viscosity lower layer Density Thickness of plastic upper layer Thickness of viscous lower layer Horizontal length scale Extension velocity
Symbol f(1) amount of softening c H r hp hv L V
Value 28 – 158 0 1 1021, 1022, 1023 Pa s 3000 kg m23 60 km 60 km 600 km 1 cm year21 (full rate)
118
R. S. HUISMANS & C. BEAUMONT
single weak seed (Fig. 3). Most of the results can be classified as narrow rifts as defined by Buck (1991). We review the sensitivity of deformation style to the extension velocity and the strength of the middle and lower crust.
Sensitivity of model results to rifting velocity In order to demonstrate that rifting velocity is an important control on the extension mode we show results for a reference velocity of 0.3 cm year21 and for end-member variations of velocity of 0.06 and 10 cm year21 (Fig. 4). The models have strong middle and lower crusts, and include strain softening of the frictional –plastic rheology. At the reference velocity (Fig. 4b) strain softening in the frictional-plastic parts of the lithosphere results in strong localization in a single system of ‘faults’ (strictly shear zones), resulting in an asymmetric mode of extension. Middle and lower crust has been tectonically cut out and mantle lithosphere has been exhumed by large-scale frictional detachment. The proto-margins show distinct differences resembling an upper and lower plate conjugate margin pair. The main cause of asymmetry in the model is the frictional-plastic strain softening. After an initial phase of symmetric extension, feedback of strain softening results in preferential weakening of one of the two conjugate shears. The single weak shear zone remains the weakest part in the system and results in lithosphere-scale asymmetry. During later stages thermal advection produces a hotter, ductile lithosphere and a viscous necking style, and the final breakup phase is symmetric. However, asymmetry remains from the initial stages of the model evolution. At high rifting velocities (Fig. 4a) the style of extension is markedly different. The proto-margins are essentially symmetric. We have interpreted this model behaviour in terms of the increased role of viscous rheologies in the system and a stronger viscous coupling, where higher strain rates equate to higher viscous stresses at the base of the frictional layer (Huismans & Beaumont 2002, 2003). The viscous layer promotes a distributed, symmetric style of extension, and the tendency for localization and asymmetry given by the frictional–plastic strain softening is suppressed. The effect of a low end-member value for the extension velocity, v ¼ 0.06 cm year21, is illustrated in Figure 4c. The model is strongly asymmetric throughout its evolution. At this velocity thermal conduction is more efficient than thermal advection (Peclet number ,1). Consequently, the thermal evolution in the model lithosphere is close to the conductive limit, and the rift zone is relatively cool and deforms in the frictional regime. The
frictional weak shear zone therefore remains the single major weakness in the lithospheric system that allows for ongoing asymmetric extension.
Sensitivity of model results to strength of the lower crust We next examine the sensitivity of the model behaviour to the strength of the lower crust. The same frictional strain softening occurs as in the reference model, and in addition viscous flow following a wet quartz flow law now occurs in the weaker lower and middle crust (Fig. 5a). During the initial stages the crustal asymmetry is diminished because the conjugate frictional shears sole out in the weak ductile lower crust. Asymmetry in the lower lithosphere is still apparent in the shear zone cutting the mantle lithosphere. At later stages a set of synthetic frictional-plastic crustal shear zones form on either side of the central rift zone, thereby creating blocks bounded by frictional shear zones that sole out in the lower crust and that propagate beneath the rift flanks. During the late stage of deformation thermal advection results in nearly symmetric ductile necking of the lower lithosphere. The reference model result for a strong lower crust is given in Figure 5b.
Factors controlling mode selection and asymmetry during extension These results indicate three fundamental rift modes for the type of model described above: (1) lithosphericscale symmetric rifting; (2) lithospheric-scale asymmetric rifting; and (3) asymmetric upper lithosphere rifting concomitant with symmetric lower lithosphere rifting. Strain softening is the fundamental cause of the asymmetry and in a purely frictional-plastic model this would always lead to a phase of asymmetrical extension. In more complex models that include a viscous lower crustal layer the degree of asymmetry accompanying strain softening also depends on the rate of extension and on the viscous strength of the lower crust. The following factors need to be considered in regard to the efficiency of the strain-softening feedback in the frictional-plastic layer. † Geometrical hardening (Fig. 6d) may occur as the model evolves, for example by the rotation of shear zones, so that the active shears are no longer geometrically viable. This type of hardening opposes the continued increase in asymmetry and, depending on the amount and range over which strain softening occurs, even very weak shear zones will lock and be abandoned. † The extent to which the distribution of frictional-plasticity dominates the model
GEOMETRY OF RIFTS AND RIFTED MARGINS
119
Strong Crust, Sensitivity to Velocity High Velocity, V = 10 cm/a t = 12 Ma, Δx = 120 km
(a) 0 –20
o
550 C
z (km)
–40 –60 –80
o
1000 C –100 o
1200 C
–120 400
500
600
700
800
700
800
700
800
Moderate Velocity, V = 0.3 cm/a t = 40 Ma, Δx = 120 km
(b) 0 –20
z (km)
–40
o
550 C
–60 –80 o
–100
1000 C
–120
o
1200 C 400
500
600
Very Low Velocity, V = 0.06 cm/a t = 370 Ma, Δx = 222 km (c)
0 –20
z (km)
–40 –60
o
550 C
–80 –100 –120 400
500
600 x (km)
Fig. 4. Rift-mode sensitivity to extension velocity. Strong lower crust models with frictional-plastic strain softening, showing deformed Lagrangian mesh, velocity vectors and sample isotherms, for the dashed area in Figure 3. Model layers from top down denote upper and lower crust, strong frictional upper mantle lithosphere, ductile lower lithosphere and ductile sublithospheric mantle. Scaling of quartz viscosity makes all three upper layers frictional-plastic with the same f ¼ 78. (a) High rift velocity model, V ¼ 10 cm year21 at t ¼ 12 Ma and 120 km of extension. Note symmetric extension. (b) Reference rift velocity model, V ¼ 0.3 cm year21 at t ¼ 40 Ma and 120 km of extension. Note asymmetric extension. (c) Low rift velocity model, V ¼ 0.06 cm year21 at t ¼ 370 Ma and 222 km of extension. Note strong asymmetry. Animations of model evolutions can be found at: http://myweb.dal.ca/huismans/ jgr-animations.html.
120
R. S. HUISMANS & C. BEAUMONT
Effect Strength Lower Crust (a) Weak Lower Crust, V = 0.3 cm/a t = 40 Ma, Δx = 120 km 0 –20 o
550 C
z (km)
–40 –60 –80
o
1000 C
–100 –120
o
1200 C 400
500
600
700
800
600 x (km)
700
800
(b) Strong Lower Crust, V = 0.3 cm/a t = 40 Ma, Δx = 120 km 0 –20 o
z (km)
–40
550 C
–60 –80 o
–100 –120
1000 C
o
1200 C 400
500
Fig. 5. Rift-mode sensitivity to strength of the middle and lower crust. (a) Weak lower crust model, with h(wet quartz) at t ¼ 40 Ma and 120 km of extension. No scaling of quartz viscosity makes middle– lower crust ductile. Note the asymmetric crustal necking and symmetric lithospheric mantle necking. (b) Asymmetric strong middle– lower crust model at reference velocity at t ¼ 40 Ma and 120 km of extension is given for reference (same as in Fig. 4b). Animations of model evolutions can be found at http://myweb.dal.ca/huismans/ jgr-animations.html.
lithosphere. A range of behaviours exists between models where by frictional-plasticity is the dominant process in the extending lithosphere, which exhibit the most asymmetry, and those where geometrically larger ductile regions promote symmetry. † The relative integrated strengths, as opposed to geometries, of the plastic and ductile regions. For example, ductile strength increases with strain rate, producing higher stresses at higher rifting velocities, whereas the frictional
strength is independent of strain rate. This strain-rate dependence of strength explains the velocity sensitivity (Fig. 6d). The strong ductile region suppresses the asymmetric deformation preferred by the overlying frictionalplastic layer; ductile symmetry dominates over plastic asymmetry (Buck et al. 1999, 2003; Huismans & Beaumont 2002, 2003; Huismans et al. 2005). † The thermal state of the model, which also contributes to the asymmetry by determining the
GEOMETRY OF RIFTS AND RIFTED MARGINS
(c) Geometric Distribution Brittle & Viscous Materials
(b) Geometrical Hardening
A
Hp
B
isotherm marking the brittle-viscous transition
Hp
isotherm marking the brittle–viscous transition
Bending related hardening
z
Hb
brittle layer thickness
A
B
σ yield stress
σ yield stress
yield stress brittle lithosphere
121
z
Large proportion frictional z strength
Small proportion frictional strength
(a) Main Cause of Localization: Strain Weakening S brittle lithosphere
A
B
Hp A
isotherm marking the brittle–viscous transition
z
(d) Strain Rate Sensitivity Viscous Coupling . εA strain. . ε rate
ε
B
yield stress
brittle lithosphere
A
B
brittle yield stress
Hp
S
yield stress
yield stress
B
loss of cohesive strength
loss of frictional strength
z
(e) Thermal Advection/Conductive Cooling . total strain ε
εA
yield stress
A
B
brittle-viscous transition depth
A z
increase in yield strength at constant temperature
brittle yield stress
brittle lithosphere
Hb A
B
B
Brittle Ductile Transition
z
decrease in yield strength at constant strain rate
Fig. 6. Factors affecting localization of deformation (modified after Buck et al. 1999). (a) Strain softening is the main cause of localization of deformation. Factors that reinforce or counteract localization are: (b) geometrical hardening resulting from bending counteracts continued localized deformation; (c) the geometric distribution of the strainlocalizing plastic layer, and the non-localizing viscous materials determines the relative importance of plastic and viscous domains. A large viscous domain promotes symmetric or pure shear modes. (d) Brittle–ductile coupling. The viscous layer in the coupled system promotes distributed deformation through viscous shear stress along the interface of the coupled system, where at high strain rates associated high viscous stresses resist localized deformation in the plastic layer and suppress asymmetry. (e) The relative importance of thermal advection and conduction determines the relative distribution of plastic and viscous domains as the model evolves. For low extension velocities the plastic layer retains its thickness which promotes asymmetry.
relative distribution of frictional –plastic and ductile regions as the model evolves (Huismans & Beaumont 2002, 2003; Lavier & Buck 2002). When extension is slow, the thermal Peclet number is small and the lithosphere cools as it extends, thereby continuously renewing the frictional –plastic layers as the model evolves. This promotes asymmetric behaviour by comparison with models that have a large thermal Peclet number (Fig. 6a).
Analysis of simple models using the principle of minimum energy dissipation Approximate analytical theory The first series of models displayed a range of behaviours that we interpreted in terms of factors that reinforced or opposed asymmetric rifting. Although this approach does provide a certain level of insight, there is no governing principle that guides
122
R. S. HUISMANS & C. BEAUMONT
the overall interpretation. In this subsection we simplify the models to consider only the most basic behaviour of a two-layer system, and take a different approach to the interpretation by demonstrating the use of the minimum rate of energy dissipation as a guide to model behaviour (Bird & Yuen 1979; Sleep et al. 1979; Masek & Duncan 1998). Huismans et al. (2005) used this approach, subject to the restrictions noted by Bird & Yuen (1979), to understand the factors that control localized asymmetric modes of deformation v. distributed symmetric modes by comparing the minimum dissipation analysis with equivalent finite-element models, which are themselves based on minimum dissipation. We review the basic case of a two-dimensional, plane-strain system comprising a uniform, strainsoftening, frictional-plastic layer overlying and bonded to a uniform linear viscous layer. The analysis predicts the rate of internal energy dissipation (henceforth termed the dissipation for brevity) during the early stages of extension for three possible modes, the pure shear (PS), symmetric plug (SP) and asymmetric plug (AP) modes (Fig. 7). Details of the analysis are given by Huismans et al. (2005). For this simple system there are two main contributions to the rate of internal energy dissipation. The first is from extension of the plastic layer and is the same for each of the modes except when the dissipation is reduced by strain softening of the localized shears in the SP and AP modes. The second is from the viscous layer and its size ranges from the contribution owing to pure shear, in the PS mode, to a combination of pure shear and shear of a boundary layer at the top of the viscous layer when there is differential extension between the two layers in the SP and AP modes (Fig. 7). If we consider PS as a reference mode (Fig. 7), we can then ask under what circumstances a different, lower dissipation mode exists and will be selected. It can readily be understood that any overall reduction in dissipation depends on the trade-off between the ‘gain’ of reduced plastic dissipation owing to strain softening in the SP and AP modes v. the ‘penalty’ of increased dissipation owing to development of the viscous boundary layer in the SP and AP modes (Fig. 7). By considering these trade-offs, mode selection can be predicted as a function of the layer properties and the rate of extension. The latter is important because the rate of dissipation in the plastic layer depends on the strain rate, whereas that of the viscous layer depends on the square of the strain rate, so mode selection will depend on the extension velocity, V (Fig. 7). The total internal dissipation for each mode, WT, is calculated approximately for a given choice of the
(a) Pure Shear Mode (PS) hp
Plastic Layer: C(ε) φ(ε)
hv
Viscous Layer: η
V
V
(b) Symmetric Plug Mode (SP)
V
hb
Viscous Boundary Layer
V
Pure Shear
(c) Asymmetric Plug Mode (AP)
S
hb
2V
2L
Fig. 7. The total rate of dissipation is estimated for each of the following modes of deformation: (a) pure shear; (b) symmetric plug; and (c) asymmetric plug modes. The geometry and velocity field assumed are based on simple approximations to each of these modes. A plastic strain-softening layer is overlying and bonded to a linear viscous layer. Both layers with respective thicknesses hp and hv, and total length 2L, extend by boundary velocity V applied at each side. For localized symmetric and asymmetric plug modes, shear resisting localized deformation in the viscous layers is assumed to occur in a boundary layer of thickness hb. Deformation below the boundary layer follows pure shear. For more details on the analysis see Huismans et al. (2005).
model parameter values such as layer thicknesses, layer properties, the amount of strain softening and the extension velocity, and the results are plotted as a function of, for example, the viscosity, h, of the lower layer (Fig. 8). This figure shows the dissipation for the SP and AP modes. A mode transition occurs between the AP and SP modes where the curves cross at the transition viscosity, hT. In this case, the SP mode becomes the lower dissipation mode for h . hT. Mode transitions only occur when the dissipation curves cross; in other cases one mode systematically has a lower dissipation than another for all values of the viscosity. The total dissipation, WT, of each mode is equal at a transition, therefore W1Plastic þ W1Viscous ¼ W2Plastic þ W2Viscous
ð4Þ
where subscripts 1 and 2 successively represent the AP, SP and PS modes in our example. Solving
GEOMETRY OF RIFTS AND RIFTED MARGINS
(a) A
WT WTotal
S
WT
Mode Transition
ηt
η (b)
ηt = 3 hb ρ g hp2 (sinφpss – sinφss) Lv V Viscous penalty force
(c)
(d)
(e)
ηt Lv V hb Lv η t ε Α
=
=
ΔFv
Plastic gain force
2
3 ρ g hp (sinφpss – sinφss) 2
3 ρ g hp (sinφpss – sinφss)
=
ΔFp
Fig. 8. Mode-transition criteria based on the principle of minimization of dissipation. Main results and main equations are given here. Full derivation can be found in Huismans et al. (2005). (a) Typical plot of total ˙A dissipation for asymmetric plug mode (W I ) and ˙ SI ) of deformation. For viscous symmetric plug mode (W layer viscosities lower than hT the asymmetric plug modes provides the minimum dissipation mode and the converse for higher viscosities. (b) Transition viscosity, ˙A ˙S hT, which can be derived by solving W T ¼ WT, is a simple function of viscous and plastic layer length scales, strain-softening parameters and applied boundary velocity. Lv ¼ L 2 þ 4hb (3hv – hb))/L is a geometrical constant with viscous layer length scales. (c)–(e) The equation for transition viscosity can be rewritten in terms of two differential forces: (1) the viscous penalty force, D Fv, incurred by forcing localized shear in the viscous layer as a consequence of localization in the plastic layer; and (2) the differential plastic gain force, DFp, resulting from strain softening in the plastic layer. Transition between asymmetric plug and symmetric plug modes, and between symmetric and pure shear modes of deformation depends on the trade-off between these two differential forces. See Table 2 for values of parameters used.
the set of equations in equation 5 yields analytical expressions for the transition viscosities, hT1 and hT2, which then shows how the transitions depend on the model properties (e.g. Fig. 8, equation 1, for the AP to SP transition: see Huismans et al. 2005, for derivation of the equations). These expressions are relatively simple if the analysis is
123
confined to the plastic and viscous contributions, as in equation 5, but become more complex when additional contributions, such as dissipation against gravity and layer bending, are included. It can also be seen from equation 5 that these additional terms can be ignored if their contribution to both modes at a transition is approximately equal. Under this circumstance they contribute equally to both sides of the equation and can be subtracted. The results from the approximate analysis (equation 5) demonstrate that modes are selected in the order AP, SP and PS as the viscosity of the lower layer increases and all other parameters are constant (Huismans et al. 2005). This order appears to be robust for the early phases of extension for parameter values appropriate for lithospheric-scale extension, and also applies to models in which the upper layer strain softens by loss of cohesion (Huismans et al. 2005). The effect of ignoring additional contributions to the dissipation means that the transition viscosity estimates, hT, are approximate, but the errors do not affect the selection order in this particular case (Huismans et al. 2005). The overall results of this analysis are summarized in Figure 9, which shows how the modes are distributed in relation to the plastic and viscous controls. NSS, PSS and SS are, respectively, the non-strain-softened, partially strain-softened and strain-softened strengths of the plastic layer (the corresponding values of f(1) appear in equation 1, Fig. 8), and hT1 and hT2 are transition viscosities between AP and SP, and SP and PS modes, respectively. Arrows show how models typically evolve during strain softening of the plastic layer, when the other properties of the system, such as extension velocity, remain constant. The mode selection and mode transitions can be understood from a physical perspective in terms of the competition between the ‘plastic gain’ and ‘viscous penalty’ contributions, defined above, to the dissipation. For example, for h . hT2 the mode selected is always PS (Fig. 9) because the viscous penalty incurred by the boundary layer dissipation in the SP or AP modes (Fig. 7) is always greater than the plastic gain achieved by strain softening of the localized shear zones. This result explains the perplexing observation that for this viscosity range the model will extend by pure shear even when the upper layer could develop weak localized shears. This result demonstrates that the properties of the bonded laminate are clearly different from those of its constituent layers. For hT1 , h , hT2 the strain-softened symmetric plug mode, SP2, is selected during plastic strain softening, and for h , hT1 the AP mode can be selected.
Plastic Strength, C(ε) or φ(ε)
124
R. S. HUISMANS & C. BEAUMONT
NSS
SP1
SP1
PS
Symmetric Plug
Symmetric Plug
Pure Shear
Plastic Strain Softening
PSS
SS
AP
SP2
Asymmetric Plug
Symmetric Plug
ηΤ1
PS
Pure Shear
ηΤ2
Viscosity Lower Layer Fig. 9. Rift-mode space. Evolutionary diagram of mode selection behaviour predicted by dissipation analysis and confirmed by numerical models in Figure 11. NSS, non-strain softened; PSS, partial strain softened; SS, strain-softened materials. At viscosities less then hT1 frictional-plastic strain weakening leads to the asymmetric plug mode. At viscosities larger then hT2 localization in the frictional-plastic layer is completely suppressed and distributed pure shear occurs. At intermediate viscosities, the symmetric plug mode is predicted.
However, as is explained below, the mode selection also depends on the way the modes evolve, that is, it is path dependent.
Simple two-layer numerical models The next step is to present finite-element model experiments designed to test the approximate analytical theory described in the previous section. The numerical model is the same as that used in the section on ‘Thermo-mechanical models of lithosphere extension’, but it now has the simple twolayer geometry (Fig. 10) and rheological structure (Table 2). A strain-softening frictional –plastic layer (f (1) ¼ 28–158) overlies and is bonded to the uniform constant linear viscous layer. The weak seed is placed at the base of the plastic layer. The results are shown after 40 km total extension at V ¼ 1 cm year21 for viscosities, h, ranging from 1021 to 1023 Pa s (Fig. 11). At low viscosity (Fig. 11a) the AP mode is selected. Strain softening has focused the plastic deformation onto one shear and the extension is highly asymmetric. When
h ¼ 1022 Pa s (Fig. 11b) asymmetry is suppressed and extension in the plastic layer occurs along two symmetric shears in the SP mode. At high viscosity (Fig. 11c) local deformation in the plastic layer is completely suppressed and extension of both layers occurs by the PS mode. Huismans et al. (2005) present a greater range of numerical experiments, but the three results (Fig. 11) serve to illustrate the basic behaviour.
Comparison of the analytical dissipation results with the numerical models The results of the numerical models are in qualitative agreement with the predictions of the dissipation analysis in that mode selection changes from the AP mode through SP to PS as the viscosity of the lower layer is increased (Fig. 11). The results also indicate that the transition viscosities, hT1 and hT2, fall between 1021 and 1022 Pa s, and between 1022 and 1023 Pa s, respectively, for the parameter values (Table 2) used in the models, which is in agreement with the equivalent analytical
GEOMETRY OF RIFTS AND RIFTED MARGINS
125
Numerical Model Setup 0
Plastic Layer
Weak Seed
V 60
V
Viscous Layer 120 km 0
600 km
Strain softening 15
o
φ(ε)
o
2
0.5
ε
1.5
Fig. 10. Geometry and boundary conditions for numerical experiments of extension in simple two-layer models. Upper strain-softening frictional-plastic layer and lower viscous layer are both 60 km thick. Side boundaries extend with velocity V ¼ 0.5 cm year21. Extension is seeded by a small plastic weak seed. The model has a free top surface, and the other boundaries have zero tangential stress (free slip). Sedimentation and erosion are not included in the model. There is, however, a small amount of numerical surface diffusion. The Eulerian grid has 401 and 121 nodes in the horizontal and vertical direction, respectively. Strain softening is achieved in a simple manner by a linear decrease of the angle of internal friction with deviatoric strain.
results. The finite-element model experiments not only confirm the approximate analysis for the initial mode selection, they also demonstrate that these modes persist as the model evolves. The results also confirm our assertion that the additional contributions to the dissipation, which were ignored in the analysis but are included in the full finite-element solution, do not change the order of the mode selection, thereby providing reassurance that the approximate treatment is acceptable. The results also indicate that other modes, not considered in the analysis, are not strongly preferred for this range of viscosities. We do, however, know that modes comprising a series of plugs, not just a single plug, have levels of dissipation intermediate between the SP and PS modes. In addition, it can also be shown that the mode in which the plastic layer totally detaches from the viscous layer is not selected for the range of parameters we consider because the plastic dissipation ‘penalty’ incurred by shear on the plastic detachment is much larger than any of the possible dissipation ‘gains’ in the viscous or plastic layers for this particular mode. The overall results can be interpreted in several ways, each of which offers a different insight into the mode selection. The first interpretation describes the mode selection in relation to increasing viscosity of the viscous layer. When the viscous dissipation is negligible, the AP mode is chosen because
this provides the fastest route to minimize the plastic dissipation by strain softening only one, not two, shear zones. When viscous dissipation is larger, the mode selected will change from AP to SP at the transition viscosity hT2, and to PS at hT2, as explained in the previous section. In the second interpretation the results are described in terms of the trade-off between two differential forces shown, for example, by the second equation in Figure 8. The left-hand side of the equation has dimensions of force per unit length of the model and is termed the differential ‘viscous penalty force’ incurred by deforming in the localized AP mode as opposed to the more distributed SP Mode. The right-hand side of the equation is termed the differential ‘plastic gain force’, and is equal to the reduction in the force in the plastic layer when deformed in the fully strainsoftened AP mode by comparison with the PS mode. Mode transitions occur when these two forces are equal. The physical insight provided by this requirement is that the system behaviour is determined not by the intrinsic properties of either the plastic or viscous layer alone, but instead it is determined by the trade-off between the differential penalty/gain forces. This interpretation can also expanded to consider the effect of other factors, for example, extension velocity as noted in the results of the section ‘Thermo-mechanical models of lithosphere
126
R. S. HUISMANS & C. BEAUMONT
Low Viscosity Lower Layer, V = 1 cm/a t = 4 Ma, Δx = 40 km
(a)
z (km)
0
φ(ε) = 15-2
o
60
η = 10
21
Pa s
120 0
600 km
Intermediate Viscosity Lower Layer, V = 1 cm/a t = 4 Ma, Δx = 40 km
(b)
0 z (km)
o
φ(ε) = 15-2 60
η = 10
120
22
Pa s
0
600 km
High Viscosity Lower Layer, V = 1 cm/a t = 4 Ma, Δx = 40 km
(c)
0 z (km)
o
φ(ε) = 15-2 60
η = 10
23
Pa s
120 0
x (km)
600 km
Fig. 11. Two-layer models of lithosphere extension. Upper layer strain-softening frictional-plastic, angle of internal friction f decreases from 158 to 28 as second invariant of strain ((I02)1/2 ¼ 1 ¼ 0.5–1.5) increases; cohesion c ¼ 0, lower layer uniform viscosity. Different panels are for different values of viscosity. Vertical side boundaries extend at 0.5 cm year21. Panels shown at 40 km of extension. Weak seed, f ¼ 28 in the centre at the base of frictional-plastic layer seed deformation. (a) Model with lower layer viscosity h ¼ 1021 Pa s. Note asymmetric mode is selected. (b) Model with lower layer viscosity h ¼ 1022 Pa s. Note symmetric mode is selected. (c) Model with lower layer viscosity h ¼ 1023 Pa s. Localization in the plastic layer is completely suppressed and the pure shear mode is selected.
extension’ (Fig. 4) (and also those of Huismans & Beaumont 2003) in which asymmetric, AP mode extension occurs when the rifting velocity is 0.3 cm year21, or less, but exactly the same system exhibits SP mode extension when the rifting velocity is 10 cm year21. As shown by equation 2 of Figure 8, the higher velocity modifies the viscous dissipation such that the differential viscous penalty force will be equal to the differential plastic gain force at a transition viscosity that is a factor of 33 (10/0.3 from above) smaller than that for the lower velocity, a result also implied by
equation 1 of Figure 8. This modified transition viscosity is much smaller than that for the lower extension velocity. Consequently, the rapid extension model is located in the SP part of mode space (Fig. 9), whereas the model with the lower extension velocity is located in the AP part. This result demonstrates that the transition viscosities change with model properties. At even higher extension velocities the mode would change to PS. It follows from Figure 8, equation 1 that variations in any of the parameters on the right-hand side of the equation, particularly the velocity, V, the
GEOMETRY OF RIFTS AND RIFTED MARGINS
thickness of the plastic layer, hb, and the initial and final strength of the plastic layer, rgh2b sin(fSP)/2 and rgh2b sin(fAP)/2 will modify the transition viscosity between the SP and AP modes. This result has the potential to provide insight into natural systems. The third interpretation uses the concept of ‘dominant’ and ‘subordinate’ rheologies introduced by Huismans & Beaumont (2003). This concept is qualitative but can be made more precise by considering the differential plastic gain and viscous penalty forces. Except at a mode transition (e.g. Fig. 8, equation 2) one differential force is larger and can be interpreted to imply that the corresponding rheology ‘dominates’, whereas the other is ‘subordinate’. For example, if the viscous differential force is larger, the viscous rheology dominates and forces the model to select the mode that minimizes the viscous dissipation, and the converse. The concept provides a useful physical interpretation of the model results. The preferred mode for the plastic layer acting alone would be AP, because this is the fastest way to minimize the dissipation of this layer. This mode is not selected when the viscous rheology is dominant because the ‘dominant’ rheology forces the mode selection. The same concept applies to the velocity sensitivity of the model results described in the section on ‘Thermo-mechanical models of lithosphere extension’ (Fig. 4). At V ¼ 0.3 cm year21, or less, the plastic rheology dominates and the mode is therefore AP, but at V ¼ 10 cm year21, the viscous rheology dominates and the mode is SP.
Systems with statistical heterogeneity Statistical model of inheritance from earlier tectonic deformation We now investigate systems that are characterized by the second type of heterogeneity, statistical noise in the initial strain field. The previous two sets of models used a single weak seed to localize the deformation. As noted in the subsection on ‘Role of inheritance and heterogeneity’ earlier, this approach fails to consider that the weakest point may be ignored in favour of a distribution of inherited weak heterogeneities that link together to form even weaker shear zones. Models with statistical heterogeneities in the initial strain field are designed to represent inheritance of deformation from previous tectonic phases. They then exploit the positive feedback link between the inherited strain distribution and strain softening. Specifically, the second invariant of the deviatoric strain, (I20 )1/2, is initialized with white noise that has a truncated Gaussian distribution with a
127
mean value ((I20 )1/2 mid) set to half the strain at which strain softening starts ((I02)1/2 min), and with a maximum value just below (I20 )1/2min. Therefore, there is no initial strain softening in the model, but the noise preconditions parts of the model to be on the threshold of strain softening. We have also explored other types of noise, in particular fractal noise with various fractal exponents, and have found that the first order of the model behaviour is similar for a range of noise types.
Simple two-layer models with statistical heterogeneity The results of models that are equivalent to those of the subsection on ‘Simple two-layer numerical models’, but instead have statistical noise in the initial strain field, are shown for viscosities of the lower layer ranging from 1021 to 1023 Pa s (Fig. 12). The models all extend at V ¼ 1 cm year21 and have exactly the same noise distribution. At low viscosity (Fig. 12a) the AP mode is selected. At intermediate viscosity (Fig. 12b) deformation is more distributed, strain softening having developed on several competing shear zones. At high viscosity (Fig. 12c) deformation in the plastic layer is distributed and characterized by a large number of shear zones. These behaviours can be interpreted using the minimum dissipation principle. The AP mode with highly asymmetric deformation can be adopted when the viscous penalty contribution to the dissipation is small (Fig. 12a), and in this case deformation is accommodated on only one plastic shear zone, thereby maximizing the plastic gain through rapid strain softening. At intermediate viscosity the mode is similar to SP (Fig. 11), but several shear zones have developed (Fig. 12b). Evidently, the total dissipation is best minimized by the activation of several plastic shears even though some will strain soften faster than others. In this case the major advantage is the ‘gain’ of reduced viscous dissipation by comparison with the AP and SP modes. Similar reasoning suggests the SP mode with a single plug should be selected for viscosities somewhere in the range 1021 Pa s , h , 1022 Pa s. At high viscosity, 1023 Pa s, the model (Fig. 12c) is dominated by the viscous rheology, at least in a relative sense. Some plastic gain has been achieved by strain softening on the plastic shear zones, but it is at the expense that many shears had to be created. The major gain is in the corresponding reduction in the viscous penalty because there are now only short, low strain rate, segments of the boundary layer flow between the multiple plastic shear zones, not long segments that accommodate major offsets as in the AP and SP modes.
128
R. S. HUISMANS & C. BEAUMONT
Low Viscosity Lower Layer, V = 1 cm/a t = 5 Ma, Δx = 50 km
(a)
z (km)
0
φ(ε) = 15-2
o
–60
21
η = 10 –120 0
Pa s 100
200
300
400
500
600
400
500
600
400
500
600
Intermediate Viscosity Lower Layer, V = 1 cm/a t = 8 Ma, Δx = 80 km
(b)
z (km)
0
–60
η = 10
22
–120 0
Pa s 100
200
300
High Viscosity Lower Layer, V = 1 cm/a t = 11 Ma, Δx = 110 km
(c)
z (km)
0
–60
–120 0
23
η = 10
Pa s 100
200
300 x (km)
Fig. 12. Two-layer models of lithosphere extension with distributed noisy heterogeneity. White noise with a Gaussian distribution is applied to the initial strain. Viscosity of the lower layer forms the only difference between models. (a) Asymmetric mode selected for low viscosity. (b) Multiple plug mode with several competing shear bands occurs for intermediate viscosities. (c) Pure shear mode occurs for high viscosity. Note that behaviour is similar to the single seed models in Figure 11.
The overall results indicate that this heterogeneous noise, which is below the strain-softening threshold, does not dictate the behaviour of the system and that deformation similar to the AP, SP and PS modes can occur depending on the viscosity of the lower layer. The difference is that modes with multiple plastic shear zones are now easily excited for the viscosity range hT1 , h , hT2 because the inherited strain heterogeneity places many regions of the model close to the threshold for strain softening. These new modes modify Figure 9 by occupying part of the mode space previously taken by the SP and PS modes, such that a new transition viscosity, to the multiple plastic shear modes, occurs between hT1 and hT2, and the value of hT2 is
increased. The multiple plastic shear modes will be characterized by an increasing number of shears with increasing viscosity of the underlying layer. This behaviour occurs because the viscous penalty is reduced by reducing the spacing between the plastic shears, as noted above, and in the limit of high viscosity, the new hT2, the overall deformation will approach the PS mode.
Lithosphere-scale extension with statistical crustal heterogeneity In this section we investigate the role of statistical heterogeneity in the initial strain field in
GEOMETRY OF RIFTS AND RIFTED MARGINS
lithosphere-scale models equivalent to those of the section on ‘Thermo-mechanical models of lithosphere extension’ to determine whether mode selection will be modified. We assume heterogeneity, due to earlier phases of deformation, is thermally annealed in the mantle where temperatures are high and, therefore, only apply statistical heterogeneity to the crust. Note that strain softening does not occur in the viscous (ductile) regime so that models with a strong (plastic) lower crust will experience strain softening in this region but those with a weak, ductile lower crust will not. For this series of models (I20 )1/2 is initialized with white noise that has a Gaussian distribution with a mean value ((I20 )1/2 mid) set to the strain at which strain softening starts ((I20 )1/2 min), and with the same Gaussian distribution. The other properties of these models are the same as the equivalent ones in the section on ‘Thermo-mechanical models of lithosphere extension’ except that the extension velocity is 0.5 cm year21. When the lower crust is strong (Fig. 13) extension is at first pure shear. When the incremental strain augments the initial statistical strain to the threshold for strain softening, strain softening occurs in the frictional-plastic crust and the SP mode is selected. The position of the plug depends on the particular realization of the white noise, a behaviour that is equivalent to randomly placing a weak seed. Localized deformation in the crust causes diffuse projections of the PS mode shears to develop in the upper mantle lithosphere and these forced shear zones eventually localize by strain softening in the upper frictional –plastic mantle. Feedback of strain softening results in the development of an asymmetric shear zone that penetrates most of the lithosphere (Fig. 13a) which facilitates exhumation of the upper mantle lithosphere (Fig. 13b) via the AP mode. During later stages, when thermal advection weakens the lithosphere, viscous necking becomes the dominant style of deformation (Fig. 13c). This model behaviour is very similar to its single weak seed equivalent (Figs 4b & 5b). The difference in the extension velocity, 0.5 v. 0.3 cm year21, has only a minor effect on results. A more important implication is that models with this particular set of properties behave in a similar manner and are not particularly sensitive to the difference between inherited statistical heterogeneity and a deterministic weak seed. The result is also consistent with the simple two-layer model with statistical heterogeneity and a low viscosity lower layer (Fig. 12a). Evidently, in all of these examples, which are essentially two-layer cases, the plastic gain achieved by selecting the AP mode outweighs the viscous penalty for low values of the effective mantle viscosity. For higher effective viscosities
129
of the mantle lithosphere we may expect more distributed modes of deformation like those seen in Fig. 12b, c. The style of deformation for the equivalent model with the viscously weak lower crust (Fig. 14) is markedly different, both from the model with strong lower crust and, more significantly, from the equivalent model with a single weak seed (compare Figs 5a and 14). At first deformation occurs in an overall ‘PS-type’ mode (Fig. 14a). Sets of parallel and conjugate frictional shears form in the frictional upper crust, but strain localization does not occur in the upper mantle lithosphere which has no initial statistical heterogeneity. After approximately 160 km of extension, localization in the crustal layer feeds back into the frictional –plastic upper mantle lithosphere and seeds a localized shear zone (Fig. 14b). The remaining evolution is similar to the equivalent model with a strong lower crust except that the final necking is more symmetrical and there is somewhat less exhumation of the mantle (Fig. 14c). The significance of this model is that it displays a natural progression from an early distributed wide rift mode to a late localized narrow rift mode. Although the equivalent model with the single weak seed also exhibits decoupling and differential shear between the crust and uppermost mantle (Fig. 5a), the width of the zone of distributed crustal deformation is much narrower and tends to increase as rifting progresses.
Implications for the Iberia – Newfoundland conjugate margin system First-order characteristics of the Iberia– Newfoundland conjugate margin system In this subsection we explore whether what has been learned from the models above can be applied to extension of the Iberia –Newfoundland conjugate margin system. One of the main characteristics of the Iberia –Newfoundland conjugate margin is the high degree of variability in the geometry of the margins and the location of extension with position along the margins (Figs 15 & 16). This variability may result from the competition between local inherited structures and underlying physical processes that control the mode of extension. We first attempt to identify the general features of the system because these may lead to insight into the fundamental underlying processes responsible for its formation. Secondly, we address the potential causes of the variability. Cross-sections (Keen et al. 1987a, b; Torne et al. 1994; Dean et al. 2000; Funck et al. 2003; Pe´rez-Gussinye´ et al. 2003) shown restored to
130
R. S. HUISMANS & C. BEAUMONT
(a)
t = 12 Ma, Δx = 60 km, V = 0.5 cm/a
0 –20 z (km)
–40 –60 –80 –100 –120 400
(b)
500
600
700
800
600
700
800
600 x (km)
700
800
t = 16 Ma, Δx = 80 km
0 –20 z (km)
–40 –60 –80 –100 –120 400
(c) 0
500
t = 24 Ma, Δx = 120 km
–20 z (km)
–40 –60 –80 –100 –120 400
500
Fig. 13. Strong lower-crust thermo-mechanical model of lithosphere extension with white statistical noise applied only to the crust. The model selects the narrow rift mode with strong asymmetry. Note that model behaviour is similar to the equivalent single seed model (e.g. Fig. 4b).
approximate the configuration at the time of onset of sea-floor spreading (Fig. 16) illustrate the strike variability over a distance of only 100 km. We regard the following to be key features of the Iberia–Newfoundland conjugate margin system. (1) Extension and rifting occurred during a
prolonged period of approximately 80 Ma with at least three phases of rift activity (Mauffret & Montadert 1988; Tankard & Balkwill 1989; Tankard et al. 1989; Murillas et al. 1990; Driscoll et al. 1995; Stapel et al. 1996). (2) Initial rifting occurred in a wide rift mode over an extended
GEOMETRY OF RIFTS AND RIFTED MARGINS
(a)
131
t = 24 Ma, Δx = 120 km, V = 0.5 cm/a
0 –20 z (km)
–40 –60 –80 –100 –120 400
(b) 0
500
600
700
800
600
700
800
600 x (km)
700
800
t = 32 Ma, Δx = 160 km
–20 z (km)
–40 –60 –80 –100 –120 400
(c) 0
500
t = 44 Ma, Δx = 220 km
–20 z (km)
–40 –60 –80 –100 –120 400
500
Fig. 14. Weak lower-crust thermo-mechanical model of lithosphere extension white statistical noise applied only to the crust. Model evolution: (a) wide rift mode selection, only localization in crust, very widely distributed deformation in the mantle lithosphere; (b) and (c) narrow rift mode selection, feedback of localization in crust into mantle lithosphere. Note the natural progression from wide to narrow rift mode. Model behaviour is dissimilar with equivalent decoupled single seed model (Fig. 5a).
area that was at least 400 km wide (Keen et al. 1987a, b; Tankard & Welsink 1989; Manatschal 2004). (3) The system became a narrow rift late in its evolution and continental mantle lithosphere was exhumed during the terminal phase (Boillot
et al. 1988; Louden & Chian et al. 1999; Dean et al. 2000). (4) The system is a non-volcanic rift zone, the late stages of rifting and early sea-floor spreading being remarkably a-magmatic (Louden & Chian 1999). (5) Extension occurred at very
132
R. S. HUISMANS & C. BEAUMONT
hP ass
Jeanne
C
Basin
4000 2000
B`
Fle mis
d'Arc
ia I nte
Central Transect
Ba
sin
oe sh se in or as B
H
D`
C`
B
idg e
Ca
rso
n
S. Whale Basin
Ga lic
Whale Basin
E
rio
rB
asin
46 N
Pe rid otit eR
Grand Banks
idg e
A
Flemish Cap Northern Transect A`
54°W
52°
50°
nic Oc
4000
42
F`
D
ea
E`
2000
Exhumed Mantle Lithosphere
Co Edg ntin e Cr enta us t l
Cr us
t
Southern Transect
Pe rid otit eR
44
48°
F 46°
44°
42°
40°
38°
Fig. 15. The Iberia–Newfoundland conjugate margin system. Map of the Cretaceous Central Atlantic restored to magnetic anomaly M0 time (c. 121 Ma) based on the reconstruction pole of Srivastava et al. (2000) with Newfoundland fixed to present geographic co-ordinates (modified after Preliminary Results Ocean Drilling Program, Tucholke et al. 2004). Northern, central and southern cross-sections, indicated with thick black lines, are shown in Figure 16.
low average velocities, an important point that we substantiate below. (6) Localized extension during the first two phases of rifting was limited to the crust with only minor thinning of the mantle lithosphere as evidenced by limited post-rift subsidence of the early rift basins (Stapel et al. 1996). (7) Asymmetries in crustal thickness, distribution of exhumed mantle lithosphere (Whitmarsh et al. 2001) and spatial variability of mineral chemistry of the exhumed mantle lithosphere (Muentener 2004) suggest that late-stage rifting was accommodated by an asymmetric shear that cut down into the mantle lithosphere (see also Whitmarsh et al. 2001; Manatschal 2004). Timing, duration and associated extension of the rift phases are not well constrained. We have constructed bounds on extension velocity and duration of each of the rift phases using various permissible rift-phase reconstructions and estimates of the total amount of crustal extension. The total crustal extension is estimated using the northern conjugate margin transect (Funck et al. 2003; Pe´rez-Gussinye´ et al. 2003) as a reference. Total extension of this part of the margin, prior to the onset of sea-floor spreading, ranges from 180 to 220 km, based on the cross-sectional area balance between an estimated original crustal thickness of 32 km and the present-day thickness distribution of the continental crust. This range overlaps with that based on a similar estimate of 100 –200 km derived from cross-sections in the central part of the Iberia– Newfoundland conjugate margin system (Minshull et al. 2001), although it is clear their estimate has a large uncertainty. This total extension must be apportioned among the different phases.
The timing of the rift phases can be determined approximately based on subsidence, from well data and seismic profiles (Tankard & Balkwill 1989), as a proxy for extension: Phase I:
218–205 Ma – minor relief formation in external basins; Phase II: 165–142 Ma – major relief formation in external basins (e.g. Jeanne d’Arc and associated basins, Galicia Interior Basin, Lusitanian Basin); Phase III: 142–122 Ma – Narrow rifting and breakup of the crust; Phase IV: 122–116 Ma – Asymmetric exhumation of the mantle lithosphere. Previous estimates of the extension velocity are of the order of 1.5 cm year21 full spreading velocity (Minshull et al. 2001). The approximate timing of rift phases and the value of 200 km total extension given above can be used to make simple estimates that give upper and lower bounds for the average extension velocity of the system. If the extension velocity was the same throughout all of the phases the rate was approximately 0.3 cm year21 full spreading rate. If all of the extension occurred during the latter two phases the estimated velocity would be approximately 0.75 cm year21. We regard this estimate to be an unreasonably large upper bound because phases I and II achieved significant normal faulting and basin subsidence. We therefore regard 0.5 cm year21 to be a representative upper-bound estimate of the average extension velocity during the phases when extension occurred.
GEOMETRY OF RIFTS AND RIFTED MARGINS
133
(a) Northern Transect A
Flemish Cap
A‘ B‘ PR PR
Galicia Bank
Galicia Interior Basin
Galicia
B
(b) Central Transect C J d’Arc B. Carson Bonnition B.
PR C‘
D‘ PR
Exhumed ML
Exhumed MLD
0
0
20
20
40
40 200
100
0
0
100
200
(c) Southern Transect E
E‘
Whale Basin
F‘ Exhumed ML Tagus Abysal Plain
F
0
0
20
20
40 0
100
200
300
400
500
0
100
200
40
300
(d) Model Crustal Section 0
200
400
600 x (km)
800
1000
(e) Model Application Exhumed Mantle Lithosphere
Crustal Wedge
Exterior Basins
t = 46 Ma, Δx = 230 km, V = 0.5 cm/a 0
z (km)
–50
1000 1000
–100
–150
1350
1350
–200 200
400
600 x (km)(S=0)
800
1000
Fig. 16. Compilation of crustal conjugate cross-sections based on seismic reflection data, and gravity inversion for the Iberia Newfoundland margin system, and comparison to numerical model application shown in Figure 17. Section locations indicated in Figure 15. (a) Northern transect (Murillas et al. 1990; Funck et al. 2003; Pe´rez-Gussinye´ et al. 2003). (b) Central Transect (Keen et al. 1987a, b; Keen & de Voogd 1989; Dean et al. 2000). (c) Southern transect (Keen 1987a, b; Keen & de Voogd 1989; Torne et al. 1994). (d) Crustal section from numerical model application for comparison at the same horizontal and vertical scale as the crustal sections. (e) Numerical model after 230 km of extension at 46 Ma Note the formation of exterior basins, the thin crustal wedge on top of the mantle lithosphere and exhumed mantle lithosphere in the ocean–continent transition zone.
In summary, from the discussion above, a successful model needs to meet the following constraints: † an initially wide rifting mode with extension/subsidence focused in a number of distinct, widely distributed, fault-bounded basins, with structure that varies with position along the system;
† Initial extension in these basins was mostly limited to the crust; † Abandonment of the external basins in favour of late-stage narrow, possibly asymmetric, rifting; † exhumation of the mantle lithosphere during the final phases of breakup; † An overall slow extension rate probably close to 0.5cm year21.
