Important observations and parameters for a salt water intrusion model

Sensitivity analysis with a density-dependent ground water flow simulator can provide insight and understanding of salt water intrusion calibration problems far beyond what is possible through intuitive analysis alone. Five simple experimental simulations presented here demonstrate this point. Results show that dispersivity is a very important parameter for reproducing a steady-state distribution of hydraulic head, salinity, and flow in the transition zone between fresh water and salt water in a coastal aquifer system. When estimating dispersivity, the following conclusions can be drawn about the data types and locations considered. (1) The "toe" of the transition zone is the most effective location for hydraulic head and salinity observations. (2) Areas near the coastline where submarine ground water discharge occurs are the most effective locations for flow observations. (3) Salinity observations are more effective than hydraulic head observations. (4) The importance of flow observations aligned perpendicular to the shoreline varies dramatically depending on distance seaward from the shoreline. Extreme parameter correlation can prohibit unique estimation of permeability parameters such as hydraulic conductivity and flow parameters such as recharge in a density-dependent ground water flow model when using hydraulic head and salinity observations. Adding flow observations perpendicular to the shoreline in areas where ground water is exchanged with the ocean body can reduce the correlation, potentially resulting in unique estimates of these parameter values. Results are expected to be directly applicable to many complex situations, and have implications for model development whether or not formal optimization methods are used in model calibration.     

Calculation of ground water ages--a comparative analysis

Ground water age is a fundamental, yet complex, concept in ground water hydrology. Discrepancies between results obtained through different modeling approaches for ground water age calculation have been reported, in particular, between ground water ages modeled by advection and direct simulation of ground water ages (e.g., age-mass approach), which includes effects of advection and dispersion. Here, through a series of two-dimensional (2D) simulations, the impact of water mixing through advection and dispersion on modeled 14C and directly simulated ground water ages is assessed. Impact of dispersion on modeled ages is systematically stronger in areas where water velocities are smaller and far more pronounced on 14C ages. This effect is also observed in one-dimensional models. 2D simulations show that longitudinal dispersion generally acts as a "source" of 14C, while vertical dispersion acts as a "sink," leading to apparent younger or older modeled 14C ages as compared to advective and directly simulated ground water ages. The presence of permeable and impermeable faults provides an equally important source for discrepancies, leading to major differences in modeled ages among the three methods considered. Overall, our results show that a 14C modeling approach using a solute transport model for calculating ground water age appears to be more reliable in ground water systems without faults and where water velocities are relatively high than in systems that are relatively more heterogeneous and those where faults are present. Among the three modeling approaches considered here, direct simulation of ground water age seems to yield the most consistent results in complex, heterogeneous ground water flow systems, giving a vertical age structure consistent with ages expected from consideration of the flow system.     

A comparison of solute-transport solution techniques and their effect on sensitivity analysis and inverse modeling results

Five common numerical techniques for solving the advection-dispersion equation (finite difference, predictor corrector, total variation diminishing, method of characteristics, and modified method of characteristics) were tested using simulations of a controlled conservative tracer-test experiment through a heterogeneous, two-dimensional sand tank. The experimental facility was constructed using discrete, randomly distributed, homogeneous blocks of five sand types. This experimental model provides an opportunity to compare the solution techniques: the heterogeneous hydraulic-conductivity distribution of known structure can be accurately represented by a numerical model, and detailed measurements can be compared with simulated concentrations and total flow through the tank. The present work uses this opportunity to investigate how three common types of results--simulated breakthrough curves, sensitivity analysis, and calibrated parameter values--change in this heterogeneous situation given the different methods of simulating solute transport. The breakthrough curves show that simulated peak concentrations, even at very fine grid spacings, varied between the techniques because of different amounts of numerical dispersion. Sensitivity-analysis results revealed: (1) a high correlation between hydraulic conductivity and porosity given the concentration and flow observations used, so that both could not be estimated; and (2) that the breakthrough curve data did not provide enough information to estimate individual values of dispersivity for the five sands. This study demonstrates that the choice of assigned dispersivity and the amount of numerical dispersion present in the solution technique influence estimated hydraulic conductivity values to a surprising degree.     

