5
Case Studies

INTRODUCTION

This chapter presents six case studies of uses of different methods to assess ground water vulnerability to contamination. These case examples demonstrate the wide range of applications for which ground water vulnerability assessments are being conducted in the United States. While each application presented here is directed toward the broad goal of protecting ground water, each is unique in its particular management requirements. The intended use of the assessment, the types of data available, the scale of the assessments, the required resolution, the physical setting, and institutional factors all led to very different vulnerability assessment approaches. In only one of the cases presented here, Hawaii, are attempts made to quantify the uncertainty associated with the assessment results.

IOWA

Introduction

Ground water contamination became an important political and environmental issue in Iowa in the mid-1980s. Research reports, news headlines, and public debates noted the increasing incidence of contaminants in rural and urban well waters. The Iowa Ground water Protection Strategy (Hoyer et al. 1987) indicated that levels of nitrate in both private and municipal



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 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty 5 Case Studies INTRODUCTION This chapter presents six case studies of uses of different methods to assess ground water vulnerability to contamination. These case examples demonstrate the wide range of applications for which ground water vulnerability assessments are being conducted in the United States. While each application presented here is directed toward the broad goal of protecting ground water, each is unique in its particular management requirements. The intended use of the assessment, the types of data available, the scale of the assessments, the required resolution, the physical setting, and institutional factors all led to very different vulnerability assessment approaches. In only one of the cases presented here, Hawaii, are attempts made to quantify the uncertainty associated with the assessment results. IOWA Introduction Ground water contamination became an important political and environmental issue in Iowa in the mid-1980s. Research reports, news headlines, and public debates noted the increasing incidence of contaminants in rural and urban well waters. The Iowa Ground water Protection Strategy (Hoyer et al. 1987) indicated that levels of nitrate in both private and municipal

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty wells were increasing. More than 25 percent of the state's population was served by water with concentrations of nitrate above 22 milligrams per liter (as NO3). Similar increases were noted in detections of pesticides in public water supplies; about 27 percent of the population was periodically consuming low concentrations of pesticides in their drinking water. The situation in private wells which tend to be shallower than public wells may have been even worse. Defining the Question Most prominent among the sources of ground water contamination were fertilizers and pesticides used in agriculture. Other sources included urban use of lawn chemicals, industrial discharges, and landfills. The pathways of ground water contamination were disputed. Some interests argued that contamination occurs only when a natural or human generated condition, such as sinkholes or agricultural drainage wells, provides preferential flow to underground aquifers, resulting in local contamination. Others suggested that chemicals applied routinely to large areas infiltrate through the vadose zone, leading to widespread aquifer contamination. Mandate, Selection, and Implementation In response to growing public concern, the state legislature passed the Iowa Ground water Protection Act in 1987. This landmark statute established the policy that further contamination should be prevented to the "maximum extent practical" and directed state agencies to launch multiyear programs of research and education to characterize the problem and identify potential solutions. The act mandated that the Iowa Department of Natural Resources (DNR) assess the vulnerability of the state's ground water resources to contamination. In 1991, DNR released Ground water Vulnerability Regions of Iowa, a map developed specifically to depict the intrinsic susceptibility of ground water resources to contamination by surface or near-surface activities. This assessment had three very limited purposes: (1) to describe the physical setting of ground water resources in the state, (2) to educate policy makers and the public about the potential for ground water contamination, and (3) to provide guidance for planning and assigning priorities to ground water protection efforts in the state. Unlike other vulnerability assessments, the one in Iowa took account of factors that affect both ground water recharge and well development. Ground water recharge involves issues related to aquifer contamination; well development involves issues related to contamination of water supplies in areas where sources other than bedrock aquifers are used for drinking water. This

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty approach considers jointly the potential impacts of contamination on the water resource in aquifers and on the users of ground water sources. The basic principle of the Iowa vulnerability assessment involves the travel time of water from the land surface to a well or an aquifer. When the time is relatively short (days to decades), vulnerability is considered high. If recharge occurs over relatively long periods (centuries to millennia), vulnerability is low. Travel times were determined by evaluating existing contaminants and using various radiometric dating techniques. The large reliance on travel time in the Iowa assessment likely results in underestimation of the potential for eventual contamination of the aquifer over time. The most important factor used in the assessment was thickness of overlying materials which provide natural protection to a well or an aquifer. Other factors considered included type of aquifer, natural water quality in an aquifer, patterns of well location and construction, and documented occurrences of well contamination. The resulting vulnerability map (Plate 1) delineates regions having similar combinations of physical characteristics that affect ground water recharge and well development. Qualitative ratings are assigned to the contamination potential for aquifers and wells for various types and locations of water sources. For example, the contamination potential for wells in alluvial aquifers is considered high, while the potential for contamination of a variable bedrock aquifer protected by moderate drift or shale is considered low. Although more sophisticated approaches were investigated for use in the assessment, ultimately no complex process models of contaminant transport were used and no distinction was made among Iowa's different soil types. The DNR staff suggested that since the soil cover in most of the state is such a small part of the overall aquifer or well cover, processes that take place in those first few inches are relatively similar and, therefore, insignificant in terms of relative susceptibilities to ground water contamination. The results of the vulnerability assessment followed directly from the method's assumptions and underlying principles. In general, the thicker the overlay of clayey glacial drift or shale, the less susceptible are wells or aquifers to contamination. Where overlying materials are thin or sandy, aquifer and well susceptibilities increase. Vulnerability is also greater in areas where sinkholes or agricultural drainage wells allow surface and tile water to bypass natural protective layers of soil and rapidly recharge bedrock aquifers. Basic data on geologic patterns in the state were extrapolated to determine the potential for contamination. These data were supplemented by databases on water contamination (including the Statewide Rural Well-Water Survey conducted in 1989-1990) and by research insights into the transport, distribution, and fate of contaminants in ground water. Some of the simplest data needed for the assessment were unavailable. Depth-to-bedrock information had never been developed, so surface and bedrock topographic

