Improving boundary-layer turbulence and cloud processes in CAM with a simplified higher-order turbulence closure scheme

 

Authors

Anning Cheng — IMSG
Kuan-Man Xu — NASA - Langley Research Center

Category

Modeling

Description

A major difficulty in addressing cloud feedback and climate sensitivity stems from the inadequate representation of cloud physical processes in general circulation models (GCMs) with traditional cloud parameterizations. An alternative approach is to use a cloud-resolving model (CRM) to replace cloud parameterizations in every atmospheric grid column, i.e., the multiscale modeling framework (MMF), and implement a sophisticated higher-order turbulence closure (HOC) scheme in its CRM component to better represent turbulence and low-cloud processes. This approach is shown to produce realistic global distribution of low clouds (see the companion poster by Xu and Cheng), but the computational cost increases by nearly a factor of two compared to that of the original MMF (with a first-order turbulence closure in its CRM component) and a factor of 400 compared to that of the Community Atmosphere Model (CAM) at T42 resolution. The improved low-cloud simulation and the tremendous cost of the MMF motivates us to directly implement the HOC scheme in a GCM, i.e., bypassing the CRM component of the MMF.

The new approach will require simplifications and development of an optical approach from the HOC used in CRM because of the large difference in time-stepping and horizontal scales of subgrid variability between the CRM and GCM. We are simplifying the HOC turbulence/low-cloud parameterization to a diagnostic package as one of the options and developing an optimal approach, which will combine the implicit time differencing for the equations with time-derivative terms with the diagnostic approach in the remaining equations. Preliminary results from single-column model testing and recent progress of this project will be presented in this poster.