CEHC Professor Examines Earthquake Post-Alert Messaging Following False Alarm
By Mike Nolan
ALBANY, N.Y. (Jan. 26, 2023) — In May 2020, residents of Ridgecrest, Calif. received an emergency text alert from ShakeAlert, an early warning system for the West Coast operated by the U.S. Geological Survey (USGS), telling them that a significant earthquake was detected and to take cover. However, the shaking never followed. It was a technical error.
The false alert, the first of its kind in the U.S., was followed shortly after by a post-alert message, indicating that the message was “cancelled” and “requires investigation.” Early-warning alerts are only meant to be sent to people who could feel the shaking of a magnitude 5 quake or greater.
Jeanette Sutton, an associate professor at the College of Emergency Preparedness, Homeland Security and Cybersecurity (CEHC), studies disaster and risk communication, with a focus on public alerts and warnings issued via mobile devices and social media.
The incident provided Sutton with a unique opportunity to learn more about the importance of early warning earthquake and post-alert messaging.
“The Ridgecrest area had experienced a significant series of earthquake events just 10 months prior to the false alert message,” said Sutton. “People were really on edge from that experience. When they received the false alert, many sheltered for long periods of time and were confused, even after the post-alert message was received.
“There was a real mismatch of understanding between what the alert system was, how it works and if it was truly safe to return to normalcy.”
Earthquake Alert Messaging
To better understand the public perceptions of the incident, Sutton and a team of researchers, funded by the National Science Foundation, conducted individual interviews with 40 people in the Ridgecrest community. This was followed by a series of focus groups in Southern California with individuals who were not impacted by the false alert.
The conversations were focused heavily on the post-alert message to learn about how it was perceived and what information people needed most following a false earthquake alert. The research team also shared alternative post-alert message options with the focus group.
“There is very limited research on public perceptions and responses to messages in scenarios like this one. We were able to target both individuals that were directly impacted by the earthquake false alert and those in nearby areas that could be at any time in the future,” Sutton said.
“We believe our discussions can serve as a valuable resource to help with future messaging for ShakeAlert and other early warning systems.”
Among the key takeaways included that people both with and without previous earthquake experience expressed confusion about the incident, specifically on why the false alert went out and what was being “investigated.”
The post-alert message also complimented those who acted in response to the initial alert, saying “if you protected yourself, well done,” which was largely perceived negatively. Instead, the study’s participants desired an explicit statement about their current level of safety.
While one study participant explained, “There is of course no all-clear, we don't know that there's not about to be a huge quake in the area,” others said they would have preferred a message with more definitive language such as “no further action is required” or “the incident has passed.”
“The immediacy of a post-alert message suggests that prioritizing the informational and emotional needs of the public will serve as an appropriate communication strategy,” Sutton said. “Addressing concerns about what is known about safety first, followed by organizational reputation and public education second, may provide the kind of reassurance to message receivers that the hazard has been effectively managed.”
Recommendations based on the discussions were shared in this month’s International Journal of Disaster Risk Reduction.
Future Post-Alert Research
Prior to publishing the paper, Sutton presented the team’s findings during a virtual Social Science Research Symposium in October, hosted by the USGS ShakeAlert Social Science Working Group. The symposium showcased social science research on public messaging and public perceptions of earthquake early warning systems in the U.S. and abroad.
Although they were focused on earthquake messaging, Sutton said it will also be important for future research to explore how to best optimize the language of post-alert messages across a range of other hazard events.
Other research collaborators included Michele Wood of the Department of Public Health at California State University, Fullerton and recent CEHC graduates Savanah Crouch and Nicholas Waugh.