IMPLEMENTATION OF MACHINE LEARNING FOR RAINFALL PREDICTION IN SMOKE-PRONE AREAS OF SOUTH SUMATRA
Article Sidebar
Main Article Content
Abstract
Haze caused by forest and land fires is a serious problem in South Sumatra Province. One mitigation effort that can be made is to improve the accuracy of rainfall predictions, because rain plays an important role in reducing the potential for fires. This study implements machine learning methods, namely XGBoost and ConvLSTM, to predict spatiotemporal rainfall in areas prone to haze. The results show that ConvLSTM is capable of providing better predictions than the baseline, especially during periods of haze, by considering missing data imputation and masking techniques for disrupted satellite conditions. Increasingly apparent climate change in tropical regions has had a significant impact on rainfall patterns, particularly in South Sumatra, which is one of Indonesia's main agricultural and plantation centers. High rainfall variability can lead to the risk of flooding and drought, as well as disrupting productivity in the education, health, and economic sectors. Therefore, a more accurate rainfall prediction approach is needed to support climate adaptation planning and disaster risk mitigation. This study aims to compare the performance of three approaches to daily rainfall prediction, namely the ConvLSTM-based method, XGBoost, and Persistence, using daily observation data from BMKG for the South Sumatra region for the period 1981–2020. The input variables include average air temperature (Tavg), humidity, sunshine duration, and wind speed, while rainfall is used as the prediction target. The analysis was conducted through a time series approach, statistical distribution, and model performance evaluation using the quantitative metrics Root Mean Square Error (RMSE) and Critical Success Index (CSI). The results show that the ConvLSTM model produced the highest accuracy with an average RMSE of 10 mm/day and a CSI of 0.53, which is better than XGBoost (RMSE 12 mm/day; CSI 0.48) and the persistence method (RMSE 15 mm/day; CSI 0.40). Distribution analysis indicates that light to moderate rainfall occurs more frequently, while extreme rainfall occurs sporadically. The correlation heatmap shows that rainfall has a moderate positive relationship with humidity and a negative relationship with solar radiation, while average temperature and wind play a smaller role. The main contribution of this study is to provide empirical evidence that spatiotemporal deep learning methods such as ConvLSTM are superior in modeling the complexity of tropical rainfall dynamics compared to classical machine learning approaches and simple models. These findings can serve as a basis for the development of early warning systems and interactive climate dashboards at the regional level, while enriching the literature on rainfall prediction in tropical regions using an integrative approach.
Downloads
Article Details
How to Cite
References
Ali, K., Arsyad, S., & Hendrawan, B. (2025). Mapping and visualization of research on climate change adaptation in Indonesia: Using Scopus data (2000–2023). Journal of Advanced Public Health, 11(1), 24–33.
Ariska, M., Akhsan, H. (2023). Spatiotemporal modeling of rainfall anomalies in Sumatera using ensemble machine learning approaches. International Journal of Climatology, 43(12), 5021–5035.
Ariska, M., Akhsan, H., & Muslim, M. (2022). Impact profile of ENSO and dipole mode on rainfall as anticipation of hydrometeorological disasters in the province of South Sumatra. Spektra: Jurnal Fisika dan Aplikasinya, 7(3), 127–140.
Ariska, M., Irfan, M., & Iskandar, I. (2024). Detection of dominant rainfall patterns in Indonesian regions using empirical orthogonal function (EOF) and its relation with ENSO and IOD events. Science and Technology Indonesia, 9(4), 1009–1023.
Ariska, M., Suhadi, S., & Herlambang, D. K. (2023). Empirical orthogonal function (EOF) analysis based on Google Colab on sea surface temperature (SST) dataset in Indonesian waters. Indonesian Physical Review, 6(1), 20–32.
Ariska, M., Suhadi, S., Supari, S., Irfan, M., & Iskandar, I. (2024). The effect of El Niño Southern Oscillation (ENSO) on rainfall and correlation with consecutive dry days (CDD) in Palembang city. AIP Conference Proceedings, 3052(1).
Ariska, M., et al. (2022). Prediksi perubahan iklim ekstrem di kota Palembang dan kaitannya dengan fenomena El Niño-Southern Oscillation (ENSO) berbasis machine learning. JIPFRI, 6(2), 79–86.
Ariska, M., et al. (2023). Pemodelan numerik hubungan pola curah hujan wilayah equatorial di Pulau Sumatera terhadap fenomena ENSO dan IOD. Jurnal Teori dan Aplikasi Fisika, 95–106.
Ariska, M., et al. (2023). Trend of rainfall pattern in Palembang for 20 years and link to El-Niño Southern Oscillation (ENSO). Jurnal Ilmiah Pendidikan Fisika Al-Biruni, 12(1), 67–75.
Ariska, M., et al. (2023). Efectivity of a hybrid classroom-based reflective microlearning model for prospective physics teachers. Jurnal Pendidikan MIPA, 24(3), 710–716.*
Ariska, M., et al. (2024). Education for sustainable development based of technological pedagogical and content knowledge using mixed-methods approach in physics teaching. Jurnal Penelitian dan Pengembangan Pendidikan Fisika, 10(2), 421–434.