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty maps were revised and integrated to create a new statewide depth-to-bedrock map. In addition, information from throughout the state was compiled to produce the first statewide alluvial aquifer map. All new maps were checked against available well-log data, topographic maps, outcrop records, and soil survey reports to assure the greatest confidence in this information. While the DNR was working on the assessment, it was also asked to integrate various types of natural resource data into a new computerized geographic information system (GIS). This coincident activity became a significant contributor to the assessment project. The GIS permitted easier construction of the vulnerability map and clearer display of spatial information. Further, counties or regions in the state can use the DNR geographic data and the GIS to explore additional vulnerability parameters and examine particular areas more closely to the extent that the resolution of the data permits. The Iowa vulnerability map was designed to provide general guidance in planning and ranking activities for preventing contamination of aquifers and wells. It is not intended to answer site-specific questions, cannot predict contaminant concentrations, and does not even rank the different areas of the state by risk of contamination. Each of these additional uses would require specific assessments of vulnerability to different activities, contaminants, and risk. The map is simply a way to communicate qualitative susceptibility to contamination from the surface, based on the depth and type of cover, natural quality of the aquifer, well location and construction, and presence of special features that may alter the transport of contaminants. Iowa's vulnerability map is viewed as an intermediate product in an ongoing process of learning more about the natural ground water system and the effects of surface and near-surface activities on that system. New maps will contain some of the basic data generated by the vulnerability study. New research and data collection will aim to identify ground water sources not included in the analysis (e.g., buried channel aquifers and the "salt and pepper sands" of western Iowa). Further analyses of existing and new well water quality data will be used to clarify relationships between aquifer depth and ground water contamination. As new information is obtained, databases and the GIS will be updated. Over time, new vulnerability maps may be produced to reflect new data or improved knowledge of environmental processes.

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty CAPE COD Introduction The Cape Cod sand and gravel aquifer is the U.S. Environmental Protection Agency (EPA) designated sole source of drinking water for Barnstable County, Massachusetts (ca. 400 square miles, winter population 186,605 in 1990, summer population ca. 500,000) as well as the source of fresh water for numerous kettle hole ponds and marine embayments. During the past 20 years, a period of intense development of open land accompanied by well-reported ground water contamination incidents, Cape Cod has been the site of intensive efforts in ground water management and analysis by many organizations, including the Association for the Preservation of Cape Cod, the U.S. Geological Survey, the Massachusetts Department of Environmental Protection (formerly the Department of Environmental Quality Engineering), EPA, and the Cape Cod Commission (formerly the Cape Cod Planning and Economic Development Commission). An earlier NRC publication, Ground Water Quality Protection: State and Local Strategies (1986) summarizes the Cape Cod ground water protection program. Defining the Question The Area Wide Water Quality Management Plan for Cape Cod (CCPEDC 1978a, b), prepared in response to section 208 of the federal Clean Water Act, established a management strategy for the Cape Cod aquifer. The plan emphasized wellhead protection of public water supplies, limited use of public sewage collection systems and treatment facilities, and continued general reliance on on-site septic systems, and relied on density controls for regulation of nitrate concentrations in public drinking water supplies. The water quality management planning program began an effort to delineate the zones of contribution (often called contributing areas) for public wells on Cape Cod that has become increasingly sophisticated over the years. The effort has grown to address a range of ground water resources and ground water dependent resources beyond the wellhead protection area, including fresh and marine surface waters, impaired areas, and water quality improvement areas (CCC 1991). Plate 2 depicts the water resources classifications for Cape Cod. Selection and Implementation of Approaches The first effort to delineate the contributing area to a public water supply well on Cape Cod came in 1976 as part of the initial background studies for the Draft Area Wide Water Quality Management Plan for Cape

