Jean-François Caron
Environment and Climate Change Canada
Yann Michel;
Thibaut Montmerle;
Étienne Arbogast
Centre National de Recherches Météorologiques (CNRM) / Météo-France
The 8th EnKF Data Assimilation Workshop
Improving Background-Error Covariances in a 3D Ensemble-Variational Data Assimilation System for Convective Scale NWP
Talk:
CaronJF_ImprovingBens.pptx
Following the recent development of a 3D Ensemble-Variational (3D-EnVar) data assimilation algorithm for the convective scale AROME-France NWP system, this study examined different approaches to improve the ensemble-derived background-error covariances in this new data assimilation scheme without modifying the ensemble of background generation strategy. Two variants of scale-dependent localization method that consist to apply appropriate (i.e. different) amount of localization to different ranges of background-error covariance spatial scales while simultaneously assimilating all of the available observations were examined and compared. Increasing the effective ensemble size in the representation of the background-error covariance from time-lagged members was also considered both on its own and with scale-dependent localization. The results from data assimilation cycles over a one month period showed that the scale-dependent localization approach of Buehner (2012), that include spectral localization which assumes that the covariance between the scales are zero, performed better than the more recent formulation of Buehner and Shlyaeva (2015), that avoids the complete removal of the between-scale covariances imposed in the former formulation, when the background-error covariances are derived from the most recent 25-member ensemble forecasts. Increasing the effective ensemble size to 75 members with time-lagged forecasts outperformed scale-dependent localization on its own. The largest forecast improvements were obtained when combining the two approaches.
Following the recent development of a 3D Ensemble-Variational (3D-EnVar) data assimilation algorithm for the convective scale AROME-France NWP system, this study examined different approaches to improve the ensemble-derived background-error covariances in this new data assimilation scheme without modifying the ensemble of background generation strategy. Two variants of scale-dependent localization method that consist to apply appropriate (i.e. different) amount of localization to different ranges of background-error covariance spatial scales while simultaneously assimilating all of the available observations were examined and compared. Increasing the effective ensemble size in the representation of the background-error covariance from time-lagged members was also considered both on its own and with scale-dependent localization. The results from data assimilation cycles over a one month period showed that the scale-dependent localization approach of Buehner (2012), that include spectral localization which assumes that the covariance between the scales are zero, performed better than the more recent formulation of Buehner and Shlyaeva (2015), that avoids the complete removal of the between-scale covariances imposed in the former formulation, when the background-error covariances are derived from the most recent 25-member ensemble forecasts. Increasing the effective ensemble size to 75 members with time-lagged forecasts outperformed scale-dependent localization on its own. The largest forecast improvements were obtained when combining the two approaches.
Thanks to McGill's Atmospheric and Oceanic Sciences department for hosting this web-site.