Flood-frequency prediction: For estimating floods of a specified return period at a site using annual-maximum series, which of the following methods is generally regarded as the most reliable among the choices given?

Introduction / Context:Design discharges for bridges, culverts, and spillways often require flood-frequency analysis rather than just runoff transformation methods. Statistical methods based on extreme value theory are preferred to estimate discharges for specified return periods (e.g., 50-year, 100-year floods).

Given Data / Assumptions:

• Annual maximum (or peak-over-threshold) discharge records are available.
• Objective is to estimate a quantile associated with a chosen return period.

Concept / Approach:Gumbel’s method (Type I Extreme Value) fits a theoretical distribution to annual maxima, enabling extrapolation to desired return periods. In contrast, a unit hydrograph converts rainfall to runoff (not a frequency tool), and empirical regional formulae (e.g., California method) are rough and site-limited.

Step-by-Step Reasoning:Compile annual maximum flood series.Estimate Gumbel parameters (mean and standard deviation or method of moments).Compute reduced variate and derive design flood for the target return period.

Verification / Alternative check:For many basins, regional frequency analysis (e.g., Log-Pearson Type III) may be adopted by standards. Among the listed options, however, Gumbel’s method is the most directly intended for frequency estimation of extremes.

Why Other Options Are Wrong:

• Unit hydrograph: A runoff transformation tool; not a statistical frequency method.
• California method: Empirical and regional; less reliable than analytical fitting of extremes.
• Rational method without frequency: Lacks statistical basis for return period selection.
• None of these: Incorrect because Gumbel is suitable.

Common Pitfalls:

• Confusing hydrologic transformation tools with statistical frequency tools.
• Applying empirical regional equations outside their calibration region.

Final Answer:Gumbel’s analytical method (extreme value distribution)