Integrating Weather Patterns with PMF Modeling: Insights into PM2.5 Pollution Sources and Future Applications to Ozone

Air quality is significantly influenced by weather patterns, such as wind direction, wind speed, and atmospheric dispersion, which play a direct role in driving fine particulate matter (PM_2.5) concentrations. This study employed three online monitoring instruments to conduct assessments of the chemical species in PM_2.5 during the winter season of 2020–2021. By applying the positive matrix factorization (PMF) model, pollution sources and their contributions were identified within Taipei City. To further explore the relationship between meteorological conditions and pollution, clustering techniques were employed to classify weather patterns associated with PM_2.5 levels exceeding 25 µg/m³.

The analysis revealed three distinct high-concentration weather patterns, each linked to specific pollution sources: (1) Low wind speed and poor dispersion conditions were associated with elevated traffic-related emissions, peaking at 7.4 µg/m³; (2) Strong northeast monsoon patterns resulted in relatively lower PM_2.5 levels due to reduced pollutant accumulation in the basin; and (3) Northwest wind patterns were characterized by significant contributions from coal and fuel combustion, industrial sources, and secondary aerosols, with PM_2.5 concentrations reaching up to 56 µg/m³.

This study is the first to combine source apportionment results from individual receptor sites with nonparametric trajectory analysis (NTA). Under weak wind conditions, traffic-related pollutants were observed to accumulate predominantly south of the receptor site, with maximum concentrations of 14 µg/m³. In contrast, northwest wind patterns showed notable accumulation of pollutants from civil construction and metalwork northwest of the receptor site, with concentrations reaching 8 µg/m³.

These findings highlight the critical role of weather pattern classification in understanding PM_2.5 pollution sources, offering valuable guidance for policymakers to implement effective air quality controls. Building on this foundation, future research will adapt these methodologies to explore the pollution sources of ozone and its precursors. Specifically, the new approach will integrate data from multiple monitoring stations with NTA to achieve spatial mapping of pollution source contributions. This advancement aims to provide a more comprehensive understanding of the distribution and impact of ozone and its precursors, deepening insights into the complex interactions between meteorology, atmospheric chemistry, and air quality management.