Forecasting Water Demand Using Weather Data
This report was produced for the Urban Water Research Association of Australia, a now discontinued research program.
Report no. UWRAA 30
An outline is given of the purpose and nature of forecasting and of the various methods and models used for forecasting. A model is developed for forecasting water demand for an urban area using weather data (daily maximum temperature and daily rainfall) and base usage as inputs.
The model forecasts water demand for a given day given the anticipated weather parameters for that day and the weather parameters and water consumption for the preceding day. The model is linear and the development is based on the method of ‘Recursive Least Squares’. Coefficients of the model are assumed to be time dependent.
The model was used to forecast water demand for Newcastle with a lead time of one day. The model forecasted water demand within an accuracy of plus minus 5percent.