Improving estimation of space-time dynamics of floods in Europe by combining modelled and observed flood impact data

Long-term trends in flood losses are regulated by multiple factors including climate variation, demographic dynamics, economic growth, land-use transitions, reservoir construction and flood adaptation measures. Attributing losses to any of those factors for historical flood events would require the ability to recalculate reported impacts (such as area inundated, fatalities, persons affected or economic loss) under counterfactual scenarios. Here, we present how observed flood impacts that have occurred in 42 European countries since 1950 can be compared with model simulations as a step towards climate change attribution of flood losses. Firstly, we reconstructed the potential footprints and inundation depths of individual riverine and coastal flood events. This was made possible by combining continuous simulations of river discharges (based on the LISFLOOD model) and storm surge heights (based on the Delft3D model) with flood hazard maps derived through hydrodynamic modelling. Then, the flood footprints were intersected with a set of time-varying, high-resolution exposure maps of land use, population and asset values in Europe since 1950 (based on the HANZE-Exposure v2.0 model). Modelled potential flood damage can then be evaluated against historical records of flood occurrences and their impacts. To this end, we are collecting dates, locations and, where available, impact statistics of floods in an updated HANZE-Events database. By comparing which potential floods did cause impacts, in which locations and to what magnitude, and which floods were prevented by flood protection, it will be possible to infer flood vulnerability and preparedness across time and space. Available preliminary results enable presenting the space-time dynamics of European flood damages under different exposure scenarios.