[Talk-us] BC/Washington border

[Paul Norman]

On 2013-01-09, at 12:58 PM, Clifford Snow <clifford at snowandsnow.us> wrote:

> On Wed, Jan 9, 2013 at 8:27 AM, Andrew Guertin <andrew.guertin at uvm.edu> wrote:
>> I plan to import VT town boundaries from that file at some point, but
>> was not planning on touching any of the state boundaries (except to
>> connect town boundaries to them). I'd be interested to know how well
>> that data source matches others for the state boundary, too, though.
> 
> Paul Norman got me thinking about the BC/Washington State border. Comparing Washington State's border files with TIGER data you'll find discrepancies. I emailed our state's GIS group.  Here is the response I received.
> While I cannot speak to the construction of the State polygons in the TIGER shapefile I can relate why I think the State version is more accurate.  The version created and maintained by the Washington State Department of Natural Resources utilizes positions of actual international boundary monuments that have been coordinated by the National Geodetic Survey (NGS) or by other surveys performed by licensed land surveyors.
> 
> I did some spot checking where we have monument positions and found the State line to be more accurate than the TIGER representation.  I also spot-compared the lines to existing features, such as roads running parallel to the boundary; see a screen shot below where the red line is TIGER and the green line is State.  The road in question is in fact in Canada, but the TIGER boundary zigzags from one side to the other.
> 
> It looks like there are some other differences between the Washington version and the TIGER shapefile along the borders of Idaho and Oregon as well.  For Washington we use survey data where available here as well, but I did not check as thoroughly to try and figure the differences out.
> 
> You might want to check out all of your state's boundaries against Quebec and neighboring states. I haven't done anything with Washington State. I'd say something about a rainy day project, but it has been raining here for a while and will continue until July.  

On the west coast the Canada/US border is defined as the line formed from a bunch of survey points. These points can be surveyed or seen on imagery. I used the IBC positions for the points which are far more accurate than anything I could of done. In theory they could be improved, but the accuracy limit of the DB might be an issue.

I verified the data with multiple other sources and they all agreed within the limits of the sources.

The tricky part was making the lines straight where they joined up with cities, etc.

I would not suggest using state data for the Canada/US border.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-us/attachments/20130109/e734a37e/attachment-0001.html>