Automatic plume episode identification and cloud shine reconstruction method for ambient gamma dose rates during nuclear accidents

Publication date: November 2017
Source:Journal of Environmental Radioactivity, Volumes 178–179
Author(s): Xiaole Zhang, Wolfgang Raskob, Claudia Landman, Dmytro Trybushnyi, Christoph Haller, Hongyong Yuan
Ambient gamma dose rate (GDR) is the primary observation quantity for nuclear emergency management due to its high acquisition frequency and dense spatial deployment. However, ambient GDR is the sum of both cloud and ground shine, which hinders its effective utilization. In this study, an automatic method is proposed to identify the radioactive plume passage and to separate the cloud and ground shine in the total GDR. The new method is evaluated against a synthetic GDR dataset generated by JRODOS (Real Time On-line Decision Support) System and compared with another method (Hirayama, H. et al., 2014. Estimation of I-131 concentration using time history of pulse height distribution at monitoring post and detector response for radionuclide in plume. Transactions of the Atomic Energy Society of Japan 13:119–126, in Japanese (with English abstract)). The reconstructed cloud shine agrees well with the actual values for the whole synthetic dataset (1451 data points), with a very small absolute fractional bias (FB = 0.02) and normalized mean square error (NMSE = 2.04) as well as a large correlation coefficient (r = 0.95). The new method significantly outperforms the existing one (more than 95% reduction of FB and NMSE, and 61% improvement of the correlation coefficient), mainly due to the modification for high deposition events. The code of the proposed methodology and all the test data are available for academic and non-commercial use. The new approach with the detailed interpretation of the in-situ environment data will help improving the ability of off-site source term inverse estimation for nuclear accidents.



http://ift.tt/2v2Bi9e