Boundary Layer and Deep Convective Clouds Simulated by a Multi-scale Modeling Framework

This study investigates cloud systems, especially boundary layer clouds, using a Multi-scale Modeling Framework (MMF). The Langley Research Center cloud resolving model (CRM) with an intermediately prognostic (IP) third-order turbulence closure is used to replace the cloud parameterization in each grid box of the finite-volume general circulation model (GCM) of NASA Goddard Space Flight Center in the MMF. The vertical grid spacing is 200 m near the surface and more than 2 km near 20 km with a total of 20 levels for the CRM. The horizontal grid spacing is 4 km for the CRM, with a total of 32 grid cells. The GCM is integrated from January 1 to March 1, 1998, with horizontal resolution of 2 (latitude) by 2.5 (longitude) and 32 levels in vertical direction from surface to 0.4 mb. Preliminary results show that the MMF is able to produce reasonable cloud fractions of boundary layer and deep convective clouds despite the coarse horizontal and vertical resolutions used in the CRM. The model also produces reasonable locations of the storm tracks in both hemispheres and the ITCZ. The cloud amount and location of the stratocumulus clouds near the west coast of North and South America compare well with those observed. This may be due to the sophisticated turbulence closure adopted by the CRM. The CRM used a double-Gaussian probability distribution function (pdf) to parameterize the subgrid-scale variables. The skewness of vertical velocity, total water content and liquid-water potential temperature are predicted to determine the pdf in addition to the predicted lower-order moments.