Deep Jiggles with Distant Triggers

tectonic tremor at ParkfieldThe tiny ranching town of Parkfield is California's seismic playground for scientists. This instrument package sits in Turkey Flat. Photos by Andrew Alden.

Besides earthquakes, there is a whole spectrum of energetic activity in the solid Earth. Thankfully, most of it doesn't disturb anyone's sleep. After a century of focus on jolts and jiggles, earthquake scientists have begun turning their attention to these more subtle signals. And California is one of the prime laboratories for this research.

Most big earthquakes happen in the mid-crust in a zone between 10 and 20 kilometers deep, where rocks are strongest. Above this zone, rocks are cold and brittle and tend to crack; below, they are hot and ductile and tend to stretch. Menlo Park seismologists looking at the deeper, ductile crust have put a new piece of the great puzzle into place this week.

Earthquakes of the classic type—cracks in the ground, alarums and mayhem in the human sphere—are only the best known type of seismic activity. They center around the brittle–ductile transition zone, but can be found from very near the surface down to almost 100 kilometers deep, if tectonic forces have put cold, brittle rock down there. (Deep earthquakes, which occur down to almost 700 kilometers, are a separate species.)


The new types of seismic events are small and slow murmurings within the deep crust. They involve motions that are gentler and at much lower frequencies than typical earthquakes. Think of them as less like the crisp cracking of a baguette crust and more like the quiet ripping of the bread inside.

Different varieties of these slow events have been discovered by different people in the last two decades. Their names include tremor, slow earthquakes, episodic slip events, transient creep and so on. They are hard to detect and difficult for seismologists to model. But that's what they used to say about ordinary earthquakes, and I'm sure that we will tame these creatures too. I suspect that they will eventually align in a kind of spectrum with names akin to the colors we designate in the blurry bands of the rainbow.

This week's news centers on the lively research topic of triggering: Do distant earthquakes set off local events as their seismic waves roll through? It seems obvious that they would, but science looks for proof before accepting even what seems obvious. Triggering was first accepted when the 1992 Landers earthquake in Southern California was shown to set off small quakes all the way out to Yellowstone, in Wyoming. The effects are subtle and of scientific rather than engineering interest.

The San Andreas fault is a laboratory for slow-event research; the area around Parkfield, east of Paso Robles, has been intensively instrumented since the 1970s. Recently, persistent clusters of tremor have been mapped there at depths below the earthquake zone. A paper by U.S. Geological Survey seismologist David Shelly and others published this week in Nature Geoscience notes that these ticklish spots of "ambient tremor" are sensitive to large, distant earthquakes in the right circumstances.

chile triggered tremor
Frame from Supplementary Movie 3, Shelly et al., Nature Geoscience doi:10.1038/ngeo1141. North-south profile of San Andreas fault; shaded zone and star represent slip in and hypocenter of the 28 Sept 2008 Parkfield earthquake; blue dots are earthquake events, crosses are ambient tremor locations; depth grid is 10-km intervals.

Here's an example grabbed from a Quicktime movie file showing these ambient tremor spots "lighting up" after passage of the surface waves from the February 27 2010 Chile earthquake (magnitude 8.8). Surface waves are the strongest, most damaging and slowest-moving vibrations of an earthquake; they are what makes the entire planet reverberate as long as weeks afterward. What interested the researchers about this triggered tremor is that it lasted long after the triggering waves had passed. Something was moving very slowly after the trigger left, something that continued to set off tremor until the stresses finally dissipated.

What moves slowest of all the new seismic slow events? It is creep, which I described in last week's post. Creep is known to be variable—some parts of a fault have it while others don't; its speed also varies in different locations. It's known to change speed, too. In fact, you might think of it as just an extremely slow earthquake with a frequency of years instead of seconds.

Shelly and his coauthors therefore suggest a new kind of triggered event to go along with triggered earthquakes and triggered tremor: triggered creep.

35.899 -120.433