Iceland’s volcanoes are thought to be fed by a plume rising up through Earth’s mantle.

ARCTIC IMAGES/Alamy Stock Photo

A molten puddle deep under Iceland may reveal where volcanoes get their lava

On the boundary between Earth’s core and its mantle sit 10 to 20 blobs of rock that are nothing like the rest of our world’s subterranean realm. For more than 2 decades, scientists have pondered the nature of these mysterious regions, called ultralow-velocity zones (ULVZs). Now, researchers examining one nearly 3000 kilometers below Iceland finally have an answer: They may be the partially molten roots of plumes of hot rock that slowly rise through the mantle to feed volcanoes.

That would make ULVZs deep signposts that mark the base of the world’s plumes, says Ed Garnero, a seismologist at Arizona State University in Tempe who helped discover the zones in 1996, but was not part of the new study. “They would tell you probably where you have plumes and upwellings.”

To release heat from the liquid outer core, the solid rock in Earth’s mantle moves in slow, convective swirls, like a churning pot of hot syrup. Earth scientists have long suspected that upwellings in these mantle convection currents would manifest themselves as the plumes responsible for Earth’s volcanic hot spots. Now they have started to see them—at least their upper parts—with sophisticated computer models that use the waves from large earthquakes to create CT scan–like tomographic pictures of Earth’s interior. 

The picture gets cloudier in the lower mantle, where the ULVZs live. The regions get their name from the way that earthquake waves travel so much more slowly through them. One way to explain that speed drop would be if they were partially molten. Another camp has held that the speed drops can be explained if ULVZs are made of a dense, different type of rock, perhaps enriched with iron, and chemically distinct from the rest of the mantle.

Previous studies had made tentative connections between ULVZs and the plumes underneath Hawaii and Samoa. But study author Barbara Romanowicz, a seismologist at the University of California, Berkeley, says the scene underneath Iceland provides a better picture. That’s because earthquake waves pass underneath the region from different directions and can be picked up by sensors on opposite sides of the world, unlike the Pacific islands.

Using earthquake waves picked up by arrays of sensors in the United States and China, her team better identified the position and shape of the ULVZ. They found its shape was a stubby cylinder—like a pillbox—800 kilometers across and 15 kilometers tall, more or less directly under the plume that feeds Iceland’s volcanoes, they report today in Science. She says her team’s results favor the partially molten scenario, since the other option, a chemically distinct rock, would likely have a more irregular shape and wouldn’t necessarily wind up sitting directly underneath a plume. “It’s a much more natural explanation,” Romanowicz says. “You can relate it directly to what’s going on in the plume. The temperatures are hotter.”

But Allen McNamara isn’t sure the new study rules out the chemically distinct rock scenario. A geodynamicist at Michigan State University in East Lansing, McNamara models the mantle’s slow-motion currents. And he finds that, along the core-mantle boundary, the currents are lateral, drawn toward the bases of plumes like an upside-down bathtub drain. These currents would bulldoze the dense, chemically distinct rock toward the plumes, and, over time, they could pack it into a roughly circular shape. “If you have anything dense in the lower mantle, it’s going to go along for a ride on the conveyor belt,” he says. Garnero adds that, as is the case with many phenomena, it doesn’t have to be either-or: The ULVZs could be both partially molten and chemically distinct.

Romanowicz says the debate will get resolved as pictures of the lower mantle improve. More powerful computers will allow her to use more of the high-frequency content of earthquake waves, the part that is best at illuminating shallow structures like ULVZs. Another boost, says Garnero, could come from ocean-bottom earthquake sensors. With most earthquake sensors sitting on land, two-thirds of Earth—the oceans—is a blank spot. “It’s like an ultrasound on the womb with the wand held on one spot,” he says.