Junior Research Group on Flood Forecasting and Early Warning


Challenges and objectives


Flooding is a natural occurrence. However, it is becoming a societal challenge due to increasing climate change, human intervention in the natural hydrological cycle, and intensified aggregation of values. Policymakers, scientists, flood institutions, and affected communities face the present and future task of dealing with this problem.


A key problem in the development of flood forecasting systems is accuracy and prediction time, especially with respect to the prediction of heavy rainfall-induced flood events. In particular, extreme heavy rainfall events that can lead to sudden urban flooding occur in extremely dynamic weather patterns and very rapidly. As a result, pluvial flooding and flash flooding can occur at extremely short notice. Prediction of such heavy rainfall-induced flood events requires special methodological approaches in modeling, forecasting, and effective warning.

In the field of early flood warning, there are still many issues to be addressed, especially in dealing with significant uncertainties in model predictions. Here, it is critical to determine how to evaluate these uncertainties and what additional information might be used to reduce them. Similarly, there is the question of how to effectively communicate prediction data and their uncertainties and what recommendations for action can be derived from them. Another aspect is to communicate the fact that forecasts cannot cover all contingencies and therefore complete predictability is not guaranteed.

The goal of this research group is to explore the various aspects of modeling, forecasting, and warning of flood events and to develop new model systems.


Tasks and methods of research

© J. Hofmann

Our research group is dedicated to the in-depth analysis and the possibilities of forecasting flood events (caused by heavy rainfall). In doing so, we investigate various aspects, starting from the analysis of precipitation data, through the characterization of floods, to the detailed study of the events. We use various methods, including statistical approaches, deterministic models, numerical simulations, and artificial intelligence, to study the entire process from the origin to final impact of the flood event.