H23J-01 INVITED
The influence of small dams on streambed morphology, sediment distribution and fluxes of water, heat and solutes between surface and ground water
Small dams, including natural debris dams and constructed log dams, increase the complexity of streambed morphology and control the localized sediment distribution in streams. To investigate the impact of dams on water, heat and solute fluxes in the streambed, we instrumented a 20 m section of a 2nd order semi-arid stream with temperature data loggers and in-stream mini-piezometers upstream and downstream of a 1.5 m tall constructed log dam. Hydraulic gradients and time-series temperature data in the stream, streambed and groundwater were used to qualitatively and quantitatively assess the direction and rate of water flux through the streambed daily between July of 2006 and June of 2007 and at discrete times during water quality sampling. Spatial variability of stream, streambed, and ground water quality data was used to assess the types and rates of biogeochemical processes occurring in the streambed during October, 2005, and July, 2006. Upstream of the dam, low-velocity pools allow fine sediments to settle out, restricting the flux of water, heat and solutes into the streambed, despite large, negative hydraulic gradients. Downstream of the dam, riffle sequences comprised of coarse sands and cobbles generate relatively rapid fluxes of water, heat and solutes through the shallow streambed, even under smaller hydraulic gradients. As a result, a mosaic of streambed zones with different rates of biogeochemical reactions are created by the constructed log dam and these zones persisted across different seasons of the year. The transport of heat and solutes through the streambed was unique from surface or groundwater heat and solute transport, supporting the view that the streambed is a unique hydrologic zone. Analysis of time-series temperature data was an effective tool to quantify the flux of water through the streambed with a fine temporal resolution (i.e. daily flux rates) over long periods of time (i.e. annually). Short time-scale temperature records were used to qualitatively assess flux variability with high spatial resolution. Using multiple methods to evaluate streambed dynamics revealed strong linkages between water, heat and solute fluxes.
H23J-02 INVITED
Channel-Streambed Interactions Over and Under Ice
An overwhelming set of studies in temperate streams has noted that exchanges of water, solutes, and heat across streambeds are very important to stream ecosystem structure and function. We have postulated that similar dependence is likely in streams affected by ice. In particular, streams underlain by permafrost have a finite potential aquifer with which to exchange (i.e., constrained by the depth of annual thaw), and streams that maintain an ice-cover for any period of time experience a modified hydraulic condition (compared to open water conditions) that affects channel-streambed exchanges. Here we present findings from field and modeling studies of streams underlain by permafrost and from modeling studies of ice-covered channels. Our results from these studies indicate that, 1) in general, sub-channel thaw does not limit exchange potential, rather it is the morphology of the streambed that controls exchange potential and extent, 2) these exchanges facilitate nutrient regeneration, and 3)for the same discharge, exchange fluxes are generally enhanced by the presence of an ice- cover, though the depth of exchange is diminished. These findings indicate that while ice may be an important component of these systems, streambed form and composition are primary controls on exchange processes.
H23J-03
Modeling Streambed Heating in Shallow Streams
The Oregon Department of Environmental Quality, USA is developing Total Maximum Daily Loads to address water quality concerns and threatened and endangered species habitat requirements. Approximately 940 water body segments are listed as water quality limited for temperature in Oregon. CE-QUAL-W2 Version 3 is a two- dimensional water quality and hydrodynamic model capable of modeling rivers, reservoirs and estuaries. An important aspect of modeling stream temperature is handling the short-wave solar radiation that penetrates the water surface and impacts the streambed, which can affect water temperatures under low-flow conditions. The Bull Run River-Reservoir system in Oregon is a 264 km2 watershed 42 km east of downtown Portland and serves as the city's primary drinking water source. A dynamic, three-dimensional streambed heat transfer model was developed and calibrated with field data from the Lower Bull Run River and laboratory experiment data. Model results compared well to field data from bedrock and cobble substrates. The model calibration for the cobble substrate revealed the substrate interstitial water temperature played a large role in the substrate temperatures and was necessary to calibrate the model. The streambed heating model compared well with the laboratory experiments' data in many cases. The model was compared to two analytical models and a one- dimensional model for various application cases and performed well. The streambed heating model was implemented in CE-QUAL-W2 and a sensitivity analysis examined the impacts of river flows, substrate type and streamside shading. Water temperature impacts focused on daily minimum and maximum temperatures. Increased flow rates resulted in decreased water temperature impacts from streambed heating. The largest water temperature impacts occurred with the bedrock streambed with decreasing impacts once cobble was incorporated. Increased streamside shading reduced the impact of streambed heating. General guidelines were discussed when streambed heating may be an important part of the surface water heat budget.
