A moveout-based method for the detection of weak seismic events using downhole DAS arrays

Distributed acoustic sensing offers the possibility of high-resolution passive monitoring of seismic events. Local coherency of the recorded wavefields allows for array-based processing techniques. In this study, we apply such techniques in a downhole array at the San Andreas Fault Observatory at Depth. We begin by estimating the seismic velocity along the well using both earthquake records and ambient noise interferometry. Results are compared to a conventional near-offset VSP geophone survey. DAS earthquake records also allow for S-wave velocity estimation. We utilize the computed velocity in a moveout-based detection algorithm. The method takes into account velocity along the array as well as the range of possible incidence angles of incoming events. We apply it to 20 days of data, recorded in June-July 2017. Above 70% of catalogued events in the area are detected, and a new, weak, uncatalogued event is discovered. These results set the path for DAS monitoring of natural and induced events.