A dynamic fugacity model was applied to simulate the changes of contents and transfer fluxes of hexachlorocyclohexane (HCHs) and dichloro-diphenyl-trichloroethane (DDTs) from 1950s in the environment of Hangzhou, China. The receptors are composed of air, surface water, soils, sediment and biota compartments. The model provides a method to combine loadings of HCHs and DDTs from various sources with a series of physical-chemical processes to estimate concentrations and transport fluxes of HCHs and DDTs. Model results suggested that the calculated concentrations were in line with the observed ones. The highest contents of HCH and DDT in the environment of study area were 523 t and 471 t before 1983, among which about 80.7% HCHs and 93.2% DDTs remained in the soil compartment. From 1984 to now, contents of HCHs and DDTs had decreased to about 0.07% and 0.40% of their highest amount (before 1983), and only about 0.001% and 0.014% will expect to be left in 2020 in the study area according to the model prediction. Before 1983, the main transfer fluxes of HCHs were deposition from air to soil, runoff from soil to water and diffusion from soil to air, but for DDTs the main transfer fluxes were deposition from air to soil and water, and transfer from water to sediment. From 1984 to now, runoff from soil to water and transfer from water to sediment became the dominant processes. Although a large amount of HCHs and DDTs had been applied to the study area, their residue levels in the soils were much lower than those in North China (had lesser HCHs and DDTs application than in South China) at present time, and close to other locations of South China (had similar HCHs and DDTs application level). It can be attributed to the high precipitation and temperature that enhances the processes of wet deposition, evaporation and degradation of OCPs. Sensitivities of the input parameters to the calculated concentrations were evaluated using coefficient-of-variation normalized sensitivity coefficients. The model was also subjected to uncertainty analyses using a Monte Carlo simulation. (c) 2006 Elsevier Ltd. All rights reserved.
