Burned areas dataset for the enclaves of grasslands and savannah of the Mapinguari National Park (Amazonas, Brazil)

<p><span>The present dataset includes annual mapping of fire scars for the enclaves of grasslands and savannah of the Mapinguari National Park (Amazonas, Brazil), at 30-meter spatial resolution, for the period 2000-2023. The enclave areas occupy a total of 241,000 hectares.</span></p> <p><span>The detection of burned areas was carried out using Burned Area Mapping (BAMS) algorithm (Bastarrika et al, 2014), followed by the performance of visual supervision processes and the use of active fire products to optimize the detection date of each fire event. The algorithm is applied to the Surface Reflectance series of Landsat (TM, ETM+, OLI and OLI-2) &ndash; Collection II, accessed using Google Earth Engine (Gorelick et al, 2017). Data from active fire products MCD14DL V006, VIIRS-NPP, VIIRS-NOA20 and GOES16, as well as burned area data from product MCD64A1 v006, were used to optimize the detection date of each fire scar.</span></p> <p><span>_________________________________________________________________________________________________________</span></p> <p><span>July 08, 2023 &ndash; The version 1.0 includes the period 2000-2023, with a total of 356,688.5 hectares of fire affected areas, distributed across 453 fire scars.</span></p> <p><span>_________________________________________________________________________________________________________</span></p> <p><span>The files available include: </span></p> <p><span>i) &ldquo;fire_scars_dataset.rar&rdquo;: annual burned areas vector files, in shapefile format (*.shp), projected at WGS84 UTM 20S. Each observation is an individual fire scar, with his attribute table indicating:</span></p> <p><span>- ID: Identification number for each;</span></p> <p><span>- area_ha - area of each fire scar, calculated in hectares.</span></p> <p><span>- date - detection date of the fire scar</span></p> <p><span>- date_preci - precision flag of the fire detection date:</span></p> <p><span>0 - fire date detected using only Landsat data</span></p> <p><span>1 - fire date optimized based on active fires dataset (MCD14DL V006; VIIRS-NPP; VIIRS-NOA20; or GOES16)</span></p> <p><span>ii) &ldquo;enclaves_Mapinguari_National_Park.rar&rdquo;: vector file of study area location, in shapefile format (*.shp), projected at WGS84 UTM 20S. Includes 8 enclaves of grasslands and savannah situated inside Mapinguari National Park.</span></p> <p><span>iii) &ldquo;dataset_description.pdf&rdquo;: description of the dataset.</span></p> <p><span>_________________________________________________________________________________________________________</span></p> <p><span>We thank the Conselho Nacional de Pesquisa e Desenvolvimento (CNPq) and the Instituto Chico Mendes de Conserva&ccedil;&atilde;o da Biodiversidade (ICMBIO) (process number 126772/2022-3) for the grant conceded to the second and third authors.</span></p> <p><span>_________________________________________________________________________________________________________</span></p> <p>&nbsp;</p> <p><span>References</span></p> <p>&nbsp;</p> <p><span>Bastarrika, Aitor, Maite Alvarado, Karmele Artano, Maria Pilar Martinez, Amaia Mesanza, Leyre Torre, Rub&eacute;n Ramo, and Emilio Chuvieco. 2014. &ldquo;BAMS: A Tool for Supervised Burned Area Mapping Using Landsat Data.&rdquo; Remote Sensing 6: 12360&ndash;80. https://doi.org/10.3390/rs61212360.</span></p> <p><span>Gorelick, Noel, Matt Hancher, Mike Dixon, Simon Ilyushchenko, David Thau, and Rebecca Moore. 2017. &ldquo;Google Earth Engine: Planetary-Scale Geospatial Analysis for Everyone.&rdquo; Remote Sensing of Environment 202: 18&ndash;27. https://doi.org/10.1016/j.rse.2017.06.031.</span></p>