Iconic Figures: Energy Release by Earthquakes
Scientists like stories with lots of pictures. In fact, an awful lot of what goes on during talks is more like a pre-reading child telling stories about the pictures in a picture-book than a grown-up engaging with the textual content of an article.
Some figures show up over and over again, and not just because a particular research group is really working the lecture circuit. A handful of such images, often NASA glamour shots of small satellites or asteroids, happen to be the best available images of a particular subject and are thrown up on people’s first and last slides as a branding tool. Typically, though, figures are picked up and syndicated across scienceland not because they’re pretty, but because they provide a concise summary of relevant data on a topic that is new and/or controversial. In other words, they’re a handy encapsulation of the research zeitgeist.
So this is the first in what I hope to be an occasional series of posts, each focused on explaining a particular image I’ve been seeing a lot of lately. Here’s what I’ve picked out to start:
The energy in any given earthquake can be used to deform and fracture rocks surrounding the fault, sent out as seismic waves, or turned into heat. The basic question posed by the figure above is, do large earthquakes release energy in the same way as small ones?
The traditional answer is “pretty much, yeah”. However, a raft of studies beginning in the 90s showed evidence to the contrary, and wacky science hijinks ensued. Recent research is more or less evenly divided between work that supports self-similar earthquake scaling, and work that undermines it.
The figure was first published by Satoshi Ide and Greg Beroza in 2001 (Geophysical Research Letters paper #2001GL013106). It summarizes data on seismic energy release reported by several previous studies, reanalyzed by Ide and Beroza to account for the limited frequency bandwidth captured by most seismometers. The horizontal axis shows the size of an earthquake, as represented by a quantity known as seismic moment; the top horizontal axis shows the equivalent earthquake magnitude. Note the logarithmic scale! The data displayed here spans 17 orders of magnitude, from earthquakes the size of a raindrop to earthquakes the size of the Kuiper belt. The vertical axis shows the amount of energy released in seismic waves. It’s normalized by the size of the quake, so if large and small earthquakes are the same, we would expect the data to fall on a horizontal line. Although there is quite a bit of variation, it appears from this figure that the overall trend is, indeed, horizontal, implying that a magnitude 3 earthquake is just a miniature version of a magnitude 7.
For a variety of technical reasons, the amount of sesimic energy released by an earthquake is a total bitch to determine. This is what has kept Ide and Beroza’s figure in business: if you think you’ve found a better way to determine that energy release, you put up a new version of it, with your data in red.
Being able to correctly translate our observations and theories from small earthquakes to large ones will make lab experiments more meaningful, observational results more applicable, and kitchen sinks shinier. Large earthquakes are impossible to replicate in the lab, but we can while away weeks smashing up chunks of granite under controlled conditions. Plus, nature provides us with a wealth of small earthquakes to observe while we wait for the Big One. The benefits of knowing how to scale up correctly are manifold. Some seismologists maintain that the solution to the so-called earthquake scaling problem will make us all peanut butter banana sandwiches after every earthquake, and trim off the crusts. Or maybe that’s just me.