Linking bed morphology changes of two sediment mixtures to sediment transport predictions in unsteady flows

Abstract

Flume experiments were conducted to measure bed morphology adjustments in sand/gravel and sand/silt sediment mixtures during repeated hydrographs and to link these changes to sediment transport patterns over multiple time scales. Sediment composition and hydrograph flow magnitude greatly influenced channel morphology, which impacted sediment yield, hysteresis, and transport predictions. Bed load yields were larger and more variable for the sand/silt mixture, as gravel in the sand/gravel sediment inhibited grain entrainment, limited bed form growth, and acted to stabilize the bed. Hysteresis patterns varied due to bed form and surface structure adjustments, as well as the stabilizing effect of antecedent low flows. Using half the data set, a dimensionless fractional transport equation was derived based on excess shear stress. Dimensionless reference shear stresses were estimated in two ways: as bulk values from all transport measurements and by applying a separate limb approach in which values were estimated for each limb of each hydrograph. For the other half of the data set, transport predictions with the separate limb approach were more accurate than those from six existing transport equations and the fractional relationship applied with bulk reference shear stresses. Thus, hydrograph limb-dependent dimensionless reference shear stress links changing bed morphology and sediment transport, providing a parameter to improve transport predictions during individual flood events and in unsteady flow regimes. This approach represents a framework with which to develop site-specific transport relationships for varying flow regimes, particularly in cases where detailed bed morphology measurements are not feasible and heterogeneous sediment complicates bed structure over time.

Bed Load Sediment Transport and Morphological Evolution in a Degrading Uniform Sediment Channel Under Unsteady Flow Hydrographs

“…Further regression analysis on our data is used to develop a combined hydrograph parameter ξ that accounts for the relative influence of W k and Γ HG on measured sediment yields, similar to Waters and Curran () analysis (see section ). In our case, the most appropriate form of this parameter is ξ = W k Γ HG α , where exponent α = 0.2 provides the best overall correlation ( R 2 > 0.99) to the measured W t * values (Figure c), following the power law relationship: …”

“…The observed temporal variability in bed load sediment transport in response to flow hydrographs and zero sediment supply at the upstream boundary is characterized by differential transport rates, hysteresis patterns, and bed load yields during the rising and falling limbs under a wide range of hydrograph conditions. To account for this variability, Waters and Curran () proposed a modeling approach whereby the bed load transport is predicted separately, under equivalent discharge conditions, on the rising and falling hydrograph limbs. This is obtained through evaluation of a dimensionless transport rate W * , using the Einstein‐Parker dimensionless reference shear stress approach (Parker, ; Parker et al, ), which can be written for uniform sediments in the general form: where m and n are modified transport coefficients derived from nonlinear regression and τ br * is the critical Shields stress, corresponding to the dimensionless reference transport rate W r * = q b * /τ b *3/2 = 0.002 (e.g., Parker et al, ), with q b * being the Einstein () bed load parameter (equation ).…”

The impact of inter‐flood duration on non‐cohesive sediment bed stability

“…During floods a hysteresis loop often develops in sediment flux measurements, whereby different magnitudes of bedload flux are produced on the rising and falling limbs of hydrographs for the same flow magnitude (Reid et al, 1985;Church et al, 1998;Hassan et al, 2006;Waters and Curran, 2015;Mao, 2018). During floods a hysteresis loop often develops in sediment flux measurements, whereby different magnitudes of bedload flux are produced on the rising and falling limbs of hydrographs for the same flow magnitude (Reid et al, 1985;Church et al, 1998;Hassan et al, 2006;Waters and Curran, 2015;Mao, 2018).…”

The impact of hydrograph variability and frequency on sediment transport dynamics in a gravel‐bed flume

“…Excess shear (ratio of applied shear stress and critical shear stress: τ 0 / τ c ) for the hydrographs ranges between 1.7–2.9 and 1–1.8 for the D 50bulk and D 84bulk particle sizes, respectively. Each triangular hydrograph was approximated in a series of short steps, as has been done in previous experiments simulating unsteady flow (Hassan et al, ; Mao, ; Martin and Jerolmack, ; Waters and Curran, ). The Froude number in the experiments ranged from 0.4 to 0.8 (estimated based on measured flow depths and known discharges), which is within range of the prototype streams and previous experiments studying unsteady flow (Lee et al, ; Mao, ; Waters and Curran, ).…”

“…The bed was then slowly saturated and drained to promote initial settlement of the freshly placed sediment. A period of low flow, enough to mobilize sand fractions, with no sediment feed was then initialized to provide the channel with a flow history (Waters and Curran, ; Mao, ; Redolfi et al, ). This period lasted for approximately 8 h, and was considered complete when the sand particles had rearranged such that their mobility was limited in the channel (through visual observation) and there was negligible sediment appearing in the bedload trap.…”