Improved representation of particle size and solubility in model simulations of atmospheric dispersion and wet-deposition from Fukushima
Dacre, HF; Bedwell, P; Hertwig, D; et al.Leadbetter, SJ; Loizou, P; Webster, HN
Date: 20 February 2020
Journal of Environmental Radioactivity
Elsevier / International Union of Radioecology
Radionuclides released into the atmosphere following the Fukushima Dai-ichi Nuclear Power Plant (FDNPP) accident were detected by ground-based monitoring stations worldwide. The inter-continental dispersion of radionuclides provides a unique opportunity to evaluate the ability of atmospheric dispersion models to represent the processes ...
Radionuclides released into the atmosphere following the Fukushima Dai-ichi Nuclear Power Plant (FDNPP) accident were detected by ground-based monitoring stations worldwide. The inter-continental dispersion of radionuclides provides a unique opportunity to evaluate the ability of atmospheric dispersion models to represent the processes controlling their transport and deposition in the atmosphere. Co-located measurements of radioxenon (133Xe) and caesium (137Cs) concentrations enable individual physical processes (dispersion, dry and wet deposition) to be isolated. In this paper we focus on errors in the prediction of 137Cs attributed to the representation of particle size and solubility, in the process of modelling wet deposition. Simulations of 133Xe and 137Cs concentrations using the UK Met Office NAME (Numerical Atmospheric-dispersion Modelling Environment) model are compared with CTBTO (Comprehensive Nuclear-Test-Ban Treaty Organisation) surface station measurements. NAME predictions of 137Cs using a bulk wet deposition parameterisation (which does not account for particle size dependent scavenging or solubility) significantly underestimate observed 137Cs. When a binned wet deposition parameterisation is implemented (which accounts for particle size dependent scavenging) the correlations between modelled and observed air concentrations improve at all 9 of the Northern Hemisphere sites studied and the respective RMSLE (root-mean-square-log-error) decreases by a factor of 7 due to a decrease in the wet-deposition of Aitken and Accumulation mode particles. Finally, NAME simulations were performed in which insoluble submicron particles are represented. Representing insoluble particles in the NAME simulations improves the RMSLE at all sites further by a factor of 7. Thus NAME is able to predict 137Cs with good accuracy (within a factor of 10 of observed 137Cs values) at distances greater than 10,000 km from FDNPP only if insoluble submicron particles are considered in the description of the source. This result provides further evidence of the presence of insoluble Cs-rich microparticles in the release following the accident at FDNPP and suggests that these small particles travelled across the Pacific Ocean to the US and further across the North Atlantic Ocean towards Europe.
College of Engineering, Mathematics and Physical Sciences
Item views 0
Full item downloads 0
Except where otherwise noted, this item's licence is described as © 2020. This version is made available under the CC-BY-NC-ND 4.0 license: https://creativecommons.org/licenses/by-nc-nd/4.0/