The perils of predicting earthquakes

US earthquake experts were relieved when Brian Brady withdrew his alarming prediction of massive shocks off Peru this summer. But bizarre incident, involving a specific earthquake forecast, leaves a major question unanswered -- how far should a responsible scientist go in warning of danger when the forecast is scientifically uncertain?

Brady, a geophysicist with the US Bureau of Mines, believes his laboratory studies of rock failure have shown him how to predict certain types of earthquakes. After studying the situation off Peru, he predicted two major quakes near Lima, one on Aug. 10 and a second on Sept. 15.

This was the first specific quake forecast to be made publicly by a responsible scientist. Even though few other geophysicists believed it, many Peruvians took the warning seriously. The fact that William Spence of the US Geological Survey (USGS) supported Brady reinforced the warning. Even a dismissal of it by the US National Earthquake Prediction Evaluation Council last winter did not dispel Peruvian concern.

However, when expected precursor shocks did not occur in spring and early summer, Spence withdrew his support. Five weeks later, on July 20, Brady wrote Peruvian colleagues saying he too could no longer maintain the prediction.

This should be the end of the incident. But what if it had concerned a US city, say Los Angeles, and had induced people to take precautionary measures involving some disruption and expense? Would a scientist such as Brady, issuing a warning in good faith, be liable to lawsuits when he predicted disaster did not come?

C. B. Raleigh of the USGS Earthquake Prediction Program says this is a question he and many of his colleagues are asking. Although he and most other seismologists consider quake prediction very uncertain today, they foresee a growing competence. The time may come in a decade or so when enough is known for experts at least to warn of the distinct likelihood of a major quake, if not make such a specific forecast as did Brady.

As Raleigh has explained in an editorial in Science, the forecasters could be subject to damage suits if the warning turns out to be wrong. On the other hand , were they to say nothing of their forebodings and were a quake to strike, they might be sued for neglect of professional duty.

It is a legal double bind that Raleigh says could dampen the progress of earthquake forecasting by making scientists reluctant to admit when their knowledge has passed beyond the purely research and speculative stage. Specific legal protection needs to be provided, he says.

Indeed, both the Brady episode and Raleigh's concern point up a need for the US government to develop a policy and a legal framework for handling earthquake forecasts, even though the ability to make such forecasts may not yet be in hand.