|
<br />318
<br />
<br />PALEOFLOODS
<br />
<br />"
<br />
<br />Church (1978), this approach to paleoflood analysis can
<br />be subject to large errors.
<br />The twofold division of paleoflood studies can also be
<br />considered from an ecogeomorphic perspective, Low mag-
<br />nitude, high-frequency flow events (type 1) can be studied
<br />through the use of tree ring widths as proxy records of
<br />streamflow (Stockton and Boggess, 1980, 1983; Smith and
<br />Stockton, 1981), High magnitude, low-frequency flow
<br />events (type 2) leave their direct imprint on bottomland
<br />vegetation communities. Chapter 20 by Cliff R, Hupp,
<br />"Plant Ecological Aspects of Flood Geomorphology and
<br />Paleoflood History," provides a comprehensive review of
<br />type 2 flood effects on plants. Some settings, such as
<br />Passage Creek in northwestem Virginia, provide very de-
<br />tailed ecogeoDlorphic reconstructions of floods. In general,
<br />tree ring chronologies allow paleoflood record lengths of
<br />400 to 500 years or less, as limited by the ages of extant
<br />floodplain or channel trees.
<br />The most accurate and detailed reconstructions of large
<br />paleofloods over long time periods are achieved in studies
<br />of certain stable-boundary fluvial reaches characterized by
<br />slaclcwater deposits and paleostage indicators (SWO-PSI),
<br />Chapter 21 by R. C. Kochel and V. R. Baker, "Paleoflood
<br />Analysis Using Slackwater Deposits," summarizes recent
<br />experience with SWOopS! studies. The chapter discusses
<br />paleoflood slaclcwater deposition at tributary mouths, an
<br />especially conunon field situation. Emphasis is placed on
<br />deposit interpretation for preservation, completeness,
<br />sedimentology, channel stability, and paleoflood ages, The
<br />processes of slackwater sedimentation are discussed in
<br />Chapter 8,
<br />The most important use of SWO-PS! studies is in pa-
<br />leoflood frequency analysis. Chapter 22 by R. C. Kochel
<br />entitled "Extending Stream Records with Slackwater Pa-
<br />leoflood Hydrology: Example from West Texas" illustrates
<br />this application for the Pecos and Devils rivers. The sys-
<br />tematic flood record for the Pecos illustrates the classic
<br />dilemma posed by one or two unusually large outliers in
<br />an annual flood series. In this case, the recurrence interval
<br />for the great flood of 1954 was specified by incorporating
<br />SWO-PS! paleoflood data into the flood frequency analy-
<br />sis.
<br />SWD-PS! paleoflood hydrology has benefited from two
<br />areas of recent technological advancement The first area
<br />is geochronology, which has yielded more precise age
<br />determinations for a variety of datable materials in flood
<br />deposits. Kochel and Baker in Chapter 21 discuss this
<br />topic, but it is certain that additional advances can be
<br />expected in the near future. For example, thermolumi-
<br />nescence dating (Wintle and Huntley, 1982) might be applied
<br />to windblown quartz-rich sediments that are sometimes
<br />interbedded with slackwater deposits in arid environments.
<br />Radiocarbon dating by tandem accelerator mass spectrometry
<br />(Taylor and others, 1984) has immense potential for dating
<br />paleofloods,
<br />
<br />1...,,-
<br />Ii
<br />I..""....
<br />i
<br />
<br />r-
<br />
<br />'f',
<br />
<br />/"',..-
<br />
<br />A second important area of technological advancement
<br />bas come in the use of hydraulic flow modeling to calculate
<br />paleodischarges. Chapter 23 by Jim E. O'Connor and Robert
<br />H. Webb, "Hydraulic Modeling for Paleoflood Analysis,"
<br />provides an up-ta-date review of step-backwater discharge
<br />calculations, as used in paleoflood hydrology. New computer
<br />flow models make these calculations easy to perform as a
<br />part of SWOopS! paleoflood investigations. Because hy-
<br />draulic calculations can be performed independently of the
<br />high-water indicator surveys, this approach allows a better
<br />means of specifying possible error in the analysis. The
<br />field problems of SWOopS! geomorphology and interpre.
<br />tation can be separated from analytical problems in specifying
<br />the hydraulics of a study reach,
<br />Chapter 24 by Webb, O'Connor, and Baker describes
<br />an in-depth paleoflood hydrologic study of the Escalante
<br />River in south-central Utah. Typical of many streams in
<br />the arid or semi-arid westem United States, Escalante River
<br />floods have heretofore been documented by a fragmentary ,
<br />short-term systematic record. In contrast, Chapter 24 pro-
<br />vides historical data, tree ring data, and SWD-PSI data
<br />for as many as 20 paleofloods. Hydraulic flow modeling
<br />procedures, described in Chapter 23, are used to calculate
<br />paleoflood discharges, Of particular interest are the silt
<br />lines that are interpreted as precise paleostage indicators
<br />for ancient floods. Channel stability for bedrock reaches
<br />of the Escalante River is demonstrated with historic pho-
<br />tographs,
<br />Of considerable importance for the Escalante River is
<br />the apparent clustering of flood events. Large floods occurred
<br />approximately 1000 years ago, about 500 years ago, and
<br />in historic times. The largest flows in the last 2000 years
<br />were comparable to the largest floods observed on similar-
<br />sized drainage basins in the region, The paleoflood and
<br />historical flood data were incorporated into a flood frequency
<br />analysis according to the standard procedures adopted by
<br />U,S. governmental agencies (U,s. Water Resources Council,
<br />1982).
<br />Chapter 24 underscores the importance of documenting
<br />all potential error sources in paleoflood hydrologic inves-
<br />tigation. As such studies move from the scientific to the
<br />engineering arena, there will be increased emphasis on the
<br />accuracy of risk evaluation hased on paleoflood data.
<br />Hosking and Wallis (1986) demonstrated that a single,
<br />inaccurate estimate of a large paleoflood may be of little
<br />use in improving an estimate of flood frequency, In contrast,
<br />Stedinger and Cohn (1986) find that, with proper structuring
<br />of the data, the mere knowledge of flood exceedence or
<br />nonexceedence of a censoring level may result in consid-
<br />erable improvement of flood frequency estimates. As shown
<br />in this section, SWD-PSI paleoflood studies can yield mul-
<br />tiple paleoflood estimates that are quite accurate both in
<br />age and magnitude, Stedinger and Baker (1987) consider
<br />these qualities most important in achieving useful flood
<br />frequency information from paleoflood studies,
<br />
|