Thirteen groups participated in the large-eddy simulation intercomparison. Two of the models resolved the size distributions of aerosol (based on measurements) and cloud particles, and explicitly treated cloud microphysics. The other models prescribed the cloud droplet concentration based upon the measurements, and parameterized precipitation, ignoring the sedimentation of cloud droplets. A parameterization of cloud droplet sedimentation was provided for those models, and simulations were run with and without drizzle, with and without cloud droplet sedimentation.
The majority of simulations did not match the horizontally-averaged measurements of liquid water path or precipitation, though the relationship between liquid water path and cloud-base precipitation within the ensemble generally falls within the considerable uncertainty of field measurements obtained from a variety of measurements. Adding cloud droplet sedimentation as well as drizzle tended to increase the agreement with the measured fluxes of heat and moisture, while diminishing the agreement with the measured second and third moments of vertical wind.
Including cloud droplet sedimentation as well as drizzle unanimously resulted in decreased entrainment. Cloud droplet sedimentation unanimously resulted in increased liquid water path. Drizzle resulted in increased liquid water path in some, and decreases in other, simulations; for the median of the ensemble it increased.