Based on findings from axisymmetric modeling, Persing and Montgomery (2003; PM03) proposed the term ``superintensity'' for hurricanes that are much more intense than the maximum possible intensity (MPI) suggested by the theory of Emanuel (1995; E-MPI). These ideas are summarized in a companion presentation (OPTIMAL HURRICANES: MOTIVATION AND THEORY). Presented here are results that evaluate the Carnot turbocharger model for a hurricane (PM03) in a three-dimensional modeling context. We use the CSU Regional Area Modeling System (RAMS) model with a 1-km grid spacing on the finest grid. The sophisticated radiation, surface interaction, and microphysical parameterizations employed in the standard RAMS model were simplified in a sequence of simulations until a highly-simplified form was developed to be as close as possible to the simple physical parameterizations of the Rotunno and Emanuel (1987; RE87) model (used by PM03). This RAMS simulation is thus a three-dimensional ``analog'' of the RE87 model, except the subgrid-scale diffusion and liquid condensation physical parameterizations of the original RAMS model remain.

This idealized three-dimensional simulation again shows superintensity that is only slightly weaker than that found in PM03 (83 m/s vs 95 m/s in PM03) although still much greater than E-MPI (65 m/s). Many of the structural features identified by PM03 as being associated with superintensity are again found in this simplified-RAMS simulation. The result is not peculiar to the highly simplified physics. As layer-by-layer of the sophisticated physics of RAMS is reintroduced in separate simulations, the basic result of superintensity remains in each, although with varying intensities and structure. Thus, the Carnot turbocharger (a term more general than superintensity since such a eye/eyewall interaction may be present in intensity fluctuations in weaker hurricanes) can plausibly be applied to a context closer to reality.