H43A-0350 1340h
Probability Distribution of Bed Elevation in a Bedform-Dominated Large River
Recent research on morphodynamic models for bedform-dominated rivers has strongly indicated the need to improve our knowledge of the nature of dune height and trough-scour depth variation in time and space. The probability distribution of dune-bed elevation (Ps) is a key element of these new depth-continuous models (i.e., with no distinct active bed layers) in which elevation-specific formulations for sediment transport were developed. For now, these models are theoretical frameworks and there exist only limited experimental data to implement them. This paper presents much needed data from the field. We surveyed a set of two km-long bed profiles of the Mississippi River where flow depth averaged 20 m. One profile followed the sailing line at a major river crossing and the other was located in the center of the channel. Bed profiles show multi-scale bedforms. We constructed Ps curves for the entire bed profiles and for selected stretches. We find that Ps curves are asymmetric relative to mean bed level, with a larger probability of deep bed elevations. This is comparable to previous experimental results at relatively high sediment-transport stage. Otherwise, Ps curves of entire profiles reflect the tri-dimensionality of macro-scale bedwaves that scale with channel width (1 km) and planform. The upper-limit elevation always approximates the ratio of observed mean dune height over mean flow depth. This ratio is lower for the deep dunes than for those migrating on point bars. Comparative analysis of these Ps curves with those constructed from USACE 2001 data at the same site brings some insight on sediment deposition and erosion over time. Our preliminary results also show the potential of Ps curves for studies integrating multi-scale bedform characteristics. Future research will include relating Ps curves to vertical sorting, velocity profiles and estimates of sediment transport from acoustic measurements, as well as associated stratigraphy.
H43A-0351 1340h
Toward a theory of hydraulic geometry: the principle of flow resistance for the fluvial system
One of the primary obstacles to the development of a theoretically based expression of channel scaling, or hydraulic geometry, is the identification of an appropriate formalism to adapt one-dimensional models to the description of a three-dimensional phenomenon. Within the context of the fluvial system, the most nearly stable, hence most likely channel configuration is related to the principle of flow resistance maximization. The flow resistance for the system comprises three components: grain-scale flow resistance, bedform-scale form resistance, and lastly the reach-scale form resistance, which is the primary difference between system-scale flow resistance and the more usual representations. Stream table experiments designed to test the theoretical implications associated with the maximum flow resistance criterion demonstrate that the system scale flow resistance responds as predicted. When the channel banks are as erodible as the bed, the reach-scale flow resistance is the dominant component of the system adjustment, resulting in a functional relation between the average water surface slope along the channel thalweg and the ratio of the sediment supply and the imposed discharge. This is equivalent to the well known and generally accepted graded relation between channel slope and sediment supply. When the banks are fixed, the channel slope remains nearly constant - as does the cross section shape - for a range of sediment supply rates. In this case, equilibrium seems to result from a textural modification of the bed surface and thus the grain scale or bedform flow resistance. These results are consistent with the concept of system-scale flow resistance being the key to understanding channel stability, and they indicate that the previous hypotheses, such as slope minimization, are too limited in the range of adjustments that they embrace.
H43A-0352 1340h
Sub-Reach Scale Morphological Interpretations from DEM Differencing: Accounting for DEM Uncertainty
Using repeat ground-based topographic surveys and digital elevation model (DEM) differencing to infer reach-scale sediment budgets has become a popular monitoring tool in fluvial geomorphology. However, few studies have used DEM differencing to infer more detailed sub-reach and meso-scale fluvial processes and rates. In part, this might be attributed to the recognized importance of uncertainty in representing individual surfaces and inferring `real' changes from differences. Several recent investigators have suggested applying a minimum level of detection ($_{min}$LOD) (between 10 and 30 cm), below which `real' changes are indistinguishable from noise. We present a more detailed assessment of uncertainties below typical $_{min}$LOD thresholds and compare the different morphological interpretations suggested between using or eliminating various portions of the more uncertain data. In addition, we divide the study reaches into distinct morphological units (e.g. submerged bar fronts, bar heads, channel thawlegs, sloughs, banks, etc.) and explore the respective contributions of cut, fill and predicted changes below $_{min}$LOD thresholds from each. Volumetric and surface-area distributions of elevation changes are presented to illustrate contrasting signals between morphological units. To illustrate these techniques, ground-based survey data from three real-time kinematic Global Positioning System (rtkGPS) surveys of a 801 X 272 m braided gravel reach of the River Feshie in the Scottish Highlands. Annual changes are reported between 2002 and 2003 as well as 2003 and 2004 (21,206, 51,080 and 46,000 points respectively). Although reach-averaged annual sediment transport rates are sometimes inferred from DEM differencing, the time-scales of monitoring (10$^{0}$ years) are so disconnected from the time scale of the process (10$^{-8}$ years) that the interpretation is difficult. However, meaningful metrics and rates of change at sub-reach and meso-scales beyond the reach-scale sediment budget can be inferred from DEM differencing.
