Simultaneous Inversion of Multiwindow Surface Wave Dispersion Data in 2D Layered Media

Abstract

Seismic Surface Wave Methods (SWMs) have been one of the effective tools for retrieving near-surface shear-wave velocity (Vs). However, a critical problem associated with the SWMs is that using a horizontally layered model assumption in inversion may introduce blurred errors into 2D Vs imaging due to the presence of lateral variations. In addition, the increase of investigation depth necessary for the SWMs may cause an increase in errors since a wider acquisition array is essential to acquire longer wavelength data precisely. In this study, we analyzed the relationship between the output power, calculated by an amplitude normalized frequency-domain beamforming and the local phase velocity following the ray-theory approximation. Based on this relationship, we calculated the theoretical dispersion curves of multichannel surface wave data using a local maximum search algorithm. Then, we presented an algorithm for simultaneously inverting surface wave dispersion data extracted from multi-size spatial windows. We parametrized the inverted model in a 2D layered pattern. Finally, we used synthetic and field data sets to test the effectivity of the proposed method. Both results showed that the lateral resolution of the Vs model retrieved by the simultaneous inversion is improved compared to the horizontally layered model-based inversion.