Position as PhD student (75% E13 TV-L), Department of Meteorology of Bonn University

RealPEP thrives to achieve significant improvements at all stages along the process chain from Quantitative Precipitation Estimation (QPE), Precipitation Nowcasting (QPN), numerical prediction of quantitative precipitation (QPF) and predicting discharge and potential flash floods in small- to meso-scale catchments (FFP). RealPEP will rely on a multi-sensor data exploitation platform to monitor the precipitation generating atmosphere and tackle urgent science questions to better identify mechanisms that determine the onset, location, intensity, and development of precipitating systems. Developments will be implemented for near-realtime processing in order to be able to mitigate risks to society and ecosystems.

The successful candidate will further develop, in close collaboration with the research unit, a novel precipitation ensemble nowcasting system, which exploits scaling properties of precipitation and precipitation-generating process information derived from polarimetric radars and satellites. A nationwide system based on the spatio-temporal scaling behaviour of precipitation has been implemented already and assessed. In the next step polarimetric radar signatures indicative for potential changes in precipitation generation (e.g. columns of enhanced differential reflectivity, so-called ZDR-columns) will now be exploited for refinements of the approach. A predictive recurrent neural network (PredRNN) will be used to link satellite-based information on cell initiation with their subsequent development monitored by radars.