How do hydrograph recession rate and vegetation influence the morphology of point bars in sand-bed channels?

A stream's hydrograph shows the volume of water flowing in the channel through time (Figure 1). Quantities often used to describe hydrographs include the flow magnitudes, the duration, and the rate of change between stages (Figure 1; e.g. Poff et al., 1997). Regulation of streamflow by dams has altered nearly all of these characteristics on many rivers (Figure 1d), and with them, the ability of channels to transport sediment. How do these changes impact the morphology of stream channels? Previous investigations have primarily focused on the effects of changing peak flows on mobile grain sizes and scour depths or the effects of changing minimum flow rates on habitat availability and accumulation of fine sediment. Rarely has the influence of the rate of change between flow stages on sediment transport or channel morphology been investigated. The limited research available on this topic suggests that more gradually changing hydrographs typical of snowmelt dominated streams (Figures 1b and 1c) promote better sediment sorting on the bed including the development of a coarser surface layer called armor (e.g. Hassan et al., 2006). Conversely, more rapidly changing or "flashy" hydrographs more typical of rain dominated (Figure 1a) or regulated streams (Figure 1d) result in less sorting or armor layer development (e.g. Hassan et al., 2006). Similarly, Hassan (2005) observed that bars formed in a channel subject to "flashy" hydrographs were underdeveloped with very little sorting of sediment by grain size on the surface.

The sensitivity of channel morphology to altered hydrographs is suggested by commonly documented changes downstream of dams which include channel narrowing, entrenchment, and coarsening of the grain sizes on the bed. However, it is difficult to isolate the effects of an altered hydrograph from other influences such as reduction of sediment supply by retention in reservoirs or land use changes. Due to these types of complexities in the field, flumes are commonly used to investigate river processes because systems can be modeled in a simplified way. In addition, flume experiments allow researchers to isolate a single variable of interest while holding all else constant, something that cannot be done in the field.

The Saint Anthony Falls Laboratory Outdoor Stream Lab (OSL) in Minneapolis, MN, USA (Figure 2) was used for this study to investigate how the rate of change during the receding limb of a hydrograph (recession rate) influenced the morphology of a point bar. While most flumes are rectangular in cross-section, the OSL is unique for its 40 m long, 2 m wide meandering channel and vegetated floodplain (Figure 2a). The channel is low gradient, with a maximum slope of 0.02 m/m in two constructed riffles and the substrate of the bed is predominantly sand (median grain size of 1 mm). Flow and sediment feed rates are both controlled by the user (Figure 2a).

We ran three different hydrographs to investigate the influence of recession rate on resulting bar morphology (Figure 3). We limited runs to the recession limb only to simplify the investigation and with the assumption that the recession limb has a significant influence on bar morphology because sediment will deposit in response to declining flow rate and shear stress. Further, we limited our investigation to just bar morphology because bars are significant depositional features in many stream channels and provide important habitat for a variety of plant and animal species. Similar starting conditions for each run were provided by running constant flow and sediment feed rates (0.112m³/s and 0.036 kg/s, respectively) until the channel reached equilibrium (cross-sections stabilized and downstream bar growth ceased). Recession rates of the runs were 10, 30, and 60% (flow reduction at each time step; Figure 3) to mimic natural snowmelt dominated, rain dominated, and regulated hydrographs, respectively. Despite this difference, the total estimated ability of each hydrograph to move sediment differed by <10%. However, this scaling required that the 10% recession start at a lower peak flow. At all times the sediment feed rate equaled the estimated ability of the channel to transport sediment, limiting scour or deposition resulting from an excess or lack of transport capacity, respectively. Bar topography was scanned (sonar and laser) and measured in cross-sections (Figure 2b) before, during, and after each hydrograph.

The final bar morphologies from each recession run displayed key differences. Qualitatively, the more gradual 10% recession resulted in a more "typical" bar deposit, including a well-defined flat top (Figure 4a). The bars resulting from the 30 and 60% recessions lacked a distinct top and had more irregularities in their morphology (Figure 4a). Elevation distributions differed as well (Figures 4 and 5), with more deposition at higher elevations during the 30 and 60% runs, likely because the greater peak flow rate was able to move sediment farther up the bar. In order to quantitatively compare the resulting bar tops we had to first objectively differentiate the break between the sloping side and flatter top of each bar. This was done by plotting the total bar area inundated for a range of water surface elevations. The break occurs where the area inundated increases most rapidly because the flatter region of the bar top elevation has been reached (the second derivative of the water surface elevation vs. area inundated line). After identifying bar tops in this way we found that bar top area declined as recession rate increased (Figure 6a). Similarly, all measures of bar top width (mean, mode, max and min) also declined as recession rate increased (Figure 6b). These results indicate that the top region of bar is reduced for faster recession rates, which may have important implications of riverine species who utilize this area for habitat. For example, loss of bar top area could mean loss of area for cottonwood seedling germination.

Additional studies are needed, but results from this small set of flume experiments suggest that the recession rate of hydrographs does indeed influence channel bar morphology, and possibly the morphology of other channel features. This type of information could provide guidance for designing releases from reservoirs that better preserve or restore channel morphology to provide suitable habitat. It could also be useful for predicting how channel morphologies might change as climate change alters natural hydrographs on unregulated streams (e.g. shifts from gradually changing snow dominated hydrographs to rapidly changing rain dominated hydrographs).

• Hassan, M.A. (2005). "Characteristics of gravel bars in ephemeral streams." Journal of Sedimentary Research 75(1): 29-42.
• Hassan, M.A., Egozi, R., et al. (2006). "Experiments on the effect of hydrograph characteristics on vertical grain sorting in gravel bed rivers." Water Resources Research 42: W09408. doi:10.1029/2005WR004707
• Poff, N. L., J. D. Olden, et al. (2007). "Homogenization of regional river dynamics by dams and global biodiversity implications." Proceedings of the National Academy of Sciences of the United States of America 104(14): 5732-5737.
• Young, W. J. and T. R. H. Davies (1991). "Bedload transport processes in a braided gravel-bed river model." Earth Surface Processes and Landforms 16(6): 499-511.
• Bombar, G., Ş. Elçi, et al. (2011). "Experimental and Numerical Investigation of Bed-Load Transport under Unsteady Flows." Journal of Hydraulic Engineering 137(10): 1276-1282.