7B.1
Value of Improved Hurricane Warnings: Risk Information and Factors Affecting Stated Preferences

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Wednesday, 7 January 2015: 10:30 AM
226AB (Phoenix Convention Center - West and North Buildings)
Jeffrey K. Lazo, NCAR, Boulder, CO; and A. Bostrom, R. E. Morss, J. L. Demuth, and H. Lazrus

Hurricanes Katrina in 2005 and Sandy in 2012 exemplify the significant and recurring risk posed by hurricanes. How people receive, understand, and respond to hurricane information can lead to harmful behaviors such as failing to evacuate. People's perceptions of hurricane risk are multi-dimensional and influenced by a variety of factors, from prior experience to vulnerability to risk preferences. As part of our research examining how extreme weather warnings are communicated, obtained, interpreted, and used in decision making by diverse participants in the warning and decision process we conducted a survey of hurricane vulnerable households in Florida and Texas. We report here on analysis of data collected in an online survey using Knowledge Networks KnowledgePanel in May 2012. 808 respondents completed the survey (460 in Miami-Dade and 348 in Galveston-Houston), for a 61.6% response rate. Initial results show that respondents' stated likelihood of evacuation is contingent on information received, as well as demographic and contextual differences, with evacuation more likely on receipt of an evacuation order than after seeing a threatening forecast, higher for women than men, and higher in Houston-Galveston than in Miami, for example. Controlling for other factors, stated evacuation intentions depend on trust in the information received, motives for evacuation, and perceived barriers to evacuation, but do not appear to differ by cultural ideology. We report also on a dichotomous choice experiment for programs to improve hurricane warning information across different attributes of that information including landfall location, wind speed, inland flooding risks, and storm surge information, and controlling for information recipient characteristics and perceptions. Findings identify possible improvements to hurricane risk communication to reduce ineffective protective responses, and for research to identify and resolve gaps in the risk communication process.