With the launch of the Observing Carbon Observatories (OCO-2 and OCO-3), we have entered a new era of greenhouse gas (GHG) data collection where sub-city scale data can be collected at varying times across the entire globe. An increasing focus on quantifying urban emissions of GHGs from policy makers has begun to spur new research into how best to use this unique and very high resolution data set. While, historically, this line of research has been the domain of limited domain mesoscale and regional models, an increasing understanding and respect for boundary inflow uncertainty and bias has led people to search for an alternative modeling framework that both characterizes high frequency variability of CO₂ in space and time as well as honor mass conservation requirements globally and be seamless with respect to boundaries. To monitor local anthropogenic emissions from space, the influence of atmospheric signals originating from outside the local area of interest needs to be quantified. In our first step towards building a comprehensive multi-scale CO₂ inversion system, we use free running simulations of the Ocean Land Atmosphere Model (OLAM), a variable-resolution general circulation model, to explore the signal-to-noise statistics of anthropogenic urban emissions of CO₂ versus the background inflow for approximately 40 of the largest cities across the globe. We show that signal-to-noise levels are much better in winter time than summer but also that the winter biological inflow is far from negligible, suggesting that the commonly held assumption that biology can be ignored in winter time urban emission estimates is probably incorrect. Simulated pressure-weighted column average CO₂ (XCO₂) is also used to evaluate the ability of fixed location XCO₂ measurements to provide background inflow estimates. Results show why Los Angeles, a heavily instrumented and studied urban center, is likely one of the easiest cities to observe globally from space, despite its relatively complex meteorology. Lastly, we discuss challenges and possible research paths forward to continue to advance the notion of multi-scale global CO₂ flux inversion systems capable of simultaneously optimizing local urban emissions and large-scale CO₂ transport patterns.

随着碳观测站（OCO-2 和 OCO-3）的启动，我们进入了温室气体 (GHG) 数据收集的新时代，可以在全球不同时间收集次城市规模的数据。政策制定者越来越关注量化城市温室气体排放量，这已经开始激发新的研究，探讨如何最好地使用这种独特且分辨率非常高的数据集。虽然从历史上看，这一研究领域一直是有限域中尺度和区域模型的领域，但对边界流入不确定性和偏差的日益理解和尊重促使人们寻找替代建模框架，既能表征 CO _{2 的}高频变异性在空间和时间上，并在全球范围内遵守大规模保护要求，并且在边界方面是无缝的。为了监测来自空间的局部人为排放，需要量化来自局部感兴趣区域之外的大气信号的影响。在我们建立全面的多尺度 CO ₂反演系统的第一步中，我们使用海洋陆地大气模型 (OLAM) 的自由运行模拟，这是一种可变分辨率的环流模型，探索人为的信噪统计。 CO _{2 的}城市排放与全球约40个最大城市的背景流量相比。我们表明，冬季的信噪比水平比夏季要好得多，而且冬季的生物流入远不能忽略不计，这表明在冬季城市排放估算中可以忽略生物学的普遍假设可能是不正确的。模拟压力加权柱平均 CO ₂ (XCO ₂ ) 也用于评估固定位置 XCO ₂的能力测量以提供背景流入估计。结果表明，尽管气象学相对复杂，但为什么洛杉矶是一个经过大量仪器检测和研究的城市中心，但它可能是从太空进行全球观测的最容易的城市之一。最后，我们讨论了挑战和可能的研究路径，以继续推进能够同时优化当地城市排放和大规模 CO ₂运输模式的多尺度全球 CO ₂通量反转系统的概念。