Near-Surface Shear-Wave Velocity from Train Vibrations with Distributed Acoustic Sensing
Abstract
Seismic velocity models are effective for imaging subsurface structures, yet resolving shallow features at depths less than 100 m remains challenging due to the limited availability of high-frequency signals and the need for closely spaced channels. We addressed this using distributed acoustic sensing (DAS) to image the near-surface seismic structure at high resolution, using train traffic recorded along a 14.7 km buried fiber-optic cable in Del Mar, California. The study area includes unstable coastal cliffs, where subsurface conditions vary due to differences in geologic composition and ongoing erosion. Using interferometry on train-passing recordings, we computed cross-correlation functions, estimated Rayleigh-wave phase velocities through beamforming, and inverted for 1D shear-wave velocity (VS) at each channel, which were then assembled into a 2D tomographic VS profile along the fiber. The resulting 2D VS profile reveals key geologic features, including low-velocity zones in the San Dieguito Valley and Los Peñasquitos Lagoon, corresponding to thicker alluvial and lagoonal sediments. The average VS in the upper 30 m depth (VS30) shows a strong correlation with elevation, and the low VS30 at the coastal cliff is associated with sand content and sediment thickness. In addition, diurnal strain patterns recorded by DAS indicate sensitivity to temperature variations and may help identify fiber spool locations. The VS model demonstrates the potential of DAS to provide high-resolution, cost-effective subsurface imaging and environmental monitoring in complex coastal settings.