The Regime-Dependent Benefit of a Three-Moment Bulk Rain Scheme

Two-moment microphysics schemes have been commonly used for cloud simulation in models across different scales – from large-eddy simulations to global climate models. These schemes have yielded valuable insights into cloud and precipitation processes. However, the size distributions are limited to two degrees of freedom, and thus the shape parameter is typically fixed or diagnosed. We have developed a three-moment approach for the rain category in order to provide an additional degree of freedom to the size distribution and thereby improve the cloud microphysics representations for more accurate weather and climate simulations. The approach is applied to the Predicted Particle Properties (P3) scheme. A substantial advancement over existing three-moment schemes is the explicit consideration of self-collection and breakup.

The goal of this study is to establish the rain properties and sub-cloud environments in which we can expect the largest benefits of a predicted over a parameterized shape parameter – or otherwise, in which clouds the cheaper approach may be sufficient. These regimes will depend on the dominance of different microphysical processes in the sub-cloud layer: It is known that a three-moment approach with predicted shape parameter would impact gravitational size sorting. In heavy precipitation, the interplay between rain self-collection and rain breakup is known to be crucial in shaping the size distribution. In a dry sub-cloud layer, evaporation is expected to contribute to a broad size distribution. A rain shaft model is applied to consider a wide range of atmospherically relevant rain scenarios. The investigation includes a varying cloud base height, rain intensity, as well as sub-cloud relative humidity. A systematic comparison is performed among the original two-moment scheme, the three-moment development, and a spectral bin microphysics scheme which serves as a reference. Results are interpreted in terms of the prevalent microphysical processes.