by John McPhee
In the country south of Hayward, we saw numerous streams that came down through the hills, made a right turn at the fault line, and then, after a bit, rediscovered their offset beds. We saw a deep ravine bent like a crochet hook. Like the San Andreas, the Hayward Fault seems to be stuck hard in some places (Berkeley) and in others (mainly in the south) to be creeping. A culvert curves like macaroni. A water tunnel begins to leak. Railroad tracks move. In Fremont, Moores and I climbed over some walls and fences in order to get up onto the ballast and squint down the rails of the Union Pacific, which, where it crossed the Hayward Fault, was bent into an echelon of kinks.
In 1986, a small earthquake on the Quien Sabe Fault, a deeply hidden blind thrust near the south end of the Hayward Fault, sent out elastic waves that shook open a twenty-thousand-gallon vat of cabernet sauvignon in an Almaden winery twenty-five miles from the epicenter. The vat was thirty feet high. A thunderous winefall flooded an office, broke out of the building, and poured down the road. This seems to have done it for Almaden. The Quien Sabe Fault? Like almost everybody else, Almaden had never heard of it, but the Quien Sabe Fault added insult to chronic injury. Ever since the winery was built, it had slowly been coming apart. Whereas the little Quien Sabe Fault was twenty-five miles away, the San Andreas Fault happened to run right through the building. The road outside was called Cienega, meaning “swamp,” as in reedfilled sag pond, an example of which was beside the winery. Almaden is a Spanish geological term meaning “mine,” and this location—about twelve miles south of Hollister, in the Central Creeping Zone—was no place to mine wine. Nevertheless, like so many parts of the San Andreas trace, it was an intimate and lovely valley, full of walnut orchards and olive groves and horse pastures and signs of anxious warning: “CAUTION: CHILDREN WALKING TO SCHOOL.”
This was the place where slow tectonic creep, also called aseismic slip, was first observed. The winery stands quiet now under Atlas cedars. Almaden has shut it down. Isabel Valenzuela, a caretaker who had left Mexico two weeks before, gave Moores and me a tour in Spanish. In the gloom of the great space, casks were ranked, and not a few were broken—cracked, like immense standing eggs. Through the floor a wide crack ran from one end to the other of the long rectangular building. Outside was a bronze plaque mounted on a freestanding wall of unreinforced masonry. It bore the words “San Andreas Fault has been designated a Registered Natural Landmark.”
I am from the Northeast and have never felt a destructive earthquake. The closest one to my home came in 1980, when a temblor centered in Cheesequake, New Jersey, caused no damage. My wife and I were in San Francisco in mid-October, 1989, and departed by train from Oakland, heading north. A niece in Berkeley drove us to the station and had difficulty finding it, in dark streets among warehouses on low flat bayfill ground. Up one street and down the next she searched—back and forth, around and under the Nimitz Freeway. The earthquake came more than a hundred hours after we left—an inexpressibly short span on the geologic scale, an irrelevant long one on ours. I wish I could say that I felt in my neuroplasm that it was time to go, but I was not a missing cat.
A few weeks later, when I was back in California, Moores and I were approaching the Santa Cruz Mountains from the south, and we stopped off at the Spanish mission in San Juan Bautista. Where scouts discovered two wells of water only twenty paces apart, the mission was established in 1797. Down the middle of a broad plain ran a sharply defined escarpment about fifty feet high, like the one in Hayward. It was a single long step, steeply inclined, like a grandstand. A modern geologist seeing such a break with springs or shallow wells beside it would be wondering not “Why is it there?” but “How frequent is the slippage upon it?” Moores remarked, “A fault is a good place for a well. It’s brecciated. There’s lots of porosity. Aquifers on either side are truncated, and spill into the fault.” The Franciscans built their mission on the top edge, the brink—not near, but on, the San Andreas. By October, 1800, they had begun or completed eight buildings, of adobe roofed with sedge, when earthquake after earthquake—as many as six shocks a day—brought much of the mission down.
The cool cloisters look immovable, the chapel looks repaired. Its plans (unique among the missions) called for three aisles divided by columns, but the colonnades were finished as walls. The view from the scarp is much the same: the fault is about all that has moved. In April, 1906, when San Juan Bautista was the southern end of the plate-shattering rupture, there was some destruction. Now, in 1989, there was no sign of the shaking of a few weeks before. Damage had been essentially nil in San Juan Bautista and everywhere in the fault zone to the south.
The fault scarp beside the chapel was actually used as a grandstand for dog races on a dirt track below, but the fans lost interest, and vines have grown over the seats. Beneath the bells of the mission, we sat in the bleachers and talked among the crawling vines, right on the fault, looking east toward distant mountains across the unaltered plain. Like almost all topography, it seemed to be immutable.
“People look upon the natural world as if all motions of the past had set the stage for us and were now frozen,” Moores remarked. “They look out on a scene like this and think, It was all made for us—even if the San Andreas Fault is at their feet. To imagine that turmoil is in the past and somehow we are now in a more stable time seems to be a psychological need. Leonardo Seeber, of Lamont-Doherty, referred to it as the principle of least astonishment. As we have seen this fall, the time we’re in is just as active as the past. The time between events is long only with respect to a human lifetime.”
