EPA
600/2-89-04.0
&EPA
United States
Environmental Protection
Agency
Robert S Kerr Environmental
Research Laboratory
Ada, OK 74820
EPA/600/2-89/040
July 1989
Research and Development
The Establishment of a
Groundwater Research
Data Center for
Validation of Subsurface
Flow and Transport
Models
?v.rf
'
-------
EPA/600/2-89/040
July 1989
THE ESTABLISHMENT OF A GROUNDWATER RESEARCH DATA CENTER
FOR VALIDATION OF SUBSURFACE FLOW AND TRANSPORT MODELS
by
Paul K. M. van der Heijde, Wilbert I. M. Elderhorst,
Rachel A. Miller, and Manjit F. Trehan
International Ground Water Modeling Center
Hoi comb Research Institute
Butler University
Indianapolis, Indiana
CR-813191
Project Officer
Joe R. Williams
Extramural Activities and Assistance Division
Robert S. Kerr Environmental Research Laboratory
Ada, Oklahoma 74820
U.S. ENVIRONMENTAL PROTECTION AGENCY
ROBERT S. KERR ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
ADA, OKLAHOMA 74820
-------
DISCLAIMER
The information in this document has been funded in part by the United
States Environmental Protection Agency under CR-813191 to the Holcomb Research
Institute, Butler University, Indianapolis, Indiana. It has been subjected to
the Agency's peer and administrative review, and it has been approved for
publication as an EPA document. It does not necessarily reflect the views of
the Agency and no official endorsement should be inferred. Mention of trade
names or commercial products does not constitute endorsement or recommendation
for use.
-------
FOREWORD
EPA is charged by Congress to protect the Nation's land, air and water
systems. Under a mandate of national environmental laws focused on air and
water quality, solid waste management and the control of toxic substances,
pesticides, noise and radiation, the Agency strives to formulate and implement
actions which lead to a compatible balance between human activities and the
ability of natural systems to support and nurture life.
The Robert S. Kerr Environmental Research Laboratory is the Agency's
center of expertise for investigation of the soil and subsurface environment.
Personnel at the Laboratory are responsible for management of research
programs to: (a) determine the fate, transport, and transformation rates of
pollutants in the soil, the unsaturated zone and the saturated zones of the
subsurface environment; (b) define the processes to be used in characterizing
the soil and subsurface environment as a receptor of pollutants; (c) develop
techniques for predicting the effect of pollutants on ground water, soil and
indigenous organisms; and (d) define and demonstrate the applicability and
limitations of using natural processes, indigenous to the soil and subsurface
environment, for the protection of this resource.
This report describes the activities performed at the Holcomb Research
Institute to establish the International Ground Water Modeling Center's
(IGWMC) Groundwater Research Data Center. The Data Center provides scientists
with information regarding existing datasets and assists in accessing selected
datasets. This secondary use of data reduces considerably the efforts and
costs of data acquisition for model validation and other purposes. The
datasets available from the Center or described in the Center's referral
database provide a useful basis to test newly developed theories on subsurface
flow and contaminant transport processes. Moreover, such datasets should
prove useful for evaluation of monitoring equipment, monitoring strategies,
sampling techniques of groundwater, and for evaluation of methods and
techniques of parameter estimation. This report describes the design of
centralized institutional arrangements for collecting, conditioning,
documenting, storing, and distributing datasets resulting from field research
on soil and groundwater pollution, and the establishment of a referral service
for those datasets not managed by the clearinghouse.
During the course of the project a workshop was held to evaluate the Data
Center objectives, the role of data in model validation, proposed procedures,
quality assurance, and database design. The two-day workshop involved
presentations, group discussions, and written comment on provided
questionnaires. The workshop results are reported in Appendix A.
Clinton W. Hall
Director
Robert S. Kerr Environmental
Research Laboratory
TM
-------
ABSTRACT
The International Ground Water Modeling Center (IGWMC) has established a Groundwater Research
Data Center which provides information on research datasets resulting from publicly funded field experiments
regarding soil and groundwater pollution and related laboratory bench studies, and which distributes
selected public domain datasets for testing and validation of models for flow and contaminant transport in
the saturated and unsaturated zones of the underground.
To fulfill its advisory role, the Data Center analyzes information and documentation resulting from field
and laboratory experiments and evaluates the appropriate datasets for their suitability in model testing and
validation. (The Center has identified validation as the major secondary use of such data.) To assure
consistency in the analysis and description of these datasets, and to provide an efficient way to search,
retrieve, and report information on these datasets, the Data Center has developed a computerized data
directory, SATURN, programmed independently from any proprietary software.
As secondary users of such soil water and groundwater data are highly interested in information
relevant to the assessment of data quality, a primary concern of the Center is the evaluation and
documentation of the level of quality assurance applied during data acquisition, data handling, and data
storage.
In addition to providing referral services, the Data Center distributes, on an "as-is" basis, selected,
high-quality datasets described in the data directory. The datasets of concern represent different
hydrological, geological, and geographic-climatic settings, pollutant compositions, and degrees of
contamination.
Because the quality of research data often is of great importance to the end-user, the Data Center has
adopted internal QA/QC procedures and related institutional organization tailored to IGWMC's existing,
highly successful QA/QC program in model information and software distribution.
The Center's detailed knowledge of the characteristics of a large number of subsurface flow and
transport models and research datasets allow it to serve in an advisory role for both data collectors and
modelers.
IV
-------
CONTENTS
Disclaimer Notice ii
Foreword iii
Abstract iv
List of Figures viii
List of Tables ix
Acknowledgment x
1. INTRODUCTION 1
Project Approach 4
Groundwater Research Data 5
Secondary Use of Research Data 6
Sharing Research Data 7
2. EXISTING DATASETS 9
Literature Survey 9
Preliminary Database for Information on Research Datasets 9
Evaluation of Preliminary Information on Research Datasets 11
Specific Large-scale Field Research Sites 16
Selection of Datasets for Incorporation into the Data Center 19
3. EXISTING DATA CENTERS 23
Data Centers and Data Clearinghouses 23
Previous Studies 23
Selected Natural Resources and Environmental Data Centers 26
NAWDEX 27
STORET 27
WATSTORE 29
Other Water Data Centers 29
Project Data Management 30
Data Management 30
User Access and Data Retrieval 31
Data Maintenance 31
Quality Assurance 32
Documentation 32
Standardization 34
Geographical Information Systems 34
Linking Data Management Systems 36
Data Center Maintenance 36
Conclusions 37
4. QUALITY ASSURANCE/QUALITY CONTROL FOR
RESEARCH DATASETS 39
Introduction 39
QA/QC 39
The Role of Standards and Guidelines 40
Assessment of Data Quality for Secondary Use 41
Soil Water and Groundwater Chemistry 42
Introduction 42
-------
Monitoring System Design and Sampling Programs 43
Effects of Well Construction on Sampling 44
Sampling Soil Water and Groundwater 45
Field Analyses 47
Sample Handling and Documentation 48
Laboratory Analysis 49
Data Quality Indicators 49
Laboratory QA/QC 50
Reporting Analytical Results 51
Discussion 51
Flow Information and Soil and Aquifer Parameters 52
Introduction 52
Water Level Measurement 53
Direct Measurement of Groundwater Flow 54
Vadose Zone Measurements 54
Geomechanical Measurements of Subsurface Materials 54
Determination of Aquifer Hydraulic Characteristics 55
Tracer Tests 56
Sampling Borehole Cuttings, Soils, and Cores 57
Geophysical Surveys 57
Laboratory Bench Studies 58
Data Transfer and Storage 59
Conclusion 59
5. STRUCTURE OF THE REFERRAL DATABASE AND REFERRAL FACILITY DESIGN 60
Introduction 60
Design Criteria 61
Completeness of Data 64
Balance of Information 64
Efficient Searches 65
Selected Search Strategy Criteria 66
Efficient Storage 68
Efficient Memory Usage 68
Other Design Considerations 68
Database Design 70
Structure of the Database 70
Implementation Details 73
User Interface and Database Management Programs 79
The Menu System 79
6. DETAILED ANALYSIS OF THE BORDEN DATASET 90
Introduction 90
General Project Summary and Objectives 90
Aquifer Description 92
Groundwater Flow 92
Monitoring System 92
Groundwater Quality , 92
Analysis 96
Quality Assurance 96
Movement of tne Plume 96
Retardation of the Plume 97
Sorption of Organic Solutes 97
Dataset Contents 97
Screening Results 97
Discussion 98
vi
-------
7. DATA CENTER PROCEDURES 99
Introduction 99
Database and Dataset Management 99
Information Management 101
Information Acquisition and Processing 101
Identification of Potential Datasets 101
Information Dissemination 102
Dataset Management 102
Selection, Acquisition, and Evaluation Procedures 102
Potential Problems in Acquiring Datasets 103
Dataset Preparation and Distribution 104
Services Offered by the Data Center 105
8. INTERNAL QUALITY ASSURANCE/QUALITY CONTROL 107
Introduction 107
QA Organization 107
QA Tracking 108
Referral Database 108
Data Entry 108
Information Retrieval 110
Dataset Distribution 110
Acquisition 110
Evaluation 110
Distribution 110
QA Filing 118
9. CONCLUSIONS AND RECOMMENDATIONS 119
10. REFERENCES 121
APPENDIXES
A. Workshop Report Summary 135
B. Dataset Survey 158
C. SATURN Entry Form 200
vii
-------
UST OF FIGURES
Figure 1. Structure of the IGWMC referral database systems and user interactions 62
Figure 2. Hierarchical structure of the SATURN database 71
Figure 3. SATURN database program: information levels and contents 72
Figure 4. SATURN database program: use of tables and pointers 74
Figure 5. SATURN database program: binary tree structure 75
Figure 6. SATURN database program: file structure 76
Figure 7. SATURN database program: structure of the user interface 80
Figure 8a. Main (horizontal) menu 81
Figure 8b. (Vertical) add option menu 81
Figure 8c. Change option menu 83
Figure 8d. Inquire (or search) option menu and display of existing dataset 83
Figure 8e. Report sub-menu of inquire option 87
Figure 8f. Display device menu of report sub-menu 87
Figure 8g. Report option of main menu requires site information 88
Figure 8h. Report option menu 89
Figure 8i. Display device menu of report option 89
Figure 9. The Borden landfill site 91
Figure 10. Cross-sectional view of the extent of the chloride plume
at the Borden site 93
Figure 11. Water table maps for the tracer experimental site and vicinity 94
Figure 12. Location of multilevel samplers and injection wells as of January 1986 95
Figure 13. Data Center QA organization 108
viil
-------
LIST OF TABLES
Table 1. General characterization of research studies 12
Table 2. Organizations involved in field research 13
Table 3. Funding of site investigations 13
Table 4. Key pollutants at research sites 14
Table 5. Research sites per state/continent 15
Table 6. Lithology of the research sites 16
Table 7. Preliminary information on field sites and datasets 21
Table 8. ESIS information categories for databases 25
Table 9. ESIS information categories for data systems 26
Table 10. Components of the proposed voluntary standard for abstracts
of machine-readable files (OIRA 1983) 35
Table 11. Selected search strategy criteria 67
Table 12. Descriptors included in summary print option 88
Table 13. SATURN annotation processing form 109
Table 14. SATURN search processing form 111
Table 15. IGWMC dataset tracking form 113
Table 16. IGWMC dataset evaluation form 114
Table 17. IGWMC dataset distribution form 117
be
-------
ACKNOWLEDGMENT
The authors are grateful to Margaret A. Butorac and Karen Ochsenrider for project assistance; to
Ginger Williams and Mary Willis for word processing; to James N. Rogers for manuscript editing and
production; and to Colleen Baker for graphics.
-------
SECTION 1
INTRODUCTION
Widespread concern about the protection and rehabilitation of groundwater resources is being
expressed by both the public and the scientific community. The federal government, through the U.S.
Environmental Protection Agency (EPA) and other agencies, is responding to this concern in several ways
over and above its National Ground Water Protection Strategy (EPA 1984a). Although the Strategy
emphasizes protection, it also stresses mitigation of groundwater contamination, or rehabilitation, at a
moderate number of Superfund sites. For policy development, regulation, and enforcement, and for
researching the physical, chemical, and biological principles underlying these and other areas of EPA
responsibility, the Agency and the scientific community require broad access to the most complete and
reliable data on groundwater systems.
Mathematical formulation of acquired scientific understanding forms the basis of the predictive
simulation capabilities essential to every regulatory program concerned with groundwater contamination.
At the federal level these programs are authorized under the Resource Conservation and Recovery Act
(RCRA), the Comprehensive Environmental Response, Compensation and Liability Act (CERCLA or
"Superfund"), the Underground Injection Control (UIC) regulations of the Safe Drinking Water Act, and the
Clean Water Act (CWA). Other programs focusing on groundwater operate under the authority of the
Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) and the Toxic Substances Control Act (TSCA).
The ability to predict accurately the transport and fate of potential contaminants is critical to the
success of most groundwater regulations (EPA 1985). For protecting the integrity of an aquifer or
engineered facility, monitoring of groundwater quality often is an ineffective alternative to predictive modeling.
Thus, development and adoption of methods for predicting pollutant transport and fate in the saturated and
unsaturated zones of the subsurface are key elements of the EPA's groundwater research strategy
(EPA/NCGWR 1982) and in the research programs of other agencies. The development and accuracy of
such predictive capabilities cannot take place without an equally significant effort in subsurface
characterization.
With the growing availability and use of subsurface flow and transport models, concerns regarding
their validity and accuracy have increased. Model testing, or more specifically model validation, provides
model users, decision makers, policy makers, and legal authorities with information on a model's
performance characteristics—information needed to judge the usefulness of the model results for their
problem assessments. An extensive discussion of model validation principles, definitions, and procedures
can be found for vadose zone models in Hern and Melancon (1986) and for groundwater models in general
in van der Heijde et al (1988).
As is discussed in van der Heijde et al. (1988) model, validation is basically part of the scientific
discovery process. Determining the validity of a model is in fact accepting the validity of a theory relating
quantitatively cause and effect in a observed, natural system. A comprehensive approach to model
-------
validation consists of assessing through examination and measurement the correctness of the model
concepts, mathematical formulations, and the computer code representing these mathematical expressions.
Practically, the objective of model validation is to determine how well a model's theoretical foundation and
computer implementation describe actual system behavior by comparing the results of model calculations
with numerical data independently derived from specially designed experiments or from detailed observations
of a natural system specifically selected for this purpose. In general, the parameters required for the model
calculations are estimates based on direct field observations or derived indirectly through subsequent
parameter analysis. In both cases the natural or "real" system is observed by sampling the real system
inputs (system stresses) and outputs (system responses) resulting in measured input and output values.
As these measured data are samples of the real system, they are prone to sampling and measurement
errors. Thus, model validity established by comparing calculated values with independently measured values
is always subjective.
Ultimately, the "success" of a model validation attempt depends on how accurate and complete the
independent system measurements are (which cannot be objectively determined), how small the difference
is between the results of the model calculations and the measured values, and the validity criteria accepted
by the model evaluators.
To consider a particular computer model valid under site-specific conditions, several tests covering
a wide variety of site conditions must be performed under highly controlled circumstances and using
quality-assured data. A large number of tests might be involved in the comprehensive validation of a model.
The experience gained at the International Ground Water Modeling Center shows that such extensive
validation studies are often lacking (van der Heijde et al. 1988).
This finding is confirmed by international model verification and validation studies such as INTRACOIN
(International Nuclide Transport Code Intercomparison Study, completed in 1984) and HYDROCOIN
(Hydrologic Code Intercomparison Study), organized by the Swedish Nuclear Power Inspectorate
(INTRACOIN 1986, HYDROCOIN 1987, Nicholson et al. 1987).
Level 2 of the three-level HYDROCOIN studies was aimed at the validation of mathematical models
describing the physical processes involved in groundwater hydrology by comparison of calculations with
observations and experimental measurements for five distinct cases. These cases covered heat transfer
involving thermal convection and conduction, variable density fluid flow based on a thermal convection
experiment, groundwater flow in fractured gneiss, three-dimensional regional flow in low permeable rock,
and soil water redistribution near the ground surface (Hydrocoin 1987). It was noted that comprehensive
databases sufficient for validation of complex groundwater flow models were not available and that there
is a need for experiments specifically designed and planned for validation purposes. The main problems
with the available experiments identified by HYDROCOIN were due either to incomplete parameter definition
(both spatial and temporal) or the lack of independent datasets useful for both model calibration and for
evaluation of the predictions (Nicholson et al 1987).
In recent years an increasing number of datasets derived from specially designed field experiments
have become available or are being collected as part of current research on field subsurface systems and
-------
and soil water and groundwater characteristics. Although such data are a necessary and valuable resource
for verifying theoretical concepts and for validation of models and modeling approaches, to be used
optimally they require complete documentation of collection and analytical procedures, site characteristics,
and an assessment of possible sources of error (van der Heijde et al. 1988).
In a report, the Groundwater Review Committee of EPA's Science Advisory Board concluded that
regardless of the type of model chosen, increased emphasis should be given to field testing and field
validation of each model (EPA 1985a). Data generated in association with remedial action and monitoring
Superfund sites may be used to fulfill model validation requirements. The Review Committee commented
that these data should be made available for use by other investigators. The need for extant data in the
evaluation of tracer-analysis research can be viewed in the same way. (Tracer technology remains one of
the principal approaches to obtaining field values for model parameters [Molz et al. 1985]). The Review
Committee also found that, surprisingly, the conclusions of many publicly funded research efforts are based
on data not available for peer review. Therefore, the Committee recommended that databases from field
research projects be made readily available to other groups. As costs of research and environmental
monitoring escalate, the spiraling cost of acquiring new data emphasizes the critical need for mechanisms
that facilitate access to reliable existing data.
Groundwater research programs are not restricted to the EPA. Major research on groundwater
quality issues is carried out under the auspices of the National Science Foundation and within the Water
Resources Division of the U.S. Geological Survey. Other federal programs containing significant
groundwater quality research are the Subsurface Transport Program of the U.S. Department of Energy (DOE
1985), and the Agricultural Research Program (ARS) of the U.S. Department of Agriculture.
Nongovernmental research in this area is exemplified by the Solid Waste Environmertal Studies (SWES)
program of the Electric Power Research Institute (EPRI), Palo Alto, California.
No institution currently exists for rapidly locating and searching soil water and groundwater research
databases or for standardizing data integrity and documentation of research datasets. Existing centralized
database facilities for groundwater resource management do not provide the detail and quality of data
required to successfully complete research on contaminant transport and fate. In many research projects,
the lack of rapid access to these data causes delays and money unnecessarily spent, resulting in many
incomplete model validation initiatives. The groundwater research strategy prepared by U.S. EPA and the
National Center for Ground Water Research (EPA/NCGWR 1982) states that the data accumulated through
Agency-funded research will be made available to the Agency and to the user community through
information transfer. A central data clearinghouse could acquire and distribute such data in error-free,
machine-usable form, efficiently and economically.
A report published by the National Academy Press (Fienberg 1985) drew attention to the benefits,
costs, and restrictions involved in sharing research data. As the introduction to this publication states, "An
open exchange of scientific information encourages others to engage in re-analysis of data that may detect
flaws in the original research or may lead to better conclusions as a result of improved methods of analysis."
The report continues: "the free flow of raw data can stimulate further research, particularly across disciplines,
improve methods of data collection and teaching, and foster greater scientific understanding and progress."
-------
The need for better documentation and information on and access to existing groundwater research
data can be met by the establishment of a centralized data facility.
Such a central data center or clearinghouse can provide valuable information on completed or
ongoing research projects which have a major data acquisition component. Furthermore, such a facility
could acquire and distribute research data in error-free, machine-usable form, efficiently and economically.
In addressing this need, the Holcomb Research Institute of Butler University, with support from the
U.S. Environmental Protection Agency, has established the Ground Water Research Data Center within the
framework of the International Ground Water Modeling Center (IGWMC). The new Data Center provides
information and referral services regarding datasets resulting from publicly funded field and laboratory
research on soil and groundwater pollution. In addition, the Data Center has established procedures for
selecting, evaluating, documenting, and redistributing such datasets. Creation of the Data Center is
expected to lead to additional protocols for error checking, documentation, accessing, and transferring this
kind of research data, and for acknowledging the rights that researchers have vested in their data.
PROJECT APPROACH
The project consisted of two phases: (1) determination of the scope and design of the Data Center,
and (2) development of facilities and implementation of operational procedures and organizational
framework.
The first phase consisted of five elements: analysis of data needs and potential users; survey and
analysis of existing datasets; assessment of quality assurance (QA) requirements; determination of computer
and other facilities for an operational data center; and operational design of the Data Center.
The analysis of soil and groundwater research data needs and the identification of potential users of
high-quality, well-documented datasets provided guidance, justification, and motivation for the development
of the Data Center. To determine the required level-of-effort and to obtain baseline information for the design
of the Data Center facilities, the availability and status of a number of groundwater datasets resulting from
publicly funded research have been evaluated. Current practices in collecting, handling, storing,
documenting and distributing these datasets have been studied.
Other data centers utilizing high-quality environmental research and monitoring, datasets have been
contacted to benefit from their experience in such areas as dataset acquisition, data handling, and quality
assurance procedures. Specifically, issues related to the invested rights of researchers involved in the data
collection have been discussed.
Quality assurance (QA), an essential task for a central data distribution facility, must be incorporated
on two levels: (1) the quality of the datasets of interest needs to be determined and documented; and (2)
-------
adequate quality assurance procedures need to be established for the operation of the Data Center in such
areas as dataset evaluation, referral, management, and transfer.
To determine the level of detail required for the Data Center in the evaluation of the quality of
prospective datasets, an inventory have been made of standards (existing and under development) and
current accepted practices as documented in the open literature and technical guidance of regulatory
agencies.
Based on the findings in phase 1, the institutional structure for the Data Center has been determined
and the database framework created. Two types of database have been developed: (1) a directory-type or
referral database containing descriptive information on datasets available from the Data Center or from other
sources; and (2) a database containing the datasets selected for distribution by the Data Center. Information
resulting from the dataset survey in phase 1 has been incorporated in the referral database.
Arrangements have been made to protect dataset integrity in their transfer from their generators to
the Data Center and from the Data Center to secondary users. Furthermore, quality assurance procedures
have been implemented for data handling, storing, archiving, and backup. Different levels of implementation
are distinguished, dependent on the quality and extent of the datasets, the level of documentation, and the
importance of the data. Technical support for format and transfer medium, and to a limited extent for the
analysis of the data, will be provided; the level of support will depend on the implementation level selected.
Policies have been developed regarding such issues as proprietary rights, conditional use, potential liabilities,
and other legal and ethical issues.
As a part of the International Ground Water Modeling Center, the Data Center's activities will be
subject to annual review by the IGWMC Policy Board and the International Technical Advisory Committee
(ITAC).
GROUNDWATER RESEARCH DATA
Data on groundwater quality and quantity are characterized in both the spatial and temporal domains.
Two major types of data are distinguished: site-specific data, and generic, site-independent data. It should
be noted that in this report the term groundwater is used for water in both the saturated and unsaturated
zones of the aquatic subsurface.
Certain kinds of site-specific data are constant for the time period under consideration, but may vary
from location to location. Other site-specific data might show a significant time-dependent behavior.
Collection of such data is generally aimed at identifying regional patterns during a certain time period or at
studying the time variability at specific locations (Steele 1985). These objectives of site-specific data
collection may change during the operation of the data collection network, due to changes in management
needs, technology, and institutional arrangements. Subsequently, the design and operation (when and
where to sample or measure, and which variable to measure) may be altered. Such variability certainly
applies to research data networks, which are often project-oriented and of relatively short duration. Data
-------
management, thus confronted with a widely varying set of data characteristics, must be flexible enough to
handle them efficiently.
Because water in the underground often moves quite slowly, abiotic or biotic transformations may
represent significant attenuation processes in the transport and fate of pollutants. The presence of such
processes results in a significant increase in data requirements for the predictive analysis of water quality
(as opposed to the data requirements for water quantity problems). Much of this additional data is generic
and can be established off-site in controlled laboratory or field experiments in combination with relevant site
characteristics. Such generic, site-independent data on specific chemicals (the second type of data
mentioned above) are increasingly available from research on the basic processes that govern contaminant
transport and fate, and are crucial for successful application of computer-based prediction techniques in
specific hydrogeologic environments.
At the beginning of many research projects requiring data acquisition, the establishment of efficient
data management practices is often more difficult than anticipated. Traditionally, researchers have had
almost total control over the form and documentation of their data; even contractual requirements for data
in machine-processible form have had little effect on the ultimate availability and utility of most data. In
addition, control by funding agencies over procedures and quality of data collection, storing, and distribution
to a large number of institutions, requires extensive organizational arrangements and additional personnel.
This is especially true when securing the collected data for distribution after the initial research has been
completed and the original research staff is no longer available, or when no funding is available for
continuing data management at each individual site.
Datasets for use in transport and fate modeling studies require a high level of detail concerning soil
and aquifer properties, density of data points, contaminant behavior, and qualitative data descriptors.
Specific data requirements for subsurface models include the need to define precisely the units of measure
of each input value; for example, point versus averaged values (Hern and Melancon 1986).
Whenever a model study is performed, the quality of the data used is an issue of concern. Data
quality is often critical in model validation due to the sensitivity of most models to changes in certain
parameters. Although a given field investigations may result in a large amount of data, the usefulness of
the study site for model validation is determined to a large extent by the quality of the data, as reported in
the data documentation. However, often the data documentation is lacking in detail, especially with respect
to data quality.
SECONDARY USE OF RESEARCH DATA
Current scientific research confirms that many researchers take advantage of existing high quality
field studies to test their models and support their research findings. For example, research data sets from
the Borden Air Force Base Landfill site in Canada have been used for evaluating monitoring equipment,
developing sampling strategies, researching chemical and biological subsurface activity, verifying theoretical
-------
concepts and models, evaluating parameter estimation methods, and testing remedial action technologies
(Cherry 1983).
The need for soil and groundwater model validation has been stressed by governmental agencies,
consultants, and individual researchers. The Science Advisory Board, in its review of the EPA Groundwater
Research Program (EPA I985a), stated that "it is important for EPA to screen computer models and test
them for accuracy... increased emphasis should be given to field testing and field validation of the models."
Many researchers have found that available data are not sufficient to accurately formulate mathematical
statements of real-world physical problems (Ross et al. 1982), while Moran and Mezgar (1982) identified the
adequacy of datasets as the single most important factor in site-performance model verification and
validation. Yet in many cases the modeler must fill in data gaps with estimated, interpolated, and
extrapolated values. Moran and Mezgar (1982) concluded that emphasis needs to be placed on the
development of a long-term database, one specifically designed for model validation.
Chu et al. (1987) were unable to use field data in their evaluation of data requirements for groundwater
contaminant transport modeling due to the complexity of natural aquifers and the absence of extensive field
data representing the true conditions of the aquifer. Such concerns were also expressed by the EPA
Science Advisory Board, which recommend that "the needed increase of research on the basic processes
that govern the transport and fate of contaminants in groundwater should go together with the establishment
of data bases for field application" (EPA I985a).
Although some extensive datasets are available for validation purposes, their present form may cause
additional problems. Gelhar (1986), in a review of available field datasets, found that very few sites exist
where data have been collected in sufficient detail to be useful for the verification of a new stochastic theory
regarding dispersion in aquifers. The site selected, the Borden Air Force Base in Canada, could provide data
sufficient only to evaluate part of the proposed theoretical framework. A significant part of this particular
study concerned the analysis of the contents and structure of the selected dataset, determining which of the
data to use, and reformatting the selected data in preparation of the verification computations.
A recent EPA groundwater protection data-requirements study (EPA I987a) stressed the importance
of improved access to existing soil water and groundwater data and of lowering the transaction costs
associated with obtaining and using such data. The report indicates that knowledge about and access to
the large volume of groundwater data being generated from federal programs and state initiatives is limited,
because the data are managed by many organizations and are stored in many different locations, files, and
formats. In addition, relatively few of these soil water and groundwater datasets are computerized, and a
central cataloguing facility is lacking. Although the study's conclusions are concerned with all groundwater
data useful in the protection of groundwater resources, they apply equally well to research data.
SHARING RESEARCH DATA
Availability and accessibility of environmental research data are discussed in a wide variety of
environmental literature (Armentano and Loucks 1979, EPA 1985a, Olson and Millemann 1985). Reviews of
-------
data availability indicate that many researchers give little thought to the use of their data other than for
immediate research purposes (Armentano and Loucks 1979). The appraisal by researchers of the
importance of data accessibility is reflected in their approach to data management. Many consider it an
administrative chore to be handled separately from the research, usually at the end of the study (Armentano
and Loucks 1979). Other investigators show a keen awareness of the importance of data management both
for their own use and the use of others.
Sharing data from detailed groundwater monitoring studies and laboratory bench studies is a subject
of concern both economically and with respect to the advancement of scientific research. Due to the
ever-increasing cost of field studies and the extensive sampling periods required for transport and fate
studies, it has become essential to share groundwater data so that unnecessary duplication can be avoided.
Sharing data not only produces cost benefits; it "reinforces open, scientific inquiry; permits verification,
refutation, or refinement of original research results; stimulates improvements in measurement and data
collection methods; allows more efficient use of resources spent on data collection; encourages
interdisciplinary use of data; and strongly discourages the uncommon, but nevertheless serious, problem
of fraudulent research" (NRC 1985).
As participants at a CODATA Workshop pointed out (Hopke and Massart 1986), "One of the most
critical impediments to using existing data sets is the lack of knowing that they are there." Even when
sources of data are known, data access is often complicated, delays occur in obtaining available data, and
documentation is rarely available or is very limited (Armentano and Loucks 1979, Hopke and Massart 1986).
A comprehensive referral center as represented by the IGWMC Groundwater Research Data Center,
focusing on selected datasets for groundwater model validation and testing, will help to avoid situations
where datasets of value to many potential users go unrecognized and therefore unused.
8
-------
SECTION 2
EXISTING DATASETS
Many groundwater research projects result in field and laboratory data. These studies are conducted
by a variety of organizations and for many different reasons. In addition, groundwater management studies
(e.g., site characterization in preparation of remedial action) often produce large sets of data. Much of the
research data is discussed and referred to in the open scientific literature. However, most of the time only
a part of the original data is presented, often in interpreted form, and field notes, laboratory work sheets, and
quality control information are absent.
This section presents the results of a preliminary survey of existing groundwater research datasets
performed to provide the IGWMC Groundwater Research Data Center with basic information for the
development of a referral database and of procedures for information acquisition and distribution. It
discusses objectives, type, and extent of the studies from which the data resulted, provides a summary site
characterization, lists the kind of data collected, discusses form of and accessibility to the data, its
documentation level, the history of primary and secondary use of the datasets, and lists the organizations
that funded and collected the data.
LITERATURE SURVEY
In planning the Groundwater Research Data Center, a literature survey was conducted to identify
relevant research projects and resulting datasets. The search, which was initiated using IGWMC's
computerized literature databases JUPITER (hydrology, geology, water quality, and water management) and
MOON (groundwater models), augmented with information from other sources, resulted in bibliographic
information on sites possibly useful for further analysis.
Data obtained from this literature survey were entered into a database using the REFLEX
(Borland/Analytica Inc.) database system. Selection of sites for the REFLEX database was based on their
potential for use in the validation of transport and fate models.
PRELIMINARY DATABASE FOR INFORMATION ON DATASETS
As the project developed and as the level of detail required to describe the research datasets
increased, a modified database system was developed using R:BASE V (Microrim, Inc.) database software.
-------
This database contains four files:
(1) general site/dataset information
site name
project description
location
type of contaminants or tracers
source(s) of funding
research organizations involved
project classification
number of model applications
(2) bibliographic reference information
author(s)
year published
title
source/publisher
(3) geographical information
state
county
city
(4) aquifer/soil information
material
geologic name/soil type
geologic time period
confined/unconfined
A project classification (see file 1 above) was set up to identify "prime" datasets of interest for model
testing and validation. The RESEARCH1 classification includes those site studies primarily initiated for
research purposes and which have a high degree of quality control and documentation. The RESEARCH2
classification includes those studies that may have originated as groundwater contamination investigations
but which evolved into data collection to study researchable questions. Studies initiated for research are
also included in this class if they were not as detailed as sites in the RESEARCH 1 class or if not enough was
known about them to justify their classification as RESEARCH1 sites. RESEARCH3 projects originated as
groundwater contamination investigations and their data have been used in groundwater modeling studies
and for research studies, without significant additional data gathering. The RESEARCH4 class denotes those
studies implemented as site evaluations for waste disposal or for mining operations, generally lacking the
level of detail and/or quality control needed for research. The SUPERFUND class includes sites listed on
the National Priorities List for which significant characterization have taken place.
10
-------
EVALUATION OF PRELIMINARY INFORMATION ON RESEARCH DATASETS
As of December 31, 1988, 159 sites or studies have been identified for their potential for model
validation and entered in the R:BASE database. Appendix B contains a listing of the sites included in this
database. These studies have not been screened for the level of detail present in the datasets or for their
general adequacy for validation of groundwater flow and transport models. The characteristics of these 159
studies are summarized in Tables 1-6. Many of these research sites have certain features in common, such
as relatively simple geohydrology, location on federal land (usually military installations), or a known history
of contamination. These tables are discussed briefly in the following section.
In general, the common purpose of these data collection projects is to better understand physical,
chemical, and biological subsurface processes, thus providing a basis for efficient protection of the
subsurface environment and for effective remediation of contaminated sites. In addition, the studies
identified have distinct objectives related to the framework in which they have been conducted. Similarly,
many of the studies at Superfund sites have as their objectives conceptualization of the hydrogeologic
system, analysis of the type and extent of the contamination, and evaluation of alternative remedial actions.
The purpose of field studies for the siting of municipal and industrial waste disposal facilities is to provide
the data for planning and design. However, many research studies are characterized by highly specific goals
such as the evaluation of sorption characteristics for a particular combination of aquifer material and
chemical composition, or the identification of spatial variability of hydrodynamic parameters.
Table 1 gives keyword characterizations of the various research sites analyzed. It contains information
on the hydrogeologic conditions, kind of pollutants, source of contaminants, processes studied, and special
interests.
11
-------
TABLE 1. GENERAL CHARACTERIZATION OF RESEARCH STUDIES
General characterization Number of Sites
fractured rock 9
unsaturated zone 8
soil 11
groundwater pollution 20
radioactive waste 17
fly ash 11
hydrocarbons 10
landfill/waste disposal 14
tracer test 11
flow/advection 5
biodegradation 5
geochemistry 8
physical processes 5
monitoring 4
drilling 3
geotechnics 2
(strip) mining 5
miscellaneous 6
not described 5
Tables 2-6 provide additional detail regarding the main site characteristics.
Table 2 list the organizations involved. It clearly shows that most of the activities at the various sites
are performed either by universities or research organizations. The U.S. EPA and the USGS are actively
involved in many of these projects. Often, EPA joins with others or functions as coordinator of multi-agency
projects. The USGS performs many of its own field research studies. Consultants are somewhat less
involved in such projects, which might indicate that most of the research is performed by universities and
research organizations.
12
-------
TABLE 2. ORGANIZATIONS INVOLVED IN FIELD RESEARCH
Projects executed Projects executed
Organization by one In cooperation with
organization other organizations
EPA
DOE
EPRI
USGS
Research Institutes
Industry
Consultants
Universities
States
Miscellaneous
Unknown
4
3
2
17
22
6
16
25
1
0
5
18
2
2
15
21
11
14
26
9
2
Most research projects in the survey are sponsored by the EPA, while DOE and USGS also fund a
considerable number of research projects. Fewer projects surveyed are sponsored by industries, although
industries are often involved in pollution problems.
TABLE 3. FUNDING OF SITE INVESTIGATIONS
EPA
DOE
USGS
EPRI
States
Industry
Universities
Miscellaneous
Unknown
Completely funded
Projects
34
17
15
3
3
2
0
11
47
Partially funded
Projects
15
4
3
8
4
6
4
4
13
-------
Of the research sites focused on the study of pollutant behavior, most deal with existing radioactive
waste or with potential of future storage of such waste. When present In the subsurface, these relate to
nuclear power plants or nuclear material production facilities for national defense. Other sites identified
contain fly-ash, oils/hydrocarbons, halogenated organics, miscellaneous organics, pesticides, and
insecticides.
TABLE 4. KEY POLLUTANTS AT RESEARCH SITES
Pollutant type Number of
sites
Inorganics
(heavy) metals 6
fly ash 12
conservative tracers 3
radioactive wastes 28
miscellaneous inorganics 11
Organics
halogenated organics 9
oils/hydrocarbons 15
pesticides 5
solvents 3
aldicarb (insecticide) 8
miscellaneous organics 19
Organics and Inorganics
landfill leachate 6
sewage 2
chemical waste 4
VOC 2
miscellaneous 5
Does not apply 21
14
-------
TABLE 5. RESEARCH SITES PER STATE/CONTINENT
State/ Number
continent of
sites
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
District of Columbia
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
1
0
3
0
11
5
0
1
0
5
3
0
3
5
3
1
0
1
0
0
0
2
1
4
2
1
1
2
4
1
State/
continent
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
Canada
Africa
Asia
Australia
Europe
South America
Number
of
sites
1
3
4
4
1
10
2
0
2
2
1
3
0
4
4
1
0
0
6
0
4
3
10
0
4
0
8
0
Laboratory studies: 20
15
-------
The number of sites per state varies considerably. The state with the most sites identified is California
(11). The large number of sites in North Dakota results from extensive pollution from mine wastes.
Most of the sites are characterized by porous media. However, 13 sites are found in regions with
fractured rock formations. At another 19 sites, research has been performed in the unsaturated zone.
TABLE 6. UTHOLOGY OF THE RESEARCH SITES
Lithology Number of sites
sand and gravel 6
sand 6
sand, gravel, silt, and clay 31
shale, limestone, sandstone 7
granite 4
total fractured materials 13
total soils and/or unsaturated zone 19
miscellaneous 10
does not apply/unknown 81
Some typical fractured rock studies identified include one at the Fanay-Augeres uranium mine in
France, the Creux De Chippis tracer study in the unsaturated zone at a complex site in Switzerland, and the
Underground Research Laboratory (URL) facility of Atomic Energy of Canada Limited, situated in granite rock
near Lac du Bonnet, Manitoba.
Many of the studies identified are soil column studies performed in the laboratory, using undisturbed
or repacked soil samples taken from specific field sites. These studies were performed for a variety of
reasons, such as determining the chemical attenuation in soil of inorganic and organic contaminants,
evaluating the effects of cracking on flow and contaminant transport, studying dispersion in unsaturated
soils, determining nitrate migration, evaluating the effect of stratified soils on flow and transport of
contaminants, and determining the hydraulic parameters in the unsaturated zone.
SPECIFIC LARGE-SCALE FIELD RESEARCH SITES
An increasing number of field research projects take place in a multidisciplinary and multitask
framework. At some of these sites different studies may have taken place, conducted by various teams of
16
-------
the same research organization or even independently by various research institutions. These research
facilities are sometimes called field experimental stations or sites. An excellent overview of experimental
field studies can be found in Anderson (1987).
The Lawrence Livermore National Laboratory is the site of various studies of groundwater
contamination with VOCs (volatile organic compounds). The site contains alluvial sediments of highly
variable permeability. The Auburn University field study site at Mobile, Alabama, is the location of several
single and double-well tracer tests in stratified aquifers. These studies, supported by the U.S. EPA, involved
methods to determine vertical velocity distributions.
EPRI's Solid Waste Environmental Studies (SWES) program is also a potential source for research
datasets. The program was set up to study the release of solutes from utility waste disposal sites and their
subsequent transformation and transport in the subsurface environment. The Macro Dispersion Experiment
(MADE) is a tracer study with seven different tracers. The study site is located at the Columbus Air Force
Base, Mississippi, and is in a sand, clay, and gravel aquifer.
Currently, a field study underway at Moffett Air Force Base, California, focuses on in situ remediation
of an aquifer system polluted with halogenated aliphatic compounds. The study, conducted by a Stanford
University team, includes testing of biostimulation techniques and Bromide and TCE transport experiments.
The geohydrology of this site consists of interlayered coarse and fine sediments in a shallow, confined
aquifer, an alluvial aquifer, and a buried stream channel.
The database contained 23 sites concerned with radioactive or nuclear waste contamination, mostly
focused on the effects of subsurface nuclear waste storage (both high- and low-level nuclear waste) such
as deep percolation of radionuclides and the geochemistry of the waste and the geologic environment.
