3

I'm working with the shapefiles for the US census 1999 msa files which can be downloaded here:

https://www2.census.gov/geo/tiger/PREVGENZ/ma/ma99/cm_sa_99_shp.zip

Unfortunately, there is no CRS attached with these shapefiles. After reading other questions, it seems that guess and check for the CRS is the best strategy. EPSG 4326 seems to work well, but after loading the shapefile with the R sf package, there seems to be some slight distortion in the shapefile after using sf::st_set_crs with 4326. My question is should I worry about this distortion and is there a better CRS (or method to find the CRS)?

Improve this question

1 Answer 1

Reset to default
5

US census data is almost always in NAD 1983 (EPSG:4269). You'll see the same distortion with 4269. Both 4269 and 4326 are geographic coordinate reference systems. They are probably displayed using Plate Carrée or a variant of Plate Carrée. That projection treats the angular units as if they are linear units so often looks stretched east-west.

Improve this answer
1
  • Thanks for this. I emailed the census bureau and they said, "All of our geographic data files are unprojected data. We do not use a specific coordinate reference system, but simply provide the coordinates which you can project in any format you choose." I also noticed that in later years they use 4269.
    – chandler
    Commented Jul 20, 2017 at 19:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.