Page 4 of 12
Where predictions like these leave me is with a shock of possibility, an almost physical sensation that makes me wonder what I'm looking at, as if I can't quite see the connections even though I feel them all the same. It's here, however, that the split between science and superstition opens up like a deep ground fissure, a vast and insurmountable seismic divide. The USGS, after all, refuses to give such predictions any credence; even Curtis, who describes herself as having "a very open mind about the whole thing," says that although she believes some people may be sensing something, it's vague, "like a vibe or a vibration," and not really useful, while the seismologists, for their part, dismiss the predictors and their theories as a waste of time. Perhaps no one is as vehement on the subject as Lucile Jones, the Pasadena office's scientist-in-charge, who is best known for the late-night press conference after the 1992 Joshua Tree earthquake at which she spoke to reporters while holding her own tired and frightened child in her arms. More than anything, Jones believes, this image has made her a lightning rod for various earthquake obsessives, as if she were a benign, maternal figure who might listen to their seismic fantasies -- or, as she puts it, with something between a smile and a grimace, an "earthquake mom."
Jones may be right about that; most of the predictions in Linda Curtis' archive are addressed to her, and many assume a revealing tone of familiarity, referring to her as "Lucy" or "Dr. Lucy," as if they'd been written to a friend. Some go even further -- during the late 1980s and early 1990s, she was the target of a bizarre series of letters, cards and hand-drawn maps, often inscribed in an eerie invented language, from a self-proclaimed prophet named Donald Dowdy who predicted all-out seismic apocalypse with himself as the trigger. The situation came to a head in March 1993, when Dowdy sent Jones a map of Los Angeles on which he'd written, "One of these days, I'm going to cut you into little pieces," a lyric from the Pink Floyd song "One of These Days." After the FBI was called in to investigate, Dowdy wrote Jones to explain that the lyric referred to "THE BIG ONE" and that he meant her no harm, but the memory still leaves a bitter taste. "I see prediction as having to do with severe psychological issues for a lot of people," Jones says, sitting at a conference table in her large, neat office on the second floor of the USGS. "On the one hand, it's about the need to take control, to exert control over an uncontrollable situation. But the truth is that we deal with hundreds of these people, and most of them get obsessed with their own theories and only want to be proven right. They don't want to have to measure their ideas against the same standards that I do, which contributes to a lot of misinformation about how earthquakes work."
The ironic thing about Jones' comments is that, as she'll freely tell you, even the best-informed seismologist would be hard-pressed to explain the way an earthquake works; it's one of the great unresolved geologic mysteries, and a major reason why many scientists doubt prediction will ever be a viable tool. This, in turn, suggests another irony, since throughout the 1970s and early 1980s, prediction was seriously pursued and was considered an achievable goal. The most famous prediction story is that of the 7.3 earthquake that struck Haicheng, China, on February 4, 1975; nine and a half hours before the shaking started, an alert was issued by the provincial government, which based its warning on a number of factors, including several days of foreshocks, as well as "changes in ground water which were usually changes in the level or color of the well water . . . [and] the appearance of a low ground fog." Yet if Haicheng represents the first documented, officially predicted earthquake, the success has not been replicable, and despite a wide range of experiments, it remains the only one. More common are the USGS's experiences in places like Palmdale, where in the mid-1970s a large ground deformation called the Palmdale Bulge was briefly thought to indicate mounting stress on the San Andreas, or Parkfield, a town in Central California where magnitude 6 temblors were found to take place at roughly 22-year intervals -- until, that is, the Survey moved monitoring equipment into position and, between 1988 and 1993, declared a series of earthquake warnings, none of which ever panned out.
Part of the problem with prediction, according to Jones, has to do with people's expectations; for this reason, Parkfield, despite yielding much valuable information, is considered a failure, while the Chinese efforts, which have never again lived up to Haicheng, are regarded as a success. "I could predict unlimited earthquakes in Southern California," Jones says, "as long as I allowed myself an unlimited number of false alarms. That's what the Chinese do; they don't think false alarms are relevant. But I don't believe, in Los Angeles, you can have false alarms without losing credibility. Two false alarms and you're dead meat." Jones has a point, but as always with geology, there's more than one side to the story, and for a sense of that, all you need to do is talk to Paul Silver, a staff scientist at the Carnegie Institution of Washington, who continues to seek a viable prediction model, focusing on "precursors," like those seen at Haicheng. Silver is an experienced observer of such phenomena; in 1992, he co-authored a paper tracing "the co-seismic response" between Calistoga's Old Faithful Geyser and several Northern California temblors, including the Loma Prieta earthquake that shook the San Francisco Bay Area in 1989. According to his data, one to three days before an earthquake, ground-water variations alter the geyser's eruption intervals, which return to normal once the earthquake has passed. "It's suggestive of a causal connection," Silver says with classic scientific understatement, but while he hesitates to give this too much import, he does have some ideas about what it means. He and many of his Carnegie Institution colleagues, he admits later, believe that earthquakes can be predicted: "At this stage we're nowhere near that, but even if Haicheng is the only verifiable prediction, one is more than zero. It has been done."