Abstract
Analysis of three separate urban data sets has identified significant weekend reductions of PM10, PM2.5 and PM2.5-10 mass concentrations compared to weekday values. Significant weekday-weekend differences were also found for elemental carbon in the fine size fraction, and calcium in the coarse size fraction. There was no significant weekday-weekend difference in the case of sea salt or compounds of nitrate or sulphate. Given the paucity of day of the week-specific activity data for sources other than road traffic, it has been assumed that non-traffic sources of particles are uniform throughout the week, allowing estimation of the traffic contribution to PM concentrations. While this approach may neglect a weekday-weekend difference in particulate matter from construction sources, the effect of such sources is believed to be small. Two methods (by calculation and regression) of deriving the non-vehicle contribution to the urban increment above the regional background were compared. Both resulted in values of 2.1-2.7 mu g m(-3) (with an estimated error of around 2 mu g m-3) for the non-vehicle urban increment in PM10 measured by TEOM (gravimetric equivalent) at some urban background sites in London. The annual average traffic contribution to the urban increment is estimated as 1.6-4.4 mu g m-3 for background sites in London, except Bloomsbury where the estimated traffic contribution is higher and that for non-traffic sources is lower. Use of the same method at the rural Harwell site leads to an estimate of 1.8 mu g m(-3) for the traffic contribution. (C) 2008 Elsevier Ltd. All rights reserved.