Multicompartmental fate of persistent substances. Comparison of predictions from multi-media box models and a multicompartment chemistry-atmospheric transport model.

BACKGROUND, AIM AND SCOPE Modelling of the fate of environmental chemicals can be done by relatively simple multi-media box models or using complex atmospheric transport models. It was the aim of this work to compare the results obtained for both types of models using a small set of non-ionic and non-polar or moderately polar organic chemicals, known to be distributed over long distances. MATERIALS AND METHODS Predictions of multimedia exposure models of different types, namely three multimedia mass-balance box models (MBMs), two in the steady state and one in the non-steady state mode, and one non-steady state multicompartment chemistry-atmospheric transport model (MCTM), are compared for the first time. The models used are SimpleBox, Chemrange, the MPI-MBM and the MPI-MCTM. The target parameters addressed are compartmental distributions (i.e. mass fractions in the compartments), overall environmental residence time (i.e. overall persistence and eventually including other final sinks, such as loss to the deep sea) and a measure for the long-range transport potential. These are derived for atrazine, benz-[a]-pyrene, DDT, alpha and gamma-hexachlorocyclohexane, methyl parathion and various modes of substance entry into the model world. RESULTS AND DISCUSSION Compartmental distributions in steady state were compared. Steady state needed 2-10 years to be established in the MCTM. The highest fraction of the substances in air is predicted by the MCTM. Accordingly, the other models predict longer substance persistence in most cases. The results suggest that temperature affects the compartmental distribution more in the box models, while it is only one among many climate factors acting in the transport model. The representation of final sinks in the models, e.g. burial in the sediment, is key for model-based compartmental distribution and persistence predictions. There is a tendency of MBMs to overestimate substance sinks in air and to underestimate atmospheric transport velocity as a consequence of the neglection of the temporal and spatial variabilities of these parameters. Therefore, the long-range transport potential in air derived from MCTM simulations exceeds the one from Chemrange in most cases and least for substances which undergo slow degradation in air. CONCLUSIONS AND PERSPECTIVES MBMs should be improved such as to ascertain that the significance of the atmosphere for the multicompartmental cycling is not systematically underestimated. Both types of models should be improved such as to cover degradation in air in the particle-bound state and transport via ocean currents. A detailed understanding of the deviations observed in this work and elsewhere should be gained and multimedia fate box models could then be 'tuned in' to match better the results of comprehensive multicompartmental transport models.