Factors Contributing to Bias and Variability of Convective and Stratiform Properties with Various Cloud Microphysics Schemes for a Squall Line MCS Case

A cloud microphysics intercomparison study of a mid-latitude mesoscale squall line is performed using the Weather Research & Forecasting (WRF) model at 1-km horizontal grid spacing with eight cloud microphysics schemes to understand specific processes that lead to the large spread of simulated cloud and precipitation at cloud-resolving scales. For convective core properties, we find that the majority of schemes overestimate convective updraft speed and radar reflectivity aloft, and updraft speed and precipitation have a significant spread across the schemes. The spread in simulated updraft velocity correlates well with the spread of simulated cold pool intensity and latent heating rate. Differences in ice-related parameterizations control updraft velocity variability but not precipitation variability. For the stratiform precipitation, we find that stratiform precipitation variability across schemes is larger than the variability of convective core properties, and the majority of the simulations underestimate stratiform areal coverage and rain rates. The underestimation of stratiform rain rates mainly result from underestimation of frequencies of light and moderate surface rain rates, corresponding to the underestimation of the ice particle number concentration over the 0.2-2 mm diameter range. Stratiform precipitation area (SPA) correlates with convective detrainment flux but is modulated by the hydrometeor size and fall speed. Increasing the frequency of coupling with large-scale forcing from every 3 h to 1 h leads to an increase in SPA by about 30%. Model uncertainties in microphysics and observational needs for gaining insight in improving parameterizations are discussed.