Laserfiche WebLink
weight of the water (pghS). In this case, Sf is estimated by solving the one-dimensional equation for <br />gradually varied flow: <br />z <br />Sf = dH = - d (z + h + °- (6) <br />dx dx t 2g <br />where H is the total energy, z is the bed elevation, u is the mean velocity in the downstream (x) <br />direction. Equation 6 is solved using the step-backwater method, a trial-and-error procedure that is <br />described in many hydraulics texts (e.g. Henderson, 1966; Dingman, 1984), and can be easily <br />programmed on spreadsheets. In this study, the step-backwater method was used with cross <br />section and water-surface elevation data from short (--500 m) reaches to develop relations between <br />r" and Q, which were then used to estimate discharges for initial motion and bankfull flow. <br />At scales much greater than individual pools and riffles (> 1000 m), the variations in channel form <br />have much less of an effect on the boundary shear stress. In this case, the flow can be considered <br />uniform, and the water-surface slope S, can be used in place of Sf in (6). This assumption was <br />adopted herein to calculate reach-average values of r*. <br />To estimate the discharge required to initiate bed load transport (termed the critical discharge, Q,) <br />we assume uniform flow, and combine the continuity equation (Q = w h u, where w is the flow <br />width), with the Manning equation (u = h73 Sf 1'2/n, where n is a resistance coefficient), to get <br />w h5/3 S112 <br />Q = C f- <br />n <br />(7) <br />Here, h, is the depth required to produce the assumed critical dimensionless shear stress at each <br />cross section, calculated from (4) and (5) using the reach-average D50. To estimate the bankfull <br />discharge Qb, we used the same equation (7) with the bankfull depth hb and slightly lower values of <br />Manning's n. On the basis of the results obtained in calibrating the step-backwater model, we <br />selected values of n ranging from 0.035 in upstream reaches to 0.025 in downstream reaches. <br />Equation (7), after some algebraic manipulation involving the term S, can be rearranged to give a <br />uniform-flow equation for the average boundary shear stress as a function of discharge: <br />r = PghS = Pg(Qn)0.6S0.7 <br />W <br />(8) <br />The above equation indicates that for constant n and w, r increases to the 0.6 power of Q and the <br />0.7 power of S; graphically, eqn. 8 plots as a concave-down curve, with a slope that gradually <br />decreases as Q increases. However, if any of the variables in (8) change with discharge, then the <br />relation between r and Q may differ from the uniform-flow case. For example, if S increases with <br />Q, the relation formed by (8) will be steeper and more nearly linear; conversely, if S or n decrease <br />with Q, or if w increases, the relation will be flatter and more concave. Reach characteristics <br />determine whether S, n or w change much with discharge, but whatever the case, the shape of the Q- <br />r relation gives an indication of the potential change in shear stress for a given increment of water <br />discharge: a near-linear relation indicates that r increases in proportion to Q, whereas a concave <br />relation indicates that r increases more slowly as Q increases. Equation 8 thus represents a base <br />case against which we can compare modeled discharge-shear stress relations in short reaches where <br />the assumptions of uniform flow, or constant w, n and S, generally do not hold. <br />18