Title :
The effect of counting statistics on the integrity of deconvolved gamma-ray spectra
Author :
Crossingham, Grant ; Dallimore, Matthew ; Ramsden, David
Author_Institution :
Symetrica Ltd., Southampton, UK
Abstract :
Symetrica´s spectrum-processing software improves both the spectral-resolution and sensitivity of scintillation spectrometers. This paper demonstrates the robustness of this algorithm when applied to spectra having poor statistical quality. The error in the location of the deconvolved 609keV full-energy peak, within a 226Ra spectrum; was less than 2keV for a peak area containing fewer than 150 counts. For the raw data, this accuracy was only achieved for peak areas greater than 1300 counts. Processing the spectra in this way also improves the accuracy, by factor of between 2 and 2.6, with which the full-energy peak areas could be measured. This paper has also shown that, by using the processed data, a hidden 137Cs source having a count rate at the level of just 2% that of a 60Co masking source, was always detectable in the limited number of spectrum recorded.
Keywords :
deconvolution; gamma-ray spectra; scintillation counters; counting statistics; deconvolved gamma-ray spectra integrity; scintillation spectrometers; Area measurement; Calibration; Energy measurement; Gas detectors; Position measurement; Radiation detectors; Robustness; Solid scintillation detectors; Spectroscopy; Statistics;
Conference_Titel :
Nuclear Science Symposium Conference Record, 2003 IEEE
Print_ISBN :
0-7803-8257-9
DOI :
10.1109/NSSMIC.2003.1352136