Associação entre os métodos SCS-CN e GRADEX para cálculo de vazões máximas

• Ano de defesa: 2016
• Autor(a) principal: Taina Ulhoa Mota
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais
• Palavras-chave em Português: Hidrologia Modelos; Recursos hídricos Desenvolvimento; Engenharia sanitária; Bacias hidrográficas
• Link de acesso: https://hdl.handle.net/1843/BUOS-AVRJPM

The lack of systematic streamflow monitoring with adequate extension, periodicity, temporal resolution and quality is the Brazilian reality in many watersheds where hydraulic structures or floodplains occupation is necessary. The engineering, then, is limited to simple rainfallrunoff models to estimate the peak discharges and flood volumes such as the deterministic method SCS-CN. Although the SCS-CN model is widely accepted in the world, several authors have questioned the results obtained for watersheds that have physical characteristics significantly different from those originally studied. An alternative probabilistic method for calculating the maximum flood volume in watersheds with short systematic streamflow data is the GRADEX method, which extrapolates the flood volume frequency distribution using rainfall series. Although this is a consolidated method that allows the calculation of more realistic flood volumes, GRADEX method is seldom applied in Brazil mainly because of the difficulties found in solving its initial hypotheses. The present research suggests, therefore, the association between both methods, aiming a methodology of easier application and that reduce the uncertainties involved in the determination of maximum peak discharges. In order to perform it, a case study for the Serra Azul river basin, at the location of Juatuba MG, is presented based on 12 years of continuous rain and runoff data and 14 additional years of rain data. This association occurs in the definition of the lower and upper limits of the probability density function of runoff deficit, hR(r), existing variables in the GRADEX method, based on the CNASYMPTOTIC concept proposed by Hawkins (1993) in adaptation of the SCS-CN method. The thirteen modelled scenarios, based on the largest rainfall-runoff monitored events, show the complexity involved in understanding the response phenomenon of a given watershed to rainfall events, resulting in several possibilities for extrapolation of the flood volume frequency distribution. The methodology gave a range of results that best support the definition of the basin saturation condition and, consequently, the calculation of the maximum peak discharges compared to the originally proposed SCS-CN and GRADEX methods