Environmental Vadose Zone Hydrology Posters
Presiding: D DiCarlo, USDA/ARS National Sedimentation Laboratory; R M Holt, University of Mississippi; M J Nicholl, University of Nevada, Las Vegas
H13B-01 1330h
The Effect of Calcite Crystallinity on the Dielectric Permittivity of Calcic Soils
Measurement of soil dielectric properties, ε are widely used to estimate water content in soils from remote sensing data and from in situ soil sensors such as time domain reflectometry. Investigations have also explored the use of dielectric measurement to estimate other soil properties such as cation exchange capacity, soil aggregation, and particle size analysis. Understanding the dielectric properties of soils is also required for the improved interpretation of data from methods such as ground penetrating radar and microwave remote sensing. The mineral permittivity also plays an important role in geochemical dissolution and precipitation. Models used to estimate water content from soils often assume a value of 5 for the mineral permittivity ε. However, calcite (CaCO3), a major constituent of some arid and semi-arid zone soils, has a permittivity of 8-9, nearly twice the permittivity of quartz (ε=4.6). However, tabulated permittivity values for minerals generally refer to pure crystalline samples, crystals formed by crystal growth in the absence of impurities and without isomorphic substitutions; this is not the most likely scenario for pedogenic calcite. We studied 4 soils, with micaceous mineralogy, but with two soils having 40% pedogenic calcite. We also measured the permittivity of Iceland Spar calcite (ε=9.1) and a microcrystalline calcite (ε=8.3), and used atomistic modeling to account for differences in permittivity based on the crystal density. We found dielectric permittivities for our soils to be in the range of 5.8 to 6.6, higher calcite contents resulting in increased permittivity. The estimated permittivity of the calcite in the soils was 7.4-7.9, lower than the highly crystalline samples. We estimate, for a soil with a porosity of 0.5 that assuming a permittivity of 5 instead of 6.6 will result in an overestimation of water content of about 1% at saturation. This demonstrates that a large quantity of pedogenic calcite (40%) in soil is unlikely to cause substantial error in the determination of water content using standard calibration equations.
H13B-02 1330h
Proposal of a Methodology for Comparing Electromagnetic Soil Water Content Sensors
Performance differences in the growing number of electromagnetic (EM) sensing systems designed to predict soil water content from dielectric permittivity estimates suggests the need for a standardized characterization methodology. The objectives of this study were to i) develop a methodology for characterizing sensing system performance using well defined and readily available reference dielectrics and ii) suggest a standard for comparison of sensing systems estimating permittivity. Standards are based on fluids of known or measurable frequency-dependent permittivity, which provide a homogeneous system for immersion of sample-scale EM probes. Measurement frequencies for broadband sensing systems can be inferred from correlated network analyzer and sensing system measurements or from manufacturer suggestions. Fluids were selected to provide surrogate soil-related effects (e.g. relaxation) occurring both within and outside of the effective measurement frequency range of common systems. Test conditions included dielectrically relaxing (R) and non-relaxing (NR) as well as electrically conducting (C) and non-conducting (NC) liquids and combinations thereof. No suitable combination of relaxing and conducting (R-C) dielectric fluid was found in this study but remains a goal of future work because it represents the more difficult and often common condition for EM sensing system measurements in soils containing relaxing and conducting elements. Criteria for qualitative assessment of sensing systems include measurement frequency and accuracy, susceptibility to variations in salinity, dielectric relaxation and temperature and spatial variation in probe sampling area. The approach outlined here was applied to seven different EM systems with results being presented in a companion oral presentation by Blonquist et al.
