In the Rhine-Meuse delta in the south-western part of the Netherlands, the morphology of the river branches is highly dependent on the erodibility of the subsoil. Erosion processes that were initiated after closure of the Haringvliet estuary branch by a dam (in 1970), caused a strong incision of several connecting branches. Due to the geological evolution of this area the lithology of the subsoil shows large variations in highly erodible sand and poorly erodible peat and clay layers. In this study is shown how the geological information can be used to create 3D maps of the erodibility of the sub-soil, and how this information can be used to schematize the sub-soil in computational models for morphological simulations. Local incisement of sand patches between areas with poorly erodible bed causes deep scour holes, hence increasing the risk on river-bank instability (flow slides) and damage to constructions such as groynes, quays, tunnels, and pipelines. Different types of mathematical models, ranging from 1D (SOBEK) to quasi-3D (Delft3D) have been applied to study the future development of the river bed and possible management options. The results of these approaches demonstrate that models require inclusion of a bookkeeping-layer approach for sub-soil schematization, non-uniform sediment fractions (sand-mud), tidal and river-discharge boundary conditions, and capacity-reduction transport modeling. For risk-reducing river management it has been shown how the development of the river bed can be addressed on a large scale and small scale. For instance, the use of sediment feeding and fixation of bed can be proposed for large-scale management, while monitoring and interventions at initiation of erosion can be proposed as response to small-scale developments that exceed predefined intervention levels.