摘要
With a booming expansion of urbanization, urban water consumption (WC) attracts increasing concerns in developing countries worldwide, particularly for megacities. In this study, an urban WC model for Shenzhen, a rapidly developing flagship megacity in South China from a small agrarian fishery village since 1979, was built up to simulate the WC changes (1994-009) with aim to formulate local water resources management strategies. Basically, the model was constructed using a variety of methods including a back-propagation artificial neuron network (BP-ANN), a quadratic polynomial model, a regression and auto-regressive moving average combination model, and a Grey Verhulst model. Simulation of the WC was conducted using a multiple regression forecasting model and a BP-ANN model. The results from these two models showed that the BP-ANN model is outperformed. Subsequently, a series of social–economic and demographic scenarios were formulated to project WC (2011-020) with uncertainty analysis. The results suggest that the total WC will increase slower and slower over the decade. It might approach a saturated threshold soon after 2020. Scenarios of WC incorporating uncertainty analysis aiming to provide reliable prediction results constitute the highlight of this study. This study will be beneficial to formulate appropriate sustainable development strategies of water resources for similar megacities in South China.