An official website of the United States government.

We've made some changes to If the information you are looking for is not here, you may be able to find it on the EPA Web Archive or the January 19, 2017 Web Snapshot.

RadNet Search User Guide

RadNet BannerRadNet iconThis User Guide is designed to help users effectively navigate and search the RadNet database. It describes the database input fields and output fields, as well as the data normalization quality control that was conducted. Some technical knowledge on radiation, radiochemistry, health physics, or related topics is useful for understanding and using the RadNet data. See the National Air and Radiation Environmental Laboratory website for additional information on RadNet, and the EPA Radiation website for general radiation information. Clicking on the field names below will access the data model description (or metadata) for that field.

Update on Federal Radiation Milk Sampling

November 3, 2014: The U.S. Environmental Protection Agency is no longer conducting RadNet milk sampling. The U.S. Food and Drug Administration has the authority for food safety, including monitoring radiation in milk. Learn more about FDA's Total Diet Study Exit and Toxic Elements in Food and Foodware, and Radionuclides in Food Exitprograms. 
EPA completed its final quarterly milk sample collection in April 2014 and the results from analyzing those samples, along with historical milk sampling results, will continue to be available on Envirofacts.

RadNet Search Input Fields:

Location refers to the geographic point at which a given sample was taken. State, City/State, EPA region, or Monitoring Station can be selected in the simple search. Station locations are dispersed across the U.S. and its territories, providing broad geographic and population coverage. Over the decades of operation, however, some changes have occurred regarding the RadNet stations.

Caution! When selecting state or region, many thousands of records can result, depending on the medium, nuclide/radiation, and date range selected. This is especially true for a search combination that includes a large state or region (many station locations), air sampling (the most frequently sampled medium), gross beta (the most frequently analyzed radiation type), and the full date range.

Several general points should be noted about these stations:

  • RadNet sampling location may have changed over the years. EPA depends on volunteer operators (generally State employees) to collect the ambient radiation samples and send them to NAREL for analysis. Station operators may change due to modifications in state and local organizational structure and personnel. When the responsibility for performing these duties is transferred to another volunteer, location of the station may change within the city or town. In this case, the results continue to be reported in the database from the same location. If the station is relocated to a different city or town, the data is entered into the database as a new location.
  • States may request that RadNet stations be established, relocated, or terminated, resulting in changes in location or total numbers of monitoring stations.
  • Exact sampling location within a city is historically unknown. In the past, the latitude and longitude of the sampling stations has not been tracked, due partly to the voluntary nature of RadNet. This historical database therefore reflects a latitude and longitude at city centroid for each monitoring station. Today exact station location of each station is being tracked. This information will be reflected in future updates.
  • In some cities there is more than one RadNet monitoring station for the same medium. In such cases, the locations will have the same city centroid locational information (latitude and longitude; only important for the customized search); however, they are differentiated using a station-specific numbering system.

The table below, Changes in the Number of Sampling Stations at Selected Time Intervals, provides an overview of the changes over time in the number of sampling stations for each medium.

Changes in the Number of Sampling Stations at Selected Time Intervals

aSome data collected prior to 1973 also are included in this database.

bThe isotope 85Kr was rarely detected and thus is not included in this version of the database.

RadNet media are either air particulates, precipitation, drinking water, surface water, or pasteurized milk. The sections below provide discussion of the media sampled in RadNet and changes that have taken place over time.

Caution! When selecting a frequently sampled medium (e.g., air) in combination with a large geographical area (region), frequently analyzed nuclide/radiation type (gross beta), and the full date range, many thousands of records could result. The table below, Changes in the Radiochemistry Analysis and Analytical Frequency at Selected Time Intervals, provides the regularly defined schedule for monitoring specific radionuclides in each media. This schedule changes during alert mode.

Air Particulates. Air particulates are collected continuously at field stations using a high-volume air sampler. Air filters are replaced twice per week, either on Monday and Thursday or Tuesday and Friday. Gross radiation activity in particulate matter contained in the filters is initially measured in the field following a 5-hour period to permit decay of radon daughters that may be attached to collected particles. The filters are then sent for laboratory analyses. Initially, a charcoal-based filter was used in the air sampler; however, this filter was discontinued by the manufacturer in the early 1980s. Consequently, the filter was replaced with a new type of fabric filter. Based on a NAREL study of the new filter, the change from the charcoal based filter to the fabric filter is not expected to have any significant effect on the consistency of the data collected before and after the filter changed.

