Analogous to the CSOIL model, which estimates human exposure to contaminated soils, an exposure model called SEDISOIL has been developed to quantify human exposure to sediments. When using this model it should be kept in mind that calculations are not very accurate, certainly when measurements in surface water and on fish have not been performed. SEDISOIL can be used for determinating the urgency of remediation in the case of seriously contaminated soils.

This report describes a procedure to develop a risk assessment methodology for contaminated sites with respect to the risk of dispersal in groundwater. The methodology was originally intended for landfills, but is for example also usable for risk assessment of contaminated industrial sites and sludge depots or to determine the priority of remediation. A Geographic Information System database can be constructed from available databases and preliminary modelling studies, which should then be ranked according to a fuzzy logic system, and incorporated into a Decision Support System. The final product, the Decision Support System, is a flexible instrument for making environmental policy decisions where the available data is vaguely defined and/or incomplete.

In general, pesticides which are liable to leach into groundwater at concentrations beyond the guideline concentration will not be authorised in the Netherlands. However, in the Long-term Crop Protection Plan adopted by the Dutch Government in 1991, it is stated that these pesticides can still be registered if the transformation rate in the saturated zone is sufficiently high. The transformation rate is considered to be 'sufficiently high' if the threshold concentration of 0.1 mg dm-3 is not exceeded at a depth of 10 m below the soil surface after a travel time of four years. The procedure for evaluating the behaviour of pesticides in the saturated zone will be discussed here, along with the secondary conditions for experimental research. In view of the absence of more precise knowledge on the behaviour of pesticides in the saturated zone, this procedure is considered the best achievable at the moment. The evaluation procedure is incorporated in the Decree on Environmental Criteria for Pesticides as an amendment to the Dutch Pesticide Act of 1962. The travel time of four years, as stated in the long-term crop protection plan, is adequate for the evaluation procedure; the travel time of water in the zone 1 to 10 m below the soil surface is always longer than four years, except for groundwater in the immediate vicinity of abstraction wells and a few sandy areas not in agricultural use.

In view of a revision of the Dutch Soil clean-up guideline in this report proposals are presented for human toxicologically based C- standard values for soil and groundwater, calculated from human toxicological guideline values and an exposure analysis. In this report the exposure analysis model CSOIL is presented and discussed. This model has been developed to quantify the exposure through the different exposure routes which are important in case of soil contamination. Moreover, the uncertainties are discussed with reference to this model. The model is not only used for the derivation of C-standard values, but is also suitable for actual exposure analysis.

A preliminary investigation was conducted on the occurrence of particular pesticides in the soil. The aim was to gain insight into the appearance and behaviour of pesticides in soil from the plough layer in order to support the admission policy. Seven pesticides were chosen for determination on the basis of their accumulation properties and their market share. Four pesticides were found in the plough layer: atrazine, fenpropimorph, parathion and simazine. The amounts measured did hardly exceed the detection limit, with the exception of atrazine which showed relatively high levels for a number of plots, but also a decline in time. Paraquat, pencycuron and pirimicarb could not be detected in any of the plots. The measured amounts correspond reasonably well with the amounts calculated with the simulation model PESTLA.

The aim of this Programming Study on 'Behaviour of Accumulated Contaminants in relation to Land Use Changes and Recoverability of Ecosystems' was to determine: 1) the available knowledge and 2) the short-, medium- and long-term research, needed to develop a risk assessment for the effects of mobilisation of nutrients and contaminants due to land-use changes from agriculture to nature or forest conservation. The most relevant changes in soil condition, which can take place after afforestation, are concluded to be acidification and increase in dissolved organic carbon. The development of wetlands can lead to decreases in redoxpotential and acidification. These changes in soil condition can result in mobilisation of phosphate and heavy metals. Risks from organic micropollutants are probably lower ; however, as yet not much is known about this. Increased contaminant mobility can lead to a decrease in decomposition, an increase in internal concentrations of toxicants in soil fauna and secondary poisoning of terrestrial fauna. The study also concludes that risk assessment for land-use changes will require integration of existing models. A 'quick-and-dirty' model integration is recommended as a 1 to 2-year research programme, along with research on existing nature-development projects to be used as model systems for medium-term research programmes. For long-term research priority will be given to monitoring, process research and model integration.

