Avaliação do método Silveira com uso de dois modelos chuva-vazão para estimativa de disponibilidade hídrica em pequenas bacias com pequena amostragem de vazão

Abstract

The lack of hydrological data in small basins creates uncertainties that compromise the water resources management. Nowadays, the hydrometeorological network in Brazil mainly covers the large basins, difficulting the calibration of hydrological models in the conventional manner which usually requires a long period of monitoring. Aggravating the problem, there is the fact that the most of studies of hydrological regionalization do not have applicability to small basins. In this context, Silveira (1997) developed an alternative methodology for this demand, which is based on a combination of a simplified rainfall-runoff model with a reduced sampling of flow - at least three in periods without rainfall. So, its use is appropriate to determine a confident flow duration curve in the lower branch, providing subsidies for water resources management. Since the use of the methodology is growing, the objective of this study is to verify the applicability of this methodology to estimate median and minimum streamflow using two rainfall-runoff models in no data basins. To reach the purpose, four basins with small drainage areas and different land use were studied, two located near the campus of UFSM (Santa Maria / RS) and two in the city of Rosario do Sul / RS. In these basins, the methodology developed by Silveira was applied using two different rainfall-runoff models: i) MPB2, the simplified model developed in the Silveira methodology; and ii) IPH2. For this, from a series of observed flow were selected periods of drought – six days without rainfall (three days with flow measurements, intercalated). To these samples were adjusted the rainfall-runoff models, for subsequent generation of flow duration curve from each adjustment. Thus, the resulting flow duration curve of each event was compared to the flow duration curve of whole series of observed discharge, through statistical analysis and determination of confidence intervals. As a result, all basins showed a small estimation error with 90% confidence in the determination of flows with 90% and 95% of duration (Q90% e Q95%), which are normally used in procedures for granting rights of usage. Therefore, the tested methodology produced satisfactory results to aid the management of water resources. As the two rainfall-runoff models showed acceptable results for determination of minimum flows, it is suggested to adopt the MPB2 model, which consists of a simplified model and easy to apply, while IPH2 model requires greater caution of use as considerations described in this dissertation.