Wildfire spread prediction using geostationary satellite observation data and directional ROS adjustment factor

Area of Science:

• Wildfire Science
• Remote Sensing
• Computational Modeling

Background:

• Accurate wildfire spread prediction is crucial for effective response and resource allocation.
• Existing simulators often face limitations in real-time accuracy due to data integration challenges.
• Geostationary satellite data offers near-real-time observations valuable for dynamic wildfire modeling.

Purpose of the Study:

• To develop a data-driven approach for enhancing regional wildfire spread prediction accuracy.
• To introduce a novel algorithm utilizing geostationary satellite data and a genetic algorithm for Rate of Spread (ROS) estimation.
• To propose a Directional ROS adjustment factor for more precise wildfire perimeter simulation.

Main Methods:

• Utilized near-real-time geostationary satellite wildfire detection data.
• Employed a genetic algorithm to estimate a uniform ROS adjustment factor for the FARSITE simulator.
• Developed and applied a Directional ROS adjustment factor based on fuel model location.
• Validated the methodology using the 2020 Creek Fire in California.

Main Results:

• The proposed algorithm significantly improved wildfire spread prediction accuracy.
• The Directional ROS adjustment factor yielded more precise wildfire perimeter predictions compared to uniform factors.
• The data-driven approach demonstrated enhanced reliability in wildfire response planning.

Conclusions:

• Novel data-driven approach enhances wildfire spread prediction accuracy using satellite data.
• Directional ROS adjustment factor improves perimeter modeling in FARSITE simulations.
• Methodology supports targeted wildfire response strategies, especially when high-resolution data is scarce.