Boundary Conditions at the Soil-Atmosphere Interface from Meteorological Data Using HYDRUS-1D
Evaluating the water contents and soil temperatures in the vadose zone is important for water management of agricultural fields. For accurate numerical analyses of the simultaneous movement of water and heat, the boundary condition at the soil-atmosphere interface is critical. The potential evapotranspiration can be calculated using the FAO recommended Penman-Monteith combination equation or Hargreaves equation. While Penman-Monteith equation requires information about net radiation, wind speed, minimum and maximum temperatures and relative humidity, Hargreaves equation requires only minimum and maximum temperatures. Since actual evaporation from soils and transpiration from crops are affected by water contents in the root zone, their evaluation requires simultaneous calculations of water flow and root water uptake in soils. It is well known that water and/or vapor flow and heat transport processes are closely coupled and strongly affect each other, their simultaneous interactions have to be considered. The surface energy balance equation is another surface boundary condition to estimate the soil heat flux from the net radiation, the sensible heat flux, and the latent heat flux, and coupled movement of liquid water, water vapor, and heat in the subsurface can be calculated. We implemented in the HYDRUS-1D software package the Penman- Monteith equation, the Hargreaves equation, and the surface energy balance equation as the surface boundary condition. Various combinations of meteorological data and crop data can be entered in the graphical user interface. These three types of boundary condition are evaluated using measured data collected at field sites.
A Side-by-side Comparison of Eddy Covariance and Bowen Ratio Evapotranspiration From a Northern Great Basin Ecosystem
Development of new ground water resources in rural lands surrounding urban areas of the arid western United States has been identified as a key to maintaining urban growth. The extent and rate at which ground water can be sustainably extracted, while minimizing environmental impacts, depends to a large degree on how much of the existing resource escapes back to the atmosphere via the process of evapotranspiration (ET). ET is the sum of water that evaporates (E) from soil surfaces and is transpired (T) from plant leaves. This study was conducted in a northeast/southwest trending closed basin, bounded by mountains on all sides, with no surface water outflow. The playa which is located in the center of the basin encompasses 59,600 hectares, while phreatophytic plants bordering the playa encompass 20,000 hectares. ET data was collected from a sagebrush/greasewood plant community from October 2004 to August 2008 by the Bowen Ratio technique. Eddy covariance ET monitoring equipment was installed in April of 2007 and a side-by-side comparison was conducted from April to October, 2007. At the end of the study period a 25% difference in ET was calculated between the eddy covariance and Bowen ratio techniques. The major factors for the differences in ET from the two different methods will be presented.
Variation of Surficial Soil Hydraulic Properties Across Land Uses in the Southern Blue Ridge Mountains
Soil hydraulic properties have been shown to affect watershed hydrology by influencing pathways and transmission rates of precipitation to stream networks, and human land use has been shown to influence these soil properties. Particle size distribution, saturated hydraulic conductivity, bulk density, and water holding capacity were measured at 90 points (30 points in each land use category of forest, lawn, and pasture) in a 900km2 area in the North Carolina Blue Ridge. Forest soils demonstrated markedly lower bulk densities and higher infiltration rates, and water holding capacities than lawn and pasture soils, which did not differ. Mean values for each property were (forest = F, lawn = L, pasture = P): saturated hydraulic conductivity (cm/h) – F=7.7, L=1.1, P=1.2; bulk density (g/cm3) – F=0.8, L=1.2, P=1.2; water holding capacity (%) – F=72, L=42, P=39. Particle size distributions did not significantly differ among land use classes or parent materials, and the differences between the hydraulic properties of forest vs. nonforest soils were attributed to compaction associated with land management practices. These results suggests that widespread conversion of forest to other land uses in this region will be accompanied by decreased infiltration and increased overland flow, potentially significantly altering water budgets and leading to reduced baseflows.
Ecohydrologic Modeling of Hillslope Scale Processes in Dryland Ecosystems
Dryland ecosystem processes are governed by complex interactions between the atmosphere, soil, and vegetation that are tightly coupled through the mass balance of water. At the scale of individual hillslopes, mass balance of water is dominated by mechanisms of water redistribution which require spatially explicit representation. Fully-resolved physical models of surface and subsurface processes require numerical routines that are not trivial to solve for the spatial (hillslope) and temporal (many plant generations) scales of ecohydrologic interest. In order to reduce model complexity, we have used small-scale field data to derive empirical surface flux terms for representative patches (bare soil, grass, and tree) in a dryland ecosystem of central Kenya. The model is coupled spatially in the subsurface by an analytical solution to the Boussinesq equation for a sloping slab. The semi-analytical model is spatially explicit driven by pulses of precipitation over a simulation period that represents many plant generations. By examining long-term model dynamics, we are able to investigate the principles of self-organization and optimization (maximization of plant water use and minimization of water lost to the system) of dryland ecosystems for various initial conditions and climatic variability. Precipitation records in central Kenya reveal a shift to more intense infrequent rain events with a constant annual total. The range of stable solutions of initial conditions and climatic variability are important to land management agencies for addressing current grazing practices and future policies. The model is a quantitative tool for addressing perturbations to the system and the overall sustainability of pastoralist activities in dryland ecosystems.
Topographic Controls on Spatial Patterns of Soil Texture and Moisture in a Semi-arid Montane Catchment with Aspect-Dependent Vegetation
Soil moisture exerts significant control over the partitioning of latent and sensible energy fluxes, the magnitude of both vertical and lateral water fluxes, the physiological and water-use characteristics of vegetation, and nutrient cycling. Considerable progress has been made in determining how soil characteristics, topography, and vegetation influence spatial patterns of soil moisture in humid environments at the catchment, hillslope, and plant scales. However, understanding of the controls on soil moisture patterns beyond the plant scale in semi-arid environments remains more limited. This study examines the relationships between the spatial patterns of near surface soil moisture (upper 5 cm), terrain indices, and soil properties in a small, semi-arid, montane catchment. The 8 ha catchment, located in the Cache La Poudre River Canyon in north-central Colorado, has a total relief of 115 m and an average elevation of 2193 m. It is characterized by steep slopes and shallow, gravelly/sandy soils with scattered granite outcroppings. Depth to bedrock ranges from 0 m to greater than 1 m. Vegetation in the catchment is highly correlated with topographic aspect. In particular, north-facing hillslopes are predominately vegetated by ponderosa pines, while south-facing slopes are mostly vegetated by several shrub species. Soil samples were collected at a 30 m resolution to characterize soil texture and bulk density, and several datasets consisting of more than 300 point measurements of soil moisture were collected using time domain reflectometry (TDR) between Fall 2007 and Summer 2008 at a 15 m resolution. Results from soil textural analysis performed with sieving and the ASTM standard hydrometer method show that soil texture is finer on the north-facing hillslope than on the south-facing hillslope. Cos(aspect) is the best univariate predictor of silts, while slope is the best predictor of coarser fractions up to fine gravel. Bulk density increases with depth but shows no significant relationship with topographic indices. When the catchment average soil moisture is low, the variance of soil moisture increases with the average. When the average is high, the variance remains relatively constant. Little of the variation in soil moisture is explained by topographic indices when the catchment is either very wet or dry; however, when the average soil moisture takes on intermediate values, cos(aspect) is consistently the best predictor among the terrain indices considered.
Effect of soil water content on soil thermal conductivity under field conditions
Knowledge of the thermal properties of soils is required in many areas of engineering, meteorology, agronomy, and ecosystem and soil science. Soil thermal conductivity varies in time and space, since it is influenced by soil properties as well as soil temperature and moisture conditions. We use the one dimensional heat conduction equation in conjunction with two-year data measured in a grass-covered field in North Carolina Piedmont to estimate soil thermal conductivity and to investigate how it is impacted by water content. In agreement with laboratory experiments reported in the literature, our results suggest that under dry conditions soil thermal conductivity increases across a relatively narrow range of soil water contents, above which a further increase in water content does not significantly change thermal conductivity. However, when soil approaches saturation, heat transfer is further improved, a fact not previously noted. This nonlinear behavior is consistent with the formation at high water contents of a continuous film of liquid water in soil aggregates of mineral and organic matter.