Precipitation. Precipitation samples are collected as rain, etc. events occur. The precipitation collection unit is a 0.5 square meter area fiberglass collector draining into a five gallon plastic bucket that serves as a reservoir for the precipitation. A monthly composite sample is prepared in the laboratory from the samples received from the field stations. Also, a composite of the March, April and May monthly composites is analyzed.

Drinking Water. Drinking water samples are collected at the tap from water supplies serving major population centers. The tap selected should be one that is frequently used and provides water that has had no on-site treatment such as water softeners or filters. Drinking water has been sampled at a quarterly frequency since inception of the RadNet program in 1973.

Surface Water. Surface water was collected as a grab sample from a river, lake, or other body of water, generally downstream from a nuclear facility. The sampling site should be easily accessible during all seasons of the year. The physical nature of the site should be such that water can be collected from the main body of water (not stagnant areas) in a depth sufficient to prevent disturbing bottom sediment and incorporating it into the water sample. The surface water sampling program was terminated in March 1999, since it was redundant with state sampling programs around nuclear facilities.

Pasteurized Milk. Pasteurized milk samples are collected from dairy plants, which account for 80 percent of the milk consumption for a major population center. A gamma scan is performed monthly on individual samples, and analyses of some nuclides is performed annually on composite samples. The milk samples were taken monthly from 1973 to 1999. In March 1999, the sampling frequency was changed to quarterly.

Nuclides/Radiation (Analytes)
The analytes are either a non-specific type of radiation (e.g., gross alpha) or a specific radionuclide (e.g., uranium-238). (For more information on radionuclides and radiation type, see the EPA radiation basics web pages.) Over the past three decades, EPA has continued to upgrade its analytical radiochemistry and sampling techniques to ensure the RadNet data base is an increasingly rich source of information.

Caution! When selecting a frequently analyzed nuclide/radiation type (e.g., gross beta) in combination with a large geographical area, frequently sampled medium, and the full date range, many thousands of records could result. The table below, Changes in the Radiochemistry Analysis and Analytical Frequency at Selected Time Intervals, summarizes the nuclides/radiation analyzed and the key changes that have taken place.

Date Range
The search form automatically uses the full RadNet sampling period (e.g., 1978 to the present) unless the user specifies otherwise. Because of changes in station location over time and the sporadic nature of some of the results (e.g., for nuclides that are only analyzed if gross beta exceeds a threshold), then most first time queries for a given location, medium, and nuclide/radiation combination should use this default range in order to capture the available data.

Caution! When selecting the full date range in combination with a large geographical area, a frequently sampled medium, and frequently analyzed nuclide/radiation type, many thousands of records could result. The table below, Changes in the Radiochemistry Analysis and Analytical Frequency at Selected Time Intervals, summarizes the key changes that have taken place over time.

Units of Measure
Radiation can be expressed in several different types of units, depending on the sample, the analyte, and other factors. The units used in this version of the RadNet data are picoCuries per liter (pCi/L), attocuries per cubic meter (aCi/m3), pCi per cubic meter (pCi/m3), grams per liter (g/L, which is used only for total potassium (K) and calcium (Ca)), and disintegrations per minute per gram of carbon (DPM/gC, which is only used to report the ratio of C-14/total C). The traditional U.S. unit of measure for radiation and the default output option for the Simple Query is pCi (either per L or m3 in this case). A pCi is equivalent to 3.7E-2 disintegrations per second (DPS). The International System (SI) unit of measure and the output option for the Simple Query is the becquerel (Bq; also either per L or m3). A Bq is equivalent to 1 DPS. Thus, 1 pCi equals 0.037 Bq. Results for non-radioactive elements (e.g., for potassium-40 and calcium in g/L) also are available.

RadNet Simple Query Report Output Fields:

Location is the city or cities in or near where the samples were taken. For surface water samples, the name of the surface water body is also provided. Monitoring stations for a given medium occasionally are moved within a city, and in the past these movements were not tracked. If a monitoring stations was moved to a new city in a state, a new location is provided. More information

Media samples are air, precipitation, drinking water, surface water, or pasteurized milk. More information

Sampling Date
The date the sample was collected. For air, the sample is taken over several days, and end of collection date is used to represent sampling date. For Strontium-90 in milk during the period 1960 through July 1978, sample specific collection dates were not published for routine samples. For data presentation purposes of this data, the individual or the composite sample for that month is assumed to be "collected" on the 15th of the month. For non-routine samples collected during 1976 and 1977 (i.e., increased sampling due to Chinese atmospheric nuclear weapons tests) specific collection dates are available and are used.

This field refers to the analytical procedure that was used on the sample. The specific procedure depends on several factors, including the medium and the radionuclide or radiation type of interest.

