Development of a deep learning prototype for operational probabilistic wind gust forecasting (T04# (T03))

• Subject Area: Atmospheric Science
• Term: since 2023
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 257899354

Postprocessing ensemble forecasts from numerical weather prediction to correct systematic errors has become standard practice in research and operations. In the past years, a key focal point in statistical post-processing has been the use of modern methods from machine learning. Despite promising results in case studies, hardly any of the novel methods have been used in an operational context thus far. The overarching aim of this Transfer Project is to combine expertise and experience from KNMI and W2W to help overcome remaining methodological and practical challenges for a successful transfer from research to operations. With a focus on improving probabilistic forecasts of severe wind gusts, we will develop a deep learning-based postprocessing model tailored to the specific needs of operational weather prediction at KNMI which will provide probabilistic predictions of threshold exceedances at relevant warning levels and arbitrarily defined regions. The developed model will be implemented in a prototype forecast product at KNMI, where forecasters will be trained to use it during their shifts in a test phase.

• DFG Programme: CRC/Transregios (Transfer Project)
• Subproject of: TRR 165: Waves to Weather
• International Connection: Netherlands
• Applicant Institution: Ludwig-Maximilians-Universität München
• Business and Industry: Royal Netherlands Meteorological Institute
• Project Head: Dr. Sebastian Lerch