CASE STUDIES SHOWING THE UTILITY OF A VARIETY OF METHODOLOGIES TO CHARACTERIZE THE MAGNITUDE AND FREQUENCY OF PALEOFLOODS IN THE ASSESSMENT OF FLOOD HAZARDS (Invited Presentation)

For the last 20 years, the Bureau of Reclamation has been using a variety of methodologies to evaluate extreme floods in the assessment of flood hazard for dams in their inventory. The primary focus of the studies has been on developing magnitude and frequency information of paleofloods from slackwater deposits and non-exceedance bounds. Evidence of paleofloods is commonly preserved in the Holocene geologic record, however, the methodologies needed to accurately estimate the age and discharge for these events can be less than straightforward. Problems include developing good age control on flood deposits and terrace surfaces, interpretations of and differentiation between fluvial deposition and soil horizon development, the accurate depiction of the channel geometry to estimate peak discharges, and the data requirements and uncertainties associated with various hydraulic modeling techniques.

Examples of approaches used to help resolve some of the age-related issues include using information derived from old written or verbal accounts and local historical societies, historical artifacts, archaeological information, dendrochronology and palynology, micro-charcoal analysis, and soil stratigraphy. Results of standard radiocarbon analysis remains one of the most useful techniques for determining the age of Holocene deposits with the usual contamination and inheritance issues, but results seem to be improved by paleobotantical identification of samples before submission for age analysis. Peak discharge estimates are commonly made using a variety of hydraulic models and topographic information derived from direct channel surveys, photogrammetry, or LiDAR data. However, making discharge estimates for prehistorical floods in areas where the channel has been modified historically can be difficult. Discharge estimates in these situations have been successfully made using pre-modification topography derived from older historical photographs or plane-table topographic surveys. Estimates made without good high water marks or with a poor understanding of peak stage associated with flood deposits can be refined in models where cross reference of model output in the form water depth, velocity, shear stress, and flow patterns are compared to the stratigraphic data and geomorphic evidence.