An automated calibration of the SWMM with the role of groundwater component in a rural area

SWMM (storm water management model) has been widely used in the world as a typical model for runoff analysis of urban areas. The calibration of the model, however, may be an obstacle to easy application. We developed an automated calibration module of the SWMM linked with the SCE-UA (shuffled complex evolution-University of Arizona) that is a global optimization algorithm developed by Duan (1991). Another purpose of the study is highlighting importance of the groundwater component for watershed runoff event simulations in rural basins. We used the objective function of root mean square error in a calculated hydrograph. Applying an evolutionary algorithm, the SCE-UA decides a better set of parameters in the SWMM for runoff simulation. The number of parameters for estimation was twenty-two including ten parameters in the groundwater module. We applied the calibration model to two basins. One is the urban catchment of the Guro 1 Pumping Station Basin and the other is the rural catchment of the Milyang Dam Basin. The calibration results for the watershed runoff event simulation of the Guro 1 Pumping Station Basin were fairly good, even when the groundwater module was not included. On the other hand, calibration results for the watershed runoff event simulation of the Milyang Dam Basin were bad, when the groundwater module was not included. When we derived the calculated runoff hydrograph including the groundwater computation for the Milyang Dam Basin, the results were well fitted into the observed data. Therefore, it is required to include the groundwater component, when one applies the SWMM to a rural area for a runoff simulation. The automated calibration can lessen tedious efforts in a model calibration. (The grant No. 2011-0015225 from the National Research Foundation of Korea)