Decadal Change in Groundwater Quality

About 140 million people—almost one-half of the Nation's population—rely on groundwater for public supply, and demand for groundwater for public supply, irrigation and agriculture continues to increase.

This mapper shows how concentrations of pesticides, nutrients, metals, and organic contaminants in groundwater are changing during decadal periods across the Nation.

Tracking changes in groundwater quality and investigating the reasons for these changes is crucial for informing management decisions to protect and sustain our valuable groundwater resources.

IMAGE: An image of a USGS scientist collecting data.

Learn More

Evaluating Decadal Changes in Groundwater Quality

Groundwater-quality data were collected from 5,000 wells between 1988-2001 (first decadal sampling event) by the U.S. Geological Survey's National Water-Quality Program. Samples are collected in groups of 20-30 wells with similar characteristics called networks. About 1,500 of these wells in 67 networks were sampled again approximately 10 years later between 2002-2012 (second sampling event) to evaluate decadal changes in groundwater quality. Between 2012 and 2022 (third sampling event), a subset of these networks was sampled again, allowing additional results to be displayed on the web page: Decadal changes in groundwater quality. This is the seventh iteration of data added to the website. With the additional data, it is possible to evaluate changes in water quality between the 2nd and 3rd sampling events for 76 networks, changes in water quality between the 1st and 3rd sampling events for 63 networks, and changes across all 3 sampling events for 60 networks. Samples were obtained from monitoring wells, domestic-supply wells, and some public-supply wells before any treatment on the system.

Groundwater samples used to evaluate decadal change were collected from networks of wells with similar characteristics. Some networks, consisting of domestic or public supply wells, were used to assess changes in the quality of groundwater used for drinking water supply. Other networks, consisting of monitoring wells, assessed changes in the quality of shallow groundwater underlying key land-use types such as agricultural or urban lands. Networks were chosen based on geographic distribution across the Nation and to represent the most important water-supply aquifers and specific land-use types.

Decadal changes in groundwater quality were assessed for these constituents:

Inorganics
Arsenic
Boron
Chloride
Fluoride
Iron
Lithium
Manganese
Molybdenum
Nitrate
Orthophosphate
Radium
Sodium
Strontium
Sulfate
Total Dissolved Solids
Uranium

Organics
Atrazine
Chloroform
Deethylatrazine
Dieldrin
Methyl tert-butyl ether
Metolachlor
Prometon
Simazine
Tetrachlorothene
Toluene
Trichloroethene
1,2-Dibromo-3-chloropropane (DBCP)

For more information about constituents and why they were chosen click here.

Online Mapper Showing Decadal Changes in Groundwater Quality

Decadal changes in concentrations of nutrients, metals, and pesticides and other organic contaminants in groundwater were evaluated in a total of 89 networks across the Nation by comparing changes between selected sampling events.

Decadal changes in median concentrations for a network are classified as large, small, or no change in comparison to a benchmark concentration. For example, a large change in chloride concentrations indicates that the probability of the test is ≦ 0.10 and the median of all differences in concentrations in a network is greater than 5 percent of the chloride benchmark per decade. For chloride, which has a Secondary Maximum Contaminant level of 250 milligrams per liter, this would mean the change in concentration exceeded 12.5 milligrams per liter (mg/L), or 5 percent of the benchmark.

Decadal Groundwater Quality Networks

230 networks were sampled from 1988 to 2001 to assess the status of the Nation's groundwater quality. Each dot on the above map represents the center point of a network of about 20 to 30 wells. Networks sampled in the first sampling event only are shown in green. There were 67 networks resampled from 2002 to 2012 to assess decadal changes in groundwater quality. Networks sampled from 2012 to 2022 and at least one previous sampling event are shown in orange. Networks sampled in the first and second sampling events but are no longer being sampled are shown in gray.

Related Resources

This software has been approved for release by the U.S. Geological Survey (USGS). Although the software has been subjected to rigorous review, the USGS reserves the right to update the software as needed pursuant to further analysis and review. No warranty, expressed or implied, is made by the USGS or the U.S. Government as to the functionality of the software and related material nor shall the fact of release constitute any such warranty. Furthermore, the software is released on condition that neither the USGS nor the U.S. Government shall be held liable for any damages resulting from its authorized or unauthorized use.

Version 4.0.1

Updates to Decadal Changes in Groundwater Quality

This web mapping application is based on a data release that updates Decadal changes in groundwater quality: U.S. Geological Survey Web page, Lindsey and others, 2018. The initial data release documented decadal changes in groundwater quality for 67 networks that had been sampled by the National Water-Quality Program. The results were based on analysis of 1,520 wells in the 67 networks that had been sampled during 1988-2001 and again between 2002-2012. The U.S. Geological Survey's National Water-Quality Assessment Project ended in 2021, but monitoring is continuing as a part of the National Water Quality Program. The table below shows the numbers of wells and networks that can be compared across any 2 or 3 time periods:

	Sample pairs in Decade 1 and 2	Sample pairs in Decade 1 and 3	Sample pairs in Decade 2 and 3	Sample sets in Decade 1, 2, and 3	Number of wells or networks with at least 2 decadal sampling pairs
Numbers of wells sampled	1,520	1,202	1,659	1,617	2,277
Numbers of Networks sampled	67	63	76	60	90