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty Cod (CCPEDC 1978a). This effort used a simple mass balance ratio of a well's pumping volume to an equal volume average annual recharge evenly spread over a circular area. This approach, which neglects any hydrogeologic characteristics of the aquifer, results in a number of circles of varying radii that are centered at the wells. The most significant milestone in advancing aquifer protection was the completion of a regional, 10 foot contour interval, water table map of the county by the USGS (LeBlanc and Guswa 1977). By the time that the Draft and Final Area Wide Water Quality Management Plans were published (CCPEDC 1978a, b), an updated method for delineating zones of contribution, using the regional water table map, had been developed. This method used the same mass balance approach to characterize a circle, but also extended the zone area by 150 percent of the circle's radius in the upgradient direction. In addition, a water quality watch area extending upgradient from the zone to the ground water divide was recommended. Although this approach used the regional water table map for information on ground water flow direction, it still neglected the aquifer's hydrogeologic parameters. In 1981, the USGS published a digital model of the aquifer that included regional estimates of transmissivity (Guswa and LeBlanc 1981). In 1982, the CCPEDC used a simple analytical hydraulic model to describe downgradient and lateral capture limits of a well in a uniform flow field (Horsley 1983). The input parameters required for this model included hydraulic gradient data from the regional water table map and transmissivity data from the USGS digital model. The downgradient and lateral control points were determined using this method, but the area of the zone was again determined by the mass balance method. Use of the combined hydraulic and mass balance method resulted in elliptical zones of contribution that did not extend upgradient to the ground water divide. This combined approach attempted to address three-dimensional ground water flow beneath a partially penetrating pumping well in a simple manner. At about the same time, the Massachusetts Department of Environmental Protection started the Aquifer Lands Acquisition (ALA) Program to protect land within zones of contribution that would be delineated by detailed site-specific studies. Because simple models could not address three-dimensional flow and for several other reasons, the ALA program adopted a policy that wellhead protection areas or Zone IIs (DEP-WS 1991) should be extended upgradient all the way to a ground water divide. Under this program, wells would be pump tested for site-specific aquifer parameters and more detailed water table mapping would often be required. In many cases, the capture area has been delineated by the same simple hydraulic analytical model but the zone has been extended to the divide. This method has resulted in some 1989 zones that are 3,000 feet wide and extend 4.5

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty miles upgradient, still without a satisfactory representation of three-dimensional flow to the well. Most recently the USGS (Barlow 1993) has completed a detailed subregional, particle-tracking three-dimensional ground water flow model that shows the complex nature of ground water flow to wells. This approach has shown that earlier methods, in general, overestimate the area of zones of contribution (see Figure 5.1). In 1988, the public agencies named above completed the Cape Cod Aquifer Management Project (CCAMP), a resource-based ground water protection study that used two towns, Barnstable and Eastham, to represent the more and less urbanized parts of Cape Cod. Among the CCAMP products were a GIS-based assessment of potential for contamination as a result of permissible land use changes in the Barnstable zones of contribution (Olimpio et al. 1991) and a ground water vulnerability assessment by Heath (1988) using DRASTIC for the same area. Olimpio et al. characterized land uses by ranking potential contaminant sources without regard to differences in vulnerability within the zones. Heath's DRASTIC analysis of the same area, shown in Figure 5.2, delineated two distinct zones of vulnerability based on hydrogeologic setting. The Sandwich Moraine setting, with deposits of silt, sand and gravel, and depths to ground water ranging from 0 to more than 125 feet, had DRASTIC values of 140 to 185; the Barnstable Outwash Plain, with permeable sand and fine gravel deposits with beds of silt and clay and depths to ground water of less than 50 feet, yielded values of 185 to 210. The DRASTIC scores and relative contributions of the factors are shown in Tables 5.1 and 5.2. Heath concluded that similar areas of Cape Cod would produce similar moderate to high vulnerability DRASTIC scores. The CCAMP project also addressed the potential for contamination of public water supply wells from new land uses allowable under existing zoning for the same area. The results of that effort are shown in Plate 4. Results In summary, circle zones were used initially when the hydrogeologic nature of the aquifer or of hydraulic flow to wells was little understood. The zones improved with an understanding of ground water flow and aquifer characteristics, but in recognition of the limitations of regional data, grossly conservative assumptions came into use. Currently, a truer delineation of a zone of contribution can be prepared for a given scenario using sophisticated models and highly detailed aquifer characterization. However, the area of a given zone still is highly dependent on the initial assumptions that dictate how much and in what circumstances a well is pumped. In the absence of ability to specify such conditions, conservative assumptions,

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty FIGURE 5.1 Contributing areas of wells and ponds in the complex flow system determined by using the three-dimensional model with 1987 average daily pumping rates. (Barlow 1993) such as maximum prolonged pumping, prevail, and, therefore, conservatively large zones of contribution continue to be used for wellhead protection. The ground water management experience of Cape Cod has resulted in a better understanding of the resource and the complexity of the aquifer

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty FIGURE 5.2 DRASTIC contours for Zone 1, Barnstable-Yarmouth, Massachusetts. system, as well as the development of a more ambitious agenda for resource protection. Beginning with goals of protection of existing public water supplies, management interests have grown to include the protection of private wells, potential public supplies, fresh water ponds, and marine embayments. Public concerns over ground water quality have remained high and were a major factor in the creation of the Cape Cod Commission by the Massachusetts legislature. The commission is a land use planning and regulatory agency with broad authority over development projects and the ability to create special resource management areas. The net result of 20 years of effort by many individuals and agencies is the application of

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty TABLE 5.1 Ranges, Rating, and Weights for DRASTIC Study of Barnstable Outwash Plain Setting (NOTE: gpd/ft2 = gallons per day per square foot) (Heath 1988) Factor Range Rating Weight Number Depth to Water 0-50+ feet 5-10 5 25-50 Net Recharge Per Year 10+ inches 9 4 36 Aquifer Media Sand & Gravel 9 3 27 Soil Media Sand 9 2 18 Topography 2-6% 9 1 9 Impact of Vadose Zone Sand & Gravel 8 5 40 Hydraulic Conductivity 2000+ gpd/ft2 10 3 30         Total = 185-210 TABLE 5.2 Ranges, Rating, and Weights for DRASTIC Study of Sandwich Moraine Setting (NOTE: gpd/ft2 = gallons per day per square foot) (Heath 1988) Factor Range Rating Weight Number Depth to Water 0-100+ feet 1-10 5 5-50 Net Recharge Per Year 10+ inches 9 4 36 Aquifer Media Sand & Gravel 8 3 24 Soil Media Sandy Loam 6 2 12 Topography 6-12% 5 1 5 Impact of Vadose Zone Sand & Gravel 8 5 40 Hydraulic Conductivity 700-1000 gpd/ft2 6 3 18         Total = 140-185 higher protection standards to broader areas of the Cape Cod aquifer. With some exceptions for already impaired areas, a differentiated resource protection approach in the vulnerable aquifer setting of Cape Cod has resulted in a program that approaches universal ground water protection.

