Simulating the Development of Seismic Anisotropy within an Unstructured Mesh, Computational Modelling Framework: Application to Transform Fault

Abstract:

Seismic anisotropy, generated through the Lattice Preferred Orientation (LPO) of anisotropic minerals can yield some of the most direct constraints on both regional and global mantle flow. Interpretations of seismic anisotropy, however, requires knowledge of the relationship between the nature of the LPO and the convective flow field, insights into which can be derived from geodynamical models.

To predict seismic anisotropy, we utilize a modified version of D-Rex, a software package which calculates the development of LPO of olivine-enstatite crystal aggregates within steady-state flow models. For our purposes, the standard D-Rex package was re-written in python, parallelised and expanded to 3 dimensions. This allowed us to implement D-Rex as a post-processing package for Fluidity, a multi-purpose, finite-element, adaptive, unstructured mesh computational modelling framework. This presentation will summarise the steps and challenges involved with implementing our modified 3-D D-Rex package, as well as validation against 2-D and 3-D seafloor spreading models. Finally, applications to a 3-D transform fault model will be presented, with anisotropy results compared to published observations.