Number 76 March 1996 ISSN 1062-9149
David K. Mueller, Pixie A. Hamilton, Dennis R. Helsel, Kerie J. Hitt, and Barbara C.
U.S. Geological Survey
In 1991, the U.S. Geological Survey (USGS) began full implementation of the National Water-Quality Assessment (NAWQA) Program. The objectives of the program are to:
Figure 1: National Water-Quality Assessment (NAWQA) Program study units
The data interpretation presented in the report from which this article is excerpted (Mueller et al., 1995) is based on nutrient analyses of ground and surface water samples collected through 1992 in the first 20 NAWQA study units (study units that began in FY 1991, see Figure 1). Therefore, the "national" analyses presented are based on 20 large areas, not the entire land surface of the United States. Although the results will undoubtedly improve as the NAWQA Program increases in spatial coverage, the range of conditions in the first 20 study units is sufficiently diverse to warrant consideration of these analyses as national in scope.
The nutrient synthesis describes broad regional and national patterns in nutrient concentrations, putting results of local investigations into a larger perspective, as well as showing influences that consistently affect nutrient concentrations throughout the country. Nutrients (nitrogen and phosphorus) were selected for analysis because they have been a long-standing national water quality issue and were identified as a local issue by water managers within each of the study units.
The nutrient synthesis involved analysis of existing water quality data collected before initiation of sampling in 1993 by the first 20 NAWQA study units. It is, therefore, a retrospective analysis, intended to determine pre-existing water quality conditions. Water-quality data were obtained by study-unit personnel from the USGS National Water Information System and from State and local water agencies. Some of the data were contained in the U.S. Environmental Protection Agency's (U.S. EPA) national data base, STORET. Sites were selected that had adequate samples during the period 1970-92 and that best represented land uses in the study unit. Ground-water data were supplemented by data from the Delmarva Peninsula NAWQA pilot study and the USGS-Toxic Substances Hydrology Program.
Regional and national comparisons of nutrient concentrations in ground and surface water were based on characteristics of the location of a ground-water sampling well or of the drainage basin upstream from a surface-water sampling site. Characteristics in the ancillary data set included land use, population, aquifer composition, and soil type.
Latitude and longitude of the sampling sites were the primary ancillary data used for location of ground-water and surface-water sample collection. In addition, latitude and longitude of surface-water sites were used to define the upstream drainage basin contributing to streamflow at the sites. Depth of the well and depth to water were used to determine the location of ground-water sampling below the land surface. Streamflow at the time of surface-water sampling was used to indicate the hydrologic condition during sample collection.
Classification of land use at ground-water sampling sites and upstream from surface-water sampling sites was based on categories described by Anderson et al. (1976). For ground-water sites, land-use settings were determined for individual well locations from USGS land-use and land-cover digital (GIRAS) data (Fegeas et al., 1983). Most of the GIRAS data were collected during 1970-80. Since that time, some areas of the Nation have experienced land-use changes. However, because of the time required for water movement from the land surface to the underlying aquifer, previous land use is relevant to current ground-water-quality assessments.
For surface-water drainage basins, where the traveltime from source areas to sampling sites is not as long, the GIRAS data were considered potentially obsolete and, therefore, inadequate for relating land use to nutrient concentrations in runoff. Classifications of surface-water sites were made based on the Anderson Level I categories (Anderson et al., 1976), aggregated as follows:
The national assessment of nutrients in ground water was based on analyses of approximately 12,000 water samples collected from wells in 18 of the 20 NAWQA study units and 5 supplemental study areas. Ground-water samples were selected for the national retrospective data set using specific criteria. Samples from wells around known or suspected areas of contamination were eliminated, as were samples from wells clustered in a small local area. Only one analysis (generally the most recent) from each well was included. Concentrations measured in the field were eliminated.
Nitrate was the only nutrient considered in the analysis of the ground-water data set because:
Data used in the national analysis of nutrients in surface water were obtained from data sets compiled for each study unit. Samples were limited to one per month. Sites were included only if they had a minimum of 25 samples with nutrient analyses. Sites that had an upstream drainage area of less than 5 square miles and samples collected from streamflows of less than 1 cubic foot per second were excluded.
Sites with very large (greater than 10,000 square miles) drainage areas, which are probably affected by many upstream factors, including land uses, were categorized as Large-Integrator sites and were analyzed separately. The resultant national retrospective data set included more than 22,000 samples from more than 300 sites.
