6.2    Implementation and Early Tests of a PDF Parameterization in       WRF

Larson, Vincent E., Carsten Harlass, and Jan Hoft of University of Wisconsin, Milwaukee

Currently, WRF does not predict subgrid-scale cloud fraction. However, predicting cloud fraction may aid, for instance, the representation of microphysics in shallow cumulus clouds, even at convection-permitting grid spacings.

 

Unfortunately, parameterizing subgrid cloud fraction --- due to, e.g., cumulus clouds --- poses difficulties at convection-permitting grid spacings.  For instance, at finer grid spacings, there is a breakdown in assumptions for idealized cumulus mass-flux models, such as the assumptions of steady state and small areal coverage.

 

We have implemented into WRF a parameterization of subgrid variability ("CLUBB") that is based on the assumed probability density function (PDF) method.  It predicts the subgrid PDF of heat content, moisture, and vertical velocity.  Thereby it permits the diagnosis of subgrid cloud fraction.  CLUBB is a prognostic parameterization and does not make any assumption about the areal coverage of cloud or updrafts.

We test WRF-CLUBB by simulating a case of marine stratocumulus that was observed off the coast of Chile during the VOCALS field experiment.  In this simulation, CLUBB alone parameterizes all clouds, including stratocumulus, shallow cumulus, and deep cumulus over the Amazon.  Thereby CLUBB is a unified parameterization for clouds.  We perform various sensitivity studies and analyze CLUBB's performance in these simulations.