Avaliação espectral de solos de áreas alagáveis do noroeste do Paraná

• Ano de defesa: 2011
• Autor(a) principal: Chicati, Marcelo Luiz
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Estadual de Maringá; Brasil; Departamento de Agronomia; Programa de Pós-Graduação em Agronomia; UEM; Maringá, PR; Centro de Ciências Agrárias
• Palavras-chave em Português: Análise espectral; Sensoriamento remoto; Solos; Classificação; Mapeamento do solo; Solo de várzea; Linha do solo; Norte do Paraná (Estado); Brasil.; Lowland soils; Remote sensing; Spectral analyses; Mapping; Soil line; Northwest Paraná; State; Brazil.; Ciências Agrárias; Agronomia
• Link de acesso: http://repositorio.uem.br:8080/jspui/handle/1/1110

This work airned to evaluate the use of rernote sensing for soil classes discrimination and their attributes using spectral reflectance data obtained in laboratory and orbital leveis for fiooded plains. The region chosen for study is located in the northwestern of Paraná state and is comprised in part of the Environmentai Protection Area of the Paraná River lsiands and floodplains in a portion predominantly populated by Iowland soils. In this area were carried out sou sarnpling by rigid rnesh rnethodoiogy and were submitted to physical, chemicai and mineralogical analysis at the Universidade Estadual de Maringá. The spectral analyses were taken in spectroradiometer from 350 to 2500nm in the Escola Superior de Agricultura "Luiz de Queiroz" at Piracicaba. This information set, combined with the regional cartographic base enabled the maps elaboration for assessing the spectral data efficiency in sou units' separation. The results showed the occurrence of ten sou classes in the region, being Ultisois the rnost representative. lt was also found that among ali the soils present, the main differences concerned the color and, to a lesser extent, the predorninantiy sandy texture. The spectral curves confirrned the great similarity in the behavior of lowiand soils, the majority poor in iron oxides, mainly of gray color for the reduction of iron and with little influence of organic rnatter. The data from these spectral curves were tested by rnethods like "soil line" and then observed that the orbital information not obtained satisfactory correlations. From this it was obtained equations for sou classes discriminating and tested their respective correlations with the actual data. The correlation test of spectral data to distinguish the attributes was also performed and had most of their coefficients evaiuated positively. Finally, rnaps produced with the spectral data were tested and found satisfactory by the appropriate chi-square. So the main conclusion was that the rnethod is efficient for 10w lands when using laboratory spectral data, however, does not reflect the same resuits with orbital data.