Earth's Inner Core Is Running a Tad Faster Than the Rest of the Planet

Richard A. Kerr

The claim that Earth's inner core was getting ahead of itself seemed odd at first. Why should a 2440-kilometer solid iron ball spin faster than its 3000-kilometer-thick shell of mantle rock? Well, some computer simulations showed the molten-iron outer core dragging the inner core around by the magnetic field generated in the outer core. Still, seismologists had problems with measurements of the inner core's excess spin.

Now, 9 years later, the original claimants are back with persuasive evidence that the inner core really is spinning faster than the rest of the planet. Not as fast as it first seemed, but possibly fast enough to help probe the nature of Earth's layered interior.

The authors reduced the two sources of error bedeviling the original estimate of the inner core's rotation rate. One was the exact location of earthquakes near the South Sandwich Islands in the far South Atlantic Ocean. These moderate quakes send seismic waves down through the inner core and up to a seismograph in College, Alaska.

Thanks to a woodlike grain to the crystalline iron of the inner core, waves passing through it may slow down or speed up, depending on where they pass through. If the inner core rotates faster than the rest of the planet, quakes striking the same place years or decades apart will send out waves that take slightly different paths through the core. Waves from South Sandwich quakes would arrive in Alaska a little sooner than they did the time before, revealing the inner core's "superrotation."

Unfortunately, travel times to Alaska depend not only on the amount of inner core rotation but also on the quake's exact location. But seismologists can't tell precisely where such remote quakes are. So Zhang and colleagues hunted for a pair of quakes that have identical signatures in their seismograms. For the wave shapes to match, the two quakes must take place less than a kilometer apart and they probably overlap. Knowing that such doublets are so close to each other, the group could calculate that the travel time of the waves had changed 0.0090 second per year.

The other source of error is the uneven grain of the inner core. Nine years ago, this grain variation wasn't known, but Zhang and colleagues have mapped it using a technique introduced by seismologist Kenneth Creager of the University of Washington, Seattle. That information enabled them to calculate a superrotation of 0.3° to 0.5° per year, or about 900 years for the inner core to gain one full revolution on the rest of the planet. That's about a third as fast as Song and Richards's initial estimate of 1996 and a tenth of some later estimates. Seismologists are generally impressed. "This paper removes any lingering doubt as to whether the inner core is rotating at a different rate than the mantle," says Creager.

Researchers also seem to be homing in on the size of the excess rotation. Seismologist Guy Masters of Scripps Institution of Oceanography in La Jolla, California, has gauged inner core rotation at 0.1º per year, using an independent method that involves the quake-driven, bell-like ringing of the planet. "I'm happy with 0.2° [or] 0.3°" per year, he says, a range within the error of his estimate. Researchers can now consider what the observed superrotation says about Earth's interior or changes in the length of a day. It might help test computer simulations of how the outer core generates the magnetic field, says geophysicist Bruce Buffett of the University of Chicago, Illinois. That's a lot for a little extra spin.

Science 26 August 2005:
Vol. 309. no. 5739, p. 1313