H13D-1534
"Sticky Business": the Influence of Surface Biofilm on Particle Deposition and Infiltration in Streams
Fine particulate matter is an important component of many streambed processes. For example, the deposition and infiltration of fine inorganic sediment has been repeatedly shown to degrade benthic habitat for fish and other organisms (Hynes, 1970). In contrast, fine organic particles are a significant source of carbon to benthic organisms (Webster et al., 1987). The mechanisms and rates of particle entrainment and deposition are highly complex and ill-predicted by simple physical relations. For example, a number of field studies have shown that measured rates of particle deposition often differ from still-water particle settling velocities calculated from particle size and density (e.g. Cushing et al., 1993). Several studies have proposed that adhesion of particles to surface biofilm may explain why deposition rates are faster than predicted, but few have investigated this phenomenon (Battin et al., 2003). In addition, although biofilms have been shown to significantly alter near-bed and interstitial flow velocities (Dodds and Biggs, 2002), the effect of these changes on particle depositionhas not been explored. Biofilm is pervasive in rivers and streams throughout the world, thus it may play an important, and heretofore underestimated, role in the deposition of fine particles to the streambed. This study tests the hypothesis that biofilm amount and structure may alter fine particle deposition, entrainment, and infiltration by either direct adhesion or by changes to near-bed hydraulics. A series of experiments are being conducted in a small recirculating flume to test how the amount and structure of surface biofilm influences the water column distribution, surface deposition, and infiltration of fine particles under two different flow levels (‘high' and ‘low'). Two types of surface biofilm are being tested: open-weave, filamentous assemblages and low-profile, mucilaginous forms; both are compared to a reference substrate without biofilm. Natural rocks hosting filamentous assemblages are collected from nearby channels and used to replace the surface layer of flume substrate. Low-profile forms are cultivated on a cobble-gravel substrate in an artificial stream system and tested at each of three growth stages. Immediately following a dose of fine particles (< 125 µm ground silica), short-term rates of particle deposition are determined by continuous measurements of near-bed particle concentrations. Hourly vertical concentration, grain size profiles, and 3-D velocity profiles are measured for the duration of the experiment (8h). Infiltration rate is determined from bed samplers sealed from the flow at 1, 2, 4, and 8 hours. At the end of each experiment, three surface samples are collected and analyzed for ash-weight, ash-free dry weight, and chlorophyll a. Four stratified bed samples and samplers are removed, dried, sieved and weighed to determine the amount and depth of fine particle infiltration into the bed with time and surface condition. Preliminary results indicate that particle deposition is significantly enhanced by biofilm presence relative to surfaces without biofilm; this effect is greater for mucilaginous forms. However, near-bed shear stresses and velocity profiles are not significantly altered by either mucilaginous or filamentous forms; bed topography has a dominant effect. Results also indicate that particle deposition is more strongly influenced by surface condition (e.g. particle size, biofilm presence) than by flow level or duration. Particle infiltration below the surface, however, is more a function of pore space and flow level than biofilm coverage on the streambed surface.
H13D-1535
Assessing Streambed Recharge Under a Wide Range of Flow Conditions Using Time-series Analysis of Streambed Thermographs
We evaluate downward streambed seepage in Corralitos Creek, a second-order stream in central coastal California, using differential gauging and time-series analysis of long-duration, streambed temperature records. Discharge and thermal data were collected throughout the water year at multiple locations along a 9.73-km experimental reach. Differential discharge measurements were made at various locations along the reach using the velocity-area method andusing continuous gauges installed at fixed stations. The analytical method used to interpret streambed thermal data uses changes in the phase and amplitude of subsurface diurnal temperature variations to infer the direction and rate of seepage once per day. Differential discharge data indicate that Corralitos Creek loses 0.1 to 0.2 m 3/s when channel discharge into the reach is \?1.5 m3/s, with most of the loss occurring where the stream is in direct contact with permeable aquifer units. Higher channel discharges occur mainly in the winter rainy season, during and soon after precipitation events. Differential gauging is not practical in Corralitos Creek during high flow events, but seepage rates have been calculated from thermal data in the area of greatest differential discharge throughout the water year. Streambed seepage rates calculated from thermal data are -1 to -4 m/day at times when differential discharge data can be collected, and similarly large, negative seepage rates continue during periods of high discharge, when differential discharge measurements are not practical. The area of rapid streambed seepage goes dry for periods of weeks to months late in the water year, but seepage rates of up to -1 m/day are documented soon after the streambed rewets, indicating a rapid transition from dry to actively-recharging conditions in the streambed. Geochemical data indicate that much of the documented streambed seepage along this section of Corralitos Creek recharges regional aquifers. When summed overthe entire water year, channel losses on a short reach of Corralitos Creek (watershed area ~60 km2) comprise up to 20% of the sustainable yield of the Pajaro Valley Ground Water Basin (area ~390 km2).
