Earth and Planetary Science Letters 325–326 (2012) 10–20

Contents lists available at SciVerse ScienceDirect

Earth and Planetary Science Letters

j ourna l homepage: www.e lsev ie r .com/ locate /eps l

E–W extension at 19 Ma in the Kung Co area, S. Tibet: Evidence for contemporaneousE–W and N–S extension in the Himalayan orogen

Mayumi Mitsuishi a,⁎, Simon R. Wallis a, Mutsuki Aoya b, Jeffrey Lee c, Yu Wang d

a Department of Earth & Planetary Sciences, Graduate School of Environmental Studies, Nagoya University, Nagoya 464-8601, Japanb Institute of Geology and Geoinformation, National Institute of Advanced Industrial Science and Technology (AIST), Central 7, Tsukuba 305-8567, Japanc Department of Geological Sciences, Central Washington University, Ellensburg, WA 98926, USAd Geologic Labs Center, China University of Geosciences, Beijing 100083, China

⁎ Corresponding author.E-mail addresses: mitsuishi.mayumi@nagoya-u.jp,

mitsuishi-mayumi@jogmec.go.jp (M. Mitsuishi).

0012-821X/$ – see front matter © 2011 Published by Edoi:10.1016/j.epsl.2011.11.013

a b s t r a c t

a r t i c l e i n f o

Article history:Received 5 March 2011Received in revised form 5 November 2011Accepted 7 November 2011Available online xxxx

Editor: T.M. Harrison

Keywords:southern Tibetnormal faultingage of extensionremote sensing

Active faulting in southern Tibet consists of N–S trending extensional faults and linked strike-slip faults,which are an expression of regional E–W extension. A second type of extensional deformation associatedwith N–S movement is also recognized. This extension is expressed as a series of shear zones and normalfaults in the High Himalayas – the Southern Tibetan Detachment System – and mid-crustal rocks exposedin metamorphic domes. Reported constraints on the timing of movements associated with these two phasesof extension indicate that N–S extension predates the onset of E–W extension. However, only a few studieshave provided clear constraints on the timing of E–W extension and the extent to which the two kinemati-cally distinct domains of extension were contemporaneous is unclear.The Kung Co fault in southern Tibet is a major N–S trending normal fault. The associated E–W extension is lo-cally expressed as high-strain ductile deformation. Both field and microstructural observations show that thisdeformation occurred synchronously with granite intrusion. Previously reported U–Pb zircon dating showsgranite crystallization took place at around 19 Ma, implying that ductile E–W extension in the Kung Co areawas also active at around 19 Ma. This is the oldest documented example of E–W extension in Tibet andshows that E–W extension was at least locally contemporaneous with N–S extension to the south at shallowercrustal levels. Simultaneous mid-crustal N–S extension and upper crustal E–W extension may be explained bysouthward flow of Tibetan crust with a divergent radial component.

© 2011 Published by Elsevier B.V.

1. Introduction

Active tectonics in the central part of the Tibetan Plateau is domi-nated by a linked set of NE–SW and NW–SE trending strike-slip faultsand N–S trending normal faults (Fig. 1) (Blisniuk et al., 2001; Taylor etal., 2003). The normal faulting acts to reduce the overall height of thePlateau despite the ongoing convergence of India with Asia and to-gether with the strike-slip faulting reflects a regional E–W extension(e.g. Armijo et al., 1986). The onset of normal faulting in central Tibetmarks a time when the horizontal deviatoric stresses in the plateauchanged from compressional to extensional andmanyworkers indicatethis corresponds closely to the time when maximum elevation wasobtained in the region (Cook and Royden, 2008; Molnar et al., 1993).

In southern Tibet there is also abundant evidence for an earlier phaseof extension but in a N–S direction. This extension resulted in the forma-tion of a major set of E–W striking, north-dipping normal faults andshear zones known as the Southern Tibetan Detachment System

lsevier B.V.

(STDS), which emplaced an overlying mainly unmetamorphosed Tethy-an sequence onto metamorphic rocks of the high Himalayas (Burchfielet al., 1992; Burg and Chen, 1984).

Age constraints indicate that the main phase of the N–S extensiontook place between 23 and 12 Ma (Coleman and Hodges, 1995;Edwards and Harrison, 1997; Guillot et al., 1994; Murphy andHarrison, 1999; Searle et al., 1997, 1999; Vannay et al., 2004;Viskupic et al., 2005; Walker et al., 1999; Zhang and Guo, 2007). Leeand Whitehouse (2007) propose that N–S mid crustal deformation inthe Mabja dome area dated at around 35 Ma is also related to move-ments on the STDS, but deformation of a similar age in the Changgoarea is interpreted by other workers as related to N–S shortening(Larson et al., 2010). The onset of E–W extensional faulting is thoughtto be between 17 and 7 Ma (Blisniuk et al., 2001; Cottle et al., 2009;Dewane et al., 2006; Edwards and Harrison, 1997; Garzione et al.,2003; Hager et al., 2009; Harrison et al., 1995; Hintersberger et al.,2010; Lee et al., 2011; Murphy et al., 2002; Stockli et al., 2002; Thiedeet al., 2006) or even as young as 4 Ma (Mahéo et al., 2007) and signif-icantly younger than movements on the STDS. Distinct kinematics anddifferent age ranges for the two types of extensional structures indicatethat they are related to two distinct tectonic events (e.g. Beaumont et

Fig. 1.Digital elevation model (Gtopo 30 data) of the Tibetan Plateau andmajor active faults. Location of faults modified from Blisniuk et al. (2001) and Taylor et al. (2003). Numbers areestimates of initiation age for E–W extension based on emplacement of N–S striking dykes (light-gray filled double triangles; Williams et al., 2001) and interpreted slip along normalfaults (dark-gray filled double triangles), for Leo Pargil (Hintersberger et al., 2010; Thiede et al., 2006), Gurla Mandhata (Murphy et al., 2002), Thakkola (Garzione et al., 2003), TangraYumco (Dewane et al., 2006), AmaDrimeMassif (Stockli et al., 2002), Xainza (Hager et al., 2009), ShuangHu (Blisniuk et al., 2001), Yadong (Edwards andHarrison, 1997), Nyainqentanglha(Harrison et al., 1995), Gulu (Stockli et al., 2002), and Kung Co (this study) regions. The white-filled double triangles represent the active age for N–S extension associated with SouthTibetan Detachment System for Zanskar (Walker et al., 1999), Sutlej (Vannay et al., 2004), Silving (Searle et al., 1999), Thakkola (Coleman and Hodges, 1995), Manaslu (Guillot et al.,1994), Langtang (Searle et al., 1997), Everest (Murphy and Harrison, 1999; Viskupic et al., 2005), Dinggye (Zhang and Guo, 2007), and Yadong (Edwards and Harrison, 1997)regions.

