Local subsurface characterization with seismic wavefield gradients

SUMMARY In the context of detecting shallow, localized seismic velocity variations, we assess the sensitivity of wavefield gradients, specifically normal strain and rotation, relative to traditional seismological observables such as displacement, velocity and acceleration. We begin with a simple single-scattering analysis of the displacement wavefield and its gradient, and introduce the Proximity Field Test (PFT) as a straightforward and effective tool for subsurface detection. We then perform 3-D elastic simulations with SEM46, a spectral-element code modified to directly output strain and rotation, and analyse two case studies involving localized shallow velocity changes in a homogeneous medium. In the first case, P- and S-wave velocities and density are varied by 10 per cent relative to the background, whereas in the second case, the local velocity decrease is 70 per cent. By comparing waveforms in the reference medium with those obtained after introducing the heterogeneity, we show that the joint analysis of displacement and gradient measurements enables efficient detection of subsurface anomalies. Finally, we validate the approach with a field experiment combining geophones and DAS measurements, aimed at detecting a buried concrete foundation associated with an approximate 70 per cent positive velocity contrast in the shallow subsurface. The anomaly is clearly identified with minimal processing, demonstrating the practical potential of the proposed method.