Tertiary Structural Evolution of the Gangdese Thrust System, Southeastern Tibet

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 99, NO. B9, PAGES 18,175-18,201,SEPTEMBER 10, 1994

Tertiary structural evolution of the Gangdese thrust system, southeastern Tibet

An Yin, T. Mark Harrison,1 F.J. Ryerson,2 Chen Wenji, 3 W.S.F. Kidd,4 andPeter Copeland 5

Abstract. Structuraland thermochronological investigations of southernTibet (Xizang) suggestthat intracontinental thrusting has been the dominantcause for formationof thickened crustin the southernmostTibetan plateau since late Oligocene.Two thrustsystems are documentedin this study:the northdipping Gangdese system (GTS) andthe youngersouth dippingRenbu-Zedong system (RZT). West of Lhasa,the Gangdesethrust juxtaposes the Late Cretaceousforearc basin deposits of the LhasaBlock (the Xigaze Group)over the Tethyan sedimentaryrocks of the Indianplate, whereas east of Lhasa,the faultjuxtaposes the Late Cretaceous-Eocene,Andean-type arc (the Gangdesebatholith) over Tethyan sedimentary rocks. Near Zedong, 150 km southeastof Lhasa,the Gangdesethrust is markedby a >200-m-thick myloniticshear zone that consistsof deformedgranite and metasedimentary rocks. A major southdipping backthrust in the hangingwall of the Gangdesethrust puts the Xigaze Groupover Tertiaryconglomerates and the Gangdese plutonics north of Xigaze andwest of Lhasa. A lower age boundfor the Gangdesethrust of 18.3+0.5Ma is given by crosscuttingrelationships. The timingof slip on the Gangdesethrust is estimatedto be 27-23 Ma from 40Ar/39Ar thermochronology,and a displacementof at least46+9 km is indicatednear Zedong. The ageof the Gangdesethrust (GT) is consistentwith an upperage limit of-24 Ma for the initiation of movementon the Main Centralthrust. In places,the youngerRenbu-Zedong fault is thrustover the traceof the GT, obscuringits exposure.The RZT appearsto have beenactive at circa 18 Ma but hadceased movement by 8+1 Ma. The suturebetween India andAsia hasbeen complexly modifiedby developmentof the GTS, RZT, and,locally, strike-slipand normal fault systems.

Introduction Himalaya [Wang et al., 1983; Burg, 1983], no Tertiary, crustal-scalethrusts have been reportedfrom the Lhasa Block. Various strain accommodation mechanisms have been However, geologic relationships, in particular the attributed to the Indo-Eurasia collision during the Tertiary, juxtapositionof Tethyan continentalmargin depositsagainst includingstrike-slip faulting [Tapponnieret al., 1982], Andean-typeplutonic rocks (the Gangdesebatholith), suggest thrusting[Argand, 1924; Powell and Conaghan,1973; Dewey that a southdirected, intracontinental thrust systemmay have and Burke,1973; Dewey et al., 1989], and a combinationof been active in southernTibet subsequentto collision [e.g., both that vary in spaceand time during the protracted Searle et al., 1987]. In light of thermochronologicalresults from southeasternTibet that appearto requiresuch a structure, collision [Harrison et al., 1992a]. It has long been we previously proposed[Harrison et al., 1992a] the existence appreciatedthat at leasta portionof theTertiary convergence hasbeen accommodated by majorintra-continental thrusts in of a fault we termed the Gangdesethrust system(GTS). We presenthere the resultsof field observationsconducted during theHimalaya [LeFort, 1986; Mattauer, 1986; Bouchez and the summer of 1992, together with further Pgcher, 1981]. Althoughsyncollisional thrusts have been thermochronologicalanalyses, that confirm our predictions mappedwithin Tethyan strata on the northernflank of the regardingthe GTS and constrainits timing, structuralstyle, and kinematic history. Implicationsof the GTS for the Indo- Asia collision are discussedand a tectonicinterpretation for •Departmentof Earthand Space Sciences and Institute of the Tertiary evolutionof the southernTibetan plateauand the Geophysicsand PlanetaryPhysics, University of California, Los Higher (= Greater)Himalaya is proposed. Angeles. 2Instituteof Geophysicsand Planetary Physics, Lawrence Geologic Setting LivermoreNational Laboratory, Livermore, California. 3Instituteof Geology,State Seismological Bureau, Beijing. Major tectonic elements in southernTibet (Plate 1) consist 4Departmentof Geological Sciences, State University of New of a pre-earlyTertiary thrustbelt alongthe northernmargin of York at Albany. the Lhasa Block (the accreted continental block bounded on $Departmentof Geosciences, University of Houston,Houston, the south and north by the Indus-Tsangpo and Banggong Texas. sutures, respectively) [Allegre et al., 1984; Dewey et al., 1988], an Andean-type Mesozoic to early Tertiary igneous belt (the Gangdesebatholith and Linzizong volcanics), a late Copyright1994 by the AmericanGeophysical Union. Cretaceous-earlyTertiary forearc basin (the Xigaze Group), and the ophiolitic Indus-Tsangposuture zone (Plate 1, inset) Paper number94JB00504. that separatesrocks in the Lhasa Block from rocks of Indian 0148-0227/94/94JB-00504505.00 affinity. The presenceof the Gangdesebatholith and the 18,175 18,176 YIN ET AL.' GANGDESE THRUST SYSTEM

