농업소유역의 홍수유출량 추정을 위한 단위도 모형 비교연구

Abstract

In this study, three different unit hydrograph methods (Snyder, SCS, Clark) in the HEC-HMS were compared to find better fit with the observed data in the small agricultural watershed. Baran watershed, having 3.85km² in size, was selected as a study watershed. The watershed input data for HEC-HMS were retrieved using HEC-GeoHMS which was developed to assist making GIS input data for HEC-HMS. Rainfall and water flow data were monitored since 1996 for the study watershed. Fifty five storms from 1996 to 2003 were selected for model calibration and validation. Three unit hydrograph methods were compared with the observed data in terms of simulated peak runoff, peak time and total direct runoff for the selected storms. The results showed that the coefficient of determination (R²) for the observed peak runoff was 0.8666～0.8736 and root mean square error, RMSE, was 5.25～6.37 m³/s for calibration stages. In the model validations, R2 for the observed peak runoff was 0.8588～0.8638 and RMSE was 9.57～11.80 m³/s, which were slightly less accurate than the calibrated data. The simulated flood hydrographs were well compared with the observed data. SCS unit hydrograph method showed best fit, but there was no significant difference among the three unit hydrograph methods.