Deformation in the lowermost mantle beneath Australia from observations and models of seismic anisotropy

Abstract

Observations of seismic anisotropy near the core‐mantle boundary may yield constraints on patterns of lowermost mantle flow. We examine seismic anisotropy in the lowermost mantle beneath Australia, bounded by the African and Pacific Large Low Shear Velocity Provinces. We combined measurements of differential splitting of SKS‐SKKS and S‐ScS phases sampling our study region over a range of azimuths, using data from 10 long‐running seismic stations. Observations reveal complex and laterally heterogeneous anisotropy in the lowermost mantle. We identified two subregions for which we have robust measurements of D″‐associated splitting for a range of ray propagation directions and applied a forward modeling strategy to understand which anisotropic scenarios are consistent with the observations. We tested a variety of elastic tensors and orientations, including single‐crystal elasticity of lowermost mantle minerals (bridgmanite, postperovskite, and ferropericlase), tensors based on texture modeling in postperovskite aggregates, elasticity predicted from deformation experiments on polycrystalline MgO aggregates, and tensors that approximate the shape preferred orientation of partial melt. We find that postperovskite scenarios are more consistently able to reproduce the observations. Beneath New Zealand, the observations suggest a nearly horizontal [100] axis orientation with an azimuth that agrees well with the horizontal flow direction predicted by previous mantle flow models. Our modeling results further suggest that dominant slip on the (010) plane in postperovskite aggregates provides a good fit to the data but the solution is nonunique. Our results have implications for the mechanisms of deformation and anisotropy in the lowermost mantle and for the patterns of mantle flow.

A Library of Elastic Tensors for Lowermost Mantle Seismic Anisotropy Studies and Comparison With Seismic Observations

“…Bottom image shows a stereoplot showing predicted S wave anisotropy for the best fitting orientations (same as magenta dot above) in which colors show S wave anisotropy (%) and black bars are predicted fast splitting directions. White bars show the actual splitting observations (Creasy et al, 2017) for comparison. Black arrows give geographic reference.…”

“…In addition to testing the range of polycrystalline elastic tensors from our library, we also tested single‐crystal elastic tensors against observations contained in each of these data sets. Elasticity scenarios based on single crystals are often used to model observations of shear wave splitting (e.g., Creasy et al, 2017; Ford et al, 2015); therefore, we wanted to compare how well a set of seismic observations fits both an elastic model based on a single crystal and its corresponding polycrystalline elastic tensor. For the ferropericlase single crystal, only the data sets of western United States and Caribbean are consistent with this elasticity tensor; in contrast, the polycrystalline VPSC ferropericlase tensor can fit every data set.…”

“…Ford et al (2015) observed shear wave splitting for ScS, SKS (red arrows), and SKKS (orange arrows) beneath the Afar plume. Creasy et al (2017) observed shear wave splitting for ScS, SKS, and SKKS beneath the southwestern edge of Australia and beneath New Zealand. For SKS and SKKS phases, we exaggerate the length of the D ″ portion of the raypath by four times for visibility.…”

“…To test whether a given elastic tensor is consistent with a given set of D ″ anisotropy observations, we carried out forward modeling based on the framework used by Ford et al (2015) and Creasy et al (2017). First, we assume that the anisotropic structure is laterally homogenous within each region and thus that each of the individual raypaths is sampling the same anisotropy.…”

Lowermost Mantle Anisotropy Beneath Africa From Differential SKS‐SKKS Shear‐Wave Splitting

“…It is well known that upper mantle anisotropy makes the primary contribution to SK ( K ) S splitting worldwide (e.g, Becker et al, ), and contributions from the lower mantle are a second‐order effect. Some SK ( K ) S studies of lower mantle anisotropy explicitly correct waveforms for the effect of splitting due to upper mantle anisotropy and attribute the remaining signal to the lowermost mantle (examples include Long & Lynner, ; Ford et al, ; and Creasy et al, ; Lynner & Long, , also carried out upper mantle corrections for a small subset of their stations). We now consider the nature of likely contributions from upper mantle anisotropy to our data set, as this informs our strategy for the interpretation of our SKS ‐ SKKS data set.…”