Summaries - Research
Back An Adaptive and Scalable Unified CG/DG Nonhydrostatic Atmosphere Model
|Division||Graduate School of Engineering & Applied Science|
Giraldo, Francis X.
Kozdon, Jeremy E.
|Sponsor||Office of Naval Research (Navy)|
In this project, we propose to continue the development of the Nonhydrostatic Unified Model of the Atmosphere (NUMA). NUMA is called unified for many reasons. For instance, NUMA is unified by the fact that within a single framework both limited-area and global problems can be considered. Recently NUMA has been extended to include a unified formulation of the continuous and discontinuous Galerkin (CG/DG) methods for the discretization of the spatial differential operators. Currently, NUMA is being further unified to exploit CPU- and GPU-based machines using a single code base. The unified nature of NUMA decreases code duplication, which in turn improves code maintainability and is one of the features that sets it apart as a next-generation model. In this project, we will further unify NUMA through the continued development of algorithms that can be used in both the CG and DG modes of NUMA.
Examples of such algorithms include: 1) static and dynamic adaptive mesh refinement,
2) stabilization methods based on diffusion or hyper-diffusion that can be applied equally to both CG and DG, 3) positivity-preserving schemes that work for both CG and DG, and 4) implicit-explicit methods (with preconditioners) that work for both CG and DG. Finally, we wish to develop a new unified moist convection scheme based entirely on CG/DG methods with the intent to show that such an approach will yield a much more flexible and accurate way of handling physics within element-based Galerkin models.
|Keywords||Adaptive Mesh Discontinuous Galerkin MOISTURE Refinement SPECTRAL ELEMENTS Scalable WEATHER PREDICTION|
|Publications||Publications, theses (not shown) and data repositories will be added to the portal record when information is available in FAIRS and brought back to the portal|