Seismologists Seek Clues to the Future in Fault’s Past

Tom Rockwell is standing in a freshly dug hole the size of a human grave. To his left is a pile of unexploded shells and mortars that could rip him to shreds. To his right are two hulking forms, the burned-out remnants of tanks.

Rockwell barely notices. Instead, he’s staring at the dirt walls of a trench dug along the edge of the 25-mile Hector Mine fault, which literally tore this Marine Corps base apart on Oct. 16. The 7.1 magnitude earthquake was a complete surprise, hitting a fault not even deemed important enough to name.

That a little-studied and seemingly unimportant fault could unleash such a massive quake shows how far scientists are from being able to predict when and even where earthquakes will occur. Most geologists have given up trying to look directly into the future.

But there is one place left to look: the past.

So, during a recent break from the almost constant live-fire exercises here, Rockwell and a handful of fellow geologists descended on the base for their first intimate view of this latest and biggest rupture.

The dozen trenches dug along the fault will yield a record--written in subtle layers of rock and sand--that will show how often this fault ruptured in the past, how powerful those prehistoric quakes were and how far they tugged the earth along each time.

What these paleoseismologists find will no doubt help explain the perplexing seismic caldron beneath the Mojave Desert. They also represent the first baby steps toward turning quake prediction into something resembling a science.

“It may,” said Rockwell, a professor of geology at San Diego State University, “really tell us a much bigger story about how the earth behaves.”

A chief question is whether faults rupture in distinct patterns. Geologists no longer believe the old dogma that earthquakes release tension along a fault and then repeat themselves in predictable intervals after more strain accumulates.

A series of trenches along the San Andreas fault revealed a puzzling and irregular pattern of quakes: Three dated in the years 671 through 797, then none for two centuries. Three more in 997 through 1100, then none for 246 years. Another two in 1346 and 1480 and then three centuries of inactivity until violence erupted once again in 1812 and 1857.

Until this discovery, the recorded history of San Andreas activity, spanning little more than 100 years, was far too short to reveal such patterns.

“It’s inconceivable to me now that we would try to understand a fault without understanding its history,” said Kerry Sieh, a professor of geology at Caltech. Sieh pioneered the technique of dating prehistoric earthquakes while a graduate student at Stanford in the mid-1970s and has since compiled a long and detailed record of the behavior of the San Andreas.

At the time, the idea that you could glean precise information about past earthquakes from traces in the dirt was deemed so ludicrous that Sieh got a failing grade for proposing it as a research project. “Nobody thought it was possible--or interesting,” he said.

But times have changed. In 1997, Rockwell and geologists Charlie Rubin and Scott Lindvall dug a trench in Altadena’s Loma Alta Park. They found evidence that the Sierra Madre fault at the base of the San Gabriel Mountains was capable of producing earthquakes as large as magnitude 7.6. This “wake-up call,” as seismologists called it, showed that a catastrophic earthquake near the heart of Los Angeles was possible, even though such a quake hadn’t occurred in recorded history.

“Looking backward allows you to look forward,” Lindvall said. Illuminating as such studies are, many paleoseismologists have moved past analyzing single trenches or faults. What Rockwell wants now is nothing less than to understand the complexities of how faults behave as a system.

It’s clear, from the Hector Mine quake and many others, that the working of faults is far more complex than seismologists once imagined. Some faults appear to act within networks, the movement of one influencing or triggering earthquakes on another. And earthquake activity appears to come in clusters in a specific region, with large areas rumbling in bursts of seismic activity and then lying silent for centuries. The Mojave appears to be in an earthquake cluster now.

“A fundamental thing we’re trying to get a handle on is how faults interact with each other,” said Ken Hudnut, a geophysicist at the U.S. Geological Survey office in Pasadena. “To me, that’s the most important scientific question out there.”

Clues Buried With Bombs

Field geologists are used to roughing it. Riding horse-drawn carts to inspect faults in northern Turkey? Driving three days over bad Mongolian roads in aging Russian jeeps? Eating sheep and potatoes every day for a month? All part of the job.

Still, the project in the Mojave offers special challenges.

The base is occupied year round by Marines who continually run simulated warfare exercises that light up desert nights with ground artillery, tank fire and hails of bombs from aircraft. Twentynine Palms is scheduled tightly: 50,000 Marines must run through the training each year. A brief training lull around Easter Sunday gave geologists a six-day window to scoop out a trench and look for seismic clues--a project that normally takes closer to a month.

The geologists were allowed on the base only after they learned the fundamentals about unexploded bombs, so they’d be able to avoid suspect shells and mortars that litter remote parts of the base.

Early spring temperatures often top 100 degrees, and getting lost on the Rhode Island-sized base is a distinct possibility. So the Marines also offered desert survival tips to the geologists: Move only at night if lost; breathe through your nose to minimize water loss; rub urine on your body to stay cool.

Geologists navigated the unmarked base using topographic maps and global positioning system units. They checked in with base headquarters around the clock.

“The Marines were extremely hospitable,” said Karl Gross, the U.S. Geological Survey geologist in charge of the trip’s complex logistics. Camping conditions were less hospitable: hot open desert with no shade during the day and screaming wind and frigid temperatures at night.

Still, the field site was a geologist’s dream. The fault lay exposed for miles and miles.

Holes are scooped out with backhoes, then cleared of dust with brooms, brushes, even leaf blowers. “It sounds funny--cleaning dirt walls--but cleaning makes the layers more clear,” said Heidi Stenner, a U.S. Geological Survey geologist, as she brushed away dirt to expose sediment layers underneath.

Some sites--like one of Sieh’s trenches along the San Andreas--have layers so distinct that geologists compare them to wedding cakes. The layers of dark peat and lighter sand stand out as distinctly as frosting, cake and jam.

Many of those same patterns are embedded in the dusty layers underneath Twentynine Palms. Here, though, they are much harder to see. The soils come in various shades of sand, tan and beige that blur, at first glance, into an indistinguishable brown haze.

Yet to the practiced eye, features begin to emerge. A row of dirt filled with large pebbles suddenly ends. A layer of gray gravel steps up six inches. Some distinctions are obvious. Others, like Rorschach ink blots, are open to interpretation.

“A lot of it is in the eye of the beholder,” said David P. Schwartz, director of the U.S. Geological Survey’s San Francisco Bay Area Earthquakes Hazards Project. “There’s a real tendency to put in earthquakes that aren’t there,” although once quakes are established, dating them is precise.

As one of the first geologists to dig trenches along faults in the 1970s, Schwartz is on a mission to turn the young field of paleoseismology into more of a science and less of an art.

Identifying past earthquakes within the layers is just the first step. Geologists then use radioactive dating techniques to determine exactly when those earthquakes occurred and thus begin the search for past patterns that might help divine the future. “One of the things they teach you in introductory geology is that the present is the key to the past,” Schwartz said. “I think the past is the key to the future.”

Geologists call the Hector Mine quake a “7.1 to enjoy” because no one was killed in the sparsely populated region around the epicenter.

The Los Angeles Basin has a similar geology, crisscrossed by a latticework of faults both known and unknown, Rockwell said. The same quake in Los Angeles could easily have killed thousands and caused billions of dollars in property damage.

So Rockwell races between trenches in the Mojave Desert, not only to get done before mortar fire returns, but to nudge along the new field of paleoseismology that could help save lives in the future.

“Out here we can study them,” Rockwell said. “If we can understand how fault systems work here, we can apply it to the system of faults under L.A., which are difficult to study because they’re all under concrete.”