The Great Quake: How the Biggest Earthquake in North America Changed Our Understanding of the Planet

by Henry Fountain

  In the mid-1960s, however, there were still plenty of “stabilists” around. Some accepted the concept that there was upwelling of magma at the ridges but had a much harder time with the idea that the oceanic crust was recycled at the continental margins. Instead, they thought that the magma just added to the earth’s crust, making the planet bigger. This “expansion theory” had first been proposed in the 1930s and, through various permutations, still had adherents.

  In early 1963, a young English graduate student at the University of Cambridge came up with an idea that was, as he described it later with typical British understatement, a “rider” to Hess’s hypothesis. The student, Frederick Vine, had been inspired to study earth science eight years before when, as a fifteen-year-old in West London, he’d cracked open a geography textbook while studying for his O-level exams and saw a diagram illustrating how well the coasts of South America and Africa fit together. It was the same congruence of coastlines that had inspired Wegener. In reading the text that accompanied the illustration, Vine had learned that although there was a theory that explained this close fit by proposing that the continents had once all been together but had since come apart, geologists had no idea if it was true. Vine thought it remarkable that something as basic as this was not known. He went on to study geology at Cambridge as an undergraduate and in early 1962 attended a conference at the school at which Harry Hess, as guest speaker, gave a talk on seafloor spreading. Vine was hooked on the idea of trying to prove that the hypothesis was correct.

  In the graduate program, Vine was assigned to work with a geology professor, Drummond Matthews. One of Matthews’s responsibilities was to obtain geomagnetic data; Vine’s job, as his research assistant, was to help interpret it. Beginning in the 1950s, scientists at several institutions, including Lamont Geological Observatory in New York, had been collecting data about the magnetism of the seafloor. They relied on a fact about rocks that had been known for centuries: many, notably volcanic ones that contain a lot of iron, are magnetic. They become magnetized when they form, cooling from extreme heat, and groups of atoms within them line up with the earth’s magnetic field. As the rocks harden, that alignment is locked in. The rocks then contain a permanent record of the orientation of the magnetic field—of the direction to the magnetic pole—and the field’s intensity at the time the rocks formed. This can be measured using an instrument called a magnetometer.

  Researchers in the field of what became known as paleomagnetism had been studying rocks collected on land in this way, using lab-based magnetometers. In the mid-1950s, Keith Runcorn, a physicist at Newcastle University in England, looked at data from rocks from different time periods in Europe and concluded that the magnetic pole had migrated over millions of years—what was called polar wandering. Later, he compared that data with similar studies of North American rocks and realized that a better explanation was not so much that the magnetic pole had wandered, but that the rocks—and thus the continents themselves—had moved. This spurred renewed interest in Wegener’s concept of continental drift among Runcorn and others. Eventually, with enough data from rocks around the world, these scientists were able to accurately infer how the continents had shifted in relation to one another, findings that added to the growing evidence that Wegener had been onto something.

  To collect data from the seafloor, scientists from Lamont and elsewhere trailed a magnetometer off the back of a ship as the vessel tracked back and forth across a survey area. The magnetic maps that resulted were intriguing, as some areas showed stronger magnetism than others. Often these formed a pattern—narrow strips of the seafloor, alternating strong and weak, like the stripes of a crosswalk.

  Scientists studying these maps were not sure what the patterns represented. They had known for some time that rocks can exhibit what is called reverse polarity. That is, the locked-in magnetism in some rocks is oriented one way; in others it is aligned in the opposite direction. Hypotheses had been proposed to account for this, including what seemed like a radical one: that from time to time the earth’s magnetic field flips, and that the magnetic pole, what we call magnetic north, becomes magnetic south. Then at some point it flips back again. This was much more than polar wandering; this was polar reversal. Rocks that formed during a period when the magnetic field was oriented as it is today would have one kind of polarity. Those that formed when the field had flipped would have reverse polarity.

  The idea that the earth’s magnetic field reverses was not widely accepted—for one thing, no one had a satisfactory explanation as to why it would happen—but it began to gain more credence as more data suggested that the field had flipped often over millions of years. But it was still a contentious issue when Vine arrived for graduate study in the fall of 1962. As it happened, Matthews was away on a research cruise aboard the HMS Owen, an old Royal Navy minesweeper, conducting magnetic surveys across a stretch of the Indian Ocean that included a midocean ridge. When Matthews came back later in the fall, Vine looked at the data. Like that from other surveys, it contained patterns of anomalous readings.

  Vine had a thought. What if these readings represented areas of alternating polarity, polarity that was locked in as—following Hess’s reasoning—magma oozed out of the midocean ridge and cooled to become new seafloor? As seafloor kept being produced over millions of years and moved toward the continental margin, reversals of the earth’s magnetic field would create stripes of seafloor of alternating polarity. That would account for the crosswalk-like appearance of the magnetic maps.

