Dilemma at Quake Center: Just When to Sound a Public Alarm

By JAY MATHEWS

June 12, 1988 12 AM PT

The Washington Post

PASADENA — Lucile Jones, the U.S. government’s chief earthquake watcher in Pasadena, had been on duty since 2 a.m. worrying about a series of small quakes moving closer to the huge San Andreas Fault. At 5:16 a.m. the needles swung off the scale, signaling that a 6-magnitude temblor had occurred near a spot where more than 150 years of seismological stress was waiting to be unleashed.

The date was Nov. 24, 1987. Was this the beginning of the “Big One,” the long-expected 8-magnitude earthquake that might pulverize Southern California? Should the public be warned?

Leap-frogging her immediate superiors, Jones telephoned John Filson, chief of the office of earthquakes, volcanoes and engineering for the U.S. Geological Survey in Reston, Va., and poured out a torrent of numbers and their possible implications.

“Why don’t you calm down, Lucy. Take a couple of breaths,” Filson said. He had had such conversations before.

Twice before, despite widespread speculation that public earthquake warnings might foment rumors, panic and chaos, Filson had agreed to pass on information that local authorities used as public forecasts, one wrong, one almost right. Both were received with no more panic than word of impending showers on the morning weather report. What would he do this time?

With such transcontinental conversations between federal earthquake scientists, the world has slipped into a new era of seismology as public policy, the beginning of a time when specialists hope that potentially lethal jolts in the Earth’s crust will be predicted like hurricanes and lives will be saved.

“There are two myths that have been fairly well dispelled,” said Walter Hays, deputy chief of research applications in the survey’s office of earthquakes, volcanoes and engineering. “One is that there would be panic” after a public earthquake forecast. “The other is that ordinary people couldn’t handle all the information you might give them,” he said.

The age of short-range earthquake forecasts dawned on June 18, 1985, a date unfamiliar to all but a few scientists and government officials and the few residents of San Diego who noticed.

On the afternoon and evening of the previous day, a small but unusual swarm of earthquakes occurred beneath San Diego, including temblors of magnitudes 4, 4.2 and 4 on the Richter scale within five hours of each other.

Tom Heaton, U.S. Geological Survey scientist in charge in Pasadena and Jones’ supervisor, was aware of her recently completed statistical study analyzing every significant quake sequence in Southern California since 1932. Jones had concluded that there was a 6% chance, about 500 times the usual probability, of a moderate earthquake occurring within five days and six miles of any temblor of magnitude 3 or above.

The San Diego quakes were stronger than magnitude 3. Moreover, the area had experienced little seismic activity for several years. And the quakes’ epicenter was directly under a major city. Heaton woke Filson with a telephone call at about 1 a.m. Virginia time and outlined the situation. After lying in bed thinking for half an hour, Filson called back and told Heaton to relay the information to state authorities.

Both men agreed that withholding the analysis would be irresponsible. Although Filson could not authorize a public warning without the survey director’s approval, the state could issue a warning independently. Like nearly all seismologists then, Filson said, he was worried about creating panic. “We didn’t want cars jamming Interstate 5 trying to get out of town,” he said.

At 10:45 p.m. on the West Coast, Heaton sent this message to the state office of emergency services in Sacramento: “There is a slight increase in the probability of a potentially damaging earthquake in the San Diego area. There is a 5% chance of a 5.0 magnitude or greater earthquake in the next five-day period.” The message was labeled “For Internal Use Only.”

Filson intended only to alert local civil defense offices and fire departments. But by morning, San Diego officials were discussing wider dissemination of the information. That afternoon “we had a 45-minute brainstorming session,” said Steve Danon of the San Diego County office of disaster preparedness.

The scientists were not using the word prediction but, Danon said, “we decided to hell with it. We treated it as a prediction, and that was exactly how we went with it.” Some officials discussed potential lawsuits if a warning caused panic or if a failure to warn caused injury, but the issue of liability was put aside.

