H41A-0284 INVITED 0800h
Modeling the Geochemistry of Red Mountain Creek, Colorado, With Implications for Premining conditions
In August of 2002, a synoptic water-quality study was conducted on Red Mountain Creek, an acid mine drainage stream in southwestern Colorado. Data from the study were used to calibrate OTEQ, a reactive solute transport model for streams and small rivers. OTEQ is formed by coupling the OTIS solute transport model with a chemical equilibrium submodel. The submodel is based on MINTEQ, a model that calculates the distribution of aqueous species under chemical equilibrium. The coupled model considers a variety of processes including advection, dispersion, transient storage, transport and deposition of water-borne solid phases, acid/base reactions, complexation, precipitation/dissolution, and sorption. Application of OTEQ to the low-flow dataset from Red Mountain Creek suggests that surface-water sources account for observed changes in stream geochemistry and that most solutes are transported conservatively throughout the study reach. Mass balance calculations and simulation results indicate that four mining-related sources account for 83, 70, and 69 percent of the observed metal loading for aluminum, arsenic, and zinc, respectively. A hypothetical estimate of premining water quality is obtained by performing an additional simulation in which the the four mining-related sources are replaced with a source that represents natural background. Simulation results suggest improved water quality under premining conditions, with increased pH, lower metal concentrations, and non-conservative transport. Despite this hypothetical improvement, dissolved metal concentrations remain elevated and pH remains below 5.0. This finding supports the idea that Red Mountain Creek was acidic and metal-rich prior to mining.
http://co.water.usgs.gov/otis
H41A-0285 INVITED 0800h
Hydrological Response of Sedum-Moss Roof
Eco-roofs are becoming popular for aesthetic reasons and also as units of stormwater systems. It is thought that such roofs with soil cover and vegetation reduces the total runoff, the peak flows and improves the quality of the roof water. Here are reported investigations of runoff from thin, 3-4 cm soil, extensive green roofs with sedum-moss in southern Sweden. The two-year study was performed on new roofs in the eco-city Augustenborg and also on nearby old vegetative roofs. The rain intensity and the roof runoff were measured with 5 min, or in some experiments with 1 min, resolution. The annual runoff from the eco-roofs was about half that from hard roofs and was close to that of small natural rivers. However, although most rainy days there was no or little runoff from the roofs, the highest observed daily runoff values were close to the daily rainfall. Runoff is initiated, when the soil is at field capacity. Thereafter the hourly runoff corresponds closely to the hourly rainfall. For short-term high intensity storms, the runoff peak is attenuated relative the rain intensity. The time of concentration for runoff was experimentally determined applying artificial rains on existing roofs and on experimental roof plots with varying slopes and using different drainage layers. The peak runoff from the roofs was found to correspond to the rain intensity over 20-30 minutes. The probability of high rain intensity is much higher than the probability of high runoff. When intensity-duration-frequency curves were constructed, runoff with 0.4 year return period corresponded to rain with 1.5 year return period. The influence of the slope of the roofs on the runoff peak was minor as was the effect of drainage layer. The vertical flow in the soil dominates the runoff process. The influence of extensive sedum-moss vegetated roofs on runoff quality was also studied to ascertain whether vegetated roofs behave as sink or source of pollutants and whether the runoff quality changes with roof age. The results show that in general vegetated roofs behave as a source of contaminants. With the exception of a 15-year old roof, the studied vegetated roofs contributed phosphate-phosphorus to the runoff. Some metals appeared in concentrations that corresponds to moderately polluted water. However, nitrate-nitrogen is retained by the vegetation and the soil.
H41A-0286 0800h
Modeling Dispersion in Tidal Channels
Dispersion in unstratified tidal channels is analogous to dispersion in streams. In a southern California inter-tidal wetland characterized by depths of roughly 1 m, a dye study illustrates non-Gaussian longitudinal mixing characteristics which are predicted using a depth-integrated numerical model that uses physically meaningful mixing parameters. Longitudinal dispersion is based on Elder's model and transverse mixing is based on observations by Ward. Mixing parameters are linearly related to the product of shear velocity and the depth. Flow resistance and dispersion are scaled by a common parameter, the bed roughness. The numerical model, based on the finite volume method, gives non-oscillatory flow and transport predictions and is stable for Courant numbers less than one and diffusion numbers less than one-half. The finite volume method facilitates use of physically meaningful mixing parameters since these may be arbitrarily small and not affect model stability.
