In a conventional climate model or general circulation model (GCM), subgrid-scale (SGS)
processes such as deep convection, shallow convection, turbulence, radiation and microphysics
are parameterized separately, so these processes can interact only indirectly through
the resolvable large-scale motions. Therefore, many important interactions among physical
processes are missing in conventional GCMs. Global Cloud-Resolving Modeling (GCRM)
provides a cloud-scale framework for these processes to interact and has been shown to produce
better climate simulations, depite of its required computer resources. However, many
critical subcloud-scale processes still need to be ”represented” in GCRMs. In this talk, I will
focus on the SGS transport problem.
Representations of SGS transport in GCRMs raise a couple new issues. Convective
motions are generally continuous across a wide range of scales; there is no spectral gap
between the scales that are resolvable in GCRMs and those that are not. Furthermore, the
motions near a typical GCRM grid cutoff are likely associated with individual clouds and are
thus crucial in transport processes. In a GCRM the problem in representing the effects of
SGS convection and turbulence resides in their local relationship with the resolvable cloudscale
field. In particular, the scale interaction near the GCRM grid cutoff may be crucial to
the SGS problem.
To study the SGS problem in GCRMs, scientists at CMMAP (Center for Multiscale Modeling
of Atmospheric Processes) carried out a benchmark large-eddy simulation of a tropical
deep convection system. This benchmark simulation resolves a broad range of scales from
the mesoscale convection system, to individual clouds, down to energy-containing turbulent
eddies. In this talk, I will show how this simulation is used as database to investigate the
SGS transport problem for GCRMs.