In February 2003 a Web site called geoforecaster.com sprang up, offering regular earthquake predictions to its subscribers. Within a week, the complaints of scientists prompted the California Board for Geologists and Geophysicists to accuse the firm of practicing without a license. What caused such an instant scientific allergic reaction?
Few things in Earth science are as hotly desired as the ability to predict earthquakes. Few things have been harder for geologists to do. And few things have made more trouble for geologists than people actually making earthquake predictions.
Psychics, cranks and self-deluded people are always doing it. I get regular emails from people predicting quakes. I treat them like spam. Charles Richter wrote in 1977, "Since my first attachment to seismology, I have had a horror of predictions and of predictors. Journalists and the general public rush to any suggestion of earthquake prediction like hogs toward a full trough." Nobody in the research community issues predictions. History tells us why in the notorious cases of Bailey Willis, Brian Brady, and Iben Browning.
One of California's greatest geologists, Bailey Willis (1857–1949) led a vigorous public campaign in the 1920s to raise awareness of earthquake hazards and safe building practices. As president of the Seismological Society of America, he spoke out often. "It will be well for architects to respect His Majesty, the Earthquake, in making their designs," he wrote. And indeed, after the Santa Barbara quake of 29 June 1925 some cities enacted better building codes. But action was too slow for him, and within a year Willis went overboard, telling audiences that he expected a severe quake in Southern California within ten years, even five. Time magazine didn’t help by exaggerating even that claim.
The insurance industry grew concerned. Los Angeles real-estate promoters saw red. One of them, a prominent trustee at Caltech, warned his peers that if scientists would not "stop their talk about the earthquake problem I for one am going to see what I can do about stopping the whole seismological game." Soon a powerful builders' association discredited Willis on scientific grounds, triggering his public and professional humiliation. Willis withdrew from seismology, and progress in building codes was stalled until the 1933 Long Beach earthquake threatened the city's schoolchildren. Today, structural geologists still revere Willis as a founding father, but Stanford University, where he led the geology department, scarcely notes his tenure there now. (Much of my information on this episode comes from C.-H. Geschwind's California Earthquakes.)
Brian Brady was a researcher for the U.S. Bureau of Mines who in 1976 predicted that enormous offshore earthquakes would occur near Peru in 1981 and 1982. His work was based on laboratory experiments, and seismologists were deeply skeptical. Still, ordinary people and government officials were worried, and South American seismologists had to expend much effort to calm the flames. When the large foreshock he had predicted did not occur Brady withdrew the rest of his prediction, but a delegation of U.S. scientists still had to fly to South America to deal with the questions from the government and press. The research community was reminded again that just like real earthquakes, false alarms can be dangerous.
The Browning Panic
Iben Browning was not a geologist at all but a retired biologist. Nevertheless, he issued a startling prediction: within two days of December 3, 1990, he said, there was a 50-50 chance that a magnitude-6.5 quake would strike the New Madrid region of Missouri. A remarkably energetic huckster, Browning raised such a public panic that businesses closed and schools in four states shut down during early December. Children and parents were needlessly alarmed, and much ink was spilled and scientists' time spent combating nonsense. Browning's theory of tidal forces—the same one used by quake predicter James Berkland—has never had significant support from the data.
The big lesson that geologists took from the Browning fiasco was that cranks must be counteracted swiftly and seriously. The geoforecaster.com case was an example of that response, even though the site's owners had far better credentials, and more real science behind them, than Browning did. In this case, science enlisted the state in suppressing quake prediction. More about state licensing of geologists on page 2.
PS: While seismologists in general turned away from prediction research in the mid-1990s, some believe the field is still promising despite its perils. Christopher Scholz, as prestigious as anyone in the field today, pressed this view in Geotimes in 1997, asking "What Ever Happened to Earthquake Prediction?"
Willis image from Géomanips site, CNRS France.