This chapterLocalization will introduce an adaptive Localizationadaptive correlation-basedLocalizationcorrelation based localization method used in the experiments in Part III. The technique is a straightforward implementation of a local analysisLocal analysis scheme where we update the parameters, or rows in the ensemble state-matrix, sequentially one by one using only the significantly correlated measurements in the update. Such adaptive localization has advantages over distance-basedLocalizationdistance based localization, particularly when conditioning the model on non-local measurements and when the localization distance is unknown. Another benefit is that an adaptive correlation-based localization does not refer to the spatial location of the updated variables or measurements. Thus, we only need the ensemble matrices of the state and the predicted measurements to determine the localization. We will present a theoretical foundation for selecting the truncation value as a function of the ensemble size and discuss some possible remedies for further incremental improvement of the method.