Comportamento espectral das culturas de milho, arroz e soja, usando imagens dos Satélites Sentinel-2, CBERS-4 e LANDSAT-8

Imagem de Miniatura
Tipo
Dissertação
Data
2020-02-03
Autores
Valencia, Lady Victoria Angulo
Orientador
Pamboukian, Sergio Vicente Denser
Título da Revista
ISSN da Revista
Título de Volume
Membros da banca
Duro, Magda Aparecida Salgueiro
Araújo, Sidnei Alves de
Programa
Ciências e Aplicações Geoespaciais
Resumo
Monitoring crop status at regional level is critical in the pursuit of sustainable Agricultural Management. One of the methodologies applied for this purpose is Precision Farming which together with the technological tools and equipment (Global Navigation Satellite Systems, Remote Sensing, Geographic Information Systems and Unmanned Aerial Vehicles) form an integrated agricultural system based on information of crop production and evolution. Since the 1970s, marked by the launch of the Landsat-1 satellite platform, the study of crop status has been defined by the relationship between the interaction of electromagnetic radiation in soil and vegetation and its response over different wavelength ranges (measured by reflectance) related to the leaf chemical composition of the crops. In this research, we analyzed the spectral signatures (wavelength-related spectral reflectance values) and the main vegetation indices (NDVI, GNDVI, SAVI, EVI) for the Corn, Rice and Soybean crops in a cultivated area of 320 hectares, located in the municipality of Tremembé (SP) for the period from November to March of the 2017/2018 and 2018/2019 cycles, through the use of multispectral images of the Sentinel-2, CBERS-4 and Landsat-8 satellites. The results show higher spectral responses (especially in the RedEdge and Near-InfraRed region) in the soybean crop, followed by rice and corn around 60 to 90 days of sowing and a decrease with the crop approach, which is due to loss nutrients and leaf chlorosis. In addition, a high correlation (0,972) in the enhanced vegetation index (EVI) between CBERS-4 and Landsat-8 satellite images is observed. The vegetation index calculation also shows that the 2018/2019 crop cycle had more homogeneous results and a higher percentage of areas with high values compared to the first cycle. In addition, the studied images show that there is a variation of reflectance due to several factors such as vegetation age, growing season, rainfall, temperature, among others. Nevertheless, reflectance graphs show great potential for monitoring nutrient and water content in crops, as well as vegetation indices indicate a good advantage in input management and field visits
Descrição
Palavras-chave
earth observation mission from the Copernicus Programme-2 (sentinel-2) , China-Brazil earth resources satellite-4 (CBERS-4) , land satellites observation program-8 (landsat-8) , sensoriamento remoto , agricultura de precisão , assinaturas espectrais , refletância espectral , índices de vegetação
Citação
VALENCIA, Lady Victoria Angulo. Comportamento espectral das culturas de milho, arroz e soja, usando imagens dos Satélites Sentinel-2, CBERS-4 e LANDSAT-8. 2020. 118 f. Dissertação (Ciências e Aplicações Geoespaciais) - Universidade Presbiteriana Mackenzie, São Paulo, 2020.