Analysis of climate and hydrological model uncertainty in climate change assessments in two Danish catchments

Uncertainty in the analysis of climate change impacts has attracted a lot of research interest. Its importance lies in the fact that decision makers require future projections with a reduced uncertainty. However, in some cases this does not happen. For example, there are studies where the projected change has a large spread and contrasting change signal, complicating the decision-making process. Therefore, it is relevant to assess the uncertainty and decrease the variability in the projections in order to provide useful information for decision-makers.

There are different sources of uncertainty. For instance, from the climate models, hydrological models or projected scenarios. Here, we assess the uncertainty associated to climate and hydrological models. We use a set of bias-corrected models from Euro-CORDEX to drive different versions of the MIKE-SHE hydrological model in order to simulate hydrological variables. We develop the study in two catchments located in western Denmark.

Initially, we assess the skill of three MIKE-SHE models to simulate the discharge and groundwater head in the Storaa catchment under changing climate conditions using a differential split sample test. Each model is ran using a different setup to describe the unsaturated zone. Then, the models are used to project changes for purpose-specific metrics that are relevant for decision-making in order to assess the uncertainty on the projected change.

Additionally, we analyze the uncertainty of climate models on the projected change for discharge, groundwater head, soil moisture and evapotranspiration in the Skjern catchment using four hydrological models with different simulation of the unsaturated zone. We focus on reducing the uncertainty in the projections considering the simulation skill of the climate models and the variability of the change that they project. Our results expect to provide methods that help understand the uncertainty from the assessments of climate change impacts and reduce the uncertainty of the projection to provide useful information for decision-making.