134
R. S. HUISMANS & C. BEAUMONT
Towards a first-order model with the characteristics of the Iberia– Newfoundland conjugate margin system In the context of the models described in this paper, the progression from a wide to narrow rift can be met by models that inherit heterogeneous crustal noise and in which the crustal noise develops into a wide region of extension characterized by weak upper crustal shears and faults that bound distinct basins. Horizontal decoupling and shear in the lower or mid crust appear to be necessary, in addition to the inherited heterogeneity, (cf. Fig. 14) because the wide rifting mode does not develop when the lower crust is strong and precludes decoupling (cf. Fig. 13). The progression from a wide to a narrow rift occurs naturally in this type of model if the uppermost mantle lithosphere strainsoftens, leading to localization of the deformation (cf. Figs 13 & 14). It also appears that models with a strong, coupled, lower crust show the greatest potential to exhume mantle lithosphere to the surface during the narrow rifting mode (e.g. Figs 5b & 13c) and that this exhumation is commonly achieved by asymmetric extension (cf. Fig. 4b, c). These behaviours are also compatible with slow rifting velocities that favour AP and SP modes over distributed pure shear, as described in the section on ‘Analysis of simple models using the principle of minimum energy dissipation’. The model shown in Figure 14 has many of the required general characteristics displayed by the Newfoundland–Iberia conjugate margin system, as listed above. However, it is a model with a weak lower crust and, therefore, fails to exhume the mantle to the surface during the late narrow rifting phase. To date, we have been unable to create a selfconsistent model that produces both initial wide rifting and later stage cooling that strengthens the viscous lower crust sufficiently to create the necessary conditions for very narrow AP or SP rifting and mantle exhumation. We suspect this inability occurs because we consider models in which extension is uniform and continuous, and, therefore, there are no tectonically quiescent periods during which significant cooling can occur. We demonstrate the effect of cooling and strengthening of the extended crust using a composite model. The rift evolution starts in a model with inherited strain heterogeneity and a weak middle and lower crust. We then simulate the effects of cooling and late-stage embrittlement of the crust by suppressing the viscous flow in the crust during Phase 2. This composite behaviour is obtained by increasing crustal viscosity by a factor 100 during Phase 2 to achieve a crust that is totally in the frictional–plastic regime. The extension velocity is
0.5 cm year21. During the 24 Ma-long Phase 1, 120 km of extension is accommodated by localization on multiple sets of frictional shear zones in the upper crust (Fig. 17a). Deformation of the mantle lithosphere occurs in a distributed mode. During Phase 2, localization in the crustal layer feeds back into the upper mantle lithosphere where it causes localization in the frictional–plastic uppermost mantle. Deformation is asymmetric (Fig. 17b) and mantle lithosphere is exhumed during the final phases of breakup (Fig. 17c). To first order, this model result is compatible with the constraints compiled for the Newfoundland–Iberia natural system (Fig. 16), particularly when the significant variability among cross sections located close together is taken into account (Fig. 16a– c). It also provides a potential explanation for this variability in the geometry of the margins, in particular the initial irregular distribution and later abandonment of early rift basins that flank the margins if these are the consequence of deformation related to inherited crustal heterogeneity. We anticipate that an equivalent fully three-dimensional (3D) model, with inherited initial heterogeneous strain, would develop sets of anastamosing rift basins during the early phase of rifting and that the structure of these basins would vary significantly along strike of the system. These preliminary results lead us to the proposition that the early stages of rifting in margins of this type are disorganized, reflecting inherited lithospheric characteristics. Only later in the rifting does the system overcome this inheritance and become organized in the form of a narrow rift.
Summary and conclusions This article has focused on models of extensional processes with application to lithospheric extension during rifting and the formation of rifted continental margins. Even though rifts and rifted margins have been studied intensively for more than 30 years, a fundamental understanding of the factors that control the extensional process, the distribution of extension and the associated rift and rift margin geometry is still lacking. The purpose of the model studies reported here and in associated articles (Huismans & Beaumont 2002, 2003; Huismans et al. 2005) is to test a limited range of conceptual models that combine distributed viscous–plastic flows with feedback mechanisms, specifically plastic strain softening, that lead to localized deformation. This localization allows the models to represent faults and shear zones approximately. In our earlier research, the development of localized shears was in part deterministic and was triggered by a priori inclusion of small weak regions, ‘weak seeds’, in
GEOMETRY OF RIFTS AND RIFTED MARGINS
(a) 0
135
t = 24 Ma, Δx= 120 km, V = 0.5 cm/a
–20 z (km)
–40 –60 –80 –100 –120 400
500
600
700
800
600
700
800
600 x (km)
700
800
Restart with Strong Lower Crust (b) 0
t = 32 Ma, Δx = 160 km
–20 z (km)
–40 –60 –80 –100 –120 400
(c)
500
t = 40 Ma, Δx = 200 km
0
–20 z (km)
–40 –60 –80 –100 –120 400
500
Fig. 17. Model application to the Iberia– Newfoundland conjugate margin system. (a) Initial weak lower-crust thermo-mechanical model of lithosphere extension with statistical noise applied only to the crust. Wide-rift mode selection, only localization in the crust, very widely distributed deformation in the mantle lithosphere. (b) and (c) Model restarted with a strong lower crust, simulating a phase of cooling and strengthening of the lower crust. Narrow rift mode selection, feedback of localization in the crust into the mantle lithosphere. Note the progression from wide to narrow rift mode and model selects a narrow rift mode with strong asymmetry and exhumation of mantle lithosphere during the final stages of rifting.
136
R. S. HUISMANS & C. BEAUMONT
the models. In this paper we have reviewed those results and expanded the research to include the effect of distributions of weak heterogeneities that take the form of white noise in the initial strain field and can be interpreted to represent strain, or in a wider sense damage, that is inherited. The effects of the viscous-plastic layered rheology combined with the inherited initial strain lead to more general representations of the type of deformation that can occur during lithospheric extension. We believe this approach, which combines a continuum model based on plastic and viscous rheologies with strain softening that results from a statistical distribution of inherited heterogeneity, provides a potential explanation for the general features of the Newfoundland–Iberia conjugate margins. These margins appear to have had an early phase of distributed disorganized crustal extension, and only later became organized into a narrow rift that exhumed continental mantle lithosphere during terminal extension prior to the onset of sea-floor spreading. We wish to thank the organizers of the IMEDL workshop for the invitation to participate. This research was funded through an ACOA–Atlantic Innovation Fund contract, and an IBM-Shared University Research grant. C. Beaumont was funded by the Canada Research Chair in Geodynamics. C. Beaumont also acknowledges support through the Inco Fellowship of the Canadian Institute for Advanced Research. Numerical calculations used software developed by P. Fullsack. We thank R. Buck, L. Lavier and an anonymous reviewer for the constructive and thorough reviews.
References B EHN , M. D., L IN , J. & Z UBER , M. T. 2002. A continuum mechanics model for normal faulting using a strainrate softening rheology: implications for thermal and rheological controls on continental and oceanic rifting. Earth and Planetary Science Letters, 202, 725–740. B IRD , P. & Y UEN , D. A. 1979. The use of the minimum dissipation principle in tectonophysics. Earth and Planetary Science Letters, 45, 214–217. B OILLOT , G., C OMAS , M. C. ET AL . 1988. Preliminary results of the Galinaute cruise; dives of the submersible Nautile on western Galicia margin, Spain. Proceedings of the Ocean Drilling Program, Scientific Results, 103. Ocean Drilling Program, College Station, TX, 37– 51. B OILLOT , G., B ESLIER , M. O. & C OMAS , M. 1992. Seismic image of undercrusted serpentine beneath a rifted margin. Terra Nova, 4, 25–33. B OS , B. & S PIERS , C. J. 2002. Frictional-viscous flow of phyllosilicate-bearing fault rock: microphysical model and implications for crustal strength profiles. Journal of Geophysical Research, 107, 2028, doi:10,129/ 2001JB000301. B RUN , J. P. & B ESLIER , M. O. 1996. Mantle exhumation at passive margins. Earth and Planetary Science Letters, 142, 161 –173.
B RUN , J. P. & G UTSCHER , M. A. 1992. Deep crustal structure of the Rhine Graben from DEKORP– ECORS seismic reflection data: a summary. Tectonophysics, 208, 139 –147. B UCK , W. R. 1991. Modes of continental lithospheric extension. Journal of Geophysical Research, 96, 20161–20178. B UCK , W. R., L AVIER , L. L. & P OLIAKOV , A. N. B. 1999. How to make a rift wide. In: W HITE , R. S., H ARDMAN , R. F. P., W ATTS , A. B. & W HITMARSH , R. B. (eds) Philosophical Transactions–Royal Society. Mathematical, Physical and Engineering Sciences, 357. Royal Society of London, 671–693. B UCK , W. R., L AVIER , L. L. & B ABEYKO , A. 2003. A numerical model of lithospheric extension producing fault bounded basins and ranges. International Geology Reviews, 45, 712 –723. C HE´ RY , J., Z OBACK , M. D. & H ASSANI , R. 2001. An integrated mechanical model of the San Andreas fault in central and northern California. Journal of Geophysical Research, 106, 22,051– 22,066. C ONNOLLY , J. A. D. & P ODLADCHIKOV , Y. Y. 2000. Temperature-dependent viscoelastic compaction and compartmentalization in sedimentary basins. Tectonophysics, 324, 137 –168. D EAN , S. M., M INSHULL , T. A., W HITMARSH , R. B. & L OUDEN , K. E. 2000. Deep structure of the oceancontinent transition in the southern Iberia abyssal plain from seismic refraction profiles; the IAM-9 transect at 40 degrees 20’N. Journal of Geophysical Research, 105, 5859–5885. D RISCOLL , N. W., H OGG , J. R., C HRISTIE -B LICK , N. & K ARNER , G. D. 1995. Extensional tectonics in the Jeanne d’Arc Basin, offshore Newfoundland: implications for the timing of breakup between Grand Banks and Iberia. In: S CRUTTON , R. A., S TOKER , M., S HIMMIELD , G. B. & T UDHOPE , A. W. (eds) The Tectonics Sedimentation and Palaeoceanography of the North Atlantic Region. London, Geological Society, London, Special Publications, 90, 1– 28. F ULLSACK , P. 1995. An arbitrary Lagrangian– Eulerian formulation for creeping flows and applications in tectonic models. Geophysical Journal International, 120, 1– 23. F UNCK , T., H OPPER , J. R., L ARSEN , H. C., L OUDEN , K. E., T UCHOLKE , B. E. & H OLBROOK , W. S. 2003. Crustal structure of the ocean–continent transition at Flemish Cap; seismic refraction results. Journal of Geophysical Research, 108, 2531, doi:10.1029/2003JB002434. F UNCK , T., J ACKSON , H. R., L OUDEN , K. E., D EHLER , S. A. & W U , Y. 2004. Crustal structure of the northern Nova Scotia rifted continental margin (eastern Canada). Journal of Geophysical Research, 109, B09102, doi:10.1029/2004JB003008. G LEASON , G. C. & T ULLIS , J. 1995. A flow law for dislocation creep of quartz aggregates determined with the molten salt cell. Tectonophysics, 247, 1 –23. H ANDY , M. R. & S TUENITZ , H. 2002. Strain localization by fracturing and reaction weakening; a mechanism for initiating exhumation of subcontinental mantle beneath rifted margins. In: DE M EER , S., D RURY , M. R., DE B RESSER , J. H. P. & P ENNOCK , G. M. (eds) Deformation Mechanisms, Rheology and
GEOMETRY OF RIFTS AND RIFTED MARGINS Tectonics. Current Status and Future Perspectives. Geological Society, London, Special Publications, 200, 387– 407. H OPPER , J. R., F UNCK , T. ET AL . 2004. Continental breakup and the onset of ultraslow seafloor spreading off Flemish Cap on the Newfoundland rifted margin. Geology, 32, 93–96. H UISMANS , R. S. & B EAUMONT , C. 2002. Asymmetric lithospheric extension: the role of frictional–plastic strain softening inferred from numerical experiments. Geology, 30, 211–214. H UISMANS , R. S. & B EAUMONT , C. 2003. Asymmetric lithospheric extension: relative importance of frictional–plastic and viscous strain softening inferred from numerical experiments. Journal of Geophysical Research, 108, 2496, doi:10.1029/2002JB002026. H UISMANS , R. S., P ODLADCHIKOV , Y. Y. & C LOETINGH , S. A. P. L. 2001. The transition from passive to active rifting, relative importance of asthenospheric doming and passive extension of the lithosphere. Journal of Geophysical Research, 106, 11,271–11,291. H UISMANS , R. S., B UITER , S. J. H. & B EAUMONT , C. 2005. Effect of plastic– viscous layering and strain softening on mode selection during lithospheric extension. Journal of Geophysical Research, 110, doi: 10.1029/2004JB003114. I NGEBRITSEN , S. E. & M ANNING , C. E. 1999. Geological implications of a permeability – depth curve for the continental crust. Geology, 27, 1107–1110. K ARATO , S. & W U , P. 1993. Rheology of the upper mantle. Science, 260, 771–778. K EEN , C. E. & DE V OOGD , B. 1989. The continent–ocean boundary at the rifted margin off eastern Canada: new results from deep seismic reflection studies. Tectonics, 7, 107– 124. K EEN , C. E., B OUTILIER , R., V OOGD , B. D., M UDFORD , B. & E NACHESCU , M. E. 1987a. Crustal geometry and extensional models for the Grand Banks, eastern Canada: constraints from deep seismic reflection data. In: B EAUMONT , C. & T ANKARD , A. J. (eds) Sedimentary Basins and Basin-forming Mechanisms. Canadian Society of Petroleum Engineers Memoir, 12, 101–115. K EEN , C. E., S TOCKMAL , G. S., W ELSINK , H. J., Q UINLAN , G. & M UDFORD , B. 1987b. Deep crustal structure and evolution of the rifted margin northeast of Newfoundland: results from LITHOPROBE East. Canadian Journal of Earth Sciences, 24, 1537–1549. K EEN , C. E., P EDDY , C., DE V OOGD , B. & M ATTHEWS , D. 1989. Conjugate margins of Canada and Europe: Results from deep reflection profiling. Geology, 17, 173–176. L AVIER , L. L. & B UCK , R. W. 2002. Half graben versus large-offset low-angle normal fault: Importance of keeping cool during normal faulting. Journal of Geophysical Research, 107, 2122, doi:10.1029/ 2001JB000513. L OUDEN , K. E. & C HIAN , D. 1999. The deep structure of non-volcanic rifted continental margins. Philosophical Transactions of the Royal Society of London, A357, 767–804. M ANATSCHAL , G. 2004. New models for evolution of magma-poor rifted margins based on a review of
137
data and concepts from West Iberia and the Alps. International Journal of Earth Sciences, 93, 432–466. M ASEK , J. G. & D UNCAN , C. C. 1998. Minimum-work mountain building. Journal of Geophysical Research, 103, 907–917. M AUFFRET , A. & M ONTADERT , L. 1988. Seismic stratigraphy off Galicia. Proceedings of the Ocean Drilling Program, Scientific Results, 103. Ocean Drilling Program, College Station, TX, 13–30. M C K ENZIE , D. 1978. Some remarks on the development of sedimentary basins. Earth and Planetary Science Letters, 40, 25–32. M INSHULL , T. A., D EAN , S. M., W HITE , R. S. & W HITMARSH , R. B. 2001. Anomalous melt production after continental breakup in the southern Iberia abyssal plain. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volanic Rifting of Continental Margins: A Comparison of Evidence From Land and Sea. Geological Society, London, Special Publications, 187, 537– 550. M U¨ NTENER , O. & M ANATSCHAL , G. 2006. High degrees of melt extraction recorded by spinel harzburgite of the Newfoundland margin: The role of inheritance and consequences for the evolution of the southern North Atlantic. Earth and Planetary Science Letters, 252, 437– 452. M URILLAS , J., M OUGENOT , D., B OILLOT , G., C OMAS , M. C., B ANDA , E. & M AUFFRET , A. 1990. Structure and evolution of the Galicia interior basin (Atlantic western Iberian continental margin). Tectonophysics, 184, 297–319. M UTTER , J. C., L ARSON , R. L. & G ROUP , N. A. S. 1989. Extension of the Exmouth Plateau, offshore northwestern Australia: Deep seismic reflection/refraction evidence for simple and pure shear mechanisms. Geology, 17, 15–18. P E´ REZ -G USSINYE´ , M., R ANERO , C. R., R ESTON , T. J. & S AWYER , D. 2003. Mechanisms of extension at nonvolcanic margins; evidence from the Galicia interior basin, west of Iberia. Journal of Geophysical Research, 108, doi:10.1029/2001JB000901. P ROVOST , A.-S. & H OUSTON , H. 2003. Stress orientations in Northern and Central California; evidence for the evolution of frictional strength along the San Andreas plate boundary system. Journal of Geophysical Research, 108, 18. R ICE , J. R. 1992. Fault stress states, pore pressure distributions, and the weakness of the San Andreas Fault. In: E VANS , B. & W ONG , T. F. (eds) Fault Mechanics and Transport Properties of Rocks. Academic Press, London, 475–503. R IDLEY , J. 1993. The relations between mean rock stress and fluid flow in the crust: With reference to vein- and lode-style gold deposits. Ore Geology Reviews, 8, 23–37. R UPPEL , C. 1995. Extensional processes in continental lithosphere. Journal of Geophysical Research, 100, 24,187–24,215. S ALVESON , J. O. 1978. Variations in the geology of rift basins – A tectonic model. Proceedings of the International Symposium on the Rio Grande Rift. Los Alamos Science Laboratory, 82–86.
138
R. S. HUISMANS & C. BEAUMONT
S CHOLZ , C. H. 1990. The Mechanics of Earthquakes and Faulting. Cambridge University Press, Cambridge. S CHOLZ , C. H. 2000. Evidence for a strong San Andreas Fault. Geology, 28, 163– 166. S ENGOR , A. M. C. & B URKE , K. 1978. Relative timing of rifting and volcanism on Earth and its tectonic applications. Geophysical Research Letters, 5, 419– 421. S IBSON , R. H. 1990. Conditions for fault-valve behaviour, In: K NIPE , R. J. & R UTTER , E. H. (eds) Deformation Mechanisms, Rheology and Tectonics. Geological Society, Special Publications, London, 15–28. S IBUET , J.-C. 1992. Formation of non-volcanic passive margins: a composite model applied to the conjugate Galicia and southeastern Flemish Cap margins. Geophysical Research Letters, 19, 769– 773. S LEEP , N. H., S TEIN , S., G ELLER , R. J. & G ORDON , R. G. 1979. Comment on ‘The use of the minimumdissipation principle in tectonophysics’. Earth and Planetary Science Letters, 45, 218–220. S RIVASTAVA , S. P., S IBUET , J.-C., C ANDE , S., R OEST , W. R. & R EID , I. R. 2000. Magnetic evidence for slow seafloor spreading during the formation of the Newfoundland and Iberian margins. Earth and Planetary Science Letters, 182, 61– 76. S TAPEL , G., C LOETINGH , S. & P RONK , B. 1996. Quantitative subsidence analysis of the Mesozoic evolution of the Lusitanian Basin (western Iberian margin). Tectonophysics, 266, 1 –4. S TREIT , J. E. 1997. Low frictional strength of upper crustal faults; a model. Journal of Geophysical Research, 102, 24,619–24,626. T ANKARD , A. J. & B ALKWILL , H. R. 1989. Extensional tectonics and stratigraphy of the North Atlantic margins; introduction. In: T ANKARD , A. J. &
B ALKWILL , H. R. (eds) Extensional Tectonics and Stratigraphy of the North. Atlantic Margins. AAPG Memoir, 46, 1– 6. T ANKARD , A. J. & W ELSINK , H. J. 1989. Mesozoic extension and styles of basin formation in Atlantic Canada. In: T ANKARD , A. J. & B ALKWILL , H. R. (eds) Extensional Tectonics and Stratigraphy of the North. Atlantic Margins. AAPG Memoir, 46, 175– 195. T ANKARD , A. J., W ELSINK , H. J. & J ENKINS , W. A. M. 1989. Structural styles and stratigraphy of the Jeanne d’Arc Basin, Grand Banks of Newfoundland. AAPG Memoir, 46, 265–282. T ORNE , M., F ERNANDEZ , M., C ARBONELL , J. & B ANDA , E. 1994. Lithospheric transition from continental to oceanic in the West Iberia Atlantic margin. In: B ANDA , E., T ORNE , M. & T ALWANI , M. (eds), Rifted Ocean– Continent Boundaries. Kluwer, London, 247–264. T UCHOLKE , B. E., S IBUET , J.-C. ET AL . 2004. Proceedings of the Ocean Drilling Program, Initial Reports, 210, Ocean Drilling Program, College Station, TX, doi:10.2973/odp.proc.ir.210.2004. T URCOTTE , D. L. & E MMERMAN , S. H. 1983. Mechanisms of active and passive rifting. Tectonophysics, 94, 39–50. W ERNICKE , B. 1985. Uniform-sense normal simple shear of the continental lithosphere. Canadian Journal of Earth Science, 22, 108–125. W ERNICKE , B. & B URCHFIELD , B. C. 1982. Modes of extensional tectonics. Journal of Structural Geology, 4, 105–115. W HITMARSH , R. B., M ANATSCHAL , G. & M INSHULL , T. A. 2001. Evolution of magma-poor continental margins from rifting to seafloor spreading. Nature, 413, 150– 154. W ILLETT , S. D. 1999. Rheological dependence of extension in wedge models of convergent orogens. Tectonophysics, 305, 419 –435.
The role of gravitational instabilities, density structure and extension rate in the evolution of continental margins E. BUROV Laboratory of Tectonics UMR 7072, University of Pierre and Marie Curie, Paris, France Abstract: Formation of rifted continental margins is associated with localized thinning and breakup of the continental lithosphere, driven or accompanied by the ascent of the lithosphere– asthenosphere boundary. Thinning creates sharp density and viscosity contrasts and steep boundaries between cold deformed lithosphere and hot upwelling asthenosphere, thus providing conditions for the development of positive (asthenosphere) and negative (mantle lithosphere) Rayleigh–Taylor (RT) instabilities. The evolution of many continental margins (e.g. Newfoundland margin and Iberian margin) is characterized by slow spreading rates. This allows the RT instabilities to grow at the timescale of rifting. The impact of positive RT instabilities (asthenospheric upwelling) is well studied. The negative RT instabilities, associated with mantle downwelling, remain an overlooked factor. However, these instabilities should also affect the rift evolution, in particular, they may cause mantle thinning or thickening below the rift flanks. Our numerical experiments suggest that the ratio of the RT-growth rate to the extension rate controls the overall rift geometry and evolution. Even if the effect of negative RT instabilities is more important for slow extension rates of 2 5 mm year21 (Deborah number, De , 1), it is still significant for 2– 3 times higher extension rates of 2 15 mm year21 (De , 10). The numerical experiments for extension rates of 2 15 mm year21 and mantle–asthenosphere density contrasts of 10–20 kg m23 demonstrate a number of structural similarities with continental margins characterized by low De (e.g. Flemish Cap and Galicia margin). In particular, rift asymmetry results from interplay between the RT instabilities and differential stretching at De , 1. Formation of interior basins occurs at De 1 –3. The best correspondence with the observed geometry of rifted margins is obtained for chemical density contrast of 20 kg m23 and extension rate of 2 15 mm year21, which is twice that of the averaged values inferred from the observations. This suggests that margins may initially (prebreakup stage) extend at higher rates than the average extension rates characterizing rift evolution. The influence of RT instabilities is strongly controlled by extension rate, density, rheology and thermal structure of the lithosphere; this implies that we need better constraints on these parameters from the observations.
Continental margins result from rifting characterized by large coefficients of extension 4–5 , b , 15 (e.g. Bott 1971; Salvenson 1978; Cloetingh et al. 1982; Buck 1991). Rifting processes involve laterally variable thinning of ductile layers of different densities. If a denser layer (mantle lithosphere) is located on top of a lighter layer (asthenosphere), then the system is gravitationally unstable and may develop negative Rayleigh –Taylor (RT) instabilities. This possibility specifically refers to rifting, passive or active, where hot light asthenosphere ascends to the surface and replaces colder denser lithosphere. In most situations, normal lithospheric mantle is approximately 10 –30 kg m23 denser than the underlying asthenospheric layer, basically due to its colder temperature but also due to compositional differences (e.g. Stacey 1992; O’Reilly et al. 2001; Poudjom Djomani et al. 2001; Turcotte & Schubert 2002). Some authors (e.g. Houseman et al. 1981) assume even much higher total density differences of up to 100 kg m23. The density differences of the order
of 20 kg m23 are commonly accepted for the Phanerozoic lithosphere, even though there is still a debate about whether it applies to the presumably Mg-rich and -depleted cratonic lithosphere. Irrespective, volumetric seismic velocities, which are generally considered as proxy for density, are systematically higher in the lithosphere mantle than in the asthenosphere. Depending on its viscosity the mantle lithosphere therefore has the potential to sink as the result of a RT instability (e.g. Houseman et al. 1981). The growth rate of the RT instabilities is directly proportional to the density contrast and inversely proportional to the thickness and the viscosity of the upper layer (if the viscosity of the lower layer is small). Depending on that, the instabilities may be rapid or slow compared to the tectonic deformation rates. In the first case, they will influence the rift evolution, but not in the second. The conditions when the RT instabilities are slow apply to the cases when the viscosity of the mantle lithosphere is higher than 1022 Pa s, when the density contrast between the lithosphere
From: KARNER , G. D., MANATSCHAL , G. & PINHEIRO , L. M. (eds) Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, London, Special Publications, 282, 139– 156. DOI: 10.1144/SP282.7 0305-8719/07/$15.00 # The Geological Society of London 2007.
140
E. BUROV
and the asthenosphere is small, or when the interface between the mantle and the asthenosphere remains flat (e.g. Houseman & Molnar 1997). This applies to oceanic and continental plates at
normal conditions that infer smooth boundaries and smooth vertical temperature profiles and practical absence of lateral heat transport (e.g. prerift geotherm in Fig. 1a).
Fig. 1. (a) Simplified cartoon showing the interaction between the asthenosphere and mantle lithosphere during rifting. Left: the mantle lithosphere below the rift flanks becomes gravitationally unstable owing to the negative density contrast with the hot asthenosphere and the loss of mechanical resistance resulting from lateral heat transfer from the asthenosphere below the rifted zone. The asthenosphere is positively unstable because of its positive density contrast with the embedding mantle lithosphere. As the viscosity of the mantle lithosphere is exponentially dependent on temperature, there is a relatively net ‘stability level’ that separates the regions of high viscosity from thermally weakened regions of low viscosity. These regions are subject to development of rapid RT instabilities. Right: simplified yield-stress prerift envelope of the continental lithosphere and typical prerift temperature profile. (b) Left: potential density contrast (reference density at 750 8C) and viscosity of olivine mantle, as function of temperature (left). Viscosity is shown for three representative values of the background strain rate (10214, 10215 and 10216 s21). Right: time until full detachment of the unstable mantle layer, as a function of initial layer thickness (olivine rheology), computed according to Conrad & Molnar (1999). The initial perturbation amplitude is w0. Note that for a normal depth– pressure–temperature profile of mantle lithosphere, the density contrast due to thermal expansion is nearly negated by the effect of compressibility (Turcotte & Schubert 2002). This is not the case of rifting when the mantle is heated without compensatory increase in pressure.
THE ROLE OF GRAVITATIONAL INSTABILITIES
Contrary to normal lithosphere, the rifted margins such as the Newfoundland and Iberian margin (Manatshal & Bernoulli 1999; Manatshal et al. 2001; Funck et al. 2003; Pe´rez-Gussinye´ et al. 2003; Hopper et al. 2004), as well as many ‘common’ continental rifts, constitute a perfect environment for development of gravitational instabilities owing to the presence of steep density boundaries and viscosity contrasts resulting from rifting (e.g. Huismans et al. 1998) (Fig. 1a). As mentioned above, most data suggest the existence of a compositional or chemical density contrast of 0–20 kg m23 between the mantle lithosphere and asthenosphere (e.g. Stacey 1992; O’Reilly et al. 2001; Poudjom Djomani et al. 2001; Turcotte & Schubert 2002), except probably the cratonic mantle (our settings consider, however, a noncratonic lithosphere). In addition to potential chemical density contrasts, strong thermal density contrasts of 30– 60 kg m23 arise when the mantle–asthenosphere boundary is locally uplifted owing to the rifting. In this case, horizontal heat conduction becomes at least as important as the vertical heat conduction. For the temperature contrasts of 500 –1000 8C, associated with moderate coefficients of extension b ¼ 2–5 (McKenzie 1978), horizontal thermal gradients may reach 10 –20 8C km21. The resulting lateral heat flow is of the order of 30–60 mW m22, which is higher than the normal mantle heat flow (30 mW m22: Jaupart & Marschal 1999; Turcotte & Schubert 2002). The consequences of the lateral heat transfer on the density structure during the rifting has been extensively studied in previous literature (e.g. Royden & Keen 1980; Stephenson et al. 1989). These studies have provided a number of important corrections to the most common one-dimensional (1D) thermal subsidence models (McKenzie 1978). The thermo-mechanical consequences of the lateral heat exchanges were also investigated (e.g. Keen & Boutilier 1995), yet mainly in terms of their impact on the mantle viscosity and on the development of positive asthenosphere instabilities under the rift axis (transition from passive to active rifting, post-rift doming). In particular, Huismans et al. (1998) conducted a detailed study of differential stretching of inelastic lithosphere in terms of its interaction with the unstable asthenosphere. Yet, the negative mantle instabilities were beyond the scope of this model because the experiments of Huismans et al. (1998) were limited to the upper 120 km of the lithosphere instead of at least 250 km needed to account for the mantle RT instabilities below the rift flanks. The asthenosphere has a low viscosity (5 1019 Pa s) and can flow at timescales of 1 ka. The associated heat advection is instantaneous compared to the rifting timescales that exceed 0.1 –1 Ma.
141
For b ¼ 2, the temperature of the asthenospheric material advected to a depth of 70 –90 km below the stretched lithosphere is around 1330 8C. The temperature of the prerift mantle lithosphere at this depth is only 750–800 8C (e.g. Kirby & Kronenberg 1987), (Fig. 1a). Consequently, even for a moderate extension, the temperature contrast, DT, between the upraised asthenosphere and the surrounding mantle lithosphere is around 500– 600 8C. The associated thermal density contrast, Drt, is 50–60 kg m23 (Drt ¼ r0aDT, where r0 ¼ 3330 kg m23 and a ¼ 3 1025 K21). Drt sums up with the chemical density contrast Drc of 10 –20 kg m23 yielding important ‘available negative buoyancy’ (Conrad & Molnar 1999). The total density contrast may be 2–3 times higher than the contrast of 10 –20 kg m23 that is sufficient to drive mantle instabilities at time scales of 2 –10 Ma (Conrad & Molnar 1999). Of course, the initially higher density contrast vanishes as the heat propagates from the upwelling asthenosphere to the mantle lithosphere. This decrease in density contrast slows down the RT instability. Yet, as the temperature rises, the mantle viscosity decreases. The reduction of viscosity, in contrast to that of density, accelerates the RT instability. Hence, the impact of lateral heat propagation on the growth rate of instability depends on the ratio of the density change rate, @r(T)/@T, to the viscosity change rate, @m(T)/@T. The density change rate @r(T)/@T is linearly proportional to temperature: @r(T)/@T ¼ ar, where a is the coefficient of thermal expansion (a ¼ 3 1025 K21). The viscosity change rate is super-exponential: for dislocation creep, m(T) ¼ m0exp(Q/nRT), from where @m(T)/@T ¼ 2m0Q(nRT 2)21exp(Q/nRT), where Q is activation enthalpy (5 105 J mol21), n 3 and R is the universal gas constant (see Table 2). A 200 8C temperature rise from 750 to 800 8C (mechanical bottom of the lithosphere: Kirby et al. 1987) reduces the initial density contrast between the mantle and asthenosphere by 20 kg m23 (Turcotte & Schubert 2002). Yet, this temperature rise also reduces the effective viscosity by factor of 104 (!). Such reduction is sufficient to bring the system from a mechanically stable state (m ¼ 1023 – 1024 Pa s) to an extremely unstable state at m ¼ 1019 –1020 Pa s (e.g. Conrad & Molnar 1999). Consequently, temperature rise in the mantle near the lithosphere– asthenosphere boundary will accelerate the growth rate of RT instability rather then reduce it. The question on the significance of the negative RT instabilities for slow rifting is naturally related to the problem of possible variation of the extension rate during the rifting phase. The extension rates deduced for many rift systems are averaged over some important time spans. Yet, the mechanical
142
E. BUROV
response of the brittle –ductile lithosphere depends on the strain rate at each moment of time. If the extension rate changes during the active phase of rifting, the final rift structures may differ even for similar amounts of extension. No doubt, there is a large number of other factors that may crucially influence the evolution of rifted margins. In particular, inherited structures, heterogeneities, rheology or surface processes are certainly of great importance. Yet, their influence is well studied and discussed in the previous work (e.g. Kusznir & Karner 1985; Dunbar & Sawyer 1988; Bassi 1995; Burov & Cloetingh 1997; Huismans & Beaumont 2007). We thus focus our study on more overlooked factors: negative gravitational stabilities and initial density structure of the lithosphere.
Theoretical background A number of studies (e.g. Houseman et al. 1981; Fletcher & Hallet 1983; Bassi & Bonnin 1988; Houseman & Molnar 1997; Huismans et al. 1998) have investigated the development of RT instabilities in normal, extended or shortened continental lithosphere as a function of thermo-rheological parameters. Based on these studies, the first-order condition for instable behaviour is formulated as: Drgd=2tmax , d=dfold where Dr is the density contrast between the mantle and the asthenosphere, g is the acceleration due to gravity, d is the thickness of the lithosphere, tmax is its maximal mechanical strength and dfold is the characteristic decay length of ductile strength/viscosity. For mantle lithosphere (olivine) d/dfold is around 10 and Drgd/2tmax is negative or smaller than 3 (Fletcher & Hallet 1983; Bassi & Bonnin 1988). This indicates that the mantle lithosphere is instable and may need only a small perturbation to develop a gravitational instability. Rifting produces a very strong perturbation of the mantle– asthenosphere interface and increases density contrasts, thus almost inevitably leading to unstable behaviour of the mantle layer. The results of simple linear analysis demonstrate that the growth rate of instability is mainly controlled by the density contrast and the viscosity of the mantle lithosphere (Chandrasekhar 1961; Fletcher & Hallet 1983). Let us first consider a simplified case of a layer of Newtonian fluid with viscosity m, density rm and thickness d placed on top of a weak fluid layer (asthenosphere) of lower density ra. The growth rate 1/t of the developing instability w (vertical deflection of the layer) with initial amplitude w0 depends on m, on the wavelength of perturbation, l, on the thickness of
the layer, d, and on the density contrast, Dr (Chandrasekhar 1961): w ¼ w0 expðt=tÞ t ¼ 4mðDrgdÞ1 ðv þ ðshðv1 Þchðv1 ÞÞ1 Þ ðv2 thðvÞ shðv1 Þchðv1 ÞÞ1
ð1Þ
where v ¼ l/2pd ¼ (kd)21, k is a wave number. The most rapidly growing wavelength is l ¼ 2.568d, which corresponds to minimal growth time: tmin ¼ 13:04mðDrgdÞ1 : As can be seen, the growth time is directly proportional to the fluid viscosity; so that it is less than 1 Ma for viscosities of m ¼ 1020 Pa s (assuming d ¼ 100 km and Dr ¼ 20 kg m23, Fig. 1b). High (.1022 Pa s) viscosity is needed to keep mantle layer quasi-stable for 10–15 Ma. For viscosities of the order of 1020 –1021 Pa s the mantle may be destabilized at timescales of 1–10 Ma that are perfectly compatible with typical timescales of continental rifting. The viscosity of m ¼ 1020 Pa s is usually attributed to the transition zone between the mantle and asthenosphere, which suggests that this zone can be instable. A number of analytical studies have extended equation 1 to the cases of realistic non-linear rheologies (e.g. Conrad & Molnar 1999). These studies show that in the nonlinear case the growth rate is about n times higher than in the case of Newtonian rheology, where n is the power-law exponent n (n ¼ 3 for most olivine rheologies). In the case of rifted continental lithosphere, a multitude of factors such as the rheological stratification and rift geometry affect the viscosity, temperature and the growth rate of instability (Fig. 1a, which justifies the numerical approach developed in this study). For example, as the lithosphere thins, d decreases and as a result the growth rate and wavelength of new instabilities might decrease too. However, thinning results in heating of the mantle layer. Consequently, the effect of thinning on the growth rate is mitigated owing to temperatureinduced viscosity drop. The importance of RT instabilities during rift development can be characterized by the ratio of the characteristic growth time of the instability to the characteristic time of the rifting phase, tr, expressed by means of Deborah number, De: De ¼ tmin =tr ¼ 13:04m=ðDrgdtr Þ: If De 1 the instabilities do not effect rift evolution. If De 1 rift evolution is highly affected
THE ROLE OF GRAVITATIONAL INSTABILITIES
by RT instabilities. A more conventional way to define De would be to set it as a ratio of the Maxwell relaxation time tm to flow time. Yet, in our case, the development of gravitational instabilities appears to be a leading process, which becomes clear from an estimate of the ratio of tmin to tm: tmin =tm 13G=Drgd 5E=Drgd
tr ¼ L=ux where L is the width of the rift and ux is the rate of extension. Consequently, ð2Þ
We can further introduce the effective viscosity for ductile creep law: dð1nÞ=n
meff ¼ ek
ðA Þ1=n expðH=nRTÞ
where e dk ¼ Invk (eij))1/2 is the effective strain rate and A* ¼ 1/2A . 3(nþ1)/2 is the material constant, H is the activation enthalpy, R is the gas constant and n is the power-law exponent. So that: meff ¼ m0 expðH=nRTÞ
with dð1nÞ=n
m 0 ¼ ek
ðA Þ1=n :
Taking into account the thermal dependency of density and viscosity, De becomes: De ¼ 13:04 m0 expðH=nRTÞux =ððDrc
where G is the elastic shear modulus and E is the Young modulus. E 2.5G. For typical rocks E 1011 Pa. Thus, for all plausible values of Drgd (Drgd , 10 –100 MPa), tmin/tm varies from 1000 to 10 000. Hence, gravitational instabilities must be a leading long-term process and Maxwell relaxation is of secondary importance. Time of rifting, tr, can be expressed as:
De ¼ 13:04mux =ðDrgdLÞ:
143
ð3Þ
þ ar0 DTÞgdLÞ
ð4Þ
where Drc þ ar0DT ¼ Dr, Drc is the compositional density contrast, a is the coefficient of thermal expansion, r0 is the density at reference temperature and DT is the temperature change in respect to the reference temperature. Expression 4 indicates that slow extension rates enhance the impact of the gravity-driven instabilities, whereas fast spreading rates reduce them. This clearly suggests that slow extension favours instabilities and, as a natural consequence, rift asymmetry.
Numerical model We conducted a series of numerical experiments to test the stability of the extending lithosphere as a function of the extension rate, density and thermorheological structure. In these experiments we introduced a lithospheric plate composed of crustal, mantle and asthenospheric rheological layers with brittle –ductile –elastic properties inferred from rock mechanics data (Tables 1 and 2). The model area (Fig. 2) is 500 km wide and 300 km deep (250 150 grid elements); the sub-Moho (35 km depth) lithospheric mantle has a thickness of 120 km, with thickness of the strong mechanical core of about 70 –80 km (Fig. 1a).
Table 1. Notations and physical values common for all experiments (Turcotte & Schubert 2002) Parameter s, t P u m 1˙ T hc d rm ra Drc g Cp DT a
Values and units
Definition
Pa, MPa Pa, MPa m s21, mm year21 1019 – 1025 Pa s s21 8C 7 km/35 km 120 km 3330 kg m23 3310 –3330 kg m23 0 – 20 kg m23 9.8 m s22 103 J kg21 8C21 250 8C 3 1025 8C21
Stress Pressure Velocity vector Effective viscosity Strain rate Temperature Moho depth Thickness of mantle lithosphere Reference mantle density Asthenosphere density Compositional density contrast Acceleration due to gravity Specific heat Temperature contrast Thermal expansion
144
E. BUROV
Table 2. Specific rheology and related thermal parameters. Compilation by Burov et al. (2001). r is density; Q, n, A are material-dependent parameters of ductile flow laws (Kirby & Kronenberg 1987; Kohlstedt et al. 1995). Other parameters from Turcotte & Schubert (2002). Parameter
Value
All rocks l, G Lame´ elastic constants (l ¼ G) f friction angle (Mohr –Coulomb rheology) C0 cohesion (Mohr–Coulomb rheology)
30 GPa 308 20 MPa
Specific upper and weak (quartz) lower crust properties r (upper crust) n (power-law exponent) A (power-law constant) Q (creep activation enthalpy)
2800 kg m23 2.4 6.7 1026 MPa2n s21 1.56 105 kJ mol21
Specific strong middle–lower crust properties (diabase or basalt) r n (power-law exponent) A (power-law constant) Q (creep activation enthalpy)
2980 kg m23 3.4 2 1024 MPa2n s21 2.6 105 kJ mol21
Specific mantle properties (olivine) r (lithosphere) n (power-law exponent) A (power-law constant) Q (creep activation enthalpy)
3330 kg m23 3 1 104 MPa2n s21 5.2 105 kJ mol21
Thermal model Surface temperature (0 km depth) Temperature at the bottom of thermal lithosphere Thermal conductivity of crust, k Thermal conductivity of mantle, k Thermal diffusivity of mantle, x Radiogenic heat production at surface, Hs Radiogenic heat production decay depth, hr Prerift thermotectonic age of the lithosphere, tq
0 8C 1330 8C 2.5 W m21 8C21 3.5 W m21 8C21 1026 m2 s21 9.5 10210 W kg21 10 km 150 Ma
Fig. 2. Model set-up. The model box has dimensions of 500 300 km (250 150 grid elements). The initial rheology (explicit brittle–elastic– ductile) and temperature profile is shown on the right-hand side of the model. The initial weak temperature anomaly (þ50 8C) is used to localize thinning in the middle of the model.
THE ROLE OF GRAVITATIONAL INSTABILITIES
The boundary conditions are: (1) lateral velocities applied at the box sides; (2) free surface at the top of the model; and (3) restoring hydrostatic forces proportional to the vertical displacement at the basement. As pointed out by Huismans et al. (1998), whatever is the type of rifting (active or passive), the application of kinematic boundary conditions may be justified by the restricted size of the model. These conditions represent either far-field tectonic forces or the integrated effect of large-scale basal drag owing to the regional asthenospheric upwelling. We use a ‘generalized’ initial geotherm that fits well the observed data for the Newfoundland– Galicia margin (Pe´rez-Gussinye´ et al. 2001; Manatschal 2004). This geotherm provides 550 8C at Moho depth and 1330 8C at the base of the lithosphere, and falls in the middle of the interval between 400 8C at Moho depth (cratons) and 700 8C at Moho depth (young lithospheres) (e.g. Burov & Diament 1995). The localization of rifting in the middle of the model is achieved through the introduction of a small thermal anomaly (þ50 8C) at the base of the mechanical lithosphere. This anomaly dissipates soon after the initiation of rifting and has no prolonged effect for the lateral stages of rift evolution. The FLAC-like code Par(a)ovoz v9 (derived from Poliakov et al. 1993) solves the mechanical and thermal equilibrium equations in a large strain mode with no practical limitations on the imposed rheology laws because of its explicit time-marching scheme. The implementation and application of this algorithm to basin modelling is explained in detail in the abundant FLAC literature (Cundall 1989; Poliakov et al. 1993; Buck & Poliakov 1998; Lavier et al. 2000; Burov & Poliakov 2001, 2003; LePourhiet et al. 2004). For this reason, we limit its description to basic principals. Par(a)ovoz is a large strain fully explicit timemarching Lagrangian algorithm that solves Newtonian equations of motion in the continuum mechanics approximation: ˙ divs rg ¼ 0: krul
ð5Þ
These equations are coupled with the constitutive equations of practically any kind: Ds ¼ Fðs; u; ru; . . . T . . .Þ; Dt
ð6Þ
rCp @T=@t þ urT k divðrTÞ Hf ¼ 0
body forces: r ¼ r0 ð1 aDTÞ:
t ¼ S þ sn tgf where t is the shear stress and sn is the normal stress. In addition, linear cohesion softening is used for better localization of plastic deformation, 1p, (S(1p) ¼ S0 min (0, 1 2 1p/1p0), where 1p0 is 0.01). The ductile –viscous term is represented by nonlinear power law with three sets of material parameters (Table 2) that correspond to the properties of the respective lithological layers: upper crust (quartz), middle –lower crust (quartz –diorite), mantle (olivine): meff ¼ ek
ð7Þ
assuming adiabatic temperature dependency for density and a Boussinesq approximation for related
ð8Þ
Here u, s, g, k are the respective terms for velocity, stress, acceleration due to body forces and thermal conductivity. The over-dots refer to the time derivatives. The brackets in equation 5 specify conditional use of the related term: in quasistatic mode, the inertia is dumped using inertial mass scaling (Cundall 1989). The terms t, r, Cp, T, Hr, a designate, respectively, time, density, specific heat, temperature, internal heat production and thermal expansion coefficient. The terms @/@t, Ds/Dt, F are a time derivative, an objective (Jaumann) stress time derivative and a functional, respectively. In the Lagrangian framework, the incremental displacements are added to the grid co-ordinates allowing the mesh to move and deform with the material. This enables solution of large-strain problems locally using a small-strain formulation: on each time step the solution is obtained in local co-ordinates, which are then updated in the large strain mode. Solution of equation 5 provides velocities at mesh points used for computation of element strains and of heat advection urT. These strains are used in equation 6 to calculate element stresses, and the equivalent forces are used to compute velocities for the next time step. All rheological terms are implemented explicitly. The rheology model is serial viscous – elastic – plastic (Fig. 1a, right, Table 2). The plastic term is given by explicit Mohr–Coulomb plasticity (non-associative, zero dilatency angle) with linear Navier –Coulomb criterion that implies an internal friction angle f of 308 and maximal cohesion S of 20 MPa to fit the experimental Byerlee’s law of rock failure (Byerlee 1978):
dð1nÞ=n
and with the heat transport equations
145
ðA Þ1=n expðH=nRTÞ
where edk ¼ (Invk(eij))1/2 is the effective strain rate and A * ¼ 1/2A.3(nþ1)/2 is the material constant, H is the activation enthalpy, R is the gas constant and n is the power-law exponent (Table 2).
