CASA: An Innovative Tool for Near-Real-Time Source Apportionment of Atmospheric Organic Aerosols

Since organic aerosols (OA) account for a significant fraction of PM worldwide, source apportionment is essential for effective air quality mitigation and policymaking. In the present study, we developed a novel method based on a chemical mass balance and an elastic net regressor (EN-CMB), using positive matrix factorization (PMF) as prior knowledge for near real-time source apportionment of OA. EN-CMB has been integrated into a so-called continuous aerosol source apportionment (CASA) software package, which has been evaluated against state-of-the-art rolling-PMF data at three contrasted urban sites in Europe. CASA exhibits very satisfactory performance for primary OA components, with, at all sites, R ² values of 0.87-0.97 and mean bias error (MBE) of between -0.15 and 0.14 μg/m³. Secondary OA (SOA) fractions showed similar R ² values (0.81-0.97), but slightly higher MBE values (ranging from -0.71 to 0.04 μg/m³), which can be related to the complex nature of SOA and is still acceptable regarding bulk trends. CASA allows near real-time operation and, as it is open source, represents a promising example of timely and efficient air pollution management with applications in real-time air quality monitoring. The next steps will enable community-driven initiatives to improve and expand the application of such open-source methodologies across diverse regions and emission sources.