Cracking open the center of a temblor.

Cracking open the center of a temblor

Like expectant fathers of yesteryear, confined to pacing in the waiting room, seismologists cannot actually witness the birth of an earthquake. Instead, they must rely on information gathered by listening devices placed on the ground surface, far from where the action occurs.

Despite this handicap, a new generation of instruments monitoring northern California's Loma Prieta quake has given researchers their best glimpse yet of a fault beginning to rupture.

The instruments, called broadband seismometers, can record both long- and short-period waves created by earthquakes. Unlike other monitoring devices, which become overwhelmed almost instantly by intense shaking, they can continue to record seismic waves when a quake hits nearby. California scientists have deployed these seismometers for several years, but the Loma Prieta temblor of October 1989 was the first large quake to hit an area equipped with them.

The devices that recorded Loma Prieta are among the simpler broadband instruments available today, and they caught only the first few seconds of the seismic waves before succumbing to the shaking. But even that short span of data has provided unprecedented information, says David Wald of the California Institute of Technology in Pasadena. He and his colleagues presented their findings in San Francisco this week at a meeting of the Seismological Society of America.

The researchers report that Loma Prieta started with a perplexing prologue of weak, long-period vibrations. Wald describes the quake as beginning with a low groan that eventually turned into a loud chorus of both long- and short-period shaking. Scientists are still debating what happened at the beginning of the quake, says study coauthor Thomas H. Heaton, a seismologist with the U.S. Geological Survey in Pasadena. At first, the researchers thought the weak initial waves might represent a small shock preceding the magnitude 7.1 main shock. But the long-period trembling doesn't resemble a normal small earthquake, which has many more short-period vibrations, Heaton says.

Loma Prieta's type of preliminary shaking may not be unusual. Other large California earthquakes have started off with small precursor shocks immediately before the violent shaking. But because these occurred before the use of broadband instruments, it's not clear whether they had similar long-period vibrations, says Heaton.

The kind of information collected by broadband instruments "is very important for understanding how the rupture starts," says Hiroo Kanamori, a seismologist at Caltech.

The predominance of long-period waves during the first 2 seconds suggests that the fracture initially spread smoothly and slowly from the earthquake hypocenter, located about 18 kilometers below the surface. Then, Wald proposes, the rupture grew more erratically, stopping and starting in a motion that produced many more short-period vibrations.