Low Flows Hydrological Regionalization and Definition of Homogeneous Regions using Multivariate Statistical Analyses in the Uruguai River Watershed, on the Rio Grande do Sul State Side, Brazil

Low flow estimation is essential for water resource management and water quality regulation. The success in this estimate is achieved through delimited regions that have characteristics homogeneous with the region of interest, so that the minimum flows are estimated from reliable processes. The objective of this work was to identify homogeneous regions and generate regional regression equations to estimate low flows in the Uruguai River watershed, on the state of Rio Grande do Sul state side (Brazil), using multivariate statistical analysis. The Q 90 and Q 95 streamflows, representing the low flows in the basin, were obtained through fluviometric stations. Easy-to-obtain geomorphological and climatic variables were used to describe the low flows in 46 sub-basins on the Uruguay River. The homogeneous regions were delimited using cluster analysis, which were confirmed from the discriminant analysis, making the process robust and reliable. Regional equations were generated for each homogeneous region and for each low flow, where the variables area, perimeter and total summer precipitation were the ones that best described the low flows. The cross-validation used confirmed the good fit of the regional equations. It is hoped that this study will allow water resource managers to make more correct decisions in processes that need to estimate low flows in Uruguai River watershed, on the Rio Grande do Sul state side, such as grant analyzes, framing water courses and charging for water use. Regionalizacao hidrologica de vazoes minimas e definicao de regioes homogeneas usando analise estatistica multivariada na bacia hidrografica do Rio Uruguai, lado do estado do Rio Grande do Sul, Brasil R E S U M O Estimar as vazoes minimas e fundamental para a gestao dos recursos hidricos e a regulacao da qualidade da agua. Para ter exito nesta estimativa e preciso delimitar regioes que possuem caracteristicas homogeneas com a regiao de interesse, de modo que as vazoes minimas sejam estimadas a partir de processos confiaveis. O objetivo deste trabalho foi identificar as regioes homogeneas e gerar equacoes de regressao regional para estimar vazoes minimas na bacia hidrografica do Rio Uruguai, inserida no estado do Rio Grande do Sul (Brasil), usando analise estatistica multivariada. As vazoes Q 90 e Q 95 , representando as vazoes minimas na bacia, foram obtidas atraves de estacoes fluviometricas. Foram usadas variaveis geomorfologicas e climaticas de facil obtencao para descrever as vazoes minimas em 46 sub-bacias do Rio Uruguai. As regioes homogeneas foram delimitadas usando analise de agrupamento, as quais foram confirmadas a partir da analise de discriminante, tornando o processo robusto e confiavel. Foram geradas equacoes regionais para cada regiao homogenea e para cada vazao minima, onde as variaveis area, perimetro e precipitacao total de verao foram as que melhor descreveram as vazoes minimas. A validacao cruzada utilizada confirmou o bom ajuste das equacoes regionais. Espera-se que este estudo possa permitir que gestores dos recursos hidricos tomem decisoes mais corretas em processos que necessitam estimar as vazoes minimas na bacia hidrografica do Rio Urugiai, porcao do estado do Rio Grande do Sul, como analises de outorga, enquadramento dos cursos de agua e cobranca pelo uso da agua.