Data Analysis in Support of Radiation Portal Monitoring

Passive gamma and neutron detectors screen for illicit special nuclear material in vehicles crossing borders between countries. This paper illustrates how statistical analysis of archived detector data can help to evaluate special nuclear material detection probabilities and to investigate several issues, including drifting background, background gamma suppression, nuisance alarms arising from either naturally occurring radioactive material (NORM) or cosmic ray bursts, and detector quality control. Statistical techniques described include data smoothing, cosmic ray filtering of neutron alarms, quantile estimation, and pattern recognition. New data analysis of gamma energy spectra suggests that NORM recognition using deployed detectors is difficult.