Mixed-Component Ambient Noise Cross-Correlations using Distributed Acoustic Sensing Arrays and Single Point Inertial Seismic Sensors

Abstract:

While ambient noise interferometry using Distributed Acoustic Sensing (DAS) data offers great promise for high-resolution seismic imaging, the intrinsic 2D nature of many dark fiber networks limits the imaging to a 2D plane. This limitation can be overcome by combining existing point inertial sensors like geophones or broadband seismometers, with DAS to provide improved 3D coverage in ambient noise studies. We test the potential of DAS dark fiber deployments combined with local networks of single and three-component seismic sensors for passive 3D imaging using ambient noise interferometry using both synthetic tests and observed data. We evaluate the recovery of Green’s functions from cross-correlation of noise recorded by a DAS array as part of the LBNL FOSSA experiment in the Sacramento basin, California (~22 km dark fiber profile between the cities of Sacramento and Woodland) and nearby permanent seismic stations. We tested a variety of seismic sensors in our study: broadband, short period, borehole short-period sensors, and accelerometers. We recover both Rayleigh waves and Love waves in the noise cross-correlations in the entire distance range of ~3 km to ~95 km in the secondary microseism passband (~0.1-0.4 Hz). However, the signal-to-noise ratio of the recovered signals generally decreases with increasing distance. We compare the surface-wave dispersion measured on the mixed-sensor cross-correlations with that estimated using a broadband sensor deployed near the DAS array. The recovery of meaningful noise cross-correlations between the generally lower quality accelerometers and DAS at distances > 50 km is especially promising because it indicates that the method can be useful in regions with sparse seismic networks.