Proposals for Intervention Values for soil and groundwater have been derived for the fourth series of compounds. This fourth series consists of the following 15 additional compounds and compound groups: 1) metals: um, selenium, tellurium, thallium, tin; 2) aromatic hydrocarbons: monochloroanilines to pentachloroanilines, 4-chloro-methylphenol; 3) chlorinated hydrocarbons: 1,1,2-trichloroethane, dichloropropanes, 1,1-dichloroethene; 4) pesticides: MCPA and 5) other contaminants: tribromomethane, isopropanol, ethylacetate, 1,2-butylacetate. Intervention Values were derived by integrating the ecotoxicological and human-toxicological serious soil contamination concentration (ECOTOX SCC and HUM-TOX SCC, respectively). The HUM-TOX SCC is subject of this report and was derived using the CSOIL exposure model, the standard potential exposure data set, selected contaminant specific physicochemical input parameters and the Maximal Permissable Risk levels for human intake (MPRhuman). In general, the proposals for Intervention Values for groundwater and to a lesser extent the Intervention Values for soil are characterised by a limited accuracy.

Socio-cultural patterns and socio-economic circumstances of Dutch society are changing quickly. While these developments exert a great impact on future land use, scenarios allowing the effect of different spatial strategies on the environment, natural areas or landscape to be quantified are often limited. To cope with this limitation a project group was set up at the initiative of the RIVM to develop a spatial information system. the result was an integrated, spatial information system, the Spacescanner, which integrates geographical data and mathematical models to predict the effect of demographic and economic developments. Spacescanner 1.0 is a prototype simulating future landuse. The resulting maps have a spatial accuracy of 500 x 500 m. The Spacescanner integrates present and expected land-use data from the above models via the so-called allocation module for use in predicting the attraction of a location for a certain kind of land use.

In the framework of the project Area Atlas, about 50 maps, spread over about 150 digital files were collected for several types of area-specific policy. The aim of the project was to overview the existing maps and files as completely as possible. The files, including the file manipulation and presentation program ArcView version 1, have been made available on CD-ROM but exclusively to the contributors of this project. A cumulation map for area-specific policy has been derived using 25 maps suitable for this purpose. From this, it appears that at some locations a maximum of 17 different area-specific policies are applicable.

The ecotoxicological risk concept is considered to be the basis for setting environmental quality objectives for pesticides in groundwater. Ecotoxicological critical concentrations for groundwater can be derived through several exploratory approaches. These approaches will be subsequently compared (focusing mainly on the Netherlands but bringing the European Union standard for pesticides in groundwater into the relevant context ) with critically reviewed groundwater monitoring data. Furthermore, aldicarb, atrazine and MITC results of geographic modelling are ecotoxicologically evaluated. The study focuses mainly on the Netherlands but the European context is not forgotten. Many reported groundwater monitoring data are strongly correlated. Therefore the assessment of the extent of groundwater contamination should be based on the number of localities where pesticides have been detected. Shallow Dutch groundwater monitoring data reveal that aldicarb, 1,3-dichloropropene, 1,2-dichloropropane, dinoseb, dinoterb, ethoprophos and MITC are ecotoxicologically the most hazardous pesticides. Of the pesticides found in monitoring programs critical levels are lower than 0.1 4g/l for aldicarb and aldicarb sulfoxide, dinoseb, dinoterb, ethoprophos, heptachlor, MITC, parathion ethyl and pirimicarb. If ecotoxicological risk levels for groundwater are desired they should be derived from standard aquatic toxicity data, and the ecological characteristics of the system should be taken into account in the extrapolation. It was not found meaningful to develop ecotoxicity tests based on specific groundwater organisms.