Abstract
Parallel observations of the PM2.5 were carried out for one and a half years starting from April, 2009. One is the reference methods defined by the Ministry of the Environment, Japan ( described as Run 2ref) and the other is a method using a different sampling duration, filter material and conditioning humidity ( described as Run 1 ). Two PM2.5 samplers, Partisol Plus 2025 ( Thermo Fischer Scientific ) equipped with a WINS impactor made in the United States were colocated. We evaluated 358 daily samples and 51 weekly samples of PM2.5 from September, 2009 to August, 2010. As the results of the comparison of the mass concentrations and water soluble inorganic ions, the mass concentrations almost corresponded to each other ( Run 1/Run 2ref is 0.97, s.d = 0.16, n = 51 ). The monthly concentration ratios ( Run 1/Run 2ref ) from September, 2009 to February, 2010 and April, 2010 were in the range of 1.00±0.05, On the other hand, those in the other months were 0.88 - 0.93. The chloride ions and nitrate ions, which had semi-volatile properties, reasonably increased in the winter and decreased in the summer. The Run 1/Run 2ref ratio of sulfate ions, which would be stable as ammonium sulfate, were unexpectedly in the range of 0.35 - 0.82. The adsorption of sulfur dioxides as a precursor of sulfate ions on the filters and particles obtained overnight were confirmed by additional experiments. The concentration differences of the total water soluble inorganic ions exceeded those of the mass concentrations in all the seasons. The difference in the filter materials and humidity while filtering caused this phenomenon.
