6
Conclusions and Recommendations

''Of course there must be subtleties. Just make sure you make them clear."

—Billy Wilder

This final chapter of the report revisits two of the laws of ground water vulnerability and addresses the most important findings of the preceding chapters. The chapter includes advice to policy makers and managers seeking to apply vulnerability assessments in ground water protection programs and a research agenda that suggests promising directions for improved understanding of the process of ground water contamination and the prediction of vulnerability to contamination.

First Law of Ground Water Vulnerability:

All ground water is vulnerable.

Second Law of Ground Water Vulnerability:

Uncertainty is inherent in all vulnerability assessments.

The First Law says, in effect, that ground water vulnerability is a relative rather than an absolute concept. That is, an aquifer or portion of an aquifer can only be judged to be more or less vulnerable to contamination than other aquifers or other portions of a given aquifer. Furthermore, it may be necessary to consider effects on ground water quality over longer time spans and greater distances than is commonly done in vulnerability assessments. The Second Law says that both natural variability in spatial attributes and inability to specify attribute values accurately at all spatial scales of interest over a given region will impart uncertainty, often undetermined,



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 170
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty 6 Conclusions and Recommendations ''Of course there must be subtleties. Just make sure you make them clear." —Billy Wilder This final chapter of the report revisits two of the laws of ground water vulnerability and addresses the most important findings of the preceding chapters. The chapter includes advice to policy makers and managers seeking to apply vulnerability assessments in ground water protection programs and a research agenda that suggests promising directions for improved understanding of the process of ground water contamination and the prediction of vulnerability to contamination. First Law of Ground Water Vulnerability: All ground water is vulnerable. Second Law of Ground Water Vulnerability: Uncertainty is inherent in all vulnerability assessments. The First Law says, in effect, that ground water vulnerability is a relative rather than an absolute concept. That is, an aquifer or portion of an aquifer can only be judged to be more or less vulnerable to contamination than other aquifers or other portions of a given aquifer. Furthermore, it may be necessary to consider effects on ground water quality over longer time spans and greater distances than is commonly done in vulnerability assessments. The Second Law says that both natural variability in spatial attributes and inability to specify attribute values accurately at all spatial scales of interest over a given region will impart uncertainty, often undetermined,

OCR for page 170
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty to all assessments of ground water vulnerability. These laws should caution users of vulnerability assessments that ignoring uncertainty can lead to considerable error. Careful attention should be paid to the effects of these potential errors on decisions that will be informed by a vulnerability assessment. If the decisions would not change if the uncertainty were considered, then users of the assessment should have increased confidence in using its results. If the decision would change in the face of uncertainty, then the use of the assessment in making decisions would have to be viewed with caution. This pervasive, inherent uncertainty led the committee to a probabilistic, rather than deterministic, definition of ground water vulnerability: The tendency or likelihood for contaminants to reach a specified position in the ground water system after introduction at some location above the uppermost aquifer. Ground water vulnerability is not a measurable property, but a probability statement about future contamination that must be inferred from surrogate measurements. Such information, in its simplest form, may be a single parameter, such as depth to ground water. Like a weather forecast, vulnerability to contamination is best expressed as a probability of an event (e.g., 30 percent chance of rain). Yet very few of the vulnerability assessment methods discussed in Chapter 3 produce results in the form of probabilities. This report distinguishes between two types of ground water vulnerability: intrinsic vulnerability, which reflects properties that are a function of the natural setting and does not consider the attributes and behavior of particular contaminants, and specific vulnerability, which reflects factors that relate to the properties of the specific constituent(s) of concern, and possibly specific circumstances of land and chemical use (Chapter 1). Using vulnerability assessments currently available, it is fairly easy to delineate many areas of high vulnerability, difficult to say for certain that an area has very low vulnerability, and not possible to make fine gradations in between. MANAGEMENT IMPLICATIONS Ground water vulnerability assessment is a dynamic, iterative, and interactive process that must involve the cooperative efforts of policy makers, resource managers, and technical experts. Figure 1.3, Chapter 1, illustrates the dynamic interactions among the four major components of an assessment: intended purpose, approaches, required data, and management actions. Chapter 2 describes the uses of vulnerability assessments and the technical and institutional considerations that should be addressed in planning a vulnerability assessment as a tool for management. The case studies (Chapter 5) illustrate how the vulnerability assessment process is being

