Probabilistic Finite Frequency SKS Splitting Intensity Tomography: Applications to South America and the Central Appalachians

We apply a model space search approach for the tomographic inversion of SKS splitting intensity measurements for anisotropic structure in the upper mantle to data sets from Southern Peru and the Central Appalachians. While SKS splitting measurements provide powerful constraints on upper mantle anisotropy, particularly its lateral variation, they lack information about the depth distribution of anisotropy. A tomographic approach that takes into account the finite-frequency sensitivity of SKS waves has the potential to provide such depth constraints, although such an approach requires a dense array of broadband stations to be successful. We apply a Markov chain Monte Carlo with Gibbs sampling approach to avoid linearizing a highly nonlinear inverse problem for anisotropic structure. We apply this technique to dense data sets from the South American subduction zone (the PULSE and CAUGHT datasets from Peru/Bolivia) and from the Central Appalachian Mountains (the MAGIC experiment) after measuring SKS splitting intensity values. We present images of upper mantle anisotropy beneath these regions derived from SKS splitting intensity tomography, along with a series of resolution tests to verify the robustness of the models.