Hawaii, the New York Times, and Triggered Seismicity

The New York Times has an astonishingly confusing article on the idea that one earthquake could trigger others, thousands of miles away. The real meat of the answer is hidden in the last paragraph:

While strong earthquakes trigger nearby aftershocks (in Hawaii a shallower 5.8 quake occurred within minutes of the first one), they don’t affect seismic zones hundreds of miles away. So why does it sometimes seem like there’s a string of strong quakes throughout a region?

That, Dr. Wolfe said, has more to do with psychology than seismology. The United States Geological Survey estimates that every two and a half days, on average, there is a quake of magnitude 6.0 or higher somewhere around the world. Many of these take place under the oceans or in remote land areas and are little noticed. But then one occurs in a place like Hawaii and earthquakes are in the news. So for a few days or weeks, every Tom, Dick and Harry of a quake is reported, and it feels as if there must be a connection. Even though there isn’t.

But to get there, you first must wade through some improperly contextualized information about plate tectonics:

Generally, no [one earthquake doesn’t lead to another], said Cecily J. Wolfe, a seismologist at the University of Hawaii at Manoa. More specifically, the Hawaiian quake, which damaged roads, buildings and bridges and knocked out power, was a different kind of earthquake than most of those that occur around the Pacific Rim.

Earthquakes around the rim, like the Peruvian quake, result from the release of stresses that build up at the boundaries of tectonic plates. “Out here in Hawaii, we’re in the middle of the interior,” Dr. Wolfe said. There are no plates sliding past each other.

The fact that the Hawaiian earthquake was an intraplate event has absolutely nothing to do with the implausibility of it triggering a subsequent event in Peru.

Long-distance triggered earthquakes are a real (well, controversially real) phenomenon. It’s thought that both the 1992 M7.4 Landers earthquake in the Mojave Desert and the 2002 M7.9 Denali earthquake in Alaska triggered, not big earthquakes, but an increase in the rate of small earthquakes throughout the Western U.S. and Canada. No one quite understands how this happens.

The occurrence of triggered seismicity does seem to be correlated with the direction in which the triggering earthquake radiates most of its energy. It also occurs at such large distances that the changes in stress observed when the earthquake has finished (static stress) are far, far, far too small to plausibly be important; they are much smaller than the stress changes observed as the seismic waves pass by (dynamic stress). So the triggered events are most likely due to something that happens as the seismic waves pass through. Favored explanations begin with “probably these areas were close to failure anyway” and end with “um, somthing something fluids something” or “nonlinear whoozerwhatsits”.

At any rate, the reason we don’t think the Hawaii earthquake triggered seismicity elsewhere has more to do with the fact that it was small (a magnitude 6.6 earthquake radiates 31 times less energy than a magnitude 7.6 earthquake) than that it was in the middle of a plate. And magnitude 6 events occur all the damn time, especially on the Pacific Rim; there is nothing special about either the Peru or Papua New Guinea events that would make us go looking for a special explanation.


  1. Lab Lemming wrote:

    When you mean small, do you mean like, 1, 3, or 5? I was not aware of any seismic increase related to Landers, aside from the immediate aftershocks.

    I like the phrase “controversially real”, though.

  2. yami wrote:

    It’s an increase in overall number of earthquakes at a particular location, without special regard to magnitude. Maximum magnitude of triggered events apparently decreases with distance from the triggering fault, the biggest for Landers was an M5.6 in Nevada (see: Hill, D. P. et al. Seismicity remotely triggered by the magnitude 7.3 Landers, California, earthquake. Science 260, 1617 –1623 (1993). )

  3. Western Geologist wrote:

    I’m not a seismologist, but I don’t think the stresses induced by remote earthquakes are …”far, far, far too small to plausibly be important…”

    According to this paper the stress induced by the Denali quake in Long Valley was 0.06 MPa. Assuming that the induced events in Long Valley had static stress drops of 0.1 to 10 MPa (which I’m basing on Fig. 6 in this paper – I don’t know how valid this assumption is), the remotely induced stresses are 0.6 to 60% of the static stress drop.

    However as I warned at the beginning of this comment I’m not a seismologist so I could be missing something.

  4. reviewer_a wrote:

    I read the article and I thought that as such things go, it was not bad. The main point I recall from the article was that M6-7 quakes are not uncommon, and that people begin to worry about accelerating rates of big quakes and potential triggering, when they hear of a few big quakes within days of each other. People get the sense that this is uncommon when it is not, in cases where such quakes happen to occur in populous areas and get media coverage. I think the article performed a service.

    Also, of course it had inaccuracies. Have you ever been interviewed by the media? It’s often like putting your words in a blender. This example of science journalism did not strike me as egregiously bad.

  5. yami wrote:

    Perhaps I have been spoiled by the generally excellent standard of science reporting at the NYT. The main point was good, but I think it should’ve been made before the end of the article as not everyone will read all the way through.

    WG, 0.06 MPa was the dynamic stress change observed as the Denali Rayleigh waves passed through, not the static stress. I’m only reading the abstract here, but it seems the static strain offsets they observe in that paper were caused by triggered local events.

  6. Western Geologist wrote:

    yami, thanks for the response. I’m not sure that the distinction between static and dynamic stress matters (thanks for the clarification in the body of your post too). I’m asuming that the magnitude of the static stress drops corresponds to the magnitude of the differential stress that would be needed to cause an earthquake. If that’s true then the stresses induced by the Denali quake were (potentially) a significant portion of the stresses that would be required to cause an earthquake. They wouldn’t be large enough by themselves to trigger earthquakes, but in an active area like Long Valley, the addition of the Denali-induced stresses was enough to make some of the faults there fail.

    Rereading your post I think you may have said something similar when you wrote: “o the triggered events are most likely due to something that happens as the seismic waves pass through. Favored explanations begin with “probably these areas were close to failure anyway” …”, in which case my comment was based on my lack of reading comprehension.

  7. yami wrote:

    I was trying to avoid the jargon of Coulomb stresses in my post, but that’s really what I was talking about – Coulomb stress changes from Denali were too small to be important in the Western US. Dynamic stresses caused by pas

    I suspect we’re talking past each other here, but I don’t know why you would assume that the triggering stress is related to the static stress drop of the triggered event. We don’t have much constraint on the amount of tectonic or other pre-existing stress relieved by the event…

  8. Western Geologist wrote:

    I mentioned the static stress drops because I was assuming (possibly unrealistically) that the magnitude of the static stress drop was equivalent to the Coulomb stress. This could be the case if faults in general are weak. If faults in general are weak, and I’m thinking of something like the San Andreas level of weakness with a coefficient of friction of 0.1 to 0.2, then the 0.06 MPa stresses induced by the Denali event could be substantial enough to induce failure. Sorry, I certainly should have mentioned that in my original post.

    I’m embarrassed that I didn’t think of this sooner, but there may be a fairly decent constraint on the stress field at Long Valley. The strainmeters that were used to measure the Denali event were borehole strainmeters, so it’s possible that some sort of stress measurements were made after the holes were drilled.

  9. ronald wrote:

    maui had another earthquake today like last month no doubt hawaii is still unstable. how ever this morning i dident feel it was 5.something. but that 6.5 was the worst and may have more during the week. still unsettling. i do how ever know that animals do not predict earthquakes they hear and feel it seconds before humanes, i know cause caught the big quake and the animals that morning on security camera. talk about stress the big island is stressed out.

Post a Comment

Your email is never published nor shared. Required fields are marked *