A geophysicist ponders the mysteries of intraplate earthquakes.

During my first semester at college, I attended a lecture describing plate tectonics, and immediately knew that geophysics would be my major and hopefully my career. Subsequent lectures, textbooks, journal articles and, later, my own research educated me about how elegantly plate tectonics explains the processes that control the locations of most earthquakes.

However, some of the largest-known North American earthquakes — including those of 1811–12 famed for ringing church bells in Boston and changing the course of the Mississippi River — are associated with the New Madrid Seismic Zone (NMSZ) in the Southern and Midwestern United States, far from known plate boundaries. So what causes these events? Eric Calais of Purdue University in Indiana and Seth Stein at Northwestern University in Illinois present some surprising results from an examination of Global Positioning System (GPS) data from the region (E. Calais and S. Stein Science 323, 1442; 2009).

Previous studies found that the NMSZ was moving at a different rate and in a different direction from the North American Plate, implying that strain would steadily accumulate until released by a large-magnitude earthquake. But, incorporating three years' worth of extra GPS data, Calais and Stein found motions indistinguishable from those of the North American Plate, corresponding to extremely low strain rates. It is not clear what the underlying processes causing the NMSZ earthquakes are. Is strain accumulation variable over time in intraplate settings? What are the implications for hazard prediction?

The results leave me perplexed, but oddly comforted — there are plenty of mysteries left for the next generation of geophysicists. And perhaps one day a theory will elegantly explain intraplate seismicity, just as plate tectonics did for interplate seismicity.

Discuss this paper at http://blogs.nature.com/nature/journalclub