抄録
We propose a useful exposure analysis technique based on the analytical framework linking multivariable analysis and regression analysis through the classification of monitoring points using observed concentrations of 4-nonylphenol, some surrogate indexes for emission quantity and basin characteristics in Japan. In order to accomplish this work, we constructed a nationwide basin database from data of chemical emission and watershed landscape. By cross tabulation and discriminant analysis using three variables, namely, pollutant load factor, sewage treatment factor and scaling factor, we classified the monitoring sites of 4-nonylphenol concentration into five groups: ‘emission from manufacturing industry’, ‘emission from wastewater treatment plant’, ‘emission from unsewered area’, ‘dilution and removal’, and ‘the others’. Then, we employed the principal component analysis with pollutant load factor, sewage treatment factor, scaling factor, population density and urban activity to characterize the classification groups. We obtained regression equations that explained 4-nonylphenol concentrations within a confidential interval of 10 percent for samples of the ‘emission from manufacturing industry’ group. This method can be utilized for the preliminary exposure analysis of chemicals based on their basin characteristics.