A button heat pulse probe for simultaneous measurements of soil thermal properties, water content and solute concentration
The heat pulse method is advantageous for simultaneous measurements of soil thermal properties, water content and solute concentration in vadose zone applications. The heat pulse probe (HPP) usually consists of a heater needle and at least one thermistor needle. However, these needles can be deflected during installation, changing the distance between the needles and causing measurement errors. Alternatively, we present a robust button shaped HPP design, without needles. The temperature response at the center of a 12-mm diameter ring heater, attached to a 16-mm diameter and 6-mm thick plastic disk, provides for soil thermal properties and water content. Solute concentration is evaluated by measuring the electrical resistance between the ring and the center electrode. After calibration, we test the measurement capabilities of the button HPP in different soils. The different water contents and solute concentrations estimations from the probe are compared to independent measurements. Compared to the needle HPP, we find that the unique ring-heat-source configuration increases the sensitivity of temperature response to changes in soil water content, providing an improved HPP.
Hydraulic Response Time Study of Ceramic Heat Dissipation Sensors
Heat dissipation sensors are used extensively to measure soil matric potential in the range -2.5 MPa to -0.01 MPa. Water flux between the surrounding soil and the sensor changes the thermal conductivity of the porous ceramic which changes the temperature increase near the heating element. This temperature increase is used for calibration to matric potential. This work studied the hydraulic properties of the sensor/soil system and estimates the response time of the sensor for a range of initial water potentials and soil textures. With the system near hydraulic equilibrium, a gradient is imposed and the response time is the time for the system to return to near equilibrium. The system is modeled using Hydrus and the results are compared to laboratory measurements. The results demonstrate the high sensitivity of response time to soil texture. Thermal properties are used to estimate measurement error when sensor/soil flux is greater than zero.
A new Approach to Soil Water Retention (Drying Branch)
Available models of soil water retention for both swelling and non-swelling cases with all essential differences between them are, in part or totally, based on curve-fitting to relevant experimental soil water retention data. The models use either parameters of a pore-size distribution in the fitting or parameters of some mathematical approximation (different for different models) of a retention curve. At least a part of the parameters have no clear physical meaning and can only be found by fitting. As a consequence, although the models can be of practical use for engineering applications, their possibilities from the viewpoint of advancement in the physical understanding and knowledge of the links between inter- and intra-aggregate soil structure and soil water retention as a function of the structure, are, even in the best case, limited. Recently the possibility of the physical prediction of another key soil characteristic - the shrinkage curve - was shown (Chertkov, 2007a, b, 2008a, b). These works permit one to explain soil shrinkage from inter- and intra-aggregate soil structures and without fitting (after many years of the domination of curve-fitting approaches). The results of these works strengthen the certainty that the physical prediction of soil characteristics is not hopeless, but just a difficult problem. The objective of this work is to slightly "dilute" the curve-fitting domination and to suggest some physical alternative as applied to the consideration of soil water retention (drying branch) in the general case, i.e., for swell-shrink soils. The attempt to be proposed relies on the concepts and results of recent works devoted to pure clay water retention (Chertkov, 2004) and soil shrinkage (Chertkov, 2007a, b, 2008a, b). The physical model to be presented includes three parts. We consider: (i) how in the general case a soil water retention curve can be presented through water retention curves of a contributive clay, an intra-aggregate matrix, and a system of internally saturated aggregates; (ii) essential improvement of the available clay water retention model (Chertkov, 2004); and (iii) water retention of a rigid matrix that can consist of sand grains or rigid water saturated aggregates. Then, using the model, we analyze available data to substantiate it. Chertkov V.Y. 2004. A physically based model for the water retention curve of clay pastes. Journal of Hydrology. 286, 203-226. Chertkov V.Y. 2007a. The reference shrinkage curve at higher than critical soil clay content. Soil Sci. Soc. Am. J. 71(3), 641-655. Chertkov V.Y. 2007b. The soil reference shrinkage curve. Open Hydrology Journal. 1, 1-18. Chertkov V.Y. 2008a. Estimating the aggregate/intraaggregate mass ratio of a shrinking soil. Open Hydrology Journal. 2, 7-14. Chertkov V.Y. 2008b. The physical effects of an intra-aggregate structure on soil shrinkage. Geoderma. 146, 147-156.
An Improved Technique for dry Soil Moisture Release Curves to Determine Soil Mineralogical and Physical Properties
Soil moisture release curves (MRC) or moisture sorption isotherms, which relate the amount of water in soil to its water potential or water activity, have many applications in soil physics and geotechnical engineering including determining soil water flow, specific surface area, swelling potential, and clay mineralogy and activity. Although research showing MRC for various soils dates back more than 50 years, limitations with the measurement technique have made developing MRC time consuming and inaccurate, especially in dry soils. Recently, an instrument was developed to create moisture sorption isotherms for various food and pharmaceutical products. The objective of this research was to investigate its use in soils for obtaining MRC in dry soils simply and accurately. Several different soil types were tested in the instrument from pure sand to bentonite and smectite clays. From the MRC of these soils, we were able to develop good correlations between actual and derived clay activity, surface area, and swelling potential. In addition, we were able to see hysteresis in dry soil water uptake for all soils, including sand. According to our tests, this new instrument will provide a powerful tool to investigate several soil physical properties simply and accurately.
Determination of Preferential Flow Parameters by Means of Inverse Simulation of Tension Disc Experiment
The field tension and ponded infiltration experiments were conducted to estimate the soil hydraulic properties of the soils with preferential pathways (Distric Cambisol, Sumava). Zones of preferential flow were determined through analyses of photographs taken during laboratory dye tracer infiltration experiments performed on undisturbed soil samples. Connectivity, volumetric ratio and spatial development of preferential pathways were evaluated as the necessary information for numerical simulations of flow using dual-permeability approach. The field infiltration experiment was carried out in a shallow pit for a period of one day. The upper boundary condition was controlled by the tension disk infiltrometer, the propagation of a water front was monitored by two tensiometers installed in two depths below the infiltration disk. 2D axisymetric numerical simulations were conducted to evaluate the results of the experiment. Two different approaches were used: 1. Single-domain approach based on Richards' equation. 2. Dual-permeability approach based on two interacting water flow domains (matrix and preferential domains), each governed by one Richards' equation. In the first simulation, the reference parameters derived from retention curves obtained by standard pressure extractor method were taken as properties of the soil matrix. The input hydraulic parameters were subsequently inversely optimized. In the second approach, the saturated hydraulic conductivities of the preferential flow domain were optimized to fit rapid response of the tensiometers after the start of ponded infiltration. Objective function consisted of infiltration fluxes and suction pressure head data. The parameter estimator PEST coupled with the simulation code S2D_DUAL (Vogel et al.,2000) were used. Concerning the existence of preferential flow on investigated soil, the dual-permeability model gives a better picture of the flow regime. The research has been carried out within the projects VZ03 CEZ MSM 6840770005 and GACR 103/08/1552. The demonstration of the results on AGU was supported by CTU 0820411.
Development of Error Parameters for Infiltrometers
Infiltration rates are critical to hydrologists' understanding of the behavior of a soil during rainfall events, as well as for predictive estimates of runoff potential. Current infiltration studies rarely account for inherent procedural error associated with this technique due to the large numbers of samples required. The objective of this study is to establish error parameters for use with double-ringed infiltrometers. In Fall of 2007 and 2008, infiltration rates of soils determined to be statistically similar on the basis of bulk density, texture (by class), slope, and vegetation were measured and analyzed. The soils' equilibrium infiltration rates (EIRs) were used to develop error parameters for standard use of double-ringed infiltrometers. The EIRs of soils of the same textural class, as well as statistically similar bulk densities were compared, and a 95% confidence interval was established. The hypothesis is that EIRs of soils of like textural class, vegetation cover, slope, and statistically similar bulk densities will fall within one standard deviation of the mean. Additional infiltrations were then completed to assess the viability of using the determined confidence interval as a predictor range of the EIRs of similar soils.
