Atmospheric Environment 37 (2003) 219–230 Characteristics of PM 2.5 species and long-range transport of air masses at Taean background station, South Korea Z. He, Y.J. Kim*, K.O. Ogunjobi, C.S. Hong Advanced Environmental Monitoring Research Center (ADEMRC), Kwangju Institute of Science and Technology, 1 Oryong-dong, Puk-gu, Kwangju 500-712, South Korea Received 23 May 2002; accepted 4 October 2002 Abstract To investigate the characteristics of fine particle and establish source–receptor relationship between pollutant concentrations and long-range air masses transport, PM 2.5 were collected during two different sampling periods (every- 6-day samplings and intensive samplings) at Taean background station, South Korea. High correlation among PM 2.5 , SO 4 2 , NO 3  and NH 4 + suggested that SO 4 2 , NO 3  , and NH 4 + were primary ionic components of fine particles. The sum of SO 4 2 , NO 3  and NH 4 + accounted for 35%, 32% of PM 2.5 mass for every-6-day samplings and intensive samplings, respectively. Backward trajectories were used to interpret the possible air mass sources. Most of back-trajectories passed through northwestern quadrant, where one of the dominant dust-storm source regions, gobi desert is located. On average, 75% back-trajectory at 2000 m above ground level (agl) crossed through northwestern quadrant during the above two different sampling periods. Cluster memberships were also inspected in detail for better understanding of transport patterns during different sampling periods. The trajectories passed through northwestern China with low speed in winter and spring considered as ones that carried much aerosol particles to Taean. In addition, the airflow that arrived at Taean in winter and spring carried more pollutants than that in summer and autumn. r 2002 Elsevier Science Ltd. All rights reserved. Keywords: PM 2.5 ; Clustering; Cluster-mean trajectory; Correlation coefficient 1. Introduction Lagrangian trajectories, which were used to identify pathways of the air mass transport or establish source– receptor relationships of air pollutants, have enjoyed considerable popularity as an approach to determine the potential source regions of pollutants. Ulman and Saxena (1997) established a preliminary relationship between the chemical properties of the observed clouds as well as precursor aerosols and trace gases by identifying the history of the cloud forming air masses via back trajectories analysis. Brankov et al. (1998) examined the relationship between synoptic-scale atmo- spheric transport patterns using back-trajectories and pollutant concentration levels observed at three sites in USA. Most of these studies focused on the trajectory calculation techniques. Cluster analysis has only recently been applied to meteorological data after Moody and Galloway (1988) considered trajectory coordinates as the clustering variables. East Asia region is a major source of NO x and SO 2 (Kato and Akimoto, 1992), and probably other species as well (Jaffe et al., 1997). It is also projected that sulfur emissions will exceed the total emissions of North American and Europe in the year 2010 without future control of emissions (Foell and Green, 1991). At present, there are no international agreements in place to limit such growth. The mid-latitudes (25–401N) are most strongly affected by Asia dust (Uematsu et al., 1983; Prospero et al., 1989; Gao et al., 1992), which is a significant source for deep-sea sediments (Blank et al., 1985; Gao et al., 1992). Dust can be transported more than several thousands kilometers far, even from East AE International – Asia *Corresponding author. Fax: +82-62-970-2434. E-mail address: yjkim@kjist.ac.kr (Y.J. Kim). 1352-2310/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved. PII:S1352-2310(02)00834-8