Mapeamento da cana-de-açúcar em Minas Gerais

• Ano de defesa: 2019
• Autor(a) principal: Dias, Bruna Aparecida Silva
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Uberlândia; Brasil; Programa de Pós-graduação em Geografia (Pontal)
• Palavras-chave em Português: Cana-de-açúcar; Sistema de manejo; Imagens multitemporais; Conversção de uso e cobertura da terra; Geoprocessamento; Sugarcane; Harvesting mode; Multitemporal images; Use conversion and land cover; Geoprocessing; Geografia; Cana-de-açúcar- Minas Gerais; Cana-de-açúcar- manejo; CNPQ::CIENCIAS HUMANAS::GEOGRAFIA
• Link de acesso: https://repositorio.ufu.br/handle/123456789/24449; http://dx.doi.org/10.14393/ufu.di.2019.615

The expansion of crops such as sugar cane has led to numerous changes in how to use the land and its coverage throughout the country. With the growing demand for biofuels, in a few years, Brazilian production has started to lead the world ranking, reaching more than 10 million hectares planted in 2016. Thus, the present research had as its main goal to map the area occupied by sugarcane, the sugarcane in the state of Minas Gerais based on geoprocessing techniques for the year 2017. The specific goals were: a) to test classification techniques for the mapping of sugarcane; b) identify the sugarcane harvesting system (mechanized and burned); c) verify the areas of sugarcane in accordance to the state legislation; d) identify the areas of conversion of land use in sugarcane. In order to carry out the research, the following materials were used: OLI / Landsat 8 images, PROBA-V microsatellite data, MLME generated fraction images, SRTM images, SatVeg digital platform data, land use data and land cover IBGE and vector files from Minas Gerais, as well as software such as Spring, QGIS and InterImage. Based on these data, the methodological procedures consisted of image preprocessing, automatic classification and data validation. Regarding the results found, 837,829 hectares (ha) of sugarcane were identified, of which 435,803 ha were harvested by the mechanical system, 95,128 ha with pre-harvest the usage of fire and 306,898 ha not harvested during the analyzed period. In addition, the areas occupied by sugarcane with fire usage in areas of declivity below 12% were 88,414 ha, while in areas greater than 12% that would justify the use of fires was 6,714 ha. As for the conversion of land occupancy and land cover to sugarcane, more than half (53%) corresponded to agricultural areas, 21.7% areas that were previously pastures, followed by the mosaic of occupancy in the forest area with 71,032 ha (8.5%). The methodology allowed to obtain a Kappa index of 0.76 and Global accuracy of 0.96, proving promising as to the identification of sugarcane, thus identifying the areas of sugarcane, monitoring both the environmental system, economic and social changes guaranteeing the sustainability of sugarcane cultivation in the country.