earch and Applications Workshop
S98-8
THE FLORIDA TORNADOES: RESPONSE AND MITIGATION CONUNDRUM
Moderator: Don Wernly, NOAA/National Weather Service
Recorder: Dale Lehman, Woodward-Clyde
Discussants: Bart Hagemeyer, National Weather Service;
Joanne McGlown, University of Alabama, Birmingham;
Walter Peacock, Florida International University
In February and April of 1998, severe tornadoes struck with devastating effects in Florida and Alabama. Both events had advance and accurate National Weather Service (NWS) warnings and forecasts. However, many fatalities were still experienced from both events. In Florida, a series of devastating tornadoes struck the Orlando and Kissimmee areas late at night and early in the morning on February 22 and 23. The tornadoes varied in intensity from F2 to F4. Seven tornadoes touched down with limited tracks and extent of damage. There were a total of 42 fatalities; 40 of them were residents of manufactured homes. The NWS used its tornado warning decision support system with its Doppler Radar that provided 100% detection accuracy for this event. The NWS warnings provided an average lead time of 15 minutes before touchdown, and were issued an average of 23 minutes before any fatalities occurred. Most of the news media provided coverage during their 11:00 p.m. broadcasts and warned that a dangerous situation still existed.
Jefferson County, Alabama, experienced a severe tornado at about 8:00 p.m. on April 8, 1998. There were 32 fatalities from the storm; only three of these were residents of manufactured homes. The tornado was on the ground for about 30 minutes and covered approximately 30.6 miles. It varied in intensity from F1 to F5. Most people were watching television at the time and were aware of the warning. Warnings were issued via personal telephone calls, NOAA weather radio, television, and sirens. One television station had continuous weather coverage with detailed reports of the tornado's track and areas subject to imminent threat. Most of the people that died had emergency plans and were in their "safe" places when the tornado struck. It was felt that the number of fatalities was a function of the severity of the tornado and not due to ineffectual warnings.
In the post event assessment for these two events, several key items were identified that could have led to the number of fatalities that occurred. The following are these items and some lessons learned:
- Most people knew that there was a tornado watch in effect based on the earlier warnings and watches issued through the news media. However, the Florida event occurred on a Sunday night. Work schedules and school contributed to people going to bed prior to the late-night warnings and touch down of the tornadoes. Consequently, a vast majority of people did not receive the later warnings.
- Florida is a vacation area with a transient population. The warnings for the Florida event were issued at the county level and several individuals may not have had local geographic knowledge to aid them in understanding the imminent threat in their particular location.
- Warnings issued should include more definitive information beyond just the county being affected. Clearer geographic information such as landmarks and road crossings could be more effective in communicating areas in imminent danger. Additionally, instructions on where to go and what to do should be included where possible.
- Manufactured homes provide little to no protection during severe tornadoes. Consequently, emergency plans and "safe" facilities for evacuation are critical in mobile home communities. Most people in the Florida event did not have emergency plans for tornadoes. The Florida assessment team identified a number of bath houses, laundries, etc., that were of masonry construction and could have been used by manufactured home residents given the warning lead time.
- There is a strong need for reasonably priced shelters in manufactured home parks and for low cost safe rooms in on-site constructed homes. Texas Tech University in conjunction with the American Association of Wind Engineers, under funding from FEMA, has developed an in-residence shelter design that is economical.
- NOAA weather radio could have helped in getting the warning out to residents. It was noted that watch boxes and crawls on television screens were not as effective at conveying the warnings as news break-ins.
- There was some apathy detected in both communities based on severe weather systems that were experienced earlier.
- Individuals watching cable channels or non-local stations did not receive warnings. An ability to provide for cable overrides to break in with the warnings could have more effective.
- Communities prone to tornadoes should consider implementing a program to communicate what to do for new residents (e.g., include an information packet that is given to new residents when they have their telephone service activated).
Return
to Hazards Center Home Page
Return to
Index of 1998 Session Summaries
August 28, 1998
hazctr@colorado.edu