OpenStreetMap
Introduction
Later today (September 22, 2021) OpenStreetMap US is hosting a discussion on trail tagging. There has been a lot of chatter recently about how to map these, and the following summarizes my current thoughts on the matter going into tonight’s meeting. I wrote these up mainly as a way to organize my own thoughts, and I’m sure that my opinions will change after further discussion.
Definitions
A path is any way not meant for wheeled motor vehicles. A trail is path or other way signposted or recommended by the landowner for use by those who are not in wheeled motor vehicles. Paths are mapped as ways, and trails are mapped as relations.
Paths that do not exist and never existed
These paths should be removed from OSM. If remote mappers keep restoring them, it may be worthwhile to leave the line tagged with only a note.
Paths that do not exist but recently used to exist
These paths should be tagged with abandoned:highway=path where found, but ideally they should not have been added in the first place. If these paths are deleted, from my experience, locals will inevitably add them back.
Paths that exist but should not exist according to the landowner
Tag as highway=path + access=no + informal=yes where found, but ideally they should not have been added in the first place. There are arguments to be made for using a different tag than path, but none exists as of yet. Deleting these from OSM is futile as locals or remote mappers will inevitably add them back.
Paths that exist and are only to be used by the landowner and guests
Tag as highway=path + access=private where found, but ideally they should not have been added in the first place.
Paths that exist and are tolerated
Tag as highway=path + access=permissive + informal=yes.
Paths that exist and are authorized
Tag as highway=path and include a surface tag
Paths that are part of trails
Tag as highway=path + surface=* and include in a type=route relation
Trails where the pathways used to exist but no longer do
Tag as highway=path along the most frequently-used bushwhack or along the line on the map provided by the landowner and include in a type=route relation. Also, mark it up with trail_visibility=horrible or no and other tags to fully describe that there is no real path.
Discussion
Comment from Minh Nguyen on 22 September 2021 at 19:47
This sounds like another potential use case for the
not:lifecycle prefix. But anoteis good too, because not everyone would know to look fornot:highwayin the way’s raw tags.There have been some experiments with explicitly tagging disputes, but they tend to be related to boundaries or names. A trail disputed to be in existence could be tagged
disputed=yes, but unlike with boundaries, data consumers probably aren’t expecting to add a special case for this tag on any arbitrary feature, so maybe adisputed:lifecycle prefix would be more appropriate.Comment from aweech on 22 September 2021 at 20:20
Minh, for the case of paths that exist but should not exist according to the landowner, I’m talking about paths that the landowner explicitly does not allow people to go on and wishes didn’t exist, but the community does not respect the landowner’s wishes. The mountain biking community is famous for creating these paths, and they’re quite active in copying them to OSM. I think
informalis thehighwayequivalent ofdisputed. Mainly I feel like it’s weird to “dispute” something that you can see with your eyes. I have come across some cases where the existence of a public right of way or ownership of a path is actually disputed (as in, currently working its way through the courts). It’d be worthwhile to have a way to tag that, but it’d have to be tied back to the access, since that’s what’s disputed. Maybe something likeaccess:disputed=yes.But for the case of remote mappers tracing phantom paths, the
not:highwaylifecycle tag sounds like a great fit!Comment from n76 on 23 September 2021 at 14:25
I did not see this until after the 23rd. Is there a summary of the discussion I can view or read somewhere?
I have some vested interest in trail tagging as I produce my own hiking maps and have had issues determining if a way is a trail or not, what labels I should display on features, and some other things. As such I would like to keep up with discussions on the topic of tagging trails.
Comment from Glassman on 23 September 2021 at 19:06
Paths that exist but should not exist according to the landowner
In a wilderness like area having these trail can be helpful for search and rescue efforts and should remain. Your tagging suggestion is appropriate.
Paths that exist but illegally travers culturally sensitive areas
I’ve seen these in Washington State and suspect they are elsewhere. They should be removed. The key is they are illegal.
Comment from kucai on 24 September 2021 at 00:54
how different is this from service road going around a private industrial building for example? People can map it, but essentially nobody is actually allowed to drive there (and it goes around the building so you can’t really get lost). Just random thoughts.
Comment from aweech on 24 September 2021 at 01:27
Hi, n76. The meeting ended up mostly being folks from US federal agencies talking about their experiences with visitors taking OSM-derived maps as gospel. The notes from the meeting have been posted here. There was some discussion of what to do next, with the conclusion that there need to be more meetings.
Hi, kucai. I’m not quite sure what you’re trying to say, but the closest analogue from this list of categories to the situation you brought up is “paths that exist and are only to be used by the landowner and guests”. Service roads that are closed to the general public are usually tagged as
highway=service+access=private.Comment from n76 on 24 September 2021 at 02:50
I volunteer with the US Forest Service so I am aware of some of the issues with unwanted unofficial trails. On the OSM side, there are many mappers who don’t care about official status and simply “map what exists”. Arm chair mappers can make it worse: If a trail crew works hard to close off and hide an unwanted trail at either end, an arm chair mapper looking only at older aerial imagery may add it back in.
My personal solution when mapping is to draw the way and then tag it with
informal=yes,access=no,not:highway=path(orwas:highway=path,abandoned:highway=path, ordemolished:highway=path), and if the trail crews have done a reasonable job of covering/blocking the ends alsotrail_visibility=poor. In addition, adescription=*tag letting the next mapper know that they should not change that tohighway=*.I am pretty sure that hiking apps will not show a way tagged as I do which should keep the traffic down on them until they heal. For the areas that I volunteer at, I’ve set up RSS feeds to monitor change sets so I can keep track of what other mappers are doing and take appropriate action, usually just comments on the change set.
In my area recreation functions, like trail maintenance, are largely volunteer activities. Which leads to a suggestion for land managers: Work with your recreation volunteers to get a few of them involved with OSM and to “take ownership” in OSM of the trails they maintain. Ownership in the sense of verifying the area is mapped correctly both in terms of location, condition, restrictions, etc.
Comment from Kai Johnson on 11 October 2021 at 14:00
I’ve been working on BLM routes in Imperial County, CA. There are official, numbered BLM routes that traverse desert dune areas where the sand frequently shifts.
The actual tracks on the ground follow the general direction within a few hundred yards, but vary significantly depending on where the dunes can be crossed.
I could clearly tag these ways with
trail_visibility=horrible. But are there other tags that would be appropriate given how navigation on the ground varies?