Uncertainty quantification can seem abstract to the uninitiated. However, it is a concrete practice with concrete rules that are driven by concrete needs. To help give a reader a sense for this practice, we provide four examples of uncertainty quantification in action.

The first example is drawn from Dr. Balch's doctoral work: the quantification and propagation of uncertainty in a Mars atmosphere model. The second is drawn from Dr. Balch's brief stint as a research engineer contracting at Wright-Patterson: an analysis of legacy wind tunnel data in a validation context. The third is drawn from work done by Dr. Balch over the past four years at Alexandria Validation Consulting, LLC: the computation of collision risk exposure during a satellite conjunction. This project led to a revolution in how we represent the epistemic uncertainty resulting from a statistical inference. Fourthly and finally, we share a more modest project: two-sided exact confidence bounds for binomial inference that are tighter than the classic two-sided Clopper-Pearson bounds. This improved solution to a classic problem was over 50 years in the making, and it was our use of possibilistic representations of statistical inference that allowed us to bring this project home in a way that traditional statisticians could not.