11M. Mitsuishi et al. / Earth and Planetary Science Letters 325–326 (2012) 10–20

al., 2004; Blisniuk et al., 2001). There is, however, evidence that the twophases are at least in part contemporaneous. Evidence for relativelyyoung movements on the STDS comes from the Kula granite area.Here that STDS cuts b12.5 Ma granite, so slip in this area is youngerthan 12.5 Ma (Edwards and Harrison, 1997). These relationships areconfirmed by the presence of the ~12Ma Wagyela granite structurallybeneath the STDS which displays a ductile normal slip fabric (Wu et al.,1998). There is also a report of movement on the STDS as young as b17.2 ka (Hurtado et al., 2001). The presence of ~18 Ma N–S trendingdykes provides further evidence for contemporaneous N–S and E–Wextension. The age of the dykes partly overlaps with the main docu-mented phase of N–S extension and their orientation suggests that em-placement may be related to the onset of E–W extensional stresses(Williams et al., 2001). These reports indicate that the two stages of ex-tensional deformation may have been contemporaneous and open thepossibility that they are related.

In this contribution, we present evidence for ductile E–W exten-sion in the Kung Co area of southern Tibet that took place at ~19 Maand is associated with a major N–S trending normal fault. This rankswith the oldest previously known examples of E–W extension andis a rare documented example of ductile flow associated with E–Wextension. The revised older age for onset of E–W extension is con-temporaneous with STDS related N–S ductile extension in the sameregion. This study reveals a close spatial and temporal association be-tween the two extensional domains, which indicates that they are dif-ferent manifestations of a single large-scale tectonic event. Wediscuss the tectonics in the Miocene Himalayas that could havecaused development of simultaneous N–S and E–W extension.

2. Geology of Kung Co area

2.1. Regional setting

The southern boundary domain to the Tibetan Plateau is theHimalayas. Within the Himalayan belt several major north-dippingthrust systems can be defined. The most prominent of these are theMain Central Thrust (MCT) and Main Boundary Thrust (MBT). Thehanging wall of the MCT consists of the High Himalayanmetamorphicsequence, which, in turn, has an upper boundary defined by a majornorth-dipping normal fault system, the STDS, which can be tracedroughly 2000 km along the length of the main Himalayan range(Fig. 2) (Burchfiel et al., 1992). Both the Tethyan Himalayan sequencein the hanging wall of the STDS and the Higher Himalayan Sequenceare thought to consist mainly of rocks that make up the pre-collisional Indian continental margin (e.g. Hodges, 2000; Le Fort,1996).

Numerous granites are exposed in the Tethys Himalaya (Fig. 2) andradiometric dating indicates that they can be divided into an early pre-orogenic group with ages of formation around 560–500 Ma (Lee andWhitehouse, 2007; Lee et al., 2000; Schärer et al., 1986) and a syn-orogenic younger group with ages around 35–10Ma (Aoya et al.,2005; Kawakami et al., 2007; Larson et al., 2010; Lee andWhitehouse, 2007; Lee et al., 2006; Quigley et al., 2008; Schärer et al.,1986; Zhang et al., 2004a, 2004b). The Kung Co granite intrudedaround 19Ma (Lee et al., 2011) and belongs to the younger group.The Kung Co granite is cut by a N–S trending normal fault with avery clear geomorphological expression. This normal fault also cuts

Fig. 2. Tectonic map of southern Tibet (modified from Burchfiel et al. (1992)) showing locations and zircon U–Pb ages of granite of the Kangmar dome (Lee et al., 2000; Schärer et al.,1986), Kuday dome (Zhang et al., 2004a, 2004b), Kouwu dome (Zhang et al., 2004a, 2004b), Mabja dome (Lee and Whitehouse, 2007; Lee et al., 2006), Kung Co granite (Lee et al.,2011), Malashan dome (Aoya et al., 2005; Kawakami et al., 2007), and Changgo and Kung Tang domes (Larson et al., 2010). Average stretching directions (−180°bθ≤180°) ingneiss domes are shown relative to the average of the Mabja dome with clockwise angles shown as positive. ITSZ = Indus-Tsangpo Suture Zone; STDS= South Tibetan DetachmentSystem; MCT = Main Central Thrust; MBT = Main Boundary Thrust. Inset shows strains derived from GPS velocities (Allmendinger et al., 2007) and stretching lineation data(Brunel, 1986) plotted on the same DEM image as given in Fig. 1.

12 M. Mitsuishi et al. / Earth and Planetary Science Letters 325–326 (2012) 10–20

Cenozoic calcareous and siliciclastic Tethyan sedimentary units intrud-ed by the Kung Co granite and younger Quaternary sediments.

2.2. Lithological units

The Kung Co granite can be broadly classified into coarse-grainedtwo-mica granite containing K-feldspar phenocrysts, fine-grainedtwo-mica granite (Fig. 3a), and aplite (Fig. 3b). Fine- and coarse-grained varieties of granite commonly show gradational contacts.Two independent studies report U–Pb zircon dating that has beenused to estimate the age of granite emplacement. Mahéo et al.(2007) report a discordia lower intercept age of 22.0±2.7 Ma. Leeet al. (2011) present a compatible but more closely constrainedconcordia age of 19.06±0.15 Ma (2-sigma) with an MSWD of 1.6that is based on the weighted average of 207Pb-corrected ages fromrims of zircon grains.

The dominant lithology around the granite is calc schist. The calcschist commonly contains cm-scale interlayers of pelitic schist andquartzite (Fig. 3c). A mappable unit of pelitic schist is present inthe southern part of the Kung Co area. The pelitic schist containsporphyroblasts of garnet, staurolite, andalusite and locally also silli-manite (Fig. 3d). The presence of andalusite and sillimanite indi-cates relatively high temperatures and low pressure and itsrestricted development adjacent to the granite implies that itformed as the result of contact metamorphism by granite intrusion.Mappable units of quartzite were not recognized in the field, butanalysis of remote sensing imagery from the north of the studyarea reveals the presence of quartz-rich layers with km-scale folds(Fig. 4).

2.3. The Kung Co fault

The active Kung Co fault is one of the N–S trending normal faultsthat reflect the regional E–W extension in Tibet (e.g. Armijo et al.,1986). The fault can be traced from south of the Kung Co granite

northwards towards the Indus–Tsangpo Suture Zone and is probablycontinuous across the suture zone connecting with the Daggyai Tsofault making a total distance of about 250 km (Fig. 1). The fault has anoverall NNW–SSE trend but locally segments that trend both N–S andNW–SE are also present (Figs. 4 & 5). The fault dips to the W and hasa normal displacement as shown by geomorphological features suchas hanging valleys and triangular facets (Fig. 3e) on the eastern sideand the presence of a sedimentary basin on the western side. A normalsense of displacement is also shown by the rotation of clasts in gravelcut by the fault (Fig. 3f). Slicken surfaces and slickenlines are locally ob-served (Fig. 3g) and generally indicate dip slip motion. The master faultplane is exposed at the margin of the granite body (Armijo et al., 1986;Lee et al., 2011; this study) where it dips 50° to 60° W, which is signif-icantly steeper than the slope of the triangular facets.

