Navigating Samarinda's climate: A comparative analysis of rainfall forecasting models
Modeling rainfall data is critical as one of the steps to mitigate natural disasters due to weather changes. This research compares the goodness of traditional and machine learning models for predicting rainfall in Samarinda City. Monthly rainfall data was recapitulated by the Meteorology, Climatolo...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-06-01
|
| Series: | MethodsX |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2215016124005314 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Modeling rainfall data is critical as one of the steps to mitigate natural disasters due to weather changes. This research compares the goodness of traditional and machine learning models for predicting rainfall in Samarinda City. Monthly rainfall data was recapitulated by the Meteorology, Climatology, and Geophysics Agency from 2000 to 2020. The traditional models used are Exponential Smoothing and ARIMA, while the machine learning model is a Neural Network. Data is divided into training and testing with a proportion of 90:10. Evaluation of goodness-of-fit using Root Mean Squared Error Prediction (RMSEP). The research results show that the Neural Network has better accuracy in predicting rainfall in Samarinda. Forecasting results indicate that monthly rainfall trends suggest that the months with the highest rainfall occur around November to March. This research provides important implications for developing a warning system for hydrometeorological disasters in Samarinda. The superior points in this research are: • Modeling rainfall data in Samarinda City using several forecasting methods: Exponential Smoothing, ARIMA, and Neural Network. • The Neural-Network algorithm used is Backpropagation with data standardization. • Information about predicted high rainfall can be used to issue early warnings of floods or landslides. Disaster mitigation through policies to regulate water discharge based on rainfall predictions to prevent floods and drought. |
|---|---|
| ISSN: | 2215-0161 |