Physically-based resampling scheme for roughness coefficient of surface runoff considering the spatial landuse distribution was suggested for the purpose of effective operational application of recent grid-based distributed rainfall runoff model. Generally grid scale(mother scale) of hydrologic modeling can be greater than the scale (child scale) of original GIS thematic digital map when the objective basin is wide or topographically simple, so the modeler uses large grid scale.
The resampled roughness coefficient was estimated and compared using 3 different schemes of Predominant, Composite and Mosaic approaches and total runoff volume and peak streamflow were computed through distributed rainfall-runoff model. For quantitative assessment of biases between computational simulation and observation, runoff responses for the roughness estimated using the 3 different schemes were evaluated using MAPE(Mean Areal Percentage Error), RMSE(Root-Mean Squared Error), and COE(Coefficient of Efficiency). As a result, in the case of 500m scale Mosaic resampling for the natural and urban basin, the distribution of surface runoff roughness coefficient shows biggest difference from that of original scale but surface runoff simulation shows smallest, especially in peakflow rather than total runoff volume.

Abstract
 1. 서 론
 2. 자료 및 방법
 3. 시험유역 및 적용강우
  3.1. 토지이용특성에 따른 시험유역 선정 및 구분
  3.2. 강우사상
 4. 모격자 구분에 따른 유출량 산정 및 매개변수 보정
  4.1. 모격자의 구분
  4.2. 유출량 및 매개변수 보정
 5. 공간스케일 확장기법
  5.1. 지배면적법(Predominant Approach)
  5.2. 산술평균 방법(Composite Approach)
  5.3. 모자이크 방법(Mosaic Approach)
  5.5. 분석결과
 6. 기법별 수문반응 및 정량적 평가
 7. 결 론
 참 고 문 헌

• 박상식(단국대학교 토목환경공학과) | Sang-Sik Park (Department of Civil & Environmental Engineering, Dankook University)
• 강부식(단국대학교 토목환경공학과) | BooSik Kang (Department of Civil & Environmental Engineering, Dankook University) Corresponding Author