Two separate studies have used thermo geochronology to investi-gate the movement history of the Kung Co fault (Lee et al., 2011;Mahéo et al., 2007). Mahéo et al. (2007) interpret a U–Pb zircon ageof 22.0±2.7 Ma as the emplacement age of the Kung Co granite.They further interpret mica Ar–Ar and (U,Th)–He apatite dating to re-veal a period of subsequent rapid cooling lasting to 7.5 Ma. These dataalso indicate that there was a subsequent period of slow cooling untilabout 4 Ma. Mahéo et al. (2007) suggest that the temperature de-crease up to 4 Ma simply reflects cooling of the pluton. They furthernote that simple extrapolation of the present-day (U,Th)–He apatiteage–height relationships implies an unrealistically high thermal gra-dient for the crust in the area at present. To account for this apparentcontradiction, Mahéo et al. (2007) propose a phase of rapid faultingand associated cooling after the youngest recorded age of 4 Ma.These authors do not exclude the possibility of earlier movementson the fault, but earlier movement would have been at relativelyhigh temperatures and they consider the lack of evidence for ductiledeformation to indicate that the initial movement on the Kung Cofault was related to the post 4 Ma mentioned in the paper.

In a different study, Lee et al. (2011) utilize a more complete dataset of zircon and apatite (U,Th)–He ages, which when combined with

Fig. 3. Photographs showing geological features of the Kung Co lithologies. Kfs = K-feldspar; Qtzite = quartzite. (a) Boundary between coarse-grained two-mica granite with Kfscrystals up to several cm in length and fine-grained two-mica granite. (b) Aplite dykes intruding calc schist. (c) Thin pelitic schist and quartzite interlayers within the calc schist.(d) Andalusite porphyroblasts adjacent to the granite. (e) Granite triangular facets formed by normal faulting. (f) Outcrop photo showing rotated clasts in gravel cut by the Kung Cofault. (g) Two sets of slickenlines on a granite surface. The down-dip set is older.

13M. Mitsuishi et al. / Earth and Planetary Science Letters 325–326 (2012) 10–20

the mica Ar–Ar ages of Mahéo et al. (2007) show that rapid coolingbegan between ~12.5 Ma and 10.0 Ma and accelerated at 10 Ma.These results provide strong evidence that there was a major phaseof movement on the Kung Co fault at this time. Lee et al. (2011)show that the initiation of fault movement was therefore mucholder than the 4 Ma postulated by Mahéo et al. (2007). Using 2D ther-mal modeling, Lee et al. (2011) show a period of relatively slowcooling as micas closed to argon loss at ~15.5 Ma as temperaturesdropped below ~300 °C (see their Fig. 13). They attributed this slowcooling episode to the waning stages of thermal reequilibration fol-lowing emplacement of the Kung Co granite. 12.5 Ma–10 Ma there-fore marks a period when there was rapid displacement on thefault, but leaves open the possibility that there were earlier substan-tial movements along the same tectonic boundary.

3. Remote sensing and geological mapping

Satellite images of the Kung Co area were used to complementfield mapping and to produce a geological map of the area. The region

to the east of the Kung Co fault is well exposed and particularly wellsuited to geological analysis by remote sensing.

The satellite sensor used in the present study is the AdvancedSpaceborne Thermal Emission and Reflection Radiometer (ASTER)on the Terra satellite. It has three visible and near infrared bands(VNIR), six shortwave infrared bands (SWIR) and five thermal infra-red bands (TIR) with spatial resolutions of 15, 30 and 90 m, respec-tively (Yamaguchi et al., 1998).

The granite is leucocratic and contrasts with the much darker sur-rounding sedimentary rocks (Fig. 3b, d). Information from the VNIRcan, therefore, be used to define the distribution of the Kung Co gran-ite (Fig. 4a). The VNIR image of Fig. 4a comprises band 3 in red, band 2in green, and band 1 in blue.

The ASTER data can also be used to help define the distribution ofdifferent sedimentary units. The reflection spectra for the low-grademetamorphic rocks of the Tethyan sequence are clearly differentfrom those of the granite, Cenozoic Tethyan sediments, and Quaterna-ry sediments. This distinction is especially clear in the SWIR bandratio images, consisting of band ratios 7/6 in red, 6/5 in green, and

Fig. 4. ASTER satellite images used to help produce a geological map. (a) VNIR false-color image comprised of band 3 in red, band 2 in green, band 1 in blue. (b) SWIR band ratioimage comprised of band 7/6 in red, band 6/5 in green, and band 6/4 in blue. (c) Gray-scale image of carbonate index. Carbonate Index (CI)=(band 13)/(band 14) (Ninomiya andFu, 2002). (d) Gray-scale image of quartz index. Quartz Index (QI)=[(band 11)×(band 11)]/[(band 10)×(band 12)] (Ninomiya and Fu, 2002). The lighter colors correspond to anincrease in the indexes for quartz and carbonate. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

14 M. Mitsuishi et al. / Earth and Planetary Science Letters 325–326 (2012) 10–20

6/4 in blue (Fig. 4b). In this image, the granite is red, low-grade meta-morphic rocks are blue, and Cenozoic Tethyan sediments and Quater-nary sediments show a variety of colors. It is likely that the 7/6 and 6/4 band ratios in these images are most strongly influenced by thepresence of muscovite and chlorite, respectively (Perry, 1993).

The meta-sedimentary rocks around the granite can be dividedbroadly into two lithologies based on the content of the carbonateminerals. The main carbonate minerals show major differences inthe strength of the reflection spectra for the TIR bands 13 and 14.Using this feature Ninomiya and Fu (2002) have proposed a carbon-ate index:

CarbonateIndex CIð Þ ¼ band13ð Þ= band14ð Þ

This index can be used for mapping the amount of the carbonateminerals. Ninomiya and Fu (2002) also propose an alternative expres-sion to examine the amount of quartz.

QuartzIndex QIð Þ ¼ band11ð Þ � band11ð Þ½ �= band10ð Þ � band12ð Þ½ �

Fig. 4c and d was constructed by using these two mineral indices.Areas showing high CI values correlate with the distribution of calcar-eous rocks and metamorphosed equivalents observed in the field.Areas showing high QI values correlate with siliciclastic rocks andtheir metamorphosed equivalents. There are also local bands in thenorth of the area that show especially high QI values and these repre-sent quartz-rich rocks. On the ASTER images, these prominent layerslocally define km-scale folds. The distribution of the Quaternary sed-iments is estimated using a combination of topography and digitalimages.

