H41F-0343 0800h
Quantifying hyporheic exchange in streams and rivers - Experimental limitations and practical relevance
Mixing of water and solute elements between streams and groundwater is controlled by various mechanisms such as in-stream mixing, flooding, and surface water-groundwater interaction. Major problems are to theoretically describe the hyporheic exchange mechanism and to measure it from tracer experiments. The surface water-groundwater exchange can be caused by pumping due to bed-forms, stream curvature, stream-slope variations and landscape topography on various scales. This implies severe difficulties to discriminate between effects of various exchange mechanisms on breakthrough curves from tracer experiments and to generalise results to conditions other than those under which measurements were performed. This work describes how the tail of the breakthrough curves from tracer tests can be used to interpret certain types of hyporheic exchange and the limitations with respect to generalisation. Experimental data from both small streams are show that the pumping mechanism may often dominate the exchange. A field study of the Phosphorus transport due to sewage effluents from a large number of individual households into a stream network indicates the tremendous impact of the hyporheic pumping exchange on the overall watershed response. Especially for sorbing solutes, like phosphorus, the filtration in the subsurface effectively retard the transport. The difficulty in exact measurements of all mechanisms leading to surface water-groundwater exchange and filtration implies a great source of uncertainty in watershed modelling. Potential model biases due to experimental errors are analysed in the study.
H41F-0344 0800h
Diel changes in stable carbon isotope ratios and trace element concentrations in the Clark Fork River, MT.
A diel (24-hr) water sampling was conducted on the Clark Fork River near Deer Lodge, Montana from 31-July to 1-August of 2003. The Clark Fork River is located in southwestern Montana and its upper reaches have been heavily affected by historic mining and smelting activities in the Butte and Anaconda areas. River floodplains and sediment beds contain significant quantities of metals (iron, aluminum, copper, zinc, lead, cadmium.) and arsenic. Two sites about 1.2 km apart, were sampled with a mean transit time for the water of 2.5 hours between the sites. The river in the study reach is characterized by oxic conditions, moderate alkalinity, moderate biological productivity and a pH range of about 8.0 to 8.5 during the summer low water months. During the mid-summer months, water in the Clark Fork River and its tributary streams is diverted for irrigation. Some of this water returns to the main stem in a chemically modified form as surface or groundwater return flow. This greatly complicates the hydrogeology and nutrient balance of the watershed. The two samplings sites used in this study make it possible to analyze the changes in chemical and physical properties of the water as it travels between the sites. Results reported here show that there is a large degree of temporal and spatial variability across the study area. In situ instruments were deployed and hourly water samples were collected for analysis from both sites. Diel concentration cycles are documented for dissolved forms of manganese, zinc and nitrate. Particulate forms of zinc, aluminum, copper, iron and manganese are shown to undergo twenty-four hour concentration changes. Total suspended solids (TSS) also show a diel change with the mass of TSS increasing at night. Dissolved arsenic is shown to undergo a diurnal concentration cycle at both sampling sites that was out of phase by the average 2.5 hour transit time between the two sites. This arsenic fluctuation may be a pulse of As that is being advected down river from a site above the sampling area. One component of this project was to demonstrate the presence of a diel stable carbon isotope ($\delta$$^{13}$C) cycle mediated by the use and production of dissolved carbon dioxide. Aquatic plants use carbon dioxide during photosynthesis and there is a carbon isotope fractionation associated with the removal of CO$_{2}$ from the water column. This work demonstrates the presence of a diel cycle in the stable carbon isotope ratio $\delta$$^{13}$C at both sites. The magnitude of the carbon isotope cycle is significantly different at the two sites and this is correlated with the rates of photosynthesis and respiration. The difference in productivity at the two sites is associated with the difference in nutrient levels and the nitrate to phosphate ratio.
