| Summary: | The floodplains of rivers are relevant living spaces for population globally and provide favorable locations for economic development. However, these areas are commonly exposed to floods, and the increasing population together with the changes in storminess as a result of global warming mean that the risks from flooding are expected to rise. Most studies investigating the impact that climatic change has on flood risk are based on a cascade of global climate model simulations coupled with regional climate models, hydrologic models, inundation models, and flood impact models. However, this approach is subject to uncertainties. Model results are found to be sensitive to climate forcing, the structure of the underlying models, the choice of methods used for downscaling and bias correction, and the use of extreme value analysis for both current and future climate conditions. Moreover, uncertainties are expected to propagate through the model cascade. To overcome these problems, we propose a method for analyzing and mapping the sensitivity of population exposure in floodplains to changes in flood magnitude. The method is based on downward counterfactuals, namely perturbations of a selected flood scenario by increasing its magnitude, interpreted in this case as the worsening of a today’s design flood event as a result of climatic changes. The increase in the impact of a current design flood compared to its counterfactual illustrates the sensitivity to changes in hazard. We calculate the normalized gradients of the flood exposure curves, that is, the increase in the exposure and magnitude of the perturbed event relative to the exposure and magnitude of the current scenario. We test the applicability of the method on local, national, and global scale by using existing data sets, including flood hazard maps, flood protection standards, floodplain delineation, river network definition, and spatial population distribution. The gradients were found to vary remarkably across the globe and are overall smaller in the upper range of flood magnitudes that in the lower range. Based on these results, we compare the drivers of the sensitivity in different parts of the world and identify river reaches with the highest relative gradients. These river reaches might be the most affected by climate change and thus deserve an in-depth investigation of the underlying characteristics of the floodplains and the need for climate change adaptation.
|
|---|