o YIN ET AL.: GANGDESE THRUST SYSTEM 18,177

forearc sequence of the Xigaze Group suggests that 4d), tension gashes, and asymmetric boudinage structures convergencebetween the Indian and Eurasianplates prior to (Plate 4d) are abundant in the mylonitic shear zone and the continental collision was accommodatedby a north suggesta top-to-the-southsense of shear, consistentwith the dipping subduction zone beneath southern Asia. Field field relationship. Microscopic examination of kinematic observationsshow, however, that the Indus-Tsangposuture indicators in the mylonitic rocks likewise suggeststhe same zone is southdipping near Gongga(Plate 3a) [e.g.,Allegre et sense of shear. We propose below that the south-directed al., 1984],suggesting that the suturewas deformed subsequent faults near Zedong and Xigaze are both manifestationsof the to closureof the ocean(s)that onceseparated India from Asia. Gangdesethrust system that was variably active during the In this paper we argue that this south dipping contact late Oligocene to Miocene. representsa Neogenenorth directed thrust (the Renbu-Zedong Based on Chinese mapping [Wang et al., 1983] and our own fault) that postdatesthe Gangdesethrust along the southern observations,the structuralstyle in the hanging wall of these edge of the Lhasa Block. north dipping thrusts appears to vary along strike. For The Gangdese batholith is broadly characterized by example, west of Renbu, the Gangdesethrust (GT) may be exposureof Cretaceousto Tertiary volcanicand hypabyssal manifestedas two opposedthrust faults (a "pop-up"structure); rocks west of the longitudeof Renbu, and age-equivalent a north dipping thrust bounding the southern limit of the plutonic rocks east of Renbu (Plate 1). This map relation Xigaze Group and a south dipping thrust juxtaposing these indicates differential unroofing along strike: the structural forearc sediments against Tertiary conglomeratic sandstones level of pluton exposurenear Lhasa is between 10 to 15 km as and Gangdeseplutonic rocks. As Chinese mapping indicates indicatedby thermochronometryand barometry[Copeland et that this latter feature merges with the GT near Renbu, we al., 1987, 1994], whereas the widespreadpreservation of interpret it to be a backthrustof the GT (Plates 1 and 2). This volcanic rocks in this belt suggestsmuch less denudation relationship suggests that the two thrusts that bound the west of Renbu. Spatiallycorresponding to the preservation Xigaze rocks were coeval (Plate 2). Structuresin the hanging of volcaniccover in the GangdeseShan is the appearance,to wall of this backthrust are characterizedby the development the south,of the Xigaze Group (Plate 1). The tectonicbasis of of north directed imbricate and duplex thrust systems (Plate this relationship has been investigated in light of the 4e); this structural style differs from the widespread structuraland thermochronologicalstudies. We proposethat asymmetric folds in the Xigaze Group immediately above the the distribution of these rock units in southern Tibet are best GT to the south. The backthrustis cut by a series of north- explainedby a variationin structuralstyle and magnitudeof south trending andesitic dikes. Near Zedong, 200 km east of displacementalong the GTS. The presentIndus-Tsangpo Renbu, the GTS is characterized by a single major thrust sutureis not the trace of a singlefault contactmarking the (Plates 1 and 3). Spatially correspondingto this change in locus of oceanicclosure between India and Asia. Rather, it is style of deformation east of Renbu is the absence of the a composite structural feature resulting from multiphase Xigaze Group either in the hangingwall or footwall of the GT. deformationincluding both dip-slip and strike-slipfaulting There are three possibleexplanations for the absenceof the since the late Oligocene. Xigaze Group rocks east of Renbu: erosion, removal by strike-slip faulting, or thrusting beneath the Gangdese fault. Structural Geology For Xigaze Group rocks to have been eroded away above the GT requires that the Gangdese plutonic rocks to predate the forearc sediments, since nowhere is an intrusive relationship Various structural features have been observed in the observed. However, the Xigaze Group is largely Cretaceous southern Tibetan plateau during our field study, including while Gangdesegranodiorites are as young as Eocene making thrusts, strike-slip faults, normal faults, folds and foliations. the exposed granitoids in the Zedong area unlikely candidates Tectonic positions, structural style, and timing of formation for those related to the Indo-Asia collision are described as basementof the Xigaze Group. At Quxu, 60 km southwest below. of Lhasa (Plate 1), anotherlocation where Tethyanstrata directlyabut the LhasaBlock, the granitoidsare firmly dated Thrust Systems at 414-1Ma [Schgireret al., 1986]. Althoughit is possible that much of the suture zone could have been shifted to the A northdipping thrust is well exposedin theZedong and eastby strike-slipfaulting, Xigaze Grouprocks have not been Xigazeareas (Plates 1-3). In theXigaze area (Plate 2a), the identified in eastern Tibet or Yunnan. Furthermore, the northdipping fault juxtaposes the Xigaze Group over Tethyan Xigaze Group also disappearswest of Mt. Kailas [Harrison et strataof the Indianplate. Rocksdirectly below the fault are al., 1993a]where plutonic rocks are widelyexposed, making extensivelyfractured and sheared. In contrast,south verging it difficult to understandhow right-lateralstrike-slip faulting mesoscopicfolds are the dominantstructures in the Xigaze could remove both the westernand easternportions of the Groupin thehanging wall. Dominanteast-west trending and forearcsediments but leavethe centralregion exposed. This southverging folds directly above the fault suggestthat it is conjecturealso does not explain the close relationship southdirected (Plate 4a). Althoughsome of thehanging wall betweenthe Xigaze Group and adjacentvolcanic rocks of the deformationcould be subductionrelated, a substantialportion Gangdesebatholith where the forearcis preserved. Thus we of it must postdatecollision as the overlyingEocene (?) concludethat, eastof Renbu,at leastthe approximatepresent Qiuwu conglomerate[Wang et al., 1983]is alsointensely width of the Xigaze Group (-20 km) was thrust beneaththe folded. In the Zedongarea (Plate 3a), thenorth dipping thrust Gangdesebatholith. Because the Xigaze andEocene (?) Qiuwu juxtaposesthe Gangdeseplutonic rocks and metasedimentary strata have been substantiallyshortened [Wang et al., 1983], rocks over -1-km-thick Tertiary(?) conglomeratesand at least in part since the collision began, the amount of Tethyan sediments(Plate 4b). Clasts of the Tertiary(?) displacementcould be significantlygreater. If we consider conglomeratesare predominantlymarble and granite that were that 10-15 km of crusthas been eroded away from muchof the likelyderived from the hanging wall metasedimentarycountry Gangdesebatholith east of Renbu as a consequenceof late rocks intruded by the Gangdesebatholith. The thrust near Oligocene-EarlyMiocene thickening[Copeland et al., 1987, Zedongis definedby a well-exposed>200-m-thick mylonitic 1994; Harrison et al., 1992a] alonga thrustdipping 30+5 ø, shear zone dipping at between 25o-35ø to the north and an additional 17-35 km of slip along the GT is indicated. consisting of granitic rocks and marble (Plate 4c). Thus we estimate a minimum displacementon the thrust of Mesoscopicstructures including south verging folds (Plate 464-9 km. 18,178 YIN ET AL.: GANGDESE THRUST SYSTEM

o 89ø o o V Q

Kg Tg

TS TS

29ø 20' 29ø20'

I 20km I

8 89ø

(a)

c TS TV KA TV Pz-blz A ! km / -T

• pz.Mzl n • Gat o s th ust Renbu-Z don lO thrust

(b)

Plate 2. (a) Geologic map of the Xigaze area, modified from Wang et al. [1983] basedon our field observations.Q, Quaternary sediments;Ts, Tertiary sedimentaryrocks; Tv, early Tertiary volcanic rocks of the Linzizhong Formation; Kg-Tg, Cretaceousto early Tertiary granites of the Gangdese batholith;K A, Cretaceoussedimentary of the TakenaFormation; Pz-MzA, Paleozoicand Mesozoic sedimentaryrocks of theLhasa Block; Xg, XigazeGroup; Pz-Mz In, Paleozoicand Mesozoic sedimentary rocks of the Indian plate. (b) Cross section C-C' showing structural style in the hanging wall of the Gangdesethrust near Xigaze. YIN ET AL.: GANGDESE THRUST SYSTEM 18 179

i Mezhugongka B thrustbelt Pz'KA

Q Pz-KA

K T -29ø30 '

am e Pz-K A

Zedong G th

ngg

Kin

92 ø

Renbu-Zedong thrust A Pz'KAA' km Q• / km

---- o g acer ry Ga thru lO --- lO

2o Lhaaa block .uture zone (b) Mezhugongka thruet belt ZH-2, ZH-3 B l km cle.v.g,Renbu-Z•edong . ••Qthrust J Pz-K •A / • pz.KA •' •'- ß K -Tg '• '" "' -' '" ccr o1• ½• x .- lO 10 . -._-_ mm

2o 2O Lhaaa block

suture zone (c) Plate 3. (a) Geologicmap of theZedong area, modified from Wang et al. [ 1983]and based on our fieldobservations. Map symbolsare the same as those used in Plates1 and2 (um,mafic/ultramafic complex).Cross sections (b) A-A' and(c) B-B'showing the relationship between the Renbu and Gangdesethrusts. Structural position of samplesZH-2b and ZH-3 arealso shown.

The second major thrust investigatedin this study is the Mesozoic metasedimentaryrocks of the Lhasa Block in its southdipping Renbu-Zedong thrust (RZT). This fault was first footwall (Plate 5a). Note that without recognitionof the GT, mappedby Chinese geologists[Wang et al., 1983] and later the field relationshipnear Gongga would suggesta normal included in regional compilations of southern Tibetan fault relation for the Renbu-Zedong fault, because it geology [Burg, 1983;Liu et al., 1988;Kidd et al., 1988]. The juxtaposes upper structural level rocks (low-grade RZT is clearly youngerthan the GT, becauseits hangingwall metasedimentaryrocks) in the hanging wall over lower stratalocally thrust over the trace of the GT, putting both the structural level rocks (granites and metamorphicrocks of hangingwall and footwall rocksof the GT below its footwall. greenschistto amphibolitefacies) in the footwall. However, Near Gongga, south of Lhasa (Plates 1 and 2), the RZT mesoscopicstructures such as asymmetricfolds and thrusts juxtaposesvery low-gradeTethyan metasedimentaryrocks in directly above the RZT indicate that the thrust is north its hanging wall and Cretaceous granitic and Paleozoic- directed. In addition, steeply southdipping cleavageis well 18,180 YIN ET AL.: GANGDESE THRUST SYSTEM

tOYlooitic

Plate 4. (a) South verging fold in the Xigaze Group in the hanging wall of the Gangdesethrust west of Xigaze; view to the northwest.The telephonepoles in the foregroundare about 7 m high. (b) The Gangdesethrust exposednear Zedong (see Plate 3a and Figure 1 for location). Cliff-forming unit is the mylonitic granite and marble in the hanging wall of the GT. Tc, Tertiary (?) conglomeratesdipping steeply to the south in the footwall. (c) The mylonitic shear zone above the Gangdesethrust. (d) Minor asymmetricfolds and boudinagestructures in the mylonitic shear zone. (e) Duplex systemin the hangingwall of the backthrustsystem northwest of Xigaze. YIN ET AL.- GANGDESE THRUST SYSTEM 18,181

