Climate change impacts on hydrology - informing the design of the modeling chain and supporting adaptation

Abstract

As climate change continues to perturbate regional hydrology, it has become increasingly necessary for the scientific community to support water management decisions. This responsibility requires an understanding of the impact of climatic and environmental change on future hydrological conditions. Numerical models are the principal tools used to obtain quantitative information on these dynamics. In common practice, multiple models are used in succession, known as a modeling chain. To assess and simulate potential changes to a hydrological regime due to climate change, it is first necessary to acquire climate information for current and future conditions. Emission scenarios are utilized to extend current knowledge of historical greenhouse gas emission data into the future. These scenarios are then used as input to general circulation models (GCMs) to provide information about how Earth system processes are likely going to develop under the influence of climate change. However, the spatial scale of GCMs is too coarse for local-scale impact studies. Thus, regional climate models (RCMs) are used to dynamically downscale climate variables. However, even after downscaling, biases often still exist between the GCM-RCM output and local-scale observational data. Bias correction techniques are therefore employed to reduce these biases, resulting in climate variables that are suitable as input to hydrological models. Hydrological models are then calibrated and used to create projections of streamflow under the influence of climate change. This dissertation delves into the modeling approach to achieve projections of climate change impacts on water resources. An initial study reviewed and summarized the steps to perform a climate change impact analysis as a chapter contribution to the Encyclopedia of Water. The motivation of this encyclopedia chapter was to provide guidance on the modeling chain, which was previously lacking in the literature. Another study introduced a process-based approach to evaluate the performance of climate models and their bias correction. This new approach was based on the concept that climate change impacts are the result of the interactions between variables, which are not considered in typical analyses. Additionally, the sensitivity of the modeling chain was evaluated in two co-authored papers, which showed that discharge projections are not especially sensitive to varying levels of potential evapotranspiration information in arid regions. However, discharge projections were shown to be sensitive to the choice of bias correction method for two glaciated catchments within Switzerland. The choice of bias correction method translated into considerable consequences for the hydrological responses of the catchments, although differences in total streamflow were negligible. The lessons learned from these aforementioned studies were applied in a final study, which aimed to support the decision making of a Swiss hydroelectricity company. Following a user-centered approach, a tailored climate change impact analysis was carried out. This approach allowed for insightful recommendations for climate change adaptation and decision-support regarding their upcoming concession negotiations. This study highlighted the importance of accounting for a stakeholder’s specific needs when designing a climate change impact study.