Kerry Sieh was still astounded by how little was really known of the San Andreas, especially about the southern sections of the great fault. The San Francisco quake of 1906 had focused the geophysical community on the northern segment for seventy years while the potentially dangerous six-hundred-mile stretch to the south remained an obscure curiosity. Of course, to anyone who really looked to the south, the fault appeared to be relatively harmless. In San Juan Bautista, for instance (about a hundred miles south of the Bay Area), the two sides of the fault merely crept past each other at a very slow and steady pace, creating nothing more than minor temblors from time to time. True, a few buildings, curbs, and sidewalks were being wrenched apart slowly, where they sat literally on top of the fault, but the threat was predictable and tame.
Even in Parkfield (a tiny farming community located right on the fault about halfway between San Francisco and Los Angeles) the periodic earthquakes that seemed to occur every twenty-two years reached only a Richter magnitude of 6.0 (ML), and did little, if any, damage. And to the south of Parkfield was the nearly deserted Carrizo Plain, which didn’t really contain enough civilization to worry about.
But then there was Fort Tejon, south of the Carrizo Plain. It was common knowledge that a huge earthquake had struck that portion of the fault in 1857, but no one knew how far the fault had slipped during that quake, or how badly it had affected the area that was now the densely populated Los Angeles basin.3 For that matter, no one even knew if there had ever been great quakes in that area before 1857, even though that question would obviously have a bearing on what Los Angeles could expect in the future.
Considering all those elements, Kerry had decided to structure his thesis project to answer three key questions. First, he needed to know how much the fault had slipped during the 1857 Fort Tejon quake. Secondly, he wanted to find how fast the fault was slipping year by year—the average slip rate over the past few millennia (the last few thousand years). And finally, he wanted to pinpoint when past earthquakes had occurred during the previous few thousand years in order to determine if there were a regular rhythm to their recurrence. Such information could lead to earthquake prediction, at least in the long term.4
The answers to at least the first two questions lay in the offset streambeds of the Carrizo Plain. The aerial photographs he had studied showed numerous small creeks and streams that had been dislocated—wrenched apart—where they crossed the fault. Finding the right ones of those to study would be a starting point.
In the beginning, each of those streams had started as a relatively straight channel flowing from the hills to the east directly across the fault toward the west. But in the early years of each new stream, the day inevitably came when the fault suddenly moved, the western side slipping northwestward, carrying the western part of the streambed with it.
One key to the puzzle lay in discovering when those streams had been created as a straight channel across the fault. Kerry theorized that most of them dated from the tremendous outflow of water from the last ice age, about ten thousand years ago, but he had to be sure. If there was enough carbon in the upper levels of those beds to determine when the channel was first cut, he would have the date—and a starting point. Then, by measuring how far the western channel had been carried off to the northwest on the Pacific plate side over the ensuing centuries, and by dividing the distance with the number of years since the stream was created, he would have the average yearly slip rate.
That still left a problem, however. How many feet did the fault slip with each earthquake? The key to that question, he figured, lay in the offsets of the newest streams, which had been displaced only once—in 1857.
If, for instance, the 1857 quake had caused ten meters of slippage, it just might be that each time the fault moved, it moved in ten-meter increments. Kerry knew that if that were the case, he would find all the offset streams displaced from their original location in ten-meter multiples—ten-meter jumps—as well. And that was exactly what he thought he saw in the aerial photos and in Bob Wallace’s original work.
“The point,” Kerry had explained so many times, “is that if I know the fault is moving at, say, 5.5 centimeters per year on the average, and I know that each time the fault slips in an earthquake it slips a total of ten meters, then I can divide ten meters by 5.5 centimeters, which gives me 181 years. That could mean that it takes 181 years on the average for the fault to store up enough strain (at the rate of 5.5 centimeters per year) to break whatever is snagging it, which would mean that on the average the area has a great earthquake every 181 years.”
