Surface-wave dispersion spectrum inversion method applied to Love and Rayleigh waves recorded by distributed acoustic sensing

Abstract

Recently, distributed acoustic sensing (DAS) has been applied to shallow seismic structure imaging providing dense spatial sampling at a relatively low cost. DAS on a standard straight fiber-optic cable mostly records axial dynamic strain, which makes it difficult to separate the Rayleigh and Love wavefields. As a result, the mixed Rayleigh and Love wave signals cannot be used in the conventional surface-wave dispersion inversion method. Therefore, it is often ensured that the source and the cable are in the same line and only Rayleigh wave dispersion is used, which limits the constraints on structure and model resolution. We have inverted surface-wave dispersion spectra instead of dispersion curves. This inversion method can use mixed Rayleigh and Love waves recorded when the source and receiver array are not aligned. The multiple-channel records are transformed to the frequency domain, and a slant stack method is used to construct the dispersion spectra. The genetic algorithm method is used to obtain an optimal S-wave velocity model that minimizes the difference between theoretical and observed dispersion spectra. A series of synthetic tests are conducted to validate our method. The results suggest that our method not only improves the flexibility of the acquisition system design, but the Love wave data also provide additional constraints on the structure. Our method is applied to the active source and ambient noise data sets acquired at a geothermal site and provides consistent results for different data sets and acquisition geometries. The sensitivity of the dispersion spectra to layer thickness, density, and P-wave velocity is also discussed. With our method, the amount of usable data can be increased, helping deliver better subsurface images.