Tropical cyclone (TC) propagation in the presence of diabatic heating and subgrid-scale diffusion is investigated with a 16-level primitive equation hurricane model developed at BMRC. This model includes all physical processes that are currently included in operational models except the solar radiation. The study focuses on fundamental dynamics involved in the motion and intensity change of an initially symmetric vortex on a beta-plane. By coupling the hurricane model with an intermediate ocean model, the influence of air-sea interaction on TC propagation is also investigated. In order to assess quantitatively the contributions of various physical processes to TC propagation, a new diagnostic approach (potential vorticity (PV) tendency analysis) is advanced.
It is found that the vortices of lower and middle levels move to the region of maximum PV tendency and that the contributions of different physical processes to vortex propagation can be well estimated with the proposed diagnostic approach. In contrast to an adiabatic vortex, the diabatic vortex propagation is not necessarily due to horizontal PV advection by the asymmetric flow over the vortex core region. The contributions of various physical processes may vary with time and height. The horizontal PV advection and the diabatic heating are dominant processes in TC propagation. The vertical PV advection can play a role while the contribution of horizontal and vertical diffusion is negligible.--- The asymmetric diabatic heating can directly induce a positive PV tendency, which causes a vortex propagation component toward the region of maximum heating rate. On the other hand, the asymmetric diabatic heating induces an asymmetric flow field on the mesoscale. It is shown that the maximum diabatic heating rate is located in the upstream of the asymmetric flow field near the vortex center. Therefore, the influence of the PV tendency caused by the asymmetric flow field partly cancels the direct effect of the diabatic heating.
The influence of air-sea coupling on TC propagation is investigated in a resting environment. The TC in the coupled experiment turns more to the north than in the uncoupled experiment with the fixed SST. The track difference is primarily caused by the difference of the asymmetric diabatic heating fields between the coupled and uncoupled experiments. In the coupled experiment, the asymmetric diabatic heating is significantly reduced and its pattern shifts azimuthally. As a result, in the coupled experiment, the diabatic heating contributes less to the southward propagation, thus the beta drift has a larger poleward component compared with that obtained in the uncoupled experiment.