The Role of Atmosphere Feedbacks during ENSO in the CMIP3 Models. Part III: The Shortwave Flux Feedback
The Role of Atmosphere Feedbacks during ENSO in the CMIP3 Models. Part III: The Shortwave Flux Feedback
复制标题
DOI：
10.1175/jcli-d-11-00178.1
复制
发表时间：
2012-06-15
影响因子：
4.9
通讯作者：
Weller, Hilary
中科院分区：
文献类型：
--
作者：
Lloyd, James;Guilyardi, Eric;Weller, Hilary
先前使用耦合大气环流模型（GCM）的研究表明，大气模型在厄尔尼诺-南方涛动（ENSO）模型中起主导作用，而海面温度（SST）热力学阻尼的模型间差异是ENSO振幅多样性的主要贡献者。本研究对 12 个耦合模型比对项目第 3 阶段 (CMIP3) 模拟中的短波通量反馈 (alpha(SW)) 进行了详细分析，其动机是发现 alpha(SW) 是模型热力学阻尼误差的主要贡献者。为阐明 asw 偏差而开发的“反馈分解方法”表明，所有模型都低估了赤道东部太平洋海表温度的动态大气响应，导致低估了阿尔法（SW）值。云对动力学的响应偏差以及云对短波的拦截也会导致反潜战的误差。耦合模拟和相应的纯大气模拟之间的 alpha(SW) 反馈变化与平均动力学变化有关。在观测和建模的 SW 通量反馈中发现了很大的非线性，在线性计算 asw 时隐藏了这一点。在观测中，提出了两种物理机制来解释这种非线性：1）对冷海表温度异常的下沉响应比对暖海表温度异常的上升响应更弱；2）对海表温度的非线性高层云量响应。模型往往会低估短波通量反馈非线性，这与海表温度动态响应中的非线性被低估有关。本研究中提出的基于过程的方法可能有助于纠正 ENSO 模型大气偏差，最终改进 GCM 中 ENSO 的模拟。
Previous studies using coupled general circulation models (GCMs) suggest that the atmosphere model plays a dominant role in the modeled El Nino-Southern Oscillation (ENSO), and that intermodel differences in the thermodynamical damping of sea surface temperatures (SSTs) are a dominant contributor to the ENSO amplitude diversity. This study presents a detailed analysis of the shortwave flux feedback (alpha(SW)) in 12 Coupled Model Intercomparison Project phase 3 (CMIP3) simulations, motivated by findings that alpha(SW) is the primary contributor to model thermodynamical damping errors.A "feedback decomposition method," developed to elucidate the asw biases, shows that all models underestimate the dynamical atmospheric response to SSTs in the eastern equatorial Pacific, leading to underestimated alpha(SW) values. Biases in the cloud response to dynamics and the shortwave interception by clouds also contribute to errors in asw. Changes in the alpha(SW) feedback between the coupled and corresponding atmosphere-only simulations are related to changes in the mean dynamics.A large nonlinearity is found in the observed and modeled SW flux feedback, hidden when linearly calculating asw. In the observations, two physical mechanisms are proposed to explain this nonlinearity: 1) a weaker subsidence response to cold SST anomalies than the ascent response to warm SST anomalies and 2) a nonlinear high-level cloud cover response to SST. The shortwave flux feedback nonlinearity tends to be underestimated by the models, linked to an underestimated nonlinearity in the dynamical response to SST. The process-based methodology presented in this study may help to correct model ENSO atmospheric biases, ultimately leading to an improved simulation of ENSO in GCMs.