Scenario analysis of a level III multimedia model using generic and regional data

In order to realise EU legislation, the environmental risk of new and exist ing chemical substances has to be assessed. The methods for risk assessment are laid down in the Technical Guidance Documents TGD (EC 1996A). Basicall y, risk assessment consists of a comparison of predicted environmental conc entrations (exposure assessment) and concentrations which may cause adverse effects (effect assessment). With the European Union System for the Evalua tion of Substances EUSES (EC 1996B) a decision support system is available which integrates the models presented in the TGD into a Windows 95/NT-based software product. Meanwhile, EUSES has attained a far reaching and extensi ve use in governments, industry, and research institutes. For exposure assessment, various exposure models (regression equations, sim ple generic models, multimedia compartment models) are used. One of the obj ectives is to investigate the validity of model calculations and evaluate u ncertainty in models and data arising in model application. Scientific just ification of underlying equations, limitations, and respective software are tested and checked for possible errors. The project consists of several wo rk packages: comprehensive data collection and evaluation, conceptual valid ation, model simulations, sensitivity and uncertainty analyses, comparison of model results with measured values (numerical validation), appropriatene ss of models (operational validation), comparison with alternative models, and software evaluation. It is determined for which substances and for whic h environmental segments the models deliver realistic values. Within this p roject, exposure assessments are carried out for different substances which represent a wide variety of physical-chemical properties. Substances of in terest are those which are used in many products or chemical processes and which are frequently found in the environment. Due to their relatively good data availability, typical and well-known pollutants such as dioxins and P CBs are also investigated. For validation studies, in particular to compare measured concentrations with predicted ones, spatially and temporally cohe rent data sets must be available. Thus, before starting the project, an env ironmental segment of regional scale had to be chosen. The German state of North Rhine-Westphalia (approximately 34 000 km(2)) was selected as an envi ronmental segment of regional scale.