Infiltration methods affect the results of watershed runoff simulation. There are three major infiltration methods frequently used: the Green-Ampt method, the U.S. NRCS (Natural Resource Conservation Service) method, and the Horton method. The NRCS method and the Horton method were derived from experience or experiment, and the Green-Ampt method was derived from a physical mechanism of groundwater flow. Most Korean engineers prefer to use the NRCS method rather than the Green-Ampt method, as the parameters of the NRCS method are easily determined from land use and soil maps. The NRCS method, however, was derived from the soil types and the land use classifications of U.S.A., and it does not conform to the land use classifications of Korea. The purpose of the study is to find more suitable method between the Green-Ampt method and the NRCS method in watershed runoff simulation. We also describe estimation procedures of the Green-Ampt parameters.
Using the SCE-UA method, we automatically calibrated the SWMM for watershed runoff continuous simulation with the NRCS method and the Green-Ampt method for the Milyang Dam Basin, Korea. The curve number of the NRCS method and the three parameters of the Green-Ampt method were estimated from the automatic calibrations including nineteen hydrological parameters, respectively. The Green-Ampt parameters are the saturated hydraulic conductivity, the average capillary suction head at the wetting front, and the initial moisture deficit. The verification results were evaluated with the four assessment indexes and all indexes from the Green-Ampt method were better than those of the NRCS method.
A proper estimation method of the Green-Ampt parameters is needed to simulate watershed runoff from an ungagged basin. The saturation hydraulic conductivity and the wetting front suction head were estimated from digital soil maps and land use maps. Each parameter estimated from the digital maps showed very small difference from that obtained by the automatic calibrations. To verify the applicability of the estimation procedure for the Green-Ampt parameter from digital maps, the SWMM was performed for the Hoengseong Dam Basin, Korea. There was little difference between the results of automatic calibration including the Green-Ampt parameters and the results of automatic calibration excluding the Green-Ampt parameter, in which the Green-Ampt parameters were estimated from the soil maps.
The two parameters of the Green-Ampt method can be reasonlly estimated from digital soil maps and land use maps. The estimation procedure of the Green-Ampt parameters applied in the study would be helpful for a watershed runoff continuous simulation in Korea.