H23J-04 INVITED
Understanding processes in streambeds with reductive models and high-resolution data
Streambeds host coupled physical, chemical, and biological processes that can impact watershed scale processes. Fluid flow through streambeds provides the template for most of the processes since it controls transport. Unfortunately, fluid flow within streambeds may be complicated, even at the grain scale, owing to current-topography driven exchange, heterogeneity, and turbulent-porous media flow coupling. We analyze several interacting processes within streambeds using reductive numerical modeling of hydrodynamic, biogeochemical and thermal processes, while placing equal emphasis on river hydraulics. This leads to predictive models that can potentially be useful in field settings. Results from our models can be used to constrain interfacial fluid fluxes, chemical reaction rates and solute fluxes, and observed thermal patterns. However, streambeds are heterogeneous and seldom uniform. Detailed observations of permeability fields input into models indicate that in certain cases, internal control by permeability may be more important than external fluid flow forcing. Lastly, we present observations of the thermal regime of rivers and show how geomorphologic and ecologic factors could complicate streambed temperature distribution. Despite the complex physics and spatial variability of many parameters, recent developments in observational and modeling technology allows us to inspect streambeds more closely. Although the goal of the community should be to unravel the myriad roles of the streambed in a broader space and time, as well as disciplinary, context, we need to be simultaneously and firmly grounded in process-level understanding.
H23J-05
How Groundwater Discharge Influences the Time Scales of Contaminant Release From Streambed Sediments
Streambed sediments can act as long-term storage zones for organic contaminants originating from the stream water. Although contamination levels in German rivers have generally declined in recent years, streambed sediments might still be considerably contaminated. Groundwater discharge through the sediments can induce an advective contaminant transport so that contaminants are released back from the sediment to the stream water. In our study, groundwater fluxes were quantified along a reach of 220m length of a small man-made stream in the industrial area of Bitterfeld (Germany). The stream was used for waste water discharge from the chemical industry nearby until the early 1990s. The study focuses on the long-term behaviour of chlorinated benzenes (MCB, DCBs) which are the major contaminants at the study site. The groundwater fluxes were obtained with fine spatial resolution using streambed temperatures. Groundwater discharge ranged from 11.0 to 455.0 Lm-2 d-1. According to locations with high and low groundwater discharge, time-integrating passive samplers were installed in the streambed to monitor the vertical distribution of current contaminant concentrations. Batch experiments were conducted to obtain the kinetic parameters of the desorption process and the water-sediment distribution coefficients as input data for a transport model of the streambed. We applied a numerical one-dimensional advective transport model to simulate the timescales of contaminant release and their dependence on the magnitude of groundwater discharge. As initial condition a homogeneous contaminant distribution in the streambed sediment layer was assumed. The mass fluxes from the streambed to the water are determined by the flow velocity of the groundwater and the sediment water distribution coefficient of the specific substance. The long-term predictive modeling indicated that the time required to reduce the concentrations and the resulting mass fluxes to the water by 90% of the initial values will be in the scale of decades for high-discharge locations and centuries for low-discharge locations, respectively.
