<![CDATA[Rainfall in West Java during the rainy season is quite high. This causes the area around the watershed to experience flooding. However, in the dry season the Citarum watershed experiences drought. Changes in the Citarum river water discharge from time to time is not only influenced by time but also influenced by the location around it. To forecast the Citarum river water discharge data, the Space Time Autoregressive Integrated (STARI) model can be used. In this study, the STARI model was applied to the Citarum river water discharge data at all four observation sites. Based on the stationary data, it showed that the data is not stationary, so the differencing process must first be done 1 time. The identification of the order of the AR model was one because the PACF plot was truncated in lag 1. The spatial lag used in this study was the spatial lag of order 2, so the Citarum river water discharge could be predicted with the STARI model. Estimation of STARI) model parameters with a uniform weight matrix was ​estimated by the MLE method with the help of R and S-Plus 8.0 softwares. STARI model with MAPE less than 10% was used for predicting Citarum river water discharge data for the four observation locations, thus the STARI model can be recommended to predict Citarum river water discharge data.]]>
Copyrights © 2020
