Earthquake Source Parameter Estimation Using Distributed Acoustic Sensing and Frequency Wavenumber Scaling

Distributed acoustic sensing (DAS) holds great promise for seismic moment and stress-drop estimation owing to its dense spatial sampling that fosters advanced array processing techniques and the ability to average source parameter estimates along a sensing optical fiber. The main caveat in this application lies in the measurand: Although source parameter estimation requires ground motions, DAS measures strain, and data conversion is usually required. In this study, I use a strain rate to ground acceleration conversion approach in the frequency–wavenumber domain and show that it can be directly used to obtain acceleration amplitude spectra (AS). This approach is found to be equivalent to spatial integration without a colocated seismometer. The approach is applied to 44 earthquakes recorded by an optical fiber in Israel. Converted acceleration AS were calculated using short-fiber segments and fitted with a source model to estimate source parameters. Within-event parameter variabilities are found to be similar for DAS and accelerometer-derived source parameters. DAS-derived magnitudes and stress drops are slightly higher than accelerometer-derived parameters, with average DAS and accelerometer stress drops of 16.1 and 4.1 MPa, respectively. Stress drops appear to increase with seismic moment, probably due to the limited frequency range of the source parameter inversion. The results demonstrate the great potential of DAS for source studies.