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty FLORIDA Introduction Florida has 13 million residents and is the fourth most populous state (U.S. Bureau of the Census 1991). Like several other sunbelt states, Florida's population is growing steadily, at about 1,000 persons per day, and is estimated to reach 17 million by the year 2000. Tourism is the biggest industry in Florida, attracting nearly 40 million visitors each year. Ground water is the source of drinking water for about 95 percent of Florida's population; total withdrawals amount to about 1.5 billion gallons per day. An additional 3 billion gallons of ground water per day are pumped to meet the needs of agriculture—a $5 billion per year industry, second only to tourism in the state. Of the 50 states, Florida ranks eighth in withdrawal of fresh ground water for all purposes, second for public supply, first for rural domestic and livestock use, third for industrial/commercial use, and ninth for irrigation withdrawals. Most areas in Florida have abundant ground water of good quality, but the major aquifers are vulnerable to contamination from a variety of land use activities. Overpumping of ground water to meet the growing demands of the urban centers, which accounts for about 80 percent of the state's population, contributes to salt water intrusion in coastal areas. This overpumping is considered the most significant problem for degradation of ground water quality in the state. Other major sources of ground water contaminants include: (1) pesticides and fertilizers (about 2 million tons/year) used in agriculture, (2) about 2 million on-site septic tanks, (3) more than 20,000 recharge wells used for disposing of stormwater, treated domestic wastewater, and cooling water, (4) nearly 6,000 surface impoundments, averaging one per 30 square kilometers, and (5) phosphate mining activities that are estimated to disturb about 3,000 hectares each year. The Hydrogeologic Setting The entire state is in the Coastal Plain physiographic province, which has generally low relief. Much of the state is underlain by the Floridan aquifer system, largely a limestone and dolomite aquifer that is found in both confined and unconfined conditions. The Floridan is overlain through most of the state by an intermediate aquifer system, consisting of predominantly clays and sands, and a surficial aquifer system, consisting of predominantly sands, limestone, and dolomite. The Floridan is one of the most productive aquifers in the world and is the most important source of drinking water for Florida residents. The Biscayne, an unconfined, shallow, limestone aquifer located in southeast Florida, is the most intensively used

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty FIGURE 5.6c The occurrence and distribution of ground water contamination on the Island of Maui. (Map provided by Hawaii State Department of Health.) FIGURE 5.6d The occurrence and distribution of ground water contamination on the Island of Kauai. (Map provided by Hawaii State Department of Health.)

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty because (1) the chemicals are highly sorbed in soils with high organic carbon contents, (2) the chemicals are highly volatile, and (3) the water table is several hundred meters below the surface. Measured concentrations of DBCP and EDB down to 30 meters at several locations have shown the original assessment to be wrong. They have resulted in an urgent need to understand processes such as preferential flow better and to predict if the replacement chemicals used today, such as Telon II, will also leach to significant depths. Leaching of pesticides to ground water in Hawaii could take decades. This time lag could lead to a temporary false sense of security, as happened in the past and potentially result in staggering costs for remedial action. For this reason, mathematical models that permit the user to ask ''what if" questions have been developed to help understand what the future may hold under certain management options. One needs to know what the fate of chemicals applied in the past will be and how to regulate the chemicals considered for use in the future; models are now being developed and used to help make these vulnerability assessments. Approaches Researchers have embarked on several parallel approaches to quantitatively assess the vulnerability of Hawaii's ground water resources, including: (1) sampling, (2) physically-based numerical modeling, and (3) vulnerability mapping based on a simple chemical leaching index. Taken together these approaches have provided insight and guidance for work on a complex, spatially and temporally variable problem. The sampling programs (Wong 1983 and 1987, Peterson et al. 1985) have shown that the chemicals applied in the past do, in fact, leach below the root zone, contrary to the original predictions, and can eventually reach the ground water. Experiments designed to characterize the nuances of various processes, such as volatilization, sorption, and degradation, have been conducted recently and will improve the conceptualization of mathematical models in the future. The EPA's Pesticide Root Zone Model (PRZM), a deterministic-empirical/conceptual fluid flow/solute transport model, has been tested by Loague and co-workers (Loague et al. 1989a, b; Loague 1992) against measured concentration profiles for DBCP and EDB in central Oahu. These simulations illustrate that the chemicals used in the past can indeed move to considerable depths. Models of this kind, once properly validated, can be used to simulate the predicted fate of future pesticide applications. One must always remember, however, that numerical simulations must be interpreted in terms of the limiting assumptions associated with model and data errors.