Research has been conducted for the selection of possible nuclear waste repositories. The preliminary
database for this study contains six references to studies performed for the Palo Duro Basin, Texas, one
for Richton Dome site, and thirteen for Yucca Mountain, Nevada. The geohydrology at these sites generally
involves thick, unsaturated geologic sequences and deep brine aquifers. Another site extensively studied
for its potential as a nuclear waste storage facility is at Hanford, Washington. The focus of the studies is an
aquifer averaging 70 m thick and composed of glaciofluvial sand and gravel, and silts. The basis of this
aquifer system is formed by basalt.
The Idaho National Engineering Laboratory (INEL) is conducting studies on the presence and behavior
of low-level radioactive wastes in fractured basalt aquifers. Other studies at INEL include investigations of
on-site hazardous waste disposal.
Creosote wastes are a major concern of the regulatory agencies. The U.S. EPA supports studies at
a site in Conroe, Texas, involving delineation of resulting groundwater contamination, presence of microbial
degradation, and the importance of adsorption. Pensacola, Florida, is the site of many studies of creosote
contamination, sponsored by the USGS. The site is characterized by 80 m of surficial deposits consisting
of nonhomogeneous sand and gravel deposited as fluvial and deltaic sediments. Studies here include
determining the interactions of native clay minerals with organic contaminants, the presence of dissolved
17
-------
inorganic species, improving understanding of contaminant distributions in aquifers, and determining the
rates of anaerobic degradation of phenolic compounds. Ongoing research includes work on a wide range
of hydrologic, geochemical, and microbial processes affecting the distribution, movement, and fate of
organic contaminants in aquifers.
The majority of transport models assume that soil water and groundwater move through a relatively
simple porous media. Most research into subsurface processes has been conducted at sites consisting of
a relatively simple hydrogeology and soils, such as homogeneous unconsolidated sand aquifers, which are
atypical for most contaminated aquifers. Current research indicates that even these "homogeneous" aquifers
have complexities such as macroscopic layering, which must be considered for the simulation of transport
and fate of contaminants. Such considerations also apply to soils, where the existence of macropores is
a significant phenomenon. Hence, a need exists to share data collected from sites of greater complexity
such as for consolidated fractured rock aquifers or a complex sequence of layers.
Some hazardous waste sites are especially suitable for the study of particular contamination problems.
For example, the Federal Pioneer site is located in unconsolidated materials and is the site of a PCB spill;
the contamination is multiphase. Research projects here were supported by the government of
Saskatchewan and the National Research Council of Canada. Much data was collected in the unsaturated
zone.
The Borden landfill site, located on the Canadian Forces Base at Borden, Ontario, was actively used
from 1940 to 1976. The aquifer is unconfined and composed of sand of glaciofluvial origin. The contaminant
plume extends for approximately 750 m longitudinally and to the lower confining layer vertically. The landfill
has been monitored extensively and the data have been used in several referenced modeling studies. The
Borden tracer experiment is probably the first attempt at defining the three-dimensional aspects of a
contaminant event using conservative tracers as well as more complex organic compounds that exhibit
sorption effects. Other studies at the site include a microcosm simulating the behavior of VOCs.
The Cape Cod, Massachusetts, area has many contamination sites, due to its high potential for
groundwater contamination caused by the high permeability of the unconfined sand aquifer, and to its dense
population. A field site at the Otis Air Force Base at Cape Cod, Massachusetts, was chosen by the USGS
Toxic-Waste Ground-Water Contamination Program for studying contaminant transport and attenuation in
aquifers. It was selected because of its 45-year documented history, its relatively simple hydrogeologic
setting, and its similarity to many contaminant sites nationwide. The aquifer consists of sand, gravel, silt,
and clay deposited as a glacial outwash plain. A sewage leachate plume, 20 to 25 m thick, 750 to 1,100
m wide, and 3,500 m long, has received a great deal of attention. The contaminants have been transported
with groundwater while being altered by chemical reactions. A complication is the definition of the source
behavior, as contaminants have entered the aquifer at different rates at different times. Studies are being
conducted of sorption by phosphorus, transport of bacteria, rates of microbial activity, volatile organic
compound mobility, and degradation of organics, among others. A natural-gradient tracer experiment was
initiated to measure dispersion and to determine geochemical controls on nonconservative transport in a
heterogeneous aquifer.
18
-------
Another USGS field study is being conducted at Bemidji, Minnesota, the site of a crude oil spill. The
aquifer consists of glacial outwash materials, is approximately 20 m thick and is underlain by low-
permeability till. Volatile constituents are migrating through the unsaturated zone by diffusion, and studies
are attempting a comprehensive understanding of the physical, chemical, and biological processes that will
be needed to develop predictive models of contaminant mobilization, transport, and fate.
Traverse City, Michigan, is the site of an aviation gasoline spill. The U.S. EPA continues to conduct
research here concerning the biodegradation of organic pollutants, in situ biotransformation, remedial action
effects, and sorption processes. Detailed studies are attempting to describe the rates of organic
transformations and characteristics of microorganisms able to degrade the pollutants. Other EPA-sponsored
research involves the mobility of metals, especially with respect to coal-fired power plant wastes, including
four high-level contaminant sites for the study of metal transport processes.
A final example of the kind of field and laboratory experiments of interest for one of the Data Center's
primary uses, model validation, can be found in the HYDROCOIN study described in section 1. The five
level-2 validation cases selected by the HYDROCOIN group were based on information obtained in
laboratory and field experiments conducted in different countries (HYDROCOtN 1986). The first case is a
heater experiment in a borehole cluster in a quarry in Cornwall, UK, where temperature gradients were
determined and linked to convective heat transport. The second case is an analogue for variable density
fluid flow based on a laboratory experiment with thermal convection. Case 3 involves transient pumping
tests and steady-state piezometric data in a monozonitic gneiss block at the Chalk River Site of the Atomic
Energy of Canada Limited. The block of 150 x 150 x 50 meters contains five major structural discontinuities.
Hydraulic and piezometric data as well as core logging and fracture orientation statistics have been collected
in 17 boreholes. The fourth case is the Piceance basin in Colorado, a regional scale flow system of about
50 by 100km. The purpose of this case is to compare computed piezometric heads with heads obtained
from kriging the heads measured in boreholes. Finally, case 5 focuses on an irrigation experiment in Central
Valley, California. Plots of soil were ponded until steady-state saturated conditions were reached in the soil
beneath the plots. The water supply was then terminated and the change in moisture content in the soil
underneath the plots was followed in time. This information was augmented by laboratory soil characteristic
curves determined as part of the study.
SELECTION OF DATASETS FOR INCORPORATION INTO THE DATA CENTER
From the analysis of the information gathered in this preliminary database, some criteria can be derived
for the selection of datasets suitable for incorporation in the final referral database and. for characterization
of these datasets. Many suggestions resulted from the workshop discussions reported in Appendix A.
One of the workshop conclusions was that field research datasets should have the highest priority and
be the focus of the Groundwater Data Center; laboratory bench studies connected to these field research
sites should be next in priority. Independent laboratory data should have a low priority, although effective
model validation may involve laboratory-scale experiments; such data should basically be accepted on an
19
-------
"as-is" basis with no effort made to complete them. Whatever the kind of dataset acquired by the Center,
its quality will always be a major selection criteria.
As a result of extensive analysis and discussion, the Center considers the following criteria for dataset
selection:
• significance to current soil and groundwater pollution or model validation problems
• accessibility of information regarding the dataset
• quality of the data
• completeness of data
• availability in automated format or suitability for electronic filing
• permission for distribution/publicly funded data collection
• availability of documentation of the data collection and dataset
• availability of reports on peer review of data
• timeliness of dataset
Selection based on these criteria requires a standardized approach to information acquisition and
processing. The descriptors used in the final database (see Section 5) are based on the preliminary analysis
and additional experience gained in using this database, as illustrated in Table 7.
20
-------
TABLE 7. PRELIMINARY INFORMATION ON FIELD SITES AND DATASETS
Contaminant Sources
type of source
dates of activity
Studies Conducted
funding
contacts
dates of activity
list of personnel
Type of Data
parameters measured
datatype (raw, summarized, derived)
variables and units used
temporal and spatial resolution
Form of Data
dataset name
history and maintenance of dataset
technical description of files, etc.
examples
Amount of Data
Area of Study, Geographic Coverage
Original Data and Documentation Available—Reports, Articles, etc.
authors/researchers involved
Database Software Used
name
source
example
associated capabilities
Maintenance of Data
Automated format, file structure, file size
Dates
Organizations/contacts
Storage
Availability of Data
is data available for secondary use?
Accessibility of Data
automated format
remote access
transfer of files
special requests
publications
fee information
other
(continued)
21
-------
QA/QC Information
documentation
program standards
Anticipated Updates or Ongoing Studies
Cost Information of Datasets and Reports
22
-------
SECTION 3
EXISTING DATA CENTERS
This section discusses some of the issues and concerns related to the establishment and operation
of data centers focused on environmental data, or more specifically, soil water and groundwater data. The
experience gained in studying existing extant data centers, combined with past IQWMC experience, is crucial
to the successful development of the IQWMC Groundwater Research Data Center.
DATA CENTERS AND DATA CLEARINGHOUSES
The process of gathering and organizing environmental information for dissemination to a community
of users takes on several different forms and labels (not always well defined). Data and information centers
or clearinghouses may provide a wide or limited range of services.
Typical data center services include accessing databases and information directories (e.g., a referral
service), providing data abstracts, and facilitating data exchange. In addition, data centers may provide
access to or even operate integrated database clusters, assist in data integration, perform data evaluation
or analysis, and develop project-specific databases or offer advice on such developments. Often, data
centers are part of larger organizations such as governmental agencies and research institutions. Ideally,
integrated database systems contain data selected from several sources, formatted and documented to
conform to existing standards in order to facilitate regional, integrated analyses (Merrill 1985). A data center
may be extensively involved in data processing, e.g., converting datasets into machine-readable formats,
or in data interpretation and data quality assessments.
Some confusion reigns as to what constitutes a data center or clearinghouse. For example, Olson
and Millemann (1985) define a clearinghouse as a data referral center containing data descriptions but not
actual data. However, in a review of data centers Olson (1984) states that some clearinghouses provide
direct access to data while others do not.
In this report we use the terminology of the Earth Sciences Information System of the U.S. Geological
Survey (USGS 1982). Here, a clearinghouse is defined as being a directory only; a repository operating as
a directory with data; a data center as a directory containing data and providing evaluation services; and
an integrated database system as a directory containing data along with evaluation and analytical services.
Previous Studies
In the past, a number of studies have provided an overview of existing data distribution facilities and
have assessed the usefulness of data centers. Most of these studies have been based on information
collected in literature reviews, from questionnaires, and from interviews; sometimes, such studies included
23
-------
the "trial" use of the data center concept. The extent and completeness of the studies varies from
small-scope intra- or interagency studies to statewide and nationwide surveys. The following section
discusses the findings of some of these studies.
In 1982 the Committee on Data for Science and Technology (CODATA) of the International Council
of Scientific Unions (ICSU) conducted a large international survey of data sources. Over 650 data centers
and referral centers in 94 countries were identified. The study found widespread international concern over
access to numerical data and stressed the importance of clearinghouses in providing an efficient mechanism
to locate and access data resources (CODATA 1982).
In another study, Armentano and Loucks (1979) found an overlapping hierarchy of databases
throughout federal and state government, academia, and industry; the reliability and associated
documentation of these databases varied widely. They concluded that the potential contribution of these
databases to the furtherance of research, resource management, and regulatory purposes is perhaps only
"touched upon." They noted that investigators gave little thought to the use of data other than for immediate
research purposes. Databases were computerized in some cases but were often not documented and
published, resulting in limited user-accessibility. According to this study, poor coordination and lack of
planning were prominently illustrated by the number of agencies that stored their often uncomputerized
data "in-house," and that had no way to enter their data into some type of master storage system. The
study by Armentano and Loucks proposed a network of regional or thematic databases. This national
network of regional environmental data centers should be designed to meet the national need for locating
and cataloging existing data and models and for increasing their accessibility and utilization. Although such
a system would help in locating data resources, the user would still be faced with acquiring data,
standardizing formats and units of measure, and reconciling differences in spatial and temporal attributes.
In reviewing existing environmental and natural resource databases Olson (1984) focused on the type
of data available, funding statutes, and long-term database maintenance. The study was based on
information from mail surveys, telephone interviews, and the outcome of the 1983 Integrated Data Users
Workshop (Olson and Millemann 1985). Twenty-four federal agency clearinghouses, referral centers, or data
centers were identified which maintain machine-readable inventories of environmental and natural resources
data files. The study lists a variety of management approaches and data center objectives among these
facilities, including clearinghouses or referral centers, compilations of dispersed national datasets to serve
specific agency needs, and integrated database systems containing collections of datasets. A major issue
appeared to be the long-term maintenance and funding of data systems. As Olson (1984) concluded: "Either
the funding agencies will need to explicitly allocate resources for this activity or a more effective mechanism
to recover these costs from data users will need to be established."
The U.S. Geological Survey has been active in earth science data collection and dissemination for
several years. Many of the resulting data sets have been computerized. To provide rapid access to these
databases the USGS has published a number of surveys (e.g., USGS 1979,1983). This information is also
accessible through the Earth Science Information System (ESIS), a comprehensive data management facility
designed to support the coordination, integration, and standardization of data within the USGS. ESIS
maintains a central and uniform repository of detailed information on the earth science databases and
24
-------
systems and on associated data elements. The on-line retrieval capabilities of this referral system provide
for efficient multiple keyword searching in a user-friendly, menu-driven environment with user-activated help
screens and customized reporting. The descriptive characterization of databases and data systems used
by ESIS is listed in Tables 8 and 9.
TABLE 8. ESIS INFORMATION CATEGORIES FOR DATABASES
Category
Comments
Database name
Acronym
Database type
Data structure
Contact
Spatial data types
Coordinate system
Time span of data collection
Status of database
Users
Availability for use/access
Access method
Output media
Storage media
Size of database
Computer
Language
DBMS
Subject coverage
Geographical coverage
Sources of data
Documentation
Keywords or descriptors
Definition or description
short descriptive name or title
database acronym, short name, or name abbreviation
spatial, scientific, and technical; bibliographic; or mission support
logical structure of database records
name, address, and telephone of contact person
e.g., point, line, cell, polygon, grid
e.g., operational, under development, under modification, closed
(not further maintained)
known or intended users
on-line, batch, indirect, etc.
printout, magnetic tape, etc.
media where database is stored
logical records, bytes per record, expected yearly growth
make and model on which database resides and location of
computer
used in software for database management and access
database management systems used
individual or organizational sources
user manual and descriptive references
database abstract
25
-------
TABLE 9. ESIS INFORMATION CATEGORIES FOR DATA SYSTEMS
Category
Comments
System name
Acronym
Application category
Contact
Databases accessed
Date implemented
Last modified
Expected application life
Application status
Annual operating costs/budget
Availability for use/access
Access methods
Users
Frequency of input
Frequency of output
Computer
Language
Output media
Input media
Input locations
Related systems
Definition or description
common name of system
system acronym, short name, or name abbreviation
type of applications for system data
name, address, and telephone of contact person
either utilized or integral part of system
data of last major modifications or enhancements
e.g., operational, under development, under modification, closed
(not further maintained)
on-line, batch, indirect, etc.
known or intended users
make and model on which database resides and location of
computer
used in software for database management and access
printout, magnetic tape, etc.
keyboard, tapes, remote-transmission
geographic locations of system input devices
systems that are routinely linked or interfaced with this system
database abstract
Selected Natural Resources and Environmental Data Centers
Data centers designed to store and distribute environmental and natural resources data are operated
by a wide range of public and private organizations. The following discussion highlights the objectives and
services of some of these centers.
26
-------
NAWDEX
The USGS/National Water Data Exchange (NAWDEX) system was established in 1976 to link data
collectors and users for more efficient use of the nation's water data (Edwards 1978). The system contains
information on agencies collecting water resources data. As a management and information system
NAWDEX provides a wide variety of services including search and referral, facilitating direct linkages to data
sources, assisting in data format conversions, performing data quality evaluations, and providing data
manipulation software. Although NAWDEX assumes the role of a data bank for data generators unable to
respond directly to data requests, the system is not a database. The NAWDEX system includes more than
50 local assistance centers. A computerized index describes more than 350,000 sites for which water data
are available and includes descriptions of the geographic location, data collectors, types of data available
for each location, the time periods for which data are available, the major variables measured and their
measurement frequencies, and the media on which the data are stored. NAWDEX also coordinates direct
access to two major water databases: EPA's STORET and the National Water Data Storage and Retrieval
System WATSTORE of the USGS (Cardin et al. 1986). NAWDEX operates an extensive user-support system
including annual user meetings, news bulletins, and various on-line and printed information sources.
In 1983, the USGS Water Resources Division began to design and develop a National Water
Information System (NWIS) (Edwards et al. 1987). The NWIS will replace and integrate the existing USGS
water data systems, including NAWDEX an WATSTORE. The NWIS, which is planned as an interactive,
distributed data system allowing multiple use of the various facilities, is scheduled to go on-line in 1990.
The major objectives in designing the NWIS include:
• a single flexible and expandable system that is easy to use and that improves productivity and
use over existing systems.
• provisions for standardization and uniformity of data handling, data storage, and software
procedures, thus increasing the integrity of the databases and software systems.
• modular database and software systems, allowing the development of thoroughly tested and
efficient software, that are easy to change, enhance, expand, and maintain.
STORET
STORET (STOrage and RETrieval), a central, inter-agency water quality data bank, was developed in
the early 1960s by the Federal Water Pollution Control Administration (FWPCA), the predecessor of the U.S.
Environmental Protection Agency, and is administered by the EPA through its National Computer Center
(NCC) in Research Triangle Park, North Carolina. As of 1985, the original 140-site database now contains
data on more than 500,000 sampling points (EPA 1985b). The data in STORET are usually measurements
of concentration of a particular substance at a particular site defined in space (3-D) and time. The STORET
system supports both scientific and regulatory applications.
27
-------
Originally, STORE! was designed to store information on surface-water quality and sampling
procedures (Shirley 1982). This is reflected in the way it manages general site information: quality assurance
and groundwater-related information, such as aquifer name, geologic formation and age, date of well
construction, well screen interval, etc., must be entered as parametric data. Up to 5,000 parametric values
can be attached to any sample. The principal data storage file within STORET is the Water-Quality File
(WQF), which contains 1,800 unique water quality parameters. The file is updated weekly with monthly
transfers from the USGS-WATSTORE system. STORET provides the user with various options in retrieval
report format, station selection, and the selection of a specific group of samples and observations. In 1982,
more than 40 states and approximately 50 federal, regional, and interstate agencies actively used STORET,
primarily for surface water data. Increasingly, states are utilizing the STORET database for groundwater
management. Groundwater monitoring data for compliance with federal (RCRA, DIG) and other regulations
comprise most of the groundwater database in STORET (EPA 1985b).
STORET data is collected and entered into the database by a multitude of federal, state, and local
government agencies and their contractors. To a large extent, the reliability of the data depends on the level
of care employed by these organizations in the processes of sampling, laboratory analysis, and data entry.
However, many problems occur resulting from decentralization of QA/QC (Armentano and Loucks 1979).
To prevent storing of erroneous values, the system performs automatic data checking for about 200 of the
most frequently sampled water quality parameters. Locations are checked within the STORET system by
comparing the latitude/longitude entries against state and county latitude/longitude boundaries (Shirley
1982). Although STORET has some provisions for screening the data, this does not prevent data quality
problems. STORET is programmed to reject data outside a certain range, but unfortunately, data outside
such a range might still be valid, as in case of groundwater contamination. Thus, to cover such events,
the limits on the range of acceptable parameters might be set such that for other sites they do not function
as a quality control.
The large number of parameters in STORET has proven to be a source of problems and complaints.
In selecting data, users who are often not familiar with the details of the data system structure and parameter
definitions, easily err in selecting data. They find STORET cumbersome to use and sometimes supplying
the wrong kind of data or the data in units other than those requested.
An analysis of U.S. EPA data requirements (EPA I987a) recommended many initiatives to improve
STORET. For example, groundwater data that meet the current data quality standards adopted by EPA,
should be stored separately from historic data for which information on QA/QC is lacking or which do not
meet current standards. Over time this process will produce a database that meets the "good housekeeping
seal of approval" (EPA 1987a). Other improvements suggested include user-friendly access and retrieval
capabilities aimed at non-computer-oriented professionals. In addition, the study concluded that STORET
should be modified to obtain consistency with EPA's groundwater data quality requirements.
28
-------
WATSTORE
The National Water Data Storage and Retrieval System (WATSTORE) was established for the
management of water data collected through the activities of the U.S. Geological Survey and to provide an
improved means of releasing the collected data to the public. It is operated and maintained at the USGS
headquarters at Reston, Virginia. WATSTORE contains mainly data from USGS projects and state
cooperative studies. The groundwater data in WATSTORE are collected in a separate data file, GWSI
(Ground-Water Site-Inventory), with up to 270 data items per site (Mercer and Morgan 1982). Most of the
groundwater data in GWSI are raw hydrogeologic data entered directly by the data-measuring agency.
Although several field-collected parameters of water-quality data are stored in GWSI, the majority of water
quality data reside in STORET.
WATSTORE contains a computerized verification system for providing several kinds of error checks,
including syntax, compatibility, and out-of-range checks for physical, chemical, and locational consistency
(Edwards 1983). Services provided range from distribution of simple data tables to complex statistical
analyses. Text summaries presented in USGS annual reports include location, aquifer, well characteristics,
datum, remarks, period of record, and extremes. Water-quality data are transferred monthly from
WATSTORE to STORET.
Other Water Data Centers
A typical example of an integrated approach to data management is provided by the state of Texas,
where a comprehensive computerized information system has been established: the Texas Natural
Resources Information System (TNRIS 1982). This information system provides groundwater managers and
their technical advisors with interpreted or processed data and with extensive support for problem analysis.
At its inception it was expected that such centralized natural resources information system would improve
consistency in data format and quality, and would introduce and enforce data storage and documentation
standards (TNRIS 1983). The current system provides for standardized scales between and within various
categories of data, thus facilitating analysis of complex interrelationships between the various kinds of
geographic data. Through its services and facilities the system promotes increased interdisciplinary data
sharing, allows well-targeted data retrieval for specialized applications, and facilitates increased accuracy
in data interpretation by nontechnical users. The TNRIS system includes six data/information categories:
base data (including aerial photographs and other remotely sensed data, cartographic materials, and
geographic subdivision schemes for locating and assessing natural resources), meteorological data
(daily/monthly records, obtained primarily from the National Weather Service), biological or ecological data,
information on geology and land use (much of this data is not yet in machine-readable format),
socio-economic data (census information), and water data (including USGS stream flow data). Files
pertaining to groundwater include water level measurements, groundwater quality data, location and
characteristics of subsurface injection wells, well drilling logs, and a well sample and core library. The
state-wide water level network is comprised of more than 7,000 wells; the water quality monitoring network
includes over 5,000 wells. Water level measurements may be annual, biannual, bimonthly, or continuous,
and generally make use of automatic recorders.
29
-------
Most states take a less integrated approach than Texas. For example, the Nebraska Natural
Resources Data Bank Information System Data Access Manual (NNRC 1981) is a state-wide facility for
storing, processing, and retrieving basic water and soil data. Both groundwater levels and water quality
data are available. Data may be retrieved by county or specific stations. Data summary services include
number of observations, maximums and minimums, means, standard deviations, variance, variation
coefficient, and beginning and ending record dates.
Project Data Management
In some cases, project-oriented databases have reached a high level of sophistication and detail and
thus provide an useful example for other data acquisition and data management projects. One such project
is the Wisconsin Power Plant Impact Study (WPPIS) (Shacham and McLellan 1979). To make the datasets
collected during this study accessible to other scientists, investigators at the University of
Wisconsin-Madison prepared a reference volume. The description of each dataset includes name of dataset,
names of investigators, brief abstract, experimental method, dates of research, and data form. The project's
data management was designed to facilitate continuous data synthesis during the data gathering.
Procedures were put in place to prevent duplication of unnecessary data collection, to facilitate the transfer
of data and other information from one disciplinary subproject to another, and to organize the data to
facilitate their maximum use. Protocols were established to assure that data were as free from interpretation
as possible, and that data were in a format that allowed assessing uncertainties in the data independently.
In addition, the project data documentation describes data form and format in detail, provides a data sample,
and contains a QA report and other supporting documents. Such a detailed documentation makes these
datasets highly useful for secondary analysis, synthesis, or theoretical studies.
DATA MANAGEMENT
The availability of data and the form of the datasets resulting from research projects are related to
their projected use. Often, however, not much thought is given to data use beyond the project's research
objectives. To address this issue the W.K. Kellogg Biological Station (1982) requires in the planning stage
of each research project that measures be taken to facilitate later access to the collected data. In the
management of the datasets with which it is involved, the Station distinguishes between four major
responsibilities: (1) research data analysis by the data-collecting team (primary use); (2) compilation of
databases for secondary use; (3) data referral through directories and catalogs (information services that
help researchers locate and obtain datasets); and (4) archiving in data banks (established to maintain [at
least] those datasets that have no other means of long-term care).
Data management considerations should include identification of potential data (center) users;
identification of data that should be included in the database or center; alternatives for finding a suitable
method of data presentation; determination of how data is to be documented; determination of how the
center is to be described and made known to users; and selection of storage and retrieval methods.
30
-------
Furthermore, the underlying goals of the data center should be well understood by the data center
management; any changes in these goals should be clearly reflected in management considerations.
Management issues include user access and data retrieval, data maintenance, quality assurance,
documentation, standardization, geographic information systems, and system linkage. Each of these issues
is discussed separately on the following pages.
User Access and Data Retrieval
Moffett (1983) stated that masses of data without ready access and analytic capability have very
limited universal value. Unfortunately, this is often the state of most scientific research data because most
databases are being created by agencies, industry, or research organizations for a specific purpose, with
little effort to benefit users beyond the immediate need.
Armentano and Loucks (1979) encountered many problems associated with the access of data, even
from data centers with some type of access procedure in place. Case trial runs to acquire selected
information from a few existing information centers exposed problems in contacting personnel, obtaining
initial access, accessing on-line databases during prime-time hours, specifying output formats, knowing what
datasets were available, user manuals, software system changes, hardware system updates, and terminal
procedures. Identifying a principal contact within the data management organization often proved to be a
frustrating exercise and required a significant amount of time. Delays of up to three weeks were experienced
for the receipt of the requested data. Moreover, for many databases, the access mechanisms, contact
persons, and available resources change frequently. Armentano and Loucks (1979) concluded that contacts
with potential data sources should be formalized and that emphasis should be placed on determining the
accessibility of databases, identifying principal and secondary contacts, and providing complete data
documentation.
Procedures for accessing on-line data systems range from simple to highly complex, depending on
hardware and software environment, data security, and administrative concerns, among other factors. Issues
involved include obtaining computer access authorization and passwords, ensuring compatibility of user
equipment with the host system, and learning the search, retrieval, and report procedures of the particular
data management systems. Commercially offered data systems usually require some contractual agreement
up front, with charges based on connect times. Often, access to these commercial systems is enhanced
by user-friendly menus, on-line help, and extensive system manuals.
Data Maintenance
Data maintenance is the continuing process of adding and reviewing data, updating system hardware
and software, and reviewing and updating quality assurance, data access procedures, and data
documentation. Data maintenance procedures should be documented and this information should be made
available to data system users.
31
-------
Data maintenance requires permanent or long-term operation of the data center. If a data center
project represents only a one-time effort, it may already be out-of-date, or may rapidly become so. Typically,
this is the case if the data, data directory, or user instructions are available only in printed form.
Data entry, modification, and deletion procedures comprise a large part of the computer maintenance
system. Procedures in place for data entry might include automatic syntax checks (e.g., versus a subset
of the 256 ASCII-defined characters), compatibility checks within the database (e.g., depths to water cannot
exceed depth of well), and out-of-range checks (Mercer and Morgan 1982). Some databases have more
extensive error checking capabilities including checking new data for outliers and trends (e.g., Mitchell et
al. 1985).
Quality Assurance
A major issue in the secondary use of research data is quality assurance/quality control (QA/QC).
How does a data center assure the quality of the data offered? In fact, the QA/QC issue can be split up
in two related concerns: (1) the quality of the data acquired by the center and (2) the measures taken to
preserve the integrity of the data within the data center. Armentano and Loucks (1979) concluded that
quality control in database management often appears to be inadequate. They recommended that
procedures be established to certify data files and that mechanisms should be found to fund related data
center management. The researchers collecting the data and providing them to the center must also be
considered in any plan to maintain quality control.
The need for high-quality data varies by program and by decision within programs. QA documentation
often includes descriptions of monitoring equipment such as names and technical specifications (detection
limit, zero drift, accuracy, and precision). For example, the WPPIS study included a separate QA report for
each dataset, thus providing the information necessary for an assessment of the validity and reliability of the
measurements (Shacham and McLellan 1979). Typical topics addressed in these QA reports include the
objectives of the data acquisition; descriptions of the system studied, and of the methods, equipment, and
procedures used; evaluation of the equipment performance; and the names and qualifications of the
investigators.
Another example of data center QA/QC is present in the Texas Department of Water Resources'
Industrial Waste System (TNRIS 1983). Information submitted for storage is examined through edit checks
performed on most data elements to insure that they meet known characteristics and that essential
information is present; cross-checks are made to ensure that the various codes used are defined in the data
system; and related information is matched for economy of storage and to facilitate subsequent retrieval.
Documentation
The main objective of data documentation is to provide a secondary user the necessary information
to decide for whether a particular dataset is of sufficient quality. Whatever data verification techniques are
32
-------
used, the data documentation should make clear to the user what procedures have been used (W.K. Kellogg
Biological Station 1982).
Research project data sets are usually documented in abstract form and include such details as
experimental methods, physical setting, sampling duration, QA procedures, data storage methods, and data
variability. The abstract should provide a brief description of the dataset and include the purpose for
collecting the data and the context of the dataset in the research project as a whole. Armentano and
Loucks (1979) considered data to be best abstracted as raw data.
Olson (1984) concluded that dataset descriptors developed by various clearinghouses or agencies
differ significantly. Most of the descriptions used included a dataset title, abstract, spatial and temporal
coverage, and contact person; some contain detailed descriptions of individual variables within each dataset.
The Wisconsin Power Plant Impact Study recommends a twofold approach to the data documentation
(Shacham and McLellan 1979): (1) a short description of all datasets in the database for initial screening
and selection by secondary users; and (2) a more detailed description for final selection from a short list and
as initial guidance for data use. The extended version includes description of data form and format, a data
sample, a QA report, and supporting documents. It also contains information on funding agency type,
project classification, temporal data, and units of data.
In addition to printed versions, documentation may take the form of help features in the data entry
and retrieval system and text files accessible directly on the computer system used. An example is the
TNRIS "File Description Report" (TNRIS 1982) which includes the file name, kinds of data in the file, units of
measure, geographic coverage, period of record, and capabilities associated with each file (printed reports,
plots, and terminal access).
According to Hoffman (1986), dataset documentation should include dataset name, title, files, research
locations, investigator, other researchers, contact person, project, source of funding, methods, storage
location and medium, data collection time period, voucher material, processing and revision history, and
usage history. Furthermore, data files should be documented with respect to file name, constituent variables,
key variables, subject, storage location, physical size, file creation methods, update history, and summary
statistics.
The W.K. Kellogg Biological Station Workshop (1982) recommended that documentation for data
variables should cover such items as variable name, definition, units of measurement, precision of
measurement, range or list of values, data type, position in file and/or format, missing data codes, and
computational method. For data catalogs and directories, information should be included on dataset
name/title, data collection time period, types of parameter or variable, investigators, contact person, relevant
bibliographic references, dataset storage location, research location, keywords, and research site
characteristics.
33
-------
Standardization
Armentano and Loucks (1979) found that environmental and natural resource data files are generally
widely dispersed and hard to obtain, and that these files reflect a general lack of standards or coordination
(units of measure, spatial identifiers, data formats, etc.). According to a recent study of EPA's groundwater
data management requirements (EPA 1987a), one of the major problems with the current dissemination of
data is the lack of consistency and standardization in data collection, coding, and reporting among EPA
Program Offices and Regions, and the states. This lack of standards reduces the value of the data for
sharing and limits the integration of data originating from many sources. The study concludes that one of
the fundamental building blocks for improving groundwater data management is the adoption of standards
and common formats for data collection and storage.
In reporting the characteristics of datasets it is necessary to document the extent of data interpretation
used before its storage in the database. In addition, uncertainties in the data should be expressed in a clear
manner so they may be assessed independently by secondary users. Observational data should be
prepared and documented in such a manner that they can be re-analyzed independently, e.g., in terms of
a different hypothesis than that presented by the original investigators.
In the development of a data center, standards or conventions for reporting null data, missing data,
numerical values of zero, and values below detection limits must be seriously considered. Although the use
of standards will help integrate different data systems, their development and application should be carefully
considered. As Armentano and Loucks (1979) pointed out, a data center can take aggressive action to
encourage the development and use of data collection, storage, and reporting standards.
Another standardization issue raised in the past is the abstracting of machine-readable files (MRDFs).
An interagency committee on data access and use proposed a voluntary standard for use in preparing
abstracts to describe MRDFs (OIRA 1983) (see Table 10).
Geographical Information Systems
A geographical information system (GIS) is a computerized data handling and data analysis system
designed to accept large volumes of spatial data derived from various sources (Marble and Peuquet 1983).
A GIS allows efficient storage, retrieval, manipulation, analysis, and display of these data according to
user-defined specifications. A GIS is basically a combination of an efficient, dedicated database
management system, standardized data processing, procedures for analyzing relationships between the data
elements, and (graphic) display. Such systems differ from other management information systems by their
focus on spatial characteristics and relationships and in their ability to facilitate the analysis of a large
number of interrelationships among the spatial variables. They allow projections of different data types over
each other (overlays), and extensive statistical comparisons.
34
-------
TABLE 10. COMPONENTS OF THE PROPOSED VOLUNTARY STANDARD FOR ABSTRACTS OF
MACHINE-READABLE FILES (OIRA 1983)
1. Bibliographic Citation
1.1. Title, followed by the general material designation (machine-readable data file)
1.2. Statement of responsibility (authorship, sponsor, collaborator, etc.)
1.3. Place of production
1.4. Name of producer
1.5. Date of production
1.6. Place of distribution (if different from place of production)
1.7. Name of distributor, followed by qualifier (distributor)
1.8. Date of distribution (if different from date of production)
2. Date of abstract and file number
3. General description and subject matter coverage
4. Descriptors
5. Geographic coverage
6. Time coverage
7. Technical description: file structure and file size
8. Reference materials
9. Related printed reports
10. Related files
11. OMB clearance number
12. Contacts
13. Availability information
The GIS concept is interesting to environmental data centers because of its efficient handling of the
highly diverse, spatial field data of interest to environmental research and management. A GIS provides a
framework for data analysis and QA, and drives a certain level of standardization in data formatting,
reporting, and display.
Over the years various types of GIS have been developed, ranging from simple, limited-capability
systems to large multipurpose implementations. Some of these systems have been designed for a specific
management objective or area, while others are completely generic. An example of a large, dedicated GIS
is the Texas Natural Resources Information System mentioned before (TNRIS 1982, 1983). In the "Base
Data Resources" section of TNRIS, digitized water-well locations and county boundaries are present; these,
when combined with latitude/longitude grid lines, provide effective information for local managers. The
system's GIS capabilities can be used to generate products that would be difficult for the user to prepare
(e.g., because of lack of expertise) or that would require considerable time and expense.
35
-------
ARC/INFO (Environmental Systems Research Institute, Redlands, California) is a generic GIS with a wide
variety of user-specified options, available on a large range of computer environments. It is well-suited for
data management related to groundwater modeling projects (e.g., ESRI 1986). It contains routines to
perform topological overlay functions useful in building a model grid, performing advanced analysis, and
allocating spatial variables to the model nodes and cells. Model results can be put back into the GIS and
analyzed together with the existing spatial data (e.g., location of pollution sources, surface water features,
etc.)
Linking Data Management Systems
In some cases, the linkage of different data centers might be more efficient and feasible than complete
standardization or centralization. Such linkage requires extensive coordination between the system elements
to be linked, as well as a clear view of the structure and operation of each individual element. The U.S.
EPA Data Requirements Analysis (EPA 1987a) made several recommendations with regard to such system
linkages around STORET. These will improve remote access to groundwater data (in a kind of "one-stop
shopping") and will facilitate extended data exchange, transfer, and sharing among different organizations.
The benefits become clear if links are initiated between STORET and other national groundwater data
systems, e.g., WATSTORE, or selected state groundwater data systems. The central coordinating agency
(e.g., EPA) should be charged with implementation and operation of interfaces for the routine transfer of data
between all system-elements.
Another example of a linked system is the Chemical Information System (CIS), a privately operated
system of 22 databases and software analytical programs available to environmental scientists. The system
incorporates software to solve problems inherent in system linkage. These problems are generally that each
database may require use of a unique approach to extract information (search and retrieval), and that
various levels of data quality evaluation are required, as well as differences in methods of updating and
expanding the databases.
Data Center Maintenance
An important aspect of a data center is assuring its continuity. Often, this equates with assuring a
certain level of funding for an extended period (e.g., for the center's anticipated lifetime or until a
predetermined date when its utility is re-evaluated). According to Olson (1984), inadequate funding might
result in:
• reduction of the scope of data collection
• decrease in services
• delay in new data additions or less frequent updating
• delay in incorporating new technologies and system updates
• reduced access or unavailability of data for secondary users (i.e., delays in providing datasets in a
timely manner)
36
-------
• reduced quality control
• loss of experienced staff
• obsolescence of data
However, decrease in funding might be justified or compensated by better coordination with data
providers, improvements in operational procedures, and adapting to the continuing advances in computer
technology.
In general, three mechanisms are available to fund a data center's maintenance: (1) funding from
institutional overhead; (2) allocation of project funds to database management; and (3) separate cost
recovery from data users. Often, a combination of cost coverage is sought.
CONCLUSIONS
The range of information or datasets offered by different centers is determined by the particular goals
or purpose of the data center, the means by which data are obtained, and the sources of the data or
information. The requirements of management, especially for government activities, can be short-term and
may be affected by rapid changes in funding or organizational structure. The management and
organizational approaches of a center are initially developed to meet the specific objectives. Although these
objectives may change over the course of a center's operation, the center's progress toward meeting its
original goals must be rapid for a center to be successful.
Many researchers have pointed out the uselessness of data for which the quality is unknown.
Unfortunately, the computer seems to have given data an aura of authenticity it does not always deserve.
Unless a data center documents the quality of its data and is willing to provide detailed characteristics
relating to it, its utility is in doubt. Adhering to rigid QA procedures, and maintaining hardware and software
system security, can help to assure secondary users of overall data quality. However, where possible, such
procedures should not replace direct involvement of the data generators in the secondary use, or at least
their use as information source and guidance.
In order to interest potential users in the service of a data center, these services must be widely
publicized. The clarity, style, and scope of its promotion affects the extent of a center's use.
There is a need to share not only data, but also expertise on information management itself, with several
possible ways to share this expertise. One possibility is to establish training courses, .consulting services,
and internships that benefit from the experience and expertise of leaders in scientific information
management. Conferences or workshops are also of great value.
Finally, data providers are concerned about data integrity and preserving their vested rights in their
data. Armentano and Loucks (1979) suggested that a protocol for third-party use should be adopted, aimed
at preserving the original intentions of the investigator as discussed in research reports associated with the
database. Such a protocol is aimed at securing the integrity of the dataset (and the investigator) and
37
-------
preventing the out-of-context use of the data. One such rule might be that subdividing complex tables into
more than one dataset would require the approval of the original investigator.
38
-------
SECTION 4
QUALITY ASSURANCE/QUALITY CONTROL FOR
RESEARCH DATASETS
INTRODUCTION
To determine the validity of models and their underlying theory, data are used from field observations
and laboratory experiments. The research data in which the IGWMC Data Center is interested might result
from research aimed at any of several objectives: determination of pollutant concentrations from various
sources in the environment, pollutant transport and fate in the subsurface, response of organisms to
pollutants, exposure levels, effects of pollutants on human health and ecosystems, the environment, analysis
of risks and benefits, and evaluation of economic impact. These data must be scientifically valid, defensible,
and of known and acceptable accuracy and precision and, where feasible, reproducible (Stanley and Verner
1985). Thus, the quality of data needs to be described in terms such that these requirements can be met.
To be able to evaluate the quality of data, the data acquisition, data processing, and data storage
procedures must be documented. Because most researchers tend to emphasize the analysis of their data
rather than its collection and validation, important qualifying information about the data is often lacking or
hard to obtain. This problem is of special concern since many research projects focusing on observation
and quantification of natural processes employ new, often experimental data-acquisition techniques and
procedures.
