A Multivariate Statistical Model for Indirect Bomb Damage Assessment

This paper describes a multivariate statistical model for obtaining the best estimate of the point of detonation of a long-range nuclear weapon. It extends previous work by permitting greater generality in the dimensional characterization of such problems. It also introduces a new concept of weapon-system error based upon covariance matrices of the model and exhibits simple algorithms for error computation based upon this concept. Specifically, the model provides a technique for combining the a priori knowledge of weapon-system accuracy with the post-burst reports of sensory devices deployed within measuring range of the target. Formulas are derived for several configurations and an example is given illustrating application of the model to Indirect Bomb Damage Assessment.