공간분석 기법을 이용한 대기오염 개인노출추정 방안 소개 및 적용의 사례

Abstract

The objectives of this study were to introduce spatial interpolation methods which have been applied in recentpapers, to apply three methods (nearest monitor, inverse distance weighting, kriging) to domestic data (Ulsan cohort)as an example of estimating the personal exposure levels. We predicted the personal exposure estimates of 2,102participants in Ulsan cohort using spatial interpolation methods based on information of their residential address. We found that there was a similar tendency among the estimates of each method. The correlation coefficients betweenpredictions from pairs of interpolation methods (except for the correlation coefficient between nearest montitorand kriging of CO and SO2) were generally high (r=0.84 to 0.96). Even if there are some limitations such as locationand density of monitoring station, spatial interpolation methods can reflect spatial aspects of air pollutant and spatialheterogeneity in individual level so that they provide more accurate estimates than monitor data alone. But they maystill result in misclassification of exposure. To minimize misclassification for better estimates, we need to considerindividual characteristics such as daily activity pattern.