Improving rainfall estimation across Europe through advancements in radar data processing, use of crowdsourced data, and AI

• Applicants: Dr. Christian Chwala; Dr. Jochen Seidel
• Subject Area: Hydrogeology, Hydrology, Limnology, Urban Water Management, Water Chemistry, Integrated Water Resources Management ; Atmospheric Science
• Term: since 2026
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 575891624

Accurate rainfall measurements are essential for predicting floods, managing water resources, and understanding Earth’s water cycle. In Europe, weather radar networks provide good coverage and frequent updates, making them vital for monitoring rainfall. Yet, current products still struggle with accuracy. EuRadCA brings together researchers from Germany and the Netherlands to improve rainfall estimates across Europe. The project will develop new methods for combining radar data with information from rain gauges and personal weather stations, using advanced techniques like deep learning and geostatistics. This proposal is timely for three key reasons: A large archive of European radar data from the OPERA programme will soon be publicly available for the first time. Access to personal weather station data is rapidly improving. And for the first time, advances in AI and computing make it possible to train deep neural networks on radar data at a continental scale in Europe. EuRadCA focuses on four key areas. It will improve radar accuracy by developing new quality control and bias adjustment methods that work directly in the radar’s native spherical coordinate system. It will produce pixel-level quality maps to track uncertainty at each step, leading to more useful, probabilistic rainfall estimates. The project will train deep learning models on both simulated and real rainfall data to overcome the limitations of current radar-rainfall retrieval algorithms and produce more accurate radar-only and radar-gauge adjusted rainfall products. It will also develop novel ways to identify and filter suspicious data from personal weather stations, enhancing accuracy during extreme events. Finally, EuRadCA will evaluate how well these new methods perform in intense rainfall, measuring improvements in bias and uncertainty compared to existing products on the European scale. Over three years, EuRadCA will deliver high quality reproducible, open-source rainfall products and codes for both public and research use. These products will support meteorological services, flood risk management, and climate adaptation efforts. The project team includes experts from Karlsruhe Institute of Technology (KIT), University of Stuttgart, and TU Delft, working closely with national weather services DWD (Germany) and KNMI (Netherlands) to ensure practical relevance and easy access to data and computational resources.

• DFG Programme: Research Grants
• International Connection: Netherlands
• Partner Organisation: Nederlandse Organisatie voor Wetenschappelijk Onderzoek (NWO)
• Co-Investigator: Dr. Julius Polz
• Cooperation Partner: Professor Dr. Marc Schleiss