11 March 2011

Japan M8.9 quake + tsunami

Posted by Callan Bentley

I’m not usually the “report on current events” type of geoblogger, but someone needs to throw a few details up ASAP for those who are interested in the details of the big Japanese earthquake and resulting tsunami that is currently crossing the Pacific Ocean.

First off: links to other geoblogs already reporting on the event: Geotripper, Paleoseismicity, Hypo-theses, Dan’s Wild, Wild Science Journal , Eruptions, Georneys, Hudson Valley Geologist, and the Landslide Blog.

Context – If the 8.9 magnitude is confirmed, this is the fifth largest earthquake ever recorded. It is claimed (by the BBC) to be the largest earthquake to strike Japan in more than 140 years.

The earthquake was located at a depth of 17 km (relatively shallow, hence its capacity to generate a tsunami) just east of Japan’s main island, Honshu. This is a near a complicated tectonic meeting place, where four plates meet (pardon my rough sketch of the boundaries):

However, key to understanding the quake is that deep trench to the east of the epicenter. This marks the spot where the Pacific Plate is “diving” down into the Earth beneath the overriding plate, the North American Plate. This subduction zone is also the cause of northern Japanese volcanism, as dewatering of the subducted slab of oceanic crust triggers melting, and the resulting magma rises to the surface to erupt.

The main earthquake’s focal mechanism (‘beach ball’ diagram taken from here and re-colored) is shown. This isn’t a “classic” reverse fault mechanism, and (given its shallow depth) may indicate a high-angle normal fault in the overriding plate (North American Plate) while the base of the plate stays locked to the underthrusting Pacific Plate at the subduction zone itself. As we saw with the 2009 L’Aquila quake in Italy, normal faulting can happen — to destructive effect — in the complex mess of rocks and faults above a subduction zone. [update] Another possibility, as Chris notes in the comments, is that that shallowly-east-dipping plane bisecting the beach ball is the fault plane — that would make this a low-angle thrust fault. But it’s got to be a shallowly-northwest-dipping fault instead. Apparently the distance of the seismograms that were used to create this beach ball may not be of sufficient resolution to determine the dip direction, given how shallow the dip is — a few degrees tilt change, and it’s dipping the other way! We’ll see as we refine our understanding of the focal mechanism as additional data come in.

[update 2] My original beach ball was mis-colored — the new version (shown above) is correct, and I thank Kim for chiming in via the comments just now as I was typing this. I apologize for the error! Such are the traumas associated with rushing to blog about an emerging issue without sufficient coffee. (Maybe this is why I’m not usually a “current affairs” geoblogger. I’m grateful for the peer review, Kim.) The diagram is now corrected in the new version, and as the data warrant, our interpretation must change.

So, for the record: The focal mechanism is a classic “center segment compression” which suggests reverse faulting, which is what we would expect at a subduction zone most of the time.

The map of historical seismicity in the area shows that today’s quake took place in an area where there have historically been many other shallow (<35 km) foci, but also deeper quakes (between 35 and 70 km depth), and some that are even deeper than that:

The yellow and green dots are the epicenters of past earthquakes with depths that correspond to the subduction zone fault complex, while today’s quake was above those (at a shallower depth). I suppose you could say it’s part of the subduction zone fault “complex” in general, but my point is that it’s not likely the “edge” of the North American Plate per se — though such an “edge” is a somewhat artificially-designated notion anyhow. The shallow depth suggests that this earthquake is a signal of deformation taking place within the North American Plate itself.

The footage of the tsunami is striking to watch — I was surprised to see some structures burst into flame as they were being carried inland on the massive wave. Here’s a screenshot from the best video I have so far seen (BBC), annotated by me:

If there is anything worse than being hit by a tsunami, it’s being hit by a flaming tsunami. You can see that at that point, we were still calling it an 8.4; the quake has since been upgraded, as is typical for really large magnitude events. There’s another astonishing gallery of images (including more flaming tsunami shots) here. [update 3] Also check out the gallery from The Big Picture. Crazy imbricated cars, roads split down the middle, and much more.

Of course, we will all be learning more about this event, and the damage it caused, over the course of the next hours and weeks. In the meantime, the key thing is for people in low-lying areas in the Pacific Basin to move towards higher ground. The forecast shows the potential for some damaging waves:

One final thought: there was a big earthquake in New Zealand recently, and now there is a big earthquake in Japan. People will naturally enough see a correlation in timing, and ask whether the two are related. The apocalyptically-minded will doubtless make claims that these events are harbingers of the end of the world. They are not. As with the last time there were several large and damaging quake clustered closely in time (though not space), this clustering is what we would expect with a random distribution of events of different sizes over time. Sometimes the big ones come close together, and sometimes (like in all the time since the Haiti and Chile quakes until now), they don’t. And it’s not the moon. So don’t freak out. This is what the Earth does. It’s awful when people die as a result of the physical processes our planet goes through, but it doesn’t mean the planet is acting in any way unusually.

More later, if I get time to write.