|Title||Impact assessment of GOES-15 CSR and Meteosat-10 ASR in the ECMWF system|
|Series/Collection||EUMETSAT/ECMWF Fellowship Programme Research Reports|
|Authors||Lupu, C, McNally, A|
|Event Series/Collection||EUMETSAT/ECMWF Fellowship Programme|
|Place of publication||Shinfield Park, Reading|
This report summarises activities related to the evolution of the geostationary network of clear and cloudy radiances assimilated in the ECMWF operational suite and the main results from research work carried out to study the growth rate of the model humidity errors within the 4D-Var window. Following the replacement of GOES-11 by GOES-15 and the replacement of Meteosat-9 by Meteosat-10 work has been directed towards assimilating CSR from GOES-15 and ASR from Meteosat-10 in the operational ECMWF system. Monitoring experiments have shown that GOES-15 CSR data are stable, with a small bias which can be globally corrected. The biases in the ASR from Meteosat-10 are lower than those from Meteosat-9 and in other respects the quality is the same. The impact of geostationary radiances on headline forecast measures remains rather small. The feasibility of extracting information about the model error from SEVIRI observations has been investigated in assimilation experiments. The analysis of rejected observations as a function of the position of the observation in the assimilation window has indicated that both the background check and the departure in the window channel criteria rejects more data towards the end of the assimilation window. Initial trials using a relaxed background check allows the use of more observations that were previously rejected, but did not improve the standard deviation of first guess departures of active assimilated SEVIRI observations.