AS THE death toll from the earthquake in South Asia rises, the thoughts of scientists will turn once again to prediction. Yet the unhappy truth is that seismologists remain divided over whether it will ever be really possible to forecast one of these disasters. There are many who would agree with Charles Richter — of the famous magnitude scale — when he said that: “Journalists and the general public rush to any suggestion of earthquake prediction like hogs toward a full trough . . . (prediction) provides a happy hunting ground for amateurs, cranks and outright publicity-seeking fakers.”

The popular notions that animals behave oddly before tremors (such as catfish become agitated), or that the air contains magnetic fingerprints of impending doom — ideas that remain unproven — have lent an air of charlatanism to the endeavour. Nonetheless, many respectable academics devote their careers to trying to read the geological runes, hoping that improved sensing instruments, complex computer models and colossal computing power will eventually deliver the goods.

The event that struck off the coast of Sumatra is classified as a “great earthquake”, a titan that hits, on average, once every five to ten years. It is this vagueness that presents the challenge — scientifically speaking, a prediction must include a timing, a location, a magnitude and a probability. Geologists can usually do the “where” bit but it is pinning down the other factors that is tricky. One of the most carefully constructed forecasting experiments — the so-called Parkfield Prediction Experiment conducted by the United States Geological Survey — gave warning of a magnitude 6 earthquake occurring between 1988 and 1992 in the Parkfield region of California. It arrived, 12 years late, in September this year, although the magnitude was spot-on.

The fashion for forecasting took off in the Seventies, a decade that saw the first notable success. Haicheng, in China, was evacuated two days before a 7.3 quake hit. The evacuation was credited with saving more than 100,000 lives. But after other false alarms, optimism soon waned; the Chinese authorities failed to issue an alert for an even more destructive episode a year later. The Kobe earthquake hit Japan, one of the most seismically monitored nations on Earth, out of the blue. Since then many experts have said that the successes have been down to chance.

There does, however, seem to be fresh enthusiasm for this maligned science. Vladimir Keilis-Borok, an 83-year-old who heads an international team of seismologists at the University of California, Los Angeles, maintains that it can be done by looking for distinct patterns in seismic activity. He managed to predict two big earthquakes in 2003, one in California and one in Japan, months before they happened (although a third prophecy went unfulfilled).

Some scientists wish to revive the National Earthquake Prediction Evaluation Council, an American committee that can act as an honest broker in evaluating different methods. The professor himself remains unperturbed by his failure to achieve a hat-trick.

“If we made no mistakes, that would mean earthquake prediction is easy, which it is not,” he muses.

• THERE IS much anecdotal — and some academic — evidence to support the idea that people can postpone their deaths in order to pull through Christmas or other important occasions. A smattering of investigations have found, for example, that deaths among Jewish patients show a peak after Passover, and that elderly Christians enjoy a spot of tinsel and turkey before wisely departing before the January sales.

  A fresh study of cancer patients, however, casts doubt on the notion that a steely soul can triumph over a frail body. The death records of more than 300,000 patients in Ohio were scanned to see if the death rate increased immediately after Christmas, Thanksgiving and birthdays. No such spikes were recorded. The study, conducted by academics at the Ohio State University Cancer Centre, has just been published in the Journal of the American Medical Association.

  Donn Young, one of the investigators, said people should be prompt about sharing time with dying relatives: “It’s wishful thinking to think that someone can hang on through the holidays. When it comes to death, you can’t control it.”

Times Recommends

• Strip away the figleaf and reveal naysayers
• Dave’n’George: there may be trouble ahead
• I do not accept that £64,000 a year for politicians is peanuts