S • N

Plate 4. (continued) 18,182 YIN ET AL.: GANGDESE THRUST SYSTEM

Plate 4. (continued)

developedin the Tethyan sedimentaryrocks immediately (40Ar/39Ar)K= 0.0225, (38Ar/3•Ar)c1 = 1.20 x 10-2 , above the fault in the Zedong area (Plate 5b), which also (37Ar/39Ar)ca=7.00 x 10-4, and (36Ar/39Ar)ca = 2.90 x 10-•. suggestsa northward thrusting along the RZT. The RZT itself All sampleswere step heated in a Ta cruciblewithin a double- is crosscutby a N-S trendinggraben system, the Yadong-Gulu vacuumfurnace and 40Ar/39Ar isotopic measurements rift, that initiated at 84-1 Ma (T.M. Harrison et al., The performedusing a VG 1200S automatedmass spectrometer Nyainqentanghlashear zone: Implications for TibetanPlateau operatedin the electronmultiplier mode. Details of flux uplift and onset of the Asian monsoon,submitted to monitorages, neutron irradiation, step heating,and isotopic Tectonics, 1994]. analysesare given by Harrison et al. [1992b]. Tabulated results of the argon isotopic analyses, uncorrected for neutron-producedinterferences except (38Aff39Ar)ci, are given Strike-Slip and Normal Faults in Table 1 using conventionaldecay constantsand isotope Both strike-slipand normal faults are observedlocally near abundances.The ZH-3 apatiteseparate was irradiatedin the the "Indus-Tsangpo"suture in the Xigaze area. Near Renbu HIFAR reactor,Lucas Heights,Australia, and analyzedby the (Plate 1), an east-west striking, left-slip fault juxtaposes fission track method using techniquesdescribed by Green [1986]. Resultsare given in Table 2. Cretaceoussedimentary rocks of the Xigaze Group to the southagainst Tertiary(?) conglomerates to the north,which unconformablyrest on topof theGangdese granites (Plates 6a and 6b). The structuralsignificance of this fault is not clear. Zedong However, its structuralposition and faults with similar trend to the northmapped by Chinesegeologists [Liu et al., 1988] 40Ar/39Ar and fission track thermochronometrywere suggestthat it maybe partof a transferfault systemlinking undertakenon two rock samplesin the hangingwall of the GT the NE-SW trending,late MioceneNyainqentanghla normal near Zedong (Plate 3) with the intent of constrainingthe faultsystem [Pan and Kidd, 1992]with a north-southtrending timing of thrusting. Samplelocations are shownon Figure 1. normal fault systemsouthwest of Xigaze. ZH-3 is from an undeformedhanging wall granodioritetaken West of Xigaze, right-slip faults trendingbetween about approximately 10 m above the top of the mylonitic shear N40øW and N10øE are also observed (Plate 2a). These faults zone. ZH-2b is a mylonitic granodiorite from within the displacedthe GT betweenseveral hundreds to a few >200-m-thick ductile shear zone. kilometers. Locally, theseright-slip faults bend and strike TheZH-3 hornblendeyields an 40Ar/39Ar age spectrum that parallelto thetrace of theGT to forma seriesof smallpull- dropsfrom initial agesof-50 Ma to agesover the last 80% of apartbasins (Plate 2a). The age relationshipbetween the gas release of 31.6+1.5 Ma (Table 1). This latter age is right-slipand left-slip faults is not clear.However, both sets substantiallyyounger than all other dated plutons from the of faultspostdate the GT andtheir initiation and development Gangdesebatholith [e.g., Copelandet al., 1987;Honegger et maysignal a markedchange in thestyle of deformation. al., 1982] and most certainly postdatesthe onset of collision. Clearly, this spectrumis disturbedand may have experienced both40Ar* lossand uptakeof excess40Ar* makingage Thermochronometry interpretationdifficult. We tentativelyinterpret the 32-Ma age to reflect the time of passagethrough the closure isotherm. Biotite from this sampleyields an integratedtotal Timing of Movement on the Gangdese Thrust fusionage of 30.1+0.5 Ma (Table 1). Mineral separatesof K-feldspar, hornblende,biotite, and The coexisting K-feldspar yields a spectrum that is apatite were obtained from fresh hand specimens. For the characterized by ages between about 24 and 29 Ma. 40Ar/39Aranalyses, mineral separates were irradiated together Isothermalduplicate steps taken over the first 5% of gas with Fish Canyon sanidineflux monitorsfor 45 hours in the releaseshow a strong correlationon a plot of AC1/K (the H-5 position of the Ford Reactor, University of Michigan. differencein C1/K betweenthe two isothermalsteps calculated Correction factors used for interfering neutron reactions were from38Arc1/39ArK) versus A40Ar*/K (i.e., the difference in YIN ET AL.- GANGDESE THRUST SYSTEM 18,183

Tet•yan strata

Plate 5. (a) The Renbu-Zedongthrust exposed near Gonggar. It juxtaposesthe Tethyanstrata over the Gangdeseplutonic rocks. (b) Southdipping cleavage (S 1) in Tethyansedimentary rocks in the hanging wall of the Renbu-Zedongthrust near Zedong. In many places,the cleavagetranspose the beddingcompletely, making the thicknessestimate of the Tethyan stratadifficult. 18,184 YIN ET AL.' GANGDESE THRUST SYSTEM

w E

Plate 6. (a) E-W striking left-slip faults near Renbu. Note that the red-sandstonemarker bed is offset laterally by several faults. (b) Subhorizontalstriations on left-slip fault surfaces.

apparent age between the two steps) that indicate a C1- correctedin this fashionyield a smoothage gradientfrom 25 correlatedcomponent of excessradiogenic argon of 40ArE/C1 to ~19 Ma. = 1.26_+0.07x 10-5 [Harrisonet al., 1993b]. Recognitionof Togetherwith the associatedArrhenius parameters (Figures this componentallows identificationof the fraction of excess 2d and 2e), we have recoveredthe thermalhistory (Figure 2c) 40Ar* presentin eachof thefirst 14 stepsin theage spectrum through the multidiffusiondomain model [Lovera et al., (i.e.,the product of the40ArE/C1 and CI/K) andthus permits us 1989,1991,1993; Harrison et al., 1991,1992b; Richter et al., to see through the contaminatingeffects of the Cl-correlated 1991; Fitz Gerald and Harrison, 1993; Leloup et al., 1993]. excess40Ar* (Figure 2b). Althoughthe age spectrumis By correcting for Cl-correlated excess argon, we have highly irregular over the first 5% of gas release, ages obtainedage informationfrom 0.2 to 99% of 39At released YIN ET AL.- GANGDESE THRUST SYSTEM 18,185 18,186 YIN ET AL.: GANGDESE THRUST SYSTEM

oooooooooooooooooooooooooooooooo

ooooooooooooooooooooooo$oo•o•o$oo$$$$$oo$$$$$$$$ $ YIN ET AL.- GANGDESE THRUST SYSTEM 18,187 18,188 YIN ET AL.: GANGDESE THRUST SYSTEM