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty Ground water vulnerability maps and assessments of their uncertainty were pioneered at the University of Hawaii in the Department of Agriculture Engineering (Khan and Liang 1989, Loague and Green 1990a). These pesticide leaching assessments were made by coupling a simple mobility index to a geographic information system. Loague and coworkers have investigated the uncertainty in these maps owing to data and model errors (Loague and Green 1988; Loague et al. 1989c, 1990; Loague and Green 1990b, 1990c; Loague 1991; Kleveno et al. 1992; Yost et al. 1993). The Hawaiian database on soils, climate, and chemicals is neither perfect nor poor for modeling applications; it is typical of what exists in most states—major extrapolations are required to estimate the input parameters required for almost any chemical fate model. Results Sampling from wells in Hawaii has shown the concentrations of various chemicals, both from agriculture and industrial sources, which have leached to ground water in Hawaii. These concentrations, in general, are low compared to the levels detected in other states and for the most part are below health advisory levels established by EPA. In some instances contamination has not resulted from agriculture, but rather from point sources such as chemical loading and mixing areas and possibly from ruptured fuel lines. The widespread presence of trichloropropane (TCP) in Hawaii's ground water and deep soil cores at concentrations higher than DBCP was totally unexpected. TCP was never applied as a pesticide, but results from the manufacture of the fumigant DD, which was used until 1977 in pineapple culture. The occurrence of TCP illustrates that one must be aware of the chemicals applied as well as their components and transformation products. Wells have been closed in Hawaii even though the measured contaminant concentrations have been below those considered to pose a significant health risk. At municipal well locations in central Oahu, where DBCP, EDB, and/or TCP have been detected, the water is now passed through carbon filters before it is put into the distribution system. The cost of this treatment is passed on to the water users, rather than to those who applied the chemicals. The pesticide leaching assessment maps developed by Khan and Liang (1989) are intended for incorporation into the regulatory process. Decisions are not made on the basis of the red and green shaded areas for different chemicals (see Plate 3), but this information is considered. The uncertainty analysis by Loague and coworkers has shown some of the limitations of deterministic assessments in the form of vulnerability maps and provided initial guidance on data shortfalls.

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty APPLICATION OF A VULNERABILITY INDEX FOR DECISION-MAKING AT THE NATIONAL LEVEL Need for a Vulnerability Index A vulnerability index for ground water contamination by pesticides has been developed and used by USDA as a decision aid to help attain the objectives of the President's Water Quality Initiative (see Box 1.1). A vulnerability index was needed for use in program management and to provide insight for policy development. Motivation for the development of the vulnerability index was provided by two specific questions: Given limited resources and the geographic diversity of the water quality problems associated with agricultural production, what areas of the country have the highest priority for study and program implementation? What policy implications emerge from the spatial patterns of the potential for conamination from a national perspective, given information currently available about farming practices and chemical use in agriculture? Description of the Vulnerability Index A vulnerability index was derived to evaluate the likelihood of shallow ground water contamination by pesticides used in agriculture in one area compared to another area. Because of the orientation of Initiative policies to farm management practices, it was necessary that the vulnerability measure incorporate field level information on climate, soils, and chemical use. It also needed to be general enough to include all areas of the country and all types of crops grown. A Ground Water Vulnerability Index for Pesticides (GWVIP) was developed by applying the Soil-Pesticide Interaction Screening Procedure (SPISP) developed by the Soil Conservation Service to the National Resource Inventory (NRI) land use database for 1982 and the state level pesticide use database created by Resources for the Future (Gianessi and Puffer 1991). Details of the computational scheme and databases used are described by Kellogg et al. (1992). The 1982 NRI and the associated SOIL-5 database provide information on soil properties and land use at about 800,000 sample points throughout the continental United States. This information is sufficient to apply the SPISP to each point and thus obtain a relative measure of the soil leaching potential throughout the country. The RFF pesticide use database was used to infer chemical use at each point on the basis of the crop type recorded in the NRI database. By taking advantage of the statistical properties of the NRI database, which is based on a statistical survey

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty sampling design, the GWVIP score at each of the sample points can be statistically aggregated for making comparisons among regions. Since the GWVIP is an extension of a screening procedure, it is designed to minimize the likelihood of incorrectly identifying an area as having a low potential for contamination—that is, false negatives are minimized and false positives are tolerated. The GWVIP is designed to classify an area as having a potential problem even if the likelihood is small. GWVIP scores were graphically displayed after embedding them in a national cartographic database consisting of 13,172 polygons created by overlaying the boundaries of 3,041 counties, 189 Major Land Resource Areas (MLRAs), 2,111 hydrologic units, and federal lands. Three caveats are especially important in using the GWVIP and its aggregates as a decision aid: Land use data are for 1982 and do not represent current cropping patterns in some parts of the country. Although total cropland acreage has remained fairly stable over the past 10 years, there has been a pronounced shift from harvested cropland to cropland idled in government programs. The approach uses a simulation model that predicts the amount of chemical that leaches past the root zone. In areas where the water table is near the surface, these predictions relate directly to shallow ground water contamination. In other areas a time lag is involved. No adjustment was made for areas with deep water tables. No adjustment in chemical use is made to account for farm management factors, such as chemical application rates and crop rotations. The approach assumes that chemical use is the same for a crop grown as part of a rotation cropping system as for continuous cropping. Since the chemical use variable in the GWVIP calculation is based on acres of land treated with pesticides, application rates are also not factored into the analysis. Application to Program Management By identifying areas of the country that have the highest potential for leaching of agrichemicals, the GWVIP can serve as a basis for selecting sites for implementation of government programs and for more in-depth research on the environmental impact of agrichemical use. These sites cannot be selected exclusively on the basis of the GWVIP score, however, because other factors, such as surface water impacts and economic and demographic factors, are also important. For example, the GWVIP has been used as a decision aid in selecting sites for USDA's Area Study Program, which is designed to provide chemical use and farming practice information to aid in understanding the relationships among farming activities, soil properties, and ground water quality.

