Sensitivity analysis of rainfall–runoff parameters models to estimate flows

Determinate the runoff of a watershed is a challenge due to the complexity of representing all “inlets” and “outlets” involved in a rainfall–runoff model. Therefore, methodologies applied for this purpose should have a good representation of the variables that most influence in this process. One of the models used to calculate the design flow is the (USDA in Urban Hydrology for Small. Technical release, no 55 (TR-55). Soil Conservation Service. Washigton, DC, http://scholar.google.com/scholar?hl=en&btnG=Search&q=intitle:Urban+Hydrology+for+Small+watersheds#1, 1986), which considers the analysis changes in soil coverage, time of concentration (tc), and recurrence period (T). In this way, this study sought to evaluate the hydrological behavior of a watershed with an increase in soil waterproofing. These modifications were correlated with the variation of runoff coefficients (CN), modifications of the periods of recurrence indicated by the literature, and different equations of the time of concentration. Its application was carried out in the Ribeirão do Suru watershed, Santana de Parnaíba, SP, Brazil. The CN {75; 80; 85; 90} increased 3.14, 5.61, 10.90 and 15.85%, respectively. In the most critical situation, runoff was 15.85% higher in estimated CN. The variation of precipitation as a function of T (2, 5, 10, 25, 50, 100 and 500) and application of 11 time of concentration methods designed 132 hydrographs and flow values that were statistically treated in T of Student and in the Analysis of Variance (ANOVA). Except for Bransby Willians associated Cinematic Method, Dooge with Johnstone and CTH with Tsuchyia, the pairs showed degrees of correlation below 59%. The greatest correlation was observed in Jonhstone with Dooge (90%), followed by the Kinematic Method with the Soil Conservation Service Method (83%) and with Dodge (74%). As a result, it was possible to demonstrate the behavior of the SCS parameters to minimize subjectivities and revealing how each parameter impacts the flow of the watershed. Finally, the sensitivity attributed to T was the highest among the three analyzed.