http://www.geog.soton.ac.uk/users/WheatonJ/JMW_Research.asp
H43A-0353 1340h
Computational Model Of Width Adjustment In Meandering Channels
Widening in sinuous channels occurs when the retreat of outer bank exceeds the advance of the opposite bank. This paper presented an analytical model to simulate width adjustment in meandering channels by calculating bank erosion at both banks. The solution to flow field in sine-generated channels indicated that bank erosion could occur at the bank where velocity is moderate. Concerning the phenomenon of bank retreat for non-cohesive materials, two consequent processes: basal erosion and bank failure, were recognized. The rate of basal erosion was calculated as a function of the gradient of longitudinal sediment transport rate and the strength of the secondary flow. The transverse bed slope was treated as a variable increasing with channel sinuosity until it reaches its maximum value. This study showed that the rate of bank line retreat for non-cohesive materials was determined by basal erosion rate, near-bank bed degradation rate, grain size, and the difference between flow depth and bank height. The time-dependent widening processes of two meandering channels in laboratory were selected to test the model applicability. The result showed that the simulated banklines at individual time intervals closely matched the experimental measurements.
H43A-0354 1340h
Calibrating and Measuring Bedload Transport Using a Magnetic Detection System
One of the problems in bedload transport research is that no measurement technique has been commonly accepted as superior, and there are no standard protocols. There is a need for continuous bedload measurement to adequately resolve patterns in temporal and spatial variability, especially at high transport rates. Magnetic detection systems are a promising method as they can sense the movement of natural stones, and provide high frequency data in both time and space. A number of magnetic systems have been deployed in the field, but they have not been adequately calibrated. This has limited the analysis to counting the number of pulses, and not allowed confident estimations of the true amount of sediment transport, sediment texture or particle velocities. We developed a series of lab and flume experiments to calibrate the BMD system used by Tunnicliffe et al (2000). Experiments were run with both artificial and natural stones to isolate the effects of particle size, velocity and magnetic content (susceptibility and moment) on the shape of the recorded signal. A large number of experiments were conducted to cover wide range of flow conditions, particle sizes, and particle velocities. The results show that the system is sensitive enough to detect particles down to at least 8mm. Using artificial stones we were able to relate the signal amplitude, width and area to particle size, velocity and magnetic content. These results suggest that the magnetic system can be used to estimate transport rates in natural streams. Work is continuing with natural stones both in the laboratory and the field to further develop of the system. Tunnicliffe, J., Gottesfeld, A.S., and Mohamed, M. 2000. High-resolution measurement of bedload transport, Hydrological Processes, 14, 2631-2643.
H43A-0355 1340h
Multi-Scale Modelling of Bed Load Transport
There are two main types of bed load transport models. Conventional bed load equations produce a bed load flux (which can be grain size specific) for a given set of flow conditions. They commonly require empirical parameters, which makes them unreliable and difficult to transfer between contrasting environments. The second model type is the mechanistic discrete element model (DEM). In these, each grain is individually modelled and its movement calculated by resolving the forces acting on it. However, the computational requirements of DEMs limits the number and size range of the modelled grains, therefore they have yet to be applied beyond the patch scale (cm to m). One alternative is a multi-scale approach whereby the DEM method is used to produce physically meaningful parameters for bed load transport equations. DEM beds of grains with a given grain size distribution can be created, and their surfaces interrogated to define distributions of parameters such as surface grain sizes, pivoting angles, grain exposure and roughness. A limited amount of field measurement to define the range of grain sizes would allow parameterisation of bed load transport equations, specific to an area. One limitation of many DEMs is that the grains are modelled as spheres. Grain shape is an important control on grain geometry, which affects grain entrainment. The grains in natural sediments are generally far from spherical. This could restrict the ability of a DEM to replicate natural sediment, suggesting that DEM derived parameters may also not be representative. In addition the way in which the DEM bed is created can also affect the surface properties. This paper presents research which illustrates the creation of DEM beds and measures their surface properties. The same properties are measured from gravel bar surfaces from the River Feshie, using techniques including photogrammetry, load cell measurements and stone counts. The different shape of the gravels grains, and their imbrication are found to affect the surface properties. The implications of this for DEM modelling are considered.
