H44B-01 INVITED
Dynamic Adjustments in Channel Width in Response to a Forced Diversion: Gower Gulch, Death Valley National Park, California
We study the 1941 diversion of Furnace Creek Wash (drainage area 439 km2) into Gower Gulch (5.8 km2) as an experiment in the transient response of channel geometry to a large change in water and sediment discharge. We measure sequential changes in valley width using a time series of aerial photographs (1948-1995), airborne laser elevation data from 2005, and a field survey. We find that response of the system varies depending on the pre-diversion channel morphology and geology. In two steep knickzone segments, narrowing, knickpoint retreat, and bedrock incision dominates-- a detachment-limited response. In the relatively low-gradient main part of Gower Gulch, fine-grained, soft sedimentary rocks underlie the channel, and widening dominates as the large, coarse post-diversion sediment load covers the channel bed. The response in this section is transport limited, with only modest incision and adjustments in gradient. Two different processes appear to cause the channel to widen. (1) In many reaches, the stream is attacking the valley walls, as evidenced by fresh plucking and scour marks. This probably occurs because the bed in the middle of the channel is alluviated and protected, which minimizes the opportunity for vertical incision. (2) Some reaches have experienced aggradation, which widens the valley by filling it in. This occurs in places where storage space exists (splay deposits in small tributary mouths, fill terraces in the wide valleys at larger tributary mouths) or in reaches upstream of constrictions. Over long periods, the lowering rate of Gower Gulch probably depends on knickpoint retreat, but the present-day response of this non-steady-state system is a hybrid of incision and narrowing in detachment-limited reaches and widening in transport-limited reaches. This system demonstrates the importance of initial conditions and evolving channel geometry in setting the transient response of rivers.
H44B-02
Channel Width Adjustments to Flood Flows in a Highly Erodible Landscape, North Fork Toutle River, Mount St. Helens, WA
Heavy rainfall in November 2006 induced large-scale flooding and debris flows in the Pacific Northwest, most notably on drainages originating on the flanks of volcanoes. Field evidence and measurements from acoustic flow monitors suggest flooding on the North Fork Toutle River (NFTR), which originates in the crater of Mount St. Helens, was preceded by valley-spanning debris flows coupled with as much 5 m of fluvial incision in the headwaters. Peak flood discharges with a return period of approximately 20 years completely reset the width and depth of the NFTR. In this study, we document changes in channel geometry in response to the November 2006 floods by repeating measurements of width, depth, slope and grain size taken several months earlier. Because the NFTR flows through an unvegetated alluvial plain, we hypothesized that the channel would widen to dimensions determined by the flood magnitude, thereby establishing a new downstream hydraulic geometry. Repeat measurements were taken at 12 sites spaced 1-2 km apart, with reach-averaged gradients ranging from nearly 6% in the headwaters to about 2% downstream. Modeled flow estimates for the flood channel are in agreement with peak discharges estimated from regional flood-frequency analysis, increasing downstream from 40 to 150 cms. Flood flows more than doubled the width of the channel from an average of 12 m to more than 32 m during the flood, before returning to an average active channel width of 15 m by the summer of 2007. Downstream hydraulic geometry relations for channel width, w=aQb, have essentially the same exponents, b=0.49, 0.48 and 0.50, for conditions prior to, during and after the 2006 floods. This suggests that in erosive environments where channel adjustments can occur rapidly, the scaling relationship between width and discharge, w ~ Q0.5, remains roughly constant for relatively common as well as infrequent flood events. Further, despite massive erosion and channel change, the NFTR was able to readjust its channel geometry to almost the same configuration as before the flood.
