The sensitivity of lowermost mantle anisotropy to past mantle convection
James Ward,Andrew M. Walker,Andy Nowacki,James Panton,Huw Davies
DOI: https://doi.org/10.1016/j.pepi.2024.107264
IF: 2.748
2024-09-26
Physics of The Earth and Planetary Interiors
Abstract:It is widely believed that seismic anisotropy in the lowermost mantle is caused by the flow-induced alignment of anisotropic crystals such as post-perovskite. What is unclear, however, is whether the anisotropy observations in the lowermost mantle hold information about past mantle flow, or if they only inform us about the present-day flow field. To investigate this, we compare the general and seismic anisotropy calculated using Earth-like mantle convection models where one has a time-varying flow, and another where the present-day flow is constant throughout time. To do this, we track a post-perovskite polycrystal through the flow fields and calculate texture development using the sampled strain rate and the visco-plastic self-consistent approach. We assume dominant slip on (001) and test the effect of the relative importance of this glide plane over others by using three different plasticity models with different efficiencies at developing texture. We compare the radial anisotropy parameters and the anisotropic components of the elastic tensors produced by the flow field test cases at the same location. We find, under all ease-of-texturing cases, the radial anisotropy is very similar (difference <2% ) in the majority of locations and in some regions, the difference can be very large ( >10% ). The same is true when comparing the elastic tensors directly. Varying the ease-of-texture development in the crystal aggregate suggests that easier-to-texture material may hold a stronger signal from past flow than harder-to-texture material. Our results imply that broad-scale observations of seismic anisotropy such as those from seismic tomography, 1-D estimates and normal mode observations, will be mainly sensitive to present-day flow. Shear-wave splitting measurements, however, could hold information about past mantle flow. In general, mantle memory expressed in anisotropy may be dependent on path length in the post-perovskite stability field. Our work implies that, as knowledge of the exact causative mechanism of lowermost mantle anisotropy develops, we may be able to constrain both present-day and past mantle convection.
geochemistry & geophysics