Improved Artificial Neural Network-Pedotransfer Functions (ANN-PTFs) for Estimating Soil Hydraulic Parameters
ANN-PTFs have become popular means of mapping easily available soil data into hard-to-measure soil hydraulic parameters in the recent years. These parameters and their distributions are the indispensable inputs to subsurface flow and transport models which provide basis for environmental planning, management and decision making. While improved ANN prediction together with the preservation of probability distributions of hydraulic parameters in ANN training is important, ANN-PTFs have been typically found using conventional ANN training approach with the mean square error as an error function, which may not preserve the probability distribution of the parameters. Moreover, the conventional ANN training can itself introduce correlation among predicted parameters and could not preserve the actual correlation among the measured parameters. The present study describes approaches to deal with such shortcomings of conventional ANN- PTF training algorithms by using new types of error functions and presents a group of improved ANN-PTF models developed on the basis of the new approaches with different levels of data availability. In the study, the bootstrap method is used as part of ANN-PTF development for generating independent training and validation sets, and calculating uncertainty estimates of the ANN predictions. The results demonstrate the merit of the new approaches of the ANN training and the physical significance of various types of less costly soil data in the prediction of soil hydraulic parameters.
Modeling Spatial and Temporal Variability of Soil Moisture in Shallow Depths of the Vadose Zone: A Comparison of two and Three Dimensional Simulations to Capture Relevant Physical Processes
The distribution of water saturation of soils near the ground surface is of interest in various applications involving soil moisture variations due to land-atmospheric interaction, evaporation from soils and land mine detection. Natural soil heterogeneity in combination with water flux conditions at the soil surface creates complex spatial and temporal distributions of soil moisture in the near-surface vadose zone. Validation of numerical models that are designed to capture these processes is difficult due to the inherent complexities of the problem and the scarcity of laboratory data with accurately known hydraulic parameters. A few 3-D experimental studies have been performed in attempts to generate such data. However, these experiments are tedious to setup and many challenges exist in getting accurate spatially and temporally varying measurements of water saturation and pressure. As a result, most of the experimental studies simulating multiphase flow processes in the heterogeneous vadose zone are carried out in 1-D or 2-D test systems. The issue is then to determine whether results obtained in such simplified conditions capture the relevant physical processes occurring in real 3-D heterogeneous situations. A numerical study was conducted to compare the spatial and temporal variability of soil moisture in a 3-D heterogeneous synthetic aquifer with the predictions of simplified 2-D models of vertical slices of the aquifer. The heterogeneous medium is composed of five different sandy materials, with air entry pressures ranging from 9.7 to 81.8 cm and saturated hydraulic conductivities ranging from 0.597 to 0.0067 cm/s. The numerical experiment designed around a synthetic 3-D aquifer consists of (1) simulating the drainage of the synthetic aquifer, starting from a fully saturated situation, and (2) inducing evaporation at the surface after liquid drainage has ceased. We compare results from 3-D and 2-D numerical simulations at several point locations, representing a network of sampling ports for monitoring water saturation and capillary pressure. It is found that in the specific situation where an inclusion of coarse sand is surrounded by fine sand, water entrapment can occur during the drainage phase due to a lack of air availability. In a 3-D multiphase simulation, the number of channels available for air to move into materials with a low air entry pressure increases as compared to a 2-D situation. Hence, the potential occurrence of water pockets is usually lower in 3-D simulations as compared to 2-D simulations. However, in most cases, 2-D slice models can reasonably approximate the retention behavior of the 3-D heterogeneous aquifer.
Simulation of Infiltration Into Organic-covered Permafrost Soils
Infiltration into frozen or unfrozen soils is critical in permafrost hydrology, controlling active layer soil water dynamics and influencing runoff. Quantifying the infiltration process in permafrost soils is made difficult by variable ground thawing and freezing and the layered soil profile that frequently has organic soils atop mineral horizons. Moreover, harsh environments impose technical and logistic difficulties in accurately monitoring processes experimentally. Few Land Surface Models (LSMs) and Hydrological Models (HMs) have been developed, adapted or tested for frozen conditions and permafrost soils. A need exists to improve these models to better represent the hydrology of permafrost regions, which are undergoing rapid environmental change worldwide. In this study, three infiltration algorithms commonly used in LSMs and HMs were tested against detailed measurements at four sites in Canada's discontinuous permafrost region with organic soil depths ranging from 0.02 to 3 m. Continuous measurements of liquid soil water content, precipitation, air and soil temperatures, snow depth and density and ground thaw were monitored during two consecutive years. Total soil water content (frozen and liquid) was monitored using twin probe gamma attenuation at three sites. Soil infiltration and runoff components estimated from the above measurements were then used to test simulations by the three algorithms. Parameters for the hydraulic conductivity-soil water potential and soil water potential- liquid water content relations were acquired for organic soils. Three soil hydraulic parametrization schemes were also tested. All infiltration algorithms and soil hydraulic parametrization methods were coded into the Simultaneous Heat and Water (SHAW) numerical model to conduct the comparison to ensure the identical inputs, parameters and intermediate process simulations. Preliminary analysis indicates: (1) the single most important factor that controls the infiltration process at the sites is thaw depth, indicating coupled simulations of heat and water transfer to quantify the thaw/freeze status are required in infiltration simulations for permafrost regions; (2) differences among the results by different algorithms were small during the early thawing stage and became greater as the active layer developed; (3) all the three soil hydraulic parametrization methods gave the similar results when appropriate parameters were chosen. Results of this study provide guidelines and may be directly implemented in LSMs and HMs for their applications in permafrost regions.
Simulating One Dimensional Water Flow and Solute Transport in Arid Soils
Field and laboratory experiments and modeling approaches have provided us with important clues for understanding flow and transport processes in the vadose zone. Several models have been developed and tested for simulating the movement of water and solutes in porous media. However, the spatial variability of soil hydraulic properties poses serious challenges to the application of such models to the real world conditions. The objectives of this study were to investigate the applicability of a finite element model (HYDRUS) for simulating water movement in an experimental waste burial site in the arid region of southwestern Nevada, USA, and to examine the role of preferential flow in solute transport in those soils. We used the available soil water content by layers and climatic data from ADRS(Amargosa Desert Research Site) between1998 and 2000 as input for the numerical. To investigate the role of preferential solute transport, one- and two-region models of the convection-dispersion equation (CDE) were applied to the solute breakthrough data obtained from miscible displacement experiments using undisturbed soil cores. Results showed that Hydrus-1D can be used to estimate the soil moisture profiles in the arid region and the role of preferential solute transport was not significant in a situation where the soil water potential was lower than ¡§C10 cm matrix head. The results from this study may be useful to construct a coupled model for estimating the water flow and solute transport in response to changes in infiltration and evapotranspiration in arid regions.
