Simulations of Entrainment in Continental Thunderstorms: Implications for Parameterization of Deep Convective Vertical Transport

Deep moist convection is an important transport mechanism to the tropopause and lower stratosphere. The wide range of spatial and temporal scales deep convective processes span can result in some processes not being explicitly resolved in current model simulations and therefore requiring parameterization. The evaluation of parameterized convective processes such as entrainment and vertical transport is further complicated by the inaccurate timing and location of simulated convection.

To identify ways to improve convective parameterization and vertical transport, this study evaluates the relative importance of correct convective timing and location versus correct characterization of convective entrainment. This work uses a novel lightning data assimilation (LDA) technique to trigger convection within the Kain-Fritsch (KF) convective parameterization in the Weather Research and Forecasting Coupled with Chemistry Model (WRF-Chem) where lightning was observed. By improving the spatial and temporal placement of convection on the mesoscale, this technique allowed for direct comparison of parameterized entrainment profiles, cloud top height, and convective mass flux profiles to idealized WRF large eddy simulations (LESs). The method was used for evaluation of these processes within two cases observed as part of the Deep Convective Clouds and Chemistry (DC3) field campaign: a 29 May 2012 supercell over Oklahoma and a 11 June 2012 MCS over the south-central United States, allowing further comparison across convective mode.

A combination of idealized WRF-LESs and 12-km WRF-Chem simulations with parameterized convection forced by the LDA technique were used. The LESs were initialized using both an observed sounding and a model-generated sounding extracted from the 12-km WRF-Chem run just prior to convective initiation. LES entrainment was calculated by examining the mass in the updraft core of tracers initialized outside the storm core. It was then normalized by the updraft mass flux at cloud base to allow for direct comparison to KF entrainment output from the 12-km WRF-Chem run. LES entrainment rates from the observed sounding simulations were generally larger than those from the WRF-Chem sounding simulations, which agreed with the evolution of the environments in both WRF-Chem runs. The vertical profiles of the LES entrainment rates often differed significantly from the KF parameterization output, with the primary difference being the location of the secondary entrainment maximum aloft. The differences in distributions of cloud top heights and convective mass flux were also examined. Additional sensitivity tests were also conducted to understand how directly modifying the entrainment parameterizations impact resulting profiles.