Apparently the San Andreas had been snagged in the Carrizo Plain since 1857, and it seemed inevitable that someday that snag would break, allowing the fault to slip back to equilibrum suddenly in a major earthquake which would relieve the strain and start the process all over again.5
Kerry Sieh sat down on dusty rock to rest, looking around at the hills in front of him, the reddish sun now throwing long shadows over the valley. He would need all the bullheadedness and determination he could muster if he were going to succeed here. He felt deep down that this was the right thing to do and the right time to do it, and that this parched terrain would yield its secrets if probed long enough. But he had no more guarantees than a gold miner—and gold miners can grow old waiting for that one, big find.
So far all he had was one particular stream which was typically offset as it crossed the fault in such a way that he might be able to date its origin. That one held promise, although the layers of sediment were not the sort he needed to answer the other question: when the previous quakes occurred.
The same stream also had been featured by Bob Wallace in one of his papers. It would be appropriate, Kerry thought, to name it after the senior USGS geologist. Even in the middle of a semiarid wasteland, a place called Wallace Creek would confer a certain importance. Now all he had to do was dig up the answers—make Wallace Creek yield the right information.
Kerry got to his feet, balancing his bicycle and climbing on for the ten-mile trek back to his car. He would return in the morning. There were many more miles of the San Andreas to cover, and somewhere along that line were gullies he had to find.
Pallett Creek, California—July 1, 1976
“Beautiful! Look at that!”
Kerry Sieh moved closer to the vertical face of dirt his brother’s shovel had exposed, a layer cake of sediments with interbedded peat layers deposited over the last two thousand years. As with every other cut they had made for the past month, this one was replete with clearly defined breaks in the horizontal layers, and areas in which material that looked like sand had pushed up from its natural level through some of the layers above, forming what looked like an upside-down bowling pin.
“It’s a sandblow, right there!”
Rodger Sieh moved closer, puzzled over Kerry’s excitement.
“What?”
Kerry began tracing the history they had exposed.
“There, Rodger. Right there!”
The summer sun had been merciless, but the thrill of discovery was making the hardships endurable. The results from nearly two months of work were already proving to Kerry that his initial excitement was justified—excitement at finding this shady bend in an otherwise obscure creek on the edge of the Mojave Desert. It was the spot he had been searching for in the previous year, up and down the San Andreas Fault, through the Carrizo Plain, and throughout this country of arid landscape and desert vegetation. And the answers it could provide were going to be exciting.
Kerry sat nose to nose with the exposed dirt wall of Pallett Creek’s northern bank. As he looked at the near-vertical break in the strata which could be traced from just below the upper surface of the present-day meadow downward through other breaks in the alternating dark and light layers, it was rather amazing to realize that his left foot was on the Pacific plate, and his right one on the North American plate. He was in effect staring the San Andreas Fault down the throat.
On the sur
face, as he looked back to the west-northwest, the now-familiar outline of two small sloping ridges terminating in a tiny draw running from the west exactly toward his position betrayed the line where the two plates had been sliding past each other for thousands of years. It was typical rift topography. The markings weren’t as clear and stark as they were in the Carrizo Plain 125 miles to the northwest, but there was no mistaking that sort of miniature “valley.”
His practiced eye and careful precision in really looking at what lay before him had spotted a rich, recent geologic history from the very first days of digging. For a few thousand years, it seemed, up to about 1910, the spot where the brothers were standing had been a swamp. With the water table at the surface, the area was washed by the waters of several adjacent springs, and covered periodically by flash floods coming from the nearby north shoulder of the San Gabriel Mountains, waters rich with sediments that buried the grasses and plants over and over again, providing a marvelous stack of layer upon layer of carbon-rich strata which could be carbon-dated.
A free-lance surveyor named Henry Hancock had slogged through the swamp in 1855 as he slowly worked on mapping the southern boundaries of the Mojave Desert.6 At that time, as with another survey in 1904, Pallett Creek was little more than a surface stream flowing through a soggy marsh—the same as it had been for centuries.