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty The National Agricultural Statistics Service interviews farm operators in 12 major watersheds where the U.S. Geological Survey is working to measure the quality of surface and ground water resources under its National Water Quality Assessment Program. At the conclusion of the project, survey information will be combined with what is learned in other elements of the President's Water Quality Initiative to assess the magnitude of the agriculture-related water quality problem for the nation as a whole and used to evaluate the potential economic and environmental effects of Initiative policies of education, technical assistance, and financial assistance if implemented nationwide. To meet these objectives, each Area Study site must have a high potential for ground water contamination relative to other areas of the country. A map showing the average GWVIP for each of the 13,172 polygons comprising the continental United States, shown in Plate 3, was used to help select the sites. As this map shows, areas more likely to have leaching problems with agrichemicals than other areas of the country occur principally along the coastal plains stretching from Alabama and Georgia north to the Chesapeake Bay area, the corn belt states, the Mississippi River Valley, and the irrigated areas in the West. Sites selected for study in 1991 and 1992 include four from the eastern coastal plain (Delmarva Peninsula, southeastern Pennsylvania, Virginia and North Carolina, and southern Georgia), four from the corn belt states (Nebraska, Iowa, Illinois, and Indiana), and two from the irrigated areas in the West (eastern Washington and southeastern Idaho). Four additional sites will be selected for study in 1993. Application to Policy Analysis and Development The GWVIP has also been used by USDA to provide a national perspective on agricultural use of pesticides and the potential for ground water contamination to aid in policy analysis and development. The geographic distribution of GWVIP scores has shown that the potential for ground water contamination is diverse both nationally and regionally. Factors that determine intrinsic vulnerability differ in virtually every major agricultural region of the country. Whether an impact is realized in these intrinsically vulnerable areas depends on the activities of producers—such as the type of crop planted, chemical use, and irrigation practices—which also vary both nationally and regionally. High vulnerability areas are those where a confluence of these factors is present. But not all cropland is vulnerable to leaching. About one-fourth of all cropland has GWVIP scores that indicate very low potential for ground water contamination from the use of agrichemicals. Nearly all agricultural states have significant acreage that meets this low vulnerability criterion. Areas of the country identified as being in a high vulnerability group relative to potential

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty for agrichemical leaching also have significant acreages that appear to have low vulnerability. This mix of relative vulnerabilities both nationally and regionally has important policy implications. With the potential problem so diverse, it is not likely that simple, across-the-board solutions will work. Simple policies—such as selective banning of chemicals—may reduce the potential for ground water contamination in problem areas while imposing unnecessary costs on farming in nonproblem areas. The geographic diversity of the GWVIP suggests that the best solutions will come from involvement of both local governments and scientists with their state and national counter-parts to derive policies that are tailored to the unique features of each problem area. In the future, USDA plans to use vulnerability indexes, like the GWVIP, in conjunction with economic models to evaluate the potential for solving agriculture-related water quality problems with a nationwide program to provide farmers with the knowledge and technical means to respond voluntarily to water quality concerns. SUMMARY These six case studies illustrate how different approaches to vulnerability assessment have evolved under diverse sets of management requirements, data constraints, and other technical considerations. In addition, each of these examples shows that vulnerability assessment is an ongoing process through which information about a region's ground water resources and its quality can be organized and examined methodically. In Iowa, the Iowa DNR staff elected to keep their vulnerability characterization efforts as simple as possible, and to use only properties for which data already existed or could be easily checked. They assumed that surficial features such as the soil are too thin and too disrupted by human activities (e.g., tillage, abandoned wells) to provide effective ground water protection at any particular location and sought to identify a surrogate measure for average travel time from the land surface to the aquifer. Thus, a ground water vulnerability map was produced which represents vulnerability primarily on the basis of depth to ground water and extent of overlying materials. Wells and sinkholes are also shown. The results are to be used for informing resource managers and the public of the vulnerability of the resource and to determine the type of information most needed to develop an even better understanding of the vulnerability of Iowa's ground water. The Cape Cod approach to ground water vulnerability assessment is perhaps one of the oldest and most sophisticated in the United States. Driven by the need to protect the sole source drinking water aquifer underlying this sandy peninsula, the vulnerability assessment effort has focused on the identification

