From OpenStreetMap Wiki
Jump to: navigation, search

Discuss TIGER here:

Tiger 2009

Any plans for updating the Tiger 2009 files? Some people have talked about comparing the difs and then uploading the changes. The data is organized by state then by county. In addition to the roads, there are a couple other shapefiles that may be useful. There is a limited landuse area file that includes some recreation and amenities. But it doesn't appear to be complete. There is also a shapefile that includes places. There are mountains included which have been imported on a previous update that can be removed. But there are other features such as property names or locations that don't appear to have been added. What is the best way to proceed? -- User:Srmixter - 11:28, 27 October 2009

I have reviewed the Tiger 2009 data and compared it to the current Open Street Map data. The mapping software I am developing allows me to quickly flip between OSM and Tiger data. In the vast majority of areas I have compared, the Tiger 2009 data is much more accurate for local roads. In most areas where OSM has been manually updated, it is more accurate and more detailed than the Tiger data. This is especially true in the case of primary highways and freeways. Tiger's highways and freeways are usually out of date, incomplete or have large and obvious areas. If I could, I would take the highways from the OSM data and the local roads from Tiger. This is impractical because the types of roads would not match up at intersections and I have seen numerous instances where primary roads are tagged as local roads in the Tiger data.

If the 2009 Tiger data was to be incorporated into OSM, it would have to be done as part of the editing tool. I imagine that a user would be able to select a region to work in and then select the types of data to copy from Tiger to OSM. They would then have to fix the intersections manually but this would be a lot easier than rebuilding all of the roads as they must do currently.

In the private map data I am working on, I am taking the lakes, rivers, railroads and polygons from the Tiger 2009 data. I use the roads and buildings from the OSM data. I feel that having accurate primary roads is better than having more accurate local roads for my application. Also using the OSM data means that I can update it as frequently as I wish rather than having to wait a year for the next Tiger data set. -- User:grangerfx - 13:42, 10 March 2010 (UTC)

One way streetmap

Has any movement happened on the tiger files. As soon as they are imported I think it would be great to start a distributed one way project. If that were completed this would be on par with commercial data. -- User:Dcsmith77 - 07:28, 8 March 2007


Doesn't the TIGER data have rivers? There should be a river running through Reno for example.

Actually I mention this example because I was comparing our Reno map with one here which was produced as a one man effort by somebody with a python script back in 2004. I was about to pitch in on the discussion there, and mock* their puny TIGER rendering effort. But actually... hmmm.... we seem to be missing a river.

*I mean... invite them to join our better project.

-- Harry Wood 16:13, 20 December 2007 (UTC)

I had assumed that there weren't rivers in the dataset, but they show up at: [1], so there must be some government database that includes them. It would be great to have them. --Liber 01:55, 14 January 2009 (UTC)
Aaah! Check here: NHD

Tiger has good (but not amazing, at least as of 2007) data for water:

  • Inland water is well-defined topologically, in the form of gt-polygons; this is now available via the water-body shape files.
  • There used to be vector rivers for water too thin to represent as an area - I speculate the tags are probably in the all-edges file, btu I haven't worked with this since the conversion of TIGER to .shp. I'm not sure what OSM's policy is on "1-d water".
  • TIGER is by county, and each county has spatial extent. If you mark "the entire world minus every county" as ocean, you get correct land-water for any given tile size, except for Canada and Mexico, since they are excluded from TIGER.

Looking at a few points in OSM as of this writing, there is value to be had here. For example:

Indian Lake hasn't been imported, but is present in TIGER. On the other hand:

Extended waterways in Boston look okay.

So...we could take the TIGER water shape files and bring them into OSM, but we would need some kind of filtering process to decide how to deal with conflicts.

(If you are interested in this, contact me please - the open source code for X-Plane includes both shape and OSM processing tools now, so I could create a few tools to help automate this process pretty easily!) --Bsupnik 21:46, 2 April 2009 (UTC)


The tiger import uses a Rail:Spur tag for a spur lines on railroads. Unfortunately, this isn't supported by the current renderers, as the prefered tag seems to be: service:spur, perhaps with railway:rail. It would be great if there were an automated way to change this. --Liber 01:55, 14 January 2009 (UTC)

New 2008 Tiger Data

It would appear that the new data has been posted at [2] dated November 26, 2008. I was talking to a DOT professional who uses the GIS system and she stated that the new data is MUCH more accurate.

What plans are there for incorporating this new data. I would volunteer some machine cycles for the conversion. --AC7SS 17:21, 29 January 2009 (UTC)

Is that the same TIGER data described on the TIGER 2007 page? Not sure. -- Harry Wood 00:20, 30 January 2009 (UTC)
No, the 2008 data is newer and more accurate than the 2007 tiger import. I think someone who knows the paradigm and roadmap (pun intended) of OSM needs to update the TIGER page to make it clear what the connection is between OSM and TIGER data in the future. For example, I deleted a hospital from my home town that was pulled in from tiger data tagged as being 1970s vintage--to my knowledge, there hasn't been a hospital there in at least 30 years. Is that going to show up again in the future as the TIGER is imported again? --Pouletic 14:49, 25 March 2009 (UTC)
Has anyone considered using the Automatch script from geobase2osm? there might be a way to tweak it, so then it looks for all kinds of other features. This might help by detecting data that currently exists in OSM, so to not overwrite it. --acrosscanadatrails 23:42, 26 March 2009 (UTC)
Please be aware that there is significant corruption in the 2008 TIGER data. I ran into this with my own mapping project which uses TIGER data. Road points connect randomly to points on other features. Features are misidentified. In some areas, lines spike off in random directions for considerable distances. Overall the data quality is very high and I look forward to the 2009 release. This is what the TIGER site has to say about this issue:
"User Note 8: All Lines Shapefiles - Feature Distortion
In the 2008 TIGER/Line Shapefiles release we have identified 1,159 counties that contain new features that are distorted. These features appear to be pulled away from their position to a point and returned to their correct location forming a distorted edge. The problems are a result of the efforts to automatically integrate new data with existing positionally accurate features. We have not identified wide-spread issues in other counties but it is possible that there are additional counties with some distorted edges.
We are working diligently to correct these errors and anticipate a marked improvement in the 2009 TIGER/Line Shapefiles. We are continually working on updating our database and improvements will be seen with each subsequent release. In the meantime, depending on your needs, you may wish to use the prior version of the TIGER/Line Shapefiles or wait for the 2009 release which is scheduled for late summer."
--grangerfx 20:53, 10 July 2009 (UTC)