Statistical Substantiation of the Revising of Readings by the CityAir Station of PM2.5 Concentration Levels in the Atmospheric Boundary Layer of the City

Abstract

As a marker characterizing air pollution in the surface layer of the atmosphere of modern cities, the concentration level of particulate matter with a diameter of 2.5 microns or less (Particulate Matter, PM2.5) is often used. The paper discusses the practice of using a relatively cheap optical sensor, which is part of the CityAir station, to measure the concentration of PM2.5 in an urban environment. The article proposes a statistically justified correction of the primary data obtained by CityAir stations on the values of the concentration of suspended particles PM2.5 in the surface layer of the atmosphere of Krasnoyarsk. For the construction of regression models, measurements obtained from E-BAM analyzers located at the same observation posts as the corrected sensors were considered as a reference. For the analysis, primary data was used 1) from 9 automated observation posts of the regional departmental information and analytical system of data on the state of the environment of the Krasnoyarsk Territory (KVIAS); 2) from the 21st CityAir station of the monitoring system of the Krasnoyarsk Scientific Center of the Siberian Branch of the Russian Academy of Sciences. The paper demonstrates that when correcting sensor readings, it is necessary to take into account meteorological indicators. In addition, it is shown that the regression coefficients significantly depend on the season. Supervised learning methods are compared for solving the problem of correcting the readings of inexpensive sensors. Additional information on the results of data analysis, which was not included in the text of the article, is available on the electronic resource https://asm.krasn.ru/.