146
E. BUROV
The elastic parameters (Table 2) correspond to commonly inferred values from Turcotte & Schubert (2002). The reference density of the mantle lithosphere is 3330 kg/m23. The density of the asthenosphere has been varied from 3310 to 3330 kg m23.
Figures 3–6 summarize the results of numerical experiments. We have conducted several sets of experiments aimed at testing the influence of the extension rate and of the density structure on the interplay between the gravitational instabilities and tectonically induced deformation. The extension rate ux has been varied from 5 to 15 mm year21 at both sides of the system. The chemical density contrast between the mantle and the asthenosphere varied from 0 to 20 kg m23. For the experiments, we imposed a bilayer crustal structure that corresponds to a granite (quartz-dominated) upper crust with density of 2800 kg m23 and a diabase lower crust with density of 2980 kg m23 (Figs 1a & 2, Tables 1 and 2). The influence of the extension rate, the rheology and the density contrast is characterized by a single non-dimensional parameter, Deborah number (De, equation 4). Yet, owing to the depth and laterally variable viscosity, the estimated Deborah number De is highly approximate. It is computed from averaged viscosity for 300 km of extension.
the previous case, mantle instabilities develop at about 2 Ma after the onset of rifting. The mantle lithosphere is broken after 4–5 Ma of rifting. The main features of this case resume to: (1) more pronounced interior basins; (2) ridge-like geometry of the oceanized lithosphere with characteristic ridge-like increase of the depth to the isotherm of 550–600 8C with distance from the centre; and (3) important asymmetry of the margins. As in the previous case, there is a sharp transition from highly thinned (2–6 km thick) continental crust and mantle lithosphere across the margin to nearly an undeformed zone on the backside of the rift flanks. This sharp transition results from mechanical coupling of the mechanical layers in the rifted zone and from replacement of the weak continental crust by strong oceanized asthenosphere. There are two major factors defining formation of the interior basins: tensional and gravitational instabilities. For example, high tensile stresses due to faster extension rate and mechanical coupling of the crustal and mantle layers below the rift flanks locally increase the plate strength and thus favour extension in a new place. At slow extension rates (,2 5 mm year21) gravitational instabilities are more efficient. Tensional instabilities are efficient at high extension rate, which appears to be the case of this experiment, compared to the previous experiment. There are, of course, some intermediate situations when the role of both mechanisms is comparable.
De ¼ 0.3– 0.75, slow spreading rate (ux ¼ 2 5 mm year21)
De ¼ 1.5 – 10, high spreading rate (ux ¼ 2 15 mm year21)
The results of the experiments at ux ¼ 5 mm year21 and Dr ¼ 20 kg m23 are shown in Figure 3. This case is characterized by development of gravitational instabilities at 1–2 Ma and wide rifting, prior to pronounced narrow rifting, occurring after 6–7 Ma. After this time the mantle lithosphere continues to be thinned by gravitational de-blobbing concurrently with the formation of a rift basin. At b ¼ 5 (time 6– 7 Ma) the rift structure becomes strongly asymmetric and small interior basins form at the backside of one (or both) of the rift flanks. The rift asymmetry appears to be controlled by the gravitational instabilities that have different growth rates on opposite sides of the rift. As mantle viscosity is exponentially dependent on temperature, this asymmetry is reinforced due to viscosity variations caused by asymmetry in the thermal field that also results from the gravitational instabilities.
Figure 5 summarizes the results of the experiments at ux ¼ 15 mm year21 for three values of Dr (0, 10 and 20 kg m23). In all cases, the continental lithosphere is already broken up at 3 Ma after the onset of the extension (coefficient of extension, b 4–5). At 5.6 Ma mantle, upper and lower crust are widely exhumed at the surface (as it is actually observed, for example, beneath the Iberia Abyssal Plain: Whitmarsh et al. 2000). In case of zero and weak (10 kg m23) density contrast (Fig. 5a, b) the interior basins are less pronounced. In the case of higher density contrast (20 kg m23, Fig. 5c), secondary interior basins develop at rifted margins starting from 5 Ma and form a characteristic structure that resembles that of the rifted continental margins such as the conjugated Iberian and Newfoundland margins (Fig. 6). As discussed above, formation of the interior basins is obviously related to either tensional and/or gravitational instabilities (for extension rates of more than 2 15 mm year21 tensional instabilities dominate) and to the local mechanical coupling between the crustal and mantle layers. This coupling results from
Experiments and results
De ¼ 0.75– 1.5, intermediate spreading rate (ux ¼ 2 10 mm year21) The series of experiments at ux ¼ 10 mm year21 and Dr ¼ 20 kg m23 are shown in Figure 4. As in
Fig. 3. Results of the experiments for slow extension rate (ux ¼ 2 5 mm year21) and density contrast between the mantle and asthenosphere Dr ¼ 20 kg m23. Left: evolution of the material field (colour code: upper crust, salad-green; middle– lower crust, yellow; mantle lithosphere, blue; asthenosphere, marine-green; synrift sediment or otherwise reworked material, purple). Centre: temperature field. Left: logarithm of stress to strain ratio (Pa s), which is equivalent to the effective viscosity for the ductile domains. Note: (1) removal of a large part of the mantle lithosphere by RT instabilities leading to change of style of rifting; (2) strong asymmetric rifting at developed stages of extension; (3) breakup and oceanization of the lithosphere that commences at 9 Ma and is preceded by coupling of the crustal and mechanical layers within the lithosphere; (4) exhumation of small amounts of lower crust and mantle. Arrows show the position of the interior basins. Inserts correspond to 2 blow-ups of the framed areas. Purple (reworked) material most often directly overlies its source material, i.e. purple above the green layer means material is reworked from the upper crust. Purple above blue means either reworked mantle or sediment derived from the upper and lower crust.
THE ROLE OF GRAVITATIONAL INSTABILITIES 147
Fig. 4. Results of the experiments for ux ¼ 2 10 mm year21 and Dr ¼ 20 kg m23. See the caption to Figure 3 for notations. Note: (1) removal of a large part of the mantle lithosphere by RT instabilities leading to change of style of rifting; (2) strong asymmetric rifting at developed stages of extension (at the oceanization stage); and (3) exhumation of continental and oceanic mantle, and punctual exhumation of lower crustal material at ocean –continent limits. In this experiment, continental breakup and oceanization commence at 6 Ma. Arrows show the position of the interior basins. See the description of Figure 3 for other details.
148 E. BUROV
Fig. 5. (a) Results of the experiments for ux ¼ 2 15 mm year21 and Dr ¼ 0 kg m23. See the caption to Figure 3 for other notations. Note: (1) mitigated RT instabilities; (2) nearly symmetric rifting until developed stages of extension (oceanization stage); (3) less expressed interior basins; and (4) exhumation of continental and oceanic mantle, and exhumation of lower crustal material at ocean– continent limits. In this experiment, continental breakup and oceanization commence at 4 –5 Ma. See the description of Figure 3 for other details.
THE ROLE OF GRAVITATIONAL INSTABILITIES 149
Fig. 5. (b) Results of the experiments for ux ¼ 2 15 mm year21 and Dr ¼ 10 kg m23. Right: topography profiles. See the caption to Figure 3 for other notations. Note: (1) RT instabilities play a remarkable role, starting from 5 Ma; (2) nearly symmetric rifting; and (3) exhumation of continental and oceanic mantle, and exhumation of lower crustal material at ocean– continent limits. In this experiment, continental breakup and oceanization commence at 4 –5 Ma. See the description of Figure 3 for other details.
150 E. BUROV
Fig. 5. (c) Results of the experiments for ux ¼ 2 15 mm year21 and Dr ¼ 20 kg m23. See the caption to Figure 3 for other notations. Note: (1) RT instabilities below the rift flanks play an important role, starting from 2 Ma; (2) unstable extension leading to the formation of boudinage-like structures and interior basins; (3) nearly symmetric rifting; (4) formation of a strong segmented core in the middle of the rift separated by interior basins from the margin; and (5) exhumation of continental and oceanic mantle, and exhumation of lower crustal material at ocean– continent limits. At final tages remnants of the continental crust and mantle are spread in a thin layer (2 –6 km thick) over a wide area, some of them being replaced with oceanic material. In this experiment, continental breakup and oceanization commence at 2 –3 Ma. Arrows show the position of the interior basins. See the description of Figure 3 for other details.
THE ROLE OF GRAVITATIONAL INSTABILITIES 151
152
E. BUROV
THE ROLE OF GRAVITATIONAL INSTABILITIES
differential thinning accompanied by expulsion (squeezing) of the ductile crust from the thinned zones. As a result, the competent cores of mantle and crustal layers join together. This mechanical coupling locally increases the integrated strength of the lithosphere by a factor of 2 (e.g. Burov & Poliakov 2001), which favours thinning outside of the ‘hardened’ coupled zone. For higher extension rate, the rift structure is more symmetric both in the case of low- and highdensity contrast (Fig. 5a, b). This confirms the idea that high extension rate reduces the impact of gravity instabilities on the rift asymmetry. In the case of Dr ¼ 20 kg m23 (Fig. 5c) the vertical amplitude of gravity instability in the mantle part becomes important and reaches 20 –50 km after only 2 Ma of rifting. The lithosphere under both rift flanks thickens, leaving a narrowing stretched zone in between, which obviously helps to switch rifting from wide mode to narrow mode. The instability continues to develop at the exponential rate leading to de-blobbing of a part of the lithospheric mantle at the depth level corresponding to the effective viscosity of about 1021 Pa s (Fig. 5c). After this moment the mantle below the rift flanks is removed. The whole system remains roughly symmetric, with similar interior basins formed on both sides of the rift. Figure 6 compares the results of the experiment for Dr ¼ 20 kg m23 with the present-day structure of the conjugated rifted margins of Flemish Cap and Galicia (Funck et al. 2003). It reveals a number of important similarities: the geometry of the extended zones, the sharp transition from the extremely stretched crust and mantle to practically underformed lithosphere (e.g. the Flemish Cap), the geometry of the exhumed mantle and that of the exhumed lower crustal zones (e.g. Whitmarsh et al. 2000), the geometry and wavelength of the interior basins (Fig. 6b) (Galicia Interior Basin: Pe´rez-Gussinye´ et al. 2003). Even though the partial melting is
153
not included in the present experiments, there is no difficulty in predicting, when examining Figures 5c, & 6a, b, the location of the potential zones of partial melting that are seemingly compatible with those interpreted from the seismic crosssections (Funck et al. 2003).
Discussion and conclusions † The experiments demonstrate the importance of the negative RT instabilities during the formation of ‘slow’ margins and continental rifts. These instabilities are highly effective in the range of 0,De , 10, or in the range of initial spreading velocities 0–2 15 mm year21. The RT instabilities are most efficient for De 1 and spreading rates 2 10 mm year21. The instabilities largely control rift symmetry and rifting mode. For example, the transfer from wide to narrow rift mode for De , 3 or ux 2 15 mm year21 may be associated with gravitational thinning and localized downwelling of the mantle lithosphere. In the absence of specific mechanisms of softening or inherited structures, rift asymmetry may result from gravitational instabilities and differential stretching for De , 1 or ux ,2 10 mm year21. As the temperature dependence of both the viscosity and the growth rate of instabilities is exponential, a minor variation in the thickness of lithospheric layers results in asymmetric rifting in case of small extension rates. Extension at rates of less than 2 5 mm year21 produces highly unstable asymmetric rifting. † Formation of the interior basins at rifted margins, such as the Galicia Interior Basin (e.g. Pe´rezGussinye´ et al. 2003), may be linked to the interplay between the tensional and gravitational instabilities and local mechanical coupling– uncoupling of the intracrustal and mantle layers for De,10 and b . 4– 5. At slow spreading
Fig. 6. (a) Comparison of the results of the experiment shown in Figure 5c (ux ¼ 15 mm year21, Dr ¼ 20 kg m23) with the present-day structure of the Flemish Cap margin (Funck et al. 2003). Note the similarities in the sharp transition from the deformed to undeformed continental lithosphere in the geometry of the extended crustal blocks including exhumed lower crustal and mantle material. Note that the actual margin has experienced major thermal subsidence and probably additional extension or other deformation after the rifting phase. Arrows show the position of the interior basins. The dashed line on the topography plot corresponds to the basement before water loading. See the description of Figure 3 for other details. (b) Comparison of the results of the experiment shown in Figure 5c (ux ¼ 15 mm year21, Dr ¼ 20 kg m23) with the actual structure of the Galicia margin that is conjugated with the Flemish Cap margin (Funck et al. 2003). Note the similarities in the geometry of the interior basins, in sharp transition from deformed thinned crust and mantle to undeformed continental lithosphere, in the geometry of the extended crustal blocks including exhumed lower crust and mantle material. Numerical resolution does not allow localized faulting to be reproduced, but comparison with Tirel et al. 2004 (based on the same numerical code) suggests that detachment faults appear in the zones of extra-thinned crust. Note that the actual margin has experienced major thermal subsidence and probably additional extension or other deformation after the rifting phase. Arrows show the position of the interior basins. The dashed line on the topography plot corresponds to the basement before water loading. See the description of Figure 3 for other details.
154
E. BUROV
rates (2 10 mm year21), gravity-driven instabilities dominate and generate interior basins for chemical density contrasts greater than 10 kg m23. At higher extension rates, interior basins are amplified by tensional instabilities, even for small compositional density contrasts. Prebreakup thinning of the crust and mantle first results in weakening of the continental lithosphere. Yet, in later stages (b . 4 –5) it leads to localized mechanical coupling of the competent crustal and mantle layers below the rift flanks, which locally increases the integrated strength of the lithosphere (e.g. Burov & Diament 1995). As a result, the interior zones (interior basins) become relatively weaker than the zones of mechanical coupling, and become a subject of more important stretching than the coupled zones. † The experiments for ‘high’ extension rate and Dr ¼ 20 kg m23 (215 mm year21) (Figs 5c & 6) demonstrate important similarities with the observations (Funck et al. 2003). In particular, the predicted geometry of sharp transition from zones of highly stretched crust (1–6 km thick) to practically undeformed crust and lithosphere below the rift flanks resembles that of the Flemish Cap and Galicia margin. The model predicts that the extension of the continental crust may occur in an extremely wide zone. In particular, after 9 Ma, b ¼ 5–15, the continental crust (Figs 5c & 6) thins from 30–40 km to only 1–2 km over a 80– 100 km-wide zone across the modelled margin. This crust is then punctually or totally replaced with the new oceanic material. The uplifted/ exhumed mantle underlies (depth of less than 2–5 km below the bottom) or directly replaces the ultra-thinned continental and oceanic material over a 70–100 km-wide continent– ocean transition zone (e.g. Fig. 6a). The wide zones of extended continental crust (stretched to a less than 5–6 km-thick layer) produced in the model are indeed observed across the Galicia margin. The model predicts important exhumation of the continental mantle and punctual exhumation of the lower crust, in between the exhumed upper crust and mantle. In the case of instable extension, this crust may periodically alternate with the mantle (e.g. serpentinized) and the upper crustal material. The predicted vertical crustal undulations resulting from gravitational and tensional instabilities may be associated with prominent detachment faulting, as was shown (using same numerical code) for domal uplift instabilities in Tirel et al. (2004). Yet, the numerical resolution of the present model is insufficient to reproduce localized faulting. However, the predicted patterns of accumulated plastic strain suggest that stretched and
undulated upper crust deforms along the detachment faults across the margin. In the model, interior basins are formed either on both sides of the rift (high extension rate of 2 15 mm year21) or only on one side (specifically for smaller extension rates ,2 10 mm year21). This process is governed by gravitational and tensional instabilities in multilayered lithosphere. † The experiments show that it is more difficult, although not impossible, to reproduce final structure of ‘slow’ rifted margins under assumption of low extension rates (,5 mm year21) during the entire duration of the phase. In particular, slow extension rates (,2 10 mm year21) are associated with unstable asymmetric rifting, whereas the conjugated slow margins such as the Flemish Cap–Galicia margin demonstrate less asymmetry than could be expected from the models, for reported extension rates of less than 5 mm year21. Consequently, it is not excluded that the extension rate could be higher at some prebreakup stages of rifting, and then slows down later to yield the reported average rate of 5 mm year21. This does not, of course, exclude the possibility that the extension rate could increase after the breakup. Indeed, most reliable (magnetic) data refer to sea-floor spreading stage after continental breakup (e.g. Srivastava et al. 2000). The extension rate at the prebreakup stage is much less constrained. One can admit that rift systems are subject to large prebreakup variations of the extension rate. This may explain the multitude of rifting styles even for comparable average spreading rates and coefficients of extension b. An alternative explanation may infer smaller density contrasts or stronger ductile rheology. Yet, in this case the formation of interior basins will be less probable. † It is necessary to stress the primary importance of obtaining better constraints on the major parameters controlling margin evolution: (1) the compositional density contrast between the mantle and asthenosphere; (2) the prerift thermo-rheological structure; and (3) the extension rate as a function of time. These parameters are difficult to derive from the available data, and their acquisition presents a new challenge for future geological and geophysical exploration of continental margins. † Of course, the results for narrow and wide rift might infer different reaction in terms of developing instability. We here have exploited only those initial geometrical and thermal configurations that represent in some way Flemish Cap–Galicia margins. The configurations, however, look to be quite generally applicable to slow extension zones.
THE ROLE OF GRAVITATIONAL INSTABILITIES The author thanks the reviewers – R. Govers, L. Moresi and M. Pe´rez-Gussinye´ – for highly constructive comments. I also thank G. Karner and the organizers of the IMEDL workshop for the invitation to participate and the following inspiring discussions. The numerical code Par(a)ovoz v.9 is a product of Paravoz v.3 (1996) by Poliakov and Podladchikov (Poliakov et al. 1993). This study has benefited from partial support by the MEBE program.
References B ASSI , G. 1995. Relative importance of strain rate and rheology for the mode of continental extension. Geophysical Journal International, 122, 195– 210. B ASSI , G. & B ONNIN , J. 1988. Rheological modeling and deformation instability of lithosphere under extension. Geophysical Journal International, 93, 485– 504. B OTT , M. H. P. 1971. Evolution of young continental margins. Tectonophysics, 11, 319–327. B UCK , W. R. 1991. Modes of continental lithospheric extension. Journal of Geophysical Research, 96, 20,161– 20,178. B UCK , W. R. & P OLIAKOV , A. N. B. 1998. Abyssal hills formed by stretching oceanic lithosphere. Nature, 392, 272–275. B UROV , E. B. & C LOETINGH , S. 1997. Erosion and rift dynamics: new thermomechanical aspects of post-rift evolution of extensional basins. Earth and Planetary Science Letters, 150, 7 –26. B UROV , E. B. & D IAMENT , M. 1995. The effective elastic thickness (Te) of continental lithosphere: What does it really mean? Journal of Geophysical Research, 100, 3895–3904. B UROV , E. B. & P OLIAKOV , A. N. B. 2001. Erosion and rheology controls on synrift and postrift evolution: Verifying old and new ideas using a fully coupled numerical model. Journal of Geophysical Research, 106, 16,461–16,481. B UROV , E. & P OLIAKOV , A. 2003. Erosional forcing on basin dynamics: new aspects of syn- and post-rift evolution. In: N IEUWLAND , D. A. (ed.) New Insights into Structural Interpretation and Modelling. Geological Society, London, Special Publications, 212, 209– 224. B YERLEE , J. D. 1978. Friction of rocks. Pure & Applied Geophysics, 116, 615–626. C HANDRASEKHAR , S. 1961. Hydrodynamic and Hydromagnetic Stability. Oxford University Press. C LOETINGH , S. A. P. L., W ORTEL , M. J. R. & V LAAR , N. J. 1982. Evolution of passive continental margins and initiation of subduction zones. Nature, 297, 139–142. C ONRAD , C. P. & M OLNAR , P. 1999. Convective instability of a boundary layer with temperature and strain-rate-dependent viscosity in terms of ‘available buoyancy’. Geophysical Journal International, 139, 51– 68. C UNDALL , P. A. 1989. Numerical experiments on localization in frictional materials. Ingenieur-Archiv, 59, 148–159. D UNBAR , J. A. & S AWYER , D. S. 1988. Continental rifting at pre-existing lithospheric weakness. Nature, 333, 450– 452. H OPPER , J. R., F UNCK , T. ET AL . 2004. Continental breakup and the onset of ultraslow seafloor off
155
Flemish Cap on the Newfoundland rifted margin. Geology, 32, 93– 96. H OUSEMAN , G. A. & M OLNAR , P. 1997. Gravitational (Rayleigh–Taylor) instability of a layer with non-linear viscosity and convergence thinning of continental lithosphere. Geophysical Journal International, 128, 125 –150. H OUSEMAN , G. A., M C K ENZIE , D. P. & M OLNAR , P. 1981. Convective instability of a thickened boundary layer and its relevance for the thermal evolution of continental convergent belts. Journal of Geophysical Research, 86, 6135–6155. H UISMANS , R. S. & B EAUMONT , C. 2007. Roles of lithospheric strain softening and heterogeneity in determining the geometry of rifts and continental margins. In: K ARNER , G. D., M ANATSCHAL , G. & P INHEIRO , L. M. (eds) Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, London, Special Publications, 282, 111–138. H UISMANS , R. S., P ODLADCHIKOV , Y. Y. & C LOETINGH , S. 1998. The transition from passive to active rifting, relative importance of asthenospheric doming and passive extension of the lithosphere. Journal of Geophysical Research, 106(B6), 11271–11292. F LETCHER , R. C. & H ALLET , B. 1983. Unstable extension of the lithosphere: A mechancical model for Basin-and-Range structure. Journal of Geophysical Research, 88, 7457–7466. F UNCK , T., H OPPER , J. R., L ARSEN , H. C., L OUDEN , K. E., T UCHOLKE , B. E. & H OLBROOK , W. S. 2003. Crustal structure of the ocean–continent transition at Flemish Cap: Seismic refraction results. Journal of Geophysical Research, 108(B11), 2531, doi:10.1029/2003JB002434. J AUPART , C. & M ARESCHAL , J. C. 1999. The thermal structure and thickness of continental roots. Lithos, 48, 93–114. K EEN , C. E. & B OUTILIER , R. R. 1995. Lithosphere – asthenosphere interactions below rifts. In: B ANDA , E. (ed.) Rifted Ocean–Continent Boundaries. Kluwer, Norwell, MA, 17–30. K IRBY , S. H. & K RONENBERG , A. K. 1987. Rheology of the lithosphere: Selected topics. Reviews of Geophysics, 25, 1219– 1244. K OHLSTEDT , D. L., E VANS , B. & M ACKWELL , S. J. 1995. Strength of the lithosphere: Constraints imposed by laboratory experiments. Journal of Geophysical Research, 100, 17587– 17602. K USZNIR , N. & K ARNER , G. 1985. Dependence of the flexural rigidity of the continental lithosphere on rheology and temperature. Nature, 316, 138–142. L AVIER , L. L., B UCK , W. R. & P OLIAKOV , A. N. B. 2000. Factors controlling normal fault offset in ideal brittle layer. Journal of Geophysical Research, 105, 23431–23442. L E P OURHIET , L., B UROV , E. & M ORETTI , I. 2004. Rifting through a stack of inhomogeneous thrusts (the dipping pie concept). Tectonics, 23, TC4005, doi:10.1029/2003TC001584. M ANATSCHAL , G. 2004. New models for evolution of magma-poor rifted margins based on the review of data and concepts from West Iberia and the Alps. International Journal of Earth Sciences, 93, 432– 466.
156
E. BUROV
M ANATSCHAL , G. & B ERNOULLI , D. 1999. Architecture and tectonic evolution of nonvolcanic margins: Present-day Galicia and ancient Adria. Tectonics, 6, 1099–1119. M ANATSCHAL , G., F ROITZHEIM , N., R UBENACH , M. & T URRIN , B. D. 2001. The role of detachment faulting in the formation of an ocean–continent transition: insights from the Iberia Abyssal Plain. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: A Comparison of Evidence from Land and Sea. Geological Society, London, Special Publications, 187, 405– 428. M C K ENZIE , D. 1978. Some remarks on the development of sedimentary basins. Earth and Planetary Science Letters, 40, 25–32. P E´ REZ -G USSINYE´ , M., R ESTON , T. J. & M ORGAN , J. P. 2001. Serpentinisation and magmatism at nonvolcanic margins – The effect of the initial lithospheric structure. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: A Comparison of Evidence From Land and Sea. Geological Society, London, Special Publications, 187, 551– 576. P E´ REZ -G USSINYE´ , M., R ANERO , C. R. & R ESTON , T. J. 2003. Mechanisms of extension at nonvolcanic margins: evidence from the Galicia interior basin, west of Iberia. Journal of Geophysical Research, 108(B5), 2245, doi:10.1029/2001JB000901. P OLIAKOV , A. N. B., P ODLADCHIKOV , YU . & T ALBOT , C. 1993. Initiation of salt diapirs with frictional overburden: numerical experiments. Tectonophysics, 228, 199– 210. R OYDEN , L. & K EEN , C. E. 1980. Rifting process and thermal evolution of the continental margin of Eastern Canada determined from subsidence curves. Earth and Planetary Science Letters, 51, 343 –361.
P OUDJOM D JOMANI , Y. H., O’R EILLY , S. Y., G RIFFIN , W. L. & M ORGAN , P. 2001. The density structure of subcontinental lithosphere: Constraints on delamination models. Earth and Planetary Science Letters, 184, 605– 621. O’R EILLY , S. Y., G RIFFIN , W. L., P OUDJOM D JOMANI , Y. & M ORGAN , P. 2001. Are lithospheres forever? Tracking changes in subcontinental lithospheric mantle through time. GSA Today, 11, 469. S ALVENSON , J. O. 1978. Variations in the geology of rift basins; a tectonic model. Conference Proceedings of Los Alamos Scientific Laboratory, 7487, 82– 86. S RIVASTAVA , S. P., S IBUET , J.-C., C ANDE , S., R OEST , W. R. & R EID , I. D. 2000. Magnetic evidence for slow seafloor spreading during the formation of the Newfoundland and Iberian margins. Earth and Planetary Science Letters, 182, 61– 76. S TACEY , F. D. 1992. Physics of the Earth. Brookfield Press, Brisbane. S TEPHENSON , R. A., N AKIBOGLU , S. M. & K ELLY , M. A. 1989. Effects of astenosphere melting, regional thermoisostasy, and sediment loading on the thermomechanical subsidence of extensional sedimentary basins. In: P RICE , R. A. (ed.) Origin and Evolution of Sedimentary Basins and their Energy and Mineral Resources. American Geophysical Union, Geophysical Monograph, 48, 17–27. T IREL , C., B RUN , J.-P. & B UROV , E. 2004. Thermomechanical modeling of extensional gneiss dome. In: Gneiss Domes and Orogeny. Geological Society of America, Special Paper 380, 67–78. T URCOTTE , D. L. & S CHUBERT , G. 2002. Geodynamics. Cambridge University Press, Cambridge. W HITMARSH , R. B., D EAN , S. M., M INSHULL , T. A. & T OMPKINS , M. 2000. Tectonic implications of exposure of lower continental crust beneath the Iberia Abyssal Plain, northeast Atlantic Ocean: Geophysical evidence. Tectonics, 19, 919–942.
A dynamic model of rifting between Galicia Bank and Flemish Cap during the opening of the North Atlantic Ocean D. L. HARRY1 & S. GRANDELL2 1
Department of Geosciences, Colorado State University, Fort Collins, CO 80523-1482, USA (e-mail:
[email protected]) 2
Department of Geology, Adams State College, Alamosa, CO 81102, USA
Abstract: A finite-element model is used to simulate Late Jurassic– Early Cretaceous rifting between the Flemish Cap and Galicia Bank continental margins. The model results show that variations in the thickness of the continental crust on these margins at wavelengths greater than about 75 km can be explained as a consequence of the interaction of two pre-existing weaknesses in the lithosphere. A weakness in the crust, attributed to structural fabrics in the Variscan front, controls the location of crustal extension during the early stages of rifting in the model. This results in formation of a broad rift basin similar to the Galicia Interior Basin. A deep-seated weakness located 110 km further west, attributed to the thick crust beneath the central Variscan Orogen, controls the location of mantle necking. Extension in this region is initially diffuse, but accelerates and becomes more focused with time. Approximately 13 million years after rifting begins, the locus of crustal extension shifts from the region of pre-weakened crust into the region of pre-weakened mantle. This marks the end of subsidence in the Galicia Interior Basin and the onset of subsidence in the Flemish Cap and Galicia Bank marginal basins. Extension in these areas continues for another 12 million years before continental breakup. The asthenosphere does not ascend to depths shallow enough for decompression melting to begin until less than 5 million years before the onset of sea-floor spreading. The model predicts that all late-stage synrift magmatism during this period is limited to within 45 km of the rift axis, and production of melt thicknesses greater than 2 km is restricted to within 35 km of the rift axis. Mantle potential temperatures of 1250–1275 8C, 5 –30 8C cooler than normal, result in 3.1–4.5 km-thick oceanic crust at the time of breakup, in general agreement with the 2 –4 km thick crust observed adjacent to these margins.
Jurassic –Early Cenozoic rifting between the conjugate Iberia and Newfoundland continental margins progressed in a roughly south to north direction (Ziegler 1989; Srivastava et al. 1990). On the Iberian side, this resulted in three distinct rift margin segments. From south to north, these are the Tagus Abyssal Plain, Iberian Abyssal Plain and Galicia Bank segments. The Tagus and Iberia Abyssal Plain segments are both characterized by relatively wide regions of highly attenuated continental crust and what has been interpreted to be a broad area of exhumed subcontinental mantle (Pinheiro et al. 1992; Beslier et al. 1993; Whitmarsh & Sawyer 1993; Krawczyk et al. 1996; Minshull et al. 1998; Whitmarsh et al. 2000; Funck et al. 2003; Henning et al. 2004). The Galicia Bank segment exhibits a much more abrupt transition between relatively thick continental crust and oceanic crust, and a much narrower region of exhumed mantle (Boillot et al. 1980; Sibuet 1992; Whitmarsh et al. 1996). All segments of the Iberia margin, as well as the conjugate Newfoundland margin, are characterized by relatively little synrift magmatism prior to the onset of sea-floor spreading in comparison to most other rifted continental margins. Several numerical modelling studies have examined the
unusual structure of the rifted margin and the magmatic consequences of rifting on the Iberia Abyssal Plain and the conjugate portion of the Newfoundland margin (e.g. Bassi et al. 1993; Keen & Dehler 1993; Keen et al. 1994; Tett & Sawyer 1996; Harry & Bowling 1999; Bowling & Harry 2001; Minshull et al. 2001), but relatively little consideration has been given to the Galicia Bank –Flemish Cap segment and the differences between this segment and the more southerly parts of the margin. This paper describes the results of a finiteelement numerical simulation of Late Mesozoic and Early Cenozoic rifting on the Galicia Bank and Flemish Cap segments of the Iberia and Newfoundland conjugate continental margins. The modelling goal is to match the long-wavelength (.75 km) crustal thickness variations on the margins, the timing and duration of rifting, the magmatic history of the margins, the timing and location of regional shifts in extensional tectonism, and the regional subsidence history. This paper focuses on a conjugate transect between seismic profiles ISE-1 on the Iberian margin and SCREECH-1 on the Newfoundland margin (Fig. 1) (Funck et al. 2003; Henning et al. 2004; Hopper et al. 2004).
From: KARNER , G. D., MANATSCHAL , G. & PINHEIRO , L. M. (eds) Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, London, Special Publications, 282, 157– 172. DOI: 10.1144/SP282.8 0305-8719/07/$15.00 # The Geological Society of London 2007.
158
D. L. HARRY & S. GRANDELL
The modelling technique used in this paper is based on the STRCH95 two-dimensional finite-element modelling program, which is a descendant of the STRCH program developed by Dunbar (1988). STRCH95 is used to model extensional processes on a lithosphere scale. The program allows for complex initial conditions to describe the (a)
16 W
14 W
00
12 W
10 W
8 W
-4000
0
-4
pre-extension structure of the crust and rheology and temperature of the lithosphere. The simulations discussed in this paper focus on how such preexisting features might have controlled the pattern of extension and shifts in the loci of extension with time on the Galicia Bank–Flemish Cap margins, and how such extension is accommodated
6 W
-4000
-2
00
0
44 N
00 -20
42 N
GIB
40 N -4
00
0
LB
38 N
(b)
54 W 52 W 50 W 48 W 46 W 44 W 42 W 50 N -40
OB
00
-20
00
48 N
FPB JDB
46 N
Flemish Cap
FCG
Newfoundland WB
HB
CB
0 00 -4
-200 0
SWB
44 N
42 N Fig. 1. Location map. (a) Iberia margin. (b) Newfoundland margin. Solid lines are locations of model transects. Dots are DSDP and ODP drill sites. Shaded regions are interior rift basins: GIB, Galicia Interior Basin; LB, Lusitania Basin; FCG, Flemish Cap Graben; FPB, Flemish Pass Basin; JDB, Jeanne d’Arc Basin; WB, Whale Basin; SWD, South Whale Basin; HB, Horse Shoe Basin; OB, Orphan Basin; CB, Carson Basin. Contour interval is 1000 m.
A FINITE-ELEMENT RIFTING MODEL
without producing significant amounts of synrift magmatism. In the following sections we describe the geological and geophysical observations that constrain the numerical models discussed in this paper and the modelling targets. We then describe the modelling algorithm. Finally, we describe the results and discuss their implications within the context of Galicia Bank– Flemish Cap rifting in particular and rifting on non-volcanic margins in general.
159
(Minshull et al. 2001; Hopper et al. 2004). Absolute ages are based on the timescale of Gradstein et al. (2004). The first approximately 10– 12 million years of extension (Tithonian–Valanginian) involved the Galicia Interior Basin and Newfoundland Basin. The last approximately 13 –15 million years of extension was primarily focused seaward of the Galicia Bank and Flemish Cap.
Post-rift crustal structure
The history and structure of the Galicia Bank and Flemish Cap continental margins is described in detail elsewhere in this Special Publication. Here, we briefly summarize those elements of rift history and margin structure that contribute directly to constraining the numerical simulations presented in this paper.
Timing Extension on the Iberian Peninsula and Newfoundland occurred in several distinct phases. The first phase occurred in the Newfoundland and Galicia Interior basins during the Late Triassic –Early Jurassic periods (Fig. 1) (Tankard & Welsink 1987; Tucholke et al. 1989; Ziegler 1989; Murillas et al. 1990; Foster & Robinson 1993). This was followed by a period of tectonic quiescence lasting until Late Jurassic time. The second phase of extension began in the Late Jurassic epoch and lasted until late Valanginian time, when the locus of extension shifted seaward of the Flemish Cap and Galicia Bank (Tucholke et al. 1989; Murillas et al. 1990). This led to the third phase of extension, a period of lithospheric stretching and crustal thinning west of Galicia Bank and east of Flemish Cap in Hauterivian–Aptian time. This last phase of extension continued until continental breakup and formation of new oceanic crust between the time of sea-floor spreading anomalies M3 and M0 (Srivastava et al. 2000; Minshull et al. 2001; Funck et al. 2003; Hopper et al. 2004). In this paper we focus on the Late Jurassic – Early Cretaceous phases of extension. These phases involved an initial stage of extension in the Galicia Interior Basin, a later stage of extension seaward of the unextended continental crust on Flemish Cap and the slightly extended crust on Galicia Bank, and, finally, the onset of sea-floor spreading. The total duration of rifting (excluding the Triassic –Early Jurassic phase in the interior basins) lasted approximately 25 million years, beginning in Late Tithonian time (c. 146 Ma) and ending in the early Aptian age (c. 118 Ma)
A reconstruction of the Newfoundland and Galicia continental margins at the time of anomaly M0, near the onset of sea-floor spreading, is shown in Figure 2. Unextended portions of the continental crust on Iberia and Newfoundland are 30– 32 km thick (Mendes Victor et al. 1980; Banda 1988; Diaz et al. 1993). The prerift crystalline crust on the Iberia margin thins westward to less than 10 km-thick beneath the 100 km-wide Galicia Interior Basin (Pe´rez-Gussinye´ et al. 2003), increases to approximately 20 km beneath the Galicia Bank (Gonzalez et al. 1999; Pe´rez-Gussinye´ et al. 2003), and then thins progressively westward over a 100 km-wide region between the eastern edge of the Galicia Bank and the rift axis, where the oldest oceanic crust is emplaced. The prerift crust reaches a minimum thickness of 2– 5 km at its seaward limit (Chian et al. 1999; Gonzalez et al. 1999). On the Newfoundland margin, the crust is approximately 30 km thick beneath the relatively unextended Flemish Cap, and thins progressively eastward to approximately 1.5 km over an about 80 km-wide region between the eastern edge of the Flemish Cap and the rift axis (Funck et al. 2003). Mantle exposed in an approximately 10 km-wide region on the sea-floor west of Galicia Bank is thought to have been exhumed by latestage low-angle faulting during the last 1–2 million years of extension (Sibuet 1992; Pickup et al. 1996; Reston 1996; Fuegenschuh et al. 1998; Pe´rez-Gussinye´ et al. 2003). East
West Flemish Cap
Rift Axis
Galicia Bank
Galicia Interior Basin
Pre-rif t c rust Serpentinized Mantle
–100
0
Syn- and Post-rift
200
0 10 20 30
Mantle
100
Iberia
Depth (km)
Model constraints
300
Distance from riftaxis (km)
Fig. 2. Cross-section of the Newfoundland–Iberia margins at the time of anomaly M0 along the trend of seismic transects SCREECH-1 on the Newfoundland margin (Funck et al. 2003; Hopper et al. 2004) and ISE-1 on the Galicia Bank margin (Henning et al. 2004). After Funck et al. (2003).
160
D. L. HARRY & S. GRANDELL
Extension rate Extension rates on the Newfoundland and Iberia margins vary locally at the scale of a few tens of kilometres, probably reflecting different timing between movements on individual fault systems. In a regional sense, extension rates are best constrained on the Iberia Abyssal Plain. Here, regional extension rates are estimated to be approximately 10 mm year21 during rifting (Minshull et al. 2001) and 10 –14 mm year21 at the onset of seafloor spreading (Whitmarsh et al. 2001a; Russell & Whitmarsh 2003). Assuming symmetric spreading, this corresponds to a whole-extension rate of 20 mm year21 during rifting. This is consistent with area balancing of the cross-section in Figure 2. Restoration of the cross-section to a uniform crust thickness of 32 km requires an extension rate of 20 mm year21 to produce the observed amount of stretching in 25 million years.
Prerift lithospheric structure, rheology and thermal constraints Breakup between the Galicia Bank and Newfoundland occurred within a region roughly coincident with the trend of the Late Palaeozoic Variscan Orogen in this region (Capdevila & Mougenot 1988; Ziegler 1989). This is similar to the association of Mesozoic and Cenozoic rifting with the trends of other Palaeozoic orogens, including opening of the Gulf of Mexico along the Ouachita orogenic trend, opening of the North Atlantic along the Appalachian trend and opening of the Norwegian–Greenland Sea along the Caledonide trend (Fig. 3) (Ziegler 1989; Harry & Londono 2004). Focusing of extensional deformation within the Variscan Orogen was most probably a result of orogenic weakening of the lithosphere caused by the increased thickness of the crust beneath the orogen (Fig. 4). The crust in the central Variscan Orogen near the rift axis is estimated to have been approximately 34 –35 km thick prior to extension (Pe´rez-Gussinye´ et al. 2003), 2–5 km thicker than the crust in the unextended parts of the Iberia Peninsula and Newfoundland. Assuming the thermal parameters estimated by Tejero & Ruiz (2002) for the crust in the Duero Basin in the Iberian Peninsula and a simple two-layer rheological model following empirical ductile flow laws for quartz-diorite in the crust and wet dunite in the mantle (Table 1), this difference in crustal thickness would weaken the lithosphere in the central Variscan Orogen by up to 15% in comparison to the surrounding regions (Fig. 4). The strength of the nominal lithosphere predicted by this rheological model is 7.0 1012 N m21 (Fig. 4a), which is moderately weaker than the 8 1012 N m21 lithospheric strength
estimated by Tejero & Ruiz (2002) for the central Iberian Peninsula and is consistent with the modern (presumably steady-state) heat flow and thermal structure of the Iberian lithosphere estimated by Fernandez et al. (1998).
Magmatic history Both the Flemish Cap and Galicia Bank margins are flanked by unusually thin oceanic crust, ranging from 2.5 to 4 km in thickness (Whitmarsh et al. 1996; Hopper et al. 2004), suggesting a low volume of melt production immediately following continental breakup. On the Galicia margin the oceanic crust increases to normal thickness (approximately 7 km) within 15 km seaward of the oldest oceanic crust (Whitmarsh et al. 1996). At the estimated spreading half-rate of 10–14 mm year21 (Whitmarsh et al. 2001a), this suggests that the thermal regime of the mantle rapidly evolved to that of a typical mid-ocean ridge system within approximately 1–1.5 million years after the onset of sea-floor spreading. The melt production history after the onset of sea-floor spreading on the Newfoundland margin is more complicated. Here, the oldest oceanic crust is about 3–4 km thick and thins seaward to less than 1.3 km over an approximately 50 km-wide region (Hopper et al. 2004), indicating a magma-starved environment during the first approximately 3.5–5 million years of sea-floor spreading. Near-normal oceanic crust (.6 km thick) located 3 km seaward of the thinnest oceanic crust on the Flemish Cap margin (Funck et al. 2003; Hopper et al. 2004) indicates that the transition from magma-starved seafloor spreading to normal sea-floor spreading occurred in about 0.2–0.3 million years. Recovery of late synrift volcanic and plutonic rocks landward of the oldest oceanic crust on both the Newfoundland and Iberian margins suggests that the last stages of rifting were accompanied by minor amounts of magmatism, just prior to the onset of sea-floor spreading (Manatschal & Bernoulli 1999; Whitmarsh et al. 2001b; Tucholke et al. 2004). Regardless of the complexities of late rift-stage and early spreading-phase melt production, there is broad consensus that for most of the rift history the margins were amagmatic, or at least involved very little melt production. Volcanism that occurred prior to the onset of sea-floor spreading appears to have been limited in volume and space, being confined to within about 20 km of the location where oceanic crust is first produced.
Modelling targets On the basis of the above discussion, our numerical simulations of rifting between Galicia Bank and Flemish Cap are constrained by: (1) the prerift
A FINITE-ELEMENT RIFTING MODEL
161
Fig. 3. Late Permian reconstruction of the North Atlantic continents and trends of major Late Palaeozoic orogens (after Ziegler 1989).
structure of the crust, including a presumably weak and heterogeneous crust in the Variscan Orogen embedded between older strong Precambrian crust of Canada and the Iberian peninsula (Fig. 3); (2) the prerift crustal thickness, taken to be 32 km by comparison to the thickness of the modern unextended crust beneath Iberia and Canada; and (3) radiogenic heat production of the crust, taken to be similar to that of basement rocks exposed in Iberia (Table 1). The models seek to reproduce: (1) the post-rift crustal thickness variation across
the margins at length scales greater than approximately 75 km, particularly moderately extended crust beneath the Galicia Interior Basin, unextended crust beneath the Flemish Cap and slightly extended crust beneath Galicia Bank, and highly extended crust beneath the marginal rift basins; (2) the lack of magmatism during rifting; (3) the duration of rifting; (4) the amount of magmatism required to produce thin oceanic crust at the time of breakup; and (5) the subsidence history of the margin, including Late Jurassic–late Valanginian
162
D. L. HARRY & S. GRANDELL
(b)
(a)
Stress (MPa) 0
0
100
200 300 0
(d)
(c)
Stress (MPa) 100
Stress (MPa)
200 300
0
100
Temperature (°C)
200 300 0
500 1000 1500
Depth (km)
20 40 60 80 100
Net Strength 7.0 x 1012 N-m
Net Strength 6.0 x 1012 N-m
Net Strength 6.2 x 1012 N-m
32 km thick crust
35 km thick crust
Crustal weakness
120 Fig. 4. Yield strength envelopes for lithosphere based on rheological and thermal properties in Table 1 calculated at a strain rate of 10215 s21. (a) Lithosphere with nominal 32 km thick crust. (b) Lithosphere with 35 km-thick crust. (c) Lithosphere with weakened upper crust. (d) Steady-state geotherm. See text for discussion.
subsidence in the Galicia Interior Basin and late Valanginian–early Aptian subsidence seaward of Flemish Cap and Galicia Bank. The finite-element model formulation used in these simulations does not explicitly account for the formation of detachment faults, and so does not reproduce exhumation of the peridotite ridge west of Galicia Bank.
Modelling method The finite-element modelling program used in this research, STRCH95, is the most recent version of the FORTRAN 90 computer program STRCH developed by Dunbar (1988). The STRCH95 algorithm successively solves the two-dimensional heat
Table 1. Fixed model parameters* Constraint Continental crust thickness Oceanic crust thickness Lithosphere thickness Surface heat production, a0 Heat decay exponent, D Thermal conductivity, K (crust) Thermal conductivity, K (mantle) Specific heat, Cp (crust) Specific heat, Cp (mantle) Thermal expansion coefficient, a (crust) Thermal expansion coefficient, a (mantle) Extension rate Brittle yield strength, S0 Slope of brittle failure strength, B Ductile creep coefficient, A (crust) Ductile activation energy, Qc (crust) Ductile creep exponent, n (crust) Ductile creep coefficient, A (mantle) Ductile activation energy, Qc (mantle) Ductile creep exponent, n (mantle) *Symbols are defined in the text.
