RACORO continental boundary layer cloud investigations: 3. Separation of parameterization biases single-column model CAM5 simulations of shallow cumulus

Climatically important low‐level clouds are commonly misrepresented in climate models. The FAst‐physics System TEstbed and Research (FASTER) Project has constructed case studies from the Atmospheric Radiation Measurement Climate Research Facility's Southern Great Plain site during the RACORO aircraft campaign to facilitate research on model representation of boundary‐layer clouds. This paper focuses on using the single‐column Community Atmosphere Model version 5 (SCAM5) simulations of a multi‐day continental shallow cumulus case to identify specific parameterization causes of low‐cloud biases. Consistent model biases among the simulations driven by a set of alternative forcings suggest that uncertainty in the forcing plays only a relatively minor role. In‐depth analysis reveals that the model's shallow cumulus convection scheme tends to significantly under‐produce clouds during the times when shallow cumuli exist in the observations, while the deep convective and stratiform cloud schemes significantly over‐produce low‐level clouds throughout the day. The links between model biases and the underlying assumptions of the shallow cumulus scheme are further diagnosed with the aid of large‐eddy simulations and aircraft measurements, and by suppressing the triggering of the deep convection scheme. It is found that the weak boundary layer turbulence simulated is directly responsible for the weak cumulus activity and the simulated boundary layer stratiform clouds. Increased vertical and temporal resolutions are shown to lead to stronger boundary layer turbulence and reduction of low‐cloud biases. Key Points Integrated SCM‐LES framework for the evaluation of cumulus scheme in CAM5 Large low‐level cloud biases in SCAM5 traced to insufficient simulated PBL TKE Underrepresentation of shallow cumulus in SCAM5 distorts PBL cloud processes