Generally, the quality of data is evaluated by assessing its uncertainty relative to the requirements
for its specific use (which for the same data might vary, depending on the projected use). From a modeling
perspective, a distinction can be made between model validation, use of models in regulation development,
technical design, and legal enforcement (van der Heijde et al. 1988). If the data have consistency and a
small uncertainty when compared with those requirements, they are considered to be of adequate quality
(Taylor 1985). As quantitative measurements are always estimates of the true value of the parameter of
interest, the measurements must be made in such a way that the uncertainty can be posed in terms of
probability. At the same time, assessing the quality of data is a rather subjective process that depends on
the objectives, experience, and personal bias of the assessor.
QA/QC
To ensure that data collection meets project objectives, it is essential that a systematic, well-defined,
and controlled approach be taken to all steps in the data gathering process. The formulation,
implementation, and control of such an approach is the objective of quality assurance/quality control
(QA/QC). Taylor (1985) defines quality assurance as "those operations and procedures which are
undertaken to provide measurement data of stated quality with stated probability of being right." Thus, a
39
-------
primary goal in developing quality assurance in data collection programs is to ensure that such data is of
known quality.
Important to the quality of the data collected is the use of appropriate methodology, and adequate
calibration and proper usage of the equipment involved. Quality control refers to the overall system of
standards, guidelines, procedures, and practices designed to regulate and control the quality of the collected
data (Taylor 1985). Proper documentation of all elements of the data acquisition, data processing, and data
storage procedures is essential for a good quality control program.
To facilitate screening of datasets for their utility in model validation by secondary users (including
the Center's staff), this section provides background information on the various standards, guidelines,
practices, methods, and equipment applicable to the various relevant data acquisition and processing
procedures. Additional QA/QC issues are also discussed, especially with respect to proper documentation.
A distinction is made between data necessary for model-based calculations (numerical values for
model variables and parameters such as concentrations, piezometric head, transmissivities, dispersivities,
etc.) and data that are used for conceptualization only (e.g., to determine hydrogeological schematization
and type of system boundaries; this includes bore hole and morphologic sample descriptions and data from
geophysical surveys). This distinction is made because different levels of quality apply to these two data
categories. The level of detail required in peripheral data such as available through remote sensing and from
meteorological measurements depends on their use in model validation. Often, accurate data regarding
evaporation and recharge from precipitation are essential to successfully complete the testing of a model
for a particular field site.
The Role of Standards and Guidelines
Increasingly, environmental data acquisition is subject to a framework of standards, regulatory
requirements, and technical guidance. In the past the development of standards for field procedures in
groundwater measurements has received little attention. Recently, however, developments in the scientific
and regulatory arenas have accelerated the discussions on standardization and quality assurance in this
area. This recent emphasis stands in stark contrast to the field of analytical chemistry, where over the years
a considerable effort took place to standardize and regulate laboratory methods and practices (Nielson
1987).
The current developments in standardization and establishment of groundwater quality assurance
guidelines are illustrated by the activities of the Subcommittee on Groundwater Monitoring (D18.21) of the
American Society for Testing and Materials (ASTM) in developing standards for monitoring methods and
equipment (Lorenzen et al. 1986, Perket 1986. ASTM 1987a, 1987b, 1987c, 1988, Friedman 1988, Collins
and Johnson 1988). The U.S. Environmental Protection Agency (COM 1987, van Ee and McMillion 1988),
the Electric Power Research Institute (Summers and Gherini 1987), and the U.S. Geological Survey (Collins
and Johnson 1988, pp. 17-21) also study and document groundwater measurements and QA/QC
methodology.
40
-------
Assessment of Data Quality for Secondary Use
One of the most difficult issues in the secondary use of data is determining if the quality of the data
under consideration is sufficient for the intended use. Often, such determinations are hampered by
incomplete definition of the user's objectives, and show inconsistencies between the various data types
evaluated and between the data from different data sources. Even if the documentation of the data is
relatively comprehensive, it often proves to be difficult to make a quantitative assessment of the data quality.
In most cases, assessing the usefulness of a dataset based on a quantitative comparison of data uncertainty
with the limitations on uncertainty placed by their projected use, is not possible.
Describing the quality of data quantitatively requires the definition of measures for data accuracy and
precision. If these measures are not known, or when intuitive, incomplete, or inconsistent approaches have
been taken in deriving them, their value may be limited. Thus, even if measures for data accuracy and
precision are available, the manner in which the data collecting team derived those measures should be
discussed in the documentation. If such quantitative measures are absent, the data user needs to make an
assessment based on qualitative information. In this latter case circumstantial documentation (e.g., field
conditions, methodology applied, etc.) is required. However, such a qualitative assessment of data quality
by the user is often biased and subject to inconsistencies.
As the "true" value of a particular measured data item is never known, its accuracy is established by
ensuring use of the most appropriate method and applying that method as efficiently as possible. Examples
of a systematic approach in evaluating data quality can be found in Campbell and Mabey (1985), Kaplan et
al. (1985), and COM (1987). Campbell and Mabey 1985) proposed the following four-level evaluation
procedure for measured field and laboratory data:
• basis of measurement: identification of method and limitations of the method
• application of method: adherence to the method, controls and calibration, and data conversions
• statistical information: statistical methods and reproducibility
• corroborative information: independent measurements, alternate methods, and related
information.
For each of these categories, the authors discussed the measurement process points to consider.
An additional complexity in evaluating data for model validation is caused by the interpretations which
often take place between the collection of raw data and its transfer to the secondary user. Evaluating the
quality of these processed data includes scrutinizing such data manipulations. Again, proper documentation
of the data manipulations is an absolute requirement.
41
-------
Quality assurance (QA) is a major concern for the successful operation of a central data distribution
facility. This concern pertains to both the quality of the data acquired by the Center and the data handling
within the Center. To assess the quality of data considered for distribution by the Center, information needs
to be analyzed with respect to the kind of data collected and the methods of collection, both from the
perspective of adequate choice of techniques and instruments, and measurement execution.
To this purpose this chapter discusses the role of QA in data collection and handling, and briefly
reviews the major soil water and groundwater data types and related field measurement techniques. Where
possible, discussion of existing standards and guidance has been incorporated. This information served as
a basis for the development of the descriptive system used in the SATURN referral database (discussed in
section 5) and provides guidance to information sources needed in evaluating the quality of the datasets
considered for future incorporation in the Data Center.
SOIL WATER AND GROUNDWATER CHEMISTRY
Introduction
Soil water and groundwater quality at a specific location may depend on factors such as the
heterogeneity of the soil and rock, and past and present flow regimes. In designing a monitoring network
and sampling strategy, the representativeness of samples for the sampled environment is a critical issue.
Standard protocols might remove or control some sources of error and therefore set limits to the
uncertainties present; however, standard protocols do not and cannot address fundamental issues of
representativeness.
The preferred method of determining the quality of groundwater is in-situ sampling, in conjunction with
chemical analysis in a laboratory or on-site. To allow for independent assessment of data quality, the
following items should be documented:
• sampling (network and sampling design, procedures, execution, special operating conditions)
transportation, preservation, and storage
• laboratory or in-situ analysis (including calibration procedures and frequency, use of reference
and quality control standards)
• data reduction, validation, and reporting
Documentation of both sample collection and sample analysis should include: (1) statistical
descriptions providing an indication of the representativeness of the data and the level of confidence placed
on the data; and (2) standard operating procedures to ensure sampling integrity and data compatibility and
to reduce sampling and analytical error.
42
-------
Monitoring System Design and Sampling Programs
A perfectly designed monitoring system would provide the optimum number of samples for a sufficient
but not excessive amount of data to characterize the contaminant/tracer plume, and would also provide the
necessary degree of confidence in the quality of data to support its intended use (Schweitzer and Black
1985). Locating the wells optimally in a monitoring network is a challenging problem, especially taking into
consideration uncertainties in the physical system (Meyer and Brill, Jr. 1988).
The selection of materials and equipment for obtaining the appropriate information must take into
account state-of-the-art techniques and equipment as well as the necessary scope or detail for the project.
The acceptable level of uncertainty must be considered, and the quantification and documentation of this
uncertainty must be established. Decisions regarding analytical requirements for sampling should be made
in conjunction with the design of the monitoring system.
The identification of all possible sources of error is essential to the QA/QC of groundwater monitoring
data. Many researchers (Nelson and Ward 1981, Keely 1982, Schweitzer and Black 1985, Montgomery and
Sanders 1986) have recently attempted to design quantitative groundwater monitoring systems that take into
account data uncertainty and the statistical aspects of field sampling. In defining data uncertainty in terms
of how well representative observed values reflect true population characteristics, Montgomery and Sanders
(1986) propose a method to estimate this uncertainty as a function of both sampling and non-sampling
errors.
Nelson and Ward (1981) state that with careful planning an iterative procedure can be followed in
developing monitoring strategies. In documenting the design of a monitoring system, the following criteria
and considerations should be addressed:
Spatial Soil Water and Groundwater Monitoring
• Number of vertical sampling points in relation to the thickness of the aquifer or soil zone, plume
depth, and the physical character of the system
• Number of horizontal sampling points in relation to plume area and soil and geohydrological
conditions
• Surface-water-related features, drainage, infiltration, and nearby pumpage
Temporal Soil Water and Groundwater Monitoring
• Frequency of sampling
• Seasonal variations
43
-------
• Presence and effects of sorption, degradation, and (bio)chemical processes
• Velocity of contaminant plume
Sample Representation of the True Population Characteristics
• Statistical evaluation
• Soil and geohydrological conditions
• Shape and extent of plume
Methods Used for Design and Evaluation of Data Uncertainty
• Division into sampling and non-sampling errors
• Addressing spatial and temporal aspects of system
• Method of quantification
Background Water Quality
• Comparison with concentrations of contaminants or tracers
Effects of Well Construction on Sampling
The selection of well construction materials and drilling techniques is determined by the hydrogeology
of the site and by the project objectives. In particular, the drilling technique selected depends on designed
well depth and diameter, lithology present, chemical constituents being sampled, the type of monitoring
system being installed, time limitations, need for rock samples, potential effects of circulating fluid (if present)
on the formation, water quality, mobility of the drilling rig, Integrity of the borehole, ability to sample
formation waters during drilling, and economic limitations (Barcelona et al. 1985, Driscoll 1986).
Recent studies addressing the effects of well construction on the quality of groundwater samples
include Fetter (1983) and Dunbar et al. (1985). Barcelona and Helfrich (1986) and Fetter (1983) have pointed
out that casing materials may affect groundwater quality, and thus sampling, especially with respect to
certain organic chemicals.
While studying cement grout contamination in stagnant well water, Barcelona and Helfrich (1986) found
that drilling fluids may significantly affect the groundwater quality and the functioning of the well, regardless
of how often attempts to develop the well are repeated. Additional discussion regarding the selection of a
drilling method and well construction materials can be found in Barcelona and Gibb (1988) and Kerfoot
(1986).
44
-------
The documentation of the employed drilling technique should address the following items, among
others:
• drilling technique
• drilling fluid: type, quantity
• cleaning/decontamination procedures
• well development technique: volume of aquifer water evacuated, rate, time between well completion
and development
• casing materials: type, brand, specifications
« screen material: type, size, specifications
• grouting material and installation: volume, type, procedures
• gravel pack material and installation: volume, size, procedures
• construction procedures
• procedures for special situations: e.g., drilling to avoid mixing of polluted and clean water
• well logging, flow measurements, etc.
Sampling Soil Water and Groundwater
Although QC procedures for groundwater sampling should be based on proven field measurements
and sampling procedures, this is often impossible because of the wide variety of hydrogeologic and
geochemical conditions, and even though many field sampling methods are currently available, their
reliability, precision, and accuracy have not been established (Murarka and Mclntosh 1987). There is still
a lack of standardization of groundwater sampling procedures following systematic development and
controlled evaluation trials, although discussion on standards are presently taking place (e.g., Collins and
Johnson 1988). Therefore, it is essential to document these procedures and the field sampling conditions
with the incorporation of QA/QC techniques used.
The field sampling procedure must account for possible cross-contamination between different wells.
This can occur as a result of incomplete or inadequate equipment decontamination procedures. Bryden et
al. (1986) discuss the need to decontaminate purge-and-sampling equipment in the field when sampling a
number of wells at a site.
45
-------
The partial pressures of certain gases (particularly carbon dioxide, oxygen, and volatile organics),
which are often significant constituents of groundwater, may change dramatically as the water is exposed
to the atmosphere or when vacuum is applied to remove a sample. These changes may be significant in
and of themselves (e.g., loss of volatile organics), or they may produce other, equally significant changes
in groundwater quality, e.g., pH which in turn affects precipitation of other elements. Groundwater sampling
must minimize the loss of any volatile substances, especially if they are the subject of analysis. Specially
designed equipment, methods, and procedures to assure this are described by Barcelona et al. (1984),
Barker et al. 1987), and Pearsall and Eckhardt (1987).
Barcelona et al. (1985) have suggested that a truly quality-assured sampling protocol should be backed
up by laboratory and field experiments that are designed to quantify any random or systematic errors, and
that the most direct way to evaluate sampling protocols is to specify and document that class of chemical
constituents most prone to sampling error. A research tracer study (Roberts and Mackay 1986) offers an
example of documentation of a field experiment on purging requirements and the laboratory testing of a
prototype sampling station. The validation of field procedures depends on techniques already documented
and accepted and on the individual concerns of the project personnel. Any special efforts to validate field
sampling methods in order to meet project goals should also be documented.
Although sampling protocols may vary from well to well due to local hydrogeologic factors and well
construction, once they are defined, wells should be sampled consistently since trends in water quality
data can be created by changing sampling procedures (Sfawson et al. 1982).
For sampling soil water, special measures have to be taken. Vadose zone monitoring can involve
either sample material removed from the test location or in situ measurements (Earth and Mason 1984,
Everett et al. 1984a). Extraction techniques of pore fluid from soil depends on whether inorganics, organics,
or microorganisms are under investigation. Liquid sampling methods depend on whether fluid is taken from
the saturated or unsaturated regions of the vadose zone (Everett et al. 1984a). Many of the documentation
requirements listed above apply equally to soil water sampling.
Everett (1981) provides an introduction regarding instruments and methods for monitoring in the
vadose zone. Wilson (1980) presents a detailed catalog of methods for monitoring or estimating waste water
fluxes in the vadose zone, including a description of the method, principles on which it is based, advantages
and disadvantages, and pertinent references. Everett et al. (1984a) reviewed over 50 types of vadose zone
monitoring devices and methods. Everett et al. (1984b) described the factors limiting the operation of
suction-type sampling devices and discuss problems related to the representativeness of samples.
In summary, documentation of groundwater sampling should address:
Purging
• number of well volumes removed, rate removed, time between purging and sampling, consideration
of hydraulic properties, well design, and changes in color or cloudiness of water
46
-------
• continuous parameter measurements to check effectiveness of purging (e.g., to determine when
representative parameter stabilize): type, equipment used, specifications, calibration frequency,
standards or procedures used
• field /laboratory experiments made to determine purging requirements
• equipment used to purge well
• destination of evacuated water
Sampling System
• equipment and materials (pump or bailer specifications)
• consistency of sampling
• considerations for the sampling of multiphase constituents
• decontamination/cleaning of equipment
« sampling containers
• handling
• labeling
Field Analyses
Field analyses of samples effectively avoids bias in the determination of parameters for constituents
that do not store well (e.g., alkalinity and pH) (Barcelona et al. 1985). Optimal field measurements of pH,
Eh, and specific conductance should be made in the well or in a closed cell to avoid sample oxidation. Field
procedures for measuring unstable constituents are given in USGS (1977) and Barcelona et al. (1985),
among others.
An external QA program for USGS field measurements is described by Cordes (1984). It recommends
that documentation of field analysis should include a description of equipment and methods used, calibration
of instruments, field conditions, reference samples, replicate measurements taken, equipment maintenance,
and inter-comparison studies performed. Logs should detail all field measurements.
47
-------
Sample Handling and Documentation
Sample handling depends on the particular constituents to be analyzed and may include filtration,
temperature control (cooling), chemical fixatives, special sample containers, and elimination of headspace.
In some cases, no special precautions are required. Decisions concerning sample preservation and
containers should be coordinated with the laboratory where the analysis will take place and with the
monitoring team.
Details about the filtering methods can be found in Nacht (1983) and Barcelona et al. (1985), while
USGS (1977) describes container materials to be used for different groundwater constituents. A purge and
trap device for isolating volatiles is described by Semprini et al. (1987).
To assist in interpreting water quality data, the USGS (1977) and ASTM (1987b) give recommendations
for sample identification and documentation. This should include method of collection, exact location of well
or source, sample number, depth and diameter of well, casing record (including screened intervals and type
of screens), water-bearing formation(s), water level, rate of discharge and duration of pumping prior to
sampling, water temperature and other field measurements, date and time of collection, appearance, type
and quantity of preservative added, and any other relevant observations such as use of water.
Methods of sample identification should be consistent and provide for adequate numbering systems
of the sampling points and the samples. Field and laboratory notes should indicate any deviations from
prescribed labeling protocols.
The transportation of a sample may create data uncertainty from such factors as agitation, temperature,
light, and time until filtering and analyses (Montgomery and Sanders 1986). ASTM (1987b) gives
recommendations for sample shipping. Summers and Gherini (1987) outline directions for labeling sampling
containers, packing samples for transport, and sample documentation through chain-of-custody forms. Scalf
et al. (1981) provide guidelines for maintaining sample records.
Data uncertainty from sample storage can be attributed to chemicals added to preserve the sample,
filtration prior to analysis, the type of storage container, the method of storage, and changing conditions of
storage due to such effects as temperature, light, and length of time of storage (EPA 1982, Friedman et al.
1986).
Attention to cleaning procedures for sampling containers is crucial as a prerequisite for sample
collection. Pre-cleaned sample containers are detergent-washed, and acid, solvent, and deionized water-
rinsed under quality controlled conditions.
In summary, documentation of sample handling should address:
Preservation
• filtration in field (filter type used, method and equipment, time lag, filter size, prewashing of filter)
48
-------
• preservatives added (in field or lab)
• type of storage container
• transport and storage conditions
• "headspace" specifications
• temperature control
• coordination between lab and monitoring team
Sample Identification
• documentation system
• sample number and sample type
• well number, sampling point, time, and date
• location
• personnel
Laboratory Analysis
There has been a phenomenal increase in analytical sensitivity for environmental measurements over
the past decade, resulting in very low detection limits for many constituents. However, great disparities are
found between the ability of present techniques or methods to detect and measure contaminants at
ultra-trace levels and the data actually reported at very low concentration levels. In addition, it is rare to find
reported sensitivities, detection limits, or data associated with degree of uncertainty (e.g., in terms of
confidence levels or intervals) (NWQL 1986).
Data Quality Indicators-
Data quality indicators such as precision, accuracy, completeness, and detection limit of the method,
are determined through laboratory procedures for analyzing duplicate, control, blank, and spiked samples,
and through subsequent statistical analysis to quantify the results. Where possible, these data quality
descriptors are expressed and reported in terms of standard deviation, variance, and relationships between
precision and concentration levels (EPA 1983, Paulsen et al. 1988).
49
-------
Precision is the degree of agreement of repeated measurements of the same property, expressed in
terms of dispersion of test results about the arithmetical mean result obtained by repetitive testing of a
homogeneous sample under specified conditions (ASTM 1987c). The precision of a method is expressed
quantitatively as the standard deviation computed from the results of a series of controlled determinations.
Accuracy refers to the correctness of the data and defines the degree of agreement of the measurement
with the true value of the magnitude of the quantity concerned. Bias is the systematic deviation of the
average of one set of data from another. Bias is defined by ASTM (1987a) as the persistent positive or
negative deviation of the method average value from the assumed or accepted value.
Sensitivity is a measure of instrument response factor as a function of concentration. It is commonly
measured as the slope of the calibration curve (NWQL 1986).
A confidence interval is an expression of the range of values defined by upper and lower limits at a
statistically defined confidence level (NWQL 1986).
The Limit of Detection (LOD) is defined as the lowest concentration level that can be determined to
be statistically different from a blank at a specified confidence level. A distinction can be made between
instrument, method, and practical detection limits (NWQL 1986).
The Limit of Quantification (LOQ) is the level above which quantification results may be obtained with
a specified degree of confidence. It is calculated using the standard deviation and average level for the
blank. This defines the level above which quantification is reliable and also a region between MDL and LOQ,
where detection is reliable but quantification is not (NWQL 1986).
The Method Detection Limit (MDL) is defined as the minimum concentration of a substance that can
be measured and reported with 99% confidence, for which the true value, corresponding to a single
measurement, is above zero (EPA 1983). The MDL is the lowest concentration of analyte in distilled water
that a method can detect reliably and is statistically different from the response obtained from a blank carried
through the complete method including chemical extraction or pretreatment of the sample (NWQL 1986).
NWQL (1986) describes the Practical Detection Limit (PDL) as the lowest concentration of analyte in
a real sample-matrix that a method can detect reliably and is statistically different from the response obtained
from a blank carried through the complete method.
Finally, completeness is a measure for valid data obtained from a measurement system expressed
as a percentage of the amount of data that should have been collected. Completeness is of particular
importance to multi-year intensive monitoring programs.
Laboratory QA/QC-
Analytical methodologies and QA/QC procedures have been standardized and modified over a longer
period than groundwater monitoring field procedures. The quantification of data uncertainties is described
using measures of accuracy, precision, bias, sensitivity, confidence intervals, detection limits, and
50
-------
completeness. Quality control samples, including replicates, blanks, controls, spiked samples, rinse samples,
and reference materials, are used to quantify data uncertainty in the laboratory. Eggenberger (1985)
discusses fundamental principles and methods to ensure accurate and reliable laboratory analysis results.
The estimation of analytical procedure variability may be attained by duplicate analyses, by multiple analyses
of a stable standard, or by using spike recoveries. Control of analytical performance may be defined using
the repeatability—the agreement among replicate observations of the test by a single laboratory or analyst
expressed in standard deviations and the percent recovery—describing the ability of the analyst and the
procedure of the analytical system to recover a known amount of a constituent added to a natural sample
(may be obtained by spiking). EPA (1983) and ASTM (1987b) give recommendations for frequency of
duplicate samples and control samples.
The American Chemical Society (ACS), the National Bureau of Standards (NBS), the American Public
Health Association (APHA), and the Water Pollution Control Federation (WPCF) have long been concerned
with the establishment of general laboratory practices and analytical procedures in determining the chemical
composition of sampled systems. Furthermore, the American Association of Laboratory Accreditation (AALA)
has developed a program to accredit environmental laboratories (Locke 1987). To relieve secondary users
of the task of evaluating the capabilities of an analytical laboratory and their compliance with the guidance
and standards developed by these organizations, the EPA has embarked on a certification program for
laboratories involved in drinking water analyses (Hillman 1985).
Reporting Analytical Results--
Documentation of analytical measurements should provide information sufficient to support all claims
made for all the results. Documentation requires all information necessary to: (1) trace the sample from the
field to the final results; (2) describe the methodology used; (3) describe the confirmatory evidence; (4)
support statements about detectability; (5) describe the QA program and demonstrate adherence to it; (6)
support confidence statements for the data (Keith et al. 1983). The identification of the analytical method
should be made in sufficient detail to allow its specific duplication.
Measurement results should be expressed so that their meaning is not distorted by the reporting
process. Reports should make clear which results, if any, have been corrected for blank and recovery
measurements. For example, in all cases for which no analyte was detected, the expression "ND" (not
detected) and "L" (less than) should be used to report the result (NWQL 1986).
Discussion
Many available manuals, handbooks, and references describe groundwater quality monitoring and
sampling techniques (Barcelona et al. 1985, Campbell and Lehr 1973, Dunlap et al. 1977, Scalf et al. 1981,
U.S. EPA 1981, Summers and Gherini 1987, and USGS 1977). More general guidance for selecting the
appropriate measuring techniques is found in Nelson (1988), who indexes over 700 air, water, and waste
measurement methods.
51
-------
Field techniques, equipment, and instrumentation associated with groundwater quality monitoring are
rarely validated to the point that their associated errors can be quantified. Because standards for
groundwater monitoring are still in a state of development, the present discussion and guidance for the
proper consideration and documentation of these activities are solely intended to summarize the current
status of groundwater monitoring activities.
Efficient, well-documented laboratory management procedures are essential to provide quality-assured
environmental data. These procedures include communication between the sample monitoring personnel
and the laboratory personnel, participation in inter- and intra-laboratory comparison studies, procedures for
reporting analysis results, qualifications and assignments of laboratory personnel, validation of analysis
results and analytical methods, and incorporation of automated data storage systems. Efforts are being
made to establish national accreditation or certification programs for laboratories performing environmental
analyses.
I n order that secondary data users may assess the quality and accompanying uncertainties associated
with analytical data, proper documentation, including sensitivities, detection limits, precision, accuracy,
general laboratory procedures, analytical method identification, and other documentation must be available.
The ultimate user of environmental measurements must have access to data quality assessments. All major
environmental databases must be capable of accepting, storing, and retrieving these assessments with each
measurement (EPA 1983).
FLOW INFORMATION AND SOIL AND AQUIFER PARAMETERS
Introduction
The current situation with respect to the QA/QC of hydrogeological field work is comparable to that
for field monitoring of water quality and is significantly less well defined than the QA/QC status of laboratory
experiments and chemical analysis. Field conditions vary from place to place, and even at one site different
situations may occur. Furthermore, technologies to characterize field conditions vary widely in concept,
sophistication, complexity, and reliability. For these reasons regulations, standard procedures, and QA/QC
requirements are more difficult to establish than for laboratory experiments where all conditions are
controllable.
As is the case in acquiring data on the composition of groundwater, in providing data of interest to
the research community, all stages of field work and all related activities should be well documented. For
the data to be of value for secondary uses (e.g., model validation) documentation on geohydrological
characterization should address such issues as methodology, equipment used, and field conditions
encountered.
Regarding the measurements, information should be present on parameters measured, the type of
measurements and measurement methodology, field-procedures, applicable standards and adherence to
52
-------
them, the use of reference and replicate measurements, and noise-sources and noise level. Among other
areas, information should be present on frequency and total number of measurements, location of
measurements (in x,y,z coordinates), date and time of measurements, and statistical evaluation.
The equipment used for both the measurements and data recording should be described in terms of
equipment type and specifications, their calibration, maintenance records present, and inter-comparison of
equipment. Of high interest is information on the precision and accuracy of the measuring devices.
Furthermore, documentation should address the field bias present, methodology and instrument
detection limits, and data confidence limits and resolution. Often, secondary users are interested also in the
qualifications of personnel involved.
This general listing of documentation elements is valid for various types of measurement techniques.
Specific kinds of measurements may require additional documentation.
Water Level Measurement
Water level measurements may be made to determine the phreatic surface in an area, the piezometric
head in a confined or semi-confined aquifer, for analyzing aquifer tests, or to determine drawdowns in a well.
In general, these measurements are performed in a well either developed as a production or injection well
or as an observation well (Powers 1981, Driscoll 1986). The technique of measurement will generally
depend on the well type, aquifer conditions, depth of water table or hydraulic head present, frequency of
measurements, and other relevant field and logistic conditions. Water level measurements should be taken
with great care because hydraulic gradients, critical for correctly estimating groundwater flow rates and
direction, normally are determined from measured level differences between wells, which often are very
small (Bryden et al. 1986).
Important considerations in evaluating water level measurements are the density of the observation
network and the observation frequency. Statistical analyses, including geostatistics and time-series analyses,
are used increasingly in order to optimize groundwatexJevel measurement frequency and network density.
Other factors that influence the effectiveness of an observation network are the vertical piezometric head
gradients (and thus vertical flow) present in the aquifer, seasonal water level fluctuations, trends in water level
changes, unknown screen locations, unknown or excessively long screened intervals, and density differences
of the groundwater due to varying salinity. Additional considerations that need to be taken into account
include well construction integrity and extraneous geohydrological influences (e.g., pressure changes due
to nearby pumping, surface water level variations, surface water control, drainage, and recharge variations).
Important aspects of water level measurements are the selection of measurement sites, performing
well integrity checks, types of production well access (if applicable), definition and selection of measuring
points and reference points, recording devices, and determination of the optimal measurement frequency
(USGS 1977).
53
-------
Direct Measurement of Groundwater Flow
Various instruments are used to measure groundwater flow direction and velocity inside an existing
well or a specially designed well-point (USGS 1977, Kerfoot 1988). Examples of direct flow meter
measurements are presented by Kerfoot (1984). Kerfoot and Massard (1985) discuss sensitivity of flow meter
measurements for various kinds of wells and differences in construction quality. Melville et at. (1985) have
evaluated a heat-pulsing flow meter in a controlled laboratory environment.
Vadose Zone Measurements
Measurements in the vadose zone involve water affected by different forces including gravity water,
capillary water (matric), and hygroscopic water (governed by molecular attraction and removed by heating).
The measurement of soil water may involve the volume of water contained, the movement of water, and the
energy with which water is held in the soil. The soil water content Is inversely related to matric potential.
The rate of water movement for saturated and unsaturated conditions is determined by the head gradient
and the permeability. The volumetric soil water content is given as a ratio (between water-occupied pore
volume and total volume). Soil water retention is often described using moisture characteristic curves of
water content versus matric potential. Other soil parameters of importance for determining soil water
transport include soil salinity and temperature. A recent EPRI publication (EPR11987) discusses field testing
of different vadose zone measurement systems, including tensiometers, resistance cells (yielded moisture
content), and neutron probes. Information on measuring soil-water content, soil-water potential, soil-water
retention, soil-water movement can be found in USGS (1977), Everett et al. (1984a), and ASA and SSSA
(1986), among others.
The unsaturated hydraulic conductivity may be determined in the laboratory by means of steady flow
in soil columns (Nimmo et al. 1987) and in the field by the instantaneous profile method, unit gradient
methods, disc permeameter, and the air entry permeameter. Soil inhomogeneities might substantially
influence the results. Among others, Olson and Daniel (1981) discuss the measurement of the hydraulic
conductivity K in the laboratory and field (in situ) both for fine and coarse grain materials, and discuss
sources of error in laboratory tests using saturated soils. Fang and Evans (1988) describe laboratory
permeability tests and routine soil analysis tests using ASTM standard techniques. Additional discussions
of permeability and hydraulic conductivity in saturated and unsaturated rocks and soils can be found in
USGS 1977, Zimmie and Riggs (1981) and Kashef (1986).
Because many soil properties are established in the laboratory from field samples, documentation of
soil water measurements should specifically address the size of test specimens compared to scale of the
main soil features such as soil stratification, size and spacing of cracks and fissures, and other features of
secondary porosity.
Geomechanical Measurements of Subsurface Materials
Geomechanical measurements are mainly made for evaluation of soils and rock properties at
construction sites. However, many of these properties are identical to those needed in hydrogeologic
54
-------
evaluations, e.g., porosity, bulk density, particle size, and specific surface area. Other geomechanical
properties which might be of concern in model validation are clay mineralogical properties and saturation,
compaction, total organic carbon, and specific gravity of solids. Determination of various parameters is
described by USGS (1977) and Mills et al. (1985). The ASTM Annual Book of Standards series (e.g., ASTM
1988) presents the latest developments in geomechanical testing procedures; standard procedures are
described for soil sampling and field investigation as well as for determining soil texture, plasticity, and
density characteristics, soil water content, and hydrological and structural properties.
Determination of Aquifer Hydraulic Characteristics
Aquifer testing is often done by stressing the aquifer by pumping a well or a system of wells, while
observing the aquifer response by measuring drawdowns in observation points. Currently, aquifer testing
is the most widely used method for determining in situ transmissh/ity, hydraulic conductivity, hydraulic
resistance of a semi-pervious layer, and storativity. Other methods for determining hydraulic aquifer
properties in situ include slug or bail tests, and various borehole tests (auger hole tests, packer tests, tracer
tests, and permeability tests). Measurements conducted as part of an aquifer test include water level,
pumping rate, drawdown in well, and barometric pressure. Each time such a measurement is made, the
time since the beginning of the aquifer test should also be recorded.
Of great importance for an optimal analysis of aquifer test results are the proper timing of water level
measurements, duration of the tests, and number of measurements. Other considerations affecting the
results include the location of observation and pumping wells, the design and construction of the wells, and
proper recording of the raw data (Kruseman and De Bidder 1970, USGS 1977, Hamill and Bell 1986).
Different factors may affect aquifer test results, such as barometric pressure changes, natural water
level changes (in the vicinity of rivers or other surface water bodies, or resulting from variations in
precipitation-induced recharge), and abstractions from nearby wells during the test. Other factors to be
considered in evaluating the quality of aquifer test data are the possibility of improper well development
(resulting in a large skin factor) and variations in the discharge rates caused by pump problems during the
test.
The use of automated computerized systems to conduct aquifer tests and collect the data is rapidly
increasing. A major benefit of this trend is the increase in data quality, as observer inaccuracies and bias
are eliminated and high consistency during the measurements can be obtained (Way et al. 1984).
Many analysis techniques are used to determine hydraulic parameters from aquifer test results (Ferris
et al. 1962, Kruseman and De Ridder 1970, USGS 1977, Mills et al. 1985, Kashef 1986, Walton 1987). Both
time-drawdown or drawdown-distance measurements may be used in these analyses. Different equations
and solution techniques are available, depending on the geohydrological situation. Analyses of aquifer
tests is increasingly done by means of computer programs which calculate the parameters that best fit the
drawdown measurements (Clarke 1987, Walton 1987).
55
-------
Although there are many references for these types of tests and many analytical techniques have
been developed to determine aquifer parameters, there are currently no standards for performance quality
of these tests. Further complications in setting standards are caused by the wide variety of geohydrological
conditions encountered in the field.
Less expensive means of determining aquifer permeability include bailer or slug tests (i.e., the removal
from or addition of water to a well, respectively). These tests involve measurement of the static water level,
the volume of water added or removed, and the number of bailers taken during a given time period (USGS
1977, Freeze and Cherry 1979, Leap 1984). Analytical techniques are used to analyze slug test data to
determine the hydraulic conductivity representing the immediate aquifer area surrounding the test well.
Documentation of aquifer tests should include description of aquifer characteristics, including aquifer
type, the position of the well screens with respect to aquifer depth, and geometry of the aquifer layering and
boundaries.
Tracer Tests
A tracer, injected as a slug into a groundwater body, can provide an efficient way to accurately
describe the movement of groundwater and dissolved chemicals subject to such processes as diffusion,
dispersion, dilution, adsorption, decay, and other (bio)chemical interactions with the soil or rock. The tracer
must be carefully selected to support the purpose of the experiment and not to reflect hydrogeologic
conditions and tracer characteristics. In some groundwater systems, such as those in karst terrain, tracer
tests may be the only way to determine direction and travel time. Tracer tests can derive valuable
information on transport and fate characteristics of pollutants in the subsurface, especially in heterogeneous
and anisotropic aquifer systems. The kind of information that may be derived from tracer tests often
concerns the layering of the aquifer, relative permeabilities of layers, longitudinal and transverse horizontal
and transverse vertical dispersion coefficients, scale dependence of dispersion coefficients, location of the
plume, displacement of the plume, velocity of the plume, and the retardation factor. Detailed information
on tracer test procedures can be found in Smart and Laidlaw (1977), Davis et al. (1985), Roberts and
MacKay (1986), Caspar (1987), Leap and Kaplan (1988).
In addition to the issues described at the beginning of the section on groundwater levels and
groundwater parameters, documentation of tracer tests should include information regarding:
Injection System: volume injected, rate of injection, injection method, construction details
Tracers: composition, concentration, method of mixing, possible interaction with aquifer
material and resident groundwater, and special conditions
56
-------
SAMPLING BOREHOLE CUTTINGS, SOILS, AND CORES
The objective in sampling aquifer materials, soils, and various geologic zones varies widely from
project to project. Objectives include determining intrinsic permeability, hydraulic conductivity, porosity, bulk
density, or the mineralogical characteristics of the formation; characterizing grain size distribution; identifying
the lithology; and conducting geomechanical measurements.
There are many factors to consider in the sampling of soils, aquifer materials, or other solid subsurface
materials (Campbell and Lehr 1983). The purpose of the sampling should be clearly stated and the sampling
technique and equipment selected should reflect this purpose. The most likely sources of sampling error
are sample location, sample size, collection technique, preservation technique, and the number of QC
samples.
USGS (1977) presents the requirements for formation sampling, addressing such issues as the
relationship between sampling technique and drilling method selected, lithology, drilling rate, and fluid-loss
logs, and describing the identification and preservation of samples. Smith (1982) discusses examination of
borehole cuttings. Dunlap et al. (1977) provides a discussion of core taking, handling, and preparation of
laboratory analysis for organic pollutants and microorganisms. Wilson (1980) discusses solids sampling
in the vadose zone, using a variety of techniques and equipment: hand auger samplers, split spoon
sampler-barrel type augers, tube type samplers, and engine-driven augers and drilling equipment. Bruner
(1986) discusses QA/QC pertinent considerations. Flatman (1986) reports on statistical considerations for
design of soil sampling programs.
GEOPHYSICAL SURVEYS
In essence, a geophysical survey is the interpretation of variations in measured response at the earth's
surface or in a borehole resulting from certain (geo)physical forces, either naturally occurring or artificially
induced within the earth's crust. Such variations might result from differences in physical characteristics
within the earth's crust such as density, elasticity, magnetism, radioactivity, and electrical resistivity of the
underlying materials.
Geophysical surveys are being used for many different geohydrological purposes such as determining
and mapping vertical and horizontal changes in lithology, shallow or deep stratigraphy, delineating
contaminant plumes, determining depth to water table, mapping an aquifer, and detecting the location of
a fresh-water/salt water interface. Two groups of geophysical methods can be distinguished: surface
methods and borehole methods. Different techniques are chosen depending on the geohydrological setting
and the goal of the survey (USGS 1977, Campbell and Lehr 1983, Fritterman 1987, and Olhoeft 1988).
The most commonly used surface techniques include resistivity, seismic reflection, low induction
conductivity mapping, transient (or time-domain) electromagnetic sounding, and electromagnetic techniques
(Zohdy et al. 1974, USGS 1977, Benson et al. 1982, Smith 1982, Violette 1987, van Ee and McMillion 1988,
57
-------
Stierman and Ruedisili 1988, and Benson et al. 1988). White (1988) discusses the use of surface resistivity
techniques to determine hydraulic groundwater parameters.
The most common borehole techniques are spontaneous potential, electric resistivity, natural gamma,
neutron, and gamma-gamma. Often, additional caliper, acoustic, and flow logs are made (Keys and
MacCary 1971, USGS 1977, Campbell and Lehr 1973, Benson et al. 1982).
LABORATORY BENCH STUDIES
Generally, laboratory bench studies are conducted to improve the understanding of the individual
processes occurring in the subsurface and their interaction. In the past, such studies have focused on
contaminant transport and fate, including hydrodynamic dispersion, adsorption and desorption, ion
exchange, and the mechanisms, directions, and rates of chemical reactions. Other processes studied in
laboratory experiments have focused on vapor diffusion, liquid diffusion, and volatilization.
Collection methods include chemical analysis, column experiments, and leaching tests. For example,
a large volume of lab results has been compiled on the adsorption of organic compounds onto numerous
materials; however, very few results are reported on the equilibrium adsorption properties of subsurface
formation rocks. Collins and Crocker (1988) describe an experiment using both dynamic flow and static
systems. They, among others, designed experimental protocols for application in evaluating mobility,
adsorption, and degradation of hazardous organic chemicals in a simulated deep subsurface environment.
To illustrate the variety of laboratory studies performed, the following are some of the studies
performed of interest for model validation. Schiegg and McBride (1987) describe a laboratory setup to study
two-dimensional multiphase flow in porous media. Bohn et al. (1979) describe adsorption isotherms
including the Langmuir and Freundlich equations. Francis et al. (1988) discuss waste leaching tests.
Jenkins and Schumacher (1987) discuss extraction solvents for determination of volatile organics in soil.
Olsen et al. (1988) describe a lab bench study on the effects of the permanent composition on pore-fluid
movement. Crocker and Marchin (1988) conducted a lab study to simulate the adsorption and degradation
of enhanced oil recovery chemicals.
The various issues to be addressed in conducting and documenting laboratory bench studies vary
among the different types of studies. When studying pH-dependent reactions, attention should be given to
temperature fluctuations, order of reactions, ionic strength, sorption present, permeability variations for
different fluids, desired precision, and level of confidence. Documentation for research on volatile organic
compounds (VOCs) in soils should include recoveries, sorption, extraction material (fluid), purging times,
temperature, and decomposition. When the focus is on soil pH and conductivity, issues of interest include
the texture variability, water content, salinity, and possible alterations in the chemical equilibria existing
between solid matrix and ions.
