Abstract
This study used SCS-CN method to estimate the real recharge of the study area which is one of the most reasonable techniques to estimate groundwater recharge when there is no available runoff data in a watershed. From the results of tile real recharge analysis for the study area using SCS-CN method, it was analyzed that the year 1994 when the drought was severe shotted the lowest recharge of 106.3mm with recharge rate of 12.4%, and the highest recharge of 285.6mm with recharge rate of 21.8% occurred in 1990. Yearly average recharge of 213.2mm was obtained, and tile average recharge rate was 16.9%/year. KOG-FLOW model which has powerful post process functions consists of setting environments for input parameters in Korean language, and help function is added to each input data. Detailed information for each parameter is displayed when the icon is placed on the input parameters, and geologic boundaries or initial head data for each layer can be set easily on work sheet. The relative errors (R. E.) for each model's observed values and calculated values are $0.156{\sim}0.432$ in case of KOG-FLOW, and $0.451{\sim}1.175$ in case of WINFLOW, therefore it is known that KOG-FLOW model developed in this study produced results compared to observed head values.