08 November, 2009

Sunday Sensational Science

Shocking Truth About Aftershocks

We've discussed earthquakes before, and everybody's probably pretty aware of the fact that when you have an earthquake, you're probably going to have an aftershock.  Or two.  Or two dozen.  Most of us think those aftershocks will last, at most, a few days. 

New studies suggest that some aftershocks will go on - are you ready for this? - for a few centuries:
Many researchers assume that small-scale seismic activity reveals where stress is building up in the Earth’s crust — stress that can cause larger quakes in the future, says Mian Liu, a geophysicist at the University of Missouri in Columbia. However, Liu and Seth Stein of Northwestern University in Evanston, Ill., report in the Nov. 5 Nature, many moderate-sized temblors that occur far from the edges of tectonic plates could be merely the aftershocks of larger quakes that occurred along the same faults decades or even centuries ago.


Stein and Liu analyzed earthquake data gathered worldwide. For major quakes that occurred where the sides of a fault moved past each other at average rates of more than 10 millimeters per year — as the two sides of many tectonic boundaries do — aftershocks died off after a decade or so. But for faults where the sides scraped past each other at just a few millimeters per year, aftershocks lasted about 100 years, the researchers reported. The longest series of aftershocks, some which have lasted several centuries, were triggered by quakes that occurred in continental interiors along slow-moving faults.
Bet you folks in the Midwest didn't think New Madrid was sending you old news, did you?  But it certainly seems so.

Let's step back a moment and take a look at the mechanics here:
Large earthquakes are often followed by aftershocks, the result of changes in the surrounding crust brought about by the initial shock. Aftershocks are most common immediately after the main quake. As time passes and the fault recovers, they become increasingly rare. This pattern of decay in seismic activity is described by Omori's Law but Stein and Liu found that the pace of the decay is a matter of location.

At the boundaries between tectonic plates, any changes wreaked by a big quake are completely overwhelmed by the movements of the plates themselves. At around a centimetre per year, they are regular geological Ferraris. They  soon "reload" the fault, dampen the aftershocks, and return the status quo within 10 years. In the middle of continents, faults move at less than a millimetre every year. In this slow lane, things can take a century or more to return to normal after a big quake, and aftershocks stick around for that duration.

It's a tale of two faults!  Let's have a look at New Madrid, shall we?  Go ahead.  Search for photos of "New Madrid Fault."  I'll wait.

Lots of maps, not many photos, right?  That's because not a lot's going on there.  Here's the best I could do:

Ed Yong says,
Again, New Madrid proves the principle - a cluster of large earthquakes hit the area in the past thousand years, but the crust shows no sign of recent deformation according to two decades of GPS measurements. It seems that recent activity really is the legacy of centuries-old quakes, a threat that has since shut down.
In other words, not a lot going on that would show at the surface.  It's a slow, sleepy fault, despite the excitement it caused over the winter of 1811-1812.  Compare that to the San Andreas, which is bleeding obvious:

Compared to New Madrid, the San Andreas fault is a speed demon, and it shows.  There are other differences, of course - one's a transform fault where two plates are scooting past each other, the other's more of a rift type thing where North America started developing a split personality and then changed its mind(s?) - but the main thing is speed.  Cecil Turtle compared to Speedy Gonzales, shall we say.  According to the study, San Andreas locks and loads within a decade or so, leaving the aftershocks in the dust and nervous Californians waiting for the Big One.  New Madrid's still squirming around trying to get comfortable after a fairly dramatic disruption.  And every time it twitches noticeably, folks in the Midwest experience a nervous attack of their own.

The river did, after all, run backwards the last time this thing went crack.  Bound to worry folks a bit.  But according to Stein and Liu, there's nothing much to worry about - at least, not where New Madrid's concerned.  You're just in for hundreds of years of aftershocks, since the fault moves more than 100 times slower than the San Andreas.  This is good news.

And the data are beautiful:

"A number of us had suspected this," Liu said, "because many of the earthquakes we see today in the Midwest have patterns that look like aftershocks. They happen on the faults we think caused the big earthquakes in 1811 and 1812, and they've been getting smaller with time."

To test this idea, Stein and Liu used results from lab experiments on how faults in rocks work to predict that aftershocks would extend much longer on slower moving faults. They then looked at data from faults around the world and found the expected pattern. For example, aftershocks continue today from the magnitude 7.2 Hebgen Lake earthquake that shook Montana, Idaho and Wyoming 50 years ago.

"This makes sense because the Hebgen Lake fault moves faster than the New Madrid faults but slower than the San Andreas," Stein noted. "The observations and theory came together the way we like but don't always get."

This might be of some comfort to residents near the epicenter of the Hebgen Lake Quake.  Then again, it might not.  It's rather hard to feel comforted by the fact that the fault moves slower than the San Andreas, and therefore shall have aftershocks longer, when the last big quake took down a mountainside, ripped open roads, created a new lake, and left fault scarps all over the damned place, right?

And this study points to the fact that the small isn't always a foreshadow of the big:

The new results will help investigators in both understanding earthquakes in continents and trying to assess earthquake hazards there. "Until now," Liu observed, "we've mostly tried to tell where large earthquakes will happen by looking at where small ones do." That's why many scientists were surprised by the disastrous May 2008 magnitude 7.9 earthquake in Sichuan, China -- a place where there hadn't been many earthquakes in the past few hundred years.

"Predicting big quakes based on small quakes is like the 'Whack-a-mole' game -- you wait for the mole to come up where it went down," Stein said. "But we now know the big earthquakes can pop up somewhere else. Instead of just focusing on where small earthquakes happen, we need to use methods like GPS satellites and computer modeling to look for places where the earth is storing up energy for a large future earthquake. We don't see that in the Midwest today, but we want to keep looking."

Sounds like a very good idea to me.  Anything we can do to increase the chances of successful earthquake prediction could help save a lot of lives.  And it allows us to rest easier when we find out that those little temblors are just past earthquakes saying "So long, and thanks for all the fish."

No comments: