Implementation and Evaluation of MG Microphysics in FV3GFS

Microphysics schemes are usually very simple in NWP models. For example, the Zhao-Carr microphysics in the NCEP FV3GFS is a simplified one-moment scheme, forecasting the mixing ratios of water vapor and total condensate only. Some studies show that predicting rain and snow mixing ratios may improve precipitation scores because the sedimentation is more realistically simulated. Furthermore, two-moment microphysics schemes provide greater flexibility and detail for NWP models by predicting both the mass and number mixing ratios of hydrometers, allowing coupling with the aerosol processes and representation of homogenous and heterogeneous ice nucleation.

This study presents the implementation and evaluation of the Morrison-Gettelman two-moment microphysics – version 2 (MG2) in FV3GFS. In addition to predicting the mass and number of cloud liquid, ice, rain and snow, MG2 has some beneficial features: 1) unified treatment of cloud fraction in radiation and macrophysics, 2) subgrid-scale microphysics, 3) max-overlap and in-cloud precipitation fraction area, and 4) options for subcolumn microphysics. The unified treatment of cloud and subgrid-scale microphysics allows seamless coupling between the microphysics and the double-Gaussian based pdf in a simplified higher-order turbulence closure. On the other hand, the subcolumn microphysics can be unified with the 3D/subcolumn radiation. The precipitation fraction area also has the potential to be coupled with radiation. Various mean fields, time series, anomaly correlation, and root mean square will be compared with those from state-of-the-art one moment, e.g. Zhao-Carr, and the Thompson two-moment scheme in NCEP FV3GFS.