H44B-03 INVITED
The dynamics of bedrock channel adjustment: Modeling the influence of sediment supply, weathering, and lithology on channel cross-sectional and longitudinal shape
The cross-sectional form of a natural river channel controls the capacity of the system to carry water off a landscape, to convey sediment derived from hillslopes, and to erode its bed and banks. Numerical models that describe the response of a landscape to changes in climate or tectonics therefore require formulations that can accommodate changes in channel cross-sectional geometry through time. We have developed a 2D numerical model that computes the formation of a channel in a cohesive, detachment-limited substrate subject to steady, unidirectional flow. Boundary shear stress is calculated using a simple approximation of the flow field in which log-velocity profiles are assumed to apply along vectors that are perpendicular to the local boundary surface. The resulting model predictions for the velocity structure, peak boundary shear stress, and equilibrium channel shape compare well with the predictions of a more sophisticated but more computationally demanding ray-isovel model. For example, the mean velocities computed by the two models are consistent to within ~3%, and the predicted peak shear stress is consistent to within ~7%. The efficiency of our model makes it suitable for calculations of long-term morphologic change both in single cross-sections and in series of cross-sections arrayed downstream. For a uniform substrate, the model predicts a strong tendency toward a fixed width-to-depth ratio, regardless of gradient or discharge. The model predicts power-law relationships between width and discharge with an exponent near 2/5, and between width and gradient with an exponent near -1/5. Recent enhancements to the model include the addition of sediment, which increases the width-to-depth ratio at steady state by favoring erosion of the channel walls relative to the channel bed (the "cover effect"). Inclusion of a probability density function of discharges with a simple parameterization of weathering along channel banks leads to the formation of model strath terraces. Downstream changes in substrate erodibility or tectonic uplift rate lead to step-function changes in channel width, consistent with empirical observations. Finally, explicit inclusion of bedload transport allows channel width, gradient, and the pattern of sediment flux to evolve dynamically, allowing us to explore the response of bedrock channels to both spatial patterns of rock uplift, and temporal variations in sediment input.
H44B-04
Sediment Supply and the Prediction of Bedrock Channel Cross Section Evolution
The cross sectional shape of a bedrock channel reflects the channel's history of incision. Although current theory suggests that sediment supply is a dominant control on width-averaged bedrock incision rates, models predicting the cross sectional shape of bedrock channels have thus far neglected sediment supply effects. Here we present a simple model that predicts the evolution of the cross sectional shape of a bedrock channel under given conditions of sediment supply. Our model assumes that all erosion is the result of particle impacts of saltating bedload, and that bedload transport occurs at transport capacity along an active zone, the width of which is determined by the sediment supply, discharge, and channel cross sectional shape. We allow bedrock erosion to occur at locations within the active transport zone where the thickness of the sediment layer does not exceed a critical threshold. Preliminary model results are qualitatively similar to experimental observations documenting slot incision in response to a reduction in sediment supply and partial alluviation and peripheral erosion in response to a subsequent supply increase. Integrating our model predictions over the channel cross section to calculate width-averaged erosion rates indicates that the model captures the so-called `tools' and `cover' effects wherein width-averaged erosion rates reach a maximum value at an intermediate sediment supply rate. However, the specific nature of the relationship between average erosion rate and sediment supply is strongly influenced by the shape of the channel cross section. Our results may provide insight on estimating historical sediment supply rates from observed bedrock channel geometry.
H44B-05
The effects of emergent vegetation on sediment transport and channel morphology
Riparian vegetation in rivers and on channel banks can significantly influence flow, sedimentation and channel width. For example, emergent vegetation on banks will locally increase drag and sediment deposition, which may cause the channel to narrow. Alternatively, depending on the location, size, and density of vegetation within a channel, it may enhance local and lateral channel erosion. Current models cannot accurately predict sediment transport rates through fields of emergent vegetation. We conducted a set of flume experiments to gain a mechanistic understanding of the influence of vegetation on flow and sediment transport. In our experiments, sand was transported through regular arrays of cylinders for a range of cylinder densities and flow discharges. We measured the velocity field using particle imaging velocimetry (PIV) and the spatial variations in the sediment transport rates using a high-speed video camera. For a given flow velocity, an increase in cylinder density (by area) augmented the reach-averaged drag, local near-bed turbulence intensities and spatial variability in the shear stress. We incorporated these three effects into a sediment transport equation and tested a number of transport equations using our experimental data. Except for our sediment transport equation, all of the equations predicted sediment fluxes that differed from the measured values by several orders of magnitude. Sediment transport equations that account for vegetation drag but only use the reach-averaged shear stress significantly under-predicted sediment transport. Such equations do not capture the nonlinear variability in sediment flux with spatial variations in boundary shear stress. Thus, the effects of turbulence and spatial variability in the flow must be included in sediment transport predictions through vegetation. We combine our sediment transport equation with a two-dimensional flow model to predict the effects of vegetation on erosion and deposition in natural river channels. Such predictions are used to understand the effects of vegetation on channel morphology and the evolution of channel width.