H13D-1536
A Comparison of In-Channel Dead Zone and Hyporheic Zone Transient Storage Parameter Estimates Between a 1st and 5th Order Stream
A major enhancement to our understanding of how watersheds function would be the ability to discriminate between in-channel dead zone ( DZ) and hyporheic zone ( HZ) transient storage, and an evaluation of how these properties scale across stream orders. The nature of DZ storage is to display faster exchange rates with the main channel and less overall sediment contact time than HZ storage. These differences have great significance to many in-stream processes such as nutrient cycling. The combination of high slope, coarse bed material and fluvial structure endemic to many 1st order streams can provide greater forcing of hyporheic flow paths than occurs within the lower gradient 5th order streams. Conversely many 5th order reaches exhibit large side pool and back eddy DZ areas not common along 1st order streams. This study builds on existing methods to delineate the DZ and HZ from the integrated signal of a conservative solute's breakthrough curve ( BTC). Data for this comparison were collected over the summer of 2007 within the Ipswich River watershed, a basin which drains into Plum Island Sound on the north shore of Massachusetts, USA. The conservative solute NaCl was injected into both a 1st order medium gradient stream and a 5th order low gradient stream. The BTCs collected in thalwegs from the NaCl injections were simulated using a version of the solute transport model OTIS containing two zones of transient storage. Hydrometric measurements of stream velocity were used to estimate average main channel cross sectional area ( A) and DZ cross sectional area ( ASDZ) for each reach to constrain parameter estimates and avoid model equifinality between the storage zones. Initial values for the exchange rate between main channel flow and DZ storage ( αDZ) were estimated from DZ BTCs. Our results indicate that although the overall storage zone is much larger in proportion to the main channel for the 1st order reach than for the 5th order reach, the percentage of median transport time due to storage as evaluated by the FMED200 is smaller. Additionally the integrated DZ and HZ α is faster for the 5th order reach, which in agreement with our original hypotheses that DZ storage is the dominant transient storage process in high order streams. Further analysis across stream orders may reveal useful relationships between channel form and hydrologic function.
H13D-1537
Spatial and Temporal Patterns of Sediment Transport in a Small Stream
In this paper we attempt to link the distribution of shear stress and sediment sources/destinations at the reach- scale in East Creek, a small, forested, gravel-bed river in coastal British Columbia. The study reach is characterized by low sediment supply where a majority of the transported material comes from the bed of the channel. The shear stress distribution is modeled using River2D, a numerical hydraulic flow model. The model is calibrated using flow and channel topography data collected at the field site. On average, the site receives 9 to 12 rain-dominated, sediment mobilizing flood events per year, ranging from moderate flows to bankfull discharge. We identify sediment sources and destination zones by annually recovering 1400 magnetically tagged tracer particles placed in two distinct channel morphologies (rapid (plane bed) and riffle-pool). Sediment mobility depends largely on within-reach sediment storage and channel morphology. By measuring the distance of travel and depth of burial and estimating bed shear stresses, a morphological model of preferential sediment transport and storage is developed.
