Search USGS⌕ Search

Geology topics

Mary C. Hill

Publications and source records attributed to Mary C. Hill.

At least 37 records · Page 2Linked to original sources
USGS2010

Death Valley regional groundwater flow system, Nevada and California: Hydrogeologic framework and transient groundwater flow model

A numerical three-dimensional (3D) transient groundwater flow model of the Death Valley region was developed by the U.S. Geological Survey for the U.S. Department of Energy programs at the Nevada Test Site and at Yucca Mountain, Nevada. Decades of study of aspects of the groundwater flow system and previous less extensive groundwater flow models were incorporated and reevaluated together with new data to provide greater detail for the complex, digital model. A 3D digital hydrogeologic framework model (HFM) was developed from digital elevation models, geologic maps, borehole information, geologic and hydrogeologic cross sections, and other 3D models to represent the geometry of the hydrogeologic units (HGUs). Structural features, such as faults and fractures, that affect groundwater flow also were added. The HFM represents Precambrian and Paleozoic crystalline and sedimentary rocks, Mesozoic sedimentary rocks, Mesozoic to Cenozoic intrusive rocks, Cenozoic volcanic tuffs and lavas, and late Cenozoic sedimentary deposits of the Death Valley regional groundwater flow system (DVRFS) region in 27 HGUs. Information from a series of investigations was compiled to conceptualize and quantify hydrologic components of the groundwater flow system within the DVRFS model domain and to provide hydraulic-property and head-observation data used in the calibration of the transient-flow model. These studies reevaluated natural groundwater discharge occurring through evapotranspiration (ET) and spring flow; the history of groundwater pumping from 1913 through 1998; groundwater recharge simulated as net infiltration; model boundary inflows and outflows based on regional hydraulic gradients and water budgets of surrounding areas; hydraulic conductivity and its relation to depth; and water levels appropriate for regional simulation of prepumped and pumped conditions within the DVRFS model domain. Simulation results appropriate for the regional extent and scale of the model were provided by acquiring additional data, by reevaluating existing data using current technology and concepts, and by refining earlier interpretations to reflect the current understanding of the regional groundwater flow system. Groundwater flow in the Death Valley region is composed of several interconnected, complex groundwater flow systems. Groundwater flow occurs in three subregions in relatively shallow and localized flow paths that are superimposed on deeper, regional flow paths. Regional groundwater flow is predominantly through a thick Paleozoic carbonate rock sequence affected by complex geologic structures from regional faulting and fracturing that can enhance or impede flow. Spring flow and ET are the dominant natural groundwater discharge processes. Groundwater also is withdrawn for agricultural, commercial, and domestic uses. Groundwater flow in the DVRFS was simulated using MODFLOW-2000, the U.S. Geological Survey 3D finitedifference modular groundwater flow modeling code that incorporates a nonlinear least-squares regression technique to estimate aquifer parameters. The DVRFS model has 16 layers of defined thickness, a finite-difference grid consisting of 194 rows and 160 columns, and uniform cells 1,500 meters (m) on each side. Prepumping conditions (before 1913) were used as the initial conditions for the transient-state calibration. The model uses annual stress periods with discrete recharge and discharge components. Recharge occurs mostly from infiltration of precipitation and runoff on high mountain ranges and from a small amount of underflow from adjacent basins. Discharge occurs primarily through ET and spring discharge (both simulated as drains) and water withdrawal by pumping and, to a lesser amount, by underflow to adjacent basins simulated by constant-head boundaries. All parameter values estimated by the regression are reasonable and within the range of expected values. The simulated hydraulic heads of the final calibrated transient model generally fit observed heads reasonably well (residuals with absolute values less than 10 meters) with two exceptions: in most areas of nearly flat hydraulic gradient the fit is considered moderate (residuals with absolute values of 10 to 20 meters), and in areas of steep hydraulic gradient along the Eleana Range and western part of Yucca Flat, southern part of the Owlshead Mountains, southern part of the Bullfrog Hills, and the north-northwestern part of the model domain (residuals with absolute values greater than 20 meters). Groundwater discharge residuals are fairly random, with as many areas where simulated flows are less than observed flows as areas where simulated flows are greater. The highest unweighted groundwater discharge residuals occur at Death Valley, Sarcobatus Flat (northeastern area), Tecopa, and early observations at Manse Spring in Pahrump Valley. High weighted-discharge residuals were computed in Indian Springs Valley and parts of Death Valley. Most of these inaccuracies in head and discharge can be attributed to insufficient representation of the hydrogeology in the HFM and(or) discharge estimates, misrepresentation of water levels, and(or) model error associated with grid-cell size. The model represents the large and complex groundwater flow system of the Death Valley region at a greater degree of refinement and accuracy than has been possible previously. The representation of detail provided by the 3D digital hydrogeologic framework model and the numerical groundwater flow model enabled greater spatial accuracy in every model parameter. The lithostratigraphy and structural effects of the hydrogeologic framework; recharge estimates from simulated net infiltration; discharge estimates from ET, spring flow, and pumping; and boundary inflow and outflow estimates all were reevaluated, some additional data were collected, and accuracy was improved. Uncertainty in the results of the flow model simulations can be reduced by improving on the quality, interpretation, and representation of the water-level and discharge observations used to calibrate the model and improving on the representation of the HGU geometries, the spatial variability of HGU material properties, the flow model physical framework, and the hydrologic conditions.