H41F-0345 0800h
Modelling Thermal Structure and Vertical Eddy Diffusivity in Ponds With the Princeton Ocean Model
Observations of the vertical thermal and chemical structure in a shallow pond reveal that the rate of vertical transport varies significantly on the centimeter space scale and the subhourly timescale. The rates of vertical transport must be quantified on the appropriate time and space scale to understand the coupled physical and chemical dynamics of shallow (maximum depth of a few meters) aquatic systems. Empirical formulas do not exist for vertical eddy diffusivity in shallow systems where shear and density gradients are dynamic. Therefore, the Princeton Ocean Model was adapted for use in shallow systems in order to determine the vertical eddy diffusivity for use in future biogeochemical reaction and transport modelling. The heat flux terms were determined from local meteorological conditions and published bulk formulations. An open boundary condition for heat was added to the bottom to account for heat exchange with the sediments. The incident short wave radiation was divided into seven bandwidths with different attenuation coefficients. Model temperature generally agreed with data from Mirror Lake (Storrs, CT; max depth = 1.5 meters) over a range of weather conditions and measured visible light attenuation coefficients. The modeled vertical diffusivity for heat varied over five orders of magnitude. The transition from molecular diffusivity to complete turbulent mixing occurs on the hourly time scale and the cm space scale. During the modeled time periods, wind speeds are generally negligible at night when the water column mixed depth increases due to convective overturn. The magnitude of the thermal stratification is a strong function of the visible light attenuation coefficient. The Princeton Ocean Model can be applied to examine diel stratification/destratification dynamics and the subsequent effects on the biogeochemistry of shallow aquatic systems.
H41F-0346 0800h
Techniques for Interpretation of Shallow Lake Metabolism
In shallow lakes, nutrients come from two sources; aerobic respiration in the water column and anaerobic respiration processes (by HNO$_{3}$ and Fe$_{2}$O$_{3}$) typically in the sediments. Because of the water column and sediment reactions, the net observable change in oxygen and carbon dioxide in the shallow lake may or may not follow the Redfield Ratio of -138O$_{2}$:106C. We define the Process Ratio (PR) as the rate of change in dissolved oxygen concentration divided by the rate of change in total dissolved inorganic carbon (TDIC) concentration. The PR can be used to evaluate net biogeochemical pathways contributing to lake metabolism. Mirror Lake, Storrs, Connecticut, has a maximum depth of 1.5 meters. During summer, Mirror Lake is thermally stratified during the day and may or may not destratify at night. Oxygen and TDIC were measured at six depths within the water column. Significant departures from the Redfield Ratio are observed in the bottom layers of Mirror Lake when oxygen is at or near depletion. Surface PR values for the same time period are close to the Redfield value of -1.3 (within error). These observed departures from the Redfield Ratio are consistent with shallow lake metabolism. Idealized PR values were modeled using stoichiometric equations for the oxidation reactions of organic matter by HNO$_{3}$ and Fe$_{2}$O$_{3}$ as end members. These are used to identify both the nitrate and the iron reduction nutrient pathways. The reduction of iron also implies the loss of a significant adsorption surface that may result in remobilization of adsorbed metals. PR may thus serve as a diagnostic tool for highly eutrophic lakes.
H41F-0347 0800h
Estimating Effective Vertical Diffusivity in Shallow Ponds by a Constrained Flux-Gradient Method
Shallow ponds have been used to mitigate the deleterious effects of storm water run-off by acting as detention/retention basins that sequester run-off associated pollutants in sediments. Studies show that the retention efficiency of these systems can decrease over time as a result of the internal loading of nutrients/contaminants from the sediments back to the water column where they are available for export downstream. Quantifying the vertical transport of gases (down) and sediment derived materials (up) is vital to the modeling and understanding of the processes that contribute to the magnitude of internal loading. A critical parameter is the effective vertical diffusion coefficient: K$_{z}$=D$_{molecular}$ +D$_{eddy}$ (cm$^{2}$ sec$^{-1}$). The flux gradient method for estimating effective vertical thermal diffusivity has been applied with success in large lakes which undergo stratification cycles on seasonal or longer time scales. We offer a constrained version of the flux-gradient method that has been adapted for use in a shallow pond with a daily stratification cycle. The method employs heat as a tracer and assumes that transport in the face of a stable gradient is diffusive. By shrinking the spatial and temporal resolution of measurement to scales appropriate to the system of interest and carefully accounting for internal source and sink terms of heat (e.g solar radiation and sediment heat fluxes) we are able to calculate K$_{z}$ as a function of time and depth during periods of stable stratification, i.e when the pond is not vertically well-mixed. Results show the magnitude of K$_{z}$ varies from ca. 10$^{-3}$ to 10$^{-1}$ (cm$^{2}$ sec$^{-1}$) under stratified conditions depending primarily on the strength of stratification.
H41F-0348 0800h
Temporal and Spatial Variation of Chemical Water Quality in a Contour Canal.
