Deficiencies of slab models of the hurricane boundary layer

Simplified models of the tropical cyclone boundary layer have multiple uses, including for understanding the observed flow, as part of potential intensity models, for understanding the effect of the boundary layer on the rest of the storm, for risk assessment and engineering design, and for forcing storm surge and wave models. Most such models aim to diagnose the boundary-layer flow given the structure of the hurricane as a whole. Many such models further simplify the problem by azimuthal averaging, linearization, or depth averaging. Here, the focus in on depth-averaged, or slab models.

The flow from a slab model will be compared to that from the model of Kepert and Wang (2001), which fully resolves the vertical structure of the boundary layer. Identical forcing and, so far as is possible, physical parameterisations are used. The differences between the models are both large and significant. Relative to the full model, the slab model typically has too strong inflow, azimuthal flow that is too weak at large radius and too strong near the radius of maximum winds (RMW), and too strong an updraft near the RMW. In moving storms, the motion-induced asymmetry is several times too strong. The slab model also displays an unphysical sensitivity to the Coriolis parameter, while the tendency to produce oscillatory solutions near the RMW (previously noted by others), is shown to occur at large radius also. Reasons for the unrealistic performance of slab models will be given.

Recently, a novel hypothesis for tropical cyclone intensification has been advanced, that was developed based partly on results from slab model simulations. The implications of these results for that hypothesis will be discussed.