H43A-0356 1340h
Implications of Acoustic Bed Velocity Measurements for Sediment Transport and Bedform Dynamics in Large Sand-bed Rivers
Field investigations into the relationship between acoustic Doppler current profiler (ADCP) bed-velocity measurements and bedload transport rates have served to narrow the range of the possible near-bed phenomena that influence ADCP bottom-track response, and provide a glimpse into the nature of sediment dynamics in the near-bed region of a large sand-bed river. We collected about 80 bedload samples from the lower Missouri River while concurrently recording ADCP, real-time kinematic GPS, and echo-sounder data. We calculated apparent bed velocities using an approach introduced by Rennie et al. (2002), who suggested that the difference between GPS positions and ADCP bottom-track positions represents the velocity of particles moving on or near the stream bed. We found the relationship between ADCP bed velocities and measured bedload transport rates to be non-linear, with a decreasing rate of increase in the bedload transport rate for higher bed velocities. This nonlinearity appears to be generated primarily by steepening of the near-bed velocity gradient at higher transport stages, and is linked to changes in bedload particle size and bedform morphology. An analysis of near-bed acoustic backscatter profiles indicates that water bias, i.e., acoustic backscatter from faster-moving suspended sediments traveling higher in the water column, is a relatively minor contributor to the non-linearity. The acoustical returns recorded by our instrument appear to be influenced by particles traveling at a relatively consistent height above the bed, despite differences in sream flow conditions. Data collection spanned a period of approximately five-months, during which several heavy precipitation events in the region produced highly variable discharge and periodic pulses of tributary sediment inputs. We were therefore able to evaluate the effects of simultaneous changes in both sediment characteristics and hydraulic conditions on ADCP bed velocity measurements.
H43A-0357 1340h
Comparison of Three Measurement Techniques for Estimation of Sediment Transport Using Channel Morphology
Documenting changes in channel morphology has traditionally been achieved through periodic resurvey of permanently monumented cross sections. In British Columbia, 197 cross sections in eight study areas of Carnation Creek have been resurveyed annually since 1970 to measure changes in bed material storage in response to forest harvesting. Spatial interpolation between cross sections allows production of a digital elevation model (DEM), and subtraction of multi-temporal DEMs in a geographical information system (GIS) enables high-resolution estimates of changes in stored sediment. However, the quality of each DEM depends on the spacing between cross sections relative to the local variability in channel morphology. Recent advancements in digital photogrammetry enable acquisition and analysis of very large data sets with the potential for increased precision and accuracy compared to conventional ground surveying techniques. In this study, independent DEMs derived from seven cross sections and from digital photogrammetry are each compared to a high-resolution total station survey of a single riffle-pool sequence of Carnation Creek. The mean error of the DEMs derived from cross sections and from photogrammetry was -0.032 and -0.035 m, respectively. Errors in the DEM derived from photogrammetry were most apparent in relatively deep water ($>$ 0.5 m) and under overhanging vegetation. Errors in the DEM derived from cross sections were most apparent near the mid-point between cross sections and along the streambanks. Experimental removal of cross sections from the analysis suggests that errors remain relatively stable for a cross section spacing of up to about one bankfull width. The spatial distribution of errors inherent to each technique has important implications for estimates of sediment transport derived from the morphological method. Cross sections are most effective when the channel is relatively straight, bank heights are low, and the morphology is simple. Photogrammetry is most effective when water depths are relatively low, vegetation cover is minimal, and the morphology is complex.
