Crustal radial anisotropy across eastern Tibet and the Western Yangtze Craton

Phase velocities across eastern Tibet and surrounding regions are mapped using Rayleigh (8-65 s) and Love (8-44 s) wave ambient noise tomography based on data from more than 400 Program for Array Seismic Studies of the Continental Lithosphere and Chinese Earthquake Array stations. A Bayesian Monte Carlo inversion method is applied to generate 3-D distributions of Vsh and Vsv in the crust and uppermost mantle from which radial anisotropy and isotropic Vs are estimated. Each distribution is summarized with a mean and standard deviation, but is also used to identify "highly probable" structural attributes, which include (1) positive midcrustal radial anisotropy (Vsh > Vsv) across eastern Tibet (spatial average = 4.8% ± 1.4%) that terminates abruptly near the border of the high plateau, (2) weaker (-1.0% ± 1.4%) negative radial anisotropy (Vsh < Vsv) in the shallow crust mostly in the Songpan-Ganzi terrane, (3) negative midcrustal anisotropy (-2.8% ± 0.9%) in the Longmenshan region, (4) positive midcrustal radial anisotropy (5.4% ± 1.4%) beneath the Sichuan Basin, and (5) low Vs in the middle crust (3.427 ± 0.050 km/s) of eastern Tibet. Midcrustal Vs < 3.4 km/s (perhaps consistent with partial melt) is highly probable only for three distinct regions: the northern Songpan-Ganzi, the northern Chuandian, and part of the Qiangtang terranes. Midcrustal anisotropy provides evidence for sheet silicates (micas) aligned by deformation with a shallowly dipping foliation plane beneath Tibet and the Sichuan Basin and a steeply dipping or subvertical foliation plane in the Longmenshan region. Near vertical cracks or faults are believed to cause the negative anisotropy in the shallow crust underlying Tibet.