Our research of Andrew (1992) was the first successful cloud-resolving modeling study of a hurricane in which we incorporated more sophisticated cloud microphysics and other physical processes. The same model physics schemes were then applied to the modeling of Hurricane Bonnie (1998), which differed significantly from Hurricane Andrew in terms of track, intensification, storm environment, and inner-core structures. This implies that such a cloud-resolving approach is robust and it could feasibly be used to predict tropical cyclones in an operational setting.

Our current research goals are built upon our previous findings as well as the findings of other researchers. For example, we have been diagnosing the modeled four-dimensional (i.e., x, y, z, t) high-resolution data, at increments of 4-6 km and 15 min that cannot be obtained from any existing measurement, to examine the inner-core structures and evolution of the hurricanes, and gain insight into their relationship with hurricane intensity as well as their environmental conditions. In addition, after analyzing some interesting features of the mature storms, we are being motivated to study their origins, namely, from where they first appear as a weak atmospheric disturbance off the shore of West Africa.

  Please talk a little bit about your 2-part 1996 Monthly Weather Review report on oceanic cyclogenesis as induced by a mesoscale convective system: what were the major findings and implications?

In these two papers, we studied the transformation of a continental convective storm into a tropical cyclone in which the cyclogenesis occurs as the former moves offshore and deep convection develops within its associated cyclonic flow. This convective storm was responsible for the Johnstown, Pennsylvania, flash flood of July 1977. It was shown that our computerized model could reproduce the long life cycle of the transforming process up to four days. It was found that the cyclonic flow associated with the continental storm helps eliminate a low-level cold pool, and produces localized cyclonic vorticity in the lower troposphere through the organized new convection over the ocean surface, thereby transforming a cold-domed continental storm to a warm-cored, rapidly rotating tropical cyclone.

This was the first case study showing that tropical cyclogenesis occurs as a result of the upward transport of cyclonic rotation from the lower troposphere, the so-called the bottom-up mechanism. The results have important implications to the understanding of tropical cyclogenesis from typical convective storms.

In our 2006 paper entitled, "Shear-forced vertical circulations in tropical cyclones," we found that when a hurricane vortex is embedded in a vertically sheared environment, a counter-shear vertical circulation will be induced in the inner-core region with rising motion on the downshear side of the eyewall, sinking motion on the upshear side of the eyewall, and counter-shear flows across the radius of maximum wind in the vertical. This wavenumber-1 vertical motion asymmetry accounts for the development of more clouds and precipitation that are often observed on the downshear-left side of the eyewall.

The result also has important implications to the development of tropical cyclones or other atmospheric vortices in vertically sheared environments. That is, while vertical shear is still inimical to the hurricane intensification, the shear-induced vertical circulation could reduce the destructive action of the environmental shear as much as 40%, including its forced vertical tilt, and facilitate the vertical coupling of vortical flows. This vortex-restoring effect helps partly explain why some portion of environmental air is forced to flow around a tropical cyclone, more in the upper troposphere, as if it were an "obstacle," rather than flowing through it.

You are referring to the operational seasonal prediction using global circulation or climate models. This, in the US, has been officially carried out by the National Centers for Environmental Prediction. My research deals primarily with the short-term prediction of the track and intensity of a few tropical storms. I am particularly interested in studying past hurricanes that are either poorly predicted or incapably understood in order to advance our scientific research on hurricanes and the other convective storms.

I would couple our current atmospheric model with the models of ocean circulation, ocean wave, and inundation in an attempt to reproduce the roles of air-sea interaction in generating the right intensity of Katrina, including surface winds and precipitation, the right timing and right intensity of the storm surge, and the right coverage and magnitude of inundation after landfall.

The operational model predicted reasonably well the track and landfalling timing and location of the storm. Hopefully, the modeling tools so developed, after verifying against observations, can be used to predict the other landfalling storms. In addition, the computerized model data so generated would be used to provide a better understanding of various meteorological processes taking place in Hurricane Katrina (2005).

The main thrust of my current research may include the computerized modeling of developing vs. non-developing tropical cyclones, extratropical transition of landfalling hurricanes, the storm-environment interaction, and balanced dynamics in hurricanes.