Precipitação estimada por satélite para uso em modelo concentrado chuva-vazão aplicado em diferentes escalas de bacias

• Autor(a) principal: Paula, Stefany Correia de
• Data de Publicação: 2015
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Manancial - Repositório Digital da UFSM
• Texto Completo: http://repositorio.ufsm.br/handle/1/7906
• topic: IPH II; TRMM; Efeito de escala; Scale effect; CNPQ::ENGENHARIAS::ENGENHARIA CIVIL

In Hydrological rainfall-runoff models, rainfall is one of the main variables analyzed before being used as input data. However, its peculiar characteristics of occurrence make it challenging to measure, especially for developing countries, since it requires monitoring networks with high spatial and temporal resolutions. To overcome such problems, the use of remote sensing to estimate rainfall is being gradually diffused, with results considered of good reliability. In this sense, the objective was to evaluate the results generated by a rainfall-runoff modeling, using precipitation information estimated by the TRMM satellite, applied to four watersheds of different scales. To achieve this objective was used the concentrated model IPH II in the version for Windows® (WIN_IPH2 version 2, denominated in this work as WIPH2). In a first analysis, conducted for the watersheds of the rivers Turvo (1,540 km2), Ijuí (9,450 km2), Jacuí (38,700 km2) and Alto Uruguai (61,900 km2), which are located in southern Brazil, it was observed that the precipitation estimates provided by the TRMM satellite generated consistent spatialized mean rainfall in relation to the generated by rain gauges networks. Were found errors about 7.5% (Ijuí River watershed) to 15% (Turvo River, Jacuí River and Alto Uruguai watersheds). In overall, the daily mean rainfall, generated by the TRMM estimates, increased their correlation with the rain gauges information, as there was an increase in the watershed drainage area. In the modeling, it was noted that the WIPH2 showed good fit for the simulation of flows in the selected watersheds, with exception to the application in the Alto Uruguai watershed. In some cases, even though the mean rainfall estimated by the TRMM had generated greater uncertainty regarding the data generated by the different rain gauge networks, the model was able to compensate the deficiency of this information, by changing parameters within limits considered acceptable for the characteristics watershed. It was found that, in overall, precipitation estimated by the TRMM produced improvements in the flow evaluation coefficient (Nash-Sutcliffe NS, and correlation R coefficients), as there was an increase in the watershed drainage area until the Jacuí River watershed, with a decrease of the coefficients for the Alto Uruguai watershed, which was probably caused by the inadequacy of the WIPH2 concentrated model to the heterogeneous characteristics of the last watershed. In places with good rainfall monitoring coverage, the satellite estimates produced lower results than those generated by information measured by rain gauges. In places where the lack of equipment is accentuated and/or their distribution is non-uniform, the use of satellite estimates proved to be more reliable, reaching clearly superior results. In this sense, it was concluded that the TRMM satellite estimates may present as good alternatives to watersheds with scarce rainfall information measures on the ground, having great chances to show better results when compared to information of rain gauges scarce and poorly distributed.