Quake Warnings Years Off

Despite eight years and several hundred thousand dollars spent on study, Southern California remains years away from developing an early warning system for earthquakes, according to quake scientists.

In the mid-1990s, many public officials and scientists were optimistic about developing such a system for Southern California as early as 2000. The optimism stemmed from the success of a warning system in Mexico City that gave 50 seconds’ notice of a major 1995 temblor 190 miles away in the Pacific.

But a shortage of government funds and a series of second thoughts have delayed the project. Any warning system is still at least five years away, and perhaps more, said Caltech seismologist Egill Hauksson, who has been studying the idea.

The notion of an early warning system for quakes goes back to at least 1868, when a doctor wrote to the San Francisco Examiner, suggesting that by making use of the time it takes an earthquake’s waves to travel, authorities could issue a warning that would allow “all to escape from dangerous buildings before it reaches us.”

Supporters say an early warning system could let utilities shut off vulnerable systems, building managers halt elevators and residents take cover.

In a study funded by the Federal Emergency Management Agency and the state Office of Emergency Services, officials from the Los Angeles Department of Water and Power and the Metropolitan Water District told researchers that early warnings could allow them to shut off power lines before ground motion caused them to slap together. Others suggested automatically opening the garage doors of firehouses before shaking could collapse them.

But the study, which was completed in 2001 but never released, also includes warnings that early notification of a quake could cause panic and lead to lawsuits. Los Angeles County Fire Department officials told researchers, for example, that warnings given to the public could cause worried callers to overload the 911 system and “prevent real emergency calls from getting through.”

At Los Angeles International Airport, the study reported, officials expressed concern that “if people were to run out onto the airfield” to escape terminals after a warning, “it would constitute a major breach of airport security, and the entire airport would have to be shut down regardless of whether the earthquake caused any danger.”

The study cost $400,000 and was guided by Caltech. A copy was provided to The Times by Paul Flores, a former official of the state Office of Emergency Services and one of the consultants who helped prepare the report.

A decline in national and state funds has pared annual spending on early warning research to $150,000. Hauksson said $5 million to $10 million a year would be needed to fully develop and maintain an early warning system.

A successful warning system would have to solve several problems. One is that effective early warning is possible only when a quake begins a substantial distance from the urban areas that it would strike.

If a quake originated in the Imperial Valley, for example, the shaking would take roughly 70 seconds to travel 200 miles to Los Angeles. That would allow some time to act if sensors were operating and the alarm were sounded automatically.

By contrast, the shaking from the 1994 Northridge quake reached downtown Los Angeles in 8 seconds -- not enough time to allow any reasonable warning. To be effective, a warning has to allow not only enough time for action, but also some amount of time to overcome people’s initial skepticism that a warning is genuine, researchers say.

Another problem with early warnings is that researchers still cannot determine a quake’s magnitude in the first seconds. That creates the possibility of false alerts that could quickly undermine confidence in warnings and lead people to ignore them.

Sensors also have to be located in the right areas to detect a quake. In last month’s large Mexican earthquake, which hit hardest in the Pacific Coast states of Colima and Jalisco, no warning was sounded in Mexico City because there were no sensors operating near the epicenter.

Just how and to whom any warning should be disseminated also have become major issues.

Officials at Caltech, where scientists have performed much of the early warning research, were taken aback by warnings that the school could be exposed to lawsuits if it were involved in issuing warnings.

Some public agency would probably have to be involved in issuing the warnings, say those who have studied the legal issues.