New data added from the time period 2022 include: 34 wells sampled in 2 networks that could be compared to the same wells and networks sampled during 1988-2001; 65 wells sampled in 3 networks that could be compared to the same wells and networks sampled during 2002-2012; and 52 wells in 2 networks with sufficient data to evaluate changes across all three events. The results across multiple time periods allow the user to select the time period of interest for display. The comparison of 1988-2001 to 2002-2012 displays the same results as all previous versions, with exception of the errata listed below. The new options allow the user to select comparisons for new time periods: 1988-2001 to 2012-2021, 2002-2012 to 2012-2022, and also results for data collected across all three time periods. The updated map has results for 2,277 wells in 90 networks. The list of constituents analyzed was updated in 2020 to include radium, which is now included on the map for the 2nd to 3rd time interval. Previous time intervals lack sufficient data to evaluate changes in radium concentrations. The criteria for mapping, benchmarks, data preparation, and statistical analysis are described in the Technical Information tab. In order to make the evaluations of the magnitude of change comparable, the magnitude of change is displayed based the change in concentration per decade. For comparisons between the first to third sampling period (approximately two decades), the median change is divided by two. For comparisons across three or more time periods, the statistical test result of change in concentration per year is multiplied by ten.

Change from versions prior to 2022:

The network sanasus1 result for orthophosphate was previously incorrectly reported as a small decrease from the first to the second decade. This should have been reported as a small increase and is correctly reported in this update.

Constituent	Statistically significant finding of increase, decrease, or no change in concentration for this network. Details of statistical analysis are available in the documentation. Small and large changes are defined in the explanation.
Network type	Major aquifer studies target drinking water wells in a selected aquifer. Land use studies target wells that underlie areas of urban or agricultural land use in a selected aquifer, and are typically shallower than wells sampled in major aquifer studies.
Types of wells	If no qualifier is listed, network is entirely of one well type. 'Predominantly' indicates that 80 percent of the wells in the network are of that type. Mixed networks list those well types making up at least 75 percent of the wells in the network. Possible well types include commercial, domestic, industrial, irrigation, monitoring, public-supply, stock, recreational, and other.
Typical depth range	Range of well depths listed are first and third quartile for the network.
Principal aquifer	Aquifer names are from the map of the principal aquifers of the United States (U.S. Geological Survey, 2003, http://water.usgs.gov/ogw/aquifer/map.html
Regional aquifer	The local or regional name for the aquifer sampled.
Aquifer material	Aquifer materials are from the map of the principal aquifers of the United States (U.S. Geological Survey, 2003, http://water.usgs.gov/ogw/aquifer/map.html
Additional information	Lists the specific land use activity, if available.
Network code	The network name acronym used by the USGS as described in the supporting data release.
Sample dates (1^st, 2^nd, 3^rd)	The year of the first, second, and third sampling event.

Study scope

1. What is the purpose of this study?

The purpose of this study is to determine if the quality of the Nation's groundwater has become better, worse, or stayed the same during 1988-2021. Evaluating changes in groundwater quality at the decadal scale is one component of a larger effort to understand the quality of the Nation's water resources and how water quality conditions are changing with time.

2. Who completed this study?

The U.S. Geological Survey's (USGS) National Water-Quality Program (NWQP) completed this study. Sampling teams in USGS Water Science Centers across the country collected the samples and project staff analyzed the data.

3. Why evaluate changes in water quality?

Groundwater is the source of drinking water for about one-half of the population of the United States and has many other important uses, such as for irrigation. It is important to document how the quality of this vital resource is changing in order to understand how management practices and ongoing environmental stresses may be affecting groundwater quality. Degradation of this resource can happen during a long time period, and recovery from historic contamination can also take a long time.

4. Does this map represent all changes in groundwater resources across the country?

The samples used to create this map come from 27 principal aquifers that account for more than 90 percent of the groundwater used for public water supply in the United States; however, the wells that were sampled are from selected geographic parts of each aquifer and, in some cases, also represent a specific land-use type. In addition, most of the samples are from monitoring wells and domestic-supply wells, which represent the shallower parts of the aquifer. As such, these results apply to the depth zone and setting where the samples were collected but also provide insight on possible future changes in water quality in the deeper zones of these aquifers that is used for public supply.

5. What percentage of the population of the United States relies upon groundwater for their drinking water?

Groundwater is the source of drinking-water supply for 140 million people - nearly one-half of the Nation's population (Maupin and others, 2014).

6. How many wells were sampled for this study, and what time period is represented?

The dataset consists of 2,277 wells in 90 groups of wells called networks in various principal aquifers across the country from the U.S. Geological Survey National Water-Quality Program. A network typically is a group of 20-30 wells representing an aquifer (major aquifer study), or a specific depth and (or) land use (land-use study) (Lapham and others, 1995). Each network has been sampled at least two times during the periods of 1988-2001, 2002-12, and 2012-21.

7. What does the term "Decade" mean as used in the "Trend Period" selector?

The National Water-Quality Assessment project evaluates changes groundwater quality at the network level on approximately a 10-year time scale. The first decade, previously referred to in the program as a "cycle", was from 1988 to 2001 (including pilot studies). The second decade was between 2001 and 2012. And the third decade is from 2012 to 2021. Additional data will be added to the third decade as data are approved and statistical analysis completed.

