A Study Of Water Resource Demand Forecast In Hilly Loess Plateau In Ningxia

Water is the most important natural resource in nature. It is the source of life and economic blood which is considered as the important and basic resource that can guarantee the normal development of economy. While the water resource available on earth is limited, therefore, with the development of economy and society, water resource is increasingly becoming the strategic resource related to the national security. Although our country’s water resource ranks forefront in the world, the per capita water resource quantity accounts for only a quarter of the world average level. Therefore, how to plan and utilize the limited water resource reasonably at home, water demand forecast will become an important research topic.The subject of the water resource is from hilly losses plateau in Ningxia. The author tries to make a prediction of the water resource demand on the future level years in hilly losses plateau through the BP neural network model method combining with the factor analysis, and has got the believable results. First, the author classifies the water resource in hilly losses plateau according to the life water demand, agriculture water demand and industry water demand, and then analyzes the impact factors, including the natural conditions, economic development and human activities and so forth and abstract the main factor from numerous factors. After finding out the main factor, the author makes a detailed analysis of the main factor and establish the BP neural network model by using of the result of the analysis. Finally, the water resource demand is gotten.In order to guarantee the validity of the data, this passage adopts two methods to predict the water resource demand. The first method is to adopt the traditional prediction methods, such as quota method and trend method, and then make prediction on water resource demand of various industries. Forecast that the regional water demand in2015will be17466.84million cubic meter and in2020will be19222.43million cubic meter. The second method is to establish the BP neural network model in total water resource demand. Forecast that the regional water demand in2015will be17466.60million cubic meter and in2020will be19222.60million cubic meter. Through comparison and analysis, this second method is not quite differ from the traditional method. Water demand prediction absolute error in2015is-0.76million cubic meter and in2020is-0.17million cubic meter. The two methods are just different angles. In the water resource planning process, we recommend the second method, for this method is more meticulous and comprehensive to consider the water resources situations in various industries, helping the researchers to make a more reasonable water resources planning.