Calibration of Muskingum Method using Ant Colony Algorithm (Case Study: Garmrood Basin)

River and reservoir routing is very important in prediction of flood outflow from the catchment area. Therefore, accuracy of this calculation has close relationship with the increasing of accuracy in the outflow from the catchment area. The Muskingum method is a river routing methods, K and X coefficients are determined according to the type of catchment area and the river. But in most watersheds, these coefficients are considered approximately due to lack of parameters. Therefore, the aim of this study is to calibrate the Muskingum method in the Garmrood basin by determines K and X coefficients. Therefore At first, HEC-HMS software was used to calculate the flood for each sub-basin and then river routing was calculated using a computerized model. Aco algorithm is used in order to calibrate total of 36 decision variables (including 18 variables for K and 18 variables for X). Root Mean Square Error between computed and observed hydrograph data were taken as objective function. The results showed that the optimal coefficients reduced the value of the objective function about 28% and 4.2% respectively in calibration and verification periods.