Patterns of PM10 particles change in the atmospheric air of Ivano-Frankivsk city

Abstract PM is one of the main atmospheric air pollutants causing serious concern for public health. New evidence on health effects of PM has led to the publication of new WHO air quality guidelines. Spatio-temporal patterns of PM10 change are a current topic in atmospheric and environmental sciences. An equally important aspect is the study of the relationship between PM10 concentration and meteorological characteristics. Temperature, humidity, wind and atmospheric pressure can affect the concentration, distribution and transport of PM10. The database for the study included measurements of PM10 concentration, air temperature, atmospheric pressure, and relative humidity for 306 days in 2019 (from March 1 to December 31) every hour, a total of 7,344 for each parameter from the Ecocity public monitoring station in the central part of Ivano-Frankivsk city. There are no significant stationary pollution sources and highways near the observation site. Certain regularities were found in the daily variation of PM10 concentration: sharp increase in concentrations at night with a peak around midnight, decrease during the day (minimum around 16:00). The proposed sinusoidal model fairly well reflects the distribution of data in the morning hours, from 5 to 11 am. Significant fluctuations in daily concentrations (from 0 to 100 μ g/m 3 ) were observed, especially in the cold period of the year. The highest average monthly values of PM10 are in March, the lowest are in July. The current average daily limit value in Ukraine and the EU is 50 μ g/m 3 (no more than 35 exceedances per year), 442 instantaneous indicators above the norm. When analyzing the relationship between PM10 concentrations and air temperature, the regularity is observed both during the day and throughout the year (in cold periods, the concentration of particles is higher, which can be explained by lower air temperatures). The patterns identified are related to a number of factors, such as meteorological conditions, pollution sources, and atmospheric chemistry. Multivariate linear regression was used to find the mathematical relationship between the average monthly values of PM10 concentration, humidity and air temperature. The coefficient of determination R 2 = 0.67, Fisher’s test F = 70.95 and the mean square error MSE = 23.12. On the basis of this, it can be stated that the model is acceptable for estimating and forecasting temporal patterns of the concentration of PM10 particles in the atmospheric air of Ivano-Frankivsk, depending on the average monthly indicators of humidity and atmospheric air temperature.