http://www.fluvial-systems.com/aguposter04
H43A-0358 1340h
Empirical and Experimental Validation of Channel Dynamics Models
Significant advances in understanding the morphodynamics of alluvial rivers have been achieved through a combination of field monitoring, laboratory modelling and morphometric analysis. While each has shed significant insight, all have limitations: monitoring is typically localized and short-term; scaling difficult to achieve in the laboratory; and morphometric studies confounded by the complexity of linking form and process. Coupled numerical modelling of flow and sediment transport is seen as a way to relax these constraints and provide tools for investigation the evolution of alluvial systems. However, the complexity of solving mutually adjusting flow and topographic fields severely restricts the application of conventional hydraulic models at appropriate reach and decadal space and timescales. Reduced complexity approaches, such as cellular automaton (CA) and raster modelling offer alternative frameworks to explore interactions at these scales and combine computational efficiency with simplified physical theory. While such models appear capable of qualitatively reproducing complex alluvial morphologies, confidence in their predictions requires further robust model testing. This research presents an analysis of the hydraulic engines of popular CA approaches applied to braided rivers. In these, sediment flux is driven by simple raster flow routing algorithms that predict the distribution of discharge by apportioning flow between cells on basis of local elevation differences. Variants on this theme incorporate additional sophistication by allowing flow routing at high angles and applying uniform flow formulae to predict pathways on the basis of water surface slope. These schemes are applied to a high-resolution terrain and distributed flow dataset for the braided river Feshie, surveyed in July 2003 and 2004. Observations of the distributed pattern of flow depth, cross-sectional flow and allocation between anabranches are used to test predictions and Monte Carlo simulation used to assess sensitivity to terrain model errors. CA frameworks are found to perform well, with schemes based on water slope superior those driven directly by the topography. All models were, however, found to be highly sensitive to minor terrain model errors, limiting their application to low precision airborne datasets
H43A-0359 1340h
Variations and Dependency of Sedimentary Processes and Morphology of the Hudson River Estuary
The morphology of rivers and estuaries is controlled by many factors including bedrock geology, flow conditions, and sediment input. The 240 km long Hudson River Estuary crosses different types of bedrock, and its sediment transport is strongly influenced by tidal currents and strong runoff events. The small contribution of tributary input in the estuarine part of the Hudson River allows to compare the relative influence of these different factors on morphology and sedimentary processes. A comprehensive data set including multibeam bathymetry, sidescan sonar, a dense network of sub-bottom profiles, and over 500 sediment cores and grab samples has been acquired as part of the Hudson River Benthic Mapping Project funded by the New York State Department of Environmental Conservation. Based on these data we describe and compare important factors and their changes such as grain size distribution, morphology, bedrock geology, and depositional and erosional environments. We see a distinct change in grain size distribution from sand to mud-dominated sediment composition that corresponds to a morphological change from a meandering and bar-building channel to a straighter channel with large estuarine embayments. The dominating sedimentary processes vary more often than the grain size composition and there is an apparent correlation of these variations with changes in the surrounding bedrock geology and related morphology. In addition to natural variations, a canalized section of the Hudson River shows the influence of human alteration on the morphology and possibly on related sediment transport.
H43A-0360 1340h
Experiments in eruption recovery: Channel bed and sediment transport adjustments as sand inputs decline
Transient sediment loading can alter channel bed conditions and sediment transport rates which complicates the prediction of sediment yields, yet evaluation of post-eruption sediment yields remains an important component of volcanic hazard assessment. The 1991 eruption of Mount Pinatubo, Philippines, introduced 1 km$^{3}$ of loose, sandy, pyroclastic flow debris to the Pasig-Sacobia-Abacan basins. River recovery is ongoing as sediment is removed from the basin or stabilized. Declines in sand inputs appear to be driving changes in surface composition and grain mobility. From 1996 to 2003, surface grain size increased, clast structures developed, and mobility declined on the Pasig-Potrero River. To investigate how further declines in sand input may affect bed organization and