| Summary: | 碩士 === 國立臺灣大學 === 環境衛生研究所 === 103 === Traffic intensity, length of road, and proximity to roads are the most common traffic indicators in the land use regression (LUR) models for particulate matter in ESCAPE study areas in Europe. This study conducted three 14-day measurments at 20 monitoring sites in Taipei metropolis and explored what local variables can improve the performance of LUR models in an Asian metropolis with high densities of roads and strong activities of industry, commerce and construction. By following the ESCAPE procedure, we derived LUR models of PM2.5, PM2.5 absorbance, PM10, and PMcoarse (PM10－PM2.5) in Taipei. The overall annual average concentrations of PM2.5, PM10, and PMcoarse were 26.0 ± 5.6, 48.6 ± 5.9, and 23.3 ± 3.1 μg m-3, respectively, and the absorption coefficient of PM2.5 was 2.0 ± 0.4 × 10-5 m-1. Our LUR models yielded R2 values of 0. 95, 0.96, 0.87, and 0.65 for PM2.5, PM2.5 absorbance, PM10, and PMcoarse, respectively. PM2.5 levels were increased by local traffic variables, industrial, construction, and residential land-use variables and decreased by rivers; while PM2.5 absorbance levels were increased by local traffic variables, industrial, and commercial land-use variables in the models. PM10 levels were similar to PM2.5, increased by local traffic variables, industrial, commercial, and construction land-use variables. PMcoarse levels were increased by elevated highways. Road area explained more variance than road length by increasing the incremental value of 0.27 and 0.06 adjusted R2 for PM2.5 and PM10 models, respectively. In the PM2.5 absorbance model, road area and transportation facility explain 0.29 more variance than road length. In the PMcoarse model, industrial and new local variables instead of road length improved the incremental value of adjusted R2 from 0.39 to 0.60. We concluded that road area can better explain the spatial distribution of PM2.5 and PM2.5 absorbance concentrations than road length. By incorporating road area and other new local variables, the performance of each PM LUR model was improved. The results suggest that LUR model can be applicated as a useful approach for assessing and predicting exposures in the context of environmental health impact assessment or environmental policy control.
|
|---|
