Incorporating Uncertainty in Uxo Discrimination

We compare two representations of feature vector uncertainty in the context of unexploded ordnance (UXO) discrimination: a local uncertainty estimate derived from the curvature of the misfit function, and a global estimate of model uncertainty which uses multiple feature vectors to represent the range of converged models obtained in the inversion process. We show that this latter representation can find outlying feature vectors estimated from time-domain electromagnetic data.