SANTA CRUZ, CA--For the first time, scientists have directly measured the amount of heat flowing from the molten metal of Earth's core into a region at the base of the mantle, a process that helps drive both the movement of tectonic plates at the surface and the geodynamo in the core that generates Earth's magnetic field.

The boundary between the core and the mantle lies half-way to the center of the Earth, at a depth of 1,740 miles (2,900 kilometers). Seismologists are able to probe the structure of this region by studying its effects on seismic waves generated by earthquakes. The new temperature measurements, published in the November 24 issue of the journal Science, were obtained by relating seismic observations to a recently discovered mineral transformation that occurs at the ultrahigh pressures and temperatures prevailing near the core-mantle boundary. "This is the first time we've had a 'thermometer' that tells us the temperature half-way down to the center of the Earth," said Thorne Lay, professor of Earth and planetary sciences at the University of California, Santa Cruz, and first author of the paper.

"If our interpretation is right, it gives us the temperature at two different depths right above each other, so we get not just the absolute temperature but the rate at which the temperature is changing with depth, as well as laterally," Lay said. "This temperature gradient tells us the amount of heat flowing out of the core into the base of the mantle in that location."

As heat flows from the outer core into the mantle, it drives important processes in both the mantle and the core. The mantle is a thick layer of silicate rock that surrounds a dense, predominantly iron core. The outer core is molten liquid and surrounds a solid inner core about the size of the moon. The cooling of the liquid outer core results in fluid motions in the molten metal that produce electric currents, which generate the geomagnetic field.

Heating at the base of the mantle, meanwhile, drives upwellings of hot mantle material that may rise to volcanoes at the surface and contribute to the slow shifting of tectonic plates. These plates consist of the thin, rocky crust and the rigid top layer of the mantle. They float on the deeper mantle, which is solid but plastic enough to flow very slowly, and their movements trigger earthquakes and gradually change the positions of continents.

"Heat flow is the holy grail, because it tells us how much energy powers the geodynamo, and it tells us how much the mantle is being heated from below. The approach we used is the most direct method so far for getting that information," Lay said.

Lay's coauthors include John Hernlund of the Institut de Physique du Globe in Paris, Edward Garnero of Arizona State University, and Michael Thorne of the University of Alaska, Fairbanks. They applied innovative methods for analyzing seismic signals and used a supercomputer to process a large amount of high-quality seismic data, more than ever before analyzed for a localized region in the Earth. The analysis required 72,000 hours of computer time at the Arctic Region Supercomputing Center and produced very detailed seismic velocity models for the deep mantle under the central Pacific.

Their investigation also relied heavily on laboratory studies of mineral physics. Under the extreme pressures and temperatures deep in the Earth, minerals are squeezed into crystal structures not seen on the surface, except in a few specialized mineral physics labs. If scientists take the common mineral olivine and squeeze it--subjecting it to the ultrahigh pressures and temperatures associated with increasing depth in the Earth--the mineral goes through phase transitions involving sudden reorganizations of its crystal structure.

These phase transitions change the mineral's seismic properties--how fast it transmits certain seismic waves--enabling seismologists to detect where the phase transitions occur deep in the Earth. The depth of the transition tells researchers the pressure, and from that they can get the temperature based on laboratory calibrations, since the pressure at which the transition occurs depends on the temperature.

"If we detect a sudden change in the seismic properties of the mantle, we can associate that with a phase transition in the minerals, and we can use the laboratory calibrations to tell us how hot it is. But until two years ago, we never had that kind of information for the lower mantle," Lay said. In 2004, Japanese researchers working in the laboratory discovered a new form of high-pressure mineral, called postperovskite, that is likely to occur in the lower mantle. Lay and his coauthors detected the phase transition to postperovskite from its precursor perovskite in the lowermost mantle near the core-mantle boundary. Moreover, they observed that the mineral appears and then disappears with increasing depth, forming a layer or "lens" of postperovskite.

"The reason it transforms back into perovskite is that the temperature increases very rapidly right above the core--so rapidly that this high-pressure form becomes unstable," Lay said. "We also see that this layer becomes thinner as you move laterally and eventually thins out and disappears, which you would expect if you have a lateral increase in temperature."

The researchers suspect that upwelling of hot mantle material may be taking place at the edges of the lens of postperovskite. They detected the lens in the lowermost mantle southeast of Hawaii, an area where previous studies have suggested there is an upwelling hot mantle plume from near the core-mantle boundary that may be responsible for the Hawaiian Islands chain of volcanoes. The temperature at the upper boundary of the lens, where the phase transition from perovskite to postperovskite occurs, is around 2,500 kelvins (4,000 degrees Fahrenheit). At the lower boundary, where the reverse transition occurs, the temperature is around 3,500 kelvins (5,800 degrees Fahrenheit). These two points gave the researchers a temperature gradient from which they calculated the heat flow, or thermal flux: about 80 million watts per square meter. Extrapolating to the entire surface of the core gave a total heat flow of about 13 trillion watts.

"We think we are in a relatively hot region of the mantle, and cooler areas will have an even higher heat flux, so this probably sets a lower bound on the total heat flow across the core-mantle boundary. The numbers you might read in a textbook are about one-third of that," Lay said. Such a high heat flow supports the idea that the upwelling of hot plumes of mantle material from near the core-mantle boundary makes a significant contribution to mantle convection, the slow turnover of mantle material that moves tectonic plates on the surface. It also suggests that the solid inner core may be relatively young.

"The core must have been pretty hot in the past for this much heat to be still coming out, and the inner core, which is slowly solidifying from the inside out as the core cools, may be only about a billion years old," Lay said. "These implications are not well constrained, but it's amazing that you can go from detecting seismic reflections to this long-term perspective on how the whole system seems to work," he added. "It's a remarkable convergence of advances in seismology, mineral physics, and thermodynamical models of deep mantle processes."

###

This research was supported by the EarthScope and Geophysics Programs of the National Science Foundation (NSF). The high-quality seismic data analyzed in this study was obtained, in part, by the deployment of hundreds of new seismic stations in the western United States as part of NSF's EarthScope Program.