水汽和云辐射反馈对海表强迫的响应及其对全球平均表面增暖的影响

How many degrees the global-mean surface temperature would warm in the 21st century is a substantial problem in climate change with essential public attention. As an important greenhouse gas, increase of water vapor would significantly enhance the extent of global warming, while the complexity of cloud radiative effects are the most important source of the inter-model variability for surface warming prediction. Previous studies focus on the global-mean atmospheric changes and cloud physical processes, which were unable to fundamentally constrain the uncertainty in cloud radiative feedback. For this, we propose to use a novel approach for a breakthrough in the contribution of ocean-atmosphere dynamics to climate feedbacks. Regional change of sea surface temperature (SST) in climate projections has been demonstrated to control precipitation change patterns through atmospheric circulation adjustments, with significant uncertainty. We plan to diagnose its influences on the change patterns of water vapor and clouds from climate models and access their contributions to the diversities in climate feedbacks and sea surface warming. Eventually, we try to constrain the uncertainty in global surface warming prediction by using observations of interannual variability, based on exploring the relationship between global-mean and regional changes of SST.

21世纪全球平均表面温度确切升高几度，是气候变化中极其重要也备受公众瞩目的问题。随着温室气体浓度的升高，水汽的增加会显著加剧全球变暖程度，而云辐射效应的复杂性引起表面升温预测在气候模式间的差异。以往国内外研究围绕全球平均大气变化和云物理过程展开，未能从根本上约束云辐射反馈的不确定性。针对这个问题，本项目试图采用新的思路在海－气动力学对气候反馈贡献方面有所突破。已经证实气候预估中海表温度的区域变化通过大气环流来调整降水分布，并蕴含着丰富的不确定性，拟基于气候模式诊断其对水汽和云空间分布变化的影响，进一步评估它们对全球平均气候反馈和海表升温预测多样性的贡献。最终通过探讨海表温度全球平均与区域变化的相互关系，利用年际变化的观测约束全球表面升温预测中的不确定性。

气候反馈通常使用与全球平均表面温度升高相关的辐射效应进行评估；但是，这不足以理解反馈中的不确定性，会直接影响21世纪全球气候敏感度范围，从而很难确定可信的增温程度。分析表明如同降水和大气环流，模式表面温度的变化形态是气候模式中水汽、气温直减率和云辐射反馈不确定性的关键驱动力，通过热带和中高纬地区截然不同的影响方式，这种动力控制表现在模式的大气响应及其辐射效应。研究发现了热带的百分比变化效应，并提出“动力云反馈——暖者更高”范式（只有极地“暖者更低”），强调了区域气候变化在判定全球辐射反馈中不确定性中的重要性。最终确定了最敏感和有效的约束区域，指出如果针对表面增暖的空间差异，还有可能是其因子的气候态低云，尤其是在各大洋赤道和副热带的东洋盆，再加上北大西洋至巴伦支海周边区域，利用观测约束提高数值模拟和预估的精确性，则气候敏感度预估的不确定性可能会大大降低。