Sensitivity of resolution and vertical grid types on 3D overflow simulations using mpas-ocean
Summary: The Model for Prediction Across Scales (MPAS) is a climate model framework that supports unstructured, variable resolution grids. Since a primary issue in ocean modeling is the treatment of the vertical coordinate, MPAS-Ocean has been developed to allow for a variety of options in the vertical coordinate choice. The representation of overflows has been shown to be difficult at horizontal resolutions coarser than a few kilometers. Therefore, the combination of the unstructured horizontal grid and the variety of vertical grid choices available with MPAS-Ocean provides a unique approach. MPAS-Ocean is used to simulate an idealized density driven overflow using the dynamics of overflow mixing and entrainment (DOME) setup. Numerical simulations are carried out at a variety of resolutions to compare the accuracy and computational cost of increasing the vertical versus the horizontal resolution. Additionally, various vertical grid types are studied including z-level, z-level with partial bottom cells, and sigma coordinates. Entrainment and transport metrics are calculated and analyzed in order to compare the results from the various grid setups.
Collaborators: Dr. Mark Petersen - Los Alamos National Lab, Dr. Scott Reckinger - Brown University
Reckinger, S. M., M. Petersen, S. J. Reckinger, “Sensitivity of resolution and vertical grid types on 3D overflow simulations using mpas-ocean”, November/2014, APS Meeting, Division of Fluid Dynamics-Talk, San Francisco, CA.
Reckinger, S. M., Petersen, M., Reckinger, S. J., “Sensitivity of Resolution and Vertical Grid Types on 3D Overflow Simulations using MPAS-Ocean”, February/2014, Ocean Sciences Meeting-Poster, Honolulu, HI.
Reckinger, S. M., M. Petersen, S. J. Reckinger, “Understanding how numerical parameters affect dynamics in ocean models”, September 25th, 2014, Fairfield University, Department of Mathematics Colloquium, Fairfield, CT
Support: Clare Boothe Luce Professorship, Los Alamos National Lab.
back to Research.