8. How did you determine which constituents to display on the map?

More than 300 constituents were sampled at most of the wells during the two decades. Constituents were prioritized for analysis of decadal change if they met the following criteria: (1) they exceeded a human health benchmark in at least 1 percent of the wells used as a source of drinking water from public-supply wells (Toccalino and Hopple, 2010) or domestic-supply wells (Desimone, 2009), (2) they exceeded a U.S. Environmental Protection Agency (EPA) secondary maximum contaminant level (SMCL) (U.S. Environmental Protection Agency, 2018) in at least 1 percent of the wells used as a source of drinking water (U.S. Environmental Protection Agency, 2018), (3) they were among the five most frequently detected volatile organic compounds (VOCs) in the Nation (Zogorski and others, 2006), or (4) they were among the five most frequently detected pesticides in the Nation (Gilliom and others, 2006). Other constituents were added to this list based on regional importance. Radon and gross alpha (α) activity met the criteria for analysis but do not have sufficient data for analysis; thus, they are not included in the mapping tool. In all, 28 constituents were selected for analysis of decadal change and inclusion in the mapping tool. Constituents are only displayed for time periods where sufficient data were available and analytical methods were comparable across the sampling events. For a listing of the 28 constituents selected for analysis and the reason they were selected, click here.

9. Where can I find help navigating this web page?

An explanation of the arrows and icons on the map and what they mean is here. A description of how to activate map layers is here. A description of the search, zoom, and print capabilities is here.

Features of the Mapper

10. Are the data available for download?

The data used to make these maps are available for download here: https://www.sciencebase.gov/catalog/item/6476425ed34e4e58932da21f

11. What are groundwater networks and what do they represent?

Networks are groups of wells with similar characteristics. Some are designed to give a broad overview of groundwater quality in an aquifer used as a source of drinking-water supply and others are designed to examine the factors that affect the quality of shallow groundwater underlying key types of land use. Networks were chosen for decadal-scale water-quality sampling based on geographic distribution across the Nation and to represent the most important aquifers and specific land-use types. A network typically is a group of 20-30 wells representing an aquifer (major aquifer study), or a specific depth and (or) land use (land-use study) (Lapham and others, 1995). The same wells from each of the selected networks are sampled on a decadal-scale interval (Rosen and Lapham, 2008).

Study methods

12. Why are you sampling only once every decade?

Sampling one time per decade allows the project to collect repeated samples in all of the well networks across the Nation. If we sampled more frequently, we would have to sample fewer wells. Because changes in groundwater quality typically happen relatively slowly, sampling once per decade can be sufficient to capture these slow changes. Other ongoing USGS studies evaluate changes in groundwater quality at time intervals greater than and less than a decadal scale.

13. How do you determine whether or not changes in concentrations with time are statistically significant?

Two statistical tests were used to evaluate changes. The Wilcoxon-Pratt signed-rank test (Pratt, 1959) was used to make comparisons between pairs of sampling events, and the Regional Kendall test was used to make comparisons across three or more sampling events. Significant changes for each constituent are determined by a statistical test called the Wilcoxon-Pratt signed rank test (Pratt, 1959) as described in Lindsey and Rupert (2012) using the R-statistical software. The method calculates changes in concentrations at individual wells and then uses the pattern of those changes to determine whether or not there has been a statistically significant change for a well network as a whole. The Regional Kendall test (Helsel and others, 2006) evaluates trends for individual sites across three or more sampling events, then evaluates whether a consistent trend is evident at the network level. The test evaluates whether the changes are statistically significant, and also determines the slope, or the rate of change in concentration. For both of these statistical tests, a 90-percent confidence level, or a p-value of less than or equal to 0.10, is used to signify a statistically significant change. More on statistics.

14. How do you calculate statistics when reporting levels changed with time?

Because reporting levels vary with time, a maximum common assessment level (CAL) is chosen for each constituent prior to the statistical analysis. Data for the pesticide compounds atrazine, prometon, metolachlor, simazine, dieldrin, and deethylatrazine are prepared using the method described in Toccalino and others (2014). More on data preparation.

Study results

15. Which benchmarks were used to provide context for the results?

Two human-health benchmarks and one other water-quality benchmark were used in this study. The human-health benchmarks were Maximum Contaminant Levels (MCLs) developed by EPA,s Office of Water for compounds that are regulated in drinking water under the Safe Drinking Water Act and nonenforceable Health-Based Screening Levels (HBSLs ) developed by the USGS for unregulated compounds without MCLs. These benchmarks are concentrations below which contaminants are not anticipated to cause adverse human-health effects from a lifetime of exposure. Secondary Maximum Contaminant Levels (SMCLs ) are nonenforceable guidelines regarding cosmetic effects such as tooth or skin discoloration or aesthetic effects such as taste, odor, or color of drinking water.

16. Do the findings indicate if water from the wells is safe to drink?

No. The NWQP did not assess the safety of drinking water. The quality of finished drinking water is regulated by the EPA under the Safe Drinking Water Act. All of the samples included in this study, however, were collected prior to any treatment or blending that potentially could alter contaminant concentrations. As a result, the sampled groundwater represents the quality of the source water and not necessarily the quality of finished water ingested by the people served by these domestic and public wells. In addition, the sampling included some monitoring wells and other types of wells, which are not used as a source of drinking water.

17. What does it mean when large decadal increases in concentrations were identified on these maps with upwardly facing red arrows, and why do the concentration ranges for large and small arrows vary by constituent?