Prerift model value 32–35 km 3–4 km 120 km 3.1 m Wm23 12 km 2.5 Wm21 K21 3.4 Wm21 K21 875 J kg21 K21 1250 J kg21 K21 3.1 1025 K21 3.1 1025 K21 20 mm year21 60 MPa 5 MPa km21 5 10218 Pa2n s21 219 kJ mol21 2.4 4 10225 Pa2n s21 498 kJ mol21 4.5
References Pe´rez-Gusenye´ et al. (2003) Tucholke et al. (2004) Fernandez et al. (1998) Fernandez et al. (1998) Fernandez et al. (1998) Tejero & Ruiz (2002) Tejero & Ruiz (2002) Tejero & Ruiz (2002) Tejero & Ruiz (2002) Tejero & Ruiz (2002) Tejero & Ruiz (2002) Russell & Whitmarsh (2003) Bowling & Harry (2001) Bowling & Harry (2001) Bowling & Harry (2001) Bowling & Harry (2001) Bowling & Harry (2001) Bowling & Harry (2001) Bowling & Harry (2001) Bowling & Harry (2001)
A FINITE-ELEMENT RIFTING MODEL
conduction–advection –generation equation: rc
@T ¼ r KrT þ A @t
ð1Þ
and the Navier –Stokes equation for flow in a viscoplastic lithosphere r
@u ¼ rF rP þ hr 2 u @t
ð2Þ
where u is the particle velocity, T is temperature, r is the density of the rock, c is the specific heat, K is thermal conductivity, A is the volumetric heat production rate, F is body force per unit mass, P is pressure and h is viscosity. The physical properties r, c, A and K for each element in the finite-element mesh are specified by the user. Viscosity is defined by the rheology of each element, and is pressure, temperature and strain rate dependent. Viscosity is determined by empirical relations derived from experimental rock deformation studies. Initially, a uniform strain rate, 1˙, is assumed throughout the model. An effective viscosity, heff, for each element is determined from the estimated stress s and strain rate: heff ¼
s : 1_
ð3Þ
The stress in the above equation is taken to be the weaker of either a ductile (power-law creep) or plastic (a linear pressure-dependent yield criterion) rheology: 1=n 1_ eQc =nRT A
ductile:
s¼
plastic:
s ¼ S0 þ Bz
ð4Þ ð5Þ
where z is depth and R is the universal gas constant (other parameters are defined in Table 1). Once the effective viscosity is determined, the Navier– Stokes equation is solved. This results in an updated estimate of the strain rates in the model, which are used to revise the estimated effective viscosity of each element. The process is repeated until the strain rate estimates on successive iterations agree to within a convergence criterion prescribed by the user. The strain rates are then used to determine velocities at each node in the mesh, and the geometry of the model mesh is updated by stepping forward the locations of each node in the mesh a finite amount of time (time steps of 10 000 years duration are typical). The heat equation is then solved to update the thermal structure, using the new mesh geometry as a boundary condition and the thermal state of the model at the previous time
163
step as an initial condition. The new model temperatures and the strain rates from the past time step are used to calculate a revised estimate of the element viscosities to begin the next time step. Accuracy of the model solution is assessed empirically by comparing model results obtained with coarse and fine meshes and coarse and fine time step sizes. For a given family of models, a coarse mesh and time step size is initially selected. The mesh size and time step size are then halved and the model re-run. Refinement of mesh size and time step size continues until successive models produce the same result. If the model becomes highly deformed (individual elements develop large aspect ratios) a singularity develops in the linear system of equations that define the finite-element problem being solved. When this occurs, the model is remeshed. During remeshing, domain boundaries that define different materials used in the model (e.g. the model edges and the contacts between different rock types) are preserved. Within each domain, new elements are defined that have an aspect ratio as close to unity as possible. Temperatures, pressures and strain rate information within the model are preserved during the remeshing process. An important limitation to the STRCH95 program is the way in which the asthenosphere is treated. STRCH95 is a lithosphere-scale modelling program. The program does not explicitly consider convective deformation and heat transport in the asthenosphere. Rather, the asthenosphere is treated as a boundary condition on the lithosphere. For the temperature condition, we consider constant heat flux, constant temperature or adiabatic temperatures at the base of the lithosphere. The mechanical boundary condition is prescribed by Airy’s theory of isostasy, essentially treating the asthenosphere as an inviscid fluid. In terms of deformation, this is unlikely to have a major impact on model behaviour at the scales considered here as the mantle at these depths is much weaker than most of the lithosphere. However, this simple treatment of the asthenosphere may have a significant impact on melt production predicted by the model. In particular, conductive cooling of the upwelling asthenosphere beneath the rift axis (Pedersen & Ro 1992; Bown & White 1995), which is not considered here, may diminish melt production during rifting in comparison to the model predictions. Consequently, the mantle potential temperatures required in these models to match the Iberia –Grand Banks magmatic history probably underestimate the steady-state potential temperature of the asthenosphere during rifting. We therefore consider the asthenosphere temperature in these models to be a lower bound on the true mantle potential temperature. We note, however, that this
164
D. L. HARRY & S. GRANDELL
is a complex problem. Small-scale convection in the asthenosphere, also not considered in these models, may have the opposite effect, enhancing melt production during rifting and therefore leading to an overestimate of mantle potential temperatures in the models (e.g. Mutter et al. 1988; Boutillier & Keen 1999). The meshing algorithm in STRCH95 is not designed to handle highly deformed elements that result from mantle convection, so we are unable to fully address the feedback between asthenosphere thermal evolution and lithosphere deformation.
Model description A basic assumption behind the modelling strategy used here is that pre-existing strength heterogeneities in the lithosphere determine the location and distribution of loci of extension during rifting (e.g. Dunbar & Sawyer 1989). Weaknesses in the upper and middle crust (crustal weaknesses) tend to produce short-lived extensional provinces during the early stages of rifting, whereas deep-seated weaknesses (mantle weaknesses) control the location where lithospheric necking develops and, hence, the location of eventual continental breakup. If these weaknesses are offset from one another, interior rift basins (controlled by the crustal weakness) often form landward of the deep offshore marginal basin (controlled by the mantle weakness) (e.g. Harry & Sawyer 1992). The models presented here include both of these types of weakness in the lithosphere (Figs 4 & 5). Mantle weaknesses are simulated as regions of thickened crust, which is attributed to the root
beneath the Variscan Orogen prior to rifting. This is represented with a simple triangular-shaped region of increased crustal thickness. Key variables describing the mantle weakness are the width of the region of thickened crust and the amount that the crust is thickened. A crustal weakness is simulated by a region in which both the plastic and ductile yield strengths are decreased relative to the surrounding crust. In the plastic regime, this involves a reduction of S0 (equation 5). For ductile behaviour, both the temperature-dependent terms (Qc/n) and temperature-independent terms (A 21/n) are reduced by similar amounts (equation 4). The crustal weakness may be attributed to lithological variations in the upper and middle crust resulting from accretion of various exotic terranes during the Variscan Orogeny, structural fabrics associated with Variscan sutures between these terranes or to pre-existing faulting near the Variscan front (Martinez Catalan et al. 1997; von Raumer et al. 2003; Simancas et al. 2003). Key variables describing this weakness in the model are the amount by which the yield strength is decreased, the width and thickness of the weakened crust, and the position of the crustal weakness relative to the mantle weakness. All of these variables were iteratively adjusted to produce a model that best fits the geometry and rift history of the Flemish Cap and Galicia Bank margins. Other modelling parameters, including heat production, specific heat, thermal conductivity, extension rate and nominal thickness of the crust (outside the Variscan Orogen), were held fixed (Table 1). The criterion used to determine when rifting ends in the models is the requirement
T = 0°C
crust
Crustal weakness
Ux
Ux
Mantle weakness
mantle T = Tm
Isostatic Buoyancy Pressure Fig. 5. Finite-element model construction. Numbers at the top and sides indicate constant temperature and constant extension rate boundary conditions.
A FINITE-ELEMENT RIFTING MODEL
that the crust thin to approximately 2 km at the rift axis, which is similar to the thinnest continental crust on the Galicia Bank and Flemish Cap margins (Chian et al. 1999; Gonzalez et al. 1999; Funck et al. 2003).
Results Results of a typical model are shown in Figures 6 –8. The model simulates in general the present structure of the Galicia Bank–Flemish Cap conjugate margins. In particular, the model predicts extreme crustal thinning beneath conjugate offshore rift basins that form during the late stages of rifting, formation of an interior basin on the eastern shelf during the early stages of rifting, and a region of less extended crust between the offshore and interior rift basins (Fig. 7). Most of the models examined in this study produced these basic attributes. Exceptions were models in which either the crustal or mantle weakness was very small (typically involving a reduction of yield strength in the crustal weakness of less than 5% or a change in crustal thickness in the mantle weakness of less than 2 km). In those models (not shown here), the dominant weakness controls rifting from the outset, resulting in formation of a nearly symmetric rift basin centred on the dominant weakness. All other models that were examined, involving changes in yield strength in the crustal weakness ranging from 5 to 30% and changes in crustal thickness in the mantle weakness ranging from 2 to 5 km, produced results that were generally similar to that shown in Figures 6– 8. However, changes in the relative locations and magnitudes of the two weaknesses produces a wide variation in the widths and depths of the interior rift basin and deep offshore basins, the duration of extension in the interior basin, the time elapsed before breakup, the amount and pattern of crustal thinning, and the presence or absence of a region of relatively unextended crust between the deep offshore and interior rift basins. As discussed below, changes in these parameters also have a significant effect on the magmatic history of the model. The model shown in Figures 6–8 is based on the variables that provide the best fit to observations on the Galicia Bank–Flemish Cap conjugate margins (Table 2). During the first 10 million years, extension results in rapid thinning of the crust in the region encompassing the crustal weakness (Figs 6 & 7a). Crustal thinning also occurs above the mantle weakness, but at a much slower rate. Tectonic subsidence in the region of weakened crust begins almost immediately after the onset of extension, marking the first stages of formation of the interior rift basin (Fig. 7b). The region above the mantle weakness, where the deep offshore basin ultimately forms, does not subside below sea level until approximately
165
5 million years later. The two basins are separated by a structural high that undergoes moderate crustal thinning (from 32 to c. 20 km) during the first 5 million years of extension, and only a minor amount of thinning (c. 3 km) afterwards. This pattern of crustal thinning and subsidence can be understood in terms of the distribution of extension, illustrated by the model strain rates (Fig. 8). Extension in the crust is broadly distributed across the region encompassing the crust and mantle weaknesses during the first 10 million years, but is concentrated most strongly in the area of the crustal weakness (Fig. 8). Strain in the mantle is more evenly distributed across the preweakened region, and is coupled to deformation in the middle and upper crust by narrow regions of relatively high strain. Offsets in the initially vertical element boundaries show that the high strain region in the lower crust behaves as a subhorizontal detachment during this time (Fig. 6). Because extension in the mantle is initially distributed over a wide region, asthenospheric upwelling is broad and the rate of lithospheric thinning is relatively slow. As extension progresses, strain in the mantle begins to focus strongly beneath the mantle weakness. After 10 million years this deformation begins to propagate upwards, creating two regions of focused extension in the upper crust (the interior rift basin and marginal basin) that are separated by a region of relatively slow strain forming the structural high between the two basins. At this time, extension and subsidence is waning in the interior rift basin and accelerating in the marginal basin. By 15 million years extension throughout the lithosphere begins to focus exclusively in the offshore region (Fig. 8). Crustal thinning and subsidence ceases in the interior rift basin and on the structural high (Fig. 7). Because extension is more focused, lithospheric necking accelerates (Figs 6 & 8). The lithosphere beneath the rift axis thins to approximately half its initial thickness 20 million years after the onset of extension, and proceeds rapidly to breakup in the next 5 million years.
Discussion The thickness of the crust in the model generally agrees with the long wavelength (.75 km) crustal thickness variations on the Flemish Cap and Galicia Bank margins along the SCREECH-1 and ISE-1 seismic profiles (Figs 2 & 7a). The duration of the main extensional episode in the interior rift basin is approximately 13 million years, in general agreement with the late Tithonian–Valanginian phase of rifting in the Galicia Interior Basin (Tucholke et al. 1989; Murillas et al. 1990). The model predicts that deep water depths (.c. 1 km) did not become established in the offshore rift basin seaward of Galicia Bank until approximately 12 million years after the onset
166
D. L. HARRY & S. GRANDELL
Novia Scotia 0 Ma
Variscan orogen
Iberia
Weak upper crust Crust
0 32 km
Mantle
122 km 5 Ma
10 Ma
15 Ma
20 Ma
25 Ma
V.E.5:1
500 km
Fig. 6. Finite-element mesh for the model that best fits Galicia Bank– Flemish Cap rifting. Numbers on the left indicate time (million years) elapsed since onset of extension.
A FINITE-ELEMENT RIFTING MODEL
Crust Thickness (km)
(a)
40 35 30 25 20 15 10 5 0 –200 –100
0 Ma 15
10
5
20 25
0
100
200
300
400
5 –200 –100 0 100 200 300 Distance from rift axis (km)
400
Elevation (km)
(b) –1
0
5
0 10 15
1 2
20
3 25 Ma
4
Fig. 7. (a) Crustal thickness variation in the finite-element model. Bold line indicates observed thickness in the cross-section of Figure 2. (b) Elevation predicted by the finite-element model (does not include syn- and post-rift sediment loading).
of extension. Breakup is predicted in early Aptian time (c. 118 Ma), 25 million years after the onset of extension, in agreement with the estimated onset of sea-floor spreading at around anomaly M0 time (Srivastava et al. 2000; Minshull et al. 2001; Funck et al. 2003; Hopper et al. 2004). The magmatic consequences of rifting in the model are assessed in terms of the timing, volume and spatial distribution of melt produced by adiabatic decompression melting in the upwelling asthenosphere (McKenzie & Bickle 1988). As discussed previously, we neglect conductive heat loss from the rising asthenosphere, which may be significant in slow spreading systems such as the
167
Newfoundland –Iberia rift (Pedersen & Ro 1992; Bown & White 1995; Whitmarsh et al. 2001a, b), so our estimate of melt production should be considered to be an upper bound. Key parameters in the melt calculation are the shape of the ascending asthenospheric diapir, which is dictated by the pattern of lithospheric thinning in the model, and the mantle potential temperature, which is a variable in the model. A series of models were developed using mantle potential temperatures ranging from 1250 to 1300 8C as a basal boundary condition (Fig. 5). Other parameters in these models were identical to those in Tables 1 and 2. Each of these models produced patterns of extension and subsidence very similar to that shown in Figures 6–8. The only significant differences were the predicted melt production history. Maximum melt thicknesses range from 3.1 to 6.5 km at the rift axis at the time of breakup (Fig. 9). The models using a 1250 and 1275 8C mantle potential temperature produced a melt thickness of 3.1 and 4.7 km at the time of breakup, respectively. This is in good agreement with the 3 –4 km-thick oceanic crust adjacent to the Newfoundland and Galicia Bank margins (Whitmarsh et al. 1996; Hopper et al. 2004), suggesting a mantle potential temperature during rifting somewhere between these two values. As synrift conductive cooling of the asthenosphere was neglected, this should be considered a minimum estimate of the asthenospheric potential temperature prior to and during rifting. It seems unlikely, then, that the mantle potential temperature was more than about 30 8C cooler than the 1280 8C global average calculated by McKenzie & Bickle (1988). In any case, the presence of normal 7 km-thick oceanic crust within about 15 km of the oldest oceanic crust on the Galicia Bank margin (Whitmarsh et al. 1996) requires that mantle potential temperatures reached approximately 1280 8C within about 1–1.5 million years after breakup, assuming a 10–14 mm year21 sea-floor spreading half-rate (Whitmarsh et al. 2001a; Russell & Whitmarsh 2003). In the 1250 and 1275 8C potential temperature models, melt production does not begin until 2.5
Table 2. Variable model parameters Parameters Mantle weakness width (km) Mantle weakness amplitude (km) Crustal weakness width (km) Crustal weakness thickness (edges) (km) Crustal weakness thickness (centre) (km) Strength reduction in crustal weakness (%) Distance between centres of weaknesses (km) Temperature at bottom of lithosphere (8C)
Best-fit model value 120 3 80 12 26 25 110 1250, 1275, 1300
168
D. L. HARRY & S. GRANDELL
Fig. 8. Second invariant of the strain rate in the finite-element model. Strain rates range from 10215 s21 (dark blue) to 10210 s21 (red).
Melt Thickness (km)
A FINITE-ELEMENT RIFTING MODEL
8 1300 °C, 25 Ma
6
1275 °C, 25 Ma
4
1250 °C, 25 Ma 2
1300 °C, 25 Ma
0 –200 –100
0
100
200
300
400
Distance From Rift Axis (km) Fig. 9. Melt production predicted by models using mantle potential temperatures of 1300, 1275 and 1250 8C.
or 5 million years before the onset of sea-floor spreading, respectively. No melt is predicted further than 45 km landward of the rift axis in either model, and melt thicknesses in excess of 2 km is restricted to within 35 km of the rift axis. Minor late-stage synrift magmatic episodes have been documented further south on the Iberian margin (Manatschal & Bernoulli 1999; Whitmarsh et al. 2001b) and Newfoundland margin (Tucholke et al. 2004), and similar evidence in the form of small mid-ocean ridge basalt (MORB)-like volcanic edifices and gabbroic intrusions emplaced on and within exhumed subcontinental mantle has been reported on the Adrian fossil non-volcanic rifted margin exposed in the Swiss Alps (Manatschal & Bernoulli 1999; Muntener et al. 2000; Muntener & Piccardo 2003). We surmise that a minor amount of late-stage magmatism, within the range of volumes predicted by the models here, is likely to be typical of non-volcanic rifted margins such as Newfoundland–Iberia. If the mantle potential temperature at the onset of sea-floor spreading was about 1250 8C, as the model implies, what are the implications for the pre- and synrift thermal state of the asthenosphere? We consider the possibility that a layer of moderately cool mantle (c. 1250 8C potential temperature) separated the lithosphere from normal asthenosphere (1280 8C potential temperature) at the onset of sea-floor spreading. As mentioned above, this is at the low end of mantle potential temperatures required to account for the thickness of the oldest oceanic crust. How thick would this abnormally cool mantle be? The melt predictions in Figure 9 are obtained by integrating the melt fraction over the thickness of the melt column. A mantle with potential temperature of 1250 8C begins to melt when it reaches a depth of approximately 40 km according to the parameterization given by McKenzie & Bickle (1988). Hence, at the onset of sea-floor spreading, a 40 km-thick layer of relatively cool mantle would account for
169
the 2–4 km thickness of the oldest oceanic crust. Extrapolating the amount of lithosphere thinning at the rift axis at this time (90% compared to the prerift thickness) to the sublithosphere mantle suggests an original thickness of approximately 400 km for this cool mantle layer. Although this simple analysis neglects the complexity of deep mantle flow during rifting, it does provide a firstorder test of the possibility that a moderately cool layer on the scale of the asthenosphere is consistent with the melt production and lithosphere thinning history of the margin. This is generally consistent with the hypothesis of Reston & Morgan (2004), who appeal to a cool prerift asthenosphere to explain the development of non-volcanic rift margins (although we note that our model invokes a much warmer lithosphere than they advocate). A problem with the scenario described above arises when we consider the time elapsed (1–1.5 million years) between the onset of sea-floor spreading and production of approximately 7 km-thick oceanic crust. The thinning rate of the model lithosphere beneath the rift axis after 25 million years of extension is 2.5 km Ma21. If the thinning rate of the cool layer in the asthenosphere is similar, then only 2.5– 3.75 km of cool sublithospheric mantle could be removed between the onset of sea-floor spreading and production of 7 km-thick oceanic crust. Reston & Morgan (2004) recognized this problem, and advocated along-strike movement of warm asthenosphere at the end of the rifting episode to account for the transition between amagmatic rifting and production of approximately 7 km-thick oceanic crust. An alternative is that the potential temperature throughout the asthenosphere in this area was close to the global average prior to rifting, and the moderate cooling needed to explain the low melt production at the onset of sea-floor spreading is a result of synrift conductive cooling within the upwelling asthenosphere diapir (e.g. Pedersen & Ro 1992; Bown & White 1995). In such a scenario, asthenosphere temperatures in the rift neck at the onset of sea-floor spreading would be quite low at the top and gradationally increase to about 1280 8C potential temperature at depth, in such a way that the total melt production would be similar to that of a moderately cool (1250 8C) asthenosphere. Because the cool mantle would be restricted to within the rift neck and concentrated primarily at shallow depths, this would allow for a rapid transition to normal mantle temperatures and mid-ocean ridge melt production volumes soon after sea-floor spreading began. This scenario is generally consistent with the spatial and temporal distribution of magma production observed on the margins, and does not require adhoc supposition of a pre-existing thin layer of anomalously low temperature in the asthenosphere.
170
D. L. HARRY & S. GRANDELL
Summary Finite-element models simulating rifting between Galicia Bank and Flemish Cap invoke pre-existing weaknesses in the lithosphere to account for the shift in the locus of extension from the Galicia Interior Basin during the first half of the rifting episode to the deep offshore rift basin during the second half of the rifting episode. Two forms of weaknesses were examined: an upper mantle weakness, created by excess crustal thickening over a 120 km-wide region that is attributed to the central part of the prerift Variscan Orogen; and a crustal weakness created by reducing the yield strength of the upper and middle crust in an 80 km-wide region that is attributed to pre-existing structural fabrics in the eastern Variscan front. The model results were found to be robust over excess crust thicknesses ranging from 2–5 km and upper and middle crust weakening ranging from 5 to 30%. Under these circumstances, all models produced an early interior rift basin followed by a shift in extensional deformation to a deep offshore basin where continental breakup ultimately occurs. Details of the width and depth of the basins, the time at which deformation shifts from the interior rift basin to the offshore basin, and the pattern of crustal thickness at the time of breakup are determined by the relative positions, shape and magnitudes of the weaknesses imposed in the model. Model parameters in Tables 1 and 2 produced the general features of the last 25 million years of rifting on the Flemish Cap and Galicia Bank continental margins, including 13 million years of extension in the Galicia Interior Basin, 12 million years of extension in the Galicia Bank and Newfoundland Cap marginal basins, and formation of moderately extended crust on Galicia Bank. The model geometry at the time of breakup approximates structural features of greater than about 75 km wavelength on these margins. The non-volcanic nature of rifting, ultimate production of thin (2–4 km thick) oceanic crust and rapid transition to generation of normal thickness (7 km) oceanic crust within 1–1.5 million years after breakup requires a mantle potential temperature of 1250– 1275 8C, roughly 5–30 8C cooler than the global average at mid-ocean ridges. This modest amount of cooling is attributed to conductive cooling of the ascending asthenosphere during the late stages of rifting. At these mantle temperatures, the model predicts production of 3.1–4.5 km of new oceanic crust at the time of breakup. Synrift magmatism is limited to the last 5 million years or less of extension and to within 45 km of the locus of continental breakup. Melt thicknesses of greater than 2 km are restricted to within 35 km of the locus of continental breakup.
This research was supported by National Science Foundation grant OPP 0408475. We thank the organizers of the InterMargins Modelling of Extensional Deformation of the Lithosphere workshop held in Pontresina, Switzerland in July 2004 for providing a venue for discussions that led to this modelling study, and the many participants of the workshop who provided data used to constrain the models. S. Grandell received support from the Ronald E. McNair Postbaccalaurate Achievement Program for this research. The authors thank R. Buck, D. Mu¨ller and F. Tsikalas for their helpful reviews.
References B ANDA , E. 1988. Crustal parameters in the Iberian Peninsula. Physics of the Earth and Planetary Interiors, 51, 222–225. B ASSI , G., K EEN , C. E. & P OTTER , P. 1993. Contrasting styles of rifting: Models and examples from the eastern Canadian margin. Tectonics, 12, 639– 655. B ESLIER , M.-O., A SK , M. & B OILLOT , G. 1993. Ocean– continent boiundary in the Iberia abyssal plain from multichannel seismic data. Tectonophysics, 218, 383–393. B OILLOT , G., G RIMAUD , S., M AUFFRET , A., M OUGENOT , D., K ORNPROBST , J., M ERGOIL D ANIEL , J. & K ORENT , G. 1980. Ocean– continent boundary off the Iberian margin: A serpentinite diaper west of the Galicia Bank. Earth and Planetary Science Letters, 48, 23–34. B OUTILIER , R. R. & K EEN , C. E. 1999. Small-scale convection and divergent plate boundaries. Journal of Geophysical Research, 104, 7389–7403. B OWLING , J. C. & H ARRY , D. L. 2001. Geodynamic models of continental extension and the formation of non-volcanic rifted continental margins. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: A Comparison of Evidence From Land and Sea. Geological Society, London, Special Publications, 187, 511–536. B OWN , J. W. & W HITE , R. S. 1995. Effect of finite extension rate on melt generation at rifted continental margins. Journal of Geophysical Research, 100, 18,011–18,030. C APDEVILA , R. & M OUGENOT , D. 1988. Pre-Mesozoic basement of the western Iberian continental margin and its place in the Variscan belt. Proceedings of the Ocean Drilling Program, Initial Reports, 103. Ocean Drilling Program, College Station, TX, 3– 12. M ARTINEZ C ATALAN , J. R., A RENAS , R., G ARCIA , F. D. & A BATI , J. 1997. Variscan accretionary complex of northwest Iberia: Terrane correlation and succession of tectonothermal events. Geology, 25, 1103–1106. C HIAN , D., L OUDEN , K. E., M INSHULL , T. A. & W HITMARSH , R. B. 1999. Deep structure of the Ocean– continent transition in the southern Iberia Abyssal Plain from seismic refraction profiles: Ocean Drilling Program (Legs 149 and 173) transect. Journal of Geophysical Research, 104, 7443–7462. D IAZ , C. J., G ALLART , J. ET AL . 1993. A deep seismic sounding investigation of lithospheric heterogeneity
A FINITE-ELEMENT RIFTING MODEL and anisotropy beneath the Iberian peninsula. Tectonophysics, 221, 35–51. D UNBAR , J. A. 1988. Kinematics and dynamics of continental breakup. PhD thesis, University of Texas, Austin, TX. D UNBAR , J. A. & S AWYER , D. S. 1989. How pre-existing weaknesses control the style of continental breakup. Journal of Geophysical Research, 94, 7278–7292. F ERNANDEZ , M., M ARZAN , I., C ORREIA , A. & R AMALHO , E. 1998. Heat flow, heat production, and lithospheric thermal regime in the Iberian Peninsula. Tectonophysics, 291, 29– 53. F OSTER , D. G. & R OBINSON , A. G. 1993. Geological history of the Flemish Pass basin, offshore Newfoundland. AAPG Bulletin, 77, 588– 609. F UEGENSCHUH , B., F ROTZHEIM , N. & B OILLOT , G. 1998. Cooling history of granulite samples from the Ocean– continent transition of the Galicia margin: Implications for rifting. Terra Nova, 10, 96–100. F UNCK , T., H OPPER , J. R., L ARSEN , H. C., L OUDEN , K. E., T UCHOLKE , B. E. & H OLBROOK , W. S. 2003. Crustal structure of the Ocean– continent transition at Flemish Cap: Seismic refraction results. Journal of Geophysical Research, 108, 2531, doi:10.1029/2003JB002434. G ONZALEZ , A., C ORDOBA , D. & V ALES , D. 1999. Seismic crustal structure of Galicia continental margin, NW Iberian Peninsula. Geophysical Research Letters, 26, 1061–1064. G RADSTEIN , F. M., O GG , J. G. & S MITH , A. G. 2004. A Geologic Time Scale. Cambridge University Press, Cambridge. H ARRY , D. L. & B OWLING , J. C. 1999. Inhibiting magmatism on nonvolcanic rifted margins. Geology, 27, 895–898. H ARRY , D. L. & L ONDONO , J. 2004. Structure and evolution of the central Gulf of Mexico continental margin and coastal plain, southeast United States. Geological Society of America Bulletin, 116, 188– 199. H ARRY , D. L. & S AWYER , D. S. 1992. A dynamic model of extension in the Baltimore Canyon trough region. Tectonics, 11, 420–436. H ENNING , A. T., S AWYER , D. S. & T EMPLETON , D. C. 2004. Exhumed upper mantle within the Ocean– continent transition on the northern West Iberia margin: Evidence from prestack depth migration and total tectonic subsidence analyses. Journal of Geophysical Research, 109, B05103, doi:10.1029/2003JB 002526. H OPPER , J. R., F UNCK , T. ET AL . 2004. Continental break up and the onset of ultraslow sea floor spreading off Flemish Cap on the Newfoundland rifted margin. Geology, 32, 93–96. K EEN , C. E., C OURTNEY , R. C., D EHLER , S. A. & W ILLIAMSON , M.-C. 1994. Decompression melting at rifted margins: Comparison of model predictions with the distribution of igneous rocks on the eastern Canadian margin. Earth and Planetary Science Letters, 121, 403–416. K EEN , C. E. & D EHLER , S. A. 1993. Stretching and subsidence: Rifting of conjugate margins in the north Atlantic region. Tectonics, 12, 1209– 1229. K RAWCZYK , C. M., R ESTON , T. J., B ESLIER , M.-O. & B OILLOT , G. 1996. Evidence for detachment tectonics on the Iberia Abyssal Plain rifted margin. In:
171
W HITMARSH , R. B., S AWYER , D. S., K LAUS , A. & M ASSON , D. G. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 149. Ocean Drilling Program, College Station, TX, 603– 615. M ANATSCHAL , G. & B ERNOULLI , D. 1999. Architecture and tectonic evolution of nonvolcanic margins: Present-day Galicia and ancient Adria. Tectonics, 18, 1099– 1119. M C K ENZIE , D. & B ICKLE , M. J. 1988. The volume and composition of melt generated by extension of the lithosphere. Journal of Petrology, 29, 625–679. M ENDES V ICTOR , L. A., H IRN , A. & V EINANTE , J. L. 1980. A seismic section across the Tagus Valley, Portugal: possible evolution of the crust. Annales Geophysicae, 36, 469 –476. M INSHULL , T. A., D EAN , S. M., W HITMARSH , R. B., R USSELL , S. M., L OUDEN , K. E. & C HIAN , D. 1998. Deep structure in the vicinity of the Ocean– continent transition zone under the southern Iberia Abyssal Plain. Geology, 26, 743– 746. M INSHULL , T. A., D EAN , S. M., W HITE , R. S. & W HITMARSH , R. B. 2001. Anomalous melt production after continental breakup in the southern Iberia Abyssal Plain. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: A Comparison of Evidence From Land and Sea. Geological Society, London, Special Publications, 187, 537– 550. M UNTENER , O., H ERMANN , J. & T ROMMSDORFF , V. 2000. Cooling history and exhumation of lower-crustal granulite and upper mantle (Malenco, eastern Central Alps). Journal of Petrology, 41, 175–200. M UNTENER , O. & P ICCARDO , G. B. 2003. Melt migration in ophiolitic peridotites: The message from Alpine–Apennine peridotites and implications for embryonic ocean basins. In: D ILEK , Y. & R OBINSON , P. T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 68– 89. M URILLAS , J., M OUGENOT , D., B OILLOT , G., C OMAS , M. C., B ANDA , E. & M AUFFRET , A. 1990. Structure and evolution of the Galicia Interior Basin (Atlantic western Iberian continental margin). Tectonophysics, 184, 297–319. M UTTER , J. C., B UCK , W. R. & Z EHNDER , C. M. 1988. Convective partial melting, 1, A model for the formation of thick basaltic sequences during the initiation of spreading. Journal of Geophysical Research, 93, 1031– 1948. P EDERSEN , T. & R O , H. E. 1992. Finite duration extension and decompression melting. Earth and Planetary Science Letters, 113, 15–22. P E´ REZ -G USSINYE´ , M., R ANERO , C. R., R ESTON , T. J. & S AWYER , D. 2003. Mechanisms of extension at nonvolcanic margins: Evidence from the Galicia interior basin, west of Iberia. Journal of Geophysical Research, 108, 2245, doi:10.1029/2001JB000901. P ICKUP , S. L. B., W HITMARSH , R. B., F OWLER , C. M. R. & R ESTON , T. J. 1996. Insight into the nature of the ocean–continent transition off West Iberia from a deep multichannel seismic reflection profile. Geology, 24, 1079–1082. P INHEIRO , L., W HITMARSH , R. & M ILES , P. 1992. The ocean–continent boundary off the western continental
172
D. L. HARRY & S. GRANDELL
margin of Iberia–II. Crustal structure in the Tagus abyssal plain. Geophysical Journal International, 109, 106–124. R ESTON , T. J. 1996. The S reflector off Galicia: The seismic signature of a detachment fault. Geophysical Journal International, 127, 230– 244. R ESTON , T. J. & M ORGAN , J. P. 2004. Continental geotherm and the evolution of rifted margins. Geology, 32, 133–136. R USSELL , S. M. & W HITMARSH , R. B. 2003. Magmatism at the west Iberia non-volcanic rifted continental margin: evidence from analyses of magnetic anomalies. Geophysical Journal International, 154, 1–25. S IBUET , 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. S IMANCAS , J. F., C ARBONELL , R. ET AL . 2003. Crustal structure of the transpressional Variscan orogen of SW Iberia: SW Iberia deep seismic reflection profile (IBERSEIS). Tectonics, 22, 1062, doi: 10.1029/2002TC001479. S RIVASTAVA , S. P., R OEST , W. R., K OVACS , L. C., O AKEY , G., L EVESQUE , S., V ERHOEF , J. & M ACNAB , R. 1990. Motion of Iberia since the Late Jurassic: Results from detailed aeromagnetic measurements in the Newfoundland basin. Tectonophysics, 184, 229–260. S RIVASTAVA , S. P., S IBUET , J.-C., C ANDE , S., R OEST , W. R. & R EID , I. D. 2000. Magnetic evidence for slow sea-floor spreading during the formation of the Newfoundland and Iberian margins. Earth and Planetary Science Letters, 182, 61– 76. T ANKARD , A. J. & W ELSINK , H. J. 1987. Extensional tectonics and stratigraphy of Hibernia oil field, Grand Banks, Newfoundland. AAPG Bulletin, 71, 1210–1232. T EJERO , R. & R UIZ , J. 2002. Thermal and mechanical structure of the central Iberian Peninsula lithosphere. Tectonophysics, 350, 49–62. T ETT , D. L. & S AWYER , D. S. 1996. Dynamic models of multiphase continental rifting and their implications for the Newfoundland and Iberia conjugate margins. In: W HITMARSH , R. B., S AWYER , D. S., K LAUS , A. & M ASSON , D. G. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 149. College Station, TX: Ocean Drilling Program, College Station, TX, 635–648. T UCHOLKE , B. E., A USTIN , J. A. & U CHUPI , E. 1989. Crustal structure and rift-drift evolution of the Newfoundland Basin. In: T ANKARD , A. J. &
B ALKWILL , H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins. AAPG Memoirs, 46, 247–263. T UCHOLKE , B. E., S IBUET , J.-C. & K LAUS , A. & SHIPBOARD SCIENCE PARTY . 2004. Proceedings of the Ocean Drilling Program Initial Reports 210 (online). Available from World Wide Web: http://www-odp. tamu.edu/publications/210_IR/210ir.htm. VON R AUMER , J., S TAMPFLI , G. & B USSY , F. 2003. Gondwana-derived micro continents – the constituents of the Variscan and Apline collisional orogens. Tectonophysics, 365, 7– 22. W HITMARSH , R. B. & S AWYER , D. S. 1993. The ocean/ continent transition beneath the Iberia abyssal plain and continental-rifting to seafloor-spreading processes. In: WHITMARSH , R. B., SAWYER , D. S., KLAUS , A. & MASSON , D. G. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 149. Ocean Drilling Program, College Station, TX, 713–733. W HITMARSH , R. B., W HITE , R. S., H ORSEFIELD , S. J., S IBUET , J.-C., R ECQ , M. & L OUVEL , V. 1996. The ocean–continent boundary off the western continental margin of Iberia: Crustal structure west of Galicia Bank. Journal of Geophysical Research, 101, 28,291–28,314. W HITMARSH , R. B., D EAN , S. M. & M INSHULL , T. A. 2000. Tectonic implications of exposure of lower continental crust beneath the Iberia Abyssal Plain, Northeast Atlantic Ocean: Geophysical evidence. Tectonics, 19, 919–942. W HITMARSH , R. B., M ANATSCHAL , G. & M INSHULL , T. A. 2001a. Evolution of magma-poor continental margins from rifting to sea floor spreading. Nature, 413, 150– 154. W HITMARSH , R. B., M INSHULL , T. A., R USSELL , S. M., D EAN , S. M., L OUDEN , K. E. & C HIAN , D. 2001b. The role of syn-rift magmatism in the rift-to-drift evolution of the West Iberia continental margin: geophysical observations. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: A Comparison of Evidence From Land and Sea. Geological Society, London, Special Publications, 187, 107–204. Z IEGLER , P. A. 1989. Evolution of the North Atlantic – An overview. In: T ANKARD , A. J. & B ALKWILL , H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins, AAPG Memoirs, American Association of Petroleum Geologists, Tulsa, Oklahoma, 46, 111– 129.
A kinematic modelling approach to lithosphere deformation and basin formation: application to the Black Sea S. S. EGAN1 & D. J. MEREDITH2 1
Earth Sciences and Geography, School of Physical and Geographical Sciences, Keele University, Keele, Staffordshire ST5 5BG, UK (e-mail:
[email protected]) 2
Grid Technology Group, Daresbury Laboratory, Warrington, Cheshire WA4 4AD, UK (e-mail:
[email protected])
Abstract: A kinematic model of lithosphere deformation has been developed that integrates the following components: structural deformation of the crust and mantle lithosphere; thermal conditioning, perturbation and subsequent re-equilibration of the lithosphere temperature field; flexural isostatic adjustments; and surface processes, including both lateral and temporal variations in basin fill and bathymetry. This approach enables the forward modelling of extensional basin evolution in two and three dimensions followed by deformation due to subsequent extensional and compressional (i.e. inversion) events. The model has been applied to the Black Sea, which is one of the deepest basins in the world and yet it is poorly understood in terms of the mechanisms that have controlled its evolution. Although it is widely accepted that this basin was initiated by Mesozoic back-arc extension related to the subduction of the Tethys Plate to the south, most of the subsidence observed today occurred within the Palaeogene and Neogene (i.e. within the framework of the Alpine–Himalayan orogenic belt). The modelling approach described above has been used to test possible geological and geodynamic mechanisms that have controlled the subsidence history of the Black Sea. In particular, the investigation has focused on trying to explain the anomalously thick post-rift subsidence that occurred in the basin. Models assuming uniform lithosphere extension do not generate the observed thickness of sediment infill in the basin. Similarly, modelling of the compressional deformation around the edges of the basin structure does little to explain the large magnitude of subsidence within the centre of the basin. Model results show that the observed basin depths can be attained only when the total magnitude of deformation is constrained from crustal thickness changes rather than by fault displacement measurements.
Kinematic models define the dimensions of all components of a system as well as a set of relationships that describe how the parts in the system interact and evolve. This paper is concerned with the kinematic modelling of the geological and geodynamic processes that occur during the extension and compression of the continental lithosphere, but placing emphasis on how this deformation controls uplift and subsidence at the surface. The main advantage of the kinematic modelling of lithosphere-scale deformation is that it can be used to investigate the effects of processes that are poorly constrained by geophysical and field data, and to provide a better understanding of the structural, thermal and stratigraphic development of deep frontier basins where well control is poor. The first kinematic models of lithosphere extension and basin formation (e.g. McKenzie 1978; Royden & Keen 1980) combined the crustal thinning, thermal and local isostatic processes associated with the instantaneous stretching of the lithosphere to quantify surface subsidence or, sometimes, uplift. Although these models are now
considered relatively simplistic, they are based on concepts that underpin much of the lithosphere-scale basin modelling being carried out today. One of the main aims of this paper is to describe the individual processes that play a key role in deformation at lithosphere scale and how these processes have been integrated within a twodimensional (2D) and 3D kinematic modelling environment. This approach enables the forward modelling of extensional basin evolution followed by subsequent extensional or compressional deformation. The modelling approach has been applied to provide insights into the evolution of a specific case study in the geological record, namely the Black Sea. The evolution of the Black Sea region represents an interference of tectonic events over geological time in that most of the subsidence (up to 14 km) took place within the basin when the immediate surrounding regions were experiencing compressional deformation. Few modelling studies have focused on the evolution of the Black Sea such that the mechanisms that have controlled
From: KARNER , G. D., MANATSCHAL , G. & PINHEIRO , L. M. (eds) Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, London, Special Publications, 282, 173– 198. DOI: 10.1144/SP282.9 0305-8719/07/$15.00 # The Geological Society of London 2007.
174
S. S. EGAN & D. J. MEREDITH
the subsidence history of the basin remains poorly constrained (Brunet & Cloetingh 2003). The main aim of this case study is to both test and apply 2D and 3D kinematic models of lithosphere extension to obtain a better understanding of the geological and geodynamic processes that have contributed to the subsidence history of the Black Sea basin. Models have been generated to investigate the evolution of the basin in the context of uniform lithosphere extension, compressional deformation and loading at the edges of the basin, and depthdependent lithosphere extension in the lower crust and mantle lithosphere. The modelling has also been applied to the Iberian margin as part of a benchmarking exercise for the comparison of different modelling approaches carried out as part of the 2004 InterMARGINS Extensional Deformation of the Lithosphere (IMEDL) workshop (see the Appendix).
Lithosphere extension and basin formation: historical background The deformational response of a rheologically layered lithosphere to applied tensional and compressional forces is highly variable. A wide range of basin-scale phenomena have been attributed to the varying mechanisms of lithosphere-scale deformation, such as the difference in width between extending regions (e.g. Buck 1991) and the generation of asymmetric rift basins (Coward 1986). Despite the complexity of lithosphere-scale deformation, crustal thickness changes are generally modelled by either simple shear (i.e. faulting and ‘brittle’ deformation: e.g. Wernicke 1985) or pure shear (i.e. stretching or squashing: e.g. McKenzie 1978) or by a combination of the two (e.g. Kusznir et al. 1987). The pure shear and simple shear mechanisms represent end-member lithosphere deformation processes and were used as a foundation for the early models of basin formation. One of the first numerical models to link deformation of the continental lithosphere to extensional basin formation, using a kinematic approach was the McKenzie or uniform lithosphere extension model (McKenzie 1978). This model, which is shown diagrammatically in Figure 1a, assumes that the whole of the lithosphere deforms by a pure shear or stretching mechanism quantified by the b or extension factor, equivalent to 1 plus the strain. The McKenzie model is essentially 1D in that it generates results in the form of subsidence curves (Fig. 1b), which show the rift and post-rift phases that characterize many extensional sedimentary basins. The rift phase is caused by the combined effects of crustal thinning, rift-induced heating owing to raising of the lithosphere–
asthenosphere boundary and isostatic compensation. It is followed by the post-rift or thermal subsidence phase primarily driven by the gradual cooling of the lithosphere back to an equilibrated state. The McKenzie model and various derivatives, such as the non-uniform lithosphere extension model (Royden & Keen 1980; Hellinger & Sclater 1983), have been applied to many extensional basins in the geological record (e.g. Barton & Wood 1984) and can explain successfully the subsidence history of those basins that have experienced a relatively uncomplicated evolution of rifting followed by gradual post-rift/thermal subsidence. There was another significant period of advancement in lithosphere-scale modelling during the mid –late 1980s, which was mainly driven by new insights into crustal and upper mantle structure being gained from the acquisition and interpretation of deep seismic data. For example, deep seismic reflection data acquisitioned by the BIRPS group from the United Kingdom (e.g. Brewer & Smythe 1984; BIRPS & ECORS 1986) reveals the important role played by crustal faults during basin formation. The importance of faulting in lithosphere-scale deformation was further emphasised in a qualitative model presented by Wernicke (1985). This model, which is mainly based on observations made from the Basin & Range province in the western United States, assumes lithospheric deformation to be controlled by brittle failure and faulting (i.e. simple shear) in the form of a large-scale shear zone that propagates through the entire lithosphere with a dip at a low angle. The model, therefore, effectively divides the lithosphere into an upper hanging-wall plate and a lower footwall plate. As a result of simple shear, upper crustal (thin-skinned) deformation is spatially offset from zones that have experienced significant lower crust and mantle thinning. Consequently, the model predicts that a basin should be characterized by absent or minimal thermal subsidence in the region that has experienced significant upper crustal deformation. The major weakness of this model is that it fails to account for basins that have post-rift thermal subsidence sequences overlying the regions of major fault controlled subsidence. Deep seismic reflection data (e.g. Snyder & Hobbs 1999), along with studies of lithosphere deformation (Wernicke 1985), lithosphere rheology (e.g. Kusznir & Park 1987) and seismological investigations (Jackson & McKenzie 1983), have been very important in revealing the large-scale layering and structure of the continental lithosphere within various tectonic regimes. In addition, these studies suggest that lithosphere extension occurs as a hybrid of fault-controlled deformation within the brittle layer of the crust and pure shear extension within the lower crust and mantle lithosphere that
KINEMATIC MODELLING: THE BLACK SEA
(a)
time = 0
l
time
βl crust
crust
a
mantle lithosphere
a mantle lithosphere asthenosphere
(b)
175
crust
a/β mantle lithosphere
asthenosphere
Pure Shear (stretching)
asthenosphere
0.000 –0.500
Rift - induced subsidence
Depth (km)
–1.000 –1.500 –2.000
Post - rift (thermal) subsidence
–2.500 –3.000 –3.500
Beta = 2
–4.000 0
5
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 Time (Ma)
Fig. 1. The uniform lithosphere extension model (McKenzie 1978). (a) The model assumes that the lithosphere extends by pure shear and quantifies subsidence owing to crustal thinning and uplift caused by the raising of hotter material at depth nearer to the surface, along with the associated Airy isostatic compensation. The second part of the model simulates the thermal subsidence phase of basin evolution. (b) Subsidence curve generated by the McKenzie model showing basin depth following 100% lithosphere extension (i.e. b factor ¼ 2).
dominantly exhibit ductile rheological behaviour (Coward 1986; Kusznir et al. 1987). The next section describes how these developments have been extended in improved kinematic models that account more satisfactorily for the evolution of sedimentary basins.