Quantifying the Distribution and Dynamics of Managed Aquifer Recharge Using Mass- Balance and Time-Series Thermal Methods
Managed aquifer recharge (MAR) is becoming increasingly popular for supplementing fresh water resources, helping to limit ground water overdraft and improve both the quantity and quality of available water. Most MAR systems are operated above a vadose zone, and usually recharge rapidly during an initial phase of diversion, but recharge typically slows considerably within subsequent weeks to months as sedimentation, biofouling, soil compaction, drainage at the base of the wetting front, and other processes reduce the hydraulic conductance below the percolation pond. We instrumented a MAR recharge pond above a shallow aquifer in central coastal California, to quantify variations in rates and locations of recharge, and to measure changes in soil properties with time during a recharge season. Careful measurements and calculations of diversion rates, precipitation, evaporation, pond water levels, and pond volume and area were used to construct a history of whole-pond seepage. Shallow piezometers were installed in the base of the pond and instrumented with autonomous temperature sensors and loggers prior to the start of the recharge season, allowing us to use heat as a tracer of fluid flow based on time-series analysis. Autonomous pressure loggers installed in the same locations allow quantification of head gradients with time. When this information is combined with seepage rates, we can determine absolute values of the hydraulic conductance of the saturated soil at the base of the pond, including changes in these values with time. The recharge pond is 3 km2 in area and typically recharges at the start of the MAR operating season at a mean rate of 2 m/day, based on a whole-pond mass balance and bottom area, but this rate tapers off abruptly after 4-8 weeks of operation. Point-specific seepage rates vary enormously throughout the recharge cycle across the pond base, with some areas allowing recharge at rates in excess of 10 m/day, and other areas being virtually stagnant. Collection of soil samples before and after the seepage season allows assessment of accumulating fine-grained sediment during MAR, and geochemical data from the aquifer below the pond help to resolve the fraction of the pond that contributes most to recharge. Studies such as these generate improved understanding of recharge processes in general, with MAR systems providing controlled windows into subsurface conditions and processes.
Quantitative Uncertainty Analysis of Multiphase flow in Non-stationary Heterogeneous Porous Media
In this study, we develop an approach based on Karhunen-Loeve (KL) decomposition to investigate the stochastic behavior of two-phase flow (water-NAPL) in a non-stationary heterogeneous field consisting of multiple statistically homogeneous units with arbitrary shape. The statistical properties, such as mean, variances, correlation length of hydrological properties (intrinsic permeability, pore size distribution parameter, van Genuchten fitting parameter etc.) for each unit differ from each other. These spatial random functions for each unit are decomposed by the KL approach, yielding a set of eigenvalues and eigenfunctions, based on which the eigenvalues and eigenfunctions for the entire domain can be found. The mean and variances of water and NPAL phase pressure, saturation can be solved using the stochastic KLME model developed by Chen et al. (Water Resources Research, 42, 2006). Monte Carlo simulations are conducted as ¡§true¡¨ solutions to validate the accuracy of the proposed efficient stochastic approach. This study makes the stochastic multiphase flow model based on KLME approach available as a more applicable uncertainty analysis tool in practical environmental remediation site or petroleum reservoir modeled as multiphase flow field, since the natural subsurface are usually formed by irregular hydrological, geological units.
Numerical Study of Capillary Barrier Effects on Hillslopes
On hillslopes, where coarser soil material layers are overlain by finer soils, the lower layer may function as a capillary barrier, which impedes a vertical flow. As a consequence, water flows laterally in the direction of the upper boundary of the lower inclined layer (funneled flow). Capillary barriers are successfully used as alternative insulation layers in engineering systems such as waste disposal sites or cuttings in road engineering. Under natural conditions, capillary barrier effects may significantly affect the soil water regime. Under specific conditions these effects may contribute to the occurrence of large scale landslides (Guadagno et al, 2005). The formation of capillary barrier related funneled flow at an inclined slope, as a special form of preferential flow, has been studied using the simulation model S2D for selected combinations of soil layers with distinct hydraulic properties. All scenarios were modeled by means of classical Richards' theory, using single- or dual-permeability approach. Sensitivity analysis has been performed to study the triggering conditions in terms of the upper layer critical saturation under given morphological and climatic conditions. The results of the analysis could help when numerical simulations are used to analyze key aspects and rules with regard to urban planning in potentially hazardous areas. The study has been supported by the research project VZ 04 CEZ MSM 6840770005.
Evaluating locally conservative finite element methods for flow in the vadose zone
Accurate simulation of variably saturated flows can be difficult for many reasons. Soil properties often vary greatly over the problem domain. The underlying balance equations are highly nonlinear, and solutions may develop sharp fronts at the scale of interest. It is clear that straightforward application of standard Galerkin finite element methodologies is inadequate for these problems. Local oscillations can appear around sharp fronts without the use of mass-lumping, and velocity fields obtained from differentiation of pressure fields are discontinuous at element boundaries. Over the years, a number of alternatives like mixed finite elements, discontinuous Galerkin methods, or control volume finite element approximations have been proposed. Recently, stabilized conforming approximations combined with local velocity postprocessing schemes have shown promise as well. Here, we consider several locally conservative finite element strategies for full two-phase air-water flow and Richards' equation, including stabilized conforming approximations, nonconforming schemes, and discontinuous Galerkin approximations. We evaluate the methods in terms of their solution accuracy, velocity approximation, and ability to control oscillations around sharp fronts. We further compare the overall efficiency of full two-phase model formulations and Richards' equation when combined with appropriate numerical solution schemes.
Rapid Recharge to Perched-Intermediate Groundwater Zones, Pajarito Plateau, Los Alamos, New Mexico
The Los Alamos National Laboratory continuously monitors groundwater levels and surface-water discharge at over 150 locations on the Pajarito Plateau. The resulting data sets were analyzed to help identify locations where surface water and shallow alluvial groundwater (generally <30 ft) recharge deeper perched- intermediate groundwater (approximately 200 to 700 ft bgs) zones. Runoff from snowmelt and summer rainstorms recharges the shallow alluvial groundwater. The vadose zone beneath the alluvial groundwater ranges in thickness from 600 to 1200 ft. Typical estimated annual infiltration rates vary spatially from <1mm/yr to 1000 mm/yr. However, localized rapid recharge can sometimes yield transient infiltration events that are detected 200 ft below ground surface within a few days. The recharge events are identified by the observed water-level fluctuations in perched-intermediate groundwater monitoring wells. Preliminary analysis of the data suggests that the mechanism for this rapid recharge is spatially-limited fracture flow through basalt and welded volcanic tuff. The rapid recharge has been observed at three separate areas in response to both snowmelt and stormwater runoff events. One recharge area located on the west side of the plateau is associated with a densely-welded, fractured, volcanic tuff that is also possibly faulted. Surface runoff and alluvial groundwater in drainages presumably infiltrates faulted and fractured tuffs, recharging perched- intermediate groundwater at depths of 600 to 700 ft. Two other areas of rapid infiltration occur on the eastern side of the plateau where thin nonwelded tuff and/or thin alluvium in canyon bottoms overlie highly fractured basalt. In these areas surface water and alluvial groundwater easily gain access to the shallow basalt beneath the canyon floor. Infiltration then takes place along fractures and recharges perched-intermediate groundwater within deeper basalt units. Identifying these areas is important for establishing an effective groundwater monitoring network and for refining conceptual models of infiltration mechanisms and contaminant pathways.
Impact of Soil Layering on Evaporation Driven Flow and Transport in Arid Soils
Coupling between soil development, hydrologic processes, and plant growth in arid regions is not well understood. Here, we integrate field, lab and numerical investigations to study the impact of soil layering on evaporation driven flow and transport in arid soils. Specifically, two hypotheses are proposed: 1) soil horizon development may significantly impact evaporation rate and spatio-temporal chemical species redistribution in arid soils; 2) differences in layering between soils beneath plant canopies and nearby interspace (bare soils) may significantly influence evaporation-driven upward water flow and solute transport. Field samples were collected from two 1-m deep soil pits in Eldorado Valley, approximately 50 km from Las Vegas, Nevada. One soil pit was located beneath a creosote bush, the other from the adjacent interspace. The overall concentrations of K+, Na+, Ca2+, Mg2+, SO42-, and soluble salts in soils under plant canopy are much higher than those from the interspace. Salts accumulated at depths from 60 to 90 cm under the canopy; in contrast, the salt concentrations in bare soils were more uniform and may accumulate in relatively deeper layers. Soil samples taken from the pits will be packed into layered and non- layered columns, respectively, to examine the effects of soil layering on evaporation flow and transport. Evaporation rate, soil-water matric/osmotic potential, and temperature gradients in each column will be continuously monitored. Upward flow and transport in different soil layering under various conditions will be simulated using the HYDRUS model. It is expected that less soil horizon development will lead to higher evaporation rates, resulting in lower volumetric water content and higher accumulation of salts in the uppermost soil horizons.
