Investigating Changes in Flow and Transport Properties due to Bio-clogging of Porous Media from Complex Conductivity Measurements
Complex conductivity measurements (0.1-1000 Hz) were obtained in flow through sand columns inoculated with Pseudomonas aeruginosa to simulate the effect of bio-clogging on the flow and transport properties of sands. Pressure transducers installed along the side of the columns were used to monitor the changes in hydraulic conductivity. Evidence of bio-clogging was obtained from temporal biomass growth, scanning electron microscope (SEM) observations, and temporal changes in flow rate. After week one of the experiment, the inoculated sand columns showed an increase in the imaginary conductivity component concurrent with a reduction in hydraulic conductivity, a decrease in flow rate and an increase in microbial cell numbers. SEM images showed microbial cells attached to sand grains and polysaccharides joining two or more sand grains together. Analysis of breakthrough curves (BTCs) from dye tracer test conducted at the beginning and the end of the experiment showed a reduction of the porosity and the dispersion coefficient from the initial values by 16% and 50%, respectively. Empirical equations involving formation factor and imaginary conductivity component was used to calculate the hydraulic conductivity. Good agreement was obtained between the calculated hydraulic conductivity from the complex conductivity measurements and the hydraulic conductivity measured along the side of the sand columns. The increase in imaginary conductivity component can be explained by constriction of pores and narrowing of pore throats due to microbial growth and biofilm formation in the sand columns. The results of this study highlights the potential of complex conductivity measurements to validate bioclogging models used to assess the effect of biomass growth on the flow and transport properties of porous media.
Mapping Subsurface Remediation With Ground Penetrating Radar
Successful remediation of sites contaminated with dense non-aqueous phase liquids (DNAPLs) would be substantially assisted by characterisation of the evolving volume and extent of the DNAPL source zone. Ground penetrating radar (GPR) is a geophysical tool that has the ability to continuously and non-invasively sample the subsurface distribution of DNAPLs over time; however, its effectiveness for real applications remains elusive due to challenges in successful interpretation of data from contaminated sites - in either a qualitative or quantitative way. The objective of this study is to evaluate the potential of GPR to map realistic, evolving DNAPL source zones within complex subsurface environments, during remedial efforts. Power et al. 2008 (Eos Trans. AGU, 89(53), Fall Meeting Suppl., H51G-0923) reported on the development of a novel numerical simulator that integrated a multiphase flow model and GPR simulator: DNAPL3D-MT generates realistic DNAPL release scenarios, while GPRMAX simulates GPR responses. In this study, this simulator is employed to explore the ability of GPR to track DNAPL source zone evolution during remediation at the site scale. Two- dimensional, surface releases of chlorinated solvent DNAPL into heterogeneous silty sand aquifers have been conducted, including DNAPL migration, redistribution and remediation, and its continuous mapping by time- lapsed surface GPR scans. Qualitative and quantitative analysis of the results reveal that, in favourable environments, DNAPL volume reduction may be readily monitored as a function of surface location. In addition, simulations reveal that DNAPL migration into previously uncontaminated regions (e.g., as a result of DNAPL remobilization induced by pumping or surfactant flushing) may be observable. Sensitivity simulations explore the influence of site and operational parameters that may limit a GPR exploration of DNAPL sites.
