Title: Particulate air pollution sources in low-income megacities

Duration: 2022-2026
Funding Agency: Swiss National Science Foundation (SNSF)
Funding Scheme: Ambizione
Contact: Kaspar R. Dällenbach (
Grant info:

Read more: here

Particulate matter (PM) has severe impacts on climate and human health. PM’s chemical composition, in particular that of the organic aerosol (OA) fraction, is critical for understanding and predicting PM’s health effects. It is vitally important to characterize the chemical composition and the sources contributing to organic aerosol (OA), a major fraction of PM mass, in highly-polluted densely populated regions in order to alleviate its adverse health effects. OA can be directly emitted from primary emissions (POA) or secondary (SOA) formed through atmospheric processing of emitted volatile organic compounds. Unfortunately, there is a multitude of potential pollutant sources, their emissions are unidentified and their contribution to OA are unquantified. Therefore, quantifying the sources’ contribution to both POA and SOA as well as assessing the processes governing SOA formation is crucial for mitigating particulate air pollution in megacities. OA source apportionment has been limited by both measurement and data mining techniques. While a newly developed extractive-electrospray ionization ultra-high resolution mass spectrometer offers real-time near-molecular characterizations of OA with a great potential to resolve the sources of primary aerosols and secondary aerosol precursors, only very few studies have used the molecular data together with advanced data mining techniques for OA sources apportionment. So far positive matrix factorization (PMF) has been the most common data mining technique used for aerosol source apportionment. While developments focused on different ways to pre- and post-process data, PMF has not been improved since decades. When applied to molecular data, PMF resolved factors linked to SOA, which is often the most important contributor to OA. Some of these SOA factors could be related to sources (biogenic, residential heating), but other factors remained of unclear origin or were separated as a function of atmospheric age instead of formation pathway. Overall the identification of SOA sources and formation mechanisms remains challenging. Currently, the lack of new algorithms able to tap the full potential of recent near-molecular measurements remain a major bottleneck in air-quality research. Within this project, we aim to overcome the barriers of understanding the drivers of particulate air pollution. We will develop novel data mining tools for extracting the processes governing OA emission and formation based on near-molecular OA characterizations. Coupling the atmospheric observations with laboratory emission aging experiments of emissions from key sources will lead to an in-depth understanding of the chemical processes governing the formation of secondary organic particulate air pollution in polluted environments.