Towards global observations of shear wave splitting in the Earth's lowermost mantle

It is well documented that the Earth's lowermost mantle (a.k.a. D") is seismically anisotropic. This anisotropy is indicative of a long range order which may be caused by the alignment of crystals in a strained fabric due to lower mantle flow, or alternatively by the presence of shaped heterogeneities such as fine layers, or pockets of melt. The resolution of this anisotropy, and its proper interpretation, will place important constraints on the thermodynamic and mineralogical properties of this elusive region. Local studies of D" anisotropy have revealed that in some places the simplest style of anisotropy that fits the data is transversely isotropic with a tilted axis of symmetry. To date, global studies of lowermost mantle anisotropy have made the assumption that the anisotropy is transversely isotropic with a vertical axis of symmetry. This limits the capacity of such measurements to infer geodynamic or mineralogical causes. It remains an outstanding challenge to make global observations of anisotropy in D" without recourse to simplifying assumptions as to the style and symmetry axis of anisotropy. We use seismograms from the IRIS DMC (over the period 1998 to 2005) to perform S-ScS differential splitting and thus make observations of shear wave splitting in the lowermost mantle at a number of localities around the globe. Our method makes no assumptions as to the style of anisotropy. The determination from the S phase of splitting in the upper mantle beneath stations and events enables us to ensure that splitting in the upper mantle does not contaminate our observations of splitting in the lowermost mantle on the ScS phase. This work validates our method, and provides a firm platform from which to further our observations of shear wave splitting in D". Our dataset has excellent spatial coverage over the Northern Hemisphere with especially good azimuthal coverage over a large swath beneath Asia. The global nature of the dataset will enable us to measure splitting in D" over an expansive region in a globally consistent manner. Interpretation of these measurements in light of the azimuthal coverage will permit the resolution of the style and symmetry axis of anisotropy. Considerations from mineral physics, geodynamics, and waveform modelling will place new constraints on the cause of anisotropy in D", which will lead to better understanding of how this region contributes to the broader dynamic Earth.