ANCIENT QUAKES ON SAN ANDREAS STUDIED

Caltech geologists studying the prehistoric sediments of ancient lakes and riverbeds straddling the southern San Andreas fault have found that different segments of the fault rupture in major earthquakes on quite different time scales.

In papers delivered December '81 at the annual meeting of the American Geophysical Union in San Francisco, researchers Kerry Sieh and Ray Weldon described the results of studies along the San Andreas fault at two sites-near Indio and at Lost Lake near San Bernardino (see map). In one paper Dr. Sieh, an assistant professor of geology, described findings showing that near Indio the fault has not produced a great earthquake in at least 560 years.

Map

In sharp contrast, as graduate student Weldan and Dr. Sieh reported in another paper, the Lost Lake site revealed recurrence of large earthquakes about every one-and-a-half centuries, for the last six to eight events. Importantly, studies at the Lost Lake site confirmed earlier assessments that the latest great earthquake on the San Andreas fault in southern California-in 1857-almost certainly did not involve rupture of the fault as far south as San Bernardino. However, previous earthquakes on this stretch did.

The Lost Lake findings are similar to those reported by Sieh in 1978 and 1980, in which excavations at Pallett Creek and Wallace Creek farther northwest along the San Andreas showed recurrence intervals of around one to two centuries.

"We still need further excavations near India to date the last few large events in that region," said Dr. Sieh. "However, we may have a situation in which the southernmost segment of the San Andreas fault -between San Gorgonio Pass and the Salton Sea breaks very rarely, while the segment northwest of San Bernardino breaks more often."

"One might also conclude that ten meters or so of slip is stored up across this southern segment," said Dr. Sieh, "because other studies suggest that the region is being deformed at a rate of about two centimeters per year. This amount of potential slip is approximately that known to occur for the largest earthquakes along the San Andreas fault."

Dr. Sieh suggests several scenarios for the next great earthquake along the San Andreas fault. A repeat of the 1857 earthquake might occur, in which the portion of the San Andreas from near San Luis Obispo to San Bernardino might break. Or, the long-dormant stretch from San Bernardino to the Salton Sea might fail. Alternatively, some shorter portion of these segments of the fault may rupture. In a more extreme case, however, the entire locked section of the San Andreas, from San Luis Obispo all the way to the Salton Sea, might break in one event. The shaking from such an earthquake, said Dr. Sieh, might not be much greater in Los Angeles than if only a portion of the locked section ruptured. However, he said, such an earthquake would severely shake a much larger region, including all of the heavily populated regions in southern California that are adjacent to the fault.

"These studies, as with other studies of the San Andreas, show that we in southern California must continue to take quite seriously our efforts at earthquake hazard reduction," said Dr. Sieh.

Dr. Sieh's studies near Indio involved excavation into the deposits of several ancient lakes. The lakes were created centuries ago when the Colorado River overflowed its banks near what is now Yuma, Arizona, and flowed northward, flooding low-lying areas of the Coachella, Imperial, and Mexicali valleys, including the sites of Indio, El Centro, and Mexicali. Radiocarbon dating of plant remalns in the sediments revealed that the latest lakes at the site were created in the periods 1330-1480 A.D. and 1630-1700 A.D. Each time, however, the lakes evaporated after the Colorado reverted to its usual southward-flowing course to the Gulf of California, cutting off the supply of water to the ancient lakes.

Dr. Sieh excavated across the fault at the northern shoreline of these lakes and examined the offsets in sediment layers along the San Andreas fault, thus reading the record of ancient earthquakes. The sediments did show evidence of minor slippage during the past six centuries. However, the latest major disruption shown occurred before a peat layer radiocarbon-dated between 1280 and 1420 A.D. was laid down.

In contrast, the section of the San Andreas northwest of Lost Lake has experienced at least two and perhaps four earthquakes during this period. At Lost Lake, Weldon and Sieh studied ancient river and lake sediments deposited across the San Andreas fault. Past earthquakes were revealed by digging a trench where the sediments were broken by the San Andreas. Studying the offset in the ancient river banks, the scientists found the average offset rate to be between two and three centimeters per year. They also found that the latest major earthquake occurred at the site more than 200 years ago and that substantial slip associated with perhaps six to eight large events has occurred over the last millenium. The slippage along the fault amounted to an average of three to four meters per event.

This rate of large earthquake occurrence is similar to that reported previously by Sieh, as a result of studies of excavations of ancient sediment at Pallett Creek and studies of earthquake-caused offsets at Wallace Creek. The recurrence interval also agrees with studies of sediments near Frazier Park by another researcher at U.C. Santa Barbara. The Caltech studies were supported by the U.S. Geological Survey.... Caltech