Caracterização de sedimento de reservatórios por espectroscopia como suporte para prática de reuso do sedimento em região semiárida

• Ano de defesa: 2020
• Autor(a) principal: Carvalho, Thayslan Renato Anchiêta de
• Orientador(a): Medeiros, Pedro Henrique Augusto
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Reuso de sedimento; Espectroscopia; Mínimos quadrados parciais; Reservatórios de superfície; Semiárido; Brasil; Sediment reuse; Spectroscopy; Partial least squares; Surface reservoirs; Semiarid; Brazil
• Link de acesso: http://www.repositorio.ufc.br/handle/riufc/65664

Soil erosion by water yields sediment to surface reservoirs, reducing their storage capacities, changing their geometry and degrading water quality. Sediment reuse, i.e. fertilization of agricultural soils with the nutrient-enriched sediment from reservoirs, has been proposed as a recovery strategy. In this study, we characterize the sediment silted in reservoirs in the densely dammed semiarid region of Brazil by VIS-NIR-SWIR spectroscopy and assess the effect of spectral resolution and spatial scale on the accuracy of nitrogen, phosphorus, potassium, carbon, organic matter and clay prediction models. Sediment was sampled in ten reservoirs, and physical and chemical laboratory analyses were performed, as well as spectral readings. Partial least square regressions performed satisfactorily to very well in the prediction of clay, organic matter and carbon at spatial scales from micro (reservoirs < 0.1 km²) to large (catchments from 100 to 10,000 km²). Models for nitrogen, potassium and phosphorus were more unstable and performed unsatisfactorily in some situations. At the macro scale (basins > 10,000 km²), models for all the sediment attributes performed unsatisfactorily, except for clay. Coarsening spectral resolution by up to 10 nm degrades only slightly the models’ performance, indicating the potential of characterizing sediment from spectral data captured at lower resolutions, such as by hyperspectral satellite sensors. By reducing the costly and time-consuming laboratory analyses, the method helps to promote the sediment reuse as a practice of soil and water conservation.