Iran Meteorology Society (IMS) Journal of Meteorology and Atmospheric Science 2645-7261 8 Number 2 2025 12 22 Rainfall Analysis and Modeling Using the ARIMA Model (Case Study: Ahvaz) Rainfall Analysis and Modeling Using the ARIMA Model (Case Study: Ahvaz) 92 111 250624 10.22034/jmas.2026.576251.1266 FA Leila Mokaram Yazd University - Department of Geography 0009-0009-0947-9162 Ahmad Mazidi Associate Professor of Climatology, Department of Geography, Yazd University, Yazd, Iran, 0000-0003-4558-9907 Journal Article 2026 02 16 Abstract<br>Precipitation is one of the most important components of the water cycle and is considered one of the most important input components to the hydrological cycle, which plays a very important role in assessing the climatic characteristics of any region. Rainfall forecasting plays a very important role in preparedness, management, and prevention of events resulting from natural hazards. In this research, to forecast the annual rainfall of the Ahvaz synoptic station, annual precipitation data of the Ahvaz station for the statistical period of 1970 - 2019 were used. To forecast rainfall behavior, the ARIMA modeling method was utilized. To check the stationarity of the model, autocorrelation functions (ADF) and partial autocorrelation functions (PACF) were used. The best ARIMA model for forecasting the rainfall of the Ahvaz station was identified as the ARIMA (3,1,1) model. The rainfall forecast for the coming years indicates a downward trend in rainfall for the Ahvaz station. Also, in this research, to forecast the annual rainfall of the Ahvaz station, the Random Forest (RF) machine learning method was utilized, which, considering the time series of the data, had lower accuracy compared to the ARIMA model. The purpose of this model is to compare with the ARIMA time series in order to forecast annual rainfall at the Ahvaz station. Abstract<br>Precipitation is one of the most important components of the water cycle and is considered one of the most important input components to the hydrological cycle, which plays a very important role in assessing the climatic characteristics of any region. Rainfall forecasting plays a very important role in preparedness, management, and prevention of events resulting from natural hazards. In this research, to forecast the annual rainfall of the Ahvaz synoptic station, annual precipitation data of the Ahvaz station for the statistical period of 1970 - 2019 were used. To forecast rainfall behavior, the ARIMA modeling method was utilized. To check the stationarity of the model, autocorrelation functions (ADF) and partial autocorrelation functions (PACF) were used. The best ARIMA model for forecasting the rainfall of the Ahvaz station was identified as the ARIMA (3,1,1) model. The rainfall forecast for the coming years indicates a downward trend in rainfall for the Ahvaz station. Also, in this research, to forecast the annual rainfall of the Ahvaz station, the Random Forest (RF) machine learning method was utilized, which, considering the time series of the data, had lower accuracy compared to the ARIMA model. The purpose of this model is to compare with the ARIMA time series in order to forecast annual rainfall at the Ahvaz station. https://www.ims-jmas.net/article_250624_c80c35d1aaf7e81ec44cc8f605fd6903.pdf