Calibração do modelo hidrossedimentológico SWAT na bacia hidrográfica do córrego Samambaia, Goiânia - GO

• Ano de defesa: 2014
• Autor(a) principal: Veiga, Aldrei Marucci
• Orientador(a): Soares, Alexandre Kepler
• Banca de defesa: Soares, Alexandre Kepler, Siqueira, Eduardo Queija de, Santos, Roberto Ventura
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Goiás
• Programa de Pós-Graduação: Programa de Pós-graduação em Engenharia do Meio Ambiente (EEC)
• Departamento: Escola de Engenharia Civil - EEC (RG)
• País: Brasil
• Palavras-chave em Português: Calibração; Vazões; Sedimentos em suspensão; SWAT; SWAT-CUP
• Palavras-chave em Inglês: Calibration; Streamflow; Suspended sediment; SWTA; SWAT-CUP
• Área do conhecimento CNPq: RECURSOS HIDRICOS::PLANEJAMENTO INTEGRADO DOS RECURSOS HIDRICOS
• Link de acesso: http://repositorio.bc.ufg.br/tede/handle/tede/3497

This research paper focus on the analysis of SWAT model calibration in terms of flow and sediment in Samambaia River Basin, a small watershed (32.78 km2) located at Goiânia, Brazil. Streamflow and suspended sediment daily measurements have been carried out by February to December 2013 and climatic data were obtained form a weather station located inside the basin. Terrain data such as Digital Elevation Model (DEM), soil types, and land use were on the SWAT autocalibration too as well as on SWAT-CUP software, which is a specific too for automatic calibration. Initially, the simulation run in SWAT overestimated values of runoff peak and underestimated minimum discharges. However, the peaks were minimized and minimum discharges were adjusted to the observed flows after sensitivity analysis. By using different optimization schemes (GLUE, PARASOL and SUFI-2) in SWAT-CUP, an automatic calibration analysis has been done, which presented a better fit to the observed values (start with streamflow, than move to sediment). Statistical analysis using the coefficient of Nash-Sutcliff efficiency (COE) resulted in 0.80 and 0.88 for runoff and suspended sediment, respectively, which are considered good fits between simulated and observed values. The CN parameter, which is related to soil type, land use, and infiltration, showed the highest sensitivity in the calibration. After that, the alpha factor of base flow was another which showed higher sensitivity. The higher value obtained for the Manning roughness coefficient allows runoff to be damped. With regard to sediment calibration, parameters of sediment from landscape (USLE_P and USLE_C) as well as parameters of sediment from channel (SPCON and SPEXP) have been used in the calibration, once that they have shown higher sensibility.