ABSTRACT
The U.S. Geological Survey is actively involved in the development and testing of procedures for estimating flood-peak discharges at selected probabilities of exceedance. This paper describes the techniques utilized by the U.S. Geological Survey in estimating flood-peak discharges (1) at gaging stations, (2) at both rural and urban ungaged sites, and (3) at ungaged sites on streams with one or more gaging stations. The U.S. Geological Survey uses Bulletin 17B, Guidelines for Determining Flood Flow Frequency, for estimating flood-peak discharges at gaging stations. At ungaged sites on streams with one or more gaging stations, weighted estimates are made based on data at the nearby gaging station and regional estimates at the ungaged site. The regional estimates for ungaged sites are generally based on multiple-regression techniques that relate flood discharges of various exceedance probabilities to watershed, climatic, and channel-geometry characteristics. The accuracy of U.S. Geological Survey procedures for flood estimation for ungaged sites is also compared to six frequently used rainfall-runoff estimation techniques.
