SR/SSD 99-19
7-1-99

Technical Attachment

A Note on the Baroclinic Structure of
Hurricane Danny at Landfall
by
Kwan-yin Kong
Department of Meteorology
James B. Elsner
Department of Geography
Florida State University

1. Introduction

The term "baroclinic" is usually associated with wintertime mid-latitude cyclones that develop and intensify along fronts separating air masses of contrasting thermal characteristics. The mid-latitude cyclone is conceptualized as a low-pressure system with trough and ridge axes tilted westward with height. The strongest horizontal temperature gradients are found near the ground. The term "baroclinic" however, is not often associated with hurricanes. The vertical thermal structure of a hurricane may be generalized as a warm-core low-pressure system. At low levels the hurricane consists of intense cyclonic inflow while at upper levels it consists of anticyclonic outflow. There is little or no vertical tilting between the low-level inflow and upper-level outflow.

Since they typically develop in the tropics or subtropics, we tend to conclude that hurricanes are barotropic systems. Yet, fundamental dynamic principles assert that atmospheric motions (or more precisely, the rate of change of motion) are driven by temperature gradients (Dutton 1986). A baroclinic atmosphere is defined as one whose density depends on both temperature and pressure. That is, if temperature gradients exist on a constant pressure surface, the atmosphere is baroclinic. Thus, with this definition, a hurricane is actually baroclinic. An idealized warm-core hurricane may be called equivalent barotropic, since isotherms and height contours on constant pressure surfaces are parallel, and the barotropic vorticity equation is applicable.

In this paper, the structure of hurricane Danny of 1997 at landfall over Mobile Bay is examined. Danny exhibited various interesting anomalies during its lifespan. For instance, the origins of Danny can be traced to a complex of thunderstorms which developed off the Louisiana coast, perhaps as a regeneration of a mesoscale convective system which moved south from the Great Plains. Danny also remained nearly stationary over Mobile Bay while maintaining minimal hurricane intensity. The precipitation pattern associated with Danny evolved from a symmetric spiral to a single rainband which was anchored to the west of the center. As a result, more than 40 in of rain fell over western Mobile Bay underneath this rainband. After moving inland, the remnants of Danny remained remarkable intact as it traveled across the southeast United States.

The track of Danny and its remnants is plotted in Fig. 1. Danny re-intensified and produced eight tornadoes in the Carolinas (Edwards 1999) before emerging off the mid-Atlantic coast. Of interest for the present study is the fact that significant surface horizontal temperature gradients accompanied the hurricane at landfall. The presence of these gradients suggest to us an alteration of the wind field and forward motion of the storm. Danny exhibited both tropical and extra-tropical characteristics.

An awareness of non-classical signatures in landfalling hurricanes is important for forecasters in the Southern Region.

2. Analyses

a. Doppler radar

Danny was the only hurricane during the 1997 North Atlantic season to make landfall in the United States. After crossing the mouth of the Mississippi River on July 18, Danny approached Mobile Bay at minimal hurricane intensity (Category One on the Saffir/Simpson scale of hurricane destruction potential). On an animation loop of Doppler radar reflectivity, Danny's eyewall appeared as a convective spiral ring as it approached the bay. The precipitation echoes were symmetric around the center of circulation. The spiral pattern of deep convection persisted as the center of Danny entered the bay at 0900 UTC on July 19. At this time the storm was intensifying, but precipitation rates were dropping (Medlin and Blackwell 1999).

As Danny entered Mobile Bay, its forward motion slowed to an eastward drift. Meanwhile, Doppler radar observations showed the precipitation echoes on the east side of Danny continued to fade, but intense convection began to fire in the western eyewall. By 1800 UTC on the 19th, only a single rainband remained to the north and west of Danny's circulation center. Rainfall amounts associated with this stationary band were copious. Doppler radar estimates indicated a storm total in excess of 40 in fell across the west shore of Mobile Bay, resulting in record high levels of flood water along tidal rivers (Blackwell and Medlin 1999). Rain gauge observations substantially verified the radar estimates.

b. Wind and pressure fields

Danny was a compact, well-organized hurricane. Satellite images and surface pressure fields reveal the diameter of the storm was about 300 km. Surface observations suggest the wind field associated with Danny was different from most hurricanes. Wind speeds on the west side of Danny (under northwesterly flow) were stronger than those on the east side (under southeasterly flow). This is inferred from the two wind speed traces recorded at Grand Isle, Louisiana and Dauphin Island, Alabama (Fig. 2 and Fig. 3). Danny passed within 20 km of both stations. Wind speeds were higher following the passage of the center of Danny as compared with the wind speed just before passage. Doppler radar velocities show an asymmetric wind field developed after 0900 UTC on July 19, just as the center of circulation entered Mobile Bay.

c. Coastal thermal gradients

Analyses of the surface virtual temperature (Tv) field are shown in Figs. 4a-4e. It is clear from these plots the idealized warn core of air was absent from the storm. Instead, a temperature gradient of 5C/50 km or more was found along the coastline from Pensacola to Apalachicola, Florida by 1200 UTC on the 19th. This was approximately three hours after the hurricane entered the bay. The coastal temperature gradient was not found 12 hr earlier, and the warm oceanic air (Tv > 32C) at 1200 UTC was not found upstream at 0000 UTC (Fig. 4a). Thus, an apparent warming of the oceanic air occurred as the circulation of Danny approached the coastline. Note the warming occurred during the overnight hours when solar insolation was not a factor. Hence, it is likely that diabatic heating, possible latent heat release and/or fluxes of sensible and latent heat from the ocean, contributed to the warming (Kong 1998).

