Growth and Collapse of the Tibetan Plateau (Geological Society Special Publication 353)

Growth and Collapse of the Tibetan Plateau

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) RICK LAW (USA) PHIL LEAT (UK) NICK ROBINS (UK) RANDELL STEPHENSON (UK) Society Books Advisors

MIKE BROWN (USA) ERIC BUFFETAUT (FRANCE ) JONATHAN CRAIG (ITALY ) RETO GIERE´ (GERMANY ) TOM MC CANN (GERMANY ) DOUG STEAD (CANADA ) GONZALO VEIGA (ARGENTINA ) MAARTEN DE WIT (SOUTH AFRICA )

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: GLOAGUEN , R. & RATSCHBACHER , L. (eds) 2011. Growth and Collapse of the Tibetan Plateau. Geological Society, London, Special Publications, 353. MECHIE , J., KIND . R., SAUL . J. 2011. The seismological structure of the Tibetan Plateau crust and mantle down to 700 km depth. In: GLOAGUEN , R. & RATSCHBACHER , L. (eds) Growth and Collapse of the Tibetan Plateau. Geological Society, London, Special Publications, 353, 107–123.

GEOLOGICAL SOCIETY SPECIAL PUBLICATION NO. 353

Growth and Collapse of the Tibetan Plateau EDITED BY

R. GLOAGUEN and L. RATSCHBACHER Technische Universita¨t Bergakademie, Freiberg, Institut fu¨r Geologie, Germany

2011 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 10 000. 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. þ 44 (0)20 7434 9944; Fax þ 44 (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. þ 44 (0)1225 445046, Fax þ 44 (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 2011. 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 Ltd, Saffron House, 6 –10 Kirby Street, London EC1N 8TS, UK. Users registered with the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923, USA: the item-fee code for this publication is 0305-8719/11/$15.00. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. ISBN 978-1-86239-326-4 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. þ 91 11 2327-9113/2326-4180 Fax þ 91 11 2326-0538 E-mail: [email protected]

Contents GLOAGUEN, R. & RATSCHBACHER, L. Growth and collapse of the Tibetan Plateau: introduction

1

ROGER, F., JOLIVET, M., CATTIN, R. & MALAVIEILLE, J. Mesozoic –Cenozoic tectonothermal evolution of the eastern part of the Tibetan Plateau (Songpan-Garzeˆ, Longmen Shan area): insights from thermochronological data and simple thermal modelling

9

ZHANG, Q.-H., DING, L., CAI, F.-L., XU, X.-X., ZHANG, L.-Y., XU, Q. & WILLEMS, H. Early Cretaceous Gangdese retroarc foreland basin evolution in the Selin Co basin, central Tibet: evidence from sedimentology and detrital zircon geochronology

27

DUNKL, I., ANTOLI´N, B., WEMMER, K., RANTITSCH, G., KIENAST, M., MONTOMOLI, C., DING, L., CAROSI, R., APPEL, E., EL BAY, R., XU, Q. & VON EYNATTEN, H. Metamorphic evolution of the Tethyan Himalayan flysch in SE Tibet

45

LIEBKE, U., ANTOLIN, B., APPEL, E., BASAVAIAH, N., MIKES, T., DUNKL, I. & WEMMER, K. Indication for clockwise rotation in the Siang window south of the eastern Himalayan syntaxis and new geochronological constraints for the area

71

HETE´NYI, G., VERGNE, J., BOLLINGER, L. & CATTIN, R. Discontinuous low-velocity zones in southern Tibet question the viability of the channel flow model

99

MECHIE, J., KIND, R. & SAUL, J. The seismological structure of the Tibetan Plateau crust and mantle down to 700 km depth

109

RATSCHBACHER, L., KRUMREI, I., BLUMENWITZ, M., STAIGER, M., GLOAGUEN, R., MILLER, B. V., SAMSON, S. D., EDWARDS, M. A. & APPEL, E. Rifting and strike –slip shear in central Tibet and the geometry, age and kinematics of upper crustal extension in Tibet

127

KIRBY, E. & OUIMET, W. Tectonic geomorphology along the eastern margin of Tibet: insights into the pattern and processes of active deformation adjacent to the Sichuan Basin

165

PERRINEAU, A., WOERD, J. V. D., GAUDEMER, Y., LIU-ZENG, J., PIK, R., TAPPONNIER, P., THUIZAT, R. & RONGZHANG, Z. Incision rate of the Yellow River in Northeastern Tibet constrained by 10Be and 26Al cosmogenic isotope dating of fluvial terraces: implications for catchment evolution and plateau building

189

HARTMANN, K., WU¨NNEMANN, B., HO¨LZ, S., KRAETSCHELL, A. & ZHANG, H. Neotectonic constraints on the Gaxun Nur inland basin in north– central China, derived from remote sensing, geomorphology and geophysical analyses

221

KORUP, O. & WEIDINGER, J. T. Rock type, precipitation, and the steepness of Himalayan threshold hillslopes

235

Index

251

Growth and collapse of the Tibetan Plateau: introduction RICHARD GLOAGUEN* & LOTHAR RATSCHBACHER Technische Universita¨t Bergakademie, Freiberg, Institut fu¨r Geologie, Germany *Corresponding author (e-mail: [email protected])

The India – Asia collision: archetype of collision Geodynamic focus: how does crust thicken in plateaus? The Pamir – Tibet– Himalaya orogenic system (Fig. 1) has been growing as a consequence of the collision between the Eurasian and the Indian plates since the Early Eocene (e.g. Hodges 2000). While the time of formation and the magnitude of shortening across the Tibetan Plateau are relatively well constrained, its origin and evolution are disputed, untested, or unknown (e.g. Yin & Harrison 2000). For example, it has been argued that Miocene to Recent geological, geodetic, palaeomagnetic, geomorphological, and geophysical features of the eastern Plateau are best explained by flow of low-viscosity material within the middle –lower crust (e.g. Clark & Royden 2000; Burchfiel et al. 2004). This model contrasts with the classic crustal imbrication –tectonic escape model that calls for thrust –fold belt style thickening and large-scale displacement of lithospheric blocks toward SE Asia. Both models have fundamental implications for crustal rheology, vertical coupling in the lithosphere, and Plateau formation and growth. Advocates of imbrication (e.g. Meyer et al. 1998; Chen et al. 1999; Tapponnier et al. 2001) have emphasized the importance of intracontinental subduction and large-scale thrust-fold belts in accommodating shortening and generating thick crust (Fig. 2). This classic mode of crustal shortening and accompanying lithospheric mantle subduction implies relatively brittle behaviour at substantial depth, with strain concentrated along moderately-dipping, discrete shear zones. Such thrusts/folds may be accompanied by deeply penetrating strike–slip shear zones that accommodate large fractions of collisional convergence by lateral extrusion (e.g. Tapponnier & Molnar 1976; Ratschbacher et al. 1991). In the Pamir– Tibet – Himalaya orogenic system, thickening may have occurred in steps that involved the sequential growth and related thickening/uplift of up to 2000 km sized (along strike) thrust-fold belts. While the crust thickens by folding-imbrication, the mantle decouples from the crust and subducts

intra-continentally along pre-existing heterogeneities; in the case of the Pamir –Tibet –Himalaya orogenic system, these likely are the c. east-striking, Palaeozoic –Mesozoic sutures (e.g. Schwab et al. 2004). Accompanying Cenozoic magmatic belts may be related to this subduction. Plateau growth is directed toward weakly constrained margins (i.e. the Pacific subduction zones) and spares lithospheric heterogeneities; that is, the Pamir –Tibet – Himalaya system grows towards the NE and NW into Palaeozoic orogenic belts, like Qinling (NE) and Fergana/Hissar (NW), bypassing the lithospheric heterogeneity of the Tarim basin that involves cratonal crust. Pamir – Tibet– Himalaya orogen formation stepped northward across Tarim, forming the Tien Shan, mostly over the last c. 10 Ma (e.g. Sun et al. 2004). This model describes continental deformation in analogy to plate tectonics, but presumes that highly distributed deformation in the crust veils discrete deeper level mantle subduction zones. Widespread and spectacular upper crustal extension is due to orogenic collapse and lateral extrusion. The crustal flow model appeals to ductile flow of weak lower crust to achieve both Plateau uplift (e.g. Zhao & Morgan 1987) and marginal expansion (e.g. Clark & Royden 2000; Klemperer 2006). To a first order, the geodetically determined velocity field for the Pamir –Tibet –Himalaya orogenic system is similar to results from 3D crustal models that assume a viscous crust deforming above a plate-like mantle (e.g. Shen et al. 2001), suggesting that on length scales of .100–200 km the behaviour of the Tibetan crust is similar to that of a viscous fluid. A striking outcome of the geodetic work is the observation that within parts of the Plateau margins, crustal thickening and surface uplift have occurred without significant shortening of the upper crust (e.g. Wang et al. 1998). This observation, paired with findings from INDEPTH and other surveys (e.g. Zhao et al. 1993; Nelson et al. 1996; Hacker et al. 2000; Unsworth et al. 2005) that the middle –lower crust beneath most of the Tibetan Plateau is rheologically weak led to models of crustal flow. In these models, material flows from the central Tibet Plateau toward its margins within a weak channel (e.g. Clark & Royden 2000; Klemperer 2006); the driving mechanism is

From: Gloaguen, R. & Ratschbacher, L. (eds) Growth and Collapse of the Tibetan Plateau. Geological Society, London, Special Publications, 353, 1–8. DOI: 10.1144/SP353.1 0305-8719/11/$15.00 # The Geological Society of London 2011.

2

R. GLOAGUEN & L. RATSCHBACHER

Fig. 1. Topography of the Pamir– Tibet–Himalayan orogenic system with reference to the geographic – geological frame discussed in this preface.

topographic stress due to differences in elevation. Any viscous or plastic deformation may be regarded as ‘flow’ and such flow need not be channelized; it may affect the entire lithosphere, as in the ‘thin viscous sheet’ (England & Molnar 1997), or in the ‘vertically coherent deformation’ style inferred for the Tibet Plateau from seismic and geodetic data (Flesch et al. 2005). Most models addressing channel flow, however, envision a viscosity minimum at some depth that localizes horizontal flow and decouples flow at different depths (e.g. Royden et al. 1997; Beaumont et al. 2004). Currently popular models of crustal flow (e.g. Clark & Royden 2000; Beaumont et al. 2004; Klemperer 2006) suggest that orographic exhumation and gravitational potential energy drive a mostly south- and east-directed Poiseuille-type flow (flow in a pipelike channel; Klemperer 2006). Current research on Plateau formation aims to test and unite the essences of these hypotheses. For example, geophysicists trace Moho offsets and mantle flow zones attached to potential intracontinental subduction zones (e.g. INDEPTH IV project goals), geologists map, quantify, and date shortening to balance crustal thickness and crustal contraction (e.g. Kapp et al. 2005), geochemists decipher the geodynamic significance of magmatic belts (lithosphere delamination v. localized continental subduction, e.g. Ding et al. 2007), geodesists discriminate block rotations from continuous flow (e.g. Thatcher 2007), and modellers build physical models for continental subduction and largescale crustal flow driven by topographic stress and climate (e.g. Zeitler et al. 2001; Beaumont et al. 2004) (Fig. 3).

Key elements of the Pamir– Tibet – Himalayan orogenic system Almost all of what we know about deformation of the Pamir –Tibet –Himalaya orogenic system comes from investigations of the eastern part of the system, the Tibetan Plateau. Key elements are: Subduction of Indian and Asian lithosphere beneath the Tibetan Plateau. The base of the Indian lithosphere dips northward from a depth of 160 km beneath the Himalaya to a depth of 220 km just south of the Bangong suture in southern central Tibet (e.g. Kumar et al. 2006). This modest, presentday northward extent of the Indian plate requires that since the onset of collision, subducting India has lost (by slab break-off ) and overridden its own sinking mantle (e.g. Replumaz et al. 2004). Earthquakes beneath the Himalaya occur at depths between near surface and c. 100 km below sea level (e.g. Monsalve et al. 2006). The base of the Asian lithosphere is nearly horizontal at a depth of c. 170 km from central to northern Tibet. There is a vertical gap of c. 50 km between Indian and Asian lithospheres. Asian lithosphere subduction may be indicated by a prominent southdipping converter that extends downward from the Moho beneath northern Tibet (e.g. Kind et al. 2002). Magmatic belts on the Tibetan Plateau. Although several interpretations attributed Cenozoic magmatism on the Tibet Plateau to melting of enriched lithospheric mantle induced by delamination of the thermal boundary layer (e.g. Turner et al. 1996), recent work attributes its geochemistry, age, and

INTRODUCTION

3

Fig. 2. (a) Schematic cross-section of thrust-fold belt structure and speculative location of Tertiary intra-continental subduction zones along a north– south line across the Tibet at c. 888E. (b) Upper crustal thickening structures and two solutions for their middle–lower crustal accommodation. The upper panel suggests that initial (Cretaceous) northward underthrusting of the Lhasa block underneath the Qiangtang block was followed by Tertiary underthrusting of the Qiangtang block underneath the Lhasa block. In the lower panel, the entire thrust belt of the Lhasa (Amdo anticlinorium) and southern Qiangtang (Shuang Hu anticlinorium) blocks formed by long-lived northward underthrusting of the Lhasa block beneath the Qiangtang block. Partly modified from Mechie et al. (2004) and inspired by Kapp et al. (2005).

arrangement parallel to the Palaeozoic –Mesozoic sutures as indicating suture-zone reactivation and melt generation in intra-continental subduction zones (e.g. Hacker et al. 2000, 2005; Ding et al. 2007). Post-collision, up to four magmatic arcs might have been present: (1) Miocene Gangdese; (2) southern Qiangtang; (3) northern Qiangtang; and (4) southern Kunlun. Zones of crustal shortening within the Tibetan Plateau. Four Cenozoic high-strain shortening zones may occur in Tibet and were suggested to be related to the Cenozoic magmatic belts: the Himalaya, southern Qiangtang, northern Qiangtang, and the Kunlun (Fig. 2). The Tertiary Shiquanhe– Gaize –Amdo thrust system substantially modified the Jurassic –Cretaceous Bangong suture and is south of the southern Qiangtang volcanic suite (e.g. Kapp et al. 2005, 2007). The Palaeogene Hohxil–Fenghuo Shan–Nanqian thrust belt is

superimposed on the Triassic–Jurassic Jinsha suture and is north of the northern Qiangtang lavas. The Kunlun/Qaidam basin thrust-fold belt lies north of the Late Miocene –Recent magmatic belt of the south Kunlun. The areal extent of these belts is poorly constrained and there is considerable shortening between them. Their root zones are speculative. Except for the Himalaya and the Kunlun, no substantial crystalline basement is exposed along these belts. Major strike–slip faults of the Tibet Plateau. While thrust faulting accommodated most of the India– Asia convergence, strike– slip faults have also played an important role. Such faults are implicit in the indentor/tectonic extrusion model of Asian deformation (e.g. Tapponnier & Molnar 1976). It is still unclear, whether these faults penetrate the entire lithosphere (e.g. Wittlinger et al. 1998) or are limited to the crust and detached from the

4

R. GLOAGUEN & L. RATSCHBACHER

Fig. 3. Particle flow lines (magenta bands) across Tibet derived from kinematic interpretation of surface structures. They possibly reflect ductile flow of weak mid-lower crust and could achieve both Plateau uplift and marginal expansion. A particle (e.g. red arrows) flowing out of central Tibet accelerates across central Tibet, reaches its highest velocity at the neck between the syntaxes (Assam-Namche Barwa of India and Gongha of South China), and decelerates south of it. Grey arrows mark the east component of the velocity vectors at c. 928E from Zhang et al. (2004) with reference to Eurasia and the blue arrows the velocity field of the Tibetan Plateau in a Tibetan-fixed reference frame, which best depicts the Plateau-interior deformation (from Gan et al. 2007). Digital elevation model as background is from http://www.ngdc.noaa.gov/mgg/topo/globe.html. Modified from Ratschbacher et al. (2011).

underlying mantle by low-angle faults (e.g. Searle 2006; Leloup et al. 2007). ‘Rigid’ intra-continental blocks of Central Asia. The Qaidam, Tarim, Tajik, Fergana, and Junggar basins appear as islands of lesser deformation sandwiched between zones of intensive faulting. These contrasts presumably derive from differing rheologies/compositions at depth or perhaps the distribution of pre-existing weaknesses. Drainage and basins in the Tibet orogenic system. Except for its margins, Tibet is drained internally. The Himalaya is considered the foremost example of the concentration of deformation by climatically controlled surface processes (e.g. Thiede et al. 2004; Finnegan et al. 2008). Cenozoic depositional sequences along the margin of the Tarim basin and within the Tibet Plateau trace its erosion and deformation evolution: in the western Kunlun, latest Cretaceous –Oligocene deposits are marine; during late Miocene –early Pleistocene, 2–3 km

thick continental molasse indicates rapid and accelerated hinterland uplift. Undeformed middle –upper Pleistocene deposits unconformably cover the steeply dipping molasse (e.g. Jin et al. 2003). The Nima basin, along the Bangong suture provides a record of Cretaceous –late Miocene sediment accumulation in south–central Tibet (DeCelles et al. 2007). A depositional hiatus at c. 50 Ma correlates with major shortening and ignimbrite eruption. The Oligocene Nima basin accumulated coarsegrained deposits associated with reactivated thrusts. Carbon and oxygen isotope data from palaeosol carbonate indicate late Oligocene regional elevations .4.6 km, like today (Rowley & Currie 2006). The Hoh Xil basin in north –central Tibet contains c. 6 km of sediments. This basin developed in three stages: 56–32 Ma, 32 –30 Ma, and approximately early Miocene; the first two comprise piggyback basins, whereas the latter represents a stable lake that post-dates major deformation (Liu & Wang 2001).

INTRODUCTION

Addressing open questions The papers in this Special Publication of the Geological Society, London, address some of the open questions and geodynamic aspects introduced above. The book is structured in order to new aspects on the evolution of the Tibetan Plateau from its generation to its collapse. The first group of papers concerns the early Plateau evolution; research appears to approach consensus that the Plateau had a pre-India–Asia collision topography/elevation and the E-striking, pre-existing suture zones were preferably reactivated during Plateau development. Roger et al. synthesize the tectonic and thermochronological evolution of the eastern Tibet Plateau since the Triassic. Based on long-term cooling histories obtained from magmatic and metamorphic rocks of the southern SongpanGarze system and Kunlun and Yidun blocks, they show that these zones underwent slow and regionally comparable cooling during the Late Jurassic and Cretaceous. Therefore, they suggest an absence of major tectonic events between c. 150 and 30 Ma. The first deformation event occurred at c. 30 Ma and is interpreted to record initial Plateau growth with localization of deformation along known faults. The Cretaceous accretion of the Lhasa block to Asia apparently had no effect on eastern Tibet. The authors also postulate that the eastern Plateau already had significant, albeit flat (plateau-like) topography. Zhang et al. postulate that significant crustal thickening occurred before the India –Asia collision. The Selin-Co basin of the northern Lhasa block formed in a retroarc foreland position with more than 3 km of upward coarsening Lower Cretaceous strata that comprise flysch to molasse. Their petrographic analysis yielded sandstones rich in volcanic and sedimentary lithoclasts, derived from recycled orogen and magmatic-arc sources. They speculate that significant crustal thickening started in the Early Cretaceous in the Lhasa block. Dunkl et al. looked into the metamorphic evolution of the southeastern Tibetan Plateau by determining the thermal overprint of Tethyan Himalayan Sequence rocks, using Ku¨bler Index and vitrinite reflectance analyses and applying thermobarometric and geochronological (illite/muscovite K –Ar and zircon and apatite (U –Th)/He geochronology) methods. After a first metamorphic episode during the Early Cretaceous, peak metamorphic conditions were reached in the Eocene, closely following the initiation of the India –Asia collision; subsequent metamorphism, reaching into the Miocene, is weak and likely related to ongoing shortening. Liebke et al. determined palaeomagnetic remanences in sedimentary and volcanic rocks of the Siang window of southeastern Tibet and inferred

5

post-Miocene clockwise block rotations compatible with lateral extrusion of upper crustal material north of the eastern Himalayan syntaxis. The second group of papers deals with geophysical surveys aiming to localize potential mid-crustal flow in Tibet and the kinematic interpretation of grabens and strike –slip fault zones in terms of flow. Hete´nyi et al. interpret new data from the HI-CLIMB seismic experiment and argue that the locally observed ‘bright spots’ or low-velocity zones imaged by the INDEPTH I-II team (e.g. Brown et al. 1996; Makovsky et al. 1996) were incorrectly interpreted to underlie the whole southern Tibetan Plateau; instead they interpret these low-velocity zones to be associated with the active Tibetan grabens. The authors critically debate the existence of a crustal flow channel. Mechie et al. provide a seismic profile down to 700 km depth across the Tibetan Plateau. One of the most striking and counterintuitive findings is the northward deepening of the 410 and 660 km discontinuities. This implies that the upper mantle beneath northern Tibet is slower, less dense and warmer than under southern Tibet which, in turn, could provide some of the isostatic support for the high elevations albeit thinner crust in northern Tibet. Another result that has emerged from the compilation of velocity models is the existence of a petrophysical change at 30–40 km depth beneath Tibet. Based on the P-wave velocities on either side of this boundary, it is proposed that it represents the interface between the felsic upper crust and more mafic lower crust. The authors argue for the presence of partial melts within the crust but admit strong heterogeneities in their distribution. Ratschbacher et al. analyse one of the c. northtrending graben systems in Tibet and outline a kinematic model for the spatial variability of normal and conjugate strike –slip faults over Tibet; the grabens are intimately associated with strike–slip faults and both accommodate north–south shortening and eastward transport of the Tibetan crust. The distribution of strike –slip and normal faults corresponds to Liu & Yang’s (2003) model of indentation by the Indian plate and gravitational spreading of the Plateau. The north–south variability of these structures in western and central Tibet is interpreted as an effect of the distance of underthrusting by the Indian plate. On first order, these results converge with Mechie et al. concerning the north– south partition of the rheological behavior of the Plateau. The third group of papers describes the active evolution of the Tibetan Plateau. The rising field of tectonic geomorphology has contributed important data on exhumation/uplift of active orogens and geographers and climatologists have a long experience in unravelling the climate record from

6

R. GLOAGUEN & L. RATSCHBACHER

sedimentary deposits. In this context the third group articles outline that the climatic and tectonic signals are hard to discriminate in the Pamir –Tibet –Himalaya orogenic system. Kirby & Ouimet present a geomorphological analysis of the eastern margin of the Tibetan Plateau adjacent to and north of the Sichuan Basin; they use morphometric indicators to map uplift. The results suggest that uplift is concentrated along fault systems adjacent to the Longmen Shan but there is also regional uplift in areas devoid of mapped faults. The authors propose ongoing crustal thickening triggered by mid-crustal flow now located in the area of the Qinling orogen north of the Sichuan Basin, which is in line with data from thermochronology and structural geology (Enkelmann et al. 2006). Perrineau et al. present a geochronological study of terraces along the Yellow River along the eastern edge of the Plateau. Their modelled cosmogenic ages indicate erosion rates of 2–6 mm/a, implying strong and recent drainage reorganization. Their work, although focusing on the interplay between climatic and tectonic forcing, also highlights the difficulty to separate the two signals from fragmentary records. This complexity in discriminating between climatic and tectonic forcing from geomorphological data is supported by the work of Hartmann et al., who show that tectonic activity enhanced subsidence rates by a factor 3 in the Gaxun Nur Basin at the northeastern edge of the Tibetan Plateau; their work calls for carefulness when interpreting proxies that are commonly related to climate. The records from the Gaxun Nur Basin outline a distinct interplay between non-climatic and climate-triggered landforming processes that affected the depositional environment and thus require a more substantiated interpretation of lake records (e.g. Wu¨nnemann et al. 2010). Korup & Weidinger show that the often forgotten lithological factor plays an important role in the geomorphological analysis of active orogens. In focusing on the coupling between climate and tectonics, authors commonly omit the well-known lithological effect on the incision/ uplift signal. In their study of incision on both sides of the Main Central Thrust Zone in Nepal, they found no relationship between uplift and incision under comparable climatic conditions. Furthermore, they illustrate that larger mass movements occur at lower altitudes, implying that para-glacial effects must be at work and demand further analysis. This book grew out of two sessions ‘Growth and Collapse of the Tibetan Plateau’ at the EGU General assembly held in Vienna during April 2009. It is dedicated to the unselfish work of a cornucopia of reviewers that improved the papers submitted. The editors were funded in the framework of their research in Central Asia by the Deutsche Forschungsgemeinschaft.

References Beaumont, C., Jamieson, R. A., Nguyen, M. H. & Medvedev, S. 2004. Crustal channel flows: 1. Numerical models with applications to the tectonics of the Himalayan –Tibetan orogen. Journal of Geophysical Research, 109, B06406, doi: 10.1029/2003JB002809. Brown, L. D., Zhao, W. et al. 1996. Bright spots, structure, and magmatism in southern Tibet from INDEPTH seismic reflection profiling. Science, 274, 1688–1690. Burchfiel, B. C. 2004. New technology; new geological challenges. GSA Today, 14, 4– 10. Chen, W.-P., Chen, C.-Y. & Nabelek, J. L. 1999. Present-day deformation of the Qaidam basin with implications for intra-continental tectonics. Tectonophysics, 305, 165–181. Clark, M. K. & Royden, L. H. 2000. Topographic ooze; building the eastern margin of Tibet by lower crustal flow. Geology, 28, 703– 706. DeCelles, P. G., Kapp, P., Ding, L. & Gehrels, G. E. 2007. Late Cretaceous to middle Tertiary basin evolution in the central Tibetan Plateau: changing environments in response to tectonic partitioning, aridification, and regional elevation gain. Geological Society of America Bulletin, 119, 654–680. Ding, L., Kapp, P., Yue, Y. & Lai, Q. 2007. Postcollisional calc-alkaline lavas and xenoliths from the southern Qiangtang terrane, central Tibet. Earth and Planetary Science Letters, 254, 28– 38. England, P. & Molnar, P. 1997. Active deformation of Asia: from kinematics to dynamics. Science, 278, 647–650. Enkelmann, E., Ratschbacher, L. et al. 2006. Cenozoic exhumation and deformation of northeastern Tibet and the Qinling: is Tibetan lower crustal flow diverging around the Sichuan basin? Geological Society of America Bulletin, 118, 651 –671. Finnegan, N. J., Hallet, B., Montgomery, D. R., Zeitler, P. K., Stone, J. O., Anders, A. M. & Yuping, L. 2008. Coupling of rock uplift and river incision in the Namche Barwe– Gyala Peri Massif, Tibet. Bulletin of the Geological Society of America, 120, 142 –155. Flesch, L. M., Holt, W. E., Silver, P. G., Stephenson, M., Wang, C.-Y. & Chan, W. W. 2005. Constraining the extent of crust– mantle coupling in central Asia using GPS, geologic, and shear-wave splitting data. Earth and Planetary Science Letters, 238, 248–268. Gan, W., Zhang, P. et al. 2007. Present-day crustal motion within the Tibetan Plateau inferred from GPS measurements. Journal of Geophysical Research, 112, B08416, doi: 10.1029/2005JB004120. Hacker, B. R., Gnos, E. et al. 2000. Hot and dry deep crustal xenoliths from Tibet. Science, 287, 2463–2466. Hacker, B. R., Luffi, P. et al. 2005. Near-ultrahigh pressure processing of continental crust: Miocene crustal xenoliths from the Pamir. Journal of Petrology, 46, 1661–1687. Hodges, K. V. 2000. Tectonics of the Himalaya and southern Tibet from two perspectives. Geological Society of America Bulletin, 112, 324 –350. Jin, X., Wang, J., Chen, B. & Ren, L. 2003. Cenozoic depositional sequences in the Piedmont of the west

INTRODUCTION Kunlun and their paleogeographic and tectonic implications. Journal of Asian Earth Sciences, 21, 755– 765. Kapp, P., DeCelles, P. G., Gehrels, G. E., Heizler, M. & Ding, L. 2007. Geological records of the LhasaQiangtang and Indo-Asian collisions in the Nima area of central Tibet. Geological Society of America Bulletin, 119, 917– 932. Kapp, P., Yin, A., Harrison, T. M. & Ding, L. 2005. Cretaceous–Tertiary shortening, basin development, and volcanism in central Tibet. Geological Society of America Bulletin, 117, 865 –878. Kind, R., Yuan, X. et al. 2002. Seismic images of crust and upper mantle beneath Tibet: evidence for Eurasian plate subduction. Science, 298, 1219–1221. Klemperer, S. L. 2006. Crustal flow in Tibet: geophysical evidence for the physical state of Tibetan lithosphere, and inferred patterns of active flow. In: Law, R. D., Searle, M. P. & Godin, L. (eds) Channel Flow, Ductile Extrusion and Exhumation in Continental Collision Zones. Geological Society, London, Special Publications, 268, 39– 70. Kumar, P., Yuan, X., Kind, R. & Ni, J. 2006. Imaging the colliding Indian and Asian lithospheric plates beneath Tibet. Journal of Geophysical Research, 111, B06308, doi: 10.1029/2005JB003930. Leloup, P. H., Tapponnier, P. & Lacassin, R. 2007. Discussion on the role of the Red River shear zone, Yunnan and Vietnam, in the continental extrusion of SE Asia. Journal of the Geological Society, London, 164, 1253–1260. Liu, Z. & Wang, Ch. 2001. Facies analysis and depositional systems of Cenozoic sediments in the Hoh Xil basin, northern Tibet. Sedimentary Geology, 140, 251–270. Liu, M. & Yang, Y. 2003. Extensional collapse of the Tibetan Plateau: results of three-dimensional finite element modeling. Journal of Geophysical Research, 108, 2361, doi: 10.1029/ 2002JB002248. Makovsky, Y., Klemperer, S. L., Ratschbacher, L., Brown, L. D., Li, M., Zhao, W. & Meng, F. 1996. INDEPTH wide-angle reflection observation of P-to-S wave conversion from crustal bright-spots in Tibet. Science, 274, 1690– 1691. Mechie, J., Sobolev, S. V. et al. 2004. Precise temperature estimation in the Tibetan crust from seismic detection of the a– b quartz transition. Geology, 32, 601–604. Meyer, B., Tapponnier, P. et al. 1998. Crustal thickening in Gansu-Qinghai, lithospheric mantle subduction, and oblique, strike– slip controlled growth of the Tibet Plateau. Geophysical Journal International, 135, 1–47. Monsalve, G., Sheehan, A., Schulte-Pelkum, V., Rajaure, S., Pandey, M. R. & Wu, F. 2006. Seismicity and one-dimensional velocity structure of the Himalayan collision zone: earthquakes in the crust and upper mantle. Journal of Geophysical Research, 111, B10301, doi: 10.1029/2005JB004062. Nelson, K. D., Zhao, W. J. et al. 1996. Partially molten middle crust beneath southern Tibet: synthesis of project INDEPTH results. Science, 274, 1684–1688. Ratschbacher, L., Frisch, W., Linzer, H.-G. & Merle, O. 1991. Lateral extrusion in the Eastern Alps, part II: structural analysis. Tectonics, 10, 257– 271.

7

Ratschbacher, L., Krumrei, I. et al. 2011. Rifting and strike–slip shear in central Tibet and the geometry, age, and kinematics of upper crustal extension in Tibet. In: Gloaguen, R. & Ratschbacher, L. (eds) Growth and Collapse of the Tibetan Plateau. Geological Society, London, Special Publications, 353, 125– 161. Replumaz, A., Karason, H., van der Hilst, R. D., Besse, J. & Tapponnier, P. 2004. 4-D evolution of SE Asia’s mantle from geological reconstructions and seismic tomography. Earth and Planetary Science Letters, 221, 103– 115. Rowley, D. B. & Currie, B. S. 2006. Palaeo-altimetry of the late Eocene to Miocene Lunpola basin, central Tibet. Nature, 439, 677– 681. Royden, L. H., Burchfiel, B. C., King, R. W., Wang, E., Chen, Z., Shen, F. & Liu, Y. 1997. Surface deformation and lower crustal flow in eastern Tibet. Science, 276, 788–790. Schwab, M., Ratschbacher, L. et al. 2004. Assemblage of the Pamir: age and origin of magmatic belts from the southern Tien Shan to the southern Pamir and their relation to Tibet. Tectonics, 23, TC4002, doi: 10.1029/2003TC001583. Searle, M. P. 2006. Role of the Red River Shear zone, Yunnan and Vietnam, in the continental extrusion of SE Asia. Journal of the Geological Society, London, 163, 1025– 1036. Shen, F., Royden, L. H. & Burchfiel, B. C. 2001. Large-scale crustal deformation of the Tibetan Plateau. Journal of Geophysical Research, 106, 6793– 6816. Sun, J., Zhu, R. & Bowler, J. 2004. Timing of the Tianshan Mountains uplift constrained by magnetostratigraphic analysis of molasse deposits. Earth and Planetary Science Letters, 219, 239–253. Tapponnier, P. & Molnar, P. J. 1976. Slip-line field theory and large-scale continental tectonics. Nature, 264, 319–324. Tapponnier, P., Xu, Z., Roger, F., Meyer, B., Arnaud, N., Wittlinger, G. & Yang, J. 2001. Oblique stepwise rise and growth of the Tibetan Plateau. Science, 294, 1671– 1677. Thatcher, W. 2007. Microplate model for the presentday deformation of Tibet. Journal of Geophysical Research, 112, B01401, doi: 10.1029/2005JB004244. Thiede, R. C., Bookhagen, B., Arrowsmith, R., Sobel, E. R. & Strecker, M. R. 2004. Climatic control on rapid exhumation along the Southern Himalayan Front. Earth and Planetary Science Letters, 222, 791–806. Turner, S., Arnaud, N. et al. 1996. Post-collision, shoshonitic volcanism on the Tibetan plateau: implications for convective thinning of the lithosphere and the source of ocean island basalts. Journal of Petrology, 37, 45–71. Unsworth, M. J., Jones, A. G., Wei, W., Marquis, G., Gokarn, S. G., Spratt, J. E. & INDEPTH-MT TEAM 2005. Crustal rheology of the Himalaya and Southern Tibet inferred from magnetotelluric data. Nature, 483, 78– 81. Wang, E., Burchfiel, B. C., Royden, L. H., Chen, L. Z., Chen, J. S., Li, W. X. & Chen, Z. L. 1998. Late Cenozoic Xianshuihe-Xiaojiang, Red River, and Dali fault systems of southwestern Sichuan and central

8

R. GLOAGUEN & L. RATSCHBACHER

Yunnan, China. Geological Society of America Special Paper, 327, 1– 108. Wittlinger, G., Tapponnier, P., Poupinet, G., Jiang, M., Si, D., Herquel, G. & Masson, F. 1998. Tomographic evidence for localized lithospheric shear along the Altyn Tagh fault. Science, 282, 74– 76. Wu¨nnemann, B., Demske, D. et al. 2010. Hydrological cycles during the last 15 kyr in the Tso Kar lake basin, Ladakh (India), derived from geomorphological, sedimentological and palynological records. Quaternary Science Reviews, 29, 1138–1155. Yin, A. & Harrison, T. M. 2000. Geologic evolution of the Himalayan –Tibetan orogen, Annual Review of Earth and Planetary Sciences, 28, 211– 280.

Zeitler, P. K., Meltzer, A. S. et al. 2001. Erosion, Himalayan geodynamics, and the geology of metamorphism. GSA Today, 11, 4– 8. Zhang, P., Shen, Z., Wang, M., Gan, W., Bu¨rgmann, R. & Molnar, P. 2004. Continuous deformation of the Tibetan Plateau from global positioning system data. Geology, 32, 809 –812. Zhao, W. & Morgan, W. J. 1987. Injection of Indian crust into Tibetan lower crust: a two-dimensional finite element model study. Tectonics, 6, 489– 504. Zhao, W. J., Nelson, K. D. & PROJECT INDEPTH TEAM 1993. Deep seismic-reflection evidence for continental underthrusting beneath southern Tibet. Nature, 366, 6455, 557– 559.

Mesozoic– Cenozoic tectonothermal evolution of the eastern part of the Tibetan Plateau (Songpan-Garzeˆ, Longmen Shan area): insights from thermochronological data and simple thermal modelling FRANC¸OISE ROGER1*, MARC JOLIVET1,2, RODOLPHE CATTIN1 & JACQUES MALAVIEILLE1 1

Universite´ Montpellier 2, Laboratoire Ge´osciences Montpellier CNRS-UMR 5243, Place Euge`ne Bataillon, F-34095 Montpellier Cedex 05, France

2

Universite´ Rennes 1, Laboratoire Ge´osciences Rennes, CNRS-UMR 6118, Bat 15, Campus de Beaulieu, F-35042 Rennes Ce´dex, France *Corresponding author (e-mail: [email protected]) Abstract: We present a synthesis of the tectonic and thermochronological evolution of the Eastern Tibet since the Triassic. The long-term cooling histories obtained on magmatic and metamorphic rocks of the South Songpan-Garzeˆ, Kunlun and Yidun blocks are similar showing a very slow and regular cooling during Late Jurassic and Cretaceous, confirming the suspected lack of major tectonic events between c. 150 and 30 Ma. The exhumation linked to the Tertiary growth of the Tibetan Plateau initiated around 30 Ma and concentrates at the vicinity of the major tectonic structures. Exhumation rates increased again from about 7 Ma in the Longmen Shan. To interpret this very slow cooling rate between Late Jurassic and Early Cenozoic from granites of this area, we use a simple 1D thermal model that takes into account the thermal properties of both sediments and crust. The results suggest that: (1) high temperature (500 8C) can be kept over a long period of time; (2) during Cretaceous, cooling is mostly controlled by the thermal properties of sediments of continental origin; and (3) the initial Late Triassic rapid cooling rate was caused by the large thermal contrast between the granite body and the sedimentary rocks rather than by a high exhumation rate.

The tectonic evolution of northeastern Tibet presents intriguing features. First in this area, most of the rocks presently exposed consist of flyschoid Triassic sequences showing widespread deformations (the series are deformed into a wide accretionary fold– thrust belt) and very low metamorphism. This suggests that since the Triassic orogeny, this area has only been submitted to very low erosion and exhumation rates. Furthermore despite the huge distances that separate the sampled locations, the studies of South Songpan-Garzeˆ, Kunlun, and Yidun granites suggest a homogeneous thermal evolution for the whole North-Eastern Tibet. All these granite plutons exhibit a similar three stages thermal evolution, which includes: (1) a short period (,10 Ma) of rapid cooling after their emplacement during late Triassic; (2) a long period (at least 100 Ma) of thermal stability; and (3) a final Middle Tertiary exhumation. No satisfying explanation has yet been proposed to explain this original behaviour. Here using a synthesis of structural and thermochronological data combined with a simple modelling of the tectono-thermal behaviour of the

lithosphere, we discuss the relative effect of firstorder parameters that played a great part in this evolution from Triassic to present. The Triassic fold belt of NE Tibet results from interactions between the south China, north China and Qiangtang (north Tibet) blocks during the Indosinian orogeny (e.g. Pullen et al. 2008; Roger et al. 2008). It is mainly composed, from west to east, of the Bayan Har, Songpan-Garzeˆ, and Yidun (or Litang-Batang) terranes (Fig. 1). In Permian times, due to the opening of the Neotethys, the Qiangtang block, which was part of the Cimmerian continent detached from the Gondwanan continent, migrated northwards, closing the Palaeotethys (Fig. 2a). A synchronous activity along three subduction zones, Kunlun-Anyemaqen to the north, Jinsha to the south and Yushu-Batang to the east, induced the growth of a wide accretionary orogen until the end of the Triassic period. The Songpan ocean formed the eastern domain of the Palaeotethys and was divided into two main basins (Roger et al. 2008). The first one developed towards the east due to differential motion between the north China Block and the south China Block

From: Gloaguen, R. & Ratschbacher, L. (eds) Growth and Collapse of the Tibetan Plateau. Geological Society, London, Special Publications, 353, 9–25. DOI: 10.1144/SP353.2 0305-8719/11/$15.00 # The Geological Society of London 2011.

10

F. ROGER ET AL.

Fig. 1. Geographical extension of the Songpan-Garzeˆ belt. The white area represents the current extension of the Triassic belt (after Roger et al. 2008). SG, Songpan-Garzeˆ belt; BH, Bayan Har terrane; Yi, Yidun (or Litang-Batang) block;LH, Lhasa block; QT, Qiangtang block; QA, Kunlun-Qaidam block.1, rocks of the South Songpan-Garzeˆ area; 2, granites of Yidun block; 3, Xidatan gneiss; 4, Kokoxili granites; 5, blueschist of Gangma Co; 6, granites of Yushu. Numbers (1, 2, 3, 4) refer to cooling histories presented from Figures 5 and 6.

(Songpan-Garzeˆ basin) (Zhang et al. 2006). The second basin (Garzeˆ-Litang or Yidun basin) was a marginal basin developed in a back-arc setting, following the Permian rifting event (Panxi Rift), which separated the Yidun block from the south China – Indochina Craton (e.g. Roger et al. 2008, 2010; Zi et al. 2008 ). The east dipping subduction zone (Yushu-Batang) that generated the Yidun calcalkaline granites (245 –229 Ma) was thus located west of the Yidun block (Reid et al. 2007). This subduction, together with the Jinsha subduction zone further west, accommodated the closure of the wide Palaeotethys ocean. To the northwest, the Yushu-Batang subduction became intra-oceanic and joined the Kunlun-Anyemaqen subduction zone to the north. The subsequent arc-type volcanism formed the intra-oceanic Yushu arc (Fig. 2b). Along the northern and western border of the south China block, the Songpan-Garzeˆ and Litang basins were limited by a passive margin (Mattauer et al. 1992; Calassou et al. 1994; Roger et al. 2004; Zhang et al. 2006). The sedimentation in the oceanic basins was characterized by thick detrital flyschoid series (5 to 15 km), deposited through large alluvial fans. Those sediments, now forming the Songpan-Garzeˆ accretionary prism were deposited on both the oceanic crust of the basin and the surrounding thinned continental margins (Zou

et al. 1984; Sengo¨r 1985; Rao & Xu 1987; Mattauer et al. 1992; Nie et al. 1994) (Figs 2b & 3). During late Triassic– Lower Jurassic period, the Triassic Songpan-Garzeˆ ocean is limited to the north by the Kunlun–Anyemaqen–Qinling subduction zone which induced the widespread emplacement of large calc-alkaline batholiths (220 –200 Ma) (e.g. Roger et al. 2003, 2008) and controlled the development of the Kunlun and Qinling orogenic wedges. In the Qinling belt, continental subduction followed the oceanic closure, marked by late Triassic high-pressure metamorphism (Ames et al. 1993, 1996; Eide et al. 1994; Rowley et al. 1997; Hacker et al. 1998, 2006; Li et al. 2007) (Fig. 2c). South of the Songpan-Garzeˆ ocean deformation along the Jinsha subduction zone is attested by HP-LT, 220 Ma old blueschist metamorphism (Kapp et al. 2000; Pullen et al. 2008). At 210– 200 Ma old calc-alkaline Yushu granites (Roger et al. 2003) were emplaced in the complex orogenic wedge bounding the northern edge of the Qiangtang block (Figs 1 & 4). To the SE, along the eastern margin of the Yidun block, a second series of calc-alkaline granites (219 –216 Ma) was emplaced in late Triassic flysch-type sediments and volcanics (Reid et al. 2007) (Fig. 4). Subduction under both the North China block and the Qiangtang block led to the shortening of

THERMAL EVOLUTION OF EASTERN TIBET

11

Fig. 2. Geodynamic evolution of North Tibet from Permian to Tertiary (modified after Roger et al. 2008, 2010). Dark green dotted lines show location and names of future subduction zones. Green area shows the Palaeotethysian domain and its northern branches (light green). Dotted area shows the domains of Triassic high sedimentary input. Yi, Yidun (or Litang-Batang) block; IN, Indochina block; QT, Qiangtang block; QA, Qaidam block; NC, North China block; SC, South China block; AB, CD location of evolutionary sections of Figure 3(a, b). (a) Northward motion of the Qiangtang block induces the closure of the Palaeotethys ocean and the development of the Yushu volcanic arc. (b) Shortening of the northeastern branch of the Palaeotethys ocean along the north-dipping Anyemaquen subduction zone. Growth of the Songpan-Garzeˆ de´collement – fold–thrust belt: the Songpan-Garzeˆ prism is thrust onto the South China passive margin. (c) End of subductions: Remnants of the Palaeotethys (Jinsha suture) are trapped between the intra-oceanic Yushu arc/Yidun block and the Qiangtang continental margin. To the north, remnants of the northern oceanic basin (Anyemaquen suture) separate NC from the thick and wide Triassic orogenic wedge. (d) Shortening following the India– Asia collision induced south-verging continental subduction along the reactivated Jinsha suture zone. The Triassic de´collement level is reactivated. Formation of the Longmen Shan initiates c. 30 Ma.

the Songpan-Garzeˆ ocean and the development of the wide Triassic orogenic fold-thrust wedge (Figs 2–4). In the Songpan-Garzeˆ area the Triassic series are intensively folded and separated from the crystalline basement outcropping near Danba by a large-scale south-verging de´collement affecting the Palaeozoic series (Malavieille et al. 1991; Calassou 1994; Huang et al. 2003; Roger et al. 2004; Harrowfield & Wilson 2005; Zhou et al. 2008) (Fig. 4). The Triassic de´collement zone has been recognized over more than 300 km and studied in detail in Tien Wan (SW of Konga

Shan) and Danba areas (Mattauer et al. 1992; Xu et al. 1992; Calassou 1994; Harrowfield & Wilson 2005). However this structure or similar and/or associated ones can possibly be followed on thousands of kilometres from the north in the Dabie Shan and Qinling mountains to the south in the Konga Shan area (Hsu¨ et al. 1987; Mattauer et al. 1992; Xu et al. 1992; Huang & Wu 1992; Calassou 1994; Roger et al. 2004; Harrowfield & Wilson 2005). The important shortening induced by the Triassic orogeny was accommodated through major crustal thickening of the orogenic wedge

12

F. ROGER ET AL.

Fig. 3. Synthetic cross-sections showing the tectonic evolution of the Triassic belt from Jurassic– Tertiary (modified after Roger et al. 2008, 2010). See text and Figure 2 for detailed explanations.

(Malavieille et al. 1991; Calassou 1994; Roger et al. 2004) (Fig. 3). In the internal domain (close to Danba), this deformation is associated with a Barrovian type metamorphism (garnet –sillimanite – kyanite) (5–7 kbar and 400 –600 8C) dated at 220– 190 Ma (Xu et al. 1992; Huang et al. 2003; Zhou et al. 2008). Along the South China block

passive margin, crustal thickening induced the emplacement of syn to late-orogenic adakitic or I-type and S-type granitoids (220 –200 Ma) (Roger et al. 2004; Zhang et al. 2006; Xiao & Clemens 2007; Xiao et al. 2007). Later on, post-orogenic granitoids (200–150 Ma) were also emplaced into the prism (Wallis et al. 2003; Roger et al. 2004)

THERMAL EVOLUTION OF EASTERN TIBET

13

Fig. 4. Geological map of the Songpan-Garzeˆ belt (modified after Roger et al. 2010 and C.I.G.M.R. 1991) showing the U– Pb ages available for the Triassic magmatic rocks and Danba domal metamorphic terrane (Danba D.M. T.).

(Fig. 4). The isotopic composition (Nd, Sr, Pb) of most of these granitoids shows that their magma source was predominantly derived from melting of the Songpan-Garzeˆ Proterozoic basement (Yangtze block) with varying degrees of sedimentary material and negligible mantle source contribution (Roger et al. 2004; Zhang et al. 2006; Xiao et al. 2007). In the northern part of the Songpan-Garzeˆ terrane, immediately south of the Anyemaqen suture zone, the Nianbaoyeche granite (211 + 1 Ma) shows a different chemical composition (H. F. Zhang et al. 2007, 2008b) suggesting that the magma source is probably different (Fig. 4). The authors interpreted this granite as a post-orogenic A-type granite resulting from lithospheric delamination following the Triassic crustal thickening. However, the origin of

the A-type granite is still debated. C. Z. Zhang et al. (2008a) proposed that the granite was generated by northward subduction along the Jinsha subduction zone. The very different isotope signature of the Nianbaoyeche granite compared to the other Sonpan-Garzeˆ granites further south would seem instead to indicate a different origin, which might be related to the occurrence of a remnant oceanic crust below the sediments of the northern SongpanGarzeˆ basin (e.g. Roger et al. 2010) (Fig. 3). Except for few localized Tertiary and Quaternary outcrops, post-Triassic sediments are totally absent from the Songpan-Garzeˆ –Yidun fold belt (C.I.G.M.R. 1991) (Fig. 4). However, post-Triassic deformation has been recognized within the belt. During the Tertiary, it is generally admitted that

14

F. ROGER ET AL.

additional shortening and crustal thickening has been induced by the India –Eurasia collision. It resulted in the Songpan–Garzeˆ – Yidun area in a reactivation of the Triassic high angle faults, and folds, and large scale folding of the de´collement of the accretionary belt. For instance, in the Danba area the de´collement is affected by a wide Tertiary antiform with a NNW– SSE axis (Huang et al. 2003; Roger et al. 2004; Zhou et al. 2008). Still during the Tertiary, the de´collement level has been exhumed in the hanging wall of the NE–SW thrust faults of the Longmen-Shan belt that marks the transition between the Tibetan Plateau and the Sichuan basin (Fig. 3). Toward the SW, the most prominent tectonic feature is the active sinistral strike slip Xianshui He fault (Allen et al. 1991). Geochronological data on the synkinematic Konga Shan granite have shown that the fault is active since at least the Miocene (13 Ma, U –Pb on zircons) (Roger et al. 1995) (Fig. 4). All this part of eastern Tibet have episodically absorbed significant shortening since the Late Triassic to present (Chen & Wilson 1996). The amount and precise timing of post-Triassic deformation in the SongpanGarzeˆ –Yidun and Longmen Shan area are difficult to constrain especially because of the difficulty to isolate the Tertiary thermochronological signal from the protracted Late Triassic– Cretaceous thermal history (e.g. Wilson et al. 2006). Nonetheless it is generally accepted that Jurassic –Cretaceous tectonism did not modified the general Triassic architecture of eastern Tibet (e.g. Burchfield et al. 1995; Roger et al. 2004; Harrowfield & Wilson 2005; Reid et al. 2005b; Wilson et al. 2006). In the following the thermal evolution of eastern Tibet is presented and discussed in detail. First using a simple 1-D thermal modelling, we pay a special attention to Cretaceous thermal structure to estimate the parameters-set that could lead to the stable thermal regime observed over 100 Ma. Next, based on available low temperature thermochronological data we study the timing and the location of late tertiary exhumation. Finally we discuss the tectonic implications on the evolution of eastern Tibet over the last 200 Ma.

Pre-Tertiary cooling curves: 100 Ma of thermal quiescence The Songpan-Garzeˆ prism: an atypical belt? Geochronological data account for the closure temperature of each mineral/geochronological system (Dodson 1973). Figures 5 and 6 present general cooling curves drawn using ages obtained from various geochronometers applied to single samples of granites and gneisses from south

Songpan-Garzeˆ (Xu & Kamp 2000; Kirby et al. 2002; Roger et al. 2004; Zhang et al. 2006), western Kunlun (Arnaud et al. 2003; Jolivet et al. 2003; Roger et al. 2003) and Yidun arc (Reid et al., 2005a, 2007; Lai et al. 2007). The cooling curves obtained for these three Triassic granites are similar: a phase of rapid cooling immediately after the emplacement of the pluton, followed by a very slow and regular cooling during Late Jurassic and Cretaceous –Early Tertiary. Together these results suggest: (1) that the initial rapid cooling rate during the Late Triassic was induced by a large thermal contrast between the newly emplaced granite body and the sedimentary rocks rather than by a high exhumation rate; and (2) an absence of major tectonic events between c. 150 and 50 –30 Ma. The fission tracks (zircons) data (Xu & Kamp et al. 2000; Kirby et al. 2002; Lai et al. 2007) indicate that the final exhumation and Tertiary cooling is younger than 30 Ma. Richardson et al. (2008) demonstrated the occurrence of a major, general erosion event in the Sichuan basin initiating after 40 Ma and no later than 25 Ma. They rely this phase of major erosion (at least 1.3 km of sediments removed) to major modifications of the drainage pattern and especially the driving of the Middle and Lower Yangtze Rivers through the Three Gorges area. Such reorganization of the drainage pattern could also indicate the onset of topography building in northeastern Tibet. In the South Songpan-Garzeˆ area, the temperature prevailing for each sample during the thermal stability period can be directly related to their burying depth at the onset of this thermal phase. For example, the post-tectonic Markam granite, mainly derived from melting of the Songpan Garzeˆ sediments was emplaced above the de´collement level and shows a temperature of stability (i.e. the temperature from which the observed cooling curve becomes nearly flat for a long period of time, temperature corresponding to an equilibrium with the host rock) around 300 8C, whereas the syntectonic Manai granite which was emplaced inside the de´collement level shows a temperature of stability of 500 8C (Figs 5 & 6a). Finally the Yanggon and Maoergai granites were emplaced higher up in the sedimentary pile at temperatures of 200– 300 8C (Figs 5 & 6a). The cooling history of the major metamorphic zones in the ‘Danba Domal Metamorphic Terrane’ is similar to that of the Manai granite. Huang et al. (2003) interpreted these geochronological data as a two-stages metamorphic evolution in the Danba high-grade metamorphic terrane. The Barrovian metamorphic phase M1 (garnet, staurolite, and kyanite zones) (5–7 kbar and 570–600 8C) initiated at the end of the Triassic due to the collision between the South China and North China blocks.

THERMAL EVOLUTION OF EASTERN TIBET

15

Fig. 5. Compilation of geochronological data obtained on magmatic and sedimentary rocks in the South Songpan-Garzeˆ. All cooling histories display a similar trend. The grey envelope shows the cooling history of the Ma Nai and Rilonguan granites as well as of metamorphic rocks in the Danba area (white boxes). Rapid cooling after granite emplacement is followed by a very slow and regular cooling during Jurassic and Cretaceous, suggesting the absence of major tectonic events between c. 150 and 50–30 Ma. The fission tracks (TF) data indicate that the final exhumation and cooling occurred around 30 Ma. U –Pb and Rb/Sr data are from Roger et al. (2004), and the fission track data are from Xu & Kamp (2000). Danba data are from Huang et al. (2003). The striped boxes represent the data from the Markam granite. The hatched envelope shows the cooling history of the Yanggon and Maoergai granites. U–Pb data are from Zhang et al. (2006) and Kirby et al. (2002). Rb –Sr (biotite), Ar/Ar (K-feldspar) and U– Th/He (apatite and zircon) data are from Kirby et al. (2002). Closure temperatures used are estimated (excepted for some metamorphic data). Zirc., zircon; Mnz, monazite; Ti, titanite; Ap., apatite; Gr, garnet; Amp., amphibole; Ms, muscovite; Bt, biotite; K-Feld., K-feldspar.

M1 was recorded throughout the Songpan-Garzeˆ orogenic belt with peak conditions reached at 204 –190 Ma. The second higher-temperature sillimanite-grade M2 metamorphic phase (660– 725 8C) occurred locally in the northern Danba area at 164 Ma. It is marked by the occurrence of sillimanite and local migmatite and may represent a local thermal perturbation. The origin of this

event remains unclear. For Huang et al. (2003) the M2 thermal event could have been related to east –west compression during the early Yenshanian collision between the south Tibet (Lhasa terrane) and north Tibet blocks. But recently, Zhou et al. (2008) proposed that the Neoproterozoic crystalline basement of Danba was first exhumed in Middle Jurassic (159 –173 Ma) during the beginning of a

16

F. ROGER ET AL.

Fig. 6. Cooling histories of (a) the South Songpan-Garzeˆ rocks (see Fig. 5 for details of the data), (b) the Kunlun granites and gneiss and (c) the Yidun arc granites drawn from multisystem geochronology. For the Kunlun area (b), two Kokoxili sample are Triassic granites (White and black box). Data are from Roger et al. (2003) and Jolivet et al. (2003). For the Xidatan orthogneiss (grey box), the data are from Arnaud et al. (2003). These areas have a similar cooling history. For the Yidun granites, the data are from Reid et al. (2005a, 2007) and Lai et al. (2007).

THERMAL EVOLUTION OF EASTERN TIBET

major period of extensional deformation in the region generated by the collision between the north China and south China blocks. Between 138 and 30 Ma, cooling appears uniform and slow. Clustered Rb –Sr biotite ages mark a return to fast cooling rates (associated to renewed uplift and exhumation) around 30 Ma (Huang et al. 2003) (Fig. 5). The c. 100 Ma long period of thermal stability observed in all the cooling curves presented in Figure 6 suggests a lack of major tectonic event, erosion or sedimentation between c. 150 and c. 30 Ma (Xu & Kamp 2000; Jolivet et al. 2001; Kirby et al. 2002, Huang et al. 2003, Roger et al. 2003, 2004, 2010, Reid et al. 2005a,b, 2007; Lai et al. 2007; Zhou et al. 2008). This near absence of thermal change is not restricted to the cooling history of the Triassic granites but is also recorded by the metamorphic series of the Danba dome (Figs 5 & 6a) or by some Palaeozoic granites in the Kunlun belt (Xidatan orthogneiss) (Fig. 6b). Furthermore, all these samples have very different geodynamical settings: crustal thickening in Songpan-Garzeˆ, subduction related calc-alkaline magmatism in the Kunlun and Yidun (Roger et al. 2003, 2004; Reid et al. 2005a,b, 2007; Zhang et al. 2006). In the Yidun block, the emplacement of the Cretaceous granites (Chola Shan granite; Reid et al. 2007) does not affect the thermal history of the Indosinian plutons (Fig. 6c). Furthermore, after a very rapid cooling following their emplacement, those younger granites reach the same temperature of stability as the Indosinian ones. This confirms that the emplacement of the Cretaceous granites is not related to a tectonic episode and that they cannot be used to support a Cretaceous Yenshan tectonic event in Kunlun, Yidun or Songpan Garzeˆ (Dirks et al. 1994; Chen et al. 1995; Arne et al. 1997; Arnaud et al. 2003). On the northern side of the Qinling ranges to the NE, fission track and Kfeldspar 40Ar/39Ar data indicate rapid cooling and local strong exhumation (over 4 km) during Late Cretaceous, mostly associated with local basin formation and strike –slip fauling (Ratschbacher et al. 2003; Enkelmann et al. 2006). Further east, in Dabie Shan, apatite fission track ages are also Late Cretaceous. Finally, Vincent & Allen (1999) reported evidences of local Late Jurassic –Early Cretaceous north–south transtensive movements in the Hexi corridor along the eastern edge of the Qilian Shan. Evidences for Cretaceous exhumation and tectonic movements are thus obviously found east and NE of the actual Tibet plateau. However Cretaceous exhumation is generally localized, mostly associated with strike –slip faults and smallscale extensional basins. The Cretaceous sedimentation in the Sichuan basin is represented by lacustrine and aeolian series indicative of a general low sediment input (see synthesis in

17

Richardson et al. 2008). All these observations associated with the cooling curves presented above seem to indicate that, if the Yenshan event did exist, it only affected eastern China and not the Tibet plateau area. In that respect it is difficult to link this event to the Lhasa –Qiangtang collision.

Numerical modelling The thermal structure of the lithosphere depends on many factors, including the boundary conditions, surface erosion and material properties. Here we only focus our approach on the very slow cooling rate between Late Jurassic and Early Cenozoic obtained from granites of South Songpan-Garzeˆ and Kunlun. Assuming negligible horizontal heat transfer, we use a simple 1D thermal modelling that takes into account the thermal properties of both sediments and crust within these areas (Fig. 7). Furthermore we assume that the dominant thermal processes are radiogenic heat production and conductive heat transport towards the surface. The temperature T as a function of time t and depth z is thus given by the following transient heat equation  2  @T 1 @ T @T ¼ k 2 þ A  vz , r CP @t @z @z

(1)

Fig. 7. Physical model setup showing the boundary conditions and material properties used in this study. T0 is the surface temperature. qm is the upward heat flux at the base of the model. As, Ac, A0, hr, ks and hs are the sedimentary heat production, the crustal heat production, the heat production at the top of the crust, the exponential decay of the crustal heat production, the sedimentary thermal conductivity and the sediment thickness, respectively.

18

F. ROGER ET AL.

where r is the density, CP the heat capacity, k the thermal conductivity, A the radiogenic heat production and vz the vertical advection associated to exhumation. Here we study a long time period (at least 100 Ma) of very low cooling rate. The lack of post-Triassic sediments in the Songpan-Garzeˆ area, the occurrence of a complete Jurassic to Quaternary section within the present-day surrounding basins (Sichuan and Qaidam), and a Jurassic to Cretaceous sequence on the Qiangtang block suggest that this low cooling rate can be related to very low erosion rates. In the following we consider an end-member model with a steady state temperature field (@T/@t ¼ 0) and no denudation processes (vz ¼ 0). The equation 1 becomes 0¼k

d2 T þ A, dz2

(2)

that can be easily solved to interpret the Cretaceous temperature field. For sediments we assume a constant thermal conductivity ks ranging from 1–5 W m21 K21 and we take constant heat production As between 0 –5 mW m23 Following Turcotte & Schubert (2002) we assume that crustal heat production decreases exponentially with depth Ac ¼ A0 e(zhs )hr

(3)

with hs the thickness of the sedimentary layer, A0 the radiogenic heat production at the top of the crust (z ¼ hs) and hr a length scale for the decrease in Ac with depth. We take A0 ¼ 0–5 mW m23 and hr ¼ 5–15 km (Turcotte & & Schubert 2002). Integration of equation 2 using the boundary conditions T0 ¼ 0 8C at z ¼ 0 km and upward heat flux qm ¼ 15–30 mW m22 at great depth (z ! þ 1) gives for temperature field within the sedimentary layer T¼

1 As 2 qm þ A0 hr þ As hs z þ z þ T0 : ks 2 ks

(4)

First for granites of south Songpan-Garzeˆ we use this last equation to calculate the depth of both Markam mica granite and Manai and Rilonguan amphibole granites associated with Cretaceous temperatures of 300 –400 8C and 450 –550 8C, respectively (Roger et al. 2004). Based on earlier studies (Xu et al. 1992; Calassou 1994) we assume a sediment thickness hs ¼ 15–30 km. We obtained two Gaussian distributions of depth with means m300 – 4008C ¼ 12.5 km and m450 – 5508C ¼ 20 km associated with standard deviations s300 – 4008C ¼ 3 km and s450 – 5508C ¼ 5 km, respectively. Next, using these distributions we estimate the a

posteriori distribution of thermal parameters (Fig. 8). A wide range of modelling gives a good agreement between estimated and calculated temperature. We show that crustal parameters A0, hr as well as qm are poorly constrained. Not surprisingly, our result confirms that thermal properties of sediments have a primary control on the temperature field within the sedimentary layer. We obtain high heat production As ¼ 3–4 mW/m3, high thermal conductivity ks ¼ 3.5 –4.5 W/m and a sediment thickness of at least 23 km, which is consistent with the depth of emplacement of granites inferred from metamorphic studies (Xu et al. 1992). A similar approach is used to interpret Kokoxili and Xidatan granites along the western Kunlun fault, which gives Cretaceous temperature of 200– 300 and 300– 450 8C, respectively (Fig. 6). As for the Songpan-Garzeˆ, the calculated temperatures are mostly controlled by sedimentary properties. Thus our inversion cannot be used to estimate crustal parameters (Fig. 9). For sediment more than 65% of the models give a high heat production As . 3 –4 mW/m3 and a thermal conductivity ks ¼ 2–3.5 W/m. The calculated distribution of sediment thickness shows that hs is mostly controlled by the a priori information on depth, which gives a pressure ,3.5 kbar for metasediments (Harris et al. 1988). Together the results obtained from the SongpanGarzeˆ and Kunlun granites suggests that: (1) high temperature can be maintained over a long period of time; and (2) Cretaceous cooling temperature is mostly controlled by the thermal properties of sediments of continental origin.

Discussion and tectonic implications Our results demonstrate that a reasonable combination of high thermal conductivity and high heat production in the sediments is able to account for the observed Cretaceous thermal quiescence in the south Songpan-Garzeˆ terrane as well as in western Kunlun (Fig. 10). It demonstrates that whatever the nature of the crust, high thermal steady state can be maintained over such a long time period of thermal stability. The continental nature of the underlying crust within the southern part of the Songpan-Garzeˆ belt is widely accepted. The geochemistry of the I-type and adakitic-type granites clearly demonstrates that they are essentially derived from partial melting of a Neoproterozoic crystalline basement similar to the basement of the Yangtze craton (Roger et al. 2004; Zhang et al. 2006). However, within the northern part of the prism, east of the Longmen Shan, the nature of the crust is still debated as no outcrop is available (Zhang et al. 2007; Roger et al. 2008, 2010). We recently proposed that the peculiar tectonic

THERMAL EVOLUTION OF EASTERN TIBET

19

Fig. 8. A posteriori distribution of thermal parameters obtained from the inversion of Cretaceous temperature estimated in south of Songpan-Garzeˆ. As, A0, hr, ks, hs and qm are respectively the sedimentary heat production, the heat production at the top of the crust, the exponential decay of the crustal heat production, the sedimentary thermal conductivity, the sediment thickness and the upward heat flux at the base of the model. The range of each parameter is given a priori from available geological and metamorphic data.

and geodynamic setting of the Songpan-Garzeˆ basin and surrounding blocks did not allow a complete collision between the main continental domains (Roger et al. 2008, 2010). The volumes of Triassic sediments deposited in the basin are huge and the series are exceptionally thick (up to 25 km). Associated to the triangular geometry of the convergence zone between the Qiangtang, south China and Kunlun/north China blocks that thickened the series it led, during the orogenesis, to the build-up of a wide accretionary orogen. However we propose that the converging system gets blocked before the complete subduction of the SongpanGarzeˆ oceanic crust. We currently do not have enough data to build up the complete cooling curve for any granite in north Songpan-Garzeˆ. However, our physical modelling shows that the nature of the underlying crust has no or minor effect in the thermal evolution of the samples. This implies that the shape of the obtained cooling curve in north Songpan-Garzeˆ and in the other

parts of the study area should be similar. The only unknown parameter remains the depth of the granites during the thermal stability period. Furthermore our modelling suggests that the main parameter controlling the temperature stability is the sediment thickness that must remain constant over a long period of time (100 Ma). This supports the assumption that sedimentation and erosion were not active in the area from Late Triassic –Early Tertiary. The 20 –25 km of sediments required to maintain the Manai and Rilonguan samples at a temperature of 500 8C during the period of thermal stability correspond to the generally accepted sediment thickness within the Songpan-Garzeˆ prism (Mattauer et al. 1992; Xu et al. 1992; Calassou 1994; Roger et al. 2008, 2010) (Fig. 8). As already suggested above, the lower temperature of stability of the Markam granite (300 8C) is explained by the position of the granite further up in the sedimentary sequence. In the Kunlun belt the temperature of stability

20

F. ROGER ET AL.

Fig. 9. Same as Figure 8 for western Kunlun.

of the samples is c. 300 –400 8C similar to the temperature observed in Markam. In the Yidun block, the temperature of stability is much lower, c. 100 8C suggesting a thinner sedimentary cover (Figs 6 & 10).

Late Tertiary, localized reactivation of the exhumation The Tertiary deformation that affected the SongpanGarzeˆ prism, and especially the growth of the Longmen Shan range is difficult to date and characterize. From a tectonic and metamorphic point of view, the Tertiary deformation was strongly influenced by the Indosinian and older structures, so that it is difficult to unravel the effects of the Cenozoic deformation from the Triassic deformation (e.g. Calassou 1994; Roger et al. 2008). Along the Longmen Shan, the large drop in altitude between the 4000 m high plateau and the 500 m high Sichuan basin is correlated with a sharp variation in the Moho depth, from 55 to 60 km below the plateau to 40 km in the Sichuan basin (Wang et al.

2007; Xu et al. 2007; Robert et al. 2010). While part of this structure must be correlated to the Late Miocene uplift (e.g. Clark et al. 2005b) part of it could be inherited from the Triassic orogeny. Near Danba, the Triassic de´collement level is refolded by a large antiformal structure NNW– SSE axis that could correspond to Tertiary folding. Similarly within the Longmen Shan, the NE – SW-trending faults that separate the plateau from the Sichuan basin have absorbed significant shortening between Late Triassic and present (Chen & Wilson 1996). Finally, activity along the major Xianshui He strike–slip fault is attested by shearing of the syn-tectonic, 13 Ma old Konga Shan granite (Roger et al. 1995) (Fig. 11). Low temperature thermochronology (fission tracks and U –Th/He) data available in south Songpan-Garzeˆ, Yidun and Longmen Shan (Figs 6 & 11) indicate that the final exhumation and cooling occurred in the Tertiary after c. 30 –25 Ma (Arne et al. 1997; Xu & Kamp 2000; Jolivet et al. 2001; Kirby et al. 2002; Huang et al. 2003; Roger et al. 2004; Wilson et al. 2006; Lai et al. 2007). This exhumation episode can be separated in two

THERMAL EVOLUTION OF EASTERN TIBET

Fig. 10. Temperature field at depth within the sedimentary layer. Grey scale gives the ‘likelihood’ of the calculated geotherm. Top: Black squares give the estimated temperature of mica granite and amphibole granite of south Songpan-Garzeˆ. Bottom: Black squares give the estimated temperature of Kokoxili and Xidatan granites of western Kunlun.

phases: a regional moderate cooling starting c. 25 Ma followed by a second, stronger phase initiating between c. 10 and 5 Ma (Roger et al. 1995; 2004; Arne et al. 1997; Xu & Kamp 2000; Kirby et al. 2002; Wilson et al. 2006; Zheng et al. 2006; Lai et al. 2007). Figure 11 displays a compilation of the available apatite fission track data grouped in age intervals. The resulting image indicates that the Tertiary exhumation is restricted along the major tectonic features such as the Xianshui He fault or the Longmen Shan range. Away from these structures the Tertiary exhumation is nearly null or at least too weak to have been registered by the thermochronometer. For example, the Garzeˆ

21

and Daocheng granite plutons have apatite fission track ages older than 100 Ma (Lai et al. 2007) implying that these areas were not or only weakly affected by post-Jurassic exhumation (Fig. 11). Furthermore, the ages obtained from the Daocheng pluton tend to decrease while approaching the Garzeˆ and Litang thrust faults to the east. This trend is further confirmed by the ages between 10 –25 Ma obtained immediately along the active Garzeˆ and Litang strike– slip faults (Lai et al. 2007) (Fig. 11). Within the Min Shan, the apatite fission track ages range between 30– 100 Ma with the exception of one sample with an age of 122 + 12 Ma near Songpan (Arne et al. 1997). Like in the Garzeˆ –Litang area, the ages tend to be younger close to the prominent tectonic features such as the Min Shan thrust fault or the edge of the northern Longmen Shan range. (U –Th)/He data indicate that thrusting along the Min Shan thrust fault, west of Songpan is thought to initiate c. 5–3 Ma leading to renewed erosion (Kirby et al. 2002). We thus interpret the fission track ages in that region as intermediate between the Mesozoic ages obtained on the Garzeˆ and Daocheng plutons as well as in Songpan, and the truly Tertiary ages associated to the last exhumation episode. The onset of thrusting along the Min Shan fault is coeval with the second phase of increased exhumation that is only observed (with the exception of 2 samples near the Yushu-Batang suture zone) along the easternmost major tectonic structures (Xianshui He fault, Jinsha thrust, Longmen Shan thrust system and the Songpan-Garzeˆ Triassic de´collement fault) (Roger et al. 1995, 2004; Arne et al. 1997; Xu & Kamp 2000; Kirby et al. 2002; Wilson et al. 2006; Lai et al. 2007). This last exhumation episode is confirmed by the few (U –Th)/He data available along the Longmen Shan (Kirby et al. 2002; Godard et al. 2009). Within the Longmen Shan range, a total denudation of 7 –10 km is estimated for the late Cenozoic period (Arne et al. 1997; Kirby et al. 2002; Clark et al. 2005a,b). Similar amounts of late Tertiary denudation have been estimated along an east –west section across the Xianshui He fault near Kangding: west of the fault, the denudation reaches 4–6 km east of the fault and 7 –10 km west of the fault (Xu & Kamp 2000). Figure 11 also shows that the Triassic de´collement level is reactivated leading to the exhumation of the Manai and Rilonguan granites. Similarly, however based on very few data, thrusting along the Batang suture zone may also be reactivated probably because these north–south trending faults are perpendicular to the regional direction of compression. This is again indicative of the strong influence of the inherited structures on the ongoing deformation. The second exhumation phase is not imaged along the western half of the Xianshui He

22

F. ROGER ET AL.

Fig. 11. Geological map of the Songpan-Garzeˆ belt (modified after Roger et al. 2008, 2010 and C.I.G.M.R. 1991) showing the fission track ages (apatite) available for the Triassic magmatic and sedimentary rocks grouped into time spend corresponding to the main geodynamic phases. Individual data are indicated below the general envelopes. See text for discussion. Data are from Arne et al. (1997), Xu & Kamp (2000), Reid et al. (2005a), Zheng et al. (2006) and Lai et al. (2007).

fault or along the Litang and Garzeˆ faults, probably because these are nearly pure strike–slip and do not create much relief. The Tertiary exhumation (caused by compressional tectonics) thus appears mostly concentrated along the limit between the plateau and the Sichuan basin, the former being squeezed on the later since c. 30 –25 Ma with a increased activity since less than 10 Ma. The internal part of the Songpan-Garzeˆ prism is poorly exhumed as Mesozoic low temperature thermochronology ages are preserved, probably due to a very distributed Tertiary shortening (and thickening) in most part of the huge Triassic sedimentary accretionary belt

that constituted the eastern part of the Tibetan Plateau.

Conclusion Complete cooling curves of basement rocks and syn- to post-tectonic granite throughout the whole north-eastern Tibet area demonstrate that, while the Indosinian orogeny that led to the building of the Songpan-Garzeˆ thrust and fold belt was a major geodynamic episode, it has been followed by a very long period of stability. For c. 100 Ma, between Late Jurassic –Middle Tertiary, deformation as well as erosion and sedimentation were very

THERMAL EVOLUTION OF EASTERN TIBET

low and distributed. This implies that the accretion of the Lhasa block during the Cretaceous had, if any, only a very limited effect on the tectonic, metamorphic and topographical structure of north and north-eastern Tibet. Between the end of the Indosinian orogeny and the onset of the Tertiary deformation at c. 30 Ma, the Songpan-Garzeˆ area remained probably a flat and poorly drained region. While we have no indication of its absolute elevation, the lack of Jurassic – Tertiary sediments suggests that the Triassic belt already constituted a plateau higher than the surrounding basins. Finally the low temperature thermochronological data indicate that the Tertiary deformation allows exhumation only where it is strongly localized, i.e. along inherited tectonic structures. This result underlines the need of studying the pre-Tertiary tectonic evolution to understand the present-day deformation pattern and tectonic activity. The authors would like to thank all the colleagues involved in the field expeditions which took place between 1990 and 2000, particularly the leaders Z. Q. Xu, P. Tapponier and M. Mattauer. Most of these studies were supported by INSU– CNRS funding. Special thanks go to E. Kirby and an anonymous reviewer for their comments and suggestions.

References Allen, C. R., Qian, H., Wen, X., Zhou, H. & Huang, W. 1991. Field study of a highly active fault zone: the Xianshui He fault of Southwestern China. Geological Society of America Bulletin, 103, 1178–1199. Ames, L., Tilton, G. R. & Zhou, G. Z. 1993. Timing of collision of the Sino-Korean and Yangtse cratons: U–Pb zircon dating of coesite-bearing eclogites. Geology, 21, 339– 342. Ames, L., Zhou, G. & Xiong, B. 1996. Geochronology and isotopic character of ultrahigh-pressure metamorphism with implications for collision of the SinoKorean and Yangtze craton, central China. Tectonics, 15, 472–489. Arnaud, N., Tapponnier, P., Roger, F., Brunel, M., Scha¨rer, U., Chen, W. & Xu, Z. 2003. Evidence for Mesozoic shear along the western Kunlun and Altyn-Tagh fault, northern Tibet (China). Journal of Geophysical Research, 108, 2053– 2080, doi: 10.1029/2001JB000904. Arne, D., Worley, B. et al. 1997. Differential exhumation in response to episodic thrusting along the eastern margin of the Tibetan Plateau. Tectonophysics, 280, 239–256. Burchfield, B. C., Chen, Z., Liu, Y. & Royden, L. H. 1995. Tectonics of the Longmen Shan and adjacent regions, Central China. International Geology Review, 37, 661– 735. Calassou, S. 1994. Etude tectonique d’une chaıˆne de de´collement: (a) tectonique triasique et tertiaire de la chaıˆne de Songpan-Garzeˆ. (b) ge´ome´trie et

23

cine´matique des de´formations dans les prismes d’accre´tion se´dimentaires: mode´lisation analogique. PhD thesis, University of Montpellier II, France. Chen, S. F. & Wilson, C. J. L. 1996. Emplacement of the Longmen Shan Thrust-Nappe Belt along the eastern margin of the Tibetan Plateau. Journal of Structural Geology, 18, 413–440. Chen, S. F., Wilson, C. J. L. & Worley, B. A. 1995. Tectonic transition from the Songpan-Garzeˆ fold belt to the Sichuan basin, south-western China. Basin Research, 7, 235– 253. C.I.G.M.R. (Chengdu Institute of Geology and Mineral Resources) 1991. Geological Map of Xinghai-Xizang (Tibet) Plateau and adjacent areas (1:1 500 000) and Explanatory Note. Chengdu Cartographic Press, Chengdu. Clark, M. K., Bush, J. W. M. & Royden, L. H. 2005a. Dynamic topography produced by lower crustal flow against rheological strength heterogeneities bordering the Tibetan Plateau. Geophysical Journal Interior, 162, 575–590. Clark, M. K., House, M. A., Royden, L. H., Whipple, K. X., Burchfield, B. C., Zhang, X. & Tang, W. 2005b. Late Cenozoic uplift of eastern Tibet. Geology, 33, 525–528, doi: 10.1130/G21265.1. Dirks, P. H. G., Wilson, C. J. L., Chen, S., Luo, Z. L. & Liu, S. 1994. Tectonic evolution of the NE margin of the Tibetan Plateau: evidence from the central Longmen Moutains, Sichuan Province. Journal of Southeast Asian Earth Sciences, 9, 181–192. Dodson, M. H. 1973. Closure temperature in cooling geochronological and petrological system. Contribution to Mineralogy and Petrology, 40, 259–274. Eide, E. A., Mcwillians, M. O. & Liou, J. G. 1994. 40 Ar/39Ar geochronology and exhumation of highpressure to ultrahigh pressure metamorphic rocks in east-central China. Geology, 22, 601– 604. Enkelmann, E., Ratschbacher, L. et al. 2006. Cenozoic exhumation and deformation of northeastern Tibet and the Qinling: Is Tibetan lower crustal flow diverging around the Sichuan Basin? GSA Bulletin, 118, 651–671; doi: 10.1130/B25805.1. Godard, V., Pik, R. et al. 2009. Late Cenozoic evolution of the central Longmen Shan, eastern Tibet: Insight from (U-Th)/He thermochronometry. Tectonics, 28, 5, TC5009, doi: 10.1029/2008TC002407. Hacker, B. R., Ratschbacher, L., Webb, L., Ireland, T., Walker, D. & Shuwen, D. 1998. U/Pb zircon ages constrain the architecture of the ultrahighpressure Qinling-Dabie Orogen, China. Earth and Planetary Science Letters, 161, 215–230. Hacker, B., Wallis, S., Ratschbacher, L., Grove, M. & Gehrels, G. 2006. High-temperature geochronology constraints on the tectonic history and architecture of the ultrahigh-pressure Dabie-Sulu orogen. Tectonics, 25, 1 –17. Harris, N. B. W., Xu, R., Lewis, C. L. & Jin, C. 1988. Plutonic rocks of the 1985 Tibet Geotraverse, Lhasa to Golmud. Philosophical Transactions of the Royal Society of London, Series A, 327, 145 –168. Harrowfield, M. J. & Wilson, C. J. L. 2005. Indosinian deformation of the Songpan Garzeˆ fold belt, northeast Tibetan plateau. Journal of Structural Geology, 27, 101– 117.

24

F. ROGER ET AL.

Hsu¨, K. J., Wang, Q., Li, J., Zhou, D. & Sun, S. 1987. Tectonic evolution of Qinling Mountains, China. Eclogae Geologicae Helvetiae, 80, 735– 752. Huang, M., Maas, R., Buick, I. S. & Williams, I. S. 2003. Crustal response to continental collisions between the Tibet, Indian, South China and North China blocks: geochronological constraints from the Songpan-Garzeˆ orogenic belt, western China. Journal of Metamorphic Geology, 21, 223–240. Huang, W. & Wu, Z. W. 1992. Evolution of the Qin ling orogenic belt. Tectonics, 11, 371 –380. Jolivet, M., Brunel, M. et al. 2003. Neogene extension and volcanism in the Kunlun fault zone, northern Tibet: New constraints on the age of the Kunlun fault. Tectonics, 20, 1052, doi: 10.1029/ 2002TC001428. Jolivet, M., Brunel, M. et al. 2001. Mesozoic and Cenozoic tectonics of the northern edge of the Tibetan Plateau; fission-track constraints. Tectonophysics, 343, 111–134. Kapp, P., Yin, A. et al. 2000. Blueschist-bearing metamorphic core complexes in the Qiangtang block reveal deep crustal structure of the northern Tibet. Geology, 28, 19–22. Kirby, E., Reiners, P. W. et al. 2002. Late Cenozoic evolution of the eastern margin of the Tibetan plateau: Inferences from 40Ar/39Ar and (U–Th)/He thermochronology. Tectonics, 21, doi: 10.1029/ 2000TC001246. Lai, Q. Z., Ding, L., Wang, H. W., Yue, Y. H. & Cai, F. L. 2007. Constraining the stepwise migration of the eastern Tibetan Plateau margin by apatite fission track thermochronology. Science in China Series D: Earth Sciences, 50, 172– 183. Li, S., Kusky, T. et al. 2007. Collision leading to multiple-stage extrusion in the Qinling orogen: insights from the Mianlue suture. Gondwana Research, 12, 121– 143. Malavieille, J., Mattauer, M., Lancelot, J., Calassou, S., Xu, Z. & Hao, Z. 1991. Large scale de´collement in the Indosinian (Triassic) fold belt of Eastern Tibet. Terra Abstracts, 3, 258. Mattauer, M., Malavieille, J. et al. 1992. La chaıˆne triasique de Songpan-Garzeˆ (ouest Sechuan et Est Tibet); une chaıˆne de plissement-de´collement sur marge passive. Translated title: The Songpan-Garzeˆ Triassic belt of western Sichuan and eastern Tibet; a de´collement-fold belt on a passive margin. Comptes Rendus de l’Acade´mie des Sciences Paris, 314, 619– 626. Nie, S., Yin, A., Rowley, D. B. & Jin, Y. 1994. Exhumation of the Dabie Shan ultra-high-pressure rocks and accumulation of the Songpan-Ganzi flysch sequence, central China. Geology, 22, 999–1002. Pullen, A., Kapp, P., Gehrels, G., Vervoot, J. D. & Ding, L. 2008. Triassic continental subduction in central Tibet and Mediterranean-style closure of the Paleo-Tethys Ocean. Geology, 36, 351– 354. Ratschbacher, L., Hacker, B. R. et al. 2003. Tectonics of the Qinling (Central China): Tectonostratigraphy, geochronology and deformation history. Tectonophysics, 366, 1 –53, doi: 10.1016/S0040– 1951(03)00053-2.

Rao, R. B., Xu, J. F., Chen, Y. & Zhou, D. 1987. The Triassic System of the Qinghai-Xizang Plateau. Geological Publishing House, Beijing. Reid, A. J., Fowler, A. P., Phillips, D. & Wilson, C. J. L. 2005a. Thermochronology of the Yidun arc, central eastern Tibetan Plateau: constraints from 40 Ar/39Ar K-feldspar and apatite fission track data. Journal of Asian Earth Sciences, 25, 915– 935. Reid, A. J., Wilson, C. J. L. & Liu, S. 2005b. Structural evidence for the Permo-Triassic tectonic evolution of the Yidun arc, eastern Tibetan plateau. Journal of Structural Geology, 27, 119 –137. Reid, A., Wilson, C. J. L., Shun, L., Pearson, N. & Belousova, E. 2007. Mesozoic plutons of the Yidun arc, SW China: U/Pb geochronology and Hf isotopic signature. Ore Geology Reviews, 34, 88–106. Richardson, N. J., Densmore, A. L. et al. 2008. Extraordinary denudation in the Sichuan basin: Insights from low-temperature thermochronology adjacent to the eastern margin of the Tibetan Plateau. Journal of Geophysical Research, 113, B044409, doi: 10.1029/ 2006JB004739. Robert, A., Zhu, J. et al. 2010. Crustal structures in the area of the 2008 Sichuan Earthquake from geophysics and geology. Tectonophysics, 491, 205– 210. Roger, F., Calassou, S. et al. 1995. Miocene emplacement and deformation of the Konga Shan granite (Xianshui He fault, west Sichuan, China): Geodynamic implications. Earth and Planetary Science Letters, 130, 201 –216. Roger, F., Arnaud, N. et al. 2003. Geochronological and geochemical constraints on Mesozoic suturing in East Central Tibet. Tectonics, 22, 1037, doi: 10.1029/2002TC001466. Roger, F., Malavieille, J., Leloup, P. H., Calassou, S. & Xu, Z. 2004. Timing of granite emplacement and cooling in the Songpan-Garzeˆ fold belt (eastern Tibetan Plateau) with tectonic implications. Journal of Asian Earth Sciences, 22, 465 –481. Roger, F., Jolivet, M. & Malavieille, J. 2008. Tectonic Evolution of the Triassic Fold Belts of Tibet. Comptes Rendus Geoscience, Acade´mie des Sciences, Paris, 340, 180 –189. Roger, F., Jolivet, M. & Malavieille, J. 2010. The tectonic evolution of the Songpan Garzeˆ (North Tibet). Journal of Asian Earth Sciences, 39, 254– 269. Rowley, D. B., Xue, F., Tucker, R. D., Peng, Z. X., Baker, J. & Davis, A. 1997. Ages of ultrahigh pressure metamorphism and protolith orthogneisses from the eastern Dabie Shan: U/Pb zircon geochronology. Earth and Planetary Science Letters, 151, 191– 203. Sengo¨r, A. M. C. 1985. East Asian tectonics collage. Nature, 318, 15– 17. Turcotte, D. L. & Schubert, G. 2002. Geodynamics. Cambridge University Press, Cambridge. Vincent, S. J. & Allen, M. B. 1999. Evolution of the Mule and Chaoshui basins, China; implications for Mesozoic strike– slip basin formation in Central Asia. Geological Society of America Bulletin, 111, 725–742. Wallis, S., Tsujimori, T., Aoya, M., Kawakami, T., Terada, K., Suzuki, K. & Hyodo, H. 2003. Cenozoic and Mesozoic metamorphism in the Longmenshan

THERMAL EVOLUTION OF EASTERN TIBET orogen: Implications for geodynamic models of eastern Tibet. Geology, 31, 745–748. Wang, C. Y., Han, W. B. et al. 2007. Crustal structure beneath the eastern margin of the Tibetan Plateau and its tectonic implications. Journal of Geophysical Research, 112, B07307; doi: 10.1029/2005JB003873. Wilson, C. J. L., Harrowfield, M. J. & Reid, A. J. 2006. Brittle modification of Triassic architecture in eastern Tibet: implications for the construction of the Cenozoic plateau. Journal of Asian Earth Sciences, 27, 341–357. Xiao, L. & Clemens, J. D. 2007. Origin of potassic (C-type) adakite magmas: experimental and field constraints. Lithos, 95, 399– 414. Xiao, L., Zhang, H. F. et al. 2007. Late Triassic granitoids of the eastern margin of the Tibetan Plateau: Geochronology, petrogenesis and implications for tectonic evolution. Lithos, 96, 436– 452. Xu, G. & Kamp, P. J. J. 2000. Tectonics and denudation adjacent to the Xianshuihe Fault, eastern Tibetan Plateau: Constraints from fission track thermochronology. Journal of Geophysical Research, 105, 19 231 –19 251. Xu, Z. Q., Hou, L. & Wang, Z. 1992. Orogenic Processes of the Songpan-Garzeˆ Orogenic Belt of China. Geological Publishing House, Beijing. Xu, L., Rondenay, S. & Van Der Hilst, R. D. 2007. Structure of the crust beneath the southeastern Tibetan Plateau from teleseismic receiver functions. Physics of the Earth and Planetary Interiors, 165, 176–193. Zhang, C. Z., Li, B., Cai, J. X., Tang, X. C., Wei, Q. G. & Zhang, Y. X. 2008a. A-type granite and adakitic magmatism association in Songpan-Garze fold belt, eastern Tibetan Plateau: Implication for lithospheric delamination: comment. Lithos, 103, 562– 564. Zhang, H., Zhang, B., Harris, N., Zhang, L., Chen, Y., Chen, N. & Zhao, Z. 2006. U– Pb zircon SHRIMP

25

ages, geochemical and Sr–Nd–Pb isotopic compositions of intrusive rocks from the Longshan-Tianshui area in the southeast corner of the Qilian orogenic belt, China: Constraints on petrogenesis and tectonic affinity. Journal of Asian Earth Sciences, 27, 751– 764. Zhang, H. F., Parrish, R., Zhang, L., Xu, W. C., Yuan, H. L., Gao, S. & Crowley, Q. G 2007. A-type granite and adakitic magmatism association in Songpan-Garze fold belt, eastern Tibetan Plateau: Implication for lithospheric delamination. Lithos, 97, 323– 335; doi: 10.1016/j.lithos.2007.01.002. Zhang, H. F., Parrish, R., Zhang, L., Xu, W. C., Yuan, H. L., Gao, S. & Crowley, Q. G. 2008b. Reply to the comment by Zhang et al. on: ‘First finding of A-type and adakitic magmatism association in Songpan-Garze fold belt, eastern Tibetan Plateau: implication for lithospheric delamination. Lithos, 103, 565–568. Zheng, D., Zhang, P-Z. et al. 2006. Rapid exhumation at 8 Ma on the Liupan Shan thrust fault from apatite fission-track thermochronology: Implications for growth of the northeastern Tibetan Plateau margin. Earth and Planetary Science Letters, 248, 198–208. Zhou, M. F., Yan, D. P., Vasconcelos, P. M., Li, J. W. & Hu, R. Z. 2008. Structural and geochronological constraints on the tectono-thermal evolution of the Danba domal terrane, eastern margin of the Tibetan plateau. Journal of Asian Earth Sciences, 33, 414– 427. Zi, J., Fan, W., Wang, Y., Peng, T. & Guo, F. 2008. Geochemistry and petrogenesis of the Permian mafic dykes in the Panxi region, SW China. Gondwana Research, 14, 368 –382. Zou, D. B., Chen, L. K., Rao, R. B. & Chen, Y. M. 1984. On the Triassic turbidite in the southern Bayan-Har mountain region. Contribution to the Geology of the Qinghai-Xizang (Tibet) Plateau, 15, 27–39.

Early Cretaceous Gangdese retroarc foreland basin evolution in the Selin Co basin, central Tibet: evidence from sedimentology and detrital zircon geochronology QING-HAI ZHANG1,3, LIN DING1*, FU-LONG CAI1,2, XIAO-XIA XU1, LI-YUN ZHANG1, QIANG XU1 & HELMUT WILLEMS3 1

Key Laboratory of Continental Collision and Plateau Uplift, Institute of Tibetan Plateau Research, Chinese Academy of Sciences, 100085 Beijing, China

2

Institute of Geology and Geophysics, Chinese Academy of Sciences, 100029 Beijing, China 3

University of Bremen, Department of Geosciences, D-28359 Bremen, Germany *Corresponding author (e-mail: [email protected]) Abstract: The Selin Co basin in the northern Lhasa terrane includes more than 3000 m of upward coarsening Lower Cretaceous strata, and the sedimentary sequence from the flysch to the molasse indicates the evolution of a foreland basin. Petrographic analysis shows that sandstones are rich in volcanic and sedimentary lithics and most of them fall into recycled orogen and magmatic arc. Uranium– lead (U– Pb) ages were determined for 435 detrital zircons from the Lower Cretaceous strata in the Selin Co basin. Relative probability of detrital zircon ages shows the Eshaerbu Formation was rich in zircon grains with the age of 125–140 and 160–180 Ma, and the Duoni Formation was dominated by one main age cluster of 125– 150 Ma. Analysis of the potential provenances suggests the Early Cretaceous zircon grains were primarily derived from the Gangdese magmatic arc to the south. The youngest zircon ages in the lowermost exposure of the Eshaerbru Formation are c. 130 Ma, providing a maximum depositional age of sediments in the Selin Co basin. Collectively, our studies, together with previously documented Cretaceous thrusting in the Lhasa terrane, suggest the Lower Cretaceous Selin Co basin was deposited in a retroarc foreland basin. From 145–90 Ma, a retroarc foreland basin was presumed to develop in the Lhasa terrane, migrating from the south to the north. Crustal thickening, likely associated with the evolution of the retroarc foreland basin, was speculated to start in the Early Cretaceous in the Lhasa terrane.

A fundamental question in global tectonics is to determine when the Tibetan crust began to experience large-scale crustal thickening and the mechanisms that caused it. One hypothesis postulates the thickened crust of the Tibetan Plateau formed predominantly from the continental collision between India and Asia (Dewey et al. 1988; Molnar et al. 1993; Zhang 2000; Zhang et al. 2004). The alternative hypothesis suggests that significant shortening and crustal thickening had occurred prior to the India –Asia collision (England & Searle 1986; Murphy et al. 1997; Ding & Lai 2003; Kapp et al. 2005b; Leier et al. 2007a). These two opposing models mainly result from different interpretations of the evolution of Cretaceous basins within the Lhasa terrane. In the Cretaceous, the Lhasa terrane was occupied generally by the Gangdese arc to the south and sedimentary basins to the north. The Selin Co (Co ¼ Lake) basin in the northern Lhasa

terrane provides an ideal setting to investigate the tectonic history of central Tibet just prior to the India–Asia collision (Fig. 1), taken to have initiated at 50 + 15 Ma (Searle et al. 1987; Yin & Harrison 2000; Aitchison et al. 2002; Ding et al. 2005). This study was conducted to identify the provenance of the sediment in the Selin Co basin, and then to explain the Early Cretaceous evolution of the Lhasa terrane. Based on the measured stratigraphic sections and U– Pb detrital zircon geochronology, we found that the sediments deposited in the Selin Co basin were primarily derived from the Gangdese magmatic arc and the Carboniferous strata to the south and partly from the Qiangtang terrane to the north. Therefore, our results suggest that a retroarc thrust belt and foreland basin were developed in the Lhasa terrane during c. 145– 90 Ma, which likely served to thicken the Lhasa terrane crust prior to the India–Asia collision.

From: Gloaguen, R. & Ratschbacher, L. (eds) Growth and Collapse of the Tibetan Plateau. Geological Society, London, Special Publications, 353, 27– 44. DOI: 10.1144/SP353.3 0305-8719/11/$15.00 # The Geological Society of London 2011.

28 Q.-H. ZHANG ET AL. Fig. 1. Sketched geological map of the Lhasa terrane modified from Pan et al. (2004) and Kapp et al. (2005b). Abbreviations: IYS, Indus-Yarlung suture; BNS, Bangong suture (Bangong-Nujiang suture); JS, Jinsha suture; KS, Kunlun suture; SGAT, Shiquanhe-Gaize-Amdo thrust system; GST, Gaize-Selin Co thrust; GJT, Geren Co-Jiali thrust; GT, Gangdese thrust system; GCT, Great Counter thrust.

GANGDESE RETROARC FORELAND BASIN EVOLUTION

Geological background Since the Late Palaeozoic the Lhasa terrane started to rift from Gondwana, drifted northward, and eventually collided with Eurasia in the Late Jurassic – Early Cretaceous (Alle`gre et al. 1984; Dewey et al. 1988; Yin & Harrison 2000). The rifting between Lhasa terrane and India continent caused the formation of the Neo-Tethys Ocean. During the Cretaceous, Neo-Tethys oceanic lithosphere subducted northward beneath the Lhasa terrane, generating the Gangdese magmatic arc in the southern Lhasa terrane (Tapponnier et al. 1981; Alle`gre et al. 1984; Chang et al. 1986). Contemporarily, thick sequences of marine and non-marine strata were accumulated mainly in the northern Lhasa terrane (Yu & Wang 1990; Xiao & Li 2000). However, there are significant speculations about the Cretaceous tectonic setting of the Lhasa terrane. The Cretaceous strata are interpreted to have been deposited within a: (1) Gangdese retroarc foreland basin (England & Searle 1986; Yu & Wang 1990; DeCelles et al. 2007); (2) back-arc extensional basin (Zhang et al. 2004); (3) peripheral foreland basin (Leeder et al. 1988; Yin et al. 1994; Kapp et al. 2005b, 2007a; Leier et al. 2007c); or (4) composite foreland basin (Ding & Lai 2003) (Fig. 2). The Lhasa terrane was separated from the Tethys –Himalayan to the south by the IndusYarlung suture. And to the north it was bordered the Qiangtang terrane by the Bangong suture. Geologically the Lhasa terrane can be divided into the south Lhasa terrane and the north Lhasa terrane. The southern Lhasa terrane is predominantly characterized by Cretaceous –Cenozoic igneous rocks of the Gangdese arc (Burg & Chen 1984). And Jurassic –Cenozoic sedimentary strata are sporadically distributed among them. Volcanic rocks associated with the Gangdese arc are principally the Early Cretaceous Zenong Group (K1zl) in the north and the Early Tertiary Linzizong Group in the south. In addition, the Middle–Late Jurassic igneous rocks have also been reported within the Gangdese arc near the Indus –Yarlung suture (Chu et al. 2006; Dong et al. 2006; Yao et al. 2006; Zhu et al. 2008). In contrast, the northern Lhasa terrane generally crop out Upper Palaeozoic– Cretaceous sedimentary sequences (Leeder et al. 1988; Yin et al. 1988; Leier et al. 2007a). The exposed outcrops of pre-Jurassic strata are mainly Carboniferous siliciclastic sediments (Figs 1 & 3). Structurally the Selin Co basin is bound by the Shiquanhe-Gaize-Amdo Thrust system (SGAT) to the north (Kapp et al. 2005b) and the south-dipping north-verging Geren Co-Jiali Thrust system (GJT) to the south. The GJT trends NW –SE and extends from Geren Co to the west, where it is intruded by the Early Cretaceous granite and exposed along

29

with a mylonitic amphibolite near Zigui Co, to Nam Co and Jiali areas to the east. It juxtaposes predominantly the Carboniferous strata and the Zenong Group volcanic rocks in its hanging wall against the Lower Cretaceous strata of the Selin Co basin in its footwall. The fault was active by c. 130 Ma (Ding & Lai 2003). Following the deposition of the Langshan limestone, the GJT system is interpreted to have propagated northward and initiated thrusting along the Gaize-Selin Co Thrust (GST) by c. 97 Ma, which juxtaposed the limestone of the Langshan Formation in its hanging wall against the Upper Cretaceous non-marine Jingzhushan Formation in its footwall (DeCelles et al. 2007; Kapp et al. 2007a) (Figs 1 & 3).

Stratigraphy The Selin Co basin, roughly surrounded by Selin Co, Geren Co and Wuru Co, has an area of c. 8000 km2 and exposes nearly the complete Lower Cretaceous strata. Stratigraphic sections were measured and sampled near the Geren Co, Duba and Mayue area (Fig. 3). The Geren Co area made an ideal location of our investigation because a continuous section of the Lower Cretaceous rocks were exposed by the incision of snow water from the peak of surrounding mountains. Approximately 3300 m of the Lower Cretaceous rocks, including the Eshaerbu Formation (K1e), Duoni Formation (K1d), and Langshan Formation (K1l) cropped out in this area. Stratigraphically below the Lower Cretaceous rocks is the Upper Jurassic–Lower Cretaceous Rila Formation (J3-K1r), consisting of more than 1000 m of ophiolitic coarse sandstone and conglomerate at the base, and limestone at the top (Qu et al. 2003). The contact between the Eshaerbu Formation and Duoni Formation is conformable, whereas the contact between the Duoni Formation and Langshan Formation is a paraconformity (Figs 4 & 5a). In the Duba area, the Eshaerbu Formation was completely missing, and only the upper part of the Duoni Formation was preserved. However the Jingzhushan Formation (K2 j) was well exposed in the area (Fig. 5b). The outcrop in the Mayue area is poor, and the angular unconformity between the Duoni and Jingzhushan Formations can be recognized in the area.

The Eshaerbu Formation (K1e) Description. The Eshaerbu Formation is exposed only in the Geren Co area. The lower part of the Eshaerbu Formation is c. 450 m thick. It is characterized by laterally continuous, planar, parallel strata with interbedded sandstone and laminated mudstone (Fig. 6a). Individual sandstone bed is

30

Q.-H. ZHANG ET AL.

Fig. 2. Summary sketches of previously proposed tectonic models of Cretaceous basin in the Lhasa terrane. (a) Gangdese retroarc foreland basin (England & Searle 1986; Yu & Wang 1990; DeCelles et al. 2007). (b) Back-arc extensional basin (Zhang et al. 2004). (c) Peripheral foreland basin (Leeder et al. 1988; Yin et al. 1994; Kapp et al. 2005b, 2007a; Leier et al. 2007c). (d) Composite foreland basin (Ding & Lai 2003).

typically 10– 40 cm thick, and shows internal graded-bedding (Fig. 6b), parallel lamination and ripple cross-lamination. Flute casts in the sandstone beds indicate a NW-directed palaeocurrent (Fig. 6c). The middle Eshaerbu Formation consists of c. 500 m black mudstone apparently devoid of fossils and bioturbated textures. The Upper Eshaerbu Formation is c. 1000 m thick and consists of rhythmically uniform interbeddings of black,

laminated mudstone and grey siltstone (sometimes sandstone) (Fig. 6d). Individual siltstone beds coarsen upward and vary from 10 –50 cm in thickness. Small planar cross-beddings appear and palaeocurrent measurements trend towards the NE. Interpretation. The Eshaerbu Formation is interpreted to be deposited in a submarine fan and basin plain environment. The lower part represents

GANGDESE RETROARC FORELAND BASIN EVOLUTION

Fig. 3. Geological map of the Selin Co basin modified from the 1:250 000 regional geological surveys in Tibet (Nima area and Duba area, unpublished). 31

32

Q.-H. ZHANG ET AL.

