Don't Panic

+ See all authors and affiliations

Science  02 Dec 2011:
Vol. 334, Issue 6060, pp. 1183
DOI: 10.1126/science.334.6060.1183-a

The 2011 Tohoku-oki earthquake off the coast of Japan was the fifth earthquake since 2004 of magnitude 8.5 or greater (the others include the 2010 Maule earthquake off the coast of Chile and three off the island of Sumatra in Indonesia). This global recurrence rate is certainly higher than historical records indicate—the last major earthquake before these occurred in 1965 off the coast of Alaska, USA. In the past decade, several smaller-magnitude earthquakes with large numbers of fatalities in Haiti, China, Pakistan, and elsewhere have added fuel to the speculation that Earth is experiencing an era of increased seismicity, and therefore, global risk from future large earthquakes has increased. But could this simply be the result of a random variability of earthquakes across many unrelated faults and different types of plate boundaries? Michael ran three statistical tests using the record of magnitude 7.0 earthquakes since 1900 to resolve this dilemma. The tests rule out any statistically significant clustering and suggest that global triggering of large earthquakes by other earthquakes, except for aftershocks, is not occurring. Recent seismicity can be described by random and high variability of low-rate events within a Poisson process rather than clusters of related events. Global seismic hazard estimates should therefore not be adjusted to account only for the recent past; seismic risk should remain calculated based on the entire earthquake record, which, in some cases, stretches back thousands of years.

Geophys. Res. Lett. 38, L21301 (2011).

Navigate This Article