Design values for water resources projects rest on probabilistic hydrology: the return period of an event is the reciprocal of its annual exceedance probability, and this generates the related concepts of reliability (probability a design event is not exceeded during a structure's design life) and risk (its complement, the probability of at least one exceedance), which are traded off against cost and the consequence of failure. Because observed records are typically short relative to the design return period desired, frequency analysis techniques — extreme value probability distributions (Gumbel, log-Pearson Type III, log-normal) fit to annual-maximum series and characterized by a mean, standard deviation, and distribution-dependent frequency factor — are used to extrapolate design values, with confidence intervals quantifying the resulting estimation uncertainty. This belongs to engineering hydrology/hydrologic design, and connects probability theory to structural design practice via the factor of safety and safety margin.
Design values for water resources projects rest on probabilistic hydrology: the return period of an event is the reciprocal of its annual exceedance probability, and this generates the related concep…