On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
Aftershocks—larger than any quake to hit the mainland U.S. in years—continue to rattle a beleaguered Japan. Thursday's magnitude 7.1 earthquake was among the largest so far, but experts had not immediately reported major additional damage, and a tsunami warning, issued shortly after the 11:32 P.M. local time quake, was called off within two hours.*
The island nation is frequently rocked by tremors, owing to its location on the seismically active Pacific Ring of Fire. But the March 11 magnitude 9.0 quake was the largest in the country's recorded history. And even in the earthquake-ready country, most local seismologists had not expected a temblor so large.
But in addition to launching a destructive tsunami and playing a role in disabling the still-leaking Fukushima Daiichi nuclear power plant, the massive March earthquake also shook up many assumptions about the nature of the region's fault zone.
So when will the ground around Japan quiet down? The science is still shaky. But Thomas Heaton, who directs the Earthquake Engineering Research Laboratory at the California Institute of Technology, explains what might be in store for the area—and why a magnitude 7.1 aftershock is no surprise.
[An edited transcript of the interview follows.]
There had already been some pretty strong aftershocks since the March magnitude 9.0 earthquake in Japan. Is it surprising to have another large quake so long afterward?
No. It's perfectly normal behavior. The statistics of aftershocks are: on average there's one earthquake one magnitude less than the main shock and 10 that are two units less than the main shock. So the sequence so far is pretty much typical for a magnitude 9 earthquake. But that's sort of the average statistic. Every earthquake is different.
Can we expect any aftershocks larger than the magnitude 7.4 quake today?
On average, about one in 20 earthquakes is a foreshock, and there's actually a chance that this 9 was a foreshock. But it's not a very high probability given this sequence of earthquakes.
This sequence is pretty special, though, because the Japan Trench is one of the world's most active subduction zones. But historically there's been relatively little seismic activity on the trench compared to the relative activity on the fault. Because there's been little activity in the trench, the Japanese seismologists had classified it as a relatively low hazard area. But several Western seismologists had questioned that, including me.
If it's been locked for 1,000 years, then it implies a tremendous amount of accumulated strain in the region. And the earthquake doesn't seem to be big enough to account for the accumulated strain. We're still left with the question: Where do we put the remainder of the slip of the fault?
At what point do we start calling these events earthquakes rather than aftershocks—is there a time or size cutoff?
There's really no definition of what we call an aftershock or an earthquake. Every time you get an earthquake, it tends to trigger additional earthquakes in the area. Everybody debates when to stop calling it an aftershock. Some people still say that when there is a small earthquake around Missouri, it is an aftershock from the 1812 earthquake.
If there are more large aftershocks along this area, are they also likely to create tsunamis?
It turns out that magnitude 9 earthquakes always make big local tsunamis. It almost certainly moves a lot of water, but the key thing about a tsunami is the amount that is displaced. That is multiplied by a factor of 30 by every magnitude unit you go up. So a magnitude 9 is very different from a magnitude 8. Once you get to a magnitude 9, you're almost guaranteed you're going to get a giant tsunami with it.
One of the big things about this earthquake for the people in the United States is that people are now able to visualize what a giant subduction earthquake does in a developed country. We now have strong evidence that we have even stronger earthquakes in our own Pacific Northwest. All magnitude 9s have a tremendous impact on anybody who lives in the region of the magnitude 9. Between the tsunamis, the ground failure and the prolonged ground shaking magnitude 9s are a very serious business.
Given what we know about the seismic activity around Japan, can we predict where the next quake is likely to be?
It looks like most of the slip was in the northern part of the Japan Trench. And it looks like, at least for the last 20 years, the southern and northern part have been locked. So a simplistic argument would be the northern part would have had much of its strain released in this event. But it's pretty obvious that we don't understand this system very well.
So is it too simplistic to think that if one part of a fault slides in an earthquake, another part of the fault will build up more pressure and be more ready to go?
We've got a pretty simplistic model. But to be honest, when we get into the details of our understanding, you can poke a lot of holes in our standard theories pretty quickly. There are many first-order mysteries left in the process.
I get very uncomfortable when statements are made about what is or is not possible in these systems. I think our knowledge is pretty rudimentary—and possibly naive.
What can we say about the short-term seismic future for Japan and the region?
When we look at other magnitude 9s, they have many aftershocks. The average magnitude 9 has one magnitude 8. And there was one magnitude 8 that was hidden in the main earthquake data. They will have 10 magnitude 7s, 100 magnitude 6s and so on. That's the typical sequence.
After the Sumatran earthquake there was a magnitude 8.7 earthquake roughly a month or so later, and there have been many other magnitude 8s south of the Sumatran earthquake. In a sense the entire zone kind of lit up for years. So there's no scientific reason to say that this thing is over. Just on average, something like this magnitude 9 would be the biggest and you wouldn't see another one somewhere else. But we just don't know.
What are some of the big questions that still need to be answered about this region?
Does this zone primarily slip in earthquakes, or does it primarily slip in a slower process—either through steady creep or occasionally large, slower earthquake that may occur over a period of days or weeks? And we just don't know the answer to that.
I think many people were assuming since there hadn't been activity over the last 1,000 years that it must be slipping steadily or in slow earthquakes. But now that we've had this large event, it's pretty obvious that that view was wrong.
It also calls into question our approach to protecting ourselves from earthquakes. Often we characterize our earthquake hazard with some kind of probabilistic map. We prioritize how we respond based on these maps. Unfortunately, recently a lot of our most important earthquakes have been in areas that were low-hazard areas. And it to me is an interesting statistics problem. It is like trying to predict where and how large the next war will be, or where and how large the next epidemic will be. Right now what we use is like a heart attack model of risk. Our problem is more like a bird flu kind of problem.
Is there any way that engineers and seismologists can prepare for this sort of event—a massive earthquake and continued large aftershocks?
Really large earthquakes are so rare that there's no way that engineers have been able to say that they know what the ground shaking is like to know how to build for it. This is the first time we're going to have ground recordings from a magnitude 9, so we're all very interested in it.
But, to be honest, I'm a little concerned now that because we'll finally have ground recordings, people will say: "Now we know what a magnitude 9 looks like." But just because you've seen one, you haven't seen them all. Most of the rupture was a long distance from land, so it was far offshore. But in my opinion that distance really spared Japan a lot of damage.
It looks like part of the fault that's been locked for 20 years is closer to land. That part would be much closer to the buildings of Japan—between Sendai south to the Boso Peninsula.
From our perspective, one of the really interesting parts of this sequence was that the Japanese have—just in the last two years—begun operating a real-time alerting system that alerts people when shaking has commenced somewhere and is on its way to them. And the system actually worked in this earthquake. For Sendai, the warning was in the neighborhood of 10 seconds; for Tokyo it was more like a minute.
The idea that there's some new tool out there that would give you a heads-up—and some idea of what's about to happen—really is a new thing.
*Update (4/7/11): The magnitude of this earthquake was downgraded midday by the U.S. Geological Society from 7.4 to 7.1. This paragraph was changed after publication to reflect that change.