The Aftershocks That Weren't

New evidence from California --and old reports from the Midwest--indicate that some earthquakes can trigger others hundreds of miles away.

Before dawn on June 28, 1992, residents of southern California were jolted from their sleep by a magnitude 7.3 earthquake. The temblor started beneath the tiny Mojave desert town of Landers, but its effects were felt in Los Angeles, ninety miles to the west, as well as in northern California, southern Nevada, and at the Mexican border.

Earthquakes are not unusual for California, of course. The state is riddled with both large and small faults at the boundary between two tectonic plates: the North American Plate and the Pacific Plate. These two enormous pieces of the earth's crust are constantly sliding past and grinding against each other, creating an active and complicated boundary zone. Sections of faults often rupture here, earning California its reputation as earthquake country. So the powerful Landers temblor was not unusual in itself. Its far-reaching aftereffects, however, revealed some surprises.

Nearly half a century ago, seismologist Charles Richter, of Richter scale fame, observed that "an earthquake of consequence is never an isolated event." He was referring to the aftershocks that invariably follow a large temblor, or main shock. As earthquakes go, aftershocks are remarkably well behaved. In time, space, and magnitude, they follow predictable patterns that have been recognized by seismologists for decades. As a general rule, they are small, more frequent in the immediate aftermath of a main shock, and usually clustered no farther from the site of the main shock than the length of the rupture it created.

The Landers temblor, however, permanently changed the established view of earthquake sequences. In the minutes and days that followed the quake, a substantial number of smaller seismic events occurred well beyond its aftershock zone--as far away as the Lassen Peak area and at the Geysers, about sixty miles north of San Francisco. The largest of these distant events, with a magnitude of 5.4, struck in western Nevada twenty-two hours after the Landers event. Quickly dubbed remotely triggered earthquakes, these outlying seismic events seemed to be different beasts from anything seismologists had previously encountered.

Because the Landers quake was one of the first magnitude 7.0 events to be recorded by state-of the-art seismic stations throughout California, it yielded the first solid evidence of distant earthquake triggering. The authors of one research paper commented that "no previous experience would have led us to anticipate the observations of remotely triggered [earthquakes]." Earth scientists began to realize that large quakes could perturb the crust in complex ways that earlier theories could not explain.

Almost immediately after Landers, researchers began to focus on these puzzling quakes. Not only were they located too far from the main shock to be classified as aftershocks, they also couldn't even be associated with the same physical mechanisms. As the Landers earthquake helped seismologists understand, conventional aftershocks appear to result primarily from local, mechanical factors--changes in stress caused by large-scale movement of a crustal block. But as seismologist Joan Gomberg and others have shown, the outlying earthquakes of 1992 were apparently set off solely by waves, or shaking, emanating from the Landers main shock. Such waves, it seems, were capable of engendering--that is, triggering--new quakes as far as 700 miles away.

Using data not only from Landers but also from Greece and Mexico, seismologists attempted to characterize the geology of vulnerable areas and to define a threshold level of shaking needed to trigger remote events. The first models implicated something curious: bubbles. Most of the triggered quakes identified in the immediate aftermath of Landers had taken place in regions with active volcanoes or geothermal features such as geysers. A couple of models showed that disruption of the bubbles within fluid reservoirs can indeed raise stress on faults, which, the researchers argued, could then set off earthquakes. According to this scenario, triggered earthquakes were rather exotic events occurring only in highly unusual geological zones.

Last year, however, I took a new look at evidence from an unlikely time and place, and my investigation has revealed that the triggered earthquakes of 1992 were not as unprecedented--or as exotic--as they first seemed. Nearly two centuries ago, settlers along the Ohio River valley wrote reports that (had they been fully appreciated) might have helped scientists interpret what happened after the Landers earthquake. In the wee hours of the morning on December 16, 1811, the usually stable U.S. midcontinent was rocked by an earthquake so large that it caused damage as far away as coastal South Carolina. Now estimated to have had a magnitude upwards of 7.3, the temblor was the first event in an extraordinary sequence. Centered at the southeast corner of Missouri--the Bootheel region, adjoining the Tennessee-Kentucky border--the sequence was named New Madrid, after a town close to the event. The area experienced two more quakes after December 1811: events of comparable magnitude jolted the region on the mornings of January 23, 1812, and again on February 7, 1812. Historical accounts peg the final quake as the largest of the three; some refer to it as the "hard shock." Immediately following it, a section of the Mississippi River in the vicinity of New Madrid reversed course for several hours.