PROJECTING RESIDENTIAL TORNADO EXPOSURE RATES THROUGH ARCMAP DIGITIZATION TECHNIQUES AND THE EMPLOYMENT OF BOTH U.S. CENSUS AND MIGRATION PATTERN DATA

Multiple tornado studies in the past have employed U.S. Census blocks on Arcmap software as a means of estimating total numbers of individuals who may be directly impacted by a tornado event. This process has consisted of mapping tornado damage paths with U.S. census block polygons on Arcmap software. While U.S. census blocks provide total numbers of homes and population counts for a given geographic area, they do not capture the true spatial patterns of individual housing structures populating a given area. Given that tornadoes are relatively small-scale geographic features, the intersection of census blocks with synthetic tornado damage paths provide limited degrees of spatial accuracy when compiling estimates of homes that may be exposed to such events. This study outlines a proposed methodology intended to allow more precise estimates to be gauged when accounting for total numbers of homes and individuals that may be exposed to various small-scale natural disaster events. The process first consists of manually digitizing individual building footprints through satellite imagery on Arcmap software to account for actual spatial locations of residential structures; the digitization process simply requires the creation of individual residential centroid point features, which represent the mean location of each structure. Each centroid point is then spatially joined with data stemming from the respective U.S. census block polygon the centroid falls underneath. Lastly, to better gauge the likelihood of individuals being home during the time of a natural disaster event, data stemming from a national migration survey is incorporated into this study; the migration data is comprised of aggregated daily business activities stemming from over 9,000 interview subjects. The intersection of a synthetic tornado damage path with individual centroids (representing homes) consisting of census and migration statistics may ultimately yield more precise numbers of residents who may be exposed to a hypothetical tornado event.