ABSTRACT
Uttarakhand State is located in northern India and is mostly hilly, occupying an area of 53,483 km2 (20,650 mi2), of which 86% is mountainous and 65% is covered by forest. Uttarakhand witnessed significant wildfires during May 2019, which burned nearly 900 ha of the forest. One of the worst affected districts was Nainital, which is mostly comprised of oak and pine trees. This study assessed the potential of very high-resolution PlanetScope data for burnt area delineation in Nainital. To capture the fire events, we also used the active fire data from VIIRS 375 m and MODIS 1 km collection during the end of April till May 23, 2019, to identify the peak burning dates. The burnt areas derived from the PlanetScope were compared with the Landsat (30 m) data for relative comparison. Specific to the active fires, the VIIRS 375 m data captured more fires than the MODIS. The peak fires were detected on May 2, 4, 5, and 6. The Landsat-8 was available on April 21 and May 7, which were 16 days apart. In contrast, nine different images were available from PlanetScope from April 27 till May 19, and these data could capture burnt area progression more robustly than the Landsat. In the burnt areas, a clear decrease in near-infrared signal could be detected from the PlanetScope data. The results identify important differences between the Landsat (30 m) and PlanetScope (3 m) data useful for burnt area mapping.
