Conference Agenda
| Session | ||||
WG2: Multisite and multitime source apportionment
| ||||
| Presentations | ||||
11:30am - 11:45am
TU2-1: 1 Quantifying non-exhaust emissions in London using a combined source apportionment and machine learning approach 1Imperial College London, UK; 2National Centre of Scientific Research “Demokritos”, Greece; 3Paul Scherrer Institute, Switzerland Traffic remains an important source of particulate matter with non-exhaust currently estimated to make up a greater proportion of vehicle emission by mass than exhaust. The new Euro 7 standard will include non-exhaust emissions; to assess its impact it is crucial to have a good understanding of current emissions. Here we utilise the high time resolution aerosol measurements to carry out source apportionment, which is combined with machine learning and NO2/CO2 dilution approaches to estimate non-exhaust emission factors. Preliminary results show that 7.9%, 2.0% and 2.3% of PM10 at the roadside is brake, tyre & road and train wear, respectively.
11:45am - 12:00pm
TU2-1: 2 Source-dependent absorption Ångstrom exponent in the Los Angeles Basin: Multi-time resolution factor analyses of ambient PM2.5 and aerosol optical absorption 1Institute of Environmental Assessment and Water Research (IDAEA-CSIC), Spain; 2Environmental Monitoring and Reporting Branch, Ontario Ministry of the Environment, Conservation and Parks, Toronto, Canada; 3Department of Public Health Sciences, University of Rochester School of Medicine and Dentistry, Rochester, NY, USA This study applies advanced receptor modeling using the multilinear engine (ME-2) within Positive Matrix Factorization (PMF) to apportion PM2.5 sources in the Los Angeles Basin. Unlike Aethalometer optical method, this approach extracts source-specific Absorption Ångström Exponents (AAE) without a priori assumptions. A comprehensive PM2.5 chemical dataset and multi-wavelength absorption coefficients were analyzed at urban (CELA) and suburban (RIVR) sites. Five-source factors were identified, with secondary sulfate and nitrate dominating PM2.5 mass. Source-dependent AAE values varied by location, ranging from 1.24 to 3.0, highlighting differences in emission sources and atmospheric processes between traffic-dominated urban areas and suburban environments.
12:00pm - 12:15pm
TU2-1: 3 Multi-time Positive Matrix Factorization approach for enhanced source apportionment of organic aerosols from aerosol mass spectrometry and molecular speciation in two urban environments (Lyon and Bordeaux, France) 1Ineris, Verneuil en Halatte, 60550, France; 2IMT Nord Europe, Centre for Energy and Environment, Lille, 59000, France; 3LCSQA, 60550 Verneuil-en-Halatte, France; 4IGE, Univ Grenoble Alpes, Grenoble, 38400, France; 5Atmo Nouvelle-Aquitaine, Limoges, 87280, France; 6Atmo Auvergne-Rhone-Alpes, Bron, 69500, France To better characterize Organic aerosol (OA) sources, online measurements from the Aerosol Chemical Speciation Monitor and organic tracer analyses were carried out in 2019 at two urban background sites in France (Lyon and Bordeaux) and combined using the novel “multi-time” Positive Matrix Factorization (PMF) approach. This analysis separated secondary sources from biogenic and anthropogenic emissions. It also highlighted local sources, identifying cooking in Lyon and a marine source of OA in Talence, and provided a more accurate contribution from primary sources, which are underestimated in individual online PMF.
12:15pm - 12:30pm
TU2-1: 4 A source apportionment methodology joining multi-time resolution and size-segregated datasets for a better understanding of aerosol sources 1Università degli Studi di Milano & INFN-Milan, Italy; 2Università degli Studi di Milano, Italy; 3Università di Roma La Sapienza, Italy; 4Institute for a Sustainable Environment, Clarkson University & University of Rochester School of Medicine and Dentistry, USA; 5C.N.R. Institute of Atmospheric Pollution Research, Italy We developed a completely novel multi-time and multi-size resolution PMF (MTMS-PMF) implemented in a script for the Multilinear Engine ME-2 program. This cutting-edge model is an expansion of the PMF and allows the analysis of data measured at different time resolutions in multiple size classes. Moreover, both size-segregated data and PMX data can be inserted at the same time in the model. As output, the MTMS-PMF provides size-segregated chemical profiles and factor temporal contributions retrieved at the highest temporal resolution available in the dataset. The MTMS-PMF was successfully tested on a large dataset collected in the Po Valley (Ferrara, Italy).
| ||||