Petascale computing for geosciences research

Authors


Abstract

A 2005 report describes the opportunities facilitated by, and the feasibility of, establishing a leadership-class computing facility dedicated to geosciences research. A forthcoming workshop will focus on implementation of the vision outlined in this report.

The geosciences research community has progressed significantly in its ability to monitor and model components of the Earth system, and now is poised to make significant breakthroughs in understanding the system as a whole. Computational simulation, along with theory and observation, has become established as a fundamental methodology for making progress in Earth system science. Increasingly, numerical simulations are not only tested by observations, but they also provide the first glimpses of new phenomena and quantitative characterization of complex processes. In turn, simulations inspire new theoretical investigations and observational strategies. For example, properties of complex minerals at temperatures and pressures inaccessible to laboratory measurements are beginning to be derived from first-principles-based computations with accuracies sufficient for direct input to Earth models.