Geyser Can Forecast Quakes, Woman Says : Temblors: Scientists take notice after Old Faithful’s erratic eruptions preceded three in Northern California, including Loma Prieta.

Olga Kolbek may be the only person in the world who wasn’t surprised when a major earthquake ripped through the Santa Cruz Mountains on Oct. 17, 1989.

She is convinced that the Old Faithful Geyser near the Napa Valley community of Calistoga had predicted the Loma Prieta earthquake that devastated parts of the San Francisco Bay Area, just as it had at least two other major temblors in the past 16 years. Although experts have rejected her claims for years, a scientist with the Carnegie Institution has recently concluded that she may be right.

In the days before the earthquake struck, the popular tourist attraction had not been erupting as faithfully as it normally does, and Kolbek had seen that pattern before in the records she has been keeping since 1973.

She has told seismologists about her findings many times, but “they all pooh-poohed it,” she said in a telephone interview.

But now, some of them are not so sure the retired schoolteacher is wrong.

“You have to take it seriously,” said geophysicist Paul Silver of the Carnegie Institution of Washington.

Silver and his wife, Nathalie Valette-Silver, a geochemist at the University of Maryland, were vacationing at the Calistoga hot springs two years ago when they met Kolbek, who manages the privately owned Old Faithful resort. Kolbek told them of the records she has been keeping, and the two scientists took an immediate interest.

After carefully analyzing the records, the couple concluded that the interval between Old Faithful’s eruptions lengthened significantly a day or so before three large earthquakes in Northern California, including the 7.1 magnitude Loma Prieta temblor. A similar pattern occurred just before the 5.7 Oroville quake on Aug. 1, 1975, and the 6.1 Morgan Hill quake, just south of San Jose, on April 24, 1984. The fact that the performance of Old Faithful changed just before all three major quakes in that region in the last two decades is not very likely coincidental, Silver said in a telephone interview.

“We don’t think this could have happened by chance alone,” he said.

Other experts remain highly skeptical, however.

“It seems fairly incredible,” said William Bakun, who heads a San Francisco Bay Area earthquake task force for the U.S. Geological Survey.

“Paul Silver has a fine reputation,” Bakun added. “He’s not known as a kook. But this is hard to believe.”

Among other things, Bakun is troubled by Old Faithful’s distance--120 miles--from the Loma Prieta epicenter, and scientists have long believed that earthquakes are local events. Furthermore, not one of the many instruments near the epicenter recorded any change in the Earth’s crust leading up to the quake, leaving scientists wondering what kind of processes might have been taking place that would have escaped detection near the quake but shown up more than 100 miles away at Calistoga.

And that, Silver counters, may be the heart of the problem.

“Much of the attempt to predict earthquakes has been focused on where the earthquake will occur,” Silver said. The records of Old Faithful, he added, suggest that scientists ought to be looking for evidence of large tectonic strains some distance from the fault.

Silver admits that the record is not without its flaws. Old Faithful has slowed its eruptions at some times when earthquakes did not occur, and it has not been influenced by smaller quakes.

The geyser is fed by an underground reservoir, and the eruption rate depends on pressure within the reservoir, Silver said. He believes that just before a major quake, stresses within the Earth’s crust may cause the ground around the Calistoga area to deform slightly in “tectonic strain.”

That slight deformation, he said, may temporarily increase the permeability of the rocks around the reservoir, thus allowing some of the water to escape. That in turn would reduce the pressure within the reservoir and increase the interval between the eruptions.

The Carnegie Institution’s Department of Terrestrial Magnetism, where Silver works, has installed a computerized monitor at Old Faithful so that Silver can watch its every move. He believes that the record will eventually show that Old Faithful’s fidelity is, indeed, being influenced by events leading up to major earthquakes.

If it is true, “it will really be surprising,” said Bakun.

Not to Olga Kolbek.

She said that during the years she has been running the resort, the geyser has erupted on an average of about every 40 minutes, but that changes according to the seasons, particularly during periods of heavy rainfall.

But “on Aug. 1, 1975, the geyser didn’t erupt for a 2 1/2-hour interval,” she said. “I thought maybe it was going kaput.”

“Later that day, we had an earthquake 100 miles north of here,” near Oroville, she said.

She said the geyser also performed erratically just before a 1980 magnitude 7 quake off Mendocino, but Silver believes the record there is less convincing than for the other quakes.

In those days all the records were based on visual sightings, but in 1983 an infrared sensor was installed at the geyser, and since then it has compiled a precise record of Old Faithful’s behavior.

In 1984, it acted up again, followed by the Morgan Hill quake.

And finally, the day before the Loma Prieta quake hit, the interval--which had been gradually lengthening for several days--grew to an hour and 45 minutes.

Kolbek says she wasn’t surprised when the quake struck. But since the interval was not as long as it had been before other quakes, she was surprised that it was a 7.1.

“It (Old Faithful) wasn’t forecasting such a large one,” she said.

Silver is not quite as confident of Old Faithful’s reliability.

“I don’t regard it as a practical earthquake predictor,” he said. “It would have to be reliable every time, and it isn’t, but this is important because it demonstrates that there is a preparatory phase, and out of that knowledge may come something.”

A Geyser’s Signal?

The Old Faithful Geyser at Calistoga has behaved erratically before three major Northern California earthquakes; some scientists believe its performance may be linked to stresses building in the Earth just prior to a quake, but others doubt it.

* A two-hour gap before a magnitude 6.1 quake near Oroville in 1975.

* Changes in intervals recorded the day before the 6.1 magnitude Morgan Hill quake in 1984.

* Changes in eruptions the day before the 7.1 Loma Prieta earthquake in 1989 near Watsonville.