Spatio-Temporal Subsurface Variations in a Marsh Site Based on DAS Noise Interferometry

There is a growing interest in applying ambient noise processing techniques to fiber-optic arrays, which are now a mainstream tool in seismology. Traditionally applied to geophone arrays, noise-based interferometry methods have been widely used for subsurface imaging and monitoring over the last two decades. In this study, we retrieve surface wave phase velocities by cross-correlation of DAS-recorded ambient noise data to monitor the subsurface of a quiet brackish marsh site of the Loire estuary. Seasonal recordings were obtained from a DAS array consisting of multiple linear and a spiral cable layout, capturing the ambient seismic wavefield mainly influenced by natural forcing, for example, tidal activity in the river and streams (below 15 Hz), but also by anthropogenic sources. We use time-frequency weighted Phase Cross-Correlation (PCC) technique, which in addition to being efficient, reduces sensitivity to amplitude variations and emphasizes phase coherence. We observe diurnal and seasonal variation in ground water level and in noise characteristics, like amplitude and directionality. Such temporal variations provide an opportunity to monitor both (a) the changes in the subsurface medium itself, and (b) the impact of noise characteristics on surface wave retrieval. We also observe a pronounced lateral contrast of surface wave phase velocity across the marsh site, highlighting the extent of spatial variability of the subsurface in complex natural environments.