Conventional wisdom says that the New Madrid Seismic Zone—which encompasses parts of Missouri, Arkansas, Tennessee, Kentucky, Illinois, and Indiana—presents a major earthquake risk, this predicated on the fact that it produced at least three large earthquakes in a space of three months in late 1811 and early 1812. Yet the zone, named after the town of New Madrid (pronounced “MAD-red”) in southeastern Missouri, is located within the North American plate (not at a plate boundary), and seems a relatively unlikely candidate for producing a devastating quake. New geological ideas and data seem to confirm that the hazard is nowhere near as high as that of California, the part of the country to which New Madrid is sometimes compared. Nevertheless, scientific misconceptions, bureaucratic instincts, the media’s love of a good story, and predictions of doom have combined to reinforce New Madrid’s reputation as a potential apocalyptic disaster area.
With the bicentennial of the New Madrid earthquakes fast approaching, and countless anniversary events planned across the Midwest beginning early next year, heightened awareness of the earthquake risk in the region is inevitable. It also explains why Seth Stein—a professor of Geological Sciences with 35 years of experience studying earthquakes—chose this time to release “Disaster Deferred: How New Science is Changing Our View of Earthquake Hazards in the Midwest” (Columbia University Press), a book that concisely explains what we know (and don’t know) about the hazard, and encourages the formulation of sensible policy that takes account of our newfound understanding.
Stein was kind enough to speak with me by phone from his office at Northwestern University, eager to relate what he’s learned from “twenty years of thinking about New Madrid.” The bottom line, he says, “Is that any 1811/12-style earthquake is very far in the future and the hazard is nowhere near as high as being claimed.”
Let’s begin by talking about the earthquakes that struck the Midwest in 1811 and 1812.
The first thing to realize is that they were big, but nowhere near as big as the image that has gotten out there. You’ll hear claims that they were the biggest earthquakes ever to hit North America, which is not true. In fact, there are about 15 earthquakes of the same magnitude [around 7.0] somewhere in the world every year. And they certainly didn’t ring church bells in Boston—that’s an urban legend.
The earthquakes—on December 16, 1811, January 23, 1812, and February 7, 1812—occurred before the invention of the seismometer. How do we know how big they were?
By taking account of the damage pattern and comparing it to damage patterns of earthquakes that have happened more recently, where we know their magnitude from seismometers.
Which faults are part of the New Madrid system?
The ones that were active in 1811-12—we think—were the Cottonwood Grove fault, the Reelfoot fault, and the North New Madrid fault. There are a lot of other faults in that area and several thousand years ago, different ones were active. What we’re learning is that faults switch on and off. They will be active for a thousand years or so, and then inactive for several thousand years. And then other faults may become active. From a scientific standpoint, that’s the really exciting thing we’ve learned from New Madrid. It has been the key to the door that opened up a whole new understanding about how faults inside continents work.
What makes the Global Positioning System (GPS) such an important new tool in earthquake studies?
The way earthquakes work is that energy is stored up in the ground for hundreds of years. Then it’s released in an earthquake—like a spring. A seismometer tells us what happens during an earthquake, but GPS lets us see what’s happening between earthquakes. It lets us see the ground storing up that energy and deforming.
When we look at faults around the world, we see them storing up that energy. So when we first put markers in the ground and measured the position of the Midwestern fault lines we were surprised that we didn’t see any motion at New Madrid. We concluded that there’s no sign that a big earthquake is on the way.
So why do the National Seismic Hazard Maps show New Madrid equaling or exceeding the earthquake hazard of California?
Until the mid-1990s they didn’t. Then [the United States Geological Survey] changed the computer program used to produce the maps, and that drove the hazard up enormously. It used to be about a quarter to a third of that of California, which is sort of reasonable. But then for complex bureaucratic reasons they decided to change the method. Traditionally we have defined seismic hazard as the shaking we might expect on average once every five hundred years. But for New Madrid they decided to make it the biggest shaking one might expect every 2,500 years, and that raised the hazard. The question is whether it’s physically meaningful or not. The answer is: probably not.
What other factors have contributed to the New Madrid earthquake hazard getting blown out of proportion?
If you think of swine flu, or SARS, or Y2K, you’ll notice a pattern [that emerges]. Government agencies tend to exaggerate, the media likes to hype these things, and we all like to hear these stories.
What effect did Iben Browning’s infamous prediction [that a major earthquake would occur in the New Madrid seismic zone within a few days of December 3, 1990] have?
It had some effect. The way I think about it is that it set fire to logs that had been stacked there. The government had been telling people that there was a huge earthquake danger, so when someone said, “An earthquake is about to happen,” everybody was prepared to believe it. And the government has continued to warn of these apocalyptic events—that thousands of people are going to die and there will be billions of dollars of damage. They keep putting these messages out, so people have been pre-conditioned to believe them.
So the bottom line is Midwesterners don’t need to be particularly worried?
We will continue to have small earthquakes, many of which are aftershocks of 1811-12. But it doesn’t look like we need to worry about an 1811-style earthquake for many hundreds of years—perhaps even longer. The danger of earthquakes in the Midwest is about 1/10th that of California.
Based on GPS data, might there soon be a big earthquake in an area where the public doesn’t expect one?
We haven’t found one in this country, though a good example would have been the [magnitude 7.9] earthquake that happened [in May 2008] in Sichuan Province, China. The Seismic Hazard map didn’t show it as very dangerous, but if you looked at the GPS data, it showed that the strain was building. In the U.S., even though there aren’t big earthquakes now in Utah, Oregon and Washington, we can see the strain building up in the GPS data.
What advice do you have in terms of devoting resources to preparing for a major earthquake in the Midwest?
The problem is that the preparations the government wants communities and businesses to undertake are very expensive. Bringing one hospital in Memphis up to California standards would cost about 100 million dollars. Bringing all the buildings in the area up to the standard would take billions of dollars. It would be a huge tax with little benefit. Spending that money on other social needs would do a lot more good.
The moral of this is that people should calm down, and communities should think carefully before rushing to spend millions of dollars preparing for an earthquake.
Do you have any thoughts about the New Madrid bicentennial?
There is going to be a lot of hype about the bicentennial, and they are going to stage earthquake drills [including the Great Central U.S. ShakeOut on April 28, 2011]. I think everyone should have a good time with the bicentennial and enjoy it, but I don’t think there is anything worth getting excited about. Treat it like Halloween—kind of fun, but not to be taken too seriously.