As the third jolting earthquake within four hours shook San Diego the evening of June 17, again sounding an alarm 100 miles north at the U.S. Geological Survey seismic lab in Pasadena, duty scientist Tom Heaton got on the phone and jarred his boss awake in Washington.
The developing pattern in San Diego disturbed Heaton and his colleagues in their lab at the California Institute of Technology.
San Diego County traditionally has had few earthquakes and never a pattern of quake swarms such as those that occurred that day. The intensity of the quakes increased between the first and second events. The locations appeared to be directly under the metropolitan area, where significant damage could take place if a major shake were to follow.
The USGS group had in their hands a recent study of Southern California earthquakes: since 1931, 6% had been followed by larger incidents within five days.
Based on that study, and their uncertainty over what was happening in San Diego, the scientists decided to issue their first-ever statement estimating the short-term chances of a large earthquake taking place.
Heaton called California disaster officials in Sacramento the same night with an estimate that San Diego stood a 5% chance, within five days, of experiencing a quake of Richter scale magnitude greater than 5, which is capable of causing damage if located near an urban center.
The consequences of those telephone calls were an educational experience for government agencies at all levels involved in earthquake emergency planning.
The USGS learned that issuing any type of statement about earthquake probability will be viewed as a prediction, whether the scientists judge it a true prediction or not.
State emergency officials learned that such statements must be made public and the meanings explained, even if they would prefer to keep the information in-house for emergency planning.
And county disaster planners in San Diego and elsewhere learned that they must have information as quickly as possible to notify fire and police agencies and utilities, and that there must be a set plan for getting the public to automatically take elementary precautions after a warning.
"We need to have the same type of skills for earthquakes in California as we have for tornadoes in the Midwest," said William Ellsworth, chief of the USGS seismology branch. "And since you don't always have warning signs before a quake hits, you've got to have care and planning ahead of time."
Dan Eberle, manager of the San Diego County Office of Disaster Preparedness, added: "It was beneficial in boosting the general awareness of the public, but hopefully next time we'll get the information faster."
The USGS based its statement on a 1984 paper by staff seismologist Lucile Jones, who examined every quake in Southern California since 1931 that reached 3.0 on the Richter scale or greater, to see whether they had been followed by a larger quake. (The Richter scale, which measures the magnitude of earthquakes, is open-ended, with higher numbers representing logarithmic increases in intensity. That is, a 5.0 quake is 10 times stronger than a 4.0.)
In order for the first quake to be designated a foreshock (a minor quake preceding a greater one), the larger quake had to occur within five days and be located within six miles (10 kilometers) of the first.
"I found that 6% of all quakes were followed by an event larger than the first and--while a small number--it is a higher percentage than we expected," Jones said.
One-fourth of the second quakes took place within the first hour of the foreshock, and three-fourths took place within 24 hours after the initial temblor, Jones found. Half of the larger quakes were within a Richter point of the first; that is, if the first was a magnitude 4, chances were 50% that the second would be between magnitude 4 and magnitude 5, she said.
On June 17, San Diego was hit with a 4.0 temblor at 5:12 p.m., followed by a 4.2 quake at 8:22 p.m., and a third quake measuring 4.0 at 9:28 p.m. It was that third quake that precipitated Heaton's call to the USGS in Washington and the notification of state officials in Sacramento about an hour later.
"So you can see that since an hour passed by the time we issued a statement, the chance of a magnitude 5 was already down to 2.5%," Jones said. "But since we had a swarm of three events, we arbitrarily doubled the probability to 5%."
Jones' study did not look at the percentage of major quakes preceded by a swarm of foreshocks. However, she said that enough cases exist of large quakes following several foreshocks to believe there was a greater chance for an incident in San Diego.
"In this case, it was appropriate for the USGS to comment and say we want to alert you, based on a statistical model," Ellsworth said, because that kind of activity in San Diego had not been seen previously.