I have encountered a situation where there appears to be duplicate rail geometry, and I have a couple of questions…
- Is this correct?
- If not correct, how should this problem be corrected?
The track in question can be seen in the image below. I have include two links because I could only embed a single image as a new user. The geometry is not exactly the same. The first section of track goes a bit further north than the second example. There is not any overlap with other geometry on the north end of the longer track. I believe the shorter line should be deleted, but I would like some confirmation before doing something that may negatively impact the dataset. I appreciate the help.
Hello @ALTRIOSRailGuy and welcome to the forum!
I tried identifying the situation based on aerial imagery. Unfortunatly I am not sure how many rails there are. There is also this peace of rail that I can’t see on the aerial image:
Was there recently construction going on?
If there realy is only one rail and not 2 parrallel ones than there should only be one way in OSM and deleating the shorter one seems reasonable. The longer one is part of 13 relations (“Teil von”=“part of”):
So deleating this would more complicated.
The 2 rails you linked have different tags attached to them. Please make sure to check which of the tags should be kept and which should be deleated before you deleate the way.
This changeset may have intended to split the tracks at this switch but the split only took partially. Editing software used to be pretty bad about causing that sort of thing. The longer way is a member of all the relevant route relations and has more history behind it. So for convenience, I’d keep that one and transfer any relevant tags from the shorter way to the longer way before deleting the shorter way. And if some of the tags only apply up to the switch, then you can split it again.
2 Likes
Thanks for the reply. I am unsure if there is new construction there. I uncovered this part of the data because I am using the open street map data to build a network for some rail modeling software that has been developed by the National Renewable Energy Laboratory, Southwest Research Institute, and the University of Texas. Its called ALTRIOS, and it designed to simulate things like alternative locomotive powertrains in a rail network. It tries to answer things like how big should a battery locomotive be, where should I place the chargers, how much electricity is needed by location…
I am trying to improve the network building process so that we can study more corridors than the couple we have been able to put together so far. The overlapping geometry is causing some problems with the build process. The good thing is that this process has identified several places where switches are missing that I have been able to add.
I appreciate the help. I’ll keep you posted on what I find.
https://www.nrel.gov/transportation/altrios
4 Likes
In case it helps, QLever turns up 70 overlapping railways in the United States as of a couple weeks ago.
After a bit more digging, I think the link ending in 159 needs to be shortened. I had missed the siding earlier. 159 extends all the way south past the siding. The south end of 159 and 417 match up. 159 still needs to cover the gap that i have indicated with the blue arrow. I appreciate the help with this.
Thanks for the info. I am not familiar with QLever before today. That looks like a really neat tool. Thanks for sharing.
If anyone is interested, I took care of this with changeset 167031560. I appreciate the help.
Sorry for the consternation – I must have messed it up when I was trying to edit that area the first time around. Unfortunately, Altrios’ edit left a hole in the maxspeed tagging and PT relationships, so I had to go back in and fix it up further.
No need to apologize. I am very new to editing OpenStreetMap data, and this gave me an opportunity to dig into something small and correct it. I learned a lot in the process.