H41A-0287 0800h
Sensitivity Analysis of a Reactive Transient Storage Model With Streambed Sorption Applied to Experimental Field Data
The transient storage model (TSM) has been widely used in simulations of stream solute transport and fate, with an increasing emphasis on reactive solute transport. In this study we perform sensitivity analyses of a reactive solute transport model (RSTM) that couples with a conservative TSM a formulation of sorption of a solute onto streambed sediments. The simulations were performed using the OTIS (One-dimensional transport with inflow and storage) model of solute transport for rivers and streams. The RSTM is analyzed to examine its effectiveness to reliably characterize stream and storage zone solute reactions from a previously reported data set. The field data are from a transport experiment with the addition of the conservative tracer chloride and the sorbing tracer strontium in Uvas Creek (Santa Clara County, California). Sensitivities of simulations to parameters within and among reaches, parameter coefficients of variation, and correlation coefficients are computed and analyzed. Our results indicate that 1) simulated values have the greatest sensitivity to parameters within the same reach, 2) simulated values are also sensitive to parameters in reaches immediately upstream and downstream (inter-reach sensitivity), and 3) simulated values have decreasing sensitivity to parameters in reaches farther downstream. Simulations of reactive solutes are shown to be equally as sensitive to model transport parameters and model reaction parameters; this being evidence of the control of physical transport on reactive solute dynamics. Similar to conservative transport analysis, reactive solute simulations also appear to be most sensitive to data collected during the rising and falling limb of the concentration breakthrough curve (CBC). In designing reactive stream tracer experiments, these findings have significance for appropriate sampling of the CBC.
http://co.water.usgs.gov/otis
H41A-0288 0800h
Physical and Numerical Experiments to Investigate the Influence of Dead-Water Zones on the Dispersive Mass Transport in Rivers
The prediction of transport velocities, maximum concentrations and skewness of a cross sectional averaged pollutant cloud leads to strong uncertainties, because, the influence of morphological heterogeneities on the transport characteristics is not completely understood. An important type of heterogeneities are dead-water zones, such as groin fields that are intensively used in various rivers to stabilize the main channel at a certain position, to increase the water depth in the main channel for shipping purposes in low water periods and to prevent bank erosion. In order to implement new Alarm-Models like the River-Rhine-Alarm-Model, to predict the mass transport in case of pollutant spill scenarios, expensive field studies are needed to calibrate these models. Hence, the predictions in cases of different hydrological conditions is very limited and inhibits the transfer of such a model to different river systems. In the present study laboratory experiments have been performed, in order to analyze the flow and mass transport characteristics in the presence of groin fields. Large-Scale-PIV measurements to analyze the mean flow field and the turbulence characteristics at the water surface have been performed as well as concentration measurements at single dead-water zones to determine the mass exchange. Experiments with different groin field aspect ratio and inclination angles as well as varied groin field volume have been conducted. The results show, that the groin field geometry, that can be expressed with a certain shape factor, has an influence on the mass exchange properties. With the help of a Lagrangian-Particle-Tracking-Method (LPTM), based on a random walk simulation, the experimental results obtained locally at a single groin field could be translated into the transport characteristics in the far field, comprising the effect of many groin fields, by analyzing the statistics of the virtual released particles during such a simulation. This transport model has been verified with the help of analytical solutions of the 1-D advection diffusion equation. The influence of dead-water zones has been parameterized in the LPTM using a transient-adhesion-boundary that leads to a certain retardation of particles that cross this boundary. The retardation effect is related to the mean residence time of tracer in a dead-water zone. In a further step the influence of changing boundary conditions along the flow on the persistence of the skewness of the cross-sectional averaged tracer distribution in longitudinal direction has been analyzed.
