How Long Do Aftershocks Last?

The classroom treatment of aftershocks is fairly simple: An earthquake hits, and then a series of aftershocks occurs that is well summarized by three mathematical expressions called the Gutenberg-Richter relation, Bath's law and Omori's law. The first two relate the aftershocks to the mainshock in terms of number and size (G-R says their numbers rise by an order of magnitude with each unit fall in seismic magnitude, and Bath says the largest aftershock averages about 1.2 magnitude units smaller). Omori's law describes how aftershocks tail off to a steady background rate as the reciprocal of time (that is, 10 years later aftershocks occur at 1/10 the rate).

The great majority of earthquakes and their aftershocks are easy to deal with because they occur near plate boundaries. There the seismic background rate is high and aftershock sequences therefore are short, around 10 years (because the background noise swamps their long tails). But away from plate boundaries the background noise is very low, and also the physics within plates appears to favor extra-long aftershock sequences. In sum, Seth Stein and Mian Liu argue in today's Nature, in places like the New Madrid earthquake zone in the central United States, the seismicity we see today could be aftershocks from earthquakes hundreds of years ago. They wouldn't represent the buildup to the next big one.

If we don't recognize these as aftershocks, they will mess up our models for determining earthquake hazards. To places like Memphis, overstrengthening buildings based on a poor New Madrid zone model is wasted money. In more complex plate settings like China, Stein and Liu argue, even distinguishing these aftershocks would only remove one veil from the mysterious and subtle patterns of deadly quakes like the 2008 Sichuan event.

More:
About aftershocks
Earthquake basics
U.S. earthquake hazard map
China earthquake hazard map