UC San Diego
Physical modeling of stratocumulus cloud mixing processes in numerical weather prediction models
- Author(s): Yang, Handa
- Advisor(s): Kleissl, Jan
- et al.
Stratocumulus (Sc) clouds are vast sheets of convective clouds which are responsible for the colloquial “May Gray” and “June Gloom” in southern California. The most prevalent cloud type by area, Sc possess strong reflective properties which result in a net cooling effect on the Earth’s radiative balance.
Accurate prediction of the spatial coverage and lifetime of Sc reduces uncertainty in global climate response due to the associated cloud feedback effect and facilitates the integration of solar power into the electrical grid through predictions of solar energy production. The modeling of Sc in global climate (GCM) and
numerical weather prediction (NWP) models requires parameterization of physics governing Sc which consist of a complex interplay between cloud top- and surface-driven convection, radiative forcings, microphysical processes, and small-scale mixing across a stratified interface. The difficulty in characterizing these processes is compounded by the coarse resolution at which GCMs and NWPs operate.
This thesis focuses on the Weather Research and Forecasting (WRF) NWP model and its ability to predict Sc. First, different initialization techniques are compared to characterize the effect of initial cloud cover. Even when augmented with satellite-derived cloud cover, WRF predicts overly thick clouds over ocean and early dissipation times over land.
Second, the WRF planetary boundary layer (PBL) scheme, which is responsible for mixing processes in the PBL is investigated in a single-column model. A thick cloud bias was discovered to result from a cold and moist bias. Modifications were made to two PBL schemes to better account for deep mixing
due to downdrafts and entrainment mixing at cloud top, leading to more accurate cloud thickness and lifetimes.
Finally, a radiatively-driven downdraft mass flux model was developed in order to account for deep mixing in WRF through an eddy diffusivity-mass flux framework using large-eddy simulation, observational data, and turbulence theory. Many WRF PBL schemes neglect downdraft mixing, but turbulent downdrafts help couple the cloud layer with surface moisture and distribute the warm and dry air mixed in from aloft. These processes aid in sustaining the cloud, preventing the early dissipation bias frequently
observed in weather predictions.