H54A-01
Morphology and Hydraulics of Nine Sand-bed Fluvial Bifurcations from the Mossy Delta, SK: Implications for Their Stability
The distributary channel network of the Mossy Delta, SK consists of ~25 presently-active bifurcations created by channel splitting around river mouth bars during a 70 year history of delta growth. Detailed morphologic and hydraulic data from nine bifurcations are analyzed here to define the processes that determine their stability. Processes considered include bedload steering by adverse bed slopes, sediment and flow steering by secondary circulation, flow steering by inherited channel planform, and gradient advantage due to external boundary conditions. Stability of the bifurcations was determined from serial aerial photo analysis. Results indicate that the net topology and planforms of the bifurcations were set early during deposition of river mouth bars. Inherited alignment of a bifurcate channel thalweg with the main stem does not play a large role in subsequent bifurcation evolution; today roughly half of the side-channels take a greater proportion of the flow. After creation, bifurcate channels decreased in width by c. 15 %, with most of the reduction occurring in the first decade through bank accretion of scroll bars. Modern bifurcate channel depths and widths follow a hydraulic geometry scaling law (although they are wider and shallower than typical river channels), indicating a morphodynamically stable channel network. Most active bifurcations today are asymmetric; the average proportion of discharge through subordinate channels is 37 % with variation from 20 to 50 %. Local water surface slopes are flat approaching a bifurcation and steepen down the bifurcate arms, with the steeper slope in the shallower channel. All subordinate bifurcate channels possess morphologic ramps from the main channel with adverse bed slopes of 3-5 %. Although this suggests topographic steering of bedload is an important process in maintaining stability, it is not the only controlling process, because the subordinate ramp is shallower in c. 30 % of the cases.
H54A-02
Partitioning of Water Discharge by Distributary Channels in the Prograding, Wax Lake Delta, Coastal Louisiana, USA
How water and sediment is routed through distributary networks on river deltas is incompletely known and a topic of much active research. We have undertaken a study to determine the controls on partitioning of water and sediment discharge in distributary channels of the Wax Lake Delta and to connect these transport processes to the land building associated with the growth of islands that separate distributary channels from each other. Here we present first results from the field project that defines how water from the upstream primary channel is partitioned between the first set of five distributary channels. Measurements of water discharge and channel bathymetry were collected using a 22-ft research vessel equipped with an acoustic Doppler velocity profiler, a swath bathymetry profiler and dual differential GPS antennas. Wax Lake Delta is situated at the downstream end of Wax Lake Outlet, a man-made channel that diverts water and sediment from the lower Atchafalaya River, roughly 20 km upstream from Morgan City, LA. The subaerial delta has been building out into Atchafalaya Bay since roughly 1973 with a delta-front advance rate of about 0.27 km/yr. Associated with this growth has been development of a distributary network of channels that continues to evolve as the delta progrades seaward. Measurements collected in May, 2007 define properties of the upstream channel and the first set of five distributary channels. Characteristic width, depth and water discharge for the upstream channel are 420 m, 21.2 m, and 2900 m3/s. Characteristic values for width, depth and water discharge for the five distributary channels are 1) 270 m, 6.7 m, and 310 m3/s, 2) 300 m, 6.5 m, and 350 m3/s, 3) 650 m, 6.8 m, and 820 m3/s, 4) 395 m, 6.5 m, and 560 m3/s, and 5) 440 m, 6.0 m, and 440 m3/s. These data highlight a number of interesting points regarding the initial set of bifurcations. First, the transition from one to five channels is associated with a two-thirds reduction in characteristic flow depth and one-third reduction in characteristic flow velocity. Measured discharge in the five channels sums to 86% of the upstream channel. Based on field observations we propose that the remaining 14% is associated with shallow flow onto and across the weakly emergent islands. There is very little difference in characteristic channel depth even though widths of the five channels vary by greater than a factor of two. There is also very little difference in the spatially average velocity for the five channels; velocities range between 0.17-0.22 m/s. All of these channel properties will be discussed in the context of sediment transport and sedimentation within this active delta.
