Analysis of hydrometeor vertical velocities in non-drizzling to heavily drizzling stratocumulus: Simulations versus long-term observations

Stratocumulus clouds are challenging to simulate accurately owing in part to their propensity to drizzle and its strong feedbacks on boundary layer dynamics. To examine the fidelity of model drizzle properties, here we compare simulated values of mean Doppler velocity (MDV, the sum of air vertical velocity plus reflectivity-weighted hydrometeor fall speed) binned by radar reflectivity to robust measurements from profiling cloud radar. The radar observations represent 27 days of marine boundary-layer clouds from the 2009-2010 Clouds, Aerosol, and Precipitation in the Marine Boundary Layer (CAP-MBL) campaign. Simulations are performed using the DHARMA LES code with two differing microphysics schemes: (1) two-moment (bulk) microphysics and (2) size-resolved (bin) microphysics. Systematic deviations from observations across simulations are identified, which appear to depend largely upon the microphysics scheme used. In simulations using bulk microphysics, we find that average MDV (positive for downward motion) does not increase with reflectivity as in the observations. This shortcoming is attributable to the representation of drizzle in the bulk scheme. Using bin microphysics, we find anomalously high MDV at intermediate reflectivities compared with observations, a feature that shifts to smaller reflectivities for larger droplet concentrations. This feature appears to be associated with inaccurate treatment of droplet evaporation in downdrafts in the bin microphysics scheme. The limiting behaviors of the MDV distribution with reflectivity are also examined. At low reflectivities (-40 to -50 dBZ), the simulations examined thus far converge toward average MDV of about 10 cm/s (downward), which is greater than the mean of the observations but within the range of observed cases. At high reflectivities (greater than 0 dBZ), however, simulations with bin microphysics, which largely reproduce the correct trend of downward motion in terms of MDV, yield maximum MDV and reflectivity that are not as great as in the observations. We will discuss the significance of these factors for the representation of geophysical parameters such as entrainment rates, liquid water paths, and precipitation rates and their dependence on aerosol concentrations.