When there was a statistically significant change in the concentration of a constituent with time, the magnitude of the change was classified as being "large" or "small" as compared to a benchmark to provide context for the results. "large" changes in concentrations over time mean that the magnitude of the change was more than 1 percent or 5 percent of the benchmark concentration, depending on the type of constituent, meaning that concentrations in the overall group of wells are approaching a benchmark more quickly than areas having "small" changes. It does not mean that concentrations exceed a benchmark. Each assessment is for a group of 10-30 individual wells in a similar setting. Individual wells within a network may or may not have concentrations that exceed a benchmark. Within each group some concentrations are likely to be decreasing and others increasing, regardless of the direction of change for the overall group. Files that include concentrations for individual wells are available for download.

18. Where can I learn more about contaminants in public-supply wells and domestic wells?

For more information on contaminants in domestic wells see Desimone (2009) and Desimone and others (2015) and for more information on public-supply wells see Toccalino and Hopple (2010) and Eberts and others (2013). For the most recent findings on water quality in public-supply wells in principal aquifers click here.

19. What are the causes of these changes in groundwater quality?

The exact causes of changes in groundwater quality are not evaluated for every constituent and every network. Some nationwide changes, such as the banning of a chemical or introduction of a new chemical, can be documented on a large scale. Changes at the level of an individual network, however, require a focused evaluation of hydrologic conditions, groundwater age, and history of use of the contaminant. Some of these studies are here.

20. What human factors are contributing to changes in groundwater quality?

See more information on factors affecting vulnerability to contamination.

21. What natural features influence water quality?

See more information on factors affecting vulnerability to contamination.

22. Will other constituents be added to this tool?

The 28 constituents displayed were selected as the most important for statistical analysis as described in the 'Criteria for analyzing constituents' table. Constituents that meet those criteria in future sampling events will be added to the map. Because of changes in data collection and analytical methods, some constituents that meet those criteria cannot be analyzed for statistical change across all time periods.

23. The data show results through 2021. Are data still being collected and will those results be displayed?

A third decade of data collection is underway, and these results will be compared to results from the previous two decades. Data collected after 2021 will be analyzed and the results will be added to this map on an annual basis after quality checks on the data.

24. Are some areas likely to see faster changes in groundwater quality than others? Why?

Wells in areas where groundwater arrives at a well shortly after entering the aquifer will likely see concentrations change more rapidly than wells in areas where groundwater travels more slowly. Factors such as well depth and the type of aquifer material control these rates. See more information on factors affecting vulnerability to contamination.

25. What can we do to improve groundwater quality? How long will it take?

The EPA has resources on source water protection. Source water protection strategies that rely on changes in human activities and practices at the land surface to achieve water-quality objectives can take many decades to affect the quality of water in some deeper wells.

For more information on groundwater quality

26. Where is more information available about water-quality testing guidelines for domestic wells?

There are many sources of information about water-quality testing of domestic wells. Many state environmental or public-health agencies provide information and recommendations for homeowners about testing and water-quality of domestic wells. The EPA also provides such information, and provides links to many state agency Web sites on private wells. The U.S. Centers for Disease Control and Prevention provides information on water-quality testing and the health effects of selected contaminants in private wells. The U.S. Department of Agriculture, in cooperation with EPA, provides information and resources for domestic-well owners through its Farm*A*Syst/Home*A*Syst and Cooperative State Research, Education, and Extension (CREES) Program. Local health departments, in many cases, are a source of information about private wells. Recommendations for water-quality testing and other information about domestic wells also are provided by several nongovernmental organizations. Sources of information available on the internet from some of these agencies and organizations are listed below:

U.S. Environmental Protection Agency, Private Drinking Water Wells
https://www.epa.gov/safewater/privatewells/index2.html and
https://www.epa.gov/privatewells.

U.S. Center for Disease Control and Prevention, Division of Parasitic Diseases, Drinking Water, Private Well Resources
https://www.cdc.gov/ncidod/dpd/healthywater/privatewell.htm

U.S. Department of Agriculture and U.S. Environmental Protection Agency, National Farm*A*Syst/Home*A*Syst Program
https://extension.psu.edu/programs/nutrient-management/farm-a-syst

American Ground Water Trust
https://www.agwt.org

Ground Water Protection Council
https://www.gwpc.org

National Ground Water Association
http://www.ngwa.org/Pages/default.aspx and https://www.wellowner.org

National Rural Water Association
https://www.nrwa.org

Water Systems Council
https://www.watersystemscouncil.org

27. Where can I learn more information about other NWQP water-quality assessments?

Access https://water.usgs.gov/nawqa for information about the USGS NWQP and other assessments in the Project about the water resources of the Nation.

Citation and Contacts

28. How should these Web pages be cited?

Lindsey, B.D., Johnson, T.D., Privette, L.M., and Estes, N.J., 2018, Decadal changes in groundwater quality: U.S. Geological Survey Web page, https://nawqatrends.wim.usgs.gov/Decadal/

Lindsey, B.D., May, A.N., Watson, E., and Johnson, T.D., 2023, Data from Decadal Change in Groundwater Quality Web Site, 1988-2022: U.S. Geological Survey, https://doi.org/10.5066/P9YEB7FS

29. Who can I contact for more information on this study of decadal changes in the quality of water?

Bruce Lindsey, USGS Hydrologist
Email: blindsey@usgs.gov
Phone: (717) 730-6964

Search, zoom, and print capabilities

SearchingThe search option allows the user to search for a location such as a city, state, zip code, or general place name.

