Effects of an assimilation of radar and satellite data on a very-short range forecast of heavy convective rainfalls

An assimilation method of radar reflectivity and satellite data into a NWP model COSMO with a horizontal resolution of 2.8 km and its impact on a very short range forecast of precipitation are presented. The assimilation method consists in a correction of model water vapour mixing ratio using the nudging technique. The correction depends on: (i) the difference between model and observed precipitation, which is derived from radar reflectivity, (ii) whether observed clouds are classified as precipitating. The cloud classification utilises two channels (10.8 and 6.2 µm) from the Meteosat Second Generation (MSG), which are also used to estimate precipitation. Two types of corrections are examined with respect to places at which the correction is applied. First approach performs the correction at the same grid point where the difference is calculated. Second approach performs the correction at a grid point advected upwind from the point where the difference is calculated. sts of three convective events with heavy local precipitation were evaluated subjectively (by eye) and by applying the fraction skill score. The results show that the assimilation of radar information or combined radar and MSG data significantly improves precipitation forecasts for at least three forecast hours. Despite uncertainties in the relationship between MSG data and precipitation, the assimilation method that combines radar and MSG data shows comparable or better precipitation forecasts than when only radar reflectivity is used. The advantage of the approach transforming MSG measurements into precipitation is that the same assimilation technique can be applied to both radar reflectivity and satellite data.