Crustal flow in Tibet: geophysical evidence for the physical state of Tibetan lithosphere, and inferred patterns of active flow

Abstract Many seismic and magnetotelluric experiments within Tibet provide proxies for lithospheric temperature and lithology, and hence rheology. Most data have been collected between c. 88°E and 95°E in a corridor around the Lhasa-Golmud highway, but newer experiments in western Tibet, and inversions of seismic data utilizing wave-paths transiting the Tibetan Plateau, support a substantial uniformity of properties broadly parallel to the principal Cenozoic and Mesozoic sutures, and perpendicular to the modern NNE convergence direction. These data require unusually weak zones in the crust at different depths throughout Tibet at the present day. In southern Tibet these weak zones are in the upper crust of the Tethyan Himalaya, the middle crust in the southern Lhasa terrane, and the middle and lower crust in the northern Lhasa terrane. In northern Tibet, north of the Banggong-Nujiang suture, the middle and probably the lower crust of both the Qiangtang and Songpan-Ganzi terranes are unusually weak. The Indian uppermost mantle is cold and seismogenic beneath the Tethyan Himalaya and the southernmost Lhasa terrane, but is probably overlain by a northward thickening zone of Asian mantle beneath the northern Lhasa terrane. Beneath northern Tibet the upper mantle has not been replaced by subducting Indian and Asian lithospheres, and is warmer than to the south. These inferred vertical strength profiles all have minima in the crust, thereby permitting, though not actually requiring, some form of channelized flow at the present day. Using the simplest parameterization of channel-flow models, I infer that a Poiseuille-type flow (flow between stationary boundaries) parallel to India-Asia convergence is occurring throughout much of southern Tibet, and a combination of Couette (top-driven, between moving boundaries) and Poiseuille lithospheric flow, perpendicular to lithospheric shortening, is active in northern Tibet. Explicit channel-flow models that successfully replicate much of the large-scale geophysical behaviour of Tibet need refinement and additional model complexity to capture the full details of the temporal and spatial variation of the India-Asia collision.