Researchers from across Europe gathered in Hamburg from 10–12 March 2026 for a hands-on workshop exploring how deep learning can support climate science. The event was organised by the German Climate Computing Center (DKRZ) as part of the EXPECT project.
The workshop focused on applying deep learning methods to climate data, with an emphasis on building robust and reproducible workflows. Participants worked with real-world datasets and explored applications such as classification, gap-infilling, and downscaling.
Combining theoretical foundations with practical sessions, the workshop brought together scientists with a wide range of expertise, from PhD candidates and Post-Doc to senior researchers. Participants represented several institutions, including the Max Planck Institute for Meteorology, Helmholtz-Zentrum Hereon, German Weather Service, Vrije Universiteit Amsterdam, the University of Leipzig, the University of Hamburg, the University of Bremen, the University of Würzburg, the Alfred Wegener Institute, the Karlsruhe Institute of Technology, the Helmholtz Centre Potsdam, the GFZ German Research Centre for Geosciences, and the Federal Maritime and Hydrographic Agency.
Beyond technical training, the workshop created space for exchange at the intersection of climate science and machine learning, strengthening collaboration within the EXPECT consortium.