A geological map in the Kung Co area was constructed by combin-ing a series of remote sensing images with field observations (Fig. 5).The Gyirong–Kangmar thrust is a major fault that extends fromKangmar through the Kung Co area and continues to the west (Burgand Chen, 1984). The distribution of this fault in the study area istaken from Burg and Chen (1984). In addition to defining the distri-bution of the granite body and faults, the ASTER images also showthe presence of prominent layers of E–W striking quartz-rich rocksin the north of the area that are distributed both the east and westof the Kung Co fault. The large-scale folds in the quartz-rich sedi-ments indicate high-strain ductile deformation occurred extensivelyin this area. The lack of major lateral offsets of these marker horizonsacross the fault is compatible with the down slip movement indicatedby slickenlines.

4. Ductile deformation

4.1. Deformation phases

In the Kung Co area, a strong foliation is developed in themetasedimentary schists surrounding the Kung Co granite. A set ofstructures that formed during a phase of ductile deformation beforedevelopment of the dominant schistosity can be defined on thebasis of overprinting structures. These two stages of ductile deforma-tion are referred to as D1 and D2 with associated foliations S1 and S2.A younger low-strain deformation stage, D3, can also be locally recog-nized where post-D2 folds are developed. In general, therefore, al-most any tectonic foliation observed in the field can be classified aseither S1 or S2. Both foliations can be observed in most outcropswith a folded S1 that has an axial planar S2 (Fig. 6a, b). The threephases of ductile deformation are overprinted by later phases of

Fig. 5. Geological map of the Kung Co area incorporating field observations and theresults of remote sensing data analyses (Fig. 4). *1 Location of fault estimated fromtopographic map and Armijo et al. (1986). *2 Kangmar Gyirong Thrust (positionapproximate) after Burg and Chen (1984).

15M. Mitsuishi et al. / Earth and Planetary Science Letters 325–326 (2012) 10–20

brittle fracturing and faulting. Fig. 7 shows stereographic plots of thestructural elements of D1, D2, and the brittle normal faulting.

4.2. D1 deformation and post-D1 metamorphism

The oldest deformation event, D1, is best preserved farthest awayfrom the granite. The associated S1 generally dips at a low angle(Fig. 7). The L1 stretching lineation on S1 is most commonly definedby iron oxide streaks associated with pyrite. D1 structures show con-siderable variation in orientation, which is likely to be due to the ef-fects of subsequent deformation.

A microstructural criterion for differentiating between S1 and S2 isprovided by porphyroblastic plagioclase, which is widely developedin calc-schist units. The plagioclase grains commonly contain a well-preserved straight internal foliation mainly defined by graphite andmuscovite inclusions. In samples that contain open D2 microfolds,the trace of the folded S1 defines zigzag patterns due to the presenceof straight sections of S1 protected by overgrown plagioclase(Fig. 6b). This microstructure indicates that D2 post-dates the growthof the porphyroblastic plagioclase. In addition, the straight internalfoliation can be traced continuously into the external S1 implyingstatic, post-D1 growth of the host plagioclase. Therefore, the growthof plagioclase porphyroblasts occurred between D1 and D2. This rela-tionship is useful for determining whether an observed foliation is S1or S2, especially when an S2 overprinting an earlier S1 cannot be di-rectly identified in outcrop.

4.3. D2 deformation

D2 ductile deformation is the most widely developed deformation inthe area. In the southern part of the Kung Co area, D2 structures are par-ticularly strongly developed in the pelitic schist. S2 dips steeply to thewestern quadrant with mean azimuth and dip of 264° and 53°, respec-tively (Fig. 7). The L2 stretching lineation is defined by strain shadowsaround porphyroblasts and the preferred orientation of elongate mineralgrains. L2 plunges NW to SWwithmean azimuth and plunge of 255° and31°, respectively. Later deformation is only locally developed implyingthat the kinematic axes have not undergone major reorientation. Theflow direction shown by the D2 structures therefore implies that D2 duc-tile deformation is associated with a component of E–W extension.

To further examine the kinematics of D2 deformation, the sense ofshear was determined using structures developed on a variety ofscales: (i) synkinematic rotation of andalusite porphyroblasts(Fig. 6c, d); and (ii) outcrop-scale shear bands and asymmetric lenses(Fig. 6e). The results show a consistent pattern of top to the W to topto the NW shear.

D2 ductile deformation and later overprinting brittle normalfaulting show similar kinematics. Lee et al. (2011) conclude that thesimilarity in orientation of the D2 foliation, fault plane, and fracturesdeveloped within the granite in this area shows that these structuresrecord a history of progressive deformation during normal faultingfrom a depth where temperatures were high enough that graniteunderwent weak ductile deformation to shallow crustal levelswhere brittle deformation dominated.

5. Relationship between granite intrusion andductile deformation

The relationship between granite intrusion and different stages ofdeformation can constrain the timing of the associated tectonicevents. In our field studies we paid particular attention to structuresthat could help to determine the relative timing of granite intrusionand deformation. Features that were especially helpful include struc-tures of granitic dykes and contact metamorphic minerals. These fea-tures developed at the same time as the granite intrusion andobservation of ductile deformation fabrics in the same region allowsus to determine if deformation occurred before or after intrusion.

5.1. Deformation of granitic dykes

The contact zone between the granite and surrounding lithologiescontains numerous granitic dykes (Figs. 3b, 8). The dykes commonlycrosscut the D1 tectonic fabrics and we saw no evidence that granitewas affected by D1 deformation. However, granite dykes are locallyclearly affected by D2 folding (Fig. 8a) and they are also locallyboudinaged subparallel to the S2 foliation (Fig. 8b). These features in-dicate that granite was intruded after D1 and before or during D2 de-formation. Undeformed granite, that crosscuts a well-developed S2foliation within the pelitic schist, is also observed (Fig. 8c, d). Theseobservations imply that granite intrusion continued at least locallyafter D2 deformation ceased, although it had already began beforeD2 was complete. The period of granite intrusion, therefore, overlapsthat of the D2 deformation and the main phase of granite intrusionwas at least in part contemporaneous with D2 deformation. Evidencefor high T deformation of the granite is given by the presence ofchessboard subgrain microstructures in quartz (Fig. 8e) indicatingthe granite was deformed under high temperature D2 deformation(500–700 °C) (Passchier and Trouw, 2005).

5.2. Microstructures of contact metamorphic minerals

Contact metamorphism of pelitic schist around the granite body isassociated with the formation of andalusite and sillimanite. These

Fig. 6. Outcrop photograph and photomicrographs in the Kung Co area. Pl = plagioclase; Qtz = quartz; Bt = biotite; And = andalusite. (a) Representative mesostructures of calcschist showing S1 and S2. (b) Photomicrograph of axial planer S2 in pelitic schist with trace of folded S1; several plagioclase porphyroblasts that contain straight S1 composed bycarbonate are indicated (crossed nicols). (c) Photomicrograph of an andalusite porphyroblast in pelitic schist showing the continuation of the external S2 with the rotated internalfoliation (crossed nicols). The porphyroblast shows simultaneous extinction when viewed between crossed polars and there is no evidence for recrystallization or fracturing thatcould account for the observed curvature of the inclusions trails. (d) Enlargement of the boxed part of (c) highlighting the relationships of the porphyroblast and foliation.(e) Outcrop-scale shear bands developed within layers of quartzite interbedded with calc schist.

