Comparisons among rainfall prediction of monthly rainfall basis data in Aceh using an autoregressive moving average

Abstract

Climate variability especially rainfall is an important factor in observing climate change. Extreme weather events can disrupt the rice planting calendar system which ultimately causes crop failure. The planting season for each region depends on the rainfall of an area. The purpose of this study is to determine the comparison of the predicted rainfall results in Sabang, Aceh Besar, and Aceh Tengah Municipality. Historical data for 1988-2015 were taken from BMKG. The ARIMA model was divided into three groups, namely autoregressive (AR), moving average (MA), and ARIMA (autoregressive moving average) models. The best prediction model was seen from the model that has the smallest standard error estimate (S) value. The results shows that the three regions have different rainfall than the seasonal plot visualization. The highest peaks of rainfall occurred on October and November in Aceh Besar, on November and December in Aceh Tengah, and only on December in Sabang (Weh Island). The best model for predicting rainfall for January 2016-December 2020 in Sabang was ARIMA (1,0,0) (2,0,0)[12]. The prediction model for the location of Aceh Besar in the future was ARIMA (0,0,0) (2,0,0)[12]. In Aceh Tengah, the best model for predicting January 2016-December 2020 rainfall was ARIMA (1,0,0) (2,0,2)[12]. Forecasting using ARIMA model was good for short-term forecasting, while for long-term forecasting, the accuracy of the forecasting was not good because the trends of rainfall was flat.