COntinental COnvective OrganisatioN and rainfall intensification in a warming world: Improving storm predictions from hours to decades (COCOON)

今天，大气和气候科学中一些最紧迫的问题集中在雷暴如何应对大气环境的变化。极端降雨将如何随着气候变化而变化？内部风暴过程和动力学如何影响这些变化？没有什么地方比（亚）热带地区的挑战更紧迫了，那里的大型雷暴集群，所谓的中尺度对流系统（MCS）经常造成恶劣天气和洪水，但由于贫困和经济不景气，人口弹性很低。为了估计和规划未来的风暴影响，我们需要了解和模拟风暴动力学将如何响应（并且已经响应）大气变化，以及是否存在可能加剧极端降雨的内部动力学机制，而不仅仅是与温暖大气中水分增加有关的纯粹物理因素。在受影响最严重的地区，MCSs为农作物、牲畜和人类提供了至关重要的供水，占总降雨量的50-90%，但同样与影响地球仪数百万人的恶劣天气有关。随着气温继续上升，这种情况只会恶化。然而，尽管MCS的社会重要性，我们仍然不知道为什么特别是他们的亚日降雨极端可以经常超过预期的强度。最大降雨量的外部驱动因素（例如大气湿度、风切变、温度）和内部驱动因素（风暴环流、上升气流速度和大小）的相对重要性仍然不清楚，这也妨碍了我们估计全球变暖效应的能力。到目前为止，气候模式对驱动因子贡献的评估还不存在，因为分辨率约为100公里的传统全球气候模式在MCS尺度范围内表现过程时遇到了很大的困难，它们既不能明确地解决也不能令人满意地参数化，即它们没有“看到”MCS。然而，在过去十年中，在使用高分辨率（<10公里）区域对流允许（CP）模式进行气候预测方面取得了迅速进展。由于不需要依赖对流参数化，CP模式甚至比中分辨率模式产生更真实的峰值降雨强度，并且可以模拟真实的MCS。然而，即使是最先进的CP模式仍然在1- 10公里的“灰色地带”运行，那里的内部风暴环流只能部分解决。忽略次网格过程的后果仍在调查中，缺点需要仔细检查。通过将地球观测数据与新兴的最先进的CP气候模式模拟相结合，我的项目研究了对流的规模（连续云盾，嵌入式对流核心尺度，上升气流大小）如何影响MCS降雨极端和陆地上的生命期。基于地球观测数据，我的工作将发现大陆对流组织的规模在过去20-30年内是否发生了变化，以及什么过程是决定这种趋势的关键。这也将探讨MCS与陆地特征和大气环境的相互作用是否会随着对流尺度的变化而变化。我将进一步挑战CP模型与确定的过程，并开发基于过程的模型基准方法，测试如何值得信赖的CP模型在捕捉未来气候的降雨增强机制。研究结果将用于试验方法，以改进风暴临近预报，并根据多条证据改进对未来MCS降雨极端情况的估计，这些证据将至关重要地包括对流尺度。因此，我的项目将使我们对全球变暖如何驱动对流尺度变化，降雨和尺度如何联系，以及尺度信息是否可以改善对天气和气候时间尺度的极端降雨预测的理解发生重大变化。

Some of the most pressing questions in atmospheric and climate science today focus on how thunderstorms will respond to changes in the atmospheric environment. How will extreme rainfall change with climate change? And how do internal storm processes and dynamics affect these changes? Nowhere is the challenge more urgent than in (sub-)tropical regions where large thunderstorm clusters, so-called Mesoscale Convective Systems (MCSs) frequently cause severe weather and flooding, but population resilience is low due to poverty and staggering economies. To estimate and plan for future storm impacts, we need to understand and model how storm dynamics will respond (and are already responding) to atmospheric changes, and whether there are internal, dynamical mechanisms that may intensify rainfall extremes beyond purely thermodynamical considerations linked to increased moisture in a warmer atmosphere. In most affected regions, MCSs provide crucial water supplies for crops, livestock and people, contributing 50-90% to total rainfall but are likewise associated with severe weather that affects millions around the globe. A situation that will only worsen as temperatures continue to rise. And yet, in spite of the societal importance of MCSs, we still do not know why in particular their sub-daily rainfall extremes can frequently surpass expected intensities. The fact that the relative importance of external (e.g. atmospheric humidity, wind shear, temperature) and internal drivers (storm circulations, updraught speeds and size) of rainfall maxima remain unclear also hampers our ability to estimate global warming effects. Climate model assessments of driver contributions so far do not exist as conventional global climate models with coarse resolutions ~100km have major difficulties representing processes in the MCS scale range, which they can neither explicitly resolve nor satisfactorily parametrise, i.e. they do not 'see' MCSs. Over the last decade however, there have been rapid advances in the use of high-resolution (<10 km) regional convection-permitting (CP) models for climate prediction. Not having to rely on convective parametrisations, CP models produce more realistic peak rainfall intensities even compared to medium-resolution models, and can simulate realistic MCSs. However, even state-of-the-art CP models still operate in the "grey-zone" of 1-10km where internal storm circulations are only partly resolved. Consequences of the neglect of sub-grid processes are still under investigation and shortcomings need to be put under scrutiny.By combining earth observation data with emerging state-of-the-art CP climate model simulations, my project investigates how the scale of convection (contiguous cloud shields, embedded convective core scales, updraught size) affects MCS rainfall extremes and lifetimes over land. Based on earth observation data, my work will discover whether scales of continental convective organisation have changed within the last 20-30 years, and what processes are key to determining such trends. This will also explore whether MCS interactions with land features and atmospheric environments change as a function of convective scale. I will furthermore challenge CP models with the identified processes and develop process-based model benchmarking approaches, testing how trustworthy CP models are in capturing rainfall intensification mechanisms in a future climate. The findings will be used to trial methods for improved storm nowcasting and for improved estimates of future MCS rainfall extremes based on multiple lines of evidence that will crucially include convective scales. Thus, my project will bring a step-change in our understanding of how global warming drives convective scale changes, how rainfall and scales are linked, and whether scale information can improve extreme rainfall predictions on weather to climate timescales.