3-D geodynamic models of the India-Eurasia collision zone: investigating the role of lithospheric strength variation Sarah Bischoff and Lucy Flesch EAPS, Purdue University

The India-Eurasia collision zone is the largest zone of continental deformation on the Earth's surface. A proliferation of geodetic, seismic, and geologic data across the zone provides a unique opportunity for constraining geodynamic models and increasing our understanding of mountain building and plateau growth. We present a 3-D, spherical, Stokes flow, finite volume, geodynamic model of the India-Eurasia collision. Lithospheric volume is constrained by seismic data. Continuous surface velocities, inferred from GPS and Quaternary fault slip data, are used to approximate velocity boundary conditions. We assume a stress-free surface, and free-slip along the model base. Model viscosity varies with depth and is calculated assuming the laterally-varying, depth-averaged viscosities of Flesch et al. (2001) and a cratonic Indian plate. Laterally the model extends from the southern tip of India northward to the Tian Shan, and from the Pamir Mountains eastward to the South China block. Vertically the model volume extends to a depth of 100 km, and is divided into three layers: upper crust, lower crust, and upper-lithospheric mantle. We use COMSOL Multiphysics (www.comsol.com) to investigate the role of vertical viscosity variation on surface deformation by holding the dynamics constant, adjusting the viscosity substructure, and determining the resultant stress and velocity fields. Solved model surface velocities are compared to the observed surface velocities inferred from GPS and Quaternary fault slip rates. A two-layer model employing laterally-variant viscosity estimates throughout the crust and mantle is ineffective at replicating the observed force balance. The weak crustal viscosities necessary for attaining the observed clockwise rotation around the eastern Himalayan syntaxis also result in erroneous southward velocities in southern Tibet, driven by excessive gravitational collapse. Strengthening crustal viscosities balances the boundary/body forces and allows for accommodation of Indian plate motion across Tibet, but no longer produces clockwise rotation around the eastern syntaxis. The best-fit velocity magnitude and rotation solution is achieved by a full three-layer model incorporating an upper crust of intermediate strength, a weaker lower crust, and a stronger upper mantle. Our three-layer model achieves rotation around the indenter without excessive gravitational collapse. Model and observed velocities diverge slightly in the Tarim Basin, the southern Gobi, and the northern South China block. Model velocities in the Tarim Basin are shifted in an easterly direction; possibly indicating a weaker than previously assumed Altyn Tagh fault, while Gobi and South China model velocities are shifted to the north; suggesting the presence of an additional level of complexity.