Reflection and transmission of elastic waves at porous/porous elastic media saturated by two immiscible fluids
Elastic wave phenomena in porous media containing compressible viscous fluids are of considerable interest to diverse engineering applications. The behaviors of reflection and transmission of elastic waves at a plane interface between two different porous elastic half-spaces saturated by two immiscible fluids are investigated based on the theory of poroelasticity developed by Lo et al. (2005) for a two-fluid system, where inertial and viscous couplings both in an Eulerian frame of reference are considered. The amplitudes and energy ratios have been determined theoretically not only for the different reflected and refracted waves but also for different angles of incidence for P and SV waves which propagate through two different porous elastic half- spaces saturated by two immiscible fluids. It is showed numerically that the amplitudes of the second and third refracted P waves are usually much smaller than those of the other reflected and refracted P and SV waves. Especially, the amplitudes of the third refracted P waves are the least. The energy dissipation at the interface is observed, which is identical with the results of the literature studies.
Three-Dimensional pore Structure Analysis for Peat materials Using Microfocus X-ray Comuputed Tomography
Soil-pore geometry including size distribution, total and air-filled porosity, pore tortuosity and connectivity, and anisotropy of pore network strongly affects water and air flow in soils. This study investigated three- dimensional (3-D) images of pore structure for peat materials by a microfocus X-ray CT system and evaluated the soil-pore geometric characteristics using a multi-dimensional scanning line method. Undisturbed peat soil blocks were sampled from the soil surface down to 140-cm depth at Bibai wetland, Hokkaido, Japan. Soil cores (100 cm3) were taken from the blocks in the laboratory, and were used for the experiments. Microfocus X-ray scanning was conducted for the soils cores of peat materials treated under different conditions: intact, drainage at different matric potentials (pF), consolidation under different uni-axial loads. The treated samples were freeze-dried prior to the microfocus X-ray scanning. The scanned 3-D X-ray CT images were converted to black and white images (binarized), and were used for quantitative and qualitative analysis of peat materials. The visualized CT images captured well macropore networks of a pore diameter larger than around 30 μm, and showed clearly anisotropic characteristics of the macropore networks. The 3-D structural anisotropy of pore networks was evaluated using a multi- dimensional scanning line method that employs straightforward image analysis, and its results were visualized using a stereonet projection. Soil-water and gas transport parameters such as water conductivity, air conductivity, and soil-gas diffusion coefficient for the treated peat materials were also measured. The pore geometric characteristics based on measured soil-water and gas transport parameters will be combined with the results of CT image analysis.
Noninvasive Imaging of Tracer Experiments in a Soil Column
A set of tracer-infiltration experiments on soil columns by means of magnetic resonance imaging (MRI) was performed. Computed tomography (CT) was applied in order to map the spatial distribution of porous media, namely the local densities and porosities, and their variation within the soil sample under test. The CT visualisation was done in order to trace disturbances in the structure as a possible source of preferential flow. By means of MRI the flow paths during the infiltration experiment were visualized using a tracer pulse containing Ni(NO3)2 in a concentration of 0.05 mol/litre. The pulse was added under hydraulic steady state conditions. The tracer motion was monitored through its effect on the signal relaxation of 1H using a 7 Tesla vertical magnet system equipped with a 40 mm RF probe. The boundary condition at the top of the soil columns was maintained using a dripping system connected to a HPLC pump with flow rate set to 0.5 ml/min. Free outflow was used as the bottom boundary condition. The vertical component of the local velocity value was calculated after the experiment. Small disturbances in the tracer front observed during the break-through could be related to the preferential flow phenomena in combination with the air bubble entrapment. This research has been supported by research project SP/2e7/229/07 and DBU - Deutsche Bundesstiftung Umwelt.
Visualizing Rhizosphere Soil Structure Around Living Roots
The rhizosphere, a thin layer of soil (0 to 2 mm) surrounding a living root, is an important interface between bulk soil and plant root and plays a critical role in root water and nutrient uptake. In this study, we used X-ray Computerized Microtomography (microCT) to visualize soil structure around living roots non-destructively and with high spatial resolution. Four different plant species (Helianthus annuus, Lupinus hartwegii, Vigna radiata and Phaseolus lunatus), grown in four different porous materials (glass beads, medium and coarse sand, loam aggregates), were scanned with 10 ìm spatial resolution, using the microtomography beamline 8.3.2 at the Advanced Light Source, Lawrence Berkeley National Laboratory, Berkeley, CA. Sample cross section images clearly show contacts between roots and soil particles, connecting water films, air-water interfaces as well as some cellular features of the plants taproots. We found with a simulation experiment, inflating a cylindrical micro-balloon in a pack of air-dry loam aggregates, that soil fracturing rather than compaction might occur around a taproot growing in dry soil. Form these preliminary experiments, we concluded that microCT has potential as a tool for a more process-based understanding of the role of rhizosphere soil structure on soil fertility, plant growth and the water balance at the earth-atmosphere interface.
Measurements of Capillary Pressure-Saturation Relationships for Silica Sands Using Light Transmission Visualization and a Rapid Pseudo Static Method
Measurement of water saturation in porous media is essential for many types of studies including subsurface water flow, subsurface colloids transport and contaminant remediation to name a few. Water saturation (S) in porous media is dependent on the capillary pressure (Pc) which, in turn, depends on the interfacial tension between wetting and non-wetting phases, the pore size distribution of the porous material as well as the contact angle between fluids and porous media. Traditionally, measurements of the Pc-S relationship were carried out using pressure cells but these methods can be very time consuming. The objective of this research was to implement the innovative technique of light transmission visualization (LTV) to investigate the Pc-S relationships for porous media, and compare the results to an independent technique, namely a rapid pseudo-static automated method that was recently developed by Chen et al. Two dimensional (2-D) chambers constructed of two glass plates were packed with either 20/30 or 40/60 silica sand and then fully saturated with water. Images of the entire 2-D chambers were captured using a charge- coupled device (CCD) camera and analyzed pixel by pixel to determine air and water saturation in the porous media. Variable saturation of water and air across the height of the chamber was dictated by controlling the capillary pressure of the fluids. The primary drainage curve, main imbibition curve and secondary drainage curve were all measured, for the first time, using LTV and the results were compared to the results obtained by the rapid pseudo-static method. Overall, there was good agreement between the results of the two methods and variation in air entry pressure that was calculated based on Brooks-Cory model was within 2 cm for the sand sizes tested. Both techniques are capable of providing high spatial resolution data for the Pc-S relationships. Disclaimer This is an abstract of a proposed presentation and does not necessarily reflect EPA policy.
