National Library of Medicine TOXMAP Skip Navigation

How accurate are TRI locations in TOXMAP?

Overview

TOXMAP plots the location of all TRI facilities and Superfund sites using addresses and coordinates from the EPA Facility Registry System (FRS) (http://www.epa.gov/frs). The location of some of these may appear to be slightly inaccurate, because they are calculated by a geocoding estimation process. General information about the accuracy of locations in TOXMAP can be found in the TOXMAP FAQ "Why are the locations of some TOXMAP facilities slightly inaccurate?"

The following provides more detailed information about the problems with TRI location data, EPA's ongoing efforts to improve it, the limitations of FRS, and location accuracy in TOXMAP today.

Problems with TRI, Limitations of FRS

The EPA recommends use of addresses and coordinates from their Facility Registry System (FRS) (http://www.epa.gov/frs). TOXMAP only plots TRI facilities with defined FRS locations. Corrections to FRS are made year-round by EPA and are incorporated into TOXMAP annually with each year's publicly available data.

This approach avoids many errors in the TRI location data. However, there are also limitations to FRS data, such as in the case of a large facility (tens or hundreds of square miles in area) that reports to multiple EPA programs. Since FRS only uses the "best" of the reported coordinates, a more accurate location for a TRI release could be overlooked (e.g. the street address coordinate is used because it is easily identified as "best", but the TRI release actually takes place miles away).

Coordinate Accuracy Assessment

To get a better understanding of coordinate accuracy in EPA FRS data, TOXMAP annually conducts an assessment in which each facility's reported ZIP code is compared with the 5-digit ZIP code the coordinate lies within (plus a 5-mile buffer).

To translate coordinates into ZIP codes, TOXMAP uses a ZIP code file created by ESRI in 2006 (based on ZIP code boundaries from an earlier date).

The results place facilities into five categories based on the number of matching digits. Because ZIP codes follow a pattern across the US, these categories helped, to some degree, indicate accuracy. The latest results are as follows:
  • Number of facilities evaluated: 54032
  • TRI data for years: 1988-2011
  • FRS data as of: November, 2012
  • First 3 or More Digits Match: ~99.06%
  • First 1 or 2 Digits Match: ~0.25%
  • None of The Digits Match: ~0.004%
  • Not Evaluated: ~0.68%
Look for the "Coordinate accuracy suspect" and "Coordinate not evaluated" notes when viewing TRI details. These notes indicate that the facility falls in one of the last four categories above.

Some facilities have correct coordinates but are categorized with partial ZIP code matches simply because their reported ZIP code was incorrect.

Facilities categorized with "No ZIP Code Was Found" have coordinates that fall outside of five-mile buffered ZIP code boundaries. Facilities in Puerto Rico and other US territories were not evaluated.

To Report a data error to EPA, please use the "Report an Error" (http://www.epa.gov/epahome/comments3.htm) web service and click the link on that page to report a data error.