Interpreting Radial Anisotropy in Global and Regional Tomographic Models

Abstract:

With the advent of exact numerical methods for seismic wavefield computation, the resolving power of regional and global tomographic models has clearly improved over the last decade. However, a remaining problem is computational cost, which leads seismologists to only interpret the long periods in seismic waveforms, and hence to only constrain long wavelength structure. In this way, tomographic images do not represent the true Earth, but rather an effective, apparent or equivalent model that provides a similar long wavelength data fit.

In this study, we focus on the problem of artificial radial anisotropy in the upper mantle due to unmapped small scale vertical heterogeneities (e.g. layering). We propose a fully probabilistic approach to sample the ensemble of layered models equivalent to a given smooth tomographic profile. We objectively quantify the trade-off between isotropic heterogeneity and strength of anisotropy. The non-uniqueness of the problem can be addressed by adding high frequency observations such as receiver functions, able to map first order discontinuities. We show that this method enables us to distinguish between intrinsic and extrinsic (i.e. artificial) anisotropy in 1D models extracted from tomographic results.