Influence of Variable Environmental Conditions on Presence and Concentration of Energetic Chemicals Near Soil Surface in the Vadoze Zone
Many explosive-related compounds (ERCs) are found near the soil-atmospheric surface in sites containing buried explosive devices, such as landmines and unexploded ordnance, detonation-residual, and munitions residues from explosive manufacturing facilities. Accurate assessment of the fate and transport processes is essential for predicting their movement to the surface, groundwater, or any other important environmental compartment. The transport processes controlling the direction and magnitude of the movement, and chemical and physical processes controlling the fate of the chemicals vary with environmental conditions. This research addresses the effect of variable rainfall, evaporation, temperature, and solar radiation on fate and transport of 2,4,6-Trinitrotoluene (TNT), 2,4-Dinitrotoluene (DNT), and other related chemicals in partially saturated soil. Experiments have been conducted in a laboratory-scale 3D SoilBed placed inside an environmental chamber equipped with rainfall and solar radiation simulators, and temperature control settings. The SoilBed was packed with a sandy soil. Experiments have been conducted by burying a TNT/DNT source, simulating a landmine, and applying different rainfall and light radiation cycles while monitoring DNT, TNT, and other related ERCs solute concentrations temporally and spatially within the SoilBed. Experiments include different source characteristics, rainfall intensities, temperatures, and radiation cycles to evaluate their effect on the detection and movement of ERC in soils in both aqueous and vapor phases. Temporal and spatial data has been analyzed comparatively and quantitatively. Comparative analysis was developed using surfer®- and voxler®-generated images and 3D visualization models applying spatial interpolation and masking methods. Single and multi-variable statistical analysis has been employed to determine the most important factors affecting the fate, transport and detection of ERC near soil-atmospheric surfaces. Results show that rainfall, radiation, and temperature variations influence the presence, transport, and concentrations of TNT and DNT near the soil surface. Higher concentrations are observed near the end of rainfall events, both in the aqueous and gaseous phases. Higher rainfall intensity results in higher presence and concentrations. Lower TNT and DNT concentrations than their solubility limit indicate rate-limited mass transfer, dissolution limitations, and dilution processes. Radiation events and higher atmospheric temperatures result in greater presence and concentrations of DNT and TNT, indicating influence of these factors on fate and transport processes. TNT degradation by-products measured mostly in the upper segments of the SoilBed, suggest degradation processes resulting from radiation-induced conditions near the soil-atmospheric surface. Although the ERC source consists of equal mass of TNT and DNT, greater detection density and concentrations are observed for DNT. A generalized linear mixed statistical model has been applied to quantify the effect of environmental conditions on ERC detection and concentrations. The statistical analysis indicates that rainfall events and related water contents are the most influential factors affecting the presence and concentrations of ERCs in the aqueous and gaseous phase. Solar radiation, and related heat flux, is the second most influential parameter. Although atmospheric temperature influence the presence and concentration of ERCs in soils, it is the least influential parameter.
Effect of Mineral Reactions on the Hydraulic Properties of Unsaturated Soils: Model Development and Application
Precipitation/dissolution induces changes in the pore radii of water-filled pores, and, consequently, affects
flow in porous media. The selective radius shift model was developed to relate changes in mineral volume
due to precipitation/dissolution reactions to changes in hydraulic properties of unsaturated soils. The model
considers the dependency of the amount of mineral precipitation/dissolution within a pore on the local pore
volume. Furthermore, it accounts for precipitation/dissolution taking place only in the water-filled part of the
pore space. The pore bundle concept was used to relate the pore-scale process of dissolution/precipitation
to changes in macroscopic soil hydraulic properties. In the numerical model, the finite change in mineral
volume at a discrete time step leads to a discontinuous pore-size distribution, because only the water-filled
pores are affected. This pore-size distribution is converted back to a discontinuous soil moisture
characteristic to which, at every time step, a new water retention curve is fitted under physically plausible
constraints. The model equations were derived for the commonly used van Genuchten/Mualem hydraulic
properties. Together with the selective radius shift model a head-based solution of Richards' equation for
aqueous phase flow was implemented into the geochemical modelling framework PHREEQC, thereby making
available PHREEQC's comprehensive geochemical reactions. The model was applied to kinetic halite
dissolution and calcite precipitation as a consequence of cation exchange in a variety of unsaturated flow
situations. The applications showed marked changes in the soil's hydraulic properties due to mineral
precipitation/dissolution and the dependency of these changes on the water content. Furthermore, it was
shown that the unsaturated hydraulic conductivity at fixed reduced water content can even increase during
precipitation due to changes in the pore-size distribution.
Irreversibility of 2,4-Dichlorophenoxyacetic Acid Sorption onto a Volcanic Ash Soil
Pesticide sorption and desorption in soils are key processes governing fate and transport of pesticides in the soil environment. The irreversibility (or hysteresis) in the processes of pesticide sorption and desorption needs to be known to accurately predict behavior of pesticides in soil systems. 2,4-dichlorophenoxyacetic acid (2,4-D) is a widely used pesticide in agriculture fields. However, only few studies of 2,4-D adsorption onto Andosols (volcanic ash soils) have been published, and the knowledge of 2,4-D desorption onto Andosols is very limited. In this study, a volcanic ash soil sampled from a pasture site in Nishi-Tokyo, Japan was used as a sorbent in order to investigate the irreversibility of 2,4-D sorption. For comparison, a pure clay mineral (kaolinite) obtained from Clay Science Society of Japan (CSSJ) was also used. 2,4-D solutions with three concentrations (0.011, 0.022 and 0.045 mmol/L) were prepared in artificial rain water (ARW= 0.085mM NaCl + 0.015mM CaCl2) to simulate field conditions. To prepare the sample solutions, the solid mass/liquid volume ratio of 1:10 was used for both sorbents (volcanic ash soil and kaolinite). The experiments were conducted in triplicate using a batch method under different pH conditions to examine the effect of pH. Desorption was measured during a equilibration procedure: After removal of 7 mL of supernatant in the sorption step, 7 mL of ARW excluding 2,4-D was added to the sample solution after which, it was equilibrated and centrifuged. The procedure was performed sequentially three or four times to obtain a desorption isotherm. Sorption and desorption generally followed Freundlich isotherms. The results showed markedly effects of pH on 2,4-D sorption and desorption in both the soil and kaolinite, with the percentage of sorption increasing with decreasing pH whereas the percentage of desorption decreased. There was a larger adsorption-desorption hysteresis in the volcanic ash soil as compared to kaolinite. Moreover, the largest hysteresis was found under the lowest pH condition indicating that the hysteretic effect likely depended on the variation of pH. In summary, the volcanic ash soil exhibited higher sorption coefficients and hysteresis than kaolinite, likely due to soil properties such as high organic matter content and the presence of different clay minerals.
Simple estimation of minimum unsaturated contaminant travel times at Rainier Mesa and Shoshone Mountain, Nevada Test Site
In the unsaturated zone the fastest travel times frequently occur via preferential flow that bypasses the soil/rock matrix. Experimental data provide compelling evidence that minimum solute travel times through preferential paths depend primarily on whether water supply is continuous versus non-continuous in time, with little influence from matrix hydraulic properties. We employ a simple model based on this "source- responsive" paradigm to estimate minimum preferential travel times to the regional water table for nonreactive radionuclides at Rainier Mesa and Shoshone Mountain within the Nevada Test Site. The radionuclides at the site originate from underground nuclear testing within a ~1-km-thick unsaturated zone. Contaminated sources at Rainier Mesa and Shoshone Mountain that are continuously supplied include ponded water in certain tunnels, filled detention basins, and partially-filled boreholes with detonation cavities. Tunnels without ponding and unfilled detonation cavities are considered non-continuous sources supplied by percolation of precipitation. Decades of geological and hydrological characterizations provide the foundation for establishing preferential flow as a viable transport mechanism at Rainier Mesa and Shoshone Mountain Our estimated minimum travel times via preferential flow for Rainier Mesa are one to two months for a continuously-supplied source and tens to hundreds of years for a non-continuous source. Previous studies in the scientific literature conducted isotopic analysis of fracture water collected in tunnels at Rainier Mesa that indicated transit times for 400 m of transport from land surface to tunnel levels of one to 40 years. Four monitoring wells in the carbonate aquifer have not detected radionuclide levels above the drinking water standards at Rainier Mesa. Travel times for both the continuously and non-continuously supplied sources at Shoshone Mountain are twice the Rainier Mesa estimates, resulting from longer transport distances and less precipitation. A shale confining unit at Shoshone Mountain may cause a transition from preferential to matrix dominated flow, increasing estimated transport times to more than a thousand years. Concentrations of radionuclides above the drinking water standard have not been detected in a deep well or in springs at Shoshone Mountain. Preliminary results from our analysis indicate the potentially rapid nature of unsaturated-zone preferential transport (despite a semi-arid climate) for continuous sources and highlight the value of a simple modeling approach to complement traditional, more complex flow simulation approaches for contaminant travel time estimation.