Documentation of solubility experiments on radioactive elements should cover hydrolysis, carbonate
complexation, and redox reactions. Bench studies regarding the injection of hazardous waste often focus
58
-------
on the compatibility issues between the waste and resident water and rock and on factors affecting fate,
such as the effects of pH-Eh relationships for the fluids, brine concentration, clay type and clay amount,
presence or absence of iron oxide and organic complexing agents, molecular character of organic materials,
presence of an anaerobic or aerobic environment, and the plugging potential of precipitates formed.
DATA TRANSFER AND STORAGE
There are many steps within the flow of data from their acquisition in the field or in the laboratory to
their use in decision making, the most important of which are the actual data acquisition, site (or laboratory
data processing, site (or initial laboratory) data storage, site-to-office transmission (or transport), office data
storage and archiving, data processing, and data distribution. It is essential that at each of these steps, the
data are managed efficiently. To maintain data integrity attention should be given to the communication and
transmission methods and equipment used, computers systems involved, and database systems and other
software used. For each stage in this sequence procedures should be established to minimize errors and
to obtain a high level of QA/QC in data handling. An in-depth discussion of proper data handling and
storage procedures and QA/QC can be found in Sexton et al. (1987).
The effective sharing of data in an increasingly automated environment requires standards and
conventions for the identification and transmission of data. The Standard Hydrologic Exchange Format
(SHEF) is a standardized system of encoding hydrological data transmissions for both manual and
automated processing (Bissell et al. 1983). Features of SHEF include being readable by both man and
machine, supporting a wide variety of parameters and data types, having a flexible time identification, using
either SI or English units, and allowing flexible use of spaces and comments within the code text to enhance
readability.
A more detailed discussion of the QA/QC aspects of data handling is presented in Section 7.
CONCLUSION
This section has focused on quality assurance aspects of the main variables and parameters needed
in the validation of the current generation of groundwater models. Many aspects of measurement methods
have been touched upon, primarily to indicate information sources regarding current techniques, procedures,
and equipment. Many of the issues discussed apply equally to other data that might be required for specific
models. For example, QA/QC in meteorological data acquisition programs has been the topic of a number
of publications (EPA 1976, USGS 1977, Finkelstein et al. 1983, Lockhart 1985, EPA 1987b).
59
-------
SECTION 5
STRUCTURE OF THE REFERRAL DATABASE AND REFERRAL FACILITY DESIGN
INTRODUCTION
This chapter describes the operational design of the groundwater research referral database or
computerized data directory SATURN and its hierarchical database structure. Also discussed are the
SATURN database objectives, database design criteria, the choice of a relational database structure, and
the effect of these elements on the overall design approach.
The SATURN database system utilizes current microcomputer technology and relational database
management techniques to identify field research datasets suitable for the validation of groundwater and soil
water flow and contaminant transport models.
In general, a database is a structure that contains information about specific topics and that includes
mechanisms to store data or retrieve it. A database should also contain options to manipulate the data and
to display or print it in a useful form.
A referral-type database does not contain the actual data collected; it contains information
characterizing the actual data and information regarding its availability.
According to standard database design procedures, a functional analysis and a data analysis was
performed (Parker and Parker 1986). The functional analysis is a "top-down" approach to identify the needs
for information to be served by the database. The data analysis is a "bottom-up" process to obtain an
understanding of the form of the information (or data) to be incorporated in the database.
To achieve the objectives, three steps are taken (Parker and Parker 1986):
• data collection and normalization; gathering and grouping data into small, logical groups resulting
in data tables
• entity-relationship mapping; associating entities that are related to each other
• data modeling; designing a logical structure for the data elements
Based on the analysis of the needs for information on field research datasets, five types of potential
use have been identified (see Appendix A):
• model validation
- • education/training in the use of models
60
-------
• comparison of field sites
• example/guidance for collection of field data (QA/QC)
• example for the planning of other types of field studies
The first type of use is the primary objective of the database; the other uses are to be considered as spinoffs.
The content of the database is thus a consequence of its primary objective: identifying datasets that
can be used in groundwater models. Because many different types of groundwater models exist (e.g., see
van der Heijde et al. 1988), and still other types of models will be developed, the data requirements of the
models vary widely (e.g., Mercer et al. 1982).
DESIGN CRITERIA
The first step in the development of the referral database facility is identifying applicable standards and
formulating design criteria. Five tasks can be distinguished in the development and the operation of IGWMC
referral-type databases (see Figure 1):
• information analysis: analyzing objects to be described in the database and update of existing
descriptions
• information need or search request analysis: determining the appropriate search strategy and
reporting format, and performing the searches.
• system analysis: developing the database structure and programming criteria to meet the design
criteria
• programming; developing, updating and testing the computer programs, using either existing
database management systems (dbms) software or an independent programming language
• data entry; adding information to the database and updating its content
In the IGWMC information-processing approach, the information analyst functions as an interface
between information providers (e.g., data generators) and the databases. Search request analysts
intermediate between users (e.g., modelers in search of suitable datasets for their model validation efforts)
and the databases, translating the request into searchable questions.
61
-------
Interpreted
data report
annotation
entry/update
data
development
system
•*•-*• feedback
Information
Figure 1. Structure of the IGWMC referral database systems and user interactions
62
-------
The design philosophy of the referral-type databases is to maximize the performance of these different
tasks. More specifically, the databases should include options to facilitate data storage and maintenance,
data search and retrieval, and reporting. The databases should be written such that its structure and
programs can be modified and updated with minimal effort, even when the original designers and
programmers are no longer available for these tasks.
The design criteria are also formulated to assure optimal functioning of the databases for various
categories of users. These design criteria will:
• assure completeness of data
• obtain a balance of information stored
• allow intuitive operation
• facilitate optimal user-computer interaction (e.g., effective screen layout, command structure, and
command execution)
• permit efficient, useful, and "neat" reporting
• facilitate efficient searches
• facilitate efficient, multi-location updating of database content
• realize efficient internal data storage in terms of computer core memory use and mass storage
• facilitate fast operation (e.g., efficient CPU-mass storage device interaction)
• allow portability within the hardware/software environment in which the database is developed
The role of information transfer in the database design is shown in Figure 1, which depicts both the
data stream and feedback features in the setup of the IGWMC referral database systems. Note that the
descriptors required for user identification of datasets (searches) may be quite different from those needed
in the reporting (see discussion reported in Appendix A).
The Data Center will benefit from involving researchers early in the fine-tuning of its procedures and
in promoting of the use of its facilities. Later on, when a referral database is fully operational, user
experience and feedback may lead to changes in the descriptor list, database structure, or operational
procedures.
Some of the design criteria for the research dataset referral database SATURN are discussed in detail
below.
63
-------
Completeness of Data
The section on identification of potential datasets (Section 2), QA/QC (Section 4), and the analysis
of an example dataset (Section 6) provide an extensive analysis of the different kinds of data that might be
present in the various research datasets. The discussion on QA/QC provides extensive guidelines for
documentation on which the actual choice of the referral database descriptors, though less detailed, is
based.
The level of detail in the database should be sufficient to identify datasets of interest to the user.
Only descriptors needed for this purpose should be included in the referral database. If the database were
too elaborate, a number of problems could arise: (1) few if any personnel present at the Data Center would
have an overview of the entire database or would be able to provide sufficient assistance to potential users;
(2) manuals would be more difficult to prepare and update; and (3) extracting all the required information
from the original dataset and its documentation and entering it in the database would be difficult and very
time consuming.
During the 1988 Data Center workshop (Appendix A), where a draft version of the list of database
descriptors was discussed with experts from different research organizations, it was noted that the set of
descriptors should be selected to cover a reasonably detailed part of the given research areas, but that
such a list should not be too elaborate. By avoiding unnecessary complexities and detail and by assuring
coverage of a broad area of related research and field investigations, the initial problems in the operation
of the database system can be minimized.
Thus, in the initial stage of the project, the database should have a flexible structure rather than an
overly extensive or complex set of descriptors. Such flexibility provides an additional advantage: new
developments in any of the involved research areas, or changes of interest in specific aspects of research
or specific parameters, will be easy to incorporate.
Balance of Information
Although special care should be taken to select a set of database descriptors that is well balanced
for the different research areas, the different sections of the database need not contain the same level of
detail. On the contrary, the descriptors presently incorporated reflect the experience and views of the
project staff and some may prove to be quite subjective. As the main purpose of the database is to identify
datasets useful for the validation of particular kind of models, especially transport and fate models, some
research areas are covered in more detail than others by more descriptors, and with a higher level of quality
assurance/quality control. Again, the flexible, modular database structure adopted facilitates future
modifications and enhancements. The final list of database descriptors is presented in Appendix C.
64
-------
Efficient Searches
The SATURN database contains many descriptors, and though it is possible to use each of these
during a search session (using a "query-by-example" option), in practice only a few descriptors will be used
frequently. Once it is known which descriptors will be used primarily to identify the datasets of interest, this
knowledge can be used to optimize the searches. Search strategies thus identified are implemented in the
database system by using so-called pointers, thus accelerating considerably the speed of a search. Search
strategies consist of a sequence of hierarchical searches using specific search criteria, based on the most
frequently used database descriptors. The various search strategies implemented represent the different
search paths users might take to arrive at the information they are seeking. The search strategies selected
for implementing in the SATURN database system include the following categories:
1. Site Name or Geography
The user may know through a literature study or from professional and personal contacts that a
specific site is well suited for the user's model validation objectives. Such a user might want to select
the dataset by site name, city name, or possibly by (approximate) coordinates, and use the referral
database to obtain additional information to judge if the dataset indeed fulfills the user's requirements.
2. Hydrogeology/Lithology/Morphology
Physical, chemical, and to a lesser extent biological processes that affect solute transport all depend
on the lithology of the porous or fractured medium. For this reason lithology will be a primary search
criterion. Related to lithology is the geologic formation. The user may want to select datasets that
are collected for a particular formation. However, such a formation might contain different lithologies.
Moreover, formation naming often varies from region to region and the user might be unfamiliar with
some names. Therefore, such a search will usually be combined with lithology. In addition, the user
may be interested in the hydrogeological schematization and related terminology. For soils, a
somewhat different approach is taken, based on soil type using a taxonomic description.
3. Type and Source of Pollutant or Tracer
When the transport and fate characteristics of a subsurface system depend strongly on the chemical
compounds involved (e.g., type of pollutant or tracer), the search can be performed using the
chemical (s) involved as a primary search criterion. Eventually, the user may wish to search for more
general compound characteristics such as volatility, radioactivity, sorbtivity, etc..
The characteristics of a contamination source have a significant influence on transport and fate of
contaminants. Both the spatial and time-dependent aspects of pollutant transport depend highly on
the location, the time-dependent strength, and the composition of the source. Determination of a
contaminant source is often a problem in case of an extensive pollution situation, as in a multiple
source area, or where the source has been active for several decades.
65
-------
4. Extent of Database and Datatype
Determining the validity of a subsurface flow or transport model requires an extensive and detailed
dataset (van der Heijde et al. 1988). A typical example of field datasets used for this purpose results
from three-dimensional transient aquifer tracer tests or from the detailed tracking in time of well-defined
contamination plumes. Another example can be found in the one-, two-, or three-dimensional transient
tracer experiments in soils. Such experiments and investigations are characterized by the results from
the analysis of hundreds or even thousands of water samples. In addition, these experiments require
the collection of extensive information on geology, hydrology, and physical and mineralogical
properties of the system, and the chemical and microbiological processes present. For these reasons,
the user may wish to select a dataset based on such criteria as the number of chemical analyses
performed, the number of observation points, or the total monitoring time.
5. Completeness and Quality of the Dataset
In some cases, the user may be interested primarily in well documented, high quality datasets in
terms of consistency, completeness, and data quality. A general, single indicator for the QA/QC level
of each dataset would identify such datasets.
6. Previous Model Testing and Validation Applications
Some users may want to validate specific parts of their models with data that has been used for
other model validation purposes. This enables the modeler to compare his or her modeling results
with the results of others, and also an indicates the usefulness of the dataset for such model validation
purposes. Even if the model is designed for more complex systems than the dataset represents,
partial validation of the model for simplified situations could prove beneficial. Comparison of modeling
results for those datasets used for other models might allow the modeler to assess the correctness
of numerical approximations, schematizations, assumptions, discretization, boundary conditions, and
model parameters (van der Heijde et al. 1988).
Selected Search Strategy Criteria
Based on the above-mentioned considerations, a set of search criteria has been selected from the
complete list of dataset descriptors to be included in the adopted search facility. These criteria are listed
in Table 11.
66
-------
TABLE 11. SELECTED SEARCH STRATEGY CRITERIA
Geographic identification
Site name
City/town name nearby
Coordinates (approximately)
Study type
Purpose and scope
Duration
Data type
Pollutant/tracer
Type
Source
Hydrogeology/soil characteristics
Lithology
Soil type
Aquifer/soil material
Extent of database
Types of measurements
Number of measurements/sampling points for main variables in areal plane
Number of measurements/sampling points or zones in vertical direction
Total number of measurements/samples taken in saturated zone
Total number of measurements/samples taken in unsaturated zone
Timespan of measurements/sampling
Frequency of measurements/sampling
Completeness and quality of datasets
QA/QC indicator for data quality
QA/QC indicator for data consistency
QA/QC indicator for data completeness
QA/QC indicator for data documentation
Known model testing applications
Number of reported secondary dataset use projects
Characterization of application
Users should always be able to extract information of interest from the database by using other search
strategies based on their own criteria selected from the complete list of database descriptors.
67
-------
Efficient Storage
In designing the database structure, the amount of disk storage required should be controlled. The
SATURN database incorporates several features to assure efficient storage. Detailed information on the
datasets is stored in separate tables, independently called in by the program manager. As the amount of
detail available may vary considerably for the different datasets, this reduces the disk storage required at
any one time.
Another measure taken to reduce disk storage is adoption of a hierarchical structure based on the
distinction between sites, studies, and investigations. Different research teams may have conducted various
studies at a particular site, each with its own principal investigator or project leader and funding. Also,
within a single study (e.g., with a single funding source) more than one type of investigation or task may
have taken place (e.g., microbiological characterization, geophysics, aquifer testing). In a hierarchical
structure, certain information (e.g., geographic site description) needs to be entered in the system only once,
even if more then one study generating its own dataset has taken place at that site.
Efficient Memory Usage
To include a wide range of users, and to make the program more efficient, memory should be used
wisely. This has been taken into consideration in the design of the SATURN database. Portions of memory
no longer needed are de-allocated. Efficient memory usage has been realized in both code and data.
Both, the use of assembly language routines and a design modular structure ensure maximal memory and
search speed.
Other Design Considerations
In designing the database, several other considerations have been taken into account:
- Flexibility
The database design ensures that future corrections, modifications, etc. are easy to perform, along
with adjustments of the database structure, descriptors, software, manuals, and the reporting formats. This
flexibility has been achieved by using a modular design, a relational database structure, and clearly defined
relations between the different tables of the database.
- Clarity of Structure
The overall database structure should be clear to all users in order to provide easy handling, easy
understanding of the system, and to provide for easy modification.
68
-------
- Relationship with Other IQWMC Databases
For design efficiency and maintenance reasons the structure of the database for research datasets is
to a large extent identical to the IGWMC groundwater model referral system MARS (Model Annotation and
Retrieval System). Many subroutines are shared by both systems. The similarity of the two designs eases
maintenance of both systems, as the programmer/! system analyst has to be familiar with only one type of
program structure and a single programming language.
External Users
To maintain integrity of the database, a
and reporting facilities of the database system will
distribution. The complete system including
maintaining their own database or in modifying th
provide regular updates of the database content, a
database structure meets the requirement that onl}
distributed. To assure the integrity of the
maintenance of the database content and program
Center.
version of the database and executable versions of the search
be made available on a subscription basis for general
source codes will be available for researchers interested in
B accompanying software. Measures will be taken to
5 well as support in its use. The modular design of the
a part of all database system programs will be widely
database system and the quality of the information stored,
s should remain under the control of the (IGWMC) Data
- User-Friendliness
Measures have been taken to make the software as easy to use as possible. For example, "help" is
available at any point in the program through the
use of the key. Helpful messages are displayed
on the bottom line of the screen throughout the interactive sessions (e.g., data entry session, search, and
reporting session). To avoid accidental corruption of data, error checking is automatically performed
wherever possible.
Procedures have been designed and programmed to allow for highly flexible updating of the database
content while maintaining data integrity and complying with QA/QC requirements. These procedures are
designed such that any updating of existing records or adding of new records to the Data Center's master
database can be initiated by different operators e|ither at the Data Center or at any other location. These
operators work with their own version of the database (not necessarily the latest version). When the update
of their satellite database is finished, this database is transferred to the Data Center and checked against
the Center's master database. Utility programs are run to identify and isolate new records in the satellite
database not yet present in the master database, display their content for evaluation by the Data Center QA
staff, and upon acceptance renumber them (to ensure a unique site number) and add them to the master
database. Other utility programs are used to chqck the satellite database against the master database for
changes in the content of existing records. Again, records which have been modified in the satellite
database are displayed for evaluation by the Center's QA staff and upon their acceptance are integrated in
the master database. SATURN'S main program includes routines to automatically update a log file to keep
track of date, record number, and type of modifications made to the database content.
69
-------
DATABASE DESIGN
The SATURN referral database system contains information that describes available research datasets.
Every dataset is represented by so-called descriptors. Values for these descriptors are stored in tables
which are organized in such a way that for each "type" of user and for each activity an optimum
configuration is established.
Structure of the Database
A relational database consists of several interrelated tables. The relation between the different tables
is expressed by common columns in two or more tables, the so-called keys (Date 1986). Such relational
databases have proven to be the most effective and efficient type of databases and as a result, lately almost
all databases are relational. Although SATURN has been designed as a relational database, it has some
nonstandard features to avoid duplication of data fields. Specifically, the use of common data tables or
columns has been replaced by direct pointers between data tables in order to avoid searching for the keys
in the separate tables.
In SATURN, three levels of information are distinguished and stored in 13 data tables, using the
information fields identified in Section 2 and compiled in Table 12. On the first level, general information
about each site is stored (see Figure 2). At the second level, general information is stored about each study
that has resulted in a separate dataset. Because more than one study might have been conducted at a
given site, the table on study information might contain more records than the table on site information
(Figure 2). The third level of the hierarchical structure gives information on the different investigations that
might have been performed during the various studies. Level three information is stored in a general task
information table and nine separate tables according to the tasks identified. In addition to the tables
containing these three levels of information, the database contains two external tables with literature
references pertinent to a specific site, study, or task, and references regarding the use of the dataset in
model validation, respectively. The relationships between the three main tables and the two external tables
are realized through the use of pointers (see arrows in Figure 4).
Figure 3 indicates how information is divided over the three main tables. The complete descriptor list
for the database is given in Appendix C.
The SATURN database consists of 14 tables, represented by 14 separate files. The general information
table contains a record for each site in the database. The file containing this table is constructed as a binary
tree (see Figure 5) with the first record of the file the root of the tree. Each record contains two long
integers, one pointing to the root of the left sub-tree, and the other to the root of the right sub-tree. These
pointers are the positions of the records in the file. Consider the following tree: In each of the cells, the top
number represents the key for that record and the bottom two numbers represent the pointers to the two
sub-trees. When the tree is placed in a sequential file structure, the root-record is placed as the very first
record or record zero of the file. The file structure for the above tree is shown in Figure 6.
70
-------
1-
>
"§ 10
«
je
1"
i
•O
3 O « O
® »• W at:
coo.
L. .-
-I & O « S
71
-------
Level 1
gtoQrvphfe Information
tftv datt
ottwr «h» data MaibUhy
modtl AppKotfons/tttting
Level 2
study identification
irajor nxasurMnents mad*
typ* of irM«sur*m«nts/tMks p*rform«d
Level 3
1 .!»
task
1-9
i
i-
•>
fat-Shu field monitoring for water qualify
sampling and chemical analysis
tracer ttsts
water level observations
aquifer tests
toil moisture tension/water content measurements
infiltration/moisture movement measurements
•ofl parameter determination
laboratory experiments
Figured. SATURN database program: Information levels and contents.
72
-------
The other 13 tables are entered sequentially in files. Each new record is added to the end of each.
These tables contain the records for each of the following data types: study description, individual
investigation descriptions, and relevant publications.
The records are added to these tables in the following manner: A general information record is added
for each site. Then, each site may have several study records (or other type of records) associated with it.
The general information record contains pointers to its record(s) in the study (sub -) table. The other
records are handled in the same manner.
The following list contains the 14 files including the general information task file, and describes their
contents.
SITE SDB -> general information (site information)
STUDY SDB -> study information
OBS SDB -> water level observation
MON SDB -> In-situ monitoring
SAMP SDB -> sampling and analysis
TRAC SDB -> Tracer tests
TASK SDB -> general information on individual tasks
FARM SDB -> aquifer tests and parameter identification
TENS SDB -> water content/soil moisture tension measurements
MOV SDB -> infiltration/moisture movement measurement
MMT SDB -> measurement of soil parameters
LAB SDB -> laboratory experiments and measurements
USER SDB -> user references
SST SDB -> site/study/task references
Implementation Details
Each of the records relates to a data definition array which describes the fields for its corresponding
record. Each element of the array is a record containing the following pieces of information:
'Scr1 is the screen number for data entry/update. 'Row' and 'Col' constitute the screen location where
the field should be input. 'FType', the field type, can be alpha, numeric, or logical. 'Mum' and 'Size1 have
different meanings for the different field types. For alpha fields, 'Num' is the size of window in which the
field will be entered, and 'Size' is the maximum number of characters allowed. For numeric fields, 'Num'
is the size of window in which the value will be entered, and 'Size' is the maximum number of digits allowed.
For logical fields, 'Num' indicates the bit number and 'Size' is always equal to one.
The alpha fields are represented by arrays of characters. The first two bytes of the array constitute
an integer value indicating the number of useful characters in the array. This allows an alpha field to contain
73
-------
Sites
sitel
site 2
site 3
site 4
site 5
e
site 6
References
J L
Studies
site 1
study 1
site 2
•tudyl
sitel
study 2
site 3
study 1
5
site3
study 2
e
sitel
study 3
6
1 site 1
study 1
teskl
site 1
study 2
taskl
site 2
study 1
taskl
site 3
study 1
taskl
°site 1
study 3
taskl
Users
L_TL_i
j i
site 3
study 2
taskl
Figure 4. SATURN database program: use of tables and pointers
74
-------
-1
-1
-1
-1
6
-1
-1
Figure 5. SATURN database program: binary tree structure
75
-------
Record *0:
Record +1:
Record *2:
Record *3:
Record *4:
Record *5:
Record *6:
-1 -1
6
-1-1
Figure 6. SATURN database program: file structure
76
-------
up to 65536 characters. The values for alpha fields can be entered using a window smaller than the length
of the field. For example, an 80-character field can be input in a 20-character window by scrolling the
contents left or right as needed.
The numeric fields are represented by words. A word consists of two bytes, and, the value may
range from 0 to 65535.
The following key strokes are allowed for editing alpha and numeric fields during add and change:
- move left one character
- move right one character
- move to the first character
- move to the last character
- move left one word (alpha only)
- move right one word (alpha only)
- toggle insert mode (on/off)
< Del > - delete character under the cursor
< Backspace > - delete character to the left of the cursor
The logical fields are represented by bits. Each record contains an array of words. Each word has
sixteen bits and therefore can accommodate up to sixteen logical fields. A bit value of one indicates true,
and zero indicates false.
The bit number from the data definition record is a number ranging from 0 to 255. Since there are
16 bits per word, the first 16 logical values can be stored in the first element of the array, the next 16 values
in the second element, and so on. Therefore, for any given bit number, we can calculate the array element
number as the quotient of dividing the number by 16, and the bit number within that word as the remainder.
For example, to calculate the bit location of bit number 58, divide 58 by 16.
16/58
48
10
Bit 58, therefore, is the 10th bit in the fourth element of the array, marked by "X" in the following
example.
77
-------
Bit#: 1111110000000000
5432109876543210
Element* 0: 1 0000000000000000 1
- 1: | 0000000000000000 |
- 2: | 0000000000000000 |
- 3: | 00000X0000000000 |
- 4: | 0000000000000000 |
NOTE: The array elements are numbered starting with zero.
SATURN has been developed mainly with Borland's Turbo Pascal v5.0. The repetitive calculations and
time-consuming string handling routines have been developed in assembly language for the 8088/86.
The text fields are represented by character arrays. The first two characters of the array indicate the
number of useful characters in the array (i.e., length of the text); this allows text fields to be longer than 256
characters. Some of the routines (for searching, copying, etc.) are written in assembly language. As an
example, when searching a database for records that contain a specified word or phrase in one of the text
fields, the pattern matching is done by assembly language routines. Other utility routines for string
manipulation, such as finding the Nth character of a string, setting all characters of a string to upper-case,
etc., are also written in assembly language.
Two functions frequently used, MIN and MAX to find the minimum and maximum of two values, are
also developed in assembly language. These functions had to be fast because they are used by the pop-up
and pull-down menu routines. Some other routines used to save portions of the screen for displaying
menus and restoring the screen are also written in assembly language.
Every time ADD is chosen from the main menu, or Query By Example is chosen from the Inquire
menu, several data structures must be initialized. From Pascal, the initialization of large records or arrays
can be very slow. For example, to initialize a two-dimensional array, nested loops must be used, as follows:
FOR I := 1 TO ROWS DO
FOR J:= 1 TO COLS DO
Two_Dim_Array[l, J] : = 0;
78
-------
An assembly language routine is used instead of such Pascal loops. The routines uses the size of the
whole structure in bytes and the starting address in memory to fill the structure with NULL characters.
Every time a record is added or changed in the database, the files containing the modified tables are
duplicated. In an attempt to eliminate the delay from this frequent backup, two assembly language routines
are used.
User Interface and Database Management Programs
The main menu contains items that allow the user to add information, change information, query the
database, and generate reports on the sites and studies in the database. The menus are structured in the
following manner (also presented in Figure 7):
SATURN is operated from the keyboard. Commands are executed by bringing the cursor into the
field of choice and hitting the or < return> key (see Figure 8). The system responds either with
the execution of the command or with a new menu for further selection.
The Menu System
All menus in SATURN are either horizontal or vertical. The horizontal menus are displayed with all
menu items on the same line (Figure 8a), while the items in a vertical menu are displayed on consecutive
lines, starting in the same column (e.g., Figure 8b). When a menu is displayed, the first item is highlighted.
To highlight a different item, the cursor keys are used: for horizontal menus the or
keys, and for vertical menus the or keys. In either menu
type, the key highlights the first item and the key highlights the last item. Alternately,
the first character of the menu item to be selected can be used to highlight that item. Once an item is
highlighted, it can be selected by pressing the key. All menus except the main menu and the
inquire sub-menu can be removed from the screen by pressing the < escape > key. Doing so returns the
user to the previous step. The key calls up help on the menu currently active.
The main menu items are: ADD, CHANGE, INQUIRE, REPORT, and EXIT. The following section explains
the use of these options.
ADD: Adding a record to any of the 14 tables.
A menu of 5 of these database tables is displayed (Figure 8b).
79
-------
Main Menu
I I \
Add Change Inquire
— Site
1 , I
te
udy
isk
te/Study/Task
Reference
ser Reference
l l
\ \
Report Exit
QBE Backtrack Report End-Search
Summary
L_
-*r*
Figure 7. SATURN database program: structure of the user interface.
Full
Printer Disk Screen
80
-------
Add
Change
Inquire
Report
Exit
Add a dataset to the database.
Figure 8a. Main (horizontal) menu
Fl=Help
Add
Change
Inquire
Report
Exit
TABLES
Site
Study
Task
SST BeF.
User Ref
n=ttelp
Figure 8b. (Vertical) add option menu
81
-------
After a table is selected, the first screen of the record layout is displayed (according to Appendix C).
Each screen of the layout contains several fields. These fields may be one of three types: alpha, numeric,
or logical.
Field editing is allowed for both alpha and numeric fields. Each field is assigned a window smaller or
equal in length to the maximum length of the field. While entering information, any of the following keys
may be used to correct mistakes:
-help!
- move left one character
- move right one character
< home> - move to the first character
< end > - move to the last character
- move left one word (alpha only)
- move right one word (alpha only)
- toggle insert mode (on/off)
- delete character under the cursor
< backspace > - delete character to the left of the cursor
Once the information has been typed, the key is pressed to make it part of the record.
Any of the following keys can be used to move from one field to the next or from one screen to the
next:
-help!
- move to next field
< shift-tab > - move to previous field
< page-down > - display next screen
< page-up > - display previous screen
- save record
< escape > - do not save record
CHANGE: Changing the information already in the database.
Upon selecting the CHANGE option of the main menu, the same menu of 5 tables is displayed as in
the ADD option (Figure 8c). After the user selects a table and the record to be modified, the same layout
screens as in ADD are displayed, and the keys defined above are allowed for editing and moving from one
field to another and from one screen to another.
INQUIRE: Searching the database for any of the fields in the record and building a search-list.
82
-------
Add
Change
TABLES
Site
Study
Task
SSI Kef.
User Fef.
Inquire
Report
Exit
FL=Help
Figure 8c. Change option menu
Add
Change
Inquire
Report
Exit
SEARCH
Query By Example
Backtrack
Generate Report
End Search
SUBSET
0901
0192
0134
Munber
Criterion
Records round
81
Original Set
03
Reads a record for query by exanple
Figure 8d. Inquire (or search) option menu and display of existing dataset
Fl=Help
83
-------
When the user selects INQUIRE, a list of all the keys in the database is built. With each subsequent
search, a sub-list is built at the next (sub-)level. This allows the user to backtrack (Figure 8d) one level at
a time during the search (e.g., if the last level appears to be empty).
Query-by-example (Figure 8d) is a search method that allows the user to search using any combination
of fields with the specified values. When this item is selected from the INQUIRE sub-menu, a record
layout, similar to the one in ADD and CHANGE, is displayed (see for details Appendix C). The user can
mark the fields to be searched by moving to those fields and pressing the key. After a field has been
marked, the user must specify a value to be searched for.
Alpha fields are searched to see if the specified value is equal to or a sub-string of the field in the
record being searched. This allows searching for words and phrases in the alpha fields.
Numeric and logical fields are searched for the specified value.
For example, assume that the database contains five records with keys numbered one through five.
The main search list would then be:
+ .„+.„+-..+„.+.„+
Level 0: | 1 2 3 4 5 |
A search on one (or more) of the fields may result in a sub-list containing records two, four, and five.
This would produce two levels of search-list in the following manner:
+_.+...+._+—+.„+
Level 0: | 1 | 2 | 3 | 4 | 5 |
+---+—+—+—+—+
Level 1: | 2 4 5 |
A search at the third level may result in a sub-list with two records, as follows:
+—+—+—+..-+.„+
Level 0: | 1 | 2 | 3 | 4 | 5 |
+—+_.+._+—+.„+
Level 1: | 2 j 4 | 5 |
+—+—+—+
Level 2: 2 5
84
-------
If a search does not find any records that satisfy the search criteria, the user is forced to backtrack
before continuing with the search. Suppose that another search on the example database produces an
empty sub-list:
4- ... 4. ... 4. _ 4, ... 4. ... 4.
Level 0: | 1 | 2 | 3 | 4 | 5 |
+ .— + .-- 4 — 4- — 4. — . +
Level 1: | 2 | 4 | 5 |
+—+_+„.+
Level 2: | 2 | 5 |
+ — +_-+
Level 3: |
Backtracking removes the last level from the search-list, resulting in the same search-list as in the
previous step:
+ ... 4. _« 4- ... 4- ... 4- ... 4.
T — "•• T ""• T """ T ""• T
Level 0: | 1 | 2 | 3 | 4 | 5 |
+ 4. 4- 4- „ 4- ... 4.
... -f» ... -p ... -y ... -^ __• -j-
Level 1: | 2 | 4 | 5 |
+ --+—+--+
Level 2: | 2 | 5 |
+ .„+„.+
The 'Generate Report' item on the INQUIRE sub-menu (Figure 8d) generates a report for one or all of
the records in the last level. A window next to the sub-menu shows a list of the keys for records in the
last level. When 'Generate Report' is selected, an arrow is placed next to the first key in the list. The user
might press the key to generate a report for all of the records in the list. To select one record, the
arrow should be placed next to its key by using the and < cursor-down > keys (Figure 8d)
and the < enter > key pressed. Once the records to be listed in the report have been selected, there are
two more questions— the type of report and its destination. A menu of the type of reports is displayed:
'Summary' or 'Full' (Figure 8e). A list of descriptors used in the summary report is given in Table 12. The
last menu contains the three possible destinations of the report-'Printer', 'Disk File', or 'Screen' (Figure 8f).
If the user selects 'Disk File', then he/she will be prompted for a file name.
REPORT: Generating a report for a record specified by a predetermined key.
After the user enters the key for the record to be printed (Figure 8g), the same two menus as In
'Generate Report' (above) are displayed, one for the type of report (Figure 8h) and the other for its
destination (Figure 8i).
85
-------
An option exists to obtain for each annotation a hard copy of the database content, using the same
format as in the ADO and CHANGE Screens. This option is created specifically for error checking as part
of internal IGWMC QA/QC.
EXIT: Closes all database files and returns control back to the calling program (DOS).
86
-------
fldd
Change
Inquire
SEARCH
Report
REPORT TOTE
Exit
Query By Exa
Backtrack
Generate Rep
End Search
Sunnary
Full
SUBSET
088 L
819Z
0134
Munber
Criterion
Records Found
Bl
Original Set
Generate sunmary report
Figure 8e. Report sub-menu of inquire option
Fl=Help
Add
Change
Inquire
Report
Exit
SEARCH I REPORT TYPE
Query By Exa
Backtrack
Generate Rep
End Search
HIT
DEUICE
Printer
Disk File
Screen
SUBSET
088 L
aisz
8134
Nunber Criterion
Bl Original Set
Records Found
03
Send output to the printer
Figure 8f. Display device menu of report sub-menu
Fl=Help
87
-------
TABLE 12. DESCRIPTORS INCLUDED IN SUMMARY PRINT OPTION
Site name and IGWMC key number
Geographic information
Kind of pollutants/source of pollution
Purpose and duration of investigations
Lithology, main geologic formations and soil types
Hydrogeological situation (e.g. schematization, boundaries)
Aquifer and soil characteristics (e.g. porous medium/fractures/macropores)
Parameters measured
Type of data (raw, summarized, derived)
Spatial variability and resolution
Temporal information on dataset (e.g. sampling or monitoring frequency, length and completeness
of timeseries)
Technical description of data files (format, amount)
Restrictions on accessibility and secondary use
QA/QC information
Publication (key reference on both the site and the selected study)
Odd Change Inquire Report Exit
Site nunbep:
Print categorized report on datasets Fl=Help
Figure 8g. Report option of main menu requires site information
88
-------
Add
Change
Inquire
Report
REPORT TYPE
Exit
Sunnary
Full
Generate sunnary report
Figure 8h. Report option menu
Fl=Help
fldd
Change
Inquire
Report
Exit
REPORT TWE
Sun
Ful
DEUICE
Printer
Disk File
Screen
Send output to the printer
Figure 8i. Display device menu of report option
Fl=Help
89
-------
SECTION 6
DETAILED ANALYSIS OF THE BORDEN DATASET
INTRODUCTION
The Borden site is an abandoned landfill situated on a military base about 80 km northwest of
Toronto in the Canadian Province of Ontario. A large zone of shallow contamination in the sandy aquifer
underlying the base has provided opportunities for a number of research groups to perform extensive
investigations on transport and fate of pollutants In the subsurface. Various studies were aimed at detection
and delineation of the extent of certain contaminating chemicals present in the groundwater. Of particular
relevance to the present report are the detailed tracer studies of extended duration that have been
performed, resulting in various extensive, well-documented datasets (Sudicky et al. 1983, Mackay et al.
1986a,b). Many of the datasets resulting from these studies have been used for testing of groundwater
transport and fate models (e.g., Sykes et al. 1982, 1983, Huyakorn et al. 1984, Tompson and Gray 1986,
Frind and Hokkanen 1987). One of the natural gradient tracer study datasets from the Borden Site (Mackay
et al. 1986a,b) has been selected for analysis in order to develop and test the Data Center's procedures
and to evaluate its data storage and retrieval facilities.
GENERAL PROJECT SUMMARY AND OBJECTIVES
A long-term, large-scale field experiment was conducted in the saturated zone of an unconfined
sandy aquifer located at the Borden landfill site (Figure 9). The study, a collaboration between the Civil
Engineering Department of Stanford University and the School of Earth Sciences of the University of
Waterloo, consisted of a natural gradient tracer test, supported in part by the U.S. EPA. The injected tracers
included bromide (3870 g), chloride (106.60 g), bromoform (0.384 g), carbon tetrachloride (0.367 g).
tetrachloroethylene (0.361 g), 1,2-dichlorobenzene (3.97 g), and hexachloroethane (0.234 g) (Mackay et al.
1986b). The concentration data were collected to further knowledge in the areas of physical, chemical, and
microbiological processes controlling transport in the groundwater environment of the site; to test
laboratory-scale predictions of field-scale transport; and to develop a database incorporating the effects of
chemical interactions, microbiological transformations, and the spatial variability of aquifer parameters
(Roberts and Mackay 1986). By design, the organic solutes varied in mobility and potential for
biotransformation. Injection took place on August 23,1982, over a period of 14.75 hrs. The total flow rate
equaled 13.5 l/min (9 injection wells), and the total volume of the pulse solution was A 1.95 cubic meters.
There were 20 synoptic sampling episodes between 8/24/82 and 6/26/85; the number of samples analyzed
for each episode ranged from 233 to 1883. A time-series monitoring program included 12 sampling points.
Approximately 19,900 samples (solute concentration data) were collected over the 3-yr period. The
approximate sampling domain was 100 m (longitudinal), 20 m (transverse horizontal), and 2-4.5 m (vertical).
The elapsed travel time was 1038 days, with the maximum travel distance exceeding 110 m.
90
-------
Extent of Leachate
Plume (10 mg/l CD
Figure 9. The Borden landfiB site and the location of the natural gradient tracer experiment The
rectangle within the sand quarry Ilustrates the location of the transport experiment and
matches the frame of Figure 11. Also shown is the approximate extent of contamination
from the landffl In 1979, as delineated by a 10mg/l chloride isopieth. (After Roberts and
Mackay1986)
91
-------
AQUIFER DESCRIPTION
The total aquifer volume included in the experiment was approximately 3000 cubic meters. The
aquifer material can generally be described as clean, well-sorted fine to medium sand that is fairly
homogeneous, with some horizontal layering. The phreatic aquifer is underlain by low-permeability clay
layers which contain silts and pebbles. As the top of these clay layers slopes, the saturated thickness of
the aquifer decreases from ca. 20 m at the injection point to ca. 10 m at the tracer zone, which is situated
ca. 500 m downstream of the injection point (see Figure 10). Core samples revealed the properties of the
aquifer: (grain size (0.070-0.69 mm with very low clay content), bulk mineralogy (determined by X-ray
diffraction), particle density (2.71g/cm3), bulk density (1.81g/cm3), porosity (0.33 - calculated), organic
carbon (average = 0.02%), specific surface area (average = 0.8 mz/g), cation-exchange capacity (0.52 +/-
0.09 Meq/100g). Apparent dispersivity (after 11 m advection) = 0.08 m (longitudinal) and 0.03 m (horizontal
transverse) with a vertical dispersion coefficient of 10E-10 mz/sec.
GROUNDWATER FLOW
The average water table (depth below ground surface) is 1.0 m with a yearly fluctuation of
approximately 1.0 m. The horizontal hydraulic gradient ranges from 0.0035 to 0.0054 and a vertical hydraulic
gradient was not detected. During the period of the study the flow direction varied approximately 10 degrees
(see Figure 11). The hydraulic conductivity was determined from 26 slug tests, 2 falling-head permeameter
tests, and grain-size distribution analyses. The overall geometric mean hydraulic conductivity was calculated
to be 9.75 x 10E-5 m/s with a standard deviation SD = 0.62.
MONITORING SYSTEM
The jetted casing method was used to install the nine injection wells (3.8 cm ID PVC) and the
bundle piezometers (1.3 cm ID PVC centerstocks with 12-18 individual sampling tubes of 3.2mm Teflon or
polyethylene). The monitoring network consists of the bundle piezometers/multilevel samplers (assembled
in the laboratory) in a 3-D sampling array of 5000 individual sampling points (Figure 12). Two portable
sampling manifolds, each containing 14 separate sampling stations, allowed all the vertical points at a given
multilevel sampler to be sampled simultaneously.
GROUNDWATER QUALITY
The background water quality characteristics were summarized in these ranges: alkalinity (as
CaC03) = 100-250 mg/l, TDS = 380-500 mg/l, DOC < 0.7 mg/l, DO = 0.85 mg/l. and pH = 7.3-7.9.
