The goal of this work is to develop a unified probabilistic framework, based on both theory and data interpretation, to assess the uncertainty of extreme rainfall. This research is directed toward the development of a risk-based design as an alternative to the currently used quasi-deterministic Probable Maximum Precipitation (PMP) procedure. Three main tasks will be undertaken: 1) to derive the probability of exceedance associated with the PMP by assessing and combining the errors involved in each step of its computation; 2) to perform an exploratory statistical analysis of an available data base of extreme storms in order to derive mathematical descriptions of the spatial and time distribution of rainfall extremes; and 3) to explore probabilistic methods to assess the exceedance probabilities of extreme precipitation.
