Exploring Variable and Uniform Resolution Modeling Approaches Using Global MPAS-A Aquaplanet Simulations: Sensitivity to Specification of Equatorial Channel

Friday, 19 December 2014
Matus Martini1, William I Gustafson Jr1, Po-Lun Ma1, William C Skamarock2 and Laura D Fowler3, (1)Pacific Northwest National Laboratory, Richland, WA, United States, (2)Natl Ctr Atmospheric Research, Boulder, CO, United States, (3)NCAR, Boulder, CO, United States
Climate models with variable-resolution grids offer a computationally less expensive way to provide more detailed information at regional scales and increased accuracy for processes that cannot be resolved by a coarser resolution grid. This study uses the recently released version of the Model for Prediction Across Scales–Atmosphere (MPAS-A), consisting of a non-hydrostatic dynamical core and a subset of Advanced Research WRF (ARW) model atmospheric physics and Community Atmosphere Model version 5 (CAM5) cloud fraction parameterization, to investigate the potential benefits of using increased resolution in an equatorial channel. The simulations are performed with an idealized aquaplanet mode using two quasi-uniform grids, with 30-km and 240-km grid spacing, and two variable-resolution grids spanning the same grid spacing range, one with a narrow (20S–20N) and one with a wide (30S–30N) equatorial channel.

We find that for the wider equatorial channel, there is a closer correspondence with the 30-km quasi-uniform simulation. Results show that increasing resolution in the Tropics impacts both the tropical and extratropical circulation. Compared to the quasi-uniform coarse grid, the narrow-channel simulation exhibits a stronger Ferrel cell and stronger updrafts in the middle of the updraft branch of the Hadley cell. However, the total atmospheric poleward energy transports are similar in all simulations. With the CAM5 cloud fraction parameterization, the cloud microphysics shows moderate resolution dependence. The channel simulations have similar peak ITCZ precipitation as the simulation with quasi-uniform 240-km grid spacing and are only about 10% larger than the quasi-uniform 30-km simulation. All simulations have a single ITCZ. The relatively small differences in mean global and tropical precipitation rates among the simulations are a promising result, and the evidence points to the equatorial channel being an effective method for avoiding the extraneous numerical artifacts seen in earlier studies that only refined portion of the Tropics.