Constraints on Lowermost Mantle Mineralogy and Fabric beneath Siberia from Seismic Anisotropy
Orientations of candidate elastic models which fit the shear-wave splitting observed beneath Siberia. Panel A is a cartoon showing how to interpret the other panels. Plots are upper hemispheric projections of the shear plane (grey arcs) and direction (black circles) predicted by matching orientations of different elastic tensors to the polarisations observed in the data. North is up the page, as indicated, and normal to the page is vertical (i.e., radially in the Earth). The elastic tensors tested are: B and C, an LPO of MgO; D, an ideal TTI medium; E and F, MgSiO3 perovskite at 1500K and 3500K; G, post-perovskite LPO using MgGeO3; H and I, post-perovskite LPO using CaTiO3.
<p>Seismic anisotropy is an important tool for studying the nature, origin and dynamics of the lowermost mantle (D"). Here we present analysis of the seismic anisotropy beneath Siberia from shear-wave splitting measured in ScS phases. Data come from two near-perpendicular raypaths (Hindu-Kush to Northern Canada and the Kuril Arc to Germany; both at ~80 degrees epicentral distance) with close ScS reflection points on the Core-Mantle boundary (CMB). We apply differential S-ScS splitting to minimise contamination from the source and receiver side upper mantle. The two raypaths show different ScS splitting times and fast shear-wave orientations, incompatible with the VTI style of anisotropy inferred for much of the lowermost mantle. The availability of data at two azimuths give us an opportunity to better understand D" anisotropy than in previous studies. For example, the results provide the first accurate measurement of the dip of the symmetry plane. Several mechanisms have been suggested to explain lowermost mantle anisotropy, including the lattice-preferred orientation of lower mantle minerals such as perovskite or post-perovskite, or the shape-preferred orientation of inclusions of melt. In order to infer the flow regime implied by these mechanisms we use elasticities from published deformation experiments to forward model shear-wave splitting. Tomography of the region suggests a north-south trend in the geodynamics, and a model incorporating post-perovskite with a [100](010) slip system or aligned melt inclusions are most naturally compatible with such a trend. This may suggest a connection with remnant slab material from past subduction in the north Pacific.
</p><p>References
</p><p>Wookey, J., Kendall, J-M. (2008) Constraints on Lowermost Mantle Mineralogy and Fabric beneath Siberia from Seismic Anisotropy, Earth Planet. Sci. Lett., 275, 32-42
</p><p>Acknowledgements: Data came from the CNSN and SZGRF datacentres. JW was supported by a NERC fellowship grant.</p>