Aquifer Vulnerability Assessment Based on Sequence Stratigraphic and 39Ar Transport Modeling

A large‐scale groundwater flow and transport model is developed for a deep‐seated (100 to 300 m below ground surface) sedimentary aquifer system. The model is based on a three‐dimensional (3D) hydrostratigraphic model, building on a sequence stratigraphic approach. The flow model is calibrated against observations of hydraulic head and stream discharge while the credibility of the transport model is evaluated against measurements of ³⁹Ar from deep wells using alternative parameterizations of dispersivity and effective porosity. The directly simulated 3D mean age distributions and vertical fluxes are used to visualize the two‐dimensional (2D)/3D age and flux distribution along transects and at the top plane of individual aquifers. The simulation results are used to assess the vulnerability of the aquifer system that generally has been assumed to be protected by thick overlaying clayey units and therefore proposed as future reservoirs for drinking water supply. The results indicate that on a regional scale these deep‐seated aquifers are not as protected from modern surface water contamination as expected because significant leakage to the deeper aquifers occurs. The complex distribution of local and intermediate groundwater flow systems controlled by the distribution of the river network as well as the topographical variation (Tóth 1963) provides the possibility for modern water to be found in even the deepest aquifers.     

Groundwater ages in coastal aquifers

The interpretation of groundwater ages in coastal aquifers requires an improved understanding of relationships between age distributions and the processes accompanying dispersive, density-dependent flow and transport. This study uses numerical modeling to examine the influence of mixing and a selection of other hydrogeological factors on steady-state age distributions in coastal aquifers. Three methods of age estimation are compared: the piston flow age, the direct age, and the tracer-based age. These are applied to various forms of the Henry problem, as well as to three variants of a larger, hypothetical coastal aquifer. Circulation of water within the seawater wedge results in markedly higher ages in the transition zone than in the underlying saltwater or overlying freshwater. Piston flow ages show a sharp increase where the freshwater and saltwater systems meet, whereas direct- and tracer-based simulations result in a smoother age distribution, as expected. Greater degrees of mixing result in larger differences between piston flow and direct or tracer-based ages, and bring about lower ages in the saltwater wedge. Tracer-based ages are preferred over direct- and piston flow ages for comparison with field data, especially in cases with wide transition zones. Despite the relatively simple conditions used for the simulations, complex age distributions with depth were obtained. Hence, the assessment of ages in field cases will be difficult, particularly where accurate ages in the transition zone are sought.     

Cyclic hysteretic flow in porous medium column: model, experiment, and simulations

A periodic vertical movement of the groundwater table results in a subsequent cyclic response of the water content and pressure profiles in the vadose zone. The sequence of periodic wetting and drying processes can be affected by hysteresis effects in this zone. A one-dimensional saturated/unsaturated flow model based on Richards’ equation and the Mualem (Soil Sci. 137 (1984) 283) hysteresis model is formulated which can take into account multi-cycle hysteresis effects in the relation between capillary pressure and water content. The numerical integration of the unsaturated flow equation is based on a Galerkin-type finite element method. The flow domain is discretised by finite elements with linear shape functions. Simulations start with static water content and pressure profiles, which correspond to either a boundary drying or wetting retention curve. To facilitate the numerical solution of the hysteretic case an implicit non-iterative procedure was chosen for the solution of the nonlinear differential equation. Laboratory experiments were performed with a vertical sand column by imposing a high frequency periodic pressure head at the lower end of the column. The total water volume in the column, and the periodic water content profile averaged over time were measured. The boundary drying and wetting curves of the relation between water content and capillary pressure were determined by independent experiments. The simulations of the experimental conditions show a clear effect of the hysteresis phenomenon on the water content profile. The simulations with hysteresis agree well with the measurements. Computed dimensionless water content profiles are presented for different oscillation frequencies with and without consideration of hysteresis.     

