Influence of altered low cloud parameterizations for seasonal variation of Arctic cloud amount on climate feedbacks
- 주제(키워드) Arctic cloud parameterization , Climate model feedback , Radiative kernels
- 등재 SCIE, SCOPUS
- 발행기관 Springer Verlag
- 발행년도 2015
- 총서유형 Journal
- URI http://www.dcollection.net/handler/ewha/000000121369
- 본문언어 영어
- Published As http://dx.doi.org/10.1007/s00382-015-2926-1
- 저작권 이화여자대학교 논문은 저작권에 의해 보호받습니다.
초록/요약
This study investigates the alteration of climate feedbacks due to overestimated wintertime low-level cloud amount bias over the Arctic region (60°N–90°N) in a climate model. The climate feedback was quantitatively examined through radiative kernels that are pre-calculated radiative responses of climate variables to doubling of carbon dioxide concentration in NCAR Community Atmosphere Model version 3 (CAM3). Climate models have various annual cycle of the Arctic cloud amount at the low-level particularly with large uncertainty in winter and CAM3 may tend to overestimate the Arctic low-level cloud. In this study, the seasonal variation of low-level cloud amount was modified by reducing the wintertime cloud amount by up to 35 %, and then compared with the original without seasonal variation. Thus, we investigate how that bias may affect climate feedbacks and the projections of future Arctic warming. The results show that the decrease in low-level cloud amount slightly affected the radiation budgets because of a small amount of incident solar insolation in winter, but considerably changed water vapor and temperature profiles. Consequently, the most distinctive was decreases in water vapor feedback and contribution of heat transport (by −0.20 and −0.55 W m−2 K−1, respectively) and increases in the lapse rate feedback and cloud feedback (by 0.13 and 0.58 W m−2 K−1, respectively) during winter in this model experiment. This study suggests that the change in Arctic cloud amount effectively reforms the contributions of individual climate feedbacks to Arctic climate system and leads to opposing effects on different feedbacks, which cancel out in the model. © 2015 Springer-Verlag Berlin Heidelberg
more