Multisource Time-Lapse Elastic Full-Waveform Inversion Using a Target-Oriented Common-Model Strategy

Full-waveform inversion (FWI) is a powerful tool for time-lapse seismic analysis, enabling high-resolution imaging of subsurface physical properties to monitor reservoir changes during injection, production, and long-term CO2 storage. However, conventional time-lapse FWI, which relies on a parallel inversion strategy, suffers from significant artifacts due to survey non-repeatability, disrupting convergence consistency between baseline and monitor inversions. Additionally, the high computational cost remains a major challenge. To address these limitations, we propose a novel time-lapse FWI strategy—the target-oriented (TO) common-model strategy (CMS)—which strategically integrates multiple approaches. Our method combines TO FWI, which enhances model convergence in the target region to improve time-lapse accuracy, with CMS, which reduces artifacts by using an optimized starting model to guide baseline and monitor inversions toward similar convergence paths. Additionally, we employ an amplitude-encoding multi-source strategy, significantly reducing computational costs without compromising inversion accuracy. Through extensive elastic tests, we validate the robustness and effectiveness of TO CMS, demonstrating superior performance over both the conventional parallel strategy and standard CMS across various challenging scenarios—including non-repeated source positions, random noise, seawater velocity variations, and biased initial models. Notably, strong noise and seawater velocity variations can significantly impact time-lapse FWI results, highlighting the need for further investigation. Ensuring consistent multi-source parameters in time-lapse FWI can help minimize artifacts.