OCR for page 170
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty approached across the country. Clearly, the case studies demonstrate that information gained is fed back to improve and refine the resulting actions. The value of educating people must be emphasized. For example, the main purpose of the Iowa vulnerability assessment process was to inform the state's residents of the need for better pesticide and nutrient management practices to protect ground water from potential contamination. The case studies show that structured, quantitative vulnerability assessments do not necessarily fill a direct decisionmaking role, but contribute to the understanding of the scope of the problem and help create a consensus for action. Vulnerability assessments should be refined as experience grows. For this reason, models, indices, or other approaches should not be chosen without careful consideration of the factors discussed in Chapter 2 that should influence the selection and use of vulnerability assessments. Although maps are only a small component of the vulnerability assessment process, they are an inevitable, and the most visible, product and often can impart a false sense of security to the user who accepts them uncritically. For example, a user may conclude with false confidence that areas identified by assessments as having low vulnerability will provide reliably acceptable sites for land uses or activities likely to be potential sources of ground water contamination. A false negative vulnerability rating—areas shown on the map as low vulnerability that are in reality high—could result in serious contamination and related management problems. Also, false positive errors—high vulnerability areas on the map that are actually low—can lead to overly restrictive, costly, and unpopular land use requirements. Again, the consequences of false positives or false negatives, as they affect management decisions, need to be thought through before action is taken. Analysis further suggests that even if a region can be partitioned into safe and vulnerable areas, subdividing it into areas having intermediate vulnerabilities will be difficult. More categories of vulnerability in the assessment may suggest to managers an ability to construct a zoning system or site screening process of greater discrimination, but the ultimate utility of these devices is limited by both uncertainty in the vulnerability assessment and uncertainty in the evaluation of contamination risk by the land use or activity. Chapter 3 contains further discussion of the inability of existing techniques to support such discrimination. This finding departs from the expectations of both regulatory policy makers and the regulated community, who idealistically may prefer a finer discrimination among safe and vulnerable areas. In the context of differential management of ground water—with its goal of efficient use of resources—policy makers, resource managers, and land users, would use vulnerability assessments as a tool in adjusting regulatory requirements and management practices for different areas and allocating program resources. For example, regulatory requirements and management

OCR for page 170
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty practices aimed at potential ground water contaminants, such as agricultural pesticides, could vary from area to area with the results of the vulnerability assessment as displayed on a multicategory map. This strategy for ground water management allows limited resources and/or personnel to be directed toward particular areas or activities with a higher likelihood of contamination, or higher vulnerability. That is not to say, however, that other areas should not be managed; less vulnerable areas still demand some level of management. APPROACHES The vulnerability assessment methods discussed in Chapter 3 range from simple overlay approaches, to index and statistical methods, to process-based modeling approaches. A rule of thumb for currently available techniques is that the more complex and data intensive the method, the smaller the area that can be assessed. For example, detailed process-based models are often used for field and small hydrologic unit scales; overlay and indexing methods have often been applied at the larger regional and national levels. This rule of thumb, however, does not suggest the appropriate use of methods; it simply suggests how cost considerations and data availability have led to methodological preferences based on the scale of application. Statistical methods can be applied at all levels consistent with the spatial resolution of the data. In theory, chemical movement through the soil and vadose zone could be described by a model of contaminant transport and fate, but current models are not good enough for predicting where, when, and at what concentration a constituent will appear. This situation is due to spatial variability in characteristics of the landscape and properties of the media, uncertainties associated with the modeling techniques, and uncertainties involved in estimation of attributes based on available data. These difficulties are discussed extensively in Chapters 3 and 4. On a regional scale, index methods are in some sense conceptually appropriate in that they deal explicitly with the multivariate nature of the problem; however, one set of weights is not sufficient for all situations. None of these methods, even process-based models, can be validated in the usual scientific sense for vulnerability assessments because of spatial and temporal variability. This uncertainty in the ability to estimate the likelihood of future contamination will persist in the absence of noninvasive techniques for characterizing soil and ground water systems in three dimensions with respect to the parameters that affect contaminant movement through soil and the vadose zone. Despite these difficulties, inferences about the accuracy of a regional vulnerability assessment can be made through several lines of inquiry. The

OCR for page 170
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty process of testing and evaluating vulnerability assessments may involve a hierarchical approach that evolves over several stages. In fact, the most sensible applications of vulnerability assessment techniques may include explicit plans to test, review, and refine the assessment over time. Those who generate vulnerability assessments must ensure that users are aware of the uncertainties associated with the modeling scheme and data used. Policy and decision makers are left with the responsibility of making informed choices using uncertain scientific assessments. Scientists must accept the responsibility of assisting decision makers in correctly interpreting the sources of uncertainty and increasing their confidence in the results of vulnerability assessments. DATA AND DATABASES Databases and their characteristics, content, scale, limitations, deficiencies, and availability are discussed in Chapter 4. Only nationally available databases are discussed, but many states have data and machine retrievable databases that are valuable for vulnerability assessments. The appropriateness of the various databases for different levels of assessment are discussed. One of the committee's original goals was to develop a single set of parameters that are important for a national level assessment for vulnerability. However, the complexity and local nature of conditions leading to ground water vulnerability make it impossible to establish a set of parameters important in all cases. The important parameters differ in different parts of the country and in different conditions. A major constraint on vulnerability assessments is the different scales of data in the various databases. A second, serious deficiency is the uncertainty of the data in the databases. These limitations are being addressed by the Federal Geographic Data Committee and the new Spatial Data Transfer Standard; however, more emphasis on these and other similar efforts is needed. Standard national databases of good quality and understandable content are essential to ground water resource assessment and protection. The other major constraint on assessment of ground water vulnerability is the lack of digital spatial databases, particularly at the county, watershed, and field levels. Although soils and topography are mapped for most of the country, less than 10 percent are digitized. Geologic mapping at scales useful for many vulnerability assessments is limited in many areas. Good climatic data are also lacking for much of the western United States. Databases on chemical properties and chemical use continue to expand and improve, but are still lacking. The uncertainty of the data involved in the vulnerability assessment, and the uncertainty of the assessment method itself, are often not well