16 M. Mitsuishi et al. / Earth and Planetary Science Letters 325–326 (2012) 10–20

minerals form markers of contact metamorphism and their micro-structures can therefore be used to determine the relative timing be-tween ductile deformation and granite intrusion.

Andalusite-bearing pelitic schist is present in the southern part ofthe study area. The porphyroblasts do not contain D1 fabrics and therelationship with D1 could not be determined unequivocally. Howev-er, microstructural studies suggest that S1 affected by D2 folds is lo-cally preserved in the pressure shadows around andalusiteporphyroblasts (Fig. 8f, g). Andalusite grains that are stretched andboudinaged parallel to L2 are also locally well-developed (Fig. 8h, i)showing pre-D2 or syn-D2 development. More information aboutthe relationships between andalusite growth and deformation of thehost rock are given by microstructural observations. Many andalusitegrains can be divided into an inclusion-free core and a rim with alarge numbers of inclusions (Fig. 8f, g). The minerals included in therim are of similar size and shape to the matrix and show an alignmentcontinuous with the external S2 foliation but curved within the anda-lusite grain (Figs. 6c, d & 8f, g). The external foliation shows no similarcurvature. These observations show that the andalusite overgrowsthe S2 foliation whilst rotating indicating synkinematic growth dur-ing non-coaxial D2 deformation. The core regions are commonly in-clusion free and the relationship with deformation unclear, but

most of the andalusite growth was synchronous with D2 deforma-tion. There is no evidence for the presence of andalusite during D1.

Sillimanite is locally formed adjacent to the andalusite (Fig. 8j).There is no microstructural evidence that the formation of sillimaniteis associated with the breakdown of muscovite by reaction withquartz. This implies pressures between 2 and 3.5 kbar and peak tem-peratures between 500 and 650 °C; these P–T estimates are compati-ble with those given by Mahéo et al. (2007) for the contactmetamorphism. The strong alignment of the sillimanite parallel toS2 indicates that it was affected by D2. Sillimanite growth post-dates andalusite formation, so sillimanite also formed during D2deformation.

The above microstructural observations show that contact meta-morphic minerals grew during D2. This confirms the field observa-tions indicating that granite intrusion was synchronous with D2.

6. Discussion

6.1. Timing of the onset of E–W extensional deformation

Field relations, mesoscopic structures, and microstructures showthat E–W ductile flow associated with extensional deformation

Fig. 7. Stereonet plots showing orientations of D1 and D2 deformation structures, and normal fault structures measured in the outcrop. Geological map of the Kung Co area is basedon field observations and image processing of ASTER satellite imagery. The fault slip data are also reported in Lee et al. (2011).

Fig. 8. Structural features of the Kung Co granite and andalusite in the surrounding pelitic schist formed by contact metamorphism. And= andalusite; Sil = sillimanite. (a) Outcropphoto showing granite dyke folded by D2. (b) Outcrop photo showing occurrence of boudinaged granite. (c) Field photo showing undeformed granite, that cross-cuts pelitic schistwith a D2 foliation. (d) Field photo showing undeformed granite surrounding a partially detached sliver of pelitic schist that has a D2 foliation. (e) Photomicrograph of thechessboard subgrain microstructure in granite deformed by D2 (crossed nicols). (f) Photomicrograph of andalusite in pelitic schist (open nicol). Folds preserved in the strainshadow of the andalusite can be correlated with D1. The internal foliation of the porphyroblast is continuous with the external S2. (g) Enlargement of boxed part of (f) showingcurved S2 within the andalusite porphyroblast and straight S2 outside. (h) Field photo showing boudinaged andalusite stretched parallel to L2. (i) Outcrop photo showing theD2 strain shadows around andalusite. (j) Photomicrograph of aligned sillimanite adjacent to the andalusite (open nicol).

17M. Mitsuishi et al. / Earth and Planetary Science Letters 325–326 (2012) 10–20

Table 1Details of average stretching lineation in gneiss domes. See Fig. 2 for the localities. N =number of data; θ = the azimuth of the vector mean relative to Mabja dome; d0=95%confidence interval of the mean direction calculated following the procedure outlinedin Masuda et al. (1999).

Area N θ d0 References

Kangmar 109 +8.7° 9.0 Lee et al. (2000)Kampa 66 −16.5° 11.5 Quigley et al. (2008)Mabja 143 0° 5.0 Lee et al. (2004)Kung Co 31 −94.9° 17.8 This studyMalashan 56 +17.8° 14.8 Aoya et al. (2006)

18 M. Mitsuishi et al. / Earth and Planetary Science Letters 325–326 (2012) 10–20

occurred simultaneously with 19 Ma Kung Co granite intrusion in theKung Co area. That is, ductile E–W extension was occurring in theKung Co area at around 19 Ma—considerably older than most otherestimates for the onset age of E–W extension. Ductile orogen-parallel extension in the Himalayan orogen has also been reportedfrom the Leo Pargil area at 17–14 Ma (Hintersberger et al., 2010;Thiede et al., 2006) and the Ama Drime Massif area at ~13–12 Ma(Cottle et al., 2009) (Fig. 1). The presence of these similar instancesof E–W ductile extension occurring in the early Miocene evolutionof the Himalayan orogen indicates that the results from Kung Cocould represent a regionally developed phase of E–W ductileextension.

6.2. Relationship with N–S extension of the Southern TibetanDetachment System

Ductile deformation at mid crustal levels that occurred during theHimalayan orogeny is also reported from the Tethyan Himalayanmetamorphic domes (e.g. Aoya et al., 2006; Kawakami et al., 2007;Lee et al., 2004, 2006; Quigley et al., 2008). In the metamorphicdomes, the extensional direction is dominantly N–S and generallyinterpreted as linked to the STDS associated with N–S extension(Fig. 2). The distinct timing and kinematics of deformation betweenthe N–S extension and E–W extension has meant that the two phaseshave been treated as distinct with different tectonic causes (e.g.Coleman and Hodges, 1995). However, the age of the ductile E–Wflow reported in this contribution from the Kung Co area is slightlyolder than the b17 Ma estimate for the age of movement on theSTDS reported from the Qomolangma detachment in the Everestarea (Murphy and Harrison, 1999), which is approximately 100 kmto the south of Kung Co. These observations imply that a tectonicmodel is required that incorporates both E–Wand N–S flow occurringat the same time in a relatively narrow region.

