Wildfire Severity Mapping Using Landsat 8 OLI and Sentinel-2A Satellite Multispectral Data

To read the file of this research, you can request a copy directly from the author.


The increasing number of wildfires demands the high-productive, fast and cost-efficient techniques to assess the damage as a part of post-fire management. This project explores the benefits of a remote-sensing approach to defining burn severity and area of damage using free available satellite multispectral imagery from Landsat 8 OLI and Sentinel 2A satellites. The tested area is the 6006.88 ha part of the Ust-Donetsk forestry in Rostov region of Southern Russia where wildfire occurred in August 2017. The assessment and mapping were produced based on imagery data from dates before and after the fire for each sensor. Considering the correlation between field-based burn severity indices and spectral indices, differences of Normalized Difference Vegetation Index (dNDVI) and Normalized Burn Ratio (dNBR) were calculated and compared. The dNBR was taken for mapping and area counting as more reliable. The classification was done using proposed thresholds resulted in the overall accuracy is 66.2% with Kappa 0.54. The universal sequence of steps was proposed for rapid and effective post-fire severity and area assessment.

No file available

Request Full-text Paper PDF

To read the file of this research,
you can request a copy directly from the author.

ResearchGate has not been able to resolve any citations for this publication.
Full-text available
A new supervised burned area mapping software named BAMS (Burned Area Mapping Software) is presented in this paper. The tool was built from standard ArcGISTM libraries. It computes several of the spectral indexes most commonly used in burned area detection and implements a two-phase supervised strategy to map areas burned between two Landsat multitemporal images. The only input required from the user is the visual delimitation of a few burned areas, from which burned perimeters are extracted. After the discrimination of burned patches, the user can visually assess the results, and iteratively select additional sampling burned areas to improve the extent of the burned patches. The final result of the BAMS program is a polygon vector layer containing three categories: (a) burned perimeters, (b) unburned areas, and (c) non-observed areas. The latter refer to clouds or sensor observation errors. Outputs of the BAMS code meet the requirements of file formats and structure of standard validation protocols. This paper presents the tool’s structure and technical basis. The program has been tested in six areas located in the United States, for various ecosystems and land covers, and then compared against the National Monitoring Trends in Burn Severity (MTBS) Burned Area Boundaries Dataset
Full-text available
Forest fire is one of the dominant disturbances in boreal forests. It is the primary process responsible for organizing the physical and biological attributes of the boreal biome, shaping landscape diversity and influencing biogeochemical cycles. The Greater Hinggan Mountain of China is rich in forest resources while suffers from a high incidence of forest fires simultaneously. In this study, focusing on the most serious forest fire in the history of P. R. China which occurred in this region, we made use of two Landsat-5 TM (Thematic Mapper) images, and proposed to map the overall burned area and burned forest area by multiple methods. During the mapping, the fire perimeter, as well as rivers, roads and urban areas were first extracted and masked visually, and then four indices of Normalized Difference Vegetation Index, Enhanced Vegetation Index, Vegetation Fractional Cover and Disturbance Index were calculated. For each index, the optimal threshold for separating burned from unburned forest area was determined using their histograms. For comparison, threshold segmentation using single-band reflectance was performed, in addition to a Maximum Likelihood Classifier (MLC) based supervised classification of all features and forest area alone; their accuracies were also evaluated and analysed. Among all the methods compared here, mapping by EVI threshold segmentation proved to be optimal by the comparisons of overall accuracy (99.78%) and the kappa coefficient (0.9946). Finally, the calculated burned area and burned forest area were compared with the values from official statistics. Compared with the classical methods used to report official statistics on burned areas, the remote sensing-based mapping is more objective and efficient, less labour- and time-consuming, and more repeatable.