Exploring the impacts of physics and resolution on aqua‐planet simulations from a nonhydrostatic global variable‐resolution modeling framework

Abstract

The nonhydrostatic Model for Prediction Across Scales (NH‐MPAS) provides a global framework to achieve high resolution using regional mesh refinement. Previous studies using the hydrostatic version of MPAS (H‐MPAS) with the physics parameterizations of Community Atmosphere Model version 4 (CAM4) found notable resolution‐dependent behaviors. This study revisits the resolution sensitivity using NH‐MPAS with both CAM4 and CAM5 physics. A series of aqua‐planet simulations at global quasiuniform resolutions and global variable resolution with a regional mesh refinement over the tropics are analyzed, with a primary focus on the distinct characteristics of NH‐MPAS in simulating precipitation, clouds, and large‐scale circulation features compared to H‐MPAS‐CAM4. The resolution sensitivity of total precipitation and column integrated moisture in NH‐MPAS is smaller than that in H‐MPAS‐CAM4. This contributes importantly to the reduced resolution sensitivity of large‐scale circulation features such as the intertropical convergence zone and Hadley circulation in NH‐MPAS compared to H‐MPAS. In addition, NH‐MPAS shows almost no resolution sensitivity in the simulated westerly jet, in contrast to the obvious poleward shift in H‐MPAS with increasing resolution, which is partly explained by differences in the hyperdiffusion coefficients used in the two models that influence wave activity. With the reduced resolution sensitivity, simulations in the refined region of the NH‐MPAS global variable resolution configuration exhibit zonally symmetric features that are more comparable to the quasiuniform high‐resolution simulations than those from H‐MPAS that displays zonal asymmetry in simulations inside the refined region. Overall, NH‐MPAS with CAM5 physics shows less resolution sensitivity compared to CAM4.

ICON-ART 2.1: a flexible tracer framework and its application for composition studies in numerical weather forecasting and climate simulations

“…Thus, simulations ranging from a few hours to hundreds of years may become more consistent and advances of one scale or application can more easily be transferred to another. However, the construction of scale-aware physical parameterisations that can be applied over the entire range of scales remains a challenge (Rauscher et al, 2013;Chun et al, 2016). The extended modelling system ICON-ART (ICOsahedral Nonhydrostatic -Aerosols and Reactive Trace Gases; Rieger et al, 2015) is a representative of a next-generation system.…”

Resolution Dependence and Rossby Wave Modulation of Atmospheric Rivers in an Aquaplanet Model

“…For the CAM4 hydrostatic simulations, a strong poleward shift in the jet is found in the 120‐km simulations compared with that in the 240‐km simulations, which must have an impact on AR frequency (Hagos et al, ; Zhao et al, ). However, there are only subtle differences in climatological mean zonal wind among the nonhydrostatic simulations examined here (Zhao et al, , see their Figure 6), and very little variation in climatological mean water vapor as well (not shown). What does vary with resolution a great deal is the climatological time and zonal mean LWA of 300‐hPa absolute vorticity.…”

Physics–Dynamics Coupling in Weather, Climate, and Earth System Models: Challenges and Recent Progress

“…Similar sensitivities could negatively affect VR simulations that feature multiple resolutions within a single simulation. For example, a striking difference in precipitation appears inside and outside the high-resolution domains in aquaplanet VR simulations using the CAM-SE dynamical core (Zarzycki et al 2014b) or the MPAS-A dynamical core (Hagos et al 2013;Rauscher et al 2013;Zhao et al 2016). In the following, the effects of physicsdynamics coupling on the model sensitivity to spatial resolution in VR modeling are briefly explored.…”