PNNL

Abstract

Large variation in time-dependent ambient gamma-ray radiation challenges the search for radiation sources. A common strategy to reduce the effects of background variation is to raise detection thresholds, but at the price of reduced detection sensitivity. We present simple algorithms that both reduce background variation and maintain trip-wire detection sensitivity with gamma-ray spectrometry. The best-performing algorithms focus on the spectral shape over several energy bins using Spectral Comparison Ratios and dynamically predict background with the Kalman Filter.