Jan. 14, 2000


Using The Chemistry of Lava to Study Earth's Interior Motion

By Hannah Fairfield

Marc Spiegelman

A Columbia scientist has developed a new way to use geochemical data from ocean floor spreading ridges to understand the processes of the earth's mantle convection and the movement of magma beneath mid-ocean ridges, a discovery that may change the way scientists study the mantle itself as well as the geochemical evolution of the earth.

Marc Spiegelman, a geophysicist at Lamont-Doherty Earth Observatory, and Jennifer Reynolds, who received her Ph.D. from Lamont in geochemistry and is now at Duke University, collaborated on the project and published their research in the Nov. 18 issue of Nature.

"This work attempts to reconcile geophysics and geochemistry, to use the chemistry of rocks as clues to the big picture of earth's internal motion," Spiegelman said. "To do this, however, requires including the dynamics of melting rock and magma transport because these processes control how different chemical components are mixed to form the rocks that we can analyze at the surface."

At Lamont, Spiegelman developed computer models that predict the observable chemical signatures of different mantle processes, such as how the mantle rocks melt into magma and how they rise to erupt at the surface. These models--unlike any previous models of ocean spreading ridges--suggest that variations in the chemistry of lavas erupted on the seafloor can contain important information about how both magma and the mantle move.

Independent of these models, Reynolds and another Lamont geochemist, Charles Langmuir, conducted a detailed mapping and sampling survey at the ridgecrest of the fast-spreading East Pacific Rise at 12šN off the coast of Mexico, and discovered some peculiar lavas that only erupt at a distance from the ridge axis. The composition of these lavas could not be explained using standard models--but the composition is consistent with Spiegelman's predictions.

Together, Spiegelman and Reynolds compared the predictions of the models to the data and showed that the data match well with models that collect melt from a wide region (between 50 and 100 kilometers) around the ridge axis and focus it into a very narrow region only a few kilometers wide. Previous geophysical studies using seismic and electromagnetic techniques have suggested that melt is present over a wide region beneath the Pacific; however, no one before has been able to infer the direction and flow rate of the melted rock. This new study suggests that "chemical imaging" can provide new insights into deep-earth processes.

"This approach has applications to a whole host of other research areas, including the dynamics of subduction zones, mantle plumes like Hawaii and Iceland and the global geochemical evolution of the planet," Spiegelman said. "It should also have applications to other coupled fluid/solid problems such as the flow of water and oil in the Earth's crust. I've been using this approach for well over five years and I think the field is wide open."

Spiegelman admits that one of the driving motivations for this research is the pursuit of the long-term geochemical evolution of the planet. "Geochemistry provides one of the few clues to unraveling the dynamics of the solid earth over the past 4.5 billion years," he said. "However, producing consistent models that explain both chemical and physical observations remains a challenge. This work is just the beginning, but it suggests that magma dynamics could provide the crucial link for reconciling geophysics and geochemistry because it directly connects the large scale motions of our planet to the composition of the rocks we measure."