University of Siegen Develops Real-Time Hazard Maps for Heavy Rain and Flooding

13.07.2026

The fifth anniversary of the flood disaster in the Ahr Valley will be observed in mid-July. At the same time, heavy rainfall events are on the rise in many places as a result of climate change. With “RiskAware,” the University of Siegen is developing a digital tool designed to help cities and municipalities prepare for heavy rainfall and flooding. With the help of artificial intelligence, risks are expected to be identified more quickly and accurately and better assessed in the future.

Participants in the RiskAware project workshop with representatives from local governments.

The University of Siegen is developing a digital tool designed to better prepare cities and municipalities for the consequences of climate change. The “RiskAware” research project is developing a system that dynamically aggregates, analyzes, and presents information on heavy rainfall, flooding, and other water-related natural hazards in real time in an easily understandable format. The goal is not to identify risks only after a disaster has occurred, but to make them visible in advance, thereby enabling better decisions regarding civil protection and urban planning.

Although the Federal Agency for Cartography and Geodesy (BKG) already provides hazard maps for heavy rainfall, these maps only reflect the situation at the time they were created. New buildings, redeveloped areas, or other construction projects can cause existing hazard maps to no longer reflect the current situation.

New Hazard Maps in Real Time

This is exactly where RiskAware comes in. In this project, Prof. Dr. Jorge Leandro and Dr. Ingrid Althoff, along with two doctoral students from the Chair of Hydromechanics and Hydraulic Engineering at the University of Siegen, are developing an AI-supported system that links various data sources and uses them to generate up-to-date heavy rainfall hazard maps. Satellite data, terrain models, precipitation data, and information on roads, buildings, and land use are automatically integrated. Two coupled deep-learning models are designed to autonomously detect changes in urban areas and generate dynamic hazard maps in real time. The tool is initially being tested in collaboration with three cities and municipalities in North Rhine-Westphalia: the cities of Bochum and Lüdenscheid, and the district of Recklinghausen. These areas have different characteristics: Bochum is densely populated; the Recklinghausen district includes both urban and rural areas; and Lüdenscheid has steeper slopes and a rural character. Two additional cities will be used to test the system’s transferability.

Another distinctive feature: The researchers are developing the models so that they can also be applied to regions not included in the training. Additionally, the system is intended to be able to automatically assess damage in the future. This would allow municipalities to evaluate risks much more quickly than before and plan preventive measures in a more targeted manner. Among other things, the maps developed are intended to help prepare protective measures or better simulate evacuation scenarios. The actual decisions remain the responsibility of the relevant authorities and disaster management organizations. “The idea arose from the fact that heavy rainfall in Germany has caused massive damage in recent years, and according to current climate models, extreme weather events are expected to occur with increasing frequency in the future,” explains Leandro.

From Research to Practice

A prototype is scheduled to be completed by the end of the project in February 2028. The source code for the AI models will then be made publicly available via GitHub. Further development and commercialization will then be handled by the project partner company, Okeanos Smart Data Solutions GmbH. The medium-sized company from Bochum is expected to benefit directly from the project, thereby strengthening German and European competitiveness.

The “RiskAware – Development of a Generative Planning Tool for Urban Planning Aware of Heavy Rainfall Risks” project builds on previous research conducted by the Chair of Hydromechanics and Hydraulic Engineering at the University of Siegen. Previous projects have already focused on AI-supported early warning systems for heavy rain and flooding. RiskAware is funded through the GreenEconomy.IN.NRW innovation competition, which is financed by the state of North Rhine-Westphalia and the European Union. The grant amount for the University of Siegen is approximately 690,000 euros. The goal of the funding is to translate scientific findings into practical applications as quickly as possible and to make cities and businesses more resilient to the consequences of climate change.