Over the last decades, river floods in Europe seem to occur more frequently and are causing more and more economic and emotional damage. Understanding the processes causing flooding and the development of simulation models to evaluate countermeasures to control that damage are important issues. This study deals with the application of a 2D hydraulic flood propagation model for flood hazard and risk assessment. It focuses on two components: 1) how well does it predict the spatial-dynamic characteristics of floods and 2) how can the model results be transformed into a flood risk assessment. Flood risk results from the interaction of flood water with human activities. This makes flood risk assessment a multi-disciplinary endeavour: on one hand it requires good understanding of fluvial processes and flood behaviour; on the other hand a methodology is needed to quantify its impact on the socio-economic environment. In this study it is argued that in low-land areas with near-flat terrain and complex topography, traditional definitions of flood hazard and risk, based on magnitude-frequency relations, are not useful and that additional information is needed to differentiate the hazard using multiple parameters, for instance flood propagation, maximum water depths and flow velocities, duration, etc. Two-dimensional flood models are the appropriate tools for simulating flow of water to assess the consequences of terrain modifications on the flood characteristics. This is useful when flood consideration need to be included in the decision-making process and Environmental Impact Assessment studies. This study presents three case-studies. The inundation of the Ziltendorfer Niederung (Germany) during the 1997 Oder floods is used to test the 2D hydraulic model Delft-FLS. It is shown that the model yielded good results in reconstructing this flood event and that differences between predicted and observed water depths can largely be attributed to errors in the input data. In the second case-study, the model is applied to the Adige floodplain near Trento (Italy) to assess the flood consequences of a planned new embanked motorway. It is concluded that such developments in flood-prone areas can have a significant effect on the flood characteristics and will result in a redistribution of the flood risk in the area. The third case-study, the "Land van Maas en Waal" in the Netherlands, studies the effect of sub-compartments in a polder. It is concluded that compartmentalizing structures have a significant effect on the propagation of the flood through the area, the speed of water level rising and on the duration of the inundation. It is demonstrated that Spatial Multi-Criteria Evaluation (SMCE) can be used for multi-parameter flood-risk assessment to generate flood-risk maps that can be used in the decision making process. It is concluded that Delft-FLS is a powerful tool to simulate flood behaviour over complex topography and that "what-if" scenarios can help to assess the consequences of decisions and actions regarding flood characteristics. It is also concluded that for the incorporation of flood considerations in the decision-making process, the model results need to be transformed to risk. For this, SMCE could be a helpful tool.