InSAR deformation time series for an agricultural area in the San Luis Valley
The San Luis Valley (SLV) is an 8000 km2 region in southern Colorado that is home to a thriving agricultural economy. This valley is currently in a period of extreme drought, with county and state regulators struggling to develop appropriate management policies for both the surface water and the ground water. In 1998 the state of Colorado commissioned the Rio Grande Decision Support System to refine the hydogeologic characterization of the system, including the development of a MODFLOW finite difference model of groundwater flow. The main challenge in the SLV is acquiring sufficient data to characterize the spatially heterogeneous, time-varying behavior of the groundwater system. Here we apply the small baseline subset analysis (SBAS) interferometric radar (InSAR) technique to provide such data. InSAR techniques yield the deformation of Earth's surface at fine spatial resolution occurring between two satellite overflights, and SBAS permits solution for a time series of deformation maps. The measured deformation can be related to changes in the water table in underlying confined aquifers. The ability to map these changes, over time, in the SLV will provide critical information about the groundwater system. Historically, InSAR measurements have been difficult to make in agricultural areas. The change in cm-scale crop structure with time leads to signal decorrelation and the loss of useful information about surface deformation. The recently-developed SBAS method allows stable deformation estimates at certain ground points in an otherwise decorrelated time series of data. We applied this approach to data collected by the European Space Agency's ERS-1 and ERS-2 satellites over the western SLV from track 98 frame 2853 for the years 1992-2001. We used the Generic SAR (GSAR) SBAS software developed by Norut to produce time series deformation measurements for many positions across the entire SLV. We find that the 2000 km2 area captured in track 98 frame 2853 shows very high levels of correlation in areas between the center pivot irrigation circles, where the lack of water results in little surface vegetation. We extracted a time series displaying the change in deformation over the time period of 1992 to 2001, with a sampling interval of approximately 3 months. The ability to obtain such high quality temporal data across the entire SLV suggests that improved groundwater flow models describing finer- scale heterogeneities than are presently represented are possible with the integration of InSAR data. Specifically, the objective is to develop a quantitative relationship between ground surface deformations measured by InSAR and confined aquifer heads.
Land use Influence on Vadose Water and Salt Fluxes: A Geoelectrical Perspective From Central Argentina
Native dry forests continue to be replaced by rain-fed annual crops in the semi arid plains of central Argentina. The resulting changes in vegetation cover alter the partitioning of precipitation affecting recharge rates and solute transport in the vadose zone. Like similar dry forests from other continents, these deep rooted native ecosystems have virtually no drainage below the rooting zone. Their replacement by annual crops can lead to the onset of deep drainage, solute displacements, and rising water tables that could affect the long-term viability of agriculture in the region. To understand the ecohydrological implications of this vegetation change we evaluate deep subsurface water and solute distributions using soil coring and electrical resistivity imaging along a chronosequence of agricultural plots that where converted from forests 1 to 90 years ago. Our findings suggest a progressive increase of soil moisture and leaching of salts in the vadose zone following the onset of cultivation.
Mapping and Modeling Buried Tunnel Valleys
Throughout the last 20 years aquifers found in buried tunnel valleys have played an important role for the Danish groundwater supply and will continue to do so in the future. The spatial distribution of tunnel valleys along with the identification of the lithology within and around the valleys is often complex and strongly depends on data achieved from geophysical methods. Data density, resolution capacity and penetration capability of the geophysical methods are particularly important when qualified hydrogeological models of the Danish Quaternary and Tertiary succession have to be constructed. In two case studies complex systems of buried tunnel valleys have been mapped and modeled. Apart from existing borehole data, these have been based on data from various types of geophysical methods e.g. Schlumberger soundings, Transient ElectroMagnetic (TEM) soundings, offshore and onshore reflection seismics. The following topics will be evaluated based on the case studies. 1) Evaluation of mapping methods and the benefits and constraints of these in the construction of hydrogeological models. 2) Combined interpretation of the different geophysical methods in order to achieve a better understanding of the geology in the uppermost 200-300 m. 3) Evaluation of the importance of data density, resolution capacity and penetration depth. 4) 3D modeling of tunnel valleys based on geophysical data with focus on infill distribution and hydraulic parameters.