H13D-1538
Effects of In-Stream Geomorphic Structures on Stream Temperatures and Streambed Heat Flux Through Induced Hyporheic Exchange
In-stream geomorphic structures such as debris dams and log dams are common in undisturbed streams, are often installed as part of stream restoration projects, and can significantly enhance hyporheic exchange. While this exchange is important ecologically for many reasons, including affecting summer temperatures in streams that are impacted by loss of riparian shade and climate change, the thermal impact of these structures is relatively unknown. We conducted field experiments with a variable-height weir-type structure in a stream monitored with a three-dimensional temperature and water level sensor network. Results indicate that structures induce a coherent hyporheic flow cell that varies predictably with structure height. Data from the sensor network were analyzed to estimate the impact of structure presence and size on advective and conductive heat transfer processes between the stream and the hyporheic zone. These thermal process data were used to estimate the net thermal impact of structure-induced hyporheic exchange on stream temperatures. Results indicate that 1) both advection and conduction have a net cooling effect on the stream in summer, 2) advection tends to increase with structure size, and 3) streambed hydraulic conductivity determines the relative importance of advection and conduction. However, the net cooling effect of these structures via induced hyporheic exchange can be outweighed by warming effects due to exchange of heat with the atmosphere.
H13D-1539
Impact of Near-Stream Ground Water Pumping on Streambed Fluxes
A three-dimensional, multi-phase flow model of the Russian River Bank Filtration Facility near Forestville, California was developed to investigate the impact of ground water pumping from two horizontal collector wells on the streambed fluxes and the stream-water travel times to the collector wells. Simulations were conducted at different aquifer to streambed permeability ratios. When the streambed was modeled as a homogeneous layer and the aquifer beneath was saturated, the simulations showed that the streambed velocities increased across the channel, with the largest velocities on the east bank near the collector wells. At aquifer to streambed permeability ratios ranging between 10 to 325, an unsaturated region developed below the streambed near the collector wells, which had an impact on streambed velocities. Lower velocities were observed in regions where the aquifer was desaturated below the streambed compared to the cases where the aquifer was saturated. When the total volumetric streambed fluxes were calculated at different aquifer to streambed ratios, the simulations indicated that in some cases, the streambed fluxes increased as the streambed permeability decreased. However, the stream-water residence times increased and the fraction of stream-water that reached that the wells decreased as the streambed permeability decreased, indicating that a higher streambed flux did not necessarily correlate to greater recharge of stream water around the wells.
H13D-1540
Associations Between Groundwater-Surface Water Dynamics and Coaster Brook Trout Spawning Habitat in the Salmon Trout River, Marquette County, Michigan
The Salmon Trout River (STR) is the only river on the south shore of Lake Superior known to sustain a reproducing coaster brook trout (CBT) population. Related studies demonstrate that brook trout tend to select spawning sites, based on the presence of groundwater discharge into the river. The results of these studies also suggest that groundwater presence is vital to the reproductive success of CBT. Previous studies of the STR have characterized the life history strategies and ecology of CBT, but to date no study has investigated the influence of groundwater on CBT spawning habitat in the STR. We hypothesize that spatial distributions of groundwater inflows through river-bottom sediments are a critical factor in the selection of spawning sites. In this study, high-resolution data collection methods are implemented to quantify the interaction between the groundwater and surface water in order to verify the presence or absence of groundwater discharge into the river at sites that support a reproducing population of coaster brook trout. By independently inverting temperature and pressure measurements the exchange of water between groundwater and surface water can be simultaneously analyzed, permitting a more precise estimate of groundwater velocity. An array of 1.5 inch diameter PVC piezometers, are installed into the river banks and bottom sediments in two active spawning sites. Transects containing three piezometers a piece are located in the active section of each site, as well as immediately upstream and downstream of the active spawning sections. Each piezometer is equipped with multiple temperature sensors placed at incremental depths (0 to 4 feet) beneath the riverbed. Manometers are used to monitor pressure gradients between the groundwater and surface water at depths equal to the placement of temperature sensors. The study will span the course of one full year beginning in August of 2007 and ending in August of 2008. Preliminary data will be presented to show how simultaneous analysis of temperature and pressure data can be used to achieve refined estimates of groundwater velocity.