By early evening, county authorities had warned local governments and news organizations, repeating the language of the overnight message and noting that the prediction was based on “historical data.”

“One in 20 of these swarms has resulted in a damaging earthquake of 5.0 or greater,” it said, putting the figures in betting terms.

Radio and television stations carried the advisory without much comment. Newspapers, still focusing on reaction to the earthquakes of the day before, barely mentioned it. A local city editor said he does not remember anyone calling it the nation’s first official short-term prediction.

No earthquake materialized in the five-day period, but officials said they thought they had made an important discovery.

Did Not Spread Panic

“It did not spread panic. It was very well received,” said Richard Andrews, assistant director of the state office of emergency services. “It gave us some confidence that this kind of system might work.”

Jones’ statistical analysis of Southern California earthquakes, published later that year in the Bulletin of the Seismological Society of America, became a guide to planners considering earthquake warnings. But she acknowledged that she could not tell a foreshock from an isolated quake, and her data indicated that warnings might not be made in time.

“The main shock is most likely to occur within the first hour after the occurrence of the foreshock; 26% of the main shocks occur within that time,” her report said.

That information was particularly unnerving to Jones early on Nov. 24, 1987, when seismometers registered a 6.3-magnitude quake that seemed to be at the vulnerable southern end of the San Andreas Fault.

“Segmentation” is the tendency of faults to break in sections rather than along their entire length. Research indicates that the San Andreas Fault is likely to begin its next major shift at one of only a handful of points along its 600 miles. Bombay Beach at the southern end, on the northeast shore of the Salton Sea, is one such point.

A 6.0 temblor had hit the southern end of the Salton Sea at 5:53 p.m. the previous day, and the aftershocks seemed to be moving northeast toward Bombay Beach. The members of the California Earthquake Prediction Evaluation Council conferred by telephone, deciding, for the time being, just to watch for activity closer to the main fault.

The geological survey and Caltech scientists agreed to staff their monitoring equipment through the night, but the Caltech laboratory’s outmoded computer could not handle data as rapidly or efficiently as newer ones. Seismologists had to use rulers and the naked eye to calculate epicenters, causing a 15- to 30-minute delay before they knew exactly where a large earthquake had occurred. Hundreds of smaller aftershocks had to be ignored.

Jones went to work at 2 a.m. to relieve Kate Hutton, a senior Caltech seismologist who was usually the first to arrive at the laboratory after any major quake and was weary from long hours of calculating by hand.

‘Here We Go’

By 5:16 a.m., when the 6.3-magnitude quake hit, Jones was weary herself. “Oh, hell,” she remembers saying aloud. “Here we go.” She knew it would be several minutes before Caltech technician Riley Geary could tell her how close the temblor came to the San Andreas Fault. The previous line of foreshocks pointed toward the Salton Sea town of Niland, uncomfortably close to the end of the huge fault.

“I was freaked out,” Jones recalled. She telephoned Filson, then took a few breaths at Filson’s suggestion while they reviewed the possibilities and waited for Geary. Ten minutes later, he dashed into the office shouting, “It’s not Niland! It’s not Niland! It’s Superstition Hills”--to the south, 25 miles from Niland, 30 miles from Bombay Beach--too far away to trigger the Big One.

Filson would not have to authorize a warning that day.

In part because of a surge in California earthquakes last fall, Heaton has secured a new $45,000 computer system he hopes will eliminate delays in processing information from the hundreds of seismometers monitoring the ground throughout Southern California. But even that system has technical problems.

With quicker information will come more discussions over when, how and why to warn the public. The San Diego advisory, Heaton and Danon said, was designed simply to put emergency crews on alert and remind residents to have flashlights, batteries, emergency food and water in stock. “Maybe people would also know that they should be strapping down their water heaters,” Heaton said.

More thoroughgoing warnings--when to close schools and businesses, for instance--will not be considered until further experiments bear fruit, Andrews said.

“We don’t feel now,” he said, “that the scientific ability to predict earthquakes is such that we can justify suddenly altering personal daily life.”