An integrated model provides a means for understanding what is likely to happen in the future under various circumstances. But this capability must be made available in a way that is easily accessible to the user: the planner or decision maker. For this, the model must have several characteristics: adequately represent all the important processes necessary to provide the relevant policy outputs; provide information at spatial resolutions reflecting the scale of variation in the most important variables. A spatial scale also at which policy problems occur and can be adressed; operate at time scales and temporal resolutions representing realisticaly the autonomous dynamics of the system modelled. A time horizon also which is relevant for policy design, implementation and assessment; assist in the solution of loosely structured problems; deliver reliable information in a usable form; easily modifiable to represent evolving circumstances or changes in the user's needs; easy to calibrate to correctly represent new situations of interest to the user; provide a useful summary and evaluation of the results. Part of these characteristics are achieved by the choice and integration of the appropriate set of sub-models and their application at the right levels of detail. Other characteristics are achieved by the transparency of the models, the architecture and the technical implementation of the software system, the graphical user interfaces and the degree to which the DSS provides effective support for the problems in the policy field. RIKS combines all these aspects in its Decision Support Systems. Graphical Interfaces Our interfaces are essentially graphical. Data and results are displayed as dynamic maps or as graphs and charts. The model itself is represented graphically by a flow diagram -- one with which the user can interact. Clicking on a part of the diagram allows the user to specify that part of the model to his own needs -- or, if the model is already running, to view results for that part of the model. Decision Support Using the graphical interface it is easy to specify various scenarios to be investigated, or to run "what if" experiments. The models will automatically take into account the uncertainties inherent in any prediction of the future, and show a range of probabilities for possible outcomes. Pre- and Post-processing tools Additional support tools enable to interactively prepare the necessary inputs fitting seamlessly the requirements of the models, or provide an interactive multiple criteria, multiple objective evaluation of the various policies that have been simulated, in order to facilitate the choice of the most effective policy. Users The system is equally accessible both to users with little expertise in computers or modelling, and to technicians and researchers with extensive modelling experience. For the first group, the graphical user interface and associated decision support tools permit questions of policy to be addressed in a systematic, comprehensive way, without the requirement of any particular technical expertise. For the second group, the interface permits the modeller to deal directly with the technical issues which interest him, without the inconvenience of having to write or rewrite software. For both user groups, there are large savings of time and cost, and considerable satisfaction with the results.