Fig. 4. A schematic column of composite stratigraphic section in the Selin Co area, and see Figure 3 for explanations of the symbols. Note that the stratigraphic sections of the Carboniferous, Permian, and Rila Formation are not measured in detail, and their thicknesses and the contact relationships between them are not clear.

a typical flysch facies with Bouma sequence, and medium to fine, graded turbidites and the lateral continuity indicate an outer fan environment. Black mudstones in the middle are the products of settling from suspension in a basin plain. And the upper part is deposited in a fan lobe or the base of a middle fan environment, and the thickening and coarsening upwards sequence may result from the progradation of a fan lobe.

The Duoni Formation (K1d) Description. The Duoni Formation crops out in the Geren Co area, the Duba area and the Mayue area. However the complete Duoni Formation is exposed only in the Geren Co area with a thickness of c. 1200 m. The lower part of the Duoni Formation is not well exposed. Roughly, it consists of nonbedded to thick-bedded pebbly sandstone and

thin-bedded mudstone. Sandstones are poorly sorted, disorganized, and with heterogeneous clasts of volcanoclastic rocks and limestone (Fig. 6e). And the lateral continuity of sandstone beds is unclear. A layer of volcanoclastic rocks consisting of angular andesite, rhyolite and limestone appears in this part. The Upper Duoni Formation is c. 600 m thick, and consists of upward fining sequences of pebbly conglomerate, sandstone, siltstone and mudstone. A c. 100 m rhyolite layer and a thin coal layer are present at the base of this part. Conglomerates are relatively well organized with volcanic and limestone clasts. The conglomerate beds are lenticular and with basal scour surfaces. Unidirectional current ripples and small crossbeddings in the sandstone indicate a northward palaeocurrent direction. The laminated grey mudstones turn to red abruptly at the upper part (Fig. 6f). The uppermost Duoni Formation consists of the interbeddings of red mudstone and orbitolinid limestone-dominated breccia. In the Duba area, only c. 200 m thickness of the Duoni Formation is exposed and represents a analogue of the Upper Duoni Formation in the Geren Co area. It consists of upward fining sequences of grey-green conglomerate, coarse sandstone, sandstone, and laminated red mudstone. The conglomerate beds have basal scour surfaces, and the sandstone beds have trough and planar crossbeddings and unidirectional current ripples, indicating a SW palaeocurrent direction. Interpretation. The lower part of the Duoni Formation is deposited in a slope apron or the proximity of a submarine fan. The sedimentary environment of the Upper Duoni Formation is a river delta. The coal layer may indicate a marsh in the interdistributary area, and the upward fining sequences are the sediments of distributary channels in a delta plain. The red mudstones at the top reflect a dominantly oxidizing nature of the depositional and early diagenetic environment, and may indicate a very shallow marine environment. The volcanoclastic rocks and rhyolite layer in the Duoni Formation suggest the eruption of contemporary volcanoes in the surrounding area.

The Langshan Formation (K1l) Description. The Langshan Formation is well exposed but strongly eroded in the basin. It is .200 m thick, and consists of dark-grey rudist and orbitolinid-dominated mudstone, wackestone, and bioclastic packstone (Fig. 6g). Parts of the bioclastic packstone are composed of ooids, and large bivalves and gastropods. Orbitolinid foraminifera collected from the lowermost Langshan Formation consist mainly of Mesorbitolina texana and

GANGDESE RETROARC FORELAND BASIN EVOLUTION

Fig. 5. Logs of measured sections of the Lower Cretaceous rocks: (a) in the Geren Co area; and (b) in the Duba area. See Figure 3 for the location. 33

34

Q.-H. ZHANG ET AL.

Fig. 6. Photographs of the Selin Co basin: A– G are taken from the Geren Co area, H is from the Duba area. (a) Laterally continuous, parallel strata with interbedded sandstone and mudstone in the Lower Eshaerbu Formation. (b) Basal element of Bouma turbidite sequence in the Lower Eshaerbu Formation. (c) Flute casts in the Lower Eshaerbu Formation. (d) Rhythmically uniform interbeddings of black mudstone and siltstone in the Upper Eshaerbu Formation. (e) Poorly sorted sandstone with heterogeneous clasts of volcanoclastic rocks and limestone in the Lower Duoni Formation. (f) The sudden alteration of mudstone colour from grey to red, and the paraconformity between red beds of the Duoni Formation and grey orbitolinid limestone of the Langshan Formation. (g) Orbitolinid limestone in the Langshan Formation. (h) Mottled pebble-cobble conglomerate with dominant clasts of orbitolinid limestone and volcanic rocks in the Jingzhushan Formation.

GANGDESE RETROARC FORELAND BASIN EVOLUTION

Palorbitolinoides hedini, representing Aptian– Early Albian time. At the base of the Langshan Formation, red and green sandstone beds (1–2 m) are intercalated among the massive limestone. Interpretation. The depositional environment of the Langshan Formation is interpreted to be a shallow marine carbonate platform, and the sandstone interlayers indicate occasional input of terrigenous materials into the basin.

The Jingzhushan Formation (K2 j) Description. The Jingzhushan Formation is well exposed in the Duba area. It is .300 m thick, and consists of red medium-coarse sandstone and mottled pebble-cobble conglomerate with upwards coarsening sequence. Sandstone beds have large planar cross-beddings, indicating a northward palaeocurrent direction. The conglomerate is dominated with clasts of rudist, orbitolinid limestone (.70%) and small amount of sandstone and volcanic rocks (Fig. 6h). The conglomerate of the Jingzhushan Formation has been referred to as the ‘Upper conglomerate unit’ by DeCelles et al. (2007), and was derived mainly from the uplifted Langshan Formation limestone (DeCelles et al. 2007). Interpretation. Its depositional environment is interpreted to be a subaerial alluvial fan. The composition of the conglomerates indicates they are derived from the uplifted Langshan Formation, and the upward coarsening sequence suggests the progradation of an alluvial fan.

35

The Geren Co area In the Geren Co area, six samples from the Eshaerbu Formation consist of calcite-cemented sublitharenites and litharenites. The samples are characterized as subrounded to subangular, poorly-sorted monocrystalline quartz grains (c. 50–60%) with abundant fluid or mineral inclusions, and little chert (c. 8– 10%). Feldspar grains constituting ,4% of the modal compositions are predominantly plagioclase (c. 3%). Lithic grains are abundant and consist mostly of carbonate (c. 20–40%). Zircon and rutile are the common accessory minerals, and hornblende and chromite are also present. Nine samples from the Duoni Formation consist of litharenites and feldspathic litharenites. Compared with the Eshaerbu Formation, the content of volcanic lithic grains and feldspar grains (mainly plagioclase) increases. Most of the feldspar grains counted occurs within volcanic clasts. Accessory minerals are relatively rare and include zircon, pyrite and hematite.

The Duba area Four samples from the Duoni Formation near Duba consist of feldspathic litharenites. The quartz in the rocks is angular, poorly sorted, and with the content of 30 –50%. Volcanic lithic grains are c. 25%. Plagioclase is the main constituent of feldspar. The content of accessory minerals is relatively low (,1%). Samples of the Duoni Formation from Geren Co area and Duba area are rich in volcanic and sedimentary lithics, and fall into recycled orogen and magmatic arc in the Q– F–L plot (Fig. 7).

Petrographic analysis

The Mayue area

Compositions of sandstone grains were counted to analyse the potential provenance types. A standard practice for petrographic analysis was conducted by employing a modified Gazzi-Dickinson method, in which crystals larger than silt-sized within lithic grains are counted as monocrystalline grains (Dickinson & Suczek 1979; Ingersoll & Suczek 1979; Ingersoll et al. 1984). The modal framework grain compositions of 20 medium-grained sandstones were determined by point-counting standard thin sections (450 counts per slide). An important modification was counting carbonate grains as total sedimentary lithic grains (DeCelles et al. 2007). The six parameters counted from thin sections were Qm (monocrystalline quartz), Qp (polycrystalline quartz), P (plagioclase feldspar), K (potassium feldspar), Lv (total volcanic lithic grains), and Ls (total sedimentary lithic grains) respectively.

One sample of the Duoni Formation in the Mayue area is clasified as a sublitharenite. Subrounded to rounded, poorly-sorted monocrystalline quartz grains and sedimentary lithic grains dominate total modal compositions. Sedimentary lithic grains consist of carbonate and phyllites grains. The sandstone from the Mayue area lacks feldspar and volcanic lithic grains, which is different from the Duoni Formation in the Geren Co and Duba area, and falls into recycled orogen close to Q-apex in the Q –F –L plot.

Zircon U – Pb geochronology U –Pb geochronology was conducted mainly on detrital zircons to provide sediment provenance and maximum depositional ages of the Lower Cretaceous siliciclastic rocks. For more detailed

36

Q.-H. ZHANG ET AL.

Fig. 7. Ternary diagrams showing modal framework compositions of sandstone grains of the Lower Cretaceous strata in the Selin Co basin. The meanings of eight symbols in the ternary diagrams are as follows: Q, quartz; F, feldspar; L, total nonquartzose lithic grains; Qm, monocrystalline quartz; Qp, polycrystalline quartz; Lt, total lithic grains; Lv, total volcanic lithic grains; Ls, total sedimentary lithic grains.

information of experimental methods and procedures see supplementary material. Samples Z68 and Z48 were collected from the Duoni Formation in the Geren Co area, and 15 zircon grains were analysed for each sample. Sample Z68 is a volcanoclastic rock. Zircon crystals are euhedral and yielded Th–U ratios of 0.2–1.3. 10 usable zircon ages are scattered, among which 3 younger ages are 134, 135 and 144 Ma (Fig. 8a). Owing to the limited measurements of zircon grains, Z68 can not provide more information on provenance except the input of the Early Cretaceous igneous rocks. A rhyolite of sample Z48 yielded relatively large (50–200 mm), euhedral crystals with Th –U ratios of 0.6–1.2. The weighted average age of 13 concordant zircon ages is 125.8 + 2.8 Ma (2s), and is interpreted to mark the age of crystallization (Fig. 8b).

Nine sandstone samples were measured in the Geren Co area and the Duba area. By integrating U –Pb ages of detrital zircon grains from the same formation, three frequency plots were made to represent age distribution of zircon grains in the Eshaerbu Formation (sample Z06, Z83 and Z76) (Fig. 9a), the Duoni Formation in the Geren Co area (sample Z64, Z44, and Z28) (Fig. 9b), and the Duoni Formation in the Duba area (sample Z169, Z166, and Z174) (Fig. 9c). Sample T153 is the only one collected from the top of Duoni Formation near Mayue (Fig. 9d). 50 zircon grains were analysed for each sandstone sample. Most detrital zircon grains are 30–100 mm in diameter, euhedral to anhedral, and yielded Th/U ratios between 0.2– 1.4, which suggest the grains were mainly derived from igneous rocks and recycled sandstones (Wu & Zheng 2004).

GANGDESE RETROARC FORELAND BASIN EVOLUTION

37

Fig. 8. Concordia diagrams of zircon U– Pb ages from sample (a) Z68, and (b) Z48, and weighed mean ages of the youngest zircon populations from sample (c) Z06 in the Geren Co area. See Figure 5 for sample position in the measured section.

The mean age of the youngest cluster of zircon grains from sample Z06 is c. 130 Ma, which is interpreted to represent the maximum depositional age of the Eshaerbu Formation (Fig. 8c). The Eshaerbu Formation shows two largest populations with zircon ages of 125– 140 and 160 –180 Ma. The

cluster of the age between 220–290 Ma is also relatively prominent (Fig. 9a). However the dominant cluster in the Duoni Formation lies only in 125– 130 Ma in the Geren Co area (Fig. 9b). In the Duba area, there is also only one dominant cluster with the age of 125–150 Ma. But compared with

Fig. 9. Frequency plots of detrital U–Pb zircon ages from: (a) the Eshaerbu Formation in the Geren Co area, (b) the Duoni Formation in the Geren Co area, (c) the Duoni Formation in the Duba area, and (d) the Duoni Formation in the Mayue area. See Figure 5 for sample position in the measured section.

38

Q.-H. ZHANG ET AL.

its counterpart in the Geren Co area, the main cluster in the Duba area has a relatively scattered distribution (Fig. 9c). Sample T153 is considerably different from the other samples in this study due to the lack of the late Jurassic –Early Cretaceous ages except for a single 134 Ma age. The younger populations of detrital zircon ages are lying among 190 and 220 – 280 Ma (Fig. 9d).

Discussion Sediment provenance in the Selin Co basin Owing to the evidently different sandstone compositions and detrital zircon U –Pb ages between sample T153 in the Mayue area and the rest sandstones in the Geren Co and Duba areas (Figs 7 & 9), we will discuss them separately. When all nine of the sandstone samples from the Lower Cretaceous strata in the Geren Co and Duba area are combined into a single group, they yield a probability distribution of zircon ages that may be considered representative of the Lower Cretaceous rocks from the northern Lhasa terrane. And the most significant population of ages is between 125 –140 Ma, and additional significant populations are at 160 –180, 220– 290, 500 –700, 900 –1000, 1100–1200, 1900–2000 and 2400–2500 Ma (Fig. 10). The Gangdese magmatic arc is attributed to the northward subduction of the Neo-Tethyan ocean lithosphere beneath the Lhasa terrane (Alle`gre et al. 1984; Chang et al. 1986), and was active at least during the Late Cretaceous –Early Tertiary (Scha¨rer et al. 1984; Xu et al. 1985; Coulon et al. 1986; Debon et al. 1986; Harris et al. 1988; Yin et al. 1994; Yin & Harrison 2000). Several observations of Jurassic –Early Cretaceous igneous rocks in the Gangdese arc have been reported in

Fig. 10. Relative probability plot of detrital zircon U–Pb ages from all nine sandstone samples of the Lower Cretaceous strata (except sample T153 in the Mayue area) in the Selin Co basin.

the recent years (Ding & Lai 2003; Chu et al. 2006; Dong et al. 2006; Yao et al. 2006; Zhu et al. 2006, 2008) (Fig. 11a), which are also thought to be related to the Neo-Tethyan ocean lithosphere subduction (Chu et al. 2006; Dong et al. 2006). In addition, detrital zircon ages of the Ngamring Formation within the Xigaze forearc basin, which is mainly derived from the Gangdese magmatic arc (Du¨rr 1996), include widely distributed Jurassic–Early Cretaceous ages (Ji et al. 2009). Therefore we conclude that the Gangdese arc has a longer history than usually thought, and 125– 140 Ma zircon grains in the Selin Co basin are mainly derived from the Gangdese magmatic arc. Igneous rocks ranging from 160–180 Ma have been documented in the Gangdese arc, the BNS and the Qiangtang terrane (Fig. 11a) (Kapp et al. 2003b, 2005b; Schwab et al. 2004; Guynn et al. 2006). Moreover, detrital zircons of the Mesozoic strata in the Qiangtang terrane have yielded significant age populations between 160–180 Ma (Dinglin, unpublished data) and 200–300 Ma (G. Gehrels and others, unpublished data) (Fig. 11b). In addition, areas to the north of Qiangtang terrane, such as Songpan-Ganzi flysch complex and Kunlun terrane also expose 200–300 Ma igneous rocks (Fig. 11a) (Harris et al. 1988; Cowgill et al. 2003; Roger et al. 2003; Gehrels et al. 2003; Kapp et al. 2003b; Schwab et al. 2004). Hence, the 160– 180 Ma zircon grains in the Selin Co basin were likely derived from the Gangdese magmatic arc to the south and/or the BNS and the Qiangtang terrane to the north. However, 220 –290 Ma zircon grains were mainly derived from the Mesozoic rocks of the Qiangtang terrane and/or igneous rocks of the Songpan-Ganzi and Kunlun terrane to the north. Zircon grains exceeding ages of 500 Ma in the Selin Co basin are more challenging to interpret in terms of provenance, because the Mesozoic and Palaeozoic rocks in the Lhasa terrane and the Palaeozoic strata in the Qiangtang terrane could all be potential provenance areas (Fig. 11b). However, the curve of the relative probability for zircon ages .500 Ma in the Selin Co basin is similar to that of Carboniferous sandstone near Nam Co, with dominant populations at 500–600, 800–1000, 1100–1200, 1800– 1900 and 2400– 2500 Ma (Fig. 11c). In addition, large-scale outcrops of the Carboniferous sandstone are exposed to the south of the GJT. So, it is reasonable to suggest the Carboniferous sandstone as the source rock for the majority of the .500 Ma zircons to the Selin Co basin. Although detrital zircon ages in the Mayue area is relatively scarce, the dominant age cluster of 200–300 Ma indicates a lack of the Early Cretaceous zircon grains. According to a comprehensive

GANGDESE RETROARC FORELAND BASIN EVOLUTION

39

Fig. 11. U –Pb zircon ages of the Lower Cretaceous rocks in the Selin Co basin and (a) igneous rocks in the Tibetan Plateau, (b) the Palaeozoic and Mesozoic strata in the Lhasa and Qiangtang terrane, and (c) Carboniferous sandstone near Nam Co. Note that zircon ages presented in (a) is only between 100– 300 Ma. Corresponding data of igneous rocks are taken from the literatures (Harris et al. 1988; Cowgill et al. 2003; Kapp et al. 2003a, b, 2005 a, b; Gehrels et al. 2003; Roger et al. 2003; Schwab et al. 2004; Chu et al. 2006; Dong et al. 2006; Guynn et al. 2006; Volkmer et al. 2007). In (b), U–Pb data of detrital zircons from the Palaeozoic and Mesozoic rocks in the Lhasa and Qiangtang terranes are provided by G. Gehrels and other (unpublished data). The comparison in (c) is mainly focused on the ages older than 500 Ma. U– Pb data of Carboniferous sandstone near Nam Co is from Leier et al. (2007b).

summary of zircon U –Pb ages in the Tibet (Fig. 11), we conclude the Lower Cretaceous rocks in Mayue area are most likely derived from the Qiangtang terrane and/or the Songpan-Ganzi and the Kunlun to the north.

Regional palaeogeography of the Cretaceous Selin Co basin In the Selin Co basin, the maximum depositional age of the Eshaerbu Formation is c. 130 Ma while 125 Ma from a rhyolite at the top of the Duoni Formation reveals the terminational age of the Lower siliciclastic sediments. The determination of orbitolinid foraminifera of the Langshan limestone provides the age of Aptian –Early Albian. And the conglomerates of the Jingzhushan Formation were supposed to be deposited during 105 –90 Ma (DeCelles et al. 2007). Therefore the regional palaeogeography of the Selin Co basin will be reconstructed in three stages: 130– 125 Ma (Barremian); 125 –105 Ma (Aptian –Early Albian); and 105– 90 Ma (Late Albian –Turonian) (Fig. 12). During the Early Barremian, the southern Selin Co basin was occupied by submarine fan and deep marine, and the palaeogeography of the northern Selin Co basin is unknown because of the poorly exposed Eshaerbu Formation equivalents. At the Late Barremian, the depositional environment was increasingly shallowing. The southern Selin Co

basin was occupied by a deltaic and fluvial environment with volcanic eruption at the surrounding area. Terrigenous material shed into the Selin Co basin may have been derived from the north, south and east of the Selin Co basin (Fig. 12a). During the Aptian– Early Albian, marine transgression, possibly resulted from a rise in global sea level (Haq et al. 1987) and/or tectonic quiescence (Flemings & Jordan 1990), enlarged the northern boundary of the Selin Co basin to the Gaize-Selin Co thrust (Kapp et al. 2007a). Massive orbitolinid limestone dominated the northern Lhasa terrane, and carbonate platform occurred across the whole basin. Although the influx of terrigenous materials during the period was extremely limited, the appearance of sandstone interlayer among the limestone indicated occasionally input of terrestrial materials (Fig. 12b). Based on facies distribution and perpendicular relationships between palaeoshoreline and palaeocurrent directions, the Selin Co basin during the Barremian –Early Albian was likely characterized by an east –west-elongating gulf with shoreline on the south, east and north. The Qiangtang terrane, the Naqu area and the Gangdese arc are all lack of significant accumulations of the Lower Cretaceous strata (Pan et al. 2004), which may indicate that these areas were subaerial highlands at this time. During the Late Albian –Turonian, seawater was completely retreated from the Selin Co basin and

40

Q.-H. ZHANG ET AL.

Fig. 12. Palaeogeographical sketches of the Selin Co basin in c. 130–125 Ma (a), c. 125– 105 Ma (b), and c. 105– 90 Ma (c).

alluvial fans dominated the northern Selin Co basin. The Langshan Formation in the southern Selin Co basin was uplifted by the thrusting of the GaizeSelin Co Thrust (GST) and alluvial fans sourced from the Langshan Formation lead to the deposition of the Jingzhushan Formation.

Early Cretaceous tectonic setting and basin evolution in the Selin Co area Some authors have suggested that the Lhasa terrane subducted northward beneath the Qiangtang terrane, triggering the development of an early Cretaceous peripheral foreland basin in the Lhasa terrane (Leeder et al. 1988; Kapp et al. 2005b, 2007a; Leier et al. 2007c), or a back-arc extensional basin

was developed during the Early Cretaceous (Zhang et al. 2004). However, the investigation of the Lower Cretaceous rocks does not argue for, at least during 130– 125 Ma, the presence of an extensional environment in the Selin Co basin. Firstly, petrographic analysis of sandstones shows that the sediments in the basin are mainly recycled, which usually points to a compressional environment such as an orogenic source area. Secondly, the geochemical analyses of mudstone (major elements, trace elements and Nd isotope) show that the Selin Co basin are mainly derived from an Andean-type active continental margin and its geochemical compositions are similar with the average upper crust (Zhang 2008), which also argues against a rifting setting with the input of mantle-like materials (Zhang et al. 2007). Thirdly, the research of the structure of the area, characterized by some thrusts such as the GJT, the GST, and the SGAT, do not suggest an extensional environment. Moreover, the overall upwardcoarsening sequence may represent vertical juxtaposition of different depozones resulting from the lateral migration of a foreland basin (DeCelles & Giles 1996; Sinclair 1997a, b; DeCelles & Horton 2003), and hence may represent a foreland basin deposits. Palaeocurrent direction and detrital zircon geochronology analysis indicate that the Lower Cretaceous sediments in the Selin Co basin were mainly derived from the Gangdese magmatic arc and the Carboniferous outcrops to the south. Furthermore, the Geren Co area, close to the Geren Co-Jiali Thrust, has the thickest accumulation of the Lower Cretaceous strata (Yu & Wang 1990), which contradicts with the observation of northward increasing thickness of the Lower Cretaceous strata in the Lhasa terrane (Leeder et al. 1988; Zhang et al. 2004; Kapp et al. 2005b; Leier et al. 2007c). So our findings in the Selin Co basin are inconsistent with the development of a typical peripheral foreland basin. Instead, we propose that a retroarc foreland basin model provides a better explanation for the dominant source of the Selin Co basin being from rocks located to the south. The deposition of the Rila limestone and the Eshaerbu turbidities in the Geren Co area is consistent with an interpretation of the transition from the forebulge to the foredeep position within a foreland basin system (Blair & Bilodeau 1988; Heller et al. 1988; Jordan 1995; DeCelles & Giles 1996). The maximum depositional age of c. 130 Ma provided by sample Z06 is not only recorded in the Selin Co basin, but also in the Xiagangjiang Range area near the Coqin county (Volkmer et al. 2007). Coincidently, the muscovite 40Ar/39Ar age of granite in the hanging wall of GJT is also 130 Ma (Ding & Lai 2003), which is taken as the thrusting

GANGDESE RETROARC FORELAND BASIN EVOLUTION

age of GJT. Therefore, the initial development of a retroarc foreland basin in the northern Lhasa terrane could be occurred at c. 130 Ma.

A retroarc foreland basin during the Cretaceous in the Lhasa terrane Changes to the understanding of the Cretaceous stratigraphy, detrital zircon provenance and palaeogeography of the Selin Co basin presented here has motivated us to present a new model for the evolution of a retroarc foreland basin, which developed above the basement of the Lhasa terrane from 145 – 90 Ma. During c. 145 –130 Ma, owing to the northward subduction of the Neo-Tethyan ocean lithosphere, a Gangdese retroarc thrust belt and foreland basin were speculated to develop at the south of Nam Co and deposited the Chumulong Formation in the Linzhou basin. The detrital zircon grains from the Chumulong Formation have dominant ages of 140 –160 Ma (Leier et al. 2007c), which are likely derived from the Gangdese magmatic arc (Fig. 13a).

41

During c. 130–125 Ma, thrusting and shortening in the retroarc area migrated northward and the north-vergent thrusting of the GJT came into being. The Selin Co basin was accumulated by the Eshaerbu Formation and the Duoni Formation. The continued subduction of the Neo-Tethyan ocean lithosphere generated magmatism in the north of the Gangdese arc. The Zenong volcanic rocks associated with this magmatism were transported northwards, providing the Early Cretaceous zircon grains in the Selin Co basin (Fig. 13b). During c. 125–105 Ma, the fluxes of terrigenous materials to the basin were reduced by tectonic quiescence and/or climatic changes. The orbitolinidbearing limestone of the Langshan Formation was developed in the north Lhasa terrane. Minor limestone of the Penbo Member can also deposited on the top of the Chumulong Formation to the south. Flat subduction of the Neo-Tethyan ocean lithosphere during this time was continued, generating magmatism in the Nima basin (Kapp et al. 2007a) (Fig. 13c). During c. 105–90 Ma, reactivation of the Gangdese retroarc thrust belt produced a new foredeep zone in the Linzhou basin and deposited the Takena Formation atop the Chumulong conglomerate (Leier et al. 2007a; Kapp et al. 2007b). In addition, shortening along the GST to the north generated the deposition of continental red beds and conglomerates of the Jingzhushan Formation in the northern Selin Co basin and South Nima basin, which may represent wedge-top deposits at that time (Fig. 13d).

Conclusion

Fig. 13. Sketches showing the proposed tectonic history of a retroarc foreland basin in the Lhasa terrane from 145–90 Ma. See Figures 1 and 3 for explanations of the symbols. Not drawn to scale.

Our studies in the Selin Co basin provide a new perspective for the Cretaceous basin evolution and initial thickening of the Lhasa terrane. The upward coarsening sequence of the Eshaerbu Formation and the Duoni Formation were deposited in the submarine fan and delta environments, and represent an underfilled flysch stage of a foreland basin. And the orbitolinid limestones of the Langshan Formation were the products of carbonate platform during the period of tectonic quiescence. Finally the conglomerates of the Jingzhushan Formation indicate an alluvial fan environment and represent an overfilled molasse stage of a foreland basin. Petrographic analysis shows that the Lower Cretaceous sediments are mainly derived from a recycled orogen and magmatic arc. Zircons grains from the Lower Cretaceous rocks have a significant cluster of ages at 125–140 Ma, likely derived from Gangdese magmatic arc. Ages of 220–290 Ma were likely derived from the Qiangtang terrane and/or further north. Both the Gangdese magmatic arc and the Qiangtang terrane (including BNS) could

42

Q.-H. ZHANG ET AL.

have supplied 160 –180 Ma zircons to the Selin Co basin. Zircons with ages .500 Ma were likely derived from the Carboniferous rocks to the south of the Geren Co-Jiali Thrust. Generally, the Gangdese magmatic arc and Carboniferous strata to the south was the primary sediment source for the Selin Co basin. A retroarc foreland basin was supposed to develop in the Lhasa terrane during c. 145 –90 Ma, and the northward subduction of the Neo-Tethyan ocean lithosphere beneath the Lhasa terrane provided the dominant tectonic control on the basin evolution. The Early Cretaceous retroarc folds and thrust belts accompanying the retroarc foreland basin likely caused shortening and possibly an uplift of the Lhasa terrane earlier than previously thought. Our work will have profound implications for the evolution of central Tibet, contrasting with previous models presented for the Lhasa terrane. We thank Chun-Rong Diwu, Hong Zhang and Xiao-Ming Liu to help with zircon U– Pb isotopic analysis at the Key Laboratory of Continental Dynamics of NW University in Xi’an. We also thank George E. Gehrels to share his detrital zircon U– Pb age data, and we benefit a lot from the discussion with Alex Pullen and Paul Kapp. Seb Turner and another anonymous reviewer are thanked for their constructive suggestions. This project was supported by grants from National Science Fund for Outstanding Youths of China (40625008 to Ding), Chinese Academy of Sciences (KZCX2-YW-Q09-03 to Ding) and Chinese Ministry of Science and Technology (2009CB421000 to Ding).

References Aitchison, J. C., Davis, A. M. & Luo, H. 2002. New constraints on the India–Asia collision: the lower Miocene Gangrinboche conglomerates, Yarlung Tsangpo suture zone, SE Tibet. Journal of Asian Earth Sciences, 21, 251– 263. Alle`gre, C. J., Courtillot, V. et al. 1984. Structure and evolution of the Himalaya-Tibet orogenic belt. Nature, 307, 17–22. Blair, T. C. & Bilodeau, W. L. 1988. Development of tectonic cyclothems in rift, pull-apart, and foreland basins; sedimentary response to episodic tectonism. Geology, 16, 517– 520. Burg, J. P. & Chen, G. M. 1984. Tectonics and structural zonation of Southern Tibet, China. Nature, 311, 219– 223. Chang, C., Chen, N. et al. 1986. Preliminary conclusions of the Royal Society and Academia Sinica 1985 Geotraverse of Tibet. Nature, 323, 501–507. Chu, M. F., Chung, S. L. et al. 2006. Zircon U– Pb and Hf isotope constraints on the Mesozoic tectonics and crustal evolution of Southern Tibet. Geology, 34, 745– 748. Coulon, C., Maluski, H., Bollinger, C. & Wang, S. 1986. Mesozoic and Cenozoic volcanic rocks from Central and Southern Tibet: 39Ar– 40Ar dating,

petrological characteristics and geodynamical significance. Earth and Planetary Science Letters, 79, 281–302. Cowgill, E., Yin, A., Harrison, T. M. & Wang, X. F. 2003. Reconstruction of the Altyn Tagh fault based on U–Pb geochronology: role of back thrusts, mantle sutures, and heterogeneous crustal strength in forming the Tibetan Plateau. Journal of Geophysical Research, 108; doi: 10.1029/2002JB002080. Debon, F., Fort, P. L. E., Sheppard, S. M. F. & Sonet, J. 1986. The four plutonic belts of the TranshimalayaHimalaya: a chemical, mineralogical, isotopic, and chronological synthesis along a Tibet– Nepal section. Journal of Petrology, 27, 219–250. Decelles, P. G. & Giles, K. A. 1996. Foreland basin systems. Basin Research, 8, 105–123. Decelles, P. G. & Horton, B. K. 2003. Early to Middle Tertiary foreland basin development and the history of Andean crustal shortening in Bolivia. Bulletin of the Geological Society of America, 115, 58–77. Decelles, P. G., Kapp, P., Ding, L. & Gehrels, G. E. 2007. Late Cretaceous to Middle Tertiary basin evolution in the Central Tibetan Plateau: changing environments in response to tectonic partitioning, aridification, and regional elevation gain. Bulletin of the Geological Society of America, 119, 654–680. Dewey, J. F., Shackleton, R. M., Chengfa, C. & Yiyin, S. 1988. The tectonic evolution of the Tibetan Plateau. Philosophical Transactions of the Royal Society of London. Series A, Mathematical and Physical Sciences (1934– 1990), 327, 379– 413. Dickinson, W. R. & Suczek, C. A. 1979. Plate tectonics and sandstone compositions. American Association of Petroleum Geologists Bulletin, 63, 2164– 2182. Ding, L. & Lai, Q. 2003. New geological evidence of crustal thickening in the Gangdese Block prior to the India–Asian collision. Chinese Science Bulletin, 48, 1604– 1610 [in Chinese with English abstract]. Ding, L., Kapp, P. & Wan, X. 2005. Paleocene-Eocene record of ophiolite obduction and initial India–Asia collision, South Central Tibet. Tectonics, 24; doi: 10.1029/2004TC001729. Dong, Y., Xu, J., Zeng, Q., Wang, Q., Mao, G. & Li, J. 2006. Is there a Neo-Tethys’ subduction record earlier than arc volcanic rocks in the Sangri Group? Acta petrological Sinica, 22, 661–668 [in Chinese with English abstract]. Du¨rr, S. B. 1996. Provenance of Xigaze fore-arc basin clastic rocks (Cretaceous, South Tibet). Bulletin of the Geological Society of America, 108, 669–684. England, P. C. & Searle, M. 1986. The CretaceousTertiary deformation of the Lhasa block and its implications for crustal thickening in Tibet. Tectonics, 5, 1– 14. Flemings, P. B. & Jordan, T. E. 1990. Stratigraphic modeling of foreland basins: interpreting thrust deformation and lithosphere rheology. Geology, 18, 430–434. Gehrels, G. E., Yin, A. & Wang, X. F. 2003. Magmatic history of the northeastern Tibetan Plateau. Journal of Geophysical Research, 108; doi: 10.1029/ 2002JB001876. Guynn, J. H., Kapp, P., Pullen, A., Heizler, M., Gehrels, G. & Ding, L. 2006. Tibetan basement

GANGDESE RETROARC FORELAND BASIN EVOLUTION rocks near Amdo Reveal “missing” Mesozoic tectonism along the Bangong Suture, Central Tibet. Geology, 34, 505– 508. Haq, B. U., Hardenbol, J. A. N. & Vail, P. R. 1987. Chronology of fluctuating sea levels since the Triassic. Science, 235, 1156–1167. Harris, N. B. W., Ronghua, X., Lewis, C. L., Hawkesworth, C. J. & Yuquan, Z. 1988. Isotope geochemistry of the 1985 Tibet Geotraverse, Lhasa to Golmud. Philosophical Transactions of the Royal Society of London. Series A, Mathematical and Physical Sciences (1934–1990), 327, 263–285. Heller, P. L., Angevine, C. L., Winslow, N. S. & Paola, C. 1988. Two-phase stratigraphic model of foreland-basin sequences. Geology, 16, 501–504. Ingersoll, R. V. & Suczek, C. A. 1979. Petrology and provenance of Neogene sand from Nicobar and Bengal Fans, DSDP Sites 211 and 218. Journal of Sedimentary Research, 49, 1217–1228. Ingersoll, R. V., Fullard, T. F., Ford, R. L., Grimm, J. P., Pickle, J. D. & Sares, S. W. 1984. The effect of grain size on detrital modes; a test of the GazziDickinson point-counting method. Journal of Sedimentary Research, 54, 103–116. Ji, W. Q., Wu, F. Y., Chung, S. L., Li, J. X. & Liu, C. Z. 2009. Zircon U–Pb geochronology and Hf isotopic constraints on petrogenesis of the Gangdese Batholith, Southern Tibet. Chemical Geology, 262, 229– 245. Jordan, T. E. 1995. Retroarc foreland and related basins. In: Busby, C. J. & Ingersoll, R. V. (eds) Tectonics of Sedimentary Basins. Blackwell Science Publications, Oxford, 331– 362. Kapp, P., Murphy, M. A., Yin, A., Harrison, T. M., Ding, L. & Guo, J. 2003a. Mesozoic and Cenozoic Tectonic evolution of the Shiquanhe area of Western Tibet. Tectonics, 22; doi: 10.1029/2001TC001332. Kapp, P., Yin, A., Manning, C. E., Harrison, T. M., Taylor, M. H. & Ding, L. 2003b. Tectonic evolution of the early Mesozoic blueschist-bearing Qiangtang metamorphic belt, Central Tibet. Tectonics, 22; doi: 10.1029/2002TC001383. Kapp, J. L. D., Harrison, T. M., Kapp, P., Grove, M., Lovera, O. M. & Ding, L 2005a. Nyainqentanglha Shan: a window into the tectonic, thermal, and geochemical evolution of the Lhasa Block, Southern Tibet. Journal of Geophysical Research, 110; doi: 10.1029/2004JB003330. Kapp, P., Yin, A., Harrison, T. M. & Ding, L. 2005b. Cretaceous-Tertiary shortening, basin development, and volcanism in Central Tibet. Bulletin of the Geological Society of America, 117, 865 –878. Kapp, P., Decelles, P. G., Gehrels, G. E., Heizler, M. & Ding, L. 2007a. Geological records of the LhasaQiangtang and Indo-Asian collisions in the Nima area of Central Tibet. Bulletin of the Geological Society of America, 119, 917– 932. Kapp, P., Decelles, P. G. et al. 2007. The Gangdese retroarc thrust belt revealed. GSA Today, 17, 4 –9. Leeder, M. R., Smith, A. B. & Jixiang, Y. 1988. Sedimentology, palaeoecology and palaeoenvironmental evolution of the 1985 Lhasa to Golmud Geotraverse. Philosophical Transactions of the Royal Society of London. Series A, Mathematical and Physical Sciences (1934–1990), 327, 107–143.

43

Leier, A. L., Decelles, P. G., Kapp, P. & Ding, L. 2007a. The Takena Formation of the Lhasa Terrane, Southern Tibet: the record of a Late Cretaceous retroarc foreland basin. Bulletin of the Geological Society of America, 119, 31–48. Leier, A. L., Kapp, P., Gehrels, G. E. & Decelles, P. G. 2007b. Detrital zircon geochronology of Carboniferous-Cretaceous strata in the Lhasa Terrane, Southern Tibet. Basin Research, 19, 361–378. Leier, A. L., Decelles, P. G., Kapp, P. & Gehrels, G. E. 2007c. Lower Cretaceous strata in the Lhasa Terrane, Tibet, with implications for understanding the early tectonic history of the Tibetan Plateau. Journal of Sedimentary Research, 77, 809–825. Molnar, P., England, P. & Martinod, J. 1993. Mantle dynamics, uplift of the Tibetan Plateau, and the Indian monsoon. Reviews of Geophysics, 31, 357–396. Murphy, M. A., Yin, A. et al. 1997. Did the Indo-Asian collision alone create the Tibetan Plateau? Geology, 25, 719 –722. Pan, G. T., Ding, J., Yao, D. S. & Wang, L. Q. 2004. Geologic map of the Qinghai-Xizang Plateau and its neighboring regions, scale: 1:1500000. Chengdu Institute of Geology and Mineral Resources [in Chinese]. Qu, Y., Zhang, S. et al. 2003. The Late Jurassic– Early Cretaceous Rila formation, Rila formation Suor clastic rocks and characteristics of Biotas in the Yunzhug ophiolite belt, Northern Tibet. Geological Bulletin of China, 22, 959– 963 [in Chinese with English abstract]. Roger, F., Arnaud, N. et al. 2003. Geochronological and geochemical constraints on Mesozoic suturing in East Central Tibet. Tectonics, 22; doi: 10.1029/ 2002TC001466. Scha¨rer, U., Xu, R.-h. & Alle`gre, C. J. 1984. U –Pb geochronology of Gangdese (Transhimalaya) plutonism in the Lhasa-Xigaze region, Tibet. Earth and Planetary Science Letters, 69, 311– 320. Schwab, M., Ratschbacher, L. et al. 2004. Assembly of the Pamirs: age and origin of magmatic belts from the Southern Tien Shan to the Southern Pamirs and their relation to Tibet. Tectonics, 23; doi: 10.1029/ 2003TC001583. Searle, M. P., Windley, B. F. et al. 1987. The closing of Tethys and the tectonics of the Himalaya. Bulletin of the Geological Society of America, 98, 678–701. Sinclair, H. D. 1997a. Flysch to molasse transition in peripheral foreland basins: the role of the passive margin v. slab breakoff. Geology, 25, 1123– 1126. Sinclair, H. D. 1997b. Tectonostratigraphic model for underfilled peripheral foreland basins: an alpine perspective. Bulletin of the Geological Society of America, 109, 324– 346. Tapponnier, P., Mercier, J. L. et al. 1981. The Tibetan side of the India– Eurasia collision. Nature, 294, 405– 410. Volkmer, J. E., Kapp, P., Guynn, J. H. & Lai, Q. 2007. Cretaceous– Tertiary structural evolution of the North Central Lhasa Terrane, Tibet. Tectonics, 26; doi: 10.1029/2005TC001832. Wu, Y. & Zheng, Y. 2004. Genesis of zircon and its constraints on interpretation of U–Pb age. Chinese Science Bulletin, 49, 1554–1569.

44

Q.-H. ZHANG ET AL.

Xiao, X. & Li, T. 2000. Tectonic Evolution and Uplift of Qinghai-Xizang (Tibet) Plateau. Guandong Science & Technology Press, Guangzhou [in Chinese with English abstract]. Xu, R. H., Scharer, U. & Allegre, C. J. 1985. Magmatism and metamorphism in the Lhasa Block (Tibet): a geochronological study. Journal of Geology, 93, 41–57. Yao, P., Li, J., Wang, Q., Gu, X., Tang, J. & Hui, L. 2006. Discovery and geological significance of the Adakite in Gangdise Island arc belt, Xizang (Tibet). Acta Petrologica Sinica, 22, 612–620 [in Chinese with English abstract]. Yin, A., Harrison, T. M., Ryerson, F. J., Chen, W. J., Kidd, W. S. F. & Copeland, P. 1994. Tertiary structural evolution of the Gangdese thrust system, Southeastern Tibet. Journal of Geophysical Research, 99, 18175– 118201. Yin, A. & Harrison, T. M. 2000. Geologic evolution of the Himalayan –Tibetan orogen. Annual Reviews in Earth and Planetary Sciences, 28, 211– 280. Yin, J., Xu, J., Liu, C. & Li, H. 1988. The Tibetan Plateau: regional stratigraphic context and previous work. Philosophical Transactions of the Royal Society of London. Series A, Mathematical and Physical Sciences (1934–1990), 327, 5– 52. Yu, G. & Wang, C. 1990. Sedimentary Geology of the Xizang (Tibet) Tethys. Geological Publications House, Beijing (in Chinese with English abstract).

Zhang, K. J. 2000. Cretaceous palaeogeography of Tibet and adjacent areas (China): tectonic implications. Cretaceous Research, 21, 23– 33. Zhang, K. J., Xia, B. D., Wang, G. M., Li, Y. T. & Ye, H. F. 2004. Early Cretaceous stratigraphy, depositional environments, sandstone provenance, and tectonic setting of Central Tibet, Western China. Bulletin of the Geological Society of America, 116, 1202– 1222. Zhang, K. J., Zhang, Y. X., Li, B. & Zhong, L. F. 2007. Nd isotopes of siliciclastic rocks from Tibet, Western China: constraints on provenance and pre-Cenozoic tectonic evolution. Earth and Planetary Science Letters, 256, 604– 616. Zhang, Q.-H. 2008. Provenance analysis and Early Cretaceous retroarc foreland basin evolution in the Selin Co area, Central Tibet. PhD thesis, Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing. Zhu, D. C., Pan, G. T. et al. 2006. Late Jurassic– Early Cretaceous geodynamic setting in Middle-Northern Gangdese: new insights from volcanic rocks. Acta Petrologica Sinica, 22, 534 –546 [in Chinese with English abstract]. Zhu, D. C., Pan, G. T., Chung, S. L., Liao, Z. L., Wang, L. Q. & Li, G. M. 2008. Shrimp zircon age and geochemical constraints on the origin of Lower Jurassic volcanic rocks from the Yeba formation, Southern Gangdese, South Tibet. International Geology Review, 50, 442–471.

Metamorphic evolution of the Tethyan Himalayan flysch in SE Tibet I. DUNKL1*, B. ANTOLI´N2, K. WEMMER1, G. RANTITSCH3, M. KIENAST1, C. MONTOMOLI4, L. DING5, R. CAROSI4, E. APPEL2, R. EL BAY2, Q. XU5 & H. VON EYNATTEN1 1

Geoscience Center, University of Go¨ttingen, Goldschmidtstrasse 3, D-37077 Go¨ttingen, Germany

2

Institute for Geosciences, University of Tu¨bingen, Sigwartstrasse 10, D-72076 Tu¨bingen, Germany

3

University of Leoben, Peter-Tunner-Straße 5, A-8700 Leoben, Austria

4

Department of Earth Sciences, University of Pisa, via S. Maria 53, I-56126 Pisa, Italy 5

Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Shuangqing Road 18, Beijing 100085, China *Corresponding author (e-mail: [email protected]) Abstract: The metamorphic conditions and the age of thermal overprint were determined in metapelites, metaarenites and metabasites of the Tethyan Himalayan Sequence (THS) in SE Tibet using Ku¨bler Index and vitrinite reflectance data and applying thermobarometrical (Thermocalc and PERPLEX) and geochronological methods (illite/muscovite K– Ar and zircon and apatite (U – Th)/He chronology). The multiple folded thrust pile experienced a thermal overprint reaching locally peak conditions between the diagenetic stage (c. 170 8C) and the amphibolite facies (c. 600 8C at 10 kbar). Burial diagenesis and heating due to Early Cretaceous dyke emplacement triggered the growth of illite in the metapelites. Eocene collision-related peak metamorphic conditions have been reached at c. 44 Ma. During collision the different tectonic blocks of the THS were tectonically buried to different structural levels so that they experienced maximum greenschist to amphibolite facies metamorphism. Later, during Oligocene to Miocene times the entire THS underwent anchi- to epizonal metamorphic conditions, probably associated to continuous deformation in the flysch fold-thrust-system. This period terminated at c. 24– 22 Ma. Adjacent to the north Himalayan metamorphic domes, the base of the THS was metamorphosed during Miocene times (c. 13 Ma). Post-metamorphic cooling below c. 180 8C lasted until Late Miocene and took place at different times.

The northward drift of Greater India during the Cenozoic resulted in the closure of the Tethys ocean, the initiation of the India –Asia collision in the Paleocene and the subsequent uplift of the Himalayan Range (c. 55 –50 Ma; e.g. Gaetani & Garzanti 1991; Patzelt et al. 1996; Najman et al. 2005). The Himalayan arc forms an active WNW– ESE asymmetric fold and thrust belt with a main southward vergence (Fig. 1a, b). The northern member of the Himalaya is the Tethyan Himalayan Sequence (THS) which is located in the highest structural position within the orogen (Le Fort 1975; Hodges 2000). For that reason the rocks forming the THS have probably better preserved the early tectonothermal evolution of the Himalayan orogen than other tectonometamorphic units, like the Greater Himalayan Sequence, made up by mid-crustal rocks which has nearly lost its

pre-climax memory during metamorphism around 23 –17 Ma (Guillot et al. 1993; Harrison et al. 1997; Searle & Godin 2003; Godin et al. 2006). The THS can be traced along the 2500 km of the Himalayan arc between the Nanga Parbat syntaxis in the west and the Namche Barwa syntaxis in the east (Fig. 1). The Cambrian to Eocene sequence is composed of very variable lithologies, derived from different sedimentary facies zones of the former passive continental margin of the Indian plate (e.g. Brookfield 1993; Willems et al. 1996; Garzanti 1999). Some tectonic domains are altered only diagenetically and usually low-grade metamorphism was not exceeded (e.g. Fuchs 1967; Hodges et al. 1996; Crouzet et al. 2007). The aim of this study is to constrain the postsedimentary evolution of the THS from metamorphic and geochronological data. In the study

From: Gloaguen, R. & Ratschbacher, L. (eds) Growth and Collapse of the Tibetan Plateau. Geological Society, London, Special Publications, 353, 45– 69. DOI: 10.1144/SP353.4 0305-8719/11/$15.00 # The Geological Society of London 2011.

46

I. DUNKL ET AL.

Fig. 1. (a) Position of the study area in SE Tibet (red box). (b) Major structural units of Central and Eastern Himalayas. The study area (red box, see Fig. 2) situated close to the eastern termination of Tethyan Himalayan Sequence. Violet dashed line shows the Indus-Yarlung suture zone; YTR, Yarlung Tsangpo River; GCT, Great Counter Thrust; STDS, South Tibetan Detachment System; GHS, Greater Himalayan Sequence; MCT, Main Central Thrust; LHS, Lesser Himalayan Sequence; MBT, Main Boundary Thrust (simplified after Pan et al. 2004; Yin 2006). (c) Schematic cross section along the red line in Figure 1b. KT, Kakhtang thrust, position of STDS from Grujic et al. (2002), McQuarrie et al. (2008). Structures in the northern Tethyan Himalaya from Antolin et al. (2011).

area (Fig. 1b) previous metamorphic and geochronological studies have been focused mainly on the Indus-Yarlung suture zone, Great Counter Thrust and Gangdese Thrust (e.g. Yin et al. 1994; Quidelleur et al. 1997; Harrison et al. 2000; Dupuis et al. 2005). However our studied profiles are distributed along valleys south of the IndusYarlung Suture where few work has been done before. Ku¨bler Index (KI) for ‘illite crystallinity’, vitrinite reflectance data, and K– Ar dating of micron and sub-micron fractions of illite are used to constrain the degree and age of metamorphism on metapelites, slates and sandstones of the Triassic flysch, which is the dominant metasedimentary sequence of the eastern THS. Thermobarometric methods (Thermocalc and PERPLEX, Holland & Powell 1998; Connolly & Petrini 2002, respectively) were applied on metamorphosed basic dyke rocks, which experienced a greenschist to amphibolite facies overprint.

Geological setting The major tectonic structures in the Himalayan Range are from south to north, the Himalaya Frontal Thrust (HFT), the Main Boundary Thrust (MBT), the Main Central Thrust (MCT), the South

Tibetan Detachment System (STDS), the Great Counter Thrust (GCT) and the Indus-Yarlung Suture Zone (IYSZ; Fig. 1b, c; Hodges 2000; Upreti 2001 and references therein; Yin 2006). These structural discontinuities divide the Himalaya into three main units traceable along the entire mountain belt. The Lesser Himalayan sequence between MBT and MCT is composed of Proterozoic –Cambrian sediments deposited on the proximal Indian shelf. Paleocene sediments are found in the eastern part of the belt (Sto¨cklin 1980; Valdiya 1980). This sequence was deformed by thrust and folds under very low-grade metamorphic conditions (Le Fort 1975; DeCelles et al. 2001; Robinson et al. 2003; Paudel & Arita 2006). The Greater Himalayan Sequence (GHS) in the hanging wall of the northdipping MCT consists of high grade metasediments, meta-igneous rocks (Le Fort 1971; Peˆcher 1975; Grujic et al. 2002) and leucogranitic intrusions in the footwall of the STDS (e.g. Guillot et al. 1993; Searle & Godin 2003). The uppermost unit, the Tethyan Himalayan Sequence is bordered by STDS and GCT and forms the cover of the Greater Himalayan Sequence. The middle part of the THS is characterized by the outcrop of the North Himalayan gneiss domes, which contains leucogranite bodies of Early Miocene age (e.g. Lee et al. 2000; Leech 2008).

METAMORPHIC EVOLUTION IN SE TIBET

The Tethyan Himalayan Sequence The Tethyan Himalayan Sequence involves a Cambro-Ordovician to Eocene pile that crops out along c. 150 km between STDS and IYSZ (Fig. 1; Brookfield 1993; Pan et al. 2004), deposited on the passive northern margin of the Indian continent (Fuchs 1967; Willems et al. 1996; Garzanti 1999). The Gyrong-Kangmar thrust in south –central Tibet divides the Tethyan Himalaya into two subunits (Liu & Einsele 1994). The southern sub-unit, north of the STDS, is formed by more than 13 km thick Cambrian to Eocene carbonates of the former passive margin, whereas the northern sub-unit is composed of Cretaceous clastic sediments recording the separation of the Indian plate from Gondwana and the drift to abyssal conditions (Willems et al. 1996). Towards the east, in SE Tibet, Lhunze fault (Fig. 2) separates the passive palaeomargin into two sub-units (Pan et al. 2004; Aikman et al. 2008). South of it, the Tethyan sediments are composed of Cretaceous clastic rocks, Jurassic –Cretaceous marine clastic rocks intercalated with intermediate –basic volcanic rocks, and marine limestones in the hanging wall of the STDS (Fig. 2; Pan et al. 2004). The sub-domain north of the Lhunze fault is build up by turbiditic sandstones and slates – subsequently called as flysch sequence. It was deposited in abyssal and bathyal environments between Middle Triassic and Early Jurassic (Mercier et al. 1984; Pan et al. 2004; Dupuis et al. 2005). In the north the flysch is juxtaposed against rocks of the active palaeomargin (Lhasa block), ophiolites related to the IYSZ and the me´lange complex (Heim & Gansser 1939; Ratschbacher et al. 1994; Yin et al. 1994; Quidelleur et al. 1997). The me´lange complex (limestones, cherts, marbles, shales, phyllites, andesites, diorites, mafic and ultramafic bodies) was deposited on the growing Neo-Tethys ocean floor (e.g. Searle 1986). The southern border of the Lhasa block is made up of Cretaceous clastic rocks and the Gangdese granite (Yin et al. 1994; Harrison et al. 2000; Pan et al. 2004). Ophiolites are widespread distributed along the suture and related to a suprasubduction environment (Ding et al. 2005; Dupuis et al. 2005, 2006).

Mafic magmatism In SE Tibet the Tethyan flysch sequence is penetrated by basalt, diabase, gabbroic diabase, diorite, olivine websterite and lherzolite dykes (Zhu et al. 2008; Xu et al. 2009). The silica content is typically between 43–55%. The thickness of the dykes is variable; it ranges from a few metres to c. 100 m. The texture and the typical crystal size are also very variable from fine grained, nearly aphanitic to

47

very coarse-grained holocrystalline. In the ultramafic members poikilitic or cummulate textures are typical. According to zircon SHRIMP U –Pb geochronology the dykes intruded 145– 130 Ma ago (Xu et al. 2009). Some of the magmatic bodies and typically their interior parts show well preserved magmatic textures, but in many occurrences the dykes are strongly deformed and transformed to chlorite or amphibole schists. The dykes form usually dissected boudins; the margins and the contact aureoles of the dykes are mainly detached or altered to banded, chaotic, mica-rich zones.

Structural setting of the Tethyan Himalaya The entire Tethyan sequence has been folded and imbricated during several tectonic phases. In the southern sector of the study area the main deformation phase (Eohimalayan phase: Hodges 2000) is related to the Middle Eocene to Late Oligocene collision of India against Asia. It is characterized by top-to-the-south thrust faults and south-facing folds (e.g. Burg & Chen 1984; Ratschbacher et al. 1994; Carosi et al. 2007; Aikman et al. 2008; Montomoli et al. 2008, Antolin et al. 2011). During the subsequent Neohimalayan (Hodges 2000) tectonic phase, north-facing folds with a related axial plane cleavage were developed (e.g. Carosi et al. 2007; Montomoli et al. 2008; Kellett & Godin 2009, Antolin et al. 2011). In the study area these two deformation phases are well developed. The first one is widespread in the southern sector, whereas the second one is well-developed in the northern portion. Consequently, two structural domains have been distinguished by Montomoli et al. (2008) and Antolin et al. (2011). Both deformation phases (D1 and D2) are associated to folds (F1 and F2 respectively) with related axial plane foliations (S1 and S2). The S1 foliation is associated to a synkinematic recrystallization of chlorite, white mica, calcite, quartz and oxides. The strain of the D2 phase increases progressively towards the north at the point that S2 foliation transposes the S1 foliation in the northern more strained areas. S2 foliation is a crenulation cleavage in the southern sector with no dynamic recrystallization, whereas in the north, it is associated to dynamic recrystallization of chlorite, white mica, calcite, quartz and oxides (Montomoli et al. 2008; Antolin et al. 2011). Godin (2003) used cross-cutting relationships of the different structural elements and U –Pb geochronology to date the D2 fold structures in the Annapurna area as Oligocene. In central Nepal Crouzet et al. (2007) dated the D2 tectonic phase by K –Ar ages of newly formed illite with c. 30–25 Ma. The synchronous activity of the STDS and the MCT along the Himalayan arc resulted in the exhumation

48 I. DUNKL ET AL.

Fig. 2. Simplified geological map of SE Tibet (after Pan et al. 2004; Yin, 2006; Aikman et al. 2008; Antolin et al. 2011).

METAMORPHIC EVOLUTION IN SE TIBET

of mid-crustal rocks of the GHS c. 23 –17 Ma ago (Godin et al. 2006 with references). By Th –Pb ion microprobe dating on monazite grains from the Khula Kangri granite, the age of activity of the STDS was dated to be younger than 12.5 Ma in the Eastern Himalaya (Fig. 2; Edwards & Harrison 1997). The displacement along the MCT in the western Arunachal Pradesh was dated to be c. 10 Ma by U –Th ion microprobe dating of monazite inclusions in synkinematic garnets (Yin 2006). During this time interval, the northern part of the Tethyan Himalaya was thrust along the GCT to the north at about 17 Ma in the south –central Tibet and between 18– 10 Ma near Zetang (Fig. 2; Ratschbacher et al. 1994; Harrison et al. 2000). The final tectonic phase took place during Miocene times when orogen parallel (east –west) extension triggered north–south-trending normal faults in form of graben structures, cross-cutting both the Lhasa block and the Tethyan Himalaya. In the study area related structures are formed by the Yadong Gulu Graben at the western margin of the study area and the Cona Graben at longitude of c. 928E (Fig. 2; Armijo et al. 1986; Garzione et al. 2003).

Samples The study area is situated south and SE of Lhasa along a 250 km long stripe where the Triassic flysch of the Tethyan sediments form the widest belt in the Himalayan chain (Figs 1 & 2). The present study focuses on north –south profiles along valleys between Nagarze in the west (908200 E) and Gyaca in the east (928500 E). From west –east, the main studied valleys are given by the name of the major localities (length of the profile in brackets): Nagarze (85 km), Zetang (45 km), Qusum (70 km) and Gyaca (17 km). 203 samples were collected from metapelites and arenites of the Triassic flysch sequence, from the basaltic –dioritic dykes, and greenschists formed from these magmatic rocks. The metapelites are typically dark grey to black with a total organic carbon content of 2– 4 wt% and a sulphur content of 0.01–0.3 wt%. They are rich in early diagenetic pyrite cubes with crystal size up to 2 cm. Sandstone layers (grey, well-sorted quartz-litharenites, with some wackes and red quartzites) are widespread in the sequence. The deformation of the flysch assemblage varies strongly along north –south profiles (e.g. Montomoli et al. 2008; Antolin et al. 2011) and the post-sedimentary overprint ranges from the diagenetic stage to greenschist and amphibolite facies. Thus, microscopic textures and mineral assemblages of samples from different structural domains vary from purely sediment to completely recrystallized schists (see Fig. 3a –d).

49

The mafic –intermediate magmatites usually form several metres wide dykes or house-sized, dissected boudins. They experienced a very variable degree of deformation. In the internal zones of the coarsest gabbro and diorite bodies even primary mafic minerals are well preserved, but the margins of the magmatic bodies commonly show transformation to chlorite–muscovite schist. In tectonic units where the degree of metamorphism is high penetrative, ductile deformation is widespread and the entire volumes of the magmatites were transformed to greenschist or amphibole –garnet schist (see Fig. 3e, f).

Methods Ku¨bler index (‘illite crystallinity’) For Ku¨bler index (KI) estimation and K –Ar dating, the fractions ,0.2, ,2 and 2–6 mm nominal size were separated from fine grained, pelitic –silty samples as follows. After crushing the samples to c. 1 cm size, only homogeneous, fine-grained rock fragments, completely free of limonite staining, detrital mica flakes and calcite or quartz veins were selected. The hand-picked aliquots were crushed and ground in a ring mill (vibration mill equipped by hard metal inlays) for 10 to 20 s, sieved and split into two parts. The ,63 mm fractions were used to extract the illite-rich fractions of ,2 mm by settling in Atterberg cylinders. A second aliquot of ,2 mm fraction was used to separate the fraction ,0.2 mm by an ultra-centrifuge. Due to the coarse crystal size of newly grown mica, it was not possible to produce a sufficient amount of ,0.2 mm fraction in samples with a relatively high degree of metamorphism. From these samples the ,2 and 2–6 mm fractions were investigated. The mineralogical composition of all fine aliquots was examined by X-ray diffraction. ˚ The KI is the half height peak width of the 10-A illite peak in X-ray diffractograms (Ku¨bler 1990). Digital measurement of KI was carried out by step scanning (301 points between 7 –108 2Q, with scan steps of 0.0108 2Q, integration time 4 s, receiving slit 0.1 mm, automatic divergence slit) on a Philips PW 1800 diffractometer. The KI values were calculated by the IDEFIX computer program developed at the Geoscience Centre of the University of Go¨ttingen by D. Friedrich and was rewritten to FORTRAN by K. Ullemeyer (Geomar, Kiel) in 2005. To detect the expandable layers of mixedlayer minerals, measurements were carried out in ‘air dry’ and ‘glycolated’ state of the size fractions. All samples were investigated in duplicates, and the KI-values are given in D82Q as an average of the two measurements. As suggested by Ku¨bler (1967, 1968, 1990) the anchizone is limited by 0.25 and

50

I. DUNKL ET AL.

Fig. 3. Macroscopic images and microphotographs demonstrating the characteristic textures and the mineral assemblages of the Tethyan (meta)flysch sequence. (a) Bottom-view of the bedding surface of a sandstone layer from the less deformed, southernmost zone of the sequence, which experienced only diagenetic overprint (site DB-19).

METAMORPHIC EVOLUTION IN SE TIBET

0.42 D82u, respectively. The used methodology has been validated by an inter-laboratory standardization program (see e.g. Warr & Rice 1994; Kisch ´ rkai et al. 2007). et al. 2004; A

Vitrinite reflectance Apparent maximum, minimum (Rmax, Rmin) and random vitrinite reflectances (Rr) were measured on polished sections cut perpendicular to the foliation, using polarized light and plane polarized light, respectively. The measurements were carried out at wavelength of 546 nm. Reflection was recorded on fine-dispersed vitrinite particles characterized by an elongated shape, smooth surface and strong bireflectance, without any traces of oxidation and/or re-deposition.

Thermobarometry Petrographic thin sections were polished, carbon coated and analysed using a JEOL JXA 8900 electron microprobe. A tungsten filament was used as electron source. The acceleration voltage for quantitative wave length dispersive spot analyses and line scans was set to 15 kV. The beam current was adjusted to 15 nA with a beam diameter of 5 mm. The count time for the peak position for each element was 15 s. The lower and upper background was measured for 5 s each. A Phi–Rho Z matrix correction was applied on all measurements (Armstrong 1991). The following phases were used as standards for the analysed elements: olivine_SC, albite, sanidine, TiO2, hematite, anorthite, wollastonite, rhodonite, Cr2O3 and NiO. The pressure and temperature (pT) conditions for samples WE-12 and Sr-21-a were calculated with Thermocalc (Holland & Powell 1998; Powell & Holland 2006) were obtained using the average pT mode or avpT. The PC Version 3.32 of Thermocalc with the internally consistent dataset tc-ds55s, obtained from the Thermocalc resource page (http://www.metamorph.geo.uni-mainz.de/ thermocalc/index.html) was used. To convert the mineral compositions into chemical activities the

51

program AX, by T. J. B. Holland (http://rock.esc. cam.ac.uk/astaff/holland/index.html) was used. After each run of Thermocalc the activities of the phases were iteratively recalculated with AX. This was done until a best fit between the pT results of Thermocalc and the calculation conditions for AX was achieved. The crucial sfit value of the avpT mode of Thermocalc (Powell & Holland 2006) was observed during the calculations. The quoted pT values are within the sfit range. The computer program PERPLEX (Connolly & Petrinin 2002; Connolly 2005) was applied to construct a phase diagram section or pseudosections (Powell & Holland 2008) of sample SR-21-a. The pseudosection was utilized to plot the garnet composition isopleths of the four garnet end members (almandine, pyrope, grossularite and spessartine). The necessary chemical information was obtained with a whole rock XRF analyses as well as electron microprobe spot analyses of garnet cores. The intersection of four isopleths yields the range of pT conditions during the initial growth period of the garnets. This approach is discussed by J. A. D. Connolly on the PERPLEX resource page (http://www. perplex.ethz.ch/perplex_pseudosection.html).

K – Ar geochronology The argon isotopic composition was measured in a Pyrex glass extraction and purification line coupled to a VG 1200 C noble gas mass spectrometer operating in static mode. The amount of radiogenic 40Ar is determined by isotope dilution method using a highly enriched 38Ar spike from Schumacher, Bern (Schumacher 1975). The spike is calibrated against the biotite standard HD-B1 (Fuhrmann et al. 1987). The age calculations are based on the constants recommended by the IUGS quoted in Steiger & Ja¨ger (1977). Potassium is determined in duplicate by Eppendorf Elex 63/61 flame photometer. The samples are dissolved in a mixture of HF and HNO3 according to the technique of Heinrichs & Herrmann (1990). CsCl and LiCl are added as an ionization buffer and internal standard, respectively. The analytical error for the K –Ar age

Fig. 3. (Continued) Width of image is c. 15 cm. (b) Microphotograph of the sandstone layer of photo (a). The texture elements are sedimentary in character, the detrital components are not oriented and deformed, the pressure solution is hardly detectable. Width of image is 4.3 mm. (c) Typical, crenulated texture of the low-grade metapelites. S1 and S2 are respectively the foliations developed during D1 and D2 tectonic phases according to Montomoli et al. (2008) and Antolin et al. (2011). Width of image is 3.8 mm. (d) In thermal aureole of the leucogranite intrusions of the Yala Xiangbo dome rutile crystals were grown in the metapelites. Width of image is 4.3 mm. (e) Schist-parallel and unoriented growth of millimeter sized muscovite crystals in a metamorphosed dyke. Width of image is 4.3 mm. (f ) The typical mineral assemblage of the greenschist-facies metamorphosed dykes: chlorite . calcite  quartz . sphene  albite. Metamorphic minerals show a preferred orientation and define a coarse grain S1 foliation (Montomoli et al. 2008; Antolin et al. 2011). Width of image is 1.1 mm. (g) Amphibole-garnet schist (site SR-21). Width of image is 4.3 mm. (h) Microphotograph showing the deformation of Dala granite. The bands of biotite were formed in a partly-crystallized, semi-ductile state of granite. Width of image is 8 mm).

52

I. DUNKL ET AL.

calculations is given on a 95% confidence level (2s). More details of argon and potassium analyses are given in Wemmer (1991).

(U – Th)/He thermochronology Zircon and apatite crystals were concentrated by standard mineral separation processes (crushing, sieving, gravity and magnetic separation). Only clear, intact, euhedral single crystal aliquots were dated. The shape parameters were determined and archived by multiple microphotographs and used for the correction of alpha ejection (Farley et al. 1996). The crystals were wrapped in c. 1  1 mm-sized platinum capsules and degassed in high vacuum by heating with an infrared diode laser. The extracted gas was purified using a SAES Ti –Zr getter at 450 8C. The chemically inert noble gases and a minor amount of other rest gases were then expanded into a Hiden triple-filter quadrupol mass spectrometer equipped with a positive ion counting detector. Crystals were checked for degassing of He by sequential reheating and He measurement. Following degassing, samples were retrieved from the gas extraction line, spiked with calibrated 230Th and 233U solutions. Zircons

were dissolved in pressurized Teflon bombs using distilled 48% HF þ 65% HNO3 in five days at 220 8C, while apatites in 2% HNO3. The actinide and Sm concentrations were determined by isotope dilution method using a Perkin Elmer Elan DRC II ICP-MS equipped with an APEX micro-flow nebulizer.

Results Sheet-silicate mineralogy and Ku¨bler Index values The studied grain size fractions are composed mainly of illite/muscovite, chlorite, paragonite, quartz and minor albite. Kaolinite, smectite and carbonates were not detected. The illite/chlorite ratios are typically high and thus well-suited both for KI determination and for K –Ar geochronology (DB-47 in Fig. 4), but sometimes the chlorite is the dominant sheet silicate (CE-6E in Fig. 4). In some samples there is a discrepancy between the crystallinity state of illite/muscovite and chlorite, although ´ rkai indices is the correlation of Ku¨bler and A ´ rkai 1991; Potel 2007). usually very good (e.g. A In these cases we assume that well-crystallized

Fig. 4. XRD patterns of oriented, air-dried ,2 and ,0.2 mm fractions demonstrating the characteristic mineralogical composition of the analysed samples. I/M, illite/muscovite; Chl, chlorite; Q, quartz; DB-47, A typical metapelite sample, the dominant component is illite/muscovite and the amounts of Q and Chl are subordinate. CI-6E, A rare and less optimal composition, as chlorite is more abundant than illite. DB-36, Beyond the well crystallized illite the chlorite peaks are very diffuse.

METAMORPHIC EVOLUTION IN SE TIBET

illite/muscovite crystals are detrital in origin, whereas the chlorite is newly-formed at low pT conditions. Paragonite is also present in several samples (Fig. 5a). Peak deconvolution performed on the double-peak white mica XRD patterns (using

53

FITYK, Wojdyr 2007) suggest that K –Na ‘mixed ´ rkai layer phases’ are present (Livi et al. 1997; A et al. 2003). The KI values dominantly indicate an anchizonal –epizonal overprint, only one sample is in the diagenetic zone (Table 1; Fig. 6a). The good correlation between the air-dried and glycolized samples indicates that there are practically no expandable smectite layers in the majority of the samples (the only exception is sample DB-19; see Fig. 5b). This property of the crystal lattice is important for the evaluation of K –Ar ages (see below).

K – Ar ages The lack of smectite interlayers in the dated samples indicates that the proportion of exchangeable cations is low that is, crystals are closed for cation migration. Thus, we do not need to count some unpredictable Ar diffusion processes due to a non-illitic lattice. The illite/muscovite K –Ar ages range between 106–14 Ma (see Table 2, Fig. 6b). The potassiumoxide content is in each case below the stoichiometric composition of muscovite and varies between 3.4–7.1 wt%. The proportion of radiometric argon shows an even wider scatter between 13 –98%. These two parameters correlate well (Fig. 7). The samples are grouped according to the present sheet-silicate assemblage. The illite-rich samples show the highest K content and the highest proportion of radiogenic argon. The chlorite-bearing fractions contain less K and Ar*, and even smaller values are observed in the paragonite-bearing samples. Figure 7 shows high amounts of atmospheric argon in paragonite and chlorite and indicates mixing between K-bearing and K-free phases.

Vitrinite reflectance Random vitrinite reflectance values range from 1.6 –4.1%Ro. Two samples (DB-37, DB-39) contain graphitized organic matter with Rmax values above 9.8%, Table 3). The areal distribution of the measured values in the study area is presented in Figure 6c.

Metamorphic pT conditions determined by Thermocalc and PERPLEX Fig. 5. (a) Diagnostic sections of XRD patterns showing badly and well crystallized illite (samples DB-19 and DB-14, respectively) and the paragonite-bearing samples (two peaks). (b) Plot of Ku¨bler Index (KI) measured on air dried v. KI measured on ethylene glycol treated samples. For the correlation coefficient and regression parameters only the values below 0.38 were considered. The 2s error bars of the presented values are smaller than the symbols.

Thermobarometric analyses were performed on two metamorphosed dykes collected in the NE part of the study area (WE and SR sites; see Fig. 2). The mineral paragenesis of sample WE-12 is Ms, Chl, Carb, Q, /Ep and in sample SR-21a Amp, Grt, Ep, Chl, Fsp, Q, /Carb (Fig. 3e, g). Mineralogical and textural evidences indicate the absence of a retrograde transformation or complex deformation.

54

Table 1. Geographical coordinates, illite Ku¨bler Index and K – Ar age of samples wehere beyond ‘illite crystallinity’ illite geochronology was also performed. Ku¨bler Index

Sample Lat.

Elev.

Fraction

air dry

(8)

(8)

(m)

(mm)

(D82u)

DB-1

89.6575

28.4848

4329

DB-5 DB-9 DB-14

90.1567 90.3183 92.2190

28.8988 28.8850 28.6208

4957 4637 5030

DB-19

91.6215

28.9048

4139

DB-21

91.6362

28.9271

4025

DB-23

91.6419

28.9333

4040

DB-25

92.1574

29.1036

3829

DB-26

92.1574

29.1036

3829

DB-32

91.6496

28.9522

3970

DB-36

91.6764

28.9881

3858

DB-38

91.6950

29.0650

3726

,0.2 ,2 ,6 ,6 ,0.2 ,2 ,0.2 ,2 ,0.2 ,2 ,2 2–6 ,0.2 ,2 ,0.2 ,2 ,0.2 ,2 ,0.2 ,2 ,0.2 ,2 ,0.2

0.193 0.193 0.153 0.205 0.179 0.170 0.434 0.387 0.242 0.207 0.199 0.179 0.509 0.228 0.232 0.185 0.227 0.205 0.194 0.166 0.184 0.180 0.245

DB-44

91.3174

29.0537

3963

glyc.

,0.2 mm (Ma)

+2 s

0.187 0.186 0.160

35.6

1.0

0.179 0.165 0.383 0.331 0.227 0.205

35.0

1.1

88.7

1.5

55.7

1.0

0.176 0.528 0.229 0.228 0.182 0.221 0.205 0.201 0.163 0.183 0.189 0.246

Ro

,2 mm

2 – 6 mm

(Ma)

+2 s

34.6

0.7

1.98 1.85 1.63

37.9

0.9

2.04

106.5

3.3

60.1 61.61

(Ma)

0.8 0.91

1.9

31.6

2.2

46.8

1.4

32.8

1.0

24.2

1.2

74.9

1.2

32.2

1.6

33.0

1.5

71.8

1.9

33.8

0.6

22.0

1.1

(%)

3.77 73.0

23.4

+2 s

1.0

3.05

graphite

I. DUNKL ET AL.

Long.

K– Ar

DB-45

91.1048

29.0483

4129

DB-47

91.1116

29.0857

3953

DB-55

90.6353

29.1977

4685

CI-6E

91.7515

29.1947

3821

CI-9A

91.7045

29.0730

3693

92.69

28.99

10611

DE7 DE8a DE19 DR14

92.54 92.57 90.8410 92.38

28.99 29.05 29.28 28.82

11493 11015 11761 14584

DR17

9.37

28.82

14686

QU19

92.21

28.99

13946

TU6b

92.22

28.83

14913

0.225 0.265 0.237 0.282 0.265 0.205 0.183 0.187 0.177 0.185 0.187 0.173 0.165 0.172 0.160 0.171 0.171 0.160 0.153 0.150 0.160 0.174 0.165 0.148

0.225 0.267 0.241 0.291 0.253 0.204 0.184 0.184 0.174

0.16 0.18 0.17 0.17 0.22 0.16 0.15 0.200 0.160

77.8

1.5

79.0

0.9

62.2

1.3

49.0

1.8

25.7

0.9

82.7

2.4

86.6

1.1

90.4

1

63.6

0.8

56.6

0.8

30.0

1.4

2.76

2.51 1.94

27.6

21.7

0.4

22.0

0.8

34.4

0.8

51.6

0.8

14.2

0.5

0.9

48.3

1.1

13.9

0.7

0.149

Note: Italics in the columns of Ku¨bler Index indicates that paragonite is also present among the sheet silicates. In some cases it was not possible to express the KI.

METAMORPHIC EVOLUTION IN SE TIBET

DE3

,2 ,0.2 ,2 ,0.2 ,2 ,0.2 ,2 ,0.2 ,2 ,0.2 ,2 ,2 2–6 ,2 2–6 ,0.2 ,2 2–6 ,2 2–6 ,2 2–6 ,2 2–6

55

56

I. DUNKL ET AL.

Fig. 6. Areal distribution of metamorphic and geochronological results in SE Tibet. Grey area, Triassic flysch. (a) Ku¨bler Index (fraction ,2 mm), (b) K –Ar ages of illite fractions ,2 mm (in ellipses) and (U –Th)/He ages (in rectangles), (c) vitrinite reflectance. G: graphite particles recorded in the organic matter. Uncertainties of the plotted values and analytical details are in Tables 1– 5.

METAMORPHIC EVOLUTION IN SE TIBET

57

Table 2. K –Ar ages of illite fractions of fine grained (meta)pelitic members of Tethyan flysch and of coarse muscovites formed in metabasic rocks Sample

Fraction (mm)

K2O (wt. %)

40

Ar* ([nl/g]) STP

40

Ar* (%)

Age (Ma)

2 s-Error (Ma)

Illite-rich fractions from (meta)pelitic samples from the Tethyan flysch DB-1 ,2 4.93 5.56 DB-1 ,0.2 4.81 5.57 DB-14 ,2 7.11 8.79 DB-14 ,0.2 6.75 7.69 DB-19 ,2 6.09 21.56 DB-19 ,0.2 6.47 18.98 DB-21 ,2 6.44 12.70 DB-21 ,0.2 6.32 11.53 DB23 ,2 5.17 10.45 DB23 2–6 4.18 10.04 DB-25 ,2 3.44 3.61 DB-25 ,0.2 3.63 2.76 DB-26 ,2 3.65 3.92 DB-26 ,0.2 3.58 3.67 DB-32 ,2 3.81 9.00 DB-32 ,0.2 5.61 8.58 DB-36 ,2 6.49 7.15 DB-36 ,0.2 6.07 6.48 DB-38 ,2 5.24 3.74 DB-38 ,0.2 4.34 3.40 DB-44 ,2 5.32 14.53 DB-44 ,0.2 5.42 13.36 DB-45 ,2 5.51 15.76 DB-45 ,0.2 5.67 14.55 DB-47 ,2 5.65 16.90 DB-47 ,0.2 6.08 15.82 DB-55 ,2 6.05 12.63 DB-55 ,0.2 5.92 12.07 CI-6E ,2 4.83 8.96 CI-6E ,0.2 4.89 7.83 CI-9A ,2 4.27 4.17 CI-9A ,0.2 4.56 3.80 DE3 2–6 6.52 4.60 DE8a 2–6 2.91 2.07 DR14 ,2 5.92 5.31 DR14 2 –6 5.15 5.77 QU19 ,2 6.04 9.53 QU19 2 –6 5.14 8.67 TU6b ,2 5.24 2.36 TU6b 2 –6 3.86 1.77

50.3 54.3 54.7 47.0 94.4 94.0 87.6 86.5 83.9 76.3 19.6 12.9 23.1 21.8 79.5 72.9 56.0 52.0 31.0 27.7 89.0 85.1 91.7 90.4 90.6 88.5 82.1 82.5 82.8 71.2 45.7 42.7 76.7 30.1 33.2 45.4 44.7 64.3 19.1 28.4

34.6 35.6 37.9 35.0 106.5 88.7 60.1 55.7 61.6 73.0 32.2 23.4 33.0 31.6 71.8 46.8 33.8 32.8 22.0 24.2 82.7 74.9 86.6 77.8 90.4 79.0 63.6 62.2 56.6 49.0 30.0 25.7 21.7 22.0 27.6 34.4 48.3 51.6 13.9 14.2

0.7 1.0 0.9 1.1 3.3 1.5 0.8 1.0 0.9 1.0 1.6 1.9 1.5 2.2 1.9 1.4 0.6 1.0 1.1 1.2 2.4 1.2 1.1 1.5 1.0 0.9 0.8 1.3 0.8 1.8 1.4 0.9 0.4 0.8 0.9 0.8 1.1 0.8 0.7 0.5

Coarse muscovite of the greenschist (meta-basalt) WE-12 125–250 7.57 WE-10A 125–250 8.52

96.8 98.4

44.1 43.7

0.5 0.6

Site WE-12. For Thermocalc method 19 analyses of a total of 74 spots were selected. Table 4 shows the averages of the selected analyses for the Thermocalc computations as well as the cations per formula unit. The calculations of the cations per formula unit were done with AX. For the calculations the CO2-fraction of the fluid-phase was set to 10%. The obtained pT results for sample WE-12 are T ¼ 474 + 35 8C & P ¼ 6.4 + 1.6 kbar (sfit ¼ 1.64). Even though

10.90 12.16

Thermocalc estimates a 2s temperature error of 35 8C (output), because of various methodological error sources (Kohn & Spear 1991; Powell & Holland 1994), it should be replaced by a minimum error of 50 8C (Powell & Holland 2008). Site SR-21a. Thermocalc evaluations of sample SR-21-a were performed on mineral assemblages in the vicinity of garnets. According to microstructural

58

I. DUNKL ET AL.

Table 3. Vitrinite reflectance values measured on (meta)pelitic samples of SE Tibet Sample

Long. (8)

Lat. (8)

Elev. (m)

Ro (%)

sd

DB-1 DB-3 DB-5 DB-10 DB-13 DB-15 DB-16 DB-19 DB-21 DB-23 DB-28 DB-30 DB-32 DB-33 DB-42 DB-45 DB-47 DB-54 DB-55 CI-9B CI-17M CI-15B DE-16C DE-17B

89.6575 89.6640 90.1567 90.5334 92.2446 92.2190 92.1906 91.6215 91.6362 91.6419 92.0433 92.0321 91.6496 91.6564 91.3024 91.1048 91.1116 90.5926 90.6353 91.7045 91.7210 91.7107 91.7210 91.6671

28.4848 28.4967 28.8988 28.4597 28.5411 28.6208 28.7015 28.9048 28.9271 28.9333 29.1454 29.1714 28.9522 28.9649 29.0303 29.0483 29.0857 29.1888 29.1977 29.0730 29.1487 29.1275 29.1487 29.0150

4329 4341 4957 5128 4373 5030 4844 4139 4025 4040

1.98 2.01 1.85 1.63 2.40 2.04 2.20 4.09 3.77 3.05 1.66 1.96 1.84 2.01 3.01 2.76 2.78 2.36 2.51 1.94 1.86 2.02 1.88 2.00

0.27 0.29 0.30 0.16 0.19 0.33 0.26 0.24 0.21 0.31 0.27 0.25 0.24 0.30 1.59 0.36 0.28 0.26 0.22 0.21 0.25 0.18 0.21 0.21

3614 3970 3939 4092 4129 3953 4502 4685 3693 3634 3648 3634 3846

Graphite-bearing samples DB-37 91.6794 29.0099 DB-39 91.3100 29.0874 DB-39 91.3100 29.0874

3812 3655 3655

SR-21a

3389

92.86

28.95

Rmax (%)

sd

Rmin (%)

sd

N 50 22 19 24 39 14 16 30 30 50 22 19 6 5 8 50 50 50 50 22 30 5 30 30

10.4 9.85

0.98 0.84

1.19 1.43

0.38 0.73

26 18

graphite particles graphite

observations, the Grt– Fsp– Chl–Am assemblage is in equilibrium. In the surroundings of eight selected garnet crystals a total of 122 spot analyses were performed in the adjacent mineral phases. Analyses of the garnet rims were used to estimate the metamorphic conditions with Thermocalc (Spear 1995). The garnet compositions determined in the cores of the crystals forms the base for the PERPLEX method. The results of two data sets yielding the smallest 2s errors (see Table 4) are 531 8C (2s ¼ 13 8C), 9.4 kbar (2s ¼ 1.0 kbar) at sfit ¼ 0.87 and 511 8C (2s ¼ 12 8C), 10.0 kbar, (2s ¼ 0.9 kbar) at sfit ¼ 0.48, respectively. The reported 2s errors are underestimating the real uncertainty (see above), thus we use a minimum of 50 8C and 1 kbar respectively. For the PERPLEX calculations 10 spot analyses of garnet cores, with totals between 99.5 wt% and 101 wt%, were selected and averaged (Table 4). The PERPLEX garnet isopleths for sample SR-21-a are depicted in Figure 8. The pT result for this sample is estimated by 600 + 50 8C and 7.4 + 1.5 kbar.

(U –Th)/He ages Zircon and apatite helium ages (ZHe and AHe, respectively) together with the analytical details are listed in Table 5. The dated localities are plotted in Figure 6b. The amounts of actinide elements for all single crystal ZHe measurements are at least 20 times higher than the limit of detection. The ZHe ages are slightly different in the samples, but they cluster mainly in the Late Miocene (12.2 – 7 Ma), only the southernmost sample (DB-12) resulted in an Oligocene age. Only one sample of a metamorphosed dyke (WE-12) contains proper apatite crystals for He-chronology resulting in a Late Miocene AHe age.

Discussion Low-grade pelites potentially contain detrital white mica grains, carrying the signal of the crystallization conditions and age of the source region of the sediment (Hower et al. 1963; Hurley et al. 1963). The

Table 4. Average chemical and cation composition of the mineral phases used for thermobarometry. Oxygen numbers used for cation numbers are: MS: 11, Chl: 14, Carb: 6, Ep: 12.5, Grt: 12, Fsp:8 Amph: 23 WE-12 greenschist SiO2 46.33 24.61 0.02 38.49 Si 3.10 2.60 0.00 3.02

TiO2 0.17 0.04 0.01 0.08 Ti 0.01 0.00 0.00 0.01

Al2O3 35.44 23.47 0.01 25.29 Al 2.80 2.92 0.00 2.34

Cr2O3 0.00 0.08 0.01 0.01 Cr 0.00 0.01 0.00 0.00

Fe2O3 0.00 0.00 0.00 10.08 Fe3 0.00 0.00 0.00 0.60

FeO 1.04 26.18 1.23 0.61 Fe2 0.06 2.31 0.03 0.04

MnO 0.01 0.31 0.75 0.03 Mn 0.00 0.03 0.02 0.00

MgO 0.73 13.12 0.56 0.01 Mg 0.07 2.06 0.03 0.00

CaO 0.03 0.01 59.04 23.70 Ca 0.00 0.00 1.92 1.99

Na2O 1.09 0.01 0.01 0.00 Na 0.14 0.00 0.00 0.00

K2O 9.01 0.01 0.02 0.01 K 0.77 0.00 0.00 0.00

SR-21a amphibole-garnet schist: compositios used for Thermocalc Assemblage #1 Grt Fsp Chl Am Assemblage #2 Grt Fsp Chl Am Assemblage #1 Grt Fsp Chl Am Assemblage #2 Grt Fsp Chl Am

SiO2 38.15 66.34 25.66 45.08

TiO2 0.10 0.00 0.06 0.42

Al2O3 20.97 21.32 19.93 14.31

Cr2O3 0.00 0.00 0.07 0.03

Fe2O3 0.00 0.27 0.27 0.27

FeO 27.01 0.00 28.94 17.14

MnO 2.93 0.03 0.44 0.24

MgO 1.72 0.00 12.65 8.73

CaO 9.79 2.10 0.03 9.66

Na2O 0.02 10.50 0.00 2.43

K2O 0.00 0.03 0.01 0.11

37.58 68.31 24.85 45.63 Si 3.02 2.90 2.75 6.61

0.09 0.00 0.04 0.32 Ti 0.01 0.00 0.01 0.05

20.46 19.75 18.83 11.51 Al 1.96 1.10 2.52 2.48

0.03 0.01 0.00 0.04 Cr 0.00 0.00 0.01 0.00

0.10 0.00 0.00 3.30 Fe 3 0.00 0.01 0.02 0.03

27.09 0.00 28.40 12.87 Fe 2 1.79 0.00 2.60 2.10

3.13 0.00 0.42 0.17 Mn 0.20 0.00 0.04 0.03

1.77 0.00 12.86 10.38 Mg 0.20 0.00 2.02 1.91

8.97 0.24 0.02 10.01 Ca 0.83 0.10 0.00 1.52

0.02 11.67 0.01 2.02 Na 0.00 0.89 0.00 0.69

0.00 0.04 0.01 0.14 K 0.00 0.00 0.00 0.02

3.02 2.98 2.75 6.77

0.01 0.00 0.00 0.04

1.94 1.02 2.46 2.01

0.00 0.00 0.00 0.01

0.01 0.00 0.00 0.37

1.82 0.01 2.63 1.60

0.21 0.00 0.04 0.02

0.21 0.00 2.12 2.30

0.77 0.01 0.00 1.59

0.00 0.99 0.00 0.58

0.00 0.00 0.00 0.03

FeO 26.90 Fe 2 1.79

MnO 3.96 Mn 0.27

MgO 1.65 Mg 0.20

CaO 9.14 Ca 0.78

Na2O 0.00 Na 0.00

K2O 0.00 K 0.00

METAMORPHIC EVOLUTION IN SE TIBET

Mineral Ms Chl Carb Ep Mineral Ms Chl Carb Ep

SR-21a amphibole-garnet schist: average garnet core composition used for PERPLEX SiO2 37.90 Si 3.02

TiO2 0.13 Ti 0.01

Al2O3 20.56 Al 1.93

Cr2O3 0.03 Cr 0.00

Fe2O3 0.15 Fe 3 0.01

59

Mineral Grt Mineral Grt

60

Table 5. Results of (U –Th)/He geochronology Sample

Zircon DB-27 (3865 m)

MV31c (3440 m) TU-4a (4655 m)

DB-12 (4373 m)

Apatite WE-12 (4437 m)

He

U238

Th232

Sm

Ejection

Uncorr.

Ft-Corr.

vol. (ncc)

s.e. (ncc)

mass (ng)

s.e. (ng)

Mass (ng)

s.e. (ng)

Th – U Ratio

Mass (ng)

s.e. (ng)

correct. (Ft)

He-age (Ma)

He-age (Ma)

1 s (Ma)

#1 #2 #3 #1 #2 #3 #1 #2 #1 #2 #3 #4 #1 #2 #3

2.661 0.981 0.781 2.591 5.782 2.193 0.633 1.582 5.245 1.934 1.749 1.268 4.431 2.151 2.227

0.045 0.017 0.014 0.044 0.096 0.037 0.012 0.027 0.087 0.033 0.029 0.022 0.073 0.036 0.037

3.048 1.125 0.890 2.048 4.092 1.945 1.043 2.234 5.136 1.804 1.643 1.242 1.564 0.735 0.716

0.055 0.020 0.016 0.037 0.074 0.035 0.019 0.040 0.093 0.033 0.030 0.023 0.028 0.013 0.013

1.195 0.538 0.718 0.765 2.234 0.406 0.196 0.523 1.873 1.013 0.869 0.465 0.764 0.627 0.770

0.029 0.013 0.017 0.018 0.054 0.010 0.005 0.013 0.045 0.024 0.021 0.011 0.018 0.015 0.019

0.39 0.48 0.81 0.37 0.55 0.21 0.19 0.23 0.36 0.56 0.53 0.37 0.49 0.85 1.08

0.028 0.052 0.065 0.031 0.052 0.028 0.004 0.016 0.069 0.040 0.021 0.022 0.032 0.019 0.016

0.002 0.003 0.004 0.002 0.003 0.002 0.000 0.004 0.002 0.002 0.001 0.001 0.002 0.001 0.001

0.70 0.76 0.80 0.78 0.80 0.79 0.74 0.73 0.77 0.78 0.74 0.74 0.68 0.69 0.71

6.6 6.5 6.1 9.6 10.4 8.9 4.8 5.6 7.8 7.8 7.8 7.8 21.0 20.1 20.5

9.5 8.5 7.6 12.4 13.0 11.2 6.5 7.6 10.1 10.0 10.6 10.5 30.7 29.3 28.9

0.2 0.2 0.2 0.3 0.3 0.3 0.2 0.2 0.2 0.2 0.2 0.2 0.7 0.7 0.6

#1 #2 #3 #4 #5

0.001 0.001 0.008 0.011 0.026

0.000 0.000 0.001 0.001 0.001

0.001 0.001 0.005 0.007 0.013

0.001 0.001 0.001 0.000 0.001

0.004 0.005 0.018 0.029 0.066

0.000 0.000 0.001 0.001 0.002

7.52 5.33 3.86 4.47 4.99

0.119 0.160 0.553 0.846 1.752

0.011 0.015 0.050 0.077 0.158

0.78 0.67 0.71 0.76 0.76

4.4 3.8 4.6 4.4 5.0

5.6 5.6 6.4 5.8 6.6

1.9 1.5 0.6 0.4 0.4

Sample unweighted aver. (1 s.e.)

8.5

0.5

12.2

0.5

7.0

0.5

10.3

0.1

29.6

0.5

6.0

0.2

I. DUNKL ET AL.

DR13 (4523 m)

aliq.

METAMORPHIC EVOLUTION IN SE TIBET

61

distinction of the inherited and newly formed generations of sheet silicates in grain size fractions is difficult and such aliquots result in typically mixed ages. However, white micas formed in basic and intermediate magmatic rocks are exclusively metamorphic in origin. Therefore these white micas are free of any inherited signals and their K –Ar ages reflect the age of metamorphism or cooling. Thus, we will separately discuss the results yielded from metapelitic and metabasic lithologies.

Conditions and age of metamorphism of some basic dykes

Fig. 7. Plot of potassium content v. the proportion of radiogenic argon and the mineralogy of the dated sheet-silicate rich fractions. The 2s error bars of the presented values are smaller than the symbols. The DB-36 sample contains the badly crystallized chlorite (see Fig. 4).

Fig. 8. Metamorphic pressure and temperature conditions (white area) of amphibole-garnet schist of SR 21 site determined by PERPLEX method. Paired garnet isopleths demarcate the stability field of garnet that is in equilibrium with the bulk rock chemistry and mineral paragenesis. The lines show the +1 s.d. of the chemical compositions determined in the cores of the garnet crystals by multiple electron microprobe analyses (see Table 4).

Thermobarometric analyses indicate greenschist facies metamorphism for the WE sites and amphibolite facies metamorphism for the SR sites. The greenschists of the WE sites contain well developed white mica crystals (Fig. 3e). They are aligned along the main foliation, indicating crystallization during the S1 tectonic phase (Montomoli et al. 2008; Antolin et al. 2011). White mica does not occur in mafic magmatic rocks as primary phase and the texture of this folded greenschist does not show any relict magmatic element. Therefore, the muscovites are completely metamorphic in origin. In these sites S2 foliation is a crenulation cleavage not associated to dynamic recrystallization (Antolin et al. 2011). The corresponding K– Ar ages are 43.7 and 44.1 Ma. Noticeable are the high percentages of radiogenic argon (Table 2). The two samples were collected in c. 1 km distance from two distinct metamagmatite bodies of different chemistry and deformation degree. We interpret these ages as the formation age of the muscovites, recording the age of greenschist facies metamorphism that took place in a part of the THS east of the Yala Xiangbo dome. These ages resemble the emplacement age of the Dala granite (44.1 + 1.2 Ma, U –Pb zircon dating; Aikman et al. 2008), located 25 km SE of the sampled sites. At this location the Dala granite intruded into a low structural level and experienced some near-solidus deformation (Fig. 3h) that occurred simultaneously or soon after the emplacement. Similar ages (c. 44 Ma) for peak metamorphic conditions have been detected also in the underlying GHS tectonic unit using Th –Pb and U –Pb datings on monazites (Catlos et al. 2002, 2007; Carosi et al. 2010). The key observation for the amphibole-garnet schist of SR site is the difference in the calculated pressures using Thermocalc and PERPLEX methods. The Thermocalc p-estimate of the garnet rim assemblages is at least 2 kbar higher than the PERPLEX estimate of the garnet core. This indicates prograde metamorphism during garnet growth.

62

I. DUNKL ET AL.

Here it is important to note that the studied site is far away (.60 km) from the next metamorphic dome (Yala Xiangbo dome). Thus an immediate heating from the lower structural unit has to be excluded. The available samples did not allow geochronological dating, but we assume that the amphibolite facies metamorphism was co-genetic with the c. 44 Ma old greenschist facies event detected at the WE sites, discussed above.

Low-grade metamorphism of metapelitic samples The geothermometrical and geochronological data from sub-greenschist facies pelitic lithologies are not so well constrained as the results measured on orthometamorphic rocks of higher pT conditions (like the data discussed above). Anchi- and epizonal conditions result in incomplete, disequilibrium phase transformations. This fact is supported by field observations. The metapelitic samples show a very variable intensity of white-mica growth, even detectable within a single outcrop. The mica growth process is strongly controlled by the host lithology. In silty and sandy protoliths the neoformation of white mica is in a more advanced state than in neighbouring pelitic lithologies. We assume that three major factors are responsible for this difference: (1) the high permeability of the arenitic lithologies; (2) the local liberation of potassium and increase of Kþ activity during decomposition of feldspar grains and lithic fragments in arenites; and (3) the presence of dispersed organic material in pelitic lithologies probably embedding the silicate phases and reducing locally the diffusivity and mineral growth. Figure 6 presents the results obtained on metapelitic samples projected on the schematic geological map of the study area. The sample sites with the highest metamorphic degree are related to tectonic slices derived from the deeper part of the THS situated close to the Yala Xiangbo dome (e.g. site TU-6). For the proper understanding of the evolution of the studied THS area we have to evaluate the KI data in concert with the argon geochronology. In epi- and anchizonal overprinted areas the mineral transformation is usually incomplete and the individual crystallinity and age data typically show apparent values, which are actually points along mixing or transformation curves. The interpretation of individual data is difficult and can result in misleading conclusions. Thus, we do not force an individual interpretation for each single sample or sample site, but rather to process all data synoptically in order to identify the major epochs of the thermal evolution for the entire eastern THS.

The KI values are controlled by the ratio of newly formed to detrital illite. It usually increases towards finer size fractions. Using two fractions, the finer one is richer in newly formed sheet silicates and always indicates a value closer to the conditions of the latest metamorphic event (Reuter 1987; Clauer & Chaudhuri 1999). The same is true for the K –Ar ages of these size fractions. The detrital grains carry an old age signal, while the newly grown population gives always a younger K –Ar age (Hower et al. 1963; Hurley et al. 1963). By combining the KI with the K – Ar ages, the ratio of newly formed to detrital illite can be estimated, especially if the initial detrital age is known. The KI measured on different size fractions and the corresponding K– Ar ages of the metapelites in SE Tibet are plotted on Figure 9. The plot shows clear trends. The older K –Ar ages are measured in samples having a lower KI. With increasing KI the argon ages are getting younger. The ,0.2 mm fractions, being rich in newly formed illite, show less ordered illite structures and younger K –Ar ages than the ,2 mm fractions. The distances between the projection points of the two size fractions become smaller with increasing metamorphic degree. For samples showing the youngest K –Ar ages, the two fractions give indistinguishable results. We assume that in these cases both size fractions are dominated by the newly formed white mica. This convergence indicates equilibrium conditions and the ages (c. 24 –22 Ma) are considered as the cessation of illite growth. These sites are structurally controlled by the D2 deformation and S2 foliation became the main penetrative foliation (Harrison et al. 2000; Montomoli et al. 2008; Antolin et al. 2011).

Fig. 9. Plot of Ku¨bler Index v. K–Ar age measured on the same fraction. Note that ,0.2 mm fractions show smaller degree of illite crystallinity and younger K– Ar age compared to ,2 mm fractions.

METAMORPHIC EVOLUTION IN SE TIBET

For this reason the new growth of illite at c. 24– 22 Ma is most likely related to the development of the S2 foliation.

Estimation of the maximum metamorphic temperature of metapelites by organic maturation The vitrinite reflectance cannot be converted directly to temperature, because the transformation of organic material is a kinetic process (e.g. Barker & Pawlewicz 1986; Sweeney & Burnham 1990). Peak temperature estimation for the Triassic flysch was performed by means of three different algorithms, assuming different effective heating times (Fig. 10). For the selection of the most reliable curve we have to consider the effective heating time. The above outlined K –Ar ages indicate that the final metamorphism of the Triassic flysch took place in Oligocene –Miocene time, thus we exclude both long-lasting maximum temperature conditions as well as a shock-heating process. The most probable duration of the maximum temperature is between c. 5 and 15 Ma. Thus, we use the Bostick (1979) and the Sweeney & Burnham (1990) algorithms assuming 10 Ma effective heating time to estimate the range of maximum temperature (Fig. 10). Considering this conversion, the lowest vitrinite reflectance values (around Ro ¼ 1.65%) indicate c. 170 –185 8C maximum temperature. Typical reflectance values around 2–3%Ro

63

indicate c. 180–200 and 225–235 8C maximum temperatures, respectively. Above c. 4% reflectance the transformation of Ro values to temperature is not properly calibrated (see e.g. Judik et al. 2008), thus for the estimation of maximum metamorphic temperature of the Triassic flysch we have to use the inorganic mineral phases of the metamorphosed dikes of the sequence (see above).

Miocene greenschist facies metamorphism at the base of THS The sample set of site TU-6 was collected close to the detachment of the Yala Xiangbo dome, a zone that was intruded by thin leucogranitic and aplitic dykes. The dykes have only weekly developed chilled margins. This sample site represents a deep structural level of the THS. Newly grown, well developed muscovite crystals dominate the microtexture and KI indicates a highly crystalline lattice of the white mica. The muscovite K –Ar ages are the youngest in the studied sample set (14.2– 13.9 Ma; Fig. 6b). This age range is close to the c. 13.5 Ma muscovite K– Ar age reported from the Yala Xiangbo dome by Aikman et al. (2004).

Post-sedimentary metamorphic evolution of the Tethyan flysch in SE Tibet The evolution of the region is rather complex. To better describe the evolution of the eastern

Fig. 10. Estimation of palaeotemperature from vitrinite reflectance values. Three different algorithms were used for conversion assuming different effective heating times. Grey belt represents the vitrinite reflectance values (except the graphitized samples). Evaluation is in the text.

64

I. DUNKL ET AL.

Fig. 11. Synopsis of the tectonothermal evolution of the Tethyan Himalayan flysch in SE Tibet. Time intervals emphasized by grey vertical lines are the major deformation periods dated as c. 44 Ma and an Oligo-Miocene one, which terminated at c. 24–22 Ma. Different processes were resulted in white mica growth, which took place in more periods, both in diagenetic and in metamorphic conditions. Lines represent the evolutions of different tectonic blocks of THS.

