Neverre et al. (2026) Balancing drinking water security and conservation: A spatial multi-objective optimization framework for regional groundwater management under global change

Identification

• Journal: Journal of Hydrology
• Year: 2026
• Date: 2026-01-18
• Authors: Noémie Neverre, Jonathan D. Herman, Ludovic Schorpp, Sandra Lanini, Manuel Pulido-Velazquez, Yvan Caballero
• DOI: 10.1016/j.jhydrol.2026.134950

Research Groups

• BRGM, Univ Montpellier, Montpellier, France
• G-EAU, Univ Montpellier, AgroParisTech, BRGM, CIRAD, INRAE, Institut Agro, IRD, Montpellier, France
• Department of Civil and Environmental Engineering, University of California, Davis, CA, USA
• Centre for Hydrogeology and Geothermics, University of Neuchâtel, Neuchâtel, Switzerland
• Research Institute of Water and Environmental Engineering (IIAMA), Universitat Politècnica de València, València, Spain

Short Summary

This study developed a spatially-explicit multi-objective optimization framework, coupling hydro-economic and high-resolution hydrogeological models, to balance drinking water security and environmental conservation in the Roussillon plain under global change. It found that future demand cannot be fully met under drier climates without violating seawater intrusion constraints, necessitating abstraction redistribution and demand-side management.

Objective

• To develop and apply a spatially-explicit multi-objective optimization framework for regional drinking water management under global change scenarios, balancing drinking water security with resource preservation, accounting for hydrogeological dynamics, and addressing high-dimensional optimization problems.

Study Configuration

• Spatial Scale: Regional scale, applied to the plain of Roussillon (Mediterranean France), covering 90 municipalities with a hydrogeological model resolution of 300 meters by 300 meters inland and 600 meters by 300 meters at sea, totaling 70,755 cells.
• Temporal Scale: Long-term planning horizon, focusing on the year 2050, using a steady-state hydrogeological model and optimizing annual abstraction volumes.

Methodology and Data

• Models used:
  • Multi-objective optimization: NSGAII (from Platypus Python library)
  • Hydro-economic model: Drinking water supply model (Neverre, 2024), recoded in Python.
  • Hydrogeological model: Simplified 3D groundwater model of the Roussillon aquifers (Schorpp et al., 2023), a steady-state MODFLOW 6 model built with FloPy.
  • Nested optimization: Gradient-based optimization for water allocation.
• Data sources:
  • Climate projections: Ensemble of simulations from five global climate models and two downscaling methods for 2050 (RCP8.5 scenario), providing effective rainfall and natural recharge.
  • Drinking water demand projections: 2050 water demand projection from Roussillon’s Plan to Secure Drinking Water (Artelia, 2017).
  • Observed data: 2015 abstraction levels, water use, prices, and historical abstraction caps (Hydriad, 2014).
  • Hydrogeological data: Digital Elevation Model for river stages, input fluxes from previous studies, and irrigation recharge estimates based on crop needs.
  • Stakeholder input: Participatory workshops for defining demand-side management scenarios and selecting key piezometric control points.

Main Results

• Under 2050 drier climate and increased demand, it is impossible to fully satisfy future drinking water demand while meeting seawater intrusion constraints; even the least-cost solution shows a 15% (6.19 million cubic meters per year) shortage.
• Optimal abstraction redistribution is required from the Pliocene to the Quaternary aquifer (and upstream areas) to satisfy demands while avoiding seawater intrusion.
• Demand-side management measures (10% demand reduction and 85% network yield) act as a "win-win-win" strategy, significantly reducing shortages (to 4% or 1.41 million cubic meters per year) and environmental impacts on aquifers and rivers, though some coastal areas still face up to 31% shortage.
• Current regulatory abstraction caps for the Pliocene aquifer are too generous to prevent future piezometric level decline under drier climates, and there are no caps for the Quaternary aquifer.
• Relaxing seawater intrusion constraints (allowing a 0.25 meter decrease at control points) improves demand satisfaction but at the expense of increased environmental impacts on groundwater levels and river drainage, highlighting trade-offs between precaution and demand satisfaction.

Contributions

• Incorporates a spatially explicit representation of hydrogeological dynamics into hydro-economic modeling for drinking water supply, allowing for the evaluation of spatially heterogeneous decisions and impacts.
• Expands the size of the optimization problem (167 decision variables) for integrated 3D hydrogeological modeling within multi-objective optimization, addressing a complex regional-scale problem.
• Provides a generalizable methodological framework applicable to other socio-hydrosystems, particularly in coastal areas, facing challenges in balancing groundwater abstraction and drinking water allocation with environmental constraints.

Funding

• French National Research Agency (convention N◦19-CARN 0003-01)
• French Ministry of Higher Education Research and Innovation
• European Regional Development Fund (ERDF) through the Occitanie Region
• Rhone Mediterranean and Corsica Water Agency
• Perpignan Méditerranée Métropole Communauté Urbaine
• Pyrénées Orientales department council
• French Geological Survey (BRGM)
• Dem’Eaux Roussillon project

Citation

@article{Neverre2026Balancing,
  author = {Neverre, Noémie and Herman, Jonathan D. and Schorpp, Ludovic and Lanini, Sandra and Pulido-Velazquez, Manuel and Caballero, Yvan},
  title = {Balancing drinking water security and conservation: A spatial multi-objective optimization framework for regional groundwater management under global change},
  journal = {Journal of Hydrology},
  year = {2026},
  doi = {10.1016/j.jhydrol.2026.134950},
  url = {https://doi.org/10.1016/j.jhydrol.2026.134950}
}