Social science and ShakeAlert, the earthquake early warning system for the West Coast of the United States

Abstract:

As the full public rollout of ShakeAlert began in October 2019 in California, with Washington and Oregon likely following in 2021, critical questions remain about how people will understand ShakeAlert, what protective actions they know to take, and what their responses will be when the system behaves differently than they might expect (e.g. false or missed alerts). Additionally, we do not yet know if people understand what the alerts mean or how the system works. Currently, there is some but not much literature about social science and ShakeAlert: a gap this project aims to fill.

To evaluate whether the alerting system has been successful in answering these key research questions, the U.S. Geological Survey, in collaboration with its university and state partners, developed a social science initiative focusing on four goals:

1. Develop understanding of where populations are currently in terms of risk perception, protective action knowledge, and basic earthquake preparedness across Washington, Oregon, and California.
2. How best to use current social science research to inform the communication, education, and outreach (CEO) program for ShakeAlert, outlined in the 2018-2020 ShakeAlert CEO Plan.
3. Develop and test an adequate alert with sound (tone), haptics (vibrations), words, and images, as well as post-alert messaging.
4. Facilitate the development and implementation of a monitoring and evaluation plan for ShakeAlert public education and communication resources.

Projects currently underway include analysis of Closed-Circuit Television (CCTV) footage from recent earthquakes in the U.S.A. to better understand human behavior during shaking, message testing, a citizen science Wireless Emergency Alert project, adding questions to the Did You Feel it? (DYFI) questionnaire on the USGS website, and an attitudes/perceptions baseline to determine levels of understanding about ShakeAlert. This important work will help us understand what perceptions and human behavior exist currently about ShakeAlert so the public aspects of the systems can be iterated to better suit the needs of our users.