Validation of severe weather forecasts based on synthetic satellite data as part of an integrated thunderstorm forecast system

Abstract

This paper presents on-going work on a new tool that validates severe weather forecasts based on synthetic satellite data against satellite observations. The algorithm Cb-TRAM is used for tracking and monitoring objects of severe convection both in observed satellite imagery from METEOSAT SEVIRI and in synthetic satellite images forecasted by the German Weather Service (DWD) COSMO-DE model. Objects are defined by the area where convection is detected based on certain thresholds in Cb-TRAM, together with the vertical extent of the cloud cell. The validation tool compares size, displacement, overlap, intensity distribution and history of the observed and forecasted objects and thereby assesses the quality of the forecast. The functioning of the method is demonstrated for case studies of severe thunderstorms. If several forecasts of one object are available (e.g. from an ensemble forecast), this method enables the selection of the forecast that agrees best with the observation and that could then be used to estimate the future thunderstorm evolution. In particular, this method has the potential to close the gap between nowcasting and forecasting. It is part of an integrated thunderstorm forecast system for air traffic named WxFUSION (Weather Forecast User Oriented System Including Object Nowcasting) which is currently under development within the framework of the DLR project “Wetter und Fliegen”, in close collaboration with the DWD.