Counting Statistics and Ion Interval Density in AMS

Abstract

Confidence in the precisions of accelerator mass spectrometry (AMS) and decay measurements must be comparable for the application of the radiocarbon calibration to age determinations using both technologies. We confirmed the random nature of the temporal distribution of 14C ions in an AMS spectrometer for a number of sample counting rates and properties of the sputtering process. The temporal distribution of ion counts was also measured to confirm the applicability of traditional counting statistics.

Calculation of the compounded uncertainty of 14C AMS measurements

“…Uncertainty in the raw 14 C/ 12 C, or 14 C/ 13 C ratio The 14 C/ 12 C, or 14 C/ 13 C, ratio, is the result of two or three measurements, the 14 C counts over a certain amount of time and the current of the stable isotope (s) used. The ''internal uncertainty'' of the 14 [3] have demonstrated that AMS measurements, even at very low counting rates, follow the Poisson distribution.…”

Anion formation by neutral resonant ionization

“…Only cylindrical recesses were used in the model, and the more complex shapes of sputtered craters are not modeled. We have never observed variation in the measured isotope ratio from recessed samples as a function of sputter cratering, even for a 4 h sputtering that left a tiny carbon fleck surrounded by massive loss of the aluminum holder [32]. Fig.…”

An Information-Efficient Bayesian Model for AMS Data Analysis

“…The CAR model assumes Gaussianity for the individual isotope ratios. This has been measured to be the case for some AMS machines (Vogel et al 2004) and has a theoretical justification as a good approximation of the counting statistical uncertainty. The MB method assumes Gaussianity for the final result of the analysis.…”