sediment transport, we conducted a series of four experiments in a 10m $\times$ 0.7m flume at Humboldt State University. Bed material and initial slope (2%) matched conditions on the Pasig-Potrero River at a 1:4 grain size ratio. Water and sediment discharge were set such that all grain sizes were visibly mobile. Median bedload size to depth ratios covered a similar range as the Pasig-Potrero, with higher average values. In the first run, the sediment feed matched the bed distribution with 70% sand. Subsequent runs had 60%, 50%, and 40% sand in the sediment feed. We monitored water surface and bed elevations, depths, sediment output rates and grain size distributions. Periodically, the bed was scanned at 1mm resolution and photographed. Bed configuration changed dramatically between runs. At 70% sand, the bed was highly mobile, with isolated gravel clasts rolling or sliding over a moving carpet of sand. At 60% sand, most stationary gravel was found in clusters. In runs with 50% and 40% sand, armored alternate bars developed, with pulses of sand and gravel jams moving downstream. As sand content decreased, the slope increased up to 4%, raising the shear stress to transport material of greater caliber. Bedload was finer than the feed during aggradation. Surface grain size adjusted faster than slope to the changes in sand content. Friction angles measured from topographic scans show a dramatic shift in distribution peaking at $0\deg$ in the 70% sand run to a flat distribution with 60% sand, to a distribution peaking at higher angles in runs with 50% and 40% sand. The average positive angle increased from $24\deg$ in the 70% sand run, to $56\deg$ with 40% sand. This corresponds to a 1.5- to 3.5-fold increase in dimensionless critical shear stress. Reach-averaged shear stress increased with declining sand content, but Shields stress ($\tau^{*}$) decreased an order of magnitude between 60% and 50% sand content. The range of $\tau^{*}$ and dimensionless bedload transport rates in these experiments spanned the range measured on the Pasig-Potrero River during the rainy seasons in 1997-98 and 2001. The data plot in two distinct sets, and the sharp transition illustrates bed sensitivity to sand content and could help explain strongly seasonal behavior on the Pasig-Potrero River. Low sand inputs during the dry season lead to channel consolidation, incision, and armoring, with the channel returning to wide, shallow, braided conditions in the rainy season. The sharp transition also illustrates the difficulty in predicting sediment yields during recovery from sediment loading, as evolving sediment rating curves may change dramatically through time.
H43A-0361 1340h
Coupling Sediment Transport with the Dryland Environment
The hydrologic response of dryland fluvial systems is unique in comparison with that of humid fluvial systems. Infrequent, high intensity precipitation events produce the effective flow discharge, controlling channel morphology. Transport efficiency in dryland ephemeral channels is often orders of magnitude higher than in similar perennial streams with bedload transport making a relative higher contribution to overall sediment yield, which is fully transport capacity limited. Because of the infrequency of the transport events in dryland fluvial systems, empirically based datasets are rare. Acceptable predictions of sediment transport rates and yields based on available simulations models span a broad range of values. High quality field data are critical to improving simulation models. A research project was initiated in 2004 to measure and model sediment transport processed in low-order dryland channels based on detailed field data collection. The USDA ARS has monitored the Walnut Gulch Experimental Watershed near Tombstone, Arizona for over 50 years. The area is semiarid rangeland, receiving an average of 350 mm of precipitation annually. The Lucky Hills watershed, an 11.4 acre sub-watershed, is intensively monitored for precipitation and run-off. Two traversing slot-samplers and pit traps have been installed to measure sediment transport within its main channel. Historical local research and a decade of comprehensive cross-section surveys provide information on channel morphology change. Historic runoff and sediment data are analyzed to define relationships between the two. Findings are coupled with recent field surveys of channel morphology, bed substrate, affecting vegetation, and evidence of hillslope processes to provide a greater understanding of regional landscape evolution. Even in a small, relatively uniform watershed like Lucky Hills, controlling processes are highly complex and vary dramatically in time and space. However, results show a strong correlation of sediment transport rates with the nuances of the disharge event hydrograph and the physical properties of local substrate available for entrainment. The wealth of historical data available at Walnut Gulch is a rare and valuable asset, helping to overcome the inherent difficulties of study within dryland fluvial geomorphology.
