Sediment Layer May Forecast Greatest Earthquakes
Researchers at Yale and the University of Washington report that great earthquakes, like the 2004 Sumatra earthquake, may be caused by the build up of sediment on top of subduction zones, suggesting a new way to forecast these most severe earthquakes.
Subduction zones are the boundaries where two tectonic plates collide — one plate pushes over and one pushes under the other. The most severe earthquakes in recent history — in Indonesia in 2004, Alaska in 1964, Chile in 1960 and the Pacific Northwest in 1700 — occurred at subduction zone faults. In the United States, there are subduction zones along the Aleutian margin of Alaska and the Cascadia margin bordering the west coast of the Pacific Northwest.
Schematic cross section of subduction zone showing the relationship of the plates to sediment, the surrounding ocean and land mass.
“Seismologists have long known that the motion of the plates at subduction zones can be smooth and steady in some areas, and sticky and unsteady in other areas,” said Mark Brandon, professor of geology and geophysics at Yale and senior author on the paper appearing in the February issue of the journal Geology.
Earth plates move, and earthquakes ensue in these subduction zones, but some quakes are far more damaging than others. Doctoral student Christopher W. Fuller and associate professor Sean D. Willett at the University of Washington, along with Brandon at Yale, believe they have found a key to identifying specific areas within a subduction zone that will produce the most severe damage when they rupture.
The Earth’s surface is laced with about 32,000 miles of subduction zones, and the motion along the margins of the zones averages a slip of about two inches per year. However, where the margins stick and then rip apart into earthquakes, displacements have been as much as 65 feet, over many hundreds of miles, in a matter of only tens of minutes, according to Brandon.
The team used computer simulations to determine how the upper plate deformed above a subduction zone fault. “The leading edge of the overriding plate will continually deform, much like snow in front of a snow plow blade,” according to Brandon. “As one plate moves under another and the upper plate deforms, it breaks up adhesion on the subduction fault and reduces its ability to generate greater earthquakes.”
Past research has shown that as a subducting plate slides beneath an upper plate, stress builds where the plates meet and stick, and the upper plate warp creates a wedge and a bowl-shaped depression, called a forearc basin. Beneath the sea, this basin fills with sediment that empties from nearby rivers. It appears that the most severe subduction zone earthquakes occur in areas where such sediment-filled basins are found.
These current simulations showed, however, that when sediment was deposited on top of the overriding plate, it reinforced the edge of the plate and caused it to “stick,” directly above where the earthquake would happen so that it no longer deformed internally. The researchers speculate that this allows the subduction zone to remain at rest for longer periods of time and thus to “stick,” making it more prone to earthquake events.
“This phenomenon is analogous to a mayonnaise jar in the refrigerator. The lid opens easily if you use it every day. But if you open the jar infrequently, adhesion will make it difficult to open,” said Brandon. “A sharp tap on a counter breaks the adhesion and the lid opens with a quick spin. In the Earth, the earthquake marks the break down of adhesion on the subduction zone.”
“Over millions of years, the sediment typically piles to great depths, from a half-mile to nearly two miles, and in rare cases it might reach three miles deep,” said Fuller, the lead author on the paper. “The increased weight of the sediment stops deformation from occurring.”
This modeling could have implications for forecasting areas within a subduction zone, such as Cascadia, great earthquakes are the most likely to occur. But the work is not applicable to every subduction zone because each has different characteristics. “You have to understand the nature of basins and how they work in each area before you can use them as an interpretive tool,” Fuller said.
A release from the University of Washington on this paper can be found online. This research was funded by the National Science Foundation.
Citation: Geology: (February 2006)