Wildfire Susceptibility Mapping in China Combining Machine Learning, Deep Learning, and Transformer-Based Models
Earth · 13.07.2026 · 7 (4) · 119
DOI: 10.3390/earth7040119 ↗Kurz gefasst
Fachartikel in der Zeitschrift Earth, online erschienen am 13.07.2026. Verfasst von Uroš Durlević und 7 weiteren Autor:innen. Die Arbeit nennt 131 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Ministry of Science, Technological Development and Innovation, RUDN University.
Abstract (Originalsprache)
Long-term wildfire susceptibility mapping represents a significant component of disaster prevention and the protection of human communities, public health, and local ecosystems. In this study, a wildfire inventory was developed through multi-sensor fusion of satellite data (MODIS and VIIRS), comprising 153,305 fire events across China for the period 2001–2024. In addition to historical incidents, 14 predictive variables were processed, representing geomorphological, climatological, hydrological, vegetative, and anthropogenic conditions. This study evaluates long-term spatial wildfire susceptibility based on long-term mean environmental and climatic conditions. Methodologically, the research applies six models from machine learning (ML), deep learning (DL), and transformer-based approaches: Random Forest (RF), Extreme Gradient Boosting (XGBoost), Deep Neural Network (DNN), Fourier Multi-Layer Perceptron (F-MLP), Kolmogorov–Arnold Network (KAN), and Feature Tokenizer (FT) Transformer. The results were integrated into an ensemble susceptibility map with a spatial resolution of 500 m using Geographic Information Systems (GIS), indicating that 7.4% of China’s territory is classified as having a very high wildfire susceptibility. In addition to the national-scale assessment, a local differentiation was conducted across 34 province-level divisions, revealing that Fujian Province (86.8%) and the Guangxi Zhuang Autonomous Region (82.9%) had the largest shares of areas classified as high and very high wildfire susceptibility. Performance evaluation under spatial block-based validation demonstrated that the Random Forest model achieved the highest predictive power, with an area under the curve (AUC) of 87.8%, followed by XGBoost (87.3%) and Fourier MLP (86.6%). Based on the combined SHAP (Shapley additive explanations) analysis of all applied models, soil moisture, elevation, and terrain slope were identified as the most influential factors affecting wildfire occurrence in China. Overall, the findings contribute to more effective wildfire prevention and risk management strategies at both the local and national levels.
Förderung
- Ministry of Science, Technological Development and Innovation
- RUDN University
Zitieren
APA
Durlević, U., Ilić, V., Radovanović, M. M., Pešić, A. M., Petrović, M. D., Milenković, M., Jovanović, J. M., & Atasoy, E. (2026). Wildfire Susceptibility Mapping in China Combining Machine Learning, Deep Learning, and Transformer-Based Models. Earth, 7(4), 119. https://doi.org/10.3390/earth7040119
Vancouver
Durlević U, Ilić V, Radovanović MM, Pešić AM, Petrović MD, Milenković M, et al. Wildfire Susceptibility Mapping in China Combining Machine Learning, Deep Learning, and Transformer-Based Models. Earth. 2026;7(4):119. doi:10.3390/earth7040119
BibTeX
@article{Durlevi2026,
author = {Durlević, Uroš and Ilić, Velibor and Radovanović, Milan M. and Pešić, Ana Milanović and Petrović, Marko D. and Milenković, Milan and Jovanović, Jasmina M. and Atasoy, Emin},
title = {Wildfire Susceptibility Mapping in China Combining Machine Learning, Deep Learning, and Transformer-Based Models},
journal = {Earth},
year = {2026},
volume = {7},
number = {4},
pages = {119},
doi = {10.3390/earth7040119}
}
EUVIMED