Investigating the magnitude of lower crustal flow and impact on surface deformation patterns in Tibet using 3-D geodynamic models

Differential flow in the lower crust of Tibet has been invoked to explain features in the region, including uniform plateau elevation, crustal thickness/topographic gradients, and uplift without observed shortening. Here, we use 3-D finite element modeling to test impacts of assumed lower crustal viscosities on deformation patterns in the India-Eurasia collision zone. We simulate instantaneous lithospheric deformation with Stokes flow using COMSOL Multiphysics (www.comsol.com). Our model geometry ranges eastward from the Pamir to Sichuan, northward from the southern tip of India to the Tien Shan, and vertically downward from the Earth's surface to 100 km below sea level. We divide model geometry into four domains: Indian lithosphere, Eurasian upper crust, lower crust, and upper mantle. Seismic and magnetotelluric study results guide inclusion of subducted Indian and Burma slabs along with our targeted weak lower crust. Within the larger Eurasian lower crust domain, weak lower crust is restricted to a zone bounded clockwise by the Himalayan Frontal Thrust, Karakorum, Altyn-Tagh, Kunlun, Longmen Shan, and onset of lower elevations along the plateau's southeastern margin. From top to bottom, vertical bounds of the zone are constrained by a constant 20 km below sea level and the shallower of either the top of the Indian slab or Moho. Strength is approximated via 3-D maps of effective viscosity constrained by the vertically-averaged lithospheric estimates of Flesch et al. [2001]. We forward model lower crust effective viscosities on the order of 1018 to 1022 Pa•s and inspect resulting horizontal and vertical deformation patterns. Results suggest that effective viscosities of less than 1020 Pa•s are required for both appreciable differential mass flux through lower crustal flow as well as decoupled lower crustal flow from the upper crust or mantle. Movement of the lower crust is partitioned within weaker fault zones. Effective viscosities of 1020 Pa•s or less produce pronounced patterns of surface subsidence in Qiangtang and uplift in eastern Lhasa and Longmen Shan inconsistent with observations. Solutions show lower crust strength impacts surface stress style with weaker strengths leading to regions of dominant extension separated by compression in the east central Tibetan Plateau.