According to the distribution of aftershocks, the fault was about five hundred miles long from the northeast to the southwest and more than one hundred miles wide from the southeast to the northwest. This made it a huge fault. But given the size of the earthquake, it had to be huge. Energy had to have been building up, for a long time, along the fault before it slipped on March 27.
How had the energy built up? As the oceanic crust slid underneath the continental crust, friction in the zone where the two met would cause the continental crust to be compressed. It would be as if the oceanic crust were tugging on the continental crust and squeezing it. Plafker thought he had seen evidence for this on Middleton Island. He’d seen young rocks that had clearly been folded and otherwise deformed, indications that they had been under compression. The earthquake had occurred when the strain between the two types of crust became so great as to overcome the friction between them. This resulted in the sudden release of an enormous amount of stored-up energy over the wide area of the fault.
Plafker had his fault, and he had his earthquake mechanism. It all made sense to him. It explained both the uplift and the horizontal movement: when the fault ruptured, the continental crust rebounded like a spring, up and out, in the opposite direction from the compression. This crust movement also explained the secondary faults on Montague Island, which came to be called splay faults (and were later determined to be the cause of some of the destructive tidal waves). And as he had figured out before, the movement explained the subsidence that occurred to the northwest of the uplift zone.
This was no ordinary thrust fault: it was so big, and produced such enormous effects when it ruptured, that just calling it one seemed inadequate. It would eventually come to be known as a “megathrust” fault.
—
Press’s paper, written with a Caltech undergraduate, David Jackson, appeared in Science—a publication of the American Association for the Advancement of Science and the most prestigious general science journal in the country—on February 19, 1965. Plafker had been generally aware of the findings, so he wasn’t shocked to see them in print. Yet he still couldn’t understand why Press had come to the conclusions he had.
According to their analysis, a near-vertical dip-slip fault ran southwest to northeast for close to five hundred miles along the hinge line. It began, they calculated, about 10 miles below the surface. In itself that wasn’t unreasonable, Plafker thought, and it would explain why there was no surface evidence of the fault. But Press and Jackson proposed that their fault extended to depths of 60 to 120 miles. This, as they acknowledged in the paper, was an order of magnitude deeper than any known earthquake fault.
To Plafker, the extreme depth should have been a clue that perhaps their fault concept was incorrect. For one thing, the data showed that the deepest aftershocks originated only about 25 miles down. If the fault was as deep as Press and Jackson said it was, some of the aftershocks should have occurred at greater depths.
Plafker kept working on his ideas, convinced more than ever that he was on the right track. He began writing what would turn out to be, by the standards of academic journals, a long paper—more than three thousand words, with numerous photographs and diagrams—and submitted it to Science.
The work, “Tectonic Deformation Associated with the 1964 Alaskan Earthquake,” was published on June 25, four months after Press’s paper. Plafker was the sole author, and he laid out the evidence he and others had collected, most notably the more than eight hundred measurements of uplift and subsidence. He detailed the use of barnacles and other marine organisms as reference marks. He wrote about the large tidal waves and what was known about their cause, and about the secondary faults that had been discovered on Montague Island.
The last page and a half of the thirteen-page paper was devoted to discussing the main fault and other aspects of how the quake originated. Quietly, with the kind of dry, clinical language befitting a scholarly paper, Plafker dismantled Press and Jackson’s arguments. In describing the changes in land level he and others had recorded, he noted that “these measurements showed no abrupt changes of level indicative of vertical fault displacement.” In discussing the idea that, of the two possible solutions, a near-vertical fault was the correct one, he wrote that “the hypothesis appears to pose more problems than it answers.”
The most serious of these problems, Plafker continued, was the absence of any surface displacement at the fault line. There was no scarp or other feature in the earth where the fault was presumed to be, but there was plenty of displacement to either side of it.
He then pointed out that Press and Jackson, “in an elegant analysis of the displacements,” had shown that the uplift and subsidence could be accounted for by a fault that began ten miles below the surface. This was the extra-deep fault that to Plafker seemed unbelievable. He cited four “serious objections” to it:
1. A near-vertical fault of this size and destructiveness should have broken the surface somewhere.
2. It should have produced land-level changes opposite of what actually occurred. The land to the northwest should have been uplifted, and that to the southeast should have subsided.
3. It was far deeper than any known fault.
4. Aftershocks occurred on only one side of it.
In place of Press and Jackson’s hypothesis, Plafker laid out his own, that the fault in question was a low-angle thrust fault. He also described the mechanism by which energy had been stored in it: “The postulated stress pattern could result from progressive underthrusting of the oceanic crust and mantle beneath the continental margin.”
The main weakness of the thrust-fault argument, he pointed out, was that while it could account for the uplift and the seaward lateral movement, it had a harder time accounting for the subsidence. But Plafker detailed his idea that the subsidence might be “a secondary effect resulting from elastic deformation” when the land was pulled and stretched.