H13B-03 1330h
Spatial and Temporal Variations in Electrical Conductivity in North Mississippi Loamy Soils
The use of electromagnetic induction (EM) to characterize and quantify spatial and temporal variations in soil properties is appealing due to low operational costs, rapid measurements, and device mobility. EM methods are sensitive to soil electrical conductivity, which can vary with soil moisture, clay content, soil salinity, and the presence of electrically conductive minerals. We are evaluating the controls on EM response in loamy soils present at the University of Mississippi (UM) Soil Moisture Observatory (SMO). The 5 acre SMO is located in a former agricultural field at the UM Field Station, a 740 acre tract of land located 11 miles from the UM campus in Oxford, Mississippi. EM responses were surveyed along transects and randomly chosen locations using a Geonics EM38. The apparent electrical conductivity (EC) of the soil was determined in both a vertical and horizontal dipole position, which correspond to deep (~1m) and shallow (~0.5) measurements, respectively. We find that EC is primarily controlled by soil physical and chemical properties under moderately dry conditions. Under wetter conditions, EC shows good correlation with soil moisture content. EC variograms show similar spatial structures at different times. However, EC variability increases under wet conditions. These results imply that pore surface conduction dominates under dry conditions, while pore water conduction becomes more important under wet conditions. Variogram similarity suggests that spatial variations in EC are strongly influenced by spatial variations in soil properties regardless of soil moisture conditions. These relationships may allow the development of a predictive model for soil moisture based on EC measurements in North Mississippi loamy soils.
H13B-04 1330h
An Assessment of Analytical Model Predictions of Pumping-Induced Vadose Zone Response in an Unconfined Aquifer Using Neutron-Derived Soil Moisture Profiles
Analytical models used to analyze hydraulic head data from pumping tests represent drainage from above the water table by means of boundary conditions defined along the water table. While these boundary conditions do not explicitly describe the details of flow in the vadose zone, they do predict the bulk vadose zone response (i.e., cumulative drainage flux and excess storage). This bulk response can be determined directly from soil moisture content profiles measured during a pumping test. A comparison of the predicted and observed bulk vadose zone response was performed using hydraulic head data and soil moisture content profiles obtained during a seven day pumping test at CFB Borden, Ontario. Detailed soil moisture profiles were acquired with a thermal neutron moisture probe in six observation wells at varying radial distances throughout the test. Three different analytical models (one assuming instantaneous drainage and two types of delayed drainage using exponential decay) were used to predict the bulk vadose zone response from the hydraulic head measurements. These predictions significantly overestimated the observed cumulative drainage flux in the vicinity of the pumping well. The exponential decay models predicted a relatively rapid dissipation of the excess storage; the observed excess storage exhibited little, if any, decay throughout the entire pumping test. Our results clearly show that boundary conditions used in these analytical models do not adequately account for the details of flow in the vadose zone.
H13B-05 1330h
Using Airborne Geophysics to Improve the Management of Produced Water from Coal Bed Natural Gas Extraction in the Powder River Basin
The Powder River Basin (PRB) of Wyoming and Montana has seen a boom in drilling for coalbed natural gas (CBNG), the natural gas contained in coal seams. Large quantities of water are coproduced during the extraction process. The water is currently managed by land application (irrigation), returned to shallow groundwater aquifers via infiltration basins, directly discharged to ephemeral or perennial streams, or injected into the deep subsurface via injection wells. At present, there are over 28,000 CBNG wells permitted or drilled in the PRB and it is estimated that another 50,000 to 100,000 new wells will be drilled in the future. Produced water management is a major challenge to the oil and gas industry as well as federal and state regulators. The purpose of this study was to evaluate the use of airborne electromagnetic (AEM) methods for the large-scale mapping of vadose zone properties. The base maps derived from the AEM data show the location of conductive anomalies within the vadose zone. These conductive anomalies have been identified as conditions related to soil properties, geologic features, saturated areas, and seepage zones. In the PRB, the data can be used to identify suitable locations for constructing impoundments in areas that avoid highly conductive soils where infiltrating water may leach salts through the vadose zone and into shallow aquifers. Hydrologic changes within the vadose zone were evaluated by completing an AEM survey in 2003 and 2004 over two coincident spatial areas. The data were analyzed to determine statistical relationships between the data sets, in particular data outliers which may represent areas of significant change between each year. Some outliers plot near areas of CBNG development. Ultimately, it is hoped that the information from these surveys will identify cost effective treatment or disposal options for produced water that address both production and environmental issues.
