A Crack in the Edge of the World

by Simon Winchester

  But, leaving aside the scimitar-like line it cuts across the scenery, it is the right-lateral displacement of the rivers running across the fault that presents it in an even more dramatic light. Such oddities show the San Andreas not merely as a static entity but as something that shifts: E pur si muove.

  To the east of the plain rise the low hills of the Temblor Range, and from their upperworks course innumerable rivulets—usually dry—known locally as washes. On those rare occasions when rain does fall, these washes fill up with water and form raging streams, which rush downhill, carving deep canyons from the loose and friable soil as they do so. The dry ravines that are left behind are known as arroyos, and, were the hillsides perfectly geologically stable, one would regard such things as merely pretty, the serried ranks of parallel valleys proceeding from summit to valley floor, peculiar only in their number and symmetry.

  But the edge of the hills marks the line of the fault. The Carrizo Plain happens to be on the Pacific Plate, and the Temblor Range is on the North American Plate. Every so often the Pacific Plate jerks itself northward, and when this happens it takes with it the paths of the arroyos, just where they begin to run across the plain. Where this has happened the effect is startling: The ravine starts off running downhill, then suddenly turns ninety degrees and begins to flow along a straight line parallel both to the edge of the range and to the valley floor—or, to put it another way, parallel to the line marking the joined edges of the two plates. It goes on like this for a few hundred feet—and then with equal suddenness it turns back ninety degrees, flowing in its original direction once more.

  And every stream does the same—in echelon, every single stream exhibits the double jog, each in the same direction, each serving as a marvelous reminder of the fact that the fault moves and of the direction in which it does so.

  This is not to say that the movement along the fault is horizontal only—which is what a cursory look at the offset streams of the Carrizo Plain might imply. There is a vertical component, too—the Temblor Range, for example, has been caused by vertical uplift along the fault. (The reason it exists at all is that the fault does not exactly coincide here with the direction of plate motion: It is slightly skewed to one side, and as a consequence of this skewing the material bunches up, like a pleat in a carpet, as the plates move alongside each other.) But in this part of the San Andreas the horizontal component far outweighs the vertical, by a factor of between 10 and 20 to 1.

  One of the most dramatically offset of these Carrizo Plain streams—a stream that has become the poster child for all San Andreas phenomena, so clear-cut is its display—has been named Wallace Creek, honoring the field geologist who has perhaps been most prominently associated with the fault, Robert Wallace of the U.S. Geological Survey.

  It was Robert Wallace who reversed a notion among some geologists that had become fashionable during the mid-twentieth century: This was, that since the newly invented machine called the seismograph gave all the necessary answers about earthquakes, there was little need for geologists to go out into the field to study them. For years around the middle of the century, quakes were primarily in the purview of white-coated technicians, people sitting in air-conditioned laboratories working their bloodless mathematical wizardry to determine the nature of the spasms that occasionally afflicted the earth. Robert Wallace changed all that: Armed with a hammer, a compass, a series of very good maps, a tent, and endless cans of beef stew*—which he liked to eat cold—he went back out among the rocks and worked for years to delineate the fault as it charged southward from the Carrizo Plain.

  He mapped the places where the fault exhibited what has come to be known as the Big Bend—the long and lazy turn it executes midway along its length, which gives the San Andreas its approximate overall shape of a boomerang. To create this bend, the line first turns through thirty-five degrees to the left—to the southeast, that is—then lazes back again toward the right, until finally, sixty miles of zigzagging contortion later, it settles straight, back on its original track.

  The turn is made as it traverses an unusual set of mountains called the Transverse Ranges—unusual because, unlike all other Californian mountain ranges, these mountains slice across the state from west to east, and do not parallel the coastlines and the faults, as do the Sierra and the coast ranges, as well as all the lesser ranges like the Temblors and the San Gabriels.

