Spatio-temporal Evolution And Prediction Of Water Consumption In Yangtze River Delta Urban Agglomeration Based On Nighttime Light Data

Water resources are essential basic resources in the process of human survival and economic social development.The water demand of human beings is increasing,but the scarcity of available water resources on the earth and the imbalance of its spatial distribution make many regions in the world face serious water crisis.As an important intersection of the Belt and Road Initiative and the Yangtze River Economic Belt,the Yangtze River Delta urban agglomeration plays an important role in China’s economic development.However,in recent years,the rapid growth of population and economy and environmental pollution have intensified the contradiction between supply and demand of water resources,resulting in the shortage of water resources,which greatly limits the comprehensive and sustainable development of the Yangtze River Delta Urban Agglomeration.Therefore,it is of great reference value for the Yangtze River Delta urban agglomeration to explore the spatial-temporal evolution characteristics of water consumption and its driving factors,and to reasonably predict its future water consumption,so as to rationally formulate the overall planning of water resources and effectively cope with the pressure of water resources in the future development process.Based on the night light data from 2000 to 2018,combined with water consumption data,this paper establishes a spatial model of urban water consumption in the Yangtze River Delta urban agglomeration.On this basis,the standard deviation ellipse and exploratory spatial analysis method are used to study the spatial-temporal evolution characteristics of water consumption in the Yangtze River Delta urban agglomeration,and the main driving factors of water consumption in the Yangtze River Delta urban agglomeration are explored by using the geographical detector and factor analysis method.In addition,in order to predict the water consumption of the Yangtze River Delta urban agglomeration,this paper also constructs the random forest,support vector machine and BP neural network model,and the random forest,support vector machine and BP neural network model based on factor analysis.By comparing the prediction accuracy of the six models,the optimal water consumption prediction model is determined.Combined with grey prediction model and comprehensive consideration of economic factors,water price,industrial structure and water saving factors,five scenarios are set up to predict the water consumption of the Yangtze River Delta urban agglomeration from 2019 to 2030.The main conclusions of this paper are as follows:(1)From 2000 to 2018,the spatial distribution of urban water consumption in the Yangtze River Delta urban agglomeration changed from close to the’east-west’direction to the’northwest-southeast’direction,and the center of gravity shifted to the southeast as a whole.The spatial agglomeration of water consumption showed a trend of increasing first and then decreasing,but the overall spatial agglomeration of water consumption was still strengthened,and the spatial distribution of water consumption shows a pattern of high water consumption in core cities such as Shanghai and Nanjing and southeast coastal cities.In addition,the county scale water consumption of the Yangtze River Delta urban agglomeration has a significant positive spatial correlation.The high-value clustered areas are mainly distributed in the central region,and the low-value clustered areas are distributed in the edge region.(2)From the overall point of view,the level of urban economic development,price and industrial structure,water-saving measures are the main factors that affect the water consumption of Yangtze River Delta urban agglomeration,and from the part point of view,there are great differences among the main factors that affect the water consumption of high-speed developing,medium-speed developing and low-speed developing cities.In addition to the level of urban economic development has an important impact on all types of cities,the tertiary industry ratio and water-saving investment are the main factors that affect the water consumption of high-speed developing cities,the awareness of water saving is the main factor that affects the water consumption of medium-speed developing cities,and secondary industry ratio is the main factor that affects the water consumption of low-speed developing cities.(3)Among the six water consumption prediction models constructed,the average relative error of FA-SVM model is 2.13%in the fitting stage,1.00%in the prediction stage and 1.95%for the whole sample data.The model has the highest prediction accuracy,the best generalization performance and the overall performance is better than the other five models.Therefore,FA-SVM model is selected as the water consumption prediction model of the Yangtze River Delta urban agglomeration.(4)The water consumption of the Yangtze River Delta urban agglomeration in2030 is 140.61,171.78,123.52,118.17 and 130.35 billion m~3 under the benchmark model,the rapid economic growth scenario,the structural adjustment scenario,the water saving priority scenario and the comprehensive scenario,respectively.Compared with the benchmark model,the water consumption under the rapid economic growth scenario increases by 22.17%,and the water consumption under the structural adjustment scenario,the water saving priority scenario and the comprehensive scenario decreases by 12.15%,15.96%and 7.30%,respectively.Among them,the water saving priority scenario can significantly save water resources.Therefore,the Yangtze River Delta urban agglomeration should focus on improving industrial water use efficiency and increasing water saving investment,supplemented by water price and industrial structure adjustment.