Chemical water quality is a highly variable aspect of any water body. Historically numerous researchers have investigated the chemical variability of rivers, streams and wetlands, artificial water bodies such as canals have been largely neglected. Canals are typically hydraulically characterised by low flows and a lack of mixing processes. This can potentially lead to significant spatial variability in water chemistry, and as a result many canals in the UK regularly fail water quality targets at specific locations. Recent changes to UK legislation, following the European Water Framework Directive (2000/60/EC), have resulted in canals being subject to achieving `good ecological status'. In the case of canals, what constitutes `good ecological status' is largely unknown and little expertise is available since historically canal management has not been driven by chemical and ecological quality targets. Consequently, there is an urgent need for new research to determine the main factors influencing canal water quality and their ecological status. This research presents results from a study based on a UK contour canal, the Union Canal in central Scotland. The Union Canal typically demonstrates spatially and temporally variable levels of dissolved oxygen (DO) and orthophosphate (PO4-P): simultaneously, seasonal and diel fluctuations of DO and PO4-P are pronounced at a small number of locations. During 1995, minimum levels of DO along the canal length ranged from 9mgl-1 in Edinburgh to as low as 2mgl-1 approximately 20kms away, this then rose again to 8mgl-1 after a further distance of 2km. These acutely low levels of DO are coupled with events of excessive PO4-P up to 0.235mgl-1:10 times greater than those normally found in rivers, causing localised eutrophication and extensive fish kills. To determine the cause of the `hot spots' of poor water quality found on the Union Canal, simultaneous investigations of the hydraulic regime, spatial and temporal water quality variation and the canal's biological status were carried out. Velocity metering in the canal identified extremely low flow rates ~0.15m3s-1. A tracer testing procedure for the canal's low flow conditions was designed and implemented which identified a lack of rapid dispersion processes with D~0.133m3s-1. Water quality sampling consisted of a year-long programme of high frequency temporal and spatial sampling along the canal length. Observations demonstrate significant variability, with widely differing measurements of DO as little as 5m apart. In addition, spot samples of water quality taken from individual incoming field drains showed PO4-P concentrations up to 2mgl-1, with a predominance of nutrient bound clay and silt sediments that ultimately settle on the canal bed. Due to low dispersion rates, residence times for pollutants are long and field drains, in combination with navigational activity, may well be one of the primary causes of raised nutrient levels at some locations. This research has shown that canal water quality is highly spatially and temporally variable; far in excess of the variability normally found in river systems. This is mainly determined by a lack of hydraulic mixing and the presence of small quantities of incoming runoff water of very low quality. Whilst low in volume, incoming sediment from the drains appears to strongly influence the nearby canal water quality. These results have important consequences both for future monitoring strategies of canals and management of their gradual ecological improvement.
H41F-0349 0800h
Phosphorus Treatment in Spatially Distributed Isolated Wetland Systems of the Okeechobee Basin, FL
A first-order degradation, one-dimensional steady state water flow model was developed to simulate phosphorus (P) treatment efficiency of spatially distributed isolated wetlands (IWs) typical of those in the Lake Okeechobee Basin, FL. Field measurements of groundwater gradients, coupled with surface water hydro- and P chemographs from ditched IWs, added insight to the dynamics and mechanisms involved in P treatment and transport in the landscape as well as the soundness of some model assumptions. This region is of interest because it is currently the largest non-point exporter of P to Lake Okeechobee, which supports wildlife, recreation, and contributes to deep-aquifer recharge. Isolated wetlands may exhibit sporadic hydraulic connectivity with nearby surface water bodies; however, groundwater connectivity may often play a leading role in water and contaminant transport pathways. Three parameters were evaluated in the watershed-scale treatment of P: 1) spatial distribution of wetlands in the landscape, 2) areal coverage of water in the landscape, and 3) wetland P degradation coefficient. A system of two wetlands was evaluated for five different spatial arrangements. The areal coverage of wetlands in the landscape was varied from 0 to 16 percent of the entire landscape. The P decay coefficient for the wetland was varied between 0.01 and 10 [yr-1]. Maximum treatment of groundwater P at the watershed outflow boundary ranged between 20 and 28 percent, based on the independent treatment contribution of each wetland across the water flow field. Measured groundwater gradients revealed that, during rainfall events, groundwater entered the IWs from nearly all directions and comprised a significant portion of the surface water outflow from the IWs. Phosphorus chemographs showed a relatively high initial mass flux out of the wetland at the beginning of the wet season. Simulations show that under steady state non-ditched water flow conditions in the Lake Okeechobee Basin, spatially distributed IWs may promote P reduction from agriculturally impacted waters. Field data demonstrated that ditched IWs are less likely to reduce P loading because event-driven surface water outflows decrease water retention in the wetlands.