Tornadoes, although the subject of numerous studies, remain somewhat of a mystery.  While there have been improvements in the prediction of tornadoes and tornado outbreaks, it is still unknown how tornadoes actually form and why one particular storm will produce a tornado while a similar storm will not.  Recent studies have suggested that the rear flank downdraft (RFD) of a supercell is crucial to the development of a tornado.  Markowski et al. (2002) analyzed 30 supercell storms (18 tornadic and 12 non-tornadic) and found that, on average, the tornadic supercell storms had RFDs with pseudoequivalent potential temperature (θ_ep) values closer to those observed at the surface in the storm's environment than did the nontornadic supercells.  This was particularly true in the case of strong tornadoes [defined by Markowski et al. (2002) to be tornadoes of ≥ F2 intensity or tornadoes that persisted longer than 5 minutes].

Previous numerical simulations of RFDs only produced ‘cold' downdrafts [e.g., Schlesinger (1975), Wilhelmson and Klemp (1978), Klemp et al. (1981)].  This could be due to the absence of a layer of reduced lapse rate, or cap, in the storm environment which can deter air aloft from descending to the surface.  Askelson et al. (2004) found that the θ_ep deficit at the surface was minimized in a storm environment that had a cap just above the boundary layer.  This cap prevented upper level air with low θ_ep values from descending to the surface, thus minimizing surface θ_ep deficits and creating an environment that, according to Markowski et al. (2002), is favorable for tornadogenesis.

Askelson et al. (2004) used a simplified downdraft model that did not consider pressure perturbation forcing.  The model used in this study is the Weather Research and Forecasting (WRF) model.  The WRF model is a three dimensional model that allows both hydrometeor and pressure-perturbation forcing of the RFD.  This study expands on the work of Markowski et al. (2002) and Askelson et al. (2004) by investigating (a) the characteristics of the pre-storm environment that will produce RFDs with relatively high θ_ep and (b) if the results obtained in Askelson et al. (2004) remain valid in a full scale cloud model that considers pressure perturbation forcing.  In addition, the effects of microphysical parameterizations on the thermodynamic structure of the RFD are examined.