Caracterização e modelagem de solos arenosos e seus atributos químicos, físicos e físico-hídricos por sensoriamento remoto

• Ano de defesa: 2020
• Autor(a) principal: Raimo, Luis Augusto Di Loreto Di
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Mato Grosso; Brasil; Faculdade de Agronomia e Zootecnia (FAAZ); UFMT CUC - Cuiabá; Programa de Pós-Graduação em Agricultura Tropical
• Palavras-chave em Português: Espectroscopia; Frações areia; Pedologia; Micromorfometria; Retenção de água; CNPQ::CIENCIAS AGRARIAS; Spectroscopy; Sand fractions; Pedology; Micromorphometry; Water retention
• Link de acesso: http://ri.ufmt.br/handle/1/3513

Despite the complexity of management, high susceptibility to erosion, extensive areas of coverage and growing inclusion as crop areas, with the expansion of agricultural frontiers, studies of sandy soils are rare. Remote sensing, both terrestrial and orbital, has proven great potential for soil characterization, with low investments of time and resources, but it is predominantly limited, until now, to clay soils. In view of the above, the objective of the present study was to characterize pedological classes and chemical, physical and physical-hydric attributes of sandy soils via terrestrial and orbital remote sensing. To complete this objective, two data sets were used. One of them, composed of 29 profiles located in the west and center-south of Mato Grosso state, and another, by 216 sampling points collected from 0 to 20 cm and located in a predominantly sandy agricultural area in the southeast of Mato Grosso. Chemical, physical and physical-hydric analyses were performed by traditional methods. Information was also obtained via remote sensing, by spectroscopy, at Vis-NIR-SWIR (VNS) and MIR wavelengths, and by X-ray fluorescence (pXRF). The orbital sensing data were obtained from exposed soil images (Synthetic Soil Image) of Landsat 5 and Sentinel 2, and simulations of Landsat 5, Sentinel 2 and Terra (ASTER). The integration of VNS, MIR and pXRF information allows pedologically complex characterizations and conclusions about sandy soils in a short period and with low investments in analysis and reagents. VNS and MIR spectra and pXRF data, used separately and/or in association, presented high predictive potential for clay, sand, sand subfractions, cation exchange capacity, pH, base saturation, organic matter, field capacity and permanent wilt point. For both orbital and terrestrial imagery, the reflectance intensity of sandy soils is inversely proportional to the particle diameter of the predominant sand sub-fraction. Orbital data, obtained 800 km away from the earth's surface, present high accuracy for the prediction of sand, clay and sand subfractions of sandy soils. Water retention of sandy soils is highly influenced by the roughness of sand particles. VNS and, especially, MIR spectra, have high capacity to identify different patterns of sand particle roughness in sandy soil samples. This study presents new approaches and potential uses for spectroscopy applied to soil science, utilizing the fast characterisation of sandy soils, with low cost and without the use of reagents.