Past Awardee

Dynamics of Catastrophic Volcanic Eruptions and Dusty Gas Flows in "Urban Canyons"

Susan Kieffer
Susan Kieffer

College: Liberal Arts and Sciences
Award year: 2005-2006

A number of the most dangerous volcanoes in the world have had, and are capable of future, highly explosive eruptions that occur when the flank of a volcano collapses. Upon collapse, the pressure on a high-temperature reservoir is reduced and the reservoir material then explodes outward--like an exploding steam boiler. Such volcanoes are not rare, and are often densely populated because of their location and fertile soils. As many as 2,000,000 people are endangered by Vesuvius in Italy, and both human tragedies and global-scale economic disruptions could occur when Mt. Rainier near Seattle next erupts.

These flank eruptions produce hot, particle-laden flows, broadly called "lateral blasts" that decimate landscapes over tens or even hundreds kilometers, traveling up, down and ridges and canyons. Specific related flows are called "pyroclastic flows" and "pyroclastic surges". At a smaller scale, the same dynamics were observed as the debris and gases from the collapse of the Twin Towers expanded through the "urban canyons" of New York City on September 11, 2001. These volcanic and urban flows are geometrically complex, requiring a fully three-dimensional representation, and are variable enough that steady-state solutions are not adequate: waxing and waning stages must be considered in order to understand the consequences of the eruptions. Only massive computer simulations can help us understand these flows that are rarely observed and not easily monitored or simulated in the laboratory.

The conditions at Mount St. Helens in 1980 produced a relatively small lateral blast, but it was so well documented that the data provide a chance to validate theoretical models. Models of these flows require solution of complex, nonlinear equations by massive computational techniques, and visualization of the results in a meaningful way requires similar sophisticated capabilities. With such simulations the dynamics of the flows can presented to scientists, disaster managers, and the public in the same way that hurricane and tornado visualizations have been presented. Few places in the world have the facilities, or people, to run such simulations: UIUC is one.

As a NCSA Faculty Fellow, I would plan to explore complex non-linear fluid dynamics models for such eruptions, and develop visualization of the results of the simulations for science, hazards planning, and public awareness. The focus focus on two volcanoes: Mount St. Helens and Mount Vesuvius, with the intention to model Mt. Rainier, and the collapse of the Twin Towers in the future.

For volcanic eruptions, the code ROCFLU has been validated against a laboratory analog experiment, and preliminary two- and three-dimensional simulations for volcanic scales are presented. The proposed work would extend the simulation to multiple gas phases, particle loading of the volcanic gas phase, and fully-three dimensional applications. Results will be published in peer-reviewed literature, and using sophisticated animations. Potential funding agencies for future work are the National Science Foundation, NSF, the Army Research Lab, the Federal Emergency Management Agency, and Homeland Security.