Comparisons between measurements and predictions of Rayleigh wave amplification across the contiguous United States

Abstract

We present local surface-wave-amplification maps spanning the contiguous United States for Rayleigh waves between 35 s and 125 s using data recorded on the USArray between 2006 and 2015. We isolate the effect of local structure from those of the earthquake and propagation using a previously developed method based on ratios of amplitudes measured at adjacent stations. To assess the ability of our technique to resolve surface-wave amplification, we perform a parallel synthetic-tomography experiment. We determine amplification from a large dataset of SPECFEM synthetic seismograms calculated for USArray stations and compare the measurements with direct predictions of local amplification. Correlations between synthetic and predicted amplification are high, with correlation coefficients ranging from 0.77 at 40 s to 0.95 at 100 s, indicating that we are able to resolve well local variations in amplification, particularly at long periods. The remaining differences between synthetic and directly predicted amplification maps suggest the influence of finite-frequency effects on surface-wave amplitudes. Observed Rayleigh wave amplification factors reflect local elastic structure in the region surrounding each station and could be used as a complementary constraint to image the structure of the Earth's crust and upper mantle.