Ariska, M., et al. (2024). Exploration of seasonal dynamics of tropical Indo-Pacific Ocean during dry, wet and neutral years in Indonesia using composite method. Jurnal Penelitian Fisika dan Aplikasinya (JPFA), 14(1).
Bochenek, B., & Ustrnul, Z. (2022). Machine learning in weather prediction and climate analyses—applications and perspectives. Journal of Environmental Climate Research, 13(2), 180. https://doi.org/10.3390/atmos13020180
Breiman, L. (2001). Random forests. Machine Learning, 45(1), 5–32.
Chen, L., & Han, B. (2023). Machine learning methods in climate prediction. Journal of Climate Science and Technology, 29(5), 500–525. https://doi.org/10.3390/app132112019
Chen, L., & Han, B. (2023). Machine learning methods in weather and climate survey. Applied Sciences, 13(2), 1201. https://doi.org/10.3390/app132112019
Darmastowo, M., Putra, M., Handoko, D., & Rosid, M. S. (2023). Rainfall estimation in equatorial region using weather radar-based machine learning. In Proceedings of the 2023 International Seminar on Applied Technology, Information and Communication (iSemantic) (pp. 266–270).
Dayal, K., & Deo, R. (2023). Application of hybrid neural networks for rainfall estimation in Japan. Hydrometeorology, 28(2), 200–218. https://ieeexplore.ieee.org/document/7848588
Derot, J. (2024). Improved climate time series forecasts by machine learning using signature methods. Environmental Modelling & Software, 168, 105731.
Djajadi, M. I. J., & Phinn, S. (2025). Machine learning and Bayesian method for monitoring and forecasting Jayawijaya’s tropical glacier change. In Proceedings of Climate Change AI.
Ferreira, S., & Wanderley, H. (2025). Changes in the distribution of precipitation with the potential to cause extreme. Atmosphere, 16(4), 358. https://doi.org/10.3390/atmos16040358
Hikouei, I. S. (2023). Using machine learning algorithms to predict groundwater level changes across an Indonesian tropical peatland. Science of the Total Environment, 872, 162312.
Lestari, P., & Nurrahman, R. (2022). Application of XGBoost for rainfall intensity prediction in tropical climate regions. Journal of Hydrology: Regional Studies, 41, 101089.
Muhammad, F. R., Lubis, S. W., & Setiawan, S. (2020). Impacts of the Madden–Julian Oscillation on precipitation extremes in Indonesia. Atmospheric Research, 246, 105129.
Niculae, A. M., Pascu, A. M., & Stefan, A. (2024). Assessing the role of machine learning in climate change research: A bibliometric and thematic analysis. Sustainability, 16(24), 11086.
Nugrahani, E. H., et al. (2024). Sensitivity and feature importance of climate factors and evaluation of different machine learning models for predicting fire hotspots in Kalimantan, Indonesia. International Journal of Artificial Intelligence, 13(2), 2212–2225.
Pratama, R., & Wiratama, A. (2024). Comparative analysis of ANN and LSTM for predicting extreme rainfall in equatorial Indonesia. Journal of Water and Climate Change, 15(2), 329–339.
Putra, A. P. S., Andriani, N., Ritonga, A. F., & Ariska, M. (2024). Analisis curah hujan di Papua Barat menggunakan metode Empirical Orthogonal Function (EOF). Jurnal Online Fisika, 9(3), 26–31.
Puspasari, R. L., Yoon, D., Kim, H., & Kim, K. W. (2023). Machine learning for flood prediction in Indonesia: Providing online access for disaster management control. Economic and Environmental Geology, 56(1), 65–73.
Rahmawati, D., & Nugroho, Y. (2023). Long-term sea level rise prediction using deep learning models in the Indonesian maritime continent. Ocean & Coastal Management, 239, 106478.
Rasyid, M., & Hadi, M. P. (2020). Predicting El Niño-Southern Oscillation impacts on rainfall using machine learning models in Eastern Indonesia. Theoretical and Applied Climatology, 142(3), 1273–1285.
Saputra, R., & Purnomo, H. (2021). Climate variability and flood hazard mapping in Sumatra using machine learning. Natural Hazards, 108(2), 1281–1298.
Silva, T., & Santos, C. (2022). Climate indices-based analysis of rainfall spatiotemporal variability. Water, 14(4), 2190. https://doi.org/10.3390/w14142190
Siregar, H., & Nabila, R. (2022). Evaluation of machine learning models for predicting tropical cyclone occurrences in Indonesian waters. Meteorological Applications, 29(5), 2105.
Sun, H., & Yin, H. (2024). Integrating the functions ecological network sustainability under climate change scenarios. Climate Change Journal, 39(2), 400–420. https://doi.org/10.1007/s10980-024-01799-y