October 6, 2003

Geologist Awarded NSF Grant To Explore
Earth's Core By Computer

ernon Cormier's work takes him to the center of the earth. But unlike Jules Verne or the Hollywood heroes of the recent movie, The Core, his visits there are virtual.

Cormier, one of two geophysicists on the UConn faculty, uses supercomputers at centers around the country, such as those at the University of Texas, Austin, and at the San Diego Supercomputer Center, to analyze small-scale data and perform the extensive computations that his models of material - or fabric, as geophysicists call it - at the earth's center require.

Vernon Cormier, a professor of geology, with a poster about his research. Cormier studies the earth's core. Photo by Dollie Harvey

Now, a $60,000 National Science Foundation (NSF) equipment grant is enabling him to purchase, among other things, a Beowulf-style computer, a 20-processor cluster computer that will make it possible for him to do high-speed computing on the UConn campus. In addition, NSF has awarded him a four-year, $360,000 grant to support his research on earth structure.

Fast computing is the tool of the trade in geosciences research today. The NSF and a consortium of academic institutions are forming a Geosciences Network, GEON, to build a "cyberinfrastructure" that will network even the super-fast supercomputer centers around the country. It will enable geoscientists to better analyze the complex data sets and 3-D models they work with.

The problems Cormier works on used to take a day of time on 32 processors. Now, with faster computers working in parallel configurations, the same problem can be solved in a couple of hours. Parallel computers work in tandem on separate "slices" of the same model to solve the equations he needs for his research.

The crux of his work is exploring the fine structure of the core-mantle boundary, the place nearly 2,000 miles into the earth where the mantle - the layer of earth below the crust - meets the core. Some of the most puzzling mysteries of geosciences reside there. Among them are questions Cormier asks, such as do deep mantle plumes originate at the core-mantle boundary? Plumes are the hot uprisings of thick mantle liquid that can lead to volcanic eruptions. Do slabs of tectonic plates that slide underneath each other descend as far as the core-mantle boundary, as his research suggests, or do they stop half way?

He also studies the inner core boundary, where the liquid outer core of the earth meets the iron-solid inner core. The earth has a stronger magnetic field than any of the other terrestrial planets because of the movement of liquid in the outer core, he explains. The fluid motion of the outer core, stirred by the solidification of the inner core, makes a magnetic field much as an electric current would. Understanding this liquid-solid interface and how the core solidifies will lead to a better understanding of the earth's magnetic field, he says.

Unlike the plot of The Core, where "terranauts" descend into the earth to fix a disastrous failure of the earth's magnetic field, Cormier explores the earth by computer. He points out that it is impossible to even bore very far into the earth. The pressure is so great that it would be like trying to dig a hole at the beachÑthe sand fills in as quickly as you can dig - and the temperature rises so high that any instrument would melt. (Luckily for them, The Core's geophysicists had access to an indestructible made-for-movies substance, "inobtainium.")

Real-world geophysicists, on the other hand, rely on seismic waves to "see" what is going on beneath us. Studying the inner earth by seismology may be likened to studying the human body through ultrasound, Cormier says. Seismic readings are taken at stations around the world by the Global Seismographic Network, part of a consortium funded by NSF, of which UConn is a member. The data are recorded digitally and stored on disks. In order to model the waves and learn about their structure, Cormier relies on supercomputers to solve the wave equation important to seismologists.

He studies the changes in the shape of the seismic waves interacting with the solid core - focusing on such fine detail as the length of crystals. The plumes he studies may be too narrow to be seen well by studying seismic wave velocity, so he must model wave shape. To study wave structure, he looks for seismic waves that are "not too big, not too small," generated from earthquakes that would be classified at around 5.5 to 6.5 on the Richter scale.

While the questions he explores may seem esoteric, the results of deep earth research are relevant to learning more about the earth's water, atmosphere, and climate. The analytical techniques used in his research can be applied to other problems, including oil and mineral exploration and medical imaging. What he learns can also tell us something about how the earth has evolved and where it's headed, he says.

His theoretical work has been applied to a practical application - monitoring underground nuclear tests for the Comprehensive Test Ban Treaty, which the U.S. Senate approved but the United States did not sign. A network of seismic stations headquartered in Vienna monitor underground activity for the treaty. In a three-year study for an agency of the Department of Defense, Cormier's research group developed models for each of the stations, leading to a correction in determining the arrival of seismic waves at the stations.

Cormier's research group includes a postdoctoral researcher and two Ph.D. students. He has been on the UConn faculty for 16 years and formerly was a senior research scientist at MIT and a postdoctoral researcher at Harvard and the University of Colorado. He received his Ph.D. at Columbia University and his bachelor's degree at Caltech. Besides his appointment as a professor of geology and geophysics, he is on the physics faculty and is a visiting professor at MIT.

Cormier says geophysics is a good field for students, because they become marketable from their knowledge of computing. Geophysics students typically must learn three operating systems, parallel computing, and two or three computer languages.

Aside from that, the field is fun, he says. "It's very rich, earth science. There are all kinds of things you can do with it."