8274 - Laserfiche WebLink

Home Browse Search

flume experiments and field observations are used to infer two different phases of transport: initial motion and significant motion. The initial motion phase denotes the onset of bed load transport at the critical dimensionless shear stress z*r. In this phase, very few particles on the bed are moving and bed load transport rates are very low (Wilcock and MacArdell, 1993; Andrews, 1994). This phase is nonetheless important for maintaining micro-scale habitats because it marks the point at which coarse particles start moving and fine sediment begins to be flushed from the bed (Kondolf and Wilcock, 1996; Wilcock et al., 1996). The lower limit for this phase is a matter of some debate. Data compiled by Buffington and Montgomery (1997) suggest that particles equal to the median grain size, D50, begin moving at a value of z*r = 0.030. Milhous (1998) proposes a lower value of 0.021, while flume experiments by Wilcock (1998) indicate that the r'c for gravel particles varies from 0.01 to 0.045, depending on the amount of sand on the bed. In the present study, we varied r', from 0.030 in strata 7-11, to 0.025 in strata 6-3, and 0.020 in strata 2. Our reasons for adjusting the value of r`,- are discussed later, in the section on transport thresholds. The second transport phase, significant motion, is characterized by continuous movement of most all particles on the bed. A lower limit for this phase is not well established. In a previous study, Pitlick et al. (1999) reasoned that significant motion should occur at flows approaching bankfull, since these flows shape the channel. Using data from reaches near Grand Junction, they found that bankfull flows produced an average r* of 0.047, which is about 1.5 times the assumed value of r`C (Pitlick et al., 1999). We followed a similar convention in this study, although our field data from the reaches below Westwater Canyon (strata 6-2) suggest that that bed-Toad transport stages corresponding to 1.5f, may occur at flows less than bankfull. There are two other important points to note about (4) and its application. The first point concerns whether the D50 is representative of the bed as a whole, and the implication that if the D50 begins to move, all sizes begin to move (a condition termed equal mobility). We will not review this debate here, except to note that in most gravel-bed rivers, framework grains of most all sizes are entrained by flows ranging from about 1/2 to 2/3 of the bankfull discharge. The debate over equal mobility has been resolved by the experiments of Wilcock and McArdell (1993), who showed that, at low shear stresses (0.030 < r' < 0.045), the number of coarse particles entrained from the bed is very low in comparison to the number of fine particles entrained, resulting in a bed load that consists of mostly fine particles. At higher shear stresses (rk > 0.060), fine and coarse particles are entrained at about the same rate, resulting in nearly equal proportions of fine and coarse bed load. The range in shear stress over which this transition occurs is roughly the same as that defined by our initial motion and significant motion transport stages. The second point to note about eqn. (4) is that the shear stress r should be formulated in a way that is appropriate for the particular problem. In the general case, the near-bed shear stress is i=PgRS, (5) where R is the hydraulic radius and Sf is the friction slope. In channels with high width-to-depth ratios, R is nearly equal to the mean depth, h, hence these variables are substituted for each other. An important practical issue with respect to eqn. 5 concerns the formulation of the teen Sf . At the scale of individual features such as pools, riffles and runs (distances of, say, less than 1000 m), natural undulations in bed topography and channel width cause the flow to accelerate or decelerate, which increases or decreases the shear stress produced solely by the downstream component of the 17