Collaborative Research: North American Warm-season Extremes in a Changing Climate: Large-scale Drivers and Local Feedbacks

在美国大部分地区，气候变化主要是通过其对暖季极端事件（如洪水、热浪、火灾和干旱）的影响来感受到的。 基本热力学表明，这些事件的严重性和频率应该增加，例如，在气候变暖的情况下，最热的热浪可能会变得更热，而风暴强度可能会增加，因为温暖的空气含有更多的水分。 热力学论证有所帮助，但影响极端事件的全套过程是广泛的，涉及广泛的空间尺度，从数公里规模的雷暴到驱动天气系统的半球规模的急流。 广泛的尺度范围使使用天气和气候模型研究极端事件变化的工作变得复杂，因为即使在最大的计算机上，在所有相关空间尺度上模拟所有相关过程（发生在全球变暖的数十年过程中）也是不切实际的。 气候模型可以模拟数十年甚至数百年的整个全球气候系统，但分辨率太粗（网格间距可能为 100 公里），无法代表强烈风暴的规模。 特别是，它们没有捕获中尺度对流系统（MCS），而中尺度对流系统是美国大陆上大部分恶劣天气的原因。 另一种称为伪全球变暖（PGW）的方法使用高分辨率模型来模拟观察到的极端事件，并通过修改环境条件来重复模拟以代表变暖的气候。 PGW模拟非常有价值，但它们只能考虑气候变化如何影响特定事件的严重程度，因此无法研究极端事件发生频率的变化。 此外，PGW 模拟通常使用区域模型进行，因此不能正确代表半球尺度大气环流变化的影响。该项目开发了一种研究气候变暖中极端事件变化的方法，该方法解决了多尺度问题，并能够检查极端事件频率和其他汇总统计数据。 首先，高分辨率全球模型，跨尺度预测模型（MPAS）用于模拟过去30年（1990年至2019年）的天气和气候。 该模型的网格间距为 15 公里，能够代表 MCS。 其次，在这次“自然运行”中识别极端事件，并通过修改海面温度和其他表面条件来重新模拟，以代表未来的变暖。 这些修改是使用耦合模型比对项目 (CMIP) 的气候模型模拟生成的。 重新模拟是仅具有全局域的 PGW 的一种形式，因此可以在整个空间尺度范围内检查强度的变化。 第三，使用 CMIP 模型输出生成一组 30 个暖季（5 月至 11 月）MPAS 模拟来代表未来的气候变化。 暖季模拟遵循 PGW 方法，但整个季节持续时间意味着模拟不会遵循特定事件，而是显示极端事件的典型季节如何因温暖条件而变化。 这些模拟需要解决的一个问题是北美上空急流的变化对洪水和热浪的影响，因为气候模型通常显示美国大陆上空的急流风速降低，而向北和向南的速度增加。鉴于极端事件的破坏性影响以及关于极端事件变化的更好信息对指导决策​​的价值，这项工作具有社会和科学意义。 该项目还为五名研究生和一名本科生研究助理提供支持和培训。 该项目中生成的模拟可供研究社区使用，并且输出的简化版本托管在 JupyterHub 上，以便通过 Jupyter Notebooks 为参与该项目的大学研究人员提供访问权限。 外展活动是通过北卡罗来纳州自然科学博物馆 (NCMNS) 的初级馆长计划进行的，该计划针对对野外生物学和保护感兴趣的高中生。 学生们收集当地天气事件及其影响的实地测量结果，包括昆虫爆发、霉菌、洪水和大雨的其他后果。 活动指南是根据这些活动创建的，并通过美国地球科学教师协会传播。该奖项反映了 NSF 的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

Throughout much of the US climate change will be felt largely through its effects on warm season extreme events like flooding rains, heat waves, fires, and droughts. Basic thermodynamics suggests that the severity and frequency of these events should increase, for instance the hottest heat waves are likely to get hotter in a warming climate and storm intensity is likely to increase because warmer air holds more moisture. The thermodynamic arguments help but the full suite of processes that affect extreme events is extensive and involves a broad range of spatial scales, from the multi-kilometer scale of thunderstorms to the hemispheric scale of the jet streams that drive weather systems. The broad scale range complicates efforts to study extreme event change using weather and climate models, as a brute force effort to simulate all the relevant processes at all the relevant spatial scales, occurring over the decades-long progression of global warming, is not practical even on the largest computers. Climate models can simulate the full global climate system for decades and even centuries but at resolutions too coarse (perhaps 100km grid spacing) to represent the scales of intense storms. In particular they do not capture the mesoscale convective systems (MCSs) which account for much of the severe weather over the continental US. An alternative approach called pseudo-global warming (PGW) uses a high-resolution model to simulate an observed extreme event, and the simulation is repeated with modifications to the ambient conditions to represent the warmer climate. PGW simulations are quite valuable but they only allow consideration of how climate change affects the severity of specific events, thus they do not enable research on changes in the frequency of occurrence of extreme events. Also, PGW simulations are typically conducted using regional models and thus do not properly represent the effects of changes in the hemispheric-scale atmospheric circulation.This project develops a methodology for looking at extreme event change in a warming climate which addresses the multi-scale issue and enables examination of extreme event frequency and other aggregate statistics. First, a high-resolution global model, the Model for Prediction Across Scales (MPAS) is used to simulate the weather and climate of the past 30 years (1990 to 2019). With a grid spacing of 15km the model is capable of representing MCSs. Second, extreme events are identified in this "nature run" and resimulated with modifications to sea surface temperatures and other surface conditions to represent future warming. The modifications are generated using climate model simulations from the Coupled Model Intercomparison Project (CMIP). The resimulations are a form of PGW only with a global domain, so that changes in intensity can be examined accounting for the full range of spatial scales. Third, a set of 30 warm season (May to November) MPAS simulations using CMIP model output is generated to represent future climate change. The warm season simulations follow the PGW approach but the full season duration means that the simulations do not follow particular events but instead show how a typical season of extreme events changes due to warmer conditions. One issue to be addressed with these simulations is the effect of changes in the jet streams over North America on floods and heat waves, as climate models typically show a reduction in jet-level wind speed over the continental US with increases in speed to the north and south.The work is of societal as well as scientific interest given the damaging effects of extreme events and the value of better information on extreme event change to guide decision making. The project also provides support and training to five graduate students and an undergraduate research assistant. Simulations generated in the project are made available to the research community, and reduced versions of the output are hosted on a JupyterHub to provide access to researchers at the universities participating in the project through Jupyter Notebooks. Outreach is conducted through the Junior Curator program North Carolina Museum of Natural Sciences (NCMNS), a program for high school students interested in field biology and conservation. The students collect field mesaurements of local weather events and their impacts, including insect outbreaks, mold, flooding, and other after-effects of heavy rain. Activity guides are created based on these activities and disseminated through the National Association of Geoscience Teachers.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.