This regularized inversion technique using ML-prior (Li et al., 2021) combines the seismic and well log data, to help improve the resolution and accuracy of time-lapse elastic waveform inversion (TL-EFWI) in the target zone. To derive the prior model, we train a deep neural network (DNN) to learn the mapping relation between the seismic estimation and the facies interpreted from well logs. We then apply the trained network to the target inversion domain to predict a prior model. We then perform another time-lapse inversion in which we fit the simulated data difference to the redatumed data difference, as well as, fit the model changes to the prior model.
Applying a target-oriented strategy through elastic redatuming to reduce the computational cost by focusing our inversion on a target zone.
Incorporating well information to the time-lapse EFWI by regularization. The required prior model is predicted from a trained deep neural network (DNN), which identifies the mapping relation between an initial inverted model and facies interpreted from well logs.