Spatiotemporal Evolution and Prediction of the Water Ecological Footprint in Chinese Super-megacities

The harmonized development of water resources and cities is of great importance for the high-quality development of regions and water security. Study of water ecological footprint (WEF) has important theoretical and practical significance for the sustainable utilization of water resources. This study utilizes an improved WEF model and panel data from Chinese super-megacities to analyze the spatiotemporal evolution characteristics and variation trends of the WEF through the application of kernel density estimation, variable coefficient model and gray neural network model. The research findings are as follows. (1) The WEF per capita of 21 super-megacities shows significant variations from 2003 to 2020, along with large inter-annual fluctuations. This disparity seems to be widening spatially. The regions with the highest WEF per capita are concentrated in 12 cities, where the annual average value exceeded 0.006 hm2/cap. Besides, these super-megacities are predominantly located in the northeast. The water ecological carrying capacity (WEC) of super-megacities reflects the supply capability of water resources and is declining during the study period, with an average multiyear decline rate of 1.2%, highlighting the urgent need to enhance water production capacity. (2) Seventeen super-megacities are experiencing water resource ecological deficits and severe ecological imbalances, with the deficit primarily concentrated in the northeastern region. According to sustainability evaluation indicators, water resources in 13 super-megacities are undergoing unsustainable development, particularly with Beijing, Tianjin and Zhengzhou reaching Level I and being in a moderately unsustainable state. Research has shown that the sustainable utilization of water resources is more advanced in southern regions compared to northern regions. (3) Economic development, industrial structure, technology level and environmental regulations have a significant differentiated effect on WEF per capita across various city levels. (4) The prediction results of the gray neural network demonstrate a decrease in the WEF per capita of seven representative super-megacities from 2023 to 2027, showing significant variations across different regions.