Résumé
Ensuring long-term drinking water security is a critical challenge for water managers worldwide, fueled by multiple pressures including increasing water demand, deteriorating water quality, and reduced resource availability with climate change. This study addresses the need for integrated and spatially-explicit modeling tools for drinking water management at regional scale. It develops a methodological framework to support management strategies that balance drinking water security and environmental objectives, by combining drinking water supply hydro-economic modeling with high-resolution hydrogeological modeling and multiobjective optimization techniques. The coupling makes it possible to account for the spatial impacts of abstractions on groundwater levels, river-groundwater interaction, risks of salinization and satisfaction of drinking water demand, while optimizing their distribution and allocation. The framework is applied to the plain of Roussillon case study (Mediterranean France), providing insights for water management and demonstrating the scalability of the approach to a large-scale problem (167 decision variables). Results indicate optimal abstractions should be redistributed from the Pliocene toward the Quaternary aquifer (and toward upstream areas) to satisfy drinking water demands while avoiding seawater intrusion. However, results show that it will not be possible to fully satisfy future demand under drier climate without violating seawater intrusion constraints. Implementing demand-side management measures is a win-win-win strategy that reduces water shortages and environmental impacts to aquifers and rivers. Results also suggest that current regulatory abstraction caps are too generous to prevent future decline in piezometric levels under drier climate. The developed methodological framework can generalize to other basins and support analyses of adaptation strategies.