Evaluating ARIMA Models for Short-Term Rainfall Forecasting in Polewali Mandar Regency

Abstract

This study aims to forecast rainfall in Polewali Mandar Regency using the ARIMA model. This is a quantitative study that uses secondary data, specifically monthly rainfall data (in mm) from January 2008 to December 2020, obtained from the NERC EDS Centre for Environmental Data Analysis. Two ARIMA models were tested: ARIMA(0,1,1)(0,1,1)[12] and ARIMA(1,1,1)(0,1,1)[12], with model selection based on the Akaike Information Criterion (AIC), which balances model fit and complexity. The AIC calculation revealed that the ARIMA(1,1,1)(0,1,1)[12] model had a lower AIC value (1677.33) compared to the ARIMA(0,1,1)(0,1,1)[12] model (1678.16), making ARIMA(1,1,1)(0,1,1)[12] the preferred model. Using this model, the forecasted rainfall for the next five months is as follows: January 2021: 279.8745 mm, February 2021: 238.2206 mm, March 2021: 237.1745 mm, April 2021: 349.3206 mm, and May 2021: 336.0976 mm. These forecasts provide valuable information for water resource management, agricultural irrigation planning, and disaster mitigation related to rainfall. The study emphasizes the importance of selecting the appropriate model to improve forecasting accuracy.