-I

II II YIN ET AL.' GANGDESE THRUST SYSTEM 18,189

ß o

o II 18,190 YIN ET AL.: GANGDESE THRUST SYSTEM YIN ET AL.' GANGDESE THRUST SYSTEM 18,191

Table 2. ZH-3 Apatite FissionTrack Dating Results Age of Backthrusting CalculatedUsing a Zeta of 353 for NBS GlassSRM 612 A lower limit on the age of the backthrust northwest of Xigaze (Plate 2) is constrained by dating of a north-south Parameter Value trending andesitic dike (sample XH-10B2) that cuts both the backthrust and its hanging wall structures. K-feldspar from this rock yields a complex spectrumwith ages of--18 Ma over Numberof grains 12 the first 60% of gas release spectrum that then rise to >500 Ps,x 106cm '2 0.351 Ma (Table 1), undoubtedlydue to excessargon. Oscillations Ns 199 in agein theearly release again correlate with release of 38C1 Pi'x 106cm '2 0.565 and yield an 40ArE/C1 = 2.5+_0.5x 10-5 . Following Ni 3207 correction for Cl-correlated excess argon for steps up to (O•/Oi) 0.062+0.005 900øC (beyond which point no argon release from fluid P(X2) 43.2 inclusions is indicated, an age of 15.4_+1.5 (1•)Ma is r 0.772 calculated between 15 and 65% 39At released. pDx106 cm -2 1.447 About 1 km east of this dike, a granitic sill can be traced ND 154 within the Xigaze Formation for several kilometers. K- Age. Ma + lo 16+l feldspar from this sill (sample XH-11), corrected for C1- Mean tracklength. mm 12.78+0.39 correlatedexcess argon using 40ArE/C1 = 1.27_+0.04x 10-5, Number of confined tracks 49 yieldsan ageof 14.0_+1.0Ma between15 and90% of 39At S.D.. pm 2.77 release (Table 1) which is statisticallyindistinguishable from XH-10B2 K-feldspar. The coexisting XH-11 hornblende The parametersare Ps,spontaneous track density; N s, numberof yields an age over the last 60% of gas releaseof 18.3_+0.5Ma, spontaneoustracks counted; Pi' inducedtrack density; N i, numberof indicating that the K-feldspar ages are cooling dates. Since inducedtracks counted; ps/Pi, mean of individualcrystal ratios (+ 1 o) for no andesitic magmatism of this age has previously been thoseanalyses giving P(X2) < 5%; P(X2),probability of obtaining reportedin southeasternTibet, it is reasonableto assumethat observed•2 valuefor v degreesof freedom(v is thenumber of crystals thesetwo intrusivesare coeval and date a phaseof extensional minus one); r, correlationcoefficient between individual crystal track magmatism that occurred at 18.3_+0.5Ma. We concludethat counts(N s and Ni); PD' trackdensity measured in externaldetector backthrusting had terminated in this region by that time. adjacentto theglass dosimeter during irradiation; ND, numberof tracks This is consistent with our hypothesis that the pop-up countedin determiningPD; mean confined track length (+1o); S.D., structure is a manifestation of the Gangdese thrust, and that standarddeviation of the track lengthdistribution. thrusting ended during the early Miocene. The magmatism may reflect incipient collapse following late Oligocene-early Miocene crustal thickeningvia the GT. which correspondsto a temperaturerange of 160ø to --350øC. A cobble from the conglomerate in the footwall of this Uncertainty of the K-feldspar derived thermal history is thrust was dated in the hope of improving the Eocene (?) age estimatedto be _+25øCand -+0.2 Ma [Richter et al., 1991]. A constrainton the Qiuwu molasse[Wang et al., 1983] and thus secondsplit of ZH-3 wasrun (40ArE/C1 = 1.44_+0.25x 10 -5) narrowing the possible duration of movement on the backthrust. K-feldsparfrom a graniticcobble (sample XH-9A) but detailed information is obscured by the fact that the yields an age, once corrected for Cl-correlated excess argon samplewas fused after only 63% of the gas had beenreleased. (40ArE/C1= 7.9_+0.1x 10-5),of--90 Ma. Howeverthis age, Nonetheless,it yieldsa preciseage over the initial gas release which correspondsto the oldest known plutonism in the of 26.28_+0.06(ls) confirmingthat rapid coolingoccurred at Gangdese batholith [Schiirer et al., 1984], does not further that time. Also shown on Figure 2 are the coexisting hornblendeand biotite40Ar/39Ar agesplotted assuming constrainthe age of the Qiuwu. closuretemperatures (Tc) of--550øC [McDougalland Harrison, 1988] and 370+50øC (diffusion length scale, r = 0.34 mm Discussion [Copeland et al., 1987]; activation energy, E = 47.0 kcal/mol,frequency factor, D O= 0.077cm2/s [Harrison et al., Crustal Thickening in Southern Tibet 1985]; and coolingrate, dT/dt = 100øC/m.y.),respectively. Fissiontrack analysisof the ZH-3 apatiteyields an age of In our earlier tectonic model [Harrison et al., 1992a], we 16+1 Ma with a meantrack lengthof 12.8_+0.4gm (Table 2). proposedtwo scenariosfor the rapid cooling of the eastern Together,these results correspond to an age of 18 Ma for the Gangdese batholith in the early Miocene recorded by transitof this samplethrough the 110_+10øCisotherm (Figure thermochronometry:(1) movement on a north dipping 2). normal fault along the northern margin of the Gangdese The age spectrumfor ZH-2b K-feldsparis alsosuggestive of batholith causedrapid tectonicdenudation; (2) movementon a rapid cooling through the closure temperatureinterval at 27 north dipping thrust along the southern margin of the Ma (Figure 3). The old agesin the late stagesof gas release Gangdesebatholith brought the Gangdesebelt upward and are likely not geochronologically meaningful but rather southwardonto the Indian continent(the hanging wall was reflect excessargon. We have arbitrarily chosennot to use rapidly eroded during or immediately after the thrust resultsbeyond 90% 39At releasein the thermalcalculations. emplacement). Our field observations,together with earlier No Cl-correlatedexcess argon component was detectedin this Chinese mapping, suggestthat no significant normal faults sample. The log (r/to) plot of ZH-2b differs from ZH-3 in are present along the northern margin of the plutonic belt, being convex upward rather than concave indicating a ruling out hypothesis1. Insteadon the northernmargin we continuous (rather than discrete) distribution of subgrain have observed south dipping thrusts in three separate sizes, possibly reflecting deformation-related locations that are possibly backthrusts of the GT. recrystallization. The cooling histories of both rocks Recognitionof the GTS relegatesother mechanisms(climate directlyabove the GT are showntogether in Figure 3. change, tectonic denudation by normal faulting) [e.g., 18,192 YIN ET AL.: GANGDESE THRUST SYSTEM

91.880 91.92o 91.960 92.000

32.44e Pz-Mz

Pz-Mz . -• Pz-Mz

. Qal ß

Gangdesethrust ß shear zone 21 ß

24 Qal 34 . .o K-Tg ß" Plate4d Plate ...! 26 . .--- - ...- 32.38 e 22 )•9•Pz-Mz.•30 Q1 T$--r'- 40 --,-- 20 34 6• • syn • Tin Tin 26

32.34e •

Tin 65• i i

Figure 1. Simplified geologic map (see Plate 3a for map location) showingthe locationsof samplesZH-3 andZH-2b relativeto the traceof the Gangdesethrust. Also shownare the positions andangles from whichPlates 4b, 4c, and4d weretaken. Rockunits: Pz-Mz, metasedimentaryrocks of theLhasa block in thehanging wall of theGangdese thrust; K-Tg, Cretaceous to Tertiary granite (= Gangdesebatholith); ml, melangecomplex including highly shearedultramafic rocks and isoclinallyfolded cherty shale and fine-grainedcherty arenite; T s, Tertiaryconglomerate containing clastsof both the Gangdesegranite and Tethyan sedimentarystrata; syn, syeniteof unknownage and originin the footwallof the Gangdesethrust; T TM, Triassic strata of Tethyansedimentary sequence,they beddingis mostly transposedby axial cleavagedue to isoclinal folding; Qal, Quaternaryalluvium deposits. Foliation symbols:solid trianglesindicate gneissic or mylonitic foliations,and open symbolsindicate axial cleavagesuch those in the hangingwall of the Renbu- Zedongthrust. R-Z thrust,Renbu-Zedong thrust.

England and Molnar, 1990] to a supportingrole in explaining Regime 2. Sufficiently far above the thrust surface, the mid-Tertiary transition in cooling rate. cooling is dominated by erosional denudationin responseto We have in earlier papers emphasizedthe role of erosional thrust-relatedcrustal thickening. Again, becausethickening denudation in cooling midcrustal rocks following is diachronous, midcrustal rocks in the south will tend to be lithosphericthickening [Copelandet al., 1987; Richter et al., uplifted and denudedbefore rocksin the north. 1991; Harrison et al., 1992a]. An additionalcooling process Regime 3. A transition region exists between regimes 1 implicit in the GTS hypothesisis refrigerationof the hanging and 2 in which cooling is due to both processes. wall as a consequenceof thrusting [e.g., $hi and Wang, Near the thrustsurface (regime 1) we see evidencefor rapid 1987]. The interplay between these two boundaryconditions cooling beginningat 27 Ma that we interpret to result from (i.e., denudationat the upper boundary,lateral heat flow at the refrigeration by the cold Tethyan sediments. This event is base) producesthree mid-Tertiary thermal regimeswithin the observednot only at Zedong, but also 60 and 120 km further Gangdesethrust plate. west along strike, respectively,at Samye and Quxu (Plate 1) Regime 1. A zone exists directly above the thrust surface [Copeland et al., 1994]. This concordanceof ages within whose thickness is a function of the thermal time constant for regime 1 is perhaps the best evidence that thrusting was the inferred duration of thrusting (maximum duration -5 underwayat 27 Ma. Note that the observedpattern of cooling m.y.). In this zone, the thermal evolution is dominatedby in the Gangdesebatholith is the oppositeof what would occur heat loss to the footwall. Becausearrival of the footwall ramp if the RZT had been the dominant structural control. is diachronousacross the Gangdesebatholith in the direction The geological relationshipscharacterizing the rocks east of thrust transport,this conductivelength scale (i.e., the zone of Zedong are similar in at least two respectswith those refrigeratedby the footwall) might vary from ~10 km above ~1100 km to the west. In the vicinity of Mount Kailas, the the thrust surfacein the south to only a few kilometersin the absenceof the Xigaze strata is again correlatedwith an early north. Miocene uplift/unroofing pulse suggestingthat the GTS may YIN ET AL.: GANGDESE THRUST SYSTEM 18,193

32 -2

30 ZH-3 -3 28 -4 26 -5 24

22 -6

i i i I i 2O -7

-8

[] K-spar measured -9

ß ß K-spar model -10

-11

-12 7 9 11 13 Atmospheric-corrected 22 10000/T (øK) C!-corrected 1.4 model

16 1.0 o 2 4 6 Cumulative% 39Arreleased

0.6 hornblende I / I

0.2 -. d K-sparmodel ' / ' apatite C E o -0.2 15 20 25 30 35 0 20 40 60 80 100 Age(Ma) Cumulative% •Ar released Figure2. ZH-3(a) 40Ar/39Ar spectrum and modeled fit; (b) an expanded view of the first 8% of gas releaseshowing the change to theage spectrum from correction for Cl-correlatedexcess 40Ar*; (c) the calculatedthermal history (the portionconstrained by the K-feldsparis shownby the thick line); (d)the Arrhenius plot calculated from 39Ar diffusivities together with the model fit; and(e) the log (r/to) plot. Modelfit parametersare, E = 41.7kcal/mol; log (Do/r2) and volume fractions of the eight domainsare (1) 7.86/0.0005,(2) 5.37/0.023, (3) 5.28/0.041, (4) 2.41/0.361, (5) 2.40/0.155,(6) 1.10/0.285,(7) 1.09/0.133,(8) -0.83/0.0015.Also shownin (c)are 40Ar/39Ar agesfor biotite and hornblende andan apatite fission track result (see text for T c information).

