Boston University | Center for Computational Science
HomeNews and EventsResearchEducationPeopleSeminarsFacilitiesContact Us

CCS Seminar
Professor Paul Hall
Boston University - Department of Earth Sciences
FRIDAY - February 29, 2008
12:00 noon
Physics Research Building - Room 595
3 Cummington Street

"Computational Geodynamics: Exploring the Earth's Deep Interior from the Comfort of the Computer Lab"

The development of the theory of plate tectonics in the late 1960's revolutionized our understanding of the solid Earth, providing a framework through which phenomena such as earthquakes and volcanoes can be understood as the surface manifestation of convective motion in the Earth's deep interior. Over the 40 years since the advent of this new paradigm, the exquisitely slow motions of the Earth's rocky mantle have been investigated using a variety of both geophysical and geo-chemical methods, and they are now thought to be intimately tied to the origin and evolution of life, as well as to a number of mass extinction events that nearly destroyed life on Earth. However, because the Earth's interior is physically inaccessible, we are unable to directly observe convection in the mantle, and therefore many basic questions about the morphology and timescale of these motions remain unresolved. In an effort to answer these questions, Earth scientists have increasingly turned to computational fluid dynamics (CFD) to develop models of mantle convection.

By treating the mantle as a highly viscous fluid, using material properties obtained from high-pressure, high-temperature mineral physics experiments and applying boundary and initial conditions derived from geophysical and geochemical observations, it is possible to produce meaningful CFD models of flow in the Earth's mantle. This type of computational modeling poses many challenges that are not normally encountered within more traditional applications of CFD. For example, the relevant physical processes being modeled cover a vast range of length scales (from the flow of magma along the boundaries of individual mineral grains (L < 0.000001 m) to the dimensions of individual convective cells (L > 1,000,000 m)), and pressures (from 0 GPa at the surface to >100 GPa at the base of the mantle). Furthermore, ductile deformation within the mantle is governed by a range of mechanisms, from diffusion creep (e.g., Coble, Nabarro-Herring) to dislocation creep, resulting in an effective viscosity that is highly non-linear and varies with temperature, pressure, composition, grain size and strain rate. Finally, mantle minerals undergo a variety of both solid-solid and solid-liquid phase transitions as they move around within the mantle, significantly altering their physical properties quite abruptly over very short distances. By way of illustration, a case study using a finite element CFD model in conjunction with a Lagrangian particle method to model the interaction between a thermo-chemically buoyant mantle plume and a mid-ocean ridge system (an analog for the creation of the Easter Island - Salas y Gomez seamount chain) will be presented.











Page last updated on February 19, 2008. Please send comments to Cheryl Endicott

copyright © 2006, Center for Computational Science | Boston University , MA, 02215