Over-Alerting: Episode 4

GEOGRAPHY

In our research on over-alerting, we found scholars who wrote about the issues of mis-located alerts, polygons, and how geography affects public perceptions of message relevance. 

In the paper by Sara McBride, Ann Bostrom, et al on earthquake early warnings, they pointed to mis-located alerts as a risk related to technology - that is, when technological issues result in alerts going to places where people don't feel shaking. They can be perceived as over-alerting because the alerts go beyond the boundaries of the threat.

In the paper by Trainor, Nagele, Philips, and Scott, on tornado warnings, they described perceptions of false alerts related to their location in the warning polygon. Receiving an alert in a location where you don't personally experience a tornado, does not mean it was false for everyone. In fact, it may not be false at all - but it did include more people than all who were directly affected.

In our interviews and surveys with emergency managers and the public, they also described geographical issues.

Geographical over-alerting can occur when there are 1) technological problems (such as a WEA being issued far beyond the area of threat but within the vicinity of a cell-tower) as well as 2) issues related to the boundaries selected by the sender (such as an entire state being alerted to an event).

Both of these types of over-alerting touch on message RELEVANCY because they require a message receiver to determine whether the threat relates to them and if they need to take an action to protect themselves or others. 

Perhaps one of the clearest recent examples of relevancy related to geography comes from the January 2025 wildfires in L.A. When a WEA got "stuck" in the system, it echoed beyond the original polygon where it was sent, alerting people in different parts of the county for many hours after the alert was cancelled. This was a failure of technology that resulted in people being notified to a threat that was not relevant to them. 

While there have been improvements to WEA related to location-based geofencing which increases geo-targeting capabilities, gaps remain in making this consistent across locations, carriers, and device users. There is also a recognition by the FCC that statewide alerting contributes to perceptions of over-alerting, resulting in their recommendation to allow some alerts to be sent silently (a workaround which will result in new problems). 

In the meantime, what is the solution? There is ONE strategy that can be applied to every message that is issued: Write complete messages that clearly indicates the location of threat.