havebeen active over at least1100 km alongstrike [Liu et 12_+6mm/yr, which is similar to the presentrate of al., 1988; Harrison et al., 1993a]. Variations in the convergence at the southern edge of the collision zone magnitudeof displacementalong strike of theGT maybe due [Molnarand Deng, 1984'Avouac and Tapponnier,1993]. to the transferof displacementto thrustsin the Tethyan Himalaya(= SouthernTibetan Zone), irregularity of the Relationship of the GTS to Previous northern boundary of the Indian continent, or different Thermochronometry curvaturesof the Gangdesearc and the northern Indian continental margin. We have previouslypointed to thermochronological resultsin the Gangdesebatholith as evidencethat this belt in Minimum Slip and Slip Rate Along the Gangdese Thrust southernTibet experiencedrapid denudationin responseto crustalthickening during the late Oligocene-earlyMiocene [Copelandet al., 1987, 1994; Richter et al., 1991; Harrison The rate of slip along the GT can be calculatedfrom its et al., 1992a, 1993a]. Our presentobservations appear to minimumdisplacement (46+9 km) dividedby the durationof providea mechanismthat explainsthe patternof cooling thrusting,estimated from thermochronometry to be 4+2 m.y. agesthroughout the Gangdesebatholith. Assuminga flat- This latter figure is based on considerationsof the time rampgeometry for the GT, with the flat followingthe brittle- differencebetween the onsetof thrust-relatedcooling near ductile transitionat -15 km depth,the tip of the footwall Zedong(at 27 Ma) and the time at whichuplift-related ramp would propagatenorthwards at the samerate as the GT, denudationwas initiated -60 kmto thenorth (at circa23 Ma), estimatedto be at least12 mm/yr. At thisrate, the Gangdese describedin the next section[Copeland et al., 1987, 1994; batholithnear Lhasa, originally restingon the thrust flat, Richteret al., 1991]. Thissuggests a minimum slip rate of would not be transportedup the ramp of the GT (and thus 18,194 YIN ET AL.: GANGDESE THRUST SYSTEM

ZH-2b 35

• K-•parmeasured I • - - -K-•par model

-8 []K-spar measured -9 ß K-spar model b -10

-11 i • I ] I • 0 20 40 60 80 100 7 9 11 13 Cumulative% •Ar released 10000/T (øK)

1.8

ZH-3 thermal histol• 1.4 K-spar measured ZH-2b K-spar model ,,. K-spar model

1.0

0.6

l 0.2

½ d

I , , • 100 i i I J I i i , I .... I 0 20 40 60 8O lOO 15 20 25 30 Cumulative% •Ar released Age (Ma) Figure 3. ZH-2b K-feldspar(a)40Ar/39Ar spectrumand modelledfit; (b) the Arrheniusplot calculatedfrom 39At diffusivitiestogether with the modelfit, (c) log (r/ro) plot,and (d) calculated thermal history (heavy line). Also shown is the thermal history derived for ZH-3. Model fit parametersare: E = 46.5 kcal/mol;log (Do/r2) andvolume fractions of the ninedomains are (1) 5.80/0.108, (2) 4.88/0.028, (3) 4.87/0.247, (4) 4.86/0.091, (5) 4.11/0.222, (6) 3.38/0.124, (7) 2.61/0.060, (8) 2.49/0.061, 0.57/0.059. uplifted and subsequentlydenuded) until 23-22 Ma. The assumptionsregarding the GT fault geometryare correct,a widespread preservation of precollisional Linzizong total displacementof 60 kin wouldbe implied. volcanic rocks along the northern margin of the Gangdese Thermochronologicalresults from two regime 2 locations Shan [Liu et al., 1988] and the relatively modestamount of in the Gangdesebatholith, about 30 km northof Quxu (Plate post-22Ma denudation(and thusthickening) in the vicinity 1), reveal clear evidenceof rapid cooling beginningat about of Lhasa (<4 km) [Copeland et al., 1994] suggestthat the 20 Ma [Copelandet al., 1987;Richter et al., 1991]. These northerntip of the footwall ramp did not extendbeyond the data were interpretedto reflect enhanceddenudation due to northernmargin of the Gangdesebatholith. Thus we infer isostacy-drivenuplift. Because there is a-2 m.y. lag that movement on the GT terminated at 23-22 Ma. If our between the onset of rapid denudation and significant

Plate 7. Cross-sectionalviews of the tectonic developmentof the southernTibetan plateau and Higher Himalaya. (a) 100-60 Ma: North dipping subductionconsuming the oceaniclithosphere betweenIndia and Asia. (b) 60-50 Ma: Volcaniceruptions that producedthe Linzizhongvolcanics. (c) 50-35 Ma: Subductionof the thinnednorthern margin of the Indian continentbeneath the thin crustof the southernmargin of the LhasaBlock that is madeof mostlythe accretionarywedge. (d) 35-27 Ma: Further convergencebetween India and Asia along the suturezone that produced thickenedcrust in the southernmargin of the LhasaBlock. At about27 Ma, the Gangdesethrust beganto develop.(e)-(h) 27-20 Ma: Structuralevents occurred during this period are (from older to younger),(1) movementalong the Gangdesethrust that brought the Gangdeseplutonic rocks over the Tethyansediments and the developmentof the backthrustthat putsthe Xigaze strataover the Gangdeseplutonic rocks and Tertiary conglomerates, and (2) the developmentof the Renbu-Zedong thrust as a back thrust of the Main Central Thrust and the thrust toe of the High Himalayan normal fault system. YIN ET AL.: GANGDESE THRUST SYSTEM 18,195

100-60 Ma

Cretaceous thrust belt Xlgaze Group- Forearc Basin Indian Plate • •hasaBlockGangdeseCret•/eousBathollth lOO km

/ Subduction Complex Qlangtang Block

(a) 60-50MaI Cretaceous thrust belt Indian Plate Xlgaze Group Gangdese Bathollth AsianPlate•• Tv• CretaceousSeaway lOO km

Qlangtang Block

(b) Xlgaze Group Cretaceous Seaway IndianPlate•gdese Bathollth TV • A,lan Plate • • / lOO km

future emslonal surface (c)