Tethyan Himalayan Sequence we compiled therefore the known thermal and tectonic events and the assumed mineral transformations and illite forming periods in a scheme (Fig. 11). The burial history is reconstructed from the subsidence curve of Jadoul et al. (1998) estimating the minimum amount of burial of the Triassic flysch. In their study the sedimentary sequence was interpreted as a near-shore facies assemblage, thus, according to the observed facies differences, the thickness of the total Mesozoic pile in the study area is probably much higher. In the assumed depth of at least 4 to 5 km, the burial diagenesis can already generate new clay mineral assemblages (Meunier & Velde 2004), thus we consider the

sedimentary burial as the first illite-forming epoch (Fig. 11). The oldest argon ages of the coarsest size fractions, dominated by detrital mica, are much younger than the typical Precambrian mica ages of the Indian basement rock, being presumably the source area of the flysch (Gaetani & Garzanti 1991). They are even younger than the c. 250– 210 Ma age of sedimentation. On the other hand, the oldest detrital (mixed) ages are older than the Eocene (c. 44 Ma) greenschist facies metamorphic event detected in the metabasic rocks. The lack of pre-Mid Cretaceous detrital mica ages indicates that there was a post-sedimentary reset older than the Cenozoic collision of THS. The most plausible

METAMORPHIC EVOLUTION IN SE TIBET

candidate for the post-Triassic, pre-Eocene thermal reset of the detrital mica Ar-age is the high heat flow during Early Cretaceous magmatism of the region (Zhu et al. 2008; Xu et al. 2009). We assume that at c. 140 Ma a significant part of the THS was in anchi- and epizonal conditions and the emplacement of the dykes and the formation of their hydrothermal aureoles triggered the formation of white mica in the pelites (Fig. 11). During Eocene times the THS experienced locally greenschist and probably also amphibolite facies metamorphism. This metamorphic event was related to the underthrusting of tectonic units during the Himalayan collision. A part of our data measured on metapelites shows a very weak overprint, thus several tectonic blocks-mainly in the southern part of the THS-occupied only shallow depths during the Eocene subduction. During Oligo-Miocene times the ongoing shortening deformed the northern zones of the THS (Harrison et al. 2000; Montomoli et al. 2008; Antolin et al. 2011). The formation of illite is common, but the intensity of the deformation and the mineral growth is very variable (Fig. 11). The majority of the KI and K– Ar data from the THS were formed in disequilibrium conditions and actually they are the results of pre-Neogene and Neogene events (see e.g. Fig. 9). This indicates that the maximum thermal overprint during Miocene times usually did not exceed anchizonal–epizonal conditions. After the Oligocene to Miocene stacking, the pile of the Tethyan Himalaya was penetrated by the Yala Xiangbo dome and associated leucogranitic dykes (TU-6 site) reached greenschist facies conditions. The youngest K –Ar ages are around 14 Ma and they were measured on samples from the deepest part of the sequence, probably close to the basal detachment (represented by gray line in Fig. 11). However, this overprint was local, because in the main part of the THS the illite K –Ar ages typically show only partial reset and the newly grown illite has a low degree of crystallinity. The obtained ZHe ages (between 30 and 7 Ma) are interpreted as cooling ages. Samples from the northern part of THS indicate a complete reset in the Miocene and prove that the currently exposed level of the THS was situated deeper than the c. 180 8C isotherm until the Late Miocene. The oldest ZHe age (sample DB-12: 39.6 Ma) was measured on a site at the southern margin of the Triassic flysch belt, which experienced only a diagenetic overprint. This Oligocene He-age and the weak overprint indicate that some tectonic blocks in the southern part of the flysch belt were in a shallow position, both during Eocene subduction and during Oligo-Miocene shortening (represented by dashed line in Fig. 11).

65

The only sample that contains datable apatite crystals yields a 6 Ma AHe age. This single datum and the calculated c. 70 8C closure temperature suggest an average post-Miocene cooling rate of c. 10 8C/Ma.

Conclusions † The different tectonic blocks of the THS in SE Tibet experienced a thermal overprint between c. 170–600 8C. † The Tethyan Triassic flysch sequence registers four tectonothermal events. (1) Early Cretaceous. Due to the subsequent events its direct dating is not possible. Nevertheless from the hot-spot related magmatism penetrating the region and from the missing pre-Mid Cretaceous illite/muscovite argon ages, we assume an Early Cretaceous period of high heat flow resulting in the formation of illite in the metapelites. (2) Eocene (c. 44 Ma). The early Himalayan granites (Dala granitoids; Aikman et al. 2008) intruded during or slightly before the greenschist and locally probably even amphibolite facies metamorphism. Maximum temperature and pressure conditions of c. 600 8C and 7.8 kbar indicate that a part of the THS was subducted to midcrustal levels. This metamorphic event is probably contemporaneous with the collision related deformation phase (e.g. Godin 2003; Carosi et al. 2007) or Eohimalayan phase (Hodges 2000; Guillot et al. 2003). (3) Oligo-Miocene (terminated at c. 22 Ma). The northern part of the THS, from Zetang to the east, experienced anchi- to epizonal metamorphism with a of deformation and thermal alteration. We assume that this process was associated to a crustal shortening period that probably terminated c. 22 Ma ago. This north–south shortening phase can be correlated to the D2 tectonic phase as defined by Godin (2003), Kellett & Godin (2009), and Antolin et al. (2011), or to the Neohimalayan phase of Hodges (2000). (4) Miocene (between c. 18 and 13 Ma). The very base of the Tethyan Himalayan Sequence in the surroundings of the Yala Xiangbo dome experienced a greenschistfacies overprint. The formation of white micas or the cooling below their argon closure temperature took place c. 13 Ma, caused by the emplacement and exhumation of the Yala Xiangbo dome.

66

I. DUNKL ET AL.

† Zircon and apatite (U –Th)/He ages indicate that the post-metamorphic cooling history lasted until Late Miocene times. The final cooling was not coeval in the whole THS, the northern zones experienced a later cooling probably induced by the exhumation of the hanging wall of the south-dipping Great Counter Thrust along its backthrust plane. The published thrusting ages of the MCT, STDS (see Godin et al. 2006) and GCT (Yin et al. 1994; Ratschbacher et al. 1994; Quidelleur et al. 1997; Harrison et al. 2000) suggest that these three tectonic structures where active during the same time interval. † Methodologically, the present study showed the important role of the lithological constraints on the development of metamorphic minerals. The growth of white mica and garnet was hindered by high organic content and low permeability in the metapelites, while the metaarenitic lithologies always contain much coarser and well-developed metamorphic minerals. Consequently, textures and sometimes even the paragenesis of meta-arenites and meta-tuffs indicate higher metamorphic conditions than the adjacent metapelites. ´ rkai who is not only a living index but also an To Dr Pe´ter A always helpful mentor. The authors are grateful for the aid during fieldwork to our Tibetan drivers Puchum, Tawa, Nobu (Lhasa), and to the Chinese students Xu Xiaoxia, Xu Qiang and Zhang Qinghai (Beijing) who partly joined the field work. Many thanks for careful sample preparation to U. Grunewald and I. Ottenbacher (Go¨ttingen). The final version of the manuscript was benefited by the helpful comments of two referees. This work was funded by the German Research Foundation (DFG) and is part of the Priority Programme ‘Tibetan Plateau: Formation, Climate, Ecosystems (TiP)’.

References Aikman, A., Harrison, T. M. & Lin, D. 2004. Preliminary results from the Yala-Xiangbo leucogranite dome, SE Tibet. Himalayan Journal of Sciences, 2, 91. Aikman, A., Harrison, T. M. & Lin, D. 2008. Evidence for early (.44 Ma) Himalayan crustal thickening, Tethyan Himalaya, southeastern Tibet. Earth and Planetary Science Letters, 274, 14–23. Antolin, B., Apell, E., Montomoli, C., Dunkl, I., Ding, L., Gloaguen, R. & El Bay, R. 2011. Kinematic evolution of the eastern Tethyan Himalaya: Constraints from magnetic fabric and structural properties of the Triassic flysch in SE Tibet. In: Poblet, J. & Lisle, R. (eds) Kinematic Evolution and Structural Styles of Fold-and-Thrust Belts. Geological Society, London, Special Publications, 349, 99– 121. ´ rkai, P. 1991. Chlorite crystallinity: an empirical A approach and correlation with illite crystallinity, coal rank and mineral facies as exemplified by Palaeozoic

and Mesozoic rocks of northeast Hungary. Journal of Metamorphic Geology, 9, 723–734. ´ rkai, P., Faryad, S. W., Vidal, O. & Balogh, K. A 2003. Very low-grade metamorphism of sedimentary rocks of the Meliata unit, Western Carpathians, Slovakia: implications of phyllosilicate characteristics. International Journal of Earth Sciences, 92, 68– 85. ´ rkai, P., Sassi, F. P. & Desmons, J. 2007. Very low- to A low-grade metamorphic rocks. In: Fettes, D. & Desmons, J. (eds) Metamorphic Rocks: A Classification and Glossary of Terms: Recommendations of the International Union of Geological Sciences Subcommission on the Systematics of Metamorphic Rocks. Cambridge University Press, Cambridge, UK, 36–42. Armijo, R., Tapponnier, P., Mercier, J. L. & Tonglin, H. 1986. Quaternary extension in southern Tibet: field observations and tectonic implications. Journal of Geophysical Research, 91, 13803–13872. Armstrong, J. T. 1991. Quantitative elemental analysis of individual microparticles with electron beam instruments. In: Heinrich, K. F. J. & Newbury, D. E. (eds) Electron Probe Quantification. Plenum, New York, 261–315. Barker, C. E. & Pawlewitcz, M. J. 1986. The correlation of vitrinite reflectance with maximum temperature in humic organic matter. In: Buntebarth, G. & Stegena, L. (eds) Lecture Notes in Earth Sciences, Paleogeothermics. Springer Verlag, 5, 79–93. Bostick, N. H. 1979. Microscopic measurement of the level of catagenesis of solid organic matter in sedimentary rocks to aid exploration for petroleum and to determine former burial temperatures – A review. In: Scholle, P. A. & Schluger, P. R. (eds) Aspects of Diagenesis, Society of Economic Paleontologists and Mineralogists Special Publication. Tulsa, Oklahoma, 26, 17–44. Brookfield, M. 1993. The Himalayan passive margin from Precambrian to Cretaceous. Sedimentary Geology, 84, 1 –35. Burg, J. P. & Chen, G. M. 1984. Tectonics and structural zonation of southern Tibet, China. Nature, 311, 219–223. Carosi, R., Montomoli, C. & Visona`, D. 2007. A structural transect in the lower Dolpo: insights on the tectonic evolution of Western Nepal. Journal of Asian Earth Sciences, 29, 407–423, doi: 10.1016/ j.jseaes.2006.05.001. Carosi, R., Montomoli, C., Rubatto, D. & Visona`, D. 2010. Late Oligocene high-temperature shear zones in the core of the higher Himalayan crystallines (Lower Dolpo, Western Nepal). Tectonics, doi: 10.1029/ 2008TC002400. Catlos, E. J., Harrison, T. M., Manning, C. E., Grove, M., Rai, S. M., Hubbard, M. S. & Upreti, B. N. 2002. Records and evolution of the Himalayan orogen from in situ Th–Pb ion microprobe dating of monazite: eastern Nepal and western Garhwal. Journal of Asian Earth Science, 20, 459 –479. Catlos, E. J., Dubey, C. S., Marston, R. A. & Harrison, M. T. 2007. Geochronological constraints across the main central thrust shear zone, Bhagirathi River (NW India): implications for Himalayan tectonics. Geological Society of America Special Paper, 419, 135 –150.

METAMORPHIC EVOLUTION IN SE TIBET Clauer, N. & Chaudhuri, S. 1999. Isotopic dating of very low-grade metasedimentary and metavolcanic rocks: techniques and methods. In: Frey, M. & Robinson, D. (eds) Low-grade Metamorphism. Blackwell, Oxford, 202– 226. Connolly, J. A. D. 2005. Computation of phase equilibria by linear programming: a tool for geodynamic modeling and its application to subduction zone decarbonation. Earth and Planetary Science Letters, 236, 524–541. Connolly, J. A. D. & Petrini, K. 2002. An automated strategy for calculation of phase diagram sections and retrieval of rock properties as a function of physical conditions. Journal of Metamorphic Geology, 20, 697–708. Crouzet, C., Dunkl, I., Paudel, L., Arkai, P., Rainer, T. M., Balogh, K. & Appel, E. 2007. Temperature and age constraints on the metamorphism of the Tethyan Himalaya in Central Nepal: a multidisciplinary approach. Journal of Asian Earth Sciences, 30, 113–130, doi: 10.1016/j.jseaes.2006.07.014. DeCelles, P. G., Robinson, D. M., Quade, J., Ojha, T. P., Garzione, C. N. & Copeland, P. 2001. Stratigraphy, structure and tectonic evolution of the Himalayan fold-thrust belt in western Nepal. Tectonics, 20, 487–509. Ding, L., Kapp, P. & Wan, X. 2005. Paleocene–Eocene record of ophiolite obduction and initial India– Asia collision, south central Tibet. Tectonics, 24, TC3001, doi: 10.1029/2004TC001729. Dupuis, C., He´bert, R., Dubois-Cote´, V., Wang, C. S., Li, Y. L. & Li, Z. J. 2005. Petrology and geochemistry of mafic rocks from me´lange and flysch units adjacent to the Yarlung Zangbo suture zone, southern Tibet. Chemical Geology, 214, 287–308. Dupuis, C., He´bert, R., Dubois-Cote´, V., Guilmette, C., Wang, C. S. & Li, Z. J. 2006. Geochemistry of sedimentary rocks from me´lange and flysch units south of the Yarlung Zangbo suture zone, southern Tibet. Journal of Asian Earth Sciences, 26, 489–508. Edwards, M. A. & Harrison, T. M. 1997. When did the roof collapse? Late Miocene north– south extension in the high Himalaya revealed by Th–Pb monazite dating of the Khula Kangri granite. Geology, 25, 543–546. Farley, K. A., Wolf, R. A. & Silver, L. T. 1996. The effects oflong alpha-stopping distance on (U– Th)/He ages. Geochimica et Cosmochimica Acta, 60, 4223–4229. ¨ sterreichische Fuchs, G. 1967. Zum Bau des Himalaya. O Akademie der Wissenschaften, MathematischNaturwissenschafliche Klasse, Denkschriften, 113, 1–211. Fuhrmann, U., Lippolt, H. J. & Hess, J. C. 1987. Examination of some proposed K–Ar standards: 40Ar/39Ar analyses and conventional K –Ar-Data. Chemical Geology, 66, 41–51. Gaetani, M. & Garzanti, E. 1991. Multicyclic history of the northern Indian continental margin (northwestern Himalaya). American Association of Petroleum Geologists, Bulletin, 75, 1427– 1446. Garzanti, E. 1999. Stratigraphy and sedimentary history of the Nepal Tethys Himalaya passive margin. Journal of Asian Earth Sciences, 17, 805– 827.

67

Garzione, C. N., DeCelles, P. G., Hodkinson, D. G., Ojha, T. P. & Upreti, B. N. 2003. East–west extension and Miocene environmental change in the southern Tibetan plateau: Thakkhola graben, central Nepal. Geological Society of America Bulletin, 115, 3 –20. Godin, L. 2003. Structural evolution of the Tethyan sedimentary sequence in the Annapurna area, central Nepal Himalaya. Journal of Asian Earth Sciences, 22, 307 –328. Godin, L., Grujic, D., Law, R. D. & Searle, M. P. 2006. Channel flow, extrusion and exhumation in continental collision zones: an introduction. In: Law, R. D., Searle, M. P. & Godin, L. (eds) Channel Flow, Ductile Extrusion and Exhumation in Continental Collision Zones. Geological Society, London, Special Publications, 268, 1– 23. Grujic, D., Lincoln, S., Hollister, L. S. & Parrish, R. R. 2002. Himalayan metamorphic sequence as an orogenic channel: insight from Bhutan. Earth and Planetary Science Letters, 198, 177–191. Guillot, S., Peˆcher, A., Rochette, P. & LeFort, P. 1993. The emplacement of the Manaslu granite of central Nepal: field and magnetic susceptibility constraints. In: Treloar, P. J. & Searle, M. P. (eds) Himalayan Tectonics. Geological Society, London, Special Publications, 74, 413– 428. Guillot, S., Garzanti, G., Baratoux, D., Marquer, D., Maheo, G. & de Sigoyer, J. 2003. Reconstructing the total shortening history of the NW Himalaya. Geochemistry, Geophysics, Geosystems, 4, 1064 doi: 10.1029/2002GC000484. Harrison, T. M., Ryerson, F. J., Le Fort, P., Yin, A., Lovera, O. M. & Catlos, E. J. 1997. A late Miocene–Pliocene origin for the Central Himalayan inverted metamorphism. Earth and Planetary Science Letters, 146, E1–E8. Harrison, T. M., Yin, A., Grove, M. & Lovera, O. M. 2000. The Zedong Window: a record of superposed Tertiary convergence in southeastern Tibet. Journal of Geophysical Research, 105, 19,211–19,320. Heim, A. & Gansser, A. 1939. Central Himalaya. Geological observations of the Swiss expedition 1936. Me´moires de la Socie´te´ Helve´tique des Sciences Naturelles, 7/31, 1 –245. Heinrichs, H. & Herrmann, A. G. 1990. Praktikum der Analytischen Geochemie. Springer Verlag, Berlin. Hodges, K. V., Parrish, R. R. & Searle, M. P. 1996. Tectonic evolution of the central Annapurna Range, Nepalese Himalayas. Tectonics, 15, 1264– 1291. Hodges, K. V. 2000. Tectonics of the Himalaya and southern Tibet from two perspectives. Geological Society of America Bulletin, 112, 324–350. Holland, T. J. B. & Powell, R. 1998. An internallyconsistent thermodynamic dataset for phases of petrological interest. Journal of Metamorphic Geology, 16, 309– 344. Hower, J., Hurley, P. M., Pinson, W. H. & Fairbairn, H. W. 1963. The dependence of K– Ar age on the mineralogy of various particle size ranges in a shale. Geochimica et Cosmochimica Acta, 27, 405–410. Hurley, P. M., Hunt, J. M., Pinson, W. H. & Fairbairn, H. W. 1963. K– Ar age values on the clay fractions in dated shales. Geochimica et Cosmochimica Acta, 27, 279 –284.

68

I. DUNKL ET AL.

Jadoul, F., Berra, F. & Garzanti, E. 1998. The Tethys Himalayan passive margin from Late Triassic to Early Cretaceous (South Tibet). Journal of Asian Earth Sciences, 16, 173–198. ´ rkai, P. & Judik, K., Rantitsch, G., Rainer, Th. M., A Tomljenovic´, B. 2008. Organic metamorphism in metasedimentary rocks from Mt. Medvednica (Croatia). Swiss Journal of Geosciences, 101, 605–616. Kellett, D. A & Godin, L. 2009. Pre-Miocene deformation of the Himalayan superstructure, Hidden valley, central Nepal. Journal of the Geological Society, London, 166, 261 –275. ´ rkai, P. & Brime, C. 2004. On the caliKisch, H. J., A bration of the illite Ku¨bler index (illite ‘crystallinity’). Schweizerische Mineralogische und Petrographische Mitteilungen, 84, 323–331. Kohn, M. J. & Spear, F. S. 1991. Error propagation for barometers: 1. Accuracy and precision of experimentally located end-member reactions. American Mineralogist, 76, 128– 137. Ku¨bler, B. 1967. La cristallinite´ de l’illite et les zones tout a` fait supe´rieures du me´tamorphisme. E´tages Tectoniques, Colloque de Neuchaˆtel 1966. Edition de la Baconnier`e, Neuchaˆtel, 105–122. Ku¨bler, B. 1968. Evaluation quantitative du metamorphisme par la cristallinite´ de l’illite. Bulletin du Centre de Recherches, Pau SNPA, 2, 385– 397. Ku¨bler, B. 1990. ‘Cristallinite´’ de l’illite et mixed-layers: bre`ve re´vision. Schweizerische Mineralogische und Petrographische Mitteilungen, 70, 89– 93. Le Fort, P. 1971. Les formations cristallophyliennes de la Thakkhola. Recherches ge´ologiques dans l’Himalaya du Ne´pal, re´gion del Thakkhola. Edition du CNRS, Paris. Le Fort, P. 1975. Himalayas, the collided range. Present knowledge of the continental arc. American Journal of Science, 275-A, 1 –44. Lee, J., Hacker, B. R. et al. 2000. Evolution of the Kangmar Dome, southern Tibet: structural, petrologic, and thermochronologic constraints. Tectonics, 19, 872– 895. Leech, M. L. 2008. Does the Karakoram fault interrupt mid-crustal channel flow in the western Himalaya? Earth and Planetary Science Letters, 276, 314–322. Liu, G. & Einsele, G. 1994. Sedimentary history of the Tethyan basin in the Tibetan Himalayas. Geologische Rundschau, 83, 32–61. Livi, K. J. T., Veblen, D. R., Ferry, J. M. & Frey, M. 1997. Evolution of 2:1 layered silicates in low-grade metamorphosed Liassic shales of Central Switzerland. Journal of Metamorphic Geology, 15, 323–344. McQuarrie, N., Robinson, D., Long, S., Tobgay, T., Grujic, D., Gehrels, G. & Ducea, M. 2008. Preliminary stratigraphic and structural architecture of Bhutan: implications for the along strike architecture of the Himalayan system. Earth and Planetary Science Letters, 272, 105–117. Mercier, J. L. et al. 1984. La collision Inde-Asie coˆte´ Tibet. In: Mercier, J. L. & Li, G. C. (eds) Mission Franco-Chinoise Au Tibet 1980: E´tude Ge´ologique Et Ge´ophysique De La Crouˆte Terrestre Et Du Manteau Supe´rieur Du Tibet Et De L´Himalaya. Editions du Centre National de la Recherche Scientifique, Paris, France, 29, 341–350.

Meunier, A. & Velde, B. 2004. Illite: Origins, Evolution and Metamorphism. Springer, 286. Montomoli, C., Appel, E., Antolin, B., Dunkl, I., El Bay, R., Lin, D. & Gloaguen, R. 2008. Polyphase deformation history of the ‘Tibetan sedimentary sequence’ in the Himalaya chain (South-East Tibet). Himalayan Journal of Sciences, 5, 91. Najman, Y., Carter, A., Oliver, G. & Garzanti, E. 2005. Provenance of Eocene foreland basin sediments, Nepal: constraints to the timing and diachroneity of early Himalayan orogenesis. Geology, 33, 309– 312. Pan, G., Ding, J., Yao, D. & Wang, L. 2004. Geological map of Qinghai-Xizang (Tibet) Plateau and Adjacent Areas (1:1,500,000). Chengdu Institute of Geology and Mineral Resources, China Geological Survey. Chengdu Cartographic Publishing House. Patzelt, A., Li, H., Wang, J. & Appel, E. 1996. Paleomagnetism of Cretaceous to Tertiary sediments from southern Tibet: evidence for the extent of the northern margin of India prior to the collision with Eurasia. Tectonophysics, 259, 259– 284. Paudel, L. P. & Arita, K. 2006. Thermal evolution of the Lesser Himalaya, central Nepal: insights from K-white micas compositional variation. Gondwana Research, 9, 409–425. Peˆcher, A. 1975. The Main Central Thrust of the Nepal Himalaya and the related metamorphism in the Modi Khola cross-section (Annapurna Range). Himalayan Geology, 5, 115– 131. Potel, S. 2007. Very low-grade metamorphic study in the pre-Late Cretaceous terranes of New Caledonia (southwest Pacific Ocean). Island Arc, 16, 291– 305. Powell, R. & Holland, T. J. B. 1994. Optimal geothermometry and geobarometry. American Mineralogist, 79, 120– 130. Powell, R. & Holland, T. J. B. 2006. Course Notes for ‘THERMOCALC Short Course’. Sao Paulo, Brazil, on CD-ROM. Powell, R. & Holland, T. J. B. 2008. On thermobarometry. Journal of Metamorphic Geology, 26, 155 –179. Quidelleur, X., Grove, M., Lovera, O. M., Harrison, T. M. & Yin, A. 1997. Thermal evolution and slip history of the Renbu-Zedong thrust, southeastern Tibet. Journal of Geophysical Research, 102, 2659– 2679. Ratschbacher, L., Frisch, W., Liu, G. & Chen, C. 1994. Distributed deformation in southern and western Tibet during and after the India– Asia collision. Journal of Geophysical Research, 99, 19,917–19,945. Reuter, A. 1987. Implications of K–Ar ages of wholerock and grain-size fractions of metapelites and intercalated metatuffs within an anchizonal terrane. Contribution to Mineralogy and Petrology, 97, 105–115. Robinson, D. M., DeCelles, P. G., Garzione, C. N., Pearson, O. N., Harrison, T. M. & Catlos, E. J. 2003. Kinematic model for the Main Central thrust in Nepal. Geology, 31, 339– 362. Schumacher, E. 1975. Herstellung von 99,9997% 38Ar fu¨r die 40K/40Ar Geo-chronologie. Geochronologia Chimia, 24, 441 –442. Searle, M. 1986. Structural evolution and sequence of thrusting in the High Himalayan, Tibetan-Tethys and Indus suture zones of Zanskar and Ladakh, Western Himalaya. Journal of Structural Geology, 8, 923– 936.

METAMORPHIC EVOLUTION IN SE TIBET Searle, M. P. & Godin, L. 2003. The south Tibetan detachment and the Manaslu leucogranite: a structural reinterpretation and restoration of the AnnapurnaManaslu Himalaya, Nepal. Journal of Geology, 111, 505–523. Spear, F. S. 1995. Metamorphic Phase Equilibria and Pressure– Temperature –Time Paths. Mineralogical Society of America, Washington DC, 1– 799. Steiger, R. H. & Ja¨ger, E. 1977. Subcommission on geochronology: convention on the use of decay constants in geo- and cosmochronology. Earth and Planetary Science Letters, 36, 359–362. Sto¨cklin, J. 1980. Geology of Nepal and its regional frame. Journal of the Geological Society, London, 137, 1 –34. Sweeney, J. J. & Burnham, A. K. 1990. Evaluation of a simple model of vitrinite reflectance based on chemical kinetics. Bulletin American Association of Petroleum Geologists, 74, 1559– 1570. Upreti, B. N. 2001. Stratigraphy, structure, and tectonic evolution of the Himalayan fold-thrust belt in western Nepal. Tectonics 20, 487 –509. Valdiya, K. S. 1980. Geology of the Kumaun Lesser Himalaya. Wadia Institute of Himalayan Geology, Dehra Dun. Warr, L. N. & Rice, A. H. N. 1994. Interlaboratory standardisation and calibration of clay mineral crystallinity and crystallite size data. Journal of Metamorphic Geology, 12, 141– 152. Wemmer, K. 1991. K/Ar-Altersdatierungsmo¨glichkeiten fu¨r retrograde Deformationsprozesse im spro¨den und

69

duktilen Bereich – Beispiele aus der KTB-Vorbohrung (Oberpfalz) und dem Bereich der Insubrischen Linie (N-Italien). Go¨ttinger Arbeiten zur Geologie und Pala¨ontologie, 51, 1 –61. Willems, H., Zhou, Z., Zhang, B. & Gra¨fe, K. U. 1996. Stratigraphy of the Upper Cretaceous and lower Tertiary strata in the Tethyan Himalayas of Tibet (Tingri area, China). Geologische Rundschau, 85, 723– 754. Wojdyr, M. 2007. FITYK: A curve fitting and data analysis program. World Wide Web address: http://www. unipress.waw.pl/fityk/ Xu, X., Ding, L., Qiang, X., Fulong, C., Qinghai, Z., Liyun, Z. & Qingzhou, L. 2009. Tectonic implications of the Ultramafic Dykes in Southeastern Tibet. Chinese Journal of Geology, 44, 1012–1024. Yin, A. 2006. Cenozoic tectonic evolution of the Himalayan orogen as constrained by along– strike variation of structural geometry, exhumation history, and foreland sedimentation. Earth-Science Reviews, 76, 1– 131. Yin, A., Harrison, T. M., Ryerson, F. J., Wenji, C., Kidd, W. & Copeland, P. 1994. Tertiary structural evolution of the Gangdese thrust system, southeastern Tibet. Journal of Geophysical Research, 99, 18, 175– 18, 201. Zhu, D., Mo, X. et al. 2008. Petrogenesis of the earliest Early Cretaceous mafic rocks from the Cona area of the eastern Tethyan Himalaya in south Tibet: interaction between the incubating Kerguelen plume and the eastern Greater India lithosphere? Lithos, 100, 147– 173.

Indication for clockwise rotation in the Siang window south of the eastern Himalayan syntaxis and new geochronological constraints for the area URSINA LIEBKE1*, B. ANTOLIN1, E. APPEL1, N. BASAVAIAH2, T. MIKES3,4, I. DUNKL3 & K. WEMMER3 1

Department of Geosciences, University of Tu¨bingen, Sigwartstrasse 10, D-72076 Tu¨bingen, Germany 2

Indian Institute of Geomagnetism, Kalamboli Highway, New Panvel, Navi Mumbai 410218, India

3

Geoscience Centre Go¨ttingen, Goldschmidtstrasse 3, D-37077 Go¨ttingen, Germany

4

Present address: Institute of Geosciences, Goethe University Frankfurt, Altenho¨ferallee 1, D-60438 Frankfurt am Main, Germany *Corresponding author (e-mail: [email protected]) Abstract: Palaeomagnetic, rock magnetic and geochronological investigations were carried out on the Abor volcanics of Arunachal Pradesh, NE India. A Late Palaeozoic formation age for part of the Abor volcanics cannot be excluded based on K– Ar whole rock dating. Low-temperature thermochronometers – zircon (U–Th)/He and fission track analyses – yield a maximum burial temperature of c. 150–170 8C during Late Miocene. ZFT thermochronology of the Yinkiong and Miri Fms. indicates a post-Paleocene and post-Jurassic deposition age, respectively. This infers that the volcanic rocks intercalating or intruding them are not part of the Late Palaeozoic sequence but represent one or more, latest Cretaceous to Tertiary event(s). Therefore the Abor volcanics are connected to at least two separate events of volcanism. From palaeomagnetic sites, two characteristic magnetic remanence components were separated: a low-coercivity-component demagnetized below 20 mT and a high-coercivity-component demagnetized between 15 and 100 mT. Fold tests support a secondary origin of both components. Thermochronological and rock magnetic analyses indicate a low-grade overprint event between India–Asia collision and Miocene, which probably represents the time of remanence acquisition. The high-coercivitycomponent shows a trend of clockwise declinations, which is likely related to vertical-axis rotations of the eastern Himalayas due to eastward extrusion of the Tibetan Plateau.

The Himalayan mountain range, created by collision of the Indian and Eurasian continents during Early Eocene, provides a unique opportunity for studying processes of continent–continent collision (Yin & Harrison 2000). The Himalaya extends between the Nanga Parbat Syntaxis at the western end and the Namche Barwa Syntaxis at the eastern end. The area around the Namche Barwa Syntaxis plays a key role in understanding geodynamic processes related to the eastward extrusion of the Tibetan Plateau. Ongoing indentation of India into Eurasia has led to considerable tectonic deformation of the Eurasian continent (Molnar & Tapponnier 1975). Clockwise tectonic rotations in the eastern part of the Himalayas are predicted by several numerical models and laboratory experiments in which the rigid Indian plate penetrates into the continuum of Asia (e.g. Houseman & England 1993;

Beaumont et al. 2001; Cook & Royden 2008). Palaeomagnetic investigations on Cretaceous – Miocene rocks in the wider region north and west of the Namche Barwa Syntaxis supported these clockwise rotations (e.g. Otofuji et al. 1990, 1998; Funahara et al. 1992, 1993; Huang & Opdyke 1993; Sato et al. 1999, 2007; Yang et al. 2001). In addition, velocity vectors obtained from global positioning system (GPS) clearly display a clockwise rotational pattern around the Namche Barwa Syntaxis for the present aseismic surface movements (e.g. Chen et al. 2000; Mattauer 2002; Zhang et al. 2004; Shen et al. 2005; Sol et al. 2007). Most of the previous studies were performed north or east of the Namche Barwa Syntaxis. Due to political and infrastructural restrictions and very dense vegetation, the area south of the syntaxis is relatively unexplored. The present study provides new

From: Gloaguen, R. & Ratschbacher, L. (eds) Growth and Collapse of the Tibetan Plateau. Geological Society, London, Special Publications, 353, 71– 97. DOI: 10.1144/SP353.5 0305-8719/11/$15.00 # The Geological Society of London 2011.

72

U. LIEBKE ET AL.

palaeomagnetic and geochronological data on the Abor volcanic and adjacent sedimentary rocks of the Siang window in Arunachal Pradesh, NE India, located just south of the Namche Barwa Syntaxis. The age of the Abor volcanic rocks is highly controversial (e.g. Gansser 1964; Jain & Thakur 1978; Tripathi & Chowdhury 1983; Bhat 1984; Acharyya 1994, 2007; Kumar 1997). We present new geochronological and palaeomagnetic results that offer insight in the regional structure of the eastern Himalayan syntaxis area and the extrusion of the Tibetan Plateau.

Geological setting The Arunachal Himalaya covers c. 350 km along strike of the eastern Himalayan fold-thrust Belt (Fig. 1) and extends from the eastern border of

Bhutan to the Dibang and Lohit valleys at the eastern border of India. Many controversies exist about the classification, the distribution, and the age of rock units. Three of the main tectonostratigraphic units of the Himalayan orogen can be observed within the Arunachal Himalaya (Fig. 1b, c; Yin 2006). From north to south they are: the Greater Himalayan Crystalline Complex between the South Tibetan Detachment System (STDS) and the Main Central Thrust (MCT), the Lesser Himalayan sequence between the MCT and the Main Boundary Thrust (MBT), and the Siwaliks in the footwall of the MBT (Fig. 1; see Hodges et al. 2000; Yin 2006 for a review). Figure 2a shows the main stratigraphic units of the eastern Arunachal Himalaya. In the eastern Himalaya, the MCT is broadly folded (e.g. Gansser 1983). Dating of monazite inclusions in garnet from the MCT zone in western Arunachal indicates

Fig. 1. The study area (green box) at three different scales: (a) Sketch of India and surrounding regions; (b) Simplified geological map of the Himalaya (modified after Ding et al. 2001; Pan et al. 2004; Yin et al. 2006). (c) Relief map (http://srtm.csi.cgiar.org/) around the Siang window with approximate trace of the main structural elements from Ding et al. (2001) and Acharyya (2007). MBT, Main Boundary Thrust; MCT, Main Central Thrust; STDS, South Tibetan Detachment System; GCT, Great Counter Thrust; LHS, Lesser Himalayan Sequence; GHCC, Greater Himalayan Crystalline complex; rectangle shows study area.

THE SIANG WINDOW OF ARUNACHAL PRADESH

73

that the MCT was active c. 10 Ma ago (Yin et al. 2006). The Se La Group between the STDS and the MCT consists of high grade gneisses and schists of the Greater Himalayan Crystalline Complex. The STDS is well recognized further east in the Siang river section (Acharyya & Saha 2008). The Lesser Himalaya is composed of the Late Proterozoic Bomdila Group and the Late Palaeozoic Lower Gondwana Group. In the eastern part of Arunachal Pradesh the Lesser Himalaya is made up of mostly Phanerozoic rocks (Kumar 1997). The Neogene – Early Quaternary molasse-type sediments in the footwall of the Main Boundary Thrust (MBT) belong to the Siwaliks (Kumar 1997).

clinopyroxene phenocrysts. Formation of chlorite and epidote locally in the glassy matrix indicates a weak post-emplacement hydrothermal and/or diagenetic transformation of the volcanic rocks (Fig. 3). Locally, the rock units within the window are dissected by several subsidiary normal faults oblique to both, the MBT and the North Pasighat Thrust (NPT) (Kumar 1997; Acharyya 2007). They are bound by the Bomdila Group to the north, east and west. To the south, the rocks units of the Siang window are truncated against Neogene Siwaliks sediments across the NPT (Fig. 2; Kumar 1997; Acharyya 2007).

The Siang window: tectonostratigraphy

Previous age constraints of the Abor volcanic rocks

The present study focuses on the Abor volcanic rocks which crop out at the core of the Siang window. The Siang window encompasses c. 8000 km2 of the eastern part of the Arunachal Himalaya (Figs 1a, c & 2a). According to several authors (e.g. Sengupta et al. 1996; Gururajan & Chowdhury 2003; Acharyya 2007), the Siang window comprises antiformally folded and up-arched thrust sheets, which are made of Lesser Himalayan rocks (Fig. 2b). It has been proposed that the window was evolved during the process of southward exhumation of the Greater Himalayan Crystalline Complex and southvergent thrusting over the Tertiary Himalayan foreland basin (Acharyya 1998; Acharyya & Sengupta 1998). The antiformal duplex structure, which breached the MBT and passively folded overlying Himalayan nappes, was possibly formed by compressive tectonics at the Eastern Syntaxis (Acharyya 2005). Two sedimentary units are exposed at the core of the Siang window: the upper Yinkiong Fm and a lower unit (Miri Fm) composed of mainly white quartzites with occasionally coarser bands containing jasper. The Yinkiong Fm is composed of an alternating sequence of dark grey sandstone to siltstone and green and red shale with rare orthoquartzites and volcanic rocks (Jain & Tandon 1974). Its foraminiferal assemblage indicates an Early to Mid Eocene age (Acharyya 1994). The quartzites of the lower unit were lithostratigraphically correlated with the Middle Palaeozoic to Lower Permian Miri Quartzite exposed in adjacent areas (Jain & Thakur 1978; Bhat 1984). Acharyya (1994) reported a foraminiferal assemblage from limestone bands in the upper part of the quartzites and proposed a Palaeogene age for at least this upper part of the Miri Fm. The Abor volcanic rocks are exposed as lava flows, sills, dykes, volcanic breccias, and metatuffs (Jain & Thakur 1978) and comprise hard, massive, or amygdaloidal basaltic to andesitic rocks. Thin section petrography of our samples reveals rather fresh feldspar and

Age and evolution of the Abor volcanic rocks have been a matter of great controversy. Based on lithological and geochemical similarities, many authors correlated them to the Permo-Triassic Panjal Trap volcanism (e.g. Gansser 1964; Bhat 1984; Bhat & Ahmad 1990; Kumar 1997), associated with the break-up of Gondwana. Jain & Thakur (1978); however, assumed a Precambrian to Middle Palaeozoic age for the volcanic rocks. Their arguments were based on: (a) the NNW– SSE structural trend of the lithological units within the Siang district, which is thought to be connected to the Mid Palaeozoic orogeny in this part of the Himalayas; (b) the presence of quartzite, slate, basalt, and other rock fragments in diamictites of the Gondwana belt, interpreted to have been derived from the MiriSiang Group and the Abor volcanic rocks; and (c) agglomeratic volcanic rocks and tuffs intercalating pebbly mudstones assigned to the Gondwana belt in the NE Himalaya. These arguments have been challenged early on by Chowdhury (1979) on basis of biostratigraphical and structural data. Gansser (1974) and Le Fort (1975) have also cast doubt on a Mid Palaeozoic orogeny in this area. Tripathi et al. (1979, 1981b) were the first to report Early Eocene foraminifera and plant imprints closely associated with the Abor volcanic rocks and from the Yinkiong Fm. This was supported by the discovery of Upper Paleocene to Lower Eocene nummulitic limestones in the Yinkiong Fm (Tripathi et al. 1981a). In other sections, marine molluscs and sporomorph assemblages of Gondwana affinity document a Permian age for the sedimentary rocks associated with the volcanic formations (e.g. Singh 1981; Tripathi & Chowdhury 1983; Sinha et al. 1986; Prasad et al. 1989). It has therefore been assumed that two different sets of volcano-sedimentary sequences occur in the Siang window, an Early Tertiary and a Permian one (e.g. Singh 1984; Singh & De 1984; Singh & Malhotra

74

Fig. 2.

U. LIEBKE ET AL.

THE SIANG WINDOW OF ARUNACHAL PRADESH

75

Fig. 3. Photomicrographs of thin sections. (a) glassy groundmass of sample AbV-24; (b) chloritization of a clinopyroxene in sample AbV-1.

1987). Tripathi et al. (1988) also postulated two different units of the Abor volcanic rocks: an older Permian unit exposed around Rotung and a younger Eocene unit exposed around Geku. However, mapping revealed that these units join together (Singh 1993). A slate boulder with Permian marine bivalves was reported close to the Dalbuing area, whereas Lower –Mid Eocene foraminifera were recorded from the Yinkiong Fm (Sinha et al. 1986; Singh 1993). According to Acharyya (1994, 2007) the deformed slate with Permian fossils possibly occurs as a thrust sliver close to the MBT and NE of Dalbuing. Acharyya (1994) dated calcareous quartzites and limestone bands directly underlying the Abor volcanic rocks to be Late Paleocene –Early Eocene in age by larger foraminifera. According to him, the Abor volcanic rocks were erupted in-between the deposition of the Late Paleocene– Early Eocene quartzites and the Yinkiong Fm and therefore range in age from Upper Paleocene –Early Eocene. If this scenario is valid, volcanism would be fairly contemporaneous with the collision of the Indian and Asian plates during Early Eocene probably triggered by

either adiabatic decompression following crustal thickening (Sengupta et al. 1996) or the thermal anomaly related to slab break-off following the collision (Acharyya 2007).

Sample treatment and analytical procedures Core samples from 35 sites were drilled with a portable rockdrill. Orientation was generally recorded by a magnetic compass; no sun compass was required because of the relatively low remanence intensity. The core samples were cut into standard specimens of 2.5 cm diameter and 2.2–2.3 cm length. Palaeo- and rock magnetic measurements were carried out at Tu¨bingen University; geo- and thermochronological analyses were performed at Go¨ttingen University. The natural remanent magnetization (NRM) of two pilot samples (twin samples) was demagnetized by alternating field and thermal treatment. Remanence directions were measured using a 2 G Enterprises SQUID magnetometer 755R. Alternating

Fig. 2. (a) Simplified map of the Siang window (modified after Acharyya & Saha 2008). 1, Upper and Middle Siwaliks; 2, Lower Siwaliks; 3, Rocks exposed at the core of the Siang window (undifferentiated); 4, Miri Quarzite; 5, Low-medium grade meta-argillite (Proterozoic); 6, Late Palaeozoic metasediments of the Lower Gondwana Group; 7, High grade rocks and gneisses of the Greater Himalayan Crystalline Complex (Se La Group); 8, Bombdila Group (low-grade quartzite-dolomite metasediments of the Buxa Fm); 9, Yang Sang Chu Fm (Proterozoic); 10, Trans-Himalayan granitoids and gneisses; 11, Mafic and ultramafic rocks (ophiolites). Structural features: 12, Major thrust; 13, Fault; 14, Axial trace of synclinal fold; 15, Axial trace of anticlinal fold; 16, Strike–slip fault. Abbreviations: MCT, Main Central Thrust; MBT, Main Boundary Thrust; NPT, North Pasighat Thrust; STDS, South Tibetan Detachment System. The Dibang and Lohit Valleys are located off the map further east. (b) Geological map of the core of the Siang window showing palaeomagnetic site locations of this study and position of geochronology studied site Db (‘drift boulder’). The map was modified after Acharyya & Saha (2008); structural data are taken from Singh (1993). The legend is the same as in Figure 2a; rocks exposed at the core of the Siang window are differentiated to: 3a, Yinkiong Fm; 3b, Abor volcanic rocks; 3c, Quartzites. Note that all palaeomagnetic sites were sampled from the Abor volcanic rocks; the location of some sites within different stratigraphic units is either due to imprecise drawing of the map or imprecise GPS data. (c) Structural cross section along AB (modified after Acharyya 2007).

76

U. LIEBKE ET AL.

field demagnetization (AfD) was performed using an automatic 3-axes degaussing system integrated in the SQUID magnetometer; 15 steps with a maximum applied field of 100 mT were applied. For thermal demagnetization (ThD) 13 steps with 25–100 8C temperature increments and a maximum temperature of 700 8C were performed. Heating was done in an ASC scientific furnace (model TD-485 C). In order to detect possible changes in the magnetic mineralogy, the bulk magnetic susceptibility was measured after each step of heating using a Kappabridge KLY-2 (Agico). The demagnetization results indicate that both AfD and ThD are suitable to separate remanence components. As AfD is more convenient, the natural remanent magnetization (NRM) of eight to nine specimens per site was progressively demagnetized by alternating field cleaning. In addition, ThD of NRM was applied to eight specimens of five sites in order to relate magnetic components to specific ferro(i)magnetic minerals. To determine the magnetic mineralogy one specimen of 19 representative sites was applied to stepwise acquisition of isothermal remanent magnetization (IRM) using a pulse magnetizer MMPM9 (Magnetic Measurements Ltd) with a maximum field of 2.5 T. The intensity of IRM was measured with a Minispin spinner magnetometer (Molspin Ltd). Subsequently, the IRM of three orthogonal components (Lowrie 1990) was thermally demagnetized. Susceptibility v. temperature curves from room temperature to maximum 700 8C were measured for 20 specimens using a CS-3 heating device coupled with a Kappabridge KLY-3 (Agico). Anisotropy of magnetic susceptibility (AMS) was measured for three sites using Kappabridge KLY-2. The saturation IRM (SIRM) of eight specimens per site was determined (using the Minispin) after applying a magnetic field of 2.5 T. For zircon thermochronology the samples were crushed, sieved and treated by the common heavy liquid and magnetic separation processes. The zircon crystals were embedded in PFA Teflon and polished in five steps by diamond suspensions and etched by the eutectic melt of NaOH – KOH at a temperature of 220 8C. Neutron irradiations were performed in the research reactor of the Technical University of Munich (Garching). The external detector method was used (Gleadow 1981). After irradiation the induced fission tracks in the mica detectors were revealed by etching in 40% HF for 40 min at 21 8C. Track counting was made with a Zeiss-Axioskop microscope-computer-controlled stage system (Dumitru 1993), with 1000 magnification. The FT ages were determined by the zeta method (Hurford & Green 1983) using age standards listed in Hurford (1998). The error was calculated by using the classical procedure, that is, by

double Poisson dispersion (Green 1981). Calculations and plots were made with the TRACKKEY program (Dunkl 2002). For zircon (U –Th)/He chronology single crystal aliquots were dated; only intact, euhedral crystals with minor inclusions were selected. The shape parameters were measured and archived by multiple microphotographs; the alpha ejection factors (Ft, see in Farley et al. 1996) were determined by the constants of Hourigan et al. (2005). The crystals were wrapped in c. 1  1 mm-sized platinum capsules and degassed in high vacuum by the heating of an infra-red diode laser (Reiners 2005). The extracted gas was purified using a SAES Ti –Zr getter at 450 8C. The chemically inert noble gases and a minor amount of other rest gases were then expanded into a Hiden triple-filter quadrupol mass spectrometer equipped with a positive ion counting detector. Beyond the detection of helium the partial pressures of some rest gases were continuously monitored (H2, CH4, H2O, Ar, CO2). He blanks (c. 0.0003 and 0.0008 ncc 4He) were estimated using the same procedure on empty Pt tubes (cold and hot blanks, respectively). Crystals were checked for degassing of He by sequential reheating and He measurement. Following degassing, samples were retrieved from the gas extraction line, spiked with calibrated 230Th and 233U solutions, and dissolved in pressurized Teflon bombs using distilled 48% HF þ 65% HNO3 in five days at 220 8C (applying a slightly modified procedure of Evans et al. 2005). Spiked solutions were analysed by isotope dilution method using a Perkin Elmer Elan DRC II ICP-MS equipped with an APEX micro-flow nebulizer. Sm, Pt and Zr concentrations were determined by external calibration. The oxide formation rate and the PtAr– U interference were always monitored, but their effects were negligible on the concentration of actinides. For K –Ar dating the samples were carefully cleaned and selected in order to avoid altered parts or veins. The selected pieces were crushed and gently pulverized in a ball mill to reach an appropriate grain size of approx. 100 mm to prevent the loss of Ar during preparation. In case of the shale sample, the ,2 mm and ,0.2 mm size fractions were separated from suspension by gravity settling (Stokes’ Law) and ultra-centrifugation. The samples investigated suffer from a trapped, atmospheric contamination (up to 70% of the 40Ar) which can not be avoided in most volcanic rocks (s. discussion in McDougall & Harrison 1999). Therefore, errors amount up to 4.6% (2s). The argon isotopic composition was measured in a Pyrex glass extraction and purification line coupled to a VG 1200 C noble gas mass spectrometer operating in static mode. The amount of radiogenic 40Ar was determined by isotope dilution method

THE SIANG WINDOW OF ARUNACHAL PRADESH

using a highly enriched 38Ar spike from Schumacher, Bern (Schumacher 1975). The spike is calibrated against the biotite standard HD-B1 (Fuhrmann et al. 1987). The age calculations are based on the constants recommended by the IUGS quoted in Steiger & Ja¨ger (1977). Potassium was determined in duplicate by flame photometry using an Eppendorf Elex 63/61. The samples were dissolved in a mixture of HF and HNO3 according to the technique of Heinrichs & Herrmann (1990). CsCl and LiCl were added as an ionization buffer and internal standard, respectively. The analytical error for the K –Ar age calculations is given on a 95% confidence level (2s). Details of argon and potassium analyses for the laboratory in Go¨ttingen are given in Wemmer (1991).

Results Rock magnetic results Magnetic mineralogy could be deduced from IRM acquisition curves, thermal demagnetization of three-component IRM, and the thermo-magnetic behaviour of susceptibility. Results from representative samples prove magnetite, Ti-rich

77

titanomagnetite, and hematite as the dominant ferro(i)magnetic minerals. High temperature thermomagnetic runs were done for powdered samples and crushed samples. Powdered samples were ground to a fine rock powder, whereas crushed samples consisted of rock particles with diameters of about 2–5 mm. Heating curves of powdered samples reveal magnetite by a decrease in magnetic susceptibility just below its Curie temperature (Fig. 4a). Cooling curves are partly irreversible indicating oxidation of magnetite to hematite during heating; deviating Curie temperature during heating and cooling can be related to temperature hysteresis due to different heat capacities of the sample and the thermoelement. Heating curves of crushed samples show an additionally decrease below 400 8C (Fig. 4b), probably caused by Tirich titanomagnetite. Because Ti-rich titanomagnetite is only meta-stable at room temperature and collapses on application of mechanical stress or heating (O’Reilly 1984), it has been likely destroyed during grinding of the powdered samples. For analysis of IRM acquisition curves, the irmunmix2_2_1 and IRM_CLG1 software (Kruiver et al. 2001) were utilized, which are based on cumulative log-Gaussian analysis (CLG). By means of

Fig. 4. Susceptibility v. temperature curves of abv 29– 8–1; (a) heating and cooling curve of a powdered sample; (b) heating curve of a crushed sample; the noisy character of the heating curve is probably due to relatively high volume differences in the rock particles. Max sus, maximum susceptibility.

78

U. LIEBKE ET AL.

Fig. 5. Results of Cumulative log-Gaussian analysis of abv 19– 9– 1; (a) linear IRM acquisition plot; (b) gradient acquisition plot; (c) standardized acquisition plot.

these programs coercivity parameters of different components are approximated by the applied magnetic field which causes half of the SIRM to be acquired (B1/2). IRM acquisition curves and ThD of three-component IRM show variable results for different samples reflecting the complex rock and palaeomagnetic behaviour. Three different components were distinguished based on their coercivity spectra (Fig. 5, Table 1) and blocking temperatures. A low coercivity component (Clow) was defined by B1/2 values below 70 mT. B1/2 values between c. 70 and 100 mT were allocated to an intermediate coercivity component (Cint) and B1/2 values above 380 mT to a high coercivity component (Chigh). IRM acquisition curves and thermal demagnetization of three-components IRM confirm the presence

of magnetite, Ti-rich titanomagnetite, and hematite (Fig. 6). A decrease of magnetization just below 575 8C indicates magnetite and a more continuous decrease over larger temperature intervals below 400 8C may be related to Ti-rich titanomagnetite. Because the Curie temperature of Ti-rich titanomagnetite increases during low-temperature oxidation, different oxidation stages within Ti-rich titanomagnetite grains broaden their (un)blocking temperature intervals, which results in a more continuous decrease of IRM intensity. Clow can be related to magnetite with some contribution of multidomain titanomagnetite, and Cint to pseudo-single domain or single domain Ti-rich titanomagnetite. A decrease of IRM intensity just below 675 8C in the 2.5 T fraction indicates that Chigh is residing

Table 1. Results of cumulative log-Gaussian analysis of IRM acquisition curves; the relation of values within brackets to Clow or Cint is uncertain Specimen

abv 1-12-1 abv 6-6-1 abv 8-3-2 abv 12-2-1 abv 14-7-1 abv 15-10-1 abv 19-9-1 abv 20-9-1 abv 23-10-1 abv 25-3-1 abv 27-10-2 abv 28-8-1 abv 29-4-1 abv 31-8-1 abv 35-5-2

Clow

Cint

Chigh

S-ratio

%

B1/2 (mT)

%

B1/2 (mT)

%

B1/2 (mT)

– 40 – (100 (45 (79 (70 (100 21 52 (50 (98 13 59 –

– 35 – 43) 69) 54) 42) 55) 56 60 67) 32) 29 55 –

(72 60 (79 – – – – – 79 48 – – 87 41 (91

100) 88 72) – – – – – 102 91 – – 85 96 100)

28 – 21 – 55 21 30 – – – 50 2 – – 9

562 – 384 – 752 730 703 – – – 646 575 – – 631

0.45 0.95 0.69 1 0.08 0.69 0.54 0.99 0.94 0.99 0.26 0.97 0.98 0.87 0.65

THE SIANG WINDOW OF ARUNACHAL PRADESH

Fig. 6. IRM acquisition curves and thermal demagnetization of three-components IRM of (a) abv 19–9 –1 (strong hematite contribution), (b) abv 12–2 –1, and (c) abv 31–8-1 (predominant contribution of magnetite and Ti-rich titanomagnetite).

79

80

U. LIEBKE ET AL.

in hematite (Fig. 6a). Contribution of different components to IRM varies considerably between single specimens. About half of the samples contain dominantly magnetically softer components without a detectable content of hematite. Different contribution of Cint to the NRM within these specimens is displayed by the S-ratio (IRM20.3T/SIRM), which ranges from 0.865 –0.997. Saturation is reached between about 200 and 500 mT. Because of its high spontaneous magnetization Ms, the coercive force of magnetite is shape dominated and has a theoretical maximum of about 300 mT for the saturation field (for indefinite long needles). Therefore saturation fields higher than 300 mT are an indication for the presence of Ti-rich titanomagnetite. Ti-rich titanomagnetites with a composition around Fe2.4Ti0.6O4, typical for basalt, are dominated by stress anisotropy and the coercivity can be increased well above 300 mT by internal stresses (Appel 1987). For the remaining samples lower S-ratios (,0.7) and a significant contribution of Chigh to the IRM was observed. Saturation fields of these samples have values 1 T which verifies a considerable contribution of hematite (Table 1). For samples with significant content of hematite (Chigh) only one lower coercive component can be analysed by CLG distribution, and its relation to Clow or Cint is uncertain because of limited resolution. Figure 7 displays bivariate plots of NRM and SIRM v. magnetic susceptibility k. The parameters show high variations between single sites. NRM values are very low for basalts which could be an indication for strong alteration and therefore a secondary origin of the NRM. A positive relationship between k and NRM or SIRM is given for SIRM values above about 5  103 mA m21 and NRM values above about 50 mA m21. This indicates very different concentrations of ferro(i)magnetic

Fig. 7. Bivariate plots of NRM and SIRM v. magnetic susceptibility.

minerals. A complex nature of NRM is furthermore supported by a strong variation of NRM/ SIRM ratios.

Palaeomagnetic results Remanence directions were determined from straight segments in the Zijderveld-diagram using principal component analyses. For the component with highest coercivity also stable end point directions were used. The demagnetization behaviour within single sites was mainly smooth and easy to analyse; directions of characteristic components in general show good grouping (Fig. 8). In contrast, site mean directions obtained by Fisher statistics (Fisher 1953) vary to a high extent. Basically AfD of NRM shows three different components (Fig. 9): a low-coercivity component (LC-C) which is demagnetized at field intervals between 0 and 20 mT, a high-coercivity component (HC-C) demagnetized at fields between 15 and 100 mT, and an ultra-high coercivity component (UC-C) defined as the direction remaining at 100 mT. In accordance with results of rock magnetic investigations, we relate the LC-C to magnetite with a possible contribution of multi-domain titanomagnetite, the HC-C to Ti-rich titanomagnetite, and the UC-C to hematite. Relative contribution of different components to NRM varies to a high extent between single sites. In most sites, the HC-C represents the highest contribution (Fig. 9a). To a minor degree, the LC-C or UC-C are more dominant (Fig. 9b, c). All components show partly normal and reverse directions. For statistical analysis, reverse polarity directions were inverted. ThD of NRM applied to sites with more than about 75% of magnetization left after AfD reveals a high-temperature component (HT-C) unblocking in the range of 500–580 8C and an ultra-high temperature component (UT-C) unblocking above 600 8C (Fig. 10a). These components are most

Fig. 8. Equal area stereoplots of single sites (geographical coordinates): (a) LC-C of abv 5; (b) HC-C of abv 12. Black cross, mean direction; black circle, 95% confidence limit.

THE SIANG WINDOW OF ARUNACHAL PRADESH

Fig. 9. Zijderveld-diagrams and intensity curves of alternating field demagnetization for representative samples (geographical coordinates): (a) abv 28-7-1; (b) abv 14-1-1; (c) abv 11-6-1.

81

82

U. LIEBKE ET AL.

Fig. 10. Zijderveld-diagrams and intensity curves of thermal demagnetization for representative samples (geographical coordinates): (a) abv 28-7-1; (b) abv 31-8-2.

likely related to magnetite and hematite, respectively. In few specimens also a low-temperature component (LT-C) demagnetized below 450 8C occurs, which can be allocated to Ti-rich titanomagnetite (Fig. 10b). Comparison of AfD and ThD reveals coincidence of the UC-C and UT-C, which confirms that both are related to hematite. No distinct correlation of other components was found. In addition, three sites showing significant results by AfD were thermally demagnetized to obtain further relationships between components. Most specimens of these sites reveal a high increase in magnetization between 500 and 625 8C, probably indicating new formation of magnetite due to high temperature oxidation of Ti-rich titanomagnetite. This is confirmed by a strong increase in magnetic

susceptibility at the same temperatures. Unfortunately remanence directions of the newly formed magnetite obscure directions of initial (titano)magnetite and therefore no characteristic remanence directions could be isolated. Site mean directions and statistical parameters are listed in Table 2. AMS measurements were done in order to exclude that remanence directions were significantly influenced by magnetic anisotropy. A generally low degree with P-factors (Jelinek 1977) ranging from 1.007–1.031, and a random distribution of the principle susceptibility axes confirm a low influence of magnetic anisotropy. Site mean values of the UC-C show a rather random distribution and therefore are not further discussed. Figure 11 shows equal area stereoplots

Table 2. Site mean directions and statistical parameters of LC-C and HC-C

Site

lat. (8N)

long. (8E) 0

28819.776

94858.116

abv 2

28819.2060

94859.3230

abv 3

28819.3880

94859.6210

abv 4

28820.5560

94859.3920

abv 5

28820.8220

95800.1830

abv 6

28820.8580

95800.4970

abv 8

28823.2990

95804.6650

abv 9

28823.0300

95804.5130

abv 11

28821.1720

95802.8340

abv 12

28844.6870

94852.7920

abv 13

28843.8440

94855.4320

abv 14

28843.0500

94857.4480

abv 15

28842.6410

94857.9030

abv 18

28822.5180

95803.8540

abv 19

28822.3530

95803.7730

abv 20 abv 21

28821.0830 28821.0830

95802.8540 95802.8540

abv 22

28821.0830

95802.8540

Nm

Ni (nor./rev.)

LC-C HC-C LC-C HC-C LC-C HC-C LC-C HC-C LC-C HC-C LC-C HC-C LC-C HC-C LC-C HC-C LC-C HC-C LC-C HC-C LC-C HC-C LC-C HC-C LC-C HC-C LC-C HC-C LC-C HC-C HC-C LC-C HC-C LC-C HC-C

9 9 8 8 9 9 8 8 8 8 9 9 9 8 8 8 10 10 10 10 8 8 8 8 8 8 8 8 9 9 9 8 8 8 8

5 (5/0) 7 (5/2) 7 (7/0) 6 (6/0) 7 (4/3) 8 (8/0) 5 (5/0) 6 (6/0) 6 (6/0) 8 (8/0) 5 (2/3) 7 (7/0) 5 (4/1) 7 (6/1) 5 (0/5) 8 (8/0) 10 (10/0) 8 (0/8) 3 (3/0) 10 (10/0) 7 (7/0) 8 (3/5) 8 (6/2) 8 (2/6) 5 (5/0) 4 (3/1) 8 (5/3) 8 (5/3) 3 (0/3) 6 (6/0) 9 (0/9) 7 (2/5) 8 (7/1) 5 (5/0) 8 (0/8)

Bedding dip dir./dip

In situ Dec. (8)

Inc. (8)

Tilt corrected Dec. (8)

Inc. (8)

a95

k

– – – – – – – – – – – – – – – – – – – – 100/45 – – – – – – – – – 077/21 078/22

26.1 43.1 160.2 1.7 135.6 68 340.8 351 0.7 89.1 65.7 109.9 296.2 126.7 102.5 90 357.3 140.2 107.1 30.4 24.8 135.1 14 355.5 352.9 329.3 355 266.8 109.8 220.6 155.9 327.4 207.5 55.6 33.8

49.6 10 251.4 42.3 233.3 26.4 57.5 84.6 36.3 32.8 23.8 33.5 226.6 31.2 12.1 15.5 36.3 36.5 222.7 39.8 19.4 9.4 46.5 41.8 244.2 26.3 25.3 -9.4 7.3 46.9 75.5 68.3 65.1 71.6 69.2

– – – – – – – – – – – – – – – – – – – – 34 139.6 – – – – – – – – 109.1 21.3 152.3 29.5 22.6

– – – – – – – – – – – – – – – – – – – – 3.7 227.1 – – – – – – – – 62.4 65.5 70.3 34.9 30

10.4 24.7 43.2 8.5 37 11.9 45.8 22 11.1 13.2 18.4 7.5 41.3 4.3 21.7 11.9 13.6 16.3 49.4 8 45.4 22.4 30.8 18 61.9 5.2 39.8 30.5 89.3 6.6 2.2 29.8 13.3 14.8 9.2

55.1 7.1 2.9 62.7 3.6 22.6 3.7 10.2 37.3 18.5 18.2 65 4.4 193.7 19 22.7 13.5 12.5 7.3 37.6 2.7 7.1 4.2 9.2 2.5 309.6 2.4 3.8 3 104.1 532.8 5 18.3 27.8 37.5

015/40

83

(Continued)

THE SIANG WINDOW OF ARUNACHAL PRADESH

abv 1

0

C

84

Table 2. Continued

Site

long. (8E)

C

Nm

Ni (nor./rev.)

abv 23

28825.6990

95815.8820

4 (4/0) 7 (7/0)

abv 24 abv 25

28825.434 28825.2450

0

95815.666 95814.6520

abv 26 abv 27 abv 28

28822.8890 28821.9960 28821.6540

95813.8020 95813.2940 95813.0200

abv 29

28815.6930

95812.8410

abv 30

28812.2730

95813.5470

abv 31

28810.4050

95801.3200

abv 32

28808.4480

96805.5680

abv 33

28808.1770

95806.1620

abv 34

28808.0830

95808.1150

abv 35 abv 36 abv 37

28808.2790 28814.4070 28811.3320

95808.7250 94859.3400 95800.2770

abv 38

28809.0870

94803.8720

8 8 10 8 8 8 10 10 8 8 9 9 9 9 9 9 8 8 8 8 9 9 10 10

8 (5/3) 8 (8/0) 10 (8/2) 3 (2/1) 7 (6/1) 8 (7/1) 10 (3/7) 10 (1/9) 5 (5/0) 6 (0/6) 9 (5/4) 8 (0/8) 9 (6/3) 9 (3/6) 9 (9/0) 9 (7/2) 6 (4/2) 6 (6/0) 6 (6/0) 8 (7/1) 9 (9/0) 8 (4/4) 4 (4/0) 10 (10/0)

abv 39

28812.7280

94814.0550

LC-C HC-C – LC-C HC-C HC-C HC-C LC-C HC-C LC-C HC-C LC-C HC-C LC-C HC-C LC-C HC-C LC-C HC-C LC-C HC-C HC-C LC-C LC-C HC-C LC-C HC-C –

8 8

0

Bedding dip dir./dip

In situ Dec. (8)

110/18 – – – – – – 065/78 – – – – – 220/51 – – – 302/38 – 320/18 – – – – – – – –

22.2 49.5

61.6 74.2

48.8 92.3 253.6 230.4 61.7 216.8 74 338.4 205.7 282.3 338.7 57.6 53 132.2 264.2 13.6 27.3 201.1 212.3 312.8 28.6 79.9 354.5 50.1

22.8 37.5 28.3 2 64 28.9 2.6 25.4 36.3 217.3 8.5 7.1 20.6 20.3 211.6 36 8.8 24.6 11.1 82.5 67.4 52 38.4 76.5 62.6

Inc. (8)

Tilt corrected Dec. (8)

Inc. (8)

a95

k

51.3 80.8 – – – – – 60.6 171.4 – – – – 333.7 36.3 – – 272.2 12.1 21.1 205.5 – – – – – – –

56.2 60.9 – – – – – 249 60.5 – – – – 3.5 68.7 – – 4 24.1 16.7 19.2 – – – – – – –

7.9 17.7

34.9 12.6

27.6 15.6 20.4 2.4 26.8 5.6 28.3 29 40.5 4.3 25 7 38.6 42.4 35.1 9.1 16.7 5.4 19.2 11 12.5 10.3 32.2 8

5 13.5 6.6 108.8 6 98.6 9.2 3.7 4.5 243.2 5.2 63.8 2.7 2.4 8.8 32.8 5.3 125.8 13.1 26.2 17.9 29.6 9.1 377.9

lat., latitude; long., longitude; Nm, number of specimens measured; Ni, number of specimens included in statistics; nor., normal polarity; rev., reverse polarity; dip dir., dip direction; Dec., declination; Inc., inclination; a95, 95% confidence angle; k, precision parameter; missing site numbers are due to sampled pilot sites of the Yinkiong formation; the bedding was obtained by adjacent sediments.

U. LIEBKE ET AL.

lat. (8N)

THE SIANG WINDOW OF ARUNACHAL PRADESH

85

Fig. 11. Equal area stereoplots of site mean directions (geographical coordinates) for the (a) LC-C and (b) HC-C. Only sites with k . 10 and a95 , 25 are displayed. Expected palaeodirections for the study area for Early Miocene (no. 1), Early Permian (no. 2), and Upper Carboniferous (no. 3) are also shown (calculated from the APWP of Besse & Courtillot (2002) for 1 and McFadden & McElhinny (1995) for 2&3), as well as the present day Earth’s magnetic dipole field (no. 4). Directions marked by ‘R’ were inverted in Figure 12 (a). N, number of specimens.

of remanence directions residing in the LC-C and HC-C, as well as expected palaeodirections for possible remanence acquisition ages and the present day Earth’s magnetic dipole field. In Figure 11 density plots of the LC-C and HC-C are displayed. In the case of the LC-C a trend of northpointing directions with inclinations slightly higher than the present day Earth’s magnetic dipole field (46.88) occurs. The HC-C yields a trend to clockwise rotations and a second trend of very steep inclinations (.608) which fits to the expected palaeofields of Upper Carboniferous–Early Permian (Fig. 12b). Few sites of the LC-C also show inclinations .608. Bedding data could be obtained only at site locations where the contact between volcanic rocks and sediments was exposed, that is, at nine sites. Because of the complex structure of the area and insufficient structural data, results of fold tests are ambiguous. The McFadden (1990) fold test and the inclination-only fold test (Enkin & Watson 1996) were applied to the HC-C and the LC-C of all sites with structural control, separated for components with higher inclinations (sites abv 20– 23) and lower inclinations (sites abv 13, abv 28, abv 31, abv 33, abv 34). In all cases the McFadden (1990) fold test did not reach the 95% significance level. The inclination-only fold test was used to exclude effects due to different vertical axis rotations. Determination of the k-value was performed after Enkin & Watson (1996), whose method is also applicable to steep inclinations. Results for the sites with lower inclinations and with higher inclinations are similar and support a

Fig. 12. Density plots of site mean directions in geographical coordinates (lower hemisphere) of (a) HC-C (inverted directions: abv 28, abv 30, abv 34; see Fig. 11b); (b) HC-C without inverted directions; (c) LC-C. Numbers of expected palaeofields and the present day Earth’s magnetic field are explained in Figure 11. Only sites with k . 10 and a95 , 25 were used.

secondary origin of both the LC-C and the HC-C (Fig. 13).

Geochronology and thermochronology The geochronology of basaltic lithologies is difficult, due to the typically low abundance of radiogenic elements and the chemical instability of the datable phases such as plagioclase, mafic silicates and glass. The collected pilot samples are typically fine grained, hampering the separation of pure mineral phases. We have performed whole rock K –Ar geochronology on five basalt samples and additionally on one shale sample from the Yinkiong Fm. A relatively fresh large basalt boulder from a river bed in the Changsin valley north of the Siang river towards Sibbum was dated to Late Carboniferous age (319 + 15 Ma). Further basaltic samples yield ages between 87.2 + 1.3 and 24.9 + 0.4 Ma. Fine mineral fractions (,2 mm and ,0.2 mm) of the shale sample yielded 62.3 + 0.7 and 46.8 + 0.7 Ma, respectively (see analytical details in Table 3).

86

U. LIEBKE ET AL.

samples (Fig. 14). The zircon (U –Th)/He ages range between 71 and 9 Ma; they are considerably younger than the ZFT ages.

Discussion Age constraints on the Abor volcanism and deposition of the sediments

Fig. 13. Results of the inclination-only fold tests (Enkin & Watson 1996) for: (a) all sites with higher inclinations; and (b) all sites with lower inclinations.

Zircon fission track (ZFT) and (U –Th)/He (ZHe) low-temperature thermochronology were performed on two grey quartz wacke samples of the Yinkiong Fm and on a quartzite sample of the Miri Fm, in order to constrain the provenance signatures and the post-effusive thermal history of the Siang window. In each case, ZFT and ZHe analyses were performed on zircon aliquots from the same sample. The results of ZFT age determinations are listed in Table 3. The single grain ages show a wide scatter; the chi-square tests fail and the values of dispersion are also high. The central ages calculated for the samples have therefore no any geological meaning. In the samples the single grain ages form relatively distinct clusters around 150 and 260 Ma (AbV-11b), 70 and 260 Ma (YS-7) and c. 56 Ma (YS-16a) indicating the mixed character of the

Due to the dense vegetation the contact between sedimentary rocks and Abor volcanic rocks is rarely exposed, which makes it difficult to determine the stratigraphic position of the volcanic rocks. Intercalation of the Abor volcanic rocks with the sediments of the Yinkiong Fm, as described by Jain & Thakur (1978), Singh & Kumar (1990), Singh (1993), and Acharyya (2007) was not observed within the sampling area, thus our geochronological results cannot directly constrain the Late Paleocene to Early Eocene age of the Abor volcanism (cf. Acharyya 1994). One basalt sample yields a Late Carboniferous whole rock K –Ar age (319 + 15 Ma), which, within analytical uncertainty, agrees with the biostratigraphic age assignment based on mollusc and palaeofloral assemblages (Singh 1981; Sinha et al. 1986). This age datum provides the first geochronological support for Late Palaeozoic volcanism in the Siang Valley. Other basalt samples yield K –Ar ages between 87.2 + 1.3 and 24.9 + 0.4 Ma. These results are interpreted as partly reset ages resulting from argon loss during the partial transformation of the sampled volcanic rocks. This process is reflected also in the formation of epidote and chlorite in the feldspars and the glassy matrix of some samples. In addition, rock magnetic analyses reveal advanced low-temperature oxidation of Ti-rich titanomagnetite. Finally, major element geochemical patterns also indicate a posteffusive oxidation (Sengupta et al. 1996). According to the low-temperature chronometers the transformation of the rock-forming phases and the oxidation of the Fe-oxides took place under diagenetic conditions. Zircon fission track ages form well defined age clusters. The individual zircon grains in the age clusters have similar uranium content, thus the metamictization-controlled partial rejuvenation of the young populations is not probable. We interpret the age clusters as Permian, Jurassic, Late Cretaceous, and Paleocene cooling or formation ages of the zircon-bearing rocks of the source area of the sediment (Fig. 14). The temperature range of ZFT partial reset is around 220–260 8C; the ZFT age distribution of the sandstone samples clearly show that during their thermal evolution they remained below this threshold. The young age clusters of the Yinkiong

THE SIANG WINDOW OF ARUNACHAL PRADESH

87

Fig. 14. Zircon fission track and (U–Th)/He single grain age distributions of the sandstone samples.

Fm samples preclude the sedimentation age of the sandstones being Permian, and documents that their deposition occurred after c. 56 Ma. The youngest age cluster of the Miri Quartzite at c. 150 Ma

does not support the widely suggested Proterozoic to Permian ages for this unit. Very probably, the major source of both the Yinkiong and Miri sediments was located on the overriding Asian plate

88

U. LIEBKE ET AL.

Table 3a. Results from geochronology: (a) K –Ar dating of the basalts and shales Sample Drift boulder AbV-1 AbV-14 AbV-20 AbV-24 YS 10b ,2 mm YS 10b ,0.2 mm

Lithology

K2O (wt. %)

40 Ar* (nl/g) STP

Hyaloclastic basalt Basalt Basalt Basalt Basalt Shale Shale

0.11 0.49 0.93 1.88 2.01 3.50 3.80

1.24 1.411 1.540 1.522 1.693 7.150 5.810

40

Ar* (%)

Age (Ma)

2 s-Error (Ma)

30.62 72.56 70.64 67.29 37.12 86.82 81.10

319.4 87.2 50.6 24.9 25.9 62.3 46.8

14.8 1.3 0.7 0.4 0.7 0.7 0.7

Samples used for geochronology are indicated by ‘AbV’ or ‘YS’ to distinguish them from samples used for palaeomagnetic analyses (‘abv’). Sample numbers indicate the sampling site, the site locations of the AbV samples are given in Table 2. Locations of the YS sites are adjacent to AbV-8 (YS 7) and AbV-11 (YS 10). No volcanic rocks are exposed near site YS 16 in the Sirnyuk Valley at Jegging (28831.5780 N/95802.9700 E). The drift boulder was located at 288180 N/958110 E.

and/or in the Gangdese thrust belt; the Jurassic to Cretaceous thermal events retained in the detrital zircon age spectra can clearly not be connected to the Indian Plate. The Yinkiong Fm underwent burial soon after its deposition as indicated by the c. 47 Ma K –Ar age of the shale in its 0.2 mm size fraction – the age of diagenetic illite growth – from which the youngest detrital ZFT age clusters do not differ significantly. Our results infer that volcanic bodies intruding the Miri quartzites or interbedded with the Yinkiong Fm are clearly postPermian, and pertain to a volcanism latest Cretaceous to Early Tertiary in age. We therefore conclude that the Siang window exposes the products of more than one major volcanic event, which confirms the model put forward earlier by Singh (1984), Singh & De (1984) and Singh & Malhotra (1987).

Age of remanence acquisition The single crystal zircon (U –Th)/He ages show an extremely wide scatter between latest Cretaceous and Late Miocene (Table 3b). A part of the ZHe ages are considerably younger than the youngest age cluster in the ZFT ages and even younger than the base of the Siwalik Group below the MBT. Thus, the ZHe ages cannot be cooling ages of the source terrains of the sediment; the ZHe thermochronometer underwent partial degassing after the sedimentation. This process results in a wide scatter in ages, because the diffusion of He (and thus the closure temperature) depends on the size, shape, content of U & Th, and their distribution, as well as alpha dosage of the individual crystals. Reiners et al. (2004) determined the closure temperature of the ZHe thermochronometer to be c. 180 8C. The dated samples were close to this threshold, but definitely below it. From the distribution of the ZHe ages, from the size of the dated zircon crystals, as well as based on the agreement of the Cenozoic

K –Ar ages between basalt and shale samples we postulate that the Siang window rocks (Miri Fm. Abor volcanic series, Yinkiong Fm.) shared a largely common thermal history since Palaeogene times and experienced a maximum burial temperature around 150– 170 8C for the Late Miocene time. Applied fold tests (McFadden 1990; Enkin & Watson 1996) suggest a secondary formation of both the LC-C and the HC-C. The occurrence of Ti-rich titanomagnetite constrains the possible peak temperature of metamorphism to 350 8C as at higher temperatures Ti-rich titanomagnetite would have decayed into magnetite and ilmentite near phases (Gromme´ et al. 1969; Tucker & O’Reilly 1980). Few studies of the metamorphic evolution of adjacent areas were made. Goswami et al. (2009) report results of structural and metamorphic analyses on rocks of the Lesser and Greater Himalayan sequences in the western Arunachal Himalaya and Gururajan & Chowdhury (2003) report a low-grade metamorphic event in the Lesser Himalaya of the Lohit valley located south-east of the Siang window. In both cases timing is poorly constrained. Ding et al. (2001) report high grade metamorphic events in the Namche Barwa Syntaxis at about 160, 65, 40, and 11 Ma. According to Booth et al. (2009) the Namula thrust separates high-grade rocks at the core of the Namche Barwa Syntaxis to the north from lower grade rocks to the south. The younger K –Ar ages of the present study (87.2 – 24.9 Ma) are very likely partly reset ages and reveal an overprinting event which occurred probably between the India–Asia collision and Early Miocene. This would fit to high-grade metamorphism in the Namche Barwa Syntaxis at 40 Ma, which is related to the early stages of the India– Asia collision (Ding et al. 2001). In addition, Dunkl et al. (2007) reported a metamorphic event at c. 24 Ma in the Triassic flysch of the Tethyan Himalayan in SE Tibet, which also fits to the

Sample

aliq. vol.

He s.e.

mass

U238 s.e.

Th232 mass s.e.

Th/U

mass

Sm s.e.

Ejection correct.

Uncor. He-age

Ft – Cor. He-age

1s

(ncc)

(ncc)

(ng)

(ng)

(ng)

(ng)

ratio

(ng)

(ng)

(Ft)

(Ma)

(Ma)

(Ma)

YS 16a

#1 #2 #3

1.534 2.087 0.813

0.026 0.036 0.015

0.861 0.650 0.812

0.016 0.012 0.015

0.258 0.198 0.536

0.006 0.005 0.013

0.30 0.30 0.66

0.014 0.011 0.051

0.001 0.001 0.003

0.71 0.74 0.74

13.8 24.8 7.2

19.5 33.4 9.7

0.5 0.8 0.2

AbV-11b

#1 #2 #3

5.163 5.662 2.096

0.086 0.094 0.036

0.851 1.009 0.761

0.015 0.018 0.014

0.340 0.677 0.876

0.008 0.016 0.021

0.40 0.67 1.15

0.016 0.029 0.020

0.001 0.002 0.001

0.74 0.76 0.75

45.8 40.0 17.9

61.6 52.9 23.9

1.4 1.2 0.5

YS 7

#1 #2 #3

0.308 1.092 5.328

0.006 0.019 0.089

0.322 0.571 0.761

0.006 0.010 0.014

0.241 0.192 0.825

0.006 0.005 0.020

0.75 0.34 1.08

0.041 0.005 0.021

0.003 0.000 0.001

0.69 0.67 0.64

6.7 14.7 46.0

9.8 21.8 71.4

0.2 0.5 1.6

Numbers in italics: rough uncorrected data. Numbers in bold: age, which can be interpreted in context with geological evolution.

THE SIANG WINDOW OF ARUNACHAL PRADESH

Table 3b. Zircon(U –Th)/He data

89

90

U. LIEBKE ET AL.

Table 3c. Zircon fission track results obtained on the sandstone samples from Arunachal Sample YS-16a YS-7 AbV-11b

Crystal

RhoS

(Ns)

RhoI

(Ni)

RhoD

(Nd)

Chi-sq. P (%)

Disp.

Central Age

+

1s

30 25 25

80.2 123.8 150.2

(2595) (2446) (2687)

39.5 41.8 30.7

(1276) (826) (550)

7.97 7.94 8.39

(4200) (4200) (4200)

0 0 0

0.70 0.65 0.25

100 144 256

+ + +

13 20 19

Track densities (RHO) are as measured (105 tr/cm2); number of tracks counted (N) are shown in brackets. RhoD and Nd are track densities and number of tracks in the CN2 detector. Chi-sq P(%): probability obtaining Chi-square value for n degrees of freedom (where n ¼ no. crystals-1). Disp., dispersion, according to Galbraith & Laslett (1993). Central ages calculated using dosimeter glass: CN 2.

possible timing of metamorphic overprint within the Siang window. Our new geochronological results suggest that the regional thermal overprint in the Siang window did not exceed the diagenetic stage.

The clockwise rotation trends of the HC-C would fit to a clockwise rotation trend reported from palaeomagnetic investigations on Cretaceous to Miocene rocks north and east of the Namche Barwa Syntaxis (Fig. 15 and Table 4; Otofuji et al.

Fig. 15. Map of the eastern part of the Himalayas (modified after Dupont-Nivet et al. 2002) showing the declination maximum (HC-C; density .2.3) of the Abor volcanic rocks (grey area marked by ‘A’). Additionally, rotations with respect to stable Eurasia are displayed by the black arrows marked with numbers. The rotations were determined by the difference between expected declinations (APWP of Eurasia; Besse & Courtillot 2002) and observed declinations of previous palaeomagnetic studies; references are given in Table 4.

THE SIANG WINDOW OF ARUNACHAL PRADESH

91

Table 4. Results of previous palaeomagnetic investigations on the eastern Tibetan Plateau No.

locality

lat. (8N)

long. (8E)

Age

Do

De

Drot (Do 2 De)

a95

ref.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Ya`an Markam Yuanmou Chuxiong Yongping Hekou group Lanping Basin Xining – lanzhou Fenghuoshan Yushu Tuoluo Jungong Xining Lanping Basin Xiao Qaidam (HTC) E Bo Liang

30.1 29.7 25.9 25 25.5 39 25.6 36.2 34.5 33.2 35.3 34.7 36.5 25.8 37.4 38.7

103 98.6 101.7 101.5 99.5 99.6 100.5 103.5 92.8 96.7 98.6 100.7 102 99.4 95.3 92.8

Cr lCr uCr Cr lCr lCr uCr lCr Ec Mi Mi Mi Ec uCr Mi Og

2.1 48.2 26.9 44.6 42 41.9 6.9 44.7 25.5 35.6 20.1 19.8 29.3 28.1 0.7 8

10.1 16 9.6 9.7 10.5 16.9 9.5 10.7 9.6 5.3 8.8 5.2 12.8 9.5 5.7 10.6

28 32.2 17.3 34.9 31.5 24 22.6 34 15.9 30.3 11.3 14.6 16.5 18.6 25 22.6

11.3 8.8 3.6 10.7 15.7 5.1 8.6 5.1 6 9 27.1 6 13.2 2.4 5.8 5.1

1 1 2 3 4 5 6 7 8 9 9 9 9 10 11 11

lat., latitude; long., longitude; D, declination; o, observed; e, expected; rot, rotation; a95, 95% confidence limit; ref., reference; Mi, Miocene; Og, Oligocene; Ec, Eocene; Cr, Cretaceous; l, lower; u, upper. References: 1: Otofuji et al. (1990), 2: Otofuji et al. (1998), 3: Funahara et al. (1992), 4: Funahara et al. (1993), 5: Huang et al. (1992), 6: Huang & Opdyke (1993), 7: Halim et al. (1998), 8: Halim et al. (1998) and Lin & Watts (1988), 9: Cogne´ et al. (1999), 10: Yang et al. (2001), 11: Dupont-Nivet et al. (2002).

1990, 1998; Funahara et al. 1992, 1993; Huang & Opdyke 1993; Sato et al. 1999, 2007; Yang et al. 2001). The LC-C shows a trend of remanence directions which are close to the present day Earth’s magnetic dipole field. These directions indicate a more recent component probably formed during alteration of the rocks. The trend to high inclinations (.608) of the HC-C could be due to: (a) tilting around horizontal axis after remanence acquisition; or (b) a primary remanent magnetization of Late Palaeozoic age. Remanence directions residing in HC-C are plotted in Figure 16 on a structural map of the Siang window. The sites with high inclinations are marked by orange circles. No clear correlation to structures on the map can be identified. However, six of the sites with high inclinations (abv 4, abv 20, abv 21, abv 22, abv 35, abv 38) can be connected by an approximately NNW–SSE trending line, which is the general trend of most fold axis and faults. Whether this is an indication for a larger structure (i.e., the axis of the anticline of the Siang window) is questionable as it is difficult to explain why adjacent sites (with lower inclinations) should not have been tilted in the same way. Steepening of the inclinations could rather be due to small scale local tilting of the rocks. A possible explanation could be a system of duplex structures, as it is common in the Lesser Himalaya and the Siwaliks (e.g. Srivastava & Mitra 1994; DeCelles et al. 2001). The second possibility would be a primary magnetization with a Late Palaeozoic remanence age. Expected Late

Palaeozoic palaeodirections yield inclinations between 2698 and 2788 (Fig. 11). Therefore a primary magnetization acquired during emplacement of the Abor volcanic rocks at Late Palaeozoic could explain the high inclinations. However, applied fold tests indicate a secondary origin of remanent magnetization for the high-inclination sites; results of the fold tests are based on four sites with high inclinations only and should be regarded with caution. The Siang window was affected by several deformation events (Jain & Tandon 1974; Singh 1993; Kumar 1997) and the sites are clearly not positioned on two different limbs of one single fold as would be the ideal case for a fold test. Therefore a primary magnetization of the high inclination sites cannot completely be ruled out. Similar rock magnetic properties of the sites with low and high inclinations are not supportive for a partly primary origin. Clarifying the origin of the high-inclination sites is not possible within the present study and calls for further structural and palaeomagnetic investigations.

Structural implications and clockwise rotation Figure 16 demonstrates that the complex pattern of remanence directions is probably due to strong tectonic deformation of the area. The partly high variation in declination and/or inclination of nearby sites indicates small scale structural deformation.

92

U. LIEBKE ET AL.

Fig. 16. Structural map of the Siang window (modified after Acharyya & Saha 2008) displaying observed declinations, inclinations, and values of a95 of remanence directions residing in the HC-C (geographical coordinates).

There is no clear relation between structures on the map and remanence directions. However, a number of sites connected by a NNE–SSW-trending line show similar remanence directions, which supports the general trend of an anticline structure of the Siang window. The density plot of in situ directions residing in HC-C (Fig. 12a, b) reveals a trend to clockwise rotations. The degree of clockwise rotations is not

confined to a certain magnitude, but spreads over a larger interval with a maximum between about 60 and 1258 (density .2.3). The spread in declinations can be partly explained as an artefact created by tilting around sub-horizontal axes after remanence acquisition, but may also indicate that remanence acquisition stretched out over a larger time interval (following India–Asia collision, as long as until the Late Miocene), or that the area consists of sub-units

THE SIANG WINDOW OF ARUNACHAL PRADESH

with different degree of clockwise rotation. We assume that the clockwise rotation trend is attributed to the clockwise rotation trend of the eastern Himalayas, which was observed by several palaeomagnetic investigations on Cretaceous to Miocene rocks north and east of the Namche Barwa Syntaxis and is interpreted as the result of ongoing indentation of India into Eurasia (Fig. 15). Increasing clockwise rotations from the western to the eastern part of the central Himalayas were also obtained in a study of secondary Late Eocene to Early Miocene remanences of Tethyan Himalayan meta-sediments (Schill et al. 2004). Several numerical models and laboratory experiments on the deformation due to India – Asia collision and ongoing indentation of India into Eurasia have predicted clockwise tectonic rotations around the eastern corner of India (e.g. Houseman & England 1993; Beaumont et al. 2001; Cook & Royden 2008). Different models were developed to describe tectonic evolution and uplift of the Tibetan Plateau, such as the ‘block model’, which assumes that the collision zone is comprised of several lithospheric blocks (e.g. Tapponnier et al. 1982; Peltzer & Tapponnier 1988; Avouac & Tapponnier 1993), or those suggesting the existence of a ductile lower crust (e.g. Houseman & England 1996; Royden et al. 1997; Beaumont et al. 2001; Shen et al. 2001). To reveal the relative importance of different models for the evolution of the Tibetan Plateau, palaeomagnetic investigations within the eastern part of the Himalayas are crucial. Our study provides the first such results in the area just south of the Namche Barwa Syntaxis. The continuously distributed maxima of clockwise rotations show that clockwise vertical-axis rotation prevailed at least over the period of remanence acquisition. In comparison to observed rotations in surrounding western or northern areas, the degree of rotation is relatively high (Fig. 15). In consideration that the sampling location is very close to the core of the Namche Barwa Syntaxis, the high degree of rotation is probably due to the high stress rate. Further investigations with a focus on structural control are required in order to get a better insight into the evolution of the eastern Himalayas and the regional tectonic structure of the Siang window.

(2)

(3)

(4)

(5)

(6)

Conclusions (1)

Detrital zircon fission track thermochronology of the Yinkiong and Miri Fms indicates a post-Paleocene and post-Jurassic age of deposition, respectively, inferring that: (a) the sediment source was located on the overriding Asian plate and/or in the Gangdese thrust belt; and (b) the Abor volcanic rocks

93

intercalating or intruding these sediments are not part of the Upper Palaeozoic sequence as widely proposed in the literature; they must reflect one or more distinct, additional, latest Cretaceous to Tertiary volcanic event(s). The Siang window rocks (Miri Fm., Abor volcanic series, Yinkiong Fm.) likely shared a common thermal history since Palaeogene times. Early Miocene K –Ar ages of basalt samples are interpreted as partly reset ages representing a post-effusive thermal overprint. Incomplete reset of zircon (U –Th)/He ages suggest maximum temperatures of c. 150– 180 8C for Late Miocene, implying that the post-depositional thermal overprint did not exceed diagenetic conditions. The detected transformation in magnetic mineralogy is well in line with these results. Whole rock K –Ar dating of a basalt drift boulder near Sibbum yields an age of 319 + 15 Ma and provides geochronological support for the presence of Upper Palaeozoic rocks in the Abor volcanic series. It can be not excluded that sites with high inclination represent a primary origin during late Palaeozoic times. However, similar rock magnetic properties of all sites are not supportive for variable remanence acquisition ages. The HC-C shows a trend to clockwise rotations. The inclination only fold test (Enkin & Watson 1996) supports a secondary origin of the remanence component. We assume that remanence acquisition of all sites with directions matching with the clockwise rotation trend occurred during the low temperature overprinting event(s) within in the period of India–Asia collision and Late Miocene. Palaeomagnetic data presented in this paper are compatible with the models of the Tibetan Plateau extrusion. The rotation is not confined to a certain magnitude. The HC-C displays a range of declination values between maxima at about 608 and 1258. This indicates that remanence acquisition was performed over a larger time interval during progressive clockwise rotation, or that the area consists of sub-units with different degree of rotation. The complex pattern of remanence directions clearly reveals ongoing tectonic deformation after remanence acquisition. Remanence directions support a general NNE– SSW trend of regional structures. The palaeomagnetic results of this paper do not provide an unambiguous interpretation; however, allow some insights into the kinematics of the crust in this geologically highly unexplored area. Further structural and palaeomagnetic

94

U. LIEBKE ET AL.

investigations are necessary to unravel the complex regional structure of the Siang window. Support on the geological investigations by G. Fuchs and P. Blisniuk is kindly acknowledged. We also thank E. Schill, K.V.V. Satyanarayana, P.B. Gawali, and V. Purushotham Rao for field work assistance and sampling. This study was supported by the German Research Foundation (DFG) and by grants from the Indian Institute of Geomagnetism. We thank S. K. Acharyya and C. Crouzet for critical review of the manuscript.

References Acharyya, S. K. 1994. The Cenozoic foreland basin and tectonics of the eastern Sub-Himalaya: problem and prospects. Himalayan Geology, 15, 3– 21. Acharyya, S. K. 1998. Thrust tectonics and evolution of domes and the syntaxis in Eastern Himalaya, India. Journal of Geological Society of Nepal, 18, 1 –17. Acharyya, S. K. 2005. NE India: a complex terrain of accreted continental blocks. MGMI, Proc. Sem. Mineral & Energy Resources of Eastern & Northeastern India. 29– 54. Acharyya, S. K. 2007. Evolution of the Himalayan Paleogene foreland basin, influence of its litho-packet on the formation of thrust-related domes and windows in the Eastern Himalayas – A review. Journal of Asian Earth Sciences, 31, 1 –17. Acharyya, S. K. & Saha, P. 2008. Geological setting of the Siang Dome located at the Eastern Himalayan Syntaxis. Himalayan Journal of Sciences, Special Issue extended abstracts, 5, 16–17. Acharyya, S. K. & Sengupta, S. 1998. The structure of the Siang Window, its evolution and bearing on the nature of Eastern syntaxis of the Himalaya. National Academy of Science Letters, 21, 177 –192. Appel, E. 1987. Stress anisotropy in Ti-rich titanomagnetites. Physics of Earth and Planetary Interiors, 46, 233– 240. Avouac, J. P. & Tapponnier, P. 1993. Kinematic model of active deformation in central Asia. Geophysical Research Letters, 20, 895–898. Beaumont, C., Jamieson, R. A., Nguyen, M. H. & Lee, B. 2001. Himalayan tectonics explained by extrusion of a low-viscosity crustal channel coupled to focused surface denudation. Nature, 414, 738–742. Besse, J. & Courtillot, V. 2002. Apparent and true polar wander and geometry of the geomagnetic field over the last 200 Myr. Journal of Geophysical Research, 107, 2300–2329; doi: 10.1029/2000JB000050. Bhat, M. I. 1984. Abor volcanics: further evidence for the birth of the Tethys Ocean in the Himalayan segment. Journal of the Geological Society, London, 141, 763– 775. Bhat, M. I. & Ahmad, T. 1990. Petrogenesis and the mantle source characteristics of the Abor volcanic rocks, eastern Himalayas. Journal of the Geological Society of India, 36, 227 –246.

Booth, A. L., Chamberlain, C. P., Kidd, W. S. F. & Zeitler, P. K. 2009. Constraints on the metamorphic evolution of the eastern Himalayan syntaxis from geochronologic and petrologic studies of Namche Barwa. Geological Society of America Bulletin, 121, 385–407. Chen, Z., Burchfiel, B. C. et al. 2000. Global Positioning System measurements from eastern Tibet and their implications for India/Eurasia intercontinental deformation. Journal of Geophysical Research, 105, 16215–16227. Chowdhury, J. R. 1979. Abor Volcanics of the Arunachal Himalaya: a discussion. Journal of the Geological Society of India, 20, 45–46. Cogne´, J. P., Halim, N., Chen, Y. & Courtillot, V. 1999. Resolving the problem of shallow magnetizations of Tertiary age in Asia: insights from paleomagnetic data from the Qiangtang, Kunlun and Qaidam blocks (Tibet, China), and a new hypothesis. Journal of Geophysical Research, 104, 17715– 17734. Cook, K. L. & Royden, L. H. 2008. The role of crustal strength in shaping orogenic plateaus, with application to Tibet. Journal of Geophysical Research, 113, B08407, doi: 10.1029/2007JB005457. DeCelles, P. G., Robinson, D. M., Quade, J., Ojha, T. P., Garzione, C. N., Copeland, P. & Upreti, B. N. 2001. Stratigraphy, structure, and tectonic evolution of the Himalayan fold-thrust belt in western Nepal. Tectonics, 20, 487– 509. Ding, L., Zhong, D. L., Yin, A., Kapp, P. & Harrison, T. M. 2001. Cenozoic structural and metamorphic evolution of the eastern Himalayan syntaxis (Namche Barwa). Earth and Planetary Science Letters, 192, 423–438. Dumitru, T. A. 1993. A new computer-automated microscope stage system for fission-track analysis. Nuclear Tracks and Radiation Measurements, 21, 575–580. Dunkl, I. 2002. TRACKKEY: a Windows program for calculation and graphical presentation of fission track data. Computers and Geosciences, 28(1), 3 –12. Dunkl, I., Lin, D. et al. 2007. Diagenetic and Metamorphic Overprint and Deformation History of Permo-Triassic Tethyan Sediments, SE Tibet. 22nd Himalaya– Karakorum –Tibet Workshop, 22–25 May 2007, Hong Kong, abstract volume. Dupont-Nivet, G., Butler, R. F., Yin, A. & Chan, X. 2002. Paleomagnetism indicates no Noegene rotation of the Qaidam Basin in northern Tibet during Indo-Asian collision. Geology, 30, 263– 266. Enkin, R. J. & Watson, G. S. 1996. Statistical analysis of palaeomagnetic inclination data. Geophysical Journal International, 126, 495 –504. Evans, N. J., Byrne, J. P., Keegan, J. T. & Dotter, L. E. 2005. Determination of Uranium and Thorium in Zircon, Apatite, and Fluorite: application to Laser (U–Th)/He Thermochronology. Journal of Analytical Chemistry, 60, 1159–1165. Farley, K. A., Wolf, R. A. & Silver, L. T. 1996. The effects of long alpha-stopping distance on (U– Th)/ He ages. Geochimica et Cosmochimica Acta, 60, 4223– 4229. Fisher, R. A. 1953. Dispersion on a sphere. Proceedings of the Royal Society of London A, 217, 295–305.

THE SIANG WINDOW OF ARUNACHAL PRADESH Fuhrmann, U., Lippolt, H. J. & Hess, J. C. 1987. Examination of some proposed K–Ar standards: 40Ar/39Ar analyses and conventional K –Ar-Data. Chemical Geology. (Isotope Geoscience Section), 66, 41– 51, Amsterdam. Funahara, S., Nishiwaki, N., Miki, M., Murata, F., Otofuji, Y. & Wang, Y. Z. 1992. Paleomagnetic study of Cretaceous rocks from the Yangtze block, central Yunnan, China: implications for the India–Asia collision. Earth and Planetary Science Letters, 113, 77–91. Funahara, S., Nishiwaki, N., Murata, F., Otofuji, Y. & Wang, Y. Z. 1993. Clockwise rotation of the Red River fault inferred from paleomagnetic study of Cretaceous rocks in the Shan–Thai –Malay block of western Yunnan, China. Earth and Planetary Science Letters, 117, 29–42. Galbraith, R. F. & Laslett, G. M. 1993. Statistical models for mixed fission track ages. Nuclear Tracks and Radiation Measurements, 21, 459– 470. Gansser, A. 1964. Geology of the Himalayas. Wiley-Interscience, New York. Gansser, A. 1974. The Himalayan Tethys. Rivista Italiana di Paleontologia e Stratigrafia Memoria, 14, 393– 411. Gansser, A. 1983. Geology of the Bhutan Himalaya. Denkschriften der Schweizerischen Naturforschenden Gesellschaft, 96. Birkha¨user Verlag, Basel. Gleadow, A. J. W. 1981. Fission-track dating methods: what are the real alternatives? Nuclear Tracks, 5, 3–14. Goswami, S., Bhowmik, S. K. & Dasgupta, S. 2009. Petrology of a non-classical Barrovian inverted metamorphic sequence from the western Arunachal Himalaya, India. Journal of the Asian Earth Sciences, 36, 390–406. Green, P. F. 1981. A new look at statistics in fission track dating. Nuclear Tracks, 5, 77–86. Gromme´, C. S., Wright, T. L. & Peck, D. L. 1969. Magnetic properties and oxidation of iron-titanium oxide minerals in Alae and Makaopuhi lava lakes, Hawaii. Journal of Geophysical Research, 74, 5277–5293. Gururajan, N. S. & Chowdhury, B. K. 2003. Geology and tectonic history of the Lohit valley, Eastern Arunachal Pradesh, India. Journal of Asian Earth Sciences, 21, 731–741. Halim, N., Cogne´, J. P. et al. 1998. New Cretaceous and early Tertiary paleomagnetic results from XiningLanzhou basin, Kunlun and Qiangtang blocks, China: implications on the geodynamic evolution of Asia. Journal of Geophysical Research, 103, 21025– 21045. Heinrichs, H. & Herrmann, A. G. 1990. Praktikum der Analytischen Geochemie. 669, Springer. Hodges, K. V., Geissman, J. W. E. & Glazner, A. F. E. 2000. Tectonics of the Himalaya and southern Tibet from two perspectives, Special focus on the Himalaya. Geological Society of America Bulletin, 112, 324– 350. Hourigan, J. K., Reiners, P. W. & Brandon, M. T. 2005. U–Th zonation-dependent alpha-ejection in (U–Th)/He chronometry. Geochimica et Cosmochimica Acta, 69/13, 3349–3365. Houseman, G. & England, P. 1993. Crustal thickening v. lateral expulsion in the Indian–Asian continental collision. Journal of Geophysical Research, 98, 12 233 –12 249.

95

Houseman, G. & England, P. 1996. A lithospheric thickening model for the Indo-Asian collision. In: Yin, A. & Harrison, T. M. (eds) The Tectonic Evolution of Asia. Cambridge University Press, New York, 3 –17. Huang, K. & Opdyke, N. D. 1993. Paleomagnetic results from Cretaceous and Jurassic rocks of South and Southwest Yunnan: evidence for large clockwise rotations in the Indochina and Shan–Thai–Malay terranes. Earth and Planetary Science Letters, 117, 507– 524. Huang, K., Opdyke, N. D., Li, J. & Peng, X. 1992. Paleomagnetism of Cretaceous rocks from eastern Qiangtang terrane of Tibet. Journal of Geophysical Research, 97, 1789– 1799. Hurford, A. J. 1998. Zeta: the ultimate solution to fissiontrack analysis calibration or just an interim measure? In: Van den Haute, P. & De Corte, F. (eds) Advances in Fission-Track Geochronology. Kluwer Academic Publishers, Dordrecht, 19– 32. Hurford, A. J. & Green, P. F. 1983. The zeta age calibration of fission-track dating. Chemical Geology (Isotope Geoscience Section), 41, 285– 312. Jain, A. K. & Tandon, S. K. 1974. Stratigraphy and structure of the Siang district, Arunachal (NEFA), Himalaya. Himalayan Geology, 4, 28–60. Jain, A. K. & Thakur, V. C. 1978. Abor volcanics of the Arunachal Himalaya. Journal of the geological Society of India, 19, 335–349. Jelinek, V. 1977. The statistical theory of measuring anisotropy of magnetic susceptibility of rocks and its application. Geofyzika Brno, 1– 88. Kumar, G. 1997. Geology of Arunachal Pradesh. Geological Society of India, 217. Kruiver, P. P., Dekkers, M. J. & Heslop, D. 2001. Quantification of magnetic coercivity components by the analysis of acquisition curves of isothermal remenent magnetisation. Earth and Planetary Science Letters, 189, 269–276. Le Fort, P. 1975. Himalayas: the collided range. Present knowledge of the continental arc. American Journal of Science, 275-A, 1– 44. Lin, J. & Watts, D. R. 1988. Palaeomagnetic results from the Tibetan Plateau. Philosophical Transactions of the Royal Society of London A, 327, 239–262. Lowrie, W. 1990. Identification of ferromagnetic minerals in a rock by coercivity and unblocking temperature properties. Geophysical Research Letters, 17, 159– 162. Mattauer, M. 2002. New GPS data in China: a key for a better understanding of the Cainozoic tectonics of Asia. Comptes Rendus Geoscience, 334, 809– 810. McDougall, I. & Harrison, T. M. 1999. Geochronology and Thermochronology by the 40Ar/39Ar Method. Oxford University Press, New York. McFadden, P. L. 1990. A new fold test for palaeomagnetic studies. Geophysical Journal Interior, 103, 163– 169. McFadden, P. L. & McElhinny, M. W. 1995. Combining groups of paleomagnetic directions or poles. Geophysical Research Letters, 100, 316–317. Molnar, P. & Tapponnier, P. 1975. Cenozoic Tectonics of Asia: effects of a Continental Collision. Science, 189, 419–426.

