A Markov chain Monte Carlo approach to forward modeling anisotropic receiver functions – Application to the eastern North American craton

P-to-S receiver functions have the potential to place strict constraints on Earth structure. It particular, backazimuthal variations in P-to-SH converted phases can be used to infer the orientation of anisotropic symmetry axes at depth. Many prior studies have used a trial-and-error approach to forward modeling of receiver functions, making it impossible to explore model space in a systematic and quantitative manner. Here we present a Markov chain Monte Carlo algorithm for forward modeling of receiver functions. Synthetic examples with complex Earth structure are used to test the viability of the algorithm.

We use this methodology to investigate the anisotropic character of the seemingly pervasive mid-lithospheric discontinuities (MLDs) beneath the central United States. Previous receiver function results have revealed multiple mid-lithospheric discontinuities that correspond to sharp gradient in anisotropy, with a N- to NW- orientation of anisotropy in the upper mantle lithosphere. To place more quantitative constraints on Earth structure, as well as to bolster previous results, we apply the MCMC method to actual receiver function data from station WCI in the central United States. Ultimately, model space searches such as the MCMC algorithm presented here provide a framework for forward modeling of receiver functions to obtain estimates on Earth structure, and the orientation of anisotropy at depth, in a systematic and quantitative manner.