California, Nevada↗
USGS2009

Sensitivity analysis, calibration, and testing of a distributed hydrological model using error‐based weighting and one objective function

We evaluate the utility of three interrelated means of using data to calibrate the fully distributed rainfall‐runoff model TOPKAPI as applied to the Maggia Valley drainage area in Switzerland. The use of error‐based weighting of observation and prior information data, local sensitivity analysis, and single‐objective function nonlinear regression provides quantitative evaluation of sensitivity of the 35 model parameters to the data, identification of data types most important to the calibration, and identification of correlations among parameters that contribute to nonuniqueness. Sensitivity analysis required only 71 model runs, and regression required about 50 model runs. The approach presented appears to be ideal for evaluation of models with long run times or as a preliminary step to more computationally demanding methods. The statistics used include composite scaled sensitivities, parameter correlation coefficients, leverage, Cook's D, and DFBETAS. Tests suggest predictive ability of the calibrated model typical of hydrologic models.

Water Resources Research↗
USGS2009

Effects of simplifying fracture network representation on inert chemical migration in fracture-controlled aquifers

[1] While it is widely recognized that highly permeable ‘large‐scale' fractures dominate chemical migration in many fractured aquifers, recent studies suggest that the pervasive ‘small‐scale’ fracturing once considered of less significance can be equally important for characterizing the spatial extent and residence time associated with transport processes. A detailed examination of chemical migration through fracture‐controlled aquifers is used to advance this conceptual understanding. The influence of fracture structure is evaluated by quantifying the effects to transport caused by a systematic removal of fractures from three‐dimensional discrete fracture models whose attributes are derived from geologic and hydrologic conditions at multiple field sites. Results indicate that the effects to transport caused by network simplification are sensitive to the fracture network characteristics, degree of network simplification, and plume travel distance, but primarily in an indirect sense since correlation to individual attributes is limited. Transport processes can be ‘enhanced’ or ‘restricted’ from network simplification meaning that the elimination of fractures may increase or decrease mass migration, mean travel time, dispersion, and tailing of the concentration plume. The results demonstrate why, for instance, chemical migration may not follow the classic advection‐dispersion equation where dispersion approximates the effect of the ignored geologic structure as a strictly additive process to the mean flow. The analyses further reveal that the prediction error caused by fracture network simplification is reduced by at least 50% using the median estimate from an ensemble of simplified fracture network models, and that the error from network simplification is at least 70% less than the stochastic variability from multiple realizations.

Water Resources Research↗
USGS2008

SIM_ADJUST -- A computer code that adjusts simulated equivalents for observations or predictions

This report documents the SIM_ADJUST computer code. SIM_ADJUST surmounts an obstacle that is sometimes encountered when using universal model analysis computer codes such as UCODE_2005 (Poeter and others, 2005), PEST (Doherty, 2004), and OSTRICH (Matott, 2005; Fredrick and others (2007). These codes often read simulated equivalents from a list in a file produced by a process model such as MODFLOW that represents a system of interest. At times values needed by the universal code are missing or assigned default values because the process model could not produce a useful solution. SIM_ADJUST can be used to (1) read a file that lists expected observation or prediction names and possible alternatives for the simulated values; (2) read a file produced by a process model that contains space or tab delimited columns, including a column of simulated values and a column of related observation or prediction names; (3) identify observations or predictions that have been omitted or assigned a default value by the process model; and (4) produce an adjusted file that contains a column of simulated values and a column of associated observation or prediction names. The user may provide alternatives that are constant values or that are alternative simulated values. The user may also provide a sequence of alternatives. For example, the heads from a series of cells may be specified to ensure that a meaningful value is available to compare with an observation located in a cell that may become dry. SIM_ADJUST is constructed using modules from the JUPITER API, and is intended for use on any computer operating system. SIM_ADJUST consists of algorithms programmed in Fortran90, which efficiently performs numerical calculations.