Navigating the mapThe scroll wheel on the mouse can be used to zoom, and in the upper left area of the tool are "+" and "-" icons to zoom in and out, respectively.

Home LocationThe crosshair button will zoom the map to your location, and the home button will center the US on the map.

Print MapThe print map tab creates a .pdf file of the current map extent, including an explanation.

Explanation

Explanation displayThe Explanation box can be collapsed to display more of the mapped area.

Using the ExplanationThe expanded Explanation box explains the arrows and circles that appear at the network centroid, displaying the statistical results.

The size of the arrow indicates the magnitude of change for networks with statistically significant changes.

Networks without any statistically significant change are displayed as a solid circle.

Networks that have been sampled, but have insufficient data for the selected trend period- constituent combination are not displayed unless the "all networks" box in the map layers menu is selected. When that option is selected, networks with no data or insufficient data for the trend period-constituent combination are shown with a gray circle; this could be because fewer than 10 pairs were available or because a constituent was sampled in one sampling period but not in the other.

Map selection options

Changing the basemap styleThe "Basemaps" sidebar allows a user to show various geographic information or landscape imagery as a background layer beneath the trend results.

Changing or adding map layersThe "Map Layers" sidebar allows a user to toggle laysers that sit on top of the basemap.

Inorganic vs OrganicSelect either "Inorganic" or "Organic" based on the Constituent you would like to view.

Selecting a constituentThe Constituent group selection allows a user to choose a constituent group and then choose a specific constituent from within that group. See About -> Learn More for a list of constituents.

Toggling Decades of the Trend PeriodThe Trend Period selection allows a user to choose a Constituent and then view the decades within each Constituent.

Download the Data

Mapping Criteria and Benchmarks This interactive mapping tool displays decadal changes in concentrations. From among the more than 300 constituents sampled during the three decades, up to 28 constituents were prioritized for analysis for decadal change based on the following criteria:

The constituent exceeded a human health benchmark in at least 1 percent of the wells used as a source of drinking-water public supply wells (Toccalino and Hopple, 2010) or domestic supply wells (Desimone, 2009),
The constituent exceeded a U.S. Environmental Protection Agency (EPA) secondary maximum contaminant level (SMCL) (U.S. Environmental Protection Agency, 2018) or taste threshold in at least 1 percent of the wells used as a source of drinking water,
The constituents were among the five most frequently detected volatile organic compounds (VOCs) in the Nation (Zogorski and others, 2006),
The constituents were among the five most frequently detected pesticides in the Nation (Gilliom and others, 2006), or
Constituents of special or regional interest.

For the two constituents included in the analysis based on regional importance, orthophosphate does not have a health benchmark but is important for ecological health of surface-water systems; 1,2-Dibromo-3-chloropropane (DBCP) was not widely used nationwide, but it is important in regions where it was used in the past. Radon and gross alpha (α) activity met the criteria for analysis, but do not have sufficient data for analysis; thus, they are not included in the mapping tool.

Because of changes in sampling design and laboratory methods, results for pesticides, VOCs, and orthophosphate are not included in comparisons beyond the second decade; however, the results from previous pairs of samples for these compounds are still displayed. Thus comparisons of changes between the first and second decadal sampling events display results for 26 constituents, but, with the addition of radium, comparisons between the second and third decade display results for 15 constituents.

Benchmarks used to prioritize constituents include the following: EPA maximum contaminant levels (MCLs) (U.S. Environmental Protection Agency, 2018), USGS Health-Based Screening Levels (HBSLs) (Norman et al., 2018), and nonregulatory SMCLs (U.S. Environmental Protection Agency, 2018). Of these benchmarks, only MCLs are enforceable (regulatory) drinking-water standards; all but SMCLs are human-health benchmarks. For more information about these benchmarks, click here. For a listing of the 28 constituents selected for analysis and the reason they were selected, click here.

Data Preparation Steps In preparation for statistical analysis, environmental water-quality data for selected sites are retrieved from the USGS National Water Information System database. A common assessment reporting level (CAL) as described by Gilliom and others (2006) is chosen for the data analysis, usually the lowest reporting level that still retains the maximum amount of data for analysis. Analysis is completed for networks with at least 10 pairs of samples. Once a CAL is selected for a given constituent, all nondetections with a reporting level greater than the CAL are deleted from the dataset. All nondetections and reported values less than the CAL are recoded to a unique value selected to specifically represent values below the CAL. The value used for recoding is typically slightly less than the CAL, but its exact value does not affect the statistical analysis, which calculates results using the ranking of values relative to each other rather than using the actual values themselves. The selection of a CAL and the recoding are done to make correct comparisons among nondetections and between nondetections and low-level detections; for example, if a CAL of 0.2 is selected, reported results of < 0.1 and < 0.2 are recoded as 0.19 before statistical analysis so that the statistical program does not interpret these two nondetections as different values and also to distinguish both from a reported value of 0.2. Reported results of 0.17 (a detection) and < 0.2 (a nondetection) are also recoded as 0.19 before statistical analysis because it is not possible to determine if the two values differ. The CAL and the value used for recoding nondetections are reported for each constituent in the "readme" tab of the data files. Data for the pesticide compounds atrazine, prometon, metolachlor, simazine, dieldrin and deethylatrazine (a degradate of atrazine) are prepared using a different method, as described in Toccalino and others (2014). Differences in data preparation for pesticide compounds and degradates include that concentrations were adjusted for recovery and that nondetections were replaced with a single value less than the lowest detection (rather than a value less than the CAL). For methyl tert-butyl ether, the CAL was determined for each pair rather than for the entire data set. The CAL that was selected for use in analyzing changes from decade 1 to 2 was also used to analyze changes across other time periods so that the results are comparable.

