Collaborative Research: Subgrid-scale Models for Large-eddy Simulation of Cloud Formation and Evolution

云对地球的能量平衡很重要，是气候和天气的调节者。对气候和天气预报的云形成、云量、降水等的估计都是通过数值模拟来完成的。尽管在计算方面取得了进步，而且在可预见的未来，用于天气和气候预测的真实大气模拟的特点将是，其中很大一部分能量、热量和蒸气通量等无法得到解决，必须建立模型。该项目的目标是在加州大学伯克利分校、斯坦福大学和NCAR科学家的合作下，为湿大气大涡模拟（LES）模拟代码中的子网格和子滤波尺度开发、验证并向社区提供改进的、物理上更真实的湍流模型。将开发两个新的亚网格尺度（SGS）模型集（一个使用动态方法，另一个使用线性代数模型），用于云模拟的所有元素，即动量，热量，水的液体和蒸汽相，霰，微物理中的预测方程等。这些SGS闭包改进了以前边界层模拟中的平均场和高阶统计量，但尚未应用于云。SGS模型将在明确的滤波和重建框架内构建，这减少了数值误差，并提供了湍流应力的更物理表示。智力优势：该项目旨在为未解决的领域（包括动量、热量、水蒸气、液体和其他标量通量）提供显著改进的模型，以对正在建模的物理过程产生更深入的理解，并在实际云的测试用例模拟中验证这些模型。另一个目标是澄清在大气模拟的未知领域/灰色地带使用这些子程序的有效性，即网格分辨率区域，其中对流热气流等流动特征部分得到解决，部分得到子网格解决。随着数值模拟覆盖越来越多的长度尺度，这个区域正成为越来越大的挑战。SGS闭包在未知领域的作用仍然很大程度上未被探索，特别是在云的情况下。在前人SGS模型研究的基础上，本研究将(1)建立新的湿大气SGS方程集，(2)将其应用于先验试验，然后(3)对晴空对流边界层、有无降水的信风积云、浅积云和深对流等场尺度情况进行模拟。这些模拟将用于评估方程集的准确性和效率，并评估模型在TI（或灰色地带）中的性能。更广泛的影响：有两个领域的更广泛的影响。首先，这项工作的成功完成将在模拟中改进云的产生和演化预测。因为工作所基于的代码在国际上被广泛使用，这将是社区的主要利益。鉴于云的形成和行为的准确预测是天气和气候预测的关键因素，特别是降雨，这项工作有可能对整个天气领域产生重大影响。以往的经验表明，新的SGS模型也可以很容易地推广到其他规范，这将进一步扩大这项工作的影响。其次，该项目旨在通过与NCAR的合作，为博士后研究人员提供广泛和高质量的培训，他们将受益于伯克利和斯坦福的建模专业知识以及NCAR的建模和微物理专业知识。此外，斯坦福大学的一名本科生将参与该项目，以补充博士后研究员的工作。

Clouds are important to the earth's energy balance and a regulator of both climate and weather. The estimation of cloud formation, cloud cover, precipitation, etc., for both climate and weather prediction is accomplished with numerical simulations. Despite computational advances and for the foreseeable future, simulations of the real atmosphere for weather and climate prediction will feature results in which a significant fraction of the energy, heat and vapor fluxes, etc., will not be resolved and must be modeled. This project's goal is to develop, validate and make available to the community improved and more physically realistic turbulence models for the subgrid- and subfilter-scales in moist-atmosphere large-eddy simulation (LES) simulation codes, in collaboration between UC Berkeley, Stanford University, and NCAR scientists. Two new subgrid-scale (SGS) model sets (one using dynamic methods and one using a linear, algebraic model) will be developed for all elements of a cloud simulation, i.e., momentum, heat, water's liquid and vapor phases, graupel, the prognostic equations in the microphysics, etc. These SGS closures have improved mean fields and higher-order statistics in previous boundary layer simulations, but have yet to be applied to clouds. The SGS models will be constructed within the explicit filtering and reconstruction framework, which reduces numerical errors and provides a more physical representation of turbulent stresses. Intellectual Merit: The project aims to provide significantly improved models for the unresolved fields (including momentum, heat, water vapor, liquid, and other scalar fluxes), to yield deeper understanding of the physical processes being modeled, and to validate those models in test case simulations of realistic clouds. The other goal is to clarify the validity of using such subroutines in the Terra Incognita [TI] / Gray Zone of atmospheric simulations, i.e., the zone of grid resolution in which flow features such as convective thermals are partly resolved and partly sub-grid. This zone is becoming an ever greater challenge as numerical simulations cover more and more length scales. The role of SGS closures in the Terra Incognita is still largely unexplored, particularly in the case of clouds. Building upon prior research on SGS modeling, the research will (1) create new SGS equation sets for the moist atmosphere, (2) apply them in a priori tests and then (3) carry out simulations of field-scale situations covering clear convective boundary layers, trade-wind cumulus with and without precipitation, shallow cumulus, and deep convection. These simulations will be set up to assess the performance of the equation sets for their accuracy and efficiency and to assess model performance in the TI (or Gray Zone). Broader Impacts: There are two domains of broader impacts. First, successful completion of this work will yield improved predictions of cloud generation and evolution in the simulations. Because the code on which the work is based is widely used internationally, this will be a major benefit to the community. Given that accurate prediction of cloud formation and behavior is a critical element in weather and climate prediction and, in particular, rainfall, the work has the potential for significant impact across the weather domain. Previous experience suggests that the new SGS models will be easily transported to other codes as well, which will further broaden the impact of this work. Second, this project aims through its collaboration with NCAR to give broad and high quality training to a postdoctoral researcher, who will benefit from the exposure to the modeling expertise at Berkeley and Stanford and the modeling and microphysics expertise at NCAR. In addition, a Stanford undergraduate student will work on the project to complement the work of the postdoctoral researcher.