Extracting subsurface information based on extremely short period of DAS recordings

Abstract

Distributed acoustic sensing (DAS) is a newly developed technology, and it is getting more attention in the energy industry, civil engineering investigations and geo-hazard studies due to its dense sensor spacing, low cost of fiber cables and high repeatability for monitoring. Long-term monitoring by DAS brings us huge amount of data, processing of which is laborious and time consuming. We use the quarry blasts data recorded by the Stanford DAS array to demonstrate that reliable information can be extracted using extremely short recordings. We apply denoising, missing data interpolation, seismic interferometry and dispersion analysis to the DAS data. The average velocity changes along the DAS line are obtained using two segments of 20-second recordings. The average velocity changes are consistent with the nearby basement construction schedule. Using pieces of 100-second recording, the phase velocity changes with frequency have been obtained. The dispersion results are consistent with those obtained using month-long data recordings. These consistent results suggest that reliable information can be extracted using an extremely short period of DAS recordings due to anthropogenic activities. Presentation Date: Monday, September 16, 2019 Session Start Time: 1:50 PM Presentation Time: 4:20 PM Location: 221C Presentation Type: Oral