Overview
| DOI | https://doi.org/10.3334/ORNLDAAC/1694 |
|---|---|
| Version | 1 |
| Project | |
| Published | 2020-02-14 |
| Updated | 2020-02-17 |
| Usage | 297 downloads |
Description
This dataset provides maps at 30-m resolution of landscape surface burn severity (surface litter and soil organic layers) from the 2014-2015 fires in the Northwest Territories and Northern Alberta, Canada. The maps were derived from Landsat 8 Operational Land Imager/Thermal Infrared Sensor (OLI/TIRS) imagery and two separate multiple linear regression models trained with field data; one for the Plains and a second for the Shield ecoregion. Field observations were used to estimate area burned in each of five severity classes (unburned, singed, light, moderate, severely burned) in six stratified randomly selected plots of 10 x 10-m in size across a 1-ha site. Using this five class scale a burn severity index (BSI) for each 1-ha site was calculated using multiple weighted and averaged field parameters. Pre- and post-fire phenologically paired Landsat 8 images were used to model the five discrete severity classes using midpoints as breaks.
Science Keywords
- HUMAN DIMENSIONS
- NATURAL HAZARDS
- WILDFIRES
- BURNED AREA
- BIOSPHERE
- VEGETATION
- LITTER CHARACTERISTICS
- LAND SURFACE
- SOILS
- ORGANIC MATTER
- BIOSPHERE
- ECOLOGICAL DYNAMICS
- FIRE ECOLOGY
- FIRE OCCURRENCE
Data Use and Citation
This dataset is openly shared, without restriction, in accordance with the EOSDIS Data Use Policy. See our Data Use and Citation Policy for more information.
Data Files
Sign in to download files.
Companion Files
Sign in to download files.
Dataset Companion Files
Dataset has 2 companion files.
- severity_training.csv
- NWT_Burn_Severity_Maps.pdf