Flooding in marsh areas caused by climate change – sensitivity to the accuracy of sea level projections and bias correction procedure

One of the most sensitive areas to climate change impacts are coastal-near and low-relief areas. While these coastal zones often are places of high population density and important infrastructure, they are also particularly exposed to multiple future climate change hazards such as sea level rise, intensified storm surges, rising groundwater and high-intense precipitation events. Thus, elevated risks of overbank spilling, dam breaks and flooding events are expected in a future climate in temperate wet regions such as Denmark. To mitigate and adapt to such elevated risks, prediction of future flooding events is vital, and here hydrological modelling is an essential tool. However, such impact evaluations are subject to a range of uncertainties such as emission, climate and sea level prediction uncertainties as well as impact model uncertainty.

In this study, we investigate the uncertainties of predicted flooding from a river course and the groundwater system using a hydrodynamic model coupled with a 3D groundwater model. The model is forced with different climate scenarios and inputs from two different approaches of downscaled global sea level rise implemented using an indirect and direct bias correction procedure. The indirect procedure uses a simple delta change approach perturbing future climate and sea level change on observations, while the direct method uses bias-corrected climate model data, corresponding ocean model run data (for internal oceanic response to climate change), combined with global sea level change. This approach makes it possible to investigate the relative importance (sensitivity) of the flood prediction of sea level projection and bias correction procedure as well as the identification of past and future origin (surface or groundwater) of flooding events.