Variable Structural Style Along the Karakoram Fault Explained Using Triple-Junction Analysis of Intersecting Faults

Variable structural style along the Karakoram fault explained using triple-junction analysis of intersecting faults

N.S. Raterman* E. Cowgill Department of Geology, University of California, Davis, California 95616-8605, USA Ding Lin Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing, People’s Republic of China

ABSTRACT Co fault system likely initiated between 10 and small (<100-km-long) faults and have been geo- 3 Ma to accumulate 25–32 km of total left sep- graphically restricted to the San Andreas and Structural style along the active, NW-strik- aration. We suggest that the Gozha–Longmu Red River fault systems. In addition, the origin, ing, right-slip Karakoram fault in western Co fault system formed during late Miocene and thus general kinematic signifi cance of fault Tibet ranges from transpression in the north to Pliocene structural reorganization of the junctions such as that defi ned by the intersection (37° to 34°N) to transtension in the south (34° southwestern Altyn Tagh and southern Kara- of the San Andreas and Garlock faults, remains to 32°N). This transition in structural style koram fault systems to allow eastward migra- an open question (e.g., Bohannon and Howell occurs at a 27-km-wide bend in the fault. Our tion of the Tibetan Plateau and northward 1982; Spotila and Anderson, 2004). This defi - new neotectonic mapping has documented migration of the Pamir syntaxis. ciency in the understanding of the geometric the long-asserted structural linkage between and kinematic evolution of intersecting major the ENE-striking Gozha–Longmu Co fault Keywords: strike-slip faults, Himalayan orog- faults warrants further investigation of natural system and the similarly oriented active, left- eny, kinematics, fault zones, triple junctions. examples of such systems. slip Altyn Tagh fault to the northeast. This A number of very large active strike-slip mapping also indicated that the restraining INTRODUCTION fault systems cut the Indo-Asian collision zone bend in the Karakoram fault is located where (Fig. 1, inset), the largest active continental oro- this fault intersects the Gozha–Longmu Co Understanding the pattern of deformation that gen on Earth. As a result, this collision serves as extension of the Altyn Tagh fault to the west. occurs where two or more fault systems inter- an excellent natural laboratory to investigate the Additional observations from remotely sensed sect is a basic problem in determining how fault geometric and kinematic evolution of intersect- imagery suggest that the total left-separation systems help accommodate continental defor- ing major faults (Armijo et al., 1989; Molnar and along the Gozha–Longmu Co fault system is mation during orogenesis. Several studies have Tapponnier, 1978; Peltzer et al., 1989; Tappon- 25–32 km. We use the new neotectonic map- addressed this problem. Bohannon and Howell nier and Molnar, 1977). These structures have ping and published slip rates to develop a (1982) examined the intersection between the strike lengths in excess of 1000 km and cumula- simple kinematic model for the main active San Andreas and Garlock faults and argued that tive displacements of several hundred kilometers faults in western Tibet to explore the genetic slip on the Garlock fault may have infl uenced or more (Armijo et al., 1989; Molnar and Tap- relationship between slip along the Gozha– the formation of the Big Bend and the Big Pine ponnier, 1975; Molnar and Tapponnier, 1978; Longmu Co fault system and the geometric fault. Work by Wang et al. (1998) demonstrated Peltzer et al., 1989; Tapponnier and Molnar, and kinematic evolution of the Karakoram that slip along the Xianshuihe–Xiaojiang fault 1977). One of the clearest zones of intersection fault. This model combines published geodetic resulted in an ~60-km-wide defl ection of the between two major strike-slip faults within this and Quaternary slip rates with the known Red River fault. Additional studies of the San collision zone occurs in western Tibet, where fault geometries and demonstrates that the Andreas system and Eastern California shear the 325°-striking, right-slip Karakoram fault lies transition from transpression to transtension zone have used geodetic data to show that fault near the southwest end of the Gozha–Longmu along the Karakoram fault can be explained intersections are often associated with increased Co extension of the 070°-striking, left-slip Altyn by differential motions between the NW Hima- strain rates (King and Cocco, 2001; Snay et al., Tagh fault (Fig. 1) (e.g., Peltzer et al., 1989). The laya, the Tianshuihai terrane, and the Tibetan 1996) and may result in transient strain accumu- active, left-slip Gozha–Longmu Co fault system Plateau. These motions produce bending and lation, secular variation in fault slip rates, and separates the Tianshuihai terrane to the north- transtension along the central and southern earthquake clustering (e.g., Peltzer et al., 2001). west from the Tibetan Plateau to the southeast Karakoram, respectively, and movement of Still other studies have shown that fault inter- (Fig. 1) (Liu, 1993). Evaluation of the extent to the Tibetan Plateau at a rate of 6–13 mm/ sections play fundamental roles in how fault which the Altyn Tagh and Karakoram faults may yr toward the east-southeast relative to the systems geometrically evolve and transfer strain infl uence one another requires investigation of Pamirs. We also fi nd that the Gozha–Longmu throughout an orogen (Ando et al., 2004; Spotila the geometric and kinematic evolution of the and Anderson, 2004; Van der Woerd et al., 1999). zone of intersection between them (Avouac and *E-mail: [email protected]. Most of these previous studies have focused on Tapponnier, 1993; Molnar and Tapponnier, 1975,

Geosphere; April 2007; v. 3; no. 2; p. 71–85; doi: 10.1130/GES00067.1; 7 ﬁ gures; 1 table, 1 online map.

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/3/2/71/893719/i1553-040X-3-2-71.pdf by guest on 27 September 2021 Raterman et al.

K 75°E o 81°E 2004a; Phillips et al., 2004; Searle, 1996). Its n

S vertical extent remains to be determined. h

a Tarim Basin n China Figure 2 Previous work along the Karakoram fault has

Altyn Tagh F. established that the late Miocene to recent slip W Pamirs estern Kunlun Shan direction along the fault north of its intersec- Karak 36˚N ax Fault tion with the Gozha–Longmu Co fault (34.5°N latitude) differs from that observed to the south. ~150 km Tianshuihai separation of the Along the northern Karakoram fault, transpres- Baltoro granite o F. ha C sional deformation is suggested by Neogene to Goz Tajikistan S out h K Quaternary fault strands that have both thrust ai las T Co F. hr Longmu and strike-slip kinematics (Searle et al., 1998) NW Himalaya us t and apatite fi ssion-track ages that indicate 5 Ma Pakistan Tibetan Plateau to recent rapid exhumation of the ranges fl ank- High Himalay Exhumed Pangong Maximum extension ing the fault (Foster et al., 1994). In contrast, Range (Baltoro directions 85o ± 28o 33°N Granite + Tangtse Leucogranite) transtensional deformation is suggested along Shi quan a he F. the southern portion of the Karakoram fault K a ra by active normal faults and tension gash data k o ra o o (Murphy and Burgess, 2006; Murphy et al., m 89 ± 5 F . 2000, 2002; Ratschbacher et al., 1994), as well as reconstructed offset terraces and moraines S S o ou u th t K h ai (Brown et al., 2002; Chevalier et al., 2005). T las ib Thr eta ˚ ust n De Although this variation in structural style is a tac Gurla Mandhata hm e India nt fi rst-order characteristic of late Cenozoic slip 81°E 0 km 100 along the Karakoram fault, the cause of this variation remains poorly understood. Figure 1. Simplifi ed map of major Cenozoic structures in western portion of India- Here we investigate the extent to which this Asia collision zone. Blue corner ticks delineate area shown in dynamic Web-based map pronounced variation in structural style along (http://dx.doi.org/10.1130/GES00067.s1, which is shown in simplifi ed form in Figure 2. the Karakoram fault may be due to the relative Arrows with wedges indicate the mean extension directions (arrow) and error (wedge) motions between the NW Himalaya, the Tian- compiled from measurements of normal faults and tension gashes near the southern Kara- shuihai terrane, and the Tibetan Plateau (Fig. 1) koram fault (Ratschbacher et al. [1994] in the north and Murphy et al. [2000] in the south). and associated slip along the Gozha–Longmu Co Map is simplifi ed from Murphy et al. (2000) and Phillips et al. (2004). Black earthquake extension of the Altyn Tagh fault. We present new focal mechanisms are from Harvard Centroid Moment Tensor (CMT) catalog (http://www. (Fig. 2, red faults) and compiled (Fig. 2, orange, seismology.harvard.edu/CMTsearch.html). Green earthquake focal mechanisms are from yellow, and blue faults) neotectonic mapping of Armijo et al. (1986). Inset shows location of Figure 1 in context of major structures within the region encompassing the intersection of the the India-Asian collision zone. western Altyn Tagh and Karakoram faults. We then review previously published work on the slip rates of these faults and combine them with 1977; Peltzer and Tapponnier, 1988; Tapponnier 87°E to 90°E, the fault is thought to extend to the the newly mapped fault geometries to construct and Molnar, 1976; Tapponnier et al., 1982). base of the lithosphere (Wang et al., 2003; Witt- a simple kinematic model of the intersection of The Altyn Tagh fault (Fig. 1) extends for over linger et al., 1998; Zhao et al., 2006). The Kara- the Altyn Tagh and Karakoram faults using the 1200 km and separates the Tibetan Plateau to the koram fault extends over 1000 km and separates principles for plate triple junctions (McKenzie south from the Tarim Basin to the north (Mol- the Pamirs and NW Himalayas to the southwest and Morgan, 1969). Although our analysis is nar et al., 1987; Peltzer et al., 1989; Tapponnier from the Tibetan Plateau to the northeast. Total broadly similar to that of Liu (1993), here we and Molnar, 1977). Total left-slip along the fault right-slip along the Karakoram fault is disputed, are focused on understanding along-strike vari- exceeds 450 km (Cowgill et al., 2003; Peltzer et but it appears to be between 150 and 500 km ation in structural style along the Karakoram al., 1989; Ritts and Biffi , 2000; Ritts et al., 2004; along the central portion of the fault, which lies fault and the role of the Longmu Co fault, if any, Yue et al., 2001, 2004a, 2004b, 2005). From between 34° to 36°N latitude (Lacassin et al., in producing this variation. Our work suggests

Figure 2. Geometry and kinematics of active faulting in western Tibet at the Karakoram–Altyn Tagh intersection. (A) Simplifi ed version of the dynamic, Web-based neotectonic map (which can be found at http://dx.doi.org/10.1130/GES00067.s1). The map was compiled from our analysis of Landsat Thematic Mapper (TM) and Enhanced Thematic Mapper Plus (ETM+) imagery, previous research (Armijo et al., 1986; Peltzer et al., 1989; Liu, 1993; and Avouac and Peltzer, 1993), and our own fi eld observations. Base Landsat data are ETM+ from http://onearth.jpl.nasa.gov/. Note the absence of geomorphic evidence for active structures within the interior of the Tianshuihai terrane. (B) Plots showing 10-km-wide topographic swaths perpendicular to the northern and southern Karakoram fault. The gray band shows the difference between the minimum and maximum elevation at each point along the profi le, and the black line represents the mean elevation within the swath. The swaths shown are representative of the topography across the entire northern and southern stretches of the fault.

