Predicting spatio-temporal wildfire propagation with dynamic firebreaks
Abstract
Wildfire management strategies increasingly demand accurate predictive models that integrate real-time intervention measures. Despite advances in machine learning (ML) for wildfire modelling, existing approaches largely overlook the role of firebreak placement. In this work, we present the first deep learning-based predictive model for simulating spatio-temporal wildfire propagation with dynamic firebreaks. Utilizing a Convolutional Long Short-Term Memory (ConvLSTM) architecture, the model captures both the spatial and temporal complexities of wildfire spread while incorporating data on firebreak positioning and effectiveness. Our training dataset, derived from Cellular Automata (CA) simulations, integrates key geophysical parameters and human intervention strategies, including temporary and permanent firebreaks. Model validation across three major wildfire events in California demonstrates robust performance, with significant accuracy gains in scenarios involving strategic firebreak placement. This integration of movable firebreak placement into a wildfire spread model provides a tool for improving real-time wildfire management efforts.
Related
- Explainable AI with spatio-temporal transition modelling for predicting urban land use transitions in the Kolkata metropolitan area
- Wildfire risk for main vegetation units in a biodiversity hotspot: modelling approach in New-Caledonia, South Pacific.
- A machine learning model for early season wildfire risk in Ghana’s Guinea savannah assessing discrimination, operational prevalence, spatial transferability, and CMIP6 climate scenarios
- FireDataForge: A Unified Framework for Multi-Source Wildfire Data Retrieval and Integration
- The July 2026 North American wildfire smoke event was three times deadlier due to climate change
BibTeX
@article{zheng2026predicting,
title = {{Predicting spatio-temporal wildfire propagation with dynamic firebreaks}},
author = {Jiahe Zheng and Zhengsen Xu and Rossella Arcucci and Sandy P Harrison and Lincoln Linlin Xu and Sibo Cheng},
journal = {Natural Hazards and Earth System Sciences},
year = {2026},
month = jun,
doi = {10.5194/nhess-26-2871-2026},
url = {https://hal.science/hal-05838431v1},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Predicting spatio-temporal wildfire propagation with dynamic firebreaks
AU - Jiahe Zheng
AU - Zhengsen Xu
AU - Rossella Arcucci
AU - Sandy P Harrison
AU - Lincoln Linlin Xu
AU - Sibo Cheng
PY - 2026
DA - 2026/06/17
JO - Natural Hazards and Earth System Sciences
DO - 10.5194/nhess-26-2871-2026
UR - https://hal.science/hal-05838431v1
AB - Wildfire management strategies increasingly demand accurate predictive models that integrate real-time intervention measures. Despite advances in machine learning (ML) for wildfire modelling, existing approaches largely overlook the role of firebreak placement. In this work, we present the first deep learning-based predictive model for simulating spatio-temporal wildfire propagation with dynamic firebreaks. Utilizing a Convolutional Long Short-Term Memory (ConvLSTM) architecture, the model captures both the spatial and temporal complexities of wildfire spread while incorporating data on firebreak positioning and effectiveness. Our training dataset, derived from Cellular Automata (CA) simulations, integrates key geophysical parameters and human intervention strategies, including temporary and permanent firebreaks. Model validation across three major wildfire events in California demonstrates robust performance, with significant accuracy gains in scenarios involving strategic firebreak placement. This integration of movable firebreak placement into a wildfire spread model provides a tool for improving real-time wildfire management efforts.
SN - 1561-8633
SN - 1684-9981
ER -