Monitoring changes in subsurface seismic properties caused by heavy rains with a roadside DAS by targeted interferometry

Many cities worldwide are threatened by flooding and sea level rise because of climate change. Early detection of potential threats is essential for safeguarding human lives and housing, as well as critical infrastructures. Continuous monitoring of the subsurface by analysis of surface waves recorded by roadside DAS systems that exploit preexisting telecommunication fibers can provide useful information at a low cost. In urban areas, vehicles transiting on city streets generate large amounts of broad-band (2-25 Hz) surface waves that propagate in the subsurface and can be readily used for continuous monitoring. We designed and successfully tested a targeted interferometry workflow capable of generating high signal-to-noise (SNR) virtual source gathers.

 

The first step in our workflow is to identify the path of vehicles transiting on roads close to the fiber cable. That can be accomplished by following the low-frequency strain signal caused by the quasi-static elastic deformation of the ground. The path is tracked using an algorithm based on Kalman filters. Because long and heavy vehicles generate lower frequency surface waves, we could also perform space deconvolution of the quasi-static signal to estimate the vehicle length and number of axels to improve the signal-to-noise ratio (SNR) at low frequencies. The second step is to perform targeted seismic interferometry in the time-space windows where the surface waves generated by the tracked vehicles are strongest. We found that about two hundred vehicles were sufficient to synthesize high SNR virtual source gathers. The temporal resolution of the measurements is thus of the order of hours, depending on traffic intensity and the SNR required by the specific monitoring application. We used the virtual gathers to generate phase and group velocity profiles. We also measured the amplitude decay as a function of offsets and frequencies to estimate surface-wave attenuations. The analysis of DAS data continuously recorded before, during, and after the heavy rains in California during the 2022-23 winter showed that seismic velocities decreased and attenuation increased as the rains saturated the ground, and they rebounded as the ground dried up. The reliability and high spatial-resolution of the measurements (order of tenth of meters) also enabled us to observe the difference in the seismic response between paved and lawn areas.