Near-surface S-wave velocity and attenuation structure from full-waveform inversion of distributed acoustic sensing data

Abstract

Distributed acoustic sensing (DAS) technology is rapidly developing for seismic applications because of advantages such as dense spatial sampling, low-cost, and easy installation. At the Field Research Station of the Containment and Monitoring Institute (CaMI.FRS) in Newell County, Alberta, Canada, seismic field experiments with DAS were carried out for carbon capture and storage monitoring. Compared to geophone measurements, the DAS-recorded surface-waves do not suffer from spatial aliasing and contain lower frequencies. Fullwaveform inversion (FWI) was applied to image the near-surface Swave velocity (VS) and attenuation (quality factor QS) structures using the DAS-recorded surface-waves. Compared to traditional surfacewave analysis, we show that FWI intrinsically incorporates fundamental and high-order modes and provides high-resolution VS model that resolves lateral variations. The low-frequencies in DAS data help overcome the cycle skipping problem of FWI. Then, a root-mean-square amplitude difference misfit function is used to recover the near-surface QS model. Strong QS anomaly is resolved at the near-surface. With the inverted VS and QS models, the synthetic data match with the observed data closely in both amplitude and phase.