Bill Menke's BLOG Page: Prediction Predicament.

aftermath of the quake

Very few of my fellow seismologists seem to have any sympathy for Giampaolo Giuliani, the technician at Italy's National Institute of Nuclear Physics who made an earthquake prediction a month or so before an actual earthquake devistated the town of L'Aquila in central italy and killed 293 people. Giuliani was indicted a couple of days before the April 6, 2009 quake for "spreading rumors". Giuliani's prediction was based on his observing, as part of his physics lab activities, an increase in the rate of seepage of radon gas from the ground. But frankly, I know very little about his prediction, because he was forced to withdraw all information about it from public view.

Giuliani's use of radon in an earthquake forcasting setting is by no means unique. Seismologists have monitored it on and off for the last thirty years, hoping that it would have predictive value, but with equivocal results. In a few cases, such as with an earthquake in Iceland* in 2000, radon levels spiked a few months before the quake. In many other cases, no such premonitory signal occurs. With no access to the details, I have no way of assessing the credibility of Giuliani prediction.

Nevertheless, I believe that Italian authorities would have done well to have given Giuliani's prediction a formal evaluation by a board of scientists specifially charged to recommend an appropriate official response. The US has such an entity, the National Earthquake Prediction Evaluation Council. Not withstanding claims made to the press by my fellow scientists, that earthquake predictions are made all the time, (and presumably by kooks), this Council has evaluated only a handful of predictions in its twenty year history, all made by people with arguably-relevant credentials.

Many of my colleages also have used the opportunity to assert the impossibility of earthquake prediction. Now I'll be the first to agree that we do not know how to predict earthquakes. And I'll even agree that earthquake prediction has proven to be a very hard problem. But we don't know that it's impossible. Until we have some sort of rigorous proof that it is impossible, I believe that we should keep on trying.

Some of my colleages have also decried expenditures on prediction efforts, which are after all not a sure thing, when techniques for improving earthquake preparedness (e.g. reinforcing buildings) are well-developed and have demonstrable value. Rather than argue this point, I will note that medical researchers are still eager to pursue AIDS vaccine research, even though they are all aware that HIV transmission prevention techniques (e.g. abstinence, condoms) are widely available, and even though many acknowledge progress towards an effective vaccine to be slow. And perhaps they will be up to the challenge.

I will make my own prediction, of sorts. If someone ever discovers how to predict earthquakes, he or she will not be a seismologist. We seismologist's just aren't trying hard enough.

*Einarsson, P. et al, Radon Changes Associated with the Earthquake Sequence in June 2000 in the South Iceland Seismic Zone, Pure and Applied Geophysics 165, 63-74, 2008.