US Emission Facilities Land Cover Area Derived At Parcel Scale

This dataset includes US industrial facilities from the Environmental Protection Agency's (EPA) 2017 National Emissions Inventory (NEI), combined with location data from the EPA's Facility Registry Service and land cover classes from the United States Geological Survey's (USGS) National Land Cover Data (NLCD). These land cover classes are measured in square meters at the parcel scale. The matching of facility parcels was done using a tiered approach to enhance spatial accuracy. The parcel data was provided by Homeland Infrastructure Foundation-Level Data (HIFLD) US Parcel Data. Because this parcel data is proprietary, the parcel geometries and fields were removed from the final dataset. However, centroid latitude and longitude coordinates were derived to allow spatial joins with publicly available parcel data.

This dataset is organized by unique EPA NEI facilities data fields. Unique facility observations are identified by the field cleaned_name which represents the concatenated address and/or place name for the facility (dependent on data availability). Each row represents a facility matched to parcel scale land cover information and are associated with the unique identifier MatchID. NLCD land cover fields are described as the total area in meters squared of each facility parcel. Please see data README file for more information on individual data fields. 

Known Limitations

Parcels are matched to the facility and in some cases multiple facilities are matched to the same parcel. Data users may want to omit these multiple match parcels and there is a data flag called multi_match that enables this.

Approximately 15% of the dataset includes facilities where the latitude/longitude coordinates are over 200 meters from the matched parcel. Spot checking these instances revealed that, in many cases, the facility latitude and longitude locations did not accurately match the street address, city, or postal zip code associated with the facility. These instances are flagged as a potential source of inaccuracy and can be removed at the users discretion.