需水预测是需水管理的基础与难点，已成为学术界与生产部门关注的热点。以往国内外需水预测研究往往“重总量，轻结构”，对各行业需水机理与需水层次认识不足，预测成果经不住实践检验。本项目在解析层次化需水——刚性、弹性和奢侈需水内涵的基础上，分析各行业用水变化特征与规律，剖析社会经济发展内生动力与外部环境变化对需水的影响机制，揭示各行业水资源层次化需求机理，并基于隶属度区间分布函数建立各行业层次化需水评判模型；引入全面刻画社会经济发展全景图的CGE模型，新增水源替代模块与水资源总量控制模块，构建层次化需水预测CGE模型；以用水竞争最强烈的北京市为例，设置模拟方案集，模拟规划水平年社会经济发展及29行业分水源分层次需水情况，进行需水预测成果合理性检验，提出北京市多水源分用户分层次需水预测方案。预期成果对丰富需水预测理论具有一定的科学意义，同时为水资源精细化配置、水资源综合规划与管理提供重要决策依据。

The water demand forecasting is the basic work and difficult problem of water demand management, it has become a focused issue in academic circles and production departments. Many researches on water demand forecast have been conducted both at home and abroad, however, there are many deficiencies in the existing research, such as, "over- estimate volume and underestimate structure“ , little understand of water demand mechanism and demand hierarchy, these leads to the deviation of the forecasting results from reality, this deviation in water demand have misled some water resource planning. Therefore, this research will take the Beijing city, which have strongest competition in water use, as an case region and conduct following studies: (1) define hierarchical water demand, including rigid demand, elastic demand and luxury demand, analyze the characteristics and law of water use in various industries, analyze influence mechanism of the internal driving force of social-economic development and external environment on water demand, finally reveal the mechanism of hierarchical water demand in various industries. A evaluation model of hierarchical water demand was established based on the interval distribution function of membership degree; (2) develop the computable general equilibrium (CGE) model, which can detail social-economic development and reflect the whole process of water use, consume, and drainage. The water production and supply sectors is refined, this research newly add water source replacement module and total water consumption control module. The CGE model for hierarchical water demand prediction is constructed; (3) design various scenarios and simulate social-economic development status and hierarchical water demand in 29 sectors under different scenarios from 2012 to 2030, test the rationality of water demand prediction results. Finally, this research form a water demand forecast scheme that could reflect the demand of different hierarchy, different sources and different users. The expected results have certain scientific significance for enriching the theory of water demand prediction, and provide important decision basis for the water refined allocation, water resources comprehensive planning and management.