FLOOD05412 - Laserfiche WebLink

Home Browse Search

46 GEOMORPHICALLY EFFECTIVE FLOODS Geomorphic work in fluvial systems has been variously defined to be represented by the rate of sediment movement [Wolman and Miller, 1960], or as the mass of sediment transponed through a venical distance in unit time [Caine, 1976]_ Depending on the interpretation of what constitutes geomorphic 'work', and the relative magnitude of forces acting on the landform during the time of interest, slow and persistent processes that generate low forces may appear to predominate [Wolman and Miller, 1960; Andrews, 1994]. or large and rare floods that generate large forces may be considered to be most significant [Baker, 1977]. The recognition that some really large, rare floods may not have long-lasting effects, or cause long-term changes in channel and valley morphology [Costa, 1974; Moss and Kochel, 1978; Huckleberry, 1994] led to the realization that the absolute magnitude or force of a geomorphic process is not the sole factor responsible for the resulting landforms, nor their perseverance. Other contro1\ing factors include landsurface resistance [Bull, 1979; Graj, 1979; Brunsden, 1993]. the frequency and ordering of effective processes [Beven, 1981]. and the rate of recuperative processes fo1\owing formative events [Costa, 1974]. This holistic view of what constitutes an effective event in geomorphol- ogy is we1\ captured in the benchmark paper by Wolman and Gerson [1978]. There are, at present, no simple measures of flow and effect that have been used consis- I tently in describing the interaction of floods and the landscape. --- 1.1 The significance of flood-flow duration Our investigations of the recent floods from two small dam failures in Washington and Oregon inspired us to furtheJ. consider the imponance of flood-flow duration with respect to the geomorphic effectiveness of floods. Although these floods had extremely high instantaneous values of shear stress and stream power, they produced few or no geomorphic changes in downstream valleys or channels. The purpose of this paper is to document the role of flood duration as an obvious, but often ignored, critical factor_ Flood duration can affect geomorphic response to large flows in several ways. Long flow duration may be necessary to saturate channel banks before they will fail, or aid in the wetting and subsequent expansion of floodplain soils, with concomitant reduction in shear strength. Some finite amount of time may be necessary to break down floodplain vegetation or erode through the cohesive, root- strengthened top strata, after which erosion of less-cohesive substrata can proceed more rapidly. Also, sediment entrained from hillslopes or channels requires time to be transponed onto floodplain surfaces, especially if it travels as bedforms. Flow duration, in addition to flow magnitude and frequency, stream power, resistance of the land surface, and the restorative and recuperative processes between effective eyents, determines whether a large discharge event is geomorphically effective. Flow duration can be a key to understanding how floods with lower values of peak discharge, shear stress, or stream power, can have greater geomorphic impact in some alluvial channels than floods with larger instantaneous values. 2. INSIGHT FROM FLOODS FROM THE FAILURE OF SMALL DAMS Floods resulting from the failure of small dams in upland areas can offer a unique perspective into the influence of a high-magnitude event on steep channels and floodplains in small basins. Dam failures in upland areas involve a precisely known volume of water being intro- duced to a channel at a point location. Dam failures also offer a mechanism for the creation of floods far larger than possible from snowmelt or rainfall-runoff, and that may be unprecedented in the recent or pasi geological history of the basin [Jarrett and Costa, 1986]. Such small-dam failures occur frequently, and many go unreponed in the literature. Recent documented examples include rainfall- induced failure of seven eanhfl1\ gravity dams in 1977 near Johnstown, Pennsylvania [Hoxit and others, 1982]. and three eanhfl1\ gravity dams in 1989 at Fayetteville, North Carolina [Mason and Caldwell, 1992]. The recent failures of two small upland dams in Washington and Oregon present the opportunity to evaluate the role of stream power and flood duration on geomorphic effectiveness in downstream channels and floodplains. Both dams failed rapidly and nearly instantaneously released their stored water down small, steep, upland channels and floodplains. 2.1 The failure of Reservoir No.3" Cenrralia, Washington Reservoir No.3 is a small concrete-lined water-supply reservoir for the city of Centralia, Washington (Figure 1). On Oct. 5, 1991, the bedrock hillslope under the southwest side of the reservoir suddenly failed, and instantaneously released 13,250 m' of water down a small steep valley that led to the eastern edge of the city of Centralia. Two houses were destroyed, four city blocks were flooded, and 400 people were evacuated (Figure 2) [Costa, 1994]. The reason for the failure is believed to have been a landslide in the silty sandstone bedrock beneath the reservoir, caused by some combination of (a) seepage from .-..