Test of the volume-of-fluid method on simulations of marine boundary layer clouds

The impact of using grid-averaged thermodynamic properties (i.e., neglecting their subgrid variability due to partial cloudiness) to represent forcings for condensation or evaporation has long been recognized. In particular, numerical difficulties in terms of spurious oscillations and/or diffusion in vicinity of a cloud-environment interface have been encountered in most of the conventional finite-difference Eulerian advection schemes. This problem is equivalent to the inability of models to accurately track the cloud boundary within a grid cell, which eventually leads to spurious production or destruction of cloud water at leading or trailing edges of clouds. This paper employs a specialized technique called the "volume-of-fluid" (VOF) method to better parameterize the subgrid-scale advection process that accounts for the transport of material interfaces. VOF also determines the actual location of the partial cloudiness within a grid box. Consequently, relevant microphysical parameterizations in mixed cells can be consistently applied in "cloudy" and "clear" regions. The VOF technique is incorporated in a two-dimensional hydrodynamic model to simulate the diurnal cycle of the marine stratocumulus-capped boundary layer. The fidelity of VOF to advection-condensation processes under a diurnal radiative forcing is assessed by comparing the model simulation with data taken during the ISCCP FIRE observational period as well as with results from simulations without VOF This study shows that the VOF method indeed suppresses the spurious cloud boundary instability and support, a multiday cloud evolution as observed. For the case without VOF, the spurious instability near the cloud top causes the dissipation of the entire cloud layer within a half of a diurnal cycle.