文摘
A multimedia mass balance model (MMM), POPsME, wasevaluated for its performance to assess the fate andtransport of 2,3,7,8-substituted PCDD/Fs in Seoul and aneighboring area (150 km × 150 km) in Korea. As part ofthe evaluation, the effects of the grid size with thehomogeneous mixing assumption were estimated on thesign and magnitude of the prediction bias. The predictioncompared with the monitoring data generally within 1order of magnitude. In the presence of a concentrationgradient within individual cells of 30 km × 30 km, the predictionbias for air and soil tended to increase with the distancefrom the sources. It was observed that the ratio of thepredicted to the measured values could vary more than 4times with the location of the monitoring site. Use ofthe large grid size in POPsME (and perhaps other MMMs)resulted in overpredictions for a major portion of themodel domain at the cost of underprediction for limitedareas near the sources. Beyond the source areas, a substantialcontribution of the total prediction bias originated fromthe congeners with a large atmospheric loss rate and smallTEF value. On a TEQ basis, therefore, the magnitude ofthe overpredictions could be significantly reduced ascompared to that of the underprediction for the sourceareas.