CARNEGIE INSTITUTION OF WASHINGTON
For release Wednesday, 6:00 P.M. EST, 15 January, 1997
Contact news office at 202-745-0780; tmcdowell@ciw.edu, or
Cecily Wolfe at 508-289-3290; cwolfe@whoi.edu, or
Sean Solomon at 202-686-4370, ext. 4444;solomon@dtm.ciw.edu

Imaging a Mantle Plume

A team of scientists from Carnegie's Department of Terrestrial Magnetism and the University of Iceland have produced the first high-resolution seismic image ever obtained of an active mantle plume. Former fellows Cecily Wolf, Ingi Bjarnason, and John Van Decar,* along with DTM director Sean Solomon, have "mapped" the upper mantle directly beneath Iceland, obtaining the first dimensions of Iceland's plume.

Geologists believe that plumes are jets of hot material that arise from deep within the mantle (perhaps even from the core-mantle boundary). As these rising plumes approach the surface, partial melting occurs, and the resulting magma migrates to the surface to produce volcanic eruptions. Plumes are thought to be fixed in the mantle and thus do not move with the Earth's lithospheric plates. Since the concept of mantle plumes was introduced, nearly three decades ago, plumes have figured prominently in theories of mantle convection. They have also been used to explain "hotspot" regions of volcanism. Hotspots occur both in regions where plates are spreading apart, such as Iceland, and, more commonly, in regions within plates, far from plate boundaries, such as the volcanic chain of Hawaii. While such hotspot islands have been mapped by geologists, until now no one had determined what a mantle plume looks like at depths below the surface.

The DTM results, reported in the January 16 issue of Nature, show that the plume beneath Iceland is a hot, narrow jet of upwelling material some 150 km in radius and extending at least 400 km in depth. The temperature of the plume appears to be at least 200 degrees C hotter than the surrounding mantle. The center of the plume lies directly beneath an elevated region of central Iceland presently covered by the Vatnajökull galcier; volcanic eruptions beneath this glacier last fall resulted in severe flooding.

One reason why plumes have not previously been imaged is that they are generally believed to be narrow, less than 300 km across, and images of Earth's mantle obtained from global studies of seisimic waves have been at too coarse a resolution (1000 km) to see them. The Iceland data, however, were obtained from an array of portable seismometers emplaced directly above the proposed hotspot site. The array measured relative travel times of seismic waves produced by dozens of earthquakes over a three-year period (1994-1996). Iceland offers a relatively large land area for station siting (unlike most hotspot regions, which are oceanic seamounts or islands). It is, thus, an ideal location for such an experiment.

The results support a model contrary to that suggested by several recent numerical studies modeling mantle flow and melting beneath Iceland. Some of these numerical studies had predicted a broad, cool plume 300 km in radius and only 75 degrees C hotter than the surrounding mantle.
The experiment began in 1993, when DTM researchers began emplacing an array of 15 portable seismometers across Iceland. The goal of the ICEMELT experiment, now a collaborative effort with the Science Institute of the University of Iceland, was to ascertain the detailed upper-mantle structure beneath Iceland. Since then, the researchers have determined that travel times of seismic P- and S- waves arriving at the seismometers are significantly delayed in an area some 150 km in radius directly below Iceland. (Primary and secondary waves travel through the mantle. They travel in different ways and yield independent results, and are an important tool for probing the Earth's interior.) The data showed that a cylindrical low-velocity seismic region extends down at least 400 km beneath central Iceland.

The authors note that numerical models of plume flow consistent with the new seismic observations suggest that far too much melt should be produced in the ascending mantle for the known thickness of the crust beneath Iceland. The DTM data support a suggestion, put forward earlier by others, that much of the melt must thus channel along the mid-Atlantic Ridge to the south and north of the hotspot. Cecily J. Wolfe is now at the Woods Hole Oceanographic Institution. John C. VanDecar is now an editor at Nature magazine in London. And Ingi Th. Bjarnason is now at the Science Institute of the University of Iceland.