Multi-scale variations and empirical modeling of particulate matter in response to meteorological drivers

Understanding the temporal dynamics of particulate matter (including total suspended particulates designated as TSP, inhalable coarse particles designated as PM10 and fine particulate matter designated as PM2.5) and their response to meteorological conditions is crucial for effective air quality management. This study investigates the multi-scale characteristics of particulate matter concentrations and develops empirical models linking them to meteorological factors (temperature, humidity, and wind speed) in a coal yard environment. Results show that particulate concentrations exhibit strong temporal variability, with higher levels observed in autumn and winter. Correlation analysis across hourly, daily, monthly, and seasonal scales reveals robust inter-pollutant consistency and increasing meteorological influence with broader time scales. Wind speed showed a pronounced negative correlation with PM2.5 at the seasonal level, while humidity positively correlated with PM2.5, highlighting their roles in dispersion and secondary formation processes. Empirical models constructed for each pollutant demonstrated excellent predictive performance, with Pearson correlation coefficients exceeding 0.92, low bias, and normally distributed residuals. Model parameter distributions aligned with atmospheric expectations: Temperature effects were nonlinear and skewed, and humidity responses remained stable and symmetric. The findings contribute to the understanding of meteorological–pollutant interactions and provide a reliable framework for high-resolution air quality forecasting and control strategies.