The characteristics of transport and transformation of SO₂, NOx and O₃ in northeast Asia have been investigated by using a comprehensive regional air quality model (RAQM) driven by a meteorological model MM5. A study period of 1-15 March 2002 has been selected due to the availability of intense observation of chemical species for both ground and upper levels. Model results have been compared against observational data to provide insights into the strength and weakness of the model's ability and the evolutionary features of chemical species. Validation shows a good skill of this model system in reproducing most of the key features in long-range transport, but apparent bias still remains due to a series of uncertainties from either emission estimates, prescribed parameters, or inherent model limitations. In general, this model shows a better skill for SO₂ and O₃ than for NOx. Large discrepancy occurs between the observed and calculated NOx concentration at higher levels, with the model results being much lower. A series of sensitivity tests have been conducted to investigate the potential affecting factors and it is found that the inaccuracy or incompleteness in currently used emission inventories could be a most likely cause for such discrepancy. Long-range transport from Asian continent to the western Pacific is pronounced in springtime. Substantially high concentrations of SO₂ and NOx in or above the boundary layer (0.5~3.0 km) over the Yellow Sea suggest an important pathway for long-range transport in northeast Asia.

Recent developments in satellite remote sensing technologies resulted in the ability to observe major pollution events such as dust and smoke around the world on a daily basis. Satellite imagery can sometimes detect long-range transport episodes. In this paper, a high CO episode at remote GAW station, Mt. Waliguan, detected by MOPITT CO dataset during the end of April 2002, is described. CO concentrations above 600 hPa almost doubled on 27 April and CMDL surface sample measurements also observed this significant CO enhancement. Using NCEP data, satellite fire products data and backward trajectory model we suggest that this high CO episode of 27 April is not a local pollution event, but that it is due to long-range transport from active biomass burning and biofuel burning areas located in the border areas of Pakistan and India. The trajectory cluster analysis shows that the origins of 5-day backward trajectories, for air masses reaching Mt. Waliguan station, at all altitudes, mainly overlap with the fire spot locations detected by TRMM data and biofuel burning in India.