Graduation Year


Document Type




Degree Name

Master of Science (M.S.)

Degree Granting Department

Geography, Environment and Planning

Major Professor

Jennifer Collins, Ph.D.

Committee Member

Joni Downs, Ph.D.

Committee Member

Margarethe Kusenbach, Ph.D.


Disasters, Hazards, Tornadoes, Vulnerabilities


The purpose of this study is to model and determine significant predictors of tornado death index values, and to investigate these significant predictors and what makes people vulnerable to tornado fatalities through expert interviews. This study also provides an understanding of the study participant’s perceptions of their county’s vulnerability to tornado fatality and demonstrates a true integration of methods and fields by studying geographic, meteorological, and sociological phenomena by use of quantitative and qualitative methods. The study consists of two parts: 1) A quantitative exploration of variables hypothesized to predict Tornado Death Index (TDI) values, 2) A qualitative investigation to further understand what leads to higher tornado fatalities. For the quantitative portion of the study descriptive statistics and multiple linear regressions were run on TDI values. It was predicted that several tornado characteristic, demographic, housing type and characteristic, religious, region, rural vs. urban, and potential casualty variables were significant predictors of TDI values. For the qualitative portion of the study a highest order emergency manager was interviewed, coding was done and themes, sub-themes, and categories emerged, and quotes that demonstrated the themes and categories were examined.

Overall, significant predictor variables of TDI are tornado frequency, tornado width, ages 35-44, percent born in the Northeast, percent rural housing units, and potential casualties. As tornado width, and percent of rural housing units increases TDI increases (positive relationship), whereas as tornado frequency, ages 35-44, being born in the Northeast, and potential casualty increases TDI decreases (negative relationship). In the interview, age, cultural beliefs, and mobility challenges were found to increase risk to tornado fatality. It was also suggested that differences in tornados may exist between the Midwest and the South in terms of tornado development, duration, and warning lead-times. Finally, vulnerability can be reduced by educating the public, and reaching out to vulnerable populations and their caregivers.