Adaptive and accuracy-aware multiple data assimilation in a three step framework

This paper presents a three-step framework that unifies various adaptive variants of the ensemble smoother with multiple data assimilation (ES-MDA) method. ES-MDA is an algorithmic approach for solving inverse problems in reservoir engineering and related fields by gradually transitioning a prior ensemble to a posterior ensemble. The framework demonstrates how different implementation strategies of the multiple data assimilation process affect both accuracy and computational cost, providing a systematic understanding of existing methods.