Integrated kinematic modelling of lithosphere extension and compression This section describes the major processes that occur during lithosphere extension and compression along with some of the approaches that are used to numerically model them. These processes are then integrated into a kinematic model of continental lithosphere deformation.
Mechanical thinning or thickening of the crust owing to movement along crustal faults Extensional movement along crustal faults generates the foundations of basin structure. It also causes crustal thinning and, therefore, negative loading of the lithosphere, which responds by
flexural isostatic rebound. In contrast, reverse movement along crustal faults generates increased surface topography (i.e. crustal thickening) as well as constituting a downwards-acting load on the lithosphere, which responds by flexural isostatic subsidence. Fault configuration within the model is defined in terms of the number of faults required, and their relative spacing and orientation (i.e. synthetic or antithetic). All of the faults are assumed initially to terminate or detach at mid –lower crustal levels in order to maintain compatibility with evidence from deep seismic reflection data (e.g. Hall et al. 1984; BIRPS & ECORS 1986), seismological investigations (e.g. Jackson & McKenzie 1983) and rheological modelling of the lithosphere (e.g. Kusznir & Park 1987). There are several methods, mainly based on geometric constructions, that can be used to simulate deformation caused by fault movement. One the most popular of these constructions is the vertical shear or Chevron construction (Verrall 1982). It assumes that each vertical thickness of hanging wall is displaced laterally by the same amount of extension, usually defined in terms of a heave value. The next assumption is that any unsupported
176
S. S. EGAN & D. J. MEREDITH
part of the hanging wall following extension deforms such that it collapses directly downwards onto the underlying fault and footwall, which are assumed to be rigid. Geometric methods such as the Chevron construction and derivatives, like the inclined shear method (White et al. 1986), are very suitable for numerical modelling applications because they give reasonably accurate results for lithosphere-scale deformation and are computationally efficient.
Mechanical thinning or thickening of both crust and mantle lithosphere by a regionally distributed pure shear mechanism Although the deformation caused by fault movement plays an important role during lithosphere deformation, seismic data and rheological studies indicate that the lithosphere is mostly too ductile below depths of about 15– 20 km to support major fault structures. Instead, the lower crust and parts of the mantle are thought to deform by movement along a network of shear zones that split and merge over relatively short distances (e.g. Reston 1988). One approach that can be used to model this type of regionally distributed deformation is pure shear, stretching under extension and squashing under compression (Kusznir et al. 1987). Both the lateral position and the width of the pure shear deformation can be defined in the model. The overall deformation by both faulting and pure shear is initially area balanced in the profile in order to avoid mass balance problems.
Disturbances caused to the lithosphere temperature field Lithosphere extension and compression causes an overall heating and cooling, respectively. For example, reverse movement along crustal faults causes the emplacement of relatively hot hangingwall material onto that of a cool footwall, creating a cooling of the geotherm at shallow depth. Compressional deformation by pure shear thickens the lithosphere and also reduces the geothermal gradient. Within the model the lithosphere temperature field is represented as a 2D or 3D grid. Each node on the grid is assigned a predeformational temperature. The temperature of each grid node is then modified according to the amount of deformation experienced.
Re-equilibration of the temperature field after deformation The lithosphere temperature field experiences gradual thermal recovery after a major tectonic
event. For example, in the context of extension the lithosphere cools back to an equilibrated state. The model calculates both lateral and vertical heat flux by the mechanism of conduction. The physical properties of the lithosphere are such that the thermal recovery process takes of the order of 100 million years but follows an exponential decay path.
Flexural isostatic compensation of tectonic loads Crustal thinning and thickening, and thermal perturbations caused by tectonic activity, impose loads on the lithosphere that have to pass through an isostatic filter to give a compensated amount of subsidence or uplift observed at the surface. The flexural isostatic response of the lithosphere to loading is calculated within the model in order to generate a realistic surface or basement profile and underlying crustal structure. The methodology used to model flexure assumes that the lithosphere behaves as a continuous elastic plate, which is in equilibrium under the action of all applied loads (Turcotte & Schubert 2002). The flexing properties of the lithosphere are defined by its flexural rigidity (Walcott 1970; Watts 2001), which is, in turn, set in the model by the parameter effective elastic thickness (Te).
Surface processes The model contains simple algorithms to simulate basin infill and the erosion of uplifts. Isostatic adjustments are calculated in response to these processes such that sedimentary infill enhances subsidence within a basin, whereas erosion has the combined effect of reducing the size of the uplifts and unloading the lithosphere, which responds by regional isostatic uplift or rebound (Egan & Urquhart 1993). A comprehensive description of all theoretical aspects of the above model are presented in Kusznir & Egan (1989) and Egan & Urquhart (1993), and will not be repeated here. Figure 2, however, illustrates some of the parameters included in the model calculations and examples of results. A typical starting condition for the modelling is shown in Figure 2a, which illustrates a regional crosssection of undeformed lithosphere. The crustal component of this lithosphere is assumed to be 35 km thick with a density of 2800 kg m23, while the mantle lithosphere is assumed to be 90 km thick with a density of 3300 kg m23. The modelled lithosphere is thermally conditioned with an equilibrated geotherm, which has a surface temperature of 0 8C and a temperature at the lithosphere–asthenosphere boundary of 1333 8C. All of these parameters can
KINEMATIC MODELLING: THE BLACK SEA
177
Fig. 2. An integrated kinematic model of lithosphere deformation. (a) Cross-section of undeformed lithosphere showing initial values of parameters used in the model calculations. This model can be applied to simulate; (b) lithosphere extension and associated basin formation; (c) lithosphere shortening and associated thrust-belt and foreland basin formation, or a combination of the two.
be varied according to the tectonic scenario to be modelled. The effects of extending this lithosphere are shown in Figure 2b. The model shows how the crust has accommodated extensional deformation by movement along a sequence of faults, while the mantle lithosphere has been deformed by regionally distributed pure shear. The boundary
between upper crustal deformation and that in the lower crust and mantle lithosphere is determined by a detachment depth, which is equivalent to the concept of necking depth presented by other authors (e.g. Kooi et al. 1992). The model shows a basement profile consisting of a sequence of closely spaced half-grabens with relative uplift of the footwall owing to flexural isostatic processes
178
S. S. EGAN & D. J. MEREDITH
(Weissel & Karner 1989; Egan 1992). Extension has also caused heating of the lithosphere, which subsequently has cooled to generate gradual subsidence. The effects of this can be seen in the model by a post-rift stratigraphic sequence that blankets the underlying fault blocks and synrift sequence. Lithosphere shortening is represented by the model in Figure 2c, which shows a ‘piggy-back’ style of thrusting adjacent to a foreland basin that has been generated mainly by flexure in response to crustal thickening caused by compressional tectonics. In addition, the model is sufficiently versatile to include multiple extensional and compressional events.
Application of integrated kinematic modelling of lithosphere extension and compression to investigate possible subsidence mechanisms within the Black Sea basin The modelling approach described in the previous section has been applied to the Black Sea, which is a deep, semi-isolated marine basin located north of Turkey and south of Ukraine and Russia. The main aim of the work described in this section has been to use kinematic modelling to try and decipher the role of the main subsidence mechanisms within this poorly understood basin.
An overview of the regional structure and stratigraphy of the Black Sea basin The Black Sea basin covers an area of approximately 423 000 km2, has a present maximum bathymetry of 2200 m and comprises the western and eastern Black Sea sub-basins, which are separated by the Mid-Black Sea high (Fig. 3). The origins of the basin go back to a phase of back-arc extension during the mid-late Cretaceous when northwards subduction was occurring beneath modern-day Turkey representing the closure of Tethys Ocean (Hsu¨ 1977; Letouzey et al. 1977; Zonenshain & Le Pichon 1986; Go¨rur 1988; Okay et al. 1994). Extension continued throughout the upper Cretaceous, perhaps leading to continental separation and the formation of oceanic crust. However, most of the 12–14 km of subsidence experienced by the central parts of the basin took place when the surrounding regions were experiencing compressional deformation. This is demonstrated by Alpine– Himalayan thrust belts, including the Pontides, Balkanides, Crimean Mountains and Greater Caucasus (Fig. 3), that surround the Black Sea.
Regional cross-sections are presented (see Fig. 3 for location) to illustrate the structure and stratigraphy within the western and eastern Black Sea basins. The section presented in Figure 4a shows structure and stratigraphy from offshore Turkey into the centre of the western Black Sea. It is a section constructed from the depth-converted interpretation of a regional seismic line acquired by BP and TPAO. Interpreted horizons have been constrained by data from a well (Akc¸akoc¸a-1) located in the southern part of the section, which penetrated down into rocks of Cretaceous age at a depth of about 2000 m. The original seismic data are confidential, but a detailed description of its interpretation and analysis can be found in Robb (1998). The cross-section is dominated by major extensional faults, generated by an Albian– Aptian rift phase, which has produced half-graben structures separated by areas of relative footwall uplift. Go¨ru¨r (1988) relates this rift phase to the opening of the western Black Sea due to back-arc extension overlying a northerly dipping subduction zone. However, the magnitude of extension associated with this rift phase was not very great and amounts to about 3 km of heave on the faults observed in the section (approximately 2% overall extension). The deposition of middle–upper Cretaceous synrift sequences was followed by a broad, thin blanket of Paleocene deposits, which from its general distribution is interpreted as a post-rift thermal sag phase. The section becomes more complicated in the Eocene with the contractional reactivation of extensional faults and the development of new reverse faults at the basin margin. In contrast, the northern part of the cross-section shows that the centre of the Black Sea was subsiding very rapidly from the Palaeogene to the present, with the deposition of over 12 km of sediment. The Ukrainian Black Sea region is dominated by basin inversion as shown by the section in Figure 4b. The sequence of tectonic events that have generated the overall structure and stratigraphy exhibited in this section are, however, very similar to those already described in the context of the Turkish section. Again, there is evidence of rifting during the middle Cretaceous. The extensional evolution of this part of the basin was terminated during the Eocene by an inversion phase. Inversion continued into the Oligocene, generating uplift and erosion of the earlier syn- and post-rift sequences. A foreland basin-type sequence was generated in the Miocene as a response to the crustal thickening and flexure caused by the inversion tectonics. Finally, the Neogene sequence indicates a dramatic increase in subsidence towards the centre of the Black Sea, which has also caused the proximal parts of the shelf to subside flexurally. The deeper structure of the western Black Sea basin is a matter of controversy because of an
Fig. 3. Black Sea location map showing major tectonic elements (adapted from Robinson 1997) and location of cross-sections used for modelling study of the basin. The rectangle positioned within the eastern Black Sea shows the approximate extent of the 3D model (see Figs 10 & 11).
KINEMATIC MODELLING: THE BLACK SEA 179
Fig. 4. Sections showing structural and stratigraphic components across the western Black Sea (see Fig. 3 for location information). (a) Depth section across the Turkish part of the basin. (b) Two-way time (TWT) section across the Ukrainian part of the basin. (c) Composite TWT section across the central (western) Black Sea (adapted from Finetti et al. 1988) based on the interpretation of several regional seismic lines (vertical dashed lines indicate the boundaries between the individual seismic lines).
180 S. S. EGAN & D. J. MEREDITH
KINEMATIC MODELLING: THE BLACK SEA
absence of well data and inadequate seismic reflection data at depth. The section presented in Figure 4c illustrates the overall structural and stratigraphic style in the central part of the basin, and is based upon interpreted seismic data presented by Finetti et al. (1988). This section shows that the central western Black Sea basin is characterized by a flat-lying sedimentary sequence, which reaches depths of about 14 km. Although the seismic data do not show deep structure, Finetti et al. (1988) suggested that the basin-fill sequence has been deposited on flat-lying basement, probably basaltic in composition. Evidence from gravity and seismic refraction data (Neprochnov et al. 1970; Belousov 1988) also suggests that western Black Sea basement may consist of oceanic crust. There is, however, no direct evidence as to whether the density structure determined from these data is due to oceanic crust or to highly attenuated and/ or metamorphosed continental crust. Four regional-scale sections across the eastern Black Sea are presented in Figure 5 (see Fig. 3 for locations). These sections are derived from the interpretation of seismic reflection lines acquired by BP and TPAO that have been interpreted by BP personnel and subsequently modified (Meredith & Egan 2002). The original seismic data are confidential. Section D (Fig. 5a) shows structure and stratigraphy from onshore Turkey, into the centre of the eastern Black Sea and finally onto the Russian shelf in the NE of the region. As for the western part of the basin, the section is dominated by extensional faults. In particular, the Archangelsky and Shatsky ridges represent regionally uplifted footwall blocks to extensional faults that form the southern and northern continental slope regions, respectively. This interpretation is in agreement with Banks et al. (1997), who suggest that the Shatsky ridge defines the northern rift margin of the eastern Black Sea. Similar to the western Black Sea, fault displacement measurements indicate that the magnitude of extension was not very great and amounts to about 60 km of heave on the faults observed in section D. This equates to an overall magnitude of extension of approximately 13% (i.e. b ¼ 1.13). Thickness variations across the oldest horizons in the section suggest that rifting occurred later than in the western sub-basin, possibly some time in the Paleocene or Eocene. This timing correlates with a late Paleocene –early Eocene rift event suggested by Banks & Robinson (1997) and Spadini et al. (1997), or early –middle Eocene rifting inferred by Kazmin et al. (2000). The onshore and shelf regions of the eastern Black Sea become more complicated at the end of the Eocene with evidence of compressional deformation. Compressional tectonic features are best developed in the NE part of section D (Fig. 5a)
181
where the Tuapse fold belt, which forms the southern limit of the Great Caucasus mountain belt, shows reverse movement along basement faults. Immediately adjacent to this thrust belt is the Tuapse Trough, which is superimposed onto the footwall block forming the Shatsky ridge. The deeper structure of the central part of the eastern Black Sea basin is a matter of controversy because of the general poor quality of seismic reflection data at depth. The sections presented in Figure 5 show a similar structural and stratigraphic style to the western Black Sea in that the central part of the basin is characterized by a ‘layer-cake’ stratigraphy, which is present to depths of over 12 km.
The western Black Sea basin – application of a 2D kinematic approach The three cross-sections described in Figure 4 have been combined to produce a regional cross-section that extends from the Ukrainian to Turkish margins of the western Black Sea (Fig. 6). Possible processes that have contributed to the subsidence history exhibited by this section are explored below using the kinematic modelling approach described in the previous section on ‘Integrated kinematic modelling of lithosphere extension and compression.’ Has the western Black Sea evolved due to uniform lithosphere extension combined with loading from the surrounding thrust belts?. Figure 7 shows a model of the western Black Sea using the methodology described in the previous section. The overall dimensions of the model, along with major fault structures, are based upon the regional north –south section across the western Black Sea presented in Figure 6. In addition, deformation by faulting has been balanced to regionally distributed stretching at depth to represent uniform lithosphere extension. The model results show clearly that it is not possible to reproduce Black Sea subsidence with extensional deformation constrained by the magnitude of faulting. The magnitude of observed deformation as determined from extensional movement along fault structures amounts to 6.75 km total heave (equivalent to b of approximately 1.01) across the section. In addition, this deformation is mainly confined to the basin margins and can only account for overall subsidence in these parts of the basin. However, the model shows less than 1 km of subsidence in the central part of the basin, which is mainly produced by thinning of the lower crust, thermal subsidence (modelled for a 100 million years time period) and sediment loading in response to pure shear deformation. This amount falls far short of the 14 km of total subsidence indicated by seismic data.
182
S. S. EGAN & D. J. MEREDITH
Fig. 5. North–south sections across the eastern Black sea (see Fig. 3 for locations) showing a dominance of compressional tectonics within the Turkish and Russian onshore and near-shelf regions. Extensional tectonics dominates both northern and southern continental slope regions. The deep basin, however, shows a thick, flat-lying sedimentary sequence that is mostly devoid of structure.
KINEMATIC MODELLING: THE BLACK SEA
183
Fig. 6. Composite section (adapted from Robb 1998) across the western Black Sea based on the combination of the individual sections presented in Figure 4.
It is possible that a significant component of post-rift subsidence was caused by compressional tectonics. More precisely, the loading generated by the thrust belts surrounding the Black Sea may have caused the development of foreland basins, which constructively interfered to produce one large basin. Cobbold (1993) refers to this type of structure as a ‘push-down basin,’ while Graham (1995) calls them ‘Central Asian type basins,’ which have the characteristics of being ‘relatively symmetric, relatively fixed and commonly contain enormous thicknesses of sediment.’ The effects of loading from surrounding thrust belts have been modelled in the context of the western Black Sea. The model profile presented in Figure 7c represents the extensional evolution of the basin followed by compression at the margins occurring at a modelled time of 50 Ma to represent the onset of deformation due to the Alpine–Himalayan orogeny. The magnitude of both extensional and compressional deformation is constrained by the fault movement measured from the cross-sections in Figures 4 & 6. A lack of data prevents widening of the model to reproduce compressional deformation observed in the Pontides and Ukraine. However, it is very evident that the flexural subsidence generated in response to the formation of the thrust belts deepens the original extensional basin at its edges, rather than in the centre. The model also shows how crustal thickening has caused uplift of pre-, syn- and postrift sequences at positions 10 and 500 km along the section. These uplift structures can also be observed in the data, but they are not subaerial. In addition, the model does not generate sufficient flexural subsidence in the centre of the basin. Crustal thinning beneath the western Black Sea. Although there appears to be a lack of extensional faulting within the western Black Sea, it cannot be
discounted that there are more extensional faults than resolved by the seismic data and that the magnitude of deformation assumed within the models presented so far is an underestimation. The model results presented in Figure 8 are also based on uniform lithosphere extension. However, the magnitude of extension has been increased significantly compared to the models presented in Figure 7. This is because the magnitude of deformation has been calculated using crustal thickness rather than basement faulting. The cross-section presented in Figure 8a has been adapted from that in Figure 6 to show the relative thickness of syn- and post-rift stratigraphy. The boundary between these sequences is placed at the division between the Cretaceous and Paleocene, as shown in Figures 4 & 6. The section also shows an estimation of the depth to basement (i.e. base synrift). Although this boundary cannot be accurately determined from the data available, it is assumed to occur at the base of the Upper Cretaceous (Go¨ru¨r 1988; Spadini et al. 1996). An estimation of Moho topography is also displayed and is based on commercial gravity data and published material (e.g. Neprochnov et al. 1970, 1974; Belousov 1988). The prerift thickness of the crust in the Black Sea region is open to debate. However, Spadini et al. (1996) propose that the western Black Sea formed by the rifting of the Moesian-European platform, which has a characteristic crustal thickness of 35 km. From the section presented in Figure 8a it is possible to determine the magnitude of crustal thinning or thickening relative to a regional average and, in turn, to calculate a sequence of b values (i.e. 1 þ strain: McKenzie 1978) across the basin. This b distribution has been used to constrain the magnitude of extension and compression evident in the model presented in Figure 8b, which assumes uniform deformation of the lithosphere based entirely on a pure shear mechanism. The
Fig. 7. Model profile based on the cross-sections presented in Figures 5 & 6. (b) is a vertical exaggeration of (a) to show the details of basin structure and its infill. The model represents uniform lithosphere extension, whereby deformation by faulting is balanced to regionally distributed stretching at depth. Although it is possible to explain the overall magnitude of subsidence at the basin margins by this mechanism, model results show that it is not possible to reproduce the large amount of subsidence observed in the central part of the basin. (c) Model profile investigating the possibility that Black Sea subsidence can be explained by loading caused by adjacent thrust belts.
184 S. S. EGAN & D. J. MEREDITH
KINEMATIC MODELLING: THE BLACK SEA
185
Fig. 8. (a) Cross-section adapted from Figure 6 to show syn- and post-rift stratigraphy, top basement and Moho depth. Thinning or thickening of the crust (Co0 ), relative to a regional average of 35 km (Co), has been used to quantify the magnitude of deformation across the profile in terms of a b distribution. (b) Model based on the magnitude of deformation calculated in (a) assuming uniform lithosphere deformation. The model shows overall subsidence that is comparable with that observed in the western Black Sea. However, the relative thicknesses of syn- and post-rift sequences do not match observations.
186
S. S. EGAN & D. J. MEREDITH
model shows an overall basin depth that is almost comparable with that across the western Black Sea. Maximum subsidence predicted by the model is 13.5 km. The value of elastic thickness for the Black Sea region at the time of rifting and during post-rift evolution is open to speculation. Model calculations have assumed an effective elastic thickness (Te) of 5 km at rifting and 10 km for the post-rift phase based on a sensitivity test of the parameter in terms of it generating the most realistic combination of basin geometry and subsidence. These values are also compatible with a number of rift settings (e.g. Barton & Wood 1984; Kusznir et al. 1991; Van Wees & Cloetingh 1994). Although the model includes the effects of infilling the basin with sediment (density ¼ 2500 kg m23), a bathymetry, rising to a maximum depth of 3 km in the centre of the basin, has been maintained so as not to overestimate the loading and subsidence caused by sediment infill. The main difference between the model profile and observed data concerns the proportions of syn- and post-rift stratigraphy. The model shows a synrift sequence that reaches a thickness of 3.5 km. This value is not compatible with the data interpretation in Figure 8a, which suggests that the basin fill is mainly post-rift and that the synrift sequence is of the order of 1–2 km in thickness. Possible effects of subsurface loading. In this subsection we investigate the possibility as to whether the subsidence history of the western Black Sea basin can be explained by depthdependent extension, such that the deformation of the mantle lithosphere by stretching exceeds that in the crust due to faulting. The importance of depth-dependent stretching was identified by the numerical modelling of two-layer lithosphere stretching carried out by Royden & Keen (1980) and Hellinger & Sclater (1983). More recently, depth-dependent stretching has been used to explain the discrepancies between upper and lower lithosphere extension within continental margin settings (e.g. Roberts et al. 1997; Kusznir et al. 2004). The model profiles presented in Figure 9 have again used a faulting configuration based on the regional north– south section across the western Black Sea, but the extension has been increased in the lower crust and mantle lithosphere (i.e. below the assumed detachment depth at 20 km). The model presented in Figure 9a has assumed extension in the lower crust and mantle lithosphere to be given by a maximum b value of 8.5 (at x ¼ 300 km), which is compatible with the overall magnitude of crustal thinning. b is decreased following a sinusoidal distribution to no deformation over a lateral distance of 210 km either side. The pattern of subsidence exhibited in this model is
similar to that exhibited in the regional section. More specifically, the model shows a very thin synrift phase, analogous to the small thickness of middle –upper Cretaceous synrift deposits observed in the basin, and a large thickness of post-rift deposition, analogous to the thick Palaeogene and Neogene sequences observed in the basin. However, overall basin depth is too shallow compared to the observed data for the basin. The modelling approach used to generate Figure 9a assumes a fixed detachment depth of 20 km, whereby the upper crust above this depth is thinned by faulting as observed from the data, while the lower crust is thinned by pure shear according to the estimated thinning of the crust (see Fig. 8a). The assumption of a fixed detachment depth limits the magnitude by which the crust can be thinned overall and explains the relatively low basin subsidence generated in the model profile. The model presented in Figure 9b attempts to reconcile the observed magnitude of fault-controlled extension with the observed attenuation of the whole crust. The scenario modelled is one where the detachment or necking depth is allowed to progress towards the surface rather than being fixed at mid-crustal levels. The detachment zone may migrate throughout the evolution of a rift or during multiple rifts because the thickness of the brittle and ductile zones would constantly change during deformation in response both to the changing thickness of the crust and to temperature perturbations (Kusznir & Park 1987). In addition, Driscoll & Karner (1998) show an example of how an eastdipping detachment model of the northern Carnarvon Basin, Australia has been used to account for the observed discrepancy between the amount of upper crustal and lower lithosphere deformation. The model presented in Figure 9b shows how thin early fault-related synrift sediments are generated during an initial stage of rifting. Depth-dependent stretching then thins the crust (from its base) during a second rift stage. Maximum basin subsidence is about 13.5 km, which is comparable to that observed in the western Black Sea. Similarly, the model exhibits a very thin synrift stratigraphic thickness, large post-rift thickness and a realistic bathymetry. These results suggest that enhanced mantle extension may have played a role in the evolution of the western Black Sea.
The eastern Black Sea basin – application of a 3D kinematic modelling approach A weakness with 2D basin models is that the Cartesian co-ordinate that extends perpendicular to the plane of the section is assumed to be both constant and infinite in extent in terms of its physical
KINEMATIC MODELLING: THE BLACK SEA
187
Fig. 9. (a) Model based on depth-dependent stretching, whereby the lower crust and mantle lithosphere have been extended more than the upper crust. This model is comparable with the data in that it shows a relatively thin synrift sequence overlain by a thick post-rift sequence. (b) Model that reconciles the observed magnitude of fault-controlled extension with the observed attenuation of the whole crust. The scenario modelled is one where the detachment or necking depth is allowed to progress towards the surface rather than being fixed at mid-crustal levels.
properties. As a result, lateral discontinuity in structure and thermal perturbation outside the plane of section is ignored. This has significant implications on the flexural compensation of loads and on the true geometry of the plane of the modelled section. In three dimensions, flexural interactions modify the spatial pattern and extent of basement subsidence and/or uplift. A 3D lithosphere-scale basin model has been developed in order to investigate how regional interactions between a range of geological and geodynamic processes control basin subsidence and stratigraphy through time. This model has been applied to provide insights into the regional subsidence history of the eastern Black Sea. The methodology developed is similar to that in the 2D modelling whereby a range of parameters can be independently controlled and defined in order to simulate basin development through time whilst accounting for multiple rift phases, progressive sediment infill and reproduction of bathymetry. The following subsection describes this modelling
approach and the additional considerations required when modelling geological and geodynamic processes in three dimensions. † A set of crustal thickness profiles (e.g. Fig. 8a) are constructed initially. † For each model time step, the tectonically induced subsidence (thermally and mechanically derived) is calculated and flexurally compensated for by assuming an initial water infill. Loading owing to changes in crustal thickness, thermal effects, sediment infill and bathymetry are compensated for by using a 3D flexural algorithm modified after Hodgetts et al. (1998). † Perturbation and/or re-equilibration of the lithosphere temperature field is calculated for each model time step. † In order to simulate the observed proportion of sediment infill, the basin is then filled with sediment to a potential maximum bathymetry. The extra flexural subsidence generated from the initial sediment infill is then optionally filled
188
S. S. EGAN & D. J. MEREDITH
with sediment or water ensuring that variations in the infill density are compensated for at all times. † If necessary, this procedure is repeated using a different potential maximum bathymetry until an approximate representation of the proportion of sediment infill is obtained. Owing to the lack of data generally available on the regional palaeobathymetry, this method provides an alternative approach to filling the basin to a ‘known’ palaeobathymetry and also provides an estimate of how the basin’s sediment –water interface has evolved through time. A detailed description of the theoretical aspects of this 3D modelling approach can be found in Meredith (2003). The cross-sections presented in Figure 5 have been used to provide ‘seed-lines’ for a 3D model that covers an area of 52 000 km2 of the Turkish and central regions of the eastern Black Sea (see Fig. 3 for location). A b profile has been calculated for each of the cross-sections based upon variation in crustal thickness. b values have been calculated at 4 km intervals along each profile and, in turn, placed into a 3D spatial model as to coincide with the relative positions of each cross-section. Figure 10 shows a 3D model representation of the eastern Black Sea assuming evolution by uniform lithosphere extension of a 35 km-thick crust. The model has been simulated over a period of 65 Ma. Effective elastic thickness is assumed to be 5 km during instantaneous rifting, evolving to 10 km during the post-rift subsidence phase. The model also reproduces and regionally compensates for temporal changes in both stratigraphic thickness and bathymetry. Post-rift sediments have been subdivided into early- and late-stage sequences in order to replicate the Paleocene –Middle Miocene sequence and the anomalously thick Upper Miocene–Quaternary sequence. Total subsidence is shallower in the 3D model than is observed. Maximum basin depth is 12 km, whilst maximum depth in the model is 8.5 km. A lack of subsidence is further depicted in the model by the elevation of the Archangelsky and Andrusov Ridges, which, according to the data, have accumulated significant thicknesses of Quaternary sediment. This reduction in basin depth in the 3D model partly results from the increase in out-of-plane flexural support provided by the MidBlack Sea High. This is illustrated in Figure 10 by the increase in basement elevation when moving westerly toward the Mid-Black Sea high. Variations in prerift lithospheric configuration, including crustal thickness, can strongly control the total magnitude of subsidence produced during
and after extension. Estimation of the prerift crustal thickness remains widely open to conjecture owing to the long succession of rifting and collisional events that have preceded eastern Black Sea basin development. If the prerift crust had been thicker than the commonly assumed thickness of 35 km, the total subsidence derived from the apparent magnitude of extension would be enhanced. According to Sengo¨r et al. (1985), the prerift crust of western Turkey was thickened by early Palaeogene compression to an estimated thickness of 50 –55 km. Kazmin et al. (2000) also propose that the prerift basement of the eastern Black Sea was analogous to that in the eastern Pontides, which is known to have undergone strong Paleocene compression and thickening (Okay & Sahintu¨rk 1997). This evidence suggests that the thickness of the prerift crust may have been in the order of 40 –45 km. Consequently, models have been produced in order to investigate rifting of a thickened prerift crust. Figure 11 shows a 3D model representation of the eastern Black Sea assuming evolution by uniform lithosphere extension of a 45 km-thick prerift crust. Raising the initial crustal thickness value to 45 km produces extension across the whole model and a maximum b value of 4.2 in the central basin. Model results indicate that rifting of a thickened crust generates sufficient subsidence to house the great thicknesses of post-rift sediments including the present-day deep bathymetry. Three-dimensional reproduction of the basin depth, and syn- and post-rift stratigraphic architectures are directly comparable with that observed. In addition, model results suggest that the anomalously thin synrift sediments and thick post-rift sequences could have resulted from underfilling of the basin due to prolonged sediment starvation followed by late stage post-rift sedimentation.
Discussion There has been much speculation within the literature as to the cause of the large magnitude of subsidence within the Black Sea. In this study we have used 2D numerical modelling of lithosphere extension and compression to test some of the subsidence mechanisms proposed for the western part of the basin. Readers should refer to Meredith & Egan (2002) for the application of a similar approach to the eastern Black Sea. Model results show clearly that it is not possible to explain subsidence in the Black Sea by uniform lithosphere extension coupled with sedimentary infill, when the magnitude of extension is constrained by the observed faulting of basement. Furthermore, it is very unlikely that the lack of extensional structures evident within the central part of the basin can be entirely
Fig. 10. Application of the 3D kinematical modelling approach to the eastern Black Sea. Extension of a 35 km-thick crust simulates the main structural elements and a synand post-rift stratigraphy that has a near-comparable ratio to those observed. The basin model, however, is shallower than observed partly due to the increased flexural support produced by the Archangelsky Ridge, which is orientated obliquely to the direction of the data sections.
KINEMATIC MODELLING: THE BLACK SEA 189
Fig. 11. Three-dimensional model of the eastern Black Sea. Extension of a thicker (45 km) prerift crust maximizes synrift subsidence, and produces a syn- and post-rift stratigraphy that is directly comparable in terms of thickness and architecture to that observed in the eastern Black Sea.
190 S. S. EGAN & D. J. MEREDITH
KINEMATIC MODELLING: THE BLACK SEA
attributed to a lack of resolution within the currently available data. Walsh et al. (1991) have shown that typically 10–40% of total fault-related extension may not be observed on regional-scale seismic profiles. However, the difference between magnitude of observed fault controlled extension and overall deformation of the crust is over 200% in the context of the Black Sea. It is now widely accepted that the relative thickness of syn- and post-rift sequences within a basin can be influenced by the nature of the extensional deformation experienced by the lithosphere. More precisely, the relatively thick post-rift sequence within the Black Sea could be explained by depthdependent extension whereby the magnitude of deformation within the mid–lower crust and mantle lithosphere is much greater than that experienced by the upper crust. Modelling of this deformational regime (Fig. 9) is compatible with the subsidence history of the Black Sea in terms of the overall magnitude of subsidence as well as generating a relatively small amount of synrift subsidence followed by a much thicker post-rift sequence. It is also compatible with the suggestion that the Black Sea is floored by oceanic crust because enhanced mantle extension raises the lithosphere asthenosphere boundary to a shallow depth resulting in adiabatic melting within the mantle lithosphere and the potential to generate several kilometres of melt thickness (McKenzie & Bickle 1988). It is beyond the capabilities of the modelling used in this study simulate the formation of oceanic crust. In addition, there is little constraint from data or the published literature on how oceanic crust may have formed in the Black Sea region or, indeed, what its relationship is with the surrounding (continental) margins. However, the possible occurrence and mechanisms of generation of oceanic crust within the Black Sea requires further investigation. The observed discrepancy between the amount of upper crustal extension by fault displacements and overall crustal thinning in a variety of sedimentary basins has been identified by a number of authors (e.g. Moretti & Pinet 1987; Ziegler 1992; Driscoll & Karner 1998; Kusznir et al. 2004). Evidence of crustal stretching is often absent or indicates only less than 10–15% stretching (Artyushkov 1992). A number of problems exist with regards to the validity of the enhanced extension mechanism. In particular, space problems arise if there is exaggerated stretching of the lower crust and mantle lithosphere (e.g. Rowley & Sahagian 1986). A number of possible mechanisms may provide a solution to this apparent complication. Royden & Keen (1980) have argued that magmatic intrusion into the crust may account for some of the inherent space problems. According to Brun & Beslier (1996), mantle exhumation at passive margins and in the core complex of
191
intracontinental basins may also accommodate the space problem associated with enhanced extension of the mantle lithosphere. Watcharanantakul & Morley (2000) have also recently argued that enhanced stretching could be accomplished by subduction roll-back and slab steepening of a subducting plate and have applied this mechanism to the syn- and post-rift modelling of the Pattani Basin, Thailand. In addition, lower crustal flow has also been advocated as a mechanism for explaining the observed discrepancy between the amount of upper crustal extension and extent of crustal thinning (e.g. Bertotti et al. 2000). At geological timescales, weak ductile lower crust may behave as a ductile fluid channel able to flow from areas of high lithostatic stress to areas of lower lithostatic stress (Block & Royden 1990; Westaway 1998). However, predicting the flow behaviour of the mid and lower crust remains uncertain, and in terms of numerical modelling requires further investigation. A 3D kinematic modelling approach illustrates the importance of simulating regional-scale basin subsidence and flexure in three dimensions in order to have an appreciation of out-of-plane flexural support of loads. Application of the 3D kinematic modelling approach to the eastern Black Sea suggest that extension of a thickened (40–45 km) prerift crust is necessary in order to generate the necessary amount of subsidence. Indeed, a number of studies support the occurrence of compressional tectonics related to Tethys subduction occurring in the region before the onset of (back-arc) extension. Consequently, it is argued that extension of a continental crust 5–10 km thicker than average is a potential mechanism that may certainly account for the overall subsidence and basin depth.
Summary A kinematic model of lithosphere deformation has been developed that includes the complex interaction, in two and three dimensions, of structural, thermal, isostatic, rheological and surface processes to simulate the evolution of extensional basins and thrust belt– foreland basin couplets. Application of this kinematic modelling approach to the western Black Sea indicates that it is not possible to reproduce subsidence within the basin when the magnitude of lithosphere extension is based on the amount of fault-controlled deformation. However, modelling of the Black Sea in which the magnitude of deformation is constrained using crustal thinning/thickening generates amounts of total subsidence that are comparable with that observed. These models rely on a depth-dependent extension mechanism to reconcile the observed (small) magnitude of faulting with overall attenuation of the
192
S. S. EGAN & D. J. MEREDITH
crust. In addition, a 3D modelling approach has been applied to the eastern Black Sea and shows that the magnitude of total subsidence is significantly reduced when accounting for a realistic bathymetry, infill and regional flexure, even when the overall magnitude of deformation is constrained by the observed crustal thinning. The observed subsidence has to be accounted for by additional subsidence mechanisms, such as the extension of thickened crust. The main conclusions from the application of the modelling is that for highly extended margins and basin structures the observed basin depth can only be attained when the total magnitude of crustal thinning is constrained from crustal thickness changes rather than by fault-displacement measurements. Nevertheless, reconstruction of the seismically visible faults simulates a comparable basement topography to that observed. Consequently, while it may be regarded that upper crustal faults account for the apparent magnitude of extension at the level of the basement, processes occurring deeper in the crust and mantle may thin the crust without appreciably deforming the faulted basement topography. Processes such as lower crustal flow, penetrative pure shear, lateral flow of weak ductile crustal layers between stronger layers, and a component of subseismic scale deformation may combine to produce whole crustal thinning without significant mismatches in upper and lower crustal stretching factors. Clearly, these processes require further investigation both through modelling and observation.
Appendix: InterMARGINS Extensional Deformation of the Lithosphere (IMEDL) benchmarking exercise: application of a kinematic modelling approach to the Iberian margin Introduction In this section the 2D kinematic modelling approach described in the section on ‘integrated kinematic modeling of lithosphere extension and compression’ in the main part of the paper has been applied to the Iberian margin as part of a benchmarking exercise carried out for the 2004 InterMARGINS Extensional Deformation of the Lithosphere (IMEDL) workshop. The aim of this benchmarking exercise was to establish the strengths and weaknesses of different modelling approaches (i.e. dynamic, kinematic, inverse, etc.) when applied to the same case study consisting of a combined section across the Iberia and Newfoundland margins, including their respective ocean–continent transitions. The database made available for the exercise consisted of multichannel seismic reflection data,
refraction data, published material and Ocean Drilling Program (ODP) drilling data sets. Our main aim, as part of this benchmarking exercise, was to show how a kinematic modelling approach can be applied to provide insights into the geological and geodynamic processes that control overall subsidence history and basin architecture of a passive margin setting.
Data constraints The modelling carried out has focused on the Iberia margin and is mainly based on a regional seismic reflection line, Iberia Seismic Experiment – 1 (ISE-1), which was made available for the IMEDL benchmarking exercise. The geological evolution of the Iberian margin is described in a number of publications (e.g. Pinheiro et al. 1996; Whitmarsh et al. 1996) and will not be repeated here. Seismic line ISE-1, which was acquired in 1997, is shown in Figure A1a. It starts about 11 km to the west of the Portuguese– Spanish coastline, crossing the Galicia Interior Basin, the Galicia Bank, the deep Galicia Basin and the Peridotite Ridge. The section has been interpreted to show the sea bed, top basement and crustal faults (Fig. A1b). In order to start building the first model, the magnitude of heave has been quantified for each of the faults that have been identified (Fig. A1c). The heave along each fault has been measured as the distance from the footwall cut-off to the point of contact between basement and the fault. No attempt has been made to reconstruct the footwall tops to take into account material lost through erosion so it is acknowledged that the heave values measured are likely to be an underestimate. Total extension, as given by the sum of all heave values, is 104.5 km, which approximates to about 50% extension (b ¼ 1.5).
Application of kinematic modelling One of the main observations to arise from the modelling of the Black Sea basin in the main part of the paper is that it is not possible to use the observed magnitude of faulting to constrain deformation throughout the lithosphere. This is partly because the data acquisition and interpretation process does not reveal all of the faults that exist at a variety of scales within the crust. However, there is a dilemma in that many basins within the geological record show a mismatch between the deformation due to faulting and overall thinning of the crust that is too great to explain by a lack of identification of faults within the available data. This mismatch is apparent across the Iberian margin where the b value determined from the fault heave values is about 1.5, which leads to postextensional crustal thickness of about 23 km. In contrast, seismic refraction data (Whitmarsh et al. 1996) indicate that the crust is about 2– 4 km thick at the western edge of the section, which is indicative of b values of 8 or 9. One possible scenario that can reconcile the observed low magnitude of fault deformation with high attenuation
KINEMATIC MODELLING: THE BLACK SEA
193
Fig. A1. (a) Iberia Seismic Experiment – 1 (ISE-1) regional seismic reflection line that was made available for the 2004 IMEDL benchmarking exercise. The x-scale is defined in CDP locations, which are 12.5 m apart, giving the section a total length of about 325 km. (b) The section has been interpreted to show the sea bed, top basement and crustal faults. (c) The magnitude of heave has been measured for each of the faults that have been identified.
194
S. S. EGAN & D. J. MEREDITH
Fig. A2. Model based on reflection line ISE-1 assuming a migrating detachment depth (4 –15 km). It shows a realistic basement geometry and overall pattern of relative uplift and subsidence across the margin.
of the crust determined from geophysical data is a migrating detachment or necking depth (Braun & Beaumont 1989). This depth determines the relative importance of upper crustal thinning due to faulting relative to lower crustal thinning due to a more regionally distributed process. It is possible therefore that during the evolution of the Iberian margin there has been a change from a deep detachment depth and therefore dominance of crustal thinning due to faulting to a shallow crustal necking depth where there is a dominance of lower crustal thinning. The driving mechanism for this migration of the detachment depth can be explained by an increase in the geothermal gradient as the rifted margin evolves (e.g. Spadini et al. 1996). Figure A2 shows a model profile based on a migrating detachment depth. The deformation due to faulting is based on that interpreted from the ISE-1 section (Fig. A1b, c) and assuming a fault detachment depth of 15 km, which is compatible with evidence from deep seismic reflection data (e.g. Hall et al. 1984), seismological investigations (e.g. Jackson & McKenzie 1983) and rheological modelling of the lithosphere (e.g. Kusznir &
Park 1987). The crust is assumed to have a predeformational thickness of 35 km (Pinheiro et al. 1996; Manatschal 2004). A second phase of extension is then modelled with the necking depth positioned at a depth of 4 km, and the lower crust and mantle lithosphere is gradually thinned from no deformation in the east to a maximum B value of 8 at the western edge of the section in order to reproduce the observed attenuation of the crust. The model has been allowed to run for a simulated time of 200 million years to represent the first phase of rifting in the lower Jurassic. For simplicity, the basin has been assumed to fill to sea level with sediment during rifting, whereas the post-rift phase has been water filled to avoid overloading and overdeepening the basement. The elastic thickness of the lithosphere is taken to be 5 km at rifting, which is compatible with a number of active rift settings (e.g. Barton & Wood 1984; Kusznir et al. 1991; Van Wees & Cloetingh 1996) gradually rising to 20 km at the end of the model evolution in response to cooling of the lithosphere. The model shows distinct uplift of the footwall blocks covered by a broad blanket of thermal subsidence.
KINEMATIC MODELLING: THE BLACK SEA
195
Fig. A3. (a) Composite section across the Iberian margin showing structure and stratigraphy with respect to depth (adapted from Pinheiro et al. 1996). The location map shows the near proximity of this section to seismic line ISE-1. (b) Modified model (cf. Fig. A2) whereby any subsided part of the basin following rifting has been half-filled with sediment and allowed to subside flexurally. Any remaining subsidence is filled with water. This model shows a close comparison to the structural and stratigraphic cross-section presented in Figure A3a.
The resultant model profile exhibits both a realistic basement geometry and overall pattern of subsidence across the margin. A published section across the Iberian margin by Pinheiro et al. (1996) has been used to make a detailed comparison with the model results (Fig. A3a). This section shows the variation of structural and stratigraphic components with respect to depth and has some
overlap with seismic line ISE-1. A comparison between the western parts of this section and the model profile in Figure A2 shows that the modelled basement depth is reasonably accurate with the tops of the fault blocks at depths of 4– 5 km and basement within the graben structures at depths of 6– 8 km. However, there are mismatches between the model results and interpretation. In particular,
196
S. S. EGAN & D. J. MEREDITH
the modelled water depth is too low and the sediment thickness is too great. This incompatibility can be reduced by decreasing the amount of infill of the basin. A new model has been generated in Figure A3b using an identical set of parameters to that shown in Figure A2, except that any basin subsidence generated immediately following rifting has been half-filled with sediment, while any remaining component flexural subsidence has been filled with water. Although the modified model shows a close match to the western part of the section, the Galicia Bank region is too deep in the model. This mismatch can be explained by an overestimation of the magnitude of fault-controlled extension in the region. Alternatively, a closer match could also be obtained by modifying the flexural strength of the lithosphere. The eastern part of the model shows a reasonably accurate geometry and depth profile across the fault blocks, apart from the exaggerated footwall uplift structure at the edge of the profile. This feature is not present in the geological section possibly owing to either erosion or by additional extension off-section to the east.
Summary A kinematic model of lithosphere extension has been included within a benchmarking exercise to compare the strengths and weakness of different modelling methods in the context of the Iberia margin. Twenty four crustal faults have been interpreted in seismic section ISE-1 across the Iberia margin, exhibiting a total horizontal extension of about 105 km (b ¼ 1.5). This magnitude of fault-controlled deformation has to be reconciled with the overall attenuation of the crust to generate a realistic magnitude of subsidence, which can be explained by a gradually shallowing necking depth. We would like to thank BP and TPAO for providing data for this study. S. Egan is also grateful for the opportunity to have taken part in the 2004 IMEDL workshop. The authors would also like to thank N. Christie-Blick, A. Goodliffe and an anonymous reviewer for their constructive comments during the review process of this manuscript.
References A RTYUSHKOV , E. V. 1992. Role of crustal stretching on subsidence of the continental crust. Tectonophysics, 215, 187–207. B ANKS , C. J. & R OBINSON , A. G. 1997. Mesozoic strikeslip back-arc basins of the western Black Sea region. In: R OBINSON , A. G. (ed.) Regional and Petroleum Geology of the Black Sea and Surrounding Region. AAPG Memoir, 68, 53– 62. B ANKS , C. J., R OBINSON , A. G. & W ILLIAMS , M. P. 1997. Structure and regional tectonics of the Achara–Trialet fold belt and the adjacent Rioni and Kartli foreland basins. In: R OBINSON , A. G. (ed.) Regional and Petroleum Geology of the Black Sea and Surrounding Region. AAPG Memoir, 68, 19– 38.
