驱动预测过程的几乎所有变量都存在不确定性:人口和家庭构成、就业和工业增长、客户部门的收入和用水特征、保护目标和绩效，以及天气。准确预测的关键是隔离或指定每个变量的影响，以便将随时间变化的变量引起的需水量变化准确地纳入预测中。本文的目的是简要讨论需求预测的主要方法，然后集中于衡量保护绩效，并将保护目标纳入长期需求预测。

There is uncertainty in virtually all variables that drive the forecasting process: population and household formation, employment and industrial growth, income and water use characteristics by customer sectors, conservation targets and performances, and weather. The key to accurate forecasting is to isolate or specify the impacts of each variable so that changes in water demand attributable to changes in the variables over time can be accurately included in the forecasts. The objective of this paper is to briefly discuss the major methods of demand forecasting, and then concentrate on measuring conservation performance and integrating conservation targets into long-run demand projections.