Almost two years after convincing the scientific community that most of the Earth's mantle is pretty uniform in composition, researchers at MIT -- in two related articles in the March 19 issue of Science magazine -- proposed a model that may explain why the mantle seemed to comprise two dissimilar and separate sections to begin with.
The researchers hope that their new "hybrid convection model" will lay to rest nagging questions about the nature of the mantle -- a 3,000-kilometer-thick layer of hot, molasses-like moving rock between the crust and the core -- and at the same time introduce new ways to think about the Earth's heat-transfer system as a whole.
"We expect that this model will form a new framework for further investigations of the chemical and thermal evolution of our planet," said Robert D. van der Hilst, associate professor of earth, atmospheric and planetary sciences and co-author of both studies.
For almost 50 years, scientists have debated whether the heat transfer called convection occurs throughout the entire mantle at once -- creating a huge mixing pot of essentially the same stew -- or separately in the upper mantle, which extends from near the surface to about 660 kilometers in depth, and the lower mantle, from 660 to about 2,880 kilometers. The second scenario would mean that like oil and water, there are two chemically distinct sections of the mantle that almost never mix.
Using computer simulations and mountains of data to create a kind of CAT scan of the Earth, the researchers demonstrated that previous evidence for separate upper and lower mantles may be explained by certain goings-on in the very depths of the mantle -- an area about 1,000 kilometers from the molten core.
In this area, shifts in densities due to increased quantities of iron and silicon, partially offset by skyrocketing temperatures, may account for minute, previously unexplained differences in the composition of magmas. Researchers have long noticed these differences in the mid-ocean ridges and ocean islands, where, after being heated deep within the planet, the mantle reveals itself in volcanic eruptions.
Professor van der Hilst's co-author on one of the Science papers is MIT graduate student Hrafnkell K������rason; the other paper has Louise H. Kellogg of the University of California at Davis and Bradford H. Hager, the Cecil and Ida Green Professor of Earth, Atmospheric and Planetary Sciences at MIT as co-authors. The latter paper offers simulations of what the circulation within the layers of the mantle would be like if the effects of the bottom layer's massive heat production, composition and density are taken into account.
On the basis of a wide range of evidence from geophysics and geo-chemistry, the researchers argue that a transition in the mantle's structure and composition occurs in the middle of the lower mantle at about a depth of 1,700 kilometers, and that elusive "reservoirs" of high radioactive heat production and distinctive chemical composition reside in the bottom 1,000 kilometers of the mantle.
"We realize that this is a first-order model, but it's more realistic than ones in the past," said Professor van der Hilst, who also holds a Kerr-McGee career development professorship. "A lot of evidence can be explained with this model, but we still don't know much about the ultimate origin and nature of this layer. We hope to provoke a lot of interest on the topic in multiple disciplines."
To unravel the chemical and thermal evolution of the Earth, scientists need to understand how the planet loses the heat that is generated internally by radioactive decay and liberated by cooling from the planet's previously hotter state. This process of convection causes the massive plates that cover the planet to move slowly across the Earth's surface. Scientists agree that the plates slowly sink into the mantle when they collide at island arcs, but it has been difficult to determine how deep into the mantle the plates sink.
At about 650 kilometers, there is a well-documented discontinuity in the properties of the mantle that some scientists believe forms a logical boundary between the upper and lower mantles. The amount of heat flowing out of the Earth is much greater than can be accounted for by the production of heat in the upper mantle, and the chemical composition of basalts erup-ted at midoceanic ridges suggests that large fractions of the mantle have been prevented from mixing since the Earth's formation 4.5 billion years ago. These facts have been used to suggest that the mantle convection occurs in separate layers, with a boundary coinciding with the discontinuity.
But in 1997, Professor van der Hilst and colleagues presented definitive evidence that tectonic plates sink from Earth's surface more than halfway into the lower mantle, thus swinging the pendulum toward a view of a homogeneous, single mantle.
Although the 1997 milestone ruled out long-term segregation of the upper and lower mantle across a barrier at 660 kilometers, the whereabouts of the "reservoirs" of material enriched in heat producing and other geochemically distinctive elements remained enigmatic, said Professor van der Hilst. If the whole mantle overturned, everything within it would have to be homogeneous, which would not explain the trace elements and noble gas isotopes that indicated that different materials existed in different parts of the mantle.
"By measuring the heat flow coming out of the Earth and comparing it to the amount produced by the decay of radioactive material, we came to the conclusion that significant parts of the mantle contain much more heat-producing stuff than others," Professor Hager said.
In their new model, the researchers require a difference in mantle makeup -- not at 650 kilometers, but at around 1,700 kilometers. In fact, they argue that perhaps only the bottom 1,000 kilometers of the mantle contain anomalous chemical reservoirs of heat-producing elements.
Professor van der Hilst, a seismologist, gathers and interprets millions of pieces of data that provide a kind of three-dimensional CAT scan of the inner Earth. By analyzing the speed of seismic waves that propagate through the interior of the planet, he infers information about the properties of the materials through which they pass.
When the ocean floor sinks into the mantle, for instance, it stays cold for a time and this temperature difference is reflected by changes in the speed of the seismic waves that pass through it. "You can see a lot of detail" this way, he said.
One of his observations was that when the ocean floor sinks down to 1,700 kilometers, it starts to change. "The simple pattern that is related to plate motion at the Earth's surface disintegrates," Professor van der Hilst said. "There's something going on at those depths that you can't explain by differences in temperature alone. There's strong evidence that there is a change in composition, possibly in iron content." These heavier, denser materials would be prevented from mixing with the mantle as a whole, thus providing reservoirs of heterogeneous material within a largely homogeneous mantle.
This work is funded by the National Science Foundation and the David and Lucile Packard Foundation.
A version of this article appeared in the March 31, 1999 issue of MIT Tech Talk (Volume 43, Number 24).
