Deep Generative Networks for Detecting Anomalous Events in the Water Cycle (D05)

• Subject Area: Atmospheric Science ; Image and Language Processing, Computer Graphics and Visualisation, Human Computer Interaction, Ubiquitous and Wearable Computing
• Term: since 2022
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 450058266

This project aims to identify extreme events in simulated water cycle components by developing novel deep generative networks that detect anomalous events in simulated data. Since the detected anomalous events will be data-driven, they will not always co-occur with extremes like droughts. We will therefore develop novel methods based on deep learning that predict the impact of anomalous events like agricultural droughts. Furthermore, we will use the developed approaches to investigate the impact of anthropogenic drivers on anomalous events.

• DFG Programme: Collaborative Research Centres
• Subproject of: SFB 1502: Regional Climate Change: Disentangling the Role of Land Use and Water Management
• Applicant Institution: Rheinische Friedrich-Wilhelms-Universität Bonn
• Project Heads: Privatdozentin Dr. Petra Friederichs; Professor Dr. Jürgen Gall