Assessing Forest Fire Severity: Tools from Satellites to Soil

Determining fire severity—the degree of ecosystem alteration by fire—requires a multi-tool approach across scales:

1. Fire Behavior Prediction Systems

Models (e.g., FARSITE, FlamMap, BEHAVE) integrate weather, fuels, topography, and ignition to predict rate of spread, flame length, fireline intensity, crown fire potential. Used pre- and during-fire for tactical decisions.

2. Satellite Remote Sensing

• Active fire detection: MODIS, VIIRS (375 m–1 km) for real-time perimeter.
• Burn severity mapping: dNBR (differenced Normalized Burn Ratio) from Landsat/Sentinel-2 (10–30 m) compares pre/post NIR/SWIR reflectance. Classes: unburned, low, moderate, high severity.
• BAI (Burn Area Index), MIRBI for specific ecosystems.

3. Aerial Assessments

• Manned aircraft: IR line scanners, high-res RGB, LiDAR for 3D structure loss.
• Drones (UAVs): cm-resolution orthomosaics, thermal, multispectral—ideal for targeted areas.

4. Ground-Based Composite Burn Index (CBI)

Field plots score substrate (soil char, duff consumption), herbs, shrubs, trees (crown scorch, mortality, bole char). Calibrates satellite indices.

5. Soil Analysis

Post-fire sampling for hydrophobicity, nutrient loss (N, P), microbial biomass, pH, aggregate stability. Indicates recovery trajectory and erosion risk.

6. Post-Fire Vegetation Monitoring

Repeated plots, LiDAR, hyperspectral track regeneration, invasive colonization, carbon recovery over years.

Integration of these tools—validated by field data—produces severity maps guiding rehabilitation (seeding, mulching), salvage logging decisions, and long-term monitoring.