GEOLOGY, May 2008 363ABSTRACTThe recent dramatic increase in seismic anisotropy and surface global positioning system (GPS) data for central Asia permits a com-prehensive examination of the mantle’s role in mountain building. A joint analysis of 178 shear-wave-splitting and ~2000 GPS observations using a new technique reveals that the crust and lithospheric mantle deform coherently, arguing for crust-mantle mechanical coupling during deformation. The observed spatial variations in anisotropy refl ect the large-scale pattern of lithospheric deformation, as well as a change in deformational style from simple shear on the Tibetan Plateau transitioning to pure shear in surrounding regions.Keywords: anisotropy, lithospheric deformation, global positioning system, central Asia.INTRODUCTIONHow mountains form is one of Earth science’s basic questions. Prog-ress has been made in understanding crustal properties of continental lithosphere in the mountain-building process, i.e., the anomalous thick-ness and high topography created by the process of continent-continent collision. The role of the lithospheric mantle in orogeny, however, is less clear, as refl ected in the range of mantle properties found in current models for orogenic deformation (England and Houseman, 1986, 1989; Royden et al., 1997; Tapponnier et al., 1982; Thatcher, 2006; Meade, 2007). If the lithospheric mantle deformation could be measured directly and compared to the surface, it would constitute an important means of assessing the mantle’s role in orogens. Such a means is now avail-able through the joint analysis of surface deformation data and mantle seismic anisotropy (Flesch et al., 2005). A recent increase in the number of both types of data in central Asia and conceptual advances in their joint analysis provide strong support for mechanically coupled orogenic lithosphere throughout this region.GEOPHYSICAL DATAMantle anisotropy is constrained with 178 SKS shear-wave split-ting observations, 73 from previously published studies (McNamara et al., 1994; Huang et al., 2000; Flesch et al., 2005; Lev et al., 2006; Sol et al., 2007) and 105 new ones analyzed in this study obtained from portable deployments in Yunnan (2002) (Chang et al., 2006), eastern Tibet and Sichuan (2004–2006, Chinese Earthquake Administration, Carnegie Institution of Washington, Saint Louis University, Multimax Corpora-tion), and permanent stations from Chinese regional and national networks (Zhao et al., 1997) (Fig. 1; GSA Data Repository Table DR11). We have used standard methods (Silver and Chan, 1991; Wolfe and Silver, 1998) to calculate individual splitting observations of the fast polarization direction, φ, and delay time, δt, as well as station stacks. For our own obser vations, only those stations where splitting could unambiguously be detected were used. We proceed assuming that the seismic anisotropy arises from a single homogenous mantle layer because: (1) we checked our data for evidence of two anisotropic layers by the analysis of back-azimuthal variations in both φ and δt (Silver and Savage, 1994), but found none (Figs. DR1–DR3; see footnote 1), and (2) several studies have concluded that the crust cannot account for SKS splitting observations in central Asia based on studies of crustal splitting (McNamara et al., 1994; Herquel et al., 1995; Sherrington et al., 2004; Frederiksen et al., 2003; Ozacar and Zandt, 2004; Karalliyadda et al., 2007).For the surface deformation fi eld, we have used ~2000 global posi-tioning system (GPS) observations (Abdrakhmatov et al., 1996; Bendick Geology, May 2008; v. 36; no. 5; p. 363–366; doi: 10.1130/G24450A.1; 2 fi gures; Data Repository item 2008091.© 2008 The Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or [email protected] Data Repository item 2008091, method equations, uncertainty de-scriptions, Figures DR1–DR6, and Tables DR1–DR2, is available online at www.geosociety.org/pubs/ft2008.htm, or on request from [email protected] or Documents Secretary, GSA, P.O. Box 9140, Boulder, CO 80301, USA.Evidence for mechanically coupled lithosphere in central Asia and resulting implicationsChun-Yong Wang Institute of Geophysics, China Earthquake Administration, Beijing 100081, People’s Republic of ChinaLucy M. Flesch Department of Earth & Atmospheric Sciences, Purdue University, West Lafayette, Indiana 47907, USAPaul G. Silver Carnegie Institution of Washington, Washington, DC 20015, USALi-Jun Chang Institute of Geophysics, China Earthquake Administration, Beijing 100081, People’s Republic of ChinaWinston W. Chan Multimax Corporation, 1441 McCormick Drive, Landover, Maryland 20774, USAFigure 1. Shear wave splitting data set used in this study. See also Table DR1 (see footnote 1). Orientation of lines gives fast polariza-tion direction, φ, and delay time, δt, given by length of bar, according to legend. These represent several previous portable deployments, one recent deployment, as well as the analysis of permanent sta-tions from a variety of networks. Black lines represent large left-lateral faults, and green lines represent right-lateral faults. Plot is shown in degrees east and north. (X-X—Xianshuihe-Xiaojiang fault, R-R—Red River fault, T—Tibet, S—Szechwan, Y—Yunnan, BNS—Bangong-Nujiang suture).364 GEOLOGY, May 2008et al., 2000; Calais et al., 1998; Chen et al., 2000; Heki et al., 1999; Michel et al., 2001; Shen et al., 2000, 2001; Simons et al., 1999; Wang et al., 2001; Yu et al., 1999; Zhang et al., 2004; Zhu et al., 2000; http://sideshow.jpl.nasa.gov/mbh/series.html) and Quaternary fault slip rates (England and Molnar, 1997) to quantify the velocity, Vc, velocity gradient tensor, Lc, and strain-rate tensor, ε˙Yc, following the method of Haines and Holt (1993) and Haines et al. (1998) (Fig. DR4).JOINT ANALYSIS OF SHEAR-WAVE SPLITTING AND GPS DATAWe directly compare mantle anisotropy with predicted fast polari-zation directions, φc, from the surface deformation fi eld. We assume A-type lattice preferred orientation (LPO) in olivine, where the olivine a-axis is parallel to the fi nite-strain maximum shear direction for simple shear (Zhang and Karato, 1995), and the fi nite-strain extension direction for pure shear (Nicolas et al., 1973). For the nearly vertically propagating shear waves we are using, and assuming that the a-axis concentration is subhorizontal, φc will have the same