Comparisons Between Array Derived Dynamic Strain Rate (ADDS) and Fiber-Optic Distributed Acoustic Sensing (DAS) Strain Rate

Abstract

Distributed acoustic sensing (DAS) strain rate and particle velocity can be compared through approximate scaling with medium velocity. We instead performed a direct comparison between array derived dynamic strain (ADDS) rate and DAS strain rate for six frequency bands. The PoroTomo project at Brady's Hot Springs, Nevada, deployed a 240-geophone 3C array co-located with fiber-optic DAS system and 8.7 km of buried cable. We selected subsets of the geophone array to create four smaller arrays and computed ADDS. The horizontal components of the ADDS were rotated into the direction of the fiber-optic cable and then compared with the observed DAS strain rates. From three example regional earthquakes of local magnitudes 2.9, 4.1, and 4.3, the ADDS are found to be coherent with DAS for frequencies ≤1 Hz. For frequencies >1-Hz, this correlation decays quickly. Small differences between linear and areal dynamic strains at 1-Hz suggest poor signal-to-noise or localized strain that is perturbed by shallow heterogeneities compare to the average strain propagating across the geophone array. The implication is that around 1-Hz, straight fiber DAS is measuring axial strain along the fiber and can provide good approximations to translational particle motions. However, above 1-Hz, DAS becomes more sensitive to shallow velocity gradients that can be beneficial for geophysical imaging yet becomes a limitation for traditional seismic analysis methods depending on absolute amplitude and phase from translational particle motions.