Zhang et al. (2025) Integrating deep learning and multi-objective optimization for floodwater utilization: a coordinated surface water-groundwater regulation framework for groundwater recovery

Identification

• Journal: Journal of Hydrology
• Year: 2025
• Date: 2025-12-12
• Authors: Libin Zhang, Yonggen Zhang, Shi Lian, Xinwang Li, Ping Feng, Lutz Weihermüller
• DOI: 10.1016/j.jhydrol.2025.134749

Research Groups

• State Key Laboratory of Hydraulic Engineering Intelligent Construction and Operation, Tianjin University, China
• Institute of Surface-Earth System Science, School of Earth System Science, Tianjin University, China
• Hebei Provincial Hydrologic Survey and Research Center, China
• Hebei Institute of Water Science (Hebei Province Dam Safety Technology Center, Hebei Province Levee Sluice Technology Center), China
• Agrosphere Institute IBG-3, Forschungszentrum Jülich GmbH, Germany

Short Summary

This study developed a novel deep learning and multi-objective optimization framework to integrate flood control, water storage, and groundwater recovery through coordinated surface water-groundwater regulation. It demonstrated that optimizing reservoir operations and implementing managed aquifer recharge effectively reduced flood risk and water scarcity while promoting groundwater recovery.

Objective

• To develop a multi-objective optimization framework employing deep learning to integrate flood control, water storage, and groundwater recovery, ensuring ecological flow and downstream flood safety, specifically addressing the challenge of coupling physical models with multi-objective algorithms for groundwater recovery.

Study Configuration

• Spatial Scale: Study area (not explicitly named in the provided text).
• Temporal Scale: 2023 flood season (June to September).

Methodology and Data

• Models used: Deep learning, multi-objective optimization framework, 3D groundwater numerical model.
• Data sources: Multi-scenario simulations for reservoir operations, assessment of managed aquifer recharge (MAR) impact. (No explicit external data sources like satellite, observation, or reanalysis are mentioned in the provided text).

Main Results

• Increasing the flood limited water level (FLWL) reduced average reservoir flood risk by 84.9 % and water scarcity by 61.9 %.
• Raising FLWL weakened the inverse relationship between flood risk and water scarcity, improving individual objectives and reducing conflicts for balanced optimization.
• Maintaining continuous ecological river flow promoted groundwater recovery despite reduced total river discharge.
• Managed aquifer recharge (MAR) at 300 cubic meters per day (m³ d⁻¹) achieved effective groundwater recovery in 17.6 % of the study area, with a maximum recovery of 0.46 meters.

Contributions

• Presents a novel framework coupling deep learning, multi-objective optimization, and 3D groundwater modeling.
• Enables optimized surface water-groundwater regulation for enhanced floodwater utilization and groundwater recovery.
• Addresses the challenge of integrating groundwater systems into flood utilization strategies, moving beyond surface water-centric approaches.

Funding

• The provided text does not contain a funding section.

Citation

@article{Zhang2025Integrating,
  author = {Zhang, Libin and Zhang, Yonggen and Li, Jianzhu and Lian, Shi and Li, Xinwang and Feng, Ping and Weihermüller, Lutz},
  title = {Integrating deep learning and multi-objective optimization for floodwater utilization: a coordinated surface water-groundwater regulation framework for groundwater recovery},
  journal = {Journal of Hydrology},
  year = {2025},
  doi = {10.1016/j.jhydrol.2025.134749},
  url = {https://doi.org/10.1016/j.jhydrol.2025.134749}
}