Data analysis methods used were graphical and statistical. Data distributions were displayed graphically by using truncated boxplots (Helsel and Hirsch, 1992). Because many plotted distributions did not appear to be normal and differences in standard deviation were obvious from variations in spread, nonparametric statistical methods were chosen to test for differences among data groups. The Kruskal-Wallis test was used to test for differences in median values. This test was calculated by performing a one-way analysis of variance on the ranks of the data. If a significant difference among data-group medians was indicated, individual differences were evaluated by applying Tukey's multiple comparison test (Helsel and Hirsch, 1992) to the rank-transformed data.
Relations between nutrient concentrations and selected ancillary data were investigated by using nonparametric correlation analysis (Helsel and Hirsch, 1992). This technique identifies monotonic, although not necessarily linear, correlation. All statistical computations used in the data analysis were made by using commercially available statistics-software procedures (SAS Institute, 1990).
More types of ancillary data were available for ground-water sampling sites than for surface-water sites. For this reason, the ground-water data interpretation was organized to present differences in nutrient concentrations related to a variety of ancillary factors. In the surface-water section, the relation between nutrient concentrations and upstream land use was emphasized.
Analysis of the national retrospective data set indicated significant differences among nitrate concentrations in samples collected from different types of water-supply wells. Median nitrate concentrations in more than 5,600 samples ranged from 0.2 mg/L for public-supply wells to 2.4 mg/L for irrigation and stock wells. Concentrations of nitrate exceeded the drinking-water MCL (maximum contaminant level, which is 10 mg/L) most frequently in samples collected from irrigation and stock wells and in only about 1% of the samples collected from public-supply wells.
The significantly higher nitrate concentrations in samples collected at stock wells might be caused by watering, feeding, or corralling animals near the well. In addition, many stock wells are old and poorly maintained, allowing nitrates from manure to percolate down the well to the water table.
Broad regional patterns in the quality of ground water were evident. Nitrate concentrations were significantly higher in samples from the Northeastern, Northern Plains, and Pacific States than in samples from other regions.
Several factors affect the concentrations of nitrate in wells with regard to national patterns, including land use, depth below land surface, hydrogeologic setting, soil hydrologic group, depth to water, and type of agricultural land use. Findings of the nutrient analysis which are associated with three of these factors are discussed below.
Land Use: Nitrate concentrations were significantly different in ground water beneath different land-use settings. Key findings of the national analysis of ground-water quality beneath different land-uses include the following:
Direct comparison of nutrient concentrations in ground and surface water is limited using the national retrospective data sets because the sample collection was not designed for this purpose. However, a regional comparison illustrates some general patterns that deserve additional analysis as the NAWQA Program progresses. A regional pattern of nitrate concentrations in agricultural areas is shown in Figure 2. Median concentrations are plotted, with study units arranged from the Northeast on the right side of the figure to the West Coast on the left side. High concentrations in ground water occur in parts of the Northeast, the West, and along the West Coast. Many study units in these regions also have high nitrate concentrations in surface water. Ground- and surface-water nitrate concentrations generally were low throughout the Southeast and Midwest. The sole exception was high concentrations in surface water in the White River study unit (Indiana). Agricultural areas in that study unit are extensively tile-drained, which diverts seepage from the ground water and provides a quick path for nutrient-rich runoff to reach surface streams.
Figure 2: Median nitrate concentrations in ground and surface water from selected NAWQA study units (see Figure 1 for study unit names and abbreviations)
Nitrate concentrations in ground water were elevated primarily in agricultural areas and were highest in shallow ground water. Because of the time involved for ground water to move vertically in some areas, the full effect of nitrogen availability might not be noted in some aquifers for many years. Likewise, the effects of implementing management practices to improve water quality might not be evident for a number of years.
Nitrate concentrations also were elevated in surface water downstream from agricultural areas but were lower than those in ground water. Regional and local factors, such as soil and geologic characteristics, can affect nitrate concentrations and the partitioning of nitrate between ground and surface waters. Regional patterns and local characteristics could be useful in identifying areas of potential nitrate problems.
Nitrate concentrations exceed the MCL for drinking water in about 21% of the wells that tap the upper 100 feet below land surface in agricultural areas. Generally, nitrate does not pose a health risk for residents who drink water from deeper confined and bedrock aquifers. Nitrate in ground water is a greater concern in rural domestic-supply wells than in urban public-supply wells. Nitrate concentrations exceeded the MCL in only 1% of the sampled public-supply wells, mostly in agricultural areas of the Midwest and Northwest. Concentrations in surface-water samples rarely exceeded the MCL.