92
-------
UJ
O
z
<
h-
(0
&
o
a*
to
I
O
m
o
0>
_2
a
o
2
o
-------
Rgure 11. Water table maps for the tracer experimental site and vicinity (after Roberts and Mackay
1986)
94
-------
• MuMwl S«mpl«r»
4 Infection W«n$
I I 1
J I
20 40
Y(m)
60
80
-2
-a
:::
• • • •
; « :;;»•; •;*;:;;;; < •
: : : t:s:;:::Utt:s • :
: : : tt:::h::::;:t * :
• • t •••*•••••»**•• ! •
• ••••••••*l**»* • *
• »»*•••••••*•!• t *
••••*••••••*•• ; •
ill
20
40
60
60
100
120
Figure 12. Location of multflevel samplers and Injection weOs as of January 1986: plan view (top);
vertical distribution of sampling points (bottom) projected onto cross-section AA' (vertical
exaggeration * 4.6) (after Roberts and Mackay 1986)
95
-------
ANALYSIS
The time between collection (including filtering and storing in containers) and analysis of the water
quality samples (including transportation) was approximately three weeks. Total analyses time for a single
sampling period (up to 1800 samples) was less than two weeks. Samples were trucked from the site to the
University of Waterloo and air-freighted to Stanford University for analysis. Inorganic samples were analyzed
with an automated ion chromatograph and organic samples were analyzed with two gas chromatographs.
QUALITY ASSURANCE
Duplicate samples were collected from one sampling point at each multilevel device (bundle
piezometer). A prototype sampling station was tested in the laboratory and field tests were performed to
determine the vertical spacing integrity and appropriate volume of sample and purging requirements. The
injection system, field sampling equipment and protocols, and analytical methodologies for the project were
documented in detail. Monitoring and sampling materials, containers, and techniques were selected with
consideration to sample representativeness, especially with respect to the volatile organic tracers. The
criteria for the injection concentration levels are as follows: levels were to exceed 100 times the background
or analytical detection limit, whichever was greater; the conservative tracer concentrations were to be kept
as low as possible to minimize the density contrast between the injected pulse and the native groundwater;
and the organic solute concentrations were set to yield roughly equal peak areas in the gas chromatographic
analysis in order to achieve similar sensitivity for all compounds. Several groups of data with questionable
validity are discussed in the documentation (Roberts and Mackay 1986).
MOVEMENT OF THE PLUME
To derive measures of mass, mean velocity, and dispersion, the zeroth-, first-, and second-order
spatial moments were defined over the volume of the plume (Freyberg 1986). The zeroth-order spatial
moment measures the mass of the respective solutes. The first-order spatial moments measure the location,
movement, and velocity of the center of mass of the solute plume. The vertical displacement of the plume
is small. The vertical component of the mean solute velocity vector is negligible. The second-order spatial
moments define the spatial covariance tensor, which measures the spread of a plume about its center of
mass. This is a measure for the dispersion of the plume. Evidence was found of what commonly is called
"scale-dependent" dispersion. The role of growth of the covariance over time is probably not linear, as
would be predicted by the classic advection-dispersion equation with constant effective parameters. Also,
the effect of heterogeneities in aquifer characteristics on dispersion could be distinguished.
96
-------
RETARDATION OF THE PLUME
Due to sorption of the halogenated organic solutes, the different plumes of these organic solutes
are retarded in varying degrees (Roberts et al. 1986). The retardation factors increased with elapsed time
in a manner that suggests deviation from local equilibrium.
The retardation estimates obtained from time series sampling at particular points agreed well with
estimates based on periodic high-resolution spatial sampling.
SORPTION OF ORGANIC SOLUTES
Batch laboratory experiments with core samples have been used to determine sorption-based
retardation factors of the halogenated organic solutes (Curtis et al. 1986).
The isotherms measured were all essentially linear throughout the concentration ranges studied and
appeared nearly reversible. The retardation factors inferred from the observed distribution coefficients were
generally in good agreement with observed temporal and spatial field data.
DATASET CONTENTS
The dataset of the Stanford/Waterloo natural gradient tracer experiment at the Borden site is
currently available from the International Ground Water Modeling Center (IGWMC) on an "as-is* basis. The
tape being distributed contains:
• sampling dates
• x,y,z coordinates of sample location
• concentrations of the seven constituents of the injected tracer solution.
An extensive abstract and additional information on QA/QC-procedures, data formats, field
conditions, and bibliographic references, etc., are available from the IGWMC's SATURN database (see
Section 5). The documentation consists of a report from the data collection team (Roberts and Mackay
1986). This report includes an appendix on the experimental database tape documentation.
SCREENING RESULTS
After analyzing the extensive literature documenting the experiment and the resulting dataset, the
Data Center's prototype dataset annotation form was updated and finalized (see Appendix C). This form
functioned both as a checklist for the Center's dataset evaluation procedure and to evaluate the test dataset.
97
-------
The referral database descriptors do not address the variability of parameters in detaH. The dataset
description allows only for describing whether or not a parameter was measured, and whether or not
information is available on procedures, equipment, and methods used, and on documentation. This means
that for most measurements no indication exists regarding:
• number of measurements
• spatial variability of parameters
• time dependence of parameters
• standard deviation of parameters
• actual parameter values
The "Remarks" field was used to indicate that information on this issue was available.
For large, detailed datasets, the SATURN dataset description format (dataset annotation) might prove
insufficient to fully cover the extent of the available data. In that case, the modular, flexible program
structure of SATURN allows for easy modification and updating of the data files by incorporating additional
fields in the dataset record. Future use of the database should make clear whether or not additional
descriptive information of datasets should be incorporated.
DISCUSSION
Analyzing the Stanford/Waterloo natural gradient tracer experiment provided important feedback
regarding the referral database descriptors. The main corrections concerned the level of detail in the
descriptor list that was necessary to concisely describe the available data and the manner in which to allow
for inclusion of additional information not represented by descriptors.
In the future, updates to the current descriptor list may prove necessary as the interests of the
research and modeling community change.
98
-------
SECTION 7
DATA CENTER PROCEDURES
INTRODUCTION
In order to obtain the best possible operation of the Data Center, procedures have been established
to ensure proper (referral) database and dataset management, acquisition and annotation of dataset
information, acquisition and evaluation of selected datasets, dissemination of information, and distribution
of datasets. These procedures are discussed in the following sections. The QA/QC procedures for these
operation are discussed in Section 8.
DATABASE AND DATASET MANAGEMENT
Computer data management procedures should emphasize data integrity and security, whether for
referral information or actual datasets. This can be done by developing and enforcing strict data processing
procedures that include authorization rules specifying that certain tasks be performed only by a specified
group of users. External users, for example, are not allowed to write, modify, or delete data from the
database or have direct access to the files containing the actual data of distributable datasets. Another
important procedure would require routine data backup, thus allowing recovery of the database or the
dataset files in the event that their content is corrupted, destroyed, or lost.
Although database management covers many topics, the present section discusses security,
integrity, and recovery.
Database security is the protection of the database or its software against unauthorized or illegal
access, either intentional or accidental. Database security can be provided in three basic ways (Kroenke
1977):
• Encryption: storing data in an encrypted format which will only be 'translated' by the system
when authorized users work with the database system
• Definition of subschemas or views: referring to the information that a user is allowed to see
• Authorization rules: limiting access privileges to the database, both internal and external, allowing
only certain tasks (e.g., database design, programming, adding, updating, and deleting data,
performing searches and selecting information for display or reporting, and implementation of
an accounting system), using passwords and audit logs
Integrity of the database management system (dbms) is the mechanism that ensures that both the
data in the database and changes to the data follow certain rules. For example, transmissivity, permeability
99
-------
and aquifer thickness should not be less than zero. Another integrity-assuring method is the use of spelling
checkers for text-field data entry (e.g., bibliographic information). Either the dbms or the accompanying
software should enforce integrity constraints. Integrity is also important when data are transferred between
computer systems. For example, when tapes containing data files have been through many undocumented
manipulations, their integrity might not have been preserved. Data transfer integrity might be preserved by
comparing the transferred data with the original data, by returning a copy of the tape to the author for
verification; this can be done by computer, using existing software.
Recovery is the mechanism for recovering the database in the event that it is damaged in any way.
Generally, two mechanisms assure this: backups and journaling.
In backup procedures, hard copies of both the database and the software are stored at a physical
location away from that of the database, as in a fire-resistant storage place in the same building and one
somewhere else. If calamity should befall the host computer, or if the data files are corrupted (as through
operator error, hardware failure, or a so-called computer virus), the backup version assures that no valuable
information will be lost. In order to maintain an up-to-date version of the database in each of these backup
locations, they must be replaced at regularly scheduled intervals. Such a schedule is determined by the
frequency of updating the database or the data files and the cost of replacing the information lost between
backup and time of data loss. Measures need to be in place to check the integrity of the backup
immediately after it has been prepared.
For the IGWMC Data Center, internal backups of datasets will be replaced after each new dataset
that has been downloaded on the computer system. In addition, the internal backup procedure calls for
backup as part of the computer facility's own backup procedure (e.g., every month). Backups will be
stored for at least three years as will the backups of every third year prior to this period.
To prevent program alteration, additional measures can be taken, including separation of program
source codes from the user-accessible environment, using only executable images of the software. In
addition, periodically refreshing the operational software (both database management software and
application programs) by reloading from backup might prevent unauthorized program alteration.
As the SATURN referral database will be widely distributed, complete with search and
report-generating software, procedures are adopted to prevent data corruption off-site while maintaining the
integrity of database content. To protect this content, only compiled versions of the search and
report-generating software will be widely distributed, and database users will be provided periodically with
updated data files and application programs to replace their previous versions. On request, database users
will receive the complete SATURN version, as it is public domain software. In that case, they need to assure
themselves of the integrity of the database content. Specific software under development by the Data
Center will allow user-made modifications of the database content to be checked against the master
database at the Data Center and when accepted to be incorporated in the master database.
Journaling is keeping a log or journal (preferably automatically) of all the activities that update or
modify the database. In case of a computer or database system crash, the system manager can always
100
-------
restore the latest correct version of the database. At the IGWMC, journaling is part of QA/QC procedure
and is currently being automated.
INFORMATION MANAGEMENT
Information Acquisition and Processing
Procedures have been established for collecting and processing information on field research sites
which might provide datasets suitable for model validation. These procedures have been tested using the
information collected during the first phase of the Data Center project. The procedures have drawn on the
experience of the IGWMC in modeling and information processing.
Identification of Potential Datasets—
The IGWMC Data Center staff continuously collects and analyzes information on datasets they have
identified. The initial information may come from open literature or from presentations and discussions at
conferences, workshops, and other meetings, or may be obtained directly from researchers.
Once a dataset of interest is located, additional information is collected from the research team that
collected the data, and from pertinent literature, to enable the Center's staff to include the dataset in the
SATURN database. In selecting a dataset for inclusion in the referral database, special attention is given to
the quality of the available data, to whether or not standard procedures were used during data collection,
to the extent of the dataset, to the kind of chemicals and processes involved, and to existing restrictions on
data transfer and documentation distribution.
To assure consistency in the evaluation of the site and dataset information and in the data entered
in the referral database, a standardized form, the SATURN data entry form, has been designed (see
Appendix D). A complete dataset annotation includes comments made by the original research team and
the IGWMC staff, as well as bibliographic references regarding its collection, interpretation, and primary and
secondary use. After detailed evaluation of the field site and the data collected by the Center's staff, a
SATURN data entry form is filled out, and the data is entered in SATURN (see Figure 1 in Section 5).
To ensure that the dataset description is correct and complete, a full report of the stored information
is verified with the dataset contact person, if identified.
To collect information on field sites where large data collection projects have taken place, or where
a large number of separate investigations have been conducted, the Center might choose a project
approach in cooperation with outside experts.
Once all the information describing a dataset is entered in the referral database, the information is
checked for errors (see also related QA/QC in Section 8).
101
-------
Periodically, the database content will be reviewed and updated where necessary, following the
procedures described above.
Information Dissemination—
The Center distributes the information stored in the SATURN database in three ways: through
published overviews, distribution of database files and search and report programs, and in response to field
support requests. Regularly, the Center will publish, in the open literature, reviews and overviews of the
datasets described in the referral database. Major outlets include journals, conference proceedings, and
IGWMC's own publication series. For those who need assistance in locating detailed datasets for model
validation, the IGWMC staff is available for consultation by telephone, correspondence, or visits to the
Center.
To provide rapid, direct access to the database, remote access by telephone will be made available;
or, the user may obtain a distributed version of the database system. This distributed version is available
on a subscription basis, to allow the Center to provide the user with updated content and application
software.
The experience of the Center in disseminating model information indicates that persons contacting
the Center for such information are not familiar with the variety of models available, the detail in which
these models are described, or with the selection process. In such cases, IGWMC staff provides assistance
in helping define user needs and determining the most efficient search and report strategy to meet those
needs. The same situation is expected to occur with requests for dataset-related information.
If the search is to be performed by an IGWMC staff member, care should be taken that the proper
descriptors are used as a search criterion. A retrieval form will be filled out by the requestor or by the
Center's staff. A report containing information on the search and a description of the selected datasets will
be checked for errors by the Center's staff before is sent to the requestor. From this information the user
might select a dataset for a specific problem. Eventually IGWMC staff members can play an advisory role
in this decision, as they are familiar with the usefulness of the different datasets for specific testing and
validation purposes. If the desired dataset and additional information are available from the Data Center,
they can be sent to the user on request.
To evaluate the efficiency of its information dissemination process, the Center's staff will maintain
a record of the information requests received and of the follow-up by the Center.
DATASET MANAGEMENT
Selection. Acquisition, and Evaluation Procedures
The Center may receive unsolicited datasets, or it may actively seek out a dataset for incorporation
in its distribution package.
102
-------
Selection of datasets to be distributed by the Center is determined by distribution criteria, priorities,
and availability of staff time and funding. As discussed in Section 2, the dataset selection criteria adopted
by the Center include:
• its significance to current groundwater pollution or model validation problems
• accessibility of information regarding the dataset and its collection
• presence of sufficient documentation (might be expanded and improved by Data Center)
• completeness of dataset (as raw data and/or processed data)
• quality of the data
• current format (electronic format or hard copy); if data are important and only available as hard
(paper) copy, the Data Center might transfer them to an electronic storage medium
• timeliness of the data
• potential to obtain permission for distribution (e.g., if data collection has been publicly funded)
The first step in the selection process is screening descriptive information available in the SATURN
database. When the Center considers a dataset for possible distribution, the staff contacts the dataset
author (or custodian) to obtain detailed information regarding the dataset status, its availability, restrictions
in distribution, its format, and pertinent documentation. If the Center decides to distribute a dataset, it will
obtain written permission from the dataset author, custodian, or responsible agency or institute.
The Center will ensure that any dataset to be obtained is uncorrupted. Whenever possible, an
integral copy of the electronically transferred dataset will be returned to the author for automatic verification
of its content, or the dataset provider will be asked to prepare a duplicate for verification at the Center.
Upon receiving the dataset, the Center will copy the tape or disk content into a computerized
master-directory and will store the original tape or disk in a safe place. An identical procedure will be
followed for the documentation. In any case, the Data Center will acknowledge the receipt of the material
and follow the QA/QC procedures detailed in Section 8.
Potential Problems In Acquiring Datasets
Various problems, limiting data availability and accessibility, can occur in acquiring datasets. Such
problems include the presence of mixed formats, data "tied-up" in projects, and data available only in
hard-copy form ("shoebox filing"). Obtaining data from research projects often means spending significant
effort coercing the data custodian to release the data and to provide the necessary data documentation.
Another problem is that the datasets pertinent to a particular site often result from different investigations
and thus reside with different custodians.
103
-------
Furthermore, researchers often are reluctant to share data, even when their collection was made
possible through public funding, because they want to make absolutely sure that their data contain no errors.
During the discussions at the Data Center workshop (see Appendix A) it was noted that "you must cover
yourself completely before the release of data.* After data is collected during a research project, much effort
is required to document, reformat, and acknowledge the data before they are released. There may be
internal authorship battles which inhibit the release of data and documentation. Furthermore, a researcher
often wants to get as much out of the data as possible. It is difficult to collect data and analyze it at the
same time—the data may not be available for a long time, as when they are not released for two to three
years after their collection. Even after data is made available in principle, its accessibility may be limited by
any of the aforementioned factors. Specifically, a problem occurs when researchers responsible for the
initial data collection claim that secondary data use, as in model validation, constitutes an infringement of
their authorship rights; some researchers consider the results of such validation as their own research finding
even if they have not contributed to the theoretical development of the model. The Data Center encourages
model developers in search of field datasets to validate their models, to team up with the data generators
to ensure optimal and correct use of the data. Dealing with these issues, the Data Center follows the general
IGWMC approach and does not take a confrontational position, but will accomodate the data generators'
concerns where possible. In any case, secondary users are asked to give full credit to the source of the
data and inform the data generators of this secondary use. Arrangement of financial support for data
generators documenting their data before release for secondary use and for their participation in the
secondary use should be made in advance.
The Data Center will develop a protocol for secondary data use as discussed in the last paragraph
of Section 3. This protocol will address such issues as protecting the integrity of the dataset, prevention of
misuse or misinterpretation of the data, adhering to the original intent of the data collectors, the role of data
collectors in secondary use, and proper acknowledgment of the data generator's work. The protocol must
be accepted by the data user before the Data Center will release the data for secondary use.
Dataset Preparation and Distribution
When the dataset is acquired, the Center's final evaluation will be made of its quality and
completeness and the completeness of its documentation. The outcome of the Center's final evaluation will
be added to the distributed documentation.
As part of the distribution process, the computerized data files will be prepared for easy
reproduction. The transfer medium will be a tape or disk. If the dataset has been altered by the Center, the
distributed files will contain both the original "unaltered* data and the modified data. Each tape or disk wPI
be labeled as to its content and format specifications and will be accompanied by a listing of the files and
their content. Furthermore, the Center will distribute with the dataset a listing of all pertinent information
regarding the site and the dataset (from the SATURN database) and a copy of the pertinent (core-)
documentation.
104
-------
Initially, each dataset will be distributed on an as-is basis. The dataset requestor should check the
integrity of the dataset received. For primary datasets, those that IGWMC judges to be of high quality, the
Center will eventually summarize the error-checking conducted by data collectors and perform its own
extensive error-checking. However, such evaluations are time-consuming and require additional funding.
If additional funding becomes available, the dataset formats and documentation might be improved and
additional data restructuring, analysis, and reporting might be provided.
Additional information, such as field forms, notes, etc., are not likely to be distributed by the Data
Center. However, if these are available, the same strict regulations and procedures will apply as for the
other distributed materials.
In distributing datasets the Center will apply the same safeguards as it does in acquiring datasets
from researchers. Therefore, the Center will advise dataset requestors of proper verification procedures.
SERVICES OFFERED BY THE DATA CENTER
To cover the costs of the operation of the IGWMC Research Data Center, seven types of activities
need to be distinguished:
- dataset identification and annotation
- dataset selection and acquisition
- dataset preparation (for distribution) and documentation
- dataset transfer
- technical assistance (in quality assessments, dataset use evaluations, reformatting, etc.)
- information dissemination
- internal software maintenance
Two of these activities directly involve secondary data users: (1) information dissemination on
research datasets, and (2) distribution of selected datasets. Information disseminatiorvwill be two-fold: (1)
by responding directly to written or telephone requests; and (2) through the distribution on a subscription
basis of the dataset referral system SATURN. The fee structure for these services w3l be based on the costs
related to the preparation of the information or the datasets in requestor-specified format and the costs
related to the transfer of the materials to the requestor. ASCII flat files will be available for most systems
without the need to reformat, and the data will be available on magnetic tape or diskette. For standard
preparation and transfer activities a fixed fee will be determined.
105
-------
The cost of information analysis and storage In the SATURN system and the maintenance of
SATURN software will be covered in part from overhead on the services directly related to the use of this
system, and to a large extent from future contracts and agreements with funding agencies, especially the
U.S. EPA, focused on its maintenance and utilization.
To fulfil one of the major objectives of the Center, that of distributing existing datasets efficiently,
separate funding is required to cover identification, acquisition, and preparation of datasets for distribution.
This is especially important if the selected dataset is not yet available on an electronic medium, if
documentation is lacking or incomplete, or if extensive interaction with the dataset collecting team is required
in readying the dataset for secondary use. In the first place, such funding needs to be sought from the
agency that funded the original research or from agencies interested in the secondary use of these datasets.
One possibility is for the Data Center to act as subcontractor to and in cooperation with the data collectors
to complete the transfer and final documentation of the data as the last phase of a funded research project.
If this funding mechanism cannot be realized, much of the potential of the Data Center concept will be
unfulfilled.
106
-------
SECTION 8
INTERNAL QUALITY ASSURANCE/QUALITY CONTROL
INTRODUCTION
As discussed in Section 4, quality assurance at the Research Data Center is a functional methodology
for the documentation, filing, and control of technical reports, data, and computer programs prepared
and/or distributed by the Center. The data handled by the Center can be divided into two groups: (1)
numerical data generated in field and laboratory bench studies and managed by the IGWMC Groundwater
Research Data Center, and (2) informational data stored in the Center's referral database SATURN.
This section describes the safeguards taken at the Center to insure that the quality standards adopted
are applied through a variety of mechanisms. The applicable standards and procedures have been
described in Sections 4 and 7, respectively.
The primary objective in implementing QA practices at the IGWMC Data Center is to concurrently
document all data-related activities in order to provide evidence that standards of quality have been
maintained. This objective implies the concepts of integrity, traceability, and accountability.
QA ORGANIZATION
The QA organization of the IGWMC Data Center is integral with the QA organization of the IGWMC
as a whole. There are two levels in the IGWMC QA framework: (1) a permanent organization complete
with QA management policies, goals, and objectives, and (2) project QA organization where general QA
policies and assignments are detailed toward project objectives. A separate IGWMC QA Manual presents
the charter of the Center's permanent QA organization, defining each element of the organization and
outlining its responsibility. Where possible, the persons responsible for QA are independent from those
responsible for its operational activities such as data management and software development and
maintenance.
The hierarchy of responsibility for Data Center activities within the IGWMC is shown in Figure 13 and
is designed with routine QA activities embedded in the technical chain-of-command. This approach differs
from possible structures separating all QA activities. The QA coordinator has a key role outside other QA
responsibilities; this role is currently assumed by the IGWMC Office Directors, whereas the daily QA activities
of the staff are an integral part of the technical work. This implies that QA is not something new or
additional, but rather a part of a permanent, comprehensive approach to file maintenance and reporting
through the chain-of-command. The QA coordinator is brought into a project or task as needed to clarify
QA issues, audit project (or task) files for QA compliance, and coordinate to a certain extent any
inter-division activities.
107
-------
Figure 13. Data Center QA organization
office director > QA coordinator
I I
v v
division coordinator internal and external
| review (ITAC)
v
project or task leader
Responsibilities:
• Project (or task) leader is responsible for adequate QA files; or, if project is the direct responsibility
of the division coordinator, the latter should maintain the QA files.
• QA coordinator audits all QA files and forms; at the close of the project the coordinator ensures
that all folders have been filed (archived).
• After project ends QA control takes over (e.g., adequacy of documentation, form of filing or
archiving, establishing retention period)
• The International Technical Advisory Committee (ITAC) provides advice and technical and scientific
review of the IGWMC programs, services, and products.
QA TRACKING
The internal QA tracking is an important aspect of quality control (QC). The Center uses a series of
forms to ensure that the QC is consistent for all models, information systems, and datasets it maintains and
to provide staff guidance with respect to the QA/QC procedures to be followed. Each form will be subjected
to final QA review by the IGWMC Director before the QA file is completed.
Referral Database
Data Entry—
To analyze a dataset for the SATURN referral database a standard annotation form is completed (see
Section 7). To ensure consistent and adequate completion of this form and subsequent data entry and
filing of the pertinent materials, an annotation-processing tracking form is used (see Table 13). Through this
form the Data Center tracks the completion of the annotation, the entering of the data in the database, the
internal verification of the entered data, the verification of the dataset analysis with the dataset author(s) or
custodian, and the filing of the pertinent materials. For each of the checked activities, the operator, the
date,' and the initials of the checking person are required.
108
-------
Table 13. SATURN ANNOTATION PROCESSING FORM
HD Is Head, IGWMC Data Division
GENERAL INFORMATION
Site Name:
Site Number: Study Number:
ANNOTATION FORM / DATA ENTRY INFORMATION
Encircle: New / Update Log (Up ) date:
Task Operator Date Check
Form completed:
Annotation entered:
Printouts checked:
Sent to Author:
Received from author:
Modifications made:
Hard copy filed:
Comments:
QA Review (IGWMC Director): (date) (initials)
INTERNATIONAL GROUND WATER MODELING CENTER
Holcomb Research Institute, Butler University
Indianapolis, Indiana, USA
109
-------
Information Retrieval—
To document the execution of a search request and the subsequent IGWMC response, a search
processing form is used (see Table 14). The major elements of this form are the primary and secondary
search criteria requested (criteria that have to be met and criteria that would be useful to the requestor,
respectively), and an administrative check-off list. For each of the checked activities, the operator, the date,
and the initials of the checking person are required.
Dataset Distribution
Acquisition—
To document the acknowledgment of the receipt of datasets and proper internal response, the Center
uses a dataset tracking form (see Table 15). For each of the checked activities, the operator, the date, and
the initials of the checking person are required.
Evaluation—
To ensure that a dataset received at the Center is evaluated according to established procedures (see
Section 7), a dataset evaluation form is used. This form documents completion of the major evaluation steps
and key findings of the Center staff member performing the particular evaluation (see Table 16). The main
categories are:
• the extent of the data, their completeness and usefulness
• the dataset documentation
• file structure and data formats
• QA/QC documented
• Assessment of the required level of support (e.g. explanation, additional documentation,
reformatting)
The form also documents the follow-up of the evaluation and the Center's decision with respect to
distribution to secondary users. For each of the checked activities, the operator, the date, and the initials
of the checking person are required.
Distribution—
To document the transmittal of a dataset and appropriate documentation the Center uses a dataset
distribution form (Table 17). This form is both a check-list to ensure that all pertinent materials have been
included, and a proof of transmittal. Therefore, each dataset distributed by the Center will have a dedicated
transmittal form, listing the form in which the dataset has been distributed and the specific items (e.g.,
number of disks, type of tape, list of files, complete listing of printed documentation, etc.). For each of the
checked activities, the operator, the date, and the initials of the checking person are required.
Each dataset distributed by the Center will be accompanied by the statement, "Enclosed are the items
you have requested and which are described below. Please, report to us your acceptance of the materials,
or any problem encountered during installation."
110
-------
Table 14. SATURN SEARCH PROCESSING FORM
HD Is Head, IQWMC Data Division
GENERAL INFORMATION
Information requested by:
Name:
Organization:
Department:
Address:
City:
Telephone: ( )
Date of request:
State/Country:
ZIP/Postal Code:
Form of request: Written / Telephone
SEARCH REQUEST DESCRIPTION
Main Search Criteria/Problem Statement (First Search Level):
(continued)
111
-------
TABLE 14. (continued)
Other Search Criteria (Second Search Level):
Requested Reporting Format: Summary / Full
Task Operator Date Check
Search performed:
Reporting checked:
Sent to requestor:
Copy of report filed:
Comments:
QA Review (IGWMC Director): (date) (Initials)
INTERNATIONAL GROUND WATER MODELING CENTER
Holcomb Research Institute, Butler University
Indianapolis, Indiana, USA
112
-------
TABLE 15. IGWMC DATASET TRACKING FORM
HD is Head, IGWMC Data Division
Dataset Name/Acronym: IGWMC Key.
Form (encircle): tape / disk(s) Number of tapes/disks:
Documentation Included (encircle): yes / no
IGWMC Addressee: Date:
Task Operator Date Check
Passed on to HD:
IGWMC Director notified:
Transmittal letter filed:
Internal review completed:
Authors informed of results:
Dataset baselined/archived:
QA file updated:
Comments:
QA Review (IGWMC Director): (date) (initials)
INTERNATIONAL GROUND WATER MODELING CENTER
Holcomb Research Institute, Butler University
Indianapolis, Indiana, USA
113
-------
TABLE 16. IGWMC DATASET EVALUATION FORM
HD is Head, IGWMC Data Division
GENERAL INFORMATION
Dataset Name/Acronym: IGWMC Key:
Site Name:
Number: Study Number:
Received for evaluation: (date)
Contact Person:
Organization:
Address:
City: State: ZIP:
EXTENT OF DATA/COMPLETENESS/USEFULNESS
Evaluation performed by: Date:
Evaluation:
DOCUMENTATION
Evaluation performed by: Date:
Evaluation:
(continued)
114
-------
FILE STRUCTURE/DATA FORMATS
Evaluation performed by:
Number of files:
Evaluation:
QA/QC APPLIED
Evaluation performed by:
Evaluation:
REQUIRED LEVEL OF SUPPORT
Evaluation performed by:
Evaluation:
General Comments:
TABLE 16. (continued)
Date:
Total storage required:
Mbytes
Date:
Date:
(continued)
115
-------
TABLE 16. (continued)
ADDITIONAL INFORMATION REQUESTED
Topic
Date Requested
Date Received
HD-Check
Documentation:
QA/QC:
Error checking:
(Reformatting:
Author support:
Restrictions:
DISTRIBUTION STATUS
Date
HD-Check
To be distributed by IGWMC: yes / no
Documentation distribution ready:
Packing list prepared:
Annotation completed/updated:
Notification Indianapolis Office:
Notification Delft Office:
Announcement IGWMC Newsletter:
Restrictions:
QA Review (IGWMC Director):
(date)
_(initials)
INTERNATIONAL GROUND WATER MODELING CENTER
Holcomb Research Institute, Butler University
Indianapolis, Indiana, USA
116
-------
TABLE 17. IGWMC DATASET DISTRIBUTION FORM
HD is Head, IGWMC Data Division
To:
Name:
Organization:
Department:
Address:
City: State/Country:
Telephone: ( ) ZiP/Postal Code:
Date of Request: Form of Request: Written / Telephone
Name of Staff Person Handling Request:
Enclosed materials:
o Magnetic Tape (9-track)
Unlabeled o / other:
1600bpi o / other:
ASCII o / EBCDIC o
o Floppy Disk
#
51/4' Double-sided/double density (360Kb)
51/4" Double-sided/high density (1.2Mb)
31/2' Double-sided/double density (720Kb)
31/2" Double-sided/high density (1.44Mb)
Task Operator Date Check
Tape/diskette(s) generated:
Documentation prepared:
Packing list checked:
Invoice prepared:
Final check:
Comments:
QA Review (HD): (date) (initials)
INTERNATIONAL GROUND WATER MODELING CENTER
Holcomb Research Institute, Butler University
Indianapolis, Indiana, USA
117
-------
QA RUNG
The IGWMC QA filing system is a three-tiered administrative structure consisting of archiving pertinent
materials (e.g., baselined computer codes, datasets, documentation, and other products, either in electronic
form or as printed material), QA folders, and QA control files. To organize the QA folders, a hierarchical
filing system is used (e.g., divisions project or task->folder), linking the files to projects or divisions of the
Center. For example, in addition to the separate document folders prepared for each record in the IGWMC
databases (e.g, containing all the past and present versions of a model or dataset annotation and related
correspondence and notes), a separate QA tracking form is completed for filing in the QA task folder for the
database of concern (MARS or SATURN) together with relevant QA information (e.g., audit reports, form
reviews, etc.).
For each model or dataset received by the IGWMC, the Center maintains a separate folder (or group
of folders) containing technical information specific to the model or dataset, including pertinent
documentation, references to related technical reports, copies of technical memorandums and technical
letters, internal and external technical reviews of published or delivered reports, and a hard copy of the
computer codes, code input/output, or raw data received from external sources (complete with transfer
correspondence).
In addition, a separate QA filing system is in place using QA control files for each of the models or
datasets. These files contain the QA forms, QA audits, problem reports and subsequent remedial action,
etc. In the near future this QA control system wfll be implemented on a microcomputer using a database
management system (see Johnson et al. 1987).
For easy identification each model or dataset has its own identification number (IGWMC key), date,
and short description.
118
-------
SECTION 9
CONCLUSIONS AND RECOMMENDATIONS
Sharing data resulting from detailed field experiments and laboratory bench studies is important to
the furtherance of scientific research. It stimulates interdisciplinary use of data and it enables verification,
refutation or refinement of original research results. Added to this is an economical factor as many
experimental studies, especially those carried out in the difficult-to-access subsurface environment, require
complex facilities, elaborate instrumentation and equipment, and are often of prolonged duration.
Increasingly, large, multi-disciplinary field studies are conducted to enhance our understanding of the
complex processes that govern the transport and fate of contaminants in the subsurface. These studies are
often accompanied by laboratory experiments and laboratory testing of field samples. Such studies often
are the result of cooperation between various funding agencies and research groups, each with its own
missions and objectives. Increasingly, one objective of these studies is to promote their secondary use,
and funding agencies are often prepared to provide additional financial support to this purpose. Moreover,
various national and international forums have indicated the need for high quality datasets in order to
perform testing and validation of subsurface contaminant transport and fate models, and to validate
theoretical concepts on which these models are based.
As most research projects and their supporting organizations are not set up to distribute the elaborate
datasets resulting from current field experimental research, the International Ground Water Modeling Center
(IGWMC) in response to this need has established a comprehensive referral facility for selected,
well-documented research datasets. Through this service the IGWMC hopes to prevent situations where
datasets of value to many potential users go unrecognized (often, along with the researchers who generated
the data) and therefore unused.
The datasets of interest to the Center should satisfy various selection criteria, such as accessibility of
pertinent information, availability in automated format, adequate level of documentation, adequate quality
of the data, acceptability of dataset distribution restrictions, and their significance for the main objectives of
potential secondary use.
The secondary user of environmental measurements must have access to information relevant for their
assessment of data quality. Therefor, the evaluation and documentation of the level of quality assurance
applied during data acquisition, data handling and data storage will be a primary concern for the newly
established IGWMC Research Data Center.
A key element of the Data Center's activities is the referral database SATURN. This database, which
will be distributed, is designed to provide IGWMC staff and other users efficient access to important
descriptive information on the content and quality of groundwater research datasets, available either through
the Center or directly from the data originators. In order to design an efficient data directory type of data
base, a list of major dataset descriptors has been developed, facilitating both rapid analysis and
119
-------
characterization of datasets by the Center's staff and fast search and retrieval its users. To allow future
modification of database structure, descriptors and software, the database (programmed in Pascal and
Assembly languages) has a flexible, transparent design.
In the future, additional detail in describing dataset quality and a system of dataset quality levels might
prove an useful improvement. This extra effort should focus on the documentation by the data originators
of information on project QA not included in publications.
Because often data originators and their supporting organization don't have the facilities or institutional
framework to actually distribute datasets and their documentation routinely, the IGWMC Research Data
Center also distributes selected datasets. Wherever possible, this distribution will be done in consultation
with and with the consent of the data originators. The Data Center will encourage publicly funded data
generators to share their data with other researchers for scientific use.
Although currently, the Center only distributes datasets on a "as-is" basis, expanding its services in
the future might be considered to provide processed data according to user's specification. These additional
services might include data reformatting, providing graphics, performing (geo)statistic analysis, and providing
additional analysis regarding the utility and quality of datasets for user's specific objectives.
As the quality of the research data often is of great importance to the end-user, QA/QC aspects of
data collection, handling, transfer, and storage are not only an issue for the data originators, but also for the
Research Data Center. Internal QA/QC procedures and related institutional organization has been tailored
to the existing, highly successful QA/QC program in model information and software distribution of the
IGWMC. It is expected, that in the near future many of the internal QA/QC procedures of the IGWMC will
be automated.
For this new facility to optimally function, groundwater research groups and their funding agencies
should provide it with the necessary information, and funding should be made available to the Center to
maintain and update the database. It is hoped that this new activity of the IGWMC will provide groundwater
modelers with a new efficient means to improve the confidence placed by regulators and decision-makers
in modeling as a powerful analytic tool.
120
-------
SECTION 10
REFERENCES
American Society of Agronomy (ASA) and Soil Science Society of America (SSSA). Methods of Soil
Analysis. 2 Vols., Agronomy Monograph #9, Madison, Wisconsin, 1986.
American Society of Testing and Materials (ASTM). Standard Definitions of Terms Relating to Water. In:
1987 Annual Book of ASTM Standards, Vol. 11.01, Water (1), Philadelphia, Pennsylvania, 1987a.
American Society of Testing and Materials (ASTM). Standard Practices for Sampling Water. In: 1987 Annual
Book of ASTM Standards, Vol. 11.01, Water (1), Philadelphia, Pennsylvania, 1987b.
American Society of Testing and Materials (ASTM). Standard Practice for Determination of Precision and
Bias of Applicable Methods. In: 1987 Annual Book of ASTM Standards, Vol. 11.01, Water (1).
Philadelphia, Pennsylvania, 1987c.
American Society of Testing and Materials (ASTM). 1988 Annual Book of ASTM Standards, Vol. 4.08 Soil
and Rock, Building Stones; Geotextiles. Philadelphia, Pennsylvania, 1988.
Anderson, M.P. Field Studies in Groundwater Hydrology—A New Era. Reviews of Geophysics
25(2):141-147, 1987.
Armentano, T.V., and O.L Loucks. Ecological and Environmental Data as Under-Utilized National Resources:
Results of the TIE/ACCESS Program. The Institute of Ecology (TIE), Indianapolis, Indiana, 1979.
98 pp.
Barcelona, M.J., J.P. Gibb, J.A. Helfrich, and E.E. Garske. Practical Guide for Ground-Water Sampling. EPA
600/S2-85/104, U.S. Environmental Protection, Agency, R.S. Kerr Environmental Research Laboratory,
Ada, Oklahoma, 1985.
Barcelona, M.J., and J.P. Gibb. Development of Effective Ground-Water Sampling Protocols. In: A.G.
Collins and A.I. Johnson, eds. Ground-Water Contamination: Field Methods. ASTM STP 963, Am.
Soc. for Testing and Materials, Philadelphia, Pennsylvania, 1988. pp. 17-26.
Barcelona, M.J., and J.A. Helfrich. Well Construction and Purging Effects on Ground-Water Samples.
Environmental Science and Technology, 20(11):1179-1184, 1986.
Barcelona, M.J., J.A. Helfrich, E.E. Garske, and J.P. Gibb. A Laboratory Evaluation of Ground Water
Sampling Mechanisms. Ground Water Monitoring Review, 4(2):32-41, 1984.
121
-------
Barker, J.F., G.C. Patrick, L Lemon, and G.M. Travis. Some Biases in Sampling Multilevel Piezometers for
Volatile Organics. Ground Water Monitoring Review, 7(2):48-54, 1987.
Barth, D.S., and B.J. Mason. Soil Sampling Quality Assurance User's Guide. EPA-600/4-84-043. U.S.
Environm. Protection Agency, Environm. Monitoring Systems Lab., Las Vegas, Nevada, 1984.
Benson, B.C., R.A. Glaccum, and M.R. Noel. Geophysical Techniques for Sensing Buried Wastes and
Waste Migration. National Water Well Assoc., Dublin, Ohio. 1982.
Benson, R.C., M. Turner, P. Turner, and W. Vogelsong. In Situ, Time Series Measurements for Long-Term
Ground-Water Monitoring. In: A.G. Collins and A.I. Johnson, eds. Ground-Water Contamination:
Field Methods. ASTM STP963. Am. Soc. for Testing and Materials, Philadelphia, Pennsylvania, 1988.
pp. 58-72.
Bissell, V.C., P.A. Pasteris, and D.G. Bennett. The Standard Hydrologic Exchange Format (SHEF) for
Operational Hydrological Data, Hydrological Applications of Remote Sensing and Remote Data
Transmission, IAHS Publ. 145. Internat. Assoc. for Hydrological Sciences, Exeter, UK, 1988. pp.
181-188.
Bohn, H.L, B.L McNeal, and G.A. O'Connor. Soil Chemistry. John Wiley and Sons. New York, 1979. 329
PP-
Bruner, R.J., III. A Review of Quality Control Considerations in Soil Sampling. In: C.L Perket (ed.), Quality
Control in Remedial Site Investigation, ASTM STP 925, Am. Sc. for Testing and Materials, Philadelphia,
Pennsylvania, 1986. pp. 35-42.
Bryden, G.W., W.R. Mabey, and K.M. Robine. Sampling for Toxic Contaminants in Ground Water. Ground
Water Monitoring Review, 6(2):67-72,1986.
Campbell, M.D., and J.H. Lehr. Water Well Technology. McGraw-Hill, New York, New York, 1973. 681 pp.
Campbell, J.A., and W.R. Mabey. A Systematic Approach for Evaluating the Quality of Ground Water
Monitoring Data. Ground Water Monitoring Review, 5(4):58-62, 1985.