H13D-1541
Quantifying Radiative Inputs to Headwater Streams in Recently Harvested Forests
Removal of streamside forest canopy during timber harvest can lead to large increases in the amount of shortwave (solar) radiation reaching the stream surface. Shortwave radiation increases may be partially offset by longwave (thermal) radiation decreases due to the decrease of the vegetation canopy. Rapid recovery of herbaceous streamside vegetation in recently harvested areas may act to negate the increased shortwave radiative inputs caused by tree canopy reduction. A quantifiable difference in the radiative inputs affecting streams was found by measurement of incoming shortwave and longwave radiation. Arrays of 10 shortwave and 2 longwave radiation sensors were installed in first-order stream systems for 24-hour periods of high radiation loading to quantify the contribution of incoming solar and thermal radiation in headwater systems. Sensors were installed in unimpacted forested reaches, a partial cut reach (~50% canopy removal), densely vegetated and sparsely vegetated 5 year-old clearcut reaches, and a simulated freshly cut (absent of any streamside vegetation). Results indicate that thermal radiation is the dominant source of radiation in all treatment reaches. All-wave incoming radiation was reduced by 36 to 39% in the undisturbed reaches, compared to open areas. The partial cut area and heavily vegetated 5-year old clearcut were very similar to intact forested reaches, with 37% and 38% all-wave reduction relative to open areas. The sparsely vegetated 5-year old clearcut exhibited a 16% reduction in all-wave radiation, and the simulated fresh clearcut exhibited a 9% reduction, relative to open areas, due to shading by overhanging banks. Results from this study indicate that recently clearcut headwater streams are able to recover to pre-harvest shading characteristics if adequate regrowth of herbaceous vegetation occurs.
H13D-1542
Modeling Streambed Hyporheic Exchange Using a Spectral Scaling Based Pumping Model
Modeling solute transport in rivers is critical to evaluating the transport of contaminants, nutrients, and other water-borne constituents, and thus is inherent to the study of ecosystems and water quality. Our objective is to enable prediction of hyporheic exchange at the bedform-to-reach-scale based on readily measurable system characteristics. We employ a spectral scaling approach as the basis for a generalized analysis of topography- induced exchange in river systems. The model includes the lateral hyporheic zone in addition to the flow directly beneath the streambed. This approach encompasses a larger range of scales than is normally considered in predictive exchange modeling, including subsurface flow induced by very small scale bedforms to much larger features such as meanders. The primary input parameters for modeling are in-stream velocity, sediment permeability and porosity, and detailed measurements of the stream channel topography. The primary outputs are a flow path analysis, water exchange flux across the sediment boundary, and subsurface residence time distribution, which can be compared against field data. Having spatially explicit information allows us to evaluate the contributions of various classes of streambed features in overall hyporheic exchange, a key advantage over the more empirical approach of conducting a stream tracer experiment. The solution method involves Fourier fitting of the topography followed by calculation of the boundary head distribution and then the subsurface head and velocity fields. A sophisticated geometric transformation is required to accurately represent areas of high pressure on the upstream edge of submerged topographic features in meandering channels. We have implemented a Schwarz-Christoffel conformal mapping procedure for this purpose. This method provides consistent results regardless of the orientation of the stream, as well as a reasonable estimation of the three- dimensional boundary head distribution. Lateral (floodplain) exchange is captured by superimposing the Fourier solution for the in-channel hyporheic exchange on a finite-difference solution for broader stream-groundwater interactions. The model is applied to several highly detailed two- and three-dimensional datasets from laboratory flumes, and also to one field site in a headwater agricultural stream.
