Statistical-dynamical methods for scale dependent model evaluation and short term precipitation forecasting (STAMPF)

 

Application PHASE 3

Summary

 

The object of the project is a scale dependent evaluation of precipitation forecasts of the DWD weather prediction models (COSMO-DE/EU) in relation to dynamical and statistical parameters as well as cloud properties. For different IOP episodes the Dynamic State Index and a newly developed precipitation activity index of clouds will be used and compared with classical parameters like CAPE and vertical velocities. Theoretical concepts like IPV-Thinking and turbulent cascades of energy and enstrophy will be used for explaining the narrow filament-like rainbands and to identify physical processes responsible for deficiencies in QPF. For evaluation the model-independent FUB-analyses of 7, 2.8 and 1 km will be further improved by assembling the unique data sets of GOP and COPS field campaign. The analysis scheme will combine cloud statistics from Meteosat-8, disdrometer and radar information. Special emphasise is placed to the comparison of lowland  and orographic induced precipitation with respect to the separation of stratiform and convective rainfall. Here, statistical evaluation parameters of precipitation forecasts are scaling exponents of distribution functions and information entropy. The novel precipitation activity index of clouds will be tested with observed and modelled precipitation amounts in order to attribute a precipitation probability to phenomenological cloud types. This approach could be applied for data assimilation techniques.