H54A-03
Avulsion and Bifurcation Stability; the Relative Roles of Internal and External Processes in a Holocene Deltaic Distributary Network
Avulsion, the abandonment of all or part of a channel belt in favor of a new course, and hence bifurcation stability, is controlled by both internal and external processes. External controls comprise the boundary conditions that are unaffected by local evolution (climate, base level, tectonics). Internal controls evolve due to internal system dynamics (e.g. bar and meander dynamics, delta lobe switching). The Rhine-Meuse delta is without doubt the best studied delta regarding Holocene avulsion processes. Based on studies of this delta, an overview is given of 1) external and internal processes influencing different aspects of avulsion and bifurcation stability, 2) the relative importance of these processes in time and space, and 3) observations that can be used to differentiate internal and external causes of avulsion. The following avulsion parameters are quantified, and their external and/or internal nature are discussed: period of activity of channel belts (period between beginning and ending sedimentation of the river channel), interavulsion period (time between successive avulsions of a channel belt), avulsion duration (time between initiation of a new channel and complete abandonment of the previous channel=bifurcation stability), avulsion frequency (number of avulsions per time interval in a given area), and avulsion location. In the Rhine-Meuse delta, the period of activity of the Holocene channel belts varies considerably, but shows no significant trend over time, whereas external factors changed. This suggests that this parameter is mainly internally controlled. The average interavulsion period increased from 8000 cal yr BP to 2800 cal yr BP, and decreased since then. Significant fluctuations occurred on a shorter time scale. A maximum variability in the interavulsion period occurred between 3200 and 1800 cal yr BP. This was a time, when large channel belts gradually came into existence. During this period the avulsion frequency also reached a maximum. The long-term increasing trend in interavulsion period is related to the decreasing rate of sea level rise. The decreasing interavulsion period since 2800 cal yr BP can be explained by increased discharge and sedimentation, resulting in an increase in the number of avulsions. The avulsion duration fluctuates between less than 200 and 1250 cal years and averages 335 cal years. The avulsion duration shows no significant trend over time and remained constant until at least 1900 cal yr BP. During an avulsion sequence, avulsion sites shift progressively upstream with a simultaneous decrease in interavulsion period. The sequences can be explained as a result of continued growth of alluvial ridges and increasing cross-valley slopes upstream of avulsion locations. New channel belt segments down-valley from avulsion locations have low natural levees and a low probability of avulsion; therefore avulsion sites tend to shift upstream until the apex of the delta is reached. The next avulsion can then occur far downstream again (Mackey & Bridge, 1995). Seven avulsion sequences may be present in the dataset of the Holocene Rhine-Meuse, suggesting a periodicity of ~500-600 yr. Each avulsion sequence coincides with a peak in the avulsion frequency. The periodicity of ~500 years in the avulsion frequency is proposed to be internally controlled. However, there also is a long-term trend in the avulsion frequency that can be related to sea level rise, and (after 3000 C-14 yr BP) to climate change and human influence. Avulsion locations are related to the external factors relative sea level rise, local tectonics, and changes in discharge and sediment load.
H54A-04
Experiences in 2D Numerical Modeling of River Bifurcations
We present our experiences with morphological computations using a 2D depth-averaged finite-difference model for river bifurcations. These computations reveal new details of the importance of a proper computational gid and a proper representation of transport processes of graded sediment. The basic structure of the computational grid, irrespective of fine-tuning, determines the possibility to reproduce flow separation at sharp bank angles as well as the errors due to grid distortion in the key area of the bifurcation. Inclusion of floodplain flows requires that the grid at the bifurcation be split into subgrids with corresponding domain decomposition techniques to solve the equations. Application to the Rhine bifurcation at Pannerden casts doubt on the common wisdom that the active-layer thickness for graded sediment is equal to half the dune height, or somewhat larger due to sporadic deeper troughs. A larger active layer must be adopted on larger time scales, because the thickness depends on all kinds of local bed level fluctuations generated by discharge variations. Examples are the sand waves generated by fast changes due to the flooding of floodplains, and the stage-dependent variations of the transverse bed slope in river bends.
H54A-05
Controls on the Alluviation of Oxbow Lakes by Bed Load as Observed Along the Sacramento River of California
As the products of meander cutoff that are widespread within many floodplains, oxbow lakes can affect the ability of rivers to migrate across their valleys, as well as physical and chemical exchanges between the river and floodplain environment. The particular functions of the oxbow lake are determined by the manner it is filled by sedimentation. Although the alluviation of oxbow lakes has been observed in natural settings and generalized by means of rules in planform evolution models, no theory exists to explain how oxbow lakes are filled because the controls on the process have not been widely studied or physically interpreted. Utilizing existing theory and field data from lakes of the Sacramento River, we examined the controls on oxbow alluviation by bed load and found that the transport of bed material through a channel abandoned by cutoff is highly sensitive to the orientation of the abandoned-channel entrance. In particular, the diversion angle, the angle between the approaching active- channel flow and the abandoned-channel entrance, is a direct control on discharge through the abandoned channel, and thus can significantly reduce boundary shear-stress and limit the transport capacity of bed load. The higher the angle, the more greatly reduced is the capacity to transmit bed material, and the more quickly the channel is hydraulically disconnected as diverted bed load rapidly aggrades the channel entrance. In contrast, the lower the angle, the longer the duration the channel remains hydraulically connected, and the more likely it will experience filling and narrowing by bed load because sufficient flow allows for the downstream and transverse transport of bed material. Our findings from the Sacramento River compare well to observations from other large meandering rivers and may explain why some lakes are terrestrialized within decades of cutoff, whereas others remain as open-water habitat for considerably longer.
