Modeling of Seismic Anisotropy Observations Reveals Plausible Lowermost Mantle Flow Directions Beneath Siberia

Observations of seismic anisotropy just above the core-mantle boundary are a powerful way to understand the dynamics of the lowermost mantle. Here we present models of seismic anisotropy in the lowermost mantle beneath Siberia based on new and previously published body wave observations. We compile a set of measurements based on a variety of data types, including the differential splitting of SK(K)S-PKS and S-ScS phases and polarities of PdP and SdS reflections off the D '' discontinuity. We carry out ray theoretical forward modeling of these data sets in combination, using a novel approach that allows for tighter constraints on the geometry of seismic anisotropy than would be possible using a single data type. Observations for each seismic phase alone only provide limited information; by combining different data types into one modeling approach, we can constrain D '' seismic anisotropy more tightly. We test a range of plausible mechanisms for seismic anisotropy, including a variety of candidate minerals (bridgmanite, post-perovskite, and ferropericlase), dominant slip systems, and orientations, and consider both single-crystal elasticity and elastic tensors based on polycrystalline texture modeling. In general, we find that post-perovskite (with either a [100](010) or [100](001) dominant slip system) or bridgmanite models provide the best fit to the observations. The best-fitting models suggest a dominant shear direction to the southwest or northeast direction at the base of the mantle. This is consistent with flow directed toward the Perm anomaly, potentially driven by slab remnants impinging on the core-mantle boundary.

Automated summary

Observations of seismic anisotropy just above the core-mantle boundary provide insights into the dynamics of the lowermost mantle. By combining different data types in modeling approaches, constraints on the geometry and mechanisms of seismic anisotropy can be tightened, with post-perovskite or bridgmanite models showing the best fit to observations. The best-fitting models suggest a dominant shear direction to the southwest or northeast at the base of the mantle, consistent with flow towards the Perm anomaly driven by slab remnants at the core-mantle boundary.