Skip to content

BAS service specifications

icon

This service takes as input a pair of calibrated optical datasets from the same mission and estimates burnt areas using the Normalized Burn Ratio (NBR) index. The datasets must have nir and swir22 bands (e.g. Sentinel-2).

đź“• The tutorial of the BAS service is available in this section.

Service Description

The Burned Area Severity Analysis (BAS) service takes as input a pair of calibrated optical datasets containing nir and swir22 spectral assets and estimates the severity of burnt areas. It provides as output a raster with a color gradient based on RBR value representing burnt area severity.

The burn severity describes how the fire intensity affects the functioning of the ecosystem in the area that has been burnt. The observed effects often vary within the area and between different ecosystems (Keeley, 2009)1. The burn severity can also be described as the degree to which an area has been altered or disrupted by the fire2.

The Burned Area Severity Analysis methodology integrated into this service is based on the Normalized Burn Ratio (NBR) index (Key and Benson, 2006)3 and its derivatives: the delta Normalized Burn Ratio (dNBR) from Miller and Thode, 20144, and the Relativized Burn Ratio (RBR) from Parks et al., 20145. Such indexes are described below.

Cover

Figure 1 - Overview BAS highlighting burnt areas over large wildfires in Chile. This product is showing a classification of the RBR index in yellow to red tones, derived from a pair of Landsat-9 OLI Calibrated Datasets acquired on 10 Jan 2023 (pre-event) and 27 Feb 2023 (post-event) over Quillón and Chillán areas.

Normalized Burn Ratio and delta Normalized Burn Ratio

The Normalized Burn Ratio (NBR) index developed by Key and Benson, 20063 is defined as the normalized difference between the nir and swir22:

\[ NBR = { { nir - swir_{22} } \over { nir + swir_{22} } } \]

The delta Normalized Burn Ratio (dNBR) from Miller and Thode (2014) is defined as:

\[ dNBR = NBR_{prefire} - NBR_{postfire} \]

Relativized Burn Ratio

The Relativized Burn Ratio (RBR) from Parks et al.,20145 is defined as:

\[ RBR = { dNBR \over { NBR_{prefire} + 1.001 } } \]

Also, the processor generates from the pair of input optical images a couple of Shortwave infrared (swir22, nir, red) RGB composites, for both pre- and post-event images.

Workflow

The first step of the BAS processing service creates a collocated stack between the pre and post-event acquisitions of the assets red, nir and swir22.

This collocated stack covers the common area between the pre and the post-event acquisitions.

In case of inputs with different spatial resolution, the higher resolution is used.

The second step applies the band expressions listed below:

\[ NBR_{pre} = { { nir_{pre} - swir_{ {22 pre} } } \over { nir_{pre} + swir_{22 pre} } } \]
\[ NBR_{post} = { { nir_{post} - swir_{ {22 post} } } \over { nir_{post} + swir_{22 post} } } \]
\[ dNBR = NBR_{pre} - NBR_{post} \]
\[ RBR = { dNBR \over { NBR_{prefire} + 1.001 } } \]

The third step generates an RGB composite with:

\[ Red\ channel = swir_{22 post} \]
\[ Green\ channel = nir_{post} \]
\[ Blue\ channel = red_{post} \]

In the last step the overview-bas asset is built by considering only pixels satisfying the following condition: RBR>=0.3. After the normalization to the 8-bit range, the color representation of the RBR index is made using a pre-defined graded yellow-red color bar (YlOrRd Matplotlib color palette). BAS overview legend is shown below.

bas-overview-legend

Warning

In this service water and cloud masking are not performed.

Note

Output products, such as the RBR index, can be masked in further post-processing outside the platform (e.g. in QGIS using external available water masks).

Input

The BAS processor requires in input a pair of Calibrated Datasets, derived in a systematic fashion with the Optical Products Calibration service from geocoded L1 or L2 products of supported sensors. Both Optical Calibrated Datasets shall include the following single band assets:

  • red,

  • nir,

  • swir22.

Examples of such missions are:

  • Sentinel-2,

  • Landsat-8, Landsat-9,

  • Worldview-3 (only if swir bands are given).

The pair is typically made by pre- and post-event datasets.

Parameters

The BAS service requires a specified number of mandatory and optional parameters. All service parameters are listed in the below Table 1.

