The seismic non-pocalypse

A post by Chris RowanBefore everyone was distracted by Eyjafjallajokull disrupting air travel and tongues alike, the geo-worry of the moment was not volcanoes, but earthquakes. So far this year there have been six earthquakes of magnitude 7 or greater, including the a devastating magnitude 7.0 quake in Haiti, a truly enormous tsunami-generating magnitude 8.8 off the coast of Chile (although fortunately, the tsunami hysteria was greater than the actual tsunami), and a 7.2 in northwest Mexico. Most recently, last week’s magnitude 6.9 in western China/Tibet also caused much damage and many casualties.
Whenever the world experiences what seems like a run of especially damaging earthquakes, people start to wonder if it has some long-term significance, with some seeing precursors of their impending apocalypse of choice. The USGS has clearly been fielding some questions about the perceived uptick in seismic activity, as last week they released a statement that in terms of earthquake frequency 2010 is so far on course to be a relatively normal year.

Scientists say 2010 is not showing signs of unusually high earthquake activity. Since 1900, an average of 16 magnitude 7 or greater earthquakes – the size that seismologists define as major – have occurred worldwide each year. Some years have had as few as 6, as in 1986 and 1989, while 1943 had 32, with considerable variability from year to year.

With six major earthquakes striking in the first four months of this year, 2010 is well within the normal range. Furthermore, from April 15, 2009, to April 14, 2010, there have been 18 major earthquakes, a number also well within the expected variation.

I thought this point could be best illustrated visually, so I ran a search through the USGS/NEIC earthquake catalogue for earthquakes of magnitude 7 or greater that have been recorded globally since 1973 (the starting point of the database). The results are plotted below. In the last 28 years, there have been on average around 13 such ‘significant’ earthquakes a year, with a magnitude 8 occuring about every year and a half. This average rate is marked by the grey line on the plot: if we extraplolate the six major earthquakes recorded in the first four months, 2010 is on course to experience 18 major earthquakes, a little above average but well within the variability shown by the whole dataset (and it’s actually closer to the centennial average of 16 major quakes a year reported by the USGS above).

Significant global earthquakes 1973-2010
(Click for a larger version.)

However, looking at earthquake frequencies may not be the best way to examine this issue. What we’re really interested in is the total energy being released by these earthquakes, and because the magnitude scale is logarithmic, a single magnitude 8 is responsible for releasing many times more energy than a magnitude 7. In fact, an increase of one unit of earthquake magnitude, which corresponds to a 10-times increase in the amplitude of the earthquake waves recorded by a seismometer, corresponds to >a 32-times increase in the total seismic energy released. If you look at the frequency distribution on the graph above with that fact in mind, it immediately becomes clear that the relatively infrequent magnitude 8 and 9 earthquakes actually release far more energy than all those magnitude 7s. A plot of annual seismic energy release clearly shows this: there is a lot more year-on-year variability, because in some years there are one or two magnitude 8s and in others there are none, and there are too few magnitude 7s to make up the difference. The year that really stands out is 2004: it was only a slightly above average year in terms of earthquake frequency, but one of those quakes was the magnitude 9.1 Boxing Day earthquake off Northern Sumatra, which released as much energy as 1500 magnitude 7s, all in one go.

Global seismic energy release 1973-2010
(Click for a larger version.)

Interestingly, 2010 does stand out a bit on this plot thanks to that magnitude 8.8 off Chile, which is, behind the Sumatra earthquake, the 2nd biggest earthquake recorded since 1973. Indeed, if you’re willing to squint a bit, you might hypothesise that there does seem to be an increase in the rate of seismic energy release in the last 5 years. However, the significance of this is easily refuted by adding a couple of events from just before the NEIC catalogue starts: the 1960 Chile earthquake (magnitude 9.5), and the 1964 Prince William Sound earthquake (magnitude 9.2).

