Shallow Cumulus Sensitivity to Aerosol within a Fixed Meteorology Framework
In order to isolate aerosol induced changes in cumulus mixing that are solely due to microphysical-dynamical interactions and not from mean-state thermodynamic instability changes caused by aerosol-cloud-precipitation feedbacks, this study uses a new approach of forcing shallow cumulus clouds in large eddy simulations (LESs). Nine (9) LESs with systematic variations in aerosol concentration and model domain size are initialized with the well-studied trade cumulus regime of the Barbados Oceanographic and Meteorological Experiment (BOMEX). However, rather than using the standard BOMEX forcing functions for each of the nine (9) simulations, which can result in different mean thermodynamic states when variations in aerosol concentration are imposed, the horizontal mean states of the following four (4) model prognosed, conserved variables are held fixed: liquid potential temperature (θl), total water (qt), zonal wind (u) and meridional wind (v). This guarantees that all variations of the cloud populations and their role in mixing are strictly the result of local microphysical-dynamical changes that result from changes in aerosol concentrations and not from changes to bulk conditional instability.
Results from the nine (9) simulations show aerosol invigoration of cumulus mixing near and above cloud top, where larger magnitudes of θl and qt tendencies occur with increasing aerosol concentration. The larger qt tendencies indicate that in more polluted environments the cumulus clouds detrain more water and transport it higher into the atmosphere. Conversely, due to increased precipitation from enhanced collision-coalescence in the more pristine environments, more cold pools are produced that enhance mixing within the boundary layer. Additionally, the cloud populations increase in number and updraft strength, but decrease in size and precipitation production as aerosol concentration increases. The variation in LES domain size shows reduced cumulus sensitivity to changes in aerosol concentration and thus highlights the importance of using a sufficiently large domain size for numerical studies.