Toward the end of the paper Plafker trod lightly, describing his findings as speculation and noting that much more research was needed. But he also left little doubt that he believed his ideas about the fault and the origin of the earthquake were correct
Plafker had had two things going for him, he thought. One was that he had seen the effects of the earthquake for himself. He strongly believed in the importance of observation; there was no substitute for it. He had practically lived and breathed the earthquake for months, traveling all over the affected area, talking to people who had lived through it and measuring the land level changes himself. Before that, earlier in his career, he’d seen things in the field, notably in Bolivia, that had informed his thinking as well. So when he started analyzing the data, he had an intuitive understanding of what made sense and what didn’t. Press and Jackson’s solution just didn’t fit with what he’d seen and experienced.
Second, Plafker was convinced that seafloor spreading and its related concepts were real. He felt that his findings fit well with those concepts, certainly much better than Press’s did. A near-horizontal thrust fault fit much more easily with the idea of oceanic crust sliding underneath continental crust than did the near-vertical fault that Press proposed. But Plafker’s conviction was stronger than that. As he saw it, the only way to understand what happened in south-central Alaska on March 27, 1964, was to accept the ideas that had originated with Alfred Wegener half a century before and had been altered and adapted by the likes of Harry Hess, Fred Vine and others.
They were ideas that would soon be further altered and adapted and supported by the work of scientists around the world. In a few years—thanks in part to Plafker’s work—they would become fully accepted, and much better known, as the theory of plate tectonics.
The invitation was a little unusual. It was the spring of 1967, and George Plafker was being asked to make a presentation at one of the biggest scientific conferences of the year—of geophysicists. The forty-eighth annual meeting of the American Geophysical Union was to be held in Washington in mid-A
pril, over four days at the Sheraton-Park Hotel, the sprawling convention facility overlooking Rock Creek Park. It was not lost on Plafker that he, a geologist who spent much of his time in the field looking at rocks, would be speaking at a session with nine other scientists who spent much of their time at a desk looking at plots of seismic or magnetic data. Among those nine were some of the world’s leading thinkers on the subject of the workings of the earth’s crust, including the man responsible for the theory of seafloor spreading, Harry Hess. Plafker might feel a little out of his league—he would say later that he was the “token” geologist among the presenters—but he was flattered that the geophysicists thought he had something important to say.
His paper in the journal Science hadn’t exactly caused a sensation when it was published in 1965. Not that most people expected it would, but even those who might have been disposed to see great significance in it—after all, the paper quietly implied that the Alaska earthquake was a real-world validation of the ideas that Alfred Wegener and Hess and others had talked about all these years—were busy looking for their own proof of those ideas.
The publication in 1963 by Fred Vine and Drummond Matthews of their hypothesis that oceanic crust consists of sections, or stripes, of alternating magnetic polarity had set off a rush to find data that confirmed it. Much of that data was already available, and much of it had been obtained by researchers at the Lamont lab, where being a nonbeliever in seafloor spreading had not prevented the director, Maurice Ewing, from establishing a vigorous program to conduct magnetic surveys of the oceans. Analysis of the data over the next few years led to a sea change, of sorts, at Lamont, with more scientists joining the mobilist camp. Even Ewing showed signs of softening his position. By 1967, several papers by Lamont researchers had been published that validated the hypothesis.
Besides paleomagnetism, research in other fields contributed to growing acceptance of the seafloor-spreading idea. Another Lamont researcher, Lynn Sykes, had studied earthquake data to better characterize the midocean ridges. Bryan Isacks and Jack Oliver, both from Lamont as well, reported on evidence for subduction in seismological data from the vicinity of Tonga, in the South Pacific. In 1965, J. Tuzo Wilson, a geophysicist at the University of Toronto, proposed the idea of transform faults, between segments of the midocean ridges, recognizing that crust could sometimes move in ways that didn’t result in its being created or consumed. And in 1966, Dan McKenzie of the University of Cambridge applied thermodynamic principles to studying the mantle and came up with a more detailed mechanism of crust movement.
With momentum growing, the Geophysical Union decided to devote a full day of sessions to seafloor spreading and related subjects at its 1967 meeting. But there were many other subjects on the program: the conference included a who’s who of geoscience delivering talks on cutting-edge research in fields like meteorology, hydrology, volcanology and planetary science. Among those giving presentations was Frank Press, who had moved from Caltech to the Massachusetts Institute of Technology and had turned some of his attention to the moon. He was on the bill at two sessions on lunar seismicity on Monday and Tuesday, including one with his old adviser Maurice Ewing of Lamont.
The sessions on seafloor spreading were scheduled for Wednesday. The morning session included talks about seafloor magnetics, sediments, gravity studies and other subjects. Plafker’s presentation was one of ten in the afternoon. He would go fourth, and like the others among the first nine speakers would have twenty minutes. Hess, whose ideas were the star of the show, would go last and have thirty minutes.
Plafker’s talk was titled “Possible Evidence for Downward-Directed Mantle Convection Beneath the Eastern End of the Aleutian Arc.” It was a further discussion of what he’d stated in his 1965 paper, and although the title used the word possible, Plafker left no doubt about what he believed, that the Alaska earthquake was evidence of “downward directed mantle convection,” or subduction, of oceanic crust.
