Estimativa de área de soja e milho cultivado no Estado do Paraná utilizando-se do perfil espectro-temporal de índices de vegetação

• Ano de defesa: 2013
• Autor(a) principal: Souza, Carlos Henrique Wachholz de
• Orientador(a): Mercante, Erivelto
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Estadual do Oeste do Parana
• Programa de Pós-Graduação: Programa de Pós-Graduação "Stricto Sensu" em Engenharia Agrícola
• Departamento: Engenharia
• País: BR
• Palavras-chave em Português: Sensoriamento Remoto; Máscara de Culturas; EVI; NDVI; WDRVI
• Palavras-chave em Inglês: Remote Sensing; Crop Mask; EVI; NDVI; WDRVI
• Área do conhecimento CNPq: CNPQ::CIENCIAS AGRARIAS::ENGENHARIA AGRICOLA
• Link de acesso: http://tede.unioeste.br:8080/tede/handle/tede/2609

The use of remote sensing technology has been studied as a way to make the current system of monitoring and crop forecasting in Brazil more efficient, dynamic and reliable. One of the difficulties found in the use of medium spatial resolution images as MODIS (250 Meters) is that the spectral profiles of temporary crops, as soybean and corn, may present similar curves, difficulting the separation of these cultures at the time of classification of the seeded areas. In this sense, the aim of this work was analyzing the pattern of temporal profiles, from the vegetation index (VI) EVI (Enhanced Vegetation Index), NDVI (Normalized Difference Vegetation Index) and WDRVI (Wide Dynamic Range Vegetation Index), obtained by the MODIS images for the crops of corn and soybean in the crop years of 2010/2011 and 2011/2012 in the state of Paraná. The aim was performing the spectral separation of these cultures to make its mapping. The applied methodology allowed the discrimination of areas with soybean and corn (masks) for each crop year. The areas of the masks were extracted and compared with SEAB official data, finding adjustments in "R ²" between 0.89 and 0.94 for soybean and from 0.43 to 0.83 for corn. For the Willmott coefficient (d) values were between 0.85 to 0.87 for the soybean crop and 0.63 to 0.76 for corn. The accuracy of spatial masks using images with high spatial resolution achieved the best results with the IV WDRVI with overall accuracy (OA) = 86% and = 0.78, and Kappa Index (KI) IV EVI with OA and KI = 83% = 0.74. Based on these results, it can be conclude that the proposed methodology is promising and may be used for mapping of these crops in the estimation of the state area.