(In a 1983 paper called “Large Scale Thin-Skin Tectonics,” Seeber addressed the possibility that crustal deformations have occurred on an areal and cataclysmic scale never imagined or described. “Our direct view of geologic phenomena has been severely limited by the relatively short span of history and by the relatively small vertical extent of outcrops,” he wrote. “In many respects we only have a two-dimensional snapshot view of the geologic process. Moreover, the interpretation of geologic data was probably influenced by the psychologic need to view the earth as a stable environment. Manifestations of current tectonism were often perceived as the last gasps of a geologically active past. Thus, subjected to the principle of least astonishment, geologic science has always tended to adopt the most static interpretation allowed by the data.”)
Earthquakes in the six-to-seven range occurred at least once a decade in the San Francisco Bay Area between 1850 and 1906. Afterward, that segment of the San Andreas Fault was essentially quiet for more than eighty years, a not very significant exception being the 5.5 event near Mussel Rock in 1957. While strain accumulated along the San Francisco Peninsula, whole lifetimes passed, so the principle of least astonishment, which works to a fare-thee-well in a place like New York, seemed to be working even here. In recent days, of course, the newspapers had been full of comment suggesting that least astonishment was no longer a principle in the Bay Area. Withal, there was an undercurrent of implication that it had not died and would come back.
Jerry Carroll, in the San Francisco Chronicle:
There is no greater betrayal than when the earth defaults on the understanding that it stay still under foot while we go about the business of life, which is full enough of perils as it is.
Stephanie Salter, San Francisco Examiner:
A traumatic experience … started in the depths of the earth and wreaked damage all the way to the depths of the psyche … . Or maybe the truth is, earthquake time is the most real time of all, a time when all the bull ceases and the preciousness of life is understood most acutely.
Herb Caen, the venerable columnist of the San Francisco Chronicle, who had seen his share of accumulating strain:
[This is] a headstrong, careless city dancing forever on the edge of disaster … . We realize afresh the joys and dangers of living here, and we reaffirm our belief that it is worth the gamble, however great … . We have been validated as San Franciscans.
A few miles up the trace,
we looked across the Pajaro Valley at the high notch where the fault slices into the southern extremities of the Santa Cruz Mountains. The east side was Oligocene shale, Moores said, and on the west side was a quartz diorite of the mid-Cretaceous. The two formations differed by at least sixty million years and by who knows how many miles of sliding offset. It was as if an apple and a pear had both been vertically sliced in half, and two of the differing halves had been placed together to make the mountains. The lookoff where we stood was at the northern end of the Gabilan Range. The Pajaro River, narrow and slow, ran westward toward the ocean through a topographic gap that punctured the mountains. Indians of the eighteenth century informed the arriving Spanish that the small stream in that huge gap had not always been so modest, that it had once been the outlet of the interior rivers, also draining the bays—the role now played by the Golden Gate. The Indians were right, Moores said. Never mind that they may not have suspected that the whole of the coastal country was moving northwest, occluding what lay to the east. Geologists have described the Pajaro Valley as “one of the most seismically active regions in the coterminous United States.” On April 18, 1906, a freight train crossing the valley was thrown off the tracks.
The San Andreas Fault was exceptionally smelly where it crossed the Pajaro River and went into the Santa Cruz Mountains. Highway 129 follows the right bank there. From the fault zone—a landslide of sedimentary hash about a hundred and twenty yards wide—a dozen sulphur springs were pouring. Galvanized pipes had been driven into the springs, causing the water to spout onto the roadside and color it yellow-cream.
Just to the west was Watsonville, which looked like a French battle town in 1944. The eighty-six-year-old St. Patrick’s Church, until recently a steepled brick structure with four spires, stood in its own red scree. That it stood at all was remarkable. Its crosses were aslant, its buttresses denuded, its brick gone in swatches from the walls. Like cartoon lightning, jagged cracks descended through the brick that remained. Two spires were stretched out in the church parking lot. Where Highway 1 crossed Struve Slough on concrete columns, the columns had punctured the pavement and now protruded upward like standing stones. Wide acreages in the town center had been bulldozed bare and brown. Ford’s Department Store (1851) was totally destroyed. Countless buildings were shuttered with plywood or wrapped in chain-link fencing. There were fissures in cement-block walls. Two hundred and fifty houses were off their foundations, many of them crushed like foam cups. There were tents. There were cellar holes where houses had been. In a single long moment at the edge of town, a million apples had fallen to the ground.
California building codes that involve seismic requirements were first written in 1933. They covered school buildings and nothing else. San Francisco did not extend such codes to other structures until the late nineteen-forties, when they appeared in the laws of virtually all communities around the bays and of many around the state. While most buildings are still “pre-code,” what is most remarkable is how effective the codes have been. In the 1989 Loma Prieta earthquake, sixty-two people died. In an earthquake of similar magnitude in Armenia in 1988, fifty-five thousand died. In Mexico City in 1985, ten thousand died. In the Iranian earthquake of 1990, fifty thousand died. The difference may lie partly in luck, in site, in relative intensity, but largely it lies in building codes, and the required or suggested strengthening of existing structures. Certain vulnerabilities notwithstanding, California seems to know what it is up against, and what to try to do about it. Never mind that in October, 1989, twenty-one thousand homes and commercial buildings were cracked, crumpled, or destroyed, and nature’s invoice for a few moments of shaking was six billion dollars.
