Single-source rapid-repeat time-lapse elastic FWI based on the DAS-VSP data

Xiaohui Cai, Kristopher A. Innanen, Donald C. Lawton

Last year, to track the transient changing levels of CO₂, we employed elastic full waveform inversion (EFWI) to identify anomalies in single-source rapid-repeat time-lapse VSP data. The time-lapse EFWI inversion outcomes, relying on differences in data norms, have demonstrated their ability to identify and assess changes in time-lapse data caused by the injection of CO₂. However, the data norm difference response only in the geophone data instead of the distributed acoustic sensing (DAS) data. In theory, the injection of CO₂ into the formation is expected to change the elastic parameters, resulting in noticeable changes in the difference between the shot gathers before and after the injection. This change typically manifests as distinct upgoing wavefield data. Therefore, we develop an inversion strategy that relies on the difference of upgoing DAS data This strategy aims to determine whether it can effectively detect transient changes based on DAS data during the injection of CO₂.