dNBR (burn severity)
The difference between a pre-fire and a post-fire normalized burn ratio, with the published severity ranges it is usually read against.
dNBR is 0.48, in the moderate-high severity range.
- Severity class
Provenance
- Computed by
- raster.index.dnbr 1.1.0, core 0.1.0
- Model
- dNBR = NBR(pre-fire) - NBR(post-fire); classes from the Key and Benson (2006) table LA-2 ranges, read on dNBR x 1,000 to the nearest whole number
- Accuracy
- Exact arithmetic. The class ranges are a published starting point, not a determination of severity.
- Notes
- None
- Cites
- Key, C. H., and Benson, N. C., USDA Forest Service, Rocky Mountain Research Station, Landscape Assessment (LA): sampling and analysis methods, in FIREMON: Fire Effects Monitoring and Inventory System
Something look off?
How we got thisFormula, worked example, sources, and proof
Model: dNBR = NBR(pre-fire) - NBR(post-fire); classes from the Key and Benson (2006) table LA-2 ranges, read on dNBR x 1,000 to the nearest whole number
Accuracy: Exact arithmetic. The class ranges are a published starting point, not a determination of severity.
When to use this: Use this once you have an NBR from before a fire and one from after it: the drop between them is what burn-severity mapping is based on, and it separates ground that burned from ground that was already bare. The class it names is the range the value falls in.
Limitations: The ranges come from western US conifer forests and do not transfer unchanged to other ecosystems; severity mapping normally calibrates them against field plots, and often uses the relativized form (RdNBR) where pre-fire cover varies. Both images need the same sensor, season, and processing, or the difference measures the images rather than the fire.
Worked example: A moderate-high severity burn. Source: Key and Benson 2006: a drop of 0.48 falls in the moderate-high range (0.44 to 0.66). It is golden test vector v001, and every build checks the tool still gives its answer within its tolerance.
You enter
- Post-fire NBR
- 0.13
- Pre-fire NBR
- 0.61
You get
- dNBR
- 0.48
- Severity class
- moderate-high severity
Review: Not yet independently reviewed by a GIS professional.
Last verified: 2026-09-22, when a maintainer last confirmed this tool's sources at the issuer. See the sources ledger.
Status: version 1.1.0, core 0.1.0. See this tool in the verification report.
Changes
- 2026-09-24, fixed: dNBR (version 1.1.0) now reads the Key and Benson severity table the way it is printed, in whole units of dNBR times 1,000, so a drop of exactly 0.10 (such as 0.37 to 0.27) is low severity, not unburned by a rounding hair; the same holds at every class edge. Values beyond about -550 or +1,350, which the table's authors attribute to clouds or misregistration rather than fire, are now flagged. The tool is now stable. Changelog
Checked against: 24 golden test vectors (download the test vectors, each with its source and tolerance). See how results are checked and every source.
Constants this tool assumes
Values you do not enter, and where each comes from.
- dNBR classification scale
1000— Landscape Assessment (LA): sampling and analysis methods, in FIREMON: Fire Effects Monitoring and Inventory System- High-to-low post-fire regrowth threshold
-250 scaled dNBR— Landscape Assessment (LA): sampling and analysis methods, in FIREMON: Fire Effects Monitoring and Inventory System- Low-regrowth-to-unburned threshold
-100 scaled dNBR— Landscape Assessment (LA): sampling and analysis methods, in FIREMON: Fire Effects Monitoring and Inventory System- Unburned-to-low-severity threshold
100 scaled dNBR— Landscape Assessment (LA): sampling and analysis methods, in FIREMON: Fire Effects Monitoring and Inventory System- Low-to-moderate-low severity threshold
270 scaled dNBR— Landscape Assessment (LA): sampling and analysis methods, in FIREMON: Fire Effects Monitoring and Inventory System- Moderate-low-to-moderate-high severity threshold
440 scaled dNBR— Landscape Assessment (LA): sampling and analysis methods, in FIREMON: Fire Effects Monitoring and Inventory System- Moderate-high-to-high severity threshold
660 scaled dNBR— Landscape Assessment (LA): sampling and analysis methods, in FIREMON: Fire Effects Monitoring and Inventory System- Lower suspect-value bound
-550 scaled dNBR— Landscape Assessment (LA): sampling and analysis methods, in FIREMON: Fire Effects Monitoring and Inventory System- Upper suspect-value bound
1350 scaled dNBR— Landscape Assessment (LA): sampling and analysis methods, in FIREMON: Fire Effects Monitoring and Inventory System
Sources
- Landscape Assessment (LA): sampling and analysis methods, in FIREMON: Fire Effects Monitoring and Inventory System, Key, C. H., and Benson, N. C., USDA Forest Service, Rocky Mountain Research Station, General Technical Report RMRS-GTR-164-CD, pages LA-1 to LA-55. NBR and dNBR, and table LA-2: ordinal severity levels and example dNBR ranges (scaled by 1,000), offered as a starting point rather than a rule; values beyond about -550 and +1,350 are likely anomalies.
Terms
- NBR: normalized burn ratio
- The difference between near-infrared and the longer shortwave-infrared band over their sum. It falls sharply over recently burned ground. Source: Landscape Assessment (LA): sampling and analysis methods, in FIREMON: Fire Effects Monitoring and Inventory System