H13D-1543
Spatial and Temporal Variability of Hydraulic Properties in the Russian River Streambed, Central Sonoma County, California
The reach of the Russian River flowing through Sonoma County, CA, is important to fisheries and recreations, as well as being essential to the water resources infrastructure of the county. An improved understanding of the manner in which streambed sediments impact rates of ground-water recharge is essential in optimizing withdrawals without increasing potential impacts on fishery habitats and recreational needs. Temporal and spatial variations of flux and vertical hydraulic conductivity (Kv) were measured in the streambed along the Russian River at multiple locations. In-situ flux and Kv measurements were made using a modified seepage meter equipped with piezometers during monitoring events performed in June 2003, September 2003, and March 2004. Additionally, bulk sediment samples were collected during the monitoring events to characterize the grain size distribution of the streambed. Three different streambed locations (near-bank, midpoint, and thalweg) were monitored and sampled at five different sample locales in a 20-km reach of the Russian River. Vertical hydraulic conductivity of the streambed ranged from 8.55x10-5 cm/sec to 1.52x10-1 cm/sec. Significantly (p<0.05) higher values of Kv were found near the banks of the Russian River, and Kv increased (30% to an order of magnitude) after the winter storm season of 2004. Flux varied from -240 to 600 cm/day, which indicates both gaining and losing reaches of the stream occur in our study area. These findings will assist in developing a MODFLOW ground-water flow simulation that incorporates the variable streambed conductance values determined along this reach of the Russian River.
H13D-1544
The Influence of Sediment Size on Dissolved Inorganic Nitrogen in Hyporheic Mesocosms
Dissolved inorganic nitrogen (DIN) is composed of nitrate, nitrite and ammonia. The availability of DIN in streams is important to water quality and to stream ecosystems in general because at low concentrations DIN can limit primary productivity, while at high concentrations DIN contributes to eutrophication. An important location of nitrogen transformations in streams is the interstitial spaces between streambed sediments. As surface water passes transiently along shallow subsurface flow paths, the activities of interstitial biofilms and invertebrates alter the form and availability of nitrogen. Heterotrophic consumers release nitrogen as ammonia. Nitrifying bacteria further transform ammonia to nitrate under aerobic conditions. While under anaerobic conditions denitrifying bacteria can transform nitrate to nitrogen gas. Studies in natural streambeds have suggested the size of streambed sediments can influence whether the streambed serves as a net source or sink for DIN in transient sub-surface water. We tested whether the size distribution of streambed sediments in isolation from the effects of streambed topography and groundwater upwelling could alter DIN production or uptake as stream water passed transiently along interstitial flow paths. We filled pipes with contrasting sediment mixes, and placed them in a stream so that surface water flowed through the sediments. After an incubation period allowing biofilm development, we measured nitrate + nitrite and ammonia in interstitial water drawn at several distances along the pipes. These measurements revealed that both sediment types were net sources of DIN, with the majority of DIN production occurring near the beginning of subsurface flow paths. Concentrations of DIN were greater in the finer sediments at any given distance along the pipes. However, higher surface water exchange rates through the coarser sediments caused net DIN production in the coarse sediments, and therefore DIN release to surface water, to equal DIN production in the fine sediments.
H13D-1545
Assessment of Stream-Aquifer Interactions Using A Three Dimensional Analytical Solution
We developed a three-dimensional semi-analytical solution of groundwater flow from an adjacent water body to an aquifer with finite vertical dimension. The developed solution improves previous two-dimensional solution by considering approximated geometry of upper-lying water body which is conventionally over-simplified. The solution is derived based on the method of Green's function where production rule and the method of superposition are used. In the solution, vertical Green's function is derived in the Laplace domain and numerical inversion is applied to estimate the value in time domain. Then convolution is used with time varying source function and Green's functions of horizontal directions. The developed solution can be applied to any hydrological setting such as two neighboring water body with finite conductance (ex. surface-groundwater interactions, two aquifer interactions, impact of canal to surrounding aquifers, etc.). In the analysis using the developed solution shows that interfacial storage is not important in practical applications. Also, we found that early-time responses of aquifers with finite and infinite vertical dimension behave identical.
