Fire extent and severity maps

Our scientists have developed a semi-automated approach to mapping fire extent and severity, in collaboration with the NSW Rural Fire Service.

Fire extent and severity mapping (FESM) uses satellite imagery and machine learning to deliver timely maps of fire severity. These maps help us:

  • classify and model fires
  • understand relationships between fuels and fire behaviour
  • interpret how a fire has changed the landscape
  • provide information about the damage a fire has caused to the tree canopy and understory
  • gather evidence about how far each fire has spread
  • support on-ground actions, fire management and conservation planning.

Before FESM was developed, fire severity information at a landscape scale was not routinely captured in New South Wales.

When maps are produced

Maps for individual fires can be produced once a fire event has stabilised. We also produce summary maps annually after each fire season to investigate changes in vegetation trends over time. This enables our scientists to better understand how future fire events may unfold and the potential impacts of these events on the environment.

The FESM state-wide map and datasets for the 2017–18, 2018–19 and 2019–20 seasons are available on the Sharing and Enabling Environmental Data (SEED) portal or by searching ‘FESM SEED’. FESM is part of collaborative research being undertaken as part of the NSW Bushfire Risk Management Research Hub  to improve fire management strategies and reduce the risk bushfires pose to people, property and the environment.

Fire extent and severity mapping image

Mapping method

Since 2018, we’ve been developing a mapping system to deliver near-real time maps of fire severity to help classify and model fires.

A prototype operational system was deployed in December 2019 to provide rapid-response severity mapping to support the Department’s operations during the emerging black summer bushfire crisis. A refined, fully operational system was launched in July 2020 that:

  • delivers mapping in near-real time
  • links with the Rural Fire Service IT infrastructure
  • automates the application of the modelling to new fires.

Our scientists are leading continued refinement and development of FESM, which involves the NSW Rural Fire Service and Natural Resources Commission. The program will benefit from capturing and incorporating additional training data from future fires.

The Supporting fire management with the fire extent and severity maps fact sheet contains more information about the fire mapping methodology.

Fire severity classes

The FESM system designates areas into standardised fire severity classes to facilitate comparison of different fires across the landscape. This helps land managers and researchers understand how fires have affected the landscape and informs:

  • on-ground conservation
  • fire management actions
  • recovery efforts.

The algorithm uses ecologically meaningful definitions of ‘severity’ to define levels of canopy scorch and consumption. This enables us to:

  • gain a more nuanced understanding of the impact of fires on the environment
  • provide information about the potential impact on species and habitats.

Pixel Value

Severity class

Description

Photo interpretation cues (false colour infra-red)

Percentage foliage fire affected

0

Unburnt

Unburnt surface with unburnt canopy

Dark red (live understorey) between the dark red tree crowns

0% canopy and understorey burnt

2

Low

Burnt surface with unburnt canopy

Dark grey (burnt understorey) between the dark red tree crowns

>10% burnt understorey
>90% green canopy

3

Moderate

Partial canopy scorch

A mixture of green, orange and brown colours in tree canopies

20-90% canopy scorch

4

High

Complete canopy scorch (+/– partial canopy consumption)

No green or orange, but an even brown colour in tree canopies

>90% canopy scorched
<50% canopy biomass consumed

5

Extreme

Complete canopy consumption

Mostly black and dark grey, largely no canopy cover

>50% canopy biomass consumed

Table note: Fire severity classification ruleset based on high-resolution aerial photo interpretation, adapted from McCarthy et al. (2017) and Collins et al. (2018).

An annual report summarising the previous fire season is published each year. Each report provides an outline of analyses for the previous fire season and makes comparisons to the preceding fire seasons. The annual report is accompanied by a data spreadsheet, and can help governments, fire managers, conservation and landscape ecology researchers understand and respond to environmental effects of fire on the landscape.

The Fire extent and severity mapping annual report for the 2019–20, 2018–19 and 2017–18 fire years was released in 2021.

The FESM state-wide map and dataset for the 2019–20 season is available on the SEED portal or by searching for ‘FESM SEED’. Maps are delivered as zip files including rasters (10m pixel size) in GeoTiff (.tif) and ERDAS Imagine (.img) format. The rasters are viewable in standard GIS software (ArcGIS, QGIS).

This work has been led by scientists from the Department and reflects extensive collaboration with the NSW Rural Fire Service, the NSW Bushfire Risk Management Research Hub and the NSW Natural Resources Commission over several years of research and testing.

Bowman DMJS, Williamson GJ, Gibson RK, et al. 2021 The severity and extent of the Australia 2019–20 Eucalyptus forest fires are not the legacy of forest management. Nature Ecology and Evolution, https://doi.org/10.1038/s41559-021-01464-6

Collins L, Griffioen P, Newell G, Mellor A 2018 The utility of random forests in Google Earth Engine to improve wildfire severity mapping. Remote Sensing of Environment, 216, 374–384 https://doi.org/10.1016/j.rse.2018.07.005

Gibson RK, Danaher T, Hehir W, Collins L A remote sensing approach to mapping fire severity in south-eastern Australia using sentinel 2 and random forest, Remote Sensing of Environment, 240, 111702 https://doi.org/10.1016/j.rse.2020.111702

McCarthy G, Moon K, Smith L 2017 Mapping fire severity and fire extent in forest in Victoria for ecological and fuel outcomes, Ecological Management and Restoration, 18(1), p54–65 https://doi.org/10.1111/emr.12242