  In a paper published in 1963, what came to be known as the Vine-Matthews-Morley hypothesis (recognizing that a Canadian scientist, Lawrence Morley, had come up with the idea independently at roughly the same time) offered a neat way to prove the seafloor-spreading hypothesis. The data that Vine and Matthews offered wasn’t clear-cut, so the stabilists, especially, were skeptical. But if seafloor spreading actually occurred, further analysis of magnetic data—and by then there was a lot of it—should show stripes of seafloor of alternating polarities, and these stripes should be roughly symmetrical on either side of a midocean ridge. What’s more, analysis of the age of rock samples from the seafloor should show the stripes becoming progressively older the farther they are from the ridge.

  —

  By the fall of 1964, when George Plafker started working on his data in earnest in his office in Menlo Park, no such proof of the Vine-Matthews-Morley hypothesis—and thus of Hess’s seafloor-spreading hypothesis—had been published. But Plafker wasn’t waiting for proof. By then, he was becoming something more than a field geologist. In the months after the earthquake he had learned a lot of geophysics—in particular, seismology—and he would learn a lot more. He was aware of the disagreements between stabilists and mobilists and found the arguments put forward by the mobilists compelling. But he also had an intuitive sense that they were right. He kept going back to his time in Bolivia and the dropstones he’d seen on the eastern side of the Andes. He knew they had to have been deposited by glaciers, and it seemed to him that the glaciers could have come from nowhere else but another continent—one that had once been attached to South America but had long since drifted away.

  Plenty of scientists weren’t convinced that Hess and others were right, especially in the United States. Lamont Geological Observatory had a number of researchers who were skeptical of “spreaders and drifters,” as the mobilists were sometimes called, even though the lab had produced much of the data regarding the physical and magnetic characteristics of the ocean floors. Lamont’s director, Maurice Ewing, once asked a colleague before a talk about seafloor spreading, “You don’t believe all this rubbish, do you?” and as is not uncommon in science, some lower-level researchers took their cues from those above them. There were also stabilists at Menlo Park, geologists who took a rather narrow view: they didn’t see evidence for the movement of continents in their own work and thus rejected the idea.

  But some noted researchers at the Survey were firmly in the mobilist
camp, and Plafker was fortunate that one occupied a nearby office. His name was Allan V. Cox.

  Cox wasn’t just any researcher. Just three years older than Plafker, already by 1964 he was one of the foremost experts on paleomagnetism in the world, one whose work had been cited by Vine and Matthews in their paper the year before. A native Californian, Cox had dropped out of the University of California at Berkeley early during his freshman year, had served in the merchant marine for several years, had worked as a field assistant studying glaciers in Alaska and then had been drafted into the army. Upon being discharged he returned to Berkeley, eventually graduating with a degree in geology. After more time in Alaska, in the mid-1950s he came back to Berkeley once again for his doctorate, studying the magnetism of rocks under a faculty adviser who was a firm believer in continental drift. In 1959, with his PhD in hand, he’d been hired at the Geological Survey. Cox, with another Survey scientist, Richard Doell, quickly waded into the subject of magnetic field reversals. Their idea was to look at magnetized rocks around the world and determine their ages using the latest dating techniques, to put together a timeline for when the magnetic field had reversed.

  It was an ambitious undertaking, involving a lot of fieldwork (and they had to persuade the Survey to hire another scientist who was an expert in dating). But they’d published a paper in 1963 that established a partial timeline of reversals—a proof of concept, as it were—based on analysis of samples from six lava flows in California.

  Plafker was familiar with their work. In fact, he had been responsible for inadvertently ruining Doell’s summer a couple of years before, when Doell had come to Alaska to find rocks for his magnetic studies. Plafker was helping a colleague map rock formations in the Wrangell Mountains with an eye on potential deposits of copper ore, and Doell had worked out of their camp. There were layers of volcanic rock in the area dating back to the Permian period, more than 250 million years ago, which would be perfect for his research. He’d spent weeks collecting samples while Plafker and his colleague were doing their mapping work. Then, toward the end of the summer, Plafker came across a layer of shale-like rock below the volcanic layers. In looking at the shale closely, he noticed it contained fossil shells that were easily recognizable as a certain kind of extinct clam—but one from the Triassic period, less than 250 million years ago. If the shale was less than 250 million years old, that meant that Doell’s volcanic layers, above the shale, must be younger, and not from the Permian, as he had thought. Thanks to Plafker’s find, weeks of Doell’s work went out the window.

  In the fall of 1964, Cox was still in the thick of his paleomagnetism research for the Survey, though in a few years he would leave for a professorship at Stanford, where he’d work for two decades until his death in 1987. Plafker thought Cox was brilliant and a bigger thinker than most of the other scientists at the Survey, who seldom looked much beyond the quadrangle they were mapping or the basin they were studying. Plafker would go to Cox’s office for brief conversations about what he was finding in the data, the conclusions he was coming to. Cox was an unabashed mobilist and offered Plafker encouragement. So did George Gryc, the Alaska branch chief who had sent him to Anchorage after the earthquake. Gryc thought it great that Plafker had thrown himself into this work, when so many others would have complained that it was a career diversion.