Elevated concentrations of ammonia and phosphorus in surface water occur primarily downstream from urban areas. During the 1970s, concentrations in surface water were high enough to warrant concerns about toxicity to fish and accelerated eutrophication in several parts of the United States. Recent improvements in wastewater treatment have decreased ammonia concentrations downstream from some urban areas, but the result has been an increase in nitrate concentrations. This condition limits the direct threat of toxicity but does not change the potential for eutrophication.
Information on environmental factors that affect water quality is useful to quickly and efficiently identify drainage basins with the greatest vulnerability for nutrient contamination and to delineate areas where ground- or surface-water contamination is most likely to occur. Determining where water quality problems are most likely to occur is the key to devising appropriate watershed-management strategies. These findings imply that management strategies need to incorporate some flexibility in different regions. For example, watershed management in the Southeast, where abundant rainfall, highly organic soils, ditching, and vegetation can minimize nitrate loading to ground water, generally will be different than watershed management in the Northeast, Midwest, and West, where ground water is more vulnerable to contamination. Understanding the regional distribution and key environmental factors that affect nutrient concentrations in ground and surface water is critical to implementing and evaluating Federal, State, and local programs designed to manage and protect water resources.
Dennis R. Helsel, Hydrologist U.S. Geological Survey 413 National Center, Reston, VA 22092 Tel: 703-648-5713 Fax: 703-648-6693 email: email@example.com
Copies of the report on which this article is based (WRI Report 95-4031) are available free from the NAWQA office at USGS until June 15, 1996, or until gone. Request a copy by email to firstname.lastname@example.org, or by fax at 703-648-6693. After June 15, 1996, copies may be purchased for $15.75, including postage, from USGS Earth Science Information Center Open-File Reports Section, Box 25286, Mail Stop 517, Denver Federal Center, Denver, CO 80225, Tel: 303-202-4200, Fax: 303-202-4695
Information regarding the National Water-Quality Assessment (NAWQA) Program is available on the Internet via the World Wide Web at:
Anderson, J.R., Hardy, E.E., Roach, J.T., and Witmer, R.E. 1976. A land use and land cover classification system for use with remote sensor data. U.S. Geological Survey Professional Paper 964, 28 p.
Fegeas, R.G., Claire, R.W., Guptill, S.C., Anderson, K.E., and Hallam, C.A. 1983. Land use and land cover digital data. U.S. Geological Survey Circular 895E, 21 p.
Hallberg, G.R., and Keeney, D.R. 1993. Nitrate, in Alley, W.M., ed., Regional ground-water quality. New York Van Nostrand Publishing Co., p. 297-322.
Helsel, D.R., and Hirsch, R.M. 1992. Statistical methods in water resources. Elsevier, Amsterdam, 522 p.
Mueller, D.K., Hamilton, P.A., Helsel, D.R., Hitt, K.J., and Ruddy, B.C. 1995. Nutrients in Ground Water and Surface Water of the United States -- An Analysis of Data Through 1992. Water-Resources Investigations Report 95-4031. U.S. Geological Survey, Denver, CO. 74 p.
SAS Institute. 1990. SAS/STAT user's guide (version 6, 4th ed.) SAS Institute Inc., Cary, NC. 2 vol., 1,686 p.
U.S. EPA. 1986. Quality criteria for water 1986. U.S. Environmental Protection Agency Report 440/586001, Office of Water, Washington, D.C..
Purpose: Provide technical and scientific support for nonpoint source (NPS) watershed projects with long-term land treatment and monitoring components.
Further Information: Patricia Lietman, U.S. Geological Survey, Tel: 717-730-6960, Fax: 717-730-6997, email: email@example.com
NWQEP NOTES is issued bimonthly. Subscriptions are free (contact: Publications Coordinator at the address below or via email at firstname.lastname@example.org). A list of publications on nonpoint source pollution distributed by the NCSU Water Quality Group is also available with each issue of the newsletter.
I welcome your views, findings, information, and suggestions for articles. Please feel free to contact me.
Judith A. Gale, Editor
Water Quality Extension Specialist
North Carolina State University Water Quality Group
615 Oberlin Road, Suite 100
Raleigh, NC 27605-1126