A fast data-matching approach for Snowflake DAS VSP datasets at CaMI

He Liu, Xin Fu, Xiaohui Cai, Daniel O. Trad, Kristopher A. Innanen

Vertical seismic profile (VSP) is a powerful technology for monitoring CO₂ injection and sequestration. Applying full-waveform inversion (FWI) to VSP data can provide images and time-lapse models of physical properties in the reservoir where CO₂ is injected. The biggest challenge in the application of FWI to time-lapse datasets is data matching. In the Snowflake DAS VSP datasets acquired at CaMI in 2018 and 2022, we observe no time-lapse data difference by subtracting two preprocessed DAS VSP gathers with well-repeated acquisition geometry, which results in a failure of imaging time-lapse physical parameter change. To deal with this problem, we propose a fast data-matching approach, aiming for a successful time-lapse FWI using DAS VSP data.