Gas Diffusion Coefficient in Variably Saturated Peat Soil: Development and Tests of Predictive Models
The soil-gas diffusion coefficient (Dp) and its dependency on air-filled porosity (ε) govern gas diffusion and reaction processes in soil. Accurate Dp(ε) prediction models for variably saturated peat soils are needed to evaluate vadose zone transport and fate of greenhouse gases such as methane in peaty wetlands. In this study, we measured Dp on undisturbed peat soil samples at different soil-water matric potentials, and developed new, linear and nonlinear expressions for describing and predicting Dp(ε). The new Dp(ε) models together with existing Dp(ε) models were tested against both measured data and independent data sets from literature. Twelve undisturbed 100cm3 peat soil cores were taken between the soil surface and down to 30-cm depth at Bibai wetland, Hokkaido, Japan. The soil cores were initially saturated with water, and drained at given matric potentials, pF=1.0, 1.5, 1.8, 2.0, 3.0, and 4.1 (where pF equals to log | Ψ | , Ψ: the soil-water matric potential in cm H2O), using the hanging water and pressure plate methods. At each matric potential, simultaneous measurements of volume shrinkage, soil-water retention, and Dp were conducted. Literature datasets of Dp(ε) for peat soil cores taken from different areas within the same wetland, specifically 12 samples from Iiyama and Hasegawa (2005) and 12 samples from Iiduka et al. (2008), were also used. A total of 191 measurements of Dp(ε) at pF ≤ 2.0 were applied for developing new Dp(ε) models for pF ≤ 2.0 where effects of shrinkage on Dp were assumed negligible. By modifying 3 existing Dp(ε) models, the Buckingham (1904) model, the Macroporosity-Dependent Model (MPD; Moldrup et al., 2000), and the Penman-Call model (Moldrup et al., 2005), we suggested 3 new Dp(ε) expressions for peat soil. In the Buckingham-based Dp(ε) model, a variable X(εƒw relationship (where X is the pore connectivity factor) derived from measurements was introduced in the Dp(ε) expression. In the Penman-Call-based Dp(ε) model, new expressions for the model parameters defining the linear Dp(ε) relationship, the slope of Dp(ε)/D0 and the threshold air-filled porosity where gas diffusion ceases due to complete water blockage, were derived from measured data. In the MPD-based Dp(ε) model, a new Dp,100(ε100) relationship for peat soil (where Dp,100 and ε100 are the gas diffusion coefficient and the air-filled porosity at pF2.0) was introduced in the model. To validate the new Dp(ε) models, we tested the models against independent data for peat soil samples from Freijer (1994). The new Dp(ε) models, except for the Penman-Call-based model, predicted well the independent data, and the Buckingham-based Dp(ε) model performed the best among the existing and newly-developed models.
Simulation of Carbon-14 Migration Through a Thick Unsaturated Alluvial Basin Resulting from an Underground Nuclear Explosion
Yucca Flat is one of several areas on the Nevada Test Site that was used for underground nuclear testing. Extensive testing performed in the unsaturated and saturated zones have resulted in groundwater contamination and surface subsidence craters in the vicinity of the underground test areas. Simulation of multiphase 14C transport through the thick Yucca Flat alluvial basin was performed to estimate the magnitude of radionuclide attenuation occurring within the unsaturated zone. Parameterization of the 14C transport in the multiphase flow and transport simulator (FEHM) was verified with experimental data collected from a large unsaturated soil column experiment. The experimental data included 14C as a radio-labeled bicarbonate solution, SF6 gas, and lithium bromide solution breakthroughs. Two representative simulation cases with working points located at shallow and deep depths relative to the water table were created to investigate the impact of subsidence crater-enhanced recharge, crater-playa areal extent, gas-phase partitioning, solid-phase partitioning, and a reduced permeability/porosity compressed zone created during the explosion on 14C transport. The representative shallow test had a detonation point located 175 m below land surface, and the deep test had a working point 435 m below land surface in a 500 m deep unsaturated zone. Carbon-14 transport is influenced by gas-phase diffusion and sorption within the alluvium. Gas-phase diffusion is an attenuation mechanism that transports 14C gas as 14CO2 throughout the unsaturated zone and exposes it to a large amount of soil moisture, resulting in dilute concentrations. The simulations indicated that the majority of the 14C inventory remains in the unsaturated zone over a 1,000-year time period after detonation because gas-phase diffusion moves the bulk of the 14C away from the higher recharge occurring in crater playas. Retardation also plays a role in slowing advective aqueous phase transport to the water table.
Vadose Zone Chromium(VI) Immobilization using Foam-Delivered Calcium Polysulfide (CaS5)
Calcium polysulfide (CPS; CaS5) is a remedial amendment that can be used to reduce and immobilize hexavalent chromium [Cr(VI)] in the vadose zone. However, how to deliver CPS to the vadose zone is a challenge. Water-based CPS delivery to the vadose zone is a concern because of the high mobility of Cr(VI). The water-based delivery will easily mobilize the pollutant therefore contaminate the underlying aquifer. Furthermore, the preferential flow of the CPS solution in the vadose zone due to the formation heterogeneity is very difficult to overcome, resulting in bypassing of zones with lower permeability. Foam has unique transport properties in the vadose zone that enable mitigation on the mobilization of mobile contaminants and enhance the sweeping over heterogeneous systems. Thus foam-facilitated transport is promising for the remedial amendment CPS delivery to the vadose zone to overcome the mentioned difficulties in water-based delivery. The delivery of calcium polysulfide to the vadose zone using foam and the immobilization of Cr(VI) via reduction by the foam-delivered CPS was studied in a series of batch and column experiments. Batch tests were conducted to select the foam-generating CPS-surfactant solutions, to determine the solution foamability and the reducing potential of CPS-containing foams, and to study the influence of foam quality, surfactant concentration, and CPS concentration on foam stability. Column experiments were performed to test the foam delivery of CPS to sediments under conditions similar to field vadose zone, to study the foam transport and interaction with sediments, and to determine the extent of Cr(VI) immobilization using this novel delivery approach. CPS-containing foams with high reducing potential were prepared based on the batch tests. Sediment reduction by foam-delivered CPS was observed in the columns. Massive mobilization of Cr(VI) from sediments occurred when CPS was delivered in aqueous solution. The Cr(VI) mobilization was minimized in column tests when CPS was delivered by foams, resulting in significant Cr(VI) in-situ immobilization. These results demonstrated for the first time that foam injection can be successfully used for CPS delivery and that foam-delivered CPS can be applied for Cr(VI) immobilization in contaminated vadose zones. Due to its advantage over water-based delivery, the foam delivery technology can also be applied for the delivery of other remedial amendments, such as citrate, acetate, and polyphosphate to the vadose zone for metals, radionuclides, and organic contaminants remediation.
Experimental Evaluation of Gas-Phase Transport and Reactivity of Two Organophosphate Compounds in Unsaturated Porous Media
An experimental study was undertaken to evaluate the feasibility of using organophosphate compounds that can be transported in the gas phase as a source of phosphorus for mineral formation (e.g. apatite) and contaminant sequestration in deep unsaturated zones. Previous work by others with gaseous phosphate compounds utilized triethyl phosphate (TEP) for bioremediation. In the current study we used both TEP and another chemically similar compound, dimethyl methylphosphonate (DMMP) that has a higher saturation vapor pressure. Batch abiotic degradation experiments in aqueous solutions with and without sediment (in both oxic and reducing conditions) indicate that both TEP and DMMP are very recalcitrant. Slow conversion from organic-to inorganic-P forms occurred (<5% in 3 months) under high temperature (80° C) and highly alkaline pH conditions. TEP and DMMP biodegradation to PO4 was found to be minimal over a similar time period using concentrated solutions of in situ microbes with no other growth substrates present. Gas transport studies using FTIR spectroscopy show that these compounds also adsorb very strongly to unsaturated sediments from the Hanford Site, to the extent that no breakthrough was observed even after >1000 pore volumes of gas exchange and complete dessication of the sediments. Methanol production was observed during the gas transport experiments, indicating that the lack of observed breakthrough of the original organophosphate compounds was attributable to both adsorption and reaction processes. FTIR reflection spectroscopy and microprobe analyses were performed to identify and quantify adsorbed species and possible mineral formation.
