Simulação da vazão e uso e ocupação da terra na Bacia do Alto e Médio Rio Teles Pires

• Ano de defesa: 2012
• Autor(a) principal: Onga, Naomi
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Mato Grosso; Brasil; Instituto de Ciências Humanas e Sociais (ICHS); UFMT CUC - Cuiabá; Programa de Pós-Graduação em Geografia
• Palavras-chave em Português: Modelagem hidrológica; Uso e ocupação da terra; NGFlow; Curve number; CNPQ::CIENCIAS HUMANAS::GEOGRAFIA; Hydrological modeling; Land use and cover
• Link de acesso: http://ri.ufmt.br/handle/1/1290

The advance of the agricultural frontier towards north in the state of Mato Grosso favored by soils, climate, topography and water availability has resulted in the predominance of large-scale crop-farming production systems, which has change vegetation cover and therefore the hydrological cycle. To evaluate the predictability of these ongoing alterations, NGFlow, a semi-distributed precipitation-runoff model was applied, developed specifically for data-poor regions and implemented in a Geographical Information System. During the re-engineering and extension of the original application, the Curve Number method was included in the model to allow runoff simulations as a function of soil types and land use and cover. Monthly water balance and runoff were simulated for the watershed of the Upper and Mid Teles Pires river between 1994 and 2003, with and without using the CN module. Model performance was evaluated through the comparison of observed and simulated discharge at six river gauges. Both simulations, with and without the CN module showed strong overestimates of runoff in 1994, a year with higher precipitations than the average. In 1997, a dry year, strong underestimates were observed. In 2003, a year of average precipitation, best model performance was obtained, showing only slight underestimates. The evaluation of monthly runoff shows that the un-calibrated model tends to underestimate runoff under base flow conditions and that the use of the CN module does not systematically improve performance of the un-calibrated model. Obtained performance indicators such as the Root Mean Square Error, the Efficiency Coefficient of Nash e Sutcliff and the Coefficient of Determination show that the NGFlow model must be calibrated for operational use in monthly runoff simulations in the basin.