The correlations of the toxic air pollutants for all three stations were indicated by their correlation coefficients, as shown in Table 4.3a. High correlation coefficients implied that the same sources of toxic air pollutants are responsible for the readings at all the stations. The source(s) could be a single source or mixed sources that appeared at the same time.

The majority of correlation coefficients for the selected VOCs at the Central/Western station were higher than those for the Tsuen Wan station. The poor correlation for VOCs at Tsuen Wan could be explained by the mixed sources of VOCs, like traffic and industrial activities for instance. In addition to traffic emissions, industrial pollution emissions also form a serious source.

At the Central/Western station, the selected VOCs were mainly from vehicle emissions. At the PolyU station, all of the selected VOCs were highly correlated (r=0.64 to 0.95), with the exception of 1,3-butadiene. This suggests that at the PolyU station, with the exception of 1,3-butadiene, the other four VOCs come mainly from the same sources. The concentrations of 1,3-butadiene detected at the PolyU station were low, therefore the uncertainties or variations were comparably large. The correlation between acetaldehyde and formaldehyde (the Carbonyls) at all three monitoring stations were high(r=0.53 to 0.82), indicating that they came mainly from the same emission sources at all the stations. There was also a strong correlation between VOCs (except methyl chloride) and carbonyls at PolyU station, indicating that they too might originate from the same emission sources.

However, there was a poor correlation between carbonyls and PAHs at PolyU station. The poor correlation could be explained by the mixed sources of PAHs that did not appear at the same time or some sources of PAHs emitted sporadically or only at specific times, such as emissions from kitchens. Levels of PAHs at the Central/Western and Tsuen Wan stations showed strong correlations but poor correlations were generally found at the PolyU station, except for the correlation between benzo(a)pyrene and chrysene (r=0.66). That indicates that other than on-road vehicles, there were other sources of PAHs surrounding the PolyU station, especially for anthracene and chrysene. But for the other two stations, PAHs might have come from similar source types. Benzene correlated well with PAHs at the Central/Western and Tsuen Wan stations, but not at the PolyU station. Again, the pollutant sources for benzene and PAHs were different at the PolyU station, but the sources for benzene and PAHs are likely to have been of the same types at the other two stations. At the Central/Western station, the main source of benzene and PAHs is on-road vehicles, while at the Tsuen Wan station, in addition to on-road vehicles, industrial pollutant emissions were dominant sources of benzene and PAHs.