Spatial characterization of land surface energy fluxes and uncertainty estimation at the Salar de Atacama,Northern Chile

We use Advanced Spaceborne Thermal Emission and Reflection radiometer (ASTER) data to estimate spatial energy flux and evaporation distributions at the Salar de Atacama, a playa in Northern Chile. Our approach incorporates ASTER surface kinetic temperature, emissivity, and reflectance data, ground-based meteorological measurements, and empirical parameters. Energy flux distributions are estimated using either spatially constant or spatially distributed values of model parameters, with spatially distributed parameters assigned separately to each land cover category in an image classification. We test the sensitivity of energy budget calculations to state variable and parameter values by conducting Monte Carlo simulations for regions with ground energy budget measurements. Results show that assigning spatially distributed model parameters via land cover classifications yields significant improvements to ground and sensible heat flux predictions. Latent heat fluxes cannot, however, be predicted with sufficient accuracy to allow estimation of area-integrated evaporative moisture loss at this low-evaporation playa.     

Using high resolution tracer data to constrain water storage,flux and age estimates in a spatially distributed rainfall‐runoff model

Models simulating stream flow and conservative tracers can provide a representation of flow paths, storage distributions and mixing processes that is advantageous for many predictive purposes. Compared with models that only simulate stream flow, tracer data can be used to investigate the internal consistency of model behaviour and to gain insight into model performance. Here, we examine the strengths and weaknesses of a data‐driven, spatially distributed tracer‐aided rainfall‐runoff model. The model structure allowed us to assess the influence of landscape characteristics on the routing and mixing of water and tracers. The model was applied to a site in the Scottish Highlands with a unique tracer data set; ~4 years of daily isotope ratios in stream water and precipitation were available, as well as 2 years of weekly soil and ground water isotopes. The model structure was based on an empirically based, lumped tracer‐aided model previously developed for the catchment. The best model runs were selected from Monte Carlo simulations based on dual calibration criteria using objective functions for both stream isotopes and discharge at the outlet. Model performance for these criteria was reasonable (Nash–Sutcliffe efficiencies for discharge and isotope ratios were ~0.4–0.6). The model could generally reproduce the variable isotope signals in the soils of the steeper hill slopes where storage was low, and damped isotope responses in valley bottom cells with high storage. The model also allowed us to estimate the age distributions of internal stores, water fluxes and stream flow. Average stream water age was ~1.6 years, integrating older groundwater in the valley bottom and dynamic younger soil waters. By tracking water ages and simulating isotopes, the model captured the changes in connectivity driven by distributed storage dynamics. This has substantially improved the representation of spatio‐temporal process dynamics and gives a more robust framework for projecting environmental change impacts. Copyright © 2016 The Authors Hydrological Processes Published by John Wiley & Sons Ltd.     

Estimation of Salt Water Upconing Using a Steady‐State Solution for Partial Completion of a Pumped Well

A new steady‐state analytical solution to the two‐dimensional radial‐flow equation was developed for drawdown (head) conditions in an aquifer with constant transmissivity, no‐flow conditions at the top and bottom, constant head conditions at a known radial distance, and a partially completed pumping well. The solution was evaluated for accuracy by comparison to numerical simulations using MODFLOW. The solution was then used to estimate the rise of the salt water‐fresh water interface (upconing) that occurs under a pumping well, and to calculate the critical pumping rate at which the interface becomes unstable, allowing salt water to enter the pumping well. The analysis of salt water‐fresh water interface rise assumed no significant effect on upconing by recharge; this assumption was tested and supported using results from a new steady‐state analytical solution developed for recharge under two‐dimensional radial‐flow conditions. The upconing analysis results were evaluated for accuracy by comparison to those from numerical simulations using SEAWAT for salt water‐fresh water interface positions under mild pumping conditions. The results from the equation were also compared with those of a published numerical sharp‐interface model applied to a case on Cape Cod, Massachusetts. This comparison indicates that estimating the interface rise and maximum allowable pumping rate using the analytical method will likely be less conservative than the maximum allowable pumping rate and maximum stable interface rise from a numerical sharp‐interface model.     

