Differences between PM2.5 chemical sources at semi-closed transport environments: comparing bus depots and metro platforms.

Chemical composition and source contributions of PM2.5 were investigated at a public-restricted bus depot in Valladolid (Spain) and two metro stations (Alto dos Moinhos and Quinta das Conchas) in Lisbon (Portugal) using positive matrix factorization (PMF). Locally derived sources dominated PM2.5 in both environments, with OC-rich combustion emissions prevailing at the bus depot and Fe-rich wear processes at the metro stations as major components. Trace element signatures (ng m-3) differed accordingly, with Zn dominating at the bus depot, reflecting tire wear and maintenance, and Mn prevailing at the metro stations, indicative of metro wear (e.g. rail-wheel-brake wear). Seven sources were resolved at the bus depot, led by mineral/construction (27.6%), followed by secondary sulfate (15.8%), tire wear (14.8%), exhaust (11.5%), washing/resuspension (11.4%), maintenance (9.9%), and brake wear (8.9%). However, both metro stations were dominated by metro wear emissions (∼45-46%), with additional contributions from sea spray (∼16-17.5%), secondary sulfate (8.4-16.5%), and mineral/construction (11.5-20.6%). Traffic-related infiltration was identified only at Alto dos Moinhos (8.1%), whereas biomass burning was resolved at Quinta das Conchas (9.7%). The different source profiles indicate the need for specific exposure-control strategies targeting indoor emissions, particularly combustion- and activity-related sources at the bus depot and mechanically generated metal-rich wear particles at the metro stations. Source-specific responses to air purifier operation suggest that filtration can complement source-control measures, although its effectiveness may depend on particle characteristics, spatiotemporal distribution, and continuous source generation. These findings support targeted air quality management in semi-closed transport environments.