Estimation of pomegranate evapotranspiration for orchard management using artificial neural network

The study has been undertaken to investigate the utility of artificial neural networks (ANNs) for comparison of weekly values of pomegranate evapotranspiration (ETp) estimated by ETr and Kc approach and that of estimated by ANNs for 1st to 5th year pomegranate orchards. Feed forward network has been used for prediction of ETr using resilient back-propagation method. For the purpose of this study, the meteorological data of 25 years from 1983 to 2007 were used as input. For training of ETp, crop coefficient values were also taken as input to the networks along with meteorological parameters. The results revealed that, the values of correlation coefficient between estimated ETp using crop coefficient approach and ANN estimated ETp which are 0.998, 0.999, 0.9999, 1 and 1 for 1st year to 5th year pomegranate orchards, respectively and the value of MSE 0.0021, 0.0128, 0.0628, 0.0043 and 0.0001 for 1st year to 5th year pomegranate orchards, respectively. Based on comparison, it can be concluded that in the study ANN model is suitable for prediction of ETp.