This project focuses on advancing high-resolution wildfire modeling for Korea to better understand, predict, and manage the growing risks associated with climate change. By integrating advanced fire behavior algorithms with human and socio-environmental factors, the project aims to deliver new scientific insights and provide operationally relevant tools to support evidence-based wildfire policy and management.
© Nui Malama | Pexels
Gangwon Province, with over 82% of its area covered by forests, has historically experienced some of Korea’s most devastating wildfires. Due to its ecological and geographical characteristics, Gangwon plays a critical role in Korea's wildfire management efforts.
Wildfire occurrence in Korea is driven by a combination of natural and human-induced factors. Dry spring conditions, strong seasonal winds, agricultural burning, forestry activities, and other anthropogenic drivers significantly increase wildfire risk. The increasing frequency and intensity of recent wildfire events highlight the urgent need for advanced analytical frameworks and predictive modeling tools tailored to Korea’s regional conditions.
The project is led by the FLAM team within the Advanced Forest Ecosystems (AFE) Group and seeks to strengthen Korea’s capacity to anticipate extreme wildfire events, improve long-term risk projections, and support adaptive responses under changing climatic and socio-economic conditions.
We seek to lay the groundwork for regionally tailored wildfire prediction and management models that can later be scaled up for national application. The project is designed to incorporate socio-economic factors and climate projections into wildfire risk assessments to support adaptive responses to climate change.
The project integrates high-resolution wildfire spread modeling with climate, environmental, and socio-economic drivers to improve understanding of wildfire dynamics and enhance predictive capability. The work is centered on the development and application of a regionally tailored FLAM model for Korea, with a focus on extreme fire behavior, spotting, and re-ignition processes.
Key objectives include:
- Developing methodologies to incorporate human and socio-economic factors into the FLAM framework.
- Applying the integrated FLAM model to simulate long-term climate change impacts and adaptation scenarios, thereby strengthening strategic wildfire planning and response capabilities.
- Introducing stochastic simulation concepts into FLAM to reproduce extreme wildfire events.
- Harmonizing high-resolution GIS datasets to incorporate anthropogenic activities into the wildfire frequency and magnitude predictions.
A central technical innovation of the project is the explicit representation of ember-driven spotting and secondary ignitions. This work addresses current limitations in operational wildfire models, which typically lack the ability to simulate repeated spotting events and cascading fire spread.
Key technical enhancements include:
- Develop an algorithm in Python to probabilistically estimate spotting frequency and distance under varying environmental conditions
- Implement a re-ignition probability module at spotting locations, utilizing ignition algorithms from IIASA’s FLAM model
- Address the current limitations of the KFS model, which lacks the ability to simulate ember-induced re-ignition and cascading fire spread
- Leverage the FLAM model’s ignition probability functions, which have been validated in global and national-scale studies, to enhance the Korean model's capability to simulate fire propagation across firebreaks and non-burnable areas
- Provide improved fire behavior prediction under dry and high-risk conditions through simulation of repeated spotting events
Partners
- OKTA Holdings
- Jeonju University
- Korea Forest Fire Management Service Association (External Consultant)
- East Coastal Forest Fire Center of Gangwon Province (External Consultant)
This project has been funded by Korea Forest Service (KFS).