Comparison of simulated radioactive atmospheric releases to citizen science observations for the Fukushima nuclear accident

• Published in: Atmospheric environment (1994) 2019-02, Vol.198, p.478-488
• Main Authors: Hultquist, Carolynne, Cervone, Guido
• Format: Article
• Language: English
• Subjects: Crowdsourced; Dispersion model; Fukushima; HYSPLIT; Radiation
• ISSN: 1352-2310; 1873-2844
• DOI: 10.1016/j.atmosenv.2018.10.018

Citizen science data from the Safecast project were shown to provide a reliable estimation of the spatial distribution of concentrations of elevated radiation levels around Fukushima when compared to government data. A comparison is presented between the HYSPLIT Lagrangian atmospheric transport and dispersion (T&D) model and a reflected Gaussian model to both government and Safecast contributed measurements. The advantage of contributed data with respect to the government data is that they are collected over a long period of time and have a larger spatial coverage. First, the Safecast contributed measurements are compared to aerial surveys completed by the Department of Energy (DOE) and the National Nuclear Security Administration (NNSA). Then the HYSPLIT T&D model is run to simulate the nuclear release using high resolution terrain and meteorological data. A Gaussian dispersion model is also run for comparison using meteorological data observed at the time of the accident. The results of both models and observed data are decay corrected to December 2016 in order to use a larger quantity of contributed measurements in the comparison. The comparison of areas of elevated radiation shows that the citizen science observations align with the prediction of models representing dynamic behavior of radionuclides dispersed in the environment. This paper shows that citizen science data can be used to validate and potentially better calibrate atmospheric T&D models. •Comparison of observations and models reveals new insights on spatial patterns.•Unique aspects support the need to use multiple sources and not ignore limitations.•Citizen science is particularly ideal for monitoring in populated areas of interest.•Citizen science can act as an early warning system and be integrated with models.