72 Geosphere, April 2007

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/3/2/71/893719/i1553-040X-3-2-71.pdf by guest on 27 September 2021 Triple-junction explains variable structural style along the Karakoram fault

78°E 80°E 82o A Strike-slip Fault Fault (Armijo et al., 1986) Fault (Avouac and Hot Normal Fault an Fault Peltzer, 1993) 37.5°N Tarim Thrust Fault Fault (Liu, 1993) 37˚30' Fault (Confident) Fault (Peltzer et al., 1989)

Fault (Approximate) Fault (This Study) N Kunlun Pediment Fault (?)

0 50 100 km N Karakax ? Fault Fig. 3A Altyn Tagh Fault 36˚N Fig. 3B 36°N Tianshuihai No geomorphic evidence for ? internal deformation ′ A Zone of inferred ozha Co Fault distributed deformation G Fig. 4A from Wright et al., 2004 Fig. 3C A B′ ′ Fig. 4B B C 34.5˚N Fig. 3D Longmu Co Fault 34.5°N C ~27 km bend in the K Fig. 4C Karakoram fault at its arakoram F intersection with the ′ Longmu Co Fault D Fig. 3E E′ Tibet NW Himalya D ault F′ E 78°E F 80°E 82°E

B Profiles across the Karakoram Fault north of the bend A A′ B B′ C C′ 7000 7000 Karakoram F. 7000 Karakoram F.

Max. 6000 6000 6000 Karakoram F.

5000 5000 5000 Elevation (M) Elevation (M) Mean Elevation (M)

4000 4000 4000 Min.

3000 3000 3000 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 Distance (Km) Distance (Km) Distance (Km) Profiles across the Karakoram Fault south of the bend D D′ E E′ F F′ 7000 Karakoram F. 7000 Karakoram F. 7000 Karakoram F.

6000 6000 6000

5000

5000 Elevation (M) 5000 Elevation (M) Elevation (M)

4000 4000 4000

3000 3000 3000 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 Distance (Km) Distance (Km) Distance (Km)

Geosphere, April 2007 73

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/3/2/71/893719/i1553-040X-3-2-71.pdf by guest on 27 September 2021 Raterman et al.

that rigid-block kinematics provides an accurate geometrically linked with the Altyn Tagh fault, Background and Methods fi rst-order depiction of continental deformation and (3) to identify and map in detail the inter- in this region. section between the Karakoram and Gozha– The previously published neotectonic maps Longmu Co faults. Although Peltzer et al. (1989) from western Tibet fall into two broad cat- NEOTECTONIC MAPPING hypothesized that the Gozha–Longmu Co fault egories: (1) regional compilation maps (e.g., system links with the Altyn Tagh fault, detailed Armijo et al., 1986, 1989), which generally Several important studies have reported neo- maps of this area have not been reported; thus, include minimal observational data in sup- tectonic mapping from western Tibet (Armijo this linkage remained to be defi nitively estab- port of the reported mapping, and (2) detailed et al., 1986, 1989; Peltzer et al., 1989), and lished until this study. Likewise, detailed map- maps of 10–50-km-long reaches of the faults, the most extensive work has been reported in a ping of the intersection between the Karakoram in which the geomorphic criteria used to Ph.D. thesis by Liu (1993). To build on this work and Gozha–Longmu Co faults also remained deduce the fault geometry and kinematics are and better determine the geometry and kinemat- to be reported. Confi rmation of these linkages clear but must be extrapolated over large dis- ics of active faulting in western Tibet, we com- suggests that the major faults of western Tibet tances to infer the kinematics along the entire piled a new neotectonic map of the area using form a fault circuit, and this justifi es the use of length of the fault (e.g., Liu, 1993; Peltzer et the published maps, shown as orange, yellow, the triple junction approach to construct an ini- al., 1989). The neotectonic mapping presented and blue faults in Figure 2, and new results from tial kinematic model of western Tibet. The new in this study (Web Map and Fig. 2) builds our own analysis of remotely sensed imagery compilation is provided as an interactive, Web- on this previous work in three ways. First, it locally augmented with fi eld mapping, shown as based map (located at http://dx.doi.org/10.1130/ offers the opportunity to simultaneously view red lines in Figure 2. Our work had three aims: GES00067.S1 and henceforth referred to the both the mapping and the primary data from (1) to reproduce the earlier results, (2) to deter- as the Web Map), and it is shown in simplifi ed which the mapping was derived over the whole mine if the Gozha–Longmu Co fault system is form in Figure 2. area, thereby permitting users to independently evaluate the geomorphic criteria we used to generate the map. This capacity is particularly important for regions that we interpret to show minimal geomorphic evidence of active faulting, such as in the center of the Tianshuihai terrane (Fig. 1). Second, the level of mapping detail we report is comparable to that in the detailed maps presented by Liu (1993) and Peltzer et al. (1989), but it spans an area comparable to that covered in summary form by the regional work of Armijo et al. (1986). To identify geomorphic evidence of active faulting between 33° and 37°N latitude and 76° and 83°E longitude, we initially used the Real- Time Interactive Mapping System (RIMS), a new interactive three-dimensional (3-D) visu- alization and analysis tool that allows extrac- tion of surface information through real-time, georeferenced vector mapping on a virtual terrain model from any perspective (see Fig. 3). In RIMS, bands 7, 4, and 2 from a mosaic of Landsat 7 Thematic Mapper (TM) (http://glcf. umiacs.umd.edu/index.shtml) scenes with a resolution of 28 m/pixel were draped over a Shuttle Radar Topographic Mission digital Web Map: The full-scale, Web-based, version of the neotectonic map shown in simplifi ed elevation model (DEM), which has a resolu- form in Figure 2a. This interactive map can be found at http://dx.doi.org/10.1130/GES00067. tion of ~90 m/pixel. File-size limitations in S1. The original mapping was compiled from Landsat Thematic Mapper and Landsat RIMS necessitated a second iteration of map- Enhanced Thematic Mapper Plus (ETM+) remotely sensed imagery in the Real-Time Inter- ping at a higher resolution in Google Earth active Mapping System (RIMS) (Bernardin et al., 2006) and Google Earth (http://earth. Plus! (http://earth.google.com) using bands 1, google.com). Display of the remotely sensed imagery and neotectonic mapping is accom- 2, and 3 from a mosaic of Landsat 7 Enhanced plished using a freely available open-source Flash-based application titled WorldKit (http:// Thematic Mapper Plus (ETM+) scenes (http:// www.worldkit.org). The scale of the map can be changed by left clicking on the magnifying onearth.jpl.nasa.gov/), pan-sharpened using glasses with the plus (+) and minus (-) signs. Panning on the map is accomplished by either panchromatic band 8 to increase the spatial holding the left mouse button and dragging in the map window or by using the directional resolution of these bands to 14.25 m/pixel. arrows. Click on the checkbox next to the label “Mapping On” to view the imagery with or Where possible, the kinematics of previously without the neotectonic mapping. Click on the button titled “Full Extent” to return to the unmapped faults were determined using stan- full map scale. See the text for further explanation of the geomorphic criteria used to create dard geomorphic criteria (e.g., Molnar and the map. Please note that the map might take several minutes to load. Tapponnier, 1978).

74 Geosphere, April 2007

Active Faulting in Western Tibet characterize the deformation in the vicinity complex set of left-stepping structures in the of 83°E (Web Map and Fig. 3A). To the west, vicinity of Longmu and Sumxi lakes transfers The dynamic Web-based map (Web Map) the active trace of the Altyn Tagh fault splays displacement along the Gozha Co fault to the and Figure 2 support previous work indicating and becomes less discrete as it approaches the left-slip Longmu Co fault, as was also shown by strike-slip motion along four major fault zones. intersection between the Karakax and Gozha– Liu (1993). In particular, the Altyn Tagh, Karakax, Gozha– Longmu Co faults between 81°E and 83°E (Web Our mapping along the Gozha–Longmu Longmu Co, and Karakoram faults dominate Map) in a geometry similar to that mapped pre- Co fault system also appears to constrain the active faulting in the region. Geomorphic evi- viously (Chinese State Bureau of Seismology, total slip along the system to 25–32 km. Along dence of large dip-slip structures occurs in only 1992). The Karakax fault, commonly referred to three widely separated reaches of the system, two areas. The fi rst is along the northern mar- as the western extension of the Altyn Tagh fault gin of the Western Kunlun Shan (Figs. 1 and 2), (e.g., Peltzer et al., 1989), strikes 110° and forms where we infer the location of the blind Hotan the southern boundary of the Western Kunlun thrust from the northernmost extent of the region Shan (Web Map and Fig. 2). We differentiate the A Altyn Tagh Fault (View to NE) of recent incision, as described by Avouac and Karakax and Altyn Tagh faults here because they Peltzer (1993). The second case occurs in the have signifi cantly different regional orientations. southeastern portion of the mapped area (Fig. 2), Along the Karakax fault, left-laterally offset fl u- where diffuse N-S–striking normal faults are the vial terraces and glacial moraines clearly mark predominant active structures (e.g., Tapponnier the trace of the fault within the Karakax valley et al., 1981). Most of the active deformation (Ding et al., 2004; Peltzer et al., 1989; Ryerson occurs along the margins of the Tianshuihai et al., 1999) (Web Map and Fig. 3B). However, terrane, within which we found no clear geo- the active trace of the Karakax fault becomes B Karakax Fault (View to NE) morphic evidence of active internal deformation poorly expressed for ~150 km between ~79°E, (Web Map and Fig. 2). This lack of geomorphi- where it leaves the Karakax valley, and ~81°E, cally expressed active faulting contrasts with where it joins one of the main splays of the Altyn recent Interferometric Synthetic Aperture Radar Tagh fault (Web Map). (InSAR) measurements (Wright et al., 2004), The 070–090°–striking Gozha–Longmu Co suggesting that the observed InSAR signal may fault system (Liu, 1993) extends from the Altyn result from atmospheric interferences, as sug- Tagh fault in the northeast to the Karakoram gested by Wright et al. (2004), or from recent fault in the southwest (Web Map). The north- C Ghoza Co Fault (View to SE) initiation of rapid slip along a fault system that eastern end of the fault system lies between is poorly expressed geomorphically. The geo- 81.5°E and 83°E, where NE-SW–striking faults morphic setting of this locality on the high and associated with predominantly extensional arid Tibetan Plateau makes it unlikely that the earthquake focal mechanisms (Fig. 1) mark its InSAR data refl ect a long-lived, fast-slipping trace. This set of faults merges with the diffuse fault that has no geomorphic expression because splays of the southwestern Altyn Tagh fault and the pace of geomorphic resurfacing outpaces the appears to transfer left-slip from the Altyn Tagh rate of surface deformation. fault to the Gozha Co fault. As the Web-based Along the Altyn Tagh fault, well-defi ned fault map (Web Map) and Figure 3C indicate, drain- scarps, shutter ridges, and both beheaded and ages crossing the Gozha Co fault show sinistral defl ected drainages that show left- separations separations. To the southwest, a geometrically D Northern Karakoram Fault (Map View)

Figure 3. Geomorphic indicators used to determine the kinematics of major fault systems in western Tibet. In all cases the active fault trace is indicated with red triangles, disturbed drainages are outlined in blue, and offsets are indicated by dashed bars. Locations are indicated in Figure 2. (A) Left-laterally displaced ﬂ uvial terrace risers along the Altyn Tagh fault. (B) Left-laterally displaced glacial moraine along the Karakax fault. (C) Left-laterally displaced stream channels along the Gozha Co extension of the Altyn Tagh fault. (D) Right- laterally displaced ﬂ uvial terrace riser within the bend that separates the northern and southern portions of the Karakoram fault. Red arrows point along fault traces. (E) Right- N laterally displaced glacial moraines (dashed black lines) and drainages (blue lines) along the southern Karakoram fault. A normal fault lies at the base of the faceted range front, E Southern Karakoram Fault (View to W) as indicated by the red dashed line decorated with ticked balls. Panel A shows Landsat Enhanced Thematic Mapper Plus (ETM+) imagery (http://onearth.jpl.nasa.gov/) draped over the SRTM90 digital elevation model as viewed in Google Earth Plus!. Panels B, D, and E show Landsat Thematic Mapper (TM) imagery (http://glcfapp.umiacs.umd.edu/index. shtml) draped over the Shuttle Radar Topographic Mission (SRTM90) digital elevation model as viewed in the Real-Rime Interactive Mapping System (RIMS). The topography is exaggerated by 2× in the virtual terrain views shown in panels A, B, D, and E.