Analytical power series solutions to the two-dimensional advection–dispersion equation with distance-dependent dispersivities

As is frequently cited, dispersivity increases with solute travel distance in the subsurface. This behaviour has been attributed to the inherent spatial variation of the pore water velocity in geological porous media. Analytically solving the advection–dispersion equation with distance-dependent dispersivity is extremely difficult because the governing equation coefficients are dependent upon the distance variable. This study presents an analytical technique to solve a two-dimensional (2D) advection–dispersion equation with linear distance-dependent longitudinal and transverse dispersivities for describing solute transport in a uniform flow field. The analytical approach is developed by applying the extended power series method coupled with the Laplace and finite Fourier cosine transforms. The developed solution is then compared to the corresponding numerical solution to assess its accuracy and robustness. The results demonstrate that the breakthrough curves at different spatial locations obtained from the power series solution show good agreement with those obtained from the numerical solution. However, owing to the limited numerical operation for large values of the power series functions, the developed analytical solution can only be numerically evaluated when the values of longitudinal dispersivity/distance ratio e_L exceed 0·075. Moreover, breakthrough curves obtained from the distance-dependent solution are compared with those from the constant dispersivity solution to investigate the relationship between the transport parameters. Our numerical experiments demonstrate that a previously derived relationship is invalid for large e_L values. The analytical power series solution derived in this study is efficient and can be a useful tool for future studies in the field of 2D and distance-dependent dispersive transport. Copyright © 2008 John Wiley & Sons, Ltd.     

Flow in unsaturated random porous media,nonlinear numerical analysis and comparison to analytical stochastic models

This work presents a rigorous numerical validation of analytical stochastic models of steady state unsaturated flow in heterogeneous porous media. It also provides a crucial link between stochastic theory based on simplifying assumptions and empirical field and simulation evidence of variably saturated flow in actual or realistic hypothetical heterogeneous porous media. Statistical properties of unsaturated hydraulic conductivity, soil water tension, and soil water flux in heterogeneous soils are investigated through high resolution Monte Carlo simulations of a wide range of steady state flow problems in a quasi-unbounded domain. In agreement with assumptions in analytical stochastic models of unsaturated flow, hydraulic conductivity and soil water tension are found to be lognormally and normally distributed, respectively. In contrast, simulations indicate that in moderate to strong variable conductivity fields, longitudinal flux is highly skewed. Transverse flux distributions are leptokurtic. the moments of the probability distributions obtained from Monte Carlo simulations are compared to modified first-order analytical models. Under moderate to strong heterogeneous soil flux conditions (σ²_y≥1), analytical solutions overestimate variability in soil water tension by up to 40% as soil heterogeneity increases, and underestimate variability of both flux components by up to a factor 5. Theoretically predicted model (cross-)covariance agree well with the numerical sample (cross-)covarianaces. Statistical moments are shown to be consistent with observed physical characteristics of unsaturated flow in heterogeneous soils.©1998 Elsevier Science Limited. All rights reserved     

A hybrid artificial neural network-numerical model for ground water problems

Numerical models constitute the most advanced physical-based methods for modeling complex ground water systems. Spatial and/or temporal variability of aquifer parameters, boundary conditions, and initial conditions (for transient simulations) can be assigned across the numerical model domain. While this constitutes a powerful modeling advantage, it also presents the formidable challenge of overcoming parameter uncertainty, which, to date, has not been satisfactorily resolved, inevitably producing model prediction errors. In previous research, artificial neural networks (ANNs), developed with more accessible field data, have achieved excellent predictive accuracy over discrete stress periods at site-specific field locations in complex ground water systems. In an effort to combine the relative advantages of numerical models and ANNs, a new modeling paradigm is presented. The ANN models generate accurate predictions for a limited number of field locations. Appending them to a numerical model produces an overdetermined system of equations, which can be solved using a variety of mathematical techniques, potentially yielding more accurate numerical predictions. Mathematical theory and a simple two-dimensional example are presented to overview relevant mathematical and modeling issues. Two of the three methods for solving the overdetermined system achieved an overall improvement in numerical model accuracy for various levels of synthetic ANN errors using relatively few constrained head values (i.e., cells), which, while demonstrating promise, requires further research. This hybrid approach is not limited to ANN technology; it can be used with other approaches for improving numerical model predictions, such as regression or support vector machines (SVMs).     

