Wavelength dependent light absorption as a cost effective, real-time surrogate for ambient concentrations of polycyclic aromatic hydrocarbons

Many monitoring stations used to assess ambient air concentrations of pollutants regulated by European air quality directives suffer from being expensive to establish and operate, and from their location being based on the results of macro-scale modelling exercises rather than measurement assessments in candidate locations. To address these issues for the monitoring of polycyclic aromatic hydrocarbons (PAHs), this study has used data from a combination of the ultraviolet and infrared channels of aethalometers (referred to as UV BC), operated as part of the UK Black Carbon Network, as a surrogate measurement. This has established a relationship between concentrations of the PAH regulated in Europe, benzo[a]pyrene (B[a]P), and the UV BC signal at locations where these measurements have been made together from 2008 to 2014. This relationship was observed to be non-linear. Relationships for individual site types were used to predict measured concentrations with, on average, 1.5% accuracy across all annual averages, and with only 1 in 36 of the predicted annual averages deviating from the measured annual average by more than the B[a]P data quality objective for uncertainty of 50% (at −65%, with the range excluding this value between + 38% and −37%). These relationships were then used to predict B[a]P concentrations at stations where UV BC measurement are made, but PAH measurements are not. This process produced results which reflected expectations based on knowledge of the pollution climate at these stations gained from the measurements of other air quality networks, or from nearby stations. The influence of domestic solid fuel heating was clear using this approach which highlighted Strabane in Northern Ireland as a station likely to be in excess of the air quality directive target value for B[a]P. [Display omitted] •Data from the UV and IR channels of aethalometers were investigated as a surrogate PAH measurement.•The relationship observed was non-linear but fairly consistent across site types in the UK.•These relationships predicted annual PAH concentrations with a 1.5% accuracy, on average.•Predictions of PAH concentrations were made at stations in the UK where there is no monitoring currently.•This technique provides a cheap and quick method to optimise the siting of PAH monitoring stations.