In an effort to improve the representation of microphysics fields in high-resolution tropical cyclone simulations, this research evaluated the statistics of microphysics fields from 1.67-km simulations of Hurricane Dennis (2005) using the MM5 mesoscale model. Simulated microphysics fields were analyzed at two discrete time periods - “early” (18:00Z July 5th – 00:00Z July 6th) and “late” (06:00Z July 7th – 12:00 Z July 7th) - when Dennis intensified from a weak tropical storm to a category 1 hurricane in the observations. Means and distributions were calculated for various microphysics fields including vertical velocity, reflectivity, and ultimately hydrometeor (i.e. graupel, rain, snow) concentrations. At the forefront of these statistical analyses, the use of contoured frequency by altitude diagrams (CFADs) conveyed the most representative illustration of statistical differences in simulated microphysics fields between the “early” and “late” periods.

These statistical analyses show that the MM5 simulation exhibits a unique microphysical evolution. From the CFADs of vertical velocity (confined to convective regions), it was determined that a small percentage (.1% to .01%) of large updrafts (i.e. isolated convective cores) between 15 m/s and 30 m/s exist during the “early” period; this localized, intense convection is virtually absent (< .01%) during the “late” period of the simulation. Higher mean graupel mixing ratio during the early period (by ~ .5 g/kg) throughout the upper troposphere suggests that isolated convective cores contribute to the majority of the transport of super-cooled liquid droplets above the melting level (~4.5 km). It is hypothesized that an associated temporal variability of latent heat release potentially explains the MM5's deficiency in replicating Dennis's intensification. These differences in microphysics between Dennis's “early” and “late” organization exhibit a temporal trend in the model that will, in future analyses, be statistically compared with observed fields.