96

U. LIEBKE ET AL.

O’Reilly, W. 1984. Rock and Mineral Magnetism. Blackie, Glasgow and London; Chapman & Hall, New York, 220. Otofuji, Y., Inoue, Y., Funahara, S., Murata, F. & Zheng, X. 1990. Palaeomagnetic study of eastern Tibet-deformation of the Three Rivers region. Geophysical Journal International, 103, 85–94. Otofuji, Y., Liu, Y., Yokoyama, M., Tamai, M. & Yin, J. 1998. Tectonic deformation of the southwestern part of the Yangtze craton inferred from paleomagnetism. Earth and Planetary Science Letters, 156, 47–60. Pan, G., Ding, J., Yao, D. & Wang, L. 2004. Geological map Qinghai-Xizang (Tibet) Plateau and Adjacent Areas (1:1500000). Chengdu Institute of Geology and Mineral Resources, China Geological Survey. Chengdu Cartographic Publishing House. Peltzer, P. & Tapponnier, P. 1988. Formation and evolution of strike–slip faults, rifts, and basins during the India–Asia collision: an experimental approach. Journal of Geophysical Research, 93, 15 085–15 117. Prasad, B., Dey, A. K., Gogoi, P. K. & Maithani, A. K. 1989. Early Permian plant microfossils from the intratrappean beds of Abor Volcanics, Arunachal Pradesh, India. Journal of the Geological Society India, 34, 83–88. Reiners, P. W. 2005. Zircon (U–Th)/He Thermochronometry. Reviews in Mineralogy & Geochemistry, 58, 151– 179, Mineralogical Society of America. Reiners, P. W., Spell, T. L., Nicolescu, S. & Zanetti, K. A. 2004. Zircon (U –Th)/He thermochronometry: He diffusion and comparisons with 40Ar/39Ar dating. Geochimica et Cosmochimica Acta, 68, 1857– 1887. Royden, L. H., Burchfiel, B. C., King, R. W., Wang, E., Chen, Z., Shen, F. & Liu, Y. 1997. Surface deformation and lower crustal flow in eastern Tibet. Science, 276, 788–790. Sato, K., Liu, Y., Zhu, Z., Yang, Z. & Otofuji, Y. 1999. Paleomagnetic study of middle Cretaceous rocks from Yunlong, western Yunnan, China: evidence of southward displacement of Indochina. Elsevier Journal, 165, 1– 15. Sato, K., Liu, Y., Wang, Y., Yokoyama, M., Yoshioka, S., Yang, Z. & Otofuji, Y. 2007. Paleomagnetic study of Cretaceous rocks from Pu’er, western Yunnan, China: evidence of internal deformation of the Indochina block. Earth and Planetary Science Letters, 258, 1– 15. Schill, E., Appel, E., Crouzet, C., Gautam, P., Wehland, F. & Staiger, M. 2004. Oroclinal bending v. regional significant clockwise rotations in the Himalayan arc-constraints from secondary pyrrhotite remanences. In: Sussman, A. J. & Weil, A. B. (eds) Orogenic Curvature: Integrating Paleomagnetic and structural analyses. Geological Society of America, Colorado, Special Paper, 383, 73– 85. Schumacher, E. 1975. Herstellung von 99,9997% 38Ar fu¨r die 40K/40Ar Geochronologie. Geochronologia Chimia, 24, 441– 442. Sengupta, S., Acharyya, S. K. & De Smeth, J. B. 1996. Geochemical characteristics of the Abor volcanic rocks, NE Himalaya, India: nature and early Eocene magmatism. Journal of the Geological Society, London, 153, 695– 704.

Shen, F., Royden, L. H. & Burchfiel, B. C. 2001. Large-scale crustal deformation of the Tibetan Plateau. Journal of Geophysical Research, 106, 6793– 6816. Shen, Z.-K., Lu¨, J., Wang, M. & Bu¨rgmann, R. 2005. Contemporary crustal deformation around the southeast borderland of the Tibetan Plateau. Journal of Geophysical Research, 110, B11409, doi: 10.1029/ 2004JB003421. Singh, S. 1984. A reappraisal of Yinkiong Formation with reference to Dalbuing area, East Siang district, Arunachal Himalaya. Indian Minerals, 38, 34–38. Singh, S. 1993. Geology and tectonics of the Eastern Syntaxial Bend, Arunachal Himalaya. Journal of Himalayan Geology, 4, 149– 163. Singh, S. & De, A. K. 1984. An account of Lower Tertiary rocks of East and West Siang districts, Arunachal Himalaya. Seminar on Recent Advances in Cenozoic Geology of North Eastern Region of India. Abstracts Book, Dibrugarh. Singh, S. & Kumar, G. 1990. Basic volcanism in the Arunachal Himalaya and their stratigraphic positions. Semn. Tectonics and Metallogeny of Ophiolites and recent advances in geology of northeast India. Manipur, 28 (abstract). Singh, S. & Malhotra, G. 1987. A note on the basic volcanics of the Siang valley, Arunachal Himalaya. Indian Minerals, 41, 60–63. Singh, T. 1981. Age and faunal affinity of the Garu Formation, Arunachal Pradesh. Himalayan Geology, 10, 263–270. Sinha, N. K., Satsangi, P. P. & Misra, U. K. 1986. Paleontology of Permian and Eocene rocks of Siang District, Arunachal Pradesh. Records of the Geological Survey of India, 114, 53– 60. Sol, S., Meltzer, A. S. et al. 2007. Geodynamics of the southeastern Tibetan Plateau from seismic anisotropy and geodesy. Geology, 35, 563–566. Srivastava, P. & Mitra, G. 1994. Thrust geometries and deep structures of the outer and lesser Himalaya, kumaon and Garhwal (India): implications for evolution of the Himalayan fold-and thrust belt. Tectonics, 13, 89–109. Steiger, R. H. & Ja¨ger, E. 1977. Subcommission on Geochronology: convention on the Use of Decay Constants in Geo- and Cosmochronology. Earth Planetary Science Letters, 36, 359–362. Tapponnier, P., Peltzer, G., Le Dain, A. Y., Armijo, R. & Cobbold, P. 1982. Propagating extrusion tectonics in Asia: new insights from simple experiments with plasticine. Geology, 10, 611– 616. Tripathi, C. & Chowdhury, R. J. 1983. Gondwanas of Arunachal Himalaya. Himalayan Geology, 11, 73–90. Tripathi, C., Dungrakoti, B. D. & Ghosh, R. N. 1979. Note on discovery of Nummulites from Dihang valley, Siang district, Arunachal Pradesh. Indian Minerals, 33, 43–44. Tripathi, C., Ghosh, R. N., Gupta, P. D., Malhotra, G. & Dungrakoti, B. D. 1981a. Foraminifera from the Siang District, Arunachal Pradesh. In: Sinha, A. K. (ed.) Contemporary Geoscientific Researches in Himalaya. Bishan Singh Mahendra Pal Singh, Dehra Dun, 1, 231–242.

THE SIANG WINDOW OF ARUNACHAL PRADESH Tripathi, C., Chowdhury, R. J. & Das, D. P. 1981b. Discovery of Tertiary plant fossils from Geku Formation of Dihang valley, Siang district, Arunachal Pradesh. In: Sinha, A. K. (ed.) Contemporary Geoscientific Researches in Himalaya. Bishen Singh Mahendra Pal Singh, Dehra Dun, 1, 225–230. Tripathi, C., Gaur, R. K. & Singh, S. 1988. A note on the occurrence of Nummulitic limestone in East Siang district, Arunachal Pradesh. Indian Minerals, 35, 36–38. Tucker, P. & O’Reilly, W. 1980. The laboratory simulation of deuteric oxidation of titanomagnetites: effect on magnetic properties and stability of thermoremanence. Physics of the Earth and Planetary Interiors, 23, 112–133. Wemmer, K. 1991. K– Ar-Altersdatierungsmo¨glichkeiten fu¨r retrograde Deformationsprozesse im spro¨den und duktilen Bereich – Beispiele aus der KTB-Vorbohrung (Oberpfalz) und dem Bereich der Insubrischen Linie (N-Italien). Go¨ttinger Arbeitskreis Geologie/Pala¨ontologie, 51, 1 –61, Go¨ttingen.

97

Yang, Z., Yin, J., Sun, Z., Otofuji, Y. & Sato, K. 2001. Discrepant Cretaceous paleomagnetic poles between Eastern China and Indochina: a consequence of the extrusion of Indochina. Tectonophysics, 334, 101– 113. Yin, A. 2006. Cenozoic tectonic evolution of the Himalayan orogen as constraint by along-strike variation of structural geometry, exhumation history, and foreland sedimentation. Earth Science Reviews, 76, 1– 131. Yin, A. & Harrison, M. 2000. Geologic evolution of the Himalayan–Tibetan Orogen. Annual Review of the Earth and Planetary Sciences, 28, 211–280. Yin, A., Dubey, C. S., Kelty, T. K., Gehrels, G. E., Chou, C. Y., Grove, M. & Lovera, O. 2006. Structural evolution of the Arunachal Himalaya and implications for asymmetric development of the Himalayan orogen. Current Science, 90, 195–206. Zhang, P.-Z., Shen, Z., Wang, M., Gan, W., Bu¨rgmann, R. & Molnar, P. 2004. Continuous deformation of the Tibetan Plateau from global positioning system data. Geology, 32, 809–812.

Discontinuous low-velocity zones in southern Tibet question the viability of the channel flow model ˆ ME VERGNE2,3, LAURENT BOLLINGER4 & ¨ RGY HETE´NYI1,2*, JE´RO GYO RODOLPHE CATTIN2,5 1

Department of Earth Sciences, ETH Zu¨rich, Zu¨rich, Switzerland

2

Laboratoire de Ge´ologie, Ecole Normale Supe´rieure – CNRS UMR8538, Paris, France 3

Ecole et Observatoire des Sciences de la Terre, Strasbourg, France

4

Laboratoire de De´tection et de Ge´ophysique, CEA, Bruye`res-le-Chaˆtel, France

5

Laboratoire Ge´osciences Montpellier, Universite´ Montpellier 2, Montpellier, France *Corresponding author (e-mail: [email protected]) Abstract: Low-velocity zones (‘bright spots’) imaged by the INDEPTH seismic experiment in southern Tibet are extensively interpreted as widespread partial melt within the crust, which has given a strong support for the channel flow model. These suggest that a continuous seismic lowvelocity zone underlies Tibet on the large scale. Here we take advantage of the Hi-CLIMB seismic experiment which includes a dense south–north profile and a lateral 2D seismic network to assess the vertical and the horizontal extension of low-velocity zones in southern Tibet. Several approaches including migration, amplitude analysis and waveform inversion of receiver functions are performed to detect crustal low-velocity zones using this new seismological dataset. Our results reveal localized and discontinuous low-velocity zones in Tibet. They indicate that the vertical extension of the low-velocity zones is about 10 km, and their maximum horizontal length appears to be c. 50 km. Our study suggests a partial correlation between the location of these low-velocity zones and the spatial distribution of Tibetan grabens. These results, especially the non-continuity of low-velocity zones, together with the observed regular value of mean crustal VP/VS ratio, question the existence of widespread partial melt of the southern Tibetan crust and, therefore, the viability of the channel flow model.

The Tibetan Plateau and the Himalayas are the results of the well-known ongoing convergence between the Indian and Eurasian plates (e.g. Molnar et al. 1973; Molnar & Tapponnier 1975; Patriat & Achache 1984; Molnar & Stock 2009). However, the accommodation of this convergence, especially the localization of the deformation within the Tibetan Plateau is still a matter of debate. Two end-member models are commonly proposed. On the one hand most of the deformation occurs along the active faults that bound the major rigid Tibetan blocks (e.g. Avouac & Tapponnier 1993). On the other hand the convergence rate is mostly accommodated by internal deformation within the blocks related to the viscous behaviour of the continental lithosphere (e.g. Houseman & England 1993; England & Molnar 1997). Global Positioning System measurements are key data to discuss these two models. However, as previously mentioned by Thatcher (2007), due to the nonhomogeneous GPS data coverage (e.g. Zhang et al. 2004), the distinction between block and continuum models depends upon the scale of the

study area. A complementary approach is the use of heat flow data to constrain the rheological behaviour of the crust according to its thermal structure. These measurements give a high Tibetan heat flow (between 61– 319 mW m22 in southern Tibet), which favours viscous behaviour and thus may support the continuum model. However, heat flow data are only available at a few locations on the plateau (Fig. 1), and only a few of them are considered to be of good quality (e.g. Pollack et al. 1993; Fig. 1). Thus the analysis of heat flow gives only spatially limited constraints on the rheological behaviour of the Tibetan crust. A major contribution to our knowledge of the physical properties of the plateau in southern Tibet comes from the passive and active seismological experiments, like the INDEPTH projects (International Deep Profiling of Tibet and the Himalayas; Zhao et al. 1993; Nelson et al. 1996; Haines et al. 2003). Active seismic surveys conducted during INDEPTH II in the Yadong-Gulu rift (Fig. 1) show unusually large negative amplitudes in near vertical reflections (i.e. larger than 10 dB above

From: Gloaguen, R. & Ratschbacher, L. (eds) Growth and Collapse of the Tibetan Plateau. Geological Society, London, Special Publications, 353, 99– 108. DOI: 10.1144/SP353.6 0305-8719/11/$15.00 # The Geological Society of London 2011.

100 G. HETENYI ET AL. Fig. 1. Topographic map of southern Tibet, with an overlay of the original map from Armijo et al. (1986, 13 864, Fig. 28) in black: their map shows Quaternary normal fault traces (solid lines) bounding grabens, the associated depressions (horizontally striped areas), and lakes (horizontally striped areas with white background). Geo-referencing of the maps was performed using river and lake shapes in the central area. Roman numerals point to seven main rift systems discussed in Armijo et al. (1986), number VI being the Yadong-Gulu rift. Blue circles and diamonds mark broadband seismological stations from the Hi-CLIMB experiment (phase 2), from the main array and from the lateral deployment, respectively. Yellow line shows the location of the migrated profile on Figure 2. Heat flow data and quality (squares) are from Pollack et al. (1993). YTS, Yarlung-Tsangpo Suture. MBT, Main Boundary Thrust. The location of Figures 2 and 5 are shown for reference. Inset shows the location of this map in Asia.

NO WIDESPREAD PARTIAL MELT IN TIBET

101

Fig. 2. Map of main array stations and projected profile, with Quaternary normal fault traces, depressions and lakes in red (see Fig. 1 for symbol explanation and location of this map). Note that the map has been rotated hence north is to the right. The profile of migrated P-to-S receiver functions (projected on to the north–south-trending yellow line at 858E) is from the Hi-CLIMB main array (Hete´nyi 2007; Na´beˇlek et al. 2009; Hete´nyi et al. 2010). The profile shows localized blue-over-yellow (positive-over-negative) regions that are interpreted as low-velocity zones in the upper crust (highlighted by dashed green ellipses). The blue (negative) zone beneath them is the trace of the Main Himalayan detachment (see references above). Note that the graben at station H108 (Tak Kyel Co graben) as well as the one around 31.78N (Dawa Co graben) are located above a low-velocity zone.

background) (Brown et al. 1996), and strong P-to-S converted amplitudes in three-component wideangle reflections (Makovsky et al. 1996). The converters that generate these unusually strong signals have been named ‘bright spots’. Moreover, their location fits with the top of mid-crustal low-velocity layers deduced from the analysis of receiver functions (RF) at broad-band stations installed along the same profile (Kind et al. 1996). The tops of these low-velocity zones (LVZ) are usually modelled as a solid-fluid interface at c. 15 km depth (Makovsky et al. 1996). The nature of the crustal fluids is mostly given to be a granitic magma (e.g. Brown et al. 1996), while some suspect the presence of free aqueous fluids as well (Makovsky & Klemperer 1999). Magnetotelluric studies, performed during the INDEPTH II experiment (Chen et al. 1996) and during other measurements elsewhere on the plateau (Wei et al. 2001; Unsworth et al. 2005), indicate the existence of an electrically conductive zone in the southern Tibetan crust that also suggests the presence of an interconnected fluid phase. However, despite some attempts to

quantitatively describe these conductive features (e.g. Li et al. 2003), there is no consensus on their origin (zones of partial melt or of aqueous fluid accumulation), nor on their thickness. In summary, geophysical observations show that in southern Tibet there exist zones with significant seismic P-wave velocity decrease in the upper crust (at c. 15 km depth) that coincide with the top of an electrically conductive zone. The ‘classical’ and most extensively held interpretation of these geophysical observations is that a thick and widespread partially melted layer exists in the southern Tibetan crust (e.g. Nelson et al. 1996; Brown et al. 1996; Kind et al. 1996; Wei et al. 2001; Li et al. 2003; Unsworth et al. 2005). The view that this layer behaves as a fluid on the scale of Himalayan deformation gives strong support to a model in which the deformation of Tibet is driven by a channel-like viscous flow within the crust. This model is supported by arguments from field observations (e.g. rift flank topography in Masek et al. 1994) and numerical models (e.g. Royden 1996) and, since then, it

102

G. HETENYI ET AL.

became a popular mechanism to explain the evolution of deformation of the plateau (e.g. Clark & Royden 2000; Beaumont et al. 2001; Shen et al. 2001; Jamieson et al. 2004; Clark et al. 2005). A typical set of characteristics of a channel flow model is a viscosity reduced to the 1017 –1019 Pa s range, a channel thickness of 15 km, and an average flow speed of 80 mm a21 (Clark et al. 2005). Therefore, the expected seismic signature of such a feature can well be a continuous LVZ beneath the plateau with a pronounced velocity decrease on its top, as suggested by Nelson et al. (1996) and modelled by Hyndman & Shearer (1989). The limitations of the INDEPTH II results and therefore of the viability of the channel flow model come from the data coverage itself. Due to the practical difficulties to perform geophysical measurements in Tibet most INDEPTH II data were collected along the Yadong-Gulu rift, which may have specific lithospheric properties. A major drawback coming from the generalization of local INDEPTH II observations is therefore that it is uncertain whether partial melt is restricted to riftsystems, such as the Yadong-Gulu rift (Fig. 1), or occurs beneath all Tibet (Yin 2005). This study aims to assess whether regions of seismic wave velocity reduction are truly widespread across the southern Tibetan Plateau, or simply restricted to rift systems as the one where INDEPTH II data were collected. We take advantage of the Hi-CLIMB seismic experiment which includes a dense south –north profile and a lateral 2D seismic network in southern Tibet composed of stations located both in and away from individual grabens forming the rift-systems (see Fig. 1 for location and definition). First we apply passive seismic methods to assess the vertical and the horizontal extension of the LVZs in southern Tibet. Next, we compare our results to previous studies concerning the distribution and continuity of LVZs. Finally, we discuss the potential physical causes of LVZs, the relation of these to the rift systems, as well as the implications on the channel flow model and on the behaviour of the Tibetan crust.

Data processing Seismological data used in this study have been acquired in the frame of project Hi-CLIMB. Details concerning the Hi-CLIMB experiment, including deployment and data, can be found in Hete´nyi (2007), Na´beˇlek et al. (2009) and Hete´nyi et al. (2010). This broad-band seismology experiment was carried out to study the continental lithosphere and mountain building in the Himalaya–Tibet region. Here we focus on the second phase of the experiment, during which 113 stations were deployed for 15 months in southern and central Tibet (Fig. 1). The south –north-trending linear

profile near 858E has a station spacing of c. 8– 10 km, which allowed us not only to image the lower part of lithosphere at high resolution, but also to have sufficient ray-path crossings at shallow depth to study the upper crust in one continuous image. The lateral deployment covers the region between this array and the former location of INDEPTH stations in the Yadong-Gulu rift (Fig. 1). Here the station spacing is sparser (c. 35 km in average), but the spatial distribution of the stations is better than in INDEPTH, leading to a potentially better resolution of the lateral extent of the structures at depth. Also, having some of the stations in grabens and some of them away allows the examination of the effect of grabens on the Tibetan crustal structure. To study the structure of the crust below these stations we used receiver function analysis, a passive source method that detects waves from distant, teleseismic earthquakes that were converted from P- to S-wave on interfaces beneath the observing stations (Langston 1977). The converted waves arrive with a delay compared to the direct P-wave, and this delay is proportional to the depth of the interface where the conversion occurs, as well as to the velocity structure between the interface and the surface. The polarity of the converted P-to-S wave is positive when the impedance (product of seismic velocity and bulk density) increases with depth across the interface, and is negative when the impedance decreases. The amplitude of the converted P-to-S wave is proportional to the impedance change across the interface. Receiver functions are more suitable for detecting an eventual LVZ than active source seismology, since the illumination of structures is from below instead of above. Furthermore, the lower frequency content of RFs (typically 0.05 to c. 1–2 Hz) allows detection of the bottom of the LVZ as well, even when this is more likely to be a gradient than an interface. However, because of their lower frequency content, the vertical spatial resolution of the RFs will be relatively poorer. To detect and image LVZs within the crust we apply three different methods including: migration; amplitude analysis; and waveform inversion of receiver functions.

Migration This method is suitable when neighbouring stations are close enough to have incoming rays crossing at the depth of interest. This depth is about twice the spacing of the stations, so we apply this method to the dense linear array of stations only. The migration itself is a process that relocates the amplitude of conversions in the receiver function from time to space, and hence draws a continuous image of velocity contrasts along a profile beneath the stations. Here we use the common conversion

NO WIDESPREAD PARTIAL MELT IN TIBET

103

Here we take advantage of the proportionality between the amplitude of a converted wave on the RF and the impedance change across the interface that causes the conversion. Searching for LVZs, we look for the minimum amplitude of each individual RF between 1.0–2.5 s after the P-wave arrival. In case of an ordinary velocity-structure of the upper crust [VP ¼ 5.8 km s21, VS ¼ 3.36 km s21 in the iasp91 velocity-model (Kennett & Engdahl 1991)], this time range corresponds to conversions originating from interfaces between c. 8 and 20 km depth. Even with an eventual sedimentary cover or deviation in mean velocity that may further delay waves, we consider that this time frame is suitable to detect conversions from the upper crust. An example for a negative amplitude peak on a RF is shown on Figure 3b (left curve). This method is applied to the 2D lateral deployment, where the station spacing is too sparse to perform migration. Data selection is detailed in Appendix A.

L2-norm sense. Inversion requires good quality data and, depending on the method, considerable computation time. Here we apply two inversion approaches: a semi-linearized approach that is more likely to find local minima; and a more robust nonlinear method (for details on methods and processing, see Appendix B). The semi-linear inversion was applied to all stations on data selected as described above. At stations with the more stable results, the more time consuming non-linear inversion was also performed. Two examples of RF waveforms and inversion results, showing one station with, and one without, evidence for an upper crustal LVZ (that is, a decrease and then an increase of velocity with depth) are shown on Figure 3. The velocity –depth curves obtained by inversion are used first qualitatively and then quantitatively. The reasons for this are that the quality of the instruments and deployment sites in the lateral array was not as good as in the main array, and that the semi-linearized inversion method is not very robust. Thus we visually determine whether a LVZ in the upper crust is present or not in the velocity –depth curve obtained by inversion at each station, and identify bright spots when a VS decrease of at least 0.5 km s21 is apparent. The results of this selection are used together with the amplitude analysis, and are reported on Figure 4.

Waveform inversion

Spatial extent of low-velocity zones

The inversion of receiver functions aims to find a velocity –depth model that generates the synthetic RF closest in shape to the observed one in the

The obtained migrated profile of receiver functions across the Lhasa Block is shown on Figure 2. A few localized, strongly negative anomalies dominate the

point method (e.g. Dueker & Sheehan 1997) and processing as detailed in Na´beˇlek et al. (2009) with RFs filtered to 1 Hz maximum frequency. The resulting image is shown on Figure 2 and discussed in the next section.

Amplitude analysis

(a)

(c) 0

H108

20 Depth (km)

30 40

80

30 40

V 2

H159

70

20 5

10

60

10

H108

50

RF Data NA inv. Semi-lin.

S 3

4

H159 0

Waveforms 0

Depth (km)

Delay after P wave (s)

10

(b)

50 60 VS

70 3.5

4

4.5

Fig. 3. Observation of low-velocity zones by receiver functions (RF) in waveforms (b) and velocity– depth profiles obtained by inversion (a, c). Subfigures (a) and (c) show the best model obtained by semi-linearized (green) and non-linear (NA) (blue) inversions. Grey curves show the best 100 (best 1% of all tested) models from the non-linear inversion, and red the average of these. The examples depict one station (a, H108) with (red arrow) and one station (c, H159) without an underlying low-velocity zone (see Fig. 1 for their locations, and Fig. 2 for the location of station H108 on the migrated profile). See text for more discussion and Appendix B for more details on inversion.

G. HETENYI ET AL.

upper crust, which we identify as the seismic bright spots since: (1) they represent a significant drop in seismic velocity (we assume similar behaviour in VS as in VP); (2) they exhibit a velocity decrease with depth (negative amplitude); and (3) they are imaged in the same depth range as the bright spots seen by the INDEPTH experiment (i.e. c. 15 km deep). The negative anomalies are often underlain by a positive anomaly, usually of lower amplitude, and hence represent a gentler positive velocitygradient with depth. These features form the top and bottom of LVZs. Horizontally the LVZs have a maximum length of c. 50 km, which might be an apparent length as the LVZs may extend beyond the aperture of the imaging rays perpendicular to the profile. One can observe that LVZs lie close to north–south grabens (Fig. 2), especially between 29.58N and 308N (Tak Kyel Co graben) and around 31.78N (Dawa Co graben). There are also one or more LVZs between c. 30.6–31.28N, which are not related to a graben. Other zones, for example between 30–30.68N, have no significant velocity variations within the upper crust. The thickness determined for these three LVZs does not exceed 10 km. To increase the spatial coverage of our study, we look at the lateral stations, where the minimum amplitudes are combined with the analysis of the inversion results (Fig. 4): an LVZ-observation is represented in darker grey. This combination shows a good correlation between the presence of LVZs and the most strongly negative amplitude

(a)

anomalies. It allows the data to be split into two distinct groups (Fig. 4b). Although there is an overlap in the range of minimum amplitudes between the two sets, they form two distinct distributions. Stations where an LVZ is present possess a statistical peak at 20.050, and those where an LVZ is absent possess a peak at 20.015 (Fig. 4b). Finally, to investigate whether the observed LVZs are laterally continuous or not, the results are reported in map-view (Fig. 5). The main and most striking observation is that there is no continuous pattern in the distribution of LVZs. Thus independently from their physical cause, they do not form a coherent sheet or continuum beneath southern Tibet. By comparing the distribution of LVZs with that of extensional grabens, it can be argued that the loci of grabens are partly controlled by the presence of LVZs. A majority but not all (7 out of 9) stations located in individual grabens see an underlying LVZ. However, 3 out of 5 stations lying in the axis of main rift-systems but outside of mapped grabens (outlined in Armijo et al. 1986) are not underlain by an LVZ, showing that LVZs do not form a continuous feature along-strike of the riftsystems. There are LVZ observations outside of riftsystems (see next section), but 8 out of 13 stations located here have a regular velocity structure beneath them with no underlying LVZ. These observations can be interpreted together with the INDEPTH experiment: (1) stations located

ID number of lateral stations

18 19

8 13 910 11 14

1 2 4

16 15

0.00

26 25

36

27 23 2021

−0.02

37 38 −0.04

32 34 35

−0.06 −0.08 −0.10

31 LEGEND LVZ absent LVZ present

station No. histogram

8 28

28

−0.12 −0.14

RF minimum amplitude between 1.0-2.5s delay time

104

(b)

Frequency 0

1

2

3

4

0.00 −0.02 −0.04 −0.06 −0.08 −0.10 −0.12 −0.14

Fig. 4. Distribution of minimum amplitudes of receiver functions from lateral stations (see blue diamonds on Fig. 1 for locations) within the 1.0–2.5 s delay time range. This time window after the P-wave arrival is expected to detect upper-crustal velocity variations. LVZs seen by RF inversions are reported in colour code: observation of LVZ in dark grey shows correlation with more negative amplitudes, and their absence in light grey shows correlation with the less negative amplitudes. Numbers corresponding to stations with usable three-component data are reported one after another (a). The histogram (b) shows that the distribution of these amplitudes is non Gaussian, and corresponds to one group of stations without and one group with observed upper crustal LVZs.

NO WIDESPREAD PARTIAL MELT IN TIBET

105

Fig. 5. Low-velocity zone observations in southern Tibet. Earlier observations based on INDEPTH II data are shown in orange: lines (Makovsky et al. 1999), squares (Yuan et al. 1997), and circles (Mitra et al. 2005). Stations from the lateral deployment of the Hi-CLIMB experiment (diamonds) are reported with a colour code indicating whether an LVZ is observed or not. Rift systems, grabens and associated depressions, as well as lakes are from Armijo et al. (1986) (see Fig. 1 for symbol definition). See text for discussion and interpretation. Stations in grabens: 1, 4, 15, 23, 31, 32, 35, 36, 37. Stations in rift-systems but outside grabens: 14, 25, 26, 28, 38. Stations outside of rift-systems: 2, 8, 9, 10, 11, 13, 16, 18, 19, 20, 21, 27, 34.

in grabens are likely to detect underlying LVZs; (2) the additional spatial coverage given by the Hi-CLIMB lateral stations shows that LVZs are localized; and (3) at the scale of the study area, 400  200 km2, LVZs do not form a continuous crustal feature. Therefore it is more plausible that there is no generalized physical process in the crust, such as widespread partial melt. This is also supported by the observation that the average crustal VP/VS ratio throughout the sampled area in Tibet (Hete´nyi 2007; Na´beˇlek et al. 2009; Hete´nyi et al. 2010) shows values close to the worldwide average for continental areas (Zandt & Ammon 1995), and not high values expected for large amounts of partial melt (Watanabe 1993).

Potential nature of the fluids and the viability of channel flow The question of what causes the LVZs observed in Tibet and what are the respective amounts and roles of magmatic and aqueous fluids has been broadly discussed in the literature (e.g. Makovsky & Klemperer 1999; Li et al. 2003). Makovsky & Klemperer (1999) demonstrated that the observed INDEPTH II bright spot reflections are likely to be generated by 10% volume of free aqueous fluids in the Tibetan middle crust. One can infer that these free aqueous fluids may

come from metamorphic dehydration reactions at depth. A regional contribution from below the LVZs may be derived from underplating of the Indian lower crust and its eclogitization liberating fluids at depth around the Yarlung-Tsangpo suture and further north beneath the Lhasa block (Hete´nyi et al. 2007). If the free fluids are not all driven away by the overlying Main Himalayan thrust, they can migrate upwards toward the LVZs. Free aqueous fluids may also come from above, as infiltrations of meteoric waters along grabenbounding normal faults. Whatever the scenario, these aqueous fluids do not preclude the existence of magmatic fluids and may even help to initiate or to further enhance localized melting. However, for most authors, this process is a trivial addition to models dominated by the presence of magmatic fluids (e.g. Nelson et al. 1996). Although it has not been unequivocally demonstrated, the presence of magma bodies within the Tibetan crust coinciding with the LVZs is suggested, given the locally high heat flow and young volcanism at surface (e.g. Brown et al. 1996). In this study, we argue against thick and widespread partial melt; in the meantime, given the thickness of the crust (c. 80 km) and the thickness of the LVZs (c. 10 km), the possibility of having thin layers of molten rock in the crust with high VP/VS ratio cannot be ruled out. A quantitative characterization of the LVZs seems feasible by joint application of P-to-S and S-to-P receiver

106

G. HETENYI ET AL.

functions (Wittlinger et al. 2009) yielding an independent estimate for VP/VS ratio of the LVZs that would be characteristic of either aqueous or magmatic fluids (Watanabe 1993). This method at present implies potentially large errors due to the thinness of the LVZs, and is not in the scope of our study. The origin of aqueous fluids that enhance or initiate local partial melt within the crust can be speculated from our results and Figure 5. There are 4 stations in 4 grabens that do not detect an underlying LVZ (stations 14, 25, 36, 37), but in all of these grabens there is another station beneath which an LVZ can be observed (respectively stations 1, 28, 31, 32). This distribution is nonconclusive: it can either mean localized melting at depth followed by rifting initiation and rapid propagation away at surface, or a tectonic event followed by fluid circulation and localized partial melting. However, there are 5 stations in areas where no grabens have been previously mapped that detect an underlying LVZ (stations 2, 16, 20, 21 and 34): this favours a thermally driven rifting in southern Tibet. Summarizing our results from the channel flow perspective, our images show seismic LVZs that are of comparable thickness that channels in crustal flow models (c. 10 km scale). The cause of these LVZs can still be both aqueous fluids and partial melt: a 10-km thick layer with a few percent partial melt embedded in an 80-km thick crust does not significantly raise the average VP/ VS-ratio. In the meantime, what is clearly shown by the spatial distribution of LVZ observations is that a continuous sheet of partial melt beneath Tibet is highly implausible. Therefore the mechanical viability of the channel flow model as a governing force of the plateau’s general deformation is strongly questioned. Instead, both LVZs and rifting appear to be localized zones of weakness that help to maintain Tibet’s deformation broad on the crustal scale, and potentially contribute to the difficulty in distinguishing between block and continuum deformation models on the large scale.

Conclusions We have focused on the spatial distribution of LVZs in southern Tibet using new seismological data and various analyses of receiver functions. Our study has shown that these LVZs are limited to 10 km in thickness, and that they do not form a continuous layer beneath the plateau. Indeed, these LVZs occur mostly in and also away from extensional grabens, without any evident and common connectivity pattern. These results, as well as normal average crustal VP/VS-ratios, question the thick

and widespread partial melt scenario that was suggested to characterize the southern Tibetan crust, and which is the major pillar of channel flow models. Our study has pointed out the importance of denser spatial sampling of geophysical data. At the present stage of data coverage in Tibet, acquiring new heat-flow data, both in and away from grabens, would be of invaluable help to build a comprehensive picture of the thermal state of the lithosphere, and therefore of the geodynamics and growth of Tibet. We are grateful to the numerous people have who contributed to the data acquisition in the field, and to John Na´beˇlek for the seismology data. Many thanks to Richard Gloaguen and Lothar Ratschbacher for promptly editing this volume. We warmly thank the constructive criticism of reviewers Ian Watkinson and Michele Zucali, including the most thorough grammatical corrections we have ever seen in a review. Maps were generated with the Generic Mapping Tool (Wessel & Smith 1991). Feedback from several people following an ETH-seminar and a 2009 EGU-talk helped to focus the conclusions. This work was supported by Swiss National Science Foundation grant 200020107889.

Appendix A Data selection for RF amplitude analysis and inversion Receiver functions were calculated as indicated in Na´beˇlek et al. (2009) but with 2 Hz maximum frequency content. First an automatic, then a visual selection process was applied to keep the best quality traces at each station. The criteria included: (1) selection of M  6.4 events in the epicentral distance range of 30–958; (2) exclusion of numerous small aftershocks closely following large magnitude events; (3) exclusion of traces based on signalto-noise ratio of the original three-component data; and (4) automatic examination of the receiver function shape based on the time of its peak amplitude. As a result, 123 traces (between 1 and 16 at each station) were kept at the 27 lateral stations, and stacked for amplitude analysis. Further 10 stations of the lateral deployment did not produce good quality data in the analysed frequency band.

Appendix B Inversion methods and details Semi-linearized inversion (Ammon et al. 1990) tries a limited number of velocity models by modifying one or more initial models, the choice of which is crucial to the success of the inversion. The depth range to be modelled is divided into depth-slices of fixed thickness and constant velocity. The VP/VS ratio is fixed, which is a simplifying

NO WIDESPREAD PARTIAL MELT IN TIBET assumption, but effectively reduces the number of parameters to invert, and hence calculation costs. Thus the only parameter to recover is the shear-wave velocity in each layer. In order to stabilize the semi-linear inversions, a steplike resolution procedure was followed, starting from low frequencies (0.05 Hz) and thick layers (10 km), and ending with high frequencies (0.5 Hz) and thin layers (2.5 km). At each step, either the frequency was increased or the thickness was decreased. Finally, the highest frequency solution was low-pass filtered and compared to the RF filtered similarly, which allowed an estimate of the robustness of the procedure. The final high-frequency results are step-like velocity– depth curves, such as the ones in green on Figure 3. The drawback of this method is (1) the use of constant VP/VS ratio, which could imply an error on the shape of the velocity-structure, and (2) the strong dependency on the initially assumed model. However, as the average crustal VP/VS-ratio is in the regular range (Hete´nyi 2007; Na´beˇlek et al. 2009; Hete´nyi et al. 2010) this error is considered to be minor. For the non-linear (stochastic) inversion scheme, the Neighbourhood Algorithm (NA) was applied (Sambridge 1999). This technique has the advantage of: (1) varying both VP and VS, as well as the thickness of the layers; and (2) performing a more complete search in the parameter space, that in general helps to avoid falling in local minima. The number of layers has to be fixed in advance, and the range of variation for each parameter to invert has to be bound as well. It also allows the presence of velocity-gradients. The cost for these advantages is increased computation time compared to the semilinearized approach. In our study, the initial model was taken from the semilinear inversion result, and the VP/VS ratio was allowed to vary between 1.70– 1.90. The inversions have been run using 100 iterations with 100 velocity-models in each (a total of 10 000 models) to reproduce traces filtered up to 0.5 Hz. In the final results, the best 100 (best 1% of all tested) velocity-models (in grey), their average (red), and the best model (blue) are shown (Fig. 3). The velocity-models resulting from the inversions were not always similar between low and high frequencies, and/or between the semi-linearized scheme and the NA. Only at least partially coherent results were used for the interpretation concerning the velocity-structure, and hence the presence of low-velocity zones.

References Ammon, C. J., Randall, G. E. & Zandt, G. 1990. On the non-uniqueness of receiver function inversions. Journal of Geophysical Research, 95, 15 303 –15 318. Armijo, R., Tapponnier, P., Mercier, J. L. & Han, T. L. 1986. Quaternary extension in southern Tibet: field observations and tectonic implications. Journal of Geophysical Research, 91, 13803– 13872.

107

Avouac, J.-P. & Tapponnier, P. 1993. Kinematic model of active deformation in Central Asia. Geophysics Research Letters, 20, 895– 898. Beaumont, C., Jamieson, R. A., Nguyen, M. H. & Lee, B. 2001. Himalayan tectonics explained by extrusion of a low-viscosity crustal channel coupled to focused surface denudation. Nature, 414, 738– 742. Brown, L. D., Zhao, W. J. et al. 1996. Bright spots, structure, and magmatism in southern Tibet from INDEPTH seismic reflection profiling. Science, 274, 1688– 1690. Chen, L., Booker, J. R., Jones, A. G., Wu, N., Unsworth, M. J., Wei, W. & Tan, H. 1996. Electrically conductive crust in Southern Tibet from INDEPTH magnetotelluric surveying. Science, 274, 1694– 1695. Clark, M. K. & Royden, L. H. 2000. Topographic ooze: building of Tibet by lower crustal flow. Geology, 28, 703– 706. Clark, M. K., Bush, J. W. M. & Royden, L. H. 2005. Dynamic topography produced by lower crustal flow against rheological strength heterogeneities bordering the Tibetan Plateau. Geophysical Journal International, 162, 575–590. Dueker, K. & Sheehan, A. 1997. Mantle discontinuity structure from midpoint stacks of converted P to S waves across the Yellowstone hotspot track. Journal of Geophysical Research, 102, 8313– 8327. England, P. & Molnar, P. 1997. Active deformation of Asia: from kinematics to dynamics. Science, 278, 647– 650. Haines, S. S., Klemperer, S. L. et al. 2003. INDEPTH III seismic data: from surface observations to deep crustal processes in Tibet. Tectonics, 22, 1001, doi: 10.1029/2001TC001305, 2003. Hete´nyi, G. 2007. Evolution of deformation of the Himalayan prism: from imaging to modelling. PhD thesis, E´cole Normale Supe´rieure, Universite´ ParisSud XI. Hete´nyi, G., Cattin, R., Brunet, F., Vergne, J., Bollinger, L., Na´beˆlek, J. L. & Diament, M. 2007. Density distribution of the India plate beneath the Tibetan Plateau: geophysical and petrological constraints on the kinetics of lower-crustal eclogitization. Earth and Planetary Science Letters, 264, 226– 244, doi: 10.1016/j.epsl.2007.09.036. Hete´nyi, G., Cattin, R. & Vergne, J. 2010. Geodynamics of Tibet and the Himalayas – Geophysical advances from imaging and modelling. LAP Lambert Academy Publisher, Saarbrucken. ISBN 978-3-8383-2353-4, 1– 408. Houseman, G. & England, P. 1993. Crustal thickening v. lateral expulsion in the Indian– Asian continental collision. Journal of Geophysical Research, 98, 12233–12249. Hyndman, R. D. & Shearer, P. M. 1989. Water in the lower continental-crust: modeling magnetotelluric and seismic-reflection results. Geophysical Journal International, 98, 343 –365. Jamieson, R. A., Beaumont, C., Medvedev, S. & Nguyen, M. H. 2004. Crustal channel flows: 2. Numerical models with implications for metamorphism in the Himalayan– Tibetan orogen. Journal of Geophysical Research, 109, B06407.

108

G. HETENYI ET AL.

Kennett, B. L. N. & Engdahl, E. R. 1991. Traveltimes for global earthquake location and phase identification. Geophysical Journal International, 105, 429– 465. Kind, R., Ni, J. et al. 1996. Evidence from earthquake data for partially molten crustal layer in southern Tibet. Science, 274, 1692–1694. Langston, C. 1977. Corvallis, oregon, crustal and upper mantle receiver structure from teleseismic P-waves and S-waves. Bulletin of the Seismological Society of America, 67, 713– 724. Li, S. H., Unsworth, M. J., Booker, J. R., Wei, W. B., Tan, H. D. & Jones, A. G. 2003. Partial melt or aqueous fluid in the mid-crust of southern Tibet? Constraints from INDEPTH magnetotelluric data. Geophysical Journal International, 153, 289–304. Makovsky, Y. & Klemperer, S. L. 1999. Measuring the seismic properties of the Tibetan bright spots: evidence for free aqueous fluids in the Tibetan middle crust. Journal of Geophysical Research, 104, 10795– 10825. Makovsky, Y., Klemperer, S. L., Ratschbacher, L., Brown, L. D., Li, M., Zhao, W. J. & Meng, F. L. 1996. INDEPTH wide-angle reflection observation of P-wave-to-S-wave conversion from crustal bright spots in Tibet. Science, 274, 1690– 1691. Masek, J. G., Isacks, B. L., Gubbels, T. L., Fielding, E. J. & Browaeys, J. 1994. Erosion and tectonics at the margins of continental plateaus. Journal of Geophysical Research, 99, 13941–13956. Mitra, S., Priestley, K., Bhattacharyya, A. K. & Gaur, V. K. 2005. Crustal structure and earthquake focal depths beneath northeastern India and southern Tibet. Geophysical Journal International, 160, 227– 248. Molnar, P. & Stock, J. M. 2009. Slowing of India’s convergence with Eurasia since 20 Ma and its implications for Tibetan mantle dynamics. Tectonics, 28, TC3001, doi: 10.1029/2008TC002271. Molnar, P. & Tapponnier, P. 1975. Cenozoic tectonics of Asia – effects of a continental collision. Science, 189, 419– 426. Molnar, P., Fitch, T. J. & Wu, F. T. 1973. Fault plane solutions of shallow earthquakes and contemporary tectonics in Asia. Earth and Planetary Science Letters, 19, 101–112. Na´beˇlek, J., Hete´nyi, G. etal. & THE HI-CLIMB TEAM 2009. Underplating in the Himalaya-Tibet collision zone revealed by the Hi-CLIMB experiment. Science, 325, 1371– 1374, doi: 10.1126/science.1167719. Nelson, K. D., Zhao, W. J. et al. 1996. Partially molten middle crust beneath southern Tibet: synthesis of project INDEPTH results. Science, 274, 1684– 1688. Patriat, P. & Achache, J. 1984. India –Eurasia collision chronology has implications for crustal shortening and driving mechanism of plates. Nature, 311, 615–621. Pollack, H. N., Hurter, S. J. & Johnson, J. R. 1993. Heat flow from the Earth’s interior: analysis of

the global data set. Reviews of Geophysics, 31, 267–280. Royden, L. 1996. Coupling and decoupling of crust and mantle in convergent orogens: implications for strain partitioning in the crust. Journal of Geophysical Research, 101, 17679– 17705. Sambridge, M. 1999. Geophysical inversion with a neighbourhood algorithm – I. Searching a parameter space. Geophysical Journal International, 138, 479–494. Shen, F., Royden, L. H. & Burchfiel, B. C. 2001. Large-scale crustal deformation of the Tibetan Plateau. Journal of Geophysical Research, 106, 6793– 6816. Thatcher, W. 2007. Microplate model for the present-day deformation of Tibet. Journal of Geophysical Research, 112, B01401, doi: 10.1029/2005JB004244. Unsworth, M. J., Jones, A. G., Wei, W., Marquis, G., Gokarn, S. G. & Spratt, J. E. 2005. Crustal rheology of the Himalaya and southern Tibet inferred from magnetotelluric data. Nature, 438, 78–81. Watanabe, T. 1993. Effects of water and melt on seismic velocities and their application to characterization of seismic reflectors. Geophysical Research Letters, 20, 2933– 2936. Wei, W. B., Unsworth, M. et al. 2001. Detection of widespread fluids in the Tibetan crust by magnetotelluric studies. Science, 292, 716– 718. Wessel, P. & Smith, W. 1991. Free software helps map and display data. EoS Transactions American Geophysical Union, 72, 441 and 445– 446. Wittlinger, G., Farra, V., Hete´nyi, G., Vergne, J. & Na´beˇlek, J. 2009. Seismic velocities in Southern Tibet lower crust. A receiver function approach for eclogite detection. Geophysical Journal International, 177, 1037–1049, doi: 10.1111/j.1365246X.2008.04084.x. Yin, A. 2005. Cenozoic tectonic evolution of the Himalayan orogen as constrained by along-strike variation of structural geometry, exhumation history, and foreland sedimentation. Earth-Science Reviews, 76, 1 –131. Yuan, X., Ni, J., Kind, R., Mechie, J. & Sandvol, E. 1997. Lithospheric and upper mantle structure of southern Tibet from a seismological passive source experiment. Journal of Geophysical Research, 102, 27491–27500. Zandt, G. & Ammon, C. J. 1995. Continental crust composition constrained by measurements of crustal Poisson’s ration. Nature, 374, 152– 154. Zhang, P. Z., Shen, Z., Wang, M., Gan, W. J., Burgmann, R. & Molnar, P. 2004. Continuous deformation of the Tibetan Plateau from global positioning system data. Geology, 32, 809– 812. Zhao, W. J., Nelson, K. D. & PROJECT INDEPTH TEAM. 1993. Deep seismic-reflection evidence for continental underthrusting beneath southern Tibet. Nature, 366, 557– 559.

The seismological structure of the Tibetan Plateau crust and mantle down to 700 km depth J. MECHIE*, R. KIND & J. SAUL Deutsches GeoForschungsZentrum – GFZ, Sections ‘Geophysical Deep Sounding’ and ‘Seismology’, Telegrafenberg, 14473 Potsdam, Germany *Corresponding author (e-mail: [email protected]) Abstract: A seismic velocity cross-section down to 700 km depth beneath the Tibetan Plateau has been constructed. Beneath the cover layer, felsic rocks rich in a quartz exist down to 15– 25 km depth. Beneath these depths, temperatures are probably high enough for ductile flow and partial melting to occur. The velocity increase across the boundary at 30– 40 km depth marks the interface between felsic upper crust and more mafic lower crust. Crustal thickness is greatest (c. 74 km) south of c. 31.58N, where Indian lower crust forms the basal layer. Northwards, crustal thickness decreases to c. 66 km around 338N, before increasing to c. 70 km beneath northern Tibet. Crossing the Kunlun, the crust thins to c. 54 km beneath the Qaidam basin. High-velocity, dense, cold Indian lithospheric mantle extends northwards until about the Banggong-Nujiang suture, where it downwells to 350–400 km depth. The lithosphere –asthenosphere boundary occurs at 160– 225 km depth. The apparent northwards deepening of the 410 and 660 km discontinuities implies that the upper mantle beneath northern Tibet is slower, less dense and warmer than under southern Tibet which, in turn, could provide some of the isostatic support for the high elevations in northern Tibet where the crust is thinner than under southern Tibet.

An overview of both controlled and passive source seismic studies on the Tibetan Plateau up to 2004, including some excellent location maps, has been provided by Gao et al. (2005). Controlled source seismic studies on the Tibetan Plateau began in 1958 with the recording of a near-vertical incidence reflection test profile by Chinese scientists in the Qaidam basin in the northeastern corner of the plateau. The first controlled source study on the high part of the plateau, including elevations above 4000 m, was a wide-angle reflection/refraction profile completed by Chinese scientists in 1977 between Yadong and Damxung across the IndusYarlung suture in southern Tibet (Fig. 1, Teng et al. 1983; Yin et al. 1990). The Sino-French set of wide-angle reflection/refraction profiles in southern Tibet in 1981, was the first international programme involving controlled source studies on the plateau (Hirn et al. 1984a, b). This group completed a further phase of controlled source studies in 1998 in the northeastern part of the plateau (Galve´ et al. 2002a). The other major international programme involving controlled source studies on the plateau is INDEPTH (INternational DEep Profiling of Tibet and the Himalaya), which to date has completed four phases of fieldwork across the plateau (Zhao et al. 1993, 2001, 2008; Nelson et al. 1996). The thick, often hot crust of the Tibetan Plateau presents a challenge for obtaining high quality

controlled source seismic data. A rather typical data example, from phase three of the INDEPTH project, shows a record section from a wide-angle reflection/refraction profile on the plateau (Fig. 2a). This record section displays the vertical component of ground motion in which the compressional (P) wavefield can mainly be seen. Despite the rather large station spacing of 5–10 km, several phases can be recognized (Zhao et al. 2001). In wide-angle reflection/refraction profiling, the 60–80 km thick crust beneath the plateau means that the critical distance corresponding to the angle of critical refraction and beyond which the Moho reflection, PmP, can be recognized due to its high amplitudes, generally exceeds 200 km (Fig. 2a), whereas the crossover distance, beyond which the uppermost mantle refraction, Pn, can be observed, generally exceeds 300 km. As a consequence, wide-angle reflection/refraction profiles in Tibet have to be 400–500 km long which, in turn, means that large shots need to be executed in order that the seismic energy carries such long distances. The often hot crust means that the seismic quality factor, Q, is often low (e.g. McNamara et al. 1996; Reese & Ni 1996; Phillips et al. 2000; Xie et al. 2004), which again means that shots with large charge sizes are required. On the other hand, as there is little industry on the Tibetan Plateau, cultural noise due to humanbeings is generally low as long as one is far enough away from the Lhasa to Golmud highway and

From: Gloaguen, R. & Ratschbacher, L. (eds) Growth and Collapse of the Tibetan Plateau. Geological Society, London, Special Publications, 353, 109–125. DOI: 10.1144/SP353.7 0305-8719/11/$15.00 # The Geological Society of London 2011.

110

J. MECHIE ET AL. 85˚

40˚ height in metres 9000

90˚

100˚

95˚

40˚

ATF

Tarim basin

NQF Qilian belt

7500

35˚

H Qaidam basin

6000 5500 5000 4500 4000

G

KF

NKT P NKF

SGT

JS

35˚

SKF

F

R

J

S

Qiangtang terrane L

T 0

I

B

E

T

D BNS

E

C

N Lhasa terrane

30˚

30˚

K Lhasa

A B

IYS

Tethyan belt MCT

M

HHC MFT

85˚

90˚

95˚

100˚

Fig. 1. Location map of controlled source (CS) and passive source (PS) experiments in Tibet, the results of which are used or discussed in this study. Black lines show profiles which have a CS component. Dashed black lines show CS fan profiles. Red lines show profiles which have only a PS component. Dashed red lines show the positions of cross-sections through regional P-wave or surface wave tomography models. Red dots show the locations of the permanent station at Lhasa as well as the stations of the Sino-American Tibetan Plateau broadband experiment used in this study (Owens et al. 1993; Owens & Zandt 1997). The red diamond marks the position of where lower crustal xenoliths were found during the third phase of the INDEPTH project (Hacker et al. 2000). Key: A, Yadong–Nam Tso CS profile & INDEPTH I þ II CS and PS profiles (Teng et al. 1983; Yin et al. 1990; Zhao et al. 1993; Nelson et al. 1996); B, Sino-French CS profiles in S. Tibet, including fan profiles shown as dashed lines (Hirn et al. 1984a); C, Sino-French CS profile (Sapin et al. 1985; Zhang & Klemperer 2005); D, INDEPTH III CS and PS profile (Zhao et al. 2001; Mechie et al. 2004); E, Coqen– Lugu CS profile (Kong et al. 1996); F, Tuotuohe– Golmud CS profile (Lu & Wang 1990); G, Golmud– Ejin Qi CS profile (Cui et al. 1995; Gao et al. 1995); H, Tarim– Qaidam CS profile (Zhao et al. 2006); J, Sino-French CS and PS profiles in NE Tibet, including fan profile (Galve´ et al. 2002a, 2006; Jiang et al. 2006); K, Sino-French PS profile in S. Tibet (Hirn et al. 1995); L, Sino-French PS profile in N. Tibet (Herquel et al. 1995; Wittlinger et al. 1996); M, HIMNT PS experiment (Schulte-Pelkum et al. 2005); N, Hi-CLIMB PS profile (Wittlinger et al. 2009); P, pilot experiment of Sino-U.S. northeastern Tibetan Plateau seismic (NETS) project (Shi et al. 2009); R, cross-section through regional P-wave tomography model (Li et al. 2008); S, cross-section through regional surface wave tomography model (Friederich 2003); MFT, Main Frontal Thrust; MCT, Main Central Thrust; HHC, High Himalayan Crystallines; IYS, Indus-Yarlung suture; BNS, Banggong-Nujiang Suture; JS, Jinsha Suture; SGT, Songpan-Ganz Terrane; KF, Kunlun Fault; NKF, North Kunlun Fault; SKF, South Kunlun Fault; NKT, North Kunlun Thrust; NQF, North Qaidam Fault; ATF, Altyn Tagh Fault.

railway line. The often low Q also means that shear (S) waves are often difficult to record with controlled sources. A notable exception on the plateau is in northeastern Tibet where the Sino-French group recorded good S-waves (Galve´ et al. 2002a).

Fortunately, local earthquakes occur quite frequently on the plateau and data from such earthquakes, especially if they occur close to a recording profile, can be used to supplement the controlled source data. Again, a data example from phase three of

SEISMOLOGICAL STRUCTURE OF TIBET

(a)

111

12 11 10

PmP

Time-Distance/7.0 [s]

9 8

Pi2

Pi2P

7 6

Pi1P

5 4 3

Pg

2 1 0 –350

–300

–250

–200

–150

–100

–50

0

Distance [km]

N

S

Time-Distance/4.041 [s]

(b) 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

SmS

Sn

Sg

Sg

–50

N

0

50

100

150

Distance [km]

200

250

300

S

Fig. 2. (a) Controlled source seismic data from a shot along the INDEPTH III profile. The record section reduced with a velocity of 7 km s21 shows the vertical component of P-wave motion in which each trace is normalized individually and bandpass filtered from 2 to 12 Hz. Key: Pg, first arrival refraction through the upper crust; Pi1P, reflection from the boundary in the upper crust at 20– 30 km depth; Pi2P, reflection from the top of the lower crust; Pi2, first arrival refraction through the lower crust; PmP, reflection from the Moho. (b) Seismic data from a nearby earthquake recorded along the INDEPTH III profile. The record section reduced with a velocity of 4.04 km s21 shows the north– south horizontal (approximately radial) component of S-wave motion in which each trace is normalized individually. Key: Sg, refraction through the upper crust; SmS, reflection from the Moho; Sn, refraction through the uppermost mantle.

the INDEPTH project is shown (Fig. 2b). In this case, the record section displays the north –south horizontal (approximately radial) component of ground motion from a nearby earthquake. In this section, the S wavefield can mainly be seen (Mechie et al. 2004). Recording of earthquakes with broadband seismometers began on the Tibetan Plateau in 1991 with the Sino-American Tibetan Plateau broadband experiment which lasted until 1992 (Owens et al.

1993; Owens & Zandt 1997) and the start of recording at the permanent station LSA in Lhasa (Fig. 1). In 1992 the Sino-French group began temporary passive source projects on the plateau with an experiment from the Lesser Himalayas in Nepal to the Qiangtang terrane (Hirn et al. 1995). This group continued with experiments in 1993 in the northern part of the plateau from the Qiangtang terrane to the Qaidam basin (Herquel et al. 1995; Wittlinger et al. 1996), in 1995 across the Altyn Tagh fault from the

112

J. MECHIE ET AL.

northern Qaidam basin to the Tarim basin (Herquel et al. 1999), in 1998 in northeastern Tibet from the Jinsha suture to the Gonghe basin along strike to the east from the Qaidam basin (Vergne et al. 2002) and finally in 2001 in western Tibet from the Lhasa terrane to the Tarim basin (Wittlinger et al. 2004). Project INDEPTH has incorporated passive source recordings in phases two to four of its programme (Kind et al. 1996; Huang et al. 2000; Zhao et al. 2008). Another major international collaboration between scientists from China, Nepal, USA and France was the Hi-CLIMB experiment which ran from 2002–2005 with three phases of deployment covering the region from the Ganges foreland in Nepal to the Qiangtang terrane in central Tibet (Na´beˇlek et al. 2009; Wittlinger et al. 2009). The above described experiments, except for the fourth phase of the INDEPTH project, mainly consisted of linear arrays with only a few offline stations. As such they are good for receiver function profiling. Rather typical receiver function examples, again from the third phase of the INDEPTH project, are shown (Fig. 3). In these examples, the Moho between 60 and 70 km depth can be easily recognized (Fig. 3a) as well as the 410 and 660 km discontinuities marking the top and bottom of the mantle transition zone (Fig. 3b). Experiments which had a significant number of stations distributed in a 2D array include a Chinese experiment crossing the northwestern margin of the Tibetan Plateau from 1998–1999 (Kao et al. 2001), the HIMNT experiment extending from the Ganges plains in Nepal to the Indus-Yarlung suture in southern Tibet from 2001 –2003 (Schulte-Pelkum et al. 2005), two experiments from 2003–2004 around the eastern syntaxis and crossing the SE margin of the plateau (Sol et al. 2007) and the fourth phase of the INDEPTH project from 2007– 2009 in northern Tibet (Zhao et al. 2008). With their areal distribution of stations these experiments are good for locating local earthquakes and for 3-D tomography studies.

Crustal structure Much of our knowledge about the seismological structure of Tibet comes from the transect along the Lhasa to Golmud highway, as this is logistically the easiest part to access. Many controlled and passive source experiments have been carried out along this transect and one of the main aims of this study is to collate the information from all these experiments and construct a cross-section of the crust and upper mantle across the Tibetan Plateau. In constructing the cross-section, information has also been taken from studies up to 400 km to the west and east of the main Lhasa to

Golmud transect. Information from studies in far western Tibet (e.g. Wittlinger et al. 2004) and across the northern and eastern plateau margins has not been included in the cross-section. The information included in the present study has been gleaned from the results of studies published up until October 2009. The crustal part of this crosssection (Fig. 4) presents an update and extension of the cross-section of Moho depths across the Tibetan Plateau presented by Zhao et al. (2001). The present cross-section shows the Moho dipping down to the north beneath southernmost Tibet and then flattening out at a depth of c. 74 km below the surface beneath the northern part of the Tethyan belt and the southern Lhasa terrane. It then shallows by c. 8 km beneath the northern Lhasa terrane and the southern Qiantang terrane before deepening slightly by c. 4 km beneath the northern Qiangtang and Songpan-Ganz terranes to the vicinity of the Kunlun fault. This variable but smooth course of the Moho is similar to that presented by Zhao et al. (2001) and also that of Kind et al. (2002) based on a receiver function image. It also shows similarity to the distribution of Moho depths obtained from an inversion of gravity data by Braitenberg et al. (2000). An alternative view of the Moho structure of southern Tibet was presented by Galve´ et al. (2002b) based on controlled source, wide-angle fan recordings and a receiver function image derived from data gathered by the Sino-French group alone. In this section, Moho depths between 28 – 338N vary from 50 –80 km. In addition, the section shows much more abrupt changes in Moho depths than the section shown here and that shown by Kind et al. (2002), who also incorporated the data gathered by the Sino-French group in creating their receiver function image. To some extent the difference in the receiver function images of Galve´ et al. (2002b) and Kind et al. (2002) may lie in the way the data are processed. There is also the question of whether or not there is a significant, abrupt change in crustal thickness in the vicinity of the Jinsha suture. An abrupt change of c. 20 km in Moho depth in the vicinity of the Jinsha suture was first proposed by Herquel et al. (1995). Two of the studies utilized here show an abrupt shallowing of the Moho by c. 10 km to the north, while the other two studies used here, show no significant, abrupt step. One of the studies which shows an abrupt shallowing is based on controlled source, wide-angle, mainly fan recordings c. 400 km to the east of the main Lhasa to Golmud transect (Jiang et al. 2006) while the other is based on a receiver function image along the main Lhasa to Golmud transect (Vergne et al. 2002). One of the studies that shows no significant, abrupt step in the vicinity of the Jinsha suture is that of Kind et al.

SEISMOLOGICAL STRUCTURE OF TIBET

(a)

113

0 20 40

Depth [km]

60 80 100 120 140 160 180 200 30.5

31.0

31.5

32.0

32.5

33.0

33.5

34.0

Latitude [°N]

35 40

350

45

400

50

450

55

500

60

550

Depth [km]

Delay time [s]

(b)

600

65

650

70

700

75

750

80 30

31

32

33

34

Latitude [°N] Fig. 3. (a) Receiver functions for the crust and uppermost mantle along the INDEPTH III profile (after Saul 2003). Red and yellow colours indicate positive amplitudes whereas blue and green colours indicate negative amplitudes. The model used to derive the common conversion point depth section was an average of the P-wave velocity model of Zhao et al. (2001), in conjunction with a variable Poisson’s ratio determined from the conversions and their multiples. The width of the horizontal smoothing operator (triangle below the section) is 20 km. (b) Receiver functions for the mantle along the INDEPTH III profile, displayed in the form of a common conversion point time section (after Saul 2003). Red and yellow colours indicate positive amplitudes whereas blue and green colours indicate negative amplitudes. The depth scale to the right of the section has been derived using the iasp91 model (Kennett & Engdahl 1991) and correcting for the effects of the thick crust and high topography beneath the Tibetan Plateau. The width of the horizontal smoothing operator is 120 km.

114

J. MECHIE ET AL. Lhasa terrane

HHC STD Tethyan belt IYS

BNS

JS SGT KF

Qiangtang terrane

NKT

QB

NQF

Qilian belt

Latitude (°N) 28

29

30

31

32

33

34

35

36

37

38

0

0 E

6.0 −20

R

YZR D

6.37

Depth (km)

−40

5.9, 0.21 ABQT E

6.3

6.2

5.9−6.1

R 6.6

7.25 7.0, 0.29 F

C

ILC P

N

−70 A −80

B

6.4

6.2

6.0

6.0

6.1

6.3

6.65 L

6.7 M

J

6.7

−50 −60

H K U

G

−30 −40

6.7 V

TLC 6.3, 0.25 6.8 6.9, 0.20

−70

K

S

−80

N

P

−20

L

K

6.26 E

6.1

6.3

K

6.1−6.3 6.8, 0.29

6.6

6.05 6.0

6.4

S

−60

6.26

H

6.65 6.5, 0.29

C

6.2

6.0, 0.29

−10

L H

TUC

E

6.30

6.3

A

−50

6.1, 0.23

C

MHT

IUC

5.8, 0.22

6.0 BS BS A Vs−LVL

−30

6.0

Depth (km)

C −10

T

Vertical exaggeration 5:1

−90

−90 0

100

200

300

400

500

600

700

800

900

1000

1100

1200

2.0 1.9 1.8 1.7 1.6 1.5

N E X

K

T

S

N

K

S 0

100

200

300

400

2.0 1.9 1.8 1.7 1.6 1.5

X

W

500

600

700

800

K 900

1000

1100

VP/Vs

VP / VS

Distance (km)

1200

Distance (km)

Fig. 4. Cross-section showing the main features of crustal structure as derived from seismological observations beneath Tibet and whole crustal VP/VS (P-wave velocity/S-wave velocity) ratios beneath Tibet. Large numbers denote whole crustal average P-wave velocity in km s21. Small numbers denote average P-wave velocity in km s21 and, if available, Poisson’s ratio for individual layers. Black lines, P-wave velocities and annotations: controlled source observations along the Lhasa–Golmud transect. Blue lines, P-wave velocities and annotations: controlled source observations about 400 km west or east of the Lhasa–Golmud transect. Red lines, P-wave velocities and annotations: passive source (mainly receiver function) observations along the Lhasa– Golmud transect. Green lines: passive source (receiver function) observations about 200 and 400 km west and 400 km east of the Lhasa– Golmud transect. Stippled bands represent the main seismic boundaries within the Tibetan crust. Data, with letters following the references keyed to the measurements shown in the figure, are from Zhao et al. (1993), Brown et al. (1996), A; Hirn et al. (1984a), B; Teng et al. (1983), Yin et al. (1990), C; Makovsky et al. (1999), D; Zhao et al. (2001), Mechie et al. (2004), E; Sapin et al. (1985), Zhang & Klemperer (2005), F; Kong et al. (1996), G; Lu & Wang (1990), H; Li et al. (2004), J; Galve´ et al. (2002a), Galve´ et al. (2006), Jiang et al. (2006), K; Cui et al. (1995), Gao et al. (1995), L; Zhao et al. (2006), M; Kind et al. (2002), N; Mitra et al. (2005), P; Schulte-Pelkum et al. (2005), R; Wittlinger et al. (2009), Na´beˇlek et al. (2009), S; Vergne et al. (2002), T; Zhu & Helmberger (1998), U; Shi et al. (2009), V; Galve´ et al. (2002b), W; Rodgers & Schwartz (1997, 1998), X. Key: see Figure 1 and STD, South Tibetan Detachment system; QB, Qaidam Basin; BS, bright spots; YZR, Yarlung Zangbo Reflection; ABQT, a– b Quartz Transition; MHT, Main Himalayan Thrust; IUC, Indian Upper Crust; ILC, Indian Lower Crust; TUC, Tibetan Upper Crust; TLC, Tibetan Lower Crust.

(2002), who use essentially the same data as used by Vergne et al. (2002). Beneath the Qaidam basin crustal thickness is 50–55 km and there is evidence that the crustal thickness increases in the Qilian belt to the north of the Qaidam basin. Just how abrupt the change in crustal thickness is between the Songpan-Ganz terrane and the Qaidam basin and where exactly the change is located is still a matter of ongoing study within for example, the INDEPTH IV project (Zhao et al. 2008). The most abrupt changes that have so far been invoked are 10 –15 km in depth over a horizontal distance of c. 5 km (Fig. 4).

The cover, consisting mainly of sedimentary rocks and weathered crystalline basement, attains thicknesses of 2.5–7 km across the plateau (Fig. 4). North of the North Kunlun Thrust the sediments in the Qaidam basin reach thicknesses of more than 10 km. Beneath the cover, the top layer of the upper crust has average velocities of 5.8 – 6.1 km s21. In the northern Lhasa and southern Qiangtang terranes, where shear waves (S-waves) have also been observed (Sapin et al. 1985; Zhao et al. 2001; Zhang & Klemperer 2005) average S-wave velocities in the top layer of the upper crust are high with respect to the average P-wave

SEISMOLOGICAL STRUCTURE OF TIBET

velocities, resulting in low VP/VS (P-wave velocity/S-wave velocity) ratios or Poisson’s ratios (s) (Fig. 4). The low Poisson’s ratios of 0.21 –0.23 are indicative of felsic rocks rich in quartz in the a state (Sobolev & Babeyko 1994; Christensen 1996). South of the Kunlun fault, average P-wave velocities in the layer below the boundary within the upper crust at 20 –30 km depth, range from 6.0–6.6 km s21. Beneath the southern Qiangtang and northernmost Lhasa terranes, Mechie et al. (2004) proposed that the boundary at 20 –30 km depth represents the a–b quartz transition (ABQT in Fig. 4). South of c. 318N, bright spots identified at 15–20 km depth below the southern Lhasa terrane (BS in Fig. 4) have been proposed to represent the top of regions of partial melt (Brown et al. 1996; Nelson et al. 1996). As further evidence for this fact, it was noted that the bright spots lie at the top of regions of crustal low S-wave velocities (VS-LVL in Fig. 4, Kind et al. 1996) and high electrical conductivities (Chen et al. 1996; Unsworth et al. 2005). Beneath the Indus-Yarlung suture Makovsky et al. (1999) modelled the Yarlung Zangbo Reflection (YZR in Fig. 4) as representing the top of an ophiolitic slab c. 5 km thick and having a high P-wave velocity of c. 7 km s21 based on refracted phases. At the southern end of the cross-section the prominent reflection identified as the Main Himalayan Thrust (MHT in Fig. 4, Zhao et al. 1993; Brown et al. 1996; Nelson et al. 1996) dips down to the north from c. 25 –43 km depth. The Main Himalayan Thrust (MHT in Fig. 4) represents the boundary along which India is presently underthrusting Tibet. Beneath much of the Lhasa terrane a prominent convertor c. 20 km above the Moho convertor has been identified in the data from the main Lhasa to Golmud transect (Yuan et al. 1997; Kind et al. 2002). In the models of Yuan et al. (1997) for the data from the INDEPTH project and the permanent station at Lhasa, this layer has average S-wave velocities ranging from 4.0–4.3 km s21, which are typical of lower crustal material. In the data from the Hi-CLIMB project c. 400 km to the west of the main Lhasa to Golmud transect, a similar convertor has been identified which, because of the more extensive data coverage to the south, actually extends from at least 100 km south of the Indus-Yarlung suture to just north of the middle of the Lhasa terrane at c. 20–30 km above the Moho convertor (Na´beˇlek et al. 2009; Wittlinger et al. 2009). The 20 km thick layer between the two convertors south of the Indus-Yarlung suture under the Hi-CLIMB profile has been interpreted to represent Indian lower crust (Na´beˇlek et al. 2009; Wittlinger et al. 2009). North of the Indus-Yarlung suture under the Hi-CLIMB profile, this layer has been interpreted to represent a layer of eclogites with an

115

S-wave velocity of 4.73 km s21 and VP/VS ratio of 1.69 (Wittlinger et al. 2009). One problem with this model is that the S-wave velocity contrast producing the upper convertor is c. 1.2 km s21 and a similar S-wave velocity contrast is required to produce the lower convertor as both converted phases have similar amplitudes. However, an upper mantle S-wave velocity of c. 5.9 km s21 is unrealistic. An alternative model for this layer north of the Indus-Yarlung suture beneath the Hi-CLIMB profile has been provided by Na´beˇlek et al. (2009). In this model, derived from receiver function waveform inversion; this layer has an average P-wave velocity of c. 6.9 km s21. This model and the models of Yuan et al. (1997) seem more realistic. In the data from the HIMNT project about midway between the Hi-CLIMB profile and the main Lhasa to Golmud transect, a lower crustal conversion 20–30 km above the Moho convertor has also been recognized extending from the Lesser Himalaya to the Indus-Yarlung suture (Schulte-Pelkum et al. 2005). South of the High Himalaya in Nepal, Moho conversions are stronger and the velocity of the lower crust is lower than in the Tethyan belt north of the High Himalaya, where lower crustal velocities are greater than 7.0 km s21. This northwards increase in lower crustal velocity has been interpreted in terms of partial (c. 30%) eclogitization of the Indian lower crust beneath southern Tibet (SchultePelkum et al. 2005). The caveat here is that the northwards increase in lower crustal velocity may just represent a change in properties of the Indian lower crust beneath southern Tibet due to for example, a terrane (block) boundary. Nevertheless, if this convertor does represent the top of Indian lower crust, irrespective of whether this crust has undergone no, partial or whole eclogitization, then Indian lower crust exists beneath the Tibetan Plateau at least as far north as about the middle of the Lhasa terrane (Fig. 4). From the results of controlled source seismic studies, it can be seen that a boundary exists in the Tibetan crust at c. 40 km depth beneath the southern part of the plateau, rising to c. 33 km depth beneath the northern part of the plateau. Below this boundary, the average P-wave velocity is 6.6– 7.0 km s21, except in two cases, and thus this boundary is taken to be the top of the lower crust beneath Tibet. Information regarding the velocity in this layer can also be gleaned from xenoliths collected during the third phase of the INDEPTH project from the northern Qiangtang terrane at c. 34.48N, 89.28E (Fig. 1, Hacker et al. 2000). The mineralogies of the xenoliths indicate that the lower crust is at ambient temperatures of 800– 1100 8C and is mainly composed of anhydrous metasedimentary rocks, less mafic granulite, and rare

116

J. MECHIE ET AL.

amphibolite and clinopyroxenite (Hacker et al. 2000). Calculations of the physical properties of the xenoliths based on the method of Sobolev & Babeyko (1994) yielded P-wave velocities ranging from 6.1–7.2 km s21 with an average of 6.6 km s21 and Poisson’s ratios ranging from 0.258–0.272 with an average of 0.266 (Hacker et al. 2000). In one of the cases where the average P-wave velocity is outwith the 6.6–7.0 km s21 range, the seismic line trended east– west just south of the BanggongNujiang suture and in this case the average P-wave velocity was estimated to be c. 7.2 km s21 by Zhang & Klemperer (2005). In the other case, the seismic line trended NW–SE just south of the Jinsha suture and in this case the average P-wave velocity and corresponding Poisson’s ratio were estimated to be c. 6.3 km s21 and 0.25 respectively by Galve´ et al. (2006). Further north an extremely low Poisson’s ratio of 0.20 was estimated for the lower crust of the Songpan-Ganz terrane by Jiang et al. (2006). In this case Jiang et al. (2006) discussed the possibility of crustal anisotropy affecting the Poisson’s ratio estimation. These low Poisson’s ratios for the northern Qiangtang terrane and Songpan-Ganz terrane in NE Tibet are indicative of a felsic composition for the lower crust of this region (Galve´ et al. 2006; Jiang et al. 2006). Two further parameters which characterize the crustal structure of the Tibetan Plateau are the average crustal P-wave velocity and the average crustal VP/VS ratio or Poisson’s ratio. Each study which has measured the average crustal P-wave velocity in Tibet has found that it is lower than the global average of 6.45 + 0.21 km s21 (Christensen & Mooney 1995), with values ranging from 5.9– 6.4 km s21 (Fig. 4). Values measured by controlled source seismic studies range from 6.2 –6.4 km s21, except beneath the Qaidam basin where the large thickness of low-velocity sediments lowers the average crustal P-wave velocity to 6.0– 6.1 km s21. Values measured using earthquake sources range from 5.9–6.3 km s21, with lower values in the Lhasa terrane in the south and higher values in the Qiangtang terrane in the north (Rodgers & Schwartz 1997, 1998). With respect to the average crustal VP/VS ratio, Kind et al. (2002) show the variation of this parameter across the Tibetan Plateau from 28.5–36.58N, derived from receiver function data (Fig. 4). There seems to be a tendency for lower values of this parameter away from the centre of the plateau towards its southern and northern margins. The greatest variation or discrepancy in this parameter seems to occur in the Lhasa terrane at c. 318N, where values ranging from 1.68 –1.91 have been determined from data sets spaced only c. 200 km apart. Analysis of surface wave dispersion data has also provided information on the S-wave velocity

structure of the crust beneath the Tibetan Plateau. Early studies showed that the average crustal S-wave velocity below Tibet had low values of 3.4–3.5 km s21 and that there were indications for a low-velocity zone at intermediate crustal depths (Chun & Yoshii 1977; Chen & Molnar 1981; Romanowicz 1982). With the advent of temporary, dense, broadband deployments on the plateau, it has been possible to make regional S-wave models using surface waves. Thus Cotte et al. (1999) derived crustal S-wave models using surface waves for the Tethyan belt south of the Indus-Yarlung suture and the southern Lhasa terrane north of the suture and found a low-velocity zone in the lower crust north of the suture but a high-velocity layer in the lower crust south of the suture and a sharp transition at the suture. A similar analysis by Rapine et al. (2003) required a low-velocity zone in the middle crust of the western portion of southern Tibet (Lhasa terrane) south of the Banggong-Nujiang suture but did not require any crustal low-velocity zone in the western portion of northern Tibet (Qiangtang terrane) north of the suture. Further, these authors derived average crustal S-wave velocities for the western portion of the Lhasa terrane of 3.49 km s21 and for the western portion of the Qiangtang terrane of 3.45 km s21, and also found that the S-wave velocity in the lower crust beneath the western Lhasa terrane is c. 0.2 km s21 faster than under the western Qiangtang terrane. Another line of study which provides information on the rheological state of the crust beneath the Tibetan Plateau is the seismicity within the plateau. Using the stations from the INDEPTH III project in central Tibet Langin et al. (2003) were able to locate 267 local earthquakes. 99% of these earthquakes had focal depths less than 25 km, and the locations of the few deeper events were poorly constrained. This analysis confirmed the results from earlier studies (e.g. Molnar & Chen 1983; Molnar & Lyon-Caen 1989; Zhao & Helmberger 1991; Randall et al. 1995) which had also found that all earthquakes within the plateau occurred at focal depths of less than 25 km, except for one event at about 90 km depth in the NW corner of the plateau and two events also at 85–90 km depth beneath southernmost Tibet. A recent study by Priestley et al. (2008) has also found evidence for deep events in the NW corner of the plateau and beneath southernmost Tibet. With respect to the deep events beneath southernmost Tibet between 86 and 908E in the region of the main Lhasa-Golmud transect, they all occur south of 308N, which is south of the northern end of the strong lower crustal convertor identified in the receiver function data (Yuan et al. 1997; Kind et al. 2002; Schulte-Pelkum et al. 2005; Wittlinger et al. 2009; Na´beˇlek et al. 2009).

SEISMOLOGICAL STRUCTURE OF TIBET

The final line of study which will be discussed and which provides information on the rheological state of the crust beneath the Tibetan Plateau is attenuation. Back in the seventies Ruzaikin et al. (1977) noted that the crustal shear wave phase, Lg, could not be observed for paths which had crossed the Tibetan Plateau. Later studies (e.g. McNamara et al. 1996) have recognized that Lg can propagate up to distances of at least 600 km within the plateau, but that for paths crossing the plateau margins, the phase is severely attenuated such that it can often not be recognized anymore. For 1 Hz Lg data, estimates of the quality factor, Q, provide values of 60 –400 for the plateau itself (McNamara et al. 1996; Reese & Ni 1996; Phillips et al. 2000; Xie et al. 2004). Along the INDEPTH profiles in southern and central Tibet, the Q value estimated for Lg at 1 Hz is 60 –100 north of the Indus-Yarlung suture and then increases south to values higher than 300 at about 200 km south of the suture (Xie et al. 2004). In central Tibet the P-wave refracted through the upper crust, Pg, and the P-wave reflected from the Moho, PmP, can be well observed and the S-wave refracted through the upper crust, Sg, can also be quite well observed (Sapin et al. 1985; Mechie et al. 2004; Zhang & Klemperer 2005). However, the S-wave reflected from the Lhasa terrane

HHC STD Tethyan belt IYS

BNS

117

Moho, SmS, is much more difficult to observe (Sapin et al. 1985; Mechie et al. 2004; Zhang & Klemperer 2005) and Sapin et al. (1985) noted that the attenuation of S-waves in the crust below c. 15 –25 km depth in central Tibet is greater than that of P-waves. It thus seems reasonable to postulate that most of the attenuation of the Lg phase beneath Tibet must be occurring at depths greater than c. 20 km. From controlled source wide-angle reflection/refraction seismic data south of the Indus-Yarlung suture, Hirn & Sapin (1984) determined the Q value for P-waves in the crust below 30 km depth to be c. 125. From similar data collected in northeastern Tibet, Galve´ et al. (2006) derived the Q value for P-waves in the crust to be at least 7000 below c. 40 km depth in the northern Qiangtang terrane and to be between 1000–2000 below c. 30 km depth in the Songpan-Ganz terrane. Attenuation of the S-wave refracted through the uppermost mantle, Sn, will be discussed in the following section.

Mantle structure The mantle part of the cross-section (Fig. 5) shows a somewhat less cluttered picture, due to the fact that there are fewer results for the deeper parts of the Qiangtang terrane

JS SGT KF

NKT

QB

NQF

Qilian belt

Latitude (°N) 28

29

30

31

32

33

34

35

36

37

38

0

0

Moho B

8.32, 4.59 H

Depth (km)

−200

LAB

F G

−300

D

E

8.03 D

C

ALM F LAB ILM ~2 s shear wave splitting in E−W to NE−SW direction A D

−100 −200 −300

−400

−400

C

410 −500

−500

−600

Depth (km)

−100

−600

C

660

−700

−700

Vertical exaggeration 1:2 −800

−800 0

100

200

300

400

500

600

700

800

900

1000

1100

1200

Distance (km)

Fig. 5. Cross-section showing the main features of mantle structure as derived from seismological observations beneath Tibet. Pn and Sn velocities are from McNamara et al. (1995). Black lines (B, C, E) are boundaries where velocity increases downwards identified mainly from P receiver function images. Red lines (F) represent the lithosphere– asthenosphere boundary (LAB) where velocity decreases downwards identified from S receiver function images. Blue to green (red) shaded regions (A, D, G, H) are high (low) velocity anomalies identified from teleseismic P-wave delay time or surface wave tomography studies. The region of substantial shear wave splitting between about 328N and the Kunlun Fault is indicated. Other data, with letters following the references keyed to the measurements shown in the figure, are from Wittlinger et al. (1996), A; Kosarev et al. (1999), B; Kind et al. (2002), C; Tilmann et al. (2003), D; Shi et al. (2004), E; Kumar et al. (2006), F; Li et al. (2008), G; Friederich (2003), H. The apparent spreading of the high velocity region, G, at greater depths is due to the horizontal exaggeration. Key: see Figures 1 and 4 and ILM, Indian Lithospheric Mantle; ALM, Asian Lithospheric Mantle; 410 & 660, Upper mantle discontinuities at c. 410 and 660 km depth.

118

J. MECHIE ET AL.

section. In fact, nearly all the results are from observations made at stations along the main Lhasa to Golmud transect. Immediately below the Moho, from regional observations of the Pn and Sn phases, the P-wave velocity of the uppermost mantle beneath southern Tibet is faster than beneath northern Tibet (McNamara et al. 1995). Whereas observations of the Sn phase in southern Tibet give an uppermost mantle S-wave velocity estimate of c. 4.6 km s21 in this region, this phase cannot be observed at high frequencies above c. 1 Hz beneath northern Tibet (Ni & Barazangi 1983; McNamara et al. 1995). These observations imply higher uppermost mantle temperatures beneath northern Tibet than beneath southern Tibet (McNamara et al. 1995). A tomographic study of teleseismic P-wave delay times from the Sino-French passive source experiment in the northern part of the plateau revealed a large slow velocity anomaly between 34– 368N between depths of 200 –350 km (A in Fig. 5, Wittlinger et al. 1996). A similar study based on data from the third phase of the INDEPTH project in central Tibet revealed a large fast velocity anomaly c. 100 km wide just south of the Banggong-Nujiang Suture extending essentially from the Moho down to c. 400 km depth (D in Fig. 5, Tilmann et al. 2003). This anomaly was interpreted in terms of fast, cold downwelling Indian lithospheric mantle (Tilmann et al. 2003). A crosssection at c. 898E (P in Fig. 1) through a regional model derived using P-wave travel-time tomography, by Li et al. (2008) shows a high velocity anomaly in the upper mantle beneath southern Tibet, extending about as far north as the BanggongNujiang Suture (G in Fig. 5). North of c. 308N this anomaly seems to extend into the mantle transition zone to depths of c. 550 km. However, west of c. 87.58E and east of c. 91.58E this anomaly does not appear to penetrate into the mantle transition zone (Li et al. 2008). Surface wave tomography has also indicated high S-wave velocities in the upper mantle beneath southern Tibet (Griot et al. 1998; Friederich 2003; Priestley et al. 2006). The cross-section (R in Fig. 1) along the main LhasaGolmud transect through the model of Friederich (2003) reveals high S-wave velocities between 100– 300 km depth beneath southern Tibet extending north until about the Banggong-Nujiang suture (H in Fig. 5). However, for northern Tibet the picture is less clear. Whereas, Friederich (2003) found average-to-low S-wave velocities in the upper mantle down to c. 660 km depth beneath northern Tibet, Griot et al. (1998) and Priestley et al. (2006) found evidence for high S-wave velocities down to 250 –300 km depth. Shear wave splitting studies have also revealed differences in mantle properties beneath the Tibetan Plateau. Shear wave birefringence of teleseismic

SKS phases (shear wave which has passed through the core as a compressional P-wave) is regarded as a proxy for anisotropy in the mantle beneath a region. Negligible shear wave birefringence has been found beneath India, the central Tethyan belt and the southern Lhasa terrane (McNamara et al. 1994; Hirn et al. 1995; Sandvol et al. 1997; Chen ¨ zalabey 1998) and moderate values of splitting &O less than 1 s have been found in the northern Lhasa terrane and the Qaidam basin (McNamara et al. 1994; Hirn et al. 1995; Sandvol et al. 1997). In contrast, strong shear wave splitting up to 2 s has been found north of 328N beneath the northernmost Lhasa terrane, the Qiangtang terrane and the Songpan-Ganz terrane (McNamara et al. 1994; Huang et al. 2000). Beneath the northern plateau the fast polarization direction varies from east – west to NE– SW (McNamara et al. 1994; Huang et al. 2000). Taken in conjunction with for example, the Pn and Sn observations and the body and surface wave tomography results discussed above, the anisotropy measurements are consistent with a model in which hot and weak upper mantle beneath northern Tibet is being extruded eastwards between the advancing Indian lithosphere to the south and the Qaidam and Tarim lithosphere(s) to the north and west (Huang et al. 2000). GPS measurements on the Tibetan Plateau (Wang et al. 2001) document that the surface of the plateau south of the Kunlun fault is moving eastwards with respect to both Eurasia and India. Some of the anisotropy exhibited by seismic stations on the Tibetan Plateau is thought to originate in the crust (e.g. Vergne et al. 2003; Ozacar & Zandt 2004; Sherrington et al. 2004). Beneath the central part of the plateau the fast polarization direction is east –west in the middle to lower crustal layers (Sherrington et al. 2004). Thus the GPS measurements and the crustal anisotropy analysis indicate that, in addition to the upper mantle, the crust of the Tibetan Plateau south of the Kunlun fault is also being extruded eastwards. The thin black lines in the cross-section (Fig. 5) represent conversions, identified in P-receiver functions, from boundaries with a downwards velocity increase (Kosarev et al. 1999; Kind et al. 2002; Shi et al. 2004) whereas the red line represents the conversion from the lithosphere– asthenosphere boundary (LAB) at depths of 160–225 km, and identified in S-receiver functions (Kumar et al. 2006) with a downwards velocity decrease. The dipping convertors in the mantle lithosphere have all been interpreted as evidence for intra-continental subduction. The ILM was interpreted by Kosarev et al. (1999) to represent Indian Lithospheric Mantle subducting northwards beneath southern Tibet. Similarly, the ALM was interpreted by Kind et al. (2002) to represent Asian Lithospheric Mantle

SEISMOLOGICAL STRUCTURE OF TIBET

subducting southwards beneath northern Tibet, while the south-dipping convertor at c. 328N was interpreted to represent southward intra-continental subduction of Tibetan lithosphere along the Banggong-Nujiang suture (Shi et al. 2004). One of the most spectacular results of the receiver function studies from Tibet is the image of the 410 and 660 km discontinuities bounding the mantle transition zone (Fig. 3b, Kind et al. 2002). These two boundaries are continuous and run more or less parallel to each other beneath the whole of the plateau. This implies that there is no subducting slab penetrating into the mantle transition zone beneath the plateau, which is at variance with the tomography results of Li et al. (2008), and that temperature variations within the mantle transition zone are negligible. In particular, the piercing points where the waves penetrate the mantle transition zone for the receiver function example shown here (Fig. 3b) lie at the same geographical locations as where the high velocity anomaly of Li et al. (2008) is postulated to extend into the mantle transition zone. The two boundaries are apparently about 20 km deeper beneath the northern part of the plateau than beneath the southern part. This could be due to a south-to-north decrease in the S-wave velocity in the upper mantle above the transition zone by c. 5% which, in turn, could be

119

due to a corresponding northwards increase in upper mantle temperature by c. 300 8C (Kind et al. 2002).

Synoptic cross-section Based on the crustal and upper mantle crosssections (Figs 4 & 5), a synoptic cross-section down to 700 km depth beneath the Tibetan Plateau has been produced (Fig. 6). In the south the section is based on the one derived by Nelson et al. (1996). South of the Indus-Yarlung Suture, Tethyan belt cover rocks and seismic basement (TBC in Fig. 6) overlie the South Tibetan Detachment system (STD in Fig. 6). High Himalayan Crystalline rocks of Indian affinity occupy the footwall of the South Tibetan Detachment system. The Main Himalayan Thrust is the present-day active plate boundary along which India underthrusts the High Himalayan Crystalline rocks. Whereas Indian upper crust is shown to be added to the trailing edge of the High Himalayan Crystalline rocks, Indian lower crust is shown to underlie the plateau until north of the middle of the Lhasa terrane. The upper boundary of the Indian lower crust is marked by the prominent conversion c. 20 –30 km above the Moho conversion observed on three receiver function profiles over an east –west distance of c. 400 km between 85 –898E (Yuan et al. 1997; Kind et al. 2002;

Fig. 6. Synoptic cross-section for Tibet based mainly on the results shown in Figures 4 and 5. Note the change of scale at 75 km depth. The apparent spreading of the ILM at greater depths is due to the horizontal exaggeration. Key: see Figures 1, 4 and 5 and TBC, Tethyan Belt seismic basement; TLM, Tibetan Lithospheric Mantle; yellow colour, cover rocks; dark red, low velocity anomaly in upper mantle beneath northern Tibet (Wittlinger et al. 1996); stippled areas, regions of greater concentrations of partial melt.

120

J. MECHIE ET AL.

Schulte-Pelkum et al. 2005; Na´beˇlek et al. 2009; Wittlinger et al. 2009). The fact that the conversion is prominent indicates that the Indian lower crust beneath southern Tibet has a high S-wave velocity ranging from 4.0 –4.3 km s21 (Yuan et al. 1997). DeCelles et al. (2002) have correlated this high velocity basal crustal layer beneath southern Tibet with Indian cratonic lower crust and have further suggested that Greater Indian (Himalayan) lower crust extends all the way north to the Kunlun fault. Below the Moho, high-velocity Indian lithospheric mantle underlies the Indian lower crust. Irrespective of whether the deep seismicity beneath southernmost Tibet occurs in the Indian lower crust or the Indian lithospheric mantle, or even both, the evidence points to the fact that the Indian lower crust and the Indian lithospheric mantle have high velocities and are dense and cold. Based mainly on the tomographic images of Tilmann et al. (2003), Friederich (2003) and Li et al. (2008), Indian lithospheric mantle extends northwards to about the Banggong-Nujiang Suture where it downwells to about the base of the upper mantle at about 400 km depth (Tilmann et al. 2003). At variance with the results of Li et al. (2008), Figure 6 shows that Indian lithospheric mantle does not penetrate into the mantle transition zone as evidenced by the continuous 410 km discontinuity beneath the whole of the Tibetan Plateau in the receiver function images (Fig. 3b, Kind et al. 2002). North of the Indus-Yarlung Suture, beneath the layer of cover rocks, the upper crust down to at least 15–20 km depth generally has P-wave velocities of 5.8–6.2 km s21. In the central part of the plateau from 30.5 –348N, where S-waves have also been observed, Poisson’s ratios are low, indicative of felsic rocks rich in quartz in the a state (Sobolev & Babeyko 1994; Christensen 1996). Beneath the central part of the plateau Mechie et al. (2004) postulated that the boundary where the a to b transition in quartz takes place could be identified at 20–30 km depth (ABQT in Fig. 6). As the ABQT is a precise temperature indicator, this, in turn, indicates temperatures of 700 8C at about 20 km depth beneath the southern Qiangtang terrane and 800 8C at c. 30 km depth beneath the northern Lhasa terrane. Just below this boundary temperatures would then be high enough to cause partial melting of the felsic rocks as is also indicated by a region of extremely high electrical conductivities, low seismic velocities and high Poisson’s ratios (Mechie et al. 2004; Solon et al. 2005). Further south, bright spots identified in near-vertical incidence and wide-angle reflection data, very low S-wave velocities and extremely high electrical conductivities again indicate a region of partial melting beneath the southern Lhasa terrane below c. 15 km depth along the main Lhasa to Golmud transect

(Brown et al. 1996; Chen et al. 1996; Kind et al. 1996; Makovsky et al. 1996; Nelson et al. 1996). As the extremely high electrical conductivities extend from 77 –928E, a region of partial melting has been interpreted to be present for at least 1000 km along-strike beneath the southern Lhasa terrane (Unsworth et al. 2005). From the receiver function analysis of the data from the Hi-CLIMB project about 400 km to the west of the main Lhasa to Golmud transect, Na´beˇlek et al. (2009) also recognized conversions of limited horizontal extent (,50 km) indicating a downwards velocity decrease at about 15 km depth beneath the southern Lhasa terrane. Na´beˇlek et al. (2009) also presented an average P-wave velocity depth profile for the southern Lhasa terrane determined by receiver function waveform inversion, that contains a low velocity zone between c. 15–35 km depth. Nevertheless, Na´beˇlek et al. (2009) concluded that these conversions indicating a downwards velocity decrease do not form the top of a partially molten middle and lower crust. On the balance of the evidence though, it is thought that at least quite extensively, below c. 15 km depth in the southern Lhasa terrane, 30 km depth in the northern Lhasa terrane and 20 km depth in the southern Qiangtang terrane, temperatures high enough to cause partial melting of the felsic rocks exist. This is consistent with the brittle –ductile transition occurring in the upper 25 km of the crust beneath the plateau as indicated by the general confinement of earthquake focal depths to the upper 25 km of the crust beneath the plateau (Langin et al. 2003). It is also consistent with models involving ductile flow within the deeper (middle and/or lower) crust beneath the Tibetan Plateau as proposed by for example, Zhao & Morgan (1987); Clark & Royden (2000) and Klemperer (2006). One of the main results which has emerged from this compilation of velocity models is the existence of a boundary at 30 –40 km depth beneath Tibet. Based on the P-wave velocities on either side of this boundary, it is proposed that this boundary represents the interface between the felsic upper crust and more mafic lower crust. This boundary generally occurs at c. 40 km depth beneath the Lhasa terrane in southern Tibet and then rises to c. 33 km depth beneath northern Tibet and the Qaidam basin. Due to the more mafic composition of the lower crustal layer, it is possible that the regions of partial melting beneath southern Tibet do not extend into this layer. Only beneath the northern Qiangtang terrane do the extremely high electrical conductivities begin at depths which are greater than the depth of this boundary (Wei et al. 2001; Unsworth et al. 2004). On the other hand, the northern Qiangtang terrane (Fig. 1) is also where Hacker et al. (2000) found evidence based on xenoliths for

SEISMOLOGICAL STRUCTURE OF TIBET

a hot, but nevertheless dry, lower crust. The caveat is, of course, that there are significant lateral variations in the properties of the lower crust beneath the northern Qiangtang terrane, as the locality where the xenoliths were found, is c. 300 km west of the profile where the electrical conductivities were measured. As the profile where the electrical conductivities were measured is along the main Lhasa to Golmud transect, the synoptic crosssection shows a region of partial melts in the lower crust beneath the northern Qiangtang terrane (Fig. 6). The boundary at 30 –40 km depth beneath Tibet has not been recognized in receiver function studies, which implies that the S-wave velocity contrast across this boundary is small. In the case of the INDEPTH controlled source profile in central Tibet, for which both P and S-wave velocity models were obtained, the S-wave velocity contrast across this boundary was c. 20–40% of the S-wave velocity contrast across the Moho (Mechie et al. 2004). Thus it is perhaps not surprising that this boundary has not been recognized in receiver function studies. With a value of c. 74 km, the crust is thickest beneath southern Tibet where Indian lower crust is present. Beneath northern Tibet it is at least 4 km thinner. With a value of 66 km, the crust is thinnest under the southern Qiangtang terrane. This north – south variation in crustal thickness is somewhat similar to that obtained from analysis of gravity data by Braitenberg et al. (2000). As the Kunlun mountains are crossed the topography decreases northwards from 4.5– 5 km elevation to about 3 km in the Qaidam basin. A concomitant northwards decrease in crustal thickness occurs from c. 70 km beneath the high elevations of the northern plateau to c. 54 km under the Qaidam basin, where sedimentary thicknesses in excess of 10 km are also present. Just how sharp this transition in crustal thickness is, is a matter of ongoing study within for example, the fourth phase of the INDEPTH project. As the surface elevations in northern Tibet are at least as high as those in southern Tibet, it is perhaps somewhat surprising that the crust is somewhat thinner below northern Tibet than under southern Tibet, bearing in mind that the plateau is more or less in isostatic equilibrium (Molnar 1988; Jin et al. 1994). This state of affairs implies that the crustal density and/or the upper mantle density is lower beneath northern Tibet than beneath southern Tibet. A lower upper mantle density beneath northern Tibet with respect to southern Tibet could support the hypothesis that the upper mantle beneath northern Tibet is hotter than beneath southern Tibet (Ni & Barazangi 1983; Molnar 1988). This is in turn consistent with the observations of a lower Pn velocity beneath northern Tibet (McNamara et al.

121

1995), a region with a strongly attenuated highfrequency Sn phase under northern Tibet (Ni & Barazangi 1983; McNamara et al. 1995), strongly developed shear wave anisotropy under northern Tibet (McNamara et al. 1994; Huang et al. 2000), a large low-velocity anomaly in the upper mantle beneath northern Tibet (Wittlinger et al. 1996) and a possibly apparent deepening of the 410 and 660 km discontinuities beneath northern Tibet with respect to southern Tibet (Kind et al. 2002). Within the upper mantle the position of the lithosphere –asthenophere boundary (LAB) sometimes seems to be at odds with other results, especially from tomographic studies (Fig. 5). For example, the LAB north of 318N penetrates into the highvelocity bodies (D and H in Fig. 5) interpreted to represent Indian lithospheric mantle beneath southern Tibet. South of 318N the LAB lies well within the high-velocity body of Li et al. (2008) although it coincides quite well with the base of the high-velocity body of Friederich (2003). On the other hand, beneath northern Tibet the LAB passes mainly below the features in the upper mantle above 200 km depth and it also runs above the low-velocity body of Wittlinger et al. (1996). In the synoptic cross-section the thickness of the Indian lithospheric mantle has been kept more or less constant as it downwells into the upper mantle at c. 318N (Fig. 6, in which the vertical exaggeration of 1:2 should be noted). South of c. 318N, the lithospheric thickness as defined by the high-velocity body of Friederich (2003) and the position of the LAB has been favoured over that as defined by the high-velocity body of Li et al. (2008). The gap in the LAB in the synoptic cross-section (Fig. 6) is somewhat larger than that in the mantle crosssection (Fig. 5). However, the concept of Indian lithospheric mantle downwelling through this gap is the same in both cross-sections. The differences which probably exist in the upper mantle between northern and southern Tibet seem to disappear by the depth at which the transition zone is reached. The continuity and parallelism of these two discontinuities beneath the whole plateau as shown in the receiver function image of Kind et al. (2002) indicates that the mantle is normal and undisturbed at the level of the transition zone and that the orogeny has negligible effects at depths greater than c. 400 km. This is at variance with the tomography results of Li et al. (2008) which suggest that between 87.5–91.58E the transition zone beneath the plateau is affected by the India– Asia collision. It has been the aim of this paper to provide a two-dimensional cross-section of the crust and mantle down to 700 km depth beneath the Tibetan Plateau. It is hoped that this cross-section will provide a link between the many individual regional

122

J. MECHIE ET AL.

studies on the plateau and the three-dimensional distribution of various boundaries and velocity anomalies which are beginning to emerge from for example, whole-scale plateau studies (see e.g. Li et al. 2008). A test of how good the cross-section is would be to use it as a model recovery test with the geometries of the various experiments which have been carried out to date across the plateau. The authors participation in projects in Tibet has been made possible by funding from the Deutsche Forschungsgemeinschaft and the Deutsches GeoForschungs Zentrum – GFZ and the provision of instruments from the Geophysical Instrument Pool of the Deutsches Geoforschungs Zentrum – GFZ. The help of Gao Rui is acknowledged in providing access to Chinese journals describing the results from some of the controlled source profiles carried out by Chinese scientists. Discussions with Lothar Ratschbacher are also acknowledged. Comments by James Ni, Richard Gloaguen and two anonymous reviewers also improved the manuscript.

References Braitenberg, C., Zadro, M., Fang, J., Wang, Y. & Hsu, H. T. 2000. The gravity and isostatic Moho undulations in Qinghai-Tibet plateau. Journal of Geodynamics, 30, 489– 505. Brown, L. D., Zhao, W. et al. 1996. Bright spots, structure, and magmatism in southern Tibet from INDEPTH seismic reflection profiling. Science, 274, 1688–1690. Chen, L., Booker, J. R., Jones, A. G., Wu, N., Unsworth, M. J., Wei, W. & Tan, H. 1996. Electrically conductive crust in southern Tibet from Indepth magnetotelluric surveying. Science, 274, 1694–1696. Chen, W.-P. & Molnar, P. 1981. Constraints on the seismic wave velocity structure beneath the Tibetan plateau and their tectonic implications. Journal of Geophysical Research, 86, 5937– 5962. ¨ zalabey, S. 1998. Correlation between Chen, W. P. & O seismic anisotropy and Bouguer gravity anomalies in Tibet and its implications for lithospheric structure. Geophysical Journal International, 135, 93–101. Christensen, N. I. 1996. Poisson’s ratio and crustal seismology. Journal of Geophysical Research, 101, 3139–3156. Christensen, N. I. & Mooney, W. D. 1995. Seismic velocity structure and composition of the continental crust: a global view. Journal of Geophysical Research, 100, 9761– 9788. Chun, K.-Y. & Yoshii, T. 1977. Crustal structure of the Tibetan plateau: a surface-wave study by a moving window analysis. Bulletin of the Seismological Society of America, 67, 735–750. Clark, M. K. & Royden, L. H. 2000. Topographic ooze; building the eastern margin of Tibet by lower crustal flow. Geology, 28, 703–706. Cotte, N., Pedersen, H. et al. 1999. Determination of the crustal structure in southern Tibet by dispersion and amplitude analysis of Rayleigh waves. Geophysical Journal International, 138, 809– 819.

