Investigating the Core-Mantle Boundary and ULVZ Topography with Synthetic FD Seismograms for 3-D Axi-Symmetric Geometries: Predictions and Data

We are interested in quantifying the effects of core-mantle boundary (CMB) and ultra-low velocity zone (ULVZ) topography on diffracted seismic phases. Such topography is important due to possible focusing/defocusing of energy that may strongly perturb the wavefield. In particular, we model the effects of topography on diffracted core phases such as SPdKS, which is often used to infer the presence, location and structure of ULVZs, which may be of partial melt origin. We model P/SV-wave propagation using a 3-D axisymmetric finite difference (FD) algorithm. The axisymmetric approach is used as we are able to produce synthetic seismograms with dominant frequencies on the order of 0.2 Hz to model detailed regional structures and test important aspects of the model space. Models tested include sinusoidal CMB topography and a range of ULVZ models, including isolated non-periodic features varying in shape from ring-shaped structures with Gaussian or dome shaped cross-sections, to structures with broader boxcar- (or mesa-) shaped cross-sections. We consider ULVZ models with sharp as well as gradational wave-speed transitions. We use Vs:Vp reductions of 1:1 and 3:1, the latter of which is appropriate for the partial melt scenario. Additionally, we study the effects of source versus receiver side ULVZ geometries for the seismic phase SPdKS. These results are compared to broadband data available from the Fast Archive Recovery Method (FARM) database. We utilize a global dataset of deep focus events with simple impulsive source mechanisms for the epicentral distance range of 100 to 130 deg. While CMB and ULVZ topography greatly expands the model space which already contains significant trade-offs, fixing ULVZ velocity and density perturbations allows discussion of possible topographical scenarios. Gradational wave-speed transitions reduces the magnitude of SKS pre-cursors predicted by sharp structures, which have not been observed in data. We also produce high resolution snap shots of the wave propagation in our structures, allowing delineation of additional arrivals due to complex structures.