  Plafker’s task was essentially this: how to describe what had happened beneath Alaska beginning at 5:36 p.m. on March 27 in a way that best accounted for the observed effects, primarily the changes in land level around the state. What mechanism would cause such massive uplift and subsidence over tens of thousands of square miles of land?

  The key, he knew, was to understand the major fault. The two faults he’d seen on Montague Island had to be only secondary features. But where precisely was this main fault, how big was it and, most important, how was it oriented? He’d seen no evidence of it on the land surface, but there were clues in the data.

  From what Plafker knew of seismology, his choices of a type of fault were limited. Faults can be complicated, but they can be divided into three basic categories:

  Strike-slip, in which one side, or block, slips past the other horizontally.

  Dip-slip, in which one block moves past the other vertically, or nearly so.

  Thrust, in which one block rides over the other at a low angle.

  In the case of the Alaska earthquake, Plafker could rule out a strike-slip fault. It was extremely difficult to see how rocks sliding against each other horizontally could cause either side to rise up or sink, especially by the amounts measured over such a wide area of south-central Alaska.

  As for the other two types of faults, “first-motion” studies provided some clues. A first-motion study analyzes the waves from an earthquake to determine the fault plane, the angle that the fault makes in the earth. The problem with these kinds of studies is that they normally produce two possible solutions rather than one. With the Good Friday quake, the analysis suggested that the fault could be at a steep, nearly vertical angle—a dip-slip fault—or at a low, nearly horizontal one—a thrust fault.

  Plafker knew that a well-known geoscientist—no less than the head of the seismology lab at the California Institute of Technology, in Pasadena—was also working on the earthquake and had determined that the fault must be a steep, dip-slip fault. In the geology and geosciences community, as in most scientific communities, word gets around as to what people are working on. But Plafker also knew this because the scientist Frank Press had asked Plafker for his data, which he had shared.

  Five years older than Plafker, Press was, like him, a native New Yorker who had graduated from City College. But unlike Plafker, who had found a job after graduation and headed west, Press had continued his education at Columbia, getting a doctorate in geophysics in 1949. His adviser had been Maurice Ewing, who guided Press in designing an improved type of seismograph. Press remained at Columbia until 1955, when he moved to Caltech, which at the time was considered the world’s leading institution for research in seismology. Two years later he was named director of the Seismological Laboratory, a somewhat surprising choice given his relative youth. But Press, while an extraordinary scientist—among many accomplishments, he was at the forefront in understanding the seismic signatures of earthquakes versus atomic bomb tests and had helped design the worldwide network of seismographs that measured the Good Friday earthquake—also had a head for policy making. In the early 1960s he had served on the President’s Science Advisory Committee and had been a delegate to international meetings in Geneva and Moscow on limiting nuclear weapons testing. In 1962 and 1963 he was president of the Seismological Society of America.

  Although Plafker had readily shared his data with Press—it was the government’s, after all—as he’d heard through the grapevine about Press’s conclusions, he found them difficult to accept. How could a near-vertical fault produce the kind of large elevation changes on either side of it that had been seen across such a huge swath of south-central Alaska? To Plafker, it couldn’t. If a dip-slip fault had had such a widespread effect, surely there would be some sign of the fault on the surface, like a scarp where one side of the fault was now higher than the other.

  But Press was an important and highly respected scientist, and in the groupthink that sometimes affects science, plenty of lesser scientists felt that his explanation must be the correct one. Plafker, an unsung field geologist who became involved only because of his knowledge of Alaska, who until a few months before had known almost nothing about seismology, felt trepidation going against such an authority. But he was slowly figuring out a mechanism for the earthquake that differed from Press’s and, with the encouragement of Cox and others, felt his solution was correct.

  Plafker thought the alternative fault orientation—a fault that dipped at a low angle to the northwest—was the right one for two main reasons.

  One had to do with the aftershocks. As he’d learned before, the vast majority of these occurred within the
land that had uplifted. But the depth of the strongest aftershocks—about 130 that were of Richter magnitude 5.0 or higher—was even more of a clue. These tended to be deeper the closer they were to the hinge line. That tendency would be expected if, as Plafker thought, the fault was dipping to the northwest at a low angle, toward the hinge line.

  But the other reason had more to do with what Plafker had come to learn, and accept, about seafloor spreading. If crust that made up the floor of the Pacific Ocean was sliding and sinking beneath the crust that made up Alaska and the rest of the North American continent, it wasn’t hard to envision that this was happening at a shallow angle. There’d be no reason to think, for example, that the heavier ocean crust would dive sharply down when it met the continent. Rather, a slow, gentler, more gradual sinking made intuitive sense.

  So to Plafker’s thinking, the earthquake fault was a thrust, not a dip-slip, fault. It occurred where the oceanic crust and continental crust were in contact, as the former was sliding, or subducting, under the latter. Given that the fault was nearly horizontal, the fact that no surface trace of it had been discovered now made sense.