Historical clues from the San Andreas.

The Loma Prieta earthquake of 1989 caused major damage in the San Francisco Bay area, but its force paled in comparison with the great 1906 shock in that region. Geologists have now uncovered signs that a comparable megashock struck the Bay area around 1650. The finding may help scientists predict when the next great quake will shake this segment of the San Andreas fault.

Trenches dug across the San Andreas fault in the Santa Cruz mountains reveal evidence of two earthquakes in this region: the 1906 event and one from the mid-1600s, reports David Schwartz of the U.S. Geological Survey in Menlo Park, Calif. A trenching study conducted at Point Arena, several hundred kilometers to the north, also shows signs of two quakes from the same periods. At both sites, scientists used carbon-dating techniques. The evidence from these two widely separated sites suggests that the entire length of the fault between the Santa Cruz mountains and Point Arena may have broken at once in a great earthquake around 1650, Schwartz says.

When this long stretch of the San Andreas ruptured in 1906, the resulting earthquake reached an estimated magnitude of more than 8.0. But scientists have never before found geologic evidence of an earlier great quake along the same section. The new work suggests that such megashocks may recur every 250 years or so - a figure that agrees with mathematical calculations of the time between large shocks in this region

The evidence does not prove that a large earthquake ruptured this entire stretch at once, Schwartz cautions. Carbon dating is not precise, so different sections of the fault may have ruptured separately If so, the Santa Cruz segment may have slipped in a strong- but not "great" - quake in 1650, followed 20 years later by another strong quake near Point Arena.