Estimating the upper tail of flood frequency distributions

Procedures for estimating recurrence intervals of extreme floods are developed. Estimation procedures proposed in this paper differ from standard procedures in that only the largest 10–20% of flood peaks are explicitly used to estimate flood quantiles. Quantile estimation procedures are developed for both annual peak and seasonal flood frequency distributions. The underlying model of flood peaks is a marked point process , where represents time of occurrence of the jth flood during year i and the mark represents magnitude of the flood peak. Results from extreme value theory are used to parameterize the upper tail of flood peak distributions. Quantile estimation procedures are applied to the 92-year record of flood peaks from the Potomac River. Results suggest that Potomac flood peaks are bounded above. The estimated upper bound is only 20% larger than the flood of record.

[1]  Laurens de Haan,et al.  Sample extremes: an elementary introduction , 1976 .

[2]  M. Bryson.  Heavy-Tailed Distributions: Properties and Tests , 1974 .

[3]  Richard L. Smith.  Threshold Methods for Sample Extremes , 1984 .

[4]  E. J. Gumbel,et al.  Statistics of Extremes. , 1960 .

[5]  J. Hosking.  Testing whether the shape parameter is zero in the generalized extreme-value distribution , 1984 .

[6]  P. Prescott,et al.  Maximum likeiihood estimation of the parameters of the three-parameter generalized extreme-value distribution from censored samples , 1983 .

[7]  A. Karr,et al.  Flood Frequency Analysis Using the Cox Regression Model , 1986 .

[8]  Two extreme value processes arising in hydrology , 1976 .

[9]  J. R. Wallis,et al.  Estimation of the generalized extreme-value distribution by the method of probability-weighted moments , 1985 .

[10]  P. Todorovic,et al.  Stochastic models of floods , 1978 .

[11]  M. Woo,et al.  Prediction of annual floods generated by mixed processes , 1982 .

[12]  M. A. Benson.  Evolution of methods for evaluating the occurrence of floods , 1962 .

[13]  J. Delleur,et al.  A cluster model for flood analysis , 1983 .

[14]  A. Walden,et al.  Maximum likelihood estimation of the parameters of the generalized extreme-value distribution , 1980 .

[15]  L. Duckstein,et al.  On the joint distribution of the largest flood and its time of occurrence , 1976 .

[16]  Maximum Likelihood Estimates for the Parameters of Mixture Distributions , 1984 .

[17]  R. Serfling.  Approximation Theorems of Mathematical Statistics , 1980 .

[18]  E. Zelenhasić,et al.  A Stochastic Model for Flood Analysis , 1970 .

[19]  J. Pickands.  Statistical Inference Using Extreme Order Statistics , 1975 .

[20]  Richard L. Smith.  Maximum likelihood estimation in a class of nonregular cases , 1985 .

[21]  W. DuMouchel.  Estimating the Stable Index $\alpha$ in Order to Measure Tail Thickness: A Critique , 1983 .

[22]  M. Bryson,et al.  Effect of tail behavior assumptions on flood predictions , 1980 .