| Literature DB >> 31980935 |
Elias Rodrigues da Cunha1,2, Celso Augusto Guimarães Santos3, Richarde Marques da Silva1, Vitor Matheus Bacani4, Paulo Eduardo Teodoro5, Elói Panachuki6, Naelmo de Souza Oliveira6.
Abstract
In the last 30 years, the growth of the agriculture and livestock industries in the Cerrado biome has caused severe changes in land use and land cover (LULC), and areas previously occupied by native vegetation are changing to agricultural monocultures (e.g., soybean or corn) and/or pastures. Thus, the objective of this study was to analyze the LULC changes for the years 1986, 1999, 2007, and 2016 based on Landsat time series and object-based image analysis (OBIA) for the Prata River Basin. Twelve LULC classes were mapped: riparian forest, cerrado, swampy grasslands, wetlands, semideciduous forest, pasture, agriculture, fallow agricultural land, barren land, eucalyptus, water bodies, and burnt area. The classifications presented results with an overall accuracy of more than 93% and a kappa coefficient of 0.92. In 2007, the pasture class had the highest increase in area (48.5%), with a total area of 118.32 km2 of Cerrado biome vegetation converted to pasture, and the classes banhado, riparian forest, swampy grasslands, and cerrado had the greatest reductions in area (41.58%, 29.67%, 25.44%, and 21.63%, respectively). More precisely, the wetlands class underwent the greatest decrease under the advancement of pasture in the studied period (- 36.2%). These changes are due to factors favorable to agropastoral practices, such as a flat relief and soil with good agricultural suitability. Graphical abstract.Entities:
Keywords: Cerrado biome; Deforestation; Land use/land cover change; Object-based image analysis; Remote sensing
Mesh:
Year: 2020 PMID: 31980935 DOI: 10.1007/s10661-020-8093-9
Source DB: PubMed Journal: Environ Monit Assess ISSN: 0167-6369 Impact factor: 2.513