Determination of the source location of anthropogenic radionuclides collected in Finland and Sweden in June 2020 using a multi-technology analysis.

In June 2020, observations of anthropogenic radionuclides in Estonia, Finland, and Sweden that were not related to any acknowledged environmental release led to a comprehensive investigation on the source and cause of the unusual emissions. Several of the observed radionuclides were on the list of Comprehensive Nuclear-Test-Ban Treaty (CTBT) relevant radionuclides as an indicator of a potential nuclear test, and warranted detailed investigation. While analysis of aerosol samples coupled with Atmospheric Transport and Dispersion Modelling (ATDM) is a standard approach for environmental particulate releases, several new techniques were employed to better characterize the samples that allowed for useful inferences to be made. These inferences were crucial in forming the ultimate hypothesis for determining the facility type and location of the release.

Uncertainty and source term reconstruction with environmental air samples.

Environmental air sampling is one of the principal monitoring technologies employed for the verification of the Comprehensive Nuclear-Test-Ban Treaty (CTBT). By combining the analysis of environmental samples with Atmospheric Transport and Dispersion Modelling (ATDM), and using a Bayesian source reconstruction algorithm, an estimate of the release location, duration, and quantity can be computed. Bayesian source reconstruction uses an uncertainty distribution of the input parameters, or priors, in a statistical framework to produce posterior probability estimates of the event parameters. The quality of the event reconstruction directly depends on the accuracy of the prior uncertainty distribution. With many of the input parameters, the selection of the uncertainty distribution is not difficult. However, with environmental samples, there is one component of the uncertainty at the interface between sample measurements and the ATDM that has been overlooked. Typically, a much smaller volume or quantity of material is sampled from the much larger domain represented in the ATDM. By examining the response of a dense network of radionuclide detectors on the West Coast of Canada during the passage of the Fukushima debris plume, an initial estimate of this uncertainty was determined to be between 20% and 30% depending on sample integration time.