2017 AirToxScreen: Assessment Results

About the 2017 Assessment

This page provides the emissions data, ambient concentrations and health-effect results for the 2017 Air Toxics Screening Assessment, or AirToxScreen. You can also find links to supporting files and web pages with more information about these results.

EPA developed AirToxScreen as a screening tool for state, local and tribal air agencies. AirToxScreen’s results help these agencies identify which pollutants, emission sources and places they may wish to study further to better understand any possible risks to public health from air toxics. Often, more localized studies are needed to better characterize local-level risk. These studies often include air monitoring and more detailed modeling.

The following files will help you navigate the AirToxScreen results:

2017 AirToxScreen: Technical Support Document provides technical information on the assessment, including and emissions, modeling, and risk calculations. We've also included a folder containing supplemental data files referred to in the TSD and a document describing how we calculated ethylene oxide emissions from commercial sterilizers for this assessment.
2017 AirToxScreen: Emissions Update Document lists some of the known issues and updates that were not addressed during the review of the 2017 assessment, including information on some facilities about which EPA is aware that have reduced emissions of certain air toxics since 2017.
2017 AirToxScreen: Explanation of Data Elements provides descriptions of the many fields in the results files of AirToxScreen.

2017 AirToxScreen Emissions

The 2017 AirToxScreen used the 2017 National Emissions Inventory (NEI) as a starting point. The NEI is EPA’s comprehensive estimate of air pollution emissions from sources across the country. NEI supports air quality modeling and other activities within EPA’s Air Toxics Program. We updated these data from comments provided by state, local and tribal agencies during the AirToxScreen review. The resulting emission files, accessed below, reflect the emissions used in AirToxScreen.

2017 NEI Technical Support Document provides more details on the 2017 NEI.

NOTE: Emission inventories submitted by state, local and tribal agencies vary in the level of detail and completeness. For this reason, you should not compare AirToxScreen risk estimates between states or regions without considering these inventory differences.

2017 Modeled Ambient Concentrations, Exposures and Risks

The following tables present the EPA's 2017 AirToxScreen ambient concentration, exposure concentration and risk estimates across all 50 U.S. states, Puerto Rico, the Virgin Islands, and the District of Columbia.

You can view these data in several ways. Microsoft Excel- and Access-format files below contain data at national, state, county and census tract levels for each or endpoint. We show total risks and hazards, then break out the contribution for each NATA source type (for example, stationary point sources, on-road gasoline vehicles, etc.).

The 2017 AirToxScreen National Respiratory Risk by Tract Source spreadsheet presents noncancer for 14 target organ systems, or “endpoints.” Past assessment results suggest that the respiratory endpoint (the effect of air toxics on the lungs and the rest of the respiratory system) usually drives noncancer health effects. But we provide results for all 14 endpoints in the results spreadsheet. Information on these endpoints, such as the critical concentration used for each, can be found in the Supplemental Data files for AirToxScreen.

In the 2017 AirToxScreen results, results for stationary sources are broken into two groups: “point” and “nonpoint” sources. This reflects the way we modeled each source.

Each point source’s exact latitude and longitude coordinate is in the AirToxScreen source inventory. We used these locations in modeling. Large industrial complexes often have many individual point sources. We also modeled some smaller sources, such as dry cleaners, as point sources.

For other smaller sources, we may not have an exact location in the AirToxScreen inventory. We modeled these as nonpoint sources. Emissions from homes, such as wood-burning stoves and fireplaces or solvent emissions, are examples of nonpoint sources.

We usually inventory nonpoint sources by county. But a more precise location gives better modeling results. So we divided each county into smaller, square “grid cells,” then assign nonpoint emissions to each cell using population or another method that realistically distributes the emissions across the county (these are called “surrogates;” see the AirToxScreen Technical Support Document for more details).

We used a similar approach to assign most mobile source emissions. An exception is for airports, included as a nonroad source type under mobile sources. We modeled airports using their actual locations.

Nationwide Results

National cancer risk summaries:

National noncancer hazard index summaries:

Pollutant Specific Results

After you select a pollutant, press download to open or save the associated MS Access ZIP file. Please note that file sizes range from 5MB to 79MB.

State Summary Files

After you select a state, press download to open or save the associated MS Access ZIP file. Please note that file sizes range from 1MB to 115MB.