Remote sensing scripts for mapping prescribed fire boundaries at Texas nature preserves using satellite imagery. Developed during a science internship with The Nature Conservancy β Texas Chapter.
The scripts calculate the Normalized Burn Ratio (NBR) and differenced NBR (dNBR) from satellite imagery captured before and after a controlled burn. This allows ecologists to:
- Identify which areas within a preserve were actually burned
- Assess burn severity using USGS classification thresholds
- Export fire boundary polygons as shapefiles for land management records
- Monitor post-fire vegetation recovery using NDVI (Sentinel-2 script)
| File | Satellite | Resolution | Best for |
|---|---|---|---|
landsat8_nbr.js |
Landsat 8 (NASA/USGS) | 30m | Historical burns (2013+), large preserves |
sentinel2_nbr.js |
Sentinel-2 (ESA) | 10m | Recent burns (2017+), finer detail |
- Open the Google Earth Engine Code Editor
- Paste the contents of either
.jsfile into a new script - Set the configuration variables at the top:
PRESERVE_NAMEβ must match theInterest_Nfield in your asset- Date ranges for before and after the burn
- Click Run
- Uncomment the
Exportblock to save fire polygons to Google Drive
- NBR:
(NIR - SWIR) / (NIR + SWIR)β highlights burned areas - dNBR: difference between pre- and post-fire NBR for burn severity classification
- NDVI: tracks vegetation health for post-fire recovery monitoring
- Cloud masking: QA60 bitmask for Sentinel-2, metadata filtering for Landsat 8
- Clymer Meadow Preserve (October 2013) β Landsat 8
- Roy E. Larsen Sandyland Sanctuary (March 2023) β Sentinel-2