Long-term reductions in regional pollutants contributed to decreased PM2.5 concentrations in Taiwan from 2017 to 2023

Long-term monitoring of PM_2.5 mass concentrations and chemical composition provides essential insights into the temporal variation of air pollutants and serves as a basis for evaluating the effectiveness of emission control strategies. Although the Taiwan Ministry of Environment (MOE) has established a nationwide air quality monitoring network comprising over 75 stations, routine measurements of PM_2.5 chemical composition are not included. To address this gap, this study collected PM_2.5 samples every six days from 2017 to 2023 at six MOE stations—Hualien, Banqiao, Zhongming, Douliu, Chiayi, and Xiaogang—to analyze their chemical composition. The results indicate that sulfate (SO₄^2-), nitrate (NO₃^-), and organic carbon (OC) were the predominant components, jointly accounting for over 50% of the PM_2.5 mass. All three species exhibited decreasing trends across the six sites during the study period. In 2023, compared to 2017, SO₄² concentrations decreased by 1.09–2.06 μg m^-3 (20–42%), NO₃^- by 0.27–2.32 μg m^-3 (17–36%), and OC by 0.63–1.90 μg m^-3 (27–43%). Positive Matrix Factorization (PMF) analysis resolved six major source factors: “Regional pollution,” “Mixed secondary pollution,” “Mixed primary pollution,” “Oil combustion,” “Sea spray,” and “Suspended dust.” The “Regional pollution” and “Oil combustion” sources were strongly associated with transboundary pollution. Notably, the “Oil combustion” factor exhibited a marked decline starting in 2020, coinciding with the implementation of the International Maritime Organization's global sulfur cap (IMO 2020), which limited the sulfur content in marine fuels. Over the seven-year period, the contributions from “Regional pollution” and “Oil combustion” decreased by an average of 29% and 76% across the six stations, respectively. In contrast, “Mixed primary” and “Mixed secondary” pollution were more closely linked to local sources, particularly traffic emissions. The “Mixed primary pollution” factor showed a strong correlation with CO (r = 0.65), with correlation coefficients exceeding 0.7 at the Douliu , Xiaogang, and Zhongming stations, indicating a significant influence from primary traffic emissions. While “Mixed secondary pollution” decreased significantly by 40% on average across the six sites over the last seven years, “Mixed primary pollution” showed only a marginal decline of 5%. In summary, the overall decline in PM_2.5 concentrations from 2017 to 2023 can be attributed to reductions in regional pollution and secondary aerosols. However, the stagnation in reducing mixed primary pollution highlights a critical gap, suggesting that future control strategies must prioritize stricter regulations on primary pollutants, encompassing both traffic and industrial emissions.