FireDataForge turns one MTBS fire ID into a stack of harmonized, analysis-ready geospatial layers (perimeters, FRP, terrain, fuels, weather, and more) from many sources.
-
Updated
Oct 19, 2026 - Python
FireDataForge turns one MTBS fire ID into a stack of harmonized, analysis-ready geospatial layers (perimeters, FRP, terrain, fuels, weather, and more) from many sources.
Remote-sensing experiments on wildfire burn-severity mapping, cross-event transfer, and pre-fire prediction using Landsat, MTBS, and CanLaBS.
Calibrated dNBR thresholds match a U-Net for burn severity on 31 unseen California and Oregon wildfires, scored against MTBS analyst maps
This repository contains the logic associated with the research manuscript: "Roadless status under The Roadless Area Conservation Rule is not associated with increased wildfire in the US National Forest System".
To associate your repository with the mtbs topic, visit your repo's landing page and select "manage topics."