Comparison of five methodologies to apportion organic aerosol sources during a PM pollution event.

This study presents a comparison of five methodologies to apportion primary (POA) and secondary organic aerosol (SOA) sources from measurements performed in the Paris region (France) during a highly processed PM pollution event. POA fractions, estimated from EC-tracer method and positive matrix factorization (PMF) analyses, conducted on measurements from PM ₁₀ filters, aerosol chemical speciation monitor (ACSM) and offline aerosol mass spectrometry (AMS), were all comparable (2.2-3.7 μg m ^-3 as primary organic carbon (POC)). Associated relative uncertainties (measurement + model) on POC estimations ranged from 8 to 50%. The best apportionment of primary traffic OA was achieved using key markers (EC and 1-nitropyrene) in the chemical speciation-based PMF showing more pronounced rush-hour peaks and greater correlation with NO _x than other traffic related POC factors. All biomass burning-related factors were in good agreement, with a typical diel profile and a night-time increase linked to residential heating. If PMF applied to ACSM data showed good agreement with other PMF outputs corrected from dust-related factors (coarse PM), discrepancies were observed between individual POA factors (traffic, biomass burning) and directly comparable SOA factors and highly oxidized OA. Similar secondary organic carbon (SOC) concentrations (3.3 ± 0.1 μg m ^-3 ) were obtained from all approaches, except the SOA-tracer method (1.8 μg m ^-3 ). Associated uncertainties ranged from 14 to 52% with larger uncertainties obtained for PMF-chemical data, EC- and SOA-tracer methods. This latter significantly underestimated total SOA loadings, even including biomass burning SOA, due to missing SOA classes and precursors. None of the approaches was able to identify the formation mechanisms and/or precursors responsible for the highly oxidized SOA fraction associated with nitrate- and/or sulfate-rich aerosols (35% of OA). We recommend the use of a combination of different methodologies to apportion the POC/SOC concentrations/contributions to get the highest level of confidence in the estimates obtained.
Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
(Copyright © 2020 Elsevier B.V. All rights reserved.)