Collaborative Research: Enabling Cloud-Permitting and Coupled Climate Modeling via Nonhydrostatic Extensions of the CESM Spectral Element Dynamical Core

地球系统模型（esm）的计算量非常大，其组件横跨大气、海洋、陆地、生物圈和冰。这些模式通常以过于粗糙的分辨率运行，无法明确表示雷暴复合体、热带气旋和地形强迫降水等天气系统。然而，现在的计算能力足以使用几公里的网格间隔在全球范围内对这些系统进行明确的模拟，尽管模拟的时间是几周到几个月，而不是在较低分辨率下可能需要几十年到几百年。在这个项目中，研究小组将加强由美国国家科学基金会国家大气研究中心（NCAR）领导的社区地球系统模型（CESM），使其具备作为全球风暴解决模型（GSRM）运行的能力。GSRM版本的CESM作为研究天气-气候界面的工具具有巨大的潜力，特别是因为CESM拥有一个全球开发人员和用户社区，并且是美国学术研究界使用的主要模型。作为该项目的一部分，多名学生和早期职业科学家将在大气科学和高性能计算的交叉点接受培训，增强科学进步所急需的劳动力。该项目的主要目标是为CESM的模拟奠定基础，CESM由包括社区大气模型在内的各个模型组成，目前正在开发第七版（CAM7），将于2024年发布。CAM7的动力核心采用谱元（SE）方法求解流体静力近似下的流体运动方程，其中竖直动量方程中忽略加速度项。该项目将把为能源部（DOE）能源百亿亿次地球系统模型（E3SM）开发的非流体静力SE动力核心移植到CESM/CAM7中。SE动态内核的第二个更具实验性的版本也将被移植，旨在运行在高级计算架构上。然后，研究人员将运行CAM7来测试重现较小规模现象的能力。具体来说，该项目将：•将DOE的非流体静力、基于fortran的SE动力核心移植到CESM框架中。•通过nsf资助的EarthWorks项目提供的支持gpu的物理包，将实验性的c++版本SE内核集成到CESM中。•以中尺度对流系统和热带气旋为例，测试和演示新的CESM配置的能力。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Earth System Models (ESMs) are very computationally intensive to run, with components spanning the atmosphere, ocean, land, biosphere, and ice. These models are typically run at resolutions which are too coarse to explicitly represent weather systems like thunderstorm complexes, tropical cyclones, and terrain-forced precipitation. However, computing power is now sufficient to allow explicit simulation of these systems on a global domain using grid spacings of a few kilometers, albeit for simulations of weeks to months rather than the decades to centuries possible at lower resolutions. In this project, the research team will enhance the Community Earth System Model (CESM), led by the NSF National Center for Atmospheric Research (NCAR), with the capability to run as a Global Storm Resolving Model (GSRM). The GSRM version of CESM has tremendous potential as a tool for research on the weather-climate interface, particularly as CESM has a worldwide community of developers and users and is the primary model used by the US academic research community. As part of the project, multiple students and early career scientists would be trained at the intersection of atmospheric science and high-performance computing, enhancing a workforce that is crucially needed for scientific advancement.The main objective of the project is to lay the groundwork for simulations with CESM, which is comprised of individual models including the Community Atmosphere Model, for which version 7 (CAM7) is currently under development for release in 2024. CAM7's dynamical core uses the Spectral Element (SE) method to solve the equations of fluid motion under the hydrostatic approximation, in which the acceleration term is neglected in the vertical momentum equation. This project will port the nonhydrostatic SE dynamical core developed for the Department of Energy’s (DOE’s) Energy Exascale Earth System Model (E3SM) into CESM/CAM7. A second, more experimental version of the SE dynamical core meant to run on advanced computing architectures will also be ported. The researchers will then run CAM7 to test the ability to reproduce smaller-scale phenomena. Specifically, the project will: • Port DOE’s nonhydrostatic, Fortran-based, SE dynamical core into the CESM framework.• Integrate an experimental C++ version of the SE core into CESM with a GPU-enabled physics package from the NSF-funded EarthWorks project.• Test and demonstrate the capabilities of the new CESM configuration using mesoscale convective systems and tropical cyclones as exemplars.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.