  The fault crosses this range at a remote mountain cwm called the Tejon Pass—so named because the Spanish lieutenant who discovered it at the beginning of the nineteenth century found a dead badger, el tejón, at the mouth of the canyon—and then passes on into the achingly dry wilderness of the Mojave Desert, close to the city of Palmdale. Robert Wallace was one of the first to work out just how far the fault seemed to have slid in these parts over its millions of years of life. As a brash young graduate student he declared that at Palmdale* the two sides had shifted 75 miles apart from each other—a figure that, in those more conservative, pre-tectonic-plate-theory days, was so improbably large that his colleagues chortled in disbelief. They should not have bothered, since Wallace was, of course, almost right—at least so far as orders of magnitude were concerned. The two sides of the San Andreas Fault have moved at least 250 miles apart since it was born—sometimes slowly and steadily, sometimes through the savage interruptions of earthquakes.

  California’s other truly great earthquake occurred in these parts, close to the Tejon Pass, in the spring of 1857. It was an event that was at least as big, and may well have been even larger, than the quake that was to wreck San Francisco nearly half a century later.

  It occurred on April 9, and it had a supposed summary magnitude of 8.25—just a notional amount less than that later ascribed to San Francisco. Its intensity (which is not the same as magnitude, the distinction being explained in the appendix) is thought to have been in the region of VII, also rather less than the VIII and IX that were experienced in San Francisco. Some calculations insist that it was a greater event than San Francisco—that more energy was released and that the displacement of the ground was far more extensive. But there is so little by way of historical record—much less than that of the New Madrid Earthquake of 1811, mentioned before—that it has to be regarded as an event of less historic importance. It cannot be classed as one of the most significant earthquakes of all time—which 1906 most certainly was, because it ruined so much and killed so many. But it was nonetheless very, very impressive.

  Its epicenter was actually nowhere near Tejon Pass. It has since been shown to have been back in Parkfield. But since virtually no one lived in Parkfield in 1857, and since Tejon had an army base, Fort Tejon, this bleak mountain pass has ever since been put forward to enjoy or to suffer the notoriety.

  The base had been created three years before, ostensibly as a means of controlling—the army’s word—the resident Indian population. It was little more than a handful of adobe huts huddled among the valley oak trees, beside a stream. The only thing that was unusual about the fort was its small and temporary population of camels, which had been brought to Texas from Egypt and Turkey because some military planner thought they might be useful; they proved worse than useless, however. The Texans got rid of them and sent them to California, and the army in California had no real idea what to do with them either, and they soon vanished. Some still associate Fort Tejon with something called the U.S. Army Camel Corps: There was never, in fact, any such thing.

  Just after eight on the Friday morning of January 9, 1857, an almighty earthquake rolled down onto Tejon from the north. Witnesses speak of huge wavelike shakings of the earth; and, though some speak of up to three full minutes of shaking, an unprecedented duration, most agree that it was just some forty or fifty seconds’ worth of nightmarish movement that wrecked all the army huts, tore most of the trees from the earth, and killed a woman at the nearby Reed’s Ranch. The local Kern River ran backward; fish were thrown hundreds of yards from where they swam in Tulare Lake; long zigzag cracks appeare
d in the ground at San Bernardino; massive ridges, five feet high and fifteen feet across, rose and started to snake through fields; artesian wells suddenly failed; the Los Angeles River was hurled out of its bed and began, if only briefly, to flow along another channel; and up on the Carrizo Plain the fault jerked so dramatically that many of the rivers coursing down from the Temblors were thrown off course by as much as thirty feet in a matter of microseconds.

  The event was felt all across Southern California. It was not felt at all north of Parkfield, perhaps because of the more lubricated nature of the ever-moving fault up there. Had it struck in modern times, it would have caused dreadful damage forty miles away in Los Angeles. But, as it was, only two people (the rancher, and one other man in a village plaza) were killed; and the 4,000 people who lived in the sprawling village that was Los Angeles got little more than a jostling.

