Characterisation and quantification of the sources of PM10 during air pollution episodes in the UK

Data for concentrations of PM(10) and gaseous pollutants from sites in the UK Automatic Urban and Rural Network have been examined during periods of elevated concentrations of PM(10). The ratios of concentrations of PM(10) to those of the other pollutants were used to determine the most probable source of the additional particles. The hypothesis is that because the concentrations of PM(10) were divided by those of the other pollutants, the ratio should decrease when PM(10) and the other pollutants have a common source. Conversely, the ratio should increase when the sources are different. During episodes where road traffic was the most probable source of the additional particles, the ratios of concentrations of PM(10) to carbon monoxide and oxides of nitrogen did decrease, but the comparable ratios for sulphur dioxide and ozone increased. In contrast, during episodes known to have been caused by construction activity, all these ratios increased. This is taken to show that the basic hypothesis is valid. For prolonged episodes, it was possible to use data averaged over the total duration of the episode for the purposes of source identification. For sporadic construction, or other short-duration episodes, it was necessary to use time series data. The data have also been used to calculate the differences between hourly average concentrations of pollutants measured during episodes and long-term hourly average concentrations. These have been used to model the additional PM(10) during air pollution episodes associated with construction activities and road traffic emissions. This confirms the lack of relationship between PM(10) and other pollutants during construction works. During episodes arising from road traffic emissions, there was good agreement between measured and modelled additional concentrations of PM(10) when an appropriate factor, F, related to the contribution of road traffic emissions to PM(10) at different site types was applied. The values used were 0.2 (Suburban), 0.3 (Urban Background/Urban Centre), and 0.5 (Roadside), representing 20%, 30%, and 50% contributions from road traffic, respectively.