Sometime after 1904, however, floodwaters carved a channel through the swamp, eroding the layers drastically, slicing out a gorge thirty feet deep in an ever-widening streambed, which at some point had acquired the name Pallett Creek from a pioneer named George Pallett who had settled the area in the 1860’s. When the elevation of the streambed dropped, the water table dropped as well, draining the swamp, and the layers of saturated soil dried out in the space of a few years as the desert grasses reclaimed the land. Cottonwood trees, which needed constant moisture, followed the waterline down, growing in the stream channel but forsaking the upper surface, where the marsh had been. By the time a small biplane with an aerial photographer flew over in 1930, the stream gorge had filled with fast-growing cottonwoods and willows while the upper surface had become a desertlike meadow standing, hot and dusty, where the willows and marsh grasses of the soggy swamp had been just three decades before.
Now, with a handful of unimpressed horses grazing the slim supply of grasses in that meadow, the two Sieh brothers chopped away at the remains of that dried marsh while below in the stream channel the summer sounds of locusts and unseen buzzing insects accompanied the visual richness of the cottonwoods that typically responded to hot desert breezes by spreading their cottonlike seeds in the air, creating an ocasional incongruous desert “snowstorm.”
The climate simply was not fit for human habitation, the brutal temperatures hardly noticed by Kerry, but increasingly onerous to his new wife, Laurie, whose brief honeymoon (following their May 29, 1976, marriage) had ended in the Mojave Desert at Pallett Creek. By late June their lives were revolving around the arid former swamp—with Kerry and his brother Rodger (a divinity student) attacking the streambank by hand with shovels during the day while Laurie, by her own description, served as chief cook and bottle washer. Because they were working with very limited funds, the motel rooms they were renting in nearby Palmdale were a luxury, but also a necessity. By July, Laurie was spending the blast furnace days hugging the air conditioner in one of the tiny rooms while her husband and brother-in-law worked on at Pallett Creek, Kerry getting more excited with each sunrise.
He had hoped for a place with multiple layers going back a few hundred years. Pallett Creek’s dried banks had that and more. He had needed carbon-rich plant material in each important layer. The strata were filled with peat layers, rich in carbon 14. He had needed a place in which previous earthquakes had ripped open whatever surface existed at that time, and which rapidly building sediments had covered and preserved. Pallett Creek, again, was more than he’d dreamed of finding. And of course, all those requirements had to be fulfilled by a spot whose sediments and telltale strata were on the San Andreas Fault. At Pallett Creek, the main fault trace of the San Andreas ran right through the dried-up swamp, less than five yards from the lip of the creek gorge, where it paralleled the stream for a distance before cutting diagonally across the channel.
It was, in a word, ideal—or at least as close to it as he could have imagined.
Kerry was also beginning to feel more confident about the answers he had found in the Carrizo Plain, with which he might be able to answer at least two of his three key thesis questions. At the end of the previous summer he had found evidence that the western side of the fault had slipped 10 meters (approximately 30 feet) in the 1857 Fort Tejon earthquake. To the south, however, closer to Palmdale, the offsets in the San Andreas were in multiples of only 4.5 meters. That would indicate that the Palmdale section could store only about 5 meters of accumulated strain before it would jump forward, producing a major earthquake in the process, while the Carrizo Plain section could store 10 meters of strain before breaking.7
Did that also mean that the San Andreas Fault, which appeared to be a single rip between the two great plates, was moving at different speeds in different locations? No one knew as yet. Deep in the earth beneath the fault the two plates continued to move at a steady rate past each other according to the accepted theory. The brittle surface, however, was a different story. At each point along the fault the western side would slip forward in a jerky motion of sudden release and slippage, followed by periods of no movement, followed by still another buildup of strain, ending in sudden release and slippage, each cycle causing approximately the same magnitude of earthquake. (What governed how much strain could be contained before a snagged area broke was basically the strength of the snag.) In addition, when all the distances through which the fault had slipped were added up and divided by the number of years, the result yielded the average movement per year—the same rate, of course, as the movement of the plate at great depths.
