Identification of Burned Areas and Severity Using SAR Sentinel-1

In this letter, we performed investigations on the potentiality of the Sentinel-1, C-band synthetic-aperture radar (SAR), for the characterization and mapping of burned areas and fire severity. To this aim, we focused on fire occurred on July 13, 2017, in Metaponto (South of Italy). Both VH and VV p...

Full description

Saved in:
Bibliographic Details
Published in:IEEE geoscience and remote sensing letters 2019-06, Vol.16 (6), p.917-921
Main Authors: Lasaponara, Rosa, Tucci, Biagio
Format: Article
Language:English
Subjects:
Backscatter
Burn severity
burned areas
C band
Classification
Data acquisition
Data analysis
Data processing
fire
Fires
Laser radar
Mapping
Noise reduction
Polarization
Radar
remote sensing
SAR (radar)
Sensors
Sentinel-1
Speckle
Synthetic aperture radar
synthetic-aperture radar (SAR)
Timber industry
Vegetation mapping
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this letter, we performed investigations on the potentiality of the Sentinel-1, C-band synthetic-aperture radar (SAR), for the characterization and mapping of burned areas and fire severity. To this aim, we focused on fire occurred on July 13, 2017, in Metaponto (South of Italy). Both VH and VV polarizations were considered. Radar Burn Difference (RBD) and radar burn ratio (RBR) were computed between Sentinel-1 data acquired before and after the fire using both single- and time-averaged scenes (to reduce speckle noise effects). The most marked differences between burned and unburned areas were observed in the VH polarization of both RBD and RBR. The novelty of our approach is based on the use of three steps data processing devised to identify different levels of fire severity without using fixed thresholds. The burned areas are first: 1) highlighted using the ratio between multitemporal data set acquired before and after the fire occurrence; 2) further enhanced by Getis-Ord spatial statistic; and 3) finally, categorized using ISODATA unsupervised classification. The approach herein proposed pointed out that: 1) the time-averaged ratio of VH polarization of Sentinel-1 well perform in mapping burned area and 2) the use of Getis-Ord spatial statistic coupled with ISODATA unsupervised classification suitably captures the diverse levels of burned severity as confirmed by in situ assessment.
ISSN:1545-598X
1558-0571
DOI:10.1109/LGRS.2018.2888641