New theory predicts less shaking at faults

SAN JOSE, Calif. — For years California has been building to withstand a major earthquake, pouring billions into retrofit projects to keep vital structures from collapsing should the Big One strike.

What if some of that money could have been saved?

Drawing on data from major earthquakes around the world — including the 1999 earthquake in Turkey, the 1999 Chi-Chi quake in Taiwan and the 2002 Denali quake in Alaska — scientists have determined that the ground near a fault doesn't shake as violently during a big event as they once thought. In fact, the seismic waves may be at least 25 percent less intense.

It's a controversial theory that could have far-reaching implications.

If the ground motion patterns in those faraway events apply to California, it might mean building codes for some structures could be relaxed.

But while engineers would love to save money for their clients by building to lower standards, they're conservative by nature and likely to resist any dramatic revision of the codes.

"There's an issue of science and an issue of politics," said Norm Abrahamson, an engineering seismologist at Pacific Gas & Electric who came up with new models for ground motions in major earthquakes. He presented his surprising results last week at a San Francisco conference commemorating the 100th anniversary of the 1906 earthquake.

The U.S. Geological Survey, which is responsible for creating earthquake hazard maps that are used to set building codes, has been holding workshops to discuss the new models.

It's applied them to a new hazard map for the Pacific Northwest, resulting in a 15 percent decrease in expected shaking for Seattle, said Arthur Frankel, coordinator of earthquake-effects research for the USGS.

The survey will do the same for California in September, and in November will meet with engineers to discuss the implications of any changes.

"There will be a lot of back and forth and a lot of debate," Frankel said.

"We want to have the wider community look at these and judge them: Do we use them? Give them partial weight? You don't want to understate the hazard and not provide the level of protection — or over-estimate, which would not be cost-effective."

In related work presented at the conference Tuesday, engineering seismologists Ivan Wong and Walt Silva used the new models to simulate what could happen in a repeat of the 1906 quake on the San Andreas Fault.