In recent years, several events have occurred in which radionuclides were detected by radiological filter stations in a number of European countries without any knowledge of the origin of those radionuclides. In such cases, there is a need to locate potential release sites by inverse dispersion modelling techniques. However, if the release site is actually known, or if a potential release site has been localized by inverse methods, then there is an additional need to estimate the release rates from this location as a function of time for the various radionuclides detected.
Two kinds of radiological monitoring networks exist, viz. filter and gamma stations. Filter stations detect radionuclide-specific time-average concentrations with high accuracy and low thresholds, however with long averaging periods (e.g. between one and seven days). Gamma stations produce time-average radiation doses in which the radionuclides are lumped together, and with lower accuracy and higher thresholds, however with short averaging periods (down to 10 minutes).
While in the SLIM NKS project, methodologies have been developed to localize an unknown source of radionuclides dispersed in the atmosphere and detected by a radiological monitoring network, the SOCHAOTIC project develops methodologies, suited for operational use, by which a characterization of the source, whose location is known, can be derived, i.e. to estimate the temporal release profiles of the radionuclides detected.
For operational use, nuclear decision-support systems should be extended with modules handling and analysing such monitoring data automatically, e.g. by employing the European Radiological Data Exchange Platform (EURDEP), and conveying the data together with the geographical coordinates of the release point to the national meteorological centre accompanied by a request to estimate the temporal evolution of the release rates.
In the first year of SOCHAOTIC (2021), the following results are obtained:
Case studies identified and selected, the ETEX-1 and the October 2017 case of Ru-106 in Europe. In addition, an artificial case is produced by running a dispersion model forward and calculating average concentrations at filter stations and gamma dose rates at nearby gamma stations.
Methods for estimation of the temporal release profiles are developed, implemented and described.
Deterministic numerical weather prediction (NWP) model data are derived from the European Centre for Medium-Range Weather Forecasts (ECMWF) as well as the non-hydrostatic high-resolution Harmonie model, corresponding to the selected cases.
Quality-controlled measurement data of ground-level concentration are obtained.
The methods for source term characterization are applied by using the DERMA, MATCH, SILAM and SNAP atmospheric dispersion models to the selected cases.
Results are intercompared.
In the second year of SOCHAOTIC, the following results will be obtained:
A meteorological ensemble prediction system (EPS) will be set up and run by DMI in hindcast mode on sufficiently large geographical domains for the selected cases. Resulting meteorological data will be distributed to FMI, MET Norway and SMHI.
The inverse methods developed in the first year of SOCHAOTIC will be applied by using DERMA, MATCH, SILAM and SNAP for each NWP model ensemble member.
Inverse methods will be applied to a set of likely source locations based on methods from the SLIM project and the first phase of SOCHAOTIC.
Ensemble statistical methods will be developed and applied.
The results will be studied, discussed and intercompared.
Added value of using an ensemble approach will be investigated and described.
An interface to the ARGOS nuclear DSS will be described.
The results of SOCHAOTIC will be reported as an NKS report and published in a peer-reviewed scientific journal.