B ARTON , P. & W OOD , R. 1984. Tectonic evolution of the North Sea basin: crustal stretching and subsidence. Geophysical Journal of the Royal Astronomical Society, 79, 987– 1022. B ELOUSOV , V. V. 1988. Structure and evolution of the Earth’s crust and upper mantle of the Black Sea. Bolletino di Geofisica Teorica e Applicata, 30, 109– 196. B ERTOTTI , G., P ODLADCHIKOV , Y. & D AEHLER , A. 2000. Dynamic link between the level of ductile crustal flow and style of normal faulting of brittle crust. Tectonophysics, 320, 195– 218. B IRPS & E CORS , 1986. Deep seismic profiling between England, France and Ireland. Journal of the Geological Society, London, 143, 45– 52. B LOCK , L. & R OYDEN , L. H. 1990. Core complex geometries and regional scale flow in the lower crust. Tectonics, 9, 557–567. B RAUN , J. & B EAUMONT , C. 1989. A physical explanation of the relation between flank uplifts and the breakup unconformity at rifted continental margins. Geology, 17, 760– 764. B REWER , J. A. & S MYTHE , D. K. 1984. MOIST and the continuity of crustal reflector geometry along the Caledonian– Appalachian orogen. Journal of the Geological Society, London, 141, 105 –120. B RUN , J. P. & B ESLIER , M. O. 1996. Mantle exhumation at passive margins. Earth and Planetary Science Letters, 142, 161– 173. B RUNET , M.-F. & C LOETINGH , S. 2003. Integrated PeriTethyan Basins Studies (Peri-Tethys Programme). Sedimentary Geology, 156, 1–10. B UCK , W. R. 1991. Modes of continental lithospheric extension. Journal of Geophysical Research, 96, 20,161–20,178. C OBBOLD , P. R. 1993. Sedimentary basins and crustal thickening. Sedimentary Geology, 86, 77–89. C OWARD , M. P. 1986. Heterogeneous stretching, simple shear and basin development. Earth and Planetary Science Letters, 80, 325– 336. D RISCOLL , N. W. & K ARNER , G. D. 1998. Lower crustal extension across the Northern Carnaravon basin, Australia: Evidence for an eastward dipping detachment. Journal of Geophysical Research, 103, 4975–4991. E GAN , S. S. 1992. The flexural isostatic response of the lithosphere to extensional tectonics. Tectonophysics, 202, 291– 308. E GAN , S. S. & U RQUHART , J. M. 1993. Numerical modelling of lithosphere shortening: application to the Laramide orogenic province, western USA. Tectonophysics, 221, 385 –411. F INETTI , I., B RICCHI , G., D EL B EN , A., P IPAN , M. & X UAN , Z. 1988. Geophysical study of the Black Sea. Bolletino di Geofisica Teorica ed Applicata, 30, 197–324. G O¨ RU¨ R , N. 1988. Timing of opening of the Black Sea. Tectonophysics, 147, 247–262. G RAHAM , R. H. 1995. Asian analogues for Precambrian tectonics? In: C OWARD , M. P. & R IES , A. C. (eds) Early Precambrian Processes. Geology Society, London, Special Publications, 95, 271– 289. H ALL , J., B REWER , J. A., M ATTHEWS , D. H. & W ARNER , M. R. 1984. Crustal structure across the Caledonides from the WINCH seismic reflection profile: influences on the evolution of the Midland
KINEMATIC MODELLING: THE BLACK SEA Valley of Scotland. Transactions of the Royal Society of Edinburgh: Earth Sciences, 75, 97– 109. H ELLINGER , S. J. & S CLATER , J. G. 1983. Some comments on two-layer extensional models for the evolution of sedimentary basins. Journal of Geophysical Research, 88, 8251– 8269. H ODGETTS , D. & E GAN , S. S., W ILLIAMS , G. D. 1998. Flexural modelling of continental lithosphere deformation: A comparison of 2D and 3D techniques. Tectonophysics, 294, 1 –20. H SU¨ , K. J. 1977. Tectonic evolution of the Mediterranean basins. In: N AIRN , A. E. M., K ANES , W. H. & S TEHLI , F. G. (eds) The Ocean Basins and Margins, Volume 4A The Eastern Mediterranean. Plenum, New York, 29–75. J ACKSON , J. A. & M C K ENZIE , D. P. 1983. The geometrical evolution of normal fault systems. Journal of Structural Geology, 5, 471– 482. K AZMIN , V. G., S CHREIDER , A. A. & B ULYCHEV , A. A. 2000. Early stages of evolution of the Black Sea. In: B OZKURT , E., W INCHESTER , J. A. & P IPER , J. D. A. (eds) Tectonics and Magmatism in Turkey and the Surrounding Area. Geological Society, London, Special Publications, 173, 235– 249. K OOI , H., C LOETINGH , S. & B URRUS , J. 1992. Lithospheric necking and regional isostasy at extensional basins. Journal of Geophysical Research, 97, 17,553– 17,571. K USZNIR , N. J. & E GAN , S. S. 1989. Simple-shear and pure-shear models of extensional sedimentary basin formation: application to the Jeanne d’Arc basin, Grand Banks of Newfoundland. AAPG Memoir, 46, 305–322. K USZNIR , N. J. & P ARK , G. 1987. The extensional strength of the continental lithosphere: its dependence on geothermal gradient, crustal composition and thickness. In: C OWARD , M. P., D EWEY , J. F. & H ANCOCK , P. L. (eds) Continental Extensional Tectonics. Geological Society, London, Special Publications, 28, 35– 52. K USZNIR , N. J., K ARNER , G. D. & E GAN , S. S. 1987. Geometric, thermal and isostatic consequences of detachments in continental lithsphere extension and basin formation. In: B EAUMONT , C. & T ANKARD , A. J. (eds) Sedimentary Basins and Basin Forming Mechanisms. Canadian Society of Petroleum Geologists Memoir, 12, 185–203. K USZNIR , N. J., M ARSDEN , G. & E GAN , S. S. 1991. A flexural cantilever simple-shear/pure-shear model of continental lithosphere extension: application to the Jeanne d’Arc basin, Grand Banks and Viking Graben, North Sea. In: Y IELDING , A. M. & F REEMAN , B. (eds) The Geometry of Normal Faults. Geological Society, London, Special Publications, 56, 41– 60. K USZNIR , N. J., H UNSDALE , R. & R OBERTS , A. M. 2004. Timing of depth-dependent lithosphere stretching on the S. Lofoten rifted margin offshore mid-Norway: Pre-breakup or post-breakup? Basin Research, 16, 279–296. L ETOUZEY , J., B IJU -D UVAL , B., D ORKEL , A., G ONNARD , R., K RISTCHEV , K., M ONTADERT , L. & S UNGURLU , O. 1977. The Black Sea: A marginal basin. Geophysical and geological data.
197
In: B IJU -D UVAL , B. & M ONTADERT , L. (eds) International Symposium on the Structural History of the Mediterranean Basins. Editions Technip, Paris, 363– 376. M ANATSCHAL , G. 2004. New models for evolution of magma-poor rifted margins based on a review of data and concepts from West Iberia and the Alps. International Journal of Earth Sciences, 93, 432–466. M C K ENZIE , D. P. 1978. Some remarks on the development of sedimentary basins. Earth and Planetary Science Letters, 40, 25– 32. M C K ENZIE , D. P. & B ICKLE , M. J. 1988. The volume and composition of melt generated by extension of the lithosphere. Journal of Petrology, 29, 625–679. M EREDITH , D. J. 2003. 2-D and 3-D computer modelling of lithosphere dynamics and sedimentary basin formation – Application to the eastern Black Sea. PhD thesis, Keele University. M EREDITH , D. J. & E GAN , S. S. 2002. The geological and geodynamic evolution of the eastern Black Sea basin: insights from 2-D and 3-D tectonic modelling. Tectonophysics, 350, 157– 179. M ORETTI , I. & P INET , B. 1987. Discrepancy between lower and upper crustal thinning. In: B EAUMONT , C. & T ANKARD , A. J. (eds) Sedimentary Basins and Basin Forming Mechanisms. Canadian Society of Petroleum Geologists Memoir, 12, 233–239. N EPROCHNOV , Y. P., K OSMINSKAYA , I. P. & M ALOVITSKY , Y. P. 1970. Structure of the crust and upper mantle of the Black and Caspian Seas. Tectonophysics, 10, 517– 538. N EPROCHNOV , Y. P., N EPROCHNOVA , A. F. & M IRLIN , Ye. G. 1974. Deep structure of the Black Sea basin. In: D EGENS , E. T. & R OSS , D. A. (eds) The Black Sea – Geology, Chemistry and Biology. AAPG Memoir, 20, 35–49. ¨ . 1997. Geology of the O KAY , A. I. & S AHINTU¨ RK , O eastern Pontides. In: R OBINSON , A. G. (ed.) Regional and Petroleum Geology of the Black Sea and Surrounding Region. AAPG Memoir, 68, 291–311. O KAY , A. I., S ENGOR , A. M. C. & G O¨ RU¨ R , N. 1994. Kinematic history of the opening of the Black Sea and its effect on the surrounding regions. Geology, 22, 267–270. P INHEIRO , L. M., W ILSON , R. C. L., P ENA DOS R EIS , R., W HITMARSH , R. B. & R IBEIRO , A. 1996. The western Iberia margin: A geological and geophysical overview. In: W HITMARSH , R. B., S AWYER , D. S., K LAUS , A. & M ASSON , D. G. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 149. Ocean Drilling Program, College Station, TX, 3 –23. R ESTON , T. J. 1988. Evidence for shear zones in the lower crust offshore Britain. Tectonics, 7, 929–945. R OBB , N. A. 1998. An integrated structural, sequence stratigraphic and geodynamic analysis of the western Black Sea. PhD thesis, Keele University. R OBERTS , A. M., L UNDIN , E. R. & K USZNIR , N. J. 1997. Subsidence of the Voring Basin and the influence of the Atlantic continental margin. Journal of the Geology Society, London, 154, 551–558. R OBINSON , A. G. 1997. Tectonic elements of the Black Sea region. In: R OBINSON , A. G. (ed.) Regional and Petroleum Geology of the Black Sea and Surrounding Region. AAPG Memoir, 68, 1– 6.
198
S. S. EGAN & D. J. MEREDITH
R OWLEY , D. B. & S AHAGIAN , D. 1986. Depth-dependent stretching: A different approach. Geology, 14, 32– 35. R OYDEN , L. & K EEN , C. E. 1980. Rifting processes and thermal evolution of the continental margin of eastern Canada determined from subsidence curves. Earth and Planetary Science Letters, 51, 343– 361. S¸ ENGO¨ R , A. M. C., G O¨ RU¨ R , N. & S AROG˘ LU , F. 1985. Strike-slip deformation basin formation and sedimentation. In: B IDDLE , K. T. & C HRISTIE -B LICK , N. (eds) Strike-slip Faulting and Basin Formation. Society of Economic Paleontologists and Mineralogists, Special Publications, 37, 227– 264. S NYDER , D. & H OBBS , R. 1999. The BIRPS Atlas II: A Second Decade of Deep Seismic Reflection Profiling. Geological Society, London, CD-ROM Publication. S PADINI , G., R OBINSON , A. & C LOETINGH , S. 1996. Western versus Eastern Black Sea tectonic evolution: pre-rift lithospheric controls on basin formation. Tectonophysics, 266, 139–154. S PADINI , G., R OBINSON , A. & C LOETINGH , S. 1997. Thermomechanical modelling of Black Sea basin formation, subsidence and sedimentation. In: R OBINSON , A. G. (ed.) Regional and Petroleum Geology of the Black Sea and Surrounding Region. AAPG Memoir, 68, 2nd edn. 19–38. T URCOTTE , D. L. & S CHUBERT , G. 2002. Geodynamics (2nd edn). Cambridge University Press, Cambridge. V AN W EES , J. D. & C LOETINGH , S. 1996. 3D flexure and intraplate compression in the North Sea basin. Tectonophysics, 266, 343–359. V ERALL , P. 1982. Structural Interpretation with Applications to North Sea Problems. Course Notes No. 3. Joint Association for Petroleum Exploration Courses. W ALCOTT , R. I. 1970. Flexural rigidity, thickness and viscosity of the lithosphere. Journal of Geophysical Research, 75, 3941–3954.
W ALSH , J. J., W ATTERSON , J. & Y IELDING , G. 1991. The importance of small-scale faulting in regional extension. Nature, 351, 391– 393. W ATCHARANANTAKUL , R. & M ORLEY , C. K. 2000. Syn-rift and post-rift modelling of the Pattani Basin, Thailand: evidence for a ramp-flat detachment. Marine and Petroleum Geology, 17, 937– 958. W ATTS , A. B. 2001. Isostasy and Flexure of the Lithosphere. Cambridge University Press, Cambridge. W EISSEL , J. K. & K ARNER , G. D. 1989. Flexural uplift of rift flanks due to tectonic denudation of the lithosphere during extension. Journal of Geophysical Research, 94, 13,919–13,950. W ERNICKE , B. 1985. Uniform-sense normal simple shear of the continental lithosphere. Canadian Journal of Earth Sciences, 22, 108– 125. W ESTAWAY , R. 1998. Dependence of active normal fault dips on lower crustal flow regimes. Journal of the Geological Society, London, 155, 233 –253. W HITE , N. J., J ACKSON , J. A. & M C K ENZIE , D. P. 1986. The relationship between the geometry of normal faults and that of the sedimentary layers in their hanging-walls. Journal of Structural Geology, 8, 897–909. W HITMARSH , R. B., W HITE , R. S., H ORSEFIELD , S. J., S IBUET , J., R ECQ , M. & L OUVEL , V. 1996. The ocean-continent boundary off the western continental margin of Iberia: Crustal structure west of Galicia Bank. Journal of Geophysical Research, 101, B12, 28,291–28,314. Z IEGLER , P. A. 1992. Geodynamics of rifting and implications for hydrocarbon habitat. In: Z IEGLER , P. A. (ed.) Geodynamics of Rifting, Volume III. Thematic Discussion. Tectonophysics, 215, 221– 253. Z ONENSHAIN , L. P. & L E P ICHON , X. 1986. Deep basins of the Black Sea and Caspian Sea as remnants of Mesozoic back-arc basins. Tectonophysics, 123, 181–211.
Early kinematic history of the Goban Spur rifted margin derived from a new model of continental breakup and sea-floor spreading initiation D. HEALY1 & N. J. KUSZNIR2 1
Department of Earth Sciences, University of Durham, Durham DH1 3LE, UK (e-mail:
[email protected])
2
Department of Earth & Ocean Sciences, University of Liverpool, Liverpool L69 3GP, UK Abstract: Deformation of continental lithosphere leading to breakup and the initiation of seafloor spreading can be described in terms of an upwelling divergent flow field within the continental lithosphere and asthenosphere. A new model (SfMargin) of rifted continental margin formation using this upwelling divergent flow field has been developed that successfully predicts lithosphere depth-dependent stretching and mantle exhumation at rifted margins. The kinematics of this flow field has important consequences for the bathymetric (subsidence) and thermal history of continental margin lithosphere. We apply the new model to a profile through the Goban Spur on the eastern Atlantic continental margin. Forward modelling yields parameter values describing the kinematics of the flow field leading to continental breakup and sea-floor spreading initiation that are consistent with the known amagmatic history of this margin. We employ a grid search method to systematically explore model parameter space and provide an assessment of the sensitivity of the model to the kinematic flow field parameter values. The preferred forward model parameters coincide with low values of misfit with respect to observed present-day bathymetry and free air gravity anomalies. Our results reveal the early kinematic history of the Goban Spur margin as characterized by a ratio of upwelling rate to half-spreading rate of unity, consistent with that expected for the initiation of sea-floor spreading at a non-volcanic rifted margin and only a modest amount of prebreakup lithospheric stretching (b ¼ 1.5). These results conform with recent observations of depth-dependent stretching and serpentinized mantle exhumation at this amagmatic rifted continental margin.
The early kinematic evolution of rifted continental margins plays a major role in the burial and thermal histories of the sediments deposited thereon. Recent mantle convection modelling by Nielsen & Hopper (2002) has shown the importance of the mantle upwelling velocity in determining the volume of magmatic material produced during continental breakup and sea-floor spreading initiation. Existing models of rifted continental margin formation (e.g. Le Pichon & Sibuet 1981) have been based on an extrapolation of the McKenzie (1978) intracontinental rift-basin model, with stretching factors (b) tending to infinity as the continental lithosphere necks and finally ruptures. However, more recent observations of depth-dependent stretching at rifted continental margins (Driscoll & Karner 1998; Davis & Kusznir 2004; Kusznir et al. 2004) and serpentinized mantle exhumation (Pickup et al. 1996; Manatschal et al. 2001; Whitmarsh et al. 2001) are not explained by existing models of rifted continental margin formation. A new model of rifted continental margin formation (Kusznir & Karner 2007) has been developed based on the assertion that thinning of continental lithosphere leading to breakup and sea-floor intiation
occurs in response to an upwelling divergent flow field within continental lithosphere and asthenosphere. An upwelling divergent flow field provides the kinematic basis for ocean ridge models (e.g. Spiegelman & McKenzie 1987; Su & Buck 1993), but can also be applied to the thinning of continental lithosphere leading to breakup and the initiation of sea-floor spreading. In the new model of rifted continental margin formation proposed by Kusznir & Karner (2007) the upwelling divergent flow field within continental lithosphere and asthenosphere, which leads to thinning, breakup and sea-floor spreading initiation, is defined kinematically (Fig. 1a). The model also includes a contribution to lithosphere thinning from prebreakup depth uniform (pure shear) lithosphere stretching. The modelling described in this paper seeks to recover values of two key kinematic parameters that exert a major influence on the early history of rifted continental margins: the ratio of the upwelling velocity (Vz) to the divergence velocity, or half-spreading rate (Vx), and the prebreakup pure shear lithosphere stretching factor (b). Using these optimum parameter values the new rifted margin model can then be used to predict crustal
From: KARNER , G. D., MANATSCHAL , G. & PINHEIRO , L. M. (eds) 2007. Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, London, Special Publications, 282, 199– 215. DOI: 10.1144/SP282.10 0305-8719/07/$15.00 # The Geological Society of London 2007.
200
D. HEALY & N. J. KUSZNIR
(a)
x
vx
vx
z=0
vz z
x=0 (b)
(c)
Steady-state breakup flow Vx
Vx Vz
Upwardly propagating breakup flow Vx = 0
Vx = 0
Vx = 0
Continental crust Continental lithosphere
Vx > 0 Before breakup Vx
Sea Floor Spreading
Vz
Before breakup Vx
Vz
Vx
Asthenosphere
Sea Floor Spreading
Vz
Vx Crust
Lithosphere
After breakup
After breakup
Asthenosphere
Fig. 1. Parameters used to describe different effects of model flow fields in the SfMargin model on predicted lithosphere structure. (a) The 2D model, half-space symmetrical about x ¼ 0, showing stream lines for upwelling divergent flow. Vx is the horizontal velocity (or half-spreading rate) and Vz is the vertical velocity in this corner flow solution (Batchelor 1967). (b) For steady-state flow during breakup, the upwelling divergent flow field is immediately active over the whole thickness of the lithosphere from t ¼ 0. (c) For upwardly propagating flow during breakup, the divergent flow field progressively moves up through the lithosphere section. Before the zone of divergent flow propagates to the surface Vx, above the zone of divergent flow (dotted line), is 0 and greater than 0 within the divergent zone streamlines for a pure shear flow field typical of the McKenzie rifting model. (b) Lithosphere cross-section at rupture produced by pure shear flow. Note the lack of mantle exhumation or depth-dependent stretching predicted by this model. (c) Streamlines for the upwelling divergent flow field implemented in SfMargin. The prescribed velocities, Vx and Vz, are constant along the top and left sides, respectively, and control the corner flow solution. (d) Lithosphere structure at breakup predicted by upwelling divergent flow where Vx ¼ Vz. Note the exhumation of lithospheric mantle at the ocean– continent transition. (e) Streamlines for an upwelling divergent flow field where Vz Vx. (f) Predicted lithosphere structure when Vz Vx. Note the depth-dependent stretching at the continental margin.
and lithospheric thinning, subsidence and heat flow histories for the margin transect, and is of direct importance to hydrocarbon exploration in these frontier areas. The method employs observations derived from publicly available bathymetry and gravity anomaly datasets (Sandwell & Smith 1997; IOC, IHO & BODC 2003). The technique does not depend on detailed geological data such as palaeobathymetry indicators derived from hydrocarbon wells (cf. White 1993; White & Bellingham 2002), and can therefore be applied to
any continental margin in advance of a potentially expensive drilling program.
A new model for the formation of rifted continental margins The formulation of the new model of rifted margin formation (SfMargin) is described in greater detail in the companion paper by Kusznir & Karner (2007). The upwelling divergent flow field is described using analytical stream functions defined by
KINEMATIC HISTORY OF THE GOBAN SPUR
an isoviscous corner flow model (Batchelor 1967). SfMargin uses a coupled fluid mechanics and thermal advection –diffusion solution to describe the upwelling divergent flow field of continental breakup and sea-floor spreading. This flow field is applied to a two-dimensional (2D) template of continental crust, continental lithosphere and asthenosphere, with initially uniform thickness and temperature distribution, and this flow field is symmetrical about the spreading (z-) axis (Fig. 1a). Fluid flow boundary conditions are imposed as kinematic constraints along the surface (top) and spreading axis (left) edges of the model space. The fluid flow field is used to advect lithosphere material and lithosphere and asthenosphere temperature. Note that the upwelling divergent flow field is applied after an optional amount of instantaneous prebreakup pure shear stretching. Thermal solutions are computed using a finite-difference method. Local (Airy) isostasy is assumed in order to calculate bathymetry from crustal thickness and lithosphere temperature. Crustal and lithosphere parameter values used in this study are listed in Table 1. Gravity anomalies are computed from the bathymetry, crustal structure and lithosphere temperature. The upwelling divergent flow field can be imposed on the modelled section in two distinct ways. Steady-state breakup flow imposes the divergent flow field over the full vertical thickness of the model from the beginning (Fig. 1b). As a consequence, all parts of the model are subjected to non-zero values of Vx from t ¼ 0. In contrast, upwardly propagating breakup flow starts with the divergent flow field in the asthenosphere only and then progressively migrates up through the lithosphere section (Fig. 1c). At any instant above the zone of divergent flow, the half-spreading rate Vx is 0. When the divergent flow field reaches the surface, the flow field resembles the steady-state breakup flow field. Initiating the divergent flow field in these different ways has important
201
consequences for the evolution of the rifted margin lithosphere. The key kinematic parameters controlling the fluid flow field are the vertical upwelling velocity, Vz, the horizontal half-spreading rate, Vx, and the prebreakup lithospheric stretching factor, b. Modelling by Nielsen & Hopper (2002) shows how early flow velocities during continental breakup and seafloor spreading initiation can differ from the later steady-state kinematics of sea-floor spreading. The SfMargin model allows for a short transient period (typically 2 –20 Ma) of different velocities at the early stage of sea-floor spreading. This paper focuses on recovering optimum values for the ratio of these early velocities and the prebreakup stretching factor, using observations of present-day bathymetry and gravity.
Application of SfMargin to the Goban Spur profile The Goban Spur region is considered to be a typical non-volcanic (amagmatic) and relatively sedimentstarved rifted continental margin (Horsefield et al. 1993; Watts & Fairhead 1997; Bullock 2004; Bullock & Minshull 2005). Figure 2 shows the location of the Goban Spur margin and the bathymetry and gravity observations along the profile modelled in this study. Depth-dependent lithosphere stretching has been observed for the Goban Spur rifted margin (Davis & Kusznir 2004) and mantle exhumation has been inferred (Bullock 2004; Bullock & Minshull 2005). Goban Spur shows (Fig. 3) a region of the order of 75 – 100 km width adjacent to the ocean –continent transition in which stretching of the upper continental crustal stretching is substantially less than that of the whole crust and lithosphere (Davis & Kusznir 2004). Extension and thinning of the upper crust has been measured using fault heaves from seismic reflection data; of the whole crust using
Table 1. Model parameters used in this study Symbol
b Vx Vz
Description Initial crustal thickness Initial lithosphere thickness Initial asthenosphere temperature Density of crust Density of mantle Density of sea water Goban Spur rifted margin age Prebreakup stretching factor Half-spreading rate Upwelling rate
Value and units 30 km 125 km 1300 8C 2800 kg m23 3300 kg m23 1000 kg m23 120 Ma cm year21 cm year21
202
D. HEALY & N. J. KUSZNIR
Fig. 2. Location maps for the modelled Goban Spur profile. (a) Bathymetry data from GEBCO (IOC, IHO & BODC 2003) and (b) gravity data from Sandwell & Smith (1997) are shown from the UK continental shelf out to the mid-Atlantic ridge. Our modelled profile runs parallel to the fracture zones, i.e. parallel to the inferred direction of ocean opening.
crustal basement thinning derived from crustal structure utilizing wide-angle seismology and gravity studies; and of the whole lithosphere from post-breakup subsidence using flexural backstripping of postbreakup stratigraphic data. A detailed description of stretching and thinning measurements and their analysis to determine depth-dependent lithosphere stretching is given in Davies & Kusznir (2004). We apply the new model of continental breakup and sea-floor spreading initiation to recover the kinematic flow field parameter values for this margin using the SfMargin to model observed bathymetry and gravity anomaly profiles. The bathymetry values are taken from GEBCO (IOC, IHO & BODC 2003) and the free air gravity data is from Sandwell & Smith (1997). Our selected profile across this margin is parallel to the fracture zones in the oceanic crust (Fig. 2) and we infer that this line is parallel to the oceanic opening vector. Observations are extracted from the published datasets along the profile line at a constant, arbitrary
horizontal spacing (c. 20 km). Both the bathymetry and gravity observations are affected by noise, i.e. features not modelled by SfMargin, particularly over the oceanic part of the profile. Figure 3 shows the bathymetry and gravity observations along the profile modelled in this study. Note the spikes in bathymetry and gravity anomaly between 250 and 1000 km horizontal offset from the ridge, caused by unmodelled roughness on the ocean floor. We discard observations close to (,250 km) the mid-Atlantic ridge as too noisy. The key features of a rifted continental margin are clearly present, including a distinct continental slope and sharp shelf break in the bathymetry, and a coincident dipole in the free air gravity anomaly.
Model sensitivities to kinematic parameter values The results of any forward modelling are inherently non-unique, so we have explored the SfMargin
KINEMATIC HISTORY OF THE GOBAN SPUR
203
Fig. 3. Depth-dependent lithosphere stretching has been observed on the Goban Spur rifted continental margin. Stretching and thinning factors for the upper crust are significantly less than those of the whole crust and lithosphere within 75– 150 km of the ocean– continent transition (adapted from Davis & Kusznir 2005).
model sensitivities to the kinematic parameter values through a suite of constrained models.
much sharper OCT. These geometric differences in the subsurface architecture are reflected in the distinct patterns of thinning factors shown in Figure 4g, h.
Initiation of upwelling divergent flow In the new model, the upwelling divergent flow field is applied after an optional amount of instantaneous pure shear lithosphere stretching. The flow field can be applied instantaneously to the whole thickness of the lithosphere (steady-state breakup flow) or allowed to migrate from the base of the lithosphere to the surface (upwardly propagating breakup flow). Figure 4 shows the effects of these two distinct flow regimes on the model predictions compared to the observations from Goban Spur. In each of these models, b ¼ 1.5 and Vz/Vx ¼ 1. The predicted bathymetry and gravity profiles (Fig. 4a–d) are similar for both breakup flow regimes, and match the observations fairly closely. The most important difference lies in the predicted lithosphere crosssections (Fig. 4e–f). For steady-state breakup flow, continental lithospheric mantle is exhumed at the ocean–continent transition (OCT), whereas upwardly propagating breakup flow generates a
Varying the prebreakup stretching factor, b Holding the value of the velocity ratio Vz/Vx constant at 1.0, we varied the model value of b (prebreakup stretching factor) to explore its effect on the model bathymetry and gravity predictions. For steady-state breakup flow, the effect of varying b can be seen in Figure 5. Increasing the value of b from a physical minimum of 1.0 for no stretching to 5.0, when the lithosphere has been thinned to one-fifth of its original thickness, leads to a marked steepening of the continental slope (Fig. 5i) and a broadening and deepening of the gravity dipole (Fig. 5j). Overall, a similar progression is observed in the case of upwardly propagating breakup flow (Fig. 6). In comparison to the equivalent models using steady-state breakup flow, the net effect of upwards propagation of the flow field, for a given b, tends to flatten the continental slope (compare Fig. 6g to 5g) and reduce
204
D. HEALY & N. J. KUSZNIR
Fig. 4. Model results comparing a steady-state breakup flow field with an upwardly propagating flow field. In all cases the margin age is 120 Ma, prebreakup stretching b ¼ 1.5 and the Vz/Vx ratio for the early sea-floor spreading phase was set at 1.0. (a) Model bathymetry (red) compared to observations (blue) for the steady-state flow field. (b) Bathymetry for the upwardly propagating flow field. (c) Gravity anomaly (free air over sea, Bouguer over land) for the steady-state case. (d) Gravity anomaly for the upwardly propagating case. (e) Predicted lithosphere cross-section for the steady-state case. Base lithosphere (red), Moho (blue) and base upper crust (green) are shown for the continent. Note the exhumed lithospheric mantle at the ocean– continent transition (OCT). (f) Cross-section for the upwardly propagating flow field case. Note the lack of mantle exhumation. (g) Predicted thinning factors (1 2 1/b) for the upper crust (green), whole crust (blue) and whole lithosphere (red) for the steady-state breakup flow case. (h) Thinning factors for the upwardly propagating flow field case.
KINEMATIC HISTORY OF THE GOBAN SPUR
205
Fig. 5. Model sensitivity to changing values in prebreakup pure shear b on the predicted bathymetric profiles and gravity anomalies, for the case of steady-state breakup flow. Model results shown in red, observations in blue. Note that b ¼ 1 (a, b) implies zero prebreakup pure shear and b ¼ 1 (i, j) means that pure shear stretching ruptured the lithosphere before the upwelling divergent flow field started. With the Vz/Vx ratio held constant at 1.0, higher values of b lead to a steeper, more angular continental slope in the bathymetry signal and a broader, higher amplitude dipole in the gravity anomaly at the OCT.
206
D. HEALY & N. J. KUSZNIR
Fig. 6. Model sensitivity to changing values in b for the case of upwardly propagating breakup flow. Compared to the corresponding results for steady-state breakup flow (Fig. 5), the predicted profiles have a greater wavelength and a lower amplitude at the OCT.
KINEMATIC HISTORY OF THE GOBAN SPUR
the amplitude of the gravity dipole at the OCT (compare Fig. 6h to 5h).
Varying the velocity ratio, Vz/Vx We now consider the effect of varying the velocity ratio of the 10 Ma early sea-floor spreading phase, with a constant prebreakup stretching factor of 1.5.
207
For steady-state breakup flow, the model sensitivity to variations in Vz/Vx from 0.5 to 5.0 is shown in Figure 7. The lower limit of 0.5 is close to the analytical lower limit of the stream-function cornerflow solution of 2/p for passive sea-floor spreading (Phipps Morgan 1987). Progressively higher values of Vz/Vx generate shallowing and eventual uplift (Fig. 7g) of the continental margin, and a
Fig. 7. Model sensitivity to changing values in the ratio Vz/Vx for the early sea-floor spreading phase (duration: 10 Ma) on the predicted bathymetric profiles and gravity anomalies, for the case of steady-state breakup flow. Model results shown in red, observations in blue. Bathymetric profiles in the left column, gravity anomalies in the right column. With a constant prebreakup b ¼ 1.5, higher values of Vz/Vx predict shallowing (compare 7a with e) and eventual emergence (see g) of the continental crust at the OCT, and a degradation of the gravity dipole.
208
D. HEALY & N. J. KUSZNIR
corresponding flattening of the gravity dipole at the OCT. Model sensitivity to changing the velocity ratio in the case of upwardly propagating breakup flow is shown in Figure 8. In comparison to steady-state breakup flow, higher values of the
velocity ratio (Fig. 8g, h) predict significant uplift of the margin and a very distinct and complex gravity anomaly at the OCT, reflecting a continental margin with a laterally heterogeneous density distribution.
Fig. 8. Model sensitivity to changing values in the ratio Vz/Vx for the case of upwardly propagating breakup flow. At low values of the ratio (a –d) there is little effect on the profiles compared to the corresponding steady-state breakup flow models. However, at higher values (e– f), the predicted uplift of the margin increases and the gravity anomalies become very distinct.
KINEMATIC HISTORY OF THE GOBAN SPUR
Exploring parameter trade-offs with a multidimensional grid search Least-squares method in a grid search algorithm The SfMargin model encapsulates a complex geodynamical process and displays a non-linear relationship between the various parameters and the predicted data. To further assess the validity and robustness of our forward model results, and to explore potential trade-offs among the model parameter values, we ran a grid search over discrete ranges of parameter values. We vary two-parameter values in the grid search, the ratio of Vz to Vx and b, with all other parameters held at constant values (see Table 1). The velocity ratio varied between 0.5 and 5.0, covering a range of viable upwelling and half-spreading rates. The prebreakup stretching factor b parameter is constrained to the range 1.0 – 5.0, with 1.0 as a physical minimum (no stretching) and 5.0 exceeding the likely maximum for the lithosphere beneath Goban Spur. The SfMargin model is executed for each combination of parameter values in this 2-space, and the results are stored as a suite of trial models. To compare the results of a trial model with synthetic observations we compute the mean square misfit, m: m¼
N 1X ðobsi triali Þ2 N i¼1
ð1Þ
where N is the number of data points, obsi is the i-th observation and triali is the i-th prediction. For a trial model represented by a single combination of parameter values, we calculate the misfit between model predictions and a set of observations. Note that a grid search algorithm makes no a priori assumptions about the distribution of misfit minima and we apply no smoothing or filtering to the data or the misfit function.
Investigating kinematic parameter trade-offs with a synthetic model To assess the interplay of the kinematic flow field parameters we used the least-squares grid search method with synthetic observations generated by a previous run of the forward model. Perfect, zero noise data should result in a zero misfit for the trial model at the appropriate combination of parameters. Using a previous forward model with parameter values Vz/Vx ¼ 2 and b ¼ 3, we generated synthetic bathymetry and free air gravity observations. We calculate separate misfits for both bathymetry and gravity datasets for each trial
209
model. Figure 9 shows contour plots of the mean square misfit calculated for bathymetry (Fig. 9a) and gravity (Fig. 9b) datasets using a steady-state breakup flow field. For these synthetic observations, the grid search method successfully located a single misfit minimum of zero within the parameter constraints at precisely the parameter combination used by the original forward model. Furthermore, the location of this minimum was the same for both bathymetry and gravity datasets. The same process was repeated using an upwardly propagating breakup flow field, and the contoured misfit surfaces for bathymetry and gravity data are shown in Figure 9c, d, respectively. A single misfit minimum of zero was located at the correct forward parameter combination for each dataset. These misfit surfaces generated using synthetic observations allow us to explore the trade-off between the kinematic parameter values of the Vz/Vx ratio and b. For both breakup flow fields, the low misfit region (pale colours in Fig. 9a–d) is an elongate narrow valley with its long axis lying along b ¼ 3 (the forward model parameter value used to generate the synthetic observations). The misfit is therefore more sensitive to variations in b compared to Vz/Vx.
Grid search method applied to observations from Goban Spur The least-squares grid search method has also been applied to observed bathymetry and gravity data from the Goban Spur profile. For these uncorrelated and noisy data, we use a weighted mean-square misfit defined as: m¼
N 1X ðobsi triali Þ2 s2 N i¼1
ð2Þ
where s is the standard deviation of the measurements. For the bathymetry data we took s as +250 m (IOC, IHO & BODC 2003), and for the gravity data we took s as +5 mGal (Sandwell & Smith 1997). Weighting the least-squares misfits for bathymetry and gravity observations allows us to form the sum and define a (dimensionless) composite misfit. Figure 10 shows the contoured misfit surface computed from the bathymetry (Fig. 10a) and gravity (Fig. 10b) observations measured along the Goban Spur profile. The composite misfit (weighted bathymetry misfit þ weighted gravity misfit) is also shown (Fig. 10c). The misfit surface topography is broader and flatter in all of these plots when compared to their synthetic equivalents (compare Fig. 9a, b with Fig. 10a, b). The noise on the bathymetry and gravity observations (see
210
D. HEALY & N. J. KUSZNIR
Fig. 9. Plots of contoured mean square misfit from the grid search. The misfit was calculated with respect to synthetic observations generated by a previous run of the forward model with kinematic parameter values of Vz/Vx ¼ 2 and b ¼ 3. Within the constrained parameter ranges a single misfit minimum is clearly present for both bathymetry and gravity datasets. A grid search using a steady-state breakup flow field generated the misfits shown in (a) and (b), with the corresponding plots for the upwardly propagation flow field shown in (c) and (d). The misfit minimum occurs at the same parameter combination and is precisely 0 in all cases. Trade-offs between the parameters are discussed in the text.
profiles in Fig. 3) degrades the misfit throughout the model parameter space, although the misfit minima are slightly sharper for the case of upwardly propagating breakup flow (compare Fig. 10a–c with Fig. 10d–f). In some plots there is no single misfit minimum, although there are distinct regions of low and high misfit. Using the measured data, the preferred parameter combination for Goban Spur closely coincides with the regions of lowest misfit (paler grey shades in Fig. 10). The plots shown in Figures 9 & 10 show that it may be possible to invert measured bathymetry and gravity observations to obtain the initial kinematic flow field parameters for a given rifted continental margin. The presence of misfit minima within the model parameter space, even for noisy observations, should enable us to apply formal minimization methods (e.g. Powell’s method, downhill simplex,
etc.) to locate the optimum model parameter combination. Work is under way to develop and apply this technique to other datasets, for example measured upper crustal stretching values, and to other rifted continental margins, for example Iberia, Grand Banks and Flemish Cap.
Reference frame issues In calculating the misfit for a given trial model, we transform the predicted data values into the reference frame of the observations by applying a horizontal (x-) shift and a vertical (z-) shift. The x-shift is calculated from correlation (e.g. Press et al. 1992), by selecting the lag value (x-offset) with the highest correlation between the trial model predictions and the observations.
KINEMATIC HISTORY OF THE GOBAN SPUR
211
Fig. 10. Plots of contoured mean square misfit from a grid searches using observed bathymetry and gravity data from Goban Spur. For the case of steady-state breakup flow shown in (a) – (c) the misfit surface has relatively low relief, although lower values of misfit are concentrated towards the lower left corner of each plot, corresponding to low values of Vz/Vx and b. For upwardly propagating breakup flow shown in (d) – (f) pronouned relief exists on the misfit surface, implying higher sensitivity to the parameter values of this flow field. High values of Vz/Vx are ruled out by the high misfits to the top of each plot, and for realistic values of b for the Goban Spur (,2) a low value of Vz/Vx is more likely.
212
D. HEALY & N. J. KUSZNIR
In general, we can never know the exact horizontal velocity (half-spreading rate) history of a margin and this correlation technique allows us to bring the model predictions and observations together in a systematic fashion.
Discussion and summary A newly developed forward model, SfMargin, explains recent observations at rifted continental margins and serves as a predictive geodynamic tool for the hydrocarbon industry operating in frontier areas. Using publicly available bathymetry and gravity data from the Goban Spur margin, forward modelling yields parameter values for the early kinematic history of this margin consistent with observations of depth-dependent lithosphere stretching. A sensitivity analysis among the kinematic parameters supports the results of the forward modelling. A least-squares grid search method employing synthetic observations allows us to assess the trade-offs between the parameter values. The same grid search method applied to observations of bathymetry and gravity measured along the Goban Spur profile
yields a misfit minimum almost coincident with that derived from forward modelling. The observation of depth-dependent stretching at Goban Spur and other non-volcanic and volcanic margins (Davies & Kusznir 2004; Kusznir & Karner 2007) rules out a dominantly depth-uniform (pure shear) mechanism for thinning continental lithosphere leading to continental breakup and sea-floor spreading initiation. Instead, the dominant process reponsible for continental lithosphere thinning leading to seafloor spreading initiation is an upwelling divergent flow field within continental lithosphere and athenosphere. Models using this mode of continental lithosphere thinning are able to satisfactorily predict observed bathymetry and gravity anomalies. While a contribution to continental lithosphere thinning from a depth-uniform (pure shear) mechanism may exist, it is not believed to be the dominant lithosphere deformation process leading to breakup. A region of exhumed serpentinized mantle of the order of 70 km wide, separating thinned continental crust and oceanic crust, has been inferred for at Goban Spur margin from seismic refraction and reflection studies by Bullock & Minshull (2005), and is summarized in Figure 11. The exhumation
Fig. 11. Seismic refraction and reflection studies on Goban Spur suggest the existance of a region of exhumed serpentinized mantle of the order of 70 km wide separating thinned continental crust and oceanic crust (from Bullock & Minshull 2005).
KINEMATIC HISTORY OF THE GOBAN SPUR
of continental lithospheric mantle is predicted by the steady-state pattern of breakup flow shown in Figure 4e; in contrast, the upwardly propagating flow field pattern model shown in Figure 4f predicts a much sharper ocean –continent transition with no exhumation of continental lithosphere mantle. Intuitively, the upwardly propagating flow field pattern seems more physical and likely to occur; however,
20
100
0
Distance (km) 200 300
400
500
Depth (km)
while this model does satisfactorily predict observed bathymetry and the free air gravity anomaly, it does not predict the exhumation of continental lithospheric mantle. The possibility exists that the dominant contribution to mantle exhumation at nonvolcanic margins is melt suppression during seafloor spreading initiation (rather than continental lithospheric mantle exhumation) analogous with
20
continental crust
–20 –40
–100
–100
–120
Depth (km)
(a) 100
0
200
300
400
500
–120
20
0
0
–20
–20
–40
–40
–60
–60
–80
–80
–100
–100
(b)
–120
100
0 water
Depth (km)
400
500
–60
continental lithospheric mantle
–80
200
300
400
500
20
0
100
200
300
400
500
(e) 0
100
200
300
400
500
continental margin
0
–20
–20
–40
–40
–60
–60 –80
oceanic lithosphere &
–100 asthenosphere
(d)
–120
0
–120
Distance (km) 200 300
–40
–80
–80
100
–20
Moho
–60
20
0
0
0
20
213
(c)
–100
(f)
–120
Fig. 12. Development in time of continental lithosphere thinning and bathymetry predicted by the upwelling divergent flow model of continental breakup and sea-floor spreading initiation compared for steady-state and upwardly propagating upwelling divergent flow patterns. The model with steady-state flow field (a)– (c) ruptures continental crust before continental lithospheric mantle, and leads to the exhumation of a broad region of continental lithosphere mantle. The model with an upwardly propagating flow field (d)– (f) ruptures continental lithospheric mantle before the continent crust, and results in no exhumation of continental lithospheric mantle.
214
D. HEALY & N. J. KUSZNIR
mantle exhumation during ultra-slow sea-floor spreading at the Gakkel and South West Indian ridges (Dick et al. 2003; Michael et al. 2003). Another important difference between the steady-state and upwardly propagating flow field pattern models (Fig. 4) of continental lithosphere thinning is in the time sequence of the rupture of continental crust and lithospheric mantle. The steady-state flow field predicts that the continental crust ruptures before the continental lithosphere mantle, while the upwardly propagating flow field predicts that the continental lithospheric mantle ruptures before the continental crust. These are illustrated in Figure 12. At present it is not known which model best describes the natural process: new observations are required to answer this question. Further work will employ measured upper crustal stretching factors as observations, and model the inferred conjugate margin in the western Atlantic around Flemish Cap. SfMargin has been developed as part of the NERC Ocean Margins iSIMM (integrated Seismic Imaging and Modelling of Margins) project. The iSIMM team comprises researchers from the Universities of Liverpool and Cambridge, Badley Geoscience and Schlumberger, and is funded by a consortium of NERC, DTI, Agip, BP, Amerada Hess, Anadarko, ConocoPhillips, Shell, Statoil, and WesternGeco. iSIMM investigators are R.S. White, N.J. Kusznir, P.A.F. Christie, A.M. Roberts, A. Chappell, J. Eccles, R. Fletcher, D. Healy, N. Hurst, Z. Lunnon, C.J. Parkin, A.W. Roberts, L.K. Smith, R. Spitzer and V. Tymms. The GMT software of Wessel & Smith (1998) was used to produce some of the figures. The authors gratefully acknowledge discussions with iSIMM team colleagues, constructive reviews by G. Houseman, J. Nunn and an anonymous reviewer and the editorial patience of G. Karner, all of which significantly improved this manuscript.
References B ATCHELOR , G. K. 1967. An Introduction to Fluid Dynamics. Cambridge University Press, Cambridge. B ULLOCK , A. D. 2004. From continental thinning to seafloor spreading: a geophysical study of rifted margins southwest of the UK. PhD thesis, University of Southampton. B ULLOCK , A. D. & M INSHULL , T. A. 2005. From continental extension to seafloor spreading: crustal structure of the Goban Spur rifted margin, southwest of the UK. Geophysical Journal International, 163, 527– 546, doi:10.1111/ j.1365-246X.2005.02726.x. D AVIS , M. & K USZNIR , N. J. 2004. Depth-dependent lithospheric stretching at rifted continental margins. In: K ARNER , G. D., T AYLOR , B., D RISCOLL , N. W. & K OHLSTEDT , D. L. (eds) Rheology and Deformation of the Lithosphere at Continental Margins. Proceedings of NSF Rifted Margins Theoretical Institute, Volume 1. Columbia University Press, New York, 92–136.
