Seismic anisotropy of western Mexico and northeastern Tibet

In this dissertation, characteristics of upper mantle anisotropy, using shear wave splitting techniques, for two distinct tectonic provinces are presented.;In the first part, in western Mexico, the Rivera and Cocos plates subduct beneath the North America plate constituting a young subduction setting where plate fragmentation and capture is occurring today. We characterize the upper mantle anisotropy from SKS and local S phases from the data collected by the MARS experiment (MApping the Rivera Subduction zone) and by two stations of the Mexican Servicio Sismologico National. SKS shear-wave splitting parameters indicate that the fast directions of the split SKS waves for the stations that lie on the central and southern Jalisco block are approximately trench normal. Fast polarizations of these phases also follow the convergence direction between the Rivera Plate and Jalisco block with respect to the North America plate. S-wave splitting from slab events show a small averaged delay time of about 0.2 sec for the upper 60 km of the crust and mantle. Therefore, the main source of anisotropy must reside on the entrained mantle below the young and thin Rivera Plate. Trench-oblique fast SKS split directions are observed in the western edge of the Rivera Plate and western parts of the Cocos slab. The curved pattern of fast SKS split directions in the western Jalisco block and the Rivera-Cocos gap indicate 3-D toroidal mantle flow, around the northwestern edge of the Rivera slab and Rivera- Cocos gap. This behavior profoundly affects finite strain field in the northwestern edge of the Rivera slab and the mantle wedge. The shear wave splitting results support the idea that the Rivera and Cocos plates not only moved in a down-dip direction but also have recently rolled back towards the trench and the Colima rift is intimately related to the tearing between the Rivera and Cocos plates.;In the second study, the tectonic enviroment of the northeastern Tibetan plateau is considered. Shear wave splitting measurements using teleseismic SKS and SKKS phases recorded by the ASCENT (A Seismic Collaborative Experiment in Northeastern Tibet) and INDEPTH-IV (International Deep Profiling of Tibet and the Himalaya, Phase IV) experiments reveal significant anisotropy in north-eastern Tibet with a large delay time of up 2.2 sec, indicating that anisotropy exists in both the lithospheric and asthenospheric mantle. The coherence between fast polarization directions of split core phases and the left-lateral slip on eastern-striking, southeastern-striking and southern-striking faults in eastern Tibet as well as the surface velocity calculated from GPS data support the idea that left-lateral shear strain is the predominant cause of the orientation of the upper mantle petrofabrics. The left-lateral motion can be best understood as a manifestation of north-striking right-lateral simple shear exerted by the eastern edge of the underthrusting Indian plate as it penetrates into Eurasia, as well as the bending of the Eastern Himalayan Syntaxis (EHS) by the foundering Burma-Andaman-Sumatra slab. Two plausible competing models are proposed for the flow of asthenosphere. In the first, the deforming lithosphere gliding over the passive asthenosphere induces flow of the asthenosphere. In the second, the asthenosphere beneath northeastern Tibet is flowing eastward in an asthenosphere channel that lies between the Ordos plateau and Sichuan basin, and around the EHS as it is being compressed between the advancing Indian continental lithosphere and the thick Tarim and Qaidam lithospheres to the north. Delay times from stations in the EHS have a maximum of 1.3 sec suggesting that although most anisotropy is residing in the lithosphere, some may be associated with flow of the asthenosphere. The retreating Burma slab induces flow that is toroidal and located exclusively around the northern edge of the slab. The curved fast directions of split shear waves for stations in the EHS are consistent with the toroidal flow pattern as well as the rotational deformation of the overlying lithosphere. It is suggested that the foundering Burma plate may also play an important role in bending the EHS in the late Cenozoic time.