Effects of cumulus parameterizations on tropical cyclone potential vorticity structure and steering flow

The potential vorticity (PV) structure of tropical cyclones (TCs) is characterized by a cyclonic PV tower surmounted by a lens of anticyclonic PV. The advection of this cyclonic PV tower determines the motion of the TC. The steering flow of TCs is typically defined as the layer-averaged wind field in a column encompassing the TC circulation. It has been observed that the steering flow of weaker (stronger) TCs is defined by a layer-averaging of the wind field over a relatively shallow (deep) depth. Additionally, it has been observed that weak (intense) TCs are typified by a shallow (deep) PV tower. The structure of the PV tower is dependent upon the distribution of latent heating within the TC; the PV tower of a modeled TC is therefore dependent upon the cumulus parameterization scheme used in the model. As a consequence, the choice of the cumulus parameterization scheme could influence a modeled TC's track due to the interactions between the steering flow and the parameterization-dependent cyclonic PV tower.

In this study, we explore the role that the choice of cumulus parameterization plays in the modeled PV structure and related motion of a TC. Multiple NCAR/Penn State MM5 model simulations of Hurricane Andrew (1992) are performed. An ensemble of simulations is performed using the same initial condition but different cumulus parameterization schemes. For each cumulus parameterization, a quantitative diagnosis of PV tendency is performed. Additionally, we analyze the best-fit steering flow column for the TC at each model output time. We then evaluate the correlation between changes in PV structure and the depth of steering flow columns. Finally, we investigate a PV-weighted steering flow and its accuracy compared to the non-weighted best-fit steering flow column.