High resolution simulation of tropical storm Ivan (2004) in the Southern Appalachians: Role of planetary boundary-layer schemes and cumulus parametrization

The Weather Research and Forecasting (WRF) model was used to simulate the evolution of tropical storm Ivan (2004) in the southeast United States using both the Yonsei University (YSU) and Mellor-Yamada-Janjić (MYJ) boundary-layer parametrizations. In contrast to tropical cyclone (TC) simulations over the ocean, the effect of the surface layer becomes secondary for a dissipating hurricane along its terrestrial track. Although these two schemes can reproduce Ivan reasonably well, our results suggest that the mixing properties for damped mechanical turbulent conditions (weakly stable) are strongly underestimated by both parametrizations. This underestimation impacts the thermodynamic properties of the storm, leading to significant differences in the storm areal extent and the simulated precipitation fields. Suggestions for further improvements are provided. An evaluation of the impact of using or not using a convective parametrization, specifically the Kain-Fritsch (KF) scheme, at 3 km grid spacing shows marginal impact on storm coverage, intensity and precipitation, except for the presence of widespread light rainfall in the Piedmont east of the mountains when the KF is employed. Analysis of the thermal structure of the simulated storm indicates that, in the inner-storm region, the KF is either not activated or primarily produces (parametrized) shallow convection. As a result, the net heating tendency associated with adiabatic and diabatic processes is almost unaltered inside the storm, together with a nearly equivalent surface momentum sink, leading to similar storm areal extent and intensity. Light rainfall to the east of the mountains can be due to the trigger mechanism of KF, which depends on boundary-layer convergence, forcing parametrized deep convection near the coast, where surface roughness changes enhance convergence.