Parameter Description Required Default value
Post-event optical calibrated dataset Post-event Optical calibrated dataset with red, nir and swir22 assets YES
Pre-event optical calibrated dataset Pre-event Optical calibrated dataset with red, nir and swir22 assets YES
Area of Interest Area of interest expressed in WKT NO

Table 1 - Service parameters of the BAS processor.

Output

The BAS processor generates the following products:

  1. N=6 co-located single band TOA/BOA reflectance assets (red, nir, swir22) from the two input Calibrated Datasets.

  2. Single band NBR index (pre-event),

  3. Single band NBR index (post-event),

  4. Single band dNBR index,

  5. Single band RBR index,

  6. BAS overview as 4 bands RGBA,

  7. SIR overviews for pre- and post-event as 4-band RGBA in 8-bit, plus the three 8-bit single-bands for Red, Green, and Blue channels.

Warning

The BAS service does not perform a masking of permanent waters into products. Thus false alarms may occur over water surfaces.

BAS Product Specifications can be found in the below tables.

Long Name Normalized Burn Ratio pre-event
Short Name nbr_pre
Description Normalized difference between nir and swir22 assets of the pre-event acquisition
Processing level L1 / L2
Data Type Float32
Band Single
Format COG
Projection EPSG:4326 - WGS84
Units N/A
Valid Range [-1 , 1]
Fill Value N/A

Long Name Normalized Burn Ratio post-event
Short Name nbr_post
Description Normalized difference between nir and swir22 assets of the post-event acquisition
Processing level L1 / L2
Data Type Float32
Band Single
Format COG
Projection EPSG:4326 - WGS84
Units N/A
Valid Range [-1 , 1]
Fill Value N/A

Long Name Delta Normalized Burn Ratio
Short Name dnbr
Description Difference between the pre and post Normalized Burn Ratio
Processing level L1 / L2
Data Type Float32
Band Single
Format COG
Projection EPSG:4326 - WGS84
Units N/A
Valid Range [-2 , 2]
Fill Value N/A

Long Name Relativized Burn Ratio
Short Name rbr
Description Relativized Burn Ratio using the pre and post Normalized Burn Ratio
Processing level L1 / L2
Data Type Float32
Band Single
Format COG
Projection EPSG:4326 - WGS84
Units N/A
Valid Range [-2 , 2]
Fill Value N/A

Long Name Burned Area Severity overview
Short Name overview-bas
Description RGBA composite from a reclassification of RBR index using predefined color operations (from yellow to red shades), including transparency.
Processing level L1 / L2
Data Type Unsigned 8-bit Integer (UInt 8)
Band 4
Format COG
Projection EPSG:4326 - WGS84
Units N/A
Valid Range [1 - 255]
Fill Value 0

Long Name Shortwave Infrared RGB composite
Short Name overview-sir-pre, overview-sir-post
Description RGBA composite (Shortwave Infrared), including transparency, for both pre- and post-event images.
Processing level L1 / L2
Data Type Unsigned 8-bit Integer (UInt 8)
Band 4
Format COG
Projection EPSG:4326 - WGS84
Units N/A
Valid Range [1 - 255]
Fill Value 0

References

  1. Keeley, J. E. (2009), "Fire intensity, fire severity and burn severity: A brief review and suggested usage", International Journal of Wildland Fire, 18(1), 116–126. Available at: https://pubs.er.usgs.gov. ↩

  2. UN Spider Knowledge Portal - Normalized Burn Ratio. ↩

  3. Key, C. H. and Benson, N. C. (2006), “Landscape Assessment (LA): Sampling and Analysis Methods”, USDA Forest Service Gen Tech. Rep RMRS-GTR-164-CD. FIREMON Fire effects monitoring and inventory System. Available at: https://www.fs.usda.gov. ↩↩

  4. Miller J. and Thode A. (2007), “Quantifying burn severity in a heterogeneous landscape with a relative version of the delta Normalized Burn Ratio (RdNBR)”, Remote Sensing of Environment, 109, 66-80. DOI: 10.1016/j.rse.2006.12.006. ↩

  5. Parks S. A., Dillon G. K., Miller C. A. (2014) "A New Metric for Quantifying Burn Severity: The Relativized Burn Ratio" Remote Sens. 6, no. 3: 1827-1844. DOI: 10.3390/rs6031827. ↩↩