H23J-06
The role of the streambed zone on the fate of nitrate in the Morgan Creek, MD watershed
Understanding the watershed-scale fate of nitrate introduced through agricultural practices requires an understanding of the physical and chemical processes taking place during ground water and surface water interaction within the highly reactive streambed zone. The Morgan Creek, MD watershed is one of several agricultural watersheds being studied as part of the Agricultural Chemicals Sources, Transport, and Fate Topical Study of the U.S. Geological Survey's National Water-Quality Assessment Program. Approximately half of the annual streamflow of Morgan Creek, MD is attributed to ground-water inflow. This ground water enters the creek by flowing through the streambed, as well as discharging through seeps at the edge of the streambed. Comparison of baseflow stream-water chemistry to ground-water chemistry shows that concentrations of relatively non-reactive elements such as silica and magnesium change very little as water flows through the streambed zone. However, there is a substantial decrease in nitrate concentrations as ground water passes through the organic-rich, low-oxygen denitrifying zone of the streambed and discharges to the creek. Preliminary mass balance analysis suggests that nearly 80% of the nitrate present in near-stream ground water is removed as the ground water discharges to the stream. Although the majority of fertilizer-applied nitrate is believed to be removed in the soil and shallow unsaturated zones by the combined processes of crop uptake and harvest, volatilization, and denitrification, the attenuation of nitrate concentrations in the streambed zone significantly reduces nitrate loads in the creek. Thus, quantifying the watershed-scale mass budget of nitrate requires characterizing and quantifying ground-water discharge and nitrate concentration changes in the streambed zone.
H23J-07
Spatial Distribution of Nitrate Flux from the Streambed of a Low-Relief Coastal Catchment on Virginia's Eastern Shore
Nitrate is a pervasive chemical contaminant in groundwater. The fate of groundwater nitrate is not well understood. We study the fate of groundwater nitrate as it discharges through the streambed of Cobb Mill Creek. The Cobb Mill Creek catchment is located on the eastern shore of Virginia and provides an attractive setting for studying some elements of the nitrogen cycle due to large inputs of agriculturally-derived nitrogen. The catchment is 496 hectares with land uses dominated by forested and agricultural areas. Concentrations of nitrate in the groundwater of Cobb Mill Creek are often close to or exceeding the US EPA drinking water standard of 10 ppm nitrate-N. Concentrations of nitrate in stream water are almost an order of magnitude lower. Previous work has indicated that denitrification in the streambed of Cobb Mill Creek is the most likely cause for the removal of nitrate in groundwater, but removal is spatially variable. We installed dense grids of seepage meters in several reaches of Cobb Mill Creek to investigate spatial patterns of nitrate removal across the channel cross section. We measured the rates of groundwater seepage and the concentrations of nitrate and chloride in groundwater discharging from the streambed. The data indicate that certain areas of the streambed are dominated by vertical upwelling of deep groundwater where denitrification is the primary cause for nitrate removal. Other portions of the streambed receive groundwater that flows through the riparian zone where plant uptake and evapo-concentration may also be important in nitrate dynamics.
H23J-08
The Streambed as a Dynamic Membrane in the Watershed
Improved understanding of the streambed filtering capacity is essential for predicting and mitigating anthropogenic impacts on stream and ground-water quality. This presentation discusses the extent and composition of the streambed, and introduces the concept of the streambed as a dynamic membrane, periodically changing character while continually filtering water flowing through its matrix. The physical extent of a streambed is described as sediments in the watershed under direct influence of an active stream, often forming over paleo streambeds and overbank materials comprising the alluvial aquifer, but also forming thin layers on bedrock in upland locations. The active streambed may consist of a single particle size, such as fine sand, or consist of a wide range of particle sizes of both inorganic and organic material. In alluvial settings, the streambed is often similar in general thickness and porosity to adjacent soils. The initial streambed filtering capacity is controlled by the particle-size characteristics of both the bulk properties and the stratigraphic layering of the matrix, and supplemented by a microbiological and invertebrate community that rapidly populates the matrix to complete the composition of streambed. Once a biogeochemical streambed matrix is established within an alluvial system, the quality of water passing through the streambed is transformed, such that stream and ground- water quality are distinct in their chemical and microbial composition. The physical matrix, geochemistry, and the resident biology may change seasonally (due to scour, deposition, and changing hydraulic, chemical, and thermal boundary conditions), such that streambed properties are temporally dynamic. Thus, as a system- regulated filter, the streambed may be viewed as a dynamic membrane, which facilitates engagement of membrane science for characterizing the streambed filtering capacity.