Improving the Performance of Water Demand Forecasting Models by Using Weather Input

Literature shows that water demand forecasting models which use water demand as single input, are capable of generating a fairly accurate forecast. However, at changing weather conditions the forecasting errors are quite large. In this paper three different forecasting models are studied: an Adaptive Heuristic model, a Transfer/-noise model, and a Multiple Linear Regression model. The performance of the models was studied both with and without using weather input, in order to assess the possible performance improvement due to using weather input. Simulations with the models showed that when using weather input the largest forecasting errors can be reduced by 11%, and the average errors by 7%. This reduction is important for the application of the forecasting model for the control of water supply systems and for anomaly detection.