Cardin, C.W., J.E. Moore, and J.M. Rubin. "Water Resources Division in the 1980's." A Summary of Activities
and Programs of the U.S. Geological Survey's Water Resources Division 1986, tJSGS Circular 1005,
1986. 79pp.
COM Federal Programs Corporation. Data Quality Objectives for Remedial Response Activities; Development
Process. EPA/540/G-87/003, U.S. Environm. Protection Agency, Off. of Solid Waste and Emergency
Response, Washington, D.C., 1987.
122
-------
Cherry, J.A. (ed). Migration of Contaminants in Groundwater at a Land Fill: A Case Study. Journal of
Hydrology, 63(1/2): 1983.
Chu, W., E.W. Strecker, and D.P. Lettenmaier. An Evaluation of Data Requirements for Groundwater
Contaminant Transport Modeling. Water Resources Research, 23(3):408-424,1987.
Clarke, D. Microcomputer Programs for Groundwater Studies. Elsevier Science Publishers, Amsterdam, The
Netherlands, 1987.
Collins, A.G., and M.E. Crocker. Laboratory Protocol for Determining Fate of Waste Disposed in Deep Wells.
EPA Project Summary, EPA/600/S8-88/008, U.S Environmental Protection Agency, CERI, Cincinnati,
Ohio, 1988.
Collins, A.G., and A.I. Johnson (eds.). Ground-Water Contamination: Field Methods. ASTM STP 963, Am.
Soc. for Testing and Materials, Philadelphia, Pennsylvania, 1988.
Committee on Data for Science and Technology (CODATA). International Council of Scientific Unions,
Project Report, Paris, France, 1982.
Cordes, E.H. Ground Water Instrumentation Program of the U.S. Geological Survey. Ground Water
Monitoring Review, 4(4):103-114, 1984.
Crocker, M.E., and LM. Marchin. Adsorption and Degradation of Enhanced Oil Recovery Chemicals. In:
A.G. Collins and A.I. Johnson (eds.), Ground-Water Contamination: Field Methods. ASTM STP 963,
Am. Soc. for Testing and Materials, Philadelphia, Pennsylvania, 1988. pp. 358-369.
Curtis, G.P., P.V. Roberts, and M. Reinhard. Sorption of Organic Solutes and Its Influence on Mobility. In:
P.V. Roberts and D.M. Mackay (eds.), A Natural Gradient Experiment on Solute Transport in a Sand
Aquifer, Technical Report 292, Department of Civil Engineering, Stanford University, Stanford,
California, 1986. pp. 124-144.
Date, C.J. An Introduction to Database Systems, Vol. 1. Addison-Wesley Publishing Co.
Davis, S.N., D.J. Campbell, H.W. Bentley, and T.J. Flynn. Groundwater Tracers. National Water Well
Association, Worthington, Ohio, 1985.
Driscoll, F.G. Ground Water and Wells. 2nd Edition, Johnson Division, Saint Paul, Minnesota, 1986. 440
pp.
Dunbar, D., H. Tuchfeld, R. Siegel, and R. Sterbentz. Ground Water Quality Anomalies Encountered During
Well Construction, Sampling and Analysis in the Environs of a Hazardous Waste Management Facility.
Ground Water Monitoring Review, 5(3):70-74,1985.
123
-------
Dunlap, W.J.. J.F. McNabb, M.R. Scalf, and R.L Cosby. Sampling for Organic Chemicals and
Microorganisms in the Subsurface, EPA 600/2-77-176, U.S. Environmental Protection Agency, R.S.
Kerr Environmental Research Laboratory, Ada, Oklahoma, 1977.
Edwards, M.D. NAWDEX: A Key to Finding Water Data. U.S. Geological Survey, National Water Data
Exchange, Reston, Virginia, 1978. 15 pp.
Edwards, M.D. Water Data and Information Exchange Programs of the U.S. Geological Survey. Proceedings
of Conference on Improving Access to On-line Water Information, Cornell University, Center for
Environmental Research, Ithaca, New York, 1983. pp. 30-39.
Edwards, M.D., A.L. Putnam, and N.E. Hutchinson. Conceptual Design for the National Water Information
System. U.S. Geological Survey Bulletin 1792, Washington, D.C., 1987.
Eggenberger, L.M. Establishing an Environmental Quality Assurance Program. In: J.K. Taylor and T.W.
Stanley (eds.), Quality Assurance for Environmental Measurements, ASTM STP 867, Am. Soc. for
Testing and Materials, Philadelphia, Pennsylvania, 1985. pp. 391-406.
Electric Power Research Institute (EPRI). Field Evaluation of Instruments for the Measurement of
Unsaturated Hydraulic Properties of Fly Ash, EPRI EA-5011, Palo Alto, California, 1987. 86 pp.
Environmental Systems Research Institute (ESRI). Final Report, San Gabriel Basin Geographic Information
System. U.S. Environmental Protection Agency, Environmental Monitoring Systems Laboratory, Las
Vegas, Nevada, 1986.
Everett, LG. Monitoring in the Vadose Zone. Ground Water Monitoring Review, 1(2):44-51, 1981.
Everett, L.G., E.W. Hoylman, LG. Wilson, and LG. McMillion. Constraints and Categories of Vadose Zone
Monitoring Devices. Ground Water Monitoring Review, 4(1):26-32, 1984b.
Everett, LG., LG. Wilson, and E.W. Hoylman. Vadose Zone Monitoring for Hazardous Waste Sites.
Pollution Technology Review No. 112, Noyes Data Corporation, Park Ridge, New Jersey, I984a. 358
pp.
Fang, H.Y., and J.C. Evans. Long-Term Permeability Tests Using Leachate on a Compacted Clayey Liner
Material. In: A.G. Collins and A.I. Johnson, eds. Ground-Water Contamination: Field Methods. ASTM
STP 963, Am. Soc. for Testing and Materials, Philadelphia, Pennsylvania, 1988. pp. 397-406.
Ferris, J.G., D.B. Knowles, R.H. Browne, and R.W. Stallman. Theory of Aquifer Tests. Water-Supply Paper
1536-E, U.S. Geological Survey, Reston, Virginia, 1962.
Fetter, C.W., Jr. Potential Sources of Contamination in Ground-Water Monitoring. Ground Water Monitoring
' Review, 3(2):60-64, 1983.
124
-------
Finkelstein, P.L, D.L Mazzarella, T.J. Lockhart, W.J. King and J.H. White. Quality Assurance Handbook for
Air Pollution Measurement Systems, Volume IV: Meteorological Measurements, EPA 600/4-82-060,
Office of Air and Radiation, Research Triangle Park, North Carolina, 1983.
Fienberg, S.E., M.E. Martin, and M.L Straf, eds. Sharing Research Data. National Academy Press,
Washington, D.C., 1985. 240 pp.
Flatman, G.T. Design of Soil Sampling Programs: Statistical Considerations. In Quality Control in Remedial
Site Investigation—Hazardous and Industrial Solid Waste Testing, Gary L Perket (ed.), ASTM STP
925, Philadelphia, 1986. pp. 43-56.
Francis, C.W.. M.P. Maskarinec, and D.W. Lee. Physical and Chemical Methods for the Characterization of
Hazardous Wastes, Oak Ridge National Lab., Tennessee, 1988. 27 pp.
Freeze, R.A., and J.A. Cherry. Groundwater. Prentice Hall, Inc., Englewood Cliffs, New Jersey, 1979. 604
PP-
Freyberg, D.L. Spatial Moments and the Advection and Dispersion of Non-Reactive Tracers. In: P.V.
Roberts and D.M. Mackay, eds., A Natural Gradient Experiment on Solute Transport in a Sand Aquifer,
Technical Report 292, Department of Civil Engineering, Stanford University, Stanford, California, 1986.
pp. 50-92.
Friedman, D. Waste Testing and Quality Assurance. ASTM STP 999, Am. Soc. for Testing and Materials,
Philadelphia, Pennsylvania, 1988.
Friedman, L.C., LJ. Schroder, and M.G. Brooks. Recovery of Several Volatile Organic Compounds from
Simulated Water Samples: Effect of Transport and Storage. Environ. Sci. Techno!., 20(8):826-829,
1986.
Frind, E.O., and G.E. Hokkanen. Simulation of the Borden Plume Using the Alternating Direction Galerkin
Technique. Water Resources Research, 23(5) :918-930,1987.
Fritterman, D.V. Examples of Transient Sounding for Ground-Water Exploration in Sedimentary Aquifers.
Ground Water, 25(6):685-€92, 1987.
Gaspar, E. (ed.). Modern Trends in Tracer Hydrology, 2 volumes, CRC Press, Inc., Boca Raton, Florida,
1987. 312 pp.
Gelhar, L.G. Stochastic Prediction of Dispersive Transport at Hazardous Waste Sites. U.S. Environmental
Protection Agency. R.S. Kerr Environmental Research Laboratory, Ada, Oklahoma, 1986.
Hammill, L, and F.G. Bell. Groundwater Resource Development. Butterworths, London, UK, 1986.
125
-------
Hern, S. and S. Melancon. Vadose Zone Modeling of Organic Pollutants. Lewis Publishers, Inc., Chelsea,
Michigan, 1986.
Hillman, J.J. Certification of Local Laboratories Analyzing Drinking Water in Nonprimacy States. In: J.K.
Taylor and T.W. Stanley (eds.), Quality Assurance for Environmental Measurements, ASTM STP 867,
Am. Soc. for Testing and Materials, Philadelphia, Pennsylvania, 1985. pp. 103-109.
Hoffman, J.L A Computer Databse for Overview of Ground Water Pollution Investigations. Ground Water
Monitoring Review, 6(1):76-79, 1986.
Hopke, P.K., and D.L. Massart. Environmental Data Management; Directions for Internationally Compatible
Environmental Data. Proceedings CODATA Workshop, International Council of Scientific Unions, Paris,
France, 1986.
Huyakorn, P.S., A.G. Kretschek, R.W. Broome, J.W. Mercer, and B.H. Lester. Testing and Validation of
Models for Simulating Solute Transport in Ground-Water. GWMI 84-13, International Ground Water
Modeling Center, Holcomb Research Institute, Butler University, Indianapolis, Indiana, 1984.
HYDROCOIN. Progress Report No. 4., January - August 1986. Swedish Nuclear Power Inspectorate,
Stockholm, Sweden, 1986.
HYDROCOIN. The International HYDROCOIN Project; Background and Results. Nuclear Energy Agency,
Organisation for Economic Co-operation and Development (OECD/NEA) Paris, France, 1987.
INTRACOIN. Final Report Levels 2 and 3, model validation and uncertainty analysis. Report SKI 86-2,
Swedish Nuclear Power Inspectorate, Stockholm, Sweden, 1986.
Jenkins, T.F., and P.W. Schumacher. Comparison of Methanol and Tetraglyme as Extraction Solvents for
Determination of Volatile Organics in Soil. CRREL-SR-87-22, Cold Regions Research and Engineering
Lab., U.S. Army Corps of Engineers, Hanover, New Hampshire, 1987. 36 pp.
Johnson, G.L, J.S. Ford, and LE. Michalec. PC-QTRAK: An Automated Tracking System for Environmental
Quality Assurance Activities. EPA/600/D-87/138, U.S. Environmental Protection Agency, Air and
Energy Engineering Laboratory, Research Triangle Park, North Carolina, 1987.
Kaplan, E., J. Naidu, M. Hauptman, and A. Meinhold. 1985. Guidebook for the Assembly and Use of Diverse
Ground Water Data. BNL-37356, Brookhaven Nat. Lab., Dept. of Applied Science, Brookhaven, New
York, 1985.
Kashef, A-A. I. Groundwater Engineering. Mc-Graw Hill Inc., New York, New York, 1986.
Keely, J.F. Chemical Time-Series Sampling. Ground Water Monitoring Review, 2(4):29-38,1982.
126
-------
Keith, L.H., W. Crumlett, J. Deegan, Jr., RA Libby, J.K. Taylor, and G. Wentler. Principles of Environmental
Analysis. Analytical Chemistry, 55:2210-2218,1983.
W.K. Kellogg Biological Station. Data Management at Biological Field Stations. Workshop Report, prepared
for National Science Foundation, Michigan State University, Hickory Corners, Michigan, 1982. 46 pp.
Kerfoot, W.B. Darcian Flow Characteristics Upgradient of a Kettle Pond Determined by Direct Ground
Water Flow Measurement. Ground Water Monitoring Review, 4(4):188-192, 1984.
Kerfoot, W.B. Monitoring Well Construction, and Recommended Procedures for Direct Ground-Water Flow
Measurements Using a Heat-Pulsing Flowmeter. In: A.G. Collins and A.I. Johnson, eds., Ground-
Water Contamination: Field Methods. ASTM STP 963, Am. Soc. for Testing and Materials, Philadelphia,
Pennsylvania, 1988. pp. 146-161.
Kerfoot, W.B., and V.A. Massard. Monitoring Well Screen Influences on Direct Flowmeter Measurements.
Ground Water Monitoring Review, 5(4):74-77, 1985.
Keys, W.S., and L.M. MacCary. Application of Borehole Geophysics to Water Resources Investigations.
U.S.G.S. Techniques of Water Resources Investigations, Book 2, Chapter E1, U.S. Geol. Survey,
Washington, D.C., 1971.
Kroenke, D. Database Processing. Science Research Associates, Inc., Chicago, Illinois, 1977.
Kruseman G.P., and N.A. De Ridder. Analysis and Evaluation of Pumping Test Data. Bulletin 11, Internat.
Inst. for Land Reclamation and Improvement, Wageningen, The Netherlands, 1970.
Leap, D.I. A Simple Pneumatic Device and Technique for Performing Rising Water Level Slug Tests. Ground
Water Monitoring Review, 4(4):141-146, 1984.
Leap, D.I., and P.G. Kaplan. A Single-Well Tracing Method for Estimating Regional Advecth/e Velocity in a
Confined Aquifer: Theory and Preliminary Laboratory Verification. Water Resources Research,
24(7):993-998, 1988.
Locke, J.W. Laboratory Accreditation. Environmental Science and Technology, 21(4):332-333, 1987.
Lockhart, T.J. Quality Assurance of Meteorological Measurements. In: J.K. Taylor and T.W. Stanley, eds.,
Quality Assurance for Environmental Measurements, ASTM STP 867, Am. Soc. for Testing and
Materials, Philadelphia, Pennsylvania, 1985. pp. 253-259.
Lorenzen, D., R.A. Conway, LP. Jackson, A. Hamza, C.L Perket, and W.J. Lacy, eds. Hazardous and
Industrial Solid Waste Testing and Disposal; 6th Volume. ASTM STP 933. Am. Soc. for Testing and
Materials, Philadelphia, Pennsylvania, 1986.
127
-------
Mackay, D.M., W.P. Ball, and M.G. Durant. Variability of Aquifer Sorption Properties in a Field Experiment
on Groundwater Transport of Organic Solutes: Methods and Preliminary Results. J. of Contaminant
Hydrology, 1:119-132,1986a.
Mackay, D.M., D.L Freyberg, and P.V. Roberts. A Natural Gradient Experiment on Solute Transport in a
Sand Aquifer, 1. Approach and Overview of Plume Movement. Water Resources Research,
22(13) :2017-2029, 1986b.
Marble, D.F., and D.J. Peuquet. Geographic Information Systems and Remote Sensing. In: Manual of
Remote Sensing, 2nd Edition, Volume I., American Society of Photogrammetry, Falls Church, Virginia,
1983. pp. 923-958.
Mercer, M.W., and C.O. Morgan. Storage and Retrieval of Ground-Water Data at the U.S. Geological Survey.
Geological Survey Circular 856, Reston, Virginia, 1982. 9 pp.
Mercer, J.W., S.D. Thomas, and B. Ross. Parameters and Variables Appearing in Repository Siting Models.
NUREG/CR-3066, U.S. Nuclear Regulatory Commission, Washington, D.C., 1982.
Melville, J.G., F.J. Molz, and 0. Guven. Laboratory Investigation and Analysis of a Groundwater Flowmeter.
Ground Water, 23(4):486-495, 1985.
Merrill, D. Practical Applications of Integrated Data Approaches. In Proceedings of the 1983 Integrated Data
Users Workshop. ORNL, CONF-831117, Oak Ridge National Laboratory, Oak Ridge, Tennessee,
1985. 94 pp.
Meyer, P.O., and E.D. Brill, Jr. A Method for Locating Wells in a Groundwater Monitoring Network under
Conditions of Uncertainty. Water Resources Research, 24(8): 1277-1282,1988.
Mills, W.B., D.B. Porcella, M.J. Ungs, S.A. Gherini, K.V. Summers, D.A. Haith, L Mok, G.L Rupp, and G.L
Bowie. Water Quality Assessment: A Screening Procedure for Toxic and Conventional Pollutants-Part
II. EPA/600/6-85/002b. U.S. Environmental Protection Agency, Cincinnati, Ohio, 1985.
Mitchell, P.J., R.S. Argo, S.L Bradymire, and C.A. Newbill. Hanford Ground-Water Data Base Management
Guide and User's Manual. Battelle Pacific Northwest Laboratory, Richland. Washington, 1985.
Moffett, T.B. Where Have All the Ground Water Data Gone—Into the Files Never to Return? In Proceedings
of the Sixth National Ground Water Quality Symposium, Sept. 22-24,1982, Atlanta, Georgia. National
Water Well Association, Dublin, Ohio, 1983. pp. 116-126.
Molz, F.J., J.G. Melville, 0. Guven, R.D. Crocker, and K.J. Matteson. Design and Performance of Single Well
Tracer Tests at the Mobile Site. Water Resources Research. 21(10): 1497-1502, 1985.
128
-------
Montgomery, R.H., and T.G. Sanders. Uncertainty in Water Quality Data. In: A.M. El-Shaarawi and R.E.
Kwiatkowski (eds), Statistical Aspects of Water Quality Monitoring, Elsevier, Amsterdam, The
Netherlands, 1986. pp. 17-29.
Moran, M.S., and LJ. Mezgar. Evaluation Factors for Verification and Validation of Low-Level Waste Disposal
Site Models. DOE/OR/214QO-T119, Oak Ridge National Laboratory, Oak Ridge, Tennessee, 1982.
Murarka, IP., and D.A. Mclntosh. Solid-Waste Environmental Studies (SWES) Description, Status, and
Available Results. EPRI-EA-5322-SR, Electric Power Research Inst., Palo Alto, California, 1987.
NWQL (National Water Quality Laboratory). Quality Assurance in the National Water Quality Laboratory,
NWQL and Canada Centre for Inland Waters, Burlington, Ontario, Canada, 1986. 98 pp.
NWQL (National Water Quality Laboratory). Brochure, Canada Centre for Inland Waters, Burlington, Ontario,
1986.
Nacht, S.J. Ground-Water Monitoring System Considerations. Ground Water Monitoring Review, 3(2):33~39,
1983.
National Research Council (NRC). Sharing Research Data: A Scientific Responsibility. News Report
(6): 15-18, 1985.
National Water Quality Laboratory (NWQL). Quality Assurance in the National Water Quality Laboratory.
Environment Canada, Canada Centre for Inland Waters, Burlington, Ontario, Canada, 1986.
Nebraska Natural Resources Commission (NNRC). Nebraska Natural Resources Data Bank Information
System, Data Access Manual, 2nd ed. Nebraska Natural Resources Commission, Lincoln, Nebraska,
1981. 348pp.
Nelson, M.R. Index to EPA (Environmental Protection Agency) Test Methods. EPA/901/3-88/001, U.S.
Environm. Protection Agency, Region 1, Boston, Massachusetts, 1988.
Nelson, J.D., and R.C. Ward. Statistical Considerations and Sampling Techniques for Ground-Water Quality
Monitoring. Ground Water, 19(6):617-625,1981.
Nicholson, T.J., T.J. McCartin, P.A. Davis, and W. Beyeler. NRC Experiences irj HYDROCOIN: An
International Project for Studying Ground-Water Row Modeling Strategies. Proceedings GEOVAL 87,
Stockholm, Sweden, April 4-9,1987. Swedish Nuclear Power Inspectorate, Stockholm, Sweden, 1987.
Nielson D.M. Developing Standards for Ground Water Monitoring-lmproving the State-of-the-Science.
Editorial. Ground Water Monitoring Review, 7(4):4-5,1987.
129
-------
Nimmo, J.R.. J. Rubin, and D.P. Hammermeister. Unsaturated Flow in a Centrifugal Field: Measurement of
Hydraulic Conductivity and Testing of Darcy's Law. Water Resources Research, 23(1): 124-134.1987.
Office of Information and Regulatory Affairs (OIRA). Procedure for Preparation of Abstracts for Public Use
of Statistical Machine-Readable Data Files. Proposed Voluntary Standard. Office of Management and
Budget, Washington, D.C., 1983. 49 pp.
Olhoeft, G.R. Geophysics Advisor Expert System. EPA/600/X88/257. U.S. Environm. Protection Agency,
Environm. Monitoring Systems Lab., Las Vegas, Nevada, 1988.
Olsen, H.W., T.L Rice, and R.W. Nichols. Measuring Effects of Permeant Composition on Pore-Fluid
Movement in Soil, In: A.G. Collins and A.I. Johnson, eds., Ground-Water Contamination: Field Methods.
ASTM STP 963, Am. Soc. for Testing and Materials, Philadelphia, Pennsylvania, 1988. pp. 331-342.
Olson, R.E., and D.E. Daniel. Measurement of the Hydraulic Conductivity of Fine-Grained Soils In: T.F.
Zimmie and C.O. Riggs (eds.), Permeability and Ground Water Contaminant Transport, ASTM STP 746,
Am. Soc. for Testing and Materials, Philadelphia, Pennsylvania, 1981. pp. 18-64.
Olson, R.J. Review of Existing Environmental and Natural Resources Data Bases. ORNL/TM-8928, Oak
Ridge National Laboratory, Oak Ridge, Tennessee, 1984.
Olson, R.J., and NT. Millemann, eds. Proceedings of the 1983 Integrated Data Users Workshop. USGS
National Center, Reston, VA. ORNL, CONF-831117, Oak Ridge National Laboratory, Oak Ridge,
Tennessee, 1985. 94 pp.
Parker, D.G., and S.C. Parker. Local Water Resource Information Management System. Publication No. 120,
Arkansas Water Resources Research Center, University of Arkansas, Fayette, Arkansas, 1986.
Paulsen, S.G., C.L Chen, K.J. Stetzenbach, and M.J. Miah. Guide to the Application of Quality Assurance
Data to Routine Survey Data Analysis. EPA/600/4-88/010, U.S. Environm. Protection Agency,
Environm. Monitoring Systems Lab., Las Vegas, Nevada, 1988.
Pearsall, K., and D.A.V. Eckhardt. Effects of Selected Sampling Equipment and Procedures on the
Concentration of Tricloroethylene and Related Compounds in Groundwater Samples. Ground Water
Monitoring Review, 7(2):64-73, 1987.
Perket, C.L, ed. Quality Control in Remedial Site Investigation: Hazardous and Industrial Solid Waste
Testing; 5th Volume. ASTM STP 925, Am. Soc. for Testing and Materials, Philadelphia, Pennsylvania,
1986.
Powers, J.P. Construction Dewatering. John Wiley & Sons, New York, New York, 1981.
130
-------
Roberts, P.V., M.N. Goltz, and D.M. Mackay. Retardation Estimates and Mass Balances for Organic Solutes.
In: P.V. Roberts and D.M. Mackay, eds., A Natural Gradient Experiment on Solute Transport in a
Sand Aquifer, Technical Report 292, Department of Civil Engineering, Stanford University, Stanford,
California, 1986. pp. 93-111.
Roberts, P.V., and D.M. Mackay, eds. A Natural Gradient Experiment on Solute Transport in a Sand Aquifer.
Technical Report 292, Department of Civil Engineering, Stanford University, Stanford, California, 1986.
153 pp.
Ross, D., D. Sharma, and W.W. Shuster. Modeling the Fate of Toxic Organic Materials in Aquatic
Environments. EPA-600/S3-82-028, U.S. Environmental Protection Agency, Environmental Research
Laboratory, Athens, Georgia, 1982.
Scalf, M.R., J.F. McNabb, W.J. Dunlop, R.L Cosby, and U.S. Fryberger. Manual of Ground-Water Quality
Sampling Procedures. National Water Well Association, Dublin, Ohio 1981. 93 pp.
Schiegg, H.O., and J.F. McBride. Laboratory Setup to Study Two-Dimensional Multiphase Flow in Porous
Media. PNL-SA-14988, Battelle Pacific Northwest Lab., Richland, Washington, 1987.
Schweitzer, G.E., and S.C. Black. Monitoring Statistics. Environmental Science and Technology,
19(11): 1026-1030, 1985.
Semprini, L, P.V. Roberts, G.D. Hopkins, and D.M. Mackay. A Field Evaluation of In-Situ Biodegradation
for Aquifer Restoration. EPA/600/2-87/096, U.S. Environm. Protection Agency, R.S. Kerr Env. Res.
Lab., Ada, Oklahoma, 1987. 53 pp.
Sexton, K.G., H.E. Jeffries, J.R. Arnold, T.L Kale, and R.M. Kamens. Validation Data for Photochemical
Mechanisms. EPA/600/3-87/003, Atmospheric Sciences Res. Laboratory, Research Triangle Park,
North Carolina, 1987.
Shacham, S., and H. McLellan. Abstracts of the Data Base—Wisconsin Power Plant Impact Study.
Preliminary draft, Environmental Research Laboratory, U.S. Environmental Protection Agency, Duluth,
Minnesota, 1979. 109 pp.
Shirley, P.A. Use of STORET as a Database for Ground-Water Quality Management. In: Proceedings of the
Sixth National Ground-Water Quality Symposium, Sept. 22-24, Atlanta, Georgia,'National Water Well
Association, Dublin, Ohio, 1982. pp. 131-141.
Slawson, G.C. Jr., K.E. Kelly, and LG. Everett. Evaluation of Ground Water Pumping and Bailing Methods-
Application in the Oil Shale Industry. Ground Water Monitoring Review, 2(l):27-32,1982.
Smart, P.L, and I.M.S. Laidlaw. An Evaluation of Some Fluorescent Dyes for Water Tracing. Water
' Resources Research, 13(1):15-33,1977.
131
-------
Smith, S., ed. Ground Water Hydrology for Water Well Contractors, National Water Well Association, Dublin,
Ohio, 1982. 288 pp.
Stanley, T.W., and S.S. Verner. The U.S. Environmental Protection Agency's Quality Assurance Program.
In: J.K. Taylor and T.W. Stanley, eds., Quality Assurance for Environmental Measurements, ASTM
STP 867, Am. Soc. for Testing and Materials, Philadelphia, Pennsylvania, 1985. pp. 12-19.
Steele, T.D. Strategies for Water Quality Monitoring. In: J.C. Rhoda, ed., Facets of Hydrology, Vol. II. John
Wiley & Sons, Chichester, United Kingdom, 1985.
Stierman, D.J., and LC. Ruedisili. Integrating Geophysical and Hydrogeological Data: An Efficient Approach
to Remedial Investigations of Contaminated Ground Water. In: A.G. Collins and A.I. Johnson, eds.,
Ground-Water Contamination: Field Methods. ASTM STP 963, Am. Soc. for Testing and Materials,
Philadelphia, Penn., pp. 331-342, 1988.
Sudicky, E.A., J.A. Cherry, and E.O. Frind. Migration of Contaminants in Groundwater at a Landfill: A Case
Study, 4. A Natural Gradient Dispersion Test. J. of Hydrology, 63:81-108, 1983.
Summers, K.V., and S.A. Gherini. Sampling Guidelines for Groundwater Quality. EPRI EA-4952, Electric
Power Research Institute, Palo Alto, California, 1987.
Sykes, J.F., S.B. Pahwa, R.B. Lantz, and D.S. Ward. Numerical Simulation of Flow and Contaminant
Migration at an Extensively Monitored Landfill. Water Resources Research, 18(6): 1687-1704, 1982.
Sykes, J.F., S.B. Pahwa, D.S. Ward, and R.B. Lantz. The Validation of SWENT, A Geosphere Transport
Model. In: R. Stepleman et al. (eds.), Scientific Computing. IMACS/North-Holland Publishing
Company, Amsterdam, The Netherlands, 1983.
Taylor, J.K. What is Quality Assurance?. In: J.K. Taylor and T.W. Stanley (eds.), Quality Assurance for
Environmental Measurements, ASTM STP 867, Am. Soc. for Testing and Materials, Philadelphia,
Pennsylvania, 1985. pp. 5-11.
Texas Natural Resources Information System (TNRIS). File Description Report, TNRIS-002, Texas Natural
Resources Information System, Austin, Texas, 1982. 33 pp.
Texas Natural Resources Information System (TNRIS). TNRIS-A Progress Report, Texas Natural Resources
Information System, Austin, Texas, 1983. 59 pp.
Tompson, A.F.B., and W.G. Gray. A Second-Order Approach for the Modeling of Dispersive Transport in
Porous Media, 3. Application to Two Porous Media Problems. Water Resources Research,
22(13):1959-1971, 1986.
132
-------
U.S. Department of Energy (DOE). Research Plan Subsurface Transport Program. DOE/ER-01563. Office
of Health and Environmental Research, Ecological Research Division, Washington, D.C., 1985. 43 pp.
U.S. Environmental Protection Agency (EPA). Quality Assurance Handbook for Air Pollution Measurement
Systems: Volume I - Principles. EPA 600/9-76-005, Office of Air and Radiation, Research Triangle
Park, North Carolina. 1976.
U.S. Environmental Protection Agency (EPA). Procedures Manual for Groundwater Monitoring at Solid
Waste Disposal Facilities. SW-611, Office of Solid Waste. Washington. D.C.. 1981.
U.S. Environmental Protection Agency and National Center for Ground Water Research (EPA/NCGWR).
A Ground Water Research Strategy. Ground Water Research Branch, R.S. Kerr Environmental
Research Laboratory, Ada, Oklahoma, 1982. 56 pp.
U.S. Environmental Protection Agency (EPA). Handbook for Sampling and Sample Preservation of Water
and Wastewater. EPA-600/4-82-029. Environm. Monitoring and Support Lab., Cincinnati, Ohio,
1982a.
U.S. Environmental Protection Agency (EPA). Guidelines for Assessing and Reporting Data Quality for
Environmental Measurements. Environm. Monitoring and Support Lab., Cincinnati, Ohio, 1983.
U.S. Environmental Protection Agency (EPA). Ground Water Protection Strategy. Office of Ground Water
Protection, Washington, D.C., 1984a. 73 pp.
U.S. Environmental Protection Agency (EPA). Review of the EPA Ground Water Research Program. Office
of the Administrator, Science Advisory Board, Washington, D.C., 1985a. 55 pp.
U.S. Environmental Protection Agency (EPA). Methods for the Storage of RCRA (Resource Conservation
and Recovery Act) Ground-Water Monitoring Data on STORET - Users Manual (Revised). Office of
Solid Waste, Washington, D.C., 1985b.
U.S. Environmental Protection Agency. Ground-Water Data Requirements Analysis for the Environmental
Protection Agency. Office of Ground-Water Protection and Office of Information Resources
Management, I987a.
U.S. Environmental Protection Agency (EPA). On-site Meteorological Program Guidance for Regulatory
Modeling Applications. EPA-450/4-87-013, Office of Air and Radiation, Research Triangle Park, North
Carolina, 1987b.
U.S. Geological Survey (USGS). National Handbook of Recommended Methods for Water-Data Acquisition,
Chapter 2: Groundwater (revised). Office of Water Data Coordination, Reston, Virginia, 1977. 177 pp.
133
-------
U.S. Geological Survey (USGS). Scientific and Technical, Spatial, and Bibliographic Data Bases and
Systems of the U.S. Geological Survey, 1979. Geological Survey Circular 817, Reston, Virginia, 1979.
U.S. Geological Survey. ESIS User Manual. USGS, Information Systems Division, Reston, Virginia, 1982.
van der Heijde, P.K.M. A.I. El-Kadi, and S.A. Williams. Groundwater Modeling: An Overview and Status
Report. GWMI 88-10, International Ground Water Modeling Center, Holcomb Research Institute,
Butler University, Indianapolis, Indiana, 1988.
van Ee, J.J., and LG. McMillion. Quality Assurance Guidelines for Ground-Water Investigations: The
Requirements. In: A.G. Collins and A.I. Johnson (eds.) Ground-Water Contamination: Field Methods.
ASTM STP 963, Am. Soc. for Testing and Materials, Philadelphia, Pennsylvania, 1988. pp. 27-34.
Violette, P. Surface Geophysical Techniques for Aquifer and Wellhead Protection Area Delineation.
EPA/440/12-87/016, U.S. Environm. Protection Agency, Office of Ground-Water Protection,
Washington, D.C., 1987.
Walton, W.C. Computer Programs for Pumping Test Analysis, Groundwater Pumping Tests; Design and
Analysis. Lewis Publishers, Inc., Chelsea, Michigan, 1987.
Way, S.C., C.R. McKee, and H.K. Wainwright. A Computerized Ground-Water Monitoring System. Ground
Water Monitoring Research, 4(1):21-25, 1984.
White, P.A. Measurement of Ground-Water Parameters Using Salt-Water Injection and Surface Resistivity.
Ground Water, 26(2): 179-186, 1988a.
Wilson, L.G. Monitoring in the Vadose Zone: A Review of Technical Elements and Methods. EPA
600/7-80-134, Environm. Monitoring Systems Lab., Las Vegas, Nevada, 1980.
Zimmie, T.F., and C.O. Riggs, eds. Permeability and Groundwater Contaminant Transport. ASTM STP 746,
Am. Soc. for Testing and Materials, Philadelphia, Pennsylvania, 1980.
Zohdy, A.A.R., G.P.Eaton, and D.R. Mabey. Application of Surface Geophysics to Ground-Water
Investigations. Techn. of Water Resourc. Investig., Chapter D1, Book 2, U.S. Geological Survey,
Washington, D.C., 1974.
134
-------
October 1988
APPENDIX A
SUMMARY OF "WORKSHOP ON THE ESTABUSHMENT OF A
GROUNDWATER RESEARCH DATA CENTER FOR
VALIDATION OF GROUNDWATER MODELS'
135
-------
SUMMARY OF "WORKSHOP ON THE ESTABLISHMENT OF A
GROUNDWATER RESEARCH DATA CENTER FOR VALIDATION OF GROUNDWATER MODELS"
August 9-10, 1988
1. INTRODUCTION
On August 9-10,1988, a workshop was held at the Holcomb Research Institute, Indianapolis, Indiana,
on the establishment of a groundwater research data center for validation of groundwater models. This
workshop is part of the EPA-funded two-year agreement with the Holcomb Research Institute, "Development
and Operation of a Groundwater Research Data Center for Model Validation," funded through the R.S. Kerr
Environmental Research Laboratory, Ada, Oklahoma. The workshop was attended by eleven persons,
representing three EPA environmental research laboratories (Ada, Las Vegas, and Athens), the USGS (local
and national), academic institutions, and Holcomb Research Institute.
The workshop goals were to discuss and evaluate Data Center procedures and the results of the work
performed under the two-year agreement, with special focus on the formal description of datasets, quality
assurance, the structure of the database (SATURN), and the role of research data in model validation. The
two-day workshop involved presentations, group discussions, and written comment on provided
questionnaires. A workshop notebook included graphs and peripheral information. Questions and
comments concerning all aspects of the Data Center were addressed during the workshop.
The following summary of the workshop describes the presentations given and the issues, concerns,
and suggestions expressed by the participants.
2. PURPOSE AND SCOPE
2.1. INTRODUCTORY COMMENTS (Paul van der Heijde)
Many researchers have found that data for the validation of complex models are not sufficiently
accessible. A recent EPA study (U.S. EPA 1987) revealed a need to improve access to groundwater data
f
and to lower the transaction costs associated with obtaining and using such data. The EPA study stressed
the need to coordinate and standardize the large volume of groundwater data (and its storage) presently
being generated and stored by many organizations in many different locations, files, and formats. The high
costs of field research studies make it vital to share data and avoid duplication of research (U.S. EPA 1985).
136
-------
To address these needs, the Holcomb Research Institute (HRI) is developing a Groundwater Research
Data Center, with the support of the U.S. Environmental Protection Agency. The Data Center activities are
taking place within the framework of the International Ground Water Modeling Center (IGWMC) and HRI's
Data Center. The main tasks of the Groundwater Research Data Center will be the collection, conditioning,
documentation, storage, and distribution of datasets resulting from research on groundwater pollution, and
the establishment of a referral service for those datasets not managed by the Data Center. The Center will
meet the need for better documentation of, information on, and access to existing groundwater research
data, and will concentrate on those datasets that might be useful for groundwater model validation studies.
2.2. DISCUSSION
The following issues and concerns were expressed by the workshop participants after the presentation
and in response to the questionnaire for this section of the workshop.
2.2.1. Model Validation
Model validation may be defined from many different viewpoints, e.g., management, legal, research,
and model development. Validation involves determining and evaluating the differences between model
predictions and field observations under a variety of conditions and stresses. It was pointed out that model
validation can be seen as an iterative process, or at least a process of continued learning during the course
of model development. Generally, a model developer will test the model to a certain extent and will often
claim partial validation. Many professionals feel that reliable field data are essential for model validation,
backed up by process-related laboratory data.
One practical purpose of model validation is to convince potential users that the performance of a
model is acceptable for the conditions tested. Furthermore, a model's validity is often construed on the
basis of the number of successful applications. Frequent use enhances credibility and reliability as model
users find existing coding "bugs" and as limitations in applying the code to various situations are explored.
Model development often takes a long time; consequently, continuity during this phase is often a
concern. Funding may be discontinued during the development process, personnel come and go, and
interest may wane or change focus. Frequently, model documentation is delayed because of poor planning
or a lack of funds allocated to the task. Due to these factors, a model often becomes available before its
validity is established.
To illustrate some of the obstacles delaying validation, the validation of MINTEQ (a hydrochemical
speciation model) was discussed. During the course of validating this model, the participants in the
137
-------
validation process were of different backgrounds (statisticians, geohydrdogists, laboratory bench scientists,
etc.) 'and were hard to convince of the model's overall validity because each was concerned with a different
aspect of Its performance. However, despite disagreement on the final validity or MINTEQ, it was generally
agreed that it is the best model currently available for this type of problem.
Management generally has a limited understanding of models and model validation issues. Therefore,
it would be useful to present a range of reviewed and tested models to assist in the analysis of management
alternatives. Management in turn should have procedures in place to assure that a model's application is
within the range of its established validity. For example, the USGS has a review system in place for
applications of models: any application model must be checked with the original developer or another
authority on that model. A "band of error" must be defined, meaning that the outcome of validation
calculations should be within a predefined range from independently measured values. It should be noted
that, although the USGS prefers not to use the term "validation," determining model reliability is a key
concern of the USGS.
2.2.2. Purpose of the Referral Database
A number of possible uses of the database are:
(1) datasets for model validation
(2) datasets as example for other types of field studies
(3) datasets for education in model use
(4) comparison studies of research datasets
(5) guidance for collection of research data.
The participants discussed and weighed the relative importance of each of these uses. The consensus
was that the Center should focus on data that can be used in relation to modeling activities. The use of
datasets as examples for other research data acquisition projects is specially important because of the lack
of good data collectors. It was noted that the first type of use should be the Center's primary focus, and
that uses 2 and 3 are of secondary importance. Point 4 was rated important by some who thought that it
might be useful for gathering statistical information. Most participants felt that the last type of use is least
important.
2.2.3 Sharing Research Datasets
Different views were expressed concerning the scientific value of the storage and hence third-party*
use of data. The sharing of datasets is essential for cooperative studies and will facilitate the testing of
138
-------
hypotheses, and will prevent unnecessary duplication of effort and redundant data. Workshop participants
agreed that these shared datasets will not lead to new groundwater research "breakthroughs," as these will
have already occurred in the initial studies.
2.3. CONCLUSION
The participants agreed that the Data Center can function, as part of the International Ground Water
Modeling Center, to facilitate and promote effective model validation and to improve understanding of the
limitations of field validation of models. Another role the Data Center might play is that of providing a
reference base for model parameters to be used in model application studies. Furthermore, the Center's
products might be useful for teaching purposes and as examples for future data studies. Although these
services and functions might be important spin-offs of the Center, its primary responsibility is providing a
referral service for datasets. Of lower priority is its role in the distribution of research datasets. Finally,
through its accumulated expertise, the Center might provide dataset analysis services to third parties. The
workshop participants felt that initially, the Center should focus on a national audience, leaving acquisition
of datasets from abroad for a later phase in its development.
3. DATA REQUIREMENTS FOR SELECTED GROUNDWATER MODELS
3.1. INTRODUCTORY COMMENTS (Stan Williams)
As the focus at the Data Center is on groundwater quality, an overview was given of (1) processes
that influence transport and fate of contaminants in groundwater, and (2) related data requirements.
