| 英文摘要 |
The purposes of this project are to setup the PM supersite in southern Taiwan, to compare the consistency among monitors to be used in the supersite, to measure the PM concentrations and chemical compositions in southern Taiwan, to evaluate the feasibility to analyze the particle compositions from the β–gauge filter, and to evaluate the impact of festival on the local air quality. The PM supersites in sourthern Taiwan include four monitoring stations with the core site at DaLiu, Kaohsiung and three auxiliary sites at ChoTao, Kaohsiung, ChanJen, Kaohsiung, and ChiuChou, Pingtung. The monitoring items include PM2.5 and PM10 mass concentrations, particulate sulfate, nitrate, and EC/OC concentrations, particle size distributions, precursors NH3, photochemical products H2O2, NOy and nephelometer. The intercomparisons showed that the correlation coefficient and relative absolute difference percentage between RP8400N and their average were 0.98 to 0.99 and 12 to 14% and those for RP8400S were 0.96 to 0.99 and 4 to 12%. Therefore, good agreements were found among the four RP8400s. The correlation coefficient between dichotomous and RP8400N and RP8400S were 0.88 and 0.94 and those for RP2000 were both 0.89. However, the relative absolute difference percentage between dichotomous and RP8400N and RP8400S were 67% and 22% and those for RP2000 were 74% and 24%. Among the three routine monitoring sites, the greatest PM concentration was at DaLiu and the major chemical species were all the same at three sites and they were sulfate, nitrate, and ammonium, implying the significant contributions from secondary aerosol. For the feasibility of analyzing the β–gauge filter, only sulfate, nitrate, and chloride showed statistically significant results; the results for other species were not statistically different from blanks. The effects of the traditional festival for DaChi MaZu on air quality were due to the increase in mobile sources and the PM concentrations could be twice greater and those for CO and NO concentrations were
|