3D-Gain
Realistic 3D Atmospheric Reconstruction for Generative AI Nowcasting of Precipitation and Irradiance using Remote Sensing and In-Situ Data
3D-Gain develops a unified, AI-driven model for realistic three-dimensional reconstruction of the atmosphere. By integrating multi-modal observations such as satellite imagery, weather radar, all-sky cameras and in-situ data, the project creates physically consistent 3D representations of atmospheric processes.
An encode–diffuse–decode framework combined with self-supervised and physics-informed learning addresses the lack of 3D ground truth while quantifying reconstruction uncertainty. The resulting models enable simultaneous nowcasting of precipitation and solar irradiance, supporting applications in renewable energy integration and flood forecasting.
Other projects
GRIDMARK – Generating Reproducible Insights through Data Benchmarking for AI in Energy Systems
Transforming energy systems toward climate neutrality: Distribution grids have the potential to be catalysts for the energy transition. Unfortunately, most Distribution System Operators lack the resources to fully monitor their systems. Therefore, there is an urgent need for more high-quality data, particularly to develop and test machine learning models.
GROOVY
HiGh ContRast DichrOic ReflectiOn EUV MicroscopY
A new table-top EUV reflection microscope enables lens-less, polarization-sensitive nanoscale imaging, revealing surfaces and magnetic structures with extreme resolution
BenthicAI
Illuminating invisible life in the Wadden Sea
Underwater cameras, sonar and AI detect burrowing animals from seafloor traces, enabling non-invasive mapping of marine species and habitats to support ocean conservation