Joint Seasonal/Annual Hydrologic Frequency Analysis

Abstract

Hydrologic frequency analysis, as commonly practiced, focuses on estimation of return periods associated with annual maxima or minima of str eamflow magnitudes, precipitation depths, etc. However, in some applications, it is desirable to perform joint (i.e., simultaneous) frequency analyses for seasonal as well as annual events. This paper presents arguments for performing joint seasonal/annual frequency analyses and highlights problems that are encountered in such analyses. Methods for coping with the estimation problems that are encountered in such analyses are also proposed. Introduction Hydrologic frequency analysis is the generic term given to statistical methods directed to estimation of the return period or exceedance probability associated with a hydrologic variable of a fixed magnitude. Such analyses also may be applied to estimate the magnitude of a hydrologic variable associated with a fixed exceedance probability. Traditionally, frequency analyses have focused mainly on annual maximum streamflows (either instantaneous or averaged over a fixed duration), annual minimum streamflows (usually averaged over a duration of 7 days), and annual maximum precipitation depths (accumulated depths over various durations). Fundamental assumptions underlying frequency analyses are that the data are independent and identically-distributed (iid). In other words, the data should be representative of a stochastically stationary process (no time trends, as might be caused by drainage basin urbanization or gradual climatic change), and the data should be homogeneous (all data should arise as a result of a common causative mechanism). Heterogeneous annual streamflow maxima could arise due to floods in some years being caused by precipitation events, and in other years being caused by snowmelt runoff. Heterogeneous precipitation data could be caused by annual