Studies relied on atmospheric PM2.5 data to estimate health risks and sources, but its representativeness for indoor air remains unclear. This study placed an atmospheric PM2.5 sampler on the rooftop of a three-story building, and indoor and outdoor (balcony) samples were collected from study houses within a 5 km radius of the atmospheric sampling site during low and high PM2.5 seasons to evaluate the representativeness of atmospheric PM2.5 data. The average PM2.5 concentrations were 15.4 ± 7.1 μg/m3 (indoor), 18.9 ± 7.0 μg/m3 (outdoor), and 13.9 ± 6.1 μg/m3 (atmospheric). Atmospheric PM2.5 concentrations were significantly associated with indoor and outdoor PM2.5, but outdoor PM2.5 concentrations were higher. Source identification revealed that traffic-related emission was a major contributor to PM2.5 across all sites and seasons, while long-range transport from China was another source during the high PM2.5 season based on lead isotope ratios. Crustal elements and construction dust, identified through positive matrix factorization, were higher in both indoor and outdoor air than in the atmosphere. Elements from industrial or traffic emissions showed similar concentrations across different sampling sites. The average ratios of indoor to outdoor or atmospheric PM2.5 concentrations were 0.84 ± 0.27 and 1.18 ± 0.37, respectively. In conclusion, atmospheric PM2.5 can be used to estimate PM2.5 exposure and element concentrations from industrial or traffic emissions in indoors; however, it underestimates the contributions of crustal elements and construction dust. Ambient PM2.5 samples from outdoor or atmospheric environments should be considered when comparing their influence on indoor air across studies.
Keywords: Atmosphere; Construction; Indoor air; Pb isotope; Positive matrix factorization.
Copyright © 2025 Elsevier Ltd. All rights reserved.