May 21 - July 30
How Various Modes of Communication Impacted Sheltering Decisions of Lee County, Alabama, Tornado Survivors
Elizabeth Leslie, Daphne LaDue, Lara Mayeux, and Jayelene Bryant
What is already known:
- In the United States, 70% of all tornado fatalities occur in a residential structure; the Southeast United States sees a higher rate than the national average.
- Known factors that affect vulnerability in the Southeast US include a higher proportion of mobile and manufactured homes, higher population density throughout rural areas, risk perception and decision making of members of the public.
- The pilot of a new study to assess structural engineering combined with survivor stories provided the opportunity to study how survivors’ severe weather warning knowledge and communication affected sheltering decisions in the 03 March 2019 tornado.
What this study adds:
- Interviews of direct survivors of the 03 March 2019 Lee County, Alabama EF4 tornado were inductively coded to understand knowledge and communication prior to taking shelter.
- People at 24 of 27 homesites had more than one type of communication mode before they took action. Nearly as many saw, heard, or felt the tornado before seeking shelter.
- These results confirm the importance of empirically identifying the best sheltering options in various home types for a larger, two-year study.
Abstract:
On 03 March 2019 an EF4 tornado struck Lee County, Alabama killing 23 people. Shortly thereafter a team of social scientists and engineers traveled to the damage path to pilot a protocol for an upcoming two-year study to combine interviews of direct survivors with the engineering assessment and larger wind context of residential structures. In the United States, 70% of all tornado fatalities occur in a residential structure; the Southeast United States sees a higher rate than the national average. Known factors that affect vulnerability include a higher proportion of mobile and manufactured homes, higher population density throughout rural areas, and lower or non-personalized perception of the risk of violent tornadoes. This paper focuses on survivors’ knowledge and the communications they received prior to making sheltering decisions. On two trips, 38 participants were interviewed at 27 homesites. These interviews were transcribed and coded both inductively and deductively for communication modes. The coding was aimed at learning how survivors knew about the tornado before it struck, if they did. The resulting major communication mode codes were the chance of tornadoes, TV coverage, a friend or relative, phone alerts, outdoor warning device, and weather radio. Of these 27 homesites, it was found that people at 24 of them had more than one type of communication mode before they took action. In addition, 23 of the 27 homesites had a non-human source: the tornado itself. Participants at all but two homesites sheltered in place; those two did not seek shelter, unaware of the tornado.