Tracking and nowcasting of convective cells using remote sensing data from radar and satellite

Rad-TRAM, a new tracking and nowcasting algorithm based on radar data, has been developed recently. Rad-TRAM uses the same tracking algorithm as the cloud tracker Cb-TRAM which is based on a pyramidal image matcher taking into account the spatial scale of detected cells (Zinner et al., 2006). Using the European radar composite of the German Weather Service (DWD) cells of high reflectivity representing regions with heavy precipitation and hail are identified using a threshold criterion. These are analysed in parallel to convective cells detected by Cb-TRAM for three case studies where a multitude of thunderstorm tracks has been detected over Europe. The results from both tracking algorithms have been evaluated visually as well as statistically. The statistical analyses of the thunderstorm tracks focused on the spatial extent and time duration of the convective systems detected by the two algorithms. Spatial overlap between cells detected by Rad-TRAM and cells representing mature severe convection detected by Cb-TRAM is found in more than half of all investigated cases. This is a realistic result because the cells detected by Rad-Tram also include precipitation patterns with non-convective origin due to the threshold criterion of 37 dBZ while Cb-TRAM only considers convective clouds. In addition to this analysis the quality of short range forecasts of heavy precipitating cells up to one hour provided by Rad-TRAM have been investigated. It is shown that these forecasts perform on average better than corresponding forecasts based on persistence. Overall, the results from this study suggest that a combined use of both tracking systems for clouds as well as for precipitation cores provides a more complete figure of convective cells at hand and might be useful for operational nowcasting systems in real time conditions.