A dynamic probability density function treatment of cloud mass and number concentrations for low level clouds in GFDL SCM/GCM

A Dynamic multi-variate Probability Density Function (Dyn. PDF) parameterization has been incorporated in the GFDL Single Column Model (SCM). This Dyn. PDF parameterization is a higher order turbulence closure scheme coupled with a statistical representation of sub-grid variability in vertical velocity, temperature, and moisture. With the Dyn. PDF scheme, the GFDL SCM successfully predicts cloud fraction and cloud water content for three representative albeit challenging cloud regimes, indicating that the Dyn. PDF scheme is able to parameterize cloudy boundary layer processes reasonably well.

By using the sub-grid variations in vertical velocity, we enhanced the capability of the Dyn. PDF parameterization to represent the activation of cloud droplets. The turbulent transport and evaporation of cloud droplets have been taken into account. The predictions of the droplet number concentration have been evaluated against reference large eddy simulations, and are in good agreement with the reference simulations.

We have conducted high and low vertical resolution simulations, and found that a hybrid configuration, where the Dyn. PDF scheme uses high resolution and the SCM uses low resolution, can largely reproduce the promising results using high resolution in both the Dyn. PDF parameterization and the SCM.