Forecasting Rainfall in Pangkalpinang City Using Seasonal Autoregressive Integrated Moving Average With Exogenous (Sarimax)

Abstract

Changes in extreme rainfall can cause disasters or losses for the wider community, so information about future rainfall is also needed. Rainfall is included in the category of time series data. One of the time series methods that can be used is Autoregressive Integrated Moving Average (ARIMA) or Seasonal ARIMA (SARIMA). However, this model only involves one variable without involving its dependence on other variables. One of the factors that can affect rainfall is wind speed which can affect the formation of convective clouds. In this study, the ARIMA model was expanded by adding eXogen variables and seasonal elements, namely the SARIMAX model (Seasonal ARIMA with eXogenous input). Based on the analysis that has been carried out, the prediction of rainfall in Pangkalpinang City, Bangka Belitung Islands Province can be modeled with the SARIMAX model (0,1,3)(0,1,1){12} for monthly rainfall and SARIMAX (0,1,2 )(0,1,3){12} for maximum daily rainfall. When compared with the actual data and previous studies using ARIMAX, the SARIMAX model is still better in the forecasting process when compared to the ARIMAX model. If viewed based on the AIC value of the SARIMA model, the SARIMAX model is also more suitable to be used to predict rainfall in Pangkalpinang City.

Predicting particulate matter PM2.5 using the exponential smoothing and Seasonal ARIMA with R studio

“…2.2. The Seosenal Autoregressive Integrated Moving Average (Seasonal ARIMA ) Seasonal ARIMA differs from ARIMA models in that it contains seasonal characteristics of time series [7], and is an extension of ARIMA model [11]. The notation commonly used for Seasonal ARIMA is [6,7,12,13]:…”