期刊名称：Journal of Advances in Modeling Earth Systems
电子版ISSN：1942-2466
出版年度：2021
卷号：13
期号：6
页码：e2020MS002239
DOI：10.1029/2020MS002239
出版社：John Wiley & Sons, Ltd.
摘要：General circulation models (GCMs) are typically run with coarse vertical resolution. For example, the Energy Exascale Earth System Model (E3SM) has a vertical resolution of about 200 m in the boundary layer, which is far too coarse to resolve sharp gradients often found in the thermodynamic fields capping subtropical marine stratocumulus. In this article, we present a series of multiyear atmosphere only simulations of E3SM version 1 where we progressively increase the vertical resolution in the lower troposphere to scales approaching those often used in large eddy simulation (LES). We report marginal impacts in regards to the simulation of boundary layer clouds when vertical resolution is moderately increased, yet find significant positive impacts when the vertical resolution approaches that typically used in LES (∼10 m). In these experiments, there is a marked change in the simulated turbulence and thermodynamics which leads to more abundant marine stratocumulus. However, these simulations are burdened with excessive computational cost. They are also subject to degradations in overall climate metrics due to time step sensitivities and because some processes and parameterizations are sensitive to changes in the vertical resolution. Plain Language Abstract Models that are used to simulate and predict climate often have trouble representing specific cloud types, such as stratocumulus, that are particularly thin in the vertical direction. It has long been speculated that one of the reasons for this deficiency relates to coarse vertical resolution used in these models. In this study, we increase the vertical resolution to scales that previous process oriented studies suggest are needed to represent these cloud types. We find that increasing the vertical resolution is a necessary ingredient toward simulating stratocumulus, though deficiencies remain and the simulations are computationally expensive.
