Expected Moments Alogrithms for Flood Frequency Analysisby William L. Lane,
Timothy A. Cohn,
Document Type: Proceeding Paper
Abstract: The Expected Moments Algorithm (EMA) is a new but conceptually simple approach for the treatment of nontraditional data in flood frequency analysis. Nontraditional data includes censored data sets, historical information, and data other than the annual peak flows of normal gauged systematically collected streamfiow data. The EMA is designed for use in conjunction with the method of moments estimation of parameters for flood frequency distributions. EMA is a highly efficient approach for capturing the information contained in historical flood data, particularly paleoflood data. EMA is not limited in applicability to only historical information, but also will be useful in the treatment of a variety of other nontraditional forms of data. These include truncated data sets, various forms of censored data, both low and high outliers, and zero flows for some distributions such as log Normal and log Pearson. Simulations have shown EMA to be comparable in efficiency to the Maximum Likelihood estimation procedure, but with significant computation advantages (Cohn et al., 1996). The same approach is easily modified to be applicable to some other forms of parameter estimation, such as L Moments.
Subject Headings: Hydrologic data | Flood frequency | Data collection | Frequency analysis | Parameters (statistics) | Frequency distribution | Information management | Algorithms
Services: Buy this book/Buy this article
Return to search