Simulations of an isolated two-dimensional thunderstorm : sensitivity to cloud droplet size and the presence of graupel

Abstract

Cloud Resolving Models (CRMs) which are used increasingly to make operational forecasts, employ Bulk Microphysics Schemes (BMSs) to describe cloud microphysical processes. In this study two BMSs are employed in a new Nonhydrostatic σ-coordinate Model to perform two hour simulations of convection initiated by a warm bubble, using a horizontal grid resolution of 500 m. Different configurations of the two BMSs are applied, to test the effects of the presence of graupel with one scheme (2-configurations) and of changing the cloud droplet sizes in the second scheme (4-configurations), on the simulation of idealised thunderstorms. Maximum updrafts in all the simulations are similar over the first 40 minutes, but start to differ beyond this point. The first scheme simulates the development of a second convective cell that is triggered by the cold pool that develops from the outflow of the first storm. The cold pool is more intense in the simulation with graupel because of melting of graupel particles, which results in relatively large raindrops, decreases the temperature through latent heat absorption, causing stronger downdrafts, which all contribute to the formation of a more intense cold pool. The second scheme simulates the development of a second cell in two of its configurations, while two other configurations do not simulate the redevelopment. Two configurations that simulate the secondary redevelopment produce a slightly stronger cold pool just before redevelopment. Our results show that small differences in the microphysics formulations result in simulations of storm dynamics that diverge, possibly due nonlinearities in the model.