The ductile extensional deformation in the Kung Co area is spatiallyclosely associated with brittle faulting. The ductile brittle transition isgenerally at a level of a few kilometers. The structures developed in theKung Co area record this transition and suggest that at least the laterpart of the ductile deformation occurred at a similar depth. A shallowcrustal level for the D2 deformation is confirmed by the low-pressure es-timates for the concomitant contact metamorphism (b3.5 kbar equiva-lent to b13.2 km for ρ=2.7 g/cm3). Contemporaneous N–S ductileextension is recorded in the cores of many of the Tethyan Himalayandomes (Kawakami et al., 2007; Lee et al., 2000, 2004). Pressure estimatesin the domes show that the N–S ductile extension took place at deeperlevels than that observed in Kung Co. This difference indicates thatthere may be a link between structural level and kinematics of deforma-tion: the deeper levels were associated with N–S flow, whereas the shal-lower levels were associated with E–W flow. A similar change instretching directionwith crustal depth has been documented in the Sew-ard Peninsula of Alaska (Amato and Miller, 2004).

6.3. Model to account for contemporaneous N–S and E–W extension

Copley and McKenzie (2007) present a model for mid-crustal flowbeneath the Tibetan Plateau with spreading of the Tibetan crustsouthwards over the Indian shield. The Indian Shield is thought to ex-tend up to 33°N beneath the Plateau (Jin et al., 1996), and most of theN–S faulting occurs to the south of this. In this model, gravitationalflow over the Indian Shield is driven by the surface slope – or moreprecisely, the gravitational potential energy – of the HimalayanOrogen. The curvilinear form of the Orogen results in a radial compo-nent of flow, which causes orogen-parallel extension that could be re-lated to the E–W extension of southern Tibet. Recent divergentmovement is shown clearly in strain patterns derived from GPS data(Allmendinger et al., 2007) and also in the pattern of stretching line-ations in the Himalayas (Brunel, 1986) (Fig. 2). With the exception of

the Kangmar area, the stretching directions associated with N–S midcrustal flow in the southern Tibetan metamorphic domes show a sim-ilar radial component to flow (Fig. 2, Table 1).

We propose, therefore, that southward spreading of Tibetan mid-crust has been associated with a radial component of flow since theMiocene around 19 Ma, and this caused the onset of the observedE–W extension at shallower levels (Fig. 9). Our proposal is consistentwith the observation that E–W extension is better developed in thesouth of the Bangong–Nujiang Suture than the north. Divergent crust-al flow is also seen in the SE of the India–Asia collisional domain tothe south of the Sichuan Basin (Zhang et al., 2004a, 2004b). Exten-sional shear zones related to orogen-parallel extension roughly per-pendicular to this flow have been documented (Jolivet et al., 2001)and these may be analogous to the features we describe here.

6.4. Comparison with other models

Orogen-parallel extension due to the underthrusting of a rigid In-dian continental margin (DeCelles et al., 2002; Kapp and Guynn,2004) makes a similar prediction that extension will be concentratedin southern Tibet. However, this model does not account for the rela-tionship between the curvature of the Himalayan arc and extensiondirection that we highlight here. Large-scale underthrusting of aslab of Indian basement beneath Tibet is also likely to result in the de-velopment of a significant surface slope (Jiménez-Munt and Platt,2006); a feature which is not observed. Other models can also ac-count for east–west extension in Tibet such as regional ductile flowof the mid crust into a region of weak crust (Cook and Royden,2008) and large-scale removal of the lithospheric mantle in an overallnorth–south compressive regime (Molnar et al., 1993). However,these models do not specifically address the stronger developmentof extensional structures in the south. Our observations do not ex-clude either of these proposals for the regional tectonics. However,the evidence we present for linked upper crustal and mid-crustal de-formation does not support the idea that the upper crust can betreated as a rigid upper lid to a channelized flow (Beaumont et al.,2004). We also note that the Kung Co fault shows a marked changeto very rapid movement at around 10 Ma (Lee et al., 2011), which iscompatible with an additional tectonic process starting to operate atthis time.

7. Conclusions

The Kung Co area in southern Tibet contains a major N–S trendingnormal fault, associated with ductile deformation that records E–Wextension with a normal sense of displacement. Field and microstruc-tural observations show the ductile deformation is synchronous withgranite intrusion that occurred at 19 Ma. This is one of the oldestdocumented examples of orogen-parallel E–W extension in Tibet.The age of ductile E–W extension reported here overlaps with N–Sductile flow in the same region that is associated with extensional de-formation along the Southern Tibetan Detachment. Southwardspreading of mid-crust with a radial component can account for the

Fig. 9. 3D tectonic model of contemporaneous E–W and N–S extension in southernTibet. Southward spreading of mid-crust caused E–W extension at shallower levels.The depth of the boundary between the upper and middle crust is estimated frommetamorphic pressures associated with E–W and N–S extension in the Kung Co andMalashan areas, respectively. The lower boundary of the middle crust is estimatedfrom the pressure at which N–S extension occurred in the Kangmar and Mabja areas.Lin = stretching lineation.

19M. Mitsuishi et al. / Earth and Planetary Science Letters 325–326 (2012) 10–20

contemporaneous N–S extension at relatively deep levels and E–Wextension at shallower levels.

Acknowledgments

This study was financially supported in part by the Grants-in-aidfor Scientific Research grants 13573011 and 20403014 awarded toSW and a grant from Central Washington University awarded to JL.We thank Y. Yamaguchi for advice on ASTER remote sensing analysis.The authors are also grateful to the Earth Remote Sensing Data Anal-ysis Center (ERSDAC-Japan), for the delivery of ASTER images in theframe of the project “Mapping and interpretation of the geology ofsouthern Tibet” (ASTER Announcement of Research Opportunity —

Agreement No. H140125). This work is also benefited from discus-sions with members of the petrology group of Nagoya University.We also acknowledge the helpful input from careful reviews by KyleLarson and Alexander Robinson as well as the editorial assistance ofMark Harrison.

References

Allmendinger, R.W., Reilinger, R., Loveless, J., 2007. Strain and rotation rate from GPS inTibet, Anatolia, and the Altiplano. Tectonics 26, TC3013.

Amato, J.M., Miller, E.L., 2004. Geologic map and summary of the evolution of theKigluaik Mountains gneiss dome, Seward Peninsula, Alaska. Geol. Soc. Am. Spec.Pap. 380, 295–306.

Aoya, M., Wallis, S.R., Terada, K., Lee, J., Kawakami, T., Wang, Y., Heizler, M., 2005.North–south extension in the Tibetan crust triggered by granite emplacement.Geology 33, 853–856.

