| Texto completo | |
| Autor(es): |
da Silva, Gabriel Maximo
;
Arai, Egidio
;
Shimabukuro, Yosio Edemir
;
de Souza, Anielli Rosane
;
Hoffmann, Tania Beatriz
;
Dutra, Andeise Cerqueira
;
Martini, Paulo Roberto
;
Duarte, Valdete
;
IEEE
Número total de Autores: 9
|
| Tipo de documento: | Artigo Científico |
| Fonte: | 2022 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM (IGARSS 2022); v. N/A, p. 4-pg., 2022-01-01. |
| Resumo | |
This article presents a land use and land cover (LULC) classification map using Random Forest algorithm in the Sao Paulo State (Brazil), and an assessment of burned areas using two products (MCD64A1 and MapBiomas Fire). The method uses Landsat Operational Land Imager (OLI) time series images from January to December of 2020. We performed the classification class by class considering: water, urban area, forest formation, sugarcane, agriculture, forest plantation and pasture. For each class, we used different spectral bands and image fraction according to the best response for the class. For 2020, the top three areas mapped in Sao Paulo State were pasture (40.49%), sugarcane (24.74%) and forest formation (20.60%). Comparing the two burned area products, MCD64A1 mapped more burned areas as it uses MODIS images combined with 1 km active fire observations with higher temporal resolution than MapBiomas Fire. About 60% of the burned areas mapped in 2020 occurred in the sugarcane class. The results show the importance of land use and land cover classification for better understanding fire-prone classes given the spatial distribution. It turns as an environmental tool for environmental strategies of planning and monitoring burned area assessment over regional scales. (AU) | |
| Processo FAPESP: | 19/19371-5 - Análise espaço-temporal do uso e cobertura da terra no estado de São Paulo utilizando técnicas de sensoriamento remoto |
| Beneficiário: | Yosio Edemir Shimabukuro |
| Modalidade de apoio: | Auxílio à Pesquisa - Regular |