H13B-06 1330h
Information Content and Complexity of Simulated Soil Water Flux Series
The accuracy-based performance measures may not suffice to discriminate among soil water flow models. Comparing complexity of the model structures does not lead to quantifiable results. The objective of this work was to attempt using information theory parameters to discriminate between simulated time series of soil water fluxes obtained from different models for the same site. The Richards equation-based model HYDRUS-1D and a water budget-type model MWBUS were used to simulate one-year long observations of soil water contents and infiltration fluxes at various depths in a 1.5-meter deep loamy Eutric Regosol in Bekkevoort, Belgium. We used the metric entropy and the mean information gain as information content measures, and the effective complexity measure and the fluctuation complexity as complexity measures. To compute the information content and complexity measures, time series of fluxes were encoded with the binary alphabet; fluxes greater (less) than the median value were encoded with one (zero). Fifty Monte Carlo simulation runs were performed with both models using hydraulic properties measured along a trench. The two models had the similar accuracy of water flux simulations. Precipitation input data demonstrated a moderate complexity and relatively high information content. Model outputs showed distinct differences in their relationships between complexity and information content. Overall, more complex simulated soil flux time series were obtained with the HYDRUS-1D model that was perceived to be conceptually more complex than the WMBUS model. An increase in the complexity of water flux time series occurred in parallel with the decrease in the information content. Using both complexity and information content parameters allowed us to discriminate between the soil water models that gave the same accuracy of soil water flux estimates.
H13B-07 1330h
Fate and Transport of CL-20 and RDX in Unsaturated Laboratory Columns
This research examines the fate and transport of two explosive compounds, Hexanitrohexaazaisowurtzitane (CL-20) and Hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX) in unsaturated laboratory columns. The transport and fate of these compounds were studied under saturated and unsaturated conditions in three natural soils: coarse sand, sandy loam, and a silt loam. Unsaturated column experiments were conducted using an ultra-centrifugation method. Sorption and degradation parameters were determined by moment analysis and hydrodynamic parameters were assessed with a two-region flow model. Differences in these parameters were evaluated as a function of water content. The fate and transport of CL-20 is highly dependent on 1) the soil type and 2) the compound's residence time in the soil and 3) water content of the media. Sorption of CL-20 was rate-limited. CL-20 degradation in saturated columns produced a half-life of as much as 22hr, but in unsaturated columns the degradation rate increased considerably, producing a half life of as little as 2hr. The fate and transport of RDX are also affected by the soil type, but sorption appeared to be instantaneous. Degradation of RDX was negligible. Our results suggest that at very low water content immobile water regions may become (at least in effect) isolated water regions and significantly alter the retardation of the tracer. In the sandy loam, there was as much as a 20-fold over-prediction of the retardation factor in the unsaturated saturated columns when predicted by Kd values derived from saturated columns. In the coarse sand, Kd values derived from saturated columns over-predicted retardation in the unsaturated columns by as much as 30%. In the silt loam, retardation factors were over-predicted by as much as 80%. At very low water contents, predictions of tracer behavior become very difficult because of changes in the flow regime that cannot be directly accounted for.
H13B-08 1330h
Evaluating Groundwater Pesticide Loading Using 1-D Vadose Zone Models
A major pathway for pesticide entry into groundwater is through the vadose (unsaturated) zone. In the absence of reliable monitoring data, computer models are often used to estimate pesticide concentrations in ground water. Models that attempt to provide usable values for predicted groundwater pesticide concentrations need to incorporate several crucial parameters, including information on soil properties, hydrologic conditions, geomorphology, meteorological data, water table depths, crop types, and pesticide usage patterns. Preferential flow (e.g., through macropores) within the vadose zone must also be considered. In this study, three models, PRZM, RZWQM, and LEACHP, were evaluated for the purpose of estimating groundwater pesticide loading through the vadose zone. Methods, such as using deep vadose zone pore water chemical concentrations as a surrogate for shallow groundwater concentrations, and modeling/integrating macropore flow contributions to groundwater, were explored. Values obtained from these models were compared with field-collected groundwater data to determine the suitability of methodology and relative accuracy of each predictive model. Preliminary data indicate that, at least under near-steady-state conditions (that is, in the absence of major rainfall events that would trigger substantial macropore flow), vadose zone pore water concentrations generally reflect shallow groundwater concentrations. This abstract has been subjected to review by EPA's Office of Pesticide Programs and was approved for submission. Approval does not signify the contents reflect the views of EPA, nor does mention of trade names or commercial products constitute endorsement or recommendation for use.