D ICK , H. J., L IN , J. & S CHOUTEN , H. 2003. An ultraslow-spreading class of ocean ridge. Nature, 426, 405– 412. D RISCOLL , N. W. & K ARNER , G. D. 1998. Lower crustal extension across the Northern Carnarvon basin, Australia: Evidence for an eastward dipping detachment. Journal of Geophysical Research, 103, 4975– 4991. H ORSEFIELD , S. J., W HITMARSH , R. B., W HITE , R. S. & S IBUET , J.-C. 1993. Crustal structure of the Goban Spur rifted continental margin, NE Atlantic. Geophysical Journal International, 119, 1– 19. IOC, IHO & BODC. 2003. Centenary Edition of the GEBCO Digital Atlas. Published on CD-ROM on behalf of the Intergovernmental Oceanographic Commission and the International Hydrographic Organization as part of the General Bathymetric Chart of the Oceans, British Oceanographic Data Centre, Liverpool. K USZNIR , N. J. & K ARNER , G. D. 2006. Continental lithospheric thinning and breakup in response to upwelling divergent mantle flow: application to the Woodlark, Newfoundland and Iberia margins. In: K ARNER , G. D., M ANATSCHAL , G. & P INHEIRO , L. M. (eds) Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, London, Special Publications, 282, 389– 419. K USZNIR , N. J., H UNSDALE , R. & R OBERTS , A. M. 2004. Timing of depth-dependent lithosphere stretching on the S. Lofoten rifted margin offshore mid-Norway: pre-breakup or post-breakup? Basin Research, 16, 279–296. L E P ICHON , X. & S IBUET , J. C. 1981. Passive margins: a model of formation. Journal of Geophysical Research, 86, 3708– 3720. M ANATSCHAL , G., F ROITZHEIM , N., R UBENACH , M. J. & T URRIN , B. 2001. The role of detachment faulting in the formation of an ocean–continent transition: insights from the Iberia Abyssal Plain. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: Evidence From Land and Sea. Geological Society, London, Special Publications, 187, 405– 428. M C K ENZIE , D. P. 1978. Some remarks on the development of sedimentary basins. Earth and Planetary Science Letters, 40, 25–32. M ICHAEL , P. J. & L ANGMUIR , C. H. ET AL . 2003. Magmatic and amagmatic seafloor generation at the ultraslow-spreading Gakkel ridge, Arctic Ocean. Nature, 423, 956– 962. M U¨ LLER , R. D., R OEST , W. R., R OYER , J.-Y., G AHAGAN , L. M. & S CLATER , J. G. 1997. Digital isochrons of the world’s ocean floor. Journal of Geophysical Research, 102(B2), 3211–3214. N IELSEN , T. K. & H OPPER , J. R. 2002. Formation of volcanic rifted margins: are temperature anomalies required? Geophysical Research Letters, 29, 2022– 2025. P HIPPS M ORGAN , J. 1987. Melt migration beneath midocean spreading centers. Geophysical Research Letters, 14, 1238– 1241. P ICKUP , S. L. B., W HITMARSH , R. B., F OWLER , C. M. R. & R ESTON , T. J. 1996. Insight into the nature of the
KINEMATIC HISTORY OF THE GOBAN SPUR ocean-continent transition off West Iberia from a deep multi-channel seismic reflection profile. Geology, 24, 1079–1082. P RESS , W. H., T EUKOLSKY , S. A., V ETTERLING , W. T. & F LANNERY , B. P. 1992. Numerical Recipes in FORTRAN: The Art of Scientific Computing. Cambridge University Press, Cambridge. S ANDWELL , D. T. & S MITH , W. H. F. 1997. Marine gravity anomaly from Geosat and ERS 1 satellite altimetry. Journal of Geophysical Research, 102(B5), 10039– 10054. S PIEGELMAN , M. & M C K ENZIE , D. 1987. Simple 2-D models for melt extraction at mid-ocean ridges and island arcs. Earth and Planetary Science Letters, 83, 137–152. S U , W. S. & B UCK , W. R. 1993. Buoyancy effects on mantle flow under midocean ridges. Journal of Geophysical Research, 98(B7), 12,191–12,205.
215
W ATTS , A. B. & F AIRHEAD , J. D. 1997. Gravity anomalies and magmatism along the western continental margin of the British Isles. Journal of the Geological Society, London, 154, 523–529. W ESSEL , P. & S MITH , W. H. F. 1998. New, improved version of the Generic Mapping Tools released. EOS, Transactions of the American Geophysical Union, 79, 579. W HITE , N. 1993. Recovery of strain rate variation from inversion of subsidence data. Nature, 366, 449–452. W HITE , N. & B ELLINGHAM , P. 2002. A two-dimensional inverse model for extensional sedimentary basins, 1, Theory. Journal of Geophysical Research, 107, (B10), 2259– 2279. W HITMARSH , R. B., M ANATSCHAL , G. & M INSHULL , T. A. 2001. Evolution of magma-poor continental margins from rifting to sea-floor spreading. Nature, 413, 150–154.
The boundary between continental rifting and sea-floor spreading in the Woodlark Basin, Papua New Guinea A. M. GOODLIFFE1 & B. TAYLOR2 1
Department of Geological Sciences, University of Alabama, Tuscaloosa, Alabama, USA (e-mail:
[email protected])
2
School of Ocean and Earth Science and Technology, University of Hawaii, Honolulu, Hawaii, USA
Abstract: Seismic reflection, bathymetry, acoustic imagery and magnetic data are presented that encompass the boundary between rifting of the Papuan continent and westward propagating seafloor spreading in the Woodlark Basin. West of the spreading tip, the southern margin is characterized by large fault blocks, which were tilted to the south on north-dipping normal faults during the current rifting phase, and graben further south where previous rifting failed. The northern margin is devoid of large offset normal faults and its subsidence from near sea level requires synrift flow of the lower crust. The margin asymmetry primarily reflects across-strike differences in the prerift geology, morphology and rheology. Horst and graben with 3 km relief below sea level occur adjacent to the spreading tip. Although heat flow data imply greater crustal thinning to the south of Moresby continental seamount, seismicity shows that deformation is currently focused on the Moresby normal fault that bounds its northern side. Based on logging results from ODP Leg 180 and semblance analysis, depth conversion of seismic reflection data show that Moresby Fault has a dip 298, which is compatible with an earthquake slip plane of 308–338. Directly east of Moresby Seamount, the first spreading segment is identified on the basis of magnetic, bathymetry and sidescan data to comprise two ring dykes in the west and Cheshire Seamount in the east. A seismic reflection line crossing the western ring dyke shows an intrusive body that cuts Moresby Fault at depth. Currently, intrusive rocks have only reached the surface near the ring dyke, leaving flat-lying synrift sediments, sills (?) and surface lava flows overlying the main intrusive body. The boundary between continental rifting and sea-floor spreading in the Woodlark Basin is spatially abrupt, although the processes overlap in time. The progression from rifting to spreading is characterized by a decrease in sedimentation as the margins are progressively thinned, subside below sea level, are eroded less and trap sediments in proximal basins. The post-rift sedimentation near the continent– ocean boundary is hemipelagic and drapes breakup topography without a distinct breakup unconformity. Synrift sediments, deposited above the 8.4 Ma rift onset unconformity and prior to breakup, are characterized by rotated sections (with parallel or diverging reflectors) as well as by ponded sections that onlap bounding fault blocks where local extension has ceased. Splayed reflectors representing sediment deposition on rotating fault blocks do not characterize the full synrift section, but only those times and places where extension is localized.
The transition between continental rifting and seafloor spreading marks the boundary between two fundamentally different modes of strain accommodation. How this transition takes place is poorly understood at all scales. Much of our present knowledge is based on observations from old and cold passive margins where the location of the conjugate margin is ambiguous and the basement is buried by thick sediments or voluminous magmatism (Holbrook & Kelemen 1993; Driscoll & Karner 1998). Active rifting–spreading boundaries are found only in a small number of locations. The Red Sea and the Gulf of Aden, both in cratonic continental settings, are currently forming as a result of rifting of the Arabian Peninsula from Africa (Cochran 1981; Martinez & Cochran 1988). The
Woodlark Basin and the Gulf of California (Moore 1973; Taylor et al. 1995, 1999; Fletcher & Munguia 2000) are also the result of continental breakups, but the tectonic history is made more complex by the interaction with recently active subduction systems. Several kinematic models have been proposed to explain basin-scale variations in features such as continental margin width, breakup style and spreading centre segmentation by varying the amount of extension prior to the initiation of sea-floor spreading (Courtillot 1982; Vink 1982; Martin 1984). Finite-element models have been used to match the observed data by varying the thickness and composition of the crust, the geotherm, the extension rate, and the strength and distribution of pre-existing
From: KARNER , G. D., MANATSCHAL , G. & PINHEIRO , L. M. (eds) Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, London, Special Publications, 282, 217– 238. DOI: 10.1144/SP282.11 0305-8719/07/$15.00 # The Geological Society of London 2007.
218
A. M. GOODLIFFE & B. TAYLOR
weak zones (Bassi et al. 1993; Dunbar & Sawyer 1996; Harry & Bowling 1999; Lavier & Buck 2002; Corti et al. 2003; Van Wijk & Blackman 2005). Modelling studies have been hampered by a lack of accurate kinematic constraints on parameters such as strain rate, duration of rifting preceding rupture, and prerifting continental width and crustal thickness. The evolution of rift architecture through time is typically described as a transition from broadly distributed extension, where large crustal blocks rotate along faults that sole out at depth or transition to ductile deformation, to a narrow zone of focused extension that allows rapid lithospheric thinning and melt generation (Cochran 2005). In contrast, South Atlantic passive continental margins (Karner et al. 2003) are characterized by the stacking of regional synrift sag basins. Large, late-stage rift faults in the South Atlantic margins do not control the regional development of synrift accommodation and are thus a small part of the extension process. Late-stage subsidence during the rifting– spreading transition is usually explained by interpreting the margin being studied as the upper plate in a simple shear model. This has resulted in the ‘upper plate paradox’ (Driscoll & Karner 1998), where both conjugate margins have been interpreted as the upper plate. This paradox can be avoided by using lower crustal and upper mantle deformation to create subsidence during late-stage continental extension (Karner et al. 2003). Subsidence may also originate from cooling and contraction of anomalously hot asthenosphere in the region of deformation (Buck 2004) or be the result of an upwelling divergent flow field within continental lithosphere and asthenosphere (Kusznir et al. 2004). The processes involved in the initiation and growth of a spreading centre in a continental rift have been examined on a large scale at several locations. In the Red Sea, continuous sea-floor spreading in a welldefined axis in the south becomes discontinuous to the north, forming discrete deeps interpreted as cells of sea-floor spreading (Bonatti 1985). The regular spacing of the deeps and the similar spacing of discontinuities in the southern Red Sea suggest that both may be caused by asthenospheric upwelling, the wavelength of which is controlled by the ratio of the viscosities and the layer thicknesses. Once a diapir reaches the surface at the centre of a segment it continues to grow along strike, eventually merging with its neighbours to create a continuous spreading centre. During rifting, adjacent segments may be offset by accommodation zones that eventually form the site of oceanic transform faults (Cochran & Martinez 1988). In the Woodlark Basin, two distinct modes of sea-floor spreading initiation have been identified. One involves propagation of sea-floor spreading into the rifting continent; the other nearly simultaneous nucleation of
sea-floor spreading along more than 100 km-long rift segments (Taylor et al. 1999). The former results in a continent–ocean boundary (COB) that is highly oblique to early margin structures, abyssal hill fabric and magnetic isochrons. In the later case the COB, abyssal hill fabric and magnetic isochrons are all parallel. Other studies of the geological and dynamic processes by which sea-floor spreading initiates in a continent (Rosendahl 1987; Hayward & Ebinger 1996) demonstrate that many of the details are highly debated. One shortcoming of all these studies is that they have not determined how the spreading centre first ruptures the surface and how this location is related to the late-stage continental faulting. The Woodlark Basin is perhaps the only place where the COB has been located regionally at a kilometre scale. Sediment cover is generally thin, and the opening kinematics, strain history and margin subsidence have been intensely investigated (Taylor et al. 1995, 1999; Goodliffe 1998; Martinez et al. 1999; Taylor & Huchon 2002). This affords us an unprecedented view of the detailed processes involved in: (1) continental rifting directly preceding the initiation of a spreading centre; (2) the mechanics of spreading centre emplacement; and (3) the early evolution of a spreading centre.
Spreading history and Euler poles Rifting of the Papuan Peninsula started at approximately 8.4 Ma (Taylor & Huchon 2002), and led to separation of the once contiguous Woodlark and Pocklington rises that bound the Woodlark Basin (Fig. 1). The oldest oceanic lithosphere in the Woodlark Basin, formed during magnetic chron 3A.1 (Taylor et al. 1999) (c. 6 Ma: Cande & Kent 1995), is found at the eastern end of the basin adjacent to the southern margin; any older oceanic lithosphere has been subducted under the Solomon Islands. Reconstructions based largely on fracture zone trends (Goodliffe 1998) show that since magnetic chron 3A.1 the majority of basin opening can be described as a rotation about a single Euler pole (1478 þ 18–28E, 9.38S + 0.28S, 4.2348 Ma21 Fig. 2), implying that rifting initiated simultaneously throughout the basin. Abyssal hill fabric, magnetic isochrons, a major low-angle fault bordering Misima Island (Hill 1995), faults bounding Milne Bay graben (Jongsma 1972) and Goodenough Fault, a structure bounding the southern side of Goodenough Basin (Fang 2000) (Figs 1 & 2), are all co-polar with this Euler Pole. Sea-floor spreading propagated in a stair-step fashion from the east into rifting crust at an average rate of approximately 140 km per Ma21 (Taylor & Exon 1987), reaching Moresby Seamount, where the rifting to spreading transition is
BOUNDARY BETWEEN RIFTING AND SPREADING
219
Fig. 1. Regional setting of the boundary between continental rifting and sea-floor spreading in the Woodlark Basin. The conjugate Woodlark and Pocklington Rises have separated over the last 8.4 Ma as a result of rifting the Papuan Peninsula and the westward propagation of sea-floor spreading (accreted oceanic crust is outlined by a solid black line; the spreading centre is depicted by a black line with white centre). Dashed black lines indicate active strike-slip motion. Teleseismicity is shown by black circles.
currently found. The 500 km-long active spreading centre, located on the basis of bathymetry and acoustic imagery, comprised five main segments at the start of the Brunhes chron (0.78 Ma). An up to 228 counter-clockwise reorientation of the spreading centre at 80 ka (Goodliffe et al. 1997) in response to a shift in the Euler pole at 0.52 Ma (Goodliffe 1998) (Fig. 2) resulted in the subdivision of many of the spreading segments into shorter segments that are now oblique to the Brunhes isochron. The modern Euler pole (128S, 1448E, 2.4378 Ma21: Fig. 2) is defined largely by the offset of pre-existing fracture zones and is close to the Australia– Woodlark Euler pole (11.288S, 147.6.28E, 2.82 + 0.188 Ma21) (Wallace et al. 2004) determined by simultaneously inverting GPS data and earthquake slip vectors with transform orientations. Spreading segment 1 extends from the current spreading tip at 1518400 E–152851.60 E and comprises three subsegments, 1a, b and c (Fig. 2). Seafloor spreading on this segment nucleated between 1518450 E and 152851.60 E at approximately 0.8 Ma (Taylor et al. 1995; Goodliffe 1998). At the resolution of the available data and analysis techniques the COB is coincident with the Brunhes isochron
and parallel to great circles about the pre-0.52 Ma Euler pole (Fig. 2). The change in Euler pole resulted in the subdivision of segment 1 into two shorter segments and an approximately 88 rotation. Unlike segments 2– 5, segments 1b and 1c are closer to co-polar with the pre-0.52 Ma Euler pole and the Brunhes isochron than to the current pole. To the north of segment 1, the seismically active Egum Graben (Figs 1 & 2) cross-cuts normal faults in the northern margin of segment 1 and has a trend that is co-polar with the present Euler pole. Although this may be coincidental, it may mark the reactivation of a favourably trending structure by the change in Euler pole. This might also explain the concurrent sea-floor spreading and continental rifting at this longitude (Taylor et al. 1995). The fault bounding Egum Graben has been interpreted as a continuation of the strike-slip fault separating the eastern Woodlark Basin from the Solomon Sea (Taylor et al. 1999). This transtensional fault system may bound a third (Trobriand) plate, including the margin to the north of Moresby and Egum Graben, Egum Atoll and Woodlark Island (Davies et al. 1984), and form a triple junction with
220
A. M. GOODLIFFE & B. TAYLOR
Fig. 2. Map of major faults seen at the surface in the vicinity of the rifting–spreading transition (a). The axes of the active neovolcanic zones are shown by thick solid lines and the limit of oceanic crust is dashed. Sea-floor magnetization illuminated from the north by bathymetry (b). Oceanic crust formed since the start of the Brunhes chron (0.78–0 Ma) on segment 1 has a positive magnetization. At the resolution of the data the continent– ocean boundary and the 0.78 Ma isochron are coincident. Small and great circles about the present Euler pole (128S, 1448E, 2.4378 Ma21, solid grey lines) and the pre-0.52 Ma Euler pole (1478 þ 18–28E, 9.38S + 0.28S, 4.2348 Ma21,
BOUNDARY BETWEEN RIFTING AND SPREADING
spreading segment 1a and the Papuan rifts. Given the complex tectonic setting of segment 1a it is not clear that it is opening around the same Euler pole as the spreading segments further east.
The continental margins proximal to the initiation of sea-floor spreading Using swath bathymetry data and three representative six-channel seismic lines (see Appendix 1 for details of the data used and processing steps), we first describe the margin structures that surround sea-floor spreading segment 1. The margin of the Woodlark Basin south of segment 1 and Moresby Seamount (Fig. 2) is characterized by large fault blocks, trending on average east–west but locally NE–SW, that are typically tilted to the south and overlain by isolated sedimentary basins that record varying degrees of rotations (Fang 2000) (Fig. 3). Poor basin interconnectivity and an absence of sampling make attempts at dating fault motion purely speculative. A scarcity of seismicity in this region (Abers et al. 1997; Ferris et al. 2002) (Fig. 1) indicates that there is currently little fault motion. An analogy has been made to oceanic rift propagation, with many of the fault blocks rotated to the NE about vertical axes by shear between laterally offset centres of extension (Mutter et al. 1996). Six-channel seismic (SCS) line MW9304-30 (Figs 2 & 3) crosses the current focus of deformation to the west of Moresby Seamount where sea-floor spreading has not yet initiated. At the southern end of the line there are isolated sediment ponds that record tilting to the south of at least two large crustal blocks. The northward tilting of an adjacent progradational sequence illustrates that deformation is locally more complex. A graben centred on SP900 that marks the deepest point at this longitude widens to the east to become the South Moresby Graben (SMG). Further north a series of three fault blocks tilted to the south on north-dipping normal faults sit in the westward extension of Moresby Seamount and North Moresby Graben (NMG). Some complex basement geometries (for example, at SP1560) suggest locally more complex deformation, possible involving unresolved south-dipping faults. A fourth ridge (SP1900) is an uplifted horst block with a thick synrift sedimentary sequence. To the north, the largely undeformed northern margin with its thick synrift sedimentary sequence dips to the south.
221
The stratigraphy here is contiguous with that imaged to the east and sampled during ODP Leg 180 (Taylor & Huchon 2002). North of SP2500, prominent intrabasement reflectors below the synrift sediments dip to the north. These are believed to originate from sills within the approximately 66 Ma diabase and metadiabase basement (Brooks & Tegner 2001; Monteleone et al. 2001). SCS line MW9304-50, 18.5 km to the east, crosses the SMG, Moresby Seamount, and NMG. Owing to the near coincidence with this line of multichannel seismic (MCS) line EW9510-1366 and the ODP Leg 180 drilling transect, the stratigraphy and basin history of Moresby Seamount and the northern margin is very well known at this longitude. In summary, this involved gradual subsidence from sea level and tilting to the south of the northern margin, with early focusing of extension to the north and south of Moresby Seamount and its subsequent isostatic rise between late bounding faults (Taylor & Huchon 2002) (Fig. 4). It is important to note that, although the northern margin is bounded in the south by the major north-dipping Moresby Fault, the wavelength of subsidence is too great to be due to that fault alone. The absence of large-offset normal faults on the northern margin means that other mechanisms such as flow of the lower crust must be called on to explain its subsidence. The ODP drilling transect did not extend to the southern section of SCS line MW9304-50, leaving the details of the evolution of this part of the basin unexplored. SMG (Fig. 2) is deepest directly to the south of the peak of Moresby Seamount. Here it is subdivided into two parts separated by an north– south trending ridge that plunges to the north. To the east of the ridge the graben floor reaches 3150 m. Low-amplitude positive magnetization implies that rifting may have been accompanied by magmatism extending west from spreading segment 1b. The graben floor is subhorizontal and sediment covered, with the exception of localized regions of high acoustic backscatter in the north indicative of fault scarps (Fig. 2c). A 4 0.1 km linear feature on the sea floor at 1518420 E, 98570 S is interpreted as a lava flow. To the west of the northplunging ridge the graben floor is characterized by generally smooth sea floor reaching depths of 3000 m. A somewhat rougher sea floor close to Moresby Seamount is likely to be the result of debris flows. Consistently high values of heat
Fig. 2. (Continued) dashed grey) are shown. The 0.78 Ma boundary is co-polar with great circles about the pre-0.52 Ma Euler pole, whereas Egum graben is co-polar with great circles about the current Euler pole. Bathymetry illuminated from the north and overlain by acoustic imagery (c) shows the location of the neovolcanic zone (black ¼ high acoustic backscatter) relative to the main fault trends in the vicinity of the rifting– spreading transition. The location of the six-channel seismic lines described in this paper are shown along with ODP sites 1108, 1117 and 1118.
222
A. M. GOODLIFFE & B. TAYLOR
Fig. 3. Six-channel seismic (SCS) lines MW9304-30 (top), MW9304-50 (middle) and MW9304-70 (bottom) across the continental margins in the vicinity of the rifting– spreading transition. The rotation of back-tilted fault blocks is recorded locally by the fanning of sediment packages in half graben that form small isolated basins. The southern margin is dominated by these tilted fault blocks, whereas the northern margin is largely unfaulted, but subsided and south tilted. From SCS line MW9304-30 in the west to MW9304-70 in the east there is a focusing of extension first at North and South Moresby Graben and then on spreading segment 1a. The blue line marks the base of the synrift sediments. ODP Site 1118 is projected from its true location 1.5 km to the east. ODP Site 1117 is projected from its true location 0.7 km to the west. The location of the SCS lines is shown in Figure 2.
flow, reaching a peak of 254 mW m22, were recorded in SMG (Goodliffe et al. 2000). This is in contrast to a maximum of 100 mW m22 recorded in NMG. SCS and MCS data in SMG (Fig. 3) do not satisfactorily image basement. Nevertheless, gravity data indicate a similar amount of crustal thinning as NMG (Goodliffe et al. 1999b). To the south, between SP500 and SP1200 on SCS line MW9304-50, north-dipping normal faults bound crustal blocks that have been tilted to the south. Further to the south at SP390 a southdipping normal fault forms the northern margin of a shallow rift basin, the eastward extension of which is seen in the form of a series of southdipping faults on SCS line MW9304-70. The southern side of this rift basin was not imaged
owing to its proximity to the reefs that top Pocklington Rise. Further east, between 1518500 E and 1528300 E, a scarp dipping approximately 308S is seen in the bathymetry (Figs 1 & 2). This fault may form the northern boundary of a basin seen to the west of Misima Island. Still further to the south a large (.160 km long and c. 40 km wide) basin is bounded by reefs (Fig. 1). The absence of seismicity and the location of this basin away from the focus of current activity suggest that it is an early (failed) rift that predates the rifting that led to breakup and the initiation of sea-floor spreading on segment 1. To the north of SP250 on SCS line MW9304-70 there are at least three large fault blocks that have been tilted to the south by motion on north-dipping
km
N VE = 3
3
1
3
1
2
0
2
0
1
0
1116
Moresby Seamount
1114 40
1118 50
North Moresby Graben
30
1109 60
70
8.4 Ma
5.5 Ma
3.8 Ma
3.2 Ma
2.55 Ma
1.6 Ma
0.0 Ma
80
0
2
0
4
2
1115
1
2
0
2
0
2
Fig. 4. Interpreted line drawing of MCS line EW9510-1366 (top panel), 1.5 km to the east of SCS line MW9304-50. Reconstructions at 1.6, 2.55, 3.2, 3.8 and 5.5 Ma (lower panels) use constraints from seismic stratigraphy and ODP drilling (magnetobiostratigraphy and palaeowater depths). Adapted from Taylor & Huchon (2002).
5
3
km
km
1
km
10
km
0
km
km
km
BOUNDARY BETWEEN RIFTING AND SPREADING 223
224
A. M. GOODLIFFE & B. TAYLOR
normal faults. North of SP800 fault geometries become unclear, possibly because of extreme fault block rotation and subsequent cross-cutting by the late-stage north-dipping faults indicated by the small sea-floor offsets. Locally, the sea floor is underlain by horizontally stratified sediments. This close to the western end of segment 1b, the presence of lava flows, dykes and sills cannot be ruled out. Basement and fault plane reflectors on the small ridge, continuous with Moresby Seamount to the west and centred on SP1600, suggest that it is underlain by south-tilted fault blocks. Directly to the north the sea floor has high acoustic reflectivity and is interpreted to be floored by recent lavas (Fig. 2c). Small ridges on the sea floor at SP1830 and SP1920 are part of an approximately 4 km-diameter circular feature on the sea floor. Below the region of high sea-floor reflectivity there are no coherent reflectors. At SP2195 and SP2350 major south-dipping normal faults downdrop the thick synrift sedimentary sequence of the northern margin to the south. As in SCS lines MW9304-30 and MW9304-50, the margin further north is tilted to the south and shows no evidence of major fault offsets. The SCS lines, especially MW9304-30 and MW9304-50, show marked margin asymmetry (Fig. 3). The northern margin slopes gradually to the south with little evidence of large-offset normal faults until the NMG is reached. In the south large fault blocks are rotated. At face value this could imply a simple shear origin for the margin (Wernicke 1985; Lister et al. 1991), but a detachment fault of sufficient size to explain the wavelength of the subsidence has not been identified. More probably, the structural asymmetry is a function of the preexisting continental morphology–geology–thickness (forearc basin v. mountainous continental arc and back-arc) combined with lower crustal flow (Goodliffe et al. 1999a). Between 1518550 E and 1528500 E the fabric on the southern and northern rifted margins close to the continent–ocean boundary (COB) is broadly parallel to the N848W-trending COB (Fig. 2). This is consistent with the initiation of sea-floor spreading almost simultaneously along segment 1 (Taylor et al. 1999). The normal faults bounding the margin north of segment 1 are cut by Egum Graben. This more than 100 km-long structure trends N738E from Moresby Seamount and is the focus of ongoing seismic activity with many .Mw6 earthquakes (Taylor et al. 1995, 1996; Abers et al. 1997). To the east the rift narrows and shallows towards the platform surrounding Woodlark Island. At 1528130 E Egum Graben is 17.5 km wide and bounded by scarps with approximately 408 dips. The main rift floor, at a depth of 2500 m, is 600 m below the rift shoulders. A
nested 4.1 km-wide and 50 km-long graben on the southern side of Egum Graben locally reaches a depth of 2600 m and terminates in the west against Cheshire Seamount. There are small edifices to the south of Egum Graben at 1528100 E, 98500 S that may be volcanic in origin, but are no longer active (Fig. 2c). At 1528400 E, 108180 E the southern margin is cross-cut by a WNW-trending COB (Fig. 2). Seafloor spreading on segments 1 and 2 propagated past one another, forming an overlapper that surrounded and rotated counter-clockwise a sliver of continental crust (centred at 1528400 E, 108100 E) (Goodliffe 1998). Since 80 ka, sea-floor spreading on segment 2 has rotated and jumped to the south. Propagation of the new spreading into the margin has created a small re-entrant at 1528450 E, 108250 S (Fig. 2c). Kinematic reconstructions at the longitude of Moresby Seamount (Fang 2000; Taylor & Huchon 2002) of faulting and subsidence north of 108S show that many of the faults (with dips ranging from 308 to 608) that are active today were active from at least 5.5 Ma (Fig. 4). Between 5.5 and 1.6 Ma, extension on the major faults only added up to 7 km, compared to more than 12 km between 1.6 Ma and the present day. However, similarities in sedimentation rate between ODP sites 1108, 1109, 1118 and 1114 (Takahashi et al. 2002) show that these sites were in a common tectonic setting with no large depositional barriers separating them until at least the late Pliocene (Robertson & Sharp 2002). Although the rift basins in this region were growing, fault growth was being outpaced by sedimentation. The post-1.6 Ma increase in extension rate at the longitude of Moresby Seamount (Taylor & Huchon 2002) and a dramatic decrease in sedimentation on the northern margin at 1.2 Ma (Takahashi et al. 2002), implying that the rapid subsidence of NMG and uplift of Moresby Seamount had cut off sedimentation to the north (Sharp & Robertson 2002), both indicate strain localization on the rifts surrounding what would become sea-floor spreading segment 1. When sea-floor spreading initiated on segment 1 at about 0.8 Ma, the neovolcanic zone extended only from 1518450 E to 1528400 E (Goodliffe 1998). Beyond these limits, strain accommodation and crustal thinning was still not sufficiently focused for spreading to initiate. To the east strain localization was achieved by propagation of the neovolcanic zone, forming a spreading centre that overlapped with segment 2 (Goodliffe 1998). The neovolcanic zone did not propagate to the west, remaining stalled at 1518480 E for approximately 0.7 million years. Strain accommodation on the faults directly to the west of the spreading tip must therefore have still remained distributed, at
BOUNDARY BETWEEN RIFTING AND SPREADING
225
least within the confines of the segment 1 rift. This produced the horst and graben terrain that is centred on Moresby Seamount and continues westward to the north of Normanby Island (Figs 1 –3). Within this faulted terrain, earthquakes indicate that current deformation is localized on the NMG (Fig. 1), probably at least since the nucleation of spreading on segment 1a about 0.1 Ma.
Faulting directly ahead of the westernmost spreading segment NMG has been a focus of MCS surveys and ODP Leg 180 drilling aimed at understanding continental breakup and the evolution of a shallow-angle normal fault. We reproduce here just four of the tens of MCS lines across the structure (Figs 5–7). In addition, a line drawing interpretation of MCS line EW9510-1366, which passes through the primary transect of drill sites, is shown in Figure 4 and all of the EW9510 MCS lines are published (Goodliffe et al. 1999b). The MCS data show that NMG is an asymmetric graben bounded to the south by Moresby Fault and to the north by an antithetic normal fault system. The width of the graben between the sea-floor traces of the bounding faults varies from approximately 12 km NNE of Moresby Seamount to more than 20 km on either side, varying as a function of the curvilinear to zig-zag fault patterns (Figs 2 & 5). MCS line EW9510-1369 crosses the NMG near its narrowest (Fig. 5) and images Moresby Fault continuing at depth beneath the antithetic fault and a downdropped keystone basement block (Fig. 6). Basement drilled on both sides of NMG beneath the rift-onset unconformity (Fig. 4) comprises diabase and, in the footwall of Moresby Fault at ODP Site 1117, gabbro with a crystallization age of 66 + 1.5 Ma (Brooks & Tegner 2001; Monteleone et al. 2001). Heavily faulted Pliocene–Pleistocene synrift sediments cap Moresby Seamount, which rises to almost 150 m bsl (below sea level) (Taylor & Huchon 2002) (Figs 3 & 4). The northern border fault of the NMG is a growth fault, dipping approximately 548S, with basement offsets of about 1.5 km but surface offsets of only approximately 150 m (Figs 4 & 6). Whereas its growth matched that of Moresby Fault early in the development of the rift, it has been far exceeded in the Quaternary by the faults that bound Moresby Seamount (Fig. 4). The NMG is contiguous with the trough containing spreading segment 1a. It shares a common northern border fault with Egum Graben, but has a deeper axis. From the western end of the neovolcanic zone the sea floor climbs 150 m over a distance of approximately 5 km before descending to the deepest part of the NMG at 3250 m (Fig. 5) where
Fig. 5. Bathymetry (a) in the vicinity of the rifting– spreading transition. Key features, including two ring dykes on the seafloor at the western end of segment 1a, are labelled. Cheshire Seamount is a volcanic construct at the eastern end of segment 1a. Sea-floor magnetization (b) for the same area (see Fig. 2 for the colour scale). Profiles along which the horizontal gradient was calculated are shown in grey. Black dots mark the location of gradient maxima. Insets show the magnetization (blue) and horizontal gradient (green) along the profiles A1– A2 and B1 –B2. The separation of the maxima is a measure of the width of the causative body assuming a tabular geometry.
the hanging-wall sediments are tilted south and have collapsed the most (Fig. 4). ODP Site 1108 (Fig. 2) was drilled in the bathymetric saddle on MCS line EW9510-1374 (Fig. 6). It and other drill sites at the foot of the Moresby Fault could not reach the fault at depth owing to the presence of coarse, unconsolidated talus shed from the slope. The synrift sediments north of the NMG are approximately 1 km thick and thin northwards, but those within the graben are more than 2 km thick and subparallel to basement (Figs 4 & 6). The graben basement is almost horizontal on MCS lines EW9510-1369 and EW9510-1374 but dips gently south on MCS lines EW9510-1366 and EW9910-11. Basal onlap is rarely observed. An exception occurs south of SP7200 on MCS
226
A. M. GOODLIFFE & B. TAYLOR
1
200.0
SP 2500
300.0
5000
7500
400.0 10000
500.0
12500
600.0
15000
700.0
17500
20000
EW9510-1369
800.0 900.0 22500 25000 27500 Offset (m)
V.E at 1500 ms-1 = 2.93
Two-way traveltime (seconds)
2
3 Moresby Fault
4
5
6
7 8
N
S 200.0
100.0
SP
Offset (m) 2500
300.0
5000
7500
10000
400.0 12500
500.0 15000
600.0 17500
Site 1108
EW9510-1374
700.0 20000
800.0
22500
25000
V.E at 1500 ms-1 = 2.93
Two-way traveltime (seconds)
2
Deformation zone around antithetic fault
3
Moresby Fault 4
5
6
7
S SP Offset (m) 0 1
N
6700.0
6800.0
2500
5000
6900.0
7000.0
7100.0
7200.0
7300.0
7400.0
7500.0
7500
10000
12500
15000
17500
20000
22500
V.E at 1500 ms-1 = 2.93
EW9910-11
Two-way traveltime (seconds)
2
3
Moresby Fault 4
5
6
7
S
N
Fig. 6. MCS lines EW9510-1369 (top), EW9510-1374 (middle) and EW9910-11 (bottom) across the North Moresby Graben to the west of the rifting– spreading boundary (see Fig. 5 for locations). The MCS lines image the shallow-angle Moresby Fault and the high-angle, graben-bounding, antithetic fault system (on which the profiles are aligned). The blue line marks the base of the synrift sediments and the location of ODP Site 1108 is shown. For clarity, only some of the larger-offset normal faults are drawn. The NMG widens eastward; only the western line, EW9510-1369, images the Moresby Fault passing under the antithetic fault.
Fig. 7. MCS line EW9510-1369 (see Fig. 6a for time section) converted to depth using a velocity model for the synrift sediments derived from ODP results and semblance velocities. A velocity of 4.5 km s21 is assigned to basement at the base of the synrift sediments. Below this a gradient is applied such that the velocity is 6 km s21 at 8 s TWT. In this depth section the apparent dip of Moresby Fault is approximately 298 and of the antithetic fault is 548.
BOUNDARY BETWEEN RIFTING AND SPREADING 227
228
A. M. GOODLIFFE & B. TAYLOR
line EW9910-11 (Fig. 6) where sediment onlap to the north is likely to have been the result of basement tilting to the south associated with early motion on the Moresby Fault. The thick NMG sedimentary fill is cut by myriads of small-offset, north- and south-dipping, high-angle normal faults. For clarity, most of these faults are omitted on the interpretations in Figure 6, but the surface traces of the larger ones are shown in Figure 2. Sediment reflectors lose coherency as Moresby Fault is approached, usually above the point at which the basement reflector intersects the fault. This is probably the result of fault drag, extensional collapse, fluid flow and deposition of slump material. Moresby Fault was first proposed to be active on the basis of teleseismic waveform inversions of normal faulting earthquakes in the vicinity of Moresby Seamount that had shallow-dipping nodal planes (Abers 1991). At least two of the focal mechanisms demonstrate normal faulting at dips of 258– 358, consistent with EW9203 MCS data (Abers et al. 1997). Evidence of water multiples in a water depth of 3 km indicated that the location of one of the largest earthquakes, on a plane dipping 31.68 + 1.38N, is beneath the deepest part of the NMG and coincident with the Moresby Fault. An initial estimate of the dip of Moresby Fault made using MCS lines EW9510-1369 and 1374 was 278 + 38 towards 0158 (Taylor et al. 1999). Velocity data collected by logging and laboratory measurements during ODP Leg 180 allow a more accurate determination of the dip of Moresby Fault. Here we present a depth-converted version of MCS line EW9510-1369 using realistic velocities for the synrift sediments in NMG. This line has a number of desirable characteristics: (1) Moresby Fault and the south-dipping antithetic fault intersect at depth; (2) a distinct and continuous reflector is imaged along the length of the Moresby Fault; and (3) in two-way travel-time both faults have a distinct kink that is due to velocity pull-up where they descend below the keystone basement block that should be removed by depth conversion. Using the velocity model derived in Appendix 2, MCS line EW9510-1369 was converted to depth (Fig. 7). Moresby Fault has an average apparent dip of 298N and that of the antithetic fault is 548S. The average strike of the antithetic fault system is east– west (Figs 2 & 5) and hence its apparent dip on this line is true. In contrast, Moresby Fault has a sinuous fault trace (Figs 2 & 5), depth contours indicate a curviplanar form and varying strike down-dip (Taylor et al. 1999), and the direction of slip may have changed during its long history (Fig. 4). If an upper fault strike of 121.58, determined from 11 200 m-spaced EW9910 MCS lines across the NE face (Fig. 5), is
used then the apparent dip equates to a true dip of 388. However, if the long-term slip paralleled small circles about the main Euler pole of opening (i.e. N78E, Fig. 2) then the Moresby Fault dip in this direction is 29.28, and in the 0158 direction used by Taylor et al. (1999) the dip is 29.98. Fault dips of 308 –338 are compatible with the slip plane of the large earthquake located beneath the NMG (Abers et al. 1997). That the approximately 308 – 338 Moresby Fault is active has been the subject of much controversy. Theory states that if the coefficient of friction is that of typical rock (0.6–0.8) and s1 is near vertical then a normal fault will dip near 608 (Anderson 1951) or 458 for the case of finite slip (Forsyth 1992; Buck 1993). Lower-angle faults either may be explained by rotation of the stress field (Yin 1989; Parson & Thompson 1993), weakening through near lithostatic pore-fluid pressures within the fault zone (Byerlee 1990) or fault gouge materials with a lower coefficient of friction (Hill & Thatcher 1992). A primary goal of ODP Leg 180 was to penetrate Moresby Fault and investigate the mineralogy and stress state to determine why this fault is active. Although the active segment of Moresby Fault was not penetrated, 12 m of talc –chlorite–serpentinite clayey fault gouge with coefficients of friction of 0.21 –0.3 (Kopf et al. 2003) was recovered at Site 1117 along a segment of the fault exposed at the sea floor. Underneath, a poorly recovered approximately 50 m-thick section of mylonites (recording plastic deformation) and ultracataclastites (recording intermittent plastic and brittle deformation) was recovered from the quartz-gabbro basement. Analysis of samples from Site 1117 and other proximal sites has demonstrated the importance of frictional heating and hydrothermal fluid flow in the evolution of the Moresby Fault (Kopf et al. 2003). Channelling of fluids along this fault is evinced by the greater permeability of the fault gouge, especially in the direction parallel to the tectonic fabric, than the overlying sediments (Kemerer & Screaton 2001; Kopf 2001; Stover et al. 2001). The continuation of the fault gouge along the Moresby Fault beneath the NMG is indicated by inverse models of the MCS data that include a 4.3 km s21, approximately 33 m-thick layer along the fault with isolated pockets of a 1.7 km s21 lower velocity material (Floyd et al. 2001). Given the friction coefficients of the fault gouge and the presence of fluids in the fault zone, either of which alone would make motion on the Moresby Fault feasible (Abers 2001), strain accumulation on this fault is no longer an issue. The initiation of motion along an approximately 308 fault remains problematic, for although full flexural modelling of the evolution of rift systems shows the importance of rotation from initially high to low dips (Buck
BOUNDARY BETWEEN RIFTING AND SPREADING
1988), rotating the Moresby Fault to higher dips simultaneously rotates the large, contemporary, southdipping normal fault within Moresby Seamount to low dips (Fig. 4) (Taylor & Huchon 2002). The abundant pre-existing structures in this once convergent regime (Davies & Jaques 1984) may play a role.
Morphology and structure at the point of continental breakup Sea-floor spreading segment 1a occupies the southern half of the eastern NMG and obliquely truncates the SW end of Egum Graben. It is identified by a narrow band of high-amplitude positive magnetization and a coincident region of sea floor with high acoustic backscatter extending from 1518400 E to 151852.250 E (Figs 2 & 5). At the eastern end of the segment, the 8 km-diameter Cheshire Seamount, bisected by an east–west cleft, rises more than 850 m above the surrounding sea floor to a depth of 1750 m. Recent lava flows extend down-slope to the east but are unlikely to be underlain by extensive intrusives given the reduced amplitude of the magnetization. Directly to the west of Cheshire Seamount, and at depths of 3050 and 3250 m, are two ringshaped structures about 4 km in diameter that are in the centre of the Brunhes magnetization high (Figs 2 & 5). Around both constructs, high acoustic backscatter reveals lava flows that have pooled against the base and extended a short distance up the slopes of the surrounding fault-bounded ridges. Hyaloclastites were found in the sediments 5 km to the NW at ODP Site 1108 (Shipboard Scientific Party 1999b). They are high Na2O (4.3%), high TiO2 (2.8%) basaltic andesites (SiO2 ¼ 53%) that are light rare earth element (LREE)-enriched (chondrite-normalized La/Yb ¼ 3.5) (Lackschewitz et al. 2001). Similar rocks have been dredged from Cheshire and Franklin seamounts; the former also includes dacites (Binns & Whitford 1987; Dril et al. 1997). Other rocks from the west end of segment 1a are normal mid-ocean ridge basalts (N-MORBs) with indications of low degrees of melting (at MgO ¼ 8%, Na2O ¼ 3.1%, FeO ¼ 8.1% and TiO2 ¼ 1.14%) (Binns & Whitford 1987). Thus, the mantle sources of segment 1 lavas include N-MORB, arc-like and enriched components (Dril et al. 1997). The extent of the intrusive body that fed the segment 1a lavas can be derived from the magnetization anomaly. Assuming that the intrusion is broadly tabular, the edges of the body will be coincident with peaks in the horizontal derivative of the anomaly (Goodliffe 1993). A series of north–south profiles across the magnetization anomaly were extracted and the horizontal derivative calculated and maxima plotted. Five profiles centred on two locations (A1–A2, B1 –B2: Fig. 5b) were stacked to remove
229
noise before the maximum horizontal derivative was calculated. The horizontal derivative for the western part of segment 1a displays distinct peaks that lie on either side of a body that is between 4.1 and 5.7 km wide. Closer to the Cheshire Seamount the calculated width of the body increases to 7 km, which suggests that segment 1a propagated to the west. Using the current Euler pole and an intrusive body width of 5 km we calculate that sea-floor spreading began on the western part of spreading segment 1a by 0.11 Ma. This is considered an upper limit as some of that width may originate from overprinting rather than strain accommodation. For example, if spreading initiation coincided with the 80 ka reorientation of the entire Woodlark Basin spreading centre (Goodliffe et al. 1997) then only 3.6 km of the intrusion would represent strain accommodation. MCS line EW9910-13 crosses the western end of segment 1a, providing a snapshot of the development of this young accretionary plate boundary (Fig. 8). On this MCS line, the Moresby Fault and the antithetic fault that together bound the NMG are readily identified, as are the thick synrift sediments above the keystone basement block. Between SP8300 and SP8500, however, there is a region of incoherent reflectivity below 5 s two-way time (TWT) against which the Moresby Fault, the synrift sediments and the keystone basement reflectors terminate (Fig. 8). We interpret this incoherent region to be the expression of the Brunhes axial intrusion. It correlates closely with, but is about 800 m north of, the limits of the intrusive body determined from the magnetization anomaly (Figs 5 & 8). Those limits exactly match the position of three sea-floor ridges that, with the bathymetry data, we interpret as volcanic piles above a ring dyke (northern two ridges) plus another dyke to the south, fed from the intrusion below (Fig. 8). Above Moresby Fault at the southern end of MCS line EW9910-13 there are few coherent reflectors, perhaps owing to deposition of sediments in the form of slumps originating on Moresby Seamount (probably similar to talus found at ODP Leg 180 Site 1108) (Shipboard Scientific Party 1999a) (Fig. 8). To the north of the intrusive body the synrift sedimentary sequence is offset at SP7925 by a south-dipping normal fault. This fault is imaged on MCS line EW9910-11 to the west and continues to the east as the major northern bounding fault of Egum Graben. The synrift sequence is interpreted to be continuous with and have a depositional history similar to that sampled to the west during ODP Leg 180 (Shipboard Scientific Party 1999a). The evidence presented above indicates that a major intrusive body underlies the centre of
230
A. M. GOODLIFFE & B. TAYLOR
Fig. 8. Uninterpreted (top) and interpreted version (bottom) of MCS line EW9910-13 (see Fig. 5 for the location). The currently forming continent –ocean boundary involves igneous intrusions penetrating the shallow-angle normal fault of an asymmetric rift graben. Moresby Fault is cut by an approximately 5 km-wide intrusive body that is overlain by subhorizontal sediments that have been cross-cut by a ring dyke.
segment 1a. With a width of 5 km it is assumed that it is today accommodating extension in a similar fashion to a mature spreading centre, although the intrusion has only broken the surface in the form of localized dykes. If the intrusive body is similar to a mature spreading centre it may take the form of a sheeted dyke sequence that is analogous to those observed at ophiolites and inferred to exist below modern spreading centres. Although there is no evidence of an active magma chamber, the sheeted dyke sequence may be underlain by massive
gabbros. Although there are extensive lava flows at the surface, the MCS data suggest that they overlie an approximately 500 m-thick (assuming a velocity of 1800 m s21) extension of the synrift sedimentary sequence seen to the north. As the intrusive body does not reach the surface, extension above it must be transferred into the overlying synrift sediments. The well-stratified reflectors show little evidence of brittle faulting, implying either a dominance of soft-sediment deformation or the offset of extension to either side of the intrusion. Much of the sediment
BOUNDARY BETWEEN RIFTING AND SPREADING
in this bathymetric low may have been deposited rapidly and recently.