Global seismic energy release including Chile and Alaskan earthquakes

Effectively, the entire 27 years’ worth of seismic energy release recorded in the NEIC catalogue only represents about three quarters of the energy released by just these two earthquakes. So, if there was ever a time to announce the oncoming seismic apocalypse, the early 60s would have been the time to do it – even if you would then have had to cope with the embarrassment of us still being here. The problem is that the seismic rhythms of the planet operate over timescales much longer than a year, or even a decade. Over hundreds and thousands of years, there is some regularity to the seismicity in a particular area, driven as it is by tectonic forces that can remain relatively constant over millions of years. But we humans don’t see things on that broad temporal scale: instead we are stuck right in the temporal guts of a random process, giving the occasional large earthquake much more significance that it actually has in the grand scheme of things.
There is a slightly more subtle point here too. The seismic energy released so far this year was mostly due to just one earthquake: the magnitude 8.8 off Chile. Whilst it caused $30 billion of damage and killed more than 500 people, in terms of human impact it was dwarfed by the earthquake in Haiti – despite that quake releasing less than 1/500th of the energy. As Kim explained a while back, Port-au-Prince was right next to a shallow rupture, so the shaking due to the earthquake was extremely intense; in Chile, the fault rupture was deep and offshore, and by the time it had reached the towns and cities on the coast, the seismic energy had spread out, resulting in a wider region of strong, but not quite so intense, shaking.
So really, worrying about rates of large earthquakes is a little beside the point. For human populations and infrastructure it’s not so much a question of how much energy is being released, but where that release is taking place.

Categories: earthquakes, geohazards, geology

Comments (11)

  1. Lab Lemming says:

    Chris, you last sentence misses the most importnat point: Chile has modern earthquake-resistant building codes that are on par with those of Japan and California. Haiti barely has a government or an economy, but less the institutions capable of creating an enforcing codes and the infrastructure to build to them. I would be surprised if th eaccelerations measured in Haiti were that much greater than in Conception. What is clear is that the city was not built to tolerate the same level of shaking.

  2. Chris Rowan says:

    Chile shakemap: maximum intensity VIII (8) over a large area
    Haiti shakemap: maximum intensity IX (9) but over a very small area that unfortunately included Port-au-Prince
    Robust building codes are of course an important part of managing earthquake risk (as I have stated many times before) but in this case the point I was trying to make was simply that a small, close earthquake can be just as, if not more, dangerous than a large, distant one.

  3. Terry Wallace says:

    I enjoy your blog; it covers a wide range of topics, and covers them well! I appreciate your analysis on the seismicity, but there are many seismologists that think the last ten years are actually quite unusual. There have been 18 moment magnitude 8 earthquakes in the decade, far higher than any decade in the 20th century. There are some usual events like the 2007 Solomon Islands event in which there was a magnitude 8.1 followed 20 minutes later by a magnitude 8.0 that does not show up in most earthquake catalogues. There are lots of statistical tests to examine if this cluster of magnitude 8s is truly unusual – and indeed if earthquakes are assumed to be a Poisson process then maybe the decade is not exceptional. However, extreme event analysis suggests that it may be significant. There are several research groups working on this hypothesis right now. The interesting question then becomes “why”? Even using earthquake energy, it would appear that something is strange about the last decade. On idea is that very large earthquakes trigger other events – perhaps a super cluster every 50 years or so (the Aleutians, Chile and Alaska in the middle of the 20th century). Of course, the record is so short (and perhaps incomplete) this is speculative. The point is that we actually don’t understand the earthquake cycle well, and in the last decade our thoughts on things like earthquakes triggering other earthquakes in the far field has changed significantly. It is a great time to be a seismologist.
    Terry Wallace

  4. Mark says:

    Is there a case for the reporting equipment getting better/more of them since the 70′s as well?

  5. Chris Rowan says:

    Terry – thanks for your perspective. I feel that a decade is still too small a time window to say anything meaningful about trends in seismicity, and that even a century may be too short. But then, I’m a tectonics person, not a seismologist, so I guess I would naturally tend that way…
    Mark – we’re only looking here at the very biggest earthquakes, so improved instrumentation might help with locating and characterising them, but not detecting them.

  6. eddie says:

    A wee questionl when calculating the energy totals of various earthquakes, do you take all aftershocks in the total?

  7. Chris Rowan says:

    No. Just the main shocks. But the aftershocks are much too small to make much appreciable difference.

  8. John says:

    Thank you. That was very informative, and very interesting.

  9. John says:

    I’m curious to know if there’s any correlation between rising sea levels (and diminishing glaciers), to the frequency and magnitude of earthquakes.

  10. Sergio says:

    Be carefull when saying things like this: “…magnitude 8.8 off the coast of Chile (although fortunately, the tsunami hysteria was greater than the actual tsunami)… the tsunami killed nearly 150 people and destroyed partially ot totally, six towns… What did you need?… still more damage and fatalities to void calling hysteria on what happened?

  11. Allan L??pez says:

    The recent Chilean tsunami was real and plenty of destruction probes it. It was not a scizofrenia but real schizosphere .
    Allan L??pez