Abstract
The London Marylebone Road monitoring site is a roadside supersite adjacent to a highway carrying 80-90,000 vehicles per day on six lanes in a street canyon. Data from the Automatic Urban and Rural Network (AURN), Black Carbon Network, Automatic Hydrocarbon Network, Heavy Metal Network, Particle Size and Number and Particle Composition Network covering the period from 2009 to 2018 were analysed to determine short-term (diurnal, weekly and seasonal) and long-term variations and geographic source attribution. The contribution of roadside emissions relative to background sites (roadside increment) to the pollution climate was also investigated. The long-term trend analysis shows significant decline in regulated pollutants such as the particulate matter fractions (4.0%; 3.93%/yr for PM2.5 and PM10 respectively) and gas phase pollutants associated with vehicular emissions (5.5%; 1.22% and 2.1%/yr for CO, NOx and NO2 respectively), although concentrations of SO2 and O-3 have remained relatively constant over the years. Equivalent Black carbon (eBC) and total particle number count have also declined over the years whereas the heavy metals show mixed results (only Cu, Ni and V shows significant downward trends). The inorganic ionic components of the PM10 fraction, elemental (EC) and organic carbon (OC) and the volatile organic compounds all generally show declining trends over the period. Assessment of the diurnal variations shows elevated concentrations of the particulate matter fractions, the nitrogen oxides, CO and SO2 at periods corresponding to the traffic rush hours, whereas O-3 peaked in the afternoon when there is less titration due to lower NO concentrations. The diurnal pattern of eBC and Total Particle Number Count are similar to NOx and show strong traffic influence. Cl, Mg, K and Na levels show no systematic pattern throughout, with their presence likely controlled by meteorological conditions, and Ca showing high concentrations in the afternoon because of resuspension of deposited dust resulting from turbulence created by vehicular movement, and eroded road-surface material. Ammonium and nitrate show their lowest concentrations during the day when the temperature is high, probably reflective of their semi-volatile nature, with sulphate producing a peak around mid-day. The VOCs, with the exception of ethane, give the bimodal peaks typical of traffic related emission in the diurnal plots and their pattern is more similar to CO than the other traffic emitted gaseous pollutants. Ethane is associated with leakages from gas supply pipes. The weekday plots show weekday (Monday - Friday) increases in traffic-related pollutants and a decline over the weekends due to lower traffic volumes, with the reverse observed for O-3. K and the marine aerosol components show relatively similar concentrations on all days of the week, while Ca, NH4+, NO3- and SO42- all show a weekday maximum and decline over the weekend. The pollutants show seasonal variations; O-3 shows a springtime maximum, with the traffic-emitted pollutants (NO, NOx, CO, EC, OC etc) giving a winter maximum due to increase in emission, lower mixing depth and poor dispersion. Particulate matter fractions and total particle count show lower concentrations in summertime reflective of the semi-volatile nature of some components.
Ca shows less seasonal variability, with marine aerosol components showing a maximum winter concentration driven by higher wind speed conditions. NH4+, NO3- and SO42- show lowest levels in summer and maximum springtime concentrations. The traffic-related VOCs show a summertime minimum and wintertime maximum, while isoprene shows increased concentrations during summertime. The street canyon circulation causes the sampling of North London air on northerly winds, but enhanced traffic pollution when winds have a southerly component. There is a sizeable roadside increment above the local background for both exhaust (particulate matter mass fractions, particle number and eBC) and non-exhaust emissions (heavy metals). However, roadside increments of inorganic species, which include Ca, marine aerosol components and the secondary particulate matter components, are not significant, indicating that they are mainly controlled by regional transport of the pollutants. Polar plots show strong local contributions for the regulated gas phase pollutants and the carbon components (EC and OC), with O-3 concentrations enhanced mainly from the northerly direction. The long-range contribution from mainland Europe to the particulate matter fractions is significant and occurs mainly as secondary aerosol. Ratios of OC/EC in particles have shown a steady increase due to a more rapid reduction of EC than OC.