  Matters were somewhat more serious in Santa Barbara, then a pretty coastal hamlet of 2,500 thirty miles west of Tejon. Villagers, who were accustomed then as now to an idyllic setting with equable temperatures, balmy weather, and no more than the gentle plashing of Pacific waves, streamed out onto the streets, panicky and terrified at something they could hardly have imagined. In huge numbers they fell to their knees and struck beseeching attitudes (a mode of behavior that seems now to have been supplanted by a need to turn on CNN), then waited anxiously until the swelling vibrations fell away and the cascades of aftershocks abated. The local newspaper turned to doggerel to explain the dreadful majesty of the moment: “How awful is the thought of the wonders under ground / Of the mystic changes wrought in the silent, dark profound.”

  The implications of what was truly an awe-inspiring event at Fort Tejon go some way beyond the simple matter of the quake’s enormous magnitude. The geometry of the San Andreas Fault’s Big Bend has an effect on the local topography that is very complicated and still being properly worked out. But it boils down to one reality: that because the Pacific Plate is still pushing northward on this part of the fault—moving in the same direction, in other words, here as everywhere else on the 750 miles of its length—and because the Big Bend thrusts a prow-shaped bulge out into its path, the effect of the movement is here not simply a sliding-along-the-side affair; it is also a pushing-up-from-beneath kind of movement, a movement that tends to lift the prow of the northbound plate up somewhat and produce a range of hills in front of it.

  Uplift like this was noticed up in Parkfield, where another section of the fault had become very slightly misaligned from the plate boundary—and the Temblor Range, as we have seen, was thrown up as the result of that. Here at Tejon the misalignment was evidently more spectacular: Rather than throwing up a low range of hills on just one side of the boundary, the plate in these parts seems to thrust itself head-on into the rock mass, right into the prow-shaped bulge of California, and has lifted up hills that would in due course become the Transverse Range, hills that look like a giant raft rising up onto the plate’s advancing bow wave. The analogy is an imperfect one—the precise nature of what is going on below these mountains is still being analyzed—but in its essentials it holds.

  There are other faults working here, too—very ancient faults with names such as the Clemens Well–Fenner–San Francisquito Fault and the San Gabriel Fault, which displaced the surface rocks for scores of miles, tens of millions of years ago; and shallow thrusting faults that never break surface today but that all the time are helping to accommodate the compression of today’s plates as they ram themselves together, breaking and buckling like the crush zones in modern cars or the water-filled “impact attenuators” you see at dangerous road intersections.

  And it would be idle to pretend that matters get any less complex as the fault spears ever onward, southeastward, passing as it does so along the zone that separates the San Gabriel Mountains from Antelope Valley (where there are countless supersecret defense establishments, making and testing costly warplanes and missiles), past the suburban sprawls of San Bernardino and Loma Linda and Redlands, then through the San Gorgonio Pass with its throbbing forests of wind generators, and along the Coachella Valley, where the elderly rich like to live out their final years in luxury and perfect weather, in communities like Palm Springs, Rancho Mirage, Indian Wells, and Palm Desert.

  The fault underlies all of these places. When Andrew Lawson wrote his earthquake report in 1908, he supposed that the San Andreas in fact petered out here, but his confusion can be excused. For in this area the fault ceases to be a neat and tidy trolley track, or even a pair of tracks running in parallel; it is very tricky to recognize. It does not appear to cause much by way of local seismic problems, nor does it seem to determine the local lie of the land—which is, in any case, a mess of mountains and valleys spreading in all sorts of directions at once.

  But in fact the fault, or the fault zone, continues on its merry way for at least another 150 miles—the first part of which is unutterably confusing, with a significance that must have been easy for a 1908 geologist to miss. Subsurface maps of this particular region, now that much of the fault geography has been worked out, show that the fault does indeed continue, but that it has the crazed aspect of a spider’s web, or of a car’s broken windshield. Newly identified sister faults hiss and sidle out from the main line, faults with names like the Pinto Mountain, the Garnet Hill, the Vincent, and the Arrowhead.

