Assimilating infrared brightness temperatures using an ensemble data assimilation system and a cloud-dependent bias correction scheme

This study presents results from data assimilation experiments in which clear and cloudy sky infrared brightness temperatures from the SEVIRI sensor onboard the Meteosat Second Generation (MSG) geosynchronous satellite were assimilated at convection-resolving horizontal resolutions. The infrared satellite observations were assimilated in the regional-scale COSMO model using the Kilometer-scale Ensemble Data Assimilation (KENDA) system being developed at the German Deutscher Wetterdienst (DWD). Cloud-dependent bias correction values were obtained for the clear sky grid points, and for low, mid, and high clouds, using observation-background (OMB) bias errors from a 3.5-day passive monitoring period prior to the start of the assimilation experiments. Results from a 12-hour assimilation period with hourly assimilation of the infrared observations showed that the cloud height dependent bias correction scheme was able to successfully remove the OMB bias from most cloud types and that the infrared observations led to large improvements in the cloud and moisture fields and neutral impact on the temperature and wind fields. The monitoring and assimilation results also indicate that a cloud-dependent bias correction scheme is necessary when assimilating clear and cloudy sky infrared brightness temperatures.