REDUCTION OF BACKGROUND BY HIGHER ORDER STATISTICS WITH NMIS

Measurements that accumulate the rate of real coincidence between multiplets of detection events (groupings of arbitrary order, e.g., one event, two events, three events, etc.) can yield spurious results if background events arise from processes (e.g., spontaneous fission or neutron spallation) that themselves produce correlated multiplets. This is particularly true if this background varies significantly over time or from one location to another, as it often does in operating facilities, i.e., those not specifically designed to support experimental radiation measurements but that instead rely upon the support of precise radiation measurements for, e.g., NMC&A. In particular, both the quantity and location of radioactive material in weapons facilities changes frequently and unpredictably, and so the background due to the presence (or absence) of this material is completely out of the control of the radiation measurement analyst. Furthermore, numerous Nuclear Materials Identification System (NMIS) measurements have revealed that background often contains mutually correlated events even in the complete absence of material (e.g., 240Pu) with a significant spontaneous fission rate. The technique subsequently described removes the effects of such self-correlated background from active NMIS measurements. It could be adapted to other active radiation measurements.