Assessing water management vulnerability under future climate scenarios: the case of the reservoirs in the Seine River basin

The high stakes surrounding the availability of water resources under future climate raise the need to make relevant climate-water projections available to water managers. However, beyond this information, there is also the need to account for water management current practices and future needs in terms of adapting operations to balance water availability and demand under future climate and water use pressures. This isparticularly the case when dealing with reservoirs and their ability to regulate floods and droughts.

This study aims at setting up a full modelling chain, from climate to reservoirs management, to address water availability and operational management needs under future climate and water demand conditions. Our case study is the Seine River basin in France and its four upstream reservoirs which are operated to regulate flows up to the city of Paris. We first co-designed with stakeholders the “what-if” scenarios to investigate, combining possible future climate and management states. We then set up a simplified modelling approach to allow us to address the stakeholder’s needs. This modelling framework relies on a semi-distributed rainfall-runoff model, coupled to different management scenarios (reservoir rule curves) and forced by contrasted climate projections. The model was calibrated and validated using historical climate, hydrological and reservoir operation datasets, while also relying on stakeholder consultation. The relevance of current reservoir management operations in a future climate where the intensity and frequency of low flows might increase was investigated. We used the results of the national EXPLORE2 project, which downscaled 17 pairs of RCM/GCMs from CMIP5 over France. We used the outputs of the national EXPLORE2 project, which downscaled 17 pairs of RCM/GCMs from CMIP5 over France.

Our results show that the modelling framework accurately reproduces the relative influence of the reservoirs on the downstream discharges over the historical period. The influence on low-flow support is on average around 15%, with a maximum influence that can reach up to 50% during summer and downstream up to Paris. We also show that this can be a good indicator to quantify the influence of reservoir management on low flow conditions under future climate.

This work received funding from the Horizon Europe under grant agreement No. 101059372 (STARS4Water project).