H54A-06
The Interplay Among Vegetation, Bifurcations, Avulsions, and Channel Morphology in an Experimental Channel
A series of laboratory experiments investigating the interactions between vegetation and braiding demonstrate the interplay between bifurcations, avulsions, channel morphology, and sediment storage and release as the system evolves from unvegetated braided to single-thread with a vegetated floodplain. Vegetation led to net storage of sediment and super-elevation of channels in the experiments. This unstable condition caused channels to avulse, steering the flow into nearby steeper floodplain channels. In contrast with the braided system, where channel switching is a nearly continuous process with persistent flow shifting among multiple channels following slight gradient differences, the vegetated state produced less frequent and more distinct avulsions. Because of the added difficulty of cutting a new path across vegetated surfaces, a channel avulsion required a larger super-elevation which led to abandonment of the old channel and relatively complete diversion of flow to the new one. Channel abandonment at bifurcations under low flow conditions played a key role in vegetation expansion and the incorporation of channels into the evolving floodplain. The steady state system produced quasi-stable states: a dynamic exchange of flow and sediment occurred between anabranches separated by a stable vegetated island. At different times one channel contained the majority of the flow. As the dominant channel filled in with sediment, flow was diverted to the secondary channel and the secondary channel deepened. http://www.geo.umn.edu/orgs/seds/
H54A-07
The Effect of Alternating Bars Migration on River Bifurcation Dynamics
Recent theoretical analysis, field and laboratory observations pointed out that fluvial bifurcation show an intrinsic instability, leading to the establishment of an unbalanced flow and sediments distribution in the downstream branches. The existence of equilibrium configurations has been proved, which mainly depend on the hydraulic and morphologic conditions of the upstream flow. However, flow and sediment transport in braided networks are highly unsteady, so that the bifurcation can hardly reach an equilibrium configuration. One of the main causes of temporal fluctuations is the migration of alternate bars in the upstream channel, that can affect and control the flow partition in the distributaries. We analysed the bar – bifurcation interactions by experimental and analytical investigations. We performed a set of flume experiments on a Y shaped fixed banks and movable bed bifurcation. Laboratory results show that bar formation in the upstream channel perturbs the discharge distribution with a series of fluctuations strictly related to the period of bar migration. Four different behaviours have been identified, characterised by small perturbations of the equilibrium state (balanced or unbalanced), by the occurrence of large fluctuations or by the closure of one of the distributaries. The character of the bifurcation is controlled by the amplitude and speed of alternate bars that directly influence the amplitude and period of discharge oscillations. Consequently, at large values of the aspect ratio (high bars) and low sediment mobility (slow bars) the bifurcation dynamics is likely to be dominated by bars migration. Extending the one-dimensional model proposed by Bolla Pittaluga et al. (2003), we introduce the effect of bars migrating in the upstream channel. In the present model, the bifurcation is forced with spatial crosswise fluctuations of feeding conditions, in order to reproduce the transverse distribution of sediment and water of an alternate bar pattern as predicted by the weakly nonlinear theory of Colombini et al. (1987). In this way, the model reproduces bed perturbation and discharge fluctuations at the inlet of the downstream channels. For different bar characteristics, correspondent to different flow conditions in the upstream channel, the model is able to qualitatively reproduce the four behaviours detected in the experiments. The obtained results underline the complexity of the bar-bifurcation interaction, due to their similar evolution time scales.
H54A-08
Flow Structure in a Bifurcation: CFD modeling and Validation
There has been a recent growth in scientific interest concerning the role of river bifurcations as key nodes within fluvio-deltaic systems and in braided rivers, with the realization that they are critical in the very process of braiding itself. The geometrical characteristics of the bifurcation largely control flow and sediment partitioning between the two distributaries, the inherited downstream flow structure and thus potentially evolution of the subsequent bifurcation morphology. Furthermore, it has been shown that certain geometrical configurations induce flow instabilities that allow formation of a downstream mid-channel bar through symmetric forcing. However, to date, we have a poor process understanding of flow dynamics within such bifurcations and their downstream distributaries. In order to begin to address these issues, here we examine flow structure in a bifurcation through application of a time-averaged Computational Fluid Dynamics model, using an RNG k-å turbulence model, a non-equilibrium wall treatment and a second order accurate numerical solver. The domain was meshed using a multi- (two-) block mesh approach where the inflow channel comprises two connected meshes, which divide at the bifurcation, and become symmetrical in the two distributaries. This thus provides a good boundary representation in the distributaries, as well as a better flow field simulation in the vicinity of the bifurcation, where flow partitioning is sensitive to the mesh design. The numerical experiment replicates concurrent research where a bifurcation has been examined in an identical-geometry physical model that provided the boundary conditions and model validation data. This paper will present details of i) the model and its validation, and ii) subsequent numerical experiments where modifications were made to the distributary channel width, slope and bifurcation angle, that aimed to gain an insight into the dominant geometric characteristics of the bifurcation and their influence on flow structure and potential bifurcation stability.