Spatiotemporal Analysis of Climate–Forest Fire Correlation in Peninsular Malaysia Using SARIMAX and Ensemble Models
Journal of Data Science and Intelligent Systems · 28.07.2026 · 4 (4) · 489-507
DOI: 10.47852/bonviewjdsis62028499 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Data Science and Intelligent Systems, online erschienen am 28.07.2026. Verfasst von Grace Kian Hwai Ling und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Forest fires in Peninsular Malaysia continue to threaten biodiversity, public health, and economic activity, with risks most evident in peat swamp forests where prolonged dry periods increase fuel flammability. This study analyzes 22 years of fire occurrence (2001–2023) across peat swamp forest regions in Pahang, Selangor, Johor, and Terengganu to examine long-term climate–fire relationships and to develop predictive tools for fire-risk assessment. The modeling framework was organized into three sequential stages. Seasonal Autoregressive Integrated Moving Average with Exogenous Regressors (SARIMAX) was applied for temporal modeling, while an ensemble framework combining Generalized Linear Models and Maximum Entropy estimated fire occurrence probability. A hybrid system subsequently linked SARIMAX-derived climatic projections with ensemble classifiers to generate forward-looking estimates. The results reveal a recurring seasonal pattern, characterized by a clear March peak and noticeably higher fire activity during drought years, including 2005 and 2014. SARIMAX produced the strongest temporal accuracy (average Mean Absolute Scaled Error = 0.1902), demonstrating substantially lower forecast error (around 81%) than a simple seasonal repetition benchmark. Spatial probability modeling showed strong climatic separability at the district–monthly scale (area under the curve = 1; Continuous Boyce Index = 0.67). Projection outputs suggest a gradual redistribution of hotspot intensity from Sabak Bernam toward inland districts such as Pekan, although these estimates remain conditional on projected climatic patterns. Overall, the integrated framework offers a climate-informed foundation for operational monitoring and longer-term management of peat swamp forest fires in Peninsular Malaysia. Received: 28 December 2025 | Revised: 9 June 2026 | Accepted: 17 June 2026 Conflicts of Interest The authors declare that they have no conflicts of interest to this work. Data Availability Statement The data that support the findings of this study are openly available on Zenodo at https://zenodo.org/records/11542164. Author Contribution Statement Grace Kian Hwai Ling: Conceptualization, Methodology, Software, Formal analysis, Investigation, Resources, Data curation, Writing – original draft, Writing – review & editing, Visualization, Project administration. Weng Howe Chan: Conceptualization, Validation, Writing – review & editing, Supervision, Project administration.
Zitieren
APA
Ling, G. K. H., & Chan, W. H. (2026). Spatiotemporal Analysis of Climate–Forest Fire Correlation in Peninsular Malaysia Using SARIMAX and Ensemble Models. Journal of Data Science and Intelligent Systems, 4(4), 489-507. https://doi.org/10.47852/bonviewjdsis62028499
Vancouver
Ling GKH, Chan WH. Spatiotemporal Analysis of Climate–Forest Fire Correlation in Peninsular Malaysia Using SARIMAX and Ensemble Models. Journal of Data Science and Intelligent Systems. 2026;4(4):489-507. doi:10.47852/bonviewjdsis62028499
BibTeX
@article{Ling2026,
author = {Ling, Grace Kian Hwai and Chan, Weng Howe},
title = {Spatiotemporal Analysis of Climate–Forest Fire Correlation in Peninsular Malaysia Using SARIMAX and Ensemble Models},
journal = {Journal of Data Science and Intelligent Systems},
year = {2026},
volume = {4},
number = {4},
pages = {489--507},
doi = {10.47852/bonviewjdsis62028499}
}
EUVIMED