NSF StormSPEED
The 'Storm-resolving SPEctral Element Dycore' for CESM3
Mission of the project
This NSF StormSPEED project, funded by the U.S. National Science Foundation (NSF), lays the groundwork for future-generation, cloud-permitting community modeling efforts for climate and Earth system applications with the Community Earth System Model (CESM) from the NSF National Center for Atmospheric Research (NCAR). Over the project period (2024-2027) we will
integrate existing nonhydrostatic (NH) versions of the Spectral Element (SE) dynamical core from the Department of Energy into CESM3’s Community Atmosphere Model (CAM) version 7 (CAM7) to enable cloud-permitting coupled climate simulations,
advance CESM’s readiness for the newest high-performance computing (HPC) architectures with accelerators, e.g., Graphical Processing Units (GPUs), and
test and demonstrate CESM’s new scientific capabilities with CAM7 SE-NH via a model hierarchy and selected scientific exemplars.
Two exemplars serve as science drivers for this project. They focus on (i) the interactions of extreme, convectively-driven precipitation over the Continental United States (CONUS) with the large-scale flow and (ii) the impact of multi-scale ocean-atmosphere interactions on tropical precipitation and cyclones. We thereby build bridges between the mesoscale and large-scale scientific communities and provide scientific insights into the impacts of mesoscale motions on the climate system.
Climate modeling at the kilometer scale requires nonhydrostatic dynamical core designs that will be integrated into the CESM climate model.
Integrating GPU technology into the CESM climate model will pave the way for future, computationally affordable climate simulations at the kilometer scale on modern computer architectures.
Kilometer-scale modeling empowers the project team and the CESM community to explore the upscale effects of mescoscale atmospheric and oceanic flow features on weather systems and climate patterns.
The Dynamical Core Model Intercomparison Project (DCMIP) and its associated summer school for students and postdocs sheds light on the design of climate models and their fluid flow component, the so-called dynamical core.
Questions?
Contact cjablono@umich.edu to get more information on the project