FLOOD04278 - Laserfiche WebLink

Home Browse Search

6 M. Grimm er aL /Geomorphology ()() (1995) ()(J()....O{)() flood study is a sensitivity analysis of factors affecting the accuracy of discharge estimates (Jarrett and Malde, 1987; Jarrett and Way thomas, 1995). This type of anal- ysis provides information used to determine a range of possible discharges and the most probable discharge and associated hydraulic conditions corresponding to the PSI. A series of computational runs was made to test the sensitivity of the step-backwater analysis. In the vari- ous runs, different roughness coefficients, channel geometries, and gradients were considered. Contrac- tion and expansion coefficients were not varied from the standard values of 0 for contracting reaches and 0.5 for expanding reaches because recent studies indicate that these coefficients have relatively little influence on discharge estimates (Jarrett and Malde, 1987; J.O, Shearman, U.S. Geological Survey, pers. commun., 1992). Values of Manning's coefficient were selected onsite based on guidelines for selecting n values in natural channels and floodplains (Barnes, 1967; Jarrett, 1985; Hicks and Mason, 1991) and then varied over a reasonable range of values to assess the effect on com- puted discharge. Changes in channel geometry also were analyzed by running a scour-and-fill scenario for the Lower Bear Creek site. Because this site is confined by bedrock walls and channel bed, the thickness of the fill could be estimated relatively easily and removed from cross sec- tions in the discharge computations. 3.2. Geochronology Where possible, we sampled the flood deposits for organic materials suitable for radiocarbon analysis and used the characteristics of on-site vegetation to estimate minimum ages for floods. Radiocarbon samples con- sisted of allochthonous charcoal and wood that could have been reworked from older deposits. The samples thus provide a maximum limiting age for the associated flood sediments (Blong and Gillespie, 1978; Baker and Pickup. 1987). One radiocarbon sample was collected at each of the four primary sites. Although tree growth is sparse on the flood bars in Bear Creek basin, we sampled one tree at Lower Bear Creek and one tree at Turkey Creek. These trees were sampled with an increment borer to determine the approximate tree age by counting annual growth rings (Phipps, 1985). The trees provide minimum limiting ages for the flood deposits (Costa, 1978;Hupp,1988). 3.3. Coarse-sediment characteristics The grain-size distribution of sediments coarser than 2 mm was assessed at 19 locations throughout the Bear Creek basin. Twelve of these locations were paired sites immediately upstream and downstream from tributary junctions with Bear Creek. The Wolman (1954) sam- pling technique was used to measure the b-axis diam- eter of 100 clasts at each site. A sampling site was a single depositional flood bar, where present, or an in- channel bar and channel-bed sediment where flood bars were absent. The sampling sites are noted in Table 1. Sites were chosen to minimize human or land-use effects on in-channel grain-size distributions. The pres. ence of granitic clasts at all sampling sites indicated that lithologic control was not a limiting factor on grain size. 4, Results 4.1. Discharge estimation Table 2 summarizes estimated paleoflood dis- charges, largest indirect discharges, and competence calculations for the four primary sites, The indirect discharges in Table 2 for the Upper Bear Creek, Cub Creek, and Turkey Creek sites were determined by U.S. Geological Survey personnel immediately after each flood, using the slope-area method, because the gage was destroyed by the flood (Follansbee and Sawyer, 1948; station records for USGS gages 06711000 and 06710400), (We were unable to locate the original station records for the 1896, 1933, 1934, and 1938 floods on Lower Bear Creek.) Slope-area estimates of discharge generally are too large for high-gradient (>0.002) streams (Jarrett, 1986; Quick, 1991). In addition, if a flood in the range of 240 m' s -1 occurred at the Lower Bear Creek site in 1896, as indicated by the gage notes, flood deposits should verify this. Flood evidence, however. indicates a paleodischarge of approximately 113 m' s ", assuming that the top of the boulder bar represents the water-surface elevation. Table 2 clearly indicates the decrease in unit discharge with increasing elevation. Journal: GEOMOR Article: 368