Story URL: http://news.medill.northwestern.edu/chicago/news.aspx?id=209748
Story Retrieval Date: 8/23/2014 12:31:47 AM CST

Top Stories
Features
QUAKE1

The April 2009 L'Aquila earthquake reduced buildings to rubble and killed over 300 people.  Prosecutors argued that had scientists not downplayed the risk of a major earthquake, residents would not have stayed in their homes during the quake.


Illinois seismologist warns against Italian example of convicting scientists

by Will Grunewald
Oct 30, 2012


QUAKE7

Courtesy of Wang-Ping Chen

Seismologist Wang-Ping Chen is featured here while monitoring chemical blasts at a coal mine in Indiana.

Seismologist Wang-Ping Chen warns that the threat of criminal liability for inaccurate scientific predictions could undermine governments' emergency preparedness.

Chen brings decades of experience researching earthquakes at the University of Illinois and elsewhere to his perspective on an Italian court's conviction of six scientists and engineers for manslaughter for failing to correctly gauge the risk of a major earthquake.

Chen said that criminalizing incorrect predictions will undermine the public role of scientists.

“If the government and scientists are too concerned about liability, we are likely to end up with stifling bureaucracy that produces no results at all,” says Chen, who holds a doctorate in geophysics from Massachusetts Institute of Technology.

In April 2009, an earthquake devastated the central Italian city of L'Aquila, after weeks of smaller tremors felt in and around the city. More than 300 people died in the rubble, and prosecutors argued that had experts from Italy's National Commission for the Forecast and Prevention of Major Risks not downplayed the danger, residents would have sought safety outside of buildings. 

All four scientists, two engineers and also a government expert face six years in prison, pending appeal, in the sentence handed down last week.

Reaction from the scientific community was swift. Guardian columnist and astrophysicist Stuart Clark likened the present trial to the trial of Galileo. This ruling will stifle the study of earthquakes just as the 1633 ruling stifled the study of the cosmos, he wrote. And an editorial in Nature summed up the scientific community's general consensus: “The ruling is perverse and the sentence ludicrous.”

In a Q and A, Chen offers his own assessment of the Italian court's decision and explains the scope and limitations of earthquake forecasting.

Chen headed the University of Illinois Urbana-Champaign geology department from 2007-2011 before accepting a professorship at Zhejiang University in China where he continues to study earthquakes.  He researches rupture processes that occur during earthquakes, and uses earthquake-generated seismic waves to illuminate Earth’s internal features – a process he compares to studying human anatomy with a CT scan.

What is your reaction to the guilty verdict returned last week and this idea of holding scientists criminally liable for wrong predictions?


My personal view is that this is a very delicate balancing act. If the government and scientists are too concerned about liability, we are likely to end up with stifling bureaucracy that produces no results at all. It is extremely important to point out that the essence of science is discovery of the unknown. As such, there is always a risk of failure. This is a feature common to all creative endeavors, including arts, literature and even sports.

Why did the scientists from the National Commission in Italy downplay the significance of the small tremors that turned out to be precursors to the major L’Aquila earthquake?

If you look at earthquakes in many different places and over a long period of time, in most cases small earthquakes do not lead to a major, destructive event. So this is simply “bad luck” based on statistical estimates. Is it more sensible to bet on the outcome with the highest probability or the lowest? Overall, we’d certainly be in a lot more trouble if we worked against probability.

Are probabilities as close as scientists can get to forecasting a major earthquake?

Along boundaries of tectonic plates, we know very well where earthquakes will occur, but not how soon or how large. Generally we are guided by historical records. The thinking is that if earthquakes occurred in a certain region in the past, they will recur in the same general place. However, there are cases where this line of reasoning doesn’t jibe with independent evidence. In those cases we are really stuck.

So with extent of information available, did the scientists from the National Commission in Italy do anything wrong on a scientific level?


This is a very complex case and hindsight is always 20-20. Probability and other statistical estimates have a very important tacit assumption: the number of trials must be very large. For example, it is impossible to predict the life expectancy of any individual, but the average life expectancy of, say, all the people who are at age 50 in the U.S. can be predicted very precisely. The insurance companies know this very well; otherwise they would be out of business very quickly.

Then there is the all-important difference between scientific research and public policy. Even in the unrealistic situation of perfect scientific knowledge, public policy decisions do not automatically follow good science. There are other deciding factors in play, mainly economic considerations and political climate. In fact, it is very difficult to transfer important scientific findings into public policy.

Could they have done anything differently?

If I were in the shoes of the L’Aquila scientists, I probably would have made the same decision. The reason is two-fold. First, the odds of a really large earthquake happening are small, so without any other reliable information, one must not overreact. Second, a false alarm can have dire consequences: traffic jams and accidents, financial loss, general panic. In the worst case scenario, riots and other destabilizing acts could occur.

It seems like public expectations – or at least the prosecutor’s expectations of what earthquake experts can predict – were maybe too high. What can scientists do to reconcile expectations with reality for the general public?

In the past two decades, the scientific community has placed a lot of emphasis on public outreach. We could always do more and do a better job, of course. Unlike entertainment or sports, science is usually not fun to watch or easy to digest. But anyone with some curiosity and a clear head should be able to understand the essence of any scientific issue – provided we do a good job of explaining it.

Would you want to work as a seismologist in Italy anytime soon?


I’d have to pass up the opportunity of enjoying living by the Mediterranean, tasting good wine and fine cuisine. Fortunately, the world is still large enough that you can usually find a good place to work.

And what about here in Illinois . . . What’s the risk of a significant earthquake here?


For those of us who lived in Illinois for several decades, we all felt earthquakes more than once. These events are widely felt but not terribly damaging, unless you happen to be near ground-zero. The most famous sequence of strong earthquakes in the Midwest occurred in the winter of 1811-12 near the tiny community of New Madrid, MO. There is continual discussion regarding the cause of those earthquakes and the probability of their recurrence – a highly contentious topic that would take more time to get into than we have here.

It saddens me, though, when people listen to quacks instead of finding balanced views and making informed decisions. Back in 1990 a gentleman named Iben Browning made quite a stir and lots of money by “predicting” another major earthquake in the New Madrid region. Or course, the event did not occur as predicted and his audiences did not get a refund either.

(Note: Browning was a zoologist with a penchant for claiming predictive powers over volcanic eruptions and earthquakes. When Browning died the following year, The Southeast Missourian remembered him as “a scientist with vague and inflated credentials.”)

So, will scientists ever be able to predict earthquakes?

Never say never. There are only two possible outcomes. Either we succeed someday, or we show that earthquake prediction with any certainty is impossible. Until then, some of us will keep trying.