As the day progressed, hurricane Danny continued to exhibit a baroclinic structure near the surface. However, the spatial pattern of the isotherms changed. A cool pool of air began to emerge over and to the west of the circulation center (Fig. 4d). This corresponded to the development of the western rainband. The cool pool persisted as Danny began moving eastward during the evening. Evaporative cooling of rain and sensible heating of the air around the circulation may explain the change in the spatial pattern of the isotherms.

3. Discussion

From the above analyses, it is evident that observed thermal, wind, and precipitation patterns indicate a character change in the structure and dynamics of hurricane Danny at landfall. With the presence of the coastal thermal gradient, we invoke the idea of a sea breeze circulation. In this particular case, the oceanic air was warmer than the continental air, implying an induction of a land breeze perpendicular to the coastline. The presence of this asymmetric circulation is supported by the asymmetric wind distribution around the center of Danny, where the northwesterly winds to the west of the center are stronger than the southeasterly winds to the east. The motion of the hurricane cannot explain this asymmetric wind distribution since the storm is stationary over the bay. The coastal thermal gradient and land breeze circulation would, however, suggest slantwise ascent of the warm oceanic air to the east of Danny, similar to the slantwise ascent of the warm oceanic air to the east of Danny, similar to the slantwise ascent over the warm front to the east of a mid-latitude cyclone. The collapse of the precipitation pattern over this region of the storm does not support this claim.

Since we acknowledge that changes in atmospheric motions are fundamentally driven by horizontal thermal gradients, this study emphasizes the importance of baroclinity of a hurricane in influencing the wind field and forward motion. Plausible processes of concern are latent heat release, evaporative cooling, sensible and latent heat fluxes from the ocean and land surfaces. The land-sea contrast can lead to spatial differential heating, and thus can have an appreciable impact on the structure and motion of a slow-moving landfalling hurricane. Hurricane Danny at landfall can be described as a hybrid cyclone, exhibiting both tropical and extra-tropical characteristics. In the case of tornado-producing cyclones over Florida, Hagemeyer (1997, 1999) notes that hybrid cyclones have tropical characteristics at low levels and extra-tropical influences aloft (e.g., a shortwave at 500 mb). This is in contrast to hurricane Danny which exhibited baroclinic boundary layer structure, which may have contributed to the longevity of its remnants over the southeast.

4. Forecast implications

The investigation of tropical cyclones at landfall is one of the highest NWS research priorities. In the case of Danny, 40 in of rain fell over Mobile Bay in just a few hours. If that had occurred a few miles west, over downtown Mobile, the effects of this otherwise minimal hurricane likely would have been considerably worse - in terms of flood damage, if not deaths as well. A better understanding of why storms behave the way they do upon landfall, with emphasis on very small scale features such as described in this brief study, should help forecasters better anticipate the

potential for localized heavy rain, tropical storm-induced tornadoes, and other damaging effects associated with storm landfall.

Acknowledgments

A version of this paper was presented at the AMS 23rd Conf. On Hurricanes and Tropical Meteorology in Dallas, Texas. The work was part of an M.S. thesis in the Department of Meteorology at Florida State University. Special thanks go to Jeffrey Medlin (SOO, NWSO Mobile) for a review of an earlier draft. This research was supported through the NOAA/NWS Cooperative Institute for Tropical Meteorology at FSU.

References

Blackwell, K.G. and J.M. Medlin, 1999: Doppler-observed eyewall replacement cycles in hurricane Danny near the Alabama coast and corresponding fluctuations in storm intensity, organization, and precipitation. Preprints of the 23rd Conference on Hurricanes and Tropical Meteorology, Dallas, Texas, 321-324.

Dutton, J.A., 1986: The Ceaseless Wind: An Introduction to the theory of atmospheric motion. Dover, New York, 617 pp.

Edwards, R., 1999: Tornado production by exiting tropical cyclones. Preprints of the 23rd Conference on Hurricanes and Tropical Meteorology, Dallas, Texas, 485-488.

Hagemeyer, B.C., 1997: Peninsular Florida tornado outbreaks. Wea. Forecasting, 12, 399-427.

Hagemeyer, B.C., 1999: El Niño and significant tropical and hybrid cyclone tornado events in Florida. Preprints of the 23rd Conference on Hurricanes and Tropical Meteorology, Dallas, Texas 415-418.

Kong, K.Y., 1998: Landfall and post-landfall structure of hurricane Danny and the possibility of retrieving vertical velocity profiles from rawinsonde soundings. M.S. thesis, Department of Meteorology, Florida State University, 92 pp.

Medlin, J.M. and K.G. Blackwell, 1999: A persistent supercell associated with prolonged excessive rainfall within hurricane Danny's eyewall over Mobile Bay. Preprints of the 23rd Conference on Hurricanes and Tropical Meteorology, Dallas, Texas, 196-199.