519.21The paper considers models and approaches to the analysis and decision making under catastrophic risks. It is shown that the design of optimal robust strategies for the flood risk management can be approached as a stochastic spatially explicit optimization problem combining the goals and constraints of various agents such as producers, farmers, individuals, governments, insurers, reinsurers, and investors. The approach is illustrated with a case study on catastrophic flood risks, which shows the importance of an appropriate combination of ex-ante and ex-post structural and financial measures.