OpenStreetMap

Crazy TIGER

Posted by SimonPoole on 5 May 2014 in English (English)

Anybody that has spent any amount of time fixing TIGER data in the US has seen all the artifacts that we have all come to love, over- and underruns in corners, lost fixes after being inside, wiggles where the surveyors stopped to read road signs, crosscountry non-existing residenials and so on. Today I saw one thing I hadn't before:

(blue is the corrected version)

Comment from kusmi on 5 May 2014 at 12:38

Looks like he/she held the GPS device wrong by 90° ;-)

Hide this comment

Comment from dalek2point3 on 5 May 2014 at 16:53

When TIGER was imported, something called the TIGER accuracy improvement project was only half done. These are the counties where you are less likely to find errors like this, and I'd be willing to bet that Arkansas had not been touched by the TIGER accuracy when it was imported into OSM. While its OK to criticize TIGER, it should also be noted that better TIGER exists, we just didnt use it properly.

Hide this comment

Comment from dalek2point3 on 5 May 2014 at 16:54

Oops -- forgot to include a link to the TIGER completion map:

Imgur

Hide this comment

Comment from SimonPoole on 5 May 2014 at 17:37

@dalek2point3 undoubtably TIGER is a great resource, and more recent releases are clearly a lot better for a large parts of the USA (there are still numerous areas where it is rather bad to say the least). But, as you point out, it is clearly not TIGER projects fault that we imported the data at a such early date. Further I don't think it is in any way fair to blame the importers, there seems to have been a community consensus that it was a good thing at the time.

What we shouldn't do however, is let political correct positivism get in the way of learning the many lesson we should learn from the TIGER import. For example don't hurry to import data when there is no immediate need, it is not going to go away and the next release will be better.

Hide this comment

Comment from nfgusedautoparts on 7 May 2014 at 12:37

i've seen stuff like this in WV. i think there was a lot of hand drawing of maps in the field followed by data entry without supporting aerial imagery in some areas. that would account for the good topology and bad localization.

Hide this comment

Comment from kc0nlh on 8 May 2014 at 04:13

Wow! As the saying goes stuff happens.

Hide this comment

Leave a comment

Parsed with Markdown

  • Headings

    # Heading
    ## Subheading

  • Unordered list

    * First item
    * Second item

  • Ordered list

    1. First item
    2. Second item

  • Link

    [Text](URL)
  • Image

    ![Alt text](URL)

Login to leave a comment