Review of nonlinear Kalman, ensemble and particle filtering with application to the reservoir history matching problem

Abstract

This chapter reviews the recent advances in applying different Bayesian filtering approaches to realistic assimilation problems in geosciences. We discuss the similarities and differences of these approaches, and present how to build data assimilation schemes that are efficient at reasonable computational costs. We also study the performances of these different filters in an application to the history matching problem in a synthetic, twodimensional,
oil-water reservoir model.