H13B-09 1330h
Landfill Gas Effects on Evapotranspirative Landfill Covers
The performance of an evapotranspirative landfill cover can be adversely affected by transport of landfill gases to the plant root zone. Healthy plant communities are critical to the success and effectiveness of these vegetated landfill covers. Poor vegetative cover can result in reduced transpiration, increased percolation, and increased erosion regardless of the thickness of the cover. Visual inspections of landfill covers indicate that vegetation-free areas are not uncommon at municipal waste landfills. Data from soil profiles beneath these areas suggest that anaerobic conditions in the plant-rooting zone are controlling plant distribution. On the same landfill, aerobic conditions exist at similar depths beneath well-vegetated areas. The movement of methane and carbon dioxide, generated by degradation of organic wastes, into the overlying soil cover displaces oxygen in the root zone. Monitoring data from landfills in semi-arid areas indicate that barometric pumping can result in hours of anaerobic conditions in the root zone. Microbial consumption of oxygen in the root zone reduces the amount of oxygen available for plant root respiration but consumption of oxygen and methane also produce water as a reaction byproduct. This biogenic water production can be on the order of centimeters of water per year which, while increasing water availability, also has a negative feedback on transport of landfill gases through the cover. Accounting for these processes can improve evapotranspirative landfill cover design at other sites.
H13B-10 1330h
How do Sandy Palaeochannels Affect Shallow Groundwater Flow in a Clay-Dominated Vadose Zone under Flood Irrigation?
Palaeochannels, or prior streams, are common landscape features in much of the irrigated areas of the Northern Murray-Darling Basin, Australia. Prior research has indicated that these features are likely to amplify irrigation water losses due to their sandier textures. However, little is known about the pathways and flux of this water after it infiltrates and how changes in soil properties and sedimentary layering govern this movement. This poster utilises electromagnetic induction surveys to supplement soil core information to model a palaeochannel system using the three-dimensional groundwater simulation package PMWIN-MODFLOW. Hydraulic properties were estimated using pedotransfer functions of soil physical properties. Based on the distribution of observed soil properties and the uncertainties related to the hydraulic property estimation, several realisations were created and Monte Carlo-type simulations were run to identify model sensitivities to the empirically-derived input parameters. Irrigation and rainfall top boundary inputs were applied to simulate water movement through the system. Model outputs are analysed using simulated observation wells and particle tracing, showing perched water tables within the sandy palaeochannel following stress periods. Contrasts in hydraulic conductivity, soil layering, and slope of the channel had a greater effect on water movement under irrigation than under natural rainfall conditions.
H13B-11 1330h
A Toolbox of Models for Evaluating Appropriateness of Infiltration Predictions in Coupled Surface and Subsurface Flow Applications
A number of coupled surface and subsurface flow models include Richards' equation (RE) solutions to simulate flow through the vadose zone, but the question of the appropriate application of RE at the watershed scale remains unclear. Published applications of RE in watershed-scale modeling often use coarse vertical resolutions to decrease the computational burden. In contrast, investigations demonstrate that small vertical cell sizes, on the order of 1 cm, are required near the soil surface, to accurately simulate RE infiltration fluxes in watersheds. A compromise is sought to satisfy both the computational burden and accuracy issues within the confines of available hydrological data for this scale of problem. Research indicates that conditions commonly exist in watersheds wherein basic assumptions of RE applicability are violated or where RE is unable to reproduce unstable infiltration behavior. Thus, physical and computational factors motivate the search for alternatives to RE in simulating the quantity and rate of infiltration that occur in coupled surface and subsurface flow systems. Research is underway to identify a toolbox of infiltration models to evaluate appropriateness for the scale and scope of the physical processes being simulated. The selection of toolbox models must also be sensitive to the fidelity of available data and the site hydrologic, hydrogeologic, and topographic conditions. Results to date in the identification of the range of application of toolbox models will be presented.
