Impacts of Graupel Parameterization on Idealized and Real Case Simulations of Squall Lines

To better simulate storms and associated precipitation within atmospheric models, accurate representation of sub-cloud processes is needed. In the Thompson microphysics parameterization in the Weather Research and Forecasting (WRF) model, graupel and hail are represented by a single hydrometeor category (hereafter, referred to as the graupel category) that only predicts mass mixing ratio and therefore sets particle density to a constant value. This is not realistic given that the densities of graupel and hail are known to vary greatly between and within storms. This study assessed the sensitivity of two simulated squall lines to the prescription of graupel density using the WRF model. One case was an idealized simulation, while the other was a “real” simulation set up to represent an observed squall line. The range of graupel density was varied from 200 kg m-3 to 800 kg m-3, representing particles more characteristic of soft graupel to those of hard hail, respectively. As the density of graupel particles was decreased from being hail-like to graupel-like, the idealized simulation showed a faster squall line with less graupel and more cloud ice. In particular, the lower density graupel case had a notable increase in the graupel melting rate, which resulted in more latent cooling and therefore a more intense cold pool and a faster storm propagation speed. The “real” case exhibited similar trends in graupel melting rate and ice production, but the sensitivity of its cold pool to graupel density was not as strong. This paper will present the results of changing graupel density in these two simulations and the implications this may have on improvements that could be made to the graupel parameterization in the Thompson microphysics scheme.