Analysis of long-range transport influences on urban PM₁₀ using two-stage atmospheric trajectory clusters

There is evidence that long-range transport of natural and/or anthropogenic particles can have a negative impact on urban air quality. Certain European cities may fail to comply with the currently implemented 24-h PM₁₀ limit value due to the additional impact of remote sources of particles, which are not controllable at local level. For that reason, reliable methodologies identifying long-range transport patterns and quantifying their contribution to urban PM₁₀ levels are required. In this study, a two-stage clustering methodology was developed and applied to back trajectories arriving in three European cities: Athens, Madrid and Birmingham, which experience large, moderate and small numbers of daily PM₁₀ episodes, respectively. The atmospheric trajectories used in this analysis were computed with HYSPLIT model (NOAA) for a 3-year period. The two-stage cluster analysis has the advantage of providing highly disaggregated trajectory clusters, which proved to correspond to significantly different PM₁₀ levels. In the case of Madrid and Birmingham, clusters of trajectories originating from North Africa and continental Europe, respectively, were strongly associated with the highest PM₁₀ levels recorded in selected monitoring stations. In Athens, long-range transport patterns and trans-boundary influences were less evident, which highlighted the more substantial impact of local emission sources. Finally, two simple statistical indexes were proposed for assessing PM₁₀ episodes associated with different clusters of trajectories. This study has established a practical methodology for examining the impact of long-range atmospheric transport on local air quality, without resorting to high resource-demanding source apportionment techniques.