35-27 Ma Xlgaze Group Indian Plate angde•eBathollthTV Asian Plate \

100km

future Gangdese thrust

(d) 18,196 YIN ET AL.: GANGDESE THRUST SYSTEM

initiation of cooling in the midcrust [Copeland et al., 1987; between the MCT and NHF; using an elastic approximationit Richter et al., 1991], crustal thickening at these locations could be shown that the state of stressis critically dependent can be inferred to have occurred at 23-22 Ma. The on pore-fluid pressuresboth along and within the wedge- coincidence between this age and the presumed time of shaped hanging wall of the MCT. Within the model arrival of the GT in the northernGangdese batholith appears framework, the onset of partial melting [LeFort, 1986] above to implicate it as the thickeningmechanism responsible for the thrust zone produced the parent of the magma that later the observed pulse of late Oligocene-early Miocene uplift. crosscutsthe NHF. Anatexis within the hanging wall would Because virtually all active crustal scale thrusts around the act as a tremendous sink for the volatiles evolved during Tibetan plateau surface at elevations < 2000 m, we may metamorphism of the lower plate, thereby reducing the pore presume that the southeasternTibet was at an elevation fluid pressurealong the thrust and creating stressconditions between 0 and 2000m during the development of the in the hanging wall favorable for initiation of normal Gangdesethrust system. faulting. This could explain the virtual synchroneity of the MCT and the earliest phase of movement on the NHF. Once Age Constraints on Movement of the Main the NHF was initiated, continued motion at a low angle is Central Thrust likely if a high porefluid pressurein the fault zoneis present [e.g., Axen, 1992]. The Main Central thrust (MCT), generally believed to be North-south extensionin the Higher Himalaya between the the earliest of the crustal scale thrustsdeveloped on the Indian early Miocene and Pliocene may have been partially or continent, is thought to have experienced between 100 and completely compensated by motion on the RZT. L. 250 km of south directed displacement[e.g., Gansser, 1981; Ratschbacher et al. (Distributed deformation in southern and Lyon-Caen and Molnar, 1983, 1985; Bouchez and P•cher, western Tibet during and after the India-Asia collision, 1981]. Although cooling ages of minerals from the shear submitted to Tectonophysics, 1993) obtained a 17.5+0.9 Ma zone place an approximate lower age limit of-17 Ma (see K-Ar age from white mica filling a tension gash in the RZT Hubbard and Harrison [1989] for a review), details of the shear zone near Renbu. Although possibly a cooling age, the timing of thrusting, particularly of initiation, remain poorly low grade of metamorphismaffecting these rocks likely did known. From thermobarometric and thermochronological not exceed the white mica closuretemperature of-350-400øC. considerations, Hubbard and Harrison [1989] argued that The circa 18 Ma age is consistentwith the known timing of metamorphismof the MCT in responseto tectonismwas in the NHF. Our proposed structural model suggeststhat the progress at 20.9+0.3 Ma near Mt. Everest. Parrish and extension was restricted to the Higher Himalaya and Hodges [1993] concludedthat ductile deformationwithin the synchronouswith both contraction south (MCT) and north MCT near Annapurna terminated at 22+1 Ma. The oldest (RZT) of the normal fault system. established age of anatectic material thought to be derived from MCT related activity is the 24-Ma Makalu leucogranite [Schiirer, 1984]. Copelandet al. [1990,1991] concludedthat Role of Precollisional Thrusting in the anatexis due to thrusting was probably largely completed by Formation of the Tibetan Plateau 18 Ma, suggestinga similar age for terminationof large-scale slip. Whether the presentlyhigh elevationof the Tibetan plateau Our thermochronologicalresults from the GT near Zedong was caused by the Indo-Eurasia collision alone, or was may bear on the age of initiation of the MCT. As discussed contributedto by crustal thickening prior to the collision, is above, we infer from geological and thermochronological a controversial [Kidd et al., 1988; England and Houseman, considerations[Copeland et al., 1987; Richter et al. 1991; 1986; England and Searle, 1986] issue that is central to Harrison et al., 1992a] that the GT was active between 27 and determining the crustal mass balance during collision [e.g., ~23 Ma. It is generally believed that the initiation of major Le Pichon et al., 1992; England and Houseman,1986; Richter thrust motion in the Himalaya has moved sequentiallytoward et al., 1992]. Our preliminary mapping in the Chu0qin area, the foreland during the collision [Gansser, 1981]. If so, this south-central Tibet (Figure 1), suggeststhat crust in this and other lines of evidence mentioned above constrain the regionwas thickenedby at least40-50% over a belt 40-70 km timing of the most significant MCT activity to be between wide by thrusting and folding. There, Paleozoic-Mesozoic about 24 and 21 Ma. stratawere involved in southwardthrusting and southverging folding. Thrustswere intrudedby an Early Cretaceousgranite (T. M. Harrison, unpublished data, 1993) and rest Relationship Between the MCT, North Himalayan unconformably below Paleocene Linzizong volcanic strata. Fault, and Renbu-Zedong Thrust Similar structures are also observed north of Lhasa in the Mezhugongka area (Plate 1), where the folded Cretaceous It is possiblethat the MCT, the North Himalayan(normal) Takena Formatic>n is overlain unconformably by the fault (NHF), and the RZT are related productsof the same PaleoceneLinzizong volcanics. Depositionin southernTibet evolving mechanicalsystem. In the Rongbuk valley, an of widespread Cretaceous clastics, containing medial earlyphase of the North Himalayannormal fault system[Burg Cretaceousmarine beds,may have occurredin a forelandbasin et al., 1984;P•cher, 1991;Burchfiel et al., 1992]is cut by an producedby loading of thrust sheetsto the north after closure essentiallyundeformed granitoid that hasbeen precisely dated of ocean(s) between the Lhasa and Qiangtang blocks (Plates at 21.7+0.5 Ma [cf. Copeland et al., 1988; Parrish, 1990; 7a and 7b). Somewhat thickened crust may have remained T.M. Harrison et al., Identification of inherited monazite in from this convergentevent in the northernpart of the Lhasa the Manasluleucogranite by 208Pb/232Thion microprobe block at the time of initiation of the Indo-Asian collision. dating,submitted tc .•fature,1994], placinga lower age limit on fault motion. Th,asthe earliestphase of movementappears Why Did the Gangdese Thrust Begin Movement to have been prior to 21.2 Ma, possibly coeval with our 20 m.y. After the Initial Collision? estimate made above fcr the movement on the MCT. Later movement within the NHF certainly postdates -11 Ma Although somewhat uncertain, it seems likely that the [Maluskiet al., 1988] and may be relativelyrecent [Burchfiel collision of India with southernAsia was underway by about et al., 1992]. Yin [1993] proposed a mechanical link 50 Ma [Dewey et al., 1989; ?eltzer and Tapponnier, 1988]. YIN ET AL.' GANGDESE THRUST SYSTEM 18,197

I 27-20Ma- "A" ! Xlgaze Group TM .Kg-Tg Gangdese thrust Gangdesethrust• xg

s n t

Ductile Flow?

Subductlon Complex oceanic crust of the / Indian Plate I 27 - 20 Ma- "B"I Xlgaze Group (e) Renbu-Zedongthruat/ TM K

ß .n

oceanic crust of the •lndlan Plate 27-20 Ma -"C" I Xlgaze Group MCT STDS GT K (• Indian FIZT/ TM Aalan Plate

oceanic cruat of the Indian Plate J20Ma - PresentI GangdeseBathollth (g) MCT Xlgaze Group Indian• RZT• Plate GT/VK AalanPlate

oceanic crust of the Indian Plate (h)

