Abstract

Natural calamities like floods and droughts are most frequently caused by waterlogging. The organizations that needs to make decisions in the case of a disaster or incident prevention might benefit from using weather forecasts. The weather prediction includes a forecast for rainfall. The prediction of rainfall has been approached in a number of ways, including physical, statistical, and hybrid approaches. The objective of this study is to predict the amount of rainfall using the best forecasting model, and then to evaluate that model's performance. Karbala was chosen as the research area and the monthly rainfall data were taken from the Iraqi Meteorological Department and covered the years from January 2012 to December 2023. It was discovered that Karbala's rainfall pattern had a trend and seasonality throughout the year. To predict the rainfall data in this study, the Holt-Winters, Box-Jenkins, and hybrid methods were suggested. Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), and Mean Absolute Error (MAE) are performance measures that were used to assess the models' performance. It was discovered that the Holt-Winters model yields the most precise result and can be used to predict future rainfall.