Need data about population density by zip code and/or Long/Lat
May 2, 2014 6:28 AM   Subscribe

I'm working on a project that requires very roughly guesstimating how many people live in a 1000' diameter circle, placed in arbitrary locations in the US. My best methodology for doing so is to take the lat/long, approximate the nearest zip code, and then use Census data to determine what the likely population density is. Any ideas?

I've spent 2 days searching for this information (for free) and the US Census (both decennial and annual) doesn't appear to have this data readily available. While I'm at it, my lat/long -> zip/zcta conversion is rough-bordering-on-nonexistent, if there are any suggestions regarding that.
posted by duende to Science & Nature (7 answers total) 4 users marked this as a favorite
You can find the area of each zip code (and the lat/lon info) here. You can find populations here. That should give you what you want to know.

I admit I am accepting the claims of these data sources at face value. I have not actually tried to work with them.
posted by ubiquity at 6:45 AM on May 2, 2014

Are you not trying to do this with a GIS application? I'd think this should be fairly easy to do in GIS using US Census shapefiles and a selection tool.
posted by ursus_comiter at 6:48 AM on May 2, 2014 [2 favorites]

You would need to use census data at the block level, but that may not be granular enough for the 1000-foot level. (Zip codes won't do it, for sure.) In rural areas, many blocks are much larger than 1000 feet. Population data is not available on a smaller scale, such as the coordinates of every residence, which is what you'd need to really do this right. If you click on Census Tracts here, you can download the full tracts description file including latitude and longitude for each tract. I believe this is going to give you the coordinates of the center of the tracts. You can get into more detailed shape files here. You can match this up with the population by census tract, which I believe you can pull via DataFerrett.
posted by beagle at 6:52 AM on May 2, 2014 [1 favorite]

I'm fairly certain ZIP code data will fail you in achieving your aim as they're not always homogeneous in population density throughout the ZIP code.

Let's say you arbitrarily selected my area. It's fairly rural. Now, my ZIP code also includes some small city areas for which the population density is much higher, and those areas are certainly more than 1000' from my house -- they're more like 5 or 6 miles away. Heck, for some houses near me, a 1000' diameter circle centered on the house wouldn't even take you off of the same person's farm, giving you a population density of something like 2-5 people, 50 head of cattle, 3 goats, and a horse. Move to urban areas of the same ZIP code and you'd get a population density of perhaps 150 people in a 1000' diameter circle.
posted by tckma at 7:05 AM on May 2, 2014

This is the type of thing for which GIS is perfect. I'm a GIS specialist, I can calculate this for you if you'd like - just memail.
posted by troika at 7:28 AM on May 2, 2014 [1 favorite]

Use the ACS data, and the data ferrett application to pull the data you need ( to avoid downloading 500 gig files).

Memail if you want more advice, and I'll message you when I get home. At the moment this stuff is my job (Economic labor research).
posted by jjmoney at 9:43 AM on May 2, 2014

This previous question might help. The Missouri Census Data Center has some great tools.
posted by Dansaman at 11:16 AM on May 2, 2014

« Older Help me find Game of Thrones theme song parodies   |   How do we use Google Drive properly? Folders... Newer »
This thread is closed to new comments.