Multidimensional water-economic-ecological joint optimization for Manas River Basin, China, using a dual-population co-evolutionary algorithm

The Manas River Basin in the Xinjiang Uygur Autonomous Region is a typical arid oasis in Northwest China, where water is scarce. A multidimensional water-economic-ecological joint optimization model was constructed and solved using an improved hybrid algorithm combining Multi-Objective Particle Swarm Optimization and Non-dominated Sorting Genetic Algorithm II. The rationality of the solution was evaluated using a comprehensive water-resource optimization evaluation system. New Hydrological Insights on the Study Region: (1) A water supply-demand balance analysis revealed that the Manas River irrigation area is projected to experience a water deficit of 6.47 × 10⁷ m³ in 2025, which will increase to 1.89 × 10⁸ m³ by 2030. (2) The hybrid algorithm demonstrated greater stability in the search performance within three-dimensional spaces. It could avoid local optima while ensuring global search capability. Dynamic weight allocation was adopted during the optimization process to reduce the subjectivity associated with manual weight adjustments. (3) In the optimal scheme for the planning year 2025, the Manas River irrigation area remained below the control thresholds, achieving a surplus of 4.02 × 10⁶ m³ of water and generating an economic net benefit of 7.27 × 10⁹ yuan. By 2030, the water deficit is projected to decrease by 76.85 %, with a remaining deficit of 4.38 × 10⁷ m³ , and an economic net benefit of 7.48 × 10⁹ yuan. These findings provide technical support for regional water resource optimization. • Hydrological data from Manas River Basin, Xinjiang Region, China, in recent ten years. • The water demand of different water departments in the Manas River irrigation district in the planning year is estimated. • A three-dimensional joint water resources optimization model was established. • The MOPSO and NSGA-II optimization algorithm is used. • The evaluation system is used to screen the optimal scheme.