Geosphere, April 2007 75

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/3/2/71/893719/i1553-040X-3-2-71.pdf by guest on 27 September 2021 Raterman et al.

bedrock markers appear to show total separa- located at 34.5°N latitude and is ~27 km wide, mechanisms (Fig. 1) and observations of faults tions of ~25 km (Fig. 4A), ~32 km (Fig. 4B), as measured perpendicular to the trace of the with thrust and strike-slip motions (Searle et al., and ~28 km (Fig. 4C). We identifi ed these sepa- Karakoram fault outside of the bend. To the 1998) indicate transpressional kinematics, fur- rations by matching contacts between regions north of this intersection, the occasional occur- ther supporting this morphologic distinction. As of differing spectral response (and thus color) rence of a distinct fault trace showing right-lat- the following kinematic analysis demonstrates, in the Landsat image on opposite sides of the erally offset terraces weakly defi nes the trace of the geomorphic and kinematic differences fault. Although we have not completed bedrock the Karakoram fault. Conversely, to the south, between the northern and southern Karakoram mapping of these areas to confi rm the proposed fault scarps that simultaneously expose steep fault near its intersection with the Longmu Co correlations, our experience in the area com- triangular facets and right-laterally offset gla- fault have signifi cant implications. paring our fi eld mapping with color changes cial moraines attest to oblique, right-normal in equivalent Landsat images suggests that the motion (Fig. 3E) and clearly defi ne the trace COMPILATION OF SLIP RATES separations shown in Figure 4 likely refl ect true of the Karakoram fault. As Figure 2 indicates, bedrock displacements. Therefore, 25–32 km fault-perpendicular topographic profi les across In addition to fault geometries and slip direc- constitutes the fi rst rough estimate of total left the northern and southern reaches of the Kara- tions, we must also know the slip rates before slip along the Gozha–Longmu Co fault system. koram fault also highlight the differences in the we can construct a kinematic model of western The Karakoram fault strikes 325° and extends geomorphic expression of the fault north and Tibet. The slip rates on the Karakoram, Altyn from near the Kongur Shan in the north to the south of the bend. In particular, the northern Tagh, and Karakax faults as well as the rate Gurla Mandhata detachment in the south Karakoram fault lies in the bottom of a deep of thrusting in the Western Kunlun Shan are (Fig. 1). The Karakoram fault is characterized axial valley that is fl anked on both sides by steep all particularly critical. Because geodetic and by two geomorphically distinct regions that are topography, whereas the southern Karakoram Quaternary rates along these structures show separated by a restraining double bend in the fault lies within an axial valley that is both shal- systematic differences, we use both in our kine- fault near its intersection with the Longmu Co lower and wider than that to the north. Along matic analysis. Table 1 presents a compilation fault (Web Map; Figs. 2 and 3D). This bend is the northern Karakoram fault, earthquake focal of known slip rates for these structures.

~25 KM

0 10 kmN B ~32 KM

0 10 kmN

C 0 10 kmN

~28 KM

Figure 4. Landsat Enhanced Thematic Mapper Plus (ETM+) imagery showing three apparent left-lateral bedrock separations along the Gozha–Longmu Co fault. Left panel shows the uninterpreted Landsat image, and right panel shows a simpliﬁ ed geologic map of inferred bedrock units and their separated contacts. See Figure 2 for locations. On each simpliﬁ ed map, the contacts that are apparently separated are indicated by a black arrow, and the active fault trace is shown in red. Distances are measured as a straight line from arrow to arrow using the scale provided.

76 Geosphere, April 2007

Reported slip rates along the Altyn Tagh and TABLE 1. PUBLISHED SLIP RATES USED IN VELOCITY TRIANGLES Karakoram faults appear to vary signifi cantly, Rate (mm/yr) Error Time frame Source depending on the method used to obtain them. (mm/yr) For instance, reconstructions of cosmogenically Western Kunlun Shan dated (10Be) fl uvial terraces and moraines yield 2.0 2.0 Geodetic Shen et al. (2001) slip rates of 26.9 mm/yr along the central Altyn 4.5 3.0 Quaternary Avouac and Tapponnier (1993) Tagh fault (Meriaux et al., 2004) and 10.7 mm/ Karakax fault yr (Chevalier et al., 2005) or 4 mm/yr (Brown et 5.0 5.0 Geodetic Wright et al. (2004) al., 2002) along the southern Karakoram fault. In 7.0 3.0 Geodetic Shen at al. (2001) contrast, slip rates determined using global posi- 6.0 2.9 tioning system (GPS) measurements for the Altyn 17.5 5.5 Quaternary Ryerson et al. (1999) Tagh and Karakoram faults are 9 mm/yr (Bendick Altyn Tagh fault et al., 2000; Shen et al., 2001) and 4 mm/yr (Jade 9.0 4.0 Geodetic Wallace et al. (2004) et al., 2004), respectively. We favor the geodetic 7.4 1.0 Geodetic Xiong et al. (2003) rates because they better explain the timing of 9.0 2.0 Geodetic Shen et al. (2001) exhumation of mid-crustal rocks along the south- 6.0 2.0 Geodetic Chen et al. (2000) ern Karakoram fault, more closely approximate 9.0 2.0 Geodetic Chen et al. (2000) the Cenozoic-averaged slip rates for these faults 9.0 5.0 Geodetic Bendick et al. (2000) 5.6 1.6 Geodetic Zhang et al. (2004) (Cowgill et al., 2003; Lacassin et al., 2004a; Phil- 7.9 1.1 lips et al., 2004; Searle, 1996; Yin et al., 2002; 17.8 3.6 Quaternary Meriaux et al. (2005) Yue et al., 2001, 2004a), and do not involve dis- 17.5 2.0 Quaternary Xu et al. (2005) puted age determinations for offset features as 26.9 6.9 Quaternary Meriaux et al. (2004) discussed by Brown et al. (2005), Cowgill (2007), 15.0 5.0 Quaternary Washburn et al. (2003) and England and Molnar (2005). 19.3 2.4 Karakoram fault ANALYSIS OF TRIPLE JUNCTION 3.4 5.0 Geodetic Jade et al. (2004) KINEMATICS 1.0 3.0 Geodetic Wright et al. (2004) 11.0 4.0 Geodetic Banerjee and Bürgmann (2002) 5.1 1.2 Conceptual Background 4.0 1.0 Quaternary Brown et al. (2002) 10.7 0.7 Quaternary Chevalier et al. (2005) Published literature contains a range of 18.0 8.0 Quaternary Liu (1993) kinematic models of continental deformation, 12.5 2.5 Quaternary Liu (1993) including two-dimensional (2-D) rigid-block 11.3 2.1 models (e.g., Jezek et al., 2002), microplate models that use Euler poles to describe relative motions between plates on a sphere (e.g., Peltzer and Tapponnier, 1988). The new neo- plates using both geodetic and Quaternary slip Avouac and Tapponnier, 1993; Liu, 1993; Pelt- tectonic mapping described herein and shown rates. Our analysis (1) predicts the direction and zer and Saucier, 1996; Replumaz and Tappon- as red lines on the Web-based map and Figure 2 rate of slip along the Gozha–Longmu Co fault nier, 2003), and continuum models that employ confi rms the linkage between the Altyn Tagh, system, (2) explains along-strike variation in the a wide range of initial and boundary conditions, Karakax, Gozha–Longmu Co, and Karakoram kinematics of the Karakoram fault, and (3) pro- as well as constitutive relationships (e.g., Bird faults and justifi es the triple junction approach of duces a testable kinematic model for the evolu- and Piper, 1980; England and McKenzie, 1982; our initial kinematic model. Kinematic analysis tion of the junction between the Karakoram and Vilotte et al., 1982). The simplest of such kine- of isolated triple junctions has proven successful Longmu Co faults. matic models are 2-D, “fl at-earth” models, in in explaining spatial and temporal variations in The analysis employs the following fi ve which discrete block boundaries are used to fault kinematics in both oceanic and continental assumptions: (1) The blocks are perfectly rigid. approximate complex fault systems. Here we deformation zones (e.g., Dickinson and Snyder, (2) The relative velocity along fault segments start with this most basic approach to explore 1979; Kleinrock and Morgan, 1988). does not change, or the blocks do not rotate. the fi rst-order kinematics of western Tibet. In Using the fault geometries and structural style (3) Motion along the Altyn Tagh and Karakax particular, we aim to evaluate if relative motions from the new neotectonic mapping reported here, faults is parallel to the regional strike of these between the NW Himalaya, the Tianshuihai we simplifi ed the mapped region into four inter- faults (i.e., pure strike-slip). (4) Multiple strands terrane, and Tibet can produce both the along- nally rigid blocks: Tarim, Tianshuihai, Tibet, and of the Gozha–Longmu Co fault system are strike variation in structural style along the the NW Himalaya. By using published slip rates simplifi ed into a single fault to defi ne a bound- Karakoram fault and the bend at its intersection for several of the block-bounding faults, we then ary between the Tibet and Tianshuihai blocks. with the Longmu Co fault. solved for all of the relative motions between (5) No estimate exists for the magnitude of A critical component of such a block model is the blocks using the principles of plate triple transpression across the northern Karakoram the intersections of major block-bounding faults, junctions (i.e., velocity triangles) (McKenzie fault; thus, only the azimuth of the fault trace and which should defi ne triple junctions, and the and Morgan, 1969). The discrepancy in geodetic the magnitude of the slip rate are left to defi ne kinematic evolution of these junctions, which and Quaternary slip rates has signifi cant impli- relative motion between Tianshuihai and the should produce diagnostic variations in struc- cations for this analysis, and for this reason, we NW Himalaya. Motion along the northern Kara- tural style along these faults (e.g., Liu, 1993; calculated the relative motions between the four koram fault is assumed to be pure strike-slip, but

Geosphere, April 2007 77

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/3/2/71/893719/i1553-040X-3-2-71.pdf by guest on 27 September 2021 Raterman et al.

