 |
Center for Nonlinear Studies |
Many real-world applications of science and engineering revolve around a decision that must be made. These decisions are supported by science and observations, but the science and observations are often plagued by uncertainty. This situation, as well as potential pitfalls, will be illustrated by studying a typical decision-support scenario at a contaminated groundwater site. Subsequently, a historical perspective will be presented on the interplay between uncertainty, mathematics, science, and computation. Finally, a robust decision-support framework will be discussed that combines Bayes’ theorem with non-probabilistic methods. The non-probabilistic methods are employed to circumvent issues that arise in applying Bayes’ theorem to messy problems. An application to geologic CO2 sequestration will be considered.
Host: Jeffrey D. Hyman