Thinking outside the polygon: a study of tornado warning perception outside of warning polygon bounds

When the National Weather Service (NWS) issues a tornado warning, the alert is rapidly and widely disseminated to individuals in the general area of the warning. Historically, the assumption has been that a false-negative warning perception (i.e., when someone located within a warning polygon does not believe they have received a tornado warning) carries a higher cost than a false-positive warning perception (i.e., when someone located outside the warning area believes they have received a warning). While many studies investigate tornado warning false alarms (i.e., when the NWS issues a tornado warning, but a tornado does not actually occur), less work focuses on studying individuals outside of the warning polygon bounds who believe they received a warning (i.e., false-positive perceptions). This work attempts to quantify the occurrence of false-positive perceptions and possible factors associated with the rate of occurrence. Following two separate storm events, Oklahomans were asked whether they perceived a tornado warning. Their geolocated responses were then compared to issued warning polygons. Individuals closer to tornado warnings or within a different type of warning (e.g., a severe thunderstorm warning) are more likely to report a false-positive perception than those farther away or outside of other hazard warnings. Further work is needed to understand the rate of false-positive perceptions across different hazards and how this may influence warning response and trust in the National Weather Service.