Within coupled hydrological simulation systems, taking socio-economic processes into account is still a challenging task. In particular, systems that aim at evaluating impacts of climatic change on large spatial and temporal scales cannot be based on the assumption that infrastructure, economy, demography and other human factors remain constant while physical boundary conditions change. Therefore, any meaningful simulation of possible future scenarios needs to enable socio-economic systems to react and to adapt to climatic changes. To achieve this it is necessary to simulate decision-making processes of the relevant actors in a way which is adequate for the scale, the catchment specific management problems to be investigated and finally the data availability. This contribution presents the DeepActor approach for representing such human decision processes, which makes use of a multi-actor simulation framework and has similarities to agent-based approaches. This DeepActor approach is embedded in Danubia, a coupled simulation system comprising 16 individual models to simulate Global Change impacts on the entire water cycle of the Upper Danube Catchment (Germany, 77,000 km2). The applicability of Danubia and in particular the DeepActor approach for treating the socio-economic part of the water cycle in a process-based way is demonstrated by means of concrete simulation models of the water supply sector and of the domestic water users. Results from scenario simulations are used to demonstrate the capabilities and limitations of the approach.