This project collected PM2.5 for chemical speciation at the air quality monitoring stations of Taiwan Environmental Protection Administration (TEPA) located in the northern, middle, and southern part of Taiwan from July 2015 to December 2016. Field work of seasonal collections were accomplished for six seasons with seven consecutive daily collections in each season and regular collections in 18 months with one daily collection in each month both synchronized at the three chosen stations. In addition, 32 mission oriented collections were completed including five in Taitung background area, 16 in Douliu highly polluted area, and 11 in Chiayi highly polluted area. On the whole, the seasonal PM2.5 mass concentrations varied consistently with that of TEPA stations indicating the results were representative in seasons in the study areas. Seasonal variations of PM2.5 mass concentrations across all stations showed that the lowest concentration occurring in summer with subsequent autumn elevation but alternating the roles of low and high among winter, spring, and autumn. Certain PM2.5 chemical components are known to vaporize, while quartz-fiber filter is subject to absorb gas interferences during sample collection. For the usage of a single filter in collecting PM2.5 water-soluble inorganic ions, underestimations of NH4+, NO3-, and Cl- were found ranging from 11%-37%, 21%-91%, and 41%-90%, respectively. In contrast, overestimations ranging from 0%-19% of organic carbon (OC) concentrations were resulted from collections without using the back-up second and third filters for interference corrections. The highest ranking of speciation percentages of PM2.5 across all stations and seasons are SO42- in summer, SO42- or OC in autumn, relatively divided in winter, and back to SO42- in spring. It is noted that NO3- is the highest ranking of PM2.5 speciation percentages at the southern Chiayi, Xiaogang, and Qianjin stations in winter. During the whole PM2.5 collection period, PM2.5 speciation percentage of NO3- was increased from 7% to 16% from the comparison between low (19 μg m-3, n=154) and high PM2.5 concentration samples (52 μg m-3, n=59) and that of the other components were reduced or unchanged. It implies that the sources discharging NOx, NO3- precursor, need to be controlled stringently. Among analyzed concentrations of metal elements across all stations, Na and K are ranked the upper two. It implies that sea breeze, dust suspension, and biomass burning are important sources. The computed values of enrichment factor (EF) across all stations show that contributions from crustal suspended dusts, coal burning, and traffic emissions are major emitting sources. In analyzing all pollution events with PM2.5 concentrations greater than 35 μg m-3, most were found caused by regional transport, bad dispersion, and photochemical reactions. Summarizing positive matrix factorization(PMF) modeling on TEPA monitoring stations in the northern, middle, and southern part of Taiwan, “secondary aerosol” was found the most important pollution category. PM2.5 speciation percentages of “secondary aerosol” at Banqiao, Zhongming, and Xiaogang stations were 30.7%, 28.6%, and 32.3%, respectively. In multiple regression analyses for ambient visibility in the stations of northern, middle, and southern part of Taiwan, NH4+ representing PM2.5 ammonium nitrate and ammonium sulfate is the major independent variable in all three study areas. Meanwhile, relative humidity and ambient temperature of meteorological factors play a role in reducing and increasing ambient visibility, respectively. The “secondary aerosol” and “fossil-fuel boilers and secondary sulfate” become major independent variables when PM2.5 chemical components are replaced by PMF factors in visibility regression modeling. In considering NO3- enhancement in high concentraion events, active control on local sources in reducing precursor emissions will be an effective measure to achieve PM2.5 air quality standard. In the international sampling experience and newly developed techniques of PM2.5 chemical speciation, 30 papers were reviewed to classify into PM2.5 chemical speciation enhancement, PM2.5 chemical speciation monitoring, haze observation, and source apportionment of PM2.5 chemical speciation. For the archived data format of PM2.5 chemical speciation, suggestions are proposed to store basic PM2.5 chemical speciation in the first tier and put extra detected PM2.5 chemical components from other specific methods or computations made on the data of the first tier into the second tier. In addition, a technical guide for the collection and analysis of PM2.5 chemical speciation is also proposed for the establishment of standard method.

PM2.5 chemical speciation monitoring, spatial and temporal characteristics of PM2.5 chemical speciation, PM2.5 source apportionment and visibility influencing factors