|
<br />. hillslopes. In general, erosion potential increases as slope steepens. Steeper slopes usually require
<br />comparatively small amounts of rainfall before substantial erosion occurs, which depends on the type of
<br />soil, vegetation cover, and infiltration rate. For example, as little as lto 2 in. of rain in a few hours can
<br />produce hillslope erosion on bare, poony drained soils on steep slopes. Thus, a lack of rill erosion is a
<br />good indicator that intense rainfall is uncommon. Extensive rill and deep gully erosion are extremely
<br />common below 7,500 ft in Colorado. For example, gullies up to 6 ft deep formed on hillslopes in the Big
<br />Thompson River Basin during the rainstorm of July 31, 1976 (McCain and others, 1979) and up to 7 ft
<br />deep on a hard-surfaced road in neany level areas during the 1921 Penrose, Colorado rainstorm (U.S.
<br />Department of Agriculture, 1921).
<br />
<br />Flood Depth and Paleo stage Indicator (HWM-PSI) Relation
<br />
<br />One of the most critical factors related to the reconstruction and accuracy of paleoflood discharge is the
<br />relation between PSis and actual flood height or high-water marks (HWMs), In mountainous-river basins,
<br />the most commonly used PSI is the maximum height of bouldery, flood bars, which result from deposition of
<br />flood-transported sediments. At present, the elevation of tbe top of the flood bar is assumed to represent
<br />the minimum stage of flood waters responsible for depositing the sediments (Jarrell, 1987, 1990b; Jarrell
<br />and Waylhomas, in press; Grimm, 1993; Pruess, 1996; Brien, 1996). The best way to assess the
<br />validity of this assumption is to document and determine the relation between the elevation of PSis and
<br />. HWMs for floods in streams. Although documenting this relation has been a goal of my research for the
<br />past 10 years, few PSI-HWM sludies have been done because of the infrequent nature of flooding.
<br />However, in 1995, near to and record snowpack occurred throughout much of the mountains in Colorado,
<br />with much of the accumulation from March through eany June. The resulting near to and record runoff
<br />transported and deposited sediments as flood bars and slackwater deposits.
<br />
<br />On site visits were made to streams throughout Colorado during the high-flow season of 1995, Visual
<br />observations of flood-bar formation and flagging 1995 HWMs were made on many rivers in Colorado. In
<br />addition, substantial rainfall in April and May, at lower elevations in the foothills and eastem plains of
<br />Colorado, resulted in bankfull flows in foothills streams and moderate out-of-bank flooding along the South
<br />Plalle River, Peak flows in some basins had recurrence intervals of 50 years to more than 100 years.
<br />Substantial mobilization, transport, and deposition of channel sediments (sand, gravel, cobble, and
<br />boulders) as flood bars and slack-water deposits, After snowmelt runoff receded, extensive post-flood
<br />surveys were made. Sixty cross-section and 1995 water-surface elevation surveys of the new flood
<br />bars were made on 45 streams in Colorado and Nebraska. In addition, PSI-HWM data from several other
<br />large floods (eg., McCain and others, 1979; Jarrell and Costa, 1986) including the 1996 Glen Canyon Dam
<br />'~Iood release" in the Colorado River in Arizona (Anderson and others, 1996) also were compiled and
<br />incorporated into the data set. These streams have drainage areas that range from less than an acre to
<br />about 24,000 mi2 (Colorado River at Glenwood Springs; South Plalle River north of Denver; South Plalle
<br />. River at North Plalle, Nebraska). Stream gradient at these sites ranges from about 0.0015 to 0.35 fVft.
<br />Sediments on the surveyed flood bars ranged in size from sand-size particles to boulders greater than 10
<br />
<br />16
<br />
|