Statistical Analysis Two statistical tests were used to evaluate changes. The Wilcoxon-Pratt signed-rank test (Pratt, 1959) was used to make comparisons between pairs of sampling events, and the Regional Kendall test was used to make comparisons across three or more sampling events. The use of the Wilcoxon-Pratt signed-rank test is described in Lindsey and Rupert (2012). The method first calculates changes in concentrations at individual wells and then uses the pattern of those changes to determine whether or not there has been a statistically significant change for a well network as a whole. Because the R-statistical program cannot analyze networks if all the data are tied (no differences in any pair), networks with all ties are assumed to have no significant change. The Regional Kendall test (Helsel and others, 2006) evaluates trends for individual sites across three or more sampling events, then evaluates whether a consistent trend is evident at the network level. The test evaluates whether the changes are statistically significant, and also determines the slope, or the rate of change in concentration. For comparisons across three or more time periods that were evaluated using the Regional Kendall test, the annual change in concentration - which is calculated by the test - is multiplied by 10 to determine the decadal change in concentration. For both of these statistical tests, a 90-percent confidence level, or a p-value of less than or equal to 0.10, is used to signify a statistically significant change.

Results of the statistical analysis for each well network were classified as indicating a statistically significant increase, a statistically significant decrease, or no significant change. Results were further classified as being large or small changes to provide context for the results. A statistically significant change for an individual network is displayed on the mapping tool by an arrow pointing up or down. Each symbol on the map represents a network of multiple wells. To provide context, the median change between the first and second sampling events was calculated for each well network with a statistically significant change, and the median was compared to the benchmark (MCL, SMCL, or HBSL).

The changes in concentration across a decade, or near-decadal period were used to evaluate the magnitude of change. For comparisons that used the Wilcoxon-Pratt signed-rank test, the median of differences in concentrations are calculated for networks with statistically significant changes. For comparisons across three or more time periods that were evaluated using the Regional Kendall test, the annual change in concentration - which is calculated by the test - is multiplied by 10 to determine the decadal change in concentration.

For inorganic constituents, if the median of all differences in concentrations was greater than 5 percent of the benchmark, the change was considered large. If the change was less than or equal to 5 percent of the benchmark, then the change was considered to be small. For organic compounds, if the median of all differences in concentrations was greater than 1 percent of the benchmark, the change was considered large, and if the change was less than or equal to 1percent of the benchmark, then the change was considered to be small. This approach provides a way to distinguish very small but statistically significant changes from changes that are of a larger magnitude.

Organic constituents are treated differently than inorganic constituents because the organic constituents are generally introduced to the environment as a result of human activity, whereas most of the inorganic constituents are found naturally at some level. In some cases, networks had statistically significant changes, but the median change between sampling events was zero. In those cases, the data were analyzed graphically to determine if the change was a decrease or an increase, and the magnitude of change was considered to be small.

Networks that have been sampled, but have insufficient data for the selected trend period- constituent combination are not displayed unless the "all networks" box in the map layers menu is selected. When that option is selected, networks with no data or insufficient data for the trend period-constituent combination are show with a gray circle; this could be because fewer than 10 pairs were available or because a constituent was sampled in one sampling period but not in the other.

References Cited

DeSimone, L.A., 2009, Quality of water from domestic wells in principal aquifers of the United States, 1991-2004: U.S. Geological Survey Scientific Investigations Report 2008-5227, 139 p. Also available at https://pubs.usgs.gov/sir/2008/5227/.]

DeSimone, L.A., McMahon, P.B., and Rosen, M.R., 2014, The quality of our Nation's waters-Water quality in Principal Aquifers of the United States, 1991-2010: U.S. Geological Survey Circular 1360, 151 p., [Also available at https://dx.doi.org/10.3133/cir1360.]

Eberts, S.M., Thomas, M.A., and Jagucki, M.L., 2013, The quality of our Nation's waters-Factors affecting public-supply-well vulnerability to contamination-nderstanding observed water quality and anticipating future water quality: U.S. Geological Survey Circular 1385, 120 p. [Also available online at https://pubs.usgs.gov/circ/1385/.]

Gilliom, R.J., Barbash, J.E., Crawford, C.G., Hamilton, P.A., Martin, J.D., Nakagaki, Naomi, Nowell, L.H., Scott, J.C., Stackelberg, P.E., Thelin, G.P., and Wolock, D.M., 2006, The quality of our Nation's waters-Pesticides in the Nation's streams and ground water, 1992-2001: U.S. Geological Survey Circular 1291, 172 p. [Also available at https://pubs.usgs.gov/circ/2005/1291/

Helsel, D.R., Mueller, D.K., and Slack, J.R., 2006, Computer program for the Kendall family of trend tests: U.S. Geological Survey Scientific Investigations Report 2005-5275, 4 p. [Also available at https://pubs.usgs.gov/sir/2005/5275/pdf/sir2005-5275.pdf

Lapham, W.W., Wilde, F.D., and Koterba, M.T., 1995, Ground-water data-collection protocols and procedures for the National Water-Quality Assessment Program-Selection, installation, and documentation of wells, and collection of related data: U.S. Geological Survey Open-File Report 95-398, 71 p. [Also available at https://pubs.usgs.gov/of/1995/ofr-95-398/.]

