No fault in Parkfield prediction.

No fault in Parkfield prediction

Parkfield, the tiny California town that straddles the San Andreas fault, is where seismologists are waiting for an earthquake to happen (SN:4/13/85,p,228). Because magnitude 6 earthquakes have rattled the Parkfield fault segment about every 22 years since 1857, and since the last quake was in June 1966, scientists predict that the next quake will occur in 1988, with a five-year margin for error.

Some recent studies, however, have cast doubt on this forecast. At the meeting of the American Geological Union last December, two papers suggested that possible errors in the sites of quakes prior to 1922 mean that the Parkfield "clock" is not as regular as some think.

But now, in a paper in the Sept. 26 SCIENCE, two researchers at the U.S. Geological Survey in Menlo Park, Calif., offer another line of evidence in support of the 1988 prediction. Paul Segall and Ruth Harris studied geodetic lines spanning the San Andreas near Parkfield. They used changes in the Lengths of these lines, which have been monitored since 1959, to calculate the the amount of strain in the fault before and after the 1966 quake. By comparing the strain that has built up since the 1966 quake with that released by the quake, they hoped to provide a test of the Parkfield prediction that does not rely on the periodicity of earthquakes.

"Our results suggest that the strain released in the 1966 earthquake will most likely recover between 1984 and 1989," they write, "although it is possible that this will not occur until 1995." When it is recovered, "sufficient elastic strain will be stored for a [magnitude 6] earthquake to rupture the Parkfield fault segment."