A machine learning model for predicting PM_2.5 and nitrate concentrations based on long-term water-soluble inorganic...

In this study, long-term variations in the concentrations of PM_2.5, water-soluble inorganic salts (WIS), and gaseous precursors measured by a roadside air quality monitoring station were investigated from 2017 to February 2021 to examine the formation mechanism of secondary inorganic PM_2.5. A new machine learning model using WIS data as input variables was further developed to predict PM_2.5 and nitrate concentrations for source tracing and effective control strategy development. The results showed that a reduction in the NO_x concentration under VOC-limited O₃ formation regime could offset the consumption of OH and O₃, causing an increase in secondary NO₃^- and PM_2.5 formation during fall and winter seasons. A good agreement was obtained between the predicted and measured PM_2.5 values, with R², root mean square error (RMSE), and mean absolute error (MAE) values of 0.81, 6.81 μg/m³, and 5.10 μg/m³, respectively. The nitrate ([NO₃^-]) prediction model could predict ∼59% of the atmospheric nitrate concentration. The sensitivity analysis of the input variables in the present model further revealed that NO₃^- and VOC were two important pollutants dominating the variation trend of PM_2.5. It is recommended that decision makers should focus more on the reduction of VOC and O₃ to reduce secondary PM_2.5 formation during winter in central Taiwan. Real-time measurements of the chemical composition of PM_2.5, taken as the regulatory air quality monitoring items are needed in the future.