An evaluation of tornado siren coverage in Stillwater, Oklahoma: Optimal GIS methods for a spatially explicit interpretation. (March 2016)

• Record Type: Journal Article
• Authors: Mathews, Adam J.; Ellis, Emily A.
• Is Part Of: Applied geography. Volume 68(2016:Mar.)
• Page Start: 28
• Page End: 36
• Publication Date: 2016-03
• Subjects: Tornado sirens -- GIS -- Spatial analysis -- Severe weather risk -- Stillwater -- Oklahoma; Geography -- Periodicals; Human geography -- Periodicals; Human ecology -- Periodicals; 910
• DOI: 10.1016/j.apgeog.2016.01.007
• Languages: English
• ISSNs: 0143-6228

Abstract

Abstract: Severe weather occurrences, especially tornadoes, are costly to communities in terms of loss of life and property damage. Cities located within Tornado Alley are especially at risk of being affected by tornadic activity and thus have a pressing need for a warning system. The use of strategically-placed outdoor tornado sirens is a common approach to notify residents of approaching severe weather. However, it remains difficult to ensure complete coverage (i.e. every resident within an audible distance of a siren) of such a system within a city. This study uses two distinct GIS-based methodological approaches using U.S. Census population data and siren locations to assess the extent city residents are safeguarded by the existing tornado siren network in Stillwater, Oklahoma. The first approach utilizes spatially unaltered (aggregated) census data whereas the second approach proposes a disaggregation technique incorporating building polygons to improve the spatial detail. The first approach revealed that 26% of city residents are not covered by the siren network, whereas the second approach estimated that 4–14% are not covered. The second approach provided a more precise account of Stillwater's vulnerable population due to its spatially explicit nature. Graphical abstract: Highlights: A GIS disaggregation method to estimate per-building population was developed. Building population and census block population were used to assess siren coverage. Per-building population estimates supported finer scale results. 4–14% of Stillwater's population is not within an audible distance of a siren. Euclidean distances from buildings to the nearest sirens were assessed citywide.