Matching solute breakthrough with deterministic and stochastic aquifer models

Two different deterministic and two alternative stochastic (i.e., geostatistical) approaches to modeling the distribution of hydraulic conductivity (K) in a nonuniform (sigma2ln(K)) = 0.29) glacial sand aquifer were used to explore the influence of conceptual model selection on simulations of three-dimensional tracer movement. The deterministic K models employed included a homogeneous effective K and a perfectly stratified 14 layer model. Stochastic K models were constructed using sequential Gaussian simulation and sequential i ndicator simulation conditioned to available K values estimated from measured grain size distributions. Standard simulation software packages MODFLOW, MT3DMS, and MODPATH were used to model three-dimensional ground water flow and transport in a field tracer test, where a pulse of bromide was injected through an array of three fully screened wells and extracted through a single fully screened well approximately 8 m away. Agreement between observed and simulated transport behavior was assessed through direct comparison of breakthrough curves (BTCs) and selected breakthrough metrics at the extraction well and at 26 individual multilevel sample ports distributed irregularly between the injection and extraction wells. Results indicate that conceptual models incorporating formation variability are better able to capture observed breakthrough behavior. Root mean square (RMS) error of the deterministic models bracketed the ensemble mean RMS error of stochastic models for simulated concentration vs. time series, but not for individual BTC characteristic metrics. The spatial variability models evaluated here may be better suited to simulating breakthrough behavior measured in wells screened over large intervals than at arbitrarily distributed observation points within a nonuniform aquifer domain.     

A practical model for mobile, residual, and entrapped NAPL in water-wet porous media

Flow of nonvolatile nonaqueous phase liquid (NAPL) and aqueous phases that account for mobile, entrapped, and residual NAPL in variably saturated water-wet porous media is modeled and compared against results from detailed laboratory experiments. Residual saturation formation in the vadose zone is a process that is often ignored in multifluid flow simulators, which might cause an overestimation of the volume of NAPL that reaches the ground water. Mobile NAPL is defined as being continuous in the pore space and flows under a pressure gradient or gravitational body force. Entrapped NAPL is defined as being occluded by the aqueous phase, occurring as immobile ganglia surrounded by aqueous phase in the pore space and formed when NAPL is replaced by the aqueous phase. Residual NAPL is defined as immobile, nonwater entrapped NAPL that does not drain from the pore spaces and is conceptualized as being either continuous or discontinuous. Free NAPL comprises mobile and residual NAPL. The numerical model is formulated on mass conservation equations for oil and water, transported via NAPL and aqueous phases through variably saturated porous media. To account for phase transitions, a primary variable switching scheme is implemented for the oil-mass conservation equation over three phase conditions: (1) aqueous or aqueous-gas with dissolved oil, (2) aqueous or aqueous-gas with entrapped NAPL, and (3) aqueous or aqueous gas with free NAPL. Two laboratory-scale column experiments are modeled to verify the numerical model. Comparisons between the numerical simulations and experiments demonstrate the necessity to include the residual NAPL formation process in multifluid flow simulators.     

