TIME SERIES MODELING WITH SEASONAL AND EXTERNAL FACTOR ON RAINFALL DATA AT KEMAYORAN SOUTH JAKARTA
Rainfall is part of our daily lives that gives positive and negative impacts on our lives. I????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/42743 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | Rainfall is part of our daily lives that gives positive and negative impacts on our lives. I????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????reason, accurate prediction of rainfall is necessary as it will give us a wider knowledge for the upcoming events in this world.
This research is focused on predicting the rainfall in Kemayoran, South Jakarta. It is composed of two models. The first model is formed through a time series model, Seasonal Autoregressive Integrated Moving Average (SARIMA). This model is based on the monthly rainfall data of Kemayoran that ranges between January 2000 up to June 2019 (222 observations). The creation of this model was formed by conducting the assumption of stationarity. The stationarity was achieved by transforming the rainfall data using the square root transformation. Based on the results, it is concluded that the model SARIMA (2,0,2)(2,0,1)12 produced the best estimated and predicted rainfall data.
In addition, external factors such as temperature, humidity, sunlight, and wind velocity were studied using analysis of correlation to find out if these factors affect the rainfall and to determine the behavior between each factors. The second model, which is the regression model, was formed by using the stepwise regression method, being the rainfall as the response, and the SARIMA combined with the external factors as the predictor. The results indicated that rainfall is affected by temperature and humidity.
To get the final result, the regression model was compared with the SARIMA model. The comparison showed that the regression model is well-estimated and represented with accurate prediction, therefore, making the regression model the best one. |
|---|