PREDIKSI AWAL MUSIM HUJAN DI WAINGAPU PROVINSI NUSA TENGGARA TIMUR MENGGUNAKAN METODE ARIMA

Abstract

The beginning of the rainy season can be predicted by various methods such as the Autoregressive Integrated Moving Average (ARIMA). Occurrence of the onset of the rainy season (AMH), the erratic impact on various sectors, especially in the agricultural sector often results in crop failure. Therefore, the aim of this study is to improve the accuracy of predictions for the start of the rainy season. In this study daily rainfall data, the beginning of the rainy season data is obtained by calculating daily rainfall data using the Liebmann method. The best ARIMA model (3,1,0) with the equation y = 0,3162−1 + 0,1284−2−0,188−3 + 0,7434−4−0,934 is used for July, August and DMI data in August are considered as input and prediction error value ARIMA as a target. The beginning of the rainy season prediction results based on ARIMA, the results of testing and evaluation obtained values ​​of r = 0.14 and RMSE = 32.53.