Abstract

Abandoned flood plains or stream terraces that flank many streams may range in age from several hundred to tens of thousands of years. Terrace surfaces are underlain by stream-transported flood plain sediment, and thereby are quite easily modified by shallow flood inundation. During extreme flood events, streams in the western U.S. are often profoundly altered and many reaches exhibit clear evidence of erosion and deposition. If ages can be obtained for preserved and flood-modified surfaces, the surfaces become conservative datums for the magnitude and the recurrence of large floods. Likewise, the absence of features indicative of inundation provides positive evidence of non-occurrence of a threshold flood stage through time.

Causey Dam impounds water in the mountainous upper 210 km² of the South Fork Ogden River basin. The failure of Causey Dam during the calculated Probable Maximum Flood (PMF) could potentially lead to the failure of Pineview Dam, with attendant large economic damages and loss of life. This paleoflood study was initiated to reduce the range of uncertainty in estimates of the return periods of large floods.

A study reach downstream of Causey Dam was identified where paleostage indicators and datable deposits are well-preserved and the river channel is bedrock controlled. The study reach is bounded by late Pleistocene and Holocene surfaces that are about 5 m and 2 m, respectively, above the low-water channel. Because of the high gradient of the river, the up to 1 m of fine-grained deposits that underlie the Holocene surface are highly susceptible to erosion and channeling if overtopped by more than several centimeters of flow. Preliminary results from step-backwater modeling indicate that a discharge larger than 115 m³/sec would overtop Holocene surfaces that are a minimum of 2,500 years old. The absence of stratigraphic and geomorphic evidence for modification of this surface demonstrates that no discharge of this size or greater has occurred since this surface stabilized. For comparison, the largest recorded annual peak discharge is 53 m³/sec in May 1952.

The only way to obtain reliable probability estimates of rare flood events is to study the history of extreme floods on a given stream or in a given region. Paleoflood hydrology is the only event-based method for extending the length of the flood record as required to make realistic estimates of the probability for extreme floods. The calculated PMF at Causey Dam has a peak discharge of about 3,075 m³/sec. A discharge of this magnitude would overtop the Pleistocene surfaces in the study reach, but this has clearly not occurred since these surfaces stabilized more than 10,000 years ago. Similarly, if a flow comparable to the spillway capacity of Causey Dam, 214 m³/sec, had occurred, it would have overtopped the Holocene surfaces by about 1.5 to 2.5 m. There is no evidence for either of these events in the geologic record.

The limits on peak discharge provided by the non-inundation of fluvial surfaces can be combined with annual records of peak discharge to obtain robust estimates of the return periods of large floods. For the South Fork Ogden River, inclusion of paleoflood bounds in frequency analyses indicates that the return period for a discharge of 214 m³/sec, comparable to spillway capacity at Causey Dam, is substantially greater than 10,000 years.