Crustal anisotropy in northeastern Tibetan Plateau inferred from receiver functions: Rock textures caused by metamorphic fluids and lower crust flow?

•We analyzed 35 stations using receiver functions and forward-modeled five stations.•Anisotropy concentrates in three isolated zones in the upper, middle and lower crust.•Horizontal symmetry axes are inconsistent with typical LPO of foliated crustal rocks.•LPO of metamorphic veins or channelized metamorphic is an alternative explanation.

The crust of Tibetan Plateau may have formed via shortening/thickening or large-scale underthrusting, and subsequently modified via lower crust channel flows and volatile-mediated regional metamorphism. The amplitude and distribution of crustal anisotropy record the history of continental deformation, offering clues to its formation and later modification. In this study, we first investigate the back-azimuth dependence of Ps converted phases using multitaper receiver functions (RFs). We analyze teleseismic data for 35 temporary broadband stations in the ASCENT experiment located in northeastern Tibet. We stack receiver functions after a moving-window moveout correction. Major features of RFs include: 1) Ps arrivals at 8–10 s on the radial components, suggesting a 70–90-km crustal thickness in the study area; 2) two-lobed back-azimuth variation for intra-crustal Ps phases in the upper crust (< 20 km), consistent with tilted symmetry axis anisotropy or dipping interfaces; 3) significant Ps arrivals with four-lobed back-azimuth variation distributed in distinct layers in the middle and lower crust (up to 60 km), corresponding to (sub)horizontal-axis anisotropy; and 4) weak or no evidence of azimuthal anisotropy in the lowermost crust. To study the anisotropy, we compare the observed RF stacks with one-dimensional reflectivity synthetic seismograms in anisotropic media, and fit major features by “trial and error” forward modeling. Crustal anisotropy offers few clues on plateau formation, but strong evidence of ongoing deformation and metamorphism. We infer strong horizontal-axis anisotropy concentrated in the middle and lower crust, which could be explained by vertically aligned sheet silicates, open cracks filled with magma or other fluid, vertical vein structures or by 1–10-km-scale chimney structures that have focused metamorphic fluids. Simple dynamic models encounter difficulty in generating vertically aligned sheet silicates. Instead, we interpret our data to support the hypothesis of vertical metamorphic-fluid domains whose alignment is determined by shear motion within the boundary layers of crustal channel flow.