H13B-12 1330h
Natural Barriers of the Geosphere at Yucca Mountain, Nevada
Geological repositories designed to isolate high-level radioactive waste need natural and engineered barriers that prevent or slow the release of radioactive elements into the accessible environment to acceptable regulatory limits. Under the U.S. Nuclear Regulatory Commission's (NRC's) regulations, a barrier is any material, structure, or feature that prevents or substantially reduces the rate of movement of water or radionuclides from the repository to the accessible environment. In 1982, Congress passed the Nuclear Waste Policy Act which directed the NRC to include multiple barriers in regulating geologic disposal of high-level radioactive waste. Accordingly, as provided in 10 CFR Part 63, the NRC's regulations for Yucca Mountain require a repository to include multiple barriers to ensure the system is robust and not wholly dependent on any single barrier. Any potential license application to construct a repository at Yucca Mountain must identify the multiple barriers (both natural and engineered), describe the capabilities of each barrier, and provide the technical bases for the capabilities of the barriers. The NRC believes that understanding the capability of the repository's component barriers improves understanding of the overall system. The objective of this paper is to discuss potential natural barriers of the geosphere at Yucca Mountain and describe the NRC regulatory requirements for such barriers. To better understand the natural barriers of the geosphere, it helps to divide the barriers into groups of features and their associated processes. Natural barriers, i.e., barriers not constructed by man, ideally include processes that delay the transport of radionuclides from reaching the accessible environment or limit the amount of water that can seep from a ground surface to the depth of an underground repository. Natural barriers at Yucca Mountain may include: topographic influences on precipitation runoff; soil and plants influences on evaporation and transpiration; effect of surface bedrock characteristics on infiltration; influences of unsaturated zone rocks above the repository on quantity and characteristics of downward flowing water; and the effects of the repository tunnel wall geometry and rock characteristics on seepage into the tunnels. Unsaturated zone rocks below the repository may influence water and radionuclide migration into either fractures or the rock matrix, where processes such as matrix diffusion and sorption can retard radionuclide movement. Properties of different saturated zone rock units may slow the radionuclide flow rate while structural features within the saturated zone rocks (faults, heterogeneities) control water flow rate and direction. The saturated zone alluvium may reduce the water velocity while radionuclides sorption onto the alluvium can further delay radionuclides from reaching the accessible environment.
H13B-13 1330h
Prediction of Pore Geometry Based on Air Permeability Measurements Using a Genetic Algorithm
Pore size distributions can be used to determine fluid retention and permeability relationships of porous media. However, the available test methods are time-consuming and expensive. This study consists of experimental and numerical work. In the experimental part we determine air permeability of a sandstone rock core as a function of water content. In the numerical part we present an extension of the capillary model, which was modified include capillaries composed of sections with different diameters. An optimization scheme that uses a genetic algorithm to predict the best possible pore size distribution from the air permeability measurements was developed.
H13B-14 1330h
Impact of Liquid Configuration and Flow Regimes on Macroscopic Transport Properties in Unsaturated Porous Media: A Lattice Boltzmann Study
The effect of flow regime and fluid morphology (distribution and spatial arrangement) on macroscopic transport properties (effective diffusion coefficient, hydraulic conductivity) was analyzed using a multi-component lattice Boltzmann model. Various flow regimes for two-phase flow in porous media have been defined: stable displacement, capillary fingering, and viscous fingering. The dominance of one regime over another in a porous medium of interest is controlled by the relative magnitudes of gravity, viscous, and capillary forces, which can be quantified with three parameters: Bond number Bo, capillary number Ca, and their difference, Bo-Ca. It has been shown that macroscopic transport properties in porous media are highly dependent on fluid configuration. Since the three flow regimes exhibit very different fluid morphologies, it seems likely that flow regime will have a significant effect on diffusion and hydraulic conductivity. Accordingly, a series of imbibition simulations in each of the flow regimes was carried out by varying Ca and Bo. Results were compared to experimentally-derived images from the literature. Unsaturated hydraulic conductivity curves and effective diffusion coefficient curves were generated as a function of water saturation for each flow regime and compared to theoretically-determined curves. Significant differences are seen in the curves for the fingering regimes compared to the stable displacement and theoretical curves.
