Seismic anisotropy (i.e. seismic waves travel in different directions at different speeds) in the deeper earth was discovered in the mid-sixties and was soon interpreted in a qualitative way as a result of crystal alignment during convection (LPO). This concept since became generally accepted. More recently strong anisotropy and heterogeneity was documented in the lowermost mantle adjacent to the metallic and liquid core. This enigmatic D" zone is both a thermal and chemical boundary layer characterized complex dynamic processes that are reflected in many intriguing seismic observations. Much progress was recently achieved in mineral physics, to characterize elastic and deformation properties of lowermost mantle minerals including the post-perovskite phase. Advances in geodynamic modeling now allow us to track the strain evolution during mantle convection. As a result, there are now precise ways to compute synthetic seismograms in a 3D anisotropic earth down to body wave frequencies.
Our proposed study follows on preliminary work started 2 years ago, and is focused on the forward modeling of LPO anisotropy in D", with the goal of combining tools and observations developed by geodynamicists, seismologists and mineral physicists, in order to gain better understanding of the origin of seismic anisotropy in D", and determine which microscopic and macroscopic processes may be at play. In our work to-date, we have set up a multi-step procedure for this purpose, which combines five modeling ingredients in a logical chain: (1) For a particular hypothesis regarding mantle dynamics, geodynamical models provide information on the macroscopic strain deformation accommodated by individual packets of mantle material. (2) This strain deformation information is then used as boundary conditions for numerical models that calculate the resulting mineralogical texture (i.e. LPO) within a polycrystalline mineral aggregate. (3) Seismic elastic constants, determined from mineral properties and preferred orientations, are applied to numerous mineral aggregates throughout the region of interest, (4) followed by forward seismic modeling through the 3D elastic anisotropic model acquired from steps 1-3. (5) Resulting models and seismic waveforms are compared to available seismic observations. Our initial results illustrate that we can use macroscopic observations to constrain plausible constituents and deformation mechanisms. So far, our work has focused on 2D models of a subducting slab reaching the core-mantle boundary (CMB) and its subsequent spreading along the CMB. We here propose to extend our set of geodynamic tools to the 3D case, which will allow us to explore the predicted distribution of different kinds of anisotropy at the base of the mantle in a more realistic framework and confront the resulting seismic wavefields to available broadband seismic data. We will perform experiments and theoretical mineral physics computations. This project promotes interdisciplinary work and cross-education in the fields of geodynamics, mineral physics and seismology among the PI's, participating students and postdoctoral associates, stimulating learning to solve complex scientific problems by sharing different expertise. It will also provide suggestions for future seismic experiments targeted at better characterizing anisotropy in D".
