ROME Earthquakes like the one that killed more than 200 people in Italy this week are still impossible to predict, and a local scientist's claims to have done so should be treated with caution, geophysicists say.
Gioacchino Giuliani is at the centre of a debate about the limits of seismology after Italian officials shrugged off his warnings last month that a devastating earthquake in the central Abruzzo region was imminent.
In fact Giuliani, who works at the National Institute of Physics, was even reported to police for spreading panic.
While many Italians are now more than ready to listen to whatever Giuliani has to say, geophysicists in Europe and the United States remain sceptical of his claims to have discovered an effective early-warning system.
Although scientists say it's easy to say where big quakes are likely to happen, pinning down the timing can't be done, at least not yet.
"It's a very humbling field to be in ... We cannot predict earthquakes," said Ross Stein, a geophysicist at the U.S. Geological Survey in Menlo Park, California.
"To predict an earthquake, whatever it is that we're using would have to be a reliable indicator ... And no one has gotten close to that."
Giuliani based his forecast on emissions of radon gas. The theory goes that the gas is released as the fault line adjusts itself before a major earthquake.
"We can know when an earthquake is coming," Giuliani said in one of the flurry of interviews he has given to Italian media over the past 24 hours.
"To those who have not believed me, I'll just say that a scientist's dictionary should not contain the word 'impossible'. History has proven that to us."
But Giuliani isn't the first to look at radon gas. It's been studied since the 1970s and no research has yet proven it as a reliable detection tool, scientists told Reuters.
"There isn't a definitive link between radon gas measurements and earthquake occurrence," said Brian Baptie at the British Geological Survey. "Sometimes people have measured radon gas and no earthquakes have occurred, and vice-versa."
Giuliani's forecast was far from perfect. He believed the quake would have struck the town of Sulmona, which is more than 50 km (30 miles) south of L'Aquila. He also got the date wrong, predicting the quake would strike several days earlier.
The head of Italy's Civil Protection Agency, Guido Bertolaso, told reporters that if they had listened to Giuliani, they probably would have evacuated the residents of Sulmona to L'Aquila just in time for the earthquake.
Still, Giuliani's claims have been a headache for Prime Minister Silvio Berlusconi, who has told Italians that seismology still can't "predict 'the when' of quakes."
"What were we supposed to do, take 300,000 people and tell them to stay outside their homes for an open-ended period? There's no way," Berlusconi told RAI state television.
Radon gas is just one of many avenues explored by scientists, who have looked at everything from how cockroaches behave on fault lines to sudden changes in water levels and temperature as possible tools to predict earthquakes.
Stern said if Giuliani wanted to be taken seriously by the scientific community, he should publish his work and put it up to review.
"This physicist from Italy has every right to pursue something, as do the cockroach people. But we all have to subject our work to the same standards of review," Stern said.
(Additional reporting by Gavin Jones; Editing by Jon Boyle)