The Future of Data Storytelling for Precipitation Prediction in the Dead- Sea-Jordan Using SARIMA Model

Abstract

This research presents a comprehensive study focused on precipitation prediction for the Dead Sea region utilizing the Seasonal Autoregressive Integrated Moving Average (SARIMA) model. The investigation seeks to interpret the accuracy and reliability of the SARIMA model's predictions by comparing them with predictions derived from climate modeling techniques. The evaluation is based on key performance metrics, including Mean Squared Error (MSE), Mean Absolute Error (MAE), and Root Mean Squared Error (RMSE). Additionally, the paper examines the SARIMA model's predictive capabilities through a comparison with actual observations spanning the period from 2010 to 2022. The obtained results reveal an MSE of 12.84593, an MAE of 2.34407, and an RMSE of 3.584123 for this period. Significantly, the SARIMA model surpasses the predictions of prominent climate models (CMIP6), namely ACCESS_CM2, Earth3_Veg, GISS_E2, and HadGEM3, based on comparative performance assessments. The findings emphasize the robustness of the SARIMA model in capturing the essence of the observations and predicting precipitation patterns, not only through its superior performance against climate models but also through its alignment with actual observations. This study contributes to a deeper understanding of precipitation prediction in the Dead Sea region and underscores the potential of the SARIMA model in enhancing forecasting accuracy for hydrological and climatic investigations