If the average slip rate per year was the same everywhere along the fault, then a five-meter-jump-per-earthquake rate for the Fort Tejon section would mean that great quakes occurred there, on the average, twice as often as in the Carrizo Plain.
At first Kerry had become very excited when offsets on the Carrizo Plain suggested an average slippage of only 1.5 centimeters per year. “That would mean …” he had figured, rather excitedly, “that the San Andreas only has great quakes about once every seven hundred years, and since the last one was only one hundred and eighteen years ago, that means Southern California has already had its big earthquake and would be out of danger.” Even at half that cycle—350 years—the Fort Tejon area, too, would be still for centuries to come.
It wasn’t, however, that simple, or that fortuitous upon further examination. The evidence he was still evaluating in the Carrizo Plain and at Wallace Creek pointed to a slip rate anywhere from 1.5 to 3.5 centimeters per year, and while that would be slower than the rate indicated by the marine geological studies (5.5 cm/year), the interval of recurrence of great quakes in the area could be on the order of only a couple of hundred years or less. The story would have to come from detailed studies of the offsets in the Carrizo, as well as the actual record of past quakes at Pallett Creek. And to get the latter information, he had to find “the” spot.
Kerry entered the fall semester of 1975 poring over aerial photographs, looking for offsets to the south that were as clean and definite as Wallace Creek, but were located south of Palmdale just north of the San Gabriel Mountains, which bordered the northern Los Angeles area. There were quite a few, he noticed, that had offsets of 130 to 150 meters. If the slip rate of 1.5 centimeters per year (which he still thought might be accurate for the Carrizo Plain) held up as the slip rate all along the San Andreas, and if the streams on the north slope of the San Gabriel Mountains had come into being about ten thousand years ago at the end of the last ice age, offsets of 130 to 150 meters would be about right. (Kerry had made plans to return to Wallace Creek in
the fall to begin the work of carbon dating and measuring offsets, but his initial figures on average slip all were speculation.)
Sometime in the fall, before returning to the Carrizo Plain, he had driven down the back roads south of the small desert town of Pearblossom toward the fault, finding one promising site several miles away from Pallett Creek. Just as in the Carrizo, he needed enough carbonaceous material in the upper surface of the landscape bordering a stream to carbon-date, with the hope that the resulting date would give him the starting point at which the new creek channel had been cut straight across the fault. From there he could measure the distance to the offset streambed on the other side of the fault and figure out the average yearly slip rate.
Levi Noble, the geologist who in the 1920’s had originally raised the radical idea that the San Andreas might have slipped a total of thirty-five miles since its creation, had mapped much of the fault in the same area. When Kerry looked at one of those maps, he was startled to find Noble’s comment that the Pallett Creek drainage contained “interbedded sands, gravels, and peats” at the fault crossing. That was exactly what he needed, and in November 1975 he had driven back to the area to climb the barbed-wire fence by Pallett Creek and take some peat from the top of the streambank (and an old embedded log from the bottom) for carbon dating. This, he thought, might be the ideal place to measure the offset streambed. If the carbon dating could tell him when the stream had first cut a channel in the swamp, he could get an average slip rate.
The creek appeared to have a very typical, characteristic jog where it met the fault, with an offset of 130 meters. Kerry returned to Stanford convinced that the top layer would be dated at ten thousand, and the bottom log at perhaps twenty thousand years.
And he was thunderstruck when the results came in.
The top layer was only two hundred years old; the bottom only two thousand.
Kerry was devastated! The best-looking streambed in the area, and the dates were useless. Obviously Pallett Creek had not been wrenched 130 meters apart in only two hundred years; that would be an incredible rate of 65 centimeters per year, 5.41 centimeters per month, or 1.8 millimeters per day! You could set up bleachers, sell tickets, and bring in crowds to watch the daily slippage if such a rate were realistic. Obviously it was not. Obviously the Pallett Creek samples were a bust, and he had found the wrong layers, missing the ones which could tell him how many thousands of years it had been since the creek first crossed the fault. Obviously …
On Shaky Ground Page 22