Closed‐Form Approximations for Two‐Dimensional Groundwater Age Patterns in a Fresh Water Lens

Simple closed‐form approximations are presented for calculating the steady‐state groundwater age distribution in two‐dimensional vertical cross sections of idealized fresh water lenses overlying salt water, for aquifers that are vertically semi‐infinite and of finite thickness. The approximations are developed on the basis of existing one‐dimensional analytical solutions for travel‐time calculation in fresh water lenses and approximate streamline formulations. The two‐dimensional age distributions based on the closed‐form solutions match convincingly with numerical simulations. As expected, notable deviations from the numerical solution are encountered at the groundwater flow divide and when submarine groundwater discharge occurs. Ratios of recharge over hydraulic conductivities are varied to explore how the magnitude of the deviations changes, and it is found that the approximate closed‐form solutions perform well over a range of conditions found in natural systems.     

Use of spatially distributed water table observations to constrain uncertainty in a rainfall–runoff model

The Generalised Likelihood Uncertainty Estimation (GLUE) methodology is used to investigate how distributed water table observations modify simulation and parameter uncertainty for the hydrological model TOPMODEL, applied to the Sæternbekken Minifelt catchment in Norway. Errors in simulating observed flows, continuously-logged borehole water levels and more extensive, spatially distributed water table depths are combined using Bayes' equation within a `likelihood measure' L. It is shown how the distributions of L for the TOPMODEL parameters change as the different types of observed data are considered. These distributions are also used to construct corresponding simulation uncertainty bounds for flows, borehole water levels, and water table depths within the spatially-extensive piezometer network. Qualitatively wide uncertainty bounds for water table simulations are thought to be consistent with the simplified nature of the distributed model.     

Ground water budget analysis and cross-formational leakage in an arid basin

Ground water budget analysis in arid basins is substantially aided by integrated use of numerical models and environmental isotopes. Spatial variability of recharge, storage of water of both modern and pluvial age, and complex three-dimensional flow processes in these basins provide challenges to the development of a good conceptual model. Ground water age dating and mixing analysis with isotopic tracers complement standard hydrogeologic data that are collected and processed as an initial step in the development and calibration of a numerical model. Environmental isotopes can confirm or refute a priori assumptions of ground water flow, such as the general assumption that natural recharge occurs primarily along mountains and mountain fronts. Isotopes also serve as powerful tools during postaudits of numerical models. Ground water models provide a means of developing ground water budgets for entire model domains or for smaller regions within the model domain. These ground water budgets can be used to evaluate the impacts of pumping and estimate the magnitude of capture in the form of induced recharge from streams, as well as quantify storage changes within the system. The coupled analyses of ground water budget analysis and isotope sampling and analysis provide a means to confirm, refute, or modify conceptual models of ground water flow.     

Key features of artificial aquifers for use in modeling contaminant transport

Two large-scale (9.5 m long, 4.7 m wide, 2.6 m deep), three-dimensional artificial aquifers were constructed to investigate the influence of spatial variations in aquifer properties on contaminant transport. One aquifer was uniformly filled with coarse sand media (0.6 to 2.0 mm) and the other was constructed as a heterogeneous aquifer using blocks of fine, medium, and coarse sands. The key features of these artificial aquifers are described. An innovative deaeration tower was constructed to overcome a problem of the aquifers becoming blocked with excess air from the ground water source. A series of tracer injection experiments were conducted to test the homogeneity of the first aquifer that was purposely built as a homogeneous aquifer and to calculate values of aquifer parameters. Experimental data show that the aquifer is slightly heterogeneous, and hydraulic conductivity values are significantly higher down one side of the aquifer compared to the mean value. There was very good agreement in estimated dispersivity values between the plume area ratio methods and the curve fitting of tracer breakthrough curves. Dispersivity estimates from a full areal source injection (12.2 m2) experiment using a 1D analytical model were higher than estimates from a limited source injection (0.2 m2) experiment using a 3D model, possibly because the 1D model does not take account of the heterogeneity of hydraulic conductivity in the aquifer, thus overestimating dispersivity. Transverse and vertical dispersivity values were about five times less than the longitudinal dispersivity. There was slight sorption of Rhodamine WT onto the aquifer media.