대기 (Atmosphere)
- 제18권1호
- /
- Pages.71-81
- /
- 2008
- /
- 1598-3560(pISSN)
- /
- 2288-3266(eISSN)
GIS와 PRISM을 이용한 고해상도 격자형 강수량 추정
Estimation of High Resolution Gridded Precipitation Using GIS and PRISM
- 신성철 (공주대학교 대기과학과) ;
- 김맹기 (공주대학교 대기과학과) ;
- 서명석 (공주대학교 대기과학과) ;
- 나득균 (기상청 기후 정책과) ;
- 장동호 (공주대학교 응용수학과) ;
- 김찬수 (공주대학교 지리학과) ;
- 이우섭 (공주대학교 대기과학과) ;
- 김연희 (공주대학교 대기과학과)
- Shin, Sung-Chul (Department of Atmospheric Science, Kongju National University) ;
- Kim, Maeng-Ki (Department of Atmospheric Science, Kongju National University) ;
- Suh, Myoung-Suk (Department of Atmospheric Science, Kongju National University) ;
- Rha, Deuk-Kyun (Climate Policy Division, KMA) ;
- Jang, Dong-Ho (Department of Geography, Kongju National University) ;
- Kim, Chan-Su (Department of Applied Mathematics, Kongju National University) ;
- Lee, Woo-Seop (Department of Atmospheric Science, Kongju National University) ;
- Kim, Yeon-Hee (Department of Atmospheric Science, Kongju National University)
- 투고 : 2008.01.28
- 심사 : 2008.03.25
- 발행 : 2008.03.01
초록
In this study, in order to estimate high resolution precipitation with monthly time scales, Parameter-elevation Regressions on Independent Slopes Model (PRISM) was modified and configured for Korean precipitation based on elevation, distance, topographic facet, and coastal proximity. Applying this statistical downscaling model to Korean precipitation for 5 years from 2001 to 2005, we have compiled monthly grid data with a horizontal resolution of 5-km and evaluated the model using bias, root mean square error (RMSE), and correlation coefficient between the observed and the estimated. Results show that bias, RMSE, and correlation coefficient of the estimated value have a range from 0.2% to 1.0%, 19.6% (June) to 43.9% (January), and 0.73 to 0.84, respectively, indicating that the modified Korean PRISM (K-PRISM) is reasonably worked by weighting factors, i.e., topographic effect and rain shadow effect.