Radial Anisotropy Beneath the Continental U.S. From Surface Wave Phase Velocities

Abstract We assemble the highest‐resolution data set of Love wave phase velocities measured across the continental U.S. to date, combining observations from ambient noise and teleseismic earthquakes to span a period range between 10 and 120 s. We jointly invert this data set and Rayleigh wave phase velocity maps for 3‐D models of isotropic shear velocity and radial anisotropy . To rigorously investigate the necessity of radial anisotropy, we use as our starting model the results from Shen and Ritzwoller (2016, https://doi.org/10.1002/2016jb012887 ), which were derived from Rayleigh wave data. We find that the starting model consistently underpredicts the observed Love wave phase velocities and fits well the Rayleigh wave phase velocities, highlighting the need for radial anisotropy to explain both data sets simultaneously. Our inverted model fits the Love and Rayleigh wave data sets well. We show that the crust in the western and easternmost U.S. is characterized by strong anisotropy, with a clear connection between positive anisotropy and regions of Cenozoic and Mesozoic extension, and between negative anisotropy and orogenesis. The central U.S. crust is mostly isotropic. We find positive upper‐mantle anisotropy across the U.S., with the exception of the Colorado Plateau, and show that it may be linked to diverse geodynamic processes including horizontal mantle flow in the asthenosphere and melt sills in the lithosphere.