Nuclides/Radiation (analytes)
The full, descriptive name of an analyte, either a non-specific type of radiation (e.g., gross alpha), or a specific radionuclide (e.g., uranium-238). More Information

The amount, activity, or concentration measured. Amounts can be zero, non-detect (ND), positive or negative. A negative result occurs when random effects and limitations in the measurement process cause the measured value for the sample to be less than that of the laboratory blank or background, which is subtracted from the sample measurement. Although negative radioactivity is physically impossible, the inclusion of negative results allows better statistical analysis of the data.

Combined Standard Uncertainty
The Combined Standard Uncertainty (CSU) is the estimated standard deviation of the result. The CSU can be used to estimate the "confidence interval" around the amount. Thus, if y is the "best estimate" of measured radioactivity, then the 95% confidence interval around y ranges from about y-(2xCSU) to y+(2xCSU). For additional guidance on the derivation and use of CSU, see the ERAMS Quality Assurance Manual and Guidelines for Evaluating and Expressing the Uncertainty of NIST Measurement Results (25 pp, 176 K, About PDFExit. For Strontium-90 in milk during the period of 1960 through July 1978, the uncertainties were reported in Radiological Health Data and Reports to be +/- 2 pCi/L (two sigma) for Sr-90 concentrations less than 20 pCi/L, and 10% of the concentration for Sr-90 concentrations greater than 20 pCi/L. Sample specific uncertainties are available for data beginning in October 1975 (two sigma).

Minimum Detectable Concentration (MDC)
The Minimum Detectable Concentration (MDC) reflects the ability of the analytical process to detect the analyte for a given sample. The MDC is the activity concentration for which the analytical process detects the radioactive material in a give sample that provides a 95% chance that the radioactive material will be detected.

See the MDC FAQ of the Multi-Agency Radiation Survey and Site Investigation Manual (MARSSIM) for additional detail.

Units of Measure
The default units used in the Simple Query output are picoCuries per liter (pCi/L), attocuries per cubic meter (aCi/m3), pCi per cubic meter (pCi/m3), and disintegrations per minute per gram of carbon (DPM/gC), which is only used to report the ratio of C-14/total C. If the output option of International System of units (Si) was selected, then becquerels (Bq) rather than pCi are displayed. More Information

Changes in Radiochemistry Analysis and Analytical Frequency at Selected Time Intervals

aRefer to the actual data to obtain exact termination dates.

bRefer to the actual data to determine specific changes in frequency.

cEach sample is a monthly composite of all samples collected within the EPA region for January, April, July, and October.

dEach sample is a monthly composite of all samples collected within the EPA region for January, April, and October.

Data Normalization, Quality Control, and Upload to Envirofacts:

Some of the data available on this site was collected as far back as 1960, although much of the data has been collected since 1978. Prior to including the data on this site, EPA conducted various steps involved with normalizing the data, conducting quality control (QC), and uploading the data to this website. These steps are described below.

Pre-1978 ERAMS and Pre-ERAMS Data

Some of the data available on this website were collected under a variety of programs that preceded ERAMS (which was established in 1973 and was renamed RadNet in 2005). These earlier programs included the Radiation Alert Network, the Tritium Surveillance System, the Interstate Carrier Drinking Water Network, and the Pasteurized Milk Network. See Legal Basis and RadNet Timeline for the history of the pre-ERAMS data.

The data reported between 1974-1978 and pre-1974 available on this website is in summary (i.e., compiled or averaged) form. Individual sample measurements for this data no longer exist. The post 1978 data is reported as individual sample measurements. Earlier data underwent QC procedures at the time it was summarized prior to 1973 and is provided on this website "as is." Currently, only a portion of the pre-1974 data has been converted to electronic format and is made available on this website. As additional pre-1974 data is converted from hard copies, it will be made available on this website.

RadNet Data

The RadNet data has undergone rigorous QC, beginning prior to sample collection (e.g., during the preparation of sample collection materials), through the field and laboratory analysis phase, to final data quality checks prior to uploading the data to this website. Sample collection and analysis QC is described in the ERAMS Quality Assurance Manual (June 2001). Several data normalization steps—for example, mathematical conversions to make the units consistent within the same medium/radionuclide combination (e.g., pCi/L) have been conducted on the data. Given the age of much of the data, however, and the different QC and normalization processes that have been applied over time as techniques and knowledge improved, several additional data quality checks and filters have been applied to the data prior to the upload to this website. The data omitted by these steps represents less than one percent of the total data.