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty and delineation of the primary recharge areas for the major aquifers. This effort began with a simple mass balance approach which assumed even recharge within a circular area around each drinking water well. It has since evolved to the development of a complex, particle-tracking three-dimensional model that uses site-specific data to delineate zones of contribution. Bolstered by strong public concern, Cape Cod has been able to pursue an ambitious and sophisticated agenda for resource protection, and now boasts a sophisticated differential management ground water protection program. In Florida, ground water resource managers rely on a combination of monitoring and vulnerability assessment techniques to identify high recharge areas the develop the state ground water protection program. Overlay and index methods, including several modified DRASTIC maps were produced to identify areas of ground water significance in support of decision making in state land acquisition programs aimed at ground water protection. In addition, several monitoring networks have been established to assess background water quality and monitor actual effects in areas identified as highly vulnerable. The coupling of ground water vulnerability assessments with monitoring and research efforts, provides the basis of an incremental and evolving ground water protection program in Florida. The programs to protect ground water in California's intensely agricultural San Joaquin Valley are driven largely by compliance with the state Pesticide Contamination Prevention Act. The California Department of Pesticide Regulation determined that no model would be sufficient to cover their specific regulatory needs and that the available data bases were neither suitable nor reliable for regulatory purposes. Thus, a ground water protection program was built on the extensive existing pesticide use reporting system and the significant ground water monitoring requirements of the act. Using farm sections as management units, the state declares any section in which a pesticide or its degradation product is detected as a pesticide management zone and establishes further restrictions and monitoring requirements. Thus, the need to devise a defensible regulatory approach led California to pursue a mechanistic monitoring based approach rather than a modeling approach that would have inherent and difficult to quantify uncertainties. In contrast, the approach taken in Hawaii involves an extensive effort to understand the uncertainty associated with the assessment models used. The purpose of this is to provide guidance to, but not the sole basis for, the pesticide regulation program. The combined use of sampling, physically-based numerical modeling, and a chemical leaching index has led to extensive improvements in the understanding of the fate of pesticides in the subsurface environment. Uncertainty analyses are used to determine where additional information would be most useful.

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty Finally, USDA's Ground Water Vulnerability Index for Pesticides illustrates a national scale vulnerability assessment developed for use as a decision aid and analytical tool for national policies regarding farm management and water quality. This approach combines nationally available statistical information on pesticide usage and soil properties with a simulation model to predict the relative likelihood of contamination in cropland areas. USDA has used this approach to target sites for its Area Study Program which is designed to provide information to farmers about the relationships between farm management practices and water quality. The results of the GWVIP have also indicated that, even at the regional level, there is often an mix of high and low vulnerability areas. This result suggests that effective ground water policies should be tailored to local conditions. REFERENCES Au, L.K.L. 1991. The Relative Safety of Hawaii's Drinking Water. Hawaii Medical Journal 50(3): 71-80. Barlow, P.M. 1993. Particle-Tracking Analysis of Contributing Areas of Public-Supply Wells in Simple and Complex Flow Systems, Cape Cod, Massachusetts. USGS Open File Report 93-159. Marlborough, Massachusetts: U.S. Geological Survey. Britt, J.K., S.E. Dwinell, and T.C. McDowell. 1992. Matrix decision procedure to assess new pesticides based on relative ground water leaching potential and chronic toxicity. Environ. Toxicol. Chem. 11: 721-728. Cape Cod Commission (CCC). 1991. Regional Policy Plan. Barnstable, Massachusetts: Cape Cod Commission. Cape Cod Planning and Economic Development Commission (CCPEDC). March 1978a. Draft Area Wide Water Quality Management Plan for Cape Cod. Barnstable, Massachusetts: Cape Cod Commission. Cape Cod Planning and Economic Development Commission (CCPEDC). September 1978b. Final Area Wide Water Quality Management Plan for Cape Cod. Barnstable, Massachusetts: Cape Cod Commission. Department of Environmental Protection, Division of Water Supply (DEP-WS). 1991. Guidelines and Policies for Public Water Supply Systems. Massachusetts Department of Environmental Protection. Gianessi, L.P., and C.A. Puffer. 1991. Herbicide Use in the United States: National Summary Report. Washington, D.C.: Resources for the Future. Guswa, J.H., and D.R. LeBlanc. 1981. Digital Models of Ground water Flow in the Cape Cod Aquifer System, MA. USGS Water Supply Paper 2209. U.S. Geological Survey. Heath, D.L. 1988. DRASTIC mapping of aquifer vulnerability in eastern Barnstable and western Yarmouth, Cape Cod, Massachusetts. In Appendix D, Cape Cod Aquifer Management Project, Final Report, G.A. Zoto and T. Gallagher, eds. Boston: Massachusetts Department of Environmental Quality Engineering. Horsely, S.W. 1983. Delineating zones of contribution of public supply wells to protect ground water . In Proceedings of the National Water Well Association Eastern Regional Conference, Ground-Water Management, Orlando, Florida. Hoyer, B.E. 1991. Ground water vulnerability map of Iowa. Pp. 13-15 in Iowa Geology, no. 16. Iowa City, Iowa: Iowa Department of Natural Resources.

