Evaluating the Roles of the Ocean, the Atmosphere, and External Forcing in Atlantic Multi-Decadal Variability

赤道以北大西洋的海面温度（SST）显示出一种变化形式，整个盆地缓慢变暖和变冷，这被称为大西洋多年代变率（AMV）。 AMV 可以用大西洋多年代际振荡 (AMO) 来表征，它是平滑后北大西洋（赤道至北纬 55 度）的平均去趋势海表温度。 AMO大约100年的记录显示了两个振荡周期，1900年代初期气温较低，1920年代至1960年左右变暖，1960年代至千禧年左右再次降温，此后变暖。 这些温度变化对周围陆地地区很重要，因为海温升高与美国东南部和北非降雨量增加、飓风活动增加以及欧洲和北非冬季变暖有关。 当AMO处于凉爽阶段时，通常会出现相反的情况，例如最近的寒冷时期发生的萨赫勒干旱。AMV的驱动机制尚不清楚。 大西洋经向翻转环流 (AMOC) 强度的变化是 AMO 缓慢波动最常被引用的解释，故意降低 AMOC 强度的模拟显示出类似 AMO 的海温响应。 但 PI 和合作者最近的工作表明，在不代表 AMOC 的模型中可以产生类似 AMO 的变异性。 他们研究了气候模型模拟，其中海洋模型被代表上层海洋的非循环“板”所取代，这些模拟产生了与标准气候模型类似的 AMO 海表温度模式和行为。 其他工作表明，AMO 变化可能是由人为气溶胶和火山气溶胶等外部强迫因素产生的。在该项目中，PI 进行了一系列数值模型实验，旨在解决外部强迫以及海洋和大气过程在驱动 AMO 变化中的作用，并进一步致力于了解 AMO/AMV 对周围陆地的影响。 这项工作解决了三个一般性问题：1）海洋对大西洋盆地的数十年变化有多大贡献？造成这种情况的主要海洋过程是什么？2) 大气的贡献有多大，包括外部强制的成分变化以及云辐射反馈？3) 大西洋盆地周围 AMV 产生影响的原因是什么？这些问题可以通过诊断热预算分析和数值实验以及旨在隔离或消除特定过程的模型配置层次结构相结合来解决。 除了上面讨论的标准气候模型和“平板”海洋模型配置之外，PI 还使用海洋模型配置，其中每个海洋柱模拟垂直混合，但仍不模拟海洋环流（请参阅 AGS-1650209）。 另一种配置允许海洋环流（包括 AMOC），但不允许表面风的变化影响环流，从而排除了风和洋流共同作用产生 AMO 的解释。 关于外部强迫作用的研究使用了马克斯·普朗克研究所的气溶胶数据集，其中分析指定的羽流代表来自主要排放中心的气溶胶。鉴于 AMO 变化与周围陆地表面气候（包括影响美国东海岸的飓风活动）之间的联系，这项工作具有社会和科学意义。 这项工作还促进了与免费提供的社区地球系统模型相关的模型层次结构的开发和使用，并且这些模型配置将提供给研究界。 此外，这项工作还为一名研究生提供支持，并包括向 K-12 学生和公众进行推广。

Sea surface temperatures (SSTs) over the Atlantic north of the equator show a form of variability in which the entire basin slowly warms and cools, which is referred to as Atlantic Multidecadal Variability (AMV). AMV can be characterized by the Atlantic Multidecadal Oscillation (AMO), formed as the average detrended SST in the North Atlantic (equator to 55 North) after smoothing. The roughly 100-year record of the AMO shows two cycles of oscillation, with cooler temperatures in the early 1900s, warming from the 1920s to about 1960, cooling again from the 1960s to around the millenium, and warming after that. These temperature variations matter for the surrounding land areas as warmer SSTs are associated with increased rainfall in the southeast US and North Africa, increased hurricane activity, and warmer winters in Europe and North Africa. Generally opposite conditions prevail when the AMO is in its cool phase, for example the Sahel drought occurred during the most recent cold period.The driving mechanisms of AMV are not well understood. Variations in the strength of the Atlantic meridional overturning circulation (AMOC) are the most commonly invoked explanation for the slow fluctuations of the AMO, and simulations in which AMOC strength is deliberately reduced show an AMO-like SST response. But recent work by the PIs and collaborators suggests that AMO-like variability can be produced in models which do not represent the AMOC. They looked at climate model simulations in which the ocean model was replaced by a non-circulating "slab" representing the upper ocean, and these simulations produced the AMO SST pattern and behavior similar to standard climate models. Other work suggests that AMO variability can be generated by external forcing factors such as and anthropogenic aerosols and volcanic aerosols.In this project the PIs perform a series of numerical model experiments designed to address the roles of external forcing and oceanic and atmospheric processes in driving AMO variability, with further effort devoted to understanding AMO/AMV impacts on surrounding landmasses. The work addresses three general questions:1) How much is the ocean contributing to multi-decadal variations in the Atlantic basin? What are the dominant ocean processes responsible for this?2) How much is the atmosphere contributing, including externally forced changes in composition, as well as cloud-radiative feedbacks?3) What are the causes of the impacts associated with the AMV around the Atlantic basin?These questions are addressed using a combination of diagnostic heat budget analysis and numerical experiments with a hierarchy of model configurations designed to isolate or eliminate particular processes. In addition to the standard climate model and "slab" ocean model configurations discussed above, the PIs use an ocean model configuration in which each ocean column simulates vertical mixing but ocean circulation is still not simulated (see AGS-1650209). Another configuration allows ocean circulation (including the AMOC) but does not allow changes in the surface winds to affect the circulation, thereby ruling out explanations in which winds and ocean currents act together to generate the AMO. Research on the role of external forcing uses an aerosol dataset from the Max Planck Institute in which analytically specified plumes represent aerosols from the major emission centers. The work has societal as well as scientific relevance given the connections between AMO variability and surface climate on surrounding landmasses, including hurricane activity that affects the US east coast. The work also promotes the development and use of a hierarchy of models connected to the freely available Community Earth System Model, and these model configurations will be made available to the research community. In addition, the work supports a graduate student and includes outreach to K-12 students and the general public.