Satellite-based thunderstorm tracking, monitoring and nowcasting over South Africa

Kurzfassung

In this study DLR’s cloud tracker Cb-TRAM (Cumulonimbus TRAcking and Monitoring) is adapted, applied and validated over South Africa for the first time. Cb-TRAM is a fully automated algorithm to detect, track, and nowcast thunderstorms based on data from the METEOSAT SEVIRI (Spinning Enhanced Visible and Infra-Red Imager) instrument. The high resolution visible (HRV), a water vapour (WV 6.2 μm), and two infra-red (IR 10.8 μm and IR 12.0 μm) channels are combined in order to determine three different development stages: on-set of convection, i.e. first development of growing cumulus clouds, (2) rapid development, i.e. strong cooling at cloud top, and (3) mature stage, i.e. thunderstorms reaching or overshooting tropopause levels. During daytime the HRV is used to detect the most active convective updraft regions by exploring the visible texture in the image. During nighttime or in regions where the HRV is not available the texture in the WV 6.2 μm is used to isolate convective turbulence. The tracking is based on the geographical overlap between detected cells and their first guess patterns derived from preceeding time steps. Both, the first guess patterns as well as nowcasts (very short range forecasts) up to 60 minutes are based on a pyramidal image matching algorithm which provides fields of displacement vectors accounting for differential cloud motion. Over South Africa, it was necessary to adapt Cb-TRAM to the HRV sector of the Southern Hemisphere which moves with the position of the sun. In several case studies, the Cb-TRAM detected and nowcast thunderstorm cells will be compared to radar and lightning observations. The performance of Cb-TRAM over South Africa will be demonstrated, and the potential of Cb-TRAM to provide early warnings of thunderstorms and related damage in this region will be dicussed.