Tang, Meng

Data assimilation in subsurface systems is challenging due to the large number of flow simulations often required, and by the need to preserve geological realism in the calibrated (posterior) models. In this work we present a deep-learning-based surrogate-modeling framework, ref…