Read our COVID-19 research and news.

Under pressure. This serpentine was fractured after being subjected to the high pressures and temperatures found in the mantle.

Dried-Out Rocks Shake It Up

The origin of earthquakes that occur far below Earth's surface, including some of the most powerful ever recorded, has puzzled geologists for years. Now researchers may have found the answer by creating smaller versions of these quakes inside a laboratory.

When one of Earth's oceanic plates collides with a continental plate, the thicker continent forces the thinner, denser ocean crust into the molten mantle below. This happens along the coasts of Oregon, Washington, Chile, Japan, and many other places. As the slab of crust slowly dives, earthquakes continuously rattle it to depths of about 600 kilometers. The conundrum is that earthquakes require brittle rock to fracture, but below 50 kilometers or so temperatures and pressures are so intense that the crust deforms like soft plastic instead of breaking.

Geologists suspected that the extreme conditions might dry out the rock and make it more brittle. To test this idea, geophysicist David Dobson of University College London and his colleagues at the University of Bayreuth in Germany experimented with serpentine in the lab. They subjected this water-bearing mineral, commonly found in the upper mantle, to conditions similar to those found deep in the mantle.

Dobson reports in today's issue of Science that at mantlelike pressure and temperature, chemical reactions do indeed force water out of the serpentine. As the reaction continues and water builds up in the pores of the rock, pressure mounts. Finally, the water fractures the rock. For just this instant, the soft, plastic rock behaves brittly, just as soft Silly Putty breaks apart if it is jerked apart. During the reaction, Dobson's team measured elastic waves analogous to the seismic waves that shake Earth's crust. If enough pressure were to build up from the dehydration in the crust, it could potentially create a large earthquake as deep as 200 kilometers, says Dobson.

Dehydration could be a good explanation, but only if there is enough serpentine in the crust and upper mantle, says geologist Derrill Kerrick of Pennsylvania State University, University Park. This is a contentious issue, he says, and more work needs to be done before he will be convinced.

Related sites
The Science paper
Department of Earth Sciences, University College London