H23G-01 INVITED 13:40h
Power-law behavior of transient storage residence time distributions from a multi-scale stream tracer test in Lookout Creek, Oregon
We quantified the residence time distribution (RTD) of transient storage at increasing spatial scales in Lookout Creek at the HJ Andrews Experimental Forest, Oregon. A longitudinal tracer test was conducted with an in-stream injection of rhodamine WT over 78 hr, followed by monitoring at 7 locations (1-14 km downstream) for 5 months. Three additional tracer tests were carried out over shorter reaches, between each of the major stream confluences, to further quantify transient storage and to separate geomorphic variability from other possible scale-dependent factors. The results clearly show power-law RTDs (i.e., $c \sim t^{-k}$), but not a single, "universal" RTD or scaling relationship. The exponent, $k$, ranges from 1.35 to 2.0, and appears to decrease downstream. Observed changes in the power-law RTD may result from characteristic changes in channel morphology or the degree of geomorphic complexity, which are related to stream size. Downstream reaches exhibit a larger delay between the peak concentration and power-law tailing compared to upstream reaches, possibly indicating a smaller hyporheic zone relative to stream size.
H23G-02 13:55h
Hydrodynamic interactions of free-flowing fluids and pore-fluids in bedforms
The physical and chemical complexity of the interface between porous bed interstitial water and surface water, sometimes referred to as the {\it hyporheic zone} in river-aquifer systems, has yet to be understood in detail. But, we do know that ecologically and environmentally significant processes occurring in these zones control the distribution of solutes, colloids, and dissolved gases from ripple to global scales. In fact, previous model computations show that the entire ocean volume could be recycled through such systems in 14,000 years. The biochemical processes occurring in these areas are mediated by fluid flow, as mass diffusive flux is usually orders of magnitude smaller than that of advective transport. Thus, a holistic view of these systems necessarily begins with a comprehensive knowledge of the hydrodynamics. The coupled fluid flow in open areas and their underlying porous bedforms are examined in this study. We numerically simulate steady viscous flow in these coupled systems by solving the Navier-Stokes and continuity equations that govern the free area and then use the pressure solution from this domain as a boundary for the porous subsurface domain governed by the groundwater flow equation. Numerical experiments were used to determine fundamental relationships between bedform geometry, free area Reynolds number ({\it Re}), exchange zone depths ({\it d}), and total fluxes through the bed surface ({\it q}). The results are as follows: 1) {\it d} and {\it Re} are functionally related through a Michaelis-Menten saturation-growth -like model; 2) {\it q} and {\it Re} follow a quadratic relationship; 3) {\it d} scales linearly with bedform length ({\it L}) and nonlinearly and non-monotonically with bedform height ({\it H}); 4) {\it q} increases as bedforms get steeper (high {\it H}/{\it L}); 5) the relative location of the bedform crest becomes a more important factor as {\it Re} increases; {\it d} is minimized when the crest is in the middle of the bedform and is maximized when the crest is near the downstream end of the bedform; 6) {\it q} is minimized and stabilizes when the crest is near the downstream end of the bedform; 7) eddy reattachment points correspond to subsurface flow divides; flow cells may cross between bedforms; 8) subsurface velocities drop dramatically with depth; the velocity at a point located directly underneath the crest and at the same elevation as the trough is less than 3% of the velocity at the crest.
H23G-03 14:10h
Hyporheic Exchange in Gravel-Bed Rivers with Pool-Riffle Morphology: A 3D Model
The hyporheic zone is a saturated band of sediment that surrounds river flow and forms a linkage between the river and the aquifer. It is a rich ecotone where benthic, hyporheic, and groundwater species temporarily or permanently reside. Head gradients along the streambed draw river water into the hyporheic zone and expel pore water into the stream. This process, known as hyporheic exchange, is important for delivering nutrients, oxygen and other solutes to the sediment, and for washing away waste products to support this ecotone. It is an essential component of the carbon and nitrogen cycles, and it controls in-stream contaminant transport. Although hyporheic exchange has been studied in sand-bed rivers with two-dimensional dune morphology, few studies have been conducted for gravel-bed rivers with three-dimensional pool-riffle geometry. The hyporheic zone of gravel-bed rivers is particularly important for salmonids, many of which are currently at risk world wide. Salmon and trout lay their eggs within the hyporheic zone for incubation. After hatching, the alevins live in the gravel before emerging into the stream. The upwelling and downwelling hyporheic fluxes are intense in these streams due to the highly permeable sediment and strong head variations forced by shallow flow over high-amplitude bed forms. Moreover, gravel-bed rivers show a wide range of flow regimes that change seasonally and have strong effects on hyporheic exchange. To study this exchange, we used four sets of pool-riffle geometries in twelve recirculating flume experiments. We kept a constant bed-form wavelength, but changed the bed-form amplitude and imposed three discharges, covering a wide range of hydraulic and geometric characteristics. Hyporheic exchange was predicted from a three-dimensional model based on bedform-induced pumping transport, where the boundary head profile is the pressure head distribution at the sediment interface, measured with an array of mini-piezometers buried within the streambed. Hyporheic flow was modeled as a Darcy's flow, confining the turbulent effects to a thin boundary layer at the sediment surface. Predicted values of hyporheic exchange were compared with measured values determined from both salt and fluorescein tracers. Observed and predicted values of exchange show good agreement, suggesting that the advective process induced by the bed forms is the major mechanism for exchange of surface and subsurface waters. In field studies, it is easy to measure the water elevation, so we tested the performance of the water-surface elevation as the boundary condition for hyporheic flow, and found agreement with the measured exchange only for low bed-form amplitude and high flows (i.e., when topographic effects are minimized).
