Determination of Effective Parameters in Estimating Reference Crop Evapotranspiration Using Artificial Neural Networks (Case study: Lorestan province)

However, several methods exist for calculation of reference crop evapotranspiration (ETo) but the FAO- 56 Penman- Monteith (FAO- 56 PM) method has been recommended by the Food and Agriculture Organization of the United Nations (FAO) as the standard equation. This method is difficult to use because it requires several weather parameters and complex calculations. On the other, over the last decades Artificial Neural Network (ANNs) have shown a good ability for modeling complex and nonlinear systems. The present study was carried out to investigate the sensitivity of the reference crop evapotranspiration to climate parameters using ANNs in Lorestan province. For this purpose in period 10 years (2001 – 2010) daily ETo were calculated using FAO-56 PM method based on weather data daily in the eight weather stations in Lorestan province. Then an Artificial Neural Network was designed with 18 scenarios. Combinations of six weather parameters (maximum and minimum air temperature, maximum and minimum relative humidity, wind speed and daily sunshine hours) which are required to calculate ETo with using FAO-56 PM method were considered as inputs and calculated ETo as output of the ANN in various scenarios. The results of this study showed that increasing the number of data in the input layers will not necessarily lead to improved outcomes of intelligence models. In case of weather data limitation, scenario 13 which was used maximum temperature and wind speed as input layer showed reliable results.