Traditional flood frequency analysis is based on the statistical evaluation of annual observed maximum flood peaks, which usually are assumed to be independently and identically distributed. However, there is evidence that persistence in climate, the influence of regional sea surface temperature variations on the manifestation of large-scale circulation modes, and catchment memory may play a role in modifying the frequency of hydrological extremes.

These findings imply to distinguish between unconditional flood risk that can at best provide estimates of long-term flood risk and conditional flood risk that considers climatological and hydrological factors. If the conditional and unconditional flood risk differ significantly, there may be potential to exploit the links between climate state/catchment state and flood magnitude and to develop seasonal forecasting models of flood risk.

Our research project is part of the current efforts to get a “mechanistic” understanding of conditional flood risk that explicitly recognizes catchment memory and climate variations. The goal is to build a robust methodology for Europe, which will be replicable in other regions with available climate and catchment data.

Joint research initiative between GFZ, German Research Centre for Geosciences, Potsdam, Germany and AXA Global P&C