H43A-0362 1340h
Dynamic Sediment Modeling in Iowa Streams and Rivers: A Case Study at Walnut Creek
Deep channel bed incision and severe channel bank erosion, which have strong effects on the evolution of channel and watershed morphology, are becoming serious problems in natural rivers and streams in Iowa as a result of wide distribution of loess soil material, agricultural activity, river training and human intervention. Consequent high sediment concentration can also cause low water quality and jeopardize aquatic habitat. Dynamic modeling of sediment transport in rivers and streams provides a useful tool for monitoring, controlling and forecasting the morphology change and water quality in channels and watersheds. In order to gain insight into sediment transport process, a dynamic sediment model is built for a 7-mile segment of Walnut Creek in Jasper County, Iowa. This creek was intensively surveyed by Iowa Geological Survey Bureau (IGSB) as part of the Walnut Creek Nonpoint Source Monitoring Project. Besides channel geometry data from the survey, hydraulic and sediment data were collected at two gauges upstream and downstream operated by USGS. A software GSTARS3 developed by USGS is adopted to model both channel bed incision and bank erosion which are typical phenomena in Iowa. The dynamic sediment model is calibrated using channel bathymetry data from recent survey conducted by IGSB. Finally, based on forecasting of flow and sediment discharge time series at the upstream and stage time series at the downstream, a sediment forecasting model is developed to see if the stream can go back to the clarity and morphology of original creek. The study on this small surveyed and controlled creek will benefit our research in other Iowa rivers and streams.
H43A-0363 1340h
Adjustments of Bed Sediment Texture to Variations in Shear Stress in High Gradient Streams
The sediment supplied to mountain streams consists of a mix of grain sizes ranging from sand to boulders. In most streams the sediment supplied quickly separates into two modes: A coarse mode that is retained on the bed surface, forming a mobile armor layer, and a fine mode that goes into temporary storage in the bed, forming the substrate. Interactions between the flow, the bed surface, and the substrate help to balance the sediment supply between neighboring stream reaches. This study investigates trends in surface and substrate grain sizes in relation to reach-scale geomorphology using data from gravel-bed streams in Colorado and Utah. The data set includes more than 100 co-located measurements of surface and substrate sediment, bankfull channel geometry and reach-average slope. Slopes at the study sites range from 0.0003 to 0.07; bankfull depths range from 0.2 to 5 m; and bankfull widths range from 2 to 200 m. The data indicate that both the bed surface and the substrate become more poorly sorted as the bankfull shear stress increases. The decrease in sorting with increasing stress reflects a greater separation between surface and substrate grain-size parameters (D84 and D50). The ratio of surface D50 to substrate D50 increases from about 2 in reaches with low shear stress to about 4 in reaches with high shear stress. Similar trends are observed in the surface D84 and substrate D84. Supposing the size distribution of the load is similar to the substrate, as proposed in some models of bed load transport, our results suggest that retention of the coarsest fraction of the sediment supplied to headwater reaches limits the mobility of the finer sizes which constitute the bulk of the load. The high potential for transport of fine grains is therefore offset by the sheltering provided by large grains.
H43A-0364 1340h
Evolution of the size distribution of fine suspended sediments during downstream transport: Role of stream-subsurface exchange
In the classical view of fine sediment transport and deposition in streams, particles are expected to be removed simply by direct sedimentation onto the streambed. However, recent research has demonstrated that fine sediments can propagate into pore spaces in the streambed due to hyporheic exchange and be removed from suspension by a combination of physical and chemical (cohesive) processes. This behavior can significantly alter fine sediment size distributions during downstream transport because the mobility of fine particles and their attachment to bed sediment grains are both functions of particle size. To illustrate the effects of these processes, we simulated the stream-subsurface exchange and deposition of suspended sediments with a bimodal size distribution under different flow conditions. We also applied this approach to analyze the results of two laboratory flume observations of suspended sediment deposition. Model simulations clearly show that the rate of particle deposition increases with increasing particle size. Thus, larger particles are preferentially removed from mixtures, resulting in a fining of the suspension over time. The temporal evolution of the particle size distribution is also a strong function of the stream flow rate. These results clearly demonstrate the effects of stream-subsurface exchange on the temporal evolution of the suspended sediment size distribution during downstream transport.
H43A-0365 1340h
A Sediment Transport Based Geomorphic Analysis of the Skykomish River Braided Reach to Identify "Restoration" Opportunities
A study is underway to assess the driving sediment transport and hydraulic processes influencing channel changes in a braided reach of the Skykomish River, Washington. The reach is located below a steeper, confined section of the river and has likely been geomorphically active since the last glaciation. Bankfull widths range between roughly 100-350 m within, and drainage area is 1,500 km2 above, the study reach. Analyses have been conducted at the reach scale, and include development of a sediment transport model, historical photograph and survey overlays, and an accounting of where sediment deposition, channel shifts and avulsions, and side channel connection are most and least likely to occur over reasonable design life spans (e.g., between 10 and 50 years). The ultimate goal of the analysis is to identify suitable locations for projects that will enhance, restore, or protect fish habitat as well as protect infrastructure, while considering constraints posed by channel hydraulic, sediment transport/deposition, and stability characteristics. Our strategy is to determine what types of projects are best suited for different locations in the reach based on the analysis results. The results can then be used to prioritize and estimate costs for project alternatives.
