Ancient Floods, Modern Hazards: Principles and Applications of Paleoflood Hydrology

ANCIENT FLOODS, MODERN HAZARDS: PRINCIPLES AND APPLICATIONS OF PALEOFLOOD HYDROLOGY

P. Kyle House, Robert H. Webb, Victor R. Baker, and Daniel R. Levish, Eds., 2002, 385 pp., $75.00, hardbound, American Geophysical Union, ISBN 0-87590-354-1

The science of paleoflood hydrology has evolved dramatically over the past three decades. This volume, the outgrowth of the Second International Paleoflood Conference in 1999, consists of 21 contributions that provide a background to the field, an overview of principles and methods of paleoflood hydrology, and case studies that illustrate the application of a broad range of techniques for research in geomorphology and hydrology.

The origins of paleoflood hydrology can be traced to attempts to understand the magnitude of the large, catastrophic late Pleistocene floods in the Channeled Scabland of eastern Washington (Bretz 1929), and site-specific studies of flood effects on rivers, for example the early work of Jahns (1947) on the disastrous floods of 1936 and 1938 in the Connecticut River valley. Subsequently, Victor Baker and his students began a series of investigations designed to test the validity of these geomorphic techniques for extending modern streamflow records. As Baker notes in this volume, these early attempts were criticized for their lack of precision in determining both the magnitude of the flood event as well as the precise timing of prehistoric floods. While this book provides a brief review of the development of the concepts of paleoflood hydrology, its greatest contribution is to demonstrate the increased sophistication in the merging of geological and hydrological techniques in the analysis of paleoflood evidence. As this volume illustrates, the paleoflood record provides a powerful long-term database of flood magnitude and frequency that has important applications in fluvial geomorphology, flood hydraulics, flood meteorology, and river management.

Several chapters address one of the most important questions of paleoflood hydrology: How accurate are the geological indicators of paleostage when compared to more short-lived high-water mark indicators? Jarrett and England demonstrate that on high-gradient small watersheds, careful analysis of a variety of stage indicators can yield accurate discharge estimates. These stage estimates can then be used as a basis to model the three-dimensional flow at the peak discharge, as described in the chapter by Denlinger et al.

Slackwater deposits are impressive accumulations of fine-grained sediment in low-velocity zones adjacent to channels. Several contributions illustrate their value in building long-term flood chronologies, both along arid region rivers in the Southwest-for example, on the Verde River in Arizona (House et al.) and on the Paria River in Utah (Webb et al.)-but also on humid region streams in the Appalachians (Kite et al). The use of flood deposits preserved in caves that are hydraulically connected to river channels is an innovative use of this methodology (Springer), particularly in climatic settings where the preservation of slackwater deposits is problematic. Slackwater deposits may not be the most accurate paleostage indicators, but the stratigraphy of the deposits, particularly when analyzed on a regional basis, can provide a long-term view of flood frequency not otherwise available. Redmond et al. show that it is possible to assemble a long-term regional paleoflood record that can be correlated to climate on timescales that are important for flood risk analysis.

As slackwater deposits increase in elevation, they serve as a self-censoring indicator of the next largest flood, a fact that truncates the recording ability of the depositional surface, while at the same time providing an important time interval for estimating frequency. Blainey et al. demonstrate that combining this paleoflood data with traditional log-Pearson-type-III frequency distributions does in fact improve the accuracy of flood distributions when compared to estimates from gauging station records alone.

Levish analyzes self-censoring data developed from slackwater deposits, as well as what he terms "paleohydrologic bounds," in which the time period of nondeposition on a geomorphic surface can be incorporated into traditional flood frequency analysis, an artful treatment of the blending of paleoflood hydrology with stochastic hydrology. Interestingly, this approach developed out of the very real problem of attempting to define the risk analysis for Bureau of Reclamation dams in the western United States where the standard for design was the Probable Maximum Precipitation, a design that was economically unattainable. Determining the paleohydrologic bounds for these rivers allowed a more realistic approximation of flood frequency for extreme-magnitude floods.

The book concludes with a perspective on the geology and geography of floods by O'Connor, Grant, and Costa that categorizes the largest floods on Earth, which are the result of dam failures. The largest of these are the result of ice dams associated with the Pleistocene ice sheets that dramatically rearranged the preglacial river systems. In contrast, meteorological floods from rainfall are nearly two orders of magnitude smaller, with the greatest rainfall floods associated with the largest tropical river basins. Further analysis of more than 22,000 gauging records in the United States, an update of the envelope curve of maximum floods as a function of drainage area (Crippen and Bue 1977), demonstrates the nonlinear increase in discharge with drainage area, a function of the scale of storms in North America. When viewed from this perspective, it suggests natural limits to the size of floods on Earth.

The editors here have compiled a timely and comprehensive review of the field of paleoflood hydrology. The organization of the chapters is well thought out, and the multidisciplinary approach to the subject provides the reader a broad sampling of current research and applications. The individual chapters are well referenced and can serve as a starting point for more in-depth study. Without a doubt, the book will be an indispensable reference for students and professionals who are interested in understanding the hydrology and geomorphology of extreme floods.

[Reference]

REFERENCES

Bretz, J. H., 1929: Valley deposits immediately east of the Channeled Scablands of Washington. J. Geol., 37, 393-427.

Crippen, J. R., and C. D. Bue, 1977: Maximum floodflows in the conterminous United States. U. S. Geological Survey Water Supply Paper 1887, 52 pp.

Jahns, R. H., 1947: Geologic features of the Connecticut Valley, Massachusetts, as related to recent floods. U. S. Geological Survey Water Supply Paper 996, 158 pp.

[Author Affiliation]

Peter C. Patton is a professor of earth and environmental science at Wesleyan University in Middletown, Connecticut.