OCR for page 170
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty represented in assessment products. In most cases, the major product is the map of the area that portrays the results of the assessment. Often, these maps oversimplify the results of an assessment or include too much information and, therefore, confuse or mislead the user. The ineffective use of maps to portray the results of assessments is due to a combination of poor definition of the purpose of the map, poor assessment of the knowledge of the user, poor cartographic skills of the preparer of the map, and the amount of time and effort it takes to prepare the complex and often multiple maps required to represent the data. If a ground water vulnerability assessment is to be useful, the map must present the results in a clear, understandable fashion so that the user can reach appropriate conclusions. Also, users must commit to vulnerability assessments the time and attention necessary for informed decision-making. By carefully reviewing each of these factors and making suitable choices, the responsible specialists can prepare effective vulnerability assessment maps. With the availability of geographic information systems (GIS) software in recent years, digital information arising from vulnerability assessments can be easily displayed on a very sophisticated map without displaying the actual quality of the assessment. Innovation by the user and GIS industry, associated with improved assessment methodologies and uncertainty analyses, will prove most useful for depicting uncertainties associated with the vulnerability assessments portrayed on these same maps. RESEARCH AGENDA The committee's evaluation of vulnerability assessments led to identification of a body of research needs, many of which are specified here in general terms. This research agenda is divided into four categories: fundamental understanding of transport and fate processes, database improvements, geoprocessing and display improvements, and improvements in assessment methods. No order of priority or relative need is reflected in the following. Fundamental Transport and Fate Processes Develop a better understanding of all processes that affect the transport and fate of contaminants. A vulnerability assessment is only as good as the information/knowledge available at the time. Lack of understanding of the factors that affect the transport and fate of the contaminant in the environment decreases the certainty associated with an assessment. Establish simple, practical, and reliable methods for measuring in situ hydraulic conductivities of the soil and the unsaturated and saturated zones. Develop simple, practical, and reliable methods for measuring

OCR for page 170
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty in situ degradation rates (e.g., hydrolysis, methylation, biodegradation), and develop methods for characterizing changes in degradation rate as a function of other physical parameters (e.g., depth in soil). Develop methods for scaling measurements that sample different volumes of porous material to provide equivalent measures. This information is of primary importance in determining contaminant fate and transport in the soil. Emphasis should be placed on developing methods that are relatively inexpensive. Develop improved approaches to obtaining information on the residence time of water along flow paths and identifying recharge and discharge areas. It is important to protect recharge zones from contamination. Additional research into methods that provide the necessary information should be encouraged. For example, methods that use environmental isotopes may be useful and, therefore, should be developed further and evaluated in this context. Databases Develop unified ways to combine soils and geologic information in vulnerability assessments. A tendency exists to consider only soil or only geologic information in vulnerability assessments. Both are important and need to be integrated in assessing vulnerability. Improve the chemical databases which are currently the source of much uncertainty in vulnerability assessments. It has been shown that for some measures of ground water vulnerability, the largest component of uncertainty involves the chemical aspects of transport. For example, the sorption process (expressed by "chemical" as Koc and "soil" as foc) has been found to produce large uncertainty in vulnerability assessments using the Attenuation or Retardation Factor approaches. The uncertainty in foc could be reduced by incorporating this parameter more systematically into current soil survey sampling. Determine the circumstances in which the properties of the intermediate vadose zone are critical to vulnerability assessments and develop methods for characterizing the zone for assessments. Research is needed to identify environmental situations where the reference or compliance surface must be below the root zone and where the base of the root zone is adequate. At present, soil surveys contain large amounts of information that can be used in vulnerability assessments, but very few data exist on the hydrologic, geochemical, and microbial properties of the intermediate vadose zone (the unsaturated zone below the root zone). Criteria need to be developed that can help to establish when the properties of the intermediate vadose zone will have little effect on vulnerability. For situations where the intermediate vadose zone cannot be ignored, methods should

