메뉴 건너뛰기
.. 내서재 .. 알림
소속 기관/학교 인증
인증하면 논문, 학술자료 등을  무료로 열람할 수 있어요.
한국대학교, 누리자동차, 시립도서관 등 나의 기관을 확인해보세요
(국내 대학 90% 이상 구독 중)
로그인 회원가입 고객센터 ENG
주제분류

추천
검색
질문

논문 기본 정보

자료유형
학술저널
저자정보
저널정보
한국기상학회 Asia-Pacific Journal of Atmospheric Sciences 한국기상학회지 제41권 제1호
발행연도
2005.2
수록면
89 - 99 (11page)

이용수

표지
📌
연구주제
📖
연구배경
🔬
연구방법
🏆
연구결과
AI에게 요청하기
추천
검색
질문

이 논문의 연구 히스토리 (2)

초록· 키워드

오류제보하기
A neural network model as a tool to downscaling has been adopted to derive winter river-basin precipitation from coarse GCM results. A neural network can be an effective alternative to traditionally statistical downscaling techniques, since they can be handle unknown nonlinear behavior existing in variable such as precipitation. The model has the multi-layer perceptron structure, consisting of the input layer, hidden layer, and output layer, and employs back-propagation algorithm to adjust the connection weight in a way that the error between desired output and actual output is minimized. We used the CAM2 (Community Atmosphere Model Ver.2) 10 members winter prediction results for the period of 1979.12-2002.2 as the input dataset. We designed six-times experiment, of which the total period was split for training of 19-year and validation of 4-year. For the input data to the neural network model, we intended to select the principal components based on Empirical Orthogonal Function (EOF) analysis which is effective method not only to identify the main large-scale atmospheric circulation pattern but also to capture most of the variance with orthogonality. In assessing the performance of neural network, we compared the network results with the GCM results as well as observation to be converted into the river-basin precipitation. The results from neural network successfully reduced the error, although the error from neural network is depended on GCM results. In case of training period, network error is almost less than half compared to GCM results. On the other hand, validation results show a varying performance for individual precipitation episode. Based on cross-validation results, the neural network approach has strong possibility for statistical downscaling of the river-basin precipitation from GCM simulation results.

목차

Abstract
1. 서론
2. 자료 및 분석 방법
3. 신경망 모델의 구조 및 학습 알고리즘
4. 유역평균 강수량 산출을 위한 신경망 모델 구축
5. 결과 분석
6. 결론 및 토의
감사의 글
참고문헌

참고문헌 (28)

참고문헌 신청

함께 읽어보면 좋을 논문

논문 유사도에 따라 DBpia 가 추천하는 논문입니다. 함께 보면 좋을 연관 논문을 확인해보세요!

이 논문의 저자 정보

이 논문과 함께 이용한 논문

최근 본 자료

전체보기

댓글(0)

0

UCI(KEPA) : I410-ECN-0101-2009-453-016237525