B18. P3's Impact in RRM Simulation of MCSs

Title

Impact of a New Cloud Microphysics Parameterization to Simulations of Mesoscale Convective Systems in E3SM Regionally Refined Model (RRM) Framework

Authors

Jingyu Wang* (PNNL), Jiwen Fan (PNNL), Zhe Feng (PNNL), Kai Zhang (PNNL), Erika Roesler (SNL), and Benjamin Hillman (SNL)

Abstract

Mesoscale convective systems (MCSs) play important roles in the hydrological cycle and general circulation because they are the largest in the family of deep convective clouds and have a major contribution to global precipitation. Traditional global climate models (GCMs) with coarse horizontal resolution (~100 km) fail to simulate MCSs. The pursuit of more accurate and detailed representations of climate processes promotes the need for finer model resolution, thereby establishing the possibility of simulating MCSs in GCMs. Besides resolution, cloud microphysics used in the GCMs generally do not consider convective microphysics (e.g., rimed particles), which is an important precipitation process in mixed-phase and deep convective clouds. In this study, we (1) use a regionally refined mesh with 0.25° grid spacing for the atmosphere and land components within the Energy Exascale Earth System Model (E3SM) atmosphere model to perform regional high-resolution simulations over the contiguous United States, and (2) explore the impact of using a newly developed Predicted Particle Properties (P3) cloud microphysics scheme for E3SM, which is physically more appropriate for simulating deep convective cloud microphysics than the original scheme (MG2, Morrison and Gettelman, 2015). To examine the MCS properties, an observationally-based tracking algorithm is applied to the 0.25° simulations and observations for intercomparison during the period of March-April-May 2011. Results show that despite reasonable agreement in total precipitation, the model simulates insufficient MCS precipitation and excessive non-MCS precipitation compared to the observations.  Additionally, the diurnal cycle of MCSs is out of phase. We find that the underestimation of MCSs by the model is a result of much smaller predicted area of convective systems and lower precipitation rates. The simulation with P3 predicts higher hourly rain rates and larger area of convective systems, resulting in more MCSs and a higher total MCS precipitation compared to MG2, agreeing better with the observations. The larger rain rates predicted by P3 is mainly a result of the melting rimed particles. We show that the larger convective system areas by P3 are associated with much more cloud ice produced and much stronger updraft motion, which is possibly a result of stronger microphysics feedback to the dynamics. The impact of the microphysics parameterization is expected to be larger at higher resolution such as 3 km, which will be examined in our future work.