estimation of daily evaporation using of artificial neural networks (case study; borujerd meteorological station)

evaporation is one of the most important components of hydrologic cycle.accurate estimation of this parameter is used for studies such as water balance,irrigation system design, and water resource management. in order to estimate theevaporation, direct measurement methods or physical and empirical models can beused. using direct methods require installing meteorological stations andinstruments for measuring evaporation. installing such instruments in various areasrequires specific facilities and cost which is impossible to be specified. panevaporation is one of the most popular instruments for direct measuring. in thisresearch, by using daily temperature, relative humidity, wind velocity, sunshinehours, and evaporation data in meteorological station and neural network model,daily evaporation is estimated. network training using daily data takes three yearsand network testing takes one year in which data is standardize for training andtesting the model. in this model, a feed forward multiple layer network with ahidden layer and sigmoid function is used. the results show the suitable capabilityand acceptable accuracy of artificial neural networks in estimating of dailyevaporation. best model for estimation of evaporation is ann (5-4-1), it have mse0.006716 and r2 0.725398. artificial neural networking is one of the methods forestimate evaporation. in this method can use in any area that have only maximumand minimum data for estimate evaporation.