Attention was given to factors that influence model selection, governing equations, mathematical solution
techniques, and specific data requirements for the various types of models. A discussion of model
limitations and current research developments was included.
Model selection is influenced by factors such as availability of the model, extent of its use, its user-
friendliness, hardware dependence, level of review and testing applied to the model, and availability and
quality of model documentation.
Generally, two governing equations are employed in a model for solute transport: one for groundwater
flow and one for solute transport. The principal parameters in the groundwater flow equation are the
hydraulic conductivity or transmissivity, specific storage or storat'rvity, and sometimes fluid or matrix
compressibility, fluid density, and dynamic viscosity. The solute transport equation usually includes terms
for advection, dispersion, and solute sinks or sources. Nonconservath/e transport of species is often
139
-------
controlled by chemical processes such as adsorption, ion exchange, dissolution-precipitation, and
biochemical transformation. Radioactive decay may take place. The parameters that enter into the solution
of a solute transport equation include groundwater velocity, longitudinal and transverse dispersivity,
distribution coefficient, and chemical half-life. Initial conditions for these solutions consist of values for the
respective dependent variables; i.e., piezometric surface or solute concentration. Boundary conditions for
the flow equation may include a combination of specified head, specified flux, and head-dependent flux
boundaries. Boundary conditions for transport are usually specified concentration, specified solute flux, and
concentration-dependent flux.
Some issues limiting the usability of available solute transport models include lack of flexibility for
spatial discretization, accuracy limitations caused by vertical averaging (used in many 2-dimensional models),
inability to handle multiple fluid problems adequately, and, in general terms, the lack of understanding of
basic processes related to describing the behavior of real-world groundwater systems. New developments
in groundwater transport modeling include modeling of multiphase systems, incorporation of biological
decay mechanisms in existing transport formulations, coupling of transport models with geochemical models,
refining the theory of dispersion, and increased capability in describing groundwater flow and transport in
variable saturated media and in fractured rock or macroporous soil.
To illustrate the type of data required, handouts were presented to workshop participants.
3.2. DISCUSSION
The following issues and concerns were expressed by the workshop participants in discussion after
the presentation and in response to the questionnaire for this section of the workshop.
3.2.1. Use of Available Data
There is a wide range of quality and testing of the many variables used in the more complex solute
transport models. To apply such models, investigators usually take "a potpourri of different data types"
from more than one source. Even for a single type of parameter, major differences exist in the quality of
data. For example, some equilibrium constants are NBS standards, while many others frequently used are
not (yet) accepted as a standard. Furthermore, there is a range of possible resolutions and accuracy in data
used to test models. This requires that data consistency must be addressed in the context of model
validation.
Data currently collected are not always suitable for model testing because of incompatible spatial or
temporal resolution, sufficient accuracy, or because the wrong type of data has been collected. In many
140
-------
cases, existing measurement technologies are inadequate to provide the necessary level of detail.
Recognizing the limitations posed by existing measurement technologies is important in determining the
usefulness of datasets for the purpose of model validation.
Currently, and in the future as more sophisticated models are developed, greater quantities of certain
data types will be needed in much more detail, these include data on adsorption and other geochemical
processes, biodegradation, multiphase flow and transport, and mineralogical and biological characteristics
of aquifer material. As an increased emphasis on risk assessment is foreseen, statistical descriptions of
system properties become increasingly important. Therefore, more attention should be given to the natural
spatial variability of parameters and characterization of correlation scale by increasing the spatial/temporal
resolution of data. In the meantime, despite these severe limitations, we must use existing or measurable
data for any validation.
3.2.2. Limitations in Field Measurement Technology
Theoretical and experimental scientists often work separately and isolated from each other. This
frequently causes problems, as system properties needed for a successful model aren't always specified
by the model developer. For example, some of the parameters required by theory are not accessible in
the field. Furthermore, problems may be encountered with the scale of measurements and the
dimensionality of the processes involved. It was stated during the workshop that an "overzealous" approach
often is taken to model development, and that in such cases, more planning and thought regarding the
development of parameter estimation methods is justified. Among other considerations, future data
collection technologies should focus on better control of the aquifer volume for which water samples are
collected and on better methods for in-situ measurement of chemical constituent concentrations.
4. CASE STUDY: HISTORY OF STANFORD/WATERLOO TRACER STUDY
AND DATASET FOR THE BORDEN SITE
4.1. INTRODUCTORY COMMENTS (Doug MacKay)
The workshop overview of the Stanford/Waterloo Tracer Study focused on data collection, project
documentation, and undocumented problems/snags during the project. Of particular of interest to
participants were the candid statements concerning the quality assurance/quality control (QA/QC) actually
applied. There was much concern during the design stage of the project about the loss of the volatile
solutes during tracer injection. One incident of poor sampling technique where the purging rate was too
high resulted in discarding 1600 samples. This error was detected during review of laboratory analysis
141
-------
results when low concentrations of the volatile organics were noticed. Another problem concerned the time
taken for sample labeling during the first sampling episode; as a result, a simpler labeling system was
adopted for the next sampling sessions.
As part of the project, twelve core samples were taken (described in a thesis by Meredith Deranth,
University of Waterloo) for experimental laboratory studies. Meanwhile, Barker and Patrick launched a BTX
experiment at the source site while the field coring for the tracer study was still in progress. As a
consequence, the coring sequence-timing was changed and delayed. The hydraulic conductivity values
used for the tracer study were taken from previously published studies by Sudicky (University of Waterloo).
Detailed head data were collected with the monitoring network but these were never published. Gary
Hopkins (Stanford University) was in charge of technical facilities needed for the project. All project data
was entered onto sheets, punched into the computer, and subsequently checked by Gary Hopkins. Stanford
University was also involved in other studies at the Borden Site. Publications relatinggarding to these
experiments are available through Paul Roberts of Stanford University. For the University of Waterloo
studies, Sudicky can be contacted.
In practice, field experiments are often done before one knows how to do them. To fulfill the EPA
QA project plan requirements, researchers often adopt an attitude of "you say what you need to say," but
during the actual study "you do what you need to do," as most data collection procedures in research
studies are still in an experimental stage or even under development as part of the study objectives.
4.2. DISCUSSION
The following suggestions, issues, and concerns were expressed by the workshop participants after
the presentation and in response to the questionnaire for this section of the workshop.
4.2.1 Information Usually Not Included in Documentation
Many aspects of field data collection projects are rarely included in the project reporting and QA.
Typical omissions are changes in approach, preliminary conclusions, changes in funding, changes in
workplan, changes in staff, a brief written chronology of project events, key contacts for future project
information, any unforseen project changes, site closures or other site activities, site anthropology, reasons
for project changes, previous contaminant exposures, actual QA/QC procedures, and explanation of how
the monitoring system was designed. Other omissions may include statistical techniques used for sampling
design, problems that may have occurred in sampling, previous experimental work in the area that might
have impacted physical properties, biases of the individuals performing the experiment, qualifications of the
142
-------
field team and laboratory analytical team, and lists of available project data. The type and amount of such
Information should be site- and project-specific.
4.2.2. Thoroughness in Reporting Field Activities
Workshop participants expressed a wide range of opinions concerning thoroughness in the reporting
of field activities. Most participants felt that it is generally easier to accept data from studies in which they
are personally involved, but that other datasets must have enough peripheral information to be acceptable
for their own use. Much crucial information is often lacking in the actual project reporting and data
documentation. As one participant said, "It is just not possible to include all of the potentially pertinent facts
in most publications when editors and reviewers are constantly seeking terseness," and that descriptive,
subjective information is often edited out. There was agreement that researchers must always be cautious
in using third-party data.
4.2.3. Difference between What Occurs in the Field and What Is Reported
EPA and other agencies have many requirements and guidelines for studies performed by or for them.
These requirements often include extensive QA procedures in preparation for administrative audits. QA is
the project manager's responsibility, starting with development of workplans and QA plans to be submitted
to the funding agency. However, during collection of field data, researchers often depart from these
stringent plans, and major differences frequently occur between planning and practice. This is not
necessarily bad, as field studies are conducted in order to allow unbiased observation of phenomena; if a
test is too constrained by expectations, observations may be of limited value and important phenomena
missed. In any case, such deviations from QA plans should be documented to allow assessment of the
reasons for and validity of the deviations, and thus the value of the data acquisition effort. Recognizing that
such deviations are probably more the rule than the exception in conducting field experiments, a practical
level of QA should be defined and field activities should be honestly reported.
5. IDENTIFICATION OF POTENTIAL AND EXISTING DATASETS
5.1. INTRODUCTORY COMMENTS (Rachel Miller)
Different types of field research studies, endeavoring to describe different processes, were presented
to workshop participants. A literature search was performed to gather information on datasets that might
be of interest for model validation purposes. Information obtained was entered into a preliminary database
containing approximately 150 such datasets. Several reports generated from the database were presented
to the workshop participants for review. Various searches were conducted, using this database to identify
143
-------
datasets with certain characteristics (e.g., fractured rock, tracer tests, certain organizations involved, and so
forth). At this stage of Data Center development, decisions must be made now to proceed with the
gathering of detailed information. Issues involved include the means by which additional information should
be collected, the considerations and criteria to be used in selecting datasets for incorporation in the Center's
database, and identification of other possible sources of information on datasets, sites, and research
projects. It should be noted that although a number of potentially interesting research sites have been
identified, little is known about the accessibility and availability of the actual datasets.
5.2. DISCUSSION
The following Issues, concerns, and suggestions were expressed by the workshop participants in
discussion after the presentation and in response to the questionnaire for this section of the workshop.
5.2.1. Field Datasets versus Laboratory Bench Datasets
Some participants expressed their opinion that the Data Center should not compile data resulting from
independent laboratory studies. As reasons for this advice, they pointed to the large number of laboratory
bench studies, the variable quality of these studies ("fraught with problems'), and that the experts in this field
are already aware of the really good datasets. However, many model developers have expressed their
interest in this type of data for model evaluation purposes. They contend that a model cannot be properly
validated without validation of its individual components. Laboratory datasets might prove useful for such
partial evaluation. It was generally concluded that laboratory bench data should have low priority; rather,
the Center's prime concern should be field research datasets followed by laboratory bench studies
connected to research sites. Initially, the Center should be concerned only with lab bench studies from high
quality field sites with which Center personnel are familiar.
5.2.2. Dataset Selection Criteria
It is important to establish criteria for both the selection of datasets, with which annotated descriptions
will be included in the referral database, and for datasets that will be distributed by the Center. It was
»'
suggested that the Center consider only field datasets that have been documented in peer-reviewed
publications. In an initial screening, the Center should select only those datasets that have been reviewed
by others, e.g., by such agencies as the USGS, having their own internal review procedures for data
collection.
144
-------
It is important to focus on those datasets collected primarily for examination of scientific hypotheses
(research data). Datasets for which no information is available on data quality should not be considered.
Time and resources should initially be concentrated on those sites on which the most comprehensive and
accessible field work have been performed and for which extensive and timely data are avaiable. If
collecting information on a specific site proves cumbersome, the effort should receive a lower priority.
Datasets already used in modeling studies and datasets with extensive first-glance data should be included.
The willingness of a data generator to have his/her data included in the referral database should be a major
requirement for the selection of datasets. However, datasets for selected distribution should pass certain
predefined QA criteria. In screening the datasets the Center should concentrate on the high quality material
that is too voluminous to be published in the peer literature. Logistically, dataset availability will be one of
the Center's first screening criteria.
5.2.3 Data Needed
Because each researcher or modeler has a special area of interest, a wide range of data types should
be available for groundwater quality model validation. Some of the data identified include natural and forced
gradient tracer tests, especially with organic or inorganic contaminants, and controlled experiments with
remedial action technology. The data needs of the model designed for such experiments will dictate the
type of data to be collected. Datasets should include groundwater elevations, spatial hydraulic conductivity
and transmissivity, well locations, pumpage rates, contamination source characterization, and porosity,
among other components. Of frequent interest is information on land use and population. Data resolution
and quantity will be a function of the scale of study.
5.2.4. Additional Existing Datasets
The following are datasets that the participants felt might be of interest for the validation of groundwater
models.
• Rocky Mountain Arsenal-current UCLA-conducted experiments; however, the data wffl not be
available for a while.
• University of Waterloo experiments on NAPL behavior in the subsurface
• Picatinny Arsenal, NJ-a pollution site where heavy metals, TCE, and other degreasers are present.
(Contact Tom Imbrogiotta, USGS District Office, NJ.)
• Wentsmith AF Base-a JP4 (jet fuel) spill (Contact T. Ray Cummings, USGS District Office, Ml.)
145
-------
• Hazardous Waste Ground-Water Task Force database-(EPA, Washington, DC) currently containing
data from about 58 RCRA facilities.
6. REFERRAL DATABASE STRUCTURE AND SEARCH CRITERIA CRITERIA
6.1. INTRODUCTORY COMMENTS (Wilbert Elderhorst)
If model testers/validators are to use the Center's referral database to identify datasets suited to their
particular purposes, the database must be well structured. Data entry, data retrieval, and maintainability and
portability of software and data content must all be considered. The database will be structured using
interrelated tables; its design will reflect efficient use of computer memory storage requirements and software
execution speed.
For information analysis and data entry purposes, forms will be designed containing all descriptors
identical to the database use interfaces. These forms will be partitioned into separate sections covering
different topical areas of field measurements, site characterization, and operational characteristics of the
datasets. This partitioning will facilitate collection of information scattered among various studies related
to the same site.
Major search strategies have been determined and considered in the design of the database.
Additional criteria are being used to develop dataset reports from the database information.
The referral database will contain information on a number of datasets collected at different
experimental sites. This information is collected using a standardized descriptive analysis and reporting
system focused on the major characteristics of the datasets (e.g., indicating which parameters are measured
or determined, what information is available concerning methods and equipment used, organizations
involved in the project data quality, geohydrological conditions present, etc.).
Each dataset will be described in a uniform way by a hierarchical set of descriptors, thus assuring
consistency among the dataset descriptions anticipated. The selection of descriptors is a subjective
process. The referral database should contain enough information to allow a searcher to determine the
usefulness of a dataset for a specific application. The database should not contain information that is not
crucial and that will be used only sparingly or not at all. It is evident that time and resources are related to
the complexity of the system designed.
146
-------
In a model validation example presented, it was shown that a well documented field experiment
(pumping test) may yield an elaborate and detailed list of descriptors. However, the use of such detailed
descriptors for all research data categories of potential interest would make the database too elaborate.
Therefore, limitations must be imposed and criteria chosen to describe the datasets efficiently. A well-
balanced list of descriptors for each of the specific research areas related to field experimentation on
groundwater quality is essential for effective dissemination of information to a wide variety of users.
As the Data Center becomes operational, the staff will regularly evaluate its progress. Based on
operational experience gained and user feedback expected, the database design may be modified. To
facilitate such maintenance, the database should have a flexible structure that is efficiently programmed
and well documented.
A draft dataset annotation format with about 400 descriptors was presented and comments were
solicited.
6.2. DISCUSSION
The following issues, suggestions, and concerns were expressed by the workshop participants after
the presentation and in response to the questionnaire for this section of the workshop.
6.2.1. Structure
The selection of the database descriptors and structure is intimately related to the main purpose of
the database. Therefore, establishing the level of detail for the descriptors follows from a careful analysis
of that purpose. Because input data are prerequisite to modeling, descriptors reflecting model data
requirements should receive major attention. Model developers need detailed, process-oriented information.
Local regulatory analysis requires GIS-type surveys; on a national level, there is a regulatory need for
characterization of uncertainty to be used in generic Monte-Carto-type simulations. With this information
present, the Center's referral database could be used by regulators to compare sites for similar
hydrogeologic environments in order to evaluate such issues as network design and/or remedial action
alternatives. However, model validation should be the primary purpose of the datasejs to be included.
In developing its database descriptors, the Center should be concerned with consistency in
terminology across disciplines and should consider standard data formats. It should address terminology
problems such as the existing disagreement over the use of such terms as "field capacity* or "water
retention." The same applies to the units used in different disciplines for the same parameters. Selection
147
-------
of units may involve the description of different techniques or procedures used in determining the values of
the specific parameter. The development and publication of a data dictionary would be helpful.
The complex datasets collected at many sites require a database structure that handles a variety of
data types such as the different bulk densities for different soil areas within a site, complex aquifer systems
varying in nature from site to site, or parameters whose values are by definition dependent on the method
of measurement.
In designing the database, a basic structure should be defined along with a top-level set of 15 to 20
basic descriptors. Further detailing of such levels can be based on feedback from users. The structure
should be easy to understand and flexible enough to facilitate future modifications; the level of detail must
not stand in the way of its main purpose. Database reporting should include a one-page abstract of the
researcher's original intent and use of the data. Descriptions of the region, type of site, and type of
contaminants included in the first general level of the database would be useful. The reports generated
should give all information of interest to the individual user and should be intelligible (not cryptic or in
symbolic language).
The design criteria for database structure should:
(1) achieve completeness with respect to relevant information
(2) ensure balance of information between sections
(3) facilitate efficient searches
(4) provide efficient storage (minimizing storage requirements)
(5) provide useful and complete reporting.
Doug Mackay proposed the following list of basic descriptors/categories as essential for the first level
of the database structure:
148
-------
General (abstract, original purpose, duration of
project)
Data Collectors
principal contacts)
organizations involved
ModeKs) tested or testable
fluid flow
solute transport/transformation
biologic exposure
chemical speciation
exposure/risk prediction
Site Location
area
general location
Site Characteristics
Climate
Surface
hydrology
geography
botany
Unsaturated zone - hydrogeology
Saturated zone
Primary (Field) Experimental Data
spatial network
parameters measured
sampling methods
analytical methods
QA/QC procedures
dataset description
Related (Field) Data
pump tests
geophysical methods
meteorological measurements
other
149
-------
The category for experimental data should include descriptors that give a general impression of the
dataset/site, including the water quality, engineering aspects, and any problems at the site.
The following are basic descriptors suggested for the laboratory datasets:
Single phase tests Process-related lab tests
2-phase system Property-related tests
partition coefficients
degradation rates, etc.
6.2.2. Level of Detail
The EPA Data Requirements study (U.S. EPA 1987) included a meeting to develop a "Required List
of Data Elements." Many iterations were needed to determine the most Important descriptors from a list
of thousands of suggested terms. A main guideline in determining which descriptors to use is that the
level of detail must be sufficient to quickly access datasets of interest.
Care must be taken so that the database will not exceed the "critical mass." An overly complex
database might cause problems such as lack of personnel familiar enough with the database to use it, user
manuals too complex to create and update, and certain basic information obscured by detail.
7. QA/QC: EVALUATION AND DOCUMENTATION OF DATASET QUALITY
7.1. INTRODUCTORY COMMENTS (Rachel Miller)
A major concern in the use of experimental data for model validation is data quality. There is need
for high quality datasets for use in model testing and validation. Quality assurance for groundwater data
collection projects is increasingly receiving attention. Of special importance is the long duration of these
projects and the large scale of other activities. The EPA requires QA project plans for'projects carried out
by or for them to be prepared during the design of the study. In collecting data, different QA approaches
are taken by the various organizations involved. These differences may be in the form of various standards
set and applied, different regulatory requirements for collection and documentation, or in technical guidance
provided. Currently, various organizations (e.g., ASTM) are formulating standards for groundwater field
measurements. Quantitative measures of data quality, such as precision, accuracy, and detection limits, are
used mainly In laboratory analysis. Data synthesis and reporting also may affect quality of reported data,
150
-------
and QA of these activities must be considered. (The Data Center is preparing a document, "Development
of Criteria for the Evaluation and Annotation of Research Datasets.") Examples of existing QA procedure
documentation were presented for various data acquisition activities including well logs, well construction
diagrams, and analyzed-synthesized data (aquifer tests, geophysical data). Points for further discussion are
how the Center should incorporate QA descriptors into the referral database and dataset distribution, and
what minimal level of QA/QC documentation is minimally required.
7.2. DISCUSSION
The following Issues, concerns, and suggestions were expressed by workshop participants in
discussion and in response to the questionnaire for this section of the workshop.
7.2.1. Evaluation of Dataset Quality
Many organizations such as the USGS and EPA already have extensive internal/external review
procedures in place to ensure the quality of research data collection and synthesis. The USGS also has
a standard procedure for the release of data; sometimes data are reviewed and modified after being
released. This is the case if a subsequent user detects inconsistencies; this would be of concern to the
Center, and a mechanism should be in place to keep track of such incidents. It was noted that a distinction
should be made between data quality for datasets in the referral database, and datasets distributed by the
Center. In the latter case, the additional QA should be applied by the Center. For datasets included in the
referral database, at least a rudimentary QA check should be performed. However, the user should have
the responsibility for determining if the data is of sufficient quality to meet the user's needs. The extent of
documentation available will be a major factor in a judgment of dataset quality. The database should contain
information concerning all procedures followed and equipment used. This is only feasible if these
procedures have been documented by the original researcher.
In evaluating the quality of datasets, the Center should be flexible, since many different levels of
QA/QC are practiced for collection of different data types. Above all, practicality and common sense should
be employed in establishing QA criteria. For example, a good way to test geophysical contractors is to have
a test course for geophysical measurements (e.g., buried targets). A test such as this may ensure better
t
quality than reviewing the documented use of Standard Operating Procedures (SOPS) by the field teams.
151
-------
7.2.2. Documentation of Dataset Quality
Workshop participants shared the opinion that a classification system using a limited number of data
quality levels should be employed in the referral database. The EPA discerns six levels of QA, from a basic
research category to a 16-point QA documentation (including chain of custody procedures). The flagging
of data elements should be considered and a system of qualitative descriptors (from rigorous to superficial
QA) adopted. Again it should be noted that QA classifications for the referral database differ from those
used in dataset distribution.
Somehow, the practice of field data collection should be evaluated by the Center. The Center should
collect information on the innovative, unplanned, and often undocumented procedures used during data
collection-especially when data collection is part of a research project. This information is crucial to
datasets planned for distribution by the Center. Although the database should not include too much detail
on data quality, information on the QA/QC is an essential part of the datasets to be distributed.
7.2.3. Field Notes and Forms
The Center should document field forms and procedures, but not field notes. It is more important to
document these for distributing high quality datasets, but is of lower priority for the referral database. A
method of evaluation might be to select some example forms and check them for accuracy and consistency.
The user could then track these forms down if needed. Field notes would not be of much value to the
Center; they are really of value only to the original researcher because the level of detail and the form in
which they normally exist require that the original researcher interpret them.
7.2.4. Field Practice
QA problems often occur in field projects; errors and omissions are made on a regular basis.
Difficulties inherent to field experimental work are frequently related to rush and fatigue (for example, the
1600-sample set that Mackay's group had to scrap). It is clear that actual data collection differs markedly
from the "standard practice" procedures described in text or workplans. Furthermore, there is a lack of
statistics on the amount and spatial arrangements of the data collected. It was agreed^hat more emphasis
should be placed on the QA/QC of data collection from the beginning of research activities, including the
statistics on the type of data collected.
152
-------
7.2.5. Role of the Center in QA/QC Practice
The Center might provide researchers/data collectors with guidelines for understanding QA concerns.
The Center could compare QA procedures in studies identified as generally good to procedures in
questionable studies; however, rigorous QA/QC may be too cumbersome, leading to few data and an
uninterpretable overall study. The Center could compare QA/QC plans with QA/QC actually done; as one
participant stated, "Perhaps we should stop the charade and home in on the really important matters."
In evaluating applied QA/QC, the Center might take an administrative view, focusing on datasets that
have been collected in projects that met formal EPA QA/QC requirements, or a more personal view,
providing information on the investigators, methods used, and past reputations among peers so that the
user could judge standards for the datasets.
8. DATA CENTER PROCEDURES
8.1. INTRODUCTORY COMMENTS (Wilbert Elderhorst)
In order to obtain effective and efficient operation of the Data Center, procedures should be established
ensuring proper database management, acquisition of data, distribution of data, and description of datasets.
Database management procedures should focus on assuring data integrity and security. This can be done
by adopting authorization rules specifying that certain tasks can be performed only by a specific group of
users. External users, for example, are not allowed to write or delete data from the database. Another
important procedure would require routine data backup, allowing recovery of the database in the event that
its content is corrupted.
Acquisition procedures must be established in order to obtain proper data transfer from data providers
to the Data Center. This involves the design and use of standard terminology and forms. The actual data
entry can be performed only by authorized Center personnel. Once all data that describe a dataset are
entered in the referral database, they should be checked for errors. Furthermore, the data transfer media
(floppy, disks, and tape), and optional reformatting procedures, must be determined. f
Documentation should describe various report-generating options available in the database system.
This would allow the user to obtain printed information from the referral database to the extent and detail
desired. In addition to printing procedures, additional information should be available concerning user
guides, availability of additional data, and reports describing the database structure.
153
-------
Distribution procedures should contain restrictions on distribution such as proprietary rights or
liabilities. When a dataset is accepted by the Data Center for distribution, these restrictions should be taken
into account to avoid future problems. The Data Center should acquire and distribute only datasets that are
public domain and do not have restrictions imposed on their general use.
8.2. DISCUSSION
The following issues, concerns and suggestions were expressed by the workshop participants in
discussion after the presentation and in response to the questionnaire for this section of the workshop.
8.2.1. Dataset Integrity
The incorporation of procedures for maintaining data integrity is essential. In the discussion, an
example was brought up in which certain data tapes were not decipherable. The tapes had been through
many nondocumented manipulations and their integrity had not been preserved; hence, the Center must
ensure that data is in its original form. A method for assuring integrity might be to compare the data on tape
with the original data by sending completed forms and data output back to the originator to check for error;
this can be done by computer, using existing software. Communication procedures should be established
between the data source and the Center, and these procedures should allow the Center's staff to discover
any inconsistencies in the data. In this context, problems within the STORET database system of EPA were
discussed, particularly in relation to ranges of pH-related ion concentration values found to be unreasonably
high.
These kinds of problems are comparable to those encountered by the IGWMC staff in the distribution
of models. In such cases the Center's policies require that, if a user complains, the Center analyzes the
problem and takes some course of corrective action, or takes the model out of distribution and informs the
user of specific problems.
It was suggested that initially, the buyer should check data integrity; but for primary datasets, those
that IGWMC judges to be of high quality, IGWMC should eventually summarize error-checking conducted
by data collectors and conduct its own error-checking. However, extensive evaluations are time-consuming
e
and require additional funding. For information entered in the referral database, or for datasets containing
information entered by IGWMC, error-checking should be mandatory.
154
-------
8.2.2. Services Offered by the Data Center
The Center would provide related services as funding or cost-recovery allows and as deemed
necessary for the distribution of datasets and for information concerning them. These services might include
guiding the deciphering of automated formats, providing a level of reliability for datasets, providing a
summary table of the dataset, and reformatting at cost. ASCII flat files will be available for most systems
without the need to reformat, and the data will be available on magnetic tape or diskette. It was generally
agreed that although interpretation of datasets is not the responsibility of the Center, use could be made of
standardized machine tests or screening (the Center must be cautious, as this approach has caused
problems in other systems such as STORET). It was thought that variability of different dataset parameters
would be of interest to data users. Summary tables could include extreme values, means, SDs, outliers,.
.. (ciding with QA claims). The Data Center should include user support and the Center should incorporate
procedures to keep track of the status, updating, and user needs relating to the datasets it refers or
distributes.
8.2.3. Cost of Services
Several logistical issues were identified in the areas of funding and data transfer. The Center needs
funds to distribute data efficiently and effectively, especially if extra formatting or documentation is required.
Ideally, the funding agencies for data collection should provide budgets for dataset formatting,
documentation, and preparation for distribution (as EPA did for the Stanford/Waterloo tracer study).
Datasets should be distributed at cost to the requestor (after formatting, and documentation by the Center).
If the original data-provider (e.g., a funding agency) subsequently requests the reformatted data, and when
this is not covered by an existing agreement, the provider must pay for these additional services.
8.2.4. Problems with Acquiring Datasets
Various problems can occur in acquiring datasets. Personal experiences with data availability and
accessibility involve problems with mixed formats, data "tied-up" in projects, and much data that is available
only in hard-copy form ("shoebox filing"). To obtain results from research studies, a secondary user is often
forced to spend significant effort identifying the source and in coercing the data custpdian to provide the
necessary information, often an exercise in sheer perseverance. For EPA projects, datasets should be
obtained through the project officer of the data acquisition study. A knowledge of the reporting mechanisms
of different agencies might be helpful in this respect.Datasets for model testing and validation often require
data from different sources, since most datasets are incomplete in themselves or are a part of a large group
effort.
155
-------
Researchers often are reluctant to share data because they want to make absolutely sure that it
contains no errors. "You must cover yourself completely before the release of data." After data is collected,
much effort is required to document, reformat, and acknowledge the data. There may be internal authorship
battles which inhibit the release of data and documentation. Furthermore, a researcher often wants to get
as much out of the data as possible. It is difficult to collect data and analyze it at the same time—the data
may not be available for a long time, as when they are not released for two to three years after their
collection. There are many ways that researchers can hang on to data; basically, "Researchers just aren't
going to give data until they are ready." Even after data is made available in principle, its accessibility may
be limited by any of the aforementioned factors.
For the referral database, initially qualified Center personnel should complete the forms using existing
dataset documentation, followed up by confirmatory interviews with or reviews by the data collectors. A
team approach will probably be required, including a database manager, hydrogeologist, and modeler. As
there are legal limitations on questionnaire mailings in studies performed with government funding, telephone
contacts may be the best way to perform these surveys.
9. SUMMARY OF THE DATA CENTER WORKSHOP
Workshop participants agreed that there might very well be a "cascade" of research datasets in the
future. Currently, many USGS research studies are being conducted at contamination sites, some in
cooperation with DOD (such as the Picatinny site). Also, DOE, NSF, EPA, and other agencies are funding
extensive field studies. It was estimated that the number of quality field datasets of eventual interest to the
Center might be on the order of 50-100.
The workshop provided an excellent opportunity for the Center's staff to clarify understanding of the
Center's objectives and the best ways to implement them. The workshop was held at a critical stage of
the project, thus providing the necessary feedback for the final design of the Center's operation. As the
Data Center becomes operational, information on new datasets will become available for the Center to
process. The Center must decide on a practical system of gathering site information and balancing these
activities with the resource made available to it.
Overall, the workshop members offered much encouragement toward the growth and future value of
the Center. The Data Center would have limited value as a separate entity, but its affiliation with the IGWMC
makes it viable and highly useful. Continued professional contacts with other organizations must be
maintained and pursued.
156
-------
The workshop served well in refocusing on problem areas in the Center's development. Many key
issues have been addressed and the Center's staff is highly motivated to continue its work.
10. REFERENCES
U.S. Environmental Protection Agency. Report on the Review of the Environmental Protection Agency's
Ground Water Research Program, Ground Water Research Review Committee, Science Advisory
Board. U.S. EPA. Washington, D.C., 1985.
U.S. Environmental Protection Agency (EPA). Ground-Water Data Requirements Analysis. EPA Office of
Ground-Water Protection, Washington, D.C., 1987.
157
-------
APPENDIX B
DATASET SURVEY - February 1989
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
735 MW MICHIGAN CITY
coal-fired utilities, groundwater
contamination
north-central Indiana
fly-ash, coal storage pile
leachates, heavy metals
dunal region
sand
DOE
Univ. of Notre Dame
2
SITE NAME: ALKALI LAKE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
chemical waste disposal
south-central Oregon
chlorophenolic compounds
fractured porous soil, playa
gravel, sand, silt, clay
EPA
OGC, Univ. of Waterloo
1
SITE NAME: ALLEN SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
fly ash disposal ponds
Gason County, North Carolina
fly ash leachate
weathered bedrock, intrusive dikes
sand, silt, diorite
EPRI, EPA
TetraTech Incorp., Arthur
D.Little(ADL)
-0-
158
-------
SITE NAME: ARMY CREEK
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
landfill leachate migration
Wilmington, Delaware
landfill leachate
-0-
-0-
USGS, EPA
USGS, Del. DNREC
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
ATLANTIC COAST
contamination of heterogeneous near
shore aquifer
Atlantic Coastal Plain of South
Carolina
Various
fluvial to marginal marine
sand, clay
DOE
Savannah River Lab
1
SITE NAME: BABYLON LANDFILL
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
landfill contaminant plume
south-east New York City, Long
Island
landfill leachate
glacial outwash plain
sand, gravel, silt
USEPA
USGS
3
SITE NAME: BAILLY GENERATING STATION
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
fly-ash ponds, dewatering plan
north-west Indiana, lakeshore
fly-ash pond leachate
Cowles Unit, dunal region
sand
Nat. Park Service(NPS)
NIPSCO, USGS
1
159
-------
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
BARNWELL SITE
evaluate potential for deep
percolation of radionuclides
Barnwell, South Carolina
radionuclides
-0-
-0-
USGS
USGS
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
BATTELLE NWLAB1
adsorption of aromatic nitrogen
bases in subsurface, laboratory
study
Battelle NW Labs
aromatic nitrogen bases
-0-
-0-
-0-
Battelle PNL
-0-
SITE NAME: BAYVIEW PARK
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
landfill, fractured porous bedrock
Burlington, Ontario
trichloroethane, landfill leachates
Queenston Fm.-Paleozoic
fractured shale
EPA, NSERC-Canada
OGC, Univ. of Waterloo
1
SITE NAME: BEALE AFB
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
monitoring of waste sites,
geophysical survey t
Marysville, California
varied wastes, fuels, landfills,
photo, battery
-0-
-0-
-0-
AeroVironment, Inc.
-0-
160
-------
SITE NAME: BEATTY SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
low-level radioactive waste site
Beatty, Nevada
low-level radioactive waste
alluvial fan, flood-plain depos
coarse-grained sand, gravel
-0-
USGS
1
SITE NAME: BEMIDJI
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
crude oil contamination in
groundwater, multi-phase
Bemidj i, Minnesota
crude oil, hydrocarbons
glacial outwash, till
sand, silt, clay, gravel
USGS
USGS
1
SITE NAME: BISCAYNE AQUIFER
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
9 hazardous waste sites on EPA
NPList
Florida
landfill and industrial waste,
hydrocarb, solvents, sewage
sedimentary sequence-Florida
limestone, sandstone
EPA-Super fund
EPA, CH2M Hill
-0-
161
-------
SITE NAME: BORDEN LANDFILL
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
contaminant plume, tracer tests
NW of Toronto
liquid radioactive waste,
(strontium)
Quaternary
sand, clay, silt
Nuc. Fuel Waste Man. Pr.
Chalk R. Nuc. Lab., Nat. Hy. Res.
Inst. Canada
-0-
SITE NAME: BORDEN SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
tracer test
NW of Toronto, Canada
chloride, bromide, 5 halogenated
organic chemicals
Quaternary
MF Sand
EPA-RSKERL
Stanford, Waterloo
9
SITE NAME: CAMP LEJUENE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
gas spill, Navy-Air Force Base,
microbial action
-0-
gas, hydrocarbons
-0-
-0-
-0-
Univ. of North Carolina
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
CANADA URL
in situ geotechnical experiments,
plutonic rock body, fract.flow
Lac du Bonnet, Manitoba
radioactive waste disposal
-0-
-0-
Can. Nuclear WMP
Atomic Energy of Canada Limited
99
162
-------
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
CANON CITY SITE
uranium ore processing, tailings
disposal, groundwater contamination
Canon City, Colorado
TDS, sulfate, molybdenum, uranium,
selenium
Tertiary, Cretaceous
alluv, SS, SH
USEPA
USGS, USNRC, Alliance Technologies
Corp.
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
CAPE COD
sewage treatment plant, colloidal
trans, of pollutants
Otis AFB, SE Massachusetts
sewage
Pleistocene
sand, gravel, silt, clay
USGS, MDEQEng., RSKERL
USGS
2
SITE NAME: CASTLEGUARD KARST
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
tracer study in karst hydrology
Alberta, Canada
-0-
-0-
-0-
-0-
Univ. of Western Ontario, London,
Ontario
-0-
163
-------
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
CEDARSAUK SITE
flyash landfill, groundwater
contamination
southeastern Wisconsin
sulfates, calcium, magnesium
glacial drift
sand, gravel, clay, dolomite
USDI-WR Cen., Uni. Wis
Univ. of Wisconsin, Wise. Elec.
Power Co.
-0-
SITE NAME: CENTER COAL SITES
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
environmental effects of fly ash
disposal and FGD wast
Center, North Dakota
fly ash, FGD waste
-0-
-0-
DOE
MMRRI
-0-
SITE NAME: CHEM-DYNE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
solvent reprocessing site
Hamilton, Ohio
solvents
glacial
fill, sand, silt, clay, till
EPA-Reg. 5-Chicago
EPA, RFWESTON Inc., E&E, CH2MHill,
OGC
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
CHROMIUM LABI
chemical attenuation of chromium in
soils, solution in groundwater
-0-
chromium
-0-
-0-
EPRI
EPRI
-0-
164
-------
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
CONCORD STATION
investigations for hazardous waste
disposal
Concord, California Naval Weapons
hazardous waste
-0-
-0-
-0-
Army Eng. Waterways Exp. Station,
Vicksburg, MS
-0-
SITE NAME: CONROE SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
aerobic biodegradation, creosote
plume
Conroe, Texas
chloride, organic compounds,
creosote
unconsolidated Pliocene
sand, sandy clay, clay
EPA, NCGWR, AMOCO Fnd.
Rice University
-0-
SITE NAME: CRATER LAKE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
hydrological effects of test
drilling
Winema National Forest, Crater Lake
drilling fluid/mud
-0-
-0-
-0-
LBL
-fl-
-------
SITE NAME: CREUX DE CHIPPIS SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED: -0-
NUMBER OF MODEL APPLICATIONS:
tracer study in a stony field soil
Sierre, Switzerland
chloride, bromide tracers
colluvial, glaciofluvial
sand, gravel, clay
Swiss Fed. Inst. Tech.
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
DAWSONVILLE SITE
fractured flow, contaminant
transport
Dawsonville, Georgia
-0-
-0-
-0-
EPA
Georgia State University, USGS
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
DEAF SMITH
candidate site for radioactive
waste disposal
Texas
radioactive waste
-0-
-0-
DOE
DOE
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
DOUGHERTY PLAIN
pesticides trans, and fate,
agricultural non-point source
Albany, Georgia t
pesticides, aldicarb, metachlor,
bromine
-0-
-0-
EPA
EPA-Athens, USGS
99
166
-------
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
ELRAMA SITE
scrubbing waste disposal, coal
strip-mine spoils
Washington County, west-central Pen
-0-
alluv., Pennsylvanian rocks
shale, siltstones
EPRI, EPA
Tetra Tech Incorp., Arthur D.
Little (ADL)
-0-
SITE NAME: ETIWANDA FIELD SITE
DESCRIPTION: unsaturated zone dispersion
experiment
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
-0-
tracers: chloride, boron, nitrate,
bromacil
field soil
-0-
SOURCE(S) OF FUNDING: EPA, EPRI
ORGANIZATIONS INVOLVED: EPRI, Univ. of CA Riverside
NUMBER OF MODEL APPLICATIONS: -0-
SITE NAME: EUROPE SOIL
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
nitrite migration through a loess
soil
northwestern Europe
nitrite
-0-
-0-
-0-
ORGANIZATIONS INVOLVED: Catholic Univ. of Louvain, Belgium
NUMBER OF MODEL APPLICATIONS: -0-
167
-------
SITE NAME: FANAY-AUGERES
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
uranium mine, nuclear waste storage
research, fracture flow
Limous in, France
radioactive waste disposal
-0-
granite
DOE, AFEE
LBL, French Bureaude Rechercles
Geologiqueset Minieres
1
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
FEDERAL PIONEER
chemical spill, transformer
manufacturing
Regina, Saskatchewan
Polychlorinated Biphenyls(PCBs),
TCB, multi-phase
recent unconsolidated, glacial,
Cret. bedrock
soil, clay, silt, till, sand, grav.