Discussion Breakup model We present a simplified model for the formation of the current rift-spreading boundary in the Woodlark Basin (Fig. 9). The initial condition, prior to continental extension, would resemble the central
(a)
50 km
(b)
Trobriand Forearc VE = 4.5
2 0 -2 -4
Woodlark Rise
MS
50 km
Failed Rift
VE = 4.5
2 0 -2 -4
kilometres
Pocklington Rise
Papuan Peninsula today, with an orogenic mountain chain of 1–3 km elevation, crust about 40 km thick, arc volcanoes spaced along the northern coastal plain and a broad shallow-marine forearc basin south of the Trobriand Trench (Taylor et al. 1999). At 8.4 Ma (Taylor & Huchon 2002) this inherently asymmetric region started to undergo north–south extension, lowering the topographic profile (Fig. 9a) and exhuming the metamorphic core complexes of the D’Entrecasteaux Islands (Fig. 1) to the west. Successive generations of
kilometres
Papuan Peninsula
231
(c)
25 km
VE = 1.5
Synrift sediments
kilometres
Tilted Fault Blocks
2 0 -2 -4 -6 -8 -10
North Moresby Graben Moresby Fault Dyke
Antithetic fault
(d)
VE = 1
(e)
VE = 1
Ring dyke
(f )
VE = 1
Fig. 9. The evolution of continental rifting through sea-floor spreading in the western Woodlark Basin. The Papuan Peninsula (topography based on a north– south profile across the Papuan Peninsula at 148.88E: (a) has undergone broadly north–south extension, subsiding to current depths at the longitude of Moresby Seamount (b). Southern graben represent former loci of extension that are now failed rifts. The southern margin is characterized by large fault blocks rotated by high-angle faults. The northern margin has subsided with little evidence of large-offset faults (c). At about 1.2 Ma extension focused on South and North Moresby Graben. More recently, extension has focused at the NMG and, ultimately, to Moresby Fault. A single dyke was intruded into the fault and the overlying sediments causing motion on the up-dip segment of Moresby Fault to cease (d). Progressively growing in width (e), the intrusion is today about 5 km wide but has only reached the surface in the form of localized dykes (f) and flows.
232
A. M. GOODLIFFE & B. TAYLOR
normal faults thinned the palaeo-Papuan Peninsula, resulting in subsidence below sea level to form the Pocklington Rise. To the north, the palaeo-Trobriand forearc (now Woodlark Rise) has tilted to the south and subsided by up to 3 km, reversing the northward-directed Miocene drainage pattern (Fig. 9b). One model for the growth of the metamorphic core complexes in the D’Entrecasteaux Islands (Fig. 1) suggests that, with the aid of fluids liberated from the subducted Solomon Sea Plate, the lower crust of the northern margin is more capable of ductile flow than the margin to the south (Martinez et al. 2001). The paucity of upper crust normal faults in the northern margin implies that crustal thinning must be primarily achieved through means other than brittle faulting. The bulk of the thinning may take place through the removal of the lower crust (Driscoll & Karner 1998; Buck 2004; Kusznir et al. 2004). In contrast, brittle failure of the upper crust is more prevalent to the south. The margin asymmetry need not reflect lithospheric-scale simple shear, but instead is a function of the prerift geology, morphology and rheology. The broadly distributed extension of the palaeoPapuan Peninsula included locally focused extension that produced large graben in the southern margin that are now failed rifts (Figs 1 & 9b). Tilted fault blocks overlain by rotated sediment packages in broadly equal-sized graben (SCS line MW9304-30, Fig. 3) show that extension only focused on the axial graben very recently, even though results from ODP Leg 180 (Taylor & Huchon 2002) show that the latter formed early in the rift history (Fig. 4). A broad heat flow anomaly, that is centred in the northern part of the SMG but also covers the NMG, reflects maximum crustal thinning across this axial system (Goodliffe et al. 2000). Today, as evinced by seismicity (Fig. 1), extension has largely ceased on the southern faults, and is concentrated on the shallow-angle north-dipping Moresby Fault (Taylor & Huchon 2002). MCS lines EW95101366 (Fig. 4), EW9510-1369, EW9510-1374 and EW9910-11 (Figs 6 & 7) reveal the final phase of current rifting prior to breakup. MCS lines EW9510-1366 and EW9910-11 show synrift sediments in the NMG dipping to the south owing to slip on the Moresby Fault, whereas the keystone basement block and overlying sediments in the NMG are less rotated on the intervening lines. The stage where extension is primarily accommodated by accretion is represented by MCS line EW9910-13 (Fig. 8). To reach this stage, we envision that Moresby Fault was cut by a dyke swarm that gradually grew in width (Fig. 9d, e). When the first dykes cut through the fault and were emplaced into the overlying synrift sediments, motion on the up-dip portion of Moresby Fault ceased and/or shifted upwards into the sediments. A transfer fault/zone may have formed to connect accretion on spreading segment 1a to normal
faulting on Moresby Fault in the NMG. The NE flank of Moresby Seamount may be such a transfer fault; its NW–SE trend is consistent with small circles about the current Euler Pole (128S, 1448E, 2.4378 Ma21, Fig. 2b). If the width of the intrusive body represented only strain accommodation through accretion, the Moresby Fault would exit the intrusive body at the same height on the northern side as it entered on the southern side. Although there is some ambiguity in the location of the continuation of Moresby Fault to the north of the intrusion, this is clearly not the case (Fig. 8). Although we maintain that the majority of the intrusive body represents strain accommodation through accretion, which formed the associated magnetic anomaly (Fig. 5), a portion of the width may represent overprinting of the original country rock, perhaps through a process similar to stoping. Alternatively, the intrusion may not be continuous, but instead have intervening unresolved slices of country rock between dykes. For the purposes of our schematic model, we have arbitrarily assigned approximately 25% of the intrusion width to these processes (Fig. 9e, f). As the intrusion grew, sediments and possibly sills were added to the centre of the graben. A ring dyke was intruded and reached the sea floor (Fig. 9e) leading to eruptions that produced the extensive lava flows observed (Fig. 2c). Continued extension, vertical propagation of the intrusive body and horizontal growth of the ring dykes will soon construct an edifice at the sea floor on the western (as already exists on the eastern) part of spreading segment 1a. As shown here, the currently forming continent– ocean boundary involves igneous intrusions penetrating the principal (and shallow-angle) normal fault of an asymmetric rift graben. Further east in the Woodlark Basin, breakup along spreading segment 2 occurred within symmetric graben bounded by high-angle faults (Taylor et al. 1999). But in each case the magnetic, bathymetric and seismic data reveal that the boundary between continental rifting and sea-floor spreading in the Woodlark Basin is locally very sharp—notwithstanding that regionally the processes overlap in time. This is in stark contrast to breakup of the Newfoundland and Iberia conjugate margins associated with ultra-slow spreading in the early north Atlantic where transition zones up to 150 km wide separate clearly defined oceanic and continental crust (Tucholke et al. 2004).
Margin stratigraphy and unconformities MCS imaging calibrated by ODP drilling results allow us to observe the stratigraphic details of continental breakup in the western Woodlark Basin. How do these images and results compare to stratigraphic models of continental margin development?
BOUNDARY BETWEEN RIFTING AND SPREADING
Falvey (1974) proposed a model to explain typical patterns of margin uplift and subsidence as a function of the thermal, erosional and metamorphic processes attendant on splitting a continent by emplacing and then cooling asthenosphere. He observed that two angular unconformities commonly separate pre-, syn- and post-rift sediments. He proposed that ‘the breakup unconformity is caused by erosion during the final uplift pulse associated with prebreakup upwelling in the mantle’. He recognized that ‘this unconformity is more localized than the rift-onset unconformity, being difficult to define in troughs’ and that ‘continental breakup and the onset of sea-floor spreading is probably not instantaneous’. The existence of two unconformities bounding synrift strata has found widespread application in rifted margin studies (Driscoll et al. 1995) even though, in the particular example Falvey used to illustrate the general model (the Otway Basin of South Australia), the proposed rift-onset unconformity actually correlates with the time of breakup and initiation of marine sedimentation, and the real rift-onset unconformity occurred much earlier (Hegarty et al. 1988). In the case of the western Woodlark Basin, there is a regional rift-onset unconformity. It truncated mafic igneous basement at ODP sites 1109 and 1118, produced a 4 Ma hiatus in the shallow-marine forearc basin sediments at ODP Site 1115 (Figs 3 & 4), and generated the late Miocene ‘unconformity A’ in the Nubiam-1 well and region north of the D’Entrecasteaux Islands (Francis et al. 1987; Taylor & Huchon 2002). The drilling and MCS studies show the reversal of the northward Miocene forearc drainage as the Papuan continent was stretched and thinned, and the northern margin (Woodlark Rise) progressively subsided southward toward the active rifts, from paralic to shelf to the present bathyal water depths (Fig. 4) (Taylor & Huchon 2002). Concurrently, the former landmass to the south was faulted and thinned to produce the graben and reefs of the Pocklington Rise (Fig. 1). Early synrift sediments were subsequently tilted on rotating fault blocks, but later sediments of the southern margin were ponded in half-graben as the early rifts in the south were abandoned, even as deformation continued further north focused on the axial rifts (Figs 3 & 9). Therefore ‘synrift’ sediments (those deposited between rift onset and breakup) can be characterized by rotated sections (with parallel or diverging reflectors) as well as by ponded sections that onlap bounding fault blocks where local extension has ceased. Splayed reflectors representing sediment deposition on rotating fault blocks do not characterize the full synrift section, but only those times and places where extension is localized. Also, in the common case where breakup propagates, the time represented by the synrift sediments and the timing of any breakup unconformity will vary
233
along strike: by 6 Ma in the Woodlark Basin. Furthermore, as Falvey (1974) noted, sedimentation in the deep water parts of margins is likely to be conformable across the time of breakup. In the Woodlark Basin, the progression from rifting to breakup and then initial spreading is characterized by a decrease in sedimentation as the margins are progressively thinned, subside below sea level, are eroded less and trap sediments in proximal basins (Figs 3 & 4). The post-rift sedimentation near the continent– ocean boundary is hemipelagic and drapes breakup topography. A distinct breakup unconformity is not observed.
Conclusions † The continental margins asymmetry at the point of current breakup – high-angle normal faults bounding tilted crustal blocks in the south in contrast to a wide tilted margin in the north with only small-offset normal faults – cannot be explained by the observed shallow-angle northdipping detachment between them. Rather, the margins asymmetry is a function of prerift geology, morphology and rheology. Removal of the lower crust through ductile flow is the preferred mode of crustal thinning of the Woodlark Rise, the palaeo-Trobriand forearc. In contrast, on the Pocklington Rise to the south, where the high mountains of the Papuan Peninsula existed 8.4Ma ago, brittle faulting of the upper crust is more prevalent. † Initially, major upper crust normal faulting was distributed from the southern margin to what is now South and North Moresby Graben. Sequentially, the locus of faulting narrowed northwards, leaving large graben (foci of former extension that are now failed rifts) and tilted fault blocks in the southern margin. † At about 1.2 Ma, extension focused at NMG and SMG where rapid subsidence was accompanied by rapid uplift of the intervening Moresby Seamount. This was a precursor to the initiation of sea-floor spreading at approximately 0.8 Ma on segment 1. Extension ceased on the large faults bounding tilted blocks directly to the south. † Sea-floor spreading on segment 1a initiated at approximately 0.08 Ma in the along-strike continuation of the NMG and more specifically through the north-dipping Moresby Fault. Ahead of the spreading tip, extension has focused on the active Moresby Fault (Abers et al. 1997; Taylor & Huchon 2002). Although motion on this normal fault was once controversial owing to its low dip, fault gouge with a low coefficient of friction and evidence of high pore fluid pressures make motion on this fault unremarkable (Taylor & Huchon 2002). Depth conversions of MCS line EW9510-1369 based
234
A. M. GOODLIFFE & B. TAYLOR
on ODP results and semblance velocities (Goodliffe et al. 1999b) indicate that the true fault dip is 298 and compatible with an earthquake slip plane of 308 –338 (Abers et al. 1997). † The normal faults bounding the northern side of spreading segment 1, co-polar with great circles about the pre-0.52 Ma Euler pole (Fig. 2), are crossed by the Egum Graben. The faults bounding the northern side of Egum Graben form the antithetic faults in the NMG and may be continuous with a major transform fault system that separates the Solomon Sea and Woodlark lithosphere to the east. If so, this would make the current rifting–spreading transition a triple junction between the Trobriand, Woodlark and Australian plates. Alternatively, the fact that the bounding faults of Egum Graben are co-polar with great circles about the present-day Euler pole suggests the possibility that this graben represents a new rifting trend, perhaps reactivating pre-existing weaknesses. † MCS line EW9910-13 records an early evolutionary step in the initiation of a spreading centre. An intrusion that cuts an along-strike continuation of Moresby Fault is connected to the surface via a ring dyke through which lavas have erupted onto the sea floor. Thus, the currently forming continent–ocean boundary involves igneous intrusions penetrating the principal (shallowangle) normal fault of an asymmetric rift graben. The magnetic, bathymetric and seismic data reveal that the boundary between continental rifting and sea-floor spreading in the Woodlark Basin is spatially very sharp – notwithstanding that temporally the processes overlap. † Synrift sediments (those deposited between rift onset and breakup) can be characterized by rotated and/or diverging sections, as well as by ponded sections that onlap bounding fault blocks where locally extension has ceased. A breakup unconformity is not generally observed at the deep-water boundary between rifted continental and accreted oceanic crust, but rather there is a decrease in sedimentation as the margins are progressively thinned, subside below sea level, are eroded less and trap sediments in proximal basins.
Appendix 1: Data and processing The data presented herein come primarily from three marine geophysical surveys conducted between 1993 and 1999 (Goodliffe et al. 1999b) and include singlechannel seismic (SCS) data, multichannel seismic (MCS) data, swath bathymetry, acoustic imagery and magnetic data. Generic mapping tools (GMT) (Wessel & Smith 1995) was used comprehensively throughout this study both for display and analysis.
Bathymetry and acoustic imagery data HAWAII-MR1 swath bathymetry and sidescan data from a 1993 R/V Moana Wave survey cover the western Woodlark Basin from 1508 500 E to 1548 450 E. Hydrosweep multibeam bathymetry data from R/V Maurice Ewing cruises in 1995 and 1999 provided additional detail in the region of Moresby Seamount. Beyond the coverage of the swath datasets, wide-beam profiler data from the National Geophysical Data Center (NGDC) database and digitized soundings from Australian Navy charts were incorporated. Land topographic data are from the Shuttle Radar Topography Mission (SRTM) 90 m grid cell dataset. Final bathymetry grids were produced with a 0.0018 grid size.
Magnetic data Magnetic data were collected along north–south 5 nm-spaced tracks during the regional 1993 R/V Moana Wave survey and during R/V Maurice Ewing surveys in 1995, 1999 and 2000 in the vicinity of Moresby Seamount. As the magnetic anomalies formed along an east–west spreading centre are highly skewed, the shipboard magnetic and bathymetry data were gridded at 0.018 and used to derive a solution for the intensity of magnetization of the sea floor following the technique of Macdonald et al. (1980). The magnetization inversion assumes a 1 km-thick source layer conforming to the sea floor. As this solution produced a reversal pattern in qualitative agreement with that from the total field anomalies, no annihilator (Parker & Huestis 1974) was added to the solution. Areas shallower than 1.2 km were masked to satisfy the requirement that the magnetic source does not come within one grid cell of the observation surface (Parker & Huestis 1974). Although the resulting magnetization map can be interpreted directly using the geomagnetic polarity timescale (Cande & Kent 1995), the oceanic crust in the region of interest is mostly formed during the Brunhes chron.
Marine seismic data The marine seismic data used in this study comes from three surveys. In 1993 six-channel seismic (SCS) data were collected during an R/V Moana Wave survey in the Western Woodlark Basin using a 150 m-streamer and a 120 cubic inch (inch3) airgun source. Shots spaced at 37.5 m yielded twofold 12.5 m-spaced CMPs. A 1995 R/V Maurice Ewing survey included a site survey for ODP Leg 180 (Taylor et al. 1996; Goodliffe et al. 1999b). To better image the Moresby Fault and document the subsidence of the northern margin of the Woodlark Basin, a 196-channel 4.9 km-streamer was deployed. The first line, designed for deepest penetration, was shot using a tuned 8460 inch3 20 airgun array. Shots every 20 s yielded a shot spacing of approximately 50 m and 48-fold 12.5 m-spaced CMPs, a record length of
BOUNDARY BETWEEN RIFTING AND SPREADING 16.384 s and a sample interval of 4 ms. Additional lines were shot using 10 airguns alternating in the same 20 airgun array. Shots every 12 s yielded a shot spacing of approximately 25 m and 60–75-fold 12.5 m-spaced CMPs with a record length of 9.116 s and a sample interval of 4 ms. In 1999 the R/V Maurice Ewing shot 13 dip lines oriented NNE– SSW, perpendicular to the strike of the fault (Taylor et al. 1999) and one strike line. Eleven of the dip lines, spaced 200 m apart, crossed Moresby Graben in the vicinity of ODP Site 1108. The two remaining dip lines crossed the western tip of the spreading centre to the east. Using a 48-channel 1.2 km-streamer, a 1395 inch3 tuned six airgun array was fired every 25 m, giving a 24-fold CMP interval of 12.5 m. The record length was 8192 ms and sample interval was 2 ms. The marine seismic data were processed using ProMAX. Data processing included trace editing, filtering, prestack and post-stack deconvolution, velocity analysis, inside and outside mutes, normal moveout corrections, stack and migration. The multichannel seismic (MCS) data presented herein have also undergone partial prestack migration.
Appendix 2: Velocity model and depth conversion Four sites were drilled and logged in the synrift sedimentary section. At Site 1108, sampled to a depth of 480 m, density and porosity measurements showed that 385 m of sediments have been removed from the top of the section through faulting and/or erosion (Shipboard Scientific Party 1999b). The termination of reflectors at the sea floor south of SP420 on MCS line EW9510-1374 and south of SP400 on MCS line EW9510-1369 (Fig. 6) confirms that this is a reasonable interpretation and any synrift sediment velocity model must take this into account. To the north of the NMG logging and laboratory measurements of velocity were obtained at sites 1109, 1115 and 1118. Depth v. velocity curves were derived for each of the three sites, refined using vertical seismic profile checkshots, and verified through the generation of synthetic seismograms (Goodliffe et al. 2001). Site 1118, on the margin just north of the NMG (Figs 2 & 4), sampled 860 m of synrift sediments. Numerous MCS lines (Goodliffe et al. 1999b) show that the sediments sampled on the northern margin can be correlated with those in NMG close to MCS lines EW9510-1369 and EW9510-1374. Here, the offset across the northern bounding fault is small, enabling the velocity– depth curve for Site 1118 to be used in the graben. As we only have synrift sediment velocity information to a depth of 801 m, the best estimate of interval velocity for deeper synrift sediments in the graben comes from semblance analysis on line EW9510-1369 where the strata are continuous and subhorizontal. The composite interval velocity curve predicts a velocity of approximately 3.5 km s21 at about 2.5 km below the sea floor (the approximate sediment thickness in the graben). Shipboard laboratory measurements of the basement
235
diabase consistently gave velocities of approximately 6 km s21 (Shipboard Scientific Party 1999a). As fracture porosity of this lithology is often high, we assigned a velocity of 4.5 km s21 to the basement at the base of the synrift sediments and applied a gradient such that the velocity would be 6 km s21 at 8 s TWT directly below the centre of the graben. Extrapolating our one-dimensional velocity model to two dimensions in most places entailed hanging the sediment and basement velocity functions from the sea floor and basement, respectively. However, south of SP400, where reflectors terminate at the sea floor, the composite velocity curve was suspended at 3.54 s TWT, effectively increasing the sediment velocity at the water–sediment interface and closely matching the velocities of the sediments recorded at Site 1108. The water column was assigned an average velocity of 1494 m s21.
References A BERS , G. A. 1991. Possible seismogenic shallowdipping normal faults in the Woodlark– D’Entrecasteaux extensional province, Papua New Guinea. Geology, 19, 1205–1208. A BERS , G. A. 2001. Evidence for seismogenic normal faults at shallow dips in continental rifts. In: WILSON, R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: A Comparison of Evidence From Land and Sea. Geological Society, London. Special Publications, 187, 305–318. A BERS , G. A., M UTTER , C. Z. & F ANG , J. 1997. Shallow dips of normal faults during rapid extension: Earthquakes in the Woodlark– D’Entrecasteaux rift system, Papua New Guinea. Journal of Geophysical Research, 102(B7), 15,301– 15,318. A NDERSON , E. M. 1951. The Dynamics of Faulting. Oliver & Boyd, London. B ASSI , G., K EEN , C. E. & P OTTER , P. 1993. Contrasting styles of rifting: Models and examples from the eastern Canadian margin. Tectonics, 12, 639–655. B INNS , R. A. & W HITFORD , D. J. 1987. Volcanic rocks from the western Woodlark Basin, Papua New Guinea. Proceedings of the Pacific Rim Congress, 87, 525– 534. B ONATTI , E. 1985. Punctiform initiation of seafloor spreading in the Red Sea during transition from a continental to an oceanic rift. Nature, 316, 33– 37. B ROOKS , K. & T EGNER , C. 2001. Affinity of the Leg 180 dolerites of the Woodlark Basin: geochemistry and age. In: H UCHON , P., T AYLOR , B. & K LAUS , A. (eds) Proceeding of the Ocean Drilling Program, Scientific Results, Leg 180, 1 –18. [CD-ROM] Ocean Drilling Program, College Station, TX. B UCK , W. R. 1988. Flexural rotation of normal faults. Tectonics, 7, 959– 973. B UCK , W. R. 1993. Effect of lithospheric thickness on the formation of high- and low-angle normal faults. Geology, 21, 933– 936. B UCK , W. R. 2004. Consequences of asthenospheric variability on continental rifting. In: K ARNER , G. D.,
236
A. M. GOODLIFFE & B. TAYLOR
T AYLOR , B., D RISCOLL , N. W. & K OHLSTEDT , D. L. Rheology and Deformation of the Lithosphere at Continental Margins. Columbia University Press, New York, 1– 30. B YERLEE , J. D. 1990. Friction, overpressure and fault normal compression. Geophysical Research Letters, 17, 2109–2112. C ANDE , S. C. & K ENT , D. V. 1995. Revised calibration of the geomagnetic polarity timescale for the Late Cretaceous and Cenozoic. Journal of Geophysical Research, 100(B4), 6093– 6095. C OCHRAN , J. R. 1981. The Gulf of Aden: Structure and evolution of a young ocean basin and continental margin. Journal of Geophysical Research, 86, 263– 287. C OCHRAN , J. R. 2005. Northern Red Sea: Nucleation of an oceanic spreading center within a continental rift. Geochemistry Geophysics Geosystems, 6, Q03006, doi:10.1029/2004GC000826. C OCHRAN , J. R. & M ARTINEZ , F. 1988. Evidence from the northern Red Sea on the transition from continental to oceanic rifting. Tectonophysics, 153, 25– 53. C ORTI , G., V AN W IJK , J., B ONINI , M., S OKOUTIS , D., C LOETINGH , S., I NNOCENTI , F. & M ANETTI , P. 2003. Transtition from continental breakup to punctiform seafloor spreading: How fast, symmetric and magmatic. Geophysical Research Letters, 30, 6-1–6-4. C OURTILLOT , V. 1982. Propagating rifts and continental breakup. Tectonics, 1, 239– 250. D AVIES , H. L. & J AQUES , J. L. 1984. Emplacement of ophiolite in Papua New Guinea. In: G ASS , I. G., L IPPARD , S. J. & S HELTON , A. W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications, 13, 341 –350. D AVIES , H. L., S YMONDS , P. A. & R IPPER , I. D. 1984. Structure and evolution of the southern Solomon Sea region. BMR Journal of Australian Geology and Geophysics, 9, 49–68. D RIL , S. I., K UZMIN , M. I., T SIPUKOVA , S. S. & Z ONENSHAIN , L. P. 1997. Geochemistry of basalts from the western Woodlark, Lau and Manus basins: implications for their petrogenesis and source rock compositions. Marine Geology, 142, 57–83. D RISCOLL , N. W. & K ARNER , G. D. 1998. Lower crustal extension across the Northern Carnarvon basin, Australia: Evidence for an eastward dipping detachment. Journal of Geophysical Research, 103(B3), 4975–4991. D RISCOLL , N. W., H OGG , J. R., C HRISTIE -B LICK , N. & K ARNER , G. D. 1995. Extensional tectonics in the Jeanne d’Arc Basin, offshore Newfoundland: Implications for the timing of breakup between Grand banks and Iberia. In: S CRUTTON , R. A., S TOKER , M. S., S HIMMIELD , G. B. & T UDHOPE , A. W. (eds) Sedimentation and Paleogeography of the North Atlantic Region. Geological Society, London, Special Publications, 90, 1 –28. D UNBAR , J. A. & S AWYER , D. S. 1996. Threedimensional dynamical model of continetal rift propagation and margin plateau formation. Journal of Geophysical Research, 101(B12), 27,845–27,863.
F ALVEY , D. A. 1974. The development of continental margins in plate tectonic theory. The APEA Journal, 14, 95–106. F ANG , J. 2000. Styles and distribution of continental extension derived from the rift basins of Eastern Papua New Guinea. Ph.D. thesis, Columbia University, New York. F ERRIS , A., Z ELT , B. C., A BERS , G. A., F LOYD , J. S., T AYLOR , B., M UTTER , J. C. & L ERNER -L AM , A. 2002. Seismicity of the Woodlark–D’Entrecasteaux Rift System in Eastern Papua New Guinea. Eos Transactions of the American Geophysical Union, 83, Fall Abstract S71C-1122. Meeting Supplement. F LETCHER , J. M. & M UNGUIA , L. 2000. Active continental rifting in southern Baja California, Mexico; implications for plate motion partitioning and the transition to seafloor spreading in the Gulf of California. Tectonics, 19, 1107–1123. F LOYD , J. S., M UTTER , J. C., G OODLIFFE , A. M. & T AYLOR , B. 2001. Evidence for fault weakness and fluid flow within an active low-angle normal fault. Nature, 411, 779– 783. F ORSYTH , D. W. 1992. Finite extension and low angle normal faulting. Geology, 20, 27– 30. F RANCIS , G., L OCK , J. & O KUDA , Y. 1987. Seismic stratigraphy and structure of the area to the southeast of the Trobriand Platform. Geo-Marine Letters, 7, 121–128. G OODLIFFE , A. M. 1993. A gravity and magnetic study in the vicinity of Novarupta, Katmai National Park, Alaska. M.S. Thesis, University of Alaska Fairbanks, Fairbanks. G OODLIFFE , A. M. 1998. The rifting of continental and oceanic lithosphere: observations from the Woodlark Basin. PhD thesis, University of Hawaii at Manoa, Honolulu. G OODLIFFE , A. M., T AYLOR , B., M ARTINEZ , F., H EY , R., M AEDA , K. & O HONO , K. 1997. Synchronous reorientation of the Woodlark Basin spreading center. Earth and Planetary Science Letters, 146, 233–242. G OODLIFFE , A. M., M ARTINEZ , F. & T AYLOR , B. 1999a. A reconstruction of the margins of the Woodlark Basin: Predictions of pre-rifting continental thickness. Eos, Transactions of the American Geophysical Union, 84(46/Supplement), 1064. G OODLIFFE , A., T AYLOR , B. & M ARTINEZ , F. 1999b. Data Report: Marine geophysical surveys of the Woodlark Basin region. In: H UCHON , P. & T AYLOR , B. (eds) Initial Reports of the Ocean Drilling Program, Leg 180, 1–20.[CD-ROM]Ocean Drilling Program, College Station, TX. G OODLIFFE , A. M., M ARTINEZ , F., T AYLOR , B., L EWIS , T. J., T AYLOR , A. E. & S CREATON , E. J. 2000. The geothermal signature of continental breakup: results from the Woodlark Basin. Eos, Transactions of the American Geophyiscal Union, 81, F1114. G OODLIFFE , A. M., T AYLOR , B. & K ARNER , G. D. 2001. Correlations between seismic, sogging and core data from ODP Leg 180 sites in the western Woodlark Basin. In: H UCHON , P., T AYLOR , B. & K LAUS , A. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, Leg 180, 1– 25. [CD-ROM] Ocean Drilling Program, College Station, TX.
BOUNDARY BETWEEN RIFTING AND SPREADING H ARRY , D. L. & B OWLING , J. C. 1999. Inhibiting magmatism on nonvolcanic rifted margins. Geology, 27, 895–898. H AYWARD , N. J. & E BINGER , C. J. 1996. Variations in the along-axis segmentation of the Afar Rift system. Tectonics, 15, 244–257. H EGARTY , K. A., W EISSEL , J. K. & M UTTER , J. C. 1988. Subsidence history of Australia’s southern margin; constraints on basin models. AAPG Bulletin, 72, 615–633. H ILL , D. P. & T HATCHER , W. 1992. An energy constraint or frictional slip in misoriented faults. Bulletin of the Seismological Society of America, 82, 883–897. H ILL , E. J. 1995. Extensional deformation of continental crust prior to breakup and sea-floor spreading: A study of Misima Island, Papua New Guinea. Geological Society of Australia Abstract, 40, 68– 69. H OLBROOK , W. S. & K ELEMEN , P. B. 1993. Large igneous province on the US Atlantic margin and implications for magmatism during continental breakup. Nature, 364, 433–436. J ONGSMA , D. 1972. Marine geology of Milne Bay, eastern Papua New Guinea. Bureau of Mineral Resources of Australia Bulletin, 125. K ARNER , G. D. 2005. Structural and depositional styles of the syn-rift systems of the West African and Brazilian margins: The need for new quantitative approaches for modeling basin subsidence. In: Return to Rifts – The Next Wave? Fresh Insights into the Petroleum Geology of Global Rift Basins. Geological Society, London. K ARNER , G. D., D RISCOLL , N. W. & B ARKER , D. H. N. 2003. Syn-rift regional subsidence across the West African continental margin: the role of lower plate ductile extension. In: A RTHUR , T. J., M ACGREGOR , D. S. & C AMERON , N. R. (eds) Petroleum Geology of Africa: New Themes and Developing Technologies. Geological Society, London, Special Publications, 207, 105– 129. K ARNER , G. D., M ANATSCHAL , G. & P INHEIRO , L. M. 2007. Imaging, mapping and modelling continental lithosphere extension and breakup: An introduction. In: K ARNER , G. D., M ANATSCHAL , G. & P INHEIRO , L. M. (eds) Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, London, Special Publications, 282, 1– 8. K EMERER , A. & S CREATON , E. 2001. Permeabilities of sediments from Woodlark Basin: implications for pore pressures. In: H UCHON , P., T AYLOR , B. & K LAUS , A. (eds) Proceeding of the Ocean Drilling Program, Scientific Results, Leg 180, 1–14. [CDROM] Ocean Drilling Program, College Station, TX. K OPF , A. 2001. Permeability variation across an active low-angle detachment fault, western Woodlark Basin (ODP Leg 180), and its implication for fault activiation. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: A Comparison of Evidence From Land and Sea. Geological Society, London, Special Publications, 187, 23–41. K OPF , A., B ERMANN , J. H., D EYHLE , A., R OLLER , S. & E RLENKEUSER , H. 2003. Isotopic evidence (B, C, O) of deep fluid processes in fault rocks from the active Woodlark Basin detachment zone. Earth and Planetary Science Letters, 208, 51– 68.
237
K USZNIR , N., H UNSDALE , R. & R OBERTS , A. M. 2004. Timing of depth-dependent lithosphere stretching on the S. Lofoten rifted margin offshore mid-Norway: pre-breakup or postbreakup? Basin Research, 16, 279– 296. L ACKSCHEWITZ , K. S., B OGAARD , P. V. D. & M ERTZ , D. F. 2001. 40Ar/39Ar ages of fallout tephra layers and volcaniclastic deposits in the sedimentary succession of the western Woodlark Basin, Papua New Guinea; the marine record of Miocene– Pleistocene volcanism. In: W ILSON , R. C. L., W HITMARSH , R. B., T AYLOR , B. & F ROITZHEIM , N. (eds) Non-volcanic Rifting of Continental Margins: A Comparison of Evidence From Land and Sea. Geological Society, London, Special Publications, 187, 373–388. L AVIER , L. L. & B UCK , R. 2002. Half graben versus large-offset low-angle normal fault: Importance of keeping cool during normal faulting. Journal of Geophysical Research, 107, 8-1–8-16. L ISTER , G. S., E THERIDGE , M. A. & S YMONDS , P. A. 1991. Detachment models for the formation of passive continental margins. Tectonics, 10, 1038–1064. M ACDONALD , K. C., M ILLER , S. P., H UESTIS , S. P. & S PIESS , F. N. 1980. Three-dimensional modeling of a magnetic reversal boundary from inversion of deep-tow measurements. Journal of Geophysical Research, 85(B7), 3670–3680. M ARTIN , A. K. 1984. Propagating rifts; Crustal extension during continental rifting. Tectonics, 3, 611–617. M ARTINEZ , F. & C OCHRAN , J. R. 1988. Structure and tectonics of the northern Red Sea: Catching a continental margin between rifting and drifting. Tectonophysics, 150, 1– 32. M ARTINEZ , F., T AYLOR , B. & G OODLIFFE , A. 1999. Contrasting styles of seafloor spreading in the Woodlark Basin: Indications of rift-induced secondary mantle convection. Journal of Geophysical Research, 104(B6), 12,909– 12,926. M ARTINEZ , F., G OODLIFFE , A. M. & T AYLOR , B. 2001. Metamorphic core complex formation by density inversion and lower-crust extrusion. Nature, 411, 930– 934. M ONTELEONE , B., B ALDWIN , S. L., I RELAND , T. R. & F ITZGERALD , P. G. 2001. Thermochronologic constraints for the tectonic evolution of the Moresby Seamount, Woodlark Basin, Papua New Guinea. In: H UCHON , P., T AYLOR , B. & K LAUS , A. (eds) Proceeding of the Ocean Drilling Program, Scientific Results, Leg 180, 1– 35. [CD-ROM] Ocean Drilling Program, College Station, TX. M OORE , D. G. 1973. Plate-edge deformation and crustal growth, Gulf of California structural province. Geological Society of American Bulletin, 84, 1883– 1906. M UTTER , J. C., M UTTER , C. Z. & F ANG , J. 1996. Analogies to oceanic behavior in the continental breakup of the western Woodlark basin. Nature, 380, 333– 336. P ARKER , R. L. & H UESTIS , S. P. 1974. The inversion of magnetic anomalies in the presence of topography. Journal of Geophysical Research, 79, 1587–1593. P ARSON , T. & T HOMPSON , G. A. 1993. Does magmatism influence low-angle normal faulting? Geology, 21, 247– 250.
238
A. M. GOODLIFFE & B. TAYLOR
R OBERTSON , A. H. F. & S HARP , T. R. 2002. Geochemical and mineralogical evidence for the provenance of mixed volcanogenic/terrigenous hemipelagic sediments in the Pliocene–Pleistocene Woodlark Backarc Rift Basin, southwest Pacific: Ocean Drilling Program Leg 180. In: H UCHON , P., T AYLOR , B. & K LAUS , A. (eds) Proceeding of the Ocean Drilling Program, Scientific Results, Leg 180, 1– 53. [CDROM] Ocean Drilling Program, College Station, TX. R OSENDAHL , B. R. 1987. Architecture of continental rifts with special reference to East Africa. Annual Reviews of Earth and Planetary Sciences, 15, 445– 503. S HARP , T. R. & R OBERTSON , A. H. F. 2002. Petrography and provenance of volcaniclastic sands and sandstones recovered from the Woodlark Rift Basin and Trobriand Forearc Basin, Leg 180. In: H UCHON , P., T AYLOR , B. & K LAUS , A. (eds) Proceeding of the Ocean Drilling Program, Scientific Results, Leg 180, 1 –58. [CD-ROM] Ocean Drilling Program, College Station, TX. SHIPBOARD SCIENTIFIC PARTY . 1999a. Leg 180 Summary. In: T AYLOR , B., H UCHON , P. & K LAUS , A. (eds) Proceedings of the Ocean Drilling Program, Initial Reports, Leg 180, Ocean Drilling Program, College Station, TX, 1– 77. SHIPBOARD SCIENTIFIC PARTY . 1999b. Site 1108. In: T AYLOR , B., H UCHON , P. & K LAUS , A. (eds) Proceedings of the Ocean Drilling Program, Initial Reports, Leg 180. Ocean Drilling Program, College Station, TX, 1– 134. S TOVER , S. C., S CREATON , E. J., L IKOS , W. J. & G E , S. 2001. Data Report: Hydrologic characteristics of shallow marine sediments of Woodlark Basin, Site 1109. In: H UCHON , P., T AYLOR , B. & K LAUS , A. (eds) Proceeding of the Ocean Drilling Program, Scientific Results, Leg 180, 1 –22. [CD-ROM] Ocean Drilling Program, College Station, TX. T AKAHASHI , K., C ORTESE , G. ET AL . 2002. Summary of revised age assignments for ODP Leg 180. In: H UCHON , P., T AYLOR , B. & K LAUS , A. (eds) Proceeding of the Ocean Drilling Program, Scientific Results, Leg 180, 1 –13. [CD-ROM] Ocean Drilling Program, College Station, TX. T AYLOR , B. & E XON , N. F. 1987. An investigation of ridge subduction in the Woodlark –Solomons region: Introduction and overview, In: T AYLOR , B. & E XON , N. F. (eds) Marine Geology, Geophysics and
Geochemistry of the Woodlark Basin–Solomon Islands. Circum-Pacific Council for Energy and Mineral Resources Earth Science Series, Houston, TX, 1– 24. T AYLOR , B. & H UCHON , P. 2002. Active continental extension in the western Woodlark Basin: a synthesis of Leg 180 results. In: H UCHON , P., T AYLOR , B. & K LAUS , A. (eds) Proceeding of the Ocean Drilling Program, Scientific Results, Leg 180. Ocean Drilling Program, College Station, TX, 1– 36. T AYLOR , B., G OODLIFFE , A., M ARTINEZ , F. & H EY , R. N. 1995 Continental rifting and initial seafloor spreading in the Woodlark Basin. Nature, 374, 534–537. T AYLOR , B., M UTTER , C., G OODLIFFE , A. & F ANG , J. 1996. Active continental extension: the Woodlark basin. JOI/USSAC Newsletter, 9, 1 –4. T AYLOR , B., G OODLIFFE , A. M. & M ARTINEZ , F. 1999. How continents break up: insights from Papua New Guinea. Journal of Geophysical Research, 104(B4), 7497– 7512. T UCHOLKE , B. E., S IBUET , J.-C. ET AL . 2004. Margins of the Newfoundland–Iberia Rift; ODP Leg 210 explores the Newfoundland Basin. JOIDES Journal, 30, 20–24. V AN W IJK , J. W. & B LACKMAN , D. K. 2005. Dynamics of continental rift propagation: the end-member modes. Earth and Planetary Science Letters, 229, 247–258. V INK , G. E. 1982. Continental rifting and the implications for plate tectonic reconstructions. Journal of Geophysical Research, 87(B13), 10,677– 10,688. W ALLACE , L. M., S TEVENS , C. ET AL . 2004. GPS and seismological constraints on active tectonics and arccontinent collision in Papua New Guinea: Implications for mechanics of microplate rotations in a plate boundary zone. Journal of Geophysical Research, 109, B05404, doi:10.1029/2003JB002481. W ERNICKE , B. 1985. Uniform-sense normal simple shear of the continental lithosphere. Canadian Journal of Earth Sciences, 22, 108– 125. W ESSEL , P. & S MITH , W. H. F. 1995. New version of the generic mapping tools released. Eos, Transactions American Geophysical Union, 76(33), 329. Y IN , A. 1989. Origin of regional rooted low-angle normal faults: A mechanical model and its tecotonic implications. Tectonics, 8, 469– 482.
Nature of the continent – ocean transition zone along the southern Australian continental margin: a comparison of the Naturaliste Plateau, SW Australia, and the central Great Australian Bight sectors N. G. DIREEN1, I. BORISSOVA2, H. M. J. STAGG2, J. B. COLWELL2 & P. A. SYMONDS2 1
Continental Evolution Research Group, University of Adelaide, Adelaide, SA 5005, Australia (e-mail:
[email protected]) 2
Geoscience Australia, Canberra, ACT 2601, Australia
Abstract: We document the interpretation of three crustal sections from coincident deep seismic reflection, gravity and magnetic data acquired on Australia’s southern margin: one section from the Naturaliste Plateau and the Diamantina Zone; and two in the Great Australian Bight (GAB). Interpretations are based on an integrated study of deep multichannel seismic, gravity and magnetic data, together with sparse sonobuoy and dredging information. All interpreted sections of the margin show a transition from thinned continental crust, through a wide continent ocean transition zone (COTZ). In the GAB the transition is to slow sea-floor spreading oceanic crust that dates from breakup in the Campanian (c. 83 Ma); in the Naturaliste–Diamantina margin the earliest oceanic crust is undated. The COTZ on these margins is geologically and geophysically complex, but interpretation of all data, including dredge hauls, is consistent with the presence of a mixture of modified continental lower crust, breakup related volcanics and exhumed continental mantle. Serpentinized detachment faults are not well imaged, but have been inferred from high-amplitude magnetic signatures interpreted to arise from magnetite associated with the hydration of peridotites. Alternative models for the structure of the COTZ, involving either mafic underplating or aborted sea-floor spreading, have been explored, but are considered unlikely on this margin. Similarity in the final architecture of these margins has major implications for the nature of rifting in the Southern Rift System, and may point to the entire 4000 km-long system being non-volcanic in character. Second-order differences in geometry and morphology of the two areas studied are unlikely to be a function of strain rate. Instead, they probably reflect complexities owing to the multiple tectonic events that occurred during final Gondwanide fragmentation. The most dramatic of these is the impact of hotspot activity in the Kerguelen Plateau, which commenced some 50 Ma prior to final breakup in that sector.
Non-volcanic, or magma-poor, continental margins are defined by features such as a wide, largely amagmatic, continent–ocean transition zone (COTZ – defined below); unroofed mantle peridotite rocks exposed at the palaeo sea floor; relatively flat-lying, serpentinized detachment faults; typically slow (55– 20 mm year21) to ultra-slow (20– 1 mm year21) full spreading rates upon initiation of sea-floor spreading; and low volumes of magmatic rocks (Boillot & Froitzheim 2001; Whitmarsh et al. 2001). These type of margins have been documented in relatively few places around the globe. The principal example is off Iberia (Pickup et al. 1996; Whitmarsh et al. 2001) and its Newfoundland conjugate (Hopper et al. 2004). Other examples have been described in the Labrador Sea (Chian &
Louden 1994; Chalmers & Pulvertaft 2001), Woodlark Basin (Robertson et al. 2001), Great Australian Bight (Sayers et al. 2001) and East Antarctica (Colwell et al. 2006). Of these examples, the few studied with integrated geophysical data – usually deep multichannel seismic reflection (MCS), seismic refraction, gravity and magnetics – exhibit significant geometric and geophysical complexity both across and along the margin, and especially in the COTZ. We follow Colwell et al. (2006) in defining two terms that are often used confusingly in studies of rifted continental margins. † Continent –ocean transition zone (COTZ): a region on the continental margin that lies between the outboard edge of highly attenuated,
From: KARNER , G. D., MANATSCHAL , G. & PINHEIRO , L. M. (eds) Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, London, Special Publications, 282, 239– 263. DOI: 10.1144/SP282.12 0305-8719/07/$15.00 # The Geological Society of London 2007.
240
N. G. DIREEN ET AL.
unequivocal continental crust and the inboard edge of unequivocal oceanic crust. The COT includes both sedimentary and magmatic components in proportions that vary both along and across the margin, and may include areas of failed sea-floor spreading. † Continent–ocean boundary (COB): this marks the inboard edge of unequivocal oceanic crust. The complexity of non-volcanic margins includes exposure of magnetized (Zhao 2001) exhumed continental mantle peridotite–serpentinite complexes, and volcanic rocks at or near the sea floor (Abe 2001; Desmurs et al. 2001; Hebert et al. 2001) within the COTZ. Exhumation of serpentinized peridotite along extensional fault footwalls may lead to the creation of linear magnetic anomalies that may be misinterpreted as sea-floor spreading anomalies, as shown by Sayers et al. (2001). Another documented complexity of ‘amagmatic’ margins is the capping of attenuated continental crust in the COTZ by thick, magnetized basalt piles sourced from the locus of break-up (Colwell et al. 2006). These can, again, lead to the creation of apparently linear magnetic anomalies that may be mistaken for those associated with seafloor spreading. These kinds of complexities mean that there is potential for the outer parts of some non-volcanic margins to have remained unrecognized in the geological record – especially those where the distribution of continental and oceanic crust has been interpreted only on the basis of potential field data. This possible bias in interpretations may mean that the percentage of continental margins classified as non-volcanic margins may have been underestimated. This, in turn, could lead to a diminution of the importance of the mechanical processes that dominate non-volcanic margins in many recent generalized models for rifting and breakup. In this paper we present a new integrated interpretation of deep MCS, gravity and magnetic data across the southern margin of the Naturaliste Plateau and adjacent Diamantina Zone (the Naturaliste –Diamantina margin), and a new interpretation of data from the central Great Australian Bight margin. Both of these margins form part of the Mesozoic Southern Rift System (SRS: Stagg et al. 1999) and exhibit the characteristics of nonvolcanic rifted margins, including a wide COTZ (Sayers et al. 2001). The purpose of this study is to evaluate the degree of along-strike complexity within a single non-volcanic rifted margin system, at scales commensurate with the evolving palaeoplate boundary. This complexity, and its distribution, form important constraints on evaluating ranges of parameters
that are incorporated into thermo-mechanical models for continental rifting and breakup.
Study area The western and southern margins of Australia (Fig. 1) are tw