Approaches to long-term forecasting of urban water demand in China
Forecasting long-term urban water demand is very important in water resources planning and management. In particular, faced with the threat of urban water scarcity, strategies and policies are required, and these need to be based on reliable water demand forecasts. However, there are many problems involved in forecasting long-term water demand, such as limited knowledge about the relationship between water use and the factors affecting it, uncertainties over the future and assumptions employed, the availability of water use data, etc. In China, little effort has been devoted to water demand forecasting, although water resources planning has been widely undertaken, and urban water shortage is currently a serious problem. In the light of the above, an overall approach to forecasting long-term urban water demand forecasting in China was undertaken. After reviewing the literature, the Chinese urban water demand system was analysed systematically, in terms of the four water use sectors: residential, industrial, agricultural and commercial. Based on the results revealed by the analyses, a system dynamic simulation model was built for forecasting longterm urban water demand. A case study has also been carried out to apply the model and to evaluate its performance. Compared to static models that have been developed in the literature, the system dynamic simulation model that has been developed in this study is superior in terms of the following aspects: (1) it clearly takes the time variable into account; (2) the system dynamic model allows alternative forecasts to be obtained easily and explicitly; and (3) the step-by-step procedures used in the system dynamic simulation give explicit and clear statements about the changing processes of the explanatory variables rather than simply accepting them as inputs.