Floods are dangerous natural hazards that can threaten lives, livelihoods, and property. Without information on flood risk, the planning, design, and management of infrastructure along rivers would be forced to proceed without quantitative scientific analysis. Flood-frequency analysis provides information about the magnitude and frequency of flood discharges based on records of annual maximum instantaneous peak discharges collected at streamgages. In essence, flood-frequency analysis is used to estimate the probability of flooding at specific river locations.

To ensure consistency in flood frequency analysis, federal agencies have worked together to produce updated Guidelines for Determining Flood Flow Frequency, Bulletin 17C. Through the auspices of the Advisory Committee on Water Information (ACWI) and its Subcommittee on Hydrology (SOH), the previous guidelines were updated to reflect advances in statistical methods since the last update was published as Bulletin 17B in 1982. Additional information on the process for developing Bulletin 17C can be found in the Archive of Public Review Materials.

The Bulletin 17C guidelines retain the basic statistical framework used previously, allowing consistency with previous studies. It also incorporates the following advances:

• A more generalized representation of flood data, allowing description of flood peaks as intervals. This allows more robust incorporation of data based on historical information or paleoflood analysis.
• The Expected Moments Algorithm (EMA) is adopted as an improved method of moments approach to fitting the Log-Pearson Type III distribution to flood peaks. EMA is able to make use of the new data representation.
• A generalization of the Grubbs Beck test used in Bulletin 17B. The new method is called the Multiple Grubbs Beck Test (MGBT) and allows multiple potentially influential low floods (PILFs) to be identified.
• Corrected confidence intervals for the flood frequency curve. The Bulletin 17B computations were acknowledged to be a simplified and incomplete representation of the confidence intervals.
• Use of new methods for estimating regional skew and it uncertainty.

Information on application of these new methods can be found in the new Bulletin 17C document, available here.

For additional information on the basics of flood frequency analysis, see also
“The 100-Year-Flood-It’s All About Chance," a fact sheet written by the U.S. Geological Survey.

Bulletin 17C Home page