  It was close by this point that the San Andreas may have had its first publicly acknowledged workout. No small amount of mystery attends the event, but it took place in the summer of 1769, at the very beginning of Spanish settlement of California, when an expedition of militarily supported Franciscans, led by Gaspar de Portolá, was pushing north from San Diego, hunting for an overland route to Monterey. On July 28, while they were camped beside the Santa Ana River, they were interrupted by a severe earthquake that, the explorer noted, “lasted about as half as long as the Ave Maria.”

  A little more can be gathered from the diaries of two other members of the expedition: They were perplexed by the aftershocks that went on for a full week. Juan Crespi, one of the diarists, recorded feeling as many as a dozen shocks a day while his team was winding its way around the southern hills of Los Angeles and down into the San Fernando Valley.

  From all this anecdotal evidence, and from comparisons that can be made with more recent earthquakes, it appears likely that what Gaspar de Portolá and his party felt was a series of magnitude 6 quakes that had resulted from movement on either the San Jacinto Fault or, more probably, the San Andreas Fault. If so, then the fault that has caused so much mayhem during all of California’s existence showed itself to be capable of great fury to the very first explorers of the region. They were given a warning—one that they and all who visited subsequently chose pointedly to ignore.

  AFTER ALL THE topographic confusions to the east of Los Angeles, the San Andreas splits itself into two parts near Palm Springs, then repairs itself again near the town of Indio and becomes one once more. And finally, following this, after all of the excitements back at the Parkfield drilling site, after the ample confusions of the Big Bend, and the terrible complexities that are apparent as its licks its way around to the east of Los Angeles, it arrives in the city of Indio itself.

  There is little that is memorable about the most southerly town on the fault’s long track; it is no more than a scalding hot little railway community, a town where they have festivals to celebrate the dates they grow and the tamales they cook, and precious little else. The fault steals through without remark and without much evidence of its passing. It realigns itself slightly, kicks back onto its customary course, and resumes its orderly journey down toward its southern end.

  And then, quite without ceremony, it disappears. It vanishes away in a muddy little field, beside one of the more unusual physical phenomena that is to be found anywhere in the country.

  The Salton Sea is an enormous brackish lake, thirty-odd miles long and fifteen wide, that was created in 1905 as a result of a very foolish
and very avoidable accident. At the turn of the century a firm called the California Development Company built a series of large irrigation canals in far Southern California to divert water from the Colorado River, which was then very close to its outlet in the gulf known in Mexico as the Sea of Cortés. The basic idea was to help local farmers on the edge of a very-low-lying part of the Imperial Valley called the Salton Sink (a relic of an earlier inland sea, as it happens) to grow fields of well-watered asparagus and broccoli. The company would charge the farmers, the farmers would sell their products, and everyone, in the classically American way, would make money.

  Except that, unhappily, on one memorably unfortunate day, the levees protecting the trunk canal broke, and the entire flow of the Colorado River poured down into Salton Sink—which was more than 220 feet below sea level, at almost the same negative altitude as Death Valley. The waters kept coming and coming for more than a year, until finally railway wagons loaded with boulders plugged the hole in the levee and the newly created Salton Sea stopped filling and began to do what it has been doing ever since—evaporating.

  The lake today is an odd, vaguely unpleasant place, rich with a strange smell that somehow mixes heat with dankness, rimmed with broken-down towns made of rusting trailers, with a reputation for the widespread manufacture of that particular nasty and highly addictive drug crystal methamphetamine, known variously as ice, Tina, Tish, or crank. The locals are seemingly obsessed with constant stories of death (as in fishermen drowning during sudden storms, birds perishing in their hundreds of thousands, beaches made exclusively from the crushed bones of fish skeletons, various species of flora and fauna dying out as the level of salinity, which is already close to that of the Pacific Ocean, keeps on climbing in the hot and pitiless sunshine).