[Imports-us] [Talk-us] Rail westerly

stevea steveaOSM at softworkers.com
Wed Dec 31 19:24:01 UTC 2014


Michael Patrick writes:
(about the Oak Ridge National Laboratory: ... Railroad Network data).

I found these at:

http://www-cta.ornl.gov/transnet/RailRoads.html

And have downloaded (~10 megabytes) a zipped shapefile of the entire 
network (as well as the simplified 7 megabyte one which "omits 
inactive lines and contains only current operators, but incorporates 
interlines as network links.")

I knew there had to be something like this in the public domain, and 
I say thank you very much, Michael.

I examine these in JOSM right now.  First they need to be unzipped, 
and it looks like the (provided on that web page) PRJ file to change 
from WGS 84 (default) to either NAD 27 or 83 projection is required. 
I haven't done that to these data in the instant case, but I've 
fiddled these before and I think it is doable.

Results in JOSM (after many seconds of load time) -- and JOSM MUST 
have the Shapefile plug-in -- do indeed display a nationwide network 
of rail lines.  A sample one I chose (UP's Coast line in California, 
which I believe I have gotten mostly correct in OSM recently) has 30 
rather cryptic (at first blush) tags, but these indeed look to be 
usable data.  Geographically, yes, the rail line looks "about 
correct" though the tag structure will have to be seriously 
harmonized to become something to import into OSM.

This starts to move (quickly) into the direction of a major import 
(and all its required vetting, etc.) into OSM.

I ask others to help me determine the suitability of whether we might 
want to use these data.  I imagine a fair bit of work would be 
required to harmonize the 30 tags into those we might deem 
appropriate for USA rail in OSM, as well as strategies for conflating 
them with existing TIGER rail data.  It's a big, big, BIG job.  On 
the other hand, I could see small segments in these data that 
interest local mappers being used to confirm names or actual track 
locations for existing data on a line-by-line basis, too.

Thanks for really good discussion about this, Charlotte stepping up 
to make a wiki page, Michael's reference to these data and the great 
volunteer, cooperative and collaborative spirits we find in OSM here 
in the USA.

SteveA
California



More information about the Imports-us mailing list