Extracting Multimodal Surface-Wave Dispersion Curves from Ambient Seismic Noise Using High-Resolution Linear Radon Transform

Abstract

Abstract In the past two decades or so, ambient noise tomography (ANT) has emerged as an established method for imaging subsurface seismic velocity structures. One of the key steps in ANT is to extract surface-wave dispersion curves. The predominant approach for subsurface shear-wave velocity structure inversion involves utilizing fundamental-mode surface waves in ANT. Nevertheless, a notable challenge encountered is the issue of nonuniqueness when employing the dispersion information of fundamental-mode surface waves to invert for shear-wave velocity models. The inclusion of higher-mode dispersion curves in the inversion offers several benefits, including the reduction of nonuniqueness, enhancement of inversion stability, and decreased dependence on the initial model. In this study, we illustrate the applicability of the high-resolution linear radon transform method (HRLRT) for extracting multimodal surface-wave dispersion energy from ambient seismic noise data. We apply HRLRT to both the synthetic noise data and real data recorded by USArray. Our results of applications show that the HRLRT method can extract multimodal surface-wave dispersion information. Compared with established methods such as the frequency–Bessel transform and multicomponent frequency–Bessel transform, the HRLRT exhibits an advantage in suppressing “crossed” artifacts and the second-/third-type artifacts caused by sparse spatial sampling, and the resulting dispersion energy from HRLRT has narrower peaks, meaning high resolution of dispersion curves based on the HRLRT method.