Assimilation of extrapolated radar reflectivity into a NWP model and its impact on a precipitation forecast at high resolution

The aim of this paper is to determine whether the assimilation of extrapolated radar reflectivity can improve very short range precipitation forecasts in the summer season. The COSMO NWP model, with a horizontal resolution of 2.8 km, is applied, and radar reflectivity data are assimilated using a water vapour correction method. Two model runs are compared: the first run assimilates observed radar reflectivity, and the second run assimilates observed and extrapolated radar reflectivity. The extrapolation is performed by the COTREC method, and the length of extrapolation is 1 h. Hourly precipitation forecasts are compared and evaluated subjectively, both visually and by the Fractions Skill Score for three events with severe convective precipitation in the model domain. The results of the comparisons indicate that the assimilation of extrapolated radar reflectivity in most cases improves the accuracy of precipitation forecasts for the second and third lead hours. In general, hourly precipitation forecasts by the COSMO NWP model with the assimilation of radar reflectivity show reasonable skill for the first and the second hours.