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty Hoyer, B.E., J.E. Combs, R.D. Kelley, C. Cousins-Leatherman, and J.H. Seyb. 1987. Iowa Ground water Protection Strategy. Des Moines: Iowa Department of Natural Resources. Kellogg, R.L., M.S. Maizel, and D.W. Goss. 1992. Agricultural Chemical Use and Ground Water Quality: Where Are the Potential Problems? Washington, D.C.: U.S. Department of Agriculture, Soil Conservation Service. Khan, M.A., and T. Liang. 1989. Mapping pesticide contamination potential. Environmental Management 13(2):233-242. Kleveno, J.J., K. Loague, and R.E. Green. 1992. An evaluation of a pesticide mobility index: Impact of recharge variation and soil profile heterogeneity. Journal of Contaminant Hydrology 11(1-2):83-99. Land Acquisition Advisory Council (LAAC). 1991. Ground Water Resources Committee Final Report: Florida Preservation 2000 Needs Assessment. Tallahassee, Florida: Department of Environmental Regulation. 39 pp. Lau, L.S., and J.F. Mink. 1987. Organic contamination of ground water: A learning experience. J. American Water Well Association 79(8):37-42. LeBlanc, D.R., and J.H. Guswa. 1977. Water-Table Map of Cape Cod, MA. May 23-27, 1976, USGS Open File Report 77-419, scale 1:48,000. Loague, K. 1991. The impact of land use on estimates of pesticide leaching potential: Assessments and uncertainties. Journal of Contaminant Hydrology 8: 157-175. Loague, K. 1992. Simulation of organic chemical movement in Hawaii soils with PRZM: 3. Calibration. Pacific Science 46(3):353-373. Loague, K.M., and R.E. Green. 1988. Impact of data-related uncertainties in a pesticide leaching assessment. Pp. 98-119 in Methods for Ground Water Quality Studies, D.W. Nelson and R.H. Dowdy, eds. Lincoln, Nebraska: Agricultural Research Division, University of Nebraska. Loague, K., and R.E. Green. 1990a. Comments on "Mapping pesticide contamination potential," by M.A. Khan and T. Liang. Environmental Management 4:149-150. Loague, K., and R.E. Green. 1990b. Uncertainty in Areal Estimates of Pesticide Leaching Potential. Pp. 62-67 in Transactions of 14th International Congress of Soil Science. Kyoto, Japan: International Soil Science Society. Loague, K., and R.E. Green. 1990c. Criteria for evaluating pesticide leaching models. Pp. 175-207 in Field-Scale Water and Solute Flux in Soils, K. Roth, H. Flühler, W.A. Jury, and J.C. Parker, eds. Basel, Switzerland: Birkhauser Verlag. Loague, K.M., R.E. Green, C.C.K. Liu, and T.C. Liang. 1989a. Simulation of organic chemical movement in Hawaii soils with PRZM: 1. Preliminary results for ethylene dibromide. Pacific Science 43(1):67-95. Loague, K., T.W. Giambelluca, R.E. Green, C.C.K. Liu, T.C. Liang, and D.S. Oki. 1989b. Simulation of organic chemical movement in Hawaii soils with PRZM: 2. Predicting deep penetration of DBCP, EDB, and TCP. Pacific Science 43(4):362-383. Loague, K.M., R.S. Yost, R.E. Green, and T.C. Liang. 1989c. Uncertainty in a pesticide leaching assessment for Hawaii. Journal of Contaminant Hydrology 4:139-161. Loague, K., R.E. Green, T.W. Giambelluca, T.C. Liang, and R.S. Yost. 1990. Impact of uncertainty in soil, climatic, and chemical information in a pesticide leaching assessment. Journal of Contaminant Hydrology 5:171-194. Maddox, G., and J. Spicola. 1991. Ground Water Quality Monitoring Network. Tallahassee, Florida: Florida Department of Environmental Regulation. 20 pp. Maddox, G., J. Lloyd, T. Scott, S. Upchurch, and R. Copeland, eds. 1993. Florida's Ground Water Quality monitoring Program: Background Hydrogeochemistry. Florida Geological Survey Special Publication #34. Tallahassee, Florida: Florida Department of Environmental Regulation in cooperation with Florida Geological Survey.

OCR for page 135
Ground Water Vulnerability Assessment: Contamination Potential Under Conditions of Uncertainty National Research Council (NRC). 1986. Ground Water Quality Protection: State and Local Strategies. Washington, D.C.: National Academy Press. Oki, D.S., and T.W. Giambelluca. 1985. Subsurface Water and Soil Quality Data Base for State of Hawaii: Part 1. Spec. Rept. 7. Manoa, Hawaii: Water Resources Research Center, University of Hawaii at Manoa. Oki, D.S., and T.W. Giambelluca. 1987. DBCP, EDB, and TCP contamination of ground water in Hawaii. Ground Water 25:693-702. Olimpio, J.C., E.C. Flynn, S. Tso, and P.A. Steeves. 1991. Use of a Geographic Information System to Assess Risk to Ground-Water Quality at Public-Supply Wells, Cape Cod, Massachusetts. Boston, Massachusetts: U.S. Geological Survey. Peterson, F.L. 1972. Water development on tropic volcanic islands—Type example: Hawaii. Ground Water 5:18-23. Peterson, F.L., K.R. Green, R.E. Green, and J.N. Ogata. 1985. Drilling program and pesticide analysis of core samples from pineapple fields in central Oahu. Water Resources Research Center, University of Hawaii at Manoa, Special Report 7.5. Photocopy. Southwest Florida Water Management Districts (SFWMD). 1991. The Bluebelt Commission. Brooksville, Florida: Southwest Florida Water Management Districts. U.S. Bureau of the Census. 1991. Statistical Abstracts of the United States: 1991, 111th edition. Washington, D.C.: U.S. Government Printing Office. Wong, L. 1983. Preliminary report on soil sampling EDB on Oahu. Pesticide Branch, Div. of Plant Industry, Department of Agriculture, State of Hawaii. Photocopy. Wong, L. 1987. Analysis of ethylene dibromide distribution in the soil profile following shank injection for nematode control in pineapple culture. Pp. 28-40 in Toxic Organic Chemicals in Hawaii's Water Resources, P.S.C. Rao and R.E. Green, eds. Ser. 086. Honolulu: Hawaii Inst. Trop Agric. Hum. Resources Res. Exten. University of Hawaii. Yost, R.S., K. Loague, and R.E. Green. 1993. Reducing variance in soil organic carbon estimates—soil classification and geostatistical approaches. Geoderma 57(3):247-262