Geodetic slip rates Quaternary slip rates

Step 1: Determine Tarim / WESTERN KUNLUN SHAN Tianshuihai Relative Movement

5 +/- 3 mm/yr 18 +/- 6 mm/yr WESTERN KUNLUN SHAN o o 24 +/- 5 103o +/- 6o 2 +/- 2 mm/yr 6 +/- 3 mm/yr o o 24o +/-5o) 103 +/- 6

TIANSHUIHAI TARIM 7 +/- 3 mm/yr 19 +/- 5 mm/yr o o TIANSHUIHAI 265 +/- 14 TARIM 268o +/- 11o

Step 2: Determine Tianshuihai / Tibet Relative Movement

TIBET TIBET 8 +/- 1 mm/yr 3 +/- 1 mm/yr 202o +/- 12o 31o +/- 33o 8 +/- 3 mm/yr TIANSHUIHAI 19 +/- 2 mm/yr 328o +/- 16o 202o +/- 12o TARIM 7 +/- 3 mm/yr 85o +/-14o

19 +/- 5 mm/yr TIANSHUIHAI 88o +/- 11o

TARIM Step 3a: Determine India / Tibet Relative Movement

6 +/- 1 mm/yr HIMALAYA HIMALAYA 97o +/- 19o TIBET

5 +/- 1 mm/yr 13 +/- 2 mm/yr o o 319o +/- 15o 119 +/- 12 3 +/- 1 mm/yr TIBET 11 +/- 2 mm/yr 211o +/- 33o TIANSHUIHAI 319o +/- 15o

8 +/- 3 mm/yr Step 3b: Determine India / Tibet 148o +/- 16o Relative Movement with 10 TIANSHUIHAI degrees of oblique convergence across the N Karakoram Fault HIMALAYA 6 +/- 1 mm/yr o o HIMALAYA 89 +/- 24 TIBET TIBET 5 +/- 1 mm/yr 11 +/- 2 mm/yr o o 329 +/- 15 329o +/- 15o

3 +/- 1 mm/yr 13 +/- 2 mm/yr o o TIANSHUIHAI 211o +/- 33o 117 +/- 11

2 mm/yr 8 +/- 3 mm/yr 148o +/- 16o TIANSHUIHAI

78 Geosphere, April 2007

Figure 5. Velocity triangles showing relative motions between the four blocks outlined in Figure 2, as derived using geodetic (left side) and Quaternary (right side) slip rates. Numbers next to each vector indicate the rate and azimuth of the slip direction (±2σ errors). Black arrows show average rates derived from the published data shown in Table 1. The colored arrows are resultant vectors calculated at each step, and they represent the average of all possible combinations of published fault slip rates. In step 1 (top pair of triangles), we derive Tarim-Tian- shuihai relative motion (green vector) using published values (black vectors) for the shortening rate across the Western Kunlun Shan and the slip rate along the Karakax fault. In step 2 (middle pair of triangles), we obtain Tianshuihai-Tibet relative motion (blue vectors) using the Tarim-Tianshuihai motion (green vector) calculated in step 1 and the published slip rates for the Altyn Tagh fault (black vector). In step 3 (bottom pair of triangles), we calculate NW Himalaya–Tibet relative motion (red vector) using the Tianshuihai-Tibet relative motion (blue vector) calculated in step 2 and the published slip rate along the northern Karakoram fault (black vector). See the text and Table 1 for additional information and explanation.

the effect of transpression along this segment of along the Altyn Tagh and Karakoram faults. The second step yielded the average for the fault is also considered (Foster et al., 1994; These large discrepancies result in an average possible Tianshuihai-Tibet relative motions, Searle et al., 1998). derived motion that, when combined with the shown as blue vectors in Figure 5. To do this, average of the known slip rates, does not result we combined the individual Tarim-Tianshuihai Velocity Triangles in a closed velocity triangle. vectors computed in step 1 (averages of which In the fi rst step, we determined average are shown as green vectors in Fig. 5) with indi- In order to use the slip rates and fault geom- Tarim-Tianshuihai relative motion (green vector vidual slip-rate values along the Altyn Tagh etries to calculate relative block motions, we in Fig. 5) by combining range-perpendicular fault and an assumption of pure left slip along calculated the set of velocity triangles shown shortening values across the Western Kunlun the Altyn Tagh fault to constrain the azimuth in Figure 5, which proceeded along the lines Shan with left-slip values along the Karakax of the Tarim-Tibet vector (average Tarim-Tibet of a similar analysis conducted by Liu (1993). fault, and the average values are shown as black vectors are shown in black in Fig. 5 and are Velocities derived from the published data vectors in Figure 5. Velocity triangles derived listed in Table 1). Because this step yielded shown in Table 1 are shown in black and rep- using geodetic and Quaternary slip rates yielded average Tianshuihai-Tibet relative motion, resent the arithmetic averages of the fault slip Tarim-Tianshuihai rates of 6.7 ± 2.9 mm/yr and it also predicted the rate and direction of slip rates published for the Hotan, Karakax, Altyn 18.93 ± 5.1 mm/yr, respectively (Fig. 5). Rela- along the Gozha–Longmu Co fault system. In Tagh, and Karakoram faults. The colored values tive to Tianshuihai, Tarim moves toward 265° particular, the blue vectors in the velocity tri- were derived according to the methods outlined ± 14° and 268° ± 11° in the geodetic and Qua- angle shown in step 2 of Figure 5 predict that in the remainder of this section and give average ternary velocity triangles, respectively. Tibet moves at either 2.8 ± 1.1 mm/yr toward relative motions between Tarim and Tianshui- hai (green), Tianshuihai and Tibet (blue), and Tibet and the NW Himalaya (red and purple). We used the following approach to account for A Initial B Final uncertainty in the choice of vectors when con- structing a velocity circuit. At each step, we fi rst Tianshuihai Tianshuihai used all possible combinations of individual values for the two known legs to compute all pos- ~28 km ~25 km sible individual values for the unknown third leg NW Himalaya ~32 km by assuming that each individual circuit closes. Future NW Himalaya 85o ± 28o Gozha-Longmu Co We then computed the arithmetic mean of these No Bend in Fault System individual solutions to determine the derived Karakoram Fault Tibet Tibet Present day bend 89o ± 5o vectors that are shown for each step in Figure 5. width = ~27 km Thus, while the circuit for each individual solution was closed, the combinations of these aver- C Time needed to restore the 27 km bend in the Karakoram Fault age velocities shown in Figure 5 do not always Active Fault Trace Slip-rate type Longmu Co Fault slip-rate (Tianshuihai-Tibet relative movement) Time to restore bend close because these velocity triangles are not Inactive Fault Trace Geodetic 3 +/- 1 mm/yr 10 +/- 3 m.y. Bedrock Separations actual solutions. They are a comparison of the average known slip rates with the average of all Quaternary 8 +/- 3 mm/yr 3 +/- 3 m.y. N 0 100 km unknown vectors that could be determined from Figure 6. Plate boundaries after restoring the ~27 km bend in the Karakoram fault using the slip rates in Table 1. The extent to which the block motions derived from geodetically determined slip rates within a NW Himalaya–fi xed triangles shown in Figure 5 close thus refl ects reference frame. (A) Proposed initial fault geometry derived by restoring the bend. (B) Fi- the extent to which the data from which they nal (present) fault geometry. (C) The time needed to restore the 27-km-wide bend in the are derived are homogeneous (closed) or het- Karakoram fault using the velocities determined in step 3 of Figure 5. These times also erogeneous (open). Thus, the unclosed triangles provide estimates for the initiation age of the Longmu Co fault. We calculated the initiation derived for Tianshuihai-Tibet and Tibet–NW age of the Longmu Co fault by retrodeforming the present plate geometry using the rates Himalaya motion using Quaternary slip rates in and azimuths presented in Table 1 and the velocity-triangle analysis shown in Figure 5 until steps 2 and 3 of Figure 5 result from the large the Karakoram fault was straight along its entire length. Sources for the mean extension discrepancies in reported Quaternary slip rates directions are the same as those in Figure 1.

Geosphere, April 2007 79

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/3/2/71/893719/i1553-040X-3-2-71.pdf by guest on 27 September 2021 Raterman et al. 0 N 39° 33° 2 81° 81° Tikilik Fault Co Fault 4 Hotan Fault Gozha–Longmu Gozha–Longmu 6 Transpression and Transpression along the transtension northern and southern Fault, Karakoram respectively C

Karakax Fault Karakax Kumtagh Fault Kumtagh 8 ? Time (Ma.) Time C between C between

7–5 Ma. Onset of 10 Main Pamir Thrust Main Pamir cooling along renewed rapid rapid renewed the Karakoram the Karakoram to 0° fault from 160° from fault km Transtensional le and Tirrul (1991), Dunlap et al. (1998), Tirrul le and S Karakoram Fault Fault S Karakoram 12 f major faults (black lines) at the indicated f major 100 75° System Late Miocene - Present nal panel shows the location of political borders 14 Transpressional oling history that is shown by the map above. See 39° 33° N Karakoram Fault Fault N Karakoram Kongur Extensional Kongur 0 ca. 7 Ma - 0 ca. C

600 200 400 Temperature (°C) Temperature 0 . N ? 2 C at 11.3 Ma C at 11.3

4 Tikilik Fault Karakax Fault Karakax the Karakoram fault fault the Karakoram Cooling of rocks along Cooling of rocks below 350° below 6 8 Time (Ma.) Time ? Kumtagh Fault Kumtagh 10 Pure strike-slip Pure strike-slip along the entire Karakoram Fault km 12 Pure strike-slip Karakoram Fault Fault Karakoram 100 Main Pamir Thrust Main Pamir Middle - Late Miocene 14 0 ca. 11 Ma - 7 ca. B

600 200 400 Temperature (°C) Temperature 0 N ? 2 4 Tikilik Fault Karakax Fault Karakax 6 8 Time (Ma.) Time ? 10 at ca.15 Ma. Formation of the Formation Karakoram Fault Fault Karakoram km Karakoram Fault Fault Karakoram Transpressional 12 Main Pamir Thrust Main Pamir 100 Middle Miocene

14 0 Transpression Transpression along the entire Karakoram Fault A 15 Ma - 11 ca. Fault Thermal along the Karakoram history of units exposed

600 200 400 Temperature (°C) Temperature text for further explanation. further text for Figure 7. Middle Miocene to Holocene reconstruction of the western India-Asia collision zone. Maps show the proposed position o of the western India-Asia collision zone. Maps show proposed 7. Middle Miocene to Holocene reconstruction Figure emphasizes the portion of co The gray shaded region et al. (1994). Searle et al. (1998), Phillips (2004), and Foster time period with the exception of the last panel, which shows the present position of the faults. Gray dotted-dashed line in ﬁ time period with the exception of last panel, which shows present Sear along the Karakoram fault as compiled from plots show the cooling history of rocks Temperature-time spatial reference. for

