Scary beauty surrounds Cameron Barrows. He works in lush groves of fan palms that erupt like mirages from moonscape terrain. Hot springs bubble beneath them. Sand dunes drift nearby.
"It's an amazing place," said Barrows, director of the Coachella Valley Preserve east of Palm Springs. The 20,000-acre sanctuary owes its splendors to the San Andreas fault, the frightening part of the bargain.
Many scientists say the Coachella Valley is where the 750-mile San Andreas seems most prone for an epic earthquake, a monster that would be enormously more powerful than the recent temblors in San Simeon, Calif., and Bam, Iran.
"There's not a lot we can do about that," Barrows said with a resigned smile. He was outside the preserve's visitors center, a 1930s log cabin that sits directly above the fault. "I'm not a worrying type person."
As the possible generator of the feared Big One, the San Andreas once dominated the quake worries of Californians. But that was before "subsidiary faults" in locales such as Loma Prieta and, especially, Northridge reordered popular anxieties. They flattened buildings and buckled interstates while the San Andreas remained relatively quiet, as it has since the great San Francisco quake of 1906.
Now, with the 10th anniversary of the Northridge temblor approaching, and after much study of those second-tier faults, scientists again are highlighting the San Andreas as the rupture without rival -- a slumbering beast napping on borrowed time.
"The primary fault in California -- the big dog -- is the San Andreas, and it's important for people to remember that," said Doug Yule, a geologist at Cal State Northridge, which was badly damaged in the Jan. 17, 1994, disaster. "The San Andreas will produce the largest earthquakes."
Yule and his colleagues have dug trenches along the southern section of the fault to carbon-date its buried fissures in hopes of determining just how "pregnant" it is. Their best guess: The San Andreas, from the Salton Sea to San Bernardino, is at term.
Two hundred fifty miles to the north -- and 35 miles from the epicenter of the magnitude 6.5 San Simeon quake Dec. 22 -- investigators in Parkfield are preparing to drill a $20.5-million hole into the San Andreas.
The idea behind the 2.4-mile-deep probe, the first of its kind, is to capture a live quake on a battery of monitors -- and perhaps advance the balky business of quake predictions.
The San Andreas also is a principal subject of the new Plate Boundary Observatory, a $100-million array of global positioning stations, strain meters and ground-motion detectors.
The unfolding project will chart inching shifts in the landscape that result from the grinding collision of the Pacific and North American tectonic plates, the crash that created the fault.
All those efforts have returned the San Andreas to the pages of science journals, quake probability reports, even travel publications -- as a foreboding curiosity that helped shape California in more ways than one.
"The tourists play golf, then they come out here to see the San Andreas," Barrows said.
He was walking through the preserve's Thousand Palms Oasis, in the jungle shadows of 70-foot trees. Wooden planks cover the path and 80-degree springs gurgle underfoot. The water nurtures the fans, California's only native palm.
The springs are propelled upward by the San Andreas, which is otherwise undiscernible to the untrained eye.
"People are always disappointed it isn't this huge chasm in the ground," Barrows said.
Instead, the San Andreas reveals itself in geological magic tricks: gullies that turn gravity-defying corners, and abrupt changes in the desert floor, with uplifted bedrock yielding to gravels that are blown into Arabian-style dunes.
Every one of the phenomena is a sign of looming calamity, said Sally McGill, a geologist at Cal State San Bernardino. Her classrooms are within a mile of the San Andreas.
"This is as close at it gets to a heavily populated area in Southern California," McGill said.
She was setting out on a short hike to the fault, whose presence is marked by an unlikely picket of trees thriving along arid foothills. "I do worry about it. This is a dangerous place."
The San Andreas last slipped in the region 191 years ago. That is 40 years beyond the average interval for the southern segment, based on estimates that stretch back 12 centuries.
Geologists arrived at the calculations through paleoseismology, a fairly new technique that dates prehistoric quakes. Scientists dig into the fault to look for layers of peat and sand.
The strata time-stamp cracks -- give or take 50 years -- that quakes opened to the sunlight and that flood sediments filled later.
It is one of the disciplines that has made strides since Northridge, although the fault responsible for that quake resists paleo-detective work because it never broke the surface. The San Andreas is a proven crust-buster.
"Over the past 10 years, we've put a lot of effort into the L.A. Basin," McGill said. "Now it's time to put more attention on the San Andreas."
She was climbing a low ridge where the October wildfires burned away enough brush to leave fault impressions bare.
Her camera raised, she paused at a gully that stopped halfway down a hillside, as if it had hit an invisible wall. McGill pointed to a spot about 20 feet away where the gully reappeared. Its lower leg had been shunted to the right in shifting of the San Andreas.
"I am interested in those other faults, but the San Andreas is the only one capable of producing a magnitude 8 earthquake," said McGill, who stashes bottled water under her desk, just in case. "I would expect a 7 here, or a 7.5. A 7.9 is possible."
A 7.9 would release about 65 times the energy of the 6.7 Northridge quake, which killed 57 people and caused $40 billion in damage.
That doesn't mean a 7.9 would be 65 times more destructive. Magnitude is a deceptive measurement. It denotes the physical length of a quake: The more miles of splintering on the fault, the higher the magnitude.
The Northridge quake struck along a 10-mile fault. The San Andreas can break for 200-plus miles, as it did in the 1906 quake, a 7.8 that devastated San Francisco.
In a given location, the shaking near the epicenter of a moderate quake can be as severe as that unleashed by a huge fault slip farther away.
But a giant quake on the San Andreas would thunder over a broader area and last longer, pounding repeatedly into buildings, dams and bridges.
There is "good news" about the fault, experts like to say. Most of the San Andreas is remote from major cities in Southern and Central California.
