CAREER: A Universal Framework for Large-Eddy Simulation of Atmospheric Boundary Layer Flow Over Complex Terrain

大气边界层（ABL）通过天气和空气质量直接影响人类生活。关于这一薄的大气层还有许多有待了解的地方，部分原因是它受到地形和陆地表面特征变化的强烈影响。这些变化可能导致热强迫流动，如在平静条件下的山谷风和在地形阻塞的强风下的大气旋涡和背风波。了解这种流动特征的物理动力学需要详细的现场观察和高分辨率的数值模拟。大涡模拟（LES）是模拟ABL流动最有前途的数值技术之一，因为它允许通过空间滤波来控制湍流长度尺度，从而将大的、可分解的运动从子滤波尺度的湍流运动中分离出来。虽然主要用于理想流动条件的模拟，但LES有可能普遍适用于复杂地形的ABL研究。智力优势：本研究将探讨关键的创新步骤，以扩展LES的可扩展性，使其可以从区域到非常精细的尺度。这将允许LES有效地应用于复杂地形上的流动，并对边界层流动过程有更深入的了解。控制湍流长度尺度、改善侧向边界强迫、生成集合模拟以及减少由于亚网格表面粗糙度参数化和陡峭地形表示而导致的数值误差的新策略将为复杂地形上的一般LES应用创建一个通用框架。该框架将与观测数据分析一起用于研究地形诱导的流动特征，如山波、旋翼和山谷风环流。PI在研究陡峭山区的边界层流动以及开发LES的数值方法和湍流闭合方案方面有丰富的经验。本文提出了一项综合教育和推广计划，通过一个互动模拟网站激发K-12学生和公众对大气边界层现象的兴趣，并为本科生和研究生提供数值模拟经验。学生们将参加加州大学伯克利分校劳伦斯科学馆（LHS）的公共推广项目，通过暑期助教奖学金和一门新的大涡模拟研究生课程，该课程将直接与在线展览相结合。本课程还将通过提供数值方法的实践训练来加速研究生的研究进展。更广泛的影响：拟议的LES通用框架将通过从中尺度到城市尺度的嵌套模拟，使所有尺度的流量特征无缝集成成为可能。通过这个新的数值框架获得的关于复杂流动的新知识将极大地改进用于预测天气、空气污染、污染物扩散和区域气候的模式。此外，对可能发生在山区的大气环流和强烈重力波的了解将对航空安全产生重要影响。在先进的研究数值模型中实现的新计算方法将通过持续的公开发布的代码更新达到广泛的用户社区。通过这项研究将培训两名研究生，并通过加州大学伯克利分校的SUPERB项目为代表性不足的少数族裔工程专业学生提供两名本科研究助理奖学金。研究成果将通过期刊出版物和会议报告传播。拟议的教育工具的可用性将通过劳伦斯科学厅（LHS）的主要网站、PI的实验室和部门网站以及美国气象学会公报上的一篇文章进行宣传。交互式建模网站有可能触及成千上万的个人，并可以在所有年龄段的课堂设置中使用，向学生介绍天气过程和数值建模。数百名额外的游客，主要是K-12学生，将参加由PI和学生提供的LHS博物馆地板演示。

The atmospheric boundary layer (ABL) directly influences human life through weather and air quality. Much remains to be understood about this thin atmospheric layer, partly because it is strongly affected by variations in topography and land-surface characteristics. These variations can lead to thermally-forced flow such as valley winds under calm conditions and atmospheric rotors and lee waves under strong winds with topographic blocking. Understanding the physical dynamics of such flow features requires detailed field observations and high-resolution numerical simulations. Large-eddy simulation (LES) is one of the most promising numerical techniques for modeling ABL flow because it allows control of turbulent length scales through spatial filtering to separate large, resolved motions from subfilter-scale, turbulent motions. Although primarily used for simulations of idealized flow conditions, LES has the potential to become universally applicable to ABL studies over complex terrain. Intellectual merit: This research will investigate key, innovative steps to extend the scalability of LES so it can be used from regional to very fine scales. This will allow LES to be effectively applied to flow over complex terrain and lead to greater insight into boundary layer flow processes. New strategies to control turbulent length scales, improve lateral boundary forcing, generate ensemble simulations, and reduce numerical errors due to subgrid surface roughness parameterizations and representation of steep topography will create a universal framework for general LES applications over complex terrain. This framework will be used together with observational data analysis to study terrain-induced flow features such as mountain waves and rotors and valley wind circulations. The PI has extensive prior experience with boundary layer flows over steep, mountainous terrain as well as with developing numerical methods and turbulence closure schemes for LES. An integrated education and outreach program is proposed to excite K-12 students and the general public about atmospheric boundary layer phenomena through an interactive modeling website and to provide numerical modeling experience to both undergraduate and graduate students. Students will participate in this public outreach program with UC Berkeley's Lawrence Hall of Science (LHS) through summer assistantships and a new graduate course on large-eddy simulation that will be directly coupled to the online exhibit. This course will also accelerate graduate research progress by providing hands-on training in numerical methods. Broader impacts: The proposed universal framework for LES will make possible seamless integration of flow features at all scales through nested simulations from the mesoscale to the urban scale. New knowledge of complex flows gained with this new numerical framework will dramatically improve models used to predict weather, air pollution, contaminant dispersion, and regional climate. Furthermore, insights into atmospheric circulations and intense gravity waves that can occur in mountainous areas will have important implications for aviation safety. New computational methods implemented in an advanced research numerical model will reach a broad community of users through continued public releases of code updates. Two graduate students will be trained through this research and two undergraduate research assistantships will be provided for under-represented minority engineering students through the SUPERB program at UC Berkeley. Research results will be disseminated through journal publications and conference presentations. The availability of the proposed educational tools will be advertised through the main Lawrence Hall of Science (LHS) website, through the PI's lab and department websites, and through an article in the Bulletin of the American Meteorological Society. The interactive modeling website has the potential to reach thousands of individuals and can be used in classroom settings at all age levels to introduce students to weather processes and numerical modeling. Hundreds of additional visitors, primarily K-12 students, will attend LHS museum floor demonstrations given by the PI and students.