Ground Water Modeling Investigation Report↗
USGS2007

Effective groundwater model calibration: With analysis of data, sensitivities, predictions, and uncertainty

Methods and guidelines for developing and using mathematical models Turn to Effective Groundwater Model Calibration for a set of methods and guidelines that can help produce more accurate and transparent mathematical models. The models can represent groundwater flow and transport and other natural and engineered systems. Use this book and its extensive exercises to learn methods to fully exploit the data on hand, maximize the model's potential, and troubleshoot any problems that arise. Use the methods to perform: Sensitivity analysis to evaluate the information content of data Data assessment to identify (a) existing measurements that dominate model development and predictions and (b) potential measurements likely to improve the reliability of predictions Calibration to develop models that are consistent with the data in an optimal manner Uncertainty evaluation to quantify and communicate errors in simulated results that are often used to make important societal decisions Most of the methods are based on linear and nonlinear regression theory. Fourteen guidelines show the reader how to use the methods advantageously in practical situations. Exercises focus on a groundwater flow system and management problem, enabling readers to apply all the methods presented in the text. The exercises can be completed using the material provided in the book, or as hands-on computer exercises using instructions and files available on the text's accompanying Web site. Throughout the book, the authors stress the need for valid statistical concepts and easily understood presentation methods required to achieve well-tested, transparent models. Most of the examples and all of the exercises focus on simulating groundwater systems; other examples come from surface-water hydrology and geophysics. The methods and guidelines in the text are broadly applicable and can be used by students, researchers, and engineers to simulate many kinds systems.

Book↗
USGS2007

Model calibration and issues related to validation, sensitivity analysis, post-audit, uncertainty evaluation and assessment of prediction data needs

When simulating natural and engineered groundwater flow and transport systems, one objective is to produce a model that accurately represents important aspects of the true system. However, using direct measurements of system characteristics, such as hydraulic conductivity, to construct a model often produces simulated values that poorly match observations of the system state, such as hydraulic heads, flows and concentrations (for example, Barth et al., 2001 ). This occurs because of inaccuracies in the direct measurements and because the measurements commonly characterize system properties at different scales from that of the model aspect to which they are applied. In these circumstances, the conservation of mass equations represented by flow and transport models can be used to test the applicability of the direct measurements, such as by comparing model simulated values to the system state observations. This comparison leads to calibrating the model, by adjusting the model construction and the system properties as represented by model parameter values, so that the model produces simulated values that reasonably match the observations.

Book chapter↗
USGS2007

MODFLOW-2005, The U.S. Geological Survey Modular Ground-Water Model - Documentation of the Multiple-Refined-Areas Capability of Local Grid Refinement (LGR) and the Boundary Flow and Head (BFH) Package

This report documents the addition of the multiple-refined-areas capability to shared node Local Grid Refinement (LGR) and Boundary Flow and Head (BFH) Package of MODFLOW-2005, the U.S. Geological Survey modular, three-dimensional, finite-difference ground-water flow model. LGR now provides the capability to simulate ground-water flow by using one or more block-shaped, higher resolution local grids (child model) within a coarser grid (parent model). LGR accomplishes this by iteratively coupling separate MODFLOW-2005 models such that heads and fluxes are balanced across the shared interfacing boundaries. The ability to have multiple, nonoverlapping areas of refinement is important in situations where there is more than one area of concern within a regional model. In this circumstance, LGR can be used to simulate these distinct areas with higher resolution grids. LGR can be used in two-and three-dimensional, steady-state and transient simulations and for simulations of confined and unconfined ground-water systems. The BFH Package can be used to simulate these situations by using either the parent or child models independently.