Plate 7. (continued) 18,198 YIN ET AL.: GANGDESE THRUST SYSTEM YIN ET AL.: GANGDESE THRUST SYSTEM 18,199

However,the southernmostthrust system of the LhasaBlock, irregular topography of the Moho due to this crustal theGTS, apparently did not begin to moveuntil about 27 Ma shorteningevent may have been smoothedby ductile flow in or-23 m.y. aftercollision began. This observationappears the lower crust (Plate 7f). Late in the 27-20 Ma period, the to contradictthe prediction of Englandand Houseman [1986] Main Central Thrust (MCT) and the North Himalayan (normal) that deformation due to indentation of India should have fault (NHF) developed.The RZT that postdatesthe GT appears startedfirst along the southernmargin of the Eurasian to have developedcoevally with the MCT and early motion on continent.Why thenwas significant deformation along the the NHF; its hanging wall may have formed as a backthrustof southernmargin of Asiadelayed for so long? It is possible the MCT and the thrust toe of the NHF accommodating the that tectonicextrusion along major strike-slipfaults, such as local extension strain in the Higher Himalaya (Plates 7d and the Red River ductile shear zone [Peltzer and Tapponnier, 7g). Further northward thrusting along the Renbu-Zedong 1988;Tapponnier et al., 1982],and crustal shortening along fault obscuresthe trace of the GT and the suture along which majorcrustal-scale thrusts farther north of thesuture [Chang the ocean(s)between India and Asia was consumed(Plate 8e). et al., 1986] permittedthe deformationfront to migrate southwardfrom the KunlunShan region south of the Qiadam Conclusions basintowards the southernmargin of the LhasaBlock (rather than the reverse). Structural and thermochronological investigations conducted in southern Tibet reveal that intracontinental Tectonic Model thrustinghas been the dominantcause for the formationof the southeasternTibetan (Qinghai-Xizang)Plateau since the late Oligocene. Two major thrustsystems are documented:the On the basis of the geologic relationships and north dipping Gangdesesystem and the younger south thermochronologicalanalyses discussed above, we proposea tectonic model for the evolution of the southeastern Tibetan dippingRenbu-Zedong thrust system. West of Renbu,the Gangdesethrust juxtaposes the Xigaze Group over the plateau since the late Cretaceous, shown in both cross- Tethyansedimentary rocks of the Indianplate, whereas east of sectional(Plate 7) and map views (Plate 8). Note that Figure 7 Lhasa, the fault juxtaposes the Gangdesebatholith over depictsonly the tectonichistory of the GTS eastof Renbu. Tethyan metasedimentaryrocks. Near Zedong,the GT is Subduction of the Indian plate beneath the Eurasian markedby a >200-m-thickmylonitic shear zone that consists continentbetween 100 and 50 Ma producedan Andean-margin of deformedgranite and metasedimentaryrocks. A major system in southernmost Asia that includes the Gangdese southdipping backthrust in the hangingwall of the GT puts batholith and Linzizong volcanics as the volcanic arc, the the Xigaze Group over Tertiary conglomeratesor the Xigaze strataas forearc deposits,and possiblyan accretionary Gangdeseplutonic rocks in places. The styleof deformation complex (Plate 7a). The northern edge of the Indian in theXigaze Group is characterizedby widelydeveloped folds continent, a passive margin since the Paleozoic, may have that are southverging directly above the GT to the southand had an irregular shape resulting in differential shortening north vergingfolds and north directedduplex systems above strain within the Lhasa block (Plates 1 and 8a). Intense the backthrustto the north. The timing of movement on the volcanic eruptionsfrom the Gangdesearc (Plate 7a) caused GT is constrained to be between 27 and -23 Ma based on deposition of the Linzizong volcanics along the southern 40Ar/39Ar thermochronologywith a minimumdisplacement margin of Asia largely between 60 and 50 Ma [Pan, 1992]. estimatedto be 46+9 km giving a minimum slip rate along Initial contact betweenthe Indian and Asian continentsbegan the GT of 12+6 mm/yr. The youngerRZT is thrustover the at -50 Ma (Plate 7a). Between 50 and 35 Ma, the thin crust of traceof the GT in many localitiesobscuring its exposure.The the northern Indian continent was subducted beneath the age of the RZT is constralnedto be between23-8 Ma. The accretionarycomplex (Plate 7a). Becausethe northernmargin suture zone between India and Asia has been complexly of India had relatively thin crust prior to collision, plate modifiedby the regionallydeveloped GT andRZT systemsand convergenceduring this period of 50 to 35 Ma may have been locallydeveloped left-slip and normalfault systems. partially accommodatedalong the north dipping suture zone and produced no significant mountain belts in southernmost Tibet and northern India (Plate 7a). The rest of the Acknowledgments. We thank Li Qi (State convergencecould have been accommodatedby thrusts and SeismologicalBureau), Yiao Zhong-Fu (Xizang Bureau of strike-slip faults further to the north and east, such as the Red Geology), and Barshang Tseren (Xizang Seismological River ductile shear zone, during this period. In map view Bureau)for assistingin the field, Wang Jo and Li (Niang Ci (Plate 8b), the proposed irregular geometry of the Pi) Shi-fu for their careful driving, Dave Fosterfor the fission- northeastern Indian continent would have caused, in this trackage, Matt Heizlerfor assistancewith the 40Ar/39Ar small region, a diachronouscollision of India with Asia (i.e., measurements,Mike McWilliams and an anonymous reviewer collision began first to the east and propagatedto the west). for useful suggestionsto improve clarity of the manuscript Further convergencebetween India and Asia occurredduring and G. Lister for microscopicexamination of the senseof 35 to 27 Ma along the north dipping suture and may have shearindicators. We are grateful for the leadershipof Zhou produceda low-elevation mountain range along the southern Xinhua, who skillfully organized the field operation. This margin of the Asian continent. The GT first developed researchwas supportedby NSF grants EAR-9118125 and between 27 and 23 Ma. The style of deformationwithin the EAR-9118827, the Department of Energy, Institute of GTS changed along strike, perhaps as a responseto the Geophysicsand PlanetaryPhysics of LawrenceLivermore irregular geometry of the northern edge of the Indian National Laboratory, and the Chinese National Natural continent(i.e., a single thrust (GT) east of Renbu, and a pop- Science Foundation grant 49173161. up structurewest of Renbu) (Plates 8b and 8c). During this interval, the GT east of Renbu cut through the Gangdese References batholith (Plate 7e) and carried the plutonic rocks over the Xigaze forearc deposits(Plates 7f and 8c). Underthrustingof the Xigaze Group and the Tethyan sedimentscaused rapid Allegre, C.J., et al., Structureand evolutionof the Himalaya- cooling of rocks directly above the GT. Movement along the Tibet orogenic belt, Nature, 307, 17-22, 1984. Gangdesethrust contributed to creationof the Gangdeserange Argand, E., La tectonique de l'Asie, Proc. 13th Int. Geol. along the southern margin of the Tibetan plateau. The Congr., 7, 171-372, 1924. 18,200 YIN ET AL.: GANGDESE THRUST SYSTEM

Avouac, J.P., and P. Tapponnier, Kinematic model of active Zagros, Hindu Kush, Himalaya-GeodynamicEvolution, deformation in central Asia, Terra Abstr., 5, 252-253, GeodynamicSer., vol. 3, edited by H.K. Gupta and F.M. 1993. Delany, pp. 111-121, AGU, Washington,D.C., 1981. Axen, G.J., Pore pressure, stress increase, and fault Green, P.F., On the thermo-tectonic evolution of Northern weakening in low angle normal faulting, J. Geophys. England:Evidence from fission track dating, Geol. Mag., Res., 97, 8979-8991, 1992. 123, 493-506, 1986. Bouchez, J.L. and A. P•cher, Himalayan Main Central Thrust Harrison, T.M., I. Duncan, and I. McDougall, Diffusion of pile and its quartz-rich tectonites in central Nepal, 40Ar in biotite:Temperature, pressure and composition Tectonophysics, 78, 23-50, 1981. effects, Geochim. Cosmochim. Acta, 49, 2461-2468, Burchfiel, B.C., and L.H. Royden, North-south extension 1985. within the convergentHimalayan region, Geology, 13, Harrison,T.M., O.M. Lovera and M.T. Heizler, 40Ar/39Ar 679-682, 1985. results for multi-domain sampleswith varying activation Burchfiel, B.C., Z. Chen, K.V. Hodges,Y. Liu, L.H. Royden, energy. Geochim. Cosmochim. Acta, 55, 1435-1448, C. Deng, and J. Xu, The South Tibetan Detachment 1991. System, Himalayan orogen: Extension contemporaneous Harrison, T.M., P. Copeland, W.S.F. Kidd, and A. Yin, with and parallel to shorteningin a collisional mountain Raising Tibet, Science, 255, 1663-1670, 1992a. belt, Spec. Pap. Geol. $oc. Am., 269, 1-41, 1992. Harrison, T.M., W. Chen, P.H. Leloup, F.J. Ryerson, and P. Burg, J.P. (Compiler) Carte g6ologiquedu suddu Tibet, scale, Tapponnier, An early Miocene transition in deformation 1:500,000, Cent. Natl. de la Rech. Sci., Paris, 1983. regime within the Red River fault zone, Yunnan, and its Burg, J.P., M. Brunel, D. Gapais, G.M. Chen, and G.H. Lin, significance for Indo-Asian tectonics, J. Geophys.Res., Deformation of leucogranites of the crystalline Main 97, 7159-7182, 1992b. Central Sheet in southernTibet (China), J. Struct. Geol., Harrison, T.M., P. Copeland, S. Hall, J. Quade, S. Burner, 6, 535-542, 1984. T.P. Ojha, and W.S.F. Kidd, Isotopic preservation of Chang, C. et al., Preliminary conclusionsof the Royal Himalayan/Tibetan uplift, denudation, and climatic Society and Academia Sinica 1985 geotraverseof Tibet, histories in two molasse deposits, J. Geol., 101, 159- Nature, 323, 501-507, 1986. 177, 1993a. Copeland,P., T.M. Harrison,W.S.F. Kidd, X. Ronghua,and Harrison, T.M., M.T. Heizler, and O.M. Lovera, In vacuo Z. Yuquan, Rapid early Miocene accelerationof uplift in crushingexperiments and K-feldspar thermochronometry, the GangdeseBelt, Xizang-southernTibet, and its bearing Earth Planet. Sci. Lett., 117, 169-180, 1993b. on accommodation mechanisms of the India-Asia Hodges, K., B.C. Burchfiel, Z. Chen, T. Housh, D. Lux, R. collision, Earth P;anet. $ci. Lett., 86, 240-252, 1987. Parrish, and L.H. Royden, Rapid early Miocene tectonic Copeland, P., R.R. Parrish, and T.M. Harrison, Identification unroofing of the metamorphic core of the Himalaya: of inherited radiogenic Pb in monazite and its Evidence from the Qomolangma(Everest) region, Tibet, implications for U-Pb systematics,Nature, 333, 760-763, Geol. Soc. Am. Abstr. Programs, 23, 372, 1991. 1988. Honegger, K., V. Dietrich, W. Frank, A. Gansser,M. Th6ni, Copeland,P., T.M. Harrison,and P. LeFort,Age and cooling and V. Trommsdorff, Magmatism and metamorphismin history of the Manaslu granite: Implications for the Ladakh Himalayas (the Indus-Tsangposuture zone), Himalayan tectonics, J. Volcanol. Geotherm. Res., 44, Earth Planet. Sci. Lett., 60, 253-292, 1982. 33-50, 1990. Hubbard,M.S., andT.M. Harrison,40Ar/39Ar age constraints Copeland,P., T.M. Harrison, K.V. Hodges,P. Mar6ujol, P. on deformationand metamorphismin the MCT Zone and LeFort, and A. P•cher, An early Pliocene thermal Tibetan Slab, eastern Nepal Himalaya, Tectonics, 8, 865- perturbationof the Main Central Thrust, Central Nepal: 880, 1989. Implications for Himalayan tectonics,J. Geophys.Res., Kidd, W.S.F., Y. Pan, C. Chang,M.P. Coward-,J.F. Dewey, A. 96, 8475-8500, 1991. Gansser, P. Molnar, R.M. Shackleton, and Y. Sun, Copeland,P., Y. Pan, T.M. Harrison, W.S.F. Kidd, M. Roden, Geological mapping of the 1985 Chinese-British and W. Chen, Thermal evolution of the Gangdese (Xizang-Qinghai) Geotraverse route, Philos. Trans., R. batholith, southern Tibet: A history of episodic Soc. London, Ser. A, 327, 287-305, 1988. unroofing, Tectonics, in press, 1994. LeFort, P., Metamorphism and magmatism during the Dewey, J.F., and K. Burke, Tibetan, Variscan and Precambrian Himalayan collision, in Collision Tectonics, edited by basement reactivation: Products of continental collision, M.P. Coward and A.C. Ries, Geol. Soc. Spec. Publ. J. Geol., 81, 683-692, 1973. London, 19, 159-172, 1986. Dewey, J.F., R.M. Shackleton, C. Chang and Y. Sun, The Leloup, P.H., T.M. Harrison, F.J. Ryerson, W. Chen, Q. Li, tectonic evolution of the Tibetan Plateau, Philos. Trans. P. Tapponnier, and R. Lacassin, Structural,petrological R. Soc. London, A 327, 379-413, 1988. and thermal evolution of a Tertiary ductile strike-slip Dewey, J.F., S. Cande, and W.C. Pitman, Tectonic evolution shear zone, Diancang Shan, Yunnan, J. Geophys.Res., of the India-Eurasiacollision zone, Eclogae Geol. Helv., 98, 6715-6743, 1993. 82, 717-734, 1989. Le Pichon, X., M. Fournier, and L. Jolivet, Kinematics, England, P., and G. Houseman, Finite strain calculations of topography, shortening, and extrusion in the India- continentaldeformation, 2, Comparisonwith the India- Eurasia collision, Tectonics, 11, 1085-1098, 1992. Asia collision zone, J. Geophys.Res., 91, 3664-3676, Liu, Z.Q., et al., GeologicMap of Qinghai-XizangPlateau and 1986. its Neighboring Regions (in Chinese), scale, 1:500,000, England,P., and P. Molnar, Surfaceuplift, uplift of rocks,and Chengdu Inst. Geol. and Min. Resour., Acad. Sini., Geol. exhumationof rocks, Geology, 18, 1173-1177, 1990. Publisher, Beijing, 1988. England, P., and M. Searle, The Cretaceous-Tertiary Lovera, O.M., F.M. Richter, and T.M. Harrison,40Ar/39Ar deformationof the Lhasa block and its implicationsfor geothermometry for slowly cooled samples having a crustal thickening in Tibet, Tectonics, 5, 1-14, 1986. distribution of diffusion domain sizes, J. Geophys. Res., Fitz Gerald,J.D., and T.M. Harrison,Argon diffusion domains 94, 17,917-17,935, 1989. in K-feldspar, I; Microstructures in MH-10, Contrib. Lovera, O.M., F.M. Richter, and T.M. Harrison, Diffusion Mineral. Petrol., 113, 367-380, 1993. domainsdetermined by 39Arrelease during step heating, J. Gansser,A., The geodynamichistory of the Himalaya, in Geophys. Res., 96, 2057-2069, 1991. YIN ET AL.: GANGDESE THRUST SYSTEM 18,201

