Balancing California's quake budget.

Seismologists rarely bear good news, but residents of Southern California might welcome the iconoclastic views of David D. Jackson and Yan Kagan of the University of California, Los Angeles.

At a meeting of the American Geophysical Union in Baltimore last week, Jackson and Kagan attacked a central assumption of seismic forecasting--the idea that any individual fault segment repeatedly produces earthquakes of a characteristic size. If their criticism proves valid, the Los Angeles region may not suffer a widely expected surge in the frequency of magnitude 6.5 to 7.0 earthquakes, tremors similar in size to those that hit Northridge, Calif., and Kobe, Japan, recently.

"The good news is that perhaps we haven't just been lucky with the magnitude 7.0s. Perhaps the historic rate of earthquakes is about right, and that's what we should expect in the future," said Jackson.

This optimistic outlook does come with a price, warned the researchers. If Southern California faults fail to produce numerous magnitude 7.0 tremors, they must eventually--within a few centuries--serve up a great 8.0 to 8.5 earthquake, much larger than what most scientists have considered possible.

The UCLA seismologists reached this conclusion after testing studies that used the characteristic earthquake hypothesis for hazard forecasts. According to the hypothesis, geologists can deduce the largest earthquake possible on a particular fault segment by determining the segment's length and how much it moved in previous quakes. For the approach to work, fault segments must generate similar earthquakes over and over.

Jackson and Kagan analyzed four studies focusing on Southern California and the Pacific region. In all cases, the theory called for more quakes than had actually occurred.

Jackson played a key role in one of these studies, issued by the Southern California Earthquake Center (SCEC) in January. The SCEC report concluded that Southern California should have experienced twice as many magnitude 7.0 and larger quakes since 1850 as it has (SN: 1/21/95, p.37).

To explain that discrepancy, Jackson and Kagan reasoned that the

characteristic earthquake hypothesis sets too low a limit on the size of potential tremors. For instance, the SCEC report included the possibility of earthquakes as large as magnitude 7.8, the size of the last Big One in Southern California, in 1857. If even stronger quakes strike the region, that would reduce the number of smaller ones--because it takes 30 magnitude 7.5 tremors to equal the energy of a magnitude 8.5 jolt.

Jackson and Kagan envision fault segments acting randomly instead of producing characteristic earthquakes. Sometimes several segments join together to create a great quake, other times they break independently. The researchers even question whether faults have identifiable segments. To test their theory, Jackson and Kagan plan to look at quake frequencies in other parts of the world.

Such arguments have not won over David P. Schwartz of the U.S. Geological Survey in Menlo Park, Calif., who formulated the characteristic earthquake hypothesis in the early 1980s. "The overall body of observations is in support of there being characteristic earthquakes," he says.

While Schwartz agrees that some fault segments can join to produce larger earthquakes, he maintains that most faults cannot produce the giant jolts called for by Jackson and Kagan. If Schwartz is right, regions such as Southern California cannot escape an increasingly shaky future.