H43A-0366 1340h
Effects of Bedload and Suspended Load on the Separation of Sands and Fines in Mixed Sediments
A high-shear stress straight flume commonly used to measure cohesive sediment erosion rates has been modified to include downstream bedload traps. The modified flume, called ASSET Flume, can then be used not only to measure erosion rates, but also to analyze and quantify the modes of transport. The modified flume was used to study erosion rates of quartz particles ranging in size from 19--1250~$\mu$m in order to validate the design and operation. Natural sediments from three locations were studied using the modified flume. Fine sediments with little or no sand eroded as aggregates. These aggregates maintained their integrity in the flume channel and moved as bedload into the traps. Natural sediments that included high percentage of sand also eroded as aggregates. However, these aggregates quickly fell apart with the extent of their disintegration proportional to the initial sand content. The sand moved as bedload and fell into the traps while fine particles separated from the sand and moved predominately in suspension. The ability to accurately quantify the modes of transport and the properties of the transported material by means of direct measurement is unique. These measurements are extremely important in determining whether dredge material made up of mixed sediments is suitable for near-shore placement and beach replenishment. {\st Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States Department of Energy's National Nuclear Security Administration under contract DE-AC04-94AL85000.}
H43A-0367 1340h
Geometric Properties of Bifurcating Delta Distributary Channels
The process of channel bifurcation lies at the heart of delta function and form, yet the geometries and stabilities of distributary diffluences, and the hydrodynamic conditions that give rise to them, cannot be predicted. To better understand this process we have collected the distributary network topologies of 26 deltas representing a broad range of climates, sediment and vegetation types, and river discharges. Channel widths, channel lengths from diffluence to diffluence, diffluence angle ($\alpha$), and the length-width ratio of diffluence pairs were measured from Landsat 5 images with 30 meter per pixel resolution using GIS tools. This resolution precluded channels narrower than 100 m; we also eliminated distributary channels that rejoined downstream. Diffluence order is here defined as the number of diffluences upstream of the current diffluence. Channel width was non-dimensionalized by width of 0$^{th}$ order channel and channel length was non-dimensionalized by the width of that channel. Results show that the distributary channel diffluence angles in our dataset are normally distributed with $\bar${$\alpha$} = 78$\deg$ and $\sigma$ = 26$\deg$ (N = 540). Channel widths and lengths are log normally distributed with $\bar{x}$ = 0.40 (N = 340) and 21.50 (N = 195), respectively. The channel width and length ratios of the bifurcate arms are square root-log normally distributed with $\bar{x}$ = 0.60:1 (N = 170) and 0.57:1 (N = 68), respectively. A statistically significant positive correlation (R$^2$ = 0.57; N = 195) exists between dimensional channel length (L) and width (W) of bifurcates such that $L = 13.33W$^{1.08}$, consistent with scaling theory. Partitioning of these properties within a delta was examined by binning the data according to diffluence order. When binned and averaged by diffluence order, diffluence angles show no trend, however there is a well-defined decrease in channel width, channel length with increasing order. Channel width and length ratios show the same trend, but are more scattered. When dimensional channel width and length are binned by order and fitted to a power law, length becomes a progressively higher order function of width with increasing diffluence order. A rational theory explaining these values remains to be proven. The diffluence angle is not easily predicted by the dynamics of turbulent plane jets, and the average width and length ratios of the bifurcate arms are opposite the ratios predicted by the theory of Bolla Pittaluga et al. (2003) for braid-bar diffluences.