Lovera, O.M., M.T. Heizler, and T.M. Harrison, Argon Richter,F.M., O.M. Lovera,T.M. Harrison,and P. Copeland, diffusion domains in K-feldspar, II; Kinetic propertiesof Tibetan tectonicsfrom a single feldsparsample: An MH- 10, Contrib. Mineral. Petrol., 113, 381-393, 1993. applicationof the 40Ar/39Ar method,Earth Planet. Sci. Lyon-Caen, H., and P. Molnar, Constraintson the structureof Lett., 105, 266-276, 1991. the Himalaya from an analysisof gravity anomoliesand a Richter,F.M., D.B. Rowley,and D.J. DePaolo,Sr isotope flextural model of the lithosphere,J. Geophys.Res., 88, evolution of seawater:The role of tectonics,Earth Planet. 8171-8191, 1983. Sci. Lett., 109, 11-23, 1992. Lyon-Caen, H., and P. Molnar, Gravity anomalies,flexure of Sch•irer,U., Theeffect of initial230Th disequilibrium on the Indian plate, and the structure,support and evolution young U-Pb ages:The Makalu case, Himalaya,Earth of the Himalaya and Ganga Basin, Tectonics,4, 513-538, Planet. Sci. Lett., 67, 191-204, 1984. 1985. Sch•irer,U., R. Xu, andC. Allegre,U-Pb geochronology of Maluski, H., P. Matte, and M. Brunel, Argon 39-Argon 40 Gangdese (Transhimalayan)plutonism in the Lhasa- dating of metamorphic and plutonic events in the North Xigaze region, Tibet, Earth Planet. Sci. Lett., 69, 311- and High Himalayas belts (southern Tibet-China), 320, 1984. Tectonics, 7, 299-326, 1988. Sch•irer,U., R. Xu, andC. Allegre,U-(Th)-Pb systematics and Mattauer, M., Intracontinental subduction,crustal stacking agesof Himalayanleucogranites, south Tibet, Earth Planet. wedge and crust-mantle decollement, in Collision Sci. Lett., 77, 35-48, 1986. Tectonics, edited by M.P. Coward and A.C. Ries, Geol. Searle,M.P., et al., The closingof Tethysand the tectonicsof Soc. Spec. Publ. London, 19, 35-60, 1986. the Himalaya, Geol. Soc.Am. Bull, 98, 678-701, 1987. McDougall, I., and T.M. Harrison, Geochronology and Shi, Y., and C.Y. Wang,Two-dimensional modeling of the; Thermochronologyby the 40Ar/$9ArMethod, 212 pp., P-T-t pathsof regionalmetamorphism in simpleoverthrust Oxford University Press,New York, 1988. terrains, Geology, 15, 1048-1051, 1987. Molnar, P., and Q. Deng, Faulting associated with large Tapponnier,P., G. Pelltzer,A.Y. Le Dain, R. Armijo,and P. earthquakesand the averagerate of deformationin Central Cobbold, Propagating extrusion tectonics in Asia: New and Eastern Asia, J. Geophys. Res., 89, 6203-6227, insightfrom simpleexperiments with plasticine,Geology, 1984. 10, 1339-1384, 1982. Pan, Y., Unroofing history and structural evolution of the Wang,X., Z. M. Li, andQiangba-Xiyiao, Geologic Map of the southernLhasa Terrane, Tibetan Plateau: Implications for Xigaze-Zedong Region (in Chinese), scale, 1:100,000, the continental collision between India and Asia, Ph.D. Xizang (Tibet) with report,568 pp., Xizang Bur. Geol., thesis,395 pp., State Univ. of N.Y. at Albany, 1992. Lhasa, 1983. Pan, Y., and W.S.F. Kidd, Nyainqentanglhashear zone: A late Yin, A., Mechanicsof wedge-shapedfault blocks,I; An elastic Miocene extensional detachment in the southern Tibetan solutionfor compressionalwedges, J. Geophys.Res., 98, Plateau, Geology, 20, 775-778, 1992. 14,245-14,256, 1993. Parrish, R.R., U-Pb dating of monazite and its applicationto geological problems, Can. J. Earth Sci., 27, 1426-1431, 1990. ChenW., Instituteof Geology,State Seismological Bureau, Parrish, R.R., and K.V. Hodges, Miocene (22+1) Beijing100029, People's Republic of China. metamorphism and two stage thrusting in the Greater P. Copeland,Department of Geosciences,University of Himalayan sequence,Annapurna Sanctuary, Nepal, Geol. Houston, Houston, Texas ?7204-5503. $oc. Am. Abstr. Programs, 25, A174, 1993. T. M. Harrisonand A. Yin, Departmentof Earthand Space P•cher, A., The contact between the higher Himalayan Sciencesand Institute of Geophysicsand Planetary Physics, cystallines and the Tibetan sedimentaryseries: Miocene Universityof California,Los Angeles, CA 90024-1567. large-scale dextral shearing, Tectonics, 10, 587-598, 1991. W. S. F. Kidd, Departmentof GeologicalSciences, State Peltzer, G., and P. Tapponnier, Formation and evolution of Universityof NewYork at Albany,Albany, NY 12222. strike-slip faults, rifts, and basinsduring the India-Asia F. $. Ryerson,Institute of Geophysicsand Planetary Physics, collision: An experimentalapproach, J. Geophys.Res., LawrenceLivermore National Laboratory, Livermore, CA 93, 15,085-15,117, 1988. 94550. Powell, C. M., and P.G. Conaghan,Plate tectonicsand the Himalayas, Earth Planet. $ci. Lett., 81, 1431-1450, (ReceivedApril 26, 1993;revised December 6, 1993; 1973. acceptedFebruary 15, 1994.)