Bin and Bulk Schemes Are More Alike Than They Are Different

Bin and bulk microphysics schemes are two primary methods to model a collection of cloud droplets. The most prominent difference between these two schemes is their structural difference: bin schemes specify the number of droplets in each size bin, while bulk schemes predetermine a functional form for the droplet size distribution. In practice, however, they are also different at another level: how physics is parameterized, and separating these two levels of difference is difficult. Here, to control the variable, we will use a novel microphysics scheme (Arbitrary Moment Predictor, or AMP) that was specifically designed to have the same physics parameterization as a bin scheme, while modeling the droplet size distribution the same way as a bulk scheme. Using AMP as a proxy for bulk scheme will thus guarantee that any difference shown in AMP and bin simulation results is a consequence of their different structural design. We will present the results from 1D and 2D simulations that test a wide range of physical conditions, e.g., initial altitudes, cloud water content, updraft profiles, and aerosol concentrations. We find that the overall difference between AMP and bin schemes to be minor. They have similar albedo (average difference <4%), optical depth (<8%), liquid water path (15%), and peak precipitation rate (15%) regardless of the initial conditions. The difference between the two schemes is especially small when condensation is the only mechanism for droplet growth, unless a significant amount of cloud water grows to rain size. When the schemes do diverge, collision-coalescence is the most likely reason. However, collision-coalescence does not always cause the schemes to diverge, but rather only when autoconversion proceeds slowly. Sedimentation of mass is somewhat too slow whereas sedimentation of number is somewhat too fast. This may contribute to the small differences in the peak precipitation rate. Although differences are generally small, this study nonetheless helps us better understand what physical processes are sensitive to the fixed functional form and thus gives us guidance as to how to effectively improve bulk schemes.