STRAT-MSB - Quantifying the Role of Stratospheric Mean State Bias on Representations of Climate

拟议的工作将寻求量化平流层平均状态偏差在驱动平流层和对流层对低平流层温度扰动的敏感性的模型间扩散中的作用。以往的文献着重指出了全球环流模型捕捉对流层上层和平流层下层过程，包括水蒸气向平流层输送的能力方面的主要不足。这对整个大气层的辐射收支有重大影响，并以平流层环流和对流层动力变率为反馈。这个项目将利用一个新的模型内部比较项目的结果，在该项目中，温度趋势已经在五个最先进的气候模型的平流层下部固定。早期结果表明，尽管存在相同的扰动，但每个模型都表现出独特的热响应，这导致了不同的地表气候响应。在这些发现的推动下，该项目将扩大分析，以确定平流层和对流层模式间的差异，包括水蒸气输送和翻转环流的变化。这些差异的来源将使用一种新的理想化平流层模型来探索，以确定响应的多大部分可能归因于模型的平流层平均状态。结果将受到平流层廓线观测的限制。这项工作的结果将通过量化平均状态偏差相对于过程级参数化或数值方法的重要性来指导模型开发。了解模型行为中的不确定性来源，能够更好地解释当前对未来变暖的预测，并对国家和全球范围的适应和缓解规划产生影响。

The proposed work will seek to quantify the role of stratospheric mean state bias in driving inter-model spread in stratospheric andtropospheric sensitivity to temperature perturbations in the lower stratosphere. Previous literature has highlighted majorinadequacies in the ability of global circulation models to capture processes in the upper troposphere and lower stratosphere,including the transport of water vapor into the stratosphere. This has major implications for the radiative budget of the entireatmosphere and feeds back on stratospheric circulation and tropospheric dynamical variability. This project will leverage the outputsof a novel model inter-comparison project in which a temperature tendency has been fixed in the lower stratosphere of five state ofthe art climate models. Early results indicate that in spite of an equivalent perturbation, each model exhibits unique thermalresponses which result in differing surface climate responses. Motivated by these findings, this project will extend analysis tocharacterize both the stratospheric and tropospheric inter-model differences, including changes to water vapor transport and theoverturning circulation. The source of these differences will be explored using a novel idealized stratospheric model to identify whatfraction of the response may be attributed to a model's stratospheric mean state. Results will be constrained by observations of thestratospheric profile. Outcomes of this work will guide model development by quantifying the importance of mean state bias ascompared to process level parameterizations, or numerical methods. Understanding sources of uncertainty in model behaviorenables improved interpretation of current future warming projections with implications for national and global scale adaptation andmitigation planning.