80 Geosphere, April 2007

030.9° ± 32.9° or 8.3 ± 2.7 mm/yr toward mal component of motion along the southern Tarim, Tianshuihai, and Tibet calculated from 302.7° ± 16.1° relative to Tianshuihai, depend- Karakoram fault. Extrapolation of this calcula- them (Fig. 5) yields closed velocity triangles ing on the use of either geodetic or Quaternary tion indicates that motion along the southern within their 2σ uncertainties. The triangles rates, respectively. Because the orientations Karakoram fault would be entirely normal at derived from the Quaternary slip rates for the of the relative motion vectors differ between ~55º of oblique convergence across the northern same faults, however, yielded velocity triangles these two solutions, they make signifi cantly Karakoram fault; however, the velocity triangle that failed to close, suggesting that either the different predictions regarding the kinematics would no longer close. assumption of perfectly rigid blocks is fl awed or of the 070°–090°–striking Gozha–Longmu Co the slip-rate determinations are inaccurate. Nev- fault system. In particular, the motion of Tibet DISCUSSION ertheless, Brown et al. (2005) suggested that the toward 031° relative to Tianshuihai in the geo- Quaternary slip rate determined by Chevalier detic solution predicts that the Gozha–Longmu The new neotectonic mapping presented in et al. (2005) for the Karakoram fault could be Co fault system should have left-reverse the Web-based map and Figure 2 outlines three revised to 4–5 mm/yr, and England and Mol- transpressional kinematics. In contrast, the previously unrecognized observations about the nar (2005) and Cowgill (2007) suggested that motion of Tibet toward 303° relative to Tian- active deformation of western Tibet. First, at its the slip rate along the central Altyn Tagh fault shuihai in the Quaternary solution predicts that southwestern end, the Altyn Tagh fault bifurcates determined by Meriaux et al. (2004) could be the fault should have right-reverse transpres- into two geometrically and kinematically com- revised to 9–10 mm/yr. These modifi cations sional kinematics, which confl icts with the plex fault networks, each of which is ~100 km bring the Quaternary slip rates into accordance observed kinematics reported here. along strike. These fault zones strike roughly with the geodetic ones. These data suggest that The third step yields average NW Himalaya– parallel to the Karakax and Gozha–Longmu Co our assumption of perfectly rigid blocks is a Tibet relative motions (red vectors in Fig. 5) faults (Fig. 2 and Web Map). Second, the south- valid approximation for western Tibet, and we using the individual solutions computed for the western end of the Longmu Co fault coincides proceed with the assumption that the results Tianshuihai-Tibet vector determined as part of with a 27-km-wide bend in the Karakoram fault. obtained from the geodetic slip rates present step 2 (the average of which is shown as the This bend is associated with a distinct change in the most accurate description of late Cenozoic blue vectors in Fig. 5), the individual slip-rate topography along the strike of the Karakoram deformation of this region. values from along the Karakoram fault, and an fault (Fig. 2B). Third, bedrock markers evident To explore the relationship between relative assumption of pure right slip along the northern in Landsat imagery are displaced 25–32 km block motions in western Tibet and the geom- Karakoram fault to constrain the azimuth of the along the Gozha–Longmu Co fault system. etry and kinematics of the Karakoram fault, Tianshuihai–NW Himalaya vector (the average Taken together, these new observations, previ- we developed the reconstruction shown in Fig- Tianshuihai–NW Himalaya vectors are shown ously published work, and the kinematic analy- ure 6, which is in a NW Himalaya–fi xed refer- in black on Fig. 5 and are listed in Table 1). The sis presented herein support four conclusions: ence frame. To generate this reconstruction, we resulting red vector in step 3a of Figure 5 shows (1) The Altyn Tagh fault is geometrically and started with the present fault geometry and then the average predicted motion of Tibet relative to kinematically linked with both the Karakax and retrodeformed the bend in the Karakoram fault the NW Himalaya across the Karakoram fault Gozha–Longmu Co faults, although such link- by sliding the Tibet and Tianshuihai blocks back south of its junction with the Longmu Co fault. ages are both complex and poorly expressed along their respective velocity vectors as derived Relative to the NW Himalaya, the geodetic and geomorphically. Peltzer et al. (1989) and the using the geodetic triangle shown in step 3 of Quaternary models predict that Tibet moves at Chinese State Bureau of Seismology (1992) Figure 5. Implicit in this approach is an assump- 6.3 ± 0.1 mm/yr toward 096.7° ± 18.4° or 13.3 both reached a similar conclusion on the basis of tion that the plate kinematics remained constant ± 1.9 mm/yr directed toward 118.6° ± 11.6°, their own work and the absence of other suitable during formation of the bend. respectively. structures in the region. (2) The lack of geomor- This reconstruction matches four fi rst-order Although both previous work and our obser- phic evidence for active deformation within the characteristics of the neotectonic map (Fig. 2), vations attest to transpressional deformation interior of the Tianshuihai terrane, and the con- earthquake focal mechanisms, and previ- along the northern Karakoram fault, the amount centration of active faulting into narrow zones ous work (Fig. 1): (1) The Karakoram fault is of fault-perpendicular shortening is uncon- along its margins with Tarim, the NW Himalaya, transtensional south of its intersection with the strained. In an attempt to evaluate the effect and Tibet, suggest that the Tianshuihai region Gozha–Longmu Co fault system; however, the of this transpression on NW Himalaya–Tibet has behaved essentially as a rigid block during magnitude of transtension could vary depend- relative motion, we also considered an alter- the Quaternary. (3) The position of the 27-km- ing upon the amount of transpression across the native solution for the third step in which we wide bend in the Karakoram fault at its inter- northern Karakoram fault. (2) The intersection arbitrarily assumed 10º of oblique convergence section with the Longmu Co fault suggests that of the Karakoram and Longmu Co faults coin- across the northern Karakoram fault. This solu- these two features are genetically related. (4) The cides with a restraining double bend. (3) The tion is indicated by step 3b in Figure 5. Incor- Karakoram fault is transpressional to the north of Gozha–Longmu Co fault system is transpres- poration of 10° of oblique convergence across this bend but transtensional to the south. sional. (4) Faults linking the Gozha Co and the northern Karakoram fault has no signifi cant As mentioned already, the discrepancy Altyn Tagh faults are slightly extensional. effect on the NW Himalaya–Tianshuihai rela- between geodetic and Quaternary slip-rate Previous geologic work along the Karakoram tive velocity vector in the Quaternary velocity determinations for the Karakax, Altyn Tagh, fault is consistent with our kinematic analysis triangle. In contrast, addition of a component of and Karakoram faults resulted in a signifi cant of the Karakoram–Gozha–Longmu Co triple oblique convergence along the northern Kara- difference in the calculated velocity triangles. junction (Fig. 5). In comparison to the predicted koram fault to the geodetic velocity triangle The combination of average geodetic slip rates east-west extension between NW Himalaya and predicts a more easterly motion of Tibet relative for the Western Kunlun Shan, Karakax, Altyn Tibet, measurements of tension gashes and nor- to the NW Himalaya at 6.3 ± 0.1 mm/yr toward Tagh, and Karakoram faults and the averages of mal faults along the southern Karakoram fault 088.8° ± 23.7°, thereby increasing the fault-nor- all possible solutions for the relative motions of and along extensional jogs of the Shiquanhe

Geosphere, April 2007 81

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/3/2/71/893719/i1553-040X-3-2-71.pdf by guest on 27 September 2021 Raterman et al.

fault (Fig. 1) indicate a mean extension direc- that are not well expressed in the geomorphol- fault began at ca. 13 Ma by assuming that the tion of 85° ± 28° (Ratschbacher et al., 1994). ogy or GPS data but that have moved faster in average geodetic slip rate of 6 mm/yr (Shen et Similarly, Murphy et al. (2000) and Murphy and the past and/or integrate to regionally signifi cant al., 2001; Wright et al., 2004) has been constant Burgess (2006) found a mean extension direc- deformation rates, then our analysis could be in throughout the development of the ~80-km- tion of 089° along the southernmost portion of error. Nevertheless, the concurrence between long left-lateral defl ection of the Karakax river the Karakoram fault. We also suggest that the the geodetic and geologic slip rates along the where it crosses the Karakax fault (Ding et al., units in the Pangong Range, which lies along Altyn Tagh and Karakoram faults, as discussed 2004). Arnaud et al. (2003) presented additional the Karakoram fault between 34.5°N and 33°N in the previous section, supports our assumption 40Ar/39Ar potassium feldspar thermochronologic (Fig. 1), may have been partially exhumed (Rol- that the geodetic slip rates are grossly similar to data to support activity along the Karakax fault land and Pecher, 2001) due to transtension along the long-term geologic rates in this area. during this time. the northernmost segment of the southern Kara- In accordance with the fi ndings of Dunlap et Figure 7B shows the late Miocene confi gu- koram fault. Strands of southern Karakoram al. (1998), we propose that the Karakoram fault ration of the western India-Asia collision zone. fault that truncate glaciers have down-to-the- has experienced three main phases of deforma- Potassium feldspar 40Ar/39Ar thermochronologic east kinematics, consistent with transtension tion, which we argue were genetically linked data indicate very slow cooling during this time along this mountain front (Searle et al., 1998). with the structural evolution of the southwestern period (Dunlap et al., 1998), and 40Ar/39Ar mus- Our kinematic analysis, and in particular the termination of the Altyn Tagh fault and other covite cooling ages of ca. 11 Ma on either side reconstruction shown in Figure 6, can also be major structures in the western India-Asia colli- of the Karakoram fault indicate that both sides of used to predict the initiation age, total slip, and sion zone. Nevertheless, we must emphasize that the fault were at equivalent structural levels at this slip rate along the Longmu Co fault. To do this, the schematic reconstruction shown in Figure 7 time, suggesting pure strike-slip motion (Fig. 7B, we assume that the bend in the Karakoram fault is largely based on temporal associations from plot) (Searle and Tirrul, 1991; Searle et al., 1998). and slip along the Gozha–Longmu Co fault are isolated regional studies along the major struc- Currently, no evidence exits to support a similar genetically related and that the present kinemat- tures in the region; thus, the extent to which these cessation of vertical exhumation along the other ics of the triple junction remained steady during structures are kinematically linked in a regional major structures in the western India-Asia colli- bend formation. As Figures 2 and 6 indicate, the deformation zone, and therefore have a cause sion zone, and the cause for this transition along bend in the Karakoram fault at 34.5°N is ~27 km and effect relationship, remains undetermined. the Karakoram fault remains to be determined. wide when measured perpendicular to the strike Figure 7A shows the middle Miocene con- Finally, Figure 7C shows the proposed late of the fault outside the bend region. The size of fi guration of the western India-Asia collision Miocene–Pliocene to Holocene confi guration this bend is in striking agreement with the esti- zone. Initial slip along the Karakoram fault was of the western India-Asia collision zone. Apa- mates of 25–32 km for total left separation along likely transpressional on the basis of 40Ar/39Ar tite fi ssion-track cooling ages from the northern the Gozha–Longmu Co fault system presented cooling ages from amphibole, muscovite, bio- Karakoram fault indicate that vertical exhuma- in Figure 4. Depending on whether we use the tite, and potassium feldspar from the Baltoro tion began at ca. 5 Ma along the northern Kara- geodetic or Quaternary velocity triangles, the granite and the Pangong gneisses, which indi- koram fault (Foster et al., 1994). The 40Ar/39Ar kinematic model in Figure 4 predicts that the cate rapid cooling from 17 to 13 Ma related to potassium feldspar thermochronology (Dunlap bend has widened at either 2.8 ± 1.1 mm/yr or vertical exhumation along the Karakoram fault et al., 1998), 40Ar/39Ar cooling ages (Rolland 8.3 ± 2.7 mm/yr, respectively. Extrapolation of (Fig. 7A) (Dunlap et al., 1998; Searle and Tirrul, and Pecher, 2001), and apatite fi ssion-track these rates backward from the present-day to 1991; Searle et al., 1998). Furthermore, initial cooling ages (Arnaud, 1992) from the southern smooth the entire 27 km bend yields an age for slip along the Karakoram fault likely began in Karakoram fault suggest that renewed vertical the onset of bending of either 9.6 ± 2.8 Ma, if the vicinity of the Baltoro granite, the unit with a exhumation began somewhat earlier along this geodetic rates are used, or 3.2 ± 2.5 Ma, using maximum undisputed displacement of ~150 km reach of the fault at ca. 8 Ma (Fig. 7C, plot). Quaternary rates. Importantly, this analysis does along the Karakoram fault (Phillips et al., 2004; We propose that this renewed vertical exhuma- not specify which fault deforms the other: left Searle, 1996; Searle et al., 1998). U-Pb zircon tion is related to the ca. 9 Ma initiation of the slip along the Gozha–Longmu Co fault system ages from leucocratic dikes in the Pangong Gozha–Longmu Co fault system as indicated can be viewed as causing deformation of the Range, which are the youngest rocks known to by our kinematic analysis. Therefore, vertical Karakoram fault as a passive marker where the be cut by the fault (Phillips et al., 2004), con- exhumation along the northern Karakoram fault two faults intersect, or the formation of a bend in strain the initiation age of the Karakoram fault relates to transpressional motion, whereas verti- the Karakoram fault could have triggered devel- to 15.68 ± 0.53 to 13.73 ± 0.28 Ma. It could be cal exhumation along the southern Karakoram opment of the Gozha–Longmu Co fault system. as old as 30 Ma, however, depending on whether fault relates to transtensional motion. This Based on these fi ndings, Figure 7 outlines U-Pb zircon ages from the same region refl ect sequence differs from the scenario outlined by a new synthesis of the Miocene to Holocene synkinematic (Lacassin et al., 2004a, 2004b) or Searle et al. (1998) and Dunlap et al. (1998), evolution of the major structures of the western prekinematic (Searle and Phillips, 2004) mag- who suggested that renewed vertical exhuma- India-Asia collision zone and their relation to matism. Apatite fi ssion-track cooling ages from tion was related to transpressional deformation the deformational and thermal history of rocks the eastern Pamirs and Western Kunlun Shan along the entire length of the Karakoram fault. exposed along the Karakoram fault. This syn- indicate that exhumation began during this time This change in kinematics along the Karakoram thesis extrapolates the results of the geodetic (Sobel and Dumitru, 1997) and suggest that the fault and initiation of the Gozha–Longmu Co velocity triangles from the previous section over Main Pamir and Kumtagh fault system initiated fault system was also contemporaneous with a the past 20 m.y.; therefore, the proceeding con- at this time. The Main Pamir–Kumtagh fault change in the sedimentation rate in the Tarim clusions are speculative since they assume that system may have linked with the Tikilik fault Basin to the north of the Western Kunlun Shan the geodetic slip rates are similar to the long- to the north of the Western Kunlun Shan (Cow- at ca. 5 Ma (Zheng et al., 2000). We infer that term geologic rates. In particular, if there are gill, 2001; Sobel and Dumitru, 1997). Finally, this change was related to initiation of slip along numerous, slowly slipping faults within the area we infer that Cenozoic slip along the Karakax the blind Hotan thrust and cessation of slip along

