Transforming our understanding of climate shifts in the North African dust belt and upskilling Earth System Models to simulate them

随着地球在未来几十年变暖，全球降雨模式将发生变化，造成不确定的区域结果。这个问题在北非最为严重。世界上最大的炎热沙漠和最大的大气尘埃来源撒哈拉沙漠在20世纪增长了10%。在半干旱的萨赫勒地区和地中海边境地区，年降雨量很少，年际降雨量变化很大，数千万人生活在极端缺水的环境中，有时会造成毁灭性的后果，如20世纪末的萨赫勒干旱。这些都是了解驱使北非气候过去变化的机制的主要动机。然而，仪器记录很短（约150年），地质和考古数据显示北非气候有能力发生更极端的变化。由于地球绕太阳运行的轨道逐渐变化，气候系统在千年尺度上的变化，北非过去的气候在比今天更干燥、更多灰尘的条件和湿润的绿色撒哈拉时期（GSPs）之间反复变化，当时沙漠被河流和湖泊转变为植被繁茂的景观，生活着河马和我们的早期祖先（最近一次是在非洲湿润期，AHP，约14.5至5000年前，ka）。是什么导致了这些过去的转变？它们对未来意味着什么？对于这些问题，我们缺乏令人信服的答案，因为用于预测未来变化的计算机模型无法再现古数据所显示的极端变化。现有古资料的一个主要弱点是它们对机械控制的诊断有限。北非有两种不同纬度和季节的降雨机制，即夏季季风和冬季西风风暴路径，但我们对它们在时间和空间上对数据中记录的湿度-干旱和植被的深刻变化的贡献知之甚少。这些模式的一个主要弱点是，我们不知道它们缺乏技巧是否反映了缺少抑制气候强迫响应的过程和反馈，或者它们是否充分代表了关键过程，但由于不确定的参数值仅仅是根据当代观测目标进行评估而确定的，因此它们无意中与稳定性紧密联系在一起。我们建议改变我们对北非沙尘带气候变化的理解，并利用我们最近取得的突破，提高模型的技能，使其能够成功地模拟气候变化（见跟踪记录）。我们将开发撒哈拉风暴区的代用气候记录，梳理过去夏季季风和冬季风暴路径降雨的变化（见目标）。我们将配置一个ipcc级别的地球系统模型（ESM），该模型显示出撒哈拉沙漠绿化的新能力，并且在计算上足够便宜，可以对参数空间进行足够的采样，以对抗我们的新代理记录（见目标）。这将使我们能够发展对极端变化的机械理解，并将我们的发现转化给那些正在开发下一代esm的科学家，这些esm将用于预测未来几十年的未来变化（见学术受益人）。

As Earth heats up in the coming decades, global rainfall patterns will shift with uncertain regional outcomes. Nowhere is this problem more acute than in North Africa. The world's largest hot desert and biggest source of atmospheric dust, the Sahara, grew by >10% in the 20th century. In the semi-arid Sahel and Mediterranean borderlands, where annual rainfall amounts are low and inter-annual rainfall variability is high, tens of millions of people live in extreme water stress with sometimes devastating consequences as during the late 20th century Sahel drought. These are major incentives to understand the mechanisms driving past variability in the climate of North Africa. However, the instrumental record is short (~150 yrs) and geological and archaeological data reveal a capacity for far more extreme shifts in North African climate. Paced by Earth's gradually changing orbit around the Sun and millennial-scale variability in the climate system, North Africa's past climate has shifted repeatedly between drier and dustier conditions than today and humid green Sahara periods (GSPs) when the desert was transformed into a well vegetated landscape crosscut by rivers and lakes, populated by hippopotamuses and our early ancestors (most recently during the African Humid Period, AHP, ~14.5 to 5 thousand years ago, ka).What caused these past shifts? What do they imply for the future? We lack convincing answers to these questions because the extreme shifts indicated by the palaeo-data are not reproduced by the computer models used to predict future change. A main weakness of the existing palaeo-data is their limited diagnosis of mechanistic control. North Africa has two latitudinal and seasonally distinct rainfall regimes, the summer monsoon and winter westerly storm track, but we know too little of their temporal and spatial contributions to the profound changes in humidity-aridity and vegetation recorded in the data. A main weakness in the models is that we do not know whether their lack of skill reflects missing processes and feedbacks that dampen the response to climate forcing or whether they adequately represent the key processes but are inadvertently hard-wired for stability because uncertain parameter values are determined solely by evaluation against contemporary observational targets. We propose to transform our understanding of climate shifts in the North African dust belt and to upskill models to successfully simulate them by capitalizing on recent breakthroughs that we have made (see Track Record). We will develop proxy climate records for a bellwether Saharan region to tease apart past variability in summer monsoon and winter storm-track rainfall (see Objectives). And we will configure an IPCC-class Earth System model (ESM) that shows novel power to green the Sahara and is sufficiently computationally inexpensive to permit adequate sampling of parameter space to confront our new proxy records (see Objectives). This will allow us to develop a mechanistic understanding of extreme change and to translate our findings to those scientists who are developing the next generation of ESMs that will be used to predict future change over the coming decades (see Academic Beneficiaries).