USING PALEOFLOOD AND NON-EXCEEDANCE DATA TO BETTER COMMUNICATE FLOOD HAZARD INFORMATION FOR RISK ASSESSMENT

An important aspect for better understanding flood hazard is in the characterization of the magnitude and frequency of large floods. Evidence of extreme pre-historical floods preserved in the geologic record and of long-term landscape stability can be used to estimate the flood hazard. A study of the Kern River in southern California was undertaken by the Bureau of Reclamation for the U.S. Army Corps of Engineers as a pilot study to illustrate the advantages of incorporating paleoflood data in hydrologic hazard assessment.

Five study reaches along the Kern River were examined and data from 18 stratigraphic sites including soil stratigraphic information and radiocarbon ages were collected to document the sedimentological characteristics of the fluvial deposits. Soil properties and radiocarbon ages on detrital charcoal samples recovered from discrete stratigraphic units provide age information on the timing of deposition and provide a measure of landscape stability. A two-dimensional depth-averaged hydraulic model was used to make peak discharge estimates for paleofloods preserved in the stratigraphic record and for the non-exceedance bounds. These discharge estimates were made using topographic information derived from photogrammetry and LiDAR data. The hydraulic model was calibrated to high water marks documented for several extreme historical floods. Hydraulic model results were cross referenced with the stratigraphic data to allow estimates of peak discharge to be refined given changes in water depth, velocity, shear stress, and flow patterns over a range of channel roughness.

The stratigraphic record preserves evidence for at least two paleofloods. The magnitude of each paleoflood appears similar to or slightly smaller than the largest historical floods. Hydraulic model estimates suggest a range in discharge from about 115,000 ft³/s to about 225,000 ft³/s. Radiocarbon ages on correlative deposits indicate that these paleofloods occurred between 1025 and 1425 years ago. In addition to the evidence of paleofloods, there is also evidence of longer term stability provided by a soil formed on bedrock that provides the basis for a non-exceedance bound. Hydraulic model estimates of the peak associated with this non-exceedance bound suggest a range in discharge from 265,000 ft³/s to about 300,000 ft³/s.