RSS Feeds

Email a Friend Printer Friendly

Study reconciles long-standing contradiction of deep-earth dynamics

August 25, 2005

New databases give researchers a look into processes inside the Earth's mantle

Researchers at Columbia University's Lamont-Doherty Earth Observatory recently resolved a long-standing contradiction about the workings of the deep Earth. For years, many geochemists have argued that parts of the deep mantle remain unchanged since the formation of the Earth, whereas many geophysicists and geodynamicists have held that the entire mantle has been convecting (moving and mixing) over geological time.

Based on a synthesis of data on global oceanic magmatism, Cornelia Class and Steven L. Goldstein show that the evidence actually favors whole-mantle convection, with the deepest parts of the Earth affected by the tectonic processes that occur at the surface. Their study will appear in the August 25 issue of the journal Nature.

"For thirty years scientists have been debating whether there is a layer in the mantle that has remained unchanged since the formation of the Earth," said Class, a Doherty Associate Research Scientist. "The new on-line databases made it possible for the first time to reevaluate the geochemical arguments based on a complete synthesis of global data on oceanic basalts. We found that the strongest evidence previously put forth in favor of a layered mantle actually indicates the opposite is true."

The question of whether the Earth's interior operates on a "layered" or "whole-mantle" model is central to scientists' understanding of how the Earth loses its internal heat. The main process of heat loss occurs through melting of the mantle to form magma. If the layered model is correct, then a large portion of the deep earth never melts and never reaches the surface. Evaluations of seismic waves generated by earthquakes indicate that continental and oceanic plates sink all the way to the core-mantle boundary, an observation that supports whole-mantle convection. However, evidence from trace amounts of helium in lavas have been interpreted as requiring that the mantle is composed of layers that are isolated from each other.

When magma is erupted by volcanoes, helium and other gasses from the mantle are expelled to the atmosphere. Unlike other gases, the helium is so light that it is lost forever to space. As a result, the Earth's inventory of 3He, the light isotope of helium, is considered "primordial," dating from the time of the formation of the planet. Indications of a high proportion of primordial helium in ocean island lavas, like those found in Hawaii, have been taken as evidence for a layer in the deep mantle that has never been melted and, hence, never degassed.

"This result adds to growing evidence that most of Earth's mantle has been subject to the same forces that drive the movements of Earth's crust," said Sonia Esperanca, a Program Director in the National Science Foundation's Division of Earth Sciences, which funded the research.

Class and Goldstein's re-evaluation of this concept of the inner Earth was based on their work with two new databases that for the first time compile all of the published data on the geochemistry of oceanic volcanism around the world: the Petrological Database of Ocean Floor Basalts (PetDB, based at Lamont) and Geochemistry of Rocks from the Oceans and Continents (GEOROC).

It has long been known that the upper mantle sources of basalt found at mid-ocean ridges, formed by sea floor spreading, have been previously melted to form oceanic and continental crust. The new global data synthesis demonstrates that the ocean island lavas that are chemically most like mid-ocean ridge basalt also contain the highest primordial helium signal. As a result, this helium signal actually indicates previous processing by plate tectonics, rather than a primordial mantle source. Class and Goldstein conclude that helium must be degassed inefficiently to the atmosphere through volcanic processes and enough remains in the mantle during melting to give the false impression that the deep mantle is primordial.

"Our results mean we can dispense once and for all with the argument that the helium data require a primordial layer in the mantle, whose existence has been difficult to reconcile with the rest of what we know about how the Earth works," said Goldstein, a professor of Earth and Environmental Sciences at Columbia University and member of the Lamont-Doherty senior staff. "The implications of our work will be hotly debated, but I expect these new observations to change the way we view deep-Earth dynamics."

The Earth Institute at Columbia University