Subduction of the Indian Plate and the Nature of the Crust Beneath Western Tibet: Insights From Seismic Imaging

The northward extent of subducted Indian plate is a fundamental component of hypotheses explaining deformation and magmatism within the Tibetan Plateau. Yet, these aspects of the plate are debated in west Tibet. Here we report a new three-dimensional lithospheric structure of seismic velocity and radial anisotropy under west Tibet constructed from Rayleigh and Love wave phase velocity maps at periods of 20-167 and 20-125 s, respectively. Our results show the Indian lithospheric mantle to be subhorizontally subducted under west Tibet across the Bangong-Nujiang suture to the Qiangtang terrane, as indicated by a prominent fast velocity anomaly accompanied with positive radial anisotropy (Vsh > Vsv). We find a positive spatial correlation of this result with information on the distribution of late Cenozoic potassic-adakitic rocks in western Tibet. Additionally, we show that the midcrust of west Tibet is characterized by an anomalously low shear wave velocity (3.2-3.4 km/s at similar to 30-km depth) and positive anisotropy, which is consistent with an estimated similar to 3% fraction of partial melt. We suggest that the midcrust of this region is capable of flowing and that its three-dimensional structure shows it to extend south of the Karakoram fault (KKF), a shear zone interpreted as a barrier to crustal flow. Instead, our results are consistent with the KKF embedded in weak middle crust along with several other structures that display a pattern of distributed deformation in the western portion of the Tibetan Plateau. Plain Language Summary We have constructed a 3-D image under west Tibet, one of the most remote areas in Tibet, with seismic waves traveling on surface of the Earth. Our lithospheric image provides direct evidence of seismic velocity and radial anisotropy to outline the variable geometry and northern extent of the subducted Indian plate. We find that the Indian lithospheric mantle is subhorizontally subducting across the Bangong-Nujiang suture to the Qiangtang terrane, which holds a positive spatial correlation with the information on the distribution of late Cenozoic potassic-adakitic rocks in western Tibet. More importantly, a similar to 3% fraction of partial melts is estimated to exist in the middle crust (similar to 30-km depth) of west Tibet. These partial melts could be connected to form a weak midcrustal layer prone to flow. The crustal flow is likely propagating southwards, penetrating the Karakoram fault, and could have exhumed to the surface in the Himalaya terrane of southwest Tibet. The dynamic processes could provide an explanation for the production of leucogranite in this region. In this case, Karakoram fault is interpreted as an upper crustal fault embedded in weak middle crust, failing to block the southward propagation of the midcrustal flow.