Planning of infrastructure and disaster risk mitigation measures is carried out under significant uncertainties. In particular, the effects of climate and anthropogenic change are associated with large uncertainty, but most extreme events are poorly understood even under stationary conditions. To address these uncertainties, the designers/decision makers can follow different strategies, such as conservative designs, flexible (adaptable) designs or delaying the decisions to later times when more information is available. The optimal adaptation strategy depends on multiple factors such as system type, regulatory framework, the degree and type of uncertainty and the amount of learning that is possible in the future. Based on quantitative Bayesian decision models, we derive general recommendations on optimal design approaches. Our results show, for example, that flexible systems are especially beneficial in cases where uncertainty is high and where the learning effect in the near future is expected to be significant. They are typically less advantageous in the context of a risk-based decision framework (where the aim is to find a balance between the residual risk and cost) than in a rule-based regulatory framework (where safety requirements are prescribed, e.g. that flood protection must be designed against a 100-year event). In the former case, it is often more effective to add safety margins, which is a no-regret strategy.     «

Planning of infrastructure and disaster risk mitigation measures is carried out under significant uncertainties. In particular, the effects of climate and anthropogenic change are associated with large uncertainty, but most extreme events are poorly understood even under stationary conditions. To address these uncertainties, the designers/decision makers can follow different strategies, such as conservative designs, flexible (adaptable) designs or delaying the decisions to later times when mo...     »