Aoya, M., Wallis, S.R., Kawakami, T., Lee, J., Wang, Y., Maeda, H., 2006. The Malashangneiss dome in south Tibet: comparative study with the Kangmar dome withspecial reference to kinematics of deformation and origin of associated granites.Geol. Soc. London, Spec. Publ. 268, 471–495.

Armijo, R., Tapponnier, P., Mercier, J.L., Han, T.L., 1986. Quaternary extension inSouthern Tibet — field observations and tectonic implications. J. Geophys. Res.91, 13803–13872.

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–Tibetanorogen. J. Geophys. Res. 109, B06406.

Blisniuk, P.M., Hacker, B.R., Glodny, J., Ratschbacher, L., Bi, S.W., Wu, Z.H., McWilliams,M.O., Calvert, A., 2001. Normal faulting in central Tibet since at least 13.5 Myr ago.Nature 412, 628–632.

Brunel, M., 1986. Ductile thrusting in the Himalayas: shear sense criteria and stretchinglineations. Tectonics 5, 247–265.

Burchfiel, B.C., Chen, Z., Hodges, K.V., Liu, Y., Royden, L.H., Deng, C., Xu, J., 1992. The South Ti-betan detachment system, Himalayan orogen: extension contemporaneous with andparallel to shortening in a collisionalmountain belt. Geol. Soc. Am. Spec. Pap. 269, 1–41.

Burg, J.P., Chen, G.M., 1984. Tectonics and structural zonation of Southern Tibet, China.Nature 311, 219–223.

Coleman, M., Hodges, K., 1995. Evidence for Tibetan Plateau uplift before 14-myr agofrom a new minimum age for east–west extension. Nature 374, 49–52.

Cook, K.L., Royden, L.H., 2008. The role of crustal strength variations in shapingorogenic plateaus, with application to Tibet. J. Geophys. Res. 113, B08407.

Copley, A., McKenzie, D., 2007. Models of crustal flow in the India–Asia collision zone.Geophys. J. Int. 169, 683–698.

Cottle, J.M., Jessup, M.J., Newell, D.L., Horstwood, M.S.A., Noble, S.R., Parrish, R.R.,Waters, D.J., Searle, M.P., 2009. Geochronology of granulitized eclogite from theAma Drime Massif: implications for the tectonic evolution of the South TibetanHimalaya. Tectonics 28, TC1002.

DeCelles, G.P., Robinson, M.D., Zandt, G., 2002. Implications of shortening in the Himalayanfold-thrust belt for uplift of the Tibetan Plateau. Tectonics 21. doi:10.1029/2001TC001322.

Dewane, T.J., Stockli, D.F., Hager, C., Taylor, M., Ding, L., Lee, J., Wallis, S.R., 2006. Timingof Cenozoic E–W extension in Tangra Yum Co - Kung rift, south-central Tibet. EosTrans. AGU, 87 (52), Fall Meet. Suppl., Abstract T34C-04.

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 theKhula Kangri granite. Geology 25, 543–546.

Garzione, C.N., DeCelles, P.G., Hodkinson, D.G., Ojha, T.P., Upreti, B.N., 2003. East–westextension and Miocene environmental change in the southern Tibetan plateau:Thakkhola graben, central Nepal. Geol. Soc. Am. Bull. 115, 3–20.

Guillot, S., Hodges, K., Lefort, P., Pecher, A., 1994. New constraints on the age of theManaslu Leucogranite — evidence for episodic tectonic denudation in the CentralHimalayas. Geology 22, 559–562.

Hager, C., Stockli, D.F., Dewane, T.J., Gehrels, G., Ding, L., 2009. Anatomy and crustalevolution of the central Lhasa terrane (S-Tibet) revealed by investigations in theXainza rift: Geophysical Research Abstracts, EGU General Assembly, 11.EGU2009-11346-11341.

Harrison, T.M., Copeland, P., Kidd, W.S.F., Lovera, O.M., 1995. Activation of theNyainqentanghla Shear Zone — implications for uplift of the Southern TibetanPlateau. Tectonics 14, 658–676.

Hintersberger, E., Thiede, R.C., Strecker, M.R., Hacker, B.R., 2010. East–west extension inthe NW Indian Himalaya. Geol. Soc. Am. Bull. 122, 1499–1515.

Hodges, K.V., 2000. Tectonics of the Himalaya and southern Tibet from two perspectives.Geol. Soc. Am. Bull. 112, 324–350.

Hurtado, J.M., Hodges, K.V., Whipple, K.X., 2001. Neotectonics of the Thakkhola grabenand implications for recent activity on the South Tibetan fault system in the centralNepal Himalaya. Geol. Soc. Am. Bull. 113, 222–240.

Jiménez-Munt, I., Platt, J.P., 2006. Influence of mantle dynamics on the topographicevolution of the Tibetan Plateau: results from numerical modeling. Tectonics 25,TC6002. doi:10.1029/2006TC001963.

Jin, Y., McNutt, M.K., Zhu, Y.S., 1996. Mapping the descent of Indian and Eurasian platesbeneath the Tibetan Plateau from gravity anomalies. J. Geophys. Res. 101,11275–11290.

Jolivet, L., Beyssac, O., Goffe, B., Avigad, D., Lepvrier, C., Maluski, H., Thang, T.T., 2001.Oligo-miocene midcrustal subhorizontal shear zone in Indochina. Tectonics 20,46–57.

Kapp, P., Guynn, J.H., 2004. Indian punch rifts Tibet. Geology 32, 993–996.Kawakami, T., Aoya, M., Wallis, S.R., Lee, J., Terada, K., Wang, Y., Heizler, M., 2007.

Contact metamorphism in the Malashan dome, North Himalayan gneiss domes,southern Tibet: an example of shallow extensional tectonics in the TethysHimalaya. J. Metamorph. Geol. 25, 831–853.

Larson, K.P., Godin, L., Davis, W.J., Davis, D.W., 2010. Out-of-sequence deformation andexpansion of the Himalayan orogenic wedge: insight from the Changgo culmination,south central Tibet. Tectonics 29, TC4013.

Le Fort, P., 1996. Evolution of the Himalaya. In: Yin, A., Harrison, T.M. (Eds.), TheTectonics of Asia. Cambridge University Press, New York, pp. 95–106.

Lee, J., Whitehouse, M.J., 2007. Onset of mid-crustal extensional flow in southern Tibet:evidence from U/Pb zircon ages. Geology 35, 45–48.

Lee, J., Hacker, B.R., Dinklage, W.S., Wang, Y., Gans, P., Calvert, A., Wan, J.L., Chen, W.J.,Blythe, A.E., McClelland, W., 2000. Evolution of the Kangmar Dome, southern Tibet:structural, petrologic, and thermochronologic constraints. Tectonics 19, 872–895.

Lee, J., Hacker, B., Wang, Y., 2004. Evolution of North Himalayan gneiss domes:structural and metamorphic studies in Mabja Dome, southern Tibet. J. Struct.Geol. 26, 2297–2316.

