Full text loading...
This study presents a joint 4D seismic amplitude and time-shift (JAT) inversion framework designed to enhance time-lapse seismic time-shift estimation for monitoring of CO2 storage in depleted gas reservoirs. Conventional inversion workflows suffer from error propagation due to inaccurate time-shift estimation, particularly in the presence of wavelet interference and strong impedance contrasts. The proposed JAT method addresses this by simultaneously inverting 4D seismic amplitude and time-shift data, constrained by a well-log calibrated petro-elastic model (PEM), to improve physical consistency and reduce artefacts. The inversion cost function couples amplitude and kinematic changes through the PEM coefficient, ensuring that velocity changes are geologically plausible and aligned with observed seismic responses.
Synthetic case studies demonstrate a substantial improvement in inversion accuracy: time-shift and velocity change NRMSEs are reduced to 7.7% and 7.5%, respectively, compared to 14.0% and 11.3% from traditional nonlinear inversion (NLI). The JAT method effectively suppresses spurious time-shifts and amplitude artefacts, especially near the top and base of the reservoir. Closed-loop validation confirms the internal consistency of the approach. These improvements translate into more reliable monitoring of plume migration and reservoir integrity. Preliminary results from a North Sea case study further validate the method’s practical applicability and uplift in interpretational reliability.