P, S, and Rayleigh wave tomography of the southwestern U.S. upper mantle

We present new 3-D P and S velocity tomography focused on the southwestern U.S. upper mantle. Body wave tomography models are derived from teleseismic P and S travel-time residuals measured from waveforms recorded by the SCSN and NCSN broadband and short-period arrays, USArray stations, and other temporary deployments. The body wave travel-time tomography method uses frequency dependent 3-D sensitivity kernels to map residual times, measured in multiple frequencies, into velocity structure and recently advanced crust thickness and velocity models to better isolate the mantle component of residual times. In addition to separate Vp and Vs body wave models, we jointly invert the two datasets for Vp/Vs perturbations by imposing a smoothness constraint on the dlnVs/dlnVp field. This joint inversion helps identify mantle volumes where non-thermal effects, such as partial melt, contribute strongly to imaged velocity structure. In an effort to improve resolution of upper mantle heterogeneity and understand differences between recent body wave and surface wave tomography, we iteratively perform separate inversions of S body wave travel-times and Rayleigh wave phase and amplitude data using the results of the previous inversion as a 3-D starting model for the next inversion. The initial 3-D starting model for the sequence of inversions is a new Rayleigh wave tomography model of the southwestern U.S. crust and upper mantle. A slightly modified version of the finite-frequency Two Plane Wave Tomography method is used to invert Rayleigh wave phase and amplitude data from USArray and SCSN stations. This conceptually simple attempt at a regional-scale joint inversion exploits the complementary characteristics of body wave and surface wave sensitivity kernels and allows quantitative assessment of how well a single 3-D Vs model can fit both datasets. Collectively, the suite of tomography models provide improved constraint on the form and amplitude of southwestern U.S. upper mantle heterogeneity including high-velocity anomalies beneath the southern Great Valley, Transverse Ranges, and south-central Utah and low-velocity anomalies beneath the Salton Trough, Walker lane, St. George volcanic trend, and the San Francisco Peaks. Reaching consensus between body wave and surface wave imaging will be an important step toward resolving these mantle structures at a level that is diagnostic of the geodynamic processes responsible for associated surface geologic activity.