On a recent development in stochastic inversion with applications to hydrogeology

Abstract

We comment on a recent approach to spatial stochastic inversion, which centers on a concept known as "anchors" and conducts nonparametric estimation of the likelihood of the anchors (along with other model parameters) with respect to data obtained from field processes. The approach is called "anchored inversion" or, less desirably, "method of anchored distribution". Conceptual and technical observations are made regarding the development, interpretation, and use of this approach. We also point out that this approach has been superseded by new developments.