High-pressure water triggers tremors

Science News, Nov 21, 1998 by Sarah Simpson

Tiny droplets of water may seem too inconsequential to break rock, but three Yale University geologists now say that water trapped far below Earth's surface can set off earthquakes.

Deep underground, heat and pressure expel water from the crystals of certain minerals, transforming one structure into another. "The classical view of metamorphism is that fluid is released very slowly and continuously in teeny, tiny amounts over millions of years," says Jay J. Ague, a coauthor of the report appearing in the Nov. 15 Geophysical Research Letters. "What we show is that once one of these [transformations] gets going, a lot of water can be released over decades or hundred-year timescales."

If water is squeezed that quickly out of crystals 10 to 20 kilometers below the surface, it soon fills pores in the rock and is left with nowhere to go, Ague says. When its pressure overcomes the weight of the overlying rocks, the water blasts some escape routes. Cracking rocks already under immense strain in a fault zone is a good way to start an earthquake, he points out.

Ague and his colleagues Jeffrey Park and Danny M. Rye wondered whether this process could trigger mysterious earthquakes such as Loma Prieta. That 1989 California quake started nearly 18 km underground, deeper than seismologists thought possible along the San Andreas Fault. Most tremors there occur in the top 12 km of the crust as opposite sides of the fault slip past each other. Deeper rocks are so hot that strain does not tend to build up: Like Silly Putty, they flow when pushed slowly but snap if stretched abruptly.

In the Yale team's proposed mechanism, water would strain the rocks so rapidly that they would break even at that depth. "The [Loma Prieta] trigger was at the very deepest place possible, which makes the idea worth entertaining," says Kevin P. Furlong of Pennsylvania State University in State College. Still, he asks whether water can be released rapidly enough to break rock that is pliable enough to flow.

Water might diffuse peacefully through the rock if it were released gradually as the minerals converted from one type to another. The Yale scientists' analysis, however, indicates that rock being heated holds water beyond the equilibrium temperature for the mineral transformation and then releases it quickly. They modeled reactions in which the mineral serpentine loses its water to become olivine. Overstepping the critical temperature by as little as 5 [degrees] C can generate enough pressure to shatter the rocks, Park observes.

Ague admits there's more work to be done. "There are hundreds of reactions that can take place, so we've really only scratched the surface," he says.

If the researchers' theory proves right, its implications don't stop after a single rumble. "If the fault seals up after an earthquake, the reactions keep going and fluid pressure builds up again," Ague says, "so this process could happen over and over."