Evaluation and prediction of water resource carrying capacity in Bortala Mongolian Autonomous Prefecture based on the GRA-TOPSIS-PSONN model

Study region Bortala Mongolian Autonomous Prefecture in Xinjiang. Study focus To support water management in water-stressed regions, this study developed an integrated “assessment-diagnosis-prediction” framework for evaluating Water Resources Carrying Capacity (WRCC) in Bortala Prefecture. A water resource-society-economy-ecology indicator system was constructed, and indicator weights were determined by combining the Entropy Weight Method and CRITIC method through game theory. The GRA-TOPSIS model was used to assess WRCC from 2013 to 2022 and diagnose key obstacle factors, while a two-stage neural network model was employed to predict WRCC trends during 2023–2030. This study fills a gap in WRCC research for arid inland regions by linking historical assessment, obstacle diagnosis, and future prediction within a unified framework. New hydrological insights for the region WRCC in Bortala Prefecture increased from 0.40 in 2013–0.49 in 2022, but remained mostly at the Critical Carrying Capacity level, indicating limited room for further water-resource development. Water production modulus, ecological water-use proportion, and per capita water resources were the most frequent limiting indicators. Forecast results show that WRCC will fluctuate between 0.371 and 0.509 during 2023–2030, with clear regional differentiation, suggesting that future water security depends more on adaptive allocation and demand regulation than on water-supply expansion alone.