Columbia University New York, N.Y. 10027 Office of Public Information (212) 854-5573
Scientists have discovered why the world's largest earthquakes occur in some places where the earth's crustal plates collide but not in others. The question has puzzled earth scientists for 25 years.
The answer is a great "sea anchor force" at work in the earth's thickly fluid mantle below regions where the plates converge, scientists at Columbia University's Lamont-Doherty Earth Observatory and the University of Chile report in the November issue of the Journal of Geophysical Research.
The earth scientists found that when an oceanic plate meets another plate and bends down into the relatively soft mantle below, it does so at a shallow angle if the two plates are moving toward each other, but at a steep angle if they are moving in the same direction over the underlying mantle. The sea anchor force, long used by mariners to aim their ships in stormy seas, determines the angle of descent, or "subduction," into the mantle, they found. Shallow dips intensify earthquake-causing friction at the critical zones where the plates rub together. Deep dips relieve the friction, so that earthquakes do not occur.
The research was conducted by Christopher Scholz, a geophysicist at Lamont-Doherty, Columbia's earth sciences research institute in Palisades, N.Y., and Jaime Campos, a seismologist at the University of Chile in Santiago.
A sea anchor is a drogue, or canvas bag, that has great resistance to being pulled through water. In a storm, sailors throw one over the bow on the end of a cable. When the wind blows against the ship, the sea anchor is forced upward through the water to a shallow angle as it retards backward movement. When the wind blows with the ship, the sea anchor is forced downward through the water until it hangs straight down.
The descending plate moves through the viscous mantle like a sea anchor through water, report the two scientists. When the upper plate is moving over the mantle toward a descending plate, the downgoing slab is forced up through the mantle into a shallow angle of descent--like the sea anchor of a ship against the wind. When the two plates are moving in the same direction over the mantle, the sea anchor force pushes against the slab and bends it down steeply--like the anchor of a ship going with the wind.
"Since the plate tectonics revolution of the 1960's, scientists have known that the world's largest earthquakes occur at subduction zones, at the frictional interface where the two plates rub together," Professor Scholz said in an interview. "More than 80 percent of the world's earthquake energy is released in these places. What has remained unexplained, however, is that, for some reason, some of these subduction zones are silent. They accommodate the plate motion without generating significant earthquake activity."
For example, along the coast of Chile and western Mexico, the Pacific seafloor is pushing eastward against the South American continental plate, which is advancing westward over the mantle. The oceanic plate slides under the upper plate at a shallow angle. The sea anchor force exerts pressure against the zone where the two plates rub together. Because friction is high, sliding occurs in intermittent jerks that generate large earthquakes, like the one that struck Mexico in mid-October. The compression at the interface also helps raise up the Andes and the Sierra Madres on the upper plates.
In the western Pacific near the Mariana Islands, on the other hand, the descending Pacific plate and the upper Philippine Sea plate are moving in the same direction over the mantle. The sea anchor force pushes the Pacific plate
nearly straight down, relieving friction where the plates meet. When the stress falls below a critical level, sliding occurs smoothly, without jerks. Thus, earthquakes are rare and small at this subduction zone.
"The sea-anchor force works like a switch," Professor Scholz said in an interview. "If the upper plate is moving toward the trench, or not moving, it forces the subducting plate closer to the horizontal, the stress at the contact zone is high, and earthquakes are turned on. If the upper plate is moving away, the subducting plate moves downward, the stress is relieved, and at a certain point, the earthquakes are turned off."
The new theory by Professor Scholz and Dr. Campos thus also explains the 25-year-old mystery of why back-arc basins form in some places around the globe but not in others. The retreating Philippine plate is pulling away and rifting apart. This allows molten material to rise and create new ocean floor behind an arc of volcanic islands--a so-called back-arc basin. But in other subduction zones, the upper plate is not pulling away rapidly enough, so back-arc spreading does not occur, the scientists say.
Writing in JGR, the scientists said: "The model we have offered is so very simple that one wonders, at least in retrospect, why it has not been suggested earlier, for this is a fairly old problem. One possibility is that it involves something rather counterintuitive."
Scientists and others had thought that the massive wall of the subducting slab could not easily move horizontally through the viscous mantle. "However, this is just not true," Professor Scholz and Dr. Campos said: The slab moves almost as easily across the mantle as it does down into it.
To arrive at their conclusion, Professor Scholz and Dr. Campos worked together at the Laboratoire de Sismologie at the Institut de Physique du Globe de Paris, where Professor Scholz spent a sabbatical year and where Mr. Campos worked on his doctoral thesis. Theoretically viewing the plates and mantle as a simple mechanical system, they calculated all the physical forces at work.
The scientists then applied their model to the real world, including the subduction zone between the Pacific and Philippine plates, which has confounded the best previous theory to explain deep-dipping subduction. Scientists had proposed that older oceanic lithosphere, which has lost more of its nascent heat, is colder, denser and more disposed to sinking. But although the subducting Pacific plate is the same age everywhere it intersects with the Philippine plate, it produces areas where large earthquakes and back-arc spreading do or don't occur--all on the same plate boundary.
Professor Scholz's and Dr. Campos' model accounts for this range of difference. It explains, for example, why large earthquakes occur near Guam but not anywhere farther north on the same plate boundary and why back-arc spreading occurs behind the Mariana Trench but not farther north along the Izu-Bonin Trench. Applying the model to the rest of the globe, they found that it explains where earthquakes and back-arc spreading occur in more than 80 percent of the world's 29 subduction zones. Of the remaining zones, half had additional complications unaccounted for by the model. About 10 percent still remain unexplained.
The findings explain great earthquakes in Japan, under which the Pacific plate is descending at a shallow angle. In the United States, the model predicts that the Juan de Fuca plate, which is slowly subducting beneath the Pacific Northwest, is capable of generating earthquakes greater than magnitude 8. Washington, Oregon and northern California have not suffered such a quake in historical times, Professor Scholz said, but the intervals between great quakes in this region may be stretched out longer than others.
Professor Scholz's research was supported by the National Science Foundation, IPGP, College de France and Centre National Recherche Scientifique. Dr. Campos' research was supported by the French and Chilean governments and the University of Chile.
11.6.95
18,758