Mapping Wildfire Ignition Probability Using Sentinel 2 and LiDAR (Jerte Valley, Cáceres, Spain)

Area of Science:

• Environmental Science
• Remote Sensing
• Forestry

Background:

• Wildfires pose a significant environmental threat, exacerbated by climate and socioeconomic factors.
• Effective wildfire management relies on accurate, up-to-date spatial data.
• Existing mapping methods may lack the required detail or timeliness for optimal response.

Purpose of the Study:

• To develop and validate a protocol for precise, high-resolution thematic map updating for wildfire management.
• To create a reliable cartographic base for calculating wildfire ignition probability.
• To provide a tool for informed decision-making in wildfire prevention and suppression.

Main Methods:

• Integration of Sentinel 2A satellite imagery (5-day revisit) with Light Detection and Ranging (LiDAR) data (0.5 points/m² density).
• Development of a decision tree model to extrapolate vegetation maps to fuel model maps.
• Validation of vegetation mapping accuracy using kappa coefficient.

Main Results:

• Achieved 88% accuracy (kappa = 0.86) in vegetation mapping.
• Successfully extrapolated vegetation data to create fuel model maps.
• Established a fast, reliable protocol for generating essential wildfire management cartography.

Conclusions:

• The developed protocol enables precise updating of high-resolution thematic maps for wildfire management.
• The generated fuel model maps serve as a critical cartographic base for ignition probability assessment.
• This approach supports strategic planning for silvicultural treatments, resource allocation, and infrastructure protection.