Lee, J., McClelland, W., Wang, Y., Blythe, A.E., McWilliams, M., 2006. Oligocene–Miocenemiddle crustal flow in southern Tibet: geochronology of Mabja Dome. Geol. Soc.London, Spec. Publ. 268, 445–469.

Lee, J., Hager, C., Wallis, S.R., Stockli, D.F., Whitehouse, M.J., Aoya, M., Wang, Y., 2011. Middleto late Miocene extremely rapid exhumation and thermal reequilibration in the KungCo rift, Southern Tibet. Tectonics 30, TC2007. doi:10.1029/2010TC002745.

Mahéo, G., Leloup, P.H., Valli, F., Lacassin, R., Arnaud, N., Paquette, J.L., Fernandez, A.,Haibing, L., Farley, K.A., Tapponnier, P., 2007. Post 4 Ma initiation of normal faultingin southern Tibet. Constraints from the Kung Co half graben. Earth Planet. Sci. Lett.256, 233–243.

Masuda, T., Kugimiya, Y., Aoshima, I., Hara, Y., Ikei, H., 1999. A statistical approach todetermination of a mineral lineation. J. Struct. Geol. 21, 467–472.

Molnar, P., England, P., Martinod, J., 1993. Mantle dynamics, uplift of the TibetanPlateau, and the Indian monsoon. Rev. Geophys. 31, 357–396.

Murphy, M.A., Harrison, T.M., 1999. Relationship between leucogranites and theQomolangma detachment in the Rongbuk valley, south Tibet. Geology 27, 831–834.

Murphy, M.A., Yin, A., Kapp, P., Harrison, T.M., Manning, C.E., Ryerson, F.J., Ding, L., Guo,J.H., 2002. Structural evolution of the Gurla Mandhata detachment system,southwest Tibet: implications for the eastward extent of the Karakoram faultsystem. Geol. Soc. Am. Bull. 114, 428–447.

Ninomiya, Y., Fu, B., 2002. Quartz index, carbonate index and SiO2 content indexdefined for ASTER TIR data. J. Remote Sens. Soc. Jpn. 22, 50–61.

Passchier, C.W., Trouw, R.A.J., 2005. Microtectonics, 2nd ed. Springer, Berlin–Heidelberg,p. 366.

20 M. Mitsuishi et al. / Earth and Planetary Science Letters 325–326 (2012) 10–20

Perry, J., 1993. An investigation into the use of ASTER (Advanced Spaceborne ThermalEmission and Reflection Radiometer) data for geologic mapping in the PowderRiver Basin, Wyoming [M.S. thesis], Bowling Green, Ohio, Bowling Green StateUniversity 121.

Quigley,M.C., Liangjun, Y., Gregory, C., Corvino, A., Sandiford,M.,Wilson, C.J.L., Xiaohan, L.,2008. U–Pb SHRIMP zircon geochronology and T-t-d history of the Kampa Dome,southern Tibet. Tectonophysics 446, 97–113.

Schärer, U., Xu, R.H., Allegre, C.J., 1986. U–(Th)–Pb systematics and ages of Himalayanleucogranites, south Tibet. Earth Planet. Sci. Lett. 77, 35–48.

Searle, M.P., Parrish, R.R., Hodges, K.V., Hurford, A., Ayres, M.W., Whitehouse, M.J.,1997. Shisha Pangma leucogranite, south Tibetan Himalaya: Field relations,geochemistry, age, origin, and emplacement. J. Geol. 105, 295–317.

Searle, M.P., Noble, S.R., Hurford, A.J., Rex, D.C., 1999. Age of crustal melting, emplacementand exhumation history of the Shivling leucogranite, Garhwal Himalaya. Geol. Mag.136, 513–525.

Stockli, D.F., 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 ap-atite (U–Th)/He data. Geol. Soc. Am. Abstr. 182–188.

Taylor, M., Yin, A., Ryerson, F.J., Kapp, P., Ding, L., 2003. Conjugate strike-slip faultingalong the Bangong–Nujiang suture zone accommodates coeval east–westextension and north–south shortening in the interior of the Tibetan Plateau.Tectonics 22. doi:10.1029/2002TC001361.

Thiede, R.C., Arrowsmith, J.R., Bookhagen, B., McWilliams, M., Sobel, E.R., Strecker, M.R.,2006. Dome formation and extension in the Tethyan Himalaya, Leo Pargil,northwest India. Geol. Soc. Am. Bull. 118, 635–650.

Vannay, J.C., Grasemann, B., Rahn, M., Frank, W., Carter, A., Baudraz, V., Cosca, M., 2004.Miocene to Holocene exhumation of metamorphic crustal wedges in the NW

Himalaya: evidence for tectonic extrusion coupled to fluvial erosion. Tectonics23. doi:10.1029/2002TC001429.

Viskupic, K., Hodges, K.V., Bowring, S.A., 2005. Timescales of melt generation and thethermal evolution of the Himalayan metamorphic core, Everest region, easternNepal. Contrib. Mineral. Petrol. 149, 1–21.

Walker, J.D., Martin, M.W., Bowring, S.A., Searle, M.P., Waters, D.J., Hodges, K.V., 1999.Metamorphism, melting, and extension: age constraints from the high Himalayanslab of southeast Zanskar and northwest Lahaul. J. Geol. 107, 473–495.

Williams, H., Turner, S., Kelley, S., Harris, N., 2001. Age and composition of dikes inSouthern Tibet: new constraints on the timing of east–west extension and itsrelationship to postcollisional volcanism. Geology 29, 339–342.

Wu, C.D., Nelson, K.D., Wortman, G., Samson, S.D., Yue, Y.J., Li, J.X., Kidd, W.S.F.,Edwards, M.A., 1998. Yadong cross structure and South Tibetan Detachment inthe east central Himalaya (89°–90° E). Tectonics 17, 28–45.

Yamaguchi, Y., Kahle, A.B., Tsu, H., Kawakami, T., Pniel, M., 1998. Overview of AdvancedSpaceborne Thermal Emission and Reflection Radiometer (ASTER). IEEE Trans.Geosci. Remote. Sens. 36, 1062–1071.

Zhang, J.J., Guo, L., 2007. Structure and geochronology of the southern Xainza-Dinggyerift and its relationship to the south Tibetan detachment system. J. Asian Earth Sci.29, 722–736.

Zhang, H.F., Harris, N., Parrish, R., Kelley, S., Zhang, L., Rogers, N., Argles, T., King, J.,2004a. Causes and consequences of protracted melting of the mid-crust exposedin the North Himalayan antiform. Earth Planet. Sci. Lett. 228, 195–212.

Zhang, P.Z., Shen, Z., Wang, M., Gan, W.J., Burgmann, R., Molnar, P., 2004b. Continuousdeformation of the Tibetan Plateau from global positioning system data. Geology32, 809–812.