Winter flooding in the Elbe - antecedent conditions vs seasonal weather

In winter 2023/24 along the Elbe River catchment in Germany, river flows were considered high or severe (European Flood Awareness System), and flooding impacted approximately 1.6 million people across Europe (International Disaster Database). Winter river floods are driven by both immediate seasonal weather events and the antecedent conditions accumulated over the preceding summer and autumn. Understanding the relative contributions of antecedent conditions and weather, as well as their interplay, to a location’s flood risk is essential for effective flood risk management. Using a lumped hydrological model and historical records, we will show that the elevated river flow along the Elbe was primarily influenced by high antecedent conditions. Using the tools and capabilities of NVIDIA’s Earth-2 platform we are able to create a large range of AI-generated weather events and combine them with different antecedent conditions for the winter 2023/24 season. We then use our hydrological models to assess the wide range of plausible river flooding. This approach allows us to answer the question of how severe flooding across the Elbe would have been in winter 2023/24 if major storms had occurred, and the likelihood of this scenario happening in future years.