Author(s): Ezio Todini
Linked Author(s):
Keywords: Predictive Uncertainty; Emulation Uncertainty; Flood Forecasting; Uncertainty Processors
Abstract: This work aims at discussing the role and the relevance of “predictive uncertainty” in flood forecasting and flood emergency management. Predictive uncertainty, is here defined as the probability of occurrence of a future value of a predictand (such as water level, discharge or water volume) conditional on prior observations and knowledge as well as on all the information that can be obtained on that specific future value, which is typically embodied in one or more hydrological/hydraulic model forecasts. After clarifying the conceptual difference between “emulation uncertainty” (commonly used when dealing with model verification) and “predictive uncertainty” (which is used when forecasting future outcomes) three presently available uncertainty processors, are introduced and discussed in terms of their benefits and limitations. The paper concludes by demonstrating, on the basis of an operational flood warning system on the Po river in Italy, the benefits that may arise from the incorporation of predictive uncertainty into the decision making process.
Year: 2010