nei2014

Retrieving the 2014 NEI data

Data for the air quality modeling version of the U.S. EPA's 2014 National emissions inventory is available for download from an EPA FTP server. A description of the included data is available here.

This repository includes a script—download.go—that downloads the data and preparing it for use. After installing the Go language compiler, the script can be run with the command go run download.go -dir="/path/to/download" where /path/to/download is the location of the directory where the data should be downloaded to. Downloading the data may take a while.

This repository also includes the additional file surrogate_specification_2014.csv. This file is combined and edited version of surrogate specification files that can be downloaded from the FTP site which has been edited to replace missing shapefiles with existing replacements and combine US, Canada, and Mexico surrogates in one place. Improvements to this file or advice regarding the locations of the missing files are welcome.

Finally, this directory in includes a configuration file—cstref_2014.toml—that specifies how the files can be used to processed the 2014 NEI. The configuration file assumes that a $nei2014Dir environment variable has been set to the directory where the data files were downloaded to (/path/to/download in the example above).

Required manual changes

After running the download.go script, some additional changes need to be made manually:

• The following line should be added to ge_dat/gridding/mgref_onroad_us_2014platform_03oct2016_nf_v2.txt:

000000;2201610080;222
000000;2202420080;222
000000;2202210080;239
000000;2202310080;239
000000;2201320080;239
000000;2205000062;239
000000;2202520080;222
000000;2205320080;241
000000;2205210080;239
000000;2201420080;222
000000;2201510080;242
000000;2202430080;202
000000;2202510080;201
000000;2201520080;222
000000;2201540080;239
000000;2202620080;244
000000;2201430080;201
000000;2202610080;222
000000;2201530080;244
000000;2202320080;244
000000;2201000062;239
000000;2201210080;239
000000;2202410080;244
000000;2205310080;239
000000;2203420080;222
000000;2202530080;244
000000;2202540080;239
000000;2201110080;239
000000;2201310080;239
000000;2202000062;239

• Delete the line starting with COUNTRY_CD from SmokeFlatFile_ONROAD_20160910.csv.

• Delete the leading "1" from each record in the PRUID attribute column of the shapefile /home/chris/data/2014_nei_data/Canada_2010_surrogate_v1/Non_NAESI/SHAPEFILE/pr2001ca_regions_simplify.shp. So 159000 should become 59000.

• The Canadian census division file that comes with the data (Canada_2010_surrogate_v1/NAESI/SHAPEFILE/gisnodat.shp) is unnecessarily large (making it unnecessarily difficult to create surrogates) and the ID codes have the same problem as listed above. To fix this, the download script will download an alternative shapefile: lcd_000b16a_e.shp. Before this shapefile can be used however, you need to add an additional attribute column called FIPS that consists of the first two characters of the attribute CDUID, then a zero, then the final two characters of CDUID. In QGIS, this can be done in the "field calculator" with the expression: concat(substr(CDUID,1,2),'0',substr(CDUID,3,2)).