Key points are not available for this paper at this time.
In an earlier study, Bensen et al. measured surface wave dispersion curves from ambient noise using 203 stations across North America, which resulted in Rayleigh and Love wave dispersion maps from 8–70 s period and 8–20 s period, respectively. We invert these maps in a two-step procedure to determine a 3-D shear wave velocity model (VS) of the crust and uppermost mantle beneath much of the contiguous US. The two steps are a linearized inversion for a best fitting model beneath each grid node, followed by a Monte Carlo inversion to estimate model uncertainties. In general, a simple model parametrization is sufficient to achieve acceptable data fit, but a Rayleigh/Love discrepancy at periods from 10 to 20 s is observed, in which simple isotropic models systematically misfit Rayleigh and Love waves in some regions. Crustal features observed in the model include sedimentary basins such as the Anadarko, Green River, Williston Basins as well as California's Great Valley and the Mississippi Embayment. The east—west velocity dichotomy between the stable eastern US and the tectonically deformed western US is shown to be abrupt in the crust and uppermost mantle, but is not coincident in these regions; crustal high velocity material tends to lap over the high velocities of the uppermost mantle. The Rayleigh/Love discrepancy between 10 and 20 s period is crustal in origin and is observed in a number of regions, particularly in extensional provinces such as the Basin and Range. It can be resolved by introducing radial anisotropy in the lower or middle crust with VSH > VSH by about 1 per cent.
Bensen et al. (Mon,) studied this question.