Modelagem da vazão na bacia hidrográfica do rio Araguari, em Minas Gerais

• Ano de defesa: 2016
• Autor(a) principal: Moraes, Márcia Regina Batistela
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Uberlândia; Brasil; Programa de Pós-graduação em Qualidade Ambiental
• Palavras-chave em Português: Qualidade ambiental; Bacias hidrográficas - Minas Gerais; Recursos hídricos - Desenvolvimento - Minas Gerais; Hidrologia; Vazão de referência; Q7,10; Q95; Q90; Vazão mínima; Low-flow; Minimum flow; CNPQ::CIENCIAS EXATAS E DA TERRA::GEOCIENCIAS
• Link de acesso: https://repositorio.ufu.br/handle/123456789/17864; http://doi.org/10.14393/ufu.di.2016.504

This study aimed to model the low-flows, long-period mean flow, annual minimum flow, annual and monthly high-flow in five Araguari River Basin’s gaging stations, in Minas Gerais, Brazil. Regardless of, this study also aimed to compare low-flows estimates and modeled long-period mean flow with those calculated in the bibliographies Deflúvios Superficiais de Minas Gerais and Atlas Digital das Águas de Minas. The studied gaging stations were: Desemboque, PCH Pai Joaquim/Fazenda Boa Vista, Ponte BR-452, Capim Branco and Letreiro. It was verified that the high-flow estimate was not possible, in any of the studied stations. The annual series of seven days minimums could be estimated only in function of annual minimum flow, from the 95th percent and 90th percent exceedances and from the mean flow in july. The regression models generated proved to be capable of estimate the annual series of seven days minimums with security and low error values. It was also verified, a very high discrepancy in the annual series of 7-day minimums with 10-year recurrence interval, calculated with annual series of 7-day minimums observed data, in relation to estimation through Deflúvios Superficiais de Minas Gerais (mean relative error of -65.18%) and Atlas Digital das Águas de Minas (mean relative error of -42.14%). Concerning the 95th percent exceedances, it was possible to estimate it only in function of annual minimum flow, from the 90th percent exceedances and from the mean flow in july. The regression models proved to be capable of estimate the 95th percent exceedances with security and low error values. It was also verified, a very high discrepancy in the values of that calculated flow with observed data, in relation to estimation through Atlas Digital das Águas de Minas (mean relative error of -31.23%). In relation to the 90th percent exceedances, it was possible to estimate it in function of annual minimum flow and the mean flow in july. The flows with 90th percent exceedances estimated through Atlas Digital das Águas de Minas, compared to those calculated with observed data, reached mean relative error of -28.85%. The flow with 90th percent exceedances estimated through the regression models in function of annual minimum flow and the mean flow in july, the error was 19.35%. The regression models to estimate the annual minimum flow in function of the mean flow in july reached relative errors of 0.20% and -11.59%. The annual mean flow estimated through Atlas Digital das Águas de Minas, compared to those calculated with observed data, presented relative error of -31.95% in the Ponte BR-452 station. In the Letreiro station the error was zero. On the other hand, when the annual mean flow was estimated through the regression models in function of the mean flow in march, this error was 3.15% and -1.42%, in the Ponte BR-452 and Letreiro stations, respectively.