Contact: Robert Sanders
University of California, Berkeley

Millimeter deviations from the expected wobble of the Earth’s axis are giving geophysicists clues to what happens 1,800 miles underground, at the boundary between the Earth’s mantle and its iron core.

A new theory proposes that iron-rich sediments are floating to the top of the Earth’s core and sticking like gum to the bottom of the mantle, creating drag that throws the Earth’s wobble off by a millimeter or two over a period of about 18.6 years.

“The wobble is explained by metal patches attached to the core-mantle boundary,” explained Raymond Jeanloz, professor of geology and planetary science at the University of California, Berkeley. “As the outer core turns, its magnetic field lines are deflected by the patches and the core fluid gets slowed down, just like mountains rubbing against the atmosphere slows the Earth down.”

The theory, first proposed by Bruce A. Buffett of the Department of Earth and Ocean Sciences at the University of British Columbia, also explains a peculiar slowing of seismic waves that ripple along the core-mantle boundary.

Buffett laid out the theory at the December meeting of the American Geophysical Union and in an article with Jeanloz and former UC Berkeley post-doctoral fellow Edward J. Garnero, now at Arizona State University’s Department of Geological Sciences in Tempe, in the Nov. 17 issue of Science. Much of the work was done while Buffett was on sabbatical at UC Berkeley.

The wobble values that the theory explains have been adopted by the International Astronomical Union as its standard for calculating the position of the Earth’s axis into the past as well as the future.

As the Earth spins on its axis the moon and sun tug on its bulging equator and create a large wobble or precession, producing the precession of the equinoxes with a period of 25,800 years. Other periodic processes in the solar system nudge the Earth, too, creating small wobbles – called nutations – in the wobble. The principal components of the nutation are caused by the Earth’s annual circuit of the sun and the 18.6 year precession of the moon’s orbit.

While these nutations have been known for many years, extremely precise geodetic measurements of the pointing direction of the Earth’s axis have turned up unexplained deviations from the predicted nutation.

An annual deviation that lagged behind the tidal pull of the sun first suggested to Buffett 10 years ago that strange processes may be going on at the boundary between the mantle, made up of viscous rock that extends 1,800 miles below the crust, and the outer core, which is thought to be liquid iron with the consistency of water. The inner core, made of very pure, solid iron, rotates along with the outer core, dragging the Earth’s magnetic field with them.

“The Earth is getting pulled and tugged at regular periods, but we observe a difference in the way the Earth responds to these tugs and pulls and what we predict,” Buffett said. “One of the ways you could explain that is by having some dissipation in the vicinity of the core-mantle boundary as the fluid moves back and forth relative to the mantle. But the viscosity of the fluid core is comparable to water, and having water slosh back and forth relative to a rigid mantle wasn’t going to produce the kinds of dissipation we needed to see.”

He hit on another way the rotating core could dissipate energy: via electrical drag.

Based on experiments Jeanloz had performed on the chemistry of rocks at the high temperatures and pressures characteristic of the core-mantle boundary, Buffett suggested that silicon-containing minerals would float to the top of the liquid outer core, carrying iron with it. Together they would form an iron-rich, porous sediment at the mantle boundary that would stick to the mantle, settling into depressions.

Because the Earth’s core rotates about a slightly different axis than the mantle (due to
the tug of the Sun and Moon), the core’s magnetic field is dragged through the mantle, passing unhindered because the mantle does not conduct electricity. The porous, iron-containing sediment stuck to the mantle, however, would resist the rotation of the magnetic field, creating just enough tug to perturb the Earth’s rotation.

“As the core rotates it sweeps the magnetic field with it, which easily slips through the mantle with no resistance,” said Buffett. “But if the bottom of the mantle has conductivity, then it’s not so easy to slip the magnetic field lines through the mantle. The magnetic field tends to stretch and shear or pull out right across the interface. That generates currents, and those currents damp out the motion and create the kind of dissipation we need to explain this lag in response.”

The sediment layer would have to be less than a kilometer thick (about half a mile) in order to have the observed effect, and would probably cover only patches of the outer core.

Support for the idea that a thin layer of iron-rich silicates may be plastered to the underside of the mantle came from the work of Arizona State University’s Garnero and his colleagues, who use seismic waves to probe the mantle and core. They had observed very thin layers at the core-mantle boundary in which seismic waves slow to a crawl. Using Buffett’s ideas, Garnero modeled what a thin silicate layer would do to seismic waves and found agreement with the data.

The team subsequently predicted where these patches are located, based on where seismic waves slow down substantially and where they do not.

“Think of it as a fuzzy boundary between the mantle and the core, with patches perhaps 10 to 20 kilometers across and up to a thousand meters thick,” Jeanloz said.

The rising sediment eventually would squeeze out the iron, leaving the silicate sediments tucked to the bottom of the mantle as the iron falls toward the solid iron inner core. The rising of the silicate contaminants and the subsequent fall of metallic iron would create a convection in the outer core consistent with what geologists think to be the source of the core’s magnetic field. Thus, the rising sediments and falling iron could rev up the Earth’s dynamo.

“In one of the popular models, created by Gary Glatzmaier and Paul Roberts, the dynamo is powered mainly by the growth of the inner core as light elements get excluded and float up through liquid iron, driving convection that powers the dynamo,” Buffett said. “If this idea about sediments is right, the sediments would add a component to drive flow from the top down. This is going to have a pretty important effect on the style of fluid motions in the core, and even in the way in which the magnetic field gets generated.”

The silicates stuck to the mantle also might be caught up in mantle convection and carried to the surface, accounting for reports of core material in lava erupting from hot spot plumes like that under Hawaii.

Though Buffett first proposed his theory 10 years ago in his PhD thesis, the data to prove it were not available. In particular, long-term measurements were needed to accurately determine an out-of-phase anomaly in the 18.6 period wobble.

“Now, with more than 20 years of data, we can confirm that the discrepancy is there and is explained very nicely by the Earth’s magnetic field causing friction at the bottom of the mantle,” Jeanloz said.


The work was supported by the National Science Foundation, the University of California Institute of Geophysics and the Natural Sciences and Engineering Research Council of Canada.

Note: Raymond Jeanloz can be reached at 510-642-2639 or Bruce Buffett is at 604-822-3466 or Edward Garnero is at 480-965-7653 or For images, check out