AIRCRAFT OBSERVATIONS OF HURRICANES

Principal Investigator: H.E. Willoughby
Collaborating scientist(s):
M. E. Rahn
Objective: To understand the dynamics of tropical cyclone intensity change and motion.
Rationale: Since hurricanes are too small to be resolved with conventional instruments, aircraft and radar are the tools of choice to observe the mesoscale and convective scale phenomena that control their evolution and motion.
Method: Analysis of ground based radar images, flight--level aircraft data, and satellite images to reconstruct Hurricanes on the mesoscale and vortex scale.
Accomplishment: Tropical Storm Bob formed east of the Bahamas on 16 August 1991 and became a hurricane the next day. It tracked northward and then northeastward, passed east of Cape Hatteras, made landfall in Rhode Island, returned to the North Atlantic over Newfoundland, and eventually dissipated at sea. In the time between Bob's formation and landfall, USAF reconnaissance aircraft flew 12 sorties with a total of 84 entrances or exits of the eye, to which NOAA research aircraft added 8 penetrations or exits during 2 sorties. Although these observations are supported by PPI radar images only during the NOAA flights and when Bob was within range of the Hatteras radar, they depict Bob's vortex--scale evolution in unprecedented detail.

Between 1100 UT on 17 August and 1600 UT on 19 August, Bob formed 5 convective rings ( Figure), denoted by Roman numerals. Ring I was the initial eyewall. Ring II surrounded ring I at the initial time, but dissipated at 0000 UT on 18 August without replacing it. Around ring I and the remnants of ring II, rings III and IV formed: respectively at 2100 UT, 3 h before ring II dissipated, and at 0000 UT, the time of ring II's dissipation. Ring III contracted and replaced ring I at 0300 on 19 August, simultaneously with ring V's formation around all previous rings. At the time of the eyewall replacement the strongest wind anywhere in the hurricane was 54 m/s. An aircraft radar image from 0500 UT shows clearly ring V surrounding ring III and the last traces of ring IV. At landfall, 1800 UT on 19 August, rings III, IV, and V remained, with ring V apparently about to replace ring III.

The minimum sea--level pressure (MSLP) fell from 996 hPa to 950 hPa during this time. The potential minimum sea-level pressure computed from the sea-surface temperature was below 940 hPa until 00 UT on 20 August. Then it rose to above 1000 hPa by 1600 UT on the 20th as Bob passed over cold water north of the Gulf Stream. The storm intensified slowly throughout the time its actual minimum pressure was above the potential value. The only interruption occurred between 0000 and 0600 UT on 18 August, when the MSLP remained constant at 980 mb, coincident with dissipation of ring II. Bob continued its general intensifying trend until an hour or two after the storm moved over colder water.

In summary, Bob formed 5 convective rings, one dissipated, one was replaced by a surrounding ring, and three survived to landfall, although one of these would probably have been lost to replacement had landfall not intervened. The ring interactions showed no strong correlation with intensity changes. This departure from previous experience probably occurred because Bob was less intense than the other documented cases and had so many convective rings that none of them came to dominate the dynamics.

Key reference:
Willoughby, H. E., and P. G. Black, 1995: Hurricane Andrew in Florida: Dynamics of a Disaster. Bulletin Amer. Meteor. Soc., 77, 543-549.

Last modified: 8/9/96