Estudo comparativo da aplicação do modelo soil and water assessment tool (swat) e métodos usuais de regionalização de vazões mínimas de referência Q90

• Autor(a) principal: Ferreira, Thiago Gomes
• Data de Publicação: 2019
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da Universidade Federal do Espírito Santo (riUfes)
• Texto Completo: http://repositorio.ufes.br/handle/10/11055
• topic: Water producion; Flows; Reference flows; SWAT; Q90; Produção de água; Vazões; Vazões de referência; Engenharia sanitária; 628

Estimating flows in hydrographic basins is a constant problem in hydrological studies and management of water resources. Specifically, when analysing the reference flow Q90, several methodologies are found to determine these values, however, this theoretical path is surrounded by uncertainties and laborious efforts. Eletrobrás is one of the several methods of determination of Q90. It was proposed in 1985 and it has been widely used throughout Brazil. In Espírito Santo authors such as Euclydes et al. (2007) and Ferreira (2010) applied this method to the Timbuí River basin, which consists of a watercourse of significant length and potential for human supply, but it has only one stream gaging station. On the other hand, with the technological advance, methods such as SWAT (Soil and Water Assessment Tool) have been widely used to determine flows in several locations, including Espírito Santo. Therefore, this study aims to analyze the Q90 results obtained through the flow regionalization equations already developed by the authors cited previously and the results from the flow continuity curve obtained from SWAT with an uncalibrated simulation to a region that includes the Valsugana Velha-Montante fluviometric station (sub-basin 115) and to a location near the mouth of the river (sub-basin 207), where there was no monitoring. This study is divided into three steps, the first one is the hydrological modelling with SWAT, which presents the input data of the model (climatic, physiographic and fluviometric variables), the sensitivity analysis of the parameters of the model through the values of p-value and t-stat and the calibration attempts with a Nash-Sutcliffe Coefficient equal to 0.17. The second step is used to determine the Q90 flow rate with SWAT data, where the distribution of the simulated flows between the simulation period (01/01/2007 to 12/31/2015) in a which allowed the determination of the flow rate, which was equal to or exceeded 90% of the time (Q90). Finally, in the third step, the equations of Euclydes et al. (2007) and Ferreira (2010) were applied to the basins studied. As a conclusion, the values obtained by the two methodologies were compared and it was found that the values obtained by the SWAT simulation diverge from those obtained through the regionalization equations for the subbasins. This comparative measure was performed with the aid of Relative Error, a method widely used for comparative analysis in hydrological models.