A new grid-scale model simulating the spatiotemporal distribution of PM2.5-PAHs for exposure assessment

Abstract

Exposure to polycyclic aromatic hydrocarbons (PAHs) associated with ambient air particulate matter (PM) poses significant health concerns. Several modeling approaches have been developed for simulating ambient PAHs, but no hourly intra-urban spatial data are currently available. The aim of this study is to develop a new modeling strategy in simulating, on an hourly basis, grid-scale PM2.5-PAH levels. PM and PAHs were collected over a one-year time frame through an established air quality monitoring network within a metropolitan area of Taiwan. Multivariate linear regression models, in combination with correlation analysis and PAH source identification by principal component analysis (PCA), were performed to simulate hourly grid-scale PM2.5-PAH concentrations, taking criteria pollutants and meteorological variables selected as possible predictors. The simulated levels of 72-h personal exposure were found to be significantly (R=0.729**, p<0.01) correlated with those analyzed from portable personal monitors. A geographic information system (GIS) was used to visualize spatially distributed PM2.5-PAH concentrations of the modeling results. This new grid-scale modeling strategy, incorporating the output of simulated data by GIS, provides a useful and versatile tool in personal exposure analysis and in the health risk assessment of air pollution.