عنوان مقاله [English]
Water is one of the most importance sources for provision of the human needs that plays a vital role in current life. Therefore awareness about water demand is important for necessary policy making for demand management. In this paper we modeled Tehran water daily demand by nonlinear Artificial Neural Network and linear process of ARMA for a 7 years period and forecasted the urban water daily demand for 10 days. The effective factors on urban water daily demand in designing of neural network are temperature (minimum, average, maximum), week days, holidays and special days. Results present that artificial neural network have higher power than ARMA in forecasting the Tehran city water daily demand on the basis of indicators of forecast accuracy valuation.