H23G-04 14:25h
Modeling Hyporheic Flux Along a Second-Order Semi-arid Stream: Red Canyon Creek, Wyoming
Models of near-stream hyporheic exchange flows are difficult to prepare because geomorphic stream features and adjacent subsurface characteristics both affect groundwater-surface water interaction. Inverse models of the results of in-stream tracer tests characterize net short time-scale hyporheic exchange along reaches, but not the actual physical processes driving the exchange. In contrast, numerical groundwater flow models simulate near-stream and hyporheic flow driven by hydraulic gradients from a physical process perspective. In this paper, we present a three-dimensional MODFLOW model of hyporheic exchange along a lower riparian reach of Red Canyon Creek, Wyoming. We calibrated the model results to hydraulic head measurements from $>$ 30 monitoring wells, piezometers, in-stream mini-piezometers, and to changes in stream discharge measured by in-stream tracer tests. We also simulated hyporheic flow paths with MODPATH (a particle-tracking package), from which we obtained residence times of water parcels in the hyporheic zone. Hydraulic gradients around in-stream flow obstructions, such as beaver dams, and through meander bends, cause most near-stream hyporheic exchange (residence time $<$30 days). Hyporheic residence times $<$10 days occur only along flow paths around beaver dams. We also simulated stream solutes moving into the subsurface with MT3D, a solute transport package, and operationally defined the hyporheic zone as places where solute concentrations were equal to or greater than 10% of the stream water concentration after a 10-day model simulation. The results of this modeling agreed with MODPATH; solutes move both horizontally and vertically from streams into the subsurface behind debris dams, which create hydraulic steps in the subsurface and surface flow systems.
H23G-05 14:40h
Application of Time-Series Method for Estimating Surface Water - Groundwater Interaction from Streambed Thermal Records
Established methods for estimating seepage from streambed thermal methods, including the use of forward models, can be time consuming, generally include relatively short time periods, and may require independent determination or calibration of hydraulic properties. We apply a newly developed method for interpretation of streambed thermal data, using long records from multiple depths. We use forward models to define relationships between seepage rate and thermal response with depth, generate type curves, and apply modeled relationships to measured temperature records to generate long-term estimates of streambed seepage. Spectral analyses of streambed thermal records from different depths show clearly that the diurnal period has the greatest power, and that there is generally strong coherence between thermal records collected from different depths at a single location. The frequency content of propagating temperature signals does not change significantly with depth, but variations in the phase and amplitude of temperature changes are well explained by coupled heat and fluid flow across the streambed. Data are filtered with a cosine-taper, band-pass filter to extract the dominant diurnal signal prior to analysis. Data from models used to generate type curves are run through the same filter to minimize errors associated with filter leakage. We have completed a series of forward models under different conditions to assess the sensitivity of the method to relative and absolute sensor depth, sediment thermal properties, abrupt changes in seepage rate, irregular thermal variations at the stream-sediment interface, and multi-directional subsurface flow. For the smallest sensor spacing, the largest range of seepage rates can be derived using the amplitude ratio. In contrast, for the largest relative depth, the largest range of seepage rates can be derived using the phase shift. Abrupt changes in seepage rate that are shorter than one day tend to be smoothed during the filtering process, as expected, but longer, more gradual changes are well resolved. Given typical streambed thermal properties, the time series method for quantifying seepage works best for specific seepage rates on the order of -3.0 to 2.5 m/day (negative down), which should allow use in most natural systems.