82 Geosphere, April 2007

the Tikilik fault, which may have resulted from former. The intersection of the Karakoram and Eldridge Moores for helpful comments that improved the initial drafts. Rebecca Bendick, Jim Spotila, and reorganization of the southwestern Altyn Tagh Gozha–Longmu Co faults shares characteristics an anonymous reviewer provided helpful comments fault. The initiation of normal faulting along the of intersecting strike-slip faults in both classes. on an earlier version of this work. We would like to Kongur Shan extensional system at ca. 8 Ma Like the San Andreas–Garlock intersection, the thank Jim Spotila and Clark Burchfi el for their com- (Robinson et al., 2004) also coincided with the surface trace of the through-going Karakoram ments on the present version of this work. Finally, we would also like to thank Nick Kent-Basham for help inferred initiation age of the Gozha–Longmu fault clearly truncates the Gozha–Longmu Co during the fi eld work. Co fault system and the onset of renewed cool- fault system, similar to the intersection of the ing along the Karakoram fault. Furthermore, San Andreas and Garlock faults. Yet, in this REFERENCES CITED the late Miocene reorganization of the western case, slip on the Gozha–Longmu Co and bend- India-Asia collision zone outlined in Figure 7C ing of the Karakoram fault appear to be geneti- Ando, R., Tada, T., and Yamashita, T., 2004, Dynamic evolution of a fault system through interactions between occurred at approximately the same time as the cally related to one another in a way that is fault segments: Journal of Geophysical Research–Solid onset or rejuvenation of normal faulting across similar to the intersection of the Red River and Earth and Planets, v. 109, p. B05303. the entire Tibetan Plateau (Pan et al., 1993; Yin, Xianshuihe–Xiaojiang faults. The reasons for Armijo, R., Tapponnier, P., Mercier, J.L., and Tong-Lin, H., 1986, Quaternary extension in southern Tibet: Field 2000; Yin et al., 1999; Dunlap et al. 1998). such behavior remain to be determined, but it observations and tectonic implications: Journal of Geo- In addition to the Karakoram–Longmu Co seems likely that an important factor in explain- physical Research, v. 91, p. 13,803–13,872. Armijo, R., Tapponnier, P., and Tonglin, H., 1989, Late fault junction, a number of other intersecting ing such differences in deformational behavior Cenozoic right-lateral strike-slip faulting in southern strike-slip faults are known. Examples of simul- at fault intersections stems from variations in Tibet: Journal of Geophysical Research–Solid Earth taneously active, major (along-strike length crustal structure related to variations in both and Planets, v. 94, p. 2787–2838. Arnaud, N., 1992, Apports de la thermochronologie 40Ar/ 39Ar >50 km), strike-slip faults that intersect include crustal thickness and thermal gradient. In par- sur feldspath potassique à la connaissance de la tec- the San Andreas and Garlock faults (Bohan- ticular, intersecting faults in which slip along tonique Cenozoique d’Asie [Ph.D. thèse]: Clermont- non and Howell, 1982; Spotila and Anderson, one fault deforms the surface trace of the other, Ferrand, France, Université Blaise Pascal, 300 p. Arnaud, N., Tapponnier, P., Roger, F., Brunel, M., Schärer, 2004), the Blackwater and Garlock faults in such as the Blackwater-Garlock, Red River– U., Chen, W., and Xu, Z., 2003, Evidence for Meso- the Eastern California shear zone (Peltzer et al., Xianshuihe, and Karakoram–Longmu Co inter- zoic shear along the western Kunlun and Altyn-Tagh fault, northern Tibet (China): Journal of Geophysical 2001), the Red River and Xianshuihe–Xiaojiang sections, all occur in locations with thick crust Research, v. 108, p. 2053, doi: 10.1029/2001JB000904. faults in eastern Tibet (Wang et al., 1998), and and/or anomalously high thermal gradients and Avouac, J.P., and Peltzer, G., 1993, Active tectonics in the North and East Anatolian faults in eastern thus may be underlain by weak, subhorizontal southern Xinjiang, China: Analysis of terrace riser and normal fault scarp degradation along the Hotan-Qira Turkey (S¸engör et al., 1985). Although more detachment horizons within the middle crust fault system: Journal of Geophysical Research, v. 98, work on these systems is needed to advance (e.g., Burchfi el et al., 1989). p. 21,773–21,807. our understanding of the crustal-scale geom- Avouac, J.P., and Tapponnier, P., 1993, Kinematic model of active deformation in central Asia: Geophysical etry of these intersections and their kinematic CONCLUSIONS Research Letters, v. 20, p. 895–898. evolution, current work suggests that intersect- Banerjeee, P., and Bürgmann, R., 2002, Convergence across the northwestern Himalaya from GPS measurements: ing strike-slip faults fall into two broad classes: Our results have several broad implications for Geophysical Research Letters, v. 29, no. 13, 1652, doi: those with surface traces that intersect, in which studies of large intercontinental strike-slip fault 10.1029/2002GL015184. case one fault truncates the other, and those for zones and how they interact on time scales less Bendick, R., Bilham, R., Freymueller, J., Larson, K., and Yin, G., 2000, Geodetic evidence for a low slip rate in which the surface traces cross but in detail fail to than 1 million years. First, our results highlight the Altyn Tagh fault system: Nature, v. 404, p. 69–72, intersect. Although the surface trace of the San the importance of accounting for potential strain doi: 10.1038/35003555. Andreas fault clearly intersects and truncates infl uences of intersecting faults when studying Bird, P., and Piper, K., 1980, Plane-stress fi nite-element models of tectonic fl ow in southern California: Physics the Garlock fault, and Bohannon and Howell along-strike variability of deformation along of the Earth and Planetary Interiors, v. 21, p. 158–175, (1982) proposed that slip along the Garlock major fault zones. The Longmu Co fault appears doi: 10.1016/0031-9201(80)90067-9. Bohannon, R., and Howell, D., 1982, Kinematic evolution of fault has played a role in altering the geometry to exert a fi rst-order control on the structural style the junction of the San Andreas, Garlock, and Big Pine of the San Andreas fault, there seems to be mini- along the Karakoram fault, which may have sig- faults: California Geology, v. 10, p. 358–363. mal evidence in support of this model (Spotila nifi cant implications for along-strike differences Brown, E.T., Bendick, R., Bourlès, D.L., Gaur, V., Molnar, P., Raisbeck, G.M., and Yiou, F., 2002, Slip rates of the and Anderson, 2004). In contrast, the surface in determinations of the slip rate and total magni- Karakorum fault, Ladakh, India, determined using cos- traces of the Blackwater and Garlock faults tude of displacement along the Karakoram fault. mic ray exposure dating of debris fl ows and moraines: do not appear to directly intersect one another, Second, our results suggest that local changes Journal of Geophysical Research, v. 107, p. 2192– 2205, doi: 10.1029/2000JB000100. although the traces of both faults continue on in the geometry and kinematics of a fault sys- Brown, E.T., Molnar, P., and Bourles, D.L., 2005, Comment either side of the inferred zone of intersec- tem could indicate important changes occurring on “Slip-rate measurements on the Karakorum fault may imply secular variations in fault motion”: Science, tion, so the faults do, at some level, appear to throughout an entire orogen. This appears to be v. 309, no. 5739, p. 1326. be mutually crosscutting. InSAR data indicate true to the fi rst order for the late Miocene of the Burchfi el, B.C., Deng, Q., Molnar, P., Royden, L., Wang, Y., that slip on one fault directly impacts slip on the western India-Asia collision zone. Finally, the Zhang, P., and Zhang, W., 1989, Intracrustal detachment zones of continental deformation: Geology, v. 17, other (Peltzer et al., 2001). The surface traces results demonstrate the applicability of a rigid- p. 748–752, doi: 10.1130/0091-7613(1989)017<0448: of the Red River and Xianshuihe–Xiao jiang block conceptualization of continental crust and IDWZOC>2.3.CO;2. faults likewise do not directly intersect one kinematic analyses in obtaining a fi rst-order Chen, Z., Burchfi el, B.C., Liu, Y., King, R.W., Royden, L.H., Tang, W., Wang, E., Zhao, J., and Zhang, X., 2000, another, although, again, these traces do cross understanding of continental deformation. Global Positioning System measurements from eastern and continue beyond the zone of fault intersec- Tibet and their implications for India/Eurasia intercon- ACKNOWLEDGMENTS tinental deformation: Journal of Geophysical Research, tion. In this instance, the surface trace of the Volume 105: Issue, v. B7, p. 16215–16228. Red River fault is defl ected by an amount that Chevalier, M.L., Ryerson, F.J., Tapponnier, P., Finkel, R.C., This work was supported by National Science is equivalent to the total slip on the Xianshuihe– Van der Woerd, J., Li, H.B., and Liu, Q., 2005, Slip- Foundation grant EAR-0310415 and the University rate measurements on the Karakorum fault may imply Xiaojiang fault (Wang et al., 1998), suggest- of California–Davis Durrell Fund. We would like to secular variations in fault motion: Science, v. 307, ing that slip along the latter has deformed the thank Ryan Gold, Mike Murphy, Alex Robinson, and p. 411–414, doi: 10.1126/science.1105466.

