You are here:
Software tools for the Generation of MPAS Horizontal Meshes to Support Global to Regional Coupled Meteorology and Air Quality Modeling
Citation:
Bullock, R. Software tools for the Generation of MPAS Horizontal Meshes to Support Global to Regional Coupled Meteorology and Air Quality Modeling. U.S. Environmental Protection Agency, Washington, DC, 2021.
Impact/Purpose:
The US EPA is developing an Advanced Air Quality Modeling System (AAQMS) capable of simulating meteorology and air quality on global to regional to local scales. This system is based on a relatively new meteorological model called the Model for Prediction Across Scales (MPAS). MPAS discretizes the atmosphere in the horizontal dimensions on a mesh of polygons and offers gradual refinement of these horizontal elements to avoid problems associated with abrupt changes in model resolution. MPAS developers at the National Center for Atmospheric Research (NCAR) offer a limited number of horizontal meshes tailored for use in simulating physical meteorology. However, air quality modeling often imposes additional requirements on the horizontal model structure than purely meteorological models. Therefore, it was critical that the US EPA be able to generate appropriate MPAS horizontal meshes to support the present development and future application of the AAQMS. The AAQMS will be significantly improved by being better able to resolve pollutant emissions, transport, transformation, and deposition at all locations where those processes are known to be important. MPAS mesh refinement can now be tailored specifically to satisfy air quality modeling needs over global to regional to local scales.
Description:
We obtained from NCAR the software they used previously to generate MPAS horizontal meshes. NCAR warned that this software requires a considerable length of time (up to months) to complete the generation of highly refined meshes. Its application involves a multi-step process that requires programmer intervention at each step. We developed new Linux command scripts to automate this multi-step process. The software from NCAR only generated mesh refinement areas that are circular is shape. The software was augmented to allow elliptical refinement areas of variable size and stretching factor. This enabled to creation of an elliptical refinement area to better encompass the continental United States. The new software also provides the capability to have more than one focus area for mesh refinement. This has already been used to create an MPAS mesh with separate refinement areas over North America and Europe which provides a unique capability to simulate intercontinental source-receptor relationships. This research effort endeavored to reduce the computational time requirement, simplify the entire mesh generation process, and enable more complex patterns of mesh refinement to better suit the needs of advanced air quality modeling at EPA and any of its partners around the world. Modified Fortran and C programming and new Linux command scripts were developed to generate MPAS horizontal meshes suiting the needs of air quality modeling while providing a faster computational solution and improved guidance to the user throughout the mesh generation process.