Detection of Temporary Increases in Storm-induced Riverine Radiocaesium Discharge Using a NaI(Tl) Scintillation Detector

We developed a method for monitoring riverine radiocaesium discharge following heavy rainfall using a waterproof NaI(Tl) scintillation detector system. The system continuously obtained gamma-ray spectra; peak radiocaesium counting rates declined following rainfall because of the shielding effect from the increased water volume, but spiked just after heavy rainfall, likely because of radiocaesium associated with muddy water runoff. The ratio between the two major peaks for radiocaesium and the relationship between fluctuations in region-of-interest channels and temperature were consistent with previous studies, validating the measurements. Additionally, we proposed a procedure using a smoothed second-order derivative filter to detect significant changes into increasing trends in peak counting rates. This procedure is applicable to real-time monitoring and will be improved by adjusting smoothing parameters and integrating water-quality indicators such as turbidity.