Geosphere, April 2007 83

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/3/2/71/893719/i1553-040X-3-2-71.pdf by guest on 27 September 2021 Raterman et al.

Chinese State Bureau of Seismology, 1992, The Altyn Tagh Morphochronologic evidence from Cherchen He and fault system, northwest China: Geological Society of active fault system (in Chinese with English abstract): Sulamu Tagh: Journal of Geophysical Research–Solid America Bulletin, v. 112, p. 61–74, doi: 10.1130/0016- Beijing, China, Seismology Publishing House, 319 p. Earth and Planets, v. 109, B06401. 7606(2000)112<0061:MOPMJB>2.3.CO;2. Cowgill, E., 2001, Tectonic Evolution of the Altyn Tagh– Meriaux, A.S., Tapponnier, P., Ryerson, F.J., Xu, X.W., King, Ritts, B.D., Yue, Y.J., and Graham, S.A., 2004, Oligocene- Western Kunlun Fault System, Northwestern China G., Van der Woerd, J., Finkel, R.C., Li, H.B., Caffee, Miocene tectonics and sedimentation along the Altyn [Ph.D. thesis]: Los Angeles, California, University of M.W., Xu, Z.Q., and Chen, W.B., 2005, The Aksay Tagh fault, northern Tibetan Plateau: Analysis of the California, Los Angeles, 311 p. segment of the northern Altyn Tagh fault: Tectonic Xorkol, Subei, and Aksay basins: The Journal of Geol- Cowgill, E., 2007, Riser reconstructions and their impact geomorphology, landscape evolution, and Holocene ogy, v. 112, p. 207–229, doi: 10.1086/381658. on secular variation in rates of strike-slip faulting: slip rate: Journal of Geophysical Research-Solid Earth, Robinson, A.C., Yin, A., Manning, C.E., Harrison, T.M., Revisiting the Cherchen He site along the Altyn Tagh v. 110, p. B04404, doi: 10.1029/2004JB003210. Zhang, S.H., and Wang, X.F., 2004, Tectonic evolution fault, NW China: Earth and Planetary Science Letters Molnar, P., and Tapponnier, P., 1975, Cenozoic tectonics of of the northeastern Pamir: Constraints from the north- (in press). Asia—Effects of a continental collision: Science, v. 189, ern portion of the Cenozoic Kongur Shan extensional Cowgill, E., Yin, A., Harrison, T.M., and Wang, X.-F., p. 419–426, doi: 10.1126/science.189.4201.419. system, western China: Geological Society of America 2003, Reconstruction of the Altyn Tagh fault based Molnar, P., and Tapponnier, P., 1977, Relation of the tec- Bulletin, v. 116, p. 953–973, doi: 10.1130/B25375.1. on U-Pb ion microprobe geochronology: Role of tonics of eastern China to the India-Eurasia collision; Rolland, Y., and Pecher, A., 2001, The Pangong granulites of back thrusts, mantle sutures, and heterogeneous application of slip-line fi eld theory to large-scale the Karakoram fault (western Tibet): Vertical extrusion crustal strength in forming the Tibetan Plateau: Jour- continental tectonics: Geology, v. 5, p. 212–216, doi: within a lithosphere-scale fault?: Comptes Rendus de nal of Geophysical Research, v. 108, p. 2346, doi: 10.1130/0091-7613(1977)5<212:ROTTOE>2.0.CO;2. l’Academie des Sciences, Serie Ii Fascicule à Sciences 10.1029/2002JB002080. Molnar, P., and Tapponnier, P., 1978, Active tectonics of Tibet: de la Terre et des Planetes, v. 332, p. 363–370. Dickinson, W.R., and Snyder, W.S., 1979, Geometry of triple Journal of Geophysical Research, v. 83, p. 5361. Ryerson, F.J., Peltzer, G., Tapponnier, P., Finkel, R.C., Meri- junctions related to San Andreas transform: Journal of Molnar, P., Burchfi el, B.C., Liang, K.U., and Zhao, Z., 1987, aux, A.-S., Van der Woerd, J., and Caffee, M.W., 1999, Geophysical Research, v. 84, p. 561–572. Geomorphic evidence for active faulting in the Altyn Active slip rates on the Altyn Tagh fault Karakax valley Ding, G.Y., Chen, J., Tian, Q.J., Shen, X.H., Xing, C.Q., and Tagh and northern Tibet and qualitative estimates of its segment: Constraints from surface exposure age dat- Wei, K.B., 2004, Active faults and magnitudes of left- contribution to the convergence of India and Eurasia: ing: Eos (Transactions, American Geophysical Union), lateral displacement along the northern margin of the Geology, v. 15, p. 249–253. v. 80, p. F1008. Tibetan Plateau: Tectonophysics, v. 380, p. 243–260, Murphy, M.A., and Burgess, W.P., 2006, Geometry, kinemat- Searle, M.P., 1996, Geological evidence against large- doi: 10.1016/j.tecto.2003.09.022. ics, and landscape characteristics of an active transten- scale pre-Holocene offsets along the Karakoram Dunlap, W.J., Weinberg, R.F., and Searle, M.P., 1998, Kara- sion zone, Karakoram fault system, southwest Tibet: fault: Implications for the limited extrusion of the korum fault rocks cool in two phases: Journal of the Journal of Structural Geology, v. 28, p. 268–283, doi: Tibetan Plateau: Tectonics, v. 15, p. 171–186, doi: Geological Society of London, v. 155, p. 903–912. 10.1016/j.jsg.2005.10.009. 10.1029/95TC01693. England, P., and McKenzie, D.P., 1982, A thin viscous Murphy, M.A., Yin, A., Kapp, P., Harrison, T.M., Din, L., and Searle, M.P., and Phillips, R.J., 2004, A comment on “Large- sheet model for continental deformation: Geophysi- Guo, J., 2000, Southward propagation of the Karakorum scale geometry, offset, and kinematic evolution of the cal Journal of the Royal Astronomical Society, v. 70, fault system into southwest Tibet: Timing and magnitude Karakoram fault, Tibet” by R. Lacassin et al.: Earth p. 295–321. of slip: Geology, v. 28, p. 451–454, doi: 10.1130/0091- and Planetary Science Letters, v. 229, p. 155–158, doi: England, P., and Molnar, P., 2005, Late Quaternary to 7613(2000)028<0451:SPOTKF>2.3.CO;2. 10.1016/j.epsl.2004.07.044. decadal velocity fi elds in Asia: Journal of Geophysical Murphy, M.A., Yin, A., Kapp, P., Harrison, T.M., Manning, Searle, M.P., and Tirrul, R., 1991, Structural and thermal Research–Solid Earth and Planets, v. 110, p. B12401. C.E., Ryerson, F.J., Lin, D., and Guo, J., 2002, Structural evolution of the Karakorum crust: Journal of the Geo- Foster, D.A., Gleadow, A.J.W., and Mortimer, G., 1994, Rapid evolution of the Gurla Mandhata detachment system, logical Society of London, v. 148, p. 65–82. Pliocene exhumation in the Karakoram (Pakistan), southwest Tibet: Implications for the eastward extent Searle, M.P., Weinberg, R.F., and Dunlap, W.J., 1998, revealed by fi ssion-track thermochronology of the K2 of the Karakoram fault system: Geological Society of Transpressional tectonics along the Karakorum fault gneiss: Geology, v. 22, p. 19–22, doi: 10.1130/0091- America Bulletin, v. 114, p. 428–447, doi: 10.1130/0016- zone, northern Ladakh: Constraints on Tibetan extru- 7613(1994)022<0019:RPEITK>2.3.CO;2. 7606(2002)114<0428:SEOTGM>2.0.CO;2. sion, in Holdsworth, R.E., Strachan, R.A., and Dewey, Jade, S., Bhatt, B.C., Yang, Z., Bendick, R., Gaur, V.K., Mol- Pan, Y., Copeland, P., Roden, M.K., Kidd, W.S.F., and Har- J.F., eds., Continental Transpressional and Transten- nar, P., Anand, M.B., and Kumar, D., 2004, GPS mea- rison, T.M., 1993, Thermal and unroofi ng history of sional Tectonics: Geological Society [London] Special surements from the Ladakh Himalaya, India; prelimi- the Lhasa area, southern Tibet—Evidence from apa- Publication 135, p. 307–326. nary tests of plate-like or continuous deformation in tite fi ssion-track thermochronology: Nuclear Tracks S¸engör, A.M.C., Görür, N., and Saroglu, F., 1985, Strike-slip Tibet: Geological Society of America Bulletin, v. 116, and Radiation Measurements, v. 21, p. 543–554, doi: faulting and related basin formation in zones of tec- p. 1385–1391, doi: 10.1130/B25357.1. 10.1016/1359-0189(93)90195-F. tonic escape: Turkey as a case study, in Biddle, K.T., Jezek, J., Schulmann, K., and Thompson, A., 2002, Strain Peltzer, G., and Saucier, F., 1996, Present-day kinematics and Christie-Blick, N., eds., Strike-Slip Deformation, partitioning in front of an obliquely convergent of Asia derived from geologic fault rates: Journal of Basin Formation, and Sedimentation: Tulsa, Okla- indenter: European Geophysical Union Stephan Muel- Geophysical Research, v. 101, p. 27,943–27,956, doi: homa, Society of Economic Paleontologists and Min- ler Special Publication Series, v. 1, p. 93–104. 10.1029/96JB02698. eralogists Special Publication 37, p. 227–264. King, G.C.P., and Cocco, M., 2001, Fault interaction by Peltzer, G., and Tapponnier, P., 1988, Formation and evo- Shen, Z.-K., Wang, M., Li, Y., Jackson, D.D., Yin, A., Dong, elastic stress changes: New clues from earthquake lution of strike-slip faults, rifts, and basins during the D., and Fang, P., 2001, Crustal deformation along the sequences: Advances in Geophysics, v. 44, p. 1–38. India-Asia collision; an experimental approach: Jour- Altyn Tagh fault system, western China, from GPS: Kleinrock, M.C., and Morgan, J.P., 1988, Triple junction nal of Geophysical Research–Solid Earth and Planets, Journal of Geophysical Research, v. 106, p. 30,607– reorganization: Journal of Geophysical Research–Solid v. 93, p. 15,085–15,117. 30,621, doi: 10.1029/2001JB000349. Earth and Planets, v. 93, p. 2981–2996. Peltzer, G., Tapponnier, P., and Armijo, R., 1989, Magni- Snay, R., Cline, M., Philip, C., Jackson, J., Feng, Y., Shen, Lacassin, R., Valli, F., Arnaud, N., Leloup, P.H., Paquette, tude of late Quaternary left-lateral movement along Z., and Lisowski, M., 1996, Crustal velocity fi eld near J.L., Haibing, L., Tapponnier, P., Chevalier, M.L., the north edge of Tibet: Science, v. 246, p. 1285–1289, the Big Bend of California’s San Andreas fault: Journal Guillot, S., Maheo, G., and Xu, Z.Q., 2004a, Large- doi: 10.1126/science.246.4935.1285. of Geophysical Research, v. 101, p. 3173–3186, doi: scale geometry, offset and kinematic evolution of the Peltzer, G., Crampe, F., Hensley, S., and Rosen, P., 2001, 10.1029/95JB02394. Karakorum fault, Tibet: Earth and Planetary Science Transient strain accumulation and fault interaction Sobel, E.R., and Dumitru, T.A., 1997, Thrusting and exhuma- Letters, v. 219, p. 255–269, doi: 10.1016/S0012- in the Eastern California shear zone: Geology, v. 29, tion around the margins of the western Tarim Basin during 821X(04)00006-8. p. 975–978, doi: 10.1130/0091-7613(2001)029<0975: the India-Asia collision: Journal of Geophysical Research, Lacassin, R., Valli, F., Arnaud, N., Leloup, P.H., Paquette, TSAAFI>2.0.CO;2. v. 102, p. 5043–5063, doi: 10.1029/96JB03267. J.L., Li, H.B., Tapponnier, P., Chevalier, M.L., Guillot, Phillips, R.J., Parrish, R.R., and Searle, M.P., 2004, Age Spotila, J.A., and Anderson, K.B., 2004, Fault interaction S., Maheo, G., and Xu, Z.Q., 2004b, Reply to com- constraints on ductile deformation and long-term slip at the junction of the Transverse Ranges and Eastern ment on “Large-scale geometry, offset and kinematic rates along the Karakoram fault zone, Ladakh: Earth California shear zone: A case study of intersecting evolution of the Karakorum fault, Tibet”: Earth and and Planetary Science Letters, v. 226, p. 305–319, doi: faults: Tectonophysics, v. 379, p. 43–60, doi: 10.1016/ Planetary Science Letters, v. 229, p. 159–163, doi: 10.1016/j.epsl.2004.07.037. j.tecto.2003.09.016. 10.1016/j.epsl.2004.07.045. Ratschbacher, L., Frisch, W., and Liu, G., 1994, Distrib- Tapponnier, P., and Molnar, P., 1976, Slip-line fi eld theory Liu, Q., 1993, Paléoclimat et contraintes chronologiques sur uted deformation in southern and western Tibet dur- and large-scale continental tectonics: Nature, v. 264, les mouvements récent dans l’Ouest du Tibet: Failles ing and after the India-Asia collision: Journal of p. 319–324, doi: 10.1038/264319a0. du Karakorum et de Longmu Co–Gozha Co, lacs en Geophysical Research, v. 99, p. 19,917–19,945, doi: Tapponnier, P., and Molnar, P., 1977, Active Faulting and pull-apart de Longmu Co et de Sumxi Co [Thèse de 10.1029/94JB00932. Tectonics in China: Journal of Geophysical Research, Doctorat]: Paris, Université Paris VII, 360 p. Replumaz, A., and Tapponnier, P., 2003, Reconstruction of v. 82, no. B20, p. 2905–2930. McKenzie, D.P., and Morgan, W.J., 1969, Evolution of the deformed collision zone between India and Asia Tapponnier, P., Mercier, J.L., Armijo, R., Han, T., and Zhou, J., triple junctions: Nature, v. 224, p. 125–133, doi: by backward motion of lithospheric blocks: Journal of 1981, Field evidence for active normal faulting in Tibet: 10.1038/224125a0. Geophysical Research–Solid Earth and Planets, v. 108, Nature, v. 294, p. 410–414, doi: 10.1038/294410a0. Meriaux, A.S., Ryerson, F.J., Tapponnier, P., Van der Woerd, p. ETG1.1–ETG1.24. Tapponnier, P., Peltzer, G., Le Dain, A.Y., and Armijo, J., Finkel, R.C., Xu, X.W., Xu, Z.Q., and Caffee, M.W., Ritts, B.D., and Biffi , U., 2000, Magnitude of post–Middle R., 1982, Propagating extrusion tectonics in Asia: 2004, Rapid slip along the central Altyn Tagh fault: Jurassic (Bajocian) displacement on the Altyn Tagh New insights from simple experiments with plasti-