Lindsey, B.D., and Rupert, M.G., 2012, Methods for evaluating temporal groundwater quality data and results of decadal-scale changes in chloride, dissolved solids, and nitrate concentrations in groundwater in the United States, 1988-2010: U.S. Geological Survey Scientific Investigations Report 2012-5049, 46 p. [Also available at https://pubs.usgs.gov/sir/2012/5049/.]

Maupin, M.A., Kenny, J.F., Hutson, S.S., Lovelace, J.K., Barber, N.L., and Linsey, K.S., 2014, Estimated use of water in the United States in 2010: U.S. Geological Survey Circular 1405, 56 p. [Also available at https://pubs.usgs.gov/circ/1405/.]

Norman, J.E., Toccalino, P.L., Morman, S.A., 2018, Health-Based Screening Levels for evaluating water-quality data (2d ed.). U.S. Geological Survey web page, accessible at https://water.usgs.gov/water-resources/hbsl/, doi:10.5066/F71C1TWP.

Pratt, J.W., 1959, Remarks on zeros and ties in the Wilcoxon signed rank procedures: American Statistical Association Journal, v. 54, no. 287, p. 655-667. [Also available at https://www.jstor.org/stable/2282543.]

Rosen, M.R., and Lapham, W.W., 2008, Introduction to the U.S. Geological Survey National Water-Quality Assessment (NAWQA) of ground-water quality trends and comparison to other national programs: the Journal of Environmental Quality, v. 37, no. 5, Supplement, p. S-190-S-198. [Also available at https://dx.doi.org/10.2134/jeq2008.0049.]

Toccalino, P.L., and Hopple, J.A., 2010, The quality of our Nation's waters-Quality of water from public supply wells in the United States, 1993-2007-Overview of major findings: U.S. Geological Survey Circular 1346, 58 p. [Also available at https://pubs.usgs.gov/circ/1346/.]

Toccalino, P.L., Gilliom, R.J., Lindsey, B.D., and Rupert, M.G., 2014, Pesticides in groundwater of the United States-Decadal-scale changes, 1993-2011: Groundwater, v. 52, Supplement S1, p. 112-125. [Also available at https://dx.doi.org/10.1111/gwat.12176.]

U.S. Environmental Protection Agency, 2013, Human health benchmarks for pesticides-2013 update: U.S. Environmental Protection Agency, Office of Water, EPA-820-F-13-019, 2 p. [Also available at https://www.epa.gov/sites/production/files/2015-10/documents/hh-benchmarks-factsheet.pdf.]

U.S. Environmental Protection Agency, 2018, 2018 edition of the drinking water standards and health advisories: U.S. Environmental Protection Agency, Office of Water, EPA 822-F-18-001, 20 p. [Also available at https://www.epa.gov/sites/production/files/2018-03/documents/dwtable2018.pdf.]

U.S. Geological Survey, variously dated, National field manual for the collection of water-quality data: U.S. Geological Survey Techniques of Water-Resources Investigations, book 9, chaps. A1-A10, accessed July 31, 2009, at https://water.usgs.gov/owq/FieldManual/.]

Zogorski, J.S., Carter, J.M., Ivahnenko, Tamara, Lapham, W.W., Moran, M.J., Rowe, B.L., Squillace, P.J., and Toccalino, P.L., 2006, The quality of our Nation's waters-Volatile organic compounds in the Nation's ground water and drinking-water supply wells: U.S. Geological Survey Circular 1292, 101 p. [Also available at https://pubs.usgs.gov/circ/circ1292/

Table 1 lists the chemical constituents that met the criteria for a statistical analysis of decadal-scale changes in concentrations in groundwater between Decade 1 (1988-2001) and Decade 2 (2002-2012), and Decade 3 (2013-2021) of the National Water-Quality Assessment (NAWQA) Project, which transitioned into the National Water Quality Program in 2022. Mapped constituents met one of the four following criteria:

Constituents that exceeded a Maximum Contaminant Level or other human-health benchmark in more than 1 percent of public- or domestic-supply wells¹ ² ³, or
Constituents that exceeded a Secondary Maximum Contaminant Level or taste threshold in more than 1 percent of public- or domestic-supply wells¹ ² ³, or
The five most frequently detected pesticides and volatile organic comounds (VOCs) in groundwater⁴ ⁵, or
Constituents of special or regional interest.

Table 1. Constituents meeting analysis criteria, results mapped

[μg/L, micrograms per liter; mg/L, milligrams per liter; SMCL, Secondary Maximum Contaminant Level]

