High-resolution directional passive surface wave dispersion imaging based on multiple signal classification

Passive surface wave imaging techniques have been widely used in near-surface geophysics due to their low cost, convenience, and non-intrusive. However, with the existence of strong directional noise generated by earthquakes, vehicles, ships, and other human activities, there is a large inhomogeneity in the spatial distribution of the ambient noise, which leads to an overestimation of the phase velocity of the surface waves. In addition, restrictions such as experimental site regulation do not allow for long time noise acquisition. We propose a velocity-multiple signal classification (V-MUSIC) method to analyze the surface wave phase velocity and azimuth in short noise segments. It is based on the principle of orthogonality between the surface wave signal and the noise. Synthesis tests and field examples have demonstrated that V-MUSIC provides significant resolution improvements in both azimuth estimation and dispersion imaging, enabling the acquisition of high-resolution, unbiased dispersion images in ultra-short records.