[Statistical Characteristics of Urban Changzhou PM_2.5 Based on k-means Analysis]

Statistical analysis methods was utilized to investigate the variations of monthly average concentrations of the six basic pollutants (SO₂, NO₂, CO, O₃, PM_2.5 and PM₁₀) of six national standard monitoring sites from 2013 to 2014 in urban Changzhou. The results showed that, except for O₃, SO₂, NO₂, CO, PM_2.5 and PM₁₀ concentrations were all high in winter and low in summer. The relationship between particulate matter and wind speed showed, with increasing wind speed, the concentration of PM_2.5 reduced. However, the concentration variations of PM₁₀ were complicated and when wind speed increased, its concentration started to go down and then elevated. Fast-cluster analysis (k-means) and the index of SWV & DIV were used to classify the six basic pollutants into four clusters, and then the relationship between gaseous pollutants and PM_2.5in each cluster was emphatically discussed by statistical analysis method. Four clusters were assigned to fossil fuel combustion emissions (cluster1), O₃ and secondary aerosols (cluster2), incomplete combustion emissions and regional haze (cluster3), urban city "background" (cluster4). Incomplete combustion cluster accounted for the smallest percentage of urban Changzhou and city "background" was cluster of urban Changzhou with the largest contribution. k-means analysis results also showed that PM_2.5 had complex sources in urban Changzhou.