Investigating the effect of climate fluctuations on apple yield (case study: Semiram and Urmia station)

the statistics of the ECMWF database were used for the observation data of the two stations of Semiram and Urmia during a 21-year period (1996-20016).In order to investigate the effects of climate fluctuations, the daily data of dynamic micro-rotation of the CORDEX project was used for the output of the ICHEC-EC-EARTH model under the RCP8.5 and RCP4.5 radiative forcing (RCP) scenarios for the period (2017-2037). By using the data of the stations and the outputs of the micro scale model, and by using the perceptron neural network and linear regression, the performance was simulated. Then, to evaluate the efficiency of the models, R, R2, MSE, RMSE, and NRMSE statistics were used, and the non-parametric Menkendall test and age slope were used for the performance trend. The result of comparing the output of artificial neural networks with the linear regression model shows that the error rate of the neural network is less and the simulated results are close to the real observations to a very acceptable extent. The phenological stages, including bud blooming to fruit ripening in the stations under both scenarios, and in all the phenological stages in the future period will be completed earlier than the base period, and the length of the growth period will also decrease. The amount of future yield in Urmia station under RCP4.5 and RCP8.5 scenarios respectively yield 3.7 and 2.2 tons per hectare and in Semiram station yield 1.4 and 3 respectively tons per hectare will decrease. The results show that in the future in the study areas, with the change in the time of occurrence of the length of the growth period, all the phenological stages as well as the declining performance of apple trees will be subject to climatic fluctuations