H13D-1546
Subsurface Stream Processes and Spatiotemporal Controls on Nutrient Cycling Upper San Pedro River, AZ
In semi-arid lotic systems, little is known about the mechanisms that control nutrient variability and availability. Many studies have demonstrated that streambed upwelling zones are important for stream nutrient cycling while other evidence indicates that episodic floods carry high concentrations of organic matter and nutrients; as a result these nutrients are reflected in subsurface waters of the river and later return to the stream through upwelling. The main goal of this research is to answer the following question: How does the spatial and temporal variability of surface water nutrients change in relation to subsurface processes and time since last flood? To address this question, chemical measurements together with paired surface and sediment water temperature measurements were conducted using a nested hierarchy approach on a 10km river segment at various stages of post flood succession. Sulfate to chloride ratios of stream waters indicate that during non-flood periods the contributions of groundwater to stream flow become more dominant over winter or summer flood flows as the time since flood increases. A near stream spring found only in November 2006 sampling campaign shows that the sulfate to chloride ratios closely resemble baseflow water but with nitrate concentrations that where similar to the highest nitrate concentration in November samples. In addition, we have found that the spatial variability and availability of nitrate increases as time after flood increases and that this variability appears to be influenced by upward subsurface hydrologic flux. Moreover, location and spatial scale appears to influence these relationships; at larger scales the variability of nutrients is controlled and dampened by in-stream processes while at smaller scales is more variable and controlled by the direction of water exchange. Our results suggest that during post-flood succession upwelling waters, like the spring found on November 2006, develop into the dominant control of stream water chemistry in the San Pedro River.
H13D-1547
Determining Water Fluxes Across Surface Water/Groundwater Interfaces Using Temperature Data and Numerical Models
Daily water releases from an upstream dam on the Deerfield River create significant differences in stage of up to .5 meters. These differences cause the river to change from a gaining stream to a losing stream on a daily basis. In this study, we use Thermochron iButtons to measure changes in groundwater temperature to assess water fluxes between the river and surficial aquifer. The causes for the temperature changes observed in the groundwater are (1) diurnal heating of surface water as air temperatures increase, and (2) changes in temperature caused by an influx of reservoir water from dam releases. We deploy a pressure transducer with iButtons which are inserted into the riverbed and riverbank at 10 cm intervals to create thermal cross sections. To account for heterogeneity in the riverbed material, the iButtons are installed along three cross section lines. One line is in coarse sand and gravel, another is in sand and silt, and the third, which is the control, is on an unregulated river that flows into the Deerfield. The control, which is downstream of the first two sites, only experiences temperature variations caused by diurnal air temperature changes. The temperature data are used to calibrate coupled numerical models of groundwater flow and heat transport. Through these models we can derive hydraulic conductivity estimates of the riverbed material, groundwater velocities during release events, and water fluxes across the surface water-groundwater interface. Groundwater velocity and flux estimates are useful for evaluating mass transport mechanisms and nutrient cycling in the near stream environment.
H13D-1548
Impacts of stream-aquifer connectedness on bank infiltration, seepage, and solute transport during semi-arid flooding
Transient storage of flood water in stream banks during high flow conditions plays a unique role in semi-arid environments. In ephemeral streams, water infiltrated during flood events is a significant component of the water balance, is essential for sustaining low flows, and plays a critical role in sustaining a riparian habitat. This work examines the role of steam-aquifer connectedness in affecting differences of floodwater distribution in the stream banks. We consider a range of vadose zone depths from closely-connected streams, common to humid areas, to increasingly disconnected streams, common to semi-arid areas. Specifically, we address the question of how stream-aquifer connectedness affects infiltration rate, quantity of seepage, and distribution of solutes. We use the variable saturation code HYDRUS 2-D to simulate transient flow and solute transport. These simulations show a significant change in water distribution for a short change (5 m) in water table depth. In particular, during flooding, cumulative infiltration increases sharply with increasing depth of the water table. During recession, total bank seepage decreases sharply with increasing depth of the water table. Capture zones determined from particle tracking show that only a small area that extends laterally above and below the base of the stream contributes to seepage. Solute transport in connected systems is predominantly horizontal through the shallow root zone. However, for thick vadose zones, a small fraction of solutes in the area of the root zone will be affected by infiltrating water, and those that are mobilized will move predominantly toward the water table, rather than back toward the stream. The maximum transitional depth between humid-like connected behaviors and those in a disconnected system is only 5 m. This implies that riverine ecosystems under pressure from various water demands can expect significant changes in water distribution in the root zone area for even modest lowering of the water table.