Automation of the inverse calculation of the atmospheric transport model as a part of a system for analyzing unknown emission sources

Abstract

In this paper, software tools for automating the inverse calculation of the atmospheric transport model were developed as a part of the pilot version of the system for analyzing unknown sources of atmospheric pollution in the case of their detection by monitoring networks. By min-imizing the defined quality function, the probability of the source location at a certain geo-graphical point, its duration, and time of onset depending on the location, together with the vol-ume of the emission, can be analyzed. The source-receptor function is calculated by using the well-known atmospheric transport model FLEXPART in the inverse calculation mode. Auto-mated calculations of the atmospheric transport model are carried out on the Cloud Computing Platform of the Ukrainian National Grid Infrastructure by creating a virtual machine for the se-ries of FLEXPART calculations. In the future, it is planned to automate the creation and dele-tion of virtual machines performing calculations. Testing of the developed algorithms was car-ried out based on meteorological conditions during the wildfires in the Chornobyl Exclusion Zone in 2020 and data generated from measurements taken at one of the stations near Kyiv on April 18–19, 2020. During the test simulation, the coordinates of the source and the amount of Cs-137 emissions were considered unknown. Satisfactory results were obtained by comparing the estimated coordinates of the source and the volume of emissions with the corresponding real values. It is shown that the system can timely and sufficiently accurately analyze the most im-portant characteristics of possible unknown sources of atmospheric emissions. The results of the study confirm the potential importance of the obtained results for use in real-life situations and help in the identification and analysis of possible sources of radioactive contamination. The developed methods and algorithmic tools have no limitations regarding the geographical region of calculations and can be used both in the case of emissions in Ukraine and abroad.