Techniques and Methods↗
USGS2007

OPR-PPR, a computer program for assessing data importance to model predictions using linear statistics

The OPR-PPR program calculates the Observation-Prediction (OPR) and Parameter-Prediction (PPR) statistics that can be used to evaluate the relative importance of various kinds of data to simulated predictions. The data considered fall into three categories: (1) existing observations, (2) potential observations, and (3) potential information about parameters. The first two are addressed by the OPR statistic; the third is addressed by the PPR statistic. The statistics are based on linear theory and measure the leverage of the data, which depends on the location, the type, and possibly the time of the data being considered. For example, in a ground-water system the type of data might be a head measurement at a particular location and time. As a measure of leverage, the statistics do not take into account the value of the measurement. As linear measures, the OPR and PPR statistics require minimal computational effort once sensitivities have been calculated. Sensitivities need to be calculated for only one set of parameter values; commonly these are the values estimated through model calibration. OPR-PPR can calculate the OPR and PPR statistics for any mathematical model that produces the necessary OPR-PPR input files. In this report, OPR-PPR capabilities are presented in the context of using the ground-water model MODFLOW-2000 and the universal inverse program UCODE_2005. The method used to calculate the OPR and PPR statistics is based on the linear equation for prediction standard deviation. Using sensitivities and other information, OPR-PPR calculates (a) the percent increase in the prediction standard deviation that results when one or more existing observations are omitted from the calibration data set; (b) the percent decrease in the prediction standard deviation that results when one or more potential observations are added to the calibration data set; or (c) the percent decrease in the prediction standard deviation that results when potential information on one or more parameters is added.

Techniques and Methods↗
USGS2007

MMA, A Computer Code for Multi-Model Analysis

This report documents the Multi-Model Analysis (MMA) computer code. MMA can be used to evaluate results from alternative models of a single system using the same set of observations for all models. As long as the observations, the observation weighting, and system being represented are the same, the models can differ in nearly any way imaginable. For example, they may include different processes, different simulation software, different temporal definitions (for example, steady-state and transient models could be considered), and so on. The multiple models need to be calibrated by nonlinear regression. Calibration of the individual models needs to be completed before application of MMA. MMA can be used to rank models and calculate posterior model probabilities. These can be used to (1) determine the relative importance of the characteristics embodied in the alternative models, (2) calculate model-averaged parameter estimates and predictions, and (3) quantify the uncertainty of parameter estimates and predictions in a way that integrates the variations represented by the alternative models. There is a lack of consensus on what model analysis methods are best, so MMA provides four default methods. Two are based on Kullback-Leibler information, and use the AIC (Akaike Information Criterion) or AICc (second-order-bias-corrected AIC) model discrimination criteria. The other two default methods are the BIC (Bayesian Information Criterion) and the KIC (Kashyap Information Criterion) model discrimination criteria. Use of the KIC criterion is equivalent to using the maximum-likelihood Bayesian model averaging (MLBMA) method. AIC, AICc, and BIC can be derived from Frequentist or Bayesian arguments. The default methods based on Kullback-Leibler information have a number of theoretical advantages, including that they tend to favor more complicated models as more data become available than do the other methods, which makes sense in many situations. Many applications of MMA will be well served by the default methods provided. To use the default methods, the only required input for MMA is a list of directories where the files for the alternate models are located. Evaluation and development of model-analysis methods are active areas of research. To facilitate exploration and innovation, MMA allows the user broad discretion to define alternatives to the default procedures. For example, MMA allows the user to (a) rank models based on model criteria defined using a wide range of provided and user-defined statistics in addition to the default AIC, AICc, BIC, and KIC criteria, (b) create their own criteria using model measures available from the code, and (c) define how each model criterion is used to calculate related posterior model probabilities. The default model criteria rate models are based on model fit to observations, the number of observations and estimated parameters, and, for KIC, the Fisher information matrix. In addition, MMA allows the analysis to include an evaluation of estimated parameter values. This is accomplished by allowing the user to define unreasonable estimated parameter values or relative estimated parameter values. An example of the latter is that it may be expected that one parameter value will be less than another, as might be the case if two parameters represented the hydraulic conductivity of distinct materials such as fine and coarse sand. Models with parameter values that violate the user-defined conditions are excluded from further consideration by MMA. Ground-water models are used as examples in this report, but MMA can be used to evaluate any set of models for which the required files have been produced. MMA needs to read files from a separate directory for each alternative model considered. The needed files are produced when using the Sensitivity-Analysis or Parameter-Estimation mode of UCODE_2005, or, possibly, the equivalent capability of another program. MMA is constructed using