OCR for page 170
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty be developed that allow systematic and inexpensive measurement of the properties of the zone required in vulnerability assessments. Perhaps in these instances, these data could be incorporated into a database consistent with the soils database of the Soil Conservation Service. One reviewer of this report commented that "Until we have a better quantitative handle on the biological processes, I think we have to assume that everything that escapes the root zone is going to ground water." Establish in the soil mapping standards of USDA's Soil Conservation Service an efficient soil sampling scheme for acquiring accurate soil attribute data in soil mapping unit polygons and documenting the uncertainty in these data. A need exists to better characterize the inclusions of other soil types in soil mapping units, including fractional area of included soil and distribution of inclusions. The quality of a vulnerability assessment is very dependent on the data employed. The uncertainty or variability of soil attribute data is critical in determining the uncertainty of the assessment. Equally important is the location and quantity of inclusions of material of differing types. In some settings, this knowledge may be as important as knowledge of preferential flow paths. For example, if a coarse sandy soil is included in a soil mapping unit (polygon) dominated by silty loam, the ground water vulnerability may depend more on the included sandy soil than on the dominant soil in the area. Establish reliable transfer functions for estimating in situ hydraulic properties, using available soil attribute data (e.g., bulk densities, particle-size distributions, etc.). Develop ways to determine the additional uncertainty arising from the use of transfer functions in ground water vulnerability assessments. Since the cost of sampling is large, methods that allow other, easy-to-obtain data to be used as surrogates for the required information are desirable. At some locations, such methods may be used to provide additional information when the required parameter has been measured. When these methods are developed, the uncertainty of using them should be determined. It may also be possible that using surrogate information, in comparison with using only the primary data, would reduce the overall uncertainty of results. Geoprocessing and Display Develop methods for merging data obtained at different spatial and temporal scales into a common scale for vulnerability assessment. It is highly unlikely that all data will be collected at the same spatial or temporal scale, especially data collected by different agencies for differing purposes. Therefore, it is very important to develop methods that permit data collected

OCR for page 170
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty at one scale to be transformed to a scale appropriate for a given vulnerability assessment. Improve analytical tools in GIS to facilitate the integration of assessment methods with spatial and attribute databases and the computing environment. Establish more meaningful categories of vulnerability for assessment methods. The issue of vulnerability classifications must be addressed before colors can be placed on a map in any meaningful way. The committee doubts such classifications can be used effectively unless the scheme used to develop them has some relevance to the assessment objective and they provide a valid measure of differences in the vulnerability of ground water. Assessment Methods Determine which processes are most important to incorporate into vulnerability assessments at different spatial scales. To determine ground water vulnerability accurately, the dominant processes at a given scale need to be identified and methods developed for characterizing them that can be used in modeling approaches. Obtain more information on the uncertainty associated with vulnerability assessments and develop ways to display this uncertainty. Methods are needed that can identify and differentiate among more sources of uncertainty. It is vital to provide information on the uncertainty of a vulnerability assessment. Current methods, however, only provide lower bounds on the uncertainty since they only take account of uncertainty from specific sources. As more effort is directed toward reducing uncertainty from known sources (e.g., sparse data), other uncertainties (e.g., model uncertainty) need to be evaluated. The task may prove formidable, since determining absolute uncertainty implicitly assumes some knowledge of absolute truth. Develop methods for accounting for soil macropores and other preferential flow pathways that can affect vulnerability. These investigations should include evaluations of the uncertainty in methods and measurements as they affect the assessment. Routes of transport that circumvent the porous media have a profound effect on transport and are difficult to quantify. Knowledge of these types of pathways could drastically alter the interpretation of an assessment made with traditional methods. Currently, an extensive research effort is devoted to the development of methods for characterizing and modeling the preferential flow process. However, there is no satisfactory method for predicting the effects of this mechanism. Because the ramifications of preferential flow are so large, additional research in this area is highly recommended.

OCR for page 170
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty Develop methods for incorporating process-based, statistical, and qualitative information into an integrated or hybrid assessment. Efforts should focus on developing rigorous approaches for making use of all available information so as to decrease uncertainty. Identify counterintuitive situations leading to a greater true vulnerability than commonly perceived. For example, develop greater understanding of the circumstances in which low-permeability materials that overlay aquifers can transmit contaminants to ground water. Some geohydrologic systems have characteristics that make them appear to have low vulnerability to contamination. Changes in management, however, may circumvent these characteristics, increasing the system's vulnerability. Likewise, some low-permeability materials overlying aquifers may transmit contaminants more easily than commonly perceived because of interconnected fracture systems. These counterintuitive situations typically would not be explicitly characterized in vulnerability assessments, generally because of the simplicity of current methods. Therefore, it would be helpful to document some common counterintuitive situations to warn decision makers and analysts of potential errors in assessing vulnerability.

OCR for page 170
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty This page in the original is blank.