Performance assessment of Copula functions in estimation of rainfall losses and rainfall-runoff modelling (Case study: Kasilian Watershed)

One of the most important challenges of researchers in rainfall-runoff modelling is the estimation losses and extracting excess rainfall, which can affect on accuracy of the model and hydrograph characteristics. In the present study, at first the infiltration index of was estimated based on the depth of runoff and rainfall and also effective rainfall time. Then, instantaneous unit hydrograph was computed through Russo Method and the dimensions of direct runoff hydrograph were determined for 20 rainfall- runoff events. Then in order to increase the precision in estimating the dimensions of obtained hydrograph, the value of penetration index was estimated based on bivariate distribution resulted from infiltration index and one of rainfall characteristics which obtained from Copula function. For this purpose, at first the correlation between infiltration index of , characteristics of rainfall hyetograph and runoff hydrograph was calculated, then the best- fitted marginal distribution on each variable was specified. Finally, Galambos function was chosen as the best copula function for creating bivariate distribution for pairs of infiltration index of and rainfall height, infiltration index of and maximum velocity, infiltration index of and average flow and also infiltration index of and velocity. Therefore, the hydrograph dimensions were obtained for each event. Comparing the various dimensions of computed and observed hydrograph by copula function in Russo method showed that the infiltration index of can be estimated more accurately by using the copula function.