On the Sensitivity of WRF Microphysics to Aerosol Concentration

Changes in atmospheric aerosol concentrations could alter the nature of deep convective clouds through microphysical effects. Deep convection usually entails heavy precipitation, which has a large impact on the heat and moisture distributions in the atmosphere. It could also alter the local radiative budget through anvil clouds, as well as the local, regional, and large-scale circulations. Thus, it is important to understand how deep convection changes with aerosol concentration. Some of the observational studies have shown increases in the amount of precipitation with increased aerosol loading; however, modeling studies are less consistent and do not agree on the sign of precipitation change. This study uses the Weather Research and Forecasting (WRF) model as a cloud resolving model with fine (1km) resolution and simulates a quarter-circular shear supercell throughout its lifetime with eight different aerosol loadings. Varying either cloud droplet concentration or the number of activated cloud condensation nuclei mimics variations in aerosol concentrations. The simulations run with three microphysics schemes so that the differences between their responses are examined as well. We show that the sensitivity of precipitation to aerosol concentration is heavily dependent on the microphysics scheme used. For the schemes that are sensitive to aerosol loading, riming of snow, or the production of graupel is the key factor in determining the amount of precipitation, as cold rain dominates over warm rain in the simulated convective clouds. Snow riming is a function of cloud droplet size and number, which produce two competing effects; graupel formation is enhanced either when few but large droplets rime efficiently (the size effect), or when small but many cloud droplets are present (the number effect). In addition to this idealized study, the WRF simulations of continental deep convection with real meteorological data are done, in order to confirm the findings from the idealized case study.