William Case, Junior, Computer Science, College of STEM
Dr. Awatif Amin, Johnson C. Smith University
Often, tornadoes are viewed with a great deal of mystery and thought of as somewhat unpredictable. Using information that is readily accessible to the public, this research project seeks to determine the conditions and metrics that should be analyzed to prepare for tornadoes, more accurately predict their outcomes, and subsequently build resilient communities. In addition, this project looks at the level of emphasis and resources that should be allocated based on a tornado’s Fujita rating. Our research focused on Florida, Iowa, Colorado, Nebraska, Missouri, Texas, Oklahoma, Kansas, Illinois, and Alabama. We used the data mining techniques of decision tree and linear discriminant analysis. Our sources were tornadohistoryproject.com, usa.com, and the USDA Forest Service. The research findings suggest that although the majority of the l public’s attention is focused on catastrophic tornadoes, weaker storms may produce more cumulative losses over time. In addition, our findings indicate there may be a relationship between tree canopy density and tornado formation.