Abstract
In this paper, sequential importance sampling is used to assess the impact of observations on an ensemble prediction for the decadal path transitions of the Kuroshio Extension. This particle-filtering approach gives access to the probability density of the state vector, which allows the predictive power—an entropy-based measure—of the ensemble prediction
... read more