84 Geosphere, April 2007

cine: Geology, v. 10, p. 611–616, doi: 10.1130/0091- Altyn Tagh fault: Science, v. 282, p. 74–76, doi: Yue, Y.J., Ritts, B.D., Graham, S.A., Wooden, J.L., Gehrels, 7613(1982)10<611:PETIAN>2.0.CO;2. 10.1126/science.282.5386.74. G.E., and Zhang, Z.C., 2004a, Slowing extrusion tecton- Van der Woerd, J., Tapponnier, P., Meyer, B., Xu Xiwei, Wright, T.J., Parsons, B., England, P.C., and Fielding, E.J., ics: Lowered estimate of post–early Miocene slip rate for Ryerson, F.J., and Meriaux, A.-S., 1999, The Altyn 2004, InSAR observations of low slip rates on the the Altyn Tagh fault: Earth and Planetary Science Letters, Tagh–Tanghenan Shan thrust triple junction: Kine- major faults of western Tibet: Science, v. 305, p. 236– v. 217, p. 111–122, doi: 10.1016/S0012-821X(03)00544-2. matic constraints and implications for the growth of the 239, doi: 10.1126/science.1096388. Yue, Y.J., Ritts, B.D., Hanson, A.D., and Graham, S.A., Tibetan Plateau, in Sobel, E., Appel, E., Strecker, M., Xiong, X., Park, P.-H., Zheng, Y., Hsu, H., and Han, U., 2003, 2004b, Sedimentary evidence against large strike-slip Ratschbacher, L., and Blisniuk, P., eds., 14th Annual Present-day slip-rate of Altyn Tagh Fault: Numerical translation on the Northern Altyn Tagh fault, NW Himalaya-Karakorum-Tibet Workshop: Kloster Ettal, result constrained by GPS data: Earth: Planets and China: Earth and Planetary Science Letters, v. 228, Germany, p. 161. Space, v. 55, p. 509–514. p. 311–323, doi: 10.1016/j.epsl.2004.10.008. Vilotte, J.P., Daignieres, M., and Madariaga, R., 1982, Xu, X.W., Tapponnier, P., Van Der Woerd, J., Ryerson, F.J., Yue, Y.J., Graham, S.A., Ritts, B.D., and Wooden, J.L., 2005, Numerical modeling of intraplate deformation: Simple Wang, F., Zheng, R.Z., Chen, W.B., Ma, W.T., Yu, Detrital zircon provenance evidence for large-scale mechanical models of continental collision: Journal of G.H., Chen, G.H., and Meriaux, A.S., 2005, Late Qua- extrusion along the Altyn Tagh fault: Tectonophysics, Geophysical Research–Solid Earth and Planets, v. 87, ternary sinistral slip rate along the Altyn Tagh fault and v. 406, p. 165–178, doi: 10.1016/j.tecto.2005.05.023. p. 10,709–10,728. its structural transformation model: Science in China, Zhang, P.-Z., Shen, Z.-k., Wang, M., Gan, W.-j., Bürgmann, Wallace, K., Yin, G., Bilham, R., 2004, Inescapable slow slip ser. D, Earth Sciences, v. 48, p. 384–397. R., Molnar, P., Wang, Q., Niu, Z.-j., Sun, J.-z., Wu, J.- on the Altyn Tagh fault, Geophysical Research Letters, Yin, A., 2000, Mode of Cenozoic east-west extension in c., Sun Hanrong, and You Xinzhao, 2004, Continuous v. 31, p. L09613.1–L09613.4. Tibet suggesting a common origin of rifts in Asia dur- deformation of the Tibetan Plateau from global posi- Wang, E., Burchﬁ el, B.C., Royden, L.H., Liangzhong, C., ing the Indo-Asian collision: Journal of Geophysical tioning system data: Geology, v. 32, p. 809–812, doi: Jishen, C., Wenxin, L., and Zhiliang, C., 1998, Late Research–Solid Earth and Planets, v. 105, p. 21,745– 10.1130/G20554.1. Cenozoic Xianshuihe–Xiaojiang, Red River, and Dali 21,759, doi: 10.1029/2000JB900168. Zhao, J.M., Mooney, W.D., Zhang, X.K., Li, Z.C., Jin, Z.J., Fault: Systems of Southwestern Sichuan and Central Yin, A., Kapp, P.A., Murphy, M.A., Manning, C.E., Har- and Okaya, N., 2006, Crustal structure across the Altyn Yunnan, China: Geological Society of America Special rison, T.M., Grove, M., Ding, L., Deng, X.G., and Tagh Range at the northern margin of the Tibetan Pla- Paper 327, 108 p. Wu, C.M., 1999, Signiﬁ cant late Neogene east- teau and tectonic implications: Earth and Planetary Wang, Y.X., Mooney, W.D., Yuan, X.C., and Coleman, R.G., west extension in northern Tibet: Geology, v. 27, Science Letters, v. 241, p. 804–814, doi: 10.1016/ 2003, The crustal structure from the Altai Mountains p. 787–790, doi: 10.1130/0091-7613(1999)027<0787: j.epsl.2005.11.003. to the Altyn Tagh fault, northwest China: Journal of SLNEWE>2.3.CO;2. Zheng, H., Powell, C.M., An, Z., Zhou, J., and Dong, G., Geophysical Research–Solid Earth and Planets, v. 108, Yin, A., Rumelhart, P.E., Butler, R., Cowgill, E., Harrison, T.M., 2000, Pliocene uplift of the northern Tibetan Pla- p. ESE 7-1. Foster, D.A., Ingersoll, R.V., Zhang, Q., Zhou, X., Wang, teau: Geology, v. 28, p. 715–718, doi: 10.1130/0091- Washburn, Z., Arrowsmith, J.R., Dupont-Nivet, G., Feng, X., Hanson, A., and Raza, A., 2002, Tectonic history of the 7613(2000)028<0715:PUOTNT>2.3.CO;2. W.X., Qiao, Z.Y., and Chen, Z.L., 2003, Paleoseismol- Altyn Tagh fault system in northern Tibet inferred from ogy of the Xorxol segment of the central Altyn Tagh Cenozoic sedimentation: Geological Society of America fault, Xinjiang, China: Annales de Geophysique, v. 46, Bulletin, v. 114, p. 1257–1295, doi: 10.1130/0016- p. 1015–1034. 7606(2002)114<1257:THOTAT>2.0.CO;2. Wittlinger, G., Tapponnier, P., Poupinet, G., Jiang, M., Shi, Yue, Y., Ritts, B.D., and Graham, S.A., 2001, Initiation and MANUSCRIPT RECEIVED 6 SEPTEMBER 2006 D., Herquel, G., and Masson, F., 1998, Tomographic long-term slip history of the Altyn Tagh fault: Interna- REVISED MANUSCRIPT RECEIVED 8 DECEMBER 2006 evidence for localized lithospheric shear along the tional Geology Review, v. 43, p. 1087–1093. MANUSCRIPT ACCEPTED 8 DECEMBER 2006

Geosphere, April 2007 85

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/3/2/71/893719/i1553-040X-3-2-71.pdf by guest on 27 September 2021