Cloud-Resolving Modeling Of An MC3E Squall Line Case:Intercomparison Of Simulations With Different Microphysics Schemes

Modeling mesoscale convective systems(MCSs)properly and precisely is one of the greatest challenges in atmospheric sciences research.Although models benefit from finer grid resolution,cloud-resolving simulations still have deficiencies in reproducing the observed internal kinematic and microphysical structure of MCSs.This study conducts the cloud-resolving modeling of a squall-line MCS case that occurred during MC3E field campaign using the WRF model with eight cloud microphysics schemes.Various observational datasets are employed to evaluate simulations from the perspective of dynamics,thermodynamics,and microphysics to understand main causes for model biases.Potential reasons are also explored for the variability of simulated convective and stratiform properties across different microphysics schemes.The main conclusions of this study are as follows:Comparisons with a multi-Doppler radar wind field retrieval suggest that simulations significantly overestimate convective updraft velocity above the melting level.The simulations also underestimate convective updraft area at upper levels,while producing small but intense updraft cores aloft regardless of the microphysics scheme.This bias is likely related to the inadequate representation of diffusion and mixing in the model.Simulated updraft velocity is sensitive to the chosen microphysics scheme.Differences in simulated updraft intensity correlate well with differences in both simulated buoyancy and low-level vertical perturbation pressure gradient accelerations.The magnitude of simulated low-level vertical perturbation pressure gradient correlates well with cold pool intensity,while cold pool intensity is strongly controlled by evaporation rate.The magnitude of simulated buoyancy correlates well with total latent heating released from microphysical processes,while condensation and riming processes might be the main factors leading to the variability in simulated total latent heatingThe upper-level updraft velocity is reduced by half from the full microphysics runs to the no-ice runs,and the variability in simulated updraft velocity is reduced by more than half.This indicates that ice-related microphysical processes could significantly enhance convective strength,and are a major contributor to the variability in simulated updraft velocity.Ice-related microphysical parameterizations can cause the convective updraft intensity variability among the simulations through two mechanisms:(1)increasing differences in evaporation rate and thus cold pool intensity,and(2)increasing differences in latent heating and thus buoyancy.Most simulations overestimate stratiform IWC above 7-km altitude but underestimate stratiform IWC right above the melting level,while the latter bias in IWC leads to the underestimation of RWC below 3-km altitude.All simulations fail to simulate the observed increase of IWC as ice approaches the melting level.The biases in the magnitude of IWC at upper levels may be related to detrainment that is too high as a result of overestimated convective intensity.The discrepancy between observed and simulated IWC profiles may be resulted from a combination of aggregation that is too weak and sedimentation that is too fast in the simulations.All simulations underestimate descending motion in the stratiform region below 3-km altitude and fail to reproduce the observed connection between RWC and downdraft This may be a result of differences in the structure of the rear-inflow jet in observations and simulations.The observed jet slowly descends through the stratiform region and intensifies the mesoscale downdraft,thus leads to significant rain evaporation.Simulated jets descend quickly and remain at relatively lower altitudes,such that a robust mesoscale downdraft is not present and rain evaporation may be limited.The majority of simulations underestimate total stratiform precipitation due to the underestimation of stratiform precipitation area.The simulated stratiform precipitation area is positively correlated with convective condensate detrainment flux but is modulated by hydrometeor type,size,and fall speed.Stratiform precipitation area is also sensitive to the update frequency of large-scale forcing.Simulated stratiform precipitation area increases by 17-25%when the lateral boundary condition update frequency is increased from every 3 h to every 1 h.The magnitude and variability of stratiform precipitation among schemes are ultimately related to the magnitude and variability of ice particle mass fluxes above the melting level.Since the detrained condensate from the convective region is the primary source for ice in the stratifom region,so ice particle mass fluxes are significantly impacted by the kinematic and microphysical properties of convective regions.These results suggest that accurate simulation of convective regions is critical to better simulating stratiform precipitation,and therefore,future observations should target convective dynamics and microphysics such that models can be more extensively evaluated and improved.