Adaptive tuning of background error and satellite radiances observation error for operational variational assimilation

This paper proposed a new implementation of error tuning algorithm based on the posterior diagnosis of variational assimilation system by applying a iterative fixed-point procedure. The method is shown to be feasible in a practical tuning of background and observation error for operational variational assimilation. The iteration process converges very rapidly and the results are reasonable .The results are illustrated by utilizing the Chinese Meteorological Administration (CMA) new generation of NWP system GRAPES (Global/Regional Assimilation PrEdiction System) regional-meso version over the East Asia domain. The impacts of the tuning of background and satellite radiances observation error on analyses and forecasts are shown and discussed. The results are very promising.