Seismologist Giacchino Giuliani desperately tried to warn the people of L'Aquila, Italy ahead of the earthquake that left 150 people known dead. The Italian government removed his forecast from the internet and stopped him from alerting the people.
Vans with loudspeakers drove around the town a month ago telling locals to evacuate their houses after Giuliani, from the National Institute of Astrophysics, predicted a large quake was on the way, prompting the mayor's anger. Giuliani, who based his forecast on concentrations of radon gas around seismically active areas, was reported to police for "spreading alarm" and was forced to remove his findings from the Internet.
My heart goes out to the people of Italy. Devastation struck a picturesque town, killing over one hundred, while most people were sleeping.
Thirty years ago today I was working hard at UCLA to complete my doctoral dissertation in geochemistry on my research on soil radon gas level changes before and after earthquakes on the San Jacinto fault in California. The results of my research were positive, but funding for earthquake prediction research was cut after a string of disappointing results from geophysical methods. When Dave Jackson (who sat on my dissertation committee) proved the "Palmdale Bulge" was just measurement error and that the "Big One" was not likely to hit Los Angeles anytime soon, the money dried up. I moved onto a more secure area of study, nuclear waste disposal research.
The old unreinforced masonry construction failed. Modern codes mandate steel reinforcement "rebar" for brick, block and concrete structures in areas with earthquake hazards. Much of the damage in Italy would have been prevented in modern buildings, but the historic buildings didn't stand a chance against a M6.3 earthquake. L'Aquila was a beautiful traditional town that could not be readily retrofitted.
California has strictly enforced seismic zoning on new buildings since the Long Beach earthquake of 1933. Public buildings in California built before 1933 have been retrofitted to meet standards. California's leaders decided that engineering, not earthquake prediction, was the best way to save lives and limit property damage. Italy, with its large number of historic structures, has not relied on building codes. That's why the death and destruction in L'Aquila is so severe.
When I was doing earthquake research I frequently thought about what I would do if I thought I had evidence that might indicate that a large earthquake was imminent. I realized that my methods and data weren't tested enough to be reliable. Talking to the press or the public would not have been appropriate without more research. I doubted that one method, such as radon concentration or flux measurements, would be reliable by itself. I thought that multiple lines of converging data would be needed to predict earthquakes.
Today, many seismologists doubt, based on the theory that an earthquake has a small source area at considerable depth, that earthquakes can ever be predicted. However, I am not one of them. I think that some earthquakes, such as deep focus earthquakes which are hundreds of kilometers deep, cannot be predicted. Others, such as the M6.3 earthquake in Italy, which happened on a relatively shallow fault that intercepts the ground surface, may be predictable, in my opinion. I will try to explain why.
Giacchino Giuliani may have forecast this earthquake. Careful review of his data and his methods will be needed to evaluate his warning. The problem Mr Giuliani faced was that his forecasting method was imprecise. He was unable to predict the time precisely enough for it to be a useful prediction. However, some of the criticism coming in from geophysicists is unhelpful.
Reuters adds that some Italian authorities are still not persuaded that Mr. Giuliani’s warning was anything other than a coincidence, including Enzo Boschi, the chairman of Italy’s National Institute for Geophysics and Vulcanology:
As the media asked questions about the authorities’ alleged failure to safeguard the population ahead of the quake, the head of the National Geophysics Institute dismissed Giuliani’s predictions.
"Every time there is an earthquake there are people who claim to have predicted it," he said. "As far as I know nobody predicted this earthquake with precision. It is not possible to predict earthquakes."
Mr. Giuliani warned the public of an impending earthquake and was censored. Clearly, he did not claim after the fact that he had predicted the earthquake. He predicted the location well enough to warn the town. That's no mean feat. "It is not possible to predict earthquakes" is a statement of belief that does not appear to be consistent with Mr. Giuliani's censored warning. It appears to be an effort to discredit and censor Mr Giuliani.
Censorship is inconsistent with the scientific method.
Good science depends on fully open publication and discussion of research results. While it may be true that people come out of the woodwork claiming that they can predict earthquakes, the work of a legitimate scientific investigator cannot be legitimately dismissed because of the false claims of flaky people.
The New York Times published saw fit to put on line "news" that wasn't fit to print.
In 1995, The Times reported that changes in radon gas levels "do not precede all earthquakes and cannot be used as a basis for issuing warnings to the public." We just added more information to the post below based on an interview with Ross Stein of the United States Geological Survey, who called radon gas measurement one in a series of "great white hopes" for earthquake prediction that has been cast aside by most scientists.
Mr. Stein should know that radon is a clear odorless gas. To date all efforts to predict earthquakes have been "great white hopes". The same used to be true for weather prediction. And the theory of continental drift was cast aside by most scientists. Until Mr. Giuliani's work is thoroughly reviewed the correct answer is.
I don't know.
I have found supporting evidence that may help explain why Mr Giuliani observed increased radon levels before the earthquake in Italy.
The M6.3 earthquake was occurred on a normal fault in Central Italy. Normal faults happen where the earth is under tension, it's stretching, and one side of the fault drops down by gravitational force. It's like a landslide but larger.
I found a paper which may help explain why radon gas flux (and likely the flux of other gasses such as CO2) would have increased before the earthquake. Migration of "overpressured" fluids before the earthquake may have triggered it.
Tectonophysics 348 (2002) 155– 168 "Normal faulting, transcrustal permeability and seismogenesis in the Apennines ( Italy)", Francesca Ghisetti a,*, Livio Vezzani baDipartimento di Scienze Geologiche, Universita` di Catania, Corso Italia 55, 95129 Catania, Italy bDipartimento di Scienze della Terra, Universita` di Torino, Torino, Italy
Within this wedge, (i) crustal thickness in excess of 25–30 km, (ii) growth of spaced normal faults that penetrate the crust down to seismogenic depths, and (iii) preservation of thrust structures not yet dissected by extensional fracturing favour maintenance of overpressuring conditions at depths of 10–15 km that may eventually trigger episodic seismic slip during 5 <Ms < 7 normal faulting earthquakes.</p>
Geological, geophysical and geochemical data suggest a close interconnection between amount of crustal stretching, depth of detachment of normal faults, and fluid circulation. In particular, the change of tectonic regime from folding and thrusting to normal faulting appears to drive rising levels of vertical permeability that have increased proportionally with length of time elapsed since onset of extension. Progressive build-up of extensional strain, leading to linkage of different fault systems in the stretched and thinned upper crust, appears to control redistribution and mixing of overpressured fluids that (i) migrate from mantle melts infiltrating the lower crust and (ii) are trapped in low permeability compartments of the thrust belt.
In this framework, it is plausible that domains of fluid overpressuring in the upper crust trigger formation of new normal faults and reactivation of older thrust faults at depths of 15–10 km, with consequent eastward migration of the belt of maximum seismic release with time.
Earthquakes in central Italy may be predictable given enough time, effort and dedication. The M6.3 earthquake was a terrible disaster but Mr. Giuliani's warning may be a bold step forward.
George F. Birchard, PhD, 6April, 2009, Sanford, NC
