Modeling Tehran municipal water consumption applying seasonal models

Determining future water consumption of cities and planning proper capacity for municipal water supply system (transfer network and purification complexes) revealed the importance of applying behavioral models and forecasting water consumptions in cities. Establishment of public and private administrations and big industries in Tehran's and encompassing trade and political centers beside population growth, showed the necessity of reconstructing municipal water infrastructures. Present study applied SARIMA model for modeling and forecasting municipal water consumption of Tehran city. Hence, using monthly time series of Tehran municipal water consumption during April 1998 till March 2010, seasonal unit root test was applied. Results showed all non-seasonal and seasonal unit roots exist in mentioned monthly time series. Therefore by choosing proper filter, fitting SARIMA models was considered. Considering the mean error of 2.32 percentages and high forecasting power of this model, forecasting municipal water consumption of Tehran city for 2011 could provide constructive results for municipal managers.