1
|
Srivastava D, Daellenbach KR, Zhang Y, Bonnaire N, Chazeau B, Perraudin E, Gros V, Lucarelli F, Villenave E, Prévôt ASH, El Haddad I, Favez O, Albinet A. Comparison of five methodologies to apportion organic aerosol sources during a PM pollution event. Sci Total Environ 2021; 757:143168. [PMID: 33143914 DOI: 10.1016/j.scitotenv.2020.143168] [Citation(s) in RCA: 3] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Subscribe] [Scholar Register] [Received: 08/08/2020] [Revised: 10/12/2020] [Accepted: 10/14/2020] [Indexed: 06/11/2023]
Abstract
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 PM10 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 NOx 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.
Collapse
Affiliation(s)
- D Srivastava
- Ineris, Parc Technologique Alata, BP 2, 60550 Verneuil-en-Halatte, France; CNRS, EPOC, UMR 5805 CNRS, 33405 Talence, France; Université de Bordeaux, EPOC, UMR 5805 CNRS, 33405 Talence, France.
| | | | - Y Zhang
- Ineris, Parc Technologique Alata, BP 2, 60550 Verneuil-en-Halatte, France
| | - N Bonnaire
- LSCE - UMR8212, CNRS-CEA-UVSQ, 91191 Gif-sur-Yvette, France
| | - B Chazeau
- Paul Scherrer Institute (PSI), 5232 Villigen, Switzerland
| | - E Perraudin
- CNRS, EPOC, UMR 5805 CNRS, 33405 Talence, France; Université de Bordeaux, EPOC, UMR 5805 CNRS, 33405 Talence, France
| | - V Gros
- LSCE - UMR8212, CNRS-CEA-UVSQ, 91191 Gif-sur-Yvette, France
| | - F Lucarelli
- University of Florence, Dipartimento di Fisica Astronomia, 50019, Sesto Fiorentino, Italy
| | - E Villenave
- CNRS, EPOC, UMR 5805 CNRS, 33405 Talence, France; Université de Bordeaux, EPOC, UMR 5805 CNRS, 33405 Talence, France
| | - A S H Prévôt
- Paul Scherrer Institute (PSI), 5232 Villigen, Switzerland
| | - I El Haddad
- Paul Scherrer Institute (PSI), 5232 Villigen, Switzerland
| | - O Favez
- Ineris, Parc Technologique Alata, BP 2, 60550 Verneuil-en-Halatte, France
| | - A Albinet
- Ineris, Parc Technologique Alata, BP 2, 60550 Verneuil-en-Halatte, France.
| |
Collapse
|