Metodologia da Curva Número: Determinação de parâmentros para sub-bacias hidrográficas em Porto Alegre-RS

Abstract

The Curve Number method of the Natural Resources Conservation Service (NRCS) stands out as one of the most frequently used in runoff calculation based on rainfall events, which is fundamental to the design of hydraulic works. However, the recommended values for CN and the initial abstraction ratio (λ) remain uncertain and are discussion points in the method, taking into consideration that it was initially conceived using hydrologic data for basins in the United States. Therefore, the present study verified the values of CN and λ for six watersheds of Arroio Dilúvio in Porto Alegre – RS using monitored hydrologic data. With that goal, the performance of five different methods for the definition of the parameter values for runoff estimation were compared: CN based on the NRCS recommended tables, CN fitted for a λ value of 20%, fitted for a λ value of 5%, λ determined using the event analysis method and λ determined using the model fitting method. The results indicate that the tabulated CN values, in general, were lower than the fitted for a λ value of 20%, and that the method resulted in the worst general performance for runoff estimation. The initial abstraction ratio values obtained using the event analysis showed considerable dispersion in some drainage basins, yet, for 83,33% of the events, the observed value of λ is lower than 20%, being closer to 5%. Between the predetermined values of λ, the one that produced the best performance after the CN fitting was 5%. On the other hand, the best general performances for runoff estimation were obtained for λ calculated using the event analysis and the model fitting methods for each drainage basin. In the first, the median values of the initial abstraction ratio ranged from 0,003 to 0,089 between the drainage basins, while the median values of CN ranged from 26,58 to 95,56. In the second, λ ranged from zero to 0,011, and the median values of CN ranged from 22,5 e 95,0. The value variation of the best fitted λ and CN for each drainage basin emphasizes the importance of regional calibration of these parameters in order to add reliability and security to the design of hydraulic projects. Nevertheless, the value of Pbias for the adjusted parameters aren’t within the accepted limits for the Cascata-I basin, which reinforces the necessity for future widening and deepening of the study.