Mantle anisotropy beneath the Tibetan Plateau: evidence from long-period surface waves

We investigate long-period surface waves (T> 100 s) propagating across the Tibetan Plateau, based on the seismic records from a PASSCAL portable array and nearby Chinese Digital Seismic Network (CDSN) stations. Significant quasi-Love waveform anomalies, associated with fundamental Rayleigh-Love coupling, are consistently observed, suggesting that strong lateral gradients in azimuthal anisotropy exist beneath the plateau. More intense surface-wave scattering is observed at intermediate periods (T=50 s), which suggests that the cause of the long-period quasi-Love waves lies in the upper mantle, not the thickened crust of the plateau. A detailed analysis of the data indicates that the long-period waveform anomalies are generated beneath the central Tibetan Plateau, where the structural trend implies deep deformations induced by continental collision. The absence of quasi-Love anomalies at the westernmost station of the Program for Array Seismic Studies of the Continental Lithosphere (PASSCAL) array suggests that the east-west extent of this mantle deformation is limited. Some, but not all, of the quasi-Love observations are consistent with SKS splitting observations. Both sets of observations predict a strong gradient in anisotropic properties in central Tibet. However, the quasi-Love waveforms are absent from records collected in the northern plateau for northerly propagation paths, which is not consistent with a gradient at the northern edge of the plateau, as suggested by SKS studies. This discrepancy indicates that the simple models used to interpret both bodyand surface-wave data may be inadequate. Either significant P-wave anisotropy exists under the plateau, or the S-wave anisotropy does not possess a uniformly horizontal symmetry axis, or strong east-west gradients in anisotropy bias our data-synthetic comparison.