|| Checking for direct PDF access through Ovid
Many studies have focused on the relationships of particulate matter between indoor, outdoor and personal exposure; however, considerable uncertainties remain regarding the portion of indoor particles and personal exposure of ambient origin. As part of the Particle Exposure Assessment for Community Elderly (PEACE) study in Tianjin, China, we have further interpreted the relationships between personal, residential indoor, outdoor and community PM10 (particulate matter with aerodynamic diameters of less than 10 μm). Comparisons of the chemical compositions of PM10 samples were performed using the coefficient of divergence (COD). A robust regression method, least-trimmed squared (LTS) regression, was used to estimate the infiltration factors of PM10 from residential outdoor to indoor environments based on the particulate component concentrations. Personal exposures of ambient origin were also estimated. A relatively good correlation was found between the personal and indoor PM10 samples with respect to chemical composition. The infiltration factors (Finf) of the residential indoor-outdoor PM10 were 0.74 ± 0.31 (mean ± SD) in summer and 0.44 ± 0.22 in winter, with medians of 0.98 and 0.48, respectively. The residential outdoor contributions to the indoor environments were 87 ± 55 μg/m3 in summer and 80 ± 54 μg/m3 in winter, with medians of 75 μg/m3 and 61 μg/m3, respectively. The personal exposures of ambient origin were 92 ± 44 μg/m3 in summer and 89 ± 47 μg/m3 in winter, with medians of 81 μg/m3 and 80 μg/m3, respectively. This study indicated that the infiltrations in an urbanized area in North China exhibited a seasonal difference: the residential outdoor contributions to residential indoor environments were larger in summer due to the higher use of natural ventilation. The personal exposures of ambient origin were comparable during the different seasons, whereas those of non-ambient origin were higher in summer than in winter.