An alternative conception of PM 10 concentration changes after short-term precipitation in urban environment

Abstract In the article, a linear model is presented which describes a reduction of PM 10 mass concentration in relation to the type of precipitation and water vapour content in the air. The model was built using covariance analysis. In studies of PM 10 concentration changes, the results of 247 observations were used, which were carried out in the urban area. Concentration changes were archived during short-term (30 min) convection and large-scale rainfalls. For the determination of PM 10 mass concentration, the reference method was used. To describe changes in PM 10 concentration in the air after precipitation, a series of linear models were created, in which the explanatory variables were: the level of PM 10 before precipitation, absolute humidity of the air and the precipitation type. The criteria for the best model selection were its linearity, zero mean residual value and high coefficient of determination. Additionally, normality and homoscedasticity of the model residuals were tested. The formulated linear model is a useful tool for predicting changes in the concentration of particulate matter following short-term precipitation in urban areas. There is a need to check the utility of models for other fractions of suspended particles.