COPING WITH UNCERTAINTIES IN FLOOD MANAGEMENT

摘要

Main elements of risk-based design of flood management are considered: flood as a load or exposure, resistance/capacity of flood control works, and consequences of possible flooding (economic, human life, ecological). All the three elements can be described with much uncertainty stemming from limited information and imprecise models/definitions. As a consequence, the classical risk-based design is often incorrect and cannot be defended, because it may lead to over or under design; both resulting in economic, social and ecological losses as well as unwarranted human life hazards. Statistical/probabilistic methods have difficulties to encode such uncertainties. Fuzzy logic, as an alternative, is shown to be possibly better applicable to the present situation. It is shown how that methodology can be used to find the best flood management alternative under uncertainties in view of satisfying - the often conflicting - economic, human life and ecological objectives.