Outlier Detection and Missing Value in Seasonal ARIMA Model Using Rainfall Data

Abstract Forecasting the trend of rainfall is a difficult task in meteorology and environmental sciences. Statistical approaches from time series analysis provide another way for rainfall prediction. The ARIMA model incorporating seasonal characteristics, which is devoted to as seasonal ARIMA model was presented. The time series data are the monthly rainfall data from 2006 to 2016. The model was denoted as SARIMA (1,1, 1) (0, 1, 1) 12 in this study. A serious problem in analyzing rainfall data is what to do when missing or extreme values occur, perhaps as a result of a breakdown in automatic counting equipment. We can analyze the stability and the correlation of the time series. The aim of this paper is to attempt look at ways of explaining this problem by using the residuals from a fitted SARIMA model. The most successful method in finding outliers and unique them from other events, being less expensive than case deletion. In our result, the model fitted the data well and the stochastic seasonal variation was successfully model. Seasonal ARIMA model was a proper method for modeling and forecasting the time series of monthly rainfall data.