In Situ Monitoring of Soil Solution Nitrate in Saturated and Unsaturated Sandy Soil
A lack of in-situ instrumentation limits continuous monitoring of soil solution concentration to evaluate environmental (contaminants) and agricultural management (plant nutrients) practices. We developed a prototype soil solution monitoring technique, to measure long-term in-situ nitrate concentration, consisting of an in-situ stainless-steel porous cup, with real time concentration measurements using a UV fiber-optic sensor. The measurement technique does not require soil solution extraction, but is based on in-situ soil solution equilibration by diffusion between the porous cup and the surrounding medium. The technique is presented for nitrate solution at different soil moisture status using new designed of solution samplers. Analytical solutions are presented to evaluate solute diffusion coefficients, as controlled by a variety of soil water contents. The principles of operation are demonstrated for diffusion a saturated and unsaturated Oso Flaco Sand, illustrating the potential application solution samplers in a soil environment
Mitigating Vadose Zone Nitrogen Transport Under Land Use Change and Urbanization
The discovery of large accumulations of nitrate within the vadose zones of many desert ecosystems coupled with land use change from urbanization in these areas may be having a detrimental effect on the ground water quality, often the source of public water supplies of these regions. Land use change can result in the initiation or increase in aquifer recharge (from over-irrigation, leaking pipes, wastewater discharge, etc.) and has the potential to mobilize the observed stores of accumulated nitrate in the vadose zone. This research focuses on mitigation options to reduce mobilization of nitrate in the vadose zone by stimulating denitrification reactions during transit through the vadose zone prior to reaching the underlying aquifers. Laboratory experiments using typical vadose zone materials from Spanish Springs, Nevada, conducted in 1 meter columns have been prepared with a nitrate rich soil layer (1000 ppm KNO3, labeled with 2% KNO3 - N15 isotopic tracer) designed to simulate a nitrate accumulation zone. All columns are irrigated with treated wastewater (effluent) at a rate of 0.5 cm/day to simulate excess irrigation of urbanized parklands. In addition to a control column, one column's irrigation is augmented with dextrose (C6H12O6) designed to provide sufficient carbon sources, when combined with higher water content, to promote microbial denitrification. A third column is treated with a compost and soil mixture at the surface of the soil to provide an alternative method of producing dissolved organic carbon to be advected with the infiltration to the nitrogen storage regions within the column. All columns are instrumented with volumetric water content probes, tensiometers, and soil solution samplers. Initial results from soil solution analysis indicate the possibility of nitrate reduction by as much as 50 % of the initial concentration, after only 14 days of residence time within the dextrose amended column. Analysis will be presented comparing the treatments as well as analysis of dissolved carbon and N-15 analysis of biological reduction.
Laboratory Testing of Foundry Sands as Bulking Agents for Porous Media Filters Used to Treat Agricultural Drainage Waters
Foundry sands are industrial byproducts that may have potential application as bulking agents that when mixed with small amounts of more chemically reactive materials (i.e. sulfur modified iron, fly ash, etc.) can be used to produce porous media filters capable of removing contaminants from agricultural drainage waters. Foundry sand bulking agents are attractive primarily as a low cost means to maintain the hydraulic efficiency of a filter. Secondarily, the foundry sands themselves may have some capacity for removal of agricultural nutrients and pesticides from water. Consequently, a laboratory study was initiated to quantify hydraulic efficiency and agricultural contaminant removal abilities of six foundry sands. Of the six foundry sands tested, all were obtained in central Ohio, three from iron casting foundries, two from steel casting foundries, and one from an aluminum casting foundry. Hydraulic efficiencies of the foundry sands were assessed by measuring hydraulic conductivity with twice replicated falling-head permeability tests. Batch tests were employed to evaluate foundry sand potential to treat water containing nitrate and phosphate nutrients, along with the pesticide, atrazine. Five of the six foundry sand samples had measured hydraulic conductivity values from 7.6 x 10-3 cm/s to 3.8 x 10-2 cm/s, which is in the range of hydraulic conductivity values found for clean sand. The one foundry sand that was an exception had much lower measured hydraulic conductivity values of 2.75 x 10-5 cm/s and 5.76 x 10-5 cm/s. For the batch tests conducted, none of the nitrate was removed by any of the six foundry sands; however, conversely, almost all of the phosphate was removed by each foundry sand. Batch test atrazine removal results were much more varied. Compared with baseline batch tests, one foundry sand removed two thirds of the atrazine, one foundry sand removed about one half of the atrazine, three foundry sands removed about a third of the atrazine, and one foundry sand removed none of the atrazine. Overall, these laboratory results generally indicate that foundry sand bulking agents can be used to maintain the hydraulic efficiency of porous media filters used to treat agricultural drainage waters, and as an added benefit, the foundry sand bulking agents may even be able, by themselves, to remove some nutrient and pesticide contaminants.
Determination of Fracture Flow at the Coles Hill Uranium Deposit in Pittsylvania County, Virginia Using Electrical Resistivity and Cross Borehole Methods
Virginia Tech researchers are currently conducting a comprehensive investigation on the Coles Hill uranium deposit in Pittsylvania County, Virginia. This high-grade hydrothermal deposit is believed to be the largest known accumulation of uranium in the United States. This multi-faceted project includes geophysical, structural, hydrological, mineralogical, geochemical, and petrological studies to characterize the deposit in situ and to develop a genetic model to describe the origin of the deposit. The deposit is located to the west of the Chatham Fault, which separates the Mesozoic metasedimentary basin to the east from the host Leatherwood Granite. The hydrogeology of the Coles Hill area is typical of the Piedmont region, with ground- water flow confined to fractures in the crystalline bedrock underlying a shallow regolith. Eleven electrical resistivity profiles were obtained that identify potential subsurface permeable fractures within the ore body and possibly along and across the Chatham fault. Initial interpretations of the resistivity data show several low resistivity zones that trend west to east toward the Chatham fault along inferred perpendicular fault traces, possibly indicating fracture pathways for ground-water flow. Monitoring wells will be established along these profiles to perform borehole geophysical logging and cross-borehole testing to further describe the fracture properties and related hydraulic characteristics of the site. Understanding the amount and locations of groundwater flow at Coles Hill is critical for developing a water usage and dewatering plan should the deposit be mined.
Comparative Analysis of Kinematic Approximation and Richards Equation Models for Subsurface Flow on Complex Hillslopes
Generalized solutions for the kinematic wave equation for subsurface flow have recently been derived for hillslopes of arbitrary geometry by introducing two dimensionless geometric parameters α and ε which define the hydrologic similarity between hillslopes with respect to their characteristic response (Norbiato and Borga, 2008). These solutions are derived by using a second order polynomial function to describe the bedrock slope and an exponential function to describe the variation of the width of the hillslope with hillslope distance. In this presentation we assess the behavior of this simple, one- dimensional model in comparison with a fully three-dimensional Richards equation model for a series of free drainage scenarios. For different values of saturated hydraulic conductivity, we specify the range of values of the two dimensionless geometric parameters α and ε for which the generalized solution is valid. Special attention is given to the discretization and setup of the boundary and initial conditions.