http://www.uvka.de/univerlag/volltexte/2004/11/
H41A-0289 0800h
Rejuvenating the Largest Treatment Wetland in Florida: Tracer Moment and Model Analysis of Wetland Hydraulic Performance
The Orlando Easterly Wetland (OEW), the largest municipal treatment wetland in Florida, began operation in 1987 mainly for reducing nutrient loads in tertiary treated domestic wastewater produced by the city of Orlando. After more than ten years of operation, a decrease in total P removal effectiveness has occurred since 1999, even though the effluent concentration of the wetland has remained below the permitted limit of 0.2 mg/L,. Hydraulic inefficiency in the wetland, especially in the front-end cells of the north flow train, was identified as a primary cause of the reduced treatment effectiveness. In order to improve the hydraulic performance of the OEW and maintain its efficient phosphorus treatment, a rejuvenation program (including muck removal followed by re-vegetation) was initiated on the front-end cells of the north flow train in 2002. The effectiveness of this activity for the improvement of hydraulic performance was evaluated with a tracer test and subsequent moment and model analyses for the tracer resident time distribution (RTDs). Results were compared to similar tracer tests conducted prior to rejuvenation activities. The models included one-path tank-in-series (TIS), two-path TIS, one-dimensional transport with inflow and storage (OTIS), plug flow with dispersion (PFD), and plug flow with fractional dispersion (PFFD). The hydraulic performance was characterized by both wetland hydraulic efficiency and the spreading of tracers. The results demonstrated that the rejuvenation considerably improved the hydraulic performance in the restored area. Also presented is a comparison of the wetland response between both bromide and lithium tracers, and the determination of the complete moments of residence time distributions (RTD) in cell-network wetlands.
H41A-0290 0800h
Determining In-Channel Transient Storage by Comparing Solute Transport in a Bedrock Channel - Alluvial Channel Sequence, Lookout Creek Basin, Oregon, USA
Current stream tracer techniques do not allow separation of in-channel (e.g., eddies) and out-of-channel (hyporheic) transient storage, yet this separation is important to understanding stream biogeochemical processes. We characterize in-channel transient storage with a rhodamine WT solute tracer experiment in a 304-m cascade-pool type bedrock reach with no hyporheic zone. We compare the solute breakthrough curve (BTC) from this reach to that of an adjacent 367 m alluvial reach with significant hyporheic exchange. In the bedrock reach, transient storage has an exponential residence time distribution with a mean residence time of 3.0 hr and a ratio of transient storage to stream volume of 0.14, demonstrating that at moderate discharge, bedrock in-channel storage zones provide a small volume of transient storage with substantial residence time. In the alluvial reach, though pools are similar in size, transient storage has a power-law residence time distribution with a mean residence time of $>$ 100 hr (estimated at nearly 1200 hr) and a ratio of storage to stream volume of 105. Because the in-channel hydraulics of bedrock reaches are simpler than alluvial step-pool reaches, the bedrock results are probably a lower end-member on volume and residence time, and demonstrate that in-channel storage may be appreciable in some reaches.
H41A-0291 0800h
Longitudinal Dispersion In Rivers With Floodplains
A new method is developed for predicting the longitudinal dispersion coefficient in rivers with floodplains. It is found that the longitudinal dispersion in natural rivers with floodplains is significantly enhanced and it displays complicated behavior. When discharge Q is smaller than the bankfull one, the longitudinal dispersion coefficient Kx increases with Q continuously. A discontinuity characterized by a sudden increase in Kx occurs when flow exceeds bankfull level corresponding to the bankfull discharge Qb. The extent of the discontinuity of dispersion coefficients depends on the ratio of the water surface width of the floodplain flows to the top width of the main channel. The dispersion coefficient Kx decreases rapidly and nonlinearly with increase of discharge when Q is greater than Qb and then reach a minimum value Kmin at a discharge Qmin. Kx increases linearly when the discharge increases further. A linear equation of regression for estimating the longitudinal dispersion coefficient is presented for flood flows with Q being greater than Qmin. Moreover, an expression is derived for determining the minimum longitudinal dispersion coefficient Kmin. These findings are of practical significance for the management of floods and river water quality. The predictions of the new method are consistent with field measurements.