Cui, Z., Li, Q., Wu, C., Yin, Z. & Liu, H. 1995. The crustal and deep structures in Golmud – Ejin Qi GGT. Acta Geophysica Sinica, 38, 15– 28. DeCelles, P. G., Robinson, D. M. & Zandt, G. 2002. Implications of shortening in the Himalayan fold-thrust belt for uplift of the Tibetan plateau. Tectonics, 21, 1062, doi: 10.1029/2001TC001322. Friederich, W. 2003. The S-velocity structure of the East Asian mantle from inversion of shear and surface waveforms. Geophysical Journal International, 153, 88–102. Galve´, A., Hirn, A. et al. 2002a. Modes of raising northeastern Tibet probed by explosion seismology. Earth and Planetary Sciences Letters, 203, 35–43. Galve´, A., Sapin, M. et al. 2002b. Complex images of Moho and variation of Vp/Vs across the Himalaya and South Tibet, from a joint receiver-function and wide-angle reflection approach. Geophysical Research Letters, 29, 2182; doi: 10.1029/2002GL015611. Galve´, A., Jiang, M. et al. 2006. Explosion seismic P and S velocity and attenuation constraints on the lower crust of the North– Central Tibetan Plateau, and comparison with the Tethyan Himalayas: Implications on composition, mineralogy, temperature, and tectonic evolution. Tectonophysics, 412, 141–157; doi: 10.1016/j.tecto.2005.09.010. Gao, R., Cheng, X.-Z. & Ding, Q. 1995. Preliminary geodynamic model of Golmud-Ejin Qi geoscience transect. Acta Geophysica Sinica, 38, 3– 14. Gao, R., Lu, Z., Li, Q., Guan, Y., Zhang, J., He, R. & Huang, L. 2005. Geophysical survey and geodynamic study of crust and upper mantle in the Qinghai-Tibet plateau. Episodes, 28, 263–273. Griot, D.-A., Montagner, J.-P. & Tapponier, P. 1998. Phase velocity structure from Rayleigh and Love waves in Tibet and its neighboring regions. Journal of Geophysical Research, 103, 21215– 21232. Hacker, B. R., Gnos, E. et al. 2000. Hot and dry deep crustal xenoliths from Tibet. Science, 287, 2463– 2466. Herquel, G., Wittlinger, G. & Guilbert, J. 1995. Anisotropy and crustal thickness of Northern-Tibet. New constraints for tectonic modeling. Geophysical Research Letters, 22, 1925–1928. Herquel, G., Tapponnier, P., Wittlinger, G., Mei, J. & Danian, S. 1999. Teleseismic shear wave splitting and lithospheric anisotropy beneath and across the Altyn Tagh fault. Geophysical Research Letters, 26, 3225– 3228. Hirn, A. & Sapin, M. 1984. The Himalayan zone of crustal interaction: suggestions from explosion seismology. Annales Geophysicae, 2, 123–130. Hirn, A., Lepine, J.-C. et al. 1984a. Crustal structure and variability of the Himalayan border of Tibet. Nature, 307, 23– 25. Hirn, A., Nercessian, A. et al. 1984b. Lhasa block and bordering sutures – a continuation of a 500-km Moho traverse through Tibet. Nature, 307, 25–27. Hirn, A., Jiang, M. et al. 1995. Seismic anisotropy as an indicator of mantle flow beneath the Himalayas and Tibet. Nature, 375, 571–574. Huang, W.-C., Ni, J. F. et al. 2000. Seismic polarization anisotropy beneath the central Tibetan Plateau. Journal of Geophysical Research, 105, 27979– 27989.

SEISMOLOGICAL STRUCTURE OF TIBET Jiang, M., Galve´, A. et al. 2006. Crustal thickening and variations in architecture from the Qaidam basin to the Qang Tang (North – Central Tibetan Plateau) from wide-angle reflection seismology. Tectonophysics, 412, 121– 140; doi: 10.1016/j.tecto.2005.09.011. Jin, Y., McNutt, M. K. & Zhu, Y. 1994. Evidence from gravity and topography data for folding of Tibet. Nature, 371, 669– 674. Kao, H., Gao, R., Rau, R.-J., Shi, D., Chen, R.-Y., Guan, Y. & Wu, F. T. 2001. Seismic image of the Tarim basin and its collision with Tibet. Geology, 29, 575–578. Kennett, B. L. N. & Engdahl, E. R. 1991. Traveltimes for global earthquake location and phase identification. Geophysical Journal International, 105, 429–465. Kind, R., Ni, J. et al. 1996. Evidence from earthquake data for a partially molten crustal layer in southern Tibet. Science, 274, 1692– 1694. Kind, R., Yuan, X. et al. 2002. Seismic images of crust and upper mantle beneath Tibet: Evidence for Eurasian plate subduction. Science, 298, 1219–1221. Klemperer, S. L. 2006. Crustal flow in Tibet: geophysical evidence for the physical state of Tibetan lithosphere, and inferred patterns of crustal flow. In: Law, R. D., Searle, M. P. & Godin, L. (eds) Channel Flow, Ductile Extrusion and Exhumation in Continental Collision Zones. Geological Society, London, Special Publications, 268, 39– 70. Kong, X., Wang, Q. & Xiong, S. 1996. Comprehensive geophysics and lithospheric structure in the western Xizang (Tibet) Plateau. Science in China (Series D), 39, 348–358. Kosarev, G., Kind, R., Sobolev, S. V., Yuan, X., Hanka, W. & Oreshin, S. 1999. Seismic evidence for a detached Indian lithospheric mantle beneath Tibet. Science, 283, 1306– 1309. Kumar, P., Yuan, X., Kind, R. & Ni, J. 2006. Imaging the colliding Indian and Asian lithospheric plates beneath Tibet. Journal of Geophysical Research, 111, B06308, doi: 10.1029/2005JB003930. Langin, W. R., Brown, L. D. & Sandvol, E. A. 2003. Seismicity of central Tibet from project Indepth III seismic recordings. Bulletin of the Seismology Society of America, 93, 2146– 2159. Li, C., van der Hilst, R. D., Meltzer, A. S. & Engdahl, E. R. 2008. Subduction of the Indian lithosphere beneath the Tibetan plateau and Burma. Earth and Planetary Science Letters, 274, 157– 168; doi: 10.1016/j.epsl.2008.07.016. Li, Q., Peng, S., Gao, R., Guan, Y. & Fan, J. 2004. Deep tectonic background of the 8.1 Ms earthquake in the East Kunlun. Acta Geoscientica Sinica, 25, 11– 16. Lu, D. & Wang, X. 1990. The crustal structure and deep internal processes in the Tuotuohe-Golmud area of the north Qinghai-Xizang plateau. Bulletin of the Chinese Academy of Geological Sciences, 21, 227–237. Makovsky, Y., Klemperer, S. L., Ratschbacher, L., Brown, L. D., Li, M., Zhao, W. & Meng, F. 1996. Indepth wide-angle reflection observation of Pwave-to-S-wave conversion from crustal bright spots in Tibet. Science, 274, 1690–1691.

123

Makovsky, Y., Klemperer, S. L., Ratschbacher, L. & Alsdorf, D. 1999. Midcrustal reflector on Indepth wide-angle profiles: An ophiolitic slab beneath the India– Asia suture in southern Tibet? Tectonics, 18, 793– 808. McNamara, D. E., Owens, T. J., Silver, P. G. & Wu, F. T. 1994. Shear wave anisotropy beneath the Tibetan Plateau. Journal of Geophysical Research, 99, 13655–13665. McNamara, D. E., Owens, T. J. & Walter, W. R. 1995. Observations of regional phase propagation across the Tibetan plateau. Journal of Geophysical Research, 100, 22215–22229. McNamara, D. E., Owens, T. J. & Walter, W. R. 1996. Propagation characteristics of Lg across the Tibetan plateau. Bulletin of the Seismological Society of America, 86, 457–469. Mechie, J., Sobolev, S. V. et al. 2004. Precise temperature estimation in the Tibetan crust fom seismic detection of the a –b quartz transition. Geology, 32, 601– 604; doi: 10.1130/G20367.1. Mitra, S., Priestley, K., Bhattacharyya, A. K. & Gaur, V. K. 2005. Crustal structure and earthquake focal depths beneath northeastern India and southern Tibet. Geophysical Journal International, 160, 227– 248; doi: 10.1111/j.1365–246X.2004.02470.x. Molnar, P. 1988. A review of geophysical constraints on the deep structure of the Tibetan Plateau, the Himalaya and the Karakorum, and their tectonic implications. Philosophical Transactions of the Royal Society of London, A, 326, 33–88. Molnar, P. & Chen, W.-P. 1983. Focal depths and fault plane solutions of earthquakes under the Tibetan plateau. Journal of Geophysical Research, 88, 1180– 1196. Molnar, R. & Lyon-Caen, H. 1989. Fault plane solutions of earthquakes and active tectonics of the Tibetan Plateau and its margins. Geophysical Journal International, 99, 123 –153. Na´belek, J., Hete´nyi, G. et al. & The Hi-Climb Team. 2009. Underplating in the Himalaya-Tibet collision zone revealed by the Hi-CLIMB experiment. Science, 325, 1371– 1374; doi: 10.1126/science.1167719. Nelson, K. D., Zhao, W. et al. 1996. Partially molten middle crust beneath southern Tibet: Synthesis of project Indepth results. Science, 274, 1684– 1688. Ni, J. & Barazangi, M. 1983. High-frequency seismic wave propagation beneath the Indian Shield, Himalayan Arc, Tibetan Plateau and surrounding regions: high uppermost mantle velocities and efficient Sn propagation beneath Tibet. Geophysical Journal of the Royal Astronomical Society, 72, 665– 689. Owens, T. J. & Zandt, G. 1997. Implications of crustal property variations for models of Tibetan plateau evolution. Nature, 387, 37– 43. Owens, T. J., Randall, G. E., Wu, F. T. & Zeng, R. 1993. Passcal instrument performance during the Tibetan Plateau passive seismic experiment. Bulletin of the Seismological Society of America, 83, 1959– 1970. Ozacar, A. A. & Zandt, G. 2004. Crustal seismic anisotropy in central Tibet: Implications for deformational style and flow in the crust. Geophysical Research Letters, 31, L23601; doi: 10.1029/2004GL021096.

124

J. MECHIE ET AL.

Phillips, W. S., Hartse, H. E., Taylor, S. R. & Randall, G. E. 2000. 1 Hz Lg Q tomography in central Asia. Geophysical Research Letters, 27, 3425–3428. Priestley, K., Debayle, E., McKenzie, D. & Pilidou, S. 2006. Upper mantle structure of eastern Asia from multimode surface waveform tomography. Journal of Geophysical Research, 111, B10304; doi: 10.1029/ 2005JB004082. Priestley, K., Jackson, J. & McKenzie, D. 2008. Lithospheric structure and deep earthquakes beneath India, the Himalaya and southern Tibet. Geophysical Journal International, 172, 345–362; doi: 10.1111/ j.1365– 246X.2007.03636.x. Randall, G. E., Ammon, C. J. & Owens, T. J. 1995. Moment tensor estimation using regional seismograms from a Tibetan Plateau portable network deployment. Geophysical Research Letters, 22, 1665– 1668. Rapine, R., Tilmann, F., West, M., Ni, J. & Rodgers, A. 2003. Crustal structure of northern and southern Tibet from surface wave dispersion analysis. Journal of Geophysical Research, 108, 2120; doi: 10.1029/ 2001JB000445. Reese, C. C. & Ni, J. F. 1996. Attenuation of coda waves in southern Tibet. Geophysical Research Letters, 23, 3015–3018. Rodgers, A. J. & Schwartz, S. Y. 1997. Low crustal velocities and mantle lithospheric variations in southern Tibet from regional Pnl waveforms. Geophysical Research Letters, 24, 9–12. Rodgers, A. J. & Schwartz, S. Y. 1998. Lithospheric structure of the Qiangtang Terrane, northern Tibetan Plateau, from complete regional waveform modeling: Evidence for partial melt. Journal of Geophysical Research, 103, 7137–7152. Romanowicz, B. 1982. Constraints on the structure of the Tibet plateau from pure path phase velocities of Love and Rayleigh waves. Journal of Geophysical Research, 87, 6865– 6883. Ruzaikin, A. I., Nersesov, I. L., Khalturin, V. I. & Molnar, P. 1977. Propagation of Lg and lateral variations in crustal structure in Asia. Journal of Geophysical Research, 82, 307– 316. Sandvol, E., Ni, J., Kind, R. & Zhao, W. 1997. Seismic anisotropy beneath the southern Himalaya-Tibet collision zone. Journal of Geophysical Research, 102, 17813– 17823. Sapin, M., Wang, X.-J., Hirn, A. & Xu, Z. X. 1985. A seismic sounding in the crust of the Lhasa block, Tibet. Annales Geophysicae, 3, 637–646. Saul, J. 2003. Untersuchung der seismischen Struktur von Zentral-Tibet und Indien mit teleseismischen Breitbandregistrierungen. PhD thesis, Free University, Berlin. Schulte-Pelkum, V., Monsalve, G., Sheehan, A., Pandey, M. R., Sapkota, S., Bilham, R. & Wu, F. 2005. Imaging the Indian subcontinent beneath the Himalaya. Nature, 435, 1222– 1225; doi: 10.1038/ nature03678. Sherrington, H. F., Zandt, G. & Frederiksen, A. 2004. Crustal fabric in the Tibetan plateau based on waveform inversions for seismic anisotropy parameters. Journal of Geophysical Research, 109, B02312; doi: 10.1029/2002JB002345.

Shi, D., Zhao, W. et al. 2004. Detection of southward intracontinental subduction of Tibetan lithosphere along the Bangong-Nujiang suture by P-to-S converted waves. Geology, 32: 209– 212. Shi, D., Shen, Y., Zhao, W. & Li, A. 2009. Seismic evidence for a Moho offset and south-directed thrust at the easternmost Qaidam–Kunlun boundary in the Northeast Tibetan plateau. Earth and Planetary Science Letters, 288, 329–334; doi: 10.1016/ j.epsl.2009.09.036. Sobolev, S. V. & Babeyko, A. Y. 1994. Modelling of mineralogical composition, density and elastic wave velocities in the unhydrous rocks. Surveys in Geophysics, 15, 515–544. Sol, S., Meltzer, A. et al. 2007. Geodynamics of the southeastern Tibetan Plateau from seismic anisotropy and geodesy. Geology, 35, 563– 566; doi: 10.1130/ G23408A.1. Solon, K. D., Jones, A. G. et al. 2005. Structure of the crust in the vicinity of the Banggong-Nujiang suture in central Tibet from Indepth magnetotelluric data. Journal of Geophysical Research, 110, B10102; doi: 10.1029/2003JB002405. Teng, J., Sun, K., Xiong, S., Yin, Z., Yao, H. & Chen, L. 1983. Deep seismic reflection waves and structure of the crust from Dangxung to Yadong on the Xizang Plateau (Tibet). Physics of the Earth and Planetary Interiors, 31, 293– 306. Tilmann, F. & Ni, J. & Indepth III Seismic Team. 2003. Seismic imaging of the downwelling Indian lithosphere beneath central Tibet. Science, 300, 1424– 1427. Unsworth, M., Wei, W. et al. 2004. Crustal and upper mantle structure of northern Tibet imaged with magnetotelluric data. Journal of Geophysical Research, 109, B02403; doi: 10.1029/2002JB002305. Unsworth, M. J., Jones, A. G., Wei, W., Marquis, G., Gokarn, S. G. & Spratt, J. E. & the Indepth-mt Team. 2005. Crustal rheology of the Himalaya and Southern Tibet inferred from magnetotelluric data. Nature, 438, 78– 81; doi: 10.1038/nature04154. Vergne, J., Wittlinger, G. et al. 2002. Seismic evidence for stepwise thickening of the crust across the NE Tibetan plateau. Earth and Planetary Science Letters, 203, 25– 33. Vergne, J., Wittlinger, G., Farra, V. & Su, H. 2003. Evidence for upper crustal anisotropy in the Songpan-Ganze (northeastern Tibet) terrane. Geophysical Research Letters, 30, 1552; doi: 10.1029/ 2002GL016847. Wang, Q., Zhang, P.-Z. et al. 2001. Present-day crustal deformation in China constrained by Global Positioning System measurements. Science, 294, 574–577. Wei, W., Unsworth, M. et al. 2001. Detection of widespread fluids in the Tibetan crust by magnetotelluric studies. Science, 292, 716– 718. Wittlinger, G., Masson, F. et al. & Lithoscope Kunlun Team. 1996. Seismic tomography of northern Tibet and Kunlun: Evidence for crustal blocks and mantle velocity contrasts. Earth and Planetary Science Letters, 139, 263– 279. Wittlinger, G., Vergne, J. et al. 2004. Teleseismic imaging of subducting lithosphere and Moho offsets

SEISMOLOGICAL STRUCTURE OF TIBET beneath western Tibet. Earth and Planetary Science Letters, 221, 117– 130. Wittlinger, G., Farra, V., Hete´nyi, G., Vergne, J. & Na´beˇlek, J. 2009. Seismic velocities in Southern Tibet lower crust: a receiver function approach for eclogite detection. Geophysical Journal International, 177, 1037–1049; doi: 10.1111/j.1365– 246X.2008. 04084.x. Xie, J., Gok, R., Ni, J. & Aoki, Y. 2004. Lateral variations of crustal seismic attenuation along the Indepth profiles in Tibet from Lg Q inversion. Journal of Geophysical Research, 109, B10308; doi: 10.1029/ 2004JB002988. Yin, Z., Teng, J. & Liu, H. 1990. The 2D crustal structure study in the Yadong-Damxung region of the Xizang plateau. Bulletin of the Chinese Academy of Geological Sciences, 21, 239– 245. Yuan, X., Ni, J., Kind, R., Sandvol, E. & Mechie, J. 1997. Lithospheric and upper mantle structure of southern Tibet from a seismological passive source experiment. Journal of Geophysical Research, 102, 27491– 27500. Zhang, Z. & Klemperer, S. L. 2005. West-east variation in crustal thickness in northern Lhasa block, central Tibet, from deep seismic sounding data. Journal of Geophysical Research, 110, B09403; doi: 10.1029/ 2004JB003139.

125

Zhao, J., Mooney, W. D., Zhang, X., Li, Z., Jin, Z. & Okaya, N. 2006. Crustal structure across the Altyn Tagh Range at the northern margin of the Tibetan plateau and tectonic implications. Earth and Planetary Science Letters, 241, 804–814; doi: 10.1016/ j.epsl.12005.11.003. Zhao, L.-S. & Helmberger, D. V. 1991. Geophysical implications from relocations of Tibetan earthquakes; hot lithosphere. Geophysical Research Letters, 18, 2205– 2208. Zhao, W. & Nelson, K. D. & Project Indepth Team. 1993. Deep seismic reflection evidence for continental underthrusting beneath southern Tibet. Nature, 366, 557– 559. Zhao, W., Mechie, J. et al. 2001. Crustal structure of central Tibet as derived from project Indepth wideangle seismic data. Geophysical Journal International, 145, 486–498. Zhao, W., Brown, L. et al. 2008. Seismology across the northeastern edge of the Tibetan plateau. EOS Transactions, American Geophysical Union, 89, 487. Zhao, W.-L. & Morgan, W. J. 1987. Injection of Indian crust into Tibetan lower crust: A two-dimensional finite element model study. Tectonics, 6, 489– 504. Zhu, L. & Helmberger, D. V. 1998. Moho offset across the northern margin of the Tibetan Plateau. Science, 281, 1170– 1172.

Rifting and strike –slip shear in central Tibet and the geometry, age and kinematics of upper crustal extension in Tibet LOTHAR RATSCHBACHER1*, INGRID KRUMREI1, MARLI BLUMENWITZ2, MARTIN STAIGER2, RICHARD GLOAGUEN1, BRENT V. MILLER3,4, SCOTT D. SAMSON3, MICHAEL A. EDWARDS1 & ERWIN APPEL2 1

Geowissenschaften, Technische Universita¨t Bergakademie Freiberg, 09599 Freiberg, Germany 2

Geologie, Abteilung Geophysik, Universita¨t Tu¨bingen, 72076 Tu¨bingen, Germany 3

Earth Sciences, Syracuse University, Syracuse, NY 13244-1070, USA

4

Geosciences, Texas A&M University, College Station, TX 77843-3148, USA *Corresponding author (e-mail: [email protected])

Abstract: The youngest deformation structures on the Tibet Plateau are about NNE-trending grabens. We first combine remote-sensing structural and geomorphological studies with structural field observations and literature seismological data to study the Muga Purou rift that stretches at c. 868E across central Tibet and highlight a complex deformation field. ENE-striking faults are dominated by sinistral strike–slip motion; NNE-striking faults have normal kinematics and outline a right-stepping en-echelon array of grabens, also suggesting sinistral strike–slip; along NW-striking fault sets, the arrangement of grabens may indicate a dextral strike– slip component. Thus, in central Tibet, rifts comprise mostly grabens connected to strike–slip fault zones or are arranged en-echelon to accommodate sinistral wrenching; overall strain geometry is constrictional, in which NNE–SSW and subvertical shortening is balanced by WNW–ESE extension. The overwhelmingly shallow earthquakes only locally outline active faults; clusters seem to trace linkage or propagation zones of know structures. The earthquake pattern, the neotectonic mapping, and the local fault– slip analyses emphasize a distributed, heterogeneous pattern of deformation within a developing regional structure and indicate that strain concentration is weak in the uppermost crust of central Tibet. Thus, the geometry of neotectonic deformation is different from that in southern Tibet. Next, we use structural and palaeomagnetic data along the Zagaya section of southern central Tibet to outline significant block rotation and sinistral strike–slip SE of the Muga Purou rift. Our analysis supports earlier interpretations of reactivation of the Bangong– Nujiang suture as a neotectonic strike–slip belt. Then, we review the existing and provide new geochronology on the onset of neotectonic deformation in Tibet and suggest that the currently active neotectonic deformation started c. 5 Ma ago. It was preceded by c. north–south shortening and c. east–west lengthening within a regime that comprises strike–slip and low-angle normal faults; these were active at c. 18–7 Ma. The c. east-striking, sinistral Damxung shear zone and the c. NE-trending Nyainqentanghla sinistral-normal detachment allow speculations about the nature of this deformation: the ductile, low-angle detachments may be part of or connect to a mid-crustal de´collement layer in which the strike– slip zones root; they may be unrelated to crustal extension. Finally, we propose a kinematic model that traces neotectonic particle flow across Tibet and speculate on the origin of structural differences in southern and central Tibet. Particles accelerate and move eastwards from western Tibet. Flow lines first diverge as the plateau is widening. At c. 928E, the flow lines start to converge and particles accelerate; this area is characterized by the appearance of the major though-going strike–slip faults of eastern-central Tibet. The flow lines turn southeastward and converge most between the Assam– Namche Barwa and Gongha syntaxes; here the particles reach their highest velocity. The flow lines diverge south of the cord between the syntaxes. This neotectonic kinematic pattern correlates well with the decade-long velocity field derived from GPSgeodesy. The difference between the structural geometries of the rifts in central and southern Tibet may be an effect of the basal shear associated with the subduction of the Indian plate. The boundary between the nearly pure extensional province of the southern Tibet and the strike–slip and normal faulting one of central Tibet runs obliquely across the Lhasa block. Published P-wave tomographic imaging showed that the distance over which Indian lithosphere has thrust under Tibet decreases from west to east; this suggests that the distinct spatial variation in the mantle structure along the collision zone is responsible for the surface distribution of rift structures in Tibet. Supplementary material: Containing supporting data is available at http://www.geolsoc.org.uk/ SUP18446. From: Gloaguen, R. & Ratschbacher, L. (eds) Growth and Collapse of the Tibetan Plateau. Geological Society, London, Special Publications, 353, 127–163. DOI: 10.1144/SP353.8 0305-8719/11/$15.00 # The Geological Society of London 2011.

128

L. RATSCHBACHER ET AL.

The youngest tectonic structures of the interior of the Tibetan Plateau are c. NNE-trending rifts (Fig. 1). Recent studies suggest that these rifts contain both normal and strike–slip faults and differ in geometry and extension rate in southern and central Tibet: Taylor et al. (2003) traced an east-trending zone of conjugate strike–slip faults across central Tibet, with NE-striking sinistral and NW-striking dextral faults north and south of the Bangong–Nujiang suture, respectively. These strike–slip faults appear kinematically linked with rifts north and south of the conjugate fault belt. The most distinct rifts occur in the Lhasa block and the northern Himalayan foreland (Fig. 1; Armijo et al. 1986, 1989; Mercier et al. 1987; Ratschbacher et al. 1994); in the Qiangtang terrane, they are less prominent. Molnar & LyonCaen (1989) and Langin et al. (2003) showed that focal mechanism solutions comprise mostly normal faults in southern Tibet and a combination of strike –slip and normal faults in central Tibet. Blisniuk & Sharp (2003), working on Quaternary slip rates along normal faults in central Tibet, proposed that extension rates are lower in central than in southern Tibet. Yin (2000) suggested that the spacing of the rifts decreases from the Himalaya to central Tibet and speculated that this difference may reflect the south –north decrease in crustal thickness across the Plateau (e.g. Zhao et al. 2001; Kind et al. 2002). Also the initiation of rifting is discussed controversially: Blisniuk et al. (2001) suggested that rifting in central Tibet initiated earlier or at c. 13.5 Ma. Similar ages (18.3– 13.3 Ma) stem from c. north-trending dykes and mineralized tension veins in the Lhasa block and the northern Himalayan foreland (Yin et al. 1994; Coleman & Hodges 1995; Williams et al. 2001). Discrimination of north–south differences in the late-stage structures across Tibet is held back by the incomplete knowledge of neotectonic deformation in central Tibet. In their study on conjugate strike–slip faults along the Bangong–Nujiang suture zone, Taylor et al. (2003) mapped the Yibug Caka rift (Fig. 1b); here, we study a morphologically well-defined neotectonic structure to its east, the Muga Purou rift (Figs 1 & 2). Its southern termination lies near the western tip of the dextral Gyaring Co fault, a segment of the Karakorum– Jiali fault system (Armijo et al. 1989); it stretches north to the southwestern segments of the Kunlun fault zone (Jolivet et al. 2003). We combine remotesensing structural and geomorphological studies based on LANDSAT, SPOT and CORONA images with structural and geomorphological field observations and published seismological data, and highlight a complex deformation field dominated by half grabens and strike–slip faults. Next, structural and palaeomagnetic data along the

Zagaya section outline significant vertical-axis block rotation and sinistral strike –slip shear SE of the Muga Purou rift, supporting Taylor et al.’s (2003) interpretation of reactivation of the Bangong–Nujiang suture as a neotectonic strike – slip belt. We review existing constraints and provide new U –Pb, Rb– Sr, K(Ar)– Ar ages on the onset of neotectonic deformation in southern and central Tibet. Together with a compilation of kinematic data from central and southern Tibet, we suggest that the currently active neotectonic deformation started c. 5 Ma ago. It was preceded by c. north–south shortening and east–west lengthening within a regime that comprises strike–slip and kinematically connected low-angle normal faults; these were active at c. 18–7 Ma. Finally, we propose a kinematic model that traces neotectonic particle flow across Tibet and speculate on the origin of structural differences in the neotectonic deformation pattern in southern and central Tibet.

Muga Purou rift The c. 330 km long Muga Purou rift consists of several NE- and NNE-striking grabens (Figs 1 & 2). Along its southern and central parts, graben structures are well-developed with recent fault activity traceable on satellite images (Fig. 2); towards the north, extensional structures are less developed.

Southern Muga Purou rift Both the Yibug Caka and the Muga Purou rifts do not cross the Bangong –Nujiang suture (Fig. 1b). Along its southern end, the Muga Purou rift splits into two segments with a weakly-developed western and a well-developed eastern graben. The latter is characterized by high relief, with mountain ranges rising c. 1 km above the basin floor (Fig. 3). Contrary to the grabens further north (see below), this basin developed within a symmetric graben. The overall trend of the basin and its defining faults changes from NNE to NE from south to north (Fig. 3a). The eastern boundary fault corresponds to an abrupt increase in slope, suggesting significant throw. Along the footwall, a series of triangular facets separate short and steep valleys (Fig. 3a). The largest facets are degraded; they are outlined by the watersheds of the valleys between the smaller and mainly active facets. Along the range-front of the glacier covered mountains of the western side of the basin, the slope contrasts are smoother than along the eastern range (Fig. 3a). Here, the range-bounding fault is inactive, as fault scarps are degraded; extension is accommodated along short (350–1500 m) fault

Fig. 1. (a) Colour-coded slope map of Tibet highlighting the contrast between the well-expressed rifts in southern Tibet and the distributed deformation consisting of sinistral strike– slip faults and grabens in central Tibet. Colour-shaded relief map of the inset gives major tectonic blocks of Tibet and the bounding sutures (from south to north Yarlung– Tsangpo, Bangong–Nujiang, Jinsha). These sutures are outlined as low-slope areas. (b) Neotectonic map of Tibet with main study areas (Muga Purou rift, Zagaya section) and names used in the text; numbers locate Figures that outline details of the rift geometry. Modified mainly from Armijo et al. (1986, 1989), Murphy et al. (2002), Murphy & Copeland (2005), Taylor et al. (2003), Wang et al. (2008), and own mapping and interpretations.

RIFTING AND STRIKE –SLIP SHEAR IN TIBET 129

Fig. 2. (a) Muga Purou rift: the Shuang Hu graben is situated in its central part and forms the most prominent neotectonic feature. The rift consists of further four basins toward the south that are all occupied by lakes and two basins further north. (b) The northern Shuang Hu normal fault system. Fault traces are outlined by scarps, relief contrasts, triangular facets, and end points of valleys or gullies. A–C, fault indices; E, eastern boundary of the graben. (c) Range-front faulting along the southern Shuang Hu graben. The range-front fault follows in its northern and southern parts the base of degraded triangular facets; in its central part it is inactive. Most of the fans show at least one older depositional surface that has a lighter surface colour than the active surface. The active fault has migrated into the basin.

130

L. RATSCHBACHER ET AL.

Fig. 3. NNE-striking eastern graben along the southern termination of the Muga Purou rift. (a) The central part of the eastern graben boundary is characterized by a sharp linear contrast between the rugged topography of the range and the flat valley floor. The valleys entering the graben at the southern and northeastern boundaries are long in contrast to the valleys in the central part. The central part of the western graben boundary reveals less steep, better developed alluvial fans than on the eastern side and an inactive sedimentation surface. The active faults bounding this surface toward the east are both normal and strike– slip. (b) Detail framed in (a) showing superposition of normal and strike– slip faults. Slope is to the ESE. Black arrows mark V-shaped surface ruptures suggesting strike– slip kinematics. A and B mark faults cutting fan heads.

segments c. 100 m closer towards the graben centre. Some scarps are outlined by a ‘V-shaped profile’, probably indicating strike–slip components (arrows in Fig. 3b). Others step down toward the east, revealing a significant normal-slip component. Fans show sedimentation along their distal side

and incision of the fan head. Several scarps cut fan heads, suggesting recent activity (sites A and B in Fig. 3b). Longer (WNW– ESE) but narrower (NNE –SSW) fans and higher mountain peaks occur in the north and may indicate a southward tilting of the footwall block. The western

RIFTING AND STRIKE –SLIP SHEAR IN TIBET

graben-bounding fault splays into three strands in the northwestern part of the basin (Figs 2 & 3a); the westernmost, topographical highest strand separates the mountain range from the large alluvial piedmont, the two eastern strands cross the fans. The western and eastern grabens of the southern Muga Purou rift are linked with the central rift by several basins (Fig. 2). Toward the north, these basins are bounded by east-striking mountain ranges, following folds and thrusts affecting Mesozoic rocks. The Cenozoic extensional structures formed at high angles (c. 60–908) to these structures with normal faults striking north and NNE. Strike – slip offsets along these faults cannot be deduced from morphological features, but the alignment of these NNE-striking grabens to the overall trend of the Muga Purou rift suggests a sinistral component for the southern rift section.

Shuang Hu Graben In the central part of the Muga Purou rift, the c. 50 km long and c. 10 km wide asymmetric Shuang Hu graben has topographic relief of c. 1 km along the western range front. The graben comprises a complex fault zone along its western side (Fig. 2b, c) and minor faults along its eastern side. In the northern part, the western fault zone is c. 3.5 km wide and comprises three major strands of normal faults (A, B, C in Fig. 2b), which become younger from west –east, as basin sediments and alluvial fans are progressively less incised from west to east. Fault B divides the alluvium into an upper western part and a lower eastern part; the former is deeper incised than the latter. The active sedimentation plain lies east of fault C. Gullies and small valleys often start along the fault scarps of fault B. The youngest fault C offsets Quaternary river terraces. At the northern and southern end of the fault zone, faults B and C merge into a singular range front fault A. Triangular facets are best developed along the southern range front fault, setting a sharp contrast between the more incised landforms of the footwall and the actively aggrading landforms in the hanging wall; the corresponding topographic relief of c. 1 km suggests a significant dip–slip component. Valleys with wine-glass geometry at the fault trace separate the faceted spurs and are also consistent with dip–slip motion. A complex pattern with small fresh and larger degraded or inactive facets suggests a heterogeneous fault–slip history along fault A. From offset Mesozoic strata, Yin et al. (1999) estimated a total normal slip offset of c. 7 km across the southern segment fault A, and inferred that that the initiation age of the fault may be no older than c. 4 Ma.

131

Large alluvial fans, forming a bajada, characterize the NE-striking mountain front in the southwestern part of the Shuang Hu graben (Fig. 2c). Between the mountain tops and the valley floor lie c. 1 (north) and c. 1.5 km (south) of relief. The range-front fault follows in its northern and southern parts the base of degraded triangular facets; it is inactive in its central part (A in Fig. 2c). Here, Blisniuk et al. (2001) observed at four locations fault-zone mineralization and obtained geochronological data that suggest range-front fault activity at or before c. 13.5 Ma and reactivation at c. 4 Ma (see below). East of the inactive range-front fault, the fan heads are incised and active sedimentation occurs along the lower reaches of the alluvial fans. Most of the fans show at least one older depositional surface that has a lighter surface colour on the satellite image than the active surface (Fig. 2c). Another abandoned fault strand runs along a line marking strong and moderate erosional incision at the upper fan heads. A well-developed fault associated with clear and fresh scarps separates the incised and active fan portions. Using U-series dating, Blisniuk & Sharp (2003) dated pedogenic carbonate pebble-rinds from two well-preserved terrace surfaces offset by this fault and obtained minimum ages of c. 16.4 and c. 233 ka for deposition of the two terraces (see below). The fault pattern in the Shuang Hu graben may indicate stronger extension in its southern than northern part, with stronger footwall uplift, significant basin-ward migration of faults, and a lake-occupied basin at lower elevation in the hanging wall in the southern rift segment.

Structural field data from the Shuang Hu normal fault system We collected fault–slip data across the northwestern Shuang Hu graben from outcrops of strata of known or assumed age and uniform lithology (Fig. 4). Slip sense was deduced from kinematic indicators such as offset markers, fibrous minerals grown behind fault steps, Riedel shears, and en-echelon tension gashes and fractures. Together with the fault data, we recorded the orientations of bedding, tension gashes and fibrous minerals growing within veins. Indications of multiple slip were recorded, and the relative chronology was used for separation of heterogeneous raw data fault sets into subsets. We used the computerprogram package of Sperner et al. (1993) and Sperner & Ratschbacher (1994) to calculate the orientation of principal stress axes and the reduced stress tensors (Angelier 1984). Out of this package, we obtained stress axes by the ‘pressuretension (P –B– T) axes’ method and calculated

132 L. RATSCHBACHER ET AL. Fig. 4. Structural data, idealized columnar section through the stratigraphy, orientations of kinematic axes, types of solutions, and stress geometries from the Shuang Hu graben. Structural field data of Carboniferous to Quaternary rocks comprise deformation structures related to Cenozoic stress fields (grey background: likely Oligocene –Miocene; white background: likely Late Miocene–Recent). The grey stereograms (lower hemisphere, equal area) portray c. north– south shortening and mostly c. east– west lengthening (see rose and shape-factor diagrams) that is mostly accommodated by conjugate strike– slip faulting, which predates the neotectonic normal faulting with c. east–west stretching (white stereograms). Faults are drawn as great circles and slip arrows with heads pointing in direction of displacement of the hanging wall. As errors in slip-sense determination have severe effects on stress calculations, degrees of certainty were assigned to each observation expressed in the head style of the slip arrows: full: certain; open: reliable; half: unreliable; without head: very poor slip-sense determination. Numbers 1 to 3 are the orientations of principal stresses, s1  s2  s3. s, bedding; s1, first foliation; tf, tension fracture; tg, tension gash.

RIFTING AND STRIKE –SLIP SHEAR IN TIBET

133

Fig. 4. Continued.

stress tensors by the ‘numerical dynamic analysis’. The quality and the quantity of field data determined the selection of the method used for calculation. The P– B–T axes method was used with scarce data. Table 1 lists the location of stations and the parameters of the deviatoric stress tensor. The supplementary material details the lithostratigraphy of the analysed sites and summarizes deformation

observations. Figure 4 displays the structural data, an idealized columnar sections through the stratigraphy of the Shuang Hu graben area, orientations of the kinematic axes, type of solutions, and stress geometries [reduced stress tensor, R ¼ (s2 2 s3)/ (s1 2 s3)]. North–south shortening within the footwall of the Shuang Hu normal fault system is recorded by

134

Site

Lithology

Shuang Hu graben 7R3 Red sandstone (Neogene) 14R1 Fluvial conglomerate (Quaternary) 15R1old Siltstone-conglomerate (Miocene?) 15R1you Siltstone-conglomerate (Miocene?) 15R1ew Siltstone-conglomerate (Miocene?) 15R2 Metavolcanic rocks (Jurassic) 15R3old-tg Shale with gypsum (Pliocene) 15R3you-tg Shale with gypsum (Pliocene) 15R3 Shale with gypsum (Pliocene) 15R4 Fluvial conglomerate (Quaternary) 15R5 Fluvial conglomerate (Quaternary) 16R1old Cataclasite of sandstone and dolomite 16R1you Cataclasite of sandstone and dolomite 16R2-3 Volcanoclastic sandstones (Oligocene?) 16R4-5c Shale, phyllite, quartzite (?Permian, Carboniferous) 17R2-3you Greenschists, marble, conglomerate 17R4 Marble (?Permian) 18R1old Limestone, marl, sandstone (Jurassic) 18R1you Limestone, marl, sandstone (Jurassic) 18R1you-tg Limestone, marl, sandstone (Jurassic) 18R2a Red conglomerate (?Miocene)

Latitude

Longitude

Method

33824.093 0 33818.2860 33813.742 0 33813.7420 33813.7420 33814.5220 33814.3760 33814.3760 33814.3760 33812.980 33812.579 33808.4660 33808.4660 33808.530 0 33809.2 0

88840.269 0 88.48.9380 88844.305 0 88844.3050 88844.3050 88842.0520 88842.6890 88842.6890 88842.6890 88845.7130 88846.2050 88845.0810 88845.0810 88844.646 0 88843.5 0

n.c. n.c. NDA NDA n.c. NDA n.c. n.c. P–B–T n.c. n.c. P–B–T P–B–T NDA NDA

33816.617 0 33817 0 33822.614 0 33822.6140 33822.6140 33822.637 0

88840.29 0 88840 0 88847.989 0 88847.9890 88847.9890 88847.653 0

NDA NDA NDA NDA n.c NDA

#

s1

s2

s3

F

R

168 148

0.4 0.6

248

0.4

20 19

0.2 0.1

17 15 24 188

0.6 0.5 0.5 0.5

10

0.4

030 00 33 32 11 11

131 64 212 12

10 10

034 03 075 74 315 00 008 49

77

149 69

006 17

99 55 10 10 55

003 21 321 66 191 16 145 24

223 60 118 22 288 22 034 38

88 12 12 19 19 23 22

003 20 193 15 193 06 126 69

137 62 022 74 098 41 001 13

12 12

168 04

076 14

210 38

075 00 302 26 304 11 045 00 111 11 300 27 250 20 272 13 090 00 120 00 111 17 212 09 068 62 258 42 266 18 284 02 290 48 267 17 243 36 274 75

L. RATSCHBACHER ET AL.

Table 1. Location of stations and parameters of the deviatoric stress tensor

18R2b 18R2you 19R1-2old 19R1-2you 19H46-48

Red conglomerate (?Miocene) Red conglomerate (?Miocene) Red conglomerate (?Miocene) Red conglomerate (?Miocene) Metavolcanics, slate (?Permian)

88847.653 0 88847.6530 88841.7 0 88841.70 88842 0

31852.2620

91841.9320

32811.4270

91842.4490

0

32804.572 32804.572 0 32804.70 c. 328120 c. 328120

0

89840.001 89840.001 0 89839.50 c. 898360 c. 898360

NDA NDA NDA NDA NDA

99 15 15 88 88 10 10

231 44 140 70 004 11 001 47 000 17

048 46 008 14 123 68 221 35 251 46

140 02 274 15 270 19 116 21 104 39

23 158 16 19 29

0.6 0.5 0.2 0.5 0.4

NDA

10 10

178 73

358 17

268 00

10

0.5

NDA

44

129 86

019 01

288 04

12

0.6

NDA NDA NDA

16 16 11 11 66

203 17 139 04 184 78

295 04 047 28 015 12

036 73 235 61 284 02

11 11 11

0.4 0.4 0.5

NDA P–B–T

88 11 11

030 60 200 01

170 10 099 86

272 10 290 02

09

0.5

Abbreviations: P–B –T, pressure –tension axes method; NDA, numeric dynamic analysis technique; n.c., principal sub-horizontal stress orientation obtained from visual inspection of the data. Note: In the measurement column (#) first number is number of measurements, second is number of measurements used for calculation. For s1 2 s3, azimuth (first number) and plunge (second number) of the principal stress axes are given. The stress ratio R is (s2 2 s3)(s1 2 s3)21 (where 1 is uniaxial confined extension, 0 is uniaxial confined compression). The fluctuation F gives the average angle between the measured slip and the orientation of the calculated theoretical shear stress. Italics denote syn- to post-Miocene c. north –south shortening.

RIFTING AND STRIKE –SLIP SHEAR IN TIBET

Pung Co and Zagaya grabens 4R1 Retrograde amphibolite, gneiss, Kfs-dyke (Pung Co rift) 4R2 Hydrothermally altered granitoid (Pung Co rift) 5R1old Eocene red bed sandstone (Zagaya rift) 5R1you Eocene red bed sandsteone (Zagaya rift) 5R1-2 Eocene volcanic and Neogene mudstone (Zagaya rift) 6-E8 Eocene red beds (Zagaya rift) 6-E9 Eocene red bed conglomerate (Zagaya rift)

33822.637 0 33822.6370 33821 0 338210 33816.2 0

135

136

L. RATSCHBACHER ET AL.

strata as young as Miocene. Open to locally tight folding is associated with mostly strike– slip faulting and was observed in all stations in the footwall of the Shuang Hu graben (Fig. 4). s1 and s3 calculated from the data in these stations trend 011+208 and 107+178, respectively (10 sites). s2 is mostly subvertical and in thrust solutions the stress-ellipsoid shape factor is oblate with vertical and along strike lengthening. A younger event also effects the upper part of Miocene– Pliocene Kangtog Formation (mudstones with gypsum; unpublished Chinese Survey mapping and own observations, Fig. 4); this event includes the active faults. Most faults are normal; s3 trends 086 + 298 (16 sites). At one site in the Kangtog Formation, overprinting generations of tension gashes and fibre demonstrate two consecutive extension events; s3 of the older and younger events trend 1208 and 708, respectively. At the southern end of the northern Shuang Hu normal fault system, we studied a valley transect across the range-front fault. Close to the southern termination of fault A (16R1, Fig. 4), a 5 to .10 m thick cataclastic fault zone in red sandstone and locally dolomite is unconformable overlain by fluvial conglomerates. A fluvial terrace relict in the fault zone contains components from the cataclastic fault zone; the terrace relict is cemented by carbonate and dips up to 138. An older event is dominated by sinistral strike –slip on NE-striking faults (s1 ¼ 0038, s3 ¼ 1118). A younger event shows normal and strike– slip faults with dominant north-strike (s1 ¼ 3218, s3 ¼ 2128), suggesting a clockwise rotation of s3 in time. Faults at stations 16R2-3 affect probably Oligocene red volcanoclastic and conglomeratic sandstone covering the hanging wall of the Jurassic Falong detachment (Kapp et al. 2003); the station is in the footwall of the range-front fault. The fault pattern may be a mixture of several events, that is, north– south shortening by folding and conjugate strike –slip faulting with WNW–ESE extension. A valley section further north crosses the Quaternary faults of the bajada, the range-front fault, and the footwall of the Shuang Hu graben. The upper part of the Kangtog Formation outcropping in the footwall is only affected by east–west lengthening. Here, it consists of gypsum-rich lake beds and predominantly siltstones with local sandstone channels. Stable deposition of the siltstones is indicated by undisturbed, thin beds. The 1–3 cm thick siltstone beds show cross bedding with flow to the SSE. The locally 5–10 m thick sandstone channels show flow toward SW, load casts, and slumping on a 1– 10 cm scale. Stress directions calculated from fault–slip data correspond to those inferred from tension gash– gypsum fibre pairs (s3 ¼ 2728); however, frequent curving fibres may

indicate a counterclockwise rotation of s3 in Neogene time (s3old ¼ 1208, s3young ¼ 708, Fig. 4, 15R3). Assuming a planar, rotating normal fault model (e.g. Twiss & Moores 1992), we calculate local extension of up to 145% from average bedding and normal fault dips. Site 15R1 stretches along the river gorge from the inner range front at fault A to 700 m upriver (Fig. 4) and again encompasses the Kangtog Formation; here, it comprises from bottom to top marly sand-siltstones, massive conglomerates, and shale and evaporites. The conglomerates contain granite and limestone pebbles up to 70 cm in diameter and were shed from NW, as indicated by cross bedding, ripples and slump folding; their likely source are the sedimentary rocks and granitoids of the Qiangtang anticlinorium west of Shuang Hu (Kapp et al. 2000). Deformation increases toward the range-front fault zone and includes an older strike –slip fault dominated event (Fig. 4, 15R1old) and a younger one, in which normal faults dominate (15R1you); both have a c. 1208 trend of s3. Bedding dips up to 258 toward WNW, probably as a result of the normal faulting (15R1); average bedding and normal fault dips would imply local extension of up to 130%. Faulting is accompanied by minor fluid activity; faults are in general polished and grooved surfaces are coated with limonite. Cataclasite is locally developed. Tension gashes are filled with blocky carbonate and locally malachite. Assuming that the 130% extension value is representative for the c. 30 km wide Shuang Hu graben and that normal faulting initiated at 4 Ma (see below) would yield a maximum extension rate of 4 mm a21. At the range front, a major fault zone juxtaposes conglomerate (toward W) against sandstones and shale (toward E; 15R1 – map view sketch, Fig. 4). The interaction of sinistral and dextral faults of the strike –slip fault event may have caused the local en-echelon range-front segmentation. Fractures in pebbles of the alluvial fan deposits collected at two sites close to fault C (15R4 and 15R5) constrain the s3 trend at 758 (15R4) and 1278 (15R5). At site 14R1, conglomerates of a cemented post-glacial alluvial fan are topped by a recent soil horizon, which is unfaulted; this indicates the lack of historic earthquakes. The conglomerates show possibly earthquake-induced syn-sedimentary deformation structures (slumping, sagging of pebble-rich layers) and are cut by tension fractures; assuming a purely extensional opening, s3 trends c. 748. Stations 18R1 and 18R2 are located in Jurassic rocks that are unconformably overlain by conglomerates of the likely Miocene part of the Kangtog Formation; these in turn are covered by coarse fluvial conglomerates. These stations are close to and along the range-front fault. Here, it has stepped westward, changed to NNW-strike, is little

RIFTING AND STRIKE –SLIP SHEAR IN TIBET

expressed morphologically, and does not show a recent scarp. An early fault set is related to folding of the Jurassic and Miocene rocks along c. east– west axes. A young normal faulting event has the largest faults, which include shallowly-dipping ‘detachment’ faults with likely considerable displacement. Most of the fault– slip data trace the NNW-strike of the range-front fault and have a dextral slip component. West of the topographic crest of the Shuang Hu graben footwall, we encountered distributed normal faulting that overprints folding and faulting related to c. north– south shortening. Map-scale oblique–slip normal faults deform the monocline outlined by massive conglomerates (Kangtog Formation) along the northern rim of the glaciercovered dome east of the Shuang Hu graben. These faults outline a horst formed early in the extension history; extension later concentrated on its eastern side, forming the Shuang Hu half graben.

Neotectonic structures west of the Shuang Hu graben West of the Shuang Hu graben, the surface trace of a sinistral strike– slip fault marks the east– west axis of a large shallow basin (Fig. 5a). A c. 2 km long

137

and c. 0.7 km wide sag is interpreted as a small pullapart basin (‘pa’ in Fig. 5b), underpinning the sinistral strike –slip interpretation. Another pull-apart structure may be outlined by the largest lake of the basin further to the east. Flow directions of rivers and degraded fault scarps suggest that the basin started to form at its northwestern margin. The eastern end of the strike –slip fault links with an extensional system with NE-striking faults. There, the basin floor lies deeper than along the strike – slip fault. Triangular facets west of the NE-striking fault trace and large alluvial fans in the piedmont characterize the steep northwestern boundary of this area. Kapp et al. (2003) related these faults to the Qiagam normal fault system, which bounds the western margin of a Quaternary lacustrine basin against schists and gneisses of the Qiangtang anticlinorium. The southeastern boundary of the basin shows a shallower relief than the northeastern boundary, suggesting weak fault activity. These normal fault system marks the termination of the strike –slip fault; the faults curve toward the receding block, suggesting that the strike– slip fault turns into an imbricate fan of faults (Fig. 5c). The antithetic SE-dipping fault fits into this fan when extension is large enough to generate a normal fault also in the hanging-wall block. Large northand NNE-striking valleys west of the active

Fig. 5. (a) Strike–slip fault system west of the Shuang Hu graben. (b) Pull-apart (pa) basin along the fault indicates sinistral kinematics. (c) The eastern normal faults mark the termination of the strike–slip fault zone in (a) and are interpreted as an extensional imbricate fan.

138

L. RATSCHBACHER ET AL.

imbricate normal faults may represent an older extensional termination of the strike –slip fault that became inactive in an early state of extension; there, only narrow valleys and poorly-developed basin formed.

Neotectonic structures of the northern Muga Purou rift North of the Shuang Hu graben, young structures are less prominent. They are isolated and cannot be combined in one large regional rift structure. Mainly east-striking, 5300–5600 m high mountain ranges, tracing folds and thrusts in Mesozoic rocks, dominate the landscape (Fig. 6). Small basins form local base levels at 4.9 –5 km a.s.l. Geomorphological indicators for c. NW-striking faults

bounding these basins are the truncation of the mountain ranges at high angles, and high relief contrasts between ranges and basin margins with triangular facets and alluvial fans. Most of the larger faults dip southwestward, forming half grabens, and are situated at the northeastern side of valleys and basins. We mapped these faults as mostly normal. Individual faults can be grouped in three major clusters (1–3 in Fig. 6) that form grabens; their c. NNW-trend and en-echelon arrangement suggest a dextral strike –slip component along the entire NW-striking structure. North of 348, young extensional structures are scarce; only a few rift structures are identifiable on the satellite images. Around 888050 E, 348250 N, a 8 km wide and 40 km long basin is surrounded by low topographic relief (4.8–5.4 km, mainly at 5.1 km; Fig. 7). The shallow lake consists of three,

Fig. 6. NW-striking fault system north of the Muga Purou rift. The faults cut c. east-trending fold-thrust structures at high angle and produce small basins that are used by rivers or filled with lakes. Basin at 2 is bound at is southwestern side by a NE-dipping fault that is lined with hot springs. Lingga Co seems to be tectonically affected, outlined by steep shorelines and a deep canyon entering the lake; a fault to its east bounds a huge floodplain. Fault clusters 1 to 3 outline three grabens that suggest a dextral strike–slip component along the entire NW-striking structure by their NNW-trend and en-echelon arrangement.

RIFTING AND STRIKE –SLIP SHEAR IN TIBET

139

Fig. 7. Major sinistral strike–slip fault zone north of the Muga Purou rift. In the NW, a discrete range-bounding fault is outlined by a sharp lineament, defined by changes in surface reflectivity, a boundary between erosion and sedimentation sites, and a change in slope. Dextral faults bound different parts of the central lake; they were used to infer the sinistral slip sense of the range bounding fault.

rhomb-shaped sub-basins. This pattern suggests a series of en-echelon dextral strike –slip faults. The northern boundary of the basin is fault controlled with a sharp lineament between the basin sediments in the southeast (dark grey on Landsat TM) and the bedrock in the northwest (light grey). The eastern part of the northern boundary lacks a pronounced lineament. Because of the low relief between the basin and the ranges to the northwest and the en-echelon dextral strike–slip faults outlining

the sub-basins of the lake, this large northern fault should have a significant sinistral component. At the southern side of the basin, evidence for faulting is sparse; faults are indicated by the abrupt termination of E-striking bedrocks ridges (Fig. 7). To the east, around lake Dogai Coring (Figs 1b & 8), rare NW-striking normal faults can be traced by linear shorelines and river courses, and fault scarps; the latter affect Late Cretaceous –Tertiary red beds of the Dogai Coring fold–thrust belt. NW of

Fig. 8. (a) The area SE of Dogai Coring shows moderate relief. NW-trending neotectonic structures are outlined by the drainage system and linear lake shorelines. (b) Field photograph; trace of the normal faults marked with 1 and 2 in (a).

140

L. RATSCHBACHER ET AL.

Dogai Coring, Cenozoic structures are weakly developed; relief is moderate. Cenozoic volcanic rocks (5–9 Ma, unpublished own whole rock 40 Ar/39Ar ages) cover discordantly the northern part of the Dogai Coring thrust-fold belt (vp, Fig. 9a). Faults with clear dip–slip movement, as indicated by relief contrast and triangular facets, trend north or east (G, H, F in Fig. 9a). The single normal fault G forms a half graben at the western end of a large plane (B) in the hanging wall. Geomorphological features indicating strike –slip motion occur only along NE-striking faults (e.g. E in Fig. 9a). The area around 358200 N, 888350 E is dominated by a continuous, east-trending fault (Fig. 9b), which is accompanied by moderate to low relief. South of a large lake (L, Fig. 9b), the fault cuts a NW-striking, 5.3 km high mountain range; the lake itself lies at 4.8 km. The distinct narrow trace and the lack of relief across the fault suggest strike– slip motion. However, it was not possible to determine the slip sense with the help of geomorphological markers; there are neither offset ridges nor beheaded streams. Rivers along the fault curve into and shortly follow the fault trace, but there is no consistent offset pattern. Lineaments further north are interpreted as faults striking east. All these faults

are inferred to be part of the western termination of the Kunlun fault zone.

Kinematics of faulting Earthquake epicentres Seismicity data of central Tibet, used herein, are from project INDEPTH III seismic recordings (Langin et al. 2003), and the USGS-PDE and Harvard-CMT catalogs. Although several of the documented events can be associated with faults and graben systems traced in this study (Fig. 10), most of the earthquake clusters cannot. Langin et al. (2003) described two event clusters (1 and 2, Fig. 10) located in the southern and central Muga Purou rift. Cluster 1 may be associated with the southern Muga Purou branch, although the cluster lies to the east of the major eastern graben-bounding fault. Cluster 2 was ascribed to the Shuang Hu normal fault system; however, this cluster lies between the Shuang Hu graben and the eastern end of the sinistral strike– slip fault and graben system west of the Shuang Hu graben, a mountainous region devoid of active faults recognizable on the satellite images. The events have depths

Fig. 9. (a) Structures NW of Dogai Coring (Fig. 1b). A volcanic platform (vp) covers discordantly the northern part of the Dogai Coring fold-thrust belt. Faults G and H constitute two north-striking grabens, other grabens strike c. east. (b) Large strike– slip fault have inferred sinistral motion due to focal-mechanism evidence along strike further west; the fault is interpreted as part of the western termination of the Kunlun fault system. Rivers crossing the fault do not reveal a uniform offset pattern to independently establish slip-sense.

RIFTING AND STRIKE –SLIP SHEAR IN TIBET

141

Fig. 10. Distribution of earthquakes between 298300 N, 868E and 358300 N, 928E plotted on slope map of the area. Although several of the documented events can be associated with mapped faults and graben systems, most of the earthquake clusters cannot. Clusters seem to trace linkage (e.g. 2) or propagation zones of the know structures (e.g. 3, 4). The earthquake pattern emphasizes the distributed, heterogeneous pattern of deformation within a developing regional structure and indicate that strain concentration is weak in the uppermost crust of central Tibet.

between 5 and 18 km and the faults east and west of the cluster do not lie within the estimated horizontal error of 1.7 km. The hypocentres might be associated with a mid–upper crustal detachment fault in which the active structures root. Event clusters 3 and 4 east the Muga Purou rift and cluster 5 to its northeast do not correlate with mapped surface faults (Fig. 10). Clusters 3 and 4 occupy the projected intersection of the northern continuation of the NE-striking Shuang Hu graben with the southeastern prolongation of the NW-striking normal fault system north of it; thus, these clusters may mark the propagation and interference of these

segments of the Muga Purou rift. Cluster 6 seems to trace the westward propagation of the northernmost strike –slip fault that we mapped and that may be associated with the southwestern tip of the Kunlun fault system.

Seismotectonics Figure 11a shows lower hemisphere projections of 41 focal mechanisms for the same area as Figure 10, taken from the INDEPTH III data set and the USGS-PDE and Harvard-CMT catalogs. Generally, the focal mechanisms indicate a

142

L. RATSCHBACHER ET AL.

Fig. 11. (a) Focal mechanisms plotted for the same areas as in Figure 10 (298300 N, 868E and 358300 N, 928E). They mostly indicate a combination of normal and strike– slip faulting; the normal-fault solutions outline NNE-trending faults with varying components of sinistral strike– slip, and the strike– slip solutions indicate sinistral motion, mainly along NE-striking faults. The focal mechanisms corroborate the field findings that normal faults in central Tibet have variable sinistral strike–slip components. (b) Fault–slip data solutions from the Shuang Hu graben plotted as ‘beach balls’ for comparison with (a).

combination of normal and strike–slip faulting, in agreement with the statements of Yin et al. (1999) and Blisniuk et al. (2001) that large north- to NE-trending normal faults in central Tibet have sinistral strike –slip components. In detail, the normal-fault solutions outline NNE-trending faults with varying components of sinistral strike –slip. The strike –slip solutions indicate sinistral motion, mainly along NE-striking faults. The few focal mechanisms falling within the NW-striking fault array northwest of the Shuang Hu graben (a to d,

Fig. 11a) show normal and reverse kinematics with a sinistral component on east- and NE-trending faults. The two focal mechanisms from the PDE catalog (e and f; grey) underline the normal kinematics. North of 358N, focal mechanisms indicate sinistral strike–slip faulting, supporting our inference that faults mapped in this area are part of the southwestern termination of Kunlun fault zone (Fig. 11a). Figure 12a plots P- and T-axes derived from the INDEPTH III, PDE, and CMT focal-mechanism

RIFTING AND STRIKE –SLIP SHEAR IN TIBET

143

Fig. 12. Lower hemisphere, equal area projections of P – T axes computed from INDEPTH III, PDE, and CMT focal mechanism solutions (a) compared with those derived from the fault– slip data of the two events recognized within the Shuang Hu graben (b); only the young event can be directly compared with the focal mechanism data. The pattern traces the heterogeneous nature of Late Cenozoic upper crustal deformation in central Tibet.

data. Most of the P-axes are arranged along a great circle trending NNE; the main cluster centres at 012 34 (+78 in trend, s1), close to the India –Asia convergence direction inferred from GPS-geodetic data (c. 0208, Zhang et al. 2004). However, the INDEPTH III data also contain NW –SE shortening solutions; the cluster centres at 121 03 (+128 in trend, s1). The T-axes indicate ENE –WSW lengthening; s3 centres at 100 07 (+118 in trend), but again there is a great circle with vertical and NE-trending axes. The major cluster is at high angle to the strike of normal faults in the Shuang Hu graben. The P–T-axes calculated from the fault –slip data collected along the Shuang Hu normal fault system match the focal-mechanism data (Fig. 12b), emphasizing the heterogeneous nature of the faulting in central Tibet. Only the data of the younger event, which includes active faults, may be directly compared with the earthquake data. Nevertheless, also during the older event, the T-axes indicate ENE– WSW lengthening, suggesting that even during this

Miocene and older pre-rifting phase (see above), NNE –SSW shortening was mainly compensated by tangential stretch.

Sinistral strike – slip south of the Muga Purou rift along the Bangong – Nujiang suture zone: the Zagaya section Taylor et al. (2003) traced a c. east-trending zone of (E)NE-striking, sinistral strike– slip faults across southern central Tibet that is located north and along the Bangong –Nujiang suture; these strike– slip faults appear kinematically linked with the central Tibetan rifts. Here, we analyse the near-field deformation of one of these strike –slip fault zones structurally and palaeomagnetically.

Geology of the Zagaya section The Zagaya river drains the area NE of Siling Co (Figs 1b & 13a). The studied section comprises

144

L. RATSCHBACHER ET AL.

Fig. 13. (a) Siling Co area and Zagaya section with the northern limit (dashed line) of me´lange rocks of the Bangong– Nujiang suture, approximately defining the boundary of the Lhasa and Qiangtang blocks, interpretation of neotectonic structures, and sample locations. (b) Overview geological map with structural data and palaeomagnetic sample sites. Lower hemisphere, equal area stereograms with B, fold axis; s0, bedding; tf, tension fracture. Fault symbols as in Figure 4.

RIFTING AND STRIKE –SLIP SHEAR IN TIBET

the northernmost Lhasa block and the Bangong– Nujiang suture zone. The sequence consists of reddish sandstones interbedded with conglomerates and dark red to black silt- and mudstones. It is imbricated with ophiolitic me´lange, including serpentinite and serpentinized ultramafic material, marble and chert breccia layers. To the south, best exposed along the northern bank of Zagaya river, pale grey, beige and mauve mud-, silt- and fine sandstones occur. Based on lithostratigraphic correlations, the red-bed unit is Eocene, part of the Niubao Formation; the mostly fine-grained section along the Zagaya river is Neogene (Smith & Yu 1988), although chronostratigraphic evidence for this age assignments is lacking. Structures visible on satellite images are major strike– slip faults traceable for up to 200 km north of Siling Co (Fig. 1b). In the study area, two welldefined, NW-dipping normal faults have lake-filled basins in the hanging wall and blocks of Eocene and older rocks in the footwall; toward north, they abut against a likely sinistral, east-trending fault (Fig. 13a). Structures comprise thrusts and folds with variably oriented, moderately to steeply westplunging axes (Fig. 13b) that are overprinted by the normal and strike –slip faults. Thrusting emplaced the Eocene red-bed unit onto the ophiolitic me´lange. We estimated c. 60% shortening from section restoration (line-length balancing) and considering outcrop-scale second-order folding and faulting. Local fault– slip data record grossly north–south shortening in the fold-thrust belt; the later c. east–west extension requires a sinistral strike– slip component along the graben bounding faults. The supplementary material details the lithology and deformation observations of the analysed sites. Table 1 gives the parameters of the deviatoric stress tensor calculated from fault –slip data sets.

Palaeomagnetic analysis We used a small-scale palaeomagnetic study to address possible rotational deformation in the nearfield of the strike– slip faults NE of Siling Co. Rock magnetic properties. Key elements of the rock magnetic and remanence investigations are displayed in Appendix 1; here we present a summary. We sampled 15 sites within Eocene and Neogene strata with 150 core specimens; Table 2 presents the results. The anisotropy of magnetic susceptibility (AMS) was determined by a KLY-2 kappabridge (AGICO). The AMS ellipsoids show oblate and prolate shapes with a degree of anisotropy (P 0 , 1.03) that is typical for weakly deformed sediments. Prolate sites mainly occur in thrust imbricates, thus indicate higher strains than the oblate sites, which likely trace undisturbed sedimentary

145

fabrics. Isothermal remnant magnetization (IRM) was imparted by a MMPM9 pulse magnetizer. IRM acquisition and thermal demagnetization curves of the specimens can be divided into two groups: Magnetite dominates all Neogene sites as shown by relatively low saturation fields. Small amounts of a harder magnetic carrier, likely hematite, occur, because only 80% of saturation is reached at c. 300 mT, while complete saturation needed 1.5 T. Hematite dominates all Eocene sites, as complete IRM saturation is not reached at maximum fields of 2.5 T. This identification of magnetic minerals is supported by the respective Curie temperatures (TC) found during thermomagnetic measurements of susceptibility (k – T curves): Neogene samples exhibit heating curves with a major decrease of susceptibility at c. 580 8C, thus identifying magnetite; the characteristic decrease at TC(hematite) suggests that only hematite exists in the Eocene samples. Remanence measurements and analyses. In 12 of 15 sites, natural remnant magnetization (NRM) analysis revealed proper demagnetization behaviour (Table 2). For remanence measurements, a 2G Enterprises SQUID magnetometer 2G760R (noise level ,0.01 mA m21 for 10 ccm samples) was used. Alternating field (AF) demagnetization was performed by a 2 G degausser attached to the magnetometer. A shielded MMTDM furnace was used for thermal demagnetization. A characteristic remnant magnetization (ChRM) in Neogene rocks is demagnetized by AF treatment between 10– 60 mT; it is carried by magnetite or magnetically softer hematite, or a mixture of both components, in agreement with the IRM and k –T investigations. For most samples, a residual NRM component is remaining after AF demagnetization and only in some cases the NRM could be removed completely. The residual magnetization is carried by a component with higher coercivity; according to the IRM experiments, it is hematite. The samples with a ChRM carried mainly by magnetite, hematite, or a mixture of both exhibit a stable behaviour in the Zijderveld plots; the demagnetization path represents a straight line and, with some exceptions, points to the origin. The Eocene red beds were demagnetized thermally: in most sites a viscous component, accounting for up to 20% of the NRM intensity and with randomly distributed directions, was found below c. 300 8C. After its removal, a stable ChRM carried by hematite could be separated. In 5 sites the ChRM consists of a hematite component, which unblocks between 350–680 8C; again, this is in agreement with the IRM and k – T data. The demagnetization path in the Zijderveld plots is directed toward the origin. Another 5 sites show two ChRM components: ChRM1 is

146

Table 2. Magnetic results of sites from the Zagaya section, Bangong – Nujiang suture Sitename

Geographic coordinates Latitude (N)

Lithology

N/n

RC

Component

Longitude (E)

0

0

32804.614 32804.8050 32804.8400 32804.8340 32804.8240 32804.2150 32812.4390

89840.109 89838.4040 89838.2460 89838.1780 89838.1020 89836.9200 89835.5580

Eoc. Neog. Neog. Neog. Neog. Neog. Eoc.

Red sandst. Brown silt-/sandst. Green/red siltst. Grey/green siltst. Green sandst. Green/red silt-/sandst. Red sandst.

32812.4690

89835.5730

Eoc.

Red sandst.

10/9

32812.4700 32812.0360 32811.4500

89835.5700 89835.0190 89835.2050

Eoc. Eoc. Eoc.

Red sandst. Red sandst. Red sandst.

10/9 10/ – 10/9

32810.3460

89834.1630

Eoc.

Grey/red sand-/siltst.

10/9

0

32809.587

0

89833.158

Eoc.

Red/brown sandst.

10/9

32810.8380 32810.7510

89834.9250 89834.8610

Eoc. Eoc.

Grey/red sand-/siltst. Grey/red sand-/siltst.

10/9 10/9

10/9 10/10 10/8 10/5 10/ – 10/ – 10/9

H M M M (M) (M) H H H H H (H) H H H H H H H H

300 – 650 8C 0 – 25 mT 0 – 120 mT 0 – 50 mT – – 300 – 600 8C 640 – 690 8C 350 – 580 8C 650 – 690 8C 325 – 690 8C – 325 – 600 8C 650 – 690 8C 350 – 580 8C 630 – 690 8C 400 – 620 8C 630 – 690 8C 350 – 680 8C 325 – 670 8C

Bedding

Bedding corrected directions

dip direction/dip

decl.

incl

k

a95

105/52 106/21 090/18 094/24 (113/19) (196/57) 184/76

056.4 023.4 021.4 022.8 – – 325.2 316.0 324.0 347.8 334.1 – 332.8 347.5 010.8 018.8 018.9 000.4 347.5 351.8

42.0 32.6 31.8 28.9 – – 44.8 47.6 20.3 21.7 16.1 – 35.3 29.6 39.8 39.3 28.1 32.1 35.0 46.2

24.1 44.6 27.0 37.0 – – 11.9 3.2 14.3 27.1 60.1 – 11.9 28.2 62.2 30.7 14.4 37.3 29.8 9.5

10.7 7.3 10.9 15.4 – – 19.5 34.3 14.1 10.1 7.2 – 19.5 9.9 7.1 10.1 13.8 8.5 9.6 17.5

166/42 003/42 (184/68) 342/49 220/85 ot 104/45 015/82 198/83

Abbreviations: RC, remanence carrier; H, hematite; M, magnetite; N, number of measured specimens per site; n, number of specimens included into statistics; decl., declination; incl., inclination; k, precision parameter; a95, 95% confidence limit; ot, overturned bedding; Eoc., Eocene; Neog., Neogene; 2 comp., two magnetic components.

L. RATSCHBACHER ET AL.

5S1A 5S2B 7S21K 7S22L 7S23M 7S24N 8S25K (2 comp.) 8S26L (2 comp.) 9S27K 9S28L 9S29M (2 comp.) 10S30K (2 comp.) 10S31L (2 comp.) 11S32K 11S33L

Age

RIFTING AND STRIKE –SLIP SHEAR IN TIBET

demagnetized between 350 and 600 8C, ChRM2 between 630– 680 8C. The demagnetization path in the Zijderveld plots is separated between the two straight segments of ChRM1 and ChRM2, the latter pointing toward the origin. Because IRM and thermomagnetic investigations imply that the remanence of both components is carried by hematite, ChRM1 and ChRM2 have to be explained by hematite of different grain size. For all samples included in the statistical evaluation, we achieved single specimen remanence directions with well-grouped site means and k-values between 9 and 62 (Table 2). As only normal polarity is observed and geomagnetic field reversals are frequent in the Tertiary, the remanences likely were acquired over a short time period. Fold tests indicate remanence acquisition prior to folding with best grouping of directions occurring at 93% of unfolding (fold test after McFadden 1990). Inclinations from the Zagaya sites show 10 –308 lower values (Table 2) than expected for stable Eurasia (Besse & Courtillot 2002). In this way our data match studies from Tertiary red beds in central Asia that have revealed this ‘low inclination anomaly’ (e.g. Gilder et al. 1996; Cogne´ et al. 1999). The discrepancy between predicted and

147

observed palaeolatitudes is mainly explained by non-rigid behaviour of the Eurasian plate in the Tertiary (Cogne´ et al. 1999). Latitudinal differences between sites amount to 158 in the Zagaya section and could be explained by primary inclination flattening during sedimentation and diagenesis or long wavelength folding prior to remanence acquisition. Some indication for the latter comes from the fold test that indicates that 7% of folding occurred before remanence acquisition. To investigate the spatial distribution of rotations, the calculated declinations of the Eocene Zagaya sites were projected onto a north–south line, perpendicular to the northern strike –slip fault (Fig. 14a). The amount of counterclockwise rotation with respect to the declination of the southernmost site 31L increases linearly northward (Fig. 14a). The maximum relative rotation amounts to 558 between sites 31L and 26L; the gradient of rotation per distance is 8.18/km. Site 31L was taken as reference, as it represents the farthest distance to the belt of sinistral fault zones to the north and thus is likely least influenced by near-field deformation along them. In addition, the 31L-site declination is close to the predicted post-Cretaceous declination for stable Eurasia (Besse & Courtillot 2002); it deviates on average 48 counterclockwise form the Neogene

Fig. 14. (a) Spatial distribution of site mean declination directions projected onto north– south line that ends at the strike–slip fault at the northern termination of the Zagaya section (inset). The amount of rotation increases linearly toward north. (b) Schematic comparison of two possible cases for the chronological order of geological events: top, folding predates block rotation; bottom, folding postdates block rotations. The observed linear trend suggests folding pre-dated block rotation.

148

L. RATSCHBACHER ET AL.

sites further SE (excluding site 1A that deviates considerably). Three geological events likely affected the study area: (1) sedimentation and remanence acquisition; (2) folding and faulting with rotation of remanence vectors; and (3) block rotation around vertical axes. Two cases for the chronological order of events are compared in Figure 14b: Plot A presents the measured data; in plot B, the declination distribution was broadened graphically, providing an illustration for the case that folding would have post-dated block rotation. In these plots, the ordinate carries only ‘recent’ information about localities, whereas the values on the abscissa fluctuate with different geological histories and thus carry information about the chronological order of ‘ancient’ events. Intense and heterogeneous folding, as observed along the Zagaya section, post-dating block rotation, would have broadened the declination range (plot B) and prohibited a well-defined linear trend; again, we infer that folding pre-dated block rotation. Tectonic models. The models of Figure 15a, modified after Nelson & Jones (1987), show potential counterclockwise declination and block rotations within a sinistral fault zone. The models illustrate domino-style block rotation with antithetic shear (A); continuous simple shear (B); small blocks with variable internal rotation (C). Models A with

only two distinct declinations and B with the northernmost declinations approaching the strike–slip fault asymptotically do not trace the observed declination change. Model C correctly predicts the observed pattern but the small block model with variable internal rotations on first sight badly matches the map-scale geological pattern (Fig. 13). Our meso-scale field data lend some support to the interpretation that rotations along the Zagaya section may be explained by a small-scale rheological continuum in the brittle state: fold axes show variable plunge and both fold-axes and bedding trends vary by .908; shortening directions derived from fault–slip analysis also differ considerably; even in the Neogene rocks southeast of the Eocene section, shortening directions locally vary up to 458. Only the extension directions calculated from late normal faults and tension gashes are constant, suggesting that these are late features. Our favoured solution to explain the post-folding deformation in the Zagaya section is a two-stage evolution: distributed deformation in the near-field of a sinistral shear zone is followed by normal faulting that breaks up the continuum in blocks (Fig. 15b). Thus, we suggest that the Bangong– Nujiang suture zone NE of Siling Co was reactivated postcollisional by both localized and distributed sinistral strike –slip deformation. From the Zagaya section alone, the distributed sinistral offset reflected by

Fig. 15. (a) Models explaining sinistral shearing of an initial square modified from Nelson & Jones (1987): A, domino-style block rotation with antithetic shear; B, continuous simple shear; C, small blocks with variable internal rotation. Comparisons of modelled declination pattern with the one observed along the Zagaya section excludes models A and B. (b) Sketch of the two post-folding deformation phases along the Zagaya section suggested in this study; stage 1, internal deformation in the near-field of a major strike– slip zone; stage 2: normal faulting breaks up the continuum into blocks.

RIFTING AND STRIKE –SLIP SHEAR IN TIBET

the declination trend amounts to c. 10 km. Our analysis supports the inference of significant sinistral wrenching along the Bangong–Nujiang suture zone (Taylor et al. 2003).

Age of neotectonic deformation in central and southern Tibet In the northern Shuang Hu graben, Blisniuk et al. (2001) obtained minimum ages for the onset of extension in the Muga Purou rift. They observed at four locations along the inactive range-front fault 2–8 m thick, calcite-dominated fault-zone mineralization, separating vein-cemented fault breccias of Early Mesozoic rocks of the graben shoulder in the west from Miocene –Pliocene conglomerates to the east; the mineralized unit is laterally continuous over tens to hundreds of metres. The mineralization along the fault demonstrates that it post-dates the onset of slip along the range front. It consists of coarse-grained calcite with rare accessory minerals, including muscovite. A Rb –Sr isochron, comprising muscovite, quartz with muscovite, fluid, and carbonate inclusions, calcite, and whole rock, yielded an age of 13.5 + 0.2 Ma; incremental laser heating 40Ar/39Ar analyses of quartz grains rich in muscovite suggest that mineralization occurred at 13.5 + 1.4 Ma, but a complex crystallization history starting at 14.4 + 0.4 Ma is possible. Coarse-grained muscovite from another sample yielded a weighted mean 40Ar/39Ar age of 4.0 + 0.2 Ma. These data are most simply interpreted as range-front fault activity at or before c. 13.5 Ma and reactivation at c. 4 Ma. Assuming that the range-front fault in the northern Shuang Hu graben formed at the same time as its southern pendant, its estimated 7 km offset (Yin et al. 1999) transfers to slip rates of c. 0.5 or c. 1.75 mm a21 for slip initiation at c. 14 or 4 Ma, respectively; horizontal extension rate would be c. 0.35 or 1.24 a21, taking the average 458 dip of the range-front fault. Slip rates were higher given the observed 20–508 obliquity of striae on the fault; this translates to 0.53 to 2.72 a21 for the minimum (7.45 km offset in 14 Ma) and maximum (10.9 km offset in 4 Ma) values. Extrapolating an active slip-rate estimate based fault-scarp degradation, Yin et al. (1999) inferred an up to 4 Ma age for the onset of the active Shuang Hu range-front fault south of the site studied by Blisniuk et al. (2001); the segments of this active, southern fault form an en-echelon array, also suggesting a sinistral oblique –slip component. East of the inactive range-front fault dated by Blisniuk et al. (2001), Blisniuk & Sharp (2003) dated pedogenic carbonate pebble-rinds, collected from two well-preserved terrace surfaces that are cut by an active fault; they employed U-series

149

dating. The terraces are vertically offset by c. 1.3 and c. 14.8 m. Minimum ages of c. 16.4 and c. 233 ka for deposition of the two terraces indicate average vertical displacement rates of c. 0.08 mm a21 during the last c. 16.4 ka and c. 0.064 mm a21 during the last c. 233 ka. Blisniuk & Sharp (2003) inferred 0.3 mm a21 as a maximum estimate for the rate of cumulative vertical offset by all normal faults of the Shuang Hu graben during the late Quaternary. To compare the radiometric ages obtained in the Shuang Hu graben with those for east –west extension elsewhere in Tibet, we compiled the available data from southern and central Tibet (Fig. 16; Table 3) and added so far unpublished dating from our previous structural work in Tibet (Fig. 17; Tables 4 –7). We grouped the data into three categories: (1) dykes and mineral filled tension gashes; (2) ductile strike–slip and low-angle normal faults; and (3) brittle high-angle normal faults. The regional distribution of ages is depicted in Figure 18a. Roughly north-trending dykes have been dated between 8.7 and 19.3 Ma (Yin et al. 1994; Williams et al. 2001; Figs 16 & 18; Table 3); these comprise the dykes crossing the Xigaze forearc rocks of the Yarlung –Tsangpo suture and the Gangdese arc on the Lhasa block. We dated one north-trending granite-porphyry dyke near Gyanze within the northern Tethyan self sediments (U –Pb on xenotime and monazite; 10.8 + 0.3 Ma; Fig. 17a; Table 4), and Kapp et al. (2005) obtained the so far youngest age of a dyke from a NE-trending leucogranite intruding the basement of the Nyainqentaghla range (8.7 + 1.0 Ma). The occurrence of these dykes has been attributed to the onset of east –west crustal extension (e.g. Williams et al. 2001). Here, we agree with Mahe´o et al. (2007) that the c. northstrike of the dykes only indicates that the minimal horizontal stress axis (s3) was c. east –west at the time of intrusion; this does not imply crustal extension. The dykes may trace a wrenching event with subhorizontal s1. A similar interpretation holds for the c. 14 Ma hydrothermal micas from an extension vein, sampled close to the Thakkhola graben (Coleman & Hodges 1995). We obtained a maximum age for the onset of east –west extension along the northern Yadong graben (Fig. 18) from the 11.5 + 0.4 Ma age of the Kari La granite east of the graben, which is cut by normal faults (U– Pb on xenotime and monazite; Fig. 17a; Table 4). There is a group of ages dating ductile shear zones (Figs 16 & 18; Table 3). Nearly purely normal detachments were dated at c. 11 Ma (Dinggye, Zhang & Guo 2007; Kali et al. 2010) and at c. 8 Ma (Nyainqentaghla, Harrison et al. 1995; Gurla Mandata, Murphy et al. 2002); Kapp et al.s’ (2008) youngest mylonitic leucogranite

150

L. RATSCHBACHER ET AL.

Fig. 16. Compilation of data for the age of onset of strike– slip and normal faulting in central and southern Tibet. The data are grouped into the categories (a) dykes and mineral-filled tension gashes; (b) ductile strike–slip and low-angle normal faults; (c) brittle high-angle normal faults. The data base is given and referenced in Table 3. See text for discussion.

(U– Pb zircon, 8.9 + 0.2 Ma) provides a similar maximum age for the onset of east –west extension along the ductile Lunggar rift detachment (Fig. 16). Dextral shear along the Karakorum–Jiali fault was dated at c. 15 and c. 8 Ma (Table 3). Edwards & Ratschbacher (2005) provided a detailed kinematic and rheological analysis of the sinistral Damxung shear zone that connects the Nyainqentaghla detachment with the (brittle) Gulu rift (Fig. 18a); deformation temperature reached at most 350 8C. Here, we report a series of K –Ar ages from synkinematic, fine-grained white mica

(sericite) that allows precise dating of lowtemperature mylonitization (Table 5): seven regional dispersed samples yielded 8.7 + 0.6 Ma, overlapping the age of onset of deformation along the Nyainqentaghla detachment. Onset of brittle faulting along the Tibetan rifts is overwhelmingly dated at c. 5 Ma (Figs 16 & 18a; Table 3). We again agree with Mahe´o et al. (2007) that the oldest sediments (11–9.6 Ma) in the Thakkhola graben may not be related to rifting but rather to piggy-back deposition or motion along the South Tibetan detachment. Blisniuk et al.s’ (2001) c. 14 Ma hydrothermal mica may in fact date a breakaway fault above a still unexposed ductile detachment that is connected to a strike–slip fault; strike –slip solutions are characteristic for prePliocene deformation in central Tibet (see above). The c. 13 Ma event in the Tangra Yum Co rift has structurally not yet been detailed (Dewane et al. 2006). Here, we report a c. 5 Ma onset of brittle high-angle normal faulting along the Ringbung graben, a segment of the northern Yadong rift (Fig. 18a; Ratschbacher et al. 1994 for structural details). The Ringbung leucogranite (7.2 + 0.2 Ma U –Pb zircon, xenotime, monazite age, Fig. 17a; Table 4) occupies the western flank of a c. NW-dipping graben and is cut by normal faults. The granite cooled rapidly to c. 6.5 Ma (Rb –Sr and Ar/Ar white mica; Fig. 17c, d; Tables 6 & 7). After a period of slow cooling, rapid cooling resumed at c. 5 Ma (Rb– Sr biotite and Ar/Ar Kfeldspar low-temperature plateau and multi-domain modelling; Fig. 17b–d; Tables 6 & 7). Harrison et al.’s (1995) data from the Nyainqentaghla detachment (Damxung) show a similar c. 5 Ma resumption of rapid cooling (Fig. 17e); for both, the Ringbung and Damxung grabens, we interpret this rapid cooling event as indicating onset of brittle highangle normal faulting.

Discussion and conclusions Geometry of neotectonic deformation in central and southern Tibet Structural mapping along and north of the Muga Purou rift revealed three fault trends: ENE, NNE, and NW (Fig. 19). The ENE-striking faults (set 1 in Fig. 19) are dominated by sinistral strike –slip motion, confirmed by both focal mechanisms and field data (Figs 4 & 11). The NNE-striking faults (set 2) have mostly normal kinematics and outline a right-stepping en-echelon array of grabens, which also suggest sinistral strike– slip. Along the NW-striking fault sets (set 3), the morphology along individual faults suggests mainly normal kinematics but the arrangement of grabens may

RIFTING AND STRIKE –SLIP SHEAR IN TIBET

151

Table 3. Age of onset of strike–slip and normal faulting in central and southern Tibet Age (Ma)

Error (Ma)

Reference

8

1

Harrison et al. (1995)

8.7 15

0.6 3

This study Lee et al. (2003)

11 11 14

1 2 1

Kali et al. (2010) Zhang & Guo (2007) Murphy & Copeland (2005)

14.5

1.5

Phillips et al. (2004)

9 14 15 8 8 4 4 4.5

2 1.5 2 2 2 0.5 2 1

Murphy et al. (2002) Valli et al. (2007) Dunlap et al. (1998) Dunlap et al. (1998) Kapp et al. (2008) Maheo et al. (2007) Yin et al. (1999) Harrison et al. (1995)

5 6 13 6 14 4 4 3

1 1 1 1 1 1 1 1

This paper Stockli et al. (2002) Dewane et al. (2006) Dewane et al. (2006) Blisniuk et al. (2001) Blisniuk et al. (2001) Jessup et al. (2008) Valli et al. (2007)

2.5 10

1 1

Armijo et al. (1986) Garzione et al. (2003)

5

2

Garzione et al. (2003)

18.3 13.8 13.8 13.3 17.3 19.3 15.4 14.2 10.8 8.7

2.7 3.5 0.3 0.8 1.9 1.5 1.5 1 0.3 1

Williams et al. (2001) Williams et al. (2001) Williams et al. (2001) Williams et al. (2001) Williams et al. (2001) Williams et al. (2001) Yin et al. (1994) Coleman & Hodges (1995) This study Kapp et al. (2005)

indicate a dextral strike–slip component (Fig. 7); this is supported by the regional WNW- and NNEtrends of the T- and P-axes (Fig. 12), respectively, which require a dextral strike– slip component along this set. The dextral strike –slip component is, however, substantiated neither by local geomorphological nor focal mechanism data. Clear dextral strike– slip kinematics only appears south of the Bangong–Nujiang suture along the various segments of the Karakorum–Jiali fault zone (Fig. 1b; southernmost fault in Fig. 19). North of

Shear/fault zone and type of motion Nyainqentanghla normal detachment Damxung sinistral shear zone Parlung, Puqu (Jiali) dextral shear zone Dinggye normal shear zone Dinggye normal shear zone Gurla Mandate-Humla transtensional fault system Central Karakoram fault, transpressional Gurla Mandata normal detachment Ayilary range, transtensional Pangong range, dextral transpression Pangong range, dextral transpression Lunggar rift, normal detachment Kung Co rift Shuang Hu rift Nyainqentanghla, range front, normal Ringbung rift Gulu rift, normal Tangra Yum Co rift, normal Tangra Yum Co rift, normal Shuang Hu rift, normal Shuang Hu rift, normal Dinggye fault, normal Gar-Baer basin, Karakorum fault, transtension S-Tibetan rifts, normal Oldest sediments in the Thakkhola graben Rift related sediments, Thakkhola graben Southern Tibet Southern Tibet Southern Tibet Southern Tibet Southern Tibet Southern Tibet Southern Tibet Dinggye Gyanze Goring La

Rheology Ductile Ductile Ductile Ductile Ductile Ductile Ductile Ductile Ductile Ductile Ductile Ductile Brittle Brittle Brittle – ductile Brittle Brittle Brittle Brittle Brittle Brittle Brittle Brittle Brittle

Dyke Dyke Dyke Dyke Dyke Dyke Dyke Tension vein Dyke Dyke

the Bangong–Nujiang suture zone, the rifts seem to branch off the northwestern end of the strike – slip faults of the Karakorum –Jiali fault zone and have a significant sinistral strike –slip component (Taylor et al. 2003 and this study). The trace of the Yibug Caka rift is S-shaped with two prominent sinistral strike–slip and graben segments. ENEtrending strike–slip segments that connect grabens are better developed and longer along the Yibug Caka than the Muga Purou rift; deformation appears less distributed in the former.

152 L. RATSCHBACHER ET AL. Fig. 17. New geochronology constraining the age of onset of strike– slip and normal faulting in southern Tibet. (a) U–Pb zircon, monazite, and xenotime geochronology of three leucogranites intruding the Tethyan series of the northern Himalayan foreland; see Table 4 for locations and analytical details. (b) The 40Ar– 39Ar results from K-feldspar Tr210 (Ringbung leucogranite): age spectrum, Arrhenius plot, log (r/r0) plot, and thermal history. Grey bar shows initiation of rapid cooling at c. 5 Ma. (c) Rb–Sr white mica and biotite isochrons and 40Ar– 39Ar white mica and biotite age spectra of sample Tr210 (Ringbung leucogranite). (d) Cooling curves for the Ringbung and Kari La leucogranites; 40 Ar – 39Ar data for the Kari La granite are from Copeland (1990). (e) Cooling paths (from Harrison et al. 1995) from the Nyainqentanghla detachment showing a second phase of rapid cooling at c. 5 Ma. Data processing and plotting used ISOPLOT (Ludwig 2003).

Fig. 18. (a) Neotectonic map of Tibet displaying the following data: (i) Ages of onset of neotectonic normal and strike– slip faulting/shearing; ductile refers to dominant temperature-activated flow. For references to these data see Table 3. Ductile shear zones are mostly strike–slip or low-angle detachments connected to strike–slip-shear zones (e.g. oblique–normal Nyainqentanghla (Damxung) detachment connected to the sinistral Damxung shear zone). Brittle refers to co-seismic fracture and cataclastic flow. Basin and vein mark age estimates from sedimentary record and tension-gashes filled with hydrothermal minerals. (ii) ‘Beach balls’ illustrate the three-dimensional stress/strain state at each node using lower hemisphere stereonet projections similar to that for earthquake focal mechanisms; the shaded quadrant is compressional/contractional and contains s1. The data summarize 1 to 15 local sites, surveyed by or calculated from Armijo et al. (1986, 1989), Mercier et al. (1987), Ratschbacher et al. (1992, 1994), Yin et al. (1999), Murphy et al. (2002), Edwards & Ratschbacher (2005), Taylor et al. (2003), Kapp et al. (2005, 2008), and this study. The Pung Co detachment and rift data are given in Fig. 24, Table 1, and the supplementary material. The yellow line separates the normal fault dominated region in southern Tibet from the normal-and-strike–slip dominated region in central Tibet; the boundary cuts NW-trending obliquely across the Lhasa block. The pink line delimits the normal-and-strike–slip dominated region in central Tibet from the strike– slip-and-reverse fault province in northern Tibet. (b) Predicted deviatoric stresses in the upper crust (at 10 km depth) of the Tibetan Plateau from Liu & Yang’s (2003) numerical modelling; it shows that a combination of indentation of the Indian plate and gravitational spreading of the elevated Plateau can produce the present field. Coloured lines give depth contours of the Indian plate from P-wave tomographic imaging (Li et al. 2008). The distance over which Indian lithosphere has thrust under Tibet decreases from west to east and via stress coupling (Liu & Yang 2003) can explain the surface distribution of rift structures in Tibet (compare with the yellow line in (a)). (c) Particle flow lines (magenta bands) across Tibet. A particle (e.g. red arrows) flowing out of central Tibet accelerates across central Tibet, reaches its highest velocity at the neck between the syntaxes (India, South China), and decelerates south of it. Grey arrows mark the east component of the velocity vectors at c. 928E from Zhang et al. (2004) with reference to Eurasia and the blue arrows the velocity field of the Tibetan Plateau in a Tibetan-fixed reference frame, which best depicts the Plateau-interior deformation (from Gan et al. 2007). Digital elevation model as background of (b) and (c) is from http://www.ngdc.noaa.gov/mgg/topo/globe.html.

Table 4. U–Pb isotopic data for zircon, xenotime, and monazite for leucogranites, Tibetan shelf north of the Himalaya Total1 Total2 Com.2 Sample/fraction (number of grains)

Weight (mg)1

U (ng)

206

Pb2/ Pb

206

Pb3/ Pb

206

Pb3/ U

% Error4

207

Pb3/ U

Ages (Ma) 207

% Error4

Pb3/ Pb

% Error4

206

Pb/ U

207

Pb/ U

207

Pb/ Pb

Pb (pg)

Pb (pg)

371.5 674.3 38.3 21.6 306.0

30.0 40.0 10.5 4.8 1.9

93 120 286 252 10264

0.114 0.108 3.162 3.510 27.733

0.0019531 0.0019313 0.0018138 0.0017899 0.0160803

0.354 0.349 1.570 0.801 0.116

0.0112396 0.0113399 0.0115708 0.0114191 0.1245000

1.390 1.310 1.800 1.020 0.145

0.041738 0.042585 0.046268 0.046271 0.055950

1.290 12.58 11.35 2242.4 1.210 12.44 11.45 2191.8 0.837 11.68 11.68 11.62 0.605 11.53 11.53 11.74 0.085 102.8 118.7 450.4

187.4 168.3 212.3 646.2 235.6 50.4 47.2 78.9 24.9

14.0 15.0 16.0 49.0 4.4 2.8 0.8 2.2 1.9

111 134 150 148 3428 1127 3581 4224 804

0.118 0.187 0.184 0.186 9.764 8.004 13.989 6.211 5.626

0.0018706 0.0018244 0.0179920 0.0017996 0.0016886 0.0016918 0.0365627 0.0077622 0.0064170

0.670 0.502 0.460 0.681 0.167 0.436 0.251 0.235 0.631

0.0105255 0.0106085 0.0107243 0.0108245 0.0107880 0.0107820 0.2858000 0.0605055 0.0482482

1.500 1.720 0.987 1.250 0.216 0.565 0.415 0.307 0.683

0.040810 0.042172 0.043230 0.043626 0.046338 0.046223 0.056692 0.056534 0.054532

1.280 12.0 1.570 11.8 0.836 11.6 1.010 11.6 0.131 10.9 0.342 10.9 0.318 231.5 0.187 49.8 0.243 41.2

522.6 36.4 40.6 29.2 11.3 45.1 30.9 37.8

49.0 3.2 2.6 1.5 1.7 5.2 2.5 4.5

220 1122 1280 1973 596 794 1350 719

0.382 13.340 15.604 16.810 8.757 10.135 15.831 6.874

0.0012994 0.0011385 0.0011255 0.0011395 0.0011290 0.0011113 0.0011071 0.0011034

0.275 0.471 0.233 0.281 0.668 0.513 0.419 0.509

0.0072988 0.0072512 0.0071685 0.0073261 0.0072636 0.0070931 0.0070520 0.0070279

0.591 0.612 0.379 0.371 0.785 0.722 0.506 0.586

0.040739 0.046194 0.046194 0.046629 0.046663 0.046205 0.046197 0.046194

0.505 0.369 0.285 0.230 0.383 0.479 0.265 0.270

204

208

238

235

206

238

8.37 7.33 7.25 7.34 7.27 7.17 7.13 7.11

235

10.6 10.7 10.8 10.9 10.9 10.9 255.3 59.6 47.8

206

2299.4 2216.2 2154.3 2131.8 15.2 9.3 479.6 473 393

7.38 2304 7.34 7.74 7.25 7.75 7.41 30.27 7.35 32.02 7.18 8.34 7.14 7.89 7.11 7.73

r5

0.405 0.416 0.799 0.808 0.808 0.523 0.431 0.537 0.594 0.793 0.797 0.643 0.793 0.935 0.522 0.799 0.661 0.785 0.874 0.749 0.852 0.887

RIFTING AND STRIKE –SLIP SHEAR IN TIBET

Kari-La granite 6 (28853.9890 N, 090814.1080 E) (A) Mnz, fragment (1) 0.010 17.60 (B) Mnz, rhomb. (2) 0.01 32.57 (C) Xno, green tetradhed. (1) 0 16.18 (D) Xno, tiny green tetrahed. (3) 0 9.53 (E) Zrc, clear needles (2) 0 20.25 Gyangze aplite dyke 6 (28850.760 N, 89838.980 E) (F) Mnz, large rhomb. (1) 0.01 10.05 (G) Mnz, small rhomb. (5) 0.010 14.16 (H) Mnz, small rhomb. (4) 0.008 18.23 (I) Mnz, small rhomb. (12) 0.03 56.17 (J) Xno, large green tetrahed. (1) 0.01 140.0 (K) Xno, small green tetrahed. (1) 0 29.07 (L) Zrc, clear needle (1) 0 1.33 (M) Zrc, clear needles (3) 0.01 9.63 (N) Zrc, clear needle (1) 0 3.59 Rengbung granite 7 (29806.730 N, 90803.340 E) (O) Mnz, green ovoid (1) 0 113.9 (P) Xno, clear tetrahed. (1) 0 32.22 (Q) Zrc, clear needles (5) 0.01 36.80 (R) Zrc, clear needles (6) 0.01 26.50 (S) Zrc, clear needles (1) 0 9.48 (T) Xno, small clear tetrahed. (2) 0 39.48 (U) Xno, clear tetrahed. (1) 0 28.58 (V) Xno, clear tetrahed. (1) 0 31.94

Atomic ratios

1

Weight estimated from measured grain dimensions and assuming reasonable densities, c. 20% uncertainty affects only U and Pb concentrations. Corrected for fractionation (0.1 + 0.05%/amu – Faraday; 0.18 + 0.07%/amu – Daly) and spike. Corrected for fractionation, blank, and initial common Pb. 4 Errors quoted at 2s. 5207 Pb/235U – 206Pb/238U correlation coefficient of Ludwig (1989). 6 Corrected for common Pb with assumed composition: 204Pb:206Pb:207Pb:208Pb ¼ 1:18.67:15.76:39.32. 7 K-Feldspar common Pb correction: 204Pb:206Pb:207Pb:208Pb ¼ 1:19.10:15.76:39.60. 2 3

153

154

Table 5. K –Ar age data of mylonitic rocks from the Damxung shear zone, northern Yadong – Gulu rift

6R2a 6R2b 6E1b 7R2a 7R3a 1E-4/8-II 1E-4/8-III

Lithology

Latitude

Longitude

Mineral

K, %+1s

Phyllite Quartzitic phyllite Phyllitic rhyolite Phyllitic phenocrystic rhyolite Phyllitic rhyolite Phyllite Phyllite

30836.5810 30836.5810 30835.6510 30835.2820 30835.4690 308360 308360

91811.1320 91811.1320 91811.4350 91806.9230 91806.8520 918350 918350

sericite sericite sericite sericite sericite sericite sericite

3.29 + 0.04 4.42 + 0.04 5.45 + 0.05 6.64 + 0.06 4.09 + 0.04 4.40 + 0.04 4.36 + 0.04

40

Ar, ng/g+1s 2.00 + 0.15 2.69 + 0.15 3.45 + 0.15 3.90 + 0.20 2.20 + 0.15 2.75 + 0.75 2.68 + 0.15

Age (Ma)+2s 8.8 + 0.6 8.7 + 0.6 9.1 + 0.6 8.5 + 0.6 7.8 + 0.7 9.0 + 0.6 8.9 + 0.6

Note: Details about lithology, deformation, and deformation temperature are given in Edwards & Ratschbacher (2005). Lithology describes pre-deformation protolith. Content of radiogenic 40Ar was measured by isotope dilution with 38Ar as an isotope tracer. Potassium was determined by flame photometry; its concentration was determined twice in each sample. Analyses by Y.D. Pushkarev, Russian Academy of Sciences, St. Petersburg.

L. RATSCHBACHER ET AL.

Site

Sample

Mineral

TR210 TR210 TR210

Wm Bt Kfs

40

Ar/39Ar data from the Ringbung leucogranite; location 29806.73 0 N, 090803.34 0 E J

Weight (Mg)

Grain size (mm)

TFA (Ma)

IA (Ma)

MSWD

40

Ar/36Ar

0.0042568 0.0032860 0.0013686

2.7 5.0 64.9

63 – 250 63 – 250 63 – 250

6.47 + 0.03 6.09 + 0.03 10.26 + 0.10

6.33 + 0.09 6.03 + 0.09 4.43 + 0.05

5.5/1.9 1.6/2.3 5.5/2.3

297 + 1 298 + 14 281 + 3

WMA (Ma)

Steps used

% 39Ar used

6.32 + 0.03 6.04 + 0.03 4.35 + 0.04

1 – 11/14 3 – 9/17 4 – 15/38

84 82 18

Abbreviations: J, irradiation flux parameter; TFA, total fusion age (uncertainty reflects only analytical precision); IA, isochrone age; MSWD, mean square weighted deviation, which expresses the goodness of fit of the isochron; WMA, weighted mean age; isochron and weighted mean ages are based on fraction of 39Ar listed in the last two columns. Other abbreviations: Wm, white mica; Bt, biotite; Kfs, potassium-feldspar. Source: Analyses by Lothar Ratschbacher, Stanford University.

RIFTING AND STRIKE –SLIP SHEAR IN TIBET

Table 6. Summary of

155

156

L. RATSCHBACHER ET AL.

Table 7. Summary of Rb–Sr data from the Ringbung leucogranite; location 29806.73 0 N, 090803.34 0 E Sample

Mineral

TR210

Wm Fsp (mixture) Bt Kfs

87

Rb/86Sr

291.809 6.769 805.79 16.08

+ (%)

87

1 1 1 1

Sr/86Sr

0.76457 0.73851 0.79421 0.73921

+ (%)

Age (Ma)

+ (Ma)

MSWD

0.05 0.02 0.05 0.02

6.45

0.23

0.54

4.91

0.12

0.06

Mineral abbreviations as in Table 6. Source: Analysis by Marion Tichomirowa, TU Bergakademie Freiberg.

