Prediction of El Niño La Niña & Indian Ocean Dipole Phenomena Using the ARIMA Model by Prospective Teachers in Indonesia

Abstract

The high number of losses and casualties experienced by the Indonesian people is caused by various factors. One of the factors lies in the people who lack knowledge in dealing with climate phenomena. Student-teacher candidates must take an important role in solving climate problems. One solution is to predict climate phenomena. This study uses the ARIMA model to predict the El Niño La Niña & Indian Ocean Dipole phenomenon. The research method is the ARIMA model prediction steps using student worksheets. (1) know the model by determining ACF and PACF. (2) parameter model estimation determines ARIMA on predetermined ACF and PACF. The ARIMA parameter model is the slightest Mean Squared Error (MSE) level. (3) the diagnostic model of the smallest MSE. (4) a measure of predictive accuracy. And (5) Argumentation. Results In the IOD calculation, the correlation between the three-month prediction period is 0.92 and for six months is 0.96. This is the same as in the Nińo 3.4 correlation data. The value is close to 1, which is very accurate and can be used. The predicted value of the IOD index was expected. The resulting values ​​ranged from 0.01 to 0.17. This means that students are proficient in making climate predictions. Because students are enthusiastic and passionate about studying climate, it is a new science