Ben Goldacre, The Guardian, Saturday 19 June 2010

On the 6^th of April 2009, an earthquake registering 5.8 on the richter scale hit the town of L’Aquila in Abruzzo, Italy. This was a tragedy, and hundreds of people died. It would be great if we could have firm predictions about every risk whose rare but tragic outcome cannot be accurately predicted, whether it is a flu outbreak, a murder, an illness, or an earthquake. Most of us recognise that this is impossible.

But some find it harder to accept. The L’Aquila Prosecutor’s office have now leapt into action. They have a Commissione Grandi Rischi, after all – a “Commission on Big Risks” – and it’s full of seismologists. If these people can’t predict an earthquake, then what’s the point of them? And so these seismologists are now being indicted and investigated for manslaughter, on account of their failure to warn the population that an earthquake was coming.

You can join various Fellows of various Royal Societies in protesting about this case at qurl.com/quake. Clearly the Italian government would rather be informed by scientists who are happier to throw caution into the wind and make claims in excess of the evidence. Oddly enough, though, that did actually happen, in the week before the earthquake.

Gioacchino Giuliani is a laboratory technician who became convinced that he was able to predict earthquakes by measuring the emission of radon from the ground. He ignored the doubts of seismologists – he has never published his theories or evidence in an academic journal – and invested in several measuring devices to let him make his predictions.

Shortly before the earthquake struck, Giuliani became convinced that something serious was coming. He began desperately trying to warn the public, even posting a video on YouTube explaining his theory, and warning people to evacuate their houses urgently. Vans loaded with loudspeakers were driven around the town to spread the warning. Giuliani tried in vain to persuade the mayor that he was right.

But they did not heed this warning: instead, the local government reported him to the police for spreading unnecessary panic and alarm, forced him to remove his warnings from the internet, and forbad him from telling anyone anything about the coming earthquake.

In reality, of course, Giuliani made a lucky guess (and he was out by 55km). Nothing has changed, and there is still no reliable or validated way to predict an earthquake. Because of this, seismologists around the world are united in explaining that the best way to protect your population is not through an impossible early warning system, but rather by investing in preparedness, to mitigate against the damage done by one rare, unpredictable, horrific outcome.

So you use seismic hazard maps to reliably work out where the risk is greatest, rather than when. You change the specifications in your building code so that homes are less likely to collapse and crush people to death. You insist on retrospective modifications to existing structures. You make sure your population are educated in what to do when the worst happens, and you prepare your emergency services with enduring supplies of the appropriate equipment.

If, in a political emergency, you find you have failed to do all this to universal satisfaction: then you can charge some scientists with manslaughter. But ideally this should be a last resort.

==========================
==========================

Below is some footage of a recent presentation on risk appraisal by the academic initially approached by the Italian government to head their Commission on Big Risks.