And scientists generally believe the 1906 cataclysm vented enough stress to spare fault-straddling San Francisco another mammoth quake for decades to come.
Then again, the San Andreas runs through the fast-growing Inland Empire and Antelope Valley, which last felt the fault's fury in 1812 and 1857. This is the bad news.
"Here's an 8.0 on the San Andreas," said Bob Tincher, a San Bernardino water engineer who was poring over a quake-hazard map on his desk.
He tapped a finger on red-colored sectors of high groundwater. They spread across much of the city, including downtown.
"Red is high risk," he said.
In a magnitude 8 temblor, streets in the red could be rendered quicksand by liquefaction as the quake briefly scrambled saturated soils. Buildings could come crashing down.
San Bernardino is trying to head off the peril. Tincher drove from his office to a shopping center. In the parking lot, behind a barbed wire-tipped fence, a pump as big as a pickup truck was pulling water from the ground and sending it through pipes and storm drains to the Santa Ana River.
The pump and others like it are lowering the water table to keep San Bernardino on solid moorings (Orange County is buying the excess water).
The city is halfway toward its quake-ready goal of siphoning out 25,000 acre-feet of water annually, the amount used by 50,000 households.
"It's kind of a unique problem," Tincher said.
Also unique is the drilling planned for Parkfield, a tiny Monterey County town where visiting fault sleuths sometimes outnumber the population.
Lately, 350 scientists from the U.S. Geological Survey, Stanford University and other institutions have been gearing up to drive a seismometer-lined shaft into and across the San Andreas.
A 1.4-mile pilot hole already has been sunk close to the fault to test equipment. Its seismometers found that the San Simeon quake had moved the fault one-twenty-fifth of an inch or less. The San Andreas typically slides 1 inch a year in the Parkfield area.
"There's a little chance the San Simeon quake is going to have a significant impact on the San Andreas," said Steve Hickman, a U.S. Geological Survey geologist and a designer of the Parkfield project. "But we'll keep an eye on it."
Digging of the main "well" in Parkfield -- the geologists are borrowing oil technology -- will begin this summer. The probe will target the "creeping section" of the San Andreas, which has delivered the most regular pattern of moderate quakes on the fault: six in the magnitude 6 range since 1857, the last in 1966.
That record of punctuality led seismic forecasters in the 1980s to predict another quake within several years. It has yet to strike, but scientists remain convinced that Parkfield has secrets to spill.
"We're going to be able to instrument a fault right where a quake is happening," said Andy Snyder, a U.S. Geological Survey geologist.
He had spent the afternoon checking on the multiplying numbers of geo-gizmos planted in the blond pastures and oak-canopied hills around Parkfield: strain and creep meters, global positioning stations and magnetometers.
Nowhere else on the planet is a fault so intently watched.
"This is where you could capture an earthquake," Snyder said.
He was standing in a weedy field beside the pilot hole's wellhead, a green hunk of metal that resembles a large fire hydrant. A nearby shed shelters banks of computers that clock every shudder in the hole.
In the 2.4-mile-deep shaft, the coin-size seismometers will track shaking and tilting while devices called transducers gauge fluid pressures. Geologists theorize that increases in day-to-day fault deformations and changes in subterranean fluids precede a quake.
"It's the information you need to go to the next step," Snyder said.
Tom Taylor, a geophysics scholar from Duke University, emerged from the shed and talked more about the next step: reliable quake predictions.
"Right now it's a voodoo science," he said. "I want to find out if you can predict them."
So does Robert Uhrhammer, a researcher at UC Berkeley's seismological laboratory.
He is working on the Plate Boundary Observatory, part of the National Science Foundation's EarthScope program, which includes the Parkfield experiments.
"In the last 10 years, there have been considerable advances in paleoseismic studies and global position systems," he said. "Earthquake probabilities are much more formal now and much more reasonable."
The observatory -- it's a widely cast installation of sensors, not a building -- will collect data from points along the San Andreas and through the Cascadia subduction zone to Canada.
The subduction zone is where the North American and Juan de Fuca plates meet. Subduction quakes are vastly more violent -- magnitude 9s are possible -- but far less frequent than those on the San Andreas.
Such events occur in subduction zones once or twice every 1,000 years; the most recent on the Cascadia was in 1700.
Uhrhammer said the observatory would help scientists determine where the San Andreas might be lurching toward a 1906-strength quake.
He doesn't expect a repeat in the Bay Area anytime soon.
On the northern San Andreas, "it could well be another century or so before you get another 1906 event," Uhrhammer said.
The outlook is not as reassuring down south.
Yule, the Cal State Northridge geologist, has found evidence in his paleoseismic trench that a massive quake could be in store for the Coachella Valley -- and north into San Bernardino and beyond.
But like all fault excavators, he notes, "We just don't know."
Yule and two of his students were tip-toeing along a ledge of the terraced trench, a gash cut by backhoe, 12 feet deep by 150 feet long in the San Gorgonio Pass.
Rumblings on the San Andreas carved out the pass. Southern California's highest peaks -- San Gorgonio and San Jacinto -- struggle for supremacy across the ever-windy plain. The fault and its strands are forcing two mountain ranges into each other.
The walls of Yule's trench show scars of the battle. Lines and swirls of peat demark ancient quakes. They sketch a calendar of 4,000 years, but Yule is zeroing in on a shorter period.
Sediment veins in his trench indicate that the valley has not experienced a momentous quake since 1800 or 1700. Either date would mean the San Andreas here is past its average rest.
"If a quake happened right now, we'd be 'faulted,' " Yule said as he eyed the cleaved rock and soil. "It'd be pretty exciting."
He wasn't smiling.