South of the Bangong–Nujiang suture zone, neotectonic structures are contained in a narrow belt with dextral (the Karakorum–Jiali fault zone) and subordinate sinistral (e.g. Damxung, YiemaLa) fault/shear zones, and large, approximately NNE-striking normal faults (Fig. 1b; Armijo et al. 1986); these normal faults have up to kilometric offsets and are arranged along seven principal rift zones. The en-echelon arrangement of some of the rift segments indicates minor strike– slip in addition to dip–slip extension (e.g. Yadong-Gulu rift in south-eastern Tibet; Ratschbacher et al. 1994). The northern extremities of six of the rifts meet with strike–slip faults of the Karakorum–Jiali fault zone and have been interpreted to branch off the southeastern extremities of this dextral wrench zone (Armijo et al. 1986; Taylor et al. 2003). The south Tibetan normal faults do not follow ancient structures or fault zones in the crust but cut them at high angle (Armijo et al. 1986). If one takes the two well-studied rifts in central Tibet (Yibug Caka and Muga Purou) as representative, the geometry of neotectonic deformation is different in central and southern Tibet (Figs 1b & 18a). Prominent rift structures in central Tibet are shorter, less continuous, more closely spaced and with less relief and slope (Fig. 1a) compared to the southern Tibetan rifts; the relief contrasts amount to 1.5 km in central and 2 km in southern Tibet. The rifts consist of short graben segments connected via (extensional) strike– slip segments. The central Tibetan rifts have an overall NE-strike, more similar to the northern Tibetan Altyn Tagh fault system than to the southern Tibetan rifts. Normal faulting appears not only to be more diffuse and subdued in central than southern Tibet but also to be subordinate to strike –slip faulting (compare Taylor et al. 2003); this difference is most apparent in the fault-plane solutions. Focal mechanisms indicate that, in general, normal faulting is the primary mode of deformation between the Himalaya and the Karakorum –Jiali fault; to its north, strike –slip faulting becomes dominant. The seismic activity and the number of active faults decrease north of the Muga Purou rift, especially between 348 and 358N. Around Dogai

Fig. 19. Overview of central Tibet around the Muga Purou rift with generalized fault trends. The ENE-striking faults (set 1) are dominated by sinistral strike–slip motion, confirmed by both focal mechanisms and field data (cf. Figs 4 & 11). The NNE-striking faults (set 2) have mostly normal kinematics and outline a right-stepping en-echelon array of grabens, which also suggest sinistral strike–slip. Along the NW-striking fault sets (set 3), the morphology along individual faults suggests mainly normal kinematics but the arrangement of grabens may indicate a dextral strike– slip component (cf. Fig. 7); this is supported by the regional WNW- and NNE-trends of the T- and P-axes (cf. Fig. 12), respectively, which require a dextral strike– slip component along this set.

RIFTING AND STRIKE –SLIP SHEAR IN TIBET

Coring and northeast of it, there appears to be a corridor of no seismicity. High seismic activity resumes at c. 358N and is associated with the belt of major strike– slip faults in northern Tibet that appear to accommodate active deformation. Although the overwhelmingly shallow earthquakes (most of the hypocentres are located less than 10 km below the surface, Langin et al. 2003) locally outline active faults, clusters in particular seem to trace linkage or propagation zones of the know structures. The earthquake pattern together with the neotectonic mapping and local fault-slip analyses emphasize the distributed, heterogeneous pattern of deformation within a developing regional structure and indicate that strain concentration is weak in the uppermost crust of central Tibet.

Kinematics, dynamics, and age of neotectonic deformation in Tibet Both the southern and central Tibetan rifts do not cross the Bangong–Nujiang suture zone, which appears to host a belt of conjugate strike –slip fault sets that accommodate east –west extension and north–south shortening (e.g. Taylor et al. 2003; Figs 1b & 18a). In central Tibet, the rifts comprise mostly grabens connected to strike–slip fault zones or arranged en-echelon to accommodate sinistral wrenching; overall strain geometry is constrictional, in which NNE–SSW and subvertical shortening is balanced by WNW –ESE extension (Mercier et al. 1987; Taylor et al. 2003). A few localities indicate a counterclockwise rotation of the principal extension direction in time. The south Tibetan rifts are nearly pure extensional. It appears that in both, central and southern Tibet, rifting was preceded by a shortening event that accommodated overall north– south contraction principally by sinistral strike –slip; this is the ‘shortening and strike– slip related extension’ event in the Shuang Hu graben in central Tibet and in southern Tibet (cf. Ratschbacher et al. 1994); this may include the ‘ductile’ solutions, encompassed by strike –slip shear zones and subhorizontal normal detachments (see below). Figure 18a summarizes the current data base on regional fault kinematics/dynamics based on fault-slip analysis. The data are plotted as ‘beach balls’ (shortening grey), known from focal mechanisms, comprise a summary of one to 15 local sites, discriminate between the ‘shortening’ event defined above and the ‘rifting’ event, and represent the ‘long-term’ neotectonic pattern in contrast to the ‘short-term’ pattern, derived from earthquake focal mechanisms. Although the fault-slip based pattern is still based on insufficient data, the ‘rifting’ pattern compares well to the earthquake derived one (see e.g. Fig. 2 of Taylor & Yin 2009);

157

this suggests that the current kinematics has been active since at least 4 Ma (see above). The c. east-striking Damxung shear zone and the c. NE-striking Nyainqentanghla sinistral-normal detachment allow speculations about the geometry of ductile east –west stretching in the southern Tibet crust. The ductile, low-angle detachments may be part of a mid-crustal de´collement in which the strike–slip shear zones root; they may not be related to crustal extension with s1 subvertical. This may apply to the Gurla Mandata detachment that occupies a releasing bend between the Karakorum and Gurla Mandata– Hunla dextral shear zones (Murphy & Copeland 2005) and to the Nyainqentanghla detachment that occupies the extensional tip of the Damxung shear zone (Edwards & Ratschbacher 2005); the latter may connect to the Jiali shear zone via a still unexposed low-angle detachment. We suggest that the ‘high-temperature’ deformation (ductile flow structures and hydrothermal vein fillings) that shows c. east –west stretching is

Fig. 20. (appendix 1). Representative curves for IRM (isothermal remanent magnetization) acquisition and thermal demagnetization of sites from the Zagaya section. (a) Site 22L, contribution of magnetite to the magnetic spectra. (b), (c) sites 29M and 32 K, hematite is dominating magnetic phase. SIRM, saturation IRM.

158

L. RATSCHBACHER ET AL.

related to the regional c. north–south shortening, which predates the active crustal extension by brittle normal faulting. This about Miocene event is characterized by strike–slip faults/shear zones underlain by a regional de´collement (see above). We found similar yet undated kinematics, that is,

sinistral-oblique low-angle brittle –ductile normal faulting overprinted by high-angle normal faulting in the Pung Co rift along the Bangong –Nujiang suture (Fig. 18a; data are given in Fig. 24 (Appendix 2) and Table 1). The currently available geochronology constrains this event between c. 18 and 7 Ma;

Fig. 21. (Appendix 1). Representative demagnetization behaviour of natural remnant magnetization of sites from the Zagaya section. (a) Alternating field demagnetization of sites 21 K and 22L, demagnetization of magnetite between c. 10 and 60 mT. (b) Thermal demagnetization of sites 27 and 32 K, unblocking of one hematite component at c. 680 8C. (c) Thermal demagnetization of sites 29M and 31L, two hematite components are demagnetized between 400 and 630 8C, and above 650 8C.

RIFTING AND STRIKE –SLIP SHEAR IN TIBET

the classic high-angle normal-fault Tibetan rifting likely started at c. 6– 4 Ma. Liu & Yang’s (2003) numerical modelling of the present stress field within the Tibetan crust showed that a combination of indentation of the Indian plate and gravitational spreading of the elevated Plateau

159

can reproduce the present field. The first-order correspondence of the modelled stress field (Fig. 18b) and the observed ‘stress’ field based on fault–slip analysis (Fig. 18a) warrants further discussion of the structures in central and southern Tibet based on east-directed material flow. The geometry of

Fig. 22. (Appendix 1). Orthogonal vector (Zijderveld) plots of demagnetization behaviour of sites from the Zagaya section. H, V, horizontal, vertical projection. (a) Alternating field demagnetization of sites 21 K and 22L, magnetite is demagnetized between c. 10 and 60 mT. (b) Thermal demagnetization of sites 27 and 32 K, unblocking of one hematite component at c. 680 8C. (c) Thermal demagnetization of sites 29M and 31L, two hematite components (1, 2) are demagnetized between 400 and 630 8C, and above 650 8C. All data bedding corrected.

160

L. RATSCHBACHER ET AL.

Fig. 23. (Appendix 1). Fold tests on 8 sites from the northwestern area of the Zagaya section. Fold tests are negative at the 99% level. After bedding correction, declinations show a larger spread than inclinations. ChRM1 directions (see text).

the Indian plate and the rotational component of Tibetan deformation around the Assam –Namche Barwa syntaxis (northeastern corner of India) impose a northward increasing clockwise rotational component also at the longitude of the Yibug Caka and Muga Purou rifts. This appears to be reflected by the change of c. east-trending extension in the central southern Tibet rifts to more SE-trending extension in the central Tibetan rifts (compare the fault–slip derived stress solutions in Fig. 18a with the modelled solutions of Fig. 18b). The yellow line in Figure 18a separates the normal fault dominated region in southern Tibet from the normal-and-strike –slip dominated region in central Tibet; the boundary cuts NW-trending obliquely across the Lhasa block. The pink line (Fig. 18a) delimits the (still badly defined) northern boundary of the normal-and-strike –slip dominated region in central Tibet from the strike –slip andreverse fault province in northern Tibet, principally represented by the Altyn Tagh, Kunlun and Ganzi – Xiangshuihe fault zones. Figure 18c uses the derived kinematics (see also Taylor et al. 2003; Taylor & Yin 2009) and develops a plateau-wide

Fig. 24. Structural data from the Pung Co ductile detachment and rift. For fault–slip data legend see captions to Figure 4; for location Table 1. See Appendix 2.

kinematic model for Late Cenozoic deformation, extending the one proposed by Ratschbacher et al. (1996) for eastern Tibet and western Sichuan. There are three major backstops: India, Tarim, and South China (Sichuan basin). By analysing the fault pattern and fault–slip data in terms of kinematics one may construct particle flow lines across Tibet (Fig. 18c). Particles accelerate and move eastwards from western Tibet. Particle flow lines first diverge as the plateau is widening, and thus particle acceleration is modest; this may be reflected in the structural geometry, comprising rifts and strike–slip faults connected to grabens. The centre of the flow channel is traced by the belt of conjugate strike –slip faults that follows the Bangong–Nujiang suture. At about 928E, the flow lines start to converge toward the neck between the Assam–Namche Barwa syntaxis and the strike –slip-and-reverse fault belt of the Qaidam basin and particles accelerate; this area is characterized by the appearance of the major through-going strike –slip faults of eastern-central Tibet, the Kunlun, Ganzi –Xiangshuihe and Jiali fault zones. The flow lines turn southeastward and converge most between the Assam–Namche Barwa and Gongha (the western edge of the South China block) syntaxes. Here the particles reach their highest velocity. The flow lines diverge south of the cord between the syntaxes. A particle flowing out of central Tibet thus accelerates across central Tibet, reaches its highest velocity at the neck between the syntaxes, and decelerates south of it. In this interpretation, the velocity-rate changes in eastern Tibet and western Sichuan are reflected in the structures as follows (Ratschbacher et al. 1996, Fig. 18c): The Batang –Litang conjugate fault zone occupies the centre of the neck. The increase in extension rates just south of the Batang –Litang fault interaction, inferred from the presence of basins only south of the intersection, may reflect the area of strongest velocity-rate change. The velocity decrease south of the cord connecting the syntaxes causes north– south contraction and east –west extension in the Panxi rift and elsewhere in Yunnan. This neotectonic kinematic pattern correlates well with the decade-long velocity field derived from GPS-geodesy. Figure 18c includes the east component of the velocity vectors at c. 928E from Zhang et al. (2004) with reference to Eurasia (grey arrows) and key elements of the velocity field of the Tibetan Plateau in a Tibetanfixed reference frame, which best depicts the Plateau-interior deformation (blue arrows; from Gan et al. 2007). Three additional features of the central and southern Tibetan deformation pattern are relatively well explained by the kinematic model: (1) southern Tibet rifts have a dextral component where the flow lines diverge in the west

RIFTING AND STRIKE –SLIP SHEAR IN TIBET

and a sinistral component where they converge in the east; (2) in northern central Tibet, rift and strike– slip structures vane, as there is a zone of little differential eastward motion between central Tibet and the strike–slip-and-reverse fault domain of northern Tibet; and (3) the location of the two areas (Shuang Hu graben, Blisniuk & Sharp 2003; Yadong–Gulu rift, Armijo et al. 1986) where Quaternary east –west extension rates have been determined within the Tibetan flow field also explains their differences; slower motion should occur in the west (Shuang Hu) than east (Yadong–Gulu). What causes the difference between the structural geometries of the rifts in central and southern Tibet, that is, the broad belt of distinct, well separated rifts in southern Tibet and the complex, distributed graben-and-strike –slip structures in central Tibet? Liu & Yang’s (2003) modelling of indentation of the Indian plate and gravitational spreading of the elevated Plateau produced the present field under certain boundary conditions. One condition is basal shear associated with the subduction of the Indian plate: a moderate basal shear stress south of the Yarlung –Tsangpo suture zone leads to weaker compression within the upper crust in the Himalaya and southern Tibet. This would explain the pure normal mechanisms and the better development of the rifts in southern Tibet (Liu & Yang 2003). If basal shear associated with subduction of the Indian plate under Tibet enhances upper crustal extension, why is then the boundary between the nearly pure extensional province of the southern Tibet rifts and the strike–slip and normal faulting province of central Tibet running obliquely across the Lhasa block and to the Yarlung –Tsangpo suture (yellow line in Fig. 18a)? Using P-wave tomographic imaging, Li et al. (2008) showed that the distance over which presumed (continental) Indian lithosphere has thrust under Tibet decreases from west, where it probably underlies most of the Plateau, to east, where it extends no further than the Yarlung–Tsangpo suture zone (Fig. 18b); much of central and eastern Tibet is underlain by lithosphere of Asian origin. We suggest that this distinct spatial variation in the mantle structure along the collision zone is responsible for the surface distribution of rift structures in Tibet. We thank Bi Shiwen, Peter Blisniuk, Bill Kidd, the late Dough Nelson, Wan Jiang, and Wu Zhenhan for joint fieldwork. LR thanks Mike McWilliams for access to the Stanford Ar-laboratory and Bradley Hacker for supervision. Oscar Lovera helped with MDD modelling of K-feldspar sample Tr210. Marion Tichomirowa and Yury Pushkarev made Rb– Sr and K –Ar dating possible. We thank Alexander Webb and Michael Taylor for helpful reviews and Phil Leat for editorial handing. Funded by DFG-grants RA 442/12 and 27 in the framework of the INDEPTH program.

161

Appendices Appendix 1 Palaeomagnetic results from the Zagaya section: documentation of rock magnetic and remanence investigations The rock magnetic and remanence investigations are detailed in Blumenwitz (2001) and Staiger (2004). Here, we present key elements in Figures 20–23.

Appendix 2 Structural data from the Pung Co ductile detachment and rift For location of stations and parameters of the deviatoric stress tensor, see Table 1; stratigraphic and deformation details are given in the Supplementary material.

References Angelier, J. 1984. Tectonic analysis of fault slip data sets. Journal of Geophysical Research, 89, 5835– 5848. Armijo, R., Tapponnier, P., Mercier, J. L. & Han, T. 1986. Quaternary extension in southern Tibet: field observations and tectonic implications. Journal of Geophysical Research, 91, 13 803–13 872. Armijo, R., Tapponnier, P. & Han, T. 1989. Late Cenozoic right-lateral strike–slip faulting in southern Tibet. Journal of Geophysical Research, 94, 2787– 2838. Besse, J. & Courtillot, V. 2002. Apparent and true polar wander and geometry of the geomagnetic field over the last 200 Myr. Journal of Geophysical Research, 107, doi: 10.1029/2000JB000050. Blisniuk, P. M. & Sharp, W. D. 2003. Rates of late Quaternary normal faulting in central Tibet from U-series dating of pedogenic carbonate in displaced fluvial gravel deposits. Earth and Planetary Science Letters, 215, 169–186. Blisniuk, P. M., Hacker, B. R. et al. 2001. Normal faulting in central Tibet since at least 13.5 Myr ago. Nature, 412, 628–632. Blumenwitz, M. L. 2001. Rock- and Paleomagnetic investigation of a pilot section across the Bangong– Nujiang Suture zone (Central Tibet). Diplomarbeit, Universita¨t Tu¨bingen, 59pp. Cheng, J. & Xu, G. 1986. Geologic map of the Gaize sheet. Tibet Bureau of Geology and Mineral Resources, scale 1:1.000.000, 369 p. (in Chinese). Cogne´, J., Halim, N. & Chen, Y. 1999. Resolving the problem of shallow magnetizations of tertiary age in Asia: insights from paleomagnetic data from the Qiangtang, Kunlun, and Qaidam blocks (Tibet, China) and a new hypothesis. Journal of Geophysical Research, 104, 17,715–17,734. Coleman, M. & Hodges, K. 1995. Evidence for Tibetan plateau uplift before 14 Myr ago from a new

162

L. RATSCHBACHER ET AL.

minimum age for east– west extension. Nature, 374, 49–52. Copeland, P. C. 1990. Cenozoic tectonic history of the southern Tibetan Plateau and eastern Himalaya: evidence from 40Ar/39Ar dating. PhD thesis, State University of New York at Albany, 397pp. Dewane, T. J., Sto¨ckli, D. F., Hager, C., Taylor, M., Ding, L., Lee, J. & Wallis, S. 2006. Timing of Cenozoic E– W Extension in the Tangra Yum Co– Kung Co Rift, South-Central Tibet, AGU fall meeting. Dunlap, W. J., Weinberg, R. F. & Searle, M. P. 1998. Karakoram fault zone rocks cool in two phases. Journal of Geological Society of London, 155, 903– 912. Edwards, M. A. & Ratschbacher, L. 2005. Seismic and aseismic weakening effects in transtension: field and micro-structural observations on the mechanics and architecture of large fault zone in southeastern Tibet. In: Bruhn, D. & Burlini, L. (eds) High Strain Zones in Nature and Experiment. Geological Society, London, Special Publications, 245, 109–141. Gan, W., Zhang, P. et al. 2007. Present-day crustal motion within the Tibetan Plateau inferred from GPS measurements. Journal of Geophysical Research, 112, B08416, doi:10.1029/2005JB004120. Garzione, C. N., DeCelles, P. G., Hodkinson, D. G., Ojha, T. P. & Upreti, B. N. 2003. East–west extension and Miocene environmental change in the southern Tibetan Plateau: Thakkhola graben, central Nepal. Geological Society of American Bulletin, 115, 3 –20. Gilder, S. A., Zhao, X., Coe, R. S., Meng, Z., Courtillot, V. & Besse, J. 1996. Paleomagnetism and tectonics of the southern Tarim basin, northwestern China. Journal of Geophysical Research, 101, 22,015– 22,031. Harrison, T. M., Copeland, P., Kidd, W. S. F., Yin, A. & Lovera, O. M. 1995. Activation of the Nyainqentanghla shear zone: implications for uplift of the southern Tibetan Plateau. Tectonics, 14, 658– 676. Jessup, M. J., Newell, D. L., Cottle, J. M., Berger, A. L. & Spotila, J. A. 2008. Orogen-parallel extension and exhumation enhanced by denudation in the transHimalayan Arun river gorge, Ame Drime massif, Tibet-Nepal. Geology, 36, 587 –590. Jolivet, M., Brunel, M. et al. 2003. Neogene extension and volcanism in the Kunlun fault zone, northern Tibet: new constraints on the age of the Kunlun fault. Tectonics, 22, 1052, doi:10.1029/2002TC001428. Kali, E., Leloup, H. et al. 2010. Exhumation history of the deepest central Himalayan rocks, Ama Drime range: Key P– T– d– t constraints on orogenic models. Tectonics, 29, TC2014, doi:10.1029/ 2009TC002551. Kapp, P., Yin, A. et al. 2000. Blueschist-bearing metamorphic core complexes in the Qiangtang block reveal deep crustal structure of northern Tibet. Geology, 28, 19–22. Kapp, P., Yin, A., Manning, C. E., Harrison, T. M., Taylor, M. H. & Ding, L. 2003. Tectonic evolution of the early Mesozoic blueschist-bearing Qiangrang metamorphic belt, central Tibet. Tectonics, 22, 1043, doi: 10.1029/2002TC001383. Kapp, J. L. D., Harrison, T. M., Kapp, P., Grove, M., Lovera, O. M. & Ding, L. 2005. Nyainqentanghla

Shan: a window into the tectonic, thermal, and geochemical evolution of the Lhasa block, southern Tibet. Journal of Geophysical Research, 110, B08413, doi:10.1029/ 2004JB003330. Kapp, P., Taylor, M., Stockli, D. & Lin, D. 2008. Development of active low-angle normal fault systems during orogenic collapse: insight from Tibet. Geology, 36, 7 –10. Kind, R., Yuan, X. et al. 2002. Seismic images of crust and upper mantle beneath Tibet: evidence for Eurasian plate subduction. Science, 298, 1219–1221. Langin, W. R., Brown, L. D. & Sandvol, E. A. 2003. Seismicity of central Tibet from project INDEPTH III seismic recordings. Bulletin of the Seismological Society of America, 93, 2146–2159. Lee, H.-Y., Chung, S.-L. et al. 2003. Miocene Jiali faulting and its implications for Tibetan tectonic evolution. Earth and Planetary Science Letters, 205, 185–194. Li, Ch., van der Hilst, R. D., Meltzer, A. S. & Engdahl, R. 2008. Subduction of the Indian lithosphere beneath the Tibetan Plateau and Burma. Earth and Planetary Science Letters, 274, 157–168. Liu, M. & Yang, Y. 2003. Extensional collapse of the Tibetan Plateau: results of three-dimensional finite element modeling. Journal of Geophysical Research, 108, 2361, doi:10.1029/ 2002JB002248, 2003. Ludwig, K. R. 1989. A computer program for processing raw Pb– U– Th isotope data. U.S. Geological Survey of Open File Report, 88–557, 39p. Ludwig, K. R. 2003. Isoplot 3.01. A Geochronological Toolkit for Microsoft Excel. Berkeley Geochronology Center Special Publications, 4, 1– 70. Mahe´o, G., Leloup, P. H. et al. 2007. Post 4 Ma initiation of normal faulting in southern Tibet. Constraints from the Kung Co half graben. Earth and Planetary Science Letters, 256, 233 –243. McFadden, P. L. 1990. A new fold test for palaeomagnetic studies. Geophysical Journal International, 103, 163–169. Mercier, J.-L., Armijo, R., Tapponnier, P., CareyGailhardis, E. & Han, T. L. 1987. Change from Tertiary compression to Quaternary extension in southern Tibet during the India-Asia collision. Tectonics, 6, 275 –304. Molnar, P. & Lyon-Caen, H. 1989. Fault-plane solutions of earthquakes and active tectonics of the Tibetan Plateau and its margins. Geophysical Journal International, 99, 123– 153. Murphy, M. A. & Copeland, P. 2005. Transtensional deformation in the central Himalaya and its role in accommodating growth of the Himalayan orogen. Tectonics, 24, doi: 10.1029/2004TC001659. Murphy, M. A., Yin, A., Kapp, P., Harrison, T. M. & Manning, C. E. 2002. Structural and thermal evolution of the Gurla Mandhata metamorphic core complex, southwest Tibet. Geological Society of American Bulletin, 114, 428– 447. Nelson, M. R. & Jones, C. H. 1987. Paleomagnetism and crustal rotations along a shear zone, Las Vegas Range, southern Nevada. Tectonics, 6, 13–33. Phillips, R., Parrish, R. R. & Searle, M. P. 2004. Age constraints on ductile deformation and long-term slip

RIFTING AND STRIKE –SLIP SHEAR IN TIBET rates along the Karakoram fault zone, Ladakh. Earth and Planetary Science Letters, 226, 305–319. Ratschbacher, L., Frisch, W., Chen, C. & Pan, G. 1992. Deformation and motion along the southern margin of the Lhasa block (Tibet) prior to and during the India– Asia collision. Journal of Geodynamics, 16, 21–54. Ratschbacher, L., Frisch, W., Liu, G. & Chen, C. 1994. Distributed deformation in southern and western Tibet during and after the India-Asia collision. Journal of Geophysical Research, 99, 19,917– 19,945. Ratschbacher, L., Frisch, W., Cheng, C. C. & Pan, G. 1996. Cenozoic deformation, rotation, and stress in Eastern Tibet and Western Sichuan. In: Yin, A. & Harrison, T. M. (eds) The Tectonic Evolution of Asia, Rubey Volume, Cambridge University Press, Cambridge, 227– 249. Smith, A. B. & Xu, J. 1988. Palaeontology of the 1985 Tibet Geotraverse, Lhasa to Golmud. Philosophical Transactions of Royal Society of London, A 327, 53– 105. Sperner, B., Ratschbacher, L. & Ott, R. 1993. Fault-striae analysis: a Turbo Pascal program package for graphical presentation and reduced stress tensor calculation. Computers & Geosciences, 19, 1361–1388. Sperner, B. & Ratschbacher, L. 1994. A Turbo Pascal program package for graphical representation and stress analysis of calcite deformation. Zeitschrift deutsche Geologische Gesellschaft, 145, 414– 423. Staiger, M. 2004. Paleomagnetic and structural investigations in unknown areas of central Tibet: a study on block rotations and crustal shortening. Doctoral thesis, Universta¨t Tu¨bingen, Tu¨bingen. 67pp. Stockli, D., Taylor, M., Yin, A., Harrison, T. M., D’Andrea, J., Kapp, P. & Ding, L. 2002. Late Miocene– Pliocene inception of E –W extension in Tibet as evidenced by apatite (U– Th)/He data. Geological Society of American Annual Meeting, 34, 411. Taylor, M. H. & Yin, A. 2009. Active structures of the Himalayan– Tibet orogen and their relationships to earthquake distribtion, contemporary strain field, and Cenozoic volcanism. Geosphere, 5, 199–214. Taylor, M. H., Yin, A., Kapp, P., Ryerson, F. J., Sto¨ckli, D. & Ding, L. 2003. Active conjugate

163

strike–slip faulting along the Bangong–Nujiang suture zone accommodates coeval east–west extension and north– south contraction in the interior of the Tibetan Plateau. Tectonics, 22, 1044–1065. Twiss, R. J. & Moores, E. M. 1992. Structural Geology. W.H. Freeman & Company, New York, XII. Valli, F., Arnaud, N. et al. 2007. Twenty million years of continuous deformation along the Karakorum fault, western Tibet: a thermochronological analysis. Tectonics, 26, TC4004, doi:10.1029/ 2005TC001913. Wang, S., Fan, C., Wang, G. & Wang, E. 2008. Late Cenozoic deformation along the northwestern continuation of the Xianshuihe fault system, Eastern Tibetan Plateau. Geological Society of American Bulletin, 120, 312–327. Williams, H., Turner, S., Kelley, S. & Harris, N. 2001. Age and composition of dykes in southern Tibet: new constraints on the timing of east– west extension and its relationship to post-collisional volcanism. Geology, 29, 339–342. Yin, A. 2000. Mode of Cenozoic east– west extension in Tibet suggesting a common origin of rifts in Asia during the Indo-Asian collision. Journal of Geophysical Research, 105, 21 745– 21 759. Yin, A., Harrison, T. M., Ryerson, F. J., Chen, W., Kidd, W. S. F. & Copeland, P. 1994. Tertiary structural evolution of the Gangdese thrust system, southeastern Tibet. Journal of Geophysical Research, 99, 18,175–18,201. Yin, A., Kapp, P. A. et al. 1999. Significant late Neogene east–west extension in northern Tibet. Geology, 27, 787– 790. Zhang, J. & Guo, L. 2007. Structure and geochronology of the southern Xainza–Dinggye rift and its relationship to the south Tibetan detachment system. Journal of Asian Earth Sciences, 29, 722–736. Zhang, P., Shen, Z., Wang, M., Gan, W., Bu¨rgmann, R. & Molnar, P. 2004. Continuous deformation of the Tibetan Plateau from global positioning system data. Geology, 32, 809–812. Zhao, W., Mechie, J. et al. 2001. Crustal structure of central Tibet as derived from project INDEPTH wideangle seismic data. Geophysical Journal International, 145, 486–498.

Tectonic geomorphology along the eastern margin of Tibet: insights into the pattern and processes of active deformation adjacent to the Sichuan Basin ERIC KIRBY1* & WILLIAM OUIMET2 1

Department of Geosciences, Penn State University, University Park, PA 16802, USA 2

Department of Geography and Center for Integrative Geosciences, University of Connecticut, Storrs, CT 06269, USA *Corresponding author (e-mail: [email protected]) Abstract: We present a review and synthesis of the tectonic geomorphology along the eastern margin of the Tibetan Plateau adjacent to and north of the Sichuan Basin. Re-evaluation of spatial variations in the form of fluvial longitudinal profiles provides a refined image of the distribution of anomalously steep channels. Three new analyses demonstrate that these variations in channel steepness reflect variations in the locus and rate of differential rock uplift. First, measurements of channel width along trunk streams reveal systematic co-variations in channel hydraulic geometry and slope that suggests channels are dynamically adjusted to spatial variations in erosion rate. Second, recent determinations of the functional relationship between channel steepness indices and erosion rate allow a quantitative estimation of erosion rate from channel profile form. Third, comparison of rock uplift patterns to variations in the distribution of slip associated with the 2008 Wenchuan earthquake confirms that channel gradients reflect differential rock uplift. Our analysis suggests that reactivated fault systems adjacent to the Sichuan Basin are primarily responsible for accommodating differential rock uplift, but that rock uplift northward along the margin is not associated with active faults and is likely sustained by flow and thickening in the deep crust.

The eastern margin of the Tibetan Plateau adjacent to the Sichuan Basin comprises one of the most enigmatic mountain belts in the world. Collectively referred to as the Longmen Shan, the range front along this margin of the plateau stands .4000 m above the Sichuan Basin, yet present-day shortening rates are slow, less than a few millimetres per year (e.g. King et al. 1997; Shen et al. 2005). Despite crustal thickness measurements that exceed 60 km (Xu et al. 2007), Cenozoic shortening across the Longmen Shan is limited to a few tens of kilometres (Burchfiel et al. 1995; Hubbard & Shaw 2009). Although rocks along the margin have been exhumed from 8–10 km depth since the mid-late Miocene (Kirby et al. 2002; Godard et al. 2009b), the Sichuan Basin has not been subjected to any significant flexural loading during the Cenozoic (Burchfiel et al. 2008). Collectively, these observations are inconsistent with conventional models of mountain building that consider crustal thickening to occur primarily by shortening on fault systems in the upper levels of the crust. Rather, is has been proposed that Cenozoic development of high topography and thick crust in this region was largely driven by flow in the deep crust (Royden et al. 1997; Royden et al. 2008). This view has been challenged recently by workers who contend that the role of deformation

in the upper crust along the Sichuan Basin has been underappreciated (e.g. Hubbard & Shaw 2009). The catastrophic Wenchuan earthquake of 2008 ruptured two of the range-bounding thrust faults along the margin of the Sichuan Basin (Xu et al. 2009; Zhang et al. 2010) and clearly attests to ongoing deformation along the reactivated Longmen Shan thrust belt (Jia et al. 2009). This event highlights outstanding questions – does the pattern of active differential rock uplift across the plateau margin reflect deformation associated with slip on fault systems in the upper crust, or does the development of topography require a role for thickening at deeper levels of the crust? One means of understanding the patterns of rock uplift at the surface is to utilize geomorphic and exhumational proxies (e.g. Molnar 1987). In eastern Tibet, these data have been used to argue that deformation is localized along the topographic front of the plateau (e.g. Kirby et al. 2003), and, alternatively, that the pattern of fluvial incision reflects a regressive wave of erosion, with minimal influence from differential rock uplift across the margin (Godard et al. 2010). In this review, we synthesize and evaluate arguments for the pattern of erosion across the margin and its relationship to development of topography along the margin of the plateau adjacent to and

From: Gloaguen, R. & Ratschbacher, L. (eds) Growth and Collapse of the Tibetan Plateau. Geological Society, London, Special Publications, 353, 165–188. DOI: 10.1144/SP353.9 0305-8719/11/$15.00 # The Geological Society of London 2011.

166

E. KIRBY & W. OUIMET

north of the Sichuan Basin. Our goal is to place arguments regarding the distribution of active deformation along this margin of the plateau in the context of new data that have become available in the past decade. To complement this synthesis, we present three new sets of observations and analyses that contribute to our understanding of the geomorphological evolution of this margin of the plateau. First, we present an updated analysis of channel longitudinal profiles across the region, and exploit recent empirical data sets (e.g. Ouimet et al. 2009) that allow us to make preliminary interpretations of the patterns of erosion. Second, we present new data on channel width for several of the larger river systems draining the plateau margin, and we argue that channel hydraulic geometry exhibits a dynamic adjustment to differential rock uplift in this landscape. Third, we compare the pattern of erosion predicted by geomorphological analyses to the distribution of slip in the 2008 Wenchuan rupture and show that some of the complexities of the rupture pattern are sensible in light of long-term

uplift pattern. Finally, we explore the implications of the spatial association among regions of steep channels, high erosion rates, and deep exhumation for the tectonic evolution of this margin, and identify future directions for research.

Background The topography along the eastern margin of the Tibetan Plateau has been argued to reflect Cenozoic crustal thickening driven by outward flow of lower crust from beneath the high central Tibetan Plateau (Royden et al. 1997; Clark & Royden 2000). North and south of the Sichuan Basin (Fig. 1), the plateau notably lacks a distinct topographic edge. Elevations gently decrease from .4000 m on top of the plateau to regional elevations of ,1000 m over distances of several thousand kilometres (Clark & Royden 2000). In contrast, the topographic escarpment directly adjacent to and north of the Sichuan Basin comprises one of the

Fig. 1. Topographic characteristics of the eastern margin of the Tibetan Plateau adjacent to and north of the Sichuan basin. (a) Inset map for location. (b) Topographic base is a shaded-relief map generated from SRTM topographic data (nominal resolution c. 90 m). Black lines show the primary active faults along the plateau margin, and blue lines show the locations of major river systems (labelled in white text). (c and d) Swath topographic profiles (i– i0 and j –j0 ) are modified after Kirby et al. (2002) and show the maximum, mean, and minimum elevations along a 75 km wide swath centred on the cross-section lines shown.

TECTONIC GEOMORPHOLOGY IN EASTERN TIBET

steepest of any of the margins of the plateau; mean elevations descend from c. 4000–500 m over distances of ,50 km (Fig. 1). Peaks along the ranges bounding the margin reach 6000–7000 m, and the great relief developed along this margin has been suggested to reflect Cenozoic thickening of the crust (e.g. Kirby et al. 2002; Clark et al. 2005). Recent geophysical surveys in the region confirm that high topography reflects the distribution of thick crust beneath the plateau. Receiver function analysis suggests that the Moho beneath the Tibetan Plateau west and SW of the Sichuan Basin is c. 60 km depth (Xu et al. 2007). Moho depths are as great as c. 65 km beneath the Longmen Shan proper and decrease abruptly to c. 45 km beneath the Sichuan Basin (Zhang et al. 2009; Robert et al. 2010). Simple isostatic models utilizing these variations in crustal thickness are consistent with the gravity anomalies across the Longmen Shan (Burchfiel et al. 2008; Robert et al. 2010), and consistent with the absence of a foreland basin. Moreover, recent acquisition of magnetotelluric data across the plateau west of the Sichuan Basin suggests the presence of high-conductivity channels in the middle to lower crust (Bai et al. 2010), consistent with the presence of a fluid-rich (and therefore weak) crust. Second-order variations in the topography of the plateau margin exist adjacent to and north of the Sichuan Basin. The high ranges of the Qionglai Shan and Longmen Shan trend NE and coincide directly with the margin of the Sichuan Basin (Fig. 1). This region is deeply dissected by the Min Jiang (note, there are several words for river in Chinese, including ‘jiang he’ and ‘qu’), and the headwaters of its major tributaries the Hei Shui He and Somang Qu tap far into the plateau (Fig. 1). Farther north, however, the high topography trends north–south along a narrow range, the Min Shan, and the drainage systems of the Fu Jiang head along the crest of this range (Fig. 1). North of the Min Shan, the topographic margin becomes broader as it cuts orthogonally across the east– west structural grain of the West Qinling. The watershed of the Bailong Jiang also extends far into the plateau north and west of the Min Shan (Fig. 1). We will argue that these topographic differences directly reflect the influence of crustal anisotropy during crustal thickening and Cenozoic development of the plateau.

Geology of the Longmen Shan and Min Shan regions Adjacent to the Sichuan basin, the present-day topographic margin of the plateau coincides with the Longmen Shan thrust belt, a Mesozoic transpressional orogen that developed along the

167

margin of the Yangzte craton during the Triassic (Dirks et al. 1994; Burchfiel et al. 1995). This thrust belt involves crystalline basement, now exposed in topographic massifs along the edge of the Sichuan Basin (Fig. 2), as well as the overlying sedimentary wedge of Neoproterozoic through Permian passive margin deposits. To the west, the orogen involved deep basinal deposits of the Triassic Songpan–Ganzi flysch sequence (Sengo¨r et al. 1993; Zhou & Graham 1996; Weislogel et al. 2006). All of these tectonostratigraphic sequences were shortened and emplaced atop the Yangtze craton during collision with the North China block (Chen et al. 1994a), generating a deep foreland basin along the orogen (e.g. Chen & Wilson 1996). Overlapping relationships of Jurassic sediments with the frontal fault system along the northern margin of the basin (Fig. 2) indicate that these structures have remained inactive since Mesozoic time (Burchfiel et al. 1995; Jia et al. 2009). The Longmen Shan orogen merges with the east –west-trending Qinling orogen north of the Sichuan Basin (Fig. 2). In this region of the Min Shan, Triassic sequences are dominated by shallowwater carbonate rocks that contrast sharply with the deeper water clastic facies to the south and west (Burchfiel et al. 1995). These carbonate platform sequences are generally thought to have developed on a palaeogeographical high (Xue Shan platform) that was incorporated and deformed in the West Qinling –Longmen Shan orogen during Triassic time (Burchfiel et al. 1995). Some of the faults that currently bound the margin of the plateau may have had an ancestry related to this structural and palaeogeographical high (Huya fault – Fig. 2).

Active faults along the plateau margin Along the Longmen Shan, structures within the Mesozoic thrust belt form an imbricate fan of steep, east-vergent thrust faults that shallow and appear to merge with a de´collement at c. 20 km depth beneath the plateau (Burchfiel et al. 1995; Hubbard & Shaw 2009; Jia et al. 2009). Several of these faults are presently active, including the Yingxiu –Beichuan fault (Chen et al. 1994b; Densmore et al. 2007), the primary structure responsible for the 2008 Wenchuan earthquake (Xu et al. 2009), and segments of the Guanxian –Anxian fault (locally referred to as the Pengguan fault). Slip rates on these structures, although not preciselydetermined, appear to be on the order of 1–2 mm a21 (Densmore et al. 2007). In addition, several folds in the southwestern Sichuan Basin are associated with faults at depth (Burchfiel et al. 1995; Hubbard & Shaw 2009) and are likely active, although definitive evidence is lacking.

168

E. KIRBY & W. OUIMET

Fig. 2. Generalized geological map of the eastern margin of the Tibetan Plateau showing primary tectonostratigraphic divisions. Geology modified after Kirby et al. (2002). Heavy lines represent major structures, and light lines represent minor structures. Thrust faults are noted with teeth on the hanging-wall block; normal faults with ball on the hanging-wall block; and strike–slip faults with arrows. Anticlines in Sichuan Basin denoted with thin arrows. P, Pengguan Massif; Y– B fault, Yingxiu–Beichuan fault; W –M fault, Wenchuan– Maowen fault.

TECTONIC GEOMORPHOLOGY IN EASTERN TIBET

West of the Precambrian Pengguan massif (Fig. 2), the NE-trending Wenchuan– Maowen fault is a steep to subvertical fault marked by thick gouge. Most workers consider this structure to be active (Chen et al. 1994b), although it did not rupture during the 2008 Wenchuan earthquake (Xu et al. 2009) and its role in the active deformation field is uncertain. Farther west, active thrust faults are recognized on the plateau (Longriba fault, Fig. 2), but little is known about their history or slip rate (Xu et al. 2009). Along the Min Shan segment of the plateau margin, a pair of active fault systems trend north– south along either side of the range (Fig. 2). To the west, the Min Jiang fault system consists of a series of steep, east vergent reverse faults (Chen et al. 1994b) that appear to have relatively low slip rates (,1 mm a21; Kirby et al. 2000). To the east, the Huya fault is also a steeply west-dipping reverse fault that was responsible for a series of c. M7 earthquakes in 1976 (Jones et al. 1984). The Huya fault lies at the foot of the Min Shan and, along with several small, unnamed faults to the south (Fig. 2), likely contributes to the ongoing uplift of the Min Shan (Kirby et al. 2003). Within the West Qinling, north of the Min Shan, the only recognized active structures are east –westtrending strike –slip faults (Fig. 2). The southern of these, the Tazang fault, appears to be an eastward continuation of the Kunlun fault (Chen et al. 1994b; Kirby et al. 2000; Van der Woerd et al. 2002), but the slip rates appear to be low, ,1 mm a21 (Kirby et al. 2007a). The northern fault, the Bailong Jiang fault, also has relatively slow Quaternary slip rates, ,1–3 mm a21 (Yuan et al. 2007).

Rates and patterns of Cenozoic exhumation Most of the evidence for the growth of the plateau in eastern Tibet rests on thermochronological data that indicate rapid cooling and exhumation of rock during the Late Cenozoic. Following Mesozoic orogeny, much of the region now subject to deep exhumation and high relief experienced extremely slow rates of cooling (c. 18/Ma) throughout the Cretaceous –Early Tertiary (Kirby et al. 2002; Roger et al. 2011), that must reflect relatively slow exhumation (order of 10 –20 m/Ma) and low relief at this time. A pronounced increase in cooling rate during the Late Cenozoic was first observed in apatite fission-track data that suggest an increase in cooling and exhumation during the past c. 20 Ma (Arne et al. 1997; Xu & Kamp 2000). Detailed work utilizing multiple thermochronometers in the Longmen Shan suggests that the onset of rapid cooling occurred at c. 9–12 Ma (Kirby et al. 2002); similar to recent estimates obtained from age-elevation transects along the

169

Longmen Shan range front (8–10 Ma, Godard et al. 2009b). Notably, the timing of the onset of rapid cooling and exhumation along the margin of the Sichuan Basin is similar to that suggested by apatite fission-track length models in the West Qinling (4–9 Ma, Enkelmann et al. 2006), and age-elevation transects in the deep canyons of the southeastern Tibetan Plateau (9–12 Ma, Clark et al. 2005; Ouimet et al. 2010). Thus, it appears that the onset of rapid cooling during the Late Miocene heralds the development of high relief and rapid erosion along the margins of the plateau. Exhumation subsequent to the Late Miocene, however, appears to vary in the magnitude across the margin of the plateau. To the north and east of the plateau margin, in the West Qinling region, exhumation was apparently quite limited (c. 1– 2 km), not extensive enough to completely reset ages in the apatite fission-track system (Enkelmann et al. 2006). Within the Sichuan Basin, east of the plateau margin, apatite fission-track data from Mesozoic basin sediments (Arne et al. 1997) exhibit ages similar to or older than depositional ages (Fig. 3). Recent studies suggest that exhumation appears to have started somewhat earlier in the Tertiary (c. 40 Ma, Richardson et al. 2008), perhaps in response to drainage integration along the Yangtze river (Richardson et al. 2010). Partial resetting of some fission-track ages suggest that up to several kilometres of material may have been removed from the basin, although the spatial distribution of this remains poorly-characterized (Arne et al. 1997; Richardson et al. 2008). West of the plateau margin, apatite fission-track samples from high elevation near or atop the plateau surface exhibit Mesozoic ages (Xu & Kamp 2000), suggestive of relatively limited exhumation. Similar results were obtained in southeastern Tibet, where apatite (U –Th)/He ages are as old as Cretaceous (Clark et al. 2005). In the vicinity of the Sichuan Basin, apatites from the headwater regions of the Hei Shui He and Min Jiang exhibit early Miocene (U –Th)/He ages and Late Cretaceous – early Tertiary fission-track ages (Fig. 3). In addition, zircon (U –Th)/He ages appear to range from Late Cretaceous –early Tertiary (Fig. 3), suggestive of relatively limited exhumation from the plateau surface (Arne et al. 1997; Kirby et al. 2002). In contrast to these regions east and west of the plateau margin, samples from the region of highest topographic relief along the Longmen Shan and Min Shan reveal cooling from c. 180–200 8C during the Late Miocene (Kirby et al. 2002; Godard et al. 2009b). Samples currently exposed at the surface were sufficiently deep in Late Miocene time to completely reset zircon (U –Th)/He ages (Fig. 3) and to perturb Ar retention in K-feldspar (Kirby et al. 2002). These data suggest that the

170

E. KIRBY & W. OUIMET

Fig. 3. Summary of existing thermochronological data from adjacent to the Sichuan Basin. Background image shows topographic relief, measured within a circular window, with a radius of 5 km. Data include apatite fission-track (AFT) ages (Arne et al. 1997) and (U–Th)/He ages from apatite and zircon (Kirby et al. 2002; Godard et al. 2009b). Note that Late Miocene– Pliocene AFT and zircon He ages suggest a locus of deep exhumation (c. 6–10 km) along the topographic front of the plateau. Regions to the west and east exhibit older ages and were apparently subject to lesser exhumation during the Late Cenozoic.

TECTONIC GEOMORPHOLOGY IN EASTERN TIBET

topographic front of the plateau margin has experienced significantly greater exhumation that regions west of the plateau margin (Fig. 3). Late Miocene exhumation of basement rocks along the plateau margin (Pengguan massif) appears to have been on the order of 7–10 km (Kirby et al. 2002; Godard et al. 2009b). Although the width of this region of deep exhumation is not particularly wellcharacterized by existing samples, it appears to be on the order of 60 –80 km in the Longmen Shan and somewhat narrower (30– 40 km) along the Min Shan (Fig. 3).

Patterns of Late Pleistocene erosion One of the more striking aspects of the geomorphology of the plateau margin adjacent to the Sichuan Basin is the coincidence among the region of high topographic relief and high peak elevations, the region of deep Cenozoic exhumation, and a region of steep bedrock river profiles (Kirby et al. 2003). This spatial coincidence led Kirby et al. (2003) to argue that steep channels and high relief were adjusted to ongoing differential rock uplift that must be have been sustained over Late Cenozoic time by mass influx at depth. Recently, Godard et al. (2010) have argued that this same pattern reflects a regressive wave of transient incision migrating headward into the plateau. In this section, we review and update analysis of the longitudinal profiles of river systems draining the eastern margin of the Tibetan Plateau. We then present two new data sets that together argue that spatial variations in channel profile forms reflect a dynamic adjustment to variations in erosion rate. In a subsequent section, we consider the implications for patterns of differential rock uplift along the plateau margin.

Channel profile analysis The interpretation of the longitudinal profiles of fluvial systems has long been utilized as a qualitative indicator of tectonic activity (e.g. Hack 1973; Seeber & Gornitz 1983). In principle, the geometry of incising channel systems in tectonically-active landscapes adjusts such that the erosional efficiency achieves a long-term balance with the rate of differential uplift of rock (defined relative to a fixed, external baselevel; Molnar & England 1990). A well-established body of literature now exists and provides a theoretical framework for how the mechanics of erosion in fluvial systems influences the functional relationship between channel gradients and rock uplift (e.g. Howard & Kerby 1983; Howard et al. 1994; Whipple & Tucker 1999, 2002; Tucker 2004). The reader is referred to the

171

review by Whipple (2004) for a comprehensive discussion of fluvial incision models and their sensitivity to tectonic forcing. Nearly all interpretations of channel profiles rely on an index of channel gradient as a means to explore variations relative to an expected profile shape. Although early workers assumed simple models of exponential or logarithmic profiles (e.g. Hack 1973), more recent work suggests that most fluvial profiles are well-described by a power-law relationship between local channel gradient (S) and contributing drainage area (A), of the form (Flint 1974), S ¼ ks Au

(1)

where ks is a measure of channel gradient, referred to as the channel steepness index (e.g. Whipple & Tucker 1999) and u describes the rate of change of channel gradient with drainage area (a proxy for discharge), referred to as the concavity index. Equation 1 predicts a linear relationship between local slope and drainage area in logarithmic space, where u is the slope of a regression on such a plot and ks is the intercept between the regression and the ordinate axis (e.g. Wobus et al. 2006a). However, in order to effectively compare steepness indices among a suite of channels, one must surmount any autocorrelation between regression slope and intercept. There are two approaches to this, one that simply calculates a reference slope (Sr) at a constant reference drainage area (Ar) (Sklar & Dietrich 1998), and one that utilizes a reference concavity (uref) to calculate a normalized channel steepness index (ksn) (Snyder et al. 2000). Numerous studies have now shown these two methods to be essentially equivalent means of characterizing variations in channel profile form among different channels (Kirby et al. 2003, 2007b; Wobus et al. 2006a). For a more in-depth discussion, the interested reader is referred to the review by Wobus et al. (2006a). Numerous models of fluvial incision into bedrock predict power-law formulations similar in form to equation 1 (Willgoose et al. 1991; Howard et al. 1994; Whipple & Tucker 1999, 2002). Steadystate realizations of these models predict that: (1) the concavity index should be independent of rock uplift rate; and (2) channel steepness should monotonically increase with increasing rock uplift rate, as erosional efficiency increases with increasing gradient (Whipple 2004). However, there are numerous complicating factors that may influence the quantitative relationship between channel steepness and rock uplift rate. These include: (1) orographic variations in the distribution of precipitation and runoff (Roe et al. 2003); (2) non-linearities in the incision process, including the influence of thresholds for

172

E. KIRBY & W. OUIMET

incision (Tucker & Bras 2000; Snyder et al. 2003; Tucker 2004); (3) adjustments in the extent and influence in sediment grain size, alluvial cover and/or hydraulic roughness (Sklar & Dietrich 1998, 2004, 2006, 2008; Turowski et al. 2007); (4) adjustments in channel hydraulic geometry (Duvall et al. 2004; Finnegan et al. 2005; Stark 2006; Wobus et al. 2006b, 2008; Whittaker et al. 2007a); and (5) possible influence of debris-flow incision (Stock & Dietrich 2003, 2006). In addition, rock mass quality can have a pronounced effect on channel profile shape (Moglen & Bras 1995; Stock & Montgomery 1999; Lave´ & Avouac 2001; Duvall et al. 2004). Despite these outstanding questions and studies showing that incision rates can be modulated by sediment flux (e.g. Johnson et al. 2009), many empirical studies have demonstrated a positive, monotonic correlation between channel profile form and erosion rate in simple field sites where the patterns and rates of tectonic activity are known (e.g. Snyder et al. 2000; Kirby & Whipple 2001; Lague & Davy 2003; Duvall et al. 2004; Wobus et al. 2006a), or where erosion rates can be measured independently (Safran et al. 2005; Harkins et al. 2007; Ouimet et al. 2009; DiBiase et al. 2010). Moreover, applications of these techniques to landscapes where underlying deformation patterns are not known have been shown to have some utility in illuminating both patterns of differential rock uplift (Kirby et al. 2003, 2007b; Wobus et al. 2003; Hodges et al. 2004) and in discriminating transient behaviour (e.g. Crosby & Whipple 2006; Harkins et al. 2007; Whittaker et al. 2007b, 2008). Collectively, these studies indicate that channel profiles can provide important constraints on the rates and patterns of differential rock uplift across erosional landscapes (e.g. Lave´ & Avouac 2001). The eastern margin of Tibet was the subject of one of the first applications of these approaches to understanding the pattern of differential rock uplift (Kirby et al. 2003). These authors examined c. 120 channels draining the Longmen Shan and Min Shan regions and observed systematic differences among channels in each of these regions. Channels draining across the Longmen Shan exhibited downstream increases in channel gradient and convex profiles, whereas channels draining the Min Shan had extremely concave profiles, with high gradients in the headwaters that rapidly decreased away from the plateau margin (cf. Fig. 7 of Kirby et al. 2003). Notably, these patterns persisted across a range of lithologies, and changes in profile form did not correspond to lithological boundaries. Moreover, tributary channels showed the same spatial differences in steepness index, and Kirby et al. (2003) argued that these patterns

were sustained by high rates of differential rock uplift that was localized along the plateau margin. Their study was subject to several limitations, however, that reflected the early development of the analysis of channel profiles. In particular, Kirby et al. (2003) were forced to use tributary junctions to delimit the extent over which to define a channel steepness index. This restricted the analysis to long spatial reaches, primarily along major channels, which in turn led to averaging across reaches of varying gradient. In the next section, we present an updated analysis of the channels in the Longmen Shan and Min Shan that circumvents some of these issues.

Spatial patterns of channel steepness in eastern Tibet Using the Shuttle Radar Topography Mission 3-arc-second topographic data (nominal pixel resolution of c. 90 m), we determined the normalized channel steepness (ksn) for every tributary within the major watersheds draining into the Sichuan Basin from the Longmen Shan, Min Shan, and West Qinling regions (Fig. 1). Guided by previous work that demonstrated that channels in this region exhibit a fluvial scaling only above a critical drainage area between 1–10 km2 (Kirby et al. 2003), we excluded portions of the channel network with drainage areas less than 10 km2. Channel network extraction was accomplished following techniques described by Wobus et al. (2006a). We removed spikes in the elevation of each profile, and smoothed the raw data using a moving average with a window size of 2 km. To affect a regular sampling of the channel profile, normalized steepness and concavity indices were calculated on a 5 km interval along the channel. Normalized steepness indices were calculated using a reference concavity of 0.45; this value was chosen both because it is typical of concavities in landscape undergoing uniform rock uplift (Kirby & Whipple 2001) and for easy comparison to published studies (e.g. Wobus et al. 2006a; Ouimet et al. 2009). Increases or decreases in steepness index along a river or between adjacent rivers thus represent deviations from the typical steady-state profile and could be interpreted in terms of spatially variable uplift or transient response of the landscape to a perturbation. Finally, we interpolated channel steepness indices onto a regular grid with an inverse-distance weighted average of all values within a 30 km radius. The results of this analysis are shown in Figure 4. Overall, our results confirm previous studies (Kirby et al. 2003; Godard et al. 2010), and reveal a region of steep channels that is broadly coincident

TECTONIC GEOMORPHOLOGY IN EASTERN TIBET

173

Fig. 4. Channel steepness indices for the Sichuan margin of the Tibetan Plateau. (a) Watershed of the Min Jiang and major tributaries considered in the analysis. (b) Map of channel steepness indices calculated at 5 km intervals along all major channels (drainage areas .10 km2). (c.) Interpolated map of average steepness indices calculated within a moving circular window (radius of 15 km). (d) Example profiles showing the distribution of channel steepness along individual trunk rivers. The Fu Jiang and Jian Jiang drain the eastern flank of the Min Shan and exhibit ksn values that systematically decrease from c. 250 –300 m0.9 in their headwaters to ,50 m0.9 in the Sichuan Basin. In contrast, ksn values in the headwaters of the Min Jiang/Hei Shui He and Somang Qu are c. 100 m0.9 and increase downstream to maxima of 300–500 m0.9 in the Longmen Shan.

with the region of high topography adjacent to and north of the Sichuan Basin (Fig. 4). Although the first-order pattern of channel steepness is similar to the study of Kirby et al. (2003), our re-analysis

allows us to significantly refine the spatial distribution of channel steepness (Fig. 4c). As before, the steepest channels occur within the Longmen Shan, immediately west of the Sichuan Basin. They are

174

E. KIRBY & W. OUIMET

not confined to the frontal escarpment, however, but extend c. 50 km west of the topographic margin. Furthermore, we observe a fairly high degree of correlation between steepness indices along major trunk rivers with tributaries (Fig. 4b), whereas previous studies appeared to show that trunk streams were somewhat less steep than corresponding tributaries (cf. Fig. 8 of Kirby et al. 2003). Given that our new analysis does not differ from that of Kirby et al. (2003), except in the interval over which channel steepness indices are calculated, we attribute this difference to the relatively coarse characterization of channel reaches in that early work. In a similar fashion, the region of steep channels extends northward from the Longmen Shan and follows the front of the Min Shan away from the margin of the Sichuan Basin (Fig. 4c). Although this general pattern is consistent with the study of Kirby et al. (2003), our revised analysis reveals subtle differences. For instance, the region of high channel steepness follows a fairly narrow zone in the southern part of the range, but then widens toward the north (Fig. 4c). In the next section, we consider the implications of this pattern for the active tectonics of the region. The primary differences in the pattern of channel steepness between these regions of the plateau margin are perhaps best illustrated with example profiles (Fig. 4d). Channels in the Min Shan exhibit steepness indices of c. 300 m0.9 along their headwater reaches (Fu Jian and Jin Jiang, Fig. 4d); steepness indices progressively decrease eastward, away from the plateau margin, to values of ,50 m0.9 at the margin of the Sichuan Basin. In contrast, rivers in the Longmen Shan exhibit values of ksn in their headwaters of c. 100 m0.9; these channels steepen to values approaching c. 300– 350 m0.9 near the plateau margin (Heishui He and Somang Qu, Fig. 4d). Channels in the Longmen Shan do appear to be somewhat steeper, with individual reaches achieving ksn values as great as 500 m0.9. Regionally, however, the overall picture is one of a broad region of steep channels, both trunk streams and tributaries, coincident with the high ranges that define the plateau margin. In the following sections, we address the question of to what degree these steep channels reflect variations in incision rate across the plateau margin.

Dynamic adjustment of channel width One of the canonical aspects of fluvial geomorphology is the nearly ubiquitous observation that the width of self-formed fluvial channels scales with the square root of average discharge (Leopold & Maddock 1953; Leopold et al. 1964). Although originally developed in alluvial rivers, similar scaling relationships are observed in many bedrock

systems (Montgomery & Gran 2001; Whipple 2004). A growing body of work, however, suggests that bedrock channels may adjust their hydraulic geometry in response to variations in incision rate (e.g. Duvall et al. 2004; Finnegan et al. 2005; Amos & Burbank 2007; Whittaker et al. 2007a). Although the process by which this occurs is incompletely understood, recent work builds upon a rational theory developed for alluvial channels (e.g. Parker 1978a, b; 1979) and suggests a central role for the sediment load conveyed in ‘mixed’ bedrock-alluvial channels (e.g. Stark 2006; Wobus et al. 2006b, 2008; Turowski et al. 2007, 2008), one that may be strongly influenced by the frequency of extreme discharge events (Hartshorn et al. 2002). In order to ascertain to what degree trunk river systems in eastern Tibet exhibit variations in width that deviate from simple hydraulic scaling, we measured channel widths along several of the primary channels draining the eastern margin of the plateau. Widths were surveyed with a laser rangefinder at c. 2 km intervals along the Hei Shui He, lower Min Jiang, Somang Qu, and Wolong in the Longmen Shan during 1997–1998 and 2003– 2006, and along the Fu Jiang and Jian Jiang in the Min Shan during 1997–1998 (Fig. 5). In all cases, the width was defined as the high-flow condition, marked by the lower limit of vegetative cover and highest river polish visible on outcrops. Although we do not know precisely what flow stage this level represents, we presume that it reflects floods that recur frequently enough to prevent vegetative colonization of channel banks. In this landscape, this may be as little as 2–3 years. Although the lack of detailed hydrological studies along these rivers precludes a comprehensive analysis, the results of our surveys reveal systematic deviations in hydraulic geometry scaling that correlate with changes in river gradients across this margin of the plateau (Fig. 5). In particular, both tributaries of the Min Jiang, the Hei Shui He and Somang Qu, exhibit scaling relationships with exponents significantly lower than 0.5 (Fig. 5), signifying that increases in channel width downstream are less than expected from simple hydraulic scaling (e.g. Montgomery & Gran 2001). In contrast, rivers draining the eastern flank of the Min Shan (Fu Jiang and Jian Jiang) exhibit scaling relationships with exponents significantly greater than 0.5 (Fig. 5), suggesting that width is increasing downstream more rapidly than expected. It is notable that widths and channel gradients appear to co-vary along these rivers – the broad increase in channel steepness as rivers enter the range front is associated with narrow channels, whereas the rapid decrease in channel steepness away from the Min Shan is associated with wide channels. Thus,

TECTONIC GEOMORPHOLOGY IN EASTERN TIBET

Fig. 5. Hydraulic geometry of trunk rivers in eastern Tibet. Map shows the location of all survey points. (a– d) Scaling relationships for channel width. Circles show measured values, and boxes show log-bin averages.

175

176

E. KIRBY & W. OUIMET

to a first-order, these patterns appear to confirm the inference that channels are dynamically adjusted to match variations in rock uplift rate along this margin (e.g. Kirby et al. 2003). The co-variance between channel width and gradient begs the question of whether a more sophisticated scaling rule can explain this behaviour as a response to spatially non-uniform rates of rock uplift (e.g. Finnegan et al. 2005). Assuming spatially uniform and constant hydraulic roughness, and a self-similar channel cross-section, Finnegan et al. (2005) derived a simple relationship that suggests that channel width (W ) should scale as W / Q5=8 S3=16

(2)

where Q is the channel-forming discharge and S is channel slope. These authors showed that this formulation both accounts for width changes along rivers with typical concave-up profiles as well as explains co-variance of width and slope along convex reaches of the Tsangpo river, in southeastern

Tibet (Finnegan et al. 2005). Moreover, recent modelling efforts (Wobus et al. 2008) arrive at a similar functional form starting from a different set of assumptions. To evaluate whether this model explains our data from eastern Tibet, we compare the predictions of both standard hydraulic scaling (W / Q 0.5) and that of equation 2. Results are shown in Figure 6 for two of the rivers in our study area. Our simple analysis assumes a direct relationship between the channel forming discharge event and drainage area in the watershed. Although the hydrology of these rivers is not well known, global databases of precipitation (e.g. New et al. 2002) suggest that the effects of orographic variations in precipitation are probably minimal in this region. Nonetheless, we regard this analysis as preliminary. For both rivers in the Longmen Shan and Min Shan, the scaling relationship in equation 2 (Finnegan et al. 2005) more closely predicts the variation of channel width along trunk rivers, relative to the simple scaling (Fig. 6).

Fig. 6. Channel width v. streamwise distance for the Min Jiang/Hei Shui He (a) and Jian Jiang (b). Black line is the longitudinal profile of the river. Solid diamonds represent channel widths surveyed in the field. Grey squares represent channel widths predicted by a simple scaling of hydraulic geometry, and open squares represent widths predicted by the scaling relationship of Finnegan & others (2005). See text for details.

TECTONIC GEOMORPHOLOGY IN EASTERN TIBET

Overall, our results reveal a strong correlation between channel gradients and width that suggest a dynamic adjustment to regional forcing. Although the steep gradients and narrow widths in the Longmen Shan could represent a transient wave of incision (e.g. Godard et al. 2010), such a mechanism cannot easily explain highly-concave profiles in the Min Shan region that exhibit the opposite patterns of channel steepness and width. To test whether these differences in channel form between the Longmen Shan and the Min Shan reflect spatially-variable tectonic forcing (as argued previously, Kirby et al. 2003), we explore the relationship of channel profiles to independent measures of erosion rate and rock uplift rate.

Functional relationship between channel steepness and erosion rate Empirical data sets. The past decade has also seen widespread application of cosmogenic isotopes to problems in landscape evolution (e.g. Gosse & Phillips 2001). In particular, inventories of cosmogenic 26Al and 10Be in modern alluvial sediment can provide a means of estimating average erosion rates across watersheds (Bierman & Steig 1996; Granger et al. 1996), provided that sediment is wellmixed and representative of the upstream catchment. The latter constraint implies that sampling of catchments dominated by landslide-driven erosion requires large enough basins to ensure that sediment is well-mixed (e.g. Niemi et al. 2005). Applications of this technique to actively-eroding

177

mountain ranges across the globe have begun to develop empirical relationships between channel steepness indices and erosion rates in the Bolivian Andes (Safran et al. 2005), in the San Gabriel Mountains of California (DiBiase et al. 2010), along the Yellow River (Harkins et al. 2007), and in the canyons of southeastern Tibet (Ouimet et al. 2009). The latter two studies hold particular significance for interpretation of channel profiles in this contribution, as they were conducted in adjacent regions, with similar climate and rock types as the Sichuan margin of the Tibetan Plateau. In southeastern Tibet, Ouimet et al. (2009) sampled 65 basins from major tributaries of the Yangtze River spanning a range of rock types from greywacke to granite. Excluding outliers and glaciated catchments near the Gongga Shan massif (Ouimet et al. 2009), the data show a remarkable correlation between the average channel steepness of the upstream catchments and basin-wide erosion rate (Fig. 7). Similar relationships have been observed within the Anyemaqen Shan, in the headwater reaches of the Yellow River (Harkins & et al. 2007). Finally, Godard et al. (2010) sampled a few catchments within the Longmen Shan itself that we include in this compilation. In order to maintain consistency among all data sets, we re-analysed channel steepness indices from both the Yellow River and Longmen Shan samples following techniques of Ouimet et al. (2009). We also excluded large trunk streams sampled by Godard et al. (2010), because of potential issues with dams and recent anthropogenic disturbance in these catchments.

1000

Erosion Rate (m Ma-1)

800 600

Granite and crystalline basement Ouimet et al., 2009 Godard et al., 2009

400

Greywacke and schist Ouimet et al., 2009

200

Harkins et al., 2007

0 0

100

200

300

400

500

Normalized Channel Steepness - ksn (m ) 0.9

Fig. 7. Functional relationship between channel steepness index (ksn) and erosion rate compiled from studies in eastern Tibet. Data are from Ouimet et al. 2009 (circles), Harkins et al. 2007 (triangles), and Godard et al. 2010 (squares). Average ksn for watersheds from the latter two studies were calculated following methods described in Ouimet et al. 2009. Data are categorized by substrate lithology; dark shades (black and dark grey) represent watersheds underlain by Mesozoic granites and/or crystalline basement rocks, and light shades (light grey and white) represent watersheds underlain by greywacke (primarily Triassic Songpan–Ganzi flysch), schist, and other sedimentary rocks. Shown are two possible fits to the entire data set; solid line is a linear model, and the dashed line represents the best-fit exponential model.

178

E. KIRBY & W. OUIMET

Fig. 8. Role of lithology on channel steepness index. Map shows the bedrock geology in the frontal portions of the Longmen Shan. Crystalline basement in the Pengguan and Xuelongbao massifs are shown in pink. Geology after 1:200 000 sheets published by the Ministry of Geology and Mineral Resources. (a –i) Channel profiles of tributaries to the Min Jiang, with channel steepness index (ksn) values calculated at 1 km intervals. Note the invariance of ksn with substrate lithology – channels a, b, and c are contained entirely within the sedimentary rocks on the NW flank of the massif, channels c, d, and e are transitional, and channels f– i are developed entirely within basement rocks.

TECTONIC GEOMORPHOLOGY IN EASTERN TIBET

When combined with the data from southeastern Tibet (Ouimet et al. 2009), all three studies define a positive correlation between ksn and erosion rate. The data exhibit a fair degree of scatter, particularly at high erosion rate (Fig. 7), and it is not known whether this reflects a degree of unsteadiness in erosion, perhaps associated with landslides (e.g. Niemi et al. 2005) or whether it reflects adjustments in channel bed state, incision process, and/ or hydraulic roughness that may accompany adjustments in channel slope. Nonetheless, the correspondence among all three data sets, including data from the Longmen Shan itself, indicates that channel gradients adjust to accomplish differences in erosion rate in eastern Tibet. Thus, to a first-order, these relationships provide definitive evidence that variations in the pattern of channel steepness indices in eastern Tibet do indeed reflect variations in erosion rate (e.g. Kirby et al. 2003). The exact nature of this relationship remains somewhat uncertain. Ouimet et al. (2009) argued for a non-linear relationship, where the rate of increase of channel steepness indices slows with increasing erosion rate, suggesting more efficient erosion in steeper channels. Although thresholds associated with the detachment of bedrock and transport of sediment, provide a compelling explanation for such behaviour (e.g. Snyder et al. 2003), other studies have argued for a linear relationship between ksn and erosion rate (Safran et al. 2005; Wobus et al. 2006a; Harkins et al. 2007). The combined data from eastern Tibet appear to be reasonably explained by either model (Fig. 7); an exponential fit to the data explains c. 66% of the variance, whereas a linear fit explains c. 58%. Thus, although we have a slight preference for the non-linear relationship, in what follows we use both as end-member constraints on the distribution of erosion rates in eastern Tibet. The role of lithology. Variations in rock mass strength and quality present one potentially significant complication to the functional relationship between channel steepness and erosion rate. Rock strength is well-known to limit the ability of bedrock streams to incise, as both the intact tensile strength and the degree of fracturing influence detachment thresholds (e.g. Howard 1980; Stock & Montgomery 1999; Whipple et al. 2000; Sklar & Dietrich 2001). In addition, the longevity of large caliber sediment downstream can also drive channel gradients toward a transport-limited condition (e.g. Sklar & Dietrich 1998; Lave´ & Avouac 2001; Whipple & Tucker 2002; Duvall et al. 2004; Johnson et al. 2009). In eastern Tibet, recent experimental determination of the rate of pebble abrasion (Godard et al. 2010) suggests fairly large variations in the wear

179

rate of intact rock. These authors inferred from this result that lithological variations in the range might significantly impact the relationship between channel gradients and incision rate. In particular, their results suggest that crystalline rocks comprising the Pengguan massif near the plateau margin and Mesozoic plutons scattered throughout the ranges resist wear to a greater degree than the Palaeozoic –Mesozoic sedimentary rocks (see Fig. 8 of Godard et al. 2010). Although they acknowledge that these results do not consider the degree of fracturing of the intact rock, Godard et al. (2010) used them to calibrate a fluvial incision rule based on fluvial shear stress. Notably, application of this model predicts large variations in incision rate on either flank of the Pengguan massif, with high rates along the northwestern flank of the massif, where Palaeozoic rocks outcrop, and significantly lower rates in the Precambrian basement, which is exposed along the eastern flank and in the southern part of the massif (see Fig. 13 of Godard et al. 2010). We argue that this pattern is likely an artifact of the calibrated incision rule, and that lithological variations do not exert as strong a control on incision rates as implied by this study. Our arguments stem from observations that variations in stream gradients are not significantly correlated to lithological variations. First, regional measurements of erosion rate (Ouimet et al. 2009) do not show a strong dependence on lithology (Fig. 7). Although we see a general tendency that steeper channels are associated with crystalline lithologies, both crystalline rocks and sedimentary rocks lie on the same trend. Notably, samples from the Pengguan massif itself (Godard et al. 2010) exhibit variations in erosion rate from c. 0.3 mm a21 to 0.7 mm a21 (Fig. 7), variations that span c. 50% of the data set. Likewise, samples from slowly eroding basins exhibit similar channel steepness indices, independent of lithology (Fig. 7). Thus, although we are mindful of the scatter in existing data, we tentatively conclude that lithological variations play a second-order role in the distribution of channel gradients in eastern Tibet. In addition, as argued previously (Kirby et al. 2003), changes in the gradients of major trunk rivers do not correspond to lithological variations. The steepening of channels entering the Longmen Shan from the west begins well upstream of the Wenchuan –Maowen fault (Fig. 4) and thus upstream of the crystalline basement. Likewise, streamwise decreases in channel steepness indices along the Fu and Jian rivers draining the Min Shan (Fig. 4) occur largely within Palaeozoic though Mesozoic sedimentary rocks that have similar resistance to wear (Godard et al. 2010). The lack of an obvious correlation between channel gradients and substrate lithology exists on

180

E. KIRBY & W. OUIMET

a smaller scale as well. We illustrate this point with nine tributary channels of the Min Jiang (Fig. 8), all of which are less than 10 km in length. Five of these drain the NW flank of the massif; three are entirely within the Neoproterozoic–Palaeozoic sedimentary rocks, and two drain across the basement– sedimentary cover contact (Fig. 8). All exhibit relatively smooth profiles, with average ksn values (sampled over 1 km channel segments) ranging from c. 285 –330 m0.9 (Fig. 8). Channels that drain granitic lithologies within basement rocks nearby span the same range of values, from c. 250 – 340 m0.9 (Fig. 8). All of these channels drain directly into the Min Jiang, and thus should be subject to the same forcing at their local base levels. The absence of any discernable influence of bedrock lithology on channel steepness is somewhat counter-intuitive, as most data lead to an expectation of shallower gradient systems in weaker rock (e.g. Duvall et al. 2004). We tentatively conclude that processes other than rock resistance to wear provide the first-order control on channel gradients in this massif. The sedimentary rocks along the NW flank of the massif were not subject to the tremendous landslides that occurred within basement rocks during the 2008 Wenchuan earthquake (Ouimet 2009). We suspect that extensive fracturing within the basement massif leads to macroscopic strength reduction (e.g. Molnar et al. 2007; Clark & Burbank 2010), and that the resistance to erosion of these rocks is significantly less than one would infer from abrasion experiments (Godard et al. 2010). Overall, although we consider it likely that lithology exerts some manner of influence on channel gradients in the study area we believe that existing data suggests this is a second-order feature. To our thinking, the regional patterns in channel steepness and erosion rate most likely reflect a dynamic adjustment of channel gradients to variations in tectonic forcing. This is supported by observed variations in channel width along trunk streams, which do not appear strongly dependent on substrate lithology, as well as by empirical correlations between basin-wide erosion rate and channel steepness indices. Although rock strength must play a role, it is likely dominated by fracture distribution and the susceptibility of the landscape to infrequent failures by landslides (e.g. Ouimet 2009).

Patterns of differential rock uplift The argument that channels in eastern Tibet are dynamically adjusted to erosion rate leads directly to the inference that the distribution of channel steepness along the plateau margin may be a reasonable proxy for the distribution of erosion. However,

the degree to which these patterns reflect the distribution of differential rock uplift (e.g. Kirby et al. 2003) v. regressive erosion of a plateau margin (Godard et al. 2010) is still a matter of debate. Here we consider the perspective provided by both long-term patterns of exhumation, and short-term variations in rock uplift associated with the 2008 Wenchuan earthquake. We utilize the empirical relationships between channel steepness indices and erosion rates presented in Figure 7 to develop estimates of the spatial patterns of erosion along both the Longmen Shan and Min Shan sections of the plateau margin (Fig. 9). Given the uncertainties associated with lithological differences, as well as the variance in these data, we consider these maps preliminary. Nonetheless, even though the absolute magnitude of erosion rate estimates may be somewhat uncertain, we believe that relative differences in erosion rate across the region are likely correct.

Spatial variations in slip during the Wenchuan earthquake The Wenchuan earthquake of 2008 provides an opportunity to assess the degree to which inferred patterns of erosion rate are related to short-term variations in differential rock uplift associated with co-seismic slip. In addition to the sheer size of this event (Mw c. 7.9, Ji & Hayes 2008), several aspects of the rupture itself are noteworthy. First, the rupture involved a complicated array of faults, including both a c. 250 km long surface rupture along the Yingxiu–Beichuan fault, as well as a shorter (c. 60 km) rupture along the Pengguan fault (Xu et al. 2009; Zhang et al. 2010). Second, the geometry of the fault systems that were activated during the event varied along-strike; in the south, the Yingxiu–Beichuan and Pengguan faults are listric structures that dip 50 –708 at the surface and shallow to merge with a de´collement at depth (Zhang et al. 2010), whereas in the north, the Yingxiu– Beichuan fault is a subvertical structure (Zhang et al. 2010). Third, the rupture itself involved oblique –reverse slip, but the magnitude of co-seismic displacements varied considerably along-strike. Numerous field observations of the surface rupture (e.g. Li et al. 2009; Lin et al. 2009; Xu et al. 2009; Zhang et al. 2010), finite fault inversions of seismic wave radiation (Ji & Hayes 2008; Ji et al. 2008), leveling data (Zhang et al. 2010), and joint inversion of GPS and InSAR geodetic data (e.g. Shen et al. 2009) all indicate that the southern portion of the rupture exhibited up to 6 m of thrust-sense displacement, whereas the northern portion was characterized by c. 1–2 m of dip–slip displacement. Although

TECTONIC GEOMORPHOLOGY IN EASTERN TIBET

181

Fig. 9. Maps of predicted erosion rate in eastern Tibet, based on the empirical relationship between channel steepness index (ksn) and erosion rate (Fig. 6). Although the difference in predicted erosion rates between a linear model (a) and an exponential model (b) is not large, each makes somewhat different predictions for the width of the locus of high erosion rate along the plateau margin.

maximum slip along the primary rupture was associated with fault junctions near Yingxiu and Beichuan (Shen et al. 2009, Fig. 10a), when one considers slip along the Pengguan fault, the total thrust-sense slip exceeds c. 4–5 m along most of the southern 150 km of the rupture (Fig. 10a). North of Beichuan, the thrust component of slip decreases rapidly toward the northern tip of the rupture. Comparison of the distribution of co-seismic slip to the distribution of high erosion rate inferred from the geomorphic analysis reveals a striking correspondence. The region of large vertical displacement coincides with the region of steep gradients in modelled erosion rate along the eastern flank of the Pengguan massif (Fig. 10b). Toward the NE, the rupture propagated into a region characterized by a lower erosion rate. Notably, we see no significant difference in channel steepness indices (or model erosion rates) on either side of the fault in this region, suggesting that there has been little active differential rock uplift across this structure in recent times (Kirby et al. 2008). Whether this pattern of differential vertical motion along the fault system has been sustained over geological timescales is difficult to assess. The magnitude of differential exhumation across the fault system provides a loose constraint on the relative vertical motion. Along the margin of the

Sichuan Basin, differential exhumation between the Pengguan massif and the Sichuan Basin suggests the Beichuan and Pengguan faults likely have accomplished at least 5–8 km of throw during the Late Cenozoic (Godard et al. 2009b). Farther north, data are sparse (Fig. 3), but a single apatite sample from the northwestern side (hanging wall) of the Yingxiu–Beichuan fault yielded an Oligocene age (Fig. 3), suggestive of relatively limited late Cenozoic exhumation (Arne et al. 1997). Fission-track ages from the Sichuan Basin immediately to the east are unreset (Fig. 3), and thus, it appears possible that the northern segments of this fault system have experienced less cumulative throw than southern segments, although additional work will be required to confirm this. Overall, the correlation among the distribution of throw observed during the 2008 Wenchuan earthquake (Shen et al. 2009), the pattern of channel steepness indices and short-term erosion rates (Fig. 10), and the tentative distribution of longterm exhumation (Fig. 3) suggests that the northeastern segments of the Beichuan fault are characterized by relatively limited differential vertical motion. Although the correspondence between channel steepness indices and the spatial distribution of slip in the event was noted previously (Kirby et al. 2008), the updated geomorphic analysis presented here, as well as improvements in the

182

E. KIRBY & W. OUIMET

Fig. 10. Relationship between differential rock uplift and slip during the 2008 Wenchuan earthquake. (a) Co-seismic slip distribution during the earthquake. Graph shows dip– slip displacement from the joint InSAR and GPS inversion of Shen et al. 2009. Data represent the reverse component of slip in the upper 4 km of the crust as a function of distance along the rupture. Slip on the Yingxiu–Beichuan fault shown as solid circles, and slip on the Pengguan fault shown as open circles. Line shows the approximate vault of the sum of surface displacement on these two faults. Arrow shows the location of the slip maxima near the town of Beichuan. (b) Relationship between rupture pattern and pattern of steep channels/high erosion rate. North and east of Beichuan, the vertical component of the rupture decreases systematically, consistent with the location of the rupture relative to the locus of inferred rock uplift.

geodetic inversions (Shen et al. 2009), argue even more strongly that channel profile steepness indices provide a measure of the long-term patterns of differential rock uplift in eastern Tibet.

Coincidence of short-term and long-term loci of exhumation As noted previously (Kirby et al. 2003), one of the most striking results is the co-location of the locus of high erosion rate inferred from channel profiles and the degree of exhumation suggested by the distribution of young cooling ages in low-temperature thermochronological systems (Arne et al. 1997; Kirby et al. 2002; Godard et al. 2009b). Rocks in both the Longmen Shan and Min Shan have been exhumed from 7 –9 km depth in the past 10– 15 Ma (Kirby et al. 2002; Godard et al. 2009b), yet these are still the topographically-highest parts of the plateau margin (Fig. 1). It seems evident that some mass flux at depth is required to both generate and sustain high relief along the margin from Miocene to present. Moreover, it is intriguing to note the similarity in erosion rates measured over these time intervals. Average exhumation rates over the past c. 10 Ma of c. 300 –600 m/Ma provide a reasonable explanation of existing thermochronological data (Kirby et al. 2008; Godard

et al. 2009b), and our estimates of present-day erosion rates appear to range from c. 300–400 m/ Ma along the locus of steep channels (Fig. 9). Although the data allow for the possibility that exhumation rates have slowed somewhat in the past several Ma (e.g. Godard et al. 2009b), limited data precludes a definitive assessment of this hypothesis at the present time. Given uncertainties in both data sets, it is just as plausible that exhumation rates have remained quasi-steady over the development of the plateau margin. Regardless of this detail, some degree of mass influx at depth is required to maintain high topography and to sustain high erosion rates along the plateau margin.

Implications for active tectonics in eastern Tibet We have argued above that the spatial correspondence between channel steepness indices and erosion rates along the eastern margin of the Tibetan Plateau suggests that channels are dynamically adjusted to spatial variations in tectonic forcing across the range. Moreover, variations in erosion rate appear to have been sustained by differential rock uplift, relative to base level in the Sichuan Basin, since Miocene time. Here we consider the

TECTONIC GEOMORPHOLOGY IN EASTERN TIBET

implications of these rates of rock uplift for the development of the plateau and ongoing tectonism along the Sichuan basin. Although we previously argued that there was little association between the pattern of differential rock uplift and active faults (e.g. Kirby et al. 2003), our recent analysis suggests a slightly refined perspective. We see somewhat varied behaviour along different portions of the plateau margin that suggest a more complicated association between active faults and the uplift of rock.

Role of faults adjacent to the Sichuan Basin In the Longmen Shan region, the difference in channel steepness across the Yingxiu–Beichuan fault system must largely reflect long-term differential rock uplift associated with this fault system. Although we are not yet able to draw quantitative conclusions about the long-term rates of uplift associated with this fault, our empirical relationships suggests differences in erosion rate of at least c. 300 m/Ma (Fig. 9) between rocks in the hanging wall and rocks in the footwall of this fault system. Given similar differences in exhumation across this fault system in the past 10 Ma (Kirby et al. 2002; Godard et al. 2009b), this likely provides a minimum constraint on the average vertical component of fault slip during the late Pleistocene. This fault system is known to have relatively low throw rates during the Holocene (,1 mm a21; Densmore et al. 2007) and rather long recurrence intervals (.2000– 4000 a, Densmore et al. 2007; Burchfiel et al. 2008; Shen et al. 2009). We find the correspondence between geomorphology and long-term fault slip encouraging; although not yet sufficient for hazard evaluation, analysis of stream profiles appears to hold promise in identifying regions of active differential rock uplift. Perhaps as striking as this is our result that there appears to be little difference in the rates of differential rock uplift across the Wenchuan– Maowen fault system (Fig. 9). This fault system represents a major break in the geology of the region; Triassic rocks west of the fault belong to the deep potions of the Songpan-Ganzi basin, whereas those to the east have affinities to the Yangtze craton and the Sichuan Basin (e.g. Chen et al. 1994a; Burchfiel et al. 1995). It was clearly a major structure during the Mesozoic, when it accommodated left-lateral shear between these two tectonostratigraphic provinces (Dirks et al. 1994). However, our results suggest that it does not accommodate significant variations in the vertical deformation field at present, consistent with the fact that lowtemperature thermochronometers yield similar ages on either side of the fault (Godard et al. 2009b). Of course, this does not rule out the

183

possibility of lateral slip; future work will need to focus on this aspect of hazard in the region. The fact that significant and ongoing differential rock uplift is apparently distributed for c. 75 – 100 km inboard of the frontal faults along the margin of the Sichuan Basin presents a challenge to any model of crustal thickening driven entirely by shortening in the upper crust (e.g. Hubbard & Shaw 2009). There is little evidence in the surface geology for faults that might accommodate high rates of rock uplift, particularly if sustained for c. 10 Ma (Godard et al. 2009b). Although one might be tempted to call upon crustal duplexing to drive localized crustal thickening, we do not see geometric variations in fabric patterns at the surface consistent with this. Rather, the architecture of the thrust belt is more consistent with an imbricate fan, with thrusts rooted in a deep de´collement (Burchfiel et al. 1995; Hubbard & Shaw 2009; Jia et al. 2009). Given that slip along the Yingxiu– Beichuan fault appears to have accommodated much of the differential exhumation during Cenozoic development of the plateau (Kirby et al. 2002; Godard et al. 2009b), we find it likely that high rates of rock uplift well inboard of the plateau margin must be sustained by mass influx in the lower or middle crust (e.g. Clark et al. 2005; Godard et al. 2009a).

Active faults along the eastern flank of the Min Shan North of the Sichuan Basin, along the flank of the Min Shan, we observe a more complicated distribution of active faults. Two of these systems, however, appear to lie near or along gradients in differential rock uplift (Fig. 9). The Huya fault probably plays a similar role as the Yingxiu– Beichuan fault farther south, as it sits at the foot of the Min Shan and appears to bound the eastern flanks of a narrow region of high rock uplift (Fig. 9). As noted previously, this fault experienced a series of earthquakes in 1976 (e.g. Jones et al. 1984). Although the mapped trace is rather short, only c. 60 km, several smaller faults of similar orientation exist south. These also lie along the uplift-rate gradient (Fig. 9), and we consider it likely that they are active. The degree to which these faults are linked with the Huya farther north, is an outstanding question. East of the Min Shan proper, the Qingchuan fault strikes NE and forms a prominent linear trace across the landscape. Deflected drainages suggest a rightlateral component of active displacement (e.g. Burchfiel et al. 1995), although we are not aware of any reliable estimates of slip-rate. Our analysis suggests that the western end of this fault system

184

E. KIRBY & W. OUIMET

is coincident with gradients in rock uplift (Fig. 9), which suggests the fault may express a vertical component of displacement as well.

Crustal thickening in the Qinling region The region of our study area north of the Min Shan shows the weakest association between active faults and the differential rock uplift rate field (Fig. 9). In this region, the structural grain trends east –west, parallel to the West Qinling orogen (e.g. Enkelmann et al. 2006; Ratschbacher et al. 2003), and the active faults (Tazang fault, Bailong Jiang fault; Fig. 2) are left-lateral strike –slip faults (Kirby et al. 2007a; Yuan et al. 2007) that appear to reactivate structures along this grain. Strikingly, the uplift-rate field inferred from channel profiles trends north–south, cutting a swath orthogonal to active faults, but parallel to the topographic margin of the plateau. Although the exact rates of differential rock uplift within this region are somewhat uncertain (Fig. 9), the orientation, location, and kinematics of the Bailong Jiang and Tazang faults are not able to explain the pattern of differential rock uplift nor the fact that the topographic margin of the plateau trends north–south. Thus, the fact that we do not observe any upper crustal structures that may be responsible for active rock uplift strongly implicates processes in the deep crust in driving ongoing deformation. Overall, our analysis reveals a spatial association between active differential rock uplift and upper crustal structures along the Sichuan Basin. In this region of the Longmen Shan, reactivated Mesozoic fault systems (e.g. Jia et al. 2009) appear to accommodate most of the ongoing development of high topography at the plateau margin (Hubbard & Shaw 2009). However, in regions distant from the Sichuan Basin, within the West Qinling and within the plateau west of the Longmen Shan, ongoing differential rock uplift appears to persist in the absence of upper crustal faults. In these regions, one must appeal to deformation and thickening in the deep crust (and perhaps the mantle) to explain the coincidence of high topography, deep exhumation, and sustained rock uplift.

Outstanding questions Our analysis of the erosional and deformation patterns during plateau development raises several questions and suggests directions for future research. We identify three major questions. First, to what degree is the active differential rock uplift field sustained by erosional exhumation (e.g. Godard et al. 2009a), and to what degree is it driven by thickening of lower crust (e.g. Clark

et al. 2005)? Although at one level the difference between these models comes down to semantics, in that rock uplift is always ‘driven’ by exhumation (e.g. Molnar 2009), the degree to which flow of lower crust plays a dynamic role in sustaining topography (e.g. Clark et al. 2005) remains a firstorder question along all margins of the plateau (e.g. Beaumont et al. 2001). The evolution of the Sichuan margin poses the additional challenge that exhumation has not yet exhumed rocks that would record conditions of lower crustal flow (although this may have been accomplished in the Gongga Shan massif, at the SW corner of the basin; Roger et al. 1995). Definitive tests of these hypotheses are probably still just beyond our reach at the present time, but new estimates of crustal rheology derived from analysis of post-seismic velocities (e.g. R. Burgmann, pers. comm., 2009), coupled with densified networks of continuous GPS may allow this to be addressed in the coming years. Second, how has exhumation rate varied over the development of the plateau margin? This question is central to ascertain to what degree erosion drives ongoing convergence (Godard et al. 2009a). At present thermochronological data are sparsely sampled and exhibit a fair bit of scatter (Kirby et al. 2002; Godard et al. 2009b). In principle, improvements in analytical techniques and interpretation of low-temperature thermochronometers (e.g. Ketcham et al. 2009) make this problem tractable with additional data. Finally, what is the wavelength of ongoing rock uplift and how does it vary with distance away from the Sichuan Basin? Testing this will likely require measuring differential rock uplift at multiple timescales. Fluvial terraces hold promise to map out patterns of deformation along the Min Shan (e.g. Lave´ & Avouac 2000), and short-term measurement of vertical deformation from leveling lines would provide data crucial to confirming patterns inferred from our geomorphic analysis. We note that, in principle, knowing the wavelength of incision and rock uplift over shorter timescales could provide insight into the proper form of the functional relationship between channel steepness indices and incision rate. Our research in eastern Tibet has been supported by the Continental Dynamics and Tectonics programs at the National Science Foundations. Our thinking has benefited from discussions with numerous individuals, notably Kelin Whipple, Marin Clark, Peter Molnar, and Doug Burbank. Reviews by Mikael Attal and an anonymous reviewer led to significant improvements in the manuscript. We thank Richard Gloaguen and Lothar Ratschbacher for their patience and editorial guidance. Kirby acknowledges support as a Scholar-in-Residence at the Earth and Environmental Systems Institute (Penn State) during the writing of this manuscript.

TECTONIC GEOMORPHOLOGY IN EASTERN TIBET

References Amos, C. B. & Burbank, D. W. 2007. Channel width response to differential uplift. Journal of Geophysical Research-Earth Surface, 112, F02010; doi: 10.1029/ 2006JF000672. Arne, D., Worley, B. et al. 1997. Differential exhumation in response to episodic thrusting along the eastern margin of the Tibetan Plateau. Tectonophysics, 280, 239–256. Bai, D., Unsworth, M. J. et al. 2010. Crustal deformation of the eastern Tibetan plateau revealed by magnetotelluric imaging. Nature Geoscience, 3, 358– 362. Beaumont, C., Jamieson, R. A., Nguyen, M. H. & Lee, B. 2001. Himalayan tectonics explained by extrusion of a low-viscosity crustal channel coupled to focused surface denudation. Nature, 414, 738–742. Bierman, P. & Steig, E. J. 1996. Estimating rates of denudation using cosmogenic isotope abundances in sediment. Earth Surface Processes and Landforms, 21, 125–139. Burchfiel, B. C., Chen, Z., Liu, Y. & Royden, L. H. 1995. Tectonics of the Longmen Shan and adjacent regoins, central China. International Geological Review, 37, 661– 735. Burchfiel, B. C., Royden, L. H. et al. 2008. A geological and geophysical context for the Wenchuan earthquake of 12 May 2008, Sichuan, People’s Republic of China. GSA Today, 18, 4– 11. Chen, S. F. & Wilson, C. J. L. 1996. Emplacement of the Longmen Shan Thrust-Nappe belt along the eastern margin of the Tibetan plateau. Journal of Structural Geology, 18, 413– 430. Chen, S. F., Wilson, C. J. L., Luo, Z. L. & Deng, Q. D. 1994a. The evolution of the Western Sichuan Foreland Basin, southwestern China. Journal of Southeast Asian Earth Sciences, 10, 159 –168. Chen, S. F., Wilson, C. J. L., Deng, Q. D., Zhao, X. L. & Luo, Z. L. 1994b. Active faulting and block movement associated with large earthquakes in the Min Shan and Longmen Mountains, northeastern Tibetan Plateau. Journal of Geophysical Research, 99, 24,025– 24,038. Clark, M. K. & Royden, L. H. 2000. Topographic ooze: Building the eastern margin of Tibet by lower crustal flow. Geology, 28, 703– 706. Clark, M. K., House, M. A., Royden, L. H., Whipple, K. X., Burchfiel, B. C., Zhang, X. & Tang, W. 2005. Late Cenozoic uplift of southeastern Tibet. Geology, 33, 525–528. Clarke, B. A. & Burbank, D. W. 2010. Bedrock fracturing, threshold hillslopes, and limits to the magnitude of bedrock landslides. Earth and Planetary Science Letters, 297, 577–586. Crosby, B. T. & Whipple, K. X. 2006. Knickpoint initiation and distribution within fluvial networks: 236 waterfalls in the Waipaoa River, North Island, New Zealand. Geomorphology, 82, 16 –38. Densmore, A. L., Ellis, M. A., Li, Y., Zhou, R. J., Hancock, G. S. & Richardson, N. 2007. Active tectonics of the Beichuan and Pengguan faults at the eastern margin of the Tibetan Plateau. Tectonics, 26, TC4005; doi: 10.1029/2006TC001987. DiBiase, R. A., Whipple, K. X., Heimsath, A. M. & Ouimet, W. B. 2010. Landscape form and millennial

185

erosion rates in the San Gabriel Mountains, CA. Earth and Planetary Science Letters, 289, 134–144. Dirks, P. H. G. M., Wilson, C. J. L., Chen, S., Luo, Z. L. & Liu, S. 1994. Tectonic evolution of the NE margin of the Tibetan Plateau; evidence from the central Longmen Mountains, Sichuan Province, China. Journal of Southeast Asian Earth Sciences, 9, 181– 192. Duvall, A., Kirby, E. & Burbank, D. 2004. Tectonic and lithologic controls on bedrock channel profiles and processes in coastal California. Journal of Geophysical Research-Earth Surface, 109, F03002; doi: 10.1029/2003JF000086. Enkelmann, E., Ratschbacher, L. et al. 2006. Cenozoic exhumation and deformation of northeastern Tibet and the Qinling: Is Tibetan lower crustal flow diverging around the Sichuan Basin? Geological Society of America Bulletin, 118, 651–671. Finnegan, N. J., Roe, G., Montgomery, D. R. & Hallet, B. 2005. Controls on the channel width of rivers; implications for modeling fluvial incision of bedrock. Geology, 33, 229–232. Flint, J. J. 1974. Stream gradient as a function of order, magnitude, and discharge. Water Resources Research, 10, 969 –973. Godard, V., Cattin, R. & Lave, J. 2009a. Erosional control on the dynamics of low-convergence rate continental plateau margins. Geophysical Journal International, 179, 763–777. Godard, V., Pik, R. et al. 2009b. Late Cenozoic evolution of the central Longmen Shan, eastern Tibet: Insight from (U –Th)/He thermochronometry. Tectonics, 28, TC5009; doi: 10.1029/2008TC002407. Godard, V., Lave, J., Carcaillet, J., Cattin, R., Bourle`s, D. & Zhu, J. 2010. Spatial distribution of denudation in Eastern Tibet and regressive erosion of plateau margins. Tectonophysics, 491, 253– 274. Gosse, J. C. & Phillips, F. M. 2001. Terrestrial in-situ cosmogenic nuclides: theory and application. Quaternary Science Reviews, 20, 1475– 1560. Granger, D. E., Kirchner, J. W. & Finkel, R. C. 1996. Spatially averaged long-term erosion rates measured from in situ-produced cosmogenic nuclides in alluvial sediment. Journal of Geology, 104, 249–257. Hack, J. T. 1973. Stream profile analysis and streamgradient index. Journal of Research of the U.S. Geological Survey, 1, 421–429. Harkins, N., Kirby, E., Heimsath, A. M., Robinson, R. & Reiser, U. 2007. Transient fluvial incision in the headwaters of the Yellow River, northeastern Tibet, China. Journal of Geophysical Research, 112, F03S04; doi: 10.1029/2006JF000570. Hartshorn, K., Hovius, N., Dade, W. B. & Slingerland, R. 2002. Climate-driven bedrock incision in an active mountain belt. Science, 297, 2036–2038. Hodges, K. V., Wobus, C., Ruhl, K., Schildgen, T. & Whipple, K. X. 2004. Quaternary deformation, river steepening, and heavy precipitation at the front of the higher Himalayan ranges. Earth and Planetary Science Letters, 220, 379– 389. Howard, A. D. 1980. Thresholds in river regimes. In: Coates, D. R. & Vitek, J. D. (eds) Thresholds in Geomorphology. Allen and Unwin, Boston, 227–258.

186

E. KIRBY & W. OUIMET

Howard, A. D. & Kerby, G. 1983. Channel changes in badlands. Geological Society of America Bulletin, 94, 739– 752. Howard, A. D., Seidl, M. A. & Dietrich, W. E. 1994. Modeling fluvial erosion on regional to continental scales. Journal of Geophysical Research, 99, 13 971– 13 986. Hubbard, J. & Shaw, J. H. 2009. Uplift of the Longmen Shan and Tibetan plateau, and the 2008 Wenchuan (M ¼ 7.9) earthquake. Nature, 458, 194– 197. Ji, C. & Hayes, G. 2008. Finite fault model – preliminary result of the May 12, 2008 Mw 7.9 Eastern Sichuan, China Earthquake. U.S. Geological Survey National Earthquake Center. Ji, C., Shao, G., Lu, Z., Hudnut, K., Jiu, J., Hayes, G. & Zeng, Y. 2008. Rupture history of 2008 May 12 Mw 8.0 Wen-Chuan earthquake: Evidence of slip interaction. Eos Transactions, American Geophysical Union, 89, Fall Meet. Suppl., Abstract #S23E-02. Jia, D., Li, Y. et al. 2009. Structural model of 2008 Mw 7.9 Wenchuan earthquake in the rejuvenated Longmen Shan thrust belt, China. Tectonophysics, 491, 174– 184, doi: 10.1016/j.tecto.2009.08.040. Johnson, J. P. L., Whipple, K. X., Sklar, L. S. & Hanks, T. C. 2009. Transport slopes, sediment cover, and bedrock channel incision in the Henry Mountains, Utah. Journal of Geophysical Research, 114, F02014, doi: 10.1029/2007JF000862. Jones, L. M., Han, W., Hauksson, E., Jin, A., Zhang, Y. & Luo, Z. 1984. Focal mechanisms of the Songpan earthquakes of August 1976 in Sichuan, China. Journal of Geophysical Research, 89, 7697–7707. Ketcham, R. A., Donelick, R. A., Balestrieri, M. L. & Zattin, M. 2009. Reproducibility of apatite fission-track length data and thermal history reconstruction. Earth and Planetary Science Letters, 284, 504– 515. King, R. W., Shen, F. et al. 1997. Geodetic measurement of crustal motion in southwest China. Geology, 25, 179– 182. Kirby, E. & Whipple, K. 2001. Quantifying differential rock-uplift rates via stream profile analysis. Geology, 29, 415–418. Kirby, E., Whipple, K. X., Burchfiel, B. C., Tang, W. Q., Berger, G., Sun, Z. M. & Chen, Z. L. 2000. Neotectonics of the Min Shan, China: Implications for mechanisms driving Quaternary deformation along the eastern margin of the Tibetan Plateau. Geological Society of America Bulletin, 112, 375– 393. Kirby, E., Reiners, P. W. et al. 2002. Late Cenozoic evolution of the eastern margin of the Tibetan Plateau; inferences from (super 40) Ar/ (super 39) Ar and (U–Th)/He thermochronology. Tectonics, 21, 1001; doi: 10.1029/2000TC001246. Kirby, E., Whipple, K. X., Tang, W. Q. & Chen, Z. L. 2003. Distribution of active rock uplift along the eastern margin of the Tibetan Plateau: Inferences from bedrock channel longitudinal profiles. Journal of Geophysical Research-Solid Earth, 108, 2217; doi: 10.1029/2001JB000861. Kirby, E., Harkins, N., Wang, E., Shi, X., Fan, C. & Burbank, D. 2007a. Slip rate gradients along the eastern Kunlun Fault. Tectonics, 26, TC2010; doi: 10.1029/2006TC002033.