Techniques and Methods↗
USGS2006

Progress made in groundwater flow and transport modeling

A special session, “Automated Parameter Identification and Sensitivity Analysis in Groundwater Flow and Transport,” was held at the Spring AGU meeting last May to explore recent scientific and technological developments in this expanding research area. The problems and approaches discussed are crucial to any field of Earth science in which models are used to represent physical processes and simulated values are compared with measured quantities. Although methods for automated parameter identification and sensitivity analysis were first developed at about the same time as numerical groundwater models, they are rarely used. Presentations at this special session clearly demonstrated the potential contributions of these methods for more insightful groundwater model development and calibration.

Eos, Transactions, American Geophysical Union↗
USGS2006

Reply to comment by F. Molz et al. on “Investigating the Macrodispersion Experiment (MADE) site in Columbus, Mississippi, using a three‐dimensional inverse flow and transport model”

W e are pleased to have the opportunity to discuss further what we think are important met hods and results presented by Barlebo et al. [2004] . W e are especially pleased to enter into a public discussion with four scientists who we regard with such high esteem concerning an important topic of mutual interest.

Georgia↗
USGS2006

A guide for using the transient ground-water flow model of the Death Valley regional ground-water flow system, Nevada and California

This report is a guide for executing numerical simulations with the transient ground-water flow model of the Death Valley regional ground-water flow system, Nevada and California using the U.S. Geological Survey modular finite-difference ground-water flow model, MODFLOW-2000. Model inputs, including observations of hydraulic head, discharge, and boundary flows, are summarized. Modification of the DVRFS transient ground-water model is discussed for two common uses of the Death Valley regional ground-water flow system model: predictive pumping scenarios that extend beyond the end of the model simulation period (1998), and model simulations with only steady-state conditions.

Open-File Report↗
USGS2006

MODFLOW-2005, the U.S. Geological Survey modular ground-water model - documentation of shared node local grid refinement (LGR) and the boundary flow and head (BFH) package

This report documents the addition of shared node Local Grid Refinement (LGR) to MODFLOW-2005, the U.S. Geological Survey modular, transient, three-dimensional, finite-difference ground-water flow model. LGR provides the capability to simulate ground-water flow using one block-shaped higher-resolution local grid (a child model) within a coarser-grid parent model. LGR accomplishes this by iteratively coupling two separate MODFLOW-2005 models such that heads and fluxes are balanced across the shared interfacing boundary. LGR can be used in two-and three-dimensional, steady-state and transient simulations and for simulations of confined and unconfined ground-water systems. Traditional one-way coupled telescopic mesh refinement (TMR) methods can have large, often undetected, inconsistencies in heads and fluxes across the interface between two model grids. The iteratively coupled shared-node method of LGR provides a more rigorous coupling in which the solution accuracy is controlled by convergence criteria defined by the user. In realistic problems, this can result in substantially more accurate solutions and require an increase in computer processing time. The rigorous coupling enables sensitivity analysis, parameter estimation, and uncertainty analysis that reflects conditions in both model grids. This report describes the method used by LGR, evaluates LGR accuracy and performance for two- and three-dimensional test cases, provides input instructions, and lists selected input and output files for an example problem. It also presents the Boundary Flow and Head (BFH) Package, which allows the child and parent models to be simulated independently using the boundary conditions obtained through the iterative process of LGR.

Techniques and Methods↗
USGS2005

Guidelines 13 and 14—Prediction uncertainty

An advantage of using optimization for model development and calibration is that optimization provides methods for evaluating and quantifying prediction uncertainty. Both deterministic and statistical methods can be used. Guideline 13 discusses using regression and post-audits, which we classify as deterministic methods. Guideline 14 discusses inferential statistics and Monte Carlo methods, which we classify as statistical methods.

Book chapter↗
USGS2005

Calibrating transient and transport models and recalibrating existing models

The methods presented in Chapters 3 to 8 are applicable to models of any system. However, there are special considerations when applying the methods to certain types of models. This chapter discusses three types of models that are of special interest to many scientific and engineering fields: transient models, transport models, and existing models that are to be recalibrated.

Book chapter↗