Characterizing Fractured Zones in Urban Karst Geology Using Leaky Surface Waves From Distributed Acoustic Sensing

Urban karst geology poses significant geohazard risks, most notably sinkholes and surface depression stemming from soluble and fractured bedrock that is prone to dissolution and collapse. However, mapping and characterizing these hazards using traditional geophysical surveys in cities is challenging due to dense infrastructure and high levels of human activity. In this work, we demonstrate how distributed acoustic sensing (DAS), deployed via preexisting telecommunication fiber-optic cables, can be leveraged to detect fractured weak zones in a populated setting. By recording traffic noises, we are able to conduct large-scale, cost-effective, and minimally intrusive subsurface investigations. Our workflow integrates ambient noise interferometry with advanced signal enhancement techniques, specifically frequency-wavenumber (F-K) filtering and bin-offset stacking. F-K filtering isolates wavefields traveling in opposite directions to suppress localized noise, while bin-offset stacking further enhances signal coherency by superposing channels with common offsets. The resulting Noise Cross-correlation Functions exhibit unique inverse-dispersion patterns that signify the presence of leaky surface waves generated by a low-velocity half-space. We invert the corresponding dispersion curves to derive a 2D S-wave velocity model, highlighting a prominent low-velocity anomaly indicative of a fractured zone. To confirm the karstic nature of this anomaly, rock physics modeling is employed to estimate spatial variations in fracture density, revealing marked heterogeneity in the fractured zone. Our findings underscore the power of DAS-based ambient noise interferometry for delineating karst features and diagnosing potential sinkhole risks in urban environments. By exploiting widely available fiber-optic networks, this approach significantly broadens the practicality of near-surface geohazard mapping at the city scale.