Considering the impact of observation error correlation in ensemble square-root Kalman filter

Covariance localization in the ensemble transform Kalman filter based on an augmented ensemble

“…In ensemble-based data assimilation, observation error and forecast error must be resolved. For the study of observation error, please see the article by Zang and Wang (2019). The statistical accuracy of forecast error is extremely important for any ensemble data assimilation scheme, since the forecast error covariance matrix (the f P matrix, hereinafter) is often sufficiently accurate and high-rank covariance matrices is prohibitive.…”