Opening doors to the core, and more.

Opening doors to the core, and more

Advances in the tracking and analysis of underground earthquakewaves are enabling seismologists to focus on the planet's innards as never before. Recently researchers made the first maps of the earth's core, showing that it is not a uniform, smooth body (SN: 7/5/86, p. 10).

This picture of a bumpy core is supported by a new studydone at Caltech in Pasadena. Olafur Gudmuundsson and his colleagues conclude that the radius of the outer core varies with latitude and longitude, so that the core-mantle boundary is made up of mountains and valleys extending as much as 10 kilometers above or below the boundary. The researchers think there are highs under eastern Australia, Central America, south central Asia, the North Atlantic Ocean and the Northeast Pacific Ocean, and lows under the Southwest Pacific Ocean, the East Indies, Europe and Mexico. They suspect that this topography is caused by the circulation patterns of the mantle.

But seismic evidence is not the only information scientistshave about the inner earth. Caltech's Bradford Hager, Mary Ann Speith at the Jet Propulsion Laboratory in Pasadena and their colleagues have examined two other variables: changes in the day length and in the earth's magnetic field. Scientists have long known that the earth's rotation rate, which determines the length of day, can be sped up or slowed down by winds blowing on mountains. Similarly, "winds," or flow patterns of the earth's liquid outer core, can change the rotation rate by pushing on whatever bumps may be on the core-mantle boundary; the bigger the bump, the greater the effect.

Using the flow patterns determined by other researchersfrom changes in the magnetic field (which is generated by core flow), Speith's group indeed found that the topography of the core-mantle boundary could be affecting the length of day. But in order to be consistent with the length-of-day data, says Hager, the relief at the boundary must be 10 to 100 times smaller than that proposed by Gudmundsson's group.

To reconcile all three data sets, Hager prefers instead a modelin which there are continent-like structures at the base of the mantle made up of a material less viscous than, and perhaps chemically different from, the overlying mantle. These structures would not extend very far into the core, satisfying the length-of-day constraints, and because they would slow down seismic waves passing through them, they could also account for variations in seismic speeds at the core-mantle boundary seen by seismologists.

Going even deeper into the core, Harvard researchers usingtwo types of seismic data have found that seismic waves traveling parallel to the earth's rotation axis go faster than those moving in the earth's equatorial plane. "This means that the crystals inside the [solid] inner core are aligned nonrandomly, in a way that is related to the direction of rotation," says John H. Woodhouse, who works with Adam M. Dziewonski, Domenico Giardini, Xiang-Dong Li and Andrea Morelli. Perhaps as the inner core crystallizes from the liquid outer core, the crystals are mechanically aligned by the fluid flow in some preferred direction, he says. Alternatively, material could be steadily circulating in the inner core in a way that aligns the crystals.

All these studies are just scratching the surface of the innerearth. What excites many scientists is that different branches of geosciences are coming together to address some fundamental, longstanding questions about the planet, such as its structure and how the flow patterns in the core generate the magnetic field.