Geographic constraints on rapid comprehension of tornado warning information

Abstract

Sketch-mapping activities with in-person interviews were conducted with 65 participants in Alabama and Mississippi to explore how geographic awareness impacts risk perception for tornadoes. Geographic awareness metrics represented three variables: 1) the distance between their perceived home location and actual home location (P-A) on a blank map, 2) the correct percentage of counties (POC) labeled that bordered their home county, and 3) an assessment of the accuracy of their county shape (SHP) from their drawing. Results showed that participants who are more competent in one area of geographic awareness will also have some competency in the other measures. Accurate knowledge of tornado path direction was statistically significant with P-A distance, POC, SHP, and number of roads drawn. The demographic characteristics of participants and their geographic awareness scores were used with Principal Components Analysis and Cluster Analysis to detect which traits of participants were associated with better geographic awareness. Four distinct groups were identified with two groups having low or poor geographic awareness. The traits associated with these groups are nuanced. Of the four groups, the lowest ranking poor local geography group was characterized by the highest rate of home ownership, and was ranked second oldest, second in time lived in the county, and second in education, with higher racial and ethnic diversity than two groups. This suggests that experience, age, and education do not necessarily translate into knowing geographic information about the county or city. People with low geographic awareness are at a disadvantage when making protective action decisions during tornado warnings. More research is needed to examine the robustness of these results to determine best strategies for communicating information.