Forecasting of Tea Yield Based on Energy Inputs using Artificial Neural Networks (A case study: Guilan province of Iran)

The objective of this study was the exploring relation between energy inputs and tea yield using artificial neural network (ANN) in the Guilan province of Iran. For this purpose, the energy use pattern was determined by collection data from 30 tea farmers using face-to-face questionnaire method in the many village of studied region. The results indicated the tot al energy consumption and yield of tea production were 46144.04 MJ ha -1 and 8419.47 kg ha -1 , respectively. The highest share of energy consumption was belonged to nitrogen with 50.84%. In this study, the energy indices covering energy use efficiency, energy productivity, specific energy and net energy were calculated at 0.18, 0.23 kg MJ -1 , 4.38 MJ kg -1 and -37724.57 MJ ha -1 , respectively. Moreover, the share of energy forms including direct, indirect, renewable and non -renewable energies was found to be as 42.96%, 57.04%, 28.34% and 71.66%, respectively. For forecasting of tea yield based on energy inputs, ANN model developed by Back propagation algorithm in this study. The results illustrated the ANN model with 7-13-13-1 architecture had the best condition for predict of tea yield. With respect to ANN model, R2, RMSE and MAPE was computed as 0.968, 0.105 and 0.006, respectively. In the last section of this study, sensitivity analysis was applied by ANN for robustness of evaluated mode. The results disclosed the farmyard manure had the highest rate of sensitivity among all inputs.