H23G-06 14:55h
Spatial variability of induced ground-water recharge beneath the Russian River, California
The Sonoma County Water Agency extracts water from the alluvial aquifer adjacent to and beneath the Russian River via large-volume Ranney-type collector wells. To aid in this extraction, the stage of the river is increased approximately 3 meters by an inflatable dam. In addition, raising the dam allows water to be diverted into infiltration basins that are located adjacent to the river. Removal of aquifer water induces large fluxes from surface water to ground water through the beds of the infiltration basins and the river. Total extraction during maximum summer withdrawals via five collector wells indicates an average flux from surface water to ground water through the riverbed and infiltration basins of 153 cm/d. Measurements of flux using in-river and in-pond piezometers, diurnal sediment-temperature data, and seepage meters, indicate that actual seepage fluxes are spatially variable and large seepage fluxes are concentrated in specific locations, some of which may not be intuitive. For example, we expected greatest induced seepage fluxes to occur above laterals that extend beneath the river and deliver water to a collector well. Seepage flux along a transverse transect of the riverbed that was located above laterals from one of the collector wells averaged 10 cm/d. At the same time, seepage flux along a transect that was 500 m upstream, and farther from the influence of the collector-well system, averaged 40 cm/d. Seepage fluxes from the central portion of one of the recharge basins averaged 3 cm/d whereas seepage fluxes near the margin of that infiltration basin averaged 250 cm/d. Seepage fluxes derived from in-stream-piezometer Darcy calculations were surprisingly consistent with seepage fluxes derived from seepage-meter measurements. Seepage fluxes derived from temperature measurements were slightly less comparable to the piezometer and seepage-meter measurements. The 121 cm/d average of all seepage-flux measurements was similar to the spatially-integrated rate (153 cm/d) based on the volume of water extracted from the river by the pumping wells divided by the affected area of the riverbed and the flooded infiltration ponds.
H23G-07 15:10h
On the interaction between the Lower Jordan River and the shallow aquifer system of the Jordan Valley
The study describes the hydrology and chemistry of the shallow groundwater along the west side of the Lower Jordan River in its north section, by means of piezometers installation. We have installed 7 piezometers at different locations near the west bank of the Jordan River and added five reference points inside the river itself. The immediate purpose of these installations is to measure the groundwater level, to calculate the water gradient, to estimate the amount of water that enters the river, to sample the shallow groundwater, and provide a chemical analysis of these samples. The study area was divided into three fields, north, middle, and south where in each field a number of piezometers where drilled and one or two reference points were placed in the river. The highest gradient of the hydraulic head was measured in the north field and reached 5 m over a distance of ~500 m. This may be explained by the alluvial fan of the largest stream in the area. In the middle field the gradient of the hydraulic head was between 0.6 and 0.8 m over a distance of 170 - 380 m and in the south field we measured the smallest hydraulic gradient that was less then 0.1%. The chemical characteristics of the river show a chloride decrease from ~1900 mg/L to ~1500 mg/L and then a slight increase to ~ 1600 mg/L. Since the river is not only affected by the groundwater but also by surface inputs from the west and the east, and by intensive pumping, the interpretation and the comparison between the river data and the piezometers is complex. Never the less, there is a significant difference between the three fields. Where the chloride concentration in the north field is lower than that of the river the chloride concentration in the other two fields is higher than in the river. As the gradient of the hydraulic head found in the north field is high it is likely that some part of the river modification may be attributed to the sampled water in the piezometers. On the other hand, the very high chloride concentration found in the other two fields, have only little effect on the river concentration since no significant convection influx exists and most of the processes are diffusive and slow. The sulfate concentration increases in the river from 380 mg/L to 450 mg/L. Except for one piezometer located in the middle field, all water samples that were pumped from the piezometers show higher sulfate concentration then those in the river. This significant result supports the hypothesis that the river sulfate increases due to its interaction with the surrounding groundwater system.
H23G-08 15:25h
A Site-Specific Analysis of Hyporheic Mass Transfer and Residence Times
Hyporheic exchange is an ecologically important process, controlling the nutrient supply in the upper river bed sediments and with this benthic habitat quality. To analyze the controlling parameters and their influence on mass transfer and residence times, hydraulic exchange at a riffle-pool-sequence in the River Lahn, Germany, was analyzed using HEC-RAS to simulate the surface water flow as a boundary for the subsurface flow and MODFLOW, MODPATH, and MT3DMS to reproduce the transport in the subsurface. Field studies provide the basic data for input to these models. Hydraulic heads within the interstices, solute transport, and residence times of surface water in the subsurface were simulated. To reproduce spatial and temporal changes in fluvial sediment properties, we determine sediment parameters for different surface water flow rates and distribute the parameters smoothly over the longitudinal profile of the river reach. The sediment parameters are calibrated to fit measured hydraulic heads and tracer throughflow curves from the pool-riffle-sequence. The results show that although the surface water flow plays an important role for the hyporheic exchange at the study site, the influence of the sediment properties on the exchange is larger. About the same percentage of surface water enters the hyporheic zone for each of the different flow conditions. Densely packed sediments reduce the exchange. Anisotropy increases the hyporheic exchange for low and medium surface water flow and has the opposite effect for high flow. Residence times were short at higher flow and long at low flow. Finally, we investigated the effect on the exchange processes of modifications of the river bed in a hypothetical case study. The effect of the local replacement of sediment by small cross-section sills was analyzed in comparison to the natural river bed conditions. In general, the presence of the sills increased the exchange and focused it around the sills.