Press, who was finished with his own presentation duties, was in the audience for Plafker’s talk. Afterward, Press approached him and, by Plafker’s recollection, acknowledged that his own 1965 paper proposing a vertical earthquake fault had been wrong and that Plafker’s analysis had been correct. Plafker was stunned. Here was one of the most renowned scientists in the country telling him—a geologist among geoscientists—that he’d gotten it right.
Press wasn’t the only distinguished researcher to approach Plafker after his talk. Clarence Allen, who had taken over as interim director of the Caltech Seismological Laboratory after Press left for MIT, came up to Plafker with a suggestion: Why don’t you look into the 1960 earthquake that struck the coast of Chile? The conditions there appeared to be similar to Alaska: oceanic crust colliding with continental crust, in this case South America. If your ideas about the Alaska quake are correct, Allen told Plafker, you should find confirmation in Chile.
The earthquake, which had struck about one hundred miles offshore in the midafternoon of May 22, 1960, was generally considered the strongest ever recorded. It was centered about 350 miles from the capital, Santiago; the nearest city of any size was Valdivia, and it was devastated.
The Chile quake was like the Alaska one, only more so. The shaking lasted longer (ten minutes by some estimates), more houses were destroyed (thousands in Valdivia alone) and more and larger tidal waves did even more damage, both in Chile and hours later across the Pacific. A wave estimated at thirty-five feet hit Hilo, Hawaii, killing sixty-one, and on the Japanese islands of Honshu and Hokkaido nearly two hundred people died. The death toll in Chile has never been ascertained, but at least 1,600 people, and probably more, were killed—as in Alaska, mostly by water. Even in 1967, some parts of coastal Chile were still struggling to recover.
Plafker had leaped at the idea of studying the quake, and Allen had the power to help him obtain a grant for the work so he could take time off from the Geological Survey. With funding secured, he went to Chile in 1968, at first talking to fishermen along the coast about tidal changes there. His Spanish, still impeccable from his time in Guatemala and Bolivia, proved invaluable. Then he’d chartered a fishing boat, the Atun, with a Chilean crew, to measure uplift and subsidence directly among the many small islands that dot the Chilean coast. The work was not as straightforward as that in Alaska—much of the elevation shifting had occurred underwater—but in a paper published in 1970, Plafker described the same basic changes he’d seen in Alaska. There was a zone of uplift to the west and one of subsidence to the east that together encompassed a huge area. He identified a “zone of faulting” that ran roughly six hundred miles north to south and was at least thirty-five miles wide. As in Alaska, he identified the fault as a thrust fault dipping only slightly from horizontal (although the direction was different because off Chile the seafloor was moving almost due east, not to the northwest as in Alaska). As Clarence Allen had suspected, the Chile quake was essentially the same as the Alaska one: the result of oceanic crust sliding, or subducting, beneath continental crust.
Plafker and a colleague, James Savage, wrote a paper describing the findings. By the time it was published in the Bulletin of the Geological Society of America, more scientists were becoming comfortable with referring to those sections of crust as “plates.” The ideas of Wegener, Hess and others were increasingly being accepted and combined into what was becoming known as the theory of plate tectonics.
One of the talks given at the American Geophysical Union meeting in 1967—one that was somewhat overlooked at the time—had played a role in this increasing acceptance. It was given during the morning session by Jason Morgan, a young geophysicist from Princeton, and was the last talk before the noon break. Because of the timing, many in the audience left before he started, wanting to beat the long lines at lunch. Morgan also changed the subject of his talk—rather than discussing how ocean trenches were formed, the topic that had been published in the meeting program, he announced that it would instead be about “rises, trenches, great
faults and crustal blocks.” Over the next twenty minutes, Morgan proceeded to lay out, essentially, the theory of plate tectonics. This was believed to be the first time that the term plate tectonics was used in a public setting.
Morgan’s was a work of synthesis, which he described in greater detail in a paper submitted in August 1967 and published in 1968. Dan McKenzie, who had moved to the University of California at San Diego, and a colleague, Robert Parker, published a paper in December 1968 describing the same general ideas.
Simply put, the theory holds that the outer layer of the earth consists of a number of rigid sections, or plates—scientists at first identified about a dozen large ones, though smaller ones were discovered later—that are in constant slow motion with respect to one another. Upwelling of hot magma at ridges is the engine that drives the motion.
It’s not that simple, of course. There would still be plenty of work to do to flesh out the theory—indeed, plenty of work is still being done today. And it would be years before it was completely accepted, for all intents and purposes. But in its own way, the theory of plate tectonics is now considered as consequential as Darwin’s theory of evolution (although plate tectonics was the work of many people, not one man). The theory shows that Alfred Wegener’s basic idea, that the surface of the earth is dynamic, was correct, even if some of his specifics were wrong (for one thing, it’s not just the continents that are moving—all of the earth’s crust is in motion). But plate tectonics explains much more than crustal movement. It is a unifying theory for all the geological features and processes that humans have wondered about for centuries; they can all be seen through its lens. It accounts for mountains and rift valleys, for volcanic eruptions, for hot spots and ocean trenches and tall undersea mountains.
The Great Quake: How the Biggest Earthquake in North America Changed Our Understanding of the Planet Page 24