Kirby, E., Johnson, C., Furlong, K. & Heimsath, A. 2007b. Transient channel incision along Bolinas Ridge, California: Evidence for differential rock uplift adjacent to the San Andreas Fault. Journal of Geophysical Research, 112, F03S07, doi: 10.1029/ 2006JF000559. Kirby, E., Whipple, K. & Harkins, N. 2008. Topography reveals seismic hazard. Nature Geoscience, 1, 485–487. Lague, D. & Davy, P. 2003. Constraints on the long-term colluvial erosion law by analyzing slope–area relationships at various tectonic uplift rates in the Siwalik Hills (Nepal). Journal of Geophysical Research, 108, 2129, doi: 10.1029/2002JB001893. Lave´, J. & Avouac, J. P. 2000. Active folding of fluvial terraces across the Siwaliks Hills, Himalayas of central Nepal. Journal of Geophysical Research, 105, 5735– 5770. Lave´, J. & Avouac, J. P. 2001. Fluvial incision and tectonic uplift across the Himalayas of central Nepal. Journal of Geophysical Research, 106, 26 561– 26 591. Leopold, L. B. & Maddock, T., jr 1953. The hydraulic geometry of stream channels and some physiographic implications. USGS Professional Paper, 252. Leopold, L. B., Wolman, M. G. & Miller, J. P. 1964. Fluvial Processes in Geomorphology. W.H. Freeman and Co., San Francisco. Li, C.-Y., Wei, Z.-Y., Ye, J.-Q., Han, Y.-B. & Zheng, W.-J. 2009. Amounts and styles of coseismic deformation along the northern segment of surface rupture of the 2008 Wenchuan Mw 7.9 earthquake, China. Tectonophysics, 491, 35–58; doi: 10.1016/ j.tecto.2009.09.023. Lin, A., Ren, Z. & Jia, D. 2009. Co-seismic groundshortening structures produced by the 2008 Mw 7.9 Wenchuan earthquake, China. Tectonophysics, 491, 21–34; doi: 10.1016/j.tecto.2009.10.027. Moglen, G. E. & Bras, R. L. 1995. The effect of spatial heterogeneities on geomorphic expression in a model of basin evolution. Water Resources Research, 31, 2613– 2623. Molnar, P. 1987. Inversion of profiles of uplift rates for the geometry of dip–slip faults at depth, with examples from the Alps and Himalaya. Annales Geophysicae, 5B, 663–670. Molnar, P. 2009. The state of interactions among tectonics, erosion, and climate: A polemic. GSA Today, 19, 44–45; doi: 10.1130/GSATG00GW.1. Molnar, P. & England, P. 1990. Late Cenozoic uplift of mountain ranges and global climate change: chicken or egg? Nature, 346, 29– 34. Molnar, P., Anderson, R. S. & Anderson, S. P. 2007. Tectonics, fracturing of rock, and erosion. Journal of Geophysical Research, 112, F03014, doi: 10.1029/2005JF000433. Montgomery, D. R. & Gran, K. B. 2001. Downstream variations in the width of bedrock channels. Water Resources Research, 37, 1841– 1846. New, M., Lister, D., Hulme, M. & Makin, I. 2002. A high-resolution data set of surface climate over global land areas. Climate Research, 21, 1– 25. Niemi, N. A., Oskin, M., Burbank, D. W., Heimsath, A. M. & Gabet, E. J. 2005. Effects of bedrock

TECTONIC GEOMORPHOLOGY IN EASTERN TIBET landslides on cosmogenically determined erosion rates. Earth and Planetary Science Letters, 237, 480–498. Ouimet, W. B. 2009. Landslides associated with the May 12, 2008 Wenchuan earthquake: Implications for the erosion and tectonic evolution of the Longmen Shan. Tectonophysics, 491, 244– 252; doi: 10.1016/ j.tecto.2009.09.012. Ouimet, W. B., Whipple, K. X. & Granger, D. E. 2009. Beyond threshold hillslopes: Channel adjustment to base-level fall in tectonically active mountain ranges. Geology, 37, 579– 582. Ouimet, W., Whipple, K., Royden, L., Reiners, P., Hodges, K. & Pringle, M. 2010. Regional incision of the eastern margin of the Tibetan Plateau. Lithosphere, 2, 50– 63. Parker, G. 1978a. Self-formed straight rivers with equilibrium banks and mobile bed. Part 1. The sand-silt river.Journal of Fluid Mechanics Digital Archive, 89, 109–125. Parker, G. 1978b. Self-formed straight rivers with equilibrium banks and mobile bed. Part 2. The gravel river. Journal of Fluid Mechanics Digital Archive, 89, 127–146. Parker, G. 1979. Hydraulic geometry of active gravel rivers. Journal of Hydraulic Engineering, 105, 1185–1201. Ratschbacher, L., Hacker, B. R. et al. 2003. Tectonics of the Qinling (Central China): tectonostratigraphy, geochronology, and deformation history. Tectonophysics, 366, 1– 53. Richardson, N. J., Densmore, A. L. et al. 2008. Extraordinary denudation in the Sichuan Basin: Insights from low-temperature thermochronology adjacent to the eastern margin of the Tibetan Plateau. Journal of Geophysical Research-Solid Earth, 113, B04409; doi: 10.1029/2006JB004739. Richardson, N. J., Densmore, A. L., Seward, D., Wipf, M. & Yong, L. 2010. Did incision of the Three Gorges begin in the Eocene? Geology, 38, 551–554; doi: 10.1130/G30527.1. Robert, A., Zhu, J. et al. 2010. Crustal structures in the area of the 2008 Sichuan earthquake from seismologic and gravimetric data. Tectonophysics, 491, 205– 210; doi: 10.1016/j.tecto.2009.11.010. Roe, G. H., Montgomery, D. R. & Hallet, B. 2003. Orographic precipitation and the relief of mountain ranges. Journal of Geophysical Research, 108, 2315 (ETG 15:1–12). Roger, F., Calassou, S. et al. 1995. Miocene emplacement and deformation of the Konga Shan granite (Xianshui He fault zone, west Sichuan, China): Geodynamic implications. Earth and Planetary Science Letters, 130, 201– 216. Roger, F., Jolivet, M., Cattin, R. & Malavieille, J. 2011. Mesozoic–Cenozoic tectonothermal evolution of the eastern part of the Tibetan plateau (SongpanGarzeˆ, Longmen Shan area): insights from thermochronological data and simple thermal modeling. In: Gloaguen, R. & Ratschbacher, L. (eds) Growth and Collapse of the Tibetan Plateau. Geological Society, London, Special Publications, 353, 9– 25. Royden, L. H., Burchfiel, B. C., King, R. W., Chen, Z., Shen, F & Liu, Y. 1997. Surface deformation and

187

lower crustal flow in Eastern Tibet. Science, 276, 788– 790. Royden, L. H., Burchfiel, B. C. & van der Hilst, R. D. 2008. The geological evolution of the Tibetan plateau. Science, 321, 1054– 1058. Safran, E. B., Bierman, P. R., Aalto, R., Dunne, T., Whipple, K. X. & Caffee, M. W. 2005. Erosion rates driven by channel network incision in the Bolivian Andes. Earth Surface Processes and Landforms, 30, 1007–1024. Seeber, L. & Gornitz, V. 1983. River profiles along the Himalayan Arc as indicators of active tectonics. Tectonophysics, 92, 335 –367. Sengo¨r, A. M. C., Cin, A., Rowley, D. B. & Nie, S. Y. 1993. Space–time patterns of magmatism along the Tethysides. Journal of Geology, 101, 51–84. Shen, Z. K., Lu, J. N., Wang, M. & Burgmann, R. 2005. Contemporary crustal deformation around the southeast borderland of the Tibetan Plateau. Journal of Geophysical Research-Solid Earth, 110, B11409; doi: 10.1029/2004JB003421. Shen, Z. K., Sun, J. et al. 2009. Slip maxima at fault junctions and rupturing of barriers during the 2008 Wenchuan earthquake. Nature Geoscience, 2, 718– 724. Sklar, L. S. & Dietrich, W. E. 1998. River longitudinal profiles and bedrock incision models: Stream power and the influence of sediment supply. In: Tinkler, K. J. & Wohl, E. E. (eds) Rivers Over Rock: Fluvial Processes in Bedrock Channels: Geophysical Monograph 107. Washington, DC, American Geophysical Union, 237 –260. Sklar, L. S. & Dietrich, W. E. 2001. Sediment and rock strength controls on river incision into bedrock. Geology, 29, 1087– 1090. Sklar, L. S. & Dietrich, W. E. 2004. A mechanistic model for river incision into bedrock by saltating bed load. Water Resources Research, 40, W06301; doi: 10.1029/2003WR002496. Sklar, L. S. & Dietrich, W. E. 2006. The role of sediment in controlling steady-state bedrock channel slope: Implications of the saltation– abrasion incision model. Geomorphology, 82, 58–83. Sklar, L. S. & Dietrich, W. E. 2008. Implications of the saltation–abrasion bedrock incision model for steady-state river longitudinal profile relief and concavity. Earth Surface Processes and Landforms, 33, 1129– 1151. Snyder, N. P., Whipple, K. X., Tucker, G. E. & Merritts, D. J. 2000. Landscape response to tectonic forcing: Digital elevation model analysis of stream profiles in the Mendocino triple junction region, northern California. Geological Society of America Bulletin, 112, 1250– 1263. Snyder, N. P., Whipple, K. X., Tucker, G. E. & Merritts, D. J. 2003. Channel response to tectonic forcing: field analysis of stream morphology and hydrology in the Mendocino triple junction region, northern California. Geomorphology, 53, 97–127. Stark, C. P. 2006. A self-regulating model of bedrock river channel geometry. Geophysical Research Letters, 33, L04402; doi: 10.1029/2005GL023193. Stock, J. & Dietrich, W. E. 2003. Valley incision by debris flows: Evidence of a topographic signature.

188

E. KIRBY & W. OUIMET

Water Resources Research, 39, 1089; doi: 10.1029/ 2001WR001057. Stock, J. D. & Dietrich, W. E. 2006. Erosion of steepland valleys by debris flows. Geological Society of America Bulletin, 118, 1125– 1148. Stock, J. D. & Montgomery, D. R. 1999. Geologic constraints on bedrock river incision using the stream power law. Journal of Geophysical Research, 104, 4983–4993. Tucker, G. E. 2004. Drainage basin sensitivity to tectonic and climatic forcing: Implications of a stochastic model for the role of entrainment and erosion thresholds. Earth Surface Processes and Landforms, 29, 185–205. Tucker, G. E. & Bras, R. L. 2000. A stochastic approach to modeling the role of rainfall variability in drainage basin evolution. Water Resources Research, 36, 1953–1964. Turowski, J. M., Lague, D. & Hovius, N. 2007. Cover effect in bedrock abrasion: A new derivation and its implications for the modeling of bedrock channel morphology. Journal of Geophysical Research-Earth Surface, 112, F04006; doi: 10.1029/2006JF000697. Turowski, J. M., Hovius, N., Wilson, A. & Horng, M. J. 2008. Hydraulic geometry, river sediment and the definition of bedrock channels. Geomorphology, 99, 26–38. Van der Woerd, J., Tapponnier, P. et al. 2002. Uniform postglacial slip-rate along the central 600 km of the Kunlun Fault (Tibet), from 26Al, 10Be, and 14C dating of riser offsets, and climatic origin of the regional morphology. Geophysical Journal International, 148, 356– 388. Weislogel, A. L., Graham, S. A., Chang, E. Z., Wooden, J. L., Gehrels, G. E. & Yang, H. 2006. Detrital zircon provenance of the Late Triassic Songpan– Ganzi complex: sedimentary record of collision of the North and South China blocks. Geology, 34, 97–100. Whipple, K. X. 2004. Bedrock rivers and the geomorphology of active orogens. Annual Review of Earth and Planetary Sciences, 32, 151–185. Whipple, K. X. & Tucker, G. E. 1999. Dynamics of the stream-power river incision model: Implications for height limits of mountain ranges, landscape response timescales, and research needs. Journal of Geophysical Research-Solid Earth, 104, 17661– 17674. Whipple, K. X. & Tucker, G. E. 2002. Implications of sediment-flux-dependent river incision models for landscape evolution. Journal of Geophysical Research-Solid Earth, 107, 2039; doi: 10.1029/ 2000JB000044. Whipple, K. X., Hancock, G. S. & Anderson, R. S. 2000. River incision into bedrock: Mechanics and relative efficacy of plucking, abrasion, and cavitation. Geological Society of America Bulletin, 112, 490–503. Whittaker, A. C., Cowie, P. A., Attal, M., Tucker, G. E. & Roberts, G. P. 2007a. Bedrock channel adjustment to tectonic forcing: Implications for predicting river incision rates. Geology, 35, 103– 106. Whittaker, A. C., Cowie, P. A., Attal, M., Tucker, G. E. & Roberts, G. P. 2007b. Contrasting transient and steady-state rivers crossing active normal faults: new field observations from the Central Apennines, Italy. Basin Research, 19, 529 –556.

Whittaker, A. C., Attal, M., Cowie, P. A., Tucker, G. E. & Roberts, G. 2008. Decoding temporal and spatial patterns of fault uplift using transient river long profiles. Geomorphology, 100, 506– 526. Willgoose, G., Bras, R. L. & Rodriguez-Iturbe, I. 1991. A coupled channel network growth and hillslope evolution model. 1. Theory. Water Resources Research, 27, 1671–84. Wobus, C. W., Hodges, K. V. & Whipple, K. X. 2003. Has focused denudation sustained active thrusting at the Himalayan topographic front? Geology, 31, 861–864. Wobus, C., Whipple, K. X. et al. 2006a. Tectonics from topography; procedures, promise, and pitfalls. In: Willett, S., Hovius, N., Brandon, M. & Fisher, D. (eds) GSA Special Paper – Tectonics, Climate and Landscape Evolution. Geological Society of America (GSA), Boulder, CO, 398, 55–74. Wobus, C. W., Crosby, B. T. & Whipple, K. X. 2006b. Hanging valleys in fluvial systems: Controls on occurrence and implications for landscape evolution. Journal of Geophysical Research, 111, F02017; doi: 10.1029/2005JF000406. Wobus, C. W., Kean, J. W., Tucker, G. E. & Anderson, R. S. 2008. Modeling the evolution of channel shape: Balancing computational efficiency with hydraulic fidelity. Journal of Geophysical Research-Earth Surface, 113, F02004; doi: 10.1029/2007JF000914. Xu, G. & Kamp, P. 2000. Tectonics and denudation adjacent to the Xianshuihe Fault, eastern Tibetan Plateau: Constraints from fission track thermochronology. Journal of Geophysical Research, 105, 19 231–19 251. Xu, L. L., Rondenay, S. & van der Hilst, R. D. 2007. Structure of the crust beneath the southeastern Tibetan Plateau from teleseismic receiver functions. Physics of the Earth and Planetary Interiors, 165, 176–193. Xu, X. W., Wen, X. Z., Yu, G. H., Chen, G. H., Klinger, Y., Hubbard, J. & Shaw, J. 2009. Coseismic reverseand oblique-slip surface faulting generated by the 2008 Mw 7.9 Wenchuan earthquake, China. Geology, 37, 515–518. Yuan, D., Lei, Z., He, W., Xiong, Z., Ge, W., Liu, X. & Liu, B. 2007. Textual research of Wudu earthquake in 186 B.C. in Gansu Province, China and discussion on its causative structure. Acta Seismological Sinica, 20, 696–707. Zhang, P., Wen, X., Shen, Z. K. & Chen, J. 2010. Oblique, high-angle, listric-reverse faulting and associated development of strain: The Wenchuan earthquake of May 12, 2008, Sichuan, China. Annual Reviews of Earth and Planetary Sciences, 38, 351 –380. Zhang, Z., Wang, Y., Chen, Y., Houseman, G. A., Tian, X., Wang, E. & Tend, J. 2009. Crustal structure across Longmenshan fault belt from passive source seismic profiling. Geophysical Research Letters, 36, L17310; doi: 10.1029/2009GL039580. Zhou, D. & Graham, S. A. 1996. Songpan– Ganzi complex of the west Qilian Shan as a Triassic remnant ocean basin. In: Yin, A. & Harrison, T. M. (eds) The Tectonic Evolution of Asia. Cambridge University Press, Cambridge, 281–299.

Incision rate of the Yellow River in Northeastern Tibet constrained by 10Be and 26Al cosmogenic isotope dating of fluvial terraces: implications for catchment evolution and plateau building A. PERRINEAU1,2, J. VAN DER WOERD2, Y. GAUDEMER1, JING LIU-ZENG3, R. PIK4, P. TAPPONNIER1, R. THUIZAT2 & ZHENG RONGZHANG5 1

Laboratoire de Tectonique, Institut de Physique du Globe de Paris – UMR 7154, 4 place Jussieu, 75254 cedex 05, Paris, France 2 Institut de Physique du Globe de Strasbourg – UMR CNRS/UDS 7516, E´cole et Observatoire des Sciences de la Terre, University of Strasbourg, 5 rue Rene´ Descartes, 67084 Strasbourg cedex, France 3

Institute of Tibetan Plateau Research, Chinese Academy of Sciences, 18 Shuang Qing Rd, PO Box 2871, Beijing 100085, China 4

Centre de Recherche Pe´trographiques et Ge´ochimiques – UPR1167, 15 rue Notre-Dame des Pauvres, 54501 Vandoeuvre-le`s-Nancy, France 5

China Earthquake Administration, Beijing, China

*Corresponding author (e-mail: [email protected]) Abstract: Unlike other large rivers flowing out of Tibet, the Yellow River escapes from the plateau towards the NE crossing no less than five NW– SE striking, actively growing ranges and intervening basins. Thick Plio-Quaternary deposits and fluvial terraces testify to a phase of aggradation and sediment infill up to the average surface elevation (3200–3250 m a.s.l.) of the Gonghe, Guide and Qinghai Lake basins. A set of seven main terraces across the Gonghe Basin suggests progressive down-cutting of the Yellow River carving the 500 m deep Longyang gorge at the basin exit. 10Be and 26Al concentrations in quartz of surface and sub-surface samples of four terraces constrain the timing of incision by determining the burial age of the deposit and the exposure age of its surface. Modelling the depth dependence of the 10Be concentration and the 26 Al/10Be ratio allows us to constrain the onset of the ongoing phase of incision to 120–250 ka. These ages suggest long-term incision rates between 2– 6 mm a21. Together with the present morphology of the Yellow River terraces across the Gonghe basin and the Longyang gorge, our results imply rapid river catchment evolution and interaction between river dynamics, tectonic and climate in northeastern Tibet.

How and when the Tibetan Plateau reached its present altitude and shape are important to investigate as the topography influences atmospheric circulations, thus the climate, but also because it may be used to constrain models of Tibetan Plateau evolution (e.g. Tapponnier et al. 1982; England & Houseman 1986; Peltzer & Tapponnier 1988; Houseman & England 1993; Royden et al. 1997; Fluteau et al. 1999; Tapponnier et al. 2001; Liu-Zeng et al. 2008). In the stepped growth model of Tibet of Tapponnier et al. (2001), northeastern Tibet is the youngest part of the plateau that formed from Pliocene time to present (Meyer et al. 1998; Me´tivier et al. 1998). The region is characterized by narrow actively growing ranges (e.g. Tapponnier et al. 1990; Meyer et al. 1998;

Van der Woerd et al. 2001) that separate flat rapidly filling closed sedimentary basins (Meyer et al. 1998; Me´tivier et al. 1998; Van der Woerd 1998). Such process may have been active in the central part of the Tibetan Plateau, leading to the formation of a high topography at c. 5000 m a.s.l. From this point of view, the northeastern Tibetan region is a key place to unravel processes of the formation of the Tibetan Plateau. Understanding the different roles of tectonics, climate and the evolving drainage is thus important to constrain the formation of the present topography of Tibet (e.g. Clark et al. 2004; Liu-Zeng et al. 2008). The Yellow River, together with the Jinsha, Lancang (Mekong), and Nu (Salween) rivers, drains the Tibetan Plateau on its eastern rim (Fig. 1). While

From: Gloaguen, R. & Ratschbacher, L. (eds) Growth and Collapse of the Tibetan Plateau. Geological Society, London, Special Publications, 353, 189–219. DOI: 10.1144/SP353.10 0305-8719/11/$15.00 # The Geological Society of London 2011.

190 A. PERRINEAU ET AL. Fig. 1. Seismotectonic map of the upper reaches of the Yellow River across northeastern Tibet. Fault modified from Van der Woerd (1998), Meyer et al. (1998), Tapponnier et al. (2001). Seismicity from Iris catalogue, focal mechanisms from Harvard. Gonghe 1990 M 7 earthquake isoseismal from Chen et al. (1996). Topographical background is contours extracted from SRTM DEM with a 30 s precision.

INCISION RATE OF YELLOW RIVER IN NE TIBET

its source is located on the high summer monsoon influenced central plateau, like the other large rivers, it does not escape the plateau to the SE but is captured to the north after a hairpin loop in the Zoige (Roergai) basin (Fig. 1; e.g. Harkins et al. 2007). As it flows back to the NW, north of the Anyemaqen Shan, it enters regions of lesser precipitation that barely feed the river. This particularity makes the river sensitive to any north–south shift of the northern monsoon limit, which depends on climate variations (e.g. Benn & Owen 1998; Joussaume 1999). Both the fact that the river crosses a relatively dry northeastern Tibet allowing long-term preservation of the terraces and its high sensitivity to climate change probably in part explains the long sequence of fluvial terraces found all along the river from the Anyemaqen to the Ordos, in northern China. The terraces of the Yellow River, near Gonghe (Fig. 2) or near Lanzhou (e.g. Li 1991; Wang et al. 2010), have been commonly interpreted both to testify the first occurrence of the Yellow River in NE Tibet and as a record of regional plateau surface uplift (e.g. Li 1991; Li et al. 1997; Harkins et al. 2007). In this study, we will show that the terraces instead more likely indicate rapid incision of the river after a phase of damming and deposition of the Yellow River. In fact, probably better than anywhere else in Tibet, the terraces of the Yellow River across the Gonghe basin illustrate a process of plateau building that involves damming of the drainage by active range growth and rapid filling of closed sedimentary basins (e.g. Meyer et al. 1998; Me´tivier et al. 1998). In this paper, we describe the terraces of the Yellow River in the Gonghe basin. We present the 10Be and 26Al exposure ages obtained on four different terrace levels that allow the constraint of both the incision history of the Yellow River across the Gonghe basin and the uplift rate of the surrounding ranges. We then discuss the implications on the basin history, and on the local and regional tectonics.

Geological and geomorphological setting The Yellow River, basins and ranges From the Zoige basin, east of Anyemaqen and south of the Kunlun fault to the Gobi platform, over a length of 500 km (or river length of .1000 km), the elevation drop of the Yellow River reaches c. 2500 m as it crosses several sedimentary basins (Tongde, Gonghe, Guide, Xunhua, Linxia), all separated by actively growing ranges (Heka Shan, Waligong Shan, Kongma Shan, Laji Shan) (Figs 1– 3). Along this stretch, the river crosses no less than 3 active strike–slip fault systems and

191

c. six active thrusts implying interactions of river dynamics and active tectonics. Across both the Kunlun and Haiyuan left-lateral strike–slip faults (Fig. 1), the Yellow River is deflected 80– 90 km left-laterally (e.g. Gaudemer et al. 1989, 1995; Van der Woerd et al. 2002). Right-lateral movement along the Haiyen fault that connects with the more easterly-trending Riyue-Laji-Kongma Shan thrusts contribute to the range growth along the northeastern border of the Guide basin. The Tongde basin is bounded to the north by the south-vergent Heka Shan (Fig. 1; Harkins et al. 2007). The Gonghe basin is limited to the north and south by active thrusts. In the south, thrusts are north-vergent in the west and south-vergent in the east, and their activity is attested by the magnitude Mw 6.9 1990 seismic event (Chen et al. 1996). In the north, the Qinghainan Shan thrusts are south-vergent to the west and are attested by folded and uplifted alluvial fans and terraces (Van der Woerd 1998) and northvergent to the east (Waligong Shan). Intensely faulted and folded Neogene series in front of the Ryue-Laji Shan (Fang et al. 2005) together with detrital zircon ages and their provenance from the Laji Shan (Lease et al. 2007) imply major shortening phases during the Plio-Quaternary along the northern Guide basin. Additional but less important thrusting also occurs along the western Linxia basin (Fang et al. 2003; Zheng et al. 2003), across the Lanzhou thrusts (Wang et al. 2010) and across the Tianjing-Mibo Shan north of the Haiyuan fault (Gaudemer et al. 1995). Most of the elevation change of the Yellow River occurs between Zoige and Linxia and the steepest river gradients are found in the gorges across the Waligong and Kongma Shan (Fig. 3). At the northeastern exit of the Gonghe basin, the Yellow River crosses the Waligong mountains carving the 500 m deep Longyang Gorge (Longyangxia in Chinese) (Figs 2 & 3). Since 1986, an artificial dam has been built at the entrance of the Longyang gorges forming the c. 100 m deep Longyangxia Reservoir (water level at 2568 m in October 2004). The NW–SE-trending Waligong Shan is composed of two massifs, the western and eastern Waligong, each one bounded to the east by active west- to SW-dipping thrusts (Fig. 2). The western Waligong thrust juxtaposes the Waligong granite on top of red sandstones of Late Miocene – Pliocene age (Qinghai Geological Bureau 1988; Pan 1994; Fig. 4e), similarly to the thrust of the eastern Waligong at the Guide basin western margin. The most striking evidence for ongoing uplift is the impressive Longyang gorges carved by the Yellow River as it crosses the Waligong range (Fig. 4a, b). The slope of the Yellow River increases dramatically at the entrance of the Longyang gorges to c. 0.58 (Fig. 3). The height of the bedrock cliff

192

A. PERRINEAU ET AL.

Fig. 2. (a) SPOT satellite image mosaic of eastern Gonghe basin. White rectangle is Figure 2b. (b) CORONA image showing the northern part of the basin aross Yellow River terrace flights, Waligong Shan and Longyang gorge.

reaches c. 300 –500 m in the gorge (Fig. 4). Other evidence for active tectonic uplift are given by the abraded bedrock abandoned above the gorge. At present, the bedrock slopes towards the west as expected if it were back-tilted to the west due to uplift of the Waligong on top of a west-dipping thrust (Fig. 5).

The Yellow River terraces across the Gonghe Basin The Gonghe Basin is a NW– SE-trending 250  50 km wide sedimentary basin (Figs 1 & 2). It is bounded to the north by the actively growing Qinghainan Shan and to the south by the

INCISION RATE OF YELLOW RIVER IN NE TIBET

193

Fig. 2. Continued. (c) Digital elevation model (SRTM, 90 m, GeomapApp) of most parts of the Gonghe, Guide and Qinghai Lake basins. Yellow River terraces clearly show up in central part of figure as wide fan shape stairs. Inset is elevation data histogram of same area. The tight peak of elevation around 3200 m a.s.l. is the wide plateau surface corresponding to surface elevation of the Gonghe and Qinghai Lake basins, as well as part of the Guide and Tongde basins.

Ngola Shan (or Gonghenan Shan) and the Heka Shan. It is filled with up to 1200 m of Quaternary sediments (Qinghai Geological Bureau 1988; Me´tivier et al. 1998). Semi-arid climatic conditions prevail over most of the basin (annual precipitation average 250 –350 mm; Domroes & Peng 1988) and sand dune fields occupy the eastern part of it (Figs 2 & 4c). The basin is almost entirely internally drained except in its eastern part where it is crossed by the Yellow River. In contrast with the Qinghai Lake basin located just north of the Qinghainan Shan (Fig. 1) and occupied by the largest freshwater lake of China (4500 sq. m and about 27 m deep; Qin 1992), only small lakes are visible across the Gonghe basin. Evidence for higher lake levels can be seen in abandoned lake shore lines or lake deposits (e.g. around Dalian lake or Chaka salt lake; Fig. 6; Van der Woerd 1998). Despite their proximity, both basins have slightly different present climatic regimes characterized by dryer conditions in Gonghe (Domroes & Peng 1988). As it crosses the Gonghe basin and incises through the sediment pile (Fig. 4d), the Yellow

River has abandoned a set of terraces that form huge fan shaped stairs across the entire 50 km width of the basin (Fig. 2). The freshness of the terraced landscape (Fig. 4c) together with the fact that regressive erosion from the Yellow River gorge itself has not yet captured the central drainage of the Gonghe basin indicate that the Yellow River incision occurred recently as suggested by several authors (e.g. Li 1991; Me´tivier et al. 1998; Van der Woerd 1998; Zhang et al. 2004). The large, up to 10 km wide and 40–50 km long, terraces across the basin are mostly flat with slopes toward the north comparable to the present Yellow River slope (0.15–0.38) in the same area (Fig. 3a, c). From the present river bed to the top we numbered the main and largest terrace levels T0 –T7 (Figs 2 & 3). We designate the highest and largest terrace T7 knowing that it also corresponds to the regional surface level of basins around 3200–3250 m a.s.l. (Figs 2 & 3). This surface can be easily recognized on satellite images or on digital elevation models (SRTM DEM) as it forms flat areas that merge smoothly with the piedmont bajadas of the surrounding ranges (Figs 2 & 7a). A similar surface

194

A. PERRINEAU ET AL.

100 (d) 37

101

QIN

102 37

GHA I LA BAS KE IN

Huang Shui

Xining

Qinghai Nan

Shan

R iy W ali

ue

Gonghe go ng

GU

T5 T'6 T6

T4

T'3

T1

ID

G

T2 T3

Guide

ow Ye ll

GD E

gm

a

ive

r

36

Sh

an

SI N

N

BA

Ye llow R

SI

HE

Heka TON

Ko n

BA

NG

an

E

O

Riv er

36

T7

Sh

Shan B A SI N

35 100

35 102

101

Gonghe basin (e)

W

3250 m (x20) 3000

Yellow River T7 618 m

T'6

T6

524 m

505 m

T5 466 m

T4 447 m

T'3

T3

391 m 408 m

2750

T2 274 m

T1 173 m

Lake level

T0

2500

2250

E T7

35 m

10 km

Fig. 2. Continued. (d) Geomorphological map of the Yellow River terraces in the eastern Gonghe basin. White filled circles are sampling site on terraces T7 and T5. (e) Topographic cross-section of terraces following black line of Figure 2d (projection N858E).

INCISION RATE OF YELLOW RIVER IN NE TIBET

195 Slope (°)

(a) 0,6

0,6

0,4

0,4

0,2

0,2

0,0

0,0

(b)

North Anyemaqen South Heka Anyemaqen Shan Kunlun Fault

4000

Zoige

Jingun

+

9 m 19 ka C14 10m 18 ka C14

Tongde

65 m 30 ka C14240m 90/110 ka OSL

T7

Altitude 4000 (m) (x100)

Gonghe Guide

T7

T7

Xunhua

140 m 140±10ka OSL

60m T2 40 ka C14/37 ka OSL 6m T1 9 ka C14

3000

Waligong Kongma

3000

2000

2000 900

800

700

600

500

400

300

200

100

0

50

100 150

Distance along Yellow river (km) 4000

(c)

South Anyemaqen Tongde

Waligong

Heka Shan

T7 140 m 140±10ka OSL

Gonghe

Kongma Guide

Altitude (m) Xunhua (x100)

T7

T7

6

6' 6 54 3' 3

WLB

3000

5

2 1

WRB 4

0 3 2

2000

100

0

50

100

150 km

Fig. 3. (a) Slope variations along the Yellow River. River slope reaches its maximum values (.0.58) across the Longyang and Kongma gorges. (b) Profile of Yellow River bed from the southern Anyemaqen to Xunhua Basin. Red lines are active faults at basin margins (strike–slip, normal or reverse). Vertical scale is exaggerated 100 times. Profiles were extracted from SRTM DEM. River profile across Longyangxia Reservoir from Chinese 1:50 000 topographic maps. Position and terrace ages upstream from Gonghe from Harkins et al. (2007). (c) Enlarged Yellow River profile between Tongde and Xunhua, with projection of the Gonghe terraces (relative heights from Fig. 2), upper T7 regional top fill, and samples locations. Height of terraces in Guide basin from Fang et al. (2005). Colours of terrace threads in different basins are arbitrary.

level correlates with the top of the sediment fill in the Guide and Tongde basin (Fothergill & Ma 1999; Fang et al. 2005; Harkins et al. 2007). Interestingly, this level is also the highest palaeo-lake stand of the Qinghai lake basin to the north

(c. 3250 m; e.g. Lister et al. 1991; Fang et al. 2005; Rhode et al. 2007, 2010; Liu et al. 2010). T7 is clearly recognized as a fluvial terrace deposit at the entrance of the Gonghe basin north of Heka Shan (Figs 2 & 7b). T7 is now covered by

196

A. PERRINEAU ET AL.

Fig. 4. (a) Field photo of the Longyang gorges, view towards the east. (b) Steep gorge cliff on right bank of river incising flat abraded bedrock surface, view towards the south. (c) Terrace T7/T6 riser inside Gonghe basin (see Fig. 2), view towards the north. Riser forms a wind shelter where dunes accumulate (barkanes). (d) Yellow River gorge in the south of the Gonghe basin, view towards north. River can be seen cutting trough the sediment fill and northern bedrock sliver of Heka Shan (e.g. Me´tivier et al. 1998). Flat incised surface is level T7. (e) View to west at the steep thrust contact between the Mesozoic granodiorite of the western Waligong and the Tertiary red sandstones in the Yellow River gorge.

INCISION RATE OF YELLOW RIVER IN NE TIBET

197

Fig. 5. (a) Panoramic view of Waligong gorge. (b) View to west, towards Longyangxia Reservoir. In front, rim of bedrock strath can be seen. Note westward sloping flat top surface above the gorge. (c) View to east of stepped slope of western Waligong near the gorge that corresponds to the progressive entrenching and narrowing of the Yellow River bed. (d) View to west of the terraces in southwestern flank of Waligong Shan.

198

A. PERRINEAU ET AL.

Fig. 6. Google Earth view of the Gonghe and Qinghai basins with present lakes or lake remnants (salt flats, abandoned shore lines) highlighted.

a 2 m thick layer of aeolian loess overlying a 15 cm thick reddish palaeosol (B-horizon; Fig. 8a, b; Pan 1994). The abandonment of T7 marks the onset of incision of the Yellow River in the Gonghe basin. The other major terrace levels (T6 –T1) are also composed of fluvial pebbles and gravels and are devoid or only covered by a thin loess cover across the Gonghe basin. Clear steep risers up to several tens of metres high separate the main terrace levels (Figs 2d & 4c). In several places, as can be clearly seen on satellite images (Fig. 2), these risers disappear under wind-blown sand dune fields because of favourable accumulation at the north–south-trending risers which are almost perpendicular to the main NW blowing winds. For a long time the Yellow River terraces have been recognized as geomorphic markers of the active tectonics of northeastern Tibet during the Quaternary (e.g. Li 1991; Li et al. 1997; Lu et al. 2004; Sun 2005). From Lanzhou to Gonghe, their Early Pleistocene –Late Pleistocene ages (e.g. Li et al. 1997) have been interpreted to represent phases of regional plateau uplift. In contrast, Harkins et al. (2007) have recently proposed, from a geomorphological analysis of the upper Yellow River gorge and its tributaries between Tongde and Zoige basins that the Yellow River propagated upstream due to headward incision. From OSL and C-14 ages of low terraces 10–140 m above the present river bed in Tongde and upstream that range from 9–140 ka, (Fig. 3), they infer an onset

of incision from the top of the basin fill in Tongde at about 400–500 ka (Craddock et al. 2010). In more details, their results imply the upward propagation of a wave of incision starting around 500 ka in Tongde with incision rates of c. 2.5 mm a21 that slowed down as incision progressed to less than 1 mm a21 at present (Harkins et al. 2007; Fig. 3). The terraces in Gonghe have not been directly dated so far. Thermoluminescence (TL) dating of loess covering seven terraces in Linxia give ages that range from 10– 160 ka (Li et al. 1997) and due to the similar setting of terraces in Gonghe, the oldest of these ages has been attributed to the highest terrace in Gonghe (Li 1991; Li et al. 1997; Fothergill & Ma 1999; Zhang et al. 2004). However, from palaeomagnetostratigraphical studies in Guide and Linxia, it has been argued that the highest terrace in Guide postdates closely the shallowest dated sediments at 1.8 Ma (Fang et al. 2005), and that the Yellow River started incising in Gonghe at c. 1.1 Ma (Yan et al. 2004), in agreement with the inference of a post 500 ka incision of the Yellow River in Tongde and the upward propagation of a wave of incision (Harkins et al. 2007). It is thus necessary to directly date the terraces, to determine their age of deposition and abandonment to constrain the incision history of the Yellow River and the rates of tectonic uplift of the narrow ranges cross-cut by the river.

INCISION RATE OF YELLOW RIVER IN NE TIBET

199

Fig. 7. (a) View towards south of terrace T7 surface, (see location in Fig. 2). Surface T7 merges smoothly with range slopes at the margin of the basin. (b) View towards north of T7 on the right bank of the Yellow River, in southern Gonghe basin. In front, gravel quarry that allowed access to first 10 m of terrace deposit.

200

A. PERRINEAU ET AL.

Fig. 8. (a) The top 10 m of terrace T7 in a gravel quarry on the right bank of the Yellow River, north of Heka Shan (see location in Fig. 2). The terrace conglomerate is covered by a 2 m thick loess deposit. Yellow vertical line (hanging tape) shows position of depth profile. Schematic section is shown on left with position of samples. Single pebbles, amalgamated samples and sand lenses were collected from 2 –10 m depth. (b) Detail of palaeosol between loess cover and conglomerate. (c) Detail of sand lenses.

INCISION RATE OF YELLOW RIVER IN NE TIBET

Terrace dating, results and interpretation Cosmogenic nuclide dating method Cosmogenic nuclides are produced in situ by the interactions between secondary cosmic ray particles and surface rocks. Cosmogenic nuclides accumulate in rocks as long as they stay at/near to the surface of the Earth. 10Be and 26Al are mainly produced in quartz from O and Si by spallation and muonic capture, and the ubiquity of this mineral makes it a favoured target. The concentration of cosmogenic radioactive nuclides in a rock, N, is given by: dN ¼ P  lN; dt

(1)

where l is the decay constant of the nuclide considered and P is the production rate, which depends on the geographic location and altitude. For sub-surface samples, the production rate decreases exponentially with depth following:  r (2) P(z) ¼ P(0) exp z ; L where P(0) is the production rate at the surface, z the depth of the sample, r the density of the overlying material and L the attenuation length of the interacting particle. Analysing deep (shielded) samples may give access to the amount of inheritance. Indeed, samples situated below several times the penetration depth are shielded from cosmic radiations and their cosmogenic nuclide concentrations thus correspond to nuclides accumulated before being deposited, that is, the inherited component. When inheritance and erosion are assumed to be negligible, then the minimum exposure age of a sample is given by:   1 1  lN : (3) tmin ¼ ln l P Taking into account erosion, inheritance and production by muons, the number of 10Be atoms grows with time according to: N(z, Dt) ¼ N0 (z, Dt) exp (lDt) 2 3   r 7 6 Pn þ4 r exp  L z 5 n lþ 1 Ln      r 1 Dt  1  exp  l þ Ln 2 3   r 6 Pm 7 þ4 r exp  L z 5 m lþ 1 Lm      r 1 Dt ,  1 exp  l þ Lm

(4)

201

where N0(z, Dt) is the inherited component, (n, m) subscripts for spallation by neutron and muons, respectively. Samples were processed in the Cosmogenic Nuclides Laboratory of the University of Strasbourg following standard methods described in Kohl & Nishiizumi (1992). 10Be/9Be and 26Al/27Al ratios were determined by accelerator mass spectrometry (AMS) at the ASTER (CEREGE, Aix-en-Provence) facility except two samples (Qi4-29-1 and Qi429-19, Table 1), which were measured at the Center for Accelerator Mass Spectrometry at Lawrence Livermore National Laboratory (Table 1). Natural 9Be has been assumed to be not present in our samples after a set of samples has been checked by ICP-MS. The total 9Be concentration is thus determined by the amount of carrier added to the sample. The 26Al concentrations were measured in a subset of samples to better constrain the exposure history of the terraces. As the stable isotope 27Al is naturally present in quartz minerals, sample content of natural 27Al was measured by ICP-MS before addition of 27Al carrier. 10Be and 26Al have different decay constants and production rates, but the same production profiles. They can thus be analysed together and compared. In case of a simple exposure history, without a period of burial, the 26Al/10Be nuclide ratio approximates 6.75 (Balco 2009). Nuclide concentrations were determined from the AMS measurements and the amount of corresponding stable isotope (Table 1). Ages were modelled using the CRONUS online calculator (Balco et al. 2008) version 2.2 (April 2009). Sub-surface sample concentrations have been modelled by fitting the sample concentrations following equation (2) (see also Granger & Smith 2000) and using the constant surface production rate determined by the CRONUS calculator (see details below). We used the revised 10Be decay constant of 5.1 + 0.26  1027 a21 of Nishiizumi et al. (2007) and the 26Al decay constant of 9.83  1027 a21, following Balco et al. (2008). Penetration depths used for depth profiles ages calculation are 160 g cm22 for neutrons and 1500 g cm22 for muons (Granger & Smith 2000; Gosse & Phillips 2001). We used densities of 2.65 for quartz, 2.7 for granite, 2.5 for a mix gravel/pebble/cobble deposit, 1.8 for soil, 1.4 for loess. No topographic shielding correction was needed given the extent and flatness of the terraces. All other data used in the calculator, as advised by Balco et al. (2008) and Frankel (2010), are given in Table 1. The largest analytical uncertainty comes from the production rate estimation (Gosse & Phillips 2001). Balco et al. (2008) estimated this uncertainty to approximate 10%. It is however larger for depth samples because uncertainties on penetration

202

Table 1.

10

Be and 26Al analytical results of surface and sub-surface samples of the Yellow River terraces in the Gonghe basin

Sample name

Qi4-29-27b† †

Qi4-29-27c

†

Qi4-29-27d

Qi4-29-19* Qi4-29-9 Qi4-29-13 Qi4-29-11 Qi4-29-11-2‡

Latitude (8N)

Longitude (8E)

Elevation (m a.s.l.)

Quartz cobble

36.17584

100.52143

2920

0

3

Quartz cobble Quartz cobble Quartz cobble

36.17567 36.17520 36.17532

100.52181 100.52123 100.521

2920 2920 2918

0 0 0

4 8 5.5

Quartz cobble Quartz cobble Quartz cobble

36.17555 36.17566 36.17595

100.52113 100.52179 100.52154

2920 2919 2920

0 0 0

Quartz cobble 8 amalgamated pebbles 13 amalgamated pebbles 32 amalgamated pebbles 55 amalgamated pebbles

35.70043 35.70043

100.29636 100.29636

3196 3194

35.70043

100.29636

35.70043

Quartz cobble Quartz cobble Quartz cobble Quartz cobble

err

26 Al concentration (at g21)

err

[26Al]/ [10Be] ratio

6 8 5

2.98E þ 06 3.12E þ 06 2.98E þ 06 3.26E þ 06 3.27E þ 06 3.34E þ 06 3.05E þ 06 3.74E þ 06 2.84E þ 06

8.44E þ 04 8.87E þ 04 8.57E þ 04 9.37E þ 04 9.27E þ 04 9.77E þ 04 9.43E þ 04 1.02E þ 05 8.52E þ 04

na na na na na na na na na

na na na na na na na na na

200 200 –220

5 2

9.55E þ 05 1.53E þ 06

2.56E þ 04 4.39E þ 04

4.64E þ 06 na

5.33E þ 05 4.86 na na

0.57 na

3194

200 –220

2

1.32E þ 06

4.04E þ 04

na

na

na

100.29636

3194

200 –220

1.5

1.47E þ 06

6.43E þ 04

8.88E þ 06

4.31E þ 05 6.03

0.39

35.70043

100.29636

3194

200 –220

1

1.54E þ 06

5.53E þ 04

9.11E þ 06

3.94E þ 05 5.91

0.33

35.70043 35.70043 35.70043 35.70043

100.29636 100.29636 100.29636 100.29636

3194 3194 3194 3194

Weighted mean: 230 4 240 4.5 260 5.5 320 6

2.13E þ 05 3.80E þ 05 na 1.11E þ 05 na

0.25 0.23 na 0.37 na

1.4615E þ 06 + 0.1031E þ 06 1.07E þ 06 2.85E þ 04 6.33E þ 06 2.13E þ 06 7.31E þ 04 8.77E þ 06 7.66E þ 05 2.83E þ 04 na 3.97E þ 05 1.61E þ 04 2.43E þ 06 4.00E þ 05 1.59E þ 04 na

na na na na na na na na na

err

na

5.89 4.12 na 6.12 na

na na na na na na na na na

A. PERRINEAU ET AL.

Terrace T5 Qi4-30 Qi4-30-2‡ Qi4-33 Qi4-34 Qi4-36 Qi4-36-2‡ Qi4-38 Qi4-39 Qi4-40 Terrace T7 Qi4-29-1* Qi4-29-27a†

Depth (cm)

10 Be Sample thickness concentration (at g21) (cm)

Sample description

Sand lens Sand lens Quartz cobble Quartz cobble Quartz cobble Sand lens Quartz cobble

35.70043 35.70043 35.70043 35.70043 35.70043 35.70043 35.70043

100.29636 100.29636 100.29636 100.29636 100.29636 100.29636 100.29636

3194 3194 3194 3194 3194 3194 3194

370 390 400 615 675 750 480

2 2 6 5 9 2 6

7.85E þ 05 5.05E þ 05 6.18E þ 05 2.35E þ 05 1.90E þ 06 4.45E þ 05 5.38E þ 05

2.74E þ 04 1.75E þ 04 2.45E þ 04 9.48E þ 03 6.54E þ 04 1.87E þ 04 2.04E þ 04

3.79E þ 06 2.46E þ 06 3.66E þ 06 1.34E þ 06 9.67E þ 06 2.25E þ 06 3.26E þ 06

1.51E þ 05 9.65E þ 04 1.62E þ 05 6.88E þ 04 4.12E þ 05 2.03E þ 05 1.35E þ 05

4.83 4.87 5.93 5.69 5.09 5.06 6.06

0.25 0.26 0.35 0.37 0.28 0.50 0.34

Granitic bedrock

36.12209

100.93520

2680

0

4

Granitic bedrock Granitic bedrock Cobble Cobble Cobble Cobble Cobble

36.12221 36.12184 36.12196 36.12203 36.12212 36.12212 36.12214

100.93571 100.93532 100.93521 100.93511 100.93505 100.93505 100.93503

2680 2683 2680 2680 2682 2681 2681

0 0 0 0 0 0 0

3.5 3.5 8 8 8 4 4

1.28E þ 06 1.40E þ 06 1.18E þ 06 1.90E þ 06 1.91E þ 06 1.89E þ 06 1.85E þ 06 1.90E þ 06 1.89E þ 06

3.72E þ 04 5.74E þ 04 3.44E þ 04 7.73E þ 04 5.46E þ 04 5.43E þ 04 6.26E þ 04 6.11E þ 04 6.28E þ 04

na na na na na na na na 1.17E þ 07

na na na na na na na na 5.83E þ 05

na na na na na na na na 6.16

na na na na na na na na 0.37

Quartz vein in bedrock Granitic bedrock Cobble

36.13491

100.95922

2925

0

8

5.61E þ 05

1.91E þ 04

na

na

na

na

36.13455 36.13449

100.95926 100.96021

2934 2938

0 0

3.5 4

36.13405

100.96105

2940

0

3

1.82E þ 06 2.64E þ 06 2.58E þ 06 2.49E þ 06

6.16E þ 04 9.86E þ 04 8.39E þ 04 7.56E þ 04

na na na 1.30E þ 07

na na na 5.30E þ 05

na na na 5.23

na na na 0.27

36.12972

100.9967

2367

0

3.09E þ 05

1.24E þ 04

1.63E þ 06

7.63E þ 04 5.28

0.32

Qi4-60 Qi4-61 Qi4-61-2‡ Qi4-62 Granitic bedrock Yellow River sand Qi4-43A River sand

na, not analysed. *AMS measurements at CAMS at Lawrence Livermore National Laboratory. † See details of sample characteristics in Figure 9. ‡ Duplicate of the same sample after quartz leaching.

INCISION RATE OF YELLOW RIVER IN NE TIBET

Qi4-29-14 Qi4-29-15 Qi4-29-16 Qi4-29-21 Qi4-29-22 Qi4-29-23 Qi4-29-24 Terrace WRB Qi4-47 Qi4-47-2‡ Qi4-48 Qi4-49 Qi4-50 Qi4-51 Qi4-52 Qi4-53 Qi4-54 Terrace WLB Qi4-59

203

204

A. PERRINEAU ET AL.

Fig. 9. Mass distribution plot (probability density function) of amalgamated pebbles of samples Qi4-29-27a, b, c and d collected about 2.1 + 0.1 m deep below T7 surface. Number of pebbles, average mass and approximate size are indicated. Concentrations of 10Be measured in amalgams are shown (see Table 1) together with weighted mean.

depths (6%), sample depth (0.5%) and density of overlying material (2–3%) must be added. Final uncertainties on production rates are then comprised between 10 and 12%. Uncertainties from the chemical processing are low (1–2%). Uncertainties from the AMS measurement are comprised between 1.8 and 3.5%. Total analytical uncertainties on ages, not considering modelling errors, thus range from 11–13% (Table 1). We sampled four different terrace levels for cosmogenic dating, T7 and T5 in the basin, and two strath terrace levels in the gorge, WRB (Waligong right bank) and WLB (Waligong left bank) (Fig. 2, Table 1).

Cosmogenic nuclide measurements Terrace T7 samples Stratigraphy. On the southern side of the Gonghe basin, north of Heka Shan and on the right bank of the Yellow River, clean exposure of the top 10 m of the T7 level can be seen in a gravel quarry used for road constructions (Figs 2d, 7 & 8). The terrace conglomerate is a thick fill deposit, composed of gravels, pebbles and cobbles with a few 5–50 cm thick and several metres long sand lenses (Fig. 8c). It is covered by a 15 cm thick reddish palaeosol (Fig. 8b), itself covered by c. 2 m of aeolian loess (in the following sections, layer and sample depths are given relative to present ground surface, i.e., the top of the loess deposit). The conglomerate pile is marked by an angular

unconformity at c. 3.5–4 m depth. In an east – west section, the c. 58 west-dipping pebble layers and sand lenses are intersected by overlying horizontal layers. In the north–south section, the angular unconformity disappears, but a clear layer transition can be seen in the section approximately at the same depth (Fig. 8). The two conglomerate units will be called C1 and C2 from bottom to top in the following sections. The top 20 cm of the C2 conglomerate, just below the palaeosol, is marked by a visible increase in the amount of whitish quartz-vein centimetric pebbles (Fig. 9; Table 1). Our field observations and inspection of the quarry cliffs have not allowed us to distinguish any other sedimentation interruption or other unconformities. Note that the latter would be difficult to detect since there is no obvious change in the nature and origin of the material deposited. Therefore, in the following discussion, we will consider the section analysed as a sequence of two depositional events (assuming that other sedimentation interruptions, if present, lasted only for short periods compared to the whole life time of the terrace), followed by the formation of a palaeosol, itself followed by the deposition of loess. Sampling. From depths of 2– 7.5 m we collected 16 quartz-rich samples for cosmogenic exposure dating along two nearby vertical profiles (Fig. 8, Table 1). At a depth of 2 m, a cobble was found above the palaeosol but below the loess cover (sample Qi4-29-1). Below the palaeosol, in unit C2, we collected a large number of centimetre-sized quartz-vein pebbles at a depth of 2.10 + 0.10 m. These pebbles were sorted by size and analysed as 4 separated amalgamated samples (samples Qi429-27a, b, c and d; Table 1; Fig. 9). Four additional samples are individual quartz-rich cobbles (quartz vein or granitoid) from 2.3 to 3.2 m. In unit C1, three sand lenses were sampled at 3.7, 3.9 and 7.5 m (samples Qi4-29-14, 15 and 23, respectively; Table 1). The 4 remaining samples are individual quartz-rich cobbles at 4.00, 4.80, 6.15 and 6.75 m (samples Qi4-29-16, 24, 21 and 22; Table 1). Among those 16 samples, 13 were analysed for 26 Al (Table 1). 10 Be nuclide concentrations. 10Be concentrations range from 235 000 to 1 500 000 at g21 (qtz) for most of the samples, except 2 that have highest concentrations of about 2 millions at g21 (qtz) (Fig. 10a, Table 1). Lowest concentrations are those of cobble samples Qi4-29-21 and Qi4-29-11 with, 235 000 and 405 000 at g21 (qtz), respectively. The sand lens samples have larger concentrations of 445 000, 505 000 and 785 000 at g21 (qtz). The large concentrations of 1 500 000 at g21 (qtz) is reached for samples right below the

INCISION RATE OF YELLOW RIVER IN NE TIBET

(a)

(b)

N 10Be (x 10e6 at/g quartz) Depth below present surface (cm)

0

1

2

3

4

5

6

7

0

(3 )

(2)

(1)

100

Loess

1

200

13 300

9

19

15

400

600

6,0

7,0

(1) (3 ) 1

27 c d

(2)

19 11

9 14 400

C1

15

16 24

500

cobble sand amalgamated outlier not considered in model

21 22

700

23

21

600

22 700

23 800

800

N 10Be (x 10e6 at/g quartz)

(c) 0

Depth below present surface (cm)

5,0

300

14 16 24

500

N26Al/N10Be 4,0

200

C2

11

3,0 0

100

Paleosol

27

205

1

0

100

2

5

19

27

16

15ka

C2

15ka

C1

24

21 22

700

23

7

Paleosol 9

11 14 15

6

(1,150)

Loess

1 13

600

4

(50,10)

300

500

(10,120)

(25, 70)

200

400

3

cobble sand amalgamated outlier not considered in model

800

Fig. 10. (a) Plot of 10Be concentrations of sub-surface samples of terrace T7 (see Table 1). Schematic stratigraphy of the section is shown. (b) 26Al/10Be ratio as a function of depth for a subset of samples of Figure 10a. Models 1, 2 and 3 are fitted to both the 10Be and the 26Al/10Be ratio (see text and Table 2). (c) Models showing the impact of various loess accumulation scenarios, first number is duration of loess deposit, second number is duration of exposure at the end of loess deposit. Conglomerates C1 and C2 where exposed each during 15 ka (Table 2).

soil-conglomerate interface, samples Qi4-27a, b, c and d (Table 1). The whole sequence of 10Be concentrations measured in the different samples do not show a monotonous exponential decrease with depth but imply either sequential deposition (e.g. Schaller et al. 2002; Matmon et al. 2009) or a variable inheritance between samples (e.g. Anderson et al. 1996; Repke et al. 1997; Me´riaux et al. 2004, 2005; Le Dortz et al. 2009). In particular the different types of samples collected, individual cobbles, sand lenses or amalgamated pebbles may have various pre-exposure components (e.g. Oskin et al. 2008). However, within each unit, C1 and C2, the sample concentrations show an exponential

decrease with depth linked to their respective exposure histories (Fig. 10). 26 Al nuclide concentrations. The 26Al concentrations were measured in 13 samples (Table 1). The concentrations range from 2 –9 millions at g21 (qtz) and imply 26Al/10Be ratios that vary between 4.12 to 6.06 (Figs 10 & 11; Table 1). Low 26 Al/10Be ratio values usually result from differential decay of 10Be and 26Al caused by shielding from cosmic rays after a period of nuclide accumulation (Bierman 1994). None of the samples have typical simple exposure surface 26Al/10Be ratio of 6.75 implying either shielding of the section or

206

A. PERRINEAU ET AL.

10 const ant exposure - no erosion steady-state e rosion N26AI/N10Be

500 ka -27c -54 -19 -62 -16 -27d -15 -22 -43A -1 -23 -14 -9

-21

1 Ma

2 Ma

T7 - sand samples T7 - cobbles T7 - Qi4-29-1 base of loess Yellow river modern sand Waligong terraces

1.0E+05

1.0E+06

1 Ma

1

100 ka

10 k

a

3 Ma

1.0E+04

-24

-11

1.0E+07

1.0E+08

1.0E+09

N10Be Fig. 11.

26

Al/10Be ratio v. 10Be for a subset of samples from the depth profile of terrace T7 (Table 1).

a significant component of inherited nuclides. Disregarding a few anomalous samples, such as, Qi4-29-1 that is probably a reworked cobble found above the palaeosol, Qi4-29-9 and Qi429-22 that have high 10Be concentrations (large inheritance) and the sand samples, the 26Al/10Be ratio decreases slightly from 6 to 5.7 between 2 m to 6.15 m depth (Fig. 10b). The sand lenses have lower 26Al/10Be ratios of about 4.8 to 5.06, which may indicate incorporation of material with inherited low 26Al/10Be ratios. We note that this might also be the case of present river sands, which have a ratio of about 5.28 (sample Qi4-43A, Table 1). We will jointly model (see below) the 10 Be concentration profile and the 26Al/10Be ratio to constrain the age of deposition and abandonment of the T7 level. Terrace T5 samples Geomorphological setting. Terrace T5 is one of the largest terraces SW of Gonghe situated about 150 m below T7 (Figs 2, 3 & 12). It is flat in its northern part where the surface is characterized by a loose gravel and pebble pavement (Fig. 12). At places, sand dunes (barkanes) travel southeastwards across the terrace with the probable consequence of incorporation of modern sand below the pavement and abandonment of sandy patches at the terrace surface. We selected the wide paved surface because such surfaces are usually characterized by low erosion rates and are thus well suited for surface exposure dating (e.g. Van der Woerd et al. 1998, 2006; Me´riaux et al. 2004; Matmon et al. 2009).

Sampling and measurements. We sampled 7 well embedded cobbles on terrace T5 (Fig. 12). The 7 samples have close 10Be concentrations on the order of 3  106 at g21 quartz (Table 1). This high nuclide concentration, which is twice the concentration of samples Qi4-29-27 from 2 m depth in T7, suggest that: (1) T7 must have been shielded quite rapidly after deposition; and (2) the T5 terrace surface has never been shielded; and (3) inheritance and/or erosion, if present, have similar effects on the samples of T5. We will use the weighted mean of the 7 sample concentrations (3.21 + 0.30  106 at g21 (qtz); Table 1) as the 10 Be nuclide concentration of terrace T5 surface. Waligong terrace samples Geomorphology. It is across the Waligong ranges that the Yellow River has cut its most impressive gorge over a length of 30 km and a total drop of water level of 200 m (Figs 2 & 3). Across the western Waligong, the Yellow River has carved a deep gorge with walls reaching about 500 m (Fig. 3a, b). The flat top of the gorge mimics the overall shape of the western Waligong (Figs 4 & 5), and slopes towards the west into Longyangxia reservoir (Fig. 2). In fact, in more details, the top of the gorge is formed by the progressive down cutting of the river resulting in abandonment of a sequence of strath terraces carved in the bedrock, which are covered by a thin deposit of fluvial gravel and cobbles (Figs 5 & 13). These terraces now slope towards the west indicating that the western Waligong is actively growing and has progressively back tilted the terraces. While the main

INCISION RATE OF YELLOW RIVER IN NE TIBET

207

Fig. 12. (a) View to east of terrace T5 surface, showing a loose pavement of gravels and pebbles. Note absence of loess cover (see location in Fig. 2). At places, terrace T5 is covered by sand dunes and grass patches. Surface cobbles were collected for 10Be cosmogenic dating in very flat areas with no sign of soil development or local incision. (b–h) close-up of cobbles sampled for surface exposure dating.

208

A. PERRINEAU ET AL.

Fig. 13. (a) View of Waligong left bank terrace (WLB) just above Longyang gorges (see Fig. 2 for location). The terrace is a smooth strath carved in granitic bedrock of Waligong Shan covered by a thin layer (.1 m) of gravels and pebbles. Both granitic bedrock strath and overlaying pebbles were targeted for 10Be exposure dating. (b) Schematic section with sample locations on WLB. (c) View of Waligong right bank (WRB) terrace with sample locations. Similarly to WLB, the strath terrace is smooth and covered by a thin layer of pebbles and cobbles. (d) Schematic section of WRB with sample locations. Samples Qi4-47 to -49 are bedrock, samples Qi4-50 to 54 are cobbles.

INCISION RATE OF YELLOW RIVER IN NE TIBET

terrace levels can be clearly depicted in the landscape (Fig. 5c, d), their shape narrows near the gorge and their riser join with a low angle the gorge rims so that they become unclear near the gorge and we were unable to number and map them precisely. Sampling and measurements. We sampled two sites on each side of the gorge (Figs 2 & 13), Waligong right bank (WRB) and Waligong left bank (WLB), located at 2680 and 2940 m a.s.l., respectively. Given the difficulties of terraces mapping near the gorge rim, these two sites are not located within the regional Gonghe basin terrace stratigraphy. Elevation only is not sufficient as we suspect uplift to have occurred since their abandonment (see below). At both sites we sampled the bedrock as well as the overlying cobbles, 4 samples at WLB and 8 samples at WRB (Fig. 13, Table 1). On terrace WLB, samples have 10Be concentrations that range from 0.56 to 2.6  106 at g21 (qtz) (Table 1). The lowest concentration (0.561  106 at g21 (qtz)) is found in a piece of a highly fractured quartz vein sampled near the gorge cliff edge (sample Qi4-59, Fig. 13a). Qi4-60 is a bedrock sample near the cliff edge and has a concentration of 1.82  106 at g21 (qtz). Both remaining samples have higher concentrations, Qi4-62 is a protruding bedrock high (Fig. 13a) and Qi4-61 is a cobble from the top of the terrace conglomerate cover (Fig. 13a). We think that the average concentration of these two latter samples of 2.57 + 0.11 at g21 (qtz) (Table 1) corresponds to the nuclide concentration accumulated since terrace abandonment and that the lower concentrations in the other samples are due to removal of the conglomerate cover and erosion of the fragile quartz vein outcrop. On WRB terrace, samples have 10Be concentrations that range from 1.2 to 1.9  106 at g21 (qtz) (Table 1). The 35% difference in concentration among the WRB samples may be explained by the relative positions of the samples in the section sampled (Fig. 13c, d). It appears that samples with the lowest concentration, Qi4-48 and Qi4-47 (Fig. 11c) are bedrock samples close to the gorge cliff, which are no more covered by the terrace conglomerate. Sample Qi4-49 belongs to a bedrock high (Fig. 13c) standing at the same level as the pebbly conglomerate top and show a similar 10Be concentration as the 5 surface cobbles Qi4-50 to 54 (Table 1). The distribution of concentrations of the WRB samples is thus best explained by removal of the conglomerate cover near the gorge cliff. The 10Be concentration average of 1.89+ 0.02 e 6 at g21 (qtz) of the 6 samples with highest concentrations correspond to the nuclides accumulated since terrace abandonment.

209

Age modelling Model age of T7. We modelled jointly the 10Be concentration and the 26Al/10Be ratio in the depth profiles (e.g. Anderson et al. 1996; Granger & Smith 2000) considering a deposition scenario in three steps corresponding to the stratigraphy observed in the field (Figs 8 & 10). Modelling results are presented in Figure 10 and take into account progressive increase in muongenic production at depth following analytical solutions described in Granger & Smith (2000) and calibrated to our site (Table 2). Disregarding samples with large inheritance and the sand samples, model 1 is a best fit adjusting all the 26Al/10Be ratios, and models 2 and 3 are limiting models toward short and long exposure times, respectively. In general, both the 10 Be concentration and the values of the 26Al/10Be ratio imply shielding by the loess during more than 150 ka (model 1). Involving fast erosion rates of the loess (0.01 cm a21, model 3) very long time of loess shielding may be implied (400 ka) with the necessity to have an original loess thickness reaching several tens of metres. Because the 2 m-thick loess deposit on top of the terrace is a shielding factor for cosmic rays (factor of decrease in production rate amounts to c. 6) (e.g. Hetzel et al. 2004; Matmon et al. 2009) we have to consider its mode and timing of deposition. The presence of a palaeosol below the loess (Fig. 8) implies that the terrace surface remained free of loess during a certain time, that is, quasi without any shielding. Without independent chronological constraints on the loess age, several loess deposition modes can be modelled. In the first mode, the 2 m thick loess is deposited instantaneously (or rapidly compared to the age of the terrace), in the second mode it was deposited continuously during a certain time (Fig. 10c). Some constraints can be added from additional observations in Gonghe and elsewhere. As we mentioned above, loess is absent on the lower Yellow River terraces across the Gonghe basin, particularly on T5, so that the age of T5 may be the time when loess deposition stopped in the area (see below). In addition, 1 to 2 m high loess sections described in northern Tibet generally show progressively younger ages towards the top (e.g. Peltzer et al. 1988; Van der Woerd et al. 2002; Hetzel et al. 2004; Rhode et al. 2007) over a time period of a few thousands of years. Another possible factor of shielding variation is involved if the loess cover is reworked by wind ablation implying a period of loess deposition followed by a period of loess erosion (model 3). Assuming that sub-surface samples cannot lose any 10Be, except by radioactive decay, the lowest concentrated and deepest samples in a depth-profile (below c. 1.5 m, Anderson et al. 1996) determine

210

Table 2. Depth profile model parameters (see Fig. 10) Inheritance (at g21 quartz)

Exposure of C1 (ka)

Erosion rate (cm a21)

Exposure of C2 (ka)

Erosion rate (cm a21)

Duration of loess deposit (ka)

Exposure loess (ka)

Erosion rate (cm a21)

Total duration of exposure (ka)

1

235 000

15

0

45

0.005

1

150

0.0001

210

2

445 000

15

0

20

0

1

80

0

115

3

235 000

15

0

50

0

400

0.01

465

Models of Figure 10c

235 000 235 000 235 000 235 000

15 15 15 15

0 0 0 0

15 15 15 15

0 0 0 0

10 70 120 150

0 0 0 0

90 125 160 181

50 25 10 1

A. PERRINEAU ET AL.

Model

INCISION RATE OF YELLOW RIVER IN NE TIBET

Model age of T5. Without independent constraints on erosion and inheritance for terrace T5, these parameters have to be estimated. Inheritance may be taken similar as for terrace T7 (235 000 –445 000 at g21quartz) or similar to the

present Yellow River sand 10Be concentration of c. 300 000 at g21quartz (Table 1). Regarding erosion, studies have been carried out in the Gonghe basin following intensification of agriculture and its consequence on desertification. Using the 137Cs technique to assess erosion in drylands and sand shifting areas, erosion rates of about 1– 2 mm a21 have been estimated over the last c. 40 years (Yan et al. 2002; Zhang et al. 2003). These values of erosion are rather high (Lal et al. 2003) and may only reflect the consequence of human activities. Lower values of 0.003–0.01 mm a21 were estimated for bedrock outcrops throughout Tibet (Lal et al. 2003). Again these values should be taken as maximum erosion rates for a flat alluvial terrace. The age of T5 calculated from the weighted mean concentration of the surface samples 3.21+0.30  106 at g21 (qtz), is 106 + 6 ka using a constant production rate model (Lal 1991; Stone 2000). Both the inheritance of 300 000 at g21quartz and a maximum erosion rate of 0.002 mm a21 (Fig. 14) imply subtractive and additive corrections of 15–20%, respectively. Without evidence that erosion rates may be larger, we thus conclude that the minimum age of T5 is 106 + 15 ka. Age model of Waligong. To calculate the age of the terraces we need to take into account the change in elevation due to their tectonic uplift (e.g. Hetzel et al. 2004; Ruszkiczay-Ru¨diger et al. 2005a, b) (Fig. 15). Waligong Shan is composed of two subparallel ranges at the junction between the eastern termination of the Qinghainan and the Laji Shan closing the northeastern corner of the Gonghe basin (Fig. 2). The western Waligong is a 15 km wide and 50 km long arcuate range trending roughly NE– SE (Figs 1 & 2). Its asymmetric topographical shape, with a steep NE and eastern flank under the summital crest that reaches 3500 m a.s.l. above a west dipping thrust and a western flank 5000000

/a

m

m

/a

m

4000000

1

00

0 =

N10Be (at/g quartz)

the minimum inherited component of the samples. We have thus considered the inheritance given by sample Qi4-29-21 (i.e. 235 000 at g21, Table 1) in most models, and show only one alternative model (model 2) with an inheritance of 445 000 at g21 fitting the profile to the two sand samples with the lowest concentration (Qi4-29-23 and 15; Fig. 10, Table 2). Note that the present Yellow River is carrying sand that has a 10Be concentration of about 300 000 at g21quartz (Table 1; see also Harkins et al. 2007) that may be considered as the average inherited 10Be component of the alluvial material transported by the river, assuming it is constant over time and not variable with material size. While below the palaeosol, the fits to the data mostly depends on the choice of the inheritance value, above the palaeosol, the fits mostly depends on the onset of loess shielding (Fig. 10a, Table 2). These models also show that, in all cases, deposition of conglomerate C2 occurred shortly after deposition of conglomerate C1 (10– 15 ka). Similarly, deposition of the loess cover occurred shortly (0–25 ka) after deposition of conglomerate C2. As shown in Figure 10c (Table 2), when a longlasting loess deposition process is considered, shielding starts to be efficient when this process lasts less than 20 ka. This implies that shielding by the loess must have started a few thousands of years after the terrace has been abandoned and that the loess must have been deposited in a short time. Note that this is fully in agreement with the observation, that the terraces below T7 are not loess covered, so that they must have been abandoned after the loess deposition process ceased in the region and on T7 in particular. Sample Qi4-29-1, sampled above the palaeosol and below the loess may be used as an additional constraint. It is clearly a reworked cobble as it is not emplaced in a fluvial deposit, but it must have been reworked at the time of the formation of the palaeosol and before being capped by the loess. As can be seen in Figure 10a, models that are adjusted to this sample imply a minimum exposure of C2 of 10–15 ka. The same duration would be needed to account for soil formation above C2 and some surface reworking. The 10Be concentration of this sample implies exposition under the loess during 150 ka. To summarize, T7, or more precisely the top C2 conglomerate may have been deposited c. 200 ka, and definitively incised by the Yellow River c. 150 ka, some time before the 2 m loess cap accumulated on top of C2.

211

=

0.

m

m/a

3m

.00

=0

/a

= 0.006 mm

3000000

2000000

1000000

0 0

50000

100000

150000

200000

250000

300000

Time (years)

Fig. 14. Minimum exposure age range of T5 and influence of erosion. Erosion of about 0.001 mm a21 cannot be excluded, but erosion of 0.003 mm a21 may be a maximum (see text for details).

212

A. PERRINEAU ET AL.

(a)

4000

3000

Altitude (m) (x25)

2000

50

0

50

(b)

4000

3000

2000

50

0

50

(c)

4000

3000

2000

50

0

50

(d)

4000

3000

2000

50

(e)4000

0

50

Waliguan Guide

Gonghe

T7

T7

3000

6' 53'46 3 2 1 0

Kongma Altitude (m) (x25)

6 5

WLB WRB

4 3 2

2000 50

0

50

Fig. 15. (a –d) Schematic reconstruction of progressive incision of Yellow River across uplifting Waligong range. Phased incision leads to the formation of large stepped terraces across the basin, while strath terraces formed across Waligong bedrock are progressively uplifted and tilted. (e) Present-day section at same scale.

that slopes c. 3.58 westward, (Fig. 4e) indicate that it is an east-vergent crustal anticline. The eastern Waligong is less elevated but larger (20–30 km), trends NW–SE and makes up the western limit of the Guide basin. Similarly to the western Waligong, it has a steep eastern rim marked by a thrust at its base (Pan 1994) and grows as an east-vergent anticline (Fig. 2). As can be seen in the field (Fig. 13) the strath terraces are carved into the Waligong Mesozoic granodiorite (Geological map of Qinghai 1988). A simple model (Fig. 15) accounting for the tilt of the Waligong Shan and uplift of the terraces

implies possible changes in elevation of 80 m for WRB terrace and of about 300 m for WLB terrace (Fig. 15). For WRB, the in-situ production rate changes from 25.73 at g21 a21 at 2600 m to 27.04 at g21 a21 at 2680 m. This change in production rate of c. 10% remains small with regard to other uncertainties and may be neglected. For WLB the production rate change is more significant, varying from 26.6 at g21 a21 at 2640 m to 31.93 at g21 a21 at 2940 m, that is, a change of 20%. The real mean production rate of this terrace lies certainly between these end-values and we will take the mean between these bounds with a large error of 15%. The terrace ages are thus 74 + 10 ka for WRB and 90 + 15 ka for WLB. The maximum changes in elevation mentioned above imply maximum uplift rates for Waligong Shan between 1–3 mm a21. However, this model does not take into account the low 26Al/10Be ratio (5.23 + 0.27) obtained for sample Qi4-62 (WRG), which is a bedrock sample near the gorge cliff (Fig. 13). This ratio implies that the surface bedrock needs to be exposed a first time (first cosmonuclide accumulation), then shielded during several hundred thousand years (ratio decreases; Fig. 11) and then re-exposed until present. The cobble cover, likely deposited by the river, covered the surface before being progressively eroded to the present state (Fig. 13). The beryllium concentrations of the cobbles would thus reflect the exposure duration since the river left, but bedrock concentrations are not totally related to this last phase of aggradation and erosion. The surface would thus be exposed since c. 80 ka. Initiation of uplift likely began shortly afterwards at a minimum rate of 0.2 mm a21. Despite the fact that the 10Be concentrations of the cobbles are higher on WLB than on WRB because of their higher elevation and production rate, simple calculations show that this scenario would lead to 10Be concentrations close to 2.0  106 at both sites, with a difference of c. 2.0  105 atoms instead of the c. 6.0  105 measured (Table 1).

Discussion Onset age of incision in the Gonghe Basin Whether the beginning of river incision corresponds to the first appearance of the Yellow River in the basin is an open question. Terrace sequences along the Yellow River differ markedly from Gonghe to Lanzhou. As explained above, terraces in the different basins seem completely unrelated to each other (ages, loess thickness, elevations). We cannot simply rely on these terraces to compare our chronology in the Gonghe Basin, but the timing of river incision should be coherent between the different basins (Harkins et al. 2007).

INCISION RATE OF YELLOW RIVER IN NE TIBET

Several authors have proposed that incision in the Gonghe Basin began at c. 150 ka (e.g. Li 1991; Pan 1994). Evidences for a major change along the Yellow River 150 ka ago have been summarized by Zhang et al. (2004). Periods of high regressive erosion from the Longyang gorges and the Sanmen gorges (in the middle reaches of the Yellow River north of the Ordos plateau, inset, Fig. 1) are reported around that time (Zhang et al. 2000; Wang et al. 2002). This period of erosion in the Longyang gorges could have followed the folding of the fluvial sediments of the Gonghe formation (Li 1991). An increased sedimentation in the marine delta of the river could be related to this period of high erosion and to the climate warming of MIS-5 (Imbrie et al. 1984). However, it has been proposed that incision began c. 500 (+200) ka ago in the Tongde Basin (Harkins et al. 2007; Craddock et al. 2010), following the last 1.8 Ma lacustrine deposit near the top of the section in Guide (Li et al. 1997). A direct implication is that incision in the Gonghe Basin began at the same time or before, but not after, which means that T7 must be much older than results from our investigations suggest. This can only be achieved if a significant amount of erosion or a long period of shielding occurred and may imply that all terraces are significantly older. A 140 ka (OSL) terrace located 140 m above the Yellow River is located at an elevation of c. 2830 m in the Tongde basin (Harkins et al. 2007; Craddock et al. 2010). For

213

comparison, an equivalent terrace in the Gonghe Basin would be near to T1, which is expected to be much younger than 140 ka.

Incision rates of the Yellow River in the Gonghe Basin Terrace ages are plotted against their elevation on Figure 16. The minimum and maximum long-term incision rates are c. 2 and c. 6 mm a21, respectively, with a mean rate of c. 4 mm a21. This mean rate fits well with the ages of T5 and T7. However, for all data it is needed to account for changes in the incision rate. The incision rate between the formation of T7 and T5 depends strongly on the age of T7, if erosion on T5 remains small. A period of high incision is needed between T5/WLB and WRB. The incision rate may have reached 9 mm a21 (10 mm a21 maximum, if no elevation changes are taken into account for the Waligong terraces). Our data suggest that incision was first slow and then accelerated between c. 100–50 ka, before slowing down again. Such fast incision would have favoured the steep Waligong gorges formation. A mean incision rate of 4 mm a21 is higher than those mentioned in other places along the Yellow River in the literature, 0.7 mm a21 since 70 ka near Lanzhou (Wang et al. 2010), between 0.75 –1 mm a21 since 166 ka near Linxia (Li et al. 1997) for the downstream part, between

Elevation (m)

120 ka

250 ka

T7

mm

/a

180 ka

/a

( = 0.0038 mm/a)

/a

3.9

mm

180 ka

T5

( = 0.0048 mm/a)

WLB

9m m/

/a

150 ka

106 ka

/a

a

6.

0

m m

/a

3000

mm

9.1

3.7

3.7

mm

/a

2.6

mm

m

0 4.

m

/a

2.3

2500

0

mm

50

WRB

Huang He

100

150

200

250 ka

Fig. 16. Incision rate of the Yellow River deduced from terrace ages and elevations. Coloured boxes are allowed age and elevation ranges. Yellow River elevation taken at Waligong gorges entrance (¼ WRB site). Several scenarios and associated incision rates are proposed. For Waligong terraces, rates are mean rates calculated from the middle of boxes. Maximum and mean total incision rates 6.0 and 4.0 mm a21, in orange.

214

A. PERRINEAU ET AL.

0.9–1.0 mm a21 for the last 140 ka in the Tongde Basin (Harkins et al. 2007). But it is in agreement with the 6 mm a21since 93 ka in the Guide Basin proposed by Pan (1994).

Climatic correlations and formation of the set of terraces Climatic fluctuations play a key role in shaping fluvial terraces. Correlation of terrace formation periods with climatic events is however not straightforward, because uncertainties inherent to the cosmogenic nuclides dating method often prevent from determining a precise correlation. Transitional periods between glacial and interglacial periods seem nevertheless to play a major role in forming fluvial terraces. These periods would be favourable for incision, while aggradation would occur during stable climatic phases, possibly both glacial and interglacial (e.g. Pan et al. 2003; Vandenberghe 2003). Loess in northeastern China was mainly deposited during glacial periods, while palaeosols developed during interglacial phases, alternating loess and palaeosol layers being correlated to climatic fluctuations (e.g. Ding et al. 2002). The fact that T7 gravels are directly overlain by a palaeosol and this palaeosol by a loess cover suggests that this terrace was abandoned during a deglaciation period and covered by loess during an a posteriori glacial event (Gao et al. 2008; Pan et al. 2009). No such climatic indicators are available for the other terraces, although it is reasonable to assume that their formation followed a similar process, that is, a formation as a consequence of dry/cold to wet/warm climate transition. The terraces in the Gonghe Basin could have been formed during glacial periods and abandoned at the transition with warmer periods, when the river high incision power enabled it to break dams formed by the Waligong Shan.

Evidence for a Gonghe-Guide-Qinghai palaeo-lake Small lakes can be found in several places in the present Gonghe basin (Fig. 6); Chaka Lake, in the NW, has a present elevation of 3060 m; Dalian Lake, at an elevation of 2850 m (Yan et al. 2002); Duolonggou lake, north of Gonghe city, at an elevation of 3194 m. Lacustrine sediments can be found at several places in the Gonghe basin and at Dalian Lake for instance. These small lakes and sediments may be remnants of a larger lake occupying the bottom of the Gonghe basin (Fig. 17). Such a lake may have covered adjacent intra-mountainous basins. It has been reported that sedimentary sequences in the Guide Basin were deposited

under lacustrine conditions (Li et al. 1997; Pares et al. 2003; Fang et al. 2005). Moreover, it is striking that the present level of the Qinghai Lake is exactly at the same altitude as Duolonggou lake and as T7 sampling site (3200 m). Qinghai lake exists since c. 0.5 Ma (Yuan et al. 1990) and its present maximum depth approximates 20 m (Colman et al. 2007). Reconstructions of Qinghai lake water level palaeo-elevations lead to various chronologies (Colman et al. 2007; Madsen et al. 2008; Liu et al. 2010; Rhode et al. 2010). The lake reached a maximum elevation of 3260 m (palaeoshorelines older than 30 ka) and its level was particularly high during the penultimate glacial maximum (c. 150 ka). Other lakes in the region seemingly experienced the same high-stand event, the nearest in distance being Gahai Lake in the Qaidam Basin (Fan et al. 2010). The Chinese literature discusses the existence of an eastern Qinghai palaeo-lake covering the Qinghai, Gonghe, Xinghai and Xining basin (e.g. Pan 1994). It is based on a possible stratigraphic continuity between sediments in the Guide and Qinghai Basin. Such a lake would have broken into several parts due to the recent tectonic activity. Tectonic activity along the Qinghainan Shan fault likely began at the end of the Miocene or at the beginning of the Pliocene (Me´tivier et al. 1998; Meyer et al. 1998; Van der Woerd 1998; Fang et al. 2005; Lease et al. 2007) separating the Qinghai and Gonghe Basin. A connection between the two basins was possible at the southeastern extremity of Qinghai Lake when the pass may have been at a lower elevation (present elevation: 3350 m). The Qinghai Basin and the western Gonghe Basin are now internally drained. Madsen et al. (2008) proposed that Qinghai Lake began to form during the Middle or Late Pleistocene when the Riyue Shan uplift separated the Qinghai Basin from the Yellow River drainage system. Uplifting mountain ranges crossing the river course would have acted as boundaries for the eastern Qinghai palaeo-lake, near the Longyang gorges (Gonghe Basin exit), Shanba gorges (Guide Basin exit) or Jishi Gorges (Xunhua Basin exit) (Fig. 17). The change from a lacustrine to a fluviatile environment would in this case correspond to the breaching of the topographic barriers and emplacement of the modern course of the Yellow River across the basins (Pares et al. 2003; Craddock et al. 2010). However, no large lacustrine deposit can be found at present to attest more clearly the existence of such a lake, in particular at our sampling site of T7. Processes that lead to the formation of the wide palaeo-regional base-level around 3200 m and of lakes present at this elevation stills need to be investigated, as well as the possible past connections between the different basins.

INCISION RATE OF YELLOW RIVER IN NE TIBET Fig. 17. Topographical contours at 3150, 3200 and 3250 m a.s.l. showing possible extensions of a regional palaeolake or closed basin in the area of Qinghai and Gonghe basins before their separation due to lake capture by Yellow River and growth of the Qinghainan Shan. Lake surface elevation may have reached 3200–3250 m a.s.l. This mapping, drawn from the present topography, does not account for large tectonic and morphological changes. 215

216

A. PERRINEAU ET AL.

Conclusion The Gonghe Basin is one of the widest basins of northeastern Tibet. It contains noteworthy wide terraces that testify to an important and recent water level decrease. Our results indicate that terraces were formed at the transition between glacial and interglacial periods. The upper one probably dates from the penultimate glacial maximum (MIS-6; Imbrie et al. 1984; Thompson et al. 1997). Obtained ages show that incision occurs at a mean rate of c. 4 mm a21. Higher incision rates during some periods may be related to Waligong Shan uplift, although the tectonic signal remains difficult to separate from the river incision. Our results highlight the consequences of mountain growth and the interplay between tectonic and climatic processes in building the Tibetan Plateau. We provide chronological constraints supporting that this part of the plateau evolved rapidly in recent time from an internally drained but interconnected set of closed basins to an externally river drained plateau margin. Field access and field trip support was provided by Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing, and the Chinese Earthquake Administration, Beijing, and its local bureau in Xining. We acknowledge support from INSU-CNRS (programmes Relief de la Terre), from CNES-SPOT Image (Tectoscope program, image acquired through ISIS program). The 10Be measurements were performed at the ASTER AMS French national facility (CEREGE, Aix-en-Provences), which is supported by the INSU-CNRS, the French Ministry of Research and Higher Education, IRD, and CEA. Two measurements were also made at Center for Accelerator Mass Spectrometry at Lawrence Livermore National Laboratory. We thank Ecole et Observatoire des Sciences de la Terre (Strasbourg, France) and Universite´ de Strasbourg for support and making possible the operation of the cosmogenic laboratory of Institut de Physique du Globe de Strasbourg (UMR 7516) and Laboratoire d’Hydroge´ologie et de Ge´ochimie de Strasbourg (UMR 7517).

References Anderson, R. S., Repka, J. L. & Dick, G. S. 1996. Explicit treatment of inheritance in dating depositional surfaces using in situ 10Be and 26Al. Geology, 24, 47–51. Balco, G. 2009. 26Al– 10Be exposure age/erosion rate calculators: update from v. 2.1 to v. 2.2. World Wide Web Address: http://hess.ess.washington.edu/math/ docs/al_be_v22/al_be_docs.html. Balco, G., Stone, J., Lifton, N. & Dunai, T. 2008. A complete and easily accessible means of calculating surface exposure ages or erosion rates from 10Be and 26 Al measurements. Quaternary Geochronology, 3, 174– 495. Benn, D. & Owen, L. A. 1998. The role of the Indian summer monsoon and the mid-latitude westerlies in Himalayan glaciation: review and speculative

discussion. Journal of the Geological Society, London, 155, 353 –363. Bierman, P. R. 1994. Using in-situ cosmogenic isotopes to estimate rates of landscape evolution: a review from the geomorphic perspective. Journal of Geophysical Research, 99, 13 885– 13 896. Chen, Y. T., Xu, L. S., Li, X. & Zhao, M. 1996. Source process of the 1990 Gonghe, China, earthquake and tectonic stress field in the northeastern Qinghai-Xizang (Tibetan) plateau. Pure and Applied Geophysics, 146, 697–715. Clark, M. K., Schoenbohm, L. M. et al. 2004. Surface uplift, tectonics, and erosion of eastern Tibet from large-scale drainage patterns. Tectonics, 23, TC1006, doi: 10.1029/2002TC001402. Colman, S. M., Yu, S.-Y., An, Z. & Henderson, A. C. G. 2007. Late Cenozoic climate changes in China’s western interior: a review of research on Lake Qinghai and comparison with other records. Quaternary Science Reviews, 26, 2281– 2300. Craddock, W. H., Kirby, E., Harkins, N. W., Zhang, H., Shi, X. & Liu, J. 2010. Rapid fluvial incision along the Yellow River during headward integration. Nature Geoscience, doi: 10.1038/NGEO777. Ding, Z. L., Derbyshire, E., Yang, S. L., Yu, Z. W., Xiong, S. F. & Liu, T. S. 2002. Stacked 2.6-Ma grain size record from the Chinese loess based on five sections and correlation with the deep-sea d18O record. Paleoceanography, 17, doi: 10.1029/ 2001PA000725. Domroes, M. & Peng, G. 1988. The Climate of China. Springer, Berlin. England, P. C. & Houseman, G. A. 1986. Finite strain calculations of continental deformation: 2. Comparison with the India-Asia collision zone. Journal of Geophysical Research, 91, 3664–3676, doi: 10.1098/ rsta.1988.0135. Fan, Q. S., Lai, Z. P., Long, H., Sun, Y. J. & Liu, X. J. 2010. OSL chronology for lacustrine sediments recording high stands of Gahai Lake in Qaidam Basin, northeastern Qinhai-Tibetan Plateau. Quaternary Geochronology, 5, 223–227. Fang, X., Garzione, C., Van der Voo, R., Li, J. & Fan, M. 2003. Flexural subsidence by 29 Ma on the NE edge of Tibet from the magnetostratigraphy of Linxia Basin, China. Earth and Planetary Science Letters, 210, 545– 560. Fang, X., Yan, M. et al. 2005. Late Cenozoic deformation and uplift of the NE Tibetan Plateau: evidence from high-resolution magnetostratigraphy of the Guide Basin, Qinghai Province, China. GSA Bulletin, 117, 1208– 1225. Fluteau, F., Ramstein, G. & Besse, J. 1999. Simulating the evolution of the Asian and African monsoons during the past 30 Ma using an atmospheric general circulation model. Journal of Geophysical Research, 104, 11 995 –12 018. Fothergill, P. A. & Ma, H. 1999. Preliminary observations on the geomorphic evolution of the Guide Basin, Qinghai Province, China: implications for the uplift of the northeast margin of the Tibetan Plateau. In: Smith, B. J., Whalley, W. B. & Warke, P. A. (eds) Uplift, Erosion and Stability: Perspectives on Long-term Landscape Development.

INCISION RATE OF YELLOW RIVER IN NE TIBET Geological Society, London, Special Publications, 162, 183– 200. Frankel, K. L. 2010. Terrestrial cosmogenic nuclide geochronology data reporting standards needed. EOS, Transactions, American Geophysical Union, 91, 31– 32. Gao, H., Liu, X., Pan, B., Wang, Y., Yu, Y. & Li, J. 2008. Stream response to Quaternary tectonic and climatic change: evidence from the upper Weihe River, central China. Quaternary International, 186, 123–131. Gaudemer, Y., Tapponnier, P. et al. 1995. Partitioning of crustal slip between linked, active faults in the eastern Qilian Shan, and evidence for a major seismic gap, the “Tianzhu gap”, on the Western Haiyuan Fault, Gansu (China). Geophysical Journal International, 120, 599– 645. Gaudemer, Y., Tapponnier, P. & Turcotte, D. L. 1989. River offsets across active strike– slip faults. Annales Tectonicae, III, 55– 76. Gosse, J. C. & Phillips, F. M. 2001. Terrestrial in situ cosmogenic nuclides: theory and applications. Quaternary Science Reviews, 20, 1475–1560. Granger, D. E. & Smith, A. L. 2000. Dating buried sediments using radioactive decay and muogenic production of 26Al and 10Be. Nuclear Instruments and Methods in Physics Research, B172, 822 –826. Harkins, N., Kirby, E., Heismath, A., Robinson, R. & Reiser, U. 2007. Transient fluvial incision in the headwaters of the Yellow River, northeastern Tibet. Journal of Geophysical Research, 112, F03S04, doi: 10.1029/2006JF000570. Hetzel, R., Tao, M. et al. 2004. Late Pleistocene/ Holocene slip rate of the Zhangye thrust (Qilian Shan, China) and implications for the active growth of the northeastern Tibetan Plateau. Tectonics, 23, TC6006, doi: 10.1029/2004TC001653. Houseman, G. & England, P. 1993. Crustal thickening v. lateral expulsion in the Indian–Asian continental collision. Journal of Geophysical Research, 98, 12 233 –12 249. Imbrie, J., Hays, J. D. et al. 1984. The orbital theory of Pleistocene climate: support from a revised chronology of the d18O record. In: Berger, A., Imbrie, J., Hays, J., Kukla, G. & Saltzman, B. (eds) Milankowitch and Climate. Reidel, the Netherlands, 269 –305. Joussaume, S. 1999. Climat d’hier a` demain, CNRS Edition, CEA. Kohl, C. P. & Nishiizumi, K. 1992. Chemical isolation of quartz for measurements of in-situ-produced cosmogenic nuclides. Geochimica et Cosmochimica Acta, 56, 3583– 3587. Lal, D. 1991. Cosmic ray labeling of erosion surfaces: in situ nuclide production rates and erosion models. Earth and Planetary Science Letters, 104, 424–439. Lal, D., Harris, N. B. W. et al. 2003. Erosion history of the Tibetan Plateau since the last interglacial: constraints from the first studies of cosmogenic 10Be from Tibetan bedrock. Earth and Planetary Science Letters, 217, 33–42. Lease, R. O., Burbank, D. W., Gehrels, G. E., Wang, Z. & Yuan, D. 2007. Signatures of mountain building: Detrital zircon U/Pb ages from northeastern Tibet. Geology, 35, 1239–1242.

217

Le Dortz, K., Meyer, B. et al. 2009. Holocene right-slip rate determined by cosmogenic and OSL dating on the Anar fault. Geophysical Journal International, 179, 700– 710; doi: 10.1111/j.1365–246X.2009.04309.x. Li, J. 1991. The environmental effects of the uplift of the Qinghai-Xizang Plateau. Quaternary Science Reviews, 10, 479– 483. Li, J.-J., Fung, X.-M. et al. 1997. Magnetostratigraphic dating of river terraces: rapid and intermittent incision by the Yellow River of the northeastern margin of the Tibetan Plateau during the Quaternary. Journal of Geophysical Research, 102, 10 121– 10 132. Lister, G. S., Kelts, K., Chen, K., Yu, J. & Niessen, F. 1991. Lake Qinghai, China: closed-basin lake levels and the oxygen isotope record for ostracods since the latest Pleistocene. Paleogeography, Paleoclimatology, Paleoecology, 84, 141– 162. Liu, X. J., Lai, Z. P., Fan, Q. S., Long, H. & Sun, Y. J. 2010. Timing for high lake levels of Qinghai Lake in the Qinghai-Tibetan Plateau since the last interglaciation based on quartz OSL dating. Quaternary Geochronology, 5, 218– 222. Liu-Zeng, J., Tapponnier, P., Gaudemer, Y. & Ding, L. 2008. Quantifying landscape differences across the Tibetan plateau: implications for topographic relief evolution. Journal of Geophysical Research, 113, F04018, doi: 10.1029/2007JF000897. Lu, H., Wang, X. et al. 2004. Geomorphologic evidence of phased uplift of the northeastern Qinghai-Tibet Plateau since 14 million years ago. Science in China, Series D Earth Sciences, 47, 822 –833. Madsen, D. B., Haizhou, M., Rhodes, D., Brantigham, P. J. & Forman, S. L. 2008. Age constraints on the late Quaternary evolution of Qinghai Lake, Tibetan Plateau. Quaternary Research, 69, 316 –325. Matmon, A., Simhai, O. et al. 2009. Desert pavementcoated surfaces in extreme deserts present the longest-lived landforms on Earth. Geological Society of America Bulletin, 121, 688 –697, doi: 10.1130/ B26422.1. Me´riaux, A.-S., Ryerson, F. J. et al. 2004. Rapid slip along the central Altyn Tagh Fault: Morphochronologic evidence from Cherchen He and Sulamu Tagh. Journal of Geophysical Research, 109, B06401. Me´riaux, A.-S., Tapponnier, P. et al. 2005. The Aksay segment of the Northern Altyn Tagh fault: tectonic geomorphology, landscape evolution and Holocene slip-rate. Journal of Geophysical Research, 110, B04404. Me´tivier, F., Gaudemer, Y., Tapponnier, P. & Meyer, B. 1998. Northeastward growth of the Tibet Plateau deduced from balanced reconstruction of two depositional areas: the Qaidam and Hexi Corridor basins, China. Tectonics, 17, 823 –842. Meyer, B., Tapponnier, P. et al. 1998. Crustal thickening in Gansu-Qinghai lithospheric mantle subduction and oblique, strike– slip controlled growth of the Tibet Plateau. Geophysical Journal International, 135, 1– 47. Nishiizumi, K., Imamura, M., Caffe, M., Southon, J., Finkel, R. & McAnich, J. 2007. Absolute calibration of 10Be AMS standards. Nuclear Instruments and Methods in Physics Research B, 258, 403– 413.

218

A. PERRINEAU ET AL.

Oskin, M., Peng, L., Shelef, E., Strane, M., Gurney, E., Singer, B. & Zhang, X. 2008. Elevated shear zone loading rate during an earthquake cluster in eastern California. Geology, 36, 507 –510, doi: 10.1130/G24814A.1. Pan, B. 1994. Research upon the geomorphologic evolution of the Guide Basin and the development of the Yellow River. Arid Land Geography, 7, 43– 50. Pan, B., Burbank, D., Wang, Y., Wu, G., Li, J. & Guan, Q. 2003. A 900 ka record of strath terrace formation during glacial-interglacial transitions in northwest China. Geology, 31, 957–960. Pan, B., Su, H., Hu, Z., Hu, X., Gao, H., Li, J. & Kirby, E. 2009. Evaluating the role of climate and tectonics during non-steady incision of the Yellow River: evidence from a 1.24 terrace record near Lanzhou, China. Quaternary Science Reviews, 28, 3281– 3290, doi: 10.1016/j.quascirev.2009.09.003. Pares, J. M., van der Voo, R., Downs, W. R., Yan, M. & Fang, X. 2003. Northeastward growth and uplift of the Tibetan Plateau: magnetostratigraphic insights from the Guide Basin. Journal of Geophysical Research, 108, doi: 10.1029/2001JB0012349. Peltzer, G. & Tapponnier, P. 1988. Formation and evolution of strike–slip faults, rifts, and basins during the India-Asia collision: an experimental approach. Journal of Geophysical Research, 93, 15 085– 15 117. Qin, B. 1992. The hydrological characteristics and water budget of Qinghai lake drainage basin, Chin. J. Oceanology and Limnology, 11, 314– 320. QINGHAI GEOLOGICAL BUREAU. 1988. Geological Map of Qinghai-Xizang. Ministry of Geology, Geological Press, Xining. Repka, J. L., Anderson, R. S. & Finkel, R. C. 1997. Cosmogenic dating of fluvial terraces, Freemont River, Utah. Earth and Planetary Science Letters, 152, 59–73. Rhode, D., Haiying, Z., Madsen, D. B., Xing, G., Brantingham, P. J., Ma, H. & Olsen, J. W. 2007. Epipaleolithic/early Neolithic settlements at Qinghai Lake, western China. Journal of Archaeological Science, 34, 600–612. Rhode, D., Haizhou, M., Madsen, D. B., Brantingham, P. J., Forman, S. L. & Olsen, J. W. 2010. Palaeoenvironmental and archaeological investigations at Qinghai Lake, western China: geomorphic and chronometric evidence of lake level history. Quaternary International, 218, 29–44. Royden, L., Burchfiel, B. C., King, R. W., Wang, E., Chen, Z. L., Shen, F. & Liu, Y. P. 1997. Surface deformation and lower crustal flow in eastern Tibet. Science, 276, 788– 790, doi: 10.1126/ science.276.5313.788. Ruszkiczay-Ru¨diger, Zs., Fodor, L., Bada, G., ¨ ssy, Sz., Horva´th, E. & Dunai, T. J. 2005a. Lee´l-O Quantification of Quaternary vertical movements in the central Pannonian Basin: a review of chronologic data along the Danube River, Hungary. Tectonophysics, 410, 157–172. Ruszkiczay-Ru¨diger, Zs., Dunai, T. J., Bada, G., Fodor, L. & Horva´th, E. 2005b. Middle to late Pleistocene uplift rate of the Hungarian Mountain Range at the Danube Bend (Pannonian Basin)

using in situ produced 3He. Tectonophysics, 410, 173–187. Schaller, M., von Blanckenburg, F., Veit, H. & Kubik, P. W. 2002. Influence of periglacial cover beds on in situ-produced cosmogenic 10Be in soil sections. Geomorphology, 49, 255– 267. Stone, J. O. 2000. Air pressure and cosmogenic isotopes production. Journal of Geophysical Research, 105, 23 753–23 759. Sun, J. 2005. Long-term fluvial archives in the Fen Wei Graben, central China, and their bearing on the tectonic history of the India– Asia collision system during the Quaternary. Quaternary Science Reviews, 24, 1279– 1286. Tapponnier, P., Peltzer, G., Armijo, R., Le Dain, A.-Y. & Cobbold, P. 1982. Propagating extrusion tectonics in Asia: new insights from simple experiments with plasticine. Geology, 10, 611– 616. Tapponnier, P., Meyer, B. et al. 1990. Active thrusting and folding in the Qi Lian Shan, and decoupling between the upper crust and mantle in northeastern Tibet. Earth and Planetary Science Letters, 97, 382–403. Tapponnier, P., Xu, Z. Q., Roger, F., Meyer, B., Arnaud, N., Wittlinger, G. & Yang, J. S. 2001. Oblique stepwise rise and growth of the Tibet Plateau. Science, 294, 1671–1677. Thompson, L. G., Yao, T. et al. 1997. Tropical climate instability: the last glacial cycle from a QinghaiTibetan ice core. Science, 276, 1821–1825. Vandenberghe, J. 2003. Climate forcing of fluvial system development: an evolution of ideas. Quaternary Science Reviews, 22, 2053–2060. Van der Woerd, J. 1998. Couplage cine´matique entre de´crochements et chevauchements actifs dans le nord du Tibet. Croissance du Plateau Tibe´tain. PhD thesis, University Paris VII. Van der Woerd, J., Klinger, Y., Sieh, K., Tapponnier, P., Ryerson, F. J. & Me´riaux, A.-S. 2006. Long-term slip rate of the southern San Andreas Fault from 10 Be– 26Al surface exposure dating of an offset alluvial fan. Journal of Geophysical Research, 111, B04407, doi: 10.1029/2004JB003559. Van der Woerd, J., Tapponnier, P. et al. 2002. Uniform postglacial slip-rate along the central 600 km of the Kunlun Fault (Tibet), from 26Al, 10Be and 14C dating of riser offsets, and climatic origin of the regional morphology. Geophysical Journal International, 148, 356–388. Van der Woerd, J., Xu, X. et al. 2001. Rapid active thrusting along the northwestern range front of the Tanghe Nan Shan (western Gansu, China). Journal of Geophysical Research. 106, 30 475– 30 504. Wang, P., Jiang, H., Yuan, D., Liu, X. & Zhang, B. 2010. Optically stimulated luminescence dating of sediments from the Yellow River terraces in Lanzhou: tectonic and climatic implications. Quaternary Geochronology, 5, 181–186. Wang, S., Wu, X., Zhang, Z., Jiang, F., Xue, B., Tong, G. & Tian, G. 2002. Sedimentary records of environmental evolution in the Sanmen Lake Basin and the Yellow River running through the Sanmenxia Gorges eastward into the sea. Science in China (Series D), 47, 596–608.

INCISION RATE OF YELLOW RIVER IN NE TIBET Yan, M., Pare´s, J. M., van der Voo, R., Rea, D. K. & Fan, X. 2004. Pleistocene magnetostratigraphy of the Gonghe Basin, NE Tibetan Plateau: headward incision of the Yellow River after 1.8 Ma. American Geophysical Union, spring meeting 2004, abstracdt GP31A-17. Yan, P., Shi, P., Gao, S., Chen, L., Zhang, X. & Bai, L. 2002. 137Cs dating of lacustrine sediments and human impacts on Dalian Lake, Qinghai Province, China. Catena, 47, 91–99. Yuan, B. Y., Chen, K. Z., Bowler, J. M. & Je, S. S. 1990. The formation and evolution of Qinghai Lake. Quaternary Sciences, 3, 233–243 (in Chinese with English abstract). Zhang, C., Gong, J., Zou, X., Dong, G., Li, X., Dong, Z. & Qing, Z. 2003. Estimates of soil movement in a study area in Gonghe Basin, north-east of

219

Qinghai-Tibet Plateau. Journal of Arid Environments, 53, 285 –295. Zhang, Z., Wang, S., Wu, X., Jiang, F. & Li, X. 2000. Age and environmental implications of the Sanmen Basin piercing in the middle reaches of Yellow River. Journal of Geosciences of China, 2, 42–44. Zhang, Z., Wang, S., Yang, X., Jiang, F., Shen, J. & Li, X. 2004. Evidence of a geological event and environmental change in the catchment area of the Yellow River at 0.15 Ma. Quaternary International, 117, 35–40. Zheng, D., Zhang, P., Wan, J., Li, C. & Cao, J. 2003. Late Cenozoic deformation subsequence in northeastern margin of Tibet: Detrital AFT records from Linxia Basin. Science in China, Series D, 46, 266– 275.

Neotectonic constraints on the Gaxun Nur inland basin in north – central China, derived from remote sensing, geomorphology and geophysical analyses ¨ LZ3, ¨ NNEMANN1,2, SEBASTIAN HO KAI HARTMANN1*, BERND WU 3 4 ANNA KRAETSCHELL & HUCAI ZHANG 1

Freie Universitaet Berlin, Department of Earth Sciences, Berlin, Germany

2

Nanjing University, School of Geographic and Oceanographic Sciences, Nanjing, China

3

Universitaet Kiel, Leibniz-Institut fu¨r Meereswissenschaften, IFM-GEOMAR, Kiel, Germany 4

State Key Laboratory of Lake Science and Environment, Nanjing Institute of Geography and Limnology, Chinese Academy of Sciences, Nanjing, China *Corresponding author (e-mail: [email protected]) Abstract: The endorheic Gaxun Nur Basin (GNB, local name: Ejina basin), which is located north of the Tibetan Plateau between the tectonic stress fields of the Qilian Shan and the Gobi-Tienshan, has evolved as a large inland basin filled with deltaic sediments during the past 250 ka. Here we present selected examples of geomorphological, sedimentological and geophysical evidence of tectonic activity and discuss a possible time frame of selected occurrences. We used medium-scale geomorphological mapping supported by analyses of Landsat ETM images, Corona images and an Aster Digital Terrain Model (Aster-DTM), combined with field surveys, dated sediment sections, and geophysical investigations using electromagnetic methods. The spatio-temporal distribution of radiocarbon-dated lake sediments within the northern GNB indicates a non-even distribution of neotectonic activity with west– east increasing amplitude of subsidence rates from 0.8– 1.1 m/ka in the western part and more than three times higher rates in the eastern part. Our data indicate that tectonic has strongly amplified climate-induced environmental changes and may be regarded as an example of non-climatic pulses affecting lake-hydrology and basin development.