H43A-0368 1340h
Measurement and Analysis of the Spatial Organization of Bed Particles In Boulder-Bed Streams
Reaches of boulder-bed streams that do not exhibit step-pool morphology are not well characterized, but they can extend several kilometers in length. These reaches often appear in rivers as a transition state from boulder-bed step-pool morphology to gravel-bed riffle-pool morphology. These reaches do not have clearly-defined morphological structures such as a steps or riffles, but the spatial organization of bed particles may still affect the hydraulics of the streams. We measured a set of 6 reaches where we examined the placement of particles within the reach. These data indicated that bed particles are not randomly arranged on the streambed and that bed particle arrangement significantly enhanced bed stability. Detailed bed particle arrangements, however, are not a practical measure for comparison among many stream channels. We examined, therefore, whether aspects of the grain size distribution could be used for bed characterization. In boulder-bed streams, the distribution of bed particles in the upper tail size ($>$ 84%) is more heterogeneous than in lower percentiles. In some cases, the largest particles are not mobile and are often protrude from the flow at discharges less than bankfull. For these reasons, the sum of the large ($>$ 84%) particles sampled across the stream emerges as a good choice to represent a horizontal roughness length, which affects flow in the channel cross-section. This sum of particles greater than 84% is referred to as topsum in this research. The topsum metric is intended as a horizontal length scale to represent the potential size of a step structure and constrictions in the channel width. We examined 50 field sites on boulder and cobble bed streams with varying particle size and sorting, we found topsum to be linked to both particle size and sorting. It also provides a horizontal roughness length scale that can be used to explain some of the variations in flow resistance among reaches.
H43A-0369 1340h
Numerical modeling of interactions between lake level head difference, river hydrodynamics and bathymetric change in the St. Clair River
The recorded drop in the difference between lake levels on Lake Huron and Lake St. Clair, and possible relationships to historical changes in the flow regime of the St. Clair River is investigated. An evaluation of the possible causes of lake level change included consideration of tectonic uplift, net basin supply change, channel bed scour and sedimentation, and changes to flow through the St. Clair River (changes in hydraulic geometry, ice, vegetation, scour due to ships, exposure of underlying erodible material). Numerical modeling was performed to assess the interactions between historical bathymetric change on river flows the in St. Clair River. The RMA2 hydrodynamic model developed by the U.S. Army Corps of Engineers was used for this investigation. The model was calibrated against stage and flow measurements in the river. Two sets of historical bathymetric data for the St. Clair River (1948 and 2000) were used for modeling runs. The model had open boundaries at the upstream and downstream limits, and both were controlled by lake levels. Various model scenarios were tested to investigate the impacts of bathymetric change and lake level head difference change on river flow patterns and vice-versa. The model output indicates that there has been a significant increase in conveyance throughout the 20th century, which may at least partly explain a drop in head difference between the lakes. Changes to river cross-section (in the last 50 years) have contributed to the acknowledged ongoing decline in the head difference between Lakes Huron-Michigan and St. Clair-Erie, whilst changes in net basin supply have contributed to discharge fluctuations in the river channel. The interactions and feedbacks between these different factors are discussed.
H43A-0370 1340h
Flow Resistance Partitioning in Step-Pool Channels
Increased understanding of controls on flow resistance and of how resistance is partitioned between different sources is essential for advancing understanding of physical processes in steep channels. Step-pool stream channels are an important category of steep channels in which flow resistance is created by large clasts, spill over step-pool bedforms, large woody debris, and other factors. In order to measure resistance partitioning between grains, steps, and woody debris, we manipulated variables contributing to flow resistance in step-pool channels via a series of laboratory flume runs. A factorial design, whereby total resistance was measured for flume runs with and without grains, steps, and woody debris, and at multiple slopes and discharges, was employed. This provided a means of quantifying the relative contributions of grain resistance, spill resistance, and debris resistance to total resistance, represented here by Darcy-Weisbach friction factor. Independent estimates of resistance partitioning were developed based on calculations of grain resistance, from Keulegan log-law relations, and of debris resistance, based on calculations of drag force associated with the cylinders used here to represent woody debris. Calculations of resistance partitioning indicated that spill resistance and debris resistance were responsible for the largest components of total resistance, and that grain roughness was a small component of total resistance. The relative contributions of grain, spill, and debris resistance depended on discharge, with debris resistance dominating at higher discharges, and debris density, with similar contributions from spill and debris components at low debris densities and greater debris roughness at higher debris densities. The large contributions of resistance associated with woody debris and spill over steps suggests that Keulegan-based methods of calculating flow resistance may substantially underestimate total resistance in step-pool channels. In addition, significant interaction effects observed between steps, grains, and debris in a factorial ANOVA provide insight into the potential errors in simple additive approaches to resistance partitioning.