Burned Area Reflectance Classification (BARC)

Written by: Grant Hamilton

Other Names

BARC

Description

Burned Area Reflectance Classification (BARC) is a method using remotely sensed data to assess the severity of fire damage to vegetated areas developed by the US Forest Service and US Geological Survey. It is premised on the contrast in near infrared (NIR) and mid infrared (MIR) reflectance between pre-fire vegetated surfaces and post-fire denuded surfaces: “Imagery collected over a forest in a pre-fire condition will have very high near infrared band values and very low mid infrared band values. Imagery collected over a forest after a fire will have very low near infrared band values and very high mid infrared band values” (USFS Remote Sensing Applications Center, n.d.).

barc_spectral_response.png

Source: USFS

• (NDVI)

• (EVI)

• (NBR)

Calculating Burn Severity Using BARC

Burn severity is calculated using landsat_etm_7 and/or landsat_8 imagery. The normalized_burn_ratio (NBR) is used to assess a fire’s severity:

((Band 4 – Band 7) / (Band 4 + Band 7)) x 1000 = NBR

where Band 4 includes wavelengths from 0.76-0.90 µm (NIR) and Band 7 includes wavelengths between 2.09-2.35 µm (SWIR). If pre-fire imagery is available the differenced Normalized Burn Ratio (dNBR) is used:

Pre-fireNBR – Post-fireNBR = dNBR

Higher dNBR indicate more severe damage. Areas with negative dNBR values may indicate increased vegetation productivity following a fire. “Non-Processing Areas” include parts of the image obscured by clouds or smoke or water bodies.

Accessing Datasets

Recent and Near Real-time Data – Burned Area Emergency Response (BAER)

Datasets for recent (generally in the current year) and ongoing fires cannot be downloaded on demand. The user must submit a request which is processed by the Burned Area Emergency Response (BAER) team. Delivery of datasets may take some time depending on the length of the request queue. BAER data is intended for federal and state incident managers. BAER’s imagery request queue is available at http://svinetfc6.fs.fed.us/birch/.

Historic Data (1984 – 2011) – Monitoring Trends in Burn Severity (MTBS)

Only fires of greater than 1,000 acres in the western conterminous United States (CONUS) and greater than 500 acres in the eastern CONUS are recorded. Data also covers Alaska, Hawaii, and Puerto Rico. Fires on both private and public property are recorded. Historic data can be downloaded on demand from the Monitoring Trends in Burn Severity (MTBS) website at http://mtbs.gov/compositfire/mosaic/bin-release/download.html.

Explanation of BARC Datasets

As they are derived from Landsat 7 and/or Landsat 8 imagery, BARC datasets have a resolution of 30m.

BAER 8-bit datasets contain unclassified values of for use by BAER teams in the field. These datasets are referred to as BARC-256 by federal agencies.

MTBS datasets are available as mosaics of the CONUS, Alaska, Hawaii, or Puerto Rico. They are in ERDAS Imagine (.img) format. MTBS dataset values are classified into four levels of fire damage severity using a variety of methods, e.g. Jenks Natural Breaks; wherever possible the Landsat TM derived datasets are calibrated with field data (Hudak, et al, 2004). These datasets are referred to as BARC-4 by federal agencies. MTBS suggests the following symbology:

Value 1 = Very Low or Unburned (Dark Green)
Value 2 = Low (Cyan)
Value 3 = Moderate (Yellow)
Value 4 = High (Red)

Values of 0 indicate “Background/No Data”, values of 5 indicate “Increased Greenness/Increased Vegetation Response”, and values of 6 indicate “Non-Processing Area Mask”.

barc_ex.png

Damage severity of the 2011 Abrams Fire on the eastern slope of the Organ Mountains in New Mexico. This map was created in ArcMap 10 using the MTBS mosaic and suggested symbology.

Technical References

Hudak, A.T., Robichaud, P.R., Evans, J.B., Clark, J., Lannom, K., Morgan, P., Stone, C. 2004. Field validation of Burned Area Reflectance Classification (BARC) products for post fire assessment. In Greer, J.D, ed. Remote sensing for field users; proceedings of the tenth Forest Service remote sensing applications conference, April 5–9; Salt Lake City, UT. http://www.treesearch.fs.fed.us/pubs/23530.

  • Discusses calibrating Landsat TM derived dNBR values to field-collected data.

USFS Remote Sensing Applications Center. n.d. BARC – Frequently asked questions. http://www.fs.fed.us/eng/rsac/baer/barc.html.

  • Includes a tutorial on calibrating BAER-256 datasets to field-collected data.

Datasets

Comments are closed.