H44B-06
Regional Variation in Landscape Controls on the Width of Wadeable Streams Across the Conterminous United States
We examine regional variations in landscape controls on the width of wadeable streams across the conterminous United States using a unique dataset from the U.S. Environmental Protection Agency's (USEPA) Wadeable Streams Assessment (WSA). The WSA dataset includes a probability sample of over 1,300 randomly selected stream reaches (drawn from the 1:100,000 scale USGS digitized stream network) and over 500 hand- picked reference sites covering 48 states sampled between 2000 and 2004. The scaling of bankfull stream width with drainage area varies significantly among the nine ecological regions (ecoregions) defined for the WSA: width increases more rapidly with basin area in the humid Eastern Highlands (encompassing the Northern and Southern Appalachians and the Ozark Mountains) and the Upper Midwest (Great Lakes region) than for the West (both mountainous and xeric areas), the southeastern Coastal Plain, and the Northern Plains (the Dakotas and Montana). Stream width increases least rapidly with basin area in the Temperate Plains (cornbelt) and Southern Plains (Great Prairies) in the heartland. Besides basin area, key predictors of channel width included particle size and precipitation, although the relative importance of these factors varies among ecoregions. Precipitation was only significant (p < 0.01) as a predictor of bankfull width in the Western Mountains, the cornbelt, and to a lesser degree the Great Prairies. Width was significantly positively related to bed material size (p < 0.01) in the mountainous and xeric West, the cornbelt, and the Southern Appalachians, but was only weakly related or unrelated elsewhere. Preliminary analysis suggests that riparian vegetation cover is an important predictor of channel width only in the Northern Plains and Great Lakes regions, and that width does not seem to be related to measures of woody debris abundance in any region.
H44B-07
Controls on channel width in large rivers dissecting the eastern margin of the Tibetan Plateau
The eastern margin of the Tibetan plateau is characterized by large rivers dissecting regionally-elevated topography that has developed in association with the continued convergence of the Indian subcontinent and Eurasia. Here, we present an extensive dataset of channel width collected from field surveys along the region's largest rivers. These data allow us to more fully characterize channel morphology during transient landscape evolution and provide insight into the manner in which channels adjust to tectonic and lithologic variability. We combine our width data with detailed analysis of river profiles, channel gradients and direct measures of erosion and incision rates. Data reveal that rivers dynamically adjust gradient and width in response to spatial variations in uplift, erosion and rock erodibility. Width trends from rivers experiencing downstream increases in erosion rate indicate channel narrowing (W~A0.1-0.3), relative to typical scaling of width with increasing drainage area (W~A0.4-0.5). Width trends from rivers experiencing downstream decreases in erosion rate, meanwhile, indicate channel widening (W~A0.6), relative to the typical scaling. Superimposed on these general width trends are systematic variations in channel width and gradients of large rivers that reflect local differences in bedrock lithology and channel morphology, the influence of large landslides, and local non- uniform uplift related to active faults.
H44B-08
Numerical and Physical Modeling of Width Dynamics on Fluvial Fans
This paper addresses two main questions: 1) How sensitive are numerical models of fluvial landform evolution to their representation of flow width and its dynamic adjustment? 2) What are the implications of this sensitivity for our understanding of relationships between environmental change and landform evolution? These questions are examined in the particular context of alluvial fans formed by fluvial processes. We present a new model of non-equilibrium flow width on fluvial fans and evaluate its performance using results from a physical modelling study of fan formation. Our results demonstrate that this new model captures the essential controls on, and spatial and temporal changes in, flow width more successfully than existing fluvial width treatments that are commonly used in landform evolution models. We then implement this new representation of flow width within a numerical model of fan evolution in order to assess the potential for using geomorphic evidence (fan morphology and fluvial surface ages) to reconstruct past environmental boundary conditions (water and sediment supply, tectonics and base level change). Multiple (c. 50,000) numerical simulations, conducted within an uncertainty framework, are evaluated using field evidence obtained for small (length c. 1 km), coarse-grained alluvial fans in the Southern Alps, New Zealand. Results demonstrate the potential for width dynamics to drive strong positive feedbacks during fan evolution and promote equifinality in fan morphology.