Regional Frequency Analysis of Hydrologic Variables

Abstract

In its 1988 report the US National Research Council (NRC) Committee on Estimating the Probabilities of Extreme Floods identified three principles for improving flood estimation. These are (1) “substitution of space for time”; (2) introduction of more “structure” into the models; and (3) focus on extremes or tails as opposed to or even to the exclusion of central characteristics” (NRC 1988). In the more general setting of hydrologic variables, the NRC’s first principle points toward the regional frequency analysis as a valid alternative to at-site frequency analysis and its attendant shortcomings. In fact, the regional frequency analysis of variables, broadly defined as the set of methods for analyzing pooled information collected at distinct sites across a geographic region, seeks to compensate for the usually short samples of hydrologic variables with a better description of them in space. After introducing the principles of regional frequency analysis and its goals in the first section, the issue of delineating hydrologically homogenous regions is considered in Sect. 10.2. Three different general approaches to performing a regional frequency analysis are described and exemplified in Sect. 10.3. The so-called index-flood approach, together with statistical inference through L-moments, is an important development in regional frequency analysis that has received much attention in recent years and, thus, is the object of a more detailed description in Sect. 10.4. A number of examples illustrate applications of the methods described.