During the previous summer, there had been a 5.2 earthquake, in the Santa Cruz Mountains, and a 5.1 quake the year before. These could be looked upon as precursors, but precursors never become such until a large jump follows, and are therefore useless as warnings. A score of 5.2 on the Richter scale is made by an earthquake with seven hundred times less energy than the one that shattered Watsonville. Richter was a professor at Caltech. His scale, devised in the nineteen-thirties, is understood by professors at Caltech and a percentage of the rest of the population too small to be expressible as a number. Another professor at Caltech in Richter’s time—and someone who manifestly understood the principles involved—was Beno Gutenberg, who provided the data from which the scale was made. The data applied only to southern California; subsequently, Gutenberg and Richter jointly developed the worldwide scale, which has since been variously refined. Gutenberg did not see or hear well and was understandably reluctant to deal with reporters. He generally asked his young colleague Charles F. Richter to explain the scale to them. Since I have no idea how the scale works, let me say only that it is a mathematically derived combination of three scales parallel to one another: a magnitude scale flanked by scales of amplitude and distance. (Amplitude is the height of the mark an earthquake produces on a seismogram.) Where a line drawn between amplitude and distance crosses the central scale, it registers magnitude. With each rising integer on the magnitude scale, an earthquake’s waves have ten times as much amplitude and thirty times as much energy. Richter always insisted that it was the Gutenberg-Richter scale.
There is a swerve in the San Andreas Fault where it moves through the Santa Cruz Mountains. It bends a little and then straightens again, like the track of a tire that was turned to avoid an animal. Because deviations in transform faults retard the sliding and help strain to build, the most pronounced ones are known as tectonic knots, or great asperities, or prominent restraining bends. The two greatest known earthquakes on the fault occurred at or close to prominent restraining bends. The little jog in the Santa Cruz Mountains is a modest asperity, but enough to tighten the lock. As the strain rises through the years, the scales of geologic time and human time draw ever closer, until they coincide. An earthquake is not felt everywhere at once. It travels in every direction—up, down, and sideways—from its place and moment of beginning. In this example, the precise moment is in the sixteenth second of the fifth minute after five in the afternoon, as the scales touch and the tectonic knot lets go.
The epicenter is in the Forest of Nisene Marks, a few hundred yards from Trout Creek Gulch, five miles north of Monterey Bay. The most conspicuous nearby landmark is the mountain called Loma Prieta. In a curving small road in the gulch are closed gates and speed bumps. PRIVATE PROPERTY. KEEP OUT. This is steep terrain—roughed up, but to a greater extent serene. Under the redwoods are glades of maidenhair. There are fields of pampas grass, stands of tan madrone. A house worth two million dollars is under construction, and construction will continue when this is over. BEWARE OF DOG.
Motion occurs fifty-nine thousand eight hundred feet down—the deepest hypocenter ever recorded on the San Andreas Fault. No drill hole made anywhere on earth for any purpose has reached so far. On the San Andreas, no earthquake is ever likely to reach deeper. Below sixty thousand feet, the rock is no longer brittle.
The epicenter, the point at the surface directly above the hypocenter, is four miles from the fault trace. Some geologists will wonder if the motion occurred in a blind thrust, but in the Santa Cruz Mountains the two sides of the San Andreas Fault are not vertical. The Pacific wall leans against the North American wall at about the angle of a ladder.
For seven to ten seconds, the deep rockfaces slide. The maximum jump is more than seven feet. Northwest and southeast, the slip propagates an aggregate twenty-five miles. This is not an especially large event. It is nothing like a plate-rupturing earthquake. Its upward motion stops twenty thousand feet below the surface. Even so, the slippage plane—where the two great slanting faces have moved against each other—is an irregular oval of nearly two hundred square miles. The released strain turns into waves, and they develop half a megaton of energy. Which is serious enough. In California argot, this is not a tickler—it’s a slammer.
The pressure waves spread upward and outward about three and a half miles
a second, expanding, compressing, expanding, compressing the crystal structures in the rock. The shear waves that follow are somewhat slower. Slower still (about two miles a second) are the surface waves: Rayleigh waves, in particle motion like a rolling sea, and Love waves, advancing like snakes. Wherever things shake, the shaking will consist of all these waves. Half a minute will pass before the light towers move at Candlestick Park. Meanwhile, dogs are barking in Trout Creek Gulch. Car alarms and house alarms are screaming. If, somehow, you could hear all such alarms coming on throughout the region, you could hear the spread of the earthquake. The redwoods are swaying. Some snap like asparagus. The restraining bend has forced the rock to rise. Here, west of the fault trace, the terrain has suddenly been elevated a foot and a half—a punch delivered from below. For some reason, it is felt most on the highest ground.