NRCC, Saskatchewan
Univ. of Waterloo, Univ. of Alberta
-0-
SITE NAME: FIELD COAL WASTES
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
study leaching of advanced coal
process waste
-0-
advanced coal process waste
-0-
-0-
DOE, METC
NDEMRC, MMRRI
99
168
-------
SITE NAME: FLY-ASH LABI
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
attenuation capacity, fly-ash
disposal site
North Dakota
fly ash, FGD waste
-0-
-0-
DOE
NDEMRC, MMRRI, Univ. of ND
-0-
SITE NAME: FLYING J REFINERY
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
hydrocarbon contain., assessment and
remediation
Williston, ND
hydrocarbons
-0-
-0-
-0-
NDEMRC, MMRRI, EES
1
SITE NAME: FORT UNION SITES
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
spoils from uraniferous lignite
mines
North Dakota
uranium in spoils piles
-0-
1ignite beds
-0-
USGS
1
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
FOSSUM SITE
evaluate brine and oil-gas drilling
fluid disposal site
North Dakota
oil brine, oil-gas drilling fluid
glacial till
-0-
NDWRRI
NDGS, Univ. of ND, MMRRI
1
169
-------
SITE NAME: FRESNO SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
aldicarb contamination, saturated
and unsaturated soil
central California
aldicarb
-0-
-0-
-0-
Union Carbide- Research Triangle
-0-
SITE NAME: GLIL YAM SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
anthropogenic anoxification of a
deep phreatic aquifer
15 km. north of Tel Aviv, Israel
oxygen, organics
Pleistocene, Pliocene
SS, siltstone, soils, clay
Nat. Council for R&D
Technion Israel Inst. of Tech.,
Weizmann Inst. of Science
-0-
SITE NAME: GLOBE LABI
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
lab column exper. in alluv., copper
mining contain.
Globe, Arizona
dissolved metals, sulfate
-0-
-0-
USGS
USGS-AZ-CO
NUMBER OF MODEL APPLICATIONS: 99
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMB.ER OF MODEL APPLICATIONS:
GLOBE SITE
acidic groundwater contain.
copper mining
Globe, Arizona
dissolved metals, sulfate
alluv., Gila conglomerate
sand, gravel, silt, clay
USGS
USGS-AZ-CO
99
from
170
-------
SITE NAME: GRAND ISLAND
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
subsurface contain, from disposal of
munition wastes
Grand Island, Nebraska
munition wastes, RDX, TNT
-0-
-0-
-0-
Univ. of Nebraska
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
HAIFA BAY EXPERIMENTAL FIELD
tracer study for porosity and
permeabilty
Israel
radioactive tracer, Co60
-0-
M sand, SS, Sh, C sand, clay
Israel Gov., UN SpFun
TAHAL Water Planning for Israel
Ltd.
-0-
SITE NAME: HANFORD SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
radiocontaminants, sat. and unsat.
south-central Washington State
radiocontaminants, plutonium
-0-
-0-
At. Richfield, DOE
Battelle, USGS, Boeing Comp.
Services
9
171
-------
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
HESKETT STATION
site selection for long-term
disposal of coal fly-ash
Mandan, ND
coal fly-ash
-0-
-0-
Montana Dakota Util.
NDEMRC, MMRRI
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
HOE CREEK
underground coal gasification
experimental site, contain
south of Gillette, Wyoming
organic contain., phenols
sedimentary sequence
sandstone, mudstone, coal seams
DOE
DOE, LLNL, Western Research Inst.
-0-
SITE NAME: HYDE PARK LANDFILL
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
immiscible organic transport,
chemical waste disp. site
Niagara Falls, New York
chemical waste
glaciolacustrine, Lockport Fm.
silt, clay, dolomite
Nat. Sci. & Eng.Council
University of Waterloo, EPA.
1
172
-------
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
IDAHO NEL SITE1
radioactive waste, Nat.Reactor
Testing Station
SE Idaho
liquid low-level rad. waste, dilute
chem., sewage
Quaternary, Snake R. Group
basalticvol., interbd. seds.
DOE
INEL, USGS, Atomic Energy Com.
(AEC)
3
SITE NAME: IDAHO NEL SITE2
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
purgeable organic compounds in gw
Idaho National Engineering Laborato
purgeable organic compounds
Quaternary, Snake R. Group
basalticvol., interbd. sed.
-0-
INEL, USGS, DOE
-0-
SITE NAME: IDAHO SPRINGS SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
measurement of stress in rocks near
mined repositories
Idaho Springs, Colorado
-0-
metamorphic rock
-0-
USGS
USGS
99
SITE NAME: INDUST SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMB.ER OF MODEL APPLICATIONS:
VOC contamination
-0-
VOCs
-0-
-0-
EPA-RSKERL
EPA, Arizona State
-0-
173
-------
SITE NAME: JEFFREY CITY
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
uranium tailings contain., sat. and
unsat., lab studies
central Wyoming, 2.4 km northeast
radionuclides, toxic elements
alluv, dune, sed. BR, cryst. BR
sand, gravel, sandstone, granite
-0-
D'Appolonia Inc., Earth Sciences
Consultants, Inc.
1
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
JET FUEL LAB
fate of jet fuel in aquatic
sediments, lab study
-0-
jet and missile fuel
-0-
-0-
-0-
ERL, Gulf Breeze, FL
-0-
SITE NAME: KELLY AFB
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
field-testing of in-situ biological
degradation
Kelly Air Force Base
organic contaminants
-0-
-0-
-0-
Science Applications International
Corp.
-0-
SITE NAME: KOPPERS COKE PLANT
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
groundwater contamination
St. Paul, MN
toxic organic substances
St. Peter aq., glacial
sand, gravel, silt, clay
USEPA, MPCA
USGS, USEPA, MPCA
-0-
174
-------
SITE NAME: LAB3
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
chemical transformation processes
in groundwater
-0-
organic chemicals
-0-
-0-
EPA-RSKERL
EPA
-0-
SITE NAME: LAB4
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
biodegradation of xenobiotics in
subsurface environment
-0-
organic pollutants
-0-
-0-
EPA-RSKERL
EPA, Univ. of Dayton
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
LAC DU BONNET SITE
fractured flow studies, granitic
batholith
Lac du Bonnet, Manitoba Canada
radioactive waste disposal
granitic batholith
granite, crystalline rock
DOE, CNFWM program
Atomic Energy of Canada
Limited(AECL), Battelle
1
175
-------
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
LAKE HAMILTON
aldicarb residue study, citrus
groves on Florida ri'dge
central ridge area of Florida, near
aldicarb residues
-0-
soil, C sand
EPA
Union Carbide, Univ. of Florida,
FDER, FDACS
2
SITE NAME: LANSING SMITH SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
ash disposal pond
Panama City Florida, coastal plain
fly ash leachate
marine dep., Floridian Aquifer
silt, sand, clay, frac. limestone
EPRI, EPA
Tetra Tech Inc., Arthur D. Little
(ADL)
NUMBER OF MODEL APPLICATIONS: -0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
LATUQUE SITE
waste pulp liquor contain., disposal
pit
LaTuque, Quebec
tannin, lignin, organic compounds
fluvial
sand
-0-
University of Waterloo
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
LBL FRACTURE SITES
organic contaminant migration in
fractured-porous rock
Lawrence Berkley Lab, California
organic pollutants
-0-
-0-
EPA-RSKERL
EPA, LBL
99
176
-------
SITE NAME: LIPARI LANDFILL
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
landfill leachate, site remediation
adjacent to Pitman and Glassboro,
New Jersey
numerous organic compounds, ether
-0-
-0-
EPA
EPA Reg. 2, Radian Corp., Camp
Dresser & Mckee, Inc.
1
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
LIQUIFIED FUEL SPILL TEST SITE
study accidental release of various
hazardous liquids
Nevada Test Site
various hazardous liquids, liq.
gaseous fuels
dry lake bed
-0-
DOE
LLNL
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
LIVINGSTON SITE
aldicarb contamination, sat. and
unsat.
central California
aldicarb
-0-
-o-
-0-
Union Carbide- Research Triangle
-0-
177
-------
SITE NAME: LLNL SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
VOCs, groundwater migration
Livermore, California
seven VOCs
alluvial sed.
sand, gravel, clay, silt
LLNL
LLNL
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
LLNL SITE 300
tritium contain, in vicinity of
landfills and exp. facil
LLNL, California
tritium
-0-
-0-
-0-
LLNL
2
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
LONG ISLAND
aldicarb residue in soil and
groundwater, nonpoint contain.
Long Island, New York
aldicarb, nonpoint contaminants
glacial moraine, outwash
sand, gravel, clay
Union Carbide, EPA
Cornell Univ., Union Carbide, USGS,
EPA
-0-
SITE NAME: LOS ALAMOS
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
environmental monitoring studies
LANL
-0-
-0-
-0-
-0-
LANL
-0-
178
-------
SITE NAME: LOVE CANAL
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
landfill
Niagara Falls, New York
mixed landfill wastes
glacial till, Lockport Dolomite
sand, silt, clay, till
EPA-Res. Triangle Pk.
GCA Corp., GeoTrans Inc.
-0-
SITE NAME: LOVIISA POWER STATION
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
tracer meas. by radioactive
isotopes
Finland
reactor wastes
-0-
-0-
-0-
Helsinki, Finland
2
SITE NAME: MADE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
macrodispersion exper., solid
wastes, utility disposal
Columbus AFB, Mississippi
seven conservative tracers
alluv., Pleist. terrace
sand, gravel, clay, silt
EPRI
Tenn. Valley Authority, MIT
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
MALIGNE KARST
field fluorometry in a karst
aquifer system
Alberta, Canada
field fluorometry tracing
faulted rock, Can. Rocky Mtns,
-0-
-0-
Univ. of Western Ontario
-0-
179
-------
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
MANTECA SITE
aldicarb contamination, sat. and
unsat.
central California
aldicarb
-0-
-0-
-0-
Union Carbide- Research Triangle
-0-
SITE NAME: MAXEY FLATS SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
low-level radioactive waste burial
Maxey Flats, Kentucky
low-level radioactive wastes
fractured sedimentary rocks
shale, sandstone
-0-
USGS
-0-
SITE NAME: MEREDOSIA SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
fertilizer contain, plume, geochem.,
remediation
south of Meredosia, Illinois
ammonia, sulfate, potassium,
chloride
alluv., glacial, Penn. BR
sand, gravel, sed. BR
ISWS
ISWS
1
SITE NAME: MOBILE SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
tracer tests, dispersion
north of Mobile, Alabama
sodium bromide
Quaternary terrace
sand, clay
USEPA-RSKERL
Auburn Univ.
9
180
-------
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
MOFFETT AIR STATION
trichloroethylene plume,
blotransformation
San Francisco Bay, California
trichloroethylene
-0-
-0-
EPA
RSKERL-EPA, Stanford
-0-
SITE NAME: MONTOUR TEST CELL
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
leachate chemistry study, fly ash
Batelle, Pacific NW
fly ash leachate
-0-
-0-
EPA, EPRI
EPRI SWES, Battelle NW
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
MONTPELIER SITE
coal flyash landfill, ground water
contamination
near Montpelier, Iowa
sulfate, Se, As, flyash leachate
Mississippi R. bluffs
loess, till, SS
USGS, Univ. of Iowa
IL. SWS, Univ. of Iowa
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
MORENO FIELD SITE
unsaturated-zone field exper.,
ponded intro. tracers
-0-
tracers
-0-
-0-
EPA, EPRI, S. CA Edison
Univ. of CA Riverside, EPRI
-0-
181
-------
SITE NAME: NDBRINE SITES
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
oil-and-gas brine leachate
north-central North Dakota
oil-and-gas brine
glacial, Fox Hills Fro.
sand, gravel, clay, sandstone
NDWRRI
NDMMRRI, NDGS, ND State Univ.
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
NDCOAL MINE SITES
determine hydrogeochemistry of
spoils settings
North Dakota
coal mining spoils
-0-
-0-
US Bureau of Mines
MMRRI
99
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
NEBRASKA NON-POINT
nonpoint pollution for agricultural
chemicals, 6 areas
High Plains Aquifer Nebraska
agricultural chemicals
High Plains Aquifer
unconsolidated sed.
USGS
USGS
-0-
SITE NAME: NEVADA TEST SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
monitoring radioactive waste
Nevada
radioactive waste, munitions waste
-0-
-0-
DOE
DOE
-0-
182
-------
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
NEW JERSEY ARSENAL
disposal of metal-plating wastes at
an arsenal
north-central NJ
chlorinated solvents, TCE
glacial drift, till, outwash
sand, gravel, silt, clay
USGS
USGS-NJ
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
NM DESERT SITE
unsat. flow in stratified soils,
effective hydraul. conductivity
New Mexico
-0-
-0-
-0-
-0-
MIT
-0-
SITE NAME: NMSU EXP STATION
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
mes. of variability of hydraulic
param. in unsat. zone
40 km northeast of Las Cruces, New
Mexico
-0-
het. layered soil-alluvial
unsat. soil
USNRC
MIT, NM State Univ., PNL, USNRC
2
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMB.ER OF MODEL APPLICATIONS:
NORTH SWISS RIVER
accidental tritium release, infil.
of river to gw
northern Switzerland
tritium
-0-
-0-
-0-
Swiss Federal Institutes
1
183
-------
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
OGCFRACTURE STUDY
organic contaminant migration in
fractured porous rock
-0-
organic pollutants
-0-
-0-
EPA-RSKERL
OGC, EPA
-0-
SITE NAME: OHIO CORN
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
pesticide leaching in agricultural
soils
Ohio
aldicarb
-0-
-0-
EPA-ERL-Athens
EPA-Athens, Union Carbide-RTP-NC
99
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
ORACLE SITE
movement of fluids in low-perm.
fractured rock
Oracle, Arizona
-0-
-0-
granite
USGS
USGS
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
ORNL FIELD FAC
research field facility for
subsurface transport
Oak Ridge, Tennesee
-0-
-0-
-0-
DOE
ORNL
-0-
184
-------
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
ORNL FIELD SITES
monitoring of HW facil.,
macropores, spatial variability
Oak Ridge, Tennessee
chemical pollutants
Cambrian, Ordovician sequence
shale, limestone, sandstone
DOE
ORNL, Geraghty & Miller
-0-
SITE NAME: ORNL HYDROFRACTURE FACILITY
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
radioactive waste disposal,
fractured rock system
Oak Ridge, Tennessee
radioactive waste
-0-
-0-
-0-
ORNL
-0-
SITE NAME: OTIS AFB
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
natural gradient tracer study
Cape Cod, Massachusetts
bromide, lithium bromide, fluoride,
molybdenum
glacial outwash, drift
sand, gravel, clay, silt
USGS
uses
l
185
-------
SITE NAME: OVIEDO SITE
NUMBER OF MODEL APPLICATIONS:
DESCRIPTION: aldicarb contamination, sat. and
unsat.
LOCATION: northeast of Orlando, Florida
aldicarb
-0-
-0-
State of FL
EPA-Athens, Univ. of FL, State of
FL
2
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
SITE NAME: PALO DURO BASIN
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
regional, high-level radioactive
waste disposal
north-east Texas
site evaluation, radioactive waste
disposal
Ogallala, Penn, Perm, Triassic
sand, clay, gravel, sed sequence
DOE
INTERA, TBEG, SWEC, ONWI-Battelle
1
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
PALOS FOREST PRESERVE
low-level radioactive waste burial,
monitoring, remedial
north-east Illinois
tritium, radioactive waste,
metallurgical waste
glacial, Silurian dolomite
clay, silt, sand
USGS
USGS, Argonne Nat. Lab
1
186
-------
SITE NAME: PANTANO SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
virus tracer studies
Pantano, Washington
viruses
-0-
-0-
EPA
EPA-RSKERL
-0-
SITE NAME: PARADOX BASIN
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
regional, high-level radioactive
waste, Gibson Dome
south-east Utah
evaluation of potential repository
site
Hermosa & Cutler Groups
marine carb, some elastics, salt
DOE
WCC, INTERA, ONWI-Battelle
2
SITE NAME: PARRIS ISLAND
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
core study of potential
denitrification rates
Parris Island, southeastern coastal
nitrate
uncon. Pleist., Floridan aquifer
sand, silt, clay, limestone
-0-
Univ. of SC, USGS
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
PEASE AFB
analyses of screening parameters
for water quality
Pease Air Force Base, New Hampshire
TOX, TOC, O & G, phenols, hyanide,
organic pollutants
-0-
-0-
-0-
Weston Inc.
-0-
187
-------
SITE NAME: PENSACOLA
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
phenolic contamination
NW FL-western panhandle
creosote, pentachlorophenol
fluvial, deltaic sediments
Sand, clay and gravel
USGS
USGS, Am. Creosote Works,
Pensacola-City, FL. Dept. Env. R
1
SITE NAME: PICEANE BASIN
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
oil-shale development
north-west Colorado
oil-shale site evaluation
-0-
SS, SH, LS, marlstone
DOE, EPA
USGS
3
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
PLAINS SITE
movement and fate of agricultural
chemicals
Plains, Georgia
pesticides, nitrogen compounds
unconsolidated
sand, gravel, clay
USGS
USGS-GA, USDA-ARS, EPA
99
SITE NAME: POWERTON SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
fly ash disposal (6-site study)
north-central Illinois t
fly ash leachate
alluv., glacial, Penn. BR
sand, gravel, clay
EPRI, EPA
Tetra Tech, Incorp., Arther D.
Little (ADL)
-0-
188
-------
SITE NAME: PRICE'S LANDFILL
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
landfill, mixed contamination
Atlantic City, NJ
acetone, acid chloroform, hexane,
cesspool waste, oil, xylene
CohanseySnd, Kirkwood FM, Tertiar
sand, clay
EPA, ACMUA
Paulus, Sokolowski and Sartor(PSS),
NJDEP, Pinder and B
2
SITE NAME: PRINCE EDWARD ISLAND
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
pesticide contain, in groundwater
below potato fields
Prince Edward Island, Ontario
aldicarb
-0-
-0-
-0-
Natl. Research Inst., Burlington,
Ontario
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
RADIAN FIELD SITES
collect field data to validate
geochemical models
-0-
-0-
-0-
-0-
EPRI
Radian Corp.
99
189
-------
SITE NAME: REGINA SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
transformer manufacturing, PCB
spill
Regina, Saskatchewan Canada
PCBs, TCBs, Inerteen 70-30
lacustrine, glacial till
clay, silt, fine sand
-0-
-0-
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
RICHTON DOME
regional, high-level radioactive
waste
south-east Mississippi
evaluation of potential repository
site
quaterary-alluv., Tertiary
sand, gravel, silt, clay, SS, LS,
SH
DOE
INTERA, Batelle-ONWI
-0-
SITE NAME: RIFLE SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
definition of reservoir and
hydraulic fracture system
Rifle, Colorado
-0-
-0-
-0-
-0-
Sandia Ntl. Labs
-0-
190
-------
SITE NAME: ROCKY MTN. ARSENAL
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
dissolved chemical transport,
unlined disposal ponds
north-central Colorado
liquid industrial wastes, Cl
alluvial
-0-
USGS
R Mnt. A, Core of Eng., CO Dept. of
Health
3
SITE NAME: SAN JOAQUIN VALLEY
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
regional study
San Joaquin Valley, CA
-0-
-0-
-0-
-0-
Univ. of CA
-0-
SITE NAME: SAND RIDGE STATE FOREST
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
dye tracer study
west-central Illinois
fluorescent dyes
Wisconsin glacial
FM sand, C sand, gravel
USEPA-OSU
IL SWS & ENR
1
SITE NAME: SAVANNAH RIVER
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
migration of radionuclides in soil
Aiken, South Carolina
radionuclides
-0-
-0-
DOE
DOE
-0-
191
-------
SITE NAME: SEYMOUR SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
hazardous waste site
Seymour, Indiana
solvents, metal finishing wastes
alluvial, glacial
sand, gravel, clay, silt
EPA
EPA
1
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
SHEFFIELD SITE
low-level radioactive-waste burial
site
Sheffield, Illinois
low-level radioactive waste
glacial till, outwash
sand, clay, silt
-0-
USGS
1
SITE NAME: SHERCO SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
sludge/flyash waste pond
Sherburne County, Minnesota
flyash leachate
glacial, PreCambrian rock
sand, gravel, granite
EPRI, EPA
Tetra Tech Inc., Arthur D. Little
(ADL)
-0-
SITE NAME: SHULLSBURG SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
zinc-lead mine contamination
Shullsburg, Wisconsin
sulfate
Galena-Platteville Fm.
dolomite, limestone, shale
WI-WRC, Univ. WI, AGU
Univ. of Wisconsin
1
192
-------
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
SOIL MOISTURE
soil moisture diffusivity char.
loamy to silty soil
Ardeche Basin, France
-0-
-0-
-0-
-0-
Univ. of Utrecht, Netherlands
-0-
of
SITE NAME: SOILS LAB 1
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
sorption and transport of toxic
organic substance
-0-
organic chemicals
-0-
-0-
EPA-RSKERR
EPA, Univ. of Florida
99
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
STRIPA SITE
geochemistry of trace elements for
radioactive waste disposal
Stripa Mine in Sweden
radioactive wastes
fract. crystalline rock
-0-
-0-
USGS
99
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
STRIPMINE SITES
environmental effects of
stripmining sites
Montana, Wyoming, North Dakota
mining spoils
-0-
-0-
USEPA
MMRRI
99
193
-------
SITE NAME: STROUOSBURG SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
abandoned illuminating gas plant
Stroudsburg, Pennsylvania
coal tar wastes
-0-
gravel, fine silty sand
-0-
-0-
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
SWES PROJECTS
EPRI studies on trans.
utility wastes
-0-
utility wastes
-0-
-0-
EPRI
EPRI
-0-
and fate of
SITE NAME: TEXAS TECH SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
tracing migration of chemicals from
nonpoint sources
Lubbock, Texas
salt solution injection
-0-
-0-
TX Wat. Dev. Board
Texas Tech. Univ.
-0-
SITE NAME: TOKYO SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
sewage sludge applications to soils
60 km. northeast of Tokyo, Japan
sewage sludge application
-0-
-0-
-0-
Natl. Inst. for Env. Studies
Tsukuba, Ibaraki
-0-
194
-------
SITE NAME: TRACER RESEARCH SITE
DESCRIPTION:
' LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
shallow soil gas study, TCE plume
-0-
TCE, gaseous phase
-0-
-0-
-0-
Tracer Research Corp., Tucson, AZ
-0-
SITE NAME: TRAVERSE CITY
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
gasoline spill-underground storage
Traverse City, Michigan- US Coast G
gasoline
-0-
-0-
EPA
EPA-RSKERL, Rice University
-0-
SITE NAME: TWIN-CITIES
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
army ammunition plant
New Brighton, MN
toxic organic substances
glacial, PrarieduChienJordan
sand, gravel, clay, silt, dolomite
USGS, USEPA, MPCA
MN Poll. Control Agency (MPCA),
USEPA, USGS
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
UNITED CHROME
chromium plating waste, United
Chrome Products
Washington State
chromium
Quaternary alluv.
clay, silt, sand, gravel
EPA Reg. 10-Seattle
OGC, E&E
-0-
195
-------
SITE NAME: VESICOL SITE
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
hazardous waste site
western Tennessee
chemical wastes from pesticide
manufacturing
-0-
-0-
-0-
EPA, Vesicol Chem. Corp.
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
VICTORIA PROVINCE
hydrogeological study in
crystalline rock, fractures
Victoria Province, United Kingdom
-0-
crystalline basement rock
granite, gneiss
-0-
Hydrotechnica Shrewsbury, UK
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
VOLK AIR BASE
in-situ soil washing, removal of
hydrocarbons
Volk Air National Guard Base, WI
hydrocarbons, chlorinated
hydrocarbons
-0-
-0-
EPA
EPA, USAF, Mason and Hanger-Silas
Mason Co., Inc.
-0-
196
-------
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
WASTE ISOLATION PLANT
evaluate geohydrology of a setting
for nuclear waste disposal
40 km east of Carlsbad, New Mexico
nuclear waste
Permian rock, faulted area
sed. rocks, salt
DOE
DOE, USGS
-0-
SITE NAME: WELDON SPRING SITES
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
radioactive waste disposal sites
48 km west of St. Louis, Missouri
radioactive waste
carbonate rocks
-0-
-0-
USGS
-0-
SITE NAME: WESTERN INDIA
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
injection well tests in an alluvial
aquifer
western India
-0-
het. alluvial mat., layered
sand, gravel, clay
-0-
Univ. of Birmingham, Central GW
Board-Ahmedabad Indi
1
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
WESTERN PROCESSING
industrial hazardous waste, 26
priority pollutants /
Kent, Washington
TCE, 26 priority pollutants
Green R. Flood Plain
sand, silt, peaty silt, clay
EPA-MERL Cine.
Battelle
1
197
-------
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
WINDERL SITE
evaluate brine and oil-gas drilling
fluid disposal site
North Dakota
oil brine, oil-gas drilling fluid
glaciofluvial, till
sand, clay, silt
NDWRRI
NDGS, Univ. of ND, MMRRI
1
SITE NAME: WISCONSIN FRACTURE SYSTEMS
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
anisotropy, directional connec.,
porosity
60 sites throughout Wisconsin
-0-
-0-
-0-
-0-
Univ. of Wisconsin
-0-
SITE NAME: WOOD RIVER JUNCTION
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
contaminant plume, uranium recovery
plant
Wood River Junction, Rhode Island
strontium 90, radionuclides,
chemical solutes
glacial till, outwash
sand, gravel, gneiss
USGS
USGS
-0-
SITE NAME:
DESCRIPTION:
LOCATION:
CONTAMINANT OR TRACER:
GEOLOGY:
AQUIFER MATERIAL:
SOURCE(S) OF FUNDING:
ORGANIZATIONS INVOLVED:
NUMBER OF MODEL APPLICATIONS:
YAKIMA SITE
processes study of gasoline and
diesel oil contamination
Yakima, Washington
gasoline, diesel oil
unconsol.-Holocene age
sand, gravel
USGS
USGS, OGC
-0-
198
-------
SITE NAME: YUCCA MOUNTAIN
DESCRIPTION: nuclear waste storage invest.,
fractured flow
LOCATION: Nevada Test Site
CONTAMINANT OR TRACER: nuclear waste
GEOLOGY: Topopah Spring Member
AQUIFER MATERIAL: ash-flow tuffs
SOURCE(S) OF FUNDING: DOE
ORGANIZATIONS INVOLVED: Sandia Nat. Lab., LBL
NUMBER OF MODEL APPLICATIONS: 3
199
-------
APPENDIX C
ENTRY FORM FOR SATURN DATABASE
LEVEL 1: GENERAL SITE-SPECIFIC INFORMATION
Site Identification
Site Information
Geographic Information
Geohydrologlcal Site Data
Other Site Data Availability
Model Applications/Testing
LEVEL 2: INFORMATION ON PARTICULAR STUDIES AT A SITE
Study Identification/Information
Type of Measurements Made
Type of Measurements/Tasks Performed
LEVEL 3: INFORMATION ON INDIVIDUAL TASKS (OR INVESTIGATIONS)
Task Identification
Dataset Information
General Information/Documented QA/QC
Tasks:
• Field - Chemistry - In-situ Monitoring
• Field - Chemistry - Sampling and Analysis
• Task: Field - Chemistry - Tracer Test
• Field - Hydrogeology - Water Level Observation
• Field - Hydrogeology - Aquifer Test (Aquifer Parameters)
• Field • Soils - Water Content/Soil Moisture Tension
• Field - Soils • Infiltration/Moisture Movement
• Field - Soils - Measurement of Soil Parameters
• Task: Laboratory Experiments (Y/N)
PUBLICATIONS
USER REFERENCES
**********************************************************************
Instructions:
The form asks for three types of Information: numerical (#), text (text), or Boolean (Y/N). The variable type
on the form is consistent with that used in the SATURN database. If a field is not filled in, it is assumed that
the value is "no" or "0*; if no information is available the field should be left blank.
200
-------
LEVEL 1: GENERAL AND SITE-SPECIFIC INFORMATION
Site Identification
Site Name (text):
General Site Research Coordination (text)
Organization:
Department:
Address:
Contact Person (s):
Site Owner/Manager (text)
Organization:
Contact Person:
Site Information
Contamination Present (Y/N):
Source of Contamination if Present (Y/N)
Waste Disposal
Industrial Impoundment: _
Illegal Dumping: _
Septic Tanks: _
Deep Well Injection:
Industry/Municipal Landfill:
Wastewater Treatment:
/
Radioactive Waste Disposal:
Oil and Gas Field Brines:
201
-------
Accidental Pollution
Surface Spills:
Pipeline Leaks:
Leaking Storage Tanks:
Acid Mine Drainage:
Non-Point Pollution
Agricultural Chemicals:
Irrigation:
Salt-water Intrusion:
Industrial Impoundment:
Illegal Dumping:
Septic Tanks:
Deep Well Injection:
Feedlots:
In-situ Mining:
Industrial/Municipal Landfill:
Wastewater Treatment:
Radioactive Waste Disposal:
Oil and Gas Field Brines:
Accidental Pollution
Surface Spills:
Pipeline Leaks:
Leaking Storage Tanks:
Acid Mine Drainage:
Non-Point Pollution
Agricultural Chemicals:
Irrigation:
Salt-water Intrusion:
Feedlots:
In-situ Mining:
Other sources (text):
Kind of Pollutants (Y/N)
Organics
Aromatic:
Oxygenated:
Halogenated:
202
-------
Anorganlcs
Heavy Metals: Other metals: Nitrates:
Phosphates: Sulfates: Cyanides:
Chlorides: Radionuclides: Other:
Tracers:
Major Pollutants (text):
Pollution History (pollution amounts, rates; time period sources were active):
Geographic Information
Coordinates
Coordinate System (text):
Latitude (#):
Longitude (#):
Size of Site (#): Units (text):
Location (text)
Nearest city/town:
County:
State/Province:
Country:
203
-------
Elevation
Maximum Elevation (#):
Minimum Elevation (#):
Average Slope Gradient (#):
Topographical Setting (Y/N)
Flatland:
Stream Channel:
Terrace:
Dunes:
(fraction)
Depression:
Hillside:
Plateau:
Polder:
Units (m or ft):
Units (m or ft):
Valley Bottom:
Hilltop:
Marshland:
Other (text):
Geohvdroloaical Site Data
Lithology (Y/N)
Sand:
Sands and Silts:
Limestone:
Other (text):
Sand and Gravel: Sandstone: _
Clay: Till: Shale:
Igneous/Metamorphic Rock:
Short Geological Description (lithology, layer thickness, major formations, karst, fractures):
204
-------
Hydrogeology (Y/N)
Single Aquifer:
Confined:
Semi-Confined:
Unconfined (shallow water-table):
Unconfined (deep water-table):
Porous:
Fractured:
Multi-Aquifer:
Top-Layer
Spatial Continuous:
Semi-Confined:
Unconfined (shallow water-table):
Unconfined (deep water-table):
Porous:
Fractured:
Lower Layers
Spatial Continuous:
Porous:
Fractured:
Short Hydrogeological Description (schematization, aquifer names, aquifer and aquitard thickness,
karst, fracture density and orientation):
205
-------
Soil Types (Y/N)
Thin or Absent:
Non-Shrinking Clay: Shrinking days:
Clayey-Loam: Silty Loam:
Sandy Loam: Sand:
Gravel: Muck:
Peat:
Other (text):
Short Soil Description (layer names, thickness, soil type, presence of macro-pores, root density,
etc.):
Surface Water Presence at or near Site (Y/N)
Streams/Creeks: Canals/Ditches:
Lakes/Reservoirs: Impoundments/Ponds:
Sea/Ocean: Fresh-Water Estuary:
206
-------
Short Description of Surface Water Features (relative importance, how does it relates to site/plume,
flow characteristics, perennial, ephemeral, etc.):
Ground water Levels
Data Availability (Y/N)
Long-term: Short-term:
Continuous:
Observation Wells (#):
Average Depth to Water-table (#): m or ft
Average Regional Groundwater Velocity (#): m/d or ft/d
Is Vertical Flow Present? (Y/N):
Sinks/Sources in Area (Y/N)
Municipal Wellfidd: Industrial Discharge Well:
Private Weils: Shallow Recharge Wells:
- f
Deep Waste Injection Wells: Pumping/Injecting Remediation:
207
-------
Short Description of Groundwater Flow Characteristics (location and discharge rates of wells,
groundwater flow direction and rates, etc.):
Other Site Data Availability
Meteorological Data (Y/N)
Precipitation: Evaporation:
Evapotranspiration: Air Temperature:
Relative Humidity: Wind Velocities:
Other (text):
Land Use (Y/N)
Grassland: Crops: Forest:
Rangeland: Prairie: Desert:
Build Area: Roads: Waste Disposal:
Other (text):
Comments:
208
-------
Previous Model Testing and Modeling Applications
Site Is Known for Model Applications (Y/N):
Study Data Have Been Used for Model Testing (Y/N):
Model(s) Used at Site/Study (text)
Acronym and IGWMC-Key:
(Add references describing application/testing to user category)
Characterization of Applications/Testing (Y/N)
Saturated Flow:
Variably Saturated Flow:
I nf iltration/Recharge:
Buoyancy Flow:
Multi-phase Flow:
Advective Transport:
Dispersion:
Soil/Water/Solute Interaction:
Vapor Diffusion:
Gas Transport in Soils:
Biodegradation/Bioremediation:
Flow/Transport in Freezing/Thawing Soils:
Row and Consolidation/Subsidence:
Salt Water Intrusion:
Other (text):
209
-------
LEVEL 2: STUDY SPECIFIC INFORMATION
Study Identification/Information
Study Title (text):
Study Coordination (text)
Organization:
Department:
Address:
Coordinator/Pi:
Contact Person(s):
Study Period/Dates (text)
Start: End:
IGWMC Check-Date:
Maior Measurements Made (Y/N)
Solute Concentrations:
Groundwater Velocities:
Soil Moisture Content:
Infiltration/Recharge:
Soil/Rock Density:
Evaporation:
Groundwater Levels/Drawdowns:
Groundwater Flow/Dicharge Rates:
Soil Tension/Suction:
Porosity:
Precipitation:
Evapotranspiration:
Other (Text):
Type of Measurements/Tasks Performed
Primary Tasks (See Level 3 for Detail)
210
-------
Field Experiments (Y/N):
Chemistry:
In-situ Groundwater/Soil-Water Monitoring:
Groundwater/Soil-Water Sampling and Analysis:
Tracer Test:
Hydrogeology:
Water Level Observation:
Aquifer Test (Aquifer Parameters):
Soil Physics:
Water Content/Soil Moisture Tension:
Infiltration/Moisture Movement:
Measurement of Soil Parameters:
Laboratory Experiments (Y/N):
Secondary Tasks (No further detail provided)
Laboratory Tests (Y/N):
Rock Mechanical Properties:
Rock Hydraulic Properties:
Soil-Physical Properties:
Soil/Rock Chemical Properties:
Borehole (Rock) Sampling
Number of Sampling Locations (#):
Number of Sampling Points (#):
211
-------
Total Number of Samples Taken (#):
Soil Sampling
Number of Sampling Locations (#):
Number of Sampling Points (#):
Total Number of Samples Taken (#):
Geophysical Surveys (Y/N):
Surface Methods (Y/N):
Surface Resistivity:
Reflection Seismics:
Low-Induction Conductivity:
Electromagnetic Sounding:
Other (text):
Borehole Methods (Y/N):
Spontaneous Potential:
Electric Resistivity:
Gamma-Gamma:
Groundwater Velocity:
Temperature:
Dip:
Natural Gamma:
Neutron:
Row Rate:
Acoustic Logs:
Caliper:
video:
Other (Text):
Geodetic Surveying (Y/N):
Remote Sensing (Y/N):
212
-------
213
-------
LEVEL 3: TASK-SPECIFIC INFORMATION
Task Identification
Task Title (text):
Task Coordination (text)
Organization:
Department:
Address:
Coordinator/Pi:
Contact Person(s):
Task Period/Dates (text)
Start: End: IGWMC Check-Date:
Dataset Information
Dataset Name/Acronym:
Dataset Structure:
Spatial Data Type (Y/N)
Point: Line: Cell/Element: Grid:
Other:
Timespan Dataset Coverage; Dates (text)
Start: End: IGWMC Check-Date:
214
-------
Status of Dataset (Y/N)
Operational: Under Development:
Under Modification: Closed (Not Further Maintained):
Availability for Use/Access (Y/N)
Computerized: Report Only: Data Sheets Only:
Unrestricted: Proprietary: Conditional Use:
Access Method (Y/N)
On-Line: Batch: Indirect:
Transfer Media (Y/N)
Magnetic Tape: Diskette: Printout:
Electronic File Transfer:
Resident Storage Media (computer, mass storage device):
DBMS (if applicable; name, version):
Size of Dataset (#)
Logical Records:
Bytes per Record:
Expected Yearly Growth (# of records):
Datatype (Y/N)
Raw: Conditioned: Summarized: Derived:
215
-------
General Information/Documented QA/QC
Available Documentation (Y/N)
Data Report: Equipment Used:
Procedures Applied: Field Conditions:
Personnel: Data Conditioning:
Data Conversions: Data Analysis:
Data Transfer and Storage: Applied QA/QC:
Documented Dataset Assessment by Source (Y/N)
Consistency: Outliers: Accuracy:
Precision: Completeness: Reproducibility:
Procedures: Equipment: Personnel:
IGWMC Dataset and Documentation Assessment (Y/N)
Consistency: Outliers: Accuracy:
Precision: Completeness: Reproducibility:
Procedures: Equipment: Personnel:
Field Conditions: Data Handling: Data Analysis:
Data Report: QA/QC Applied:
Note: IGWMC evaluations, if present, are separately documented.
Level 3 Task: Field - Chemistry • In-sltu Monitoring
SoP-Water (Y/N): Groundwater (Y/N):
Measurements (Y/N)
Specific Conductance: pH:
Temperature: Eh:
Other (text):
216
-------
Monitoring Scenario and Network Design (text):
Procedures (text):
Equipment (text):
217
-------
Level 3 Task: Field • Chemistry - Sampling and Analysis
Soil-Water (Y/N): Groundwater (Y/N):
Sampling Locations (#):
Sampling Points (#):
Total Number of Samples Taken (#):
Sampling Frequency (text):
Primary Constituents (text):
Sampling Scenario and Network Design (text):
Procedures (text):
Equipment (text):
218
-------
Level 3 Task: Field - Chemistry - Tracer Test
Sampling Locations (#):
Sampling Points (#):
Total Number of Samples Taken (#):
Sampling Frequency (text):
Tracer(s) (text):
Derived Parameters (text):
Sampling Scenario and Network Design (text):
Procedures (text):
Equipment (text):
219
-------
Level 3 Task: Field - HvdrogeoloQv • Water Level Observation
Observation Locations (#):
Observation Points (#):
Total Number of Observations Made (#):
Observation Frequency (text):
Observation Scenario and Network Design (text):
Procedures (text):
Equipment (text):
220
-------
Level 3 Task: Field - Hvdrooeolooy - Aquifer Test (Aquifer Parameters!
Aquifer Test Locations (#):
Total Tests Made (#):
Number of Observation Wells (#):
Type of Test (Y/N)
Pump Test: Bailer: Slug: Recovery:
Other (text):
Derived Parameters (Y/N)
Transmissivity:
Horizontal Hydraulic Conductivity:
Vertical Hydraulic Conductivity:
Leakage Factor:
Storage Coefficient:
Other (text):
Aquifer Test Scenario and Network Design (text):
Equipment (text):
221
-------
Method of Analysis (text:
Level 3 Task: Reid - Soils - Water Content/Soil Moisture Tension
Observation Locations (#):
Observation Points (#):
Total Number of Observations Made (#):
Observation Frequency (text):
Procedures (text):
Equipment (text):
222
-------
Level 3 Task: Field - Soils - Infiltration/Moisture Movement
Observation Locations (#):
Observation Points (#):
Total Number of Observations Made (#):
Observation Frequency (text):
Procedures (text):
Equipment (text):
223
-------
Level 3 Task: Field - Soils - Measurement of SoH Parameters
Measurement Locations (#):
Total Measurements Made (#):
Type of Test (text):
Derived Parameters (text):
Aquifer Test Scenario and Network Design (text):
Equipment (text):
Method of Analysis (text:
224
-------
Level 3 Task: Laboratory Experiments (Y/N1
Moisture Movement in Variably Saturated Soils:
Water Movement in Saturated Porous Rock:
Water Movement in Saturated Fractured Rock:
Advective-Dispersive Solute Transport:
(Bio)chemical Transformation/Degradation:
Immiscible/Multiphase Row:
Transport of VOCs:
Type of Information Collected (text):
Technique(s) (text):
225
-------
PUBLICATIONS
Title:
Author(s):
Year:
Journal:
Volume/lssue/Pages:
Book/Report Publisher:
Title:
Author(s):
Year:
Journal:
Volume/Issue/Pages:
Book/Report Publisher:
Title:
Author(s):
Year:
Journal:
Volume/lssue/Pages:
Book/Report Publisher:
Continue on new page if necessary
226
-------
USER REFERENCES
title:
Author(s):
Year:
Journal:
Volume/lssue/Pages:
Book/Report Publisher:
Title:
Author(s):
Year:
Journal:
Volume/lssue/Pages:
Book/Report Publisher:
Title:
Author(s):
Year:
Journal:
Volume/lssue/Pages:
Book/Report Publisher:
Continue on new page if necessary
OU.S GOVERNMENT PRINTING OFFICE. 1 9 8 9-6 "» 8 -i 6 3/ 00316
227
------- |