Determination of effective weather variables on pistachio yield using C&R decision tree algorithm

In recent decades, climate variation during the pistachio growing have significantly reduced the quantity nd quslity of yield of this strategic crop of Iran. The purpose of this study is to determine the most important climatic factors affecting pistachio crop yield using C&R decision tree model in Rafsanjan region south of Iran. during the period of 1380-1399. Modeling was performed using climatic variables including wind speed, number of frost days,rainfall,total evaporation,mean relative humidity, sunshine hours, mean temperature, mean of Tmax and Tmin as input variables and pistachio yield as the target variable. Correlation coefficient, root mean square error, relative error and bias metrics were used to evaluate the model. The decision tree model was run separately for 6 months before and sfter harvest. The obtained correlation coefficients were of 0.88 and 0.86, respectively, which is acceptable for prediction of yield. The RMSE values for pre- and post-harvest periods were 521 and 558 Kg ha-1, respectively. According to the results of the decision tree, during the preharvest period, the most significant attributes were the relative humidity and the number of sunny hours, respectively and the for the second half of the year the wind speed, minimum temperature, rainfall and relative humidity were the most affecting variables, respectively.