Constituent name	Constituent class	Benchmark	Units	Why study
Arsenic	inorganic	10	μg/L	Exceeded human-health benchmark in more than 1 percent of domestic or public-supply wells
Boron	inorganic	6,000	μg/L	Exceeded human-health benchmark in more than 1 percent of domestic or public-supply wells
Chloride	inorganic	250	mg/L	Exceeded SMCL in more than 1 percent of domestic-supply or public-supply wells
Fluoride	inorganic	4	mg/L	Exceeded human-health benchmark in more than 1 percent of domestic or public-supply wells
Iron	inorganic	300	μg/L	Exceeded SMCL in more than 1 percent of domestic-supply or public-supply wells
Lithium	inorganic	10	μg/L	Exceeded human-health benchmark in more than 1 percent of domestic or public-supply wells
Manganese	inorganic	300	μg/L	Exceeded human-health benchmark in more than 1 percent of domestic or public-supply wells
Molybdenum	inorganic	40	μg/L	Exceeded human-health benchmark in more than 1 percent of domestic or public-supply wells
Nitrate (measured as nitrite plus nitrate)	inorganic	10	mg/L	Exceeded human-health benchmark in more than 1 percent of domestic or public-supply wells
Orthophosphate	inorganic	None	mg/L	Constituent of special or regional interest: Possible source of discharge to surface water bodies
Radium 226 plus Radium 228	inorganic	5	picocuries per liter	Exceeded human-health benchmark in more than 1 percent of domestic or public-supply wells
Sodium	inorganic	30	mg/L	Exceeded taste threshold in more than 1 percent of domestic-supply or public-supply wells
Strontium	inorganic	4000	μg/L	Exceeded human-health benchmark in more than 1 percent of domestic or public-supply wells
Sulfate	inorganic	250	mg/L	Exceeded SMCL in more than 1 percent of domestic-supply or public-supply wells
Total Dissolved Solids	inorganic	500	mg/L	Exceeded SMCL in more than 1 percent of domestic-supply or public-supply wells
Uranium	inorganic	30	μg/L	Exceeded human-health benchmark in more than 1 percent of domestic or public-supply wells
Atrazine	organic	3	μg/L	One of the five most frequently detected pesticide compounds in groundwater in the Nation
Chloroform	organic	80	μg/L	One of the five most frequently detected pesticide compounds in groundwater in the Nation
Deethylatrazine	organic	None	μg/L	One of the five most frequently detected volatile organic compounds in groundwater in the Nation
1,2-Dibromo-3-chloropropane (DBCP)	organic	0.2	μg/L	Constituent of special or regional interest
Dieldrin	organic	0	μg/L	Exceeded human-health benchmark in more than 1 percent of public-supply wells
Methyl tert-butyl ether	organic	20	μg/L	One of the five most frequently detected volatile organic compounds in groundwater in the Nation
Metolachlor	organic	700	μg/L	One of the five most frequently detected volatile organic compounds in groundwater in the Nation
Simazine	organic	4	μg/L	One of the five most frequently detected pesticide compounds in groundwater in the Nation
Prometon	organic	400	μg/L	One of the five most frequently detected pesticide compounds in groundwater in the Nation
Tetrachloroethene	organic	5	μg/L	Exceeded human-health benchmark in more than 1 percent of domestic or public-supply wells, and one of the five most frequently detected volatile organic compounds in groundwater in the Nation
Toluene	organic	1,000	μg/L	One of the five most frequently detected volatile organic compounds in groundwater in the Nation
Trichloroethene	organic	5	μg/L	One of the five most frequently detected volatile organic compounds in groundwater in the Nation

Table 2. Constituents met criteria, not mapped due to insufficient data

Constituent name	Constituent class	Benchmark	Units	Why study
Gross alpha (α)	inorganic	15	picocuries per liter	Exceeded human-health benchmark in more than 1 percent of domestic or public-supply wells
Radon	inorganic	1300 (Alternate 4,000)	picocuries per liter	Exceeded human-health benchmark in more than 1 percent of domestic or public-supply wells

References Cited:

DeSimone, L.A., Hamilton, P.A., and Gilliom, R.J., 2009, Quality of water from domestic wells in principal aquifers of the United States, 1991-2004-Overview of major findings: U.S. Geological Survey Circular 1332, 48 p. [Also available at http://pubs.usgs.gov/circ/circ1332/.]
Toccalino, P.L., and Hopple, J.A., 2010, The quality of our Nation′s waters-Quality of water from public supply wells in the United States, 1993-2007-Overview of major findings: U.S. Geological Survey Circular 1346, 58 p. [Also available at http://pubs.usgs.gov/circ/1346/.]
Ayotte, J.D., Gronberg, J.M., and Apodaca, L.E., 2011, Trace elements and radon in groundwater across the United States: U.S. Geological Survey Scientific Investigations Report 2011-5059, 115 p. [Also available at http://water.usgs.gov/nawqa/trace/pubs/sir2011-5059/.]
Zogorski, J.S., Carter, J.M., Ivahnenko, Tamara, Lapham, W.W., Moran, M.J., Rowe, B.L., Squillace, P.J., and Toccalino, P.L., 2006, The quality of our Nation′s waters-Volatile organic compounds in the Nation′s ground water and drinking-water supply wells: U.S. Geological Survey Circular 1292, 101 p. [Also available at http://pubs.usgs.gov/circ/circ1292/.]
Gilliom, R.J., Barbash, J.E., Crawford, C.G., Hamilton, P.A., Martin, J.D., Nakagaki, Naomi, Nowell, L.H., Scott, J.C., Stackelberg, P.E., Thelin, G.P., and Wolock, D.M., 2006, The quality of our Nation′s waters-Pesticides in the Nation′s streams and ground water, 1992-2001: U.S. Geological Survey Circular 1291, 172 p. [Also available at http://pubs.usgs.gov/circ/2005/1291/.]

Constituent Select

Trend Period

Basemaps

Map Layers

Site Information

Network Information

Aquifer Information

Explanation

Web browser warning

Constituent Select

Trend Period

Basemaps

Map Layers

Site Information

Network Information

Aquifer Information

Explanation