Modelagem hidrológica: técnicas de geoprocessamento aplicadas ao modelo SCS Soilcon Servation Service Curve Number

Abstract

For hydrology, the watershed its the primary unit of study, which dynamics its influenced by various elements. The effort of science to place these elements in interaction originated hydrological models. These models aim to simulate parts of the hydrologic cycle, for, between other objectives, dimension drainage structures. This study highlights the model developed by the Soil Conservation Service, known as Curve Number, which considers the physical parameters of the watershed to estimate the maximum runoff for a given rainfall event. In this regard, GIS techniques were applied in obtaining these parameters in order tocompare the final results with those obtained by hydrological studies conducted without the application of such techniques. The physical characteristics that compose the model were obtained from a digital elevation model hydrologically consistent generated from cartographic data available. Subsequently, by specific computational algorithms, was generated a map of the direction of water flow and, from this, the definition of the contributing area of watershed. The lenght of river was interpreted based on the topographical data at a 1:50,000 scale, and, thereafter, effectively slope calculations was performed. Remote sensing techniques were performed to obtain the classification of Landsat 7 satellite images, with reference to the soil uses predefined by SCSCN.This classification was crossed with the soil map of the region, obtainingthe value of the parameter Curve Number. Precipitation was calculatedaccording to the rainfall equation for the representative pluviometric post of the region. Thereafter, the modeling was performed to compare the use of GIS techniques with conventional methods applied. The results showed that the techniques applied returned values closer to the true about the parameters, area and length of river. The effective slope calculated in a distributed manner showed a lower value than that used in the conventional method, which greatly influenced in the calculation of time of concentration and peak. The parameter CN has been shown numerically close, however, this parameter is very sensitive to the model, contributing to the discrepancy in the comparison of thefinal results. Finally, it is concluded that the increase in accuracy of theparameters did not result in increased accuracy of the model, however, in front of the adjustments aplied to the conventional method, it does not represent the inapplicability of these techniques. Was diagnosed that the model needs to be adapted.