Simulating climate changes in Iran using artificial neural networks algorithm

Temperature and precipitation are two important meteorological parameters, especially in arid and semi-arid regions. As a result, it is necessary to know the amount of these parameters, their changes and predict these phenomena, in order to have more accurate planning in the management of agricultural, economic and etc. The aim of this study is numerical simulation and forecasting of Iran's weather changes with emphasis on two climatic parameters of temperature and precipitation. The method used in this research is artificial neural network algorithm for simulating 24-hour temperature and precipitation variables by month during a 31-year period compared to the base period. Factorial variance analysis has been used to compare the average changes in temperature and precipitation during the observed period of 1990-2020 and the simulated period of 2020-2050. The comparison results show that there is no significant difference between temperature changes and precipitation in 24-hour selected meteorological stations in Iran. The computed Partial Eta Squared is 10.6 percent for temperature and 5.7 percent for precipitation. The highest and lowest precipitation observed occurred in the months of March and July. The simulations have predicted the highest and lowest precipitation for the months of March and August. Also, the highest and lowest observed temperatures occurred in the months of July and January, respectively, and the simulations predicted the same values for the same months. The highest and lowest observed precipitation values were recorded in Rasht 109.95 and Yazd 4.36 mm, respectively. The simulations predicted the highest and lowest amount of precipitation for the same stations as 112.46 and 5.63 mm. Also, the highest and lowest observed temperatures were recorded at Bandar Abbas 26.99 and Ardabil 9.36 Celsius degrees, respectively. The simulations have predicted these values 27.10 and 9.45 Celsius degrees, for the same stations, respectively.