Seismological Detection of Azimuthal Anisotropy in the Transition Zone

The goal of this research is to determine whether azimuthal anisotropy is present in the transition zone. Mineral physics data demonstrate that wadsleyite, which is likely present in the upper transition zone, is intrinsically anisotropy. However, because the detection of seismic anisotropy at these depths is challenging, its existence in the transition zone is still a matter of debate. It is, nevertheless, an important issue since it can give us insight on the style of convection in the mantle. We apply a singular value decomposition inversion method to global azimuthally anisotropic Love wave phase velocity maps in order to constrain azimuthal anisotropy down to ~1000km depth. We use 70 different modes, fundamental and overtones up to order 5, at periods between 35s and ~175s. This gives us unprecedented sensitivity to elastic parameter G, which describes the azimuthal dependence of vertically polarized shear waves. Our preliminary results show that the best data fit is generally obtained for models that display a non-negligible amount of azimuthal anisotropy in the transition zone. Uncertainties remain regarding the amplitude and the fast direction of the anisotropy, but its presence under continents appears independent of the depth parameterization or the damping applied. Under oceans, the results are less stable with respect to damping and parametrization, and display large parameters trade-offs. This could be due to inconsistencies among the data due to a poorer azimuthal data coverage in those regions. We also tested the influence of the crustal model on the local sensitivity kernels and on the resulting models of azimuthal anisotropy. Our results show that the effect of the crust on parameter G is the strongest in the top 200km, but generally negligible at larger depths.