Analysis of a Gross Counting Decision Metric for use in Threat Detection During Cargo Container Inspection

LLNL is actively engaged in the development of a variety of advanced technologies for use in detecting potential threats in sea-going cargo containers, particularly the presence of hidden special nuclear materials (SNM). One such project is the so-called ''Nuclear Car Wash'' (NCW), which uses a high-energy neutron probe to scan the container. High-energy, β-delayed γ-rays emitted during the decay of short-lived, neutron-induced fission products are then taken as a signature of fissionable material. There are a number of different threat decision metrics that one could imagine using in conjunction with an inspection system such as the NCW; however, the most straightforward approach might be to simply compare the total number of counts that our detector records during some suitably chosen time interval to the average background signal that one would expect from a ''clean'' container during the same interval. The purpose of this report is to describe the basic statistical properties of a decision metric of this sort and outline the procedures for using it in experimental practice