Palaeoclimate research in China has often been conducted under the assumption that climates across Asia were influenced by the uplift of the Tibetan Plateau (An et al. 2001), co-controlling the evolution of the Asian monsoon system and the establishment of the Inner Asian dryland belt. Apart from intensively investigated loess profiles along the Chinese Loess Plateau (e.g. Kukla & An 1989; Porter & An 1995; Xiao et al. 1995; An et al. 2001), lake and basin studies on the Tibetan Plateau and adjacent areas have become an important issue, as they provide higher resolution climate-controlled hydrological records of local to regional dimension with feedbacks to global signatures (e.g. Gasse et al. 1996; Herzschuh 2006; Mischke et al. 2010). The major outcomes of nearly all studies indicate highly variable climate fluctuations throughout the last 125 ka BP (e.g. Wu¨nnemann et al. 2007a, b), regionally influenced by the Asian monsoon dynamics and also by westerly airflow over China, following the general trend of the northern

hemispheric insolation pattern (e.g. Herzschuh 2006; Chen et al. 2008). Many authors (e.g. Chen et al. 1995, 2003, 2008; Yu et al. 2002, 2006; Zhang et al. 2004), have interpreted Late Pleistocene and Holocene signals in proxy records from lake sediments as direct responses to climate variability with remarkable differences in timing, frequency and spatial distribution (e.g. An 2000; Holmes et al. 2009; Mischke et al. 2010). However, non-climatic factors such as tectonic activity or local catchment dynamics have not yet been considered in most of the past publications, although they may have influenced many records more intensively than previously assumed. For example, Wu¨nnemann et al. (2010) demonstrate that the water balance in the Tso Kar basin, Ladakh, Indian Himalaya, was strongly influenced by tectonic movement and local catchment processes that partly blanketed climate-induced hydrological signals. Furthermore, seismic surveys in the Donggi Cona lake basin, northeastern Tibetan Plateau,

From: Gloaguen, R. & Ratschbacher, L. (eds) Growth and Collapse of the Tibetan Plateau. Geological Society, London, Special Publications, 353, 221–233. DOI: 10.1144/SP353.11 0305-8719/11/$15.00 # The Geological Society of London 2011.

222

K. HARTMANN ET AL.

reveal major tectonic activity on the basin morphology and the hydrological system throughout the Late Pleistocene (Dietze et al. 2009). Consequently, sedimentary records from these sites display a distinct interplay between non-climatic pulses (e.g. tectonics) and climate-triggered landforming processes that affected the depositional environment significantly and thus led to a more substantiated interpretation of lake records (see Wu¨nnemann et al. 2008, 2010). At present, only few researchers take into account that many lake archives on the Tibetan Plateau and adjacent areas were developed under tectonic activities along strike –slip faults (pull apart basins, e.g. Van der Woerd et al. 2002) or in lakes that were naturally dammed (Korup & Montgomery 2008). Tectonic activities may substantially change catchment characteristics within a hydrological system (e.g. catchment size, fluvial pattern, topographic gradients). Thus, geomorphological investigation within the entire catchment of a specific research area is an essential part of any reliable palaeoenvironmental model. Consequently, we integrated various research results on tectonic and geomorphic features, sedimentary sequences, and fluvio-lacustrine

dynamics of the Gaxun Nur Basin, to demonstrate that particularly geomorphological and sedimentary features provide evidence of tectonic activities that still influence the local environmental evolution of the Gaxun Nur basin. Most of the presented results cannot be regarded as stand-alone evidence of tectonically induced dislocation and thus require multiple approaches by cross-site interpretation and geophysical methods in order to test the hypothesis of non-random occurrences.

Study area The arid Gaxun Nur Basin extends from c. 99.58 to 1028E and 40.58 to 42.58N north of the Tibetan Plateau and borders the southern slopes of Gobi-Tienshan Mountains (Fig. 1). Its catchment comprises 130 000 km2. The plains are mainly composed of gravels (gravel-type surfaces: gobi) covering up to 300 m thick lacustrine sediments (Zhang et al. 2006). The asymmetric delta, stretching over c. 300 km from the northern Qilian Shan mountain rim to the north, can be divided into two parts: an inactive eastern part including the terminal lake Juyanze

Fig. 1. Satellite image (Landsat ETM RGB 7-4-2) of the Gaxun Nur Basin and adjacent areas north of the Tibetan Plateau with location of major structural elements (red lines correspond to undifferentiated faults).

NEOTECTONIC OF GAXUN NUR BASIN

and a still active western part with the lakes Gaxun Nur and Sogo Nur at its outer edges (Wu¨nnemann & Hartmann 2002, Fig. 2). Only the latter shows a clear cone-like (convex) shape in the digital elevation model (Aster-DTM), indicating young fluvial accumulation by rivers originating in the Qilian Shan Mountains. Conversely, the eastern part is characterized by a more concave shape, which underlines currently inactive fluvial processes. The crescent-shaped series of terminal lakes at the distal part of the inland delta constitutes the modern erosion base of the Hei River. The latter originates in the Qilian Shan Mountains at the northern boundary of the Tibetan Plateau. The presently inactive drainage pattern from south to NE testifies to former fluvial input along the Hei River. Further palaeo-channels enter the basin from the western Beishan basement and from the Gobi-Tienshan Mountains in the north (Wu¨nnemann & Hartmann 2002, Fig. 1). The average elevation of the Gaxun Nur Basin is between 900 –1300 m a.s.l. At present, only the smallest lake, Sogo Nur, north of the Ejina river oasis sporadically receives water during the summer season, whereas Lake Juyanze in the east has remained dry since the Ming Dynasty some 600 years ago (Tan 1996). The largest lake, Gaxun Nur, dried up in the 1960s owing to intensified water withdrawal in the middle reaches of the Hei river system (Wu¨nnemann 1999). The Gaxun Nur Basin is part of the Alashan (Alxa) Plateau

223

between the Tibetan Plateau (TPL) and the GobiTienshan and is one of the world’s largest inland deltas. Large strike–slip fault systems of the Gobi-Tienshan-, Altyn-Tagh- and Haiyuan-fault systems partly appear as thrust faults due to the Indian collision with Asia and its northward propagation (Molnar & Tapponier 1975; Vincent & Allen 1999). They border the endorheic drainage basin. Although modern tectonic activity in the vicinity of the basin has been reported in detail (e.g. Cunningham et al. 1996, 1997; Lamb et al. 1999; Li et al. 1999; Lasserre et al. 2002; Hetzel et al. 2002; Ding et al. 2004), the basin itself has been regarded as a relative stable block between the tectonic stress fields.

Methods Geomorphological field surveys were accompanied by the use of remote-sensed data such as Landsat ETMþ, panchromatic Corona Images and digital elevation data from Aster-DTM and Landsat data. They were pre-processed with a high-pass and edge enhancement convolution filter and pansharpened by the use of the ETMþ panchromatic band. Multispectral contrast enhancement by PCA and mineral composites (ferrous minerals, bands 5/4, iron oxide, bands 3/1, clay minerals, bands 5/7; after Dogan 2009) was applied for selected areas.

Fig.2. left: Landsat image showing main features of Gaxun Nur Basin, including faults (undifferentiated, confirmed by geomorphological and/or geophysical investigation) and key sites (numbered dots); right: digital elevation model (Aster-DTM) of the entire Gaxun Nur Basin.

224

K. HARTMANN ET AL.

Differential GPS (standard error: 0.3 m) was used for accurate elevation measurements of morphological features and study sites. Geometric correction of Corona images (Kh4b, acq. 1964, c. 1.8  1.8 m spatial resolution) was performed by using ground control points derived from DGPS. Radiometric dating (AMS) on bulk samples from sections and cores was done at Leibniz Laboratory, Kiel, Germany, and Beta Analytics, Miami, USA. In several cases, twin dating of the organic carbon from the remaining lye fraction and humic acid was performed to estimate contamination within one sample. Radiocarbon ages were calibrated using OxCal 4.1 with Intcal04 (Reimer et al. 2004). Ages older than 18 ka BP were calibrated according to Hughen et al. (2004). Additionally, thermoluminescence dating (IRSL, OSL) of sandy lake sediments was performed at the Geophysical Laboratories of the Chinese Academy of Science, Beijing, China, and at the Saxonian Academy of Sciences, Freiberg, Germany. Both laboratories used feldspar grains for dating. Further chronological information refers to palaeomagnetic measurements, which has been reported by Wu¨nnemann et al. (2007a) in detail. For geophysical investigations with the transient electromagnetic method (TEM) we used a LAPTEM device (Geoinstruments) in the coincident loop configuration with a loop area of ˚. 10 000 m2 and a transmitter current of c. 7 A TEM is an active electromagnetic method using a transmitter coil to excite a static magnetic field. After turn-off of the primary field, the time derivative of the secondary induced magnetic field, which is caused by induced electric currents in the underground, is measured with a receiver coil. Transients were recorded in the time range between 100 ms and 100 ms yielding information about the conductivity structure of the subsurface in a depth range between c. 25–300 m.

Results and discussion Tectonic structures derived from satellite imagery and sections Eastern basin. A division of the Gaxun Nur Basin into two parts is supported by Aster-DTM analysis. The western part resambles a classical conical fan, whereas east of the Hei River, basin morphology changes to a concave shape with northward feathering of gravel-covered ridges. We may therefore deduce that the western part is still an active alluvial fan, whereas the eastern part appears to be inactive with respect to fluvial accumulation by the Hei River. Especially the northern Juyanze sub-basin was analysed in detail (see Mischke 2001;

Wu¨nnemann & Hartmann 2002). This basin is confined by a basement block in the north (cf. Becken et al. 2007), the alluvial fan of Hei River in the west, a north –south-striking, spit-like escarpment in the east (cf. Hartmann & Wu¨nnemann 2009) as well as up to 20 m high, feather-like gravel ridges in the south. The SW –NE-striking sinuous gravel ridges are connected to the gravel plain in the south and terminate at a diffuse ESE – WNW-striking lineament (Fig. 1). The ancient fortress Khara Khoto is situated on top of such a ridge (Fig. 2). The gravels of fluvial origin do not exceed 3 m in thickness and cover up to 20 m thick sediments of lacustrine and aeolian origin. A simplified stratigraphy similar to the uppermost part of D100 is given in Figure 3. Sedimentological investigations suggest an interchanging environment between aeolian and lacustrine to semilacustrine deposition. The well-rounded gravels show a petrography similar to the bedrock north of Juyanze. Thus, the gravels were transported by streams indicating a much higher topographic gradient than at present. At present the area under consideration is extremely flat. Any fluvial activity would result in low velocity flow, which only can transport silt and sand. Hence, gravel deposition must have occurred at a time of high energy transport and thus steeper gradient. Later on, the gravels within the inactive channels protected the underlying sediment from erosion and promoted the development of terrace-like features. Similar inverted channels indicating both hydrological changes and tectonic movement have been reported, for example, from Oman by Maizels (1990). In ancient times until the Ming Dynasty (13th– 14th century, Tan 1996), a formerly extended Lake Juyanze formed the steep slopes by littoral processes. Dating results on ostrich egg shells found in the uppermost lake sediment suggest an age of c. 37 –42 ka BP (Bergman et al. 1980). Therefore, phases of incision may have occurred since Marine Isotope Stage 3 (MIS 3). A similar stratigraphy can be found in a 13 m high cliff section some 20 km further NE, which is attached to a south– north-striking ridge in continuation of the tectonically induced Gurinai depression (for details see Ho¨lz et al. 2007). The c. 25 km long scarp divides Juyanze basin (up to 925 m a.s.l.) into a western (c. 350 km2) and an eastern lake basin (c. 320 km2) with 12 fossil beach ridges on its western side (Fig. 4). In the middle part of the scarp a fossil cliff was formed, probably during the last documented high lake level some 2000 years ago (Han dynasty: 206 bc to 220 ad, Tan 1996). The c. 13 m high cliff section (HC8, for location see Fig. 1) consists of several layers of unconsolidated aeolian sand and lacustrine clay. The upper 2.40 m thick clay layer dates to

NEOTECTONIC OF GAXUN NUR BASIN

225

Fig. 3. Corona image of a gravel-covered ridge 2 km SE of Kara Khoto. Gravels cover lacustrine and aeolian sediments (cross-section). These gravel ridges indicates at least 2 periods of tectonic subsidence with lowering of the erosional base (Hartmann 2003).

18 060 + 2000 a bp. This age supports the chronological order of the overlying 2.60 m thick layer of fluvial gravels reported by Wu¨nnemann & Hartmann (2002) and provides a maximum age of widespread gravel accumulation in the Gaxun Nur basin. Older gravel sequences of Quaternary age have not yet been found. The surface of the cliff inclines approximately 3‰ to the east, where six former beaches indicate higher lake levels and the extent of the eastern Juyanze palaeolake. As visible in the Corona image in Figure 4, the upper three beaches have been cut by local runoff. A short geophysical profile using TEM (transient electromagnetic method) provides further evidence of the tectonic origin of the structure (Fig. 4a). The four stations west of the structure (Fig. 4c, stations 1–4) show a consistent layering with an eastward dipping lower layer interface at depths between 125 m to 165 m beneath stations 1 and 4, respectively. The inclination of c. 68 apparent in the 1D model should be close to the true dip, since 2D/3D effects may be neglected for such weakly inclined layers (see Danielsen et al. 2003). The top layer above this interface is characterized by relatively low resistivity around 6–7 Vm, whereas the bottom layer exhibits resistivity above 15 Vm. The same general setting, that is, a less resistive top layer and more resistive bottom layer, is also interpreted at stations 5 and 6 east of the cliff front. Here, the inverted resistivity of the top layers is less constrained by the measured data, but generally seems to be lower (1.5– 7 Vm). The resistivity of the lower layer (15 –20 Vm) is consistent with the easterly stations 1– 4. In accordance

with the surface topography, a step up of c. 50 m is also evident for the lower layer interface between stations 4 and 5 (Fig. 4c). Generally, the resolution of the performed TEM measurements roughly starts at depths above 30 m. Therefore, the measurements are not suitable to trace faults up to the surface. However, the occurrence of isolated fluvial gravels of late Pleistocene–Holocene age is indicative of a relatively young age for the normal fault. The cross-section in Figure 5 shows the situation between site HC8 and G36 (for location see Fig. 2). Radiocarbon ages and geophysical investigations indicate the onset of subsidence between 18 – 10.7 cal. ka BP. The minimum vertical distance between the reconstructed fan surface at the cliff position and bottom of site G36 accounts for 65 m, taking the present slope gradient of c. 2‰ into consideration (Fig. 5), we then can calculate a minimum rate of c. 3.6 m/ka within the Juyanze sub-basin. This seems to be more than three times higher than the average subsidence rate of the whole Gaxun Nur basin derived from D100 core. However, the close proximity of the Juyanze subbasin to the active duplex structure of the Gurinai basin (Fig. 2) and its location in a mechanical hinge between west –east and north–south striking fault systems (Becken et al. 2007; Ho¨lz et al. 2007) may explain much higher subsidence rates. Southeastern basin. The most impressive feature of tectonic origin is located at the eastern margin of the Gaxun Nur Basin as a straight, north– south striking double lineation of which the western part

226

K. HARTMANN ET AL.

(a)

(b)

18 059 cal a BP (15 060 ± 80 a BP)

(c)

Fig. 4. (a) Corona image of site HC8 with clearly visible beach ridges east and west of the escarpment which have been measured by differential GPS. (b) Lake sediments from the uppermost part (below gravel cover) suggest an age of c. 18 ka BP. C) 1D inversion results of TEM measurements along the profile shown in (a). At each location the model with the best fit (black line) to the measured data (not displayed) and equivalent models (grey lines), which still explain measured data with an acceptable misfit, are displayed. Models are displayed continuously up to the true surface; however, it should be noted that no resolution is given in the upper c. 25 m and below c. 250 m (depth relative to stations) by the measured data.

shows a fish-bone pattern with a horsetail at its southern end (Fig. 2, no. 1). This implies releasing (extensional, cf. Twiss & Moores 1992) bends of a duplex structure between two parallel, left-lateral strike–slip faults. The opposite strike direction of bends at the western side of the western lineation appears to be restraining (contractional, cf. Twiss & Moores 1992) as a result of left lateral displacement implying two phases of formation with an opposite tension field (Woodcock & Fischer 1986;

Twiss & Moores 1992). Geological maps (Gansu Provincial Geological Bureau 1980) suggest that the restraining bends occur in mid –Late Pleistocene gravels of the Gaxun Nur Basin, whereas the releasing bends cut through Holocene sediments. Thus, a contractional displacement occurred after deposition of Late Pleistocene pavement gravels, whereas a second extensional displacement occurred during Holocene time. Their position away from local runoff indicates a long-term preservation of

NEOTECTONIC OF GAXUN NUR BASIN

227

Cross section Eastern Juyanze [m asl.]

1000 Cliff-section HC8 18 059 cal a BP

reconstructed

fan surface

> 65m

950 section G36 10 771 cal a BP

900 5000

10,000

15,000

20,000

[m] rel. distance

Fig. 5. Cross-section through the eastern Juyanze sub-basin connecting the cliff section HC8 and site G36 (cf. Hartmann & Wu¨nnemann 2009) to the eastern basement. The former fan surface has been reconstructed by bootstrap random profiling and interpolation of 24 profiles.

gravel-covered features (Hetzel et al. 2002). This structural interpretation is reinforced by multispectral analyses of vegetation cover and soil moisture. The occurrence of relatively dense vegetation and high soil moisture east of the lineation is in accordance with a slight surface dip to west, both derived from Aster-DTM (Fig. 2, no. 2). Several salt pans scatter within the inner part of this structure and indicate near-surface groundwater flow. The visually identified structures and related hydrological conditions have been confirmed by MT measurements (Ho¨lz et al. 2007) and display the occurrence of aquifer modification nearly without remarkable vertical surface deformation (Fig. 2a, b). As suggested by Geyh et al. (1996) the Gurinai structure appears to be a groundwater-shed between the Hei River aquifer and the Badain Jaran aquifer (Ho¨lz et al. 2007). Northwestern basin. A further tectonically induced feature appears at the northern margin of Gaxun Nur Basin between the dry lake basins of Gaxun Nur and Sogo Nur (Fig. 6). A large number of almost parallel channels are cut by a c. 45 km long, NW–SE-striking line (Fig. 2, no. 3), where all channels terminate without any remarkable offset. Another traceable lineation occurs along a north– south striking scarp (Fig. 2, no. 4) which consists of more than 13 m thick, stratified fluvial gravels dipping 38 to the south (Fig. 6b). At the interface of both structures, a c. 40 m wide valley outlet (Fig. 6) towards Gaxun Nur provides exposed sections of gravels at the southern slope whereas fine-grained sediments containing abundant gypsum crystals of pre-Quaternary age (Bergman et al. 1980) occur on the opposite channel cut. The latter can be traced for several kilometres eastwards, forming a cliff near Sogo Nur palaeolake. The co-occurrence of pre-Quaternary finegrained sediments and Late Quaternary gravels at the same altitudinal level indicates that a

minimum vertical offset of 7 m must have taken place after the deposition of the Late Quaternary gravel layers, taking into account that the gravels on top of the pre-Quaternary sediments are identical with those at the opposite side of the channel (Fig. 6b). The modern topography, does not provide enough gradient for transportation of such coarse-grained material, which consequently requires additional relative offset towards the erosion base in the southern area. A comparable displacement has been observed at site D122 (Fig. 6), where thick gravel layers are distorted and incline against the normal river bed topography (Wu¨nnemann 1999). This may indicate that the tectonic line (see Fig. 6, no. 1) continues for c. 2.2 km to NW and may have caused the strong diversion of the respective river mouth (see Fig. 6, no. 2). Beside a set of left lateral river offsets along the western margin of Gaxun Nur Basin (Fig. 2, no. 5), the shape of the Gaxun Nur lake appears as a set of rhombic sub-basins, whose major axis is oriented at an angle of nearly 1158 to N (cf. Fig. 1, no. 1), just like the drainage system cutting the lineation mentioned above. The southern margins of both basins are formed by two normal faults with an offset of more than 10 –15 m, morphologically modified as cliffs with up to three beach ridges. Figure 7 shows several linear features within the lake basin, running mainly parallel to the margins. The straight continuation of the fault constitutes the southeastern margins of Gaxun Nur lake with nearly the same offset as the other two normal faults. The visible bar pattern between both sub-basins occurs morphologically with an offset of 1– 5 m (derived from Aster-DTM). Inferred from its geometric shape and left-lateral offsets, the Gaxun Nur lake can be classified as a pull-apart basin. To estimate the age of offset at Gaxun Nur, two key sections are linked in a cross section (Fig. 7b). The first section (HC 21) is located at the outlet of

228

K. HARTMANN ET AL.

Fig. 6. (a) Corona image of northern Gaxun Nur showing cut-off of a wadi system at southern slopes of Gobi-Tienshan range; dotted lines indicate faults. (b) Section I45 (sketch) shows the contact zone between pre-Quaternary sediments (left) and Late Quaternary fluvial gravels (right).

NEOTECTONIC OF GAXUN NUR BASIN

229

Fig. 7. Corona image of Gaxun Nur palaeolake interpreted as a pull-apart structure within a left lateral strike slip system. The cross section displays the morphology between profiles HC21 and HC115 (Hartmann 2003).

a dry channel, cutting the cliff down to 900 m a.s.l. The lower part consists of greenish grey clay up to 905 m a.s.l covered by 3 m thick, well-rounded beach gravels and intercalated sand layers. A 60 cm thick layer of lake carbonates covers the lower gravel sequence. Both sequences are preserved by a 3.5 m thick gravel layer, morphologically formed as a beach ridge. The lake carbonate has been dated to 25 700 cal. a bp, assuming a maximum age due to unknown reservoir effects. Comparing the profile with other sites in the Gaxun Nur lake basin, the age of lake carbonate seems to be reliable within the error of age estimation. The deposition of lake carbonate indicates a freshwater lake at the end of marine isotope stage (MIS) 3. The overlying gravels were deposited afterwards, even covering lake sediments in the Gaxun Nur lake basin (HC115), assuming a considerable lake level decline during the LGM (cf. Wu¨nnemann et al. 2007b). However, the 2.40 m long section HC115 only provides salty, fine-grained lake sediments dated

to between 21 502 cal. a bp (bottom part) and 4746 + 115 and 3373 + 40 cal. a bp (upper part). The latter two ages are in reverse stratigraphic order. Gravels of fluvial origin in a comparable stratigraphic order to those in section HC21 could not be found on the basin floor. As the gravels in the uppermost part of HC21 can also be found in the northern vicinity of Gaxun Nur, with similar grain size composition and petrography, we assume that they may also occur in the basin but were covered by thick younger lake deposits. Hence we may estimate a subsidence of the basin by at least 21 m. Erosion of the gravel beds within the lake basin can be excluded owing to the lack of any appropriate gradient for fluvial activity. Despite possible erosion/deflation of ostracodbearing lake carbonates, the age of gravel deposition might be much younger than 25.7 cal. ka bp. The lowermost date of section HC115 indicates that the gravel layers cannot be older than 21.5 cal. ka bp. On the basis of this assumption we calculated a subsidence rate of .0.81 m/ka.

230

K. HARTMANN ET AL.

As any evidence for a high lake level after c. 20 cal. ka bp, which would have prevented the basin from coarse infill, can also be ruled out (Wu¨nnemann et al. 2007a, b), we conclude that young tectonic movement may have led to the sedimentary and morphological pattern described above. On the basis of our age model we assume that considerable displacements may have happened during the Holocene.

Spatio-temporal distribution of tectonic activity According to Wu¨nnemann et al. (2007b), the onset of the Quaternary basin fill may have started c. 250 ka ago with a traceable lake development at drilling site D100 (Fig. 2). The basal sediments from 228.8 –230 m depth consist of pre-Quaternary (Gansu Geological Bureau 1980) red conglomerates in a clayey matrix. From 228.8 m to c. 71 m depth, laminated silt and clay indicate a continuous subaqueous sedimentation. The respective agedepth model suggests that the core reached the

radiocarbon dating limit (c. 50 ka) at around 90 m depth. From approx. 71 m upward, lake sediment interchanged with fluvial and aeolian sediments up to c. 7 m below the surface (at 940 m a.s.l). Ages from lacustrine sediments just below the gravel cover suggest that the gravel deposits are younger than 17 cal. ka bp (see Fig. 8, column C). Based on the assumption that the lake remained on a relatively low level during that time (,20 m water depth), the calculated mean sedimentation rate of c. 1.1 m/ka is nearly in the order of subsidence. However, uncertainties with respect to an exact chronological reconstruction of depositional phases arise from reverse AMS dates and their strong differences to OSL dates (Zhang et al. 2006). High-resolution remote sensing analyses reveal that the drill site was located at a previously unknown active fault that triggered sedimentary reworking processes during possible tectonic events. Hence, our chronology of the core remains tentative and thus only can provide a general time frame within the error of dating results, although it seems to be evident that the gravel layer in the

Fig. 8. Radiocarbon ages and generalized stratigraphy of selected sediment records within the northern Gaxun Nur Basin (A: Wu¨nnemann 1999; C: Wu¨nnemann et al. 2007b; E, F: Mischke 2001; D, G: Wu¨nnemann & Hartmann 2002; B, H: Hartmann 2003; I: Hartmann & Wu¨nnemann 2009).

NEOTECTONIC OF GAXUN NUR BASIN

upper part of the core is rather young (Wu¨nnemann et al. 2007a). Radiocarbon ages of lacustrine deposits within the northern part of the Gaxun Nur Basin in a nearly west –east transect are shown in Figure 2. Yardang profile 96E27 (Fig. 8, column A) is located SW of Gaxun Nur (Fig. 1), close to the basin border. It shows a low sedimentation rate during MIS 3 and 2 for at least 20 to 40 ka bp (Zhang et al. 2006; Wu¨nnemann et al. 2007a). Holocene ages in column B of Figure 2 refer to a cliff profile at the southern shore of Gaxun Nur Lake (HC21) and to the 2 m long section HC115. Both sediment records are encompassed by MIS 2 ages from the uppermost part of cliff profile HC21 (ostracoda sample 30 cm below the gravel cover, cf. Fig. 7) and the lowermost part of HC115 (Hartmann 2003). The geomorphological situation is described in Figure 7 and discussed above. Ages from lake sediments at Sogo Nur (Fig. 8, column D) are comparable with the former ones. The two uppermost ages of MIS 3 belong to a fossil cliff section at the north-western edge of Sogo Nur. Drilling core G25 was taken from the centre of the lake. It shows a similar stratigraphic situation as at site HC115. Another short core of Sogo Nur (Fig. 8, column E) from Mischke (2001) indicates a consistent stratigraphy for the last 2000 years. Columns F and G (Juyanze North, Mischke et al. 2005 and Jingshutou, Wu¨nnemann & Hartmann 2002) refer to closely related sections. Mischke (2001) reported a set of 6 profiles, adjacent to each other. They show inconsistent stratigraphies by multiple reverse ages within the profiles. The same picture occurs some hundred metres east at a site called Jingshutou (Fig. 2, No. 6) where ages are in reverse order throughout the Holocene. The locally distorted sediment structures there have been interpreted as seismites, developed close to a geophysically (MT and TEM) investigated graben structure (Becken et al. 2007). The last two columns are affiliated with 3 sections located in the heart-shaped palaeolake Juyanze. Profile G36 (ages in Column I) plays a key role for reconstruction of regional Holocene climate evolution (Hartmann & Wu¨nnemann 2009; Herzschuh et al. 2004). Details of these sections are discussed above (see Fig. 4). Obviously, the whole transect shows an overall stratigraphic disorder in two respects. First, sediments of ages older than Holocene are concentrated in the western part of the basin, whereas MIS 3 ages are absent SE of Sogo Nur. Second, except for bottom ages in cores taken from the centre of the three palaeolakes, Holocene lacustrine sediments are located at lower elevations than older ones. These findings contradict a spatially random water supply at the distal part of an alluvial fan, as a visual inspection of the Gaxun Nur fan using the

231

respective Landsat image (Fig. 2) indicates. Except for the dating of section HC8, which is situated between the western and eastern Juyanze Lake, all dates older than LGM occur in the western and northern part of the basin, whereas sediment records of the three currently dry lake basins Gaxun Nur, Sogo Nur and Juyanze date to the Holocene. Because of the observation that all investigated sections do not provide sediments of LGM age, we assume a strong sedimentary gap between Holocene and MIS3-dated strata. Owing to the endorheic situation and shape of the basin’s proximal part, fluvial erosion and considerable deflation can be excluded to explain this gap. Hence, other processes such as tectonic movement and dislocation of geomorphic features might have been responsible.

Conclusion The presented geomorphological features are considered to be important sites for the recognition and interpretation of tectonic effects on a previously poorly studied basin structure north of the Tibetan Plateau. Dated palaeolake records and their spatial distribution provide a general trend of asymmetric fan development through time. According to previous results and the presented interrelationship between fluvio-lacustrine dynamics and morphological distortions through space and time, we may estimate mean subsidence rates of 1 m per ka, which may vary locally. At the northwestern margin of the basin, the presence of Lake Gaxun Nur suggests a local subsidence rate of this pullapart structure in the order of 0.8 m per ka. On the far eastern side of the intramontane basin basin, geomorphological analyses of palaeolake Juyanze suggest a roughly three to four times higher subsidence rate in the order of 3.6 m per ka. With respect to the low sample resolution as well as possible spatial autocorrelation errors, the results show a non-even distribution of neotectonic activity with a west –east increasing amplitude of strain rates for the last c. 25 ka. Furthermore, structural analyses of tectonic features may explain multiple age inversions within key sites in north-western China. Thus, palaeoclimatology and neotectonics seem to be inseparable in two ways for areas of active movement. † First, consideration of tectonically induced morphology within a given catchment is required for a reliable interpretation of palaeohydrological records. † Second, disturbances in dated palaeohydrological records and geomorphology together provide the chronological frame for the determination of active faulting.

232

K. HARTMANN ET AL.

However, further efforts are needed to obtain a satisfactory reconstruction of the spatio-temporal distribution pattern of tectonic movement based on reliable chronological control and geophysical measurements. This study was supported by DFG (Pa131 16/1 –4) and NSFC (40871096 & 40371117). Numerous researchers from TU Berlin, FU Berlin, University of Bayreuth, TU Freiberg, Leibniz Laboratory Kiel, Lanzhou University as well different institutions of CAS contributed to this long-lasting study. Special thanks to M. Becken and M. Hu¨ls for discussion of the results.

References An, Z. 2000. The history and variability of the East Asian paleomonsoon climate. Quaternary Science Reviews, 19, 171–187. An, Z., Kutzbach, J. E., Prell, W. & Porter, S. C. 2001. Evolution of Asian monsoons and phased uplift of the Himalaya–Tibetan plateau since Late Miocene times. Nature, 411, 62–66. Becken, M., Ho¨lz, S., Fiedler-Volmer, R., Hartmann, K., Wuennemann, B. & Burkhardt, H. 2007. Electrical resistivity image of the Jingsutu Graben at the NE margin of the Ejina Basin (NW China) and implications for the basin development. Geophysical Research Letters, 34, L09315. Bergman, F., Bexell, G., Bohlin, B., Chen, P. C., Hedin, S., Ho¨rner, N. G. & Norin, E. 1980. Sven Hedin Central Asia Atlas, memoir on Maps: The Kansu-Hohsi Corridor and the Suloho-Ochinaho drainage regions. Reports from the scientific expedition to the north-western provinces of China under the leadership of Dr. Sven Hedin – The Sino-Swedish Expedition-, 50. Etnografiska Museet, Stockholm. Chen, F., Wang, S. et al. 1995. Paleomagnetic record from RH lacustrine core in Zoige Basin of the Tibetan Plateau. Science in China, Series B, 38, 1513–1521. Chen, F. H., Wu, W., Holmes, J. A., Madsen, D. B., Zhu, Y., Jin, M. & Oviatt, C. G. 2003. A mid-Holocene drought interval as evidenced by lake desiccation in the Alashan Plateau, Inner Mongolia, China. Chinese Science Bulletin, 48, 1401–1410. Chen, F., Yu, Z. et al. 2008. Holocene moisture evolution in arid central Asia and its out-of-phase relationship with Asian monsoon history. Quaternary Science Reviews, 27, 351– 364. Cunningham, W. D., Windley, B. F., Dorjnamjaa, D., Badamgarov, J. & Saandar, M. 1996. Late Cenozoic transpression in southwestern Mongolia and the Gobi Altai-Tien Shan connection. Earth and Planetary Science Letters, 140, 67– 81. Cunningham, W. D., Windley, B. F., Owen, L. A., Barry, T., Dorjnamjaa, D. & Badamgarav, J. 1997. Geometry and style of partitioned deformation within a late Cenozoic transpressional zone in the eastern Gobi Altai Mountains, Mongolia. Tectonophysics, 277, 285–306. Danielsen, J. E., Auken, E., Jørgensena, F., Søndergaard, V. & Sørensen, K. I. 2003. The application

of the transient electromagnetic method in hydrogeophysical surveys. Journal of Applied Geophysics, 53, 181– 198. Dietze, E., Wu¨nnemann, B. et al. 2010. Basin morphology and seismic stratigraphy of Lake Donggi Cona, north-eastern Tibetan Plateau, China. Quaternary International, 218, 131–142. Ding, G., Chen, J., Tian, Q., Shen, X., Xing, C. & Wei, K. 2004. Active faults and magnitudes of left-lateral displacement along the northern margin of the Tibetan Plateau. Tectonophysics, 380, 243–260. Dogan, H. M. 2009. Mineral composite assessment of Kelkit River Basin in Turkey by means of remote sensing. Journal of Earth System Science, 118–6, 701–710. Gansu Provincial Geological Bureau 1980. Geological Report on 1:200,000 scale geological survey. Lanzhou Press., Lanzhou [In Chinese]. Gasse, F., Fontes, J. C., Van Campo, E. & Wei, K. 1996. Holocene environmental changes in Lake Bangong basin (Western Tibet). Part 4: discussion and conclusion. Palaeogeography, Palaeoclimatology, Palaeoecology, 120, 79–92. Geyh, M. A., Weizu, G. & Ja¨kel, D. 1996. Groundwater recharge study in the Gobi Desert, China. Geowissenschaften, 14, 279–280. Hartmann, K. 2003. Spa¨tpleistoza¨ne und holoza¨ne Morphodynamik im no¨rdlichen Gaxun Nur Becken, Innere Mongolei , NW China. PhD thesis, FU Berlin. Hartmann, K. & Wu¨nnemann, B. 2009. Hydrological changes and Holocene climate variations in NW China, inferred from lake sediments of Juyanze palaeolake by factor analyses. Quaternary International, 194, 28–44. Herzschuh, U. 2006. Palaeo-moisture evolution in monsoonal Central Asia during the last 50,000 years. Quaternary Science Reviews, 25, 163–178. Herzschuh, U., Tarasov, P. P., Wuennemann, B. & Hartmann, K. 2004. Holocene vegetation and climate of the Alashan Plateau, NW China, reconstracted from pollen data. Palaeogeography, Palaeoclimatology, Palaeoecology, 211, 1 –17. Hetzel, R., Niedermann, S., Tao, M., Kubik, P. W., Ivy-ochs, S. & Strecker, M. R. 2002. Low slip rates and long-term preservation of geomorphic features in Central Asia. Nature, 417, 428 –432. Holmes, J. A., Cook, E. R. & Yang, B. 2009. Climate change over the past 2000 years in Western China. Quaternary International, 194, 91– 107. Ho¨lz, S., Polag, D., Becken, M., Fiedler-Volmer, R., Zhang, H. C., Hartmann, K. & Burkhardt, H. 2007. Electromagnetic and geoelectric investigation of the Gurinai structure, Inner Mongolia, NW China. Tectonophysics, 445, 26–48. Hughen, S. L. K., Southon, J., Overpeck, J., Marchal, O., Herring, C. & Turnbull, J. 2004. 14C activity and global carbon cycle changes over the past 50,000 years. Science, 9, 202–207. Korup, O. & Montgomery, D. R. 2008. Tibetan plateau river incision inhibited by glacial stabilization of the Tsangpo gorge. Nature, 455, 786 –789. Kukla, G. & An, Z. 1989. Loess stratigraphy in Central China. Palaeogeography, Palaeoclimatology, Palaeoecology, 72, 811 –814.

NEOTECTONIC OF GAXUN NUR BASIN Lamb, M. A., Hanson, A. D., Graham, S. A., Badarch, G. & Webb, L. E. 1999. Left-lateral sense offset of Upper Proterozoic to Paleozoic features across the Gobi Onon, Tost, and Zuunbayan faults in southern Mongolia and implications for other central Asian faults. Earth and Planetary Science Letters, 173, 183–194. Lasserre, C., Gaudemer, Y. et al. 2002. Fast late Pleistocene slip rate on the Leng Long Ling segment of the Haiyuan fault, Qinghai, China. Journal of Geophysical Research, 107, 2276. Li, Y., Yang, J., Tan, L. & Duan, F. 1999. Impact of tectonics on alluvial landforms in the Hexi Corridor, Northwest China. Geomorphology, 28, 299– 308. Maizels, J. 1990. Raised channel systems as indicators of palaeohydrologic changes: a case study from Oman. Palaeogeography, Palaeoclimatology, Palaeoecology, 76, 241–277. Mischke, S. 2001. Mid and Late Holocene palaeoenvironment of the lakes Eastern Juanze and Sogo Nur in NW China, based on ostracod species assemblages and shell chemistry. Berliner Geowissenschaftliche Abhandlungen Reihe E, 35. Mischke, S., Demske, D., Wu¨nnemann, B. & Schudack, M. E. 2005. Groundwater discharge to a Gobi desert lake during Mid and Late Holocene dry periods. Palaeogeography, Palaeoclimatology, Palaeoecology, 225, 157 –172. Mischke, S., Sun, Z., Hezrschuh, U., Qiao, Z. & Sun, N. 2010. An ostracod-inferred large Middle Pleistocene freshwater lake in the presently hyper-arid Qaidam Basin (NW China), Quaternary International, 218, 74–85, doi: 10.1016/j.quaint.2009.03.002. Molnar, P. & Tapponier, P. 1975. Cenozoic tectonics of Asia: effects of a continental collision. Science, 189, 419– 426. Porter, S. C. & An, Z. 1995. Correlation between climate events in the North Atlantic and China during the last glaciation. Nature, 375, 305– 308. Reimer, P. J., Baillie, M. G. L. et al. 2004. IntCal04 Terrestrial Radiocarbon Age Calibration, 0– 26 cal kyr BP. Radiocarbon, 46, 1029–1058. Tan, Q. (ed.) 1996. The historical atlas of China. Vol.VII, The Yuan Dynasty Period, The Ming Dynasty Period. China Cartographic Publishing House, Beijing. Twiss, R. J. & Moores, E. M. 1992. Structural Geology. W.H. Freeman & Co., New York. Van der Woerd, J., Tapponnier, P. et al. 2002. Uniform postglacial slip-rate along the central 600 km of the Kunlun fault (Tibet), from 26Al, 10Be, and 14C dating of riser offsets, and climatic origin of the regional morphology. Geophysical Journal International, 148, 356–388. Vincent, S. J. & Allen, M. B. 1999. Evolution of the Minle and Chaoshui Basins, China: implications for Mesozoic strike– slip basin formation in Central Asia. GSA Bulletin, 111, 725– 742.

233

Woodcock, N. H. & Fischer, M. 1986. Strike– slip duplexes. Journal of Structural Geology, 8, 725– 735. Wu¨nnemann, B. 1999. Untersuchungen zur Palaeohydrographie der Endseen in der Badain Jaran- und Tengger Wu¨ste, Innere Mongolei, Nordwest-China. Habilitation thesis, Freie Universita¨t Berlin. Wu¨nnemann, B., Demske, D. et al. 2010. Hydrological cycles during the last 15 kyr in the Tso Kar lake basin, Ladakh (India), derived from geomorphological, sedimentological and palynological records. Quaternary Science Reviews, 29, 1138– 1155. Wu¨nnemann, B. & Hartmann, K. 2002. Morphodynamics and Paleohydrography of the Gaxun Nur Basin, Inner Mongolia, China. Zeitschrift fuer Geomorphologie N.F., Suppl.-Bd., 126, 147 –168. Wu¨nnemann, B., Hartmann, K., Altmann, N., Hambach, U., Pachur, H. J. & Zhang, H. 2007a. Interglacial and Glacial fingerprints from lake deposits in the Gobi Desert, NW China. In: Sirocko, F., Claussen, M., Sanchez-Goni, M. F. & Litt, T. (eds) The Climates of the Past Interglacials. Developments in Quaternary Sciences, 7, 323 –347. Wu¨nnemann, B., Hartmann, K., Janssen, M. & Zhang, H. C. 2007b. Responses of Chinese desert lakes to climate instability during the past 45,000 years. In: Madsen, D. B., Chen, F. H. & Gao, X. (eds) Late Quaternary Climate Change and Human Adaptation in Arid China. Developments in Quaternary Science, 9, 11–24. Wu¨nnemann, B., Reinhardt, C., Kotlia, B. S. & Riedel, F. 2008. Observations on the relationship between lake formation, permafrost activity and lithalsa development during the last 20,000 years in the Tso Kar basin, Ladakh, India. Permafrost and Periglacial Processes, 19, 341– 358. Xiao, J., Porter, S. C., An, Z., Kumai, H. & Yoshikawa, S. 1995. Grain size of quartz as an indicator of winter monsoon strength on the Loess Plateau of Central China during the last 130,000 Yr. Quaternary Research, 43, 22– 29. Yu, J. & Kelts, K. R. 2002. Abrupt changes in climatic conditions across the late-glacial/Holocene transition on the N. E. Tibet–Qinghai Plateau: evidence from Lake Qinghai, China. Journal of Palaeolimnology, 28, 195 –206. Yu, X., Zhou, W., Franzen, L. G., Xian, F., Cheng, P. & Jull, A. J. T. 2006. High-resolution peat records for Holocene monsoon history in the eastern Tibetan Plateau. Science in China (Series D), 49, 615– 621. Zhang, H. C., Ming, Q. Z. et al. 2006. Dilemma of dating on lacustrine deposits in a hyperarid inland basin of NW China. Radiocarbon, 48, 219– 226. Zhang, H. C., Peng, J. L. et al. 2004. Late Quaternary palaeolake levels in Tengger Desert, NW China. Palaeogeography, Palaeoclimatology, Palaeoecology, 211, 45–58.

Rock type, precipitation, and the steepness of Himalayan threshold hillslopes OLIVER KORUP1* & JOHANNES T. WEIDINGER2 1

Institut fu¨r Erd- und Umweltwissenschaften, Universita¨t Potsdam, D-14476 Potsdam, Germany 2

ERKUDOK Institute, K-Hof Museums, A-4810 Gmunden, Austria

*Corresponding author (e-mail: [email protected]) Abstract: Studies on the dynamic coupling between tectonics, climate, and erosion at the margins of the Tibetan Plateau entail the notion of widely occurring threshold hillslopes along the southern Himalayan front. Here we show that differences in major lithological units are sufficient to explain trends in topographical relief and mean slope gradients across the eastern Nepal Himalaya. This lithological control is manifest in modal slope gradients that serve as proxies of peak rock-mass strength, and remain strikingly invariant per rock type on both sides of the Main Central Thrust (MCT) despite comparably high amounts of monsoon precipitation. We infer that lithology rather precipitation patterns plays an important though hitherto largely neglected role in modulating threshold hillslope steepness. A nonlinear relationship between mean slope gradients and relief further indicates that significant steepening of threshold hillslopes through relief increases is more readily facilitated in low-relief terrain. Most evidence of large (.108 m3) bedrock landslides in the Himalaya clusters where topographical relief .2 km, and slope gradients are above average. This provides the mechanism necessary for rapidly lowering drainage divides and limiting hillslope heights, were the steepness of threshold hillslopes to decline via relief destruction.

The Tibetan Plateau is the topographical result of the continental collision of the Indian and the Asian tectonic plates. The Plateau has a mean elevation of c. 5 km (Fielding 2000; Tapponier et al. 2001), and its margins are bounded by major mountain ranges such as the Himalaya, the Karakoram, the Kunlun, and the Longmen Shan (Fig. 1). These ranges afford some of the most extreme terrestrial relief on Earth. Traditionally, research has focused on elucidating the timing of the formation and the development of the Tibetan Plateau over geological timescales, though an increasing number of studies have begun to explore the role of surface processes in its erosion (for a recent summary see Liu-Zeng et al. 2008). The Plateau gives rise to some of Asia’s largest rivers, such as the Indus, the Tsangpo-Brahmaputra, the Mekong or the Yellow River, and very high rates of Quaternary exhumation and erosion have been inferred for some of these rivers where they traverse the Plateau margins (Vance et al. 2003; Garzanti et al. 2004; Stewart et al. 2008). In particular the Himalayan syntaxes located at the corners of the indenter plate seem to promote tectonic aneurysms characterized by a close and spatially focused feedback between high exhumation, topographical relief, and erosion (Zeitler et al. 2001; Finnegan et al. 2008). This finding is consistent with the recognition of a causal linkage between topography and erosion (Roering et al. 2007). Based on a

study of several mountain belts, Montgomery & Brandon (2002) proposed a nonlinear correlation between erosion rate E and local topographical relief H: E ¼ E0 þ

kH , 1  (H=Hc )2

(1)

where E0 is the background rate of chemical erosion, k is a diffusivity coefficient, and Hc is a critical limiting relief. Equation 1 predicts that small increases in topographical relief by fluvial or glacial erosion require commensurately higher rates of erosion as H approaches Hc. The actual value of Hc, which serves as an asymptotic expression for E, is thought to depend largely on soil or rock strength (Montgomery & Brandon 2002). Similar relationships between erosion rates and alternative measures of catchment relief also hold for several rivers draining the margins of the Tibetan Plateau, though with varying degrees of nonlinearity (Fig. 2). This underlines the importance of quantifying nonlinear relationships and critical thresholds as rate constraints for studies that explicitly incorporate surface processes into geodynamic models of the evolution of the Tibetan Plateau and its margins (Liu-Zeng et al. 2008). Pivotal in many such models is the idea that rapid fluvial bedrock incision is the prime mover of hillslope response, and hence pacemaker of Himalayan

From: Gloaguen, R. & Ratschbacher, L. (eds) Growth and Collapse of the Tibetan Plateau. Geological Society, London, Special Publications, 353, 235–249. DOI: 10.1144/SP353.12 0305-8719/11/$15.00 # The Geological Society of London 2011.

236

O. KORUP & J. T. WEIDINGER

Fig. 1. (a) Topographical relief, derived as the maximum elevation difference in a 10 km radius from SRTM30 data at 850 m resolution, of the Tibetan Plateau and surrounding mountain ranges. Regions of highest relief include the Karakoram, the Higher Himalaya and its syntaxes, and parts of the Longmen Shan; (a)– (c) are extent of swath profiles in Figure 4a– c, respectively. Black dots are reported occurrences of 170 large (.108 m3) bedrock landslides.

relief evolution (Burbank et al. 1996; Lave´ & Avouac 2001). Before this background, the interpretation of mean slope-angle distributions serves as a commonly used surrogate to infer the widespread occurrence of threshold hillslopes, that is, those sustaining pervasive landsliding to limit their inclination to peak rock-strength characteristics rather than to rates of rock uplift and concomitant bedrock incision (Burbank et al. 1996; Montgomery 2001, Fig. 3a). Yet rarely have studies on threshold hillslopes produced any data on landslide distributions and their correlation to topographical metrics as supporting evidence. This is a major shortcoming, particularly as both the morphometric signature of threshold hillslopes and the sensitivity of any such measure to variations in orographically enhanced precipitation (Gabet et al. 2004) or rock-mass strength (Korup 2008) have remained largely unresolved. Specifically, Gabet et al. (2004) suggested that threshold hillslopes may become steeper in the wake of aridification due to a decreased susceptibility to landsliding through lower porewater pressures, and lower rates of chemical weathering (Fig. 3b). This entails that the rate of divide lowering by landslides, Ed, falls below that of channel incision, Ec. Conversely, transitions from a drier to a wetter climate could reduce threshold-hillslope steepness because of higher porewater pressures and slope instability, ultimately enhancing divide

lowering by landslides (Ed . Ec; Fig. 3c). Implicit to this interpretation is that hillslope gradients scale linearly with slope height h (approximated by local topographical relief H ), while valley spacing remains constant, and that slope failures prohibit any relief growth by wearing down drainage divides in concert with fluvial bedrock incision. This is in agreement with previous notions that large slope-clearing bedrock landslides spatially cluster in areas with high local topographical relief (Korup et al. 2007, 2010), thus maintaining threshold topography and dampening the potential for climate-induced changes to relief production (Whipple et al. 1999). Here we explore the role of rock-type variations in this proposed dynamic linkage between topographical relief, threshold hillslope steepness, climate, and landsliding. We do so by analysing the spatial variability of topography, precipitation, and large-scale slope failure in three transects across the southern margins of the Tibetan Plateau.

Methods and assumptions Measures of topographical relief typically refer to a maximum elevation range derived within a specified area. This scale dependence requires an appropriate sampling length scale. In their analysis, Montgomery & Brandon (2002) proposed a

ROCK TYPE AND THRESHOLD HILLSLOPES

Fig. 2. (a) Late Quaternary catchment-averaged denudation rates estimated from 10Be concentrations in fluvial sediments (Vance et al. 2003; Finnegan et al. 2008; Godard et al. 2010) show increases with topographical relief H10 averaged over contributing catchment area upstream of sampling locations (see Fig. 1 for locations). Black line is Equation 1 with parameter values indicated, and provides a lower bound to the data; (b) Similar positive, though more nonlinear correlation between denudation rates and total relief, that is, the difference between maximum and minimum elevation for a given sample catchment, Dadu River region, Longmen Shan (Ouimet et al. 2009). Black lines refer to Equation 1, and provide lower and upper bounds to the majority of the data points through an order-of-magnitude change in diffusivity k.

sampling radius of 5 km for obtaining local relief, H5. To fully encompass the maximum spacing between divides and valley floors of the major Himalayan river valleys that we studied (Fig. 1), we chose a sampling radius of 10 km instead. We derived the resulting local relief H10 from a SRTM30 digital elevation model (DEM) projected to Albers Conic and re-sampled to a nominal grid resolution of 850 m. We find that linear correlation with H5 remains robust at varying sample sizes (H10 ¼ 1.37H5 þ 99; units [m]; R 2 . 0.9). In order to estimate the variance of erosion rates from topography, we define an Erosion Index 1 by rearranging Equation 1 such that we normalize for effects of varying diffusivity and rates of chemical erosion, 1 ¼ (E  E0 )=k ¼

H10 : 1  (H10 =Hc )2

(2)

237

We assign Hc ¼ 6000 m because this value approximates the maximum of H10 for the Tibetan Plateau’s margins (c. 5900 m). Assuming that Hc cannot be substantially higher, we normalize 1 to a range from 0 to 1 to facilitate comparison between our study areas and to account for potential regional variations in Hc. The recently released ASTER G-DEM data allow quantifying local slope gradients at 30 m nominal resolution across the Himalaya. Though subject to localized errors and noise (Hayakawa et al. 2008), these data resolve landforms, catchment topography, and slope steepness at unprecedented detail. In order to explore potential effects of rock type on threshold topography, we digitized the scanned and orthorectified outlines of major tectonic and lithological units of eastern Nepal compiled by Schelling (1992) at 1:50 000 mapping resolution, and computed zonal statistics per major lithological unit. The regional tectonic context of these units within the Himalaya is summarized elsewhere (Hodges 2000; Yin 2006). We further used hillslope-gradient distributions to characterize threshold hillslopes following the method of Korup (2008). We focused exclusively on dissected hillslope portions devoid of major sediment storage or large-scale mass movement, and carefully restricted our analysis of distributions of local slope gradients based on .5000 sample pixels each in order to ensure that local spurious artefacts remained statistically subordinate. Systematic rate constraints on hillslope denudation in the Himalaya remain scarce, mainly because of few representative records of landslide activity (Shroder & Bishop 1998). Remote sensing data typically cover only few years to decades of slope instability, while expanding agriculture, forestry, and infrastructure in the Sub- and Lesser Himalaya have altered natural landslide occurrence with activities such as road cutting or slope regrading and terracing (Marston et al. 1998; Barnard et al. 2001). Remote sensing data above the tree line suffer from censoring effects such as shadows or lack of land-cover contrasts, making most small rock-slope failures undetectable. Therefore we compiled evidence of 170 large (.108 m3) bedrock landslides in major Himalayan rivers from a global database (Korup et al. 2007), and evidence of additional slope failures recently detected on LANDSAT, ASTER, and IKONOS imagery. The resolution of these data necessitated focussing on the largest either intact or dissected landslide deposits with areas .1 km2 preserved in high-order streams. Deposits were identified by diagnostic landform assemblages such as distinct hummocky deposits impounding lakes, or located downstream of low-gradient valley fills and braided rivers in otherwise deeply incised river valleys (Hewitt 1999;

238

O. KORUP & J. T. WEIDINGER

Fig. 3. Links between changes in hillslope height h and steepness of threshold hillslopes as a form of adjustment to fluvial bedrock incision; (a) Local relief is independent of fluvial incision rate Ec, if threshold hillslopes are at all times (S1, S2) inclined at gradient Sc, where shear stress equals hillslope strength, that is, limit equilibrium. Rate of divide lowering by landsliding Ed ¼ Ec, and hillslope sediment flux qs is solely controlled by Ec; (b) and (c) Long-term changes to hillslope water content and/or rock-mass strength controls threshold hillslope steepness, assuming divide spacing remains constant (see text for explanation); (d) Stability chart showing limit equilibrium conditions for bedrock landslides as a function of the angle of internal friction F, and dimensionless stability number c/0 gh, where c0 is apparent cohesion and g the specific weight of rock (Wyllie & Mah 2004). For example, fully drained rock slope with S ¼ 0.36 requires lower F than in fully saturated case to remain stable (‘A’, white arrow), whereas slope with tan F ¼ 0.6 may attain higher steepness in dry conditions than when fully saturated, that is, Sdry ¼ 1.2 and Ssat ¼ 0.58, respectively.

ROCK TYPE AND THRESHOLD HILLSLOPES

Korup 2006). Although many Himalayan hillslopes appear to be dotted by numerous scars of large bedrock landslides, there are only few associated deposits on the valley floors. In order to assess links between precipitation patterns and topographical relief, we standardized estimates of accumulated precipitation derived from Tropical Rainfall Measuring Mission (TRMM) 3B42 Version 6 data (http://disc.sci.gsfc. nasa.gov/giovanni/overview) at 0.258  0.258 resolution for the period 1998 –2009. Standardized precipitation P* has means of 0, and standard deviations of 1, and was projected to UTM coordinates and tension spline-interpolated to 850-m grid resolution. For exploring the linkage between lithology, rainfall, and threshold hillslope steepness, we plotted elevation, local relief, erosion index, slope gradients, and precipitation in three broad swath profiles across the Himalaya centred roughly about the Sutlej, the Kali Gandaki, and the Dudh Kosi rivers (Fig. 1).

Patterns of local relief, erosion, precipitation, and slope gradients across the Himalaya The most extreme topographical relief of the Tibetan Plateau (H10 . 5 km) is limited to the central Himalaya and the Himalayan syntaxes (Fig. 1). Broad swaths reveal that relief .2 km mainly occurs in the steep mid-reaches of major Himalayan rivers such as the Sutlej, Kali Gandaki, or Dudh Kosi, where they descend from the Plateau (Fig. 4). The erosion index 1 mimics this pattern (Equation 2), with pronounced and up to 40 km wide peak regions (1 . 0.8) that, going from the NW Himalaya to eastern Nepal, gradually shift from upstream to downstream with respect to the highest elevation (Fig. 4a, c, respectively). The spatial offset between peak values of 1 and 11-a precipitation P* is consistent with earlier observations (Bookhagen & Burbank 2006). We further note a distinct covariance of topographical relief and mean slope gradients across the Tibetan Plateau, the Himalayas, and the Gangetic Plain. This relationship is linear for low-relief parts of the landscape ( p , 1024), but becomes increasingly nonlinear at 1000 m , H10 , 2000 m before dissolving into a broad scatter for those small (,5%) fractions of landscape where H10 . 3 km (Fig. 5). The relationship between mean slope gradients and P* is less systematic, and hillslopes of comparable steepness occur in both arid Tibetan mountain ranges and the humid, monsoon-influenced Lesser Himalaya. The observed variance of P* for a given slope steepness is .2s in the western and central Himalaya, and does not indicate any clear relationship (see, e.g. locations A and A*, Fig. 4).

239

The majority of documented deposits from large bedrock landslides along the southern Tibetan Plateau margin occur as a distinct cluster in the Indus River catchment, which has been described elsewhere (Fig. 1; Korup et al. 2007, 2010). In southward draining Himalayan rivers comparable landslide evidence is limited to the Higher Himalaya, mostly below the mean elevation, but above the means of both slope gradients and local relief, particularly where H10 . 2 km, and where 1 approaches maximum values. These areas have also been subject to repeated Quaternary glaciations (Owen et al. 1998; Fort 2004). Inspection of satellite imagery and DEMs supports earlier observations that many of these rock-slope failures have detached from upper hillslopes, and in many cases involved sections of local divides or interfluves (Schramm et al. 1998; Weidinger 2006; Weidinger & Korup 2009). The variance of topography, relief, and mean slope gradients across the eastern Nepal Himalaya can be attributed to different major lithological units in a more or less systematic manner (Fig. 6a– c). Mean local relief is at a maximum (H10 . 2.5 km) in the Higher Himalayan Crystalline (HHC), especially where several kilometre-thick leucogranitic intrusions are exposed (Searle et al. 2003; rkg in Fig. 6). This lithology forms the steepest, highest, and highest-lying rock walls in the Khumbu area, and many of the highest peaks in the Himalaya (Searle & Godin 2003). In contrast, relief is only half as much in the inferred mechanically weak molasse rocks of the Sub-Himalayan Siwalik Group in the hanging wall of the Main Frontal Thrust (MFT). These rocks have mean slope gradients as low as c. 0.2, and excluding several extensive low-gradient piggy-back basins from the analysis produces mean slope gradients of c. 0.3. Paragneisses and leucogranites of the HHC attain average values of c. 0.6, but upon correcting for a similar slope-depressing effect of widespread low-gradient valley glaciers and glacial deposits (reflected in the high standard deviations, Fig. 6c), may reach average gradients of c. 1.3. Further removal of this bias by quantifying per lithological unit the fraction of area with slope gradients that are one or two standard deviations above the regional mean confirms that HHC and Tethyan sedimentary rocks are significantly steeper than most Lesser or Sub-Himalayan lithologies (Fig. 6d). For example, the outcropping area of leucogranite intrusions of the HHC contains 13% of the steepest (.2s) hillslopes in the entire eastern Himalaya, that is, three orders of magnitude more than the molasse sediments of the Upper Siwalik Group (us in Fig. 6). Density estimates of slope-gradient distributions in c. 30 km2 circular sample areas on either side of the Main Central Thrust (MCT), where gradients

240

O. KORUP & J. T. WEIDINGER

ROCK TYPE AND THRESHOLD HILLSLOPES

241

Fig. 4. Swath profiles of elevation, mean local relief, and slope gradient across the Himalaya centred about (a) Sutlej River region, NW Himalaya; (b) Kali Gandaki, central Nepal Himalaya; and (c) Dudh Kosi eastern Nepal Himalaya (see Fig. 1). Note high covariance between topographical relief and mean slope gradients (Fig. 5). Normalized Erosion Index derived from Equation 2 using mean local relief and Hc ¼ 6000 m. Most detected evidence of large bedrock landslides (locations refer to Fig. 8) in southward draining Himalayan rivers occurs in areas of above-average topographical relief and steepness. Standardized precipitation averaged over an 11-year period has two broad peaks at the range front, and where the gradient in topographical relief is highest. Vertical arrows and letters A and A* indicate identical mean slope angles with comparable standard deviations occurring at strikingly different precipitation regimes.

in topographical relief are most pronounced along the Himalaya, are remarkably uniform in terms of their modes of c. 0.7 (but not their means), particularly the Junbesi Paragneisses ( jp, Fig. 7) of the HHC. Rocks in the footwall of the MCT such as the Khare Phyllites have significantly lower modal slope gradients of c. 0.5 (kp, Fig. 7) despite a close proximity. Overall, leucogranites of the HHC Rolwaling-Khumbu-Makalu Unit (rkg, Schelling 1992) have modal slope gradients of c. 1 (Fig. 8).

Discussion: links between lithology, topographical relief, and threshold hillslope steepness Topographical relief and mean slope gradients covary systematically in three swaths across the

southern margin of the Tibetan Plateau, with peak values occurring in the Higher Himalaya (Fig. 4). Gabet et al. (2004) first highlighted a linear relationship with slope steepness and topographical relief gradually increasing from the Lesser Himalaya to the Higher Himalaya of central Nepal. Our results support these findings at a spatially much broader context and higher resolution, but demonstrate that mean slope gradients and local relief are nonlinearly related instead (Fig. 5). We infer that high-relief portions of the southern Tibetan Plateau margin, and especially those where H10 . 2 km, have lesser potential for significantly steepening their hillslopes through a given increase of relief by fluvial or glacial erosion than lower-relief areas. Judging from the substantial scatter in the few small areas of extreme relief (H10 . 4 km), we do not expect that changes in hillslope steepness and topographical relief growth or decay are detectable there.

242

O. KORUP & J. T. WEIDINGER

Fig. 5. Relationship between slope gradients derived from 30 m resolution ASTER G-DEM data averaged per 850 m cell of topographical relief based on SRTM30 data for the swath profiles shown in Figure 4. Horizontal bars in (a)–(c) are standard errors of the mean, and highest in the steepest 5% high-relief areas of the swaths (to the right of vertical dashed lines), where data scatter is substantial due to DEM noise, edge effects, and glacier cover. Black vertical arrows indicate where linear correlation between mean slope gradients and topographical relief breaks down in favour of a nonlinear trend; (d) mean slope gradients per lithological units (defined in Fig. 6, after Schelling 1992), eastern Nepal Himalaya; grey error bars are standard deviations.

ROCK TYPE AND THRESHOLD HILLSLOPES

243

Moreover, we find no evidence of invariant or statistically indistinguishable slope-gradient distributions, which may be interpreted as a diagnostic of threshold hillslopes (Burbank et al. 1996) unless there is an additional climatic or lithological control on terrain steepness. The spatial coincidence of peaks in standardized precipitation P* averaged over an 11-year period and the highest gradients in elevation and topographical relief in the Himalaya is striking, and supports the notion that rainfall is mainly orographically enhanced. Particularly near the MCT, a local precipitation peak coincides with a distinct and well-documented break in topography (Bookhagen & Burbank 2006; Wobus et al. 2006). However, the hillslopes in the footwall of the MCT formed by lithologies such as the Khare Phyllites (kp) or the Ramechap Group (rg) are consistently less steep than their nearby (,15 km) counterparts such as the Junbesi Paragneiss ( jp) the in the hanging wall in terms of both mean and modal slope gradients (Fig. 7). The slope-gradient distributions of these hillslopes share distinct uniform peaks that vary systematically with rock type. This result closely matches distributions of hillslope inclinations in landslide-prone mountain ranges of New Zealand (Korup 2008). There, the histogram peaks were also found to be invariant per rock type despite order-of-magnitude variations in uplift, precipitation, extent of Pleistocene glaciations, and contemporary land cover. The resulting interpretation that modal slope inclinations are a proxy of peak rock-mass strength also seems to be applicable to the eastern Nepal Himalaya. If all these are indeed threshold hillslopes then these significant differences in steepness occur despite comparable amounts of 2–3 m of mean annual precipitation and any concomitant seasonal fluctuations of porewater pressures due to monsoonal rainfall either side of the MCT. It follows that P* is insufficient to explain the measured differences in hillslope gradients. Moreover, arid mountain ranges on the Tibetan Plateau in the lee of the Himalayas may be as steep as parts of the monsoon-affected Lesser Himalayas (Fig. 4). The roughly inverse Fig. 6. (a) Elevation; (b) topographical relief; and (c) slope gradient ordered by descending means per extent of major tectonostratigraphic and lithological units (after Schelling 1992), eastern Nepal Himalaya: rkg: RolwalingKhumbu-Makalu leucogranites; rkp: Rolwaling-Khumbu-Kanchenjunga paragneiss; kjp: Kanchenjunga paragneiss; rkm: Rolwaling-Khumbu-Kanchenjunga migmatites; km: Kanchenjunga migmatite; ts: Tibetan Sedimentary Sequence; jp: Junbesi paragneiss; kag: Khandbari augen gneiss; kp: Khare phyllite; ntg: Narayan Than granites; dp: Dolakha phyllite; rg: Ramechap Group; dtg: Dobare Thumka granites; sag: Sisne Augen gneiss; sg: Sindbuli granites; tjg: Taplejung Group; ms: Middle Siwalik Group; skp: Sun Kosi phyllite; mc: Mahabharat Crystalline; ls: Lower Siwalik Group; us: Upper Siwalik Group. Dark grey shade indicates standard deviations. Light grey rectangle encompasses regional mean and standard deviations. Mean local relief culminates at .2.5 km in Higher Himalayan Crystalline (HHC). Mean slope gradients derived from 30 m ASTER G-DEM data vary between 0.2 and 0.6 according to prevalent rock types, and gradually increase from Sub-Himalayan Siwalik Group (Sub) to HHC. Excluding valley fills and glaciers exacerbates this trend, with mean slope gradients ranging from 0.3 to 1.3, respectively. Intermediate rock types are mainly of Lesser Himalayan Series (LHS); (d) Fraction of area covered by slope gradients greater than the regional mean (dark grey), one standard deviation (medium grey), and two standard deviations (light grey). Note log scale.

244

O. KORUP & J. T. WEIDINGER

Fig. 7. Density estimates of hillslope-gradient distributions sampled from 30 m ASTER G-DEM data in c. 30 km2 circles about the Main Central Thrust (MCT), and the Higher Himalayan Crystalline (HHC); low-gradient valley fills or sediment storage were excluded from the analysis. Dark grey histograms are bounded by Gaussian kernel density estimates (black thick lines). Light grey shaded areas indicate mean slope gradients and standard deviations. Rocks in the footwall of the MCT have modal slope gradients c. 0.5, whereas rocks in the hanging wall are consistently higher (c. 0.7; vertical black line); variations in modal slope gradients are interpreted as differing peak rock-mass strengths (Korup 2008). Leucogranite intrusions (rkg) are steepest, although sampling is compromised by too small coherent rock-slope areas. Tsergo Ri landslide data demonstrates pronounced slope-lowering effects of giant leucogranite-bearing landslide deposit (c. 1010 m3), Langtang, Nepal (Schramm et al. 1998).

ROCK TYPE AND THRESHOLD HILLSLOPES

245

Fig. 8. Local slope gradients, Khumbu Himalaya, eastern Nepal, derived from 30 m ASTER G-DEM data and superimposed on shaded topography. Despite local artefacts data resolve slopes as steep as c. 898, highlighting roughly horseshoe-shaped arrangement of peaks with extreme relief and slope steepness, formed mainly by leucogranites of the HHC (rkg in Fig. 6), and which are source areas for the dense network of large valley glaciers. Deposits from two large (.108 m3) catastrophic bedrock landslides stand out as conspicuous low-gradient (¼ bright) areas in otherwise pervasively dissected terrain.

relationship between precipitation and hillslope steepness in the Himalaya downstream of the regions of peak topographical relief switches to a direct one in the lee side of the orogen. Since the inclination of threshold hillslopes must by definition be independent of variations in rates of both rock uplift and erosion, we infer that historic precipitation records fall short of documenting a potential climatic control on threshold hillslope inclinations as was suggested by Gabet et al. (2004). Instead we propose that trends in hillslope steepness are linked to changes in major lithotectonic units across the Himalaya. In this context we note that the erosion index 1 attains comparable values on the arid Tibetan Plateau and the humid SubHimalaya, and also parts of the Lesser Himalaya (Fig. 4). However, documented rates of erosion

can be up to three orders of magnitude higher in the Siwalik Hills of Nepal (.10 mm a21; Lave´ & Avouac 2000) than on the southern Tibetan Plateau (,0.03 mm a21; Lal et al. 2003), suggesting that this striking discrepancy with respect to values of 1 may be accommodated by a commensurate order-of-magnitude variance in diffusivity k (Equation 2; see also Fig. 2b). Whether this fully reflects a lithological control on k akin to that reported for erodibility coefficients in stream-power laws (Stock & Montgomery 1999) will have to await further trial. Overall, however, the idea of a rock-type control on threshold hillslope inclination is consistent with the regional trends of topography, precipitation, and slope steepness (Fig. 4), and supports the notion of a material-dependent limit to

246

O. KORUP & J. T. WEIDINGER

topographical relief (Schmidt & Montgomery 1995). It further implies that lithological differences across the Himalaya modulate threshold hillslope gradients much more than previously assumed (Gabet et al. 2004). The argument that arid Tibetan mountain ranges, which are on average as steep as the monsoon-influenced Sub-Himalayan range-front hills (Fig. 4), may be below threshold inclinations due to a lack of significant incision further underscores the importance of rock-type variability over orographic precipitation contrasts. If the notion of a distinct lithological control on Himalayan threshold hillslope steepness is correct, then leucogranitic intrusions that form some of the tallest and steepest rock walls in the HHC today (Figs 7 & 8), should also be among the strongest rocks. The fact that in the Higher Himalaya these steep rock-slopes tower several km above large valley glaciers points to glacier erosion rather than bedrock river incision as the cause of threshold hillslopes. Thus, while our results underscore a

lithological control on threshold hillslope steepness, other climatic controls, such as repeated glacial oversteepening and subsequent debuttressing during deglaciation, will require future investigation. Large rock walls in the HHC actively erode by snow and ice avalanches, rock fall (Barsch & Jakob 1998; Heimsath & McGlynn 2008), and longrunout rock avalanches such as the one that originated from a cirque headwall at 5600 m asl in the Ama Dablam massif in 1979 (Selby 1988). Prehistoric rock-slope failures involving leucogranite can be orders of magnitude larger (i.e. 109 to 1010 m3; Figs 9 & 10), and have been reported, for example, from the Langtang and Kanjenchunga massifs, where giant catastrophic landslides have beheaded peaks, lowered divides, and obliterated valley floors with thick sheets of debris (Schramm et al. 1998; Weidinger & Korup 2009). We propose that HHC rock slopes are oversteepened in the sense that only rare and large bedrock landslides may efficiently lower their relief by detaching

Fig. 9. Sketches and field photos of Khumjung landslide, an example of how giant (.109 m3) and slope-clearing rock-slope failures contribute to wearing down drainage divides and hence limiting local hillslope relief, Khumbu area, Nepalese Himalaya (see Fig. 8 for location); (a) long section with simplified geology; (b) planform view highlighting eroded rockslide debris with basal sliding plane stranded c. 250 m above river level; (c) across-valley view from NE (cf. with panel a); and (d) view from top of the deposit SE of Khumjung.

ROCK TYPE AND THRESHOLD HILLSLOPES

247

Fig. 10. Views from Chukung Ri (5380 m asl) towards NW and east: leucogranite intrusions that have given rise to large catastrophic bedrock landslides within the HHC (Weidinger & Korup 2009); (a) debris cones fed by rock-slope failures from leucogranites forming the west ridge of Nuptse (7879 m asl); (b) leucogranites are exposed along the entire Khumbu-Khumbakarna-Himal from Cho Oyu (8153 m) in the WNW to Makalu (8475 m) in the ESE, that is, a distance of .50 km, over which steep rock walls dominate; most clastic sediments have a leucogranite provenance.

248

O. KORUP & J. T. WEIDINGER

from peak regions. The affinity of large bedrock landslides to areas of high topographical relief in major transverse Himalayan rivers (Figs 4 & 8) proves the availability of such an episodic process for maintaining threshold relief, and countering relief production driven by competing rock uplift, bedrock river incision, and glacial erosion.

Conclusions Differences in major lithological units are sufficient to explain regional trends in topographical relief and mean slope gradients, which covary in a nonlinear manner across the eastern Nepal Himalaya, and possibly elsewhere in the mountain belt. Particularly on either side of the MCT, where monsoonal rainfall is highest, threshold hillslopes formed in different rock types exhibit strikingly different modal slope gradients, which we interpret as regional proxies of peak rock-mass strength, and which may locally be modulated by rock-mass discontinuities. Leucogranites of the HHC in particular form the largest rock walls and mountain peaks in the Himalaya, and are inferred to be among the strongest lithologies in the orogen. We conclude that neither the notion of regionally uniform rock-mass strength nor that of a distinguishable precipitation control on threshold hillslopes and topographical relief across the central Himalaya is tenable. Alternative potential climatic controls through both para- and periglacial processes warrant further investigation. At the regional scale, detectable deposits of large bedrock landslides preferentially cluster in areas of extreme relief, slope steepness, and inferred erosion rates, underscoring their availability of contributing to undermining hillslopes, lowering divides, and essentially limiting topographical relief production. We thank S. Binnie, D. Scherler, R. Thiede, and an anonymous reviewer for constructive comments of an earlier version of this manuscript. OK acknowledges financial support through a Heisenberg Fellowship of the German Research Foundation (DFG).

References Barnard, P. L., Owen, L. A., Sharma, M. C. & Finkel, R. C. 2001. Natural and human-induced landsliding in the Garhwal Himalaya of northern India. Geomorphology, 40, 21–35. Barsch, D. & Jakob, M. 1998. Mass transport by active rockglaciers in the Khumbu Himalaya. Geomorphology, 26, 215– 222. Bookhagen, B. & Burbank, D. W. 2006. Topography, relief, and TRMM-derived rainfall variations along the Himalaya. Geophysical Research Letters, 33, L08405.

Burbank, D. W., Leland, J. et al. 1996. Bedrock incision, rock uplift, and threshold hillslopes in the northwestern Himalaya. Nature, 379, 505–510. Fielding, E. J. 2000. Morphotectonic evolution of the Himalayas and Tibetan Plateau. In: Summerfield, M. A. (ed.) Geomorphology and Global Tectonics. Wiley, Chichester, 201–222. Finnegan, N. J. et al. 2008. Coupling of rock uplift and river incision in the Namche Barwa–Gyala Peri massif, Tibet. Geological Society of America Bulletin, 120, 144 –152. Fort, M. 2004. Quaternary glaciation in the Nepal Himalaya. In: Ehlers, J. & Gibbard, P. L. (ed.) Quaternary Glaciations – Extent and Chronology, Part III. Elsevier, Amsterdam, 261– 278. Gabet, E. J., Pratt-Sitaula, B. A. & Burbank, D. W. 2004. Climatic controls on hillslope angle and relief in the Himalayas. Geology, 32, 629–632. Garzanti, E. et al. 2004. Sand petrology and focused erosion in collision orogens: the Brahmaputra case. Earth and Planetary Science Letters, 220, 157–174. Godard, V. et al. 2010. Spatial distribution of denudation in Eastern Tibet and regressive erosion of plateau margins. Tectonophysics, 491, 253 –274; doi: 10.1016/j.tecto.2009.10.026. Hayakawa, Y. S., Oguchi, T. & Lin, Z. 2008. Comparison of new and existing global digital elevation models: ASTER G-DEM and SRTM-3. Geophysical Research Letters, 35, L17404. Heimsath, A. M. & Mcglynn, R. 2008. Quantifying periglacial erosion in the Nepal high Himalaya. Geomorphology, 97, 5– 23. Hewitt, K. 1999. Quaternary moraines vs catastrophic rock avalanches in the Karakoram Himalaya, northern Pakistan. Quaternary Research, 51, 220–237. Hodges, K. V. 2000. Tectonics of the Himalaya and Tibet from two perspectives. Geological Society of America Bulletin, 112, 324 –350. Korup, O. 2006. Effects of deep-seated bedrock landslides on hillslope morphology, Southern Alps, New Zealand. Journal of Geophysical Research, 111, F01018. Korup, O. 2008. Rock type leaves topographical signature in landslide-dominated mountain ranges. Geophysical Research Letters, 35, L11402. Korup, O. et al. 2007. Giant landslides, topography, and erosion. Earth and Planetary Science Letters, 261, 578–589. Korup, O., Montgomery, D. R. & Hewitt, K. 2010. Glacier and landslide feedbacks to topographical relief in the Himalayan syntaxes. Proceedings of the National Academy of Sciences of the United States of America, 107, 5317– 5322. Lal, D. et al. 2003. Erosion history of the Tibetan Plateau since the last interglacial: constraints from the first studies of cosmogenic 10Be from Tibetan bedrock. Earth and Planetary Science Letters, 217, 33– 42. Lave´, J. & Avouac, J. P. 2000. Active folding of fluvial terraces across the Siwalik Hills, Himalayas of central Nepal. Journal of Geophysical Research, 105, 5735– 5770. Lave´, J. & Avouac, J. P. 2001. Fluvial incision and tectonic uplift across the Himalayas of central Nepal. Journal of Geophysical Research, 106, 26561–26591.

ROCK TYPE AND THRESHOLD HILLSLOPES Liu-Zeng, J., Tapponier, P., Gaudemer, Y. & Ding, L. 2008. Quantifying landscape differences across the Tibetan plateau: implications for topographical relief evolution. Journal of Geophysical Research, 113, F04018. Marston, R. A., Miller, M. M. & Devkota, L. P. 1998. Geoecology and mass movement in the ManasluGanesh and Langtang-Jugal Himals, Nepal. Geomorphology, 26, 139–150. Montgomery, D. R. 2001. Slope distributions, threshold hillslopes, and steady-state topography. American Journal of Science, 301, 432– 454. Montgomery, D. R. & Brandon, M. T. 2002. Topographical controls on erosion rates in tectonically active mountain ranges. Earth and Planetary Science Letters, 201, 481– 489. Ouimet, W. B., Whipple, K. X. & Granger, D. E. 2009. Beyond threshold hillslopes: channel adjustment to base-level fall in tectonically active mountain ranges. Geology, 37, 579– 582. Owen, L. A., Derbyshire, E. & Fort, M. 1998. The Quaternary Glacial history of the Himalaya. Quaternary Proceedings, 6, 91–120. Roering, J. J., Perron, J. T. & Kirchner, J. W. 2007. Functional relationships between denudation and hillslope form and relief. Earth and Planetary Science Letters, 264, 245– 258. Schelling, D. 1992. The tectonostratigraphy and structure of the eastern Nepal Himalaya. Tectonics, 11, 925–943. Schmidt, K. M. & Montgomery, D. R. 1995. Limits to relief. Science, 270, 617–620. Schramm, J.-M., Weidinger, J. T. & Ibetsberger, H. J. 1998. Petrologic and structural controls on geomorphology of prehistoric Tsergo Ri slope failure, Langtang Himal, Nepal. Geomorphology, 26, 107–121. Searle, M. P. & Godin, L. 2003. The South Tibetan Detachment and the Manaslu Leucogranite: a structural Reinterpretation and Restoration of the Annapurna-Manaslu Himalaya, Nepal. The Journal of Geology, 111, 505– 523. Searle, M. P., Simpson, R. L., Law, R. D., Parrish, R. R. & Waters, D. J. 2003. The structural geometry, metamorphic and magmatic evolution of the Everest massif, High Himalaya of Nepal– South Tibet. Journal of the Geological Society, London, 160, 345– 366.

249

Selby, M. J.. 1988. Landforms and denudation in the high Himalaya of Nepal: results of continental collision. Zeitschrift fu¨r Geomorphologie, 69, 133– 152. Shroder, J. F., Jr. & Bishop, M. P. 1998. Mass movement in the Himalaya: new insights and research directions. Geomorphology, 26, 13– 35. Stewart, R. J. et al. 2008. Brahmaputra sediment flux dominated by highly localized rapid erosion from the easternmost Himalaya. Geology, 36, 711 –714. Stock, J. D. & Montgomery, D. R. 1999. Geologic constraints on bedrock river incision using the stream power law. Journal of Geophysical Research, 104, 4983– 4993. Tapponier, P. et al. 2001. Oblique stepwise rise and growth of the Tibet Plateau. Science, 294, 1671– 1677. Vance, D., Bickle, M., Ivy-Ochs, S. & Kubik, P. W. 2003. Erosion and exhumation in the Himalaya from cosmogenic isotope inventories of river sediments. Earth and Planetary Science Letters, 206, 273–288. Weidinger, J. T. 2006. Predesign, failure and displacement mechanisms of large rockslides in the Annapurna Himalayas, Nepal. Engineering Geology, 83, 201– 216. Weidinger, J. T. & Korup, O. 2009. Frictionite as evidence for a large Late Quaternary rockslide near Kanchenjunga, Sikkim Himalayas, India – Implications for extreme events in mountain relief destruction. Geomorphology, 103, 57–65. Whipple, K. X., Kirby, E. & Brocklehurst, S. H. 1999. Geomorphic limits to climate-induced increases in topographical relief. Nature, 401, 39– 43. Wobus, C. W., Whipple, K. X. & Hodges, K. V. 2006. Neotectonics of the central Nepalese Himalaya: constraints from geomorphology, detrital 40Ar/39Ar thermochronology, and thermal modeling. Tectonics, 25, TC4011. Wyllie, D. C. & Mah, C. W. 2004. Rock Slope Engineering, Civil and Mining. 4th edn. Spon Press, London. Yin, A. 2006. Cenozoic tectonic evolution of the Himalayan orogen as constrained by along-strike variation of structural geometry, exhumation history, and foreland sedimentation. Earth Science Reviews, 76, 1– 131. Zeitler, P. K. et al. 2001. Erosion, Himalayan geodynamics, and the geomorphology of metamorphism. GSA Today, 1, 4– 9.

Index Page numbers in italics refer to Figures; page numbers in bold refer to Tables. Abor volcanics 72 age constraints 86– 88 previous 73, 75 Alashan Plateau 222, 223 Ama Dablam massif, landslides 246 amplitude analysis, low-velocity zone imaging 102, 104 Anyemaquen suture 11, 12 Arunachal Pradesh see Himalaya, Arunachal Asian Lithospheric Mantle 117, 118 attenuation, S-waves 117 Bailong Jiang fault 168, 169, 184 Bailong Jiang river 166, 167 bajada, Shuang Hu graben 131, 136 Banggong-Nujiang suture 3, 28, 116, 120 strike– slip belt 128, 157 velocity anomaly 118 Zagaya section, sinistral strike– slip faults 143–149 basins pull-apart 137 Gaxun Nur Basin 227, 229, 231 sedimentation 4, 10, 221–231 Bayan Har terrane 9, 10 Bombdila Group 73, 74 carbonates Longmen Shan 167 Selin Co Basin 32, 34, 35, 40 Cenozoic, exhumation, eastern Tibet 169–171, 181, 182 Chaka Lake 198, 214 channel flow model 100–101, 105–106 channel systems, eastern Tibet dynamic width adjustment 174–177 profile analysis 171– 172 role of lithology 178, 179– 180 steepness patterns 172– 174 and erosion rate 177– 179, 181 Chumulong Formation 41 climate and river terrace formation 214 see also palaeoclimate Cona Graben 48, 49 conductivity, electrical, southern Tibetan crust 100 Cretaceous Selin Co basin, foreland basin evolution 27– 42 Songpan–Garzeˆ belt 15, 17–18, 19 crust channel flow model 1 –2, 4, 100–101, 105– 106 seismic studies 109 Lhasa– Golmud transect 112, 114, 115 synoptic cross-section 119 –122 thickening 1 –2, 3, 11–12, 27 Qinling belt 184 thickness 121 upper–lower boundary 120 –121 see also Indian lower crust; Indian upper crust

Dalian Lake 198, 214 Damxung shear zone dating 150, 151, 154 kinematics 157 Danba de´collement zone 11, 14 Tertiary exhumation 20, 21 metamorphic terrane 14–15, 17 Daocheng granite 21 Dawa Co graben, low-velocity zones 101, 103 deformation channel flow model 100–101 neotectonic 128 –161 age 149 –161 kinematics 157–161 Muga Purou rift 128 –140 Tertiary, Songpan-Garzeˆ prism 20 Tethyan Himalayan Sequence 47, 49 Dogai Coring fold-thrust belt 139–140 drainage systems, Tibetan Plateau 4 Duba, Lower Cretaceous sediments 29, 32–38 Dudh Kosi River, relief and hillslope gradient 236, 239, 241, 242 Duolonggou Lake 198, 214 Duoni Formation 29, 31, 32, 34, 35–37, 41 dykes neotectonic dating 149, 150 Tethyan flysch 47, 49, 61– 62, 65 earthquakes broadband seismometers 110 –112 and faulting 140– 141 focal mechanisms 141–143 focal depth, and crustal structure 116, 120 see also Wenchuan earthquake erosion Late Pleistocene, channel systems, eastern Tibet 171 –182 and lithology 245– 246 and topographical relief 235– 237, 239– 241 erosion index 237, 239, 240, 241, 245 Eshaerbu Formation 29–30, 31, 32, 34, 35–37, 39, 41 exhumation Cenozoic, eastern Tibet 169 –171, 181, 182 Late Tertiary, Songpan-Garzeˆ prism 20– 22 fans alluvial Gaxun Nur Basin 224, 231 Jingzhushan Formation 35, 40 Shuang Hu graben 131, 136 Songpan-Garzeˆ Basin 10 submarine, Selin Co Basin 31–32, 39 faults brittle 151 ductile 151 Longmen Shan belt 167, 168, 169 Muga Purou rift, kinematics 140–143, 156

252 faults (Continued) normal central Tibet 156 Qiagam system 137 Shuang Hu graben 129, 136–137 strike–slip 3 –4 Muga Purou rift 128–137 dextral 151 sinistral 131, 137, 139, 143– 149, 150 –151 northern Tibet 156 Shuang Hu graben 136 west of Shuang Hu graben 137– 138 Fergana basin 4 fluid, free aqueous, Tibetan crust 105 flysch Songpan–Garzeˆ sequence 9, 10, 167, 168 Tethyan Himalayan Sequence 45–66 folding, Tethyan Himalayan Sequence 47 Fu Jiang river 166, 167 channel systems steepness 173, 174, 179 width 174, 175 Gaize-Selin Co thrust system 28, 29, 39, 40, 41 Gangdese magmatic arc 3, 27, 28, 29, 38, 41 foreland basin 30, 40–41 Garzeˆ granite pluton 21 Garzeˆ-Litang Basin see Yidun Basin Gaxun Nur Basin 222– 223 tectonic structures satellite imagery eastern basin 224–225 northwestern basin 227– 230 southeastern basin 225–227 spatio-temporal distribution 230–232 Gaxun Nur Lake 223, 227, 228, 229, 231 geomorphology, eastern Tibet 165– 184 Geren Co, Lower Cretaceous sediments 29, 32–38, 40–41 Geren Co-Jiali thrust system 28, 29, 41 Gobi-Tienshan Mountains 222, 223 Gonghe Basin 190, 194, 198 geology and geomorphology 191–192 palaeo-lake 214, 215 Yellow River terraces 192– 198, 199, 200 age modelling 209–212 and climatic fluctuation 214 cosmogenic nuclide dating 201– 212 incision age 212–213 incision rate 213– 214 lakes 193, 198, 214, 215 grabens low-velocity zones 101, 103, 105 –106 Muga Purou rift 128, 129, 130– 131 granitoids Songpan–Garzeˆ 12– 13 pre-Tertiary cooling 14–20 gravels, Gaxun Nur Basin 224–229 Great Counter Thrust 46, 66 Greater Himalayan Crystalline Complex 72–73, 74 Greater Himalayan Sequence 45, 46 Greater Indian lower crust 120 greenschist facies, Tethyan Himalayan Sequence 49, 61–62, 63, 65

INDEX Guanxian-Anxian fault see Pengguan fault Guide Basin 198 palaeo-lake 214, 215 Gurinai depression 224, 225, 227 Gurla Mandata detachment, kinematics 157 Gyangze dyke 149, 152, 153 Gyaring Co fault 128 Gyrong-Kangmar thrust 47 heat flow, Tibetan crust 99, 100 Hei River 223, 224 Hei Shui He river 166, 167 channel systems steepness 173, 174 width 174, 175, 176 Hi-CLIMB project 100, 101, 102, 104, 105, 112, 115 High Himalayan Crystalline rocks 119 lithology and relief 239, 241, 243, 244, 246, 247, 248 hillslopes, threshold 236 gradient 237, 239–241 topographical relief and lithology 241–248 Himalaya, Arunachal geological setting 72–73 Siang window 73–94 Himalaya Frontal Thrust 46 Himalayan Range 45, 46 geological background 46 relief, lithology and slope gradient 239–248 see also Tethyan Himalayan Sequence HIMNT experiment 112 Hoh Xil basin 4 Huya fault 167, 168, 169, 183 imbrication 1, 47 INDEPTH project 99, 109, 112, 115 phase I 110 phase II 99, 100, 101–102, 104, 105, 110 phase III 110, 111, 112, 113, 115, 118, 141– 143 phase IV 112, 114 India– Asia collision 1– 4, 11, 14, 27, 71 Tethyan Himalayan Sequence 47 Indian Lithospheric Mantle, subduction 2, 117, 118– 119, 120, 121 Indian lower crust 114, 115, 120, 121 Indian upper crust 119 Indus– Yarling suture 28, 29, 46 seismic studies 109, 110, 115, 117, 120 inversion, waveform, low-velocity zone imaging 102 –103, 104, 106– 107 Jian Jiang river, channel systems steepness 173, 174, 179 width 174, 175, 176 Jingzhushan Formation 29, 31, 32, 34, 35, 40 Jinsha subduction zone 9, 10 Jinsha suture 3, 11, 12 crustal thickness 112, 116 Junbesi Paragneiss 241, 243 Junggar basin 4 Jurassic, Songpan–Garzeˆ belt 15, 17–18 Juyanze palaeolake 223, 224–225, 231 Juyanze sub-basin, tectonic structures 224–225, 227

INDEX Kali Gandaki River, relief and hillslope gradient 236, 239, 240, 242 Kangtog Formation 136 Kanjenchunga massif, landslides 246 Karakorum –Jiali fault strike– slip faults 151, 156 dating 150 Kari-La granite, geochronology 152, 153 Khare Phyllites 241, 243 Khula Kangri granite 48, 49 Khumbu Himalaya, hillslope gradients 245, 246, 247 Khumjung landslide 246 Kokoxili granite 16, 18, 21 Konga Shan granite 14 shearing 20 Ku¨bler index, illite crystallinity 49, 51, 52–53, 54–55, 56, 62 Kunlun Block 3 granite 9, 16, 17– 18, 19, 21 Kunlun fault 169, 190, 191 Kunlun-Anyemaqen subduction zone 9, 10 landsliding, threshold hillslopes 236, 237–239, 244, 246, 247, 248 Langshan Formation 29, 31, 32, 34, 35, 40, 41 Langtang massif, landslides 246 Lesser Himalayan Sequence 46, 72– 73, 241, 243, 245 leucogranite, High Himalayan Crystalline, landslides 239, 241, 243, 246, 247, 248 Lhasa Block low-velocity zones 103, 105, 120 rifts 128 Lhasa terrane 3, 27–42, 28 geological background 29 seismic velocity 114– 115, 116 Selin Co, retroarc foreland basin evolution 30, 41–42 Lhasa–Golmud transect, seismic studies crustal structure 112, 114, 115 mantle structure 117– 119, 120 Lhunze fault 47, 48 Lingga Co 138 Linzhou basin 41 Linzizong Group 28, 29 Litang Basin 10 Litang– Batang block see Yidun block lithology and channel steepness index 178, 179– 180 and hillslope gradient 245–246 lithosphere– asthenosphere boundary 117, 118, 121 loess 214 Longmen Shan belt 11, 12, 14 Cenozoic exhumation 169– 171 channel systems Pleistocene erosion 172–180 steepness 173, 174 and erosion rate 177–179 width 174 –177 deformation 165 faults 167, 168, 169 geology 167 Moho depth 167 rock uplift 180– 184 Tertiary exhumation 20– 22

253

Longriba fault 168, 169 Longyang gorge 191, 192, 196 Longyangxia Reservoir 191, 197 low-velocity zones southern Tibet 99– 107 channel flow model 100–101, 105–106 spatial extent 103 –105 Lunggar rift detachment, dating 150, 151 magmatism mafic, Tethyan flysch sequence 47 Tibetan Plateau 2 –3, 41 magnetization natural remanent, Zagaya section 145, 146, 147 –148 Siang window 76, 77– 80, 91 see also palaeomagnetization Main Boundary Thrust 46, 72–73 Main Central Thrust 46, 66, 72–73 hillslope gradient 239, 241, 243, 244, 248 Main Frontal Thrust 239 Main Himlayan Thrust 114, 115, 119 Manai granite 14, 15, 16, 18, 19 mantle, structure 117– 122 transition zone boundaries 119, 121 Maoergai granite 14, 15, 16 Markam granite 14, 15, 16, 18, 19 Mayue, Lower Cretaceous sediments 29, 32, 35, 38–39 metamorphism Barrovian-type 12, 14–15 Tethyan Himalayan Sequence flysch 61– 63 metapelites, Tethyan flysch 48, 49–60, 62– 63 mica, white, metamorphism, Tethyan flysch 58, 61–62, 64 migration, low-velocity zone imaging 102 Min Jiang fault 168, 169 Min Jiang river 166, 167 channel systems lithology 178, 180 steepness 173 width 174, 176 Min Shan 166 active faults 183–184 Cenozoic exhumation 169–171 channel systems lithology 180 Pleistocene erosion 172– 180 steepness 174 width 174– 177 geology 167 thrust fault 21 Miri Formation 73, 74 age constraints 86, 88 Moho, Tibetan Plateau 112 Muga Purou rift 128– 140, 129 neotectonic deformation dating 149 geometry 150, 156–157 northern 138– 140 southern 128, 130–131 Shuang Hu graben, fault system 129, 131– 137 west of Shuang Hu graben 137–138 western and eastern grabens 130, 131

254 Namche Barwa Syntaxis, tectonic rotation 71– 72, 88, 90–94, 160 Neotethys, subduction 29, 41 Ngamring Formation 38 Nianbaoyeche granite 13 Nima basin 4, 41 Niubao Formation 145 North Himalayan gneiss domes 46, 48 North Pasighat Thrust 73 nuclides, cosmogenic, dating method 201– 212 Nuptse, landslides 247 Nyainqentaghla detachment age of deformation 150, 151, 152 kinematics 157 ophiolites, Tethyan Himalayan Sequence 47, 48 palaeoclimate research, lake sediments, Tibetan Plateau 221 palaeomagnetism eastern Tibetan Plateau 91 Siang window 80– 85, 90–92 Siling Co, strike–slip faults 145– 148 palaeosol 214 Palaeotethys, closure 9, 10, 11 Pamir–Tibet– Himalaya orogenic system 1 –4, 2 Panxi Rift 10 particle flow 4, 160– 161 Pengguan fault 167, 168, 169, 180 –181 Pengguan massif, lithology and erosion rate 178, 179 Pleistocene, Late, erosion, eastern Tibet 171–182 precipitation, and topographical relief 243, 245 Pung Co rift, faulting 158, 160, 161 Qaidam basin 4 crustal thickness 121 seismic studies 109, 110, 111–112, 116, 118 crustal depth 114, 120 Qiagam normal fault system 137 Qiangtang Block 3, 9, 10, 30 Qiangtang terrane crustal thickness 121 seismic velocity 114– 116, 117, 120 Qingchuan fault 183 Qinghai Lake basin 193, 194, 195, 198 palaeo-lake 214, 215 Qinling belt 1, 10, 167 crustal thickening 184 see also West Qinling orogen Qionglai Shan, topography 167 receiver function studies 102, 106, 112, 113, 119, 120, 121 relief, topographical lithology and hillslope gradient 241–248 measurement 236–239 rifting 128– 140 Rila Formation 29, 31, 32, 41 Rilonguan granite 15, 16, 18, 19 Ringbung graben faulting, brittle 150, 151 leucogranite, geochronology 152, 153, 155, 156 rock type see lithology Rolwaling-Khumbu-Makalu Unit 241, 243

INDEX rotation, tectonic Namche Barwa Syntaxis 71–72, 90– 94 Zagaya section 160 models 148– 149 spatial distribution 147–148 Se La Group 73, 74 sediment 4, 10 fluvial, Gaxun Nur Basin 224– 225, 228 lacustrine, Gaxun Nur Basin 221– 222, 225, 229– 231 Lower Cretaceous, Selin Co Basin 29– 42 seismic studies controlled source studies 109, 110 earthquake focal depth 116 Lhasa–Golmud transect crustal structure 112, 114, 115 –117, 119–122 mantle structure 117– 122 low-velocity zones, southern Tibet 99– 107 passive source studies 110 –112 shear wave attenuation 117 Selin Co basin 27–42 geology 31 palaeogeography 39–40 petrographic analysis 35, 36 retroarc foreland basin 41–42 sediment provenance 38–40 stratigraphy 29–35 tectonic setting and basin evolution 40–41 U–Pb geochronology 35– 38 shear wave splitting 117, 118 shear zones, dating 149–150 Shiquanhe– Gaize– Amdo thrust system 3, 28, 29 Shuang Hu anticlinorium 3 Shuang Hu graben neotectonic dating 149, 151 normal fault system 129, 131–133, 136–137, 143 Siang window 72, 74 K–Ar geochronology 76– 77, 85, 86 magnetic mineralogy 76, 77– 80 palaeomagnetism 80–85, 88, 90–91, 92 previous age constraints 73, 75 tectonostratigraphy 73 (U–Th)/He thermochronology 76, 86, 87, 88, 89 Sichuan Basin 166, 167 active faults 183 Cenozoic exhumation 169–171, 181 channel systems, Pleistocene erosion 172– 180 deformation 165 Moho depth 167 rock uplift 182– 183 Siling Co 144 strike– slip faults palaeomagnetic analysis 145– 148 remanence measurement 145–148 rock magnetic properties 145, 146 Sino-American Tibetan Plateau broadband experiment 110, 111 Sino-French seismic studies 109, 110, 111 –112, 118 Siwalik Group 72– 73, 74 lithology and relief 239, 245 slopes see hillslopes Sogo Nur Lake 223, 227, 231

INDEX Somang Qu river 166, 167 channel systems steepness 173, 174 width 174, 175 Songpan Ocean 9 –11 Songpan-Garzeˆ basin 10 Songpan-Garzeˆ flysch sequence 167, 168 Songpan-Garzeˆ fold belt 9, 10, 12, 13, 22 accretionary prism 10 Late Tertiary exhumation 20– 22 pre-Tertiary cooling 14–20, 21 crustal structure 116 pre-Tertiary cooling 14– 20, 21 numerical modelling 17– 18 South Tibetan Detachment System 46– 47, 49, 66, 72, 119 subduction Indian and Asian lithosphere 1, 2, 3, 118–119, 120, 121 north Tibet 10– 11 Sutlej River 236 relief and hillslope gradient 236, 239, 240, 242 Tajik basin 4 Tak Kyel Co graben, low-velocity zones 101, 103 Takena Formation 41 Tangra Yum Co, deformation 150, 151 Tarim basin 1, 4 Tazang fault 168, 169, 184 Tertiary, Songpan-Garzeˆ prism, exhumation 20–22 Tethyan Belt seismic basement 119 Tethyan Himalayan Sequence 45, 46, 47 deformation 47, 49 flysch 47, 50 K– Ar geochronology 51–52, 53, 54– 55, 56, 57, 62 Ku¨bler index 49, 51, 52–53, 54– 55, 56, 62 metamorphism basic dykes 61–62 greenschist facies 61– 62, 63, 65 low-grade, metapelites 62–63 maximum temperature 63 post-sedimentary evolution 63– 66 thermobarometry 51, 53, 57–58, 60 (U– Th)/He thermochronology 52, 56, 58, 59 vitrinite reflectance 51, 53, 56, 58, 63 structural setting 47–49 Thakkhola graben, age 150, 151 thermobarometry, Tethyan flysch 51, 53, 57–58, 60 thermochronology Siang window 76, 86, 87, 88, 89 Tethyan flysch 52, 54, 56, 58 thrust-fold belts 1, 3 Tibetan Lithospheric Mantle 119 Tibetan Plateau crustal shortening 3, 11– 12, 27 crustal thickness 121 drainage systems 4 eastern margin 166 Cenozoic exhumation 169– 171, 181, 182 differential rock uplift 180–182 geomorphology 165–184 Late Pleistocene erosion 171– 182 topography 166– 167

255

erosion 235 magmatic belts 2– 3, 41 northeastern, evolution 9– 23 southern margin, relief, lithology and slope gradient 235, 236, 239 –248 strike–slip faults 3 –4 synoptic cross-section 119– 122 Tongde Basin 190, 191, 213 Triassic, cooling, Songpan-Garzeˆ granites 14– 20 vitrinite reflectance 51, 53, 56, 58, 63 Waligong Shan 191– 192, 194, 196, 197 uplift 212 Yellow River terraces 203, 206, 208, 209 age modelling 211–212 incision rate 213 waveform inversion, low-velocity zone imaging 102–103, 104, 106– 107 Wenchuan earthquake 165, 167, 169 spatial variations in slip 180– 182 Wenchuan– Maowen fault 168, 169, 179, 183 West Qinling orogen 166, 167, 169, 184 Xianshui He fault 13, 14, 20, 21 Xidatan orthogneiss 16, 17, 18, 21 Xigaze forearc basin 38 Yadong-Gulu rift 49 INDEPTH II data 99, 100, 101– 102 Yala Xiangbo dome 61, 62, 63, 65 Yang Sang Chu Formation 74 Yanggon granite 14, 15, 16 Yarlung-Tsangpo Reflection 114, 115 Yarlung-Tsangpo Suture 100, 105, 161 Yellow River 189, 190, 191 geology and geomorphology 191– 192, 194, 195 terraces Gonghe Basin 192–198, 199, 200 age modelling 209– 212 and climatic fluctuation 214 cosmogenic nuclide dating 201–212 incision age 212– 213 incision rate 213 –214 lakes 193, 198, 214, 215 T5 202, 206, 207, 209 T7 196, 199, 200, 202– 203, 204– 206 Yibug Caka rift 128, 151 neotectonic deformation, geometry 156 Yidun basin 10 Yidun block 9, 10 granite 9, 10, 16, 17, 20 Tertiary exhumation 20 Yingxiu– Beichuan fault 167, 168, 180–181, 182 Yinkiong Formation 73, 75 age constraints 85–88 Yushu granites 10 Yushu-Batang subduction zone 9, 10 Zagaya River 144, 145 Zagaya section, sinistral strike– slip faults 143–149 Zenong Group 28, 29, 31, 41 Zoige Basin 190, 191