The Dilemma of Scale

Ethan Coon

(page 3 of 3)

Coon programmed the additional abilities into MLUPS in Python, a powerful computer language, and then added the components to the existing code, which was developed in the legacy language Fortran.  Coon was introduced to Python through a collaboration with staff at Argonne National Laboratory as part of his doctoral research.

The work used real-world data from the Society of Petroleum Engineers’ Tenth Comparative Solution Project — a baseline data set that exists as a testbed for developing and testing computational models.  Moulton says MLUPS is now at the proof-of-concept, desktop stage for two-phase modeling. He’s confident it will demonstrate significant advantages over competing models.

Coon presented the results of his summer research as an invited speaker at the March 2007 meeting of the Society for Industrial and Applied Mathematics (SIAM).

“It was a great experience.  The other seven speakers in my session were all tenure-track professors, and I received great feedback,” he says.  Coon made contacts with researchers at the University of Texas at Austin and the University of Bath in England while there.

The DOE CSGF practicum also was a chance for Coon to see the fine scale in Los Alamos’ scientific community.

“I really liked the ability to get out and talk with the engineers and scientists at the lab — people outside my immediate field — and to work on a larger, multidisciplinary project,” says Coon, who notes that the T-7 group is noted for its highly collaborative approach.

While in New Mexico, Coon took time to get what geologists call “ground truth” with rocks.  He scaled Sandia Mountain outside Albuquerque and also hiked to the summit of Mount Wheeler, the state’s highest peak, near Taos.  Both provided stunning vistas of the surrounding deserts and forests.

Similarly, from his scientific perch at Los Alamos, Coon was able to see the grander-scale picture that ties together his work using computational models to understand how the Earth — and things in it — moves.

“Talking with David Moulton about my thesis work really helped because he had a very different perspective on it,” Coon says.  “In the end I saw that both problems come down to a question of how you incorporate levels of detail that you can’t completely resolve in your model.”

 

« Previous       1   |   2   |   3   |   Print