Método para análise de frequência e de gestão do risco de cheias, a partir da informação hidrometeorológica, sob a condição de não-estacionariedade

• Ano de defesa: 2016
• Autor(a) principal: Jose Genivaldo do Vale Moreira
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; UFMG
• Palavras-chave em Português: Meio Ambiente e Recursos Hídricos; Saneamento; Recursos hídricos Desenvolvimento; Engenharia sanitária; Controle de inundações
• Link de acesso: http://hdl.handle.net/1843/BUOS-AUFG7W

Most statistical models used in water resources management focuses on the premise of a stationary state along the time scale. However, the nonstationary trend in hydrological time series became a topic of great importance in recent decades and of notorious acceptance by the scientific community and national and international organizations. Therefore, the tests require increasingly robust models, characterized by providing reasonable levels of precision inference, to aid in the control of impacts associated with the occurrence of hydrological extreme events, such as severe floods. In this sense, the main focus of this work is the development of an auxiliary methodology to aid decisions in flood management, under the condition of nonstationary, based on the hydro-meteorological information. The method involves the analysis of the rupture of the premise of stationary state in hydrological series through the probability of type I and type II errors, of statistical significance, which are integrated in the decision support tool for the management of flood risk. The frequency analysis, especially the techniques for the estimation of the parameters of the relevant distributions to model GEV and the incorporation of hydro-meteorological information to flood data, under nonstationary, are also highlighted in this method. Regarding the transfer of information, it was used the Gradex method, taking into account the proper applicability under such conditions in accordance with the assumptions of the baseline method. The study presents an example of application of the proposed method, from observations recorded in the municipality of Tarauacá, located in the central region of the State of Acre, in the western part of the Brazilian Amazon. The results converge to the conclusion that, under a scenario of control of the impacts of floods, the decision is to recommend the deployment of a measure of control, since the cost of regret associated with the underestimation proved to be higher than that for the overestimation. This decision goes against that which was obtained through the conventional analysis that is held under the premise of stationary.