Application of Borehole Geophysical Methods for Assessing Agro-Chemical Flow Paths in Fractured Bedrock Underlying the Black Brook Watershed, Northwestern New Brunswick
The upper Saint John River valley represents an economically important agricultural region that suffers from high nitrate levels in the groundwater as a result of fertilizer use. This study focuses on the fractured bedrock aquifer beneath the Black Brook Watershed, near Saint-Andre (Grand Falls), New Brunswick, where prediction of nitrate migration is limited by a lack of knowledge of the bedrock fracture characteristics. Bedrock consists of a fine-grained, siliciclastic unit of the Grog Brook Group gradationally overlain by a carbonate unit assigned to the Matapédia Group. Groundwater flow through the fractured bedrock is expected to be primarily influenced by the distribution and orientation of fractures in these rock units. This study demonstrates the effectiveness of the select suite of borehole-geophysical tools used to identify and describe the fractured bedrock characteristics, and assists in understanding the migration pathways of agrochemical leachate from farm fields. Fracture datasets were acquired from five new vertical boreholes that ranged from 50 to 140 metres in depth, and from three outcrop locations along the new Trans-Canada Highway, approximately two kilometres away. The borehole-geophysical methods used included natural gamma ray (GR), single point resistance (SPR), spontaneous potential (SP), slim-hole optical borehole televiewer (OBI) and acoustic borehole televiewer (ABI). The ABI and OBI tools delivered high-resolution oriented images of the borehole walls, and enabled visualization of fractures in situ, and provided accurate information on the location, orientation, and aperture. The GR, SPR and SP logs identified changes in lithology, bed thickness and conductive fracture zones. Detailed inspection of the borehole televiewer images identified 390 fractures. Equal-area stereographic and rose diagrams of fracture planes have been used to identify three discrete fracture sets: 1) steeply dipping fractures that strike 068o/248o, with fracture subsets dipping roughly 70o to 80o towards the N-NW and S-SE; 2) steeply dipping fractures that strike towards 156o/336o, with fracture subsets dipping roughly 70o to 80o towards the NE and SW; and 3) primary set of moderately dipping fractures that strike 074o/254o and dip roughly 30o to 40o towards the SE. The strike of the steeply dipping fracture sets are oriented roughly perpendicular to each other, reflecting two distinct fracture generation events. The low-angle fractures are most common and correspond to openings along bedding planes that dip roughly 38o towards 164o. This is a result of penetrating only one limb of a fold; presumably a similar set of bedding-plane openings occur along the adjacent limb of the fold, with resultant fracture dips towards the northwest. Fractures exposed in outcrops along the Trans-Canada Highway exhibit a similar orientation distribution to that observed in the boreholes. However, as expected, these exposures show a greater proportion of fractures with dips between 80o and 90o, compared to the vertical boreholes. A Terzaghi fracture probability correction was applied to the boreholes in order to account for this bias. The combined fracture datasets provide valuable information towards understanding groundwater flow and migration pathways of fertilizer leachate into the bedrock aquifer, and will lead to the development of more complex hydrogeological models.
The Shale Hills Critical Zone Observatory for Embedded Sensing and Hydrologic Simulation
The future of environmental observing systems will utilize embedded sensor networks with continuous real- time measurement of hydrologic, atmospheric, biogeochemical, and ecological variables across diverse terrestrial environments. Embedded environmental sensors, benefitting from advances in information sciences, networking technology, materials science, computing capacity, and data synthesis methods, are undergoing revolutionary change. It is now possible to field spatially-distributed, multi-node sensor networks that provide density and spatial coverage previously accessible only via numerical simulation. At the same time, computational tools are advancing rapidly to the point where it is now possible to simulate the physical processes controlling individual parcels of water and solutes through the complete terrestrial water cycle. Our goal for the Penn State Critical Zone Observatory is to apply environmental sensor arrays, integrated hydrologic models, and state-of-the-art visualization deployed and coordinated at a testbed within the Penn State Experimental Forest. The NSF-funded CZO is designed to observe the detailed space and time complexities of the water and energy cycle for the Shale Hills watershed including all physical states and fluxes (groundwater, soil moisture, temperature, streamflow, latent heat, snowmelt, chemistry, isotopes etc.). In this paper we present recent results for integrated flow modeling of the Shale Hills Watershed focusing on the implementation of a macropore flow data and real-time sensor network design. The issues of optimal numerical grids and parameter estimation is the focus of this research.