LAUSR

Previous studies have shown that immediately after large earthquakes, there is a period of increased public interest. This represents a window of opportunity for science communication and disaster-relief fundraising efforts to reach more people. However, how public interest varies for different earthquakes has not been quantified systematically on a global scale. We analyze how global search interest for the term “earthquake” on Google varies following destructive earthquakes from 2004 to 2016. We find that there is a spike in search interest after destructive earthquakes followed by an exponential temporal decay. The duration and time constant of increased search interest correlate with death toll and damages but did not correlate with earthquake magnitude, estimated population exposed to very strong shaking, and number of U.S. Geological Survey (USGS) “Did You Feel It?” (DYFI) responses. Furthermore, we obtain similar time constants of increased search interest when analyzing just the U.S. search interest following destructive earthquakes outside of U.S.A., Canada, and Mexico. This suggests that a significant portion of the increased search interest comes from people who did not feel the shaking. Our observations are consistent with more destructive earthquakes receiving more media coverage which leads to a longer duration of elevated public interest in earthquakes. Of the 73 earthquakes that resulted in an increase in global search interest that fit our selection criteria, only 11 (15%) resulted in an elevated search interest of more than a week. Therefore, to take advantage of these short durations of increased public interest, science communication and disaster-relief fundraising efforts have to act promptly following devastating earthquakes. Electronic Supplement: Tables of characteristics for the 73 earthquakes included in the study as well as exponential and powerlaw parameters for the 10 earthquakes with the longest duration of increased search interest, and figures comparing N and T with earthquake magnitude, estimated population exposed to very strong shaking (modified Mercalli intensity MMI ≥ VII), and the number of U.S. Geological Survey (USGS) “Did You Feel It?” (DYFI) responses. INTRODUCTION Science website traffic (Bossu et al., 2011, 2014; Quigley and Forte, 2017), earthquake-related Twitter messages (Earle et al., 2012), and earthquake-notification mailing-list subscriptions (Schwarz, 2004) spike and decay after large earthquakes. This illustrates that immediately after an earthquake, there is a period of increased public interest during which outreach and disaster-relief fundraising efforts might reach more people. Looking at the U.S. Geological Survey (USGS) Southern California Earthquake Hazards Program’s electronic mailing lists between 2001 and 2003, Schwarz (2004) found that the subscription rate increased above background rates for approximately two weeks following major earthquakes in California. Looking at the number of visitors at four science websites during the 2010–2012 Canterbury earthquake sequence in New Zealand, Quigley and Forte (2017) found that after large earthquakes, there is an immediate increase in website traffic followed by a power-law-characterized decay. The traffic perturbation scales with earthquake magnitude and peak ground acceleration. However, we are unaware of a previous attempt to systematically quantify how public interest varies for different earthquakes on a global scale. In this study, we analyze how global search interest for the term “earthquake” on Google varies following destructive earthquakes from 2004 to 2016. We then explore if earthquake magnitude, death toll, damages, and the population size exposed to very strong shaking affect how search interest varies following different earthquakes. GLOBAL GOOGLE SEARCH INTEREST Google processes more than 3.5 billion searches per day (Sullivan, 2013). Since 2004, GoogleTrends has been providing the daily proportion of searches on Google for any term, scaled on a 0–100 range over the time interval of interest. The search interest provided is rounded off and hence consists only of integer values. In this study, we use the global search interest for the term “earthquake” provided by GoogleTrends as a measure of public interest in earthquakes. The search interest provided is not case sensitive and includes searches such as “Nepal doi: 10.1785/0220170116 Seismological Research Letters Volume XX, Number XX – 2018 1 SRL Early Edition