This study collected PM2.5 (aerodynamic diameters equal to or smaller than 2.5 μm) regularly every six days per shift at the Banqiao, Zhongming, Douliu, Chiayi, Xiaogang, and Hualien air quality monitoring sites of Environmental Protection Administration (EPA) in Taiwan in 2020. Besides, PM2.5 was collected near traffic sources at the Taiwan Boulevard site for finding vehicle emissions characteristics. PM2.5 mass, water-soluble inorganic ions, carbonaceous contents, and metal elements were analyzed for the regular and near-source samples. Meanwhile, phthalate esters were also analyzed for the regular samples. This study investigated and derived the characteristics of temporal and spatial distributions of PM2.5 mass and chemical components (including the data of December 2019), high concentration events, specific scenarios, the apportioned source factors, and atmospheric visibility-influencing factors by utilizing the analyzed PM2.5 chemical components, air quality monitoring data, environmental factors, and the corresponding model simulation products.
The PM2.5 mass concentration of this study is consistent with that of the EPA's regular manual collection. Seasonal PM2.5 mass levels increased from east, north, to the south of Taiwan during winter, spring, and autumn. Organic carbon (OC) was the most abundant component at the Hualien and Banqiao sites and NO3- for the other sites in winter. In contrast, OC was the most abundant component at most sites in spring, summer, and autumn. SO42- lost the role as the leading component in 2020, which reflected an effective control on the combustion sources of sulfur-containing fossil fuel. The mutual high-abundance metal elements (Na and K) across sites increased in concentrations from east, north, to the south of Taiwan. The distinct metal elements exhibited source characteristics among sites. The volatilization corrections compensated 12~21% and 17~38% of NH4+ and NO3- mass losses, respectively, in 2020. The error distribution of OC concentration will spread from 89% overestimation to 23% underestimation. Similarly, positive interferences also showed no spatial distinction. Ambient levels of phthalate esters were very low, with relatively high values for DEHP, DBP, and DnOP at each site. For near-source traffic emissions at the Taiwan Boulevard site, PM2.5 mass, NO3-, OC, and EC concentrations were higher during on- than off-duty rush hours. OC representing gasoline vehicle emissions was the most abundant component during on- and off-duty rush hours. The high PM2.5 concentration days (≥ 30 μg m-3), caused by local pollution and predominantly with “residual layer” in the preceding night, were mainly distributed south to central Taiwan in winter 2020. By contrast, the number of high PM2.5 concentration days declined considerably in spring. In August 2020, the PM2.5 concentration at the Hualien site increased more than twice with the abundance of SO42-, NH4+, and crustal elements and a 10-fold increase of rare-earth element Yttrium was the most stunning discovery under the influence of the long-range transport of the erupted Nishinoshima of Japan.
By classifying PM2.5 mass concentrations of sampling days of the recent four years into high, medium, and low concentration groups, PM2.5 reduction benefits are evident for the medium concentration group (15~35 μg m-3) when NO3- precursor emission sources are in control. The results from Community Multiscale Air Quality Model (CMAQ) discovered that daytime photochemical reaction and nighttime N2O5 heterogeneous hydrolysis were the mechanism of NO3- enhancement in high PM2.5 concentration events. The stable declination of SO42- concentration during higher PM2.5 concentration period (from December to April) of the recent four years implied a reduction of fossil-fuel combustion activities from stationary sources. In contrast, reversibly increased EC concentration in 2020 and either reversibly increased or in the flat fashion of NO3- and OC concentrations in 2019 and 2020 imply that mobile sources will be the next target for improving the air quality of high PM2.5 concentration period. The samples collected after traditional festivals were rich in the tracer elements of firecracker and incense. Concentration enhancements of SO42-, NO3-, NH4+, K+, tracer elements of iron and steel, cement, and coal-burning were obvious when eastern and northern parts of Taiwan were under the influence of transboundary pollution transport in winter and spring. The richness of metal elements Ba, Pb, and Ga at the Chiayi site indicated the influences of coal-burning, iron and steel, biomass burning, and local folk activities. The estimate on secondary organic carbon indicates that a stringent control of volatile organic compounds in the area south to central Taiwan will help reduce secondary PM2.5 concentration. By comparing PM2.5 concentration during the COVID-19 pandemic period with that of the extension from the past three years, more PM2.5 reduction was found from the reduced industrial activities in the central and Yulin-Chiayi areas of Taiwan.
From source apportionment approach using positive matrix factorization (PMF), “Sulfate”, “Nitrate”, and “Vehicle emissions” were the three most significant factors in the nine resolved source factors across all sites, and the influence of vehicle emissions increased gradually at the sites north to the Douliu site. For estimating the atmospheric light extinction coefficient (bext), the results computed from the revised Interagency Monitoring of Protected Visual Environments (IMPROVE) equation revealed that sulfate and organic matter contributed to bext stably and nitrate contributed the most to the area south to the Douli site in winter and spring. Similarly, SO42-, NO3-, and OC were significant factors influencing ambient visibility from the statistical regression analysis. For the most updated sampling and measuring techniques for PM2.5 chemical components, this study reviewed 32 papers from studies in various countries in recent years.
In summary, this study consistently provides PM2.5 mass and chemical components (including phthalate esters) data from the six urban sites and near traffic-source site to derive temporal and spatial characteristics of PM2.5 in Taiwan. Additional derivations from the collected data include the characteristics of PM2.5 chemical components at specific time and sites, fractions of source factor contribution, and factors influencing atmospheric visibility. The results of this study will serve as the data cornerstone of health risk assessment and control measure delivery.