H13B-15 1330h
Influence of the Air-Entry Pressure Ratio in Bimodal, Heterogeneous Unsaturated Zones
The unsaturated zone at the Idaho National Laboratory (INL) displays bimodal heterogeneity with high air-entry pressure (AEP) sedimentary interbeds intercalated with variably fractured, low AEP basalt flows. While most past work has focused on understanding flow and transport in the basalt flows, the influence of the sedimentary interbeds is beginning to become recognized. Recent studies suggest that sedimentary interbeds may control large-scale vadose zone flow and transport behavior at the INL. We investigate the influence of high AEP interbeds (e.g., sedimentary) within a low AEP medium (e.g., basalt) using a macroscopic invasion percolation (MIP) model that includes the effects of capillary, gravity, and viscous forces. The MIP model is parameterized with spanning pressures derived from pressure-saturation curves for the low and high AEP media and hydraulic conductivities derived from a capillaric model. The MIP model is used in a series of Monte Carlo simulations to define transport pathways that interconnect a source at the land surface to a water table at depth. We find that the air-entry ratio (high AEP/low AEP) strongly influences transport pathways. At small air-entry ratios, transport pathways are short and largely vertical. As the air-entry ratio increases, transport pathways become increasingly complex reflecting the spatial distribution of interbeds. Above a critical air-entry ratio, transport pathways stabilize and become insensitive to the distribution of properties within the low AEP medium (e.g., basalt). Our results suggest that detailed characterization of INL basalts may not be required to define transport pathways through the INL unsaturated zone.
H13B-16 1330h
Network Modeling of Saturation Overshoot on Infiltration
Most, if not all, models of water flow in the vadose zone consist of a continuum approach, where the discreteness of the porous media is replaced by smoothly varying parameters. This continuum approach cannot model certain infiltrations where the saturation at the wetting front is greater than the saturation behind the front. It has been argued that this occurs because the infiltrating water fronts are sharp at the pore-scale. Here we attempt to model water flow at a sharp front by using a physically-based network model that includes viscous effects. The network model includes pore and throat elements of different shapes and sizes, and a connection topology based on geologic media. We show that by adding viscous effects that the saturation overshoot seen in infiltrations can be modeled successfully, suggesting that pore-filling mechanisms control saturation overshoot.
H13B-17 1330h
Temporal Variations in Soil Water Wetness Enhance Microbial Diversity
Surface soils host unparalleled number of microbial species even in small volumes, serving as the richest storage of microbial genus in the biosphere. Spatial and temporal heterogeneity and multitude of niches and microhabitats in the vadose zone are believed to play key roles in fostering such diversity. In this study, a simple network model is constructed to simulate heterogeneous water distribution on a rough surface providing diffusional pathways supporting microbial activity. Because liquid may redistribute according to temporal variations in water potential (soil wetness), the resulting diffusion pathways and connectivity also vary with time. Local capillary properties on the surface and water contents may response differently to the overall water potential variations. Simulations are carried out using the network model as well as random walkers representing microbial activity, and reaction-diffusion model for nutrient transport and consumption. Results show that long term microbial growth is strongly determined by local conditions (location, microtopography) rather than its physiology. This is critical for fostering coexistence of different species and enhancing diversity.
H13B-18 1330h
Periodic Response of a Sweetgum Tree to Transpiration and Environmental Interferences
During the drought of 2001, researchers at Clark Atlanta University observed large diurnal fluctuations in the circumference of old deciduous trees at a suburban forested site. Additional observations were made in 2002 and 2003. A more detailed study of this phenomenon was planned in 2004 for a single tree species. Continuous measurements of the periodic change in sweetgum tree circumference due to transpiration and environmental interferences were made during a 5-week period in the summer of 2004. The study investigated links between environmental conditions and the biological responses, both on diurnal and longer-term scales, by observing changes of the tree trunk circumference of Sweetgum (Liquidambar styraciflua), a species native to the mid-Atlantic region. Monitored environmental parameters include insolation, volumetric soil moisture and temperature profiles, atmospheric temperature, relative humidity and related meteorological variables. Six Sweetgum trees were studied for evidence of periodic patterns in their respective trunk-circumferences during significant precipitation events. Three of the trees were located in the riparian zone of a small stream, while the remaining three were situated above the riparian zone. The soil at the site was a red clay loam typical of the Georgia Piedmont region, an area extending from the foothills of the Appalachian Mountains to the coastal plain. During high precipitation events, the soil under the lowland canopy would persist as a clay slurry mud retaining the moisture days after rain had fallen. For this reason, care was taken to select trees with clear canopy exposure. Harmonic analysis and nonlinear curve fitting of the circumference changes for each tree during precipitation events of half an inch or greater per hour showed distortions in the expected diurnal response to evapotranspirative pumping.