I’m looking to download historical OpenStreetMap (OSM) data specifically for Texas, but most of the resources I’ve found only provide global datasets in .torrent format. Is there an alternative way to access historical OSM data at the state or regional level?
There are various ways to download data from OpenHistoricalMap such as Overpass Turbo, or using JOSM if you’re familiar with that.
Thanks so much for the advice! I’m trying to download building footprints for Texas as they existed in 2022. When I used Overpass Turbo, it only returned a few polygons rather than all buildings. I’m wondering if I might be missing some keywords in the query or if the data is limited in some way.
I’d really appreciate any advice!
Howdy! 
& welcome to OHM, @YimingZhang. Great to have you here.
One issue might be that there aren’t that many building footprints in Texas in OHM. There’s certainly not (last I checked) comprehensive building coverage for any time period in Texas. Maybe you could help us with that? 
Separately, if you were interested in what building footprints were included in OSM at a particular point in time, adding a date parameter to an Overpass query might work.
Or, here’s a link to the Geofabrik OSM pbf extract for Texas as of 23-01-01, which would cover everything through 2022.
I really appreciate your reply!
I’m looking to download all building polygons and road networks for Texas on an annual basis over the past several years to analyze building development over time. I was wondering—does the Geofabrik file you shared include only building points, or are polygon footprints also available? From what I’ve seen, it appears the building footprints might not be included, but I wanted to confirm. Thanks!
I haven’t checked those .pbf files myself, so I’m not sure. If they don’t include buildings, you’ll want to do some variation of the Overpass query with a date filter, or run osmosis against the geofabrik pbfs to get what you’re looking for.
That all said… unless this is an inquiry into how frequently buildings are updated in OSM, rather than an inquiry into building stock turnover in Texas, you may not find what you’re looking for through OSM.
For example, say there were no buildings in Junction, TX in 2022. But, in 2023, a hypothetical mapper imported all the buildings in Junction from a 2018 LIDAR survey, and then no one added or removed buildings in OSM since then, it might be hard to draw any conclusions about actual/real world building development from OSM extracts. Same could be said for roads, although my guess is OSM keeps up with road changes much more reliably than building changes.
My guess is that for real analysis of building development over time in Texas, you’d need to do some more in-depth analysis of building permits on a county-by-county basis.
Does that make sense?
Thanks again for the helpful context!
Just to clarify — does this mean the quality and consistency of building footprints in OSM can vary to the extent that it’s not reliable for tracking actual building expansion over time?
Also, I came across the .torrent files here — are these part of the same dataset version, or would they offer any additional insight? Appreciate your guidance!
That’s correct. It’s only tracking what the user community has updated & should not be used for anything that requires up-to-date information or real-time updates.
I’m not sure if those are the same, as it’s unclear what you are comparing - the torrents to the geofabrik extracts? Those torrents are “full history,” so those should be the most authoritative historical extracts, but you will probably have to do some processing on those that geofabrik has already taken care of.
That all said, this isn’t an area most people on OHM work with. I think you might find a more authoritative answer over on the OSM Community Forum.