Dead Time, Pile-up and Counting Statistics

An overview is presented on recent progress in the field of nuclear counting statistics; theoretical expressions are shown to predict deviations from Poisson statistics due to non-random count loss in the spectrometer set-up. Frequently encountered misconceptions in the literature and in daily practice are uncovered: the unconditional belief in the general validity of Poisson statistics, the neglect of the dependency of counting statistics on the considered fraction of the pulse spectrum, the mix-up between pulse pile-up and extending dead time, the unawareness of the influence of pile-up rejection on the counting statistics in fixed live-time measurements and also with ‘loss-free counting’. Insight is provided into the statistical properties of spectra taken with ‘loss-free counting’ and ‘zero dead time’ counting, as well as the ‘variance spectrum’ provided with the latter. Uncertainty formulas are also presented for more conventional nuclear spectrometry measurements, with different types of count loss.

POMME Stefaan; 

2006-11-30

American Chemical Society

JRC33793

0841239827,   

https://publications.jrc.ec.europa.eu/repository/handle/JRC33793,