We measured PM_2.5 in 41 underground shopping districts (USDs) in the Seoul metropolitan area from June to November 2017, and associated 18 trace elements to determine the sources and assess the respiratory risks. The PM_2.5 concentrations were 18.0 ± 8.0 μg/m³ inside USDs, which were lower than 25.2 ± 10.6 μg/m³ outside. We identified five sources such as indoor miscellanea, soil dust, vehicle exhaust/cooking, coal combustion, and road/subway dust, using factor analysis. Almost 67% of the total trace element concentration resulted from soil dust. Soil dust contribution increased with the number of stores because of fugitive dust emissions due to an increase in passers-by. Vehicle exhaust/cooking contribution was higher when the entrances of the USDs were closed, whereas coal combustion contribution was higher when the entrances of the USDs were open. Although miscellanea and coal combustion contributions were 3.4% and 0.7%, respectively, among five elements with cancer risk, Cr and Ni were included in miscellanea, and Pb, Cd, and As were included in coal combustion. The excess cancer risk (ECR) was the highest at 67 × 10⁻⁶ for Cr, and the ECR for Pb was lower than 10⁻⁶, a goal of the United States Environmental Protection Agency for hazardous air pollutants.

Underground shopping districts (USDs) are susceptible to severe indoor air pollution, which can adversely impact human health. We measured 24 volatile organic compounds (VOCs) in 13 USDs throughout South Korea from July to October 2017, and the human risk of inhaling hazardous substances was evaluated. The sum of the concentrations of the 24 VOCs was much higher inside the USDs than in the open air. Based on factor analysis, six indoor air pollution sources were identified. Despite the expectation of a partial outdoor effect, the impacts of the indoor emissions were significant, resulting in an indoor/outdoor (I/O) ratio of 5.9 and indicating elevated indoor air pollution. However, the effects of indoor emissions decreased, and the contributions of the pollution sources reduced when the USD entrances were open and the stores were closed. Although benzene, formaldehyde, and acetaldehyde exhibited lower concentrations compared to previous studies, they still posed health risks in both indoor and outdoor settings. Particularly, while the indoor excess cancer risk (ECR) of formaldehyde was ~10 times higher than its outdoor ECR, benzene had a low I/O ratio (1.1) and a similar ECR value. Therefore, indoor VOC concentrations could be reduced by managing inputs of open air into USDs.