The QC and data normalization steps for the RadNet data are described in detail below and are organized as follows: Overall; Date and Time; Decay Correction; Sample Location; Sample Size; Units; Media; and Analytes.


1. Results of analyses that have not been approved by two reviewers (with the exception of field estimates of air samples) are omitted.

2. All results that are stored in the data base with a qualifier are omitted. The qualifier is either J or R. The J qualifier indicates the result has potentially large and unquantified uncertainty or bias. The R qualifier means the result is "rejected" and considered invalid.

3. If a result record has no result and no MDC, it is omitted.

4. If a gamma result record shows zero activity and zero uncertainty, it is omitted.

5. If a field estimate result record shows nonzero uncertainty, it is omitted.

Date and Time

6. All times of day have been omitted, i.e., date and time fields have become date-only.

7. If no collection date is recorded, the sample is omitted.

8. If only one collection date is recorded (i.e., either the collection start date or the collection end date), that date is used for both the beginning and end of the collection period.

9. If the collection dates in the data base are known to be incorrect (but the correct dates have not been determined), the sample is omitted.

10. If the data base indicates that a sample was analyzed before it was collected, the results of the analysis are omitted.

11. Any result for one of the long-lived radionuclides 40K, 234U, 235U, 238U, 232Th, 230 Th, 239Pu, 2262Ra, 14C, or total potassium, K, with a result date that is not within the collection period is assigned a result date equal to the collection end date.

Decay Correction

12. For a particular set of radionuclides (7Be, 3H, 134>Cs, 137Cs, 60Co, 131I, 140Ba, 140La, 89Sr, 90Sr), all results that were not originally decay-corrected back to the sample collection date are decay corrected. If their holding (decay) time is more than three effective half-lives of the radionuclide and no activity was detected, then the result is omitted. For the same set of radionuclides, any result with a decay time longer than 10 effective half-lives is omitted whether it was "detected" or not. (This rule may be modified in the future if it is found to cause the omission of valid data.) The "effective" half-life is the true half-life for all the radionuclides except 140La, which has an effective half-life that is the half-life of its parent, 140Ba.

13. For all radionuclides, if the result date is earlier than the collection date (because of an erroneous decay correction), the result is omitted unless the radionuclide is long-lived or is 3H, 137Cs, 60Co, or 90Sr and the date discrepancy is small such that the difference in decay factors is less than 0.l%. In these latter cases the result date is set equal to the collection date.

14. 89Sr results for composites of samples collected at different times are omitted, since it is not possible to decay-correct these results properly.

15. If the result date is null and the radionuclide is not long-lived, the result is omitted. If the result date is null and the radionuclide is long-lived, the result date is set equal to the collection end date.

Sample Location

16. If no sampling location is recorded, the sample is omitted.

Sample Size

17. If the amount of sample analyzed is substantially larger than the recorded total size of the sample, the results of the analysis are omitted.

18. For air samples, if the air volume analyzed is substantially less than the total air volume sampled, the results of the analysis are omitted.


19. Measurement units for results are converted to pCi/L, pCi/m3,or g/L as appropriate. If the unit is null or is such that the conversion is impossible, the result is omitted. For individual air samples, no conversion is performed. Instead, if the unit is not already recorded in pCi/m3, the result is assumed to be erroneous and is omitted. For air filter composites, if the unit is not recorded in aCi/m3 or pCi/m3, the result is omitted.

20. Results for 14C are provided in "DPM/GC", which means "disintegrations per minute per gram of total carbon."

21. Only air samples (individual or composite) may be reported per m3. Results in activity per m3 for any other samples are omitted.

22. For (individual) air samples, if the sample size is recorded in units other than m3, the sample is omitted.

23. For air filter composites, if the amount analyzed is recorded in units other than m3 or g, the results are omitted.


24. Air samples taken from the same air sampler but for which the recorded collection intervals overlap more than a week (thus indicating a data transcription or entry error since there should be no overlap) are omitted.

25. For air samples, if the air flow rate is less than 5 m3/h or more than 90 m3/h, the sample is omitted. (The nominal air flow rate is about 40 m3/h.)

26. If the total air volume is less than 10 m3, the sample is omitted.

27. For air samples, if the collection interval is longer than 28 days, the sample and its results are omitted.

28. Field radiation screening results (e.g., FIELDEST) are provided only via the customized search, not the general search.

29. Any field radiation screening result (e.g., FIELDEST) that is not labeled as an air sample is omitted.


30. 214Bi and 214Pb results (by gamma) for all samples except individual air samples are omitted because they are radon decay products that fluctuate widely in the laboratory background.

31. If the analyte is not a valid analyte for the analysis type, the result is omitted.